224 MILLER AND SMILES : XXXVII1.-The Constitution of Disulphoxides. Part I I . By CECIL JAMES MILLER and SAMUEL SMILES. ALTHOUGH the symmetrical structure of the disulphoxides (I) has been generally accepted the evidence adduced in its favour has been incorrectly interpreted and is in fact ambiguous and many simple reactions of these substances appear clearly to indicate the thio-sulphonate structure (IT) such for example as their behaviour with zinc dust potassium sulphide sodium arsenite and mer-captans (J. 1924 125 176). Moreover this view is supported by syntheses of disulphoxides from sulphenic halides and silver sul-phinates (Zincke Annalen 1912 391 67). Since all the definite evidence thus favoured the thiosulphonate structure it was concluded that this should not be abandoned but instead should be regarded as the most probable until final proof of one structure or another was forthcoming.(I.) Ar*E*fi*Ar Ar*so,*s*Ar ArR*O*SAr (111. ) 00 (11.1 0 The experiments now described provide the necessary further proof of this unsymmetrical constitution in that (1) the action of Grignard’s reagent with the disulphoxides must be added to those already quoted as indicating this structure (2) the reduction of disulphoxides with hydrogen iodide which hitherto has been quoted as the main evidence for the symmetrical formula is not only use-less as such but accords with the thiosulphonate structure (3) the unsymmetrical character of the disulphoxides is demonstrated by synthesis. slthough the behaviour of numerous carbon compounds of sulphur with alkyl and aryl magnesium halides has been investi-gated (Lapworth J .1912 101 297; Hepworth and Clapham J., 1921 119 IlSS; Wedekind Rer. 1921 54 16041 that of the disulphoxides appears to have escaped attention. These substances are very readily attacked by the magiiesinm compounds and usin THE CONSTITUTION OF DISULPHOXIDES. PART 11. 225 equimolecular proportions of the reactants high yields of the sulphinic acids have been isolated fission of the disulphoxide taking place apparently according to the following scheme (IV) : The sulphinic acid was obtained in every case examined but the fate of the thioaryl group of the disulphoxide varied according to the nature of the magnesium compound used with magnesium inethyl iodide the methyl aryl sulphide Ar*S*CH, was commonly obtained whilst with magnesium phenyl iodide the disulphide and diphenyl also were observed.The chief point at issue is tlhe situation of the oxygen in the molecule of the disulphoxide and with the symmetrical formula it is very difficult to explain the isolation of SOY; of this as sulphinic acid. According to this symmetrical constitution it would be expected that fission of the molecule would result in the formation of a sulphoxide (V) but this has not been observed. The results of this decomposition evidently support the unsymmetrical structure and appear inexplicable on the basis of the symmetrical formula. The mode of experiment is described and the results are tabulated on subsequent pages. The fact that the disulphoxides are easily reduced by hydrogen iodide t o disulphides has been regarded as sufficient reason for rejecting the thiosulphonate structure and as proving the presence of the true disulphoxide arrangement (Hinsberg Ber.1908 41, 2836 4294; 1909 42 1278; Fries Ber. 1914 47 1195). It is assumed that the reduction is direct and since the sulphonyl group usually is not easily reduced by this reagent whilst the thionyl group is readily attacked the conclusion would seem justified but it has been pointed out (Smiles and Gibson J. 1924 125 176) that the force of the argument depends on the assumption that reduction is direct and is not accompanied by fission of the mole-cule. If however rupture does take place the reduction cannot be used as an argument against the unsymmetrical or in favour of the symmetrical formula for in either case a disulphide must be expected as the final product of the reaction.The question whether this fission takes place or not may be answered by the reduction of a disulphoxide containing different aromatic groups. I n such a case, if the unsymmetrical disulphide is the sole product (VI) the assumption of direct reduction is clearly justified but if both or even one of the symmetrical disulphides are obtlained (VII) it must be admitted that rupture of the dithio-system has taken place. (vI.) R1*SO*SO*R2 -+ R1*S-S*R2 (VII.) 2R1*S,0,*R2 -+ (RlS) + (R2S), VOL. CXXVII. 226 MILLER AND SMILES: Disulphoxides containing merent aromatic nuclei are readily prepared by an extension of the synthetical method of Zincke (Zoc.cit.) For example when the silver salts of p-toluenesulphinic acid and p-chlorobenzenesulphinic acid are treated with o-nitrophenyl-sulphur chloride the o-nitrophenyl p-toluenethiolsulp~ona.te (VIII) and o-nitrophenyl p-chlorobenzenethiolsulphonate (IX) are respec-tively 'formed. N02*C,H4*SCI + AgSO,*C,H = AgCl + NO,*C,H,*S*SO,*C,H,. (VIII.) Reduction of these substances with hydrogen iodide in the usual manner (compare Hinsberg Zoc. cit.) furnished o-nitrophenyl di-sulphide (N02*C,H,*S*), in good yield and in the case of the chloro-derivative p-chlorophenyl disulphide also was isolated. Hence it is clear that the disulphoxide system has. been broken during the process of reduction and the reaction cannot be admitted as evidence favouring the symmetrical and excluding the unsymmetrical structure.Nothing else can be deduced from this result but further information has been gained by adopting milder conditions of reduction with hydrogen iodide or by the use of dilute hyposulphite (in presence of sodium carbonate). The derivative (VIII) then yielded p-toluenesulphinic acid and the nitrophenyl disulphide the chloronitro-derivative (IX) gave p-chlorobenzenesulphinic acid and the nitro-disulphide whilst p-tolyl disulphoxide and p-chlorophenyl disulphoxide gave the corresponding sulphinic acids and disulphides or mercaptans. These results are similar to those obtained by Gutmann (Ber. 1914 47 635) on reduction with sodium arsenite. In the cases of disulphoxides containing similar aromatic groups this result does not enable a decision to be made in favour of either structure for both permit the formation of a sulphinic acid by hydrolytic fission, R*SO-SO*R + R*SO,H + R*S*OH + R*SO,*SR whilst the disulphide would result from reduction of the less stable sulphenic acids.Turning to the case of a disulphoxide containing different aromatic groups a distinction between the requirements of either structure is found. On the basis of the symmetrical arrangement the formation of two sulphinic acids may be expected, for there is no reason to suppose that hydrolysis will take place in one direction only when R1 and R2 are of similar character : RlS*OH + R2*S0,H +- R1*SO*SO*R2 -+ R1*S02H + R2*S*OH THE CONSTITUTION OF DISULPHOXIDES. PART 11. 227 On the other hand according to the thiosulphonate structure only one sulphinic acid should be formed : 2R1*S0,*S=R2 + 2R1*S0,H + (SR2)2 The latter condition accords with the result obtained by the reduction of these substances only one sulphinic acid has been isolated in each case examined.Moreover it is significant that the sulphinic acid-isolated was always the one which had been used as ifi component in the synthesis thus nitrophenyl t'oluenethiolsulphonate (VIII) yielded p-toluenesulphinic acid and o-nitrophenyl disulphide, and nitrophenyl chlorobenzenethiolsulphonate (IX) gave p-chloro-benzenesulphinic acid and the nitro-disulphide : 2CsH,C1*S0,*S*C6H4*No2 + 2C6H,C1*S02H + (NO,*C,I~,*S.),. The action of mercaptans with these disulphoxides is closely analo-gous (compare Smiles and Gibson loc.cit.). o-Nitrophenyl mercaptan yields in both cases the o-nitrophenyl disulphide and the corre-ponding sulphinic acid : CiH,.SO,* S-CGH,*NO, H S*CGH,-N02 The information gained from this extended study of the reduction of the disulphoxides therefore clearly favours the unsymmetrical structure. Taking a general review of the characteristic reactions of the disulphoxides it is seen that all yield the sulphinic acid and a product which contains the thioaryl group; these substances result from the action of zinc dust magnesium alkyl halides potass-ium sulphide mercaptans and of arsenite and other mild reducing agents. The majority of these reactions indicate an unsymmetrical structure a few are ambiguous in their import whilst none insists on the syrnmet8rical arrangement.To establish more firmly the unsymmetrical constitution disul-phoxides each containing different aromatic groups have been synthesised by alternate methods. I n the first Eeries of experiments it seemed desirable to arrange that the aromatic groups K1 and R2 should differ by the nature of the substituents present and not merely by the position of these. The more stable of the known aromatic sulphur chlorides contain the nitrc-group and it was obvious to choose one of these as a component of one method of synthesis but the chief dificulty lay in obtaining a sulphur chloride of sufficient stability without this group for use in the alternate I 228 MILLER AND SMILES : process. This was overcome by the preparation of 2 5-dichloro-phenylsulphur chloride (X) from the corresponding disulphide and chlorine.From 2-nitrophenylsulphur chloride and silver 2 5-dichloro-benzenesulphinate a disulphoxide (XIa) of m. p. 142" was obtained, whilst from 2 5-dichlorobenzenesulphur chloride and silver 2-nitro-benzenesulphinate an isomeric compound (XIIa) melting at 129" was prepared. These substances differ not only in their physical properties but also in their chemical behaviour. When 2-nitro-phenyl 2 5-dichlorobenzenethiolsulphonate (XIa) is treated with 2-nitrophenyl mercaptan the dichlorobenzenesulphinic acid and o-nitrophenyl disdphide are formed (XI) whilst the action of the same mercaptan with 2 5-dichlorophenyl 2-nitrobenzenethiol-sulphonate (XIIa) yields o-nitrobenzenesulphinic acid and a mixture of the two possible disulphides (compare Smiles and Gibson Zoc.cit.). c1 \/ (XII.) c1 (X). The action of 2 5-dichlorophenyl mercaptan with the latter di-sulphoxide yielded (XII) the nitrosulphinic acid and the tetra-chloro-disulphide. Mild reduction of these disulphoxides gave analogous results ; these are collected on a subsequent page. A second series of experiments was made with material in which the groups R1 and R2 were closely similar for it was thought that such conditions would be favourable to intramolecular change-if indeed this were possible-of one thiolsulphonate to the other or of the thiolsulphonate to the true a-disulphoxide structure the latter change having been suggested by previous workers (Hinsberg Zoc.cit.; Fries Zoc. cit.). The materials chosen were the 2 5-dichloro-and 2 5-dibromo-phenyl derivatives. Thus 2 5-dichlorophenyl-sulphur chloride (X) and silver 2 5-dibromobenzenesulphinate furnished 2 5-dichlorophenyl 2 5-dibromobenzenethiolsulphonate (XIIIa) which melted at 119" even after being heated to 150" and cooled. On treatment with 2 5-dichlorophenyl mercaptan the tetra-chloro-disulphide and 2 5-dibromobenzenesulphinic acid were formed (XTTI) the structure indicated by synthesis being thus confirmed THE CONSTITUTION OF DISULPHOXIDES. PART 11. 223 The isomeric 2 5-dibromophenyl 2 5-dichlorobenzenethiol-sulphonate (XIVa) prepared in a similar manner melted a t 125" and with 2 5-dibromophenyl mercaptan yielded the tetrabromo-disulphide and 2 5-dichlorobcnzenesulphinic acid (XIV) .I n summarising the results of these experiments it is claimed that an unsymmetrical structure must be assigned to the disulph-oxides. Two unsymmetrical formulz have been advocated the thiolsulphonate and the anhydride arrangements (I1 and 111) ; there is ample reason for discarding the latter (Fries Zoc. cit.) whilst the former accords very closely with all the facts since discovered. E x P E R I M E N T A L. The disulphoxides were obtained those containing similar aromatic groups from the sulphinic acids by the usual method and those with dissimilar groups by the reaction of a silver sulphinate with the requisite sulphur chloride as follows. A solution of the sulphur halide in dry ether was shaken a t the ordinary t'emperature or boiled with an excess of the silver sulphinate until all the sulphur halide had disappeared the t8reatment being adapted to the reac-tivity of the materials and the stability of the sulphur halide used.Generally the disulphoxide separated as the reaction proceeded ; when this was complete the solid material was collected and the ciisulphoxide was extracted from it with a suitable solvent. In the majority of the cases studied the yields were good. The following substances were obtained by this method from the stated com-ponents. 2-Kit rophen yl 4-tolu e?zethiolsulphonate 02N*C6H4*S-S O,*C ,H,Me, from o-nitrophenylsulphur chloride and silver toluenesulphinate, formed colourless prisms m. p. 97" soluble in the usual organic solvents (Found S = 20-7 ; C = 50.0 ; H = 3.7.C1,HIIO,NS, requires S = 20.69; C = 50.55; H = 3.58 yo). The alcoholic solution of this substance became blue on addition of aqueous sodium hydroxide. 2-Xitrophenyl 4-chlorobenzenethioleulp~o?~ate, 0,EC ,H,.S.S Q2*C ,H,CI , from o-nitrophenylsulphur chloride and silver p-chlorobenzene-sulphinate formed colourless prisms m. p. 123" soluble in the usual organic media (Found S = 19-6; C1 = 10-5; C = 43.3; H = 2-66. Cl2H,O,NC1S2 requires S = 19-42 ; C1 = 10.74 ; C = 43.68 ; H = 2-44 Yo). 3-Xitrophenyl 2 5-dichiorobe?azenethiolsulphonate, 0,N*CsH4.S*S0,*CGH,Cl*, from o-nitrophenylsulphur chloride and silver 2 5-dichlorobenzene-sulphinate melted a t 142" and was sparingly soluble in commo 230 MILLER AND SMILES : solvents (Found C1= 19.4; S = 17.6.C1,H,04NCl,S requires C1 = 19.47; S = 17.61 yo). Like the foregoing disulphoxides, this substance gave a deep blue solution with alcoholic sodium hydroxide. 2 5-Dichlorophenyl 2-nitrobenzenethiolsulphonate, from silver o-nitrobenzenesulphinate and 2 5-dichlorophenyl-sulphur chloride separated from hot alcohol in colourless needles, m. p. 129" (Found S = 17:6; C1 = 19.6. C1,H,O,NC1,S requires S = 17.6 ; C1= 19.47 yo) ; aqueous sodium hydroxide added to the alcoholic solution did not give the characteristic blue colour. A mixture of the two isomeric substances in approximately equal amounts melted indefinitely between 88-95'. 2 5-Dichlorophenyl 2 5-dibromobenzenethioIsulphona.te, C6H,C1,*S*S0,*C &€,Br2, from 2 5-dichlorophenylsulphur chloride and silver 2 5-dibromo-benzenesulphinate separated from alcohol in colourless needles, m.p. 119' (Found S = 13.8; C1 + Br = 48.3. Cl,H60,C12Br,S, requires S = 13.45; C1+ Rr = 48-38 %). 2 5-Dibromophenyl 2 5-dichlorobenzenethiolsulphonate, C,H,Br,-S*S0,*C6H,C12, from silver 2 5-dichlorobenzenesulphinate and 2 5-dibromo-phenylsulphur bromide separated from hot alcohol in colourless prisms m. p. 125." This melting point remained unaltered after a sample had been fused and cooled. The substance is less soluble than the isomeric disulphoxide and a mixture of the two in approxi-mately equal amounts melted indefinitely at 110-114" (Found : S = 13.4; C1 + Br = 48.5. C12H60,C1,Br,S requires S = 13.45; C1 + Br = 48.38 yo). 2 5-Dichlorophenylsulphur chloride C6H,Cl,*SC1 was prepared by saturating a concentrated solution of the corresponding disulphide in dry carbon tetrachloride with chlorine.The residue obtained after the solvent had been evaporated was kept under diminished pressure when it solidified. The product was purified by crystallis-ation from ice-cold ether when the substance was obtained in golden-yellow needles m. p. 32-33' (Found C1 = 49.8; S = 14.9. C,H3Cl,S requires C1 = 49.8; S = 15.03 %). The substance was very soluble in organic media; dilute aqueous sodium hydroxide gave the corresponding disulphide and alkali sulphinate. The 2 5-dibromophenylsulphur bromide was obtained by a similar process as a yellow crystalline material but owing to its instability attempts to isolate it in a pure condition for analysis were un-successful some loss of bromine occurring with formation of the disulphide THE CONSTITUTION OF DISULPHOXIDES.PART 11. 231 Behaviour o j Disulphoxides with Methyl and Phenyl Magnesium Halides.-A dilute solution of the disulphoxide in ether was added to a cooled solution of the magnesium compound (1 mol.) in the Magnesium Sulphinic Other substances Disulphoside. derivative. acid yo. observed. D iphenyl Methyl 63 Ph*S.Me ; PhS.SPh. Diphenyl Phenyl 70 Ph*S.Ph ; PhS-SPh ; Ph-Ph. Di-p-tolyl Methyl S7 C,H,.S-Me; C,H,-S.SC,H,. Di-p-t olyl Pheiiyl 76 C,H,-S.Ph; (C,H,.S.)2 and Di-p-chlorophenyl Methyl 90 C6H,C1.S*&~e; (C6H,C1*S*)2. Ph *Ph . 2 5 2' 5'-Tetra- 7 9 82 4 4'-Dimethoxy- 7 7 68 MeO.C,H,.S.Me ; and disul-chlorodiphcnyl tolyl 3 S'-disul- phide.phoxide same solvent. The mixture was kept for 12 hours before treatment with water. Sufficient aqueous sodium thiosulphate was then added to remove free iodine if this were present and finally excess of dilute sulphuric acid. The sulphinic acid was extracted from the ethereal solution with dilute alkali the substances remaining being separately examined. A summary of the results is given in the foregoing table. The third column shows the approximate per-centages in which the sulphinic acids were isolated. These substances were identified by comparison with authentic samples and by con-version to the methylsulphones which were similarly compared. The latter have been previously described in literature with the exception of the following.4-Chlorophenylmethylsulphone C,H,Cl*SO,*CH, colourless needles, m. p. 96" was obtained by oxidation of the sulphide and by methylation of the sodium sulphinate with methyl sulphate (Found C1= 18.7; S = 16.6. C7H70,ClS requires C1= 18-58; 2 5-Dichlorophenylmethylsulp~one C6H,Cl,*S0,*CH, prepared in a similar manner separated from hot water in needles m. p. 88" (Found S = 14-0. C7H,02C12S requires S = 14-24 yo). The identification of sulphides was generally effected by oxidation to the sulphones. Action of Mercaptans with Disu1phoxides.-The method of opera-tion was similar to that already described (Smiles and Gibson Eoc. cit.) molecular proportions of the reagents being taken in alcohol. The results are collected in the following table ; the third and fourth columns respectively show the sulphinic acid and the disulphide which were isolated from the interaction of the stated disulphoxide and mercaptan.The yields of these products were generally of the order of 80% of theory and higher when the sparingly soluble S = 16.8 %) 232 THE CONSTITUTION OF DISULPHOXIDES. 2 2’-dinitrodiphenyl disulphide was dealt with. the h t column refer to numbered formulae. Disulphoxide. Mercaptan. 2-Nitrophenyl 2-Nitrophenyl 4-toluenethiol-sulphonate (VIII) 4-chloro benzene thiolsulphonate 2 5-dichloro-benzenethiol-sulphonate (XI) phenyl 2-nitro- phenyl benzenethiol-sulphonate phenyl 2-nitro-benzenethiol-sulphonate 2 5-Dichloro-phenyl 2 5-di- phenyl bromobenzene-thiolsulphonate phenyl 2 5-di- phenyl chlorobenzene-thiolsulphonate 2-Nitrophenyl 99 (1.W 2 -Nitrophenyl 9 9 2 5-Dichloro- 2 5-Dichloro-(=I) 2 5-Dichloro- 2-Nitrophenyl (=I) 2 5-Dichloro-(XIII) 2 5-Dibromo- 2 5-Dibromo-(XW) Sulphinic acid.4-Toluene-4-Chloro benzene-2 5-Dichloro-benzene-2 -Nitro benzene-2 6-Dibromo-benzene-2 5-Dichloro-benzene-PART 11. The numerals in Disulphide. Di-o-nitrophenyl 2 5 2’ 5’-Tetre-chlorodiphenyl A mixture not separated 2 5 2‘ 5’-Tetra-chlorodiphenyl 2 5 2’ B’-Tetra-bromodiphenyl Reduction of Disulphoxide+.-Three methods of reduction were employed. In casm 1 2 and 3 of the following table where com-plete reduction with hydrogen iodide was required the conditions devised by Hinsberg (Ber.1908 41 4295) were used. Milder reduction with this reagent was effected as follows. About 2 g. of the disulphoxide 3-4 C.C. of glacial acetic acid and a few C.C. of a saturated solution of sodium bisulphite containing ten drops of hydriodic acid (d 1.9) were constantly shaken with 50-100 C.C. of light petroleum for about 4 hour the necessary time varying some-what according to the disulphoxide taken. Then aqueous sodium carbonate was added until the whole was alkaline. I n cases where the dinitro-disulphide was formed this separated almost completely from the liquid. The solid material was collected the petroleum and the aqueous portion of the liquid being separately examined. The sulphinic acid was isolated from the aqueous portion and the di-sulphide from the solid or the petroleum.Examples of this type of reduction are given in Nos. 4,5,6 and 7 of the table. Reduction with sodium hyposulphite was conducted by shaking the disulph RESOLUTION OF CHLOROSULPHOACETIC ACID ETC. 233 oxide with a slight excess of a 3-4 yo aqueous solution of the reagent to which about l/lOth of its volume of alcohol had been added the mixture being kept alkaline by the addition of sodium carbonate. The disulphide formed was isolated by solution in ether whiIst the sulphinic acid and mercaptan were obtained from the aqueous portion. When mercaptan was present this was separated from the sulphinic acid as disulphide by treating the solution of sodium salts with a current of air. Examples are shown in h‘os.8 9 and 10 of the table. 1. 2. 3. 4. 5. 6. r S. 9. 10. Disulphoxide. 2-Nitrophenyl 4-toluencthiol-sulphonate 2-Nitrophenyl 4-chlorobenzene-thiolsulphonate 2 5-Dichlorophenyl 2-nitro-benzenethiolsulphonate 2-Nitrophenyl 4-toluenethiol-sulphonate 2-Nitrophenpl 4-chlorobenzene-tliiolsulphonate Di-4-chloroplienyl disulphoxide 2-Nitrophenyl 2 5-dichloro-bcnzenethiolsulphonate Di-p-tolyl disulphoxide 2 5-Dichlorophenyl 2-nitro-benzenethiolsulphonatc 2-Nitrophenyl 2 5-dichloro-beiizenethiolsulphcnate Products isolated. 2 2’-Dinitrodiphenyl disulphide 2 2’-Dinitrodiphenyl disulphide and 4 4’-dichlorodiphenyl disulphide A mixture of disulphides; not separated 4-Toluenesulphinic acid and 2 2’-di-nitrodiphenyl disulphide 4-Chlorobenzenesulphinic acid and 2 2’-dinitrodiphenyl disulphide 4-Chlorobenzenesulphinic acid and 4 4’-clichlorodiphenyl disulphide 2 5-Dichlorobenzenesulphinic acid and 2 2’-dinitrodip~ienyl clisul-phide p-Toluenesulphinic acid and p-tolyl mercaptan 2-Nitrobenzenesulphinic acid and 2 5 2‘ 5’-tetrachlorodiphenyl cli-sulphidc 2 5-Dichlorobenzenesulphinic acid, 2-nitrothiophenol and the disul-phide I n conclusion we desire to thank the Department of Scientific and Industrial Research for a grant which has enabled one of us to take part in this work.Our thanks are also due to Mr. W. E. Wright for the derivatives of 2 5-dibromophenyl mercaptan used in these experiments. KING’S COLLEGE LONDO?;. [Received November 1 lth 1924.224 MILLER AND SMILES : XXXVII1.-The Constitution of Disulphoxides. Part I I . By CECIL JAMES MILLER and SAMUEL SMILES. ALTHOUGH the symmetrical structure of the disulphoxides (I) has been generally accepted the evidence adduced in its favour has been incorrectly interpreted and is in fact ambiguous and many simple reactions of these substances appear clearly to indicate the thio-sulphonate structure (IT) such for example as their behaviour with zinc dust potassium sulphide sodium arsenite and mer-captans (J. 1924 125 176). Moreover this view is supported by syntheses of disulphoxides from sulphenic halides and silver sul-phinates (Zincke Annalen 1912 391 67). Since all the definite evidence thus favoured the thiosulphonate structure it was concluded that this should not be abandoned but instead should be regarded as the most probable until final proof of one structure or another was forthcoming.(I.) Ar*E*fi*Ar Ar*so,*s*Ar ArR*O*SAr (111. ) 00 (11.1 0 The experiments now described provide the necessary further proof of this unsymmetrical constitution in that (1) the action of Grignard’s reagent with the disulphoxides must be added to those already quoted as indicating this structure (2) the reduction of disulphoxides with hydrogen iodide which hitherto has been quoted as the main evidence for the symmetrical formula is not only use-less as such but accords with the thiosulphonate structure (3) the unsymmetrical character of the disulphoxides is demonstrated by synthesis. slthough the behaviour of numerous carbon compounds of sulphur with alkyl and aryl magnesium halides has been investi-gated (Lapworth J .1912 101 297; Hepworth and Clapham J., 1921 119 IlSS; Wedekind Rer. 1921 54 16041 that of the disulphoxides appears to have escaped attention. These substances are very readily attacked by the magiiesinm compounds and usin THE CONSTITUTION OF DISULPHOXIDES. PART 11. 225 equimolecular proportions of the reactants high yields of the sulphinic acids have been isolated fission of the disulphoxide taking place apparently according to the following scheme (IV) : The sulphinic acid was obtained in every case examined but the fate of the thioaryl group of the disulphoxide varied according to the nature of the magnesium compound used with magnesium inethyl iodide the methyl aryl sulphide Ar*S*CH, was commonly obtained whilst with magnesium phenyl iodide the disulphide and diphenyl also were observed.The chief point at issue is tlhe situation of the oxygen in the molecule of the disulphoxide and with the symmetrical formula it is very difficult to explain the isolation of SOY; of this as sulphinic acid. According to this symmetrical constitution it would be expected that fission of the molecule would result in the formation of a sulphoxide (V) but this has not been observed. The results of this decomposition evidently support the unsymmetrical structure and appear inexplicable on the basis of the symmetrical formula. The mode of experiment is described and the results are tabulated on subsequent pages.The fact that the disulphoxides are easily reduced by hydrogen iodide t o disulphides has been regarded as sufficient reason for rejecting the thiosulphonate structure and as proving the presence of the true disulphoxide arrangement (Hinsberg Ber. 1908 41, 2836 4294; 1909 42 1278; Fries Ber. 1914 47 1195). It is assumed that the reduction is direct and since the sulphonyl group usually is not easily reduced by this reagent whilst the thionyl group is readily attacked the conclusion would seem justified but it has been pointed out (Smiles and Gibson J. 1924 125 176) that the force of the argument depends on the assumption that reduction is direct and is not accompanied by fission of the mole-cule. If however rupture does take place the reduction cannot be used as an argument against the unsymmetrical or in favour of the symmetrical formula for in either case a disulphide must be expected as the final product of the reaction.The question whether this fission takes place or not may be answered by the reduction of a disulphoxide containing different aromatic groups. I n such a case, if the unsymmetrical disulphide is the sole product (VI) the assumption of direct reduction is clearly justified but if both or even one of the symmetrical disulphides are obtlained (VII) it must be admitted that rupture of the dithio-system has taken place. (vI.) R1*SO*SO*R2 -+ R1*S-S*R2 (VII.) 2R1*S,0,*R2 -+ (RlS) + (R2S), VOL. CXXVII. 226 MILLER AND SMILES: Disulphoxides containing merent aromatic nuclei are readily prepared by an extension of the synthetical method of Zincke (Zoc.cit.) For example when the silver salts of p-toluenesulphinic acid and p-chlorobenzenesulphinic acid are treated with o-nitrophenyl-sulphur chloride the o-nitrophenyl p-toluenethiolsulp~ona.te (VIII) and o-nitrophenyl p-chlorobenzenethiolsulphonate (IX) are respec-tively 'formed. N02*C,H4*SCI + AgSO,*C,H = AgCl + NO,*C,H,*S*SO,*C,H,. (VIII.) Reduction of these substances with hydrogen iodide in the usual manner (compare Hinsberg Zoc. cit.) furnished o-nitrophenyl di-sulphide (N02*C,H,*S*), in good yield and in the case of the chloro-derivative p-chlorophenyl disulphide also was isolated. Hence it is clear that the disulphoxide system has. been broken during the process of reduction and the reaction cannot be admitted as evidence favouring the symmetrical and excluding the unsymmetrical structure.Nothing else can be deduced from this result but further information has been gained by adopting milder conditions of reduction with hydrogen iodide or by the use of dilute hyposulphite (in presence of sodium carbonate). The derivative (VIII) then yielded p-toluenesulphinic acid and the nitrophenyl disulphide the chloronitro-derivative (IX) gave p-chlorobenzenesulphinic acid and the nitro-disulphide whilst p-tolyl disulphoxide and p-chlorophenyl disulphoxide gave the corresponding sulphinic acids and disulphides or mercaptans. These results are similar to those obtained by Gutmann (Ber. 1914 47 635) on reduction with sodium arsenite. In the cases of disulphoxides containing similar aromatic groups this result does not enable a decision to be made in favour of either structure for both permit the formation of a sulphinic acid by hydrolytic fission, R*SO-SO*R + R*SO,H + R*S*OH + R*SO,*SR whilst the disulphide would result from reduction of the less stable sulphenic acids.Turning to the case of a disulphoxide containing different aromatic groups a distinction between the requirements of either structure is found. On the basis of the symmetrical arrangement the formation of two sulphinic acids may be expected, for there is no reason to suppose that hydrolysis will take place in one direction only when R1 and R2 are of similar character : RlS*OH + R2*S0,H +- R1*SO*SO*R2 -+ R1*S02H + R2*S*OH THE CONSTITUTION OF DISULPHOXIDES.PART 11. 227 On the other hand according to the thiosulphonate structure only one sulphinic acid should be formed : 2R1*S0,*S=R2 + 2R1*S0,H + (SR2)2 The latter condition accords with the result obtained by the reduction of these substances only one sulphinic acid has been isolated in each case examined. Moreover it is significant that the sulphinic acid-isolated was always the one which had been used as ifi component in the synthesis thus nitrophenyl t'oluenethiolsulphonate (VIII) yielded p-toluenesulphinic acid and o-nitrophenyl disulphide, and nitrophenyl chlorobenzenethiolsulphonate (IX) gave p-chloro-benzenesulphinic acid and the nitro-disulphide : 2CsH,C1*S0,*S*C6H4*No2 + 2C6H,C1*S02H + (NO,*C,I~,*S.),. The action of mercaptans with these disulphoxides is closely analo-gous (compare Smiles and Gibson loc.cit.). o-Nitrophenyl mercaptan yields in both cases the o-nitrophenyl disulphide and the corre-ponding sulphinic acid : CiH,.SO,* S-CGH,*NO, H S*CGH,-N02 The information gained from this extended study of the reduction of the disulphoxides therefore clearly favours the unsymmetrical structure. Taking a general review of the characteristic reactions of the disulphoxides it is seen that all yield the sulphinic acid and a product which contains the thioaryl group; these substances result from the action of zinc dust magnesium alkyl halides potass-ium sulphide mercaptans and of arsenite and other mild reducing agents. The majority of these reactions indicate an unsymmetrical structure a few are ambiguous in their import whilst none insists on the syrnmet8rical arrangement.To establish more firmly the unsymmetrical constitution disul-phoxides each containing different aromatic groups have been synthesised by alternate methods. I n the first Eeries of experiments it seemed desirable to arrange that the aromatic groups K1 and R2 should differ by the nature of the substituents present and not merely by the position of these. The more stable of the known aromatic sulphur chlorides contain the nitrc-group and it was obvious to choose one of these as a component of one method of synthesis but the chief dificulty lay in obtaining a sulphur chloride of sufficient stability without this group for use in the alternate I 228 MILLER AND SMILES : process. This was overcome by the preparation of 2 5-dichloro-phenylsulphur chloride (X) from the corresponding disulphide and chlorine.From 2-nitrophenylsulphur chloride and silver 2 5-dichloro-benzenesulphinate a disulphoxide (XIa) of m. p. 142" was obtained, whilst from 2 5-dichlorobenzenesulphur chloride and silver 2-nitro-benzenesulphinate an isomeric compound (XIIa) melting at 129" was prepared. These substances differ not only in their physical properties but also in their chemical behaviour. When 2-nitro-phenyl 2 5-dichlorobenzenethiolsulphonate (XIa) is treated with 2-nitrophenyl mercaptan the dichlorobenzenesulphinic acid and o-nitrophenyl disdphide are formed (XI) whilst the action of the same mercaptan with 2 5-dichlorophenyl 2-nitrobenzenethiol-sulphonate (XIIa) yields o-nitrobenzenesulphinic acid and a mixture of the two possible disulphides (compare Smiles and Gibson Zoc.cit.). c1 \/ (XII.) c1 (X). The action of 2 5-dichlorophenyl mercaptan with the latter di-sulphoxide yielded (XII) the nitrosulphinic acid and the tetra-chloro-disulphide. Mild reduction of these disulphoxides gave analogous results ; these are collected on a subsequent page. A second series of experiments was made with material in which the groups R1 and R2 were closely similar for it was thought that such conditions would be favourable to intramolecular change-if indeed this were possible-of one thiolsulphonate to the other or of the thiolsulphonate to the true a-disulphoxide structure the latter change having been suggested by previous workers (Hinsberg Zoc.cit.; Fries Zoc. cit.). The materials chosen were the 2 5-dichloro-and 2 5-dibromo-phenyl derivatives. Thus 2 5-dichlorophenyl-sulphur chloride (X) and silver 2 5-dibromobenzenesulphinate furnished 2 5-dichlorophenyl 2 5-dibromobenzenethiolsulphonate (XIIIa) which melted at 119" even after being heated to 150" and cooled. On treatment with 2 5-dichlorophenyl mercaptan the tetra-chloro-disulphide and 2 5-dibromobenzenesulphinic acid were formed (XTTI) the structure indicated by synthesis being thus confirmed THE CONSTITUTION OF DISULPHOXIDES. PART 11. 223 The isomeric 2 5-dibromophenyl 2 5-dichlorobenzenethiol-sulphonate (XIVa) prepared in a similar manner melted a t 125" and with 2 5-dibromophenyl mercaptan yielded the tetrabromo-disulphide and 2 5-dichlorobcnzenesulphinic acid (XIV) .I n summarising the results of these experiments it is claimed that an unsymmetrical structure must be assigned to the disulph-oxides. Two unsymmetrical formulz have been advocated the thiolsulphonate and the anhydride arrangements (I1 and 111) ; there is ample reason for discarding the latter (Fries Zoc. cit.) whilst the former accords very closely with all the facts since discovered. E x P E R I M E N T A L. The disulphoxides were obtained those containing similar aromatic groups from the sulphinic acids by the usual method and those with dissimilar groups by the reaction of a silver sulphinate with the requisite sulphur chloride as follows. A solution of the sulphur halide in dry ether was shaken a t the ordinary t'emperature or boiled with an excess of the silver sulphinate until all the sulphur halide had disappeared the t8reatment being adapted to the reac-tivity of the materials and the stability of the sulphur halide used.Generally the disulphoxide separated as the reaction proceeded ; when this was complete the solid material was collected and the ciisulphoxide was extracted from it with a suitable solvent. In the majority of the cases studied the yields were good. The following substances were obtained by this method from the stated com-ponents. 2-Kit rophen yl 4-tolu e?zethiolsulphonate 02N*C6H4*S-S O,*C ,H,Me, from o-nitrophenylsulphur chloride and silver toluenesulphinate, formed colourless prisms m. p. 97" soluble in the usual organic solvents (Found S = 20-7 ; C = 50.0 ; H = 3.7.C1,HIIO,NS, requires S = 20.69; C = 50.55; H = 3.58 yo). The alcoholic solution of this substance became blue on addition of aqueous sodium hydroxide. 2-Xitrophenyl 4-chlorobenzenethioleulp~o?~ate, 0,EC ,H,.S.S Q2*C ,H,CI , from o-nitrophenylsulphur chloride and silver p-chlorobenzene-sulphinate formed colourless prisms m. p. 123" soluble in the usual organic media (Found S = 19-6; C1 = 10-5; C = 43.3; H = 2-66. Cl2H,O,NC1S2 requires S = 19-42 ; C1 = 10.74 ; C = 43.68 ; H = 2-44 Yo). 3-Xitrophenyl 2 5-dichiorobe?azenethiolsulphonate, 0,N*CsH4.S*S0,*CGH,Cl*, from o-nitrophenylsulphur chloride and silver 2 5-dichlorobenzene-sulphinate melted a t 142" and was sparingly soluble in commo 230 MILLER AND SMILES : solvents (Found C1= 19.4; S = 17.6.C1,H,04NCl,S requires C1 = 19.47; S = 17.61 yo). Like the foregoing disulphoxides, this substance gave a deep blue solution with alcoholic sodium hydroxide. 2 5-Dichlorophenyl 2-nitrobenzenethiolsulphonate, from silver o-nitrobenzenesulphinate and 2 5-dichlorophenyl-sulphur chloride separated from hot alcohol in colourless needles, m. p. 129" (Found S = 17:6; C1 = 19.6. C1,H,O,NC1,S requires S = 17.6 ; C1= 19.47 yo) ; aqueous sodium hydroxide added to the alcoholic solution did not give the characteristic blue colour. A mixture of the two isomeric substances in approximately equal amounts melted indefinitely between 88-95'. 2 5-Dichlorophenyl 2 5-dibromobenzenethioIsulphona.te, C6H,C1,*S*S0,*C &€,Br2, from 2 5-dichlorophenylsulphur chloride and silver 2 5-dibromo-benzenesulphinate separated from alcohol in colourless needles, m.p. 119' (Found S = 13.8; C1 + Br = 48.3. Cl,H60,C12Br,S, requires S = 13.45; C1+ Rr = 48-38 %). 2 5-Dibromophenyl 2 5-dichlorobenzenethiolsulphonate, C,H,Br,-S*S0,*C6H,C12, from silver 2 5-dichlorobenzenesulphinate and 2 5-dibromo-phenylsulphur bromide separated from hot alcohol in colourless prisms m. p. 125." This melting point remained unaltered after a sample had been fused and cooled. The substance is less soluble than the isomeric disulphoxide and a mixture of the two in approxi-mately equal amounts melted indefinitely at 110-114" (Found : S = 13.4; C1 + Br = 48.5. C12H60,C1,Br,S requires S = 13.45; C1 + Br = 48.38 yo).2 5-Dichlorophenylsulphur chloride C6H,Cl,*SC1 was prepared by saturating a concentrated solution of the corresponding disulphide in dry carbon tetrachloride with chlorine. The residue obtained after the solvent had been evaporated was kept under diminished pressure when it solidified. The product was purified by crystallis-ation from ice-cold ether when the substance was obtained in golden-yellow needles m. p. 32-33' (Found C1 = 49.8; S = 14.9. C,H3Cl,S requires C1 = 49.8; S = 15.03 %). The substance was very soluble in organic media; dilute aqueous sodium hydroxide gave the corresponding disulphide and alkali sulphinate. The 2 5-dibromophenylsulphur bromide was obtained by a similar process as a yellow crystalline material but owing to its instability attempts to isolate it in a pure condition for analysis were un-successful some loss of bromine occurring with formation of the disulphide THE CONSTITUTION OF DISULPHOXIDES.PART 11. 231 Behaviour o j Disulphoxides with Methyl and Phenyl Magnesium Halides.-A dilute solution of the disulphoxide in ether was added to a cooled solution of the magnesium compound (1 mol.) in the Magnesium Sulphinic Other substances Disulphoside. derivative. acid yo. observed. D iphenyl Methyl 63 Ph*S.Me ; PhS.SPh. Diphenyl Phenyl 70 Ph*S.Ph ; PhS-SPh ; Ph-Ph. Di-p-tolyl Methyl S7 C,H,.S-Me; C,H,-S.SC,H,. Di-p-t olyl Pheiiyl 76 C,H,-S.Ph; (C,H,.S.)2 and Di-p-chlorophenyl Methyl 90 C6H,C1.S*&~e; (C6H,C1*S*)2. Ph *Ph . 2 5 2' 5'-Tetra- 7 9 82 4 4'-Dimethoxy- 7 7 68 MeO.C,H,.S.Me ; and disul-chlorodiphcnyl tolyl 3 S'-disul- phide.phoxide same solvent. The mixture was kept for 12 hours before treatment with water. Sufficient aqueous sodium thiosulphate was then added to remove free iodine if this were present and finally excess of dilute sulphuric acid. The sulphinic acid was extracted from the ethereal solution with dilute alkali the substances remaining being separately examined. A summary of the results is given in the foregoing table. The third column shows the approximate per-centages in which the sulphinic acids were isolated. These substances were identified by comparison with authentic samples and by con-version to the methylsulphones which were similarly compared. The latter have been previously described in literature with the exception of the following.4-Chlorophenylmethylsulphone C,H,Cl*SO,*CH, colourless needles, m. p. 96" was obtained by oxidation of the sulphide and by methylation of the sodium sulphinate with methyl sulphate (Found C1= 18.7; S = 16.6. C7H70,ClS requires C1= 18-58; 2 5-Dichlorophenylmethylsulp~one C6H,Cl,*S0,*CH, prepared in a similar manner separated from hot water in needles m. p. 88" (Found S = 14-0. C7H,02C12S requires S = 14-24 yo). The identification of sulphides was generally effected by oxidation to the sulphones. Action of Mercaptans with Disu1phoxides.-The method of opera-tion was similar to that already described (Smiles and Gibson Eoc. cit.) molecular proportions of the reagents being taken in alcohol. The results are collected in the following table ; the third and fourth columns respectively show the sulphinic acid and the disulphide which were isolated from the interaction of the stated disulphoxide and mercaptan.The yields of these products were generally of the order of 80% of theory and higher when the sparingly soluble S = 16.8 %) 232 THE CONSTITUTION OF DISULPHOXIDES. 2 2’-dinitrodiphenyl disulphide was dealt with. the h t column refer to numbered formulae. Disulphoxide. Mercaptan. 2-Nitrophenyl 2-Nitrophenyl 4-toluenethiol-sulphonate (VIII) 4-chloro benzene thiolsulphonate 2 5-dichloro-benzenethiol-sulphonate (XI) phenyl 2-nitro- phenyl benzenethiol-sulphonate phenyl 2-nitro-benzenethiol-sulphonate 2 5-Dichloro-phenyl 2 5-di- phenyl bromobenzene-thiolsulphonate phenyl 2 5-di- phenyl chlorobenzene-thiolsulphonate 2-Nitrophenyl 99 (1.W 2 -Nitrophenyl 9 9 2 5-Dichloro- 2 5-Dichloro-(=I) 2 5-Dichloro- 2-Nitrophenyl (=I) 2 5-Dichloro-(XIII) 2 5-Dibromo- 2 5-Dibromo-(XW) Sulphinic acid.4-Toluene-4-Chloro benzene-2 5-Dichloro-benzene-2 -Nitro benzene-2 6-Dibromo-benzene-2 5-Dichloro-benzene-PART 11. The numerals in Disulphide. Di-o-nitrophenyl 2 5 2’ 5’-Tetre-chlorodiphenyl A mixture not separated 2 5 2‘ 5’-Tetra-chlorodiphenyl 2 5 2’ B’-Tetra-bromodiphenyl Reduction of Disulphoxide+.-Three methods of reduction were employed. In casm 1 2 and 3 of the following table where com-plete reduction with hydrogen iodide was required the conditions devised by Hinsberg (Ber.1908 41 4295) were used. Milder reduction with this reagent was effected as follows. About 2 g. of the disulphoxide 3-4 C.C. of glacial acetic acid and a few C.C. of a saturated solution of sodium bisulphite containing ten drops of hydriodic acid (d 1.9) were constantly shaken with 50-100 C.C. of light petroleum for about 4 hour the necessary time varying some-what according to the disulphoxide taken. Then aqueous sodium carbonate was added until the whole was alkaline. I n cases where the dinitro-disulphide was formed this separated almost completely from the liquid. The solid material was collected the petroleum and the aqueous portion of the liquid being separately examined. The sulphinic acid was isolated from the aqueous portion and the di-sulphide from the solid or the petroleum.Examples of this type of reduction are given in Nos. 4,5,6 and 7 of the table. Reduction with sodium hyposulphite was conducted by shaking the disulph RESOLUTION OF CHLOROSULPHOACETIC ACID ETC. 233 oxide with a slight excess of a 3-4 yo aqueous solution of the reagent to which about l/lOth of its volume of alcohol had been added the mixture being kept alkaline by the addition of sodium carbonate. The disulphide formed was isolated by solution in ether whiIst the sulphinic acid and mercaptan were obtained from the aqueous portion. When mercaptan was present this was separated from the sulphinic acid as disulphide by treating the solution of sodium salts with a current of air.Examples are shown in h‘os. 8 9 and 10 of the table. 1. 2. 3. 4. 5. 6. r S. 9. 10. Disulphoxide. 2-Nitrophenyl 4-toluencthiol-sulphonate 2-Nitrophenyl 4-chlorobenzene-thiolsulphonate 2 5-Dichlorophenyl 2-nitro-benzenethiolsulphonate 2-Nitrophenyl 4-toluenethiol-sulphonate 2-Nitrophenpl 4-chlorobenzene-tliiolsulphonate Di-4-chloroplienyl disulphoxide 2-Nitrophenyl 2 5-dichloro-bcnzenethiolsulphonate Di-p-tolyl disulphoxide 2 5-Dichlorophenyl 2-nitro-benzenethiolsulphonatc 2-Nitrophenyl 2 5-dichloro-beiizenethiolsulphcnate Products isolated. 2 2’-Dinitrodiphenyl disulphide 2 2’-Dinitrodiphenyl disulphide and 4 4’-dichlorodiphenyl disulphide A mixture of disulphides; not separated 4-Toluenesulphinic acid and 2 2’-di-nitrodiphenyl disulphide 4-Chlorobenzenesulphinic acid and 2 2’-dinitrodiphenyl disulphide 4-Chlorobenzenesulphinic acid and 4 4’-clichlorodiphenyl disulphide 2 5-Dichlorobenzenesulphinic acid and 2 2’-dinitrodip~ienyl clisul-phide p-Toluenesulphinic acid and p-tolyl mercaptan 2-Nitrobenzenesulphinic acid and 2 5 2‘ 5’-tetrachlorodiphenyl cli-sulphidc 2 5-Dichlorobenzenesulphinic acid, 2-nitrothiophenol and the disul-phide I n conclusion we desire to thank the Department of Scientific and Industrial Research for a grant which has enabled one of us to take part in this work. Our thanks are also due to Mr. W. E. Wright for the derivatives of 2 5-dibromophenyl mercaptan used in these experiments. KING’S COLLEGE LONDO?;. [Received November 1 lth 1924.