Another journal article courtesy of:


BC Growers Association


Comparison of three advanced chromatographic

techniques for cannabis identification

Published in Bulletin of Narcotics, vol. XLVI, No. 2, 1994


Drug Dependence Assessment and Information Centre, Pharmacology
Laboratory, Caen Regional and University Hospital Centre, Caen, France


The development of chromatography technology, with the
increasing availability of easier-to-use mass spectrometers
combined with gas chromatography (GC), the use of diode-
array or programmable variable-wavelength ultraviolet
absorption detectors in conjunction with high-performance
liquid chromatography (HPLC), and the availability of scanners
capable of reading thin-layer chromatography (TLC) plates in
the ultraviolet and visible regions, has made for easier, quicker
and more positive identification of cannabis samples that,----
standard analytical laboratories are occasionally required to
undertake in the effort to combat drug addiction. At labora-
tories that do not possess the technique of GC combined with
mass spectrometry, which provides an irrefutable identification,
the following procedure involving HPLC or TLC techniques
may be used: identification of the chromatographic peaks
corresponding to each of the three main cannabis constituents
- cannabidiol(CBD),A-9-tetrahydrocannabinol(A-9-THC)and
cannabinol (CBN) - by comparison with published data in
conjunction with a specific absorption spectrum for each of
those constituents obtained between 200 and 300 nm. The
collection of the fractions corresponding to the three major
cannabinoids at the HPLC system outlet and the cross -checking
of their identity in the GC process with flame ionization
detection can further corroborate the identification and
minimize possible errors due to interference.
Between 1968 and 1987, references appeared in the literature to
numerous methods for identifying cannabis constituents through the use
of simple techniques involving TLC on silica gel plates with visual
detection by colour reaction [1-61, as well as more elaborate techniques
involving overpressured layer chromatography [71 and HPLC using normal
or reversed phases and detection by absorption at different wavelengths
[8 - 131 or clectrocheniical means [141, and also more complex techniques
combining capillary or packed-column GC with mass spectrometry
[15 - 191.
The subsequent appearance on the market of diode-array or
programmable variable-wavelength ultraviolet absorption detectors that
can produce tfic absorption spectrum of a detected component, and of
scanners capable of reading TLC plates in the ultraviolet and visible
regions, together with the increased availability of mass spectrometers
combined with gas chromatography, which are simpler for non-specialists
to use, means that analytical laboratories working in cooperation with the
judicial authorities in drug addiction cases can now identify cannabis
samples easily, quickly and positively.
Since the new technologies were available and the authors were in
possession of 15 hashish samples from various sources with significant
differences in the relative proportions of the three major cannabis
constituents, that is, A - 9 - tetrahydrocannabinol (,& - 9 - THC), cannabidiol
(CBD) and cannabinol (CBN), it was considered useful to compare their
chromatographic profiles and correlate the findings using three advanced
and highly efficient chromatographic techniques (GC, HPLC and TLC)
in order to determine the best strategy or strategies for reliable identifica-
tion of cannabis samples, depending on the equipment resources of the
individual laboratory.
Experimental part
The analyses were carried out on 15 samples of hashish taken from
resins of different and unknown geographical origins. Various dried
organs of cannabis plants cultivated under laboratory conditions from
hemp seed were also studied.
The HPLC apparatus consists of a Rhdodyne 7125 injector, a Merck
6000 A pump, an LDC SM 4000 programmable variable-wavelength
ultraviolet detector and a Merek D 2000 integrator.
The GC apparatus consists of a Carlo Erba Fractoyap 2900 chromato-
graph equipped with a Ros injector and a flame ionization detector, and
a Varian Star 3400 chromatograph equipped with a programmable
capillary septum injector (SPI injector) and coupled to a Varian Saturn 11
ion-trap mass spectrometer.
The TLC apparatus comprises a Camag 11 scanner connected to a
Merck D 2500 integrator.

Preparation of the samples
The extracts are obtained by ultrasound mixing (for 15 minutes) of
each of the samples, in the ratio of 10 mg of substance to 1 ml of solvent
in a mixture consisting of 90 per cent bexane and 10 per cent chloroform,
which corresponds to the mobile phase used in the HPLC process. The
extracts are ultracentrifuged for 15 minutes at 10,000     revolutions per
Chromatographic analyses
HPLC: TenmicralitresofsupernatantsarcinjectedintoaWatersil-
Porasil column (15 em x 4.6 mm); the mobile phase is made up of 90 per
cent hexane and 10 per cent chloroform; the flow rate is 2 ml per minute
and detection is ef fected at 220 nm. Three main peaks are identified, and
the three fractions that correspond respectively to those three peaks are
collected at the HPLC system outlet. The fractions are concentrated and
reinjected into the HPLC equipment to check their purity. In order to
determine the exact identity of each of the cannabis constituents, a full
spectrum is produced during the chromatography in the ultraviolet aird--
visible regions from 200 to 300 nm.
GC: One microlitre of supernatant is injected into a capillary
column (25 m x 0.32 mm) coated with methyl silicone having a phase
thickness of 0.25 it. The oven temperature is 240' C and the carrier gas
(N,) pressure 0.5 bar. Three main peaks are identified by means of flame
ionization detection, and those peaks are matched to the three peaks
obtained in the HPLC process through separate injection of each of the
three fractions recovered from the HPLC system.
By coupling the same column to the mass spectrometer, a mass
spectrum is obtained for each of the three main     cannabis constituents,
which can thus be positively identified.
TLC: Five microlitres of supernatant is spotted on a Merck 60 F 254
silica gel plate. Once the mobile phase - consisting of a mixture of
bexane, chloroform and dioxane (by volume, respectively, 89, 8.75 and
2.25 per cent) - has migrated to a height of 11 em, the plate is scanned at
220 nm. Each of the three peaks of the main peak area is identified by
comparison with each of the three fractions that have been recovered
from the HPLC process, identified in the GC and mass spectrometry
process and separately deposited on the plate. Each of the spots is
scanned from 200 to 300 nm in increments of 2 nm, with the absorption
spectrum of the substance in question reconstituted in this way.

The chromatograms obtained from the GC, HPLC and TLC processes
are grouped together in figure 1. The retention times (tR) of the three
main cannabis constituents - cannabidiol (peak marked 1), A-
9-tetrahydrocannabinol (peak marked 2) and cannabinol (peak marked 3)
- are 3.95, 4.6 and 5.15 minutes, respectively, for the GC process, and 4,
4.9 and 5.7 minutes, respectively, for the HPLC process. The retardation
factor (R@ values (the distance from the deposit to the tip of the peak
divided by the distance from the deposit to the solvent migration front)
observed in the TLC process are 0.30 for CBD, 0.27 for A-9-THC, and
0.23 for CBN.
The mass spectra of the three cannabis constituents are characteristic
(figure II): a mass peak at 314 and a base peak at 232 for CBD; a mass
peak at 314 but a base peak at 300 for A-9-THC; and a low-intensity
mass peak at 310 and a base peak at 296 for CBN.
The absorption spectra of those three constituents obtained during
the HPLC process and from the scanning of the TLC plate are comparable
(figures 111 and IV): intense absorption from 200 to 230 nm, minimum
absorption between 250 and 260 nm, and a slight increase between 270
and 280 nm for CBD; intense absorption from 200 to 230 nm, minimum
absorption at around 250 nm, and a slightly more pronounced increase
between 270 and 280 nm for A-9-THC; and less intense relative absorp-
tion between 200 and 230 nm, minimum absorption between 240 and 250
nm and maximum absorption at 280 nm for CBN.
The correlation between the areas located below, the curves formed
by the chromatographic peaks obtained in the GC, HPLC and TLC
processes (those areas being proportional to the absolute quantities of
CBD, A-9-THC and CBN), and also the ratios of those areas (which are
proportional to the relative percentage of the constituents CBDIA-9-THC
and CBNJA-9-THC) were calculated for 10 hashish dr cannabis resin
samples containing substantial quantities of each of the three main
constituents (see table). These correlations are excellent between HPLC
amid TLC, and less good, but nevertheless significant, between GC and






Mass spectrometry combined with GC allows positive identification
of CBD, 4-9-THC and CBN. The relative retention t'MCS (thC tR Of A-
9-THC being considered equal to 1) of CBD and CBN - 0.86 and 1.12,
respectively - are consistent with values cited in the liter t"re: 0.87 and
1.11 [111, 0.69 and 1.31 [171, 0.69 and 1.03 [201; and 0.87 and 1.22 [211.
Identification by comparison with published data can thus be envisaged
as a first step [121. Also, the relative retention times observed in the
HPLC process - 0.82 and 1.16 - are comparable with those obtained by
Kanter and others [101 using a standard column: 0.79 and 1.28. The
fairly characteristic absorption spectra of each of the major cannabis
constituents enable the identification to be subsequently confirmed.
The use of the TLC technique, while requiring less sophisticated
apparatus, is a more delicate operation. Separation is difficult, and the
resolution, although sufficient for visualization by means of colour test
reagents, is barely adequate for reading by a scann'br. The phase
composition must he extremely precise. Any slight excess of chloroform
moves the &-9-THC close to the CBD, whereas any slight excess of
dioxane moves it close to the CBN. However, variable wavelength
readings of the spots make it possible to reconstitute the characteristic
absorption spectra of the three major cannabis constituents and thus to
confirm the nature of each of the three cannabinoids.

The analyses carried out on the samples in the authors' possession
made it possible to verify some of the data that should not be overlooked
during cannabis sample identification and led to the following conclu-
(a) The psychotropic and behavioural effects of cannabis are
essentially associated with &-9-THC [221;
(b) It is the relative percentages of the three major constituents -
,& - 9 - THC, CBD and CBN - that distinguish fibre hemp from resin hemp
(c) CBD or CBN may be absent even from fresh samples [5, 241.
The presence of the three main constituents is not constant, and only the
presence of &-19-THC in substantial quantities qualifies a sample of
hemp as a drug;
(d) Samples taken from one and the same block of cannabis resin
display fairly comparable chromatographic profiles, indicating the fairly
good homogeneity of that form of preparation;
(e) Samplestakenfromdifferentblocksofresin,eveniforiginating
from the same batch seized by the customs authorities, may produce very
different profiles. It is thus not possible to determine the precise origin
of a sample solely on the basis of the relative proportions of &-9-THC,
CBD and CBN. Only a knowledge of the minor constituents will make it
possible to ascertain the provenance of different samples of marijuana,
hashish or other forms of cannabis [16, 17, 251;
(f) A-9-THCconcentrationdecreasesovertimeinresinandhashish
samples [26, 271. Accordingly, it will be difficult to distinguish between
a fresh fibre hemp sample and an old resin hemp sample;
(g) &-9-THCconcentrationisatitshighestinthefemaleinflores-
cences, less in the male inflorescences, lower in the leaves, and virtually
nil in the stems and seeds [261.
There is no doubt that the combination of gas chromatography and
mass spectrometry guarantees positive identification of cannabis samples.
However, for laboratories lacking that technology, it would appear that
the separation of the major cannabis constituents by means of TLC or,
better still, HPLC in conjunction with the absorption spectra of those
components, or else a combination of two chromatographic techniques
(gas or liquid), can guarantee reliable analysis findings and a virtually
irrefutable identification. In most cases the use of those chromatography
techniques in combination make it possible to overcome the problem of

errors due to interference, which are often unavoidable when a single,
simple method with comparison of retention times is used.
1.    H. Aramaki and others, "Forensic chemical study on Marihuana 1.
-    A detection method of the principal constituents by thin-layer and
gas chromatography",     Chemical and Pharmacological Bulletin,
vol. 16, No. 5 (1968), pp. 822-826.
2.    K. D. Parker and others, "Rapport pr@liminaire sur la separation et
la determination quantitative par la chromatographic sur couches
minces et la chromatographic en phase gazeuse des constituents
cannabiniques pr6sents dans la plante ou ajout6s A 1'urine', Bulletin
des stupgfiants, vol. 20, No. 4 (1968), pp. 7-12.
3.    R. A. de Zeeuw, Th. M. Malingre and F.W.H.M. Merkus,
1-Tetrahydro- cannabinolic acid, an important component in the
evaluation of cannabis products', Journal of Pharmacy and Pharma-
cology, vol. 24, No. 1 (1972), pp. 1-6.
4.    M. R. Paris and R. R. Paris, 'Importance de la chromatographic pour
I'Ctude des constituents du cannabis sativa L', Bulletin de la Soci@t@
chimique de France, vol. 1, 1973, pp. 118-122.
5.    C. E. Turner and K. Hadley, 'Constituents of Cannabis sativa L. ll:
Absence of cannabidiol in an African variant', Journal of Pharma-
ceutical Sciences, vol. 62, No. 2 (1973),-pp. 251-255.
6.    S. N.Tewari and J. D. Sharma, 'Separation and identification of
cannabinoids from Cannabis indica L. by thin-layer chromaio-
graphy", Pharmazie, vol. 34, No. 1 (1979), pp. 54-55.
7.   P. Oroszian and others, "Separation, quantitation and isolation of
cannabinoids from Cannabis sativa by overpressure layer chromato-
graphy", Journal of Chromatography, vol. 388, No. 1 (1987), pp. 217 -
8.   P. B. Baker and others, "Determination of the distribution of
cannabinoids in cannabis resin using high-performance liquid
chromatography", Journal of Analytical Toxicology, vol. 4, 1980,
pp. 145-152.

9.   R. N. Smith and C. G. Vaughan, "High-pressure liquid chromato-
graphy of cannabis: quantitative analysis of acidic and neutral
cannabinoids", Journal of Chromatography, vol. 129, 1976, pp. 347-
10.  5. L. Kanter, M. R. Musumeci and L. E. Hollister, 'Quantitative
determination of &-9-tetrahydrocannabinol and &-19-tetrahydro-
cannabinolic acid in marihuana by high-pressure liquid chromato-
graphy', Journal of Chromatography, vol. 171, 1979, pp. 504-508.
11.  D. J. Harvey, 'Chemistry, metabolism and pharmacokinctics of the
cannabinoids", in Marihuana in Science and Medicine, G. G. Nahas
and others, eds. (New York, Raven Press, 1984), pp. 37-107.
12.  D. Debruyne and others, "Identification et diff@renciation des
chanvres r6sineux et textile: utilisation conj ugu6e de HPLC et GLC
haute resolution, Bulletin des stup@fiants, vol. 33, No. 2 (1981),
pp. 45-54.
13.  1. Barni Comparini and F. Centini, "Packed -column chromatography,
high-resolution gas chromatography and high-pressure liquid
chromatography in comparison for the analysis of cannabis constitu-
ents', Forensic Science International, vol. 21, No. 2 (1983), pp. 129-
14.  1. Nakabara, 'Studies on confirmation of cannabis use. 1. Determina-
tion of the cannabinoid contents in marijuana cigarette, tar, and ash
using high-performance liquid chromatography with clectrochemical
detection', Journal of Analytical Toxicology, vol. 9, No. 3 (1985), pp.
15.  T. B. Vree and others, "Gas chromatography of cannabis constituents
and their synthetic derivatives', Journal of Chromatography, vol. 74,
1972, pp. 209-224.
16.  M. Novotny and others, "Analysis of marijuana samples from
different origins by high resolution gas-liquid chromatography for
forensic application', Analytical Chemistry, vol. 48, No. 1 (1976),
pp. 24-29.
17.  L. Stromberg, "Minor components of cannabis resin. VI. Mass
spectro-metric data and gas chromatographic retention times of
components cluted after cannabinclu, Journal of Chromatography,
vol. 121, 1976, pp. 313-322.

18.  D. J. Harvey, 'Comparison of fourteen substituted silyl derivatives
for the characterization of alcohols, steroids and cannabinoids by
combined gas-liquid chromatography and mass spectrometry',
Journal of Chromatography, vol. 147, 1978, pp. 291-298.
19.  H. Grote and G. Spiteller, 'Neue cannabinoide. 11', Journal of
Chromatography, vol. 154, 1978, pp. 13-23.
20.  M. Novotny and others, "Fractionation and capillary gas chroinato-
graphic-mass spectrometry characterization of the neutral compo-
nents in marijuana and tobacco smoke condensates", Journal of
Chromatography, vol. 238, 1982, pp. 141-150.
21.  T. B. Vree and others, "Identification in hashish of tetrahydro-
cannabinol, cannabidiol and cannabinol analogues with a methyl
side -chain", Journal of Pharmacy and Pharmacology, vol. 24, No. 1
(1972), pp. 7-12.
22.  G. G. Nahas, 'Toxicology and Pharmacology", in Marihuana in
Science and Medicine, G.G. Nahas and others, eds. (New York,
Raven Press, 1984), pp. 109-246.
23.  J. Bruncton, El@ments de phytochimie et de pharmacognosie (Paris,
Lavoisier - collection technique et documentation, 1987.
24.  P. S. Fetterman and others, 'Mississippi-grown Cannabis sativa L.:
Preliminary observation on chemical definition of phenotype and
variations in tetra- hydrocannabinol content versus age, sex, and
plant part', Journal of Pharmaceutical Sciences, vol. 61, No. 8
(1971), pp. 1246-1249.
25.  R. Brenneisen and M. A. Elsohly, "Chromatographic profiles of
cannabis of different origins: Part 1', Journal of Forensic Sciences,
vol. 33, No. 6 (1988), pp. 1385-1404.
26.  J. W. Fairbairn, J. A. Licbmann and M. G. Ronan, "The stability of
cannabis and its preparations in storage', Journal of Pharmacy and
,Pharmacology, vol. 28, 1976, pp. 1-7.
27.  D. J. Harvey, "Stability of cannabinolds in dried samples of cannabis
dating from around 1896-1905", Journal of Ethnopharmacology,
vol. 28, 1990, pp. 117-128.