1960 KENNER AND WITHAM : TOLANE DERlVATIVES FROMCCVIII.--T/ze Fwmation of Tolam Derivutives fromp- Chloq-otoluene and 3 : 4-Dicldorotoluene.By JAMES EENNER and ERNEST WITHAM.As is well known, the chlorination of toluene at its boiling pointleads to the formation of benzotrichloride. On one occasion, how-ever, Liebermann and Homeyer (Ber., 1879, 12, 1971), and laterGattermann (Ber., 1884, 17, 2601), observed the formation of asolid substance as chief product during this reaction, and identifiedit as tolane tetrachloride, which had previously been obtained byZinin by the action of phosphorus pentachloride on benzil.During the preparation of chlorobenzylidene chlorides fromp-chlorotoluene and 3 : 4-dichlorotoluene, Armstrong and Wynnesimilarly obtained products which were identified as di- and tetra-chloro-tolane tetrachlorides. The present authors have also hadthe same experience in preparing pchlorobenzotrichloride fromp-chlorotoluene, but, in common with earlier investigators of thisreaction, they have failed to find a clue to the conditions whichlead to the formation of the tolane derivative.Liebermann and Homeyer (Zoc. cit.), it is true, attributed thecondensation to sulphuric acid which had been carried over into themixture, but in the case of the authors’ experiments such anexplanation is improbable, as the gas, after being passed throughsulphuric acid, was led through a tube packed with asbestos, and,moreover, the introduction of sulphuric acid into the boilingchlorotoluene has not been found to initiate the condensation.This is the more unfortunate as the only other general methoP-CIILOROTOLUEKE AND 3 : ~-~IC€ILOI~OTOLUENE.1961of preparing tolane tetrachloride and its derivatives is that usedfirst by Hanhart (Ber., 1882, 15, 899), which consists in the removalof chlorine from benzotrichloride by the action of copper powder,and does not give satisfactory results, either as regards yield orpurity of product. Hanhart, apparently unable to isolate anydefinite substance from the tarry product, had recourse to dis-tillation, and obtained the two forms of tolane dichloride, nowrecognised as the stereoisomerides :Onufrowicz (Ber., 1884, 17, 833), who succeeded in isolating thetetrachloride, showed a t the same time that it could be convertedinto cis-tolane dichloride by distillation, thus explaining Hanhart’sresult.Later, Fox (Ber., 1893, 26, 653), without reference to Onufrowicz’spaper, carried out the same reaction with o-chlorobenzotrichloride,and distilled the product, obtaining the cis- and trans-2 : 2/-dichloro-tolane dichlorides.The authors have repeated Fox’s work in thehope of separating the tetrachloride, but have been unable to findany evidence of its production, and, as recorded on p. 1966,succeeded, as he did, in isolating only the dichlorides.As p-chlorobenzotrichloride, unlike the o-derivative, forms thetetrachloride, which can be isolated without difficulty, the oppor-tunity was taken of testing the applicability of the decompositionobserved by Onufrowicz to substituted tolane tetrachlorides.Aswas expected, 4 : 4/-dichlorotolane tetrachloride, when distilled,behaves in a precisely similar manner to tolane tetrachloride, andyields the cis-form of the dichloride.Further, the tetrachloride was submitted to reduction, and resultsof some interest were obtained. Zinin (Ber., 1871, 4, 289) showedthat the action of zinc on alcoholic solutions of tolane tetrachlorideled to the formation of the stereoisomeric dichlorides. Liebermannand -Horneyer (Bey., 1879, 12, 1971) used zinc dust in place ofzinc, and obtained the same products, whilst Lachowicz (Ber., 1884,17, 1165) subsequently arrived at the same result by reducing a,boiling concentrated acetic acid solution with iron filings.Wislicenus and Blank (Annalen, 1888, 248, 1) repeated Zinin’swork, and Eiloart (Amer.Chem. J., 1890, 12, 239), who studiedthe two dichlorides from the stereochemical point of view, upheldthe opinion already expressed by Wislicenus (Zoc. cit.), that theless soluble isomeride, with the higher melting point, was the cis-form, although it may be remarked that this is contrary to theusual rule.VOL. XCVLI. 6 1962 KEKNER AND WITHAM : TOLANE DERIVATIVES FROMThe reduction products obtained from 4 : 4’-dichlorotolane tetra-chloride in alcoholic solution by means of zinc dust vary accordingto the temperature employed, I f conducted at 50°, the productsare the two 4 : 4’-dichlorotolane dichlorides, which respectivelymelt a t 166-167O and 86--87O, and, in addition, 4: 4/-dichloro-tolane; but if carried out at the boiling point of the alcoholicsolution, the isomeride melting at 166-167O can no longer be found,the reduction products being the dichloride of lower melting pointand 4 : 4/-dichlorotolane.If the conclusions drawn by Wislicenus and Blank respectingthe cis- and trans-isomerism of the tolane dichlorides can be accepted,then it will follow by analogy that the dichlorotolane dichloride,melting at 166-167O, will be the cis-form.The entire absence ofthis isomeride from the reduction product obtained by working atthe higher temperature, and the evidence that, unlike the dichloride(m. p. 86--87O), it is readily reduced may perhaps be regarded assupporting this view.As already mentioned, tolane tetrachloride was originallyobtained by Zinin by the action of phosphorus pentachloride onbenzil, and this would doubtless provide a general method of pre-paration were it not that, as is well known, the isolation of thebenzoins which should furnish the benzils on oxidation has beenaccomplished only in a few instances.Especially is this the casewhere negatively substituted aldehydes are submitted to the benzoinreaction. The authors hoped to use o-chlorobenzaldehyde as thestarting point for the preparation of the 2 : 2/-dichlorotolane tetra-chloride, which could not be isolated in the experiments alreadymentioned, but found that when the benzoin condensation wasattempted, o-chlorobenzoic acid was the chief product. Greatersuccess attended experiments with p-chlorobenzaldehyde, foralthough the product of the benzoin condensation was a resinousmass from which the pure substance could not be isolated, oxidationwith nitric acid led to the production of the 4 : 4’-dichlorobenzil,identical with the substance obtained by hydrolysing 4 : 4/-dichloro-tolane tetrachloride with acetic acid by Liebermann and Homeyer’smet hod.It has not hitherto been possible to apply the reaction to thepreparation of 2 : 2/-dichlorobenzil, because the attempt to prepare2 : 21-dichlorotolane tetrachloride by the chlorination of the cis-dichloride merely led to the conversion of the greater part into themore stable trans-form.Chlorine is known to be a particularlyeffective agent f o r the interconversion of stereoisomerides.Liebermann and I-Iomeyer (Zoc.cit.) confirmed Limpricht”s statementthat the tolane dichlorides were stable towards acetic acid, and thP-CHLOROTOLUENE AND 3 : 4-DICRI.OROTOLt7EKE. 1963present authors found the same to hold good for the 2 : 2’-dichloro-tolane dichlorides.E X P E R I Bf EN T A L.*Armatrong and Wynne (private communication), in the pre-paration of p-chlorobenzylidene chloride from p-chlorotoluene,employed chlorine prepared from manganese dioxide and hydro-chloric acid, washed by passage through water, and dried oversulphuric acid. For each of seven operations, 126 grams of p-chloro-toluene were used, and it was found that in the first two only avery small quantity, but ih the last five, considerable amounk, of4 : 4’-dichlorotolane tetrachloride separated in a crystalline formon cooling.The rate at which the chlorine was absorbed did notseem to influence the tolane condensation: thus, in the first twoand the last two operations the figures f o r increase in weight due tochlorination and, in brackets, the duration in hours were 74 (16),71 (15), 70 (14), 67 (16)-the calculated increase in weight being69 grams.The present authors have used liquid chlorine obtained fromKahlbaum as the source of the gas, and have passed the chlorinethrough sulphuric acid (to allow of the rate being checked) andasbestos wool before introducing it into the boiling p-chlorotoluene.With one exception they have been unsuccessful in obtainingthe 4 : 4’-dichlorotolane tetrachloride.Throughout, they havechlorinated to the stage of p-chlorobenzotrichloride, and the usualdaily increase in weight was 15 grams, but on the one occasionwhen the tetrachloride was formed, 44 grams on the first day, and22 on the second, were taken up. The yield of the tetrachloridefrom 100 grams of p-chlorotoluene was 44 grams, and the motherliquor consisted of 21-chlorobenzotrichloride. All attempts t o repeatthe conditions leading to the formation of the tolane derivativehave been unsuccessful.4 : 4’ - Dic7dorotoZane tetrachloride, C,I~,C1*CCl,*CC12’C,H,CI,crystallises from chloroform in thin, well-defined, flat plates. Itmelts at 190°, and dissolves only sparingly in alcohol, but readilyin chloroform, benzene, or light petroleum :(P) 0.5923 gave 0.9367 CO, and 0.1212 H@. C =43.13; H = 2.27.0.327 ,, 0.7216 AgC1.Cl=54.58.C,,H&& requires C = 43.19 ; H = 2.06 ; C1= 54.75 per cent.* The study of 4 : 4’-diclilorotolane tetrachloride was begun by Mr. S. Parrish,B.Sc., a t the Royal College of Science. He examined its reduction prodncts,obtninillg the cis- and trans-4 : 4’-dichloroto~ane dichlorides, and prepared from it4 : 4’-dichlorobeuzil and the corresponding hydrazone. Unfortunately, he wasunable to complete the work, and the investigation has been taken up afresh by thepresent authors. Analyses prefixed by the letter P were made by Mr. Parrish.-- w. P. w.6 ~ 1964 RENNER AND WITIIA&I : TOLASE DERIVA'rIVE3 FROM3 : 4 : 31 : 4'-Tetracl~lorotok~ne tetrachloride,C6H,C12* CCl,*CCl,= C,H,C12,is formed when 3: 4-dichlorotoluene is chlorinated to the stage of3 : 4-dichlorobenzylidene chloride under conditions such thatsulphuric acid used as drying agent could not be carried forwardmechanically (Armstrong and Wynne, private communication).The tetrachloride in this case did not separate on cooling, but wasisolated by extraction with chloroform from the semi-solid residueleft after fractionation of the 3 : 4-dichlorobenzylidene chloride(b.p. 274-275O). It crystallises from chloroform in small, diamond-shaped plates, which melt at 197O:0.1214 gave 0.3037 AgC1. C1= 61.88.C,,H,Cl, requires C1= 61.97 per cent.4 : 4'-Dicldorotot?ane Tetrachloride f r o m p-Chlorobennoti-i-To prepare the tetrachloride, p-chlorobenzotrichloride (40 grams),benzene (60 grams), and copper powder (30 grams) were boiled ina reflux apparatus for about six hours until the reaction seemedto be complete.The hot solution, freed from copper powder byfiltration, deposited crystals of the tetrachloride and of p-chloro-benzoic acid, the latter of which was removed by extraction withsodium carbonate solution. The yield of tetrachloride from 207grams of the trichloride amounted to 110 grams.On crystallisation from chloroform, it was obtained in platesidentical in appearance, melting point ( 190°), and composition(C1=54*58 per cent.) with the product formed during thechlorination of boiling p-chlorotoluene.When heated at 230°, the tetrachloride decomposes with theelimination of hydrogen chloride, and at a higher temperature,about 270°, chlorine also can be detected in the escaping gas.Thisdecomposition was studied inore closely by distilling 5 grams ofthe tetrachloride under a pressure of 90-130 mm. The firstportion of the distillate was a colourless liquid, which passed over at234O, but the remainder was solid, and, when crystallised fromacetone, formed small, rhombic crystals, which melted at 166-167O.This crystalline compound was subsequently identified as cis-4 : 4f-diclzlorotolane dichloride (p. 1965) :0-192 gave 0,3712 CO, and 0.045 H,O.c I& I o i d e.C =52*72 ; H= 2.60.C14H8C14 requires C = 52.85 ; H = 2-54 per cent.Reduction of 4 : 41-Dichlorotolane Tetrachloride.The reduction was carried out by suspending the tetrachloride(5 grams) in boiling absolute alcohol (500 c.c.), and addinP-CHLOROTOLUENE AND 3 : 4-DICEILOROTOLUENE.1965gradually during two days zinc dust (4 grams) and glacial aceticacid (30 c.c.). The solution, freed from zinc dust by filtration,deposited, on cooling, very thin, nacreous flakes, amounting in allto 1.5 grams.Concentration of the alcoholic filtrate to 150 C.C. furnishedmaterial amounting in all to 1.2 grams, which consisted of long,slender, brittIe, prismatic needles, and was identified as trans-4 : 4/-dichlorotoIane dichlorrde. Further concentration gave onlysmall, ill-defined separations, mixed with zinc acetate.By carrying out the reduction a t 50°, instmead of at the boilingpoint, a, third substance, crystallising in lustrous, thin, diamond-shaped plates, was obtained in addition to the other two.It wasidentified as cis-4 : 4/-dichlorotolane dichloride, but no estimatecould be formed of the relative proportion of the two isomeridesin the reduction product.cis-4 : 4/-DicldorotoZane dichloride, C6H4C1*CC1:CC1~C6H4Cl, is lesssoluble in alcohol than the trans-isomeride. It melts at 166-167O :(P) 0.1710 gave 0.3087 AgCI.When reduced in boiling alcoholic solution with zinc dust andtrans-4 : 4~-DichZorotoZane dichloride dissolves readily in alcohoI,ItC = 52.45 ; H = 2-70.On analysii, this proved to be 4 : 4’-dichlorotolane.C1=44*66.q4H8C14 requires C1= 44.62 per cent.acetic acid, it yields 4 : 4/-dichIorotolane.and separates from solution in characteristic, prismatic needles.melts at 86-87O:(P) 0.354 gave 0.6808 CO, and 0-0861 HiO.0’3347 ,) 0.5986 AgCl.C1= 44.25.C,,H,CI, requires C= 52.83 ; H = 2.52 ; C1= 44.65 per cent.4 : 4~-DichZorotoZane, C,H,Cl-CI C*c6H4cl, is much less soluble inalcohol than the two dichlorides, and appears to be dimorphous, asit crystallises from solution sometimes in nacreous flakes, and some-times in slender needles. Usually it melts a t 175--176O, but onone occasion the melting point first observed was 153-154O, which,after solidification of the specimen and re’-fusion, rose to andremained constant at 175-176O. On analysis of two differentpreparations :(i) 0.1626 gave 0.4064 CO, and 0.0518 H,O. C = 68.16 ; H =3*54.(ii) 0-1710 ,, 0.4262 CO, ,, 0.0458 H,O.C = 67.97; H =2-97.(i) 0.3517 ), 0.4080 AgC1. C1=28*70.C,,H8C1, requires C = 68.02 ; H = 3-26 ; C1= 28-72 per cent.Cl,Hl,Cl, ,, C = 67-47 ; H = 4-05 ; C1= 28.48 ),These analyses point. to the tolane formula, but the productionof a tolane, rather than a stilbene or dibenzyl derivative, by reduc-tion of the dichlorotolaae tetrachloride seemed sufficiently remark1966 KENNER AND WITHAM: TOLANE DERIVATIVES FROMable to require confirmation. Accordingly, chlorine was passedthrough a solution of the compound in chloroform until no moreof the gas was taken up, and, after removal of the solvent, it wasfound that the product consisted of 4 : 4/-dichlorotolane tetra-chloride in practically quantitative amount.Hence the conclusionis drawn that the reduction product is, as stated, 4: 4/-dichloro-tolane.The formation of a tetrachloride during the preparation ofo-chlorobenzotrichloride from o-chlorotoluene was not observed, andas attempts to make 2 : 2~-dichlorohenzil, which would have fur-nished a means of obtaining it, were unsuccessful, recourse was hadto FOX’S process in t,he hope that it might be isolated from theproduct of the action of copper powder on o-chlorobenzotrichloride.T T ~ a t m e n t of 0-Ch lor o b en xo t r i c h I o r ide with Copper Pow d e r.o-Chlorobenzotrichloride (20 grams) was dissolved in benzene(30 grams), and heated with copper powder (15 grams) for twenty-five hours in a reflux apparatus. After filtration, repeatedextraction of the residue with benzene, and removal of the benzeneby distillation, the residue was allowed to dry on a porous platefor a fortnight.In spite of many attempts with different solvents, it was notfound possible t o extract from this residue any substance otherthan o-chlorobenzoic acid (m.p. 136O). As a similar residue frombenzotrichloride, when heated at 200° with 80 per cent. acetic acid,gave a relatively large yield of benzil, a portion of this viscidsubstance was also heated with acetic acid, but furnished onlyo-chlorobenzoic acid. Yet when the part which had not gone intosolution in the acetic acid was dried and distilled, it gave bothcis- and trans-2 : 2’-dichlorotolane dichlorides. It is difficult toresist the conclusion that the tetrachloride is not formed in thecondensation, or, if formed, suffers decomposition into thedichlorides under the conditions of the experiment, being less stablethan the 4 : 4’-dichlorotolane tetrachloride, which requires t o beheated a t above 200° before decomposition into the dichloride takesplace.To confirm Fox’s results (Zoc.cit.), a portion of the viscid productwas distilled under 11 mm. pressure. Below 210° the distillatewas colourless, and solidified on cooling. By crystallisation fromlight petroleum, it was separated into two portions, the one con-sisting of clusters of radiate needles, melting at 136O, identicalwith o-chlorobenzoic acid, and the other of prismatic crystals,melting at 172-173O, which proved to be cis-2 : 2/-dichlorotolanP-CHLOROTOLUENE AND 3 : 4-DICEILOROTOLUENE. 1967dichloride (C1= 44.73 per cent.).The later distillate was brown,and, after purification, consisted for the most part of the cis-dichloride, whilst the residue in the flask, when extracted by lightpetroleum, yielded large, transparent rhombs, melting a t 127--128O,which agreed in properties with the description given of trans-2 : 2/-dichlorotolane dichloride (C1=44.5 per cent.).When the cis-2 : 2/-dichlorotolane dichloride is reduced in boilingalcoholic solution with zinc dust and acetic acid under the con-ditions already described for the 4 : 4/-isomeride, it is convertedinto 2 : 2/-dichlorotolane (m. p. 87O). The same substance is alsoformed when the dichloride is heated with copper powder at 260O.4 : 4’-Dic?dorob enail.-When 4 : 4’-dichlorotolane tetrachloride isheated with glacial acetic acid (40 c.c.) and water (10 c.c.) at 170°for six hours, it is converted into 4 : 4’-dichlorobenzil, whichseparates in the tube in long, slender, yellow needles, and whenrecrystallised from alcohol was found to melt at 193O.Montagne(Eec. trav. chim., 1902, [ii], 21, 19) gives the melting point as200° :C=59*60; H=3.09. (P) 0.3735 gave 0-8162 CO, and 0.1039 H,O.(P) 0.2318 ,, 0.2305 AgC1. C1=25*59.C,,H,O,Cl, requires C = 60.22 ; H = 2.89 ; C1= 25.42 per cent.Hantzsch and Glover obtained this compound from the corresponding benzoin, which was formed from p-chlorobenzaldehyde bythe benzoin condensation (Ber., 1907, 40, 1519). The quantitiesof materials used and the exact conditions are not given by them,and it may be of interest to note that while studying the benzoincondensation some years ago, Miss E. S. Hooper, B.Sc., failed toisolate the benzoin from this source, although the product, whenoxidised by nitric acid, gave a 40 per cent. yield of 4 : 4’-dichloro-benzil (m. p. 19l0), which wm identical with that obtained from4 : 4r-dichlorotolane tetrachloride (private communication).The rnonohgdrazone, C6H$l-C( :N*NH*CGH5) =CO*C,H,CI, crys-tallises from alcohol in small, yellow prisms, and melts at 178O:(P) 0.0965 gave 6.4 C.C. N, (moist) at 20.5O and 758 mm. N = 7-52.C2,H,,0N2Cl, requires N = 7.59 per cent.The authors wish to express their thanks to Professor Wynne,who directed their attention to this problem, and has given themmuch valuable assistance and advice during its investigation.THE UNIVERSITY,SHEFFIELD