ma THERMAL DECOMPOSIIIQN OF fi-EEE'WL 3.mammmwgwmmvuumvg, ‘ A was RADECAL a,2.A;ch MaGRAnoN ’ ~ m... germbeqreai.s--ips.ie;;gxésfieiff mmmsrm;mmmsm_; : . f . ‘ Curtis L. .Kafl; 7 7 7 1967‘;'fr" w—vd a LIBRARY Michigan State University _ This is to certify that the thesis entitled THE THERMAL DECOMPOSITION OF t—BU‘I‘LY B-METHYL-B-PI-UENYLPERIEVULINATE, A FREE RADICAL l, 2-ACYL MIGRATION presented by Curtis L. Karl has been accepted towards fulfillment of the requirements for PhoDo degree in ChemiStry 77777.. 77. 28% Major prdfessor Date February 61 1967 0-169 ABSTRACT THE THERMAL DECOMPOSITION OF t-BUTYL S-METHYL-S-PHENYLPERLEVULINATE, A FREE RADICAL 1,2-ACYL MIGRATION by Curtis L. Karl The purpose of this investigation was to demonstrate a 1,2-acyl shift to a carbon free radical, in solution and at temperatures below 1500C. To this end, t-butyl 5-methyl-5- phenylperlevulinate (I) was prepared, and 0.25 M solutions 0 tau/ted ”>0” I of I in phenyl ether and p—cymene were decomposed at 1500. Roughly 90% of the perester decomposed during a three hour period and the carbon dioxide evolution was determined to be 71% in phenyl ether and 84% in p—cymene. The reaction products at this time consisted of small amounts of 3—methyl- 5-phenyllevulinic acid, a 25% yield of a single cage product, 5-phenyl-3,6,6—trimethyl-5-oxa-2—heptanone (II) (either sol- vent), and 45% and 77% yields of acyl rearranged products in phenyl ether and p—cymene respectively, based on moles of starting perester. No unrearranged substrate was detected in Curtis L. Karl any form other than the t-butyl ether (II), and phenyl re- arranged products accounted for less than 1% of the total products. Decarboxylation is believed to be essentially concerted with the cleavage of the perester oxygen-oxygen bond; the resulting recombination of substrate and t-butoxy radicals within the solvent cage must also be esentially concerted, or must involve strong interactions between the radicals. Furthermore, radical rearrangement is also very fast, and possibly concerted with decarboxylation. THE THERMAL DECOMPOSITION OF t-BUTYL 5-METHYL-5-PHENYLPERLEVULINATE, A FREE RADICAL 1,2-ACYL MIGRATION BY Curtis L? Karl A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1967 To my wife, ii Arla ACKNOWLEDGMENTS The author wishes to express his sincere appreciation to Professor William H. Reusch for his invaluable guidance and encouragement during this investigation. Grateful acknowledgment is extended to the National Science Foundation for providing financial assistance from January, 1964 to December, 1966, and to the National Insti- tutes of Health for assistance from January to February, 1967. Appreciation is also extended to the Dow Chemical Company for summer fellowships in 1964 and 1965. Special thanks are also extended to my parents for their encouragement throughout the years of my academic studies. iii TABLE OF CONTENTS Page HISTORICAL . . . . . . . . . . . . . . . . . . . . . 1 RESULTS AND DISCUSSION . . . . . . . . . . . . . . . 8 I. Introduction . . . . . . . . . . . . . . . . 9 II. Syntheses. . . . . . . . . . . . . . . . . . 10 A. The Starting Perester . . . . . . . . . 10 B. The Reaction Products . . . . . . . . . 12 III. Products from the Thermal Decomposition. . . 15 A. Method of Analysis. . . . . . . . . . . 15 B. Reaction Products . . . . . . . . . . . 14 1. Unreacted Perester and Products from the t—Butoxy Radical. . . . . 14 2. Carbon Dioxide . . . . . . . . . . 15 5. Cage Products. . . . . . . . . . . 17 4. Rearranged Products. . . . . . . . 2O 5. Solvent Products . . . . . . . . . 24 IV. Mechanism of Rearrangement . . . . . . . . . 24 V. Conclusion . . . . . . . . . . . . . . . . . 26 EXPERIMENTAL . . . . . . . . . . . . . . . . . . . . 28 I. General. . . . . . . . . . . . . . . . . . . 29 II. Syntheses. . . . . . . . . 50 A. Preparation of t— -Butyl- 5—methyl— -5- phenylperlevulinate . . . . . . . . SO 1. Preparation of 5— —Bhenyl- -2 —butanone SO 2. Preparation of Ethyl 5—Methy1-—5— phenyllevulinate . . . . . . . . . 31 3. Preparation of S—Methyl-S-phenyl— levulinic Acid . . . . . . . . . . 55 4. Preparation of the Enol Lactone, S-Methyl- 4-methyleno-5-phenyl-7— butyrolactone. . . . . . . . . 59 5. Preparation of t- -Butyl S-Methyl- 5—phenylperlevulinate. . . . . . . 46 iv TABLE OF CONTENTS - Continued B. Preparation of Ketone Products Expected from Thermal Decomposition. . . . . . . 1. Preparation of 4-Phenyl-2-pentan- one. . . . . . . . . . . . . . . . 2. Preparation of 5-Methyl—4-phenyl- 2-butanone . . . . . . . . . . . . 5. Preparation of 5-Pheny1—2—pentan- one. . . . . . . . . . . . . . . 4. Preparation of 5—Methyl— —5— —phenyl— 2-butanone . . . . . . . . . . . III. Thermal Decomposition Procedure. . . . . . . A. In p-Cymene . . . . . . . . . . . . . 1. Purification of p-Cymene . . . . . 2. Reaction and Analysis Procedure. . B. In Phenyl Ether . . . . . . . . . . . 1. Purification of Phenyl Ether . . . 2. Reaction and Analysis Procedure. . IV. Identification of Reaction Products. . . . . A. Identification of 5-Phenyl-5,6,6-tri- methyl—5-oxa-2-heptanone. . . . . . . . B. Identification of 4-Phenyl—5-penten-2- one . . . . . . . . . . . . . . . . . . C. Identification of 4- -Phenyl-4—penten— —2- one . . . . . . . . . . . . . . . . . D. Identification of 4-Methyl— 4-phenyl- 5- (p- ~cymyl) -2-pentanone . . . . . . . E. Identification of cis—1—Acetyl—2,4,5— trimethyl-4,5—diphenylcyclopentene. F. Identification of trans—1—Acetyl-2,4,5— trimethyl—4,5—dipheny1cyclopentene. . LITERATURE CITED . . . . . . . . . . . . . . . . . . Page 50 50 51 58 62 69 69 69 69 72 72 75 75 75 77 77 77 87 87 94 LIST OF FIGURES FIGURE Page 1. Plot of log [III]/[III]-[C02'T] versus time. . 16 2. Infrared spectrum of 5—phenyl-2-butanone (film) . . . . . . . . . . . . . . . . . . . . 52 5. Nuclear magnetic resonance spectrum of 5‘ phenyl—2—butanone. . . . . . . . . . . . . . . 54 4° Infrared spectrum of ethyl 5-methyl- -5- -phenyl- levulinate (film). . . . . . . . . . . . . . 56 5. Nuclear magnetic resonance spectrum of ethyl 5-methyl-5-phenyllevulinate. . . . . . . . . . 58 6. Infrared spectrum of 5—methyl-5-phenyllevu- linic acid (CCl4). . . . . . . . . . . . . . . 4O 7. Nuclear magnetic resonance spectrum of 5- methyl-5—phenyllevulinic acid. . . . . . . . . 42 8. Infrared spectrum of 5-methyl—4—methyleno-5- phenyl-y-butyrolactone (film). . . . . . . . . 45 9. Nuclear magnetic resonance spectrum of 5—: methyl—4-methyleno-5—phenyl—y-butyrolactone. . 45 10. Infrared spectrum of t-butyl—5—methy1-5—phenyl- perlevulinate (film) . . . . . . . . . . . . . 47 11. Nuclear magnetic resonance spectrum of t-butyl- 5—methyl~5—phenylperlevulinate . . . . . . . . 49 12. Infrared spectrum of 4—phenyl—2—pentanone (CC14) . . . . . . . . . . . . . . . . . . . . 52 15. Nuclear magnetic resonance spectrum of 4- phenyl-2-pentanone . . . . . . . . . . . . . . 54 14. Infrared spectrum of 5—methyl- -4- -phenyl— —5- buten- 2- -one (film) . . . . . . . . . . . . 55 vi LIST OF FIGURES - Continued FIGURE Page 15. Nuclear magnetic resonance spectrum of 5— methyl-4-phenylL5-buten-2-one . . . . . . . . 57 16. Infrared spectrum of 5—methyl-4-phenyl-2- butanone (film) . . . . . . . . . . . . . . . 59 17. Nuclear magnetic resonance spectrum of 5- methyl—4epheny1-2-butanone. . . . . . . . . . 61 18. Infrared spectrum of 5-phenyl—2-pentanone (film). . . . . . . . . . . . . . . . . . . . 65 19. Nuclear magnetic resonance spectrum of 5- phenyl-2-pentanone. . . . . . . . . . . . . . 65 20. Infrared spectrum of 5-methyl-5—phenyl—2- butanone (film) . . . . . . . . . . . . . . . 66 21. Nuclear magnetic resonance spectrum of 5- methyl-5-phenyl—2-butanone. . . . . . . . . . 68 22. Thermal decomposition reaction apparatus. . . 7O 25. Infrared spectrum of 5-phenyl-5,6,6-tri— methyl—5—oxa-2—heptanone (film) . . . . . . . 74 24. Nuclear magnetic resonance spectrum of 5-phenyl—5,6,6-trimethyl-5—oxa-2-heptanone. . 76 25. Infrared spectrum of 4-phenyl-5-penten-2-one (CC14) o o o o o o o o o o o o o o o o o o o o 78 26. Nuclear magnetic resonance spectrum of 4-phenyl—5-penten-2—one . . . . . . . . . . . 8O 27. Infrared spectrum of 4-phenyl-4—penten—2—one (cc14). . . . . . . . . . . . . . . . . . . . 81 28. Nuclear magnetic resonance spectrum of 4-phenyl-4—penten-2—one . . . . . . . . . . . 85 29. Infrared spectrum of 4-methyl-4-phenyl—5- (p-cumyl)—2—pentanone (CC14). . . . . . . . . 84 vii LIST OF FIGURES - Continued FIGURE Page 50. Nuclear magnetic resonance spectrum of 4— methyl—4-phenyl-5-(p-cumyl)—2-pentanone. . . . 86 51. Infrared spectrum of cis- -1- -acetyl- -2, 4, 5— tri— methyl— —4, 5- -diphenylcyclopentene (CCl4) . . . . 88 52. Nuclear magnetic resonance spectrum of cis-1- acetyl— —2, 4, 5- -trimethyl— -4, 5- -diphenylcyclo- pentene. . . . . . . . . . . . . . . . . . 9O 55. Infrared spectrum of trans-1-acetyl-2,4,5-tri— methyl—4,5-diphenylcyclopentene (CCl4) . . . . 91 54. Nuclear magnetic resonance spectrum of trans— 1— —acetyl- -2, 4, 5- -trimethy1- -4, 5- —dipheny1cyclo- pentene. . . . . . . . . . . . . . . . . . . 95 viii HIS TORICAL HISTORICAL The intensive study of free radical rearrangements com- prises a relatively new direction for organic research. In spite of its youth, sufficient evidence exists to conclude that free radicals exhibit much less tendency to undergo re- arrangements than do more electron—deficient species. In addition, free radicals show properties and reactions, includ~ ing rearrangements, which are independent of their media, and essentially independent of their mode of formation (1). The simplest type of rearrangement a free radical may undergo is that of an intramolecular 1,2—shift. RlR2R3C-CH2° ‘L > RleC—CHgRs The bulk of research on such migrations have dealt with radi- cal systems where substituents have been aryl, halogen, alkyl and hydrogen. It is important to note here that this study as well as the work mentioned above involves mono-radical sys— tems. The behavior of di-radical systems, whether induced thermally or photochemically, does not necessarily parallel mono-radical systems. This will be illustrated in a future section. Of the groups known to migrate from carbon to carbon in mono-radical systems, 1,2-aryl shifts have been most fre- quently observed. Examples include the following: A OgC- —CHCH3 % ggc— “gHCHB _—_* Products (2) 1000 CH2' CHQg g 5 Products (5) B 89- 94% _a—+@ c O-CHCHE‘ a E CH3CHCH2d 5 Products (4) CH3 4.0% C? $H3 st D p-NOch—Cfié' -——4L—$> ’C—CHZGOgN-p-————%> Products (5) I | CH3 95% ® CH3 Interesting enough, the driving force of this rearrange— ment does not appear to be entirely dependent upon conversion of a primary alkyl radical to a more stable tertiary or benzyl radical, nor on steric crowding, as evidenced by Slaugh (6): 14 14 E 21 CHECHgo ——°————>- -CH2CH2 £5 ——->- Products 2.5-5.17’o@ In addition to aryl migrations, good evidence exists for 1,2—shifts of chlorine in radical systems. Several examples reported by Nesmeyanov (7) include: c1 . .I F C12C=CHCH23 25 «<— C13CCHCH2S fl —12-——>- C12CCHCH2S e —>- 11% yield 50% yield Products 4 Cl G BrCHgC-CC13-*——z~—>- BrCHg%-CC12 —————>— Products éHe 100% ® CH3 Although evidence for 1,2—shifts of bromine is not as voluminous as that for chlorine, Skell (8) has reported several such examples. H CH3CHCH2- —«Q-—>— CH3CHCH2Br ___>. Products Br 100% @ CH3 I Cch-CHg' —_‘£"‘+ ‘ CHBCCHgBr —'—-°_7‘ PrOdUCtS I Br 100% ® CH3 The question concerning the existence of 1,2-shifts of alkyl or hydrogen has been widely debated. Currently, however, the preponderance of evidence indicates that neither alkyl nor hydrogen migrate in known mono—radical systems at ordi- nary temperatures. Walling (9) states ”In most cases where they have been reported, alternate explanations are available for the observed results, or the radical nature of the re— action is in doubt.” The arguments for and against hydrogen shifts have recently been reviewed by Reutov (10). The importance of limiting this study to mono—radicals is now evident from the work of Greene and co—workers (11). 0 q 1340 I J ——-—4Lw_e>- g + + C02 / 0014 70% 15% g /I_0 g The thermolySiscfifperoxylactone (I) results in a five-fold preference for methyl migration over phenyl migration. Further evidence for 1,2-alkyl shifts as well as hydrogen migrations is found in di-radical systems generated in photochemical reactions. If the 1,2-shifts cited above are truly intramolecular, they must go through some triangular arrangement of groups which represents either the transition state or a metastable intermediate. It is interesting to note that the only groups unambiguously known to undergo intramolecular 1,2—shifts to radical centers are those able to form three—membered cyclic intermediates. In the case of aryl migration, the bridged intermediate is subject to additional electron delocalization through resonance; in the case of chlorine, the bridged inter- mediate involves an expanded valence shell for chlorine. Recent LCAO calculations for the threeemembered transi- tion state of a 1,2—alkyl radical shift from one unactivated carbon to another, indicate a transition state endothermic by +0.4 8, or about 6-20 kcal/mole depending on the value of 464? o 8 assumed (12). There are sound quantum mechanical reasons for expecting free radical 1,2—alkyl rearrangements to be energetically less favorable than corresponding carbonium ions. The energy levels for the cyclic three atom pseudo-W- electron system representing a transition state or inter- mediate for such a rearrangement are distributed such that two electrons can be placed in bonding orbitals, as in carbonium ion cases; the extra electron in the radical case must occupy an anti-bonding orbital, resulting in less delocalization energy (15). In addition to 1,2-radical shifts of phenyl and halogen, there is growing evidence for the migration of other un- saturated functions. Several examples of 1,2-vinyl shifts have been reported (e.g., equations K and L). es es K CH3CH=CHCHCH2° —‘2'—"* CH3CH=CHCH2CH' _—'—>‘ 88% QQ Products (14) CH2 ° L e E . 5 Products (15) \ 80% Q@ Similarly, a 1,2-acyl migration, which might be thought of as the oxygen analog of vinyl migration, has been reported in two instances. OH ° 0 OH O M //LK\—*"AL—9' . -—-——%*’ Products so so ‘16) CO: 0 N —9——> .— —————>— Products 34% (E; (17) A nitrogen analog, cyano-migration, has been observed in one instance in a novel 1,4—shift. A 1,2~cyano radical shift has not been reported to date. CH3 ;____,. Products (18) Whether 1,2-radical shifts of vinyl, acyl, or cyano proceed by fragmentation and recombination, a three—centered delocalized transition state, or a three-membered intermediate, is open to debate. These mechanisms have been recently described by Reusch and co—workers (17). Although more evidence exists for 1,2-vinyl shifts than 1,2—acyl migrations, the generality of neither has been shown. Information pertaining to acyl shifts is particularly lacking. It is the purpose of this study, therefore, to attempt to induce and observe a radical 1,2-acyl shift in solution at ordinary temperatures. Information regarding competition be— tween 1,2—phenyl and 1,2—acyl shifts would also be considered valuable. RESULTS AND DISCUSS ION RESULTS AND DISCUSSION I. Introduction The radical (II) selected for this study is of interest because it incorporates in one substrate the possibility of 1,2- acyl, methyl, or phenyl shifts; intramolecular migration 0 (a g I CH3" "C':—CH2° CH3 II of any one of these groups would convert this primary radical (II) to a thermodynamically more stable benzylic and/or ter- tiary radical. An additional feature is the simplicity of this radical system which permits the synthesis of expected reaction products for comparative purposes. The t-butyl perester, t—butyl 5-methyl-5—phenylperlevulin— ate (III), was selected as the precursor for the above radical O ///;x/ \\0 £3 (3 III system. These peresters are known to decarboxylate by radical 10 mechanisms at relatively low temperatures, and are generally easy to synthesize and store. p—Cymene and phenyl ether were employed as solvents, since both are nonpolar and should favor radical reactions over those proceeding by ionic intermediates. The use of p-cymene (which is difficult to purify) rather than cumene was required by the potential ambiguity in accounting for any demethylstyrene produced in the reaction. Since p-cymene is a much better source of hydrogen atoms for radical abstraction than is phenyl ether, the average lifetime of radical inter- mediates was assumed to be greater in the latter solvent. II. Syntheses A. The Starting Perester The synthesis of t—butyl 5-methyl-5-phenylperlevulinate (III) was finally achieved by employing the following route: O O O NaOH 1. NaH, DMSO ”b Tar" ”if 2. Brcsgccfi? WOOD-Rt e e IV V VI alc. KOH /\ 0 MO\ X 1. gsoacLPy o g o o 2. Me3COOH /U>(\COOH 9' III VII Certain aspects of this synthesis merit discussion. 11 In the formation of ester VI, addition of the carbanion to the ethyl bromoacetate, i.e., reverse addition, was necessary to avoid dimer formation and the resulting low yields of ester. Also, should ketone V be contaminated with unreacted methyl benzyl ketone (IV), which is likely since the two ketones azeotrope during distillation, the resulting ester mixture is separable only by preparative v.p.c. There- fore, it is necessary to recycle V at least once to achieve pure ketone and good purity of all succeeding products. A major problem was encountered in converting 5—methyl— 5—phenyllevulinic acid (VII) to the desired perester. The original plan involved synthesis of the acid chloride (VIII) followed by reaction with the salt of t-butyl hydroperoxide, or with tmbutyl hydroperoxide in pyridine, to form the ‘ ) ‘2 l \. O I O SOC12 0 Cl 3. //\ p M O I VII VIII III perester. However, treatment of acid VII with thionyl chlor- ide gave upon distillation an 82% yield of the corresponding enol lactone (IX). Even mild acid chloride syntheses (e.g., G O \ 0 IX treatment of the salt of the acid with oxalyl chloride) failed 12 due to enol lactone formation. Although conversion of levu- linic acid or the acid chloride to enol lactones is a known reaction, it was unexpectedly facile. An alternate route to the perester from the acid was therefore sought. -One such route has been suggested by Davies (19). Application of Brewster and Ciotti's esterifi— cation procedure (20) to the synthesis of alkylidene peroxy- esters (peresters) was successful (equation Q), and this method should be applicable to ordinary perester syntheses. Indeed, acid VII was converted directly to perester III in I Q Me2C(OOH)2 + 2p-N02¢000H flag/,9;— Me2C(OOCOOgN-p)g (21) 40% yield based on moles of starting acid. No other products were detected, and unreacted acid could be conveniently re— covered and recycled. The reaction probably proceeds via a carboxylic-sulphonic anhydride intermediate. B. The Reaction Products To aid in product identification, the products arising from solvent—hydrogen abstraction by the parent radical II, as well as by 1,2-acyl, phenyl, and methyl rearranged radicals, were independently synthesized. The acyl rearranged product, 4-phenylw2wpentanone (XI), was prepared by the procedure of Munch—Petersen (22); the phenyl rearranged product, 5-methyl- 4-phenyl-2-butanone (XII), was synthesized by catalytic hydrogenation of 5-methyl-4-phenyl-5-buten—2-one; the pro- cedure of Schultz et al. (25) was employed to prepare 5-phenyl—2—pentanone (XIII), the methyl rearranged product; and the unrearranged ketone, 5—methyl-5mphenyl—2-butanone (X), was synthesized by methylation of 5-phenyln2wbutanone using NaH in dimethyl sulfoxide (see Experimental). III. Products from the Thermal Decomposition A. Method of Analysis After the acids formed in the reaction were isolated from the reaction products, a quantitative determination of unreacted perester was achieved using the method of Silbert and Swern (24). This method quantitatively converts perester to the corresponding acid, simultaneously oxidizing iodide to iodine which can be titrated with sodium thiosulfate. This method employs a ferric ion catalyst and is believed to involve the following reactions: 14 Fe+++ + I9 ‘——————€>- Fe++ + I. Fe++ + RC02O j] > Fe+++ + RCOZG + «0—1— o—i— + H20 ————7\ H0 + + H0° H0° + 19 ————>v Hoe + I. ZI- 7P I2 When this procedure was tested with t-butyl 5-methyl-5-phenyl- perlevulinate, no products were isolated other than the cor— responding acid, which was recovered in 90% yield. Although titration values for known perester concentrations were up to 20% high when the acetic acid-chloroform solvent system used in the Silbert and Swern procedure was modified by adding p-cymene or phenyl ether, this would correspond to only about 1% error in product composition for the iodine concentrations encountered in this study. After the acid formed from the perester reduction was re— moved by extraction, the solvent in the organic product mix- ture was separated by distillation, the residue was chroma- tographed on neutral alumina, and the fractions analyzed by gas chromatography (see Experimental). B. Reaction Products 1. Unreacted Perester and Products from the t-Butoxy Radical When 10 mmoles of perester III was heated at 150-2OC in p—cymene or phenyl ether (0.25 M), approximately 90% of the 15 perester decomposed within 5 hours. Since only a small amount of 5-methyl—5—phenyllevulinic acid was formed in the reaction, decarboxylation was for all practical purposes concerted with the cleavage of the oxygen-oxygen bond. The products derived from the t-butoxy radical, t-butyl alcohol and acetone, were swept into cold traps and collected. The ratios of t—butyl alcohol to acetone were 9.0 and 0.6 for decomposition reactions in p—cymene and phenyl ether, respectively, and in effect, represent the solvent‘s hydrogen donating ability. 2. Carbon Dioxide Although the primary objective of this study was that of product analysis, certain experimental data relating to the rate of carbon dioxide evolution are worth mentioning. First, the rate of C02 evolution deviated from lst order kinetics in both solvents. This deviation from lst order kinetics, the ideal kinetic order for perester decompositions, is no doubt due to induced decomposition. The inducing radical Ri may be a t-butoxy radical from a unimolecular cleavage or from a previous induced decomposition, or a solvent radical. The rate of carbon dioxide evolution was roughly twice as fast I? I? R RCOO'-‘—-+—' + R‘°"——€> R$O° + R°O_‘+—- during the first half of the reaction (0-90 min.) than during the second half (see Figure I). The magnitude of the apparent 16 I I 1* "I I I T I I —V ' ' 0.90" “ 0.80‘i A/5L / r- /@/ ' / 0.707” )9 I @/ IQ’I *_ (3.60“ I, ' d’ - /@ » .2 ’ 3 [,6] I 0.50" 6 E' " :1 ’m’ E 1- 123/ "‘ l 0.40.-. ® EV, — H I H 1’ B. , .773 j 8‘ H 0.30-— (9 E - _ [E _ O 0.20-L E1 4 (D Cymene — 4 E E Phenyl ether 0.10-- @ _ El 0" I l 1 1 I I I I 1 ‘ ‘_' I I I j l I l I l l r I 50 60 90 120 150 180 Time in Minutes Figure 1. Plot of log [III]/[III]-[C02‘f] versus time. 17 lst order rate constant for the last half of the reaction was 4 X 10-5 sec?1 in phenyl ether, and 5 x 10-5 sec._1 in p-cymeneo This is about 6-8 times slower than either the rate of lst order decomposition of t—butyl 3-phenylperpro- panonate tC) ¢CH2CH2- or t-butyl 4-phenylperbutanoate to ¢CH2CH2CH2° under comparable reaction conditions (25). These latter reactions are believed to be devoid of induced decomposition under the conditions studied. The slower rate and lower yields of C02 for the decomposition of perester III in phenyl ether might be due to the fact that the more stable cymyl radical participates more effectively in induced de- composition. It should be noted that induced decomposition can be minimized by working in very dilute solutions and/or using inhibitors; these techniques were not employed in this reaction product study. An induction period is apparent in Figure I, and is probably due to traces of oxygen in the perester and the sol- vent before addition to the reaction vessel. 5. Cage Products The major product from the perester decomposition in both p-cymene and phenyl ether was, surprisingly, the t—butyl ether XIV, 3-phenyl—5,6,6—trimethylm5-oxa-2-heptanone. Identification of this compound was based on infrared (Figure 23) and n.m.r. (Figure 24) spectra, and carbon—hydrogen analysis (see Experi— mental)o A similar recombination has been observed by Bartlett 18 o 0% XIV and Lorand (26) in the thermal decomposition of t-butyl tri- phenylperacetate. The manner in which the t-butyl ether is generated is suggested by a number of related factso First, the yields of ether XIV was the same in both solvents (see Table I). Second, little or no combination between the t-butoxy radical and a rearranged substrate radical was directly observed; however, part of products XVI and XVII may have evolved from loss of t—butyl alcohol from the acyl rearranged product (XV). Thirdly, no 3—methy1-5—phenyl-2—butanone (X) was detected in the reaction products in either solvent. 0 S //u\\;>E;>O 24~26% 25% O x: ./fi\v/¥E§ 15—14% 14% O XVI /)k<¢l§j 9% 7% O XVII /JKVJLQ§ 8% 9% XVIII 8-9% ___ xx 18% 8% XXI 20% 5% a fl bicymyls 15% ___ i—propenyltoluene 12% --- Q-oa— Me , mm 22-25% Table Io Products after three hours at 150-2OC (0.25M)° 2O radicals. This we know from the constant yield in the two 0 R/U\ o R\ T o/ Mk0 + C02 ’Ar//’ solvents, the absence or small amount of rearranged recombi- nation products, and the absence of unrearranged R-H. If R~ and -OCMe3 had significant lifetimes in a solvent cage, re- arranged ether and solvent trapping (to R—H) would be expected. Furthermore, the absence of unrearranged R-H suggests that the rearrangement of radical II is very fast and possibly con- certed with decarboxylation. 4. Rearranged Products The yield of identified rearranged products ranged from 45% in phenyl ether, to about 77% in p-cymene, based on moles of starting material. The rearranged products included mono- mers, solvent containing products, and dimers. Monomeric rearranged products were produced in yields of about 30% (based on moles of starting perester) for reactions in both cymene and phenyl ether. All three monomeric products are derived from the acyl rearranged radical, and were identi- fied as 4—pheny1~2—pentanone (XI), and the two disproportionafi tion products, 4-phenyl-3-penten-2—one (XVI), and 4-phenyl—4~ penten—Z-one (XVII). Identification of XI Was based on com— parison with an independently synthesized sampfle; the n.m.r. and infrared spectra, v.p.c. and t.l.c. retention times of the two‘ samples were identical. It should be noted that the phenyl re— arranged product XII possessed v.p.c. and t.l.c. retention 21 times identical to the acyl rearranged ketone; however, n.m.r. analysis of known samples was able to detect the presence of as little as 6% XII in XI. Therefore, any phenyl rearranged product was present in less than 1% yield and was considered negligible. The disproportionation products XVI and XVII were identi— fied from their infrared (Figures 25 and 27) and n.m.r. (Figures 26 and 28) spectra, which were consistent with the assigned structures (see Experimental). The approximately equal yields of each disproportionation product is typical of radical reactions; however, the identical yields of XI in both cymene and phenyl ether is atypical. The yield of XI should be lower in phenyl ether, and in fact only 6% of XI was directly isolated from the initial phenyl ether distil— late; the remaining 8% (Table I) was found in the distillation residue and may arise from decomposition of a dimeric, poly- meric, or solvent containing intermediate. Capture of the acyl rearranged radical by the solvent was also evident. In p-cymene, the trapped rearranged radical amounted to 8-9% yield, based on moles of starting material. Interestingly, the rearranged radical combined with the methyl end of the cymene rather than the isopropyl end. No other p-cymene substituted ketones were isolated. Thus, there 1m XVIII 22 apparently is sufficient steric hindrance between the gem- dimethyl groups of cymene, and the phenyl and methyl groups of the rearranged radical to prevent such a combination. Identification of compound XVIII was based on the infrared (Figure 29) and n.m.r. (Figure 50) spectra. Although no corresponding products were identified from the reaction in phenyl ether, one high molecular weight product, as yet uni- dentified, was isolated. The n.m.r. spectrum of this material possesses an aromatic hydrogen to methyl hydrogen ratio sug— gesting a solvent trapped radical. Dimers of the rearranged radical were of sufficiently high molecular weight to make detection by v.p.c. difficult. However, during column chromatography on neutral alumina, these dimers condensed to form the cyclopentene ketones, gig: and trans—1-acety1 2,4,5-trimethyl-4,S-diphenylcyclopentene, XX and XXI, which were easily detected and collected as colorless, O 0 /U\ U W alumna I - a O XIX XXI viscous oils by gas chromatography. Identification of com- pounds XX and XXI rested on infrared (Figures 31 and 55) and n.m.r. (Figures 52 and 54) Spectra, and carbon—hydrogen analyses. In the infrared spectra, both compounds exhibited 1 absorptions at 1670 cm“ , characteristic of d-8 unsaturated ketones. The n.m.r. spectrum of XX was similar to the Spectrum 25 of XXI with respect to the number and kinds of resonances, but dissimilar with respect to certain chemical shifts. In compound XX, the methyl and phenyl groups are eclipsed, with the planes of the phenyl groups facing each other for least steric crowding. Consequently, shielding effects associated with the aromatic rings should be very evident in the phenyl resonances in compound XX, but not in compound XXI; whereas, the methyls at C-4 and C-5 should be shielded in XXI and not in XX. Since the chemical shifts of the phenyls are moved from T 2.82 in XXI to T 5.16 in XX, while the chemi— cal shifts of the methyls at C-4 and C-5 are moved from T 8.82 and T 8.98 in XXI to I 8.55 and T 8.52 in compound XX, the assignment of stereochemistry to these compounds seems to be sound. The yields of rearranged dimers increased from 15% in phenyl ether to 58% in p-cymene, based on moles of starting perester. The yield in cymene is questionable, however, as it is unreasonably high. Furthermore, the yield of dimers should be lower in the better hydrogen donating solvent. Values closer to that for the reaction in phenyl ether would also correct the error in mass balance between carbon dioxide lost and carbon dioxide-free products. 24 5. Solvent Products Products derived primarily from solvent radicals were isolated from perester thermal decomposition in p-cymene and phenyl ether (see Table I). In the case of phenyl ether, these were the o-, m—, and p—methylated phenyl ethers; the methyl radicals no doubt originating in the decomposition of t-butoxy radicals. The total yield of methylated solvent was 22—25%, based on moles of starting perester. The reactions in pucymene yielded two types of solvent derived products, the disproportionation product, p-isopro— penyltoluene, and several dimers. Two of the dimers have been identified as compounds XXII and XXIII. The total yield of bicymyls was 16%, based on moles of starting perester. XXII XXIII IV. Mechanism of Rearrangement From the types and yields of products isolated, there is little doubt that this reaction proceeds by a free radical mechanism, and an explanation of the facile 1,2~acyl shift and the absence of phenyl migration certainly merits discus- sion. 25 At least three different modes of radical rearrangement can be formulated for the radical (II) under study. I CH3C=O " CH3‘ , glc=C\H A CH3 CH3\ //0 CH3 C=O /C\ (3:0 I o J / ° \\ C . é - - - -—\ I “—4 H c— —H CH3 CCH CH3\C__-_E:/H F 3| I {a H i 91/ \ H a H B CH3 0' II \c/ XXIV CH3\ / \ /H c—c / \ Q H C J Mode A is a two step process consisting of fragmentation to an acyl radical and an olefin followed by recombination to give the rearranged radical XXIV. Mode B is a single step, intramolecular reaction, characterized by delocalization of the unpaired electron over at least three centers in the transition state. Finally, mode C proceeds in two steps through a cyclopropoxy radical intermediate (17). Although the main objective of this study was that of product identification, some of the observations are signifi— cant to the question of the rearrangement mechanism. First, no more than trace amounts of dsmethylstyrene could have been present in the products from reactions in cymene or phenyl ether according to v.p.c. analysis. If mode A were operative, a certain amount of olefin might be expected to escape 26 recombination and should be detected in some form in the products. Second, although no experimental results in this study pertain to mode B, all known attempts to identify any "non- classical" type radical have failed (1). Third, the rearrangement of II to XXIV via mode C would theoretically be reversible, since the cyclopropoxy inter- mediate can cleave between C-1 and C-2 or between C-1 and C-5, and radical XXIV could reform the cyclopropoxy inter- mediate. Experimentally, this reaction must be essentially irreversible, since no unrearranged product from II could be detected. Irreversibility may be due to the extreme thermo- dynamic stability of the tertiary benzylic radical XXIV, particularly in relation to radical II. Furthermore, the radical derived from phenyl migration would be less stable than XXIV, which might account for the absence of phenyl rearranged products. V. Conclusion It is apparent from this study that radical 1,2macyl shifts do occur in carbon free radical systems; and, where stabilization of the rearranged radical is extensive, can compete to the exclusion of phenyl rearrangement. In the case of t—butyl 5-methyl-5-phenylperlevulinate, decarboxylation appears to be essentially concerted with the cleavage of the oxygen—oxygen bond. The resulting recombination of substrate and t-butoxy radicals within the solvent cage must also be 27 essentially concerted, or must involve strong interactions between the radicals. Furthermore, the absence of unre- arranged 5-methy1-5-phenyl-2-butanone suggests that the re— arrangement of radical II is very fast, and possibly concerted with decarboxylation. EXPER IMENTAL 28 EXPERIMENTAL I. General Spectra The infrared spectra were obtained on a Perkin—Elmer 257B grating spectrophotometer. The ultraviolet spectra were obtained on a Beckman DB spectrophotometer. Proton magnetic resonance spectra were measured on a Varian Model A-60, high resolution spectrometer. All n.m.r. spectra were obtained at 60 Mc in carbon tetrachloride, unless otherwise stated, using tetramethylsilane as an internal standard. The relative peak areas were determined by electronic integration. Gas Chromatography Vapor phase chromatographic analyses were performed with an Aerograph A-90-5P gas chromatograph. Preparative work was accomplished with either an Aerograph A-QO-P or a Perkin—Elmer Model 154 instrument. Relative peak areas were determined by the triangulation method. Microanalysis All microanalytical data were obtained from the Spang Microanalytical Laboratory, Ann Arbor, Michigan. 29 50 Melting Points Melting points were determined on a Reichert hot—stage and are uncorrected. All temperatures are recorded in Centi- grade degrees. II. Syntheses A. Preparation of t-Butyl 5—Methyl- 5-phenylperlevulinate 1. Preparation of 5—Phenyl-2—butanone Employing the procedure of Schultz et al. (25), 15.4 g. (0.10 mole) of phenyl—2-propanone (methyl benzyl ketone) was added in one portion to 4.4 g. (0.11 mole) of powdered NaOH. The mixture was stirred vigorously for 15 minutes without cooling. During this time the mixture became orange colored and the NaOH partially dissolved. Methyl iodide (17.0 g., 0.12 mole) was added drop—wise over one hour, keeping the temperature of the mixture under 550. The mixture was stirred an additional 2 hours at room temperature, then refluxed gently for 5 hours on a steam bath. Cold water (25 ml) was then added with stirring. After cooling, the mixture was ex- tracted with ether and the ether washed with water. After drying over anhydrous sodium sulfate, the solvent was evapOr rated and the remaining oil was distilled employing a 6 inch vacuum jacketed vigreux column. The product, a colorless oil, b.p. 57-420/0.5 mm, was obtained in 70-80% yield. 51 The infrared spectrum is shown in Figure 2 and the n.m.r. Spectrum,T 8.68 (doublet, J'N’7 cps, 5H), T 8.11 (singlet, 5H), I 6.50 (quartet, J/b/7 cps, 1H), T 2.79 (sing- let, SH), in Figure 5. A 2,4-DNP was prepared and crystal- lized as orange—yellow, flat needles, m.p. 172.5—175.5O (reported m.p. 170—10 (27)); a semicarbazone melted at 172-5O (reported m.p. 172Q (28)). 2. Preparation of Ethyl 5-Methyl-5-phenyl- levulinate To a nitrogen-flushed round bottom flask containing 4.45 g. (0.050 mole) of pure 5~phenyl—2-butanone in 100 ml of dry dimethyl sulfoxide, was added 0.80 g. (0.055 mole) of NaH. The mixture was magnetically stirred and heated to 60-650 for about one hour, or until remaining NaH appeared to have stopped reacting. After cooling to room temperature, the yellow—brown mixture was forced under positive nitrogen pressure through a glass delivery tube into an addition flask equipped with a pressure equalizing side arm. The filled addition flask was attached to a second nitrogen-flushed flask containing 5.51 g. (0.055 mole) of ethyl bromoacetate in 100 ml of dry di— methyl sulfoxide. Stirring was commenced and the carbanion solution was added dropwise over one hour keeping the bromo- acetate solution at room temperature with a cold water bath. After addition was complete, the pale, dirty yellow solution was stirred at room temperature an additional one-half hour before being poured into an ice-water mixture. After the 52 .,EHHmV mcocmusfllmlamcmzmlm Mo Esnuommm commumcH .mm musmflm AHIEUV mocmsvmum oomfi OOON 00mm ooom 00mm _ _ fl 4 _ (queoied) eoueqqrmsuexm .AEHHMV mcocmusfllmlamcmnmlm mo Esuuommm UmnmumcH .QN musmflm AHIEUV hocmdqmnm oom 000a .Ooma oowfi oomfi oomfi _ m . .7 x . .1 L ON 6% Om Om (queoxed) eoueanMSUEIL HO.OH 54 .mGOGMpsnlmlawcmsmlm Mo Esupowmm monocommu ospmamms HmmHosz .m mnsmflm PO.m Po.© PO.¢ b.O.N 55 water mixture was extracted with ether, the extracts were washed with water, dried over anhydrous sodium sulfate and evaporated, leaving a bright yellow oil. This was distilled through a 6 inch vacuum jacketed vigreux column, and the b.p. 90-1100/0.5 mm fraction was shown by v.p.c. analysis to contain a 65-70% yield of the desired product. Final purifi- cation was achieved by distillation with a spinning band column. The infrared spectrum (Figure 4) and the n.m.r. spectrum (Figure 5), T 8.95 (triplet, J ~/7 cps, 5H), T 8.50 (singlet, 5H), T 8.10 (singlet, 5H), T 7.10 (center of AB quartet, J’V’16 cps, 2H), T 6.02 (quartet, J’V’7 cps, 2H), T 2.72 (singlet, 5H), were consistent with the assigned structure. 5. Preparation of 5-Methyl-5—phenyl— levulinic Acid The purified ester was refluxed for 24 hours with an equal weight of KOH dissolved in a ten—fold volume of 95% ethanol. After cooling, water was added and the alcohol was removed by a rotary evaporator. The aqueous solution was washed several times with ether, then acidified with aqueous HCl. Ether ex- tracts of the acidic solution were dried over anhydrous sodium sulfate, and evaporated. The remaining yellow oil solidified upon standing, and was crystallized from a large volume of hot water in long, white needles, m.p. 96.5—97.00. The yield for the saponification and crystallization was ca. 80%. .AEHHMV muMCHH5>mHahcwcmnmlahzumfilm H%£um mo Esuuommm UmumumcH .mw musmflm AHIEUV mocwsvmum coma OOON 00mm ooom 00mm ‘ 1 _ a 56 ON 0% om Om (queoxed) eoueqatmsuexm .AEHHMV mpMCHH5>waawcmflmIMIawsumalm H>£pm mo Eduuommm UmnmumcH .Qw musmflm AHIEUV hocmsqmum oow 000a ODNfi oo¢d oowfl coma — u q 4 a a 57 f (queoxed) aoueqqrmsuex; .mumcaas>maamcwnmlmlahnumfilm ahflum mo Esuuommw monocommu Uflumcmmfi Hmmaosz .m musmflm . . 90. P0.0;“ b.O.m b.O w Po w h , _ 5L , ill. 58 59 Re-esterification of the acid gave the starting ester according to v.p.c. analysis. The acid exhibited broad infra— red absorptions (Figure 6) at 5400 and 1710 cm‘l. A deutero— chloroform solution of the acid gave the following n.m.r. spectrum (Figure 7): T 8.28 (singlet, 5H), T 8.12 (singlet, 5H), T 7.07 (center of AB quartet, J«~/16 cps, 2H), T 2.72 (singlet, 5H), and T 0.54 (singlet, 1H). Anal, Calcd for C12H1403: C, 69.88; H, 6.84 Found: C, 69.48; H, 6.80. 4. Preparation of the Enol Lactone, 5—Methyl— 4-methyleno-5-phenyj—y-butyrolactone To 5.0 g. (0.04 mole) of thionyl chloride was added with stirring 2.52 g. (0.011 mole) of 5—methyl—5—phenyllevulinic acid in several portions. No heat evolved and the acid quickly dissolved. Stirring was stopped and the mixture was allowed to stand for 17 hours at room temperature. After gently reflux- ing for one-half hour, the thionyl chloride was removed under reduced pressure. Distillation of the remaining yellow oil gave 1.74 g. (82% yield) of a colorless oil, b.p. 122.5-125.50/ 1.8 mm. The product was pure by v.p.c. analysis, and exhibited . . EtOH ultraVIolet absorptions at kmax 252,258, 264, and 277 mu (all e < 200); infrared absorptions (Figure 8) at 1810 (broad) and 1665 cm—1; and n.m.r. resonances (Figure 9) at T 8.46 (singlet, 5H), I 7.22 (center of AB quartet, szr17 cps, 2H), T 5.74 (doublet, JrV’5 cps, 1H), T 5.18 (doublet, J’V’5 cps, 1H), and T 2.65 (multiplet, 5H). A parent peak of 188 m/e was evident in the mass spectrum. 40 Coma .Awaouv UHUM UHCHH5>®HahcmchMIamnumalm mo Esuuommm UmumumcH AHIEUV mocmdvmum .mm musmflm 00mm OOON 00mm 000m A ._ _ (queoled) eoueqqrmsuelm 41 .Aaauov pflom oflcflas>maahcmnmlmuamnumEIM mo Esuuommm pmumumcH .Qw musmflm AHIEUV hocmsqmnm oom OOOH coma oowd coma Coma _ _ _ _ ._ 3 _ ION 10¢ lém 10m (queoled) eoueqqrmsuexm .UHUM UHQHH5>mHahcmsmlm1H%£umE|m mo Esuuommm mUQMGOmmH Uflumcmmfi ummaosz .s musmflm b.o.0fi Po.m Ho.m 90.4 PO.N 42 .AEHHMV mcouomaouhuSQlalahcmnmlmlocwamnme|¢lahswaIm mo Esspowmm UmnmnmcH .mm musmfl AHIEUV hocwsqmnm oomfi OOON OOMN 000m 00mm . _ _ _ _ 45 m .ON Om om (queoxed) eoueqqrmsuexm 44 .AEHHMV mCOpUmHouhuSQlalahcmnmlmnocmahnumEstamnumEIm mo Esnuommm UmnmumcH 00m 0005 OONH AHIEUV mucmsvmum OO¢H coma .Qm onsmflm oomfi _ fl JON ow [Om 10m (queoxed) eoueqqrmsuex; P0.0fi 45 .mQOHUMHonmpSQ a HthSQIMIOamahnuwfilfilamnumfilm mo Esuuommm mUQMGOmmM oflumqmmfi Hmmaosz . so.4 eo.m Po m .m musmflm PO. 46 5. Preparation of t-Butyl 5-Methyl— 5ephenylperlevulinate Employing the method of ester synthesis of Brewster and Ciotti, Jr. (20), 1.91 g. (10 mmole) of p—toluenesulfonyl chloride was added to 1.05 g. (5 mmole) of 5—methyl-5-phenyl- levulinic acid in 45 ml of dry pyridine. The mixture was stirred at room temperature for one hour, then cooled to 00 with an ice bath. Freshly distilled t-butyl hydroperoxide (0.42 g., 4.7 mmole) was added over a 5 minute period. The mixture was stirred for two and one—half hours keeping the temperature under 50. At the end of this time, the reaction mixture was poured into an ice—water mixture. After the ice melted, the mixture was extracted with ether and the solvents removed on a rotary evaporator, keeping the bath temperature below 400. The mixture was taken up in ether, washed with 10% sulfuric acid, 10% sodium carbonate, and then water. Drying over anhydrous sodium sulfate and removal of the ether left a colorless oil; traces of solvent were removed under vacuum. Small amounts of unreacted t-butyl hydroperoxide, if present, can be removed by repeating the washing procedure, or by chromatographing the perester on a Florisil column. Thin layer chromatography on silica gel gave only one spot for the pure perester after treatment with an alcoholic anisalde- hyde-H2804 developer. The perlevulinate is a viscous oil, giving typical t—butyl, ketone, and perester absorptions in the infrared (Figure 10) at 2950, 1775, 1710, 1577, 1561, and 850 cm71, and exhibiting an n.m.r. spectrum (Figure 11) 47 .AEHHMV mumcflas>maummawcmchMIahnumEIMIampsnlu we Esupommm UmumnmaH AAIEUV hocmdqwnm .moa musmflm 00mm Coma OOON 00mm 000m _ _ _ _ (queoxed) eoueqqrmsuexm 48 .AEHHMV mumcflas>maummahcmzmIMIH>£umEIMIH>uUQIu mo Esuuommm UmumnwcH com 000a ,a coma . IEUV mucmsvmum oowfi coma .noa magmas oomd _ ON 0% IOm om (queolad) eoueinmsuexl .mumcflafi>maumm Iahcmnmlmlawnpmfilmuahusnlu mo Esnpummm mocmcommu oflpmcmmfi Ammaosz .Hfi musmflm b.O.o._n PO.m b.06 HO.¢ HO.N . 1 _ _ _ 1111111 in ,gflliiiiifii 49 50 consistent with the assigned structure, I 8.90 (singlet, 9H), T 8.25 (singlet, 5H), T 8.10 (singlet, 5H), T 7.16 (center of AB quartet, Jrv 15 cps, 2H), T 2.74 (singlet, 5H). Anal. Calcd for C15H2204: C, 69.04; H, 7.97 Found: C, 68.78; H, 7.94. B. Preparation of Ketone Products Expected from Thermal Decomposition 1. Preparation of 4-Phenyl-24pentanone Methyl magnesium bromide (ca. 0.5 mole) was prepared from 6.50 g. (0.26 g.-atom) Mg turnings and 52 g. (0.54 mole) of methyl bromide in 160 ml of anhydrous ether by the method of J. Colonge and R. Marey (29). All the Mg was consumed and the blackish solution was refluxed 15 minutes to drive off any unreacted methyl bromide. Employing the procedure of Munch—Petersen (22), the Grig— nard reagent was cooled to —100 with a dry ice-carbon tetra- chloride cooling bath, and 0.25 g. of cuprous chloride added in one portion. The mixture was stirred and further cooled to -150. Ethyl cinnamate (17.4 g., 0.10 mole) in 100 ml of ether was added dropwise over one and one—half hours, keeping the temperature between -80 and —150. The cooling bath was then removed, and the reaction mixture was stirred for 2 hours at room temperature. The reaction mixture was poured into an ice—HCl mixture and extracted with ether. The ethereal solu- tion was washed with a sodium bicarbonate solution, and water, dried over anhydrous magnesium sulfate, and evaporated. 51 Distillation of the residue through a 6 inch vacuum jacketed vigreux column gave a colorless oil, b.p. 65-680/0.5 mm. Vapor phase chromatographic analysis of the distillate indi- cates a yield of ca. 60%. A sample of the product collected by preparative gas chromatography gave an infrared absorption at 1715 cm"1 (Figure 12), and n.m.r. resonances (Figure 15) at I 8.78 (doublet, J’V’6 cps, 5H), T 8.07 (singlet, 5H), T 7.42 (center of multiplet, 2H), T 6.8 (center of multiplet, 1H), and T 2.87 (singlet, 5H). A semicarbazone was prepared which melted at 154—6O after one crystallization (reported m.p. 137° (30)). 2. Preparation of 5-Methyl-4—phenyl-2— butanone a) Preparation of 5—Methyl—4—phenyl-5-buten-2-one Employing the procedure of Metayer and Epinay (51), benzaldehyde, methyl ethyl ketone, and concentrated HCl were refluxed together for 4 hours. After taking up the organic layer in benzene, washing with water, drying, and evaporating the benzene, the tarry residue was distilled under reduced pressure. The meager distillate, a yellow-orange oil, b.p. 96-1500/0.8 mm, contained several components according to v.p.c. analysis. The mixture was effectively separated by column chromatography on silica gel. The desired ketone ex- hibited infrared absorptions (Figure 14) at 1660, 1625, 1605, 1465, 1580, and 1560 cm‘l, and n.m.r. resonances (Figure 15) at T 7.89 (doublet, J’N’2, 5H), T 7.68 (singlet, 5H), and 52 oomfi .Awaoov mcocmucmmlmlawcmfimle mo Esuuommm UmnmnwcH ooom a AHIEUV muamdqmum comm _ (K ooom __ .mmfi musmflm 00mm ON 64 Om Om (queoxed) eoueqqtmsuexm .Aaaoov mcocmucmmlmlahcmzmla mo Esuuommm cmHMHMCH .Qma musmam AHIEUV hocmsqmum com coca coma ooaa coma coca _ _ _ _ c _ > 1.0m 1.04 5 5 1 .!O@ V C H om (queoled) eoueIIImsuexl 54 .mcocmucmmlmlahcmgmla mo Esupommm mUCMCOmmH oaumcmmfi ummaosz 44.8 s£.m 40.4 .ma madmam eo.m 55 .AEHHMV mCOINIcmpsalmlahcmcmnalahnumfilm mo Esuuommm UGHMHMGH coma “HIEUV wocmsqmum ooom comm ooom _ _ _ comm .maa madmam low low (queoxed) eouequMSUEIL 56 .AEHawV mCOINIc®DSQImlamcm£QIanahcumalm mo Esnuommm cmummmcH occ coca AHIEUV hocwsqmum coma ooaa coca .Qaa magmas coca _ : (queoxed) eoueqqrmsuexl H0.oa 57 .mso:Nicousnlmlahcmzmlalawnumelm mo Esnpomcm wocmcommu oapmcmme Homausz P . . o c e0 0 90.6 .ma musmam b.0.N 58 singlets at T 2.70 and T 2.58 (6H). The oxime was prepared which melted at 102.5-105O (reported m.p. 1040 (52)). b) Preparation of 5-Methyl-4—phenyl—2—butanone The above butenone was dissolved in absolute ethanol and catalytically hydrogenated with 10% palladium on charcoal at 40 psi. hydrogen and room temperature. Removal of the alco- hol gave a yellow oil which, according to gas chromatographic analysis, was solely the desired product. The compound pos- sessed an infrared spectrum (Figure 16) absorption at 1708 cm'l, and n.m.r. spectrum (Figure 17), T 8.95 (doublet, J/V’6 cps, 5H), T 8.01 (singlet, 5H), T 7.5 (center of multiplet, 5H), and T 2.88 (singlet, 5H), consistent with the assigned structure. 5. Preparation of 5-Phenyl-2:pentanone Employing the procedure of Schultz et al. (25), 15.1 g. (0.10 mole) of phenyl—2—propanone (methyl benzyl ketone) was added in one portion to 4.4 g. (0.11 mole) of powdered NaOH. The mixture was stirred vigorously for 15 minutes without cooling. During this time the mixture became yellow colored and the NaOH partially dissolved. Ethyl iodide (18.8 g., 0.12 mole) was added dropwise over one hour, keeping the temperature under 550. The mixture was stirred an additional 2 hours at room temperature, then refluxed gently for two and one—half hours on a steam bath. Cold water (25 ml) was then added with stirring. After cooling, the mixture was ex— tracted with ether and the ether washed with water. After u. 59 coma .AEHHMV mcocmpsalmlahcmcmlalahnumfilm mo Ednuommm commumcH coom AHIEUV mocmdvmum comm cocm .mma musmam comm 0N ca 00 cc (queoisd) eoueIIImSUEIL 60 com .Afiaamv mcocmusfllmlawcmcmlalH%£pmfilm MO Eduuommm cmumumcH 000a A. IEUV mocmsvmum coma ccaa come, .93 853m .,oome .om ca 00 .00 (queoxed) eoueqqrmsuexl 90.0a 61 .macawusnlmlamcmnmlalamcumel <é.<<$\ €<. . ..> . \- 90.0 ixgxig. .\ I m mo Eduuowmm mocmcommn Uaumcmmfi Hmmaosz .sa musmam 9o.m 9 . c a 9o.m _ . ..P . ..,,.. . . .L.\/-3_./>>./. .e>\/)>> . . 258?. fiS€§éfl$£§§}}7?f$$kés a?€§1ff?§33 62 drying over anhydrous sodium sulfate, the solvent was evapo— rated and the remaining oil was distilled employing a 6 inch vacuum jacketed column. The product, a colorless oil, b.p. 48—520/0.4 mm, was obtained in 56% yield. The infrared spectrum is shown in Figure 18 and the n.m.r. spectrum, I 9.22 (triplet, J'v'7 cps, 5H), T 8.2 (center of multiplet, 5H), T 6.56 (triplet, Jrv'7cps,1H), and T 2.89 (singlet, 5H), in Figure 19. A semicarbazone was prepared which melted at 187—1900 after one crystallization (reported m.p. 190-10 (23)). 4. Preparation of 5-Methyl-5-phenyl—2-butanone To a stirred solution of 1.75 g. (0.052 mole) of sodium methoxide in 50 ml of dry dimethyl sulfoxide, was added 4.10 g. (0.028 mole) of 5-phenyl-2-butanone. The solution promptly turned a brown—black color. After cooling in an ice bath, 5.70 g. (0.040 mole) of methyl iodide was added dropwise; the dark color faded to a dirty yellow in the process. The mix- ture was stirred at room temperature for one-half hour, then gently heated on a steam bath about one hour. After stirring at room temperature for another 2 hours, the mixture was cooled, and an additional 4.6 g. (0.052 mole) of methyl iodide was added. The mixture was heated for one-half hour on a steam bath, cooled, water added, and extracted with ether. The extracts were washed with water, dried, and the solvent evaporated. The remaining oil was distilled employing a 6 inch 65 coma .AEHHMV mcocmecmmlmlawcmcmlm mo Essuommm commumca ooom AHIEUV hocmsqmum comm occm .wma musmam comm _ I 0N .ca 00 cc (queoxed) eoueqqrmsuexm 64 .AEHamv maccmucmmlmlahcmflmlm mo Esuuowmm commumcH AHIEUV mocmsgmum com 000a coma ocaa coma .nma magmas coma _ _ _ _ ._ W. .cm row 00 cm (queoied) eoueqqrmsuexm 90.0a 65 .mGOGMDGmmINIamcocmnm mo Hespommm mocmcomws oaumcmme Hmmaosz 9c.m 9c.c 9c.a .ma mnsmam 9o.m 66 coma .AEHawv ®COCMDSQINlahcmnmlmiahzumalm mo Esuuommm UmumumcH 000m AHIEUV wocmsqmum 00mm occm .mcm musmam 00mm _ _ .cm .cw 0m .0m (queoled) eouequMSUEIL 67 .AEHamv maocmpsalmlawcmnmlmlamnumEIm mo Esubommm cmumumcH .nom mucosa AHIEUV %ocmsqmum owc coma coma ocda coma coma necm n:0¢ .100 2 fII .Iom (queoxed) eoueqqrmsuexl .mCOGMDDQINIamcmSQImla%£meIm mo Esuuummm monocommn oaumsmme Homaosz .am mnsmam 90.0a 90.@ 90-0 9o.a 9c.N 10):!) 68 69 vacuum jacketed column. The product, a colorless oil, b.p. 90-920/7.8 mm, was obtained in ca. 45% yield. The infrared spectrum of a sample collected by gas chromatography is shown in Figure 20, and in Figure 21, the n.m.r. spectrum: I 8.62 (singlet, 6H), T 8.22 (singlet, 5H), and T 2.78 (singlet, 5H). A 2,4-DNP was prepared and crystal— lized as golden crystals, m.p. 150.5-151O (reported m.p. 151-20 (55)). III. Thermal Decomposition Procedure A. In p—Cymene 1. Purification of p-Cymene Reagent-grade "terpene—free" p-cymene was contaminated with a sufficient quantity of terpenes to complicate perester thermal decomposition analyses. Consequently, the p—cymene was purified by the procedure of Catherine G. LeFevre et al. (54). After refluxing over sulfur powder for 2 days, p—cymene was extracted with concentrated H2804, and washed with dilute NaOH followed by a saturated NaCl solution. The dried cymene was then distilled on a spinning band column. The purified p—cymene contained no impurities detectable by v.p.c. analysis. 2. Reaction and Analysis Procedure The apparatus consisted of a 5-necked round bottom flask equipped with glass entrance and exit tubes (see Figure 22). 70 The exit tube was connected to two dry ice—isopropyl alcohol traps in series by short segments of tygon tubing. The cold traps in turn were attached with tygon tubing to two ascarite traps connected in parallel. These traps consisted of tared, ascarite-filled tygon tubing, and had short ascarite safety traps attached to prevent atmospheric carbon dioxide contami- nation. Screw—type pinch clamps preceded each ascarite trap. The entrance tube to the round bottom was connected to a source of dry, oxygen- and COg-free nitrogen. This was achieved by passing pre—purified nitrogen through scrubber units containing Fieser solution (2 units), a lead acetate solution, a barium hydroxide solution, concentrated H2804, and a drying tower containing anhydrous magnesium sulfate and KOH pellets. The reaction flask was immersed in an oil bath and magnetically stirred. \II ._, {3; Figure 22. Thermal decomposition reaction apparatus. 71 The cymene was placed in the flask, the nitrogen flow stabilized, and the oil bath heated to the desired tempera— ture. The ascarite traps were weighed; one trap was then pinched off. The perester in a small amount of solvent was added in one portion by injection through a serum cap, or for larger amounts, with a pipet or addition flask. The rate of C02 evolution was determined by weighing alternate ascarite traps at various time intervals. One trap remained connected to the system while the other was being weighed. The total C02 collected was determined from both traps. The reaction was terminated by removing the oil bath, and rapidly cooling the reaction vessel with dry ice. Ether was added, and the mixture was extracted with a 10% sodium carbonate solution to isolate any acids formed in the reaction. The ether was removed on a rotary evaporator. Unreacted perester in the residue was then analyzed by the method of Silbert and Swern (24). This procedure reduces the perester to the corresponding acid, simultaneously oxidiz- ing iodide to iodine which can be titrated. For perester con- centrations up to 5 mmoles, the residue was dissolved in 100 ml of chloroform, 20—55 ml of a saturated NaI solution was added, followed by 150 ml of .002% FeCl3°6H20 in glacial acetic acid. The mixture was allowed to stand 15-20 minutes, then stirred and rapidly titrated with a standard sodium thio— sulfate solution to the disappearance of yellow color. 72 After conversion of the perester to the acid, 50% NaOH was added (with cooling) to achieve a pH near 6. The mixture was then extracted with ether, and the ethereal solution washed with 5-10% KOH. The acids formed in the hydrolysis of unreacted perester were isolated by acidification of the base extracts followed by removal of solvents on a rotary evaporator. ffluaetherealsolution was washed with a saturated NaCl solution, dried over anhydrous sodium sulfate, and distilled. The cymene was removed from the neutral products by distil- lation through a spinning band column at reduced pressures. No reaction products distilled over with the cymene (v.p.c. analysis). After the residue was weighed and dissolved in a known volume of benzene, an aliquot was removed, and combined with an internal standard (trans-stilbene). This solution was then analyzed by v.p.c. employing a 20% SE—50 column. The remaining neutral material was chromatographed on Woelm neutral alumina, activity I. The fractions obtained were further separated and analyzed on the gas chromatograph. B. In Phenyl Ether 1. Purification of Phenyl Ether Phenyl ether was washed with 10% KOH, followed by a satu- rated NaCl solution, and dried over anhydrous sodium sulfate. After distillation,v.p.c. analysis showed no impurities. When heated to 2000 for 5 hours with nitrogen flushing, no volatiles were collected in dry ice—isopropanol traps. 75 2. Reaction and Analysis Procedure The apparatus and workup were the same as with cymene as a solvent. The spinning band distillate, however, contained ketone reaction products. Their concentration was determined by the addition of an internal standard (phenyl benzoate) and v.p.c. analysis. These products were separated by preparative gas chromatography, and identified by the usual techniques. The distillation residue was chromatographed on alumina, and the fractions were further separated and identified by v.p.c. analysis. Quantitative results were achieved by gas chromatography using trans—stilbeneemfian internal standard both before and after column chromatography. IV. Identification of Reaction Products A. Identification of 5-Phenyl-5,6,6- trimethyl—5-oxa—2-heptanone This t-butyl ether was eluted from an alumina chroma- tography column with mixtures of benzene and chloroform. The infrared spectrum of a pure sample (v.p.c.) is shown in Figure 25; n.m.r. resonances (Figure 24) were exhibited at T 8.87 (singlet, 9H), I 8.55 (singlet, 5H), T 8.10 (singlet, 5H), T 6.25 (center of AB quartet, J’V/8 cps, 2H), and T 2.77 (singlet, 5H). lAnal. Calcd. for C15H2202: C, 76.87; H, 9.42 Found: C, 77.14; H, 9.55. 74 .Afiaamv mQOCMmecINIMXOImlawnumEauulc.0.mlahcmnmlm mo Esnuommm consumcH coma 000m AHIEUV mocmdqmnm 00mm oocm QOMm .mmm musmam A» a _ (aueoxed) soueqqrmsuexl -7' 75 .AEHamv mcocmummclNIMXOImlahnumEauulm.m.mlamcmsmlm mo Esuuommm commumcH .Qmm musmam AHIEUV mocmdwmum 00m 000a coma ccaa coma ccma _ m a _ .r 0 10m 10w 10m . 10m (Iueoxed) eoueqirmsuexm .mcoaMpmmclm IMXOImlahflu®EaHulm.m.mlamcmsmlm mo Eduuommm mocmaommu oaumcmmfi Homaosz .am musmam 90.0a 9o.m 9o.m 9c.a 9o.m _ . _ 13 76 77 B. Identification of 4-Phenyl-5-penten-2-one A sample of this disproportionation product was collected by preparative gas chromatography and exhibited infrared ab- sorptions (Figure 25) at 1680 and 1600 cm'l, and n.m.r. resonances (Figure 26) at T 7.82 (singlet, 5H), T 7.52 (singlet, 5H), I 5.57 (singlet, 1H), and T 2.66 (singlet, 5H). These spectra are consistent with the assigned structure. C. Identification of 4-Phenyl—4-penten-2-one This nonconjugated disproportionation product was col- lected by preparative vapor phase chromatography and exhibited infrared absorptions at 1715 and 1600 cm":L (Figure 27), and n.m.r. resonances (Figure 28) at T 7.97 (singlet, 5H), I 6.55 (singlet, 2H), singlets at T 4.80 and T 4.45 (2H), and a singlet at T 2.69 (5H). These spectra are consistent with the assigned structure. D. Identification of 4-Methyl—4—phenyl-5- (p-cymyl)-2—pentanone This reaction product was separated from other ketone products by preparative gas chromatography. The infrared spectrum (Figure 29) showed carbonyl absorptions at 1720 and 1705 cm-1, and the n.m.r. spectrum (Figure 50) exhibited reson- ances at T 8.88 (singlet, 5H), T 8.77 (singlet, 5H), T 8.64 (singlet, 5HL T 8.25 (singlet, 5H), T 7.50 (center of AB quar- tet, er/15,cps, 2H), T 7.04 (singlet, 2H), T 5.19 (quartet, J'A/9, cps, 4H), T 2.80 (singlet, 5H). 78 coma .Avaoov mCOINIcmucmmlmlahcmnmla Mo Esuuummm cmumumcH cch 1. IEUV hucmsqmum comm ccom .mmm madman 00mm _ cm 0% 0m .0m (queoled) eoueqqrmsueil 79 com .Awaoov occumlcmucmmlmlawcmnmla mo Esupummm cmumumcH AHIEUV wocmsomum 000a coma ocaa coma .Qmm musmam coma a _ _ .cN ca .om om (queoled) eoueaitmsuexl .QQOINIQmpcmmlmlawcmcmla mo Esupommm mocmcommu oaumcmmfi Homaosz .mm musmam 90.0a 9o.m 9o.m 90.4 9o.m 80 81 coma .Awaoov occlmlcmucmmlalahcmnmla mo Esuuummw commumcH ccom AHIEOV hocmsqmum comm 000m .89N musmam 00mm T. .om ca cm om (queolad) eoueqqrmsuell 82 .Awaoov maoumlcmucmmlalamcmsmla mo Esuuommw cmumnmcH .Qbm musmam AHIEUV wocmsgmum 00m 000a coma ccaa coma coma 0 _ _ a c 3 .cm .oa om .cm (queoxed) eoueqqrmsuelm .maonmlcmpcmmlalahcmcmla mo Eduuommm moamcommu vapmcmme Homaosz .mm musmam 90.0a 9o.m 9o.m 9o.¢ 9o.m 85 84 .Aaauov mcocmncmmlmlAHNEDUIQVImlawcmsmlalahnpmfila mo Eduuommm commumcH .mmm musmam AHIEUV mocmsvmum coma 000m 00mm 000m 00mm _ A _ _ _ IoN 10a 10m 10m Ill/II! (queoxed) eoueqqrmsuexl 85 .Awauuv maocwpcmmlmlAawfisolmvImlawcmcmlalamnuofila mo Esuuommm commumcH .Qmm madmam AHIEUV mocmsgmum 00m 000a coma ccwa coma 00m _ a _ _ _ _ > 1... .ca 0m cm (aueoied) eoueqqrmsuexl .mco:89gmm INIAamasolmvImlamcmnmlalawnumela mo Ednuommm mocmcommn oapmcmme Homaosz .cm musmam 90.0a 9o.m 9o.m 90.¢ 9o.m 86 87 E. Identification of gig:1-Acetyl-2,4,5-trimethyl- 4,5-diphenylcyclopentene A pure sample of this product was collected by prepara- tive gas chromatography after having bean chromatographed on neutral alumina. The infrared and n.m.r. spectra, Figures 51 and 52, respectively, exhibited abSotptions at 1670 and 1615 cm'l, and resonances at T 8.52 (singlet, 5H), T 8.55 (singlet, 5H), T 8.22 (singlet, 5H), T 7.84 (singlet, 5H), T 7.26 (center of AB quartet, J/V'16,~’2H), T 5.16 (singlet with fine splittings, 10 1.1H). These spectra are consistent with the assigned structure. A parent peak at 504 m/e was evident in the mass spectrum. F. Identification of trans-1-Acetyl—2,4,5- trimethyl—4,5-diphenylcyclopentene A pure sample of the trans—isomer was collected by pre- parative v.p.c. after having been chromatographed on neutral alumina. The infrared and n.m.r. spectra, Figures 55 and 54, respectively, exhibited absorptions at 1670 and 1615 cm“1, and resonances at T 8.98 (singlet, 5H), T 8.82 (singlet, 5H), I 8.50 (singlet, 5H), I 7.90 (singlet, 5H), T 7.16 (center of AB quartet, Jr~/18 cps,’V’2H), and T 2.82 (broad singlet with fine splittings, 10 or 11H). Anal. Calcd for C22H24O: C, 86.80; H, 7.95 Found: C, 86.12; H, 8.06. 88 -oHosoHscmsmflanm.4nasnpmsHHIIm.a.mussumomnfiu coma oobm A. IEUV hucmsvmum comm .A¢HOUV mcmucmm mac mo Esuuommm pmumumcH ooom .mam madmam comm mm H _ om ca 0m (queoxed) eouequmSUEIL .Awaoov mcmucmm Ioaomoahcwcmacum.alamnumfiauulm.¢.N|a>um08|aimao mo Esuuummm commumcH .Qam musmam AHIEUV mucmdwmum 89 com coca coma ccaa coma coma _ fi 7 0N ow 0m 0m (queoxed) eoueqirmsuexl 90.0a 90 I, Im.4lawflpm8auulm.4.N|H%umUMIalmao mo Eds 9o.m 9c. .mcmucmmoaomoamcmflmac pommm mucmcommu oaumamme Hmmaosz m 90.4 .mm musmam 9o.m . .A4HUUV mcmucmmoHomo Iamcmcmactm 4|a>£umEaHulm.4.NIH>DmUMIaImsmup mo Esuuommm cmHMHmcH .mmm Gunman AHIEUV hocmsomum 91 coma oocN oom N a 000m comm .om 04 cm .om (queoxed) eoueqqrmsuexl 92 .A4Hoov mcmucmmoaowo lamcmcmaclm.4lawcumEauulm.4.muH>umUMIalmcmuu mo Esuuommm commumcH AHIEUV mocmsvmnm com 0009 cows , 004a 0089 .nmm magmas .ooma _ c i _ > _ cm 04 cm cm (queoxed) eouquImsuexm .mcmbcmmoaomoamnmSQaclm.4 Ia%£mealem.4.mla>pm081almcmup mo Eggpommm mommaommn oaumcmma Homaosz .4m mnsmam 90.0a 9o.m 9o.m 90.4 9o.m 95 LITERATURE C I TED 94 LITERATURE CITED (1) C. Walling, "Molecular Rearrangements," P. de Mayo, Ed., John Wiley and Sons Inc., New York, N. Y., 1965, pp. 407-455. (2) David Y. Curtin and Marvin J. Hurwitz, J. Am. Chem. Soc., 14, 5581 (1952). (5) James W. Wilt and H. Philip, J. Org. Chem., 88, 891 (1960). (4) William B. Smith and James D. Anderson, J. Am. Chem. Soc., 62, 656 (1960). (5) Christoph Ruechardt and Roland Hecht, Tet. Letters, 1962, 957. "“ (6) Lynn H. Slaugh, J. Am. Chem. Soc., 81, 2262 (1959). (7) A. N. Nesmeyanov, et al., Tet., _8, 94 (1961). (8) Philip S. Skell, et al., J. Am. Chem. Soc., 88, 504 (1961). (9) C. Walling, "Molecular Rearrangements," P. de Mayo, Ed., John Wiley and Sons Inc., New York, N. Y., 1965, pp. 416-417. (10) O. A. Reutov, Pure Appl. Chem., lj 205 (1965). (11) Frederick D. Greene, et al., J. Org. Chem., 81, 2087 (1966). (12) H. E. Zimmerman and A. Zweig, J. Am. Chem. Soc., 88J 1196 (1961). (15) J. A. Berson, et al., J. Am. Chem. Soc., 84, 5557 (1962). (14) L. K. Montgomery, et al., Abstracts, 147th National Meeting, of the American Chemical Society, Philadelphia, Pa., April 1964, p. 29-N. (15) Lynn H. Slaugh, J. Am. Chem. Soc., 81, 1522 (1965). (16) D. H. R. Barton, et al., J. Am. Chem. Soc., 88J 4481 (1961). 95 (17) (18) (19) (20) (21) (22) (25) (24) (25). (26) (27) (28) (29) (50) (51) (32) (55) (54) 96 William Reusch, et al., J. Am. Chem. Soc., _8, 2805 (1966). K. Heusler and J. Kalvoda, Angew. Chem. internal. Edit., .3, 525 (1964). Alwyn G. Davies, Butterworth and Co. "Organic Peroxides,‘ Limited, London, 1961, p. 59. James H. Brewster and C. J. Ciotti, J. Am. Chem. Soc., 11, 6214 (1955). Jr., Nicholas Milas and Aleksander Golubovic, J. Am. Chem. Soc., g1, 3361 (1959). Jon Munch-Petersen, Acta Chem. Scand., 2007 (1958). 12) E. M. Schultz, et al., (1955). J. Am. Chem. Soc., 18, 1072 L. S. Silbert and Daniel Swern, (1958). Anal. Chem., 88J 585 4, 1986 (1962). Micheal M. Martin, J. Am. Chem. Soc., John P. Lorand and Paul D. Bartlett, J. Am. Chem. Soc., 66, 3294 (1966). Heinosuke Yasuda, Sci. Research Inst. (Japan), §1, 560 (1955). Repts. J. N. Chatterjea and K. Prasad, Chem. Ber., 88, 1740 (1960). J. Colonge and R. Marey, Syntheses Coll. Vol. 601 (1965). Org. _:[_\,_I J. R. A. Pollock and R. Stevens, Eds., ”Dictionary of Organic Compounds," Vol. 4, Oxford Press, New York, 1965, p. 2709. Maurice Metayer and Noele Epinay, Compt. rend., 226, 1095 (1948). J. R. A. Pollock and R. Stevens, Eds.,"Dictionary of Organic Compounds," Vol. 4, Oxford Press, New York, 1965, p. 2294. T. E. Zalesskaya and T. B. Remizova, Zh. Obshch. Khim., ,3; (12), 5802 (1963). Catherine G. LeFevre, et al., J. Chem. Soc., 1955, 480. ”Illlllllllllllllll “((111))” (WWW 3 1293 030831311