J. Org.Chem. 1983, 48, 1364-1366
1364
from the insoluble electrophile or from retro-Claisen reactions. C-Alkylation of 15 and 23 with methyl iodide in protic solvents or allylic halides in water proceeds without complication in 6040% yield.1k17 Apparently, alkylation of 15 and 23 with nonactivated electrophiles affords less than 10% of C-alkylation.ls Our results with sulfonium salts in hydrocarbon solvents represent an improvement in all instances. These solvents were chosen to promote the soft nucleophile-soft electrophile interaction. Although some increased carbon alkylation was noted, clearly competing 0-alkylation remains a problem with these enolates. These experiments demonstrate the potential utility of sulfonium salts as electrophiles for the alkylation of 6dicarbonyl anions. These electrophiles may be valuable in transferring primary alkyl groups or in preparing wfunctionalized carbonyl systems.
Experimental Section Infrared spectra were recorded on a Beckman IR 18 AX, a Perkin-Elmer Infracord, a Pye-Unicam SP 1000, or a Pye-Unicam SP 3-200 spectrophotometer;bands yielding structural information are reported in reciprocal centimeters (cm-'), using polystyrene calibration. Nuclear magnetic resonance spectra were recorded on a Varian EM 390 instrument at 25 O C in deuteriochloroform, and the peak positions are reported in parts per million from tetramethylsilane internal standard, using multiplet (m), quartet (q), triplet (t),doublet (d),or singlet (9) to describe the multiplicity. Carbon magnetic resonance spectra were recorded on a JEOL CFT-100 instrument in deuteriochloroform, and the peak positions are assigned relative to the chloroform resonances at 78.161,76.900, and 75.619 ppm. The low-resolution mass spectra were obtained on a Finigan 4000 GCMS DS instrument with sample introduction via direct probe or through a 6-ft GC column containing 3% Dexil 300 on Supelcoport. High-resolution spectra were performed at the Biomedical Mass Spectrometry Resource, University of California, San Francisco, with sample introduction by direct probe. GC analysis was performed on a Varian 3700 gas chromatograph with an FID detector outfitted with a 6 f t X 0.25 in. glass column containing 3% Dexil300 on 100/120 Supelcoport or 3% SE 30 on 100/120 Supelcoport. The (3-dicarbonyls were commercially a ~ a i l a b l e 'and ~ the sulfonium salts prepared by standard routes.20 General Procedure A. A dry sodium enolate of the 8-dimol) prepared from dicarbonyl and sodium hydride carbonyl ( in tetrahydrofuran (THF) and evaporation of the solvent. The mol of the sulfonium salt were suspended enolate and 1.3 X in 20 mL of THF or aromatic hydrocarbon solvent (distilled from calcium hydride). This suspension was heated at reflux for 24 h. Most of the solvent was removed under vacuum, and the residue was poured onto a silica gel column. Elution with hexane afforded the solvent; hexane:ether (kl), the "C"-alkylation product; and hexane:ether (1:3), the "0"-alkylation product. (15)(a) Stetter, H. Angew. Chem. 1955,67,769-784.'Newer Methods of Preparative Organic Chemistry"; Foerst, W.; Ed.; Academic Press: New York, 1961;Vol. 2, pp 51-100. (b) Schwarz, S.; Schaumann, J.; Truckenbrodt, G.; Weber, G.; Meyer, M.; Schick, H.; Welzel, H.-P. 2. Chem. 1979,19,450-451.(c) Narayanan, K.V.; Balasubramanian, K. K.; Chandrasekaran, S.; Romain, S.; Swaminathan, S. J. Chem. SOC.C 1971, 2472-2474. (d) Lansbury, P. T.; Serelis, A. K. Tetrahedron Lett. 1978, 1909-1912. ( e ) Stetter, H.; Sandhagen, H. Chem. Ber. 1967, 100, 2837-2841. ( f ) Rosenthal, D.; Davis, K. H. J. Chem. SOC. C 1966, 1973-1976. (16)For an alternative route, see: Piers, E.; Grierson, J. R. J. Org. Chem. 1977,42,3755-3757. (17)For allylation using palladium, see: Trost, B. M.; Curran, D. P. J. Am. Chem. SOC. 1980,102,5699-5700. (18)(a) Crispin, D. J.; Vanstone, A.; Whitehurst, J. S. W. J . Chem. SOC.C 1970,10-18. (b) McIntosh, J. M.; Beaumier, P. M. Can. J. Chem. 1973,51,843-847. (c) Agosta, W. C.; Smith, A. B. J. Org. Chem. 1970, 35, 3856-3860. (d) Schick, H.; Schwarz, H.; Finger, A.; Schwarz, S. Tetrahedron 1982,38, 1279-1283. (19)Aldrich Chemical Co. (20)(a) Goethfls, E.J.; Drijvers, W.; Van Ooteghem, D.; Buyle, A. M. J. Macromol. Scz. Chem. 1973,7,1375-1390. (b)Birch, S. F.; McAllan, D. T. J . Chem. SOC.1951,3411-3416.
0022-3263/83/1948-1364$01.50/0
General Procedure B. The dry sodium enolate and the sulfonium salt were mixed in the same proportions and distilled from a Kugelrohr apparatus a t 0.013 kPa. The products were separated by the same method. Methyl 1-[4-(Ethylthio)buty1]-2-oxocyclopentanecarboxylate (7): 88% yield; IR 1750,1720 cm-'; NMR 6 1.2 (t, J Hz, 3), 2.2 (m, lo), 2.6 (m, 6), 3.7 (s,3); HRMS observed m / z 240.1204, Cl3HZ2O3S(-18) requires 240.1184. Methyl 1-[3-Methyl-5-(ethylthio)-(Z)-2-pentenyl]]-2-oxocyclopentanecarboxylate(9): 98% yield; bp 50 OC at 0.013 kPa; IR 2980, 1760, 1735, 1450, 1335, 1160 cm-'; NMR 6 1.27 (t, J = 12 Hz, 3), 1.71 (e, 3), 2.2 (m), 2.43 (4, J = 12 Hz, 2), 3.68 (9, 3), 5.10 (t, 1);HRMS observed m / z 284.144064, C15H2403S requires 284.1442. 2-Methyl-2-[4-(ethylthio)butyl]cyclopentane-ly3-dione (19) and 2-Methyl-3-[4-(ethylthio)butoxy]cyclopent-2-en-l-one (20). 19: IR 2920, 1715, 1620, 1445, 1338, 1110 cm-'; NMR 6 1.10-1.80 (m, 12 H), 2.33-2.80 (m, 4 H), 2.87 (9, 4 H); HRMS requires 228.1176. observed m / z 228.1187, C12HzoOzS 20: IR 2920, 1685,1625,1380,1340,1115 cm-'; NMR 6 1.22 (t,3 H, J = 8 Hz), 1.50 (9, 3 H), 1.63-2.00 (m, 4 H), 2.20-2.76 (m, 8 H), 4.20 (t, J = 8 Hz, 2 H); HRMS observed m / z 228.1182, C12Hzo02S requires 228.1176. 2-Methyl-2-[5-(ethylthio)-3-methyl-(Z)-pentenyl]cyclopentane-Iy3-dione(21) and 2-Methyl-3-[5-(ethylthio)-3methyl-(Z)-2-pentenoxy]cyclopent-2-en-l-one (22). 21: IR 2980,2940,2880,1762,1725,1455,1422,1380,1315,1268,1078 cm-'; NMR 6 1.02 ( 8 , 3 H), 1.24 (t, J = 8 Hz, 3 H), 1.68 (9, 3 H), 2.05-2.58 (m, 8 H), 2.62 (s,4 H), 4.90 (t, J = 8 Hz, 1 H); 13CNMR 6 215.75, 215.75, 137.73, 118.59, 56.08, 34.91, 34.91, 34.03, 31.39, 29.21, 25.41, 22.79, 17.96, 14.24; HRMS observed m / z 254.1348, CI4Hz2OzSrequires 254.1345. 22: IR 2978, 2940, 2880, 1692, 1630, 1448, 1395, 1380, 1338, 1118, 970 cm-'; NMR 6 1.30 (t, 3 H, J = 8 Hz), 1.60 (9, 3 H), 1.86 (s, 3 H), 2.20-2.70 (m, 10 H), 4.70 (d, J = 8 Hz, 2 H), 5.48 (brd t, J = 8 Hz, 1 H); MS (70 eV), m / z 254 (M'); HRMS observed m / z 254.1345, C14H2202S requires 254.1345. 2-Methyl-2-[4-(ethylthio)butyl]cyclohexane-ly3-dione (28) and 2-Methyl-3-[4-(ethylthio)butoxy]cyclopent-2-en-l-one (29). 28: IR 2940,2885,1730,1698,1455,1380,1266,1135,1026 cm-'; NMR 6 1.12-1.33 (m, 6 H), 1.33-2.10 (m, 8 H), 2.20-2.80 (m, 8 H); HRMS observed m / z 242.1341, C13Hz202Srequires 242.1334. 29: IR 2920, 1610,1380,1355,1098cm-'; NMR 6 1.23 (t, 3 H, J = 8 Hz), 1.67 (s, 3 H), 1.70-2.18 (m, 9 H), 2.20-2.70 (m, 8 H), 4.02 (t, 2 H, J = 6 Hz); HRMS observed m / z 242.1329, C13H,02S requires 242.1334. Registry No. 1, 10472-24-9; 2, 676-88-0; 3, 30680-84-3; 4, 85098-51-7;5,25684-00-8;6,696-98-0;7,83705-56-0;8,83705-60-6; 9,83705-61-7; 10, 1193-63-1; 11, 37709-42-5; 12, 15839-18-6; 13, 61783-91-3; 14, 15839-65-3; 15, 765-69-5; 16, 3883-58-7; 17, 25112-86-1; 18, 25112-87-2; 19, 85098-52-8; 20, 85098-53-9; 21, 85098-54-0; 22, 85098-55-1; 23, 1193-55-1; 24, 562-13-0; 25, 25112-91-8; 26, 25112-82-7; 27, 20643-20-3; 28, 85098-56-2; 29, 85098-57-3. N
An Improved a-Chlorination of Carboxylic Acids Robert J. Crawford T h e Procter & Gamble Company, Miami Valley Laboratories, Cincinnati, Ohio 45247 Received August 3, 1982
Although the Hell-Volhard-Zelinksy (HVZ) cu-bromination of carboxylic acids was discovered a century ago,' extension of this chemistry to a-chlorination has occurred only recently. The propensity of chlorine to undergo competing free radical reactions under HVZ conditions (1)Hell, C. Ber. Dtsch. Chem. Ges. 1881,14,891.Zelinksy, N.Ibid. 1887,20,2026. Volhard, J. Justus Liebigs Ann. Chem. 1887,242,141.
0 1983 American Chemical Society
Notes required the development of specialized procedures that favor the ionic a-substitution process.2 Ogata et al. developed an appealing procedure that entails addition of gaseous chlorine to a neat aliphatic acid a t 140 "C in the presence of a strong acid catalyst and a free radical inh i b i t ~ r . The ~ preferred catalyst is chlorosulfonic acid (10 mol % of the carboxylic acid), and the inhibitor is oxygen gas. In our hands, this technique was adequate for chlorination of short-chain acids (up to C-8) but was capricious and often unsuccessful when applied to long-chain system such as stearic acid. Ogata later revised the conditions for long-chain acids by increasing the C1S03H level to 25 mol % , although competing a-sulfonation was e n h a n ~ e d . We ~ have developed a modification of this reaction that yields a number of unexpected benefits. The crucial change is the use of 7,7,8,8-tetracyanoquinodimethane (TCNQ, 0.5 mol % ) as the free radical inhibitor in place of oxygen. Compared with 02,TCNQ affords a significantly improved reaction that is also unaffected by acid chain length. The major benefits are (1)decreased reaction time, (2) reduced C1S03H concentration (3 mol %), (3)increased yield, (4) negligible polychlorination, and (5) simplified purification.
In practice, the optimum result is obtained when the chlorine flow rate, temperature, and reaction time are controlled very precisely (see Experimental Section). The chlorination proceeds so smoothly in the presence of TCNQ that laboratory preparations ranging in scale from 0.1 to 20 mol are controlled with ease. We routinely carry out 8-12-mol runs in a 5-L reaction flask with the chlorine flow at 4-6 LImin.5 For operational convenience, we favor a chlorine rate that affords complete conversion in approximately 1h a t 150 OC. Modifying the scale requires nothing more than adjusting the gas addition rate in proportion to the moles of reagents. It is vital to have TCNQ present throughout the reaction. TCNQ is slowly degraded under these conditions,6 and the quantity used is the minimum that will survive through the end of the reaction. The TCNQ is normally added in two portions, consisting of two-thirds of the specified amount at the start and the remainder when the reaction is 75% complete. The apparatus is fitted with a dry ice condenser, which refluxes unreacted chlorine but allows hydrogen chloride to escape. This technique provides an effective signal for the end of the reaction, because a rapid drop in temperature is observed when condensing chlorine no longer reacts. The reaction must be terminated at this point to avoid the formation of polychlorinated byproducts. If the resulting a-chloro acid is solid, the hot reaction mixture is poured into a suitable solvent for recrystallization of the product. Crystalline a-chloro acids are usually obtained in yields approaching 90%. Liquid products can be isolated by direct distillation of the reaction mixture, although (2) Little, J. C.; Sexton, A. R.; Tong, Y.-L. C.; Zurawic, T. E. J.Am. 1969,91,7098. Harpp, D. N.; Bao, L. Q.; Black, C. J.; Gleason, Chem. SOC. J. G.; Smith, R. A. J. Org. Chem. 1975, 40, 3420.
(3) Ogata, Y.; Harada, T.; Matsuyama, K.; Ikejiri, T. J. Org. Chem. 1975, 40, 2960. (4) Ogata, Y.; Sugimoto, T.; Inaishi, M. Bull. Chem. SOC.Jpn. 1979, 52, 255. (5) Although TCNQ is used at only 0.5 mol %, the high cost and
limited availability of commercial TCNQ may constitute a drawback in largescale chlorinations. A practical TCNQ synthesis that overcomes this problem is reported in the accompanying paper.12 (6) TCNQ is probably consumed by side reactions that are unrelated to the a-chlorination. Attempts to isolate its degradation products have led to intractable mixtures.
J. Org. Chem., Vol. 48, No. 8, 1983 1365 yields are slightly lower (80-85%) owing to distillation losses. In either case, purity is typically 99%. a-Chloro and a-bromo carboxylic acids can be converted to a wide variety of other a-substituted acids by nucleophilic displacement. Although the chloro acids are somewhat less reactive, both halides undergo ready displacement by common nucleophilic reagents, with little or no 6 elimination.' The a-chlorination process described here is considerably faster and easier to carry out than HVZ bromination.8 In most cases, the lower reactivity of chloro vs. bromo acids does not counterbalance the greater convenience of chlorination. For this reason, we consider a-chloro acids to be the more desirable intermediates. When a-substituted acids must be prepared on a large scale, the two-step route via the a-chloro derivative offers an attractive alternative to single-step procedures involving derivatization of a-lithio carboxylate^.^ The mechanistic function of TCNQ in the chlorination is unknown. It is perhaps simplest to assume that it participates as an unusually effective free radical inhibitor.1° The effect of small amounts of TCNQ on the chlorination is so dramatic that it is tempting to consider a catalytic role, despite the lack of precedent for a convincing mechanism. The use of TCNQ as a modulator of chemical reactions has received little attention, in contrast to the vast amount of research that followed its discovery as an electron acceptor." We are investigating other novel applications of this intriguing substance. Experimental Section General Methods and Materials. Stearic acid (practical grade) and acetonitrile (ACS reagent grade) were purchased from MCB, Inc. Chlorosulfonic acid was obtained from Eastman Kodak Co. and was redistilled before use. TCNQ was prepared by the method described in the accompanying paper.12 Chlorine gas was Matheson "high purity" grade. Melting points were determined with a Thomas-Hoover apparatus and are uncorrected. The following procedure for a-chlorination of stearic acid is representative of reactions that have been run with all of the even-chain saturated fatty acids between c6 and CIS and 2ethylhexanoic acid. The procedures and results are essentially identical for all substrates. The a-chloro acids from c6 to cloare liquid a t room temperature and are purified by direct vacuum distillation of the reaction mixture. The solid a-chloro acids from C12and up can be recrystallized from either hexane or acetonitrile, with the latter being favored for the longer chain lengths. 2-Chlorostearic Acid. A 1-L, five-necked round-bottomed h k was charged with 455 g (1.6 mol) of stearic acid and was fitted with a mechanical stirrer, thermometer, dry ice condenser (with gas exit tube), and two fritted gas dispersion tubes (opposite necks). The dispersion tubes were connected via PVC tubing to a T connector, the third arm of which was connected to a chlorine gas source containing an in-line flow meter. The flow meter had been previously calibrated for chlorine flow rates of 200 and 800 ml/min. The stearic acid was melted by heating it to 80 "C with a heating mantle. After addition of 1.1g (5.4 mmol) of TCNQ and 3.2 mL (0.05 mol) of chlorosulfonic acid, the chlorine flow (7)These reactions have been studied extensively in our laboratories, and the results will be reported separately. (8) Attempts to adapt a-bromination to this procedure have been unsuccessful. (9) Petragnani, N.; Yonashiro, M. Synthesis 1982, 521. (10)The more commonly used free radical inhibitors, such as phenolic antioxidants, are much less effective than TCNQ in this reaction. Also we have studied a number of other reactions in which TCNQ functions as an inhibitor. (11)Devreese, J. T.; Evrard, R. P.; Van Doren, V. E., Eds. "Highly Conducting One-Dimensional Solids"; Plenum Press: New York, 1979. Miller, J. S.;Epstein, A. S., Eds. Ann. N.Y. Acad. Sci. 1978, 313. Perlstein, J. H. Angew. Chem.,Int. Ed. Engl. 1977,16,519. Bespalov, B. P.; Titov, V. V. Russ. Chem. Rev. (Engl. Transl.) 1975, 44, 1091. (12) Crawford, R. J. J. Org. Chem., following paper in this issue.
J. Org. Chem. 1983,48, 1366-1368
1366
was started at a rate of 200 mL/min. The mixture was stirred vigorously and heated rapidly to 145 O C . At this point (which is considered time 0 for timing purposes) the chlorine flow rate was increased to 800 mL/min. Thoughout the reaction the flow rate was held constant, and the temperature was maintained at 150 3 O C . The reaction is mildly exothermic and, in its early stages, can be kept at this temperature by using an unheated mantle as an insulator. After the conversion is more than 50% complete it becomes necessary to apply heat via the mantle. At a reaction time of 50 min a second portion of 0.56 g (2.7 mmol) of TCNQ was added. At 65 min the temperature began to drop rapidly because of condensing chlorine, and the gas flow was stopped. When the mixture had cooled to ca. 90 O C , it was poured into 1200 mL of hot acetonitrile. The product was crystallized by cooling this solution in ice with stirring. The crystals were collected by suction filtration, washed with ice-cold acetonitrile, and vacuum dried to afford 450.6 g (88%)of 2-chlorostearicacid, mp 63.5-64.5 O C (lit.13mp 64.5-65.5 "C). GC analysis of this material (as methyl ester) indicated its purity to be ca. 99%.
*
Acknowledgment. Technical assistance was provided by Mr. B. A. Banker, Mrs. N. L. Kern, and Mr. J. F. Ward and is gratefully acknowledged. Registry No. 2-Chlorostearic acid, 56279-49-3; stearic acid, 57-11-4; chlorosulfonic acid, 7790-94-5;methyl 2-chlorostearate, 41753-99-5; TCNQ, 1518-16-7;02,7782-44-7. (13)Hwang, Y.-S.; Nawab-Gojrati, H. A,; Mulla, M. S. J. Agric. Food
Chem. 1978,26,1293.
A Practical Synthesis of 7,7,8,8-Tetracyanoquinodimethane Robert J. Crawford T h e Procter & Gamble Company, Miami Valley Laboratories, Cincinnati, Ohio 45247 Received August 3, 1982
The most commonly used method to prepare 7,7,8,8tetracyanoquinodimethane (TCNQ, 3) is the original synthesis reported in 1962 by Acker and Hertlerl (Scheme I). This procedure is adequate for small-scale work but has shortcomings when applied on a large scale. As a consequence, TCNQ that is sold commercially is characteristically expensive and available only in small l o k 2 In this paper I report an improved procedure that is suitable for laboratory preparation of TCNQ in molar quantities. If a synthesis is to be truly practical, it should start with inexpensive commodity chemicals, provide a high overall yield, and be amenable to scale-up. According to the first criterion, 1,4-cyclohexanedione (1) is an intermediate rather than a starting material, and its synthesis must be included in the overall route to TCNQ. This diketone is usually prepared by hydrolysis and decarboxylation of diethyl 1,4-dioxocyclohexane-2,5-dicarboxylate, which is, in turn, obtained by self-condensation of diethyl succinate? The average yield for the two steps is 56%, and the procedure is awkward to scale up.4 While the conversion of 1 to 1,4-bis(dicyanomethy1ene)cyclohexane (BDCC, 2) is rapid and quantitative, the conventional oxidation of 2 to TCNQ suffers from (1)Acker, D.S.; Hertler, W. R. J. Am. Chem. SOC. 1962,84, 3370. (2)The largest quantity supplied by Aldrich Chemical Co. (1982-1983 Catalog) is 10 g a t $45.00. (3)Nielsen, A. T.; Carpenter, W. R. 'Organic Syntheses"; Wiley: New York, 1973;Collect. Vol. V, p 288. (4)A further drawback of this route is that only 32% of the mass of the diethyl succinate is retained in 1.
0 1
2
Scheme 11
0 0 QH
QH
3Hz, HzO e
ZNaOCI, HzO,
W-7 RaNi
AH
? OH
4
5
2
RuCli
,
2CHz(CN)z, H :, NaHCO,
2CI2, 4CsH5N e 3 CH,CN
scale limitations in the isolation and purification of the final product. When BDCC is treated with 2 equiv of bromine and 4 equiv of pyridine in acetonitrile solution, TCNQ is formed along with 4 equiv of pyridine hydrobromide, and the two products partially coprecipitate. Water is then added to dissolve the salt and precipitate all of the TCNQ. Unfortunately, TCNQ contains up to 10% of other impurities when it is isolated in this way5 and must be sublimed or recrystallized. Both purification techniques limit the practicality of large-scale preparations, the latter because of the extremely low solubility of TCNQ.6 Our procedure (Scheme 11) is designed to overcome all of these limitations. Hydroquinone (4) is used as the starting material and is converted to BDCC by a threestep, one-pot procedure using water as the solvent. BDCC is then oxidized to TCNQ by using chlorine and pyridine, a modification that allows the purification of TCNQ to be effectively built into the reaction. A detailed account of the synthesis follows. Hydroquinone can be converted to 1 by a two-step reduction-oxidation sequence in which 1,4-cyclohexanediol ( 5 ) is an intermediate. The best known procedures for these reactions are catalytic hydrogenation of hydroquinone in alcohol solution and chromic acid oxidation of 5 in acetone or acetic acid.7 Water is the preferred solvent for the conversion of 1 to BDCC because the product is insoluble and crystallizes directly from the reaction mixture.' We reasoned that the isolation of both 5 and 1would be unnecessary if all three of these steps could be carried out in water. This approach succeeded because of the availability of two rarely used synthetic techniques. The reduction of hydroquinone in an alkaline 50% water slurry is effected by hydrogenation over W-7 Raney nickel. The unique combination of hydroquinone and the W-7 catalyst was mentioned briefly by Adkins and Billica in 194@ but has not been exploited. It is an unusual case in ( 5 ) The major contaminant is an amorphous brown powder that is obtained as a residue from sublimation. This byproduct has not been identified and appears to be polymeric. Interestingly, TCNQ samples that contain this impurity do not show a significant melting point depression but melt with decomposition. (6)Although acetonitrile is the best recrystallization solvent, as much as 80 mL/g may be required to dissolve TCNQ a t the boiling point. (7) Mussini, P.; Orsini, F.; Pelizzoni, F. Synth. Commun. 1975,5,283. Kern, W.; Gruber, W.; Wirth, H. 0. Mukromol. Chem. 1960,37, 198. Sircar, J. C.; Meyers, A. I. J. Org. Chem. 1965,30,3206. Gogek, C. J.; Moir, R. Y.; Purves, C. B. Can. J. Chem. 1951,29,946. Olberg, R. C.; Pines, H.; Ipatieff, V. N. J.Am. Chem. SOC.1944,66,1096.Owen, L. N.; Robins, P. A. J. Chem. SOC.1949,320.
0022-3263I83 11948-1366gO1.50,I O 0 1983 American Chemical Society I
3
