Pyrolysis of cannabinoids: a model experiment in the study of cannabis smoking


Materials and methods
Results and discussion


Pages: 55 to 59
Creation Date: 1978/01/01

Pyrolysis of cannabinoids: a model experiment in the study of cannabis smoking

Laboratory of Organic Chemistry, State University, Utrecht (The Netherlands) F. ALIKARIDIS
Department of Biological Chemistry, School of Medicine, University of Athens (Greece)

Since cannabis products are generally administered by smoking, it is of importance to get a better understanding of the effects of the smoking process on cannabinoids. Several investigators have studied the conversions of cannabinoids during smoking. However, the information available remains limited and even contradictory. Several reasons can be indicated for this discrepancy. The pyrolytic products of the cannabinoids are masked by the bulk of the substrate which is smoked. On the other hand a relative change in the composition of the cannabinoids after smoking does not necessarily implicate a mutual conversion, which is, however, a conclusion generally observed in literature (1, 2). Moreover the experimental techniques used by different research groups are very diverse: except for the material which is to be smoked and which is subject to large variations (marijuana, hashish, cannabis extracts mixed with tobacco, which is often pretreated), the smoking-techniques also differ to a great extent (e.g. cigarette, pipe, water-pipe). The only common feature of these techniques is that cannabis is subjected to a thermal treatment.

This situation calls for additional, more fundamental insight into the processes taking place during the smoking of cannabinoids. It was hoped that comparison of the products formed by smoking cannabidiol (I, CBD) through a water-pipe- which is the general way of smoking by heavy cannabis users in the eastern countries- and the products formed by treatment of CBD under several pyrolytic conditions may contribute to reach this goal.

Materials and methods

The experimental set-up used for smoking through a water-pipe is shown in figure I.

The material smoked consisted of separate layers of powdered CBD (99.5 per cent; isolated from Fibrimon-21-hemp material (3) and Greek tobacco which was pre-extracted with cold water until a colourless extract was obtained. Per experiment a total of 4 grams of tobacco and 0.5 grams of CBD were used. The smoking machine was adjusted to the way of smoking by regular users (puff volume 60 ml, puff frequency 1 puff/min and puff-duration 3 sec.). Thus 3 fractions were obtained:

  1. The water soluble components;


Experimental device for water-pipe smoking

Full size image: 12 kB, FIGURE I
  1. The water insoluble products precipitated in the water-flask;

  2. The "sublimate": the smoke-fraction which is electrostatically precipitated.

Fraction 3 is considered to be the most important since this corresponds to the material which is inhaled by water-pipe smokers. Furthermore TLC and GLC revealed that the cannabinoidal components are mainly (87 per cent) concentrated in the sublimate, while a smaller (13 per cent) quantity of the cannabinoids - whose composition quantitatively differs and which contains mainly CBD - is detected in fraction 2.

In the construction of a pyrolysis device several parameters have to be considered. This is best illustrated by the experiments of EI-Darawy et al. (4) who studied the influence of heat on hashish extracts and cannabinoids by heating them in a sealed tube in an oven for 30 min. In all cases, after heating at 500°C, no cannabinoids could be detected. Apparently the method used is not a good model for the smoking process since in practice about 50 per cent (5) of the cannabinoids are recovered unchanged after smoking. This discrepancy is mainly due to the reaction time which is too long; moreover the volatile products formed cannot be removed from the high temperature zone as is the case in a cigarette. Pyrolysis in a stream of inert gas or air overcomes this problem.

According to Claussen et al. (6) and Fehr et al. (7) the very steep temperature profile in a burning cigarette causes an efficient flash evaporation of the cannabinoids, which is probably the reason that about 50 per cent of the cannabinoids are recovered unchanged in the smoke and the butt of the cigarette. Accordingly the pyrolysis-device must enable a quick temperature rise. In addition the cannabinoid which has to be pyrolysed should be uniformly distributed on the inner wall of the pyrolysis tube to enable an effective evaporation and removal of the pyrolysis products by the gas stream.

The pyrolysis device, constructed considering the parameters mentioned above, is shown in figure II.


Pyrolysis device

Full size image: 15 kB, FIGURE II

A = tube oven; B = pyrolysis tube (quartz); C = slide-rods; D = cold trap (CO 2/acetone).

35 mg CBD was deposited as a thin film on the inner wall of the larger diameter part of the pyrolysis tube. The tube is connected to a gas control unit and nitrogen or air is passed through (100 ml/min.); the smaller diameter part of the pyrolysis tube is inserted through the oven (600°C) and connected with the cold trap. The oven is moved fully to the right.) After five minutes the oven is moved quickly over the larger diameter part of the tube (the oven is now in the position shown in figure II). After pyrolysing for 5 minutes the tube and the cold trap are rinsed with ether.

Results and discussion

The sublimate obtained after smoking CBD through a water-pipe as well as the nitrogen and air pyrolysates of CBD were analysed using GLC, GCMS and TLC (3, 8).

Full size image: 8 kB

In the case of CBD smoked through a water-pipe about 25 per cent of the original CBD is recovered unchanged in the sublimate. The total recovery of cannabinoidal compounds in the sublimate is about 34 per cent. The main cannabinoids formed - in order of decreasing content - are Δ9-tetrahydrocannabinol (II, Δ9-THC), the rearrangement product III and cannabinol (V, CBN). Furthermore many new cannabinoidal products - not yet identified - could be detected.


Estimated composition (w per cent) of the pyrolysates and the sublimate (by GLC and GCMS)


N2 pyrolysate (per cent)

Air pyrolysate (per cent)


40 45 62
Product IV
5 1
Product III
2 2 2
Minor and unidentifiedcannabinoidalproducts
7 3 28

Both in the cases of N 2- and air-pyrolysis of CBD the total recovery is 80-90 per cent, while about 45 per cent of the original CBD is recovered unchanged. Under N 2the main cannabinoids formed - in order of decreasing content - are Δ9-THC (II), an unusual rearrangement product III, the bicyclic cannabinoid IV, and CBN (V) in contrast the major product formed under air is cannabielsoin (VI, CBE) accompanied by a small amount of CBN, while Δ9-THC is practically absent.

Comparison of these data shows that there is a good correlation between the sublimate and the N 2-pyrolysate, suggesting that the smoking of cannabinoids is to a large extent an anaerobic process.

Consideration of the processes taking place in a cigarette do indeed suggest that the smoking of cannabinoids is mainly an anaerobic process. The oxygen content of the gases behind the glowing zone of a burning cigarette is drastically decreased as result of the burning process. However, the temperature directly behind the glowing zone is sufficiently high (6, 7, 9) to cause an effective evaporation of the cannabinoids. Indeed experiments of Fehr et al. (7) show that in those cases where the cannabinoids are not spread evenly over the substrate (e.g. after injection of cannabinoids in a tobacco cigarette) their evaporation has not been completed before the glowing zone (aerobic) passes, resulting in an increased destruction.

In order to quantitate the oxygen content of the gases close behind the glowing zone, a simple device shown in figure III was used. The Cambridge filter is used to prevent the deposit of tar on the polarographic O 2-electrode. In order to measure only the gases that passed the glowing zone, the penetration of air through the cigarette-paper must be eliminated by the use of special airtight paper. A three-way stopcock enabled continuous or intermittent smoking. Thus it was shown that the oxygen content is less than 0.1 per cent (accuracy of measuring device 0.1 per cent) independent of the puff volume and puff frequency.


Experimental device for the measurement of the 02 content in a cigarette

Full size image: 10 kB, FIGURE III

A = cigarette;
C = Cambridge filter;
E = three-way stopcock;
B = rubber lips;
D = polarographic O 2 cell;
F = Mijnhardt O 2 analyser UG 62.

The existence of a relation between N 2-pyrolysis and the smoking process was shown earlier by Muramatsu et al. (10), who were able to derive a quality coefficient for the evaluation of tobacco quality in terms of taste and smell.

From the results and considerations mentioned above it can be concluded that - as far as the cannabinoids are concerned - smoking of cannabis preparations is to a large extent an anaerobic process. As a consequence nitrogen pyrolysis of cannabinoids is a good model for the smoking process.


Thanks are due to the Dutch Department of Public Health and Environmental Hygienics for supporting one of us (H.S.) and to Mijnhardt Medical Instruments, Odijk, for providing facilities for the oxygen measurements. The United Nations Narcotics Laboratory is gratefully acknowledged for their encouragement and generous provision of reference material.



M. J. Repetto and M. Menendez, J. Eur. Toxicol. 6 (3) 392, 1970.


E.B. Truitt, Pharm. Rev. 23 (4) 273, 1971.


F.J.E.M. Kuppers, Thesis, University of Utrecht, The Netherlands, 1973.


Z.I. E1-Darawy, A.M. Rizk, F.M. Hammouda and Z.M. Mobarak, Indian J. Appl. Chem. 35 (1), 9, 1972.


R. Mechoulam, "Marijuana", Academic Press, New York, 1973.


U. Claussen and F. Korte, Ann. Chem. 713, 162, 1968.


K.O. Fehr and H. Kalant, Can. J. Physiol. Pharmacol. 50 761, 1972.


H.J.W. Spronck, Thesis, University of Utrecht, The Netherlands, 1976.


A.R. Patel and G.B. Gori, Bulletin on Narcotics XXVII:3, 47, 1975.


M. Muramatsu, Y.Obi and Y. Shimaday, Nippon Sembai Kosha Chuo Kenkyusho Kenkyu Hokoku 114, 75, 1972 (Chem. Abstr. 80, 93260 and 93261, 1974).