2. Part (v). X-RAY CRYSTALLOGRAPHY By George Ferguson (Department of Chemistry The Universi&y of Glasgow Glasgow W.2 Scotland) CRYSTALLOGRAPHY has been included in the ‘Physical methods’ section of annual reports for the first time and the literature coverage is slightly different from that of the previous organic crystallography section. The aim of this report is to give details of the more important organic structures which have been determined by single-crystal diffraction methods during the year ;papers which only give barest details of constitution of e.g. natural products will be reported in other chapters and have been omitted here. More than 250 relevant papers have appeared but space requirements have necessitated omitting some of these from this review.The advent of computer-controlled diffractometers makes accurate data collection more rapid and also makes the determination of the absolute configuration’ of organic molecules more certain. More accurate data can also allow details of bonding electron distribution and hydrogen-atom positions to be determined but for precise location of hydrogen atoms the increasing availability of neutron diffraction facilities is invaluable. The crystallographic phase problem remains the major difficulty in crystal structure analysis and was overcome in 53% of the structures reported here by the heavy-atom method; ‘direct’ methods’ were described in 15% of the papers and the structures described in the remaining 32% mainly those of small molecules were solved from a knowledge of molecular geometry vector or trial-and-error met hods.It is probably timely to insert a word of warning here to non-crystallo- graphers about the significance of ‘estimated standard deviations’ of bond lengths and angles which should always be quoted in crystallographic papers. It is generally recognised that it is not safe to claim that e.g. a bond length is significantly different from a comparable value unless the difference is at least three times the estimated standard deviation. This appears to have been overlooked on occasion and the situation is further complicated because in the majority of cases the estimated standard deviations are themselves under- estimated owing to the computational approximations used in deriving them.Carboxylic Acids and Related Compounds.-Crystals of deuteriated oxalic acid dihydrate (C02D),,2D20 are not isomorphous with those of (C02H), J. M. Bijvoet Proc. k. ned. Akad. Wetenschap. 1949,52 313. See also S. Ramaseshan in ‘Ad- vanced Methods of Crystallography,” Academic Press London 1964 p. 67. For a brief review see M. Gerloch and R. Mason Ann. Reports 1966,689. George Ferguson 2H20. A three dimensional X-ray analysis3 has determined the C-€ distance as 1-539 & 0.005 8 and shown the oxalic acid and water molecules to be linked by hydrogen bonds forming a three-dimensional network. The hydrogen-bond distances O-D-.-O are 2.540 2.854 and 2.822 A. The deuterium atom positions have been determined by neutron diffraction ;4 the bond lengths in the D20 molecule are 0.946 and 0.960 A and D-0-D is 110.5".The lack of optical activity in meso-tartaric acid was usually ascribed to intramolecular compension; the two halves of the molecule were supposed to be mutually related either by a mirror plane or a centre of symmetry.X-Ray analyses of a number of crystalline modifications of the acid,5 however show that in the solid a staggered conformation (1) is adopted as is found in meso- tartrates. The lack of optical activity is thus caused by intermolecular com- pensation of the conformational antipodes. meso-PP'-Dimethyladipic acid has a crystallographic centre of symmetry6 and hence adopts a fully staggered conformation in the crystal. HO CO H Me HO@:02H C02H An X-ray arbitration7 has unequivocally established the absolute configura- tion of (+)-2-isopropyl-7-methylglutaric acid as (2) in agreement with the work of Norin,' but at variance with the conclusion of Djerassi and co-worker~.~ It follows that the configurations assigned' to ( +)-sabinene (-)-umbellulone the thujanes and their congeners are confirmed but that the absolute configuration assigned to ( -)-methylisop~legone~must now be represented as (3).A further consequence of these findings is that the abso- lute configurations of numerous derivatives of ( -)-methylisopulegone and (-)-menthone must be revised. The C-0 bond lengths in aminomalonic acid show that the molecule is + present in the crystal as the zwitterion C02H*CHNK,*C0,-.'0Molecules of F.F. Iwasaki H. Iwasaki and Y. Saito Acta Cryst. 1967,23,64. F. F. Iwasaki and Y. Saito Acta Cryst. 1967,23 56. G. A. Bootsma and J. S. Schoone Acta Cryst. 1967,22 522. E. Martuscelli Ricerca sci. 1967 37 53; P. Ganis E. Martuscelli and G. Avitabile Ricerca sci. 1966,36,689. M. R.Cox H. P. Koch W. B. Whalley M. B. Hurtshouse and D. Rogers Chem. Comrn 1967 212. T. Norim Actu Chem Scand. 1962 17 640. E. J. Eisenbraun F. Burian J. Osiecki and C. Djerassi J. Arner. Chem. SOC.,1960,82 3476. lo J. A. Kanters J. Kroon P. T. Beurskens and J. A. Vliegenthart Actu Cryst. 1966 21 990. Part (u) X-Ray Crystollagraphy eaminocaproic acid NH .[CH,] -CO,H which undergo polycondensation in the solid state to Nylon 6 are also present in the crystal as zwitterions with three hydrogen atoms 0.9 8 from each nitrogen atom." Each zwitterion forms three N-H*.*O hydrogen bonds (2.728 2.753 and 2.814 A) with adjacent molecules to form a double-layered structure.The analyses of a number of butadienes of photochemical interest have appeared. The structure12 of monomethyl-trans-trans-muconate MeO,C* CH CH * CH :CH C02H con-sists of stacks of hydrogen-bonded molecule pairs (O-H***O 2.62 A); double bonds of nearest neighbour molecules make contact along the short c axis of the crystal and are presumably responsible for the photochemical formation of cyclobutanes. Interstack contacts of 3.6 A between terminal atoms of the butadiene chain may account for initiation of polymerisation. In the dimethyl- m~conate,'~ analysis of the bond angles at the methoxycarbonyl group suggests that the axis of the methyl group makes an angle of 5" with the 0-C vector interpreted as evidence for a bent bond.Accurate analyses have also yielded precise measurements for trans-trans-~orbamide'~ and trans-trans-muconodi- nitrile.15 Two polymorphic modifications of azelaic acid HO2C*[CH2],*CO2H,have been studied.I6 The two carboxyl groups of the a-form have different dimensions [C(l)-O(1) = 1.253 C(l)-0(2) = 1.266 8 while C(9) O(3) = 1.283 and C(9)-0(4) = 1.192A). The &form has crystallo- graphic two-fold (C,) symmetry with the C-0 distances 1.231 and 1.303 A. Both compounds crystallize to give infinite hydrogen-bonded chains. Deca- trans-3-trans-7-dienedioicacid has a crystallographic centre of symmetry and the torsional angle at C(2)-C(3) (double bond eclipsed to hydrogen) is -125-5" while at C(4)-C(5) (double bond eclipsed to methylene) the torsional angle is -1-6".This is thought to be the first observation of a cis-conformation in the crystalline state for this kind of bond.17 The occurrence and authenticity of some 'very short' hydrogen bonds in which the hydrogen atom is symmetrically positioned between two oxygen atoms have been reviewed.l8 Very short hydrogen bonds are found in the isomorphous rubidiumlg and potassium hydrogen diaspirinates.,' The metal ion lies on a two-fold axis and makes contact with an oxygen atom from each of six different aspirinate residues and each aspirinate makes contact with three different metal ions.The acidic hydrogen atom is involved in a very short hydrogeh bond with 0 0 2.455 0.005 A which connects two aspirinate residues across a centre of symmetry. A neutron diffraction study of G. Bodor A. L. Bednowitz and B. Post Acta Cryst. 1967,23 482. D. Rabinovich and G. M. J. Schmidt J. Chem. SOC.(B) 1967 286. l3 S. E. Filippakis L. Leiserowitz and G. M. J. Schmidt J. Chem. Soc. (B) 1967 290. l4 S. E. Filippakis L. Leiserowitz and G. M. J. Schmidt J. Chem. SOC.(B),1967 297. l5 S. E. Filippakis L. Leiserowitz and G. M. J. Schmidt J. Chem. Soc. (B),1967 305. J. Housty and M. Hospital Acra Cryst. 1967 22 288. E. Martuscelli Acta Cryst. 1967 23 1086. J. C. Speakman Chem Comm. 1967 32. l9 S. Grimvall and R. F. Wengelin J. Chem.Soc. (A),1967 968. L. Manojlovic and J. C. Speakman J. Chem. SOC.(A) 1967,971. C* 68 George Ferguson ammonium tetroxalate has located the hydrogen atoms accurately.2 The structure involves seven distinct hydrogen bonds ranging in length (0 -* 0) from 2.899-2.472 A and the corresponding 0-H distances tend to increase inversely from 0.945 to 1-102A. The four hydrogen atoms of the ammonium ion form four nearly linear N-H.**O bonds with N-.*O lengths 2.927 2.938 2.950 and 2.974 (each +0004)8 and N-H 1.004,1.022 1.015 and 0.995 (each +0.010) A. The oxalate residues have their carbon and oxygen atoms coplanar and C-C 1.544 1549 (each +0.005),and 1549 (&0-003)A. The ionization in the di-ionized salt lithium ammonium citrate monohydrate occurs in the central and one of the terminal carboxyl groups.22 The backbone of the citrate ion is fully extended in a plane roughly perpendicular to that of the central carboxy- and hydroxy-groups.The lithium ion is tetrahedrally co- ordinated by oxygen atoms from four different citrate ions. Considerable overcrowding exists in the anion of the monorubidium salt of furantetra-carboxylic acid and the resulting strain is accommodated in the usual ways by (a)in-plane splaying-out and out-of-plane bending of the exocyclic carbon atoms and (b)rotation of the carboxyl-group planes with respect to the furan ring plane.23 There is a close approach (2.386 f0.013A) between two oxygen atoms of adjacent carboxy-groups but no hydrogen atoms were located. exo-N-Methylacetanilide has crystallographic rn-symmetry with the N- methylacetamido-group in the mirror plane and the benzene ring normal to it.24 The N-C /\distance is short (1.344A)revealing appreciable double-bond character and the C=O distance (1.263 8,) is somewhat longer than the corresponding distances in related compounds.Ethyl carbamate crystals (urethane EtO CO NH2) consist of chains of centrosymmetric hydrogen- bonded dimers (N-H-**O 2.96 A) which are interconnected by further N-H-**O hydrogen bonds. The C=O and C-NH distances are 1.221 and 1.345A. The N-C-0 angle is 123.6°.25Acrylamide crystallises in a similar fashion.’ L-Ornithine hydrochloride exists as a zwitterion NH3 + [CH,] CH(NHi) CO,-.Cl-in the solid state each nitrogen atom making three N-H***O hydrogen bond^.'^ The molecule is characterised by two planar groups the carboxy-group and the aliphatic chain.The absolute configuration of (+)-2-benzylglutamic acid hydrobromide hydrate has been determined as D. The carbon skeleton of the glutamic acid is different from other glutamic acids (which are reviewed) due to internal rotations of the C-C bonds.28 The molecular and crystal structures of the hydrochloride and hydrobromide of 21 M. Currie J. C. Speakman and N. A. Curry J. Chem. SOC.(A) 1967 1862. 22 E. J. Gabe J. P. Glusker J. A. Minkin and A. L. Patterson Acta Cryst. 1967,22 366. 23 L. L. Martin and 1. C. Paul Acta Cryst. 1967,22 559. 24 B. F. Pedersen Acta Chem. Scand. 1967,21 1415. 25 €3. H. Bracher and R. W.H. Small Acta Cryst. 1967,23 410. 26 I. V. Isakov Zhur. strukt. Khim. 1966,7 898. 27 A. Chiba T. Ueki T. Ashida Y. Sasada and M. Kakudo Acta Cryst. 1967,22,863. T. Ashida Y. Sasada and M. Kakudo Bull. Chem. SOC. Japan 1967,40 476. Part (v) X-Ray Crystallography 69 L-tryptophan have also been determined.29 The iodine-iodine distance in di-iodo-L-tyrosine dihydrate (4) (6.05 A) is shorter than was expected. The molecules are held in the crystal by an extensive hydrogen-bond net involving the carboxy- amino- and phenolic groups of the molecule and the water molecules.30 The stereochemistry of L-leucine hydrobromide has been deter- mined3' and that of L-valine hydrochloride revised.32 The crystal structures of pentane and octane have been refined.33 In pentane the average C-C bond length is 1.533 0.006 the C-C repeat distance is 2.539 f0.006 A and the average C-C-C angle is 112.1 & 0.3".The shortest intermolecular C*-C distances are 3.90 3-95 and 3.96 A. In octane the average C-C bond length is 1525 f0.007 A the C-C repeat distance is 2.545 + 0.007 8 and the average C-C-C angle is 113.3 f0.6". The shortest intermolecular C...C contact is 3-65 A. OH CH2*CH(NH,) .CO,H I0 9 Me CH Aromatic Molecules and Related Systems.-The structure3' of 1,6-epoxy[lO]annulene (5) is very similar to the carboxylic acid derivative of 1,6-methano[ 10Jann~lene.~~ The eight carbon atoms not directly associated with the bridge i.e. C(2&(5) and C(7t(lO) are essentially coplanar while the re-entrant carbon atoms C(1) and C(6) each lie 0.35 A above the plane with the oxygen atom 1.25 A from it.The C-0-C- angle is 102" compared with 99.6 % for the related angle in the carbocycle. The tendency towards bond alternation noted in the carb~cycle~~ is not seen in the epoxy[lO]annulene where all bond lengths are within two standard deviations of 1.39 A. The 29 T. Takigawa T. Ashida Y. Sasada and M. Kakudo Bull. Chem. SOC.Japan 1966,39,2369. 30 J. A. Hamilton and L. K. Steinrauf Acta Cryst. 1967,23 817. 31 E. Subramanian Acta Cryst. 1967,22 910. 32 0.Ando T. Ashida Y.Sasada and M. Makudo Acta Cryst. 1967 23 172. 33 H. Mathisen N. Norman and B. F. Pedersen Acta Chem. Scand. 1967 21,127. 34 N. A. Bailey and R.Mason Chem. Comm. 1967 1039. 35 M. Dobler and J.D. Dunitz Chim. Acta. 1965 48 1429. 70 George Ferguson bond-length distribution in the peripheral 14-membered ring of trans-15,16-diethylhydropyrene (6) is characteristically aromatic36 (the lengths are in the range 1-393-1.401 A). A measure of the non-planarity of the ring is the displacement of C(3) (0.232 A) from the plane containing the four carbon atoms bonded to the internal substituent. In the corresponding methyl derivative37 the distance is only 0.117 A. The C(6)-C(9) bond is abnormally long (1-574 A) and the angle C(6)-C(9)-C(10) (115.7") is much greater than tetrahedral. The effect of these distortions is to increase the separation of the two methyl hydrogens at C(10) from the carbon skeleton of (and presumably from the 71-electron cloud of) the 14-membered ring.The transannular bond length in azulene-1,3-dipropionic acid is 1.490 A 0.008 A strongly suggestive of a Csp2-Csp2 single bond.38 The average value of the peripheral ring bonds is 1.393 A. The internal C-C-C bond angles at C(4) C(6) and C(8) are in good agreement with each other and their average of 130.1" is greater than that 127.43 of the remaining internal angles in this ring. l-Methylamino-7- methylimino-1,3,5-cycloheptatriene (7) a typically disubstituted aminotropone- imine has essentially mm2 (C?,,)symmetry including the methyl groups which are staggered with respect to the nearest ring hydrogen at0ms.j' The molecular structure can be described in terms of a 10-electron x-system encompassing the seven-membered ring and the two nitrogen atoms.The single amino- hydrogen atom appeared in a final difference Fourier synthesis as two half- weight peaks symmetrically situated with respect to the two nitrogen atoms a finding consistent with the n.m.r. i.r. and chemical evidence that the nitrogen atoms are equivalent. The i.r. absorption spectra of a number of o-halogeno-aroyl compounds both in solid state and solution have been correlated with the crystal structures as revealed by three-dimensional X-ray analy~es.~' For ortho-substituted benzoic acids the preferred conformation in the solid state is usually that with the carbonyl group adjacent to and the hydroxy-group remote from the ortho-substituent. An analysis of the o-toluic acid structure4' shows sur- prisingly that the displacements of the carboxy- and methyl groups from the benzene plane are not large.For o-nitrobenzoic acid42 the plane of the carboxyl group and that of the nitro-group make angles of 23.4 and 54.7" respectively with the benzene-ring plane. The effect of intramolecular over- crowding causes the exocyclic carbon atom and the nitrogen atom to be dis- placed out of the aromatic plane in opposite directions by 0.209 and 0.160 respectively; the C-C and C-N bonds are also displaced sideways. The angle of tilt of the carboxy-group plane to the benzene plane in o-fluorobenzoic 36 A W. Hanson. Acta Cryst.. 1967. 23. 476. 37 A. W. Hanson Acta Cryst. 1965 18 599. 38 H. L. Ammon and M. Sundaralingam,J. Amer. Chem. SOC.,1966,88,4794.39 P. Goldstein and K. N. Trueblood Acta Cryst. 1967 23 148. 40 G. Eglinton G. Ferguson K. M. S. Islam and J. S. Glasby J. Chem. SOC.(B),1967 1141. 41 C. Katayama A. Furusaki and I. Nitta Bull. Chem. SOC.Japan 1967,40 1293. 42 T. D. Sakore S. S. Tavaie and L. M. Pant Acta Cryst. 1967,22 720. M. Kurashi M. Fukuto and A. Shimada. Bull. Chem. SOC. Japan 1967,40,1296. Part (u) X-Ray Crystallography acid is 6.7".43Overcrowding in 2,6-dimethylbenzoic acid causes the carboxy- group plane to be rotated 53"out of the benzene-ring plane and there are also small displacements (0.01 and 0-03A) of the methyl carbon atoms from the aromatic plane.44 Both polymorphs of terephthalic acid have been investigated. In one polymorph which was thoroughly refined:' the bond lengths indicate a small amount of quinonoid character in the benzene ring.The carboxy-group planes are rotated about the exocyclic C-C bond 5" out of the benzene plane. The structure of ammonium acid o-carboxybenzene sulphonate NH4+ (C6H4*C02H*S0,) consists of ammonium ions and infinite chains of o-carboxybenzene sulphonate iyns with intermolecular 0-H 0 bonds 0.. (264 A) between carboxy- and sulphonate groups.46 The carboxy-group plane makes an angle of 50.7" with the benzene plane; other effects of steric interaction include shifts of the sulphur (-0.06 A) and carboxy-carbon (+0.13 .$) from the benzene plane and deviations of ring bond-angles from normal values. Crystals of orthanilic acid NH,f *C,H4*S0 have an N--H-*O bond system which contains a 'bifurcated' bond of intra- and inter-molecular character47 (the hydrogen-atom positions were confirmed by neutron diffraction analysis of the Okl projection).The residual electron density map contains features attributable to bonding electrons; for the majority of C-C bonds these consist of dumbell shaped peaks which extend approximately 1 .$ above and below the plane of the benzene ring. P-Sulphanil- amide has been studied4* by X-ray and neutron diffraction. The bond lengths in the molecule suggest that there is a small but significant contribution of quinonoid resonance form to the structure. The distribution of residual electron density within the benzene ring is explained in terms of effects resulting from electron redistribution at bonding.The hydrogen-bond system closely resembles that found in a-sulphanilamide and the N-H**.O bonds are about 0.25 A longer than in related zwitterion compounds. The dimensions of the two structurally distinct molecules of p-aminobenzoic acid in the crystal asymmetric unit also suggest a small amount of quinoid character.49 Both amino- and carboxy-groups are displaced slightly from the benzene-ring planes. Me 1 43 G. Ferguson and K. M. S. Islam Acta Cryst. 1966,21 1000. 44 R. Anca S. Martinez-Carrera and S. Garcia-Blanco Acta Cryst. 1967 23 1010. 45 M. Bailey and C. J. Brown Acta Cryst. 1967 22 387. 46 Y. Okaya Acta Cryst. 1967 22 104. 47 S. R. Hall and E. N. Maslen Acta Cryst. 1967 22 216. 48 A. M. O'Connell and E. N. Maslen Acta Cryst.; 1967 22 134.49 T. F. Lai and R. E. Marsh Acta Cryst. 1967 22 885. 72 George Ferguson The tricarbonylchromium o-toluidine molecule adopts the eclipsed con- figuration (8) in which the carbonyl-chromium vectors point closely towards the benzene carbon atoms which are ortho and para to the amino-~ubstituent.~~ The configuration (9) of tricarbonylchromium methylbenzoate is that in which the carbonyl-chromium vectors point towards the 1-,3-,and 5-positions of the benzene ring.51 These results support the proposition that the orientation of the tricarbonylchromium moiety in a tricarbonylchromium substituted benzene reflects the electron-withdrawing or -releasing character of the benzene substi tuen t. The o-nitro-groups in 2,4,6-trinitrophenetole are rotated out of the phenyl ring one by 62" and the other by 33".52The p-nitro-group is almost coplanar with the phenyl ring (rotation angle 3") and the plane of the OEt group is almost perpendicular (88") to the phenyl ring plane.The structure of the Meisenheimer complex of 1,l-dimethoxy-2,4,6-trinitrobenzene K+ [C6H2(N02),(OMe)2]-,2H,0 has been determined.53 Although the angle C(2)-C(l)-C(6) in the benzene ring is 110",the ring is nearly planar (largest deviation 0.03 A). The nitro-groups at the 2-,4- and 6-positions are twisted 6",6" and 11" out of the benzene plane. None of the substituent groups of N-methyl-N-2,4,6-tetranitroaniline (tetryl) is coplanar with the benzene ring.54 The angles made by the nitro-group planes with the aromatic ring are 25" 23" and 44"; the nitramine-group plane is inclined at 65" to the ring.The non-planarity of the 4-nitro-group may be caused in part by the formation of an intermolecular hydrogen bond between a benzene-ring hydrogen and an oxygen atom of a neighbouring nitramine group. 1,3-Dichloro-2,4,6-trinitrobenzenehas crystallographic two-fold symmetry; the nitro-group in the 2-position is rotated 75" from the plane of the benzene ring and the 4-nitro-group through 37°.55In contrast the nitro-groups of 1,3-diamino-2,4,6-trinitrobenzene are not rotated out of the aromatic plane ;56 there are apparent intramolecular hydrogen bonds between adjacent amino- f and nitro-groups (N-*.O 2.52-260 A). The benzene ring is distorted so as to relieve the overcrowding of the planar conformation.The structure analysis of N,3-dimethyl-4-bromo-2,6-dinitroaniline shows that the N-methyl group is in contact with the 2-nitro-group thus confirming earlier n.m.r. and i.r. res~lts.~ ' The bonding to the exocyclic carbon atom in potassium p-nitrophenyl- dicyanomethide (10) is ~yrarnidal.~' The cyano-groups and the nitro-group are significantly displaced from the phenyl-ring plane in the same direction. " 0.L. Carter A. T. McPhail and G. A. Sim J. Chem. SOC.(A),1967 228. 51 0.L. Carter A. T. McPhail and G. A. Sim J. Chem. Soc. (A),1967 1619. 52 C. M. Gramaccioli,R. Destro and M. Simonetta Chem. Comm. 1967,331. 53 H. Ueda N. Sakabe J. Tanaka and A. Furusaki Nature 1967,215,956. 54 H. H. Cady Acta Cryst. 1967,23 601.55 J. R. Holden and C. Dickinson J. Phys. Chem. 1967,71 1129. 56 J. R. Holden Acta Cryst. 1967,22,545. " S. Abrahamsson M. Innes and B. Lamm. Acta Chem. Scand. 1967,21,224. 58 R. L. Sass and C. Bugg Acta Cryst. 1967,23 282. Part (u) X-Ray Crystallography K+ 9-Dicyano-2,4,7-trinitrofluorene(11) is over~rowded.~' A slight twist of the fluorene framework into a propeller shape leaves outer rings out of parallelism by 3.2". The 4-nitro-group is rotated out of the molecular plane by 19.3". The force tending to maximise the resonance energy of this group is believed to be responsible for the intramolecular overcrowding of the observed conformation the slight deviation from planarity of the fluorene framework and angular distortion in C-C-N bond angles.10-Dicyanomethylene anthrone (12) is forced by considerable steric strain into a non-planar configuration which apparently accounts for its inability to form complexes.60 The molecular conformation which closely resembles that of bianthronylidene involves a bend of 28" in the anthracene skeleton a dihedral angle of 144" between the planar dicyanomethylene group and the anthracene skeleton and a twist or non-parallelism of 5" out-of-plane between the direction defined by the 9,lO- carbon atoms and the corresponding directions in the outer rings. A clearance of 2.848 between overcrowded carbon atoms is thus achieved. In crystals of propargyl2-bromo-3-nitrobenzoate there is an intermolecular H(ethynyl)*.- O(carbony1) distance of 2.398 and the C-H -**O angle is 156".61 When the ester is in dilute solution in carbontetrachloride the C-H stretching frequency of the ethynyl group is at 3313 cm.- '.In the solid state the acetylenic C-H stretch appears at 3266 cm.-'. Both the carboxy- and the nitro-groups are rotated considerably out of the plane of the benzene ring the former by 53" and the latter by 60". A C-H***O hydrogen bond is also found in the crystal structure of ~-oxo-bis[chlorobis(pentane-2,4-dionato)titanium(1v)] chloroform solvate.62 The chloroform molecule apparently takes part in a 59 J. Silverman A. P. Krukonis and N. F. Yannoni Acta Cryst. 1967,23 1057. 6o J. Silverman and N. F. Yannoni J. Chem. SOC.(8, 1967,194. 61 J. C. Calabrese A. T. McPhail and G. A. Sim J. Chem.SOC.(B) 1966 1235. 62 K. Watenpaugh and C. N. Caughlan. Inorg. Chem. 1967.6,963 George Ferguson bifurcated C-H *-*Ohydrogen bond to two oxygens of acetylacetonate groups. The C(chloroform)***O distances are 3.21 and 3-35 A. PhCOSCPh II N-OH (13) NO The accepted structures for benzil a-(13) and p-monoxime have been con- firmed by an X-ray analysis of a p-bromobenzoate derivative of the a-mon- ~xime.~~ The cation side-chain of noradrenaline hydrochloride is maximally extended and the catechol moiety is planar.64 In crystals of 2-bromo-1,l-di- p-tolylethylene the atoms are distributed in three planes; the ethylene plane and the two tolyl rings which are rotated by 24.4" (trans to Br) and 67.9" (cis to Br).65 A variety of reducing agents convert o-benzoylbenzoic acids and their acid chlorides to 3,3'-diarylbiphthalidyls(14).Two isomers the racemic (DL) (m.p. 226") and the meso (m.p. 262") forms are produced in unequal amounts. The more abundant (meso)form has been shown to exist in the solid state in the fully staggered conformation by an analysis of the dibromo-derivative (14,R = Br).66 The red compound obtained by coupling diazotised 2-nitro-4- toluidine and acetoacetanilide at higher temperatures has been shown to be a-(4-methyl-2-nitrophenylazo)acetanilide.The molecule is planar and the configuration of the carbon atom to which coupling occurs is trig~nal.~~ The molecular arrangement in crystals of the iodide of NN'-diphenyl-p-phenylene-diamine and in the perchlorate of 4,4'-bis(dimethy1amino)diphenylamine radicals have been reported.68 The molecular configuration around N(2) in the 2,2-diphenyl-l-picrylhydrazyl radical (15) is approximately trigonal planar with the phenyl groups twisted at angles of 49 and 22".69 The bond at the digonal nitrogen is bent with a C-N-N angle of 118.5" and the attached picryl carbon is twisted 28.5" out of the plane of the trigonal hydrazine nitrogen ; this non-planarity of the hydrazyl backbone was not predicted from e.p.r.measurements. The picryl-ring plane is further inclined at an angle of 33" to the plane of the digonal nitrogen. The o-nitro-groups are K. A. Kerr J. M. Robertson G. A. Sim and M. S. Newman Chem. Comm. 1967 170. 64 D. Carlstrom and R. Bergin Acta Cryst. 1967,23 313." G. L. Casalone C. Mariani A. Magnoli and M. Simonetta Acta Cryst. 1967,22 228. H. Manohar V. Kalyani M. V. Bhatt and K. M. Kamath Tetrahedron Letters 1966 5413 V. Kalyani H. Manohar and N. V. Mani. Acta Cryst. 1967,23 272. '' C. J. Brown J. Chem. SOC.(A),1967,405. 68 K. Toman D. Ocenaskova and K. Huml. Acta. Cryst. 1967,22,29 32 69 D. E. Williams J. Amer. Chem. SOC. 1967,89,4280. Part (u) X-Ray Crystallography twisted out of the picryl ring plane by 25" and 55" while the p-nitro-group is twisted by 13". The hydrazyl N-N bond distance is 1.334 A intermediate between expected values for single and double N-N bonds. /o Me -C< ? Me-ol@) I k0 Fe I (16) In 5-acetoxy-6-methoxy-8-nitroquinoline(16) the exocyclic C-0-C angles average 117" instead of tetrahedral owing to steric repulsions between ring atoms and those of the methoxy and acetoxy-gro~ps.~~ Packing and intramolecular steric requirements force the nitro- and acetoxy-groups to rotate out of the ring plane by 59" and 79" respectively.The resonance energy of the acetoxy-group is sufficient to maintain its planarity in the overcrowded environment but the quinoline deviates slightly from planarity. The cyclo- pentadiene rings in 2-biphenylylferrocene (17) are eclipsed ; the first phenyl ring of the biphenylyl group is rotated 43" out of the cyclopentadienyl plane and the outer phenyl ring is rotated 58" out of the plane of the first phenyl ring.71 These rotations relieve the strain which would exist in a planar model for the C,H4*C6H4*C6H,group.2-Chloro-l,8-phthaloylnaphthalene(18) is folded about a line joining the carbonyl carbon atoms and is distorted by the chlorine atom. The naphthalene plane makes an angle of 42" with that of the phthaloyl ring. The strain imposed by the seven-membered ring closing the peri-positions of the naphthalene moiety is accommodated largely by an increase of the valency angle C(l j-C(S)-C(S) from 121.8" in naphthalene to 126-1"and by extension of the C(l)-C(9) and C(SkC(9) bonds from the established value (1.421A) for naphthalene to 1.443and 1-456A re~pectively.'~ 'O M. Sax and R. Desiderato Acta Cryst. 1967,23 319. '' J. Trotter and C. S. Williston J. Chem. SOC.(A) 1967 1379. 72 K. M. S. Islam and G. Ferguson J. Chem.SOC.(B),1967 1134. 76 George Ferguson The crystal structure of p-benzoquinone 2,3-,73 2,5- and 2,6-dimethyl- benz~quinone,~~ (dur~quinone),~~ 2,3,5,6-tetramethyl-1,4-benzoquinone and thymoquinone and their solid-state photochemistry have been discussed76 in terms of their packing arrangements. The structures of the dimers (cyclo- butanes and oxetans) of the 2,5- and 2,6-dimethyl-p-benzoquinonescan be related to the packing geometry in the monomer crystals of parallel double bonds (C==C C==O) with centre-to-centre distances up to 4.3 A. This cor- relation is unsatisfactory for the 2,3-dimethyl derivative. The absence of short parallel contacts in p-benzoquinone and duroquinone may account for the inability of these two quinones to yield dimers.It is postulated that the formation of polymeric material is related to the presence of short contacts between carbon atoms of nearest neighbour non-parallel >C=C < groups whose interaction leads to open-chain dimer diradicals ;these cannot terminate by ring closure but may instead initiate polymerisation. A number of centro- symmetric hydroquinones and their salts have been examined. Chloranilic acid77 (19) and its dih~drate~~ have ring geometry similar to that in p-benzo- quinone but the C-C single bonds are different in length in the two cases 1.455 and 1.501 A and 1.446 and 1.512 I$,respectively and there are two kinds of C-0 distances corresponding to C=O and C-OH bonds. By contrast the carbon ring-systems in ammonium chloranilate m~nohydrate,~' ammonium nitranilate,*' and nitranilic acid hexahydrate (hydronium nitranilate)* are not (20) in quinoidal form.In each case four C-C bonds are of equal length and two are considerably longer (mean values are 1.404 and 1.535 A 1-435 and 1-551A and 1.419 and 1.559A respectively) and the C-0 distances are ofequal length (1.248 1.220 and 1.228 8 respectively). In ammonium nitranilate the nitro-group plane is rotated 6"around the C-N bond out of the carbon plane; in nitranilic acid hexahydrate the corresponding angle is 22". -R HO c1@ 0 J' (19) (20) 53 D. Rabinovich J. Chem. SOC.(I?) 1967 140. 74 D.Rabinovich and G. M. J. Schmidt J. Chem. SOC. (4,1967,127. '' D.Rabinovich G. M. J. Schmidt and E. Ubell J. Chem. SOC.(4, 1967,131.76 D.Rabinovich and G. M. J. Schmidt J. Chem. SOC.(B) 1967,144. 77 E.R.Andersen Acta Cryst. 1967,22 188. 78 E.K.Andersen Acta Cryst. 1967,22 191. 79 E.K.Andersen Acta Cryst. 1967,22 196. E. K. Andersen Acta Cryst. 1967,22,201. E.K.Andersen Acta Cryst. !967,22,204. Part (u) X-Ray Crystallography 77 A refinements2 of the anthraquinone structure rules out the possibility of a formal transfer of one electron from carbon to oxygen in the C=O bond as proposed earlier.83 There is a strong asymmetric O-H***O hydrogen bond in 1,5-dihydroxanthraquinone (21) and electron delocalisation across the quinonoid ring as gauged by the structural parameters appears minimal.84 The anthraquinone nucleus in NN'-diphenyl-1,5-diaminoanthraquinone is planar while the substituent benzene rings are inclined to the central ring system by 62.4"." There is the possibility of intramolecular hydrogen bonding between nitrogen and oxygen atoms which are 2602 8 apart.NN'-Diphenyl- 1,8-diaminoanthraquinonehas two-fold crystallographic symmetry and the phenyl rings are inclined by 61.8"to the anthraquinone nucleus which itself is not quite planar.86 Here too there is the possibility of intramolecular hydrogen bonding between the nitrogen atoms and the adjacent carbonyl oxygen (N-H -02.578A). The anthraquinone nucleus of 1,5-dinitro-4,8-dihydroxy-anthraquinone is approximately planar but the nitro-groups are inclined to it at 88". An intramolecular O-H-*O hydrogen bond is possible (O.*-O 2549 A) but the hydrogen atom was not located.87 Other anthraquinone structures reported include the 1,2- and 1,5-dihydro~y-9,10-anthraquinones~~ and 2,3-di bromo-l,4-an thraquinone.89 The dimer l0,lO-dian thronyl formed by oxidation of anthrone has a two-fold symmetry axis perpendicular to the C(l0)-C((l0) bond. The planes of the two half molecules make an angle of 145" with each other and each half molecule has a bend of 10". The dihedral angle C(9) C(10 jC(lO')C(9') is 74". All bond lengths and angles are normal except the C(LO)-C(lO)' bond (1.60 & 0.006 A).'' Other Cyclic Molecules.-Two derivatives of cyclopropenylidene-cyclo-pentadiene (calicene) (22) have been examinedg '9 92 and yield comparable results for the molecular geometry confirming the calculation^^^ of Dewar and Gleicher which predict strong bond fixation and lack of aromatic character in the calicene system.The bond lengths in the ring of the p-bromo- phenacyl ester of 7,7-dimethylcycloheptatriene-3-carboxylic acid (thujic acid) alternate and it adopts a boat conformation. Thus a norcaradiene arrangement and the planar pseudo-aromatic structure previously proposed for the ring are precluded.94 The molecular geometry implies a small twist in 1,2- and 82 A. Prakash Acta Cryst. 1967,22,439. 83 B. V. R. Murty Z. Krist. 1960 113,445. 84 D. Hall and C. L. Nobbs Acta Cryst. 1966,21,927. 85 M. Bailey and C. J. Brown Acta Cryst. 1967,22,488. 86 M. Bailey and C. J. Brown Acta Cryst. 1967,22,493. M. Bailey and C. J. Brown Acta Cryst. 1967 22 392.J. Guilhem Bull. SOC.chim. France 1967 1656. 89 J. Gaulthier,S. Geoffre and C. Hauw Compt. rend. 1967,264 C,697. 90 M. Ehrenberg Acta Cryst. 1967,22,482. 91 H. Shimanouchi T. Ashida Y. Sasada and M. Kakudo Tetrahedron Letters 1967 61. 92 0.Kennard D. G.Watson J K Fawcett K. A. Kerr C Romers Tetrahedron Letters 1967 3885. 93 M. J. S. Dewar and G. J. Gleicher Tetrahedron 1965 21 3423. 94 R. E. Davis and A. Tulinsky J. Amer. Chent. SOC.,1966,88 4583. George Ferguson 5,6-double bonds. Considerable bond alternation has also been observedg5 in the conjugated seven-membered ring of 8,8-dicyanoheptafulvene (23). A planar (centrosymmetric) cyclobutane ring is found in trans-cyclobutane- 1,3-dicarboxylic acid96 but in the ~is-isomer~~ the ring is puckered with a dihedral angle of 149".The crystal structures of both compounds consist of chains of hydrogen-bonded molecules. In the trans isomer the C-C ring bonds are 1-56? and 1.552 A; the mean C-C ring bond in the cis-isomer is 1.554 A. In crystals of 1,4-truns-cyclohexane-dicarboxylicacid the molecule has a centre of symmetry and hence adopts the chair conformation. The mean C-C4 angle in the cyclohexane ring is 112".98 The erythro-configuration has been determined for 2-(cl-bromophenyl-~-nitroethyl)cyclohexanone obtained from the hydrolysis of the reaction between cyclohexanone enamine and j3-nitro-p-bromostyrene. The cyclohexanone ring has a chair conformation with the carbonyl oxygen nearly eclipsed to the adjacent exocyclic carbon atom.99 The main reaction product of the hydrogenation of 4,6-dinitro- pyrogallol on a rhodium-platinum catalyst is a 1,3-diamino-4,5,6-trihydroxy-cyclohexane.Its dihydrochloride has the all-cis configuration and is a 2-deoxy-cis-inosa- 1,3-diamine. loo (24) (25) (26) cis-cis-cis-1,4,7-Cyclononatetraeneis known to adopt a crown conformation and when complexed with silver nitrate to give a n-complex.of composition C9H1,(AgN0,)3 it retains this conformation having crysta1logr:phic 3/m (C3J symmetry.lo' The conformation of all-cis-1,6-dichlorocyclodeca-1,3,6,8-tetraene (24) is such that each half of the centrosymmetric molecule may be defined by two planes one through C(lO),C(l),C(2) and C(3)Band or& through C(2),C(3),C(4) and C(5) inclined at an angle of 57" to each other.'02 This 95 H.Shimanouchi T. Ashida Y. Sasada M. Kakudo I. Murata and Y. Kitahara Bull. Chem. SOC.Japan 1966,39,2322. 96 T. N. Margulis and M. S. Fischer J. Amer. Chem. SOC.,1967,89,223. '' E. Adman and T. N. Margulis Chem. Comm. 1967,641. 98 J. D. Dunitz and P. Strickler Helv. Chim. Acta 1966,49,2505. 99 M. Calligaris F. Giordano and L.Randaccio Ricerca xi. 1966,36 1333. loo J. H. Palm Acta Cryst. 1967,22,209. R. B. Jackson and W. E. Streib J. Amer. Chem. SOC.,1967,89,2539. lo2 0.Kennard D. G. Watson J. K. Fawcett and K. A. Kerr. Tetrahedron Lettcrs. 1967. 3129. Part (u) X-Ray Crystallography 79 arrangement allows for a separation of 2.5 A between the two inner hydrogens at C(5)and C(l0). The tensions exerted by the silver ions on the double bonds of the germacratriene-AgNO complex (25) and some non-planar deforma- tions have combined to distort the ring much past what can be achieved by a Dreiding model.lo3 The planes of the double bonds are approximately per- pendicular to the plane of the macrocycle (26).The results of the X-ray analysis of 1,1,5,5-tetramethylcyclodecane-8-carboxylicacid do not correspond to a reasonable molecular structure.lo4 When used as a starting point for a strain- energy minimization calculation based on semi-empirical potential functions they lead to two new conformations for the cyclodecane ring.It is shown that the X-ray results can be explained by assuming that the crystal consists of a random mixture of these two conformations.lo' 4 Br (27) (28) (2 9) Bridgehead substituted polycyclics are known to exhibit widely differing solvolysis rates. The explanation of this variation based on changes in methy- lene bridgehead angle is not acceptab1e,lo6 and is further denied by crystal structure analysis of 3-exo-(N-benzyl-N-methylaminomethyl)-2-endo-norborn-anol (27). The angles in the bicycloheptane system are considerably less than tetrahedral reflecting the strain in the system.'07 As closely similar results were obtained from a neutron diffraction study of 3-endo-phenyl-2-endo- bornanol'08 and a gas-phase electron diffraction study of the parent nor-bornane'Og it appears that the geometry of the norbornane skeleton is especially insensitive to the nature and the position of substituents.Molecular geometry rn is demanded of pseudotropine (28) by its space group.'" The N-methyl group is equatorially attached with respect to the piperidine ring which is in deformed chair form such that the separation of the ethylene bridge atoms from C(3) and the N-methyl group is 3 A. Atoms C(1) and C(1') are fairly close; the angle C(l)-N-C(l)' is only 102.5". Most of the bond angles at the carbon atoms forming the tricyclic system of 2-anilino-3-bromo- lo3 F. H. Allen and D. Rogers Chem. Comm. 1967,588. lo4 J. D. Dunitz and H. Eser Helu. Chim. Ada 1967,50 1565. lo' J. D. Dunitz H. Eser M. Bixon,and S. Lifson Helv. Chim. Acta 1967 50 1572; M. Bixon H. Dekker J. D. Dunitz H. Eser S. Lifson C. Mosselman J. Sicher and M. Svoboda Chem.Comm. 1967 360. lo' A. C. Macdonald and J. Trotter Acta Cryst. 18 243 1965; ibid. 19 456 1965. lo' A. V. Fratini K. Britts and I. L. Karle J. Phys. Chem. 1967 71 2482. lo' C. K. Johnson Abstracts Amer. Chem. SOC. Meeting Atlanta Georgia 1967. log W. C. Hamilton Ph.D. Thesis California Institute of Technology 1954. 'lo H. Schenk C. H. MacGillavry S. S. Kolnik and J. Laan Acta Cryst. 1967 23 423. 80 George Ferguson tetrahydro-exo-dicyclopentadiene(29) are somewhat smaller (mean value 104") than tetrahedral the angle at the bridge carbon atom being 94O.l" Br.C,H,COO The formulation of the tricycl0[2,1,0,0,~* 'Ipentane system has been con- firmed by an. X-ray analysis of 1,5-diphenyltricyclo[2,1,0,0,2~ 5]pent-3-yl p bromobenzoate (30).The ring system is under considerable strain there being six C-(2-42 angles of about 60°,three of about 80" and four of about 90". The C(2)-*C(4) non-bonded distance is only 1.99 A.112 Six of the C-C distances are in the range 1.50-1.54 .$; the bond common to the two three- membered ring is 1-44A but the difference may not be significant. The tricyclo- dodecyl system of 12-hydroxy-6-methyltricyclo[5,3,1,12~ 6]dodec-3-yl p-iodo- benzoate (31) adopts a distorted double-chair conformation because of mutual repulsion between methylene groups at C(4) and C(11) and between the 12- hydroxyl and C(9) methylene groups.' ' The low temperature phase of adaman- tane (32) has been reported as departing significantly from 43m symmetry.'14 However another refinement' l5 of the structure starting with molecules having 33m symmetry and tilted 9" about the c axis refines to a structure not significantly different from the new starting point.On the other hand if the refinement is started at the published structure the parameters of which correspond to rather distorted molecules a structure not significantly different from this starting point results although the final R is higher 8.7 as compared with 3.1 % for the symmetrical structure. Donohue and Goodman point out that the occurrence of a false minimum which is quite close to a true minimum is rather disturbing!'I5 c1 Cl (3 3) (35) N. Tanaka T. Aishda T. Sasada M. Kakudo and N. Kasai Bull. Chem. SOC.Japan 1967 40,1574. J. Trotter,C. S. Gibbons N.Nakatsuka and S. Masamune J. Amer. Chem. SOC. 1967,89,2792. '" G. Ferguson; W. D. K. Macrosson J. Martin and W. Parker Chem. Comm. 1967 102. C. E. Nordman and D. L. Schmitkons Acta Cryst. 1965,18 764. J. Donohue and S. H. Goodman Acta Cryst. 1967,22,352. Part (0)X-Ray Crystallography 81 An unusual cage structure (33) results when hexachlorocyclopentadiene dimerizes. A chlorosulphonate group was introduced at one of the apex carbon atoms to overcome disorder of the parent perchloro-derivative. The basic cage structure is trans consisting of four cyclopentane and two cyclo- butane rings all of which are 'in the cyclobutane rings the average bond angle is 87". The bond angles of the cyclopentane rings fall into three groups with 96" at the apex and 108" and 102" at the middle and base res- pectively.Two of the isomeric compounds of formula C12Cl12 obtained by pyrolytic reaction of perchloro-3,4-dimethylenecyclobutanehave been ex-amined. '177 ' ' Perchloro-4,8-dimethylenetricyclo-[3,3,2,0'~ ']deca-2,6-diene (34)' l7 has a crystallographic two-fold axis of symmetry. The four-membered ring is slightly distorted and consequently the conformation of the two adjoining dichloromethylene groups is about 8" deviated from the eclipsed form. The five-membered rings are also puckered the dihedral angle between them being 123". Perchloro-3,4,7,8-tetramethylenetricyclo[4,2,O,O2~ ']octane (35)' '' has a rather simple chair-like form with an approximate symmetry 2/m (C,,,) although only a centre of symmetry is demanded by the space group.Each conjugated system in the molecule is nearly planar in spite of a very close contact between the two adjacent dichloromethylene groups. The nearest approach occurs between the chlorine atoms (3.28 A) which is achieved by increasing the overcrowded C-C-Cl valency angles to 126". Systems Containing Hetero-atoms.-The 2-nitronitrosoethane dimer has the two monomer units bound in the trans-configuration through a centre of symmetry at the mid-point of the N-N bond.'lg Bond lengths are N-N 1.315 f 0-010 N-C 1-462& 0.008 and N-0 1.255 & 0-006 A. Molecular orbital calculations for the nitroso-part of the system indicate that resonance structure (36)makes a substantial contribution to bonding in C-nitroso-dimers.The S-N distances (1.58 and 1-63 .$) in 5,5-dimethyl-N-methylsulphonyl-sulphilimine (Me,S -N. S0,Me) indicate a delocalized S-N-S d bond-system.'20 The S-N-S bond angle is 116.2" the average s-0 distance 1.45 A. The CSpr-S IV bond distances (1.74 A) are rather short which can be explained with the supposition of strong hyperconjugation. A refinement of the thiourea structure with accurate intensity data has led to molecular dimensions S-C 1.720 0.009 N-C 1.340 & 0.006A S-C-N 120-5& 0.5 N-C-N 119.0 & 05°.121The hydrogen atoms are approximately coplanar with the heavy atoms. Thioglycollic acid S*[CH,*C02H12 has mirror symmetry with only the sulphur atom on the mirror plane the C-S-C angle being 96".'22 cis-2-Butene-episulphone(37) has approximately M Y.Okaya and A. Bednowitz Acta Cryst. 1967,22,11 I. 117 A. Furusaki and I. Nitta Tetrahedron Letters 1966 6027. A.Furusaki Bull. Chem. SOC. Japan 1967,40,758. 'I9 F. P. Boer and J. W. Turley J. Amer. Chem. Soc. 1967,89,1034. I2O A.Kalman Acta Cryst. 1967,22,501. 12' M.R. Truter Acta Cryst. 1967,22 556. S. Paul Acta Cryst. 1967,23 491. 82 George Ferguson symmetry. The terminal C-C bond lengths are normal (1.51 A) but the central C-C distance of 1.60 _+ 0.02 A is rather longer and the S-C distances of 1-74 and 1.73 A are a little shorter than might have been expected.'23 The other dimensions are S-0 1.41 1.44 8 and 0-S-0 120". In butadiene sulphone (2,5-dihydrothiophen-l,l-dioxide)(38) the sulphur and carbon atoms lie on crystallographic mirror planes and the molecule has mm2 (C2") symmetry.'24 Average bond lengths are S-0 1.44 S-C 1.79 C-C 1.48 C=C 1.308,;the angles are C-S-C 97",0-S-0 llo" and S-C-C 104".A roughly U-shaped cation is found in 1-p-chlorophenyl-5-isopropylbiguanide hydrochloride (paludrine) (39). The biguanide part of the cation consists of two planes of atoms each containing a CN group and intersecting at 58.9". The mean C-N distance is 1.328 8 and as there is no distinction between formal single and double bonds some x-bonding is present ;this was somewhat unexpected in view of the non-planarity of the biguanide moiety.'25 (37) s -s (41) The dihedral angle at the peroxide bond in dibenzoyl peroxide is 91"; a value of 94" was predicted on quantum chemical grounds.The 0-0 bond length is 1.46 8 and the C-0-0 angle is 110" implying that the bonding orbitals in the oxygen atom are sp3 hybrids.'26 1,2-di-(p-chlorophenoxy)ethane crystallises in a cell containing two crystallographically independent molecules which are of slightly different conformations (the OCH2-CH,O torsional angles are 66 and 81") and are located on two-fold axes in the ~rysta1.l~~ The parent compound 1,2-diphenoxyethane also has two-fold symmetry and the R. Desiderato and R. L. Lass Acta Cryst. 1967,23,430. D. E. Sands and V. W. Day Z. Krist. 1967,124,220. C.J. Brown J. Chem. SOC.(A),1967,60. 126 M. Sax and R. K. McMullan Acta Cryst. 1967,22,281. N. Yasuoka T. Ando and S. Kuribayashi Bull. Chem. SOC.,Japan 1967,40,265.Part (u) X-Ray Crystallography 83 OCH,-CH,O torsional angle is 67°.128 The torsional angle CS-SC in 2,2'-diaminodiphenyl disulphide is 87".12' Tetraethylthiuran disulphide (disulphiram Antabuse) (40) is an antioxidant that interferes with the normal metabolic degradation of alcohol in the body and a crystal structure analysis i /" has established that the molecule contains two planar -S-C-N 'cgroups which have a dihedral angle of 96". The configuration about each nitrogen atom is planar rather than pyramidal and the two C-N bond lengths adjacent to the C=S bonds are both very short 1.33 and 1.36 A. Thus there is strong evidence for the influence of the C=S bond on the C-N bond length.130 A literature survey'31 indicates that in a disulphide group the normal dihedral angle of about 90" between the valencies of the two sulphur atoms corresponds to S-S bond lengths of about 2.03 A while smaller dihedral angles give longer bond lengths.The length difference is assumed to be due partly to lone-pair repulsion which is most pronounced when the dihedral angle is 0" and partly to n-bonding which is most pronounced when the dihedral angle is 90". A graph showing this trend indicates that 2.lOA is the length of a single bond between two divalent sulphur atoms when the dihedral angle is 0". A linear bond-length-bond-order curve for sulphur-sulphur bonds in cis-planar disulphide groups is proposed based on the double-bond length of 1.89 8 and the single-bond length of 2.10 A. The S-S distances found in 4-phenyl-l,2-dithiolium iodide (2.028 It 0.010 A),132thiuret hydrochloride (2.071 f 0.004 A),1333,5-diacetamide-l,2-dithioliumbromide (2.080 f 0.005 A),134 and sodium 3,5-diacetylimino-1,2-dithioltrihydrate (2.05 0.015 A)13s show that the rings are stabilised through n-orbital delocalisation extending over the sulphur-sulphur bonds.No four atoms of the five membered ring in ~~-6-thiocetic acid (41) are coplanar. The torsional angle CS-SC is near 96" and the S-S bond length is normal (2.01 A).136 The discovery of the thio- thiophen 'no bond resonance' system through an X-ray analysis'37 of the dirnethyl derivative (42) has prompted X-ray analyses of the unsymmetrical derivative (43).13* Whereas in the symmetrical molecule the S-S distances are identical (2.35 A) in the unsymmetrical molecule S(lFS(2) 2.51 and S(3) 2.22 A.Similar results [S(l)-S(2) 2-52 and S(2)-S(3) 2.18 A] have been found for the unsymmetrical derivative (44).'39 12* N. Yasuoka T. Ando and S. Kuriyabashi Bull. Chem. SOC. Japan 1967,40,270. A. H. Gornes de Mesquita Acta Cryst. 1967,23 671. I3O I. L. Karie J. A. Estlin and K. Britts Acta Cryst. 1967 22. 273. 13' A. Hordvik Acta. Chem. Scand. 1966,20 1885. A. Hordvik and E. Sletten Acta Chem. Scand. 1966 20 1874. 133 A. Hordvik and J. Sletten Acta Chem. Scand. 1966,20 1907. 134 A. Hordvik and H. M. Kjoge Acta Chem. Scand. 1966,20,1923. 135 A. Hordvik and E. Sletten Acta Chem. Scand. 1966,20 2043. 136 I. L. Karle J. A. Estlin and K. Britts Acta Cryst. 1967 22 567.13' S. Bezzi M. Mamrni and C. Garbuglio Nature 1958,192 247. A. Hordvik E. Sletten and J. Sletten Acta Chem. Scand. 1966,20 2001. 139 S. M. Johnson M. G. Newton I. C. Paul R. J. S. Beer and D. Cartwright Chem. Comm. 1967 1 170. George Ferguson Ph PhCO OH (&;N H (45) The conformation of the bond system 0-C-C=N-in anti-fufuraldoxime is trans (45).The side chain is turned away less from the plane of the furan ring than expected which leads to a rather short intramolecular distance 2.86 A between the oxime oxygen atom and a ring carbon atom. Hydrogen bonds -N-0-H *N-0-H -link the molecules in infinite chains along the screw axis. This type of molecular association appears to be characteristic for aromatic anti-aldoximes.The nearly planar molecules of trans-P-2-furyl- acrylic acid (46) are pairwise hydrogen-bonded at the carboxyl groups (O-H***O 2.67 A); these pairs are arranged in stacks in which parallel ,C=C \/\groups make contacts of 3.84 8 and is probably responsible for the photochemical formation of the dimer of symmetry rn (p-truxinic acid analogue). U.v.-induced polymerisation may occur along a reaction path involving the exocyclic ,C=C \/groups and double bonds of the furan \ ring.14' According to their cell dimensions as well as the geometry of contact of such exocyclic groups the two modifications of trans-P-2-thienylacrylic acid belong to the p-and y-type of the cinnamic acid series and the photochemistry of the two forms is discussed in terms of their packing arrangement^.'^^ a-Chloro-6-valerolactam adopts a half-chair conformation in the solid state with the chlorine atom equatorial; the peptide linkage NH-CO has the cis-c~nfiguration.'~~ The six-membered ring of trimethylene sulphite (47) has a chair conformation with an axial S=O bond.The mean single-bonded S-0 distance is 1.60 A and the S=O distance is 1.45 A. The 0-S=O angles are 107" and the 0-S-0 angle is 100". Within experimental error the molecule has symmetry rn.144 A chair conformation is also adopted by trans-2,3-dichloro-1,4-thioxan (48)with the chlorine atom axial. Its overall geometry is midway between the conformations of the corresponding 2,3-dichloro- derivatives of 1,4-dioxane and 1,4-dithiane.'45 Thioformaldehyde trimer B.Jensen and B. Jerslev Acta Chem. Scand. 1967,21,730. 14' S. E. Filippakis and G. M. J. Schmidt J. Chem. SOC.(B) 1967,229. 142 S. Block S. E. Filippakis and G. M. J. Schmidt J. Chem. SOC.(B) 1967,233. 143 C. Romers E. W. M. Rutten C. A. A. van Driel and W. W. Sanders Acta Cryst. 1967,22,893. 144 C. Altona H. J. Geise and C. Romers Rec. 'Rav. chim. 1966,85 1197. 145 N. de Wolf C. Romers and C. Altona Acta Cryst. 1967,22 715. Part (u) X-Ray Crystallography [CH,SI3 has a chair conformation with the mean S-C distance 1.818 A.146 Chair conformations are also found in 3,3,6,6-tetrabromomethyl-1,2,4,5-tetr~xane’~’ 14* and 1-thia-4-selenocyclohexane-4,4-dibromide. II c1 S c1 (47) (48) N-0 H ,,VH (53) (54) The 1,3-dithiete ring of the desaurin (49) from acetophenone must be planar (the molecule is centrosymmetric) and the ap-unsaturated carbonyl system is in an s-cis conformation.The ten central atoms of the molecule are closely planar and associated with this is a short S***0 distance of 2.64 1$ suggesting a sulphur-oxygen attraction. 149 5-Aminotetrazole is planar and exhibits a large degree of conjugation not implied by the molecular formula (50). The bond lengths are N(l)-N(2) 1.381 N(2)-N(3) 1.255 N(3)-N(4) 1-373 N(4)-C 1,321 N(1)-C 1-329 and N(6)-C 1.577 A.150The anions of 3-substituted 5-phenylrhodamines are S-alkylated by alkyl halides and the formulation of the resulting compounds as the ‘mesionic’ anhydro-2-alkylthio- 3-R-4-hydroxy-5-phenylthiozolium hydroxides has been confirmed by X-ray analysis.The compounds are best represented as betaines with the positive charge on the nitrogen atom and the negative charge on the oxygen (51).’5’ Molecules of NN‘-bisuccinimidyl have a crystallographic two-fold symmetry axis parallel to the N-N bond and non-crystallographic 222 symmetry; the 146 J. E. Fleming and H. Lynton Canad. J. Chem. 1967,45 353. 14’ M. Schulz K.Kirsche and Hohne Chem. Ber. 1967 100,2242. 14’ L. Battelle C. Knobler and J. D. McCullough Inorg. Chem. 1967,6 958. 149 T. R. Lynch I P. Mellor. S. C. Nyburg and P. Yates Tetrahedron Letters 1967 373. K. Britts and I. L. Karle Acta Cryst. 1967 22 308. S. Abrahamsson A. Westerdahl G. Isaksson and J. Sandstrom Acta Chem. Scand. 1967,21 442 86 George Ferguson ring planes make a dihedral angle of 65°.152 In crystals of 3,3'-bisisoxazolyl (52) the molecules lie on centres of symmetry and hence have the trans-conformation.Within experimental error the molecule is planar.' 53 An analysis of 5-nitrouracil monohydrate reveals that the 2,4-dioxo-tautomeric (53) form is adopted. The dihedral angle between the plane of the pyrimidine ring and the nitro-group is 4.7". A hydrogen bond C-H*-.O to a nitro- oxygen (H-a.0 2.30 A) appears to be an important structure-determining factor.154 Thiouracil contains highly polarized thionamide groups the polarization of these groups being position-dependent. The C-S bond lengths are 1.645 and 1.685 ( f 0.006) A,suggesting an important contribution of form (54) to the structure.The packing of the molecules in the crystal lattice is dominated by intermolecular S*-*H-N bonds whose lengths (H--S2.39 and 2.78 A) appear to be a function of the degree of polarization about the sulphur atoms.' s5 R' Et (55) The tetrazene chain in the centrosymmetric structure 1,4-bis-(N-ethyl-1,2- dihydrobenzthiazol-2-y1idene)tetrazene(55) is in the trans-N-trans-trans-N form with bond lengths C=N 1-302,N-N 1-400 and N=N 1.257 A. The plane of the tetrazene chain is inclined at 4.8" to the benzthiazole ~1anes.l~~ Bond angles suggest that there is a considerable amount of strain in the five- membered sultone o-hydroxy-a-toluenesulphonicacid as compared to the six-membered soltone 2-o-hydroxyphenylethanesulphonicacid (56; R1= R2 = H).The C-S-0(3) angle 96.1" is 53" smaller than the corresponding value 101.4" in the six-membered sultone and differences are also found in the S-O(3)-C(1) angles of 108-9 and 116.9". The strain introduced into the five- membered ring may be the basis for an explanation of the difference in hydrolysis rates of the two compounds.'57 A detailed analysis has been made of another six-membered sultone 2-o-hydroxyphenyl-1-phenylpropanesul-phonic acid (56; R' = Me R2 = Ph). The heterocyclic ring is in the half-chair conformation with the phenyl group axial and cis to the equatorial methyl group. The C-S-0(3) angle is here 100.2" and the S-O(3)-C(l) angle is 117.1"; some bond lengths are S-O(periphera1) 1.431 S-O(C) 1.595 S-C 1,797 and C\p2-0 1.426A.158 G.S. D. King J. Chem. SOC.(B) 1966,1224. M. Cannas and G. Marongiu Z. Krist. 1967 124 143. lS4 B. M. Craven Acta Cryst. 1967,23,376. 15' E. Shefter and H. G. Mautner J. Amer. Chem. SOC.,1967,89 1249. lS6 R. Allmann Acta Cryst. 1967 22 246. 157 E. B. Fleischer E. T. Kaiser P. Langford,S. Hawkinson A. Stone and R. Dewar Chem. Comm. 1967 197. *" K. Bjamer and G. Ferguson Acta Cryst. 1967,23,654. Part (u) X-Ray Crystallography ~-Thioxanthen-9-ol-lO-oxide (m.p. 206") has the trans-configuration. In the solid state the S-0 bond occupies a pseudo-equatorial position while the HO-CH dihedral angle is 34O.lS9 During the preparation of thiothixene (57) a drug effective against schizophrenia two isomers cis and trans are produced but of these only one exhibits therapeutic activity and is established as the cis-isomer by an X-ray analysis The two aromatic rings of the thioxanthene moiety are inclined at 141.5".160 3-Allyltropolone reacts with bromine in carbon tetrachloride to form a colourless compound CloH,O,Br whose structure has been determined as 2-bromomethyl-2,3-dihydrofuro[2,3-6]tro-pone with double bond alternation as implied in (58).The longest and shortest bonds in the tropone ring are 1.48 and 1.36 8 respectively; the carbon atoms of the tropone ring are planar but the carbonyl oxygen is slightly off (0.07 A) this plane.161 The irradiation of N-chloroacetyl-p-0-methyl-L-tyrosine results in the production of an unusual rearrangement product whose methyl ester has been identified by X-ray methods as (59).Constraints imposed on the molecule by the five- and seven-membered rings twist the molecule from a planar conformation.162 The two nitrogen atoms of the seven-membered ring in 1,4-benzodiazepins are of different basicity. The X-ray analysis of a derivative (60)shows that one of the nitrogens is pyramidal and the other planar. The *C(CO)(NMe)*C-group is nearly planar but is rotated 50" from the adjacent benzo-group. 63 The product C6HloS3 formed when an alcoholic solution of chloroacetone saturated with hydrogen chloride is treated with hydrogen sulphide has been shown to be 2,5-dimethyl-2,5-endo-thio-l,4-dithiane (61). The angle at the 0 Q Me 159 A. L. Ternay Jun. D. W. Chasar and M. Sax J. Org.Chem. 1967,32,2465. 160 J. P. Schaefer Chem. Comm. 1967. 743 16' H. Shimanouchi,T. Ashida Y. Sasada M. Kakudo L. Murata and Y. Kitahara Bull. Chem. SOC.Japan 1967,40,779. 16' I. L. Karle J. Karle and J. A. Estlin Acta Cryst. 1967,23 494. 163 J. Karle and 1. L. Karle J. Amer Chem SOC.1967 89 804. 88 George Ferguson CI I Me (60) Me (61) bridge sulphur atom is 86.3"while at the other two the values are 95.4and 94.7".The S-C bond lengths range from 1.810-1.850A. The molecule con- forms closely to one with a two-fold symmetry axis through the bridge sulphur atom. The methyl hydrogen atoms conform to this symmetry even though they have different intermolecular environments. 64 The molecular structure of the cyclo-adduct from 4-bromo-N-methoxycarbonylazepineand tetracyano- ethylene has been confirmed (62).A refinement of the structure produces anomalous results because of the unexpected presence of a co-crystallized isomer (in approximately 15% of the molecular sites) apparently originating from cyclo-addition involving 3-bromo-N-methoxycarbonylazepineand tetra- ~yanoethy1ene.l~'More details of the molecular geometry of (63)have been reported.166 The compound of empirical formula SNCCl obtained from ammonia and trichloromethanesulphonyl chloride has been shown to be the centrosymmetric tricyclic compound (SNCCl) (64).167 The central [SNC] ring is in a chair conformation but the outer rings are in half-chair conform- ation atoms except the sulphur atom common to the central ring being coplanar.The chlorine atoms are considerably twisted out of this plane to relieve steric strain arising from their close contact (3.17A). One of the so called 'acid products' of 1-methyl-1,4-dihydronicotinamidehas twice the molecular weight of the parent compound and has been shown to be (65).168 The structure contains nine different six-membered rings and very closely A. M. O'Connell Acta Cryst. 1967 23 623. R. A. Smith J. E. Baldwin and I. C. Paul J. Chem SOC.(B),1967 112. *'' M. G. Newton J. A. Kapacki J. E. Baldwin and I. C. Paul J. Chem. SOC.(4,1967 189. A. C. Hazell Acta Chem. Scund. 1967,21,415. H. L. Ammon and L. H. Jensen Acta Cryst. 1967 23 805. Part (u) X-Ray Crystallography resembles the geometry assumed by two twistane16' molecules with one ring [shown by solid circles in (65)] in common.The skeleton bears the same relationship to twistane as congre~sanel~~ does to adamantane.171 An in- vestigation of a dibromo-derivative of eriostoic acid (66) has confirmed the chemically deduced structure and determined the molecular geometry. Small differences which exist between chemically equivalent portions of the molecule are consistent with distortion of the molecule by packing forces.'72 Triclinic tetraphenylporphyrin (67) is centrosymmetric with one pair of pyrrole groups essentially coplanar with the plane of the porphyrin ring and the other pair carrying the central hydrogen atoms inclined at k6.6"to this plane. This has the effect of increasing the separation of the central hydrogen atoms by 0.2 to 2.36 A.173The structure corresponds closely with the hybrid of the two classical resonance forms of the porphyrin molecule.The diacid species of aay6-tetraphenylporphine and apy6-tetra-4-pyridylporphinehave 3 and pseudo-3 symmetry respectively and exhibit the greatest deviation from planarity yet found in any porphyrin This distortion is almost certainly due to mutual repulsion of the four hydrogen atoms on the inner nitrogen atoms. The displacements of the porphyrin atoms from their best plane range from -0.93 to +0.87 A. The geometry and absolute configuration of ethyl-1 -thio-a-D-glucofurano- side (68) have been determined.175 The furanose ring is puckered with C(3) 0.59 8 from the plane of 0(1) C(1),C(2)and C(4),and C(2) -0.60 8 from the OH OH (68) H.W. Whitlock J. Amer. Chem. SOC.,1962,84 3412. 170 C. Cupas P. Schleyer and D. J. Trecker J. Amer. Chem. SOC. 1965,87,917. 17' S. Landa Acta Chem. Acad. Sci. Hung. 1962,31 123. M. G. Paton E. N. Maslen and K. J. Watson Acta Cryst. 1967 22 120. 173 S. J. Silvers and A. Tulinsky J. Amer. Chem. SOC. 1967,89 3331. 174 E. B. Fieischer and A. L. Stone Chem. Comm. 1967 332. 175 R. Parthasarathy and R. E. Davis Acta Cryst.,1967 23 1049. George Ferguson OH OH (73) 0 II plane of 0(1),C(1) C(3) and C(4). The lactone function C-C-0-C of D-galactono-y-lactone (69) is planar; the fifth atom of the lactone ring is 0.64 8 out of this plane forming the puckered furan-type conformation very similar to that found in furanose sugar.The C-0 bond adjacent to the carbonyl group is 0.10 8 shorter than the other C-0 bonds which do not differ significantly from a mean of 1.421 The C(ltO(1)H bond (1.3928,) in P-m-arabinose (70) is significantly shorter than the mean value 1.423 8 for the other C-OH bonds in the molecule. There is no significant difference between the two ring C-0 distances 1.44 These observations are consistent with those from other recent determinations of pyranose sugars. In J3-D-glucurono-y-lactone (71) the rings neither of which is planar are inclined to each other so that the best planes containing four atoms in each make a dihedral angle of 111*3°.178Two of the substituents are endo contrary to the rule that stable derivatives of two fused five-membered rings compounds are those with the minimum number of endo-sub~tituents.'~~ The lactone group in the molecule is not planar having a carbon atom 0.26 8 out of the plane of the C.CO.0 group which is planar.The C-0 bond adjacent to the carbonyl group is again 0.10 8 shorter than other formal single C-0 bonds in the molecule. The a-l,d-linked glucopyranose residues of methyl P-maltopyranoside (72) have the C(1) chair conformation with interatomic distances which are normal for single bonds with the exception of the C-0 G. A. Jeffrey R. D. Rosenstein and M. Vlasse Acta Cryst. 1967,22 725. "'S. H. Kim and G. A. Jeffrey Acta Cryst. 1967 22 537. S. H. Kim G. A. Jeffrey R. D. Rosenstein and P. W. R. Corfield Acta Cryst. 1967,22 733.R. D. Guthrie and J. Honeyman 'Introduction to the Chemistry of the Carbohydrate,' Oxford University Press London 1964. Part (u) X-Ray Crystallography 91 bonds. The methyl f3-glycoside bond C(l’)-O(l’)Me is 1.375A. It is significantly shorter than the mean value of 1.427 A in contrast to the a-glycosidic link joining the two glucopyranoside residues which is 1.416 A. This difference in glucosidic bond lengths appears to be correlated with relative lengths of the C-0 ring bonds which are observed equal in one ring and unequal in the other.’” The primary alcohol group in the a-anomer of the pyranose form of L-sorbose is disordered which leads to an apparent shortening of corres- ponding C-OH bonds. With the exception of those bonds the C-C and C-0 distances do not differ significantly from the mean values of 1.516 and 1.424 8 respectively.lS’ P-Lyxose (73) occurs in the conversion form leq2ax- 3eq4eq with C-C bond lengths in the range 1.509-1.538 C(1)-O(1) 1.364 & 0.006 and the other C-0 bond lengths varying from 1.399-1.435 & 0.006 I$.Nine of the fourteen bond angles in which only carbon and oxygen are involved deviate significantly from the tetrahedral value. Comparison with other nucleosides and nucleotides shows that the glycosidic bond length (1.40 A) in 5-bromo-5-deoxyuridine (74) is shorter than normal and that the conformation of the C(5’)-0(5’) bond is not that most commonly found. In 5-bromouridine the glycosidic bond length (1-49 A) and the C(5’)-O(5’) bond conformation are normaLiS3 In both compounds the uracil base is essentially planar but atom C(1’) is significantly out of the uracil plane; the sugar ring is puckered with C(2’) lying 0.59 and 0.52 I$ out of the plane of the other four ring-atoms (for the deoxyribose and ribose sugar respectively).Complexes.-The molecular adducts LiBr,2NH2 [CH2] NH2 and LiC1,2NH2-[CH2I2-NH2 are isomorphous. The structure consists of infinite molecular chains held together by NH.-X- bonds. There are no Li+-X- bonds. Each lithium atom is tetrahedrally surrounded by four amino-groups from three ethylenediamine molec~les.~ LiCl also forms a complex with 84 pyridine (py) of composition LiC1,2py,2H20. The ability of deoxycholic acid (DCA) to form well-defined molecular compounds with many organic substances is well known.With p-di-iodo benzene (PDIB) crystals of com- position 2DCA PDIB are obtained and the ‘guest’ molecule is held in channels which run parallel to the crystal c axis in the ‘host’ system. With naphthalene benzanthracene and phenanthrene the structures must be disordered. 86 The equatorial isomer of perhydrotriphenylene (PHTP) gives rise to a wide variety of inclusion compounds with different kinds of molecules ranging from those with a nearly spherical (e.g. CCI,) or planar shape (benzene) to linear molecules and macromolecules. The adducts have a channel-like structure with the PHTP molecules arranged in infinite stacks whose axes are parallel to the three-fold axis of the molecules. The structures of three different adducts S.S. C. Chu and G. A. Jeffrey Acta. Cryst. 1967,23 1038. S. H. Kim and R. D. Rosenstein Acta Cryst. 1967,22 648. A. Hordvik Acta Chem. Scand. 1966,20 1943. J. Iball C. H. Morgan and H. R. Wilson Proc. Roy. SOC.,1966 A 295,320. F. Durant P. Piret and M. van Meerssche Acta Cryst. 1967,23 780. F. Durant P. Piret and M. van Meerssche Acta Cryst. 1967,22 52. A. Damiani E. Giglio N. Morosoff R. Puliti and I. Rosen Ricerca xi.,1967,37,42. D 92 George Ferguson have been described:'87 (a) one with a n-hydrocarbon n-heptane (which is isomorphous with those containing n-ethers n-carboxylic acids n-esters and also iso-octane and CCI,) (b)that with chloroform (c) that with cyclo- hexane. In case (a) no coherence is observed between the rows of included molecules and the host structure but in (b)and (c) the presence of simple molecular ratios between the host and guest compounds leads to formation of more complex structures.In the (b) adduct two kinds of non-equivalent included molecules are present; the (c) structure results from the packing of parallel PHTP molecules arranged along helices with a small radius (0.40 A) and nine residues in two pitches with rows of regularly spaced cyclohexane molecules. Other different crystal structures have been observed with dioxan with benzene toluene or bromoform; and with some substituted polybuta- 1,4-dienes. The structural results on two different modifications of pure PHTP showing different stability at room temperature are also reported.In the 2 1 complex of 1,3,7,9-tetramethyluric acid (TMU) and coronene the hydrocarbon molecule is sandwiched between two TMU molecules. The coronene to TMU distance is 3.45 A.188 A similar separation (3.48 A) is found between the molecular planes in the 2:l complex of TMU and 3.4-benz- pyrene. This structure is characterised by a plane-to-plane alternate stacking of two TMU and one benzpyrene molecule arranged in infinite columns parallel to the h axis.'89 Molecular geometry virtually identical with that found in free hexamethylene tetramine"' is found in the hexamethylenetetramine(hex) complex CaBr hex 10H20.191p-Chloro- and p-bromophenol form 2 1192 and 1 :1Ig3 complexes with p-benzoquinone. The component molecules of the isomorphous 2 :1 complexes are stacked plane-to-plane discontinuously in groups of three each group consisting of a quinone molecule between two phenol molecules and held together by charge-transfer forces.The isomorphous structures of the 1:l complexes contain the phenol and quinone molecules stacked alternately plane-to-plane in infinite columns. The structure of the 1:1 tetracyanoethylene :naphthalene complex consists of infinite columns of alternate tetracyanoethylene and naphthalene molecules with molecular overlap such that the molecules are positioned as for a Diels- Alder reaction. The mean perpendicular separation of the molecules is 3-30 A.194. 1,2,4,5-Tetracyanobenzeneforms a 1 :1 complex with naphthalene but the structure is disordered with naphthalene molecules adopting one of the alternative orientations with equal probability.The average interplanar G. Allegra M. Favina A. Immitzi A. Colombo U. Rossi R. Broggi and G. Natta J. Chem. SOC.(B),1967 1020. 18' A. Damiani E. Giglio A. M. Liquori R. Puliti and A. Ripamonti J. Mol. Biol.,1967 23 113. 189 A. Damiani E. Giglio A. M. Liquori and A. Ripamonti Acta Cryst. 1967,23 675. L. N. Becka and D. W. J. Cruickshank Proc. Roy. SOC.,1963 A 273,435. 19' L. Mazzarella A. L. Kovacs P. De Santis and A. M. Liquori Acta Cryst. 1967,22,65. lg2 G.G. Shipley and S. C. Wallwork Acta Cryst. 1967,22 585. 193 G. G. Shipley and S. C. Wallwork Acta Cryst. 1967,22 593. 194 R. M Williams and S C. Wallwork Acta Cryst. 1967,22 899. Part (u) X-Ray Crystallography spacing in the complex is 3.43A.195 The structure of the 1:1 complex NNN’N’-tetramethyl-p-phenylenediamine and 1,2,4,5tetracyanobenzenedoes not seem to show the usual n-m interaction between the aromatic rings but indicates n+n interactions localised between the nitrogen atoms of the donor and the cyano-groups of the acceptor.The magnitude of the charge-transfer from the donor to the acceptor was estimated to be 0.24in electron units.‘96 The complex between 7,7,8,8-tetracyanoquinodimethane and bis-(8-hydroxyquinolato)-copper-(11) crystallises with the component molecules plane-to-plane so that the double bond adjacent to one dicyanoethylene group of the tetracyanoquino- dimethane molecules lies over the 5:8 positions of one donor molecule whilst the other double bond is similarly oriented with respect to the benzenoid ring of the centrosymmetrically related donor molecule.The perpendicular separa- tion of the molecules in the region of overlap is about 3.2A.l” The structure of the complex formed from dimeric 8-hydroxyquinoline and chloranil molecules differs from that predicted by the overlap and orientation principle and is remarkably similar to the bis-8-hydroxyquinolinatopalladium(11)-chloranil molecular complex. The 8-hydroxyquinoline dimer is held together by a bifurcated hydrogen-bond system and the occurence of bifurcated hydrogen bonds is di~cussed.’~~ Natural Product Structures.-That there is no change in configuration at C( 11) during the formation of desmotroposantonin from santonin has been established by an X-ray investigation of 2-bromo-( -)-P-desmotroposantonin (75).lg9.Ring B approximates to an envelope rather than the expected half- chair conformation. The five-membered lactone ring is in the usual envelope conformation with C(7)displaced by 0.52 A from the plane through the other atoms of the lactone ring. The fbepoxide configuration in palmarin derivative (76) has been established and the results of the analysis reveal a degree of con- formational distortion that might appear unreasonable from an examination of Dreiding models. The distortion in ring A is best characterised by comparing the C(14)*-*C(2) distance 2-61A).Interaction distance 3.57 with the C(9).*-C(l) between the C(14)methyl and epoxide oxygen is responsible for this and th f separation of these two centres is 2.92A.2ooThe structure and relative stereo- chemistry of bromomexicanin-E are shown in (77).The two five-membered rings are non-planar and are cis-fused to the seven-membered ring which is in a boat conformation.All bond distances and angles have values close to those expected.”l The constitution and absolute stereochemistry of E-caesalpin (78)have been determined. Rings A,B and c are fused trans-anti-trans with A and B in chair and c in half-chair conformations. The C(5)axial hydroxy- 19’ S. Kumakura F. Iwasaki Y. Saito Bull. Chem. SOC.Japan 1967,40 1826. Y. Ohashi H. Iwasaki and Y. Saito Bull. Chem. SOC.Japan 1967,40 1789. 19’ R. M. Williams and S. C. Wallwork Acta Cryst. 1967,23,448. 19’ C.K. Prout and A. G. Wheeler J. Chem. SOC.,(A) 1967,469. 199 A. T. McPhail B. Rimmer J. M. Robertson and G. A. Sim J. Chem. SOC.(B),1967 101. K. M. S. Islam G. Ferguson K. H. Overton and D. W. Melville Chem. Comm. 1967 167. 201 Mazhar-UI-Haque and C. N. Caughlan J. Chem. SOC.(B) 1967,355. George Ferguson group is involved in an intramolecular hydrogen bond (2.65 A) with the C(l) axial hydroxy-group.f02 Structure determinations of testosterone and 8P-methyltestosterone deriva- tives have shown that the conformation of the latter is clearly bent in contrast Me R (75) (76) (79) '02 A. Balmain K. Bjamer J. D. Connolly and G. Ferguson Tetrahedron Letters 1967 5027. Part (u) X-Ray CrystuZZography 95 with the ‘planar’ overall shape of the ring system of testosterone.Ring D of testosterone has the P-envelope conformation while in the P-methyl derivative a half-chair conformation is adopted.203 17~-Bromoacetoxy-9~-lOa-androst-4-en-3-one has contrary to normal 9a- 10P-steroids a zig-zag skeleton [see (79) which also shows the absolute configuration]. Ring A is a distorted half- chair; rings B and c are chair forms and ring D is in the envelope conforma- ti~n.~’~ The stereochemical course of an unusually facile D-homo-rearrange- ment of 20P-p-bromo benzensulphonyloxy- 19-nor-9p 1Oa-pregn-4-ene-3-one to 17a~-p-bromobenzenesulphonyloxy-17a-methyl- 19-nor-9P l0a-D-homo- androst-4-en-3-one (80) has been established by an X-ray The product molecule is free from major conformational distortion; ring A is in the half-chair form and rings B,C and D exhibit chair forms with average valency angles slightly greater than tetrahedral.The structure of diosgenin has been confirmed206 with the stereochemistry implied by (81). The double bond at C(5)C(6) confers a degree of planarity on the A-B ring system. The triterpene 2a-bromoarborinone has structure (82) with a 13P 14~-trans configuration of the methyl groups at the C-D ring junction. Four of the five carbon atoms of ring E are coplanar to within 005 A and C(17) is displaced 0.74 A from this plane.207 In a molecule containing fused five- and six- membered rings four atoms of the 5-membered ring are frequently coplanar while one of the two atoms common to the two rings is out of plane by approxi- mately 0.7 A.The geometry of the perhydrophenanthrene skeleton in eight steroids has been discussed in considerable detail in terms of valency and torsional angles in a paper by Geise et aL208 The geometrical details of the molecules from X-ray structure determinations are compared with those obtained from theoretical considerations on appropriately substituted cyclo- hexane and cyclohexene rings. It is shown that the use of such building material leads to a qualitative agreement. A number of interactions present in a steroid but not in a cyclohexane unit prevents a quantitative agreement. The steroid skeleton (all-trans) has a somewhat bent overall shape. The occurrence of conformational transmission effects is discussed and a number of rules con- cerning the torsional angles around junctions are given.’03 H. Koyama M. Shiro R. Sato Y. Tsukuda,H. Itazaki and W. Nagata Chem. Comm. 1967,812. 204 W. E. Oberhansli and J. M. Robertson,Heh. Chim. Acta 1967,50 53. ’05 R. T. Puckett G. Sim A. D. Cross and J. B. Siddall J. Chem. SOC.(B) 1967 783. *06 E. A. O’Donnell and M. F. C. Ladd Acta Cryst. 1967,23,460. 207 0.Kennard and L. Riva Di Sanseverino Tetrahedron 1967,23 131. 208 H. J. Geise C. Altona and C. Romers Tetrahedron 1967,23,439. George Ferguson Me (86) (87) All six-membered rings in neothiobinupharidine dibromide tetrahydrate (83) are in chair conformations and all substituents are equatorial. There is some molecular disorder in the crystals associated with an approximate non- crystallographic two-fold axis.209 The hydroxy-group in the perchlorate salt of hydroxy-P-isosparteine (84) is attached to C(7)and not the bridged atom C(8)as previously thought.The molecule is the trans-trans-isomer and all four six-membered rings are in the chair conformation. Neglecting the hydroxy- group the molecule has a two-fold axis of molecular symmetry through the bridge carbon atom.210 The stereochemical features of the steroid alkaloid tomatidine (85) have been determined from an X-ray analysis of a hydro- bromide derivative.21 ' The same stereochemistry has also been reported from an X-ray analysis of the isomorphous hydroiodide derivative.2 l2 The cyclo- hexane rings A,B and c are in the chair conformation with average angles close to 110".Four atoms of ring D are coplanar while the fifth C(14) is 0.7 A from this plane. Four atoms of ring E are also coplanar while the spiro-carbon atom C(22)is out of plane by 0.5 8 resulting in the D-E ring system being symmetrically distorted about the line of fusion. The steroidal framework is bent slightly so that bonds C(lOkC(9) and C(13bC(18) are not parallel but inclined.2 The structure of the Crotalaria alkaloid retusamine has been determined as its monohydrated bromocamphorsulphonate salt. The absolute configuration of cation and anions are shown in (86) and (87).The water '09 G. I. Birnbaum Acta Cryst. 1967,23,526. 210 J. M. H. Pinkerton and L. K. Steinrauf J. Org. Chern. 1967,32 1828. 211 0.Kennard L. Riva Di Sanseverino and J.S. Rollett J. Chem. SOC.,(C),1967,956 212 E. Hohne H. Ripperger and K. Schreiber Tetrahedron 1967,23,3705. Part (u) X-Ray CrystulEography 97 molecule forms a hydrogen-bond network linking the OH group of the retusamine cation to two of the sulphonate oxygen atoms. Of particular interest is the very long ring fusion C-N bond (1-64A) since it is readily cleaved on regeneration of the free base to give an eight-membered cyclic amino-ketone exhibiting transannular interaction between the trigonal nitrogen atom and the carbonyl group. Also of general stereochemical interest are the observations that in the anion the bromine atom is in the endo-position and the sulphonate group is attached to C(9) trans-n to the keto-gro~p.'~~ Me -0 (90) CH,OH 0 The structure and absolute stereochemistry of the hydrobromide salt of a Diels-Alder adduct of thebaine has been determined as (88).The molecular shape which is partly defined by a complex cage structure is severely distorted when compared with an idealised Dreiding model. The hydroxy-group forms an intramolecular hydrogen bond with the adjacent methoxyl oxygen. The positian of the hydroxy-hydrogen atom implies that the methoxy-oxygen is sp2 hybridised. The C-O-CH3 valency angle is 120".The other methoxy- group angle is 117".214The structure and absolute stereochemistry of a p-bromobenzoate of a dihydroanhydroacetonide derivative of taxadiene tetraol have been determined as (89).*' The cyclohexane ring trans-fused with the cycloheptane ring is in a slightly distorted chair conformation.The cyclopentene ring is envelope-shaped and cis-fused with the cycloheptane ring which adopts a twisted boat conformation. A study of the quinidine salt of ( -)-1,1'-dimethylferrocene-3-carboxylic acid has defined the absolute *13 J. A. Wunderlich Acta Cryst. 1967 23 846. 214 J. H. van den Hende and N. R. Nelson J. Amer. Chem. SOC.,1967,89,2901. 215 W. R. Chan T. G. Halsall G. M. Hornby A. W. Oxford W. Sabel K. Bjamer G. Ferguson and J. M. Robertson Chem. Comm. 1966,923. George Ferguson stereochemistry of the metalocene and confirmed the accepted absolute stereochemistry of the alkaloid quinidine.2 l6 The absolute stereochemistry and precise details of molecular geometry of gliotoxin (90) have been re-ported.217 The fused rings form a rigid molecular framework with the piper- azinedione constrained to a boat conformation by the disulphide bridge.The chirality of the CSSC group is left handed and the contribution of the di- sulphide-bridged piperazinedione system can be associated with a negative peak (at 233 mk) in the circular dichroism curves of gliotoxin sporidesmin and their derivatives containing the same system. It is therefore concluded that all these compounds have the same absolute stereochemistry. The cyclo- hexadiene ring has a skew chair conformation. An N-brosyl derivative of mitomycin A an anticancer antibiotic has been shown to have structure and absolute stereochemistry (91).218The bond lengths in the aziridine ring are normal (C-N 1.48 and 1.49 A).The aziridine nitrogen is non-basic atypical for an aziridine nitrogen atom even when part of a [3,l,O]bicyclic system and is attributable to steric interference of its fourth valence pair with the free p orbital of N(2). N(1) may thus be considered as an optically active site pro- duced by steric interference. *16 0.L. Carter A. T. McPhail and G. A. Sim J. Chem. SOC.(A),1967 365. *I7 J. Fredrichsons and A. McL. Mathieson Acta Cryst. 1967,23,439. 218 A. Tulinsky and J. H. van den Hende J. Amer. Chem. SOC.,1967,89 2905.