236 DOBBIN ON SOME REACTIONS OF XIX.-On some Reuctions of Tertiary Rutyl Iodide. By LEONARD DOBBIN. (Frankland Prize of the Institute of Chemistry.) THE research the report of which is contained in the following pages, was carried out in the University Laboratory Wurzburg under the direction of Professor Wislicenus and it comprises the investigation of three reactions of tertiary butyl iodide. Preparutiort of the Tertiary B d y l Iodide. The starting material is primary isobutyl alcohol which is converted into the corresponding primary iodide by the usual method for prepa-ration of iodides of the alcohol radicals by means of amorphous phos-phorus and iodine. On rectification the portion of the distillate col-lected for several degrees below and above the boiling point of the pure primary iodide (120.5") was employed for the further reactions to obtain the tertiary iodide.The first of these consisted in the pre-paration of isobntylene which was effected by decomposing the iodide with potassium hydrat'e. A large flask containing a concentrated alcoholic solution of potas-sium hydrate into which the iodide could be allowed to flow from a dropping-funnel was connected with an inverted condenser and from the end of the condenser a tube led to a gas-holder. The potassium hydrate solution was first heated on a water-bath until the alcoho TERTIARY BUTYL IODIDE. 2 37 began to boil when the iodide was allowed to flow in drop by drop and as the latter was decomposed the isobutylene formed passed into the gas-holder. In this way isobutylene could b e rapidly prepared, hut the product always fell considerably short of the calculaled quan-tity.Some isobutylene was also prepared by heating slightly diluted sulphuric acid with isobutyl alcohol in presence of sand or powdered talc but the quantity of gas obt'ained in this way was only a small proportion of the calculated quantity as large. quantities of polymerised hydrocarbons were always formed. The.method was given up as it was not so satisfactory as that wit,h the iodide. The collected isobu-tylene was then passed into a saturaked strongly fuming solution of hydriodic acid. The latter solution was placed in a bottle which was provided with an india-rubber stopper through which passed a single tube reaching down to the liquid. The air was first driven out of the bottle by allowing a rapid stream of isobutylene to pass through for a few seconds the stopper was then firmly placed in the bottle when absorption assisted by vigorous fihaking and cooling with ice took place very rapidly tertiary isobutyl iodide being formed.Equations will render the reactions clear :-(CH,),CH.CH;I + KOH = K'I + EZO + (CH,)2C=CH,. Primary isobutjlic Isobutylene. iodide. then :-(CH,)QC=CH + HI = (CB-,),CI.CH,. Tertiary butylic iodide. The product thus obtained was separated from the aqueous hydriodic acid solution treated with dilute potassium hydrate solution to remove hydriodic acid and' free iodine carefully dried with calcium and filtered. It was then in the condition in which it was wed for the subsequent experiments.I. D E C O M P O S I T I O N WITH V A T E R : I n the course of some experiments made in the endeavour to dis-cover it good method for the preparation of tertiary butyl cyanide, ( CH&,C(CN).CH3 a quantity of tertiary hutyl iodide was shaken up with excess of a 12 per cent. solution of hydrocyanic acid. It was found that on prolonged shaking the iodide was gradually dissolved in the other liquid until a t length a quite homogeneoss liquid was ob-tained. Excess of zinc oxide was then added to combine with the hydriodic acid and excess of hydrocyania acid present. Zinc iodide and cyanide were formed and the liquid was then slxbjected to distil-lation. It. began to boil at 63" between which point and 100" a con-VUL. XXXFII. 238 DOBBIN ON SOME REACTIONS OF siderable portion of it passed over.At 100" scarcely anything but water distilled The distillate which was a slightly yellow thickish liquid was clearly a very different body from the cjanide sought for, as it was readily soluble in water which the cyanide is not and the latiter only begins to boil at about 105'. On placing it in a mixture of snow and salt it solidified at about -7" in star-like crystalline masses, and its properties in general led to the idea that it might be the alcohol corresponding to the tertiary iodide trimethylcarbinol or rather the molecular compound which this body forms with water, Acting on this idea the liquid was dried partially by adding dried potassium carbonate and then it was completely freed from water by repeated distillation with barium oxide.ARer this treatment the body exhibited all the properties of trimethylcarbinol. It solidified in characteristic semi-transparent crystalline ma,sseB whioh melted at 25.5" and the liquid boiled a t 82-82*5". 2[(CH3),C(OH).C%] + H@. Combustions gave the following numbers :-I 10857 gram substance gave *lo74 gram OHs and -2026 gram COZ. IT. *I502 gram substance gave el596 gram OH and -3097 gram coz. Pound. c A " , Caloulated for C,H,OH. I. 11. C . . . . . . 64.86 per cent. 64-47 64-87 H . . 13.51 , 13-92 13.62 0 ,. 21.63 , -As it was believed that in this reaction the hydrocyanic acid did not play any part a second experiment was made in order to ascertain whether such was really the case. 20 grams tertiary isobutyl iodide were vigorously shaken with 50 grams distilled water On leaving the liquid for two days with occasional shaking it was fouud that the iodide had all dissolved in the water and that the liquid which at first had been very slightly acid now gave a very marked acid reaction with blue litmus paper.Excess of sodium carbonate was then added, when carbon dioxide was evolved in considerable quantity ; potassium hydrate solution was added till the coloration due to iodine had dis-appeared ; then after nearly neutralising with hydrochloric acid the liquid was distilled. It began to boil at 63" the temperature rose rapidly to 84" where it was constmh for some time and then rose gradually to 100". The distillate below 100" was as before partially dried with potassium carbonate and then completely dried by re-peated distillation from barium oxide.Determination of melting point gave 25-25-5' that of boiling point 82.5' TERTIARY BUTYL IODIDE. 239 Combust,ions gave the following nnmbers :-I. ~0925 gram substance gave ,1164 gram OH2 and .2202 gram 11. ,0935 gram substance gave -1174 gram OHz and *2219 gram GO,. co,. Calculated for C4HyOH. Found. C 64.86 per cent. 64.92 64Ti H 13-51 , 13.98 13-95 0 21-63 , - -The total quantity of pure substance obtained from this reaction was 4.2 grams the calculated quantity being 7.9 grams therefore the process would seem to give very good results as there was of course, considerable loss in repeatedly distilling the small quantity from barium oxide. The reaction may be expressed thus :-(CH,),CI.CH + HOH = HI + (CH,)z.COH.CH3.This action of water is ?nalogous to its reaction upon the compound formed by absorbing isobutylene in sulphnric acid-(CH3) 2 C (0. SO,. OH) CHS, which also gives trimethyl-carbinol but it is the first example of an iodide of an alcohol radical being decomposed by water at ordinary temperatures. The t.emperature at which the action took place was, in the first case below O" and in the second below 10" 11. DECOMPOSITION WITH ZINC OXIDE During some of the experiments made in the foregoing it was found that zinc oxide acted upon tertiary isobutyl iodide in absence of water, especially on gently heating the iodide and adding dry ziuc oxide in small quantities. If much zinc oxide wei-e added at once the reaction became rather violent isobutylene was formed in considerable quantity, and a large amount of heat was evolved.26 grams of carefully dried iodide were mixed gmdually with the calculated quantity of dry einc oxide to combine with all the iodine present the temperature being kept at about 15" by cooling with water. A slight excess of zinc oxide failed to produce any further reaction. The liquid which was previously dark brown from the pre-sence of free iodine was now nearly colourless and the peculiar odour of the iodide was replaced by a pleasant ethereal odour. It was distilled directly from the precipitated einc iodide on an oil-bath. The distillate consisted of two layers the under one of whic 240 DOBBIN ON SOME REACTIONS OF mas only a very small quantity and was found to be water formed in the reaction.After separating the latter by means of a small sepa-rating funnel the upper liquid which weighed 3.5 grams was dried with calcium chloride and distilled. 1.t began to boil at about go" but the temperature rose rapidly to 170° only 1.3 grams coming over between 140' and that point. The remainder weighing 2 grams, distilled be'tueen 170" a i d 180" the boiling point remaining constant for some time at 175". The fractions 160-170' and l70-180" were again distilled and here the principal fraction weighing 0.9 gram, distilled between 174" and 176". This fraction was used for the analyses and for the vapour-density determination. Combustions gave the following aumbers :-I. -1949 gram substanoe gave -2503 gram OH? and -6121 gram cop.IT. -1569 gram substance gave -2043 gram OH and *a27 granr coz. Calculated for (CH2)n. Found. - C 83-71 85.65 85.64 H 14.28 1426 14.46 99-99 99-91 100~10 -Fapour-densi fy Determinution (Hnf riianw,'s Jfetlwtl). Weight of substance taken . 0.1762 gram = P Observed column of vapour 106.4 C.C. = V Height of mercury column above level in trough 463.6 nisn. = H Height of barometer i49.3 , = B Boiling point of aniline 182" C. - t Temperature of room . 15' C. = t' Vapour-density required . = I) -Tension of mercury vapour at 182' C. 11.89 mm. = 'I' Mercury column corrected to t'. . . . . . . . . . . . . . . . . . . . = H' IT' = H[1 - .00018(t - t')]. H' = 463*6[1 - *00018(182 - 15)]. H' = 449.7 mm. (air = 1).D = P . (1 + -00367t) . 760 *OO12932. V(B - H' - T) -1762 . (1 + -00367 . 182') . 760 *c1012932 . 106-4(749*3 - 449.7 -m' n = D = 5-64?. D calcnlated for (C,H,) = 5-805 TERTIARY BUTYL IODIDE. 241 From these determinations there could be no doubt that the body was isotributylene the formation of which may be expressed by the following equation :-G[(CH,),CI.CH,] + 3Zn0 = 2CIzH + 3H20 + 3Zn12. It may be well to mention a paper by Lermontoff approaching this subject which has recently been published (,4ni~aZen 196 116). Lermontoff saturated tertiary isobutyl iodide with isobutylene a t -lo" 15 gra,ms absorbing 7 to 8 litres of the gas. The solution thus obtained was seded in a tube with twice its weight of calcium oxide and the whole was heated for 20 hours to 100".Excess of isobutylene must be present in order that the product may not be con-siderably reduced in quantity. After the end of the reaction the liquid product on -distilling ~ f f and frdionating proved t0 consist of two bodies isodibutylene b. p. 102-5" and isotributylene b. p. 177.5" to 178.5". Zinc oxide or magnesium oxide gave the same result. In this react'ion the isobutylene formed by the action of calcium oxide upon tertiary isobutyl iodide is supposed to unite with the iso-butylene already present forming khe polymerides and the following equations are given :-but as quanfitative results are not given it cannot be ascertained whether a greater quantity of the mixed polymerides was formed than could have been formed from the iGdide alone.In the present experiments now published for the first time the action took place at ordinary temperatures and isobutylene was not formed or if a t all only in very small quantity the weight of the dis-tillate obtained on fractionating between 20" and 140° being only 0.1 gram or about gGth of the whole. Lermontoff does not mention in her paper whether the isobutylene had all been changed into polymerides by the heating in closed tubes, which is certainly a very important point in the reaction. 111. DECOMPOSITION BY SODIUM. Professor Wislicenus observed that when tertiary bu tyl iodide was acted upon by sodium a gas was formed a considerable quantity of which was absorbed by passing i t through bromine but the unab-sorbed gas still burned with a luminous flame.At his suggestion this investigation was taken up in order to ascertain whether or not the resultant gas consisted of isobutylene and isobutane and if it did 242 DOBBIN ON SOME REACTIONS OF whether they were formed in equal volumes as would be the case if the reaction were simply the following :-2C4H,I + 2Na = 2NaI + C4H + C4H,,. The %ohme of the gas formed was first approximately measured by using an apparatus for the decomposition which the accompanying sketch represents. 10 grams tertiary butyl iodide were placed with a weighed quantity of sodium (the former being in excess) in the flask A which was then heated for several hours in a water-bath. The gas formed mixed with the air originally in the flask and condenser tube passed over into the Woulfe’s bottle B whilst water was driven from the latter into the graduated cylinder C.When the level in C ceased to rise the whole was allowed to become quite cold the levels in B and C were made to correspond and then the volume of water in C was noted as well as the temperature and the pressure of the atmo-sphere. In two experiments the results agreed almost exactly with each other. The corrected volrumes of gas obtained in the two cases calcu-lated out and compared with 1 gram sodium were 630.8 C.C. and 630.5 C.C. respectively. These very close results must however be taken with care as in neither case had all the sodium been used up it having become surrounded by a covering of sodium iodide which pre-vented further action. The quantities left unchanged were however, very small.The gas thus obtained wa,s collected in a series of tubes and was subsequently analysed. The air with which it was mixed was not estimated as only the relative proportions of tho gasea formed by th TERTIARY BUTYL IODIDE. 243 reaction were wanted and previous to analyses the gas was acci-dentally mixed with more air than it had at first contained. The analyses were made with Bunsen's apparatus and the absorp-tion tube and eudiometer were specially calibrated for these analyses. Previous to any of the experiments the whole quantity of the gas used was dried with 8 bullet of potassic hydrate. The following are the various analyses made :-I. Ahsorption of the OleJne. This was accomplished in the absorption tube by means of coke bullets saturated with fuming sulphuric acid a fresh bullet being introduced every two hours until the last one on removal after this time still fnmed strongly.A potash bullet was then introduced to remove acid fumes and free sulphuric acid. 111. 6'om.histion of the Residlcal Gas afier Absorptiort of the OleJine. The gas was brought into the eudiometer at the top of which a drop of water had been placed to saturate the gas with moisture; oxygen was then added and the mixture was exploded by a spark from an induction coil. The contraction was measured then a potash bullet was introduced to absorb the carbon dioxide formed. 111. Estimation of the Valzce of n in the OleJiite. This was done with a portion of the original gas which had not been treated with fuming sulphuric acid and the method was similar to that employed in 11.The numbers obtained from the observations were as follow : 244 DOBBIN ON SOME REACTIONS OF I. AbsoFption of OleJine JC. Gas taken . __. _. . . * After bullets of fu ming sulphuric acic and KOH . 11. Combusfion of tht Residual Gas. 1. Gas taken . . . + Oxygen . . After explosion . . . . ._, *After absorption of CO: 2. Gas taken . + Oxygen . -. After explosion. . . . . . . . "After absorption of COP 111. Palue of n ilt OleJine. Uas taken . . . . _. . . + Oxygen . . . . ,. . After explosion . . . . . . . *After absorption of C02 Ob-served volume m in. 103 -5 ):;:; 125 -3 216 -7 206 *2 203.9 129 '0 215 -1 203 *1 197 *5 144 -0 356 *7 340 '3 328 -9 -102-46 88 *OO 89'73 131 -18 224 '02 ai3.41 211 *09 134 *94 222 -40 210 -29 204.63 150 -23 365 -4.0 349 -00 337.56 -Level of mercur in trough n1111.-236 *8 236 '0 237 -8 786 -0 785 -2 785 ? 786 -5 787 -4 788 -5 786 -8 787 -I 784 *9 784 *o 784 -5 796 -3 --Tem-pera-ture. "C. -5 *o 1.6 2 *o -2 *4 - 2 - 0 -1.1 +1*0 -2 -7 -2 '6 -2 '2 -5 *6 5.4 4 -0 4 *5 1 -8 -Baro-meter. mm. 747.4 737 -4 726 -0 759 -6 759 -6 759.2 '758 -0 762 '0 762 -0 760.1 '761 -5 747 '0 74!4 -1 '742 -7 73'7.1 --Cor-rected FOlUlllt? at 0" c. an cl 1,000 mm. -61 -79 51 *68 51.48 12.59 42 -23 37 -60 36 -89 13.30 41 -53 36 -57 35 83 14 -64 111 -90 100.33 20 '45 -Calculation of Results.I n I the percentage was calculated directly. In I1 the carbonic dioxide formed was taken as representing iso-butane C4R,,. This however was only a very small quantity whilst the contraction on explosion was considerable showing that hydrogen must be present in large quantity. x = C4H10 y = H, C = contraction D = CO formed. By the equation :-2CJHIo + 1302 = 8C02 + l@I&O 1 volume C4HI0 requires for combustion 6.5 volumes oxygen forming 4 volumes CO, and the D contraction is 3.5 volumes therefore x = -. 4 * The observations marked thus denote that the gas was dry. I n all others it was saturated with moisture TERTIARY BUTYL IODIDE. 245 3.5D The contraction due to CIH, is thus - -.4 Two-thirds of the re-2(c -:‘““I 3 maining contraction is due to hydrogen therefore y = I n 111 the amount of carbon dioxide which would have been formed by the mean of the quantities of isobutane found in I1 was subtracted from the whole quantity found and the remainder was calculated as representing isobutylene 3 = C4H,, y = H, i = c&. C = contraction D = CO? formed. D’ = CO formed by the quantity of isobutane present. x being taken = mean of experiments in 11 then 423 = D’. By the equation :-C4H8 + 60 = $Go2 f 4H,O 1 volume C4H8 requires for combus-tion 6 volumes oxygen forming 4 volumes C02 and the contraction 3D-D’ 3.5T)’ 4 2 c--is 3 volumes therefore z = - D - D’ ( 4 and?/= 3 Calculated out according to these formul~~ the results obtained were :-These results were unexpected as hydrogen had not been thought In order that hydrogen should be formed it was necessary that the likely to be present.reaction should take place in this way :-2[(CH3)$I.CH,] + Na2 = 2NaI + 2[(CH3),C=CH2] + H2, but in this case hydrogen would be formed i n the proportion of one volume to two of isobutylene whereas more than an equal volume was found. This was only explicable if part of the isobutylene had become polyrnerised and to ascertain if such were the case the residual liquids from the action of sodium were examined. On mixing with water three layers were formed the lowest of which was undecom-vor,. YYXVIL. 246 DOBBIN ON SOME REACTIONS OF TERTIARY BUTYL IODIDE. posed iodide and the middle one solution of sodic iodide in water.The upper liquid which burned with a brightly illuminating flame, was separated by a pipette dried with calcium chloride and then with sodium and was then distilled. It began to boil a t 152" and all dis-tilled below 1 8 2 O the principal quantity going over between 170" and 180". The latter was distilled fractionally several times with metallic silver to remove free iodine. The endeavour to remove all the free iodine was not successful as the whole quantity of liquid at this point weighed but 0.5 gram. A small portion was obt'ained boiling between 174" and 178" which had the ethereal oil odour of isotributylene and although an analysis did not give good results owing to the presence of iodine there could be no doubt that this was the body which was formed. Thus i t would appear that by decomposing tertiary biityl iodide by means of sodium the bodies formed are isobutylene isotributylene and hydrogen with small quantities of a hydrocarbon not absorbed hy fuming sulphuric acid. It will be noted that here as in the decom-position with zinc oxide not a trace of isodibutylene was observed. One more point was calculated out viz. whether the quantity of hydrogen found was such as should be formed if the reaction took place according to the last equation but taking into account the for-mation of polymerised isobutylene and of isobutane it was found that the quantity obtained was only about 75 per cent. of the theoretical quantity supposing the sodium to have been completely combined with iodine. As however the sodium had not been entirely used up, the difference might be accounted for from that fact. I n conclusion I wish to express my sincere hhanks to Professor Wislicenus for masterly assistance rendered during the progress of the work