The complex mixtures resulting from pyrolysis of carboxylic acid salts have been analyzed by combined gas chromatography-mass spectrometry with the following results. Pyrolysis of calcium decanoate at 500 produces a homologous series of nonyl ketones, 2-undecanone through 10-nonadecanone ; normal alkanes and monoalkenes of nine carbon atoms or less are also produced. Pyrolysis of calcium decanoate deuterated at C-8, C-9, and C-10 indicates that the smaller alkyl groups in the above ketones originate from the portion of the alkyl chain nearest the carbonyl group. When pyrolyzed in the presence of ferric chloride hexahydrate (a chlorine radical donor), calcium decanoate does not produce ketones but yields only normal alkanes and monoalkenes of nine carbon atoms or less. Calcium benzoate gives benzene and diphenyl ketone upon pyrolysis at 500. The pyrolysate of calcium 2,2-dimethyloctanoate consists of about equal amounts of 2-methyl-l-octene, 2-methyl-2-octene, and 1-hexene. The above data have been interpreted in terms of a free-radical mechanism in which alkyl and acyl radicals act as initiators. hermal decomposition of salts of carboxylic acids T yields symmetrical ketones. For example, calcium acetate when heated to 350 produces acetone and calcium carbonate.lc Mixtures of salts of two different acids pyrolyze to unsymmetrical ketones in varying yields. Thus, a mixture of benzoate and decanoate salts is converted upon pyrolysis to nonyl phenyl ketone in 50% yield.2 In contrast to these straightforward transformations, there are also some anomalous cases which seem to involve a skeletal rearrangement: salts of trimethylacetic acid give a low yield of tert-butyl isobutyl ketone (rather than the expected di-tert-butyl ketone) under certain pyrolysis conditions. In an attempt to study the mechanism of this reaction (known as ketonic decarboxylation), pyrolyses of salt mixtures, one of which is labeled with 13C or 14C in the carbonyl group, were carried out by several authors. The extent of label retention in the resulting unsymmetrical ketone gave an indication of the source of the carbonyl group. Unfortunately, these data and others5-' were not conclusive; nevertheless, they formed the basis for proposing various mechanisms involving carbanions 3 , 4 f 3 5 (see reactions 1 and 2) , carbonium ions4b (see reaction 3), free r a d i c a l ~ ~ ~ > 6 (see reaction 4), and a four-center intermediate' (see reaction 5) . From the variety of mechanisms reported in the literature, it is obvious that the details of this reaction, even after a century of sporadic work, are still not well understood. In most of these investigations, the emphasis has been on the yield and nature of the major product as (1) (a) C. Friedel, Justw Liebigs Ann. Chem., 108, 122 (1858); ( b ) H. Limpricht, ibid., 108, 183 (1858); (c) R. Fittig, ibtd., 110, 17 (1859). (2) C. Granito and H. P. Schultz, J . Org. Chem., 28, 879 (1963). (3) A. L. Miller, N. C. Cook, and F. C . Whitmore, J. Amer. Chem. Soc., 72, 2732 (1950). (4) (a) J. Bell and R. I. Reed, J . Chem. Soc., 1383 (1952); (b) C. C. Lee and J. W. T. Spinks, J . Org. Chem., 18, 1079 (1953); (c) R. I. Reed, J . Chem. Soc., 4423 (1955); (d) V. D. Nefedov, M. A. Toropova, and I. A. Skulskii, Zh. Fir. Khim., 29, 2236 (1955); (e) M. Okubo and R. Goto, Nippon Kagaku Zasshi, 81, 1132 (1960); (f) L. Otvos and L. Noszko, Tetrahedron Lett., 19 (1960). ( 5 ) 0. Neunhoeffer and P. Paschke, Eer. Deut. Chem. Ges. E, 72,919 (1939). (6) W. A. O'Neill and R . I . Reed, J . Chem. Soc., 3911 (1953). (7) (a) R. I. Reed and M. B. Thomley, ibid., 3714 (1957); (b) A. M. Rubinshtein and V. I. Yakerson, Kine?. Katal., 2, 118 (1961). (1) RCOOM R+ COz + M+ R'COOM + RR'COR + MORCH,COOM RCHCOOM + H+ R'COOM + RCHCOOM -t R'CCHCOOM + MOI1 I OR R'CCHCOOM + H+ R'CCH2R + CO, + M' II 0 II I OR