Friedel-Crafts Alkylation of Aromatic Compounds with Phosphorus Estersla B

ALKYLATION OF AROMATIC COMPOUNDS WITH PHOSPHORUS ESTERS 1339

Analyse der Verbindungen: Die Einlagerungsverbin- Kalium-titandisulfid: Kalium 21,0%, Titan 33, 8%, dungen wurden vorsichtig mit HN03 aufgeschlossen. Schwefel 44,2%, Summe: 99,0%. Zusammen- Das im Fall der Wolframverbindungen dabei ausfal- setzung: Ko^TiSj,^ . lende W03 wurde alkalisch gelöst. Wolfram und Mo- Reaktionsprodukt von WS* mit Li-naphthalid (Uber- lybdän wurden als Oxinat, Titan als TiOa und Schwe- schuß) : Lithium 9,8%, Wolfram 66,3%, Schwefel fel als BaS04 bestimmt. Die Bestimmung der Alkali- 23,5%, Summe: 99,6%. Verhältnis 1 W : 2 Lili95S. metalle erfolgte flammenphotometrisch. Reaktionsprodukt WS2 mit Na-naphthalid (Überschuß) : Kalium-wolf ramdisulfid: Präparat I: Kalium 8,3%, Natrium 26,9%, Wolfram 53,1%, Schwefel 18,8%, Wolfram 66,9%, Schwefel 23,7%, Summe: 98,9%. Summe: 98,8%. Verhältnis: 1 W : 2 Na2,0S. Zusammensetzung: K0.59WS2)0 . Präparat II: Kalium 8,1%, Wolfram 67,1%, Schwe- fel 23,4%, Summe: 98,6%. Zusammensetzung:

K0.57WS2,O . Kalium-molybdändisulfid: Kalium 10,86%, Molybdän Wir danken der Deutschen Forschungsgemeinschaft 53,6%, Sdiwefel 35,5%, Summe: 99,9%. Zusam- und dem Fonds der Chemie für die Unterstützung die- mensetzung: K0!49MoS1?98 . ser Arbeit.

1 Auszug aus der Dissertation E. BAYER. Tübingen 1970. 5 T. E. HOVEN-ESCH U. J. SMID, J. Amer. chem. Soc. 87, 669 2 W. RÜDORFF, Chimia [Zürich] 19,489 [1965], [1965]. 3 H. M. SICK, Dissertation Tübingen 1959. 6 W. BILTZ, P. EHRLICH U. M. MEISEL, Z. anorg. allg. Chem. 4 C. STEIN. J. POULENARD, L. BONNETAIN U. J. GOLE, C. R. 234,97 [1934], hebd. Seances Acad. Sei. 260, 4503 [1965].

Friedel-Crafts Alkylation of Aromatic Compounds with Phosphorus Estersla b

G. SOSNOVSKY and M. W. SHENDE

Department of Chemistry, University of Wisconsin-Milwaukee, Milwaukee, Wisconsin 53201

(Z. Naturforsch. 27 b, 1339—1348 11972] ; received June 26/August 22, 1972)

Friedel-Crafts, Electrophilic Alkylation, Phosphorus Esters, Alkylbenzenes

The alkylations of aromatic compounds with trialkyl phosphites (1), dialkyl phosphites (2), and trialkyl phosphates (3) in the presence of aluminum chloride were studied involving several raction variables, such as time, ratio of reactants, nature of catalyst and solvent, and combinations thereof. Extensive disproportionation and isomerization were observed in the reaction with mono- substituted alkylbenzenes under heterogeneous reaction conditions obtained by the use of an excess of aromatic substrates. A combination of aluminum chloride —nitromethane complex and dichloromethane as solvent was used to eliminate these undesirable effects and to give homogeneous and practically non-isomerizing conditions. The scope of the reaction was studied with a number of aromatic substrates, and their relative reactivities were compared to that of benzene in competitive isopropylations with triisopropyl phosphite. The relative rates and isomer distributions showed low substrate and low positional selectivities and poor agreement with B r o w n's selectivity relationship. The substrate selectivity was somewhat higher and the positional selectivities were somewhat lower than those obtained in competitive isopropylation reactions with other isopropylating agents. The selectivity factor, SF, and partial rate factors were calculated. An electrophilic alkylation mechanism is proposed on the basis of (1) the relative rates of isopropylation, (2) the isomer distribution of dialkylated aromatics, and (3) the necessity of a strong Lewis acid in these reactions.

Friedel-Crafts alkylations of organic com- an important advance in the alkylation reaction. pounds with alkyl halides, alkenes, alcohols and a However, the alkylation reaction with alkyl esters of variety of alkylating agents are relatively old reac- inorganic acids has not been investigated fundamen- tions on the chemist's list of synthetic techniques. tally or exploited practically to the same extent as The application of alkyl esters of organic and mine- the acid-catalyzed alkylations with either olefins or ral acids to the Friedel-Crafts synthesis 2 was alkyl halides. The reason for this neglect might be the easy availability of alkyl halides and other alky- Requests for reprints should be sent to Prof. G. SOSNOVSKY, lating agents as compared to alkyl esters of in- Department of Chemistry, University of Wisconsin-Mil- waukee, Milwaukee, Wisconsin 53201, U.S.A. organic acids as starting materials.

Dieses Werk wurde im Jahr 2013 vom Verlag Zeitschrift für Naturforschung This work has been digitalized and published in 2013 by Verlag Zeitschrift in Zusammenarbeit mit der Max-Planck-Gesellschaft zur Förderung der für Naturforschung in cooperation with the Max Planck Society for the Wissenschaften e.V. digitalisiert und unter folgender Lizenz veröffentlicht: Advancement of Science under a Creative Commons Attribution Creative Commons Namensnennung 4.0 Lizenz. 4.0 International License. 1340 G. SOSNOWSKY AND M. W. SHENDE

The literature contains a survey of alkylation alkyl phosphites and trialkyl phosphates with aro- reactions with alkyl esters of inorganic acids such as matic substrates in the presence of aluminum chlo- alkyl sulfates2, sulfites2, chlorosulfites2, sulfonates2, ride are extremely exothermic. Therefore, it was chlorosulfonates2, p-toluene sulfonates2, alkyl o- necessary to add these esters to the mixture of the silicates2, alkyl carbonates2, alkyl borates2, and aromatic substrate — aluminum chloride at 5° ±2°, hypochlorites 2. However, the analogous reactions in- followed by stirring the reaction mixture at room volving the use of phosphorus esters, such as tri- temperature. alkyl phosphites (1), dialkyl phosphites (2), and A number of variables were examined in order tri alkyl phosphates (3), have received relatively to determine their effect on the reaction. It was little attention. Thus, n-'butyl phosphate3 was used found that the combined yield of mono- and diethyl to alkylate benzene in the presence of boron benzenes was very dependent on the ratio of the fluoride, and triethyl phosphate 4' 5 triisopropyl phos- ester to aluminum chloride. At least a 1:1 molar phate 4 and tributyl phosphate4 were used for the ratio of alkyl group in the alkylating ester to alu- alkylation of benzene in the presence of aluminum minum chloride is required for optimum yields chloride. These experiments were limited to benzene (Table I). In practice, a slight excess of aluminium as the substrate, and the reaction products were not chloride was used in order to compensate for losses critically examined. Definitive studies in this area during handling. Thus, the molar ratios of alumi- have become possible only since the advent of gas a chromatography. Table I. Ethylation of Benzene with One Mole Triethyl Phosphate. Effect of Variation in the molar Ratio of Alumi- Recently, a systematic attempt has been made to num Chloride to Triethyl phosphate. investigate the use of phosphorus esters in alkyla- Products13 6a b tions ' . At present a variety of phosphate and AlCl3/(C2H50)3P0 Monoethyl Diethyl phosphite esters are commercially available at a mole ratio Benzene Benzene low cost. These esters are liquids of low volatility Yield % Yield %

and, therefore, make very suitable starting materials. 0.5 11 0.9C The objectives of the present work were to study 1 16 2.6C C the scope and mechanism of alkylation reactions of 1.5 18 1.4 3.5 26 44 D various aromatic hydrocarbons with trialkyl phos- 4 26 41D phites (1), dialkyl phosphites (2), and trialkyl phosphates (3). ;l Benzene to triethyl phosphate molar ration 20:1. b Reaction time 2.5 hr. (Reaction temperature, see pp. 1340 and 1346). c As the product was formed in traces, it could not be col- RCK. Rn RO\ RO— P: pn>p-°H RO—P=Ö: lected or positively identified by glpc. It shows the same reten- R0/ RO RO^ tion time as /n-diethyl benzene d m-Diethyl benzene. 1 2 3 num chloride to alkylating ester were 2.5 to 1 and R = CH3, j-C3H- R = CH3, COH, . R=C2H5 i-C3H7 3.5 to 1 in the case of diester and triester, respectively. Results and Discussion Ferric chloride, cupric chloride, stannic chloride, zinc chloride, sulfuric acid (98%) and phosphoric The conditions under which phosphorus esters acid (85%) were examined as catalysts, but with are employed in alkylations are different from those the exception of ferric chloride, no detectable reac- used in alkylations with alkyl halides 2. One reason tion was observed, even at elevated temperatures is that the inorganic acid, formed as a by-product, and on prolonged reaction times, as evidenced by is non-volatile and cannot escape from the reaction glpc analysis. Although a reaction was observed in mixture in the manner of hydrogen halides. The the presence of ferric chloride, as evidenced by the resultant increase in the acid concentration in the evolution of hydrogen chloride, it was not possible reaction medium as the reaction proceeds probably to separate the organic phase from the inorganic causes secondary reactions, thus diminishing the by-products. A number of homologs of benzene were yield of the primary products. alkylated with triethyl phosphate and triisopropyl At an early stage in this investigation it was phosphite. The results (Table II) indicate that found that the reactions of trialkyl phosphites, di- cumene and ethyl benzene undergo dealkylation to ALKYLATION OF AROMATIC COMPOUNDS WITH PHOSPHORUS ESTERS 1341

Table II. Alkylation of Aromatic Compounds with Triethyl Phosphate and Triisopropyl Phosphite in the Presence of Alumi- num Chloride in Excess Substrate

Aromatic Alkylating Monoalkylated % Isomer Distribu- Dialkylated % Isomer Distribution Substrate Ester Products tion of Mono- Products of Dialkylated (Mole) (Mole) (Yield %) alkylated Products (Yield %) Products

Toluene3 (C2H50)3P0 Monoethyl toluenes 7 73 20 Diethyl toluene l-Methyl-3,5-diethyl (1.5) (0.1) (57) (19) benzene onlyb c d Benzene (i-C3H70)3P Cumene Diisopropyl benezenes o + p = 32 (0.75) (0.05) (60) (12) " m = 68 e Toluene a (i-C3H70)3P Cymenes trace 71 29 Diisopropyl l-Methyl-3,5-diisopropyl (0.75) (0.05) (44) toluene benzene (traces of (10) 1,2,4-isomerf) Cumene (i-C3H70)3P Diisopropyl benzenes 0 65 35 Triisopropyl benzene 1,3,5-Triiso-propyl (0.75) (0.05) (114) (27) benzene only Ethyl (i-C3H70)3P h benzene c (0.75) (0.05)

a Reaction time 4.5 hr. b Compound identified by NMR. c Reaction time 3.5 hr. d Contains trace quantities of o-isomer. c Trace quantities of o-isomer formed but could not be separated by glpc. f IR spectrum indicates 1,2,4-substitution. S Sub- strate shows dealkylation to benzene; 6.17 g benzene isolated by fractionation. h Extensive dealkylation of the substrate; among products identified on column (A) are benzene, cumene, m-diisopropyl benzene, p-diisopropyl benzene, and on column (B) m-diethyl benzene, p-diethyl benzene. In addition products corresponding to four major peaks were not identified. § Mole ratio of aluminum chloride to alkylating ester, see p. 1340. Reaction temperature, see p. 1346.

benzene. In the isopropylation of ethyl benzene a The dealkylations observed with cumene and number of aromatic hydrocarbons, such as benzene, ethyl benzene, as well as subsequent disproportiona- cumene, m-diethyl benzene, p-diethyl benzene, m- tion with ethyl benzene yielding a number of aro- diisopropyl benzene, p-diisopropyl benzene, and matic hydrocarbons is due to the use of aluminum substantial quantities of dialkylated products are chloride in amounts far in excess of a catalytic formed. quantity7a. Variation in the molar ratio of alkyl These results are in agreement with the already aromatic to alkylating ester was examined in order known fact that, depending on the alkylating agent, to find conditions which minimize the formation of the nature and amount of catalyst, and the reaction higher alkylated products. The use of a large excess conditions, dealkylations, intramolecular and inter- of aromatic hydrocarbon has been recommended 7b molecular isomerizations (disproportionation or in order to reduce polysubstitution. In the ethylation transalkylation) occur in Friedel-Crafts type of toluene with triethyl phosphate, the yield of di- alkylations of alkyl aromatics7a' 8. These reactions ethyl toluene could be reduced to trace quantities by also occur when alkyl aromatics are treated with a increasing the toluene to triethyl phosphate molar Lewis acid, e. g.. metal halide7a. In these reac- ratio to 45 : 1, or to 15 moles : 1 mole of substrate tions 7a' 9, the methyl group is much more resistant to alkyl group in the ester. The overall yield of to migration than higher alkyl groups. Toluene did monoethyl toluenes decreased beyond a 30 : 1 molar not dealkylate under these conditions. ratio of toluene to triethyl phosphate, presumably

Table III. Ethylation of Toluene with Triethyl Phosphate in the Presence of Aluminum Chloride. Effect of Variation in Molar Ratio of Substrate to Alkylating Ester

PhMe/(C2H50)3P0 Monoethyl % Isomer Distribution Diethyl Toluene % Isomer Distribution Mole Ratio Toluenes Yrield % of Monoethyl Toluenes Yield % of Diethyl Toluenes

0 m V 15: la 57 7 73 20 19 l-Methyl-3,5-diethyl benzene onlyb 30: lc 83 4 80 16 traces'1 — 45: le 74 6 77 18 tracesd —

§ Mole ratio of aluminum chloride to alkylating ester, see p. 1340. Reaction temperature, see p. 1346. a Reaction time 4.5 hr. t> Compound identified by NMR. c Reaction time 3 hr. d Product could not be collected over glpc and analyzed. 1342 G. SOSNOWSKY AND M. W. SHENDE

Table IV. Alkylation of Aromatic Compounds with Phosphorus Esters in the Presence of Aluminum Chloride. Effect of Varia- tion in Reaction Time

Aromatic Alkylating Reaction Monoalkylated % Isomer Distri- Dialkylated % Isomer Distribution Substrate Ester Time Products bution of Mono- Products of Dialkylated Products (Mole) (Mole) hr (Yield %) alkylated Products (Yield %)

o m p Cumene (i-C3H70)3P 1 Diisopropyl 0 65 35 Triisopropyl 1,3,5-Triisopropyl (0.75) (0.05) benzenes benzene benzene only (114)a (27) Cumene (i-C3H70)3P 3.5 Diisopropyl 0 65 35 Triisopropyl 1,3.5-Triisopropyl (0.75) (0.05) benzenes benzene benzene only (114)a (27) Cumene (*-C3H70)3P 24 Diisopropyl 0 65 35 Triisopropyl 1,3,5,-Triisopropyl (0.75) (0.05) benzenes benezene benzene only (94) (52) Tol uene (i-C3H70)2P(0)H 0.5 Cymenes trace'3 75 25 Diisopropyl l-Methyl-3,5- (2.0) (0.05) (67) toluene (trace0) diisopropyl benezene 13 Toluene (i-C3H70)3P 4.5 Cvmencs trace 71 29 Diisopropyl l-Methyl-3,5-diisoprop\ (0.75) (0.05) (43) toluene benzene (traces (10) of l,2,4-isomerd) Toluene (C2H50)2P(0)H 0.5 Monoethyl 27 56 17 trace e (2.0) (0.05) toluenes (61) e Toluene (C2H50)3P0 3 Monoethyl 6 77 18 trace (1.5) (0.033) toluenes (74) Toluene (CH30)2P(0)H 0.5 Xylenes 51 21 28 tracee (2.0) (0.05) (56) Toluene (CH30)2P(0)H 24 Xvlenes 26 62 12 trace e (2.0) (0.05) (57)

5 Mole ratio of aluminum chloride to alkylating ester, see p. 1340. Reaction temperature, see p. 1346. a Substrate shows dealkylation to benzene, hence, more than theoretical amount of product formed. b The o-isomer, formed in trace quantities, could not be separated by glpc. c Yield could not be calculated. d The IR spectrum indicates 1,2,4-substitution. e Yield could not be calculated. The product could not be collected or identified by glpc. because of additional losses during the isolation mers in the aluminum chloride catalyzed isomeriza- procedure (Table III and Experimental pp. 1346. tions of dialkyl benzenes 11-13a-d and in the alumi- Finally, the reaction time was examined in order num chloride catalyzed alkylations of toluene with to determine its effect on the orientation and yield ethylene14 and propylene14, and of cumene with of the monoalkylated and dialkylated products propylene 14. The evidence of whether the high m- (Table IV). Short reaction times could not be exa- isomer contents are due to direct substitution in- mined conveniently since the exothermic reaction volving a reaction of low selectivity 15a_d or to a precluded a rapid addition of the alkylating ester to concurrent or consecutive secondary isomerization the substrate — aluminum chloride mixture. The al- of the initially formed o- and p-dialkyl benzenes is kylations reported so far in Tables I, II, III, and not conclusive 7c'10'16'17. IV used an excess of the aromatic substrate as the In order to eliminate side reactions such as de- reactant and solvent, and were, therefore, particu- alkylation and to minimize polysubstitution, we found larly sensitive to isomerizations 10. The orientation that a solvent system composed of nitromethane of products in these reactions indicates the forma- and dichloromethane is advantageous for the alky- tion of considerable amounts of the m-dialkyl ben- lation reactions with trialkyl phosphites, dialkyl zenes and the 1,3,5-isomer. Due to isomerizing con- phosphites and trialkyl phosphates. Nitromethane is ditions, considerable disproportionation and trans- used to dissolve the aluminum chloride, and di- alkylation also occur to give thermodynamically chloromethane is used as the solvent for the sub- controlled equilibrium mixtures of xylenes, diethyl strate and the alkylating ester, thus giving homo- benzenes, monoethyl toluenes, cymenes and diiso- geneous conditions. The molar ratios of aromatic propvl benzenes. In the literature are found nume- substrate to aluminum chloride to alkylating ester rous examples of the formation of such equilibrium were the same as those used in the previous experi- mixtures containing high proportions of meta-iso- ments, viz., 45 : 3.5 : 1 and 30 : 2.5 : 1 in the case ALKYLATION OF AROMATIC COMPOUNDS WITH PHOSPHORUS ESTERS 1343

Table V. Alkylation of Aromatic Compounds with Phosphorus Esters a in Dichloromethane in the Presence of Aluminum Chloride —Nitromethane. Effect of Variation in Reaction Time

% Isomer Aromatic Substrate Alkylating Ester Reaction Monoalkvlated Products Distribution of Mono- (Mole) (Mole) Time hr (Yield %) alkylated Products15

o m P Toluene (/-C3H70)3P 0.5 Cymenes 43 25 32 (1.125) (0.025) (52) Toluene (i-C3H70)3P '•> Cymenes c 43 26 31 (1.125) (0.025) (69) Toluene (i-C3H70)3P 24 Cymenes 43 26 31 (1.125) (0.025) (83) 1 Ethyl benezene (i-C3H70)3P 0.5 Ethyl isopropvl benzenes' Not resolved (0.75) (0.016) (46) 1 Ethyl benzene (t-C3H70)3P 24 Ethvl isopropvl benzenes' Not resolved (0.75) (0.016) (87)" Ethvl benzene (C2H50)2P(0)H 4 Diethyl benezenes 42 26 32 (0.75) (0.025) (17) Ethyl benzene (C2H50)2P(0)H 24 Diethvl benzenes 38 29 33 (0.75) (0.025) (50) Cumene (i-C3H70)3P 18 Diisopropvl benzenes o + p = 62 (1.125) (0.025) (84) m = 38 Toluene (C2H50)3P0 20 Monoethyl toluenes 51 30 19 (0.562) (0.0125) (34) Toluene (C2H50)2P(0)H 20 Monoethyl toluenes 54 28 18 (0.75) (0.025) (38)

§ Ratios of aluminum chloride, see p. 1342. Reaction temperature, see p. 1346. a No reaction obtained between ethyl benzene and trimethyl phosphite, and toluene and trimethyl phosphite and dimethyl hydrogen phosphite, respectively, b Only trace quantities of dialkylated products were formed in these reactions. c The isomer distribution of cymenes was not resolved by glpc. The IR spectroscopic examination for variation in isomer distribution indicated no change from the isomer distribu tions of cymenes formed in experiments with reaction times of 0.5 and 24 hr, respectively. d Ethyl isopropyl benzenes were isomerically similar by IR spectroscopy. of triesters and diesters, respectively. The following is produced in alkylations with isopropyl bromide 16 weight ratios were maintained with different and propylene16 in nitromethane solvent in the reactants and their solvents: alkyl aromatic to di- presence of aluminum chloride. chloromethane, 1:1; aluminum chloride to nitro- The effect of reaction time in the alkylation methane, 1 : 3.5; and the alkylating ester to di- reactions with these esters with aluminum chloride — chloromethane, 1 : 3.5. nitromethane and dichloromethane is shown in The use of this solvent system prevented the de- Table V. Enhanced yields of alkylated products alkylation of alkyl aromatics to benzene, and the were obtained with increased reaction time. The formation of dialkylated products was reduced to a variation in reaction time had no effect on the iso- minimum. The scope of the alkylation reaction using meric composition of the products formed phosphorus esters could thus be extended to a in ethylation and isopropylation reactions. The simi- number of alkyl benzenes (Table V). In the di- larity in the isomer distributions of these dialkyl chloromethane — nitromethane solvent system, the benzenes was determined by glpc and by infrared isopropyl esters are the most reactive, ethyl esters spectroscopy using the out-of-plane carbon — hydro- have intermediate reactivity and the methyl esters gen deformation absorption bands at 13.0 — 13.6// fail to react. This result is in agreement with the for the o, at 12.3-13.3/* and 14.0- 14.8 ju for observation by SCHMERLING 18 that in the aluminum the m, and at 11.6 — 12.5 JLI for the p-isomers. The chloride — nitromethane system the secondary alkyl mixture of isomers of dialkylated benzenes formed chlorides are more reactive than the primary iso- in each experiment was collected over glpc and was mers. The isopropylation of toluene with triisopro- found on infrared analysis to be spectroscopicallv pyl phosphite using aluminum chloride — nitro- identical with a mixture of isomers formed in a methane and dichloromethane yields cymenes con- similar experiment with a different reaction time. taining a higher percentage of the m-isomer 19 than Subjecting a mixture of known composition of o-, 1344 G. SOSNOWSKY AND M. W. SHENDE m- and p-diethyl benzenes to simulated reaction lation of toluene with triisopropyl phosphite, for conditions over a period of 4 hrs did not change the comparison with similar values reported previously composition of the mixture as regards percentages for other isopropylations (Table VIII). From a of o-, m- and p-isomers. Table VI. Competitive Isopropylations a of Toluene —Benzene These results indicate that the aluminum chlo- in Dichloromethane with Triisopropyl Phosphite and Alumi- ride — nitromethane — dichloromethane solvent sys- num Chloride —Nitromethane. First Order Dependence of tem provides practically non-isomerizing reaction the Isopropylation in Aromatics §. conditions. Additional information on the activity Toluene Cymenes Cumene Observed Relative Rates of these esters was obtained from studies of com- Benzene Yield % Yield % Relative According to petitive isopropylations of toluene — benzene and re- Mole Rates First Order Ratio (ki/kn) Dependence in lated alkyl benzenes — benzene in the nitromethane — Aromatics dichloromethane solvent system. The molar ratio of combined aromatic substrate to aluminum chloride 1 :: 1 53.5 18.15 2.95 (2.99) c(L) = 2.99 1 :: lb 73.4 24.1 3.04 to triisopropyl phosphite was 45 : 3.5 : 1; thus a 1 :: 4 28.2 34.4 0.82 (0.82 (j) - 3.28 constant excess of the combined aromatic substrate 2: : 3 57.5 23.4 2.46 (2.46) (4) = 3.69 3: : 2 75.4 14.0 5.36 (5.36) (J) = 3.57 was maintained. The relative reactivities A;a,.omaticf 4: : 1 79.6 5.8 13.7 (13.7) (1) =3 3.44 20 ^benzene were calculated from the amounts of iso- Average = 3.39 propylated aromatic and cumene formed. As only trace quantities of dialkylated products were formed § Ratios of aluminum chloride, see p. 1342. Reaction temperature, see p. 1346. a Reaction time 2.25 hrs. *> Reaction time in these competitive alkylations, they did not affect 3 hrs. e The value 2.99 is an average of two values obtained the calculated &aromatjc/A:benzene results seriously 16. with 1:1 molar ratio of toluene to benzene.

The method of competitive rate determination can Table VII. Competitive Isopropylations a of Alkyl Benzenes — be used to obtain relative reactivities only if the Benzeneb in Dichloromethane with Triisopropyl Phosphite and Aluminum Chloride — Nitromethane observed relative rates are dependent on the aromatic substrates 10'16. In order to establish whether Aromatic Substrate actual competition exists between benzene and the Benzene 1.00 investigated alkyl benzenes, the relative reactivities Toluene 3.39 c in toluene — benzene mixtures were determined by Ethyl benzene 2.50d changing the concentration of toluene — benzene ?i-Propyl benzene 2.87 p-Cymene 2.54 mixtures. The results in Table VI show that the re- Chlorobenzene 0.44 lative rate remains constant and that a first order dependence on the aromatic substrate concentration 8 Ratios of aluminum chloride, see p. 1342. Reaction temperature, see p. 1346. a Reaction time 2.25 hrs. b A 1:1 molar 10 16 is indicated ' . The relative reactivities of a ratio of alkyl benzene:benzene used. c Average value from number of alkyl benzenes over that of benzene, as Table VI. d Average of two values obtained with 1:1 molar ratio of ethyl benzene : benzene. obtained by the method of competitive isopropylation, are summarized in Table VII. survey of a number of substitution reactions, BROWN The isopropylation with triisopropyl phosphite and STOCK21 selected the limiting value of the ratio shows low substrate selectivity between the two aro- (log pf)/(log m{) to be 4.04 + 0.55 for adherence matic compounds and low positional selectivity be- to the selectivity relationship. The average value tween the m and p positions of toluene. This result of 3.39 for kr/kft (Table VI) and the values of par- is plausible considering the value of 3.39 for the tial rate factors (Table VIII) indicate that the iso- relative reactivity of toluene to benzene, and the propylation reaction with triisopropyl phosphite has formation of 26% m-cymene and 31% p-cymene in a somewhat higher substrate selectivity and some- the isopropylation of toluene. The average value of what lower positional selectivities as compared to 3.39 is somewhat higher than the values of kyjk^ isopropylation reactions with other isopropylating for the isopropylation reactions involving different agents. The reaction does not show adherence to the alkylating agents 10'15d'16- 20. The values of selecti- selectivity relationship. vity factor, Sf15d' 21, and partial rate factors for o An electrophilic substitution mechanism for the (of), m (mf), and p (pf) positions were calculated isopropylation of toluene by triisopropyl phosphite from the orientation data obtained in the isopropy- is supported by the following facts: (1) the pre- ALKYLATION OF AROMATIC COMPOUNDS WITH PHOSPHORUS ESTERS 1345

Table VIII. Selectivity Factors and Partial Rate Factors for Friedel-Crafts Isopropylations of Toluene with Isopropyl Bromide, Propylene and Triisopropyl Phosphite.

Alkylating Catalyst Temp. % Isomer Distri- Partial Rate Selectivity log pj Reference Agent [°C] bution of Cymenes Factors Factor log »if

Of rri{ p{ Isopropyl Gallium bromide 25 26.2 26.6 47.2 1.45 1.47 5.20 0.548 4.28 15a bromide Isopropyl bromide Gallium bromide 25 26.2 26.6 47.2 1.52 1.41 5.05 0.554 4.72 24 a Propylene BF3-Et20 or 5 37.5 29.8 32.7 0.342 A1C13—CH3N02 2.37 1.80 4.27b 2.46 25 Propylene BF3—Et20 or 65 37.6 27.5 34.9 0.404 A1C13-CH3N02 Triisopropyl A1C13-CH3N02, 5 ± 2C 43.1 26.1 30.8 4.38 2.65 6.26d 0.373 e 1.88 present phosphite solvent CH2C12 work

a Partial rate factors calculated from relative rate data &T/A:B = 1.82 and isomer distribution of cymenes of ref. ,5a. b Calcu- lated from isomer distributions of cymenes of ref. 28 by interpolation of data at 5° and 65® to 40°. c Addition of alkylating ester to substrate + A1C13—CH3N02 at this temp, followed by stirring the crude reaction mixture at room temp. d Partial

rate factors calculated from relative rate data &T/&B=3.39 with the use of the following eqs.: Of——? x re at*ve rate x ^ 100 m x relative rate x 3 % p x relative rate x 6 and pt = e Selectivity factor calculated with the use of the eq. Sf = mf= 100 100 log (2 x % p/% m). dominant o,p-orientation with toluene; (2) the re- alkyl group and (3) it justifies the need for a 1 : 1 latively large amount of m-isomer indicating that molar ratio of the alkyl group in the ester to alu- the attacking species possesses high activity 15b and, minum chloride for optimum yields in the alkyla- therefore, low selectivity; (3) the necessity of a tion reactions. For a general reaction of the type strong Lewis acid in the reaction and the ability (RO)3P + 3 ArH + A1C13 ->• 3 RAr + A1P03 + 3 HCl of the reaction to proceed at low temperatures; the following reaction scheme is in overall agree- (4) the relative rates (^toiueneMWene) being in the ment with our observations. same range as the values reported for those reactions, e. g., alkylations, entailing electrophiles of A1C13 comparatively high activity22. The dialkyl phos- A1C13 + P (OR)3 P(OR)2OR—* phites (2) show little or none of the nucleophilic ArR + HCl + P(0R)20A1C12 reactivity of the trialkyl esters (1) 23. Furthermore, A1C13 they are only weakly acidic. It is now well established that they exist almost exclusively in the phosphonate P (OR) 20A1C12 + A1C13 -> P (OR) (OR) 0A1C12 23 form (4). ArR + HCl + P(OR) (0A1C12)2

ROv. ROx y O AICI3 >P-OH > Pf RO' ROx x H P(OR) (0A1C12)2 + A1C13 P(ÖR) (0A1C12)2 ArR + HCl + P(0A1C12)3 Phosphite form Phosphonate form 2 4 P (0A1C12) 3 A1P03 + 2 A1C13

Complex formation with a strong Lewis acid is In the alkylations with triethyl phosphate and tri- possible in (1) at the phosphorus and oxygen atoms isopropyl phosphite, the presence of the phosphate of the alkoxy group, and in (3) and (4) at the ion in the aqueous layer resulting from work-up of oxygen atoms of the alkoxy and the phosphoryl the reaction mixture was confirmed by qualitative groups. Complexation of the catalyst at the oxygen analysis. In the dichloromethane — nitromethane sol- of the alkoxy group is suggestive in all three cases, vent combination no free aluminum chloride capable as (1) the oxygen is the negative end of the phos- of direct coordination with the alkylating ester is phorus — oxygen dipole, (2) complexing at the oxy- present since aluminum chloride forms a stable com- gen atom enhances the electrophilic character of the plex 16 with nitromethane, A1C13CH3N02. Any 1346 G. SOSNOWSKY AND M. W. SHENDE interaction with the alkylating ester must, therefore, following overall conditions were maintained: injector involve the aluminum chloride — nitromethane com- temperature, 150°; detector temperature, 220°; bridge current, 150 ma; helium pressure, 50 psi. The fol- plex. Owing to the competition of complexing of lowing columns were used: Column (A) — 6 ft by aluminum chloride with nitromethane and the alky- 0.25 in., Aluminum column, 20% Ucon 50 HB 280 on lating ester, the alkylation reactions performed in 60/80 mesh acid washed Chromosorb W; Column (B) this solvent system show a marked slowness. The — 21 ft by 0.25 in., Aluminum column, 9.5% Bentone 34 + 4.5% Dow 550 on 60/80 mesh acid washed, DCMS- aluminum chloride — nitromethane complex pre- treated Chromosorb P. sumably fails to polarize the carbon — oxygen bonds Infrared spectra were obtained with a Perkin-Elmer in the methoxy groups of trimethyl phosphite and 137 spectrophotometer, and nmr spectra with a Varian dimethyl hydrogen phosphite, leading to the un- T-60 spectrometer on 10% (v/v) samples in carbon reactivity of these esters in the solvent system. tetrachloride using an internal TMS standard. Mole- cular weights were determined isopiestically on a It is also possible to consider isopropylation Hitachi Perkin-Elmer Model 115 Molecular Weight with aluminum chloride — nitromethane in dichloro- apparatus. methane solvent as a concerted nucleophilic dis- Alkylations with Phosphorus Esters in the Presence placement reaction by the aromatic substrate of the of Excess Aromatic as Reactant and Solvent (Tables I catalyst : nitromethane : alkylating ester complex. through IV). General Procedure A: The ester was ad- Such displacement mechanisms have been suggested ded as rapidly as the exothermic reaction permitted at 5° ±2° to a well stirred suspension of aluminum 15c d by BROWN and co-workers ' for alkylations in- chloride in the aromatic hydrocarbon maintaining the volving primary halides. As to the nature of the ef- reaction mixture at 5° ± 2°by external cooling. On com- fective alkylating species, the data of the present pletion of the addition the reaction mixture was stir- investigation do not allow final conclusion. The sug- red at room temperature. The total reaction time spe- cified for a particular reaction includes the time re- gested Lewis acid — alkylating ester complexes quired for addition of ester followed by stirring the may split into relatively free carbonium ions, or crude reaction mixture for the rest of the time. The they may remain in a partially polarized state and reaction mixture was then carefully poured onto react. The discussion of the carbonium ion character crushed ice with stirring. The organic phase was separated and the aqueous phase was extracted with ether is only justified on the basis of overall product (50 ml). The ether extract was combined with the main distribution. organic phase and the ether removed by distillation. The organic phase was then washed successively with Experimental 10% sulfuric acid, 10% sodium bicarbonate solution and water until neutral to litmus, dried (Na2S04), and concentrated by removing excess aromatic substrate Materials. The aromatic hydrocarbons were dried by distillation under nitrogen at atmospheric pressure. azeotropically and stored over anhydrous calcium The aromatic substrate, thus removed, was analyzed chloride before use. They were of high purity according by glpc in order to ascertain that the reaction pro- to glpc analysis. Nitromethane (spectrograde, Eastman ducts were not lost. The resultant concentrated solution Organic Chemicals) was distilled over anhydrous cal- containing the reaction products was then analyzed cium chloride, and dichloromethane (Aldrich Chemical by glpc. Co.) was distilled over calcium hydride before use. An- hydrous aluminum chloride (reagent grade, Allied Alkylations with Phosphorus Esters in the Presence Chemical and Mallinckrodt Chemical Works) was used of Aluminum Chloride— Nitromethane and Dichloro- without further purification. The phosphorus esters methane (Table V through VII). General Procedure B: were best commercial grade used without purification. To a cold (5°) solution of the aromatic hydrocarbon The following aromatic compounds were used as in dichloromethane was added a solution of aluminum authentic materials for identification of products: o- chloride in nitromethane in one portion. A solution and p-xylene (Eastman Kodak Co., White Label), of the ester in dichloromethane was then added to the m-xylene (spectrograde, Eastman Kodak Co., White resultant solution at 5° + 2° as rapidly as the exother- Label), o-, m-, and p-diethyl benzene (Aldrich Chemical mic reaction permitted maintaining the reaction Co.), m- and p-diisopropyl benzene (Aldrich Chemical mixture at this temperature by external cooling. On Co.), o-, m- and p-ethyl toluene (Aldrich Chemical completion of the addition the reaction mixture was Co.), 3,5-diisopropyl toluene (Alfred Bader Chemicals, stirred at room temperature as indicated in General Division of Aldrich Chemical Co.), 1,3,5-triisopropyl Procedure A. The reaction mixture was then carefully benzene (J. T. Baker Chemical Co.), and p-cvmene poured onto crushed ice with stirring. The organic (terpene-free, Eastman Organic Chemicals). phase was separated, washed successively with 10% Analytical Procedures. Glpc analyses were performed sulfuric acid, 10% sodium bicarbonate solution and on a Varian Aerograph A90P3 gas Chromatograph water until neutral to litmus, dried (Na2S04), and con- equipped with a thermal conductivity detector. The centrated by removing dichloromethane, nitromethane. ALKYLATION OF AROMATIC COMPOUNDS WITH PHOSPHORUS ESTERS 1347 and excess aromatic substrate by distillation under NMR. The following absorption wavelengths were nitrogen at atmospheric pressure. Dichloromethane, used for the identification of the pattern of substitu- nitromethane, and excess aromatic substrate, thus re- tion: moved, were analyzed by glpc in order to ascertain Tri-substituted Isomer Absorption Range that the reaction products were not lost. The resultant concentrated solution containing the reaction products 1.2.4-trisubstituted 11.7 — 12.4/* (strong) 11.0—11.5 p (medium) was then analyzed by glpc. 1.3.5-trisubstituted 11.0 — 12.0« (strong) Product Yields. The yields of various mono-, di- and 14.1 —15.0 p (medium) trialkylated aromatics were determined by glpc on column (A). Isomerization and Recovery Studies. A known mix- In competitive isopropylations (Tables VI and VII) ture of isomeric diethyl benzenes (4.0 g, 19.6% o, 31.0% the relative response ratios of the components present m, and 49.4% p) was prepared. To 3.70 g of the above in the concentrated solution containing the reaction mixture, dissolved in ethyl benzene (79.62 g, 0.75 mole) products to that of an added internal standard were and dichloromethane (65 ml), was added a solution of determined27. Benzene was used as an internal stan- aluminum chloride (8.33 g, 0.0625 mole) in nitro- dard. One of the pure components, the isopropylated methane (27 ml) at 5° in one portion. On completion aromatic, was previously collected over glpc for use of the addition the mixture was stirred at room tem- in the mixture of known composition. Authentic samples perature for 4 hrs. After the work-up as described in of other components were used. The yields of alkylated General Procedure B, 128.97 g of dry organic phase aromatics are based on the moles of alkylating ester was isolated. A portion of the organic phase, 35.21 g, used in each reaction, assuming that every alkyl group was concentrated by removing solvent and excess ethyl can react. benzene by distillation at atmopheric pressure under a current of nitrogen yielding 5.80 g of a solution con- Product Identification and Isomer Distributions. taining isomeric diethyl benzenes. The analysis by The concentrated solution containing the reaction pro- glpc on column (A) indicated 3.54 g (96.0%) of di- ducts obtained in procedure A or B was analyzed by ethyl benzenes. The mixture of isomeric diethyl ben- glpc on column (A). The peaks corresponding to zenes present in the concentrated solution was col- mono-, di-, and tri-substituted alkyl aromatics were lected from column (A). The recovered diethyl ben- always well separated and products corresponding to zene mixture was compared to the starting material by these peaks were collected from the same column. The IR spectroscopy and was found to be isomerically mono-substituted aromatic hydrocarbons were identi- identical within experimental limits. field by comparison of their IR spectra with those of Competitive Isopropylation of Toluene —Benzene authentic samples. with Triisopropyl Phosphite. Competitive isopropylation The mixture of di-substituted aromatic hydrocar- of toluene — benzene mixtures was carried out by bons, collected on column (A) was analyzed by IR changing the concentration of toluene — benzene mix- spectroscopy for the aromatic substitution pattern in tures from 1:1 molar ratio to 1:4, 2:3, 3:2, and 4:1. the 11.6 — 14.8 /u region. Characteristic absorption The data obtained are summarized in Table VI. wavelengths (strong absorptions), associated with each Competitive Isopropylation of Alkyl Benzenes —Ben- isomer, were used to determine the isomers in the pro- zene. These were carried out with a 1:1 molar ratio duct mixture. of alkyl benzene — benzene. The data obtained are Di-substituted Absorption summarized in Table VII. The analyses of the pro- Isomer Range ducts obtained on isopropylation of the alkyl ben- o 13.0 — 13.6 p zenes are indicated for each case. m 12.3-13.3 p Ethyl Benzene. The molecular weight of the mix- 14.0-14.8 p ture of monoalkylated products corresponded to the p 11.6 — 12.5 p molecular formula CX1H16. Found: 151, calcd. for CnH16: 148. The IR spectroscopic analysis of the The isomeric di-substituted aromatic hydrocarbons were product mixture indicated the presence of o-, m-, and then separated on column (B). Each of the isomers p-isomers. The product mixture could not be resolved was identified by peak enhancement on injection with by glpc. an authentic sample. As the authentic samples of o-cymene, m-cymene and o-diisopropyl benzene could n-Propyl Benzene. The glpc analysis for the pro- not be obtained, and as these products could not be ducts was done on column (B). The molecular weight separately collected over glpc for identification, the of the mixture of monoalkylated products corresponded peak assignment on the gas chromatogram was done to the molecular formula C12H18 . Found: 165, calcd. in these cases on the strength of the IR absorption for C12H18: 162. The IR spectroscopic examination of pattern of the mixture of isomers collected on column the product mixture failed to show clear bands due to (A). . o-, m-, and p-isomers. The product mixture could not The isomeric tri-substituted aromatic hydrocarbons be resolved by glpc analysis. were well separated on column (A), and were col- p-Cymene. The glpc analysis for the products was lected and identified by comparison of their IR spectra done on column (B). The product was identfied to be with those of authentic samples, if available, or by diisopropyl toluene, corresponding to the molecular 1348 ALKYLATION OF AROMATIC COMPOUNDS WITH PHOSPHORUS ESTERS 1348

formula C13H20 • Molecular weight found: 176, calcd. C9HUC1: 154. The IR spectroscopic examination in- for C13H20: 176. The IR spectroscopic examination dicated the presence of o-, m-, and p-isomers. The pro- showed a medium band at 11.2 /z and a strong band duct mixture could not be resolved by glpc analysis. at 12.2 n (1,2,4-substitution), and a strong isopropyl doublet at 7.3 n and 7.4 /u. The position of the in- We thank Eastman Organic Chemicals, Rochester. coming isopropyl group (o to methyl or isopropyl) N.Y.; Hooker Chemical Corp., Industrial Chemicals could not be assigned. Division, Niagara Falls, N.Y.; Mobil Chemical, In- Chlor ob enzene. The glpc analysis for the products dustrial Chemicals Division, Richman, Va.; Victor was done on column (B). The molecular weight of the Chemical Division, Stauffer Chemical Co., Dobbs Ferry, mixture of monoalkylated products corresponded to N.Y.; and Aldrich Chemical Co., Milwaukee, Wise, for the molecular formula C9HUC1. Found: 151, calcd. for supplies of various phosphorus esters.

1 (a) This investigation was supported by grants from the 14 S. H. PATINKIN and B. S. FRIEDMAN, in: "FRIEDEL- Public Health Service, U.S. Department of Health, Educa- Crafts and Related Reactions", Vol. II, pp. 147, 149, tion and Welfare, GM 16741 and from the Graduate School 171, G. OLAH, Ed., Interscience Publishers, New York of the University of Wisconsin-Milwaukee; (b) Taken in 1964. part from M. S. Thesis of M. W. SHENDE, University of 15 a) H. C. BROWN and J. D. BRADY, J. Amer. chem. Soc. 74. Wisconsin-Milwaukee, 1971. 3570 [1952] ; b) H. C. BROWN and K. L. NELSON, J. Amer. 2 F. A. DRAHOWZAL, in: "Friedel-Crafts and Related chem. Soc. 75, 6292 [1953]; c) H. C. BROWN and M. Reactions", Vol. II, p. 641, G. OLAH, Ed., Interscience GRAYSON, J. Amer. chem. Soc. 75, 6285 [1953] ; d) H. C. Publishers, New York 1964. BROWN and C. R. SMOOT, J. Amer. chem. Soc. 78, 6255 3 J. F. MCKENNA and F. J. SOWA, J. Amer. chem. Soc. 59, [1956], 1204 [19371. 10 G. A. OLAH, S. H. FLOOD, S. J. KUHN, M. E. MOFFATT, 4 N. BERMAN and A. LOWY, J. Amer. diem. Soc. 60, 2596 and N. A. OVERCHUK, J. Amer. chem. Soc. 86, 1046 [1938]. [1964], 5 B. V. TRONOV and A. M. PETROVA, Zh. Obshch. Khim. 23, 17 R. H. ALLEN and L. D. YATS, J. Amer. chem. Soc. 83, 2799 1019 [1953], [1961]. 6 a) G. SOSNOVSKY, E. H. ZARET, and W. MERTZ, Syn. 142 18 L. SCHMERLING, Ind. Eng. Chem. 40, 2072 [1948]. [1971]; b) G. SOSNOVSKY, E. H. ZARET, and M. KONIECZ- 19 G. A. OLAH, N. A. OVERCHUK, and J. C. LAPIERRE, J. Amer. NY, J. org. Chemistry 37, 2267 [1972]. diem. Soc. 87, 5785 [1965]. 7 G. OLAH, in: "Friedel-Crafts and Related Reac- 20 S. U. CHOI and H. C. BROWN, J. Amer. chem. Soc. 81, 3315 tions", Vol. I, G. OLAH, Ed., Interscience Publishers, New [1959]. York 1964, (a) pp. 36-37; (b) p. 39; (c) p. 40. 21 L. M. STOCK and II. C. BROWN, J. Amer. chem. Soc. 81, 8 D. A. MCCAULAY, in: "Friedel-Crafts and Related 3323 [1959]. Reactions", Vol. II, p. 1061, G. OLAH, Ed., Interscience 22 P. KOVACIC and S. T. MORNEWECK, J. Amer. chem. Soc. Publishers, New York 1964. 87, 1566 [1965], 9 A. W. FRANCIS, Chem. Reviews 43, 257 [1948]. 23 A. J. KIRBY and S. G. WARREN, "The Organic Chemistry 10 G. A. OLAH, S. J. KUHN, and S. H. FLOOD, J. Amer. diem. of Phosphorus", pp. 17, 18, Elsevier Publishing Co., Am- Soc. 84, 1688 [1962]. sterdam/London/New York 1967. 11 G. A. OLAH, M. W. MEYER, and N. A. OVERCHUK, J. org. 24 M. J. SCHLATTER and R. D. CLARK, J. Amer. chem. Soc. 75. Chemistry 29, 2315 [1964]. 361 [1953]. 12 G. A. OLAH, M. W. MEYER, and N. A. OVERCHUK, J. org. 25 F. E. CONDON, J. Amer. chem. Soc. 71, 3544 [1949]. Chemistry 29, 2313 [1964]. 26 C. C. PRICE, Chem. Reviews 29, 37 [1941], 13 a) R. H. ALLEN, L. D. YATS, and D. S. ERLEY, J. Amer. 27 D. J. PASTO and C. R. JOHNSON, "Organic Structure Deter- chem. Soc. 82, 4853 [1960]; b) R. H. ALLEN, T. ALFREY, mination", pp. 43, 44, Prentice-Hall, Englewood Cliffs, N. JR., and L. D. YATS, J. Amer. chem. Soc. 81, 42 [1959]; J. 1969. c) R. H. ALLEN and L. D. YATS, J. Amer. diem. Soc. 81, 5289 [1959] ; d) H. C. BROWN and J. JUNGK, J. Amer. chem. Soc. 77, 5579 [1955],