Analytical Profiles of Drug Substances and Excipients, Vol 24

Analytical Profiles
of
Drug Substances
and
Excipients
EDITORIAL BOARD
Abdullah A. Al-Badr Larry D. Kissinger
Alekha K. Dash David J. Mauo
Klaus Florey Christopher T. Riley
Lee T. Grady Timothy J. Wozniak
Dominic P. Ip
Analytical Profiles
of Drug Substances
and Excipients
Volume 24
edited by
Harry G. Brittain
Ohmeda Pharmaceutical Products Division, Inc.

100 Mountain Avenue
Murray Hill, New Jersey 07974
Founding Editor:
Klaus Florey
ACADEMIC PRESS
San Diego New York Boston London Sydney Tokyo Toronto
Find Us on the Web! http://www.apnet.corn
This book is printed on acid-free paper. @

Copyright ie 1996 by ACADEMIC PRESS, INC.
All Rights Reserved.

No part of this publication may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopy, recording, or any information
storage and retrieval system, without permission in writing from the publisher.
Academic Press, Inc.

A Division of Harcourt Brace & Company
525 B Street, Suite 1900, San Diego, California 92101-4495
United Kingdom Edition published by

Academic Press Limited
24-28 Oval Road, London NW 1 7DX
International Standard Serial Number: 1075-6280
International Standard Book Number: 0-12-260824-0
PRINTED RJ THE UNITED STATES OF AMERICA

96 97 9 8 9 9 00 01 BC 9 8 7 6 5 4 3 2 1
CONTENTS
AfJiriations of Ediforsand Coaiributors vii
Preface xi
1. Carbenoxolone sodium 1
S. Pindado, 0.1.Corrigan, and C.M 0 'Driscoll
2. Clarithromycin 45
LI. Salem
3. Crospovidone 87
E. S. Barabas and C.M. Adeyeye
4. Fluvoxamine Maleate 165

N.H. Foda, MA. Radwan, and O.A. A1 Deeb
5. Gadoteridol 209
K. Kumar, M F. Tweedle, and H G. Brittain
6. Guar Gum 243

K. Yu, D. Wong, J Parasrampuria, and D. Friend
7. Mafenide Acetate 277

A.K. Dash and S. Saha
8. Maltodextrin 3 07
MJO Mollan, Jr., andM Celik
V
v1 CONTENTS
9. Nalmefene Hydrochloride 35 1
H G. Brittain
10. Polyvinyl Alcohol 397

D. Wong and J. Parasrampuria
11. Sertraline hydrochloride 443

B.M Johnson and P.-T. L. Chang
12. Solasodine 487

G. Indrayanto, A. Syahrani, R. Sondakh,
and M. H. Santosa
13. Starch 523

A. W. Newman, R.L. Mueller, I.M. Vitez,
C.C. Kiesnowski, D.E.Bugay, W.P. Findlay,
and C. Rodriguez
14. Tobramycin 579

A. K. Dash
Cumulative Index 615

AFFILIATIONS OF EDITORS AND CONTRIBUTORS
Christiunuh M Adeyeye, Department of Pharmacy, Duquesne University,

Pittsburgh, PA 15282
Abdulluh A. Al-Budr, College of Pharmacy, King Saud University, P.O. Box
2457, Riyadh 11451, Saudi Arabia
Omur A. A1 Deeb, Department of Pharmaceutical Chemistry, College of
Pharmacy, King Saud University, P.O. Box 22452, Riyadh 11459,
Saudi Arabia
Eugene S. Burabus, ISP Corporation, 1361 Alps Road, Wayne, NJ 07470
Hurry G. Brittuin, Ohmeda Pharmaceutical Products Division, Inc., 100
Mountain Avenue, Murray Hill, NJ 07974
David E. Buguy, Bristol-Myers Squibb Pharmaceutical Research Institute,
One Squibb Drive, New Brunswick, NJ 08903
Metin celik, College of Pharmacy, Rutgers, the State University of New
Jersey, Piscataway, NJ 08855
Pei-Tei L. Chung, Central Research, Pfizer Inc., Groton, CT 06340
Owen I. Corrigun, Department of Pharmaceutics, University of Dublin,
Trinity College, Ireland
Alekhu K. Dash,, School of Pharmacy, Department of Pharmaceutical
Sciences, Creighton University, 2500 California Plaza, Omaha,
Nebraska 68178
W Paul Findluy, Bristol-Myers Squibb Pharmaceutical Research Institute,
Klaus Florey, 151 Loomis Court, Princeton, NJ 08540
Nugwu H. Fodu, Department of Pharmaceutics, College of Pharmacy,
King Saud University, P.O. Box 22452, Riyadh 11459, Saudi
Arabia
David Friend, Cibus Pharmaceutical, Inc., 200 D Twin Dolphin Drive,
Redwood City, CA 94065
Lee T. Grudy, The United States Pharmacopeia, 12601 Twinbrook Parkway,
Rockville, MD 20852
vi i
...
Vlll AFFILIATIONS OF EDITORS AND CONTRIBUTORS
Gunawan Indrayanto, Laboratory of Pharmaceutical Biotechnology,

Faculty of Pharmacy, Airlangga University, Surabaya, Indonesia
Dominic P. Ip, Merck, Sharp, and Dohme, Building 78-210, West Point, PA
19486
Bruce M. Johnson, Central Research, Pfizer Inc., Groton, CT 06340
Christopher C. Kiesnowski, Bristol-Myers Squibb Pharmaceutical
Research Institute, One Squibb Drive, New Brunswick, NJ 08903
Lurry D. Kissinger, The Upjohn Company, 7171 Portage Road, Kalamazoo,
M149001
Krishan Kumar, Bracco Research USA, P.O. Box 5225, Princeton, NJ
08520
David J Mazzo, Department of Analytical & Physical Chemistry, RhBne-
Poulenc Rorer Recherche-Developpement,20, avenue Raymone
Aron, 92 165 Antony Cedex, France
Matthew J Mollan, Jr., College of Pharmacy, Rutgers, the State
University of New Jersey, Piscataway, NJ 08855
Ronald L. Mueller, Bristol-Myers Squibb Pharmaceutical Research
Institute, One Squibb Drive, New Brunswick, NJ 08903
Ann W. Newman, Bristol-Myers Squibb Pharmaceutical Research Institute,
Caitriona M. 0 'Driscoll, Department of Pharmaceutics, University of
Dublin, Trinity College, Ireland
Jugdish Parasrampuria, Cibus Pharmaceutical, Inc., 200 D Twin Dolphin
Drive, Redwood City, CA 94065
Silvia Pindudo, National Pharmaceuticaland Biotechnology Center,
Dublin, Ireland
Muhasen A . Radwan, Department of Clinical Pharmacy, College of
Pharmacy, King Saud University, P.O. Box 22452, Riyadh 11459,
Saudi Arabia
Christopher T Riley, Room 106, Building 353, Experimental Station,
DuPont Merck Pharmaceutical Company, P.O. Box 80400,
Wilmington, DE 19880-0400
AFFILIATIONS OF EDITORS AND CONTRIBUTORS ix
Christine Rodriguez, Bristol-Myers Squibb Pharmaceutical Research

Institute, One Squibb Drive, New Brunswick, NJ 08903
Shankar Saha, Department of Biomedical Sciences, Creighton University,
2500 California Plaza, Omaha, Nebraska 68 178
Isum I. Salem, Department of Pharmacy and Pharmaceutical Technology,
University of Granada, 18071- Granada, Spain
Mulja H Santosa, Laboratory of Pharmaceutical Biotechnology, Faculty
of Pharmacy, Airlangga University, Surabaya, Indonesia
Robby Sondakh, Laboratory of Pharmaceutical Biotechnology, Faculty of
Pharmacy, Airlangga University, Surabaya, Indonesia
Achmud Syahruni, Laboratory of Pharmaceutical Biotechnology, Faculty
of Pharmacy, Airlangga University, Surabaya, Indonesia
Michael Tweedle, Bracco Research USA, P.O. Box 5225, Princeton, NJ
08520
Imre M Vitez, Bristol-Myers Squibb Pharmaceutical Research Institute,
David Wong, Cibus Pharmaceutical, Inc., 200 D Twin Dolphin Drive,
Timothy J. Wozniak,Eli Lilly and Company, Lilly Corporate Center, MC-
769, Indianapolis, IN 46285
Karen Yu, Cibus Pharmaceutical, Inc., 200 D Twin Dolphin Drive,
This Page Intentionally Left Blank
The profiling of drug substances as to their physical and analytical
characteristics remains as important today as it was when the Analytical
Profiles series was first initiated. The compilation of concise summaries
of physical and chemical data, analytical methods, routes of compound
preparation, degradation pathways, and the like, is a vital function to both
academia and industry. It is certainly fair to say that workers in the field
require access to current state-of-the-art data, and the Analytical Profiles
series has always provided information of the highest quality. For this
reason, profiles of older compounds are updated whenever a sufficient
body of new information becomes available.
The series mission was expanded some time ago to include profiles of
excipient materials, reflecting the developing situation that these materials
are coming under a degree of scrutiny which is approaching that
associated with drug compounds. These highly detailed compilations of
excipient properties and analytical methods have been well received by
workers in the field, and such profiles will continue to be sought.
Perceptive readers will note that 1995 passed without publication of an

Analytical Profiles volume, a situation caused by an insufficient number of
chapter submissions. An increasingly ominous fact of our professional life
is that scientists seem to have less time available for scholarly
contributions, a phenomenon which may be related to the current trend of
consolidation and downsizing. If this trend continues, more innovation in
analytical profiling must take place. For instance, should a prospective
author feel unable to complete an entire chapter, then the submitted
portion will be accepted and the editor will find additional authors to
complete the work. The initial submission can consist of either the
physical characteristics section or the analytical methodology section.
As always, a complete list of available drug and excipient candidates is

available from the editor. We look forward to hearing from new and
established authors, and to working with the pharmaceutical community
on the Analytical Profiles of Drug Substances and Excipients.
Harry G. Brittain
xi
CARBENOXOLONE SODIUM
Silvia Pindado"', Owen I. C~rrigan'~)
and Caitriona M. O'Driscoll*(2)
(1) National Pharmaceutical Biotechnology Center

Dublin, Ireland.
(2) University of Dublin

Department of Pharmaceutics
School of Pharmacy, Trinity College
* Author for correspondence
ANALYTICAL PROFILES OF DRUG SUBSTANCES 1 Copyright 0 1996 by Academic Press, Inc.

AND EXCIPIENTS-VOLUME 24 All rights of reproduction in any form reserved.
2 SILVIA PINDADO ET AL.
1. Introduction
2. Description
2.1 Nomenclature
2.1.1 Chemical Name
2.1.2 Nonproprietary Names
2.1.3 Proprietary Names
2.2 Formulae
2.2.1 Empirical
2.2.2 Structural
2.3 Molecular Weight
2.4 Appearance
2.5 Official Compendia
2.6 Other Compendia
3. Synthesis
4. Physical Properties
4.1 Spectroscopy
4.1.1 Ultraviolet Spectroscopy
4.1.2 Infiared Spectroscopy
4.1.3 Mass Spectrometry
4.1.4 Nuclear Magnetic Resonance (IH, I3C)
Spectrometry
4.2 X-Ray Diffraction
4.3 Optical Rotation
4.4 Thermal Methods of Analysis
4.4.1 Melting Point
4.4.2 Differential Scanning Calorimetry
4.4.3 Thermogravimetric Analysis
4.5 Hygroscopicity
4.6 Dissociation Constants
4.7 Solubility
4.8 Partition Coefficients
CARBENOXOLONE SODIUM 3
5. Methods of Analysis
5.1 Elemental Analysis
5.2 Identification
5.3 Titrimetric Analysis
5.4 Ultraviolet Spectrophotometry
5.5 ChromatographicMethods of Analysis
5.5.1 Thin Layer Chromatography
5.5.2 Gas Chromatography
5.5.3 High Performance Liquid Chromatography
5.6 Radioimmunoassay
5.7 Radioactive Labeling
6, Stability
7. Pharmacokinetics
7.1 Absorption
7.2 Distribution
7.3 Metabolism
7.4 Excretion
8. Pharmacology
8.1 Therapeutic Indications and Uses
8.3 Toxicity and Side-Effects
9. References
10. Acknowledgements
4 SILVIA PINDAW ET AL.
1. INTRODUCTION
Carbenoxolone is a triterpenoid, the ester of 18p-glycyrrhetic (enoxolone)

acid with succinic acid. The di-sodium salt, identified as carbenoxolone
sodium, is used in the treatment of gastric and duodenal ulcers (Doll et al.,
1962; Pinder et al., 1976).
2. DESCRIPTION
2.1 Nomenclature
2.1.1 Chemical Names
A. Disodium 3~-(3-carboxylatopropionyloxy)-
1 1-oxo-olean- 12-en-
30-oate (B. P. 1993).
Disodium salt of 3~-(3-carboxypropionyloxy)-11-oxo-olean-12-en-

30-oic acid (Martindale, 1993; The Pharmaceutical Codex, 1979).
B. 3-(3-Carboxy-1-oxopropoxy)-11-oxoolean-l2-en-29-oic acid.
3~-Hydroxy-ll-oxoolean-12-en-30-oic acid hydrogen succinate.
3P-Hydroxy- 1 1-0xoolean-12-en-30-oic acid 3-hemisuccinate.
3-O-(f3-carboxypropionyl)-1 1-oxo- 18p-olean-12-en-30-oic acid

(Merck Index, 1989).
2-Carboxy-ethylpropionyl glycyrrhetinic acid (British patent,

1960).
2.1.2 Nonproprietary Names
A. Carbenoxolone Sodium, Disodium Enoxolone Succinate (BANM,

USAN, rINNM) (Martindale, 1993).
B. Carbenoxolone, Glycyrrhetinic Acid Hydrogen Succinate,

Glycyrrhetic Acid Hydrogen Succinate, 18P-GlycyrrheticAcid
Hydrogen Succinate (Merck Index, 1989; Clarke, 1986)
Biogastrone, Bioplex, Bioral, Carbosan, Duogastrone, Gastrausil,

Megast, Pyrogastrone, Sanodin, Ulcus-Tablinen (Martindale,
1993).
2.1.4 Chemical Abstracts Service (CAS) Registry Numbers
7421-40-1 Carbenoxolone Sodium

5697-56-3 Carbenoxolone
2.2 Formulae
2.2.1 Empirical
Carbenoxolone Sodium
Carbenoxolone
2.2.2 Structural
Carbenoxolone Sodium Carbenoxolone

614.7 Carbenoxolone Sodium

570.74 Carbenoxolone
2.4 Appearance
Carbenoxolone sodium is a white or pale-cream colored, hygroscopic

powder, with a slightly sweet taste followed by a persistent soapy
aftertaste. It tends to adsorb to glass.
2.5 OMicial Compendia
A monograph on Carbenoxolone Sodium is included in the British

Pharmacopoeia, (1 993), and in the Chinese Pharmacopeia, (1985).
2.6 Other Compendia
Carbenoxolone sodium is included in the Pharmaceutical Codex (1979),

and in Martindale (1993). Carbenoxolone is included in the Merck Index
(1989). Clarke (1986) gives a usehl summary of physical and chemical
data.
3. SYNTHESIS.
Carbenoxolone is synthesized from glycyrrhetinic acid, the aglycone of

glycyrrhizic acid (Figure I), which may be obtained from liquorice root.
Carbenoxolone is prepared by refluxing an organic acid with
glycyrrhetinicacid in an organic solvent, or by the action of an acid
anhydride in pyridine solution. The sodium salt is prepared by
neutralization with an aqueous solution of sodium hydroxide (British
patent, 1960; U.S. patent, 1962).
The carbenoxolone free acid used in the studies conducted for this
monograph was prepared from carbenoxolone sodium, B.P., by
precipitation in hydrochloric acid. The precipitate was washed with water
and dried to a constant weight at 105OC.
CARBENOXOLONESODIUM I
OH OH
Glycyrrhizic Acid
Glycyrrhetinic Acid > Carbenoxolone
Figure 1. Synthesis of carbenoxolone.

4. PHYSICAL PROPERTIES
4.1 Spectroscopy
The ultraviolet spectrum of carbenoxolone sodium (0.005% w/v) is shown

in Figure 2. The spectrum was obtained using a Shimadzu (UV-160)
UVNIS spectrophotometerand I-cm quartz cells. The spectrum, as
obtained in the range of 230 to 350 nm in a 1:1 v/v mixture of methanol
and 0.02M sodium carbonate, exhibits a single maximum at 256 nm. The
absorbance at this maximum is about 1.O. An E 1% of 200 has been
reported in this solvent mixture (Pharmaceutical Codex, 1979). In
aqueous acid and aqueous alkali, dual wavelength maxima have been
reported at 248 nm and 257 nm, with E 1% values of 172 each (Clarke,
1986).
0.891
A
0.713
0.535 .
/
0.356 -
0.178
Y
/
0.000~ ' . . 8 . I
250 300 350

Wavelength (nm)
Figure 2. Ultraviolet spectrum of carbenoxolone sodium.

4.1.2 Infrared Spectroscopy
The infrared absorption spectnun of carbenoxolone sodium and

carbenoxolone are shown in Figures 3 and 4. The spectra were recorded
with a Nicolet 5ZDX FT-IR spectrophotometer, from compressed
potassium bromide disks. Structural assignments for some of the
characteristic absorption bands in the spectra are listed in Table 1.
Table I.
Infrared assignments for carbenoxolone sodium and carbenoxolone.
Peak Maximum (cm-') Assignment
Carbenoxolone sodium
1720 C=O stretch (ester)

1680 C=O stretch (ketone)
1550 C=O stretch (carboxylate)
Carbenoxolone
1730 C=O stretch (ester)

1710 C=O stretch (carboxylate)
1650 C=O stretch (ketone)
SILVIA PINDADO ET AL
20
2000 1800 1600 1400 1200 1000 800 600
Wavenumber
Figure 3. Infrared spectrum of carbenoxolone sodium.

2000 1800 1600 1400 1200 1000 800 600

Wavenumber
Figure 4. Infrared spectrum of carbenoxolone.

12 SLVIA PINDADO ET AL.
The fast atom bombardment (FAB) mass spectrum of carbenoxolone

sodium, shown in Figure 5 , was recorded using a V.G. analytical 70e mass
spectrometer and a FAE3 probe, with glycerol as the matrix. The
intensities are calculated relative to the base peak at m/z 115. The
spectrum shows an (M+H)+ peak at m/z 615 (relative intensity 3.56%), a
peak at m/z 637 (5.53%) corresponding to (M+Na)+, and a further peak at
m/z 659 (0.47%) corresponding to (M+2Na-H)+. Major peaks were
detected at m/z (%) 571 (5.69), 137 (62), 115 (loo), 63 (53), 41 (63).
The electron impact (EI) mass spectrum of carbenoxolone (Figure 6) was

also obtained, by electron-impact at 70 eV and 200OC, using a Finnigan
MAT Quadrupole mass spectrometer and a direct insertion probe. The
molecular ion (M+) at m/z (%) 570 (4.16%) was observed. Major peaks
were detected at d z (%) 303 (64), 262 (62), 135 (loo), 101 (64), 95 (61).
!
- : 1 x10.00 8.4E6
'I
14
12
1o!
59 1
7.4E6
6.5E6
5.6E6
4.6E6
-3.7E6
-2.886
. I .9E6
659 -9.3E5
Figure 5. FAB mass spectrum of carbenoxolone sodium.

100.0 18688.
0.
50.0
Figure 6. Electron impact (EI) mass spectrum of carbenoxolone.
4.1.4 Nuclear Magnetic Resonance (IH, 136) Spectrometry
The NMR spectra of carbenoxolone sodium and carbenoxolone were

obtained using a Bruker MSL 300.
A. Carbenoxolone sodium
The 1H-NMR spectrum for carbenoxolone sodium (Figure 7) was

obtained at a frequency of 300.13 MHz, in deuterated water. The
solution contained 3-(trimethylsilyl) propionic-2,2,3,3-d4 acid,
sodium salt (TSP) as the internal standard. A 2D H-H COSY
spectrum was also obtained (Figure 8).
Figure 7. Proton nuclear magnetic resonance spectrum of

carbenoxolone sodium.
100.0 135
18688.
0.
50.0
Figure 6. Electron impact (EI) mass spectrum of carbenoxolone.
4.1.4 Nuclear Magnetic Resonance (lH, 13C) Spectrometry
The NMR spectra of carbenoxolone sodium and carbenoxolone were

obtained using a Bruker MSL 300.
A. Carbenoxolone sodium
The 1H-NMR spectrum for carbenoxolone sodium (Figure 7) was

obtained at a frequency of 300.13 MHz, in deuterated water. The
solution contained 3-(trimethylsilyl) propionic-2,2,3,3-d4 acid,
sodium salt (TSP) as the internal standard. A 2D H-H COSY
spectrum was also obtained (Figure 8).
(PPW 5
Figure 8. 2-DH-H COSY nuclear magnetic resonance

spectrum of carbenoxolone sodium.
CARBENOXOLONESODIUM 17
, ' I ~ ' " " ' I ' " ' , " '
200 150 IOU 50
200 1% 100 50
(PPd
Figure 9. Carbon-13 nuclear magnetic resonance spectrum of

carbenoxolone sodium (top), with DEPT 135' (middle),
DEPT 90' (bottom).
(PPm)
i
n 1 1
L'
1
2 t
r. t-
2
i
t
1.
! t-
I-
I
r 3
j
I
t-
Figure 10. 2-D 'H - 'jC COSY nuclear magnetic resonance spectrum of
carbenoxolone sodium.
Figure 11. Proton nuclear magnetic resonance spectrum of

carbenoxolone.
20 SEVIA PINDADO ET AL.
Figure 12. Carbon-13 nuclear magnetic resonance spectrum of

carbenoxolone, with DEPT 1350.
/lit I
I
p,,,, , , , , , , , , , , , L
(PPm) 5 4 1
Figure 13.2-D H-H COSY niwlear magnetic resonance

spectrum of carbenoxolone.
(PPm) 100 80 60
Figure 14.2-D 'H- I3C COSY nuclear magnetic resonance

spectrum of carbenoxolone.
The 13C-NMR spectrum, with DEPT (1350 and 900), for

carbenoxolone sodium was obtained at a frequency of 75.468 MHz in
deuterated water. In this case, TMDS was used as the internal
standard (Figure 9). A 2D H-C COSY was also obtained (Figure lo).
B . Carbenoxolone
The 1H-NMR (Figure 11) and the l3C-NMR (Figure 12) spectra for
carbenoxolone were obtained in deuterated chloroform (CDC13),
using tetramethylsilane (TMS) as internal standard and the same
frequencies as for the salt. The 2D H-H COSY (Figure 13) and the
2D H-C COSY (Figure 14) were also obtained.
4.2 X-Ray Diffraction
The powder x-ray diffraction patterns of carbenoxolone sodium and

carbenoxolone were obtained on a Siemens D-500 x-ray diffractometer,
using a Cu x-ray tube, at 40 kV and 40 mA (Figure 15). It is evident that
carbenoxolone is much more crystalline in nature, since the powder
pattern for carbenoxolone sodium is almost that of an amorphous solid.
The uptake of water significantly affects the structure of carbenoxolone
sodium, as is evident from a comparison of the diffraction patterns
obtained before and after drying (Figure 16). These findings are consistent
with the hygroscopic nature of the sodium salt.
4.3 Optical Rotation
In a 1 % w/v solution of carbenoxolone sodium in equal volumes of

methanol and 0.02M sodium carbonate, rotation values of +132O to +1400
are obtained (B.P., 1993). For carbenoxolone, the specific rotation [ a ] ~ ~ 0
is +1280 in chloroform, (Merck Index, 1989).
24 SILVk PINDADO ET AL.
hJ earbenoxolone
carbenoxolone sodium
0
5 10 15 20 25 30 35
Two - Theta (Degrees)
Figure 15. X-ray powder diffraction patterns of carbenoxoIone

sodium and carbenoxolone.
rigure 16. X-ray powder diffraction patterns of carbenoxolone

sodium before and after drying.
4.4.1 Melting Point.
Melting points for the carbenoxolone reported in the literature are in the
temperature range of 291-2940 (Clarke, 1989; Merck Index, 1989). In the
case of carbenoxolone sodium, melting with degradation occurred in the
range of 290° to 300OC.
4.4.2 Differential Scanning Calorimetry (DSC)
The DSC thermograms for carbenoxolone sodium and carbenoxolone were

obtained using a Mettler DSC 20 TA 3000 system, at a scan speed of
1OoC/min. With both compounds, an initial peak was obtained at
approximately 1OOOC which is attributed to the loss of water. Degradation
of carbenoxolone sodium tended to occur from 260 to 3OO0C, while
carbenoxolone melted below 300OC.
4.4.3 Thermogravimetric (TG) Analysis
The TG thermogram of carbenoxolone sodium was obtained using a TG50

Thermobalance on the Mettler TA 3000 system, at a scan speed of
100C/min. An initial weight loss was noted at approximately 1OOOC which
corresponded with loss of water. A further drop in weight occurred
between 290-300OC, which was consistent with the events noted in the
DSC studies. The acid form, carbenoxolone, also showed weight loss
above 3OOOC indicating that it also would degrade at elevated
temperatures.
4.5 Hygroscopicity
Carbenoxolone sodium is a very hygroscopic powder, so the moisture

content should not exceed 4% w/w, as determined by Karl-Fisher titration
(B.P. 1993; The Pharmaceutical Codex, 1979).
4.6 Dissociation constants
Large differences in the values for the dissociation constants of

carbenoxolone have been reported. These are summarized in Table 2.
Table 2. Dissociation constants (pKa) of carbenoxolone
PKa Method of determination Reference

PKal PKQ
6.7 7.1 not stated Downer et al.,(1970);
Clarke (1986)
4.18 5.56 Partition Blanchard et al., (1988)
4.38 5.11 Solubility
4.7 Solubility
Carbenoxolone sodium is soluble in 6 parts of water and in 30 parts of

alcohol, and is practically insoluble in chloroform and in ether. A 10%
wlv solution in water has a pH of 8.0 to 9.2 (Clarke, 1986; Martindale,
1993). At 24 and 37OC and pH 2, the intrinsic solubility of carbenoxolone
was reported to be 1.16 and 1.63 x 10-5 Myrespectively, (Blanchard et al.,
1988). A table of estimated solubilities for carbenoxolone in the pH range
4.0 to 6.5 has also been presented (Blanchard et al., 1990).
The solubility of carbenoxolone sodium in the pH range 5.6 to 7.5 was

determined, and the results are shown in Figure 17.
The distribution coefficients for carbenoxolone at 24 OC between

chloroform and aqueous buffers have been reported as 2 (PH 7.4) and
greater than 100 (PH 1.O). The distribution coefficients between n-octanol
and aqueous buffers have been reported as 9 (PH 7.4) and exceeding 100
(PH 1.O) (Downer et al., 1970).
loo 1
x
5 6 7 8
PH
Figure 17. pH solubility profile of carbenoxolone.
Apparent partition coefficients between n-octanol and a 0.1 M citrate-

phosphate buffer at 37OC, and the fraction ionized for carbenoxolone, have
been determined at various pH values (Bridges et al., 1976). The results
are shown in Table 3, with the fiaction unionized being calculated
assuming a pKa of 6.7 for carbenoxolone.
Blanchard et al., (1988) determined the partitioning of carbenoxolone

using tritium labeled drug. In this study, the apparent partition coefficients
(APC) at different pH values (2.6 to 7.6), in an n-octanol/aqueous buffer
system at 24 OC, were used to assess the pKa values. In the pH range
studied, the APC decreased from over 600 to below 50. The true partition
coefficient was reported as 643.8. The APC was independent of initial
carbenoxolone concentration between pH 3 to 7, implying that
carbenoxolone does not self-associate in the n-octanol or aqueous phases.
Table 3.
Comparison of the apparent partition coefficients between octanol and

buffer, and the fraction of carbenoxolone ionized at various pH values
(Bridges et al., 1976).
PH Apparent partition Fraction unionized

coefficients
(octanolhuffer)
8.0 14 0.05
7.4 27 0.17
7.1 60 0.29
6.8 214 0.44
6.5 350 0.61
6.2 484 0.76
5.6 679 0.93
5.0 908 0.98
5. METHODS OF ANALYSIS.
Carbenoxolone Carbenoxolone
Sodium (%) (%I
Carbon 66.43 71.55
Hydrogen 7.87 8.83
Oxygen 18.22 19.62
Sodium 7.48
5.2 Identification
The B.P. (1993) outlines four methods of identification for carbenoxolone

sodium:
(A) The characteristic light absorption, as described in the UV

Spectrophotometry section 4.1. The E 1% (1 cm) value at 256
nm is quoted as 199.
(B) 0.1 g is dissolved in 5 mL of water, made just acidic with 2M

hydrochloric acid, stirred well, and filtered. The residue is
washed with water until the washings are no longer acidic, and
is then dried to constant weight at 105OC. The infrared
absorption spectrum of this residue must be equivalent with
the spectrum of authentic carbenoxolone, as outlined in
section 4.2.
(C) Color test: 5 mg of sample is mixed with 50 mg of resorcinol

and 2 mL of sulfuric acid (80%). The mixture is heated at
2000 for 10 minutes, cooled, poured into 200 mL of water, and
made barely alkaline with 5M sodium hydroxide. The product
should exhibit an intense green fluorescence.
(D) The material must yield the reactions characteristic of sodium

salts.
Carbenoxolone sodium may be titrimetrically assayed by non-aqueous

titration, as described in the B.P. 1993. The salt is dissolved in water,
acidified, extracted into chloroform, evaporated to dryness, and finally
reconstituted in dimethylformamide. Tetrabutylammonium hydroxide is
used as the titrant and thymol solution as the indicator.
5.4 Ultraviolet Spectrophotometry
Coleman and Parke (1963) described a method for the determination of p-

glycirrhetic acid (enoxolone) and its readily-hydrolysableesters (including
carbenoxolone), in biological fluids. The material containing the
glycyrrhetic acid or its esters is hydrolyzed with ethanolic sodium
hydroxide, the glycyrrhetic acid is extracted fiom the acidified
hydrolysate, submitted to two dimensional, thin-layer chromatography on
alumina, eluted with ethanol, and estimated spectrophotometricallyat 248
nm.
The UV absorption of carbenoxolone sodium in methanol and methanolic

0.1M NaOH were similar, with the reported wavelength maxima being
252 and 253.5. For these bands, the E 1%, 1 cm, values were 174 and
176, respectively. In aqueous solution, however, wavelength maximum
shifted to 260 nm, and the E 1%, 1 cm, value was 172 (Kracmar et al.,
1990). It was suggested that this information could be used to analyze
formulations containing carbenoxolone.
5.5 Chromatographic Methods of Analysis
Downer et al., (1970) have described a thin-layer chromatographic (TLC)

system for identifying carbenoxolone and its metabolites once these are
excreted in human bile following oral administration of the drug. Thin-
layer plates (0.25 mm) of fluorescent silica gel HF 254 were used and
developed in a solvent system containing: acetic acid-lY2-dichloroethane-
n-butanol-water (4 : 4 : 1 : 1 by volume). Carbenoxolone (Rfvalue =
0.95) was detected by a characteristic quenching of the background
fluorescence when viewed under ultraviolet light (Chromatolite lamp).
Clarke (1986) outlined a further TLC method for carbenoxolone sodium.

The method used thin-layer plates of silica gel G (0.25 mm), and a choice
of three mobile phases was given. These were ch1oroform:acetone (4:1);
ethyl acetate:methanol:strongammonia solution (8: 1 0 3 ; and ethyl
acetate. The Rfvalues for carbenoxolonewere 0.07,0.0, and 0.17 in each
32 SILVIA PINDAW ET AL.
of the respective systems, following visualization with acidified potassium

permanganate solution.
The B.P. (1993) uses a TLC method to separate carbenoxolone sodium and to
identify the presence of any related substances. In this method, silica gel F254
plates are used. The mobile phase contains ethyl acetate, methanol, water, and
13.5M ammonia (60:20: I 1:1 by volume). After removal of the plate, it is
allowed to dry in air and is examined under ultraviolet light (254 nm).
Alternatively, visualization may be performed by spraying with a 1.5%w/v
solution of vanillin in sulfuric acid (60%) and heating at 105OC for 10-15
minutes.
5.52 Gas Chromatography
A gas-liquid chromatographic procedure for the determination of

carbenoxolone in human serum has been described (Rhodes and Wright,
1974). Chromatography of methylated derivatives was performed on glass
columns containing 1% OV- 1 (dimethylsilicone gum) on a solid support of
Gas-Chrom Q (100- 120 mesh) at a temperature of 265OC. The 18 a-isomer of
carbenoxolone was used as the internal standard, and detection was by flame
ionization. The method showed a greater than +11% coefficient of variation
for all values determined, and the sensitivity was 5 mg/mL in serum. The
specificity of the method was established by combining the gas-liquid
chromatography analysis with thin-layer chromatography of prepared
standards and serum from treated volunteers.
Sanofi Winthrop, England, have developed a high performance liquid

chromatographic (HPLC) method for quantifying carbenoxolne sodium in
Pyrogastrone tablets. The column used was Spherisorb Hexyl5 mm (12.5 cm
x 0.45 cm i. d.), the mobile phase was methanol (75% v/v), water (25% v/v),
ammonium acetate (1% wh), and glacial acetic acid (0.1% v/v). The flow rate
was 2 mL per minute, and detection was by UV absorption at 254 nm.
Samples were prepared in 75:25 methanol water containing 1 to 2.5%
orthophosphoric acid (50% v/v). The retention time was approximately 1.5
minutes. The sensitivity reported for this method was 0.08 aufs.
5.6 Radioimmunoassay
Peskar et al. (1 976)developed a radioimmunoassay for carbenoxolone. [3H]-

carbenoxolonewas synthesized by reduction of 3-keto-enoxolone with sodium
borotritiide, followed by succinoylation of the resultant [3H]-enoxolone with
succinic anhydride. For preparation of the antigen, carbenoxolone was
conjugated to bovine serum albumin using the carbodiimide method described
by Goodfiiend et al. (1964). The production of antisera against carbenoxolone
was described, and their specificity and use for a radioimmunoassay were
reviewed, The sensitivity of the method in serum was 1 ng/mL.
5.7 Radioactive Labeling
Iveson et al. (1971) prepared [carboxypropionyl-1,4-14C2]-carbenoxolone,

(0.1mCi/g). This derivative was used by Bridges et al. (1 976)to assess the
gastrointestinal absorption of carbenoxolone in the rat. Samples were counted
in either a dioxane based scintillator or in a Tritox 100-toluene (I :2by volume)
scintillator containing 1% w/v butyl PPD. Radioactivity was measured by a
scintillation spectrometer, and the counting efficiency determined using [14C]-
toluene as internal standard. Blanchard et al. (1988;1990)used [3H]-
carbenoxolone sodium, specific activity 6.9mCi/mg and labeled at C-3, to
study the physicochemical properties and the absorption of carbenoxolone in
the rat. Samples were counted using a liquid scintillation counter.
6. STABILITY
Carbenoxolone is a triterpenoid, the ester of 18 p-glycyrrhetic (enoxolone)

acid with succinic acid. It is stable in neutral and acid solution, but is
hydrolyzed by alkalis into P-glycyrrhetic acid plus succinate (Parke, 1972).
7. PHARMACOKINETICS
7.1 Absorption
Carbenoxolone is rapidly absorbed following oral administration of an aqueous

solution of the sodium salt to patients, attaining high blood concentrations
(Parke et al., 1972 ). When tablets of the drug were administered to man on an
empty stomach, an initial plasma maximum occurred at 1-2 hours and another
maximum at 3-6 hours after dosage (Parke et af., 1972; Downer et al., 1970 ).
It has been suggested that the second peak is probably due to enterohepatic
circulation of the biliary-excreted conjugates. Baron et a/. (1975), however,
found no evidence for this pattern of absorption, and have reported a single
absorption peak.
When carbenoxolone was administered orally to patients subsequent to the

administration of an alkaline buffer mixture, the drug did not appear in the
blood plasma until the gastric acidity was restored, and the stomach contents
attained a pH value of less than two (where less than 0.002% of carbenoxolone
is ionized). Since carbenoxolone is a weak acid and in its non-ionized form is
highly lipid-soluble, it was suggested that the major site for absorption was the
stomach (Downer et af.,1970). The rapid absorption has been associated with
the high affinity of the drug for proteins, for although it is sparingly soluble in
acidic aqueous media (the stomach) the concentration of unbound drug in the
plasma is so low as to be undetectable. The high plasma protein binding of
this drug may therefore act to accelerate gastric absorption (Parke et af.,1972).
In contrast, Bridges et al., (1976) studied the absorption of ['4C]-

carbenoxolone from inverted rat ileum in-vitro, and from rat stomach and
illeum in-situ, and obtained greater absorption at pH values where the ionized
form predominates.
Iveson et al., (1966; 1971) reported that carbenoxolone was largely hydrolyzed
to P-glycyrrhetic acid before absorption in the rat. In contrast, Downer et al.,
(1970) reported that in man carbenoxolone was absorbed largely unchanged.
Bridges et af.,(1976) obtained no evidence of metabolism in the rat during
absorption in either the stomach or the intestine.
Bridges et al., (1976) observed an extensive tissue binding of

carbenoxolone to the inverted rat ileum in-vitro and suggested that the
high percentage of carbenoxolone accumulated in the tissue was not

entirely due to binding to tissue proteins and lipids, but also due to
precipitation in, or adsorption to, the gut sac epithelium. Tissue binding to
the ileum, in-situ, was not dependent on pH, except below pH 5.0, when
extensive tissue accumulation of carbenoxoloneoccurred because of its
low solubility. Tissue binding to the stomach increased markedly with
decrease of pH from 7.4 to 6.5, and at pH 6.5 was 80 times greater than
binding to the intestine. Contrary to the pH-partition hypothesis Bridges
et al., (1 976) reported that carbenoxolone was absorbed from the intestine,
and perhaps also from the stomach, at a rate 3.8 times faster when ionized
than in its unionized form.
The absorption of carbenoxolone was reevaluated by Blanchard et al.,

(1990) using an in-situ rat intestinal perfusion technique, in which
disappearance from the intestinal lumen, binding to the perfixed jejunal
segment, and appearance in the mesenteric (jejunal) vein were measured.
The effect of the degree of ionization on these processes was examined by
employing perfusion solutions of pH 4.0,4.4, 5.0, and 6.5. Tissue binding
was observed to be independent of pH. There was a rank order correlation
of the transfer rate of carbenoxolonewith the degree of ionization, which
indicated that carbenoxolone was absorbed faster in its ionized form. This
observation appeared to support the previously work of Bridges et al.,
(1976). However, Blanchard et al., (1990) have suggested a likely
explanation for this unusual behavior is that at the low pH values some
carbenoxolone precipitates out of solution during the perfusion
experiments, thereby reducing the driving force for diffusion across the
intestinal wall. Alternatively, ion-pairing of carbenoxolone with sodium
ion present in the pH 6.5 buffer may occur.
Food has been reported to delay the initial absorption phase of

carbenoxolone in patients with peptic ulcer when given as a single dose,
however, on subsequent accumulation of the drug after repeated
administration over 3 to 7 days, no food effect was observed. Concurrent
antacid administration, in these patients, did not significantly affect
carbenoxolone absorption (Baron et al., 1975).
Carbenoxolne is normally given in the form of tablets for gastric ulcer. It

is also available as a "position-release"capsule for duodenal ulcer, and this
formulation is designed to rupture near the pylorus as a result of gaseous
distention and deformation by peristaltic abrasion and to deliver the drug
into the duodenum. In this form carbenoxolone is as readily absorbed as

when administered by tablets, although it is not certain as to which area of
the gastro-intestinal tract is involved and that the time to rupture may vary
(Lindup et al., 1970).
7.2 Distribution
Binding to plasma proteins can influence the distribution, pharmacological

properties and excretion of drugs, particularly if the drug is very highly
bound (Meyer and Guttman, 1968). During absorption studies Downer et
ui.,( 1 970) noted high blood concentrations for carbenoxolone. This
suggested that most of the drug was in the circulating blood, and indicated
a high degree of binding of the drug to the plasma proteins.
Parke (1 972) studied the binding of the drug in-vitro to whole heparinized
plasma using an ultrafiltration technique involving centrifugation. At
therapeutic plasma levels (10-100 mg/mL), the drug was more than 99.9%
bound to plasma proteins of the male and female rat, dog, monkey and
man. Using molecular sieve chromatography on Sephadex G200
[carboxypropionyl-14C1,4]-carbenoxolone (1 00 mg/mL) in human blood
plasma was associated with the globulins (17%) and the remainder (83%)
was associated with albumin. Other in-vitro and in-vivo experiments,
using the radioactively-labelleddrug in kinetic studies and in fluorescence
determinations, have shown that carbenoxolone binds to human serum
albumin at two different classes of binding sites, with apparent association
constants of 107 and 3 x 106, respectively. This binding gives rise to a
pronounced conformational change in the albumin which appears to
enhance the binding of carbenoxolone still further.
Due to the high degree of binding of carbenoxolone to the plasma proteins,

the drug is absorbed from the gastrointestinal tract, is conjugated in the
liver and is excreted in the bile with very little appearing in the urine
(Parke, 1972). Studies with [14C]-carbenoxolone administered to rats
have confirmed that the drug is almost entirely located in the
gastrointestinal tract, the liver and the blood plasma. Experiments in
which [ 14C]-carbenoxolonewas administered intraperitoneally to rats
have shown that the radioactive drug and conjugates migrate back into the
gastrointestinal tract, in particular into the gastric mucosa. This suggested
that there may be proteins present in the stomach mucosa which have a
special affinity for carbenoxolone. No significant distribution of [14C]-

carbenoxolone into the kidney, body fat, brain, musculature or tissues,
other than that previously mentioned, was detected.
A comparative study of healthy adults and geriatric patients suggested that

protein binding of carbenoxolone was reduced in the elderly, and was
associated with lower plasma albumin concentrations (Hayes et ul., 1977).
The mean clearance, plasma half-life and volumes of distribution of
carbenoxolonewere 4.72 mLkg.hour, 16.3 hours, and 0.105 L1 kg,
respectively, in the healthy adult. This may be compared to the values of
3.28 mlkgohour, 22.9 hours, and 0.098 Lkg, respectively, which had
been obtained in elderly patients. It has been suggested that these factors
may contribute to the higher incidence of carbenoxolone side-effects in the
elderly. Lower plasma half-lives for carbenoxolone, varying between 5.6
and 10 hours, have also been reported (Thorton et uf.,1980).
7.3 Metabolism
Metabolism of carbenoxolone appears to be species dependent. Following

oral administration to the rat, carbenoxolone was reported to be hydrolyzed in
the gastrointestinal tract to the glucuronide and sulfate conjugates of p-
glycyrrhetic acid and succinate prior to absorption (Iveson et al., 1971; Parke,
1972). However, Bridges et al., (1976) found no evidence of metabolism
during absorption in the rat. In man, the ester linkage appeared to be stable
and carbenoxolone was absorbed largely unchanged (Downer et al., 1970).
Following absorption, the drug was transported, bound to plasma proteins, to
the liver and was conjugated with glucuronic acid and excreted in the bile
(Parke, 1972).
7.4 Excretion
When [14C]-carbenoxolone was administered to man, 70-80% of the

radioactivity was excreted in feces, 0.2-1% in urine, and 12-20% was excreted
in expired air as 14CO2. The radioactivity present in feces occurred as
carbenoxolone and represented the biliary excretion of carbenoxolone-30-
glucuronide, subsequently hydrolyzed in the intestine by the gut microflora,
rather than non-absorbed drug. The 14CO2 present in the expired air
represented the extent of hydrolysis of [ 4C]-carbenoxolone into P-glycyrretic

*
acid plus [ 4C]-succinate, since the latter is rapidly and completely oxidized
to 14CO2. A small amount of the radioactivity excreted in the urine was
present as urea and the remainder was likely to be derived from [14C]-
succinate for no terpenoid compounds were detected (Parke, et al., 1972;
Downer et al., 1970).
Following oral administration of [ 14C]-carbenoxoloneto rats, 60-75% of the

radioactivity was excreted as 14 C02 in the expired air, 12-35% in the feces
as carbenoxolone originating from bile, and 2% in the urine (Parke, 1972).
8. PHARMACOLOGY
8.1 Therapeutic Indications and Uses.
Carbenoxolone promotes ulcer healing and prevents ulcer relapse (Langman,

1980). The anti-ulcer activity is well established in both gastric and duodenal
ulcer patients (Cooke, et al., 1980), however, the precise mechanism of action
remains unclear. Several mechanisms of action have been proposed including
an increased level of prostaglandins, particularly E2, (Peskar, 1980; Rask-
Madsen et al., 1983). Minuz et al., (1984) reported an increase in
prostaglandin E2 activity following rectal administration,therefore suggesting
that carbenoxolone also has a systemic action. Further reported actions of
carbenoxolone which may contribute to its anti-ulcer effect include
stimulation of mucus secretion (Goodier et al., 1967; Dean, 1968) and
secretion of HCO3- into the unstirred mucus gel layer coating the gastric
epithelium (Allan and Gamer, 1980) ,promotion of mucosal cell proliferation
(Van Huis and Kramer, 198l), inhibition of mucosal cell exfoliation
(Domschke et al, 1977) and inhibition of peptic activity (Henman, 1970).
Pinder et al. (1 976) have reviewed the pharmacological properties of
carbenoxolone and the therapeutic efficacy of the drug in peptic ulcer disease.
In addition to the formulations used for oral administration, including tablets

for gastric ulcers and "position-release'' capsules for duodenal ulcer,
carbenoxolone is also used as a gel or mouthwash in the treatment of mouth
ulcers. Topical carbenoxolone has been used to treat orofacial herpes simplex
infections (Poswillo and Roberts, 1981). In-vitro, it has been shown to have
antiviral activity against various DNA and RNA viruses (Dargan and Subak-
Sharpe, 1986). Treatment with carbenoxolone sodium solution every 4 hours
as a mouthwash and gargle produced symptom relief and healing of
oropharyngeal ulceration, associated with herpes simplex virus, in HIV-
infected patients (Poswillo, 1990). Aphthous ulceration, which has been
linked with varicella zoster virus, has also been successfully treated with
carbenoxolone (Poswillo and Partridge, 1984).
Carbenoxolone also has anti-inflammatoryactivity (Finney and Somers, 1958;

Khan and Sullivan, 1968). When administered parenterally this activity was
almost one third that of hydrocortisone and it is reduced by adrenalectomy
(Sullivan, 1972).
8.2 Toxicity and Side-Effects.
Carbenoxolone sodium is a drug of relatively low acute toxicity in

animals: the LD50 intravenously in mice is about 200 mgkg and the
LD50 orally in rats is about 3.0 gkg. It produces necrosis if injected in
concentrations above I %. Chronic toxicity testing in animals, using doses
up to 40 times the therapeutic dose, produced no toxicity. Reproductive,
teratogenic and carcinogenic tests on carbenoxolone in animals showed no
significant adverse effects (Sullivan, 1972).
Carbenoxole sodium has mineralocorticoid-likeeffects and may produce

sodium and water retention and hypokalaemia (Porter, 1970). This may
cause or exacerbate hypertension, cardiac failure, weight gain, oedema,
alkalosis, and muscle weakness and damage (Davies et al., 1974;
Dickinson and Swaminathan, 1978; Ganguli and Mohamed, 1980).
9. REFERENCES
Allan, A., and Garner, A. (1980). Gut, 2,249-262.
Baron, J. H., Gribble, R. J. N., Rhodes, C., and Wright, P. A. (1975). In

Fourth Symposium on Carbenoxolone Sodium, Avery Jones, F.,
and Parke, D. V. (Eds). Butterworths, London. pp. 115-128.
Blanchard, J., Tang, L.M., and Earle, M. E. (1990). J. Pharm. Sci. 29,411-
414.
Blanchard, J., Boyle, J. O., and Van Wagenen, S. (1988). J. Pharm. Sci.
22, 548-552.
Bridges, J. W., Houston, J. B., Humphrey, M. J., Lindup, W. E., Parke, D.
V., Schillingford, J. S., and Upshall, D. G. (1976). J. Pharm.
Pharmac. 28, 117-126.
British Patent, (1960), No. 843,133.
British Pharmacopoeia, ( 1 993). British Pharmacopeial Commision

London, HMSO, p. 1 10.
Chinese Pharmacopoeia, (1985). Pharmacopeial Commision, Ministry of

Hygiene of the People's Republic of China, Beijing. p. 56-57.
Clarke, E.G.C. (1 986). Isolation and Identification of Drugs, 2nd Edn.

The Pharmaceutical Press, London. p. 430-43 1.
Coleman, T. J., and Parke, D. V. (1963). J. Pharm. Pharmac. 15,841-845.
Cooke, P. J., Vincent-Brown, A., Lewis, S. I., Perks, S., Jewell, D. P., and
Reed, P.I. (1980). Scand. J. Gastroenterol, L5 (Suppl. 65), 93-96.
Dargan, D. J., and Subak-Sharpe, J. H. (1986). J. Antimicrob. Chemother.

18 (Suppi. W), 185-200.
Davies, G. J., Rhodes, J., and Calcraft, B. J. (1974). Br. Med. J. 3,400-
402.
Dean, A. C. B. (1968). In A Symposium on Carbenoxolone Sodium,

Robson, J. M. and Sullivan, F. M. (Eds). Butterworths, London. p.
33-46.
Dickinson, R. J., and Swaminathan, R. (1978). Postgrad. Med. J. 54,836-

837.
Doll, R., Hill, I. D., Hutton, C., and Underwood, D. J. (1 962). Lancet 2,
793-796.
Domschke, W., Domschke, S., Hagel, J., Demling, L., and Croft, D. N.
(1 977). Gut s , 817-820.
Downer, H. D., Galloway, R. W., Honvich, L., and Parke, D. V. (1970).J.

Pharm. Pharmacol. 22,479-487.
Finney, R. S. H., and Somers, G. F. (1958).J. Pharm. Pharmacol. U, 613-

620.
Ganguli, P. C., and Deen Mohamed, S.(1980).Scand. J. Gastroenterol. 15

(Suppl. 65), 63-69.
Goodfriend, T. L., Levine, L., and Fasman, G. D. (1964).Science 144,

1344-1346.
Goodier, T.E.W., Horwich, L., and Galloway, R. W., (1967).Gut S, 544-

547.
Hayes, M. J., Sprackling, M., and Langman, M. J. S.(1977).Gut U,

1054-1058.
Henman, F. D. (1970).Gut 11,344-351.
Iveson, P., Lindup, W. E., Parke, D. V., and Williams, R. T. (1971).

Xenobiotica 1,79-95.
Iveson, P., Parke, D. V., and Williams, R. T., (1966).Biochem. J. UN,

28p.
Khan, M. H., and Sullivan, F. M. (1 968).In A Symposium on

CarbenoxoloneSodium, Robson, J. M. and Sullivan, F. M. (Eds).
Butterworths, London, pp. 5-14.
Kracmar, J., Kracmarova, J., Bokovikova, T. N., Ciciro, V. E., Nesterova,

G. A. Suranova, A. V. Truis, N. V. (1 990). Pharmazie 45,912-916.
Langman, M. J. S. (1 980). In Advunces in Ulcer Diseuse, Holtermuller, K.

H., and Malagelada, J. R. (Eds). Excerpta Medica, Amsterdam, pp.
406-4 1 5.
Lindup, W. E., Parke, D. V., and Colin-Jones, D. (1970). Gut U, 555-588.
Martindale. The Extra Pharmacopoeia, (1 993). 30th Edn. The

Pharmaceutical Press, London, pp. 873-874.
Merck Index, (1989). 1 lth Edn. pp. 1801.
Meyer, M. C., and Guttman, D. E. (1968). J. Pharm. Sci. 52,895-918.
Minuz, P., Cavallini, G., Angelini, G. P., Lechi, A., Brocco, G., Riela, A.,
Scuro, L. A., and Velo, G. P. (1984). Pharmacol. Res. Commun.
Ifi, 875-883.
Parke, D. V. (1 972). In Curbenoxolone in Gustroenterologv, Avery Jones,
F., and Sullivan, F. M. (Eds). Butterworths. London, pp. 19-32.
Peskar, B. M. (1980). Scand. J. Gastroenterol. 15(Suppl. 65), 109-112.
Peskar, B. M., Peskar, B. A., and Turner, J. C. (1976). J. Pharm. Pharmac.

2, 720-721.
Pinder, R. M.,Brogden, R. N., Sawyer, P. R., Speight, T. M., Spencer, R.

and Avery, G. S. (1976). Drugs 11,245-307.
Porter, G. A., (1 970). In Curbenoxolone Sodium, Baron J. H., and

Sullivan, F. M. (Eds). Butterworths. London, pp. 33-47.
Poswillo, D. E., and Partridge, M. (1984). Br. Dent. J. B,

55-57.
Poswillo, D. E. and Roberts, G. J. (1981). Lancet hi, 143-144.
Poswillo, D.E. (1 990). Lancet 8 13.
Rask-Madsen, J., Bukhave, K., Madsen, P.E.R., and Bekker, C. (1983).

Eur. J. Clin. Invest. 11,351-356.
Rhodes, C., and Wright, P. A. (1974). J. Pharm. Pharmac. X,894-898.
Sullivan, F. M. (1 972). In Carbenoxolone in Gastroenterology,Avery

Jones, F., and Sullivan, F. M. (Eds). Buttenvorths, London, pp. 3-
18.
The Pharmaceutical Codex, (1979). 1lth Edn. The Pharmaceutical Press
London, pp. 138-139.
Thornton, P. C., Papouchado, M., and Reed, P.I. (1980). Scand. J.

Gastroenterol. 1_5 (Suppl. 65), 35-38.
U. S. Patent, (1962). No. 3,070,623.
Van Huis, G . A., and Kramer, M. F. (1981). Gut 22,782-787.
10. ACKNOWLEDGEMENTS
The authors wish to thank Dr. J. O'Brien (NMR Unit, Trinity College
Dublin), Dr. P. Caplan (Mass Spectrometry Unit, University College
Dublin), Dr. M. Meegan (Department of Pharmaceutical Chemistry,
Trinity College Dublin) for their help and assistance, as well as Mr. J.
Steel and Mr. D. Proctor (Sanofi Winthrop, Newcastle Upon Tyne,
England) for the supply of carbenoxolone sodium and for information on
HPLC.
CLARITHROMYCIN
Isam Ismail Salem
Department of Pharmacy and Pharmaceutical Technology
University of Granada
1 807 1 -Granada
Spain

AND EXCIF'IENTS-VOLUME 24 All rights of reproduction in any form reserved.
46 ISAM ISMAIL SALEM
Contents
1. Introduction
2. Description
2.1. Structural and Molecular Formulae, Molecular Weight
2.2. Nomenclature
2.2.1. Generic Names
2.2.2. Chemical Name
2.2.3. Chemical Abstracts Number
2.2.4. Trade Names
2.2.5. Other Names, Abbreviations, and Drug Codes
2.3. Color, Appearance, and Odor
3. Synthesis
4.1 Powder X-Ray Diffiaction
4.2.2 Differential S c d g Calorimetry
4.3 Solubility
4.4 pHRange
4.5 Ultraviolet Absorbance Spectrum
4.6 Infrared Spectrum
4.7 Nuclear Magnetic Resonance Spectra
4.7.1 'H-NMR Spectrum
4.7.2 13C-NMRSpectrum
4.8 MassSpectrum
5.2 Identification
5.3 Thin Layer Chromatography
5.4 Structural Details
5.5 High Performance Liquid Chromatography
5.6 Microbiological Analysis
CLARITHROMYCIN 41
6. Stability
7. Pharmacokinetics
7.1 Adsorption
7.2 Bioavailability
7.3 Distribution
7.4 Elimination
8. Pharmacology
8.1 Mechanism of Action
8.2 Toxicity
9. References
1. INTRODUCTION
Clarithromycin is a new semi-syntheticantimicrobial 14-membered

macrolide exhibiting a broad in vitro antibacterial spectrum. Structurally, it
differs from erythromycin only in the substitution of an 0-methyl group for
the hydroxyl group at position six of the lactone [11, with increased tissue or
cellular penetration [2]. It has a more favorable pharmacokineticsprofile,
than erythromycin, which allows twice-daily administration and a possible
increase in compliance [3].
To improve the spectrum of activity and decrease the disadvantages

of erythromycin, a new generation of macrolide compounds has been
developed. These include azithromycin, clarithromycin,roxithromycin,
dirithromycin, micocamycin and rokitamycin. Azithromycin and
clarithromycin have been approved recently by the Food and Drug
Administration (Oct. 1991).
Clarithromycin appears to have more activity against Mycoplasma

pneumoniae and Chlamydia trachomatis [4-91. Furthermore, clarithromycin
(in combination with its microbiologically active metabolite, 14-hydroxy-

clarithromycin) has shown an additive or even synergistic activity against
Haemophilus injluenzae, a species that often is resistant of intermediate
susceptibility to erythromycin [101. The 14-hydroxy-clarithromycinitself is
twice as active as the parent compound.
The effect of combining clarithromycin with a variety of other drugs

for the treatment and prevention of disseminated M avium infection in
patients with AIDS is under investigation [l l-141. In addition, it has
demonstrated activity in vitro and in clinical infections against
staphylococci, streptococci, Haemophilus species, Campylobacter species,
Mycoplasma species, Chlamydia species, Mycobacteria species, and
Neisseria gonorrhoeae It has demonstrated activity, superior to that of
erythromycin, against Legionella pneumophilia; and is active against
anaerobes [ 15-171.
Clarithromycin was discovered and patented by Taisho

Pharmaceutical Co. Ltd. Japan (Watanabe et al., Eur. Pat. Appl. Ep 41,355
(CL. C07H17/08),09 Dec 1981; JP Appl. 80/75,258,04 June 1980, 18; US
Pat. 4,331.803), and is being marketed by Abbott Laboratories.
2. DESCRIPTION
2.1. Structural and Molecular Formulae, Molecular Weight
Molecular Formula: C38H69N013
Molecular Weight: 747.96

CLARITHROMYCIN 49
2.2. Nomenclature
2.2.1. Generic Names
Clarithromycin (BAN,USAN, rINN); clarithromycin (DCF);

claritromicina (DCIT)
2.2.2. Chemical Name
(2R,3S,4S,5R,6R,8R,1OR, 11R, 12S,13R)-3-(2,6-Dideoxy-3-

C,3-o-dimethyl-ct-~-ribo-hexopyranosyloxy)-11,12-
dihydroxy-6-methoxy-2,4,6,8,10,12-hexamethy1-9-oxo-5-
(3,4,6-trideoxy-3-dimethylarnino-P-~-xylo-
hexopyranosy1oxy)pentadecan-13-olide. This structure is
shown in Figure 1.
2.2.3. Chemical Abstracts Number
CAS-81103-11-9.
2.2.4. Trade Names
Clarithromycin is marke-zd by Abbott under the proprie uy

names, "Biaxin USA", "Klacid Switz",and "Klaricid U P .
2.2.5. Other Names, Abbreviations, or Drug Codes
6-o-Methylerythromycin;A-56268; Abbott-56268; TE-031;

Erythromycin, 6-0-methyL.
2.3. Color, Appearance, and Odor
White to off-white crystalline odorless powder. Colorless needles

are obtained when the compound is crystallized from a mixture of
1:2 chlorofoddiisopropyl ether. During the synthesis of
so ISAM ISMAIL SALEM
Figure 1. Structure of clarithromycin.

CLARITHROMYCIN 51
clarithromycin, crystals were obtained from ethanol as orthorhombic

needles, for which [aID= -90.4"at 24°C (c=l .O, CHC13solution)
[11.
3. SYNTHESIS
The original synthesis of A-56268(TE-031) was performed by

Watanabe et al. in 1981. Clarithromycin was then obtained by methylating
0,N-dibenzyloxycarbonyl-des-N-methylerythromycin A with CH31,
deblocking, and subsequent N-methylation with CH20.
The preparation of clarithromycinwas reported by Morimoto et al.

in 1985 [11. 2'-0,3'-N-Bis(benzyloxycarbonyl)-N-dimethyl-e~omycin A
was methylated with CH31and NaH in dimethyl sulfoxide-tetrahydrofuran,
and the mixture was chromatographed on a silica gel column. One of the
products obtained after the separation was hydrogenated with Pd-black in
ethanol in the presence of sodium acetate-acetic acid buffer, which was
followed by reductive methylation with formaldehyde and hydrogen. The
substance was recrystallized from chloroform-isopropyl ether to give a
mixture of clarithromycin and 6,ll -di-o-methylerythromycin A [ 11.
The selective o-methylation of the C-6 hydroxyl group of

erythromycin A was achieved by Watanabe et al. in 1990 [18], using
erythromycin 9-oxime derivatives as the starting materials to obtain
clarithromycin.
To improve the synthesis method, Watanabe et al. [19] reported a

new method for the preparation of clarithromycin via the erythromycin A
quaternary ammonium salt derivative. The reactions involved in this
synthetic pathway are shown in Figure 2, and introduce three advantages.
First, the protection of the oxime and desosamine moieties is accomplished
by the use of benzyl bromide and sodium hydride in one pot. Second, the
removal of the three benzyl groups could be carried out using the CTH
method. Finally, the high selectivity of the methylation at the C-6hydroxyl
= Benyc)
Figure 2. Preparation of clarithromycin via the erythromycin A quaternary ammonium salt derivative.
CLARITHROMYCIN 53
group is sufficiently maintained. Clarithromycin was obtained in 53%

overall yield from erythromycin A 9-oxime.
Although in a previous synthesis [181, all the intermediateswere

obtained with good crystalline properties and clarithromycin could be
obtained in high purity, a large amount of benzyl chloroformate was
required at the step where the benzyloxycarbonyl (Cbz) groups were
introduced. The use of this reagent was a distinct drawback owing to the
severe irritating action and toxicity of benzyl chloroformate.
To resolve the handling difficulty and the problem of the elimination

of benzyl groups by hydrogenation during the synthesis via erythromycinA
quaternary ammonium salt derivative [19], Watanabe et al. [20] described a
facile synthesis of clarithromycin via 2’4lylethers of erythromycinA
derivatives (Figure 3). Using this synthetic pathway, it was possible to
prepare clarithromycinin a 48% yield from erythromycin A 9-oxime
without requiring the purification of each intermediate.
4.1. Powder X-RayDiffraction
The x-ray powder diffkaction pattern of clarithromycin powder sample was

obtained using a Philips diffkactometer system (model PW1710). The
pattern was obtained using nickel filtered copper radiation (h=1 S405 A),
and is shown in Figure 4. A full data summary is provided in Table I.
To date, only one polymorph or pseudopolymorph of clarithromycin has

been detected.
2.
Figure 3. Synthesis of clarithromycinvia the 2'-silylethers of erythromycin A derivatives.

le 28 40
Scattering Angle (degrees 2-8)
Figure 4. X-ray powder diffraction pattern of clarithromycin.

Table I
Crystallographic Data Deduced from the

X-Ray Powder Pattern of Clarithromycin
Scattering Angle d-spacing Relative Intensity

(degrees 20) (14)
8.60 10.273 67.4
9.56 9.2434 100.0
10.92 8.0951 60.6
11.56 7.6483 61.4
11.96 7.3934 26.7
12.48 7.0865 13.1
13.28 6.6613 8.4
13.84 6.3930 19.6
14.12 6.2669 25.8
15.24 5.8088 41.2
16.60 5.3358 18.8
17.00 5.2111 29.1
17.44 5.0806 35.0
18.20 4.8702 19.4
18.44 4.8073 18.0
19.12 4.6378 49.0
20.00 4.4357 17.9
20.60 4.3079 22.6
21.44 4.1409 10.4
21.64 4.1031 11.2
22.32 3.9796 25.1
23.20 3.8306 16.1
25.00 3S588 16.5
CLARITHROMYCIN 57
4.2. Thermal Methods of Analysis
4.2.1. Thermogravimetric(TG)Analysis
TG thermograms were obtained using a Shimazu TGA 50H

thermogravimetric analyzer, simultaneously connected to a Fisons
Instruments Thermolab mass detector and a Nicolet TGA interface Magna-
IR 550. The system was calibrated using the latent heat of melting of
Indium. The experiments were carried out in flowing nitrogen or air (20
mL/min) at different heating rates (from 1O"C/min to 2OoC/min). The
sample sizes used ranged between 6-8 mg, and were analyzed over a
temperature range of 30°C to 650°C. Mass and IR spectrums of the gases
produced during the analysis were recorded.
The TG thermogram, and its first derivative, are shown in Figure 5

for a 7.196 mg sample of clarithromycin contained in an alumina cell. The
TG studies indicated the loss of 91.O% at temperature values above 300°C.
4.2.2. Differential Scanning Calorimetry (DSC)
The thermal behavior of clarithromycin was further examined by

DSC, using a Shimazu DSC-50 differential scanning calorimeter. The
system was calibrated with a high purity sample (5 mg) of Indium.
Clarithromycin samples of 5-6 mg were run at a scanning rate of 5"C/min,
over a temperature range of 30 to 400°C. Changes caused by fusion-cooling
processes also were studied, and the peak transition and enthalpies of hsion
were determined for all samples.
DSC curves of clarithromycin showed one endothermicpeak of

fusion, having a peak maximum of 225°C. When examined by hot-stage
microscopy, the melting of the solid was observed to take place at the same
temperature value. The DSC thermogram shown in Figure 6 shows a single,
sharp, melting endotherm with an onset temperature of 222°C. Integration
of the melting endotherm permitted an estimation of the enthalpy of fusion
(AH) for clarithromycin as -41.29 J/g.
8.00.
I
-0.00
6.00 -
4.00. -2.00
2-00.
321.09 'C --4.00
0.OOJ I I I I
4.13
Temperature ("C)
Figure 5. Thermogravimetryprofile (and its first derivative) of

clarithromycin.
CLARITHROMYCIN 59
0.8'
a7-
a6-
a5-
a4-
224.9 oc --
1 1 1 " I I I I . I I . 1 1 I I I I . , a 1 . I , I I , ' 1
im 150 2al 250 300 350 400
Temperature ("C)
Figure 6. DSC thermogram of clarithromycin, obtained at a heating

rate of S"C/min.
As noted, the exothermic decomposition peak has an onset

temperature at 295"C, and a peak temperature at 320°C. This finding, and
the data obtained from the IR and MS gas analysis, reafltirmed that the
compound can undergo melting without simultaneous decomposition.
4.3. Solubility
Clarithromycin is soluble in acetone, and is slightly soluble in

methanol and in ethanol. It is practically insoluble in water.
The solubility of clarithromycinhas been studied in different

solvents, as were the effects of pH (ranging from 2.4 to 8.4) and buffer
concentrations. A series of 0.05 M; 0.1 M phosphate buffer - water
solutions were prepared at pH 2.4,5.4,7.4, and 8. An excess of
clarithromycin was added to each medium, which were then shaken for 24
hours at 25°C. Once equilibrium was reached, the samples were centrifuged
at 10,000 rpm for five minutes. The Supernatantswere clarified by filtration
through a 0.45 pm membrane, and analyzed by HPLC (method described in
section 5.5). All assays were conducted in triplicate.
At lower pH values, it was found that the solubility of clarithromycin

exhibited a slight buffer salt effect, which was most pronounced at high pH
values (Figure 7). The solubility of clarithromycin was significantly
increased at lower pH values, while the solubility was significantly
increased when different concentrations of methanol (more than 80% v/v)
were added to the stock solutions.
4.4. pHRange
Clarithromycin is a basic substance, and the pH of its aqueous

solutions is therefore dependent on the solute concentration. This behavior
is depicted in Figures 8 and 9, which illustrate the relation of solution pH
and the concentration of clarithromycin in media consisting of 80:20 v/v
methanol-water and in pure water, respectively.
CLARITHROMYCIN 61
0 Water
C 14
I -*- Buffer 0.05M
a 12
r * Buffer 0.1M
10
i
rn
t
9 8
r
m 6
0
m L 4
0
pH = 2.4 pH = 5.4 pH =7.4 pH = 8.4
Figure 7. Effect of pH and salt concentration on the solubility of

clarithromycin.
PH
0.0 0.5 1 .o 1.5
Concentration (mg/mL)
Figure 8. Effect of clarithromycinconcentration on the apparent pH of

80:20 (v/v) methanol-water.
CLARITHROMYCIN 63
1
8.5
0.0 0.1 0.2 0.3 0.4 0.5
Concentration (mg/mL)
Figure 9. Effect of clarithromycinconcentration on the pH of water.

4.5. Ultraviolet Absorbance Spectrum
The UV spectrum of clarithromycin was obtained using a Perkin-

Elmer Lambda 5 UVNIS spectrophotometer. The spectra were scanned
from 190 to 400 nm at 60 d m i n (2 nm spectral slit width), with the
solutions being contained in 1 cm quartz cells. Solution concentrations of
2 mg/mL were used, and the data were obtained in methanol; chloroform, or
methanol-water mixtures.
Typical spectra of clarithromycin dissolved in methanol and in

chloroform are shown in Figure 10. In methanolic solution spectral maxima
were observed at 21 1 and 288 nm, while peaks at 240 and 288 nm were
detected with chloroform as the solvent.
4.6. Infrared Spectrum
The infrared spectrum of clarithromycin, obtained in a KBr pellet, is

shown in Figure 1 I . The spectral peaks have been assigned to various
molecular vibrations, and these are contained in Table 11.
4.7. Nuclear Magnetic Resonance Spectra
4.7.1. 'H NMR Spectrum
The one-dimensional proton 'H NMR spectrum of 50 mg/mL

clarithrornycin dissolved in CDCl, is shown in Figure 12. This spectrum
was recorded on a General Electric QE-300 N M R system, and was
internally referenced to TMS. Table I11 lists the 'H Nh4R spectral
assignments of clarithromycin in CDCl,.
4.7.2. 13CNMR Spectrum
Figure 13 shows the one-dimensional I3C N M R spectrum of

clarithromycin dissolved in CDCl,. This spectrum was also recorded on the
QE-300 N M R system at a solute concentration of 50 mg/mL. The spectrum
CLARITHROMYCIN 65
I
300.0 400.0
.
190.0
Wavelength (nm)
Figure 10. W absorption spectra of clarithromycin in methanol and in

chloroform.
ISAM ISMAIL SALEM
4000 3500 3000 xm 2000 1800 1600 -,im 1200 1000 800 600
Wavenumber (cm )
Figure 1 1. Infrared spectrum of clarithromycin.

CLARITHROMYCIN 61
Table I1
Infiared Spectral Assignments for Clarithromycin
Energy (wavenumbers) Assignment
1690 uC4 (Ketone carbonyl)
1730 Lactone carbonyl
1420 (N-CH,)
2780-3000 Alkane stretching peaks
3450 Hydrogen bonds between OH groups
1000-1200 -C-0-C- stretch
1340-1400 CH, groups

-ri
I
CLARITHROMYCIN 69
Table I11
Proton Nuclear Magnetic Resonance ('H-NMR)

Assignments for Clarithromycin
Chemical Shift Number of Proton Assignment

(Multiplicity)
0.842 4
1 14-CH3
1.133 S 6-CH3
2.282 S N(CH3)2
3.038 S 6-OCH3
3.20 1 dd 2'-H
3.330 S 3"-OCH,
3.676 d 5-H
3.763 d 11-H
3.782 dd 3-H
4.449 d 1'-H
4.934 dd 1 "-H
5.064 dd 13-H
was recorded at 24°C and internally referenced to TMS. The 13CNMR

spectral assignments are contained in Table IV.
4.8. Mass spectrum
Mass spectra of clarithromycin were recorded on an Hewlett-

Packard model 5988-A mass spectrometer. The CI mass spectrum, acquired
with 70 eV chemical ionization, is shown in Figure 14. The LSIMS mass
spectrum (Figure 15) of clarithromycin was obtained using the VG-70 SE
system, using 3-nitrobenzyl alcohol as a matrix. The molecular ion (M-H)
was observed at 748 d z , and some characteristic peaks are noted at d z
values of 158,590, and 116.
5. METHODS OF ANALYSIS
5.1. Elemental Analysis
The following table shows the data calculated and found for the elemental
analysis of clarithromycin.
0 cal %LEQud
carbon 61.02 60.57
hydrogen 9.30 9.13
nitrogen 1.87 1.82
oxygen 27.81 28.48
5.2. Identification
Clarithromycin may be identified on the basis of its characteristic

infrared absorption spectrum O(Br pellet method), as described in section
4.6.
r
a
a
0
-cv
0
-4-
G
-a
E
C
-m
0
h
.L
c4
(
.o
7
Table IV
Carbon Nuclear Magnetic Resonance (I3C-NMR)

Assignments for Clarithromycin.
Carbon Chemical Shift Carbon Chemical Shift

number @Pm) number (Ppm)
1 175.900 1' 102.789
2 45.021 2' 70.940
3 77.934 3' 65.501
4 39.261 4' 28.530
5 80.710 5' 68.710
6 78.361 6' 2 1.440
7 39.360 7,s' 40.2 12
8 45.184 96.045
1"
9 22 1.ooo 2** 34.835
10 37.152 3'I 72.629
11 69.009 4" 77.414
12 74.214 5I' 65.750
13 76.571 6" 18.649
14 20.950 7" 21.423
15 10.550 8" 49.43 1
16 15.910
17 8.990
18 19.698
19 17.945
20 12.237
21 15.941
22 50.579
CLARITHROMYCIN 73
50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 a / Z
Figure 14. The chemical ionization mass spectrum of clarithromycin.

Figure 15. The LSIMS spectrum of clarithromycin, obtained using 3-nitrobenzyl alcohol as the matrix.
CLARITHROMYCIN 15
5.3. Thin Layer Chromatography
A thin layer chromatographymethod was developed and used by

Morimoto et al. [11 during the synthesis of clarithromycin. The samples
were applied to TLC silanized silica gel plates, and the plates developed in
2:3 phosphate buffer (0.1 M, pH 7)-Methanol. In this system, the Rfvalue
of clarithromycin was found to be 0.42.
5.4. Structural Details
The molecular structure of clarithromycin is similar to that of

erythromycin A, and to that of (14R)-14-hydroxy-6-o-methylerythromycin
A [2 11. The absolute configuration of the asymmetric centers in
clarithromycin was determined by Iwasaki and Sugawara in 1993 [213.
5.5. High Performance Liquid Chromatography (HPLC)
A HPLC method for clarithromycinwas developed in the author’s

laboratory, based on UV detection at 2 10 nm. For the present method, the
analytical apparatus consisted of an LC-6A high-pressure pump and a
Shimadzu SPD-6A variable-wavelengthdetector. Injections were made by
SIL-1A loop (20 pL) injector. A prepacked 30 cm x 3.9 mm ID p-
bondapak CI8Waters column was used, with the back pressures ranging
between 1800 and 2000 psi. The mobile phase consisted of 65% methanol
and 35% (v/v) 0.05 M monobasic sodium phosphate. The pH of the buffer
component was adjusted to 4.0 using orthophosphoric acid, and a flow rate
of 1.O mL/min was used for all work.
Owing to the inadequate solubility of clarithromycin in the mobile

phase, serial dilutions of the drug were made by first dissolving
clarithromycin in methanol. Samples were subsequently diluted to the
desired volume with mobile phase. The standard solutions were injected
into the HPLC system five times, and average values deduced from the
mean of the five measurements. The calibration curve covered the
concentration range of 0.005 1 mg/mL to 1.08 mg/mL, and was found to be
linear with a correlation coefficient equal to 0.9960 (r2=99.20 %). The
76 ISAM ISMAL SALEM
precision of this assay was found to characterized by a relative standard

deviation of 1.72%, and the limit of detection was deduced as 0.04 pg/mL.
A typical chromatogram obtained using this method is shown in Figure 16.
Numerous HPLC procedures for clarithromycin have been reported

by other authors to identify and quantify the drug in biological samples [22,
231, and as methods for the analysis of clarithromycin and related
compounds of the synthesis [20].
A method suitable for the determination of clarithromycin in body

fluids was developed, which makes use of a Nucleosil Cs (5 pm, 250 mm x
4.6 mm ID) column and a mobile phase of 39:9:52 acetonitrile-methanol-
(0.04 M) NaH2P04 [24]. The pH of the medium was adjusted to 6.8 with
NaOH. The flow rate was set at 1.2-1.4 mL/min. Electrochemical detection
was used to monitor the analysis, with the potential of the screening
electrode being set at M.5 V and the working electrode at +0.78 f 0.04 V.
A quantification limit of 10.03 p g / d in plasma was established, and a
relative standard deviation of less than 5% was obtained.
Clarithromycin was extracted using the following procedure. 0.5

mL aliquots of plasma are transferred to clean tubes. Approximately 750 ng
of internal standard (erythromycin A 9-o-methyoxime, dissolved in 1:1
acetonitrile/ water) is added to each tube, along with 0.2 mL of 0.1 M
sodium carbonate solution and 3 mL of 1:1 ethyl acetate-hexane. The
samples are then vigorously vortexed for 1 minute, and centrifuged at 800 g
for 5 min. The organic layer is transferred to a suitable container and
evaporated to dryness at 45 "C. The residue is dissolved in 200-400 pL of
50% acetonitrile/ water, and 20 pL portions were injected into the HPLC
system. A similar extraction procedure was used with urine samples.
The simultaneous detemination of clarithromycin and related

products was realized by Morgan et al. [25]. The mobile phase consisted of
4 8 5 2 v/v acetonitrile-KH2P04(0.33 M), with the pH being adjusted to
between 5.3 and 5.5. A flow rate of 1.0 mL/min was used, and a sample size
of 50 pL was found to be appropriate. A Cis Column (5 pm, 250 x 4.6 mm
ID) was used, which was heated to 50°C. The detection wavelength chosen
Figure 16. Typical HPLC chromatogram of clarithromychdissolved in methanol-watermixtures.
was 205 nm, with the detector sensitivity being set at 0.03 AUFS. The
percent of all known compounds was obtained as the area percent, and most
identified species were detectable at the 0.1% level.
5.6. MicrobiologicalAnalysis
Serum and tissue concentrations of clarithromycin were measured by

the agar dif3kion method with Sarcina Iutea ATCC 934 1 as the test
microorganism, and with antibiotic medium No. 11 (heart infusion agar)
126,271. A validated bioassay method for clarithromycin has also been
described by Fernandes [28]. The latter assay consists of the use of
Micrococcus luteus ATCC 9341 as the indicator organism.
6. STABILITY
Clarithromycin is stable under normal storage conditions. It should

be stored in tight containers, protected &om light. It is more stable to the
effects of acid than is erythromycin A, owing to the presence of the 6-0-
methyl group which blocks the formation of the 6,9;9,12-spiroketal
derivatives responsible for the gastrointestinal imtation associated with
erythromycin use. Although, clarithromycin gradually loses its antibacterial
activity in dilute HCl solution [l], its increased acid stability leads to
improved intracellular bioactivity .
The HPLC method described in section 5.5 has been used to test the
stability of aqueous and hydroalcoholic solutions of clarithromycin prepared
during the solubility study. No degradation products were observed in these
samples when they were maintained at 4°C for 20 days.
CLARITHROMYCIN 79
7. PHARMACOKINETICS
7.1. Adsorption
Clarithromycin is stable in gastric acid, and is rapidly absorbed from

the gastrointestinal tract regardless of when it is taken. Food intake before
dosing slightly delays both the onset of absorption and slightly retards the
formation of the 14-hydroxy clarithromycin antibacterial active metabolite.
It actually appears that the bioavailability of clarithromycin is improved by
its administration with food. [29-3 11. This suggests that clarithromycin can
be taken orally (in tablet form or in suspension) without concern for timing
in relation to meals.
7.2. Bioavailability
The absolute bioavailability of clarithromycin, after oral

administration, has been reported to be approximately 55 % [30]. It has a
long serum half-life (4.9 hours), and exhibits peak serum concentrations of
2.51 p g l d within two hours after administration of a single 500 mg dose in
a fasting, healthy subject [(32-341.
The rapid first-pass metabolism of clarithromycin leads to the

formation of its active metabolite (14-hydroxy clarithromycin), which also
reaches a peak serum concentration of 2.1 p g / d within two hours after
administrating a single 500 mg dose [32-341.
Steady-state amounts (1 pg/mL) of clarithromycin and the 14-

hydroxy metabolite are reached after 2-3 days of administering a 250 dose
every 12 hours. For both compounds, the steady-state peak plasma
concentrations in children (following 7.5 mgkg every 12 hours of drug
product suspension) were 3-7 and 1-2 p g / d , respectively [33].
7.3. Distribution
Clarithromycin and the 14-hydroxy metabolite are widely distributed

into most body tissues, and reach especially high concentrationsin the lung.
Tissue concentrations exceed those of serum and because of high

intracellular concentration ,negligible accumulation is observed [35-371.
The protein binding of ciarithromycin in vitro is low, and 14-hydroxy
protein binding decreases with increasing serum drug concentration.
7.4. Elimination
Clarithromycin is largely metabolized in the liver, by the hepatic

cytochrome P-450 enzymes. The major metabolic pathway is by
hydroxylation at the 14 position, and by oxidative N-demethylation [38].
Clarithromycin and its principal metabolites are excreted in feces via bile, in
urine by renal and nonrenal mechanisms. Between 20-30% of the dose is
excreted in this way as the unchanged drug [38]. Clarithromycin follows a
one-compartment, open pharmacokinetic model, and its elimination seems
to follow nonlinear dose-dependentpharmacokinetics [38-39].
8. PHARMACOLOGY
8.1. Mechanism of action
Like the rest of the macrolide group, clarithromycinexerts its

antibacterial action by binding to the 50s ribosomal subunit of susceptible
organisms and by inhibiting protein synthesis through translocation of
aminoacyl transfer-RNA [40]. The site of action of clarithromycin seems to
be the same as that of erythromycin.
Clarithromycin, like other sixteen-membered macrolides, is a poor

inducer of mRNA and does not itself cause activation of the methylase
enzyme. It thereby retains activity against inducible bacteria in the absence
of a strong inducer [30].
The activity of clarithromycin is equivalent to between two and

fourfold that of erythromycin against all isolated tested microorganisms.
Unlike erythromycin, it generates in vivo an active metabolite (14-hydroxy
clarithromycin),which by itself often exhibits more activity against bacteria
CLARITHROMYCIN 81
than does erythromycin. The combmation of clarithromycinand its

metabolite yields a synergistic effect [4,11,13]. The compound has another
major advantage over erythromycin, its activity against Mycobucterium
Avium [41] and A4 leprae [2].
Unlike penicillin or cephalosporin antibiotics, the uptake of

clarithromycinby human neutrophils is high, leading to higher concentration
of this drug in human macrophages, lymphocytes and polymorphonuclear
leukocytes. It thereby displays major activity against intracellular
microorganisms, such as S. uureus or Legionellu. [2,43].
The more potent anti-inflammatory effects exhibited by

clarithromycinmay enhance its clinical efficacy. It has been demonstrated
that clarithromycin inhibits the production of interleukin-1 (IL-1)by murine
peritoneal macrophages, lymphocyte proliferation, and lymphocyte
transformation of murine spleen cells at low concentrations [44].
8.2. Toxicity
No toxicity was described during clinical trials, and clarithromycin

has proven to be well tolerated. The most common adverse effects have been
mild-to-moderate GI irritation.
Hepatotoxicity occurred in all animal species tested at doses two

times greater than the maximum human daily dose. Renal tubular
degeneration occurred in rats,dogs, and monkeys at doses 3-8 times greater
than the maximum human daily dose. Corneal opacity and lymphoid
depletion in dogs occurred after the administrationof 3 to 12 times the
maximum human daily dose, respectively.
Clarithromycincauses teratogenic effects in laboratory animals. No

data are available in pregnant women; so it should not be used during
pregnancy, unless no alternative therapy is appropriate [45].
9. REFERENCES
1. Morimoto, S., Takahashi, Y., Watanabe, Y. and Omura, S. (1984). J
Antibiot. X ?187.
2. Anderson, R., Joone, G. and van Rensburg, E.J. (1 988). J.

Antimicrob. Chemother. 22,923.
3. Physicians G e m . (1995). 5* Ed., Riverside, CT, p. 463.
4. Perronne, C., Gikas, A., Truffot-Pernot, C. (1991).Antimicrob.

Agents Chemother. 35,1356.
5. Femandes, P.B. (1987). Antimicrob. News. 4,25.
6. Mor, N., Vanderkolk, J., Mezo, N. and Heifets. (1994).Antimicrob.

7. Hoppe, J.E. and Eichhorn, A. (1989). Eur. J. Microbiol. Infect. Dis.

8,653.
8. Rastogi, N. and Goh, K.S. (1 992). Antimicrob. Agents Chemother.

3,2841.
9. Gorzynski, E.A., Gutman, S.I. and Allen, W (1989).Antimicrob.

10. Olsson-Liljequist, B. and Hoffman, B.M. (1991). J. Antimicrob.

Chemother. 22, Suppl. A, 11.
11. Masur, H. (1993). N. Engl. J Med. 2 2 , 8 9 8 .
12. Fattorini, L., Li, B., Piersimoni, C. (1995). Antimicrob. Agents

Chemother. 19,680.
CLARITHROMYCIN 83
13. Furney, S.K., Skinner, P.S., Farrer, J. and Orme, I.M. (1995).
Antimicrob. Agents Chemother.22, 786.
14. Perronne, C., Gikas, A., Truffot-Pernot, C. (1990). Antimicrob.

Agents Chemother.34,1508.
15. Rolston, K., Gooch, G. and Ho, D. (1989). J Antimicrob.

Chemother.23, 455.
16. Hardy, D.J., Hensey, D.M., Beyer, J.M. (1988). Antimicrob.Agents

Chemother.2 , 1 7 10.
17. Sefton, A.M., Maskell, J.P., Yong, F.J. (1988). Eur. J Clin.
Microbiol. Infect. Dis. 2,798.
18. Watanabe, Y., Adashi, T., Asaka, T. (1 990). Heterocycles 3,

2 121.
19. Watanabe, Y., Kashimura, M., Asaka, T. (1993). HeterocycZes.Xi,

243.
20. Watanabe, Y., Adashi, T., Asaka, T. (1993). J Antibiotics.%, 1163.
21. Iwasaki, H. and Sugawara, Y. (1993). Acta Cryst. 49,1227.
22. Sundberg, L. and Cederberg, A. (1 994). J Antimicrob. Chemother.

33,299.
23. Ohtake, T., Ogura, K., Iwatate, C. and Suwa, T. Chemotherapy

(1988). Xi, 192.
24. Chu, S., Senello, L. and Sonders, R. (1991). J Chromutogr.521,

199.
25. Morgan, D., Cugier, P., Marello, B. (1990). J Chromatogr.m,

351.
26. Fraschini, F., Scaglione, F., Pintucci, G. (1991). J Antimicrob.

Chemother. ;?z, Suppl. 4,61.
27. Nagate, T., Sugita, K., Miyachi, M. (1988). chemotherapy. X, 170.
28. Fernandes, P.B., Hardy, D.J., McDaniel, D. (1989). Antimicrob.

Agents Chemother.11,153 1.
29. Bahal, N. and Nahata, M.C. (1992). Ann. Phurmacother. &46.
30. Neu, H.C. (1991). J Antimicrob. Chemother.22, Suppl. A, 1.
31. Davey, P.G. (1991).J Hosp. Infect.B,Suppl. A, 29.
32. Peters, D.H. and Clissold, S.P. (1992). Drugs. 44, 117.
33. Drug Facts and Comparisons (1995). Facts and Comparisons. St.
Louis, MO, p. 2003.
34. Fraschini, F., Scaglione, F., Pintucci, G. (1991). 31St Interscience

Conference on Antimicrobial Agents and Chemotherapy (1991).
Chicago. Abstract 51 2.
35. Fraschini, F., Scaglione, F., Pintucci, G. (1991). J Antimicrob.

Chemother. 25, Suppl. A, 73.
36. Kohno, Y., Ohta, K., Suwa, T. and Suga, T. (I 990). Antimicrob.
Agents Chemother.34,562.
37. Kohno, Y., Yoshida, H., Suwa, T. and Suga, T. (1990).J

Antimicrob. Chemother.26,503.
38. Ferrero, J.L., Bopp, B.A., Marsh, K.C. (1990). Drug Metabol. and
Dispos. Ls,441.
CLARITHROMYCIN 85
39. Saito, A., Shimada, J., Ohmori, M. (1 988). Chemotherapy.36, Supp.

3. 576.
40. Petska, S. “Inhibitors of protein synthesis”, in: Molecular

mechanism of protein biosynthesis (1977). Weissbach, H. and
Petska, S. eds. New York, p. 467.
41. Dautzenberg, B., Truffot, C., Legris, S.(1991). Am. Rev. Respir. Dis.
-
144,564.
42. Franzblau, S.G. and Hastings, R.C. (1988). Antimicrob. Agents

Chemother.32,1758.
43. Femandes, P.B., Bailer, R., Swanson, R. (1986). Antimicrob. Agents

Chemother.32,865.
44. Takeshita, K., Yamagishi, I., Harada, M. (1989). Drugs Ex-. Clin.
Res. k5,527.
45. AMA Drug Evaluations (1994). American Medical Association.

USA, p. 1450.
CROSPOVIDONE
Eugene S. Barabas' and Christianah M. Adeyeye2
(1) ISP Corporation

1 3 6 1 Alps Road
Wayne, NJ 07470
(2) Department of Pharmacy

Duquesne University
Pittsburgh, PA 15282
ANALYTICAL PROFILES OF DRUG SUBSTANCES 87 Copyright Q 1996 by Academic Press. Inc.

AND EXCIPIENTS-VOLUME 24 All rights of reproductionin any form reserved.
88 EUGENE S. BARABAS AND CHRISTIANAH M. ADEYEYE
Contents
1. Introduction
1 .I Structure
1.2 Nomenclature
1.3 Polymerization
2. Methods of Preparation
2.1 Official Methods
2.1.1 Preparation without Added Crosslinking Agent
2.1.2 Preparation with Crosslinking Agent
3,l Description of the Polymer
3.2 Glass Transition Temperature
3.3 Hygroscopicity
4. Primary Uses of Crospovidone

4.1 Pharmaceutical Applications
4.1.1 Tablet Disintegrant
4.1.2 Tablet Binder
4.1.3 Miscellaneous Pharmaceutical Uses
4.2 Medical Applications
4.3 Uses in Production of Alcoholic and Non-Alcoholic
Beverages
4.3.1 Stabilization of Beer
4.3.2 Stabilization of Wine
4.3.3 Stabilization of Other Beverages and Liquids
4.4 Miscellaneous Other Uses
4.4.1 Isolation and Stabilization of Enzymes
4.4.2 Use in Agriculture
4.4.3 Use in Analytical Chemistry
4.4.4 Use in Catalysis
5. Health and Safety

5.1 Acute Toxicity
5.2 Subacute Toxicity
5.2.1 28-Day Feeding Study in Rats
5.2.3 28-Day Feeding Study in Dogs
CROSPOVIDONE 89

5.3 Teratogenicity
5.4 Pharmacokinetics
5.5 Skin and Mucous Membrane Tolerance
5.6 Pharmacology
6. Compliance with Pharmacopoeia1and Food Regulations

6.1 Identification Tests
6.1.1 Reaction with Iodine
6.1.2 Infrared Spectrum
6.2 Compendia1Testing
6.2.1 Water Content
6.2.2 Nitrogen Content
6.2.3 pH
6.2.4 Non-Volatile, Water Soluble Content
6.2.5 Heavy Metals
6.2.6 Residue on Ignition
6.2.7 Vinyl Pyrrolidone Content
6.3 Other Characteristics
6.3.1 Soluble Poly(VinylPyrro1idone)
6.3.2 Arsenic
6.3.3 Zinc
6.3.4 N,N'-Divinylimidazolidone
6.3.5 Peroxides
6.3.6 Loss on Drying
6.3.7 Surface Area
6.3.8 Particle Size Distribution
6.3.9 Bulk Density
6.3.10 Flow Properties
6.4 Microbial Limit Tests
7. Interactions of Crospovidone with Drug Substances
8. References
1. Introduction
Crospovidone is the insoluble form of polyvinylpyrrolidone, and its use in

the pharmaceutical industry as a tablet excipient (tablet disintegrant and
binder) has been widely documented. It is medically used for the
treatment of some intestinal disorders, as solubilizing excipient to improve
the bioavailability of drugs (such as steroids), and as germicides in wound
treatment. It is also commonly used as clarifier in alcoholic and non-
alcoholic beverages.
1.1 Structure
j
Crospovidone is produced by the proliferous polymerization of
vinylpyrrolidone monomer:
HZC-
H2C
\/=O
--
"
Polymerization:
7 ' ti&
\/=O
I
CH =CH2 CH-CH2 n
The earliest observation of spontaneous "popcorn" polymer formation had

been made with dimethylbutadiene by Kondakov [I]. Later Staudinger
and Huseman found a similar phenomenon with the styrene-
divinylbenzene system [2]. It was, however, Breitenbach and his
coworkers who found that numerous other monomers [3] (including
vinylpyrrolidinone [4])were also capable of proliferous polymerization.
His work contributed significantly to the elucidation of the mechanism of
this unique reaction. The product of this polymerization is a densely
crosslinked structure insoluble within the system in which it is made. It
has a very low degree of swelling and consists of a very voluminous
structure which contains many voids. The polymer has a white, opaque
appearance, quite different from the normal crosslinked polymer of
vinylpyrrolidinone. Because of the formation of the "popped" structure,
this voluminous polymer is also referred to as a "popcorn" polymer. The
structure of "popcorn" PVP resembles a polymer foam, with its void space
CROSPOVIDONE 91
not being formed by a gas or vapor but instead developed by the

polymerization process itself.
An unusual property of this polymer type is that when it is brought in

contact with more monomer, it can transform the latter into a polymer of
the same ''popcorn" structure. In the absence of the proliferating seed, the
monomer would turn into a normal soluble polymer.
In the case of the vinylpyrrolidone "popcorn" polymer, the product is

formed through a complex mechanism in which both the copolymerization
with certain in situ formed crosslinking agents and the physical
entanglement of newly forming polymer chains contributes to the
development of the "popcorn" structure.
1.2 Nomenclature
Chemical Abstract Services Registration number: 9003-39-8
Chemical Abstract Name: Crosslinked homopolymer of 1-ethenyl-2-

pyrrolidinone
Crosslinked poly(vinylpyrro1idinone) has been known under a variety of

names. Some of those have been used as "approved names" by the
regulatory authorities of different countries. The commonly used names
include:
Crospovidone
Crosslinked Polyvidone
Crosslinked homopolymer of 1-ethenyl-2-pyrrolidone
Insoluble crosslinked homopolymer of N-vinyl-2 pyrrolidone
Insoluble PVP
Polyvinylpolypyrrolidone(PVPP)
PolyvinylpyrrolidonumInsolubilis
Crosslinked poly(vinylpyrrolidinone), beside being available in technical

grades with different specifications, is sold as a pharmaceutical grade
conforming to the requirements of various national and international
Pharmacopoeias, as well as to the demands of national and international
food regulatory authorities.
Pharmaceutical grades are marketed under the trade name of Kollidon CL

by BASF (Badische Anilin and Soda Fabrik A.G.) and Polyplasdone XL
by GAF Corporation (recently changed to ISP - International Specialty
Products Corporation). The products used by the beverage industry are
sold under the trade names Divergan (BASF) and Polyclar (ISP) [5,6].
1.3 Polymerization
The proliferous polymerization takes place through a free radical

mechanism, although the presence of a free radical initiator is not always
necessary to this type of polymerization. The radicals may also develop
through the rupture of polymer chains with the combined actions of
polymerization and swelling [7], which form a great number of active
sites. As a consequence, growing chains are initiated at different sites of
the polymer chains at fixed positions, which form independently growing
centers unable to interact with each other. These new chains, which also
contain the in situ formed or deliberately added bifunctional crosslinking
monomers, get entangled with chains already formed. The overall result
of these reactions is a system having a high crosslink density.
The monomer is absorbed by the swelling of the polymer network and

converted to a part of the network by polymerization. By the continuous
repetition of this process, the existing polymer chains first become strained
due to the swelling, and then rupture. This creates new free-radical sites
which react with more monomer molecules, producing new growing
chains. The straining of the polymer structure can be observed with the
help of polarizing microscopy [8]. The degree of optical anisotropy
depends upon the chemical character of the popcorn polymer [9].
While the presence of divinyl compounds is not indispensable for the

formation of popcorn polymer [lo], they often play an important role in
the development of the structure. Higher concentrations of divinyl
compounds produces a higher degree of crosslink density and yields
higher gel strengths. On the other hand, it also produces greater number of
pendant double bonds on the polymer, which leads to greater number of
growing chains and increases the tendency for chain splitting.
The unusual mechanicochemical part of the mechanism of the proliferous

polymerization was proven very convincingly by Breitenbach and Dwovak
CROSPOVIDONE 93
[l 11. These workers used a dimethacrylate Schiff-base as a crosslinking

agent. The swelling of the polymer made in this fashion was very low,
and showed the usual anisotropy to the polarizing microscope. After the
chemical crosslinking was destroyed by the addition of 0.1M
dichloroacetic acid, most of the optical anisotropy disappeared while the
polymer remained insoluble and only the degree of swelling increased
slightly. These results clearly indicated the presence of chain
entanglement, as well as the fact that there was some increase in the
swelling. This showed that the crosslinking agent was also responsible for
the formation of the structure, at least to some degree.
The number of fiee radicals formed during the splitting of C-C bonds,
together with the formation (or addition) of bifunctional monomers, are
mainly responsible for determining the rate of growth. This usually
follows an exponential law, with a linear dependence between the
logarithm of the weight of polymer (w) and the time of growth (t):
kw = dw f dt
and
-
- kt
W wo e
This is, however, an ideal law, and is exactly obeyed only when the newly
formed "popcorn" polymer has constant growth capability and the medium
remains unchanged during the process. Generally these conditions exist
only approximately.
Pravednikov and Medvedev studied the course of the proliferous

polymerization using I4C-labeledpopcorn seed which was added to
unlabelled monomer [ 131. These workers found that at the end of the
polymerization, the original labeled seed material was quite uniformly
distributed throughout the polymer. In order to achieve the nearly uniform
distribution of the seed, a great number of C-C bonds had to be ruptured
during the polymerization, thus creating a large number of free radicals.
The free radicals formed in the course of the bond splitting must be largely
responsible for the high rate of polymerization.
It had been found that the reaction medium exerts a great influence on the
course of the reaction [141. For instance, a styrene-p-divinylbenne
system containing 30 vol. % methanol shows evidence of proliferous
polymerization in about half of the time compared to a reaction without

methanol. This effect was attributed to the dimension of the polymer coils
and its dependence upon the nature of the solvent, with a reduction in coil
dimensions favoring popcorn polymer formation. It is assumed that an
optimum range of coil dimensions exists, which can be achieved by
adding a good soivent to the monomer-polymer system for which the
polymer is completely insoluble in the monomer. In the case of N-
vinylpyrrolidinone, this good solvent is water or (to a degree) methanol.
Unlike most other systems, poly(vinylpyrro1idinone) is soluble in its
monomer. It is therefore probable that the addition of water causes only
dilution, and a reduction of chain segment density at the same conversion
1141.
The PVP "popcorn" polymer may be prepared by two different methods.

In the first method (I), vinylpyrrolidinone is heated at temperatures
exceeding 100°C in the presence of an alkali metal hydroxide and a small
amount of water [ 151.
The presence of water has been shown to be an important factor in the

formation of the "popcorn", and also influences the rate and induction
period of the polymerization. This effect is most probably due to the
swelling of the polymer coils to the dimensions which favor "popcorn"
polymer formation [16]. Beside the physical entanglement, a certain
degree of chemical crosslinking must also be responsible for the densely
crosslinked structure of the vinylpyrrolidinone "popcorn" polymer. It has
been shown by pyrolysis gas chromatography that during the process, 1-
vinyl-3-ethylidenepyrrolidinone YEP) and ethylidene-bis-3-(N-vinyl-
pyrrolidinone (EVP) are formed. The amount of these unsaturated
compounds was found to be 1.5% and 0.1%, respectively [ 171. Structures
for these two compounds are provided on the following page.
CROSPOVIDONE 95
H
H2C-C=C-CH,
HzC,~,C=O
I I
I
CH=CH2
1 -Vinyl-3-Ethylidenepyrrolidinone
The formation of EVP is possible due to the two phase system that comes
into being because of the high concentration of NaOH which is used in the
system. The aqueous phase contains the caustic and part of the
vinylpyrrolidone, and under strong agitation turns to small droplets in the
organic phase. The organic phase consists of the rest of the
vinylpyrrolidone and any EVP (which has a very low water solubility of 2
mg/mL). As the reaction progresses, the water layer becomes the
continuous phase so any EVP forming in the process is protected from the
effect of the caustic.
One mole of vinylpyrrolidone monomer and one mole vinylpyrrolidone

carbanion (from the water phase) form an anionic adduct. The adduct then
splits to a thermodynamicallymore stable bifunctional compound and the
2-pyrrolidone anion. In the strongly basic environment, the bifunctional
compound isomerizes to EVP, while the 2-pyrrolidone ion hydrolyzes to
4-aminobutyrate [181. The proposed reaction mechanism is shown as
Scheme I.
In the second method (11) of production, an aqueous solution of

vinylpyrrolidinone and a small amount of a bifunctional monomer is
heated at temperatures exceeding 100°C [191. N-N y -divinylethylene urea
Step 1. &+ NaOH -0 N h
i/
VP Monomer
d
VP Carbanion
d
i/EVP
4AB
Reaction Scheme I
CROSPOVIDONE 97
and similar acid amides carrying two unsaturated groups are suitable
bifunctionals for this polymerization [21].
Polymers made by the two synthetic methods have been shown to exhibit
identical infrared spectra, which are also similar to that of linear PVP.
This finding can be attributed to the following:
a) The large crosslink density is partly due to physical
entanglements, so the number of covalent linkages is smaller
than usual
b) The structure of the bifunctional crosslinking agent is very
similar to that of vinylpyrrolidone. Any slight difference in the
structure is not sufficient for differentiation.
The corresponding infiared spectra are shown in Figure 1.
These "popcorn'' polymers made with 1.6% of the two aforementioned

bifunctionals are much more densely crosslinked than are polymers
obtained by the copolymerization of vinylpyrrolidinone with much larger
amounts of crosslinking agents and using a free radical initiator.
Consequently, the "popcorn" polymers show considerably lower swelling
characteristics than does PVP crosslinked with conventional crosslinking
agents. These correlation is shown in Figure 2.
The difference between "popcorn" and conventionally crosslinked

polymers can be demonstrated by studying their respective glass transition
(Tg) temperatures. The Tg of linear PVP K-90 is 175"C, and the Tg of
"popcorn" PVP is only 1520°Chigher (approximately 195°C). These
"popcorn" polymers which contain about 1.6% of in situ formed
crosslinking agents have a gel-volume of 5 mL/g. On the other hand, a
conventionally crosslinked polymer made with 1.6 mol-% crosslinking
agent, which has a Tg similar to that of the "popcorn" polymers (195°C)
and a gel volume of 42 mL/g. The more than eight-fold increase indicates
the existence of a more loosely crosslinked structure. If one uses a ten-
fold higher amount (1 6 mol-%) of bifunctional crosslinking agent in order
to increase the crosslink density of the conventionally prepared polymer,
the gel volume will be 12 mL/g, but the Tg will be as high as 270°C. This
is 95°C higher than that of the linear PVP polymers [ 171. This correlation
is shown in Table 1.
Table 1
Glass Transition Temperatures of Annealed

Linear and Crosslinked PVP [ 191
Sample Ta ("C)
povidone 175
crospovidone according to method I 190
crospovidone according to method I I 195
copolymer from VP with 1.6 mol-% 195

bifunctional monomer
copolymer from VP with 16 mol-% 270

bifunctional monomer
The Tg of crystalline polymers is higher than that of the amorphous ones

of the same composition. However, X-ray scattering studies failed to
show the presence of any crystalline domains in the structure of
amorphous "popcorn" polymerized vinylpyrrolidinone.
2.1 Official Methods of Preparation
At the present time there are two official methods for the preparation of
crospovidone. One of the methods was developed by GAF (now ISP)
Corporation and consists of a mechanicochemical sequence of reactions.
In this sequence, the network structure is developed without the addition
of crosslinking agents, through the inclusion of compounds having double
hctionalities which are developed in situ and take part in the
polymerization. The other method was developed by Badische Anilin and
Soda Fabrik A.G. BASF, and utilizes a unique crosslinking agent
(divinylimidazolidone) whose chemical structure is similar to that of the
vinylpyrrolidone monomer.
CROSPOVIDONE 99
m‘1700’ wo ’ wo’ Ilw ’ womm d i s
Energy (wavenumbers)
Figure 1. Infrared spectra of linear PVP and crospovidone [171, prepared

according to the two reaction schemes of section 2.1.
EUGENE S. BARABAS AND CHRISTIANAH M. ADEYEYE
30
20.
lo-
- 1
5
1
10 15 I
20
Concentration, Bifunctional Monomer (mol-%)
Figure 2. Sedimentation volume of AIBN-initiated PVP and

“popcorn”-PVP in water [ 171.
CROSPOVIDONE 101
2.1.1 Preparation (I), without added Crosslinking Agent
A mixture consisting of 200g of N-vinylpyrrolidone and 2g of sodium

hydroxide flakes is heated for 3 hours under total reflux in a distilling flask
at reduced pressure (100 mm Hg). At this time the temperature rises from
145-156°C to 190"C, and the refluxing monomer gradually converts to a
white solid. After cooling, the polymer is slurried with water to wash out
the caustic and any unconverted monomer. The slurry is filtered, and
dried at 50-60°C in vucuo [ 151.
An alternative procedure exists where a reaction mixture consisting of

vinylpyrrolidone monomer at a concentration of about 70-90% by weight
in an aqueous strongly basic solution (containing about 0.3-1.5 % base) is
heated to a temperature of about 130-170°C under an inert gas atmosphere.
The temperature is kept at the reaction temperature for a sufficiently long
time so as to create in situ crosslinking. After that, the basic reaction
mixture is diluted with distilled water to a vinylpyrrolidone concentration
of about 5-30 %. The reaction is continued at about 100°C to form a white
crosslinked polymer of low swell volume [20].
2.2.1 Preparation (11), with Crosslinking Agent [21,22]
In a vessel having a capacity of 500 parts by volume and equipped with a

thermometer and reflux condenser, a mixture is prepared consisting of 100
parts of vinylpyrrolidinone, 100 parts of distilled water, 1 part of N,N'-
divinylimidazolidinone, and one bare-metal iron packing element (such as
a Pall ring, 15 x 15 mm). About 0.005 % of dibenzoylperoxide (based on
the vinylpyrrolidinone content) is added, and the mixture heated to 35°C.
After approximately 90 minutes small white polymer seeds are seen on the
surface of the packing element and these seeds grow visibly. The growing
mass soon projects above the level of the liquid and eventually fills the
entire volume of the vessel. During polymerization, the reaction mass
heats to its boiling point of 102°C. Vaporized water is condensed in the
reflux condenser and flows back into the vessel.
The period between the appearance of the first polymer seed and the point
at which the entire volume of the vessel is full of white, crumbly polymer
mass is about 15 minutes and takes place after consumption of all of the
I02 EUGENE S. BARABAS AND CHRISTIANAH M. ADEYEYE
liquid phase. Boiling slows down soon afterwards and eventually stops
completely.
The reaction product is removed from the vessel, washed three times with
distilled water to remove soluble portions, and dried in a vacuum oven at
80°C. The yield is 90 parts of a pure, white granular crumbly polymer,
which is sparingly swellable in water, but completely insoluble in the
usual organic solvents (such as hydrocarbons, alcohols, ethers, ketones,
organic halogen compounds, and organic nitrogen compounds. The
product is non-fusible and decomposes above 300°C.
3.1 Description of the Polymer
Crosslinked poly(viny1pyrroiidinone) is a white to off-white free-flowing

powder. It is practically odorless, although sometimes it may exhibit a
faint (but not objectionable) odor, It may have a slightly salt-like taste.
The material is hygroscopic and should be kept in tightly sealed container.
Because of its crosslinked structure, the polymer is insoluble in water and

all ordinary solvents. However, it will swell when in contact with water,
as well as with some organic solvents. Because of its insolubility, the
molecular weight of the polymer is indeterminate.
3.2 Glass Transition Temperature
The glass transition temperature (Tg) of crospovidone varies with the

method of preparation, and whether the polymer co-exists with the,
vinylpyrrolidone monomer. A summary of glass transition temperature
data was presented in Table 1.
3.3 Hygroscopicity
Because of its hygroscopic nature, reference standard material must be

kept away from atmospheric humidity and is to be dried at 105°C for one
hour before its use.
CROSPOVIDONE 103
4. Primary Uses of Crosslinked PVP
The primary pharmaceutical application for crospovidone is that of a tablet

disintegrant, although it can also function as a tablet binder. In order for
the polymer to be useful as a pharmaceutical excipient, grades of material
need to possess the following properties [36]:
a) high swelling capacity
b) high capillary activity
c) high hydration capacity
d) low bulk density
e) large specific surface area
f) rapid uptake and high moisture absorption
g) complete insolubility in water
h) no tendency to form gel on contact with water
i) high binding characteristics
j) effectiveness in tablet disintegration
k) accelerated drug dissolution rates
1) good shelf stability
In addition, the polymer must be chemically and biologically inert, as well

as non-reactive with the other ingredients of the formulation.
Furthermore, it must be non-toxic and non-irritating, either when
administered orally or when applied externally.
4.1.1 Tablet Disintegrant
Crospovidone, because of its highly hydrophilic character, rapid moisture

sorption, and good swelling properties, is widely used as a tablet
disintegrant [35]. The specific surface area of the polymer is reasonably
large (1.25 m2/g), so it has very high capillary activity and hydration
capacity. As a consequence of these properties, water is rapidly drawn
into the tablet. The water uptake stretches out the folded molecular chains
lying between the crosslinks, causing an instant expansion of the polymer.
The increase of volume creates an internal pressure exceeding that of the
tablet strength, and results in fast disintegration of the tablet body [36].
104 EUGENE S.BARABAS AND CHRISTIANAH M. ADEYEYE
According to Huttenrauch and co-workers, the mechanism of

disintegration is very complex [37]. The efficiency of a disintegrating
agent requires a low water solubility, strong hydration capacity, good
plastic deformability, and high capillary activity [38,39]. Kornblum and
Stoopak, who were the first to report on the use of PVPP as a potential
high performance tablet disintegrant concluded that a large specific surface
area and substantial hygroscopicity are also required to assure fast and
complete disintegration [ 3 5 ] .
More recent studies carried out by List and M u m indicated that

neither capillarity nor the heat of adsorption were responsible for the
degree of disintegration, but that the pressure developed during the
swelling within the system is the decisive factor [40]. This internal
pressure is developed through the quick expansion of the system caused by
the absorbed water, implying that the rate of water sorption at the early
stages was particularly important. This matter was studied by Gissinger
and Stamm. through a comparison of Crospovidone with other
disintegrants [41]. The comparative degree of water uptake by some
commercially available tablet disintegrants after one minute of contact
[42] is shown in Figure 3.
Rudnic and co-workers also studied the mechanism of disintegration [43],

and concluded that while the magnitude of the force produced by the
swelling has an underlying relevance to disintegrant action, the rate of the
growth of that force must also have a strong influence on the
disintegration process. They proposed that the rate of swelling is a
function of the rate by which the force is increasing, and can be expressed
by :
dF
__ - dV
- K * -
dt dt
where dF/dt is the rate of the force development, dV/dt is the rate of
swelling, and K is a constant for any given formulation at constant
porosity. If the porosity is high, then the physical properties of the
disintegrant (surface area, density, etc.) will be the determining factors. If
K is small, dV/dt is influenced mainly by water absorption [43].
CROSPOVIDONE 105
25 -- AC-01-SOL
Figure 3 . Comparative water uptake of disintegrants after I minute of

exposure [44].
If the force grows slowly, then the elasticity of the tablet matrix will be
allowed to adjust to the stress without a consequent structural change. If
however, the force develops rapidly, the matrix will not be able to adjust
and the will structure rupture. The capacity of the disintegrant to sorb
water and swell as a result of the absorption can be evaluated an apparatus
designed by Nogami and coworkers [42].
A low degree of water solubility, or even complete insolubility in that

medium, is one of the most important prerequisites for a well-performing
disintegrant. If a polymer intended for use as a disintegrant has any
solubility in the medium or has a loosely crosslinked structure, the initial
stages of the water interaction will yield an intractable coating on the
tablet. This can partially or completely block the small pores of the tablet,
hindering or even completely stopping water penetration into the narrow
channels.
Any slight dissolution of the polymer will result in a solution of increased

viscosity. The higher the molecular weight of the dissolved polymer, the
higher will be the viscosity of the resulting medium. The viscosity
increase will slow down the absorption of the dilute polymer solution,
resulting in slow or prematurely-ended disintegration. The fact that
crospovidone is completely insoluble in water and in other solvents
eliminates the conditions for slow or uneven disintegration.
Scanning electron micrographs show that popcorn polymerized

crospovidone consist of an amorphous structure having no crystalline
domains. The solid consists of microspherical particles, 5-10 pm in
diameter, fused into agglomerates of 350-400 pm. This sponge-like
structure allows quick and free penetration of water, with a consequent
expansion of volume. The swollen network will shrink when dried, and
will also expand again on re-wetting. Owing to the nature of their
formation, tablets are not expected to swell isotropically. Khan and
Rhodes have determined the swelling ratio of tablets, and found that the
Ratio Value (the change in thickness to the change in diameter) was higher
than one [45].
Studies by Bronnsack showed that a higher tablet hardness was generally

detrimental to its disintegration [46]. Crospovidone, however, did not lead
to this type of behavior. The change in disintegration time of tablets made
CROSPOVIDONE 107
with crospovidone showed only minimal differences as a function of

tabletting pressure. Khan and Rooke reported that while the relationship
between compressional pressure and dissolution efficiency depended upon
the type of disintegrant, tablets made with crospovidone and dicalcium
phosphate dihydrate produced an increase in dissolution efficiency with an
increased tabletting pressure [47]. With tablets containing crospovidone
and lactose, the maximum dissolution efficiency was obtained when the
compressional pressure was between 1000-2000kg/cm.
In another report, Gordon et al. observed that there was no significant

correlation between changes in tablet hardness and dissolution after
storage at elevated temperature and no substantial swelling at the same
storage condition [29]. The crospovidone was incorporated
extragranularly, intragranularly, or by even distribution into the wet
granulated tablet formulation. This would imply that the mode of
incorporation of the disintegrant, or the method of manufacture of tablets,
may play a significant role in the rate of water uptake and resulting tablet
performance [29]. Jovanovic and co-workers showed that intragranularly
incorporated crospovidone was a more effective disintegrant for antacid
tablets than when it was added extragranularly [55].
Rudnic and co-workers found that an increase in the mean particle size
enhanced disintegration (and also powder flow and dissolution), but that
tablet hardness and friability were slightly better from finer grades [48].
Studies by List and Muazzam also showed that swelling pressure and
disintegrationtime were particle size dependent [49].
The swelling characteristics of crospovidone were studied by Wan and

Prasad using a video recording technique [50]. The Ferret diameters of the
swollen particles were 40- 120pm. The large differences between the
projected area and perimeter diameters observed in the dry state were
absent following hydration because swelling resulted in smoothing of the
particle edge texture.
Ringart and Guyot-Hermann found that for crospovidone (and also for
other disintegrants consisting of rounded particles), the most effective
concentration could be calculated using:
X = 0.32 J d r / d 2 [ ( D i / 0 2 + - 11 Di/Di,J
1 ox EUGENE S. BARABAS AND CHRISTIANAH M. ADEYEYE
where d, and d, are the densities of the disintegrant and drug respectively,
D1 and D2 are the average diameters determined by microscopy, and Di, is
the diameter of the disintegrant in the disintegration medium [51].
Hennig and Schubert compared crospovidone with other starch-based

disintegrants, and found that former was the most effective with respect to
disintegration time and compression strength [52].
As reported by Gordon and Chowhan, the polymer hygroscopicity can

directly affect its ability to act as a disintegrant [23]. These workers
observed a decrease in disintegrant efficiency and dissolution rates of
directly compressed tablets, resulting from the composite hygroscopicity
of the tablet formulation [53].
According to Wan and Prasad, the presence of other excipients can

influence the water uptake, so it is sometimes difficult to correlate water
uptake with decreased disintegration times [27]. The amount of
granulating fluid (water) containing crospovidone and the presence of
other excipients can also influence the water uptake [26]. When
crospovidone was omitted fiom a sulfanilamide formulation (but
containing 2% methylcellulose as a binder), the water uptake was high and
the disintegration time was short. In the presence of 2.5%crospovidone,
the same observations were noted, but when the amount of granulating
liquid was increased longer a disintegration time was noticed. This was
attributed to a decrease in the water uptake. Film formation and a more
even distribution of methylcellulose were reported to be responsible for
these differences in water penetration [26].
Van Kamp and co-workers studied the effect of water-uptake in tablet

disintegration, and determined that crospovidone showed the highest
penetration rate among the disintegrantstested [54]. The water uptake of
this excipient was approximately five times that of its own weight.
Johnson et al. found that the solubility and hygroscopicity of crospovidone
could affect disintegration efficiencies, and reported that the greater the
overall hygroscopicity and solubility of a naproxen tablet formulation
containing different disintegrants, the greater the decrease in disintegrant
efficiency [34].
CROSPOVIDONE 109
Crospovidone used in a sulfadiazine tablet formulation prepared at

ambient temperature and humidity was reported to increase the
bioavailability of the drug. When the same formulation was prepared
under 100% relative humidity, the disintegration time increased ,the
dissolution rate decreased, all with a consequent decreased urinary
excretion [28].
Moisture sorption of tabletted phenobarbitone sodium formulations and

tensile strength of the tablets were correlated in the work of Malamataris
and Dimitriou [25]. Formulations containing 36% w/w of the drug and
exposed to 93% relative humidity showed the greatest tendency for
moisture uptake and the minimum tensile strength. Van Kamp and co-
workers found that the crushing strength, disintegration, and dissolution
properties of tablets made by wet granulation with 20% potato starch as
the disintegrant could be markedly improved when the starch was replaced
by a much smaller amount (4%) of crospovidone [56].
Phadke and Anderson carried out studies on the wet granulation of powder
blends of acetaminophen and crospovidone, using hydroxypropyl
methylcellulose (HPMC) as the binder, and found that an increase in the
level of crospovidone led to an increase in the amount of fines in the
particle distribution of the dried granules [57]. At the same time, an
inverse ratio was found between the amount of crospovidone in the blend
and the bulk density of the formula. These studies indicated that the
interference in the hydration of HPMC and the increase in the total surface
area were attributable to the presence of crospovidone [57].
Wan and Prasad also found that the use of crospovidone led to increased
disintegration times when the molecular weight of the binder (methyl
cellulose) was increased, in spite of higher degree of water uptake [50].
Obviously, the hydrophilicity of the binder plays a crucial role in
influencing disintegration time. Wan and Choong found that the
disintegration and dissolution times of the tablet were functions of the
water penetration [58]. Differences in dissolution times were due mainly
to the absorptive power of the binder (starch), with the porous capillary
network in the tablet exerting only a secondary importance. In this respect
crospovidone was effective by also reducing the hydrophobic property of
the lubricant.
Using infrared spectroscopy, Casahourisat and co-workers found that the

interaction between excipients and drug influenced the disintegration time,
but not the drug dissolution rate [59]. Any correlation between dissolution
rate and compression force was found to depend upon the chemical
composition of the drugs tested.
Petroczki found that when crospovidone used as an adjuvant for the

tabletting of sulfamidine, salicylamide, bisubsalcyilate, and terpin hydrate,
very fast disintegration was obtained [60]. This finding permitted the use
of these active ingredients in tablets, while previously they could only be
formulated as suspensions. Experiments conducted by Esteve and co-
workers showed that the dissolution kinetics of phenylbutazone tabletted
with crospovidone were first order, and that the dissolution rate was
independent of the tablet hardness [61]. Researchers at Sandoz A.G.
found that tablets containing a griseofulvin-polyethyleneglycol dispersion
and formulated with crospovidone dissolved very fast [62]. Similar tablets
made with other disintegrants (such as alginic acid, sodium starch
glycolate, or cornstarch) dispersed the fungicide much slower.
In the studies of Miseta and co-workers, the release of poorly compressible

phenylbutazone was affected by the use of various disintegrants, and
crospovidone produced the best drug release [63]. Desai et al. reported
that the use of crospovidone improved the dissolution stability of
hydrochlorothiazide (HCTZ) capsules when compared to other
disintegrants such as Explotab or corn starch [64]. This was thought to be
associated with the moisture scavenging ability of the polymer, which
prevented the formation of traces levels of formaldehyde (a hydrolysis
product of HCTZ) in the presence of excipient-related moisture. Without
crospovidone, the generated formaldehyde would interact with the gelatin
capsule shell and the corn starch, resulting in the formation of less soluble
compounds and a consequently decreased dissolution rate.
Other investigations in which crospovidone was found to be effective as a

tablet disintegrant include the work of Sakr et al. [65], Gordon et al. [66],
Baykara et al. [67], Jovanovic ef al. [68), Botzolakis and Augsburger [69],
van Kamp et al. [70], Wan and Lai [71], and Liu et al. [72].
CROSPOVIDONE 111
4.1.2 Tablet Binder
Due to its good flow properties and plastic deformability, crospovidone

has good binding properties. These properties enhance the performance of
the polymer in spray, dry, and wet granulations. The polymer can also be
used for the direct tabletting of a variety of drugs (such as sulfadiazine,
phenacetin, phenazone, or pancreatin) without the need for granulation
[73,74]. The good compatibility of crospovidone with many organic and
inorganic active ingredients (and other excipients) makes this polymer
suitable for use in all types of dosage forms [43]. Gillard found that
crospovidone was an effective tablet binder when used at concentration
levels of 5-20% together with lactose [75].
4.1.3 Miscellaneous Pharmaceutical Uses
Fast dissolving pharmaceutical preparations (e.g.,indomethacin) can be

made by formulating solid dispersions of the drug in crospovidone. The
preparation can be made by suspending crospovidone in a solution of the
drug dissolved in a low boiling solvent, followed by the evaporation of the
solvent [76]. Crospovidone has also been used in the formulation of solid
dispersions of furosemide with the goal of improving the dissolution [77].
During comparisons of various excipients (such as PVP, croscarmellose
sodium, or PH- 101 microcrystalline cellulose), crospovidone was less
sensitive to the presence of other additives. However, the effect of other
excipients depended on their levels and on the drug concentration.
It has been found that when a poorly soluble drug is mixed with a water
swellable, crosslinked polymer, after vacuum drying the product the
dissolution rate of the drug increased considerably. For example,
griseofulvin and crospovidone were vacuum dried after standing for 24
hours in methylene chloride, and then exhibited a significant increase in
the dissolution rate [78].
High energy co-grinding of 6-methylene-rosta-1,4-diene-3,17-dione (an

aromatase inhibitor only slightly soluble in water) with crospovidone gave
a product exhibiting increased wettability and dissolution rate. As
determined by thermal analysis and x-ray diffraction studies, the
crystallinity of the drug decreased simultaneously [79]. In another system
where a drug (FCE-24304) was co-ground with crospovidone, it was
found that the high-energy milling resulted in a significant improvement

of the dissolution rate of the formulation [80].
A dry emulsion of griseofulvin was prepared by using crospovidone as a

solid support and a 1:4 combination of polysorbate 80 and sorbitan
monooleate as an emulsifier. The emulsion containing 55% (by weight) of
emulsifier had excellent physical stability 18 11.
A formulation having an increased dissolution rate was prepared by dry-

blending the active drug substance with a water-swellable, hydrophilic,
crosslinked polymer (such as crospovidone) in a ballmill for an extended
period of time (2 hours) at 70 rpm. Tablets made with the usual excipients
showed good disintegration characteristics, and noticeably increased rates
of dissolution [82].
Sustained release oral formulations were successfully made for active

substances whose solubility was known to be dependent upon pH. These
oral formulations consisted of a weakly basic drug (such as dypyridamole,
cinnarizine, or ketanserin), a water swellable polymer (crospovidone), and
a gastro-resistant polymer (e.g., a cellulose derivative or acrylic polymer)
within a hydrophilic or lipophilic matrix. It was found that the
formulation released the drug at the same rate in both gastric and enteric
environments [83].
The presence of crospovidone in suppositories was found to increase the

dissolution rate and absorption of antipyretic analgesics, such as
acetaminophen [841. A crosslinked mixture of poly(vinylpyrro1idone) and
poly(viny1 pyridine oxide) was found to be an effective hydrogel matrix
for the sustained release of drugs [85].
A Japanese patent application has been filed for a drug-treated surgical

bandage made from radiation crosslinked poly(vinylpyrro1idone) [86].
Crospovidone is used in the manufacturing of a medical tape, where the
topical device consists of a mixture of eperisone or tolperisone (or their
salts), crospovidone, and a base carrier. An adhesive formulation was
prepared by mixing the drug combination with crospovidone, and then
adding this combination to a solution of 2-ethylhexylacrylate and 2-
ethylhexylmethacrylate. The stirring of this mixture was continued until a
homogenous dispersion suitable to coat the base carrier was obtained [87].
CROSPOVIDONE 113
Crospovidone was used in forming a multilayer sustained release tablet

which placed ephedrine in two phases within the system. About 50% of
the drug was released from one layer in about 20-30 minutes, while the
remainder was released more slowly over several (10-20) hours [88]. The
combination of nifedipine with crospovidone in a sustained release
formulation changed the physical form of the drug from crystalline to
amorphous. The combination of the drug with the crosslinked,
hydrophilic polymer, increased the water solubility of former significantly
[89]. Crospovidone has also been recommended also for the stabilization
of pharmaceutical suspensions [90].
A combination of infrared spectroscopic and thermal analysis studies was

used to prove that oxamniquine and praziquantel do not interact
physicochemically, either with each other or with crospovidone. This
finding proved the feasibility of making a combination of the two in a
solid dosage form [91].
Crospovidone is used in the preparation of a two-phase composite,

conductive, pressure-sensitive adhesive. The continuous phase of the
hydrophilic adhesive is a solid state pressure-sensitive compound,
ionically-conductive regardless of the amount of water present in the
phase. The discontinuous phase is made up of domains of a hydrophobic,
pressure-sensitive adhesive, which enhance adhesion to mammalian skin.
In a typical preparation, crospovidone was swollen in glycerin and an
aqueous KC1 solution, and then mixed with Robond 60 acrylic latex. The
mixture was coated on a polyester backing pretreated with E-1700 Ag ink.
The adhesive properties of the tape were good, and the average skin
impedance on human subjects is reported to be 165 kQ [92].
4.2 Medical Applications
The non-toxic character, high complexing ability, and lack of solubility

makes crospovidone suitable for a variety of medical applications, and has
been tested successfully both in human and in veterinary medicine [93]. It
was found that crospovidone is beneficial for the treatment of infectious or
chronic diarrhea caused by food poisoning, change of diet, the or
excessive use of laxatives and diarrhea following the use of antibiotics. It
is recommended also for external use in the form of ointments, such as in
114 EUGENE S . BARABAS AND CHRISTIANAH M. ADEYEYE
the treatment and cicatrization of sores caused by varicose ulcers [94]. A

patent was received also for a composition of tannin and crospovidone that
was found to be effective for the treatment of diarrhea and wounds [95].
Medroxyprogesteron acetate is known to be an effective anticancer drug in

the treatment of breast cancer and endometrial cancer, but is characterized
by very low bioavailability. It was found now that this property could be
significantly enhanced by the addition of crospovidone to formulations. A
study was carried out on 22 female breast cancer and endometrial cancer
patients, with the new oral formulation being administered twice daily in
200 mg sachets. This treatment regime was compared to the standard
formulation, consisting of a 500 mg Farlutal tablet administered twice
daily. The bioavailability of the novel combination averaged 3 1/2 times
higher than that of the standard tablet [96].
The bioavailability of medroxyprogesterone acetate (MPA) was studied

with the participation of 26 female breast cancer patients. In the
randomized crossover study, the MPA formulation (using a 200 mg sachet
in which MPA had been loaded in crospovidone) was compared to the 500
mg standard tablet. The relative bioavailability of the MPA-crospovidone
formulation was approximately three times superior to that of the standard
formulation. This discovery might have important clinical implications
for the treatment of hormone-sensitive cancer [97].
An iodine complex can be made by dry-tumbling crospovidone with

elemental iodine. These complexes are efficient germicides and
disinfectants, and can be used as antiseptic dusting powders, as rubber
glove antiseptics for physicians and nurses, as foot powders, and for skin
treatments of pets and farm animals [98]. Blood, blood derivatives, other
body tissues, fluids and cells intended for transfusion or transplantation
can be disinfected by combinations containing iodine, hydrogen peroxide
and a carrier (such as crospovidone), which react with the germicide. The
preparation kills pathogenic microbes without affecting the utility of the
tissues, fluids, or cells [99].
Crospovidone, in combination with karaya gum, was found to be effective

in the treatment of chronic constipation without organic cause [ 1001. This
combination was recommended also as topical digestive agent for the
CROSPOVIDONE 115
treatment of chronic colonic diseases, such as colitis [loll, gastritis and

diverticulitis [1021.
Crospovidone was also recommended for use in the preparation of

hemodialysis membranes for artificial kidney machines. It was suggested
that these membranes may eliminate blood clotting problems [ 103,104].
Crosslinked poly(vinylpyrro1idone)was found to be suitable for the
fabrication of hydrogel contact lenses [1051.
It has been found that cotton dust and cotton stems contain naturally
occurring components which precipitate P-lipoprotein and y-globulin
(mostly IgG) in a non-immunologic manner. Sera of textile workers and
human controls gave similar reaction with these extracts. Treatment with
crospovidone eliminated the pseudo-immune reaction, thus making the
study of the pathogenesis of byssinosis possible [ 1061.
Intestinal contents and urine excreted through an artificial outlet in the

body (colostomy devices) are treated with crospovidone to facilitate their
handling [1071. Crospovidone was found to rapidly and efficiently absorb
bilirubin. Adsorption of bilirubin onto crospovidone reaches the
saturation point in a few minutes [ 1081. Crospovidone was used for the
stabilization of prostaglandin (especially of the PGE type), with the system
being plausible in a variety of dosage forms [ 1091. A patent was obtained
by the Yamanouchi Pharmaceutical Company for stable formulations
containing prostaglandin E [1lo]. A solution of nifepidin and povidone
was absorbed onto crospovidone and then dried. The resulting powder
was gelled with water, and tabletted to yield a sustained release system
[l 1I].
Crospovidone is used for the preparation of adhesives intended for use on

oral mucosa. The adhesive consists of poly(methacry1ic acid) or alginic
acid (or their pharmaceutically acceptable salts), and crospovidone at a
level of 550% of the total polymer. The polymers made with such
composition are excellent in their adhesiveness and water resistance. The
preparations can be used for the transmucosal delivery of saliva-mediated
sustained release of drugs, as well as for the protection of injury and
diseases in the oral cavity [112].
I16 EUGENE S. BARABAS AND CHRISTIANAH M . ADEYEYE
4.3 Uses of Crospovidone in the Production of Alcoholic and Non-

Alcoholic Beverages
As does PVP, PVPP forms complexes with a wide variety of natural

products, some of which can be found in various vegetable beverages
(both alcoholic and non-alcoholic). Since "popcorn" PVP is completely
insoluble. its complexes are equally insoluble and thereby removable from
beverages (such as beer, wine, vinegar, and fruit juices).
The current FDA regulation covering crospovidone reads as [ 1 131:

"The food additive poly(vinylpyrro1idinone) may be safely used in
accordance with the following prescribed conditions:
a) The additive is a homopolymer of the purified
vinylpyrrolidinone catalytically produced under conditions
producing polymerization and crosslinking such that an
insoluble polymer is produced.
b) The food additive is so processed that when the finished
polymer is refluxed for three hours with water, five percent
acetic acid, and 50 percent alcohol, no more than 50 parts
per million of extractables is obtained with each solvent. It
is used or intended for use as a clarifying agent in
beverages and vinegar, followed by removal with
filtration."
One of the most important uses of PVPP is the colloidal stabilization of

beverages made from raw agricultural ingredients. These materials
contain proteins and phenolic compounds, which during preparation and
storage polymerize and form species which react with the proteins to form
polymeric complexes of limited solubility. Smaller molecular weight
complexes form colloidal particles whose solubility is temperature
dependent. These are soluble in the medium at ambient temperature, but
become insoluble at lower temperatures (known as the "chill haze"). If,
however, the molecular weight of the polyphenol protein complex is high,
a so-called "permanent haze" is developed which is visible even at room
temperature.
CROSPOVIDONE 117
4.3.1 Stabilization of Beer
Beer is a unique and complex food, and also a very sensitive colloid
system. It is made through a multistep process which converts agricultural
raw materials into the beverage through a series of biochemical reactions.
Beer has been known almost throughout the history of mankind and its
purity and wholesomeness has always been a concern. As early as 1516 in
"Reinheitsgebot", the ruling prince of Bavaria proclaimed that, ?here shall
upon threat of withdrawal of the brewing charter, for every beer taken and
used no other material except barley, hops and water". While this
combination allows the brewing of a good product, at that time it could not
be taken into consideration that both malted barley and hops contain a
variety of natural chemical ingredients which can and do react with each
other. Some of the reaction products have been found to be objectionable.
One of the problems is related to the stability of the beer. Beer contains as
much as 150 mg/L of phenolic compounds, both as monophenols and
polyphenols [ 1141. These compounds are collectively denoted as
"flavonoids", and contain condensed rings systems of the following
general type:
Three different types of polyphenols belong to this group, and differ in the
oxidation state of Ring B:
CYANIDIN CATECHIN
QUERCmN
Polyphenols that belong to the cyanidin group differ in the number of

phenolic hydroxyls on Ring C. Beside cyanidin, pelargonidin and
delphinidine also belong to this class of polyphenols:
do. - -
OH
PELARGONIDIN CYANIDIN DELPHlNlDlN
and are called anthocyanidins. Anthocyanidins can be transformed into

red pigments by heating in dilute HCl [1 151.
The phenolic ingredients of the beer are either monomeric or polymeric. It

is difficult to differentiate between these, but usually members of the first
group have molecular weights smaller than 1000, while the molecular
weight of the polymeric polyphenols exceeds 1000. The compounds
belonging to the two groups can usually be separated by paper
chromatography [ 1 161. These polymerized polyphenols are erroneously
referred to as tannins. Natural tannins are compounds of intricate
structure, one of which is based on aromatic hydroxy acids (such as gallic
acid or hexahydrodiphenic acid [117]), while the other type has flavonoid
building blocks (such as catechins or anticyanogens [ 1 181). These
polyphenols are quite susceptible to oxidation, although their reaction with
air is very slow. On the other hand, they rapidly oxidize with oxydase
enzymes. These are copper-containing proteins which can be found in
plant tissues. The oxidation products (orthoquinones) are very reactive,
and easily undergo condensation reactions with proteins, yielding dark-
colored polymers [119].
The polymerization of polyphenols can be illustrated by the example of

the so called "Beer Constituent #12" (derived through the dehydration of
catechin and flavanols) which through the effect of slow oxidation and fast
acid catalysis undergoes this reaction:
CROSPOVIDONE 119
The fact that B is only the dimer of A is illustrative of the complicated

structure of polymeric polyphenols [ 1201.
The other components of beer are proteins, the large part of which forms
complexes with polyphenols. The solubility of these complexes depends
upon their molecular weight and the temperature of the system. Kringstad
and Damm claim that this complex has to be in an oxidized state to form a
haze, or alternatively the polyphenols have to go through an oxidative
polymerization to become reactive with proteins and form a haze [121].
Since the haze comes into being by the reaction of proteins with oxidized
polyphenols, colloidal stability of the beer may be enhanced by removing
either one or both of the ingredients of this reaction.
Various methods had been suggested to achieve this goal, which can be
grouped into four categories [122]:
a) Preventing the oxidation either by running the whole
process under anaerobic conditions, or by the addition of
reducing agents or antioxidants (such as ascorbic acid,
sodium tetrathionate, etc.).
b) Accelerated haze formation by the addition of haze forming
agents, such as tannic acid.
c) Elimination of the proteins, achieved, for instance, by the
addition of proteolytic enzymes (such as papain). This
approach has several disadvantages, such as the enzyme
remaining in the beer, the foam stability being reduced, and

the body of the beer suffering.
d) Removal of polyphenols and polyphenol-protein
complexes. The adsorbents used first were of the
polyamide type (such as nylon 66, nylon 1 1, or perlon),
which form complexes with polyphenols in a way similar
to proteins [123]. However, the other type of adsorbents
((poly(vinylpyrro1idinone) and "popcorn" polymerized
PVPP) were superior, with PVPP having the added
advantage of complete insolubility. The use of PVPP for
this application was first proposed by McFarlane and Bayle
[123].
Chillproofing beer by the adsorption method gives a product which is

more resistant to oxidation [ 1251. The treatment with PVPP removes the
tannin precursors which are originally inactive in haze formation but
which can condense during storage to active haze forming tannins when in
contact with oxygen [126].
As it has been described before, silica gel removes most of the proteins
from beer, while PVPP is effective in binding polyphenols. Recently it
was found [ 1271 that a mixture of silica gel and PVPP prepared in an 18%
H,SO, solution of the former, which after thorough washing, drying, and
milling, gave a product which was excellent in clarifying beer. The
filtration properties of PVPP could be further improved by irradiating the
crosslinked polymer with a 5 megarad dose of electron beam [128]. While
the amount of polymer to be used, as well as the contact time necessary for
the successful removal of the polyphenols, depends upon the nature and
quality of the brewing materials. In production, the use level is generally
8-20 ghectoliter for a 24 hour contact time.
4.3.2 Stabilization of Wine
In the winemaking process, the must is fermented, and then the wine is
aged. During this time the dissolved and dispersed proteinaceous and
polyphenolic substances create a disturbance in the equilibrium of the
wine colloid system and appear in the form of haze. The degree of haze
CROSPOVIDONE 121
formation depends upon the concentration of these substances. The haze

formation is the most serious with red table wines and dessert wines,
whose production methods usually result in the most complete extraction
of polyphenolic substances fiom the fibrous parts of the grape. On aging,
these hazeforming substances settle out and can be removed to a degree by
filtration or decantation. Nevertheless, the wine treated this way remains
sensitive to temperature changes and oxidation [1291.
When exposed to air, the flavonoid polyphenols of wine can react with
oxygen, either through non-enzymatic or enzymatic routes to form
quinoids and semiquinone radicals. These can further react to form brown
polymeric pigments, which are responsible for the so called "browning"
which harms the flavor, aroma, and color of wine. They come mostly
from the skins, seeds and stems of the grape. Since these parts of the h i t
differ greatly from grape to grape, different wines show different tendency
for browning [1301. For instance, white wines generally have about 50 mg
gallic acid equivalent (GAE) per liter, but this value can be as high as
2500 GAE per liter in the case of red wines, which are fermented with the
skins [131].
As a consequence of oxidation, the wine may develop a harsh taste and

strong discoloration. In order to diminish the degree of oxidation, sulfur
dioxide or ascorbic acid (or both) are sometimes added to the wine. These
materials, however, remain in the wine and affect its wholesomeness and
natural character. Other materials, such as charcoal, bentonite, or nylon
66, had also been tried with some success for wine stabilization.
However, it was shown that PVPP was superior in preventing haze
formation [ 132,1331. Silica gel preferentially adsorbs higher molecular
weight proteins and bentonite binds proteins of lower molecular weight,
but their adsorptive performance on phenolic compounds is poor and
nonspecific. PVPP, on the other hand, is highly specific with a strong and
selective adsorbing action on tannins, leucoanthocyanins,and
anthocyanins [ 132,1341. Some researchers reported good results with the
combination of PVPP and protein-adsorbing compounds. Drboglav and
co-workers used PVPP with bentonite and &Fe(CN),, and claimed to
achieve excellent stabilization by decreasing both phenolic and
proteinaceous materials [ 1351.
Lyubchenko and co-workers clarified must and wine with a combination

of PVPP, gelatin, and silicon dioxide [ 1361. Aivazov and co-workers
found that by adding PVPP to hot bottled wine, color stability of at least 3-
6 months could be achieved [ 1371.
McKissock found that PVPP not only prevented the formation of brown
pigments, but also that this adsorbent could remove any already formed
discoloration [ 1381. Vojnovic tested various adsorbents in forced
browning studies and determined that PVPP was the most effective in
preventing browning [ 1391. In the studies conducted by Farkas and
Ruzickova PVPP did not only control browning and remove tannins, but
also eliminated the unpleasant taste of oxidized wine [1401.
Since it was known that oxidation could seriously damage the quality and
the saleability of wine, the industry applied various technical innovations
to avoid it. While the use of colder fermentation and shielding the wine
from air reduced the browning problem, another phenomenon known as
“pinking” could not be avoided by these precautions. Pinking in white
wines is most probably caused by the conversion of flavenes to red
flavylium salts through reaction with oxygen. Flavenes can be formed by
the slow dehydration of leucoanthocyanidins, which turn to brown dyes
when oxidized. However, in the absence of oxygen, flavenes can
accumulate in the wine, and during the boiling stage exposure to oxygen
turn the flavenes to red flavylium salts [141]. While PVPP is not
necessarily the only adsorbent useful with wines, Simpson and co-workers
showed that PVPP was more effective in preventing pinking than either
activated carbon or casein [ 1423.
The adsorption of phenolic ingredients was found to be dependent upon

the concentration of PVPP in the system and of the amount of hydroxyl
groups on the phenols. The adsorption takes place by means of hydrogen
bonding, and is fast and it is usually complete in less than 10 minutes
(even at temperatures as low as 3°C. On the other hand, the reaction rate
is only very slightly dependent upon temperature. For instance, a
temperature increase from 3°C to 27°C brings about only a 10% increase
in the rate of adsorption [143].
There are several other important features that are associated with the use
of PVPP. The compound has no affinity towards the aroma substances
CROSPOVIDONE 123
present in the wine, so the taste of the wine does not suffer as a
consequence of the treatment. Use of PVPP produces a dense and
compact precipitate which increases the rate of filtration. Most
importantly, PVPP can be completely removed from the wine. The use of
PVPP is particularly advantageous in the treatment of sherry wines, where
sulfur dioxide could not be used due to its effect on the yeast at levels
above 3 ppm. PVPP, however, can be used in sufficiently high amounts to
assure the stability of color without affecting the yeast.
PVPP can be used to prevent browning or pinking reactions at use-levels

of 24-72 ghectoliter. It can be used also to brighten the color and improve
the flavor of red wines at a use-level of 6-12 gl hectoliter [ 1313.
Heavy metal cations (particularly of iron, copper, zinc, tin, and cadmium)
may cause a metallic taste, undesirable color changes, or haziness in the
wine. Formerly, these cations were removed by the addition of potassium
hexacyanoferrate or calcium physiate. A patent proposes the treatment of
wine with a popcorn polymer consisting of N-vinylpyrrolidinone, and/or
vinylimidazol with N,N'divinylethyleneureaas crosslinker [ 1441. It is
suggested that treatment with these popcorn polymers eliminates the
toxicological and operational drawbacks of the other methods used for the
removal of heavy metal contaminants. The diminution of heavy metals
was found to depend upon the dosage level and the contact time. The pH
also influences the amount of heavy metals retained by the system,
although maleic acid and lactic acid were found to have no effect of the
performance of crospovidone in this particular application [1451.
4.3.3 Stabilization of Other Beverages and Natural Liquids
Juices: Although the treatment of juices with PVPP is less extensive than
that discussed for beer and wine, its use has been studied by various
researchers. PVPP produced good color stability and citric acid recovery
with elderberry, black currant and raspberry juices, and it was found that
moist, swollen PVPP was more effective than the dry adsorbent [146].
Redelinghuys received a patent for removing bitter and astringent
proanthocyanidins from juices [1471. Lejeuene and Pourrat obtained
betanin with 98% purity by passing beet juice through Dowex 50-X2 (H+)
columns and through a PVPP column [ 1481. Hums and co-workers
showed that apple juice could be stabilized by PVPP, and that the
adsorbent could be regenerated by treatment with dilute sodium hydroxide

[149]. PVPP reduced the naringin content of grapefruit juice by 78% and
it slimonin content by 17.5% [ 1501. This adsorbent has been used
successfully also for the stabilization of cider { 1513.
&: 0.1% tea solution could be made polyphenol-free by treating it at

5°C with 0.4% PVPP for less than one day [1521.
Coffee: Coffee causing no dyspepsia was produced by mixing a standard

coffee extract with PVPP, which removed the phenolic and gastric juice-
stimulating components. After the addition of antiacid materials, the
solution was freeze-dried [153]. PVPP was found to be an effective
filtering material to remove mutagens from coffee extracts [ 1541.
VineEa: Fermented vinegar is made from wine, so its stability problems

are similar to those of wine. PVPP was found to be a suitable and
preferred adsorbent for the removal of haze developed during the
manufacturing process [15 51.
4.4 Miscellaneous Other Uses
Beside the stabilization of alcoholic beverages and other natural liquids,

crosslinked poly(vinylpyrro1idinone) finds numerous applications in the
food industry, in agricultural processes, and in a variety of other uses.
These uses utilize the complete insolubility, chemical inertness, and total
lack of toxicity characteristic of crospovidone.
4.4.1 Isolation and Stabilization of Enzymes
The presence of phenolic compounds in plant tissues complicates the

extraction of enzymes from them. In the intact plant the enzyme and the
tissue are separated from each other, however, when the material is broken
up, reactions begin between the enzymes and the phenolic compounds.
The products of these reactions are quinones and tannin-type compounds,
which further react with the enzyme proteins. The enzymes modified this
way are either inactive or substantially altered [156]. To separate the
enzymes at the required purity and activity, it is necessary to remove the
phenolic compounds fiom the system. PVPP is eminently suitable for this
CROSPOVIDONE 125
purpose because it has strong affinity towards vegetable tannins and can
complex them to water insoluble entities through hydrogen bonding 11571.
In the isolation of enzymes from apples Jones and co-workers used various
grades of PVP (PVPP among them) to prevent the inhibition of the activity
of mitochondrial preparations and certain soluble enzymes [ 1581. It was
found that the effect of PVP (and PVPP) on the activity of various
enzymes of the mitochondria (such as malic dehydrogenase, pyruvic
carboxylase, and phenolase) was concentration dependent, and 1% was the
optimal concentration for all enzymes. Walker and Hulme found that
when mitochondria were incubated in the presence of various amounts of
PVP (or PVPP), the degree of oxygen uptake inhibition increased up to a
PVP concentration of 1%. Thus, 1% PVP or PVPP brought maximal
inhibition of mitochondrial phenolase and also maximum activity of
mitochondrial dehydrogenase [159].
Gustavson found that the complexes formed by PVP and PVPP with
vegetable tannins could be split by high concentrations (5-8 M) of urea or
sodium dioctyl sulfosuccinate detergent, which partially reactivated the
PVPP-inhibited mitochondria [1601. Sanderson obtained 5-dehydro-
shikimate reductase by grinding frozen fresh shoots of the tea plant with
PVPP and acid washed sand in 0.1M sodium phosphate buffer. Maximum
activity was obtained with 0.6 g of PVPP per 1 g fresh weight of tissue
[161]. Loomis and Battaille studied the isolation of active enzymes in
peppermint and other monoterpene producing plants, using PVPP to
remove the inhibitory phenolic compounds [1621. The extraction
procedure yielded mevalonic kinase and phosphomevalonic kinase, and
was carried out by grinding the fresh tissue with PVPP in liquid nitrogen.
After the addition of buffer and sodium ascorbate, the thawed solution was
purified by Sephadex G-50 gel filtration.
Kaiser and Lewis reported that in the plants of nitrate-fed Heliunthus

Annuus, the nitrate reductase activity is restricted to the roots of the plant
[1631. Using an improved extraction technique consisting of a medium
containing 2%casein and 1.5 g PVPP per each gram of material, the
leaves showed a far greater nitrate reductase activity than did the roots. It
was found also that with the addition of casein and PVPP, the glutamine
synthetase activity increased in both leaves and roots. The extraction of
proteins from plant tissues is rendered difficult by the presence of phenols
and phenoloxidase enzymes. In the extraction of Citrus madurensis

Ramamurthy and Luddens used 2-mercaptoethanol as buffer [ 1641. This
compound was added to the fresh leaf and root tissue before
homogenization. When PVPP was added to the homogenized system, the
activity and stability of glutamate synthetase and glutamicoxalacetate
transaminase were noticeably increased.
Fernandez reported that by treating raw and cooked water extracts of

Phaesolus Vulgaris beans with PVPP, the inhibition due to trypsin
inhibitor and polyphenolics could be separated with a good degree of
reliability ( ~ 0 . 9 3 [)1651. Lastra reported the use of PVPP as a support
medium of immobilized enzymes [ 1663.
4.4.2 Use in Agriculture
PVPP was tested successfully in the determination of anthocyanidine

glucosides from grapes and anthocyanins from various other plants. The
crosslinked polymer was used to separate raspberry, rhubarb and
strawberry anthocyanins [ 1671. Wilson and coworkers used PVPP
columns for the preliminary isolation of chlorogenic acids from the whole
extracts of plant tissues. This step separated the acid from other
constituents that might interfere with its gas chromatographic analysis and
eliminated the losses of phenolic compounds which usually occur when
solvent extraction or similar procedures are used [ 1681. Generally, PVPP
column eluted with methanol was found to afford an effective and rapid
purification step of plant hormone containing extracts [ 1691.
Aoki and coworkers determined chlorogenic acid in apple flesh, and also
used PVPP column chromatography with ultraviolet absorption of the
methanol extract at 320 nm [ 1701. Budini and coworkers developed a
sensitive, high performance liquid chromatographic method for the
determination of indole-3-acetic acid in the berries of Nitis vinifera using
PVPP for the removal of interfering phenolic compounds. The use of
PVPP and determination on a silica-A column gave a 97% recovery with a
standard deviation of less than 5% [171]. Bjorsten and coworkers showed
that apple-allergens are probably proteins, and that they can be extracted in
an active form only if reactions with phenolic compounds present in the
apple are inhibited. This was accomplished by incorporating PVPP in the
extraction medium [ 1721.
CROSPOVIDONE 127
PVPP was used successfully for the purification of leaf nucleotides and
nucleosides prior to chromatography. Passing extracts of the leaves of
alfalfa, cotton, grape, and orange through a PVPP column removed 59-
91% of the substances that absorbed at 230 nm and 93 to 97 % of
substances that absorbed at 320 nm. Nucleotides and nucleosides passed
rapidly through the column, while the plant phenols were retained [ 1731.
Soluble phenols were determined in experimental tobacco materials that

were produced by practices that altered the concentration of these
compounds. The estimation of phenolic compounds was based on the
extent of hydrogen bonding of phenols to PVPP [1741. An efficient
fertilizer can be obtained by adding PVPP to animal excrement, and
fermenting the mixture under aerobic conditions. Due to the highly
absorptive nature of PVPP, the water content of the excrement can be
reduced by 25% [175].
The phytohormones in Scots pine (Pinus sylvestris) and spruce (Picea

abies) can usually be extracted only with great losses of the active
ingredients. Chromatography on PVPP-Sephadex LH 20 columns allows
remarkable purification, and the collection of indol-3-acetic acid in one
fraction. The other fraction contains phenylacetic acid and most of the
known CI9giberellins [176]. PVPP was used for the removal of phenolic
compounds in forage digestibility studies. The removal appeared to
increase the digestibility of cellulose and protein in alfalfa. After
treatment with Polyclar, the digestion process increased the amount of the
McDougall-buffer soluble phenolic compounds. These materials were
insoluble prior to indigestion [1771.
Clifford, while investigating the chlorogenic acids of green coffee beans

separated as many as 48 compounds belonging to this overall class. By
using thin layer chromatography and a PVPP-calcium sulfate adsorbent
system, ferrloylquinic acids (monohydroxyphenols),caffeoylquinic acids
(dihydroxyphenols), and dicaffeoylquinic acids (tetrahydroxy-phenols)
were separated. The butanone-methylphenylketone-5O%acetic acid (5:5:4
by volume) solvent system was found to give the best separation. The
system which used several structure-specificlocating agents was proposed
as a structure-diagnostic aid [1781.
Crospovidone has been proposed for use as a disintegrant in granules

containing water-soluble pesticides. The active compounds are rapidly
dissolved from such formulations in flooded rice paddies, and disperse
readily [ 1791.
A water-soluble antioxidant was prepared from black tea leaves by

extracting these with water, and then fractionating the extract using
chromatography. One part of tea leaves was treated with 29 parts of water
at 50 psi pressure and 175OC for 75 minutes. After that, the solution was
filtered and fractionated on a crospovidone column (other similar column
fillers, such as Sephadex G25M are also suitable). The antioxidant
fractions which are suitable for food applications are clear, and contain a
faint tea odor [1 S O ] .
4.4.3 Use in Analytical Chemistry
Crospovidone is a useful material for a variety of applications in the field

of analytical chemistry. The most important among these is that of
chromatography (both column and thin layer), in which the polymer was
found to be an efficient solid phase with good separation properties. The
origin of its separation ability lies in its strong hydrogen bonding
character.
PVPP was found suitable for separating nucleic acids components from
the salts used during their isolation. The salts pass through the PVPP
phase without interaction, while the bases and nucleosides are retained
[ 18 13. PVPP was found to be effective for the separation of aromatic
amino acids (phenylalanine, tyrosine, and tryptophan), with excellent
resolution of the three components [ 1821. PVPP was successfully used for
the separation of cytosine and thymine. The separation is carried out at pH
3.5 where thymine is hydrogen bonded to the pyrrolidone carbonyl, or at
pH 10.3 where the NH-group of thymine is ionized, disrupting the
hydrogen bonding [183]. Nucleotide derivatives can be separated on
PVPP columns and eluted with water in the order of nucleotides,
pyrimidines, and purines [1841.
PVPP was found to be effective as a stationary phase in the

chromatographic separation of aromatic hydrocarbons, where the order of
elution is determined by the number of condensed rings. The method is
CROSPOVIDONE 129
being used for the characterizationof products from coal liquefaction

processes [1851.
Good results were achieved by using PVPP as a coating for thin layer
chromatography plates. Bach Marles and coworkers obtained a patent for
this application, focusing on the analysis of corticosteroids, fat -soluble
dyes, bactericides, and flavonoids [1861. This method was used for the
chromatographic analysis of paramethasone and other hydroxy
cortcosteroids, as well as for the accurate determination of other epimers
[ 1871. A mixture of corticosteroids (hydrocortisoneacetate, prednisolone
acetate, fluorocortisone acetate, fluoroprednisoloneacetate, paramethasone
acetate, and betamethasone acetate) were separated by thin layer
chromatography, using PVPP as the stationary phase and 1:1 methanol-
chloroform as the mobile phase [1881.
A patent was issued for a novel determination of reduced Vitamin C using

2,6-dichlorophenolindophenolsupported on PVPP. The reagent mixture
may be coated on adhesive tape, or may be kneaded onto a filter, for
formation of the active compound [ 1891. A PVPP substrate, coated with
immobilized 8-quinolinol, was used for the preconcentration of cadmium,
zinc, lead, copper, iron, manganese, nickel, cobalt, and chromium salts
from seawater prior to their measurement by electrothermal atomic
absorption spectrometry [1901.
Nakamura and coworkers developed a method for the quantitative

determination of nitrates and nitrites in red wine. In this method, the red
pigments and sulfur dioxide that hindered the determination of nitrates and
nitrites were eliminated by passing the wine through a PVPP column and
adding triethanolamine [1911. The colorimetric determination of the
tartrate content of red wines by the vanadate method could not be
performed without first decolorizing the wine. PVPP was found to be the
most effective decolorizing medium for that purpose [1921.
Quarmby studied the separation of phenolic acids and flavonoids by TLC,

using a combination of PVPP and Celite as the adsorbent. He studied the
effect of particle size of the plate coating components, the composition of
the solvent system and the detecting agents. The Rf values of 18
flavonoids and related compounds were determined, each in three solvent
systems. It was found that by the use of formic acid as the solvent in one
direction offers the advantage of carrying much of the non-phenolic

components of the extracts with the solvent front where they cannot
interfere with the chromatographic separation [ 1931.
Olson and Samuelson studied the column chromatography of aromatic

compounds containing hydroxyl and carboxylic acid substituents, and
efficient separation was achieved with 0.001M hydrochloric acid as the
eluent. This eluent was found to be more suitable for quantitative
evaluation than water, since acid groups gave broader elution peaks in
water. With substances containing several proton donating groups,
improved results were obtained with 5M acetic acid as the diluent. The
volume distribution coefficients were calculated from peak elution
volumes. These were determined in a large number of experiments, with
single solutes and with mixtures of several solutes using phenol as a
marker. The degree of sorption was found to increase in a direct manner
with the number of proton-donating groups on the solutes. It was also
found that compounds which contain oxygen atoms in suitable distance
can act as intramolecular hydrogen acceptors. Phenyl substituted alcohols
eluted much earlier than phenols [ 1941.
Carpenter, Siggia and Carter investigated the separation and concentration

of phenolic materials on PVPP. Infrared studies conducted on the samples
after equilibration showed differences indicative of hydrogen bonding.
The hydroxyl stretching mode of phenol shifted from 3600 cm-' to 3200
cm cm-' when adsorbed on PVPP. The carbonyl band of PVPP also
shifted from 1690 cm cm-' to 1670 cm cm-'. The maximum uptake was
dependent upon the pH of the phenol solution, and in certain cases (e.g.
catechols) it was narrow, but in other cases (e.g.naphtols) it was quite
wide. As would be expected above pH 10, no adsorption took place due to
the formation of phenolate anions. The percent uptake increased with the
number of hydroxyl groups. The order of uptake for simple phenols was
phloroglucinol> resorcinol > phenol; and pyrogallol > catechol > phenol.
However, the correlation between uptake and -OH groups was not found
to be linear. The extent of aromaticity also influenced the uptake. For
instance, naphthol was removed more vigorously then phenols, owing to
the attraction between the carbonyl group and the aromatic n-electron
system. The pKa value did not seem to influence the uptake, as shown by
the observation that adsorbed phenolics could be separated fiom the
CROSPOVIDONE 131
column by treating with 4M urea. The recovery was nearly quantitative,

although a large excess of urea was required since the hydrogen bonding
between urea and pyrrolidone is weaker than that between the latter and
phenols [195].
PVPP was found to bind flavanoids rapidly, with the binding efficiency
being found to be directly proportional to the number of hydroxyl groups
attached to the flavanoid molecule. It was determined that for flavanoids
of different types, but containing identical number of hydroxyl groups,
flavones bind better than isoflavones, which then bind better than
flavonones and dihydroxyflavonones [ 1961.
Ianniello reported a sensitive method for the determination of

phloroglucinol in aqueous solution by square wave voltametric detection.
The carbon paste electrode used for the analysis was prepared by exposing
graphite and vinylpyrrolidinone to an argon RF plasma discharge. SEM
studies showed the existence of a uniform 1 pm film on the surface of the
graphite. The film was insoluble, indicating a crosslinked structure, and
the IR spectrum of the film was similar to that of PVPP [ 1971.
Very small amounts of the growth regulator diaminozide could be

determined through its adsorption onto PVPP. The adsorption capacity
was calculated to be 16.2 mg per gram of adsorbent. Since the surface area
of the diaminozide molecule (deduced using CPR space filling models)
was estimated to be 56.2 A. This suggests the existence of a multilayer
adsorption of diaminozide on crospovidone, which was bought about by
hydrogen bonding [1981.
X-ray photoelectron spectroscopy was used to identify the location of

griseofulvin loaded within a crospovidone matrix [210]. The obtained
spectral data was related to the drug loading mechanism, and the resulting
properties of the crospovidone-griseoflvin system. It was concluded that
the drug loading mechanism took place via intramolecular diffusion.
4.4.4 Use in Catalysis
A rhodium-based heterogeneous catalyst system was prepared by stirring

PVPP with rhodium acetate in a solvent medium consisting of methanol,
acetic acid, and methyl iodide. After agitating the mixture at 190°C for 2
hours, a solid product was obtained which contained 1.062% rhodium.
This catalyst system is usehl for carbonylation reactions, such as that of
methanol conversion to acetic acid. It is capable of withstanding the
carbonylation temperature of 150°C [1991.
5. Health and Safety
Since the most important applications of crospovidone are in the

pharmaceutical and medical fields, as well as in various areas of the food
industry, a lack of biological activity and a chemical inertness are
indispensable properties for the material. Furthermore, it must not interact
with other ingredients of the system any more than that of other widely
used excipients.
5.1 Acute Toxicity
Rats and mice were fed 5 g k g (body weight) crospovidone, and no

abnormal symptoms or deaths were observed after 24,48, or 72 hours
[201]. These studies may be summarized as follows.
5.2 Subacute Toxicity
A variety of subacute toxicity studies have been performed with

crospovidone, which serve to further demonstrate the safety profile of this
material [201].
Five groups of 40 rats each were dosed for 4 weeks at 0, 1,2.5,5, or 10%
crospovidone in their feed, followed by 2 weeks of no crospovidone in the
diet. At the 10% level, a slight bodyweight decrease was observed.
Clinical chemistry and hematology, necropsy results, and other
observations showed that there were no differences from the control
animals. Crospovidone was not deposited in the small intestine mucosa or
in the mesenteric lymph nodes.
CROSPOVIDONE 133
Three groups of 40 Wistar strain rats were fed diets for 90 days at 0,2, or
10% crospovidone. No compound-related effects were noted in behavior,
appearance, food consumption, feed efficacy, bodyweight gain, clinical
studies (hematology, clinical biochemistry and urinalysis), necropsy or
histopathology results.
Three groups of 6 beagle dogs were dosed with crospovidone at 1000,

2000, or 5050 mgkg bodyweight for 4 weeks. No compound-related
effects were noted in food and water consumption, hematology, clinical
chemistry, urinalysis and histopathology. No accumulation of
crospovidone was found in the liver, kidneys or mesenteric lymph nodes.
Three groups of 8 beagle dogs were dosed with crospovidone at 300, 1200,
or 4888 mgkg bodyweight for 26 weeks. No compound-related effects
were noted in behavior, food and water consumption, growth, hematology,
clinical chemistry or urinalysis results, electrocardiography,opthamology,
auditory tests, gross or histopathology. No crospovidone deposition or
storage was found in the liver, kidneys or mesenteric lymph nodes.
5.3 Teratogenicity
Three groups of 26 pregnant SPF rats were dosed with crospovidone by

gavage (at levels of 0,1000, or 3000 mgkg bodyweight) from day-6 to
day- 15, and sacrificed on day-20 of gestation. One group was left
untreated. No changes in fetal length were observed, and there were no
significant clinical symptoms, maternal deaths, and no compound-related
necropsy findings for the dams. Conception rates and live and dead
implantations were not affected. The fetuses displayed no compound-
related skeletal or visceral abnormalities.
Two additional groups of 12 pregnant SPF rats received the same dose
levels (1000 or 3000 mgkg) from day-15 to day-21 post-parturition.
Weight gain, mortality, delivery time and litter size of the test animals
were comparable to those of the control. No macroscopic compound-

related changes were observed in the mothers. Pups of treated animals
showed similar mortality, bodyweight and behavior as did the control.
Autopsies revealed no compound-related skeletal or visceral
abnormalities.
Studies on I4C-labeledcrospovidone in rats showed almost no absorption

or gastrointestinal accumulation of orally dosed crospovidone and no
biliary excretion [202].
Six Sprague-Dawley rats received an oral dose of 250 mg 14C-labeled

crospovidone by gastric intubation. The crospovidone used had been
treated to remove polymers having molecular weights less than 10,000
daltons. Approximately 0.128% of the ''C-crospovidone was excreted in
the urine, most being excreted within the first 24 hours. SO-99% of the I4C
was recovered in the feces within the 12-24 hour period. Less than 0.1%
of the I4C was recovered in the carcass, most being in the G.I. tract. There
was no evidence of preferential binding of the I4C to any organ or tissue.
5.5 Skin and Mucous Membrane Tolerance
The studies carried on rabbits indicated no evidence of irritation in either

short action ( 1 -1 5 minute) or long action (26 hour) exposures. The
polymer was applied on the skin of the rabbit as a 50% aqueous
suspension. Eye irritation studies indicated the existence of very slight
irritation that was not stronger than that obtained with using talc under the
same conditions.
5.6 Clinical Studies (Pharmacology)
Long-term clinical studies showed that crospovidone remains unabsorbed

in the gastrointestinal tract [201]. Therefore, no inherent pharmacological
action would be expected from crospovidone when this excipient is used
CROSPOVIDONE 135
in medical or pharmacological applications, particularly when it is used in

tablets at the common usage levels of 20-80 mg per tablet.
6. Compliance with Pharmacopoeias and Food

Regulations
As shown in Table 2, the various pharmacopoeiasand food regulations

differ only in some minor details. Even these slight variations should be
removed during the harmonization process currently taking place.
All pharmacopoeias and food regulations use the same methods to identify
crospovidone. These are the reaction with iodine, and the characteristic
infrared spectrum. Non-crosslinked, water soluble polyvinylpyrrolidinone
gives a deep red color when reacted with iodine. The reaction product of
the crosslinked polymer is colorless. The absorption peaks of the infrared
spectrum are equivalent to those of linear poly(vinylpyrro1idone).
6.1.1 Reaction with Iodine
1 gram of sample is suspended in 10 mL of distilled water, to which 0.1

mL of 0.1N iodine solution is added. The suspension is shaken for 30
seconds, and then 1 mL of starch solution is added. On shaking the
solution, no blue coloration should develop.
6.1.2 Infrared Spectrum
Crospovidone can be identified by its characterizationinfixed absorption

spectrum. The spectral peaks characteristic of the polymer containing the
pyrrolidone moiety, are the triplet at 1495 cm-I, 1463 cm-', and 1420 cm-l
The results of the test conducted on an unknown must be compared to
those obtained using an authentic Reference Standard. Both the test
specimen and standard must be pre-dried in similar fashion, and analyzed
on the same instrument using the same settings.
Table 2
ldentification, SDecifications and Other Characteristics
Eur. I FCC JECFA

IR
Spectrum , Spectrum,
ation
Iodination
6.0
11-12.8 11-12.8 11-12.8
~
5-8
1.5 I --- 1.5
0.00 1 0.001 0.001
~ 0.4
0.1 I 0.1 0.1
I
o.oos**
- I - 0.0003
I --
0.00025
0.002
0.04 1 -
5.0 1 --
* There are no specifications for this value, but current production

characteristically contains less than 0.005%.
* * When used in clarifying agent for beverages and vinegar.
CROSPOVIDONE 137
The test specimen and the Reference Standard are dried under vacuum at
105°C for 1 hour, and then compressed in a potassium bromide pellet. A
typical infrared spectrum is shown in Figure 4.
6.2 Cornpendial Testing
The compendia1testing discussed in this section is that required by most

or all of the pharmacopoeias on behalf of their respective regulatory
agencies.
6.2.1 Water Content
The water content of crospovidone is determined using Karl Fischer

titration. Measurement of this quantity is important in that water may
come from the reaction medium in which the polymer is prepared, or it
may be absorbed from the atmosphere. Since crospovidone is very
hygroscopic, the amount of water absorbed from the environment can be
considerable, and will depend upon the relative humidity to which the
polymer is exposed. Water is to be considered as a reactive entity which
can influence the chemical behavior of the polymer, and act as a plasticizer
capable of affecting its performance a in pharmaceutical application.
The method suitable for determination of the water content of

crospovidone is the same as previously published for povidone [reference
203, page 5861. However, since crospovidone is insoluble in the reaction
medium, the sample matrix must be stirred continually during the titration.
6.2.2 Nitrogen Content
The crosslinked polymer consists entirely of vinylpyrrolidinone monomer

units, which each contain 12.6% nitrogen. The nitrogen content thus
serves as a method to for the assay of crospovidone. The method suitable
for determination of the nitrogen content of crospovidone is the same as
previously described for povidone [reference203, page 5931.
1 I I I I I
9500 3000 a00 ZOO0 lSQ0 1000
Figure 4. Infrared spectnun of crospovidone (courtesy ISP Corporation).

CROSPOVIDONE 139
6.2.3 pH
The pH of a crospovidone solution is determined using potentiometric

means, following the procedure previously described {reference203, page
5901. Since crospovidone is insoluble, it must be kept in suspension
during the determination.
6.2.4 Non-Volatile, Water Soluble Content
The non-volatile, water soluble content of crospovidone consists of

residual uncrosslinked poly(vinylpyrrolidinone), any reaction by-products,
and adventitious contaminants. Because of their solubility and possible
lack of chemical or biological inertness, these may cause adverse effects
during the use of crospovidone. The pharmacopoeias and food regulations
limit the allowable concentration of these substances at 1.5%.
To perform the test, 25.0 g of crospovidone is transferred to a 400-mL

beaker, to which 200 mL of distilled water is added and the suspension
stirred for 1 hour. The suspension is transferred to a 250-mL volumetric
flask, rinsed with sufficient water for transfer, and diluted to volume with
water. The bulk of the solids are allowed to settle overnight, but the
settling time must not be allowed to exceed 24 hours. Use a volumetric
pipette to transfer 100 mL of the supernatant liquid to a pressure filtration
cell. Filter about 70 mL of the supernatant through a 0.45 pm membrane
filter, protected against clogging through the use of a 3 pm membrane
prefilter. Transfer exactly 50.0 mL of the filtered solution to a tared 100-
mL beaker, evaporate to dryness, and dry at 110°C for 3 hours. The non-
volatile content is calculated using:
% Water soluble =
Wt. of residue in beaker * 5 * 100
25
6.2.5 Heavy Metals
This test evaluates the content of common metallic impurities (silver,

arsenic, bismuth, cadmium, copper, mercury, and lead) that yield
insoluble, colored precipitates when reacted with hydrogen sulfide. The
heavy metal content is expressed in terms of lead equivalents, and should
140 EUGENE S. BARABAS AND CHRISTIANAH M.ADEYEYE
not exceed 10 ppm for pharmaceutical applications. The method for

determination of heavy metal content of crospovidone is the same as
previously described for povidone [reference 203, page 5971.
The residue on ignition test measures the total amount of non-volatile

substances, whether these are soluble or insoluble. Generally, the ash
content of a crospovidone sample reflects the amount of residual salt
contained in the material. The method suitable for determination of the
residue on ignition of crospovidone is the same as previously described for
povidone [reference 203, page 5961, must not exceed 0.4%.
6.2.7 Vinyl Pyrrolidone Content
The content of N-vinyl-2-pyrrolidone monomer is determined following

the methanol extraction of this species from a polymer sample. The
methanolic solution is analyzed by reversed-phase HPLC, using a base-
deactivated C8 column and quantified by UV detection at 235 nm. The
soluble PVP is washed from the column inlet using an automated
backflushing technique. The method is applicable for samples for which
the monomer levels ranging from 0.4 to 100 ppm.
The mobile phase is prepared by pipetting 200 mL of HPLC grade

methanol into a 1000-mL volumetric flask. The contents are diluted to
volume with HPLC grade water and mixed completely. The mobile phase
should be filter and degassed before use.
The working standard solution is prepared at a concentration of 1000 ppm

by accurately weighing 0.1 gram of N-vinyl-2-pyrrolidone in a 100-mL
volumetric flask, and diluting to volume with mobile phase. Additional
standard solutions having concentration values of 100 ppm, 10 ppm, and 1
ppm are prepared from the working standard solution by appropriate
dilutions.
The sample solution is prepared by accurately weighing 2.0 grams of

crospovidone in a 8-dram vial, and pipetting 20 mL of HPLC grade
methanol into the vial containing the sample. The vial contents are placed
on an automatic shaking apparatus, and shaken for one hour at a speed of
CROSPOVIDONE 141
130 cycles per minute. After this time period, the vial is removed from the
shaker, and the supernatant is filtered through an Xydex Autovial into an
HPLC autosampler vial.
The HPLC system should be equilibrated by passing mobile phase through

the column and detector for at least one hour before analyzing the standard
and samples. The analysis is conducted using an injection volume of 20
pL, a flow rate of 1 mL/minute, and a detection wavelength of 235 nm at
0.01 AUFS. A typical run time is 10 minutes for standards alone without
column backflushing, or 60 minutes for samples and column backflushing.
During typical work, a 10 ppm standard is injected in triplicate every 6-10
samples to monitor system performance.
The concentration of N-vinyl-2-pyrrolidonemonomer is calculated using:

ppm VP = 7.0 0 (Peak Area sample) (RspmsSactor)
(grams sample)
where the response factor is obtained using:
Response factor = (Standard conce-
(Peak Area standard)
6.3 Other Characteristics
There are certain requirements are not required by the various

pharmacopoeias, but are limited by certain food regulations. The methods
to determine compliance with these regulations are described in this
section.
6.3.1 Determination of Soluble Poly(Viny1 Pyrrolidone)
It is presumed that during the preparation of crospovidone all of the

polymer chains become part of an insoluble, crosslinked structure. It
remains possible that some chains remain uncrosslinked at the completion
of the reaction, and these polymer chains would be water or solvent
soluble. Although poly(vinylpyrro1idone)is non-toxic, in certain uses the
presence of soluble PVP is undesirable.
Soluble PVP can be determined using a turbidimetric procedure. The

sample is prepared by mixing 60 grams of sample and 600 mL of distilled
water in a 1000-mL beaker, and heating the stirred mixture at 90- 1OO°C
for 3 hours. At the end of this period, the mixture is cooled and
centrifuged at 2500 rpm for 1 hour. The liquid is decanted, and sufficient
water is added to restore the volume to 600 mL.
At this point 25 grams of Dicalite 2 15 and 25 grams Hyflow Super Cel are
thoroughly mixed, and 5 grams of the mixture is mixed with 100 mL of
the sample solution. A filter bed is prepared by moistening a piece of
filter paper (#42) in a 100 mm Buchner funnel, to which about 20 grams of
the Super Cel mixture slurried in water is added. The drained water is
discarded, and the sample slurry is then filtered. 50 mL of the filtrate is
transferred to a 150 mL beaker, 10 mL of 7 1% perchloric acid is added,
and the solution is stirred. The NTU Units on the turbidimeter must be
read within one minute after the addition of the perchloric acid.
A blank solution is prepared by slurrying 100 mL of distilled water with 4

grams of the Super Cel mixture, and the mixture is filtered as just
described. 0.25 mL of 0.1% K-30 solution is added to 50 mL of the
filtrate, followed by 10 mL of 71% perchloric acid. This solution is
stirred, and its turbidity measured.
The soluble poly(viny1 pyrrolidone) in the sample is estimated using the

relation:
A * 50
B
where A is the sample reading and B is the standard reading. This
analytical procedure simulates the conditions used in breweries during the
production of beer.
6.3.2 Arsenic
The World Health Organization has set a limit where the maximum
allowable limit of arsenic in crospovidone is 3 mg per kilogram. The
method for determination of arsenic content in crospovidone is the same as
previously described for povidone [reference 203, page 61 81.
CROSPOVIDONE 143
6.3.3 Zinc
All of the major pharmacopoeias have no limit on zinc content, but the
JECFA regulations state that zinc cannot exceed 2.5 mg per kilogram. If
necessary, zinc can be determined using a spectrophotometricmethod.
The most commonly used method follows the USP general method <25 1>,
that makes use of the dithizone reagent.
6.3.4 Free N,N'-divinylimidazolidone
In some methods used to prepare crospovidone,N,N'-divinyl-

imidazolidone is used as a crosslinking agent. In those cases, the
determination of any residual bifunctional monomer is required. This is
performed using gas chromatography.
An internal standard is prepared by dissolving 200 mg of N-(3-

methoxypropy1)-pyrrolidone or 300 mg of E-caprolactam (accurately
weighed), and dissolving in 100 mL of isopropanol. The sample solution
is prepared by accurately weighing 2-2.5 grams of polymer into a 50-mL
Erlenmeyer flask, to which is added 1.O mL of the internal standard
solution. 25 mL of acetone are added, the mixture is shaken for 4 hours,
whereupon the supernatant solution is analyzed.
A calibration solution is prepared by accurately weighing 25 mg of N,N'-

divinylimidazolidone into a 100-mL volumetric flask, and diluting to
volume with isopropanol. 2.0 mL of this solution is pipetted into a 50-mL
volumetric flask, and diluted to volume with acetone. 2 mL of this solution
is transferred to a 25-mL volumetric flask, 1 mL of the internal standard
solution is added, and the contents diluted to volume with acetone.
The gas chromatography system uses a 1-meter column of 2 mm inner

diameter, packed with 5% KOH and 15% polypropylene glycol 2025
supported on 45-60 mesh kieselguhr or 45-60 mesh Celite 545 NAW. An
oven temperature of 170°C and an injector temperature of 250°C are
appropriate. Detection is made by thermal conductivity, at a temperature
is 250°C. The carrier gas is helium (flow rate of 38 mL/min), and
injection volumes of 1 pL are used.
The calibration factor is calculated using:
f = W WSf
D As,
AD
where:
WD = amount of N,N'-divinylimidazolidonetaken (mg)
Wst = amount of internal standard taken (mg)
Ast = peak area of internal standard
AD = peak area for N,N-divinylimidazolidone
The N,N'-divinylimidazolidonecontent is calculated using:
where:
CD = concentration of N,N-divinylimidazolidone(mgkg)
f = calibration factor
A, = peak area for N,N'-divinylimidazolidone
W,, = amount of internal standard added (mg)
A,, = peak area of internal standard
W, = amount of specimen taken (g)
6.3.5 Peroxides
Ti(IV) is known to form a colored complex with H202 and unhindered

hy droperoxides, and these complexes may be spectrophotometrically
determined through their absorption at 405 nm. The peroxide content of
the sample is measured after 30 minute contact time with reagent, and
concentrations established by comparison to an external calibration curve
of hydrogen peroxide standard solutions. The method is suitable for the
determination of hydrogen peroxide and hydroperoxides in crosslinked
PVP at levels ranging from 10 ppm to 1000 ppm.
The Ti(1V) reagent is prepared by dissolving 2.0 g of TiOSO, in 1 liter of

distilled water, which contains 25 mL of concentrated H2S04.
CROSPOVIDONE 145
The stock standard solution is prepared by weighing 3.30 g of 30% H202

into a 100-mL volumetric flask, and diluting to volume with distilled
water. This solution should be standardized by KMn04 titration, following
the usual procedure. A 100 ppm working standard is prepared by then
pipetting 1 mL of the diluted H202 to 100 mL. H202calibration standards
are prepared by pipetting 0,2,4,10, and 20 mL of the working standard
into 1 00-mL volumetric flasks, adding 20 mL of the Ti(1V) reagent to each
flask, and diluting to volume with distilled water. These solutions
correspond to 0,0.2,0.4, 1.O, and 2.0 mg H202 standards, respectively.
The sample is prepared by accurately weighing 2.00 g of polymer into a

100-mL volumetric flask, adding 50 mL of distilled water, and suspending
by inversion. 20 mL of the Ti(1V) reagent is added, and the contents
diluted to volume with distilled water. The solution is allowed to stand for
30 minutes, with occasional mixing during the contact time. The
suspension is allowed to settle, and then a portion of the supernatant liquid
is filtered through a 0.45 pm teflon filter. This filtered solution is in the
sample solution.
After the spectrophotometer is zeroed at 405 nm, the absorbance of the

standard and sample solutions is recorded against the blank solution. A
calibration curve is constructed, where the absorbance measured at 405 nm
is plotted against the mg H202 in the standard solutions. A linear response
is obtained, which is fitted using least-squares regression. The peroxide
content in the sample is calculated using:
(A - b) 0 1000
ppm H~ozin sample =
m o w
where:
A = absorbance of the sample solution
b = y-intercept of the calibration curve
m = slope of the calibration curve
W = sample weight, g
Besides water and monomeric vinylpyrrolidone, crospovidone may

contain other volatile contaminants associated with the monomer, such as
aldehydes, diols, imides, and their reaction products. The method suitable
for determination of loss on drying is the same as previously described for

povidone [reference 203, page 6211.
6.3.7 Surface Area
The sorption of water by crospovidone is crucial to its disintegrant

function, and the rate of sorption is certainly affected by the surface area
of the solid. The most widely used procedure for the determination of the
surface area of a powdered material is the nitrogen adsorption method
developed by Brunauer, Emmett and Teller, the B.E.T. method [204]. The
method suitable for determination of crospovidone surface area is the same
as previously described for povidone [reference 203, page 6221.
6.3.8 Particle Sue Distribution
The particle size distribution of crospovidone is important to its use in

pharmaceutical technology, with different particle sizes being utilized to
suit the needs of the formulator. The particle size distribution may be
determined by analytical sieving, optical microscopy, light-scattering, or
zone-sensing methods, although sieving is the most commonly used
method.
The sieving procedure involves the mechanical distribution of samples

through sieves of successively smaller openings, followed by
determination of the percent of the sample retained on each sieve. Since
crospovidone is sensitive toward environmental factors, these must be
standardized if reproducible results are to be obtained. Temperatures
between 70 and 85"F, and relative humidities between 45 and 48% are
considered to be acceptable ambient conditions.
The method suitable for determination of particle size distribution of

crospovidone is the same as previously described for povidone [reference
203, page 6281.
6.3.9 Bulk Density
The bulk density of a powder is defined as the ratio of its mass to the
volume it occupies, and is normally expressed in units of g/cm3 (g/mL).
The bulk density of a powder differs from the absolute density inasmuch
CROSPOVIDONE 147
as the bulk density includes the contribution of the interstitial voids as well
as the volume actually occupied by the solid portion of the particles.
Knowledge of the bulk density is important primarily due to equipment
considerations during manufacturing, handling, and storage, but is also
important to considerations of product uniformity related to differences
among the densities of the formulation constituents.
The "poured" bulk density is determined by slowly pouring a weighed

amount of powder into a graduated cylinder, and determining the volume
occupied by the solid. The "tapped" bulk density is obtained by tapping
the cylinder in a predetermined fashion until the powder volume remains
constant [205]. The method suitable for determination of crospovidone
densities is the same as previously described for povidone [reference 203,
page 6271.
6.3.10 Flow Properties
Pharmaceutical powders may be characterized as being free flowing or

cohesive ( i e . , non-free flowing), or anything in between [206]. Good flow
properties assures efficient mixing and transport of the powder, which is
necessary for the production of uniform tablets. The flow rate is
influenced by factors which affect the free movement of the particles, such
as their intrinsic adhesive properties, the electrostatic forces which develop
as a consequence of the friction between moving particles, and any
adsorbed moisture. Knowledge of the nature of moisture contained within
a powder is essential, since it can dissipate electrostatic forces, while at the
same time forming bridges among the particles. Since crospovidone is
hygroscopic, the possibility always exists that it might change its flow
properties during industrial manipulations.
For these reasons, careful control of the powder conditions is required in

order to obtain reliable flow measurements. The method suitable for
determination of the powder flowability of crospovidone is the same as
previously described for povidone [reference 203, page 6351.
6.4 Microbial Limit Tests
Owing to its use as a pharmaceutical excipient, crospovidone, the material

must be free of both gram-positive and gram-negative bacteria, as well as
yeasts and molds [207]. Owing to its cross-linked structure and

insolubility, the polymer can foster bacterial growth when wet or exposed
to high humidity conditions.
To ascertain the validity of the corresponding tests, it first must be shown

that crospovidone is free of microorganisms which can inhibit or prevent
the growth of test cultures. Crospovidone should not contain more than
100 aerobic bacteria or yeasts or fungi per gram of polymer, and must be
completely free of Escherichia coli, Pseumonas aeruginosa,
Staphylococcus aureus (no bacteria per gram of crospovidone), and
SaZrnonella (no bacteria per 10 grams of crospovidone). The method
suitable for determination of microbial content of crospovidone is the
same as previously described for povidone [reference 203, page 6481.
7. Interactions of Crospovidone with Drug Substances
Crospovidone is generally a fairly inert excipient, and the literature does

not contain a large number of papers describing interactions between the
excipient and various drugs.
Crospovidone has been reported to be incompatible with some

antineoplastic agents (mitonafide and amondide) when stored at
conditions of 45°C and 45% or 72 YOrelative humidity. The interaction
was attributed to alterations in the water uptake capacity of the polymer
[25],and was detected by differential scanning calorimetry. Study of a
possible incompatibility of famotidine and excipients such as povidone,
crospovidone, and talc also revealed interaction of the drug with
crospovidone [30].
Interaction between crospovidone and some non-steroidal

antiinflammatory drugs has also been documented. Botha and Lotter
observed an interaction between naproxen and crospovidone in a tablet
formulation [ 3 11, and also found an interaction between ketoprofen and
crospovidone [32]. Depending on the particle size of ketorolac,
interactions may occur that decrease the dissolution rate, enhanced by the
presence of magnesium stearate [33]. The interaction observed with some
drugs (such as amonafide) is due to moisture uptake [24].
CROSPOVIDONE 149
An interaction between famotidine and excipients (including

crospovidone) studied using HPLC. The existence of an interaction
between the active and crospovidone was found to affect the
chromatography of famotidine [30].
The effect of crospovidone and other excipients, such as hydroxy-

ethylcelluloseand f.3-cyclodextrinon the stability of mefloquine tablets
was studied using thin layer chromatography. The TLC system consisted
of precoated silica gel, and the developing solvents were toluene, ethanol,
ammonium hydroxide, and isopropanol. No degradation of mefloquine
was observed which could be attributed to crospovidone, indicating the
inertness of this excipient [208].
Differential scanning calorimetry (DSC) was used to study the possible

interaction between naproxen and crospovidone [3 11. A 1:1 physical
mixture of these compounds revealed the existence of a second, but
smaller, endotherm, having an onset temperature of 153°C. The presence
of an additional thermal feature not associated with either pure component
was taken to indicate the existence of an interaction between the two [311.
DSC has also been used to study the possible interaction between
crospovidone and ketoprofen or ibuprofen, and it was concluded that there
indeed was an interaction between the excipient and these drug
compounds [32]. Idrayanto et al. also used DSC to study the interaction
between famotidine and crospovidone as part of the HPLC studies [30].
In contrast, Botha and Lotter were unable to detect an interaction between

crospovidone and oxyprenolol hydrochloride [209].
150 EUGENE S . BARABAS AND CHRISTIANAH M.ADEYEYE
8. References
1. J. Kondrakow, J. Pract. Chem. 64,109 (1901).
2. H. Staudinger and E. Huseman, Ber. dt. Chem. Ges. f&16 18

(1935).
3. J.W. Breitenbach, Poly. J. 6, 119 (1974).
4. H.F. Kauffmann and J.W. Breitenbach, Angew. Makromol. Chem.,

45,167 (1975).
5. E.S. Barabas, “N-Vinylamide Polymers”, in the Encyclopedia of

Polymer Science and Technology ,Volume 17, Second edition,
John Wiley and Sons, New York, p. 212,1989.
6. V. Buhler, Kollidon, BASF Aktiengesellschaft, Feinchemie, D-

6700, Ludwigshafen, p. 129 (1992).
7. J.W. Breitenbach and H. Salek, Mubomol. Chem. 1Q8,225 (1960).
8. J.W. Breitenbach, A. Dreisinger and E. Tomschik, Monutsh. Chem.

%, 807 (1963).
9. J.W. Breitenbach and H. Burgmann, Monatsh. Chem. eZ,1399

(1966).
10. J.W. Breitenbach and H. Axmann, in Polymerization Kinetics

and Technology, Advances in Chemistry Series. No. 128. Chapter
7. N.A.J. Platzer, ed., American Chemical Society, Washington
D.C. 1973. p.110.
11. J.W. Breitenbach and H. Dworak, J. Poly. Sci. 4,1328 (1966).
12. L.M. Welch et al., Ind. Eng. Chem. 39,826 (1947).
13. A.N. Pravednikov and. S.S. Medvedev, Dokl. Akad. Nauk SSSR,
N,461 (1955).
CROSPOVIDONE 151
14. J.W. Breitenbach and H.F. Kaufhann, Makromol. Chem., 125,

2597 (1974).
15. F. Grosser, U.S. Patent 2,938,017.
16. E.V. Hort, F. Grosser and A. Schwarz, U.S. Patent 3,294,765.
17. F. Haaf, A. Senner and F. Straub, Poly. Journal, U , 143 (1985).
18. S. Tseng and P. Wolf, Private Communication, ISP Co. (1993).
19. A. Senner, F. Straub and C.H. Tschang, in Proceedings of the

International Symposium on Povidone. G.A. Digenis and J.
Ansell, eds. University of Kentucky, College of Pharmacy.
Lexington, KY, 1983, p. 20.
20. J.S. Shi and S.Y. Tseng, US. Patent 5,286,826 (Feb. 15, 1994).
21. W. Denzinger et al., U.S. Patent 4,451,582 (May 29, 1984).
22. E. Hof€inann and K. Herrle, U.S. Patent 3,759,880 (Sept. 18,

1973).
23. M. S. Gordon and Z. T Chowhan, J Pharm. Sci., 26,907 (1987).
24. A. I. Torres and M. A. Camacho, Eur. J Pharm. and Biopharm.

&,41 (1994).
25. S . Malamataris and A. Dimitriou, J. Pharm. Pharmacol., 42, 158

(1990):
26. L.S. Wan and K. P. Prasad, Pharm. Znd., fl, 105 (1989).
27. L.S. Wan and K. P. Prasad, Znt. J. Pharm 3 , 1 4 7 (1989).
28. E. Grafet al, Pharm. Znd.,B, 291 (1986).
29. M. S. Gordon et al, Znt. J. Pharm. E, 119 (1993).

30. G. Idrayanto et al., Drug Dev. Ind. Pharm., a,91 1 (1994).

31. S. A. Botha and A. P. Lotter, Drug. Dev. Ind. Pharm., 16,673
(1 990).
32. S. A. Botha and A. P. Lotter, Drug. Dev. Ind. Pharm., fi,415

(1989).
33. Z. T. Chowhan and L. H. Chi, Pharm. Technol., g(Apr.), 28

(1985).
34. J. R. Johnson et al., J. Pharm. Sci.,84,469 (1991).
35. S.S. Kornblum and S.B. Stoopak, J. Pharm. Sci. a,43 (1973).
36. Anon. Polyplasdone XL, GAF Bulletin, 2302-099R 1 (1982).
37. R. Huttenrauch, Pharmazie, 26,645 (1971).
38. R. Huttenrauch and J. Jacob, Pharmazie, 3 , 2 9 3 , 5 6 3 (1971).
39. R. Huttenrauch and U. Schmeis, Pharmazie, 25,699 (1970).
40. P.H. List and V.A. M u m , Pharm. Ind., 41,459 (1979).
41. D. Gissinger and A. Stamm, Pharm. Ind. 42,189 (1980).
42. H. Nogami et al., Chem. Pharm. Bull. U , 1450 (1969).
43. E.M. Rudnic et al., Drug. Develop. Ind. Pharm. 8, 87 (1982).
44. A.H. Bronnsack, in Proceedings of the International Symposium

on Povidone, G.A. Digenis and J. Ansell, eds. Univ. of Kentucky,
College of Pharmacy, Lexington, KY, 1983 p.471
45. K.A. Khan and C.T. Rhodes, Can. J. Pharm. Sci.,8,l (1973).
46. A.H. Bronnsack, Pharm. Ind. 40, 1255 (1982).

CROSPOVIDONE 153
47. K.A. Khan and D.I. Rooke, J. Pharmacol., 28,633 (1976).
48. E.M. Rudnic et al., Drug. Develop. Ind. Pharm. 6,291 (1980).
49. P.H. List and V.A. Muazzam, Pharm. Ind. & I,(1979).
1075
50. L.S.C. Wan and P.P. Prasad, Acta Pharm. Technol. 36,20 (1990).
51. J. Ringart and A.M. Guyot-Hermann, Drug. Develop. Ind. Pharm.

-
14,2321 (1988).
52. D. Hennig and E. Schubert, Pharmazie, 42, 725 (1987).
53. M.S. Gordon and Z.T. Chowhan, J. Pharm. Sci., 26,907 (1987).
54. H.V. Van Kamp et al., Pharm. Acta Helv., a,22 (1986).
55. M. Jovanovic et al., Pharmazie, 43, 727 (1988).
56. H.V. Van Kamp, G.K. Bolhuis, and C.F. Lerk, Pharm. Weekbl.,5,
165 (1983).
57. D.S. Phadke and N.R. Anderson, Drug. Develop. Ind. Pharm. 16,
983 (1990).
58. L.S.C. Wan and Y.L. Choong, Pharm. Acta Helv., a,50 (1986).
59. L. Casahoursat et al., Bull. Tech. Gattefosse Rep. My33 (1987).
60. M. Petrbczki, Gydgyszeriszet, 23,88,(1979).
61. J. Esteve et al., Congr. NUC.Farm. Ind.[ActaJ, I,373 (1978).
62. A.G. Sandoz, Austrian Patent 348,132 (Jan. 25, 1979).
63. M. Miseta et al., Acta Pharm. Hung., Z , 4 5 (1987).
64. D. S. Desai et al., Int. J Pharm.llQ, 257-265 (1994).

65. A. Sakr et al., Pharm. Ind., 55,953-957 (1993).
66. M. S. Gordon et al., J. Pharm. Sci,82,220 (1993).
67. T. Baykara et al., Drug Dev. Ind. Pharm., iz,2359 (1991).
68. M.D. Jovanovic et al., Farm-Obz, 58,221 (1989).
69. J. E. Botzolakis and L.L. Augsburger, Drug Dev. Ind. Pharm., 14,
29 (1988).
70. H. V. van Kamp et al., Pharm- Weekbl-Sci., 5, 165 (1 983).
71. L. S.C. Wan and W. F. Lai, STP Pharma Sciences, 4,336 (1994).
72. C. H. Liu et al., Drug Dev. Ind. Pharm., a,1911 (1994).

73. E. Graf et al., Pharm. Znd., 43, 576 (1981).
74. E. Graf et al., Pharm. Ind., 44,3 17 (1982).
75. J. Gillard, Acta Pharm. Tech., 2 , 4 (1980).
76. H. Imainuni, N. Nambu, and T. Nagai, Chem. Pharm. BUZZ.,2,

25 10 (1983).
77. J. Akbuga, Pharm. Ind., 53,857, 1991.
78. M.L. Lovrecich, Eur Pat. Appl. EP 364,944 (Apr. 25, 1990).
79. A. Martini, C. Torricelli and R. De Ponti, Int. J. Pharm. 25,141

(1991).
80. A. Martini et al., Int. J. Pharm., Z5, 141 (1991).
81. S.P. Wyas, C.P. Jain, A. Kaushik and V.K. Dixit, Pharmazie, 42,
463 (1992).
CROSFQVIDONE 155
82. U. Conte, A. La Manna and P. Giunchedi, Eur. Pat Appl. EP

468,392 (Jan. 29, 1992).
83. U. Conte and P. Giunchedi, PCT Int. Appl. WO 93 00,889 (Jul. 8,

1991).
84. A. Palmieri, T. Danson and W. Grob, Drug. Dev. Ind. Pharm. 9,

421 (1983).
85. V.N. Harsici, Biomaterials (En@ Transl.), 2,3 (1981).
86. H. Maeda et al., Japan Tokyo Koho JP 7807493 (March 18,1978).
87. M. Azuma, T. Nezama, M. Yoshida, and H. Fujimori, Eur. Pat.

Appl. EP 454,089 (Oct. 30, 1991).
88. U. Conte, A. La Manna and L. Maggi, PCT Int. Appl. WO 94

06,416 (Mar. 3 1, 1994).
89. M. It0 et al., Jpn. Kokai Tokyo Koho, JP 05,306,225 (Nov. 19,
1933).
90. V. Buehler, Ger. Offen DE 3,0056,635 (Aug. 27, 1981).
91. R.J. Prankerd and S.M. Ahmed, J Pharm. Pharmacol. 44,259

(1992).
92. T.P. Dietz, R.A. Asmus and R.Uy, PCT Int. Appl. WO 93 09,713
(May 27,1993).
93. F. Kutas et al., Hungary Teljes HU 17691 (Feb. 28, 1980).
94. A. Queuille and R. Larde, U S . Patent 3,725,541 (Apr. 3, 1973).
95. K. Laemmerhirt et al., Ger. Offen. DE 3,622,896 (Jan. 21, 1988).
96. M.C. Etienne et al., J Pharm. Sci. 8Q, 1130, (1991).

97. M.B. Kjaer, P. Jacobsen, D.M.F. Edwards, and M. Strolin-

Benedetti, Anti-Cancer Drugs, 4,437 (1993).
98. F. Grosser and J. Susko, U.S. Patent 3,136,755 (June 9, 1864).
99. E. Shanbrom, PCT Int. Appl. WO 04 00,23 1 (Jan.6, 1994).
100. J.P. Gendre, Gaz. Med. FR (France), M, 122 (1983).
101. J. Guerre and M. Neuman, Med. Chir. Dig.(France), 8,679 (1979).
102. R.E. Jeanpierre et al., Med. Chir. Dig. (France), 6/7,499 (1977).
103. M. Luttinger and C.W. Cooper, J. Biomed. Mat. Res., 1,67 (1967).
104. R.A. Markle, R.D. Falb, and R.I.Leininger, Rubber Plastics Age,
45,880 (1 964).
105. S.I. Gordon, Ger. Offen DE 2,751,453 (Jan. 6, 1978).
106. S.A. Katz et al., Chest, D , 4 t h Suppl., 45 S. (1981).
107. Chemische Fabrik Stockhausen and Cie, Ger. Offen 2,706,135

(Aug. 17, 1978).
108. D.S. Henning, G.R. Brown and E. St-Pierre, J. Znt. Artif Organs
5,373 (1982).
109. B. Streuff et al., Ger. Offen DE 3,304,867 (Aug. 16, 1984).
110. Yamanouchi Pharmaceutical Co., Jpn. Kokai Tokyo Koho, JP 84

76017 (Apr. 28,1984).
111. S. Mulligan and R.T. Sparks, S. African Patent IA 87 00738 (Sept.

30, 1987).
112. K. Yoneto, Jpn Kokai Tokyo Koho JP 051,155,760 (Jan. 22,

1993).
CROSPOVIDONE 157
113. Federal Drug and Cosmetic Act, Title 21, Chapter 1, Part 121,
1110 (1963).
114. Anon., “Polyclar AT Stabilizer in Brewing”, GAF Bulletin 2302-

082 (1974).
115. W.D. McFarlane, Tech. Quart. of the Master Brewers Assoc. of

Amer. 5,87 (19683.
116. J.W. Gramshaw, Tech. Quart. of the Master Brewers Assoc. of

Amer. 6,239 (1969).
117. R. Armitage et al., J. Chem. SOC.,1842 (1961).
118. M.J. Betts et al., Chem. Comm., 1110 (1967).
119. H.S. Mason, Advances in Enzymolog;~,

_14, 105 (1 965).
120. J.W. Gramshaw, Tech. Quart. of the Master Brewers Assoc. of

Amer. 2,167 (1970).
121. H. Kringstad and E. Damm, Proc. Eur, Brew. Conj Stockholm,

128 (1965).
122. M. Dadic and J.G. Lavale, J. Am, SOC.Brew. Chem., 41, 141
(1983).
123. W.D. McFarlane, U.S. Patent 2,088,550 (Sept. 7, 1954).
124. W.D. McFarlane and P.D. Bayle, Proc. Eur. Brew. Conv. Vienna,
278 (1967).
125. W.D. McFarlane, K.D. Thompson, and R.H. Garrat, J Am. SOC.
Brew. Chem., 102 (1963).
126. K. Steiner and H.R. Stocker, Proc. Eur. Brew. Conv. Madrid, 407
(1968)
127. C.W. Chi and M.E. Winyal, Ger. Offen DE 3,302,258 (Aug. 11,
1983).
128. J. Daugherty et al., U.S. Patent 4,792,400 (Aug. 20, 1988).
129. V. L. Singleton and P. Esau, Phenolic Substances in Grape and

Wine and Their Significance, Academic Press, New York, 1969.
130. V.L. Singleton Tech. Quart. of the Master Brewers Assoc. of

Amer., 4,245 (1967).
131. Anon., “Polyclar AT Stabilizer in Winemaking”, GAF Bulletin,

9653-003 (1975).
132. F. Drawert and G.P. Lauhus, Weinbau, Nos. 27-28 (1971).
133. N.M. Pavlenko et al., Puti Povysh. Stabil nosti Vin i

Vinomaterialov, 73 (1 982).
134. A. Caputi Jr. and R.G. Peterson, Amer. J. Enology and Viticulture,
&, 9, (1965).
135. E.S. Drboglav et al., Pishch. Prom. St. Ser. 1,11 (1980).
136. 1.1. Lyubchenko et al., Vinodel. Vinograd. SSSR, 12 (1982).
137. A.A. Aivazov et al., Lozar. Vinar. 32,13 (1983).
138. A. McKissock, Food Technol. New Zealand, 12 (1966).
139. V. Vojnovic, Kem. Ind. (Belgrad), 32, LL5 (1983).
140. J. Farkas and A. Ruzickova, Vinohrad (Bratislava), 23, 134 (1985).
141. V. L. Singleton, in Advances in Food Research, Supplement 111.

143 (1972).
142. R.F. Simpson, G.C. Miller, and G.L. Orr, Food Technol. Aust. 34,
44 (1982).
CROSPOVIDONE 159
143. A. Caputi Jr., T. Brown and M. Veda, presented at the Ann.

Meeting of the Am. SOC.Enol., San Francisco, June 26, 1969.
144. J. Detering, A. Sanner, and B. Fussnegger, U.S. Patent 5,094,867

(March 19,1992).
145. B. Fussnegger, R. Maurer. and J. Detering, Wein-Wiss, 47, 8

(1992).
146. M. Ugrinovits, Mitt. Geb. Lebensmittelunters. Hyg. ll,484 (1971).
147. H.J.P. Redelinghuys, So. African Patent 8 2,505,023 (June 23,

1983).
148. B. Lejeunne and S . Pourrat, Ann. Pharm. Fr., 22,463 (1981).
149. N. Hums et al., Fluess. Obst., 41,238 (1980).
150. G. Bohun and J.F. Drilleau, BIOS 2, 13 (1976).
151. M.O. Nisperos and G.L. Robertson, Phillip. Agric., 65,275 (1982).
152. S. Kozai and T. Hoshina, Chagyo, Kenkyu Hokoku,No.52,36

(1980).
153. C. Fuehrer and R. Franke. Ger. Offen DE 2,72 1,117 (Nov. 23,
(1978).
154. Advance Kaihatsu Kenkyusho K.K., Jpn. Kokai 8 420,128 (Feb. 1,

1984).
155. Anon., “Polyclar L, a new Ciarifling and Stabilizing Agent for

Vinegar”, GAF Bulletin AP-143 (1963).
156. D.S. Bendall et al., in Enzyme Chemistry of Phenolic

Compounds, J.G. Pridham, ed, Pergamon, London, 1963, p. 7.
157. K.H. Gustavson, Leder, M,27 (1963).

158. J.D. Jones, A.C. Hulme and L.S.C. Wooltorton, Phytochemistry 3,

210 (1964).
159. J.R.L. Walker and A.C. Hulme, Phytochemistry, 4,677 (1 965).
160. K.H. Gustavson, Svensk. Kern. Tidskr., 66,359 (1959).
161. G.W. Sanderson, Biochem. J. 28,248 (1966).
162. W.D. Loomis and J. Battaille, Phytochemistry, 5, 423 (1966).
163. J.J. Kaiser and O.A.M. Lewis, Plant Soil, 31, 127 (1984).
164. S. Ramamurthy and P . Ludders, Gartenbauwissenschafr, 46, 162

(1981).
165. R.J. Fernandez, J. Agric. Food Chem., 32,734 (1982).
166. M.C. Lastra and M. Ortega, Ann. R. Acad. Farm., 2,123 (1985).
167. R.A. Wrolstad and B.J. Struthers, J. Chromatogr. B , 4 0 (1971).
168. J.L. Wilson, W.J. Dunlap, and D.S.H. Wender, J. Chromatogr. 35,
329 (1968).
169. D.M.A. Mousdale and M. Knee, J. Chromatogr. U , 3 9 (1979).
170. S. Aoki, Y. Yohagi and T. Tamura, Nippon Shokugin Kogyo

Gakkaishi, 2, 1 10 ( 1 984).
171. R. Budin et al., Microchem. J., 22, 365 (1982).
172. F. Bjorksten et al., Allergy, Z, 671 (1980).
173. G. Guinn, Anal Biochem., 54,276 (1973).
174. R.A. Anderson, T.C. Tso and J.F. Chaplin, J. Agric. Food Chem.,
X,89 (1979).
CROSPOVIDONE 161
175. Naga-0 Soda K.K., Jpn. Kokai 58 208,189 (Dec. 3,1983).
176. S. Sandberg, A. Dunberg and P.C. Oden, Physiol. Plant, =,2 19,
(1987).
177. J.H.G. Jung and G.C. Fahey, J. Agric. Food Chem. 29,817, (1987).
178. N.M. Clifford, J. Chromatogr.,94,261 (1974).
179. T. Shimada and S. Shiboyama, Jpn Kokai Tokyo Koho JP

06,186,228 (Jul. 8, 1994).
180. R.S.C. So, Eur. Pat. Appl. EP 547,370 (June 23, 1993).
181. T.M. Dougherty and A.I. Schepartz, J. Chromatogr. 411,299

(1969).
182. T.M. Dougherty and A.I. Schepartz,J. Chromatogr.42,415

(1969).
183. T.M. Dougherty and A.I. Schepartz,J. Chromatogr., a,397

(1969).
184. J.L. Lerner, T.M. Dougherty, and A.I. Schepartz, J. Chromatogr.

3 , 4 5 3 (1969).
185. G. Goldstein, J. Chromatogr. (Amsterdam), D,61 (1976).
186. J. Bach Marles et al., Spanish Patent 378,422 (June 16, 1972).
187. J. Moreno Dalmau, J. Pla-Delfina and A. Ojeda del Pozo, J.

Chromatogr. 28,165 (1973).
188. J.M. Marcos, J. Moreno Dalmau and J. Pla-Delfina, Cienc. Ind

Farm., 9,92 (1977).
189. Norin Suisansho Shokabin Sogo Kenkyu Socho, Jpn. Kokai 57

154056 (March 20,1984).
190. S.N. Willie, R.E. Sturgeon and S.S. Berman, Anal. Chim. Acta
m , 5 9 (1983).
191. M. Nakamura, M. Hashimoto, and H. Kondo, Fakuokashi Eisei

Shikenshoho, 2,60, (198 1).
192. L.R. Mattick and C. Rice, Amer. J. Enol. Vitic., 22,297 (1981).
193. C. Quarmby, J. Chromatogr. 2 , 5 2 (1968).
194. L. Oisson and 0. Samuelson, Chromatographia, U, 135 (1977).
195. A. Carpenter, S. Siggia and S. Carter, Anal. Chem., 48,222 (1976).
196. L.W. Doner, G. Becard and P.L. Irwin, J. Agric. Food Chem., U ,
753 (1993).
197. R. M. Ianniello, Anal. Lett., 25, 125 (1992).
198. R.M. Ianniello, Anal. Chim Acta, Le3,8 1 (1987).
199. M.D. Scales, J.R. Warner and P.G. Torrance, U.S. Pat. 5,281,359
(Jan. 25, 1994).
200. F. Fusiak, K.H. C h u g , and T. Bohrer, US Pat. 5,268,117 (Dec. 7,

1993).
201. Anon., “PVPP Summary of Toxicity Information to JEFCA”. ISP,

(1994).
202. Anon., Kollidon CL. BASF Technical Data Sheet, M.5650 (1977).
203. C.M. Adeyeye and E.S. Barabas, “Polyvinylpyrrolidone”, in

Analytical Profiles of Drug Substances and Excipients. Vol. 22,
H. G. Brittain, ed., Academic Press, Inc., 1993.
204. S. Brunauer, P.H. Emmett and E. Teller, J. Am. Chem. SOC.612,309

(1938).
CROSPOVIDONE 163
205. Drug Standards, 2 , 2 2 (1 952).
206. R.L. Carr Jr., Chem. Eng., Feb. Issue, 69 (1965).
207. United State Pharmacopoeia, XXII Revision. Microbial Limit

Tests. U.S. Pharmacopoeial Convention, Inc., p. 1619.
208. S. Kitamura et al., Int. J. Pharm., 141,127 (1994).
209. S. A. Botha and A. P. Lotter, Drug Dev. Ind. Pharm., fi, 1843
(1 989).
210. F. Carli and F. Garbassi, J. Pharm. Sci.,14,963 (1985).
Acknowledgements
The authors wish to express their gratitude to Messrs. John F. Tancredi

and Louis Blecher, and Drs. Robert M. Ianniello and Chi-San Wu for their
strong support and valuable advice. Special thanks are due to Dr. Edward
G. Malawer for the sound observations and helpful suggestions with
regard to methods and content. Mr. Ira Naznitsky’s kind help in collecting
the literature pertinent to the subject is gratefully acknowledged.
The authors with to express their gratitude to Dr. Harry G. Brittain, whose
careful examination and thoughtful suggestions built important
improvements into the manuscript. Finally, the authors would like to
thank Suzanne Currie and Susan Thomas for their contribution in typing
the manuscript.
FLUVOXAMINE MALEATE
Nagwa H. Fodal, Mahasen A. Radwan2, and Omar A. A1 Deeb3
(1) Department of Pharmaceutics

(2) Department of Clinical Pharmacy
(3) Departments of Pharmaceutical Chemistry
College of Pharmacy
King Saud University
(1,2) University Center for Women Students
P.O. Box 22452, Riyadh 11459

Kingdom of Saudi Arabia
ANALYTICAL PROFILES OF DRUG SUBSTANCES 165 Copyright Q 1996 by Academic F’ress. Inc.
AND EXCIPIENTS-VOLUME 24 All rights of reproduction in any form x ~ ~ e d .
166 NAGWA H. FODA ET AL.
CONTENTS
1. Description
1.1 Nomenclature
1.1.1 Chemical Name
1.2 Formulae
1.2.1 Empirical
1.2.2 CAS Registry Number
1.2.3 Structural
1.4 Appearance
1.5 Uses and Applications
3.1 Polymorphism
3.2 Powder X-ray Diffraction Pattern
3.3.1 Melting Behavior
3.4 Hygroscopicity
3.5 Solubility Characteristics
3.6 Partition Coefficient
3.7 Ionization Constant
3.8 Spectroscopy
3.8.2 Vibrational Spectroscopy
3.8.3 Nuclear Magnetic Resonance Spectra
3.8.3.1 IH-NMR Spectrum
3.8.3.2 13C-NMR Spectrum
4.2 Spectrophotometric Methods of Analysis
4.3 Polarographic Methods of Determination
4.4.1 Thin-Layer Chromatography
FLUVOXAMINE MALEATE 167

5. Stability
5.1 Incompatibilities with Functional Groups
6. Drug Metabolism and Pharmacokinetics

6.1 Absorption
6.2 Distribution
6.3 Elimination
6.3.1 Metabolism
6.3.2 Excretion
7. Adverse Reactions
8. Drug Interactions
9. References
168 NAGWA H.FODA ET AL.
1. Description
1.1 Nomenclature
1.1.1 Chemical Name [I]
5-Methoxy-l-[4-(trifluoromethyl)-phenyl]-l-pentanone-
0-(2-aminoethyl) oxime maleate
5-methoxy-4'-(trifluoromethyl)valerophenone
(E)-O-(2-aminoethyl) oxime maleate
Dumirox, Faverin, Fevarin, Floxyfral, Maveral.
1.2 Formulae
1.2.1 Empirical: C19H25F3N206
1.2.2 CAS Registry Number: 6 1718-82-9
1.2.3 Structure
CHCOOH
II
CHCOO-
1.3 Molecular Weight: 434.4
1.4 Appearance
When obtained from acetonitrile, fluvoxamine is a white crystalline

powder [ 11.
Fluvoxamine is an antidepressant. It selectively inhibits the re-uptake of

serotonine, but has relatively little effect on noradrenaline re-uptake.
Fluvoxamine is reported to cause fewer antimuscarinic side-effects than do
the tricylic antidepressants, but its mode of action is not fully understood.
In the treatment of depression, fluvoxamine is given orally as the maleate
salt, in doses of 100 to 200 mg daily. In some patients, doses of 300 mg
daily may be required. It is recommended that daily doses exceeding 100
mg should be given in divided doses, in the manner of many other
antidepressants. Fluvoxamine has been investigated in obsessive-
compulsive disorders with reports of some benefit [2].
2. Method of Preparation
Welle and Claassen [3] postulated three methods in their patent for the
synthesis of fluvoxamine maleate. The current manufacturing method is
that described in Scheme 1. A mixture of 5-methoxy-4’-trifluoromethyl
valerophenone and 2-minooxyethylamine dihydrochloride is refluxed in
absolute ethanol, using pyridine as an acid scavenger. The maleate salt is
prepared by the addition of an equimolar quantity of maleic acid to a
solution of fluvoxamine in absolute ethanol, which is then heated until a
clear solution is obtained. The ethanol is removed, and the residue
recrystallized from acetonitrile.
3.1 Polymorphism
Differential scanning calorimetry (DSC), infrared spectroscopy, and x-ray

powder diffraction have been used to reveal the presence or absence of
polymorphism [4]. Fluvoxamine maleate was recrystallized from PEG
4000,6000,20000, Tween 60, Tween 80, and PVP (as 1% and 10% for
each). The x-ray powder patterns showed that all obtained materials were
crystalline. The DSC and powder diffraction results indicated the
existence of polymorphism but, the IR spectra were less clear.
n
CHCOOH
Scheme 1 The Method of Fluvoxamine Maleate Synthesis

FUJVOXAMINE MALEATE 171
3.2 X-ray Powder Diffraction Pattern
The x-ray powder diffraction pattern of crystalline fluvoxamine maleate

was obtained using a Phillips x-ray diffraction spectrogoniometer,
equipped with PW 1730/10 generator. Radiation was provided by a
copper target (Cu anode 2000 W, 1 = 1.5418A) high intensity x-ray tube
operated to 40 kV and 35 mA. The monochromator was a curved single
crystal unit (PW 1752/00), and the divergence and receiving slits were 1
and 0.1, respectively. The scanning speed of the goniometer (PW
1050/81) used was 0.02 degrees 2 9 per second. The goniometer was
aligned before all work using a silicone sample.
Figure 1 illustrates the obtained powder pattern of fluvoxamine maleate.

The crystallographic parameters data are listed in Table 1.
The melting range of fluvoxamine maleate is 120 - 121.5OC [11.
The differential scanning calorimetry thermogram for fluvoxamine

maleate is shown in Figure 2, and was obtained using a differential
scanning calorimeter Perkin-Elmer model DSC-4 calibrated with indium
(99.999%). The thermogram was obtained using a heating rate of
1OoC/minute, under a nitrogen atmosphere.
A single sharp endotherm was observed, having an onset of 119.4OC and a

maximum at 121.6 OC. This endotherm is assigned to the melting of the
compound, and is characterizedby a AH value of 26.42 cal/gm.
3.4 Hygroscopicity
At 90% relative humidity, the water uptake at ambient temperature is less

than 1%.
60
Figure 1 X - Ray Powder Diffraction Pattern of Fluvoxamine Maleate

Table 1
X-ray Powder Diffraction Data of Fluvoxamine Maleate
0.00. 1 1 1 I 1 1 1 I 1 1 1
40.00 x)o.oo 160.00 228.00 2 80.00 340.00

TEMPRATURE ( C >
Figure 2 Thermal Curve of Fluvoxamine Maleate
EUVOXAMINE MALEATE 175
Fluvoxamine maleate is sparingly soluble in water, fieely soluble in

ethanol and chloroform, and practically insoluble in ether.
3.6 Partition Coefficient
The following partition coefficients have been obtained:

0.4 n-heptanelwater
0.1 dichloromethane/water (PH 1)
50 dichloromethane/water (PH 12)
Fluvoxamine maleate contains only a single ionizable group, whose pKa

has been found to be 8.7.
3.8 Spectroscopy
The UV spectrum of the protonated form of fluvoxamine maleate was

obtained in 0.1 N HCl, being scanned from 190-400 nm using Pye Unicam
model PU 8850 UVNIS specptrophotometer. In this form, the compound
exhibits the characteristic UV spectrum shown in Figure 3a, characterized
by an absorption maxima at 237.9 nm. The E-1% value was found to be
301.
To obtain the UV spectrum of the deprotonated form of fluvoxamine

maleate, a sample was dissolved in IN alcoholic KOH. The W spectrum
obtained in this medium is found in Figure 3b, where the absorption
maximum was observed to shift to 256.2 nm. The intensity of the
absorption was also found to decrease, being characterized by a E- 1%
value of 230.
Wavelength (nm)
Figure 3a. Ultraviolet absorption spectrum of fluvoxamine maleate in

0.1 N HC1.
I
-
200 240 280 3 20
Wavelength (nm)
Figure 3b. Ultraviolet absorption spectrum of fluvoxamine maleate in

1 N alcoholic KOH.
3.8.2 Vibrational Spectroscopy [5]
The infrared absorption spectrum of fluvoxamine maleate, obtained in a

KBr disc, is shown in Figure 4. The spectrum was recorded on a Perkin-
Elmer model 1760X infrared spectrometer. The major observed bands and
their assignments are found in Table 2.
3.8.3 Nuclear Magnetic Resonance Spectra [5]
Both the proton nuclear magnetic resonance ( H-NMR) and carbon

nuclear magnetic resonance (13C-NMR) spectra of fluvoxamine have been
obtained in CDC13, using TMS as the internal standard. The assignments
for both the 1H-NMR and 13C-NMR spectra make use of the following
numbering scheme:
1 2 3 4 I 8
C CH2CH&H&Ht OCHI
IJ 7
NOCH26H2NH2
3.8.3.1 lH-NMR Spectrum [5]
The ambient temperature 300 M H z 1H-NMR spectrum of fluvoxamine

was obtained on a Varian XL-300 NMR spectrometer. The spectrum itself
is shown in Figure 5, and a summary of the chemical shifts and spectral
assignments is provided in Table 3.
3.8.3.2 13C-NMR Spectrum [5]
The ambient temperature 75 MHz 13C-NMR spectrum of fluvoxamine

was obtained on a Varian XL-300 NMR spectrometer. The spectrum is
shown in Figure 6, and a summary of the chemical shifts and spectral
assignments is provided in Table 4. The 13C signal multiplicities were
determined by APT and DEPT experiments.
Evidence in support of the 1H and 13C spectral assignments was obtained

from two-dimensional correlation spectroscopy (COSY) experiments. The
-
c.
Y
d
a
t- 20-
WAVENUMBER (CM-')
Figure 4 Infrared Spectrum of Fluvoxamine Maleate, KBr Disc

I80 NAGWA H.FODA ET AL.
Table 2
Vibrational Spectral Assignments of Fluvoxamine Maleate
Band Frequency Assignment

(cm-1)
3200 - 2740 (3road) Asymmetrical and symmetrical NH3 stretching;
OH stretching
~~ ~ _____
3015 - 3005 (Weak) Two bands for aryl and olefinic C-H stretching
2950 - 2880 (Weak) Two bands for asymmetrical and symmetrical
aliphatic C-H stretching
1700 (Weak) C=O stretching (-OH)
~~ _____
1640 (Weak) C=N stretching

1605 - 1450 (Broad) Asymmetrical aliphatic C-H bending, aromatic and
olefinic C=C stretching; COO- stretching, and NH3
bending
1365 (Medium) Symmetrical aliphatic C-H bending,
1335 (Strong) C-0 stretching of carboxylic acid
_____
I
C-F stretching
Two bands for ether C-0 stretching
950 - 650 Multiple bands due to the 1,4-disubstituted benzene

~
ring
::
n
n
1
Figure 5 1 H-NMRSpectrum of Fluvoxamine in CDC13 from TMS.

1xz NAGWA H. FODA ET AL.
Table 3
Chemical Shifts and Spectral Assignments of the

1H NMR Spectrum of Fluvoxamine Maleate
Chemical Shift Multiplicity Assignment Number of

(6; ppm) Protons
1.60-1.62 m C4-H, C3-H 4

2.70 s (broad) -NH2 2
2.79 t C2-H 2
>
3.04 t C8-H 2
~~
3.30 S C6-H 3
3.37 t C5-H 2
4.24 t C7-H 2
7.61 d C3',5'/H 2
7.72 d C2',6'/H 2
s: singlet d: doublet
t: triplet m: multiplet
Figure 6 C-NMR Spectrum of Fluvoxamine in CDCI3 from TMS
NAGWA H. FODA ET AL.
Table 4
Chemical Shifts and Spectral Assignments of the

1% NMR Spectrum of Fluvoxamine Maleate
I Chemical Shift 1 Assignment

(6; PPm)
t
I
- ~~ ~
23.09 I c-3
26.08 c-2
29.49 c-4
41.47
58.53
72.15
-~ ~~
76.18 c-7
125.28 - 125.43 CF3, C-4', C-3'/5'
126.49 - 126.50 C-2'/6'
139.07 c-1'
157.48 c-1
FLUVOXAMINEMALEATE 185
proton-proton (HH-COSY) and carbon - proton (CH-COSY or HETCOR)

results are shown in Figures 7 and 8, respectively.
The 70 eV electron impact (EI) mass spectrum of fluvoxamine maleate is

shown in Figure 9, and was recorded on a Finnigan Mat model 5 100 series
GCMS system. The molecular ion peak of fluvoxamine at a m/z of 3 18 is
not detectable in the EI spectrum.
The chemical ionization (CI) mass spectrum of fluvoxamine maleate is

shown in Figure 10, and was obtained on a VG ZAB-HF mass
spectrometer equipped with an Ion Tech Saddle-Field FAB gun and a
standard VG source. The gun was operated at 8 KeV. Fluvoxamine
maleate yielded a very prominent [M+H]+ ion (base peak) at m/z 3 19 in
the CI spectrum, whereas the molecular ion at m/z 3 18 was not observed.
Assignment of the possible structures of the various fragments and their
relative intensities under EI conditions are listed in Table 5.
The calculated elemental content of fluvoxamine [ 11 and fluvoxamine

maleate are given in Table 6.
Table 6. Summary of Elemental Analyses
I I
Element Fluvoxamine Fluvoxamine maleate
(w/wYo) (w/w Yo)
EK%
56.59 52.53
5.80
17.90 13.12
6.49
I 0 I 10.05 22.10
0 0
e o a
I 1 I I I I
7 6 5 4 3 2
F2 (PPM)
Figure 7 HH - COSY Spectrum of Fluvoxamine.

I
c-I
I
I I
I
I
1
d0 140 140 140 1dO 1iO id0 9b 8b 7b 6b 5b i0 30
F2 (PPM)
Figure 8 -
The Assigned CN COSY (HETCOR) Specuum of Fluvoxaminc.
2
100 .o
226
50.C
I
M
03 9
'1
216
55
Mi2 50
95
160
125
150
E 260
-P
250
Figure 9 Electron Impact Mass Spectrum of Fluvoxamine Maleate

3 3
Figure 10 Chemical Ionization Mass Spectrum of Fluvoxamine Maleate

Table 5 . Fragmentation assignments of the mass spectrum

of Fluvoxamine Maleate
Table 5 Fragmentation assignments of the mass spectrum

of Fluvoxamine Maleate (Continued)
Fragment Assignment
-
F3C--(3-fH
dH=CH CH,CH20 CH3

d z (%) Fragment Assignment


d z (%) Fragment Assignment
+
71 (87) CH,=CH CH, O=CH,
+
69( 3 ) CF3
c
55(22) CHz CH,Ch=CH,
45(77) C H 2 = 6 CH3
31(3) CH,= 6 H
1 94 NAGWA H.FODA ET AL.
4.2 Spectrophotometric Methods of Analysis
‘Two spectrophotometric assay methods have been described for the

quantitation of fluvoxamine in tablets. The first method is based on
formation of a charge-transfer complex with chloranil [6]. This
complexation with chloranil substantially enhanced the weak UV
absorption of fluvoxamine, and permitted measurement of the absorbance
at 347 nm. The second method utilized the same complexation technique
[7], but the kinetics of the complexation reaction were studied at 30 and
40OC. The reaction rate constant, and the relevant thermodynamic
activation parameters, have been calculated.
4.3 Polarographic Methods of Determination
A polarographic method has been developed for the determination of

fluvoxamine maleate in tablets [8]. The drug was extracted from tablets
with Britton-Robinson buffer solution (pH 7.4). The polarographic cell
was equipped with a Ag-AgC1 reference electrode, a Pt auxiliary electrode,
and a dropping mercury working electrode. The supporting electrolyte
was 0.1 M acetate buffer (pH 3.7), and the drop time was 0.6 seconds,
with a scanning rate of -500 mV/min.
Tuncel et al. (9) studied the polarographic behavior and the optimum
polarographic conditions for the determination of fluvoxamine using direct
current, differential pulse, and superimposed amplitude pulse [9]. The
technique was applied to a pharmaceutical dosage form.
A method for the determination of fluvoxamine in human plasma has been

described [lo]. Fluvoxamine was extracted by heptane/2-propanol, and
after evaporation of the organic layer, the residue was dissolved in 0.1 M
NaHC03 and 0.1 % 4-chloro-7-nitrobenzofurazan.The derivative was
separated by TLC using silica gel 60, with CHC13/ethyl acetate (25: 1) as
the mobile phase. The derivative was fluorimetrically detected at 546 nm,
using an excitation wavelength of 434 nm.
Fluvoxamine was determined in human plasma by electron capture gas

chromatography, with its structural analogue (clovoxamine) being used as
the internal standard. The method required derivatization and multiple
extraction steps [l 11. Dawling et al. presented a gas chromatographic
method, which uses a single extraction step and nitrogen-phosphorus
detection. The method is characterized by lower sensitivity [12].
A number of HPLC systems have been reported, which are suitable for the
identification and separation of fluvoxamine maleate have been reported
[13-201. A summary of these systems is given in Tables 7 and 8.
Keating et al. developed an assay procedure which involved the HPLC

measurement (with electrochemical detection) of plasma serotonin levels
as an indirect index of the in vivo activity of fluvoxamine [131. Haertter et
al. described an automated, column switching, HPLC determination of
fluvoxamine in plasma [141. The method involved solid phase extraction,
which was characterized by a recovery of 97-100%. Linearity in analyte
response was achieved fiom 25 to 1000 p g h , and the detection limit was
10 ng/mL. Foglia et al. determined fluvoxamine in human plasma by
HPLC with ultra-violet detection at 2 15 nm, and obtained a detection limit
of 25-400 ng/mL at a with relative standard deviation of 3.2-9.7% [16].
A sensitive one-step extraction procedure for the column liquid-

chromatographic determination of fluvoxamine in human and rat plasma
was described by Van-der-Meersch-Mougeot V et al. [ 151. In this method,
more than 99% of the mobile phase was organic in nature, and superior
sensitivity was obtained (0.5 ng/mL could be detected). The within-day
relative standard deviation was 1.8 to 5.6%, and the between-days RSD
was 14.4%. In addition, there was no interference from 28 other drugs.
Table 7. HPLC Assays for the Analysis of Fluvoxamine Using UV or Electrochemical Detections
- -
Flow G
No Rate Stationary Phase Mobile Phase Detector Remarks Nc
-1 1 Hypersil ODS Column 0.2 ml/L of sodium heptane
-
(15 x 0.46 cm) sulfmate in 0.1 M - phosphate it + 0.6V vs a serotonin as an index of in viva 13
buffer (PH 5.5): methanol (17:3) 4glAgCl ref.. activity of fluvoxamine
:lectrode
2 1.5 Hypersil MOS C8 pH 6.8- MeOH-ACN- 10 mM 214 nm Column switching and on line 14
(10 p)& (25 x 0.46 k phosphate buffer injection of plasma samples
an) Nucleosil 100 CN (188:378:235)
(5 pn)
3 1 (15x 0.39 an)Resolve MeOH-ACN-THF-H20- 254 nm Detection limit 0.5 ng/d. No 15
spherical silica (5 pan) diethylamine (9859:100:20:20:1) interference from other 28 CO-
administered drugs
16
4 1 (12 x 0.46 an) 16 mM-KH2P04 buffer @H 215 nm Determination in human
2.5)-ACN (16:9) plasma
Nucleosil C8 (5 pn)
7
5 1.5 CH3CN-0.01 M CH3COONa 240 nm Evaluation in tablets
phndapak-Cl8, (30 X
0.39 cm) (10 pn) buffer adjustcd to pH 3.5 with
CH3COOH
17
6 0.8 (10 x 0.8 cm) c18 0.05 M ammonium acetate 240 nm Determination in tablets
and ACN 40% V/V
-- Bondapack (1 0 pm)
Table 8. HPLC Assays for the Analysis of Fluvoxamine Involving Derivatization.
Flow
-
Rate
mllmin
Stationary Phase Mobile Phase Detector
I Remarks Ref
-
No
(25 x 0.46 cm) Supelcosil LC-18- ACN:1O m M F1uorimetric Derivatization by 40 mM- 18

DB (5 pm)- guard cartridge (1.5 x k phosphate pH7.2
0.32 cm) New Guard RP-8 (7 pm: (17:3) NaHC03, dansyl chloride &
acetone
0.5 (15 x 0.31 cm) Lichrosorb RP-18 ACN-2.5 mM imidizole Chemiluminescent Naphthalene-and anthracene 19 I
(5 pm) or Lichrosorb Si 60 (5 pm buffer of pH 7 (3: 1) at 418 nm 2,34aldehyde as pre-column

c
(25 x 0.31 cm) labeling reagnet for primary
s amines.
1.5 (12.5 x 0.46 cm)Hypersil ODS 45 to 65% ACN gradieni FluOrimetric Detection limit 1.5 ng/ml in 20
at 3OoC elution over 10 min. plasma (0.1M Na2C03 &
10 ml dansyl chloride solutn
10 mg/ml in acetone).
1 (25 x 0.46 cm)Zorbax SIL (7 MeOH-propan2-01 F1uorimetric at Derivatization by 0.1M 10
pm) & a precolumn (5 cm x 4.6 nm
nun) Lichrosorb Si 100 (30 pm) (125:l) at 455 nm) NaHC03 & 4chloro-7-
nitiobenzofurazan solution -
I
human plasma. -
I98 NAGWA H. FODA ET AL.
Naphthalene and anthra~ene-2~3-dialdehyde have been used as pre-column

labeling reagents (reacting at the primary amine site) in the reversed-phase
and normal-phase liquid chromatography with peroxyoxalate
chemiluminescence detection [ 191. This assay methods requires that
special care be employed during the derivatization reaction owing to the
nature of the steps involved and the precautions which need to be taken.
The method has not been applied to determinations in body fluids.
Schweitzer et al. described the fluorimetric determination of fluvoxamine

or clovoxamine in human plasma after thin-layer chromatographic or
normal phase HPLC separation [ 101. The method involved derivatization
after extraction, and was characterized by a detection limit of 1-200 ng/mL
and a recovery of about 99%.
A HPLC method for the determination of fluvoxamine in human plasma

was developed by Pommery and Lhermitte [20]. Linearity in the method
was attained in the range of 10-400 ng/mL, with a detection limit of 1.5
ng/mL. These workers found no interference from 20 other drugs. The
method recovery was 62 to 77%, with this low recovery possibly being
due to the many steps involved in extraction and derivatization.
Hagga. et al. used HPLC and charge-transfer complexation methods to

evaluate fluvoxamine maleate in tablets [7]. They obtained linearity was
over the range of 3- 100 pg/mL, with a detection limit of 0.15 pg/mL. The
recovery of fluvoxamine from tablets exceeded 99%.
Foda has described an HPLC assay for the determination of fluvoxamine

maleate in tablets [ 171. Linearity was obtained was over the range from 0.5
to 12 pg/mL, and the fluvoxamine recovery from tablets was 100%.
5. Stability
5.1 Incompatibilities with Functional Groups
The physicochemical compatibility between fluvoxamine maleate and a

number of tablet and capsule excipients were investigated by differential
scanning calorimetry [21]. Fluvoxamine maleate was found to be fully
compatible with starch, polyvinylpyrollidone, triglycerides, microfine
cellulose, and microcrystalline cellulose. The drug was found not to be

compatible with stearic acid, magnesium stearate, lactose, or sodium
carboxymethyl-cellulose.
6. Drug Metabolism and Pharmacokinetics
Fluvoxamine is a second generation antidepressant, characterized by a

potent, selective, and inhibitory activity on neuronal serotonin (5-
hyroxytryptamine, 5HT) re-uptake. In addition to its lack of effects on
other monoamine re-uptake mechanisms, fluvoxamine has little or no
effect on the neuronal function of other monoamines, and has a low
affinity for receptors of a variety of neurotransmitters. The drug is
structurally unrelated to the tricyclic group of antidepressants.
For the treatment of depression or obsessive-compulsivedisorder, the

usual daily dose is 50 to 300 mg administered once daily, or in divided
doses [22]. The administration is usually started at 50 mg/day, and is
slowly increased up to 300 mg/day in an effort to improve the patient
tolerance to the nausea and vomiting which are associated with the
initiation of fluvoxamine therapy [23-251.
6.1 Absorption
Fluvoxamine is almost completely absorbed (but relatively slowly) from

the gastrointestinal tract [26]. After the administration of single oral doses
of rapidly dissolving formulations, the peak plasma concentration (Cm,)
is usually observed within 2-8 hours. For enteric coated tablets (the
commercially available dosage form), the Cm, may be observed at 4 to
12 hours after administration. Food did not interfere with fluvoxamine
absorption after administration of rapidly dissolving formulations [27].
This finding is not necessarily applicable to the enteric-coated dosage
form, since the absorption of drugs fiom this dosage form is delayed by
food intake [28].
6.2 Distribution
Approximately 77% of fluvoxamine is bound to human plasma proteins at

plasma concentrations up to 1 mg/L [22,29]. This implies that plasma
protein binding interactions of any clinical relevance are unlikely to occur

with this drug 1301. After intravenous administration of radiolabelled
fluvoxamine to rats, the radioactivity distributes rapidly and is found in
higher concentration in most organs than found in plasma [22]. Since
fluvoxamine has not been intravenously administered to humans, its
apparent volume of distribution is estimated after oral administration as
ViF (where V = apparent volume of distribution, and F = bioavailability).
The ViF factor varies between 10 and 20 L/kg [29].
6.3 Elimination
The elimination of fluvoxamine after a single oral dose follows a

biexponential decline [26]. Following the administration of enteric-coated
fluvoxamine tablets, the mean elimination half-life (t1/2 p) was 16.9 hr in
healthy volunteers, and 23.2 hr in depressed patients [31]. The mean
elimination half-life was 22 hr after multiple dosing (50 mg twice daily for
28 days), which may be compared to 19 hr after single dosing (50 mg) in
the same subjects. There is no apparent accumulation of fluvoxamine in
plasma after multiple-dose therapy. The area under the plasma
concentration-time curves (AUC) was similar after single and multiple-
dose therapy. The AUC tended to be longer following multiple oral-dose
as compared to single-dose administration [32]. This implies non-linearity
in fluvoxamine disposition after higher single dose (exceeding 100 mgj or
multiple-dose therapy.
6.3.1 Metabolism
Following the administration of a single dose of radiolabelled

fluvoxamine, at least 11 metabolites are known to accumulate in human
urine. Nine of the metabolites have been identified by mass spectrometry,
and account for 85% of the total urinary radioactivity [33]. The metabolic
pathways of fluvoxamine in humans are depicted in Scheme 2.
The major (65%) fluvoxamine metabolite is produced by oxidative

demethylation of the aliphatic methoxy group. Lesser amounts of other
metabolites are produced by degradation at the primary amino group
(1 5%j, at both the methoxy and m i n e groups (20%), or by the removal of
the entire ethanol amhe group (10%). Two of the primary metabolites
(Compounds I and I1 of Scheme 2) do not possess any psychotropic
F , C O F-CH,-CHz-CH,-CH2-O-CH,
N
\
-
0 CH,-Mz-NH,
F,C~-CH,-CH~-CH~C-OH \c~;-c~-cH,-cH,-cH,-o-cH, F,C~$-CH;-CH~CH~~-OH

N N N
\
O-CHrM@ ‘0- CHfC-fH va ‘OH
V vi 0
Scheme 2 The Metabolic Pathways of Fluvoxamine Maleate

Table 9
Drugs Reported to Interact with Fluvoxamine
Drug Reference
of Interaction Effect of the
Interaction
Inhibition of hepatic Decrease C1. 44-45

(brom azepam, oxidation Increase t 1/2and
alprazolam) plasma
concentration
Inhibition of hepatic Increase plasma 46
metabolism concentration
Propranolol NJA Increase plasma 22
concentration
I Theophylline
~~~~~~
Competitive inhibition Decrease C1. 47
I----
of hepatic metabolism Increase plasma
concentration
Warfarin NJA Increase plasma 22
concentration
8. Drug Interactions
The recent reports of drug interaction observed during administration of

fluvoxamine to healthy volunteers and patients with CNS disorders are
summarized in Table 9. The incidences of toxicity were found to increase
as an outcome of these drug interactions. It should be mentioned that
fluvoxamine affects only the pharmacokinetics of benzodiazepines
eliminated by hepatic oxidation [44-451.
9. References
1. The Merck Index. 1lthed. Merck and Co. Inc., Rahway, New
Jersey, 1989, p. 659.
th
2. Martindale - The Extra Pharmacopoeia. 30 ed. J.E.F.
Reynolds, ed. The Pharmaceutical Press, London, 1994, p. 1450.
3. H.B.A. Welle V. and Claassen, US Patent 4,085,225, April 18,

1978.
4. O.M. Al-Gohary, N.H. Foda, and F. El Shafie, Identification of

polymorphs of fluvoxamine maleate, Die Pharmazie, 49,592-594
(1994).
5. O.A. A1 Deeb, N.H. Foda, and M.A. Radwan, unpublished results.
6. A.A. A1 Haider, M.E.M. Hagga, M.E. Alawady, and E.A. Gad-

Kariem, Spectrophotometricdetermination of fluvoxamine in
tablets based on charge-transfer complex with chloranil, Anal.
Lett., 26,887-901 (1993).
7. M.E.M. Hagga, A.A. A1 Haider, H.A. Al-Kahmees, M.E. Alawady,

and E.A. Gad-Kariem, Evaluation of fluvoxamine maleate in
tablets by high performance liquid chromatogrphic and charge-
transfer complexation methods, Saudi Pharm. Journal, 1,70-75
( 1 993).
8. K. Albert, Polarographic determination of fluvoxamine maleate in

tablets, Pz-win, 3, 59-61 (1990).
9. M.Tuncel, G . Altiokka, and Z. Atkosar, The polarographic

determination of fluvoxamine maleate, Anal, Lett., 27, 1 135-1 145
( 1 994).
10. C. Schweitzer, H. Span, and E. Mutschler, Fluorimetric

determination of fluvoxamine or clovoxamine in human plasma
after thin-layer chromatographic or high performance liquid-
chromatographic separation, J Chromatgr. Biomed. Appl. 55
(J Chromatogr.,382) 405-41 1 (1986).
11. H.E. Hurst, D.R. Jones, C.H. Jarboe, and H. DeBree,

Determination of clovoxamine concentration in human plasma by
electron capture gas chromatography, Clin. Chem. 27, 1210-12 12
( 1 98 1).
12. S . Dawling, N. Ward, and E.G. Essex, Rapid measurement of

basic drugs in blood applied to clinical forensic toxicology, Ann.
Clin. Biochem., 27,473-477 (1990).
13. J. Keating, L. Dratxu, M. Lader, and R.A. Sherwood,

Measurement of plasma serotonin by high-performance liquid
chromatography with electrochemical detection as an index of the
in vivo activity of fluvoxamine, J. Chromatogr. Biomed. Appl.,
126 (J Chromatogr.,615) 237-242 ( 1 993).
14. S. Haertter, H. Wetzel, and C. Hiemke, Automated determination

of fluvoxamine in plasma by column switching high performance
liquid chromatography, Clin. Chem., 38,2082-2086 (1992).
15. V. Van-der-Meersch-Mougeot and B. Diquet, Sensitive one-step

extraction procedure for column liquid-chromatogrphic
determination of fluvoxamine in human and rat plasma, J.
Chromatogr. Biomed. Appl., 105 (J Chromatogr.,567) 441 -449
(1991).
16. J.P. Foglia, L.A. Birder, and J.M. Perel, Determination of

fluvoxamine in human plasma by high-performance liquid
chromatographywith ultra-violet detection, J. Chrornatogr.
Biomed. Appl., 87,(J. Chromatgr., 495) 295-302(1989).
17. N.H. Foda, High performacne liquid chromatogrphic

determination of fluvoxamine maleate in tablets. J. Liquid.
Chromatgr., 18,1591-1601(1995).
18. R.H. Pullen and A.A. Fatmi, Determination of fluvoxamine in

human plasma by high performance liquid chromatographywith
fluorescence detection, J. Chromatgr. Biomed. Appl. 112,(J.
Chrornatogr.,574) 101-107(1992).
19. P.J.M. Kwakman, H. Koelewijn, J. Kool, U.A.T. Brikman, and

G.J. De-Jong, Naphthalene and anthracene-2,3-dialdehydeas pre-
column labelling reagents for primary mines using reversed and
normal-phase liquid chromatography with peroxyoxalate
chemiluminescencedetection, J. Chromatgr.,511, 155-166
(1 990).
20. J. Pommery and M. Lhermitte, High-performance liquid-

chromatographic determination of fluvoxamine in human plasma,
Biomed. Chrornatogr.,3, 177-179 (1 989).
21. N.H. Foda, Compatibility study between fluvoxamine maleate,

mebeverine hydrochloride and tablet excipients using differential
scanning calorimetry, Egypt J. Pharm. Sci., a,73-79(1992).
22. P. Benfield and A. Ward, Fluvoxamine: a review of its

pharmacodynamic and pharmacokineticproperties, and therapeutic
efficacy in depressive illness, Drugs, 32,313-334(1986).
23. L. Conti, L. Dell’Osso, and F. Re F, Fluvoxamine maleate:

double-blind clinical trial vs. placebo in hospitalized depressed
patients. Curr. Ther. Res., 43,468-479(1988).
24. A.J. Martin, V.M. Tebbs, and J.J. Ashford, Affective disorders in
general practice. Treatment of 6000 patients with fluvoxamine,
Pharmatherapeutica,5,40-49 (1 987).
25. W.K. Goodman, L.H. Price, and S.A. Rasmussen, Efficacy of

fluvoxamine in obsessive-compulsive disorder: a double-blind
comparison with placebo, Arch. Gen. Psychiatry, 46,36-44
(1989).
26. H. DeBree, J.B. Van der Schoot, and L.C. Post, Fluvoxamine
maleate: disposition in man, Eur. J. Drug Metab. Pharmacokinet.,
8, 175-179 (1983).
27. J. Van Harten, P. Van Bemmel, and M.R. Dobrinska MR,

Bioavailability of fluvoxamine given with and without food,
Bioparm. Drug Dispos., 12,571-575 (1991).
28. E. Perucca, Routes of drug administration., Med. Int., 101,4229-

4234 (1 992).
29. D.P. Doogan, Fluvoxamine as an antidepressant drug,

Neuropharmacologv, 19, 1215-1216 (1980).
30. J. Van Harten, Comparative pharmacokinetics of selective

serotonin re-uptake inhibitors, Clin. Pharmacokinet., 24,203-220
(1993).
31. U.A. Siddiqui, S.K. Chakravarti, and D.K. Jesinger, The tolerance
and antidepressive activity of fluvoxamine as a single dose
compared to a twice daily dose, Curr. Med. Res. Op., 9,68 1-690
(1985).
32. M.H. De Vries, M. Raghoebar, and I S . Mathlener, Single and

multiple oral dose fluvoxamine kinetics in young and elderly
subjects, Ther. Drug Monitor., 14,493-498 (1992).
33. H. Overmars, P.M. Scherpenisse, and L.C. Post, Fluvoxamine

maleate: metabolism in man. Eur. J. Drug Metabol.
Pharmacokinet., 8,269-280 (1983).
34. V. Claassen, Review of the animal pharmacology and

pharmacokinetics of fluvoxamine, Br. J. Clin. Pharmacol., 15
(3 SUPPI.),3498-355s (1983).
35. H.M. Ruijten, H. DeBree, and J.M. Borst, Fluvoxamine: metabolic

fate in animals, Drug Metab. Disp., 12, 82-92 (1984).
36. S. Wright, S. Dawling, and J.J. Ashford, Excretion of fluvoxamine

in breast milk, Br. J. Clin. Pharmacol., 31,209 (1991).
37. W. Wagner, B. Plekkenpol, and T.E. Gray, Review of

fluvoxamine safety database, Drugs, 43 (Suppl2), 48-54 (1992).
38. J.P. Ottervanger, P.M.L.A Van Den Bemtt, and G.H.P. de Koning,
Risk of bleeding during treatment with fluoxetine (Prozac) or
fluvoxamine (Fevarin), Ned Tijdschr Geneeskd, 137,259-26 1
(1993).
39. M.I. Wilde, G.L. Polsker, and P. Benfielf, Fluvoxamine: An

update review of its pharmacology, and therapeutic use in
depressive illness, Drugs, 46, 895-924 (1993).
40. M.A. Jenike, S. Hyman, and L. Baer, A controlled trial of

fluvoxamine in obsessive -compulsive disorder: implication for a
serotonergic theory, Am. J. Psychiatry, 147, 1209-1215 (1990).
41. W.K. Goodman, L.H. Price, and P.L. Delgado, Specificity of of

serotonin reuptake inhibitors in the treatment of obsessive-
compulsive disorder, Arch. Gen. Psychiatry, 47, 577-585 (1990).
42. W.K. Goodman, M.J. KO&, and M. Liebowitz, Efficacy of

fluvoxamine in obsessive-compulsive disorder; a second
multicentre study, Poster presented at the 18* Collegium
International Neuro-Psychopharmacologicum;Nice, France;
June 28-July 1, 1992.
43. G.K. Mallya, K. White, and C. Waternaux, Short-and long-term

treatment of obsessive-compulsive disorder with fluvoxamine,
Ann. Clin. Psychiatry, 4, 77-80 (1992).
44. J. Van Harten, R.L. Holland, and K. Wesnes, Influence of

multiple-dose administration of fluvoxamine on the
pharmacokinetics of benzodiazepine bromazepam and lorazepam:
a randomised, cross-over study, Eur. Neuropsychopharmacol., 2,
381 (1992).
45. J.C. Fleishaker and L.K. Hulst, Effect of fluvoxamine on the

pharmacokinetics and pharmacodynamics of alprazolamin healthy
volunteers, Pharm. Res. 9 (Supplement), S-295 (1 992).
46. E. Spina, A. Avenoso, and A.M. Pollicino, Carbamazepine

coadministration with fluoxetine or fluvoxamine, Ther. Drug
Monitor., 15,247-50 (1993).
47. A.H. Thomson, E.M. McGovern, and P. Bennie, Interaction

between fluvoxamine and theophylline, Pharm. J., 249, 137
(1 992).
Acknowledgements
The authors gratefully acknowledge the assistance of Dr. Faiyza El Shafie.

The skillful technical assistance of Mr. Abdel Rahman AlGhdeer is highly
appreciated.
GADOTERIDOL
Krishan Kumar,' Michael Tweedle,' and Harry G. Brittain2
(1) Bracco Research USA

P.O. Box 5225
Princeton, NJ 08520
(2) Ohmeda, Inc.

Pharmaceutical Products Division
100 Montain Avenue
Murray Hill, NJ 07974
AND EXCIPIENTS-VOLUME 24 All nghrs of reproduction in any form reserved.
1. Description
1.1 Name, Formula, and Molecular Weight
1.2 Appearance
1.3 History
2. Synthesis
3.2 NMR Spectrum
3.3 Mass Spectrum
3.4 Ultraviolet Spectrum
3.5 Luminescence Spectrum
3.6 Optical Activity
3.1 CrystallographicProperties
3.8 Relaxivity and Water of Hydration
3.9 Thermal Analysis
3.10 Hygroscopicity
3.1 1 Microscopic Characterization
3.12 Complex Formation Constant
3.13 Kinetic Inertness
3.15 Solubility
3.16 Conductivity and Osmolality
3.17 Viscosity and Specific Gravity
4.1 Elemental
4.2 Spectrophotometric
4.3 Chromatographic
4.3.1 Thin-Layer
4.3.2 High Performance Liquid
5. Stability
5.1 Solid State Stability
5.2 Solution Stability
5.3 Stability in the Presence of Endogenously Available
Ions
5.4 In Vivo Stability
5.5 Stability in Biological Fluids
6. Biological Studies
6.1 Tissue Distribution in Mice
6.3 Toxicity
7. Acknowledgements
8. References
1. Description
1.1 Name. Formula. and Molecular Weight
Gadoteridol (Gd(HP-D03A) is a nonionic contrast agent for magnetic
resonance imaging (MRI). The commercial product (ProHance) is
available as a 0.5 M sterile clear colorless to slightly yellow aqueous
solution in vials, for intravenous injection. The systematic chemical name
for Gadoteridol is 10-(2-hydroxypropy1)-1,4,7,10-tetraazacyc1ododecane-
1,4,7-triaceticacid, monogadolinium salt. The Chemical Abstracts
identification number is CAS-120066-54-8. The structural formula is:
c 17H29N407Gd Anhydrous M.W. 558.69
The elemental composition corresponding to the anhydrous form is C

36.55%, H 5.23%, N 10.03%, 0 20.05, and Gd 28.15%.
Each mL of ProHanceRcontains 279.3 mg gadoteridol, 0.23 mg

CalteridolRcalcium (calcium salt of calcium complex of HP-D03A), 1.21
mg of Trizma (as Tris acetate buffer) and water for injection. ProHance
contains no antimicrobial preservative.
1.2 Apearance
Gadoteridol is a white to off-white crystalline solid, obtained as aggregate
clumps of fine needle-like microcrystals. The compound has no inherent
odor.
1.3 History
Gadoteridol, unformulated raw material, was designed as a nonionic and

hydrophilic Gd3+chelate species, intended for use as a contrast agent in
Magnetic Resonance Imaging (MRI) [ 1-31. The contrast-enhancedMRI
images are used to differentiate diseased from normal tissue, and to
identify specific disease states (such as infarcts, abscesses, and tumors).
Agents of this type have been found to be effective as water-proton

relaxation catalysts in both aqueous solution and in human serum, and
their use greatly improves the quality of data available in such studies
[ 1-31. The relaxivity of gadoteridol is comparable to that of the ionic
agents, Gd(DTPA)'- and Gd(D0TA)-.
2. Synthesis
The synthesis of the free ligand and its Gd3+ complex gadoteridol, is
published [4,5] (scheme given below). The procedure involves protection
of one nitrogen of the macrocyle, 1,4,7,10-tetraazacyclodoecane,by
forming a novel intermediate, 1,4,7,10-tetraazacyclododecane-1-
carboxyaldehyde. Alkylation of this protected macrocycle by t-
butylbromoacetate in basic media and subsequent hydrolysis by sulfuric
acid yielded D03A. D03A was reacted with propylene oxide at pH 12 to
produce HP-D03A. Excess propylene oxide was removed under vacuum.
The crude material was purified by cation exchange, anion exchange, and
PVP resins. HP-D03A was reacted with gadolinium oxide in water to
produce Gd(HP-D03A).
3.1 Infrared Suectrum
The infrared absorption spectrum of gadoteridol was obtained in a KBr
disc, and is shown in Figure 1. One distinctive carbonyl band was
observed at 1613 cm-', and is assigned to the three unresolved, equivalent
carboxylate groups of the ligand. The numerous bands noted in the
fingerprint region can be assigned on the basis of their band energies, and
a summary of the assignments is found in Table I.
Energy (ern-')
Figure 1. Infrared absorption spectrum of gadoteridol, as obtained in a

KBr disc.
Table I. Band Positions and Assignments for the Vibrational Transitions

of Gadoteridol
Frequency (l/cm) Assignment

1085 Secondary hydroxyl group stretching mode
1323 Symmetric carbonyl stretching mode of the carboxylate groups
1384 Combination band of the carbonyl stretching (carboxylate
groups) and the symmetric methyl deformation modes
1458 Methylene asymmetric deformation
1477 Methylene symmetric deformation
1613 Asymmetric carbonyl stretching mode of the carboxylate groups
1635 Hydroxyl bending mode
2858 Methylene stretching mode
2975 Asymmetric methyl stretching mode
3.2 Nuclear Magnetic Resonance Spectrum

Owing to the high paramagnetism associated with the Gd" ion in
Gadoteridol, NMR spectra of gadoteridol cannot be obtained. However,
the 'H and 13C NMR spectra of the free ligand, HP-D03A, have been
obtained. The 'H NMR spectrum was obtained at 400 MHz in D20
solution, and is shown in Figure 2. This spectrum was internally
referenced to the HOD peak at 4.65 ppm. The protons of the C1 methyl
group were observed to resonate as a doublet at 1.13 ppm. The multiplet
peak at 4.08 ppm was assigned to the C2 proton, and the remaining
methylene protons resonate as the broad, partially overlapped multiplets
observed between 2.9 and 3.7 ppm.
c... . . ~
i
C b c m i i Shift (ppm)
Figure 2. 'H NMR spectrum of HP-D03A, obtained at 400 MHz in D20

solution, and internally referenced to the HOD peak at 4.65
PPm.
The 13C NMR spectrum was obtained at 270 MHz, and is found in Figure
3. This spectrum was externally referenced to p-dioxane at 67.6 ppm. The
C 1 carbon resonates at 2 1.6 ppm, while the C2 carbon was observed at
64.4 ppm. The carbonyl groups (C 13, C 15, and C 17 yield the peaks noted
at 172.9 and 174.6. The 12 methylene carbons yield the partially
overlapped resonances noted at 49.8,50.7,51.2,56.2,56.4, and 60.5 ppm.
Figure 3. 13C NMR spectrum of HP-D03A, obtained at 270 MHz,

and externally referenced to p-dioxane at 67.6 ppm.
3.3 Mass Spectra
Gadoteridol has been characterized using both positive and negative ion
fast atom bombardment (FAB), with the compound being dissolved in
water and thioglycerol and sputtered by 8 keV xenon atoms. These data
are illustrated in Figures 4 and 5, where it should be noted that the
characteristic isotope pattern for gadolinium is evident in all the spectra.
The relative natural abundances of gadolinium isotopes are 152Gd (0.2%),
154Gd(2.2%), 155Gd (14.8%), 156Gd (20.5%), 157Gd (15.7%), 158Gd
(24.8%), and 160Gd (21.8%).
The (M+H)+ ion (where M represents the 158Gd gadoteridol species)

occurs at 560' in the positive ion spectrum, while the corresponding (M-
H)-ion is present at 558- in the negative ion spectrum. The remaining
peaks are due to the thioglycerol used in the solvent system.
Figure 4. Positive ion, fast atom bombardment (FAB), mass spectrum of

gadoteridol.
;
u_-
(00
alr
Figure 5 . Negative ion, fast atom bombardment (FAB), mass spectrum

of gadoteridol.
3.4 Ultraviolet S-pectrum
The full UV spectrum obtained for gadoteridol (aqueous solution) is
shown in Figure 6. The observed spectrum consists of a number of sharp
transitions, each of which can be assigned on the basis of known spectral
behavior [6]. The assignments are fully described in Table 11.
Most of the absorption bands are too weak to be useful for analytical
purposes, but the 8S--> 6P band system (270 - 280 nm) is sufficiently
intense to permit its use. The distinctive triplet of absorption bands noted
in this region can be used as an identity test for Gd3' in gadoteridol or its
formulation. The most intense component of the triplet is observed at 274
nm, and its molar absorptivity is approximately 2.5 M-1cm-1.
f- 1
Wavelength (am)
Figure 6. Complete UV absorption spectrum of a 55 mM aqueous

solution of gadoteridol.
FXUVOXAMINE MALEATE 217
Table II. Absorption Bands Observed for Gd3+in Gadoteridol
32,785 305.0
35,840 279.0
35,970 278.0
36,300 275.5
36,365 275.0
36,630 273.0
36,630 273.0
39,685 252.0
40,325 248.0
40,650 246.0
40,815 245.0
--->6D5/2 40,985 244.0
3.5 Luminescence Spectrum
Gd3+can be excited through the 8S--> 6P band system, and metal-centered

luminescence can be observed. The luminescence spectrum of an aqueous
gadoteridol solution is shown in Figure 7. The strongest emission band is
observed at 3 1 1 nm, and corresponds to the 6P7/2 --> 8S7/2 transition [7].
w~veiengtb
Cnm)
Figure 7. Luminescence spectrum of a 55 mM aqueous solution of

gadoteridol. An excitation wavelength of 274 nm was used to
obtain the spectrum.
The intensity of this luminescence is sufficient that it has been used as the
basis of a variety of analytical methods.

Gadoteridol contains one center of dissymmetry, and is therefore capable
of being resolved into two enantiomers. The (S)-enantiomer is identified
as SQ-33236, while the (R)-enantiomeris identified as SQ-34208.
Gadoteridol is marketed as the racemic mixture of these, and therefore
exhibits no optical activity.
The circular dichroism (CD) spectra of SQ-33236 and SQ-34208 were

obtained [8], and are shown in Figure 8. Owing to the very small molar
absorptivity values, it was not possible to obtain the CD spectra within the
intrinsic Gd” absorption bands. The CD spectra shown in Figure 8
correspond to the chirality within the carboxylate absorption bands of the
chiral ligand, with the CD peaks being observed at 193 nm. As would be
anticipated, the CD spectra of the enantiomers were mirror images of each
other. The absolute value of the circular dichroism (A&) was found to be
0.146 M-lcm-l, while the absolute value of the molar ellipticity ([el) was
found to be 481 degrees-cm2/decimole.
.P ,a x..
W-rsl-ltb
-
<nm>
zy
Figure 8. Circular dichroism spectra of the enantiomers of gadoteridol.

Shown are data for SQ-33236 (the (S)-enantiomer) and SQ-
34208 (the (R)-enantiomer).
3.7 Crystalloeraphic ProDerties
The powder x-ray diffraction pattern of the 3.5-hydrate form of
gadoteridol was obtained using a copper source (1.54060 A) [9], and is
FUJVOXAMINE MALEATE 219
reproduced in Figure 9. A total of 36 peaks were detected at scattering

angles between 2 and 32 degrees 2-8. The two most diagnostic scattering
peaks suitable for identification were observed at 8.7 degrees 2-8 (d-
spacing of 10.20A) and at 10.1 degrees 2-8 (8.75 A), although a very
diagnostic series of scattering peaks was noted between 12 and 16 degrees
2-8. A summary of scattering angles, d-spacings, and relative intensities
associated with the most intense peaks is found in Table III.
Figure 9. Powder x-ray diffraction pattern of gadoteridol.
Gadoteridol does not appear to exhibit either polymorphism or

pseudopolymorphismwhen crystallized from water/isopropanol mixtures.
The anhydrous compound exhibited a similar powder pattern to that
shown in Figure 9 for the 3.5-hydrate. When exposed to humidity values
between 15 and 70%,gadoteridol exhibits a constant powder pattern.
The structure of gadoteridol was determined by single crystal x-ray

diffraction analysis [lo]. The space group was found to be P212121, with
cell constants of a = 16.974 A, b = 25.45 A, and c = 11.247 A. The
asymmetric unit was found to contain two crystallographically
independent ennea-coordinate gadolinium complexes, and five partially
occupied sites of hydration. These two structures are illustrated in Figure
10. In each molecule, the macrocyclic ring Gd(HP-D03A) adopts a
quadrangular [3333] conformation, in which the four nitrogen and four
oxygen donor atoms are coordinated to the central Gd3+ion. The ninth
apical site is occupied by a water molecule in the capped square antiprism
arrangement, and this finding has been confirmed using spectroscopic
methods [ 113. The nitrogens are coplanar within experimental error, as are
Table 111. Powder X-ray Diffraction Data Obtained for Gadoteridol:

Scattering Angles, D-spacings, and Relative Intensities
Scattering Angle D-spacing Relative Int.

(degrees2-8) (A, (IAomax)
8.6255 10.2468 100.00
10.0800 8.7682 97.75
10.9975 8.0387 17.58
1 1.7625 7.5175 9.7 1
12.5575 7.0433 43.19
13.0775 6.7644 37.44
13.5200 6.5440 14.84
14.0475 6.2994 51.37
14.8175 5.9738 28.36
15.7750 5.6133 12.74
16.7525 5.2879 7.40
18.9275 4.6849 9.89
19.2025 4.6184 18.30
20.1950 4.3936 18.06
22.6450 3.9235 7.40
24.7500 3.5943 31.14
25.1200 3.5422 8.03
25.9600 3.4295 12.74
26.3400 3.3809 11.75
26.4200 3.3708 10.61
26.9025 3.3 114 30.52
27.2375 3.2715 36.75
27.8875 3.1967 19.78
28.1000 3.1730 23.19
28.3950 3.1407 17.82
28.7800 3.0995 24.86
29.6400 3.01 15 22.38
30.4200 2.9361 8.68
30.8625 2.8950 11.36
31.8000 2.81 17 13.56
FLUVOXAMINE MALEATE 22 I
the four coordinated oxygen atoms of the ligand arms. The Gd3+ion lies
between these parallel planes, 1.61 %, above the nitrogen plane and 0.75 8,
below the oxygen plane. The coordinated water molecule lies 1.72 8,
above the oxygen plane, and an extensive hydrogen-bonded network joins
the complexes with the water of hydration. It is of interest that the two
independent complexes in the asymmetric unit have distereomeric
conformation.
Figure 10. Independent complexes in the asymmetric unit of gadoteridol,

as determined by single crystal x-ray diffraction.
3.8 Relaxivitv and Water of Hvdration
The catalysis of the relaxation rate of water protons of tissues is governed
by a second-order rate constant called relaxivity (rl). The T i relaxivity,
'1, of a paramagnetic metal ion was determined by the measurements of
relaxation times of several Gd3+solutions at different concentrations of
GdL at 20 MHz. The slope of a plot of l/T, vs. [Gd3+]gave the measured
relaxivity, 2or1 . The 20r1 value for Gadoteridol is 3.7 mM-1s-1 in water
[2]. The number of coordinated waters (Q) on corresponding Tb(II1)
chelates were determined by a literature procedure and the calculated

value is 1.3 [ 111.
20
F
15
. c
t
r
10
0
0 2 4 6
l o J [Gd (HP-D03A)], M
Figure 11. Plot of 1R1 vs. [Gd(HP-D03A)]. Taken from ref. 2.
3.9 Thermal Analvsis
A typical differential scanning calorimetric thermogram of gadoteridol
was obtained at heating rate of 10"C/min [9], and is shown in Figure 12.
The compound exhibits a poorly defined dehydration endotherm around
loO°C, and a much sharper dehydration endotherm around 17OoC. The
low temperature thermal event is undoubtedly associated with the loss of
the network water, while the high-temperature endotherm is certainly due
to the removal of the water coordinated directly to the Gd3+ion.
Even when run under the same temperature ramping conditions as

differential scanning calorimetry, the thermogravimetric analysis of
gadoteridol does not differentiate between the two types of water bound in
the crystalline solid [9]. As evident in Figure 13, the compound exhibits a
gradual weight loss, which is only complete by 175OC. The anhydrous
compound formed above this temperature exhibits no tendency for thermal
decomposition, even when heated up to 3OOoC.
-0.4
75 125 175 275
Temperature (“C)
Figure 12. Differential scanning calorimetry of gadoteridol
100 -
h
96-
Mass
10.9%
Loss
(%) 92.
90
Figure 13. Thermogravimetric analysis of gadoteridol.

224 NAGWA H. FODA ETAL.
Re1uIvo Humidity <%)
Figure 14. Moisture uptake of gadoteridol. as a function of

external relative humidity.
3.1 1 Microscopic Characterization

As evident in Figure 15, the component crystals of gadoteridol are ortho-
rhombic in nature, and are normally long and rod-like in appearance [ 121.
The long crystal axis is often terminated in well-defined bi-capped faces.
3.10 Hveroscopicity
Gadoteridol samples were exposed to various relative humidity conditions
(over a series of saturated salt solutions), after which the water content
was determined using thermogravimetry 181. The water sorption
isotherms are shown in Figure 14, where it is evident that a region of
constant moisture uptake exists between relative humidity values of 15 to
70%. Averaging all the weight losses within these humidity values for
both samples yielded an average weight loss of 10.17%. The calculated
water content for a 3.5-hydrate species would be 10.14%, indicating that 7
waters of hydration are shared between two gadoteridol molecules. This
conclusion is exactly what was deduced from the single crystal x-ray
diffraction studies [ 101.
Figure 15. Scanning electron photomicrograph of gadoteridol,

obtained at (a) 150x and (b) 500x magnification.
3.12 Complex Formation Constants
Gadoteridol is formed by the complexation of Gd3' with the chelating

ligand, HP-D03A. Both the ligand protonation constants of HP-D03A
and the complex stability constant of gadoteridol have been determined
[ 101. The ligand protonation constant (log K at 25.W.l0C and p= 0.1
(TMAC1) were determined from pH-potentiometric titrations and the
values are: 11.96,9.24,4.43, and 3.48. The stability constant of
Gadoteridol was determined by a spectrophotometric titration using
Arsenazo-I11as the indicator. The stability constant of gadoteriodol is
logK = 23.8, which is one of the highest formation constants measured for
a lanthanide chelate species. A knowledge of the conditional stability
constant at pH 7.4 is useful in the context of physiological conditions.
The stability constant is reduced considerably due to the competition
between hydrogen ions and the metal ion for the ligand (Fig. 16). A
calculated
2 4 6 8 10 12
PH
Figure 16. Effect of pH on the Stability Constant of Gd(HP-D03A).
Taken from ref. 2.
3.13 Kinetic Inertness
Gadoteridol is kinetically very inert under physiological conditions [ 131.
However, at lower pH, the chelate slowly equilibrates into free Gd3+ and
the free ligand. For example, the half-lives of the chelate are 30 and 3 h at
pH 2 and 1, respectively. A plot of k b s d vs. [H+] is shown in Fig 17.
30 t
20
10
I I
0 0.4 0.8 1.2
W+l, M
Figure 17. Plot of k b s d vs. [H+] for acid-assisted dissociation of
Gd(HP-D03A). Reprinted with permission from ref. 13.
Copyright 1993American Chemical Society.
Another method to verify the kinetic inertness of a metal chelate is to
study the exchange of metal ion with its metal chelate, The isotope-
exchange reaction (1) was studied [ 141 and no exchange at pH 7.4 and 3.8
was seen within 14 days [14]. Low pH conditions are required to observe
the acid-assisted exchange reaction of Gd(HP-D03A).
+ Gd(HP-D03A) <=====> '53Gd(HP-D03A)+ Gd3+

IS3Gd3+ (1)
The partition coefficients of gadoteridol in butanoVwater and
octanoVwater were determined by the traditional shake-flask method. The
values of log P are: -3.68 in octanoVwater and -1.98 in butanouwater
[ 151. In both instances, the aqueous phase was buffered to pH 7. The
negative values obtained for the compound are consistent with the
interpretation that gadoteridol is highly hydrophilic.
3.15 Solubility
Gadoteridol exhibits a wide range of solubilities in various solvents [8].
Gadoteridol was found to be freely soluble in water and methanol, soluble
in isopropanol, and much less soluble in a variety of other neat solvents.
The data are summarized in Table IV. The solution pH of aqueous
gadoteridol solutions was not found to vary strongly with concentration.
Over the range of 5 - 20 mg/mL, an average solution pH of 6.4 was
observed [8].
Table IV. Solubility of Gadoteridol in Various Solvents
Solvent mg/mL USP Definition

Water 737 Freely soluble
Methanol 119 Freely soluble
Isopropanol 41 Soluble
Dimethylformamide 10.1 Sparingly soluble
Acetonitrile 6.1 Slightly soluble
Methylene chloride 5.2 Slightly soluble
Ethyl acetate 0.5 Very slightly soluble
Acetone 0.4 Very slightly soluble
Hexane 0.2 Very slightly soluble
Toluene 0.3 Very slightly soluble
3.16 Conductivitv and Osmolalitv
The conductivity of gadoteridol solutions was measured to demonstrate
the nonionic nature of the compound. The molar conductivity (measured
in units of micro Siemens) was found to be 0.8, a value which is far below
the ranges noted for either 1: 1 or 1:2 electrolyte species [ 161. The
negligible conductivity indicates that gadoteridol does not dissociate into
ions upon dissolution, and that it remains nonionic in aqueous solutions.
The lack of ionic character was further studied through osmolality

measurements. A 0.5 M solution of gadoteridol was found to exhibit an
osmolality of only 0.6 Osmolkg H20 [ 161, which clearly places the
compound in the nonionic category.
3.17 Viscositv and Specific Gravity
The viscosity of a formulated 0.5 M solution of Gadoteridol was measured
by a rolling- ball method with the use of a PAR-200 microviscometer.
The measured viscosity values [ 161 of the product are 2.0 and 1.3 CPat
20OC and 37OC, respectively. The viscosity at body temperature, 37OC,
is very close to the viscosity of blood, which is 1.32 cP. The specific
gravity of the formulated product is 1.140.
4.I Elemental Analysis

The racemic mixture of the R- and S-isomer was synthesized by the
literature procedure as given above. The elemental analysis was
satisfactory as given below [ 101.
Analysis: Found (calculated) for C17H29N407Gd.1.3H20: C, 34.86

(35.07); H, 5.58 (5.48); and N, 9.77 (9.63)
4.2 Spectrophotometry
A qualitative and quantitative spectroscopic method for the identification
of Gd" in Gadoteridol was developed. The characteristic UV absorbance
spectra consists of a triplet of bands at 274,276, and 279 nm. This pattern
of bands was used for qualitative identification of gadoteridol. The
absorbance of the 274 nm peak (EE 2.5 M-1cm-l) was used to
qualitatively evaluate the concentration of the product.
4.3 Atomic Absorbance Spectrometry
Gadoteridol samples were analyzed for the presence of trace metal ion
concentrations, (e.g. Zn, Mn, Fe, Cu, and Cr) by Flame Atomic
Absorption Spectrometry [17]. Figure 18 and 19 show correlation curves
for Cu, Fe, Mn, and Cr and Zn, respectively. No trace metals were
detected in the samples. The limits of detection (given in parenthesis as
pg/mL) of these trace metals were as follows: Cu (2.5), Cr (5), Fe(2.5),
Zn( 1.O), and Mn( 1).
Concentration, pg/mL
Concentration, p g / d
Figure 18. Correlation curves for Figure 19. Correlation curves for
Cu, Fe, and Mn Cr and Zn.
Total calcium in ProHance, was determined by a Flame Atomic

Absorption Spectrometric method [17]. Figure 20 shows the calcium
absorbance as a function of concentration from 0 to 2.5 pg/ml. A
correlation coefficient of 0.99994 indicates good linearity over this range.
Figure 21 shows addition curve, i.e., response of assay concentration vs.

added calcium concentration in the formulation. From this curve the
concentration of total calcium was calculated as 30.3H.3 pg/mL (n=5),
which can be compared with the theoretical calcium concentration in the
formulation of 30.06 pg/mL.
I I
7
10 20 1
[Calcium], @nL Cohc. of Ca Added, p g / d
Figure 20. Plot of Calcium absorbance vs. concentration of Ca2+.
Figure 2 1. Addition Curve for the determination of Ca2+.
4.4 Chromatopraphic
4.4.1 Thin-Layer Chromatomauhy

Thin-layer chromatography was evaluated to determine free Gd3+ in
gadoteridol by the tracer technique [18]. The method is valid to determine
free Gd3+ at the 0.1% level. In the method gadoteridol was first
radiolabeled with 153Gd3+ by either the exchange method or by the true-
tracer radiolabeling method. Free gadolinium was precipitated as Gd(P0,)
by addition of KH2PO4 (KspGd(Po4)= Then the sample was
applied to an ITLC (Instant Thin Layer chromatography) plate and
developed in 50% aqueous methanol containing 10% ammonium acetate.
Free Gd3+ as Gd(P04), was found at the origin while gadoteridol moved
near the solvent front. Mass transfer and poor recovery made this method
unacceptable for determination of low concentrations of free Gd3+ in
Gadoteridol.
FLUVOXAMINEMALEATE 23 1
4.4.2 High-Performance Liquid Chromatoeraphv
4.4.2.1 Determination of free Gd3+ in Padoteridol

Free Gd” in the sample of Gadoteridol was determined by complexing free
Gd3+with EDTA, followed by separation of Gd(EDTA)-(rt = 2.9 min)
from Gd(HP-D03A) (rt =10.8 min ) by an HPLC method [19]. This
method utilizes a mobile phase consisting of 10 mM EDTA and 50 mM
Tris acetate at pH 7.4 with 2%acetonitrile as an organic modifier. A C18
reversed-phase Nucleosil column was used. A fluorescence detection
method with excitation and emission wavelengths at 274 and 3 11 nm was
used. Using this method one can determine very low concentrations of
free Gd3+as well as Gd(D03A) (rt = 5.5) and Gd(D0TA)- (rt = 4.6 min) as
other possible impurities (Fig. 22). The limit of detection were determined
as follows: 0.00066 m g / d of free Gd”, which correspond to 0.002% of
Gd(HP-D03A) in the formulation. Similarly, the limit of detection for
Gd(D0TA)- and Gd(D03A) were determined as 0.005% and 0.017%,
respectively. In a research sample, the percentages of Gd3+,Gd(DOTA),
and Gd(D03A) were determined as <0.002%, <0.004%, and 0.07%,
respectively.
4.4.2.2 Determination of the Free Ligand bv HPLC
The determination of the free ligand, HP-D03A in Gadoteridol, involves
complexation of HP-D03A by Cu(II), and subsequent UV detection (at
280 nm) of the copper complex. The linearity, reproducibility, and
recovery of the method were found to be suitable for analysis [20]. The
method utilizes a Hamilton PRP-X100, anion exchange column (15 x
0.41cm), mobile phase: 98.8%buffer (10 mM EDTA and 50 mM Tris
acetate, pH 7.4), 1% acetonitrile, and 0.2% THF, UV-Vis detector at 280
nm and flow rate of 1.0 mUmin. In the method, copper acetate, in tris
acetate buffer (no EDTA), is first mixed with Gadoteridol solution so that
free ligand reacts with Cu(I1) to form Cu(HP-D03A). Unreacted Cu(I1) is
then reacted with the EDTA (which is in excess) in the mobile phase.
Excess Cu(II) in the form of Cu(EDTA)- and Free ligand in the form
Cu(HP-D03A) are separated by HPLC with retention times of 4.5 and 2.3
min, respedtively (Fig. 23). As low as 0.002% (w/w) free ligand can be
determined with good precision.
1,.
Cd (DOTA 2 SQ 30626
21.031. I . I . I . I . I . I . I .
000 212 4.25 6.35 8.50 10.53 12.76 14.52 17.01
RT in minutes
Fig. 22. HPLC chromatogram to determine the concentration of free
Gd3+ in a sample of Gadoteridol.
Fig. 23. HPLC chromatogram of determination of HP-D03A in

Gadoteridol.
4.4.2.3 Potencv Determination
An HPLC method of satisfactory linearity and reproducibility has been
developed to obtain the identity, the characteristic retention time of the
analyte, and potency of Gd(HP-D03A) in either gadoteridol or
ProHance. The HPLC conditions used are the same as used for the
determination of free Gd3’in Gadoteridol. The fluorescence detection
method with excitation and emission wavelengths at 274 and 3 11 nm,
respectively, was used. First a linearity or calibration curve, peak area vs.
[Gd(HP-D03A)], is established. The concentration of the unknown
solution is calculated from the slope of calibration curve and peak area of
unknown solution of gadoteridol. With the use of this method the
concentration of ProHance was determined as 0.505+0.005 M.
4.4.2.4 Determination of Complexed Calcium in the
Formulation
The HPLC method to determine total complexed calcium, or the
concentration of Ca[Ca(HP-D03A)]2 ,involves the reaction of Ca[Ca(HP-
D03A)]2 with Cu2+ to form Cu(HP-DO3A). followed by separation of
Cu(HP-D03A) from the unreacted Cu2+ . In the procedure, the reaction
of known concentrations of CuC12 and Ca[Ca(HP-D03A)]2 is completed
in 10 mM Tris.HC1 buffer at pH 4.0 in 1 h. Separation of Cu(HP-D03A)
formed and unreacted CuC12 is accomplished by an HPLC method (Fig.
24). From the calibration curves of standards and the peak area of
Cu(HP-D03A) formed in the unknowns, one can determine the
concentration of Ca[Ca(HP-D03A)]2. The conditions of separation are:
PLRP-S (15 x 0.46 cm) column, 99: 1 buffer:acetonirile (buffer being 50
mM NH4H2PO4 at pH 4.0, 1.O mL/min flow rate, and 270 nm UVNis
detector.
161.56
lU.01
h
II
1..
0.00 1.2S 2.50 3.16 5.01 6.26 Z.52 8.17
RT in minutes
Figure 24. Determination of Ca” as Ca[Ca(HP-D03A)], in ProHance.

4.4.2.5 Determination of Tris in the Formulation

The concentration of Tris [Tris (hydroxymethyl)aminomethane]buffer in
ProHance was determined using ion-pair reversed-phase HPLC with
refractive index detection. The conditions used were: Partisil-5-ODs-3
(25~0.46cm) column, 50% aqueous methanol containing 7 mM
Dodecanesulfonate at pH 5.0,1.O a m i n flow rate, and Shimadzu RID-6
refractive index detector. The concentration of Tris in the formulation was
within 10051% of the concentration claimed on the packaging insert [21].
4.4.3 Ion Chromatopraphv
Simultaneous determination of sodium and potassium in gadoteridol were
accomplished by an ion chromatography method [22]. The method
employs a 25 cm Dionex HPLC-CS3 cation exchange column and an
eluant of mixture of 12 mM HC1 and 0.5 mh4 D,L-2,3-diaminopropionic
acid monohydrochloride (DAP.HC1). Detection was carried out by
conductivity with chemical suppression. Since Gadoteridol is a neutral
molecule, it gives a weak conductivity response and is not retained on the
analytical column. Peak responses of sodium and potassium were found
to be linear in the range of 0.21 to 10.9 ppm (w/v) (with correlation
coefficient =0.999548) and 0.20 to 20.1 ppm (with correlation coefficient
=0.999548), respectively (Fig. 25 and 26). The limit of detection (LOD)
and minimum quantifiable limit (MQL) are: 0.01% and 0.05% (w/w),
respectively. Samples of gadoteridol were analyzed and the percentages of
Na+ and K+ were c 0.1 and c 0.05, respectively, for the sample in which
KOH was used for pH adjustment.
Figure 25. Correlation of peak area with "a+].

I
Figure 26. Correlation of peak area with [K+].
5. Stability
5.1 Solid State Stability

The effect of temperature and humidity (-2OOC, 5OC, 22OC/80% relative
humidity, 33OC and 40OC/75% relative humidity) was studied on the bulk
material for a long period of time. The samples were analyzed, at a
regular interval for 18 months, for appearance, color, and moisture by Karl
Fischer titration, free Gd3+ level, potency and impurity index by HPLC
and TLC. No significant changes in potency or in impurity levels were
observed by either HPLC or TLC. These results indicate that Gadoteridol
is reasonably stable to heat and heat/moisture conditions.
The effect of light on the bulk material was also studied. The samples
were stored in open petri dishes under 400 and 900 foot candles of
fluorescent light and short wave UV light. Batches exposed to 400 and 900
foot candles of fluorescent light for a period of two months showed
reasonable stability. However, batches showed some light sensitivity
between two and three months under these conditions. All the batches
exposed to short wave light showed degradation within the first month as
indicated by decreased potency, increased free gadolinium levels and
increased HPLC and TLC impurity levels. Thus, solid gadoteridol
samples should not be exposed to light over extended periods of time [23].
5.2 Solution Stabilitv

The stability and potency of gadoteridol was determined over a pH range
of 6 to 9 by validating HPLC methods. Time dependent determinations of
free HP-D03A ligand, free Gd%on, and the concentration of gadoteridol
showed no signs of any dissociation or decomposition indicating that the
product is stable in solution in this pH range within 22 days [24].
5.3 Stabilitv in the Presence of Endogenouslv Available Ions
The extent of reaction of Gadoteridol in vitro with Cu2+ and Zn2+ in the
presence of phosphate was determined 1251. Due to the presence of
macrocycle in the chelate, Gadoteridol proved to be inert, while other
agents reacted with Cu2+ and Zn2+ in the presence of 66 mM phosphate
(Fig. 27)
PO*”-
GdL + M <------>Gd + ML
where M = C u o r Zn
DTPA-BMA DTPA DOTA HP-D03A
Fig. 27. Reaction of Cu2’and Zn 2+ with Some Gd3+ chelate in the

presence of phosphate in 30 min. First bar is for Cu(I1) and
second for Zn (11).
5.4 In Vivo Stability
In vivo stability of Gadoteridol was measured in terms of residual free
Gd3+ in mice post 14 days intravenous administration of radiolabeled
Gadoteridol [26,27]. Consistent with greater kinetic inertia, Gadoteridol
leaves very low (<O. 1%) residual free Gd3+ in vivo in mice (Fig. 28).
100 - 0.1% free Gd
-4 10
-
Gd ( EDTA)
u
$
ei 1 -
Gd(NP-DOBA)
2 Gd(DTPA-BMA)
ap Gd (D03A)
c
g 0.1 - Gd ( DTPA)
3 Gd( DOTA)
I Gd (HP-DOJA)
0.01 I I 1 I I I
5 60 I d 7d 14d
rnin rnin
Time Post InJectlon
Fig. 28 Plot of % ID versus time for some Gd3+ chelates.
5.5 Stabilitv in Biological
- Fluids
Measurement of relaxivity of Gadoeridol in rat serum was made as a
function of time. Several concentrations of Gadoteridol were incubated at
37OC with rat serum. No change in the relaxivity over a period of 60 min
suggested that the chelate does not have any reaction with biological
components of rat serum. This also suggested that there is no significant
dissociation of the product. [28}
6. Biological Studies
6.1 Tissue Distribution in Mice

The tissue distribution of Gadoteridol was studied following iv
administration of 0.1 to 0.25 mmovkg 153Gd-labeled Gadoteridol to
unanesthetized mice [26]. In one hour, approximately 90% material was
excreted from mice. Greater than 97% of Gadoteridol was excreted into
urine within 1 day. The primary route of excretion was found to be renal.
Eighteen normal male volunteers were studied at doses of 0.025-0.3
m o l l k g intravenous gadoteridol in phase I study, to determine safety and
pharmacokinetics. Mean distribution half-life was 0.2 h and mean
eliminatjon half-life was 1.6h [ 161.
6.3 Toxicity
In mice, the acute intravenous LD50 was 11-14 mmol/kg, and in rats the
minimal lethal dose was 10 mmolkg. In two week studies, no serious
effects were observed in mice given 3 mmol/kg or in dogs given 1.5
mmol/kg daily. In vitro, 50 mh4 formulated gadoteridol showed no
tendency to hemolyze human erythrocytes [ 161.
7. Acknowledgements
The authors wish to thank the coworkers at Bristol-Myers Squibb Co. and
Bracco Research USA, who allowed us to cite their unpublished work.
FXUVOXAMINE MALEATE 239
8. References
1. R. B. Lauffer, "ParamagneticMetal Complexes as Water Proton

Relaxation Agents for NMR Imaging: Theory and Design", Chem.
-
Rev. 87,90 (1987).
2. K. Kumar and M. F. Tweedle, "MacrocyclicPolyamino Carboxylate

Complexes of Lanthanides as Magnetic Resonance Imaging Contrast
Agents", 65,515 (1993).
3. M. F. Tweedle, "Relaxation Agents in NMR Imaging" in Lanthanide

Probe in Life. Chemical, and Earth Sciences, J. -C. G. Bunzli, G. R.
Choppin, editors, Elsevier Publishing Co. Amsterdam, the
Netherlands, 1989.
4. M. F. Tweedle, G. T. Gaughan, and J. H. Hagan, 1-Substituted -1,

4,7 -Tris Carboxymethyl - 1,4,7 -Tetraazacyclododecane and Analogs"
US Patent # 4,885 363.
5. D. D. Dischino, J. F. Delaney, J. E. Emswiler, G. T. Gaughan, J. S.

Prasad, S. K. Srivastava, and M. F. Tweedle, Synthesis of Non-Ionic
Gadolinium Chelates Useful as Contrast Agents for Magnetic
Resonance Imaging", Inorg. Chem. 30, 1265 (1991).
6. C.A. Morrison and R.P. Leavitt, "SpectroscopicProperties of Triply

Ionized Lanthanides in Transparent Host Crystals", chapter 46 in the
Handbook on the Physics and Chemistry of Rare Earths, K.A.
Gschneider and L. Eyring, eds, North-Holland Pub., Amsterdam, 1982.
7. G. Stein and E. Wurzberg, "Energy Gap Law In the Solvent Isotope

Effect on Radiationless Transitions of Rare Earth Ions", J. Chem.
Phvs. 62,208 (1975).
8. H.G. Brittain, Unpublished Work.
9. S.J. Bogdanowich, Unpublished Work.

10. K. Kumar, C. A. Chang, L. C. Francesconi, D. D. Dischino, M.

Malley, J. Gougoutas, and M. F. Tweedle,"Synthesis, Stability, and
Structure of Gadolinium(1II) and Yttrium(II1) Macrocyclic Poly(amino
carboxylates)", Inorg. Chem. 33, 3567 (1994).
1 l.X. Zhang, C. A. Chang, H. G. Brittain, J. M. Garrison, J. Telser, and

M. F. Tweedle, "pH Dependenceof Relaxivities and Hydration
Numbers of Gadolinium (III) Complexes of Macrocyclic
Aminocarboxylates", Inorg. Chem. 3 1,5597(1992).
12. J. DeVincentis, Private Communication.
13. K. Kumar, C.A. Chang, and M. F. Tweedle, "Equilibrium and Kinetic

Studies of Some Lanthanide Complexes of Macrocyclic Polyamino
carboxylates", 1norg.Chem. 32,587 ( 1993).
14. K. Kumar, A. Chang, and K. Sukumaran, Unpublished Work.
15. K. Kumar, K. Sukumaran, S. Taylor, C. A. Chang, A. D. Nunn, and

M. F. Tweedle, "Partition Coefficients (log P), and HPLC Capacity
Factors (k') of Some Gd" Complexes of Linear and Macrocyclic
Polymino Carboxylates", J. Liquid Chromatomaphy, 17,3735, 1994.
16. M. F. Tweedle and V. M. Runge, "Gadoteridol" Drugs of the Future,

17, 187, 1992.
17.S.S.Black, N. S. Lewen, and P. S. Valatin, Unpublished Work.
18. K. Kumar, K. Sukumaran, C. Allen Chang, M. F. Tweedle, and W. C.

Ekkelman, "True Tarcer Radiolabeling of Gadolinium Complex of 10-
1,4,7triacetic
(2- Hydroxypropy1)- 1,4,7,1O-tetraazacyclododecane-
acid (HP-D03A)" J. Labelled Compounds and Radiopharmaceuticals
33,473(1993).
19. J. Hagan, S. Taylor, and M. Tweedle, "Fluorescence Detection of

Gadolinium Chelates Separated by Reversed-Phase High-Performance
Liquid Chromatography", Anal. Chem. 60,514(1988).
20. D. B. Whigan and A. Schuster, Unpublished Work.
21. S. M. Riseman, Unpublished Work.

22. K. M. Varrichio, Unpublished Work.
23. S . Srivastava, Unpublished Work.
24. K. Kumar and C. A. Chang, Unpublished Work.
25. M. F. Tweedle, J.J. Hagan, K. Kumar, S . Mantha, and C. A. Chang,

Reactions of Gadolinium Chelates in the Presence of Endogenously
Available Ions”, Mag. Res. Imag. 9,409 (1991).
26. P. W. Wedeking, K. Kumar, and M. F. Tweedle, “Dissociation of

Gadolinium Chelates in Mice: Relationship to Chemical
Characterstics”, Mag. Res. Imag. 10, 641 (1992).
27. M. F. Tweedle, “Physicochemical Properties of Gadoteridol and Other

Magnetic Resonance Imaging Contrast Agents”, Invest. Radiology, 27,
S2, (1992)
28. Z. Ugwuneri and K. Kumar, Unpublished Report.

GUAR GUM
Karen Yu, David Wong,
Jagdish Parasrampuria, and David Friend
Cibus Pharmaceutical, Inc.
200 D Twin Dolphin Drive
ANALYTICAL PROFILES OF DRUG SUBSTANCES 243 Copyright 0 1996 by Academic Press. Inc.
244 KAREN YU ET AL.
1. INTRODUCTION
1.1 History, Sources, and Manufacturing
1.2 Chemistry and Grades
2. DESCRIPTION
2.2 Appearance, Color, Odor, and Taste
3.1 Solubility
3.2 Surface Properties
3.3 Hydration
3.4 Viscosity
3.5 Crystallographic Characteristics
3.6 Effect on the Crystallization of Water
3.7 Films
4. METHOD OF ANALYSIS
4.1. Colorimetric Methods
4.2 Nuclear Magnetic Resonance
4.3 Chromatography
4.3.2 Gas Chromatogaphy
5. STABILITY
5. I Thermal Degradation
5.2 Flow Degradation
5.3 Degradation by Irradiation
5.4 Electrolytes
5.5 Biodegradation
5.5.1 Microbial Growth
5.5.2 Colonic
6. APPLICATIONS
6.1 Pharmaceutical
6.2 Chromatographic
6.3 Food Industry
7. PHARMACOLOGICAL EFFECT AND TOXICITY

GUAR GUM 245
1. INTRODUCTION
1.1 History, Sources, and Manufacturing
Guar gum is derived from the ground endosperm of the guar plant,
Cyamopsis tetragonolobus, of the legumnosae family. This hardy,
drought resistant plant has been cultivated for human and animal
consumption in India and Pakistan, and was introduced to the semiarid
regions of the southwestern United States in the early part of this century.
The component which imparts its physical characteristicsis the
galactomannan content. Galactomannans are water-soluble
polysaccharides common to several types of legumes, including carob,
guar, lotus pedunculatus, lucerne (Medicagosativa var. wairau), red
clover (Trifoliumpratense var. montgomery),Sophorajaponica, and
soybean (Glycine m a ) [l]. The use of carob goes as far back as the
ancient Egyptians who prepared strips with which they bound their
mummies using carob paste. Guar itself is derived from the Indian
language Gua-ahar where ‘gau’ means cow and ‘ahar’ means food [2].
The commercial use of guar gum emerged as World War I1 brought about
a shortage of carob, also known as locust bean gum, to the textile and
paper industries in the United States.
In commercial processing of guar gum, a variety of methods are utilized to

efficiently separate the guar endosperm from the husk and germ. Methods
may include multistage grinding, flame treatment, or wet-milling. The
separated endosperm is ground to finer particle sizes and sold as guar gum.
Variations in processing techniques affect the viscosity, rate of hydration
and dispersion properties of the guar product.
Guar gum was first put to use in the paper and textile industries. It served
to improve paper strength and formation during bonding in the paper
industry, and as a thickener in textiles. Since its approval as a food
additive, this low-cost polysaccharide thickener has been extensively used
in the food industry. Aside from thickening, guar gum is also used as a
viscosity modifier, a binder of free water, a suspending agent stabilizer
and a dietary fiber source. It can be found in such food products as ice
cream, processed cheeses, cake mixes, salad dressings, meat products, and
pet foods. In the pharmaceutical and cosmetic industries, guar gum has
been utilized as a disintegrant and binder in tablet formulations, and as a
thickener in lotions and creams. Other areas of industrial applications
246 KARENYUETAL.
include mining, tobacco, explosives, water treatment, and petroleum

drilling.
1.2 Chemistry and Grades
Guar gum is a galactomannan consisting of one galactose on every other

mannose unit, yielding a ratio of approximately 2: 1 (mannose:galactose).
As a natural product, guar gum contains a variety of impurities including:
moisture, protein, acid insoluble residue, ether extractable fat, and ash.
Typically, guar gum may contain 80% galactomannan, 12% water, 5%
protein, 2% acid insoluble residue or crude fiber, 0.7% ash, 0.7% fat, a
trace of heavy metals, zero arsenic, and zero lead. The amounts can differ
from crop to crop.
A variety of guar gum grades are available from several manufacturers and
suppliers. Each supplier markets their products by certain brand names,
such as Supercol' (from Aqualon) or Jaguar@(from Rhone-Poulenc).
Guar gum may be produced as pharmaceutical, food or industrial grades,
and may range in particle size from coarse to fine.
2. DESCRIPTION
Guar gum is a polysaccharide consisting of a straight chain of D-

mannopyranose units joined by p(1+4) linkages with a side-branching
unit of a single D-galactopyranose unit joined to every other mannose unit
by a-(1--A) linkages. The structure of guar g u m is shown in Figure 1. The
side chain of galactose accounts for its cold water hydration as well as its
hydrogen-bonding activity. It hydrates and swells rapidly in cold water
forming a viscous colloidal dispersion or sol. Manufacturers of guar gum
products state that the molecular weight of guar gum, on average, can
range from 1 to 2 million. The literature has not come to the same
conclusion, with some stating that the molecular weight falls around
250,000, while others report it to be in the millions [3]. A problem that
arises is the lack of a sufficient standard for use in molecular weight
GUAR GUM 247
determination. Additionally, the molecular weights can vary from crop to

crop, season to season.
Figure 1. Segment of a guar gum molecule. The

polysaccharide backbone is composed of D-
mannopyranose units, while the pendant groups are
D-galactopyranose units
2.2 Appearance, Color, Odor, and Taste
Guar gum is obtained as a granular powder, which can be off-white to very

lightly yellow or green in color It is nearly odorless with a bland taste.
3.1 Solubility
Guar gum will disperse and swell almost completely in cold or hot water
to form a viscous sol or gel. It is slightly soluble in water, but not in
organic solvents. A variety of organic compounds are often used in the
purification of guar, particularly when required for analytical purposes
[4,51.
248 KAREN YU ET AL.
3.2 Surface Properties
Both crude and purified guar gums are surface active, and have been
shown to reduce the surface tension of water to 55 Nlm [ 6 ] . Since both
crude and pure guar differ mainly in the protein content, the surface
activity must be due to a non-protein component. The effect of
concentration on the surface tension of water is illustrated in Figure 2.
Galactomannans can be used as emulsifiers as well as stabilizers, where
they can contribute to steric stabilization and to the depletion stabilization
of oil-in-water emulsions.
3.3 Hydration
The hydration rate and optimum viscosity of guar is strongly affected by

the galactomannan content, the size and structure of the polymer
molecules, and the particle size distribution [7]. The rate of hydration is
also affected by the presence of other hydrophilic substances, the
availability of water, variations in temperature and pH, and mechanical
agitation. The maximum hydration rate of guar takes place at pH 8.0,
while the minimum is noted at 3.5. Guar solutions are usually stable over
a pH range of 3.5 to 9.0, but other parameters such as temperature, ionic
strength, and additional solutes may effect the overall stability [2].
Following ingestion, the type of meal eaten, the degree of mixing with
food, and the degree of pre- and postprandial hydration can strongly affect
the viscosity development in the stomach and small intestine [7].
3.4 Viscosity
The viscosity of aqueous solutions of guar gum is much higher than that of
many common water-soluble polymers [8-1 11. It shows very pronounced
shear thinning [7], namely a decrease in solution viscosity with an
increased rate of movement. Solutions of guar are non-Newtonian in
nature, exhibiting pseudoplastic behavior [ 121, where the viscosity varies
inversely with temperature [2]. Hydrodynamic properties of various
galactomannans have been reported by Sharman, et al. [ 13, who made
measurements of viscosity as a function of shear rate, sedimentation, and
diffusion coefficient on aqueous solutions of galactomannans. Average
molecular weight calculations were made from sedimentation and
GUAR GUM 249
70
60
50
40
30
0.01 0.10
Log C (grldl)
Figure 2. Surface tension of locust bean gum (LBG) and guar gum
solution at 25°C.
250 KAREN YU ET AL.
diffusion coefficients using the Svedberg equation. The Mark-Houwink

relationship between intrinsic viscosity and average molecular weight
were found to hold for the galactomannans despite differences in
mannose-to-galactose ratios. The dependence of intrinsic viscosity on
average molecular weight is shown in Figure 3.
Anincrease in the polymer concentration yields an increase in the

viscosity of a solution to a greater extent than it decreases the diffusivity
[ 131. The viscosity can also increase with increasing molecular weights of
guar gum. Diffusion of bovine serum albumin was shown to decrease in
the presence of increasing molecular weights of guar gum [14]. The
viscosity of a guar solution can be altered to that of a gel-like structure by
the addition of other gums (ie.,locust bean gum and carrageenan) [ 151. It
has been recommended that when the replacement of guar gum is required
in a formulation, a proportionate amount should be used based upon
viscosity rather than on a dry weight to weight replacement [16].
3.5 Crystallographic Characteristics
X-ray diffraction has been used to study of the chain conformation and
three-dimensional packing schemes of galactomannans [ 171. The unit
cells of guar gum are orthorhombic, consisting of two chain segments of
opposite polarity per unit cell. X-Ray diffraction data has also shown that
the Q dimension of the unit cell is highly sensitive to relative humidity and
galactose substitution [ 181. The measured dZoospacing corresponds to the
crystallographic plane in which cohesion is maintained when water of
hydration is added. Figure 4 demonstrates the change in the d,,, spacings
of three galactomannans as a h c t i o n of relative humidity.
3.6 Effect on the Crystallization of Water
Guar has been shown to retard the rate of crystallization of water at -3°C
[ 191. This particular characteristic forms the basis of the use of guar gum
in the ice cream industry.
3.7 Films
Solutions of guar gum have been cast to form films which are brittle in
nature [20].
GUAR GUM 25 1
1.5
1 .o
0.5
0.0
I I I I I I
).5 4.0 4.5 5.0 5.5 6.0
log (b)
Figure 3. The dependence of intrinsic viscosity, [q], on the

average molecular weight, Mww.
252 KAREN YU ET AL.
17
16
15
-
5.
14
.-m
0
C
m
(r
13
rn
12
TG
11
10 I I 1 I I I I I 1
10 20 30 40 50 60 70 80 90
Relative Humidity (%)
Figure 4. Change in the dzoospacing of guar gum (GG), locust

bean gum (LBG), and tara gum(TG) as a function of
relative humidity.
GUAR GUM 253
4. METHOD OF ANALYSIS
4.1 Colorimetric Methods
Methods for determination of galactomannan content typically involve an

extraction and purification step, followed by a gravimetric or colorimetric
analysis. The crude gum is not homogenous, but rather consists of a
mixture of galactomannansvarying in the mannose/galactoseratio and in
the molecular weight [4]. Different extraction techniques will often result
in fractions of varying galactomannan content, where the amount of
galactomannan can be determined by a phenol-sulfuric acid assay [21].
Identification of individual sugars can be carried out by incubation with
0.5MH2S04for 8 hours at 100°C, followed by neutralization and analysis
via gas-liquid chromatography.
4.2 Nuclear Magnetic Resonance
Pulsed field gradient spin-echo NMR was used to study the diffusional
properties of water in guar solutions [22]. The diffusion coeficient of
water in a non-gelling guar system was found to be independent of the
polymer concentration and the same as that in water alone. Table 1 gives
diffusion coefficients measured in guar solutions. NMR has also been
used to study the interaction of guar and borate ions, a system
characterized by the formation of strong gels [23,24].
4.3 Chromatography
HPLC has been used to separate the galactose and manose

components of guar seeds (4). A representative chromatograph is given in
Figure 5. Another application of chromatography is that of size exclusion
to determine the molecular weight of guar (3). Galactomannansbehave as
a homologous series differing only in molecular weight (1). HPLC has
also been used to evaluate the interaction of guar gum with food
components such as ascorbic acid, niacin, caffeine and phenylalanine (25).
H-bonding was found to occur with the -OH, -NH, or -COOH groups of
all these compounds except caffeine, resulting in an exothermic absorption
254 KAREN YU ET AL.
Figure 5 . Separation of galactose, mannose and myo-inositol

(internal standard) by HPLC. The sample consisted
of guar seed with free sugars removed.
GUAR GUM 255
process. Caffeine interacts in an endothermic absorption proposed to be

related to hydrophobic and or electrostatic forces.
Table 1
Diffusion Coefficients Measured in Guar Solutions
Guar Concentration I Diffusion Coefficient

(% w/w) m2 s-')
2.5 2.34
3.0 2.34
3.5 2.42
4.0 2.27
4.5 2.39
5.0 2.23
4.3.2 Gas Chromatogaphy
Gas chromatography has been used to determine the polysaccharides in

gums (including guar gum), found in food products after hydrolysis and
derivatization [26]. Foods such as dairy products, salad dressings, and
meat sauces have been tested. The process involves hydrolysis with
trifluoroacetic acid and conversion of the products to an aldonitrile acetate
derivative. Monosaccharide patterns of hydrolyzed gums are shown in
Figure 6.
5. STABILITY
5.1 Thermal Degradation
As shown in Figure 7, Arrhenius plots of guar data at different

concentrations show different rate constants, with the rate constants of
256 KAREN YU ET AL.
I
P
Y
0
I
I0
I
20
. I
30
I
40
I
I0
I
20
I
30
i;40
I
Figure 6 . Typical monosaccharide patterns, obtained from the

GC analysis of hydrolyzed gums. Xylitol was used
as the internal standard.
GUAR GUM 251
-1 3
-14
k
-*
0
-15
C
\I 0 0
-1 6
\
-1 7
2.6 2.7 2.8 2.9

103 x in( ~ - 1 )
Figure 7. Arrhenius plots for guar gum data: filled

square=0.9%; filled circle=0.8%; open
circle=0.7%; open triangle=0.6% and filled
trangle=O.5 %.
258 KAREN YU ET AL.
guar gum depolymerization being dependent on galactomannan

concentration. The temperature dependence has been shown to become
weaker with increasing gaiactomannan concentration [271. In addition, the
viscosity of guar solutions will vary inversely with temperature [2].
5.2 Flow Degradation
Shearing of guar solutions causes depolymerization to take place, leading

to a decrease in the intrinsic viscosity and in the molecular weight [27].
This effect is illustrated in Figure 8. Depolymerization will also increase
with growing wall shear stress. Table 2 gives molecular weight and
intrinsic viscosity data for heat treated guar, highly and lightly sheared
guar, and untreated guar 1273.
Table 2.
Mofecular weight and intrinsic viscosity data
Molecular Weight Intrinsic Viscosity

(dmol) (dlk)
Native Guar 700,000 11.51
Heat Treated Guar 160,000 5.87
(0.5%)
Highly Sheared Guar 450,000 6.7
(0.8%)
Lightly Sheared Guar 9.4
(0.8%)
5.3 Degradation by Irradiation
Gamma irradiation of guar powder has been shown to cause a decrease in

viscosity of guar solution at room temperature and at 80°C for 1 hour [28].
It was proposed that low doses of irradiation lead to a decrease of polymer
aggregation in solution, while a high dose of irradiation yielded
GUAR GUM 259
10
7
-
b
5!6
i
X
.-
b
W 5
Y
4
a
3
1
u
30 40 50 60 70
Figure 8. The effect of the applied shear stress on the

apparent rate constant of guar depolymerization.
Data are shown for 0.5% (filled squares) and 0.8%
(filled circles).
260 KAREN YU ET AL.
degradation by hydrolysis. Figure 9 illustrates the effects of gamma

irradiation on guar gum viscosity.
5.4 Electrolytes
Because guar is non-ionic, it has good electrolyte compatibility. However,

compounds such as borax, chromium, zirconium, calcium, and aluminum
will precipitate guar [2]. Further applications of the insolubilized guar are
not practical, since the pH and concentrations required are suitable for
human consumption. The stability of a guar solution may be improved by
adding sequestering agents such as citric, tartaric, or othophosphoric acids
at 0.25 to 0.5%.
5.5 Biodegradation
5.5.1 Microbial Growth
Being a natural product, guar gum contains a resident microbial population

that is monitored by manufacturers. At 25"C, the viscosity of guar gum
will decrease significantly over 1 week. Food grade preservatives such as
sodium benzoate, methyl and propyl parabens help to maintain molecular
integrity [2].
5.5.2 Colonic
The viscosity of a guar gum solution is reduced by 75% and the pH will
fall significantly over 40 minutes when incubated with a homongenate of
feces. Short chain fatty acids are generated, along with gases such as
hydrogen, methane, and carbon dioxide [29,30]. Figure 10 shows the
change in viscosity of guar as a function of incubation time [30]. These
findings imply that guar gum can be fermented by fecal bacteria, which
explains the flatulence that can occur after ingestion of a large amount of
guar gum. The related compound xanthan gum (which possesses a similar
p-( 1-+4) linked glucose backbone as guar), is not affected by fecal
enzymes because of dense side chains which prevent access of the
enzymes to its backbone 1301.
GUAR GUM 261
n
u 0.6
-
0.4 -
0.2 -
I
0 f
I
Figure 9. Effect of gamma irradiation on the viscosity of a 1%

guar gum solution: dispersion prepared at room
temperature (open squares) and dispersion prepared
at 80°C (open circle). The values were obtained 1h
after preparation.
262 KAREN YU ET AL.
"1
nr
Control ?Qmirn ZIh Cantml 30im 2lh
inltld tat tat Initial test M
Figure 10. Median initial control value, and intial and final test
values for viscosity of the n=6 polysaccharides.
The dashed line represents the viscosity of water
measured by this system (2.0 mPa.s).
GUAR GUM 263
Heat-sterilized feces do not change the viscosity and pH of guar gum

solutions. If bacteria-free filtrate is added to guar, the viscosity can
decrease drastically, indicating the presence of an extracellular enzyme
that can reduce the viscosity of guar gum [29,313. Table 3 summarizes the
changes in pH and hydrogen concentrations produced upon incubation
with feces at 37°C for 21 hours. Example strains of Bacteroides known to
digest guar gum include B. fiagilis, B. ovatus, B. variabilis, B. uniformis,
B. distasonis, and B. thetaiotaomicron [3 11.
Table 3
Changes in pH and hydrogen concentration produced by guar incubation

with feces at 37OC for 21 hours
PH H, conc.
Components initial final (mL/L)
guar gum alone control 8.60 8.29 0.0
untreated feces control 8.12 6.91 26.8
untreated feces withguargum 8.14 6.18 63.8
filtered feces with guar gum 8.39 8.00 0.0
autoclaved feces with guar gum 8.75 8.60 0.0
(extracted from reference 29)
6. APPLICATIONS
In the pharmaceutical industry, guar gum is used primarily as a

disintegrating and binding agent in compressed tablets. As a disintegrant,
guar gum has been found to be superior to some common disintegrants
such as corn starch, celluloses, algins, and magnesium aluminium silicate
[32]. From studies that have been conducted to evaluate the use of guar, it
has been shown that an increase in guar can lead to decreased
disintegrating ability [33], while elevated temperatures can increase
binding characteristics [34]. Guar gum as a 1.5% solution in wet
264 KARENYU ETAL.
granulation is an effective binder [35], while at 10% in a base formulation

it is a sufficient disintegrant [36]. At 3% and 5%, guar gum was shown to
retard drug release better than other excipients such as acacia, Carbopol
934, sodium carboxymethylcellulose, and Explotab [37]. Although guar
gum has been used as binder, its actual effect on the hardness of tablets is
still unclear [35]. Particle size can affect disintegration, with finer particle
sizes having greater disintegrating capabilities [38]. The reason for the
superiority of guar as a disintegrant may be related to its strong affinity for
water, with the finer grade guar being a more effective disintegrant due to
a better distribution within the tablet. This coincides with the finding that
disintegration correlates with water absorption rates [32]. Detracting from
its usefulness as a disintegrant, however, is the tendency of guar to
decrease the friability of tablets [34]. Guar gum has also been used to
thicken liquid suspensions of over-the-counter antacid products, as well as
various lotions and creams. Some research has also been conducted
regarding the use of guar in gels for ophthalmic and skin infection
preparations [39].
Many natural gums have been considered in sustained release formulations

[40] and as bioadhesives [41]. Guar has been studied during the
development of controlled release dosage forms [42-491, where it has been
shown to increase the dissolution rate of water-insoluble drugs [8,9]. One
study suggested that the release of drug from tablets is governed by the
degree of branching in the guar gum [45]. One aspect of guar usage in
controlled release formulations is its ability to act as a gel-former to retard
drug release from tablets [46,47,49]. However, under certain conditions,
the drug release can be accelerated by increasing the amount of guar gum
[48]. It was found that the gel layer formed can depend proportionally on
the ionic strength [42]. The gel layer formed by guar is not as thick as that
of other water-soluble polymers such as arabic gum, sodium alginate,
carrageenan, locust bean gum, tragacanth, pectin, methylcellulose, and
hydroxpropylcellulose [8,9,11]. For sustained release Solbutamol, guar
gum alone would not retard drug release satisfactorily and a certain
amount of hydroxypropyl methylcellulose (HPMC) had to be added to the
tablet for better results [44]. HPMC in combination with guar gum or
karaya gum exhibited a potential for the two gums as release retarding
material [50]. The dissolution profile (rotating basket assembly at 100
rpm) of isoniazid from karaya gum, guar gum, and hydroxypropyl
methylcellulose is shown in Figure 11. Additionally, Bhalla found that
combinations of HPMC, guar gum and ethylcellulose could provide tablets
GUAR GUM 265
80 -
Y 60-
3
-a
(II
2
.- 40-
-
CI
(II
i
3 20 -
0 2 4 6 8 10 12
Time [h]
Figure 1 1 In vitro release of isoniazid from various

formulations. The polymer used in Formulation A
was guar gum, that of B was Karaya gum, and that
in C was hydroxypropyl methylcellulose. The
drug:polymer ratio was 1:1.5, and 1% talc and 2%
magnesium stearate was added to each formulation.
266 KAREN YU ET AL.
with suitable characteristics and appropriate drug release profiles for

ketoprofen [Sl].
6.2 Chromatographic Applications
Guar gum has been used as an absorbent for enantiomer separation,

cross-linked as a gel-filtration media for biopolymers, and as an ion-
exchanger-base [5]. As an enantiomer separator, guar purification with
methanol and derivatization with quinine has been performed. The pair of
cis -OH groups on each anhydroushexose unit confers strong dual H-
bonding interaction between guar gum and other molecules. This property
provides for weak chiral discrimination and suggests a premise for
controlled release of bioactive materials. The addition of quinine
enhances the resolution as a chiral selector.
6.3 FOOD INDUSTRY
One of the major users of guar gum is the food industry, whereit serves
many purposes [2]. Guar efficiently modifies the behavior of water, acts
as a stabilizer, minimizes friction and aids in processing and palatability of
foods, and is a viscosity aid in the control of crystal size in saturated sugar
solutions. Some of guar’s most common uses in foods are summarized in
Table 4. With frozen desserts, guar imparts smoothness to ice cream by
promoting small ice crystal development during freezing. Guar also
contributes to the body of ice cream, improving the eating quality. In
cottage cheeses, guar gum promotes curd integrity by friction reduction or
lubricity which aids in processing. It also controls free water in finished
products to promote good storage characteristics. As found in cheese
products, guar has numerous applications. In cold packed cheese, guar
eliminates weeping and creates a more uniform texture and flavor by
controlling moisture and migration. The advantages of guar use in pet
foods and meat products are the prevention of migration during storage,
the maintenance of a homogenous state, and a reduced tendency for
product voids in the can. Guar gum also creates a gloss or sheen to pet
foods, as well as facilitates the removal from the can by reducing friction.
In baked goods, guar imparts softness and yields a more moist product,
allowing easier removal from molds and reducing crumbling. In packaged
dry mixes, such as cake mixes, guar thickens and controls viscosity, as
GUAR GUM 261
well as improves mouthfeel, prolongs shelf life due to improved moisture

retention, and permits freezing of finished cakes. Finally, in beverages
and condiments, guar improves mouth feel, facilitates homogeneous
dispersions, has a bland flavor, has appropriate hydration rate an viscosity,
and is cost effective
Table 4
Food Applications of Guar
Products Examples
Dairy ice cream, ice milk, sherbet, ices, low
fat soft serve, milk shakes, cottae
cheese, dressing, processed cheese,
cheese dips
Pet foods dry and canned pet foods
Baked goods cakes, cake icing, cheese cake, pizza
Packaged dry mixes cake mix, salad dressing mix,

breadings, instant soups, instant snacks,
instant cereal
Condiments barbecue sauce, cocktail mix, relishes,

taco sauce
Beverages juices, nectar, syrups
7. PHARMACOLOGICAL EFFECT AND TOXICITY
Guar gum has the ability to reduce the postprandial rise in blood glucose
and lower cholesterol levels, especially in patients with Type I diabetes
[52-601. There is a minimum or threshold amount of guar that is needed
268 KAREN YU ET AL.
for reducing the postprandial blood glucose rise, although larger doses do
not show additional effects [61,62]. The higher the initial cholesterol
level, the greater the reduction in cholesterol [63]. A similar effect of gum
on the absorption of hydroxyproline has also been observed [64].
The exact mechanism for the hypocholesterolemic effect of guar gum is

unknown, but there are three basic explanations. First, guar is believed to
form a extremely viscous intra-luminal gel which act as a barrier to
glucose or cholesterol diffusion, causing a delay in absorption [61,63,65-
701. At the same time, guar gum may reduce the gastric emptying rate
[71-761. Ingestion of 30 g of guar was shown to increase the transit time
by 5 1% in dogs [77]. A second explanation is the formation of a coat or a
protective layer on the epithelial surface of the intestine by guar gum,
slowing down the intestinal bulk phase diffusion [55,78,79]. This
mechanism has also been used to explain the suggested use of guar gum as
an anti-ulcer agent. These first two explanations are commonly accepted
because an increase in elimination of bile acids and steroids in stool is
observed in guar studies [53,64,80,81]. Additionally, the retardation effect
on glucose absorption disappears when guar gum viscosity is destroyed by
hydrolysis [61]. The final explanation involves the induction of the
activity of HMG CoA reductase [82] and hepatic cholesterol-7-a-
hydroxylase [83] which act to reduce serum cholesterol levels.
Various formulations and methods of administration are available

including powders, granules, capsules, biscuits and crispbread
[52,55,60,84-86]. The selection of a formulation is important because it
can affect patient compliance with regards to long-term utilization of guar.
The effective dose of guar needed for a significant reduction in plasma
total cholesterol anci LDL cholesterol is as high as 15 grams daily or 6
grams three times daily [54,55]. For better control of cholesterol in
patients with type IIa hyperlipidemia, guar gum needs to be combined
with a pharmacologic agent [87,88].
In addition to the above pharmacological effects, guar gum has been

shown to reduce fecal mutagenicity. This is because guar gum contains a
short chain fatty acid which produces butyrate. Butyrate is an anti-
proliferative and differentiating agent in cell culture lines [SS].
GUAR GUM 269
Long term use of guar gum is reported to have no adverse effects

[90], but ingestion of large amounts of guar may lead to early satiety,
bloating, flatulence, increased stool frequency [53], and a decrease in its
hypocholesterolemic effect [59]. After guar gum is enzymatically
degraded to low molecular weight galactomannan,it becomes a carbon
source for intestinal bacteria such as Ruminococcus, Bifdobacterium and
Bacteroides [91]. This accounts for the apparent increase in flatulence
upon ingestion [92]. One study showed that intake of 20 grams of guar
gum a day increased the fecal weight by 20%, explained by the increase
water-holding capacity by the gel formation of guar [93]. In addition to
the increased fecal weight, guar gum dilutes large intestinal contents and
can thus speed up transit times [89]. The viscous gel formation of guar
gum may also interfere in the absorption of other pharmacologic agents
and micronutients [84]. Finally, alcohol should be avoided during
ingestion of guar gum, since it is a non-solvent of guar [84].
270 KAREN YU ET AL.
8. REFERENCES
1. W. R. Sharman, E. L. Richards, and G. N. Malcolm, Biopolymers,

17 (1978), 2817-2833.
2. C. T. Herald, in “Food Hydrocolloids”, Vol. I11 (1982), M.

Glickman (ed.), CRC Press Inc., pp. 171-184.
3. H.G. Barth and D. A. Smith, J. Chromatogr., 206 (1981), 410-415.
4. R. W. Hansen, S. M. Bymes, and A. D. Johnson, J. Sci. Food

Agric., 59 (1 992), 4 19-421.
5. R. Mathur, S. Bohra, C. K. Narang, and N. K. Mathur, J. Liq.

Chrornatogr., 15 (1992), 573-584.
6. N. Garti and D. R. Reichman, Food Hydrocoll. 8 (1994), 155-173.
7. P. R. Ellis, E. R. Morris, and A. G. Low, Diabetic Med., 3 (1986),

490-491.
8. T. Takai, K. Takayama, N. Nambu, and T. Nagai, Chem. Pharm.

Bull., 32 (1984), 1936-1941.
9. K. Takayama, N. Nambu, and T. Nagai, Chem. Pharm. Bull., 30

(1982), 673-678.

(1982), 3013-3016.

(1982), 3 70 1-37 10.
12. N. Sarisuta and E. L. Parrott, Drug Dev. Ind. Pharm., 8 (1982),

605-616.
13. N. Sarisuta and E. L. Parrott, Drug Dev. Ind. Pharm., 9 ( I 983),

861-875.
GUAR GUM 27 1
14. P. Roger, C. Hoebler, T. Ficadiere, and P. Colonna, Food

Hydrocoll., 5 (1991), 145-146.
15. E. Costell, M. H. Damasio, L. Izquierdo, and L. Duran, Food

Hydrocoll., 6 (1992), 275-284.
16. I. A. Schlakman and A. J. Bartilucci, Drug Standards, 25 (1957),

149-154.
17. K. Veluraja and E. D. T. Atkins, Carbohydr. Res., 183 (1988),

13 1-134.
18. R. H. Marchessault, A. Buleon, Y. Deslandes, and T. Goto, J.

Colloid Interface Sci., 71 (1979), 375-382.
19. E. R. Budiaman and 0. Fennema, J. Dairy Sci., 70 (1987), 547-

554.
20. P. Parashar, S.C. Datt, A. Jain, and U. D. N. Bajpai, Polymer

Testing, 11 (1992), 185-192.
21. M. Dubois, K. A. Gilles, J. K. Hamilton, P. A. Rebers, and F.

Smith, Anal. Chem., 28 (1956), 350-356.
22. E. Brosia, A. D’ubaldo, and B. Verzegnassi, Cell. Mol. Biol.,40

(1994), 569-573.
23. 0.Noble and F. R. Taravel, Carbohydr. Res., 166 (1987), 1-11.
24. 0. Noble and F. R. Taravel, Carbohydr. Res., 184 (1988), 236-243.
25. A. L. Khurana and C . T. Ho, J. Liq.Chromatog.,13 (1990), 1559-

1569.
26. J. F. Lawrence, and J. R. Iyengar, J. Chromatogr.,350 (1985),

237-244.
27. T. D. Bradley, A. Ball, S. E. Harding , and J. R. Mitchell,

Carbohydrate Polymers, 10 (1989), 205-214.
272 KAREN YU ET AL.
28. K. King and R. Graz, Food Hydrocoll., 6 (1993), 559-569.
29. J. Tomlin, N. W. Read, C. A. Edwards, and B. I. Duerden, Br. J

Nutr., 55 ( 1 986) 48 1-486.
30. J. Tomlin, J. S. Taylor, and N. W. Read, Nutr.Rep. Int., 39 (1989),

121-135.
31. J. Tomlin, C. A. Edwards, B. I. Duerden, and N. W. Read, Proc.

Nutr. SOC.,45 (1986), 105A.
32. A. M. Sakr and H. M. Elsabbagh, Pharm. Ind., 37 (1975), 457-459.
33. E. M. Rudnic., C. T. Rhodes, J. F. Bavitz, and J. B. Schwartz,

Drug Dev. Ind. Pharm., 7 (1981), 347-358.
34. M. 2. Iqbal, M. Amin, and N. A. Muzaffar, J. Pharmacy, 1 (1979),

17-34.
35. L. E. Eatherton, P. E. Platz, and F. P. Cosgrove, Drug Standards,

23 (1955), 42-47.
36. W. Feinstein and A. J. Bartilucci, J. Pharm. Sci., 55 (1966), 332-

334.
37. H. M. Elsabbagh, A. M. Sakr, and S. E. Abd-Elhadi, Pharmazie,

33 ( 1 978), 730-73 1 .
38. A. M. Sakr and H. M. Elasbbagh, Pharm. Ind., 38 (1976), 732-734.
39. N. Udupa and L. Bhat, Drug Dev. Ind. Pharm., 18 (1992), 2197-
2205.
40. S. K. Baveja, K. V. Ranga, and J. Arora, Znd. J. Pharm. Sci., 50

(1987), 89-92.
41. M. A. Longer and J. R. Robinson, Pharmacy Int., May (1986),

114-117.
GUAR GUM 273
42. P. J. Waaler, M. Anderson, C. Graffner, and B. W. Muller, Acta

Pharm. Nord., 4 (1992), 167-170.
43. H. E. Huber, and G. L. Christenson, J Pharm. Sci., 57 (1968), 164-

166.
44. H. L. Bhalla, and Y. D. Sanzgiri, Indian J Pharm. Sci., 49 (1987),

22-25.
45. S. K. Baveja, K. V. Ranga Rao, and J. Arora, Ind. J. Pharm. Sci.,

30B (1991), 133-137.
46. H. L. Bhalla and S . 0. Gulati, Indian Drugs, 24 (1987), 338-342.
47. P. J. Waaler, C. Graffner, and B. W. Muller, Acta Pharm. Nord., 4

(1992), 9-16.
48. P. J. Waaler, K. Arnesen, C. Graffner, and B. W. Muller, Acta

Pharm. Nord., 4 (1992), 291-296.
49. P. J. Waaler, C. Graffner, and B. W. Muller, Int. J Pharm., 87

(1922), 229-237.
50. N. K. Jain, K. Kutkarni, and N. Talwar, Pharmazie, 47 (1992),

277-278.
51. H. L. Bhalla and A. A. Shah, Indian Drugs, 28 (1991), 420-422.
52. A. R. Khan, G. Y. Khan, A. Mitchel, and M. A. Qadeer, Am. J

Clin. Nutr., 34 (198 l), 2446-2449.
53. M. E. McIvor, C. C. Cummings, M. A. Van Duyn, T. A. Leo, S.

Margolis, K. M. Behall, J. E. Michnowski, and A. I. Mendeloff,
Atherosclerosis, 60 (1986), 7- 13.
54. G. A. Spiller, J. W. Farqunar, J. E. Gates, and S . F. Nichols,

Arterio. Throm., 11 (1991), 1204-1208.
55. H. R. Superko, W. L. Haskell, L. Sawrey-Kubicek, and J. W.

Farquhar, Am. J. Cardiol.,62 (1988), 51-55.
214 KAREN YU ET AL.
56. L. A. Simons, S. Gayst, Balasubramaniam, and J. Ruys,

Atherosclerosis, 45 (1982), 101- 108.
57. J. W. Anderson, A. E. Jones, and S. Riddell-Mason, J. Nutr., 124

(1994), 78-83.
58. H. Vorinen-Markkola, M. Sinisalo, and V. A. Koivisto, Am. J.

Clin. Nutr., 56 (1992), 1056-1060.
59. A. Aro, M. Uusitupa, E. Voutilainen, and T. Korhonen, Am. J.

Clin.Nutr., 39 (1984), 9 11-9 16.
60. D. J. A. Jenkins, D. Reynoids, B. Slavin, A. R. Leeds, A. L.

Jenkins, and E. M. Jepson, Am. J. Clin. Nutv., 33 (1980), 575-581.
61. I. Torsdottir, M. Alpsten, H. Anderson, and S. Einarsson, J. Nutr.,

119 (1989), 1925-1931.
62. R. C. Spiller, Pharmac. Ther., 62 (1994), 407-427
63. L. M. Morgan, T. J. Goulder, D. Tsiolakis, V. Marks, and K. G. M.

M. Alberti, Diabetologia, 17 (1979), 85-89.
64. J. M. Gee, N. A. Blackburn., and I. T. Johnson, Br. J. Nutr., 50

(1 983), 2 15-224.
65. I. T. Johnson and J. M. Gee, Gut, 22 (1981), 398.
66. F. G. Robers, A. G. Low, S. Young, H. A. Smith, and P. R. Ellis,

Proc. Nutr. SOC.,50 (1991), 72A.
67. N. A. Blackbum, J. S. Redfern, H. Jarjis, A. M. Holgate, I.

Hannings, J. H. B. Scarpello, I. T. Johnson, andN. W. Read, Clin.
Sci., 66 (1984), 329-336.
68. D. J. A. Jenkins, T. M. S. Wolever, A. R. Leeds, M. A. Gassull, P.

Haisman, J. Dilawari, D. V. Goff, G. L. Metz, and K. G. M. M.
Alberti, British Med. J., 1 (1978), 1392-1394.
GUAR GUM 215
69. T. M. S. Wolever, D. J. A. Jenkins, R. Nineham, and K. G. M. M.

Alberti, Br. J. Nutr., 41 (1979), 505-510.
70. T. K. Ray, K. M. Mansell, L. C. Knight, L. S. Malmud, 0. E.

Owen, and G. Boden, Am. J. Clin. Nutr., 37 (1983), 376-381.
71. V. Tagliaferro, M. Cassader, C. Bozzo, E. Pisu, A. Bruno, S.

Marena, P. Cavallo-Perin, L. Cravero, and G. Pagano, Diabete
Metab., 11 (1985), 380-385.
72. C. J. Leclere, M. Champ, J. Boillot, G. M. Guille, G. Lecannu, C.

Molis, F. Bornet, M. Krempf, J. Delort-Laval, and J. P. Galmiche,
Am. J. Clin. Nutr. (Suppl.), 59 (1994), 7768.
73. E. Harju, Chirurgia Intaliana, 37 (1985), 139-147.
74. S. Holt, R. C. Heading, D. C. Carter, L. F. Prescott, and P. Tothill,

Lancet, 24 (1979), 636-639.
75. S. J. French and N. W. Read, Am. J. Clin. Nutr., 59 (1994), 87-91.
76. H. Kasper, C. Eilles, C. Reiners, and J. Schrezenmeir,

Hepatogastroenterol, 32 (1985),69-71.
77. L. Bueno, F. Praddaude, J. Fioramonti, and Y. Ruckebusch,

Gastroenterology, 80 (198 l), 701-707.
78. C. D. Jensen, G. A. Spiller, J. E. Gates, A. F. Miller, and J. H.

Whittam,J. Am. Coll. Nutr., 12 (1993), 147-154.
79. B. S. Katoch, J. S. Chawla, and A. Rekib, Indian Vet. J., 48 (1971),

142-146.
80. P. Ebeling, H. Yki-Jarvlnen, A. Aro, E. Helve, M. Sinisalo, and V.

A. Koivisto, Am. J. Clin. Nutr., 48 (1988), 98-103.
81. U. Smith and G. Holm, Atherosclerosis, 45 (1982), 1-10.
82. C. Moundras, S. R. Behr, C. Demigne, A. Mezur, and C. Remesy,

J. Nutr., 124 (1994), 2179-2188.
276 KAREN YU ET AL.
83. P. D. Overton, N. Furlonger, J. M. Beety, J. Chakraborty, J. A.

Tredger, and L. M. Morgan, Br. J Nutr., 72 (1994), 385-395.
84. P. A. Todd, P. Benfield, and K. L. Goa, Drugs, 39 (1990), 917-

928.
85. P. R. Ellis, T. Kamalanathan, F. M. Dawoud, R. N. Strange, andT.

P. Coultate, Eur. J. Clin. Nutr., 42 (1988), 425-435.
86. H. S. Fuessl, G. Williams, T. E. Adrian, and S . R. Bloom, Diabetic

Med., 4 (1987), 463-468.
87. D. J. A. Jenkins, A. R. Leeds, B. Slavin, J. Mann, and E. M.

Jepson, Am. J. Clin. Nutr., 32 (1 979), 16-18.
88. H. R. Superko, Curdio. Rev. Rep., 6 (1985), 1253-1265.
89. S. A. Bingham, Proc. Nutr. Soc., 49 (1990), 153-171.
90. M. E. Mclvor, C. C. Cummings, and A. I. Mendeloff, Am. J. Clin.

Nutr., 41 (1985), 891-894.
91. T. Okubo, N. Ishihara, H. Takahashi, T. Fujisawa, M. Kim, T.

Yamamoto, and T. Mitsuoka, Biosci. Biotech. Biochem., 58 (1994),
1364-1369.
92. T. M. S. Wolever, P. T. Wal, P. Spadafora, and P. Robb, Am. J.

Clin. Nutr., 55 (1 992), 7 19-722.
93. H. Takahashi, N. Wako,T. Okubo, N. Ishihara, J. Yamanaka, and

T. Yamamoto, 3: Nutr. Sci. Vitaminol., 40 (1994), 251-259.
MAFENIDE ACETATE
Alekha K. Dash1 and Shankar Saha2
(1) Department of Pharmaceutical and

Administrative Sciences
(2) Department of Biomedical Sciences
Creighton University
Omaha, NE 68178
278 ALEKHA K. DASH AND SHANKAR SAHA
CONTENTS
1. History and Therapeutic Properties
2. Description
2.1 Nomenclature
2.2 Formula, Molecular weight, and Structure
2.3 Elemental'Composition
2.4 Appearance, Color, and Odor
2.5 Pharmaceutical Dosage Form
3. Synthesis
4. Physical properties
4.1 Infiared spectrum
4.2 1H Nuclear Magnetic Resonance Spectrum
4.3 13C Nuclear Magnetic Resonance Spectrum
4.4 Ultraviolet Absorption Spectrum
4.5 MassSpectrum
4.7 Melting Point
4.8 Solubility and Partition coefficient
4.9 X-Ray Powder Diffraction
5.1 IdentificationTests
5.3 Chromatographic Methods
6. Stability
7. Pharmacokinetics
8. Toxicity
9. Acknowledgments
10. References
MAFENIDE ACETATE 279
Mafenide is a synthetic anti-infective, which is closely related to

sulfonamides in its chemical composition. It differs structurally from the
sulfonamides in that it contains a methylene group between the benzene
ring and the amino nitrogen of the basic sulfonamide structure. Mafenide
is normally obtained as the acetate salt.
In 1939, Klarer [ 11 first applied for a German patent for mafenide acetate.
The drug is commercially available in a water-miscible cream formulation.
This sulfonamide is not inactivated by p-aminobenzoic acid, or by pus and
serum. It is effective when applied topically in the prevention and
treatment of bacterial infection associated with second and third-degree
burns. Mafenide acetate has a broad spectrum of antibacterial activity and
is effective against both Gram-positive and Gram-negative organisms, as
well as clostradia [2]. It is also reported to be active against Pseudomona
aeruginosa [3], but has no activity against fungi and viruses.
The action of mafenide is primarily bacteriostatic. Unlike other

sulfonamides, it does not compete with the essential bacterial metabolite,
p-aminobenzoic acid. This suggests that its activity may be due to either
analogous competition with another metabolite, or to an entirely different
mechanism [2]. Even though the exact mechanism of the antibacterial
activity of mafenide is not yet known, Harrison et al. [4] have shown that a
marked inhibition of protein synthesis occurs at the drug concentrations
which are used in the treatment of burn wounds.
2. Description
2.1 Nomenclature
Chemical Name: Benzenesulfonamide, 4-(aminomethyl)-

monoacetate
a-amino-p-toluenesulfonamidemonoacetate
Genekic Name: Mafenide Acetate

Trade Names: 14-Homosulfanilamide, Maphenide,

Marfanil, Mesudrin, Mesudin, Sulfamylon,
Homosulfamine, Ambamide, Neofamid,
Septicid, Emilene, Homonal, Paramenyl.
CAS Registry Number: 13009-99-9
2.2 Formula and Molecular Weight
C ~ H ~ O N ~ O ~ S . C ~(MW
H ~=O246.29)
~
Structure:
CH2NH2
2.3 Elemental Composition
The elemental composition of mafenide acetate, based on the theoretical

composition of C7H1oN202S.C2H402, is calculated as:
C = 43.89% H = 5.68%
N = 11.37% 0 = 25.98%
S = 13.02%
2.4. Appearance, Color, and Odor
Mafenide acetate is obtained as a white, crystalline powder from alcohol.

MAFENIDE ACETATE 28 1
2.5. Pharmaceutical Dosage Form
Mafenide acetate is available only as a cream formulation containing

11.2% w/w of mafenide acetate. The excipients present in this
formulation are methylparaben, propylparaben, sodium bisulfate, and
disodium edetate.
3. Synthesis
Several syntheses of mafenide acetate have been reported in the literature

[5- 101.
Miller et al. have synthesized mafenide acetate starting fromp-

cyanobenzenesulfonamide[5]. This compound was dissolved in ethanolic
hydrochloric acid, and hydrogenated using Pdcharcoal as a catalyst to
yield the hydrochloride salt of mafenide. The hydrochloride salt was then
dissolved in water, and ammonia added. The liberated base was collected,
and reacted with acetic acid to give mafenide acetate. This procedure has
been illustrated in Scheme 1.
Bergeim and Baker [7] have also synthesized mafenide following a

procedure similar to that described by Miller et al. [6].
Angyal and Jenkin [101were able to synthesize mafenide in a four step

procedure, characterized by an overall yield of 11%. Commercially
availablep-toluene sulfonyl chloride was first chlorinated using a stream
of chlorine at 160°C for 10 hours, to yieldp-chloromethylbenzene
sulfonyl chloride in 23% yield. Amination of this chlorinated product in
alcoholic ammonia gave the correspondingp-chloromethyl benzene
sulfonamide in 86% yield. The conversion of this compound to its
quaternary hexaminium salt was carried out in chloroform at 96%yield.
Finally, desaltation of the salt in refluxing alcohol and concentrated
hydrochloric acid gave mafenide an 85% yield from step 3. This
procedure has been illustrated in Scheme 2.
282 ALEKHA K.DASH AND SHANKAR SAHA
Scheme 1
Synthesis of mafnide acetate, according to the procedure of

Miller, Sprague, Kissing, and McBurney [5]
Scheme 2
Synthesis of mafnide acetate, according to the procedure of

Angyal and Jenkin [lo]
CH2C1
23% 86%
SOZCl SO2CI
HC1
______)
EtOH, 85%
4. Physical properties
The infrared spectrum of mafenide acetate was obtained in a potassium

bromide disk (0.5% w/w) using FTIR methodology, and is shown in
Figure 1 [ 1 13. The band assignments are found in Table 1.
4.2. *H Nuclear Magnetic Resonance Spectrum
The 200 mHz proton nuclear magnetic resonance spectrum of mafenide

acetate was obtained at room temperature using deuterated methanol as the
solvent [ 111, and is shown in Figure 2. The chemical shifts, multiplicities,
and the peak assignments of the characteristic proton resonances are
summarized in Table 2 (relative to trimethylsilane), and make use of this
atomic numbering system:
4.3. l3C Nuclear Magnetic Resonance Spectrum
The 200 mHz 1% nuclear magnetic resonance spectrum of mafenide

acetate was obtained in deuterated methanol at room temperature [ 1 11, and
is shown in Figure 3. The chemical shifts and assignments for the carbon
resonances are given in Table 3.
I I 1 I I I I 1
4000 -0 3000 2800 2000 i600 too0 600
WAVENUMBER (cm-
Figure 1. Infrared Spectrum of Mafenide Acetate

Table 1
Infrared Spectral Assignments for Mafenide Acetate
Energy (cm-’) Assignment
3550-3000 N-H stretching (s, br)
1542, 1519 N-H bend (scissoring)
1413 CH, (scissoring)
1319 C-N stretching
1143 Sulfonamide group
1090 N-H wagging
(br) = Broad
(s) = Strong intensity
.................................. I - - -
9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5
-I-"
Figure 2. Proton Magnetic Resonance Spectrum of

Mafenide Acetate in Deuterated Methanol
Table 2
'H NMR Spectral Assignments for Mafenide Acetate
Chemical Shift Multiplicities Number of Assignments

(p.m.) protons
8.0 2 Aromatic protons

(475)
7.55 2 Aromatic protons
(Z6)
4.12 2 Amino protons
connected to
sulfoxide
1.55 S 3 methyl protons
of acetate
d = doublet, s = singlet
Figure 3. I3C NMR Spectrum of Mafenide Acetate
in Deuterated Methanol
ALEKHA K. DASH AND SHANKAR SAHA
Table 3
13CN M R Chemical Shift Assignments for Mafenide Acetate
Chemical shift Assignments

(P.rn.1
143, 140 QuaternaryAromatic Carbons

(1%4)
131 Aromatic Carbons
(33
129 Aromatic Carbons
(2,6)
MAFENIDE ACETATE 29 1
4.4. Ultraviolet Absorption Spectrum
The ultraviolet absorption spectrum of mafenide acetate was obtained at a

concentration of 1 mg/mL in both water and in methanol [1I], and the
resulting spectra are shown in Figure 4. The absorption maxima found
between 220-267 nm demonstrates the presence of an aromatic ring in the
molecule.
The bathochromic shift in absorption maximum noted on passing fiom

methanol to water is consistent with the usual trend where more polar
solvents move n+n* bands to longer wavelengths. The solvent trend
therefore provides a strong assignment for the nature of the observed band
as being associated with the aromatic chromophore.
4.5. Mass Spectrum
Using a gas chromatographic interface, mafenide acetate was introduced

into a Finnigan INCOS-SOB quadrupole mass spectrometer, and the mass
spectrum obtained using electron impact (electron energy of 70 eV and a
source temperature of 18OOC) [111. The spectrum obtained using this
procedure is shown in Figure 5. A (M-59) molecular ion peak (at m/e =
187) was observed, and other fragment assignments are given in Table 4.
4.6. Thermal Analysis
The differential scanning calorimetry @SC) thermogram of mafenide

acetate is shown in Figure 6a. The sample was heated under a nitrogen
purge from 30-250°C in a nonhermetically crimped aluminum pan at a
rate of 10°C/minin a Shimadzu model DSC-50 thermal analysis system.
Thermogravimetric(TG) analysis of mafenide acetate was conducted
using a Shimadzu model TGA-50 system, and the TG thermogram
obtained on a sample heated at a rate of 1O"C/min is shown in Figure 6b.
DSC and TG thermograms of mafenide samples heated at 5"C/min are
shown in Figure 7, while the DSC thermogram of mafenide acetate
heated from 30-250°C in a hermetically crimped pan (at a rate of
10°C/min)is shown in Figure 8.
292 ALEKHA K. DASH A N D SHANKAR SAHA
0)
U
C
m
20
In
n
a
260.0 308.8
Wavelength (nm)
a
U
C
tu
f0
v)
a
a
288.8 3ee .0
Wavelength (nm)
Figure 4. Uttraviolet Spectra of Mafenide Acetate;

(a) in Water and (b) in Methanol
1
1
187
89
122
1 141 ,~, 356
48 68 88 lh 128 148 168 108 288 228 248 268 288 388 328 340It368
p
-
Figure 5. Mass Spectrum of Mafenide Acetate Electron Impact
Table 4
Mass Spectral Data for Mafenide Acetate
Fragment Ion mle
P~(SO~NH~)CH~NHSN+ 187
PhCH2NH 106
PhCH' 89
Ph 77
osc TG4
mY/n - x
1.0 130 30
130.30
' //-- 0330

03.30
0.3
.B3 53
.B3.53
75 30
70.30
-1.0
60.30
63 30
-2.ou jO.3C
I I t I A
3.50 130.30 200.30
Temperature ("C)
Figure 6. (a) Differential scanning calorimetry, and

(b) thermogravimetry thermograms of mafenide
acetate, obtained at a heating rate of 10"Clminute.
296 ALEKHA K. DASH A N D SHANKAR S A W
DSC TGA
10.30
1.30
1.33
1.30
1.90
1.30
3.30 100 .30 200.00
Temperature ("C)
Figure 7. (a) Differential scanning calorimetry, and

(b) thermogravimetry thennograms of mafenide
acetate, obtained at a heating rate of 5"Clminute.
f
1.30.
3.30. ,
-1 00.
-2.90.
-.:.do
4 ‘ 1
I ,
I I , *
I I
Temperature (“C)
Figure 8. Differential scanning calorimetry thermogram of

mafenide acetate, obtained at a heating rate of
10OClminute in a hermetically sealed pan.
The DSC thermogram consisted of two endothermic peaks within the

range of 160-200°C. Thermomicroscopicstudies confirmed that the first
endothermic feature (at 169°C)was due to compound melting. Thermo-
microscopic studies conducted on sample immersed in silicone oil did
not show any liberation of any bubbles, which indicates that any weight
loss which accompanied sample heating was not due to a dehydration
mechanism.
4.7. Melting point
The reported melting range of mafenide acetate is 151- 152"C[121. In the

present work, the melting point of mafenide acetate was observed to take
place at approximately 167°C [I I].
4.8. Solubility and Partition Coefficient
Mafenide acetate has been reported to be freely soluble in water and

methanol [ 131.
Using the partition coefficient computational program produced by

Advanced Chemistry Development, Inc., (Toronto, Canada), it was
deduced that log would be -0.80 k 0.23. This value is consistent with
the observed solubility data where the compound is freely soluble in polar
media, but apparently not soluble in non-polar media.
4.9. X-ray Diffraction
Mafenide acetate was determined to be quite crystalline, yielding the

powder diffraction pattern shown in Figure 9 [ 111. The powder pattern
was obtained using a wide angle x-ray dieactometer (model X D S 2000,
Scintag). The calculated d-spacings for the dieaction patterns are
provided in Table 5 .
I
In
m
0
0
ln
N
0
v
r(
0
n
Figure 9. X-Ray Diffraction Pattern of Mafenide Acetate
ai
Table 5
P wd X-Ray Diffraction Data for Mafenide Acetat ?
Relative
Peak d-spacing Intensity
No (A) (%)
1 5.92 32.8
2 5.67 100
3 4.96 35.3
4 4.87 48.0
5 3.84 44.8
6 3.75 43.8
7 3.68 39.7
8 3.21 27.8
9 3.12 26.8
10 2.77 25.9
11 2.68 58.2
12 2.42 34.3
13 2.38 21.4
14 2.30 76.1
5.1 Identification
5.1.1 Infrared absorption: The infrared spectrum of mafenide acetate,

obtained in a potassium bromide disk, can be used for identification of the
drug substance. The compound will exhibit peaks at 1592,13 16,1299,
1151, 1089, and 897 wavenumbers [141.
5.1.2 Thin-Layer Chromatography: Mafenide acetate can also be

identified by a thin layer chromatographic method, which is conducted
according to the method described in section 5.3.1 of this profile. The Rf
value of the compound under investigation must correspond with that of
the USP standard run under the identical conditions
5.1.3 Color Test: Mafenide acetate produces a violet color reaction

when reacted by the Koppanyi-Zwikker test [ 141. The sample is dissolved
in 1 mL of ethanol. Addition of one drop of a 1% wlv solution of cobalt
nitrate in ethanol, followed by 10 pL of pyrrolidine with slow agitation,
produces the violet color.
Ten mL of a 2 mg/mL mafenide solution in water is transferred into a 100-

mL volumetric flask containing 1 mL of 1N HC1, and diluted to volume
with water. Mafenide acetate reference standard solution (200 pg/mL) is
prepared by dissolving USP reference standard mafenide acetate in 0.01N
HCl. The amount of mafenide acetate in the unknown solution is
quantitated by measuring absorbances of the unknown and standard
solutions at 267 nm [131.
The purity of mafenide in pharmaceutical formulations can be quantitated

by the TLC method described in the USP [13]. A 50 mg/mL mafenide
302, ALEKHA K. DASH AND SHANKAR SAHA
acetate solution is prepared in methanol, and 5 pL of this solution is

applied to a chromatographic plate coated at 0.25 mm with a silica gel
mixture. The chromatogram is developed in a chamber containing a
mixture of ethyl acetate, methanol and isopropyl m i n e (77:20:3 v/v). The
plates are air dried, and examined under short-wavelength ultraviolet light.
A TLC method has also been reported by Harrison et al. [4], and was used
to determine the pharmacokinetics of mafenide acetate in rats. One
microliter of the sample was applied to a plate coated at 250-pm with
silica gel, and separated by ether: isopropylamine:methanol(94:3:3 v/v).
The spots were identified by their fluorescence after excitation at 254 nm.
Steyn has reported the use of a TLC method for the analysis of mafenide
in biological fluids, which uses a small volume of deproteinating serum in
the method [ 151. The supernatant is spotted on the plates, and
fluorescamine is reacted with the spots to produce a more intense
fluorescence. The relative intensities of the unknown fluorescence and
standards fluorescence are used to calculate the concentration of mafenide
in the unknown sample.
Dash and Harrison have reported an ion-pair chromatographic method for

the analysis of mafenide acetate in pharmaceutical formulations [ 161. This
method used a C column, and the detection is based on the UV
absorbance at 270 nm. The mobile phase consists of 65:35 v/v methanol-
phosphate buffer @H 5.0), and also contains 2 mM of 1-heptane sulfonic
acid as an ion-pairing agent. The mobile phase flow rate was maintained
at 1 mL/min.
6. Stability
Commercially available mafenide acetate cream should be stored in tight,

light resistant containers, and kept away from excessive heat [ 131.
7. Pharmacokinetics
Mafenide acetate is unsuitable for systemic antibacterial therapy because

of its rapid inactivation in blood. Harrison et al. have reported the
absorption and metabolism of mafenide acetate in Sprague-Dawleyrats
[4]. The absorption of the drug, when applied to burns, is greatly reduced
by the layer of lipid and protein normally present on the rat skin. Removal
of this layer produced a 16-fold increase in the peak concentration within
one hour. The drug reached the subcutaneous tissue muscle (2-3 mm
from the site of application) within 30 minutes.
Following absorption, mafenide acetate is rapidly metabolized to p -

carboxybenzene sulfonamide, and is excreted in the urine [171. This
metabolite has no antibacterial activity and is a weak carbonic anhydrase
inhibitor. Organ clearance has been reported to be logarithmic in rats,
with a 10 minute half-life for kidney tissue. Approximately 97% of the
drug is excreted in the urine within 24 hours. Mafenide and its metabolite
inhibit carbonic anhydrase, reduce the renal tubular buffering mechanism
in maintaining normal body pH, systemic acidosis, increase bicarbonate
excretion, and retention of chloride ion in the blood. Complete absence of
ammonia in the urine may occur in the case of treated burn patients having
large burn surfaces (usually over 30%). This may also lead to
hyperventilation [181.
8. Toxicity
The intravenous LD50 of mafenide acetate in the mouse is reported to be

1,580 mgkg but is 2,040 m a g in the case ofthe rat. No LD50 for oral
administration has been reported. Monkeys tolerated single oral doses of
20 gkg mafenide cream with no evidence of toxicity. Both in rats and
mice, symptoms of acute intoxication included respiratory depression,
terminal clonic-tonic convulsions, and death due to respiratory arrest
occurred within 1 to 5 minutes [19].
When 10 gikg dose of the cream formulation were applied daily to the
shaved skin of rabbits over a 3 month period, no evidence of irritation was
detected. Hematological studies did not show any abnormalities, and
slight elevation of blood glucose occurred occasionally with higher doses

of mafenide acetate [181.
9. Acknowledgment
The authors thank Mr. William Ihm (Forensic Laboratory, School of

Medicine, Creighton University) for conducting the mass spectroscopic
studies.
10. References
1. J. Klarer, German Patent 726, 386, applied for 27.1 (1939).
2. R.J. Holt, R. Murphy, and P.J. O’Donnel, Br. J. Plast. Surg., 21,
106-110 (1968).
3. R.E.M. Thompson, F.C. Path, E.W. Colley, M.C. Path, and G.J.G.
Chinnock-Jones, Br. J. Plast. Surg., 22,207-209 (1969).
4. H.N. Harrison, H.W. Bales, and F. Jacoby, J. Trauma, 12,986-

993 (1972).
5. E. Miller, J.M. Sprague, L.W. Kissing, and L.F. McBurney, J. Am.

Chem. SOC.,62,2099-2103 (1940).
6. J. Klarer, U.S. Patent 2,288,531 (1942).
7. F.H. Bergeim and W. Baker, J. Am. Chem. SOC.,66, 1455-1460

(1944).
8. M. Kusaim and Z. Yakagaku, J. Pharm. SOC.Japan, 64,51

(1944).
9. Z.F. Komokina, J. Appl. Chem. (USSR), 21,681-684 (1948).

10. S.J. Angyal and S.R. Jenken, Aust. J Sci. Res., 3A, 461-465
(1950).
11. A.K. Dash and S. Saha, unpublished results.
12. The Merck Index, 1 lthEdition, Merck and Co., Inc., Rahway, N.J.,
USA, 1989, p. 5523.
13. The United States Pharmacopeia, 22nd Revision, United States

Pharmaceutical Convention, Rockville, Maryland, 1989, pp. 784-
785.
14. Clarke's Isolation and Identification of Drugs in

Pharmaceutical,Body FIuids, and Post-mortem Material, A.C.
Moffat, ed., 2nd Edition., The Pharmaceutical Press, London,
1988, pp. 716-717.
15. J.M. Steyn, J: Chromatogr., 143,210-213 (1977).
16. A.K. Dash and J.S. Harrision, J. Chromatogr., 708,83-88 (1995).
17. J.A. Moncrief, Clin.Pharmacol. Therap., 10,439-448 (1969).
18. Mafenide Acetate Cream, Drugs, 1,434-460 (1971).
19. T.W. Skulan and J.O. Hope, Life Sci.,5,2279-2282 (1966).

MALTODEXTRIN
Matthew J. Mollan Jr. and Metin Celik
Pharmaceutical Compaction Research Laboratory

College of Pharmacy
Rutgers, The State University of New Jersey
Piscataway, NJ 08855
ANALYTICAL PROFILES OF DRUG SUBSTANCES 307 Copyright 0 19% by Academic Ress, Inc.
AND EXCIF'IENTS-VOLUME 24 All rights of reproduction in any form reserved.
308 MATTHEW J. MOLLAN JR.AND METlN CELIK
CONTENTS
1. Description
1.1 Name, Definition, Formula
1.2 Appearance
1.3 Carbohydrate Profile
2.1 Particle Morphology
2.2 Crystallographic Properties
2.4 Particle Size Distribution
2.5 Surface Area
2.6 Mercury Poroshnetry
2.7 Density
2.8 Moisture
2.9 Powder Flow
2.10 Compaction
2.11 Viscosity
3.1 Cornpendial Tests
4. Identification
4.1 Maltodextrin Saccharide Separations
4.2 Thin Layer Chromatography
4.3 Liquid Chromatography
4.4 Supercritical Fluid Chromatography
4.5 NMK Spectroscopy
5 . Acknowledgments
6. References
MALTODEXTRIN 309
1. Description
1.1 Name
Maltodextrins are also known as hydrolyzed cereal solids, and are starch
conversion products which contain a relatively small amount of dextrose
and maltose. The United States Pharmacopeia and National Formulary
[l] definition of a maltodextrin is: Maltodextrin is a non-sweet, nutritive
saccharide mixture of polymers that consist of D-glucose units, with a
Dextrose Equivalent less than 20. It is prepared by the partial hydrolysis
of a food grade starch with suitable acids andor enzymes. It may be
physically modified to improve its physical and functional
characteristics.
The chemical Abstracts identification number for Maltodextrin is AS-

9050-36-6.
Maltodextrin has a general formula of

H( C6H1005 >"-OH
and is composed of D-glucose units linked primarily by a-1-4 bonds.
Maltodextrin is listed as generally recognized as safe (GRAS)for human

consumption under 21CFR 184.1444.
Maltodextrins have both Food Chemical Codex and National Formulary

monographs.
Starch conversion products with a dextrose equivalent values above

twenty are referred to as corn syrup solids. Starch conversion products
with only a trace amount of dextrose are known as dextrins. Starch
conversion products having a dextrose equivalent value not substantially
above twenty, and containing small amounts of dextrose are known as
maltodextrins. Kanig [2] described the dextrose content of a
maltodextrin as less than about 2.4% by weight, and the amount of
maltose as less than about 9.0% by weight.
The term Dextrose Equivalent Value, D.E., is defined as the reducing

value of the hydrolysate material compared to the reducing value of an
equal weight of dextrose expressed as a percent, dry basis.
310 MATTHEW J. MOLLAN JR. AND METlN CELIK
Reducing Value of Hydrolysate Material

D.E. = x 100
Reducing Value of Dextrose
The dextrose equivalent value is an indicator of the degree of

depolymerization of starch, and the D.E. value will influence the
physical properties of the maltodextrin. The higher the D.E. value, then
the greater the extent of starch hydrolysis.
1.2 Appearance
Maltodextrins are white to off white powders or granules. Maltodextrins

are bland, odorless, with a low sweetness level. The materials are often
physically processed to improve their handling characteristics.
Maltodextrins are produced from starch, usually corn. The starch, which
is almost pure carbohydrate, is cooked or pasted to open the granule and
then hydrolyzed. Products can be made by hydrolyzing with acid or
enzymes or with a combination of acid and enzymes. After the desired
amount of hydrolysis has occurred, the reaction is stopped, the product is
filtered to remove insoluble materials, then dried.[3]. The average
molecular weight decreases as the dextrose equivalent value of the
maltodextrin increases, but even at low D.E. values, it is much smaller
than the original starch. This relative molecular weight difference
between starch and the hydrolyzed sugars gives the maltodextrins a
portion of their valuable functional properties for the food and
pharmaceutical industry.
1.3 Carbohydrate Profile
Maltodextrins will have different carbohydrate profiles depending on

their dextrose equivalent value. The carbohydrate profile of a
maltodextrin has important effects for the physicochemical properties of
the maltodextrin. For example, the low molecular weight components
will influence sweetness, viscosity, and humectant properties, while the
MALTODEXTRIN 31 1
high molecular weight components will influence solubility and solution

stability.
Maltrin@M150 [4] maltodextrin

Standard Specifications
Dextrose Equivalent 13.0 - 17.0
Carbohydrate Profile (Dry Basis)
Monosaccharides 0.7%
Disaccharides 4.5%
Trisaccharides 6.6%
Tetrasaccharides 5.3%
Pentasaccharides and Above 82.9%
Maltrin@M5 10 [4]maltodextrin
Dextrose Equivalent 9.0 - 12.0
Disaccharides 2.9%
Trisaccharides 4.4%
Pentasaccharides and Above 88. I %
Maltodextrin [5]
Dextrose Equivalent 12.1
Disaccharides 2.5%
Trisaccharides 4.0%
Pentasaccharides and Above 89.2%
Maltodextrin is generally used in chewable tablet formulations, however

the modified forms can be used in oral tablet formulations. It is also
used in tablet coating formulations. Some uses of maltodextrins which
312 MATTHEW J. MOLLAN JR. AND METIN CELIK
have been given patent protection are: US patent # 3873694 [2] for use in
direct compression tabletting; South African patent # ZA 800209 A [6]
for use in coating: and South African patent # ZA 5000209 A [7] for use
in coating.
'Tablet formulations containing a significant percentage of high D.E.

value maltodextrin must guard against moisture uptake at high
humidities since the material is hygroscopic. Tablets containing a
significant portion of maltodextrin have extended disintegration times in
water, (greater than five minutes), which may influence dissolution
behavior.
Maltodextrins are used extensively in the food industry as a moisture

conditioner, food plasticizer, crystallization inhibitor, stabilizer, carrier
and bulking agent [8]. The material is often used pharmaceutically in
solid dosage forms as a tablet fillerhinder excipient, but is usually
physically processed by spray drying, fluidized bed agglomeration, and
roller compaction to improve its physical characteristics.
Maltodextrins are generally considered not to be susceptible to undergo

the Maillard reaction, which leads to browning and discoloration. A
paper by Schmidt and Brogmann [9] did find discoloration of ascorbic
acidkodium bicarbonate effervescent tablets containing maltodextrin as a
tablet binder. The found the intensity of discoloration increased with the
degree of degradation of dextrins and maltodextrins, and was
proportional to their dextrose equivalent value.
Maltodextrins have been used as binding agents in wet granulation

processes [lo], and as direct compression binding agents [ l 1, 12, 13, 141
for tablet formulations. They have also been used as stabilizers in solid-
state emulsion systems [ 151.
The physical properties of maltodextrins are determined by their

respective dextrose equivalent value, their saccharide profile, and by the
method by which they were physically processed for use.
MALTODEXTRIN 313
Scanning electron photomicrographs of some maltodextrins are shown in

Figures 1-5. The photomicrographsare at 1OOX, 400X, and 1200X
magnification, from top to bottom, respectively. Figure 1 is of a roller
compacted maltodextrin, (Experimental Maltodextrin, Edward Mendell
Co.), and shows a very dense structure with many small particles
adhering to the surface. Figure 2 is of a spray dried maltodextrin,
(Maltrin@M5 10, Grain Processing Co.), and shows a relatively smooth
surfaced material with a generally round shape. At higher
magnifications, large pores are distinctly seen. Figure 3 is of a fluidized
bed agglomerated maltodextrin, (Maltrin@M500, Grain Processing Co.),
and appears as a very porous mass with many irregularities. Figures 4
and 5 are of the fluidized bed agglomerated maltodextrins, (Malta*Gran
TG@,and Malta*Gran lo@,ZumbroAFP Inc.) and illustrate the smooth
surface and large pores which exist from the fluidized bed agglomeration
method.
The x-ray powder diffraction pattern of the maltodextrins which had

SEM photomicrographstaken are shown in Figures 6-10. The samples
all behaved similarly, and were considered amorphous materials
according to x-ray powder diffraction.
Differential scanning calorimetry thermograms of maltodextrins did not

show any significant events such as glass transitions or melting
endotherms in the temperature range examined, 25OC to 225OC. A slight
endotherm was seen from approximately 6OoCto 110°C due to
desorption of unbound water [161.
Thermogravimetricweight loss for maltodextrins was found to have a

total volatile content with a range of 7.0%-7.4%, and the individual
FIGURE 1: Scanning Electron Photoinicrograph ot
Experimental Maltodextrin
314
FIGURE 2: Scanning Electron Photomicrograph of
Maltrin M510
315
Maltrin M500
316
Malta*Gran TG
317
FlGURE 5: Scanning Electron Photomicrograph of
Malta*Gran 10
MALTODEXTRIN 319
2.00,
i .ao.
1.60.
1.40'
1.20.
1.00-
0.80.
0.60
0.40'
0.20'
0.0 5.0 10.0 t5.0 20.0 25.0 30.0
FIGURE 6 : Powder X-Ray Diffraction Pattern of

Experimental Maltodextrin
1.80.
1.60'
1.40'
I
i .OO'
0.80.
8p%
0.60' /Ifr *w
0.40-
0.20-
.
/ / F
._,..- ,__ - 1 , - - I- , - . I
-
FIGURE 7: Powder X-Ray Diffraction Pattern of
Maltrin M510
320 MATTHEW J. MOLLAN JR. AND METIN CELlK
i.aa.
1 60-
0 a01
i
0 60i

Maltrin M500
1 40

Malta*Gran TG
MALTODEXTRIN 321
x io=
2.00
l._/
1.60'
1.40'
1.20-
0.0 5.0 10.0 15.0 20.0 25.0 90.0

Malta*Gran 10
122 MATTHEW J. MOLLAN JR.AND METIN CELIK
results are shown in Table 1. All samples behaved similarly, and no

bound waters were discernable.
2.4 Particle Size Distribution
The particle size distributions of maltodextrins available vary

considerably, depending on the method of processing the material. Very
fine grades of maltodextrin powder are available, however the particle
sizes of the physically processed maltodextrins are much larger. The
particle size distribution, by sieve analysis, of several physically
processed maltodextrins are shown in Figure 1 1. The fluidized bed
agglomerated maltodextrins had the largest particle size because they are
created from spray dried maltodextrin which is then agglomerated with
water in a fluidized bed process.
2.5 Surface Area
The surface areas of the maltodextrins was recorded by a krypton gas 3-

point BET method [ 171. The surface area analysis results are shown in
Table 2. The spray dried and fluidized bed agglomerated maltodextrins
had similar surface areas, while the roller compacted maltodextrin had a
much larger surface area. These results were consistent with the
observations made by electron microscopy, which illustrated the smooth
surfaces of the spray dried and fluidized bed agglomerated maltodextrins,
while showing the rough surface of the roller compacted maltodextrin.
2.6 Mercury Porosimetry
The measurement of pore volume and distribution of maltodextrins was

performed by the method of Ritter and Drake [ 18). Figures 12 and 13
are from the results of the incremental amount of mercury intruded into
the maltodextrin powder vs pore diameter. Figure 12 shows the fluidized
bed agglomerated maltodextrins to have similar profiles, with
Malta*Gran 1Om having a larger number of pores due to its larger mean
particle size. The spray dried maltodextrin had a smaller mean pore size
than the fluidized bed agglomerated maltodextrins. Figure 13 is the
Total Volitile Content
Maltodextrin Processing Method
from TG (%)
Experimental
roller compaction 7.0
Maltodextrin
Maltrin" M 5 10 spray drying 7.2
fluidized bed
Maltrin" M500 7.4
agglomeration
fluidized bed
Malta'Gran TG" 7.2
agglomeration
fluidized bed
Malta 'Gran 10" 7.0
agglomeration
TABLE 1: Thermogravimetric Weight Loss of the Maltodextrins

323 M A T H E W J. MOLLAN JR. AND METIN CELIK
100
90
0 ,
ir!
30
e
c4 20
10
0 I I ' / ' I ' I ' I I / ' l '
0 100 200 300 400 500 600 700 800
PARTICLE SIZE (microns)
FIGURE 11: Particle Size Distribution of the Maltodextrins

0 = Experimental Maltodextrin
= Maltrin M510
0 = MaltrinM500
r = Malta*GranTG
V = Malta*Gran 10
Surface Area
Maltodextrin Processing Method
(m2/g)
Experimenta I
roller compaction 1.66
Maltodextrin
Maltrin" M 5 1 0 spray drying 0.12
fluidized bed
Maltrin" M 5 0 0 0.12
agglomeration
fluidized bed
Malta*Gran TG" 0.10
agglomeration
fluidized bed
Malta'Gran 10" 0.14
agglomeration
TABLE 2: Surf;ice Area of the Maltodextrins

1
1 0.0
Pore Diameter (um)
FIGURE 12: Mercury Porosimetry

0 = Maltrin M510
0 = Maltrin M500
V = Malta*Gran 10
MALTODEXTRIN 321
T
t" 0.
a
1
V
0
1
U
m
0.
e
0.
.o
Pore Diameter (um)
FIGURE 13: Mercury Porosimetry: log log scale

0= Maltrin M510
0 = Maitrin M500
V = Malta*Gran 10
same data plotted on a log-log scale and illustrates the roller compacted
maltodextrins large number of small pores, which accounts for its large
surface area as compared with spray dried or fluidized bed agglomerated
maltodextrins.
2.7 Density
The densities of the maltodextrins are shown in Table 3 . True density

was determined by helium pycnometry, while tap density (100 taps) and
bulk density were determined in a 100 mL glass cylinder. The densities
of the maltodextrins differ due to their method of processing. This is
especially shown in the "true density" measurements, which does not
measure internal or ''closed" pores, only external or open pores.
2.8 Moisture
The moisture content of the maltodextrins depends on the relative

humidity of the surrounding atmosphere, and maltodextrins can sorb and
desorb significant amounts of water. Figure 14 illustrates the sorption
isotherm of the maltodextrins examined. They exhibited very similar
profiles to each other despite differences in their method of processing.
The maltodextrin powders began to "gel", and formed a pliable mass
when stored at conditions above 75% relative humidity. The United
States Pharmacopeia and National Formulary states that maltodextrin
should be stored in tight containers or well closed containers at a
temperature not exceeding 3OoC and a relative humidity not exceeding
50% [ 13. Hygroscopicity and D.E. value generally correlate, with the
degree of hygroscopicity increasing with the increase in D.E. value.
2.9 Powder Flow
Flow parameters measured for the maltodextrins included: flow through

and orifice, angle of repose, and percent compressibility [19]. The
results are shown in Table 4. The maltodextrins tested were all free
flowing, and the gravimetric method of describing powder flow showed
the greatest difference between the samples due to density
Bulk Density Tap Density True Density
Maltodextrin
(g/ml) (g/mll (g/ml)
Experimental Maltodextrin .57 .65 1.503
Maltrin M510 .50 .56 1.425
Maltrin M500 .27 .32 1.410
I I I
Malta *Gran TG .38 .44 1.424

Malta *Gran 10 .29 .35 1.417
TABLE 3: Densities of the Maltodextrins

50 -
45 -
0 10 20 30 40 50 60 70 80 90 100
ZELATIVE HUMIDITY ( X i
FIGURE 14: Sorption Isotherm of the Maltodextrins

= Maltrin M510
0 = Maltrin M500
v = Malta*Gran TG
V = Malta*Gran 10
Maltodextrin
TABLE 4: Flow Paraineters of the Maltodextrins

considerations, i.e. high density materials generally tend to flow better

than low density materials. Other very fine grade maltodextrin powders
often will not flow, and this one of the primary reasons for physically
processing maltodextrins for pharmaceutical use.
2.10 Compaction
Compression of a maltodextrin powder mass in a die causes the

formation of a compact. Compaction of maltodextrin powder, (with
0.5% magnesium stearate added as a lubricant) into a tablet was
performed at several pressures and the results are shown in Figure 15.
The maltodextrins form relatively strong tablets at modest compressional
forces, and can be considered to have good binding ability.
2.1 1 Viscosity
Maltodextrins exhibit Newtonian viscosity, and decrease in viscosity as

they are heated. They generally exhibit low viscosity at modest solids
content. Kenyon and Anderson [3] presented data of viscosity of
maltodextrin solid solutions at varied percent solids. They stated that
maltodextrins will demonstrate good solubility in the following solids
range: 5 D.E. = 30-45%; 10 D.E. = 45-55%; 15 D.E. = 5045%; and 20
D.E. = 60-75%.
3.1 Compendia1 Tests
According to the United States Pharmacopeia and National Formulary

[ 11, maltodextrin is tested according to/for: microbial limits, pH, loss on
drying, residue on ignition, heavy metals, protein, sulfur dioxide, and
dextrose equivalent.
M ALTODEXTRIN 333
275
250
w 175
0
d
0 150
Lr,
0 125
i5
z 100
2
CA
75
50
25
0
0 50 100 150 200 250 300 350 400 450
APPLIED PRESSURE (ma)
FIGURE 15: Tablet Crushing Force vs Applied Pressure of the

Maltodextrins
0 = Maltrin M510
0 = Maltrin M500
v = Malta*GranTG
V = Malta*Gran 10
334 MAlTHEW J. MOLLAN JR. AND METIN CELIK
. . .
Microbial hmlts. b e t h o d <6 1 >]
It meets the requirements of the tests for absence of Salmonella species
and Escherichia coli.
g& [method <79 1 >]

The pH is between 4.0 and 7.0, in a 1 in 5 solution in carbon dioxide-free
water.
Loss ondrv- i n5. lmethod <73 I>]

Dry at 105°C for 2 hours in a forced-air oven: it loses not more than
6.0%of its weight.
Residue on ignition: [method <281>]

The residue on ignition is not more than 0.5%.
Heavy Metals: [method I1 <23

Not more than 5 ppm heavy metals can be present
Protein:
Transfer about 10 g of Maltodextrin, accurately weighed, to an 800-mL
Kjeldahl flask, and add 10 g of anhydrous potassium sulfate or sodium
sulfate, 300 mg of copper selenite or mercuric oxide, and 60 mL of
sulfuric acid. Gently heat the mixture, keeping the flask inclined at about
a 45' angle, and after frothing has ceased, boil briskly until the solution
has remained clear for about I hour. Cool, add 30 mL of water, mix, and
cool again. Cautiously pour about 75 mL (or enough to make the
mixture strongly alkaline) of sodium hydroxide solution (2 in 5) down
the inside of the flask so that it forms a layer under the acid solution, and
then add a few pieces of granular zinc. Immediately connect the flask to
a distillation apparatus consisting of a Kjeldahl connecting bulb and a
condenser, the delivery tube of which extends well beneath the surface of
an accurately measured excess of 0.1 N sulfuric acid contained in a 50-
mL flask. Gently rotate the contents of the Kjeldahl flask to mix, and
distill until all ammonia has passed into the absorbing acid solution
(about 250 mL of distillate). To the receiving flask add 0.25 mL of
methyl red-methylene blue TS, and titrate the excess acid with 0.1 N
sodium hydroxide. Perform a blank determination, substituting pur.:
sucrose or dextrose for the test specimen, and make any necessary
correction. Each mL of 0.1 N sulfuric acid consumed is equivalent to
MALTODEXTRIN 335
1.401 mg of nitrogen (N). Calculate the percentage of N in the specimen

taken, and then calculate the percentage of protein by multiplying the
percentage of N by 6.25. The limit is 0.1%.
Sulfur dioxide
Hydrogen peroxide solution:
Dilute 30 percent hydrogen peroxide with water to obtain a 3% solution.
Just before use, add 3 drops of methyl red TS, and neutralize to a yellow
endpoint with 0.01 N sodium hydroxide. Do not exceed the endpoint.
Nitrogen:
Use high-purity nitrogen, with a flow regulator that will maintain a flow
of 200 f 10 mL per minute. Guard against the presence of oxygen by
passing the nitrogen through a scrubber, such as alkaline pyrogallol,
prepared as follows. Add 4.5 g of pyrogallol to a gas-washing bottle,
purge the bottle with nitrogen for 3 minutes, and add a solution
containing 85 mL of water and 65 g of potassium hydroxide, while
maintaining an atmosphere of nitrogen in the bottle.
Apparatus:
The apparatus (see Figure I ) is designed to effect the selective transfer of
sulfur dioxide from the specimen in boiling aqueous hydrochloric acid to
the Hydrogen peroxide solution. The backpressure is limited to the
unavoidable pressure due to the height of the Hydrogen peroxide solution
above the tip of the bubbler, F. Keeping the backpressure as low as
possible reduces the likelihood that sulfur dioxide will be lost through
leaks. Preboil vinyl and silicone tubing. Apply a thin film of stopcock
grease to the sealing surfaces of all the joints except the joint between the
separatory funnel and the flask, and clamp the joints to ensure tightness.
The separatory funnel, B, has a capacity of 100 mL or greater. The inlet
adapter, A, with a hose connector provides a means of applying
headpressure over the solution. [Note: A pressure-equalizing dropping
funnel is not recommended because condensate, which may contain
sulfur dioxide, is deposited in the funnel and the side arm.] The round-
boom flask, C, is a 1000-mL flask with three 24/40 tapered joints. The
gas inlet tube, D, is long enough to permit introduction of the nitrogen
within 2.5 cm of the bottom of the flask. The Allihn condenser, E, has a
jacket length of 300 mm. The bubbler, F, (see Figure 11) is fabricated
FIGURE I. Apparatus for Sulfur Dioxide Test

(reproduced with permission of USP)
MALTODEXTRIN 331
from glass according to the dimensions given in Figure 11. The

Hydrogen peroxide solution is contained in a vessel, G, having an inside
diameter of about 2.5 cm and a depth of about 18 cm. Circulate coolant,
such as a mixture of water and methanol (4:1) maintained at 5', to chill
the condenser.
Procedure:
Position the Apparatus in a heating mantle controlled by a power-
regulating device. Add 400 mL of water to the flask. Close the stopcock
of the separatory funnel, and add 90 mL of 4 N hydrochloric acid o the
separatory funnel. Begin the flow of nitrogen at a rate of 200 f 10 mL
per minute. Start the condenser coolant flow. Add 30 mL of Hydrogen
peroxide solution to vessel, G. After 15 minutes, remove the separatory
funnel, and transfer a mixture of 50.0 g of Maltodextrin, accurately
weighed, and 100 mL. of alcohol solution (5 in 100). Apply stopcock
grease to the outer joint of the separatory funnel, return the separatory
funnel to the tapered joint flask, and concomitantly resume the nitrogen
flow. Apply headpressure above the hydrochloric acid solution in the
separatory funnel with a rubber bulb equipped with a valve. Open the
stopcock of the separatory b e 1 to permit the hydrochloric acid solution
to flow into the flask. Continue to maintain sufficient pressure above the
hydrochloric acid solution to force it into the flask. [Note: The stopcock
may be temporarily closed, if necessary, to pump up the pressure.] to
guard against the escape of sulfur dioxide into the separatory funnel,
close the stopcock before the last few mL of hydrochloric acid drain out.
Apply power to the heating mantle sufficient to cause about 85 drops of
reflux per minute. After refluxing for 1.75 hours, remove vessel G, add
3 drops of methyl red TS, and titrate the contents with 0.01 N sodium
hydroxide VS, using a 10-mL burette with an overflow tube and a hose
connection to a carbon dioxide-absorbing tube, to a yellow endpoint that
persists for a least 20 seconds. Perform a blank determination, and make
any necessary correction (see Titrimetry <23 l>). Calculate the quantity,
in pg, of SO, in each g of the Maltodextrin taken by the formula:
in which 32.03 is the milliequivalent weight of sulfur dioxide, V is the

volume, in mL, of titrant consumed, N is the normality of the titrant, and
FIGURE 11. Bubbler (F) for Sulfur Dioxide Apparatus

(reproduced with permission of USP)
MALTODEXTRIN 339
W is the weight, in g, of Maltodextrin taken. Not more than 40 pg of

sulfur dioxide per gram of Maltodextrin can be found (.004%).
Dextrose Eauivalent;
Standard solution:
Dissolve an accurately weighed quantity of USP Dextrose RS in water,
and dilute quantitatively with water to obtain a solution having a known
concentration of about 10 mg per mL.
Test solution:
Transfer about 5 g of Maltodextrin, accurately weighed, with the aid of
hot water to a 100-mL volumetric flask, cool, add water to volume, and
mix.
Procedure:
Transfer 25.0-mL portions of alkaline cupric tartrate TS to each of two
boiling flasks. Bring the contents of one flask to boiling within about 2
minutes while titrating with Standard solution to within 0.5 mL of the
anticipated endpoint. Boil gently for 2 minutes. Continue to boil gently,
add 2 drops of methylene blue solution (1 in loo), and complete the
titration within 1 minute by adding the Standard solution dropwise or in
small increments until the blue color disappears, determined by viewing
against a white background in daylight or under equivalent illumination.
If more than 0.5 mL of the titrant was required after the addition of the
indicator, repeat the titration, adding the necessary volume of titrant
before adding the indicator. Bring the contents of the second flask to
boiling, and similarly titrate with Test solution. Calculate the Dextrose
equivalent, on the dried basis, by the formula:
[ 1OO/( 1 - 0.01A)](C,/C,)(V,N,),
in which A is the percentage Loss on drying of the Maltodextrin taken,

C, is the concentration, in mg per mL, of Maltodextrin in the Test
solution, C, is the concentration, in mg per mL, of USP Dextrose RS in
the Standard solution, and V, and V, are the titrated volumes, in mL, of
the Test solution and the Standard solution, respectively. The Dextrose
equivalent is not more than 20. [Note: This is a limit test. For
maltodextrins with lower reducing values, other procedures may give

other results.]
4. Identification
41 Maltodextrin Saccharide Separations
Maltodextrins can be considered non to weakly UV absorbing

compounds.
Maltodextrins are often characterized by their D.E. values, yet the

physicochemical properties of the maltodextrins are dependent on their
overall saccharide profile present in the final hydrolysate. Various liquid
chromatographic separation techniques for the separation of
oligosaccharides, including maltodextrins, have been recently reviewed
by Chums [20]. Typical phase systems applied are: (a) a reversed
acetonitrile + water gradient on aminopropyl- or cyanopropyl-modified
silica; (b) an aqueous mobile phase on octadecyl-modified silica; (c) an
acidic (PH 5 2) aqueous mobile phase on microparticulate cation-
exchange resins such as an Aminex HPX-22H phase [21]; and (d) anion
exchange chromatography using an alkaline (PH 2 13) aqueous mobile
phase and a sodium acetate gradient [22]. The separation of saccharides
on the basis of chain length ( DP = degree of polymerization) in which
DP 1 is glucose, DP2 is disaccharides, etc., becomes increasingly difficult
above DPlO because of the extremely low levels which are found in
maltodextrins.
4.2 Thin Layer Chromatography (TLC)
Classical TLC is of limited use in attempting to separate and identify

simple saccharides in maltodextrins. Covacevich and Richards in 1976
[23J used thin layer chromatography in the continuous mode at 30 2" *
using 20 x 20 cm precoated plates. The solvent used was n-propanol-
nitromethane-water (50:20:30). The compounds were detected by
spraying the plate with 50% sulfuric acid followed by heating at 110°C
for 30 minutes. They were able to separate isomaltodextrins up to DP9.
MALTODEXTRIN 341
Bosch-Reig et al. [24] used monodimensional TLC to separate a aqueous

solution of maltodextrins in several biological fluids. They used
cellulose plates with two different eluents: (a) n-butanoYethanol/water
(3:2:2), and (b) pyridine/ethyl acetate/acetic acidwater (5:5: 1:3). The
maltodextrins were located with silver nitrate reagent. They were able to
separate maltodextrins from DP2 to DP7 components. More recently,
the use of high performance thin layer chromatography,HPTLC, has
been used successfully in the analysis of maltodextrins. Vajda and Pick
[25] used HPTLC to separate maltodextrin hydrolyzate on HPTLC silica
plate. They used an eluent of 60% acetonitrile: 10% isopropanol; 15%
aqueous ammonia 0.3%: 15% aqueous potassium chloride 0.67M, with a
membrane pressure of 5 bars, flow rate of 0.04 mL/min, and a
development distance of 170 mm. They separated up to DP14, but were
unable to separate glucose and fructose.
4.3 Liquid Chromatography (LC)
High resolution techniques are essential when performing carbohydrate

analysis, since carbohydrates have a number of isomers and homologs
that structurally resemble one another. Since maltodextrins do not
exhibit characteristic absorbances in useful regions of the UV spectrum,
derivatization is often used. Post column labeling has the advantage of
direct injection of intact samples, but has the limitation of sensitivity of
detection due to poor yields of derivatives. Precolumn labeling is often
used because high yields of derivatives can be easily achieved. If an
m i n e type of column is used in HPLC analysis of maltodextrins, then
DPl is the saccharide which will elute first. Increases in chain length
then cause increases in elution time. If a resin based column is used,
then the high molecular weight material, high DP, will elute initially and
the DP1 will elute last.
Scobell, Brobst, and Steele [26] described an automated liquid

chromatographic system for quantitative analysis of carbohydrate
mixtures. In 1981, Scobell and Brobst [27] discussed problems in
separating oligosaccharides, and specifically when using cation-exchange
resins that the loss of resolution for higher oligosaccharides is probably
due to the unique helical structure of a-1-4 linked corn-derived
oligosaccharides. They also used clean-up procedures prior to analysis,
since salts, acids, soluble proteins, and particulate matter will interfere
with the chromatographic analysis. They found that the use of silver
form resins provided superior analysis to that of equivalent resins in the
calcium form.
Cheetham, Sirimanne, and Day 2281 used a C,,-bonded silica column

with water as an eluent at a flow rate of 1 .O mL/min. They were able to
separate maltodextrins to a DP6, although pairing was seen above DP3.
The pairs of peaks was attributed to the a and b anomers of the
oligosaccharide. They then performed a sodium borohydride reduction
which replaced each pair of peaks with a single peak. This peak was
then taken to be due to the corresponding oligosaccharide alditol.
Warthesen [29] separated maltodextrins by an HPLC column (HPX-42A

from Bio-Rad laboratories) with two precolumns, cation exchanging and
anion exchanging, and used water as the mobile phase. Detection was by
differential refractive index (RI) detector. With a flow rate of 0.5
ml/min, the chromatographic analysis was completed for a maltodextrin
in 21 minutes and gave resolution of DPl to DPlO. The lowest
detectable level for each saccharide was about 2 pg injected or 0.1%
expressed as a dry weight basis. He overcame the problems of
insolubility of some carbohydrate material, and nonlinearity of the large
molecular weight peak, by using external standards for quantitation
instead of area normalization.
Brooks and Griffin 1301 examined corn syrup solids and maltodextrins
and characterized the water soluble saccharides. The DP 1 - 10 saccharide
components were separated by using a plastic cartridge C18 Resolve
column compressed in a radial compression module (Waters Associates)
with water as the mobile phase and detection by a differential
refractometer. Pairs of peaks were obtained for most components
between DP3 and DPIO. The peaks were attributed to the a and p
anomers of the saccharides. An external glucose standard was used to
obtain percent composition. They also determined overall molecular
weight profiles by High Performance Size Exclusion Chromatography
(HPSEC) with E-HighA and E-500 pBondage1 silica gel permeation
columns (Waters Associates) connected in series. Water containing
0.15M NaCl was used as the mobile phase. The HPSEC data showed
MALTODEXTRIN 343
that as the D.E. of the hydrolysate increased, the amount of soluble high
molecular weight saccharides decreased.
McGinnis, Prince and Lowrimore [313 used reverse-phase HPLC with a

refractive index detector. A reverse phase column was used since it has
wide availability, high capacity, and can use water as an eluent.
Adjustment of the polarity of the carbohydrate was needed since their is
very little interaction between the carbohydrate and the column. They
found the best column for separation of a mixture of maltodextrins was
with a ODs-2 packed with CIS(Whatman c-18). The retention times on
the column depended on the type of sugar unit, linkage, the anomeric
configuration, and the molecular weight.
Honda et al. [32] recently found that 1-phenyl-3-methyl-5-pyrazolone

reacts with reducing carbohydrates almost quantitatively under mild
conditions to yield strongly UV-absorbing and electrochemically
sensitive derivatives. A homologous series of isomaltodextrins was
separated with precolumn labeling.
Niessen et al. [33] in 1992 used on-line liquid chromatography/mass

spectrometry (LC/MS) for the analysis of maltodextrins, and could detect
oligomers up to a DP 10. Their methodology used a mobile phase of 50
mmol/L ammonium acetate, with gradient elution with 0-50% methanol
and the use of a octadecyl column. Peak splitting occurred at higher DP
values due to separation of the two anomeric forms of the sugar.
LC/thermosprayMS analysis of maltodextrin was performed with 1O4 M
aqueous sodium acetate as the mobile phase and 20 mg injection with an
octadecyl column. The higher DP value oligomers were more difficult to
detect due to the fact that the weight percentage decreases with
increasing DP value.
4.4 Supercritical Fluid Chromatography (SFC)
The new technique of supercritical fluid chromatography (SFC) uses a

highly compressed gas as the mobile phase, with carbon dioxide as the
most widely used mobile phase in SFC. Since carbon dioxide is non-
polar, polar solutes require derivatization to enhance miscibility. Lafosse
et al. [34] in 1992 used an evaporative light scattering detector (ELSD)
with both HPLC (to separate maltodextrin components) and SFC (to
separate sugars). Maltodextrin HPLC analysis was performed by
separation on octadecyl-bonded silica with a water-methanol gradient
and detection by ELSD.
4.5 N M R Spectrametry
The use of water-elimination Fourier transform nuclear magnetic

resonance to determine the degree of derivatization of maltodextrin with
acrylic acid glycidyl ester at alkaline pH was described by Lepisto, et al.
[3 51. They used a JEOL JNM-FX 100 FT-NMR spectrometer at 100
MHz and were able to determine the number of double bonds in
acryloylated maltodextrins.
Radosta ana Schierbaum [36]used NMR relaxation times (both TI and

T2) for starch polysaccharides in solution to characterize bound and free
water. German et al. 1371used Proton Pulse NMR to study water
mobility in solutions of maltodextrins and on the sol-gel transition mode.
They found that maltodextrin gels have a micro-inhomogeneous
structure, and were able to describe the maltodextrin gelation process by
percolation theory.
McIntyre and Vogel [38] used two dimensional NMR to obtain the
complete assignment of the overlapping proton NMR spectrum of
starches, as well as maltodextrins, in D 2 0 solutions at 25'C. Mora-
Gutierrez and Baianu [39] used solid-state I3CNMR techniques, and
found significant differences between the spectra of corn and potato
maltodextrins.
5. Acknowledgments
We would like to acknowledge A. Newman, G. Young, T. Cortina, and

H. Brittain of Bristol-Myers Squibb for their work for the TG, x-ray
powder diffraction, and surface area data. We would also like to thank
the Edward Mendell Company for their generous support of (MM)
throughout the research project.
MALTODEXTRIN 345
6. References
1) United States Pharmacopeia and National Formulary, NF XVII, Sixth

Supplement, Rockville, MD, pp. 2962-2963 (1 994).
2) Kanig, J., "Direct Compression Tabletting Composition and

Pharmaceutical Tablet Produced There From", U.S. patent
3873694, March 25,1975.
3) Kenyon, M.M. and Anderson, R.J., "Maltodextrins and Low-

Dextrose-Equivalence Corn Syrup Solids", Chapter two in ACS
Symp. Ser., m, pp. 7-1 1 (1988).
4) Maltrin@Maltodextrin, Product Data Technical Bulletin, Grain

Processing Corporation, Muscatine, Iowa.
5 ) Wartman, A.; Hagberg, C.; and Eliason, M., "Refractive Index-Dry

Substance Relationships for Commercial Corn Syrups", J. Chem.
Eng. Data, 2, 459-468 (1976).
6) Porter, S.C., and Woulicki, E.J., "Maltodextrin Coating", South

African patent ZA8S00209A, August 28,1985.
7) Porter, S.C., and Waznicki, E.J., "Maltodextrin Coating", South

African patent ZA5000209A, August 28,1985.
8) Lloyd, N.E., and Nelson, W.J., Starch: Chemistry and Tec-

Second Edition, R.J. Whistler, eds., Academic Press, London,
p. 611 (1984).
9) Schmidt, P., and Brogman, B., "Effervescent Tablets: Choice of a

New Binder for Ascorbic Acid", Acta Pharm. Technol., 34. 22-
26, (1988).
10) Visavarungroj, N., and Remon, J.P., "Evaluation of Maltodextrin as

a Binding Agent", Drug Dev. Ind. Pharm., 18,1691-1700
(1992).
11 ) Parrott, E.L., "Comparative Evaluation of a New Direct

Compression Excipient, Soludex 15", Drug Dev. Ind. Pharm.,
B, 561-583 (1989).
12) Papadimitriou, E.; Efentakis, M.; and Choulis, N.H., "Evaluation of

Maltodextrins as Excipients for Direct Compression Tablets and
their Influence on the Rate of Dissolution", Int. J Pharm., 86,
131-136 (1992).
13) Mollan, M.J., and Celik, M., "Characterization of Directly

Compressible Maltodextrins Manufactured by Three Different
Processes", Drug Dev.Ind. Pharm., le , 2335-2358 (1993).
14) Munox-Ruiz,A.; Mondero Perales, M.C.; Velasco Antequera, M.V.,

and Jimenez-Castellanos, M.R., "Physical and Rheological
Properties of Raw Materials", S. T.P. Pharma Sci., I, 307-3 12
(1993).
15) Myers$., and Shively, M.L., "Solid-state Emulsions: The Effects of

Maltodextrin on Microcrystalline Aging", Pharm. Res., 14,
1389-1391 (1993).
16) Mollan, M.J.,

. . of a Roller C o w t e d Maltodextrin
for Direct Corny- * , PhD thesis, Rutgers, The State
University of New Jersey, May 1993.
17) Brunauer, S.; Emmett, P.; and Teller, E., "Adsorption of Gases,
Multimolecular Layers", J. Am. Chem. Soc., 612,309-316
(1938).
18) Ritter, H.C., and Drake, L.C., "Pore-Size Distribution in Porous

Materials. Pressure Porosimetry and Determination of Complete
Macropore Distribution", Ind. Eng. Chem. Anal. Ed. , 12, 782-
786 (1945).
19) Carr, R., "Evaluating Flow Properties of Solids", Chem. Eng.,

a(2), 163-168 (1965).
MALTODEXTRIN 347
20) Churms, S.C., "Recent Developments in the Chromatographic

Analysis of Carbohydrates", J. Chromatography, 5M, 555-583
(1990).
2 1) Hicks, K.B., and Hotchkiss, A.T., "High-Performance Liquid

Chromatography of Plant-Derived Oligosaccharides on a New
Cation-Exchange Resin Stationary Phase: HPX-22H",
JChromatography, 441, 382-386 (1988).
22) Lee, Y .C., "High-Performance Anion-Exchange Chromatography

for Carbohydrate Analysis", Anal. Biochem., .l.@
151-162,
(1990).
23) Covacevich, M.T., and Richards, G.N., Tontinuous Quantitative

Thin-Layer Chromatography of Oligosaccharides", J.
Chromatography, 129, 420-425 (1976).
24) Bosch-Reig, F.; Marcote, M.J.; Minana, M.D.; and Cabello, M.L.,
"Separation and Identification of Sugars and Maltodextrins by
Thin Layer Chromatography: Application to Biological Fluids
and Human Milk", Talanta, 3,1493-1498 (1992).
25) Vajda, J. and Pick, J., "Separation of some mono-, di-, tri-, and
oligosaccharides", 2ndProc. Int. Conf. Biochem. Sep., J. Pick
and J. Vajda, eds., 191-197 (1988).
26) Scobell, H.D.; Brobst, K.M.; and Steele, E.M., "Automated Liquid
Chomatographic System for Analysis of Carbohydrate Mixtures",
Cereal Chem., 54, 905-917 (1977).
27) Scobell, H.D., and Brobst, K.M., "Rapid High-Resolution

Separation of Oligosaccharides on Silver Form Cation Exchange
Resins", J. Chromatography, U ,5 1-64 (198 1).
28) Cheetham, N.W.H.; Sirimanne, P.; and Day, R.W., "High

Performance Liquid Chromatography Separation of Carbohydrate
Oligomers", J. Chromatography, 222,439-444 (198 1).
29) Warthesen, J.J., "Analysis of Saccharides in Low Dextrose

Equivalent Starch Hydrolysates using High Performance Liquid
Chromatography", Cereal Chem., a, 194-195 (1984).
30) Brooks, J.R., and Griffin, V.K., "Saccharide Analysis of

Corn(maize) Syrup Solids and Maltodextrins Using High
Performance Liquid Chromatography", Cereal Chem., a,253-
255 (1987).
3 1) McGinnis, G.D.; Prince, S.; and Lowrimore, J., "The Use of

Reverse-Phase Columns for Separation of Unsubstituted
Carbohydrates", J. Carbohydrate Chem., 5, 83-97 (1 986).
32) Honda, S.; Akao, E.; Suzuki, S.; Okuda, M.; Kakehi, K.; and
Nakamura, J., "High Performance Liquid Chromatography of
Reducing Carbohydrates as Strongly Ultraviolet Absorbing and
Electrochemically Sensitive 3-methyl- 1-phenyl-5-pyrazolone
Derivatives'', Anal. Biochem., .€.@, 35 1-357 (1989).
3 3 ) Niessen, W .M.A.; Van der Hoeven, R.A.M.; Van der Greef, J.,
Schols, H.A., and Voragen, A.G.J., "Online Liquid
Chromatography/Thermospray Mass Spectrometry in the
Analysis of Oligosaccharides", Rapid Commun. Mass Spectrom.,
$, 197-202 (1992).
34) Lafosse, M.; Elfakir, C.; Morin-Allory, L.; and Dreux, M.,
"Advantages of Evaporative Light Scattering Detection in
Pharmaceutical Analysis by High Performance Liquid
Chromatography and Supercritical Fluid Chromatography", J
High Resolution Chromatogr., 15, 3 12-3 18 (1992).
35) Lepisto, M.; Artursson, P.; Edman, P.; Laakso, T.; and Sjoholm, I.,
"Determination of the Degree of Derivatization of Acryloylated
Polysaccharides by Fourier Transform Proton NMR
Spectroscopy", Anal. Biochem., U, 132-135 (1983).
36) Radosta, S. and Schierbaum, F., "Polymer-Water Interaction of

Maltodextrins Part II.", Starch, a, 428-430 (1989).
MALTODEXTRIN 349
37) German, M.L., Blumenfeld, A.C., Yuryev, V.P., and Tolstoguzov,

V.B., "AnNMR Study of Structure Formation in Maltodextrin
Systems", Carbohydrate Polymers, U, 139-146 (1989).
38) McIntyre, D.D. and Vogel, H.J., "Two-Dimensional Nuclear

Magnetic Resonance Studies of Starch and Starch Products",
Starch, 42, 287-293 (1990).
39) Mora-Gutierrez, A. and Baianu, I.C., "Carbon-13 Nuclear Magnetic

Resonance Studies of Chemically Modified Waxy Maize Starch,
Corn Syrups, and Maltodextrins. Comparisons with Potato
Starch and Potato Maltodextrins", J. Agric. Food Chem., 22,
1057-1062 (1991).
NALMEFENE HYDROCHLORIDE
Harry G. Brittain
Ohmeda Inc.
Pharmaceutical Products Division
100 Mountain Avenue
Murray Hill, NJ 07974
ANALYTICALPROFILES OF DRUG SUBSTANCES 351 Copyright 0 1996 by Academic Press, Inc.

352 HARRY G. BRI'ITAIN
CONTENTS
1. Description
1.1 Nomenclature
1.2 Formulae
1.2.1 Empirical
1.2.2 Molecular Weight
1.2.3 S t r u c ~
1.3 Appearance
3.2.1 Polymorphism
3.2.2 X-Ray Powder Diffraction Patterns
3.2.3 Single Crystal Structure
3.3.1 Optical Rotation
3.4 Thermal Methods of analysis
3.4.3 Thermogavimetric Analysis
3.5 Hygroscopicity
3.6.1 Solution pH
3.6.2 Partition Coefficient
3.7 Ionization Constants
3.8 Spectroscopy
3.8.1 UVNIS Spectroscopy
3.8.2 Vibrational Spectroscopy
3.9 Nuclear Magnetic Resonance Spectrometry
3.9.1 'H-NMR
3.9.2 13C-NMR
3.10 Mass Spectrometry
NALMEFENE HYDROCHLORIDE 353
4.1 Identification
4.4 SpectrophotometricMethods of Analysis
4.5 ChromatographicMethods of Analysis
4.6 Determination in Body Fluids and Tissues
5. Stability
5.1 Solid-state Stability
5.2 Solution-Phase Stability
6. Acknowledgements
7. References
354 HARRY G . BRITTAIN
1. Description
1.1 Nomenclature
Chemical Name: 17-(cyclopropylmethyl)-4,5-a-epoxy-6-

methy~enemorphinan-3,14-diol
hydrochloride
6-desoxy -6-meth y lenenaltrexone

TM
Proprietary Names: Revex
Chemical Abstracts Number: 58895-64-0
1.2 Formulae
1.2.1 Empirical: C21H2503N.HCI
1.2.2 Molecular Weight: 375.9
1.2.3 Structural
1.3 Appearance
Nalmefene hydrochloride is obtained as a white to off-white crystalline

powder.
Nalmefene hydrochloride is a pure narcotic antagonist, an analog of

naltrexone, and is structural1 similar to naloxone. It has been developed as
TM
a parented solution (Revex ) for acute reversal of unwanted or excessive
opioid effects, particularly opioid-inducedrespiratory depression.
The need for opioid reversal occurs following opioid-based anesthetic

techniques, or opioid overdoses. Opioid agonists and antagonists are a well
established group of drugs which find extensive clinical use. Morphine and
naloxone are, respectively, the classical opioid agonist and antagonist, with
their activity being mediated through opioid receptors. Opioid receptor
antagonists, such as nalmefene and naloxone, inhibit the pharmacological
effects of administered opioid agonists and also the endogenous opioid
systems. The clinical usefulness of the opioid antagonists comes from their
ability to promptly (and selectively) reverse the effects of opioid agonists,
including central nervous and respiratory system depression.
Nalmefene is an opioid antagonist structurally similar to naloxone. Its mode

of action is thought to be by competitive affinity to opioid receptor sites.
The pharmacological actions of nalmefene are similar to those of naloxone,
but appear to be more potent and have a longer duration of action [1,2].
RevexTMis presented in ampules, having concentrations of either 0.1

mg/mL or 1 mg/mL. Owing to its longer duration of action, the drug is
administered as a single bolus dose administration. This presents a clear
advantage over naloxone, which must be administered as a continuous
intravenous infusion.
As illustrated in Scheme 1, the methods used for the preparation of

nalmefene are based on a Wittig reaction carried out on naltrexone.
In the original synthesis [3], Corey’s modified reagent [4,5] was used to
obtain the 6-methylene derivative of naltrexone. A solution of methylene-
triphenylphosphorane is prepared from sodium hydride and methyltriphenyl-
phosphonium bromide in dimethyl sulfoxide. Naltrexone also dissolved in
356 HARRY G. BRITTAIN
Scheme 1. Literature routes to nalmefene.
NH
Oxycodone Noroxycodone
N-Ccyclopropylmethylnoroxycodone
0
1 0-DemethylaWn
H
?&Noroxymotphone
NH
~
Naltrexone Nalmefene
DMSO is added, and the reaction mixture stirred at 55-60°C under a positive
pressure of nitrogen for 18 hours. Work-up of the product, and
chromatographicpurification, yielded nalmefene free base in 83% yield [3].
Attempts to scale-up the original synthetic method were not successfbl, so

an improved method for the production of nalmefene was developed [6].
Using an ethereal solvent and an alkoxide base to prepare the methylene-
triphenylphosphoranereagent, only slightly in excess of three moles of
methyltriphenylphosphoniumbromide is required, and not the 60 moles as
required by the original process. In addition, the alkoxides employed are
significantly easier and safer to handle than sodium hydride, making it
possible to perform the reaction on a larger scale [6].
Once the nalmefene free base is obtained, it is converted into the

hydrochloride salt. This last step permits the recrystallizationof the product
from water, and results in the formation of a highly purified drug substance.
When obtained via the aqueous crystallization route, the material invariably
consists of a monohydrate crystal phase.
As shown in Figure 1, crystals of nalmefene hydrochloridemonohydrate

form as thin,tabular plates. The shortness of the long axis appears to be a
result of crystal breaking, which yields the plate-like appearance of the
crystals. The largest particles are generally 25-50 pm in diameter, but
typical samples contain crystal debris as small as 10 pm in diameter.
Most large crystals were found to exhibit birefiingence, indicating the

existence of appreciable crystallinity. The thinness of the crystals was such
that only first-order birefiingence could be observed. In addition, the
crystals exhibited parallel extinction, which would place them in either a
orthorhombic or monoclinic crystal system.
358 HARRY G . BRI'TTAIN
Figure 1. Optical microscopy of naimefene hydrochloride, obtained at a

magnification of 200x.
3.2 CrystallographicProperties
3.2.1 Polymorphism
Upon crystallization from water, nalmefene hydrochloride is obtained as the

monohydrate crystal phase. The water of hydration is tenaciously held, and
is maintained through rigorous drying conditions (18 hours at 12OoC over
phosphorous pentoxide).
A stable ethanolic solvate can be obtained fiom samples recrystallized fiom

ethanol. This material is metastable with respect to atmospheric moisture,
and will sorb water upon standing. The ethanolic solvate may be desolvated
under vacuum to yield an anhydrate species, but that species is highly
unstable with respective toward atmospheric moisture and is not readily
handled.
3.2.2 X-Ray Powder Diffraction Patterns
X-ray powder diffraction data were obtained by means of a Philips model

APDl700 automated powder diffractometer, using a copper source. The
patterns were scanned between 10 and 30 degrees 2-8, at a scan rate of 0.015
degrees 2-0 per minute.
All nalmefene hydrochloride materials were found to be very crystalline in

nature, and did not yield a detectable halo due to the presence of amorphous
material, The powder pattern obtained for the monohydrate phase is shown
in Figure 2, while the associated table of scattering angles, d-spacings, and
relative intensities is provided in Table I. The powder pattern of the
ethanolic solvate is shown in Figure 3, and its summary of crystallographic
data is found in Table 11.
3.2.3 Single Crystal Structure
Difficulty was obtained in obtaining a single crystal of the monohydrate

phase whose quality was suitable for diffraction work, but the appropriate
crystals of the ethanol solvate were readily produced. A clear parallelepiped
crystal, having approximate dimensions of 0.25 x 0.20 x 0.40 mm, was
mounted on a glass fiber, and the measurements made on a Rigaku AFC5S
Figure 2. X-ray powder diffraction pattern of the monohydratephase of
nalmefene hydrochloride.
0
m
€ In
N
.Q
N
4 1
.In
rl
.m
l r l l l l 1 1 1 1 1 1 1 d
0 Q Q 0 Q 0
m Q Q 0 0 8
m Q m m Q 0
9 In w m (rJ r(
Table I
CrystallographicData Associated with the X-ray Powder Diffraction Pattern

of the Monohydrate Phase of Nalmefene Hydrochloride
Scattering Angle d-Spacing Relative Intensity

(degrees 2-0) (4 (%I
11.405 7.7714 100.0
11.960 7.3938 4.2
12.505 7.0728 4.4
13.585 6.5 128 6.2
14.090 6.2805 17.0
15.055 5.8801 3.3
15.925 5.5607 4.1
16.355 5.4155 2.3
17.715 5.0027 47.9
18.110 4.8944 16.0
18.585 4.7704 4.3
Figure 3. X-ray powder diffraction pattern of the ethanol solvate of
0 0 0 0 0 m
0 Q Q Q ID 0
Q 8 Q 0 0 Q
9 !n v 0 N rl
Table I1
Crystallographic Data Associated with the X-ray Powder Difiaction Pattern

of the Ethanol Solvate of Nalmefene Hydrochloride
36.4 HARRY G . BRllTAIN
difhctometer with graphite monochromated Cu K a radiation and a 12 kW

rotating anode generator. The data were collected at a temperature of
- 160°C using the w-28 scan technique.
The compound was found to crystallize in the monoclinic crystal system,
and the space group was determined to be P21. The unit cell was
characterized by the following dimensions:
a -
- 7.977 (2) A
b = 17.499 (2) 8,
C = 17.348 (6) A
p = 90.05 (3) O
V = 2283 (2) A3
As shown in Figure 4, the asymmetric unit is composed of two protonated

nalmefene molecules, two chloride ions, and two ethanol solvent molecules.
All bond lengths and bond angles were found to be reasonable, and the
deduced structure confiied the absolute configuration of the nalmefene
unit. A complex hydrogen bonding system was detected, which involved
the chloride ions, the hydroxyl oxygens, the protonated nitrogens, and the
solvent hydroxyl groups. One view of the crystal packing is provided in
Figure 5.
As illustrated in Scheme 1, morphine is the starting material for nalmefene,

and is a naturally derived material of known stereochemistry [5]. The o d y
stereospecific reaction involved in the synthesis is the introduction of the 14-
hydroxyl function by means of oxidation. Due to conformational constraints
in the molecule, the introduction of the hydroxyl h c t i o n proceeds with
retention of configuration. No configurational inversion at a dissymmetric
center is possible during the other steps of the synthesis.
3.3.1 Optical Rotation
The specific rotation of nalmefene hydrochloride, determined at 589 nm

(sodium D-line) and a temperature of 25"C, is -164.8", calculated on an
anhydrous basis.
Figure 4. View of the asymmetric unit deduced for the ethanol solvate of
OCL1
366 HARRY G . BRlTTAIN
Figure 5 . Packing of nalmefene units in the ethanol solvate structure.
0 0
0 0
3.4 Thermal Methods of analysis
The melting behavior of nalmefene hydrochloride was studied using hot-

stage microscopy. The dehydration of the crystals is visually observed over
the range of 92403°C. The crystals are noted to melt to a clear liquid
within the 180-185°C temperature range, but the melt discolors when the
heating is continued beyond 200°C. This finding demonstrates that the
compound melts with decomposition.
A typical DSC thermogram of nalmefene hydrochloride, obtained in a

closed pan and at a heating rate of 2"C/min, is shown in Figure 6. On the
basis of the hot-stage microscopic findings, the large endotherm observed at
113°C is assigned to the dehydration process, and the small endotherm
observed at 179°C is assigned to the melting transition. That the melting
process is associated with compound decomposition is evidenced by the
drastic change in the thermogram baseline noted after the melting
endotherm. Integration of the peak areas associated with each transition
yielded enthalpy changes of 150 mJ/mg for the dehydration endotherm and
11 r d m g for the melting endotherm.
The TG analysis of nalmefene hydrochloride consists of two regions. The

first of these is complete by approximately 113"C, and reflects the loss of
l the magnitude of this weight loss is 4.8%,
lattice water. In ~ p i c alots,
which is consistent with the anticipated water content of the monohydrate
phase. The second weight loss begins at 200"C, and is indicative of the
oxidative decomposition of the compound.
3.5 Hygroscopicity
The monohydrate phase of nalmefene hydrochloride is essentially non-

hygroscopic, and can only sorb up to 1% adventitious moisture. Material
368 HARRY G . BRI7TAIN
Figure 6. Differential scanning calorimetry thermogram of nalmefene

hydrochloride.
200-
2
0
-200.
-400-
-800
-800'
45 95 145 195
Temperature ("C)
which has been artificiallydried is very hygroscopic, and will quickly

adsorb enough water to produce the monohydrate phase.
In its protonated form, nalmefene hydrochloride is very soluble in water, but

becomes essentially insoluble once the solution pH is raised sufficiently so
as to deprotonate the compound. The hydrochloride salt also exhibits a wide
range of solubilitiesin non-aqueous media, being highly soluble in polar
solvents and insoluble in non-polar solvents. All of the solubility data are
summarized in Table 111.
3.6.1. Solution pH
The equilibrium pH of a 0.10 M solution of nalmefene hydrochloride was

determined to be 5.5, while the equilibrium pH of a 0.05 M solution was
found to be 5.9.
3.6.2 Partition Coefficient
As the solubility data in Table I11 indicate, nalmefene hydrochloride is a

very hydrophilic substance. Its octanoVwater partition coefficient was found
to be 0.075 (log Po,w= -1.125), which is consistent with its hydrophilic
character.
Nalmefene hydrochloride contains a single ionizable group. Using an

aqueous titration method, the pKa of the compound was determined to be
7.63.
3.8 Spectroscopy
3.8.1 WMS Spectroscopy
The ultraviolet absorption spectrum of nalmefene hydrochloride is shown in

Figure 7. The various aliphatic groups yield a reasonably intense absorption
maximum observed at 2 11 nm, which is characterized by a molar
absorptivity of 1977 L*mol/cm. This band is accompanied by a shoulder at
approximately 230 nm (molar absorptivity of 778 L*moVcm). Finally, the
370 HARRY G . BRI'ITAIN
Solubility Characteristics of Nalmefene Hydrochloride
Solvent System Solubility

(mg/mL)
I aqueous, pH = 2.25 128

I aqueous, pH = 5.7 1 I ~~~
131
aqueous, pH = 6.15 133
aqueous, pH = 6.25 124
aqueous, pH = 7.85 1.09
aqueous, pH = 8.5 0.18
aqueous, pH = 9.15 0.09
aqueous, pH = 10.4 0.23
methanol 319
I ethanol I 86.2
I acetonitrile I 1.07
acetone 0.23
chloroform 0.13
NALMEFBNE HYDROCHLORIDE 37 1
Figure 7. Ultraviolet absorption spectrum of nalmefene hydrochloride,

obtained in aqueous methanol at a solute concentration of 0.4
mg/mL.
2.00
1so
1.00
0.50
I -W-\*
220 240 260 280 300
Wavelength (nm)
372 HARRY G . BRl3TAIN
phenolic functional group is responsible for the weaker absorption band

located at 285 nm, which is characterized by a molar absorptivity of 194
L*moVcm.
3.8.2 Vibrationai Spectroscopy
The infia red absorption spectrum of nalmefene hydrochloride was obtained

in a IU3r pellet, and a typical spectrum is shown in Figure 8. A summary of
the band assignments is provided in Table IV.
3.9 Nuclear Magnetic Resonance Spectrometry
Nuclear magnetic resonance studies were conducted on nalmefene

hydrochloride using a JEOL GSX-270 spectrometer, operating at 270.05
MHz ('H-NMR) or at 67.8 MHz (I3C-NMR). Spectra were obtained in
DMSO-d6 at room temperature, and the resonance assignments make use of
the following numbering scheme:
19
3.9.1 'H-NMR Spectrum
The one-dimensional 'H-NMR spectrum of nalmefene hydrochloride is

shown in Figure 9. The assignments for the observed resonance bands are
provided in Table V, with additionaljustification being obtained using two-
dimensional proton-proton correlated spectroscopy (COSY, Figure lo), and
two-dimensional proton-carbon correlated spectroscopy (HETCOR, Figure
1 1). As there is severe overlap of resonances in certain instances, a range of
Figure 8. Infrared absorption spectrum of nalmefene hydrochloride,

obtained in a KBr pellet.
Table IV
Infrared Absorption Spectral Assignments of Nalmefene Hydrochloride
F
Transition Energy
r r ) 1 Assignment
1 799,814 I =CH (aromatic) I

904 -CH,
942,1031,1116 C-0 (furangroup)
r 1170 I C-OH (tertiary hydroxyl group) I
13% I 0-H (phenolic group)
r[-- 1505,1619 C=C stretch

1638 I aromatic ring modes I
3200-3600 -OH and/or NH stretch I
Figure 9. One-dimensional 'H-NMR spectrum of nalmefene
hydrochloride.
Table V
Assignments of the Resonance Bands Observed in the ‘H-NMR Spectrum of

Nalmefene Hydrochloride
I Proton
1
Chemical Shift
OPm)
6.495 (d)
2 6.63 (d)
5 5.02 (s)
I 7, 7’ I 2.02,2.52 (br)
I 8,s’ I 1.15, 1.71 (br)
I 9 I 3.82 (br, d)
I 10’10’ I 2.93’3.26 (br)
1 15, 15’ 1 1.39,2.53 (br)
I 16, 16’ I 2.45, 3.02 (br)

I 17, 17’ I 2.86,3.19 (br)
I 18 I 1.03 (m)
I 19,20 I 0.34 - 0.59 (m)
I 21,21’ I 4.77, 5.13 (s)
I 22 I 9.3 (s)
I 23 I 6.5 (s)
24 8.88 (br)
Figure 10. Two-dimensional proton-proton correlated spectroscopy

(COSY) of nalmefene hydrochloride.
P
8
Figure 1 1. Two-dimensional proton-carbon correlated spectroscopy

(HETCOR) of nalmefene hydrochloride.
-
-
L V t I 0
chemical shifts has been reported for selected protons. The -OH and NH
proton resonances were assigned by long-range coupling.
13
3.9.2 C-NMR Spectrum
The one-dimensional I3C -NMR spectrum of nalmefene hydrochloride is

shown in Figure 12. The assignments for the observed resonance bands are
provided in Table VI, with additionaljustification being obtained using two-
dimensional proton-carbon correlated spectroscopy (HETCOR, Figure 11).
Quaternary carbons were assigned by searching similar fragments on the
Sadtler database. Because of the severe overlap in the proton spectrum, and
the lack of resolution in the HETCOR experiment, the assignments for C7
and C20 are not unequivocal.
3.10 Mass Spectrometry
The mass spectrum of nalmefene hydrochloride was obtained using a VG-

Trio 2 mass spectrometer, operating in the GC-MS (EI+) mode. The
spectrum is shown in Figure 13, and the assignments are given in Table VII :
Table VII
Interpretation of the EI Mass Spectrum of Nalmefene Hydrochloride
d Z Relative Intensity Fragment

I
339 100 [MI+ - HCI 1
298 21.3
284 14.8
I 242
I 17.3
110 28.2
55 65.9
In Table VII, M denotes the nalmefene unit, C21H2503N.

Figure 12. One-dimensional 13C-NMR spectrum of nalmefene
hydrochloride.
NALMEFENE HYDROCHLORIDE 38 1
Table VI
Assignments of the Resonance Bands Observed in the 13C-NMRSpectrum

of Nalmefene Hydrochloride
Chemical Shift
Carbon
(PP@
1 1 19.08
2 117.57
3 140.76
4 143.50
5 87.16
6 145.04
7 26.79
8 3 1.33
9 61.35
10 22.97
11 120.64
12 129.15
13 46.56
14 70.35
15 26.95
16 46.73
17 56.58
18 5.73
19 2.58 or 5.12
20 2.58 or 5.12
21 111.13
Figure 13. Electron-impact mass spectrum of nalmefene hydrochloride.
4.1 Identification
The id en ti^ of nalmefene hydrochloride drug substance is most

conveniently established using infrared absorption spectroscopy.
Approximately 1 mg of nalmefene hydrochloride is thoroughly mixed with
approximately 300 mg of potassium bromide, the mixture ground until
homogeneous, and a pellet compressed from the mixture. The IR spectrum
is scanned between 400 and 4000 wavenumbers, and should exhibit maxima
only at the same energies as does the authentic standard (Figure 8). An
alternative method is to dissolve the drug substance in a small quantity of
methanol, and mix the solution with the solid potassium bromide. Afier
complete removal of the solvent, the pellet may be compressed.
A second identity test for nalmefene hydrochloride has been developed,

which uses thin-layer chromatography methodology. This test will be
discussed in section 4.5.1.
The determination of the elemental analysis of nalmefene hydrochloride

presents no unusual problems for the analyst. Carbon and hydrogen
analyses can performed by the combustion technique, while the nitrogen
analysis is performed using the Dumas method. Typical results are
presented in Table VIII.
Table VIII
Elemental Analysis of Nalmefene Hydrochloride
I %
%CC %
%HH %N % c1
%
I theoretical 1 67.10 I 6.97 I 3.73 9.43

I experimental* I 67.3
67.311 I 7.04 I 3.63 1-9 .391
9.39
*corrected for water content

384 HARRY G . BRITTAW
The assay value of nalmefene hydrochloride is established using aqueous

titration methodology. Approximately 125 mg of sample are accurately
weighed into a titration vessel, to which 125 mL of glacial acetic acid and 10
mL of 6% mercuric acetate solution (in acetic acid). The sample is titrated
potientiometrically with 0.02 N perchloric acid (in acetic acid) using
glasdcalomel electrodes. The purity of the drug substance is calculated
using:
N V (0.3759) (100)
assay -
-
W {l-”/H~O}
where
N = normality of perchloric acid titrant
v = volume of titrant consumed
W = massofsampletaken
YOH,O = decimal fiaction of water in the sample
0.3759 = milliequivalent weight of nalmefene
hydrochloride
4.4 SpectrophotometricMethods of Analysis
The magnitude of the ultraviolet absorbance at 285 is used to differentiate

between the 0.01 mg/mL and 0.1 mdmL formulations of RevexTM.
A TLC method has been developed for nalmefene hydrochloride which is

cased on the use of silica gel 60 F254 (Merck) as the adsorbent. The solvent
system is cyclohexane/chlorofoddiethylarnine (10:75:15). The sample is
prepared by weighing 10 mg of drug substance, and dissolving in 1 mL of
methanol. 10 pL of this solution is applied to the plate, and dried using a
nitrogen flow. The plate is allowed to develop to a height of approximately
12 cm, and then allowed to dry. The spot is viewed using short-wave UV
(254 nm).
A typical thin-layer chromatogram is shown in Figure 14, where the relative

retention (Rf) for nalmefene is 0.41, the Rf for naltrexone is 0.28, and the Rf
for bisnalmefene (the sole degradant product) is 0.07. The criteria used
when the TLC method is used for identification purposes is met if the
retention time of the sample is identical to that of an authentic nalmefene
hydrochloride standard.
An isocratic, stability-indicating, HPLC method has been developed to

determine the impurity profile of nalmefene hydrochloride, and the assay
content in the drug products. The analytical separation is effected using a
PrimsphereTMCIScolumn, and uses a mobile phase consisting of 20:80
acetonitrile:O.O5M phosphate buffer. The buffer contains 0.2%
triethylamine, and the pH is adjusted to 4.2 with 85% phosphoric acid.
Using a mobile-phase flow rate of 1.O mL/min, the analyte detection is
effected on the basis of the UV absorbance at 2 10 nm. Sample and
standard solutions are prepared at concentrations of 0.01 mg/mL, and the
method may be used to assay the content of either the 1.O or 0.1 mg/mL
presentations of RevexTM, as long as aliquots of the neat formulation are
diluted to 0.01 mg/mL prior to analysis.
When the method is used to obtain nalmefene concentrations (expressed as

the free base in units of mg/mL) within drug product formulations, the
assay value is calculated using:
A&m W (l.O-%H*O) P 339.4

conc.(mg/mL) = - X X
AStd 100 375.9
where:
Asam = average peak area of nalmefene in
the sample solution
AStd = average peak area of the nalmefene standard
solution
Figure 14. Thin-layer chromatogram of nalmefene hydrochloride.

Moving across the chromatogram from left to right, the spot at
furthest left is that of a 1 mg sample, and next is the standard
at the same concentration. The next two spots (at the same Rf
values) are the standard spotted at concentrations of 5 pg and
2 pg, respectively. The next two spots correspond to
naltrexone (5 pg) and bis-nalmefene ( 5 pg).
W = mass of nalmefene hydrochloride standard

taken (mg)
%H,O = the water content of the standard, expressed
as a fraction
339.4f379.5 = Conversion factor, from nalmefene
hydrochloride salt to the free base
P = Purity factor of the nalmefene hydrochloride
reference standard, expressed as a fraction
The most important use of the analytical method is in the determination of

the impurity profile of nalmefene hydrochloride. As evident in Figure 15,
excellent separation of nalmefene from its process impurities (naltrexone
and A7-nalmefene)and sole degradant (bis-nalmefene) is obtained.
Structures of these related substances are provided in Figure 16.
4.6 Determination in Body Fluids and Tissues
The first method reported method for the determination of nalmefene in

human plasma used HPLC and electrochemical detection [6]. Following
extraction of the drug substance at pH 9, the extract was chromatographed
on a reversed-phase C 18 column, and detected using a glassy carbon
electrode detector. The method sensitivity was reported to be 3 ng/mL,
after extraction of a 1 mL sample of plasma.
A specific radioimmunoassay has been developed for the quantitation of

nalmefene in human plasma [7]. Since nalmefene and naltrexone were
found to cross-react equally well with a rabbit antiserum to an albumin
conjugate of naltrexone-6-(o-carboxymethyl)oxime,it was possible to use
fHI-naltrexone as the radioligand. Plasma samples were mixed with pH 9
borate buffer, extracted into ether, and the organic phase dried prior to re-
constitution in the assay buffer. When extracting 0.5 mL of plasma, a
limit of sensitivity equal to 0.2 ng/mL of nalmefene was obtained.
A HPLC method suitable for the measurement of nalmefene in human

plasma after either oral or intravenous administration has also been
described [S]. The drug is extracted from plasma using a solid-phase
cyanopropyl column, and eluted with a 60% (v/v) acetonitrile in dilute
Figure 15. HPLC chromatogram of nalmefene hydrochloride spiked with
7
naltrexone, A -nalmefene, and bis-nalmefene.
Figure 16. Structures of the nalmefene process impurities (naltrexone and

7
A -nalmefene) and the sole degradant (bis-nalmefene).
A'-NdwCene
HO
Ndtrexone
390 HARRY G . BRIl’TAIN
sodium pentanesulfonic acid solution. The concentrated and filtered

eluent is injected onto a HPLC system, which uses a phenyl column to
effect the separation and a mobile phase consisting of 30% (v/v)
acetonitrile in dilute sodium pentanesulfonic acid solution. The analyte
species are detected by electrochemical means, using a dual-electrode
system. A signal-to-noise ratio of 4.5 was reported for nalmefene when a
1 ng/mL spiked plasma sample was analyzed.
5. Stability
5.1 Solid-state Stability
Nalmefene hydrochloride has been found to be extremely stable when

stored at 15-25°C in well closed containers, and protected from both light
and moisture. Some bulk drug substance has been stored at room
temperature for up to seven years with no detectable change in compound
purity.
5.2 Solution-Phase Stability
Using the HPLC method described in section 4.5.2, the stability of

nalmefene hydrochloride in both the 0.1 mg/mL and 0.0 1 mg/mL
formulations has been studied at a variety of temperatures for periods up to
36 months [ 1 11. The trends in nalmefene potency obtained during the
course of these studies are illustrated in Figure 17, while the formation
trend of the sole degradant (2,2’-bisnalmefene) is found in Figure 18.
The stability data were found to be interpretable using first-order kinetics,

and essentially comparable rate constants were calculated for both the
potency loss and the formation of 2,2’-bisnalmefene. Applying the
Arrhenius equation to these data, a rate constant of 0.000441 month-’ was
deduced for the reactions taking place at 25°C. When this value for the
first-order rate constant was substituted into the integrated rate expression,
it was predicted that the drug product would not exceed its impurity
specification (not more than 2.5% 2,2’-bisnalmefene can be present at the
end of the product shelf life) for approximately 4.8 years [l 11.
Figure 17. Trends in product potency observed during the stability study
of the 1.O mg/mL RevexTMformulation (2 mL fill in a 2-mL
ampule) at 4,30,40,55, and 75°C.
100
A
U 90
r
0
s
t
0 80
e
70
5 15 25 35
Time (months)
Figure 18. Formation of 2,2’-bisnalmefene observed during the stability

study of the 1 .O mg/mL RevexTMformulation (2 mL fill in a 2-
mL ampule) at 4,30, and 40°C.
2.5
1.5
7 30°C
0.5
4°C
5 15 25 35
Time (months)
In a subsequent study, the short-term stability of RevexTMwas determined

in a number of diluents commonly employed for intravenous use [ 121.
Dilutions of RevexTMwere prepared separately in 0.9% sodium chloride
injection, 0.45% sodium chloride injection, 5% dextrose injection, 5%
dextrose and 0.45% sodium chloride injection, lactated Ringer's injection,
5% dextrose and lactated Ringer's injection, and 5% sodium bicarbonate
injection. Each admixture was stored at 4"C, room temperature (21"C),
and 40"C, with samples being tested after storage at each temperature for
0,24,48, and 72 hours. The results of this study are summarized in Table
IX. Defining stability as the retention of at least 95% of the initial drug
concentration at the end of the storage period, it was concluded that the
diluted solutions of RevexTMwere uniformly stable for up to 72 hours in
all of the injectable solutions maintained at either 4"C, 21"C, or 40°C.
6. Acknowledgements
Special appreciation is given to those who contributed toward the work

described in this analytical profile, especially Linda Lafferty, Petranna
Bousserski, Glenn Diegnan, Ralph Lessor, Satish Pejaver, Camille Small,
Satya Murthy, Ashok Krishnaswami, Kamalesh Johri, George Owoo,
Patrice Rafalko, and Martin Gall.
394 HARRY G. BRTITAIN
Table IX
Stability of RevexTMin Various Injectable Solutions
Solution Storage Percent Percent Percent

condition Remaining Remaining Remaining
5% Dextrose 4uc 101.0 99.5 99.0

and Lactated RT 99.0 98.0 98.5
Ringer's 4OoC 100.5 99.5 100.0
5% Sodium 4uc 99.5 99.5 101.0
Bicarbonate RT 100 99.5 100.0
4OoC 100.5 99.5 99.0
7. References
1. D.H. Dyson, T. Doherty, G.I. Anderson, and W.N. McDonell,

Veterinary Surg., 19,398-403 (1990).
2. T.J. Gal and C.A. DiFazio, Anesthesiology, 64, 175-180 (1986).
3. E.F. Hahn, J. Fishman, and R.D. Heilman, J. Med. Chem., 18,259-

262 (1975).
4. E.J. Corey and M. Chaykovsky, J. Am. Chem. SOC.,87,1345-1349

(1 965).
5. R. Greenwald, M. Chaykovsky, and E.J. Corey, J. Org. Chem., 28,

1128-1129 (1963).
6. P.C. Meltzer and J.W. Coe, U.S. Patent 4,535,157 (1985).
7. W. Klyne and J. Buckingham, Allas of Stereochemistry, 2ndedn.,

volume 1, Oxford University Press, New York, 1978, p. 141.
8. J. Hsiao and R. Dixon, Res. Comm. Chem. Path. Pharm., 42,449-

454 (1983).
9. R. Dixon, J. Hsiao, W. T d e , E. Hahn, and R. Tuttle, J. Pharm.

Sci., 73, 1645-1646 (1984).
10. J.Z. Chou, H. Albeck, and M.J. Kreek, J. Chrom., 613,359-364

(1 993).
11. H.G. Brittain, L. Lafferty, P. Bousserski, G. Diegnan, R. Lessor, C.

Small, and S. Pejaver, PDA J. Pharm. Sci. Tech., in press.
12. S.S. Murthy and H.G. Brittain, J. Pharm. Biomed. Anal.,

submitted.
POLYVINYL ALCOHOL
David Wong and Jagdish Parasrampuria
CIBUS Pharmaceutical, Inc.
200 D Twin Dolphin Drive
AND EXCIPENTS-VOLUME 24 All rights of reproductionin any fom reserved.
398 DAVID WONG AND JAGDISH PARASRAMPURIA
1. Introduction
2. Description
3. Sources and Manufacturing
4.1 solubility
4.2 Viscosity
4.3 Moisture Sorption
4.4 Permeation of Polyvinyl Alcohol Films
4.5 Glass Transition
4.6 Crystallinity and Preparation of Polyvinyl Alcohol
Hydrogels
4.7 Surface Properties of Polyvinyl Alcohol Hydrogels
4.8 Optical Properties of Polyvinyl Alcohol Hydrogels
4.9 Mechanical Strength of Polyvinyl Alcohol Hydrogels
4.10 Water Content and Swelling Properties of Polyvinyl
Alcohol Hydrogels
4.1 1 Protein Absorption by Polyvinyl Alcohol Hydrogels
5.1 Cornpendial Methods
5.1.1 pH
5.1.4 Viscosity
5.1.5 Water-Insoluble Substances
5.1.6 Degree of Hydrolysis
5.2 Other Analysis Studies
6. Stability
7. Applications
8. Pharmacological Effects and Toxicity
9. References
POLYVINYL ALCOHOL 399
1. Introduction
Polyvinyl alcohol (CAS number 9002-89-5, and often abbreviated as

PVA) was first commercially used in Germany in the 1920's, and since
then has been applied in various areas such as textiles, paper, adhesives,
cements, and films [5]. During the last 10 years, great advances have been
made in medical and pharmaceutical technologies, and polyvinyl alcohol
application methods have also undergone significant changes.
Polyvinyl alcohol has been an article in the United States Pharmacopoeia,

and the material specifications are those listed in the official monograph
~71.
2. Description
As illustrated in Figure 1, polyvinyl alcohol is a polyhydric alcohol

containing secondary hydroxyl groups on alternate carbon atoms.
Figure 1. Structure of polyvinyl alcohol.
Polyvinyl alcohol is usually not chemically modified, nor is it cross-

linked. However, for special purposes a polymer can be obtained through
crosslinking by difunctional or polyfunctional reagents.
Polymerization and alcoholysis are well-controlled during the preparation

of different grades of polyvinyl alcohol, yielding materials characterized
by different viscosities and percentages of hydrolysis. The "percentage
hydrolyzed grades" are determined by the time used for alcoholysis, and
the "viscosity grades" are controlled by the time used for polymerization
of vinyl acetate. "Chain-breakers" are occasionally added to assist in
controlling the molecular weight (viscosity grades) of polyvinyl alcohol.
Complete alcoholysis will lead to a "fully hydrolyzed grade" [4,51.
Examples of different commercially available grades of polyvinyl alcohol

are listed in Table 1.
3. Sources and Manufacturing
The raw material for polyvinyl alcohol is vinyl acetate. Vinyl acetate is
first polymerized to polyvinyl acetate by conventional techniques such as
bulk, solution, or emulsion polymerization. Polyvinyl acetate is then
dissolved in an organic solvent (such as mixed methanol and methyl
acetate) and hydrolyzed so that the pendant acetate groups are replaced by
pendant hydroxy groups. The polyvinyl alcohol precipitates out of the
reaction medium because of its low solubility, and is then filtered, washed,
dried, and packaged [5].
Polyvinyl alcohol is an odorless, white, granular powder, that decomposes

when heated to about 200°C [63,64]. It dissolves in water, forming an
acidic solution [3]. A summary of the important physical properties is
provided in Table 2.
4.1 Solubility
Polyvinyl alcohol is soluble in water, and in most mixtures consisting of

water and a water-miscible organic solvent (such as glycerin, ethylene
glycol, or dimethyl sulfoxide) [ 11. The compound has a high solubility in
water and mixed solvents, but can form lumps that require a long time to
dissolve completely. Therefore, polyvinyl alcohol should be dissolved
according to the following procedure:
Table 1
Some Standard Grades of Polyvinyl Alcohol [3]

Table 2
Physical Properties of Polyvinyl Alcohol [3]
I Appearance White to cream granular powder
I Bulk Density 40 1bs/ft3

~~
Specific Gravity (solid) 1.27 - 1.31
Specific Gravity
1.02
(1 0% solution at 25°C)
Thermal Stability Gradual discoloration at about

100°C; darkens rapidly above
150°C; rapid decomposition above
200°C
1 Refractive Index (film) at 20°C 1.55
Thermal Conductivity 0.2 W/(m.K)
Electrical Resistivity (3.1-3.8) X 107 ohm.cm
Specific Heat 1.5 J/(g.K)
Melting Point (unplasticized) 230°C for fully hydrolyzed grades;

180-190°C for partially hydrolyzed
grades
75 - 85°C
Stability (solid) Stable when protected from

moisture
Flammability Burns similar to paper
i Stability to Sunlight Excellent

The solid is first dispersed and agitated in cold water. When all
particles are wet, then the solution temperature is raised to 85-96°C
until all particles dissolve [ 3 ] . The exact temperature for preparing
complete solutions depends on the polyvinyl alcohol concentration
and its grade.
The solubility of polyvinyl alcohol is primarily a function of its hydrolysis

percentage and viscosity, properties derived from its molecular weight.
For example, 87-89 % hydrolyzed polyvinyl alcohol is soluble in both
cold and hot water. Polyvinyl alcohol hydrolyzed over 89% dissolves
only in hot water, while 75 to 80% hydrolyzed polyvinyl alcohol is soluble
only in cold water [4].In addition to its percentage of hydrolysis, tacticity
can also influence the solubility. In water, the atactic isomer is more
soluble than is the isotactic isomer [16]. The maximum amount of solids
recommended for solutions prepared using conventional high-speed
mixers are summarized in Table 3.
4.2 Viscosity
The most important single feature of polyvinyl alcohol is its water-binding

capacity, which is commonly expressed as viscosity. These measurements
can be performed using a Brookfield or Haake viscometer, since polyvinyl
alcohol dissolves in water to give a viscous solution. The viscosity of this
solution depends on the temperature, molecular weight, and concentration
of the solute [ 3 ] . Table 4 shows the relationship between viscosity and
polyvinyl alcohol concentrations at various temperatures.
4.3 Moisture Sorption
When polyvinyl alcohol films are used in controlled-release dosage forms,

the permeability of the film to moisture, water, and drug becomes
important. Films can be obtained simply by heating a dispersion of
polyvinyl alcohol in water at 60°C for two hours, and then drying at room
temperature. Untreated polyvinyl alcohol films are readily permeable to
water and drugs, reducing its usefulness in controlled-release dosage
forms. To reduce the permeability, cross-linking or increasing
crystallinity via heating can be applied [131. Thus, studies of permeation
Table 3
Maximum Recommended Solid Amounts for Polyvinyl Alcohol

Solutions (3)
Maximum
Percent Hydrolysis Viscosity, cps (4 Recommended
Solid
Super Hydrolyzed 28 - 32 10%

(99.3%) 62-72 7%
Fully Hydrolyzed 5.5 - 6.6 20%

(98.0 - 98.8%) 28.5 - 32.5 10%
62 - 72 7%
Intermediate 14 - 17 15%
Hydrolyzed 27 -31 10%
(95.5 - 97.5%)
Partially Hydrolyzed 3.5 - 4.5 30%

(87 - 89%) 5.2 - 6.2 20%
23 - 27 10%
45 - 55 7%
(a) 4% aqueous solutions, 20°C

Table 4
Kinematic Viscosity Determination of Polyvinyl Alcohol Solutions

(W40/140, Wacker Chemical Company); n=6 191
Polyvinyl Alcohol Kinematic Viscosity Kinematic Viscosity

Concentration at 33°C at 37°C
(%) (mm2s-1) (llldS-1)
0 0.80 20.01 0.72 50.01
1 1.86 20.01 1.70 50.02
3.5 17.32 k1.10 13.22 L0.59
5 56.95 L1.91 46.23 L0.44

and absorption of polyvinyl alcohol films are important in the design of

formulations/dosage forms [ 11- 131.
Three principal forms of water are recognized in films, with the water
being designated as either tightly bound, moderately bound or free.
Tightly bound water is taken as the water bound within the crystalline
phase of the polymer, and can be determined by an intense extraction
procedure using vacuum over P202 for a week [8]. Moderately bound
water is envisioned as the water adsorbed to hydroxyl groups in the
amorphous phase of a polymer. The amount of moderately bound water
can be determined by using thermogravimetric analysis (Figure 2). Figure
2 shows a transition from 30°C to 130"C, with a 10% original film weight
loss, and a sharper transition at about 260°C with a further 70% weight
loss. The first transition corresponds to loss of moisture, while the second
transition represents film decomposition and a further loss of moisture.
Finally, the free water is that which resides in a bulk or condensed form on
the surface or within voids of films [8]. This can be determined by the
mass balance of the total moisture content obtained under static conditions
(WA), moderately bound (WM)), and tightly bound water contents (WT):
Moisture content data for a typical polyvinyl alcohol film at 25°C are
shown in Table 5.
4.4 Permeation of Polyvinyl Alcohol Films
Water-soluble additives (such as citric acid or urea) have been found to

lower the moisture diffusion through polyvinyl alcohol films, but also
increase their solubility coefficients without changing their permeability
coefficients. Usually, 5-10 wt. YOare the estimated limits of compatibility
with the film-former for the additives [ 1 11.
Heat treatment of polyvinyl alcohol has been used to increase its

crystallinity, which also reduces the water solubility, equilibrium water
content, and degree of swelling [ 13,18,19]. Without heat treatment,
polyvinyl alcohol films dissolve slowly in water. However, increasing the
heating time at i00"C decreases the membrane permeability to methylene
Figure 2. Thermogravimetric analysis of a polyvinyl alcohol film [S].
Temperature ("C)
Table 5
Moisture Content Data for a Typical Polyvinyl Alcohol

Film at 25OC [8]
Condition Moisture Content

@>
~~~
Just after casting 9.65 i 0 . 2 8
Storage at 75% relative humidity 23.24 L0.5 1

for one week
Tightly bound water 7.84 f 0.12
Moderately bound water 10.10 5 0.4
Free water 5.30

blue, as has been illustrated in Figure 3. Similarly, drug release from

polyvinyl alcohol films is slowed by increasing the time of heat treatment
(Figure 4).
4.5 Glass Transition
The differential scanning calorimetry (DSC) and thermomechanical

analysis (TMA) themograms of a polyvinyl alcohol film containing no
additives are shown in Figure 5 [lo]. Three transitions are present in the
DSC thermogram: a glass transition (Tg) at about 70°C, an unknown
endotherm (Tu) at 12O-15O0C,and a melting endotherm (Tm) between
150-215°C. Three transitions were also found in the penetration mode of
the TMA thermogram: two transitions indicating polymer softening (Ts 1
and Ts2) and another (Tm) related to melting/degradationof polyvinyl
alcohol. However, Tg transitions were not visible in the expansion mode
of the TMA thermogram. The effect of citric acid and urea on the Tg, Ts,
and Tm of polyvinyl alcohol films has been reported [lo]. Urea decreases
Tg, Ts, and Tm to a slightly larger extent than does citric acid.
4.6 Crystallinity and Preparation of Polyvinyl Alcohol Hydrogels
Polyvinyl alcohol is reported to be semicrystalline [lo]. Incorporating

additives such as urea and citric acid into the material changes its
crystallinity. The percent change in crystallinity is usually obtained by
comparing the corresponding heat of fusion obtained for pure polyvinyl
alcohol with that of material containing additives. A typical data set is
provided in Table 6 .
Polyvinyl alcohol hydrogels can be obtained by increasing the degree of

crystallinity. A semicrystalline or densified polyvinyl alcohol structure
can be enhanced by either cooling a solution consisting of water and a
water-miscible solvent, or by exposing an aqueous polyvinyl alcohol
solution to several freeze-thaw cycles. Typical data obtained for this
process are shown in Table 7 [l , 56-58]. The structure of hydrogels
formed by cooling a polyvinyl alcohol solution is porous, in which the
pore diameter can be varied by changing the concentration of the solvent
used. Scanning electron micrographs of a polyvinyl alcohol hydrogel
prepared by exposing solutions to several freeze-thaw cycles are shown in
Figure 3 . Effects of heating time at 100°C on polyvinyl alcohol

membrane permeability (P=l/Rm) to methylene blue at
37°C. Error bars represent the standard deviations [13].
0-1 r I I I I I I I t
0 20 40 60 80 100 120 140 160
Heating h e a t IOO'C, hours
POLYVINYL ALCOHOL 41 1
Figure 4. Release profiles of sulphathiazole from polyvinyl alcohol

films subjected to different periods of heat treatment at
160°C. Time periods of 1 hour (V), 2 hours (A), 3 hours
(0)and 4 hours (0) are shown.
Figure 5 . Differential scanning calorimetry and thermomechanical

analysis thermograms of polyvinyl alcohol film [ 101.
Temperature “C
Table 6
Heats of Fusion and Percent Change in Crystallinity of Polyvinyl

Alcohol Films Containing Citric Acid and Urea [lo]
Additive Citric Acid Urea
Content Percent Percent

(wt %) Crystallinity Crystallinity
Change Change
0 30.9 0.0 30.9 0.0
1 32.1 +3.8 27.7 -10.3
3 29.1 -5.8 26.2 -15.2
28.2 -8.7 22.5 -27.3
25.4 -17.8 19.4 -37.2

Table 7
Heat of Fusion and Degree of Crystallinity of Freeze-thawed Polyvinyl

Alcohol Gels after Five Freeze-thaw Cycles [57]
Thawing Time Degree of Degree of

at 25°C Heat of Fusion Crystallinity, Crystallinity,
(hours) (Jk) Dry Basis Swollen Basis
(%)
O I
47.45
I 34.2 5.4
5.5 1 29.34 21.1 3.6
5.1
9.7
Figure 6. Other ways to prepare polyvinyl alcohol hydrogels include the

cross-linking of polyvinyl alcohol by a mixture of aldehydes, or by
exposing polyvinyl alcohol to an electron beam or gamma-rays [6]. A
proposed structure for polyvinyl alcohol cross-linkages is shown in Figure
7 [61.
4.7 Surface Properties of Polyvinyl Alcohol Hydrogels
Alkylation of a polyvinyl alcohol hydrogel surface increases the percent

transmittance of the infrared absorption at 3300 wavenumbers associated
with the hydroxyl group, but decreases the absorbance of the urethane
peaks at 1690 wavenumbers (Figure 8) [2]. Such alkylation can be
quantitatively analyzed by using x-ray photoelectron spectroscopy, with
typical data being shown in Figure 9 [2]. The degree of surface
irregularity of hydrogels can be observed easily by scanning electron
microscopy (Figure 10) [11.
4.8 Optical Properties of Polyvinyl Alcohol Hydrogels
The transmittance of light through a polyvinyl alcohol hydrogel increases

with increasing concentration of dimethyl sulfoxide (DMSO) used in the
solvent mixtures associated with its preparation (Figure 11).
4.9 Mechanical Strength of Polyvinyl Alcohol Hydrogels
The mechanical properties (such as tensile strength and elongation) of

polyvinyl alcohol hydrogels prepared by freeze-thaw cycles or the cooling
method are dependent on the composition of their solvent mixtures,
polyvinyl alcohol concentrations, and crystalline polyvinyl alcohol
content. These relationships have been illustrated in Figures 12 and 13.
Table 8 compares the physical properties of polyvinyl alcohol hydrogels
prepared from a 20% solution in a mixture of water and DMSO
(20%/80%) with the properties of hydrogels made of other polymers. The
elongation at the breaking point of the polyvinyl alcohol hydrogels was
found to be approximately three times that obtained for the other
materials.
Figure 6. Scanning electron micrographs for polyvinyl alcohol gels

prepared without and with pluronic block polymer L-62
[ S S ] . Images are shown for L-62 concentrations of (a) 0%
and (b) 1.O%, where the scale bar equates to 10 pm.
Figure 7. Schematic representation of the swelling of a

semicrystalline polyvinyl alcohol network [6].
Figure 8. FTIWATR spectra of (a) polyvinyl alcohol, (b) polyvinyl

alcohol after 4 hours of reaction in dimethyl formamide,
and (c) polyvinyl alcohol after 9 hours of reaction in
dimethyl sulfoxide [2].
If 4
94
74
54
34
14
t15
97
79
61
50
43
Wave number
Figure 9. High-resolutionx-ray photoelectron spectra obtained for

the carbon 1s region of polyvinyl alcohol after 4 hours of
reaction in dimethyl formamide (upper trace) and polyvinyl
alcohol after 9 hours of reaction in dimethyl sulfoxide
(lower trace) [2].
2983 278.3
8iflding €rtergy (eV)
Figure 10. Scanning electron micrographsof polyvinyl alcohol

hydrogels obtained from water and dimethyl sulfoxide
mixed solutions of polyvinyl alcohol. The ratios of
water:DMSO are (A) 100:0, (B) 80:20, (C) 60:40, (D)
40:60, (E) 20:80, and (F) 0:lOO. The scale bar equates to 5
Pm P I .
Figure 11. Measurements of the light transmittance of polyvinyl

alcohol hydrogels at different concentrations of dimethyl
sulfoxide [ 13.
Q
0
E
E
tJa
C
a
L
V
V
t
P
.-.
0 20 40 60 ao 100
Concentration of OMSO in water (w/w%)
Figure 12. Relationship between the tensile strength (0)and

elongation (0)of polyvinyl alcohol hydrogels and their
concentration of polyvinyl alcohol (water:dimethyl
sulfoxide ratio = 2090) [l].
40 -
m
- 400
a
200
0- I I t I I
0
0 10 20 30
P V A concentratton ( % )
Figure 13. Variation of the tensile modulus of semicrystalline

polyvinyl alcohol hydrogels with the crystalline polyvinyl
alcohol content (density = 2.66 X mole/cm3) [6].
I
"E I ' I
-' I ' -I
I I
' I '
> 11
U
e -
>r
{9-
5t-
c
1
20 30
CRYSTALLINE PVA CONTENT ('I*)
Table 8
Physical Properties of Hydrogels Prepared Using

Various Polymers [11
Copolymer of
Physical Polyvinyl Polyhydroxyethyl Methylmethycrylate
property Alcohol Methacry late and N-vinyl
Pyrrolidone
Water content
(%I 78 38 78
Tensile strength
(kgkm2) 47
Elongation
("/I 500
------t
Light
transmittance 99- 100 99-100 99- 100
(?%)a
Oxygen
permeability b 44 10 46
(a) 550 nm, 0.2 mm thick in water

(b) 10-1 1 cm3 (STP) crn2/cm3 s mm Hg (35°C)
The mechanical strength of cross-linked polyvinyl alcohol hydrogel has

been found to be proportional to the content of cross-linking agents, such
as magnesium chloride, formaldehyde, or glutaraldehyde [ 191. Table 9
shows that the effect of inversely increasing cross-linking agents parallels
the effect on the compression modulus and cross-link density. As the
water content decreases by increasing glutaraldehyde concentration, the
gel becomes stronger with higher compression moduli and cross-link
densities.
4.10 Water Content and Swelling Properties of Polyvinyl Alcohol

Hydrogels
Polyvinyl alcohol hydrogels do not dissolve in water, but will swell in

volume upon contact with bulk water. After polyvinyl alcohol films or
hydrogels are swollen in buffers for a predetermined period, their water
content can be calculated using:
water content = Ww-Wd

WW
where: Ww = weight of swollen film

Wd = weight of dry film
The weight of dry film can be found by drying films/gels for 48 hours,
with the temperature being gradually increased from 40 to 70°C. In most
cases, water does not influence the crystalline regions upon swelling of
polyvinyl alcohol hydrogels [ 6 ] .
The swelling behaviors of most hydrogels can be characterized using

similar calculations. The effect of ionic strength on the swelling behavior
of heparin- polyvinyl alcohol hydrogel is shown in Table 10 [36]. It was
found that the hydrogels shrank as the ionic strength of the medium was
increased. This phenomenon can be explained by the screening effect of
salt on the charges within the gel.
4.1 1 Protein Adsorption by Polyvinyl Alcohol Hydrogels
The study of protein adsorption by polyvinyl alcohol hydrogels is

important, since these hydrogels are used as contact lenses and artificial
tissues. In Figure 14, the effect of alkylation on the amount of albumin
Table 9
Physical Characteristics of Selected Polyvinyl Alcohol Hydrogels [191
Compression Cross-link
Formulation Modulus Density Water
(kg/cm2) (gram Content
moie/c&) (%, w/w)
10% polyvinyl
alcohol; 7.8 k 0.14 1.41 x 10-4 74.7 & 1.0
0.5% glutaraldehyde
10% polyvinyl
alcohol; 68.9 k 3.1
2% glutaraldehyde
Table 10
Shrinkage of Heparin-PolyvinylAlcohol Gel after Incubation in

Solutions of Varying Ionic Strengths [36]
Relative Weight of Gel

(n=3)
~~~
0 1 .oo L 0.01
0.15 1.05 k 0.06
0.50 0.99 f.0.08
1.o 0.89 k 0.04
2.0 0.76 2 0.04
3.O 0.68 k 0.01
All solutions contain 0.01M sodium phosphate

Figure 14. Surface concentration of albumin on polyvinyl alcohol,

solvent-treated polyvinyl alcohol, and alkylated polyvinyl
alcohol after 1 hour adsorption from a 9.4 pM solution of
albumin. Results are presented as the mean k SD for the
indicated number of tubes; each tube value being a mean of
several segments [2].
adsorbed to the polyvinyl alcohol surface is illustrated [2]. It was found

that the degree of adsorption increased with increasing alkyl chain length.
The adsorption rates of three different proteins are shown in Table 11,
where it may be seen that polyvinyl alcohol hydrogel adsorbed the least
amount of protein [l].
5.1 Compendia1 Methods
According to the United States Pharmacopeia [7], polyvinyl alcohol is a

synthetic resin, for which the number of repeating units (the n factor in
Figure 1) lies between 500 and 5000. It is prepared by the 85-89%
hydrolysis of vinyl acetate.
5.1.1 pH
The pH is determined using USP general test <791>, and should be

between 5.0 and 8.0 for a 1 in 25 solution.
When the solid is dried at 110°C to constant weight, is cannot lose more
than 5.0% of its weight. The details of the method are given in USP
general test <73 1>.
When tested according to USP general test <281>,the residue on ignition

cannot be more than 2.0%.
5.1.4 Viscosity
After determining the loss on drying, weigh a quantity of undried

polyvinyl alcohol, equivalent to 6.00 g on the dried basis. Over a period
of seconds, transfer the test specimen with continuous slow stirring to
about 140 mL of water contained in a suitable tared flask. When the
Table 11
Protein Adsorption Rates of Hydrogels [l]
Water Immuno- Bovine

Hydrogel Content globulin G Serum
(%) (pg/cm2) Albumin
(vdcm2)
Polyvinyl Alcohol 78 0.074 0.005 0.195
Polyhydroxyethyl
Methacrylate 40 0.271 0.037 0.230
Copolymer of
Methylmethycrylate 78 0.889 0.169 4.991
and N-vinyl
Pyrrolidone
specimen is well-wetted, increase the rate of stirring, avoiding mixing in

excess air. Heat the mixture to 90"C, and maintain the temperature for
about 5 minutes. After that, discontinue the hearing, and continue stirring
for 1 hour. Add enough water to bring the mixture up to a total weight of
150 g, and resume stirring to obtain a homogeneous solution. Filter the
solution through a tared 100-mesh screen into a 250-mL conical flask,
cool to about 15"C, mix, and measure the viscosity according to USP
general test <911>.
5.1.5 Water-Insoluble Substances
Wash the tared 100-mesh screen used in the test for viscosity with two 25-
mL portions of water, and dry at 110°C for 1 hour. Not more than 6.4 mg
(0.1%) of water-insoluble substances can be found.
5.1.6. Degree of Hydrolysis
Transfer about 1 gram of polyvinyl alcohol, previously dried at 110°C to

constant weight and accurately weighed, to a wide-mouth, 250-mL conical
flask fitted by means of a suitable glass joint to a reflex condenser. Add
35 mL of dilute methanol (3 in 5), and mix gently to assure complete
wetting of the solid. Add 3 drops of phenolphthalein test solution, and add
0.2 N hydrochloric acid or 0.2 N sodium hydroxide (if necessary) to
neutralize. Then add 25.0 mL of 0.2 N sodium hydroxide VS, and reflux
gently on a hot plate for 1 hour. Wash the condenser with 10 mL of water,
collecting the washings in the flask, cool, and titrate with 0.2 N
hydrochloric acid VS. Concomitantly perform a blank determination in
the same manner, using the same quantity of 0.2 N sodium hydroxide VS.
Calculation of the saponijication value. Calculate the saponificationvalue
by the formula:
[(B - A) N 56.1 11 / W
where B and A are the volumes (in mL) of 0.2 N hydrochloric acid VS
consumed in the titration of the blank and test preparation, respectively. N
is the exact normality of the hydrochloric acid solution, W is the weight
(in grams) of the portion of sample taken, and 56.1 1 is the molecular
weight of potassium hydroxide.
Culculution of the degree of hydrolysis. Calculate the degree of

hydrolysis, expresses as a percentage of hydrolysis of polyvinyl acetate, by
the formula:
100 - [7.84 S / (100 - 0.075 S)]
in which S is the saponification value of the sample taken. The degree of
hydrolysis should be within 85% and 89%.
5.2 Other Analysis Studies
Filter paper treated with potassium iodide and iodine solutions has been
used to measure polyvinyl alcohol concentrations in waste water over a
concentration range of 1000 to 20000 ppm [5]. Polyvinyl alcohol can be
reacted with boric acid to yield a green complex which can be used to
detect small amounts of the compound in polyvinyl chloride resins [ 5 ] .
The molecular weight of polyvinyl alcohol is determined by gel

permeation chromatography [2 I]. Turbidity is used to assay polyvinyl
alcohol in biological media [5].
Combinations of various thermal techniques, such as differential scanning

calorimetry, thermomechanical analysis, and thermogravimetric analysis
have been demonstrated to assess the glass transition, softening, melting,
plasticization, and crystallinity and moisture content of polyvinyl alcohol
[8, lo]. Compression modulus, tensile strength and elongation are
commonly employed to assess the physical strength of polyvinyl alcohol
hydrogels and films.
X-ray diffraction and UVNIS absorption spectra have been used to

characterize the effect of tacticity of polyvinyl alcohol on its complexion
with iodine [ 141. Linear dichroism techniques and polarized
photoacoustic, absorption, and fluorescence spectra were used to study the
orientation of pigments in stretched polyvinyl alcohol films [26,15].
Nuclear magnetic resonance (NMR) has been used to investigate the
structural consistency of the polyvinyl alcohol copolymer [2 11. Fourier
transform infrared spectroscopy coupled with attenuated total reflectance
optics, x-ray photoelectron spectroscopy, and scanning electron
microscopy were shown to be useful in characterizing the surface
properties of polyvinyl alcohol hydrogels [ 1,2].
6. Stability
Polyvinyl alcohol degrades slowly at 1OO'C, but the degradation takes

place rapidly at 200°C [63,64]. The polymer is considered to be
biodegradable because of its solubility in water [ 161. The degradation
process initiates with water diffusing into the polyvinyl alcohol matrix,
followed by swelling, fragmentation, and dissolution.
An apparent first-order kinetics is observed for the hydrolysis of polyvinyl

alcohol. The hydrolysis mechanism has been found to be that of specific
hydrogen- and hydroxyl-ion catalysis, and the pH of optimal stability is
4.75 [17]. Polyvinyl alcohol also undergoes esterification and the other
reactions of a compound with secondary hydroxyl groups. At high
concentrations, it is incompatible with sulfates and phosphates [63].
7. Applications
Polyvinyl alcohol is considered an emulsifyingagent, a suspending agent,

and an emollient in pharmaceutical science [60,63-651. Consequently, it is
employed as a lubricant and protectant in various liquid preparations, such
as topical lotions, suspensions, creams, ophthalmic solutions, artificial
tears, vaginal emulsions, aerosol foams, and decongestants [65,60]. Other
applications include oral tablets and film-coated tablets, as well as films
for transdernial controlled release [60,49].
Recently, drug-containing polyvinyl alcohol microparticles have been

prepared by emulsion, spray-drying, spray-desolvation,or spray
polycondensation [53,54]. When a sparingly water-soluble drug is
embedded in polyvinyl alcohol microspheres, its dissolution rate is
enhanced by increasing its wettability and its total surface area[69]. Such
formulations require a fine dispersion of drug particles in the polyvinyl
alcohol matrix, When a water-soluble drug is embedded in polyvinyl
alcohol particles, its release is sustained by the viscous gel formed at the
microsphere surface [20,29]. If a cross-linking agent is involved in the
preparation, drug release will be inversely proportional to the amount of
cross-linking agents or the dose of initiating UV irradiation.
Polyvinyl alcohol is suitable for various applications because of its high

film strength, flexibility, toughness, abrasion resistance, and adhesive
strength. In the United States in 1976, polyvinyl alcohol was used mainly
in textile warp-sizing and finishing, adhesives, polymerization, and paper-
sizing and coating [64,5,55]. It provides a high weaving efficiency at low
levels, and stiffness for textile warp size and finish. Polyvinyl alcohol,
combined with polyvinyl acetate and starch, is used as an adhesive in bag
making, carton sealing, tube winding, and solid-board lamination. It also
facilitates polymerization in polyvinyl acetate emulsions for adhesive uses.
Polyvinyl alcohol is also used in grease-resistant paper coating and
binding for pigments in certain paper coatings. Miscellaneous other
applications include building products, packaging, chemicals, cosmetics,
ceramics, steel, automotive safety glass, and materials binding [16].
Polyvinyl alcohol finds application in medical technology and surgery.

Polyvinyl alcohol sutures have demonstrated low tissue reaction and loss
of strength. Polyvinyl alcohol is sometimes cross-linked into hydrogels
for special uses in artificial tissues, contact lenses, dressing material, and
embolisms. These materials show low protein absorption, mild tissue
reaction, and advantageous physical properties. However, they must be
washed thoroughly in a sterile solution before use since they are prepared
in formalin. In some cases, a special functional group is added to
polyvinyl alcohol hydrogels for some special functions, such as thrombin
inactivation. Some examples of polyvinyl alcohol applications in medical
areas are listed in Table 12.
8. Pharmacological Effects and Toxicity
Polyvinyl alcohol is not classified as a hazardous material according to the

American Standard for Precautionary Labeling of Hazardous Industrial
Chemicals [ANSI 2129.1- 19761, However, some studies show that
polyvinyl alcohol sponges cause local calcification, fibrosis, and sarcomas
[5,16,32]. Repeated injections of concentrated polyvinyl alcohol solutions
or suspensions can produce toxicities such as glomerulonephritis and
portal hypertension [61,62]. The oral toxicity rate and ecological
information for polyvinyl alcohol is listed in Table 13.
Table 12
Examples of Polyvinyl Alcohol Applications in Medical and

Pharmaceutical Technologies
Application Reference(s)
I Artificial tears and wetting- agents

- for contact lenses I 31,15,23,38
_ . , I
Coat for implanted intraocular lenses 28
Dressing material for donor sites and partial thickness 50
skin-loss bums
Intraluminal intervascular splints to replace sutures for 25
anastomosis
Daily-wear contact lenses 1
Artificial tissues 32-34
Heparin-polyvinyl alcohol hydrogel 2, 36
Intravaginal contraceptive barriers 24
Remover of low-density lipoprotein 27
Particles used in embolization therapy 39,40-46,68
u
Glucose responsive insulin-release system 52
Disposable rubber-like elastic electrode for the 47
electroretinogram
Material for magnetic resonance imaging (MRI) for soft 22
I
tissue
Protective material for unfixed cryostat sections
Mounting medium for preserving fecal material, ova, and
parasites
Medium for embedding hemoglobin and myoglobin for X-
I ray studies
_ _ ~
Material for freeze-fracture technique 37

Medium for electrophoretic size separation of particles 48
Material for inducing disease states of animals to serve as 61, 62
models
Protectants against fluid-mechanical injury of freely- 66
suspended cells
Supporter for immobilizing biocatalysts 67
Table 13
Toxicological and Ecological Information on

Airvol-205 Polyvinyl Alcohol
Acute Toxicity Polyvinyl Alcohol Dose
Oral LD50 (rat) 23,854 mg/kg
I Dermal LD50 (rabbit) I >7490 mgkg I

1 Inhalation L C ~ O (rat)
I 64,000 p p d 4 h
I
LC 5 0 daphnia magna 8300 m g L 96 h
r Environmental Fate
I Chemical Oxygen Demand (COD) I

~~ ~~
1800 mg/g
1 Biochemical Oxygen Demand I BOD5 = 0 - 5%; BOD = 100% I

I Biodegradability I >90% (Zehn-Wellens Test) I
9. REFERENCES
1. M. Kita, Y. Ogura, Y. Honda, S. H. Hyon, W. Cha, and, Y. Ikada,
Graefe's Arch. Clin. Exp. Ophthulmol., 228, 533-537 (1990).
2. I. Strzinar, and M. V. Sefton, J. Biomed. Mat. Res., 26, 577-592
(1992).
3. Air Products, Airvol@ Polyvinyl Alcohol, 1-12 (1993).

4. T . W. Modi, Handbook of Water-soluble gums & resins, In
Davidson, ed., 1980, McGram-Hill Book Company, pp., 20-1 to 20-
31.
5. IARC Monographs on the Evaluation of the Carcinogenic Risk of
Chemicals to Humans, 19,341-366 (1979).
6. N. A. Peppas, and E. W. Merrill, J. Biomed. Muter. Res., 11, 423-
434 (1977).
7. U. S. Pharmacopeia National Formulary, United States
Pharmacopeid Convention, Inc., 23, 1288 (1995).
8. A. 0. Okhamafe, and P. York, J. Pharm. Pharmacol., 37, 385-390
(1984).
9. C. G. Wilson, 0. Olejnik, and J. G. Hardy, J. Pharm. Phamcol.,
35, 451-454 (1983).
10. A. 0.Okhamafe, and P. York, J. Pharm. Sci., 77, 438-443 (1988).

11. A. 0. Okhamafe, and H. U. Iwebor, Phurmuzie, 42, 611-613
(1987).
12. M. J. Reimers, M. J. Cibulsky, and T. A. Barbari, J. Polymer Sci.:
Part B: Polymer Physics, 31, 537 - 543 (1993).
13. P. R. Byron, and R. N. Dalby, J. Pharm. Sci., 76, 65-67 (1987).
14. Y. Choi, and K. Miyasaka, J. Applied Polymer Sci., 48, 313-317
(1993).
15. L. J. Juszczak, B. A, Zilinskas, N. E. Geacintov, J. Breton, and K .
Sauer, Biochimica et Biophysica Acta, 1058, 363-373 (1991).
16. D. K. Gilding, Biocompatibilityof Clinical Implant, In D. F. William
ed., 1981, CRC Press, Florida, 11, pp., 209-232.
17. N. Alkhalili, and B. J. Meakin, J. Pharm. Pharmacol., 29, 35P

( 1 977).
18. L. S. C. Wan, and L. Y. Lim, Drug Dev. & Indust. Pharm., 18,
1895-1905 (1992).
19. P. K. Watler, C. H. Cholakis, and M. V. Sefton, Biomaterials, 9,
150-154 (1988).
20. B. C. Thanoo, M. C. Sunny, and A. Jayakrishnan, J. Phurm.
Pharmcol., 45, 16-20 (1993).
21. D. Miller, and N. A. Peppas, Biomaterials, 7 , 329-339 (1986).
22. I. Mano, H. Goshima, M. Nambu, and M. Iio, Magnetic Resonance
in Medicine, 3, 921-926 (1986).
23, D. J. Nelson, and R. L. Farris, Arch. Ophthalmol., 106, 484-
487 (1 988).
24. M. Chvapil, T. A. Chvapil, J. A. Owen, M. Kantor, J. B. Ulreich,
and C. Eskelson, J. Biomedical Materials Res., 13, 1-3 ( 1979).
25. K. H. Niijima, Y. Yonekawa, H. Handa, and W. Taki, J.
Neurosurg., 67, 579-583 (1987).
26. D. Frackowiak, L. Lorrain, D. Wrobel, and R. M. Leblanc,
Biochem. & Biophy. Res. Comm., 126, 254-261 (1985).
27. N. Maaskant, A. Bantjes, and H. J. M. Kempen, Atherosclerosis,
62, 159-166 (1986).
28. R. J. Olson, H. Kolodner, K. S. Morgan, H. Escapini, D. Sevel,
and H. E. Kaufman, Arch. Ophthalmol., 98, 1840-1842 (1980).
29. L. B. Wingard Jr., T. R. Tritton, and K. A. Egler, Cancer Res.,
45, 3529-3536 (1985).
30. Seidel, J., Pediatrics, 66, 132-134 (1980).
31. D. A. Benedetto, D. 0. Shah, and H. E. Kaufman, Annals of
Ophthal.,April, 537-442 (1978).
32. P. H. B. Bolton-Maggs, and R. W. Motson, Thorax, 34, 561-
562 (1979).
33. M. A. Aboulghar, and Y. El-Kateb, J. Egypt. Med. Assoc., 61,

127-134 (1978).
34. K. Tamura, H. Mizuno, K. Okada, H. Katoh, S . Hitomi, and T.
Teramatsu, Biomat., Med. Dev., Art. Org., 13 (3&4), 133-152
(1985 - 86).
35. T. Y. Teng, and H. W. Huang, Biochimica et Biophysica Acta, 874,

13-18 (1986).
36. M. F. A. Goosen, and M. V. Sefton, J. Biomedical Materials Res.,
17, 359-373 (1983).
37. D. Brown, J. Microscopy, 121, 283-287 (1981).

38. Zaki, P. Fitzgerald, J. G. Hardy, and C. G. Wilson, Pharm.
Pharmacol., 38,463-466 (1986).
39. I. A. Szwarc, C. H. Carrasco, and W. Richli, AJR, 146, 591-592
(1986).
40. K. C. Wright, J. H. Anderson, C. Gianturco, S . Wallace, and V. P.
Chuang, Radiology, 142, 351-354 (1982).
41. S. Kusano, M. Koichiro, H. Ohuchi, 0. Motohashi, and H. Atari,
Investigative Radiology, 22, 388-392 (1987).
42. J. A. Belis, and J. A. Horton, Urology, 19, 224-227 (1982).
43. C. R. Jack, G. Forbes, M. K. Dewanjee, M. L. Brown, and F .
Earnest IV, AJNR, 6, 595-597 (1985).
44. J. I. Ausman, L. H. Gold, S . M. Tadavarthy, K. Amplatz, and S . N .
Chou, J. Neurosurg., 47, 119-125 (1977).
45. A. Berenstein and D. A. Graeb, Radiology, 145, 846 (1982).
46. Sadi, H. I., J. Comp. Path., 87, 503-513 (1977).
47. Y. Honda, N. Nao-i, S. Kim, E. Sakaue, and M. Nambu,
Documenta Ophthalmologica, 63,205-207 (1986).
48. T. Guszczynski, M. M. Garner, M. Deml, and A. Chrambach,
Applied and Theoretical Electrophoresis, 2, 151-157 (1991).
49. Inactive Ingredients for Currently Marketed Products, The Interphaim
Group of Companies, Buffalo Grove, 116 (199 1).
50. J. W. L. Davies, Burns, 10, 94-103 (1984).

51. R. G. Butcher, Histochemical Journal, 14, 165-168 (1982).
52. S. Kitano, Y. Koyama, K. Kataoka, T. Okano, and Y. Sakurai, J .
Controlled Release, 19, 162-170 (1992).
53. C. Voellmy, P. Speiser, and M. Soliva, J. Pharm. Sci., 66, 631-634
( 1 977).
54. T. Y. Ting, I. Gonda, and E. M. Gipps, Pharm. Res., 9, 1330-1335
( 1 992).
55. 0. Augustine, and P. York, J. Pharm. Phurmacol., 37, 492-493

( I 985).
56. N. A. Peppas, and S. R. Stauffer, J. Controlled Release, 16, 305-
310 (1991).
57. Controlled Release, 18, 95-100
N. A. Peppas, and J. E. Scott, .I.
(1992).
58. A. Takamura, F. Ishii, and H. Hidaka, J. Controlled Release, 20,
21-28 (1992).
59. Material Safety Data Sheet (Airvol# 205 PV Alcohol), Air Products
and Chemicals, Inc.
60. In A. R. Gennaro ed., Remington’s Pharmaceutical Sciences, 1990,
Marck and Co., Inc., 18th edition, p., 1307.
61. F. A. Burgener, 0. H. Gutierrez, and G. A. Logsdon, Radiology,
143, 379-385 (1982).
62. J. E. Riviere, G. L. Coppoc, W. W. Carlton, and E. J. Hinsman,
Am J Vet. Res., 41, 502-505 (1980).
63. In S. Hickok ed., Handbook of Pharmaceutical Excipients, American

Pharmaceutical Association, 1994,2nd Edition, pp., 383-384.
64. In S. Budavari ed., The Merck Index, Merck & Co., Inc., 1989, 1lth
edition, p., 1208.
65. In J. E. F. Reynolds ed., Martindale The Extra Pharmacopoeia, The
Pharmaceutical Press, 1993, p., 1031.
66. J. D. Michaels and E. T. Papoutsakis, J. Biotechnol., 19, 241-257
(1 99 1).
POLYVINYLALCOHOL 441
67. H. Ichijo, J. Nagasawa, and A. Yamauchi, J. Biotechnol., 14, 169-

178 (1990).
68. J. J. Kepes, and W. L. Yarde, The Am. J. Surg. Path., 6 , 709-711
(1995).
69. D. Wong, B. Nguyen, and R. Bodmeier, Pharm. Res. Supp., 10,
S276 (1993).
SERTRALINEHYDROCHLORIDE
Bruce M. Johnson and Pei-Tei L. Chang
Central Research
Pfizer Inc
Groton, CT 06340
w4 BRUCE M. JOHNSON AND PEI-TEI L. CHANG
SERTRALXNE HYDROCHLORIDE
1. Introduction
2. Description
2.1. Structural Formula
2.2. Molecular Formula and Molecular Weight
2.3. Nomenclature
2.4. Laboratory Codes
3. Synthesis
4. Physico-Chemical Properties
4.1. Appearance, Color, Odor
4.2. Melting range
4.3. Solubility
4.4. Hygroscopicity
4.5. Ultraviolet Spectra
4.6. Infrared Spectra
4.7. Proton Nuclear Magnetic Resonance Spectra
4.8. Carbon- 13 Nuclear Magnetic Resonance Spectra
4.9. Mass Spectra
4.10. Optical Rotation
4.11. pH andpKa
4.12. Single Crystal X-Ray
4.1 3. Polymorphism
5.2, Ionic Chlorine
5.3. Identification
5.5, Ultraviolet Spectroscopy
5.6. Potentiometric Titration
5.7. High Performance Liquid Chromatography
5.8. Gas Liquid Chromatography
5.9. Biological Fluids
6 . Stability
7. Pharmacokinetics and Metabolism
7.1. Systemic Bioavailability
7.2. Metabolism
SERTRALINE HYDROCHLORIDE 445
1. introduction
Sertraline hydrochloride is an antidepressant for oral administration.
It is chemically unrelated to tricyclic, tetracyclic, or other available
antidepressant agents. It is a novel inhibitor of serotonin reuptake
in the brain'. The mechanism of action of sertraline is presumed to
be linked to its inhibition of CNS neuronal uptake of serotonin
(5HT).Studies at clinically relevant doses in man have
demonstrated that sertraline blocks the uptake of serotonin into
human platelets. Studies in animals also suggest that sertraline is a
potent and selective inhibitor of neuronal serotonin reuptake and
has only very weak effects on norepinephrine and dopamine
neuronal reuptake. In vifrostudies have shown that sertraline has
no significant affinity for adrenergic (alpha 1, alpha 2, beta),
cholinergic, GABA, dopaminergic, histaminergic, and serotonergic
(5HT1A, SHTlB, 5HT2)or benzodiazepine receptors; antagonism
of such receptors has been hypothesized to be associated with
various anticholinergic, sedative, and cardiovascular effects for
other psychotropic drugs. The chronic administration of sertraline
was found in animals to downregulate brain norepinephrine
receptors, as has been observed with other clinically effective
antidepressants. Sertraline does not inhibit monoamine oxidase.*
2. Description
2.1. Structural Formula

Sertraline is the S-cis enantiomer of a disubstituted
tetrahydronaphthalene. The structural formula of sertraline
hydrochloride is given below:
446 BRUCE M. JOHNSON AND PEI-TEI L. CHANG
H-N-CH,
I mHCI
Cl
Sertraline Hydrochloride
2.2. Molecular Formula and Molecular Weight

Molecular Molecular
Formula Weight
Sertraline c1 7H 17Nc12 306.2
Sertraline hydrochloride Cl7H@C13 342.7
2.3. Nomenclature
Chemical Names:
1. ( 1S-cis)-4-(3,4-dichlorophenyl)-1,2,3,4-tetrahydro-N-
methyl-1-naphthalenaminehydrochloride (CAS-79559-97-0)
Preferred use name.
2. cis-( lS,4S)-N-rnethyl-4-(3,4-dichlorophenyl)-1,2,3,4-
tetrahydro- 1-naphthalenamine hydrochloride (J. Med. Chern. 1984,
27(11), 1508)
3. cis-( 1S)-N-methyl-4-(3,4-dichlorophenyl)-1,2,3,4-
tetrahydro- 1-naphthalenamine hydrochloride (US Patent
4,536318)
4. 1-naphthalenamine,4-(3,4-dichlorophenyl)-1,2,3,4-
tetrahydro-N-methyl, hydrochloride, ( 1S-cis)- (USAN chemical
name 1)
5. (lS,4S)-4-(3,4-dichlorophenyl)-1,2,3,4-tetrahydro-N-
methyl- 1-naphthylamine hydrochloride (USAN chemical name 2)
B.A.N.: Sertraline hydrochloride
U.S. A.N. : Sertraline hydrochloride
I.N.N.: Sertraline hydrochloride
Proprietary or Trade Name: Zolofim (USA)

LustralTM(UK)
2.4. Laboratory Codes

Sertraline has the Pfizer code CP-5 1,974
Sertraline hydrochloride has the Pfizer code CP-5 1,974-01
Synthesis
The pure drug substance is manufactured from commonly available
intermediates using a four step process. The structurally significant
starting material is 4-(3,4-dichlorophenyl)-3,4-dihydro-1(2H)-
n a p h t h a l e n ~ n e which
~ s ~ ~is~reacted
~ ~ ~ ~with
~ methylamine to form
the imine. The imine, (N-[4-(3,4-dichlorophenyl)-3,4-dihydro-
1(2H)-naphthalenylidenelmethanamine), is reduced by catalytic
hydrogenation to a pair of racemic diastereomers. The racemic cis
isomers are separated from the racemic trans isomers by selective
crystallization of the hydrochloride salts. The racemic cis
enantiomers are then resolved with D(-)-mandelic acid. In the final
step the mandelate counterion is replaced by chloride to give the
finished drug substance, sertraline hydrochloride.’* lo Crystallization
conditions in the final step determine the polymorphic form
produced.
448 BRUCE M.JOHNSON AND PEI-TEI L. CHANG
4. Physico-Chemical Properties
4.1. Appearance, Color, Odor

Sertraline hydrochloride is a white to off-white, crystalline
powder having no odor. It is an irritant, contact with skin and eyes
should be avoided.
4.2. Melting range

Determination of the melting point by the capillary method
in a Buchi 510 melting point apparatus with a heating rate of 1°C
per minute on Form I shows the onset of change at about 160 to
180°C with movement of the sample and a slight color change. At
about 218°C a partial melt is observed but the sample does not
become clear until a temperature of 245 to 250°C is reached. This
is consistent with the DSC and hot stage microscopic observations.
4.3. Solubility
4.3.1. Aqueous Solubility
The solubility of a saturated solution of sertraline

hydrochloride (Form I) in distilled water at room temperature is 3.8
mg/mL. The pH of this saturated solution is 5.3.
Solubility in water is pH dependent, as the data in the

following table shows. These data were determined by preparing a
saturated solution in distilled water, with pH adjustments made by
addition of sodium hydroxide and/or hydrochloric acid, followed by
appropriate filtering, diluting, and assay by HPLC.
DH Solubility (mg/mL)
1 .o 0.89
2.0 2.98
2.9 3.88
4.2 4.07
6.0 4.46
6.2 4.93
6.3 4.93
6.5 3.32
7.2 0.72
7.7 0.21
8.8 0.02
9.8 0.006
10.2 0.004
11.1 0.004
12.1 CO.004
4.3.2. Solubility in Other Solvents
The approximate solubility of sertraline hydrochloride in

selected solvents at room temperature is given below. These data
were obtained at room temperature, with saturated solutions that
were appropriately filtered, diluted, and assayed by HPLC.
Solvent Solubility (mg/mL,)

Aqueous 0.1 N HC1 0.5 1
Aqueous 0.1 N NaOH 0.002
Ethanol 15.7
Isopropyl alcohol 4.3
Chloroform 110
Acetone 1.1
N,N-dimethy1formamide 88
Dimethylsulfoxide 147
Ethyl acetate 0.20
Acetonitrile 0.85
Methanolic 0.1 N HC1 47
Chloroform/methanol, 1: 1 134
4.4. Hygroscopicity
The hygroscopicity of sertraline hydrochloride (Form I) was
determined by exposing a sample for one week to an atmosphere of
37"C/75% relative humidity, and a sample to room
temperature/88% relative humidity. Karl Fischer titration was used
to determine water content. The results showed water contents of
<O. 1% for the 37"C/75% RH sample and 0.3% for the RT/88% RH
sample. From this data, sertraline hydrochloride is considered to be
non-hygroscopic.
4.5. Ultraviolet Spectra

The ultraviolet absorption spectra of sertraline
hydrochloride were recorded in methanol, methanolic 0.01 N
hydrochloric acid, and methanolic 0.01 N sodium hydroxide. These
spectra are presented as Figures 1,2, and 3.
Absorptivity data for sertraline hydrochloride are given

below:
Solvent &-nm Absorptivity

0.01 N HCYMethanol 265 2.58
273 3.04
28 1 1.64
Methanol 265 2.61

273 3.06
28 1 1.64
0.01 N NaOWMethanol 265 1.92

273 2.26
28 1 1.60
SERTRALINE HYDROCHLORIDE 45 1
The table below demonstrates that the absorptivity and h-

remain effectively unchanged in solutions of differing acid
strengths.
Acid Strength Lax Absorptivity

0.005 N 273 3.03
0.01 N 273 3.04
0.05 N 273 3.1 1
0.10 N 273 3.03
0.50 N 273 3.04
Sertraline hydrochloride also adheres to Beer's Law at this

wavelength, as evidenced by the following data:
Concentration (mg/mL) Lax Absorptivity

0.101 273 2.99
0.201 273 2.97
0.302 273 3.04
0.402 273 2.99
0.503 273 2.99
0.603 273 3.01
0.704 273 2.96
0.804 273 2.97
0.905 273 2.95
1.01 273 2.88
avg. k std. dev. 2.98 f 0.04
452 BRUCE M.JOHNSON AND PEI-TEI L.CHANG
A
b
S
0
r
b
a
n
C
e
Figure 1. Ultraviolet Spectrum of Sertraline HC1 in Methanol
2.0 i
A
b
S
0
r
b
a
n
C
Wawkngth (nm)
Figure 2. Ultraviolet Spectrum of Sertraline HCl in Methanolic HCl

SERTRALINEHYDROCHLORIDE 453
Wavelength (nm)
Figure 3. Ultraviolet Spectrum of Sertraline HCl in Methanolic NaOH
4.6. Infrared Spectra

The infrared spectra of sertraline hydrochloride (Form I) as
a potassium bromide disc and as a Nujol mull were recorded on a
Nicolet 5 10 FTIR spectrophotometer; the spectra in KBr and Nujol
are virtually identical except for the Nujol-related bands. The
spectra are presented in Figure 4. Assignments of absorption
maxima are given below.
Band (approx. cm-1) Assignment
3100 - 3000 (w)* Aromatic C-H stretching vibrations
3000 - 2800 (m)* Aliphatic C-H stretching vibrations
27 10 - 2500 (m) NH+ stretching vibration

2500 - 2450 (m) NH+ stretching vibration
1585 (m) Aromatic C=C skeletal in-plane

vibrations
1560 (m) C-H stretching vibration; aromatic H
1470 - 1450 (s)* Aliphatic C-H deformations, N-CH,

stretching vibration
1400 (s), 1430 (m) Asymmetric CH3 deformation
1375 (m) Symmetric CH3 deformation
1340 (m) Aromatic C-H in-plane deformations

1215 (m)
1135 (s) Ahphatic secondary amine
1060 (m)
1030 (m)
1015 (m) overtones and aromatic C-H in-plane
955 (m) deformations
930 (m)
920 (m)
825 (s)
800 (s)
790 (s) aromatic C-H out-of-plane
760 (s) deformations; C-Cl stretching
710 (m) vibrations
700 (s)
670 (s)
* (w)=weak intensity; (m)=medium intensity; (s)=strong

intensity
The dlfferent polymorphs of sertraline hydrochloride yield

different infrared spectra. Polymorphism is discussed more fully in
section 4.13.
80
70
60
x 50
T
r
a 40
n
S
m 30
I
t
t 20
a
n 10
C
a
0
-10
-20
-30 U
4000 2500 m00 1500 1000 500
Figure 4. Infrared Spectra of Sertraline HCl in Nujol and KBr

4.7. Proton Nuclear Magnetic Resonance Spectra

The 300 MHz proton nuclear magnetic resonance spectrum
of sertraline hydrochloride reference standard was recorded using a
Briicker AM300 NMR instrument. The solvent was deuterated
dimethylsulfoxide. Proton assignments are listed below, and the
spectrum is reproduced as Figure 5.
b a
CI
Chemical Shift Md tiDlicitv # of Protons Assignment

1.9 - 2.4 ppm broad, complex 4 d, e
multiplet
2.6 ppm singlet 3 a
4.1 - 4.2 ppm broad, complex 1 f
multiplet
4.4 - 4.5 ppm broad multiplet 1 C
6.7 - 6.8 ppm multiplet 1 j
7.2 - 7.4 ppm mu1tiplet 2 k, 1
7.35 - 7.5 ppm multiplet 1 1
7.5 - 7.65 ppm doublet 1 h
7.7 ppm doublet 1 g
7.75 - 8.0 ppm multiplet 1 m
9.5 - 9.8 ppm broad singlet 2 b
The peaks at 2.5 and 3.4 ppm are solvent related.
*
n
,a .:. .5 L. u ,a I
. u *. &, <. a. u u u . .* .r ,>
Figure 5. Proton NMR Spectrum of Sertraline HCl in DMSO-d6
4.8. Carbon-13 Nuclear Magnetic Resonance Spectra

The carbon- 13 nuclear magnetic resonance spectrum of
sertraline hydrochloride, Figure 6, was recorded using a Briicker
WM250 instrument. The solvent was deuterated dimethylsulfoxide.
Assignments have been allocated to the spectral lines where
possible by reference to calculated shifts" and to information
gained from low power noise and/or selective decoupling.
Cl
Chemical Shift.ppm Assignment

22.9 d
26.6 C
30.3 a
44.3 e
54.9 b
126.4 0
128.9 i
129.1 n
129.5
129.7
130.2 j
130.5 P
130.9 h,q (by intensity)
131.0 g
139.9 1
146.6 f
The multiplet at 40 ppm is due to the solvent.
Both the proton and carbon- 13 nuclear magnetic resonance

spectra are consistent with the structure for sertraline
hydrochloride.
I 1
I 'l I - I- ' ~I l
180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10
PPm
Figure 6. Carbon-13 NMR Spectrum of Sertraline HCl in DMSO

4.9. Mass Spectra

The mass spectrum of sertraline hydrochloride was recorded
using a Finnigan 45 10 mass spectrometer in the direct exposure
electron impact mode. Instrument conditions are summarized
below:
Inlet system Direct insertion probe

Accelerating voltage 1.8 kV
Electron ionization 70 eV
Ion source temperature 160°C
Resolving power lo00
Calibration perfluorotributylamine (FC-43)
(nominal mass 614)
A representative spectrum is shown in Figure 7. Structure

assignments for a number of the fragments are provided in Figure 8.
These assignments were confmed by high resolution mass
spectrometry, and the structures for the chlorine-containing
fragments are consistent with the observed isotope ratios.I2
A typical spectrum (for peaks with relative intensity >6.0%)

is summarized in the table below. The data are consistent with the
structure for sertraline hydrochloride.
SERTRALINEHYDROCHLORIDE 46 1
Electron Impact Mass Spectrum of Sertraline Hydrochloride
Mass vs Intensity (relative intensity >6.0%)
-
d Z Relative Intensity m/z Relative Intensitv
304 6.5 159 99.4
279 6.1 158 10.4
278 20.0 145 19.6
277 20.5 144 18.8
276 84.7 143 7.6
275 20.6 138.5 8.9
274 100.0 133 44.7
264 21.1 132 44.2
262 33.2 131 11.7
248 8.4 130 22.4
242 6.8 129 33.2
241 11.6 128 30.6
240 7.7 127 14.3
239 28.5 121.5 13.0
238 7.4 120.5 24.8
214 6.9 118 18.4
213 11.2 117 14.7
212 16.7 116 35.3
205 6.4 115 36.4
204 16.7 113 7.9
203 18.0 109 12.3
202 22.1 103 67.6
201 7.0 102 31.5
200 7.1 101 20.3
199 8.1 95 14.4
191 6.3 91 23.8
190 11.1 90 6.2
189 11.7 89 15.2
178 25.2 88 8.8
176 12.4 87 8.8
165 13.6 77 9.4
164 6.4 75 8.3
163 15.8 70 15.8
162 8.0 63 7.5
161 47.6 57 6.3
160 10.3 51 6.4
Figure 7. Mass Spectrum of Sertraline HCl

HN /It +N ’
I+
QC1
\
Vz
m/z 159
c1 274
C1
Figure 8. Mass Spectrometric Fragmentation of Sertraline Hydrochloride
4.10. Optical Rotation

Optical purity is determined by measuring the rotation of a
1% solution of sertraline hydrochloride in methanolic 0.05 N
hydrochloric acid. The specific rotation found for the reference
standard was 40.2".
4.11. pH and pKa

The pKa of sertraline hydrochloride as determined by
potentiometric titration in ethano1:water (1: 1, v/v) was found to be
8.5. The pKa determined in methano1:water (4060, v/v) was 8.6.
Titration of sertraline hydrochloride in water was carried out in the
presence of sodium chloride and a measured excess of HCl.
Titration with NaOH provided a curve that was evaluated by the
method of Clarke and C a h ~ n ' ~ The
. pKa in water calculated by
this method was 9.48 f 0.04.
4.12. Single Crystal X-Ray

Perhaps the single best confirmatory evidence for the
structure of the sertraline hydrochloride (Form I) molecule is
provided by single crystal x-ray. A single crystal of sertraline
hydrochloride was grown and its structure determined on a Nicolet
R3M-m x-ray spectrometer. The resultant structure as determined
is reproduced in Figure 9, which confirms the structure of sertraline
hydrochloride.
Single Crystal X-Ray
Crystallographic Analysis of Sertraline Hydrochloride
Crystal Parameters
Crystal size, mm 0.11xo.11 x0.12
Cell Dimensions a = 8.004(5) 8,

b = 8.372(5) 8,
c = 25.21(2) 8,
a = 90.0"
p = 90.0"
x = 90.0"
V = 1689.3(6) 8,3
Space group ~12121

Orthorhombic
Moleculedunit cell 4
Density observed, g/cm3 1.37
Density calculated, g/cm3 1.354
Linear absorption coefficient 49.48

466 BRUCE M . JOHNSON AND PEI-TEI L. CHANG
CI
Figure 9. Single Crystal X-Ray Structure of Sertraline HCl (Form I)

4.13. Polymorphism
Five crystalline polymorphic forms of sertraline
hydrochloride have been isolated and ~haracterized.'~ Four of the
polymorphs have been examined by infrared absorption
spectroscopy, X-ray powder diffraction, single crystal X-ray,
differential scanning calorimetry (DSC), hot stage optical
microscopy, and aqueous solubility studies. Only a few crystals of
the fifth polymorph have been isolated and the only characterization
data available are from single crystal X-ray studies and infrared
spectra. The polymorphs of sertraline hydrochloride have been
designated as Forms I, 11, 111, IV, and V. Form I represents the
lower melting polymorph which may crystallize from an acidic
solution of the compound in isopropyl alcohol, hexane, or ethyl
acetate depending on temperature and rate of crystallization.
Forms 11 and IV are metastable polymorphs which can be isolated
by rapid crystallization of sertraline hydrochloride from various
solvents (e.g., methanol, ethyl acetate, acetonitrile); however, slow
crystallizations and granulations yield polymorph Form I. Form 111
is generated from Forms I, 11, or IV by heating the solid to
temperatures above about 180°C. Granulating either Form 11, 111,
or IV in isopropanol, ethyl acetate, or hexane at 40 to 60°C causes
conversion to Form I. Form V is produced by sublimation of Form
I onto a cold finger condenser. Form I is the thermodynamically
most stable polymorph at room temperature.
The following information summarizes observations

concerning the characterization, stability, and thermal behavior of
sertraline hydrochloride polymorphs.
4.13.1. Characterization Data
Infrared Absorption Spectrophotometq
The infrared absorption spectra of Forms I, 11, 111, IV,and

V are different from one another (Figure 10). The spectra are
obtained using KBr pellets to suspend the sample during
measurement. Figure 10 displays only the region of 1650 to 400

cm-1. The region from 4000 to 1650 cm-1 is not useful for
comparing polymorphic differences. Qualitative comparison of the
infrared spectra of Form I to Forms 11, 111, IV, and V show the
following differences:
a. At approximately 740 to 750 cm-l and 1075 to 1085

cm-1 Forms 11, 111, IV, and V exhibit absorption bands of
significantly higher relative intensity.
b. At approximately 780 cm-l Forms 11, 111, and IV exhibit

major peaks. Forms I and V exhibit a major peak at approximately
790 cm-1with only a shoulder at about 780 cm-1.
c. Forms 11, 111, IV, and V show absorption bands at

approximately 870 and 520 to 540 cm-1 while no absorption is
observed for Form I at these wavelengths.
d. Form 11 shows a strong absorption band and Forms 111,

IV, and V a weak band at about 640 cm-l, while a barely detectable
absorption is observed for Form I.
e. Differences for Forms 11, 111, and IV are also observed

in wavelength and relative intensity of the absorption bands in the
region of 800 to 850 cm-1 compared to Forms I and V.
Single Crystal X-Ray Analysis
Absolute structure determination of Forms I, 11, HI, IV, and

V by single crystai X-ray analysis shows that the polymorphic forms
differ in rotational conformation at the methylamino and the
dichlorophenyl positions. Forms 111and IV are similar rotationally
but differ in their space group. Using these configurations the
crystal packing diagrams are constructed and the crystal densities
are calculated. Form I gives a density of 1.354 gkc, Form I1 gives
a density of 1.314 g/cc, Form 111gives a density of 1.3 13 g/cc,
z
T
I
a
n
S
m
I
t
t
a
n
C
8
-120 4 Y
1600 1500 1400 1300 1200 1100 1mo 900 800 700 600 500
Figure 10. Infrared Spectra of Sertraline HCl Polymorphs (Wavelength range 1650 to 400 cm-1)
470 BRUCE M. JOHNSON AND PEI-TEIL. CHANG
Form IV gives a density of 1.349 g/cc, and Form V gives a density

of 1.308 g/cc. This result supports Form I as the
thermodynamically stable form at room temperature since the most
dense crystal at a given temperature is considered to be the most
thermodynamically stabiei5.
Using these single crystal data it is possible to calculate the

theoretical powder diffraction patterns. The theoretical patterns
shown in Figure 1 1 for Forms I, II, III, and IV match the observed
patterns very closely.
X-Ray Powder Diffraction
The four polymorphs of sertraline hydrochloride for which

samples are available give distinctive X-ray powder diffraction
patterns. Each form has diffractions at unique values of 28 that are
diagnostic:
Form1 7.1, 12.7, 14.1, 15.3, 15.7,21.2,23.4, and26.3
Form11 5.4, 10.8, 14.6, 16.3, 18.1, 19.0,20.3,21.8,

24.4, and 27.3
FormIII 14.3, 15.5, 17.4, and 19.6
Form IV 15.6,22.4,25.4,28.9,31.9,and 32.1
In these powder diffraction patterns of the pure polymorphs,

each of the unique peaks is greater than about 10%of the
normalized intensity. In normal practice the 28 values remain
constant but the relative intensities may vary somewhat due to
particle size effects.
Figure 1 1 shows the powder diffraction patterns calculated

from the single crystal data. The pattern for Polymorph V is also
included here and shows a distinctive pattern as well.
SERTRALINE HYDROCHLORIDE 47 1
POLYHORPH I
Figure 11. Calculated X-Ray Powder Diffraction Patterns
4.13.2. Relative Stability of Sertraline Hydrochloride Polymorphs
Interconversion Experiments
Forms 11and TV have been made by rapid crystallization out

of solvents and Form 111has been made by heating a sample of
Form I at about 180°C for a period of several hours. When
mixtures of Form I and Forms 11, 111, or IV were granulated in
isopropyl alcohol at 50°C, the resulting solid was shown to be
Form I. Similar behavior has been observed in other solvents.
In an experiment viewed under the microscope at room

temperature, a mixture of Forms I, 11, and 111was placed on a slide
and covered with a cover slip. Isopropanol saturated with sertraline

hydrochloride was added at the edge of the cover slip and the
sample was kept wet with solvent throughout the observation.
Crystals of Form III disappeared fairly rapidly. Crystals of Form II
shrunk in size slowly while the Form I crystals increased in size. In
a similar experiment with Forms I and IV,Form IV dissolved
slowly while Form I grew. These observations are consistent with
the greater thermodynamic stability of Form I.
Solubility
The room temperature aqueous solubility of the four

polymorphs is essentially equivalent at about 4 m u d .
Polymorphs 11,III, and IV appear to be slightly more soluble than
Form I, further indicating that Form I is the most stable at room
temperature.
4.13.3. Thermal Behavior
Differential Scanning Calorimetry CDSC)
The DSC thermograms for Forms I - IV are shown in

Figures 12, 13, 14, and 15. The features in the thermograms may
be interpreted as follows:
Form I shows an endotherm with an onset temperature of

about 219°C resulting from the melting of Form I. An exotherm
immediately following the melting endotherm of Form I appears at
about 225°C caused by partial crystallization of the melt to Form
III, and a second endotherm appears with an onset at about 246°C
resulting from the melting of Form III. Rapid thermal
decomposition takes over after the final melt of Form III causing
the baseline to increase rapidly and become erratic in appearance.
On rare occasions, the DSC of Form I shows only a single melting
endotherm with an onset temperature of about 219°C. This can
occur when the sample does not have sufficient time or seed to
recrystallize as Form 111prior to reaching the Form 111melting

temperature.
In the thermograms obtained for Forms 11and IV,a very

small endotherm may be observed with an onset temperature at
about 180°C which corresponds to a solid-solid transition to Form
111. This event is followed by melting of Form 111at about 246°C.
The thermogram obtained for Form 111 shows a single

endotherm at about 246°C. Above the melting temperature of 246"
C thermal decomposition increases rapidly.
The thermal events discussed above demonstrate that Form

111undergoes only a single melting process at 246°C followed by
increased thermal decomposition after the melt has occurred.
Thermogravimetry and hot stage optical microscopy reveal that all
four forms undergo slight decomposition and sublimation at
temperatures above about 160°C.
Prolonged heating studies on the Form Worm 111

conversion in the solid state have shown that below 160°C Form I
is the preferred polymorph. Above 180°C Form I converts to Form
111in the solid state. This conversion takes place at varying rates in
the DSC experiment depending on sample characteristics such as
particle size, previous handling, or the presence of Form III seed.
Thus, in some instances the first endotherm in Form I may show a
smaller endotherm with a lower onset temperature as a result of
solid state conversion of Form I to Form 111during heating. The
solid state conversion is a very low energy transition and fails to
give a reliably detectable DSC event. A sample of Form I spiked
with 3% of Form 111(Figure 16) shows an endotherm for the solid
state transition and the absence of the Form I melting endotherm.
The Form 111endotherm appears at the expected temperature of
about 246°C. Under the conditions of this DSC experiment the
solid state transition is fast enough to cause complete conversion to
Form 111before the Form I melt temperature is reached. Similar
474 BRUCE M.JOHNSON AND PEI-TEI L.CHANG
01
Figure 12. DSC Thermogram of Sertraline HCl Form I)
I
V
c m
aa
Figure 13. DSC Thermogram of Sertraline HCl (Form II)

Figure 14. DSC Thermogram of Sertraline HCl (Form III)
Figure 15. DSC Thennogram of Sertraline HCl (Form IV)

LOO
U
LSD
-I
0.w
Figure 16. DSC Thermogram of Sertraline HCl

(Form I containing 3% of Form 111)
behavior was observed in two other samples of Form I spiked with

5% of Form I1 or 5% of Form IV.
In a related experiment, small amounts of each of Forms I,

11, and IV were heated to 185°C for 15 minutes. The infrared
spectra of each sample confmed that they had converted to Form
111.
Hot Stage Optical Microscopy
Microscopic observation of a Form I sample heated on a hol

stage is consistent with the observed DSC behavior. At about 150
to 180°C the sample begins to move about and appears to "pop".
This may be caused by the onset of decomposition, solvent release,
or sublimation. Partial melting may be observed at about 210 to
220°C with the formation of some new crystals (Form 111) as the
temperature is raised to about 235°C. (In some instances, a solid-
solid transition to Form I11 is observed at temperatures above about
180°C.) At about 245 to 250°C the newly formed crystals melt

completely. The melt solidifies very slowly upon cooling and
usually remains as a yellowish glass. This suggests that significant
decomposition has taken place.
In mixtures of Form I11 in Form I, the first melt at 210 to

220°C does not cause all of the sample to liquefy, a small number
of crystals may remain and increase in size as Form 111crystallizes
around them. Holding the temperature at about 235°C permits the
melt to recrystallize to Form III. Cooling this melt to room
temperature and subsequent reheating does not show the Form I
melt at 210 to 220°C. This suggests that the conversion to Form
111is not rapidly reversible under these conditions.
Microscopic observation of Forms 11or IV on the hot stage

shows an apparent solidkolid transition at about 180°C without
evidence of melting. Continued heating gives a complete melt at
about 245°C. The temperature region of 180 to 230°C appears
also to cause decomposition and sublimation.
Microscopic observation of Form 111 shows sample

movement similar to Form I at about 160°C followed by melting at
about 245 to 250°C. No other thermal events are seen and the
observation is consistent with the DSC behavior.
When samples are heated in silicone oil the appearance of

gas bubbles is observed starting at around 150 to 180°C suggesting
the onset of decomposition. The yellowish color of the melt above
250°C results from chemical decomposition of the sample. This
phenomenon occurs for all four polymorphic forms.
47% BRUCE M. JOHNSON AND PEI-TEI L. CHANG
4.13.4. Literature Precedent
Polymorphs of sertraline hydrochloride behave similarly to

those of gepirone hydrochloride which has been characterized and
discussed in the 1iteraturel6. Both compounds have a polymorph
which is most thermodynamically stable at room temperature which
can go through an endothermic melt and subsequently recrystallize
as a higher melting polymorph.

The elemental composition of sertraline hydrochloride was
determined by a series of microanalytical determinations. A typical
set of analyses is shown below.
Assay
% carbon 59.54 59.58

% hydrogen 5.30 5.29
% nitrogen 3.85 4.09
% chlorine, ionic 10.24 10.35
% chlorine, total 30.54 31.04
Total (C,H,N,Cl) 99.23 100.00
The results for percentage carbon, hydrogen, nitrogen, and

chlorine for sertraline hydrochloride are consistent with the
theoretical values for the formula C17HIgNC13.
5.2. Ionic Chlorine

The ionic chlorine content of sertraline hydrochloride may
be determined by potentiometric titration with a standard silver
nitrate solution, which precipitates the chloride ion as AgC1.
Titration is carried out by dissolving the sertraline hydrochloride in
ethanovwater (1: 1, v/v) and titrating using a silver measuring
electrode and a Ag/ AgCl reference electrode. The theoretical ionic
chloride content of sertraline hydrochloride is 10.35%;the amount

found for the reference standard lot was 10.24%.
5.3. Identification
Sertraline hydrochloride is identified in the bulk form using
the infrared spectrum, the TLC retention, and the HPLC retention
by comparison with a working standard. In dosage forms the TLC
retention and the HPLC retention are used for confirming identity
of drug substance.

Thin layer chromatography is used as a technique for
establishing the identity of sertraline and for examining for the
presence of process related substances and other potential
impurities. The following systems have been used with silica gel 60
F-254 (Merck Darmstadt) plates.
Solvent System Rf
- Detection
Ethyl acetate/Methanol/Ammonium acetate 0.55 1

160:40:2 v/v
Chloroform 0.00 2
Chloroform/Methanol/ammoniumacetate 0.91 2
12050:15 v/v
Detection Systems
1 Dragendorff's Reagent
2 Ultraviolet light - 254 nm
5.5. Ultraviolet Spectroscopy

Assay for sertraline content of drug substance and drug
products may be carried out using ultraviolet spectrophotometry.
480 BRtiCE M. JOHNSON AND PEI-TEI L. CHANG
Sertraline solutions are prepared in 0.01 N methanolic hydrochloric

acid at a concentration of approximately 0.05 mg/mL. Absorbance
measurements are taken at 273 nm and compared to the absorbance
of a well characterized standard.
5.6. Potentiometric Titration

Being a hydrochloride salt, sertraline hydrochloride has an
acidic proton which can be titrated with 0.5 N aqueous sodium
hydroxide in the solvent mixture ethanovwater, 1:1, v/v.
When 98.67 mg of sertraline hydrochloride reference

standard was titrated in 20 mL of the solvent mixture, a single
inflection point was observed. This was attributed to the single
acidic proton located on the arnine nitrogen of the sertraline
hydrochloride molecule. The neutralization equivalent (molecular
weight) calculated for the single inflection point was 345.5. Based
on this result, an assay of 89.3% sertraline was calculated for the
reference standard lot, which is 99.9% of theory (89.4%).
This titrimetric result confirms the stoichiometry of

sertraline hydrochloride and provides an equivalent weight value
consistent with the structure. A representative curve for this
titrimetric procedure is included in Figure 17.
5.7. High Performance Liquid Chromatography

Purity assay of sertraline hydrochloride is done using HPLC
by the external standard method by comparing chromatographic
response with a well characterized standard. An isocratic reverse
phase system using a C-18 column and UV detection at 254 nm is
suitable for both drug substance and formulated products.
Column: Waters Nova-Pak C- 18

Detector: UV at 254 nm
Mobile Phase: Acetonitrile/MethanoYSuffer 45: 15:40 v/v
Buffer: 0.05 M Acetic acid/O.OZ M Triethylamine
(aqueous)
PH
9 1 2 3 4 6 6 7 8 P 1 9 1 1 1 2 1 1 1 4
1 1 1 1 l l 1 1 1 1 l l l l 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
0 -
-
0.10 -
-
0.20 -
-
0.30 -
-
0.40 -
mL -
0.60 -
0.60 -
0.70 -
0.80 -
Figure 17. Potentiometric Titration of Sertraline HCl in EthanoVWater

(1:l)
4x2 BRUCE M. JOHNSON AND PEI-TEI L. CHANG
Metabolic studies have used HPLC to measure sertraline

and its metabolites in urine, bile, and feces using radiolabeled drug17
and a radioactivity detector and/or a variable wavelength UV
detector’*. Two different mobile phases were reported.
Column: Waters C - 18 pBondapak

Detector: Berthold radioactivity monitor or UV
Mobile Phase 1 : Acetonitrile/SO mM Sodium phosphate (pH 4.5)
5050 v/v
Mobile Phase 2: AcetoniUilelSO mM Ammonium acetate (pH 5.0)
33167v/v
Another HPLC method has been reported for measuring

sertralirie and desmethylsertraline in mouse cerebral cortex using an
internal standard method”.
Column: Versapack C- 18
Detector: UV at 235 nm
Mobile Phase: Acetonitrile/0.25 M Potassium phosphate (pH
2.7) 30:70 v/v
5.8. Gas Liquid Chromatography

Gas liquid chromatography is used to separate sertraline
from its process related substances, in particular the dechlorinated
homologs. Hydrochloride salts are converted to their free base
form by extraction of a basified solution into methylene chloride
prior to injection. The following system has been used for the drug
substance.
Column: 5% OV-17 on Chromosorb W (acid washed and

DMCS treated) packed in a 7 ft. x 0.25 in. OD
glass column
Column Temp: 225°C
Injector Temp: 250°C
Detector Temp: 300°C
Detector: Flame ionization
Gas chromatography has also been used to measure

sertraline levels in biological fluids. The drug is extracted from
plasma and derivatized with trifluoroacetic anhydride. The
derivative is chromatographed using either a mass spectrometric
detector2' or a 63Ni electron capture detector21.
GCMS Method
Column: 3% Silar 1OC on Gas Chrom Q (80 - 100 mesh)
packed in a 0.8 m x 2 mm ID glass column
Column Temp: 255°C
Transfer Temp: 290°C
Detector: Mass spectrometer
Electron Capture Method

Column: SE-54,0.25 pm film thickness, 12 m x 0.32 mm
ID capillary column
Column Temp: 165°C for 0.5 min, 20"C/min to 21OoC, hold for
12 min
Detector Temp: 300°C
Detector: 63Ni Electron Capture
5.9. Biological Fluids

The levels of sertraline and its major metabolites are
measured in biological fluids by HPLC, GC/EC, and GCMS using
the methods described above. Metabolism studies have resulted in
the identification of N-desmethylsertraline, sertraline carbamoyl-0-
glucuronide, N-hydroxysertraline, a ketone [4-S-(3,4-
dichlorophenyl)-3,4-dihydro-1(2H)-naphthalenone], and the a-
hydroxyketone [4-S-(3,4-dichlorophenyl)-3,4-dihydro-2-R,S-
hydroxy-1(2H)-naphthalenone] as well as the glucuronides of N-
hydroxysertraline and the a-hydroxyketone2*.
6. Stability
The stability of sertraline hydrochloride has been studied under
extreme challenge conditions as well as the more traditional
pharmaceutical challenge conditions. The extreme conditions
included exposing excess drug substance to refluxing water for 3
hours, refluxing 5 N hydrochloric acid for 3 hours, refluxing 5 N
NaOH for 3 hours, or 10% hydrogen peroxide for 6 hours. The
more traditional stability challenge conditions of 3 months at 50°C,
12 months at 37"C, 60 months at 30°C, or 3 months in a light
cabinet were used with different packaging systems. Careful
examination of the resulting samples revealed no significant
degradation under any of the conditions used in these studies. The
only trace degradation product identified was the ketone [4-S-(3,4-
dichlorophenyl)-3,4-dihydro-1(2H)-naphthalenone] which has also
been identified as a metabolite. Sertraline hydrochloride appears to
be a very stable compound under a variety of challenge conditions.
7. Pharmacokinetics and Metabolism
7.1. Systemic Bioavailability

In man, following oral once-daily dosing of sertraline
hydrochloride (Form I) over the range of 50 to 200 mg for 14 days,
mean peak plasma concentrations (C,) of sertraline occurred
between 4.5 to 8.4hours postdosing. The average terminal
elimination half-life of plasma sertraline is about 26 hours. Based
on this pharmacokinetic parameter, steady-state sertraline plasma
levels should be achieved after approximately one week of once-
daily dosing. Linear dose-proportional pharmacokinetics were
demonstrated in a single dose study in which the C, and area
under the plasma concentration time curve (AUC) of sertraline
were proportional to dose over a range of 50 to 200 mg.
Consistent with the terminal elimination half-life, there is an
approximately two-fold accumulation, compared to a single dose,
of sertraline with repeated dosing over a 50 to 200 mg dose range.
The single dose bioavailability of sertraline tablets is approximately
equal to an equivalent dose of solution23.
The effects of food on the bioavailability of sertraline were

studied in subjects administered a single dose with and without
food. AUC was slightly increased when drug was administered
with food but the C, was 25% greater, while the time to reach
peak plasma concentration decreased from 8 hours post-dosing to
5.5 hours.
7.2. Metabolism
Sertraline undergoes extensive first pass metabolism. The
principal initial pathway of metabolism for sertraline is N-
demethylation. N-desmethylsertraline has a plasma terminal
elimination half-life of 62 to 104 hours. Both in vitro biochemical
and in viva pharmacological testing have shown N-
desmethylsertraline to be substantially less active than sertraline.
Both sertraline and N-desmethylsertraline undergo oxidative
deamination and subsequent reduction, hydroxylation, and
glucuronide conjugation. In a study of radiolabeled sertraline
involving two healthy male subjects, sertraline accounted for less
than 5% of the plasma radioactivity. About 40-45% of the
administered radioactivity was recovered in urine in 9 days.
unchanged sertraline was not detectable in the urine. For the same
period, about 40-45% of the administered radioactivity was
accounted for in feces, including 12-14%unchanged sertralineZ4.
Desmethylsertraline exhibits time-related, dose dependent

increases in AUC (0-24 hour), C, and Cmin, with about a 5-9
fold increase in these pharmacokinetic parameters between day 1
andday 14.
Acknowlegement
The authors wish to thank our many Pfizer colleagues in the Central
Research laboratories in both Groton, CT, USA and Sandwich, Kent, UK
who have contributed to the development of sertraline hydrochloride and
to the information presented in this overview.
486 BRUCE M.JOHNSON AND PEI-TEI L. CHANG
References
'Koe, B. K.; Weissman, A.; Welch, W. M.; Browne, R. G. J. Exp. Ther. 1983,226,
686.
2Package Insert, Zoloft*, Pfizer Inc, Jan. 1992
3Quallich, G. J.; Williams, M. T. U.S. Patent 4 777 288, 1988.
'Quallich, G. J.; Williams, M. T. U.S. Patent 4 839 104, 1989.
5Williams, M. T.; Quallich, G. J. Chemistry and Industry 1990,21, 315.
6Adrian, G. P. U.S. Patent 5 019 655,1991.
'Quallich, G. J.; Williams, M. T. European Patent 0 295 050 A l , 1988.
'Quallich, G . J.; Williams, M. T.; Friedmann, R. C. J. Org. Chem. 1990,55,4971.
w e l c h , W. M.; Kraska, A. R.; Sarges, R.; Koe,B. K. J. Med. Chem. 1984,27, 1508.
'()Welch, Jr., W. M.; Harbert, C. A.; Koe, B. K.; Kraska, A. R. U.S. Patent 4 536 518,
1985.
"Ewing, D. F. Organic Magnetic Resonance 1979, 12(9), 499.
12
Sharp, T. R.; Horan, G. J.; Day, S.V.O. Proceedings of the 41st ASMS Conference 01
Mass Spectrometry and Allied Topics, San Francisco, California, 1993.
I3Clarke, F. H.; Cahoon, N. M. J. Pharm. Sci. 1987,8,611.
'*Sysko, R. J.; Allen, D. M. J. US.Patent 5 248 699, 1993.
"Haleblian, J.; McCrone, W. J. Pharm. Sci. 1969,58(8), 911.
%ehme, R. J.; Brooke, D.; Farney, R. F.; Kensler, T. T. J. Pharm. Sci. 1985, 74(10),
1041.
"Welch, W. M.; Vivieros, D. M. J. Labelled Compd. Radiopkam. 1987, 24, 987.
18
Tremaine, L. M.; Stroh, J. G.; Ronfeld, R. A. Drug Metab. Dispos. 1989, 17,58.
'weiner, H. L.; Kramer, H. K.; Reith, M. E. A. J. Chromatog. 1990,527,467.
2%ouda, H. 6.; Ronfeld, R. A.; Weidler, D. J. J. Chromatog. 1987,417, 197.
"Tremaine, L. M.; Joerg, E. A. J. Chromatog. 1989,496,423.
2?remaine, L. M.; Welch, W. M.; Ronfeld, R. A. Drug Metab. Dispos. 1989, 17, 542.
23PackageInsert, Zolofi@,Pfizer Inc, Jan. 1992
24
Package Insert, Zoloft@, Pfizer Inc, Jan. 1992
SOLASODINE
Gunawan Indrayanto, Achmad Syahrani,
Robby Sondakh, and Mulja H. Santosa
Laboratory of Pharmaceutical Biotechnology
Faculty of Pharmacy
Airlangga University
Surabaya, Indonesia
ANALYTlCAL PROFILES OF DRUG SUBSTANCES 487 Copyright 0 1996 by Academic Press, lnc.
388 GUNAWAN INDRAYANTO ET AL.
1. DESCRIPTION
1.1. Nomenclature
1.1.1. Chemical names
1.1.2. Synonym
1.2. Formula
1.2.1. Empirical
1.2.2. structural
1.2.3. CAS Registry No.
1.3. Molecular weight
1.4. Elementary composition
13.Appearance
1.6. Occurence
1.7. Use
2.1. Melting point
2.2. Specific rotation
2.3. Solubility
2.4. Dissociation constant
2.6. Spectm Properties of solasodine
2.6.1. Ultra Vioiet Spectrum
2.6.2. Absorbance Reflectance Spectrum
2.6.3. Near I n h Red Spectrum
2.6.4. Infra Red Spectrum
2.6.5. 'H-Nuc~EKMagnetic Resonance Spectrum
2.6.6. '3C-Nuclear Magnetic Resonance Spectrum
2.6.7. Mass Spectrum
3. ISOLATION OF SOLASODINE
4. BIOSYNTHESIS OF SOLASODINE
5. CHEMICAL CONVERSION OF SOLASODINE TO STEROID
HORMONES
6. BIOTRAWFORMATION OF SOLASODINE
SOLASODINE 489
7. METHOD OF EXTRACTION AND QUANTITATNE ANALYSIS

OF SOLASODINE
7.1. Method of extraction
7.2. Method of quantitative analysis
7.2.1. Titration
7.2.2. Colorimetric
7.2.3. Potentiometric
7.2.4. Gas chromatography
7.2.5. High Performance Liquid Chromatography
7.2.6. Radioimmunoassay
7.2.7. Thin Layer Chromatography - Densitometric
8. PHARMACOLOGICAL AND BIOLOGICAL ACTIVITIES

OF SOLASODINE AND ITS GLYCOSIDES
1. DESCRlPTION
1 . 1 . Nomenclature
1.1.1. Chemical names
Spirosol-5-en-3P-01; solasod-5en-3P-ol
1.1.2. Synonym
Solancarpidine; Solanidhe-s; Purapuridine
1.2. Formula
1.2.1. Empirical
C27H4,No,
1.2.2. smctural
1.2.3. CAS Registry No.

126-17-0
1.3. Molecular weight

413,6
1.4. Elementary composition

C : 78.40 96; H : 10.48 96; N : 3.39 96; 0 : 7.74 96
1.5. Appearance
A fine white odorless crystalline powder; hexagonal plates crystal
of solasodine can be recrystallized from methanol or by sublimation
in high vacuum
SOLASODINE 49 1
1.6. Occurence
Solasodine was accumulated in the species of Genus Solanum.
Solasodine was produced commercially mostly from Solanwn
laciniatwn, Solanum khasianwn and Solanwn marnmoswn (1,2,3). In
Solanum plants, solasodine was found mostly as diglycosides,
triglycosides and tetraglycosides (4). The sugar moiety of the
important glycosides are shown in the table 1.
Table 1. Sugar moiety of solasodine glycosides
Glycosides Sugar moiety I

J3+alamarginc Rhamnose 01 (1-> 4) - Glucose p (1-> 3) - R
p-Solasonine Glucose p (1-> 3) - Galactose p (1-> 3) - R I

Solamargine
Rhamnose a (1-> 2)
Rhamnose a (1-> 4)
> Glucose p (1-> 3) -R
Rhamnose a (1-> 2)
Solasonine >&lactose p (I-> 3) - R
Glucose p (1-> 3)
Solaradixine Rharnnose a (1-> 2)
>Galactose p (1-> 3)-R
p (1-> 2 G I U C Op~(1-> 3)
G~UCOS~
R = Solasodine
1.7. Use
Solasodine is used as raw material for producing steroid hormone in
pharmaceutical industry.
2.1. Melting point
200 - 202O c (5)
2.2. Specific rotation

[ aD ] at 25OC (c = 0.14 in methanol) = -98' ;
[ aD ] =-113' (in chloroform) (5)
2.3. Solubility
The solubility data of solasodine are listed in table 2
Table 2. Solasodine solubility in various solvents
Freely soluble.
Benzene Freely soluble.
Methanol 9.5
Ethanol 95 % 5.0
Acetone 3.5
n-Hexane <1.0
Water <1.0
Ether Practically insoluble'
* Data from Merck Index (5)

2.4. Dissociation constant
pK, in 60 % ethanol was 6.31 (6).

The DSC (differential scanning calorimetry) thermogram of
solasodine was obtained on a Shimadzu DT-30 Thermal Analyzer. The
instrument was calibrated with indium standard. The heating rate was
10°C/min. The thermogram is presented in figure 1.
The DSC curve revealed an endothermic residual moisture peak on a
single sharp endothermic melting peak (200OC) of solasodine.
2.6. Spectral Properties of solasodine

2.6.1. Ultra Violet Spectrum
The U V (ultra violet) spectrum of solasodine in methanol (ca. 40
ppm) was obtained on a Hewlett Packard 8452A photo diode array
spectrophotometer. It exhibits a maximum at 206 nm. The UV spectrum
and its first order derivative spectrum are presented in figure 2A and 2B.
2.6.2. Absorbance Reflectance Spectrum

The visible absorbance reflectance spectrum of solasodine spotted on
Kieselgel GF254 precoated plate (EMerck), eluted with chloroform:
Figure 1. DSC thermogram of solasodine (Sigma).
I 0.72649, I
i I A A
%z 0.43589-
2c:
20.29059-
m
<
I 0.14530/ I \ nm,
200 2.50 300 350 400
B
0.06829-
-w 0.03927
>
pO.OIa26
U
Figure 2. UV spectrum of solasodine (Sigma)in methanol (A) and

its fxst order derivative spectrum (B).
2.000
1.600
* 1.200
sr
3
0.800
8
0.400
81
O.Oo0
- 0.200 - I , . . . , . . . . I . , I . I . . . - 1 . I -.
400 450 500 550 600 650 K1
nm
Figure 3. Visible absorbance ref'lectantce spectrum of solasodine (Sigma)
on Kieselgel GF254 precoated plate (Merck). BL: base line.
methano1:diethyl amine (20:2:0.5) and Visualized by anise

aldehyde-sulphuric acid (7)spray reagent (10OoC,10 minutes), is shown
in figure 3. It shows a maximum at 385 nm. The spectrum was recorded
on a Shimadzu CS-930TLC Scanner.
2.6.3. Near Infra Red Spectrum

The NIR (near infra red spectrum) of 0.5 % solasodine in CCl, was
recorded by using a Shimadzu 365 spectrophotometer. The N I R spectrum
of solasodine is shown in figure 4. The Xmx data are presented in table 3.
Table 3. NIR characteristics of solasodine
Amax (nm) Assignment

1413 OH stretching
1527 NH stretching
2309
2352 CH stretching
2.6.4. Infra Red Spectrum

The IR (infra red) spectrum of solasodine as KBr disc (1.5mg/ 100
mg) was recorded on a Jasco 5300 FT IR spectrophotometer.
The IR spectrum of solasodine is shown in figure 5. The characteristic
bands of solasodine are given in the table 4.
Table 4. IR characteristic bands of solasodine
I
~~ ~
Frequencies (crn-') Assignment

3455 Stretching vibration of OH
3362 Stretching vibration of NH
2843,2967 Stretching vibration of CH steroid skeleton
1690 Stretching vibration of C = C bond
1675 Deformation vibration of NH
1451, 1344 Deformation vibration of methyl
1379 Deformation vibration of methylene
1060, 1239 Stretching vibration of C-QC
0
9"
0
c
5 '0
Q F!
3
ABSORBANCE
-
100.00
%T
0.001.. 1 . . . . I . . . . I I (cm-')
4600.0 4000.0 3000.0 2000.0 1ooo.o 400.0
Figure 5. IR spectrum of solasodine (Sigma) in KBr disc.

SOLASODINE 499
2.6.5. 'H-Nuclear Magnetic Resonance Spectrum

The 'H-NMR spectrum of solasodine (Figure 6A and 6B) were
determined in CDCl, at 90 MHz employing a Hitachi R-1900 FT NMR
using TMS as internal standard. The chemical shift values are presented in
Table 5.
Table 5. 'H-NMR chemical shifts of solasodine
Chemical shift (6 ppm)

Proton assignment
CDCI, Pyridine - D5
Me-18 0.82 (s) 1.05 (s)
Me-19 1.03 (s) 0.91 (s)
Me-2 1 0.98 (s) 1.13 (s)
Me-27 0.88 (s) 0.84 (s)
H-6 5.32 (br.d)
H-3a 3.50 (m) 3.80 (m)
H-16 4.27 (4;6.5,6.5, 6.5 I+) 4.43 (4;6.7, 6.7, 6.7 Hz)
H-26a, p 2.57-2.65 (m; not resolved)
H ~ u P, 2.21-2.30 (m;not resolved)
By using two-dimensional NMR technique (COSY,one bond, long range

'H-I3C chemical shift correlation and NOESY),obtained on a Bruker
AMX-500 NMR spectrometer (500MHz), Puri et al (8) could elucidated
all the proton chemical shifts of solasodine. The data are presented in the
table 6.
5x
-.. .... . . , , ..,. ._.. ' . . . ' I . . . ' -

5.00 4.00
3.00 2.00 '.OO PPM
Figure 6A. Proton NMR spectruni of solasodine (Sigma) in CDCh

501
502 CUNAWAN INDRAYANTO ET AL.
Table 6. 'H-NMR chemical shifts of solasodine (6 ppm, CDCl,)*

_e_l
Proton assignment
-
Chemical shift 1 Proton assignment Chemical shift
l-a 1.oo 15- 1.95
1-P 1.80 1543 1.25
2- 1.78 16 4.20
2-9 1.40 17 1.65
3 3.42 Me-18 0.74
4-a 2.24 Me-19 0.95
4-P 2.16 20 1.81
6 5.23 Me-2 1 0.88
7* 1.48 23 1.56
7-P 1.97 24- 1.55
8 1.58 24-P 1.35
9 0.90 25 1.48
11 1.48; 1.41+ 26- 2.58
12-a 1.10 26-P 2.52
12-P 1.68 Me-27 0.77
14 1.02
~.
+) not resolved
*) Data from Puri et al. (8)
2.6.6. '3C-Nuclear Magnetic Resonance Spectrum

The broad band decoupling 13C-NMR spectrum (Figure 7) and the J
modulation spin echo 13C-NMR spectrum (Figure 3) with 1/J = 7 p sec
of solasodine in CDCl were recorded on a Hitachi R-1900 FT NMR
(22.6MHz) using TMZ as internal standard. The spectral assignments are
presented in Table 7. The data are identical with previously reported
spectrum (9, 10).
TMS
J! L L
Figure 7. Broad band decoupled 13C-NMR spectrum of solasodine

(Sigma) in CDC13.
PPM
CH,CH3
C,CHz
CDCh
J 1 TMS
50.00 00.00
PPM
Figure 8. J modulation spin echo (1/J = 7 ItSec) I3C-NMR spectrum
of solasodine (Sigma) in CDC13.
SOLASODINE 505
Table 7. 13C-NMR chemical shifts of solasodine
Chemical shifts Chemical shifts

Carbon No. Carbon No.
(6 P P d (6 P P d
1 37.3 15 32.1
2 31.6 16 78.7
3 71.6 17 62.8
1 4 42.3 I 18 I 16.4
5 140.7 19 19.4
6 121.3 20 41.3
7 32.1 21 15.3
8 31.4 22 98.1
9 50.1 23 34.1
-~ ~~ ~~
10 36.7 24 30.3
11 20.9 25 31.4
12 39.9 26 47.7
13 40.5 27 19.3
14 56.5
2.6.7. Mass Spectrum

The MS (mass spectrum) of solasodine presented in Figure 9 and
Figure 10. were obtained by electron impact (EI) and chemical ionization
(CI) using methane as reagent on a Jeol JMS-DX-303 Mass Spectrometer.
The ionizing electron beam energy was at 70 eV. The main fragments of
solasodine are given in the table 8.
Table 8. The main fragments of solasodine
Species
El CI
M+ + 29 442 (8)
M+ + 1 414 (55)
M+ + 1 - 18(H,O) 396 (58)
M+ 413 (27)
W - 15(CH,) 398 (3)
M + - 28(C2H,) 385 (13)
C,H2,NO+ 271 (2)
a* 138 (70) 138 (10)
b’ 114 (100) 114 (27)
* see figure 1 1
The fragments of m/z 138 and 114 are characteristic for spirosolane ring of
solasodine (11, 12). The EI spectrum of solasodine is identical with
previously published spectrum (13).
MIIZ
Figure 9. EI-MSspectrum of solasodine (Sigma)

RELATIVE ABUNDANCE
SOLASODINE 509
4.
O ? 3
mlz 138 m/z 114
Figure 11. Characteristic fragmentation of solasodine.

3. ISOLATION OF SOLASODINE
In our laboratory, solasodine was isolated from fruits of SoZunum
wrightii with the following method, as reported by Indrayanto et al.(14).
Fresh ripe fruits (3 kg) were cut into small pieces and homogenized
in a blender with 4 times methanol-acetic acid 3 % until slurry was
formed. The slurry was refluxed for 30 minutes then filtered. Each 50 ml
of filtrate was treated with 5 ml HC1 conc. and hydrolyzed by refluxing
for 3 hours on boiling water bath. After neutralized with 10 %. NaOH, the
alkaloids was extracted with chloroform. The chloroform extract was
evaporated in vacuo to yield a dark brown semisolid mass (22 g),then
purified through a column of A1 0, with chloroform as the eluting
solvent. After evaporating the cgloroform, the residue was
chromatographed on silica gel 60 with benzene:acetone (15: 1) as eluting
solvent. Recrystallization with methanol gave 3.63 g pure solasodine
(TLC, UV,TR, MS).
4. BIOSYNTHESIS OF SOLASODINE
The biosynthetic pathway of solasodine followed the general
pathway of steroid biosynthesis, starting from acetylcoenzym A via the
in termediates mevalonic acid, squalene, cycloartenol and cholesterol. The
nitrogen atom was introduced through simple replacement of the terminal
hydroxy group by an amino group (15).
In solanidine the donor molecule was an amino acid arginine (16).
The main steps of the pathway (partially hypothetical) of cholesterol
to solasodine in are presented in figure 12 (15,17,18). The figure shows
that the formation of ring E can precede or after the formation of ring F.
5. CHEMICAL CONVERSION OF SOLASODINE TO STEROID

HORMONES
In order to obtain steroid hormone like materials, ring E and F of
solasodine attached to C16 and C 17 first must be removed. Solasodine can
be degraded into 16-dehydropregnenoloneacetate (16-DPA), which is an
excellent starting material for the preparation of most type of steroid
hormone (Marker degradation). In figure 13 the conversion of solasodine
to some steroid hormone is presented (19,20).
SOLASODINE 511
Acetyl-CoA Mevalonic acid Fmesyl pyrophosphate
HO #-@
Cycloartenol
Squalcne
Dormantino1
Dormanunonc
4
no 22, 26-Epimino-Ssholesten-3n,16R-diol
26-Am:no-160-hydroxycholer~rol
Verazine
I
1
HO
26-A iiiiiiodihydrodiosgenin
1
-
Etiulinr
\&."-OH n "^* ' .NO

&\ %
Solasodine
I40
Teinemine
Figure 12. The biosynthetic pathway of solasodine

512 GUNAWAN INDRAYANTO ET AL
Figure 13. Chemical conversion of solasodifie to steroid hormone.

SOLASODINE 513
6. BIOTRANSFORMATION OF SOLASODINE
Patel et al. (10) reported recently that the fungus Cunninghumella
ekguns could transform solasodine into solasod-5ene- 38, 7BIB-di01,
solasod-5ene-38,7adioland 38- hydroxysolasod-5-en-7ae. In contrast,
incubation of solasodine with fungus PenicilZiumpanrlum gave
solasod4ene-3a1e and the 6-methylsalicylicacid salt of solasodine.
By using Mycobucteriumphlei solasodine could be transformed to
4-androstene-3,17dioneand 1,4-androstadiene-3,17dione(21).
Shulz et al. (22) reported that in leaves extract of Solmum
luciniutum, solasodine converted to its glucoside by a solasodine
glucosyltranferase.
7. METHOD OF EXTRACTION AND QUANTITATIVE ANALYSIS

OF SOLASODINE
7.1. Method of extraction
For the extraction of solasodine as glycoalkaloid, the plant or cell
culture materials could be extracted with methanol ( 23), ethanol 95 %
(24,25), methano1:chloroform (2: 1) (26), 3 - 5 % aqueous acetic acid
(27,28), 5 % acetic acid in methanol (29), 2 % acetic acid in 90 % ethanol
(30) or 2 % oxalic acid (31). To determine the aglycone (solasodine) the
extracts must be hydrolyzed by 1N HC1, neutralized with NaOH or
ammonia, then extracted with organic solvent (chloroform).
To determine solasodine, the biomass also could be directly
hydrolyzed with 1N HCl (32,33) or 2N HCl in methanol (34). For
separating non polar components (eg. free sterols) , the biomass was
extracted with chloroform before hydrolized (35).
To minimize the decomposition of solasodine into solasodiene, Segal
et al. ( 36) recommended using of non aqueous low boiling point alcohols
and acid concentration not exceeding 1N. By using two phase system
consisting carbon tetrachloride and HCl as hydrolized mixture, Van
Gelder (17) could prevent the degradation of the steroid alkaloid. Oven
drying above 100°C of the plant materials also lead to solasodine loss (37).
7.2. Method of quantitative analysis

7.2.1 Titration
Method of Eldridge and Hockridge (26)
51.2 GUNAWAN INDRAYANTO ET AL.
To determine the glycoalkaloid content in Solanum prycamhum, the

dried berries was extracted with methanol :chloroform (2: l), then the
extract was mixed with 0.8 % Na,SO,, shaken in a separatory funnel
and ailowed to settle overnight. The methanol phase was collected,
dried, and dissolved in 2N H SO,. This solution was heated for 2
hours and made basic with 4fi NaOH. The aglycone was extracted 3
times with benzene. After evaporating the residue was dissolved in
methanol and titrated with a solution of 0.067 % bromophenol blue and
10 % phenol in absolute methanol, against a blank of methanol.
Method of Valovich (31)

After extraction of the glycoalkaloid with 2’% oxalic acid,
hydrolyzed with HQ, the hydrolysate was added a 50 % solution of
caustic soda to bring the pH to 9.5 for precipitating solasodine. The
solasodine was extracted with neutral chloroform, and the chloroform
solution was titrated with O.OO5N solution of n-toluene-sulphonic acid in
chloroform, with 0.1 % solution of dimethyl yellow as indicator.
7.2.2. Colorimetric
Method of Birner (25)
Finely powdered materials was refluxed with 95 % ethanol for 30
minutes, filtered, and the ethanol extract was collected. After evaporating, the
residue was hydrolyzed with 1N HCI,then neutralized with 1N NaOH. The
aglycone was comphed with methyl orange and the colored complex
extracted into chlorohrm and determined colonmetrically at 430 nm.
Method of Khdagi et al. (24)

In this method, solasodine was dissolved in 95 % ethanol, after
addition of phosphate buffer pH 7.5, a solution of 0.1 % bromothymol
blue in alcohol 50 %! was added to the mixture. The bromothymol
blue-solasodine complex was extracted using benzene, and measured on
spectrophotometer at 400 nm before 45 minutes.
Method of Nigra et al. (38).

Firstly the hydrolyzed sample was alkalinized with 60 % NaOH
then solasodine was extracted with chloroform. After addition of 2. 1i4M
of brornothymol blue in borax buffer pH 6.8 and mixed in a vortex-mixer,
SOLASODINE 515
the aqueous phase was taken out, added solution of methanolic 0.01N
NaOH to the chloroform phase, then measured at 610 nm.
7.2.3. Potentiometric
Method of Telek (27)
The steroid glycoalkaloids were extracted from freshly harvested
fruits with 2 % acetic acid and methanol. After hydrolysis, the common
aglycone solasodine was extracted in benzene. An aliquot was mixed with
an equal volume of acetone and titrated potentiometrically with 0.005 N
perchloric acid in dioxane, using glass and silver electrodes for
determination.
7.2.4. Gas chromatography

Methof of Carle (34)
In this method, the freeze dried biomass was hydrolyzed using 2N
HCl in methanol. After neutralization with ammonia 25 % , solasodine was
extracted by chloroform. An aliquot of the chloroform phase was injected
to a GC with the following condition, column glass (1/8 inch x 6 ft)
packed with 3 % SE-30 (Gas Chrom Q, 100-120 mesh); FID and injector
temperature were 30OoC; column temperature was isothermal at 250OC.
By this condition solasodine and diosgenin/tigogenin were separated and
quantitatively determined from Solmum plant materials.
Method of Indrayanto (39)

By using a glass column ( 2m x 2mm id.) packed with 3 % OV-1 on
Gas Chrom Q ( 100-120 mesh); FID and injector temperature 30OoC;
oven temperature was programmed from 200 to 280°C, S°C/minute,
solasodine, hecogenin, diosgenin and sterols (cholesterol, campesterol,
stigmasterol, sitosterol and isofucosterol) could be good separated.
Method of Van Gelder (17)

To prevent steroid alkaloid, degradation, van Gelder was used two
phase system of carbon tetrachloride and HC1 to hydrolyze the plant
materials. For analyzing the steroid alkaloids two systems of GC were
used by the author.
System 1 : glass column (lm x 2mm id.) packed with 10 % SE-30
on Chromosorb W HP (80-100 mesh); maximum operating temperature
were 325OC (isothermal), 35OoC (programmed); FID temperature 350OC;
injector temperature 325OC.

System 2 : fused capillary column of 50m x 0.22 mm id., coated
with CP-Sil 5 (film thickness 0.12 pm or CP-Sil 19 (film thickness 0.19
pm); oven temperature were 30O0C (isothermal), 325OC (programmed);
R D and injector temperature were as the same as system 1. With both
systems, solasodine, solanidine, demissidine, tomatidine, solasodiene and
solanthrene could be separated.
By using two detectors (FID and NPD), the identity of 21 steroid
alkaloids can be elucidated by retention indexes and NPD/FID response
ratio.
7.2.5. High Performance Liquid Chromatography

Some of HPLC methods in the determination of solasodine or its
glycosides that have been published are listed in table 9. By using
combination of HPLC and thermospray mass spectrometry (LC-MS), the
structure of the steroid alkaloid glycosides could be elucidated. This
method seems well suited for a screening of a complex mixtures for a
certain steroidal alkaloids (40).
7.2.6. Radioimm unoassay

Weiler et al. (46) have reported a rapid, spesific and sensitive RIA
for the determination of solasodine in vegetative and generative parts of
Soiunwn luciniatwn. The antiserum added did not react at all with sterols,
solanidanes and spirostanols, but showed strong reactivity with tomatidine
and solasodine glycoalkaloids. The assay allowed the detection as little as
0.7 ng solasodine glycosides.
7.2.7. Thin Layer Chromatography - Densitometric

Some of the published methods of Thin layer chromatography (TLC)
of solasodine are listed in table 10.
In our laboratory the solasodine content of in the vitro cultures of
S o l a w spp. was determined by densitometric method. Firstly the dried
biomass was extracted with chloroform 3 times on a vortex mixer to
remove the non polar components (eg. free sterols), then hydrolised with
2N HCl in methanol (2 hours, 75'C), After neutralised with 10N NaOH,
solasodine was taken out by chloroform. The chloroform extract was
spotted on precoated Kieselgel G 60 F254 (Merck), and chloroform :
methanol : diethylamine (20:2:0.5) was used as eluent. The solasodine
M a sample
w 240, -,17 Qne, Kasslbrataetrl. (21)
205nm 1,4 adrostadecle3,17dare,~@EO&IO
W205nm sdssodne,-ne Vogd ad.(23)
sifidcacidradal I acebnibae:wster(77.5:22.5);pH4.0 WMom

w
acet#libile :Pic eS :TEA (83:17 :0.1); pH 2.8 W205Nll
:0.1 MThkmer(9: 1) w 210 nm
IWWI6 Hurler(41)
W 213 nm Hurleretal.(42)
WMSn
pBondapakC18 I metranol:O.OlMTriskner(75:25) a-,B-,wlvcosidea daaodm
Table 9. HPLC methods for determining solasodine and its glycosides

I nHaxas:EW (1:l)
nHaxas:EtOH (1:l)
C W : M H (23:2)
clt2az:MsoH (9:l)
ctt2c12:AceQne (4: 1)
nHaxas :h b l e (1 : 1)
CHzClz :&OH :AmUc acid(= : 13: 2)
PqNoj Me0H:CHQ (1:9)
Table 10. TLC methods for Analysing of solasodine

SOLASODINE 519
spots was detected by anise aldehyde-H2S04reagent (lOO°C, 5-10

minutes). Quantitation was done by measurmg at the maximum absorbance
reflectance (385 nm). The determination of solasodine was made by
calculation with a calibration graph obtained using solasodine (Sigma) as
external standard on the same TLC plate. With this method linearity of
solasodine was achieved from 0.4 to 1.6 pg/spot (r=0.998, n=7,
V =2.1 %); LOD (limit of detection)= 0.11 pg/sjmt; LOQ (limit of
q1,?&titation)=0.31 pg/spot; 'mean accuracy @ f SD) with standar
addition method was 98.92 f 2.35 %, n = 14; RSD of precision
determination was 1.87 % (n=10).
8. PHARMACOLOGICAL AND BIOLOGICAL ACTMTIES OF

SOLASODINE AND ITS GLYCOSIDES
As early as 1965 Kupchan et al. (53) demonstrated a tumor-
inhibiting activity of the glycoside 13 solamargine. Cham et al. (54)
showed t!!at glycoalkaloid solasonine, solamargine and another mixture of
glycosides containing aglycone solasodine, had on antineoplastic activity
against Sarcoma 180 in mice. Solasodine inhibited the growth some
fungal strains, although its activity was less compared to solafloridine and
verazine (55). Some steroidal alkaloids including solasodine showed
inhibitory effect on the enzymatic conversion of dihydrolanosterol into
cholesterol (56).
The steroidal glycoalkaloid solasonine and solamargine showed a
membrane-disrupting properties of phosphatydylcholinekholesterol
liposomes at concentration > 50 pM (57).
A slight acetylcholinesteraseinhibitory activity of solasonine and
solamargine was also reported (58).
Recently Frohne (59) reported that solasodine had a cortisone like
effect such as anti phlogistic or reduction of blood vessel permeability.
Solasodine also can prevent of anaphylactic shock in guinea pig. In a
clinical trial a dose of 1 mg solasodine citrate which was given 2 times a
day showed cardiotonic effect. This clinical trial also showed that
solasodine gave a desensitization effect especially in patient with a
rheumatoid polyarthritis.
References
1. Telek,L., Delphin,H., Cabanilias,E., (1977) Economic Botany 3 1:
120-128.
2. Nigra,H.M., Alvarez,M.A., Giulietti,A.M., (1990) Plant Cell Tissue
and Organ Cultures 21: 55-60.
3. Sudiarto, Chairani,F., Rosita, S.M., Wahid,P. (1985) Jurnal Litbang
Pertanian 4: 71-76.
4. Macek, T.E., (1989) Solanum aviculare Forst., Solanum laciniatum
Ait. (poroporo): In vitro cultures and the production of solasodine, in
:Bajaj,Y.P.S. (ed.)Biotechnology in Agriculture and Forestry Vol. 7,
Springer Verlag, Berlin Heidelberg, pp.444-467.
5. The Merck Index, 11thEd.(1989),Merck & CO., Inc, p. 520.
6. Bentley, K.W. and Kirby, G.W. (1972) Elucidation of Organic
structure by physical and chemical methods, 2nd ed., Interscience, p.
565.
7. Europaiesches Arzneibuch,Band I, I1 (1976), Wissenschaftliche
Verlagsgesselschaft MBH Stuttgart Goviverlag GMBH, Frankfurt.
8. Puri, R., Wong, T.C., Puri, R.K. (1993) Magnetic Resonance in
Chemistry 31:278-282.
9. Radeglia, R., Adam, G . , Ripperger, H. (1977)'Tetrahedron Letter
11:903-906.
10. Patel, A.V., Blunden, G., Crabb, T.,4. (1994) Phytochemistry 35:
125-133.
11. Budzikiewicz, H., Wilson, J.M.. Djerassi, C. (1962) Monatshefte fiir
Chemie 93: 1033-1041.
12. Budzikiewicz, H., Djerassi, C., William, D.H., (1964) Structure
elucidation of natural products by Mass Spectrometri, Vol 11, Holden
Day Inc., San Fransisco, hndon, Amsterdam, pp. 110-120.
13. Budzikiewicz, H. (1964) Tetrahedron 20: 2276.
14. Indrayanto, G., Tutuk Budiarti, Soebahagiono, Emma, Sutarjadi
(1979) The solasodine content of Solanm grandiflonun, paper
presented in UNESCO Regional Seminar on Medicinal Plants
Bangkok, Thailand.
15. Heftmann, E. (1983) Phytochemistry 22: 1843-1860.
16. Kaneko, K., Tanaka, M.W., Mitsuhashi (1976) Phytochemistry 15:
1391.
17. Van Gelder, W.M.J.(1989), PhD thesis, Agriculture University
Wageningen
SOLASODINE 52 1
18. Tschesche, R. and Brennecke, H.R. (1980) Phytochemistry 19:

1449-1451.
19. Wall, M.E. (1986) Status of raw materials for production of oral
contraseptives in Indonesia, Paper presented in the Workshop of
Biotechnology of Steroid Compounds as Contraceptives and drugs,
Jakarta, Indonesia.
20. Sebek ,O.K. ,(1986) Production of steroid compound by fermentation,
paper presented in the Workshop of Biotechnology of Steroid
Compounds, Jakarta, Indonesia.
21. Kartasubrata,J., Loyniwati, Jarnilah, Karossi, A.T. (1993) Indonesian
Journal of Applied Chemistry 3: 61-68.
22. Schulz,D., Eilert,U., Ehmke, A. (1993) Planta Media Supplement
59: A649.
23. Vogel,H., Jatisatienr, A., Bauer, R. (1990) Angew. Botanik 64:
393-400.
24. Khafagi, S.M., Amin, S.W., Hassanin, R. (1970) Planta Medica 21:
139-141.
25. Birner, J. (1969) Journal of Pharmaceutical Sciences 58: 259.
26. Eldridge, A.C., and Hockridge, M.E. (1983) Journal Agric. Food
Chemistry 31: 1218-1220.
27. Telek, L. (1977) Journal of Pharmaceutical Sciences 66: 699-702
28. Cham, B.E. and Wilson, L. (1987) Planta Medica 59-61.
29. Hunter, I.R., Heftmann, E. (1983) Journal of Liquid Chromatography
6 ~281-289.
30. Chowdhury, A.R., Prasas, R.N., Uddin. A. (1979) Quart. Journal of
Crude Drug Research 17 : 137-138.
31. Valovich, N.A. (1965) Meditzinskaya Promyshlennost USSR 19:
4548.
32. Ehmke, A. and Eilert, U. (1993) Sulunwn dulcumuru L.:
Accumulation of steroidal alkaloids in the Plants and in different in
vitro cultures, in : Bajaj, Y.P.S. (ed.) Biotechnology and Forestry,
Vol. 21, Medicinal and Aromatic Plants VI, Springer Verlag, Berlin
Heidelberg, pp.339-349.
33. Chandler, S., Dodds, J. (1983) Plant Cell Reports 2: 69-72.
34. Carle, R. (1979) Ph D thesis, University of Tuebingen.
35. Indrayanto, G., Studiawan, H., Noor Cholies (1994) Phytochemical
Analysis 5 : 24-26.
36. Segal, R., Breuer, A., Milo-Goldzweig, I. (1978) Journal of
Pharmaceutical Sciences 67: 1169-1170.
37. Crabe, P.G., Fryer, C., (1982) Journal of Pharmaceutical Sciences

71: 1356-1362.
38. Nigra, H.M., Caso, O.H., Giulietti, A.M., (1987) Plant Cell Reports
6: 135-137.
39. Indrayanto, G., (1983) Ph D thesis, University of Tuebingen.
40. Ehmke. A., Schiebel, H.M., McDowell, M. (1987) Pharmaceutisch
Weekblad Scientific Edition 9: 232.
41. Heftmann, E., Hunter, I. R. (1979) Journal of Chromatography 165:
283-299.
42. Hunter, I.R., Walden, M.K., Heftmann, E. (1980) Journal of
Chromatography 198: 363-366.
43. Crabbe, P.G. and Fryer, C. (1980) Journal of Chromatography 187:
87- 100.
44. Indrayanto, G, Utami, W., Santosa, M.H., Syahrani, A., Diosgenin,
in: Brittain, H.G.(4) .
Analytical Profile of Drug Substances and
Excipients Vol. 23., Academic Press, in Press.
45. Ehmke, A. and Eilert, U. (1986) Plant Cell Reports 5: 31-34.
46. Weiler. E.W., Kriiger, H., Zenk, M.H. (1980) Planta Medica 39:
112-124.
47. Hunter, I.R., Walden, M.K., Wagner, J.R., Heftmann, E. (1976)
Journal of Chromatography 118 :259-262.
48. Puri, R.K. and Bhatnagar, J.K. (1975) Phytochemistry 14: 2096.
49. Willuhn, G. (1966) Planta Medica 14: 408-419.
50. Willuhn, G. and Kun-anake, A. (1970) Planta Medica 18:354-359
51. Kadkade, P.G. and Rolz, C. (1979, Phytochemistry 16:1128
52. Kartasubrata, J., Fitri, T.Y., Halomoan, V.A., Lotulung, P.,
Buchori, Karossi, A.T., (1993), Annales Bogorienses 2: 12-15
53. Kupchan, S.M., Barboitis, S.J., Know, J.R., Lau Cam, C.A. (1965)
Science 150: 1827-1828.
54. Cham, B.E., Giliver. M., Wilson, L. (1987) Planta Medica 34-36.
55. Kusano, G., Takahashi. A . , Sugiyama, K., Nozoe, S . (1987)
Chemical Pharmaceutical Bulletin 35: 4862-4867.
56. Kusano, G., Takahashi, A., Nozoe, S., Sonoda, Y.,Sato, Y. (1987)
Cheinical Pharmaceutical Bulletin 35: 4321-4323.
57. Roddick, J.G., Rijnenberg. A.L., Weisenberg, M. (1990)
Phytochemistry 29: 1513-1518.
58. Roddick, J.G. (1989) Phytocheinistry 28: 2631-2634.
59. Frohne, D. (1993) Zeitschrift fur Phytotherapia 14: 337-342.
STARCH
Ann W. Newman, Ronald L. Mueller, Imre M. Vitez, Chris C.
Kiesnowski, David E. Bugay, W. Paul Findlay, Chris Rodriguez
Bristol-Myers Squibb Pharmaceutical Research Institute
One Squibb Drive
New Brunswick, NJ 08903
525 ANN W. NEWMAN ET AL.
CONTENTS
I. Description
1.2 Types of Starch
1.3 Appearance
1.4 General Chemical Properties
3. Starch Derivatives
4.1 Structural Information
4.2 Infrared Spectroscopy
4.3 Raman Spectroscopy
4.4 Nuclear Magnetic Resonance Spectroscopy
4.5 Thermal Properties
4.6 Moisture Content
4.8 Optical Microscopy
4.9 Particle Size Measurements
4.10 Surface Area Measurements
4.1 1 Bulk Powder Properties
4.12 Hygroscopicity
5 . Methods of Analysis
5.2 Chromatography
6. Excipient Studies
6.1 Drng Compatibility
6.2 Tableting
6.3 Disintegration
7. References
STARCH 525
I. Description
1.1 Name. Formula. and Molecular Weight
Starch is a polymeric material with a molecular formula of (C6H1005)n,

where n ranges from 300 to 1000 [l]. Common starches contain two types
of D-glucopyranose polymers called amylose and amylopectin. Amylose
is a linear polymer of a-D-glucopyranosyl units linked (1-4) as shown in
Figure la. These molecules can be comprised of 100 to over 1000
glucose units [2]. Amylopectin is a branched polymer of a-D-
glucopyranosyl units containing (1-4) linear linkages and (1‘ 6 ) linkages
at the branch points, as shown in Figure 1b. This polymer is three or more
times larger than amylose [2]. Most naturally occurring starches contain
approximately 30% amylose, however, specific starches and their
properties are determined by the size and amount of each type of polymer
molecule present in the material.
The starch granules are formed by the attractive forces between the
polymeric molecules. The linear portions tend to associate into micelles
which bind the molecules together to form a somewhat ordered structure.
Models of this structure have been proposed [3], and it is known that the
structure is rigid and insoluble in water.
The Chemical Abstracts identification number is CAS-9005-25-8.
1.2 T p e s of Starch
Starch can be derived from a number of natural sources, including those

listed in Table I. It is found in various parts of the plants and several
extraction methods are used to isolate the material. The most common
type of starch used in the pharmaceutical industry is corn, although studies
with other forms have been performed [4-91.
536 ANN W. NEWMAN ETAL.
Figure 1. Structure of (a) a linear amylose starch molecule, and

(b) a branched amylopectin starch molecule.
0- 0-
OH
OH OH
n
r -
STARCH 521
Table I. Sources and Characteristics of Various Starches [ 101
Starch Type Extracted From Granule Shape Granule Size

(Pm)
Corn (Maize) Seed Round or 5 -25

polygonal
Tapioca Root Round or oval 2 - 25
Potato Root Egg-shaped 15 - 100
Wheat Seed Round or 2 - 1 0 or

elliptical 20 - 35
Sago Stem Oval or 20 - 60

egg-shaped
Arrowroot Root Oval 15 - 70
Rice Seed Polygonal 3-8
Barley Seed Round or 2 - 6 or

elliptical 20 - 35
Waxy sorghum Seed Round or 6-30

polygonal
Sweet potato Root Polygonal 10 - 25
Waxy maize Seed Round or 5 - 25

polygonal
A number of starch modifications also exist and are used in

pharmaceutical applications [ 11. Pregelatinized or compressible starch has
been chemically or mechanically processed to rupture all or part of the
granules in water. It is then dried to yield an excipient material suitable
for direct-compression formulations. Sterilizable maize starch contains
magnesium oxide (not greater than 2.2%) and has been chemically or
physically treated to prevent gelatinization upon exposure to moisture or
steam sterilization. Soluble starch results when potato or maize starch has
been chemically treated to destroy the gelatinizing ability of starch.
1.3 ApDearance
Starch is a fine white powder which is odorless and tasteless. It is

composed of very small spherical or elliptical granules. The botanical
origin of the starch material will determine the granule shape and size, and
these characteristics are summarized in Table I. When the granules are
analyzed microscopically, a distinct cleft called the hilum is observed,
which is considered the origin of granule growth.
1.4 General Chemical Properties
Starch is insoluble in alcohol, most solvents, and cold water. Alkaline

solutions, however, will degrade starch and its polysaccharide components
[lo]. Starch is relatively resistant to carbohydrases other than a-amylose
[ I 13.
When starch is suspended in water and heated to a critical point called the
gelatinization temperature, water will penetrate the granules and swell
them to produce a viscous mass. With the rising temperature, the
hydrogen bonds that hold the micellar structural units and the water
molecules in an aggregated state tend to dissociate. The dissociated water
molecules are then able to penetrate the weakened starch structure and
gradually hydrate the many hydroxyl groups along the length of the starch
molecule. Gelatinization temperatures vary from starch to starch, but
range from 60 to75"C [lo]. Starch granules will lose their characteristic
shape as gelatinization proceeds.
STARCH 529
The reaction of starch with iodine is a common identity test for starch. A
dilute solution of iodine stains starch a blue to bluish red color. It is
believed that the amylose portion complexes with iodine by forming a
helix around it [lo]. This blue color has been used both as a qualitative
and quantitative test for starch in various systems.
1.5 Uses and ApD1ication.s
Starch is used widely in the pharmaceutical industry because, among its

other properties, it is readily available, inexpensive, white, and inert. It
has been described as a tabledcapsule diluent, tablet disintegrant, and
glidant [l].
The function of starch can depend on how it is incorporated into the

formulation. Starch will function as a disintegrant when it is added in the
dry state prior to adding a lubricant. It may exhibit both binding and
disintegrant properties when it is incorporated either as a paste or dry
before granulation with other agents [ 121.
The corn kernel is made up of the tip cap, the pericarp or hull, the germ or
the embryo, and the endosperm. The endosperm contains the starch
granules embedded in protein cell walls. Corn kernels have two distinct
regions of endosperm. The floury region in the grain center has loosely
packed, rounded starch granules with a low protein content. The horny
region of the endosperm at the grain edges contains angular granules with
a high protein content. The granules are removed from the kernel during
starch preparation.
The most common preparation of starch is wet milling although dry

milling is also performed [1I]. The corn is screened to remove any cob,
sand or other unwanted material and aspiration is used to remove the light
dust and chaff. In a process called steeping, the material is placed in a vat
and the kernels are softened for milling using water, heat, and sulfur
dioxide. The steeped corn is coarsely ground in an attrition mill to break

loose the germ which is then removed in a cyclone separator based on its
density. The resulting aqueous slurry is milled a second time to release
the starch granules. The kernel suspension containing starch, gluten, and
fibers is passed through a concave screen to remove the fibers. The starch-
gluten suspension is then concentrated by centrifugation to reduce soluble
material and to separate the gluten based on its density. The starch is
concentrated again and washed numerous times using the cyclone
separator. The starch suspension may be dried (unmodified corn starch),
gelatinized and dried, or modified by chemical or physical means. After
this processing. corn starch is a white powder with a pale yellow tint and
bledching is required to achieve absolute whiteness.
3. Starch Derivatives
Starches undergo many reactions characteristic of alcohols because of the

many hydroxyl groups present in the structure. Modification of the D-
glucopyranosol units can occur by oxidation, esterification, etherification,
or hydrolysis. The resulting starch derivatives are defined by a number of
factors such as plant source, prior treatment (acid-catalyzed hydrolysis or
dextrinization), aniyloseiamylopectin ratio or content, molecular weight
distribution or degree of polymerization, type of derivative (ester, ether,
oxidized). nature of the substituent group. and physical form (granular,
pregelatinized) [ 131.
'I he degree of substitution (DS) is a common method of characterizing

starch derivatives and is a measure of the average number of hydroxyl
groups on each D-glucopyranosyl unit. It is expressed as the moles of
substitucnt per D-glucopyranosyl unit and the maximum DS is 3 since
thrct! hydroxyl groups are available in the unit for substitution. Most
conimercially produced derivatives have a DS less than 0.2. The molar
substitution is used when the substituent group reacts further with the
reagent to form a polymeric substituent. It is defined as the level of
substitution in terms of mole of monomeric units (in the polymeric
substituent) per mole of D-glucopyranosyl unit and can be larger than 3.
STARCH 53 1
Manufacture of starch derivatives includes treatment of aqueous slurries

and the dry starch. For slurries, a 35-45% aqueous suspension at pH 7 to
12 and temperatures ranging up to 60°C are commonly used. Conditions
are adjusted to prevent gelatinization and to allow recovery of the starch
derivative in granular form by filtration or centrifugation. For dry
material, the starch is treated with the required reagents by dry blending,
spraying the reagents onto the starch granules or filter cake, or by
suspending the starch in a reagent solution and then recovering the starch.
The treated powders are then heated to temperatures up to 150°C to yield
granular products with DS values up to one.
Many variations on starch derivatives exist and they have been exploited
in a number of ways in the food, paper, textile, and adhesive industries.
More detailed discussions of starch derivatives are available [11,13,14].
Some common starch derivatives are listed below.
Hydroxyalkyl Starch Ethers such as hydroxyethyl and

hydroxypropy1
Starch Phosphates such as starch monophosphates and starch
phosphate diesters
Cationic Starches such as the tertiary and quaternary
aminoalkyl ethers
Oxidized Starches made by introducing carbonyl and carboxyl
groups
Starch Acetates
Cross-Linked Starches
Acid-Modijied Starches
The physical properties of unmodified and pregelatinized corn starch from

three vendors were characterized and the materials analyzed are
summarized in Table 11. Examples of unmodified corn starch are STA-Rx
and Purity 2 1. Starch 1500 is a sample of partially pregelatinized starch
and Starch 1551 represents fully pregelatinized starch. Starch 1500 LM is
a partially pregelatinized starch with a low moisture content.
Table 11. Starch Materials Analyzed for Physical Properties
Total Volatile
Starch Type Vendor
Trade Name Content (%)
tinmodified Staley STA-Rx 10.9

corn
Unmodified National Purity 21 10.2

Corn
Pregelatinized National 1551 8.3
Pregelatinized Colorcon 1500 9.4
Pregelatinized- Colorcon 1500 LM 6.5

Low Moisture
STARCH 533
4.1 Structural Information
Starch is a semicrystalline polymer. The linear amylose molecules are

amorphous in nature, but the branched amylopectin portion has been
reported as partially crystalline. It is believed that the crystalline regions
in the starch granule are interspersed in a continuous amorphous phase
[3,13,15].
X-ray diffraction studies have shown that starch exists in three crystal
forms designated A, B, and C. These forms are dependent on the botanical
source of the starch. Pattern A is observed for cereal grain starches,
whereas pattern B is characteristic of tuber, fruit, and stem starches.
Pattern C is intermediate between the A and B patterns and has been
attributed to mixtures of A and B type crystallites [151. The A type
pattern is commonly observed for corn starch.
Single crystal x-ray diffraction data for the crystalline portion of A type
starch has been reported [16], and it was found to crystallize in a
monoclinic lattice with a = 2.124 nm, b= 1.172 nm, c=l.069 nm and
y=123.5 '. The unit cell contains 12 glucose residues located in two left-
handed, parallel-stranded double helices packed in a parallel fashion. Four
water molecules are located between these helices. Intramolecular
hydrogen bonding was not observed and the interstrand stabilization in the
double helix is attributed to O(2)...0(6)
type hydrogen bonds.
It has been reported that the B type starch also contains chains arranged in
double helices [ 171. The currently accepted hexagonal unit cell has
dimensions of a=b=l.85 nm and c= 1.04 nm. The A and B structures
differ in crystal packing of the chains and in moisture content.
Powder x-ray diffraction patterns for representative unmodified and

pregelatinized corn starches are given in Figure 2. The unmodified corn
starch was found to have some crystalline character as evidenced by the
broad peaks present. The pattern is indicative of a semicrystalline material
and is similar to the A type pattern, but a definite determination of the
form is difficult based on the quality of the pattern. The pregelatinized
Figure 2. X-ray powder diffraction patterns of unmodified (upper trace)

and pregelatinized (lower trace) corn starch
m
m
Ln
N
Q
N
v)
r(
0
rl
Ln
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Q
m - m
m ~
Q + Q D
m
9
m
m
f
~
Q
N
m
Q
m
m
m
QD
Q
~
9
m
r
1p
Q
n
N
m
Q
-
s m o
N f r l rl rl rl 4
STARCH 535
Starch 1551 material exhibits a broad amorphous halo in the powder

pattern indicating that the majority of the sample is amorphous material.
This type of pattern was expected since the processing of pregelatinized
starch destroys the crystalline portions of the granule.
4.2 Infrared Spectroscopy
Diffuse reflectance (DR) infrared (IR) spectra were acquired for

unmodified, partially pregelatinized, fully pregelatinized, and partially
pregelatinized (low moisture) corn starch. Identical IR spectra were
measured for the unmodified, partially pregelatinized, and fully
pregelatinized starch samples and are represented in Figure 3. The
partially pregelatinized (low moisture) starch sample displayed slight
shifts in the absorbance bands assigned to the glycosidic COC and coupled
CO stretch vibrational modes. The IR spectral band assignments are
presented in Table I11 [IS]. The measured absorbance bands are consistent
with the structure of starch.
4.3 Raman Spectroscopy
Raman spectra were acquired for unmodified, partially pregelatinized,

fully pregelatinized, and partially pregelatinized (low moisture) corn
starch. Identical Raman spectra were measured for the unmodified,
partially pregelatinized, and fully pregelatinized starch samples and are
represented in Figure 4. The Raman spectral band assignments are
presented in Table IV [191. The measured absorbance bands are consistent
with the structure of starch.
4.4 Nuclear Magnetic Resonance Spectroscopv
Solid-state 13Ccross polarizatiodmagic angle spinning (CPMAS) nuclear

magnetic resonance (NMR) spectra were acquired for unmodified,
partially pregelatinized, fully pregelatinized, and partially pregelatinized
(low moisture) corn starch. Identical I3C spectra were measured for each
and are represented in Figure 5.
5% ANN W. NEWMAN ET AL.
Figure 3 Infrared spectra of ( a ) unmodified corn starch, STA-Rx, (b)

partially pregelatinized corn starch, Starch 1500, (c) fully
pregelatinized corn starch, Starch 1551, and (d) partially
pregelatinized corn starch with low moisture content, Starch
1500 LM
T
i
STARCH 537
Table 111. Diffuse Reflectance IR Spectral Assignments
Wavenumber fern-') Structural Assignment

-3350 OH stretch
2929,2898 CH stretch
1452 CH, bend
1417,1335, 1302 CH bend
1364,1240,1206 OH in-plane bend
1148 glycosidic COC asymmetric
(1 166 for LM starch) stretch
1076 (1065 for LM starch), 1012 coupled CO stretch, CC
stretch, and OH bend
930 ring vibration
862 C,-group vibration
764 ring breathing vibration

709,643,606,576,524 low frequency ring
vibrations
Figure 4. Raman spectra of (a) unmodified corn starch, STA-Rx, (b)

partially pregelatinized corn starch, Starch 1500, (c) fully
1500 LM.
STARCH 539
Table IV. Raman Spectral Assignments
Raman shift (cm-') Structural Assignment

-3365 OH stretch
2932,2907 CH stretch
1459 CH,, CH, deformation
1380 CH, symmetric
deformation
1335 CH deformation (?)
1123,1081,1051 CC stretch
938,855,577,479,440 ring and lattice vibrations
Figure 5 . Solid-state 13CCP/MAS nuclear magnetic resonance spectra

of (a) unmodified corn starch, STA-Rx, (b) partially
pregelatinized corn starch, Starch 1500, (c) fully
1500 LM.
c c
STARCH 54 1
Table V. Solid-state I3CNMR Spectral Assignment
NMR Chemical Shift (ppm) Structural Assignment"

61.9 C6
72.2 c4
81.5 C2, C3, and C5
101.2 c1
~~
"Carbon numbering scheme based upon Figure 6.
13
Figure 6. Numbering system of the starch sub-unit used for the C-
NMR assignments.
OH
542 ANN W. NEWMAN ET AL
The resonance positions and structural assignments for the solid-state 13C
NMR spectrum of starch are presented in Table V. Although the CP/MAS
technique provides a “liquid-like” NMR spectrum, the broad nature of the
carbon resonances prevented spectral resolution of all the carbon signals at
a Larmor frequency of 62.89 MHZ. Solution-phase, ‘H NMR studies in
DMSO-d, have also been performed on amylose and model compounds
1201. Spectral assignments and intramolecular hydrogen bonding suggest
that the same conformation is perpetuated along the amylose chain.
4.5 Thermal Properties
Thermal analysis has also been used to characterize the structure of starch.
A melting endotherm due to the crystalline portions of starch has been
studied [21], but it is not clearly resolved in all samples due to the small
amount of crystalline material present in the samples. The transition is
also dependent on the sample preparation and moisture content of the
material. The melting point of starch has been calculated using an equation
developed by Florey [22] for polymeric systems. Using this approach, the
melting point of the pure polymer has been reported as 168“C [2 11. Using
the same equation and hot stage data, a melting point of 210°C has also
been reported [23,24].
The glass transition and gelatinization temperatures have also been studied
for starch materials [25,26]. Studies of the glass transition in wheat starch
have been performed using differential scanning calorimetry (DSC) [25].
This thermal event was found to be dependent on the moisture content of
the preparation and was ill-defined below 13% moisture. When observed,
it occurred at approximately 40°C. The onset of melting was also
dependent on the moisture content of the samples in this study and was
found to range from approximately 60 to 80°C. Gelatinization studies of
starch have also used DSC [26]. Concentrated starcWwater suspensions
produced a well-defined endotherm under suitable conditions which could
then be integrated to obtain the heat of gelatinization for various starches.
STARCH 543
A representative DSC curve collected for corn starch with a low moisture
content (9%) is given in Figure 7. The sample was analyzed as received in
a hermetically sealed pan with a pinhole to allow for controlled pressure
release. A broad endotherm is observed for the sample over a temperature
range of 50-150°C. The thermogravimetric (TG) curve in the overlay
shows a weight loss in the same temperature range as the endothermic
transition, therefore, it was designated as a dehydration. A similar curve
was also observed for the pregelatinized starch samples. The glass
transition for the sample was not expected to be clear due to the low
moisture content of the samples. The dehydration predominated with the
sample preparation used and the melt endotherm was not observed for the
corn starch sample.
To obtain information on the glass transition in pregelatinized starch, a

sample was equilibrated in a 90% relative humidity chamber for 19 days.
The volatile content of the sample was 17% after equilibration. The
sample was then placed in a hermetically sealed pan and analyzed. The
glass transition was evident at approximately 53 "C, as shown in Figure 8.
It is evident that the sample preparation is a critical factor in obtaining
specific information using DSC.
4.6 Moisture Content
The moisture content of starch has been determined using a number of

techniques including Karl Fisher measurements [27], thermogravimetry
(TG) [28], loss on drying (LOD) [l], and NMR spectroscopy [28]. The
National Formulary (NF) and British Pharmacopeia (BP) specify that
moisture contents should be less than 14, 15, or 20%, depending on the
botanical source of the material.
The moisture contents of the commercial corn starch lots were measured
using TG analysis and are given in Table 11. The unmodified corn starch
samples had moisture levels around 1O%, which is similar to that reported
for similar types of samples [13. The amount of water in the
pregelatinized samples was slightly lower than that of the unmodified corn
starch samples, ranging fiom approximately 8 to 9%. The low moisture
544 ANN W. NEWMAN ETAL
Figure 7. Differential scanning calorimetry thermogram (lower trace)

and thermogravimetry profile (upper trace) of unmodified
corn starch.
In 9
y: 9
0 4 r
a
Y
n
U
3
Y
i?
r
0
P
10G
2;
-
3
E
n
O E
:+
. 10
I
10
0
I 10
N
0
STARCH 545
Figure 8. Differential scanning calorimetry thermogram of

pregelatinized starch showing the glass transition.
0
m
r
m
t
N
v)
/ * m
0
I I I 0
?J
9
0 9
‘y
5Jh ANN W. NEWMAN ET AL.
pregelatinized starch did exhibit a significantly lower moisture content of

6.5%.
4.7 Particle MorpholoFy
A number of excellent reviews on the microscopy of starches have been

published [ 10, 29-32]. Areas covered in these references include granule
morphology, granule size, gelatinization temperatures, and staining. Our
discussion will focus on granule morphology using scanning electron
microscopy.
As summarized in Table I for a number of starches, the granule shape and

size is characteristic of the botanical origin and can be used to identify the
materials. It has been reported that the floury granules, as found for potato
and tapioca starches, tend to be larger and more regular in shape.
Descriptive terms used for these types of granules include round, elliptical,
or oval. Horny starches, such corn and rice, are usually described as
polygonal because of the angular sides of the granules caused by the close
packing of the granules in the kernel. Starches are found as individual
granules, but aggregated materials are also observed and are attributed to
the drying conditions. Extensive heat and moisture during drying will
produce a slight gelatinization of the surface of the granule and cause the
granules to adhere together to form the aggregates.
Electron microscopy has been useful in the morphological study of starch

grunules 133-351. Scanning electron microscopy (SEM) exhibits good
depth of field and gives detailed information about the surface of dry
granules. The low magnification SEM micrograph in Figure 9a shows the
uniform particle size and shape for STA-Ku. The polygonal and round
shape of' the unmodified corn starch is illustrated in Figure 9b at a higher
magnification. The surfaces of the granules are relatively smooth and
pores are not evident at this magnification. An aggregate showing some
fusion of the granulc surfaces of unmodified corn starch is shown in
Figure 1 0. The pregelatinized starch sample exhibits an entirely different
morphology, as shown in Figure 1 1. The particles are irregular and show
large pores for the majority of particles. The pregelatinization process has
STARCH 541
Figure 9a. Scanning electron micrograph of unmodified corn starch

granules, taken at a magnification of 270x.
Figure 9b. Scanning electron micrograph of unmodified corn starch

granules, taken at a magnification of 2700x.
STARCH 549
Figure 10. Scanning electron micrograph of an aggregate found in an

unmoMied corn starch sample, taken at a magnification of
1500~.
SS(1 ANN W. NEWMAN ET AL
Figure 1 1. Scanning electron micrograph of pregelatinized Starch 155 1,

taken at a magnification of I OOx.
STARCH 55 1
ruptured the granules resulting in a new particle morphology.
SEM has also been used to investigate the properties of starch in a tablet
[36]. It was found that starch possessed elastic properties during
compression, with maize and rice starch being more elastic than potato.
As the applied pressure was increased, deformation of the starch granules
was reported and could predominate at very high pressures. It was also
concluded that starch particles do not fuse together and were nearly always
surrounded by a space which contributed to the disintegration properties.
4.8Optical Microscopv
Optical microscopy has been a powerful tool in the study of starch

materials and common starches have been readily identified using this
technique. It has been suggested that starch be examined as a 0.2-0.3%
suspension in water or glycerol to obtain the best images [30,3 11. A
polarizing microscope also gives information about the starch granules.
When unmodified starch granules are observed using polarized light, two
dark lines intersecting at the hilum will form a cross or a V-shape. The
shape of the cross can be used to help identify the type of starch. One
explanation for this feature suggests that the density and distribution of
moisture throughout the granule are not uniform and the hilum contains
more moisture than the other regions [lo]. As the granules dry,stresses
are formed within the granule resulting in the bright regions observed
under the polarized light. When the starch swells or is gelatinized, the
cross is no longer visible with the polarizing microscope [32]. The
absence of this cross is a simple and accurate determination of the
presence of gelatinized granules in a starch sample.
An example of an unmodified corn starch suspended in water is given in

Figure 12a. The polygonal and round shape is evident for the granules,
but less detail is obtained when compared to the SEM micrographs. The
sample was also analyzed under crossed polarizers, and the crosses are
clearly observed for the granules, as shown in Figure 12b. When partially
pregelatinized starch 1500 is suspended in water and observed under the
optical microscope, very few details of the particles are discerned, as
552 ANN W. NEWMAN E T A L
Figure 12a Optical micrograph of unmodified corn starch, using

ordinary illumination at a magnification of 400x
STARCH 553
Figure 12b. Optical micrograph of unmodified corn starch, using crossed

polarizers at a magnification of 400x.
55-1 ANN W. NEWMAN ETAL.
Figure 13, Optical micrograph of pregelatinized Starch 1500, obtained

using ordinary illumination at a magnification of 200x.
STARCH 555
Figure 13b. Optical micrograph of pregelatinized Starch 1500, obtained

using crossed polarizers at a magnification of 200x.
shown in Figure 13a. The transparency of the particles is due to the

refractive index similarities between the sample and the solvent. The
irregular shape of the particles is evident, however, and a few intact
granules may be present. Figure 13b is a micrograph of this sample under
crossed polarizers. It is evident that the majority of the sample was
pregelatinized, however, some intact granules exhibiting crosses are also
visible in the lower left quadrant.
4.9 Particle Size Measurements
Particle size can affect the disintegration, flow, handling, and tableting
properties of these materials. Sieving is a common method for obtaining
specific size fractions for granulation and disintegration studies [37-401.
Other studies characterize the particle size distribution of the materials as
received from the vendors to investigate possible variations in the
properties of the excipient.
A wide range of granule sizes, spanning from 2 to 150 pm, has been
reported for various starches [lo]. The size of the granule is generally
expressed as the length of its longest axis in microns [3 11. The results of a
number of measurements can be expressed either as a range or as an
average size. Starches such as rice show a relatively uniform distribution,
therefore, an average size is appropriate. Other starches, such as maize,
show a wide distribution of sizes and a range would more accurately
describe the granule size. Rye and wheat starches are known to exhibit
bimodal distributions of very large and very small granules.
A number of methods for determining the particle size of various starches

have been used. For bulk powder analysis, sieving is employed for large
amounts of material [ 13. One of the most common methods for particle
size determination is optical microscopy 141,423 because it gives a direct
measurement of the individual particles. Automated systems have been
used to examine the particle sizes of starch materials with good results
[43,44]. Laser light scattering analysis has also been utilized to measure
the size of dry particles (particles in air) and suspensions (particles in
liquid) [45]. This technique is suitable in many cases, but since it is not a
STARCH 557
direct measure of the particles, the data was checked using SEM. It was
concluded that laser light scattering analysis was dependent on the model
used to fit the data and better reproducibility was obtained with samples
suspended in liquid.
The particle size distributions of starch samples from the various vendors
were collected using optical microscopy and an image analysis system.
This type of measurement was suitable for the fine particles in the sample
but large aggregates would not have been included. The mean particle
sizes are summarized in Table VI. A measure of 95% of the particles is
also given in Table VI to assess the number of large particles in the
sample. The mean particle size exists in a narrow range from 5.7 x 9.4 to
12.8 x 18.9pm. The different processing of the various starches does not
appear to have substantially changed the mean particle size of the fines.
The distributions are shown graphically in Figure 14 for the particle length
range 0-30 pm since this range included most of the particles measured.
The distribution of the particle length is similar for the five starch samples,
with the majority of the particles having a length between 4 and 20 pm.
STA-Rx and Starch 1551 appear to have more particles in the range 2 to
10 pm, whereas Starch 1500 and 1500 LM appear to have more particles
in the 10 to 20 pm range.
4.10 Surface Area Measurements
Surface area measurements of starches have been obtained by air

permeametry [9,39] or nitrogen adsorption [46]. Relatively low surface
areas ranging from approximately 0.1 to 0.5 m2/ghave been reported [9],
however, values as high as 3 m2/ghave also been observed [46].
Surface area measurements on unmodified and pregelatinized corn starch

samples were obtained using a five-point nitrogen BET analysis after
outgassing the samples at room temperature under vacuum. The values
are listed in Table VI. The surface areas range from 0.18 to 0.36 m2/g
which is in agreement with literature values reported for similar materials.
The low surface areas can be partially explained by the low porosity of
the materials as seen in the SEM photos.
Table VI. Particle Size and Surface Area Data for Starch Lots
Starch Type Mean Particle 95% Less-Than Surface Area

Size (pm) Value (pm> * (m2/g)
STA-RX 8.4 x 11.1 17.3 x 22.6 0.35
Purity 21 10.6 x 15.0 25.4 x 35.7 0.36
1551 5.7 x 9.4 15.1 x 24.7 0.18
1500 12.8 x 18.9 31.1 x 46.2 0.26
1500 LM 10.1 x 14.8 19.6 x 28.9 0.24
* Particle size for which 95% of the measured particles were smaller
than.
STARCH 559
Figure 14. Particle size distribution of commercial starches.
0'2-82
8Z-92
PZ-ZZ
zz-oz
OZ-81
81-91
91-91
P 1-2 1
ZL-01
01-8
8-9
9-P
P-Z
4.1 1 Bulk Powder Properties
Bulk powder properties are important in understanding the handling

properties of an excipient or a granulated product. A number of studies
have investigated the bulk powder properties of starch [9,42,47] and
granulations made with starch [6,12,48]. Common parameters measured
are bulk and tap density. From these values the compressibility can be
calculated using the following equation:
(tap density - bulk density)

100 x = % compressibility (1)
tap density
A classification system to evaluate the flow properties of powders has

been introduced by Can: [49,50]. A flowable powder is defined as free-
flowing and will tend to flow steadily and consistently, whereas, a
floodable powder will exhibit an unstable, discontinuous, and gushing
type of flow. The parameters in Carr's system include angle of repose,
angle of spatula, compressibility, cohesion, and dispersibility. Based on
these parameters, flowability and floodability indices are calculated to
determine the handling properties of bulk solids.
Various starches have been characterized using C a d s system and the

Hosokawa powder characteristics tester [47]. To compare the bulk
powder properties of unmodified and pregelatinized corn starches from the
various vendors, the same instrumentation and procedures were used. The
results are summarized in Table VII.
The parameters used to obtain the flowability index are compressibility

(from bulk and tap density), angle of repose, and angle of spatula. The
values obtained for these tests were similar to those previously reported
for similar materials [47]. The flowability indices of 3 1 and 36 for the two
unmodified starches are similar and indicate the materials to be very
STARCH 56 1
poorly flowing powders. The pregelatinized starches exhibited flowability

indices in a narrow range of 52-54 indicating that these materials would
also have poor flow properties. The floodability indices were evaluated
using the angle of fall, angle of difference, and dispersibility
measurements. The floodability indices for all five starches ranged from
62 to 86. Values in this range are indicative of floodable material which
will exhibit difficult handling properties for the bulk powder. Previous
data on similar materials reported poor to borderline flow characteristics
using this system [47].
The mass flow rate data are also given in Table VII for comparison. The
relatively low mass flow rates ranging from 0.04 to 1.OO g/sec demonstrate
the poor flow properties of these materials. The unmodified starch
samples exhibited the lowest flowability indices and the lowest mass flow
rates. The small particle size is largely responsible for the floodable
properties of these materials.
4.12 Hygroscopicity
Starch has been classified as a moderately hygroscopic material [5 13.

Water sorption studies have been conducted using static methods
(saturated salt solutions in closed chambers) [5 1-53], modified inverse
frontal gas chromatography [54], as well as other techniques [27,55]. The
isotherms are typically type I1 curves exhibiting hysteresis (an amount of
sorbed moisture remaining during desorption) [5 11. Hysteresis is usually
attributed to ink bottle pores in which the water cannot escape the pore due
to the constricted neck. In the case of starch, the hysteresis has been
attributed to intra- and intermolecular hydrogen bonding of water with the
hydroxyl groups of the starch molecule [51,55,56]. The extent of
hydration and swelling depends on the accessibility of the hydroxyl groups
in the starch to the water [55]. It has been suggested that the amorphous
regions are responsible for the reversible swelling of starch upon the
adsorption of water [131.
S62 ANN W. NEWMAN ET AL.
Table VII. Bulk Powder Data for Starch Lots
Property STA-Rx Purity 21 Starch Starch Starch

1551 1500 1500 LM
1.54 1.54 1.55 1.54
Bulk density 0.46 0.55 0.46 0.60 0.63

(g/mL)
Tap density 0.79 0.83 0.65 0.87 0.86

(gimL)
Compressiblity 42 34 30 31 27
(3.)
Angle of 62 62 43 36 41
Repose (deg.)
Angle of 84 77 64 72 63
Spatula (deg.)
Dispersibility 7 6 8 6 4
Cohesion 4 5 15 12 12
Angle of Fall 48 54 28 28 25
(deg.)
STARCH 563
Table VII. Bulk Powder Data for Starch Lots (continued)
Property STA-Rx Purity21 Starch Starch Starch

1551 1500 1500 LM
Angle of 14 8 15 9 16
Difference
(deg.1
Flowability 31 36 52 54 53
Index
Flowability very verypoor poor poor poor

performance poor
Floodability 68 62 72 72 86
Index
Floodability floodable floodable floodable floodable very

performance floodable
Mass Flow 0.04 0.06 0.53 0.83 1.oo

Rate (g/sec)
.%-I ANN W. NEWMAN ET AL.
The amount of water sorbed by a solid can be expressed in a number of'

ways. Many investigators report the amount of water sorbed, but the
amount of water initially in the sample is not taken into account. The
calculation of percent uptake relative to the dry weight of the sample
normalizes samples to the same initial point and makes data from various
samples comparable. The percent sorbed relative to dry weight is
calculated from equation 2.
where: W, = weight of sample at equilibration

W, = original weight of sample
A = percent moisture in original sample
A second method of reporting data is the equilibrium moisture content

(EMC) [ 1,571 which is calculated using the following equation:
EMC = x 100
P + 100
where
[ Wo x 1- A *B
100
P = X I 00 (4)
A
wo - P o x 1-
100
B = weight change at equilibrium
'The hygroscopicity of the commercial starches was investigated with an

automated moisture balance system in the range 0 to 90% RH [%I. EMC
sorption values were calculated from the raw data at each relative
STARCH 565
humidity and are summarized in Table VIII. Representative curves are

given in Figures 15 and 16.
The sorption curves for the various starch samples are similar. The Type
I1 isotherm and the hysteresis are evident for all the samples and the
weight percent sorbed at 90%RH is fairly consistent at about 20%. The
unmodified corn starch data are shown in Figure 15 for the Purity 2 1
sample. It is apparent that unmodified corn starch readily sorbs moisture
along the entire range up to 90%RH. The isotherm for the pregelatinized
starch in Figure 16 is similar, but the magnitude of the hysteresis is larger
in the range 30-80%RH. These sorption curves are in agreement with data
reported for similar types of samples [ 11.
The National Formulary [59] contains the following assays for starch:
Botanic characteristics: Corn starch is described as polygonal,

rounded, or spheroidal granules up to about 35 pm in size and
usually having a circular or several rayed central cleft.
Identification: A smooth mixture of starch and cold water is

made. Boiling water is added and the mixture is boiled gently for
2 minutes. When the product is cooled, the product is a
translucent, whitish jelly. In a second test, a water slurry is colored
reddish violet to deep blue by iodine test solution (TS).
Microbial limits: Following test method <61>, the sample meets

the requirements for the absence of Salmonella species and
Escherichia coli.
pH: A slurry of starch and water is prepared in a nonmetallic

container and agitated continuously for 5 minutes. The pH is
measured immediately to the nearest 0.1 unit. For corn starch, a
Figure 15 Moisture isotherm for unmodified corn starch, purity 2 1 .
0
- m
0
STARCH 5 67
Figure 16. Moisture isotherm for pregelatinized
~. corn starch, Starch
1551.
568 ANN W.NEWMAN ET AL.
Table VIII. Sorption EMC Values at Various Relative Humidities
Relative STA-Rx Purity 21 Starch Starch Starch

Humidity 1551 1500 1500 LM
(%I
1 3.4 4.4 3.6 3.5 3.6
10 6.0 6.3 5.6 5.4 5.4
20 8.1 8.3 7.3 7.5 6.9
30 9.5 9.6 8.6 9.0 7.9
40 10.7 10.8 9.7 10.1 9.1
50 11.9 11.9 10.8 11.2 10.7
60 13.1 13.2 12.2 12.1 12.3
70 14.6 14.9 14.1 13.2 14.1
80 16.8 16.9 16.8 15.5 16.2
90 19.9 20.1 21.5 19.2 20.5

STARCH 569
value between 4.5 and 7.0 should be obtained.
Loss on drying: Following the general test method <73 1>, the
sample is dried at 120°C for 4 hours. It should not lose more than
14.0% of its weight.
Residue on ignition: Following test method <281>, the weight of

the residue can not exceed 0.5% on a 2 g sample.
Iron: The residue obtained in the test for Residue on ignition is

dissolved in hydrochloric acid with gentle heating, diluted with
water, and mixed. This sample is diluted with water and tested for
iron according to test <241>. The limit is 0.002%.
Oxidizing substances: Starch and water are swirled for 5 minutes

and decanted into a centrifuge tube. It is centrifuged to clarify.
The clear supernatant is transferred and glacial acetic acid and
potassium iodide are added. The sample is swirled and allowed to
stand in the dark for 25-30 minutes. Starch TS is added and the
solution is titrated with sodium thiosulfate volumetric solution
(VS) until the starch-iodine color disappears. Not more than 1.4
mL of sodium thiosulfate is required.
Sulfur dioxide: Starch and water are mixed until a smooth

suspension is obtained. The solution is filtered. Starch TS is
added to the clear filtrate and the solution is titrated with iodine to
the first permanent blue color. Not more than 2.7 mL is consumed.
Organic volatile impurities: The sample should meet the

requirements for Method N in general test <467>.
5.2 Chromatomaphv
High-pressure liquid chromatography (HPLC) methods have been used by

the food [60] and paper [61] industries to analyze for starch. One method
uses acid hydrolysis of the starch to detect glucose as the starch
degradation product [60]. Good agreement was obtained with titrimetric

data. A second method converts the starch to dextrose using enzymes and
subsequent HPLC analysis [61]. The chromatographic method was found
to be more reproducible than a spectrophotometric method in quantifying
starch.
6. Excipient Studies
6.1 Drug Compatibilitv
Starch is a relatively inert material and interactions with active drug

substances do not occur often. Excipient compatibility studies of starch
and various active drugs have been performed using thermal methods of
analysis. As an example, starch has been found to be compatible with
erythromycin [62], cephalexin [63], and acetylcysteine [64] using this
method of excipient screening.
6.2 Tableting
Various properties of starch have been studied to better understand its role
in the granulation and compaction processes. It has been reported that
starches deform mostly by plastic flow, but this was found to be dependent
on the particle size, size distribution, and particle shape [38].
Characterization of wet granulated formulations made with starches from
various vendors have shown no physical differences, yet, the compaction
and ejection forces were found to be different for the formulations [48].
The effect of heat formation during compression has also been
investigated for various starches and excipients [65,66].
The effect of the binder and tablet geometry have been related to the
hardness and tensile strength of starch tablets [67]. Studies have shown
that the particle size of an excipient may also affect various tablet
properties [38,39,41,68]. A decrease in the particle size of compressible
starch has resulted in a decrease [68] and an increase [38] in tablet
strengths.
STARCH 57 1
The tabletting properties of starch in various formulations with active drug

substances have also been reported [4,37,69]. Various starches (maize,
rice, potato, wheat) were studied for use in double compressed tablets of
aspirin and it was found that corn starch exhibited the best disintegration
properties while rice exhibited the worst [4]. The compression speed used
to tablet aspirin and compressible starch showed weaker and more porous
tablets at high compression speeds [37]. A comparison of compressible
starch with regular starch in an aspirin formulation resulted in better flow
and compressibility properties for the compressible starch, as expected
from the properties of the two materials [69].
6.3 Disintegation
As discussed previously, starch is used in many formulations for its

disintegrationproperties. The mechanism for the disintegration properties
of starch has been studied and various conclusions have been published
[70-731. The most common explanation for the disintegration properties is
the swelling of the starch granules when exposed to water and it has been
proposed that amylose is the component responsible for the disintegration
properties of starch due to swelling [69]. Measurements of the granules to
quantify the amount of swelling have been performed using various
methods [56,74]. A second mechanism was proposed that suggested the
disintegrating action of starch in tablets is due to capillary action rather
than swelling [70]. A third mechanism has been proposed based on the
particle-particle repulsion forces between the tablet constituents when in
contact with water and the hydrophilic nature of starch [72].
When comparing various starches as tablet adjuvants, it was found that

disintegration time for the tablets was independent of the compressional
force, but was dependent on the type of starch [5,42] and the dissolution
method [5]. A study of compression forces showed that starch tablets had
a decrease in disintegration with an increase in tablet hardness [73].
Particle size has also been correlated with the disintegration times of
tablets containing starch [39], and it has been suggested that tight controls
on particle size can minimize batch failures [44].
7. References
1. Handbook of Pharmaceutical Excipients, American

Pharmaceutical Association, Washington, 1986, pp. 289-293.
2. McGraw-Hill Encyclopedia of Science and Technology, McGraw-

Hill Book Company, New York, Vol. 17, 1987, pp. 326-328.
3. D. R. Lineback, J Jpn. SOC.Starch Sci., 33(1):80-88 (1986).
4. M. A. F. Gadalla, M. H. Abd El-Hameed, A. A. Ismail, Drug.

Dev. Znd. Pharm., 15(3):427-446 (1 989).
5. T. W. Underwood, S. E. Cadwallader, J Pharm. Sci., 61(2) 239-

243 (1972).
6. R. N. Nasipuri, F. 0. Kuforji, Pharm-Ind., 43(10):1037-1041

(1981).
7. T. Ishizaka, H. Honda, M. Koishi, J Pharm. Pharmacol., 45:770-

774 (1993).
8. H. Yoshizawa, M. Koishi, J Pharm. Pharmacol., 42:673-678

(1990).
9. C. E. Bos, G. K. Bolhuis, H. Van Doome, C. F. Lerk, Pharm-

Weekhl-Sci-Ed, 9:274-282 (1987).
10. R. W. Kerr, ed., Chemistry and Industry of Starch, Academic

Press, Inc., New York, 1950.
1I . R. L. Whistler and J. R. Daniel, Starch. In: Kirk-Othmer

Encyclopedia of Chemical Technology, 3rd ed. (M. Grayson, ed.).
John-Wiley and Sons, New York, Vol. 21, 1983, pp. 492- 507.
STARCH 573
12. J. B. Schwartz, E. T. Martin, E. J. Dehner, J. Pharm. Sci.,

64(2):328-332 (1975).
13. A. D. French. In: Starch: Chemistry and Technology, 2nd ed. (R.L.
Whistler, J. N. BeMiller, and E. F. Paschall, eds.), Academic Press,
New York, pp. 183-247 (1984).
14. 0. B. Wurzburg. In: Handbook of Food Additives (T. E. Furia, ed.),

The Chemical Rubber Company, Boca Raton, FL, Vol. 1, 1972,
pp. 361-395.
15. W. Banks, C. T. Greenwood, Starch and its Components, John

Wiley and Sons, New York, 1975.
16. A. Imberty, H. Chanzy, S. Perez, J. Mol. Biol., 201:365-378

(1988).
17. A. Imberty, A. Buleon, V. Tran, S. Perez, Staerke, 43(10):375-384

(1991).
18. B. Casu and M. Reggiani, Staerke, 7:218-229 (1966).
19. F. R. Dollish, W. G. Fateley, F. T. Bentley. In: Characteristic

Raman Frequencies of Organic Compounds, John Wiley & Sons,
Inc., New York, 1974.
20. M. St-Jacquies, P. R. Sundararajan, K. J. Taylor, and R. H.

Marchessault, J. Am. Chem. SOC.,98( 15):4386-4391 (1976).
21. J. W. Donovan, Biopolymers, 18:263-265 (1979).
22. P. J. Florey, i;inciples of Polymer Chemistry, Cornell University

Press, Ithaca, NY, 1953, chapter 13.
23. J. Lelievre, J. Appl.Polym. Sci. , 18:293-296 (1 973).

24. J. Lelievre, Polymer, 17:854-858 (1976).
25. K. J. Zeleznak, R. C. Hoseny, Cereal Chem., 64(2):121-124

(1987).
26. D. J. Stevens, G. A. H. Elton, Staerke, 23(1):8-11 (1971).
27. M. Nduele, A. Ludwig, M. Van Ooteghem, S. T.P. Pharma Sci.,

3(5):362-368 (1993).
28. M. Tomessetti, L. Campanella, T. Aureli, Thermochimica Acta,

143:15-26 (1989).
29. 0. A. Sjostrom, Ind. Eng. Chem., 28(1):63-74 (1936).
30. T. J. Schoch, E. C. Maywald,Anal. Chem., 28(3):382-387 (1956).
31. G. E. Moss, The Microscopy of Starch. In: Examination and

Analysis of Starch and Starch Products (J. A. Radley, ed.),
Applied Science Publishers, Ltd, London, 1976, pp. 1-32.
32. T. J. Schoch, E. C. Maywald, Industrial Microscopy of Starches.

In: Starch: Chemistry and Technology, 2nd ed. (R.L. Whistler, J.
N. BeMiller, and E. F. Paschall, eds.), Academic Press, New York,
pp. 637-647 (1984).
33. D. J. Gallant, Electron Microscopy of Starch and Starch Products.

In: Examination and Analysis of Starch and Starch Products (J. A.
Radley, ed.), Applied Science Publishers, Ltd, London, 1976, pp.
33-59.
34. W. C. Mussulman, J. A. Wagoner, Cereal Chem., 45(2): 162-17 I

(1968).
35. R. F. Shangraw, J. W. Wallace, F. M. Bowers, Pharm. Tech., Oct.,

44-60 (1981).
STARCH 515
36. H. Hess, Pharm. Tech., June, 54-68 (1987).
37. G. D. Cook, M. P. Summers, J Pharm. Pharmacol., 42:462-467

(1990).
38. A. McKenna, D. F. McCafferty, J. Pharm. Pharmacol., 34:347-

351 (1982).
39. A. J. Smallenbroek, G. K. Bolhuis, C. F. Lerk, Pharm. Weekblad,

11611048-1051(1981).
40. C. E. Bos, H. Vromans, C. F. Lerk, Znt. J. Pharm., 67:39-49

(1991).
41. P. Paronen, M. Juslin, J. Pharm. Pharmacol., 35:627-635 (1983).
42. M. Juslin, P. Kahela, P. Paronen, 1. Turakka, Acta. Pharm. Fenn.,

83-93 (1981).
43. K. P. P. Prasad, L. S. C. Wan, Pharm. Res., 4(6):504-508 (1987).
44. H. G. Brittian, C. J. Sachs, K. Fiorelli, Pharm. Tech., Oct., 38-52

(1991).
45. P. Merkku, J. Yliruusi, E. Kristoffersson, Acta Pharm. Nord.,

4(4):265-270 (1992).
46. N. N. Hellman, E. H. Melvin, J Am. Chern. SOC.,74:348-350

(1952).
47. B. Vennat, S. Gross, A. Pourrat, H. Pourrat, Drug Dev.Znd

Pharm., 19(11):1357-1368 (1993).
48. M. K. Kottke, H.-R. Chueh, C. T. Rhodes, Drug Dev. Ind. Pharm.,

18(20):2207-2223(1992).
49. R. L. Carr, Chem. Eng. 72:163-168(1965).
50. R. L. Carr, Chem. Eng. 7369-72(1965).
51. D. Faroongsamg, G. E. Peck, Drug Dev. Ind. Pharm., 20(5):779-

798 (1994).
52. L. Sair, W. R. Fetzer, Ind. Eng. Chem., 36:205-208(1944).
53. S. Malamataris, P. Goidas, A. Dimiriou, Znt. J. Pharm., 6851-60

(1 99I).
54. S. W. Paik. S. G. Gilbert, J. Chrom., 351:417-423(1986).
55. B. Das, R. K. Sethi, S. L. Chopra, Isr. J. Chem., 10:963-965

(1972).
56. S. E. Wurster, G. E. Peck, D.O. Kildsig, Drug Dev. Znd.

Pharm.,8(3):343-354(1982).
57. J. C. Callahan, G. W. Cleary, M. Elefant, K. Kaplan, T. Kensler, R.

A. Nash, Drug Dev. Ind. Pharm., 8(3):355-369(1982).
58. VTI Moisture Balance System, VTI Corporation, Hialeah, FL.
59. The National Formulary, NF 18,United States Pharmacopeial

Convention, Inc, Rockville, MD, 1995,pp. 2309-2310.
60. N. P. Boley and M. J. S. Burn, Food Chem., 36:45-51 (1990).
61. N.Rirosel-Boettcher, Tuppi J., 76(3):207-208(1993).
62. H. H. El-Shattawy, D. 0. Kildsig, G. E. Peck, Drug. Dev. Znd.

Pharm., 8(6):937-947 (1982).
STARCH 511
63. H. H. El-Shattawy, D. 0. Kildsig, G. E. Peck, Drug. Dev. Ind.

Pharm., 8(6):897-909 (1982).
64. J. Kerc, S. Srcic, U. Urleb, A. Kanalec, B. Kofler, J. Smid-Korbar,

J Pharm. Pharmacol., 445 15-5 18 (1992).
65. N. A. Armstrong, A. M. Gough, I. L. Baker, Int. J. Pharm. Tech.

andprod. MJi..,
6(2):13-15 (1985).
66. D. E. Wurster, J. R. Creekmore, Drug. Dev. Ind. Pharm.,

12(10):1511-1528 (1986).
67. R.-C. Hwang, E. L. Parrot, Drug Dev. Ind. Pharm., 19(5):507-519

(1993).
68. G. Alderborn, C. Nystrom, Acta Pharm Suec., 19:381-390 (1982).
69. K. S. Manudhane, A. M. Contractor, H. Y. Kim, R. F. Shangraw, J.

Pharm. Sci., 58(5):616-620 (1969).
70. L. C. Curlin, J Am. Pharm. Assn., Sci.Ed., 44(1):16 (1955).
71. N. R. Patel, R. E. Hopponen,J Pharm. Sci.,55(10):1065-1068

(1966).
72. A. M. Guyot-Hermann, Drug Dev. Ind. Pharm., 7(2):155-177

(1981).
73. P. M. Hil1,J Pharm. Sci., 65(11):1694-1697 (1976).
74. N. N. Hellman, T. F. Boesch, E. H. Melvin, J. Amer. Chem. SOC.

74:348-350 (1952).
TOBRAMYCIN
Alekha K. Dash
Department of Pharmaceutical Sciences
School of Pharmacy and Allied Health Professions
Creighton University
Omaha, NE 68 178
ANALYTICAL PROFILES OF DRUG SUBSTANCES 579

AND EXCIPIENTS-VOLUME 24
580 ALEKHA K. DASH
TABLE OF CONTENTS
2. Description
2.1 Nomenclature
2.1.1 Chemical Name
2.1.2 Generic Name
2.1.3 Trade Names
2.4 Appearance, Color and Order
2.5 Pharmaceutical Dosage Forms
3. Synthesis
4.2 H Nuclear Magnetic Resonance Spectrum
4.5 Mass Spectrum
4.6 Thermal Behavior
4.7 Melting Point
4.8 Solubility
4.9 X-Ray Powder Diffraction Patterns
TOBRAMYCIN 581
5.2 Spectrophotometric
5.3 Chromatographic
5.3.2 High Pressure Liquid Chromatography
5.4 Biological
5.4.1 Microbiological Assay
5.4.2 Radioimmuno Assay
5.4.3 RadioenzymaticAssay
5.4.4 Fluorescence Polarization Immunoassay
5.4.5 Fluorescence Immunoassay
6. Stability, Degradation and Incompatibility
7. Pharmacokinetics
8. Toxicity
9. Acknowledgments
10. References
582 ALEKHA K. DASH
1. History and Therapeutic properties
In 1957, investigators at Lilly Research Laboratories first isolated a

streptomyces species from soil samples collected in Hermosillo (Sonora,
Mexico). Seven antibiotic factors were isolated from Streptomyces
tenebrarius (ATCC 17920) [ 11, and factor 6 was subsequently designated
as tobramycin [2]. Tobramycin is an aminoglycoside antibiotic, exhibits
bactericidal activity against a broad spectrum of bacteria, and is only active
on actively multiplying bacteria. It inhibits protein synthesis, possibly on
the 30s subunits of the bacterial ribosomes [3]. Indications of this drug
include sepsis, urinary tract infections, infections of the skin, soft tissue
infections, respiratory tract infections, erc. It is especially useful in
treatment of infections due to Pseudomonas and indolepositive Proteus.
2. Description
2.1 Nomenclature
2.1.1 Chemical Name
0-3-amino-3 -deoxy-a-D-glucopyranosy1-(1-6)-0- [2,6-diamino-

2,3,6-trideoxy-a-D-ribo-hexopyranosyl-(
1-4)]-2-deoxy-D-
streptamine.
0-3-amino-3-deoxy-a-D-glucopyranosyl-( 1-4)-0-[2,6-diamino-
2,3,6-trideoxy-a-D-ribo-hexopyranosyl-(
1-6)]-2-deoxy-L-
streptamine.
2.1.2 Generic Name
Tobramycin
2.1.3 Trade Names
Distobram, Gernebcin, Obramycin, Nebcin, Tobradistin,

Tobralex, Tobramaxin, and Tobrex.
TOBRAMYCIN 583
32986-56-4
Free base C I 8H37N509 MW=467.45

Sulfate salt (C18H37N509)2 5 H2S04 MW = 1425.4
The theoretical elemental composition of tobramycin, based on the

molecular formula C18H37N509, is: C 46.24% , H 7.98%, N 14.98%,
0 30.80% [4]. Elemental analysis of tobramycin monohydrate has
been reported by Koch et al. [5] as:
Calculated, C: 44.52%; H: 8.10%; N: 14.43%.
Found, C: 44.39%; H: 8.15%; N: 14.07%.
2.4 Appearance, Color and Odor
Tobramycin is obtained as a crystalline, white to off-white, hygroscopic

powder.
5x4 ALEKHA K. DASH
2.5 Pharmaceutical Dosage Forms
Tobramycin is available in the following dosage forms: tobramycin

injection, tobramycin sulfate injection, tobramycin ophthalmic ointment and
tobramycin ophthalmic solution.
3. Synthesis
Takagi et al. [6] have synthesized tobramycin from kanamycin B.

Kanamycin B was converted to penta-N-ethoxycarbonyl-4",6"-0-
cyclohexylidene-2"-benzoylkanamycinB. This compound (1 mol
equivalent) was then treated with excess of p-toluenesulfonyl chloride (5
mol equivalent) in pyridine at 25°C overnight to give the 3'-O-tosyl
derivative as the major product. Iodination of 3'-O-tosyl derivative (mp
149-150°C) was achieved after reacting with sodium iodide in
dimethylformamide (4.9 g NaI in 10 mL of DMF) at 100°C to produce a
unstable 3I-iodide derivative. This unstable derivative was subsequently
hydrogenated with Raney nickel and hydrogen in dioxane to give 2"-0-
benzoyl-4",6"-O-cyclohexylidene-3'-deoxy-penta-N-
ethoxycarbonylkanamycin (mp 2485250°C). This compound was
successively treated with hot 4N barium hydroxide and 50% v/v acetic acid
at 80OC to give crude 3'-deoxykanamycin B. This material was purified by
chromatography and recrystallized as a monohydrate.
The infrared spectrum of tobramycin is shown in Figure 1. The spectrum

was obtained in potassium bromide disk (0.5% w/w) using a FTIR (model
1600,Perkin-Elmer spectrophotometer. Assignments of the characteristic
bands in the spectrum are listed in Table 1 [7].
c
.d
586 ALEKHA K. DASH
Table 1
Infrared Spectral Assignments for Tobramycin
Energy (cm-1) Assignment
3400 - 3200 N-H stretching (s, br)

0-H stretching (s, br)
2910 Aliphatic C-H stretching (m)
1588 N-H bending (s)
1461 CH2 scissoring (m)
1377, 1349 0-H inplane bending vibration (m)
1032 C-N stretching ( s )

C - 0 stretching (s)
(br) = Broad
(m) = Medium intensity
(s) = Strong intensity
TOBRAMYCIN 587
4.2 'H Nuclear Magnetic Resonance Spectrum
The 200-MHz proton nuclear magnetic resonance spectrum of tobramycin

was obtained on a Bruker AM 200 NMR spectrometer, and is shown in
Figure 2. The spectrum was recorded at an ambient temperature.
Deuterated water (D20) was used as the solvent, and tetramethylsilane was
the reference standard [7].] The chemical shifts, multiplicities and peak
assignments of characteristic protons are given in Table 2, and these were
found to be close to the reported values [5].
The 15.08 MHz 13C Nuclear Magnetic Resonance spectrum of 0.3-0.5 M

aqueous solution of tobramycin were recorded in aqueous solution,
containing 5% dioxane as an internal standard. The nmr instrument
consisted of a Varian Associates DP-60 magnet working at 14 kG with an
external 19F lock. The samples were spun in 13-mm 0.d. tubes [8].
Chemical shifts and structural assignments are outlined in Table 3, and are
based on the number system given above.
Figure 2. Proton Magnetic Resonance Spectrum of Tobramycin free
base in D7O.
TOBRAMYCIN 589
Table 2
N M R Spectral Assignments for Tobramycin

Chemical Shift Multiplicities Number of Assignment
@Pm> protons
6.25 d 1 Anomeric Protons

6.05 d 1 (1' and 1")
4.15-4.85 m 10 Protons on carbon

bonded to hydroxyl
group or ether linkage
3.85-4.15 m 6 Protons on carbon

bonded to amino
groups
2.9-3.17 m 2 (es) Methylene group in

a six membered ring
2.55 9 1 (ax) Methylene group in

a six membered ring
2.25 9 1 (ax> Methylene group in

a six membered ring
d = doublet eq = equatorial
m = multiplet ax = axial
q = quartet
590 ALEKHA K.DASH
Table 3
C NMR Spectral Assignments for Tobramycin [81
Chemical Shift Assignments

@Pm>
99.2 c-1'
49.5 c-2'
34.7 c-3'
65.9 c-4'
73.1 c-5'
41.5 C-6'
50.2 c-1
35.5 c-2
49.0 c-3
86.0 c-4
74.4 c-5
87.8 C-6
99.1 c-1"
71.6 c-2"
54.2 c-3"
69.2 c-4"
71.9 c-5"
60.2 C-6"
TOBRAMYCIN 59 1
-
Owing to its saturated ring system and lack of suitable chromophores,
tobramycin does not exhibit any significant absorption between 230 and 360
[91.
4.5 Mass Spectrum
A Finnigan INCOS-SOB quadrupole mass spectrometer linked to a Hewlett-

Packard gas chromatograph using electron impact at an electron energy of
70 eV and a source temperature of 180°C was used to study the mass
spectrum and fragmentation behavior of tobramycin. Unfortunately, no
useful mass spectra were obtained, as had been reported [lo].
4.6 Thermal Behavior
The Differential Scanning Calorimetry (DSC) thermogram of tobramycin

base is shown in Figure 3. The sample was heated from 30 - 25OoC in a
nonhermetically crimped aluminum pan at a rate of 1O"C/min on a DuPont
model 950 thermal analysis system. The first endothermic peak was
attributed to compound dehydration, and was followed by the melting of the
metastable form at 164OC. The metastable forms recrystallizes to the stable
form as evidenced by the exotherm at 197SoC, and finally melts at 217OC
[111.
Thermogravimetric (TG) analysis of the base was carried out on a DuPont

model 95 1 thermogravimetric analyzer, and the resulting TG and
differential TG thermograms are shown in Figure 4. It was concluded from
this work that the commercially available sample consisted of the
monohydrate phase containing some absorbed water [l 11.
4.7 Melting Point
The metastable form was observed to melt at approximately 164OC, while

the stable form melts at approximately 2 17OC [111.
113.69-C
t63.91°C
216.78-C
I 1
I I
50 100 150 200 PI
Temperature ('13
Figure 3. Differentialscanning calorimetry thermogram of

Tobramycin free base.
0.W
0.06
100 -
-u
0.04 5
bJ
C
98- 0
-..
VI
:
w
W 0.02 ;
.LI
a.
u
0
96-
0.00
-0.02
Figure 4. (a) Thermogravimetricanalysis and (b) differential

thermogravimetricanalysis thermograms of Tobramycin free
base.
594 ALEKHA K. DASH
4.8 Solubility
'Tobramycin is freely soluble in water (1 in 1.5 parts), very slightly soluble

in ethanol (1 in 2000 parts), and practically insoluble in chloroform and
ether [9,12]. A 10% (w/v) solution of tobramycin in water has a pH of 9-1 1
[13].
4.9 X-Ray Powder Diffraction Pattern
The powder pattern data of tobramycin base was obtained using a wide
angle X-Ray diffractometer (model D500, Siemens). The powder
diffraction patterns of the two polymorphs are shown in Figure 5a and 5b.
The calculated d-spacings for the diffraction patterns are provided in Table
4 [14].
In one publication, three pKa values were reported for tobramycin as 6.7,
8.3. and 9.9 [15]. However, in another work four pKa values (6.2,7.4, 7.6,
and 8.6) were reported by Raymond and Born [ 161.
5.1 Identification [13]
Tobramycin is identified by a thin layer chromatographic method, and the

exact details of this procedure are described in the subsequent TLC
discussion (section 5.3.1).
TOBRAMYCIN 595
6.0 8.0 10 12 14 $6 I8 20 22 24
28, degrees
Figure 5. Powder x-ray diffiaction patterns of; (a) commercially

available Tobramycin free base (Form I), and (b)
Tobramycin free base heated to 208OC (Form 11).
SY6 ALEKHA K. DASH
Table 4
Powder X-Ray Diffraction Data for Tobramycin
Diffraction pattern of Diffraction pattern of

Tobramycin base (Form I) Tobramycin base (Form 11)
Peak d-Spacing Relative Peak d-Spacing Relative

No (4 Intensity No (A) Intensity
(%I
1 15.07 27 1 15.77 10
2 10.35 48 2 7.69 18
3 8.87 49 3 7.08 89
4 8.14 36 4 5.90 29
5 7.48 33 5 5.50 29
6 6.17 53 6 4.98 84
7 4.97 57 7 4.72 25
8 4.79 100 8 4.64 20
9 4.69 47 9 4.27 100
10 4.56 35 10 3.95 31
11 4.39 36 11 3.77 27
12 4.24 36 12 3.60 31
13 4.08 54
14 3.91 31
15 3.79 53
TOBRAMYCIN 591
5.2 Spectrophotometric Methods
A colorimetric method based on the reaction between tobramycin and

copper sulfate has been developed for the quantification of this compound
in injectable formulations [17]. A colorimetric method based on the
reaction between tobramycin and 2,4-dinitrofluorobenzene(Sanger's
reagent) has been reported for the quantification of tobramycin in topical
formulations [181.
A spectrofluorimetric method for the determination of tobramycin in

biological fluids using a fluorescent dihydro-lutidine derivative has also
been reported [191. The fluorescent derivative is formed by condensation of
the primary amino groups of tobramycin with acetyl-acetone and
formaldehyde under acidic conditions (pH=2.4).
Sampath and Robinson have reported a spectrophotometricmethod for the

analysis of tobramycin and compared their method with the existing
methods [20].
5.3.1 Thin Layer Chromatography [13]
Tobramycin solution is prepared in distilled water (0.6% w/v). A 3 pL

portion of this solution is applied to a silica gel (0.25-mm layer) TLC plate.
The chromatogram is developed by equilibrating the plate for 5.5 hours in a
chromatographic chamber containing a mixture of methanol, ammonium
hydroxide and chloroform (60:30:25; v/v/v). The plate is removed from the
chamber and heated at 11O°C for 15 minutes. The spots are detected by
spraying with a 1 in 100 solution of ninhydrin in a mixture of butyl alcohol
and pyridine (100:1, v/v), and tobramycin is visualized as a pink spot.
5.3.2 High Performance Liquid Chromatography (HPLC)
The very low absorptivity of tobramycin in the UV and visible region does
not permit its direct quantification at low concentrations. This problem can
be solved by derivatizing this compound with a suitable absorbance-
598 ALEKHA K. DASH
enhancing or fluorescence-producing agent. This deficiency can be

overcome through the use of either pre-column or post-column
derivatization. Various HPLC methods using pre-column [21-261 and post-
column [27,28] derivatization to quantify tobramycin in pharmaceuticals
and biological fluids have been described and are summarized in Table 5 .
5.3.2 Gas Liquid Chromatography (GLC)
A CiLC method has been described by Mayhew et al. [29] for the assay of
tobramycin in biological fluids. A silanized Pyrex column (2 m by 3 mm
id) packed with 3% OV-101 coated onto 80-100 mesh Chromosorb WAW
was utilized in this method. Nitrogen gas was used as a carrier. The
injector and detector temperature were maintained at 272' and 287OC
respectively, and a electron captured detector was used in this study.
5.4 Biological Methods
5.4.1 Microbiological assay
Various microbiological assay methods for the analysis of tobramycin have

been described [24,30,3 11. The method developed by Maitra et al. [24] used
an agar diffusion method using Bacillus subtilis (ATCC 6633) as a test
organism. The organism was grown on seeded agar at 37°C for 16-18
hours. These microbiological assays are reliable and simple but they are
time consuming and less specific.
5.4.2 Radioimmunoassay (RIA)
Radioimmunoassays have been developed for measuring tobramycin in

biological fluids [32,33]. The RIA is based on the competition between
1251-tobramycinand unlabeled tobramycin in the sample to be analyzed for
the antibody binding sites. Unbound 251-tobramycin is separated by
centrifugation, and the radioactivity of the bound tobramycin is counted and
the levels calculated from a standard curve. This method is highly sensitive
and specific.
Table 5
HPLC Methods for the Analysis of Tobramycin
Type of Derivatizing Mobile Phase Column Detection Sample Type Referenc

Derivatization Agent Mode e
Pre-column 1-fluoro-2,4- Water:acetone:acetic C18 uv Biological 21

dinitrobenzene acid (30:70:0.1; v/v/v) 30 cm x 3.9 mm (365 nm) fluids
Flow = 3 mL/min
Pre-column 2,4,6-trinitro- Acetonitri1e:phosphate CIS uv Biological 22

benzenesulfonic buffer (70:30 v/v) 30 c m x 4 mm (365 nm) fluids
acid Flow = 3 mL/min
Pre-column 1-fluoro-2,4- 75% V/V (NH4)3P04 c 18 uv Formulations 23

dinitrobenzene and 25% v/v 30cmx4mm (365 nm)
Acetonitrile
Flow = 2 mL/min
Pre-column o-phthalaldehyde, Methano1:water (72:26) C18 Fluorescence Biological 24

mercaptoethanol and 0.005 EDTA 30 cm x 4 mm (360 nm EX fluids
Flow = 1 mL/min 430 nm EM)
Pre-column o-phthalaldehyde. 250 mL of 0.5 M Tris p-Bondpak C18 Fluorescence Biological 25
mercaptoethanol buffer + 10 mL of (Waters} (360 nm EX fluids
triethylamine, and qs to 430 nm EM)
I L with methanol
Flow = 2 mL/min
Pre-column 2,4,6-trinitro- Acetonitrile5OmM CIS uv Formulations 26

benzenesulfonic Phosphate buffer 25 cm x 4.6 mm (340 nm)
acid (62:38 v/v)
Flow = 2.5 mL/min
Post-column o-phthalaldehyde, Sodium. sulfate (0.1 p-Bondpak C18 Fluorescence Biological 27

mercaptoethanol M), sodium 30 cm x 3.9 mm (340 nm EX fluids
pentasulfonate (0.02 (Waters) 4 18 nm EM)
M), and 17.4 mM
acetic acid in 1 L of
water
Flow = 2 mL/min
Post-column o-phthalaldehyde water:methanol:acetic Lichosorb 5 RP Fluorescence Biological 28

acid (99.7:0.2:0.1 mole C8 (15 cm) (340 nm EX fluids
%) containing 0.2 (Chrompack) 418 nm EM)
moles of sodium sulfate
and 0.02 moles of
sodium pentane sulfate,
Flow = 1 mL/min
TOBRAMYCIN 60 1
5.4.3 Radioenzymatic assay
Radioenzymatic assays for the assay of tobramycin in biological fluids have

been reported [34-361. The method involves the specific enzymatic transfer
of a radioactive modifying group to the drug. These enzymes are present in
organisms that carry resistant (R) factors which are responsible for the
activation of the drug. The entire reaction process is stoichiometric, and the
amount of radioactivity incorporated is proportional to the concentration of
the antibiotic. These assays are simple, accurate and precise, but the need to
work with radioactive material may pose a disadvantage for some clinical
laboratories.
5.4.4 Fluorescence polarization immunoassay
Fluorescence polarization immunoassay (FPI) is a method that combines the

principles of competitive protein binding with the principles of fluorescence
polarization, and has also been utilized to determine the tobramycin
concentration in serum [37,38].
5.4.5 Fluorescence immunoassay (FIA)
FIA uses the principle of competitive protein binding, and has been used to
quantify tobramycin in biological samples [28, 39-41]. Competitive
binding reactions are set up with fluorogenic tobramycin reagent, a limiting
amount of antibody against the drug, and the serum sample to be analyzed.
Tobramycin in the serum sample competes with the fluorogenic tobramycin
reagent for the antibody binding sites. The unbound fluorogenic reagent is
then hydrolyzed by P-galactosidase to produce the fluorescence which is
detected as the observable parameter.
6. Stability, degradation and incompatibility
Tobramycin solution in water at pH 1 to 11 was reported to be stable for

several weeks at temperatures from 5 to 37OC, and could be autoclaved
without loss of potency [12]. When aqueous tobramycin was adjusted to pH
602 ALEKHA K. DASH
1.2 with HC1 and autoclaved for 30 minutes at 120°C in sealed glass
ampules, an extra peak was observed in the chromatogram. This was
attributed to a possible degradation product [26].
Tobramycin is compatible with most available IV fluids, but is not

compatible with heparin solution. In addition, it can interact chemically
with p-lactam compounds of the penicillin, cephalosporin, and cephamycin
family [ 3 ] . This interaction depends upon the concentration and pH of both
tobramycin and f3-lactam compounds. Solutions of tobramycin sulfate and
clindamycin phosphate have been reported to be unstable in dextrose
injection [42].
The stability of tobramycin in 30 and 50% dextrose peritoneal dialysate

concentrate (PDC) fluids have been reported [43]. This study indicated that
if tobramycin is to be added to PDC fluids containing 50% wlv of dextrose,
it should be used within 12 hours of admixture.
7. Pharmacokinetics
7.1 Absorption
Tobramycin is not appreciably absorbed when taken orally, but does exert
an antibacterial effect in the intestine. When applied to the skin, the drug is
not absorbed to a degree sufficient to produce any therapeutic effects. There
is no significant absorption of the drug from the bronchi and lungs after
administration as an aerosol [44]. Studies in rabbits suggest that tobramycin
is absorbed into the aqueous humor following topical installation of 3
mg/mL solution of the drug onto the eye and absorption is greatest when the
cornea is abraded [45].
Owing to these adsorption characteristics, tobramycin should be

administered either intramuscularly (IM) or intravenously (IV). Absorption
after IM injection is rapid, with the peak serum concentration being
achieved at 20-45 minutes. Senun concentrations of tobramycin following
a single IM injection are given in Table 6 [46-511. Mean peak serum
concentrations of tobramycin following various rates of IV injections are
given in Table 7 [49,50,52,53].
TOBRAMYCIN 603
Table 6
Mean Peak Serum Concentration of Tobramycin Following

Single Intramuscular Injections
Dose Serum concentrations References

(I.ldmL)
50 mg/m2 4.6 46
100 mg 5.1 47
2.5 mgkg 7.1 48
100 mg 3.8 49
100 mg 5.2 50
40 mg 2.4 51
80 mg 3.7 51
604 ALEKHA K. DASH
Table 7
Mean Peak Serum Concentration of Tobramycin

Following Intravenous Injections
Rate of Injection Dose Peak serum References

levels (yg/mL)
1 hour infusion 1 mgkg 4.4 52
30-45 min 1 mg/kg 5.5 49
30-45 min 1.5 mgkg 6.0 49
1 hour infusion 100 mg 4.6 50
bolus injection 80 mg 11.2 53

(2.5-3 min)
bolus injection 1 mgkg 10.0 53

(2.5-3 min)
TOBRAMYCIN 605
7.2 Distribution
The distribution of tobramycin in human tissue and body fluids is

summarized in Table 8 [54-601. The half-life of the drug in serum ranges
between 1.6 and 3.5 hours in normal individuals. Ragamey et al. have
reported the apparent volume of distribution (AVD) value for tobramycin to
be 24.5 liters [50]. However, Simon et al. have reported the AVD value to
be 16.9 liters [51].
7.3 Protein binding
Using equilibrium dialysis, Ramirez-Ronda et al. have reported that

approximately 70% tobramycin is bound to plasma proteins at a
concentration of 10 mg/mL or less [61]. However, Gorden et al. [62], and
Neber et a!. [63] have reported that under conditions of physiological pH
and temperature, no serum protein binding of tobramycin occurred at a
concentration of 5 mg/mL. A similar effect was also reported by Ullmann
et al. using steady state gel filtration and frontal analysis [64].
7.4 Excretion
Tobramycin is rapidly excreted unchanged in the urine after an IM or IV

injection [65]. However, Pechere and Dugal[66] have suggested that 10%
of the drug is eliminated by extrarenal mechanisms. The renal excretion of
tobramycin takes place entirely by glomerular filtration [46,50]. The total
plasma clearance of tobramycin from IV infusion studies have been reported
to be 113.7 mL/min/1.73 m2 [50] and 87.9 mL/min/1.73 m2 [51]. The rate
of recovery of tobramycin from urine over a 6 hour period is 60% and 80-
85% during the 24 hours after injection [67,68]. During the first 6 hours
after a dose of 1 mgkg (given by infusion over a period of 1 hour), urinary
concentrations between 70 and 300 mg/mL have been reported [52].
Table 8
Distribution of Tobramycin in Various Tissues and Body Fluids
Tissue or other Dose and Routes Time Concentration References

body fluids of administration elapsed detected (pg/mL)
Breast milk 80 mg by IM 1 hour 0.60 54

8 hours 0.85 55
Amniotic fluid 80 mg by IM 3 hours 1.2 56
Cerebrospinal fluid 3-4.5 mgkg by IV <I 57
sputum 5 m g k g IM injection 1 to 3 hours 0.3 58

in three divided doses
Bile 80 mg im 1 hour 1.4 59
Palatine tonsil 80 or 160 mg im 1 hour 0.5 to 1.3 60

samples
TOBRAMYCIN 607
Table 9
LD50 of Tobramycin in Rat and Mouse [69]
Animal species Route of LD50 (mg/kg)

Administration
Rat intraperitoned 1030

Rat subcutaneous 969
Rat intravenous 104
Rat intramuscular 913
Mouse intraperitoneal 445
Mouse subcutaneous 3 67
Mouse intravenous 72.5
Mouse intramuscular 440
Pig subcutaeous 676
ALEKHA K.DASH
8. Toxicity
The LD50 values determined for tobramycin in different animal species are
given in Table 9 [69]. The LD50 for tobramycin is found to be 50-66% of
that of gentamicin in mice, and 70% of that of rats [70,71]. Lethal and
sublethal doses of tobramycin produce hyperactivity, decreased respiration
and prostration in mice, rats and guinea pigs. No noticeable changes in
behavior, appearance, or in hematological or biochemical values were
noticed in dogs when 3.75 and 7.5 mgkg of tobramycin is administered for
90 days. There was evidence of slight cloudy swelling of proximal portions
of nephrons after 30 mgkg dose of the drug for 90 days.
9. Acknowledgments
The useful suggestions of Professors. Raj Suryanarayanan (University of

Minnesota, College of Pharmacy) and Shanker L. Saha (Creighton
University) are gratefully acknowledged.
10. References
1. C. E Higgens and R. E. Kastner, Antimicrob. Agents Chemother.,

1967, 324-331 (1971).
2. R. Q. Thompson and E. A. Presti, Antimicrob. Agents Chemother.,

1967,332-340 (1971).
3. H. C. Neu, J. Infect. Diseuse., 134 (Suppl.), S3-Sl9 (1976).
4. The Merck Index, S. Budavari, ed., 1l* Edition, Merck and Co.,
Inc., Kahway, N.J., USA, 1989, p. 1499.
5. K. F. Koch and J. A. Rhoades, Antimicrob. Agents. Chemother.,

1970,309-313 (1971).
6. Y. Takegi,T. Miyake, T. Tsuchiya, S. Umezaya and H. Umezaya, J

Antibiotics., 26,403-406 (1973).
TOBRAMYCIN 609
7. A. K. Dash, unpublished data.
8. K.F Koch, J. A. Rhoades, E. W. Hagaman and E. Wenkert, J. Am.

Chem. Soc., 96,3300-3305 (1974).
9. Clarke's Isolation and Identification of Drugs in Pharmaceutical,

Body Fluids and Post-mortem Material, A. C. Moffat, ed., 2nd
Edition., The Pharmaceutical Press, London, 1988, p. 1027.
10. T. Mills I11 and J. C. Roberson, Instrumental Data for Drug

Analysis, 2"d Edition, Vol3, Elsevier, NY, 1987, pp. 2256-2257.
11. A. K. Dash and R. Suryanarayanan, Pharm. Res., 8, 1159-1165

(1991).
12. Martindale, The Extra Pharmacopoeia, J. E. F. Reynolds, ed., 29*

Edition,The Pharmaceutical Press, London, England, 1989, pp. 326-
328.
13. The United States Pharmacopeia, 22ndRevision, United States

Pharmacopeial Convention ,Rockville, MD. 1989, p. 1381.
14. A. K. Dash., Ph. D. Thesis, University of Minnesota, 1990, p. 183.
15. A. Albert and E. P. Serjeant, The Determination of Ionization

Constants, 3rdEdition, Chapman and Hall, NY. 1984, p 174.
16. G. G. Raymond and J. L. Born, Drug Intel. Clin. Pharm., 20,683-

686 (1986).
17. V. Das Gupta,J Clin. Pharm. Therapy, 13, 195-198 (1988).
18. J. A. Ryan, J Pharm. Sci.,73,1301-1302 (1984).
19. A. Csiba, J Pharm. Pharmacol., 31,115-116 (1979).
20. S. Sampath and D. H. Robinson, J. Pharm. Sci., 79,428-431 (1990).
21. D. M. Barends, C. L. Zwaan and A. Hulschoof, J. Chromatogr., 225,

417-426 (1981).
610 ALEKHA K. DASH
22. P. M. Kabra, P.K. Bhatnagar, M. A. Nelson, J. H. Wall and L. J.

Marton, Clin. Chem., 29,672-674(1983).
23) N. A. Athanikar, R. W. Jurgens Jr., R. J. SturgeonL. A. Zober, P. P.

Deluca and D. D. Papadimitriou, J. Parenteral Sci. Tech.,B, 125-
128 (1 983).
24) S. K. Maitra, T. T. Yoshikawa, J. L. Hansen, M. C . Schotz, and L.

B. Guze, Am. J. Clin.Pathol., 71,428-432(I 979).
25) S.-E, Back, I. Nilsson-Ehle and P. Nilsson-Ehle, Clin.Chem.,

25,1222-1225 (1979).
A. K. Dash and R. Suryanarayanan, J. Pharm. Biomed. Anal., 9,

237-245(1991).
J. P. Anhalt and S. D. Brown, Clin.Chem., 24, 1940-1947(1978).
E. E. Stobberingh, A. W. Houben and C. P. A. van Boven, J. Clin.

Microb., 15,795-801(1982).
29) J. W. Mayhew and S. L. Gorbach, J. Chrornatogr., 151, 133-146

(1978).
S. E. Edberg and A. Chu, Am. J. Med. Technol., 41,99-105(1975).
D. C. Shanson, C. J. Hince, and J. V. Daniels, J. Infect. Disease,l34

(Suppl), S104-SlO9(1976).
32) K. Jones, J. Reynolds,T. 0. Oei and J. W. Smith, Clin. Chem., 25,

1096-1098(1979).
3 3) A. Broughton, J. E. Strong, L. K. Pickering and G. P. Bodey,

Antimicrob. Agents Chemother., 10,652-656(1 976).
34) E. L,. Frank, S. Srinivasan, P. Labthavikul and H. C. Neu, J. Clin.

Microb.. 13,93-96(1981).
TOBRAMYCIN 611
35) R. P., McKnight and J. E. Willis, Clin. Chem., 27, 1256-1261

(1981).
36) A. S. Lewis, G. Taylor, H. N. Rowe, H. 0. Williams and M. H.

Lewis, Am. J. Hosp. Pharm., 44,568-571 (1987).
37) TDX Fluorescence Polarization Immunoassay System; Abbott

Laboratories, Diagnostics Divison, North Chicago, IL 60064.
3 8) I. A. Cohen, J. M. DeKeyser and D. M. Hyder, Am. J. Hosp. Pharm.,

42,605-609 (1985).
39) Y. G. Tsay and R. J. Palmer, Clin Chem. Actu, 109, 151-157 (1981).
40) H. T. Kames, J. C. Gudat, C. M. ODonnell and J. D. Wineforder,

Clin. Chem., 27,249-252 (1981).
41) M. Sheehan and G. Caron, Ther. Drug. Monit., 7,108-1 14 (1985).
42) D. A. Marble, J. A. Bosso and R. J. Townsend, Drug Int. Clin.

Pharm., 20,960-963 (1986).
43) K. S. Nance and G. R. Matzke, Am. J Nephrol., 4,240-243 (1984).
44) G. E. Hoff, P. 0. Schiotz and J. Paulsen, Scand. J Infict. Dis., 6 ,

333-337 (1974).
45). AHFS Drug Information 93, G. K. McEvoy ed., American Society

of Hospital Pharmacists, MD, 1993, pp. 1695-1696.
46) N. Horikoshi, M. Valdivieso and G. P. Bodey, Am. J Med. Sci., 266,

453-458 (1973).
47) W. R. Lockwood and J. D. Bower, Antimicrob. Agents Chemother.,

3,215-219 (1973).
J. Klastersky, D. Daneau and V. DeMaereaer, Clin. Pharmacol.

Ther., 14, 104-1 11 (1973).
612 ALEKHA K . DASH
49) B. R. Meyers and S. Z. Hirschman, J. Clin. Pharmacol. New Drugs,

12,32 1-324 (1972).
50) C. Regamey, R. C. Gordon and W. M. W. Kirby, Clin. Pharmacol.

Ther., 4, 396-403 (1973).
51j V. K. Simon, E. U. Mosinger and V. Malerczy, Antimicrob. Agents

Chemotherapy, 3,444-450 (1973).
H. Lode, P. Kernmerich and P. Koeppe, Antimicrob. Agents

Chemotherapy, 8, 396-401 (1 975).
53) B. C. Startford and S. Dixson, Lancet, 1, 378-379 (1974).
54) Z. Takase, H. Shirauji, M. Uchida and T Sugawa, Chemotherapy,

23, 1378- (1 975).
55) M. Uwaydah, S,. Bibi and S. Salman, J. Antimicrob. Chemother., 1,

429-437 (1975).
56) R. Motomura, N. Nakajima and S. Miura., Chemotherapy, 23, 1403-

1408 (1 975).
57) A. B. Kaiser and Z. A. McGee, Clin. Res., 22,33A (1974).
58) H. B. Hawley, R. Lewis, D. R. Swartz and D. W. Gump, Curr. Ther.

Res., 16,414-423 (1974).
59) K. Konomi,Y. Yamamoto, K. Egashira, Y. Kawauchi, T. Furusawa

and T. Hisatsugu, Chemotherapy, 23,239 (1975).
T. Iwasawa, Chemotherapy, 23, 1 168 (1 975).
C. H. Ramirez-Ronda, K. Holmes and J. P. Sandford, J. Clin.

Invest., 53,63A (1974).
62j R. C. Gordon, C. Regamey and W. M. Kirby, Antimicrob. Agents

Chemotherapy, 2 , 2 14-2 16 ( 1 972).
TOBRAMYCIN 613
63) K. G. Naber, S. R. Westenfelder and P. 0. Madsen, Antimicrob.

Agents Chemotherapy,3,469-473 (1973).
U. Ulmann, J. Antimicrob. Chemotherapy,2,213-214 (1976).
H. Ymada, T. Ichihashi and H. Kinoshita, Chemotherapy, 23,900-

907 (1975).
J.-C. Pechere and R. Dugal, J. Infect, Dis., 134 (Suppl.), S118-S124

(1976).
67) J. R. Black and R. S. Griffith, “Comparative pharmacology of

tobramycin and gentamycin in adult volunteers”, in Tobramycin:
selected proceedings from the Eighth International Congress of
Chemotherapy, Athens, 1973, Excerpta Media, Amsterdam, 1974,
pp. 24-30.
G. Jaffe, B. R. Meyers and S . Z. Hirshman, Antimicrob. Agents

Chemotherapy, 5,6 11-616, 1974.
Material Safety Data Sheet for tobramycin, USP reference standard,

United States Pharmacopeial Convention ,Rockville, MD. 1988, p.
3.
70) F. DeRosa, U. Buoncristiani, P. Capitanucci and R. F. Frongillo, J.

Int. Med. Res., 2, 100-114 (1974).
71) J. S. Welles, J. L. Emmeson, W. R. Gibson, R. Nickander, N. V.

Owen and R. C. Anderson, Toxicol.Appl. Pharmacol., 25,398-409
(1973).
CUMULATIVE INDEX
Bold numerals refer to volume numbers.
Acebutolol, 19, Bromcriptine methanesulfonate, 8, 47

Acetaminophen, 3, 1; 14, 551 Bumetanide, 22, 107
Acetazolamide, 22, 1 Bupivacaine, 19, 59
Acetohexamide, 1, 1; 2, 573; 21, 1 Busulphan, 16, 53
Allopurinol, 7, 1 Caffeine, 15, 71
Amantadine, 12, 1 Calcitriol, 8, 83
Amikacin sulfate, 12, 37 Camphor, 13, 27
Amiloride hydrochloride, 15, 1 Captopril, 11, 79
Aminobenzoic acid, 22, 33 Carbamazepine, 9, 87
Aminoglutethimide, 15, 35 Carbenoxolone sodium, 24, 1
Aminophylline, 11, 1 Cefaclor, 9, 107
Aminosalicylic acid, 10, 1 Cefamandole nafate, 9, 125; 10, 729
Amiodarone, 20, 1 Cefazolin, 4, 1
Amitriptyline hydrochloride, 3, 127 Cefotaxime, 11, 139
Amobarbital, 19, 27 Cefoxitin sodium, 11, 169
Amoddiaquine hydrochloride, 21, 43 Ceftazidime, 19, 95
Amoxicillin, 7, 19; 23, 1 Cefuroxime sodium, 20, 209
AmphotericinB, 6, 1; 7, 502 Celiprolol hydrochloride, 20, 237
Ampicillin, 2, 1; 4, 518 Cephalexin, 4, 21
Apomorphine hydrochloride, 20, 121 Cephalothin sodium, 1, 3 19
Ascorbic acid, 11, 45 Cephradine, 5, 21
Aspirin, 8, 1 Chloral hydrate, 2, 85
Astemizole, 20, 173 Chlorambucil, 16, 85
Atenolol, 13, 1 Chloramphenicol, 4, 47; 15, 701
Atropine, 14, 325 Chlordiazepoxide, 1, 15
Azathioprine, 10, 29 Chlordiazepoxide hydrochloride, 1,39
Azintamide, 18, 1 Chloropheniramine maleate, 7, 43
Aztreonam, 17, 1 Chloroquine, 13, 95
Bacitracin, 9, 1 Chloroquine phosphate, 5, 61
Baclofen, 14, 527 Chlorothiazide, 18, 33
Bendroflumethiazide, 5, 1; 6, 597 Chlorprothixene, 2, 63
Benperidol, 14, 245 Chlortetracyclinehydrochloride, 8, 101
Benzocaine, 12, 73 Chlorthalidone, 14, 1
Benzyl benzoate, 10, 55 Chlorzoxazone, 16, 119
Betamethasone diproprionate, 6, 43 Cholecalciferol, 13, 655
Bretylium tosylate, 9, 71 Cimetidine, 13, 127; 17, 797
Bromazepam, 16, 1 Cisplatin, 14, 77; 15, 796
h16 CUMULATIVE INDEX
Clarithromycin, 24, 45 Diphenoxylate hydrochloride, 7, 149

Clidinium bromide, 2, 145 Dipivefrin hydrochloride, 22, 229
Clindamycin hydrochloride, 10, 75 Disopyramide phosphate, 13, 183
Clioquinol, 18, 57 Disulfiram, 4, 168
Clofazamine, 18, 91 Dobutamine hydrochloride, 8, 139
Clofazimine, 21, 75 Dopamine hydrochloride, 11, 257
Clonazepam, 6, 61 Doxorubicine, 9, 245
Clonfibrate, 11, 197 Droperidol, 7, 171
Clonidine hydrochloride, 21, 109 Echothiophate iodide, 3, 233
Ciorazepate dipotassium, 4, 91 Econazole nitrate, 23, 127
Clotrimazole, 11, 225 Emetine hydrochloride, 10, 289
Cloxacillin sodium. 4, 113 Enalapril maleate, 16, 207
Clozapine, 22, 145 Ephedrine hydrochloride, 15, 233
Cocaine hydrochloride, 15, 151 Epinephrine, 7, 193
Codeine phosphate, 10, 93 Ergonovine maleate, 11, 273
Colchicine, 10, 139 Ergotamine tartrate, 6, 1 13
Crospovidone, 24, 87 Erthromycin, 8, 159
Cyanocobalamin, 10, 183 Erthromycin estolate, 1, 101; 2, 573
Cyclandelate, 21, 149 Estradiol, 15, 283
Cyclizine, 6, 83; 7, 502 Estradiol valerate, 4, 192
Cyclobenzaprine hydrochloride, 17, 4 1 Estrone, 12, 135
Cycloserine, 1, 53; 18, 567 Ethambutol hydrochloride, 7, 23 1
Cyclosporine, 16, 145 Ethynodiol diacetate, 3, 253
Cyclothiazide, 1, 65 Etomidate, 12, 191
Cypropheptadine, 9, 155 Etopside, 18, 121
Dapsone, 5, 87 Fenoprofen calcium, 6, 161
Dexamethasone, 2, 163; 4, 519 Flecainide, 21, 169
Diatrizoic acid, 4, 137; 5, 556 Flucytosine, 5, 115
Diazepam, 1, 79; 4, 518 Fludrocortisone acetate, 3, 281
Dibenzepin hydrochloride, 9, 181 Flufenamic acid, 11, 313
Dibucaine, 12, 105 Fluorouracil, 2, 221; 18, 599
Dibucaine hydrochloride, 12, 105 Fluoxetine, 19, 193
Diclofenac sodium, 19, 123 Fluoxymesterone, 7, 25 1
Didanosine, 22, 185 Fluphenazine decanoate, 9,275; 10,730
Diethylstifbestrol, 19, 145 Fluphenazine enanthate, 2,245; 4, 524
Diflunisal, 14, 491 Fluphenazine hydrochloride, 2, 263; 4, 5 19
Digitoxin, 3, 149; 9, 207 Flurazepam hydrochloride, 3. 307
Dihydroergotoxine methanesulfonate, 7, 8 1 Fluvoxamine maleate, 24, 165
Diltiazern hydrochloride, 23, 53 Folic acid, 19, 221
Dioctyl sodium sulfosuccinate, 2, 199 Furosemide, 18, 153
Diosgenin, 23, 101 Gadoteridol, 24, 209
Diperodon, 6, 99 Gentamicin sulfate, 9, 295; 10, 73 1
Diphenhydramine hydrochloride, 3, 173 Glafenine, 21, 197
CUMULATIVE INDEX 617
Glibenclamide, 10, 337 Lactose, anhydrous, 20, 369

Gluthethimide, 5, 139 Leucovorin calcium, 8, 3 15
Gramicidin, 8, 179 Levallorphan tartrate, 2, 339
Griseofulvin, 8, 219; 9, 583 Levarterenol bitartrate, 1, 149; 2,573; 11, 555
Guanabenz acetate, 15, 3 19 Levodopa, 5, 189
Guargum, 24, 243 Levothyroxine sodium, 5, 225
Halcinonide, 8, 251 Lidocaine, 14, 207; 15, 761
Haloperidol, 9, 341 Lidocaine hydrochloride, 14, 207; 15, 761
Halothane, 1, 119; 2, 573; 14, 597 Lincomycin, 23, 275
Heparin sodium, 12, 2 15 Lisinopril, 21, 233
Heroin, 10, 357 Lithium carbonate, 15, 367
Hexestrol, 11, 347 Lobeline hydrochloride, 19, 261
Hexetidine, 7, 277 Lomefloxacin, 23, 327
Homatropine hydrobromide, 16, 245 Lomustine, 19, 315
Hydralazine hydrochloride, 8, 283 Loperamide hydrochloride, 19, 341
Hydrochlorothiazide, 10, 405 Lorazepam, 9, 397
Hydrocortisone, 12, 277 Lovastatin, 21, 277
Hydroflumethaizide, 7, 297 Mafenide acetate, 24, 277
Hydroxyprogesteronecaproate, 4, 209 Maltodextrin, 24, 307
Hydroxyzine dihydrochloride, 7, 3 19 Maprotiline hydrochloride, 15, 393
Hyoscyamine, 23, 155 Mebendazole, 16, 291
Imipramine hydrochloride, 14, 37 Mefloquine hydrochloride, 14, 157
Impenem, 17, 73 Melphalan, 13, 265
Indapamide, 23, 233 Meperidine hydrochloride, 1, 175
Indomethacin, 13, 21 1 Meprobamate, 1, 207; 4, 520; 11, 587
Iodamide, 15, 337 Mercaptopurine, 7, 343
Iodipamide, 2, 333 Mestranol, 11, 375
Iodoxamic acid, 20, 303 Methadone hydrochloride, 3,365; 4, 520; 9, 601
Iopamidol, 17, 115 Methaqualone, 4, 245
Iopanoic acid, 14, 181 Methimazole, 8, 351
Iproniazid phosphate, 20, 337 Methixen hydrochloride, 22, 3 17
Isocarboxazid, 2, 295 Methocarbamol, 23, 377
Isoniazide, 6, 183 Methotrexate, 5, 283
Isopropamide, 2, 315; 12, 721 Methoxamine hydrochloride, 20, 399
Isoproterenol, 14, 391 Methoxsalen, 9, 427
Isosorbide dinitrate, 4, 225; 5, 556 Methylclothiazide, 5, 307
Ivermectin, 17, 155 Methylphenidatehydrochloride, 10, 473
Kanamycin sulfate, 6, 259 Methyprylon, 2, 363
Ketamine, 6, 297 Metipranolol, 19, 367
Ketoprofen, 10, 443 Metoclopramide hydrochloride, 16, 327
Ketotifen, 13, 239 Metoprolol tartrate, 12, 325
Khellin, 9, 371 Metronidazole, 5, 327
Lactic acid, 22, 263 Mexiletine hydrochloride, 20, 433
618 CUMULATIVE INDEX
Minocycline, 6, 323 Phenobarbital, 7, 359

Minoxidil, 17, 185 Phenolphthalein, 20, 627
Mitomycin C, 16, 361 Phenoxymethyl penicillin potassium, 1, 249
Mitoxanthrone hydrochloride, 17, 22 1 Phenylbutazone, 11, 483
Morphine, 17, 259 Phenylephrine hydrochloride, 3, 483
Moxalactam disodium, 13, 305 Phenylpropanolamine hydrochloride, 12,357
Nabilone, 10, 499 Phenytoin, 13, 417
Nadolol, 9, 455; 10, 732 Physostigmine salicylate, 18, 289
Naiidixic acid, 8, 371 Phytonadione, 17, 449
Nalmefene hydrochloride, 24, 35 1 Pilocarpine, 12, 385
Nalorphine hydrobromide, 18, 195 Piperazine estrone sulfate, 5, 375
Naloxone hydrochloride, 14, 453 Pirenzepine dihydrochloride, 16, 445
Naphazoline hydrochloride, 21, 307 Piroxicam, 15, 509
Naproxen, 21, 345 Polythiazide, 20, 665
Natamycin, 10, 513; 23, 405 Polyvinyl alcohol, 24, 397
Neomycin, 8, 399 Polyvinylpyrollidone, 22, 555
Neostigmine, 16, 403 Povidone, 22, 555
Nicotinamide, 20, 475 Pralidoxine chloride, 17, 533
Nifedipine, 18, 221 Prazosin hydrochloride, 18, 35 1
Nitrazepam, 9, 487 Prednisolone, 21, 415
Nitrofurantoin, 5, 345 Primidone, 2, 409; 17, 749
Nitroglycerin, 9, 5 19 Probenecid, 10, 639
Nizatidine, 19, 397 Procainamide hydrochloride, 4, 333
Norethindrone, 4, 268 Procarbazine hydrochloride, 5, 403
Norfloxacin, 20, 557 Promethazine hydrochloride, 5, 429
Norgestrel, 4, 294 Proparacaine hydrochloride, 6, 423
Nortriptyline hydrochloride, 1,233; 2, 573 Propiomazine hydrochloride, 2, 439
Noscapine, 11, 407 Propoxyphene hydrochloride, 1,301; 4,520
Nystatin, 6, 341 Propylthiouracil, 6, 457
Oxamniquine, 20, 60 1 Pseudoephedrine hydrochloride, 8, 489
Oxazepam, 3, 441 Pyrazinamide, 12, 433
Oxyphenbutazone, 13, 333 Pyridoxine hydrochloride, 13, 447
Oxytocin, 10, 563 Pyrimethamine, 12, 463
Papaverine hydrochloride, 17, 367 Quinidine sulfate, 12, 483
Penicillamine, 10, 601 Quinine hydrochloride, 12, 547
Penicillin-G, benzothine, 11, 463 Ranitidine, 15, 533
Penicillin-G, potassium, 15, 427 Reserpine, 4, 384; 5, 557; 13, 737
Penicillin-V, 1, 249; 17, 677 Riboflavin, 19, 429
Pentazocine. 13, 361 Rifampin, 5, 467
Pergolide Mesylate, 21, 375 Rutin, 12, 623
Phenazopyridine hydrochloride, 3, 465 Saccharin, 13, 487
Phenelzine sulfate, 2, 383 Salbutamol, 10, 665
Phenformin hydrochloride, 4,3 19; 5, 429 Salicylamide, 13, 52 1
CUMULATIVE INDEX 619
Salicylic acid, 23, 427 Thioridazine hydrochloride, 18, 459

Scopolamine hydrobromide, 19, 477 Thiostrepton, 7, 423
Secobarbital sodium, 1, 343 Thiothixene, 18, 527
Sertraline hydrochloride, 24, 443 Ticlopidine hydrochloride, 21, 573
Silver sulfadiazine, 13, 553 Timolol maleate, 16, 641
Simvastatin, 22, 359 Titanium dioxide, 21, 659
Sodium nitroprusside, 6,487; 15,781 Tobramycin, 24, 579
Solasodine, 24, 487 a-Tocopheryl acetate, 3, 111
Sotalol, 21, 501 Tolazamide, 22, 489
Spironolactone, 4, 431; 18, 641 Tolbutamide, 3, 513; 5, 557; 13, 719
Starch, 24, 523 Tolnaftate, 23, 549
Streptomycin, 16, 507 Trazodone hydrochloride, 16, 693
Strychnine, 15, 563 Triamcinolone, 1,367; 2, 571; 4, 521; 11, 593
Succinycholine chloride, 10, 691 Triamcinolone acetonide, 1, 397; 2, 571; 4, 521;
Sulfacetamide, 23, 477 7, 501; 11,615
Sulfadiazine, 11, 523 Triamcinolone diacetate, 1, 423; 11, 651
Sulfadoxine, 17, 571 Triamcinolone hexacetonide, 6, 579
Sulfamethazine, 7, 401 Triamterene, 23, 579
Sulfamethoxazole, 2, 467; 4, 521 Triclobisonium chloride, 2, 507
Sulfasalazine, 5, 5 15 Trifluoperazinehydrochloride, 9, 543
Sulfathiazole, 22, 389 Triflupromazinehydrochloride, 2,523; 4, 521; 5, 557
Sulfisoxazole, 2, 487 Trimethaphan camsylate, 3, 545
Sulfoxone sodium, 19, 553 Trimethobenzamide hydrochloride, 2, 55 1
Sulindac, 13, 573 Trimethoprim, 7, 445
Sulphamerazine, 6, 5 15 Trimipramine maleate, 12, 683
Sulpiride, 17, 607 Trioxsalen, 10, 705
Talc, 23, 517 Tripelennamine hydrochloride, 14, 107
Teniposide, 19, 575 Triprolidine hydrochloride, 8, 509
Tenoxicam, 22, 431 Tropicamide, 3, 565
Terazosin, 20, 693 Tubocurarine chloride, 7, 477
Terbutaline sulfate, 19, 60 1 Tybamate, 4, 494
Terfenadine, 19, 627 Valproate sodium, 8, 529
Terpin hydrate, 14, 273 Valproic acid, 8, 529
Testolactone, 5, 533 VerapamiI, 17, 643
Testosterone enanthate, 4, 452 Vidarabine, 15, 647
Tetracaine hydrochloride, 18, 379 Vinblastine sulfate, 1, 443; 21, 61 1
Tetracycline hydrochloride, 13, 597 Vincristine sulfate, 1, 463; 22, 517
Theophylline, 4, 466 VitaminD3, 13, 655
Thiabendazole, 16, 61 1 Warfarin, 14, 423
Thiamine hydrochloride, 18, 4 13 Xylometazoline hydrochloride, 14, 135
Thiamphenicol, 22, 461 Yohimbine, 16, 731
Thiopental sodium, 21, 535 Zidovudine, 20, 729
Thioridazine, 18, 459 Zomepirac sodium, 15, 673
I S B N 0-12-260824-0

Analytical Profiles of Drug Substances and Excipients, Vol 24

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Analytical Profiles of Drug Substances and Excipients, Vol 24

Încărcat de

Drepturi de autor:

Formate disponibile

Analytical Profiles

Abdullah A. Al-Badr Larry D. Kissinger

Alekha K. Dash David J. Mauo

Klaus Florey Christopher T. Riley

Lee T. Grady Timothy J. Wozniak

Ohmeda Pharmaceutical Products Division, Inc.

This book is printed on acid-free paper. @

All Rights Reserved.

Academic Press, Inc.

United Kingdom Edition published by

International Standard Serial Number: 1075-6280

International Standard Book Number: 0-12-260824-0

PRINTED RJ THE UNITED STATES OF AMERICA

AfJiriations of Ediforsand Coaiributors vii

4. Fluvoxamine Maleate 165

6. Guar Gum 243

7. Mafenide Acetate 277

10. Polyvinyl Alcohol 397

11. Sertraline hydrochloride 443

12. Solasodine 487

13. Starch 523

14. Tobramycin 579

Cumulative Index 615

Christiunuh M Adeyeye, Department of Pharmacy, Duquesne University,

Gunawan Indrayanto, Laboratory of Pharmaceutical Biotechnology,

Christine Rodriguez, Bristol-Myers Squibb Pharmaceutical Research

Perceptive readers will note that 1995 passed without publication of an

As always, a complete list of available drug and excipient candidates is

Silvia Pindado"', Owen I. C~rrigan'~)

and Caitriona M. O'Driscoll*(2)

(1) National Pharmaceutical Biotechnology Center

(2) University of Dublin

* Author for correspondence

ANALYTICAL PROFILES OF DRUG SUBSTANCES 1 Copyright 0 1996 by Academic Press, Inc.

Carbenoxolone is a triterpenoid, the ester of 18p-glycyrrhetic (enoxolone)

2.1.1 Chemical Names

Disodium salt of 3~-(3-carboxypropionyloxy)-11-oxo-olean-12-en-

3~-Hydroxy-ll-oxoolean-12-en-30-oic acid hydrogen succinate.

3P-Hydroxy- 1 1-0xoolean-12-en-30-oic acid 3-hemisuccinate.

3-O-(f3-carboxypropionyl)-1 1-oxo- 18p-olean-12-en-30-oic acid

2-Carboxy-ethylpropionyl glycyrrhetinic acid (British patent,

2.1.2 Nonproprietary Names

A. Carbenoxolone Sodium, Disodium Enoxolone Succinate (BANM,

B. Carbenoxolone, Glycyrrhetinic Acid Hydrogen Succinate,

2.1.3 Proprietary Names

Biogastrone, Bioplex, Bioral, Carbosan, Duogastrone, Gastrausil,

2.1.4 Chemical Abstracts Service (CAS) Registry Numbers

7421-40-1 Carbenoxolone Sodium

Carbenoxolone Sodium Carbenoxolone

2.3 Molecular Weight

614.7 Carbenoxolone Sodium

Carbenoxolone sodium is a white or pale-cream colored, hygroscopic

2.5 OMicial Compendia

A monograph on Carbenoxolone Sodium is included in the British

2.6 Other Compendia

Carbenoxolone sodium is included in the Pharmaceutical Codex (1979),

Carbenoxolone is synthesized from glycyrrhetinic acid, the aglycone of

Glycyrrhetinic Acid > Carbenoxolone

Figure 1. Synthesis of carbenoxolone.

4.1.1 Ultraviolet Spectroscopy

The ultraviolet spectrum of carbenoxolone sodium (0.005% w/v) is shown

250 300 350

Figure 2. Ultraviolet spectrum of carbenoxolone sodium.

4.1.2 Infrared Spectroscopy