160 FORSTER AND MICKLETHWAIT : STUDIES IN THE XV1.-Studies an the Camphane Series. Part VI. Stereoisomeric Halogen- Derivatives of a-Benxoyl- camphor. By MARTIN ONSLOW FORSTER and F~ANCES M. CC. MICKLETHWAIT. IN accordance with its unsaturated character, 1-hydroxy-2-benzoyl- camphene, the enolic form of a-benzoylcamphor, immediately decolorises a solution of bromine in an indifferent solvent. At the same time hydrogen bromide is eliminated, and if om molecular proportion of the halogen is employed, the crystalline residue obtained on evaporating the liquid has the empirical formula of benaoylbromocamphor. There is no difficulty, however, in resolving this product into two distinct substances which, although isomeric and nearly alike in chemical behaviour, are widely different in physical properties.The more soluble constituent of the mixture crgstallises from alcohol in six- sided prisms, melts at 114O, has [a],, -lO*Oo in benzene, and [a],CAMPHANE SERIES. PART VI. 161 + 10.3' in chloroform; the isomeride is deposited from alcohol in rectangular plates, melts at 214', has [.ID - 53.2' in benzene, and [aID - 19*3O in chloroform. The method of preparation, the fact that neither substance dissolves in alkalis, and the transformation of both isomerides into l-hydroxy-2- benzoylcamphene by the action of alcoholic potash, are circumstances which point to the conclusion that the compounds in questionare a-bromo- derivatives of a-benzoylcamphor, and that their physical differences are the result of a difference in configuration.Theoretical considerations, moreover, led us to expect the formation of two derivatives displaying isomerism of the cistrans-type, as indicated by the following formulae (compare Lowry, Trans., 1898, 73, 572) : 3 + It is evident that a similar explanation would account also for the production of two isomerides from enolic benxoylcamphor in the event of that substance being shown to have the alternative formula, namely, C:C(OH)*C6~, that of phenylhydroxymethylenecamphor, C,HI4<bo 7 8 possibility which is not yet excluded. Several instances of this form of isomerism in the camphor series have now been established. Leaving aside the somewhat uncertain cases of the monohalogen derivatives of camphor, there remain the isomeric chlorobromocamphors, chloronitrocamphors, and bromonitro- camphors investigated by Lowry (Trans., 1898, 73, 569 and 986), and the benzylbromocamphors described by Haller and Minguin (Compb.rend., 1901, 133, 79). Up to a certain point, the case of the benzoylbromocamphors resembles those of the four derivatives men- tioned, the difference between the two forms being, however, greater than has been observed hitherto ; but an important feature distinguishes it from those already described. In dealing with the isomeric chlorobromocamphors, Lowry records unsuccessful attempts to convert a'-chloro-a-bromocamphor into a-chloro-a'-bromocamphor by the action of heat and of acids (Trans., 1898, 73, 581). Neither in his subsequent communication nor in the paper of Haller and Minguin (Zoc. cit.) is it stated that the chloronitro- camphor, bromonitrocamphor, or benzylbromocamphor of lower melting point can be transformed into the corresponding isomeride, and it is VOL.LXXXI. &I162 FORSTER AND MICKLETEWAIT: STUDIES IN THE probable therefore that the change cannot be effected or it would have been observed. It is in this respect that the benzoylbromo- camphors differ from the foregoing disubstituted a-derivatives, for the compound having the lower melting point is readily converted into the isomeride by the action of hydrogen bromide. This transformation is the fmst recorded instance of stereoisomeric change on the part of a disubstituted derivative of camphor in which both substituents occupy the a-position. It has therefore a direct bearing on the explanation given by Marsh in accounting for the unstable character of a specimen of bromqcamphor which is described as melting at 61' (Trans., 1890, 57, 832 ; compare also Lowry, Trans., 1898, '73, 572).The validity of the explanation in question depends on the formation of an intermediate isomeride, which represents the enoIic modification of the material transformed, and, in the case discussed by Marsh, would have the formula The experiments described in this paper have led us to consider this explanation improbable. I n the fist place, it cannot be applied to derivatives of camphor of the class to which benzoylbromocamphor belongs, and secondly, there seems to be no need for my explanation so complex, several cases of stereochemical transformation being known in which there is no room for any structural change to occur.In general features, the benzoylchlorocamphors resemble the corresponding bromoderivatives very closely, the two modifications which melt at 88' and 2 1 9 O displaying similarity as regards solubility and crystalline form when compared respectively with the bromo- derivatives melting at 114O and 2149 It is noteworthy, however, that we have been hitherto unable to convert one isomeride into the other. Moreover, the action of sodium hypochlorite on enolic benzpylcamphor gives rise to a preponderating quantity of the benzoglchlorocamphor of the lower melting point, whilst the benzoylbromocamphor of the higher melting point is the almost exclusive product when potassium hypobromite is employed ; bromine dissolved in chloroform yields a mixture of the isomerides in nearly equal parts, whilst bromine and glacial acetic acid containing sodium acetate afford chiefly the benzoyl- bromocamphor of lower melting point. In describing the stereoisomeric halogen derivatives of a-benzoyl- camphor, we have adopted the convention suggested by Lowry (Zoc.cit.), so that the nomenclature of the new derivatives may be uniform with that of the unsymmetrical di-derivatives already prepared. Assuming that benzoylcamphor, with [ aID + 13'7-5O in alcohol, is an a-derivative, it will be noticed that the optical influence of the benzoyl radicle exceeds that of the chlorine atom, since a-chlorocamphor has [aID + 9 6 O in the same solvent; it may be concluded therefore that tho di-CAMPHANE SERIES.PART VI. 163 derivative which has a specific rotatory power least removed from that of camphor itself, is that which contains the benzoyl radicle in the a-position. This modification is the one which melts at 219’, having [a], + 26*2O in chloroform, and is accordingly termed a-benzoyl-a’- chlorocamphor. In the case of the bromo-derivatives, it is not so easy to decide which isomeride contains the benzoyl radicle in the a-position, because the recorded values for the specific rotatory power of benzoyl- camphor and of bromocamphor in alcohol are practically identical. There is reason to believe, however, that the specific rotatory power of benzoylcamphor a t the moment of dissolution in alcohol is lower than ID + 1 3 7 5 O , because the substance, dissolving somewhat slowly in the cold solvent, suffers partial conversion into the enolic modifi- cation, with [a], + 2 6 2 O , before it can be examined in the polarimeter ; chloroform, however, which dissolves the substance very readily, yields a solution having [ + 125O, and it is therefore probable that the optical influence of the benzoyl radicle is less powerful than that of the bromine atom, because a-bromocamphor has [.ID + 135O.I f this is the case, the modification which melts at 114O and has [aID + 1 0 ~ 3 ~ in chloroform must be called a’-benzoyl-a-bromocamphor, whilst the isomeride melting at 214O, having a specific rotatory power more remote from that OF camphor, must be regarded as having the bromine atom in the a’-position ; m. p. Camphor ........................+ 42’ ( alcohol ) - a-Chlorocampbor ............... + 96 ( ,, ) - a-Bromocamphor ............... + 135 ( ,, ) - a-Benzoylcamphor ............ + 125 (chloroform) - a‘-Benzoyl-a-chlorocamphor ... - 28 ( ,, ) 8 8 O a-Benzoyl-a’-chlorocamphor ... + 26 ( ,, ) 219 a‘-Benzoyl-a-bromocamphor ... + 10 ( ,, ) 114 a-Benzoyl-a’-bromocamphor ... - 19 ( ,, ) 214 From this table, it will be noticed that the benzoylchlorocamphor and benzoylbromocamphor supposed to contain the halogen in the a-position both melt a t the lower temperature, whilst the less readily fusible modifications are assumed to have the halogen substituted in the a’-position. EXPERIXENTAL. aa-BenxoyZ6romocamphors, C,H,,< CBr*CO*C6H, I co Twenty grams of 1 -hydroxy-2-benzoylcamphene were dissolved in chloroform and cooled in melting ice.A cold solution of 12.4 grams of bromine in chloroform was then added in small quantities at a time, &r 2164 FORSTER AND MICKLETHWAIT: STUDIES IN THE and the pale red liquid, from which hydrogen bromide was being evolved, transferred to basin and allowed to evaporate spontaneously. A previous experiment having shown that two compounds are produced by this means, the crystalline residue was divided into four fractions by extracting it successively with quantities of hot alcohol amounting to 100 C.C. (twice), 200 c.c., and 300 c.c., and allowing the solutions to cool. Fraction I, weighing 6 grams, consisted of thin, transparent needles melting somewhat indefinitely at 108-110'; a 2 per cent. solution in benzene gave [ a ] D - 15.4'; and in chloroform [aID 3.7.9'.A large proportion being readily soluble in warm, light petroleum (b. p. 50-goo), the whole fraction was extracted with this solvent ; the solution de- posited large, thin, six-sided prisms melting at 114' and giving [ a ] D - 10*Oo in benzene and [ a ] , + 10.3' in chloroform. Recrystallisation from light petroleum did not change the specific rotatory power. Fraction 11, weighing 8 grams, consisted chiefly of needles, and melted somewhat indefinitely at 109-11 1' ; a 2 per cent. solution in benzene gave [ a ] , - 19-5', and in chloroform [ a]D + 2.9'. Fraction 111, weighing 2 grams, consisted of thin, rectangular plates, beginning to shrink and to change colour a t about 185' and melting a t 210' ; a 2 per cent. solution in benzene gave [ a ] D - 52*3', and in chloroform [ aID - 18.5'.Fraction IV, weighing 3 grams, consisted of thin, rectangular plates, beginning to shrink and to change colour a t about 190°, and melting a t 214' ; a 2 per cent. solution in benzene gave [a], - 53-52', and in chloro- form [ a ] D - 19.3'. The properties of this fraction were not altered by recrystallisation from boiling alcohol. d-Benxoyl-a-brmocamphor is most conveniently prepared by dissolving 1 -hydroxy-2-benzoylcamphene in glacial acetic acid containing 14 mols. of sodium acetate and adding 1 mol. of bromine dissolved in glacial acetic acid; the white precipitate obtained on pouring this liquid into water is then collected, washed, dried, and crystallised from light petroleum. It is readily soluble in chloroform, benzene, alcohol, or light petroleum, crystallising from the last-named in large, transparent, six-sided prisms, and from alcohol in slender needles having the same crystalline form : 0.1992 gave 0*1101 AgBr.Br= 23.52. C17Hl,0,Br requires Br = 23.88 per cent. The subdjance melts at 114', but fusion is not complete until the temperature is raised to about 180'. A solution containing 0.5 gram in 25 C.C. of benzene a t 21' gave aD -24' in a 2 dcm. tube, whence the specific rotatory power [ a ] D - 10.0' ; 0.5029 dissolved in 25 C.C. of chloroform at 21' gave aD + 25', corresponding to [aID + 10.3'.CAMPHANE SERIES. PART VI. 165 a-Benzoyl-a'-bromocamphor was obtained in the following manner. One hundred grams of bromine were dissolved in an ice-cold aqueous solution containing 150 grams of potassium hydroxide, and slowly added to 20 grams of 1-hydroxy-2-benzoylcamphene dissolved in dilute potash.The sticky solid which immediately separated soon hardened, and after an interval of 12 hours was collected, washed, and recrystal- lised from boiling alcohol. The yield of aa-benzoyl bromocamphor obtained by this method is quantitative, and the product consists chiefly of the variety of high melting point. It dissolves very readily in chloroform, but only sparingly in cold alcohol or benzene, and is almost insoluble in boiling light petroleum ; it crystallises from hot alcohol in transparent, rectangular plates, begins to shrink and to change colour at about 190°, and melts at 214' t o a pale brown liquid which evolves gas. The substance may be crystallised from concentrated nitric acid without undergoing change : 0.1353 gave 0.0756 AgBr.Br = 23.77. CI7H,,O,Br requires Br = 23.88 per cent. A solution containing 0-5015 gram in 25 C.C. of benzene at 21° gave aD - 2'8' in a 2 dcm. tube, whence the specific rotatory power [a], - 53.2'; 0.6451 gram dissolved in 25 C.C. of chloroform at 21' gave U, - loo', corresponding to [a],, - 19-3O. Action of Alcoholic Potassium H3droxide on aa-Bennoyl6romocamp~~. -A specimen of a-benzoyl-a'-bromocamphor which melted at 210' and gave [.ID - 18.5' in chloroform, was heated during 4 hours in a reflux apparatus with potassium hydroxide (2 mols.) dissolved in alcohol. The liquid soon became dark brown, and on evaporation yielded a residue which dissolved completely in water.A current of well washed carbon dioxide was then passed into the aqueous solution until no further pre- cipitation occurred, and the product, after crystallisation from alcohol, was obtained in the pink octahedra characteristic of 1-hydroxy-Bbenzoyl- camphene. The same compound was obtained by reducing a'-benzoyl-a-bromo- camphor with alcoholic potassium hydroxide. Action of Bromine on 1-Benxox~-2-benxoylcampherne.-When a solution of 1-benzoxy-2-benzoylcamphene in chloroform is treated with bromine, the colour of the halogen is not immediately destroyed, but after an interval, action is found to have taken place. Ten grams of the dibenzoyl derivative were dissolved in 100 C.C. of chloroform and enclosed in a stoppered bottle with 4.4 grams (1 mol.) of bromine, After 24 hours, the colour of the halogen had almost disappeared.On allowing the liquid to evaporate, a considerable quantity of hydrogen bromide was liberated, and a crystalline residue was obtained having the odour of ethyl benzoate. The solid product,166 FORSTER AND MICKLETHWAIT: STUDIES IN THE weighing 9 grams, was exhausted with 50 C.C. of hot alcohol, and the solution deposited crystals melting a t 110-1 15' and giving [a], + 4.1' in chloroform : 0.1598 gave 0.0884 AgBr. Br = 23.54. C17Hl,0,Br requires Br = 23-88 per cent, The substance was evidently a mixture of the two aa-benzoylbromo- camphors, and by repeated cry stallisation from alcohol, a specimen of the modification of higher melting point was obtained, giving [a]D - 19.5O in chloroform. Conversion of One Isomride into the .Other.-During the first at- tempt to separate the isomerides from one another by fractional crys- tallisation, a most unexpected change of the variety of lower melting point took place. A specimen of t h a t substance, which had been re- crystallised twice from alcohol without altering the melting point, melted a t 11 1-1 12' and gave [ a ] D - 29.0' in benzene ; it was dissolved in hot alcohol, which on cooling deposited lustrous plates melting at 201--204O, and giving [a], - 51.0' in benzene. Although we have not succeeded in reproducing the conditions of this experiment, the transformation of one modification into the other can be effected by the agency of hydrogen bromide. A specimen of a'-benzoyl-a-bromocam- phor melting at 109-1 1 1' and giving [ aID + 2.9" in chloroform was finely powdered and placed in a stoppered bottle with sufficient fuming hydrobromic acid to convert it into a thin paste.The following morn- ing, water was added and the solid product filtered and washed. The substance, when dried in the desiccator, melted at about 200' and gave [.ID - 18.6' in chloroform, and when recrystallised from alcohol yielded the lustrous plates characteristic of a-benzoyl-a'-bromocamphor. Action of Bromine on a-Benxoylcamphw. In describing the a-substituted halogen di-derivatives of camphor, Lowry (Trans., 1898,73, 572) suggests that '' the production of stereo- isomeric di-derivatives is most readily explained by supposing that the action of the halogen involves addition to the enolic form of the mono- derivative." This explanation is a very probable one, and the follow- ing experiment appears to give it direct support, A specimen of ketonic a-benzoylcamphor, giving only a faint colora- tion with ferric chloride, was dissolved in cold glacial acetic acid containing sodium acetate (18 mols.) ; to this liquid, a solution of bromine (1 mol.) in glacial acetic acid was added, when it was observed that the colour of the halogen was immediately destroyed.Although it must be remembered that a small proportion of the benzoylcamphor is enolised by the solvent, it is still fair to say that the behaviour of a-benzoylcamphor towards bromine exactly resembles that of theCAMPHANE SERIES. PART VI. 167 unsaturated enolic isomeride, and it is noteworthy that the identity extends to the product of the change, which gives rise to $-benzoyl-a- bromocamphor in both cases.Having found that potassium hypobromite converts 1-hy droxy-2- benzoylcamphene into a mixture of the benzoylbromocamphors, we employed the corresponding method in preparing the benzoylchloro- camphors in preference to treating the hydroxy-compound with the free halogen. Ten grams were dissolved in dilute aqueous potassium hydroxide, cooled with fragments of ice, and treated with 200 C.C. of a solution of sodium hypochlorite containing 30 grams of available chlorine per litre. A pink, dough-like solid separated and rapidly became hard. After an interval of several hours, the product was collected, washed, and extracted with 100 C.C. of boiling alcohol, thus dividing the substance into two portions, of which the more readily soluble melted somewhat indefinitely a t 85-87' and gave [.ID - 20.6' in chloroform, whilst the residual fraction melted at 219' and gave at-BenxoyEa-chZorocamphw, obtained by recrystallising the more soluble fraction from alcohol and then from light petroleum, crystal- lises from each solvent in prisms and melts at 88': C1= 12-08, C17H,,0,Cl requires C1= 12.22 per cent. [ u ] D + 26.0'. 0-1284 gave 0.062'7 AgC1. It is readily soluble in alcohol and very freely so in chloroform, but dissolves only sparingly in light petroleum. A solution containing 0.4185 gram in 25 C.C. of chloroform a t 21' gave uD -56' in a 2 dcm. tube, whence the specific rotatory power [ aID - 27.9'. a-Bsn~yGa'-chlorochor remains after the mixture of the two isomerides has been exhausted with a small quantity of hot alcohol ; it crystallises from that solvent in plates resembling the corresponding bromo-derivative and melts at 219': C1= 12.29. C17H1902Cl requires Cl = 12.22 per cent. 0*1324 gave 0.0658 AgC1. It is freely soluble in chloroform, but dissolves only sparingly in alcohol and is insoluble in light petroleum, A solution containing 0.3973 gram in 25 C.C. of chloroform at 21' gave aD +50' in a 2 dcm. tube, whence the specific rotatory power [aID + 26.29 ROYAL COLLEGE OF SCIENCE, LONDON. SOUTH KENSINGTON, S. W.