PART1.

CHEMISTRYAND PHARMACOLOGY

CULTIVATION, EXTRACTION, AND ANALYSIS OF

C A N N A B I S SATIVA L.*

Norman J. Doorenbos, Patricia S. Fetterman, Maynard W. Quimby

and Carlton E. Turner
Department of Pharmacognosy
Research Institute of Pharmaceutical Sciences
School of Pharmacy, University of Mississippi
University, Miss. 38677

Cannabis sativa L. (marijuana) has been cultivated at the University of

Mississippi since 1968 under contract support from the National Institute of
Mental Health. The primary objective is to provide fully defined standardized
marijuana for research. Confiscated marijuana of unknown history is often
adulterated.
C . sutiva was grown on three sites in Mississippi in 1968 and on two sites in
1969 and 1970. The soil types were loess (clay) on campus (site A), sandy loam
some 300 miles south of campus (site B), and clay loam in the delta region of
the state (site C ) . Site C was used only in 1968.
Land was prepared by conventional agricultural methods and treated with
slag at the rate of two tonslacre. Seeds were planted mechanically or by hand
at depths of about one-half inch in 40-inch-wide rows with a side dressing of
13-13-13 fertilizer at the rate of 300 pounds/acre. Plantings were made in
May, June, and July. Seeds germinated in less than one week if the soil was
moist. Weeds were controlled by cultivators and hand hoes. N o herbicides or
insecticides were used. Weed control was especially important to the young
plants, since their growth was greatly inhibited by competition for light, water,
and soil nutrients. Plants were watered at least once a week when it did not rain.
Plants were usually thinned to 12-18 inches apart when 12-24 inches high.
This was best accomplished by clipping an inch above the ground, since uproot-
ing frequently set back or killed neighboring plants. This method allowed greater
branching and leaf and flower development. Plants were also side-dressed one
or two additional times during the growing season.
Many seed types have been planted in this study. One cannot help being
amazed at the many morphological variations observed among plants produced
by these seed types, and it must be concluded that Cannabis is either a genus
composed of more than 100 species or a single species that has not stabilized and
that has many variations. We prefer to look upon these plants as a single species,
Cannabis sutiva, as first named by Linnaeus. Plants within a given seed type
tended to resemble each other and grow at similar rates, although some marked
differences were occasionally observed.
Plants grew as rapidly as two feet a week during peak growth and were three
to eighteen feet in height at maturity. Branches measured from a few inches to
eight feet in length and leaves from two to twelve inches across, and the number
of leaflets per leaf ranged from one to eleven, almost always odd in number
(FIGURES 1, 2, and 3). Plants of several variants were observed to contain many

Supported by contracts PH-43-68-1307, PH-43-69-1307, and HSM-42-7&109 from the

National Institute of Mental Health, United States Public Health Service, Bethesda, Md.
3
4 Annals New York Academy of Sciences
simple leaves, and in one variant all of the leaves in many of the plants were
monophyllus (FIGURE 3). Leaves of Cannabis are typically palmately compound.
Recently, however, we have observed leaves that are pinnately compound on
an herbarium specimen from Afganistan in the herbarium of the Botany De-
partment, Field Museum of Natural History, Chicago, Ill.
C. sativa is a dioecious species, but many monoecious plants were observed
in several variants. Bees were attracted by male flowers but not by female
flowers. Pollination appears to be accomplished by air currents. We have been
unable to establish sex until flowering begins. Male plants begin shedding leaves
shortly thereafter and die after shedding pollen. Female plants lose many of
their leaves, particularly older ones, which tend to be larger in size and contain
more leaflets than younger leaves, as the fruit ("seed") matures. Each female
flower is surrounded by a floral bract (a leaflike structure), which subsequently
surrounds the maturing seed. A male and female plant are illustrated in FIGURE 4.
Plant parts of some variants were harvested throughout the growing season.
These samples and samples of marijuana produced at maturity from the various
seed types have been analyzed for certain cannabinoids. Plants matured in
August, September, and October after three to five months of growth. Most
marijuana was produced by cutting the stem of plants beneath the lowest
branches, air-drying, and hand-stripping seed, bracts, flowers, leaves, and small
stems from these plants. Most of the stem material was subsequently removed
by a ten-mesh screen, and the seed, by means of a mechanical seed separator.
This manicured marijuana contained less seed and stem material than most of
the illicit samples of marijuana that we examined.
Marijuana has been extracted successfully with hexane, petroleum ether,
benzene, chloroform, and ethanol. Of these solvents, chloroform proved to be
the most useful in extracting marijuana samples for analysis.

1. Female flowering tops of a Mexican variant of Cannubis surivu L.

FIGURE
 Doorenbos et al. : Cultivation, Extraction 5

2. Female flowering tops of an Indian variant of Cannabis saliva L.

FIGURE
6 Annals New York Academy of Sciences
A nalyses
Samples were extracted by a modification of Lerner's' method? Weighed plant
samples are extracted with chloroform, the solvent is evaporated, and the resi-
due is dissolved in ethanol and filtered and analyzed with a gas chromatograph
with use of 4-androstene-3, 17-dione as the internal standard. Analyses were
performed with use of Beckman GC-5 and GC-45 gas chromatographs equipped
with flame ionization detectors and operated isothermally at 210' C with an
inlet temperature of 230" C. At this temperature, acids rapidly decarboxylate in
the injection port to their respective phenols. The columns were 10 foot % inch
stainless steel packed with 2% OV-17 on 100/120 mesh Gas Chrom Q. Nitrogen
was used as the carrier gas at a flow rate of 30 ml/min. Peak area measurements
were made by the method of peak height X width at half-height. A comparison
of typical gas chromatograms is illustrated in FIGURE 5 .
Variants of C. sativa grown thus far have been found to be of the drug or fiber
phenotype, a method of classification first suggested by Dr. Coy Waller.2 Pheno-
type is determined by calculating the following ratio:

Phenotype ratio =
% A'-THC +
% CBN
% CBD
Levine4 noted that loss of potency of marijuana is accompanied by conversion
of A9-THC to (cannabinol) (CBN). This suggests that he-THC 4- CBN would
approximate the AO-THC content of a sample when freshly harvested. Mari-
juana with ratios greater than 1.0 is classified as the drug type and that with
ratios smaller than 1.0 as the fiber type. Most samples have been found to be
larger than 5.0 or smaller than 0.2, making classification simple. FIGURE5 il-

3. A monophyllus plant.
FIGURE
 Doorenbos et al. : Cultivation, Extraction 7

FIGURE
4. Male (left)and Female (right) plants of a Mexican variant of Cannabis sariva L.
8 Annals New York Academy of Sciences
lustrates the gas chromatograms of the drug type (Mexican) and fiber type
(Turkish).
Cannabinoid acids are analyzed as trimethylsilyl esterether derivatives, with
use of the same columns, equipment, and operating condition^.^ The silyl deriva-
tives are prepared by treating the extract with bis( trimethylsilyl)4rifluoroaceta-
mide plus 1% trimethylchlorosilane (Regis Chemical Co.) .
AO-Tetrahydrocannabinol(Lo-THC) and cannabidiol (CBD) (FIGURE6) were
found to be present to the extent of an average of 95% as their respective acids,
AO-tetrahydrocannabinolicacid A (AO-THCA (A) ) , AO-tetrahydrocannabinolic
acid B (A0-THCA ( B ) ) , and cannabidiolic acid (CBDA). This suggests that
these acids are products of biosynthesis, and the nonacids are decarboxylation
products. The acid content of year-old marijuana samples was found to be 50%
or less. Decarboxylated marijuana, with less than 5% of AO-THC and CBD
present as their acid precursors, has been produced by heating marijuana at
100' C for 75 minutes in a nitrogen atmosphere (see TABLE 1 ) . No significant
decarboxylation occurred below 80' C during the 75-minute treatment used
in this study.

9-mc

I !I CAndrosfCnr
3. 17dione

5. Gas chromatograms of two variants of marijuana.

FIGURE
 Doorenbos et al. : Cultivation, Extraction 9
P3

R=H CBD R&R!=H A9-THC

R=COOH CBDA R = COOH. R I - H A9-THCA I A )
R I H, R' = COOH A9-THCA 18)

6. Cannabinoids.
FIGURE

TABLE 1
DECARBOXYLATION
OF AO-THC ACIDSBY HUT*

Temp. % CBD o/c A 9-THC 76 of A 9-THC BS acids % CBN

250 0.14 2.0 70 0.040

50' 0.13 I .9 74 0.043
60' 0.14 1 .a 67 0.040
70' 0.14 I .9 68 0.039
800 0.14 1.9 54 0.043
90' 0.13 1.9 26 0.042
1000 0.14 1.9 2 0.04 I

Mexican variant marijuana was maintained in an oven at 75 minutes in a nitrogen atmos-

phere at the indicated temperature for this study. All temperatures shown were Centigrade.
t Proportion of AO-THC present as acids.

A comparison of marijuana produced from male, female, and monoecious

plants, when available in a given variant, demonstrates that there is no significant
difference in cannabinoid content between sexes, contrary to the widespread
belief that the active substance is found exclusively or largely in female plants
(see TABLE 2).
The cannabinoid content of plant parts has been found to decrease in the
following order:
bracts
flowers
leaves
small stems
large stems
roots and seed.
Bracts with a Ag-THC content as high as 11.4% have been analyzed.
Environmental factors, including climate, are not as important as heredity in
determining the cannabinoid content of harvested marijuana, contrary to the
widespread belief that warmer and sunnier climates produce the most potent
marijuana.
Individual plants within a given variant often differ considerably in canna-
binoid content. For example, marijuana produced from male and female plants
of a Mexican variant ranged from 1.7-7.2% and 1.5-4.8%, respectively.
10 Annals New York Academy of Sciences
TABLE 2
CANNABINOID
CONTENTS OF SOME MARIJUANA PRODUCED
AT THE
UNIVERSITY OF MISSISSIPPI, 1970

Variant Sex %CBD %A9-THC

India - I Male 2.1 0.81

Female 0.89 1.3
India - I1 Male 0.15 2.2
Female 0.12 1.2
Mexico Male 0.86 3.7
Female 0.35 3.7
Thailand- I Male 0.35 2.1
Monoecious 0.31 2.5
Thailand- I1 Male 0.15 2.4
Female 0.41 2.5
Thailand - 111 Male 0.08 3.2
Female 0.42 3.2
Korea Male 0.16 1.2
Female 0.13 0.94

Iowa Female 2.1 0.068

Russia Female 2.4 0. I6

Lebanon Female 2.0 1 .o

A C. sativa plant that allegedly had been grown in a closet with a tungsten
light bulb was recently delivered to us. It was leggy, with greenish-yellow leaves.
Marijuana harvested from this plant analyzed A9-THC, 6.8%; CBD, 0.26%;
CBN, 0.28%. It produced potent marijuana under adverse lighting conditions.
No significant differences have been observed between marijuana harvested
from plants grown at sites A and B in Mississippi, .although they differed in
latitude by more than 500 miles. These observations support others that have
been reported by our laboratory.2 A comparison of the cannabinoid content
of variants from Thailand, Minnesota, and Panama, grown at greatly differing
latitudes, further illustrates that heredity, rather than climate, is an important
factor in determining cannabinoid content (see TATBLE 3).
TWOvariants of C. sativa, a Mexican and a Turkish, have been grown for
three successive generations in Mississippi. The cannabinoid content has not
changed much in these variants (TABLE 4 ) . The values reported in TABLE 4
represent large batches from fall harvests and hence are average values for
marijuana produced from many plants. The increased As-THC content in the
third-year Turkish is thought to be the result of some cross-pollination from
plants of the drug type.
The cannabinoid content of marijuana produced at maturity from variants of
 Doorenbos et al.: Cultivation, Extraction 11
TABLE3
CANNABINOID
CONTENTVERSUS
PLACE
GROWN
Place Grown Year % CBD % A9-THC % CBN

Thailand Thailand 1969 0.14 4.8 0.12

Mississippi 1970 0.17 3.4 0.11
Minnesota Minnesota 1968 1.7 0.038 0.15
Mississippi 1969 I .6 0.077 0.14
Panama Panama 1969 0.29 3.2 0.80
New Hampshire 1970 0.38 4.0 0.01

4
TABLE
CANNABINOID
CONTENTVERSUS
GENERATION.

Sample % CBD % A9-mc % CBN

Mexican ( 1 st year) 0.31 3.7 0.1 I

Mexican (2nd year) 0.19 3.9 0.17
Mexican (3rd year) 0.18 3.7 0.22

Turkish (1st year) 3.1 0.16 0.12

Turkish ( 2nd year) 2.1 0.06 0.05
Turkish (3rd year) 1.9 0.62 0.37

* Analyses for successive generations grown in Mississippi.

R R
-'9"19 Cmpesterol =O Fliedelin
-'10"21 6 - Silosterol -OH Epifriedelanol

N - (p-hydroxy-~henylelhyl)-p-hydroxy-(r.nssiniuMmide

7. Some Noncannabinoid constituents of Cannabis sariva L.

FIGURE
12 Annals New York Academy of Sciences
C. sativa grown on campus varies widely, even though all environmental factors
are similar. Examples of the analyses of samples obtained during the 1970
growing season are presented in TABLE 2.
The number of cannabinoids and other chemical substances that have been
identified in marijuana continues to increase. The potentially harmful effects
of the presence of each of these substances should eventually be considered.
The steroids, p-sitosterol and campestrol, have been isolated from marijuana,
and the triterpenes, friedelin and epifriedelanol, from C . sativa roots in our
laboratory (FIGURE 7).5 It is likely that these triterpenes are also present in
marijuana. The presence of sterols and possibly triterpenes as well in marijuana
is of interest because of the possible conversion of these substances into carcino-
genic hydroperoxides and hydrocarbons.%
N- (p-Hydroxy-P-phen ylethyl ) -p-hydroxy-trans-cinnamide5 (FIGURE 7 ) was
also isolated from the roots of C. sativa. Marijuana is being examined for the
possible presence of this tyramine amide derivative.

References
1. LERNER, P. 1969. The precise determination of tetrahydrocannabinol in marijuana and
hashish. Bull. Narc. 21: 39.
2. FETTERMAN, P. S., E. S. KEITH, C. W. WALLER, 0. GUERRERO, N. J. DOORENBOS & M. W.
QUIMBY.1971. Mississippi grown Cannabis saliva L. A preliminary observation on the
chemical definition of phenotype and variations in the THC content versus age, sex and
plant part. J. Pharm. Sci. 60: 1246-1249.
3. FETTERMAN, P. S., N. J. DOORENBOS, E. S. KEITH& M. W. QUIMBY. 1971. A simple gas
liquid chromatography procedure for determination of cannabinoidic acids in Cannabis
safiva L. Experientia 27: 988-989.
4. LEVINE, J. 1944. Origin of cannabinol. J. Am. Chem. SOC.66: 1868.
5. SLATKIN, D. J., N. J. DOORENBOS, L. S. HARRIS,A. N. MASOUD,M. W. QUIMBY & P. L.
SCHIFP,JR. 1971. Chemical constituents of Cannabis safiva L. root. J. Pharm. Sci.
Submitted for publication.
6. SMOKING AND HEALTH REPORTOF THE ADVISORY COMMITTEE TO THE SURGEON GENERAL OF
THE PUBLIC HEALTH SERVICE.1964. U. S. Department of Health, Education, and Wel-
fare : 51-52.

Discussion
DR. ALBERT SATTIN (Case Western Reserve University, Cleveland, Ohio) :
You implied in your classification between drug type and/or fiber type, that
there was a superiority in the fiber type. Is there any data to support that? The
way you described it, rather than high and low drug type, you say “fiber type
versus drug type.”
DR. DOORENBOS: Yes, that is a good question. Generally speaking, the fiber
type were plants that are being grown for fiber, but it could be misleading to as-
sume that all fiber types are going to be low in THC. We have seen an insufficient
number of variants at this time to be certain that there is a relationship between
chemical content and the purpose for which the plant has been used.
DR.GABRIEL NAEIAS (Columbia University, New York, N . Y . ): Do you have
any data on the THC content of seed plant from Morocco and the West Indies?
DR. DOORENBOS: We have no data on seed from Morocco. We have not ob-
tained any seed from Morocco as yet. We would love to. We have some analyti-
cal data on some plants from Jamaica, and they were high in THC.
But I would hasten to point out that one cannot be certain about the can-
nabinoid content according to the part of the world it comes from. For example,
we have received two seed types from India, and one of them turned out to be
 Doorenbos et al. : Cultivation, Extraction 13
what we call the fiber type. It was also a tall plant. When it had been sent to us,
we were assured that it produced good marijuana, but it produced marijuana
somewhat like that produced in the West.
Thus far, all of the samples we have examined from the near East, from
Lebanon and Turkey, for example, have been of the fiber type.
DR. JONES: What is the most potent chemical structural type, in your
opinion?
DR. DOORENBOS: You are going to get a great deal more information about
this from some of the other speakers. But the do-THC is thought to be the ma-
terial that is responsible for what happens, and some details on what happens to
this THC and its significance are going to be fully described later on.
DR. JONES:The other question was on countries of origin. Which, in your
mind, is the country of origin that produces the most potent types?
DR. DOORENBOS: We have no real opinion as to the country of origin that
produces the most potent type, but of the illicit samples of marijuana that we
have seen, the most potent have come from Southeast Asia. I am not certain
that the plants are more potent there; it may be just a matter of technology,
because the cannabinoid content is related to the type of the plants, to the time of
harvest, and to the plant part that was selected.
DR. HARRYHERMONN (Buflalo Counry Mental Health Services, Buflalo,
N.Y.): I did some research in marijuana several years ago, growing my own
plants, and some of my observations are slightly different from yours, and I
would like you to comment on it.
I found out that growing plants not watered regularly as you said, every week,
but rather growing in a desert condition with hot weather and little water
definitely develop more rapidly, maturing faster and developing stronger po-
tency THC, just as a protective coating from drying plants that are growing with
local intensive oxygenation at the high altitudes gives protection.
Plants you just mentioned, for instance, growing in a closet somewhere, prob-
ably d o not have enough oxygen or light or ultraviolet light and also develop
protective coating of the resin. I would say that what we are considering is an
active ingredient that serves to protect the plant from unfavorable conditions.
Therefore, where we are not caring for the plants so much, they develop much
better and stronger in THC content. Did you notice this?
DR. DOORENBOS: No, but we did not do any studies with the objective of find-
ing out the effect of various environmental conditions, and so our observations
in this area are more limited. It is our impression, based on a scattering of ob-
servations that we have seen, that it is the heredity of the plant that is a dominant
thing, and that you can get potent marijuana under what would be rather adverse
conditions from an agricultural standpoint.
Now, our mission was to produce marijuana in quantity that was of good
quality and of known history. I described the conditions that we used for growing
the plant. We used good agricultural conditions, and so we got a lot of marijuana.
If you grew marijuana under the conditions you described, the plants would
not grow as tall as they did in our garden and you would not get as much
marijuana from the plant. But I am not certain that it would be more potent.
DR. HERMONN: This is exactly what I am trying to say: If you have bigger
plants that are better cultivated the quality is different, and if one wanted to
produce a good quality, maybe he should subject plants to adverse conditions,
because it is my feeling that what we consider active ingredients are in direct
proportion to what the plant has to go through in order to protect itself.
14 Annals New York Academy of Sciences
DR. DOORENBOS: We have no evidence of this. In fact, one of the concerns is
that we are producing marijuana that is being used in the illicit trade.
DR. HERMONN: Did you do any experiment subjecting seeds to some chemical
influences producing different types of plants with a high content of THC? For
instance, by putting the seeds into certain chemicals like colchicine, which some-
how changes the genetic structure, you can produce plants with a different type
of leaflets and different THC content. Did you experiment with that?
DR. DOORENBOS: No, we have not done that.
DR. ROBERTSHALER:(University of Pittsburgh School of Pharmacy and
Pittsburgh and Allegheny Crime Lab., Pittsburgh, P a . ) : One of the common
methods of analyzing for the purpose of THC in illicit marijuana samples is the
use of the Duquenois test, and I want to know, based on your observations, if
these THC acids are extracted with chloroform? Would the Duquenois test then
be based on the fact that the color reaction that is formed is due to the reaction
forming the color with the acids rather than with the THC itself? Or would you
expect that a decarboxylation would take place under reaction conditions?
DR. DOORENBOS: I cannot answer that. It would be just an opinion. Dr.
Waller has indicated that in strong acidic conditions, it should be decarboxylated.
DR. SHALER: But in normal extraction with chloroform, would you then ex-
tract the acid?
DR. DFRENBOS:Yes, you would; you would extract the acids. If you
rigorously avoid heat, then you will have very little decarboxylation, but if you
heat it at all, then you will cause decarboxylation, too.