88 TROTMAN PREPARATION OF QUATERNARY HYDROCARBONS. XV.-Preparation of Quuternary Hydrocarbons. By EDWARD RUSSELL TROTMAN. IN the course of some investigations in progress in this laboratory, certain quaternary hydrocarbons of the type (C,M,*CH,),CR,R~, were required R and R being different alkyl radicals. No such compounds are known and indeed very few quaternary hydro-carbons of any kind have been described. The interaction of Grjgnard reagents and tertiary halogen derivatives which seeme TROTMAN PREPARATION OF QUATERNARY HYDROCARBONS. 89 to offer a suitable method €or their preparation has already been studied. Gomberg and Cone (Ber. 1906 39 1461 2975) succeeded in displacing the chlorii-ie atom of triphenylmethyl chloride by various aromatic and aliphatic hydrocarbon radicals in this way.On the other hand Davis arid Kipping (J. 1911 99 300) were unable to prepare yy-dibenzyl-n-hexane (C,H,*CH,),CEtPr by the action of magnesium propyl bromide on cr-bromo- rcc-dibenzyl-propane (C,K5*CH,),CBrEt ; in every case the product consisted mainly if not entirely of an unsaturated hydrocarbcn. Spiith (Monatsh. 1913 34 1963) who made an exhaustive study of the act ion of Grigiiard reagents on organic halogen derivatives found that although quaternary hydrocarbons were generally produced, tlie yield was always poor and a mixture of numerous products was obtained. It appeared therefore that the interaction of a Grigiiard reagent ancl a tertiary halide often takes quite an unexpected course and this conclusioii is confirmed by the results described below.A preliminary study of tlie proposed method was made with [$-byorno- x-pher,yl-P-meth?llbutane a compound readily prepred from imthy! ethyl ketone. This bromide reacted with magnesium ethyl bromide 153th evolution of ethane ancl ethylene but the product appeared to consist entirely of a mixturz of the olefine, C,H,*CH:CMeEt and the saturated conipound C,H5*CH,.CH3kEt , a i d the possible presence of a small proportion of the quaternary hydrocarbon could not bt proved. Attempts to preparc a quaternary hydrocarbon by first converting the bromide into a Grignard reagent and then treating the product with ethyl bromide were also unsuccessful ; although the bromide reacted with mag-iiesiuin in thc presence of dry ether magnesium bromide separated even at 0" and a mixture of the saturated hydrocarbon and the olcfine was formed.~-B~o~?~o-sr-p7~enyl-~-benx~lp~o~ane which was easily obtained in a pure crystalline form reacted with magnesium ethyl bromide with formation of a small proportion of ct6-diphen$l-[$y-dibeib~yl-~y-~~rneth?llbutcc?2e (compare Spiith loc. cit .) : 2CnleBr(CH,Ph),$-2~~gEtBr. = (CH,Ph),~le.C~le(CH,Ph) + C4H,0 4- MgBr,. The main product however as indicated by aiialysis and by the evolution of ethane and ethylene was a mixture of ay-dipkenyl-p-methylpropane (I) (Graebe Ber. 1874 7 1627) and an olefiiie, CIoH1, believed to be ccy-dipheiiyl-p-methyl-An-propene (11) : (I). CHRle(CH,Ph) + C,H4 + MgBr FFtBr CMeBr(CH,Pli),~ls"t"f The constitution of this olefine could not be established by C,H + MgBr + CHPh:CMeCH2Ph (11.90 TROTMAN PREPARATION OF QUATERNARY HYDROCARBONS. oxidising it to a mixture of benzoic acid or benzaldehyde and benzyl methyl ketone with potassium permanganate or ozonised oxygen since only benzoic acid was obtained; this result was doubtless due to the fact that benzyl methyl ketone is itself very readily oxidised. The structure assigned to the olefine is rendered very probable by the known behaviour of other bromides similar to p- bromo- cr -phenyl- p-benzylpropane (compare Orechov and Konovalov Ber. 1912 45 861; Orechov and Meerson ibid., p. 1926). p-Bromo-a-phenyl- p-benzylpropane reacted with magnesium benzyl chloride with the formation of a very small quantity of the quaternary hydrocarbon a-phenyl- p p-dibenzylpropane.The main product however was an olefine doubtless identical with I1 : CMeBr(CH,Ph) + CH,Ph*MgCl = (11.) + MgClBr + C,H5*CH3. Unlike p- bromo- a-phenyl- p-methylbutane p- bromo- a-phenyl-p-benzylpropane did not react with magnesium even when iodine was added so that the quaternary hydrocarbon could not be prepared by modifying the procedure. In boiling ethereal or benzene solution the bromide reacted with sodium but the above-described olefine was formed. Tribenzylmethyl bromide (Schmerda Monatsh. 1909 30 387), treated with magnesium ethyl bromide gave a small yield of hexabenzylethane (Schmerda Zoc. cit.) 2(CH,Ph),CBr + 2MgEtBr = (CH,Ph),C*C(CH,Ph) + C,H, + BMgBr, but the main product was ocy-diphenyl- p-benzyl-Aa-propene (111) (CH,Ph),CBr + MgEtBr = (111.) CHPh:C(CH,Ph)2 + C,H6 + MgBr,.This olefine, which has been described by Orechov and Grinberg ( J . Russ. Chern. Phys. Xoc. 1916 48 1713) as a viscous liquid was obtained in well-formed crystals m. p. 42-43" Tribenzylmethyl bromide was also treated with magnesium benzyl chloride; it gave a small amount of tetrabenxylmethune but the product consisted mainly of 111. The action of zinc ethyl on this bromide was also in-vestigated but the above olefine was the only product. As there seemed to be indications that the halogen atom of tertiary halides is more readily displaced by a benzyl than by an ethyl group the action was investigated of magnesium benzyl chloride on p-brm-a-phenyl-p-methyzpropane. A quaternary hydrocarbon of the desired type viz.a-phenyl-p-benzyl-p-methyl-propane wits in fact produced but about 50% of the product con-sisted of a-phenyl-p-methyl-Aa-propene (Klages Ber. 1904,37,1721). The results of these experiments show that in many cases the reaction between Grignard reagents and tertiary halogen deriv-atives does not follow the normal course. Spath (Zoc. cit.) expresse TROTMAN PREPARATION OF QUATERNARY HYDROCARBONS. 91 the view that during this reaction the organic halide and the Grignard reagent undergo dissociation so giving rise to free organic radicals : It's + RlMgX = R' + R + MgX,. The existence of these radicals however is only momentary ; those which are electropositive tend t o unite together but those which are more electronegative show a tendency to form olefines and paraffins containing the same number of carbon atoms as the radicals.The basis of Spiith's classification into positive and negative radicals is not very obvious and his views do not seem to afford a very satisfactory explanation of the results summarised above. ,4 few attempts mere made to obtain dilietones of the type C,H,*CO*CR(CH,*C,H,).CO.CH in the hope that subsequent reduction might yield a quaternary hydrocarbon. BenxylbenxoyZ-acetone was prepared but on treatment with sodium ethoxide and ethyl bromide it gave w-benzylacetophenone an acetyl group being eliminated. E x P E R I M E N T A L. p-Bronzo-x-phenyl-$-methylbulane C,H,*CH,*CMeEtBr a colour-less oil with a pleasant minty odour and an extremely irritating action on the eyes was obtained in almost theoretical quantity by saturating benzylmethylethylcarbinol (Konovalov J .Rziss. Chem. Phys. SOC. 1904 36 228) with hydrogen bromide a t the ordinary temperature. It decomposed when distilled even under 11 mm. and was therefore purified by washing it with dilute sodium carbonate solution and water (Pound Br = 34.0. CIoH,,Br requires Br = 35-27'0). Action of Grignayd Beagents on p-Bromo-cc-phenyl-p-?netfLylbutane. -When an excess of magnesium ethyl bromide was added to an ethereal solution of the bromide a vigorous reaction immediately occurred magnesium bromide mas formed and ethane and ethylene were evolved. The main product was an oil practically the whole of which distilled between 195" and 205". Its behaviour towards bromine and hydrogen bromide and analyses of the brorno-deriv-atives thus obtained indicated that it was a mixture of an olefine, believed to be cc-phenyl-p-methyl-Aa-butene and a saturated hydro-carbon probably a-phenyl- p-methylbutane b.p. 203-2004' (Tafel and Hahl Ber. 1907 40 3313). Since from its boiling point, it could hardly contain an appreciable quantity of the expected quaternary hydrocarbon and analysis showed the presence of a small proportion of benzylmethylethylcarbinol the oil was not further examined 92 TROTMAN PREPARATION OF QUATERKARY HYDROCARBONS. A solution of the bromide in dry ether had no action on mag-nesium until a trace of iodine had been added; a vigorous reaction then set in and magnesium bromide separated even a t 0". The product was a mixture b.p. 195-205" apparently identical with that described above. DibenxyZmethylcarbinol (C6H5*CH,),CMe*OH prepared from magnesium benzyl chloride and ethyl acetate is a colourless, viscous sweet-smelling liquid b. p. 195-200"/14 mm. (with slight decomp.) (Pound C = 85.5; H = 8.0. Cl6Hl8O requires C = p -Bromo- cr-phen yl- p - benxylpropane (C,H,*CH,),CMeBr obtained in good yield by saturating an ethereal solution of the unpurified carbinol with hydrogen bromide and recrystallising the product from light petroleum and from alcohol formed well-defined, prismatic crystals m. p. 78.5". It loses hydrogen bromide when heated and cannot be distilled even under 0-2 mm. It is readily soluble in ether alcohol or benzene and sparingly so in light petroleum (Found Br = 27.6.Action of Magnesium Ethyl Bromide on p-Bromo-a-phengl-p- benzyl-jwopane.-A solution of the bromide in ether (1 vol.) was added slowly to magnesium ethyl bromide (4 vols.) at the ordinary temperature. The gas which was soon evolved after being freed from ether with the aid of sulphuric acid contained about 42% of ethane and 58% of ethylene. After 8 hours the reaction mkture was boiled for 2 hours. The product isolated in the usual way, was a slightly yellow oil. Its solution in alcohol (8 vols.) deposited crystals of cr8 -diphen$ p y -dibenx yl- p y dimethylbutane , (C,H,*CH,),CMe*CMe( CHz*CsH5)z, which separated from much light petroleum in colourless plates, m. p. 171"; the yield was about 0.5 g. from 40 g. of the bromide (Found C = 92.0; H = 8.3; M cryoscopic in benzene = 389.C,2Ha requires C = 91.9; H = 8.1y0; M = 418). Separation of ay-Diphenyl-p-methylpropane (I) and ( ?) ay-Di-IphenyZ-p-methyl-Aa-propene (II).-The main product obtained as an oil when the above alcoholic mother-liquor was evaporated, distilled between 290" and 300". It combined with bromine in chloroform solution but the dibromide did not crystallise and very readily lost hydrogen bromide. Oxidation or treatment with sulphuric acid having failed to separate the olefine from the saturated hydrocarbon the oil was saturated with hydrogen bromide when about 50% of it was converted into crystalline p-bromo-a-phenyl-P-benzylpropane. The unchanged oil was separated from the bromide and treated with sodium amalgam and aqueous alcohol to reduce any remaining bromide.The final 85.0; H = 8.0y0). CIsH1,Br requires Br = 27.7y0) TROTMAN PREPARATlON OF QUATERNARY HYDROCARBONS. 93 product b. p. 290-294" was ay-diphenyl-p-methylpropane (Found : C = 91.2; H = S.4. Cl,H18 requires C = 91.4; H = 8.6%). Thus the quaternary hydrocarbon a-phenyl- p-benzyl- p-methyl-butane (calc. C = 9O.S; H = 9.2%) was not among the products of the reaction. The oleJine (11) a colourless oil b. p. 294-896" was obtained by boiling the bromide with an excess of pyridine for 8 hours. It combined with bromine in chloroform solution giving a di-bromide which could not be crystallised and readily lost hydrogen bromide when it was heated (Found C = 92.2; H = 7.6. C,,H1, requires C = 92-3 ; H = 7.7%). Action of Magnesium Benxyl Chloride on p-Bromo- a-phenyl-P-benxy1propane.-The Grignard reagent (4 mols.) did not react with the bromide in ether a t the ordinary temperature.The ether was evaporated and the residue heated a t 100" for 2 hours. Tlie product isolated in the usual way was an oil a solution of which in alcohol deposited crystals of a-plie?iyl-p~-clibenxylpro~une, (C,H5*CH,),CMe ; these separated from alcohol in needle-shaped prisms m. p. 113" which were sparingly soluble in alcohol or light petroleum but dissolved more readily in other organic solvents (yield about 5%) (Found C = 91.5; H = 7.9; Af cryoscopic in benzene = 294. C,,H, requires C = 92.4 ; H = 8-Oyo ; M = 300). The main product obtained as an oil when the alcoholic mother-liquor was evaporated contained the olefine (II) but the probable presence of ay-diphenyl-P-methylpropane (I) could not be proved owing to the impossibility of separating the dibenzyl which the oil contained.Action of Mugnes ium Ethyl Bromide on Tribenzylmethyl Bromide. -Tribenzylmethyl bromide (Schmerda loc. cit.) which is most conveniently prepared by saturating an ethereal solution of tri-benzylcarbinol with hydrogen bromide was treated in benzene solution with an excess of ethereal magnesium ethyl bromide ; magnesium bromide separated at once and a gas containing oiilp a vcry small proportion of olefine was evolved. After 2 hours, the reaction mixture was gently boiled for 1 hour. The product, isolated in the usual way was a very viscous oil ; froin its solution in alcohol colourless cubical crystals gradually separated.This compound after recrystallisation from alcohol melted a t 42-43" and was doubtless ay-diphenyl-$-benzyl-4a-propene (111) (coni-pare Orechov and Grinberg loc. cit.). Its identity was established by converting it into the crystalline dibromide m. p. 127-128", and also into tribenzylmethyl bromide with the aid of hydrogen bromide. When the alcoholic mot her-liquor from the crystalline olefine was evaporated there remained an oil from which hesa 94 TROTMAN PREPARATION OF QUATERNARY HYDROCBRBONS. benzylethane was isolated by converting the olefine into tribenzyl-methyl bromide (see above) and subsequently extracting the product with alcohol. From the alcoholic extract the hydrocarbon was deposited in crystals m. p.82-83". Schmerda (loc. cit.) gives the m. p. 81-82" (Found C = 92.4; H = 797. Calc. for C44H42, Tetrabenzylmethane C( CH,*C,H,),.-Tribenzylmethy~ bromide, dissolved in benzene was added to an ethereal solution of excess of magnesium benzyl chloride the solvents were then evaporated, and the residue was heated at 100" for 2 hours. The product, isolated in the usual way was an oil from which tetrabenxyl-methane (yield 5%) was precipitated on the addition of alcohol. It separates from ether in cubic crystals m. p. 164' which are very sparingly soluble in alcohol or light petroleum but dissolve more easily in other organic solvents (Found C = 92.7; H = 7.5. C,,HZ8 requires C = 92.5; H = 7.5%). The main product was ay-diphenyl-p-benzyl-Aa-propene (111). p-Bromo-a-phenyl- p -methylpropane C,H,*C~*cMe,Br obtained by saturating dimethylbenzylcarbinol (Grignard Compt.rend., 1900 130 1324) with hydrogen bromide and washing the product with sodium carbonate solution and water is a colourless pleasant-smelling liquid which loses hydrogen bromide very readily when it is warmed and does not distil unchanged under 11 mm. (Found : Br = 36.4. CIoH,,Br requires Br = 376%). a-Phenyl-p-benzyl- p-methylpropane (CH,Ph),CMe,.-Reaction set in a t once when the preceding bromide (1 mol.) dissolved in ether was added to magnesium benzyl chloride (2 mols.); after 4 hours the reaction mixture was boiled gently for 2 hours. The product isolated in the usual way consisted of an oil which was separated by distillation into a-phenyl- p-methyl-A,-propene, b.p. 181" and a fraction of higher boiling point; the latter solidified when cold and crystallised from alcohol in prismatic needles m. p. 68-69" (Found C = 90.95; H = 9.0. C,,H,, requires C = 91.1 ; H = 8.9%). ~-Phenyl-p-benzyl-p-methylpropane boils a t 293-294" is readily soluble in most organic solvents, and has a somewhat sweet odour. BenxyZbenzoy1acetone.-The sodium derivative of benzoylacetone, prepared from its constituents in ether was isolated and boiled for 1 hour with an excess of benzyl chloride the solution was filtered and the unchanged benzyl chloride removed by distillation under reduced pressure. From an alcoholic solution of the residual dark brown oil benxylbenxoylacetone was deposited in clusters of needle-shaped crystals m.p. 60-61" (yield 50%). The compound is insoluble in aqueous potassium hydroxide and gives no coloration C = 9246; H = 7.4%) ~ ~ 0 0 ~ AND LILLEY TRANSFORMATION OF 1 ~ ETC. 95 with ferric chloride (Found C = 80-6 ; H = 6.1. C1,Hl,O, requires C = 80.95; H = 6.35%). When an alcoholic solution of benzylbenzoylacetone containing sodium ethoxide and ethyl bromide is kept at the ordinary temperature for 24 hours sodium bromide separates but the product is mainly w-benzylacetophenone. The author desires t o express his thanks to Professor F. S. Kipping F.R.S. for suggesting this research and for his interest in its progress. UNIVERSITY COLLEGE NOTTIKCHAM. [Received August 29th 1924. 88 TROTMAN PREPARATION OF QUATERNARY HYDROCARBONS. XV.-Preparation of Quuternary Hydrocarbons.By EDWARD RUSSELL TROTMAN. IN the course of some investigations in progress in this laboratory, certain quaternary hydrocarbons of the type (C,M,*CH,),CR,R~, were required R and R being different alkyl radicals. No such compounds are known and indeed very few quaternary hydro-carbons of any kind have been described. The interaction of Grjgnard reagents and tertiary halogen derivatives which seeme TROTMAN PREPARATION OF QUATERNARY HYDROCARBONS. 89 to offer a suitable method €or their preparation has already been studied. Gomberg and Cone (Ber. 1906 39 1461 2975) succeeded in displacing the chlorii-ie atom of triphenylmethyl chloride by various aromatic and aliphatic hydrocarbon radicals in this way. On the other hand Davis arid Kipping (J.1911 99 300) were unable to prepare yy-dibenzyl-n-hexane (C,H,*CH,),CEtPr by the action of magnesium propyl bromide on cr-bromo- rcc-dibenzyl-propane (C,K5*CH,),CBrEt ; in every case the product consisted mainly if not entirely of an unsaturated hydrocarbcn. Spiith (Monatsh. 1913 34 1963) who made an exhaustive study of the act ion of Grigiiard reagents on organic halogen derivatives found that although quaternary hydrocarbons were generally produced, tlie yield was always poor and a mixture of numerous products was obtained. It appeared therefore that the interaction of a Grigiiard reagent ancl a tertiary halide often takes quite an unexpected course and this conclusioii is confirmed by the results described below. A preliminary study of tlie proposed method was made with [$-byorno- x-pher,yl-P-meth?llbutane a compound readily prepred from imthy! ethyl ketone.This bromide reacted with magnesium ethyl bromide 153th evolution of ethane ancl ethylene but the product appeared to consist entirely of a mixturz of the olefine, C,H,*CH:CMeEt and the saturated conipound C,H5*CH,.CH3kEt , a i d the possible presence of a small proportion of the quaternary hydrocarbon could not bt proved. Attempts to preparc a quaternary hydrocarbon by first converting the bromide into a Grignard reagent and then treating the product with ethyl bromide were also unsuccessful ; although the bromide reacted with mag-iiesiuin in thc presence of dry ether magnesium bromide separated even at 0" and a mixture of the saturated hydrocarbon and the olcfine was formed.~-B~o~?~o-sr-p7~enyl-~-benx~lp~o~ane which was easily obtained in a pure crystalline form reacted with magnesium ethyl bromide with formation of a small proportion of ct6-diphen$l-[$y-dibeib~yl-~y-~~rneth?llbutcc?2e (compare Spiith loc. cit .) : 2CnleBr(CH,Ph),$-2~~gEtBr. = (CH,Ph),~le.C~le(CH,Ph) + C4H,0 4- MgBr,. The main product however as indicated by aiialysis and by the evolution of ethane and ethylene was a mixture of ay-dipkenyl-p-methylpropane (I) (Graebe Ber. 1874 7 1627) and an olefiiie, CIoH1, believed to be ccy-dipheiiyl-p-methyl-An-propene (11) : (I). CHRle(CH,Ph) + C,H4 + MgBr FFtBr CMeBr(CH,Pli),~ls"t"f The constitution of this olefine could not be established by C,H + MgBr + CHPh:CMeCH2Ph (11. 90 TROTMAN PREPARATION OF QUATERNARY HYDROCARBONS.oxidising it to a mixture of benzoic acid or benzaldehyde and benzyl methyl ketone with potassium permanganate or ozonised oxygen since only benzoic acid was obtained; this result was doubtless due to the fact that benzyl methyl ketone is itself very readily oxidised. The structure assigned to the olefine is rendered very probable by the known behaviour of other bromides similar to p- bromo- cr -phenyl- p-benzylpropane (compare Orechov and Konovalov Ber. 1912 45 861; Orechov and Meerson ibid., p. 1926). p-Bromo-a-phenyl- p-benzylpropane reacted with magnesium benzyl chloride with the formation of a very small quantity of the quaternary hydrocarbon a-phenyl- p p-dibenzylpropane. The main product however was an olefine doubtless identical with I1 : CMeBr(CH,Ph) + CH,Ph*MgCl = (11.) + MgClBr + C,H5*CH3.Unlike p- bromo- a-phenyl- p-methylbutane p- bromo- a-phenyl-p-benzylpropane did not react with magnesium even when iodine was added so that the quaternary hydrocarbon could not be prepared by modifying the procedure. In boiling ethereal or benzene solution the bromide reacted with sodium but the above-described olefine was formed. Tribenzylmethyl bromide (Schmerda Monatsh. 1909 30 387), treated with magnesium ethyl bromide gave a small yield of hexabenzylethane (Schmerda Zoc. cit.) 2(CH,Ph),CBr + 2MgEtBr = (CH,Ph),C*C(CH,Ph) + C,H, + BMgBr, but the main product was ocy-diphenyl- p-benzyl-Aa-propene (111) (CH,Ph),CBr + MgEtBr = (111.) CHPh:C(CH,Ph)2 + C,H6 + MgBr,. This olefine, which has been described by Orechov and Grinberg ( J .Russ. Chern. Phys. Xoc. 1916 48 1713) as a viscous liquid was obtained in well-formed crystals m. p. 42-43" Tribenzylmethyl bromide was also treated with magnesium benzyl chloride; it gave a small amount of tetrabenxylmethune but the product consisted mainly of 111. The action of zinc ethyl on this bromide was also in-vestigated but the above olefine was the only product. As there seemed to be indications that the halogen atom of tertiary halides is more readily displaced by a benzyl than by an ethyl group the action was investigated of magnesium benzyl chloride on p-brm-a-phenyl-p-methyzpropane. A quaternary hydrocarbon of the desired type viz. a-phenyl-p-benzyl-p-methyl-propane wits in fact produced but about 50% of the product con-sisted of a-phenyl-p-methyl-Aa-propene (Klages Ber.1904,37,1721). The results of these experiments show that in many cases the reaction between Grignard reagents and tertiary halogen deriv-atives does not follow the normal course. Spath (Zoc. cit.) expresse TROTMAN PREPARATION OF QUATERNARY HYDROCARBONS. 91 the view that during this reaction the organic halide and the Grignard reagent undergo dissociation so giving rise to free organic radicals : It's + RlMgX = R' + R + MgX,. The existence of these radicals however is only momentary ; those which are electropositive tend t o unite together but those which are more electronegative show a tendency to form olefines and paraffins containing the same number of carbon atoms as the radicals.The basis of Spiith's classification into positive and negative radicals is not very obvious and his views do not seem to afford a very satisfactory explanation of the results summarised above. ,4 few attempts mere made to obtain dilietones of the type C,H,*CO*CR(CH,*C,H,).CO.CH in the hope that subsequent reduction might yield a quaternary hydrocarbon. BenxylbenxoyZ-acetone was prepared but on treatment with sodium ethoxide and ethyl bromide it gave w-benzylacetophenone an acetyl group being eliminated. E x P E R I M E N T A L. p-Bronzo-x-phenyl-$-methylbulane C,H,*CH,*CMeEtBr a colour-less oil with a pleasant minty odour and an extremely irritating action on the eyes was obtained in almost theoretical quantity by saturating benzylmethylethylcarbinol (Konovalov J .Rziss. Chem. Phys. SOC. 1904 36 228) with hydrogen bromide a t the ordinary temperature. It decomposed when distilled even under 11 mm. and was therefore purified by washing it with dilute sodium carbonate solution and water (Pound Br = 34.0. CIoH,,Br requires Br = 35-27'0). Action of Grignayd Beagents on p-Bromo-cc-phenyl-p-?netfLylbutane. -When an excess of magnesium ethyl bromide was added to an ethereal solution of the bromide a vigorous reaction immediately occurred magnesium bromide mas formed and ethane and ethylene were evolved. The main product was an oil practically the whole of which distilled between 195" and 205". Its behaviour towards bromine and hydrogen bromide and analyses of the brorno-deriv-atives thus obtained indicated that it was a mixture of an olefine, believed to be cc-phenyl-p-methyl-Aa-butene and a saturated hydro-carbon probably a-phenyl- p-methylbutane b.p. 203-2004' (Tafel and Hahl Ber. 1907 40 3313). Since from its boiling point, it could hardly contain an appreciable quantity of the expected quaternary hydrocarbon and analysis showed the presence of a small proportion of benzylmethylethylcarbinol the oil was not further examined 92 TROTMAN PREPARATION OF QUATERKARY HYDROCARBONS. A solution of the bromide in dry ether had no action on mag-nesium until a trace of iodine had been added; a vigorous reaction then set in and magnesium bromide separated even a t 0". The product was a mixture b. p. 195-205" apparently identical with that described above.DibenxyZmethylcarbinol (C6H5*CH,),CMe*OH prepared from magnesium benzyl chloride and ethyl acetate is a colourless, viscous sweet-smelling liquid b. p. 195-200"/14 mm. (with slight decomp.) (Pound C = 85.5; H = 8.0. Cl6Hl8O requires C = p -Bromo- cr-phen yl- p - benxylpropane (C,H,*CH,),CMeBr obtained in good yield by saturating an ethereal solution of the unpurified carbinol with hydrogen bromide and recrystallising the product from light petroleum and from alcohol formed well-defined, prismatic crystals m. p. 78.5". It loses hydrogen bromide when heated and cannot be distilled even under 0-2 mm. It is readily soluble in ether alcohol or benzene and sparingly so in light petroleum (Found Br = 27.6. Action of Magnesium Ethyl Bromide on p-Bromo-a-phengl-p- benzyl-jwopane.-A solution of the bromide in ether (1 vol.) was added slowly to magnesium ethyl bromide (4 vols.) at the ordinary temperature.The gas which was soon evolved after being freed from ether with the aid of sulphuric acid contained about 42% of ethane and 58% of ethylene. After 8 hours the reaction mkture was boiled for 2 hours. The product isolated in the usual way, was a slightly yellow oil. Its solution in alcohol (8 vols.) deposited crystals of cr8 -diphen$ p y -dibenx yl- p y dimethylbutane , (C,H,*CH,),CMe*CMe( CHz*CsH5)z, which separated from much light petroleum in colourless plates, m. p. 171"; the yield was about 0.5 g. from 40 g. of the bromide (Found C = 92.0; H = 8.3; M cryoscopic in benzene = 389. C,2Ha requires C = 91.9; H = 8.1y0; M = 418).Separation of ay-Diphenyl-p-methylpropane (I) and ( ?) ay-Di-IphenyZ-p-methyl-Aa-propene (II).-The main product obtained as an oil when the above alcoholic mother-liquor was evaporated, distilled between 290" and 300". It combined with bromine in chloroform solution but the dibromide did not crystallise and very readily lost hydrogen bromide. Oxidation or treatment with sulphuric acid having failed to separate the olefine from the saturated hydrocarbon the oil was saturated with hydrogen bromide when about 50% of it was converted into crystalline p-bromo-a-phenyl-P-benzylpropane. The unchanged oil was separated from the bromide and treated with sodium amalgam and aqueous alcohol to reduce any remaining bromide. The final 85.0; H = 8.0y0). CIsH1,Br requires Br = 27.7y0) TROTMAN PREPARATlON OF QUATERNARY HYDROCARBONS.93 product b. p. 290-294" was ay-diphenyl-p-methylpropane (Found : C = 91.2; H = S.4. Cl,H18 requires C = 91.4; H = 8.6%). Thus the quaternary hydrocarbon a-phenyl- p-benzyl- p-methyl-butane (calc. C = 9O.S; H = 9.2%) was not among the products of the reaction. The oleJine (11) a colourless oil b. p. 294-896" was obtained by boiling the bromide with an excess of pyridine for 8 hours. It combined with bromine in chloroform solution giving a di-bromide which could not be crystallised and readily lost hydrogen bromide when it was heated (Found C = 92.2; H = 7.6. C,,H1, requires C = 92-3 ; H = 7.7%). Action of Magnesium Benxyl Chloride on p-Bromo- a-phenyl-P-benxy1propane.-The Grignard reagent (4 mols.) did not react with the bromide in ether a t the ordinary temperature.The ether was evaporated and the residue heated a t 100" for 2 hours. Tlie product isolated in the usual way was an oil a solution of which in alcohol deposited crystals of a-plie?iyl-p~-clibenxylpro~une, (C,H5*CH,),CMe ; these separated from alcohol in needle-shaped prisms m. p. 113" which were sparingly soluble in alcohol or light petroleum but dissolved more readily in other organic solvents (yield about 5%) (Found C = 91.5; H = 7.9; Af cryoscopic in benzene = 294. C,,H, requires C = 92.4 ; H = 8-Oyo ; M = 300). The main product obtained as an oil when the alcoholic mother-liquor was evaporated contained the olefine (II) but the probable presence of ay-diphenyl-P-methylpropane (I) could not be proved owing to the impossibility of separating the dibenzyl which the oil contained.Action of Mugnes ium Ethyl Bromide on Tribenzylmethyl Bromide. -Tribenzylmethyl bromide (Schmerda loc. cit.) which is most conveniently prepared by saturating an ethereal solution of tri-benzylcarbinol with hydrogen bromide was treated in benzene solution with an excess of ethereal magnesium ethyl bromide ; magnesium bromide separated at once and a gas containing oiilp a vcry small proportion of olefine was evolved. After 2 hours, the reaction mixture was gently boiled for 1 hour. The product, isolated in the usual way was a very viscous oil ; froin its solution in alcohol colourless cubical crystals gradually separated. This compound after recrystallisation from alcohol melted a t 42-43" and was doubtless ay-diphenyl-$-benzyl-4a-propene (111) (coni-pare Orechov and Grinberg loc.cit.). Its identity was established by converting it into the crystalline dibromide m. p. 127-128", and also into tribenzylmethyl bromide with the aid of hydrogen bromide. When the alcoholic mot her-liquor from the crystalline olefine was evaporated there remained an oil from which hesa 94 TROTMAN PREPARATION OF QUATERNARY HYDROCBRBONS. benzylethane was isolated by converting the olefine into tribenzyl-methyl bromide (see above) and subsequently extracting the product with alcohol. From the alcoholic extract the hydrocarbon was deposited in crystals m. p. 82-83". Schmerda (loc. cit.) gives the m. p. 81-82" (Found C = 92.4; H = 797.Calc. for C44H42, Tetrabenzylmethane C( CH,*C,H,),.-Tribenzylmethy~ bromide, dissolved in benzene was added to an ethereal solution of excess of magnesium benzyl chloride the solvents were then evaporated, and the residue was heated at 100" for 2 hours. The product, isolated in the usual way was an oil from which tetrabenxyl-methane (yield 5%) was precipitated on the addition of alcohol. It separates from ether in cubic crystals m. p. 164' which are very sparingly soluble in alcohol or light petroleum but dissolve more easily in other organic solvents (Found C = 92.7; H = 7.5. C,,HZ8 requires C = 92.5; H = 7.5%). The main product was ay-diphenyl-p-benzyl-Aa-propene (111). p-Bromo-a-phenyl- p -methylpropane C,H,*C~*cMe,Br obtained by saturating dimethylbenzylcarbinol (Grignard Compt.rend., 1900 130 1324) with hydrogen bromide and washing the product with sodium carbonate solution and water is a colourless pleasant-smelling liquid which loses hydrogen bromide very readily when it is warmed and does not distil unchanged under 11 mm. (Found : Br = 36.4. CIoH,,Br requires Br = 376%). a-Phenyl-p-benzyl- p-methylpropane (CH,Ph),CMe,.-Reaction set in a t once when the preceding bromide (1 mol.) dissolved in ether was added to magnesium benzyl chloride (2 mols.); after 4 hours the reaction mixture was boiled gently for 2 hours. The product isolated in the usual way consisted of an oil which was separated by distillation into a-phenyl- p-methyl-A,-propene, b. p. 181" and a fraction of higher boiling point; the latter solidified when cold and crystallised from alcohol in prismatic needles m.p. 68-69" (Found C = 90.95; H = 9.0. C,,H,, requires C = 91.1 ; H = 8.9%). ~-Phenyl-p-benzyl-p-methylpropane boils a t 293-294" is readily soluble in most organic solvents, and has a somewhat sweet odour. BenxyZbenzoy1acetone.-The sodium derivative of benzoylacetone, prepared from its constituents in ether was isolated and boiled for 1 hour with an excess of benzyl chloride the solution was filtered and the unchanged benzyl chloride removed by distillation under reduced pressure. From an alcoholic solution of the residual dark brown oil benxylbenxoylacetone was deposited in clusters of needle-shaped crystals m. p. 60-61" (yield 50%). The compound is insoluble in aqueous potassium hydroxide and gives no coloration C = 9246; H = 7.4%) ~ ~ 0 0 ~ AND LILLEY TRANSFORMATION OF 1 ~ ETC. 95 with ferric chloride (Found C = 80-6 ; H = 6.1. C1,Hl,O, requires C = 80.95; H = 6.35%). When an alcoholic solution of benzylbenzoylacetone containing sodium ethoxide and ethyl bromide is kept at the ordinary temperature for 24 hours sodium bromide separates but the product is mainly w-benzylacetophenone. The author desires t o express his thanks to Professor F. S. Kipping F.R.S. for suggesting this research and for his interest in its progress. UNIVERSITY COLLEGE NOTTIKCHAM. [Received August 29th 1924.