The precise determination of tetrahydrocannabinol in marihuana and hashish


Internal Standard Method1
Some results with the internal standard method


Author: Pauline LERNER
Pages: 39 to 42
Creation Date: 1969/01/01

The precise determination of tetrahydrocannabinol in marihuana and hashish

Pauline LERNER U. S. Customs Laboratory Baltimore, Maryland 21202

Today, in many countries, there is in progress an extensive scientific study of the effects of marihuana in man. Several of these studies are utilizing for the first time marihuana and hashish which have been assayed for active tetrahydrocannabinol content, in an attempt to quantify the nature of the marihuana reaction.

The two major analytical methods for the determination of tetrahydrocannabinol in marihuana and hashish are gas chromatography (1-4) and thin-layer chromatography (5-6). In a previous publication from this laboratory (4), gas chromatography with an OV-17 (phenyl methyl silicone) column was used for the quantitative determination of the two active isomers, Δ 1-3,4-trans-tetrahydrocannabinol and Δ 6-3,4-transtetrahydrocannabinol. This method utilized the pure tetrahydrocannabinol isomers as external standards. The area under the gas chromatographic peak obtained from an injection of a marihuana extract was compared with the area obtained from an injection of a standard solution of pure tetrahydrocannabinol. This method, while rapid, is inherently imprecise, since it requires complete reproducibility of injection volumes and detector responses, which, in practice, is very difficult to obtain.

The present paper is concerned with the use of dl-methadone hydrochloride as an internal standard for the gas chromatographic determination of tetrahydrocannabinol. The use of this internal standard increases the precision of the tetrahydrocannabinol determination by a factor of 10 as compared with the use of an external standard.


An F & M Model 400 gas chromatograph equipped with a flame ionization detector and a ball and disc integrator for quantitative measurements was used. A Keithley Model 417 High Speed Picoammeter was utilized in place of the F & M electrometer. The samples were run on a 2 per cent OV-17 (phenyl methyl silicone on 100 mesh Gas-Chrom-Q [Applied Science Labora- tories]) column at 210°C. A six-foot glass column having an I.D. of 4 mm. was used, with a helium flow rate of 80 ml./minute. Retention times, in minutes, were as follows:

tetrahydrocannabinol ( Δ) 6
tetrahydrocannabinol ( Δ) 1

To establish a basis for comparison, the reproducibility of the external standard method [ (4)] was tested by alternating injections of a single marihuana solution and a Δ 1-3,4-trans-tetrahydrocannabinol standard solution. A coefficient of variation of 13 per cent was obtained; this corresponds, for the mean of 4 determinations, to an uncertainty of ±13 per cent of the determined value (95 per cent probability). At the two per cent tetrahydrocannabinol level, this would amount to ±0.26 per cent tetrahydrocannabinol.

It was believed that an internal standard method could substantially reduce this uncertainty. An internal standard has been used successfully for the quantitative gas chromatographic analysis of morphine [ (7)] . Finding an internal standard for marihuana and hashish is complicated by the large number of naturally occurring minor peaks that accompany the tetrahydrocannabinol peak. The chosen internal standard, dl-methadone hydrochloride, has a retention time that is significantly different from that of any of the naturally occurring compounds in marihuana or hashish.

Internal Standard Method 1

900 mg. of air-dried marihuana or 120 mg. hashish is placed in a stoppered test tube containing 5.2 mg. of dl-methadone hydrochloride and 30 ml. of chloroform. (Larger samples can be used, of course, provided the ratio of sample, standard, and solvent remains constant.) The test-tube is placed in the refrigerator, and shaken at 10-minute intervals for 1 hour. At the end of an hour the chloroform solution is filtered off and 5 µl. of the solution is injected into the gas chromatograph. The counts from the ball and disc-integrator, which correspond to the area under the peaks, are then obtained and inserted into the formula:

The internal standard method described in this paper has been recently adopted as the official analytical method for the marihuana research contracts administered by the National Institute of Mental Health (Personal communication, Dr. Coy Waller, University of Mississippi, Consultant to N.I.M.H.).

Per cent tetrahydrocannabinol =

(W R) C( S) (F) (100) / (W S) (C R)

Where W R=Weight of the dl-methadone hydrochloride

W S= Weight of marihuana sample

C R= Counts under methadone peak

C S= Counts under tetrahydrocannabinol peak

F= Factor obtained from running a purse sample of tetrahydrocannabinol; this factor will depend on the equipment employed.

Figure 1 is a chromatogram obtained by this method. The first large peak is due to the solvent; the second large peak is the internal standard, dl-methadone, and the third large peak is Δ 1-tetrahydrocannabinol.

FIGURE 1. A chromatogram illustrating the internal standard method. The first large peak, on the left, is due to the solvent. The second large peak is the internal standard, dl-methadone, and the third large peak is Δ 1 -trans-tetrahydrocannabinol. The areas under the peaks are obtained from the integrator trace at the bottom of the chart.

Full size image: 87 kB, FIGURE 1

The reproducibility of this method was tested with a single marihuana solution. A coefficient of variation of 1.1 per cent was obtained; this corresponds, for the mean of 4 determinations, to an uncertainty of ± 1.1 per cent of the determined value (95 per cent probability). At the 2 per cent tetrahydrocannabinol level this would amount to ±0.02 per cent tetrahydrocannabinol.

The chloroform solution was found to be stable (to within 1.0 per cent) at room temperature, under fluorescent light, for 45 minutes; at room temperature, in the dark, for 6 hours; in the refrigerator for 3 days.

Some results with the internal standard method

Sampling error - twenty portions, each weighing 900 mg., were taken from a lot of carefully mixed "manicured" marihuana. 2 The coefficient of variation, due to sampling, for the tetrahydrocannabinol content, was found to be 2.8 per cent. At the 2 per cent tetrahydrocannabinol level this would amount to an uncertainty (95 per cent probability) due to sampling of ±0.11 per cent tetrahydrocanabinol.

This uncertainty could, of course, be reduced by using larger sample weights. A 10 gramme sample, run 4 times, would have an uncertainty of ±0.04 per cent tetrahydrocannabinol (95 per cent probability) at the 2 per cent tetrahydrocannabinnol level.

Deterioration of marihuana with time and temperature.

Three samples of marihuana were kept at room temperature (approximately 24 °C), for a month and analysed for their tetrahydrocannabinol content at the beginning and end of the month. The results were as follows in per cent tetrahydrocannabinol:


Beginning of Month

End of Month


Sample A
2.41 2.29
Sample B
2.07 1.99
Sample C
2.20 2.14
- 0.06

All the differences are highly significant statistically (over 99 per cent probability).

A sample of Colombian marihuana which had been kept at room temperature since its seizure 61 months ago was analysed with the following results:


per cent


"Manicured" marihuana is marihuana from which stalks and seeds have been removed by passing through a sieve. Crude marihuana, including stalks and seeds, has about a 35 per cent yield of "manicured" marihuana. Crude marihuana from which the stalks have been removed has about a 65 per cent yield of "manicured" marihuana.

No quantitative analysis was available of the original marihuana, but a qualitative gas chromatographic curve existed which indicated a large Δ 1-trans-tetra- hydrocannabinol peak and a small cannabinol peak. The qualitative curve resembled current fresh seizures of marihuana, and one could fairly conclude that in the five year period practically all the Δ 1-tetrahydrocannabinol had been converted to cannabinol. If so, and if an average original tetrahydrocannabinol content was assumed from the appearance of the original curve and the current cannabinol content, a conversion rate of approximately 5 per cent a month would be indicated. This is in rough agreement with the loss in tetrahydrocannabinol content (3 to 5 per cent) found above for marihuana which had stood one month at room temperature. 3

The conversion of tetrahydrocannabinol to cannibinol is in accordance with the hypothesis of Todd ( [ 8] - [ 9] ) and Adams [ (10)] that dehydrogenation takes place biogenically and Levine's observation [ (11)] that old hashish has low tetrahydrocannabinol and high cannabinol content.

A sample of marihuana which contained 2.32 per cent tetrahydrocannabinol and 0.17 per cent cannabinol was held at 100 °C for one month. At the end of that time no tetrahydrocannabinol could be detected and the cannabinol content had increased slightly, to 0.40 per cent. Obviously, at this temperature, there are other mechanisms at work than the simple quantitative dehydrogenation of tetrahydrocannabinol to cannabinol.

(3)Correlation of tetrahydrocannabinol content of marihuana and pharmacological activity.

One important recent development that has demonstrated the need for a precise tetrahydrocannabinol assay is Dr. Harris Isbell's observation [ (12)] that response to tetrahydrocannabinol is statistically reproducible if the same subjects are used under the same conditions. Dr. Isbell's work was done with pure tetrahydrocannabinol. He found that 25 µg./kg. when smoked, was a threshold dose; for a 70 kg. man this would be equivalent to 1.8 mg. of tetrahydrocannabinol. A usual dose was about 5 mg., while a dose of 15 mg. produced depersonalization and hallucinations.

If these amounts are translated into a "normal" marihuana cigarette weighing 325 mg., it would appear that a tetrahydrocannabinol content of 0.5 per cent would be at the threshold level; 1.5 per cent tetrahydrocannabinol would produce the "usual" marihuana symptoms; and a 4.6 per cent tetrahydrocannabinol content would correspond to a mind-rocking depersonalization.


Dr. Olav J. Braenden of the United Nations Laboratory has informed me that to avoid deterioration, the United Nations reference samples of marihuana are kept under nitrogen, at freezer temperatures.

These percentages correlate quite well with our analyses of marihuana seized in the illicit market. Usual marihuana seizures, when manicured, fall in the 1.0 to 2.5 per cent range. A sample of marihuana called " Acapulco Gold ", which had a tetrahydrocannabinol content of 2.7 per cent, was reported to be quite psychoactive (13). West Virginia marihuana, which several marihuana users claimed to have no effect whatsoever, we found to contain 0.1 per cent tetrahydrocannabinol. (It is of interest that all mature domestic marihuana samples we have analysed, which were taken from wild stands growing in West Virginia, Kansas and Minnesota, have had low tetrahydrocannabinol contents and a high content (about 1 per cent) of the physiologically inactive cannabidiol.)


I want to thank Dr. Richard L. Hively, for his gift of pure Δ 1-3,4-trans- and Δ 6-3,4-trans-tetrahydrocannabinols, and the National Institute of Mental Health, for its gift of Δ 6-3,4-trans-tetrahydrocannabinol and cannabinol.



U. Claussen, W. Borger and F. Korte, Ann. Chem. 693 , 158 (1966).


U. Claussen and F. Korte, Tetrahedron Letters No. 22 , 2067 (1967).


M. Lerner, Science 140 , 175 (1963).


M. Lerner and J. T. Zeffert, Bulletin on Narcotics, XX , 53 (1968).


F. Korte, H. Sieper, S. Tira, Bulletin on Narcotics, XVII , 35 (1965).


C. Miras, S. Simos and J. Kiburis, Bulletin on Narcotics, XVI , 13 (1964).


E. Brochmann-Hanssen and A. B. Svendsen, J. Am. Pharm. Assoc. 55 , 1134 (1963).


A. R. Todd and A. Jacobs, J. Chem. Soc. 1940 , 649 (1940).


A. R. Todd, Scientific Journal of the Royal College of Science, 12 , 37 (1942).


R. Adams, et al., J. Am. Chem. Soc., 62 , 197 (1940).


J. Levine, J. Am. Chem. Soc ., 66, 1868(1944).


H. Isbell, C. W. Gorodetzsky, D. Jasinski, U. Claussen, F. Von G. Spulak and F. Korte, Psychopharmacologia 11 , 184 (1967).


T. H. Mikuriya, Personal Communication , 7/29/68, San Francisco, California.