3211

Acta Cryst. (1977). B33, 3211-3214

Cannabidiol
BY PETER G. JONES, LARRY FALVELLO, OLGA KENNARD* AND GEORGE M. SHELDRICK
University Chemical Laboratory, Lensfield Road, Cambridge CB2 1EW, England
AND RAPHAEL MECHOULAM
The Hebrew University, Jerusalem, Israel
(Received 6 May 1977; accepted 21 May 1977)

Abstract. C21H3002, M r = 3 1 4 . 4 7 , monoclinic, P2~, The structure was solved with XCSD, which
a = 10.617 (4), b = 10.649 (5), c = 17.266 (6)A, f l = incorporates a pseudo-tangent refinement to select the
95.30(4) ° , U = 1943A 3 , Z = 4 , D x= 1 . 0 7 4 g c m -3, starting set (Roberts, Pettersen, Sheldrick, Isaacs &
p(Cu Kn) = 4.5 cm -~, m.p. 66-67°C. There are two Kennard, 1973) and the negative quartet test NQEST
independent molecules in the asymmetric unit, differing (De Titta, Edmonds, Langs & Hauptman, 1975). The
mainly in the n-pentyl side-chain conformation. The best E map established positions for 37 heavy atoms;
side chains exhibit thermal disorder. The structure was only 5 small peaks in the top 42 were false. Two atoms
refined to R = 0.068 for 2270 independent reflexions. of the cyclohexene ring were missing from one
molecule; all other missing atoms were the terminal
Introduction. Cannabidiol (Fig. 1) is one of the major portions of the n-pentyl side chains. Figures of merit
components of hashish. Its chemical structure and were NQEST -0.208, R A (a point-atom R factor based
absolute configuration have been determined on observed peaks) 0.362, Ra (Roberts et al., 1973)
(Mechoulam & Shvo, 1963; Mechoulam & Gaoni, 0-142. The second E map (NQEST-O. 188, R A 0.357,
1967). It is a potent anti-epileptic in rats (Izquierdo, Ra 0.124) was almost identical. A difference map
Orsingher & Berardi, 1973; Carlini, Leite, Tannhauser based on the E-map peaks showed all the missing
& Berardi, 1973). The X-ray structure determination atoms except two terminal side-chain atoms of one
was undertaken to investigate possible structural corre- molecule and one of the other. Isotropic least-squares
lations with other anticonvulsant drugs. These, together refinement, followed by a difference map, showed the
with potential-energy calculations, will be discussed remaining heavy atoms; when these were included, the
elsewhere (Falvello, Jones, Kennard, Mechoulam, isotropic R for all heavy atoms was 0-13. Anisotropic
Meyer & Tamir, 1977).
Cannabidiol crystallized from pentane as well shaped
colourless crystals. Intensities were determined on an
automated Syntex P2~ four-circle diffractometer with
Cu Kn radiation and a graphite monochromator, for a
crystal 0.6 × 0.2 x 0.15 mm. Cell dimensions were ob-
tained by least squares from 15 strong reflexions. 2986
reflexions were measured in the range 0 < 20 < 115°;
averaging equivalent reflexions gave 2279 unique
reflexions with F > 4a(F). Systematic absences were
0k0, k odd, implying space group P2 r
* External Staff, Medical Research Council.

7
2

i( , OH

~,/,,~ HO ~, n-CsH,,
Fig. 2. Perspective views of molecules 1 (top) and 2, omitting H
Fig. 1. Cannabidiol, showing the numbering scheme. atoms.
3212 CANNABIDIOL

Table 1. Atom coordinates (× 10 4)

x ),
Molecule 1 Molecule 2
C(1) -943 (6) 2337 (7) 4583 (4) C(1) 5706 (7) -785 (8) 328 (4)
C(2) -439 (5) 2208 (6) 3906 (3) C(2) 6181 (6) 279 (7) 591 (3)
C(3) 954 (5) 2290 (6) 3797 (3) C(3) 7062 (5) 450 (6) 1311 (3)
C(4) 1721 (5) 2766 (7) 4556 (3) C(4) 7053 (6) -729 (6) 1844 (3)
C(5) 1237 (6) 2124 (8) 5258 (3) c(5) 7158 (8) -1876 (8) 1337 (5)
C(6) - 142 (7) 2493 (9) 5330 (3) C(6) 6045 (8) -2010 (9) 737 (5)
C(7) -2362 (6) 2299 (I0) 4619 (4) C(7) 4814 (8) -859 (10) -430 (4)
C(8) 3128 (6) 2614 (9) 4533 (4) C(8) 7993 (6) -643 (7) 2543 (5)
C(9) 3897 (9) 3622 (12) 4556 (6) C(9) 7599 (9) -449 (9) 3265 (5)
C(lO) 3651 (8) 1312 (I I) 4498 (5) C(10) 9417 (8) - 7 7 0 (13) 2385 (7)
C(l') 1860 (5) 2661 (6) 2508 (3) C(l') 7707 (5) 2573 (6) 1880 (3)
C(2') 1155 (5) 3100 (6) 3096 (3) C(2') 6796 (5) 1656 (5) 1732 (3)
C(3') 659 (5) 4289 (6) 3001 (3) C(3') 5637 (5) 1906 (6) 2008 (3)
C(4') 832 (6) 5038 (7) 2359 (3) C(4') 5423 (5) 2929 (6) 2437 (3)
C(5') 1568 (6) 4585 (7) 1788 (3) C(5') 6354 (5) 3842 (5) 2592 (3)
C(6') 2044 (5) 3395 (7) 1864 (3) C(6') 7484 (6) 3660 (6) 2289 (4)
C(I") 1793 (7) 5380 (9) 1087 (4) C(l") 6123 (6) 5009 (6) 3061 (4)
C(2") 667 (10) 5760 (12) 555 (5) c(2") 4989 (8) 5760 (7) 2748 (5)
C(3") -196 (12) 4797 (14) 180 (9) C(3") 4783 (10) 6903 (8) 3237 (6)
C(4") -1344 (12) 5301 (14) -326 (7) C(4") 3572 (14) 7554 (15) 2876 (I 1)
C(5") -1974 (12) 4205 (17) -755 (9) C(5") 3202 (19) 8719 (21) 3262 (11)
O(1) 2351 (4) 1483 (5) 2581 (2) 0(I) 8879 (4) 2370 (5) 1603 (3)
0(2) -57 (4) 4821 (5) 3550 (2) 0(2) 4621 (3) 1087 (4) 1844 (2)
H(I) 403 5957 2302 H(I) 4516 3043 2664
H(2) 2574 3017 1413 H(2) 8212 4369 2369
H(3) 1573 3765 4605 H(3) 6171 -798 2106
H(4) 1298 1119 5189 H(4) 7212 -2700 1703
H(5) 1810 2408 5778 H(5) 8009 - 1800 1043
H(6) -172 3464 5507 H(6) 6274 -2694 310
H(7) -514 1907 5766 H(7) 5239 -2333 1020
H(9) - 1081 2031 3395 n(9) 5913 ! 112 259
H(11) 1306 1363 3684 H(ll) 8011 541 1140
H(12) 2417 4859 746 H(12) 6949 5602 3069
H(13) 2258 6231 1301 H(13) 5985 4725 3648
H(14) 1020 6296 90 H(14) 5128 6063 2165
H(15) 98 6359 888 H(15) 4160 5170 2733

Ilk ll

Fig. 3. Stereo packing diagram of the unit-cell contents, omitting H atoms.

refinement reduced R to 0.09, but the dimensions of the final stages of refinement 26 H atoms were included on
side chains were not chemically reasonable and the well defined C atoms at geometrically calculated
temperature factors of the terminal C atoms were high. positions ( C - H 1.08 ,/k, H - C - H 109.5°). An overall
The refinement was continued (blocked full-matrix isotropic temperature factor for H refined to 0.125 (7)
methods with each molecule as one block) with all bond ,~2; the constrained side-chain bond length refined to
lengths of the side chains constrained to be equal within 1.508 ,~. Nine reflexions showing pronounced ex-
an e.s.d, of 0-01 A. No H atoms could be discerned on tinction were omitted. The final R' =
the consequent difference map and the temperature E wV2A/E wW21FoI was 0.0752, with a corresponding
factors of side-chain C atoms remained high. In the R of 0.0682; the weighting scheme was w = 1/[a2(F) +
 CANNABIDIOL 3213

0.002F2], which gave mean values of wA 2 varying only Final atomic coordinates are given in Table 1,* and
slightly with sin 0 or IF I. A final difference map bond distances and angles in Tables 2 and 3. Diagrams
showed no peaks >0.21 e/~-3. of the structure are given in Figs. 2 and 3.

Table 2. Bond lengths (A)

Molecule 1 Molecule 2 Diseussion. The relative configuration at the asym-
c(2)-c(~) 1.336 (9) 1.304 (12) metric C atoms is consistent with the absolute
c(7)-c(1) 1.514 (ll) 1.546 (12) configuration established by chemical means
c(2)-c(3) 1.510 (10) 1.496 (9) (Mechoulam & Gaoni, 1967). The two independent
c(4)-c(3) 1.562 (9) 1.556 (11) molecules are substantially similar. Significant bond-
c(5)-c(4) 1.521 (1 l) 1.512 (12)
1.507 (1 I) 1.497 (10) length differences (>0.025 ,/~) occur for C ( 3 ' ) - C ( 4 ' ) ,
c(8)-c(4)
C(6)-C(5) 1.531 (12) 1.505 (13) O ( 1 ) - C ( I ' ) , C ( 6 ) - C ( 5 ) and C ( 8 ) - C ( 1 0 ) ; significant
C(1)-C(6) !.487 (9) 1.511 (14) bond-angle differences ( > 2 . 0 °) for C ( 8 ) - C ( 4 ) - C ( 5 )
C(9)-C(8) 1.347 (15) 1.367 (13) and C ( 6 ) - C ( 5 ) - C ( 4 ) . Differences in side-chain bond
C(10)-C(8) 1.497 (15) i.566 (13) lengths and angles are not significant; the high
O(1)-C(l') 1.360 (9) 1.391 (8)
C(3')-C(2') 1.376 (10) 1.386 (8) temperature factors, particularly towards the ends of
C(3)-C(2') 1.518 (lO) 1.515 (10) the chains, indicate thermal disorder (the melting point
C(l')-C(2') !.396 (9) 1.382 (9) is 6 6 - 6 7 ° C ) and the atom positions and derived
C(4')-C(3') 1.392 (10) 1.348 (9) parameters, therefore, cannot be reliable.
0(2)-C(3') 1-389 (8) 1.396 (8)
•400 (10) 1.394 (9) The major torsion-angle difference is for
C(5')-C(4')
C(V')-C(5') .513 1.516 C(5')-C(1")-C(2")-C(3"), which is - 5 7 . 2 and
C(6')-C(5') •365 (1 l) 1.367 (9) 179.2 ° in molecules 1 and 2 respectively. The differing
C(1')-C(6') •388 (I0) 1.387 (10) side-chain conformation is clearly seen in Fig. 2.
C(2")-C(1") -496 1.503 Potential-energy calculations (Falvello et al., 1977)
C(3")-C(2") .483 1.510
C(4")-C(Y') .529 1.542 confirm that both these represent minimum-energy
C(5")-C(4") .506 1.479 conformations.
The benzene ring in molecule 1 is planar and all
Table 3. Bond angles (o) substituents are coplanar within experimental error. In
molecule 2 the aromatic ring is twisted slightly with a
Molecule 1 Molecule 2 torsion angle o f - 4 . 1 (1.5) ° for C ( 1 ' ) - C ( 2 ' ) - C ( 3 ' ) -
C(4') and with O(1) bent out of the plane of the ring by
C(7)-C(1)-C(2) 120.9 (7) 121.7 (8)
125.3 (6) 125.9 (7) 0.11 A. A similar distortion was observed in d9-tetra -
C(I)-C(2)-C(3)
C(5)-C(4)-C(3) 109.5 (6) 107.7 (6) hydrocannabinolic acid B (Rosenqvist & Ottersen,
C(8)-C(4)-C(3) 112.7 (6) 112.8 (6) 1975) where, however, the strain could be attributed to
C(2)-C(3)-C(4) 110.8 (5) 110.8 (6) the fused tricyclic ring system. The cyclohexene rings in
C(6)-C(5)-C(4) 110.2 (6) 112.3 (8) both molecules are in the half-chair conformation; C(4)
C(9)-C(8)-C(4) 120.9 (9) 120.5 (8)
C(I0)-C(8)-C(4) 118.2 (8) 115.9 (8) is 0.32 and 0.44 A above, and C(5) 0.46 and 0.33 A
C(8)-C(4)-C(5) 112.5 (6) 115.9 (7) below, the plane of the remaining four atoms in
C(1)-C(6)-C(5) 112.1 (6) 112. l (8) molecules 1 and 2 respectively.
C(2)-C(1)-C(6) 121.8 (7) 121.4 (7) The two molecules are linked by hydrogen bonding
C(7)-C(1)-C(6) 117.3 (7) 116.9 (8)
123.6 (9) between O(1) of molecule 1 and 0 ( 2 ) of molecule 2; the
C(9)-C(8)-C(10) 120-9 (8)
C(3)-C(2')-C(1') 121.0(6) 122.0 (6) O - O distance is 2.86 A. Apart from this the crystal
C(4')-C(3')-C(2') 123.0(7) 123.0 (6) packing is very open, as indicated by the low density
C(2)-C(3)-C(2') 110.1 (5) 111.8 (6) and melting point and the high thermal parameters for
C(4)-C(3)-C(2') 112.6 (6) 112.9 (6) the side chains.
C(6')-C(1')-C(2') 121-3 (7) 122-4 (6)
O(2)-C(Y)-C(2') 121.4 (6) 120.2 (6)
0(1)-C(1')-C(2') 117.9 (6) 117.3 (6)
C(3)-C(2')-C(3') 122.5 (6) 122-6 (6) We thank the Medical Research Council for finan-
C(1')-C(2')-C(3') 116.6 (6) I15.3 (6) cial support, the Winston Churchill Foundation for a
C(5')-C(4')-C(3') 119.1 (7) 121.1 (6) maintenance grant (to LF) and the Science Research
C(6')-C(5')-C(4') 118.5 (7) 117.3 (6) Council for the provision of the diffractometer. The
C(1")-C(5')-C(4') 120.6 (8) 121.8 (6)
O(2)-C(Y)-C(4') 115.5 (7) 116.7 (6) figures were drawn with P L U T O written by Dr W. D.
C(1')-C(6')-C(5') 121.4 (7) 120.7 (7) S. Motherwell.
C(1")-C(5')-C(6') 120.8 (7) 120.8 (6)
O(1)-C(1')-C(6') 120.9 (6) 120.3 (6) * Lists of structure factors and thermal parameters have been
C(2")-C(1 ")-C(5') 117.8 (8) 114.0 (6) deposited with the British Library Lending Division as Supplemen-
C(3")-C(2")-C(1:') 120.5 (11) 112.3 (8) tary Publication No. SUP 32796 (16 pp.). Copies may be obtained
C(4")-C(3")-C(2") 115.7 (13) 107. l (10) through The Executive Secretary, International Union of
C(5")-C(4")-C(3") 107.7 (13) 116.1 (15) Crystallography, 13 White Friars, Chester CH 1 1NZ, England.
3214 CANNABIDIOL

References MECHOULAM, R. & GAONI, Y. (1967). Tetrahedron Lett. pp.

1109-1111.
CARLINI, E. A., LEITE, J. R., TANNHAUSER, M. & BERARD1,
MECHOULAM, R. & Snvo, Y. (1963). Tetrahedron, 19,
A. C. (1973). J. Pharm. Pharmacol. 25, 664-667.
2073-2078.
DE TITTA, G. T., EDMONDS, J. W., LANGS, D. A. &
ROBERTS, P. J., PETTERSEN, R. C., SHELDRICK, G. M.,
HAUPTMAN, H. (1975). Acta Cryst. A31,472-479.
ISAACS, S. W. &, KENNARD, O. (1973). J. Chem. Soc.
FALVELLO, L., JONES, P. G., KENNARD, O., MECHOULAM,
Perkin II, pp. 1978-1984.
R., MEYER, A. Y. & TAMm, I. (1977). To be published.
ROSENQVIST, E. & OTTERSEN, T. (1975). Acta Chem. Stand.
IZQUIERDO, I., ORSINGHER, O. A. & BERARDI, A. C. (1973).
B29, 379-384.
Psychopharmaeology, 28, 95-99.

Acta Cryst. (1977). B33, 3214-3215

(Triethylphosphine)(diethylethylidenephosphine)-
[ 2- methyl- 1,2-diearbadodeeaboranyl(10)]platinum(II)
BY N. BRESCIANI-PAHOR
Istituto di Chimica, Universitd di Trieste, 34127 Trieste, Italy

,: (Recieved 28 February 1977; accepted 24 April 1977)

Abstract. PtP2B10CIsH42 , M r = 587.6, triclinic, P i the block-diagonal least-squares method. After iso-
(from structure determination), a = 16.39 (8), b = tropic refinement, R = EIIFol - Ifcll/EIFol was
l l.20 (6), c = 7.98 (5) A, (t = 103.2 (1), fl = 107.4 0.152. The final anisotropic refinement reduced R to
(1), y = 90.5 (1) °, U = 1357.1 A 3, D c --- 1.44, Z = 2,
D m = 1.46 g cm -3 (by flotation), g ( M o K(0 =
55-6 c m - l . The structure was solved by Patterson and
Table 1. Fractional coordinates (× 10 a) with estimated
Fourier m e t h o d s and refined to R = 9 . 5 % for 2961
standard deviations in parentheses
independent reflexions. The 2 - m e t h y l - l , 2 - c a r b o r a n y l
group is o-bonded to Pt t h r o u g h its 1C atom. One X 11

phosphine is c o o r d i n a t e d to the metal atom through its Pt 2363 (1) 1299 (1) 897(I)
P atom, the other t h r o u g h its P atom and the first C P(1) 3651 (4) 2067 (6) 834 (11)
atom of one ethyl side group. P(2) 1547 (4) 2811 (6) 1260 (10)
B(1) 1540 (17) -1520 (29) 4 (49)
B(2) 2329 (21) -1402 (28) 2077 (53)
Introduction. Single crystals were obtained by slow B(3) 3337(20) -1331 (28) 1644 (60)
evaporation of a dichloromethane solution. Intensities B(4) 3129 (17) -1394 (26) -736 (59)
were collected on a Siemens diffractometer by the 0-20 B(5) 1585 (23) -2883 (31) -1626 (59)
B(6) 1794 (21) -2826 (27) 771 (62)
scan technique with Mo Ka radiation for a maximum B(7) 2915 (25) -2785 (30) 1803 (73)
20 angle of 54 °. All reflexions with I < 3o(I) were B(8) 3405 (22) -2703 (33) 58 (68)
rejected, the remainder being corrected for Lorentz and B(9) 2529 (25) -2787 (36) -2095 (70)
polarization effects. 2961 independent reflexions were B(10) 2399 (25) -3621 (36) -692 (60)
used in the subsequent calculations. No correction for C(1) 3816 (16) 3605 (36) 1884 (52)
C(2) 4675 (25) 4402 (33) 2104 (75)
absorption was applied, because of the small size of the c(3) 4638 (16) 1483 (26) 2005 (50)
crystal used (approximately a cylinder with a diameter c(4) 4754 (23) 1706 (36) 4019 (58)
of ca 0.2 mm). During data collection a strong decrease c(5) 3813 (20) 1844 (37) - 1533 (46)
of the standard reflexion intensity was observed c(6) 3178(31) 2554(51) -2705 (50)
C(7) 1719 (21) 4041 (33) 3290 (54)
suggesting a significant disordering of the crystal, due c(8) 937 (24) 4476 (40) 3684 (60)
to X-ray exposure, which greatly affects the accuracy c(9) 865 (20) 3443 (27) -618 (49)
of the structure analysis. This was partly avoided by C(10) 548 (24) 2510 (38) -2358 (49)
correcting the intensities, which decreased by about c(11) 1114(13) 1421(21) 1368 (34)
50% during the overall recording time with a nearly C(12) 999 (16) 1115 (31) 3007 (39)
C(13) 2482 (13) -593 (19) 498 (32)
linear trend. The structure was determined by conven- C(14) 2036 (17) -1545 (28) -1599 (46)
tional Patterson and Fourier methods and refined by C(15) 1608(28) -811 (41) -3182 (51)