1404 KING THE PRODUCTION OF CXXXI1.-The Production of Methy2 Ethyl Ketone from n-Butyl Alcohol. By ALBF~RT THEODORE KING. IN this conversion the usual series of steps has been employed of dehydration re-hydration of the resulting alkylene to the secondary alcohol and finally dehydrogenation of the latter to give the corresponding ketone : R*CH,*CH,*OH + R*CH,:CH -t R*CH(OH)*CH -+ R*CO*C€13. No details appear to have been published hitherto regarding the hydration of butylene to tlie secondary alcohol and although the first and last of the above stages have in this particular case been elsewhere described the comparative results now obtained seem worthy of being placed on record also. Dehydration of n-Bzctyl AZcolzoL-Each of the three possible butylenes, CH,- CH,*CH:CH CH,-CH:CH*CH (CH3)?WH2, a-.13-. Y-. has been identified by previous observers in the product of dehydr-ation the composition of which would appear to depend both on the nature of the dehydrator and on the temperature employed. Thus Ipatiev ( J . Russ. Phys. Chem. SOC. 1903 35 577) wit METHYL ETHYL B ~ O N E FBOM N-BUTYL ALCOHOL. 1406 alumina as catalyst a t 500° obtained a gas containing 25-30 per cent. of a-butylene the rest being y-butylene; the latter was attributed to the presence of isobutyl.alcoho1 in the material employed. Senderens (Ann. Chim. Phys. 1912 [viii] 25 449), using aluminium sulphate a t 300° obtained a-butylene with 27 per cent. of y-butylene. Le Be1 and Green obtained on dehydration with zinc chloride a product free from y-butylene and containing 80 per cent.of P-butylene and 10 per cent. of a-butylene (Bull. Soc. chinz. 1881 [ii] 35 438). I n the present investigation in which phosphoric acid on pumice was used only slight differences were observed in the composition of the product a t temperatures ranging from 280° to 400O. No y-butylene was detected and even a t 280" the amount of a-butylene present so far as was indicated by fractionation of the bromide could only be slight. This method of dehydration there-fore yielding reasonably pure B-butylene gives a much more homogeneous product than those previously described. Hydration of ButyZenes.-The hydration of the three butylenes with sulphuric acid should theoretically proceed as formulated below : CH,* CH,-CH( OH)* CH \ CH,*CH,.CH:CH, CH,*CH:CH*CH f (CH,),C:CH -+ (CH,),C*OH This conversion in the case of y-butylene has been described by Butlerov (AmaZen 1875 180 245) who also records the very slow hydration of /3-butylene but without details of the isolation of sec.-butyl alcohol.Nor is the formation of the latter from a-butylene recorded; and this point is under examination the present investigation dealing only with the behaviour of 8-butylene. The rapid absorption of @-butylene (as also of u- and y-butylene) by concentrated sulphuric acid is well known but by this action polymerisation and not hydration results. With diluted acid, the very slow action on keeping has been confirmed but liquefied 8-butylene under its own pressure agitated in contact with the acid has been found to undergo absorption with ease.Polymerisa-tion which is marked a t the higher concentrations is very slight with 78 per cent. acid and below this strength the reaction proceeds normally. In one case an absorption was observed of nearly 20 per cent. in excess of the butylene theoretically required to convert the acid t o butylsulphuric acid suggesting that the latter is capable of 3 a 1406 KING THE PRODUGTIOTT OF catalysing the direct hydration of 8-butylene. It is hoped to investigate this point further. The diluted solution on distillation readily gives the secondary alcohol in good yield. Dehydrogenation of sec.-Butyl A 1cohol.-Reduced copper is recorded by Sabatier and Sender ens as being particularly efficient in catalysing the dehydrogenation of secondary alcohols these showing a conversion without complications but incomplete owing to the reverse action also taking place in the presence of copper.They observed the formation from sec.-butyl alcohol of methyl ethyl ketone and pure hydrogen without any accessory reaction, within the temperature range of 160-300° (Ann. Chim. Yhys., 1905 [viii] 4 433 465). In the present work the reverse action above referred to was perceptible but only slight. EXPERIMENTAL. Preparation of P-Butylene. The +butyl alcohol distilled a t 116-118O. It was gently boiled in a silica distilling flask attached t o a 2.2 cm. copper tube 150 cm. in length capable of giving by electrical heating temperatures, registered a t the middle part from 280° to 400O. This was packed with fragments of ignited pumice impregnated with glacial phos-phoric acid.(This catalyst can be used indefinitely its activity being renewed occasionally by burning a piece of phosphorus a t the mouth of the tube with a current of air passing through.) The unchanged alcohol and water formed were collected in a receiver with water-cooled reflux and with a syphon tube for the disuharge of the condensate from time to time. The gaseous pro-duct after scrubbing with 60 per cent. sulphuric acid to retain any y-butylene was absorbed in cooled bromine until the latter was nearly colourless. The product was then shaken with dilute alkali dried with calcium chloride and systematically fractionated. The following are typical results: (1). (2)- (3). Temperature ..................300-320". 280-300". 400". Dried product .................. 75 grams. 114 grams. 120 grams. Find fractions :-below 155" 153-157'. 3.5 9 9 2 Y Y 9.0 Y9 8 Y 9 155-157' -157-159 60 99 92 99 98 9Y 164-161 3 9 9 3 9 9 6 99 residue 3 9s 3.5 I# 4 9 s ( 7 grams METHYL ETI~YL KETONE FROM N-BUTYL ALCOHOL. 1M7 The boiling points of the three dibromides are a- 1 6 6 O ; & cis-form 158O; tram- 161O; y- 1 4 9 ~ 6 ~ . Thus a-butylene is not present even at the lowest temperature employed in notable quantity. The low fraction possibly indicating the presence of y-butylene was not appreciably increased in (2) in which waahing with 60 per cent. acid was omitted and these lower fractions were not in sufficient quantity to detect by further fractionation the presence of y-butylene dibromide.Moreover no tert .-butyl alcohol was obtained from the 60 per cent. acid used to wash more than 100 grams of gas. On dilution and distillation it yielded only 3 grams of product which proved t o be mainly n-butyl alcohol carried over. Much of the butylene formed dissolves in the butyl alcohol which escapes dehydration and condenses in the receiver. From this solution more than 150 times its volume of butylene is expelled on raising to boiling point. The un-changed alcohol layer dried and fractionated yielded finally, 1-2 per cent. distilling below 1 1 6 O the rest passing over a t 116-118O and for continuous working the condensate was shaken with salt and the upper layer separated and used again thus conserving the dissolved butylene .No side-reaction was observed in the dehydration. Action of Sulphuric Acid on /3-Butylene. On bubbling the gas through concentrated sulphuric acid rapid absorption took place the liquid becoming warm and darkening in colour. A pale yellow upper layer formed and when this had increased to about 10 C.C. in volume it wm separated washed and dried with calcium chloride. It had a slight odour of hydrocarbon, and on distilling boiled without any constancy from 140° to 300O. The sulphuric acid layer gave on pouring into excess of water a turbid liquid from which a small quantity of oil separated which gradually darkened and became very viscous. These producta, presumably formed by polymerisation of the 6-butylene were not further investigated.No secondary alcohol was detected in the aquous portion. With diluted acid even up to 78 per cent. no appreciable absorption took place when the gas was merely bubbled through. For absorption under pressure the following procedure was used. The gas was liquefied by passing through a worm cooled in a mixture of ice and salt the outlet tube dipping under the acid contained in a stoubwalled bottle also cooled in ice and salt. This gave good condensation bubbles of gas escaping only a t rar 1408 KINU THE PRODUOTION OF intervals. The bottle was then closed by a rubber stopper wired down and after weighing to ascertain approximately the amount of butylene condensed shaken in a mechanical shaker until the butylene layer had disappeared. I n each case 50 C.C.of pure acid were used with varying dilution and the time occupied for absorp-tion was roughly noted with the following results: Concentration. Per cent. 84-0 78.0 7 6.0 76-3 72.5 70.0 64.8 50.0 100 Butylene. Grams. 17 21 23 22 22 19 24 10 22 Time taken. 5 minutes -20 9 9 40 y y 1 hour 1.6 hours 5 9 ) 8 Little ' k e c t after 15 hours. With pure acid the absorption was complete in a few seconds. On pouring into ice an oil separated which after drying with calcium chloride weighed 15 grams. It distilled from 140° to above 300° without boiling constantly a t any point like the pro-duct previously described. No secondary alcohol was present in the aqueous poition. With 84 per cent. acid the alcohol was the main product but considerable polymerisation occurred.Moreover the acid crystal-lised in the freezing mixture causing inconvenience through block-ing of the delivery tube. The freezing-point curve shows the range of concentration having sufficiently low freezing point and also giving normal reaction in reasonable time to be from 74.5 (f. p. - 20°) to 78 per cent. (f. p. -ZOO). The concentration ultimately adopted was 75 per cent. With 78 per cent. acid a trace of oil with a terpene-like odour was observable on dilution this becoming imperceptible with the lower concentrations. The rise of temperature was very marked with the higher con-centrations. The pressures generated by the liquid butylene were approximately determined by a pressure gauge as follows : 1 2 O 0.6 atm.20 0.95 ,, 30 1.4 19 40 1.9 1 , With 75 per cent. acid the temperature did not reach 40° and the operation is thus without risk. To ascertain the extent of absorption 76 per cent. acid contain-ing 92 grams of pure acid was shaken with 63 grams of- butylene (theoretical quantity= 51.5) for fifteen hours. The excess of gas was allowed to escape and any in physical solution removed b METHYL ETHYL KETONE FROM N-BUTYL ALCOHOL. 1409 exhaustion when 61 grams were found to have been chemically absorbed. About 90 per cent. of the theoretical quantity was, however usually employed. Hydrolysis of Butylsulphwric Acid. Comparative experiments were made using three portions of 50 grams from the same batch of butylsulphuric acid. These were respectively neutralised with sodium carbonate and treated with quantities of 50 and 25 grams of water.The liquids were dis-tilled and the distillates caught in graduated tubes and thoroughly shaken with salt solution and a slight excess of salt. The volumes of upper layer were 23 c.c. 23 c.c. and 22.5 C.C. respectively. With neutralisation as employed by Butlerov (Zoc. cit.) the alcohol distilled over very slowly with a large quantity of water. Dis-tillation with the lesser quantity of water gave a distillate with only a small aqueous layer the bulk being retained by the acid, but with a slightly lower yield although no ether was detected in the product. Dilution with an equal weight of water was therefore adopted. With this procedure 51 grams of butylene yielded a distillate which after salting out and drying with potassium carbonate, weighed 60 grams that is 89 per cent.of the theoretical on the crude material. From 150 grams of dried undistilled product, after careful fractionation were obtained as final fractions : below 97.5' 3 grams 131 Y9 residue 5 9 9 The fraction 97.5-99.5O was employed for conversion to ketone. ,9;:E;;F 8 Y Y Dehydrogenation of sec.-Butyt A Icohol. The procedure followed was substantially that of Sabatier and Senderens (loc. cit.). A copper tube 172 cm. long and 1.2 cm. in diameter was packed with copper oxide from wire kept in place with plugs of rolled copper gauze and enclosed in a 2.5 cm. copper tube 135 cm. long wired for electrical heating. After reduction of the wire with ethyl alcohol vapour and expulsion of the latter, sec.-butyl alcohol was distilled through from a silica flask a t such a rate that the distillate passed through a t the rate of about one drop a second.This was caught in a receiver cooled in ice-water. The hydrogen was evolved just too fast to allow the bubbles through a wash-bottle to be counted. The temperature remaine 1410 ARMSTRONG AND HILDITCH CONVERSION OF THB 8IMPLE fairly constant throughout the fluctuation on either aide of 290° being never more than 5O. In the first preparation from 100 grams of alcohol 104 of distillate were obtained. This contained water formed from some unreduced copper oxide and on fractionating 94 grams distilled a t 73-75O the mixture separating a small lower layer of water. After drying with potassium carbonate the bulk distilled a t Two hundred and fifty grams of alcohol gave an almost 79-81'. theoretical yield of following fractions : dried undistilled product which gave the 76-79' 5 grams 79-81 199 9 9 81-82 9 9 9 82-86 6 9 9 residue 1.2 Y, Thus the conversion is upwards of 90 per cent. 'The fraction boiling a t 79-81O decolorised permanganate but after stirring with about half a gram of powdered potassium per-rnanganate for a short time the colour persisted. On distilling off and drying thoroughly with potassium carbonate (the boiling point is very easily affected by traces of water) the ketone distilled almost completely within the range 80-80.6°. The yield on the small scale from the crude butylene is about 70 per cent. of the theoretical and with continuous working and recoveries this figure a t least should be reached in the complete synthesis from n-butyl alcohol. IMPERIAL COLLEGE OF SCIENCE AND TECHNOLOQS, SOUTH KENSINGTON. [Received h'owrnB( r 18th 1 91 9.