Chromatographic and Spectroscopic Data of
Cannabinoids from Cannabis sativa L.
Arno Hazekamp, Anja Peltenburg, and Rob Verpoorte
Division of Pharmacognosy, Institute of Biology, Leiden University,
Leiden, The Netherlands
Laboratoire de Toxicologie et de Chimie Forensiques, Institute
Universitaire de Me
gale, Lausanne, Switzerland
Abstract: Chromatographic and spectroscopic data was determined for 16 different
major cannabinoids from Cannabis sativa plant material as well as 2 human meta-
bolites of D
-tetrahydrocannabinol. Spectroscopic analysis included UV absorbance,
infrared-spectral analysis, (GC-) mass spectrometry, and spectrophotometric analysis.
Also, the ﬂuorescent properties of the cannabinoids are presented. Most of this data
is available from literature but scattered over a large amount of scientiﬁc papers.
In this case, analyses were carried out under standardised conditions for each tested
cannabinoid so spectroscopic data can be directly compared. Different methods for
the analysis of cannabis preparations were used and are discussed for their usefulness
in the identiﬁcation and determination of separate cannabinoids. Data on the retention
of the cannabinoids in HPLC, GC, and TLC are presented.
Keywords: Cannabis sativa, Plant material, Cannabinoids, Metabolites, Chromato-
graphic data, Spectroscopic data
In recent years a lot of research on the medical applications of Cannabis sativa L.
has been initiated, as several, mostly European countries, move towards a more
Address correspondence to Arno Hazekamp, Division of Pharmacognosy, Institute
of Biology, Leiden University, Einsteinweg 55, 2300 RA, Leiden, The Netherlands.
Journal of Liquid Chromatography & Related Technologies
, 28: 2361–2382, 2005
Copyright # Taylor & Francis, Inc.
ISSN 1082-6076 print/1520-572X online
liberal view on the use of Cannabis as a medicine.
Although more than 400
compounds have been identiﬁed in the Cannabis plant,
most studies have
focused on the effects of the cannabinoids, in particular (2 )-D
-THC). One reason is that the main pharmacological and
psychoactive effects of Cannabis have been attributed to D
instance, synthetic D
-THC (dronabinol, Marinol
) has been shown to
possess anti-emetic properties useful in cancer therapy. However, in several
medical studies the effect of D
-THC or dronabinol alone could not match the
effects of a total Cannabis preparation,
indicating there might be other
active compounds present.
More than 60 cannabinoids
identiﬁed in Cannabis, and occasionally new cannabinoids are being discov-
The chemical structures of the main cannabinoids from the Cannabis
plant are shown in Figure 1, and their physical properties are listed in Table 1.
Only a few of these cannabinoids have been studied in detail, although several
of these have been shown to possess some biological activity (reviewed by
To extend the knowledge of the therapeutic properties to cannabinoids
other than D
-THC, large amounts of pure compounds must be available.
Assessment of cannabinoids pharmacology is now almost restricted to the
few that are commercially available (e.g., D
-THC, CBD, and
CBN). Furthermore, pure cannabinoids must be available as reference
compounds for their unequivocal identiﬁcation and determination. For that
purpose, chromatographic and spectroscopic methods and data are available
from scientiﬁc literature. Although these data have been published for most
known cannabinoids during isolation and identiﬁcation experiments (see
for an overview), they are scattered over a huge amount of scientiﬁc papers.
Moreover, standardised data obtained under identical analytical conditions
have not been reported yet. As far as we know, the ﬂuorescent properties of
the cannabinoids remain largely unknown.
This report lists the main chromatographic and spectroscopic data of 16
main cannabinoids and of two of their human metabolites obtained under
identical analytical conditions. Methods were kept as straightforward,
simple, and rapid as possible. The pros and cons of each method will also
be discussed. All analyses were carried out for each cannabinoid as far as
permitted by the amount available to us.
Standards and Solvents
Reference compounds of (2 )-D
(CBN), cannabidiol (CBD), cannabigerol (CBG), (2 )-D
hydrocannabinolic acid A (THCA), cannabidiolic acid (CBDA), and cannabi-
gerolic acid (CBGA) were isolated from plant materials previously in our
H-NMR method was developed for their quanti-
A. Hazekamp et al.2362
-THC) was obtained from
Sigma. The main human metabolites of D
-THC, i.e., 11-hydroxy-THC (11-
OH-THC) and 11-carboxy-THC (THC-COOH) were purchased from
Cambridge isotope laboratories (CIL, Innerberg, Switzerland) and from
Figure 1. Structures of the cannabinoids.
Chromatographic and Spectroscopic Data of Cannabinoids 2363
Lipomed (Arlesheim, Switzerland), respectively. All these cannabinoids were
available as certiﬁed and calibrated reference standards. The other cannabi-
noids used for this study were obtained by preparative HPLC on extracts of
Cannabis sativa plant materials and identiﬁed by comparing their chromato-
graphic and spectroscopic data with literature,
[13 – 15]
and by a search in
Table 1. Physicochemical properties of the cannabinoids
# Cannabinoid Full name (description) MW (calc.)
1 d9-THC trans-(2 )-delta-9-tetrahy-
314.472 21 30 2
2 d8-THC trans-(2 )-delta-8-tetrahy-
314.472 21 30 2
3 THV trans-(2 )-delta-9-tetra-
(C3-isomer of THC)
286.418 19 26 2
4 CBD cannabidiol 314.472 21 30 2
5 CBN cannabinol 310.440 21 26 2
6 CBG cannabigerol 316.488 21 32 2
7 CBC cannabichromene 314.472 21 30 2
8 CBL cannabicyclol 314.472 21 30 2
9 THCA trans-(2 )-delta-9-tetra-
hydrocannabinolic acid A
358.482 22 30 4
10 THCA-C4 trans-(2 )-delta-9-tetra-
C4 (C4-isomer of THCA)
344.455 21 28 4
11 THVA trans-(2 )-delta-9-tetra-
(C3-isomer of THCA)
330.428 20 26 4
12 CBDA cannabidiolic acid 358.482 22 30 4
13 CBNA cannabinolic acid 354.450 22 26 4
14 CBGA cannabigerolic acid 360.498 22 32 4
15 CBCA cannabichromenic acid 358.482 22 30 4
16 CBLA cannabicyclolic acid 358.482 22 30 4
17 11-OH-THC 11-hydroxy-tetrahydrocan-
nabinol (metabolite of
330.471 21 30 3
18 THC-COOH 11-carboxy-tetrahydrocan-
nabinol (metabolite of
344.455 21 28 4
A. Hazekamp et al.2364
and mass spectra databases.
All organic solvents (analytical
or HPLC reagent grade) were purchased from J.T. Baker (Deventer,
The Netherlands) or from Fluka Chemie (Buchs, Switzerland).
Thin Layer Chromatography (TLC)
Samples in ethanol were spotted on 10 20 cm silica plates. Two different
TLC systems were used. For the non-polar system, reversed phase (C
silicagel plates F254 No. 105559 (Merck, Darmstadt, Germany) were used
with methanol/5% acetic acid 19 : 1 (v/v) as the eluent. For the polar
system, normal phase silicagel plates F254 No. 105554 (Merck, Darmstadt,
Germany) were used with chloroform/methanol 19 : 1 (v/v) as the eluent.
Plates were developed in saturated normal chambers (saturation time 15
minutes). Absorption of chromatographic spots was evaluated under UV
254 nm. General visualisation of compounds was done by spraying with
modiﬁed anisaldehyde-sulphuric acid spray reagent.
For selective visual-
isation of cannabinoids, the TLC plate was sprayed with 0.5% fast blue B
salt (o-dianisidine-bis-(diazotized)-zinc double salt) (Sigma) in water,
followed by 0.1 M NaOH.
Gas Chromatography-Mass Spectrometry (GC-MS)
To obtain GC retention times, molecular weights, and fragmentation spectra
of cannabinoids, GC-MS analyses were performed on a Varian 3800 gas
chromatograph, coupled to a Varian Saturn 2000 GC/MS apparatus. The
system was controlled with Varian Saturn GC/MS workstation version 5.2
software. The GC was ﬁtted with two different types of columns; a
Durabond fused silica capillary column (30 m 0.25 mm inner diameter)
coated with DB-1 at a ﬁlm thickness of 0.1 mm, and a similar column,
coated with HP-50þ at a ﬁlm thickness of 0.15 mm (J&W Scientiﬁc Inc.,
Rancho Cordova, CA). The oven temperature was programmed from 1008C
to 2808C at a rate of 108C/min. The oven was then kept at 2808C until the
end of a 30 min run time. The injector and detector port temperatures were
maintained at 2808 C and 2908C, respectively. Helium was used as the
carrier gas at a pressure of 65 kPa. The injection split ratio was 1/50.
Elution time of D
-THC was used as internal reference to determine the
relative retention times of all cannabinoids.
High-Performance Liquid Chromatography (HPLC)
with Diode-Array and Fluorescence Detection
The HPLC proﬁles were acquired on an Agilent 1100 series HPLC, consisting
of a G1322A solvent degasser, a G1311A quaternary solvent pump, and a
Chromatographic and Spectroscopic Data of Cannabinoids 2365
G1313A autosampler. The column was kept at constant temperature by using a
G1316A column oven. The analytical column was a Waters XTerra MS C
(2.1 150 mm, 3.5 mm) ﬁtted with a XTerra MS C
(2.1 10 mm,
3.5 mm) guard column. Light absorption and emission were detected by a
G1315B UV-diode array detector (DAD) and a G1321A ﬂuorescence
detector (FLD). The system was controlled through a Vectra VL 420 DT
computer equipped with Agilent A09.01 software. UV-spectra were
measured on-line by DAD in the range of 195 –400 nm with a slit of 2 nm.
Fluorescence (FL) spectra were recorded on-line by the FLD in the range of
280–650 nm with a step of 5 nm after excitation at 222 nm. Retention times
were expressed as relative to D
DAD and FLD data of cannabinoids were recorded under acidic con-
ditions, with a mobile phase consisting of a mixture of methanol-water
containing 25 mM of formic acid (pH + 3). The proportion of methanol
was linearly increased from 65 to 100% over 25 minutes, and then kept
constant for 3 minutes. Analysis under basic conditions was obtained with a
mobile phase consisting of a mixture of acetonitrile-phosphate buffer
(10 mM, pH 7.5). The acetonitrile concentration was increased from 40 to
100% in 25 minutes, and then kept constant for 3 minutes. After each run
the column was re-equilibrated under initial conditions for 10 minutes. The
ﬂow rate was 0.3 mL/min and the total run time was 38 minutes. All determi-
nations were carried out at 308C.
Spectrophotometric Analysis (Extinction Coefﬁcients)
Cannabinoids that were available as calibrated certiﬁed standards were diluted
to a concentration of 0.01 mg/ mL in ethanol to determine molar extinction
coefﬁcients in the range of 200 to 400 nm. A blank measurement was
obtained with ethanol. UV-spectra were recorded using a Varian Cary 1 Bio
UV-Visible spectrophotometer controlled by Cary 1/3E system software,
version 3.02. A sample cell of 10 mm was used for all measurements.
Infrared Spectroscopy (IR)
Infrared spectra of cannabinoids that were available in sufﬁcient amounts were
measured using a Perkin Elmer paragon 1000PC FT-IR instrument, which was
controlled by Perkin Elmer spectrum IR V2.00 software. Concentrated
ethanolic solutions of the cannabinoids (25 mL) were mixed with ﬁnely
ground KBr (Merck, IR-grade), and ethanol was evaporated under vacuum
for 10 minutes. After proper calibration of the apparatus, IR-spectra were
measured as an average of 4 scans in the wave number range of 500 to
. After acquisition, the spectra were smoothened by using the
A. Hazekamp et al.2366
RESULTS AND DISCUSSION
Spectroscopic and chromatographic data is shown for 18 different cannabi-
noids that were available to us. However, not all cannabinoids were
available in large enough quantities to obtain exploitable data in all
analyses that were carried out. Therefore, the presented data is not complete
for all cannabinoids.
By using two TLC-systems (polar and apolar system) in combination with fast
blue B spray reagent, it was possible to distinguish and detect all tested
compounds. The Rf-values of the cannabinoids in both TLC-systems and
their spot colour after spraying with fast blue B are shown in Table 2. The
use of fast blue B as a selective detection reagent for cannabinoids
results in differently coloured spots for some compounds. Unfortunately,
these colours also depend on the concentration of the substance and on the
presence of interfering compounds, the results must be, therefore, considered
with caution. Nevertheless, we found that fast blue B was more sensitive for
detection of cannabinoid spots than UV-detection under 254 nm. For example,
the detection limit for D
-THC was 0.5 mg/mL (2 mL spotted) with UV-
detection under 254 nm, and around 0.002 mg/mL with fast blue B detection.
The main advantages of TLC are its ability to detect all spotted
compounds, while analysing several samples simultaneously under identical
conditions within a short timeframe. Lack in selectivity can sometimes be
Table 2. Relative retention (Rf) values of the cannabinoids in a polar (silica-gel) and
) TLC-system. The colours of chromatographic spots after spraying with
the cannabinoid-selective spray reagent fast blue B (FBB) are indicated
Nonpolar TLC system
(RP-18) Polar TLC system (silica)
Cannabinoid Rf-value Cannabinoid Rf-value
Red CBDA 0.68 D
Brown CBGA 0.67 D
Orange-brown CBG 0.59 CBD 0.64
red-brown CBD 0.58 CBN 0.62
Purple CBN 0.48 CBG 0.61
-THC 0.44 CBC 0.58
-THC 0.43 THCA 0.39
Red THCA 0.40 CBDA 0.37
Purple CBC 0.37 CBGA 0.31
Purple CBCA 0.35 CBCA 0.25
Chromatographic and Spectroscopic Data of Cannabinoids 2367
overcome by the use of selective detection reagents. However, in the case of
cannabinoids it seems impossible to obtain a good separation with positive
identiﬁcation of all cannabinoids when complex mixtures (e.g., plant
extracts) are analysed. Several TLC systems are therefore needed for
tentative identiﬁcation. For instance, CBDA and CBGA or CBD and CBG,
which were not separated in the non polar system, could be distinguished
when using silica as stationary phase. On the other hand, D
-THC were found to co-elute on both systems (see Table 2). In conclusion,
TLC is very useful to rapidly screen many samples for the presence of canna-
binoids in crude plant extracts, or in eluting fractions collected after prepara-
tive chromatography. However, reproducibility of TLC depends on several
parameters, e.g., relative humidity. Compared to other separation methods,
TLC performances are also very low. Consequently, unequivocal identiﬁ-
cation of cannabinoids spots requires further methods.
Two different capillary column phases were used for GC analysis (HP-50þ
and DB-1). The HP-50þ column was a medium-polar column, resulting in
relatively longer retention times compared to the nonpolar DB-1 column.
Simultaneous injection on both columns enables the distinction of all tested
cannabinoids. Retention times (relative to D
-THC) of the analysed cannabi-
noids are shown in Table 3. All cannabinoids eluted well after other major
cannabis components like terpenoids.
Because no derivatization was used in our case, the mass-spectra obtained
by GC-MS (Figure 2) are similar for the acids and their corresponding neutral
Table 3. Relative retention time (RRT) of
cannabinoids in GC using a non-polar (DB-1)
and medium-polar (HP-50) column
GC column type
Cannabinoid RRT RRT
THV 0.885 0.902
CBL 0.922 0.907
CBD 0.942 0.935
THC-C4 0.942 0.948
CBC 0.956 0.924
-THC 0.988 0.981
-THC 1 1
CBG 1.026 1.012
CBN 1.033 1.046
A. Hazekamp et al.2368
Figure 2. Mass-spectra in the range of M/Z 50-335 obtained by LC-MS.
Chromatographic and Spectroscopic Data of Cannabinoids 2369
Figure 2. Continued.
A. Hazekamp et al.2370
cannabinoids (e.g., THCA and D
-THC). Although CBD is structurally quite
distinct from CBC and CBL, these three cannabinoids nonetheless show
similar MS spectra (compare spectra of Figure 2) with identical base peak
(m/z ¼ 231) and molecular ion (m/z ¼ 314). Also their retention times in
GC were quite similar (Table 3), but their separation is good enough to dis-
tinguish them. Cannabidiol differs from CBC and CBL with one signiﬁcant
fragment at m/z ¼ 246. A retro-Diels-Alder reaction accounts for the
formation of the minor ion at m/z ¼ 246. Subsequent loss of a methyl
fragment results in a contribution to the ion at m/z ¼ 231.
As can be
seen in Figure 2, the base peak of all tested cannabinoids (except D
doesn’t correspond to the molecular ion, but to a fragment, indicating that
these cannabinoids are easily fragmented by GC-MS.
In the absence of derivatization, the high temperature that is applied in
GC causes the decarboxylation of acidic cannabinoids to their corresponding
Since the cannabis plant mainly contains the (carboxylic-)
acidic forms of cannabinoids,
GC analysis is not the method of choice
to establish the metabolic proﬁle of a cannabinoid sample. To avoid
decarboxylation, the acids must be derivatized, e.g., by silylation or
formation of the alkylboronates.
However, a 100% derivatization yield
is difﬁcult to obtain. Moreover, we believe that thermo-degradation
(oxidation, isomerization) of cannabinoids in the injector port and column
may also occur. In the case of D
-THC, a signiﬁcant amount of D
and CBN was detected in the GC-chromatogram, whereas other analyses
(HPLC, NMR, TLC) did not show these compounds, which are known degra-
dation products of D
-THC (data not shown). Despite these problems associ-
ated with GC, it remains a very useful method for the analysis of
HPLC with UV/FLD Detection
With gradient-elution, most cannabinoids were base-line separated as sharp
peaks with excellent peak purity level, yielding fully exploitable UV and
ﬂuorescence spectra. The retention times of cannabinoids (relative to
-THC) are shown in Table 4. It is interesting to note that the relative
elution time of the acidic cannabinoids can be inﬂuenced by changing the
pH of the eluent, while the order of elution for the neutral cannabinoids
remains the same.
Notwithstanding these pH differences, the elution
order of THCCOOH and THC was not modiﬁed. In this way, overlap
between chromatographic peaks of acid and neutral cannabinoids can be
decreased by changing the elution pH.
Although the UV-spectra of the analysed cannabinoids (Figure 3a) were
left unchanged when the pH was changed from 3.0 to 7.5, the FL-spectra differ
drastically (Figure 3b). Acidic cannabinoids completely lose their ﬂuor-
escence under acidic conditions, while CBC has no ﬂuorescence under
Chromatographic and Spectroscopic Data of Cannabinoids 2371
basic conditions, and CBN has no ﬂuorescent properties at all. The ﬂuorescent
properties of the other analysed cannabinoids are not inﬂuenced by pH. The
UV absorption and FL yield in Figure 3a and b cannot be compared
because no standardised concentrations of the cannabinoids were used.
Standardised UV-spectra were obtained using a spectrophotometer (see
below and Figure 4).
In some cases, partially unresolved peaks could not be identiﬁed because
their UV and ﬂuorescence spectra were identical. This can be seen with Table 4
and in Figure 3a and b with CBD and CBG or D
-THC and D
-THC, which are
characterised by very close retention times and identical UV and ﬂuorescence
The chromophore of the cannabinoids corresponds to its substituted
phenolic ring, as this is a common structural element among the tested canna-
binoids. The UV spectrum of D
-THC is identical to that of olivetol, which
shows the same phenolic ring structure and is the precursor of D
and the other cannabinoids. The alkyl-side chain does not inﬂuence the
UV-absorbance, as there is no difference between THCA (C
and THVA (C
-side chain). The cyclization of the non-phenolic part of the
cannabinoids also has no inﬂuence on the absorbance, except when another
aromatic ring (CBN, CBNA) or a conjugated double bond (CBC, CBCA) is
Table 4. Relative retention time (RRT) of cannabinoids in HPLC using a
reversed phase column (C
) and eluent with a basic (7.5) or acidic (3) pH
Acidic HPLC system Basic HPLC system
Cannabinoid RRT Cannabinoid RRT
11-OH-THC 0.70 THC-COOH 0.26
THC-COOH 0.76 CBDA 0.34
CBD 0.76 THVA 0.36
THV 0.77 CBGA 0.40
CBG 0.78 THCA-C4 0.42
CBDA 0.82 CBNA 0.50
CBGA 0.92 THCA-A 0.51
CBN 0.93 CBLA 0.53
-THC 1.00 CBCA 0.61
-THC 1.03 CBD 0.83
THVA 1.04 CBG 0.83
CBC 1.12 CBN 0.95
THCA-C4 1.13 D
CBNA 1.21 D
THCA-A 1.25 CBC 1.08
CBLA 1.32 11-OH-THC 1.31
A. Hazekamp et al.2372
Figure 3. (a) UV-spectra in the range of 190–400 nm obtained in two HPLC-systems
with acidic and basic pH. (b) Fluorescence spectra in the range of 280 –650 nm
obtained in two HPLC-systems with acidic and basic pH.
Chromatographic and Spectroscopic Data of Cannabinoids 2373
Figure 3. Continued.
A. Hazekamp et al.2374
Figure 3. Continued.
Chromatographic and Spectroscopic Data of Cannabinoids 2375
In the case of HPLC peak overlap, the use of MS-detection coupled to
HPLC (LC-MS) and furthermore, LC-MS-MS can provide better clues about
cannabinoid structure and identity. In the acid system (pH ¼ 3), formic acid
was used to make the eluent compatible with mass spectrometry. In contrast
to HPLC-DAD or Fl, which are carried out at room temperature, LC-MS
with ionspray ionisation at a relatively high temperature (e.g., 5008 C) may
result in partial thermal decomposition of acid cannabinoids. An example of
a LC-MS separation of a large array of THC metabolites in body ﬂuids at a con-
centration of 50 ng/mL is shown in Figure 6. For separation, we used a Waters
microbore column. In contrast to GC-MS operating in the EI mode,
the mass spectra are very simple with one prominent [MH]
molecular ion and very little fragmentation. For better sensitivity, the data were
recorded in the Selected Ion Monitoring (SIM) mode. Except THC
¼ 315.2), all cannabinoids were measured in the negative ionisation
mode. The monohydroxylated (8
-OH- and 11-OH-THC) and dihydroxylated
-11-diOH THC metabolites were well resolved from the acid inactive metab-
olite (THCCOOH) and its conjugated derivative (THCCOOH-glucuronide).
Spectrophotometric Analysis (Extinction Coefﬁcients)
Very few UV-absorption spectra of calibrated cannabinoids are given in the
They are generally characterised by a few parameters
Figure 3. Continued.
A. Hazekamp et al.2376
(maxima and minima, shoulders of the UV spectra). The extinction coefﬁ-
cients are very seldom presented. Because most cannabinoids differ in their
UV with several absorption peaks, many wavelengths can be selected for
quantiﬁcation. Figure 5 shows that absorption generally decreases with
increasing wavelength. A better sensitivity can be obtained in the low
200–210 nm range, while selecting a higher wavelength will increase the
selectivity by diminishing the risk of measuring interfering compounds. In
order to get a rough estimate on the concentration of cannabinoids from any
selected wavelength, a spectrum measured at 0.01 mg/ mL between 200 and
400 nm is presented for 7 major cannabinoids (Figure 5). The extinction coef-
ﬁcients (1) at 3 different maxima are also indicated.
Infrared Spectroscopy (IR)
Infrared spectroscopy has been a common tool for the identiﬁcation and
structure elucidation of cannabinoids and derivatives in isolation and
synthesis experiments. As with UV-spectra, usually IR-spectra are reported
by presenting a few maximum absorbance peaks only. Obviously, reported
IR-spectra have been measured with a large variety of IR-spectrometers. In
Figure 4. Chromatogram of a separation and identiﬁcation of cannabinoid metab-
olites from human blood in a single chromatographic run, by using LC-MS. All canna-
binoids can be identiﬁed because of the high selectivity of the mass-detector. The top
chromatogram shows the total ion current (TIC).
Chromatographic and Spectroscopic Data of Cannabinoids 2377
Figure 5. Extinction coefﬁcients in the range of 200 –400 nm at a concentration of
0.01 mg/mL in ethanol. Absorption values at maxima or shoulders are indicated.
A. Hazekamp et al.2378
this report (Figure 6) we present the full IR-spectra of 8 common natural can-
nabinoids measured on a single modern FT-IR-spectrometer.
A growing interest in Cannabis as a source of medicinal compounds has
emerged during the last few years. Several crude preparations or synthetic
Figure 6. IR-spectra in the range of 500– 4000 cm
obtained by Fourier-transform
Chromatographic and Spectroscopic Data of Cannabinoids 2379
drugs derived from Cannabis are under development, or in the clinical
pipeline for distribution on the market. For carrying out all these investi-
gations, pharmacologically pure cannabinoids must be available in large
quantities. Reference compounds for analytical research must also be
present. Chromatographic and spectroscopic data are, therefore, a prerequi-
site for their determination and identiﬁcation.
The analytical data presented here makes it possible to positively identify
the major cannabinoids found in the Cannabis plant. Presenting all analytical
parameters measured under standardised conditions should facilitate the
Figure 6. Continued.
A. Hazekamp et al.2380
identiﬁcation of cannabinoids isolated from, or present in, Cannabis prep-
arations. Unequivocal identiﬁcation of cannabinoids cannot totally rely on
only one of the tested methods because confusion of some common cannabi-
noids always remains possible.
Finally, we believe that the use of LC-MS, and especially LC-MS-MS,
should make it possible to identify all tested cannabinoids in one single
analysis, even in the low ng/mL concentration range.
Mr Pascal Cardinal, chemist, University of Alberta, Canada, is thanked for his
fruitful discussion. The grower of certiﬁed cannabis plants, Bedrocan BV, The
Netherlands, is acknowledged for providing cannabis plant material.
1. Baker, D.; Pryce, G.; Giovannoni, G.; Thompson, A.J. Lancet Neurol. 2003, 2,
2. Turner, C.E.; ElSohly, M.A.; Boeren, E.G. J. Nat. Prod. 1980, 43, 169–234.
3. Williamson, E.M.; Evans, F.J. Drugs 2000, 60, 1303 –1314.
4. Turner, C.E. Marijuana and cannabis: research. Why the conﬂict. In Marihuana’ 84.
Proceedings of the Oxford Symposium on Cannabis; Harvey, D.J., Ed.; IRL Press:
Oxford, 1985; 31 –36.
5. Mechoulam, R. Marijuana Chem. 1970, 168, 1159–1166.
6. Joyce, C.R.B.; Curry, S.H. The Botany and Chemistry of Cannabis;J.&A.
Churchill: London, 1970.
7. Nahas, G.G. Marihuana-Deceptive Weed; Raven Press: N.Y.; North-Holland
Publishing Company: Amsterdam, 1973.
8. Ross, S.; ElSohly, M.A. Zagazig J. Pharmaceut. Sci. 1995, 4, 1– 10.
9. Grotenhermen, F. Review of therapautic effects. In Cannabis and Cannabinoids;
Grotenhermen, F., Russo, E., Eds.; Haworth Press: New York, 2003; 123– 132.
10. Zoller, O.; Rhyn, P.; Zimmerli, B. J. Chromatogr. A 2000, 872, 101–110.
11. Hazekamp, A.; Simons, R.; Peltenburg-Looman, A.; Sengers, M.; van Zweden, R.;
Verpoorte, R. Preparative isolation of cannabinoids from Cannabis sativa by
centrifugal partition chromatography. J. Liq. Chromatogr. & Rel. Technol. 2004,
27, 2421 –2439.
12. Hazekamp, A.; Choi, Y.H.; Verpoorte, R. Quantitative analysis of cannabinoids
from Cannabis sativa using 1H-NMR. Chem. Pharm. Bull. 2004, 52 (6), 718–721.
13. Lehmann, T.; Brenneisen, R. J. Liq. Chromatogr. & Rel. Technol. 1995, 18,
14. Brenneisen, R.; ElSohly, M.A. J. For. Sci. 1988, 33 (6), 1385–1404.
15. Harvey, D.J. Cannabinoids. In Mass Spectrometry: Clinical and Biomedical
Applications; Desiderio, D.M., Ed.; Plenum Press: New York, 1992; Vol. 1,
16. Pragst, F.; Herzler, M.; Herre, S.; Erxleben, B.T.; Rothe, M. UV Spectra of Toxic
Compounds. Database of Photodiode Array UV Spectra of Illegal and Therapeutic
Chromatographic and Spectroscopic Data of Cannabinoids 2381
Drugs, Pesticides, Ecotoxic Substances and Other Poisons; Verlag Dr. Dieter
Helm: Heppenheim, Germany, 2001.
17. Pﬂeger, K.; Maurer, H.H.; Weber, A. Electronic Database: Mass Spectral and GC
Data of Drugs, Poisons, Pesticides, Pollutants and their Metabolites. Parts 1– 4,
2nd Ed.; Wiley-VCH: Weinheim, Germany, 2000, Mass Spectral Libraries, Rev.
D.02.00. Agilent Technologies.
18. Electronic Database: Wiley Registry of Mass Spectral Data, 7th Ed.; John Wiley &
Sons, Inc.: and Wiley Mass Spectral Libraries, 2000, Rev. D.02.00. Agilent Tech-
19. Harvey, D.J. Biomed. Mass Spectrom. 1977, 4, 88 –93.
20. Corrigan, D.; Lynch, J.J. Planta Med. 1980, 163 –169.
21. Raharjo, T.J.; Verpoorte, R. Phytochem. Anal. 2004, 15 (2), 79 –94.
22. Shoyama, Y.; Yagi, M.; Nishioka, I.; Yamauchi, T. Phytochemistry 1975, 14,
23. Turner, J.C.; Mahlberg, P.G. J. Chromatogr. 1982, 253, 295–303.
Received October 1, 2004
Accepted March 30, 2005
A. Hazekamp et al.2382