THE GR'I‘HO ALKYLATICDN OF PHENOIS WITH AROMAYIC ALCOHOLS IN THE PRESENCE OF p-TQLUENESULFONIC ACID Thesis for fha beam at Ph. D. MICHIGAN STA'IE UNIVERSITY Herbert Bowers Rickert I962 I‘ IVA'V -l-‘ Ir‘,’ I"* -‘y‘-\_ ‘I u‘:’ 5’ ~- ' . ‘- ‘I 4' ’ . v I b I'.i...~.\..II‘ I.."" ‘V‘t/ 12‘fi..~-f'v‘-—n\{ CLII‘||\I.;..-W: ‘ I' VI 1‘ I...II DI E:’.ST LAIQH-IC, 5.11;:1'L1NXN _ I \. THE CRTIIC‘ ALFYII‘ITICN CF PIHL‘I‘ICIS WITH AROMATIC AICCHCIS IN THE PZESETCE CF p—TCIUENESULFOEIC ACID By Herbert Bowers Rickert A THESIS Submitted to Kichigan State University . in rartial fulfillment of the requirements for the deéree of DCCTCR CF PHILOSOPHY Department of Chemistry 1962 The author wishes to exoress his deep aoorecjation for the help enfl euidance of Professor Ralph L. Guile durinr the course of this investimation. ii Dedication To fly WifC and children for enoourarement and understandin: beyond the call of duty. iii ABSTRACT THE CRTHC ALKYIATIC‘T' CF PHENC‘IB WITH ARCNATIC ALCCHCLS IN THE PRESENCE OF p-TOLUENJSULFONIC ACID by Herbert Bowers Rickert A new method for the ortho alkylation of nhenols byznbmatic alcohols has been discovered. The aromatic ahxmcl and nhenol were refluyed in cyclohexane with a Irtoluenesulfonic acid catalyst to yield water of reaction and ortho-alkylated products. Smaller amounts cf the para isomers were formed. When benzyl alcohol amirmenol were the reactants, as much as 4.4 parts cm o-benzylphenol for each one part of n—benzylnhenol were obtained. Benzyl alcohol may also be used to alkylate sub~ :mituted phenols. In particular the benzylation of the three isomeric cresols and the six isomeric yylenols was studied. Good yields of ortho—benzylated products vere obtained with all the cresols. In the benzylation of the six xylenols, good yields of the ortho isomers were obtained except with 2,6-xylenol. In the latter case ortho alkylation is not possible and both the 5-znm.4-benzyl nroducts were formed. Substituted benzyl alcohols may also be used for the acid-catalyzed ortho alkylation of phenols. Phenol vms successfully alkylated with p—methyl-, p-isoprOpyl-, R) Herbert Bowers Rickert gwchloro-, p-bromo—, o-chloro-, and L,4-dichlorohenzyl zflcohols. The ortho/Dara ratios are highest (4.5- 5.5fl) for alcohols containinr electron-releasing sub— mfituents. Ortho/para ratios for alcohols containing mflbaen (electron—withdravinq) substituents are in the cnder of 2.5—5.0/1. The rate of the reaction may be determined by umasurinr the amount of water removed per unit time. Eanyl alcohols containint electron-releasing substit— Lwnts react faster than benzyl alcohol itself; alcohols unmainina electron—Withdrawinq groups react slower. A mechanism is nrooosed whereby ortho alkylation Inbceeds via a cyclic intermediate comolex. Para alkylation is ssumed to take place via free carbonium ions. Acid-catalyzed ortho alkylation may also be carried mnzat 140—15000. without a solvent in the nresence of Irtoluenesulfonic acid. The yields by this method are as good as with the solvent method. Other catalysts such as Dowex SO and benzenesulfonic achinmy also be used, but somewhat lower ortho/para ratios are obtained. The benzylated nroducts may be analyzed by means Of their infrared spectra. The C—H stretching bands atEh7-5.9;o, the 0-H deformation bands at 8.2—8.3AL and the out-of—plane C—H deformation bands at 10-1%;L \ Herbert Bowers Rickert mmithe most useful for oualitative analysis. The ihmter bands are most valuable for the quantitative estimation of isomers. There is no other mood method known for the orep- anfiflon of ortho—alkylated phenols from aromatic alcohols. 33m only other general method described for the ortho tmnzylation of phenols is that of Claisen. In this latter method benzyl halides and not the alcohols are tmaalkylatin: agents. Many substituted aromatic alcohols are now avail- flfle from the reduction of readily available aroratic acids or aldehydes by means of comnlex metal hydrides. host of the substituted benzyl alcohols used in this research were nrenared by the sodium borohydride reduction of the corresnondin: benzaldehyd s. Thus the acid-catalyzed alkylation of phenols with aromatic alcohols should orove a valuable synthetic tool for the organic chemist. II. TAPIE CP cc “V“m” ILAJLIL‘J H r J I—S :U f) U L. O #3 H 0 2 C31 fII H 01 F3 (3 AL . Benzyl Phenols Benzyl Cresols Penzvl XVIenols llhvlation of P‘erol “it“ Substituted Perzyl Alcohols 3nd Ha lides Biolo:ical activity if Crtho Penzyl Phenols C) CU 'J' m EXPEYIT 'FI‘TTAL A. Infrared Analeis 1. General letiod 2. \nalysis of Benzvl Alcohols 3. Analysis of Ponzvl Phenols 4. inalvsi of Phenyl Urethanes B. Reaients and laterials 1. Commercial Chemicals Used nithout Purification 2. Commercial Chemicals Further Purified Preraration of the Pubstitutcd Benzyl Alcohols C. Alkylations . Penzvlation of Phenol a. Penzvlation with a n—toluene sulf cnic acid catalyst b. Penzflation With mineral acid catalysts c. Penzvlaticn with a Dower 50 \N H catalyst d. Penzenesulfonic acid ca talvst e. lttemnted alkylation ' itho ut a catalyst f. Identification of products a. Penzylation of Cresols a. Benzyl alcohol and n—cresol b. Penzyl alcohol and o—cresol c. Penzvl alcohol and m-cresol iv Pare I-‘ I—4 m w \fi H e4 ‘~ ;) ‘J m m n) m PWI‘OI—‘F—J fix I‘. R ) J‘\ n s (:0 \N O 'b I )I 3 46 47 as 40 [+0 / 51 q L. 'F o (DUI ;. Penzvletion of 7vlen 13 a. Ponzylztion of L,3-vvlepcl 41 b. Pcnzylation of 2,4—xylenol 65 c. Penzylation of 3,5-yylenol 65 d. Rep""lotio of 2,5—Xylonol 7 (a. P(HEZ l.fi'ion cf 5,4«orrlenol 79 f. IN"7"1-tlcs1- no reaction p—toluene sulfonic acid Alkylation with benzyl ether To one mole of phenol and 0.5 mole dibenzyl ether was added 400 ml. cyclohexane and 0.05 mele p—tolueneC. sulfonic acid. The reaction mixture was heated to reflux teraerature, and heating was continued for seventy—four minutes during which time 8.4 ml. of water was collected. As 0.9 ml. of the water was from the catalyst, a net amount of 7.5 ml. water was formed (83% of the calculated amount). The reaction mixture was treated as before to yield 41.? g. of o—tenzylphenol and 15.2 g. of p-benzyl: phenol. Based on the actual amount of water formed, this is equivalent to a 57% yield of monobenzylnhenol and an ortho/para ratio of 2.7/1. ‘/_.;\ {fillli} H -O-CH O 2 2 80 0‘ o-benzylphenol -————————4F-and p—benzylphenol +_ n-toluene sulfonic acid .0... 44 Rearransenent of benzy1_phenyl ether B;nzy1 phenyl ether was prepared from sodium phenate and benzyl chloride by the method of Short and Stewart (75). The crude product was distilled at 5 mm. to yield a fraction of benzyl nhenyl ether boiling at 152—500. This fraction solidified to give white crystals with a r .uu. of 58.5—9.000. Short and Stewart (75) give a n.p. o . Cfi 59 C. for this compound. Examination of the infrared spectrum of benzyl phenyl ether shows the absence of an O-H stretching band at 2.75fl¢. ‘The stronr band at 8.15;L is characteristic of aromatic ethers. Benzyl phenyl ether (0.89 mole) was mixed with 400 m1. cyclohexane and 0.07 mole of p—toluenesulfonic acid. The reaction mixture was refluxed for 5 hours and 25 minutes durina which tine 0.8 m1. of water was renoved from the catalyst. The mixture was cooled and mashed in the usual manner with sodium bicarbonate solu- tion followed by water. The cyclohexane was removed under aspirator vacuum. Toward the end of the distillation some water backed up into the still not and the distillation blew out at the still head. Some material was lost, but 140.5 g. of 45 a viscous oil was left in the flask. Since 164.2 g. of benzyl phenvl ether had been used in the reaction, up to 25.7 a. of oil may have been lost. Thus, at least 85% of the crude product was saved. The viscOus oil was distilled with a two foot Viareux column to give 5.0 g. of a white solid, b.p. 105-1150C. at so 31111., identified as phenol. The residue was distilled with a three foot Vixreux column to give 24.2 Q. f o—benzylphenol and 20.2 a. of p-benzylphenol. This corresponds to an ortho/para ratio of l.2/l. The rest of the material (105.8 a.) did not distill below 16500. at 2; mr“. o—benzylphenol QOOn p-benzylphenol 0 Hg: 0 L” V. =— polybenzyl phenols n—toluene and phenol sulfonic acid Since phenol is much lower boiling than the benzyl— ated phenols, it is possible that most of the 25.7 g. thich was lost was phenol. In effect the phenol could have been rapidly steam—distilled out of the flask when the explosion occurred. If the 25.7 s. lOSt W38 phenol, then a total of L8.7 q. of 0.51 mole of phenol was pro- ddced during the rearrantement. If all the residue (195.8 g.) was dibenzylated product, it would be equiva— lent to 0.59 mole, and if trihenzylated product, equiva— lent to 0.29 mole. Thus the residue likelv contains a mixture of di~ and tribenzylated products. Of course 46 some polvmerso of the her zyl fratrent could also be present. b. Bensylation with mineral acid catalysts Two experirents were carried out in which a strong mineral acid was used as a catalyst for the alkylation of phenol with benzyl alcohol in a cvclohexane solvent. PhOSphoric acid catalyst To 500 m . cyclohexane and 1.5 moles phenol was added 0.1 mole 85% phosphoric acid. This mixture was heated to reflux temperature and 0.75 mole benzyl alcohol added over a period of eighty minutes. Durinm the reac- tion the phosphoric acid apneared to be pre se UE, d‘ as a separate phase. Heatinm was continued for three hours and forty five minutes during which time 14.7 ml. of water' as collected. The reaction mixture was treated as before and the crude product distilled to yield 11.9 g. of o-benzle phenol and 54.? a. of p—benzylphenol. CHZOH 8000. ‘_ o-benzylphenol H ”O ” p-benzylphenol 5‘ 4 .ulfuric acid catalyst To 250 ml. cvclohexane and one mole phenol was added 250 ml. of 67% sulfuric acid. A two phase system resulted. This mixture was heated to reflux temperature. 47 Benzvl alcohol (one mole) was added over a period of eichty—six ti nutes durini vigorous agitation. Heating van continued for an additional hour and the reaction nnxture allowed to cool. The upper (cyclohexane) layer RES separated and treated as before. Fractionation of the crude product gave 67.0 g. of o-benzylphenol (56% yield). Less than 5 g. of material distilled below 15100. at 5 mm. in the boilirg ranfie of o-benzylphenol CHZOH 8000. {:7 p-benzylphenol H2004 c. Bengylation with a Done; 50 catalyst Solvent method Phenol (5 moles), 1.5 moles of benzyl alcohol, 50 g. of Dowex 50, and 500 :l. of n— —he otane were mix ed and refluxed for forty- -fivc minutes until all the water from the Dowex 50 had been collected. Refluxina was cortinued for eio}ty—ei2ht minutes, but very little water.flag collected. is the reaction rate und er these conditiong was very slow, the experiment was discontinued High temperature method Phenol (2 moles) and 40 g. Dowex 50 were heated for one tour at ll7-l5OOC. at the end of which time all the water present in the catalyst had been removed. Heating 48 at 149-15100. was continued uhile one mole of benzyl alcohol was aide; over a oeriod of one hour. As soon as the last of the benzyl alcohol had been added, the meter stopped distilling. The reaction mixture was allowed to cool to 9000., ard the Dowex 50 was removed by filtration. Unreacted ohenol was reroved from the filtrate by vacuum distilla- tion. The residue was fractionated, using a three foot Vigreux column, to give 77.5 g. of o-benzylphenol and 55.9 g. of p-benzylohenol. This is a 60% yield of monobenzylphenols and an ortho/para ratio of 2.5/1. A similar run was carried out with three moles of phenol, one mole of benzyl alcohol (added over two hours and 28 minutes), and 10 g. of Dowex SO. Fraction- ation of the product yielded 95.2 g. of o-benzylphenol and 59.9 a. of n—benzylnhenol. This is a 75% yield of monobenzylphenols with an ortho/para ratio of 2.4/1. d. Benzenesulfonic acid catalyst Phenol (1.5 moles), benzenesulfonic acid (0.1 mole) and 500 ml. cyclohexane were mixed and heated to reflux temnerature. Benzyl alcohol (0.75 mole) was added over a period of two hours and 27 minutes, after which refluxing was continued for 29 minutes. The reaction mixture was washed as usual, and solvent and unreacted ohenol were removed by vacuum distillation. Fractionation of the residue yielded 48.6 g. of o—benzylnhenOl and 20.2 g. of o-benzylphenol. 49 This is a 50% yield of monobenzylphenols and an ortho/ para ratio of 2.4/1. e. Attempted alkylation without a catalyst Phenol (2 moles) was added to one mole of benzyl alcohol and the mixture heated at 168-175OC. for 41 minutes. No water was formed durina this time; therefore it was assumed no alkylation took place. f. Identification of products A sample of p—benzylphenol, prepared by the p-toluenesulfonic acid method, was recrystallized four times from liaroin (60-9000.) to give white crystals, m.p. 84.500. This corresponds to the m.p. of 8400. found by Paterno (15). The sample of p—benzylphenol obtained in this work was mixed with an authentic sample of p-benzylphenol and a melting point was taken. There was no depression of the meltinq point. The infrared spectrum of the prepared sample of p-benzylphenol was compared with the infrared spectrum of an authentic sample of p—benzylphenol. The spectra were identical. The infrared spectrum of p-benzylphenol is consis— tent with its structure. The O-H stretching band at 2'2” has a hydrogen—bonded shoulder at 2.85;; which is tYDical of phenols with no ortho substituents. The aromatic substitution band at 12.4é;¢ is typical of v .' g. . . 50 compounds which have 1,4-aromatic substitution. The recrystallized sample of p—benzylphenol was reacted with phenyl isocyanate to aive a phenylurethane vmich on recrystallization from ligroin (60—9OOC.) melted at 155.5—4.o°c. This urethane is not recorded in the literature. Analysis Calculated for C2OH17NO2 : N, 4.65 Found : N, 4.68 A sample of o—benzylphenol, prepared by the p-toluenesulfonic acid method, was distilled three times with a three foot Viareux column to yield a fraction boiling at l42-5OC. at 5 mm. This fraction solidified to yield a white solid with a m.p. of 55-40C. Claisen and coaworkers (52) give 5200. as the melting point of o-benzylphenol. The infrared spectrum of o—benzylphenol shows the expected differences from that of p—benzylphenol. The O-H stretching band at 2.75;; contains no hydrogen bonded shoulder as does the O-H band of p-benzylphenol. The ortho alkyl group decreases the intermolecular hydrosen bonding and thus causes the change in the spectrum. The 12.45%; band present in the Spectrum of P-benzylphenol is absent. Instead there is a band at 13.5£L which is typical of compounds with 1,2—aromatic substitution. The purified sample of o-benzylphenol was reacted 1.. '.4 '4‘ 51 “nth phenyl isocyanate to dive a phenylurethane, which on recrystallization from liqroin (60-9000.), melted at 120°C. Claisen and co-workers (52) give the melting point of this phenylurethane as 117.5—8.OOC. Purified o—benzylphenol (0.25 g.) was mixed with purified p—benzylphenol (0.25 g.) and ground together in a mortar with a pestle. The melting point of the ndxture was 58—420C. 2. Benzylation of Cresols a. Benqyl alcohol and p:cresol Two moles of p-cresol, 0.002 mole p-toluenesulfonic acid, and 40 ml. cyclohexane were heated for four hours at 127—4200. while adding 0.7 mole benzyl alcohol. During the reaction period 15.2 ml. of water were collect- ed by distillation. Purification and distillation of the reaction mixture yielded 87.1 g. crude 2-benzy1-4- methylphenol in 65% yield. The crude product was redis- tilled (126-80C. at 0.1 mm.) to yield white crystals of 2—benzyl-4-methylphenol, m.p. 55-600. Hickinbottom (45) records a melting point of 5600. for this compound. CH3 (DH 52 The infrared spectrum of 2-benzyl—4-methylphenol is consistent with its assigned structure. The O-H stretching band at 2‘2A¢ is of medium width with just a trace of a hydrogen-bonded shoulder. It is the same size and shape as that of 2-benzyl—5,5—dimethy1phenol, which also has one ortho benzyl group. The strong band at 12.4pc is characteristic of 1,2,4-aromatic substitution. Phenyl isocyanate was reacted with 2-benzy1-4— methylphenol to yield the phenylurethane as white crystals (from hexane), m.p. 147°C. Hickinbottom (45) records a m.p. of 146°C. for this urethane. b. Benzyl alcohol and o—cresol Three experiments were carried out in which benzyl alcohol was used for the ortho alkylation of o-oresol. One experiment was of the low temperature type, whereas the other two were high temperature alkylations. low temperature alkylation One and one—half moles of o—cresol, 0.75 mole benzyl alcohol, 0.1 mole p-toluenesulfonic acid and 500 ml. cyclohexane were placed in a one-liter flask. This mixture was heated to reflux temperature and heating was continued for two hours and 55 minutes, at the end of which time 14.7 ml. of water had been collected. Purification and fractionation yielded 59.2 g. of 2-benzy1-6-methy1phenol and 26.5 g. of 4-benzyl-2— methylphenol. 55 High temperature alkylation. Two moles of o—cresol and 0.02 mole p-toluenesulfonic acid were heated to 15000., and one mole of benzyl alco— hol added over a period of one hour and ten minutes at a.temperature of l41—55OC. Heating was continued for nine minutes at 150-200. During the reaction period 15 ml. of water was collected. Purification and fractionation yielded 85.5 g. of 2-benzyl—6-methylphenol and 50.1 g. of 4—benzyl-2- methylphenol. In a second experiment two moles of o—cresol, 0.002 mole p-toluenesulfonic acid, and 45 m1. of cyclo- hexane were heated to 15000., and 0.75 mole of benzyl alcohol added over a period of four hours and 27 minutes at a temperature of 140-1510C. Heating was continued for ten minutes at 151-2OC. During this period 15.4 ml. of water was collected by azeotrOpic distillation. Purification and fractionation yielded 68.8 g. of 2-benzy1-6-methylphenol and 21.4 g. of 4-benzy1—2- methylphenol. Identification of Products 4-Bengy1-2-methy1phenol. Distilled 4-benzy1-2- methylphenol was redistillcd to yield a fraction boiling at 162-5OC. at 1 mm. This fraction was recrystallized tWice from ligroin (60-9000.) to obtain white crystals with a m.p. of 51.0-1.5OC. Huston (4) sives a m-p. of 49-5—5O.5OC. for this compound. ”H20” CH3 Examination of the infrared spectrum of 4—benzyl- 2-methy1phenol shows a medium width O—H stretching band at 2.70-5.05‘9 with a shoulder on the right due to some hydrogen bonding. This type of band is characteristic of phenols which have one ortho methyl substituent. The shape of this band is similar to the 0H band of 4—benzy1- 2,5-dimethy1phenol. The strong band at 12.4/0 is charac— teristic of the 1,2,4-substitution of the phenolic ring. Such a band is also present in 2-benzy1-4-methylphenol. 2—Benzy1—6-methylphenol. Distilled 2—benzy1~6— methylphenol was redistilled to yield a fraction boiling at 147.50C. at 1.8 mm. This white solid had a m.p. of 49-5000. as compared to the value of 49.5-50.500. recorded by Huston and co—workers (4). CM4 Qua-- The narrow O—H band at 2.75;; is characteristic of phenols containing two ortho alkyl groups. In the aromatic region the band at 15.5%; is characteristic 55 of phenols with 1,2,5—aromatic substitution. One-tenth mole of 2—benzyl—6-methylphenol was dissolved in 100 m1. of chloroform and cooled to 200. (hm-tenth mole of bromine in 50 m1. chloroform was added over a period of 47 minutes, maintaining the temperature at 2°C. The reaction mixture was aerated to remove hydrogen bromide and washed with dilute sodium bicarbon- ate. The chloroform was removed from the solvent layer by distillation. The crude 2-benzyl-4-bromo—6-methylphenol was distilled to yield a fraction boiling at 167-800. at 0.5 mm. This fraction yielded white crystals with a m.p. of 65.5-4.OOC. Huston and co-workers (4) give a m.p. of 65-400. for this compound. (DH I5r Summary of results The results of three alkylations of o-cresol with benzyl alcohol and a p-toluenesulfonic acid catalYSt are summarized in the following table. 56 Table IV. Benzylation of o—Cresol Exp. Moles Moles Moles M1. Temp. % Ortho/ o-cresol benzyl cata— cyclo- oC Mono- para alcohol lyst hexane ° benzyl ratio cresols o/p 1 1.50 0.75 0.1 500 85 57 2.5/1 2 2.00 1.00 0.2 None 141- 58 2.8/1 155 5 2.00 0.75 0.002 45 140— 61 5.2/1 151 c. Benzyl alcohol and m—cresol Two moles of m-cresol, 0.15 mole p-toluenesulfonic acid and 650 ml. of cyclohexane were heated to reflux temperature (81°C.) and one mole of benzyl alcohol added over a period of one hour. During the reaction period 18 ml. of water were collected by distillation. Purification and fractionation of the crude product yielded 101.2 g. of a fraction containing the two O-benzylated m-cresols and 18.5 g. of 4-benzy1-5- methylphenol. This is a monobenzyl yield of 59% and an ortho/para ratio of 5.5/1- The crude 4-benzy1—5-methy1phenol was recrystallized from hexane and found to have a m.p. of 95°C. Huston 57 and Houk (5) give a melting point of 950C. for 4-benzy1— 5—methylphenol. C343 There are two ortho isomers which were isolated by Huston and Houk (5) in the benzylation of m-cresol. 0H oH /__\ CH2' CH2. CH3 ‘c H5 2-benzy1-5-methy1phenol 2—benzy1-3-methylphenol m.p. = 46-700. m.p. = 71-200. Since these melting points are considerably lower than 95QS., the assignment of 4—benzyl—5-methy1phenol would seem to be correct for the 95°C. material. In addition this latter compound is the higher boiling fraction, and such a physical property would agree with its assignment. Both of the o-benzyl isomers are lower boiling since the presence of the o-benzyl groups lowers the boiling point by decreasinq intermolecular hydroqen bonding. That this assignment of the para isomer is correct may be seen by a study of its infrared spectrum. The large hydrogen bonded shoulder in the OH band (2.7—3.£“,) 58 is typical of phenols which do not contain an ortho substituent. An O-H band of similar shape is found in the spectrum of p—benzylphenol. The crude o—benzylated m-cresol containing two possible isomers was distilled at 1 mm. to give a color- less liquid boilinq at l48.0-0.50C. When this material was placed in the refrigerator a white solid formed. The solid was removed by filtration, dried and recrys— tallized to give crystals which melted at 68-7000. Further recrystallization from hexane save white crystals, th. 74.500. These crystals must be 2—benzy1—5-methy121 phenol, which melts at 71—200. according to Huston and Houk (5). 0H m. Cflds They could not be 2-benzyl—5—methylphenol, as this isomer melts at 46—7OC. according to Huston and Houk (5). Examination of the infrared spectrum of 2—benzyl- 5-methy1phenol shows this assignment to be correct. There is a strong band present at 15.ewb which is com- pletely absent in 4-benzyl-5-methylphenol. The strong band at 12-95pc is characteristic of 1,2,5-benzene sub— stitution; for example, it is found at 15.9“, in 2~benzyl~ 6-methylphenol. Furthermore, the 1,2,4-substitution band found at l2'éflb in 4-benzyl-2—methylphenol, and at 59 12.6”, in 2—benzyl—4—methylphenol, '3 completely absent from the spectrum of E-benzyl—B—methylphenol. The mixed meltinm point of 2-benzvl-5-methylphenol and 4-benzyl—5-methylphenol was found to be 60—8500. This value shows that these two materials are different compounds. If the infrared spectrum of the crude ortho benzyl material is examined, it is easily seen that another material in addition to either 2-benzyl-5-methylphenol or 4—benzyl-5—methylphenol is present. There is a band at 9.1;“,‘which is not present in either 4-benzyl-5— methylphenol or 2—benzyl-5—methy1phenol. Furthermore in 2-benzyl—5-methylphenol there are no bands in the 11.5-12-29L reaion. Since there are considerable diff— erences in this region between the crude ortho-benzylated m-cresol and 4-benzyl—5-methylphenol, the differences must be due to a third constituent. The 2-benzyl—5-methylphenol was isolated as a crystalline material which separated from the crude ortho benzyl material on standing. A portion of the mother liquor, after separation of the 2—benzyl isomer, was treated with bromine dissolved in carbon tetra- chloride. After evolution of HBr had ceased, the carbon tetrachloride was removed by distillation and the residue recrystallized from hexane to give white crystals, m.p. 102.5—5.5°C. This is apparently 2—benzy1-4,6-dibromo- S-methylphenol, which is reported by Huston and Houk (5) 60 to have a meltinw point of 102-10500. OH I \ CH3 ESr The strong band at 15.%;¢ of 2-benzyl-5-methy1phenol can be used to measure the percent of this isomer in the crude ortho—benzylated m—cresol fraction. This band is present in the crude fraction as a shoulder on the strong 15.79“; band. The pure 2—benzyl-3-methy1phenol gives an absorbance of 56% when measured from the base line, whereas this band in the mixture has an absorbance of 9% when measured from its base line. Thus the mixture contains about 25% 2—benzyl-5-methylphenol and 75% 2-benzyl-5-methylphenol. 61 5. Benzylation of Xylenols The alkylation of the xvlenols was carried out by the same general method used for the benzylation of phenol and described on p. 55. an Benqylation of 2,5:xylenol A solution of 122.2 5. (one mole) 2,5—xy1enol and 5.8 g. (0.02 mole) p—toluenesulfonic acid in 500 m1. cyclohexane was heated to reflux temperature and the water was removed from the acid catalyst. To this solu- tion was added 54.1 g. (0.5 mole) benzyl alcohol over a period of six hours and 50 minutes. The mixture was then refluxed for an additional one hour and 48 minutes. Durinf this time 8.5 ml. of water were collected. After washinq the reaction mixture as usual, distillation yielded 70.2 g. of 6-benzyl—2,5—dimethylphenol and 10.4 g. of 4-benzyl-2,5-dimethylphenol. This corresponds to a monobenzyl yield of 76% and an ortho/para ratio of 6.7/1. CN4 OH ‘/, CH3 ' C+4 /// \ //\\(:H \\ 3 Cid 3 I \:—> [[:)[CH CH2 \. 3 //l \. The crude 6-benzyl~2,B—dimethylphenol was redistilled to yield a White solid boilina at leo-l°c. at 2 mm. and melting at 58-6OOC. This redistilled material was 62 recrystallized from ethanol-water to give white crystals with a melting point of 64.500. Analysis: Calculated for C H 0 : C, 84.88; H, 7.60 15 16 Found : C, 84.88; H, 7.66 Examination of the infrared spectrum of this compound shows a narrow C—H stretching band at 2.7;a. which is characteristic of phenols substituted in both positions ortho to the OH group. The band at 12-5ép6 is characteristic of 1,2,5,4-substituted benzenes. The phenylurethane of 6-benzyl-2,5-dimethylphenol was prepared by reaction of the phenol with phenyl isocyanate. It was recrystallized from hexane to give white crystals, m.p. 156.5-157.OOC. Analysis: Calculated for C22H20NO2 : N, 4.24 Found : N, 4.46 The crude 4—benzyl—2,5—dimethylphenol was a white solid with a m.p. of 84-800. It was recrystallized once from ethanol—water and again from hexane to give white crystals , m .p. 104.00C . Analysis: 0 , (C2,. Calculated for Cl5Hl6O . C, 84.-8, H, 7.60 Found : C, 85.58; H, 7.55 Examination of the infrared spectrum of this compound shows an O-H stretching band of medium width at 2.7%;c. Such a band is characteristic of phenols 65 substituted in one of the positions ortho to the hydroxyl group. The band at 12.55;; is characteristic of a 1,2,5,4—substituted benzene. b. Benqylation of 2,4—xylenol A solution of 100 g. (0.818 mole) 2,4-xy1enol, 59.1 g. (0.547 mole) benzyl alcohol, 9.5 g. (0.05 mole) p—toluenesulfonic acid, and 550 ml. cyclohexane was refluxed for one hour and 47 minutes. During this time 9.8 ml. of water were collected. The reaction mixture was washed as usual and the solvent and unreacted xylenol removed by distillation. The residue was distilled at 5 mm. to yield 48.5 g. of 2-benzyl—4,6-dimethylphenol, a white solid. This is a.42% yield. The solid was recrystallized repeatedly from petroleum ether (60—9OOC-), benzene, and cyclohexane, but the crystals would always "felt out" as described by Houston and Houk (5). Most of the material was lost during recrystallization, so another experiment was carried out in which the product was purified by dis~ tillation. A solution of 98. g. (0.8 mole) 2,4-xylenol and 5.8 a. (0.02 mole) p-toluenesulfonic acid in 500 ml. cyclohexane was heated to reflux temperature and the water from the catalyst removed. To this solution was added 54.1 g. (0.5 mole) benzyl alcohol over a period of four hours and 55 minutes. The mixture was refluxed 64 for an additional six hours. During this time 8.7 m1. of water were collected. The mixture was washed as usual and the solvent and unreacted 2,4—xylenol removed by distillation. The crude product solidified on coolinq. The crude product was distilled to yield 64.7 g. of a white solid boilinw at 159-16800. at 2 mm. This corresponds to a yield of 61% of 2—benzyl—4,6-dimethy1: phenol. CH4 QCHZ /' CH3 \\ CH3 The product was redistilled at 2 mm. to give 59.7 g. of a white solid boiling at l59.5~16l.5OC. and melting at 65.0-65.500. This material was again redistilled at 2 mm. to give 26.1 g. of a white solid boilinm at 160.5—1.OOC. and melting at 660C. Elkobaisi and Hick— inbottom (46) record a melting point of 6700. for 2-benzy1-4,6-dimethylphenol. The infrared spectrum of 2-benzyl-4,6-dimethy1phenol was consistent with the assianed structure. It showed a narrow O-H stretching band at 2.79;; which is charac— teristic of phenols with two alkyl groups in the ortho positions. The strong band at 12.65;¢ is characteristic of 2,4,6—trialkylphenols (l,2,5,5-benzene substitution). 65 A phenylurethane of 2-benzy1-4,6-dimethylphenol was prepared by reaction of the phenol with phenyl isocyanate. The crude phenylurethane was recrystallized from ligroin (60-9000.) to yield white crystals, m.p. 145.5-4.OOC. Elkobaisi and Hickinbottom (46) record a meltinm point of 15600. for the phenylurethane. c. Benzylation of 2,5—xy1enol Experiment No.1. A solution of 100 g. (0.818 mole) 2,5—xy1enol, 9.5 a. (0.05 mole) p—toluenesulfonic acid, and 550 m1. cyclohexane were heated to reflux tempera- ture. During a period of two hours and 29 minutes, 59.1 g. (0.547 mole) benzyl alcohol were added while refluxinm was continued. A total of 10.1 ml. of water was collected. The reaction mixture was washed as usual, and the cyclohexane was removed under reduced pressure. When the residue was distilled to remove unreacted 2,5-xy1enol, the vacuum take-off became plugged with crystals of the xylenol. To prevent this, 100 m1. of 1,2,4—trichlorobenzene was added. Since 1,2,4—trichloroc benzene has about the same boiling point as the xylenol, the mixture co—distilled without further difficulty. The crude product was distilled at 5 mm. to yield 45.1 g. of 2-benzyl-5,6—dimethy1phenol and 15.6 g. of 4-benzyl-2,5—dimethy1phenol. This corresponds to a monobenzyl yield of 51% and an ortho/para ratio of 5.5/1. 66 The crude 2-benzy1-5,6-dimethylphenol was recrys- tallized once from cyclohexane—ligroin (90-12000.) and twice from ligroin (60-9000.) to yield white crystals “mich melted at 75.0-75.500. Oti CZH3 Analysis: Calculated for 015H16O : C, 84.88; H, 7.60 Found : C, 84.84; H, 7.51 Examination of the infrared spectrum of this com- pound showed a narrow O—H stretching band at 2-7QLL which is characteristic of phenols substituted in both positions ortho to the OH group. The strong band at 12.45LL is characteristic of compounds which have 1,2,5,4-benzene substitution. A phenylurethane was prepared by reaction of 2-benzyl-5,6-dimethylphenol with phenyl isocyanate. The crude phenylurethane was recrystallized from lirroin (60-9000.) to yield white crystals, m.p. 150.0— 150.500. Analysis: Calculated for 022201102 : N, 4.24 Found : N, 4.58 The crude 4-benzy1~2,5-dimethylphenol was recrys— tallized once from benzene-cyclohexane and twice from 67 ligroin (60-9000.) to yield white crystals which melted at 58-900. /CH3 <\/ \>-CH2 Q‘OH CN43 Analysis: Calculated for Cl5Hl6O : C, 84.88; H, 7.60 Found : C, 85.08; H, 7.59 Examination of the infrared spectrum of this compound showed an O—H band of medium width which is typical of phenols substituted by an alkyl group in one of the ortho positions. The two bands at 11-59u and 11.85pz are characteristic of 1,2,4,5-benzene sub- stitution. A sample of 2-benzyl—5,6—dimethylphenol was mixed with a sample of 4-benzyl—5,6—dimethylphenol. The mixture melting point of 55—600C. shows that these are different compounds. Experiment No.2. Benzyl alcohol, 108.2 g. (one mole), 2,5-xylenol, 122.2 g. (one mole), 500 m1. cyclo— hexane, and 500 ml. 67% sulfuric acid were refluxed for two hours. The reaction mixture was washed as usual, and solvent and unreacted 2,5—xy1enol were removed by vacuum distillation. Distillation of the residue yielded 12.1 g. of 68 2—benzyl-5,6-dimethylphenol and 19.5 g. of 4—benzy1- 2,5-dimethylphenol. This is a monobenzyl phenol yield of 15% and an ortho/para ratio of 0.62/1. In addition, a fraction of 4.1 g. (2%) was obtained which boiled at 154-14400. at 1 mm., about ten degrees below the fraction containing 2-benzyl-5,6-dimethy1phenol. Examination of the spectrum of this low-boiling fraction shows that it is probably impure benzyl—2,5- dimethylphenyl ether. C. I ,___\ \\/——-\\/' C H2 " O " /—\ There is a weak O-H stretching band at Q-ch which would be due to the presence of 2—benzyl—5,6-dimethle Phenol as an impurity. This band has about 20% of the absorbance that it does in pure 2—benzyl—5,6-dimethle Dhenol, therefore the ether fraction may contain about 20% of this phenol. Four other bands present in the spectrum of 2-benzyl-5,6—dimethylphenol are also present as weak bands in the ether fraction. These are bands present at 8.1, 8.5, 9.5, and 10.5/L- There is a band at 8.85yc in the ether fraction Which is not present in the spectrum of 2—benzyl-596’ dimethylphenol. This is a strong band which may be due to the CH2—O vibration. In addition, the aliphatic 69 C-H stretching band in the 5'”"5°é/¢ region is present as a doublet in the spectrum of the ether fraction. This doublet would be consistent with the fact that in benzyl 2,5—dimethy1phenyl ether there are two kinds of aliphatic C-H bonds: the type bound only to arvl, and the type bound to aryl and oxygen. (EH3 // \ _ \——> CH2 0 / CH3 2 ‘K\\ nd type (5.5/4) 1st tYDe (5'5‘L) In 2-benzy1-5,5-dimethylphenol there is only the first type of C—H bond, bound to aryl alone. ()4 CH2— —/\ CH3 /_—\ \——— orig), .,\ 1st type - bound to aryl only (5.4/4) Only the second type of aliphatic C—H bond would be Dresert in henzyl phenyl ether. 2nd type only (5.5/L) / ,/ o 70 If the spectrum of the above compound is examined there is a C—H band at 5.5/4, but not at 5.4;(. The crude 4-benzyl-2,5-dimethylphenol was recrys- tallized from hexane to yield white crystals, m.p. 59.5-60.00C. /CJ43 CH3 Comparison of the spectrum of this sample of 4-benzyl-2,5—dimethy1phenol with a spectrum of a sample Prebared by the p—toluenesulfonic acid method showed that they were the same compound. A phenylurethane was prepared by reaction of 4—benzyl-2,5-dimethylphenol with phenyl isocyanate. Recrystallization from hexane yielded white crystals, m.p. 159.0—159.5°c. Analysis: Calculated for 022H20N02 : N, 4.24 Found : N, 4.55 71 d. Benzylation of 2,6—xylenol Experiment No.1. To 450 ml. of cyclohexane was added 100 a. (0.818 mole) 2,6-xylenol and 19.0 g. (0.1 mole) p-toluenesulfonic acid. This mixture was heated to reflux temperature, and 75.7 g. (0.7 mole) benzyl alcohol was added over a period of five hours and 50 minutes. Refluxing was continued for an addi- tional hour. During the reaction period 15.7 ml. of water were removed. The reaction mixture was washed as usual and the solvent and unreacted xylenol removed by vacuum dis- tillation. The residue was distilled at 155—6100. at 5 mm. and crystallized to give 45.4 g. of a soft white solid. This corresponds to a 51% yield of monobenzyl— 2,6-dimethylphenol. The crude monobenzyl-2,6—dimethylphenol had a melting point of 41—400. An examination of its infra- red spectrum showed.tmunhs at 12.1, 12.55, and 12.89%L. Not all these bands should be present in a 2,4,6- substituted phenol such as the expected 4-benzyl-2,6— dimethylphenol. /CH3 // )c H2 QOH CH3 Apparently the monobenzyl fraction was a mixture of isomers. The other possible monobenzyl isomers would be 5—benzyl—2,6-dimethylphenol or benzyl 2,6— dimethylphenyl ether. QH / 2‘\ H3C7/ JCHB \ \\ \CHZ (PH/3 C) The crude monobenzyl fraction was recrystallized from hexane to qive white crYstals, m.p. 51—200. The mother liquor was set aside for further examination. The crystals melted at 51-200., and had an infrared spectrum which showed a much stronaer band at 12.89“, and much weaker bands at 12.l;4 and 12.55z4. This material was further recrystallized from ethanol-water t0 Yield a compound with a m.p. of 68.0—68.500. Elko— baisi and Hickinbottom (46) report a m.p. of 6600. for 4-benzyl-2,6-dimethylphenol. The infrared spectrum of the pure material showed a still stronqer band at 12.89;c whereas the bands at 75 12.1/u and 12.55/u. had disappeared. A phenylurethane was prepared by reaction of 4-benzyl-2,6-dimethylphenol with phenyl isocyanate. The urethane was recrystallized from hexane to give white crystals, m.p. 15400. Elkobaisi and Hickinbottom (46) report a melting point of 15500. for the phenyl urethane of 4-benzyl-2,6-dimethylphenol. The mother liquor from the first recrystallization was evaporated until most of the remaining xylenol had separated as a solid. The solid was isolated by fil- tration and the small amount of mother liquor evaporated to yield a yellow oil. The infrared spectrum of this oil showed stronqer 12.l;¢ and 12'5%#0 bands and a weaker 12.%;¢ band. The yellow oil was dissolved in hexane and cooled in the refrigerator. A white solid separated and was shown by its infrared spectrum and mixture melting point to be more 4-benzyl—2,6-dimethylphenol. Solvent was removed from the mother liquor to yield a yellow oil. On standing additional crystals formed in the yellow oil. The crystals were separated and shown to be more 4-benzyl—2,6—dimethylphenol. The remaining Yellow oil was dissolved in carbon disulfide and an infrared spectrum prepared. The infrared spectrum of the yellow oil showed that the bands at l?.l/¢ and lE-Sépb were larger and the band at 12.§Lp weaker. There was now a shoulder on the 12.§;¢ band at 12.3“,. 74 The yellow oil should contain 5~benzyl—2,6-xylenol, since the other isomer was shown to be 4—benzyl—2,6— xylenol. Furthermore the band at 12.55u_ (the strongest band in this region) is characteristic of 2,5,6—sub— stituted phenols. oH i45C74;\n’CLH3 V9 H, Analysis: Calculated for 015H16O : C, 84.88; H, 7.60 Found : C, 82.55; H, 7.55 The low carbon and hydroqen values show this isomer is not pure. As only a few drone of yellow oil were obtained, further purification was not attempted. Experiment No.2. To 450 ml. of cyclohexane was added 100 a. (0.818 mole) 2,6—xylenol, 108.1 g. (one mole) benzyl alcohol, and 19.0 g. (0.1 mole) p-tolueneCL sulfonic acid. This mixture was refluxed for eight hours, during which time 17.6 ml. of water were removed. The reaction mixture was washed as usual, and the solvent and unreacted xylenol were removed by vacuum distillation. The residue was distilled at 5 mm. to give 59.6 a. of monobenzyl-2,6-dimethylphenol, a yield of 54%. A considerable amount of residue remained from the 75 distillation of the monobenzyl-2,6-dimethylphenol. This residue was distilled at 2 mm. to dive 40.7 g. of a yellow licuid boiling at 215-800. which crystallized to aive a lirht yellow solid. The solid was recrystallized from benzene—liqroin and further recrystallized from hexane to yield white crystals, m.p. 117.5—118.OOC. Analysis: Calculated for 022H220 : C, 87.57; H, 7.54 Found : C, 87.51; H, 7.46 From the boilinj point and elemental analysis of the unknown substance it is apparent that two moles of benzyl alcohol have been condensed with one mole of 2,6-xylenol. The infrared spectrum of this compound was examined and the followinm facts noted: 1) There was a strong O-H stretching band at 2.79u,. (Therefore the compound is a dibenzyl dimethyli? phenol and not an ether. The narrow width of this band is characteristic of phenols with alkyl groups in both ortho positions. 2) The band at 5.50;; (aromatic C—H) was about the same depth as the band at 5.40*¢ (aliphatic C-H). In a monobenzyl dimethylphenol the 5.50pc band is not as deep as the 5.4on band. In a dibenzyl dimethylphenol there would be ten aliphatic hydrosens and eleven aro— matic hydroqens whereas with a monobenzyl dimethylphenol 76 there are eisht aliphatic hydrogens and seven aromatic hydrosens. 5) There was no band between 12 and 15“,, thus it is unlikely that there is either para or l,2,5,5-substi- tution present. 4) The monosubstitution band at 15°79AL was much larger than the monosubstitution band of a benzyl dimethylphenol. This indicates that two monosubstituted rings are present. 5) There was a band at 10°9épc which is not present in 4-benzyl-2,6-dimethylphenol. This band is character- istic of pentasubstituted benzenes. It was shown that 4-benzyl-2,6—dimethy1phenol and probably 5—benzvl-2,6-dimethylphenol were present in the monobenzyl fraction. In View of all these facts it seems likely that this dibenzyl dimethylphenol is 5,4- dibenzyl—2,6-dimethylphenol. QH L§ CJ42 CH2 Elkobaisi and Hickinbottom (46) report a hish boiling compound melting at 118°C. which they isolated from the reaction of benzyl alcohol and 2,6-xylenol in 77 petroleum ether (60-9000.) With an aluminum chloride Cfitalet. They claim the structure of their compound to be benzyl (4-benzyl—2,6-dimethylpheny1) ether. CZH3 I ”we CH2. CH3 However it seems possible that their so-called ether was 5,4—dibenzyl-2,6-dimethylphenol. A urethane of 5,4-dibenzyl—2,6-dimethylphenol was prepared by reaction with phenyl isocyanate. Recrys— tallization of this compound from hexane gave white crystals, m.p. 184—500. Analysis: Calculated for C29H28NO2 : N, 5.34 Found : N, 5.75 78 e. Benzylation cf 5,4—xy1enol To 500 m1. of cyclohexane were added 185.2 g. (1.5 moles) 5,4-xylenol and 19.0 a. (0.1 mole) p-toluene: sulfonic acid, and the mixture was heated to reflux temperature. To this was added 108.1 a. (one mole) benzyl alcohol over a period of three hours and 57 minutes. After addition was complete, refluxinr was continued for one hour. The reaction mixture was washed as usual, and the solvent and unreacted xylenol removed by vacuum distil- lation. The residue was distilled at 5 mm. to yield 119 a. of a monobenzyl 5,4—dimethylphenol (56% yield). (DH r%\\ /‘\\ CH3 ch, c H2 / (\l Examination of the infrared spectrum of the mono- benzyl 5,4-dimethylphenol shows an O-H stretching band Of medium width at 2.7“,, characteristic of phenols substituted in one ortho position. The shape of the 0~H stretching band at 2.2»0 cannot be used to distin- auish between the two possible ortho isomers; its since Shape would be about the same for each isomer, each isomer would have one ortho substituent. 79 However, an examination of the 10'1?A¢ reaion indicates that a mixture of two isomers is present. A band at 12.55fic is characteristic of 1,2,5,4—benzene substitution and indicates the presence of 2-benzy1- 5,4-dimethy1phenol. OH . CH3 CJ+3 The bands at 11.55»L and 11°8éflt are characteristic of 1,2,4,5-substitution and indicate the presence of 2-benzy1-4,5—dimethylphenol. (DH ‘CHZ H3C , CH3 The crude product was redistilled to yield 60.9 g. o of a colorless liouid boilinm at 175.5 C- at 2.5 mm. UDon standing, a solid separated from the liquid. The solid was recrystallized from hexane to give white CTYstals, m.p. 46.5-47.00C. Analysis: Calculated for C15H160 Found : C, 84.15; H, 7.69 : C, 84088; H, 7’60 Examination of the infrared spectrum shows an ortho 80 substituted phenol with bands at 11-5§;o and 11.85;;, characteristic of 1,2,4,5-benzene substitution. The band at 12.55»; (characteristic of l,2,5,4-substitution) is absent. Apparently this solid is 2-benzyl—4,5— dimethylphenol. Phenyl isocyanate was reacted with 2—benzyl—4,5— dimethylphenol to yield a phenylurethane. Recrystalli— zation from hexane yielded white crystals, m.p. 159.0- 159.500. Analysis: Calculated for 022H20N02 : N, 4.24 Found : N, 4.57 The spectrum of the original crude product, con- taininm a mixture of the two ortho isomers, was re—examined. The band at 11-85z4. which is present in E-benzyl-4,5-dimethylphenol, can be used to measure the amount of this isomer present in the crude product. Comparison of the absorption of the bands in the spectra Of the two isomers showed the crude product to contain about 70% of 2-benzyl—4,5—dimethylphenol. The other 50% of the mixture should be 2-benzyl-5,4‘dim9thylpben01- 81 f. Benzylation of 5,5-xylenol To 400 ml. cyclohexane were added 185.2 g. (1.5 moles) of 5,5-xy1enol and 5.8 a. (0.02 mole) p—toluene: sulfonic acid, and the mixture was heated to reflux temperature. A solution of 81.1 g. (0.75 mole) benzyl alcohol in 100 m1. cyclohexane was added over a period of five hours and 40 minutes. The mixture was refluxed for an additional hour and 44 minutes. A total of r 1 .d ml. of water were removed. \N The reaction mixture was washed as usual, the layers were separated, and 1,2,4-trichlorobenzene added to the organic layer. Cyclohexane was removed by dis- tillation and then the unreacted xylenol co-distilled udth the trichlorobenzene. The residue was distilled at 0.8 mm. to dive 84.4 a. of 2-benzyl—5,5-dimethylphenol and 14.4 m. of 4-benzyl-5,5—dimethvlphenol. This is a monobenzyl yield of 64% and an ortho/para ratio of 5.9/1. A sample of 2-benzy1—5,5-dimethy1phenol was recrys_ tallized once from benzene—liaroin (90—1200C.), once from cyclohexane-limroin, and once from hexane to give white crystals, m.p. 77.5-78-500- CM4 .CHz Hac \CHa 82 Analvsis: Calculated for ClSHlBO : C, 84.88; H, 7.60 Found : C, 84.85; H, 7.61 The infrared spectrum of this compound showed an O-H band of medium width at 2.7;“;. Such a band is characteristic of a phenol with an alkyl group in one of the ortho positions. There were two bands at 11.8%;c and 12.Qa;, characteristic of l,2,5,5-benzene substitution. A phenylurethane of 2—benzy1—5,S—dimethylphenol was prepared by reaction with phenyl isocyanate. The ure- thane was recrystallized from hexane to give white crystals, m.p. 127.5-128.OOC. Analysis: Calculated for CE2H20N02 : N, 4.24 Found : N, 4.25 A sample of 4-benzyl-5,5-dimethylphenol was recrys- tallized from cyclohexane to yield white crystals, m.p. 105.0-105.5Oc. CH3 / \-CH2 / \ OH _— -—___. CH3 Buu-Hoi, Sy, and Lejeune (61) record a melting point of 112°C. for this compound. The infrared spectrum of this compound shows a wide 0-3 band characteristic of a phenol which has no 85 ortho substituents. The bands at 11.65%; and 11.9§¢¢ are characteristic of compounds which have l,2,5,5— benzene substitution. since the melting point of 4—benzyl-5,5—dimethy13 phenol was somewhat lower than recorded by Buu—Hoi, 3V, and Lejeune (61), the compound was prepared by another method. To 150 ml. of cyclohexane were added 150 ml. of 67% sulfuric acid and 61.5 g. 5,5-xylenol (0.5 mole). The reaction mixture was heated to reflux and 54.1 g. (0.5 mole) benzyl alcohol were added over a period of three hours. Heatin: was continued for an additional 50 minutes. The organic layer was separated and washed twice with water, whereupon crystals formed. Recrystallization from cyclohexane yielded white crystals, m.p. 106°C. The infrared spectrum of this sample showed it to be the same compound as the 4-benzyl-5,S-dimethylphenol prepared with a p—toluenesulfonic acid catalyst. A mixture melting point of the two samples of 4—benzy1- 5,5-dimethylphenol showed no depression. However, when a Sample of 4-benzyl—5,S—dimethylphenol was mixed with a sample of 2-benzyl—5,S—dimethylphenol, the depressed melting point (65.0—72.000.) showed that these were two different compounds. 84 4. Alkylation of Phenol with Substituted Benzyl Alcohols a. Alkylation of phenol withppemethylbenzyl alcohol To 500 m1. cyclohexane were added 61.1 g. (0.5 mole) p-tolyl alcohol, 141.1 m. (1.5 moles) phenol, and 19.0 g. (0.1 mole) p-toluenesulfonic acid. The mixture was refluxed for one hour and 27 minutes. During this time 9.6 m1. of water were collected. The reaction mixture was washed as usual, and solvent and unreacted phenol removed by distillation. The residue was distilled at 2 mm. to yield 60.4 g. of 2-(p—methylbenzyl) phenol and 11.5 g. of 4-(p-methy1: benzyl) phenol. This is a monobenzyl yield of 68% and an ortho/para ratio of 5.5/1. The crude 2-(p—methy1benzy1) phenol was redistilled at 1 mm. to yield a colorless liquid boiling at 146.5~ 148.OOC. This material was again distilled at 2 mm. to yield a colorless liquid, b.p. 156.0—6.5OC., n55: 1.5885. OH CH3. CH2 / \, Analysis: Calculated for Cl4H14O : C, 84.77; H, 7.07 Found : C, 84.78; H, 7.10 85 An infrared spectrum of this compound showed an O—H band of medium width at 2-7épto Such a band is characteristic of phenols substituted by an alkyl group in one of the ortho positions. The very strong band at 15.5%; is also characteristic of 1,2-benzene substitution. Phenyl isocyanate was reacted with 2-(p-methy1benzyl) phenol to form a phenylurethane. Recrystallization from hexane yielded white crystals, m.p. 101.5—2.OOC. Analysis: Calculated for C21H19N02 : N, 4.41 Found : N, 4.58 An infrared spectrum of the phenylurethane showed an N-H stretchinr band at 2.2;4 and a C=0 stretchins band at 5.75;. The stronm band at 15.5;9 is character- istic of 1,2—benzene substitution. Distilled 4-(p—methy1benzyl) phenol had a boilinm 0 point of 171-8000. at 2 mm. and n§5 = 1.5872. C H3 cHz— OH Analysis: Calculated for Cl4H14O : C, 84.77; H, 7.07 Found : C, 85.20; H, 7.14 An infrared spectrum of this compound shows an O-H band which has considerable hydrogen bonded character. Such a band would be present in a phenol such as 86 4—(p-methylbenzy1) phenol, which does not contain a group in the ortho position. There are bands at 12.4;‘ and 12.65u,, in the region where absorption due to 1,4- benzene substitution occurs. Phenyl isocyanate was reacted with 4-(p-methy1benzyl) phenol to form a phenylurethane. Recrystallization from hexane yielded white crystals, m.p. 98.0-8.5OC. Analysis: Calculated for 021H19N02 : N, 4.41 Found : N, 4.50 Examination of the infrared spectrum of the phenyl: urethane shows the expected N—H band at 2.95“ and the 0:0 band at 5.75}; . A small sample of the phenylurethane of 2-(p—methy13 benzyl) phenol was mixed with a small sample of the phenylurethane of 4-(p-methylbenzyl) phenol. The melting point of the mixture was 89.0—95.000. Examination of the infrared spectra of the two phenylurethanes also shows they are not the same com— pounds. The meat noticeable differences are a strong band at 9.3“,, which is present in the para isomer but not in the ortho, and a band at 9.€/L mhich.is very much stronger in the ortho than in the para spectrum. The carbon analysis for 4—(p—methy1benzy1) phenol, prepared by the p-toluenesulfonic acid method, was lower than the calculated value. This compound was therefore prepared by an alternate method. 87 To 50.5 g. (0.54 mole) phenol was added 65.6 g. (0.54 mole) p—tolyl alcohol, 150 ml. cyclohexane, and 150 ml. 67% sulfuric acid. The resultinq two phase system was refluxed for one hour and 50 minutes. The reaction mixture was cooled, the oraanic layer separated, and washed as usual with 5% sodium bicarbonate and water. solvent and unreacted phenol were removed by vacuum distillation. The residue was distilled at 1 mm. to yield 17.7 g. of 2-(p—methylbenzyl) phenol and 25.6 g. of 4-(p—methy1 benzyl) phenol. This is a monobenzyl yield of 59% and an ortho/para ratio of 0.75/l. The spectra of the two isomers were the same as those of the two isomers prepared by the p-toluenesulfonic acid method. The fraction containinm 4—(p—methy1benzyl) phenol had the followins analysis: Analysis: Calculated for Cl4H14O : C, 84.77; H, 7.07 Found : C, 84.21; H, 6.91 "13 ’._J t h. Alkylation of phenol with p-isopropylbgpzyl alcohol. y Experiment No.1. To 500 ml. cyclohexane were added 75.1 m. (0.5 mole) cf p—isopropylbenzyl alcohol, 141.1 g. (1.5 moles) of phenol, and 19.0 g. (0.1 mole) of p-tolu- enesulfonic acid. The mixture was heated to boiling and refluxed for one hour and 11 minutes, during which time 10.9 ml. of water were collected. The cooled reaction mixture was washed as usual, and the solvent and unreacted phenol removed by distil— lation. The crude product was fractionated to yield 66.5 s. of 2—(p—isopropylbenzyl) phenol and 14.0 g. of 4-(p-isopropylbenzyl) phenol. This is a monobenzyl yield of 71% and an ortho/para ratio of 4.8/1. The crude 2—(p-isopropylbenzy1) phenol was redis- tilled at 0.5 mm. to yield a colorless liouid boiling at 145.5—7.OOC. This material was aegain distilled at 0.9 mm. to yield‘a colorless liquid boilinf at 155.5- 156.500., n§50= 1.5720. Wheatley, Cheney, and Binkley o (63) aive an his = 1.5722 for this compound. CH3 OH I .. // \ ,. H0 \ >CHZ CH3 The infrared spectrum of this compound showed an C-H band of medium width at 2.7;u,. thfi18.band is 89 characteristic of phenols which have an alkyl sroup in one of the ortho positions. The very strone hand at 15.5%; is characteristic of 1,2-substituted benzenes. The band at 12.25;; should be due to the para substitu- tion on the benzyl ring. The 2-(p-isopropylbenzyl) phenol was reacted with phenyl isocyanate to form a phenylurethane. Recrystal- lization of the urethane from hexane yielded white crystals, m.p. 101.5-2.5OC. Calculated for 025112351102 : N, 4.06 Found : N, 4.15 The infrared spectrum of the phenylurethane shows an N-H band at 2.0;; and a C=0 band at 5.7;;. There is also a stronq band at 15.5pt characteristic of 1,2- benzene substitution. Crude 4—(p—isopropylhenzyl) phenol as distilled from the reaction mixture melted at 57—6100. It was recrystallized from methanol—water to yield white crystals, m.p. 71-200. 9H3 H9- (MO-OH (Chg Analysis: Calculated for C16H18O : C, 84.95; H, 8.03 Found : C, 84.85; H, 8.12 The infrared spectrun of 4-(p-isopropylbenzyl) phenol had an C—H band at 2°ZZL uhich.showed considerable hydrozen bonded character. Such a band is characteristic of phenols which have no ortho substituents. There was also a very strong band at 12.2%; with a shoulder at 12.55u,. Such a band would be present in phenols which have two types of para substitution present. Experiment No.2. To 45.7 a. phenol (0.49 mole) were added 75.0 g. p—isopropylbenzyl alcohol (0.49 mole), 150 ml. cyclohexane, and 150 ml. 67% sulfuric acid. This mixture was refluxed for one hour and 15 minutes. The reaction mixture was allowed to cool, an organic layer separated and was washed as usual. Solvent and unreacted phenol were removed by vacuum distillation. The residue was distilled at 1 mm. to yield 28.1 g. of 2-(p-isopropy1benzyl) phenol and 21.0 g. of 4—(p-isoC propylbenzyl) phenol. This is a monobenzyl yield of 44% and an ortho/para ratio of 1.3/1. The 4—(p—isopropy1benzy1) phenol fraction froze when seeded with a crystal from the previous experiment to give a soft white solid. Recrystallization from hexane yielded white crystals, m.p. 73.5-74.0CC. Comparison of the spectrum of this material with a 91 spectrum of 4—(p-isopropy1henzy1)phenol prepared by the p-toluenesulfonic acid method showed these two materials to be the same compound. The 4-(p-i80propylhenzy1) phenol prepared by the sulfuric acid method was reacted with phenyl isocyanate to form a phenylurethane. Recrystallization of the urethane from hexane yielded white crystals, m.p. 126.5—127.00C. Analysis: Calculated for 025H25N02 : l, 4.06 Found : N, 4.11 l small sample of this phenylurethane was mixed with a small sample of the phenylurethane of 2-(p—isoC propylbenzyl) phenol. The mixture melting point of A O . 92.0—95.0 C. showed the two compounds were different. c. o-Chlorobenzyl alcohol and phenol. To 500 ml. of cyclohexane were added 71.3 g. (0.5 mole) o-chlorobenzyl alcohol, 141.1 g. (1.5 moles) phenol, and 19.0 g. (0.1 mole) p-toluenesulfonic acid. The mixture was refluxed for four hours and 27 minutes, during which time 10.9 ml. of water were collected. The reaction mixture was washed as usual, and the cyclohexane and unreacted phenol removed by distillation. Fractionation of the residue yielded 55.2 g. of 2-(o-chlorobenzy1) phenol and 21.0 g. of 4—(o-chlor02; benzyl) phenol. This is a monobenzyl phenol yield of 92 68% and an ortho/para ratio of 2.5/1. The 2—(o—chlorobenzy1) phenol was redistilled to yield a colorless liquid, b.p. 157—8OC. at 1 mm., CH2 CH “50 n‘ = 1.6010. D Cfld Huston and co—workers (8) give a boiling point of 146-5100. at 5 mm. for this compound. Examination of the infrared spectrum of this com- pound showed an O—H band of medium width at 2.7/pg. This type of 0-H band is characteristic of phenols which have an alkyl substituent in one of the ortho positions. Therevemaalso a very strong band at 15.5;L which indi- cates there are two 1,2—benzene substitutions present in the molecule. Reaction of 2—(o~chlorobenzy1) phenol with phenyl isocyanate yielded a phenylurethane. Recrystallization of the urethane from hexane yielded white crystals, m.p. 127.5-s.ooc. Analysis: Calculated for C20H1601N02 : N, 4.51 Found : N, 4.27 The crude 4—(o—chlorobenzy1) phenol was recrystal- lized several times from hexane to yield white crystals, m.p. 71.0—71.500. 95 CJ Huston and co-workers (8) report a melting point of 68-9OC. for this compound. The infrared spectrum of this compound showed a wide O-H band at 2.7;” characteristic of phenols which contain no aroups ortho to the hydroxyl group. In addition to the ortho substitution band at 15.5zo from the ortho substituted C1 in the benzyl group, there is also a band at 12.4fi¢wmich indicates the benzyl group is para to the hydroxyl group on the phenolic ring. Reaction of 4-(o—chlorobenzyl) phenol with phenyl isocyanate yielded a phenylurethane which on recrystal- lization from hexane gave white crystals, m.p. 88-900. Analysis: ClNO Calculated for C2 N, 4.51 2 Found : N, 4.51 1 0116 A low boiling fraction had distilled at 150-15700. at 1 mm. Examination of this fraction showed that, in addition to about 60% 2-(o—chlorobenzyl) phenol, there was probably present about 40% of o—chlorobenzyl phenyl ether. 94 QCHz-O CI This deduction is based on the following spectral evidence: 1) The C-H stretching band has about 60% of the absorbance present in 2—(o-chlorobenzyl) phenol. Thus the impurity does not contain an O-H group. 2) There are new bands nresent at 7.3“ , 8.l/L, and 11.5éfic. All these bands are also nresent in the snectrun of benzyl phenyl ether. The 8.l/L band in particular is characteristic of all aryl ethers being due to a C-O stretching vibration. If this impurity is o-chlorobenzyl phenyl ether, then it was forned in a yield of 5%. 95 d. p-Chlorobenzyl alcohol and phenol To 500 m1. cyclohexane were added 99.8 g. (0.7 mole) p—chlorobenzyl alcohol, 211.4 a. (2.1 moles) phenol, and 9.5 a. (0.05 mole) p-toluenesulfonic acid. The mixture was refluxed for five hours and 25 minutes, during which time 14.0 ml. of water were collected. The reaction mixture was washed as usual, and solvent and unreacted phenol removed by distillation. The residue was distilled at 0.5 mm. to yield 85.7 g. of 2-(p-chlorobenzy1) phenol and 27.8 g. of 4—(p-chlor03 benzyl) phenol. This is a monobenzyl yield of 73% and and ortho/para ratio of 3.0/1. The 2-(p-chlorobenzyl) phenol was recrystallized from hexane to yield white crystals, m.p. 61—200. Huston and co-workers (R) give a melting point of 60-100. for this compound. CH4 The infrared spectrum of 2-(p-chlorobenzyl) phenol has a fairly narrow O-H band at 2.3“,. Such a band is characteristic of phenols with one ortho substituent. The band at 9oépb is also characteristic of ortho sub— stituted phenols. The very strong band at 15-3AL is due to the 1,2-suhstitution of the phenolic ring, whereas the band at 12.6%4 :ha«iic to the para substitution of the other benzene rinT. The 4-(p—chlorobenzyl) phenol was recrystallized from hexane to yield white crystals, n.p. 88.0—88.500. Huston and co—workers (9) give a melting point of 87.0-87.500. for this compound. Examination of the infrared spectrum of 4—(p—chloro benzyl) phenol showed the expected differences from the ortho isomer. There is a strong hydrogen bonded shoulder at 2.8;“; on the C-H stretching band. This shoulder is characteristic of phenols which do not contain an ortho substituent. The 8.2/4 band, present in the ortho isomer, is absent in this spectrum as is the band at 15.;u,. fiince both of these bands are due to 1,2—sub- stitution, we would expect them to be absent. In addition the band at 12.25-12.5§,p is very much stronger ') K) ince in this coxpound there are two types of 1,4- substitution present. (Ll-.042 ...o.. 97 e. p-Brorobenzyl alcohol and phenol To 503 ml. of cyclohexane were added 93.5 g. (0.5 mole) p-bromcbepzyl alcohol, 141.1 g. (1.5 moles) phenol, and 19.0 r. (0.1 mole) p-toluenesulfonic acid. The reaction mixture was refluxed for two hours and 11 minutes, durinq which time 10.5 ml. of water were collected. The reaction mixture was washed as usual and the solvent and unreacted phenol removed by distillation. Distillation of the residue yielded 65.4 a. of 2—(p—bromobenzyl) phenol and 26.2 g. of 4—(p—bromobenzyl) phenol. This corresponds to a 68% yield of monobenzyl product and an ortho/para ratio of 2.4/1. Crude 2-(p-bropobenzy1) phenol was distilled at 0.5 mm. to yield a white solid toilin: at 165—500. This material was recrystallized from hexane to give white crystals, m.p. 72-500. Huston and co-workers (60) give a meltinf point of 72-500. for this compound. CH4 Br.CHZ‘ The infrared spectrum of 2-(p-bromobenzyl) phenol has a fairly narrow C—H stretchinq band at 2.3u,. Such a band is characteristic of phenols which contain an ortho substituent. There is also a strona band at 8.3;; 98 characteristic of ortho substituted phenols. The band at 12.3“, is ctaracteristic of 1,4—benzene substitution and‘would be due to the para substitution on the other benzene rinfi. Also the band at 13-9pc would be due to the 1,2-substitution of the phenolic ring. The crude 4-(p-bronobenzyl) phenol was recrystal- lized from hexane to dive white crystals, m.p. 84-500. Wu, Guile, and Huston (6) give a melting point of 85.0—85.500. for this compound. Examination of the infrared spectrum of this com— pound showed the expected bands. There is a strong hydrogen'bonded shoulder at 2.8%;‘ on the O—H stretching band, since there are no substituents ortho to the hYdroxyl group. The band at 8.?»g, which was present in the ortho isomer, is absent as is the band at 15.9%9. There is a new band at 11.95/4, which may be due to the tYpe of para substitution present in the phenolic ring. The band at 12.4%9 is present, as in the ortho isomer, and would be due to the 1,4—substitution pattern of the other benzene rim?- 99 f. 2,4-Dichlorobenzzl alcohol an d nhenol There were added 9R.5 a. (0.5 mole) 2,4-dichloro; benzyl alcohol, 141.1 a. (1.5 moles) nhenol and 19.0 2. (0.1 mole) p-toluenewulfonic acid to 500 ml. of cycle hexane. This mixture was heated to reflux temperature and 10.9 ml. of water were collected over a period of ten hours. The reaction mixture was allowed to cool, washed as usual, and the orranic layer distilled to remove solvent and unreacted orenol. The residue from this distillation was fractionated at 5 mm. to yield 66.2 g. of 2—(£,4-dichlorobenzyl) phenol and 26.3 g. of 4-(2,4- dichlorob nzyl) phenol. This is a monobenzyl yield of 75% and an ortho/para ratio of 2.5/1. The crude 2—(2,4—diohlorobenzyl) phenol was re- distilled to yield a colorless liquid, b-P- 158°5‘ 160-50 C., which solidified on standing. The solid was recrystallized from hexane to yield white crystals, mop. 66—6700. Analvsis: Calculated for 015 M1001 O :(3,62.07;H,5-9B;Cl,28.02 Found :(3,61.76;H,4.15;Cl,28.24 lCO Examination of the infrared spectrum of this compound showed a fairly narrow O-H stretching band at 2.2pc characteristic of phenols containinj a bulky group in one ortho position. The stronw band at 15.9“, is due to the presence of l,2-benzene substitution on the phenolic rinj. There was also a strong band at 5Li%¢ characteristic of ortho-substituted phenols. It is most probably an O—H deformation band whose frequency is sensitive to the amount of hydrogen bonding. A phenylurethane of 2-(2,4-dichlorobenzyl) phenol was prepared by reaction with phenyl isocyanate. Re- crystallization from hexane yielded white crystals, m.p. 125.0-123.5°c. Analysis: Calculated for CEOHlSCl2NO2 : N, 5.77 Found : N, 4.08 Crude 4-(2,4—dichlorobenzy1) phenol was recrystal— lized from hexane to yield white crystals, m.p. 72.0- 72.500. Cit-.042 .014 C! Analvsis: Calculated for Cl5HlOC120 :(3,62-07;H,5.98;Cl,28.02 Found :(3,61.94;H,5.95;Cl,28.50 lOl Examination of the infrared spectrum showed an O-H stretching band at 2.7a with a large hydrogen bonded shoulder. Such a band is characteristic of phenols which do not contain an ortho substituent. The strongest band in the spectrum is a very strong and broad band at 12'l_12'%“" Since both 1,2,4— and 1,4-substituted benzenes absorb in this region, and since both types of substitution are present in the molecule, the size of the band is reasonable. The band at 8.5“,, which was present in the spectrum of the ortho isomer, is absent in this spectrum. ' A small sample of 2-(2,4—dichlorobenzyl) phenol was mixed with a small sample of 4-(2,4-dichlorobenzyl) phenol. The mixture meltina point of 48.5-54.500. showed these compounds were different. A phenylurethane of 4-(2,4—dichlorobenzyl) phenol was prepared by reaction with phenyl isocyanate. Re— crystallization from hexane yielded white crystals, m.p. 142.50C. Analysis: Calculated for CEOHlSClZNO2 : N, 5.77 Found : N, 5.87 102 5. Relative Rate of Phenol Benzylation by Substituted Benzyl Alcohols The relative rate of phenol benzylation by substi— tuted benzyl alcohols was determined in the followina manner: The substituted benzyl alcohol (0.5 mole), phenol (1.5 moles), p—toluenesulfonic acid (0.1 mole), and 500 ml. cyclohexane were heated rapidly to boiling and the reaction mixture was refluxed until the calculated amount of water had been collected. The reaction rate was followed by recordina the amount of distillate collected in the Dean-Stark trap and the time of the readinm. Amount of water was plotted on graph paper as a function of time, and the time at which half of the calculated water had been collected was determined. As 1.9 ml. of water was distilled from the catalyst at the start of the reaction, the reaction half time was the time at which 1.9 ml. + 4.5 ml. (0.25 mole) or 6.4 ml. of water had been collected. An experiment of this type was carried out for benzyl alcohol and five substituted benzyl alcohols. The data (ml. of water collected and time) for these six experiments are to be found in the first section of the appendix. The six graphs from the plotting of these data are also in the appendix. The reaction half times were calculated from the ?raphs and recorded in the following table. 105 Table V. Rate of Phenol Alkylation with Benzyl Alcohols “— Exp. Benzyl alcohols Half time of reaction in minutes l p-isopropyl 26 2 p-methyl 52 5 unsubstituted 56 4 p-chloro 47 5 o-chloro 150 6 2,4-dichloro 254 The structure of products formed in these reactions was established in the precedins Experimental section. For purposes of comparison, it is assumed that the products formed are similar in all cases. However this is not strictly true, as has been shown by the variation of the ortho/para ratios. The quality of the reagents used in these six experiments was the same as the quality of the reagents used in the preceding Experimental section. No attempt was made to prepare reagents of special purity, as the purpose of these experiments was not to carry out a detailed kinetic study of this alkylation, but rather to see if the half times of reaction fell in an order that was consistent with the sigma or substituent constant of the substituent on the benzyl alcohol. 104 6. Scope of the Acid-Catalyzed Ortho Alkylation In order to determine the scone of the acid-cata- lyzed ortho alkylaticn of phenols, a number of experi— ments were carried out in which a series of available substituted aromatic and aliphatic alcohols were heated to reflux temperature with benzyl alcohol. In all cases a cyclohexane solvent was used. The formation of water and its distillation during reflux indicated some type of alkylation was taking place. In some cases, where nearly the calculated amount of water for the alkylation reaction distilled, the reaction mixture was washed with sodium bicarbonate solution and distilled. a. Phenol and various alkylating agents Cinnamyl alcohol and4phenol Cinnamyl alcohol (0.75 mole), one mole of phenol, and 0.05 mole p-toluenesulfonic acid were reacted at reflux temperature in cyclohexane, and the calculated amount of water removed by distillation. The product was a dark viscous polymer. CH:CH‘CH20H OH /, 80 C. _ polymeric + ~%- \\ p—toluene products sulfonic acid 105 E-Phenzlethyl alcohol and phenol Refluxinq 0.72 mole 2-phenylethyl alcohol, 1.5 moles phenol, and 0.2 mole p—toluenesulfonic acid in cyclo- hexane aave 0.07 mole water after four hours. Very little reaction p—toluene sulfonic acid Propargyl alcohol and phenol No reaction was observed on refluxinfi one mole phenol, one mole nropargyl alcohol, and 0.05 mole p-toluenesulfonic acid in cyclohexane for 6.9 hours. 0 (DH so 0. HCEC— CHZCH + // :~ No reaction p-toluene sulfonic acid \\ Egnzhydrol and phenol Water was eliminated on refluxing 0.4 mole benz— hYdrol, 0.8 mole phenol and 0.02 mole p—toluenesulfonic acid in cyclohexane. The reaction product was washed, and solvent and unreacted materials removed by distilla— tion. The residue was distilled to give 69.1 g. of an unidentified yellow solid, b.p. 174—194°c. at 1 mm. This boilin: point would be reasonable for a con— densation product of one mole benzhydrol with one mole of phenol. and 106 a 0 80 C' Unidentified a>= productc5 b.p. p-toluene 174-194 C. at sulfonic 1 mm. acid tert-Amyl alcohol and phenol Refluxin: 0.75 mole tert—amyl alcohol, 1.5 moles phenol, and 0.1 mole p—toluenesulfonic acid in cyclo- hexane for four hours and 35 minutes yielded 14.3 ml. of water. Washing the reaction mixture as usual, removin? solvent and unreacted phenol by distillation, and distillation of the residue yielded 37.5 g. of p-tert-amylphenol, m.p. 92-300.; literature (76) m.p. 92-30c. CN4 CH5 ' —C-0H ~+ CH5 C2H5 chlohexanol and phenol CN4 80°C. g / __i \ . Except for a slight shoulder on the ll.9p¢ band of the authentic sample, the Spectra of the two samples were identical. Fraction 2 is therefore assumed to be 2-Cg(-phenylethyl) phenol. Styrene and phenol Phenol (1.4 moles), 0.1 mole p-toluenesulfonic acid, and 500 ml. cyclohexane were heated to reflux temp- erature. Styrene (0.7 mole) was added over a period of one hour and refluxing continued for an additional hour. Cit-4:042 OH O I %%\\ 80 C' ‘_ Alkylated + I " products p—toluene: \\ sulfonic acid The reaction mixture was washed as usual, and the solvent and unreacted phenol removed by distillation. The residue was distilled at 5 mm. to yield the following fractions: 109 Fraction Boiling point Weight Description 1 82—15100. 0.9 g. Colorless liquid 2 151.5—157.0°C. 44.5 g. Colorless liquid 5 157.0-160.5°C. 9.4 s. Colorless liquid 4 161.5-168.OOC. 20.0 g. Colorless liquid The amount of product (from 0.7 mole alkylatina agent) distillin: in the 151.5-168.OOC. range was 75.7 g. compared to 74.7 g. of material in the 155-169OC. range obtained in the previous experiment from 0.75 mole of alkylating agent. b. Benzyl alcohol and various substrates Benzyl alcohol and o—aminophenol N0 reaction was observed on refluxing 0.75 mole benzyl alcohol, one mole o—aminophenol, and 0.05 mole D-toluenesulfonic acid in cyclohexane for two hours. (:H2CN4 (DH -NH42 —>= N0 reaction +‘ p—toluene sulfonic acid §§nzyl alcohol and p-chlorophenol To 400 ml. cyclohexane was added 1.2 moles p-chloroC Dhenol and 0.05 mole p-toluenesulfonic acid. This mixture was heated to reflux temperature and 0.5 mole benzyl alcohol added over a period of three hours and 110 45 minutes. Refluxinm was continued for an additional one hour and 45 minutes. Durinr this time 15.7 ml. of water were removed. The reaction mixture was washed as usual, and solvent and unreacted p-chloronhenol removed by distil— lation. The residue was distilled at 1 mm. to yield 65.4 g. of a colorless liquid, b.p. 154—16000. The colorless liouid solidified on standing, to give white crystals of 2-benzyl-4-ch10rophenol, m.p. 48.5-9.OOC. Klarmann, Gates, and Shtenov (62) report a m.p. of 48.500. for this compound. CHZOH OH O OH : '+ p-toluene: ——— \\' sulfonic (Ll aCld CA Examination of the infrared spectrum shows the expected features. The O-H stretching band at 2.7;u1 is of fairly narrow width since there is one ortho substituent. There is a 1,2,4-substitution doublet at 12.1% and 12.4911, very similar to that found in 2-benzy1-4—methy1phenol. In addition, there is the expected band at 8.5/¢ characteristic of ortho substi- tuted phenols. Esnzzl alcohol and o-chlorophenol Water was eliminated (14.1 ml.) on refluxing 0.7 mole benzyl alcohol, 1.2 moles o-chlorophenol and 0.1 111 mole p-toluenesulfonic acid in cyclohexane. Washing and distillation yielded 70.7 q. of an unia entified liquid, b.p. 150—16500. at 3 mm. CH20F+ 8000. C1 Monobenzyl "“”'“"*”’ condensation p-toluene products sulfonic acid Benzyl alcohol and p—ethylphenol Experiment No.1. To 450 m1. of cyclohexane were added one mole p-ethylphenol and 0.07 mole p-toluene:; sulfonic acid. The mixture was heated to reflux temp- erature and 0.6 mole benzyl alcohol was added over a period of one hour and 49 minutes. Heating was continued for an additional 50 minutes. During this time 12.6 ml. of water were removed. The reaction mixture was washed as usual, and solvent and unreacted p-ethylphenol removed by distilla- tion. Distillation of the residue at 5 mm. yielded 55.5 a. of 2-benzy1—4-ethy1phenol, boiling at 164.5- 169.500. / Q» 4_ p-toluene ’ sulfonic \\ acid . Csz 112 Redistillation of the crude product gave a color— q 0 less liquid, b.p. 164.0-164.5OC. at 3 mm., n55 = 1.5779. Analysis: Calculated for C 0 : C, 84.88; H, 7.60 15Hl6 Found : C, 85.04; H, 7.54 Hojahn (77) reports a boiling point of 197°C. at 17 mm. for this compound. Examination of the infrared spectrum of this compound showed the expected bands. Except for position and intensity of the C-H bands, the spectrum is almost the same as that of 2—benzy1-4—methy1phenol. The 1,2,4- substitution doublet at 12.1*L and lE-éfL is at a little lower wavelength than the doublet of 2-benzy1-4-methy13 phenol. Phenyl isocyanate was reacted with 2—benzy1—4— ethylphenol to dive a phenylurethane. Recrystallization o from hexane yielded white crystals, m.p. 117.5-118.5 C. Apalysis: Calculated for C22H21N02 : N, 4.24 Found : N, 4.25 Experiment No.2. To one mole p-ethylphenol were added 2 g. of Dowex 50U—X2 (200-400 mesh) and 25 ml. of cyclohexane, and the mixture heated to 15500. Benzyl alcohol (0.4 mole) was added dropwise and heating con- tinued until 7.9 ml. of water had been removed. The reaction mixture was filtered while hot and the filtrate fractioned to yield 49.6 g. of 2—benzy1-4-ethy1phenol 115 (58% yield), b.p. 171—400. at 2 mm. The infrared spectrum of this material was compared with the spectrum of 2-benzy1—4-ethylphenol prepared in the first experiment and found to be identical. Benzyl alcohol and o—ethylphenol To 550 m1. cyclohexane were added 0.59 mole o-ethle. phenol and 0.07 mole p-toluenesulfonic acid. The mixture was heated to reflux temperature and 0.5 mole benzyl alcohol added over a period of two hours and 59 minutes. Refluxinz was continued for an additional 40 minutes and a total of 10 m1. of water were removed. Washing and distillation yielded 45.8 g. of material boiliné at 157-17500. at 5 mm. CHZOH 0H 8000. . . . ,/ J_ Unidentified C2 H5 "’ monobenzylation ‘ +- p-toluene products \\ sulfonic acid Egnzyl alcohol and anisole Water was eliminated on refluxing 0.7 mole benzyl alcohol, 1.4 moles anisole, and 0.1 mole p-toluenesul- fonic acid in cyclohexane. Distillation of the crude reaction product yielded 65.6 g. of an unidentified liquid boiliur at 137-14600. at 5 mm. 114 ' 30000 // i .1 ___ Unidentified \\ ' n—toluene: monogigggigglon sulfonic ~ - « acid Benzyl alcohol and catechol Hefluxinm one mole catechol, one mole benzyl alcohol and 0.05 mole p-toluenesulfonic acid in cyclohexane gave a rapid evolution of water. Purification and distilla— tion at 5 mm. yielded the following fractions: Fraction Boiling point Weieht Description 1 142-7800. 6.7 g. Soft white solid 2 180.0-6.50C. 27.2 g. Colorless liquid 3 187—9000. 7.8 g. Light yellow liquid 4 191-400. 15.0 g. Colorless liquid Fraction 2 was recrystallized four times from o cyclohexane to wive white crystals, m.p. 94-5-5-5 C. Analysis: Calculated for C15H1202 : c, 77.97; H, 6.04 Found : C, 78.45; H, 6.01 o C 2 ##- Unidentified p—toluene: monobenzylation sulfonic products aCld (C 15 H1202) 115 Benzyl alcohol and thiophenol As evidenced by the absence of water formation, no reaction was observed after 85 minutes refluxinq of 0.5 mole benzyl alcohol, 0.5 mole thiophenol, and 0.05 mole p-toluenesulfonic acid in cyclohexane. /CH20H 9000. Q == No reaction p—toluene: sulfonic acid Benzyl alcohol and N-methyl aniline No water was obtained after 47 minutes refluxin of 0.5 mole benzyl alcohol, 0.5 mole N-methylaniline and 0.1 p-toluenesulfonic acid in cyclohexane. CHon H N CH5 I +: 4;. No reaction IV. DISCUSSION A. Infrared Analysis of Phenols The infrared spectra of benzylated phenols has been shown to be a valuable tool for the characterization of the products obtained in this work. The three most valuable regions were the 0-H stretching bands at 2-7‘5-Q/L, the 0-H deformation bands at 8.2—8.6/L, and the aromatic substitution region from 10-13u,. l. Oxygen-Hydrogen Stretching Bands The O-H stretching band at 2.7-5.0}; consists of two parts in phenol itself, a non—hydrogen—bonded part at 2.7/4 and a hydrogen—bonded shoulder at 2°8‘5°Q%L° If a phenol is substituted (with an alkyl group) in the para position, there should be no interference with the hydrogen bonding of the O-H group. A para or meta sub— stituted phenol would have a large hydrogen-bonded shoulder. All phenolic products obtained in this research which do not have an ortho substituent have a spectrum with this type of O—H bond. The introduction of an ortho alkyl group interferes with hydrogen bonding of the O-H group for steric reasons. 116 117 All benzylated phenols which contain one ortho substit- uent have an C-H band of medium width and little or no shoulder. The introduction of two ortho groups greatly reduces the amount of hydrogen bonding of the 0-H group. It was found that all benzylated phenols with two ortho substituents have narrow O—H stretching bands. For benzylated phenols, the C-H stretching hand offers positive evidence as to whether ortho or para alkylation las occurred for a particular sample. If two ortlo isomers are possible, this region is of no value in deciding which ortho isomer is present. 2. Aromatic Substitution Region at lO-lélc The strong bands in this region are caused by out- of-plane deformation vibrations of the hydrogen atoms remaining on the ring. The nature of the substituents does not have too much influence on the position of these bands. These bands are well suited for quantitative work because of their hish intensity. Many applications have been described for the estimation of relative proportions of isomers. One such application is described by Chumaevski (78), who examined structures of o-, m—, and p—cresol as % solutions in CS2. He measured the optical density of bands at 12.55;L and 12.85;L which correspond to the absorption bands of p- and m-cresol. The o-cresol 118 is considered as a solvcpt for t?e other two isomers, and i; obtairei by difference. This wcthod can be extended to include xylenols. Whiffen and Thompson (79) have used bands in this region to determine mixtures of cresols as “e11 as mix- tures of xylenols. The analysis of phenols, cresols, xvlenols, and ethyl phenols has been described by Friedel, Peirce, and McGovern (80). These bands were used a number of times to make estimations of the isoners present in products of this research. In sore cases, such as the reaction of p-chlorobenzyl and p-bromobenzyl alcohols with phenol, both ortho and para isomers were isolated. The spectra of the pure isomers were used for the quantitative esti- mation of these isomers in each distillation fraction. Ia some cases mixtures of isomers were obtained from which one isomer could be separated as a solid. However, the other isomer could not be obtained in a pure state. Such was the case with the following mixtures: 1) Erbenzyl-5—methylphenol and 2—benzy1-5- methylphenol 2) 2-benzyl-5,4—dimethylphenol and 2-benzy1-4,5_ dimethylphenol 5) 5—benzy1-2,6-dimethylphepol and 4—benzyl-2,6- dimethylphenol The structure of the isolated solid isomer was first 119 determined. Comparison of tie soectrum of the isolated isomer with t‘at of the mixture showed the bands in the mixture belonqinq to the isomer which could not be isolated. These bands could be used to check the structure of this isomer. An estimation of the relative amounts of the two isomers could be made by measuring the strenath of a.band in the lO—léicregion of the mixture belonging to the isolated isomer. The bands in the lO—léuv resion are very useful for the qualitative identification of aromatic compounds. In monosubstituted benzenes the out-of-plane CH bending absorotion gives rise to a very strong band in the 14.0-l4.%u rezion. In 1,2-substituted benzenes the four adjacent hydrosens give rise to a band in the 13.2-15.4%( region. / Three adjacent free hvdrogen atoms yield a band with stronm absorption in the 12.5-15.Q/( region. Further reductions in the number of hydrOQens keep shiftinr the band to hiaher frequencies (lower/x values) and reduce its strength. Fare (81) has determined the spectra of El alkyl— ated phenols in 082 and CCl4 on a linear/%.scale. Many of the compounds were alkylated xylenols for which he had made correlation charts in.the lO-lé/c region. These clarts were found to be useful in determining where substitution bands should occur for similar products obtained in this research. 120 B. Benzylation of Phenol The discovery of the acid-catalyzed ortho benzyla- tion of phenols makes possible tte synthesis of o—benzyl: phenols from benzyl alcohols. In the past, o—benzylphenols have usually been prepared from benzyl chlorides and phenolic salts by the Claisen (El) method. Nchaster and Bruner (55) described the preparation of o-benzylphenol from benzyl chloride and free phenol at 125—7500., however their yields of ortho—substituted product were poorer than found with the Claisen method. When the benzyl alcohols are more readily available than the corresponding halides, the n-toluenesulfonic acid "ethod may be preferable to the Claisen method. In addition the Claisen method reouires the use of phen- olic salts and in cases where the alkaline phenol is subject to oxidation, the Claisen method may not be suitable. The Claisen method is not suitable for the direct preparation of 2,6-dihenzylated phenols. Reaction of benzyl chloride with sodium phenate gives a mono benzylated product of the free phenol. The free benzyl phenol must be converted to the phenoxide and alkylated with another mole of benzyl halide. 121 Ctflx ORB CHzX ONa (A V2©Cfiz.—2<_ _>c~2©w2© + bdsXL The o-toluenesulfonic acid method may give the dibenzylated product in one step. QHZOH OH 2 ———2- 5.22263 +' l n-toluone\ \\ sulfonic acid Aromatic alcohols are readily accessible from aro atic acids and aldehydes. iromatic acids are reduced to the correspondins alcohols by lithium aluminum hydride. COOH QHZOH /’ LililHZ+ \\ Aromatic aldehydes are reduced to the alcohols by sodium borohydride. CHO CHzOH // v \\1 NaBH4 122 The ready availability of many aromatic alcohols makes the p-toluenesulfonic acid method a valuable preparative tool. In the laboratory it is easier to handle benzyl alcohols than benzyl halides. The halides are lachry- mators and severe skin irritants, whereas the alcohols are very innocuous materials. is water is a by-product in the acid-catalyzed benzylation of phenol ()H CHZOH OH + H20 /’ /”fi p-toluene: l -+ sulfonic x] \ I acid WCHZ \/ the rate of reaction may be determined by measuring the amount of water removed at various times during the reaction. Pratt, Preston, and Draper (12) studied the alkyl- ation of phenol with benzyl alcohol in the presence of p-toluenesulfonic acid with benzene as a solvent. They reported a 28% yield of p-benzylphenol and a 50% yield of diphenyl methane. Because of the interference of the benzene solvent, paraffin and cycloparaffin solvents were used in the work carried out in this laboratory. Quite unexpectedly the main product with saturated hydrocarbon solvents was ortho and not para—benzylphenol. Despite the dozens of workers who had studied the benzyl— ation of uhenols, this method had remained undiscovered up to now, The system of phenol, benzyl alcohol and p—tolueneCi sulfonic acid is a complex one and there are a number of possible products. One possible reaction would be a kylation of the phenol to produce p-benzylphenol. CHZOH // / + 2©22© \\ p—toluene ____ sulfonic acid Reaction of p-benzylphenol with another molecule of benzyl alcohol would form 2,4-dibenzylphenol. Unless a considerable excess of phenol is used, a large amount of dialkylated product is formed. Benzyl aroups have electron—releasing properties and increase the susceptibility of an aromatic ring to electrophilic attack. Ortho and para benzylphenols are more easily alkylated than phenol; therefore dibenzyl phenols are formed in large quantities when benzyl alco- hol and phenol are reacted in molar quantities. Further— more dibenzyl phenols would be more easily alkylated than either phenol or a monobenzylphenol. If desired, tri— benzyl phenols could probably be isolated from the reac- tion of benzyl alcohol and phenol. Reaction of phenol with one mole of benzyl alcohol miaht also yield o-benzylphenol. 124 OH CHZOH QH /\ / _ /./ \> ,g/ l‘ (W ___-._,. «im CHZo: k\ 1‘ L§;;, bx/J Gubsequent reaction of o—benzylphenol with benzyl alcohol miaht give either the 2,4- or the 2,6-dibenzle; phenol. 1 Phenol and benzyl alcohol might also react to give benzyl phenyl ether. OH CHon // f + kl 2 O 222G \. \u. —— ““ Benzyl phenyl ether might rearrange to give benzyl- ated phenols. Ether formation might also take place if a molecule of benzyl alcohol reacted with a molecule of a benzylij phenol. — CH @24on OH O 2Q / I /' A l; /\ + 3 \CH\ \\ L§¥j*€+bfii__:\ 2 \——— Reaction of a molecule of benzyl alcohol with another molecule of benzyl alcohol might also take place to five dibenzyl ether. CHZOH Dibenzyl ether could then act as an alkylating agent to give benzylated phenols. Benzyl alcohol mirht react with p-toluenesulfonic acid to Rive benzyl tosylate. CHZOH $03H + 2! Q/__\CH2-203©CH3 \ CH3 Benzyl tosylate miqht act as an alkylating agent to form benzylated phenols or ethers. Benzylated benzyl alcohols might be formed by the attack of benzyl cations on benzyl alcohol. CHZOH + / \ CH2 and ©CHZOH 5 CH2© These benzylated benzyl alcohols could further react to give polymeric polybenzyl compounds 126 /‘\ CH2 "’\/___\>CHZ.CHZ (\"/ —\>CH2 <2" \CHZOH or might form dibenzyl ethers (\\ vi or benzyl phenyl ethers CHz-O <*/ \> CHZfl/‘S or react with a molecule of phenol to form alkylated C H 2 \//-—\\ phenols. /\ I l CH2< > i§§>i(:khgo/;_m§; A perusal of the Historical section of this thesis Will show that most of these products have been isolated by one or more of the many workers who studied the benzylation of phenol. 127 A study of the structure of phenol —-—>0H ——’ Mme» T shows that alkylation can occur on both oxygen and carbon. Alkylation on carbon might take place in all three (ortho, meta, para) positions. As alkylation of the ring is an electronhilic substitution the ortho and para positions are favored since they are rich in electrons. According to Kornblum and Lurie (82), the formation of para substituted phenols is indicative of a carbonium ion process. The view that para alkylation is due to carbonium ion intermediates is supported by the fact that those halides which give carbonium ions readily are the ones which most easily give para alkylation. Thus tertiary alcohols or halides give almost exclusively para substituted phenols with most acid catalysts. Price (83) reports that secondary alcohols or halides often give substantial proportions of ortho isomers with the same acid catalysts. Curtin, Crawford, and Wilhelm (84) found that benzyl halides and sodium phenoxide give carbon alkyla- tion only at the ortho position. On the other hand they found that benzhydryl chloride, which readily gives carbonium ions, yielded para-substituted products with sodium phenoyide. Busch and Knoll (85) found that 128 trityl chloride, which likewise readily gives carbonium ions, yields para-suhstituted products with sodium phenoxide. Strong acids and high concentrations of acids will tend to give carbonium ions and subsequent para alkyla- tion. Some para alkylation is observed with p—toluene: sulfonic acid, a fairly strong acid. However, p-toluene: sulfonic acid is soluble in many organic solvents and can be employed in varyinn concentration in solution as a catalyst. Kost acid catalyzed alkylations of phenols are heteroseneous and present high acid concentrations at the catalyst interface. Dowex 50 gives a higher percentage of para isomer than p-toluenesulfonic acid although it is not a stronger acid. The same ortho/para ratios are obtained in both eyperiments with Dowex 50, although in one case four times as much catalyst is used. Apparently the acid concentration at the resin bead surface is the deter- mining factor in the amount of ortho and para isomers formed. The use of other sulfonated polystyrene deriv- atives as catalysts mirht yield different results. Rodia and Freeman (86) have discussed the problem of why the para position of a phenol is usually the favored point of benzylation with acid catalysis, even though two ortho positions are available. They claim that a phenolic hydroxyl group will solvate protons to Yield ArOH2+. Since the benzyl carbonium ion also 129 carries a positive charge, mutual repulsion of these charses would hinder formation of the para isomer. Thus, the para isomer would be at a maximum at low pH. Finn and Musty (87) report that in the phenol-formaldehyde reaction the percent of ortho-linked diphenylmethane isomer increases with increasing pH. Phenol-formaldehyde resins with a high percentage of ortho-linked diphenylmethane isomer have certain desirable physical properties. It may be that p-tolueneC sulfonic acid and similar catalysts would have consider- able utility in the preparation of high ortho resins. Such utility should certainly be investigated. In this research it was found that phosphoric acid gave a larger percentage of ortho isomer than sulfuric acid in the benzylation of phenol. Phosphoric acid is a weaker acid than sulfuric and should be less likely to yield carbonium ions. In addition it may be that some of the phosphoric acid was soluble in the cyclohexane phase of the reaction mixture. The acid in the cyclo- hexane phase may tend to give ortho rather than para alkylation. The correlation of ease of carbonium ion formation with para alkylation is obscured by two factors. One factor is steric interaction in the ortho position which would be greater with the more highly branched alkylating agents that form carbonium ions more easily. The second factor would be isomerization of an initially 150 formed ortho isomer to five the corresponding para isomer. Goldsmith, Schlatter, and Toland (88) reported that ortto alkylated phenols are readily isomerized by the usual alkylation catalysts to yield the more stable para derivatives. The more highly branched alkyl groups would also be the most subject to dealkylation. Isom— erizaticn would also increase with decreasing pH. Althoush the details are obscure it seems clear ttat formation of para-substituted phenols taxes place by carbonium-ion intermediates. Hart and Eleuterio (99) show that in nuclear alkylation of phenols with optically active °<-phenylethyl chloride, the para isomer is onticallv active. A carbonium ion intermediate would of course be planar and cause racemization. Uncatalyzed thermal alkylation with olefins to five ortho-sec—alkylphenols has been known for a long time. Skraun and co-workers (90) report that heating phenol and cyclohexane at 50000. for 75 hours gives a 30% conversion to o-cyclohexylphenol, and that m-cresol may be thermally alkylated with propylene to yield a 50-60% conversion to thymol after 70 hours at 57000. Recently Goldsmith, Schlatter, and Toland (88) described the thermal alkylation of phenol at 52000. with l-butene to sivecrdmut—butylphenol and with cyclo hexene to fiive o—cyclohexylphenol. Cycloheyanol has been condensed with phenol under 151 mild acid catalysis to give the ortho isomer. Thus Gardner (91) reacted half mole quantities of phenol and cyclohexanol at 8500. for 40 minutes in the presence of polyphosphoric acid to yield 17% p-cyclohexylphenol and 22% o-cyclohexylphenol. Tertiary alkyl chlorides nyitertiary olefins can also be used for the ortho alkylation of phenols. Plank and Socolcfskk (9E) alkylated phenol with tert-butyl chloride in the presence of boron triflucride treated drying oils to sive o-tert-butylphenol. It has also been reported (95) that 20-61% yields of o—tert-butylphenol are obtained when phenol is alkylated with isobutylene in the presence of 0.6% phosphorous oxychloride at 58—5500. Goldsmith, Schlatter, and Toland (88) have found that when phenol and isobutylene are heated to 5200C. for several hours excellent yields of o-tert—butylnhenol are obtained with little or no para isomer. In addition to the well known Claisen ortho alkyla- tion of sodium phenoxides with allylic and benzylic halides, Kundiger and Pledger (94) have found that an allylic chloride will give ortho alkylation either thermally or with an acid catalyst. For example, refluxing (143-6700.) phenol and 1,1,5-trichloro—2- methyl—l—propene for seven hours gives the ortho as well as the para isomer. Similarly phenol, 1,1,5-trichloro- 2-methyl-l—propene and aluminum chloride at 55-6000. 132 give both ortho and para isomers. Hart and Eleuterio (89) report that 0<—phenylethyl chloride and phenol in the presence of potassium car- bonate yield more ortho than para isomer. Bader and Bean (95) have reacted isoprene with phenol in the presence of phosphoric acid to yield both ortho and para alkylated phenols. Aluminum phenoxide catalyzed alkylation appears to be a general method for the ortho alkylation of phenols. Two independent groups of workers, Stroh et al.(96), and Kolka et a1. (97) have reported extensive research on this reaction. Phenols and substituted phenols may be alkylated with ethylene, propylene, and isobutylene to yield mainly the mono or di—ortho substituted phenols. Thus in basic or neutral media ortho alkylation of phenols appears to be rather common, provided of course that other criteria are also met. For example, in the reaction of benzyl and allylic halides with sodium phenoxide, C—alkylation will take place in a non—polar solvent (benzene), whereas O—alkylation will take place in a polar (ethanol) solvent. In 1926 Ingold (98) suggested that the ortho alkylation of sodium phenoxides in non-polar solvents involved the reaction of the alkyl halide with associated sodium phenoxide, whereas the formation of alkyl phenyl ethers in polar solvents involved reaction of the dissociated phenoxide ion. 155 Curtin, Crawford, and Wilhelm (84) have studied the alkylation of alkali phenoxides with benzyl and allyl halides to give either alkyl aryl ethers or 2,4-cycloC hexadienones o M+ RX '4' CH3-!//l CH3 §\ They found that dienone formation was favored by non— polar solvents and ether formation by polar solvents. Their explanation is essentially a modification of the picture presented by Ingold for preferential ortho alkylation in non-polar solvents. Goldsmith, Schlatter, and Toland (88) discuss the reason for almost exclusive formation of o—tert—butle phenol in thermal alkylation. They feel there are three possible intermediates: CH3 ‘8 H l . (I; 3 0 C’CHS \xCHz :—- ‘ H H CH5 3 C 3 3 I II III They reject the ether (I) as unlikely, since tert— butylphenol ether rearranges on heatinm to give para, not ortho-tert—butylphenol. 154 The second intermediate (II) represents an ion pair which, because of the low dielectric constant of the medium, would remain closely associated until reaction occurred. They reject this since the primary carbonium ion from ethylene would be much less stable than a tertiary carbonium ion so that ethylene should alkylate phenol more slowly than isobutylene. Instead the rates are of the same order of magnitude. The concerted process involving a transition state of type III is considered best by Goldsmith et a1. (88) to explain uncatalyzed ortho alkylation of phenols. The same three types of intermediates might be considered for the ortho benzylation of phenol. In this case I is a possible intermediate since it has been shown to rearrange under the conditions of the experiment to give ortho as well as para—benzyl phenol. However, since approximately equal proportions of the two isomers were shown to be formed in the rearrangement, it is probable that I is not the sole 155 intermediate in the reaction. Transition state II could help to explain why ortho—benzyl phenol is formed in good yields in a cycle hexane solvent, but in poor yields in a more polar solvent. In a non-polar solvent the ion pair would be tightly held as the cyclohexane medium would not favor dissociation. Ortho alkylation would take place by collapse of the ion pair. Para alkylation would take place by dissociation of some of the ion pairs to give a free benzyl ion which would attack the para position of the phenol. A concerted process involving a cyclic intermediate of type III is also possible in this reaction, although it would not be the sole path since some para alkylation takes place. It does not seem probable that the ortho benzylphenol will rearrange to yield the para isomer. Goldsmith et al. (88) suggest that the strong inhibiting effect of pentane dilution culthe formation of o-tert-butylphenol may indicate more than one mole- cule of phenol is present in the intermediate complex. maxfield's (5) data suggest this also, since he showed that as the molar excess of phenol was increased, the amount of o-benzylphenol also increased. In some of the experiments reported in this thesis no hydrocarbon solvent was used, although good yields of o—benzylphenol were still realized. Since all evi- dence in the literature points to a non-polar solvent as 156 favoring ortho alkylation, it seems strange that this should be, since phenol is a fairly polar solvent. If we assume that more than one molecule of phenol is involved in the transition state, we can explain this fact. Hart and Simone (59) showed that the uncatalyzed reaction of tert-butyl chloride with phenol was first order in alkyl halide, but second to sixth order in phenol, depending on the solvents used. In addition to the intermediates shown it is possible for benzyl tosylate to be formed initially and act as an alkylating agent. Hickinbottom and Rogers (99) reported that cyclohexyl tosylate and phenol gave 55% monocyclohexylphenols with approximately equal amounts of ortho and para isomers. Anisole and cyclo— hexyl tosylate also yield 61% of cyclohexylanisoles with approximately equal amounts of ortho and para isomers. iccordinr to these data it is not necessary to have a hydrosen atom on the hydroxyl group for some ortho alkylation to take place. In these cases both ortho and para alkylation might take place via carbonium ions. This new method of ortho benzylation of phenol in the presence of p-toluenesulfonic acid yields three to four times as much ortho as para benzylphenol with cyclohexane solvent. It is more convenient than the Claisen method to carry out since the sodium salt need not be previously formed. In addition, no ether forma- 157 tion takes place as with the Claisen method. The p-toluenesulfonic acid gives ortho/para ratios superior to the McMaster—Bruner hiéh temperature reac- tion of benzyl chloride and phenol. In addition, the total monobenzylphenol yield is higher with the p-toluenesulfonic acid method. In addition to alkylation with a p-toluenesulfonic acid catalyst in a hydrocarbon solvent, the reaction may be carried out at himher temperatures without a 'hydrocarbon solvent. This hijh temperature method has several advantages over the solvent method. First, a smaller reaction vessel may be employed for equivalent quantities of primary reactants. If a large prepara- tion of benzylated phenol is desired, this advantage becomes very important as there is a practical limit to the size flask which can be handled in the laboratory. Second, considerable time is saved in the purification of the product as it is unnecessary to remove the solvent by distillation. And third, purification and recovery of solvent are eliminated. A disadvantage of the high temperature method is the constant attention required during the benzyl alco— hol addition to maintain the correct temperature. In the high temperature reaction considerably less p-toluenesulfonic acid is required as a catalyst. The ortho benzylation of phenol may also be carried out with a sulfonated polystyrene resin catalyst. Such 158 a resin is similar in structure to p-toluenesulfonic acid as it consists of a number of p—toluenesulfonic acid units connected by methylene and benzylidene bridges in the polystyrene chain. If the resin is used as a catalyst with a n—heptane solvent, the formation of water does not take place at an appreciable rate. However, if excess phenol is used 3 a solvent so that a higher temperature (15000.) of reaction can be realized, alkylation of phenol with benzyl alcohol will proceed readily. The yields of monobenzyl phenol are sliahtly better than those obtained with a p-toluenesulfonic acid catalyst, but the ortho/ para ratio is simnificantly less. Thus ortho/para ratios of 2.5/1 and 2.4/1 are obtained with a resin catalyst, whereas high temperature alkylation with p-toluenesulfonic acid yields ortho/para ratios of 2.5/1 to 4.0/1. It may be that a steric effect in a cyclic intermediate decreases the yield of the ortho isomer when a resin catalyst is used. The cyclic inter- mediate would form less readily and if present, would dissociate more rapidly to form a carbonium ion. lj? Fermation of a carbonium ion would result in para benzyl- ation. The resin catalyst method involves the simplest method of purification. The hot reaction mixture is filtered to remove the resin and the crude product fracticnated at once. No washing or solvent distilla- tion is necessary. 'Xhen the polarity of the solvent is increased by the addition of n—butyl ether to cyclohexane the ortho/ para ratio decreases sharply. This would be expected as the more polar solvent will tend to stabilize a carbonium ion responsible for para alkylation. 140 C. Benzylation of Cresols l. Benzylation of p-cresol In the reaction of benzyl alcohol with p-cresol in the presence of p—toluenesulfonic acid para alkylation is not possible and Z-benzyl—4—methylphenol is formed in good yield. CN420F4 ()H w + ,. QCHO/ \ \ CH5 63% The physical prOperties of 2-benzyl-4—methylphenol and its phenylurethane agree with those recorded in the literature. In addition the infrared spectrum shows the bands expected in the assigned structure. As the meta isomer boils himher than the ortho isomer, if 5% or more was present, a Separate fraction should have been coll— ected. The directive influence of the methyl group would favor meta orientation although the directive in- fluence of the hydroxyl group is much stronger. The reaction of benzyl alcohol and p-cresol in the presence of p-toluenesulfonic acid was carried out by the high temperature method. This reaction should be a very suitable one to carry out with a Dowex SO catalyst. Although with benzyl alcohol and phenol lower ortho/para ratios are achieved with Dowex 50 than with tosyl acid, lkl here no para alkylation is possible. Thus p-cresol could be heated with a Dowex so catalyst to 14000., benzyl alcohol added, the water of reaction collected, and the hot reaction mixture filtered to yield the crude product. 2. Benzylation of o—cresol The benzylation of o-cresol in the presence of p- toluenes ulfonic acid yields 2-benzyl— 6-methylphenol and 2-henzyl-4-methylphenol. CN4 \/ \ CH2 CH3 /CHZOH OH _ Chi 5 a=L + Chg QCHz-©/_\-o~ The range of ortho/para values (2.5/1 to 5.2/1) is lower than found in the benzylation of phenol. This is expected since now only one ortho position is open for alkylation instead of the two in phenol. Apparently the lower yield of ortho isomer in this case is a statistical matter. The himh temperature method (Exps. 2 and 5) yields better results than the low temperature (Exp. 1) method. Also the us) e of a large molar excess of o-cresol and a smaller amount of catalyst appears to yield the best 142 results in Exp. 5. The low concentration of acid would be unfavorable to the formation of carbonium ions nec- essary for para alkylation. If more than one molecule of o-cresol is present in the transition state, a larger molar excess of o—cresol would favor ortho alkylation. Huston, Swartout, and Wardwell (4) obtained only a small amount of the ortho isomer on benzylation of o-cresol by the aluminum chloride method. On the other hand, the benzylation of phenol by the aluminum chloride method (5) gives a somewhat larger amount of o-benzle. phenol. In this work it was also found that benzylation in the presence of p—toluenesulfonic acid gives somewhat more ortho isomer with phenol than with o-cresol. 5. Benzylation of m-cresol The benzylation of m-cresol is more complex since two ortho isomers as well as a para isomer are possible. They are shown below. (3H CH4 /" I CH2 'CHa OH \ 'CH3 CH2 ©CHZ© CH3 2-benzyl—5- 4-benzyl-5- 2-benzyl-6- methylphenol methylphenol methylphenol 145 Benzylation of m—cresol in the presence of p-toluene sulfonic acid by the cyclohexane method yields a fraction containinz both ortho isomers and a higher boiling mix- ture containing the para isomer. OH CHZOH OH g + /D\ ’/ _5\\ Oriho ( E: + m 7 ' CH5 ¥r6c+ion V \/ ‘CH3 2 50% There is a lower percentage of para isomer formed than in the benzylation of phenol. This may be due to steric hindrance of the meta methyl group which hinders formation of 4-benzyl-5-methylphenol but not formation of 2-benzy1-5-methylphenol. Huston and Houk (5) had found in the aluminum chloride catalyzed alkylation of m-cresol with benzyl alcohol that less para isomer was formed than in the benzylation of phenol. Since the formation of 2—benzyl—5-methylphenol would also be sterically hindered by the methyl group, it would be expected that more of the 2-benzyl-5—methyl isomer would be formed. As expected, only 25% of the ortho fraction was shown to be the 2,5-isomer by the following method: The liquid ortho fraction was cooled in the refrig- erator to yield a white solid and a colorless oil. The white solid was recrystallized and shown to be 2—benzyl- 144 5-methylphenol. Infrared analysis then showed the crude ortho fraction to contain about 25% of the 2,5-isomer. An infrared spectrum of the oil separated from the solid 2,5-isomer showed bands correspondinm to substitution patterns found in the 2,5—isomer. The presence of this latter isomer was also shown by bromination of the oil to yield 2-benzyl-4,E-dibromo-S-methylphenol, first prepared by Huston and Houk (5). 145 D. Benzylation of Xylenols l. Benzylation of 2,5—xylenol Reaction of benzyl alcohol with 2,5-xylenol in the presence of p-toluenesulfonic acid in cyclohexane yields ortho and para benzylated isomers. OF! CHZOH 0H QCHZ CH3 66% CH3 * C ”3 + —> + CH5 CH2©OH 10% cmcm Both 6-benzyl~2,5—dimethylphenol_and 4-benzyl-2,5- dimethylphenol are new compounds. Boiling points, analysis, yields, and infrared spectra are in agreement with their proposed structures. An examination of the structure of 2,5—xylenol OH / CH3 \\ (:H3 shows the directive influences of the methyl groups would cancel each other, therefore from an electronic Viewnoint the isomer distribution should be similar to that of phenol. The presence of a 5-methy1 group should decrease the amount of para isomer because of steric 146 hindrance. On the other hand there is only one ortho position open, so from a statistical viewpoint less ortho isomer should be formed. The ortho/para ratio is actually a little himher than with m-cresol, so there may be a "buttressing" effect of the 5-methyl group by the 2-methyl group. As in phenol we would expect very little of the meta substitution product to be formed. If formed, it should distill in the same fraction with 4-benzyl—2,5- dimethylphenol. No meta isomer was detected. Benzyl—2,5-dimethylphenyl ether, if formed, would distill lower than 6-benzyl-2,5-dimethylphenol. None of the ether was detected. There is also formed a high—boiling residue as in the benzylation of phenol and the cresols. This residue is probably the dibenzyl product, however no attempt was made to distill and identify the material. 2. Benzylation of 2,4-xylenol Reaction of benzyl alcohol with 2,4-xylenol in the presence of p-toluenesulfonic acid in cyclohexane yields 2-benzyl-4,6-dimethylphenol. CHon CH 9H / «CH3 .CH2 CH3 \ + :— L\ CH3 CH3 6| °/o 147 his compound was previously prepared by Elkobaisi and Hickinbottom (46) who recorded a melting point of 6700., as compared to the 6600. found in this work. The boiling point and infrared spectrum also agree with the assigned structure. The melting point of the phenylurethane was 145.5- 144.OOC. instead of 15600. as recorded by Elkobaisi and Hickinbottom. The N-H and 0:0 bands in the infrared spectra of the phenylurethanes are similar to those found in the urethane spectra of 6—benzyl-2,5-dimethyl$ phenol, 2-benzyl-5,6—dimethylphenol and 2-benzyl-5,5- dimethylphenol. There seems to be no question that this is indeed a phenylurethane and it seems possible that it is purer than that prepared by Elkobaisi and Hickinbottom. The structure of 2,4-xylenol C2H3 shows that only one ortho and no para alkylation products are possible. Two meta alkylation products are possible and of course, in opposition to the influence of the OH group, the two methyl groups favor meta instead of ortho alkylation. However meta alkylation would probably proceed by a carbonium ion mechanism as does para alkyl— ation. Therefore ortho alkylation would be favored by 148 the rechanism as well as by the directive influence. Moreover meta alkylation would be inhibited by steric hindrance of the methyl groupCs). If meta isomer were present in considerable quantity, it should have distilled as a slightly higher boiling fraction than the 2-benzyl isomer. No such fraction was collected. No benzyl 2,4-dimethylphenyl ether was detected. If present, it would have been collected as a lower boiling fraction. Again there was present a very high boiling mater- ial which was not distilled. This might contain a dibenzyl isomer such as As shown in the benzylation of 2,6-xylenol such meta alkylation does take place when the ortho positions are blocked and the two methyl groups favor the alkyl- ation. 149 5. Benzylation of 2,5-Xylenol Benzylation of 2,5-xylenol in the presence of p—toluenesulfonic acid in cyclohexane yields the following products: OH CHZOH oH CHOCHS 390/ {in + / CH3 ; CH} NJ C i H 5 $H3 :2% .CH2 .OH (3H3 Both of these isomers are new compounds. The boiling point, spectrum, and analysis showed the assigned structure of 2-benzyl-3,6-dimethylphenol to be correct. A mixture meltinx ooint showed the 2— and Q-benzyl isomers to be different compounds. The boiling point and spectrum of 4-benzyl—2,S-dimethylphenol indicated the assimned structure to be correct. An alternate method of benzylation consisted of refluxinm a mixture of benzyl alcohol, 67% sulfuric acid, 2,5-xylenol, and cyclohexane. As expected, the majority of the monobenzyl product was 4-benzyl-2,S-dimethylphenol. Comparison of its spectrum with that of the previous sample showed that they are the same compound. In the latter method of preparation a small fraction Of a lower boiling material was collected which is 150 believed to be benzyl 2,5—dimethylphenyl ether. Chg CH3 The spectrum of this fraction shows evidence for such a structure. In one other instance, the reaction of o-chlorobenzyl alcohol and phenol, the presence of an ether was detected in the alkylation products. In this latter experiment alkylation was carried out in cyclohexane with a p—toluenesulfonic acid catalyst. The detection of such a small amount of ether in this experiment is good evidence that considerable amounts of ethers were not formed in the other alkyla- tions described in this research. Examination of the structure of 2,5—xylenol CH4 ~CHI-43 CH3 shows that one ortho and a para isomer would be expected. The ortho/para ratio found in the benzylation of 2,5— Xylenol is about the same as found in the benzylation of phenol. Since in phenol two ortho positions are open to one para position, here with 2,5—xylenol para alkyl- ation must be more susceptible to steric hindrance from 151 the methyl froup than the ortho alkylation. There is also a meta position open for alkylation, although the directive influences of the methyl groups cancel each other. No meta isomer was detected. 4. Benzylation of 2,6-Xylenol In the benzylation of 2,6-xylenol both ortho positions are blocked so that no ortho alkylation is possible. Para alkylation is favored by the directive influence of the hydroxyl aroup whereas meta alkylation is favored by the directive influence of the two methyl groups. As the directive strength of the hydroxyl aroup is eoual to that of 3 to 4 methyl groups we find that both para and meta alkylation take place. (DH OF! CH3 CH3 (Matt‘s CHZOH OH . CHBQCHS _______ CH2 + H" + \ . . The separation and identification of these isomers provide an interesting example of the usefulness of infrared analysis. The existence of two isomers was first suspected on examination of the crude monobenzyl product spectrum. Bands correspondina to both l,2,5,5 152 and l,2,5,4 substitution patterns were present. As the isomers were separated by fractional crystallization, the growth of some bands and the disappearance of others were noticed as spectra of the various fractions were examined. A considerable amount of dibenzyl product was also isolated. It is believed to be the following isomer: Oti CH3 CH3 CJ+1 Q The evidence concerning its structure is given in detail in the Experimental portion of this thesis. Elkobaisi and Hickinbottom (46) also have obtained a dibenzylated 2,6—dimethylphenol with the same melting point. It is possible that these two compounds are identical. However since their compound was prepared by a different method this fact could not be proven unless their actual work was repeated. An interesting extension of this present work would be an attempt to prepare the trialkylated product. Another interestins extension would be an investi- nation of the followi nT reaction. CHon P-‘tolueneSul‘onic ackfl CH3 A monobenzyl product seems likely, however the for- mation of a dibenzyl product, containing six groups on the benzene rinm, might require more vigorous reaction conditions. CH3 ca3 154 5. Benzylation of 5,4-Xylenol Reaction of benzyl alcohol and 5,4-xylenol in cyclohexane in the presence of p-toluenesulfonic acid yields a mixture of the two possible ortho isomers. cuzos O 0;: QWQ £10 0 (3H3 53£998 31796 The yield of monobenzyl product is somewhat low as only a 50% excess of the xylenol was used In determininé which ortho isomer would be formed the directive influence of the methyl groups would cancel each other. Therefore the relative amounts of the two ortho isomers should be determined by steric considera— tions Do Apparently this is the case, since about 70% of the monobenzyl product is 2-benzyl—4,B—dimethylphenol and the other 50% is 2—benzyl-5,A—dimethylphenol. The benzylation of m—cresol also yielded two ortho isomers: (N4 Ohi 2 5 o/o 0“ “We ‘75 /' \ 3 CH3 3 + CH2 60 % 90/0 These are the only two possible monobenzyl isomers. The ortho/para ratio should be similar to that of phenol as there are two open ortho positions. However the actual ratio with 5,5-xylenol is 5.9/1. It may be that para alkylation compared with phenol is more hindered by two methyl aroups than ortho alkylation is by one methyl and one hydroxy group. This seems logical since in the alkylation of phenol, ortho alkylation is subject to hindrance from one hydroxy and no methyls, whereas para alkylation is subject to no hindrance. 156 The meltinf point of the 4—benzyl-5,S—dimethylphenol was lower than that found by Buu—Hoi, Sy and Lejeune (61). More of this isomer was prepared by the cyclohexane- sulfuric acid method but after repeated recrystalliza- tions of this latter sample, the melting point remained the same as that of the material prepared by the p- toluenesulfonic acid method. Spectra and mixed melting points showed the two samples of 4—benzyl—5,5-dimethyl$ phenol to be identical. Spectra and mixed melting points also showed the samples of 4-benzyl—5,5-dimethyl$ phenol to be a different compound than 2-benzyl-5,5- dimethylphenol. As there are no other possible mono- benzyl isomers and the Spectrum of 4-benzyl—3,5—dimethyl$ phenol amrees with its assimnment, the structure should be as indicated. 157 E. Alhylation of Phenol with Substituted Benzyl Alcohols l. Alkylation of Phenol with p-Methylbenzyl Alcohol When phenol in cyclohexane was alkylated with p-methylbenzyl alcohol in the presence of p—toluenesul— fonic acid, the following products were formed: OH CFQOH )7 \ OH ,/ CH3 _ >CH2 /l \ + ——>- \ -+- éHz CH3 5776 “‘70 CH3 Possibly both of these isomers are new compounds, although Kindler (44) mentions the preparation of a (p-methylbenzyl) phenol in his patent. He does not specify which isomer, nor does he mention any physical data. As this material is one of a number of compounds claimed to have been prepared, it may be the paper invention of a patent attorney. The use of non-existent "typical examples" is well known in the patent art. When the ortho and para benzyl isomers were solids, a mixed meltina point was taken to prove that the two compounds were indeed different. In the reaction of p-methylbenzyl alcohol with phenol, both of the mono- benzyl isomers were liquids. However, when phenyl 158 urethanes of the two isomers were mixed, the depressed meltins point clearly showed these two compounds to be different. In the reaction of p—methylbenzyl alcohol with phenol the orientation possibilities are the same as with phenol. No possibility of meta substitution should exist as it does with the xylenols. However it is possible that substituents on the benzyl ring may affect the relative rates of ortho and para benzylation. That such an effect does occur is apparent in the high ortho/para ratio of 5.5/1. Alkylation of phenol with benzyl alcohol generally gives ortho/para ratios in the ranae of 5-4/1. The methyl group on the benzyl ring of the alcohol H3C-CHZOH is an electron-releasing group. Kinetic studies on relative rates of benzylation in this work showed total alkylation was faster with p—methylbenzyl than with benzyl alcohol. Since total alkylation is faster, the rate of ortho alkylation must be increased, although the rate of para alkylation could be increased less than the ortho, remain the same as before, or be decreased. 159 2. Alkylation of Phenol with p-Isopropylbenzyl Alcohol Alkylation of phenol with p—isopropylbenzyl alcohol in the presence of p-toluenesulfonic acid yields a hiah ratio of ortho to para isomer. Chi CH OH 9H3 ‘9“ Z (:44IIIF}CJ4zx/l l 59 °/o CH:H \CH 3 lZO/o /CH CH3 CH3 The ortho/para ratio is hither than in a similar reaction of benzyl alcohol and phenol. As the isopropyl group on the benzyl rinr is electron—releasing, the rate of ortho benzvlation may be accelerated by such an effect. The same high ortho/para ratio was also noticed in the reaction of p-methylbenzyl alcohol and phenol. As there are no methyl groups present on the phenol rinm, little if any meta isomer should be formed. None was detected. The remainder of the 29% yield is prob— ably polybenzylated phenols. Identification of the ortho and para isomers seems to offer no problems. However recrystallization of the 4-(p-isopropy1benzyl) phenol did not give sufficient pure material to be used in the preparation of a phenyl urethane derivative. Therefore an additional alkylation 160 was carried out with a sulfuric acid catalyst. CH10F4 __ :: 1.5/l ratio of ortho/para isomers Surprisingly enourh, the tendency toward ortho alkvlation is so stronm with p—isopropylbenzyl alcohol that even with a sulfuric acid catalyst more ortho than para isomer is formed. 5. Alkylation of Phenol with o—Chlorobenzyl Alcohol Alkylation of phenol with o-chlorobenzyl alcohol in the presence of p—toluenesulfonic acid yields the followinr products: 0” 49% C 1420 + CHZOH —————>019/ 2 OH 5°/o QCHZ— Q-/ \ C1 Total monobenzyl products amount to 73%, about the same percentage of monobenzyl product as found in the alkylation of phenol with p-methylbenzyl and p-isoC 161 propyl alcohols. However here the ratio of ortho to para isomers is much lower. Chlorine on the benzyl rin: has an electron—with— drawing effect and appears to hinder the rate of ortho alkylation. The ortho/para ratio is lower than found in the alkylation of phenol with benzyl alcohol. This same lower ortho/para ratio is found in the alkylation of phenol with p—chlerobenzyl, p—bromobenzyl, and 2,4— dichlorobenzyl phenols. Huston and co-workers (8) found in the Claisen condensation of o—benzyl chloride with phenol that poorer yields of the ortho benzylated phenol were ob- tained than with benzyl chloride and phenol. They also found when 2-, 5-, and 4-chlorobenzyl chlorides were condensed with phenol in the presence of aluminum chloride ortho-benzylated product was formed only with the 2- benzyl chloride. Apparently in alkylation with a 2- chloro substituted ament, rate of both para and ortho alkylation is reduced. In the case of o-chlorobenzyl alcohol in the presence of p—toluenesulfonic acid, as reported here, the ortho and para alkylation rates must be retarded to the point where some (5%) of the ether is formed. 4. p—Chlorobenzyl Alcohol and Phenol Alkylation of phenol with p-chlorobenzyl alcohol in the presence of p—toluenesulfonic acid yields the followinmz OF! CH4 H OH ' C 2 OH C'.CHZ / I L' + CH2 , + ———> \ 55 °/. C! I8°/o Cl No ether was isolated in this reaction, althounh some had been found in the reaction of o-chlorobenzyl alcohol with phenol. Under similar reaction conditions an ortho/para ratio of 5.5—4.0/1 would be realized with benzyl alcohol and phenol. Thus p-chlorohenzyl alcohol gives a lower 3.0/1 ratio, as do the rest of the halosen substituted benzyl alcohols. Separation of ortho from para isomers by distilla- tion works rather well for most of the alkylations carried out in this research. However 2-(p—chlorobenzyl) phenol is difficult to separate from 4-(p—chlorobenzyl) phenol by the type of distillation used. a similar difficulty was experienced with the separation of 2-(p-bromobenzyl) phenol and 4—(p—hromobenzyl) phenol. Calculation of isomer yielis was carried out in both cases by isolation of the pure compounds and use of their spectra to analyze isomer content of the distillation fractions. 163 5. p-Bromobenzyl Alcohol and Phenol Alkylation of phenol with p-bromcbenzyl alcohol in the presence of p-toluenesulfonic acid yields the followinq: OH CHon 8|.OCHOH I 2 + CH2 -————a» 48% 20% Br The ortho/para ratio of 2.4/1 is less than that found with p-chlorobenzyl alcohol (5.0/l). A p-Br substituent has only a slightly more positive sigma value (+0.252 compared to +0.226) than p-Cl, so that the difference in ratios is not all due to differences in electronic effects on the reaction rates. The melting point of the 4—(p-bromobenzyl) phenol agrees with the value recorded by Vu, Guile, and Huston (60) for the hifih eltin” crystalline modification. The sample in this research was slowly recrystallized from more than the minimum amount of hexane necessary to dissolve the sample at reflux. Similar conditions are described by Wu, Guile, and Huston (60) for their preparation of the hish meltinz modification. None of thw low meltinq modification was obtained. 164 6. 2,4-Dichlorobenzy1 Alcohol and Phenol If phenol is alkylated with 2,4-dictlorobenzyl alcohol in the presence of p-toluenesulfonic acid the followins products are obtained: CN4 CH OH 20H OH CI.CH2 ' / c‘ \ + CH2 l + ____,.. CI \ C1 CI any. 21% CI The ortho/para ratio of 2.5/1 is lower than found with benzyl alcohol and phenol and thus fits the general pattern found with halogen substituted benzyl alcohols. As about 5% of an ether was believed to have been formed in the alkylation of phenol with o-chlorobenzyl alcohol, the distilled products in this experiment were carefully examined for the presence of an ether. There was no fraction in the ether range boilinq below the 2-(E‘,4—dichlorobenzyl) phenol fraction. Therefore the infrared spectrum of this fraction was carefully exam— ined for evidence of an ether impurity. There is a sliaht shoulder on the R.9ptband at 9.1%“ which mifht be due to a trace of ether. However as the 8.1%AL ether band is an extremely strong one, there should not be more than a small amount present. 165 F. Relative Rate of Benzylation of Substituted Benzyl Alcohols The relative rate of phenol benzylation was deter- mined with six different benzyl alcohols. The rates with these alcohols fell in the followinr order: p-iso—Pr> p-Me) H > p—Cl > o—Cl ) 2,4-o1 In aeneral the areater the electron releasing ability of the substituent on the benzyl rins, the areater the rate of alkylation. The considerable de— crease in rate of the ortho- over the para-chloro alkyl— ation may be due to steric hindrance. The followinf table shows the reaction rates as well as the ortho/para ratios for the six alcohols studied. 166 Table VI. Phenol Alkvlation with Benzyl Alcohols *— Benzyl alcohol Reaction % Monohenzyl Ortho/ Hammett half—times phenols para sigma min. ratio function p—isonronyl 26 71 4.8/1 - D-methyl 52 68 5.5/1 -O.17O unsubstituted 56 (a) (a) 0.000 o-chloro 47 75 5.0/l +0.226 o-chloro 130 68 2.5/1 - 2,4-dichloro 254 75 2.5/1 - (a) The monobenzyl yields and ortho/Dara ratios were not determined for this narticular experi— ment. However, judainq from similar experi- ments with benzyl alcohol the monobenzyl yield should be about 70% and the ortho/oara ratio about 5.5-4.0/1. It is seen that the monobenzyl yields are all about equivalent, so the exoeriments can be comoared on the basis of their reaction rates and ortho/para ratios. The reaction half times vary from 26 minutes to 254 minutes, whereas the ortho/para ratios vary only from 2.5/1 to 5.5/1. Therefore the rate of either ortho or para alkylation does not annear to remain constant while the total rate chanmes. However, the reactions with o-chloro— and 2,4—dichlorobenzyl alcohols are possibly subject to steric hindrance. If this is true, there is a good correlation between the reaction rates of the other four benzyl alcohols and their ortho/para ratios. The faster the reaction rate, the larger the 167 ortho/nara ratio. This correlation bet een reaction rate and ortho/ Dara ratios could mean the rate of para alkylation is constant while the rate of ortho alkylation increases. The reaction may proceed as follows: (6‘16 para product Con§6n* in creased ortho product CHZOH OH R or the reaction could so throueh a common intermediate: para CHZOH 5:3 OH era gafil . + . ® ’ Common /®' infirmedkab ' 'n mambo R We In this latter case step one could be subject to steric hindrance from ortho substituents and cause the large decrease in rate noted for o-chloro— and 2,4— dichlorobenzyl alcohols. It should also be noted for the latter two alcohols that their ortho/para ratios are substantially the same as that of p—chlorobenzyl alcohol, even though the rates are much slower. This means that it must be step one which is subject to steric hindrance. Pratt, Preston, and Draper (12) found that in the alkylation of anisole by substituted benzyl alcohols, 168 the reaction rates varied with the substituent on the alcohol. The rates decreased in the followins order: 01150 > CH5 > H > Cl Ttey also found ttat n-nitrobenzyl alcohol reacted very slumqishly. Their reaction rates decreased with decreasinr ability of the substituent to release electrons. The same order was found to hold true in this research. Pratt and Erickson (100) studied the etherification of triphenylcarbinol and benzhydrol with n-butyl alcohol in the presence of o-toluenesulfonic acid. They found an increase in the electron-releasing ability of the para substituent of a benzyl alcohol would increase the rate of etherificatien. The N-alkylation of anilines by benzyl alcohols in the presence of activated nickel and sodium benzylate was studied by Pratt and Frazza (101). They found the reaction rate decreased with decreasing ability to release electrons. The reverse order was found to hold true with para-substituents on the anilines. It has also been shown (102) that the rearrangement rates of alkyl p—X-phenyl ethers may be correlated by Hammett's equation using U’p+ constants. The rate decreases as simma becomes more positive. Again this is the same order as found in the present research. Thus the variation in reaction rates is about what may be expected from a study of related work in the literature. 169 G. Scope of the Acid-Catalyzed Ortho Alkylation l. Phenol and Various Alkylating Agents A number of alkylatins asepts were refluxed in cyclohexane in the presence of phenol and p-toluenesulf- onio acid. In general, the alkylatina agent should be a benzyl type alcohol, vinylos of a benzyl alcohol, or an olefin derived from a benzyl alcohol, in order for alkylation to take place under these conditions. The results are listed in the following table. 170 Trhlr VII. Phenol and Various Alkylating Agents Alkvlatinq agent Structure Results Cinnamyl alcohol E—Phenvlethyl alcohol Prooarqyl alcohol Benzhydrol tert-Amyl alcohol Cyclohexanol l-Phenvlethyl alcohol Styrene -CH=CH—CH20H HCEC-CHQOH H /C. OH /CH5 1. " "' ‘H C235 C‘gH \.N Polymeric products No reaction No reaction Monobenzyl products p-tert—amylphenol No reaction Monobenzyl products including 2(1— phenylethyl) phenol Monobenzyl products 171 If Cinnamyl alcohol were added very slowly and the acid concentration made very low, it mimht be possible to isolate alkylated phenolic products. Benzhydrol has a very fast reaction rate and yields products boilinr in the monobenzhydryl range. However, benzhydryl chloride has been found by Curtin, Crawford, and'flilhelm (84) to yield para-substituted products with sodium phenoaide. The reaction of benzhydrol with phenol should be investiaated further. Tertiary alcohols or tertiary halides give almost exclusively para-substituted phenols with most acid catalysts since they very readily form carbonium ions. This terdency toward para alkylation is so strong that p-tert-amyl phenol was the main product when tert-amyl alcohol was used as an alkylatinfi agent. Tert—amyl alcohol, rather than tert—butyl alcohol,was chosen for this exoerinent as it has a higher boiling point than tert-butyl alcohol and was less likely to be lost by dis- tillation during reaction. It would be expected that l-phenylethyl alcohol would aive good yields of ortho product because it is a homoloaue of benzyl alcohol. An ortho/para ratio of 2.1/1, lower than with benzyl alcohol, was found for the phenylethylated products. Apparently l-phenylethyl alcohol more readily forms carbonium ions to yield para alkylated products than does benzyl alcohol. Althourh the monostyrenated phenol fractions were 172 not identified, it appears from their boilinr ranges that fraction 2 would be 2—(l-phenylethyl) phenol and fraction 4 would be 4—(l-phenylethyl) phenol. If this is so, then the followin: comparison can be made: Alkylatina a. g. agent ortho para Ortho/para ratio l-Phenylethyl 50.7 24.0 2.1/1 alcohol Styrene 49.0 24.7 2.0/1 This close agreement of results suggests that both alkylations miéht proceed through the same intermediate: CHZCHZ OH Q -.» H 2 HO‘C'CHa Q + Common —=’ intermediate ———-,-PrOdUCtS Q- 0 .——_> Complex Ga 173 Z. Benzyl Alcohol and Various Substrates ‘ a number of nucleoohilic substrates were refluxed in cyclohexane in the presence of benzyl alcohol and p-toluenesulfonic acid. In general, phenols with alkyl, halo, and hydroxy arouns will undergo alkylation by benzyl alcohol. Althouak not tested, it is felt that a nitro group would make the aromatic nucleus more difficult to alkylate. Phenolic ethers will also undergo benzylation as expected. Aromatic amino groups do not themselves react and apparently inhibit the reactivity of adjacent hydroxy groups. Thionhenols are not alkylated under these conditions. Results are listed in the following table. 174 Table VIII. Benzyl Alcohol and Various Substrates ._....- ---a _. ._ fl- NucleOphile Structure Results O_Amincphenol OH NO reaction NHz C} p-Chlorophenol 2-benzyl-4-chloro phenol o-ChlorOphenol Honobenzyl products 2—benzyl—4-ethyl OH . phenol C2 H5 Monobenzyl products p—Ethylphenol o-Ethylohenol Anisole OCH3 Monobenzyl products Catechol (”4 Monobenzyl products .054 Thionhenol SF4 No reaction N-I..ethyl aniline HN'CHa No reaction /’ \\. 175 The benzylation of anisole appears to aive a mix- ture of isomers which boils closer torether than do the benzylated phenols. This is expected since the lower boilinv point of ortho-benzylphenol, as compared with para—benzylphenol, is due to the interference of the ortho—benzyl group with intermolecular hydrogen bonding of the OH groups. Perhaps the isomeric mixture of benzylated anisoles could be determined by demethylation to yield mixed benzylphenols which could then be separated by distilla- tion. Alternatively, the pure methyl ethers of ortho- and para-benzylphenols could be made by methylation of the phenols. The spectra of these pure ethers could be used as standards for the determination of relative amounts of ortho— and para-benzylanisoles. In the benzylation of catechol, if fraction 2 is the B—benzyl isomer and fraction 4 is the 4-benzyl, the ortho/para ratio would be 1.8/l. Such a result would suggest that catechol is more susceptible than phenol to para alkylation. However the presence of two adjacent hydroxyl aroups miqht sterically interfere with the path of the ortho reaction. Thioohenol is not alkylated under these conditions although this in no way precludes the fact that a reaction may take place if conditions were altered. Thiophenols have a greater tendency to form thio ethers than phenols do to form ethers. In fact, thiophenol is 176 very difficult to C—alkylate but instead usually forms S—alkylation products. One of the few exceptions is recorded by Laufer (103) who reacted thionhenol with ethylene, propylene, l-butene, 2-butene, l—pentene, cyclonentene, cyclohexene, and cyclopropane. He obtained o-alkylthiophenols, 2,6-dialkylthionhenols, and alkyl aryl sulfides. No meta alkylation and little or no para alkylation took place. Aluminum chloride and certain other Lewis acids were used as catalysts at temperatures below 3000. and in some cases as low as -7OOC. Neither N—methyl aniline nor o-aminophenol was benzylated in refluxing cyclohexane with a p-toluenesul- fonic acid catalyst. However, it is felt that conditions mirht be found under which ortho alkylation would take place. The literature contains scattered reports of similar reactions with amines. Kolka, Ecke, and Closson (104) reported the ortho alkylation of aniline by olefins using an aluminum anilide catalyst. This reaction is analomous to the reaction of Kolka et al. (97) where phenols are ortho alkylated by olefins using an aluminum phenoxide catalyst. Hart and Kosak (105) reported the ortho alkylation of aniline With styrene in the presence of anilire hydrochloride. 177 H. Mechanism of Ortho Alkylation Kuch additional wore is needed on this reaction before a mecharisn can be established. Reaction rates should be studied, kinetic orders established, the in- fluence of substituents of both alcohol and phenol studied, and possible intermediates investisated. This present work was concerned largely with the discovery of a new synthetic tool, determination of favorable reaction conditions for it, and an investiaation of the The influence of substituents on the benzyl alcohol on reaction rate was studied, but such a study helps to predict for What type of aromatic alcohols this reaction will be suitable, rather than establishes a mechanism. It was shown that alcohols havinq electron—releasing substituents should aive the best yields of ortho isomers. In another instance a benzylation of phenol was stopped when 71% of the water had been removed and the product examined for dibenzyl ether or benzyl phenyl T et or as intersediates in the reaction. Neither was found. In a further experiment benzyl phenyl ether was rearranred in the presence of p-toluenesulfonic acid and ('1' I c products determined. The ortho/para ratio of 1.2/l is much lower than obtained in alkylation of phenol with benzyl alcohol. However, in this case a better reaction method may have been to add the benzyl ether slowly to 178 a cyclohexane solution contairinr p—toluenesulfonic acid and excess phenol. Alkylation of ore mole of phenol with 0.5 mole (eeuivalent to 1.0 mole benzyl alcohol) dibenzyl ether save an ortho/para ratio of 2.7/1. This is somewhat lover than that obtained in most benzylations of phenol, hovever higher ratios may have been obtained if an excess of phenol were used. Althourh very little work has been done to elucidate a mechanism, it is possible to write one for the reaction with the understandinm that such a mechanism is tentative. This mechanism should explain certain facts: 1. why the ortho position is favored over the para position by more than a 2:1 statistical factor. 2. Why non-polar solvents favor the production of the ortho isomers, whereas polar solvents inhibit this reaction. 5. ‘flhy a hish temperature method is possible with excess phenol as the solvent. Or, to phrase it another W37, Why the ortho-alkylation is dependent on the con- centration of phenol. 4. ‘Shy the rate of total alkylation is subject to steric hindrance (as found with o-chloro— and 2,4— dichlorobenzyl alcohols). 5. Why the rate of ortho-alkylation is enhanced over para-alkylation by electron-releasing substituents on the benzyl alcohol. 179 A plausible mechanism would be the follOWing: A molecule of benzyl alcohol is protonated by the acid catalyst H G) CHZOH CHZOH ‘— .7— p—toluene sulfonic acid The protonated alcohol then reacts by two different paths, one (SNl) path leading to the para isomer and the other (8N2) leadinr to the ortho isomer OH CHZQ OH H G + CHZOH ‘Hz CH2 SNZ O-- “H ‘ H OF} OH H V "9 Q —-- p, The mechanism could be written so that the primary intermediate is an ion pair containinq the acid. 63% 0.. H cu 0 .CH 20 p—toluene 2. 3 Q 3 sulfonic acid 180 Despite the exact nature of the comaon intermediate it seers verv reasonahle to assume that it then reacts by two paths (8N1) and (3N2) as shown to give rise to both ortho and para isomers. The mechanism of this reaction and the nature of the intermediates should be further investigated. 181 I. Extensions of Present Work There are a number of areas in which this work could be extended: 1. Synthesis a. Alkylating agents The alkvlation of phenols with fluorobenzyl alcohols has never been studied. Such an area would be a logical extension of the work on the chloro- and bromobenzyl alcohols. The fluorobenzyl alcohols are readily access- ible via the fluorotoluenes CH3 COOH CHZOH P1) Paterno, E., Gazz. Paterno, E. and Fileti, M., Ga 531-3 (1975)- c., El, ago—s (1882). Renrie, F. H., J. Chem. So 190 191 A., Ber., 15, 153-4 (1939 en. 390., 52, 406-11 (1396). Gazz. chim. it91., 551. 495-5 (1995); (1965) m and Walter, W., Ann., 554, 567- 85 (1904). -}onbert, E. and Buchler, C. C., J. Am. Chem. 300., 3;, 2059-72 (192:). Io, Z. angewc :5‘ H {To C Chem., 55, 478-9 (1925). H., Ann., 445, 225—46 (1925). ‘VI, U. S. Claisen, J. and Reich, , t.,'30hrerz,1., and Sohuleman, . gg. u331.(19 é) C53 (l’cC);C ._A., on Alphen, J., Rec. trav. 0him., fig, 799- 812 (19 27). 5. F., J. Chem. 300., 1928, 528 E. L., J. Chem. 800., v. Br raun, < ‘ShC’rt , Short, 7. F. and Stewart, 19299 555- 9' Behamel, C. and Breiensehuer, H., Ber.,§2, 1568-77 (1954)- Foldi, 2., Ber., Q1, 15 O9— 15 (1923). Fever H. and Bernhauer, K., konateh., 55, 721-52 ( 9&9 . :1mcfr, r. P., U.3. 2,C16,B48 (1955); 0.1., g3, 5068 (1955) Indian Chem. 800., J} rm 3, V. N. and Dutt, 5., J. _£, 775 (1935)- 1, J. B., J Indian Chem. 800., 1;, 616—21 (1955) I Ind. En,. Chem., $0 fickaster, L. and Br1ner, J. M., 9“ 508-8 (1996). 2:9 Andrianov, K. A., J. Gen. Chem. (U.S.S.R.), £99 6718 (1955)- '46—51 (1955); C.A., Nonaoelli, fl. J. anfl Hennion, 1722-4 (1941). G. C., J. Am. Chem. 300., 55, Sen. Chem. (U. S. S,R.), 15, Teterin, V. K., J. BOX? (1145) 115 (1945); 0.1., 39, [1 1'1. 45. 4 "y ’3 ('1 \J. Illiri, C., G322. chim. ital., 22, 90u-13 (1948). Ckenev, L. C., 011th, R. R., and Binklev S. B., J. (km. Chemo SOC. , ml, (AD—61+ (1949). “ rt, H. uni Simone, J. H., J. Am. Chem. 300., 13 21. 545-52 (1949). , %. add Kurlina, I. P., Zhur. Cbehchei 20 2153—67 (1950); C.A., E5, 7074 (1951). Setkina, V. J. and F1r3.nov, D. H., Izvest Akad. Nauk S.S.S.R., Ctdel.K1im. Nauk, 1951, 91—5; W (1052)- Tarb e11, D. S. and Petronoulos, J. C., J. Am. Chem. Soc., 23, 244-8 (1952). 3 Sta. >3, G., and Nudenber, W., ch, 1. , 113, 50; (1952). Kigdler, K., Ger. Patent 824,058 (1951); C.A., 32, 7594 (1955)- and Hickinbottom, W. J., J. Chem. Elkobiiee, F. M. Soc., 1758, 2451-2. :91 0 J. 9 Jo Chemo Elkobaisi, F. K. and Hickinbottom, Kornblum, N. and Lurie, A. P., J. Am. Chem. 800., .21. 2705-15 (1959). , Anaiq fac. farm., Univ. Recife, L, .A., 53, 24495 (1960). “ ’ and Uickinbottom, W. J., J. Chem. ElkebiiSJ, . w. Paterno, F. 8rd Razzarq, C., Gazz. chim. ital., 8, 533—5 (187?). C “ Gazz. shim. ital., 51, 1, 469—74 Venturi, a. 5., (1 L01) Claisen, .fiknn o , 4"":J‘é) , L., Kremers, F., Roth, F., and Tietze, E., 210-42 (1925). 59B, 2055 (1925). Comvt. rend., 19,, 1259—42 (1954), A., J. Orfi. Chem., 12, @chorifiin, P., Ber., Ckichibabin, A., and Harlund, E. Hart, ff 596- 9 (1950)- 57. 59. fi \3 (D (j a9. 70. 71. 72. 195 eatleV, 1. B., Fithihhcn, B., Cheney, L. C., and Binklev, 3. B., J. Am. Chem. Soc., 72, 4445-5 (1?50). ‘— Bailev—flocd, V. V. and Cullinane, N. 1., Chem. and Ind., 195?, 543. Feilstcin's Handbuch fler Cr°zn1°chen Chemie, 6, II, 645. Auwers, E. 1nd Janesen, E., Ann., 495, 44-65 (1950) 1 R(V, N. E. and Iqters, W. A., J. Chem. 30c., 2255-5- Ruu-Hci, N. 9., and Iejeune, C., Bull. soc. chim. France, 1‘1 , 1757—10. Flarmnnn, E., Gates, L. 1., and Shtencv, V. A., J. Am. Chem. Soc ., 7 , ~64 ~66 (15‘4’9). ., Clteney, L. C., and Binkley, S. B., .0c., 21, 64- 66 (1 M9). Jheatley, N. B J. Am. Chem. 9 ‘HheatleYa N- B°9 Fi1t231bbcn, fl- E°, Cheney, L' C'9 and Binkley, 3. B., J. Am. Chem. Soc., _§, 1655—9 (1950). Schnlewann, V., Deutsch. med. Wochenschr., 36, 1051 (1920). eichet, J., Czech. Patent 85, 057 (1955); C.A., 59, 1C2 1 (1‘56). Rhumard, R. 8., Scan Sanit. Chemicals, 29, Do 5, l5U-',155 171 (1955 ; C-A-9 $2, 5059 {—95 5). 1161‘??? 11 1 (19525; 'heqt 16V, ‘. B., Fit zzibbon, W. B., Minor, W. F., Smith, R. R., Chenev, I. C., and Bin1fley, S. B., J. Am. Chem. )oc., Zfi, 5575- 7 (1950). Wheat1eV, H. B. and Cheney, L. C., J. Am. Chem. cc., 23, 2940-1 (1952). , Proc. Soc. Apfil. Bacteriol., _5, 38—45 1., 32, 7047 (1955)- I n \l “Pheatley, H. B., Fitzgihbon, W. E., Stiner, G. F., and Cheney, Ia C., J. Am. Chem. 800., 22, 747-9 (1957). Chai1—in , j . V . ’3n’1 BI‘OVVH, Iv. G o , Jo A111. Chemo SOC. 21, 1 2-5 (1)49) .- 174 74. Cannizzarr, 3., Ann., “224, 255 ( 934 75. 310rt, J. F. and Stewart, I. L., J. Chem. 800., 1 2), 555- 9. 76. "Hanfihook of Chemistry and Physics". 59th ed. Chemical Rubber Publishin: 00., 1958. 0.1097. 77. “0:3-n, H., Arch. Pharm., 221, 417-51 (1955). 78. Chumaevskii, N. A., Zavod. Lab., 2269, 26, 957—9. D. E. and Thcmnaon, H. H., J. Chem. 300., _ .._ I , and KcGovern, J. J., 132 . 0103., +1, 418 (1350). t “1. Pare, 5., Revue dc 1'Ir3 itJt Francais d1 Petrole, 14, 695-729 (1900). 52. Eernllu:, N. and Llric, A. P., J. Am. Chem. 800., 85. Price, C. C., er. Reactions, 5, 1 (1942). 94. Curtin, D. Y. , Cravforfl, R. J., and Wilhelm. M-, J. Am. Chem. 300., 80,1591 (195L8 ) 85. Buech, I. and Yn011, R., Ber., 22, 22 49 (1927). 76. R0113, J. 3. and Freeman, J. H., J. er. Chem. 23/! , 231 (1959) . C"7. TTinn, 3. R. and 1m mty, J. H. G., J. Soc. Chem. 1nd,, Sunnl. 18 311e NC.2, 3. 49 (1950). te N. J., and Toland, "I. QClQQV‘ith, F0 810, J01 la Pr, ; 1571 (1353) t W. C., J. CrT. Chem., 25, 89. Her+, H. and Eleuterio, H. 5., J. Am. Chem. 800., _29 516 (1)58) 00. Skraun, 3. and Beifles ,W., Ber., 6GB, 1070 (17,2 m) ,‘ a, n ‘I"'"'!" akraup, 3., So chcllkopf, K. and oerinl, A., U.3. 1.,kf1 (NC) ,%11 (18,1132). 91. Gardner, P. D., J. Am. Chem. 300., 76, 4550 (1954), 92. Iank, C. J. and Socolcfsky, J. F., U.S. 2,460,795 (1943). 195 95. Pritifih Patent 725,873 (1955). Q4. Ynndi5(r, U. G. and Pledfiers H°9 U‘3° 2’837’575 (195“)- 95. Bader, A. R. qnfl Bean, 9. C., J. Am. Chem. 800., “C, 5075 (1958). 96. Stroh, R., Rev del, R., and Hahn, H., Angew. Chem., _2 W <1“ 7). 97. KOlku, 1. J., Nepolitana, J. P., Filbey, A. H., and Ecke, G. C., J. Crm. Chem., :g, 642 (1957). 99. Inqold, C. H., Ann. Repts., 1920, 14?. 9?. Hickinbottom, H. J. anfl Rogers, 4. H., J. Chem. 300 130. Pratt, E. F. and Erickson, P. 3., J. Am. Chem. Soc., 79. 7a-8 (1356). 191. Pratt, E. F. and Frazza, E. J., J. Am. Chem. 800., 29. (~174— 9 (1954) ’ 9 102. 'ahite, i. N. Gwynn, D., Schlitt , H., Girard, C., and Fife, H. J. Am. Chem. 300., 80, 3271 (1958); Goerinm, H. . 1nd Jacobson, R. B., J. Am. Chem. 300., fig, 52 7 (1958). CE. Laufer, Q. J., Abstracts of Papers, 157th A.C.§. fleetinq, Cleveland, 0., Abril 5—14, 19’“ p.61 0. (D U. , 104. Pb 11x <3, 1. J., Fake, G. 3., and Closson, R. D., U. .VB. 2,914,646 (1957) 1C5. “f ". and cha', J. B., J. Org. Chem., fig, 1752 .t, 157)- A H5“ kg” APPEND IX I AT AND GRAPES FCR KINETIC RUNS C? SUBSTITUTED BEI‘JZYL ALCOHCLS 137 Rate of Phenol Alkylation with o-ieo-Propylberzyl Alcohol Time in minntee Ml. of water distilled 2 0.9 7 2.0 15 5.? 15 5.8 19 4.9 $4 6.1 20. 6.9 35 7.9 39 8. 46 9.5 55 10.5 65 10.6 71 10.9 Ml . H,C 3.0 6.0 5.0 3.0 2.0 1.0 reaction half time = 26 minutes l l l L l l 10 Fi ". 1 . 9111/73’19tiCT1 20 50 40 50 60 Time in Hinutes of rhenol by o—iso—nropylbcnzyl alcohol 199 Rate of Phenol Alkylation with p-Letkylbenzyl Alcohol Time in minutes K1. of water distilled # 0.6 8 1.5 15 2.5 18 5.7 22 . 4.7 25 5.5 52 6.5 37 7-5 41 8.1 47 9.0 2 9.5 59 10.1 52 10.6 69 10.9 Q.) d 3,2" 7.0 ,. reaction half time 6 OP 2 2:) ° .2 2. minutes 5.0_ \N O O I 20(1)“ 0 1.0. O O L n L L A A 10 20 50 40 50 60 we in Finutee Ti of phenol hy n-methylhcnzyl alcohol FT .2. klkylition 201 Rote of Phenol Alkvlation with Benzyl Alcohol Time in minutes $1. of Water distilled \] 0.8 15 2.5 22 5.7 28 4.9 5.9 6.8 \N \N I x \N 45 7.6 47 8.2 52 8.7 62 9.7 72 10.2 10.6 102 10.9 \D R) DE 9'" ? h \ Om- ._ "D \D O I \N O I reaction half time ll 56 minutes A P l l 20 5; 40 50 60 Time in Vinuteo Llhylaticn cf nhcnol 1y benzyl alcohol F‘r 7 C \x) R9t€ of Phenol Alkyliticn with n—Chlorobenzyl Alcohol Tire in minutes K1. of water distilled 4 0.5 8 1.0 17 2.2 22 2.9 51 4.1 57 5.0 45 6.1 60 8.1 66 8.8 71 9.5 77 10.0 :7 10.5 Q7 10.8 ml. HZO 8.0- 7.0’ reaction half time = 4 minut s 6.0 " 9 7 e 5.0' 4.0. 5.0. O . 0 COO b 1.0 b o C) O I l l l I j l 10 20 50 1+0 50 60 70 Time in Minutes Fig.4. ilkylation of phenol by p-chlorobenzyl alcohol 2C5 Rate of Phenol Alkylation with o-Chlorohcnzyl Alcohol in minutes 4 {9. \J 1111 . of 1"ater dietilled 0.4 \N h.) l-‘ H (I) TU \34 \1\]'J\\n4> \N (D -$ 01 CD \9 14 k F4 O \O 7.0 5.0 2.0 _ reaction half time = 150 minutes b (D (3 G 0 O 10 go 50 40 90 éo 7o 80 go 100 110 12o 130 140 150 Time in Minutes Fi?.5. Alkvlation of phenol by o-chlorobenzyl alcohol 2 ("1 7 Rate of Phenol Alkylation with 2,4-Dichlorobcnzyl Alcohol Time in minutes K1. of water dirtilled 8 0.6 17 1.1 54 1.8 5,3 2.5 77 2.9 107 5.5 135 4.0 19; 4.4 82 5.1 go 5.8 5; 0.0 555 94 593 8.8 434 9.5 1.1. age 7.0 - reaction __, ___ half time 0.0 _ = 254 minutes 5.0 '- 4.0 ' 5.0 '- 2.0 '- 0 O 1.0 ' I l l l l 0' 160 200 500 400 500 Time in Minutes Fi?.6. Alkvlation of phenol by 2,4-dichlorobenzyl alcohol APPENDIX II INFQXRED SPECTRA m 0 \0 210 - q.- _£8_< 3.88 9.020 -0 it“ G.“ «D‘- an“ 211 .3 “g! N. o .9182 x _ NcwnOLo EQA. 212 .3 m. m. 6:00;... .mucmnoLoZBQ - «KN 0.. m .m h 7) 2 .3 0.. .0£OU. < 3.51.03 79.2.0: Aw o. m .o + 214 3:02 < .mucmn .301“;an AW 215 Q. m. 1. 705024. _>NCmQOE0Lm -m at! m~ N v- L0£+w .xcmfinw .xucmm 0“ "'0 217 ‘— «N GD .95: m .chmm 1n? «bk "0 «in 41" a.” .3 m. N. m_ocm£m1ucmfl «.1 c - mu£+0L3 _\,Cm£a Ho q.” .0r_m£n:>NC0muo .v. m. N. : o. m _ m I: q qb ab D .4» «m 219 :3 2:. «pm 220 .3 m.- .Ocmr’A— %Ncm u. QCO£+0L3 _%Cmfl£\&\0 We 221 *— 4.2 N. ‘D .ocmfifxrimcc 1? -_%Ntmm IN 2 M. on ,. 1.0 ”q- d m. ‘- N. : acmgefteéixucmmhw 20 05.583 38:9 1D 1 “*- (in 7. ,/ J «N _ocw£}x£+mrc - N - Xucmm - V g ”a, «an '- mo up” _ocm:BE+oE-@ -lucmm - N 9. «a: um r. (I0 "In «t .ocmcfxfmrc- QuoEOLL -flsTANCmm .. N “c db“ "'5 f‘f‘ L74. up“ - up" - _ocm£m_xr+wrc-m - .xucmm 1 V "no 1+ 0,, W, ~»+ db“ ”—7 ”A .v. _ocmr_n_ trimgum. u aucwm u N up“, up” [grin LfiLCJ «PO, .8052 _>r+08_t-m.u-1~cum -0 16¢ 1. 1. I- ”gas—"=3" b b t b q 1 ¢ 1 x. v ,_.. A .. . «Q A ._. m ' fl. 1. _ r. .p. .5. . . w J H. H. , r i .. l. 1 .. ,,.. .. 9.. .._ » a . 41k 2.. ’1. 1 V., x .1 i , v-1. 229 m. u._ .9193 32.1056 um .N .. .xucmm mcmurszD 3:913. ® we m. .m a h .1.” «5m .v. “N .ocmgfifagthd -lucmm Lo. 90 ”In 1.“- 1b") .V~ m4 N~ : _ocm£n_ 3:58 6-0.4» - .chwm - N o. o. ‘m p 1b 1. p p 1’“ qu- 4b r~ '7 r I, I. L. J; "F .0Cm..£&Lx£+MCc_Uu® sq 1_%NC0m1N Ho ocflimcD 3:913. o. o to “h 1? q 0&9 «9v on 7.) 7. ./ A. .4 6.. «p9 _0Cmr_,\l__ $.11.er m .0: O .m .._%.NCOm I N 9 no» a: ,2 0m ’3 7 _. /’ [1— .v. var-Q .ocmrm_>r+m8_flu® .m. .. .chmm 1N we mcmu£+0c3 130:9 O. -~oa ... -»t~ 0m «e «in dim Z .9291_>c+mE_U-m.Nn_x~cm® - V 0. «no, «mm b $7.2, / . «Q N- db 1. .ocwcfxmeith115.0an 20 9305.023 16665. 2 no, .., .1» m 1D“ ‘\ up” ‘ 1P0 17‘ 2"? I, / 11. di- qr- qt- '5 -M.’ I __ 1.; _ 1h IF- «1- qb e. 1P4 2‘41 u: Iii-m \ Ns ‘ are 0‘ «PF 1*“) 2’44 «h d! dl c4! 1| all "F‘ «I J: «1 d! TX) \fl H ) .. v.2 +13 de \ d»: 0‘ 40> mm c0“) dbfi com ‘1? d» 1—1; f\ r‘ L... —_._ .11- *1 «r- u» .1» db >2 7). #7. a... db qr- qr- .1 r7, .r C 1b- db- .0— dr- d! fl. *1 .14 z z p. C), .3 O. '9’ dbo .0 I... ..,p 10; «wt 1'“) l4- l3 l2 IO '.-_ 64 .3 699.3 A1~C®DOcoEBD -Vdv 1v. CI qh q .ocmsh A.xNCMACLo.£o._QI-V.NVI¢. .40 metrimCD 1502.5. ‘b‘ ‘ 0.. "on "an ... 00 q. um «Ii-N ,ocmcm 2x21103522 I; IN «#0 up” h no .1 "V .or.®£AOLo_£U-¢I_xNCwm 1N .3 cum cud- 3 .0cmfi._3t-v-;ucmm -N o. m m vb Inn 0* db 1P db d d MCCOEL ;£+w XVI “xNCUm IN #0 octfwkb 125:9 «. : o. a m N 0 «pm r‘ _r' VN .5 ‘1 01111113va LIBRARY 3 1291031961“