Arch Pharm Res Vol 28, No 9, 1086-1091, 2005
Simultaneous Determination of Cannabidiol, Cannabinol, and
∆9-Tetrahydrocannabinol in Human Hair by Gas Chromatogra-
Jin Young Kim1,2, Sung Ill Suh1, Moon Kyo In1, Ki-Jung Paeng2, and Bong Chul Chung3
1Drug Analysis Laboratory, Forensic Science Division, Supreme Prosecutors’ Office, Seocho-dong, Seocho-gu,
Seoul 137-730, Korea, 2Department of Chemistry, Graduate School, Yonsei University, Seoul 120-749, Korea, and
3Bioanalysis and Biotransformation Research Center, Korea Institute of Science and Technology, Seoul 130-650,
(Received May 17, 2005)
An analytical method was developed for evaluating the cannabidiol (CBD), cannabinol (CBN),
and ∆9-tetrahydrocannabinol (∆9-THC) level in human hair using gas chromatography-mass
spectrometry (GC-MS). Hair samples (50 mg) were washed with isopropyl alcohol and cut into
small fragments (< 1 mm). After adding a deuterated internal standard, the hair samples were
incubated in 1.0 M NaOH for 10 min at 95°C. The analytes from the resulting hydrolyzed sam-
ples were extracted using a mixture of n-hexane-ethyl acetate (75:25, v/v). The extracts were
then evaporated, derivatized, and injected into the GC-MS. The recovery ranges of CBD, CBN,
and ∆9-THC at three concentration levels were 37.9-94.5% with good correlation coefficients (r2
>0.9989). The intra-day precision and accuracy ranged from -9.4% to 17.7%, and the inter-day
precision and accuracy ranged from -15.5% to 14.5%, respectively. The limits of detection
(LOD) for CBD, CBN, and ∆9-THC were 0.005, 0.002, and 0.006 ng/mg, respectively. The
applicability of this method of analyzing the hair samples from cannabis abusers was demon-
Key words: GC-MS, Hair analysis, Cannabidiol, Cannabinol, Tetrahydrocannabinol
The hemp plant Cannabis sativa is one of the most
widely abused illicit drugs in Korea (Supreme Prosecutors'
Office, 2004). A dry, pulverized green and/or brown mix of
the flowers and leaves of Cannabis sativa is usually
smoked either as a cigarette or in a pipe. Cannabis
contains more than 420 chemicals including at least 61
cannabinoids (Turner et al., 1980). Although cannabis
contains many cannabinoid components, ∆9-tetrahydro-
cannabinol (∆9-THC) is the most prominent psychoactive
cannabinoid constituent, and the potency of marijuana is
determined by its concentration . Cannabidiol (CBD) and
cannabinol (CBN), like ∆9-THC, are constituents that can
be isolated from both Cannabis sativa and cannabis
smoke (Staub et al., 1999).
∆9-THC undergoes extensive metabolism in humans
into 11-hydroxy- and 8-hydroxy-∆9-tetrahydrocannabinol
and finally to 11-nor-∆9-tetrahydrocannabinol-9-carboxylic
acid (THCCOOH), which is excreted in the form of
glucuronide conjugates with several different conjugated
species (Wall et al., 1970; Agurell et al., 1976; Hawks et
Identifying CBD, CBN, and ∆9-THC in decontaminated
hair indicates exposure to cannabis (Cirimele et al., 1996;
Baptista et al., 2002; Musshoff et al., 2002; Uhl and Sachs,
2004). Besides the parent drug, ∆9-THC, determining the
level of the main metabolite, THCCOOH, is recommended
for distinguishing passive exposure from active, intentional
ingestion. In contrast, determining the THCCOOH level
can only be performed in a separate examination using
gas chromatography-tandem mass spectrometry (GC-
MS-MS), GC-MS-negative ion chemical ionization (NCI)
with a high-volume injection or an additional clean up
before GC-MS-NCI (Mieczkowski, 1995; Uhl, 1997; Sachs
Correspondence to: Jin Young Kim, Drug Analysis Laboratory,
Forensic Science Division, Supreme Prosecutors’ Office, Seocho-
dong, Seocho-gu, Seoul 137-730, Korea
Tel: 82-2-535-4173, Fax: 82-2-535-4175
The GC-MS Analysis of CBD, CBN, and THC in Human Hair1087
and Dressler, 2000; Moore et al., 2001; Uhl and Sachs,
2004). However, these latter analytical methods are time
consuming and technically expensive. Therefore, an
analysis of hair to identify cannabis use is often restricted
to the identification of CBD, CBN, and ∆9-THC, which can
be detected by routine gas chromatography-mass spec-
The aims of this study were to establish and validate a
GC-MS method for the simultaneous determination of
CBD, CBN, and ∆9-THC in human hair, and to investigate
the chemical stability of these analytes in the alkaline
medium used for extracting the drugs from the matrix.
This GC-MS method was successfully applied to the
analysis of CBD, CBN, and ∆9-THC in the hair samples of
MATERIALS AND METHODS
The CBN and CBD were purchased from Sigma Israel
Chemicals LTD. (Jerusalem, Israel). The ∆9-THC and ∆9-
tetrahydrocannabinol-d3 (∆9-THC-d3) were purchased
from Cerilliant (Austin, TX, U.S.A.) in vials at a concen-
tration of 1000 and 100 µg/mL in methanol, respectively.
The methanol, n-hexane, isopropyl alcohol, and ethyl
acetate were purchased from J. T. Baker (Phillipsburg,
NJ, U.S.A.). All solvents were of high performance liquid
chromatography (HPLC) grade. The N-methyl-N-trimethyl-
silyltrifluoroacetamide (MSTFA), trimethylchlorosilane
(TMCS), and N-trimethylsilylimidazole (TMSI) were pur-
chased from Sigma-Aldrich Corp. (St. Louis, MO, U.S.A.).
Preparation of solutions
The CBD and CBN (10 mg) were dissolved in methanol
(10 mL) to prepare a stock solution. Subsequently, 1, 10,
and 100 µg/mL CBD and CBN solutions were prepared
by dilution. Working standards solutions of ∆9-THC (1 µg/
mL, 10 µg/mL) and ∆9-THC-d3 (1 µg/mL) were prepared
using the appropriate dilution with methanol. All of these
solutions were stored at -20oC in the absence of light until
The stability of CBD, CBN, and ∆9-THC under alkaline
digestion conditions at 95oC for 10 min was determined
by measuring the recovery of the analytes after digestion
using 1 mL of a 1.0 M NaOH solution spiked with each
analyte at three concentration levels (10, 75, and 200 ng/
mL) in five replicate experiments.
Preparation of Materials
The drug-free human hair was obtained from a 36-year-
old male volunteer, and was used as the control and
blank matrix for preparing the matrix-matched calibration
solutions for CBD, CBN, and ∆9-THC. Hair samples (n =
22) from possible cannabis abusers were included in the
analysis among the suspected abusers’ samples received
from the Narcotics Department at the Seoul District
Prosecutors' Office between June and November 2004.
These samples, weighing more than 30 mg, were
generally pulled out or cut as close as possible to the skin
from the posterior vertex. The total length of the hair
samples was measured, and any special treatment
features were noted. The samples were stored under dry
conditions at room temperature.
The hair samples were washed three times with 10 mL
isopropyl alcohol, dried in a fume hood, cut with scissors
into small fragments (<1 mm), and weighed. A hair
sample (50 mg) was then transferred to a silanized glass
test tube (16 × 100 mm) containing 50 ng of ∆9-THC-d3 as
an internal standard. The analytes were isolated from the
protein matrix via consecutive hydrolysis in a 1.0 M NaOH
solution at 95oC for 10 min in. After cooling, 5 mL n-
hexane-ethyl acetate (75:25, v/v) was added to the
incubation mixture and the analytes were extracted by
mechanical shaking for 20 min at approximately 60 cycles
per minute (cpm) and centrifuged for 5 min at 2800 × g.
The organic phase was transferred to a silanized test tube
(12 × 100 mm) and evaporated under a nitrogen stream
using a TurboVap LV evaporator. The residue was dried in
a vacuum desiccator over P2O5-KOH for at least 30 min.
The trimethylsilyl derivatives were formed by a reaction
with 40 µL of MSTFA-TMCS-TMSI (100:2:5, v/v/v) in a dry
heating block at 60oC for 15 min. An aliquot (1 µL) of the
sample solution was injected into the GC-MS.
Gas chromatography-mass spectrometry
GC-MS analysis was performed using an Agilent
Technologies 6890N GC (Palo Alto, CA, U.S.A.) coupled to
a 5973N Mass Selective Detector. The system was con-
trolled by Drug Analysis Chemstation G1701CA software
(C.00.00, Agilent Technologies). The gas chromatograph
was equipped with a capillary column (Ultra-1, 25 m × 0.20
mm i.d., 0.33 µm, J&W Scientific, Folsom, CA, U.S.A.).
The GC temperature program used is as follows: initial
temperature was 180oC, held for 0.5 min, increased to
240oC at a rate of 5 oC/min, held for 2.5 min, and then
increased to 300oC at a rate of 25 oC/min. The injector
temperature was 260oC and a splitless injection was used
with a split-valve off-time of 0.5 min. The flow rate of the
carrier gas (helium) was 1.0 mL/min. The mass spectrom-
eter was operated at 70 eV in the electron impact mode
with selected ion monitoring (SIM) for quantification. The
selected ion groups of the derivatized analytes monitored
1088 J. Y. Kim et al.
are as follows: m/z 390 and 301 (CBD); m/z 367 and 368
(CBN); m/z 386 and 303 (∆9-THC); m/z 389 (∆9-THC-d3).
Validation of analytical method
In order to evaluate the specificity of the method, a
blank hair sample was prepared, analyzed, and examined
for its response in each of the analyte and internal
standard (IS) chromatographic profiles. The linearity of the
method for CBD, CBN, and ∆9-THC was examined in the
concentration range 0.05-5.0 ng/mg. A six-point calibration
curve was established at each concentration using 50 mg
samples of drug-free human hair spiked with methanolic
solutions of analytes. The internal standard versus analyte
concentrations was determined using linear regression
analysis of the peak area ratios of the analyte. The
percentage accuracy was determined by dividing the
measured concentration by the theoretical concentration
and multiplying this result by 100. The inter-day and intra-
day validation was performed by calculating the precision
and accuracy for four concentrations (0.05, 0.2, 1.5, and
4.0 ng/mg) corresponding to the quality control (QC)
samples of the method. The limits of detection (LODs) for
CBD, CBN, and ∆9-THC were calculated based on the
injected mass giving a signal that was three times the
peak-to-peak noise of the background signal. The limit of
quantification (LOQ) was defined as the lowest concen-
tration that could be determined with a coefficient of
variation (CV) < 20%. The extraction recoveries of each
analyte were quantified at three concentrations (0.2, 1.5,
and 4.0 ng/mg) of five replicates.
RESULTS AND DISCUSSION
Gas chromatography-mass spectrometry
The chromatographic separation of CBD, CBN, and ∆9-
THC was optimized for improving the peak shape and
retention characteristics (Fig. 1). The analytes were identified
by a retention time comparison with the reference stan-
dards. The SIM chromatograms for the analytes clearly
indicated the order of elution (Table I).
Fig. 2 shows the EI mass spectra of the trimethyl-
silylated (a) CBD, (b) CBN, (c) ∆9-THC, and (d) ∆9-THC-
d3. The molecular and fragmented ions were formed in
good abundance in all cases, and were used as the
quantification ions. A derivatization reaction was performed
to convert the polar -OH groups of the analytes into
thermally stable, non-polar groups. The EI mass spectra
of trimethylsilylated CBD, CBN, and ∆9-THC in the human
hair samples were consistent with those of the reference
Passive drug absorption on hair samples must be
considered as a possible source of any false-positive
results. Therefore, an extra step was needed to remove
external contaminants before beginning the analytical
procedure. In order to confirm that CBD, CBN, and ∆9-
THC were adsorbed on the hair surface and eluted with
washing, the positive control samples from the suspected
Fig. 1. SIM GC-MS chromatograms of CBD, CBN, and ∆9-THC in
human hair. (a) blank hair, (b) hair sample spiked at 0.5 ng/mg of
each analyte, (c) positive control sample.
Table I. Retention times and characteristic ions of trimethylsilylated
(TMS) CBD, CBN, and ∆9-THC
TMS ∆9-THC-d3 (ISd)
bRetention time relative to that of the internal standard
cThe ion used for quantification is given in bold
The GC-MS Analysis of CBD, CBN, and THC in Human Hair1089
abusers were washed using the aforementioned washing
procedure and washed again with 10 mL isopropyl alcohol.
The washing efficiency of the isopropyl alcohol wash after
the normal wash of the hair samples was evaluated. None
of the target analytes were detected in the isopropyl
alchohol wash, demonstrating that the hair-washing
procedure is sufficient for decontamination.
Stability study of analytes
The results after digestion with 1 mL of 1.0 M NaOH at
95oC for 10 min at the three concentration levels (10, 75,
and 200 ng/mL) confirmed the expectation that the stability
of the analytes might be affected during the digestion
procedure. The severe digestion conditions appeared to
affect the stability of CBD (Fig. 3). Therefore, the effect of
the digestion procedure should be considered when
interpreting the analysis of cannabinoids in hair.
Evaluation of validation data
The described method was validated by determining
the selectivity, linearity, precision and accuracy, the limit of
detection (LOD), the limit of quantification (LOQ), and
recovery. After derivatization, CBD, CBN, and ∆9-THC can
be well defined chromatographically. A comparison of the
SIM chromatograms for the blank hair sample and the
hair sample fortified with the analytes and an internal
standard confirmed there were no interfering peaks near
the retention times of the analytes in the hair sample
matrix (Fig. 1).
Fig. 2. EI mass spectra of trimethylsilylated CBD, CBN, ∆9-THC,
and ∆9-THC-d3 (internal standard).
Fig. 3. The stability of CBD, CBN, and ∆9-THC after digestion in 1
mL of 1.0 M NaOH solution at 95°C for 10 min.
Table II. Validation data for the analysis of CBD, CBN, and ∆9-THC in hu-
Concentration range (ng/mg)
0.05 – 5.00.05 – 5.00.05 – 5.0
Recovery (% CV, n = 5)
0.2 ng/mg42.2 ± 5.893.4 ± 3.484.7 ± 2.4
1.5 ng/mg37.9 ± 2.494.5 ± 1.585.9 ± 1.9
4.0 ng/mg43.6 ± 0.494.0 ± 1.193.4 ± 1.8
aLinearity is determined by the linear correlation for the calibration
bLOD : Limit of detection was defined as the concentration at which the
characteristic ion was detectable on the corresponding mass
chromatogram at S/N = 3 or greater.
cLOQ : Limit of quantification was defined as the lowest concentration
on the calibration curve with a precision < 20% (% CV).
1090J. Y. Kim et al.
Tables II and III show the summarized quantitative
validation parameters. The calibration standards at the six
concentration levels for each analyte were used to con-
struct the calibration curves. The correlation coefficients
were 0.9989-0.9995 for all the analytes. The LOQ was
defined as the lowest concentration that permitted
precision within 20% (% CV). The extraction recoveries
for CBD, CBN, and ∆9-THC were determined at three
concentration levels with five replicates. The peak-area
ratios for each analyte in the samples that had been
spiked with analytes prior to extraction were compared
with the samples to which the same levels of analytes had
been added after extraction. The recoveries of the
analytes with the exception of CBD were 84.7-94.5%. The
recovery range of CBD was 37.9-43.6%, demonstrating
that the digestion procedure can influence the stability of
the analyte during alkaline digestion.
The intra-day precision and accuracy were obtained by
analyzing three replicates of four different spiked hair
samples (0.05, 0.2, 1.5, and 4.0 ng/mg). The precision
was determined by calculating the CV of the degree of
agreement between the measured and nominal concen-
trations of the spiked samples. The inter-day precision
and accuracy were determined by analyzing the same
spiked, human hair samples on each of five consecutive
days. For the intra-day (n = 3) assays, the precision ranged
from 0.8 to 17.7% for CBD, CBN, and ∆9-THC, while the
accuracy ranged from -9.4 to 14.8%. For the inter-day (n
= 5) assays, the precision ranged from 0.8 to 12.3% for
these compounds, while the accuracy ranged from -15.5
to 14.5%. These results were considered satisfactory
considering the complexity of the biological matrix.
The method described in this report was applied to the
analysis of hair samples obtained from suspected
cannabis abusers. Fig. 1(c) shows the SIM chromatogram
for a positive control sample obtained from a confirmed
cannabis abuser. Table IV shows the quantitative results
of the analytes in hair samples. The measured CBN
concentration in the hair samples was superior to those of
either CBD or ∆9-THC. The average concentration of
CBD, CBN, and ∆9-THC was 0.04, 0.36, and 0.14 ng/mg,
In 15 of the 22 hair samples, CBN was the cannabinoid
detected most frequently both alone and in association
with other cannabinoids. This is might be due to the
pyrolitic degradation of ∆9-THC to CBN when smoked
(Strano-Rossi and Chiarotti, 1999). CBD was the second
frequently detected compound. In one hair sample, only
∆9-THC was detected and in six hair samples, only CBN
was detected (Table V).
Table III. Precisiona and accuracyb for the determination of CBD, CBN, and ∆9-THC
Intra-day (n=3)Inter-day (n=5)
Mean (ng/mg) Precision (% CV) Accuracy (% bias)Mean (ng/mg) Precision (% CV) Accuracy (% bias)
CBD 0.05 0.04517.7 -9.40.047 6.2-15.5
0.20.235.2 14.8 0.224.912.1
1.51.40 1.2-6.61.39 1.8-6.7
4.0 4.045.30.9 4.083.6 2.1
CBN 0.050.047 11.9 -5.70.046 5.9-7.3
0.2 0.198.4-4.2 0.20 9.0-1.2
4.0 3.861.2 -3.53.853.3 -3.7
0.20.20 5.3 1.5 0.207.0 -1.7
1.51.520.8 1.01.53 0.80.9
4.03.89 3.3-2.8 3.942.6-1.4
aExpressed as the coefficient of variance of the peak area ratios of the analyte/internal standard.
bCalculated as [(mean calculated concentration – nominal concentration)/nominal concentration] × 100
Table IV. Quantitative results of CBD, CBN, and ∆9-THC in hair
samples of suspected cannabis abusers
22 140.06-0.270.14 0.09
aQualified data below LOD and the lower end of the calibration range
were included in the number of positive samples.
bThe concentration value below the lower end of the calibration range
at three to four times higher than the normal amount of hair was
included in measured concentration range.
The GC-MS Analysis of CBD, CBN, and THC in Human Hair1091
The GC-MS method showed a better sensitivity, good
linearity, and lower LOD and LOQ than other methods. It
also appears useful for simultaneously determining the
presence and concentration of CBD, CBN, and ∆9-THC in
human hair. The data obtained from the hair samples
cannabis of abusers indicated that CBN was the most
frequently detected cannabinoid, followed in succession
by CBD and ∆9-THC. This study also showed that the
digestion procedure can affect the stability of analytes.
Therefore, it is essential to check the stability of the
analytes under the digestion medium when interpreting
the hair analysis results.
The authors wish to acknowledge Dr. K. M. Kim and
Ms. H. K. Ryu at the Bioanalysis and Biotransformation
Research Center, Korea Institute of Science and
Technology for their excellent technical assistance.
Agurell, S., Nilsson, I. M., and Widman, M., Cannabinoids
metabolites hydroxylated in the pentyl side chain: In Nahas,
G. (Ed.), Marihuana Chemistry, Biochemistry and Cellular
Effects, Springer-Verlag, New York, pp. 141-158, (1976).
Baptista, M. J., Monsanto, P. V., Pinho Marques, E. G., Bermejo,
A., Ávila, S., Castanheira, A. M., Margalho, C., Barroso, M.,
and Vieira, D. N., Hair analysis for ∆9-THC, ∆9-THC-COOH,
CBN and CBD by GC/MS-EI. Comparison with GC/MS-NCI
for ∆9-THC-COOH. Forensic Sci. Int., 128, 66-78 (2002).
Cirimele, V., Sachs, H., Kintz, P., and Mangin, P., Testing human
hair for Cannabis. III. Rapid screening procedure for the
simultaneous identification of ∆9-tetrahydrocannabinol, can-
nabinol, and cannabidiol. J. Anal. Toxicol., 20, 13-16 (1996).
Hawks, R. L., The constituents of cannabis and the disposition
and metabolism of cannabinoids: In Hawks, R. L. (Ed.). The
analysis of cannabinoids in biological fluids, NIDA Research
Monograph Series 42. National Institute on Drug Abuse,
Rockville, pp. 125-137, (1982).
Mieczkowski, T., A research note: The outcome of GC/MS/MS
confirmation of hair assays on 93 cannabinoid (+) cases.
Forensic Sci. Int., 70, 83-91 (1995).
Moore, C., Guzaldo, F., and Donahue, T., The determination of
11-nor-∆9-tetrahydrocannabinol-9-carboxylic acid (THC-
COOH) in hair using negative ion gas chromatography-mass
spectrometry and high-volume injection. J. Anal. Toxicol., 25,
Musshoff, F., Junker, H. P., Lachenmeier, D. W., Kroener, L.,
and Madea, B., Fully automated determination of canna-
binoids in hair samples using headspace solid-phase micro-
extraction and gas chromatography-mass spectrometry. J.
Anal. Toxicol., 26, 554-560 (2002).
Sachs, H. and Dressler, U., Detection of THCCOOH in hair by
MSD-NCI after HPLC clean-up. Forensic Sci. Int., 107, 239-
Staub, C., Chromatographic procedures for determination of
cannabinoids in biological samples, with special attention to
blood and alternative matrices like hair, saliva, sweat and
meconium. J. Chromatogr. B, 733, 119-126 (1999).
Strano-Rossi, S. and Chiarotti, M., Solid-phase microextraction
for cannabinoids analysis in hair and its possible application
to other drugs. J. Anal. Toxicol., 23, 7-10 (1999).
Supreme Prosecutors' Office, White Paper on Drug-Related
Crimes (2003). SPPO, Seoul, Korea, pp. 98-122, (2004).
Turner, C. T., Elsohly, M. A., and Boeren, E. G., Constituents of
Cannabis Sativa L. XVII. A review of the natural constituents.
J. Nat. Prod., 43, 169-234 (1980).
Uhl, M., Determination of drugs in hair using GC/MS/MS.
Forensic Sci. Int., 84, 281-294 (1997).
Uhl, M. and Sachs, H., Cannabinoids in hair: strategy to prove
marijuana/hashish consumption. Forensic Sci. Int., 145, 143-
Wall, M. E., Brine, D. R., Brine, G. A., Pitt, C. G., Freudenthal, R.
I., and Christensen, H. D., Isolation, structure, and biological
activity of several metabolites of ∆9-THC. J. Am. Chem. Soc.,
92, 3466-3468 (1970).
Table V. Qualitative results from CBD, CBN, and ∆9-THC in the
twenty-two hair samples