III III IIIIII I I l I l — — — — — — — THS STLJDY OF THE FRAGMENTATION OF Di-ISOPROFYL METHYL AND DI-ISOPROPYL ETHYL CARBINOLS IN THE PRESENCE OF ALUMINUM CHLORIDE Thesis fox IItE Degree c-f M. S. h’IICHICAIQ STATE COLLEGE large Awuapara 1942‘ STUDY OF THE FRAGHENTATION 0F DI-ISOPHOPYL METHYL AND DI-ISOPROPYL ETHYL CARBINOLS IN THE PRESENCE OF ALUMINUM CHLORIDE by JORGE AWUAPARA A THESIS Submitted to the Graduate School of Michigan State College of Agriculture and Applied Science in partial fulfilment of the requirements for the degree of MASTER OF SCIENCE Department of Chemistry 1942 '53? :1; t 9'; ” / .wzrz‘ fl ’3 El autor deeea expreear su mas profundo agradecimiento 31 Dr. R. C. Huston por la enorme ayuda prestade durante el deearrollo de la presente investigacion. 331-601 \'4>\ *¢ I NTHODUCTIJN Previous work from this Laboratory (1) has shown that t-aliphatic alcohols condense with benzene in the presence of aluminum chloride to give the expected t-alkylbenzenes. Further investigation (2) has shown that the accumulation of alkyl groups on the carbon atom adjacent to the carbinol carbon has a marked depressing influence on the condensing capability of the compound. The isolation of hydrocarbons of lower molecular weight from the condensation of some t-octyl alcohols was related directly with a marked reduction in yield of octylbenzene, suggesting fragmentation rather than a direct depressive influence in the condensing capability of the alcohol. The nature of these lower molecular weight hydrocarbons was determined in only a few cases, and no generalizations could be drawn from these isolated facts. The purpose of the present investigation is to determine the nature of the fragmentation products obtained from highly €_ branched alcohols when condensed with benzene using aluminum I chloride as a catalyst, and to determine the type of alkyl groups that facilitate the formation of the lower homologs, thereby reducing the yield of the expected alkylbenzene. Three homologous alcohols were chosen for this investigation: Di-isopropyl carbinol, Wethyl Di-isOpropyl carbinol and Ethyl Di-isopropyl carbinol. HISTORICAL The condensation of an organic halide with an aromatic compound, using aluminum chloride as a catalyst, was announced in 1877 by Friedel and Crafts as "Une nouvelle methods generals de synthese d'hydrocarbures". (3) Later the Friedel-Crafts reaction has been applied extensively to many other fields of organic chemistry. The Friedel-Crafts reaction was first confined to the preparation of benzene homologs using alkyl-halides as alkylating agents, but later was shown that alkylation could be also accomplished by using olefinic hydrocarbons instead of alkylrhalides. The reaction was further extended to the synthesis of aldehydes and ketones; to the addition of olefinic acids, esters and kstones to aromatic hydrocarbons; to intramolecular ringblosures or arylalkyl compounds having an olefinic double bond; to the synthesis of polynuclear hydro- carbons, etc, etc. Dehydrating condensations were also effected with alumium chloride. The alkylating and acylating agents which have been used included carboxylic acids, esters and ethers. More recently have the alcohols achieved prominence in this dehydrating condensation reaction with benzene and other aromatic compounds. The first reported condensation of alcohols with aromatic compOunds using aluminum chloride was the work of Ref (4) who, in 1897, prepared diphenylmethane “F-Jp'filfiv by allowing benzyl alcohol to react with benzene in the presence of aluminum chloride. This work was repeated by Huston and Friedeman (5) and a 30 percent yield of diphenylmethane was reported. I In continuing this work, Huston and Friedeman condensed several mixed aliphatic-aromatic and true aromatic secondary alcohols with benzene. Huston (6) later used phenol and some of its others in place of benzene and found that benzyl alcohol gave yields of 45-50 percent of alkylated product. Huston and Sager (7) investigated the condensation of primary alcohols with benzene but were unsuccessful in condensing: methyl, ethyl, propyl, n-butyl, isobutyl, isoamyl, phenylethyl and phenylprOpyl alcohols. Later Norris and Sturgis (8) were able to condense primary alcohols with_bensene, using larger amounts of aluminum chloride and higher temperatures. According to them.the course of the I reaction of alcohols with aromatic compounds is definitely regulated by the quantity of aluminum chloride employed. This accelerating effect of excess catalyst wad high temperatures had been previously pointed out by Tsukervanic and Vikhrova (9). In a series of experiments, Huston and co-workers showed that unsaturation of the alpha-carbon atom favors condensation reactions of aliphatic and.aromatic alcohols with aromatic hydrocarbons and phenols in the presence of aluminum chloride. "574F319? This unsaturation may be due to an aromatic ring (10), a simple double bond or strain in a polymethylene ring (11). In 1936, Huston and Hsieh (12) reported the condensation of several saturated aliphatic alcohols with benzene and phenol. They succeeded in condensing secondary and tertiary alcohols but it was found that under the conditions of their experiment primary alcohols would not condense with_either phenol or benzene. . In 1934, Huston and Fox (13) condensed tertiary butyl, tertiary amyl, and the three tertiary hexyl alcohols with benzene, while Huston and Binder (14) condensed the tertiary heptyls with benzene. The condensation of the tertiary cctyl alcohols with benzene and phenol was investigated by several workers. In 1940 Huston and Wasson (15) condensed some of the dimethyl- amyl carbinols with benzene, and Huston and Guile accounted for the condensation of eight of the seventeen tertiary cctyl alcohols. Fragmentation of the carbon chain was 'found to occur when the amyl radical was highly branched. As a result of this fragmentation some alkyl benzenes of lower molecular weight were obtained. In the same year Huston and Esterdahl found that condensation of some secondary alcohols gave a mixture of products. (17). In continuing this investigation, Huston and Kaye (18) found that secondary alcohols condense with benzene to yield rearranged product or mixtures of both ’9“ F :71; 1 f6 rearranged and not rearranged products. Only isOpropyl, s~butyl and pinacolyl alcohol gave the corresponding secondary alkylbenzenes in the pure form. The first rearrangement was observed by Gustavson (19), who obtained cumene from the condensation of both normal prcpyl and isOpropyl bromide when condensed with benzene using aluminum bromide as a catalyst. Similar results were obtained by Kekule and SchrBtter.(20). This observation was the groundwork for a series of investigations on rearrangements caused by the action of aluminum chloride, but no conclusive results have been drawn to decide the actual mechanism of rearrangement. The tendency of aluminum chloride to exert a disruptive action on aromatic hydrocarbons has long been observed.by Friedel and Crafts (21). Following this observation other workers have intensively studied this effect. Anschfitz, (22,23) in studying the effect of aluminum chloride on several aromatic hydrocarbons, found that _poly-alkylated rings are dealkylated.by aluminum chloride to give lower homologs and rings containing less alkyl groups. He also observed that migration of the alkyl groups occurs in the benzene ring. Von Dumreicher (24) found that bromobenzene in the presence of aluminum chloride undergoes reaction to yield benzene and dibromobenzene. Kohn (25) and co-workers have studied the migration of halogens from phenols to aromatic hydrocarbons. In connection with this study, Copisarow and Long (26) in their study of halogen migration in dibromo- benzene, attempted to fix the nascent halogen by substituting phenol for the benzene and obtained tribromophenol. In connection with the present investigation it is important to mention the fact that aluminum chloride acts in such a way as to induce cleavage, cracking, and transfer of the cracked fragments whether to the same or other molecule. II I I, \'I IIII‘IIIIIIII’I‘I III I I THEORETICAL Various mechanisms have been advanced for the condensation of alcohols with aromatic compounds in the presence of anhydrous aluminum chloride. Unfortunately, there is much confusion in the literature with regard to both the nature of the product and the mechanism of the reaction. ‘ A I It is beyond the scope of this thesis to discuss the I. many mechanisms proposed for the Friedel—Crafts reaction; however, a discussion of the latter would be necessary to explain any mechanism for the reaction of alcohols, since it has been proved that the intermediate compounds are alkyl halides and olefins. The formation of alkyl halides has been explained in several ways. Tsukervanik and Nasarova (27), have suggested as a mechanism the formation of a complex from the alcohol and aluminum chloride, hydrogen chloride being given off simultaneously. The complex then breaks down to form an olefin which can combine with the hydrogen chloride to form an alkyl halide. The reaction may be represented as follows: Cannon 4- ALCLS...’ ALCL2(065H11) 4- I101. ALCL2(005H11) ----) 05H 4» ALCL OH 10 2 051110 0 LICL MOSHIlCL Tsukervanik's mechanism for the condensation of primary and secondary alcohols is similar, but in this case the intermediate aluminum alkoxy compound is thought to react directly with benzene prior to splitting and formation of the olefin. Thus: n-CSH7OH o ALCL3 _.___.._..) n-CSH7OALCL2 . nor. n-05H70ALCL2 4 06116 ALCLL n-csn706115 +IIOALCL2 This mechanism will not explain the rearrangements obtained when some secondary alcohols are condensed with benzene. Recently Norris and Ingraham (28) have reported the alkylation of benzene with methyl and ethyl alcohols. They found that the alcohols react with aluminum chloride to form compounds of the formula ROALCLZ, which decompose when heated to form the alkylhalide. According to them, one mole of the alcohol reacts with one mole of aluminum chloride to give the corresponding alkylhalide: 02115011 § ALCLS -———) 62H50H°ALCL3 lICL‘t CQIISO'ALCLz -—-) 02H50L+ ALOCL That alkylation with alcohols does not proceed simply through formation of the complex ROALCLZ was shown by reacting this complex with benzene. No alkylated products were obtained. Regardless of the path, the formation of alkyl halides has been proved to be an intermediate in the condensation of alcohols with aromatic compounds when aluminum chloride is used as condensing agent. An ionic type of mechanism has been suggested by Price (29) for the alkylaticn of benzene. He points out that Weryporoch and Firla (30) have clearly demonstrated.by conductance studies the formation of an ionic complex between aluminum chloride and an alkyl halide: CL CL as 00 00 I. O R:.J§:*§L:CL_R¢:2{.:§‘L:CL CL CL In this complex the carbon-halogen bond is weakened, and the compound dissociates into an electron-deficient carbonium ion and a negatively charged aluminum complex: we .CL es OCCL R":1<,:AL:CL".—..R¢and :L:AL:CL' CL CL Price assumes that the reaction is then entirely analogous to halogenation, the electron-deficient carbonium ion (R ) reacting Just as does the Bromine cation (Bra) to complete its octet by association with a pair of electrons from a double bond of the aromatic nucleus. 10 This mechanism could.be pictured.asfollows: H a . . “a. , "\L R .4. //\¥ R 4. + 173* ‘““ b‘>/‘ V” L ?1 + H .. : + \ J H Price also explains, on the basis of the cationoid mechanism, the condensation of olefins with aromatic compounds to yield alkyl benzenes. The electron-deficient catalyst in this case associates with a pair of electrons from the double bond of the olefin to give an active intermediate which is similar to that of the ordinary Friedel-Crafts reaction in that one carbon atom has only a sextet of electrons: H H R z c ' R 3 c * ::+A e " H 2 G H : C :‘A H B This carbon atom then completes its octet by association with an electron pair from the aromatic nucleus. The final step is elimination of the catalyst, followed by the shift of a proton. *Y’b I . pr; .flf K. 11 This mechanism could be pictured as follows: P"C’H+ m * —s» I x ; hearrangemont of the alkyl groups can also be explained with the caticnoid mechanism. According to Whitmore (31), the electron-deficient carbon atom in the intermediate alkyl cation should tend to be secondary rather than primary and tertiary rather than secondary. In cases where rearrangements do not occur, the existence of the alkyl cation may be too transient to allow for the migration of a proton ( or an alkyl group ) leading to a rearrangement product. PM 1.; 3;,“ 'II‘IleI‘IIIJ‘ 12 EXPERIMENTAL I. Preparation of the Carbinols Di-ieopropyl carbinol. Munch (32) first prepared this alcohol in 1909 by the reduction of di-isOprOpyl ketone with sodium.and moist benzene. Conant and Blatt (33) prepared the same alcohol by the reaction between isobutyraldehyde and isopropyl- magnesium bromide. They reported a yield of 78 percent when .25 mole of di-isOpropyl ketone was added to a theoretical .5 mole of isopropyl-magnesium bromide. Stas (34) used the same procedure successfully and also obtained the alcohol as a reduction product when di-isopropyl ketone was condensed.with ethylmagnesium bromide. Preparation of isopropyl-magnesium bromide: Isopropyl- bromide was prepared by placing 5 moles of isopropyl alcohol in a three liter flask and carefully adding 10 moles of concentrated sulfuric acid. The mixture was cooled, 400 cc. of water and 10 moles of sodium bromide was added. This mixture was refluxed for four to five hours. ‘Two layers were formed, the upper one being the iscpropyl bromide, which was separated, washed with potasium carbonate and then with water. After drying over calcium chloride it was distilled. This freshly distilled isopropyl bromide was reacted with magnesium in the usual manner to give a theoretical yield of 5 moles of isopropyl-magnesium bromide. 13 Preparation of the Carbinol: To the Grignard reagent thus prepared was added one and one-half mole of anhydrous di-iscprcpyl ketone mixed with an equal volume of anhydrous ether. The rate of addition was so regulated that the ether refluxed gently. After the ketone had been added the stirring was continued for a period of at least one hour. The reaction mixture was then decomposed with ice, and sufficient concentrated hydrochloric acid was added to barely dissolve the basic magnesium salts that were formed. The other layer was then separated, and the aqueous layer extracted twice with ether. The combined ether extracts were washed once with water, then with a dilute sodium carbonate solution and finally dried over anhydrous sodium sulfate. The ether was removed.and the alcohol distilled at reduced pressure. Yield 70 percent B'P745 156° - 137° c. B’P15 46° - 50° 0. 03” 0.8145 n23 1.4253 x.“ :$9,7/ The alpha-phenyl urethane melted at 94—960 C. l4 methyl Di—isqpropyl Carbinolg By adding di-isopropyl ketone (isobutyrone) to methyl- magnesium bromide, Stas (34) prepared this alcohol, the yield being as high as 75 percent. Preparation of methyl-magnesium bromide: Methyl-magnesium bromide was prepared.by slowly reacting methyl bromide gas with magnesium in the usual manner. The methyl bromide gas was generated in sufficient quantity by reacting solium bromide with concentrated sulfuric acid and methyl alcohol in a generator the description of which is given in G. R. Meloy's,Doctors thesis (35). Preparation of the Carbincl: To a theoretical 5 moles of mmthyl-magnesium bromide 4 moles of di-isopropyl ketone, mixed with an equal volume of anhydrous ether, was added. The procedure is the same as the one given for the preparation of di-isOpropyl carbinol. Yield 65 percent. s-p750 156° - 157° 0. B-Plz 52° - 54° 0. DEC 0.8492 4, n%0 1.;653 I ‘. k“ 15 EthylgDi-isgprcpyl Carbinol. Stas (34) first prepared this alcohol by adding di-isopropyl ketone to ethyl magnesium bromide. He reported a yield of 54 percent of ethyl di-isopropyl carbinol and 24 percent of the reduction product, di-isoprOpyl carbinol. Ethyl di-isopropyl carbinol was obtained as the single reaction product when ethyl-magesium.bromide was reacted slowly with di-isopropyl ketone cyanhydrine. I Preparation of ethyl-magnesium.bromide: The preparation of ethyl-magnesium bromide was accomplished in the same manner as the other Grignard reagents. For a theoretical 5 moles of the ethylznsgnesium bromide, 4 moles of the ketone were used. Preparation of the Carbinol: The condensation of the ketone with the Grignard reagent was carried out in the usual manner. The rate of addition was so regulated that other refluxed gently. Yield 50 percent B'P755 176-177 B'Plz 65 ~64, of? 0.81% n30 1.4Zg; L... 7?... g; g ,1 All of these carbinols are very viscous liquids. They all possess a camphor-like odor very characteristic of this group of alcohols. "FM“ ‘5 .16 II. Condensations. All condensations have been carried out in the same manner, the conditions being those more favorable for fragmentation. It has already been shown by many workers that higher temperatures decrease the yield of the main products and increase the amount of fragmentation. For this type of work, however, room temperature has been found :i'~'-‘""." 31!. to be the most reliable, since fragmentation occurred in the absence of external heat that could alter the course 9 $33525 “1E5"? .". of the reaction. Cooling has also been found to facilitate. -.~‘ ’w‘nlr‘a‘ the formation of chlorides, probably chlorides of the carbinol, thereby decreasing the yield of fragmentation products. The quantities and procedure adopted.were the ones worked out by Huston and Kaye (36) for the condensation of secondary alcohols but slight modifications have been made. Procedure. One-half mole of the alcohol, dissolved in 50 m1. of benzene, was added drop-wise and with vigorous stirring to a suspension of 50 gms. of anhydrous aluminum chloride in 400 cc. of anhydrous thicphen-free benzene. The reaction mixture was stirred for one hour after the addition of the alcohol and than permitted to stand over~ night. The reaction mixture was then poured on ice; the benzene layer was separated, washed with water, then with a dilute sodium carbonate solution and once more with water. After drying over anhydrous sodium sulfate, the benzene was removed and the residual liquid was then fractionated. 17 Di-isoprOpyl Carbinol One-half mole (57 gms.) of the alcohol was condensed with benzene at room temperature. The reaction mixture which was fractionated first at reduced pressure gave three fractions. The first two fractions were combined and refractionated at atmospheric pressure, while the third fraction was redistilled at reduced pressure. The final result was the following: Fraction 3-? Yield (gms) 1 112-118 (745mm.) 10 2 101-104 (20 mm.) 30 3 residue 10 Fraction 1 showed slight decomposition during the distillation. The presence of hydrogen chloride as a decomposition product indicated the presence of some unstable chloride. This fraction was refluxed for several hours with a dilute sodium hydroxide solution. The organic layer was then separated, dried over anhydrous sodium sulfate, and fractionated at reduced pressure. An unsaturated compound was distilled over at 83-86. In order to establish the identity of this olefin, it was oxidised with potasium dichromate (57) in acid solution. The oxidation took place very slowly, and it was necessary to boil the mixture under reflux for about 25 hours. The product was then diluted with water and extracted with other. ‘5 " 33‘5“! 18 After the ether extract was dried over calcium chloride, it was distilled at atmospheric pressure. No definite fractions were obtained, but the distillate was acid to litmus paper and possessed the same odor as isobutyric acid. , i Since the oxidation of this olefin did not give sufficient quantities of the acids in order to be identified, the physical constants of the olefin were determined, and checked closely with those given in the literature (38) for 2,4-dimethyl pentane-2, which would be the expected product when hydrogen chloride splits off from.di-isOpropyl‘chloride.' The physical constants found, were the following: 3-? 84-85 D 25 0.6961 n :5 1.4016 Fraction 2 consisted of 2,4-Dimethyl, 2-phenyl pentane as it has been established.by the melting point and.mixed melting points of its derivatives. The expected product of the condensation of this alcohol was 2,4-dimethyl-5-phenyl pentane, but evidently a rearrangement took place. This fraction was nitrated, reduced, diazotized, and hydrolyzed to the phenol (59). The alpha-naphtyl urethane, and the bensoY] ester of this phenol were prepared. Preparation of the alpha-phenyl urethane of the phenol: (40) One to two grams of the phenol was placed in a test tube wasps 19 and .5 ml. of alpha-naphtylisocyanate added. It has been found that, if the reaction is not spontaneous, addition of a few drops of a solution of trimethyl amine in other will readily catalyse it. During the reaction the contents of the tube must be protected against moisture by means of a calcium chloride drying tube. The solution was heated over a low flame for a minute, cooled inaa beaker of ice, and the sides of the tube scratched in order to induce crystallization. ,The urethane is purified by repeated.recrystallization from skellysolve or petroleum other.‘ Preparation of the Benzoyl Ester of the phenol: (41) )/I One-half gram of the phenol was dissolved in about 2 cc. of pyridine and .5 ml. of benzoyl chloride was added. After the initial reaction the mixture was warmed over a low flame, cooled and poured on to ice with stirring, and extracted with other. The ether extract was washed with acid to remove the pyridine, then with sodium.carbonate solution. After the removal of the ether the ester crystallised after standing in the ice box. The crystals thus obtained were recrystallised from.85 percent alcohol and finally from 85-89 percent acetic acid. we .lil .‘ I‘ll. 20 m-P of the alpha-naphtyl urethane : 114-115 MOP of the benzoyl ester : 70-71 The alpha-naphtyl:urethane and the benzoyl ester of iso-butyl-dimethyl-p-hydroxy phenyl methane have been prepared by Hedrick (42) and the melting points are in perfect agreement with those given above. Mixed.melting points of the bensoyl ester did not show depression. Therefore, this fraction consisted entirely of 2,4-dimethyl- 2-phenyl pentane. Yield 36 percent B-P745 216-217 Bopzo 101-102 D 90 0.8724 n 2° 1.4928 u 9.; 21 Methyl Di-isOpropyl Curbing; One-half mole (65g) of the alcohol which condensed with benzene gave 65gms. of a mixture whose composition was the following: Fraction B‘P (10 mm) Yield 1 46-50 8 2 52-54 30 3 102-105 9 4 residue 18 Fraction 1 consisted of a mixture that could be partially nitrated. About 80 percent of this mixture N was methyl di-isopropyl chloride and.about 20 percent tertiary butyl benzene. The chloride could not be isolated in pure form in order to determine its physical constants and insufficient quantity was available to prepare a derivative. However, when hydrogen chloride was split off by means of a sodium.hydroxide solution, an unsaturated compoundcould be distilled at about 9’ 115-1180 C. The boiling point for the olefin corresponding to methyl di-isopropyl alcohol is reported in the literature as 118° (43). Fraction 2 consisted entirely of tertiary butyl benzene as shown by the melting point and mixed melting point of its acetamino derivative. This derivative was PU”?! _. a, J 22 obtained by the method of Ipatieff and Schmerling (44) with some modifications: "One gram of the unknown was shaken with 5 gms. of a nitrating mixture that consisted of equal volumes of concentrated sulfuric and.mitric acids. When the evolution of heat had ceased, the mixture was poured upon cracked ice; the oily nitro layer was taken up in ether, washed twice with water, and the ether evaporated on a water bath. To the residue was added 5 gms. of mossy tin and about 20cc of ethyl alcohol. Concentrated hydrochloric acid was then added in small portions, vigorous shaking being necessary after the addition of each portion. When 5 to lOcc. of the acid had been added the reduction was almost completed. Upon compllgtion of the reduction the aqueous alcoholic solution was decanted from the excess tin into about 20 cc. of water. The solution was extracted with ether; ether was then evaporated to a small volume. Water was added and all the ether evaporated off on the steam bath. Sufficient alkali was added to the aqueous solution to dissolve most of the precipitate of tin hydroxide which formed first. The amino compound was extracted with ether or steam distilled, washed with water and the ether removed by evaporation. To the residue was added 1 cc. of acetic anhydride bringing about the crystallization of the derivative almost immediately. The excess anhydride was hydrolysed with . \ 5 cc. of water. The derivative was filtered, washed several times with water and recrystallized from dilute alcohol to constant melting point. M-P of the derivative 168-169 Mo? 0: the known ’ 169-170 Mixed melting point 168-170 Fraction 3 upon several refractionations gave a constant boiling portion at 104-105 (11 mm). The identity , of this compound has been established by C. R. Meloy (45) and proved to be 2,3,4-Trimethyl-3phenyl pentane. Yield 9.0 percent BvP749 234-256 03° , x 0.8808 nBDO 1.4968 Fraction 4 decomposed entirely when heated above 1100 at 10 mm. and it was impossible to be distilled. '5 773?: ,w‘ . I "A 24 Ethyl Di-isoprqpyl Carbinol One-half mole (723ms.) of the alcohol which condensed with benzene gave 50 gms. of a mixture whose composition, after several fractionations was the following: Fraction 3-? Yield (5&3.) 1 145-160 (750mm,) 10 Pd.“ 2 165-172 (750mm.) 2’?“ 5 188-193 (750mm,) 3 _tw 4 106-114 (10 mm.) 25 ” 5 residue 10 Fraction 1 consisted of a chloride that liberated hydrogen chloride when heated at its boiling point. This chloride was converted into the corresponding olefine by the previously described method. The olefin thus obtainedfl 1i«: was distilled over at 128° C. and some of its physical if constants were determined. A survey of the literature revealed no olefin whose physical constants checked with the above. However, the boiling point of 2,3,4-tri- methyl pentane-2 was found to be 11g” 0. It was postulated that the addition of one more carbon atom.t0 this olefin should increase its boiling point to the neighborhood of 128. In support of this postulation the calculated.molecular refraction for 2,4—dimethyl-3-ethy1 pentane-2 was compared with the observed molecular refraction of the unknown. The slight discrepancy between the observed and calculated 25 values may be due to the presence of some impurities in the olefin. Be? 128 0:0 0.7652 <3" 3 '5 3‘“ n g0 1.4120 =3 / 757’ 20 as, 1- M°R 41.083 (calc. 43.140) Fraction 2 was identified as tertiary butyl benzene by means of its acetamino derivative. The method for the preparation of this derivative has been previously described.in this thesis. Fraction 3 consisted of tertiary amyl benzene and small amounts of tertiary butyl benzene. The isolation and identification of tertiary amyl benzene has been accomplished by repeated fractionation and by'conversion or the fraction into the acetamino derivative. Before a constant melting point could be obtained.more than 10 recrystallizations were necessary. The melting point of the acetamino derivative, was a few degrees lower than the melting point reported in the literature (46). However, it did check with the melting point found by the author, and mixed melting point did not show depression. Fraction 4 upon several refractionations gave a constant boiling portion at 112—114 (10mm.). The identity of this "FIE 1'. _ _ . ‘ _... .. ,4 .-~ :t‘nnrgc. _.-. . -... _ i 26 fraction has not been established but it was assumed to ~ be the nonyl benzene; The physical constants were ”'5ff'w» determined and the calculated molecular refraction was in agreement with the observed. Yield 20 percent B'Plo 112-114 ( D25 0.8706 "V?" \.':q 1:: ' n§5 1.4950 Y' 25 . MoR . 68.21 (0810. 67.88) ,7" 2 (7' . (o I: . ‘ , A I (. . 1‘ ‘1”, .‘ r» f I “:/;}f"/ F‘H'r'f '\ iélrv'J/j “VHF 27 Methyl Di-isopropylgphenyl methane and Aluminum Chloride. To further investigate the disruptive action of aluminum chloride, this hydrocarbon was treated with aluminum chloride in the presence of a large amount of benzene. Fifteen grams of methyl di-isopropyl-phenyl methane were mixed with 100 cc. of anhydrous benzene. Ten grams of anhydrous aluminum chloride were added and the mixture refluxed on a sand bath for about 2 hours. After cooling the mixture was poured onto ice in order to hydrolize the aluminum chloride present. The organic layer was separated from.the aqueous solution and it was washed once with a dilute solution of potassium carbonate and twice with water. After drying over anhydrous sodium sulphate the mixture was fractionated at atmospheric pressure. A fraction boiling at l68-170° C. was collected and the residue began to decompose at this temperature. This fraction was redistilled and converted into the acetamino derivative. The melting point and mixed melting point of this derivative showed that this fraction was tertiary butyl benzene, which was the expected product from the fragmentation of methyl di-isoprOpyl-phenyl methane. 28 .Honooaw on» no eHoB one on edownooahxae no oHoE one so venom a oneocem 55.5 .6433 35 9:5 n wow: -euahsogsum.a H.523. 78-15pm oneness 533a...» H8330 a: o -HEoosEequJ anooosofideAafiox ensuaem Andean Hosanna". .......i...... 3 stintseouim ”2283-3 odeahh oceacom span emonoaon acted oaeah.»deoaem ,hll encasondthdeoa needeosoo Honooad )1 H qu