You are on page 1of 2


explained by an intramolecular hydrogen bonding in THC (7), and 34% unreacted olivetol.' Careful chro-
this acid competing with its intermolecular association. matography of this mixture on silica gel and elution
(13) C. P. Scott Research Fellow of the Weizmann Institute of
with 60 :40 hexane-benzene, followed by chromatog-
Science. raphy on Florisil and elution with 2 :98 ether-petroleum
(14) Predoctoral student of the Feinberg Graduate School. ether (bp 30-40), gave ( -)-A1(6)-THC (infrared nmr,
* To whom correspondence should be addressed. glc, and tlc identical with those of authentic material ;
Amnon Yogev,lS Leon Margulies,l4 Yehuda Mazur* [ a ] -245"
~ (CHC13), indicating an optical purity of
Departmetit of Chemistry, The Weizmann Institute of Science ca. 9573.8 Since (-)-Al@)-THC has already been
Rehovoth, Israel converted to ( -)-A1-THC,2~3this route constitutes yet
Received May 6, I970 another stereospecific synthesis of (-)-4.
However, it was found that in the above condensation
Hashish. A Stereospecific Synthesis of ( -)-A1- reaction, if the molar ratio of 1 was increased (1.6
and ( - )-AIcG)-Tetrahydrocannabinols1 mol/mol of olivetol), the major products formed were
A'-trans-THC (28 %) and A'-cis-THC (8) together with
Sir: other products, but no A1@)-THCwas formed (Scheme
Two stereospecific syntheses of (-)-A '(Wetrahydro- I). Similar results were obtained by the addition of 1 %
cannabinol (THC, 3) have been reported which utilize boron trifluoride etherate in methylene chloride to
the condensation of olivetol with either ( -)-verbeno12 an equimolar ratio of 1 and 2 in methylene chloride
or (+)-cis- or trans-p-mentha-2,8-dien-l-o13 in the pres- at room temperature. The mixture of products thus
ence of acids. In both cases the synthesis of (-)-A1- obtained was chromatographed on Florisil and eluted
THC (4) was achieved by addition and then elimination with 2 :98 ether-petroleum ether (30-40"). A fraction,
of hydrogen chloride to form 3. [ a ] -59"
~ (CHC13), contained 4 and 8 (60:40 ratio),
Scheme I

4 8


p c 5 H 1 1 + v c 5 H l l

iso-THC kTHC
6 7
We wish to report an entry into cannabinoids via as shown by gas chromatographic6 and nmr analysesg
carane derivatives that provides a one-step stereospe- and by conversion to A1@)-THCand iso-THCs on
cific synthesis, leading to ( -)-A1(6)-THC, and hence treatment with p-toluenesulfonic acid in refluxing ben-
to (-)-A1-THC. Furthermore, a slight change in the zene.'tfi3lo This fraction was separated by preparative
experimental conditions also provides the first one- vpc (column 3.8% SE 30 on Diatoport S ; gas, helium;
step synthesis of AI-THC (4) of high optical activity. oven temperature, 240") to give (-)-A1-THC6 with
We have found that by allowing (+)-trans-2-carene an optical purity of >SO by comparison with the
oxide4s5 (1) to react with an equimolar quantity of O R D curve of an authentic sample. This therefore
olivetol (2) in the presence of p-toluenesulfonic acid provides the first one-step synthesis of optically active
(0.05 mol/mol of olivetol) in benzene, a complex mix- A'-THC.
ture of products was obtained. This mixture con- It is interesting to note that no cannabidiol3 was
tained (glc)6 23 % A'@)-THC,7 % A'-THC, 13 % citryl- formed in any of these reactions and that the ratio
idenecannabis ( 5 ) and A8-iso-THC (a), 11 % A4(8)-iso- of trans to cis products was ca. 1.2 with p-toluenesul-
(1) Hashish. V. For part IV see R. K. Razdan and B. Zitko, (7) The yield of Al(s)-THC is 54% of theory, on the basis of olivetol
Tetrahedron Lett., 4947 (1969). reacted.
(2) R. Mechoulam, P. Braun, and Y.Gaoni, J . Amer. Chem. SOC., 89, (8) The material after silica gel chromatography was identical with 3
4552 (1967). except it had [ a ]-212"
~ (CHCh). We attribute this difficulty in ob-
(3) T. Petrzilka, W. Haefliger, and C. Sikemeier, Hela. Chim. Acta, taining optically pure 3 to our carene oxide, which was not prepared
52, 1102 (1969). from the natural A*-carene.4 Further experiments utilizing the natural
(4) Supplied by Glidden and Co., [a10 -45.0" (neat), nZ0D1.4642, bp Az-carene are planned to confirm this point.
75" ( 5 mm). It contains A3-carene oxide as impurity (estimated as 15 (9) The nmr showed a clear separation of signals at S (CDCh) 1.08 (s,
by glc and nmr). I t was synthesized from As-carene. one of the geminal methyls) and 3.18 (br, d, C-3 proton), which cor-
( 5 ) B. A . Arbuzov and A . R. Vil'chinskaya, Izr. Akad. Nauk SSSR, respond to A'-rrans-THC, and similarly 1.26 (9) and 3.56 (br) cor-
Ser. Khim., 954 (1967); cf. Chem. Abstr., 68, 22063 (1968). For a re- responding to AI-cis-THC. See K. E. Fahrenholtz, M. Lurie, and R.
action of trans-4-carene oxide in the presence of acids see von G . Ohloff W. Kierstead, J . Amer. Chem. SOC.,89, 5934 (1967).
and W. Giersch, Helc. Chim. Acta, 51, 1328 (1968). (10) Some 5 and 8 were also formed as expected.'
(6) The compounds were identified on the basis of relative retention (11) Both O R D curves were found to be parallel in the 280-600-mp
times of authentic samples and by addition of authentic samples to the region; [a]o- 131 f 2" (cyclohexane); an authentic sample of (-)4
reaction mixture with subsequent glc. had [alo - 148 f 2".

Communications to the Editor

fonic acid, whereas it was ca. 0.6 (i.e., increase of cis fact that their uv-visible and nmr spectra clearly indicate
products) with the use of boron trifluoride etherate. We the formation of the ionic species responsible for the
interpret these results as suggesting that trans- and cis- normal rearrangement, the oxygen-protonated, de-
A'-TNCs are first formed (Scheme 11) and are then
0 OH
Scheme I1 II I!


H+ &H

I1 Iv
localized cyclohexadienyl cations I11 and IV, respec-
4 f 8
t i ~ e l y . ~Only
, ~ heating at 60" for 24 hr causes partial
1,2 migration of the methyl group in I and generates,
after hydrolysis, 2-methyl-4-hydro~ybenzaldehyde.~
converted into their transformation products 3 and In the present communication we report the efficient
5 + 7, respectively. This is in complete agreement photorearrangement of ions TI1 and IV to the same
with the known acid-catalyzed transformations in can- photoproduct, 2-methyl-5-hydroxybenzaldehyde (V).
nabinoids. l 2 Furthermore, the results suggest that
Figures 1 and 2 show successive changes in the uv-
the reaction has a more concerted character when a visible absorption spectra of IV and I11 following short-
nonprotonic acid catalyst like BF, is used, whereas duration irradiations' with light of 370 and 262 nm,
acid catalysts like p-TSA, which give a more carbonium respectively, at ambient temperature. Figure 1a il-
ion character to the allylic system in the intermediate 9, lustrates the gradual disappearance of IV (A, 370 nm)
give the trans products, which are more thermodynami- with simultaneous increase in absorbance at 258 and
cally and kinetically stable. 1 1 2
298 nm. That these two absorption bands do not be-
Further work along these lines is in progress and long to the same species is shown by the subsequent
will be the subject of future communications. changes shown in Figure lb. Thus the peak at 298 nm
Acknowledgment. The authors wish to thank Dr. R. continues to increase upon further irradiation whereas
Lyle for helpful discussions and for obtaining the that at 258 nm gradually decreases after reaching a
O R D curves and Dr. L. Mir for carrying out the maximum. The final product (A,, 298 and 388 nm)
preparative vpc. This work was carried out with the seems to be photostable. The absorption spectrum of
support of Contract No. PH-43-68-1339, National a solution of IV irradiated to any per cent conversion
Institute of Mental Health. remained unchanged after standing in the dark for
extended times. This testifies that the changes in Fig-
(12) B. Yagen and R . Mechoulam, Tetrahedron L e f t . ,5353 (19693.
* Address correspondence to this author. ure 1 are photochemical with no interference from ther-
mal reactions. It is quite clear that ion IV is converted
Raj K. Razdan,* G . Richard Handrick photochemically to an intermediate species ,,A,( 258
Arthur D . Little, h c . nm) which in turn undergoes photolysis to a product
Cambridge, Massachusetts 02140 absorbing at, ,A, 298 nm. The three isosbestic points
Receiced June 25, 1970 at 248, 272, and 332 nm in Figure l b indicate that the
intermediate --c product photoconversion preserves a
constant stoichiometry and that the transformation is
void of competitive or consecutive slow reactions. On
Photochemical Rearrangement of the other hand the isosbestic point at 329 nm in Figure
Cyclohexadienyl Cations l a merely indicates a coincidental equality of extinction
Sir: coefficients at that wavelength for the three species
In the presence of sulfuric acid, 4,4-dimethyl- and Figure 2 shows changes in the absorption spectrum
4,4-diphenyl-substituted cyclohexa-2,5-dienones readily of 111 on irradiation at 262 nm. Again the photoreac-
undergo the dienone-phenol rearrangement. In con- tion is quite clean and the final product exhibits a uv-
trast, the "abnormal Reimer-Tieman" dienones 4- visible spectrum ,A,(, 298 and 388 nm) indistinguishable
methyl-4-dichloromethylcyclohexa-2,5-dienone (I) and from that obtained from IV.
6-methyl-6-dichloromethylcyclohexa-2,4-dienone (11) do Preparative-scale photolysis of either 111 or IV loF2
not rearrange under similar conditions. Furthermore, M in 96 H2S04at 17 O was accompanied by HCI liber-
they can be recovered quantitatively from concentrated
sulfuric acid by dilution with water,3 in spite of the (3) H . Budzikiewicz, Tetrahedrorz Lett., 12 (1960).
(4) E. C. Friedrich, J . Org. Chem., 33, 413 (1968).
( I ) (a) F. G . Bordwell and K . IM. Wellman, J. Org. Chem., 29, 509 (5) V. P. Vitullo, ibid., 34, 224 (1969).
(1964); (b) H . E. Zimmerman and D. I . Schuster, J . Amer. Chem. Soc., (6) T. G . Miller, ibid., 27, 1549 (1962).
84, 4527 (1962); (c) A . J. Waring, Adcan. Alicyclic Chem., 1, 207 (1966). (7) Bausch & Lomb high intensity monochromator with 125-W
(2) I(. v. Anwers and G . Keil, Ber., 35,4207 (1902). high-pressure H g arc.

Journal of the American Chemical Society 1 92:20 1 October 7 , 1970