Journal of Dental Research

http://jdr.sagepub.com/

Interfacial Chemistry of the Dentin/Adhesive Bond

P. Spencer, Y. Wang, M.P. Walker, D.M. Wieliczka and J.R. Swafford
J DENT RES 2000 79: 1458
DOI: 10.1177/00220345000790070501

The online version of this article can be found at:

http://jdr.sagepub.com/content/79/7/1458

Published by:

http://www.sagepublications.com

On behalf of:
International and American Associations for Dental Research

Additional services and information for Journal of Dental Research can be found at:

Email Alerts: http://jdr.sagepub.com/cgi/alerts

Subscriptions: http://jdr.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations: http://jdr.sagepub.com/content/79/7/1458.refs.html

>> Version of Record - Jul 1, 2000

What is This?

Downloaded from jdr.sagepub.com at PENNSYLVANIA STATE UNIV on March 2, 2014 For personal use only. No other uses without permission.
P. Spencerl ,2*, Y. Wang', M.P. Walkerl,
D.M. Wieliczka2, and J.R. Swafford1
'Department ofOml Biology, 2Department ofPediatric Dentistry,
University ofMissouri-Kansas City School of Dentistry, 650 E. 25th
St., Kansas City, MO 64108; 3Departnent ofPhysics, University of
Missouri-Kansas City, Kansas City, MO; *corresponding author,
SpencerP@umkc.edu
J Dent Res 79(7):1458-1463, 2000
ABSTRACT
To date, the dentin/adhesive (d/a) bond has
primarily been studied by morphologic analysis in
conjunction with bond strength measurement.
Although these analyses have enhanced our
understanding, numerous questions about the
chemistry have not been answered. The purpose of
this study was to determine, at the molecular level,
quantitative differences in the composition of the
d/a interface formed under "wet" bonding
conditions. The occlusal one-third of the crown
was removed from 10 extracted, unerupted human
third molars. The prepared dentin surfaces were
treated, per manufacturers' instructions, with either
Single Bond (3M) or One-Step adhesive (Bisco).
Three-micron-thick sections of the d/a interface
were cut and stained with Goldner trichrome for
light microscopy. Companion slabs were analyzed
with micro-Raman spectroscopy; the sample was
placed at the focus of a 100x microscope
objective, and spectra were acquired at 1-ptm
intervals across the d/a interface. Reference
spectra were collected on model compounds of
type I collagen and adhesive; the relative ratios of
the integrated intensities of spectral features from
adhesive and collagen were determined and
plotted as a function of wt% adhesive. The same
ratios were determined for the interface samples;
by comparing these ratios with the calibration
curve generated from the model compounds, we
determined the percent of adhesive as a function
of spatial position across the d/a interface. The
relative percent of Single Bond adhesive was <
50% throughout more than half of the hybrid
layer; One Step adhesive was ' 50% throughout
most of the hybrid. The results from this study
provide the first direct chemical evidence of phase
separation in a dentin adhesive and its detrimental
effect on the dentin/adhesive bond.
KEY WORDS: dentin, adhesive, interface,
spectroscopy, Raman.
Received August 16, 1999; Last revision December 6,
1999; Accepted February 4, 2000
1458 Downloaded from jdr.sagepub.com at PENNSYLVANIA STATE UNIV on March 2, 2014 For personal use only. No other uses without permission.
Interfacial Chemistry of the
Dentin/Adhesive Bond
INTRODUCTION
Current dentin bonding theories suggest that there are two fundamental
processes involved in bonding an adhesive to dentin. First, the mineral
phase must be extracted from the dentin substrate without damaging the
collagen matrix, and second, the voids left by the mineral must be filled with
an adhesive resin that penetrates the exposed collagen fibril network. If the
exposed collagen collapses during the bonding procedure, the porosity of the
dentin substrate is reduced, and many of the sites available for resin
penetration are eliminated (Pashley et al., 1993; Gwinnett, 1994; Eick et al.,
1995). The dentin/adhesive bond is weakened because the collapsed collagen
forms a barrier that prevents resin penetration throughout the decalcified
layer and into the unaltered dentin substrate.
It has been reported that as long as the dentin is kept fully hydrated, the
surface morphology of the demineralized layer does not change, i.e., the
water supporting the collagen matrix is not lost, and the matrix network does
not shrink (Kinney et al., 1993). Results from previous studies with a "wet"
bonding technique support these findings (Kanca, 1992; Gwinnett, 1994); the
higher bond strengths with this technique reflect the minimal collapse of
"wet" dentin collagen vs. collagen that is dried with an air syringe (Pashley,
et al., 1993). It is speculated that moist dentin provides a more porous
collagen network and thus greater infiltration of adhesive monomers
(Pashley, et al., 1993; Tam and Pilliar, 1994; Gwinnett, 1994).
To date, questions regarding dentin adhesive bonding under "wet"
conditions have generally been investigated by means of bond strength
studies in combination with morphologic analyses. Although these
techniques allow us to compare products, they provide little information on
the chemistry of the dentin/adhesive interface. The chemistry of the
dentin/adhesive interface has been investigated by micro-Raman
spectroscopy (Suzuki et al., 1991; Wieliczka et al., 1996, 1997; Lemor et al.,
1999, 2000), but with this method there has been little effort to quantify the
distribution of adhesive throughout the zone of "wet" demineralized dentin.
The two-fold purpose of this study was to determine, at the molecular level,
the composition of the dentin adhesive/hybrid layer interface formed under
"wet" bonding conditions and to quantify the diffusion of these single-bottle
adhesives into the "wet" demineralized dentin. This study tested the null
hypothesis that there was no difference in adhesive diffusion into the
demineralized dentin with new adhesive systems that apply the "wet"
bonding technique.
MATERIALS & METHODS
Specimen Preparation
Extracted unerupted human third molars stored at 4°C in 0.9% w/v NaCl containing
0.002% sodium azide were used in this study. The teeth were collected after the
patients' informed consent was obtained under a protocol approved by the UMKC
adult health sciences institutional review board. The specimen preparation has been
J Dent Res 79(7) 2000 Dentin/Adhesive Interface Chemistry
presented in detail in a previous publication (Spencer and Swafford,
1999). In brief, the occlusal one-third of the crown was sectioned
perpendicular to the long axis of the tooth by means of a water-cooled
low-speed diamond saw (Buehler, Lake Bluff, IL, USA). We created a
uniform smear layer by abrading the exposed dentin surface with 600-
grit silicon carbide under water. The prepared dentin specimens were
randomly selected for treatment with either Single Bond (Lot No.
19970204, 3M Corp., St. Paul, MN, USA) or One Step (Lot No.
029128, Bisco, Itasca, IL, USA) dentin adhesive according to
manufacturers' instructions. The dentin adhesives were polymerized for
30 sec by exposure to a visible light source (Spectrum light, Dentsply,
Milford, DE, USA). Five teeth were used per adhesive system. These
specimens were stored for a minimum of 24 hrs in water at 37°C before
being sectioned.
These treated dentin surfaces were sectioned perpendicular and
parallel to the bonded surface by means of a water-cooled low-speed
diamond saw. The resultant rectangular (10 x 2 x 2 mm) slabs were
mounted on a methacrylate support with cyanoacrylate adhesive, and
three-micron-thick sections were cut from the face of the slab by means
of a tungsten carbide knife mounted on a Polycut S "sledge" microtome
(Leica, Deerfield, IL, USA). The sections were collected on glass
microscope slides treated with Haupt's adhesive. This adhesive, a
mixture of gelatin, glycerol, and phenol, is used to keep the sections
attached to the glass slide during the subsequent staining procedures.
Differential staining was accomplished with Goldner trichrome, a
classic bone stain (Goldner, 1938). Slides with adherent stained
sections were dehydrated through ascending ethanol and xylene. The
sections were cover-slipped with mounting media and examined under
a Zeiss light microscope. The width of the exposed protein layer was
determined by direct measurement from photomicrographs whose
exact magnification was established with a stage micrometer.
Separate but adjacent slabs were prepared for micro-Raman
spectroscopy. The rectangular 10 x 2 x 2 mm slabs were placed directly
on glass microscope slides for micro-Raman spectroscopic analysis. To
remove residual saw marks, we lightly polished the surface of the slab
by hand on microclothg impregnated with a water slurry of 1-jim
alumina particles (Buehler, Lake Bluff, IL, USA), followed by brief
sonication in distilled water. The surface was then rubbed with a cotton
swab saturated with 5% sodium hypochlorite to rid the surface of
smeared protein derived from the polishing.
Instrumentation for Micro-Raman Spectroscopy
Micro-Raman spectra were recorded with a Jasco NRS 2000 Raman
spectrometer equipped with Olympus lenses and a liquid-nitrogen-
cooled CCD detector. The excitation source was an Argon laser,
operating at 514.5 nm. After passing through the bandpass filter and
condensing optics, approximately 3 mW power was incident upon the
sample.
The sample was placed at the focus of a I00x microscope
objective, and spectra were acquired at positions corresponding to 1-
jim intervals across the d/a interface by use of the computer-controlled
x-y-z stage with a minimum step width of 50 nm. The micro-Raman
spectroscopic technique offers the investigator the unique opportunity
to both acquire an image of the interface and determine the molecular
structure. The focus of the laser beam in conjunction with a 25-jtm
confocal aperture provided a spatial resolution of 1 jm. Spectra were
obtained at a resolution of 6 cmnl over the spectral region of 875 to
1785 cm' and with an integration time of 120 sec. A digital image of
the d/a interface, with demarcations identifying the position of each
Downloaded from jdr.sagepub.com at PENNSYLVANIA STATE UNIV on March 2, 2014 For personal use only. No other uses without permission.
1459
spectrum, was recorded simultaneously. Multiple sites across the
interface of each specimen were examined spectroscopically, and the
comparison of these spectra confirmed the reproducibility of the
technique. Instrument fluctuation was evaluated daily by comparison of
spectra from standards such as silicon.
Quantifying Adhesive Diffusion
Spectral data from the d/a interfaces were compared with reference
spectra of pure adhesive, demineralized dentin, and mineralized dentin.
Spectra were also collected on a series of model compounds composed
of type I collagen and adhesive. The calibration curve generated from
these model compounds was used to determine the percent of adhesive
as a function of spatial position across the dentin interface.
The model compounds were made from 1% w/v solutions of type
I collagen and adhesive. Acid-soluble rat tail type I collagen (c8897,
Sigma Type VII, Sigma, St. Louis, MO, USA) was used in this project.
The concentrations of One Step and Single Bond adhesives were
calculated by the weight loss method. One percent w/v solutions of
adhesive in ethanol and of collagen in 0.5 M acetic acid were prepared.
We obtained the model compounds of collagen and adhesive by
blending the solutions in appropriate ratios. The compositions of these
collagen/adhesive compounds were 100/0, 80/20, 70/30, 60/40, 50/50,
40/60, 30/70, 20/80, and 0/100. The mixtures were polymerized for 30
sec by exposure to a visible light (Spectrum light, Dentsply, Milford,
DE, USA), washed with distilled water, and vacuum-dried. The
polymerized dried samples measured 3 x 2 mm. The model
compounds were made in triplicate.
Raman spectra were acquired at a resolution of 6 cm-' from a
minimum of 3 different sites on each sample made from the model
compounds. A comparison of the spectra that were collected from
the 3 different sites indicated complete overlap, suggesting that there
was a homogeneous mixture of collagen and adhesive throughout
these samples.
The relative ratios of the integrated intensities of a spectral feature
from the adhesive and the carbonyl (1666 cm-') from the amide I were
determined from the spectra of the samples made from the model
compounds. The spectral feature associated with Single Bond adhesive
was the C-O-C (1113 cm-'), whereas the C=C (1610 cm-') was used for
One Step. To generate the calibration curves, we plotted the ratios of
the integrated intensities of the spectral features from the adhesive and
collagen as a function of weight percent adhesive. A linear regression
analysis was used to determine the goodness of fit.
The relative ratios of the integrated intensities of the spectral
features described above, i.e., carbonyl from the amide I and the
respective groups from the adhesives, were determined as a function of
spatial position across the d/a interface. We compared these relative
ratios with the calibration curve generated from the model compounds
to determine the corresponding weight percent of adhesive. This
comparison provided a quantitative representation of the percent of
adhesive as a function of spatial position across the d/a interface.
RESULTS
Light Microscopy
The light micrographs of separate but adjacent sections from the
specimens that were analyzed by micro-Raman spectroscopy are
presented in Figs. IA and lB. In these Goldner-trichrome-stained
sections of the d/a interface, exposed protein stained red, mineral
appeared green, protein that was partially coated with either primer
1 460 Spencer et ai. J Dent Res 79(7) 2000
Table. Spectral Identification and Tentative Assignment for Adhesives (1667) are plotted as a function of the weight percent of One Step
adhesive. The adhesive features at 1113 and 1610 cm-' were selected
Product Raman Shift (cm-') Tentative Assignment for the calibration curves because they were not affected by the
degree of polymerization, and they were distinct from the spectral
Single Bond 11 13 (A) Phenyl, C-O-C features attributable to collagen. The correlation coefficients (r) were
185 (B) Gem-dimethyl (CH3-C-CH3) 0.97 and 0.98 for Single Bond and One Step, respectively.
1230 CH-OH stretch
1289 =CH2 rock, C-O stretch Micro-Raman Spectra of Dentin/Adhesive Interfaces
1404 CH-OH, CH2-CO- stretch Spectra acquired at 1-jm intervals across the dentin/Single Bond
1450 (C) CH CH deformation
1607 (D) De?ormation of phenyl (C=C) adhesive interface are shown in Fig. 4. Based on both visual and
1634 >C=C< spectroscopic evaluation, the first spectrum was acquired from pure
1720 >C=O adhesive. The relative increase in intensity of the 961 cmn' peak (P-
One Step 1081 C-H bending O symmetric stretch) in the eighth spectrum suggests that this
1115 Phenyl, C-O-C represents the floor of the demineralized dentin. The spectral
1190 gem-dimethyl (CH3-C-CH3) intensity of the 961 cm-' peak in the ninth spectrum indicates
1262 (A) CH-OH mineralized dentin. The ratios of the relative integrated intensities
1318 (B) =CH2 rock, C-O stretch of the phenyl group (C-O-C, 1113 cm'l) from the Single Bond
1457 CH CH3 deformation adhesive and the carbonyl (1667 cm'l) from the amide I of collagen
1610 (C) De?ormation of Phenyl (C=C) were calculated from spectra such as these (Fig. 4). We determined
1637 >C=C<
1720 (D) >C=O the percent of adhesive as a function of spatial position across the
Dentin 60 (A) P-O sym. stretch interface by comparing these ratios with the calibration curve
1244 (B), 1268 Amide III generated from the associated model compounds.
1428, 1451 (C) CH2 Spectra acquired at 1-jim intervals across the dentin/One
1635, 1666 (D) Amide Step adhesive interface are presented in Fig. 5. According to both
visual and spectral examinations, the first spectrum (Fig. 5) was
acquired from pure adhesive. The relative intensity of the P-O
or adhesive stained orange, and pure adhesive was beige (Spencer group (961 cm-') in the seventh spectrum suggests that this
and Swafford, 1999). The representative photomicrograph of a represents the floor of the demineralized dentin layer. Thus, the
dentin specimen treated with Single Bond adhesive shows a distinct second through seventh spectra represent the hybrid layer. The
red zone, 4-5 jm wide, at the d/a interface (Fig. la). The red
-
eighth and ninth spectra show a relatively intense peak at 961 cm'
represents protein at the interface that was available for reaction with peak (P-O symmetric stretch) and contribution from the adhesive
the stain. The width of the zone that showed a color distinct from at 1610 cm41. The ratios of the relative integrated intensities of
either pure Single Bond adhesive or mineralized dentin, i.e., the the C=C (1610 cm-') from the One Step adhesive and the
entire red/orange zone, was - 6 jm. These results suggest that the carbonyl (1667 cm-') from the amide I were calculated from
1-2-gm-wide orange layer represents demineralized dentin that was spectra such as these (Fig. 5). To determine the percent of
protected by the diffusion of Single Bond adhesive into this area. adhesive as a function of spatial position across the interface, we
The representative photomicrograph of a Goldner-trichrome- compared these ratios with the calibration curve generated from
stained section from the specimen treated with One Step adhesive the model compounds of One Step adhesive and collagen.
shows a very narrow red line, approximately 0.5 jm wide, at the A comparison of percent of adhesive with both materials as a
d/a interface (Fig. lb). The entire zone that showed a color distinct function of position across the d/a interface is presented in Fig. 6.
from either pure adhesive or mineralized dentin was 4.6 jm. -
The second position represents the adhesive/hybrid layer interface;
These results suggest that One Step adhesive penetrated nearly 90% this position corresponds to the second spectrum in Figs. 4 and 5.
of the collagen network, i.e., adhesive diffused through 4 jim of The percent of Single Bond adhesive penetrating the demineralized
the 4.6-jim demineralized dentin layer. The color differences, i.e., dentin drops from 70% to - 20% in the hybrid layer. These results
-

the wide red area on the left and a smaller area on the right of the
micrograph, suggest inconsistent adhesive penetration along the
length of the interface.
Raman Spectroscopy-Reference Spectra
The Raman spectra for Single Bond and One Step adhesives as
well as dentin are presented in Figs. 2a, 2b, and 2c, respectively.
The tentative assignment of the spectral features for both adhesives
and dentin is presented in the Table. Vibrations of both the
inorganic and organic components are observed in the dentin
spectrum. The most intense peak at 961 cm-' (P-O symmetric
stretch) is associated with the inorganic component.
The calibration curves for the model compounds of adhesive
and collagen are presented in Fig. 3. The peak area ratios for C-O-C
(11 13)/C=O (1667) vs. the weight percent of Single Bond adhesive
are represented. Similarly, the peak area ratios for C=C (1610)/C=O
Downloaded from jdr.sagepub.com at PENNSYLVANIA STATE UNIV on March 2, 2014 For personal use only. No other uses without permission.
indicate that Single Bond adhesive penetration is < 50% throughout
more than half of the hybrid layer. In comparison, One Step
adhesive remains above 50% throughout most of the hybrid layer.
DISCUSSION
In the present study, using the intrinsic vibrational signatures
of the constituents at the interface, we generated chemical maps of
the hybrid layer at 1 jm spatial resolution. Interestingly, although
no efforts were taken in this study to remove the organic
component, we did not experience spectral interference from
fluorescence. This may be attributable in part to the improved
confocal signal detection and short integration time. We quantified
adhesive diffusion by comparing spectral features from the
chemical maps of the hybrid layer with calibration curves generated
from model compounds of adhesive and type I collagen. A similar
J Dent Res 79(7) 2000 Dentin/Adhesive Interface Chemistry1
tchnliC Lie IMas hCCel uscd to CILanItiiy the spatial disitjbutionl o01xwatcr
ill Ilic Cot11n. (MI3 ntl I(a!1 998) and cliol estelol ill cor-olnl iy arterN
x IIi, I Rotiset
II ol 1 998).
-Ilc m ic0o-Ranian spCCtIi0sCop ic ICSLilts indicated strikin-
dii l'c-tences in adhsesive penctration witli the commiiercial produLcts
1ix1sti0cate1C ill tIlis StLICd I cxor eXamieC. at approximateily 2 imiiclonis
itito the densincrali/Cel dcintin. Silile 13Blnd aIdhIcsix\c penetration
dc,tceasced to - 5' Sin-lefion adlIICsix\ C pICtLation con1tinnLCdI to
dilop0 Lilntil itiIseSLICed 0')" alt tIlC iloot- 01 tle dicinicrali/cd
dCIltill IaetVcI 'HIcsc Icsuilts sugCest tIat thc Colnttibution ilrotii thc
SitlicI 13otid adlicsixc is <50()0 thr-ouhI-out lcatrIN hallf o' the
denminctrai/CIe dcilitini Inl Contrast. the pcilctiratioll of1 Onic Step
Idhesix c xxas 5()' thitonLho0Lnt miost ol'thc dc1nmincralizedc dcntini
laxCIe I'hlC Issict o-ataniII speetiCLIse0CpiC tCsLIlts also suggcst tiat the
O(nc Stcp adhesixe ditifLsedi into the minerali/ed dentill substrate
I or cxalmlple the cighthl and ninitil spectria ol' the deontiil ine Step
ad_ieCsix c inttCl-lac shoIw t`CtiLtCs atttibuthIle to the adhesix e as well
as tclati cl\ i ntClsSc IceatLuItCs assocCtCned Withi thIte ttI IC,aI compon1lent
ill (deniti. 'I lie I-elatix c ititeinsity ot' thlcsC fCeItulres Wvithil the same1C
spceti u is siuggcsts adhesix c diill'usio inito tihe neiralli/ed denitiii.
A t -C\xioLts Is1icr-oRam1In Spectroscopic investigation ot' La
dCeiltill adheS i\ eCi ri-dLced
od hy the snissC 1is1s-11inCtUritet as Sinigle Bo[ld
has Simitlariy repoited limIiteCI penCtratiOn1 tiht-OLI oU1-It tIhC
Figure la. Light microg raph of Single Bond (SB) dentin-adhesive
interfoce stained with Gl dner trichrome: mineralized dentin (D, green),
adhesive (beige), and exposed protein (red).
Figure lb. Light micrograph of One Step (OS) dentin-adhesive interface
stained with Goldner trichrome: mineralized dentin )D, green), adhesive
(beige), protein that is partially coated with adhesive or primer
(orange), and exposed protein (red)
Downloaded from jdr.sagepub.com at PENNSYLVANIA STATE UNIV on March 2, 2014 For personal use only. No other uses without permission.
1461
..j
.6
.74.1__
r
IODliii
1785 1 600 14l00 1200 877
Ranian Shiftlcnl-'1
Figure 2A. Raman spectrum of Single Bond adhesive. Selected spect-al
features are labeled (a, b, c, d) and correlated with peak assignments
in the Table.
Figure 2B. Raman spectrurn of One Step adhesive.
Figure 2C. Raman spectrum of dentin.
denmiscrali/cd denitin (Wiclic/lka ci Il1997) 1 ot c\atilplc Itticro-
1Ralilin sp.cet--oscopic tersuilts xxitlI SeotC lih iidri Lritilpi posc l'Iris
('IM. St. Paril MN, US ) suicestscd thIat the adliesixc talitlc( tr
petictrate 3-5 pini o0'tlhc dcminlrnli/cd (iceitill .ieIoplItel)x Ith ough
thsc Catlicr study (ihitht) did lot atten[ipt to tIaLnitiN the cotiipriicits
ol'tIhC lixhi dl ItycL theI rCstilts were co.nliriscihdxmcalns of'thl liglht
microscopy tcCIhIli(lquc descrihcd in the presetit illx CstiaItiots
(Spencerann1-i S affl'ord. 1 999). listerestin Ilx b! oth Single 13otsdtid1
Scotc lihboirl MrliIti pLui_0SC lPILIs rise 3 ` 0 plhospholic tcidL to eteli tile
dcntiti. lic aggrCssix\ C IatUric ofi this etels itiax he olti Inetot
CoIt r-ibLItits c to the IiitCd te ad(h1csix c petnetrationi
lie licht illicrograpIs ol' the rI.litill Siticl_ Botid idliscsix
shoxu s a distinct red huiie approxiLitlatly 4- iim xx icc at the
ilItCtltCe 'I-I is snI_gcsts that tIlee-C is CXlposed pl oltcill at thle d a
intcl-tacc that is an aitlahic ttr rcactioil xx iti the dilllcictin] Staill. I lIe
-2-pisll-wiIe olrilc Ii tie aptleICsenAts p otetill thlint xx aspiSp-otCCtCed h
(nIilf'isioni ol Si iigle13eoIn adliesix e t Ises tils Ixcrc coileo iiiiued( bx
i1l1lN'sis of the illicro-lRaisllst spectroseopie rlata, \x istlt
rlcilolistrated a rapllid (diolp ill Sitngle 13ol(i aIdllesix pctiet altioll ill
tile 11llbid ly ci.
Ill Comipari sol, the ligilt ciriogil-thli ol the ( ioldliti-tritel toltie-
stailnle sectioll fioill the dic ltili (O)ti Stcpl aIdlesix ilitcl fineC sito\\xs a
/ileC ri OLlti`1C Stalill it C;ISuII-illg alppl OXHiWtIacl\ 4 itis I lic otnilt c
SlluugCSts thlat priotei iihx
C0101 lt (tCemtillea li/id dlentini\i
itli is laC is
CenmtCel xthJlidldSi C atCsd tIltLis is ilot axa Ilkel ir lctlectionl itiI tihe
(oioltilsl t-iclil oleic staill ( Spenerc atid SxxaifI10Id. (9 )) li tet eStitgxi
tIlh igilt illiCrOSCOpx srL_CestS' lia u (5 Ltlll Ill \\ i(ltll MO dlititl\
-
approximlaltinlg the istiller'lli/edl (deittiii \xx Ci-e ()11 Stelp adlIlssi C
Concilelt atiol is less thalsl 25' Dcrieased Inlhes i\ e pcllcti itioil alt
-
this positioll suggcsts thalt testidiI;l SisICse klaxVCI_ enIac ClIsCd1Ced tIlh
porosity of tlsC leiitill alid inhibitedr arIllssixc (dill'Lsiois. P rc\ iOIS
ant1_1t1 s Isax e sIisitliFlx- sLi-CsteLI thailt the tCSinlKual sisseatr 11a c-
Collagcn canl sexVcCclV tedLICc the poLosityto' tiC Lctitiii (PiislIc\ / 1/..
'I'9ICC
199) 1Iss anthr0- Isx puotlisi/eLd that xCi CII I Ill'dsi \ tCsiti
isisallCed tno perictl ate this xIerCI ol LlClariLIlCC. Cs`irltInll isica l Clae
colluellt the concentration of arlldssix x\ tlllut thlis Ciitical lax ci
 I1A62 Spencer et al. J Dent Res 79(7) 2000
Ravlh
I . . . . - .
2.0

1.5
EU ..

3% 1.0
4
co
a
.C 0.5
tE 21 A
4 0.0
0O 1785
1400 li00
-0.5 Raman Shiftlcm-1I 1000 877
-d I

0 10 20 30 40 50 60 70 80 Figure 4. Waterfall presentation of the micro-Raman spectra acquired at

one-micron intervals across the dentin/adhesive interface with Single Bond.
Percent of Adhesive (wt°h)
Fi ure 3. Calibration curves for the model compounds made from
adesive and type collagen. The vertical axis represents the relative
ratios of the integrated intensities of a spectral feature from the respective
adhesive and the carbonyl (1667 cm-') from the amide 1. The band at
11 3 cm-' associated with the phenyl group (C-O-C) was the spectral
feature used for Single Bond adhesive (represented by 0). The spectral
feature at 1610 cm-' associated with the deformation of the phenyl
group (C=C) was used for One Step adhesive (represented by A).
would be only a fraction of the concentration beneath this layer.
The area of red stain along the tubules in the light micrograph of
the dentin/One Step adhesive indicates that the protein lining the
dentin tubules was available for reaction with the Goldner trichrome
stain. These results suggest that either the One Step adhesive pulled
away from the walls of the tubule as it polymerized or the presence
of water along the tubule walls inhibited adhesive penetration at 1785
these sites. In either case, the uncovered protein lining the tubules 1400 1200 1000 877
was available for reaction with the differential stain. Raman Shift[cm-11
The lack of complete diffusion throughout the demineralized
dentin with adhesive systems that inhibit collagen collapse through Figure 5. Micro-Raman s ptra acquired at one-micron intervals across
"wet" bonding can be attributed to a variety of factors, including the dentin/adhesive interace with One Step.
dissimilar solvents and differences in the hydrophobicity of the
adhesive. For example, under moist bonding conditions, the
channels between demineralized dentin collagen fibrils are filled
with water (Pashley, et al., 1993; Nakabayashi and Pashley, 1998).
The only mechanism available for monomer infiltration is diffusion
of monomers into whatever solvent is in the spaces of the substrate
and along the collagen fibers. The functions of the solvent in
"single"-bottle adhesives include reducing the viscosity of the
adhesive and changing the vapor pressure of residual water within
the demineralized dentin.
The solvent for Single Bond adhesive is ethanol, while One
Step uses acetone. The vapor pressure of water is very low in
comparison with that of both acetone and ethanol. However, a
factor complicating a direct comparison of these materials based on
solvent characteristics is the inclusion of HEMA in the Single Bond
adhesive. Results from a recent study have shown that HIEMA can
dramatically lower the vapor pressure of water (Pashley et al.,
1998). The results of that study suggest that, as the HEMA
concentration increases, there is a progressive decrease in the partial
pressure of water. As the partial pressure drops, it becomes more
and more difficult for residual water to be removed from the
demineralized dentin. The hydrophobic BisGMA monomer would Figure 6. Graph of percent of adhesive as a function of spatial position
across the dentin/adhesive interface. Vertical dotted line
resist diffusing into these sites where there is residual water.
The results from the present study provide direct chemical
dentin adhesive/hybrid layer interface; corresponds to secondrepresents
spectrum
in Figs. 4 and 5.
Downloaded from jdr.sagepub.com at PENNSYLVANIA STATE UNIV on March 2, 2014 For personal use only. No other uses without permission.
J Dent Res 79(7) 2000 Dentin/Adhesive Interface Chemistry
evidence of an adhesive concentration gradient within the hybrid
layer. The results indicate that although the "wet" bonding
technique provides stability to the collagen, several other factors
interfere with adhesive diffusion into the demineralized dentin
layer. For example, in contrast to One Step adhesive, Single Bond
resisted diffusion into the moist demineralized and mineralized
dentin. It is likely that this resistance reflects the differences in the
composition of these adhesive systems, including variation in the
concentration of the hydrophobic BisGMA monomer.
The micro-Raman spectroscopic results present the first direct
chemical evidence of a probable phase separation in an adhesive
system under "wet" bonding conditions. Both the light
microscopy and micro-Raman spectroscopy results indicate that
the phase separation in concert with the residual smear layer
collagen and deep dentin etch translate to incomplete adhesive
diffusion throughout the demineralized dentin. Phase separation
with consequent interfacial disruption has been reported in
previous morphologic investigations of the "wet" bonding
technique (Tay et al., 1996).
In summary, multiple clinical and laboratory studies have
suggested that the dentin/adhesive bond is the weak link in the
adhesive restoration. Unsuccessful dentin bonding means that there
are sites at the tooth/restoration interface that are vulnerable to
hydrolytic breakdown and susceptible to penetration by bacterial
enzymes or other toxic substances (Walshaw and McComb, 1996).
To date, no one has determined the concentration or distribution of
adhesive that is required to resist the chemical and mechanical
stresses that characterize the oral cavity. The direct and quantitative
measurement of the adhesive concentration throughout the hybrid
layer presented in this study is the first step in this process.
ACKNOWLEDGMENTS
This investigation was supported in part by USPHS
Research Grants DE12252 and DE12487 from the National
Institute of Dental and Craniofacial Research, National
Institutes of Health, Bethesda, MD 20892. The authors
gratefully acknowledge Bisco Corporation and 3M, Dental
Products Division, for donating the dentin adhesive
products used in this study. A preliminary report was
presented at the 1999 meeting of the International and
American Associations for Dental Research, Vancouver,
BC, Canada
REFERENCES
Bauer NJ, Wicksted JP, Jongsma FH, March WF, Hendrikse F,
Motamedi M (1998). Non-invasive assessment of the hydration
gradient across the cornea using confocal Raman spectroscopy.
Invest Ophthalmol Vis Sci 39:831-835.
Eick JD, Robinson SJ, Byerley TJ, Chappell RP, Spencer P,
Downloaded from jdr.sagepub.com at PENNSYLVANIA STATE UNIV on March 2, 2014 For personal use only. No other uses without permission.
1463
Chappelow CC (1995). Scanning transmission electron
microscopy/energy-dispersive spectroscopy analysis of the dentin
adhesive interface using a labeled 2-hydroxyethylmethacrylate
analogue. JDentRes 74:1246-1252.
Goldner J (1938). A modification of the Masson trichrome technique
for routine laboratory purposes. Am JPath 14:237-242.
Gwinnett AJ (1994). Chemically conditioned dentin: a comparison of
conventional and environmental scanning electron microscopy
findings. Dent Mater 10: 150-155.
Kanca J 3rd (1992). Effect of resin primer solvents and surface
wetness on resin composite bond strength to dentin. Am J Dent
5:213-215.
Kinney JH, Balooch M, Marshall GW, Marshall SJ (1993). Atomic-
force microscopic study of dimensional changes in human dentine
during drying. Arch Oral Biol 38:1003-1007.
Lemor RM, Wieliczka DM, Kruger MB, Swafford JR, Spencer P
(1999). Spectroscopic and morphologic characterization of the
dentin/adhesive interface. JBiomed Optics 4:22-27.
Lemor RM, Wieliczka DM, Kruger MB, Spencer P (2000). Dentin
etch chemistry investigated with infrared and Raman
spectroscopy. JRaman Spect 31:171-176.
Nakabayashi N, Pashley DH (1998). Hybridization of dental hard
tissues. Tokyo: Quintessence Publishing Co., Ltd.
Pashley DH, Ciucchi B, Sano H, Homer JA (1993). Permeability of
dentin to adhesive agents. Quintessence Int 24:618-63 1.
Pashley EL, Zhang Y, Lockwood PE, Rueggeberg FA, Pashley DH
(1998). Effects of HEMA on water evaporation from water-
HEMA mixtures. Dent Mater 14:6-10.
Romer TJ, Brennan JF 3rd, Schut TC, Wolthuis R, van den Hoogen
RC, Emeis JJ, et al. (1998). Raman spectroscopy for quantifying
cholesterol in intact coronary artery wall. Atherosclerosis
141:117-124.
Spencer P, Swafford JR (1999). Unprotected protein at the dentin-
adhesive interface. Quintessence Int 30:501-507.
Suzuki M, Kato H, Wakumoto S (1991). Vibrational analysis by
Raman spectroscopy of the interface between dental adhesive resin
and dentin. JDent Res 70:1092-1097.
Tam LE, Pilliar RM (1994). Effects of dentin surface treatments on the
fracture toughness and tensile bond strength of a dentin-composite
adhesive interface. JDent Res 73:1530-1538.
Tay FR, Gwinnett AJ, Wei SH (1996). The overwet phenomenon:
a transmission electron microscopic study of surface moisture
in the acid-conditioned, resin-dentin interface. Am J Dent
9: 161-166.
Walshaw PR, McComb D (1996). Clinical considerations for optimal
dentinal bonding. Quintessence Int 27:619-625.
Wieliczka DM, Spencer P, Kruger MB (1996). Raman mapping of the
dentin/adhesive interface. Appl Spectrosc 50:1500-1504.
Wieliczka DM, Kruger MB, Spencer P (1997). Raman imaging of
dental adhesive diffusion. Appl Spectrosc 51:1593-1596.