ABSTRACT PHOTOCHEMHCAL REARRANGEMENT OF O’B-EPOXY KETONES. AN ELABORATION OF THE MECHANISM by Charles S. Markos A study of the photochemical rearrangements of a number of a,fi-epoxy ketones was undertaken to elaborate the mech- anism of the shift of a B-substituent to form a B-dicarbonyl compound. The order of migratory aptitudes found for 5-substit- uents is: benzhydryl and benzyl > hydrogen > methylene > methyl >> phenyl. With the exception of hydrogen, this order parallels the stability of the corresponding radical, and radical characteristics are suggested for intermediates or transition states in the mechanism. Although a fi—benzyl substituent is partly diverted to free radical products (33.,dibenzyl), the stereospecificity of the rearrangement, the migratory aptitude of B-hydrogen and the formation of a strained rearrangement product (2-acetyl-1-hydroxycyclobutene from 3-methyl-2,3-epoxy- cyclopentanone) all argue against a general fragmentation— recombination mechanism for the rearrangement. The 1,3-di- radical intermediate prOposed for these rearrangements must be very short lived, since the pulegone oxide stereoisomers rearrange stereospecifically faster than they isomerize. Studies with triplet state quenchers (piperylene) and sensitizers (acetophenone) indicate that the rearrangement Charles S. Markos occurs only from the first excited singlet state [81] of the epoxy ketone. An alternate explanation for the re- arrangement of a B~hydrogen substituent was ruled out by the absence of bimolecular hydrogen abstraction involving a triplet excited state [T1]. 4-Methyl-3,4-epoxy-2-pentanone was rapidly photo- reduced by tri-gfbutyl stannane (m = 0.65). This reaction apparently proceeds yi§_the triplet state [T1] since it was completely quenched by solvent piperylene, leaving only photorearrangement to 3-methyl-2,4-pentanedione (m = 0.03). PHOTOCHEMICAL REARRANGEMENT OF c.6-EPOXY KETONES. AN ELABORATION OF THE MECHANISM BY 9“ Charles §L Markos A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1967 1 ACKNOWLEDGMENT The author is grateful to Professor William H. Reusch for his guidance and counsel throughout the course of this investigation, and also for anznging financial support from June 1966 to August 1967. Special thanks is given to the Dow Chemical Company for a summer scholarship in 1966. ii F -“-s§£‘r;.‘f1q TABLE OF CONTENTS Page HISTORICAL AND INTRODUCTION . . . . . . . . . . . . . 1 RESULTS . . . . . . . . . . . . . . . . . . . . . . . 9 I. Photoproducts of d,B—Epoxy Ketones . . . . 9 A. Photolysis of 3-benzyl- -5, 5- -dimethyl- 2, 3-epoxycyclohexanone . . . . . . 9 B. Photolysis of 4, 4-dipheny1- -2, 3-epoxy— cyclohexanone . . . . . . . . . 10 C. Photolysis of 5, 5-dimethyl- -2, 3-epoxy- cyclohexanone . . . . . . . . . . . . 11 D. Photolysis of 2,3-epoxycyclopentanone . 12 E. Photolysis of 3-methyl-2,3-epoxycyclo- pentanone . . . . . . . . . . . . . . 12 F. Photolysis of B—pulegone oxide . . . . . 13 G. Photolysis of a-pulegone oxide . . . . . 14 H. Photoreduction of 4-methyl-3,4-epoxy-2— pentanone with tri—gfbutyl stannane. . 14 II. Relative Rates of Photochemical Reactions. . 15 A. Relative Quantum Yields . . . . . . . . 15 B. Solvent Variations . . . . . . . . . . . 17 C. The Pulegone Oxides . . . . . . . . . . 18 DISCUSSION . . . . . . . . . . . . . . . . . . . . . 25 I. Radical Character of the Rearrangement . . . -25 II. Photochemistry of the Rearrangement . . . . 28 EXPERIMENTAL . . . . . . . . . . . . . . . . . . . . 34 I. General Procedures . . . . . . . . . . . . . 34 A. Apparatus . . . . . ; . . . . . . . . . 34 B. .Melting Points . . . . . . . . . . . . . 34 C. ~Microanalysis . . . . . . . . . . . . . 35 II. Preparation of a, B-Epoxy Ketones . . . . . 35 A. Preparation of 3-benzyl- 5 5-dimethyl- 2 ,3-epoxycyclohexanone . . . . . . . . 35 1. Preparation of 5, 5-dimethyl- -3- methoxy-2-cyclohexene- -1-one . . 35 2. Preparation of 3-benzyl— 5, 5- -dimethyl- 2- -cyclohexene- -1- -one . . . . . . . ~36 3. Preparation of 3-benzyl- -5, 5-dimethyl- 2 ,31epoxycyclohexanone . . . . . . . 37 iii TABLE OF CONTENTS - Continued III. IV. E. F. Preparation of 4,4-dipheny1-2,3-epoxy- cyclohexanone . . . . . . . . . . . 1. Preparation of Stilbene glycol . . 2. Preparation of diphenyl acetaldehyde 3. -Preparation of 4, 4-diphenyl-2-cyclo- hexene-l-one . . . . . . . . . . 4. Preparation of 4, 4-dipheny1- -2, 3- epoxycyclohexanone . . . . . . . Preparation of 5, 5-diphenyl- -2 ,3-epoxy- cyclohexanone . . . . . . . . . . 1. Preparation of 5, 5- -dimethyl-2-cyclo— hexene- 1-one . . . . . . . . 2. Preparation of 5, 5-dimethy1-2,3- epoxycyclohexanone . . . . . . . . Preparation of 2,3—epoxycyclopentanone 1. ~Preparation of 2-cyclopentane-1-one 2. Preparation of 2,3-epoxycyclo-‘ pentanone . . . . . . . . . Preparation of 3-methyl-2-cyclo- pentanone . . . . . . . 1. Preparation of 3-methyl- -2- -Cyclo- pentene-l-one . . . . . . . . . 2. Preparation of 3—methyl— —2 ,3-epoxy— cyclopentanone . . . . . . . . . . Other Epoxy Ketones . . . . . . . . Photolysis of 3—benzyl-5,5-dimethyl-2,3- epoxycyclohexanone . . . . . . . . . . Photoproducts from 3-benzyl—5,5- dimethyl-2,3-epoxycyclohexanone . 1. Benzyl dimedone . . . . . 2. Dimedone . . . . . . . . . . . . 3. Dibenzyl . . . . . . . . . 4. a-Benzyl diethylether . . . . . Photolysis of Neat Samples . . . . Separation of a dimedone and benzyl dimedone mixture . . . . . . . . . Photolysis in other solvents . . . . 1. Photolysis in pentane . . . . . 2. Photolysis in 0.1M piperylene . -Mass balance of photoproducts from 3— benzyl- -5, 5- -dimethyl- -2, 3-epoxycyclo- hexanone . . . . . . . . . . . . . Photolysis of Benzyl Dimedone . . . . . ,A, B. Preparation of benzyl dimedone . . . Photolysis of benzyl dimedone iv 43 43 43 43 44 44 45 45 46 46 47 47 48 48 48 TABLE OF CONTENTS - Continued Page V. Photolysis of 4, 4-diphenyl-2, 3-epoxy- cyclohexanone . . . . . . . 49 A. Photoproducts from 4, 4-diphenyl- -2, 3- epoxycyclohexanone . . . . 49 1. 4, 4-Diphenyl-1, 3-cyclohexanedione. 49 2. 3, 3- -diphenyl- 2-hydroxymethylene- cyclopentanone . . . . . . . . 50 B. Derivatives of the photoproducts . . . 50 VI. Photolysis of 5, 5-dimethyl-2, 3-epoxy- cyclohexanone . . . . . . 51 A. Photoproducts from 5, 5-dimethyl- -2, 3- epoxycyclohexanone . . . . . . . . 52 1. Dimedone . . . . 52 2. 4, 4-Dimethyl-2-hydroxymethy1ene- cyclopentanone . . . 52 3. Basic cleavage of 4, 4-dimethyl- -2- hydroxymethylenecyclopentanone . . 52 VII. Photolysis of 2,3-epoxycyc10pentanone . . 53 A. Photoproducts of 2,3-epoxycyclopentanone 53 1. Cyclopentane-1,3-dione . . . . . . 53 2. Unknown compound . . .1. . . . . . 53 VIII. Photolysis of 3-methyl-2, 3-epoxycyclo- pentanone . . . . 54 A. Photoproducts of 3-methyl- -2 ,3—epoxy- cyclopentanone . . . . . . . . . . . 54 1. Acetic Acid . . . . . . 54 2. 2—Acetyl-1—hydroxycyclobutene . . 55 B. Preparation of 1-methoxy- 2-acetylcyclo- butene . . . . . . . . . . . . . . . 56 IX. Photolysis of 2-acetyl-1-hydroxycyclobutene 56 X. Photolysis of the Hulegone Oxides . . . . 57 A. Photolysis of B—pulegone oxide . . . . 57 B. Photoproducts of B-pulegone oxide . . 57 1. 2-Acetyl-2,5—dimethylcyclo- hexanone stereoisomers . . . . . . 57 2. afpulegone Oxide . . . . . . . . 57 C. Photolysis of a-pulegone Oxide . . . . 58 XI. Photoreduction of 4-methyl-3,4-epoxy-2- pentanone with tri-gfbutyl stannane . . 58 XII. Preparation of 3,3-dimethylcyc10pentanone. 59 A. Preparation of 3,3-dimethylcyclohexanol 59 B. Preparation of B,B-dimethyl adipic acid 59 C.; Preparation of 3,3-dimethylcyclopentanone60 V TABLE OF CONTENTS - Continued XIII. Preparation of a-Benzyl Diethylether A. Preparation of l-phenyl-Z-propanol B. Preparation of a-benzyl diethylether XIV. Preparation of Tri-gfbutyl Stannane XV. Quantitative Measurements of Photolysis Rates . . . . . . . . . . LITERATURE CITED . . . . . . . vi Page 60 60 60 60 61 81 TABLE LIST OF TABLES Effect of acetophenone on photolysis of isophorone oxide . . . . . . . . . . . Photolysis of 3-benzyl-5,5-dimethyl-2,3—epoxy- cyclohexanone in different solvents . . . . . Photolysis of 4-methyl- -3, 4-epoxy-2-pentanone; rearrangment vs reduction by tri-n-butyl stannane: effect of 2 4-pentadiene . . . Photolysis of the pulegone oxides . . . . . . Effect of acetophenone and 2, 4-pentadiene on photolysis of 5, 5— O-dimethyl— —2 ,3-epoxycyclo- hexanone . . . . . . . . . Mass spectrum of 2-acetyl-1-hydroxycyclo- butene . . . . . . . . . . . . . vii Page 19 20 21 23 24 55 Figure Page 1. Nmr Spectrum of 3-benzyl-5,5-dimethyl-2,3- epoxycyclohexanone . . . . . . . . . . . . . 62 2. Infrared spectrum of a-benzyl diethylether . 63 3. 'Nmr spectrum of quenzyl diethylether . . . . 64 4. Nmr spectrum of 3, 3- -diphenyl-2-hydroxymethylene- cyclopentanone. . . . . . . . . . . . . . . 65 5. Infrared spectrum of 3, 3- -diphenyl -2-hydroxy- methylenecyclopentanone. . . . . . . . . . 66 6. Infrared spectrum of the formalin derivative of 4,4-diphenyl-1,3-cyclohexanediOne . . . . 67 7. Infrared spectrum of 4, 4-dimethyl- -2-hydroky- methylenecyclopentanone . . . . . . . . . 69 8. Nmr Spectrum of 4, 4-dimethyl— 2-hydroxymethylene- cyclopentanone . . . . . . . . . . . . . . 71 9. Infrared spectrum of compound "X" . . . . . . 72 10. Nmr spectrum of compound "X" . . . . . . . . 74 11. Nmr spectrum of 1-acetyl-1-hydroxycyclo— butene . . . . . . . . . . . . . . . 75 12. Infrared spectrum of 2-acetyl-1-methoxycyclo- butene . . . . . . . . . . . . . . . . . . . 76 13. Nmr spectrum of 2-acetyl-1-methoxycyclobutene 77 14. Infrared spectrum of 2-acetyl-1-methoxycyclo— butene . . . . . . . . . . . . . . . . . . . 78 15. Infrared spectrum of 2—acetyl— 2, 5—dimethyl- . cyclohexanone 54 . . . . . . . . . . . 79 16. Infrared spectrum of 2-acetyl- 2, 5- -dimethyl- LIST OF FIGURES cyclohexanone 55 . . . . . . . . . . . . 80 viii HISTORICAL AND INTRODUCTION The photochemical rearrangement of a, B epoxy ketones has been known since 1918, when Bodforss (1) reported the conversion of chalcone oxide 1.to dibenzoyl phenylmethane 2_upon exposure to light (eq 1). More recently, Zimmerman (2) has shown that trangfdypnone oxide 3 is isomerized to the cis-isomer g_photochemically. (eq 2). C6H5 k\_ 0 o (1) \:>185 ’ C6H5 H |w I» Several mechanistic interpretations of the reactions described by Bodforss and Zimmerman are possible. These may be classified as (a) intermolecular abstraction of a hydrogen atom, or (b) direct 1,2-hydrogen migration. CHART _:_c_ :t h C6H5>A<' “H5 1 ”TM C6H5 V _ *- -1- > H« c6115 a > CGHS b \ c H *0 o 6 SMC‘BHS /'O‘\Kfl\c ‘R-H il/Kc H 6 CsHs CsHs 6H5 Electronic excited state ‘H- II (°) Cr 0+).(-) Jeger, Schaffner and Wehrli (3) then found that B-alkyl groups also shift in a similar manner (eq 3); consequently, mechanism (b) appeared to offer a more general rationale for the rearrangment. Additional examples of alkyl group rearrangement were reported by Johnson, Dominy, and Reusch (4) (eq 4,5). These workers also noted a remarkable and unexpected preferential shift of a B-methylene over a phenyl group as illustrated in equation (6). Independent work by Zimmerman, Cowely, 3 Tseng and Wilson (5) has disclosed a similar migrational order in the acyclic epoxy ketone 15 (eq 7). hv O O > M 2 t“ 07;? O O |l /fi——C6H5 12 CsHs (5) These facts led Reusch and Zimmerman to propose a mechanism (illustrated in Chart II) in which n, v* excita— tion of the carbonyl group causes oxirane ring cleavage (5) to an intermediate g. followed by rearrangment of a 5—substituent. CHART_I_:_[_ . O 0 * o 0 R3/\ .. 32.21241: R3/\ . )c~c-c-R1<———— )C-C-C-Rl Rz 5 a R2 5 a * * *- fi * R R3 9 9 R3 -C“-“C"="t: -R1 < /\c -c -c -R1 * > R2 5 a R2 / .12. o o 9. R3—c-c-c—R1 R2 * = - or i This mechanism successfully rationalizes the effects of substituent variations. If R1 = phenyl, oxirane ring opening to intermediate Q_is not favored because the re— active character of the excited carbonyl carbon would be delocalized by the phenyl group. In fact, Zimmerman and co—workers (5) have found that trgngfdypnone oxide 32 gives 1,3-diphenyl-3-butene-2-ol-1-one‘lla as the major photo— product. Dimethyl acrylophenone 3b similarly gave 1-phenyl- 3-methyl-3-butene-2-ol-1-one 112, This transformation ap- parently proceeds gig the well documented (6) abstraction of y-alkyl hydrogen by the triplet state of an aryl ketone, as outlined in Chart III. If R1 is an alkyl group oxirane ring opening to inter- mediate g_occurs, and the relative migrational aptitudes of the fi-substituents must be determined in a subsequent step. CHART III H H (K; I O I H2C H2C >T7<\ —— Mew“ C6H5 R O H R O H .§§ R = C6H5 l 5313 R = CH3 OH EH2 .1 H2O 7 °\C6H5 R "' C " [CH — C‘EC6H5 < x) y,_\\§/) R (L9 H H ‘\\N CH2 R - c - CH - c - C6H5 HO 0 173 R = C6H5 17b R = CH3 Lack of phenyl migration from intermediate g may be explained by assuming radical character,in QJleading either to the radical pair Q_or to a transition state having more break— ing of the C -R bond than making of Ca-R bond. In both 6 cases, the migratory aptitude of the B-substituent should parallel the stability of the corresponding radical. A fivefold preference for methyl migration y§_phenyl in the thermal decomposition of 5-methyl-B—phenyl-B-peroxy- propiolactone (eq 8) was noted by Greene Adam, and Knudsen (7) and a 1,3-diradical similar to intermediate g_was 6 suggested as an intermediate in this transformation. 0 O O O ' -CO (8) ——A_2—> C H /IJ\/ + /u\/C6H5 6H5 6 5 5 1 In stereochemically rigid steroidal epoxy ketones, Jeger and co—workers (3,8) have found similar 1,2—migrations of 5-substituents to yield B-dicarbonyl compounds in yields ranging from 10 to 80% as shown in (eq 9-16). (10) (11) (12) CHO 1 2.5.3. 22 53.9. N ~ ’\. (14) 1 hv o C. > “O O/ ’ o 1.1. 1?— (15) (16) ‘ 52.9. Rearrangement of the Al-unsaturated epoxy ketones (18 and 12) was found to proceed only by exposure to 2537 R light * excitation) and not from irradiation with light of (w. w wavelengths greater than 3100 R (n,v* excitation). The stereospecific rearrangment of epoxy ketones 21 and g§_has been cited (8) as evidence for a "synchronous" or "conceded” process (Chart IV) as opposed to a two step 8 fragmentation mechanism. However, the conformational re- strictions of medium sized rings may prevent facile equi— libration of a diradical intermediate of type Q.(Chart II). CHART IV Our objectives in this research were: (a) To deter— mine the existence and nature of intermediate QDby investi- gating the effect of radical scavengers, triplet quenchers and sensitizers on the rearrangement. (b) To determine whether the migration of a B-substituent is truly concerted. (c) To elaborate the migrational aptitudes of various 5— substituents. RESULTS I. PHOTOPRODUCTS OF Q,B-EPOXY KETONES The photochemical transformations described here were effected by a medium pressure mercury lamp equipped with a Corex filter. The yields are not optimum conversions as it was desired to obtain relatively pure materials uncontamin- ated by secondary products and polymers. A. Photolysis of 3—benzyl—5,5-dimethyl-2,3—epoxycyclo- hexanone: 0 O 6H5 (17) CsHs _...._> HO 0 + HOQO :1 :42 + C6H5/\/C6H5 + O/\ 47 48 Irradiation of an 0.029M ether solution of 3-benzyl- 5,5-dimethyl-2,3-epoxycyclohexanone for 4 1/2 hr gave a single rearrangement isomer, benzyl dimedone, in 19.4% yield. The identity of this material was demonstrated by comparison (infrared spectrum and melting point, 155-1570) with authentic benzyl dimedone, prepared by alkylation of dimedone (9). Inspection of the vpc chromatogram before and after basic extraction of the photomixture indicated 9 10 that two base-soluble compounds were present; the second proved to be dimedone, isolated as formalin dimedone in 11.7% yield. Base-insoluble compounds isolated from the photo— mixture were unchanged epoxy ketone (28.6%), dibenzyl (21.0%), and Q-benzyl diethylether (6.5%). These were purified by a combination of column and preparative vapor phase chromatography. The identities of the latter com- pounds were established by a comparison of the melting points and infrared spectra with authentic dibenzyl, and by comparing the infrared and nmr spectra with those from independently synthesized a-benzyldiethyl ether. Irradiation of 38 in pentane gave the same product mixture with the exception of the substituted ether. Products having similar vpc retention times were assumed to be benzyl pentane isomers. Irradiation of neat samples of 38 gave the stoichiometrically expected ratio of dimedone to dibenzyl of 2:1. An ether solution of benzyl dimedone (0.007M) re- mained essentially unchanged after being irradiated under the same conditions used above. B. Photolysis of 4,4—diphenyl—2,3—epoxycyclohexanone: hv ether) 0 O CH CHO o 6 CH CH 5C6115 CH‘CH 65 65 65 65 (18) 11 Irradiation of an 0.025M ether solution of 4,4-di- phenyl-2,3-epoxycyclohexanone for 3 1/2 hours gave a 47% yield of base soluble material which proved to be a 5:1 mixture of two components: 3,3-diphenyl-2-hydroxymethylene- cyclopentanone, mp 94-960, and 4,4-diphenyl-1,3-cyclo- hexanedione, mp 1930 (dec.). Unchanged epoxy ketone was recovered in 50% yield from the base insoluble residue. 3,3-Diphenyl-2-hydroxymethylenecyclopentanone was characterized by a satisfactory C, H micro analysis, infra- red (Fig. 5) and nmr (Fig. 4) spectra and a crystalline copper chelate mp 131-1330. The infrared spectrum of the copper chelate exhibits absorption at U (CCl4) 1600, max 1463 and 1340 cm—1 4,4—Diphenyl-1,3-cyclohexanedione was characterized by a satisfactory C, H analysis, and by its formaline deriva- tive which has a similar infrared Spectrum to that of formalin dimedone in the carbonyl region. C. Photolysis of 5,5-dimethyl-2,3-epoxycyclohexanone: Irradiation of an 0.048M ether solution of 5,5—di- methyl-2,3-epoxycyclohexanone for 1 1/2 hr gave 23- 34% of a 10:1 mixture of dimedone and 4,4-dimethyl-2-hydroxy— methylenecyclopentanone. Unchanged epoxy ketone accounted 12 for 42% of the starting material. The structure of 4,4—dimethyl—2-hydroxymethylene- cyclopentanone was determined from the nmr (Fig 8) and infrared (Fig 7) Spectra. Base catalyzed degradation of this compound gave a Single base-insoluble compound, identi- fied as 3,3-dimethylcyclopentanone by comparing its infra- red spectrum with that of an authentic sample. D. Photdysis of 2,3-epoxycyclopentanone: O ’3 hv O O W 0 *— 41 52 Irradiation of an 0.092M ether solution of 2,3—epoxy— cyclopentanone for 1/2 hr gave a precipitate in 4.7% yield, mp. 149-1510, which proved to be cyclopentane-1,3-dione (10). Spectrosc0pic examination was not particularly informative as this compound is quite insoluble in suitable solvents, however, absorption characteristic of a cyclic enolized B-diketone is observed at 1580 and 1660 cm-1 E. Photolysis of 3-methyl-2,3—epoxycyclopentanone: O O (21) /. L). HO\_ l\ ether 13 Irradiation of an 0.132M ether solution of 3—methyl- 2,3-epoxycyc10pentanone for three hr gave a 40% yield of 2-acetyl-1-hydroxycyclobutene mp 104-104.5°, provided con- tinuous extraction of the product was in effect throughout the reaction. If continuous extraction was not employed, the only base-soluble material isolated after long irradia- tion periods (over 2 hr) was acetic acid. However, 2-acetyl- l-hydroxycyclobutene was Shown to be photochemically inert under equivalent conditions. The rearranged product was characterized by micro- analysis, infrared, nmr, and mass Spectra. Positive iodo- form and ferric chloride tests were also obtained. Treat- ment with ethereal diazomethane gave 1-acetyl-1-methoxy- cyclobutene, and the nmr Spectrum of this derivative (Fig 13) displayed a quartet for the four methylene protons in con- trast to the broad singlet observed with the enol precursor. F. Photolysis of 6-pulegone oxide: 3 . 54 55 33 Irradiation of a 0.057M solution of B-pulegone oxide in ether for three hours gave three products: a-puiegone oxide, and the diasteromic 2,5-dimethyl—2-acetylcyclo~ hexanones. Unchanged B-pulegone oxide represented more than 90% of the product mixture. 14 a-Pulegone oxide was identified by Vpc retention time only. The two rearrangement isomers were isolated by re- moving the solvent and distilling the residue at 5mm, col- lecting 22 500 mg of forerun. Preparative Vpc gave two isomers having identical infrared spectra to authentic samples prepared by ozonolysis of a 4-methylaiggepulegone isomer mixture (11), configurations of which have been recently established (12). G. Photolysis of a—pulegone oxide: (23) hV> + + o o .- 0 s‘ F O O—r 33 55 54 43 Irradiation of a 0.25M solution of a-pulegone oxide in acetonitrile (£3 3 ml) gave fi-pulegone oxide and the same two rearrangement isomers described above. These as- signments rested on a comparison of vpc retention times with authentic samples. H. Photoreduction of 4-methyl-3,4-epoxy-2-pentanone with tri-nfbutyl stannane: O OH O (24) :>Zl\\}r’+-2(C4H9)3SnH hV> :>L\\//n\\ 15 A 0.1 molar pentane solution of 4-methyl-3,4-epoxy- 2-pentanone containing an equivalent of tri-nfbutyl stannane was irradiated for 2 hr under an atmosphere of nitrogen. Column chromatography of the products gave a 93% yield of diacetone alcohol as the only isolated product, identified by comparison (infrared spectrum) with authentic diacetone alcohol. A Similar chromatographic separation of the epoxy ke— tone and tri-gfbutyl stannane showed that no reduction occurred on the column. II. Relative Rates of Photochemical Reactions All solutions were degassed and irradiated in sealed pyrex ampules, using a merry-go-round apparatus (13). An inert internal standard was added to the solutions as an aid to vpc integration. The data thus obtained is pre— sented in Tables 1—5. A. Relative quantum yields. A comparison of the rates of rearrangment and tri- gfbutylstannane photoreduction of 4-methyl-3,4-epoxy—2- pentanone with the stannane photoreduction of cyclohex— anone to cyclohexanol (m - 1.00 assumed) gave relative quantum yields of 0.65 and 0.03 respectively for photoreduc- tion and rearrangement of the epoxy ketone. 16 OH hv _ (25) (C4H9)3SnH D - 1.00 21 .53. >LOX\I OH hv _ (24) 6" (C4H975SnH ,:>L“~’//H\\ E _ 0'65 § 55 O O § 59 The quantum yield for 3-methyl—2,4-pentanedione 59 formation was not affected by solvent changes (piperylene, pentane, acetonitrile, ether) or the presence of air. However, photoreduction of §_with tri-nfbutyl stannane was completely quenched by 6M piperylene leaving only rear- rangement to 3-methyl-2,4-pentanedione (m = 0.03). Thus it appears that photoreduction of 4-methyl-3,4—epoxy-2- pentanone proceeds only through a triplet [T1] excited state, and photorearrangment of the epoxy ketone occurs only from the Singlet [Si] excited state. Additional evidence excluding a triplet [T1] state intermediate in the rearrangement is seen in the failure of acetophenone (Et = 73.6 kcal) to sensitize the rear— rangement of isophorone oxide or of 5,5-dimethyl-2,3-epoxy- cyclohexanone. This fact contrasts with a report (14) of 17 triplet sensitization of certain steroidal epoxy ketones by compounds of lower triplet energy than acetophenone (triphenylene, biacetyl, B-dicarbonyl compounds). Sensi- tized stannane photoreduction of these substrates was also observed. B. Solvent variations A simultaneous irradiation of 3—benzyl-5,5-dimethyl- 2,3-epoxycyclohexanone in ether, acetonitrile, and 2,5-di- methyl-2,4-hexadiene demonstrated that the rate of reaction was essentially the same in all three solvents. The frag— mentation product, dibenzyl, was also produced at approxi- mately the same rate in all three solvents. This insensi- tivity to solvent variations, including triplet state quenchers, was alSO noted for the B-hydrogen rearrangment in 5,5-dimethyl-2,3-epoxycyclohexanone, and again implies the rearrangement proceeds only from the Singlet excited state [S1]. A comparison of the rearrangement rates for 3-benzyl- 5,5-dimethyl-2,3-epoxycyclohexanone (xmax 2850 A) and 4- methyl-3,4-epoxy-2-pentanone (xmax 3000 R) is not meaning- ful, as the absorption spectra of these two compounds are markedly different. All preparative scale photolyses of epoxy ketones were not affected in any detectable manner by degassing. Further- more, addition of piperylene, a triplet state quencher, to ether solutions of 3-benzyl—5,5—dimethyl-2,3-epoxycyclohexanone 18 or isophorone oxide did not affect product ratios or the overall yield. C. The pulegone oxides Simultaneous irradiation of acetonitrile solutions of fi-pulegone oxide and a—pulegone oxide gave a stereospecific rearrangement to stereoisomeric 2,5—dimethyl-2—acetylcyclo- hexanones as well as isomerization to the stereoisomeric pulegone oxide. The combined rate of rearrangement to 5- diketones was greater than the rate of isomerization in all cases. A temperature difference of 22 300 had no apparent effect on the course of the reaction. 19 Table 1. Effect of acetophenone on photolysisa phorone oxide. of iso- System Ketone(s),(M) Solvent ?;?:) % Reaction 1 isophorone,(0.3) aceto- .267 tracec nitrile b .754 0.31 1396 0.83 2 isophorone,(0.3) aceto- 267 none acetophenone,(0.5) nitrile 754 none 1396 none 3 acetOphenone,(0.3) toluene 30 3.4d 60 6.0 90 8.9 120 11.3 a450 watt mercury lamp filtered by pyrex, 0.1M naphthalene in cyclohexane, and soft glass. bProduct: 4,4-dimethyl—2-acetylcyclopentanone from compar— ison of Vpc retention times (20% SE-30) with an authentic sample. Chlorobenzene was added as an internal standard. cA ferric chloride test proved more sensitive than vpc. dDetermined by disappearance of 3200 R maximum in uv. 20 Table 2. Photolysisa of 3-benzyl-5,5—dimethyl-2,3—epoxy- cyclohexanone in different solvents. . b . Concentrat1on Solvent T1me(hr) (M) x 102 Ether 0 8.83 2 7.89 4 7.53 Acetonitrile 0 8 96 2 8.35 4 7.75 2,5-dimethyl-2,4— 0 8.88 hexadiene 2 7.75 4 7.23 a200 watt mercury lamp, Corex filter. bAnalysis by vpc (4% QF-I) with triphenylcarbinol added as an internal standard. 21 5000.05 Amwmo..mcmccmumv .I 5.5.55555.5 I- 55 555x5505omo A55.V.5 m 5555 5 Hooo.OH II om Aaom..mcmccmumv .l 5.5.55555.5 5:55:55 Amom.v.5 5M 5555 -----mmmmqm--:....M.5.m....m----:mwm ............................................... o.nv5oo.5 555.5 555 I. 555.5 555.5 5 mcmmeO5omo A55.V.w 5m 5555 -----mmmmqm ...... Mmum ...... m«m----:-:::--:-.,. ........................... U55555.5 555.5 555 A5 .55555555555 .1 555.5 5555.5 5 .mcmxmnO5umo A55.V.5 m 5555 5 5555.5 ----555q5----- 555 - 55.5555555555W U55555.5 555.5 555 55555..mcmacmum .1 555.5 5555.5 5 .mcmmeO5oao A55.V.5 5m 5555 ------55555..----55555------555 ............................................... 555.5 555.5 55 15555..mcmccmumv I. 555.5 55555.5 5 .mamx5555omo A55.v.5 5m 5555 -----555545------55555------55 ................................................ 5555.5 555.5 55 5555.5 555.5 55 U5555555 55555.5 55 “5.55555555555 .I 55.5 555.5 5 .mcmx5555omo A55.v.5 55553 555 Aposooumv. choumxv C58 dz.mpodoomV . ASV COHMMHHSCUSOU $859 ucm>aom ASV waoumx mUHSOm 5£m55 lhwcmflpmucmmlv.m mo uommmw "mamacmum HmdeLMIHHp ma SOHHUSUCH m> pcwfiwmcmuHmmH.”mcocmuammlmlmxommlv.nuahfluwalv mo mHmmaouonm .m CHQMB .mmHmEmm 03¢ mo mmmuw>¢m .HoqumnoHumu «pUSUOHmm .Honooam wcoumUMHU uuodwoumm mu .Amav msumummmm choulomlmuuwe 5 mo mm: map How Hmemz .b .m .59 on HSMGHMHm mum mad .GCOHUcmucmmlv.Nlamnumfilm "posvoumo .vumocmum Hmcumuaw mm mm UmUUm mcmNGmQOHoHnu £953 Amdmm Rmfiv Um> an m5m>HMC¢n .kuaflm xmummm AN..®:mccmumv waocmxmn 5555.5 .. 55 5555555 555.5 -55555 m 5555 U AuUSUOHmM. mmcoumxw NQHEM d2.mposwvmq sz5mcoumx mUHSOm pflmflq sz coaumnuamuaoo mEHB pcm>aom Awmscflucouv .m manme 23 Table 4. Photolysisa’b of the pulegone oxides. M Temp Time Reaction 55 [55+54] Ketone 0C (min) % [54J [isomeri— zation| 43_(.25M) 45-50 0 0C - - 100 0.16 1.00 - 200 2.16 1.91 1.50 300 2.20 4.60 1 04 400 3.69 2.61 1.04 .43 ( 25m) 18-20 0 0C - - 100 0.92 00 1.87 200 2.56 1 43 1.14 300 3.00 1 58 1.08 44_( 255) 45-50 0 0C - - 100 1.40 0.39 2.50 200 2.36 0.34 1.95 300 3.20 0.34 2.20 400 3.80 0.24 1.44 44 ( 25m) 18-20 0 0C - - 100 1.24 0.31 1.22 200 2.40 0.35 1.07 300 2.72 0.31 1.11 a200 watt mercury lamp filtered by pyrex. bAcetonitrile solution. CAnalysis by Vpc (4% QF—I) with biphenyl added as internal standard. A small amount of impurities (22 1%) with the same retention times as the rearrangement products made exact measurement difficult. 24 Table 5. Effect of gcetophenone and 2,4-pentadiene on photolysis of 5)5-dimethyl—2,3-epoxycyclohexanone. Ketone(s), (M) Solvent Tfiij %Reactionb 40, (0 .30) 2,4-pentadiene 6 14 .3C 40, (0 30) benzene 6 19.0 fl' (0 '28) benzene 6 0 .00‘1 Acetophenone,(0.42) a450 watt lamp, pyrex filter. bDisappearance of starting material was noted. Products did not give satisfactory analysis at concentration used. \ CAnalysis by Vpc (4% QF—I) with menthol added as internal standard. dAnalysis by vpc with acetophenone as standard. DISCUSSION 1. Radical Character of the Rearrangement A consideration of the B-dicarbonyl products produced by photolysis of the a,B-epoxy ketones studied in this investigation, combined with those reported previously,(1, 3,4,5) leads to the following order of migratory aptitudes for B-substituents: benzhydryl and benzyl > hydrogen > methylene > methyl >> phenyl. With the exception of hydro— gen, this order parallels the stability of the correspond— ing radicals and suggests that the fi-group assumes radical characteristics during its migration. Additional support for this view is found in the frag- mentation products, dibenzyl, dimedone, and a-benzyl di— ethylether, obtained from 3—benzyl-5,5—dimethyl-2,3-epoxy- cyclohexanone. The implication here is that the rearrange— ment either proceeds gig, or is diverted to a radical pair; and that in this case the relatively stable benzyl radical escapes the solvent cage. This homolytic fragmentation of a fi-group can be inter- preted in another manner, as illustrated for the case of 3-benzyl-5,5—dimethyl—2,3-epoxycyclohexanone: 25 26 O O\ O . OH Ph . ° hv Ph R-H Ph > > However, photolysis of this epoxy ketone in acetonitrile, a poor hydrogen atom source, gave essentially the same rate of reaction and just as much dibenzyl as photolysis in ether. Also, fragmentation products are still observed when triplet quenching dienes (0.1-9M) are added to the photolysis sample. Since bimolecular hydrogen abstraction by singlet excited states has been shown not to occur (15), this alternate mechanism may be discounted. The position of hydrogen in the order of migratory aptitudes argues against a general fragmentation mechanism, as hydrogen atoms are not normally formed in radical reac- tions. In the photolysis of 5,5—dimethyl—2,3-epoxycyclo- hexanone, an alternate mechanism involving abstraction of a fi-hydrogen could lead to the observed products. 0 . O 0 , 0 0 0 0 0 . RH R > —> > 1 1‘1 0 CHO o . O H 0 04h. . o 8 .. O —_> > 51 The most likely radical initiating species in this mech- anism is an excited carbonyl compound. Since photolysis in piperylene gave approximately the same rate of epoxy ketone disappearance as in benzene, and since no reaction occurred when acetophenone was used in an attempt to sensitize the reaction, there is no support for this mechanism. A two step radical reaction in the photolysis of 3—methyl-2,3—epoxycyclopentanone should favor either forma- tion of a six-membered heterocyclic compound or fragmenta- tion rather than the strained cyclobutane derivative. 0 ° 0 *0.” >1 ”W o o > 32 .53. O O c? % //“\\¢é’ + 0 <— 28 Stereospecific rearrangement of steroidal epoxy ketones .25 and g§_as well as gl_and g§_has also been cited (8) as evidence for a concerted mechanism rather than a two step radical fragmentation, which would be expected to lead to epimerization of C10 in g§_and 28. and C4 in gl_and.g§. The stereospecific photorearrangement of the pulegone ox- ides reported here offers the most convincing evidence for a concerted mechanism, since in this substrate the steric course of the reaction is less likely to be controlled by conformational barriers in the rest of the molecule. This is illustrated in greater detail on page 32. II. Photochemistry of the Rearrangement In recent years, several authors (16,17,18) have ad— vanced theories to account for photochemical transforma- tions of carbonyl compounds. There is general agreement that n,w* excitation of a carbonyl group produces excited singlet states which rapidly cascade to the zero vibrational level of the lowest excited singlet state [81]. This excited state may be deactivated through chemical reactions. conversion to the ground state SO by fluorscence, internal conversion and bimolecular energy transfer. In the photo— chemistry of carbonyl compounds, the most important mode of n,w* singlet state deactivation is intersystem crossing to the first excited triplet [T1]. The triplet state in turn may be deactivated through chemical transformations, conversion to So by intersystem crossing, phosphorescence, or bimolecular quenching. [SI] \/\/\ __ __J; Products Products sO :& The true photochemistry in the rearrangement of a,B-epoxy ketones centers about the formation of diradical intermediate 9 _§_the various modes of deactivation of excited states. Within these bounds, the most important competing paths should be intersystem crossing from both the triplet and singlet excited states: [TiP\V/\5> So + heat [SflW [T1]+ heat and bimolecular quenching of the excited triplet state: [T1]+Q >SO+Q*. All other modes of deactivation are too slow to compete with these (19). Both [$1] and [T1] appear to be theoretically ac- ceptable precursors for 9, From our attempts to quench or sensitize a triplet reaction, it would appear that 30 intermediate Q_is produced only from [SI]. However, other workers have reported (3,8,14) that photochemical rear- rangement of steroidal epoxy ketones is both sensitized (triphenylene, biacetyl, and B-dicarbonyl compounds were usedO, and to some extent quenched by piperylene. These results suggest that intermediate 9 is produced from both [5 1] and [Til-- A h" > [le > [T1] M ' 2 Products < The same workers (14) have also investigated tri-Ef butyl stannane photoreduction of their epoxy ketone sub— strates, and report that solvent piperylene completely quenches the rapid reduction of substrate as well as the slower rearrangment. This contrasts strikingly with ir— radiation in piperylene in the absence of stannane, which gave the rearrangement product in 15% yield. Our experiments with tri-nfbutyl stannane photoreduc— tion of 4-methyl-3,4-epoxy-2—pentanone show that photo- reduction of the epoxy ketone occurs only from the triplet excited state, a fact in accord with recent work on a similar photoreduction of acetone (15). The quantum yield for rearrangement of this epoxy ketone (m = 0.03) was unchanged by irradiation in cyclohexane, piperylene, or in piperylene and stannane. 31 Using the quantum yield for photoreduction (m = 0.65) as a measure of intersystem crossing HSiP\A,+[T1]+ heat) in epoxy ketones, demonstrates that 35% of the excited singlets must be deactivated by other means. Since the value of misc is low in comparison to acetone (misc > 0.99) (15L formation of Q_from[Sfl is left as the only at— tractive alternative for deactivation of[Sfl. If the quantum yield for [SI] > 9_ is 0.35, the low quantum yield for rearrangment (m = 0.03) is probably due to un— favorable competition between rearrangement and closing of the oxirane ring (kC > kr)° Our original objective in using tri-nfbutyl stannane was to trap diradical intermediate 9, But as photolysis of 4—methyl~3,4—epoxy—2-pentanone in piperylene and stan- nane gave no reduction, one can only assume that if 9_ exists, its lifetime is too short to permit bimolecular reactions. Good evidence for a short lived intermediate, or for a concerted process, is seen in the stereOSpecific rearrangement of the pulegone oxides where oxirane opening must be immediately followed by B-methyl migration before much rotation about the Cd-CB ‘bond can occur (k1 > k2). 32 (1 k1 g Q ‘ ——9 O O . O 54 k2 — (5 Cl 5 3 O The usual course of photochemical reductions of car- bonyl compounds begins with hydrogen abstraction by the excited state of the carbonyl oxygen, which has been at- tributed to its oxy-radical character (15). 'Hydrogen abstraction by an oxy-radical followed by oxirane ring Opening is plausible in view of the radical reduction of epibromohydrin by stannane to give allyl alcohol in high yield (20). The photoreduction of a,B-epoxy ketones does not, however, necessarily follow this path, as[Tfl could give a triplet intermediate 9} which could subsequently undergo reduction by stannane (path b). 33 Rearrangement \ Q /%‘2‘; OH O‘H If this latter mechanism (path b) operates, then inter- mediate gf is generated regardless of the presence of stannane. The facts presented in this thesis greatly re- strict the subsequent reactions of g}, and in fact argue against a significant contribution by path b, Thus, Q] can not react with dienes since no products of this nature are observed and epoxy ketone is not consumed by unknown reactions; 9] can not be converted to 9, since this would represent a triplet state precursor to rearrangement products; finally 9} would have to be very short lived (for a triplet state intermediate) to account for the slow isomerization of the pulegone oxides. EXPERIMENTAL I. General Procedures A. Apparatus All infrared spectra were obtained on a Perkin-Elmer Model 237B recording Spectrophotometer, using sodium chloride cells. The ultraviolet spectra were determined in 1 cm quartz cells using a Unicam SP.800 spectrophOh . tometer. Nuclear magnetic resonance speCtra were deter- mined in solution or neat using a Varian, A-60, high reso- lution spectrometer or a JEOL Co. C-60H high resolution instrument. All spectra were obtained at 60Mc using tetramethylsilane as an internal standard. -Mass spectra were taken on a Consolidated Electrodynamics Corp. Mass Spectrometer Type 21-103C. Vapor phase chromatography analyses were made using three different instruments: Aerograph A-90P and A—9OP3 instruments with thermal conductivity detectors using 1/4" x 6' columns, and a Varian Aerograph series 1200 Hy-Fi III instrument with a flame ionization detector using 1/8" x 6' columns. B. Melting Points Melting points were determined on a Koefler hot— stage and are uncorrected. 34 35 C. Microanalysis All microanalyses were performed by the Spang Micro- analytical Laboratory, Ann Arbor, Michigan. II. Preparation of a,B-epoxy Ketones A. Preparation of 3-benzyl-5,5-dimethyl-2,3-epoxycyclo- hexanone 1. Preparation of 5,5—dimethyl-3—methoxy-2—cyclo- hexene—l—one. A solution of dimedone (100 g, 0.715 mole) in 700 ml anhydrous methanol and 30 ml concentrated sulfuric acid was stirred under gentle reflux for 19 hours. Approximately 600 ml of the methanol was removed on a rotary evaporator. The residual solution was poured into one liter of 6% sodium hydroxide solution and extracted with three 300 ml portions of USP ether. The combined ether layers were washed with water until the wash was neutral and dried over magnesium sulfate. Filtration and removal of the ether left 65.7 grams of crude product. The aqueous portion was acidified with concentrated sulfuric acid and cooled in an ice-water bath whereupon unchanged dimedone crystallized. The crystals were collected by suction filtration, washed with water and dried in the air (yield 25 grams). 5,5-dimethyl—3-methoxy-2-cyclohexene-l-one was dis- tilled under reduced pressure and 55 g (50%) of pure pro- duct was obtained, bp 72-740/1.0 mm. 555x (CCl4) 1662, 1610 cm-1 36 2. Preparation of 3-benzyl—5,5-dimethyl-2-cyclo- hexene-l-one. Into a three-necked, 250 ml round-bottom flask equipped with a magnetic stirrer, reflux condenser and dropping funnel and having all outlets protected by drying tubes, was placed magnesium turnings (3.94 g, 0.162 mole) in 50 51 of anhydrous ether. .Benzyl chloride (20.5 g, 0.162 mole) in 50 ml of anhydrous ether was added slowly until a Spon- taneous reaction began. The flask was then placed in an ice-water bath and the remainder of the benzyl chloride solution was added dropwise. The Grignard complex was heated under gentle reflux (heating mantle) for 1/2 hour after the Spontaneous reaction had subsided. (The flask was then cooled in an ice-water bath and 50 ml of an ether solution containing 5,5-dimethyl-3-methoxy-2-cyclohexene- 1-one (25 g, 0.162 mole) was added dropwise with rapid stir- ring. The reaction mixture was stirred at reflux tempera- ture for an additional 2 hours, cooled and poured into a solution of 40 ml concentrated sulfuric acid in 200 ml of ice water. The ether layer was separated and the aqueous phase extracted twice more with 100 ml portions of ether. The combined ether layers were washed with three 75 ml portions of water and dried over magnesium sulfate. The solvent was removed on a rotary evaporator to yield 32.96 g of crude product. Vapor phase chromatography (5' x 1/4" 4% QF-l column, 195°, 50 cc/min) of the crude product showed peaks having retention times of 0.44 min (ether and 37 toluene), 2.84 min (the enol ether), 12.8 minutes (desired product) and three other peaks at 2.06 to 4.9 minutes which were later shown to be dibenzyl and the two dehydration products of the initial Grignard adduct. Integration of the chromatogram indicated a yield of 58%. The product was purified yia_Girard's T derivative (22), taking care to keep the reaction mixture at ice-water tem- perature while extracting all non-ketonic material. -Hydrol- ysis of the derivative left an orange oil which was further purified by distillation under reduced pressure to yield 15.4 9, bp 110°/0.2 mm; 43% of theory. -1 VMax (cc14) 1670 cm 3. Preparation of 3-benzyl-5,5-dimethyl-2,3-epoxy- cyclohexanone 5,5-Dimethyl-3-benzyl-2,3—epoxy-1-cyclohexanone was prepared from pure 5,5-dimethyl-3-benzyl-2-cyclohexene-1-one by the method of Wasson and House (23). An 80% yield of material distilling at 100-1050/0.05 mm was obtained. vMax (CCl4) 1720 cm_1. Nmr spectrum (Figure 1). B. Preparation of 4,4—diphenyl-2,3-epoxycyclohexanone 1. Preparation of stilbene glycol (24) Benzoin (100 g, 0.48 mole) was added to 500 ml of 95% ethanol and warmed slightly to aid solution. rTen grams of sodium borohydride was added in small portions (exothermic 38 reaction with foaming). The solution was then heated on a steam bath for 20 minutes, 500 ml of water was added and the solution was boiled for five minutes on a hot plate and filtered through a glass—wool plug into one liter of water. The glycol crystallized upon cooling and was col- lected by suction filtration and washed with water, mp 135-136°. 2. Preparation of diphenylacetaldehyde Diphenylacetaldehyde, bp 125-130°/0.5 mm, was prepared from the crude, wet glycol by the method of Henze and Leslie (25), i.e., in refluxing 25% sulfuric acid. Yield: 15.5% from benzoin. 3. Preparation of 4,4-diphenyl-2-cyclohexene—1-one 4,4-Diphenyl—2~cyclohexene-1-one, mp 91-92.5°, was prepared from diphenylacetaldehyde and methyl vinyl ketone in 53% yield by the method of Bordwell and Wellman (26). 4. Preparation of 4,4-diphenyl-2,3-epoxycyclo- hexanone 4,4-Diphenyl-2-cyclohexene-l-one (9.0 9, 0.0363 mole) was dissolved in fifty ml of 50% acetone and methanol. Hydrogen peroxide (10.5 ml of 30% solution) was added to the solution and the temperature rose from 20° to 32°. The solution was stirred in a water bath at 20° while one milliliter of 6N sodium hydroxide solution was added. »After a total reaction time of 1 1/2 hours, analysis by micro tlc 39 (silica G, benzene) showed no remaining starting material. The mixture was worked up as usual to yield a clear oil which was taken up in hot ligroin (60—900) and benzene. Slow cooling of the solution produced only an oil, but rapid cooling caused a granular white solid to precipitate, 8.0 g (83%), mp 87589.5°. ‘Further attempts to purify this material gave only the oil again. The infrared spectrum was identical to that of another sample (mp 72-73°) from a previous preparation purified by column chromatography and recrystallization from ligroin. (CC14) 1713 cm—1. 3 Max C. Preparation of 5,5-dimethyl-2,3-epoxycyclohexanone 1. Preparation of 5,5-dimethyl—2-cyclohexene-1-one 5,5-Dimethy1-3-methoxy—2-cyclohexene-1—one (15.4 g, 0.10 mole) was reduced with 1.9 g of lithium aluminum hydride by the method of Gannon and House (27) to yield 10.7 g (87%) of 5,5-dimethyl-2-cyclohexene—1-one, bp 63°/7.5 mm. (ccl4) 1675 cm’l. vMax 2. Preparation of 5,5-dimethyl-2,3—epoxycyclohexanone 5,5-Dimethyl-2-cyclohexene—1-one (10.7 g, 0.087 mole) was oxidized with 20 ml of 30% hydrogen peroxide by a modi- fication (total reaction time was 1.5 hr at 8-12°) of the method of Wasson and House (23) to yield 10.05 g (83%) of 5,5-dimethyl-2,3-epoxy—1-cyclohexanone, bp 80°/8 mm. 1 vMax (cc14) 1720 cm 40 D. Preparation of 2,3-epoxycyclopentanone 1. Preparation of 2-cyclopentene-1-one The procedure devised by Garbish (28) was used to prepare 2—cyclopentenone via the 2-bromo ethylene ketal of 1 cyc10pentanone in 26% yield. vMax (CCl4) 1707 cm- 2. Preparation of 2,3-epoxycyclopentanone 2-Cyclopentene—1-one (20.5 g, 0.25 mole) was dissolved in 230 ml of methanol and cooled to 4° in an ice-bath. One milliliter of 6N sodium hydroxide solution was added and 26 ml of 30% hydrogen peroxide was added over a period of 35 minutes, keeping the temperature between 5—100. The ultraviolet spectrum at this time showed no alpha-beta unsaturated ketone (xmax 218 mu, 6 9500). After the tem- perature had dropped back to 4°, the mixture was poured into 300 ml of water containing 95 g of sodium chloride. The aqueous mixture was extracted with three portions of ether and the combined ether extracts were washed with two 50 ml portions of saturated sodium chloride solution fol- lowed by one 50 ml portion of water. The ether solution was then dried over magnesium sulfate, filtered and the solvent removed by distillation. The residue was distilled on the spinning band column at 30 mm and gave the following fractions: 41 1 24-46° (cyclopentenone) 1.1g 2. 46-740 3. 74-790 4. 79-82° 2.0 g (contains some cyclopentenone) 5. 82° 3 .0 g 6. 82° 2.5 g (overall yield 30.5%) fractions 4-6. VMax (c014) 1750 cm—1. E. Preparation of 3-methyl-2,3-epoxycyclopentanone 1. Preparation of 3-methyl-2—cyclopentene—1—one .3—Methyl-2-cyclppentenene-l-one, bp 80-82°/18 mm, was prepared from 120 g of acetonylacetone in 52% yield by the method of Acheson and Robinson (29). The only modification was the use of saturated sodium chloride solutions for all extractions and washes. Vpc retention times of the starting material and product were the same on 4% QF-l and 20% -1 SE-30 columns. VMax (CCl4) 1707 cm 2. Preparation of 3-methyl—2,3-epoxycyclopentanone 3-Methyl-2-cyclopentene-1-one (4.8 g, 0.05 mole) was dissolved in 45 ml of methanol, twelve drops of 6N sodium hydroxide solutions was added and the solution was cooled in an ice-bath; 30% hydrogen peroxide was added in portions and ultraviolet spectra were taken after each addition by taking aliquots of 0.5-0.6 mihroliters and diluting these to 5 ml in methanol. 42 Time H202 added NaOH added Aliquot Absorbance (min) (ml) __ (Lil) - 5 0.00 12 drops 0.50 1.95 20 3.00 -------- 0.53 0.88 40 5.00 -------- 0.55 0.60 60 7.00 -------- 0.56 0.34 80 9.00 -------- 0.58 0.22 100 11.00 16 drops 0.60 0.10 120 --- -------- 0.60 0.06 140 --- -------- 0.60 0.005 The reaction mixture was poured into 100 ml of satur- ated sodium chloride solution, extracted with three portions of ether and the combined ether extracts were washed with saturated sodium chloride and dried over magnesium sulfate. The ether was removed by distillation to yield 4.6 g of crude product. Analysis by vpc (20% SE-30, 135°) showed only one peak at 4.8 min. Following the above procedure, 24 g (0.25 mole) of 3— methyl-Z-cyclopentene-1-one gave 14 g (51%) of 3-methyl- 2,3-epoxy—1-cyclopentanone distilled on a spinning band column at reduced pressure and pure according to vpc analysis. (CC14) 1745 cm-1. vMax F. Other epoxy ketones Samples of the pulegone oxides (30), isophorone oxide (31), and 4-methyl-3,4-epoxy-2-pentanone (32) were supplied by Dr. William Reusch. 43 III. Photolysis of 3-benzyl-5,5-dimethy152,3—epoxy- cyclohexanone A degassed solution of 3-benzyl-5,5-dimethyl-2,3- epoxycyclopentanone in 300 ml of anhydrous ether was ir- radiated for 4 hr with a 450 watt mercury lamp filtered by Corex. Base soluble materials were extracted with cold 5% sodium hysroxide solution. -Treatment of the basic extracts with 6N hydrochloric acid gave a white precipitate (0.39 g) which was recrystallized from acetone and water, needles mp 155-1570. Analysis of the photomixture by Vpc (4% QF-I) before and after basic extraction showed two base soluble com- pounds, unchanged epoxyketone and several smaller peaks. A. Photoproducts from 3-benzyl—5,5-dimethyl-2,3-epoxy- cyclohexanone 1. Benzyl dimedone 45 The crystalline base soluble material, mp 155-157°, was shown to be 2-benzyl dimedone by an undepressed mixed melting point with authentic 2-benzyl dimedone. .The photo- product also had an infrared spectrum identical to that of the authentic material. 2. Dimedone 46 Extraction of the ether photomixture from a similar photolysis of 3-benzyl-5,5-dimethyl-2,3-epoxycyclohexanone 44 (2.0 g) with three 20 ml portions of UM sodium acetate solu- tion followed by treatment of the combined extracts with 2.5 ml of concentrated hydrochloric acid, 25 ml water, and 10 ml of 38% formaldehyde solution gave an immediate pre- cipitate which was collected after 16 hrs to yield 0.1173 g of formalin dimedone, identified by a comparison of melting point (188.5-190°) and infrared Spectra with authentic material. 3. Dibenzyl 41 The base insoluble photOproducts.were chromatographed on silica gel and gave dibenzyl in the first fractions eluted with hexane. The product was further purified by sublimation and found identical to authentic dibenzyl by mixed melting point and comparison of infrared spectra. 4. a-benzyl diethylether 48 The base insoluble photoproducts were fractionated in a micro distillation apparatus equipped with a dry-ice trap. Analysis of the distilled material by vpc showed a mixture of dibenzyl and a more volatile compound. ‘The latter was collected on a DEGS column at 160° and proved to be a-benzyl diethylether identical to an authentic sample in its infrared and nmr spectra (Figures 2 and 3). 45 B. Photolysis of neat samples Samples of 3-benzyl-5,5—dimethyl-2,3-epoxycyclohexanone (ca. 5 ul) were placed between two glass microscope slides (optical density: 95% transmittance at 3400 R, 40% at 3000 X, 0% at 2800 X). The slides were then attached to a vycor lamp holder with rubber bands and irradiated with a 450 watt lamp through a Corex filter. At intervals of 1/2 hr the slides were removed, washed down with ether, and the solution checked by Vpc (4% QF-l, 190°). *All chromatdgrams showed a ratio of dibenzyl to dimedone of approximately 1:2. -The residues from four of the slides were dissolved in ether, extracted with cold 5% sodium hydroxide solution, washed with water and dried over magnesium sulfate. Analysis of the baserinsoluble ether solution by vpc showed that the peaks corresponding to dimedone and benzyl dimedone had disappeared. 5C. -Separation of a dimedone and benzyl dimedone mixture Solutions of dimedone and benzyl dimedone in 100 ml of ether were prepared as follows: Sample dimedone (g) benzyldimedone (g) I 0.1134 none II 0.1075 0.1612 III none 0.1655 46 Each ether solution was extracted with three 20 ml portions of 1M sodium acetate solution, followed by one 30 ml portion of cold 5% sodium hydroxide solution. 1T0 the sodium acetate extracts from each sample was added 2.5 ml of 12N (conc) hydrochloric acid, 25 ml of water, and 10 ml of 38% formaldehyde solution. Samples I and II gave almost immediate precipitates upon the addition of formaldehyde while sample III gave no precipitate after 24 hours. The sodium hydroxide extracts were acidified with 6N hydrochloric acid. Sample I gave no precipitate, Samples II and III gave immediate precipitates. The formalin dimedone precipitates were collected from samples I and II, washed with water and set to dry in the air. Sample I gave 0.0935 g(78.3%). Sample II gave 0.0899 g (80.1%). -Melting point of the formalin dimedone prepared in this way was 188.5-190°. Infared spectra of the two samples were identical. -D. Photolysis in other solvents 1. Photolysis in pentane 3—Benzyl—5,5-dimethyl-2,3-epoxycyclohexanone (2.0 g) was irradiated in 300 ml of pentane for 4 1/2 hr With a 450 watt mepcury lamp filtered by Corex. A precipitate on the walls of the apparatus was shown to be a mixture of dimedone and benzyl dimedone using the procedure under C. There was no base soluble material in the pentane solution. Vpc analysis of base insoluble residues showed dibenzyl, 47 starting epoxyketone and materials of approximately the same retention time as a~benzyl diethyl ether (assumed to be benzyl pentane isomers). 2. Photolysis in 0.1M piperylene Irradiation of an ether solution 0.015M in epoxy ketone and 0.1M in piperylene gave essentially the same product mixture as before according to Vpc analysis. E. -Mass-balance of photo—products from 3-benzyl-5,5-di- methyl-2,3—epoxy-1-cyclohexanone in diethyl ether. Integration of the vpc chromatogram gives the following results: compound at; 45 .45 .42. .26. M;w. 230 230 164 182 140 Grams a Isolated --- 0.3891 --- --- 0.1424 Integrated Area (Rel.) 630 428 145 233 254 M0193 0.00249 0.00169 0.00057 0.000092 0.00102 -Present Grams 0.57 0.3891 0.093 0.167 0.1424 Present Yield (%)b 28.6 19.4 6.5 21.0 11.7 aDimedoneanalysis: The amount of dimedone isolated (calcu— lated from formalin dimedone) was actually 0.1140 grams but the procedure has been shown to yield only about 80% of the theoretical amount. -Dividing 0.1140 by 0.80 results in the value of 0.1424 which is in good agreement with the vpc integration ratios. 48 bThese data account for 75.5% of the starting material. Unidentified materials are estimated to account for 15-20% of the starting material (vpc analysis). IV. Photolysis of Benzyl Dimedone A. Preparation of benzyl dimedone A solution of dimedone (5 g, 0.036 mole) in 8 ml of 20% potassium hydroxide containing potassium iodide (0.25 g) wasuflkylated with benzyl chloride (4.5 g, 0.036 mole) ac- cording the method of Stelter, KesSeler, and Meisel.(9); After heating under reflux for 1 1/2 hr the mixture was cooled, poured into 53 ml of.3% sodium hydroxide, ex- tracted with ether, and the basic layer treated with 6N _ hydrochloric acid to yield a yellow-white precipitate. -This was collected, washed with water and recrystallized from acetone and water; yield: 6.3 g (76%), needles, mp 155-157°. B. Photolysis of benzyl dimedone An ether solution of benzyl dimedone (0.4954 9, 0.00216 mole) under a nitrogen atmosphere was irradiated for 4 hr with a 450 watt mercury lamp filtered by Corex. The solution was extracted with five 40 ml portions of cold 3% sodium hydroxide solution, washed with water, and dried over magnesium sulfate. Base soluble extracts were treated with 6N hydrochloric acid and a white precipitate was collected and shown to be unchanged 1-benzyl dimedone. 49 Removal of the ether from the base insoluble layer gave -a small amount of residue which was combined with that from previous photolyses and micro distilled. The distillate gave (33, 10 ul) a-benzyl diethyl ether by preparative vpc (4% QF-I, 148°), identified by its infrared spectrum. V. Photolysis of 4,4—diphenyl-2,3—epoxycyclohexanone An ether solution of 4,4-diphenyl-2,3-epoxycyclohex- anone (2.0 g) was irradiated for 3 1/2 hr with a 450 watt mercury lamp filtered by Corex. The solution was extracted with four 50 ml portions of cold 5% sodium hydroxide. The combined basic extracts were saturated with sodium chloride and washed with ether, then treated with 6N hydrochloric acid to yield 0.93 g of precipitate. -The base insoluble ether layer was washed with water, dried over magnesium sul- fate and the solvent removed to yield 0.97 g of a yellow oil which crystallized on standing. This material was identi- fied by infrared and vpc analysis as the unchanged epoxy ketone. A. Photoproducts from 4,4-diphenyl—2,3-epoxycyclohexanone 1. 4,4-Diphenyl-1,3-cyclohexanedione The base soluble precipitate was crushed under 7 ml hot carbon tetrachloride and filtered to give 0.167 g of precipitate. ~The filtrate yielded a brown oil (0.78 g). The carbon tetrachloride insoluble precipitate was recrystallized from acetone and water to yield needles 50 m.p. 191-193° (dec). Anal. ‘C31C'd for C18H1602: C, 81.792'H, 6.10% Found: C, 81.80? H, 6.13% 2. 3,3-Diphenyl-2-hydroxymethylenecyclopentanone The carbon tetrachloride soluble filtrate was chromato- graphed on 10 g silica gel with 75 ml of methylene chloride. The effluent yielded a solid substance which was recrystal- lized from acetone and water, needles, mp 94—960. The infrared spectrum of this compound, 3,3-dipheny1—2-hydroxy- methylenecyc10pentanone, is shown in Figure 5. 5221. 'Calc'd for C13H1302: C, 81.79; H, 6.10% Found: C, 81.63; H, 6.15% B. :Derivatives of the photoproducts 4y4-diphenyl-2,3-epoxycyclohexanone (1.0068 g) was_ irradiated for 2 1/2 hr as before. The basic extracts were made neutral to pH paper with 6N hydrochloric acid. -Half of the cloudy suSpension was treated with formaldehyde solution to give a formalin derivative (0.0131 g, 2.6%) with an identical infrared spectrum (Figure 6) to that of the formalin derivative prepared from 4,4-diphenyl-1,3- cyclohexanedione. One fourth of the neutral suspension was treated with + 2 ml of c0pper acetate solution containing 0.2632 meqCu+ . The chelate was filtered and the filtrate iodometrically 51 determined to contain 0.2221 meq Cu++. Thus, the sample contained 0.0411 meq of chelating B-dicarbonyl compound and the yield of 3,3-diphenyl-2-hydroxymethylenecyclopentanone was calculated as 43%. In a separate experiment, 3,3~diphenyl-2-hydroxymethyl- enecyclopentanone was treated with excess copper acetate and gave a green precipitate, mp 131-1330. The infrared Spectrum of this compound exhibits absorption at V (CCl4) max 1600, 1463 and 1340 cm”1 VI. Photolysis of 5,5-dimethyl-2,3-epoxycyclohexanone Irradiation of an ether solution of 5,5-dimethyl—2,3- epoxycyclohexanone for 90 min with a 450 watt mercury lamp filtered by Corex gave the following compounds according to Vpc analysis (4% QF-I, 138°): Compound Retention Time(min) Relative Area '51 2.3 0.058 49_ 3.4 1.00 .42 8.8 0.595 The volume of solvent was reduced to approximately 20 ml and precipitated dimedone was collected by suction fil- tration. The filtrate was then extracted with cold 5% sodium hydroxide solution. The basic extract was treated with 6N hydrochloric acid and the aqueous suspension ex- tracted with ether. -The ether extract was washed with sodium chloride solution and dried over magnesium sulfate. 52 A. Photoproducts from 5,5-dimethyl-2,3-epoxycyclohexanone 1. Dimedone The precipitate from the photolysis was identified as dimedone by a comparison of its infrared Spectrum eith that of authentic material, yield: 0.44 g, 25%. In another experiment the yield of formalin dimedone by extraction with sodium acetate solutions was 31%. 2. 4,4-Dimethyl-2-hydroxymethylenecyclopentanone The ether solution of base soluble photoproducts gave an oil which upon sublimation at reduced pressure yielded crystals, mp 85-88°(dec), identified as 4,4—dimethyl-2— hydroxymethylene 51. by infrared spectrum (Figure 7) and nmr spectrum (Figure 8). This compound also exhibited a positive ferric chloride test. 3. Basic cleavage of 4,4-dimethyl—2—hydroxymethylene- cyclopentanone (33) 4,4-Dimethyl-2-hydroxymethylenecyclopentanone (200 mg) was placed in 20 ml 10% sodium hydroxide solution and heated at reflux for 1/2 hr. -The solution was cooled, ex— tracted with ether, and the ether extract was washed with saturated sodium chloride solution, and dried over magnesium sulfate. 'Removal of the solvent left a residue containing only one component by Vpc analysis. The product was shown to be 3,3—dimethylcyclopentanone by comparison of its infra- red spectrum with that of authentic material. 53 VII. Photolysis of 2,3-epoxycyclopentanone An ether solution of 2,3-epoxycyclopentanone (2-7 g, 0.0276 mole) was irradiated for 1/2 hr with a 450 watt mercury lamp filtered by Corex. A considerable amount of white precipitate had collected on the walls of the appara- tus making a longer irradiation time impractical. The precipitate was collected to give 0.1260 g of material (4.7%), mp 149-151°. The clear ether filtrate was extracted with cold 5% sodium hydroxide and gave an orange basic solution,-which upon treatment with 6N hydrochloric acid gave an orange suspension. This suspension was filtered and extracted with ether. The ether extract was washed with saturated sodium chloride solution and dried over magnesium sulfate. Removal of the solvent left approximately 1 g of an orange oil. A.‘ Photoproducts of 2,3-epoxycyclopentanone 1. Cyclopentane—1,3-dione The precipitate mp 149-151° has the same melting point as that reported for cyclopentane—1,3-dione (10). An infrared spectrum of this material shows 3;:gl3 1590 cm."1 and 1650 cm-1 indicative of-a B-dicarbonyl compound. 2. Unknown compound "X" The base soluble oil was microdistilled and the volatile material showed a major peak upon vpc inspection. 54 Preparative vpc gave enough compound for an infrared spectrum (Figure 9) and nmr spectrum (Figure 10) but no structure has been assigned from this data. VIII. Photolysis of 3—methyl-2,3-epoxycyclopentanone An ether solution of 3-methyl—2,3-epoxycyclopentanone (3.9 g, 0.035 mole) was irradiated for 2 hr with a 450 watt mercury lamp filtered by Corex. The ether solution was extracted with cold 5% sodium hydroxide solution. The basic extract was washed with ether and treated with 6N hydrochloric acid. The aqueous phase was then extracted with ether, the ether extracts washed with saturated sodium chloride solution and dried over magnesium sulfate. Re- moval of the solvent left 1.25 g residue. When a similar epoxy ketone solution was irradiated for 4 hr, only acetic acid was found as a base soluble product. In another photolysis, the product was continuously extracted by a 100 ml layer of 2% sodium hydroxide during the irradiation period. A 40% yield of base soluble material was obtained by the previous workup. A. Photoproducts from 3-methyl-2,3-epoxycyclopentanone 1. Acetic acid The base soluble material from a 4 hr irradiation of 3-methyl—2,3—epoxycyc10pentanone-was distilled at reduced 55 pressure to yield volatile material from which acetic acid was collected by Vpc. Identification was by comparison of infrared and nmr spectra with authentic acetic acid. 2. 2-Acetyl-l-hydroxycyclobutene The base sqfluble residue from photolysis of 3-methyl- 2,3-epoxycyclopentanone was sublimed at reduced pressure yielding white crystals, mp 104-104.5°. This compound was not 2-methyl-1,3-cyclopentanedione, reported mp 214—215° (34). The compound gave positive ferric chloride and iodo—l form tests. When an alcohol solution of the compound was treated with an excess of copper acetate solution, a color change from blue to green was noted, but no precipitate formed. An nmr Spectrum (Figure 11) infrared spectrum (Figure 12) and mass spectrum (Table 6): Table 6. Mass spectrum of 2-acetyl-1-hydroxycyclobutene mZe Rel. Abund. mZe Rel.-Abund. 27 83 55 89 39 60 69 89 41 100 83 46 43 71.5 112 46 all support 2-acetyl-1—hydroxycyclobutene as the structure of this compound. Anal. Calc'd for C3H1002: c, 64.27; H, 7.19% Found: c, 64.33; H, 7.23% 56 B.- Preparation of l—methoxy-Z-acetylcyclobutene An ether solution of 2-acetylcyclobutanone (0.5 g, 0.0045 mole) was treated with an excess of ethereal diazo- methane. After the yellow color had disappeared from the solution, a few drops of acetic acid were added and the solution was extracted with cold 5% sodium hydroxide, washed with saturated sodium chloride solution and dried over magnesium sulfate. The solvent was removed leaving a residue showing only one peak according to vpc analysis. Preparative vpc gave enough pure compound for an infrared spectrum (Figure 14) and nmr spectrum (Figure 13). These spectra support the structure of the methylated photoproduct as 1-methoxy-2- acetylcyclobutene. IX. Photolysis of 2-acetyl-1-hydroxycyclobutene 2-Acetyl-l-hydroxycyclobutene (0.050 g, 0.00044 mole) was placed in a pyrex test tube and dissolved in approxi- mately 3 ml of anhydrous ether. After the solution was flushed briefly with nitrogen, the tube was stoppered and attached to a vycor immersion lamp holder. Irradiation for 1 1/2 hr gave no reaction as determined by Vpc inspection. Evaporation of the solvent left a residue of pure white crystals mp 104°. 57 X. Photolysis of the Pulegone Oxides A. Photolysis of B-pulegone oxide A solution of fi-pulegone oxide (2.9 g, 0.0172 mole) in 300 ml ether was irradiated for three hours with a 450 watt mercury lamp filtered by Corex. Analysis on a 4% QF-I column (135°) indicated approximately 5% conversion to products. The solvent was removed and the residue' distilled on a short-path apparatus at 5 mm up to 90°. The condenser was rinsed with ether and combined distillates purified by preparative vpc to yield several unidentified low boiling fractions and two photorearrangement products. B. Photoproducts of fi-pulegone oxide 1. 2-Acetyl-2,5—dimethylcyclohexanone stereoisomers The two rearrangement isbmers were collected from a 4% QF-I column at 130°, retention times 4.3 and 5.4 min. The infrared spectrum (Figure 16) of the compound with a reten— tion time of 4.3 min was identical to that of the minor ozonolysis product .55 of a 4-methylfiigg-pulegone isomer mixture (11). The infrared spectrum (Figure 15) of the compound with a retention time of 5.4 min was identical to the major ozonolysis product 54, Retention times of the ozonolysis products were the same as the photoproducts. 2. a-Pulegone oxide a-Pulegone oxide was not isolated from the photomixture. Identification wasnmde by a comparison of Vpc retention 58 time with authentic a-pulegone oxide. C. Photolysis of a—pulegone oxide A solution of a-pulegone oxide 0.25M in 3 ml aceto- nitrile was irradiated in a sealed pyrex tube for 6 hr. Analysis by vpc demonstrated the presence of compounds hav- ing the same retention times as B—pulegone oxide and both stereoisomeric rearrangement products. The extent of re- action was approximately 3%. XI. Photoreduction of 4—methyl-3,4-epoxy-2-pentanone with tri-nfbutyl stannane (a) A solution of 4-methyl-3,4-epoxy—2-pentanone (1.23 g, 0.0108 mole) and tri-nfbutyl stannane (5.82 g, 0.020 mole) in 90 ml of anhydrous ether was irradiated for 2 1/4 hr with a 450 watt mercury lamp fitted with a Corex filter. The solution was agitated by a continuous stream of nitrogen. After filtering the solution, the solvent was removed by a rotary evaporator and the residue was washed with three portions of water (both diacetone alcohol and the_ starting epoxy ketone are water soluble). The combined water washes were filteredyrsaturated'wifh sodium chloride and extracted with ether to yield 23, one gram of product. This was distilled and purified by vpc (4% QF—I). Two com- ponents were present; the major (22: 80%) was diacetone alcbhol identified by its infrared spectrum, and the minor was unchanged starting material. 59 (b) A solution of 4-methyl-3,4—epoxy-2-pentanone (1.14 g, 0.01 mole) and tri—nfbutyl stannane (5.82 g, 0.02 mole) in 90 ml of pentane was irradiated as above for two hours. The solvent was removed and the residue was chromato- graphed on a silica gel column 4.5 cm in diameter by 3.5 cm in depth. The butyl tin compounds were eluted with pentane and the remaining organic material was eluted with 300 ml of ether. Most of the ether was removed on a rotary evaporator to yield 1.2 g of residue which upon vpc inspec- tion (4% QF-I, 115°) contained only ether (32, 10%) and diacetone alcohol. XII. Preparation of 3,3-dimethylcyclopentanone (11) A. 3,3-dimethylcyclohexanol 3,3—Dimethylcyclohexanol was prepared by the method of Doering (35). Dimedone (94 g) was dissolved in 100 ml of purified acetic acid and reduced over platinum oxide to give 50 g (58.5%) of the alcohol, bp 85°/13mm. B. B,B-dimethyladipic acid 5,5-dimethyladipic acid was prepared by the nitric acid oxidation of 3,3-dimethy1-cyclohexanol according to the procedure of Ellis (36). A 59% yield was obtained, mp 80-84°. 60 C. 3,3-Dimethylcyclopentanone 3,3-Dimethylcyclopentanone was prepared by the method of Pines (37). B,B—Dimethyl adipic acid (40 g) and 2 g of barium hydroxide were heated to 285°. The distillate gave 14.2 g (55%) of the cyclopentanone bp 29°/5 mm. n34 1.4322. XIII. Preparation of aebenzyl Diethylether A. 1-Phenyl-2-propanol (38) Beduction of phenylacetone (26.8 g, 0.20 mole) by an excess of lithium aluminum hydride in anhydrous ether gave l-phenyl-Z-propanol bp 98°/10 mm in 75% yield. B. a-Benzyldiethylether Ethyl iodide (10 g, 0.065 mole) was added to an anhy- drous tetrahydrofuran suspension of the sodium salt of 1- phenyl-2-propanol (6.8 g - prepared by adding an equivalent of sodium hydride to the alcohol) to yield the desired ether contaminated by the starting alcohol. After removing the solvent and a crude distillation, a-benzyl diethyl ether was collected by preparative vpc (20% DEGS, 160°) . An infrared spectrum (Figure 2) and nmr spectrum (Figure 3) were identical to those of the photoproduct. XIV. Preparation of tri-nfbutylstannane (39) Treatment of tri-gfbutyl tin chloride (112 g, 0.345 mole) in 400 ml of ether with lithium aluminum hydride 61 (5.0 g, 0.13 mole) gave tri-nfbutyl stannane distilling at 71-740/o.5 mm; 80 g, 80% yield. XV. Quantitative Measurements of Photolysis Rates Samples of the materia1(s) to be irradiated were ac- curately weighed and solutions made up volumetrically to insure known concentrations. A known quantity of a sub- stance having a suitable Vpc retention time and absorbtion Spectrum was usually added as an internal standard for vpc analysis. Piperylene and 2,5-dimethy1r2,3—hexadiene were purified by distillation, while other hydrocarbon solvents were purified by treatment with concentrated sulfuric acid fol- lowed by distillation from phosphorus pentoxide. Aceto— nitrile was Fisher certified grade and ether was Fisher anhydrous grade. Solutions were prepared for irradiation by placing 22.3 ml into 13 mm pyrex tubes which were then degassed and sealed. Irradiation was in a light equilibrating water bath (13) using either a 200 or 450 watt Hanovia medium pressure mercury lamp equipped with appropriate filters in a water cooled (or heated) vycor or quartz immersion well. Conversion to products was kept small as some of them exhibit photochemical instability. .mcocmxmzoaohohxommlm.Nlamflpmfiflplm.mlammcmnlm mo ESHpummm HEZ .H musmflm OH m w h m m w m N 62 HI EU xocmsvwam ccc ccca ccma ccva ccca ccca cccm _ q _ _ q a .Hmzumaanumac Hmucmnlo mo Enauummm UmamamcH .N madmam Tu I E U 8 mL TL. 1- 1. P. 3 U 6 D 8 S % C 3 1 ON“ ow om Ow. OOH 64 -"~-'- .Hmflumamzumaw Hmwcmfllo mo Esauummm MEZ .m madmam m N P b w \W, . I .i .4 aL.__.. 65 .mcocmucmmoachowcmamnumE>xoapmnlNiamcmflmavlm.m mo Edapommm HEZ .w madman 66 aIEU cow OOOH OONH u aw .mcocmpcwmoaumomcmahsawe Imxoawhnlmlamcmnmawlm.m mo Edauummm UmamHmGH .mocmsvmam oova a .m masmam OOGH . oowH u ,cccm 4.1 O V 'aoueqqrmsuexm om % icca EU .mucmsvwam HI ccma cccm cccm cccm cccm cccw - d . . q a u o .maoameManoaoholm.Huamcmnmanlfi.v mo m>aum>aam© GHHmanm may no Enauommm owamawcH .mm madmam «ON I 1 e w ch m T.- 3. Q 7 E 6 U o a dPnB® m nooa 68 also .mucmswmam oow coca coma oova coma coma ooom _ u a a a H O .wcoaomcmxw£0ao>UIm.Hlamcmsmanlv.¢ mo m>aum>aamw caamfiaom on» no Edauommm commamcH .Qm madmam -ON % ’aoueanmsuezm L To ‘w 00H __1_ aIEU .mocmdvmum , coma cooN comm .ccom comm cccv H 4 _ a — o ) . -iom Tu m 9 no 3 m i T. icv 3 1 e u o a dfl loo 1 now .wcocmucmmoHohomcwawzywemxoaomnlmlamnumanlv.v mo Enauummm omamamcH .mu musmam ncca aIEu .mucmdqmam com coca coma coca coma coma coo. a _ u a a a 0 com 2 5 m 2 x e w ca. m TL. I a 3 K w. m a my co co .mcocmuchOHUMUmcmaanumfihxoacmnlNlamflpwEaclw.w mo Enauoomm commamca .Qb mafimam coa 71 .wcocmucmmoaomumcmamnumemxoaowslNlaazumfiaclv.v mo Edauommm HEZ .w madmam oa o A w b. _ m N a a a J1 a o q a a a/ a 72 coma cch comm d H IE0 .mocmsomam coom comm — a .:X: vasomeoo mo Edupummm nonmawca .mm madmam cccc % 'aoueqqrmsuexm :17 ncv ncca a :80 .mocodqomam cow coca cha ccva coca coma och 1 a + d u u o .=x: QGSOQEOU mo Edauomma cmHmHmcH .Qm madmam low Tu 1 cu. slow m To 1' m m w a C 7... now coca 74 .:N: UQSOQEOU mo Esapommm HEZ .ca madmam -( I-fl \ § -0, 75 ca .mcmp590aowohxoaomnlalawmeMIN mo ESHuommm HEZ .aa madmam m m b r m N H a a a a _ iii??? 7», _.._——-———, ’—~~a ._A aIEU .mocmsvmam coca ocNa ccva coma ccma cccm comm 4 a . . a . 7 . O 3 . 8 Tu I E U S m T. I 1. 1- ‘ W‘rOfi D 3 dfi m g 2 1 cm .maUSQOachomxouomSIalamumOMIN Mo Enauummm cmamamcH .Na mascam L. cca 77 .mamUSQOaomomxozwmalalamumUMIN mo Edapommm HEZ .ma mascam ca m h m m P 4 a — 4 a a \. aIEU .mucmscmam ccoa coma cova coma coma coom _ cocm a a q a — .wcmudnoaomomxonumEIa IamumUMIN mo Esauommm commamca .oa madmam 78 aIEo .mucmdvmum com coca coma coma coma coma Doc a a a a a a o xflm rwcocwmeOaUmo IamumUMImlamnquacnm.m mo Edauommm Umamumca, .ma mascam 1cm To 1 w. s .c¢ m T. 1. m as 7 w e . M” I Om .cm as 80 aIEU .mocmsvmam 00w OOOH OONH oovH OOGH OOQH OOON 4 a u 4 a _ .mm. .mcocmxmsoa0>uahumom -cm ImlamnumEaCIm.m mo Esauommm commamca .ma madmam .cv I 1 e u S m I n. r m. .cm u o a c dp .cm 1cca H 10. 11. 12. 13. 14. 15. 16. LITERATURE CITED S. Bodforss, Ber., 51“ 214 (1918) H. E. Zimmerman, Abstracts of the 17th. National Organic ChemistryiSymposium, June 1961, Bloomington, Indiana, p. 31. O. Jeger, K. SChaffner and H. Wehrli, Pure Appl. Chem., 9” 555 (1964). C. Johnson, B. Domin and W. Reusch, J. Amer. Chem. Soc., 82“ 3894 (1963 H. E. Zimmerman and co-workers, J. Amer. Chem. Soc., §§, 947 (1964). J. G. Calvert and J. N. Pitts, Jr., Photochemistr 382-389, John Wiley and Sons, Inc., New York, (19 6). F. D. Green, W. Adam and G. Knudsen, Jr., J. Org. Chem., 31, 2087 (1966). . Je er and co-workers, Helv. Chim. Acta., 42” 2218 (1966. H. Stelter, H. Kessler and H. Meisel, Ber., 81” 1617 (1954 C. H. Depuy and E. F. Zaweski, J. Amer. Chem. Soc., 22, 3923 (1957). C. K. Johnson, Doctoral Thesis, Michigan State Uni- versity, (1963). M. R. Cox, H. P. Koch and W. B. Whalley, Chem. Comm., 212 (1967). G. S. Hammond and co-workers, J. Amer. Chem. Soc., 86, 3216 (1964). O. Jeger and co-workers, Helv. Chim. Acta. (1967) in print. P. J. Wagner, J. Amer. Chem. Soc., 82” 2503 (1967). H. E. Zimmerman, Advances in Photochemistry, Vol. 1, 183, Ed. W. A. Noyes, Jr., G. Hammond and H. N. Pitts, Jr., Interscience, New York, (1963). 81 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 82 N. J. Turro, Molecular Photochemistry, 1-88, W. A. Benjamin, Inc., New York, (1965). J. G. Calvert and J. N. Pitts Jr., Photochemistr , 293, John Wiley and Sons, Inc., New York, (1966). -N. J. Turro, Chem. and Eng. News, 84, May 8, 1967. H. G. Kuivila, Adv. in Organometallic Chem., Vol. 1, 77, F. G. A. Stone and R. West, Academic Press, New York, 1964. VK. Arakawa and M. Irie, C.A., 52, 15446a (1958). .L. F. Fieser, Organic Experiments, 99, D. C. Heath and Co., Boston, 1964. R. L. Wasson and H. 0. House, Org. Syn., Coll. Vol. 4, 552 (1963). L. F. Fieser, Organic Experiments, 216, D. C. Heath and Co., Boston 1964. H. R. Henze and W. B. Leslie, J. Org. Chem., 15” 901 (1950). F. G. Bordwell and K. M. Wellman, J. Org. Chem., 28” 2544 (1963). W. F. Gannon and H. 0. House, Org. Syn., 42” 14 (1960). E. W. Garbish Jr., J. Org. Chem., 32, 2109 (1965). R. M. Acheson and R. Robinson, J. Chem. Soc., 1127 (1952). Prepared by Phillip Mattison. Prepared by Donald Anderson. Prepared by Calvin Johnson. C. R. Hauser,-F. W. Swamer and B. I. Ringler, J. Amer. Chem. Soc., 12“ 4023 (1948). J. J. Panouse and C. Sannie', Bull. Soc. Chim. France, 1036 (1966). W. von E. Doering and F. Marshall, J. Amer. Chem. Soc., 2;, 2229 (1949). A. Ellis, Org. gyn., Coll. Vol. 1, 18, John Wiley and Sons, Inc., New York, 1946. 83 37. H. Pines, E. J. Pavlik and V. N. Ipatieff,-J. Amer. Chem. Soc., 12” 5986 (1957). 38. E. Eliel and C. A. Lukach, J. Amer. Chem. Soc., 12” 5986 (1957). 39. H. G. Kuivila and O. F. Beumel, Jr., J. Amer. Chem. Soc., g}, 1248 (1961).