THE KINETICS AND MECHANISM OF THE RACEMIZATION OF PHENYLMETHYLC ARBINYL CHLORIDE BY PHENOLS By WILLIAM LUDWIG SPLIETHOFF A THESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1953 A CKNOWLEDGMENT The author wishes to express his sincere appreciation to Dr. Harold Hart for his helpful guidance and for the many stimulating discussions throughout the course of this work. He is also indebted to his wife, Dorothy, for her assistance and sympathetic understanding which helped to make possible the .completion of this thesis. TABLE OF CONTENTS INTRODUCTION HISTORICAL PAGE 1 3 EXPERIMENTAL 10 RESULTS 28 DISCUSSION OF RESULTS 39 SUMMARY 72 REFERENCES 75 APPENDIX 77 INTRODUCTION The alkylation of the aromatic nucleus with optically active molecules has been little Investigated. Such an investigation should yield Information concerning the m e cha nism of aromatic alkylation. Furthermore, it is attractive from an experimental viewpoint, In that reactions involving optically active molecules are easily followed. It has recently been shown (l) that the uncatalyzed nuclear alkylation of phenol by optically active phenylmethylcarbinyl chloride proceeds readily at convenient temperatures and that the mixture of ortho and para-phenylethylphenols produced has a measurable optical rotation. It was decided to examine this reaction by means of a detailed kinetic study In the hope that it would result in a clearer understanding of the stereochemical aspects of nuclear alkyla tion. It became evident at the outset of the work that the reaction involved an independent racemization of the phenylmet hylcarb inyl chloride induced by the phenol present In the reaction m i x t u r e . Accordingly, it was decided to study the racemization reaction and the alkylation reaction as separate processes. It was of interest to examine the racemization of phenylmethylcarbinyl chloride in various alkylated phenols in order to evaluate the role of the phenol in the racemization reaction. Optically active phenylmethylcarbinyl chloride dis appears as a result of alkylation which is a necessary adjunct - 2 - to its racemization in all phenols having an unsubstituted ortho or para position. It was, therefore, necessary to determine the extent of the alkylation reaction under con ditions identical to those employed in the study of the racemization reaction. The data necessary for an analysis of the two reactions were obtained for phenol at three temperatures, ;p-cresol, ocresol, 2,6-dimethylphenol and mesitol. The kinetic data thus realized have been fitted by differential equations which p er mit a plausible interpretation with respect to the mechanisms of these reactions. - 3 - HISTORICAL Although numerous examples of alkylation of the benzene nucleus have been reported (2), the study of the stereochem istry of the alkylation process has been largely neglected. Price and Lund (3) and Burwell and Archer (4) have studied the alkylation of benzene with optically active sec-butyl a l  cohol in the presence of boron trifluoride, hydrogen fluoride, sulfuric acid, phosphoric acid and aluminum chloride. Optically active sec-butylbenzene with a small rotation, opposite in sign to that of the sec-butyl alcohol employed, was obtained in all cases except in the presence of aluminum chloride, which led only to racemic product. However, extensive racemization (a minimum of 9 9 .3^) also took place in the presence of all the other catalysts. This has been cited as evidence for a free carbonium ion intermediate in these reactions (5). The fact that the sec-butylbenzene had a measurable activity was inter preted as being due to the alkyl cation reacting almost simul taneously with the process of ionization. Thus, it would perhaps retain its asymmetry to such an extent as to account for the slight activity observed in the alkylated benzene. Burwell (6 ) has also shown that the alcohol is appreciably racemized in hydrogen fluoride and sulfuric acid which may partially account for the extensive racemization observed when benzene is alkylated with sec-butyl alcohol in the p r e s  ence of these c a t a l y s t s . - 4 - A reaction, which is somewhat akin to the direct alkylation of the benzene ring, alkylaryl ethers. is the rearrangement of The rearrangement of optically active sec-butyl phenyl ether in the presence of concentrated sul furic acid or zinc chloride in acetic acid leads to secbutylphenol with partial retention of optical activity (7). Both direct alkylation of benzene with sec-butyl alcohol and the rearrangement of sec-butyl phenyl ether involve scission of a carbon-oxygen bond, followed b y the union of the alkyl group thus liberated with a carbon atom of the benzene ring. The greater retention of activity in the case of the rearrangement reaction is ascribed to the fact that the rearrangement is intramolecular and the alkyl fragment never leaves the vicinity of the parent m o l e c u l e . The possibility that this rearrangement proceeded via a bimolecular (intermolecular) mechanism was excluded b y the work of Gilbert and Wallis (8 ). They showed that the r e a r  rangement of optically active sec-butyl mesityl ether in the presence of p-cresol led to completely racemic 4-methyl-2-secbutylphenol. Since this rearrangement must be intermolecular, it was concluded that the rearrangement of sec-butyl phenyl ether leading to optically active sec-butylphenol must be intramolecular. Hart (l) has recently shown that phenol is alkylated by phenylmethylcarbinyl chloride in the absence of a catalyst to yield a mixture of ortho and p a r a - a - p h e n y l e t h y l p h e n o l s . The - 5 - reaction was essentially quantitative in one hour when carried out at a temperature of 7 5 °» in the absence of a solvent. Alkylation with optically active phenylmethyl carbinyl chloride, 4l.6, under the same conditions gave the mixture of <x-phenylethylphenols with a rotation of -0.35. The fact that any alkyl chloride recovered from the reaction mixture was appreciably racemized was co n  sidered significant especially when it was observed that racemization of the alkyl chloride also took place in mesitol although much less rapidly (l). The detailed investigation, of this reaction is the subject of this thesis. A survey of the literature revealed that the racemization of optically active alkyl halides in the presence of phenols has never been studied. Indeed, investigation of the r a c e m i  zation of alkyl halides under any conditions appears to be limited. It has been observed, for example, that phenylmethyl carbinyl chloride (0.252 M) is more or less rapidly racemized in liquid sulfur dioxide, hours (9). the half-time at 0° being eighteen The racemization is unaffected by the addition of chloride ion, (as tetraethylammonium chloride) hence the p r o c  ess is not a displacement accompanied by inversion (see below) nor a fast reversible ionization followed by a rate-determining racemization of the alkyl cation. The racemization should be accelerated by chloride ion in the first case and be retarded in the second case, since the common ion effect would serve to - 6 - suppress the ionization. In addition, since analogous halides give conducting solutions in this solvent, concluded that the alkyl cation, it was CgH^-foi-CH^, was formed. Such an ion would n e c e s s a r i l y be plan a r or, at least, capa ble of assu m i n g p l a n a r i t y r e l a t i v e l y easily; hence it would lead to equal amounts of b o t h enantiomorphs u p o n r e c o m b i n a  tion with chloride ion. More recently, Hughes, Ingold and Scott (10) have suggested that the r a c e m i z a t i o n of p h e n y l m e t h y l c a r b i nyl chloride in liquid sulfur dioxide is due to a revers ible loss of h y d r o g e n chloride as indicated below. C5H5 - C H C I - C H3 *-C6 H 5 -CH = C H 2 + HC1 The reco m b i n a t i o n of the h y d r o g e n chloride w i t h the optically inactive styrene m u s t lead to the racemic m o d i f i c a t i o n of phenylmethylcarbinyl chloride. S uch a m e c h a n i s m is supported by the fact that the rate of r a c e m i z a t i o n was found to be a p  proximately equal to the rate of f o r m a t i o n of free h y d r o g e n chloride. The r a t e - d e t e r m i n i n g step in the m e c h a n i s m proposed is u n i m o l e c u l a r i n v olving the io n i z a t i o n of the alkyl chloride followed b y a fast step in w h i c h the free alkyl cat i on is s t a  bilized b y the e j e c t i o n of a proton. c6h5 chci-ch3— C6H 5 - S h - C H 3 > c 6h5 -CH3 + Cl C6 H 5 - C H = C H 2 + H + It has a lso b e e n found that the r a c e m i z a t i o n s of 2iodooctane (ll) and of p h e n y l m e t h y l c a r b i n y l b r o m i d e (1 2 ) in acetone solution are c a t alyzed b y iodide and b r o m i d e ions - 7 - respectively. The situation obtaining in this instance is somewhat different than that of the preceeding spontaneous racemization and therefore, will probably involve a different mechanism. It has been shown, quite conclusively, (13, 14) that substitution at an asymmetric carbon is always accompanied by an inversion of optical configuration. The main conclusion is that the negative ion approaches the carbon-halogen dipole at its positive end, leading to an inverted product, or, in the case of a reversible reaction such as substitution by like ions, to an ultimately racemic product. This hypothesis has been tested experimentally for the above racemizations by employing radioactive halide ions and measuring the rate at which the alkyl halide acquires radioactivity as well as the rate of racemization. Thus, in the case of the racemization of phenylmethylcarbinyl bromide by bromide ions (lithium b r o  mide) in acetone, (1 2 ), Br" + (+)-C6H5 -CHBr-CH3 -- > ( - ) -C6H 5 -CHBr-CH3 + Br" the rate of radioactive exchange should be Just one-half that of the rate of racemization, a condition which has been demonstrated experimentally within the limits of accuracy of the experimental methods. The reaction of a halide ion with an alkyl halide is, therefore, a bimolecular process involving inversion of optical configuration, which, because of the r e  versibility of the reaction leads to ultimate racemization. A somewhat similar racemization of phenylmethylcarbinyl bromide is effected b y silver bromide and molecular bromine in - 8 - carbon tetrachloride solution (1 5 ). If the alkyl halide is refluxed for one hour in the presence of bromine (0.7 equiva lents) and excess silver bromide, the recovered alkyl bromide is 99.8$ racemized. In the presence of silver bromide alone, refluxing for three hours produces only 0 .6$ racemization. In the presence of bromine alone (2.0 equivalents) three hours refluxing produces alkyl halide which is 38.4$ racemized, while nineteen hours refluxing results in 99.0$ racemization. exact nature of this racemization is not clear. The It m a y be due to a bimolecular halogen exchange such as that discussed above or It m ay be the result of a free radical reaction. RBr + B r ^ R* + B r * =>R • + Br • + Br^ >RBr A variety of metallic chlorides(HgCl2 , ZnCl^, SnCl^, BCl^j TiCl^, SbCl^), whose affinity for halide ion is d e mon strated by the complex halide ions that they readily form, racemize phenylmethylcarbinyl chloride (16). The reaction was investigated in a number of solvents and appeared to be markedly dependent upon the solvent as well as the metallic halide. The rate of racemization increases as the dielectric constant of the solvent increases. Indeed, w i t h formic acid (D = 5 9 )> the racemization proceeds rapidly without a metallic halide catalyst and is complete in about ten minutes at 25°; the addition of mercuric chloride is without effect. When mercuric chloride is employed as the metallic halide, the racemization is about fifty times slower in acetone than in - 9 - nitromethane, and about one thousand times slower In ether than in nitromethane. Lithium chloride, hydrogen chloride or tetramethylammonium chloride cause very little racemiza tion in the absence of one of the metallic halides noted above, but retard the racemization when they are added in equivalent amounts to a solution containing the alkyl halide and mercuric chloride. From these facts it is concluded that the racemization is not a bimolecular reaction involving h a l  ide ions, since mercuric chloride is about fifty times as effective as the three chlorides above and yet it is the least ionized one of the four. Instead it is probably due to the formation of a complex with the metallic halide, the formation of which is retarded in the presence of chloride ions, which readily undergo complex formation themselves. These racemizations, then, must involve an attack on the halogen by the metallic halide and are, in all probability, polymolecular, since the rates vary as a higher power (1.3 to 14) of the metallic halide concentration. - 10 - EXPERIMENTAL A. Preparation and Purification of Starting Materials 1. Resolution of dl-Phenylmethyl Carbinol 0 a 0 CH' if -C-O-CH-CgH^ . /%.-CH-0H Pyriding -C00H GH3 0 b . 0 CH' 0 CH3 »» » -C-OCH-C6H5 it Ns. -C-O-CH-CgH^ -C00H + Brucine -C00“Bru+ (+) (-) and (-) (-) 0 CHo 0 CHo ii i » t 0 ^ - C - 0 C H - C 6H5+ h o HgSOi^ jj^\|-C-OCH-C6H5 -C00“Bru 0 •COOH CH' ch3 it N^-C-OCH-CgH^ -COOH ^ ^ NaOI^ + H^O ' % v-C00~Na+ + f/^-CH-OH I -C00'Na+ (+) or (-) - 11 - a • Preparation of a-Phenylethyl Hydrogen Phthalate.-This ester was prepared by a modification of the procedure of Houssa and Kenyon (17). In a typical experiment, 296 g. (2 moles) of phthalic anhydride, 244 g. methyl carbinol and 158 g. (2 moles) of dl-phenyl- (2 moles) of anhydrous p y r i d i n e were placed in a three liter flask. (C.P.) The mixture was heated with stirring on the steam bath until it was homogeneous (ca. 1 hr.). At the end of two hours, the reaction was assumed to be com plete and the contents of the flask were poured with stirring into an excess (3 moles) of concentrated hydrochloric acid in ice. The white taffy-like substance which separated became oily at room temperature. The water layer was decanted and extracted several times with ether. The ether extracts were combined with the crude half-ester and most of the ether was removed by distillation. The residue was then extracted with 10$ sodium carbonate solution in several portions. The c om bined alkaline extracts were acidified to Congo Red with h y d r o  chloric acid and extracted with several portions of chloroform. The chloroform solution of the half-ester thus obtained was warmed on the steam bath to remove most of the chloroform. The residue, upon standing overnight, deposited crystals of a-phenylethyl hydrogen phthalate, m.p. 106-107°. Additional quantities of this half-ester were obtained from the mother liquors of this and succeeding crystallizations so that the total yield of a-phenylethyl hydrogen phthalate was usually between 75 and 85$ of the t h e o r etical. - 12 - b . Preparation of the Brucine Salt of a-Phenylethyl Hydrogen Phthalate (l8 ) .--In a typical experiment, 394 g. (l mole) of anhydrous brucine was added portionwise to a solution of 270 g. (l mole) of a-phenylethyl hydrogen phthalate in 1000 c c . of acetone. After standing a few minutes, a thick slurry of the brucine salt formed. The solution was heated to boiling and kept at the boiling point of acetone for about 15 m i nutes. The hot solution was filtered and the crystals of the brucine salt were washed with hot acetone. The filtrate deposited additional crystals upon cooling and these were com bined with the first crop. This process was repeated until 27^ g • (83^) of the levo-brucine a-phenylethyl hydrogen phthalate, m.p. 147-150°, had been obtained. The dextro-brucine a-phenyl ethyl hydrogen phthalate in the acetone mother liquors was not further isolated or p u r i f i e d . c . Hydrolysis of the Brucine Salts of a-Phenylethyl Hydrogen Phthalate.--A slurry of 200 g. (p.3 mole) of levo- brucine a-phenylethyl hydrogen phthalate in 7°0 ml. of carbon disulfide was treated with 1200 ml. of 2 N sulfuric acid. The mixture was warmed on the steam bath with stirring until com plete solution of the brucine salt was effected. The carbon disulfide layer was separated, washed once with water and dried with sodium sulfate. The carbon disulfide was removed by d i s  tillation and the levo-a-phenylethyl hydrogen phthalate thus obtained was saponified without further purification. - 13 - The acetone solution of the impure dextro-brucine <xphenylethyl hydrogen phthalate was warmed to remove the acetone and the syrupy brucine salt was hydrolyzed in a manner identical to that described above for the levo-salt. d. Saponification of Optically Active a-Phenylet Hydrogen Phthalate.--An excess of 5 N sodium hydroxide was added to the levo-a-phenylethyl hydrogen phthalate obtained in (c) and the mixture steam-distilled. The distillate was salted out, extracted with ether and dried with sodium sulfate. Distillation yielded levo-phenylmethyl carbinol, b . p .25 92 9 2 .5°^ ctp -25° to -39° (from various experiments) in 65$ yield from the brucine salt. The dextro-phenylmethyl carbinol, b . p .25 102-104°, a-Q 2 0 .6° to 3 1 .6 °, was obtained in an identical manner. 2. The Conversion of Phenylmethyl Carbinol to Phenylmethyl carbinyl Chloride Method 1 .--Following the procedure of Gerrard (1 9 ), a solution of 10.4 g. (0.05 mole) of phosphorus pentachloride in 200 m l . of carbon disulfide was added dropwise to a solution of 12.2 g. (0.1 mole) of dextro-phenylmethyl carbinol (cc^° + o 2 6 .2 , 1 = 1 , homogeneous) and 32 g. (0.4 mole) of anhydrous pyridine in 20 m l . of carbon disulfide at - 1 0 ° C . The addition required about 30 minutes and proceeded with the formation of a white precip i t a t e . 16 hours. The reaction mixture was kept at 15° for The white precipitate was then filtered by gravity, and the filtrate was washed successively with dilute ice-cold - 14 - sulfuric acid, dilute sodium carbonate solution and water. After drying with sodium sulfate, the carbon disulfide was stripped. Distillation of the residue yielded about 6 g. 25 (43$) of phenylmethylcarbinyl chloride, b . p .25 74-76°, <xD -41.3°. Method 2 .--To 9..1 ml. (0.125 moles) of thionyl chloride (Reagent Grade) in a 25 m l . round-bottomed flask was added dropwise 12.2 g. (0.1 mole) of levo-phenylmethyl carbinol (a^° -38.7 ° t 1 = 1 , homogeneous) at room temperature. After standing for 15 minutes, the mixture was distilled to yield 12.2 g. (87$) of levo-phenylmethylcarbinyl chloride, b . p .30 79-83°, -37.7° (l = 1, homogeneous). 3. The Preparation of Mesitol (2,4,6-Trimethylphenol) NO 2 CH^- -CH + in glacial HOAc t ch3 n o2 b nh2 t . CH + Sn + HC1 1 ch3 glacial HOAc CHo-^ s-CH 3 ch3- -CH3 + NaN02 + H C 1 CH CHq J -c h 3 y ch3 3 OH -CH 3 3 70° ch3 a . Conversion of Mesitylene to Nitromesitylene. This was carried out following the procedure of Powell and Johnson (20) as modified by Hart (21) except that six times the quantities listed were used. An 80$ yield of nitro mesitylene, m.p. 43-44°, was realized. b . The Reduction of Nitromesitylene to Mesidine (Aminomesitylene).--To a stirred mixture of 208 g. (1.25 moles) of nitromesitylene, 238 g. (2.0 moles) of tin turn ings and 300 ml. glacial acetic acid was added 800 ml. of concentrated hydrochloric acid in small portions with icebath cooling. The addition was complete in about 45 minutes. The reaction mixture was then heated on the steam bath for one hour. A solution of 600 g. of sodium hydroxide in 1500 ml. water was added and the mixture steam-distilled. The organic layer in the distillate was extracted from the water with ether and dried. The ether was stripped and the residue dis tilled to give 137 g. (81$) of mesidine, b . p . ^ 0 220-224°. - 16 - c. Conversion of Mesidine to Mesitol.--A solutio of 360 ml. concentrated hydrochloric acid, 225 ml. water and 137 g. mesidine was treated at 0 to 5° with a solution of 72 g. of sodium nitrite in 450 ml. of water. The excess of nitrous acid was destroyed by portionwise addition of urea until starch iodide paper gave no positive test. The solu tion of the diazonium salt was then warmed on the steam bath until no more nitrogen was evolved. The crude mesitol was extracted with ether and the ether removed by distillation. A final distillation gave mesitol, b . p . ^ o 208-210°, in 85^ yield. It was crystallized from petroleum ether to give white needles, m.p. 68-690 . It should be noted that when the excess of nitrous acid is not destroyed with urea, con siderable tar formation results and a correspondingly lower yield of mesitol is obtained. 4. The Preparation of a-Phenylethyl Derivatives of o-Cresol, p-Cresol and 2 ,6 -Xylenol 4-q-Phenylethyl-2-Methylphenol.--Equimolar amounts of o-cresol and dl-phenylmethylcarbinyl chloride were warmed on the steam bath in a 50 m l . round-bottomed flask fitted with a reflux condenser and drying tube for about two hours. The mixture was fractionally distilled at 25 mm. to remove any unreacted starting materials. The a-phenylethyl-2-methyl- phenol fraction was doubly distilled from a 10 m l . Wurtz flask and was collected as a water-white, viscous liquid, b . p . g l 581^3°, yield 50$. (22 ). The reported boiling point is 179° at 13 mm. - 17 - 4 -Methyl-2-(co-phenylethyl-)phenol.--This was prepared in a manner Identical to that just described and was also a water-white, viscous liquid, b.p.^ 165-170°, yield 60%. The reported boiling point is 175-177° at 13 mm. (22) . 4 - (a-Phenylethyl)-2,6-Dimethylphenol.--This compound was prepared by a procedure identical to the above except that the reaction mixture was warmed on the steam bath for 5 hours. The triply-distilled product, b.p.-j^g 143-145°, was a viscous, very pale yellow liquid, yield about 50#. cl6Hl8° : c * 84.8; H, 8.0. Pound: Anal. Calc'd. for C, 84.8; H, 8.4. 5 . Purification of Commercial Starting Materials Phenol.--C.P. grade phenol was triply distilled, the final distillation being carried out in a nitrogen atmos phere. The boiling range of the phenol used in the kinetic studies was 179-180° at atmospheric pressure. o-Cresol.--Water-white, C.P. grade o-cresol (Pisher Scientific Co.) was used without further purification. p-Cresol.--Practical grade g-cresol was triply distilled to give a water-white product, b.p. 199-200°, at atmospheric pressure. 2,6-Xylenol.--Edcan Laboratories1 2,6-xylenol was triply distilled in a nitrogen atmosphere to give water-white 2 ,6 -xylenol, b.p. 208-210° at atmospheric pressure which so lidified to white crystals upon cooling, m.p. 48-49°. - 18 - Solvents.--The benzene used In the rate studies was C.P. thiophene-free benzene, distilled from sodium. The p-xylene was C.P. p-xylene, distilled from and stored over sodium. The methanol used in the analytical determinations was Merck and Co. absolute methanol, C.P., acetone-free, and was used without further purification. B. Measurement of Alkylation Rates 1 • Standardization of Solutions The bromate-bromide solution used was prepared following the directions of Ruderman (23). It contained 5-6 g. of potassium bromate and 30 g. of potassium bromide in a volume of 2 liters. A 25 ml. aliquot of this solution was diluted with an equal volume of water and acidified with 5 ml. of concentrated hydrochloric acid. Ten ml. of a 10 per cent potassium iodide solution was added and the liberated iodine was titrated to the disappearance of the iodine color with standard sodium thiosulfate solution. Dilution with acetone- free methanol instead of water required the same titer of sodium thiosulfate indicating that methanol does not consume bromine under these conditions. The sodium thiosulfate solution was standardized against a weighed sample of potassium iodate using the standard analyti cal procedure (24). - 19 - 2. Analytical Procedure Phenol, p-Cresol and o-Cresol.— Stock solutions of these compounds and their respective a-phenylethyl sub stituted derivatives were prepared by dissolving a weighed sample in methanol and making the solution up to volume with 60$ aqueous m e t h a n o l . These solutions were usually about .04 M. The general procedure for determination of these phenols was as follows: An appropriate aliquot of the stock solution (or solutions in the case of mixtures) was pipetted into a 250 ml. iodine flask. A calculated excess of standard bromate- bromide solution was added, together with sufficient distilled water or methanol to make the combined volume about 40 to 50 ml. Five ml. of concentrated hydrochloric acid was then added, the flask quickly stoppered and swirled for the desired reaction time. The bromine excess was then destroyed with 10 ml. of a 10 per cent potassium iodide, solution. The iodine liberated was titrated within 30 seconds with standard sodium thiosulfate solution to the disappearance of the iodine color, or to a starch endpoint. The results are tabulated in Tables 1, 2 and 3 . 2,6-Xylenol.--Stock solutions of this compound and 4-a-phenylethyl-2,6-dimethylphenol were prepared as Just d e  scribed for the other phenols. The analytical procedure was identical with that described above except that the solution - 20 - was allowed to stand for 5 minutes after addition of the potassium iodide before titrating the liberated iodine with the standard sodium thiosulfate solution to a starch endpoint. The results are tabulated in Table 4. 3. Rate Studies a. General Procedure for Determination of the Ra of Alkylation of Phenol by Phenylmethylcarbinyl Chloride.--A preliminary qualitative experiment showed that a reaction m i x  ture approximately 25 mole $ in each reactant and 50 mole $ in benzene would give a measurable rate of reaction at 50O C . Accordingly, in a first series of determinations of the rate of alkylation, stock solutions of this concentration were p r e  pared. To 31.4-98 g. of phenol was added 30.52 ml. of dry thiophene-free benzene. (26.14 g.) The density of this solution, d^0 =0.9740, was determined using a specific gravity balance calibrated to read density directly. The solution was thus 5.65 M in phenol. In a similar manner, 44.912 g. of doubly distilled phenyl methylcarbinyl chloride was mixed with 31.37 ml. of benzene to give a solution of d^Q-^.9789 and molarity 4.36. In subsequent determinations of alkylation rates the stock solutions were prepared by dissolving a weighed amount of each reactant in £-xylene and making up to a definite volume so that the concentration of the stock solutions was exactly 5.00 M. - 21 - The reaction was carried out in sealed glass ampoules constructed from 10 inch sections of 25 mm. o.d. Pyrex tubing constricted at one end to 8 mm. for each kinetic point. in each run. A separate ampoule was used At least seven points were obtained Frequently, duplicate samples for a point were taken. Each ampoule, previously cleaned and dried overnight at 110° was loaded with 2 m l . of the phenol stock solution and chilled in a Dry Ice-acetone bath. After thorough freezing, 2 m l . of the phenylmethylcarbinyl chloride stock solution was added and the ampoule and its contents again immediately placed in the Dry Ice-acetone bath. During the cooling, the constricted opening of the ampoule was protected from foreign substances by plugging it with a short section of tight-fitting rubber tubing fitted with a clean, tight-fitting cork. The ampoules were sealed in va cuo, placed in a wire test tube rack and lowered into a constant temperature bath maintained at the desired temperature + 0.1°. The buoyant effect of the water in the bath was overcome by placing a short section of 1/4" pipe over the neck of each a m p o u l e . The time at which the ampoules were lowered into the bath was taken as zero t i m e . Each ampoule was shaken within 3 minutes after it had been placed in the bath to insure homogenity of its contents. After the desired reaction time, each ampoule was removed from the bath and immediately cooled in the Dry Ice-acetone bath. At a convenient later time, the ampoule was opened, the contents immediately diluted with - 22 - 10 m l . of benzene and transferred to a 100 m l . separatory funnel. The ampoule was rinsed with two 5 nil. portions of benzene and these were added to the contents of the separa tory funnel. The phenols present were then extracted with three successive 10 ml. portions of 5 N sodium hydroxide. Each extraction was accomplished by shaking for 15 seconds and allowing the phases to separate for 2 minutes, 1.5 minutes and 1 minute, respectively. The rafflnate phase was washed with two 10 ml. portions of distilled water and these were combined with the alkaline extrahend. The effec tiveness of this alkali extraction was demonstrated using known amounts of the p h e n o l s . The raffinate phase in these cases contained nothing but benzene after extraction. The alkaline solution was made acid to phenolphth&lein with con centrated hydrochloric acid and diluted to 250 ml. with a 60$ aqueous methanol solution. A 5 m l . aliquot of this solution was then analyzed in the manner described previously. The bromination time was 25-30 seconds. In order to determine the loss on extraction, a blank determination was carried out. Two ml. of the phenol stock solution was placed in an ampoule and treated in the same manner as an authentic kinetic sample as regards extraction with alkali and dilution with 6 C$ aqueous methanol. An a l i  quot was analyzed by the standard procedure described above and the value obtained was used as the amount of phenol present initially. - 23 - In a first series of experiments, 2.00 ml. of 5.65 M phenol solution and 2.4 ml. of 4.36 M phenylmethylcarbinyl chloride solution were the quantities used in each ampoule. The remainder of the procedure was unchanged. The results of the experiments with phenol are given in Tables 5* 6 , 7 and 8 . b. General Procedure for the Determination of th Rate of Alkylation of o- and p-Cresol by Phenylmethylcarbinyl Chloride.--Stock solutions of the cresols in benzene were prepared by dissolving a definite weight of the cresol in benzene and diluting to a definite volume with benzene so that the final concentration was 5.00 M. Two ml. of the 5.00 M cresol solution and 2 m l . of 5 .00 M phenylmethyl carbinyl chloride were placed in each ampoule using the technique just described for phenol. The ampoules were sealed at atmospheric pressure and placed in the constant temperature bath as previously described. At the desired time, each ampoule was removed from the bath and placed in the Dry Ice-acetone bath. At a convenient later time the contents were diluted with 10 ml. of methanol and transferred to a 250 ml. volumetric flask together with two 5 ml. rinse portions of met h a n o l . The solution was made up to the mark with 60$ aqueous methanol and a 5 m l . aliquot was •oaken for analysis. The bromination time in the o-cresol samples was 15 seconds and for tne p-cresol samples it was 25-33 seconds. The results are given in Tables 9 and 10. Blank determinations - 24 - on the benzene and phenylmethylcarbinyl chloride indicated that these compounds absorbed no bromine under the conditions employed in the analysis. c. General Procedure for the Determination of the Rate of Alkylatlon of 2,6-Dimethylphenol by Phenylmethyl carbinyl Chloride.--A 5.00 M solution of 2,6-dimethylphenol in benzene was prepared by dissolving a weighed sample of 2 ,6 -dimethylphenol in benzene and diluting to the proper volume. The general procedure used for the cresols was followed. Since the analytical determinations on known samples indicated considerable overbromination in the case of the 4 - (a-phenylethyl)-2 ,6 -dimethylphenol under varying conditions, it was necessary to employ an empirical method of analysis. Mixtures of known composition were subjected to bromination for a definite time and the milliequivalents of bromine consumed were plotted vs. the composition of the mixture to give a calibration curve. The analysis of the kinetic samples was then carried out in an identical manner, the per cent alkylation being determined from the calibration curve. C. Measurement of Racemization Rates 1. Blank Determination A 2.50 M solution of (+) phenylmethylcarbinyl chloride in p-xylene was placed in a 1 dm. thermostatted polarimeter tube which was kept at 50° + 0.1 by circulating water from a constant - 25 - temperature bath through it. The initial rotation was 5.82 +0.02° and this value did not change for thirty-six hours. 2. General Procedure for the Determination of the Racemization Rate of Optically Active Phenylmethylcarbinyl Chloride by Phenol, o-Cresol, p-Cresol, 2,6-Xylenol and Mesitol In all cases except with mesitol, 5 m l . of 5.00 M (+) phenylmethylcarbinyl chloride in £>-xylene and 5 m l . of a 5.00 M solution of the appropriate phenol in g-xylene were preheated at 50°, mixed and quickly poured into a thermostatted polarimeter tube through which water at the desired temperature +0.1° had been circulating for at least fifteen minutes. The time at which the solutions were mixed was taken as zero time. intervals. The rotation was recorded at the desired time In the case of mesitol, 3 m l . of 5.00 M (+) phenyl methylcarbinyl chloride solution were placed in a thermostatted polarimeter tube at 50° +0.1° followed by 3 m l . of a 5.00 M mesitol solution. Duplicate determinations were carried out in polarimeter tubes of slightly different capacity, thus one tube required slightly more than 3 m l . of 5-00 M mesitol solu tion for filling. D. Isolation of Optically Active a-Phenylethylphenols In an ampoule, constructed from a 200 ml. round-bottomed flask by constricting the neck, were placed 20.59 ml. of 5.65 M phenol stock solution (in benzene) and 24.8 ml. of 4.36 M (-) - 26 - phenylmethylcarbinyl chloride stock solution. The rotation 20 of the (-) phenylmethylcarbinyl chloride used was aD -47.7° (1 = 1 , homogeneous). The techniques used in loading and sealing the ampoule were identical to those employed in the alkylation rate studies. The ampoule was placed in a constant temperature bath at 50° + 0 .1°C. for 30 minutes. It was then chilled in a Dry Ice-acetone bath, opened and the contents diluted with 120 m l . of b e n z e n e . After transfer to a separa tory funnel, this solution was extracted with a total of 350 ml. of 5 N sodium hydroxide in several p o r t i o n s . The raffinate phase was washed with one 50 ml. and two 100 ml. portions of water which were combined with the alkaline e x t r a h e n d . This aqueous alkaline solution was acidified to Congo Red with 6 N hydrochloric acid and the phenols extracted with 50 ml. of benzene in two portions. After drying with sodium sulfate, the benzene was stripped and the phenols distilled to give 4.2 g. of phenol and 2.3 g. of alkylated phenols. The latter was doubly distilled from a 10 m l . Wurtz flask to yield a water-white viscous liquid, b.p.^ 150-160°, <xD -.75 +0.05° (l = 1 , homogeneous). The raffinate phase yielded 11.0 g. of (-) phenylmethylcarbinyl chloride, a.^ -21.1° (1 = 1, h o m o  geneous) upon distillation. Using an identical procedure to the one just described, 17.7 ml. of 5.65 M phenol solution and 21.3 ml. of 4.36 M (-) phenylmethylcarbinyl chloride solution were heated in a sealed ampoule at 50 +0. 1°C. for five hours. The extraction - 27 - of the phenols was accomplished with four 100 ml. portions of 5 N sodium hydroxide and two 100 ml. portions of water. The alkaline extract was acidified to Congo Red with 6 N hydrochloric acid, the phenols were extracted with benzene, dried and distilled as described above. The alkylated phenols, wt. 5 g.* were triply distilled from a 10 ml. Wurtz flask to yield a water-white, viscous liquid, + 0.05° (1 = 1, homogeneous). 152-155°* a j) -»25° - 28 - RESULTS The rate of alkylation of phenol, £-cresol, o-cresol and 2,6-xylenol by phenylmethylcarbinyl chloride was d e  termined . The reaction rate In each case was followed by product analysis. The principle of the method used In f ol lowing the reaction depends on the fact that the alkylated phenol formed as a reaction product must always have one less unsubstituted ortho or para position than the phenol used as a reactant. It Is well-known that phenols will undergo nuclear bromlnation only at the positions ortho or para to the hydroxyl group. Thus, while phenol Itself can be quantitatively bromlnated to yield 2,4,6-tribromophenol, a-phenylethylphenol, whether the a-phenylethyl group be ortho or para, has only two ring positions available for bromlnation. Consequently, it is possible to devise an analytical method for following the alkylation process based on this fact. The method used in this investigation consisted of adding a known excess of standard potassium bromide-potassium bromate solution to a phenol mixture, acidifying, allowing the free bromine to react for a definite time and then titrating the excess bromineAwith standard sodium thiosulfate solution. The amount of bromine consumed is a measure of the relative amount of the phenol and a-phenylethyl phenol present while the total amount of both phenols is equal to the quantity of the phenol originally present. Thus, a pair of - 29 - simultaneous equations m a y be set up and evaluated. For example, in the case of phenol itself, the following equa tions were u s e d . x + y = total moles of phenol originally present 6x + 4y = equivalents of Br2 consumed (l) where x = moles of phenol at time t y = moles of a-phenylethylphenol at time t Similar pairs of equations were formulated for the other phenols s t udi e d . In order to test the validity of this analytical approach, samples of the phenol, its a,-phenylethyl derivative and repre sentative mixtures of the two were analyzed by this m e t h o d . The results are given in Tables 1 to 4. It should be pointed out, that, under the conditions employed in the measurement of the alkylation rates, the amount of dialkylation is negligible until the reaction is more than 7°^ complete. It was necessary to develop rather exacting experimental conditions for each phenol mixture in order to obtain r e pro ducible results. Several factors effect the determination of phenol mixtures b y b r o m l n a t i o n . The bromine excess and time of reaction with the bromine must be carefully controlled or overbromination m a y occur. The overbromination is probably largely a result of side-chain bromlnation which apparently occurs to a much greater extent when primary alkyl groups are attached to the ring ortho to the hydroxyl g r o u p . vestigation, a constant In this in (within 5$) bromine excess was maintained - 30 - and the reaction time varied until reproducible results within one to two per cent of the calculated value were o b t a i n ed. The samples for kinetic points were analyzed using the conditions established with the known s a m p l e s . As can be seen from an inspection of Tables 1 and 2, the conditions for the analysis of mixtures of phenol and a-phenylethylphenols and mixtures of n-cresol and o-a-phenylethyl-j>cresol are quite similar. Reproducible analyses for mixtures of o-cresol and its a-phenylethyl derivative required a shorter reaction time to avoid overbromination. (Table 3 , Sample 4) The method as an absolute one failed for 2,6-dimethylphenol and its a-phenylethyl derivative because of extensive overbromina tion in the 4-a-phenylethyl-2,6-dimethylphenol. This compound, which has no ortho or para positions available for nuclear bromination, nevertheless consumed considerable bromine even with very short reaction times 8). (Table 4, Sample Nos. 6, 7 and Thus, even though the 2,6-dimethylphenol could be determined with good reproducibility, (Table 4, Sample Nos. 2, 4 and 5) the method was not amenable to the analysis of mixtures. An empirical method was developed in which the reaction time was arbil^arily set at twenty seconds and in which a definite bromine excess was maintained. A calibration curve in which percent of 4 -a-phenylethyl-2,6-dimethylphenol present in the mixture (Column 8, Table 4) was plotted versus the milliequivalents of bromine consumed (Column 6, Table 4) was used to determine the extent of the alkylation in the kinetic studies. - 31 - The rate of alkylation of the various phenols by phenylmethylcarbinyl chloride is given in Tables 5 to 11. The data are represented satisfactorily by the conventional second order rate expression. - ka (PhOH) ex dt (RC1) or (2) â– ^ Rate constants were calculated using the integrated form of this expression, for the case in which the initial con centrations of the reactants are equal, vis: k a = (?.) 'DJ i k __2__ .. r \ t a(a-x) The symbols are defined in Table 12. In one case, (Table 6) the initial concentrations of the phenol and phenylmethylcarbinyl chloride were not e q u a l . Here, the second-order rate constants were evaluated by using the integrated second-order expression . 2.303 .. a - log b(a-x) ^(^0 (4) where b = initial concentration of phenylmethylcarbinyl chloride and the other symbols are defined in Table 12. Optically active phenylmethylcarbinyl chloride was found to undergo racemization in the presence of phenol, p-cresol, o-cresol, 2,6-dimethylphenol and mesitol. One mechanism was found to represent the experimental data satisfactorily for all cases. A slight modification was 32 - necessary in case of 2,6-dimethylphenol and mesitol since in these cases, alkylation of the phenol is very slow and non-existent, respectively. These cases will he treated separately below. The kinetic data are represented satisfactorily by the following differential equation. _ dt = k p(R*Cl) (PhOH) + k fl p (R*C1) (A-PhOH) r H.r. + kH C 1 (R*Cl)(HC1) + ka (R*Cl)(PhOH) or (5) - ^ = k p (a)(a-x) + kA > P .(a ) U ) + kH C 1 (a)(x) Q.T3 * * + ka (a)(a-x) The symbols used are defined in Table 12. By combining like quantities, equation (5) can be simplified to ' 5t = (kP + ka )(a)(a-x) + (kA.F. + Since neither kA p nor kjjC1 is known from an independent source, the method does not permit the evaluation of each of these constants, but rather only the sum of two, which is represented by k lev . The fact that two constants, rather than one, are necessary, will be shown below. The quantity, x, can be found in terms of known quantities from equation (3). It is given by x = a.2kat 1 + akat (7) Substituting this value for x into equation (6) gives _ da. = ( dt p k jJallaL. + (a)(agk a t) a 1 + akat ^ 1 + akat (8) - 33 - Separating variables and integrating gives -In a = kp + k„ lCy ln(l + akat) + ~ (l + akat) (9) kx ln(l + aka t) + C ka which can be simplified to kp -f* ko — kv ^-y* -In a = ----- r------- ln(l + akat) + -— (l + akat) + C Ka ka When t = 0 , ct= a 0 , then C = -In a Q - ^ . ka (1 0 ) Thus, an In k P + k« - k-jf --- rf----s ln(l + akat) + ak^t (ll) Dividing both sides b y t and converting to common log arithms gives log o.0/a t kp + ka - kx log(l + akat) " ka t akx + 2.303 Equation (12) is in the form of a straight line, Y = c + dx, with the slope, d, equal to (kp + ka - k x ) and the intercept, c, equal to alcx . Since a is a constant and ka is known 2.303 from independent measurements (see above), it is possible to evaluate kp and k^. by plotting the quantity (log aQ/a/t) versus the quantity [log(l + a k at)/t] and determining the slope and intercept of the resulting straight l i n e . For simplicity, we have let Y = log aQ/a/t and X = [log(l + akat)/t]. The best straight line through the experimental points was determined by the method of least squares. Figure 1 shows a typical straight line obtained for the racemization of - 34 - Figure 1. Y vs. X fqr the Racemlzatlon of Phenylmethyloarblnyl Chloride In Phenol at 50°. 10.2 10.0 9.8 Y x 103 9.6 9.2 9.0 8.8 8.6 3.5 3.6 3.8 3.7 X x 103 3.9 4.0 - 35 - phenylmethylcarbinyl chloride in the presence of phenol at 50°. Only every other experimental point is shown for the sake of clarity in the drawing. Tables 13 to 16 give the pertinent data for the racemization of phenylmethylcarbinyl chloride in phenol, p-cresol and o-cresol. In the case of mesitol, equation (5) becomes (13) since there is no alkylation possible. Equation (13) may he integrated to g i v e : log a Q/a = akpt/2.303 (14) Here a plot of log a 0/a v s . t should give a straight line with a slope equal to akp/2.303. The data for a typical run are plotted in Figure 2 and tabulated in Table 17- The best straight line was determined by the method of least squares. It should be pointed out that although equation (l4) is typical for a first order reaction, kp, nevertheless, has the units of 1. m o l e -1 min."1 , and it is thus consistent dimensionally with the kp obtained from equation (5). Since the rate of alkylation in 2,6-dimethylphenol is extremely slow (ka = 1.1 x 10"5)^ it is possible to modify equation (5) for the racemization of phenylmethylcarbinyl chloride in 2,6-dimethylphenol. If it be assumed that a-x is essentially equal to a and x is equal to zero for the time interval over which the experimental data were obtained, then equation (5) becomes (15) Figure 2. The Racemization of Phenylmethylcarbinyl Chloride in the Presence of Mesitol or 2,6-Dimethylphenol at 50 .09 .08 .07 .06 rH .04 • - 2,6-Dimethylphenol o - Mesitol .03 2,6-Dimethylphenol k » I .96 x 10“5i. mole"1 min."1. .02 Mesitol 1.77 x 10"5i. mole"1 mi n ."1 .01 0 1000 2000 4000 Time, Minutes 3000 5000 - 37 - Integration of equation (15) gives log a Q/a = (kp + ka )at/2.303 (l6 ) Thus, a plot of log a Q/a vs. t should give a straight line with a slope equal to a(kp + ka )/2.303. in Figure 2. The best straight line was determined by the method of least squares. Table Y J . Such a plot is shown The experimental data are given in It should be pointed out that the rate constant thus obtained for the racemization in 2 ,6 -dimethylphenol deviates from the true value by a small indeterminate quantity due to the error introduced by neglecting the small amount of alkylation that does occur during the time interval involved. The activation energies for the alkylation of phenol by phenylmethylcarbinyl chloride and for the racemization of op tically active phenylmethylcarbinyl chloride in phenol have been determined over the temperature range 30 to 50° C . by means of the Arrhenius equation k = A e _AEe x p . ( 1 7 ) which may be expressed in its equivalent form log k = -AEe x p y 2 . 3 0 3 HT + log A (l8 ) A plot of the logarithm of the reaction rate constant versus the reciprocal of the absolute temperature is a straight line, the slope of which is equal to -AEexp /2.303 Rj and whose inter cept is equal to log A. readily evaluated. The energy of activation can thus be For the reactions under consideration, the best straight line through the three experimental points was determined by the method of least squares. - 38 - The entropies of activation for the same reactions were determined from the equation A = e â„¢ in which K is A h e (19) K R is the frequency factor (based on collision the Boltzmann constant theory), (I.38O x 10_1^ergs per degree), h is Planck's constant (6.624 x 1 0 " ^ erg-sec.) and R is the gas law constant (1.987 c a l . per degree per mole). This equation can be written in its equivalent form log A = ASa/ 2 .303 R + 1/2.303 + log KT/h (20) Equation (20) permits the evaluation of the entropy of activation since log A is the intercept of the straight line plotted for the evaluation of the experimental energy of activation and all the other terms are known. were obtained for a temperature of 4 0 ° C . given in Table 18. The values The results are - 39 - DISCUSSION OP RESULTS It has been shown that the alkylation of phenols by optically active phenylmethylcarbinyl chloride is accom panied by simultaneous racemization of the chloride. It was of Interest to examine these reactions in some detail in an attempt to arrive at a reasonable theoretical ex planation for the experimental observations. Accordingly, the alkylation of phenol, o-cresol, p-cresol and 2 ,6 -di methylphenol by phenylmethylcarbinyl chloride and the racemization of optically active phenylmethylcarbinyl chloride in these phenols and mesitol have been subjected to kinetic analysis. The alkylation process was followed by product analysis, employing a bromination technique which has already been dis cussed. (cf. Results) It was necessary to employ rather concentrated solutions (2.50 M) of the reactants in order to obtain measurable rates for all the cases investigated. Since -equal initial concentrations of both reactants were used in all cases and the initial concentrations were kept constant, the data permit only an evaluation of the overall order of the reaction. As has already been demonstrated, the experimental data in every case fit overall second order ki  netics, i.e., the reaction rate is proportional to the concen tration of both reactants. However, a second reaction path - 4o - appeared plausible and was, therefore, tested. The reaction may be catalyzed by the hydrogen chloride evolved. If this were the case, one would expect overall third order kinetics as indicated by the following equation: _ . d (. P. h. Q.Hj, = k a o(PhOH) (HC1 ) (HCl) dt 3 or / ^ . <5la=£l = k a ,(a-x)2x dt 3 The symbols used are defined in Table 12. variables and integrating, Upon separating (l) becomes: 2.303/a2 (log x/a-x) + log l/a(a-x) = ka ^t + C (2 ) If the reaction exhibits third order kinetics of this type, then a plot of the left-hand side of equation (2 ) versus time should yield a straight line whose slope is equal to the third order reaction rate constant, k g ^ . Such a plot for the case of the alkylation of phenol by phenylmethyl carbinyl chloride at 50° is shown in Figure 3. It is evident from this that the data do not fit a kinetic expression which incorporates hydrogen chloride catalysis. It should be pointed out, however, that this possibility is not ruled out entirely for this reaction. As has already been noted, the experimental data are obtained from rather concentrated solutions and it would thus be difficult to detect autocatalysis if it were present. Indeed, it seems likely that such autocatalysis does take p l a c e . It has been detected in the alkylation of phenol and o-cresol by trityl chloride in dilute solutions (25 ) in which case, the rate of reaction was followed by a - 41 - (2.303 log x/a-x)/a2 + l/a(a-x) Figure 3. Test of A Third Order Mechanism for the Alkylation of Phenol by Phenylmethylcarbinyl Chloride at 50°. 0 20 40 60 80 Time, Minutes 100 120 - 42 - much more precise method than the one used in this work. It was also shown that the autocatalysis effect is masked when the concentrations of reactants are high. In the present instance, however, the experimental results can be more conveniently represented and compared by employing the second order rate expression. The striking difference in the alkylation rate of phenol, p-cresol, o-cresol and 2 ,6 -dimethylphenol is prob ably the result of several factors. Figure 5 , 9, 10 and 11 contain the pertinent data. 4 and Tables Figure 4 gives rate curves calculated from the root mean square alkylation rate constants. The experimental points are shown. Figure 4 thus illustrates the closeness of fit between the experi mental points and the calculated curve. The uncertainty in each experimental point is about + two per cent and this is indicated by the size of the p o i n t . Figure phenol. It will be noted that 4 does not show the rate of alkylation for 2,6-dimethyl The alkylation rate here is extremely slow (ka = 1.1 x 10“5) in comparison to the other phenols studied and conse quently cannot be represented in this figure. Before considering the cause for the differences in the alkylation rates it will be necessary to recall a well-estab lished fact concerning substitution in the benzene ring. A. substituent, such as the hydroxyl group directs incoming groups to the ortho and/or para positions of the r i n g . On a purely statistical basis, one would expect a two to one ratio of ortho Figure 4. The Rate of Alkylation of Phenol, p-Cre3ol and o-Cresol by Phenylmethylcarbinyl Chloridetat 50°. i Phenol k i i = 3.97 x 10"3i. moleâ€* min.â€* p-Cresol ka =* 1.69 x 10~3i. mole*"* 1. min. o-Cresol k& = 3.29 x 10_^1. Alkylation moleâ€* min.-* Percent -Fr 00 80 1 100 120 Time, Minutes _ 44 - to para isomer if the hydroxyl group affects each of these positions equally. Experimentally, however, the formation of the para isomer is almost always favored and the ortho para ratio is less than two (26 ). Although the theoretical basis for such behavior is not clear at present, this situa tion obtains in the case of the alkylation of phenol by phenylmethylcarbinyl chloride, which results in an approxi mately equimolar mixture of ortho and para-a-phenylethylphenols (27* 28). The fact that phenol is alkylated at a rate about twice as fast as p-cresol (k a = 0.00397 and 0.OOI69 1. mol m i n .-1 , respectively) can be interpreted as being largely due to the lack of a free para position in p-cresol. Thus, the alkylation must be exclusively ortho, and if one can reason from the analogy of phenol itself, ortho alkylation proceeds at only about one-half the speed of para alkylation per ortho position. In addition to this primary factor, it must also be pointed out that the position ortho to the hydroxyl group in p-cresol is not as highly nucleophil\ic as in phenol because of the effect of the para methyl group. This effect is, however, very much less than the strong orienting Influence of the hydroxyl group and is probably balanced by the inductive effect of the methyl group. p-Cresol and o-cresol are identical molecules with the exception of the position of the methyl group. Thus, one would expect electrical effects in the two molecules to be identical, i.e., in both cases the hyperconjugative effect - 45 - of the methyl group serves to decrease slightly the electron density around the free para and ortho positions, while the inductive effect tends to make these positions more electro negative. In addition, the phenomenon of predominant alkyla tion in the para position should lead to faster alkylation of o-cresol than p-cresol, if it were the only effect operating. Experimentally, the rate of alkylation of o-cresol is only about one-fifth that of p-cresol, (ka =0.000329 and0.OOl69 —1 —1 1 . mol. min. , respectively) which cannot be explained by invoking either of the above facts. However, if the reaction path is one which involves the hydroxyl group in the rate-determining step, then the presence of a methyl group in the ortho position would interfere sterically, causing a slower rate of reaction. The manner in which the hydroxyl group is involved in the process is not clear from these experimental data. The results do provide strong evidence that a sterically unhindered hydroxyl group permits the alkylation reaction to proceed at a much faster rate. This fact is even more strikingly demonstrated when one con siders- the alkylation rate of 2,6-dimethylphenol. This com pound is alkylated at a rate which is only about one-thirtieth that of o-cresol and about one-three hundred and fifieth that of phenol. Rate data on 2,4-dimethylphenol would be desirable for comparison, but were not obtained in this work. This reaction could conceivably proceed via an Sn^ type of mechanism, i.e., a slow step in which the halide is ionized, - 46 - followed by a fast step in which the free carbonium ion reacts with the phenol. However, several factors argue against such a free carbonium-ion intermediate. It does not explain the decrease observed in the rate constants as the ortho positions become substituted with methyl groups. One might suppose that electron density in the para position (or free ortho position in o-cresol) is decreased because of the inhibition of normal phenol resonance by the ortho methyl groups. In the aniline series, however, such steric inhibition of resonance is not detectable unless atoms larger than hydrogen are on the nitrogen. The same is likely here for ortho methyl substituents, although ortho t-butyl substituents are large enough to inhibit even hydrogen atoms. Furthermore, resonance structures + OH i c h 3- / X , -CH-3 0-H ii -CH II such as II do not contribute greatly to the resonance hybrid in the undissociated phenol molecule because of the unfavorable charge distribution which is set up (29)* Their effect is much more pronounced in the resonance structures of the phenolate ion, which is, however, unlikely to exist in an acid solution such as prevails in the case under consideration. Of course, the hyperconjugative effect of the methyl groups also serves to make the para position less attractive to an electrophillic - 47 - reagent such as the carbonium ion involved here, but again this effect is largely balanced by the opposing inductive effect. However, it seems that to attribute the large ob served decrease in the alkylation rate constants to the above factors is certainly a procedure open to question. Very striking additional evidence against a free carbonium ion is provided by the fact that the alkylated phenols formed retain some optical activity, which would not be possible if a free planar carbonium ion were the intermediate. A situation which is more likely to obtain in this case is one in which the carbonium ion is highly solvated by the phenol molecules surrounding it. This would require a high power of rate dependency for phenol, which appears to be the case in experiments which permit the calculation of the order with respect to phenol (30). If such solvation were necessary for maximum ease of alkylation, then one would expect the alkyla tion rate to decrease sharply as the ability for the hydroxyl group to solvate is hindered sterically by bulky groups in the ortho position. This is, of course, the situation encountered experimentally. This mechanism would thus require an intimate participation by the hydroxyl group of the phenol in the ratedetermining step. Other mechanisms which require that the hydroxyl group of the phenol molecule be intimately involved in the ratedetermining step and which, thus, require maximum steric free dom of this group for maximum reaction, are possible. These - 48 - would Involve either (l) the formation of an oxonium-lon type intermediate in the rate-determining step which is transformed to the alkylated phenol in the presence of the liberated hydrogen chloride or (2 ) a concerted type mechanism in which the hydroxyl group of one phenol molecule is coordi nated to the chlorine atom of the phenylmethylcarbinyl chloride, while the para position of another phenol molecule is available for alkylation. The experimental data do not permit a final distinction as to which mechanism (s) is/are operative. They do, however, indicate the need for a mechanism which requires that the hydroxyl group be sterically unhindered for maximum ease of alkylation. Actually, it seems that there is little distinction between the latter three mechanisms in the final analysis, since all meet this requirement. The activation energy and activation entropy for the alkylation of phenol by phenylmethylcarbinyl chloride over the temperature range 30 to 5CPc. ai*e 13.8 +0.2 kcal. and -29.0 e.u. respectively. The relatively large negative entropy of activation and the low energy of activation in dicate "that a rather precise steric orientation is required in the transition state, but that a relatively easy reaction path, energetically speaking, is available once the proper steric relationships have been established (30, 31). Swain (31) has found an activation energy and entropy of activation of the same order of magnitude for the formation of methyl trityl ether from trityl chloride and methanol. He has cited this as supporting evidence for a termolecular transition state. - 49 - The racemization of optically active phenylmethyl carbinyl chloride in the presence of phenol, g-cresol, o-cresol, 2,6-dimethylphenol, and mesitol was studied kinetically. The reaction was followed by noting the change in optical rotation with time. A blank determina tion indicated no thermal racemization of the halide under the conditions of the experiments. The fact that the chloride is racemized in mesitol is highly significant. No alkylation can take place here since the ortho and para positions are blocked by methyl groups. This means that racemization can and does take place by an independent path and that alkylation is not a requisite for racemization. Thus, although racemization and alkylation are undoubtedly interrelated in the cases where both can take place, at least part of the racemiza tion observed in each case must be due to an independent reaction. This will occur simultaneously with the alkyla tion reaction when the latter is a possibility. Before a comparison of the racemization rates in the various- phenols is made, it will be well to consider the possible mechanisms which could explain the observed ex perimental facts. The experimentally observed changes in optical rotation in the cases of racemization in phenol, p-cresol, o-cresol and 2,6-dimethylphenol are not a measure of the true rate of racemization since in all these cases the optically active chloride is also being used up in the - 50 - alkylation reaction. Thus any mechanism proposed must consider the disappearance of the alkyl chloride by the alkylation process. The situation would appear to be further complicated by the fact that the alkylated phenol formed in the reaction has optical activity. Experimen tally, however, the optical rotation of the alkylated phenol is very slight compared to that of the chloride even when it was measured in the absence of a solvent. For example, when 2.50 M solutions of phenylmethylcarbinyl 20 chloride, a-p i n -47.1 / \ (1 = 1 , homogeneous), and phenol were mixed and allowed to react at 50° for thirty minutes, the a-phenylethylphenol isolated had a rotation of only a^0 -.75° +0.05° (1 = 1, homogeneous). In the kinetic studies under consideration, the phenylmethylcarbinyl chloride had an approximate rotation of <x^° 20° (1 = 1 , homogeneous), which would lead to a significantly lower rotation for the a-phenylethylphenol than that found a b o v e . In addition to this, the kinetic measurements were made in a solution 2.50 M in each reactant, which reduces the optical rotation still further. Thus, it would seem that the maximum rotation that the a-phenylethylphenol. could contribute under the conditions of the kinetic studies would be somewhat less than 0.1°. Ex periments recently carried out (3 2 ) indicate that rotations of this order of magnitude or less would be produced in the kinetic experiments involving racemization by the other phenols under consideration. Consequently, these rotations were neglected - 51 - in the consideration of rate expressions that would fit the experimental d a t a . As has already been noted (cf. Results) the racemization of phenylmethylcarbinyl chloride in various phenols can be satisfactorily explained by a rate expression in which the rate of disappearance of optically active halide is the sum of four terms, three of which represent the racemization in duced by the phenol, the alkylated phenol and hydrogen chloride, while the fourth represents the disappearance of optically active halide due to simultaneous alkylation. _ dt Thus, we have = k p (R * C 1 )(P h O H ) + k. p (R*C1)(A-PhOH) + r a .r . k H ci(R*c l )(H C 1 ) + k a (R*Cl)(PhOH) or (2 ) - ~ = k p (a)(a-x) + k A p _(a)(x) + k H C 1 (a,)(x) + d"C * â– k a (a)(a-x) The symbols for this and the following equations are defined in Table 12. In order to determine the closeness of fit of this mechanism to the experimental data, values of a at various times were calculated from the integrated form of equation (2 ). The middle curve in Figure 5 is such a calculated curve, while the points indicated are those from the experimental data. Although this mechanism (Mechanism #l) appears to fit the experimental data exceptionally well and is, thus, the mechanism used below to discuss the experimental data, it was thought desirable to test several other possible mechanisms in order - 52 - Figure 5. Test of Mechanisms 1, 5 and 6 for the Racemization of Phenylmethylcarbinyl Chloride In Phenol at 50°. Optical Rotation, Degrees 7.0 4.0 Mechanism Mechanism 2.0 1.0 Mechanism 5 0 10 20 40 30 Time, Minutes 50 60 - 53 - to determine whether the experimental data could be satisfactorily explained b y more than one mechanism. One such me c h a n i s m can be represented b y a rate expression in which the rate of disappearance of optically active halide is due to the sum of three terms, one of which represents a simple first order racemization, the second is the racemization due to hydrogen chloride p r e s  ent, while the third again represents the disappearance due to simultaneous alkylation. " = k Mathematically this becomes (R*C1) + k£c l (R*Cl)(HCl) + k (R*Cl)(PhOH) a or (3 ) †dt j IT = k ria + k - ] - I C l + ka (a ) (a _ x ) JHowever, upon closer inspection, it can be seen that the racemization rate constants of this equation can be e x  pressed in terms of those of Mechanism 1 by the following procedure. In Mech a n i s m 1, we shall let kp = k^ p (4) + A Actually, these constants (kp and k A p ) may be equal in some instances, but for the general case, tion is not justified or necessary. Also, such an a s s u m p  since the term involving the alkylation rate constant is the same for either mechanism, the following. - — it is designated b y the letter C in Equation (2) then becomes = k p (a)(a-x) + (kp-A) (<x) (x) + k H C 1 (a)(x) + C (5) _ 54 - which can be simplified to “ Qu = akPa + (k HCl “ k p + Prom a comparison of equations p )(a)(x) + C (6 ) (3 ) and (6 ), it is apparent that the kri of Mechanism 2 is equal to akp of Mechanism 1 and that k ^ ^ of Mechanism 2 is equal to kHC1 †k P + kA P of Mechanism 1. Thus, Mechanisms 1 and 2 are identical mathematically and must be distinguished from one another by consideration of the experimental facts. Mechanism 2 does not satisfactorily explain these experimental facts, since it is predicated upon the assumption that the phenol is not involved in the racemization. It does not, there fore, account for the change in rate of racemization as the structure of the phenol is changed. Mechanism 1 is, therefore, preferred over Mechanism 2. Another plausible mechanism (Mechanism 3) is one in which the disappearance of optically active halide is due to racemization induced by the phenol, a first order r ace mization and, again, the simultaneous alkylation reaction. This can be represented by _ d(R*Cl^ = k r l (R*Cl) + k p (R*Cl)(PhOH) + dt k n (R*Cl)(PhOH) or a - — — = k ria + kp(a)(a-x) + C dt 1 r (7) By a treatment similar to that just employed for Mechanism 2, it can be shown that (kri + a k p ) of Mechanism 3 must equal the akp of Mechanism 1, and that (-kp ) of Mechanism 3 must equal - 55 - (kjjci - kp + k A P >) of Mechanism 1. Thus, Mechanism 1 and 3 are identical mathematically, but Mechanism 3 is immediately reduced to absurdity since the solution of equation (7 ) leads to negative values of kp. A fourth possible mechanism, which ignores possible racemization by hydrogen chloride, requires three terms in the rate equation. Two of the terms account for the disappearance of optically active halide due to racemiza tion induced by the phenol and the alkylated phenol, while the third term again accounts for the disappearance due to simultaneous alkylation. Symbolically this becomes: d(R*Cl) = k p (R*Cl)(PhOH) + kA dt k a (R*Cl)(PhOH) (R*C1)(A-PhOH) + (8) or ^ = k p (a)(a-x) + kA P < (a)(x) + C An analysis similar to that outlined above for Mechanism 2, leads to the conclusion that the quantity, akp, of this mechanism (Mechanism 4) is equal to the quantity, akp, of Mechanism 1, while the quantity (kA p -kp) of this mechanism is equal to (kHC1 + kA p - kp) of Mechanism 1. Once again, therefore, the two mechanisms are reduced to a mathematical identity. Implicit in a consideration of Mechanism 4, however, is the fact that there is no racemization induced by hydrogen chloride and, furthermore, that the racemization induced by the alkylated phenol is greater than that due to the unalkylated phenol. (Since kA p -kp is equal to k^cl + .F. ~ and - 56 - this is a positive number, see Table 20). possible that kA p Although it is be larger than k p , it seems highly u n  likely that it should be larger by a factor of three. 20) (Table Such a large difference becomes even more unreasonable when it is noted that rate of racemization is experimentally found to decrease as the phenol involved becomes more sub stituted, i.e. more like the alkylated phenol. Mechanism 1 is to be preferred over Mechanism On this basis, 4. It was thought that perhaps the experimental data could be explained by a simpler rate expression as, for example, one in which the disappearance of optically active halide would be simply the sum of two terms; one, a first order racemization term and the other again representing the disappearance due to the simultaneous alkylation reaction. Thus, - dt = kr (R*Cl) + k a (PhOH) (R*C1) 1 ex or (9 ) - “ = kri<x + k a (a-x) (a) Upon integration, this equation gives: log a Q/a = krit/2.303 + log(l + akat) (10) Dividing both sides by t leads to: log a Q/a/t = k ri/ 2.303 + log(l + akat)/t (ll) which is in the form of a straight line, Y = c + dx, whose slope, d, is equal to one, and whose Y intercept, c, is equal to kp^/2.303. The straight line obtained for the racemization of phenylmethylcarbinyl chloride in the presence of phenol is - 57 - shown in Figure 1. That this mechanism does not fit the experimental data, is immediately evident since the slope of the straight line in Figure 1 is negative and this mechanism requires a slope of positive one. Additional evidence is afforded by the fact that the value of k r^ calculated from the value of the Y intercept, leads to calculated values of the optical rotation, a, which are widely divergent from those found experimentally. The lack of agreement is shown in Figure 5, in which the above mechanism is referred to as Mechanism 5The possibility that the racemization was one which was dependent solely on hydrogen chloride catalysis was also considered. This can be represented in the form of an equation, as follows: _ = kjjQ1 (R*C1) (HC1) + k a (R*Cl) (PhOH) (1 2 ) or †dt = k6 ci<a )<x > + ka(tx) (a - x ) This expression can be integrated in a manner similar to that described above. The integrated form is as follows: l°g a’o/<x = a^c’ kci^'/^ m3^3-k^ci_k a/Aca [log(1 + akat)](l3) If both sides are divided by t, the equation log a 0/a = akftci _ kftc1 -ka log(l + akat) t 2.303 ka t is obtained. This equation is of the same form as equation (ll) above, with the slope, d, equal to (_kHCl~ka ) and the Y ka - 58 - intercept, c, equal to akj^Q^/^ •303. It requires, then, that the second order racemization rate constant calculated from the slope and that calculated from the intercept be the same. When the equation is applied to the experimental data for the racemization of phenylmethylcarbinyl chloride in the presence of phenol at 50°, the racemization rate constant cal culated from the slope of the straight line (Figure l) is 0. OIO89 1 . m o l . --1- m i n . --1-, while that obtained from the inter cept is 0.01455 1. m o l . --1- m i n . --1-. differ by a considerable amount, Although these values the mechanism was tested further for closeness of fit to the experimental points by calculating values of a using the mean value of 0 .01272 1 . mol.--1- m i n . --1- for in equation (l4) . The deviation from the observed values is considerably greater than the experimental error as can be seen by an inspection of Figure 5. This mechanism is designated as Mechanism 6 in Figure 5. The final mechanism that was tested can be represented by a rate expression in which the disappearance of optically active halide is the sum of three t e r m s . Two of these are third order terms involving hydrogen chloride and optically active chloride with alkylated phenol and phenol respectively. The third again accounts for disappearance due to simultaneous alkylation. Mathematically, this becomes: _ d.(fi*. c l ) = k P (PhOH) (HC1) (R*C1) + k A p (A-PhOH) dt 3 * *3 (HC1)(R*C1) + k a (R*Cl)(PhOH) or (15) - -- = k p (a-x)(x)(a) + k. p (x)(x)(a) + dt 3 A *r •3 k a (a) (a-x) - 59 - Integration of (15) by methods similar to those employed in the previous cases leads to In a Q/a = 2a-2kA . P .3 + 3a2kp^ + a3kat(kp^ + kA .p.^)t 1 + akat al. ?kA.,P.3_±-. 3k P3} ln(l + ak t) + ln(l + ak_t) k„ a a a Transposing and combining like quantities gives: In <x0/a - ln(l + ak&t) = a(2kA>p<3 + 3^ ) r at 1 . a3ka(kpc; + kA.p.^)t2 [1 V akat - £ ln(l + ak^t)] + ..... 1 + aka t ' '' Multiplying both sides by (l + akat)/t2 , combining terms and converting to common logarithms gives: - [log ^ \ , £- - log (1 + akat)] = a(2kA- p cc + 3kp ) *0 3 (18) [— a 2.303 t (1 + akat) log (1 + akat)] + a3 ir (k + k )/ kat2 a p3 A -p -3 2.303 Equation (18) is now in the form of a straight line. Thus, if this mechanism is applicable to the experimental data, a plot of the left-hand side of equation (1 8 ) versus [a/2.303 t - .(1 + akat) log (1 + akat ) j shoula yleia a kafc straight line. Figure 6 shows such a plot for the race mization of phenylmethylcarbinyl chloride in phenol at 50°. The data definitely do not yield a straight line and this mechanism (Mechanism 7) is therefore discarded. ( Figure 6 . Test of Mechanism 7 for the Racemization of Phenylmethylcarbinyl Chloride in Phenol at 50°. - -5.5 -4.5 -5.° t )Aa*2 a/2.303t - (1 + alc^logU + akat)Aat2 - 61- It follows, then, that of the seven rate expressions considered, only the first explains all the experimental data satisfactorily. This expression, it will be recalled, contains three second order rate constants, the sum of which, together with a term that accounts for the disappearance of the chloride by alkylation, represents the observed change in optical rotation with t i m e . Figure 7» 8 and 9 show the curves calculated from this equation for the racemization of phenylmethylcarbinyl chloride in phenol, g-cresol and o-cresol at 50°. The points shown are those observed e x  perimentally. Since both alkylated phenol and hydrogen chloride are formed in stoichiometrically equal amounts in the reaction, they must be represented by the same quantity in the rate expression, and the corresponding rate con stants cannot be evaluated separately. However, the sum of these two constants can be evaluated and this is listed as kj^. in Table 20. The second order rate constant, k p , for the racemization reaction which depends on phenol and phenyl methylcarbinyl chloride can be determined directly and these values are also listed in Table 20. the value It will be noted that of kx for each phenol studied is at least threefold greater than the value of kp. The value of k^.p. should be of the same order of magnitude as kp and is probably less, as has already been pointed out. Thus, the greater part of kx must be the contribution of kH C 1 , which is, therefore, a very real - 62 - Optical Rotation, Degrees 7.0 Figure 7. The Racemization of Phenylmethylcarbinyl Chloride in Phenol at 50° 5.5 4.0 3.5 3.0 2.5 10 # 20 30 40 Time, Minutes 50 60 - 63 - Figure 8. The Racemization of Phenylmethylcarbinyl Chloride in o-Cresol at 50°. Optical Rotation, Degrees 7.0 5.5 5.0 4.0 3.5 3.0 2.5 40 80 120 160 Time, Minutes 200 240 Figure 9. The Racemization of Phenylmethylcarbinyl Chloride in £-Cresol at 50°. 7 6 6 5 5 4 4 3 3 2 10 20 30 40 Time, Minutes 50 60 - 65 - quantity that accounts for a great deal of the experimentally observed racemization. In considering a mechanism for the racemization of the chloride induced by the phenol (measured b y kp in the rate expression), the following facts must be satisfactorily explained: (l) The values of kp decrease markedly as the ortho positions of the phenol become substituted with methyl groups, while there is only a slight decrease when the methyl group is in the para position, (2 ) the reaction is first order with respect to phenol and first order with respect to the chloride, and (3 ) the reaction has a relatively high entropy of activation of -25 e.u. and a relatively low energy of activation of 1 5 .^ k c a l . in the case of phenol itself. Two alternate reaction paths appear plausible. In one, the transition state is represented b y a cyclic intermediate as f o l l o w s : H C6% C1J V ^ - H H s*/ H - 0 - C6H 5 This intermediate requires that a rather precise orientation of the molecules involved be established in the transition s t ate. Such a requirement receives strong support from the relatively high negative entropy of activation of -25 e.u. found for the racemization of the chloride in phenol at 40°. Bulky ortho substituents would interfere with the establish ment of such a transition state in that the approach of the - 66 - chloride molecule to the phenol molecule would be considerably more hindered. Since the formation of this intermediate is dependent upon the ease with which the molecules involved can approach one another, the measured value of kp should and does decrease as the degree of ortho substitution increases. That substitution in the ortho position is the important criterion of the rate is readily ascertained by a comparison of the kp's for mesitol and 2,6-dimethylphenol. These phenols are identical in their steric requirements in the vicinity of the hydroxyl group and one would expect, on the basis of the considerations above, that their kp's should be essentially the same. This is indeed the case as can be seen from Table 20. The fact that the kp for the racemization induced by phenol is somewhat larger than that for p-cresol can r e a  sonably be ascribed to the electrical effect of the para methyl group. Such a contention finds further support in the work of Boyd (33), who has determined the dissociation constants of various phenols. The pertinent information is reproduced in the following table. Dissociation Constants of Certain Phenols Compound Phenol m-Cresol p-Cresol o-Cresol Prom these values, Dissociation Constant 1.15 0 .98 .67 .63 x x x x 10" j-*? 1 0 â€J® 1 0 _1° it can readily be seen that the combined inductive and hyperconjugative effects of the nuclear methyl - 67 - group serve to strengthen the 0-H "bond in the c r e s o l s . The methyl group in o-cresol, of course, exerts no steric influence in the removal of a proton, thus, the dissociation constants for o- and p-cresol are essentially identical. However, when a molecule, whose steric requirements are considerably greater than those of a proton, must approach the hydroxyl group in order for reaction to take place, the presence of the bulky ortho methyl group will serve to hinder this approach. Such is the situation that obtains in the present instance. Thus, it m a y be concluded that the d i f  ference observed between the kp for the racemization in pcresol and that for the racemization in phenol, is a measure of the electrical effect of the para methyl g r o u p . On the other hand, the difference in kp for the racemization in pcresol and o-cresol is largely a measure of the steric effect of the ortho methyl group since electrical effects are essen tially the same in these two m o l e c u l e s . The cyclic intermediate proposed also accounts for the observed racemization in that the products resulting from the transition state would be styrene and hydrogen chloride. Thus an equilibrium would be established between the alkyl chloride, styrene and hydrogen c h l o r i d e . This equilibrium would be e x  pected to be largely displaced toward the left, i.e. toward the alkyl halide under the conditions of the experiment (sealed polarimeter tube). The reaction path must be one which is energetically quite favorable, since the activation energy is - 68 - only 15.4 k c a l . In other words, once the molecules have become involved in the transition state complex, the chances are excellent that they will pass over the energy "hump" to form the products above. This means that each alkyl chloride molecule that becomes a part of the intermediate will be largely transformed to the racemic modification by the above process. The second of the two possible reaction paths involves an ionization of the optically active phenylmethylcarbinyl chloride brought about b y the phenol p r e s e n t . Such an iq!lza- tion should occur most readily in the presence of an unhindered phenol, since the ability of the phenol to bring about ioniza tion is directly related to the steric freedom of its hydroxyl group. The transition state in this case would then be a solvated planar carbonium ion, which can readily stabilize itself by combining with a chloride ion or b y ejecting a proton to form sty r e n e . Should it do the latter, this path becomes nearly identical to the cyclic intermediate discussed above. In either case, the resulting phenylmethylcarbinyl chloride (formed either by direct combination of the carbonium ion with chloride ion or by subsequent reaction of the styrene with hydrogen chloride) would be a racemic modification. The present investigation unfortunately does not permit a further distinction of mechanism. however, It has been suggested, (3 4 ) that it should be possible to distinguish between these two mechanisms b y means of deuterium exchange experiments. - 69 - If the transition state involves a cyclic intermediate, then c6h 5 - CH - Cl should exchange deuterium with the phenol present, and further more, the rate of exchange and rate of racemization should be directly related. If the racemization takes place due to the formation of a planar carbonium ion, followed by combination with a chloride ion, little or no deuterium exchange would be expected. Experiments to establish this point are now in progress in these laboratories. As has already been pointed out, the values of k are in reality the sum of two rate constant, kA<p> and kH C 1 . If we can assume that the value of k ^ p> is the same as k p , then an estimate of the magnitude of kjjQ^ is available. Such an estimate is a minimum value, since k A>p>is undoubtedly less than k P . The values of kjjQj listed In Table 20 were obtained as the difference between and k p . The racemiza tion induced by alkylated phenol and measured by k^ p would, of course, be expected to proceed via the same type of Inter mediate- as that induced by phenol. The term Involving hydrogen chloride in the rate expression is second-order and, thus, it would seem that the racemization induced by the hydrogen chloride is dependent upon only the concentration of hydrogen chloride and optically active chloride. However, the fact t hat the values for k HC1 (or k^) decrease as the phenol Involved becomes more substituted in the ortho position - 70 - must also be accounted Tor in any mechanism of racemization by hydrogen chloride. It is important to note that the total phenol concentration is constant throughout the reaction insofar as the number of hydroxyl groups is concerned. Thus, it would be impossible to distinguish kinetically the effect of the phenols in the racemization induced by hydrogen chloride. Indeed, their role is essentially that of a solvent, whose prop erties differ from those of a true solvent only in that the ratio of phenol to alkylated phenol changes somewhat as the reaction proceeds. This would have a small effect on the rate of racemi zation by hydrogen chloride, but could not be detected by the present technique. It is entirely conceivable, therefore, that the transition state here involves the phenols as well as hydrogen chloride and the optically active chloride, but the determination of a kinetic order with respect to the phenols is not possible under the experimental conditions employed. The net result is, then, the second order kinetics which are found experimentally. On the basis of the above considerations, it is proposed that the hydrogen chloride induced racemization can best be represented b y a termolecular transition state involving hydrogen chloride, phenylmethylcarbinyl chloride and either the phenol or its alkylated derivative, as follows: ?6H5 â– H-Cl — ^ C H - Cl CH3 H-0-C6H 5 - 71 - This postulated termolecular intermediate is also favored energetically. The entropy of activation is n ega tive and significantly larger than that for the racemization reaction measured by k p (-47.5 e.u. and -25.0 e.u., Table 18). This means that an even more ordered arrangement is required in the present intermediate. The low energy of activation (7.3 kcal.) leads one to predict that the reaction will p r o  ceed with ease once the proper steric relationships have been established. It seems logical that a termolecular intermediate would require Just such a precise and ordered arrangement, which should become progressively more difficult to attain as the hydroxyl group becomes more h i n d e r e d . The net result will be a decrease in the value of the rate constant, which is precisely the experimental r e s u l t . Therefore, Mechanism 1 is In excellent agreement with the experimental values and the changes in the rate constants can be accounted for on the basis of reasonable theoretical considerations. - 72 - SUMMARY 1. The alkylation of phenol at 50°, 40° and 30° and of g-cresol, o-cresol and 2 ,6 -dimethylphenol at 50° "by phenylmethylcarbinyl chloride was studied kinetically by product analysis involving the bromination of the free ortho and para positions in the phenol mixtures. Second order rate constants for each alkylation were calculated. The experimental points were in satisfactory agreement with values calculated from the average second order rate constant for that reaction. 2. The alkylation rate constants decrease markedly from phenol to o-cresol to 2,6-dimethylphenol. This is attributed largely to steric hindrance of the ortho methyl groups which retards the formation of a precisely ordered transition state in which the phenolic hydroxyl group is intimately involved. 3. The rate of racemization of phenylmethylcarbinyl chloride in the presence of phenol at 50°, 40° and 30°, and in g-cresol, o-cresol, 2 ,6 -dimethylphenol and mesitol at 50° was determined. The racemization reaction was fol lowed by noting the optical rotation at measured time inter vals and was carried out in a thermostatted polarimeter t u b e . The rate of change in optical rotation with time was found to be satisfactorily expressed by a rate expression, containing - 73 - four second-order terms. One of these accounts for the disappearance of optically active chloride b y alkylation. The other three account for the racemization induced by phenol, alkylated phenol and hydrogen chloride respectively. This can be represented mathematically as g-(£ --C1) = k p (R * C 1 ) (P h O H ) + kA P (R*C1) (A-PhOH) + dt k H C 1 (R*Cl)(HCl) + k a (R*Cl)(PhOH) This expression must be modified for the cases of mesitol and 2 ,6 -dimethylphenol since in the former the alkylation cannot take place and in the latter it is negligible over the time interval stu d i e d . The values of kp and that for the sum of kA> p^ and kHC1 were determined from the above expression. Rate curves calculated from these constants were in e x  cellent agreement with the experimental points. Six other expressions were considered and found unsuitable. 4. The decrease in kp as the phenol is changed from phenol to £-cresol is attributed to the electrical effect of the methyl group. The marked decrease found in kp as the ortho positions in the phenol are substituted with methyl groups can only be interpreted in terms of a steric factor, and suggests that transition state requires maximum steric freedom for the hydroxyl group of the phenol for maximum racemization. The decrease in kHC1 for the same series is attributed to the fact that the racemization is effected in a termolecular 74 _ - transition state involving the alkyl chloride, hydrogen chloride and the phenol. Second order kinetics are ob served because the concentration of the phenol (insofar as the number of hydroxyl groups is concerned) does not change. 5. Energies and entropies of activation for the alkylation of phenol by phenylmethylcarbinyl chloride and for the racemization of phenylmethylcarbinyl chloride in the presence of phenol over the temperature range 30 to 50°C. were calculated. The relatively large negative entropies of activation and relatively low energies of activation indicate that the transition state is highly ordered in all cases, and that the reaction can proceed with relative ease, energetically, once the transition state has been established. 6 . Possible mechanisms which fit the kinetic observations are discussed. - 75 - REFERENCES 1. H. Harfcj Abstracts, A. C. S. Meeting, April 9-11, 1951, Cleveland, p. 94 m . 2. C. C. Price, Chem. Rev. 29, 37 (1941). 3. C. C. Price and M. Lund, (1990). 4. R. L. Burwell and S. Archer, ibid. 64, 1032 (1942). 5. C. C. Price, "Reactions at Carbon-Carbon Double Bonds", p. 51, Interscience Publishers, N. Y . , 1946. 6 . R. L. Burwell, 7. J. Am. Chem. S o c . 62, 3105 J. Am. Chem. Soc. 64, 1025 (1942). M. M. Sprung and E. S. Wallis, ibid. 56, 1715 (1934). 8 . W. I. Gilbert and E. S. Wallis, J. Org. Chem. 5, 184 (1940). 9. E. Bergmann and M. Polanyi, Naturwissenschaften 21, 378 (1933). 10. E. D. Hughes, C. K. Ingold and A. D. Scott, Nature 138, 120 (1936). 11. E. D.. Hughes, F. and J. Weiss, J. 12. E. D. Hughes, F. Juliusburger, A. D. Scott, B.Topley and J. Weiss, ibid. 1173, (1936). 13. E.„Bergmann, M. Polanyi and Chem. 2 0 © , l6l (1933). A. L. Szabo, Z. physik. 14. A. R. Olson and F. A. Long, (1934). J. Am. Chem. Soc. 5 6 , 1294 15. Juliusburger, S. Masterman, B. Chem. Soc. 1525, (1935). C. L. Arcus, A. Campbell and J. Kenyon, Topley J. Chem. Soc. 1510, (1949). 16. K. Bodendorf and H. Bohme, Ann. 516, 1 (1935). 17. A. J. Houssa and J. Kenyon, 18. E. Downer and J. Kenyon, J. Chem. Soc. 2261, ibid. 1156, (1939). (1930). - 76 - 19. W. Gerrard, ibid. 741, (1946). 2 0 . A. H. Blatt, "Org.. Synthesis, Coll. Vol. II", John Wiley and Sons, N. Y. (1943), p. 449. 2 1 . H. Hart, J. Am. Chem. S o o . 7 2 , 2900 (1 9 5 0 ). 22. Ng. Ph. Bui-Hoi, H. LeBihan and F. Binon, J. Org. Chem. 17, 246 (1952). 23. I. W. Ruderman, Anal. Chem. 18, 753 (1946). 24. E. Leininger and K. G. Stone, "Elementary Quantitative Analysis, A Practical Approach", Mich. State College Press, East Lansing, Mich., 1950. 25. F. A. Cassis, Jr. and H. Hart, unpublished work. 26. E. Alexander, "Principles, of Ionic Organic Reactions", John Wiley and Sons, N. Y., 1950, p. 242. 27* H. Hart, Anal. Chem., 24, 1500 (1 952). 2 8 . H. Hart, private communication. 29. Ref. 26, p. 25. 30. H. Hart and J. H. Simons, J. Am. Chem. Soc. 7 1 , 345 (1949). 31. C. G. Swain, ibid. 70, 1119 (1948). 32. H. Hart and H. S. Eleuterio, unpublished work. 33. D . R. Boyd, J. Chem. Soc., 107, 1539 (1915). 34. H. Hart, private commimication. - 77 - APPENDIX The tables referred to in the body of this thesis are contained, in consecutive order, In this appendix. TABLE 1 ANALYSIS BY BROMINATION OP PHENOL, o-(a-PHENYLETHYL)PHENOL, ■£- (a -PHENYLETHYL)PHENOL AND MIXTURES OP THESE ________ Millimoles Sample o-(a-Phenyl- p-(a-PhenylNo. Phenol ethyl)phenol ethyl)phenol 1 2 3 4 5 6 0.402 .402 .402 .402 .402 .402 0.384 .384 .384 7 8 9 10 11 12 13 14 Reaction Net Time, Meq. Meq. Meq. Br2 Sec. BrO^'-Br- SgOo= Absorbed R* 120 120 240 20 25-30 25-30 2.7160 2.7160 2.7160 2.7160 2.7160 2.7160 0.3726 .3575 .3575 .3520 .3320 .3370 2.3434 2.3585 2.3585 2.3640 2.3840 2.3790 2.92 2.93 2.93 2.95 2.97 2.96 20 20 30 1.6490 1.6490 1.6490 .1913 .1913 .1720 1.4577 1.4577 1.4770 1.92 1.92 1.95 2.20 2.06 2.09 2.05 2.09 O .1656 .1656 .1656 .1656 .3312 60 10-15 10-15 5 25-30 0.7760 .7760 .7760 .7760 1.4550 .0504 0.7256 .6803 .0957 .6904 .0856 .1007 •6Z?3 .0705 1.3845 Phenol Pound Calc'd. p 15 16 17 .396 .396 .396 .0212 .0212 .0212 .0229 .0229 .0229 20 25-30 25-30 2.7160 2.7160 2.7160 .206 .156 .156 2.510 2.550 2.550 85.3 89.8 89.8 90.0 90.0 90.0 18 19 20 .396 .396 .396 .0772 .0772 .0772 .0557 .0557 .0557 30-35 25 25 3.100 3.100 3.100 .216 .201 .212 2.884 2.899 2.888 72.6 73.9 73.0 74.9 74.9 74.9 TABLE 1 CONTINUED Millimoles Reaction Net o-(a-Phenyl- g-(a-PhenylTime, Meq. Meq. Meq. Bro Sample Sec. BrO^--BrÂâ– s2o3= Absorbed No. Phenol ethyl)phenol ethyl)phenol % Phenol R* Pound Calc'd. 21 22 23 24 0.396 .396 .396 .396 0.193 .193 .193 .193 0.209 .209 .209 .209 20 25-30 25-30 25-30 il.171 4.171 4.171 4.171 0.236 .201 .201 .201 3.935 3.970 3.970 3.970 46.4 48.7 48.7 48.7 49.7 49.7 49.7 49.7 25 .158 .278 .201 20-25 3.100 .246 2.854 24.0 24.8 * R denotes the number of reactive positions per molecule; i.e. for phenol, R = 3. (theory) TABLE 2 ANALYSIS BY BROMINATION OP g-CRESOL, 2- (a-PHENYLETHYL) 4-METHYLPHENOL AND MIXTURES OP THESE _______ Millimoles________ Sample 2-(a-Phenylethyl)No. g-Cresol 4-methylphenol 1 2 3 4 Net Meq. Bro Absorbed R % p-Cresol 15 25 25 25 1.183 1.183 1.183 1.183 0.470 .415 .405 .408 0.713 .768 .778 .775 1.85 1.99 2.02 2.01 0.2165 .2165 .2165 .2165 25 25 25 25 0.592 .592 .592 .592 .179 .170 .159 .144 .413 .422 .433 .448 0.96 .98 1.00 1.03 0.1930 .1930 .1930 .1930 5 6 7 8 Reaction Time, Meq. Meq. Sec. BrO^'-Br" ^ 2^ Pound Calc*d. 9 10 .1544 .1544 .0217 .0217 25 25 .986 .986 .328 .326 .658 .660 86.9 87.4 87.4 87.4 11 12 13 14 15 .1544 .1544 .1544 .1544 .1544 .0433 .0433 .0433 .0433 .0433 25 20 25 25 25 I.O85 1.085 1.085 1.085 1.085 .382 .392 .385 .387 .378 .703 .693 .700 .698 .707 77.8 75.3 77.3 76.8 78.8 77.8 77.8 77.8 77.8 77.8 16 17 18 19 20 .1351 .1351 .1351 .1351 .1351 .0650 .0650 .0650 .0650 .0650 25 25 35 30 30 0.986 .986 .986 .986 .986 .322 .324 .310 .316 .314 .664 .662 .676 .670 .672 66.0 65.5 69.O 67.5 68.0 67.5 67.5 67.5 67.5 67.5 TABLE 2 CONTINUED Millimoles Reaction Net 2-(a-Phenylethyl)- Time, Meq. Meq. Meq. Bro Sample 4-methylphenol Sec. BrO^-Br- s2o3= Absorbed No. p-Cresol 21 0.1158 22 23 .1158 .1158 0.0866 .0866 .0866 30 30 35 0.986 .986 .986 0.352 0.634 .3^7 .639 .342 .644 $ p-Cresol R Pound Calc!d. 56.9 57.9 59.4 57.4 57.4 57.4 TABLE 3 ANALYSIS BY BROMINATION OP O-CRES0L, a-PHENYLETHYL-2-METHYLPHEN0L AND MIXTURES OF THESE Reaction Millimoles Net a-PhenylTime, Meq. Meq. 1 Sample Meq. Br2 Sec. BrOgâ€-Br†s2°3= Absorbed No. o-Cresol ethyl-o-cresol 1 2 3 4 0.200 .186 .186 .186 15 15 15 25 0.204 .204 .204 .204 .204 5 6 7 8 9 15-20 25 27 15 15 1.183 I.O85 I.O85 1.183 0.372 .338 .344 .405 6.6902 .6902 .6902 .5916 .5916 .2867 .2847 .2796 .1925 .2026 % o-Cresol R 0.811 .746 .741 .778 2.03 2.01 2.00 2.10 .404 .406 .411 .399 .389 0.99 1.00 1.01 0.98 .96 Pound Calc'd. 10 11 .180 .360 .0200 .0400 15 15 1.183 2.268 .415 .729 .768 1.538 91.7 91.8 90.0 90.0 12 .13 .334 .334 .081 .081 15 15 2.169 2.169 .672 .669 1.497 1.500 80.3 80.8 80.5 80.5 14 15 16 .278 .278 .278 .122 .122 .122 15 15 15 1.972 1.972 1.972 .598 .612 .598 1.374 1.360 1.374 71.6 70.0 71.6 69.5 69.5 69.5 17 18 .241 .241 .163 .163 15 15 1.873 1.873 .580 .573 1.293 1.300 60.2 60.8 59.8 59.8 TABLE 4 ANALYSIS BY DOMINATION OP 2,6-DIMETHYLPHENOL, 4-(a-PHENYLETHYL)-2,6-DIMETHYLPHENOL AND MIXTURES OP THESE Millimoles Reaction Net Sample 2,6-Dimethyl- 4-(a-Phenylethyl)Time, Meq. Meq. Meq. Brg No. phenol 2,6-dimethylphenol Sec. BrO^^-Br- 8203= Absorbed 1 2 3 4 5 6 7 8 0.200 .200 .200 .200 .200 9 10 11 12 13 14 15 16 .400 .400 .380 .380 .360 .360 .340 .340 .320 .320 .300 .300 .280 .280 18 19 20 21 22 0.192 .192 .192 0 10 5 5 5 5 5 5 0.444 .444 .444 .444 .444 .444 .444 .444 0.108 .030 .037 .046 .046 .101 .127 .152 0.336 .414 .407 .398 .398 .343 .317 .292 .019 .019 .038 .038 .057 .057 .076 .076 .096 .096 .115 .115 20 20 20 20 20 20 20 20 20 20 20 20 20 20 .887 .887 .838 .838 .789 .789 .740 .740 .690 .690 .641 .641 .592 .592 .055 .061 .029 .028 .027 .027 .022 .022 .021 .021 .020 .018 .021 .022 .832 .826 .809 .810 .762 .762 .718 .718 .669 .669 .621 .623 .571 .570 % 2,6-Dimethyl- R phenol Calc'd. 0.840 1.03 1.02 0.995 .995 .890 .830 .760 100 100 95.2 95.2 90.4 90.4 85.6 85.6 80.7 80.7 75.8 75.8 70.9 70.9 TABLE 5 THE RATE OP ALKYLATION OP PHENOL BY PHENYLMETHYLCARBINYL CHLORIDE IN p-XYLENEa TEMPERATURE: 50° INITIAL CONCENTRATIONS: 2.50 M. IN EACH REACTANT Sample No. 1 2 3 4 5 6 7. Bb Time, Min. 10 20 40 60 80 120 140 0 Net Meq. Brp Consumed 1.137 1.119 1.077 1.038 0.998 .970 .942 1.178 Millimoles a-Phenylethylphenol 0.0206 .0292 .0505 .0700 .0900 .104 .118 % Alkylation x(moles/l.) k-(l. mole"1 a-x (moles/l.) a min.-l) 0.26 2.24 0.00464 10.5 .00348 14.9 2.13 .37 25.8 .00352 1.85 .65 1.61 .89 .00369 35.7 45.8 .00426 1.35 1.15 53.2 1.33 1.17 .00379 1.00 .00428 60.2 1.50 Root mean square ka = 0.00397 ± °.00C i8$ a Analyses carried out on 5 ml. aliquot obtained from dilution of the phenol mixtures to 250 ml. b Blank analysis on 5 ml. aliquot to determine meq. of bromine consumed by 10 millimoles of phenol after alkaline extraction and dilution to 250 ml. TABLE 6 THE RATE OP ALKYLATION OP PHENOL BY PHENYLMETHYLCARBINYL CHLORIDE IN BENZENEa'b TEMPERATURE: 50° PHENOL CONCENTRATION: 2.57 M. PHENYLMETHYLCARBINYL CHLORIDE CONCENTRATION: 2.38 M. Sample No. 1 2 3 4 5 6 Bc Time, Min. Net Meq. Br2 Consumed 10 30 50 60 90 120 0 1.327 1.281 1.241 1.213 1.161 1.125 1.365 Millimoles a-Phenylethylphenol 0.019 .042 .062 .076 .102 .120 % Alkylation 8.3 18.3 27.0 33.2 44.9 52.6 a-x (moles/l.) b-x (moles/1 .) kg^l. mole"1 min.-1) 2.36 2.17 0.00399 2.10 .00309 1.91 1.88 1.69 .00309 1.72 .00352 1.53 1.41 1.22 .00399 1.02 1.21 .00412 Root mean square ka =0.00366 +0.000*$ a Average of three runs. b Analyses carried out on 5 ml. aliquot from dilution of phenol mixture to 250 ml. c Blank analysis on 5 ml. aliquot to determine meq. of bromine consumed by 10 millimoles of phenol after alkaline extraction and dilution to 250 ml. Average of two determinations. TABLE 7 THE RATE OP ALKYLATION OP PHENOL BY PHENYLMETHYLCARBINYL CHLORIDE IN p-XYLENEa TEMPERATURE: 40° INITIAL CONCENTRATION: 2.50 M. IN EACH REACTANT 6a 7 8. B 20 40 60 60 90 120 120 150 150 180 240 0 1.152 1.124 1.101 1.106 1.081 1.053 1.043 1.025 1.033 0.991 .978 1.178 0.0129 .0266 .0382 .0357 .0484 .0625 .0675 .0763 .0722 .0935 .1000 a * Alkylation 6.6 13.5 19.5 18.2 24.6 31.8 34.4 :s(moles/1.) a-x (moles/1.) 0.17 .34 2.33 2.16 .47 .62 2.03 1.88 1.68 *1Q Q 18-9 ,, . 33#1 36.8 37.0 47.5 51.0 cvi * 1 2 3 3a 4 5 5a 6 Time, Min. Millimoles a-Phenylethylphenol 00 Sample No. Net Meq. Br2 Consumed k« (1 . moles-1 min."*) 0.00146 .00157 .00154 .00147 .00163 .00163 .95 1.55 .00202 1.31 1.19 1.28 1.22 .00175 Root mean square ka =0.00158 +o.0004Jt Analyses carried out on a 5 ml. aliquot obtained from dilution of the phenol mixture to 250 ml. 13 Blank analysis on a 5 ml. aliquot to determine meq. of bromine consumed by 10 millimoles of phenol after alkaline extraction and dilution to 250 ml. TABLE 8 /> THE RATE OP ALKYLATION OP PHENOL BY PHENYLMETHYLCARBINYL CHLORIDE IN p-XYLENEa TEMPERATURE: 30° INITIAL CONCENTRATION: 2.50 M. IN EACH REACTANT Sample No. 1 2 3 3a 4 5 % B* Time, Min. Net Meq. Bro Consumed 60 120 180 180 240 360 360 0 1.138 1.099 1.055 1.049 1.030 1.008 0.997 1.178 Millimoles a-Phenylethylphenol 0.0200 .0398 .0615 .0645 .0737 .0848 .0904 $> Alkylation 10.2 20.3 x(moles/l.) a-x (moles/l.) 0.26 .51 2.24 1.99 1.70 ka^1, mole_1 min.*1) 0 .000774 .000854 .00105 ii:i 32-1 -80 .00100 1.56 37.5 .94 43.2 44 5 1 n .00106 0 1.39 46.0 „ ' Root mean square ka = 0 .000955 +0.00111 a Analyses carried out on a 5 ml. aliquot obtained from dilution of the phenol mixture to 250 ml. k Blank analysis of 5 ml. aliquot to determine meq. of bromine consumed by 10 millimoles of phenol after alkaline extraction and dilution to 250 ml. TABLE 9 THE RATE OP ALKYLATION OP £-CRESOL BY PHENYLMETHYLCARBINYL CHLORIDE IN BENZENE3 TEMPERATURE: 50° INITIAL CONCENTRATIONS: 2.50 M. IN EACH REACTANT Sample No. 1 2 3 3a 4 5 5a 6 6a 7 8. Bb Time, Min. 15 30 45 45 60 90 90 120 120 150 180 0 Net Meq. Bro Consumed 0.759 .744 .726 .725 .702 .673 .689 .652 .648 .635 .619 .785 Millimoles 2-(aPhenylethyl)-4methylphenol 0.0130 .0205 .0295 .0300 .0415 .0560 .0480 .0665 .0685 .0750 .0830 % a-x k„(l. moles-1 (moles/l. ) min.-1) Alkylation x(moles/l.) 6.6 10.5 O .17 .26 2.33 2.24 O.OOI93 .00155 -38 .53 .67 2.12 .00159 1.97 1.83 .00179 .00163 1 1 :3 « - a 21.1 |8;6 26.6 p % 3 M 38.3 42.3 1.64 .00175 .86 .00166 1.54 .96 1.44 .00164 1.06 Root mean square k =0.00169 +0.0001$ a *“"â– a Analyses carried out on 5 ml. aliquot obtained from dilution of the phenol mixture to 250 ml. k Blank analysis on a 5 ml. aliquot to determine meq. of bromine consumed by 10 millimoles of phenol after transfer and dilution to 250 ml. TABLE 10 THE RATE OP ALKYLATION OP o-CRESGL BY PHENYLMETHYLCARBINYL CHLORIDE IN BENZENE8, TEMPERATURE: 50° INITIAL CONCENTRATIONS: 2.50 M. IN EACH REACTANT Sample No. 1 2 3 3a 4 4a 5 5a 6 7b Bd Time, Min. Net Meq. Bro Consumed Millimoles aPhenylethylo-cresoi Alkylation x (moles/l.) a-x (moles/l.) 60 120 180 180 300 300 360 360 480 600 0 0.774 .753 .741 .739 .716 .720 .702 .708 .684 .671 .793 O.OO95 .0200 .0260 .0270 .0385 .0365 .0455 .0425 .0545 .0610 4.8 10.1 ^•8*1 13.6 ^ 18.4 22.9 21.5 27.5 30.8 0.12 .25 .34 2.38 2.25 2.16 0.000336 .000370 .000350 .47 2.03 .000309 13.4 -,Q Q 18-9 „ 22 *2 .56 .69 .77 k (1 . mole"1 a mln.-l) 1.94 1.81 1.73 Root mean square k - a Analyses carried out on a 5 ml. aliquot obtained from dilution of the phenol mixture to 250 ml. b Blank analysis on a 5 ml. aliquot to determine meq. of bromine consumed by 10 millimoles of phenol after transfer and dilution to 250 ml. .000321 .000318 .000295 =0.000329 .00002D TABLE 11 THE RATE OP ALKYLATION OP 2,6-DIMETHYLPHENOL BY PHENYLMETHYLCARBINYL CHLORIDE IN BENZENE8, TEMPERATURE: 50° INITIAL CONCENTRATIONS: 2.50 M. IN EACH REACTANT Sample No. Time, Hours 1 2 3 4 5 6 72 96 120 144 168 240 8 Net Meq. Br;> % Consumed Alkylation 0.733 .717 .693 .665 .650 .621 x(moles/l.) a-x (moles/l.) kfl(l. mole"1 min.-l) 1.31 X 10-5 2.19 12.5 0.31 14.1 1.13 x 10-5 2.15 .35 1.12 x 10"5 2.08 .42 16.6 2.01 1.13 x 10"5 19.5 .*9 1.07 x 10-5 21.0 1.97 .53 24.0 .60 0.88 x 10"5 1.90 Root mean square ka=l.11 x 10--5 +o.0?x 10-5 Analyses determined on a 10 ml. aliquot obtained from dilution of the phenol mixture to 250 ml. Values of # alkylation are obtained from calibration curve of percent composition vs. meq. of bromine con sumed for known samples containing 0.4 millimoles total phenols. TABLE 12 TABLE OP SYMBOLS EMPLOYED IN THE RATE EQUATIONS a = initial phenol concentration In moles per liter. x = alkylated phenol or hydrogen chloride concentration in moles per liter at any time, t. a-x = phenol concentration in moles per liter at any time, t. aQ = optical rotation in degrees at zero time. a = optical rotation at any time, t. t = time in minutes. ka = rate constant in 1 . mole-! m i n .-1 for the second order alkylation of the phenols by phenylmethylcarbinyl chloride. kp = rate constant in 1 . mole-1 m i n .-1 for the second order racemization of phenylmethylcarbinyl chloride by unalkylated phenol. kA P = I*a‘^e constant in 1 . mole-1 m i n .-1 for the second order racemization of phenylmethylcarbinyl chloride by alkylated phenol. kHCl = 1>a'be constant in 1. moleâ€1 m i n .â€1 for the second order racemization of phenylmethylcarbinyl chloride by hydrogen chloride. kx = sum of kA _p _ and k ^ . ka 3 kr = rate constant in l .2 moles-2 m i n .-1 for third order alkylation of the phenols by phenylmethylcarbinyl• chloride. = rate constant for the first order racemization of phenyl1 methylcarbinyl chloride considered in Mechanisms 2, 3 and 5 - k.L,- = rate constant for second order racemization of phenylmethylcarbinyl chloride involving hydrogen chloride considered in Mechanism 2 and 6 . TABLE 12 CONTINUED kpo = rate constant for third order racemization of phenyl methylcarbinyl chloride Involving phenol. k. p o = rate constant for third order racemization of phenylA * *~5 methylcarbinyl chloride involving alkylated phenol. TABLE, 13 THE RACEMIZATION OP OPTICALLY ACTIVE PHENYLMETHYLCARBINYL CHLORIDE IN PHENOL TEMPERATURE: 50° CONCENTRATIONS: 2.50 M. IN EACH REACTANT Time, Minutes 0 3.33 5.00 6.67 8.33 10.00 11.67 13.33 15.00 16.67 18.33 20.00 21.67 23.33 25.00 26.67 28.33 30.00 31.67 33.33 35.00 36.67 Run 1 Optical Rotation, Degrees 6.59 6.14 5.94 5.70 5.47 5.30 5.13 4.92 4.77 4.60 4.44 4.25 4.09 3.91 3.80 3.63 3.50 3.38 3.23 3.07 2.90 2.79 Y x 1Q3 9.213 9.018 9.445 9.709 9.461 9.323 9.520 9.353 9.364 9.356 9.525 9.562 9.717 9.564 9.711 9.703 9.663 9.776 9.952 10.186 10.179 X x 103 4.246 4.204 4.174 4.143 4.112 4.084 4.040 4.021 3.979 3.962 3.941 3.903 3.884 3.848 3.828 3.798 3.777 3.745 3.726 3.697 3.676 Run 2 Optical Rotation, Degrees 6.51 6.10 5.91 5.70 5.51 5.34 5.16 4.96 4.80 4.63 4.49 4.31 4.14 4.01 3.87 3.73 3.59 3.47 3.32 3.20 3.08 2.97 Y x 10 8.483 8.396 8.651 8.695 8.600 8.663 8.875 8.820 8.878 8.789 8.950 9.077 9.014 9.032 9.066 9.121 9.107 9.236 9.253 9.28q 9.294 TABLE 13 CONTINUED Time, Minutes Run 1 Optical Rotation, Degrees 38.33 40.00 41.67 43.33 45.00 46.67 48.33 50.00 51.67 53.33 55-00 56.67 58.33 60.00 2.74 2.58 2.52 2.40 2.30 2.21 2.04 2.01 1.90 1.78 1.70 1.65 1.59 1.58 Y x 10^ 9.943 10.183 10.019 10.125 10.157 10.167 10.536 10.314 10.453 10.658 10.700 10.612 10.586 10.337 X x 108 3.650 3.630 3.612 3.584 3.567 3.540 3.524 3.498 3.480 . 3.458 3.440 3.418 3.401 3.383 Run 2 Optical Rotation, Degrees 2.84 2.73 2.63 2.51 2.42 2.30 2.23 2.13 2.00 1.96 1.87 1.78 1.73 1.63 Y x 103 9.397 9.438 9.446 9.555 9.551 9.681 9.625 9.704 9.921 9.775 9.849 9.936 9.866 10.023 TABLE 14 THE RACEMIZATION OP OPTICALLY ACTIVE PHENYLMETHYLCARBINYL CHLORIDE IN PHENOL INITIAL CONCENTRATIONS: 2.50 M. IN EACH REACTANT _____________Temperature =_40°____________ Time, Optical Rotation, †I Minutes Degrees Y x 10^ x x 103 0 5.00 6.67 8.33 10.00 11.67 13.33 16.67 20.00 23.33 26.67 30.00 33.33 36.67 40.00 43.33 46.67 50.00 53.33 56.67 60.00 63.33 66.67 5.15 4.89 4.80 4.72 4.63 4.55 4.48 4.33 4.18 4.02 3.88 3.72 3.63 3.48 3.37 3.25 3.13 3.02 2.88 2.77 2.67 2.55 2.47 4.502 4.582 4.545 4.622 4.608 4.542 4.519 4.533 4.612 4.616 4.710 4.557 4.641 4.605 4.613 4.635 4.636 4.733 4.752 4.755 4.821 4.786 1.686 1.684 1.677 1.674 1.667 1.661 1.650 1.641 1.631 1.621 1.612 1.602 1.592 1.583 1.574 1.565 1.557 1.548 1.539 1.531 1.523 1.514 Time, Minutes 0 16.67 20.00 23.33 26.67 30.00 33.33 36.67 40.00 46.67 53.33 60.00 66.67 73.33 80.00 86.67 93.33 100.00 106.67 113.33 120.00 126.67 133.33 Temperature = 300 Optical Rotation, Y x 103 Degrees 5.26 4.87 4.79 4.68 4.62 4.53 4.46 4.39 4.33 4.20 4.04 3.91 3.79 3.65 3.50 3.35 3.22 3.07 2.97 2.84 2.72 2.62 2.50 2.0072 2.0325 2.1745 2.1121 2.1623 2.1500 2.4415 2.1128 2.0Q43 2.1489 2.1466 2.1359 2.1642 2.2125 2.2603 2.2833 2.3390 2.3268 2.3621 2.3866 2.3896 2.4233 X x 103 1.0168 1.0130 1.0090 I.OO52 1.0013 0.9976 .9940 .9903 .9829 .9758 .9693 .9624 .9555 .9489 .9423 .9360 .9300 .9238 .9177 .9117 .9055 .9000 TABLE 14 CONTINUED Time, Minutes 73.33 80.00 Temperature = 40° Optical flotation, Y x 103 Degrees 2.21 2.06 5.010 4.974 X x 10? 1.498 1.483 Time, Minutes 140.00 146.67 160.00 166.67 Temperature = 30° Optical Rotation, Y x 103 Degrees 2.39 2.28 2.08 1.98 2.4471 2.4756 2.5181 2.5457 X x 103 0.8943 .8891 .8781 .8730 TABLE 15 THE RACEMIZATION OF OPTICALLY ACTIVE PHENYLMETHYLCARBINYL CHLORIDE IN p-CRESOL TEMPERATURE: 50° CONCENTRATION: 2.50 M. IN EACH REACTANT Time, Minutes Run 1 Optical Rotation, Degrees 0 1.67 3.33 5.00 6.67 8.33 10.00 11.67 13.33 15.00 16.67 18.33 20.00 21.67 23.33 25.00 26.67 28.33 30.00 31.67 33.33 35.00 6.55 6.43 6.31 6.18 6.07 5.97 5.83 5.72 5.59 5.47 5.38 5.20 5.10 5.01 4.91 4.83 4.69 4.60 4.47 4.35 4.30 4.23 X x 10-5 o Y x 10-3 Run 2 Optical Rotation, Degrees ~ Y x lO'5 1.840 1.826 1.8l4 1.811 1.804 1.799 1.791 1.785 1.780 1.773 1.767 1.762 1.756 1.751 1.744 1.739 1.733 1.728 1.722 1.716 1.711 4.790 4.865 4.844 4.957 4.837 5.057 5.042 5.163 5.216 5.127 5.466 5.430 5.367 5.366 5.292 5.444 5.4l8 5.527 5.614 5.482 5.429 6.54 6.41 6.25 6.14 5.97 5.88 5.75 5.62 5.54 5.42 5.31 5.18 5.08 4.93 4.83 4.73 4.63 4.53 4.44 4.32 4.23 4.14 5.228 5.916 5.478 5.939 5.544 5.591 5.642 5.407 5.438 5.427 5.523 5.485 5.662 5.641 5.628 5.624 5.630 5.607 5.687 5.677 5.674 ? TABLE 15 CONTINUED Time, Minutes Run 1 Optical Rotation, Degrees X x 10^ Y x 103 36.67 38.33 40.00 41.67 43.33 45.00 46.67 48.33 50.00 51.67 53.33 55.00 56.67 58.33 60.00 63.33 66.67 73.33 80.00 4.13 4.04 3.94 3.84 3.75 3.65 3.58 3.48 3.38 3.32 3.25 3.18 3.11 3.02 2.94 2.81 2.64 2.39 2.15 1.705 1.700 1.695 1.690 1.685 1.680 1.675 1.670 1.665 1.660 1.655 1.650 1.645 1.640 1.636 1.626 1.617 1.598 1.581 5.462 5.474 5.515 5.568 5.592 5.647 5.625 5.682 5.748 5.711 5.706 5.707 5.708 5.765 5.798 5.803 5.919 5.971 6.049 Run 2 Optical Rotation, Degrees Y x 10^ 4.04 3.95 3.86 3.77 3.68 3.58 3.49 3.40 5.702 5.714 5.725 5.7^0 5.763 5.816 5.845 5.878 3.07 3.01 2.96 2.90 2.83 2.72 2.59 6.158 6.127 6.076 6.055 6.063 6.016 6.034 TABLE 16 THE RACEMIZATION OF OPTICALLY ACTIVE PHENYLMETHYLCARBINYL CHLORIDE IN o-CRESOL TEMPERATURE: 50° CONCENTRATIONS: 2.50 M. IN EACH REACTANT Time, Minutes 0 20.00 23.33 26.67 30.00 33.33 36.67 40.00 43.33 46.67 50.00 53.33 56.67 60.00 66.67 73.33 80.00 86.67 93.33 100.00 106.67 113.33 120.00 Run 1 Optical Rotation, Degrees 6.57 6.13 6.08 6.01 5.91 5.86 5.81 5.74 5.69 5.61 5.56 5.50 5.45 5.38 5.28 5.16 5.04 4.91 4.78 4.66 4.56 4.46 4.35 X x 103 0.3543 .3538 .3534 .3529 .3522 .3518 .3515 .3510 .3505 .3500 .3493 .3490 .3487 .3478 .3468 .3460 .3451 .3441 .3433 .3425 .3415 .3407 Y x 103 1.5055 1.4432 1.4511 1.5330 1.4905 1.4560 1.4663 1.4417 1.4697 1.4502 1.4474 1.4323 1.4460 1.4239 1.4305 1.4387 1.4596 1.4851 1.4920 1.4868 1.4842 1.4925 Run 2 Optical Rotation, Degrees Y x 103 6.58 6.16 6.08 6.02 5.95 5.88 5.84 5.77 1.4325 1.4706 1.4484 1.4573 1.4650 1.4129 1.4265 5.64 1.4348 5.52 1.4303 5.42 5.26 5.15 5.07 4.94 4.81 4.69 4.55 4.48 4.35 1.4037 1.4588 1.4523 1.4150 1.4365 1.4582 1.4710 1.5018 1.4727 1.4975 TABLE 16 CONTINUED Time, Minutes 126.67 133.33 140.00 146.67 153.33 160.00 166.67 173.33 180.00 186.67 193.33 200.00 206.67 213.33 220.00 233.33 Run 1 Optical Rotation, Degrees 4.25 4.16 4.03 3.92 3.82 3.72 3.64 3.21 3.07 2.93 2.73 X x 103 Y x 103 0.3399 .3391 .3382 .3372 .3364 .3356 .3348 .3340 .3332 .3322 .3315 .3307 .3299 .3292 .3284 .3267 1.4944 1.4888 1.5200 1.5293 1.5359 1.5438 1.5390 Run 2 Optical Rotation, Degrees Y x 103 4.15 1.5015 3.96 1.5034 3.75 1.5263 3.55 1.5462 3.35 1.5707 3.21 1.5585 3.03 1.5788 1.5689 1.6092 1.5986 1.5940 1.6346 TABLE 17 THE RACEMIZATION OP OPTICALLY ACTIVE PHENYLMETHYLCARBINYL CHLORIDE IN MESITOL-p-XYLENE SOLUTION AND 2,6-DIMETHYLPHENOL-p-XYLENE SOLUTION TEMPERATURE: 50° CONCENTRATIONS: 2.50 M. IN EACH REACTANT Run 1 (Mesitol) Time, Optical Rotation, Minutes Degrees 0 990 1200 1380 1680 2280 2700 3000 3810 4i4o 5130 5.30 5.07 5.01 4.98 4.93 4.81 4.72 4.65 . 4.48 4.41 4.22 Run 2 (Mesltol) Time, Optical Rotation, Minutes Degrees 0 510 1170 6.20 6.00 5.80 1530 i860 5.65 5.55 5.40 5.35 5.30 5.18 5.14 5.05 4.85 4.72 2520 3000 3270 4200 4320 4800 5400 5880 = 1 77 y lfl-5 1 . mol®-1 Run 3 (2,6-Plmethylphenol) Time, Optical Rotation, Minutes Degrees 0 60 180 300 540 720 1460 1680 2070 2700 2940 3180 3570 6.47 6.40 6.37 6.32 6.18 6.10 5.77 5.67 5.57 5.25 5.15 5.05 4.94 TABLE 18 EXPERIMENTAL ACTIVATION ENERGIES AND CALCULATED ENTROPIES OP ACTIVATION FOR THE ALKYLATION OF PHENOL BY PHENYLMETHYLCARBINYL CHLORIDE AND FOR THE RACEMIZATION OF PHENYLMETHYLCARBINYL CHLORIDE IN PHENOL AEexp, kcal ASact> (at 313°A.) Temperature, °A. 303 313 323 ka x 105 (l. mole-1 min."1) 95.5 157 397 13.8 +0.2 -29.0 kp x lo5 (l. mole"-*- min."1) 86.3 255 400 15.4 +o.2 -25.0 kjjci x 405 438 1075 7.3 + 0.2 -47.5 (l* moleâ€'*' mln."-*-) TABLE 19 SUMMARY OF THE POSSIBLE MECHANISMS FOR THE RACEMIZATION OF PHENYLMETHYLCARBINYL CHLORIDE IN A SERIES OF PHENOLS Mechanism Rate Expression 1 _ g. Jy g.1] = k [PhOH] [R*C1 ]+ k. p[A-PhOH] [R*C1] dt A.P. kHCl ^HCl5 £R*C13 + ka tph0H^ [R*C1] 2 _ gJfl*g3J. = krl [R*C1] + k^G1 [R*C1] [HC1] + ka [R*Cl] [PhOH] dt 3 _ g [R*C1J = k dt rl * - d ,.[R*C1] = k [R*ci] dt F 5 _d 6 _ g IR*C. 1.1 = k^cl [R*C1] [HC1] + k& [R*C1] [PhOH] dt 7 _ g. ,[R*C1] s k [PhOH] [HC1] [R*C1] dt F3 K [R*C1] [PhOH] cL [R*ci] + KP + [R*C1] [PhOH] + k [R*Cl] [PhOH] a [PhOH] + kA A *F * [R*Cl] [PhOH] + k[R*Cl] [PhOH] a [R*Cl] a k [r #c i ]+ kQ [PhOH] [R*Cl] dt rl a + k, p o [A-PhOH][HCl] [R*C1] A.P.3 + TABLE 20 SUMMARY OP RACEMIZATION RATE CONSTANTS FOR RACEMIZATION OF PHENYLMETHYL CARBINYL CHLORIDE IN PHENOL, £-CRESOL, o-CRESOL, 2,6-DIMETHYLPHENOL AND MESITOL, CALCULATED FOR MECHANISM 1 Compound kP x 105 (1. mole“l min.-l) kHCl x 105 (1 . mole~l min.~l) *kx x 105 (l. mole"* min,"I) Phenol (Run 1) Phenol (Run 2) Phenol (Avg.) 435 366 400 1060 1089 1075 1495 1455 1475 £-Cresol (Run 1) £-Cresol (Run 2 ) £-Cresol (Avg.) 284 330 307 942 652 797 1225 982 1104 o-Cresol (Run 1} o-Cresol (Run 2) o-Cresol (Avg.) 98.0 96.8 97*4 218 202 210 316 299 307 2,6-Dimethylphenol 1.96 Me3itol (Run 1) Mesitol (Run 4) Mesitol (Avg.) 1*77 2.06 1.97 Phenol (40°) Phenol (30°) 255 86.3 * Sura of kHcl and kA p for Mechanism 1. • 438 405 693 491 VITA William Ludwig Spliethoff candidate for the degree of Doctor of Philosophy Major Subject: Organic Chemistry Minor Subjects: Inorganic and Physical Chemistry Biographical Data: Date of Birth: April 8, 1926, Matamora.c3, Pennsylvania Education: B.. S.. in Chemistry, The Pennsylvania State College, 1946 M. S. in Fuel Technology, The Pennsylvania State College, 1948 Additional Graduate Study, University of Michigan, 1948-50 Michigan State College, 1950-52. Experience: Graduate Assistant, The Pennsylvania State College, 1946-48 Teaching Fellow, University of Michigan, 1948-49 Research Fellow, University of Michigan, 1949-50 Graduate Assistant, Michigan State College, 1950-52. Professional Affliations: American Chemical.Society Phi Lambda Upsilon The Society of the Sigma Xi