i MOLECULAR REARRANGEMENTS 0F DERIVATIVES 0F - .HEXAMETHYLBICYCLOE3.2.0JHEPTA- 3,6 - DIEN - 2 - ONE Thesis for the Degree of M. S. MICHIGAN STATE UNIVERSITY SHIN LEE 1977 ' mexfiky MiCEiZde; gm»: University In rearrang (L3) and dichloy‘c (fl?) Her Tre Yield 01 Eandl ABSTRACT MOLECULAR REARRANGEMENTS 0F DERIVATIVES 0F HEXAMETHYLBICYCLOI3.2.0]HEPTA-3,6-DIEN-2-0NE by Shin Lee In this theses, synthesis, acid-catalyzed and photoinduced rearrangements of hexamethylbicycloi3.2.0)hepta-3,4-epoxy-6—en-2-one (13) and the synthesis and acid-catalyzed rearrangements of 4,4- dichloro-l,3,5,6,7.8-hexamethyltricyclol4.2.0.03'Slocta-7-en-2-one (g9) were investigated. Treatment of 13 with trifluoroacetic acid resulted in a high yield of a mixture of two stereoisomeric compounds to which structures lg and 12 are assigned. n [i CF3CO2 >< ' H0 CF3 1§_and 12 The mechanism proposed fOr this reaction involves protonation of the carbonyl oxygen, a circumambulatory-type rearrangement and addition of the trifluoroacetic acid. of of its The ePox also Shin Lee Irradiation of l; in ether through Pyrex resulted in the formation of §§ (70%), l§_(8%) and 33 (5%). The unusual strained cage structure of §§_is tentatively assigned on the basis of the spectral data and its subsequent thermal reaction to gg. 0 0 0H 0 \ ll. —W—+ + + 0 . 13 12 35 I. The mechanism proposed for the formation of §§ involves cleavage of the epoxide C-C bond. Mechanisms for the formation of 1;, 25 and g; are also proposed. Treatment of §Q with aqueous silver perchlorate resulted in the exclusive formation of ring expansion product 4-chloro-5-hydroxy- 1,3,5,6,7,8-hexamethylbicyclo(4.2.0]octa-3,7-dien-2-one (§§). Treatment of §Q with methanolic sulfuric acid resulted in the formation of compound fig (10%) and El (72%) which are tentatively assigned the structures shown. The mechal ambul ato r3 Shin Lee The mechanism proposed for the formation of §§ involves a circum- ambulatory type rearrangement whereas the formation of El involves a homotropylium cation rearrangement. MOLECULAR REARRANGEMENTS 0F DERIVATIVES 0F HEXAMETHYLBICYCLO[3.2.0]HEPTA-3,6-DIEN-2-0NE by Shin Lee A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemistry 1977 A i ,' :1 A" ~—4 guida Scieni SUDpOl ACKNOWLEDGMENTS The author is deeply grateful to Professor Harold Hart for his guidance and encouragement throughout the course of this study. Appreciation is extended to Michigan State University, National Science Foundation and National Institute of Health for financial support in the fOrm of teaching and research assistantships. ii INTRODL RESULTS 1. INTRODUCTION ........ . ............................................. RESULTS AND DISCUSSION ............................................ 1. EXPERIMENTAL ...................................................... 1. 2. TABLE OF CONTENTS Epoxidation of Hexamethylbicyclo[3.2.0]hepta-3,6- dien-Z-one (1) via Alkaline Hydrogen Peroxide vs m-Chloroperbenzoic Acid ................................ . Acid-Catalyzed Rearrangement of Hexamethylbicyclo- [3.2.0]hepta-3,6-dien-2-one (lg) .......................... . Addition of Dichlorocarbene to Hexamethylbicyclo- [3.2.0]hepta-3,6-dien-2-one (l) ........................... . Acid-Catalyzed Rearrangement of 4,4-dighgoro- 1.3.5.6,7,8-hexamethyltricyclo[4.2.0.0 locta- 7-en-2-one (SQ) ................................... General Procedure ................................. Epoxidation of Hexamethylbicyclo[3.2.0]hepta- 3,6-dien-2-one (l) with Alkaline Hydrogen Peroxide . Preparation of 4- Trideuterohexamethylbicyclo- [3.2.0]hepta-3,6-dien-2-one (If) .................. Epoxidation of AT ......... . ....................... . Epoxidation of Hexamethylbicyclo[3.2.0]heota- 3,4-epoxy-6-en-2-one (13) with m-Chloro- perbenzoic Acid.. ..... T? ......... . ................ . Epoxidation of ;§* with m-Chloroperbenzoic Acid... Epoxidation of Hexamethylbicyclo[3.2.0]hepta- 6,7-epoxy-3-en-2-one (19) with Alkaline Hydrogen Peroxide ....... . ...... . ...... ...... ...... iii- . Photolysis of 1; .......................................... ........ Page 10 17 25 27 35 35 37 37 37 38 39 TABLE OF CONTENTS (continued) Page 8. Epoxidation of lg_with m-Chloroperbenzoic Acid ............ 39 9. Acid-Catalyzed Rearrangement of Hexamethyl- bicyclo[3.2.0]hepta-3,4—epoxy-6-en-2-one (1;) ............. 4O 10. Acid-Catalyzed Rearrangement of lgf ...................... 41 11. Hydrolysis of 1§_and 1]..... ............................. 41 12. Photolysis of Hexamethylbicyclo[3.2.0]hepta- 3,4-epoxy-6-en-2-one (13) ................................ 41 13. Photolysis of 12f... ..................................... 43 14. Addition of Dichlorocarbene to Hexamethyl- bicyclo[3.2.0]hepta-3,6-dien42-one (1) ................... 43 15. Addition of Dichlorocarbene to 1? ........................ 45 16. Silver Perchlorate Promoted Ring Expansion of (§9) ....... 45 17. Acid-Catalyzed Rearrangement of §Q in Sulfuric Acid ............................................ 46 18. Acid-Catalyzed Rearrangement of §Qf ...................... 48 19. Treatment of g] in Trifluoroacetic Acid-d1 ............... 48 20. Treatment of g] in CH30Na/CH300 solution ................. 48 BIBLIOGRAPHY ...................................................... 49 iv LIST OF TABLES Page 1. Possible Configurations of 12, E, E and If; ................... 9 LIST OF TABLES Page 1. Possible Configurations of IO, 13, 14 and _1_5_ ................... 9 I. Poss LIST OF TABLES Page 1. Possible Configurations of _1_0_, 1g, 1_4_ and 1_5_ ................... 9 INTRODUCTION In an earlier attempt to synthesize hexamethyltropone by Dr. M. Nitta in this laboratory, compound_1 -— a valence isomer of hexamethyl- tropone -— was obtained. The reaction sequence for synthesizing 1 compound 1 was as follows , starting from the commercially available hexamethyldewarbenzene: \ / :cc1 \__J c1 I l --2-’ l c1 I \ /”—7 .3. o H... __4__, c1 ..__._>, -HCl "2° 3 1 The dichlorocarbene adduct £_was not isolated but spontaneously under- went dehydrochlorination to yield compound 3. Treatment of 3 with cold, concentrated methanolic sulfuric acid gave the bicyclic dienone_l in I 2 85% yield. The conversion of §.to.1 involves a sequence of 1,2-shifts: +- \ 1’ ‘~ Cl l . 3.1.1 f,:'——->a .SI: HO -HCl 3 1 Compound 1_showed some interesting acid-catalyzed2 (Scheme 1) and photoinduced3 (Scheme 2) rearrangements. Scheme 1 OH T“) H” )F‘r. ‘ I / —_" I j) ——> i. v: 1 4 OH H o -+ 0 ~——-z—* . , FSO3H ,, _s_ 5 3 The proposed mechanism for the fermation of §_in F503H involves a 1,2-shift to the OH-bearing carbon to give_4, which further rearranges to 5. Scheme 2 O hv -—- 0 £3 a———-’ N 5 ¥ . 1 7 8 The formation of §_by the photolysis of I can be rationalized by an electrocyclic photorearrangement followed by ring opening of the bicyclic ketene intermediate 1, An anticipated intramolecular cyclo- addition product, the tetracyclic cyclobutanone_2, was not obtained in this reaction. in! A 1‘ I 0 .9. Small-ring compounds have been notorious for undergoing molecular rearrangements, and the versatile rearrangements of compound 1_ stimulated our interest to further study the chemistry of some of its derivatives. That this would be a promising area was already clear from a study of the epoxyketone 19?. Irradiation of the g,r-epoxyketone 4 19 resulted in the fbrmation of lg_(56%)3. The proposed mechanism (Scheme 3) is consistent with the labeling results. Bond formation Scheme 3 .0 hv \\ ‘* \\ / —" \. . —’ / * * V * 351 an . o "0 0H 0 OH O C O " '"' -—9 / —9 /~ —--e + ——O t * t 11 12 between C-4 and C-6 followed by bond cleavage between C-1, C-7 and C-5, C-6 and bond formation between C-3 and C-6 could give 11. r-hydrogen abstraction fOllowed by aromatization would then lead to the observed products, pentamethylphenol and ketene. I will discuss, in this thesis, the acid-catalyzed and photo- induced rearrangements of the one-epoxy derivative of compound __l_and the acid-catalyzed rearrangements of a dichlorocarbene adduct of compound_1. RESULTS AND DISCUSSION 1. Epoxidation of Hexamethylbicyclo(3.2.00hepta-3,6-dien-2-one (1) _vja Alkaline Hydrogen Peroxide vs m-Chloroperbenzoic Acid It is well established4 that the epoxidation reactions of olefins by peracids proceed by electrophilic attack of the peracid upon the double bond. When the olefinic bond is conjugated with an electron- \C/ 'IIL_‘\6/H ‘(fi --€§ Q\. + u .r'” \61 -—R ,-—C--C~.‘ R-C-OH c / / \ . withdrawing group such as carbonyl, nucleophilic epoxidation by means of alkaline hydrogen peroxide is often used. R-CH’CH-fi-R + H-O-O- 1::22 R O H-CH=C-R ——+ R-CH-CH-C-R lg \/ II -c I o o oo "0? Thus, the epoxidation of the “'9‘9' ,r'-unsaturated ketone _l_with alkaline hydrogen peroxide or with m-chloroperbenzoic acid led to two different monoepoxides, ;§_and 19 respectively. HZOZINaOH ‘7 " m-CPBA .2, 0 /’ 10 The chemical shifts and europium shift slopes of ;§_and 19_are shown on the structural formulas. The signal at the C-4 position of 1§_and lg_was assigned by epoxidizing compound If (labeled with a CD3 group at the C-4 position). The remaining peaks are readily assigned from their chemical shifts and from their Eu-shift slopes. The configurations of these epoxides (exo or endo) were not determined, but only one isomer was formed in each case. I will return to this question of stereochemistry shortly. 2.94 (1.76) (1.02) (2.00) (g 34) 0 1.43 \ (3.20) 1.20 ° (2.41) I 1.27 1.67 1 58 // (1:00) 1'12 1.34 1'30 1.03 (1°48) (1.52) (1°00) (1.47) (i:23) The Shown 0n 1; and 1 grow) at from the comSlur: bUt only QUESLIOn (1.75) 1.43 , 1.58 I (1.00) HZOZINaOH L 7 1 e " m-CPBA O // 10 The chemical shifts and europium shift slopes of 1§_and 19_are shown on the structural formulas. The signal at the C-4 position of I; and ;Q_was assigned by epoxidizing compound If (labeled with a CD3 group at the C-4 position). The remaining peaks are readily assigned from their chemical shifts and from their Eu-shift slopes. The configurations of these epoxides (exo or endo) were not determined, but only one isomer was formed in each case. I will return to this question of stereochemistry shortly. 2.94 (1.76) (1.02) (2.00) (3-34) 0 1.43 \ (3.20) 1.20 ° (2.41) l 1.27 1.67 1 58 // (1:00) 1°12 1.34 1'30 1.03 (1°43) (1.52) (1°00) (1.47) (i 23) 13 10 The Shown on L3 and B group at from thej configura but only QUESIIOn ‘ (1.76) 1.43 \ 1.58 J (1.00) HZOZINaOH w‘7 1 a: " m-CPBA 0 ,/ 10 The chemical shifts and europium shift slopes of 1§_and 19_are shown on the structural formulas. The signal at the C-4 position of I; and 19_was assigned by epoxidizing compound If (labeled with a CD3 group at the C-4 position). The remaining peaks are readily assigned from their chemical shifts and from their Eu-shift slopes. The configurations of these epoxides (exo or endo) were not determined, but only one isomer was formed in each case. I will return to this question of stereochemistry shortly. 2.94 (1.76) (1.02) (2.00) (3°34) o 1.43 \ (3.20) 1.20 ° (2.41) ' J 1.27 1.57 1 58 I // (1:00) 1'12 1.34 1'30 1.03 (1°48) (1.52) (1°00) (1.47) (I123) 13 10 7 When an excess of m-chlorpperbenzoic acid was used, a single diepoxide 14_was formed from 1, Treatment of 19_with alkaline hydrogen peroxide also gave exclusively 14, However, treatment of 1§_with m-chloroperbenzoic acid gave a mixture of two stereoisomeric diepoxides 1§_and 14_in a ratio of 9 to 1. m-CPBA ‘ +2 14 H202/NaOH .__ H 0 The chemical shifts and europium shift slopes of 14_and 1§_are tentatively assigned as shown on the structural f0rmulas. (2.94) (1.97) (2.00) 0.9 0 (fi-gg) 0.90 0 1.20 H 0 ° 2 2 ‘ . 1.38(1.55) (5‘7 / or 9 0 (1°00) (1°47) (1.22) (1.36) 1.45(1.00) .19. 1e 10% (2.94) ( ) 1.02 2.30 o (3.20) 1°52) 0.9 I 1.27 M'CPBA 1.28(2.07) 90% 0 1.12 1.37 (1.48) (1:33) (1.00) (i:§§) 1.34(1.20) 13 15 8 The signal at the C-4 position of 14 and 1§_was assigned by epoxid- izing compound 147 (labeled with a CD3 group at the C-4 position). The remaining peaks were assigned by comparing with those of the monoepoxides from which they were derived. The configurations of these diepoxides (exo or endo) could not be determined. However, once the configuration of one of the diepoxides is determined, the configurations of the other diepoxide and the two monoepoxides 1g and 1g_will be known. There are all together four sets of possible configurations of these epoxides (Table I). Set 3 and Set 4 are least likely considering the unfavored diepoxide structure (15 in Set 3 and 14 in Set 4) with both epoxy groups having endo positions. The fact that the chemical shifts of the methyls at C-3 and C-4 of 14 (S 1.38, 1.45) are at lower field than those of 11 (8 1.27, 1.34) and 1§_(5 1.28, 1.34) is consistent with the configurations assigned in Set 1 in which the two methyls in 14, being close to the endo epoxy group, would be shifted to lower field. For this reason Set 1 is favored. However, the configuration of 14 in Set 1 does not explain why the europium shift slope of the C-7 methyl (2.00) is larger than that of the C-1 methyl (1.97). However in Set 2, the configuration assigned to 14 places the C-7 methyl group in an endo position, possibly closer to the binding site of the carbonyl group and europium shift reagent than the C-1 methyl group, offering an explanation for the larger europium shift slope of the C-7 methyl. Thus Set 2 remains a possibility. -\ O ‘- or O . .0.» . .06 g g .0.‘ E .m 2.. 3 .fl .2 .3 65:23:50 5233.. 4 £2: lO 2. Acid-Catalyzed Rearrangement of Hexamethylbicyclot3.2.0]hepta- 3,4-epoxy-6-en-2-0ne (13) Treatment of 13 with trifluoroacetic acid at 0°C for 1 hr gave a mixture of two stereoisomeric compounds to which structures 1§_and 12 are assigned. It was not possible to separate the two isomers. The two isomers: 1g and 11 structures of these isomers were assigned based on their spectral properties though the configuration of each isomer could not be ~determined. The molecular formula C15H1904F3 was confirmed by a mass spectrum (parent peak m/e 320) and elemental analysis. The ir spectrum 1 1 showed a strong 0-H stretching band at 3400 cm' ; the uc=0 at 1755 cm- and uv maxima at 214 nm (e 1,030) and 232 nm (a 590) indicate that there is no conjugation in the molecule, and the position of the carbonyl absorption is consistent with its location in a strained five- membered ring. For example, compound 182 has a carbonyl absorption at 1753 cm‘l. L/ 11 An 19 F NMR spectrum gave two signals with an area ratio of 10 to 9 corresponding to the magnetically different CF3 groups in the two isomers 1g and 12. A PMR spectrum of the mixture in CCl4 showed a. singlet for the four aliphatic methyls at 5 1.20 and another singlet for the two vinyl methyls at S 1.67. However, eur0pium shift reagent resolved the signal at 8 1.20 into two singlets, thus gave a spectrum with three singlets of equal intensity with europium shift slopes of 2.36, 1.47 and 1.00 (two vinyl methyls). These three singlets were not resolved further by shift reagent but were broadened owing to the presence of the two isomers. Use of acetone-d6 as the solvent resolved the PMR spectrum of the mixture into two sets of signals corresponding to compounds 1§_and 11, with relative areas of 10:9. Compound 1g had three equal singlets at S 1.12, 1.33 and 1.70 whereas compound 12 had three equal singlets at S 1.15, 1.20 and 1.70. When the mixture of 1§_and 17 was treated with a 7% solution of K2603 in aqueous MeOH for 4 hours the starting material was recovered quantitatively. The difficulty in the hydrolysis of 19 and 17 is consistent with the structure assigned. Although the structures of 14 and 11 are not rigorously proved, it is difficult to imagine an alternative structure which will satisfy the symmetry demands of the NMR spectra, and will also be consistent with the carbonyl absorption in the ir. A plausible mechanism for the formation of 19 and 12 from 1; in TFA is shown in Scheme 4. 12 Scheme 4 O OH \ + \ ’7 H 1,2-shift :1 / 0 / 'k 'k 13 12 OR (a) ' , H 0" _ CF c0 + 1,2 shift 1 3 2 1% _11__+ ‘\ 0 . 0 0 * * o=é * 0:0 0" 20 | (b) ___ CF3 CF3 epoxide 1 21 (1)1,2-shift walk "' (2)-H+ ‘\ OH 0 * (1)1,2-shift (2)-H+ o I V! 23 19 and 12 l3 Protonation of the carbonyl oxygen and two 1,2-shifts gave cation 29, then followed by either route (a) or route (b) could lead to the final products. In route (a), 29 was first attacked by trifluoroacetate ion to give 21. Reprotonation followed by 1,2-shift and deprotonation gave 22, which on cyclization would give 1g and 11. In route (b), 29 could undergo epoxide walk process followed by 1,2-shift and deprotonation to give 22 which could be attacked by TFA to give 1§_and 12. The first step involves the protonation of the carbonyl oxygen. In general, the acid-catalyzed rearrangement of epoxy ketones is initiated by protonation of the epoxy oxygen atoms. However, if epoxide were protonated first, ring opening would occur in such a manner as to place the positive charge remote from carbonyl group to give ion 24, The subsequent rearrangements could not lead to the products observed but would only lead to carbonium ion 2g which is unfavored by having the positive charge «.to the carbonyl group. 0 ,1 + ———> OH ——+ $.01?“ 0H A few examples of the protonation of the carbonyl oxygen in an epoxyketone are known. For example, a competitive protonation at each oxygen of the epoxyketone 24 was reported by Hart and HuangG. 14 The formation of 21 from 2g involved initial protonation at the carbonyl oxygen, a mechanism consistent with labeling experiments. 1. acyl shift 0‘" 27 An exclusive protonation at the carbonyl oxygen of an epoxy ketone had also been suggested3 in the acid-catalyzed rearrangement of compound ‘19 to 22, OH H” '\ ‘ 7 —~ +»—~0 * * * é _1_0 o __, Ct? 15 The rearrangement of 12 to 29 in Scheme 4 by two 1,2-shifts is a process closely related to the circumambulatory process which was ' postulated by Ninstein and coworkers7 to rationalize the 5-carbon degenerate rearrangement (shown by deuterium labeling) of the 7-norbornadienyl cation 22. The suggested mechanism involved a shift of C-2 from C-l to C-7 (or by symmetry, C-3 from C-4 to C-7) to form m '9 29, which through the reverse process was reconverted to 22, but with the carbon skeleton in a different sequence. In this process the bond vinyl group (C-2, C-3) maintains its identity, but circum- ambulates about the five-membered ring. Finally, there is precedent for the unusual ortho ester type of 16 functionality postulated for 1g and 12. A cvclization process of trifluoroacetic acid with epoxide similar to the present case had 8. Compound 22 was the been found in the acid treatment of 21_with TFA major product of the reaction. The driving force of this reaction is still unknown. 0 i O TFA/ether - . o°c ; 0 n 0 32 H0 053 32 When 12_labeled with a CD3 group at the C-4 position was subjected to the same acid rearrangement conditions, the products were found to be 19 and 1] with one of the four aliphatic methyls labeled. This result is consistent with the proposed mechanism. 17 3. Photolysis of 13 The extensive rearrangements of aliphatic «,8-epoxyketones upon ultraviolet light irradiation are attributable to transformations ' of the n “1* excited stateg. Compound 12 has an n 411* absorption at )max 315 nm (5 320); thus a pyrex filter was used for the irradiation of 12: The photolysis of 12 in ether through Pyrex was followed by vpc. The peak corresponding to 12 disappeared completely in about 18 hr. During that period, one major peak corresponding to 22 and two minor peaks corresponding to 12 and 24 were growing in area. However, the NMR spectrum of the crude reaction mixture did not show the signals corresponding to 22. It showed mainly the signals corresponding to another compound 22, along with those corresponding to 12 and 24. Obviously the photolysis of 12 gave 22 (70%), 12 (8%) and 24 (5%) as products but the major product 22 decomposed thermally on being subjected to gas chromatography, to give compound 22. Attempts to isolate 22 by using tlc plates failed due to its decomposition on the plates. By comparing the spectral data with those of the authentic samples, compound 12 was identified as pentamethylphenol, compound 24 as pentamethylphenyl acetate and compound 22 as l-acetyl-penta- methylcyclopentadiene. Compound 22 is tentatively assigned the unusual strained cage structure shown, on the basis of the following spectral data, and on its subsequent thermal reaction to give 22. The NMR spectrum of 22 showed that all of the methyl groups were aliphatic 18 (all signals at.8:s 1.26). Europium shift reagent removed the accidental degeneracy of the two methyls at 8 1.00 and the two methyls at S 1.26 to give a spectrum with six singlets, showing that each.methyl group was unique and that there could be no symmetry in the molecule. The ir band at 1760 cm"1 is attributed to the cyclobutanone absorption. These spectral data severely restrict the possible structures. For example, an alternative structure with all methyls aliphatic could arise by a 1,2-acyl shift (oxa-di-n-methane rearrangementlo) of 12: 1,2-acyl shift However, although 2§_has six unique methyl groups and therefore could be consistent with the NMR Spectrum, it is not consistent with a carbonyl frequency at 1760 cm'l. 19 A plausible mechanism can be deduced for the formation of 2_5_ and for its rearrangement to 22. The following mechanism is proposed for the formation of 22 from 12: 0 0. O. 0. \ \ ——) —') ."—’ V O o 'k I * * * 13 37 _ ”l 0* . E .0 9—7— I * 35 38 After electron promotion through the n 411* state, the C-C bond in the epoxy ring was broken to give diradical 2_7. Participation of the (hr-double bond formed 22. Ring closure to the di radical 22 would give 221. The feature of this mechanism that is unique is the C-C bond cleavage. Normally, one might have expected 12 to rearrange as followsgz 20 ml \ . ' ———-> I ——9 O / 0 \\ ./ I» i ('5 I 00 Indeed this was the process which was anticipated, but the properties of the product were clearly inconsistent with structure 22. However, there is precedent, though rare, for C-C bond cleavage in the ring Opening of epoxides. For example, the photolysis of compound 49 which led to the formation of the four isomers seems to involve the C-C bond cleavage of the epoxide ring as the key stepll. I—0 — +299 [fii:]:f§=//L§TO hv ._— o ——-> H COCH3 21 Three possible mechanisms to rationalize the thermal decomposition .. of 25_ to Q were considered: (A) 0 -c0 —————9 as 4_1 .43. * -— 0 .__—.__, = _ * 2 (B) O *01: -C0 22--:.4_1.-—>‘ =—:- 22 (C) The first two mechanisms (A) and (B) involve decarbonylation as the initial step. In mechanism (A), cyclopropyl ring opening of 41_ could give 42, and cleavage of the C-0 bond could give 22. In mechanism (8), ring closure of the diradical 41 would give 42. Isomerization to 44 followed by cleavage of the C-0 bond and ring contraction could give 22. Mechanism (C) involves cleavage of a C-0 bond as the initial step. Opening of cyclopropyl ring followed by cleavage of the bond «.to the carbonyl group and decarbonylation could give carbene 42, which could undergo ring expansion to give 22, Undoubtedly other mechanisms than those shown could also be devised, and some steps in the above mechanism could be coupled or concerted, rather than discrete. In order to obtain additional mechanistic information, the photo- lysis was carried out by irradiating compound 127 which was labeled at the C-4 position with a CD3 group. Product 227 lacked the NMR signal 23 of one of the methyl groups at 8 1.26, as shown by a europium-shifted spectrum. However careful examination of the NMR spectrum of compound .227 obtained from the thermal decomposition of 22f showed that the signal of the acetyl methyl group was reduced about 50% in area while the signals of all other methyls retained their full intensity. It is possible that the deuterated acetyl methyl was partly exchanged although how it was exchanged is still unknown. The labeling result rules out path (A) and (B) because both of these mechanisms would result in label (or partial label) at the vinyl methyls. Path (C), with a full label at the acetyl methyl group, is more favored to be correct. The formation of the two minor photoproducts 12 and 24 from 12 can be rationalized by the mechanisms in schemes 5 and 6. Scheme 5 involves breaking of the epoxy C-0 bond followed by ring contraction to give e-diketoneI4z. This feature of the mechanism is not too appealing, but if coupled with the next step might have a reasonable driving force due to relief of strain. Ring opening could give 42, which could easily eliminate one molecule of ketene to form 12, Compound 12 could recombine with ketene to form 24. Scheme 6 involves the formation of 22 as an intermediate. The same path as in the rearrangement of 22 in mechanism (C) to give 42, followed by ring opening could give .42, chich could eliminate ketene to give 12 and 24. Scheme 5 and 6, which give the same labeling position in 24? (at the acetyl group) by photolysis of 127, were both consistent with the experimental results. 24 Scheme 5 O at 1 I1 9- ’ Ki + HZC=C=O ——-> 25 4. Addition of Dichlorocarbene to Hexamethylbicyclo[3.2.0]hepta-3,6- dien-Z-one (I) The method of generating of dichlorocarbene by the action of, aqueous alkali on chloroform in the presence of triethylbenzylammonium 12 was used for the chloride developed by Makosza and Nawrzyniewicz addition of dichlorocarbene to compound_1. Three products, 2Q_(45%), 21 (3.8%) and 22 (3.3%) were obtained in this reaction. Compound §Q_was assigned the structure shown based on spectral properties. The molecular formula C14H180Cl2 was confirmed by the mass spectrum and elemental analysis. The NMR spectrum showed two mutually coupled vinyl methyl groups (5 1.43, 1.65), and the ir band at 1715 cm'1 and the uv maxima at 230 nm (5 2,210) and 305 (440) indicated that the carbonyl group was unconjugated. Therefore the carbene must have added to the double bond in the five-membered ring. \ :CCl Cl Cl 1 29. (45%) Compounds 21 and 22 were assigned the structures shown. They had similar spectral properties. The molecular formula CISHUOCI3 was confirmed by mass spectra and elemental analyses. NMR signals of_21 showed two mutually coupled vinyl methyls (S 1.60, 1.90) and two mutually coupled hydrogens (5 1.92, 2.16, as 8 Hz); an ir band at 1690 cm‘1 and 26 (2:33) ‘2'}? 0 (1:39) 1.57(3.12) 1.69(2.48) 2.16 1.92 1.93 1.08 1 95(1.70) (1.00) (1.05) (1.00) (1.58) 51 . 52 a uv maximum at 242 nm (2 8,700) indicated the presence of a conjugated carbonyl group in a five-membered ring. The distinction between 21 and ‘22 was based on their europium shift slopes, which showed that in.21 the two hydrogens are remote from the carbonyl group whereas in 22 the methyl adjacent to the cholrine is the one that is furthest from the carbonyl group. ‘ §1_and 22_were presumably formed by the following mechanism: O / Cl :CCl :CCl .1. .‘2" C‘ / -————+2 51 3 __ H 122 H :CCl 0 ‘ ———’ c1 —’ c1 _—’ 5-3 27 Addition of the dichlorocarbene to the double bond in the four-membered ring followed by elimination of HCl gave 22_and 24, Further attack by dichlorocarbene on the exocyclic double bond gave 21 and 22. 5. Acid-Catalyzed Rearrangement of 4,4-dichloro-I,3,5,6,7,8-hexamethyl- tricyclo(4.2.0.03'5)0cta-7-en-2-one (50) A. Silver Perchlorate Promoted Ring Expansion of 50 Nhen‘29_was treated with an excess of silver perchlorate in acqueous acetone, the dichlorocarbene adduct underwent a ring expansion reaction to give 22. The structure of 22 was assigned as shown based on spectral properties. The molecular formula C14H1902Cl was confirmed by a mass spectrum and elemental analysis. The ir band at 1660 cm"1 and uv maxima at 237 nm (2 2,700), 248 (4,650) and 285 (1,230) are consistent with a conju- gated carbonyl group in a six—membered ring. The presence of a hydroxyl group is clear from the ir band at 3460 cm"1 and from the presence of a one-proton peak in the NMR spectrum at 8 2.03. Two mutually coupled vinyl methyls (8 1.40, 1.62) were also shown by NMR, in addition to singlets for the four remaining methyl groups. 28 Compound 22 was presumbly formed by loss of a chloride ion followed by attack of water on the resulting carbocation. 0 £9. -9-———> l —-—+ §§ - + ‘C‘ c] “H B. Acid-Catalyzed Rearrangement of 50 in Sulfuric Acid When a solution of.2g in methylene chloride and methanol was treated with concentrated sulfuric acid at 0°C, two products 22 (10%) and 22 (72%) were formed within 10 minutes. Compound 22 was assigned the structure shown based on spectral properties. The molecular formula CI4H17OCI is confirmed by a mass 1.57(1.16) { I 1.87(1.00) 1.70 1.14 ( ) 1.20 A” 4.80(1.36) 4.88(1.14) 56 spectrum (parent peak m/e 238). The NMR Spectrum showed two mutually coupled vinyl methyls (S 1.57, 1.70), two vinyl protons (S 4.80, 4.88), 29 two aliphatic methyls (81.20) and a singlet vinyl methyl. The europium shift slopes shown on structure clearly indicated that the two aliphatic methyls are close to the carbonyl group whereas the other methyls and methylene group are relatively far away. The uv maxima at 217 nm (6 8,570) and 253 (11,800) indicated a conjugated system. The ir band at 1775 cm'1 is indicative of the bridged carbonyl group; compare, for example, with the.V =0 of compound 2213. 58 b’c=o 1780 cm"1 A possible route to 22 is shown in Scheme 7. Protonation of the carbonyl oxygen would give 22. The rearrangement of 22 to 29 is a circumambulatory type process similiar to the path from 12_to 29 (Scheme 4) in the acid-catalyzed rearrangement of 12. Ring opening would give ‘21. A 1,2-shift followed by deprotonation and elimination of HCl can lead to 22. Nhen‘29* (derived from dichlorocarbene addition to.1f) was subjected to the same acid conditions, the product 22f lacked the NMR signal at 8 1.87. This result is consistent with the proposed mechanism. Scheme 7 1) \ H+ l ---* / Cl2 * 29 OH C12 fl * :31 0 Cl -H+ 1 ’ / 3O 1’ + *2 _5_93 Xi”, c1 .____+ I Cl / * 91 0 l Cl -HCl ———’ / . l 56 Two possible structures 4 and 2 were considered for the major product compound 2], based on its spectral properties. The molecular formula was confirmed by a mass spectrum (parent peak m/e 238). The ir band at 1705 cm"1 is consistent with a conjugated carbonyl group in five-membered ring. The uv maximum at 237 nm (a 14,100) also indicates conjugation. The NMR spectrum and europirm shift slopes are shown on (3.01) (1.68) 1.87 1.42 O 1.67 1.67 (1.00) iilIIIIIIIDA' (2.57) C] H’J(H 2.00(1.68) (2.57) 1.67 H.'( 1.42 .00(1.68) H 1.68 \( ) 4.83(1.20) 4.83(1.20) 4.87(1.07) 4.87(1.07) A B the structures. There are four vinyl methyls (5 2.00, 1.87 and two at 1.67) each of which shows homoallylic coupling, two vinyl hydrogens . (8«4.83, 4.87) and an aliphatic methyl (8 1.42). The conjugation of the exocyclic double bond with one of the double bonds in the ring was confirmed by a deuterium exchange experiment. When 22 was treated with deuterotrifluoroacetic acid at room temperature, the NMR signals at.8 4.83, 4.87 and 1.87 disappeared in 10 min, and a europium shifted spectrum showed that the signal at 8 1.67 (Eu-shift slope 1.00) had sharpened to a singlet. This phenomenon is consistent with both structures_4_and_2. An attempt to deuterate compound 22 in CH30Na/Me00 solution failed due to the instability of 22 under basic conditions. Possible routes to_4.and_2_are shown in Scheme 8. Protonation of 29 followed by ring opening could give cation 22. A (1.6) sigmatropic shift of 22 could lead to either 22 or 24, Elimination of HCl accompanied by ring closure followed by deprotonation could lead to 4_or_2, 32 Scheme 8 32 Scheme 8 ” l -—-—-p 21 011 ' C] C, . 4.... - i—‘i 4——5 0” 01 Cl 5 A» Lei ‘, Cl * * 92 “I III 64 H‘0 a Cl L I2:- CD 33 The rearrangement of §2_to 22 or 24 is similar to the well established14 degenerate rearrangement of the bicyclo(3.1.0]hex43-en- Z-yl cation which was first reported by Swatton and HartAs. However, the degenerate rearrangement of the homotropylium cation, though symmetry allowed, has not been detected thermally16 . A photo- induced rearrangement of the Z-hydroxyhomotropylium cation 2217 has been reported recently by Childs and Rogerson. H ~ H ———o—+ ~H -70 C H 0H 0H 65 The difference in energetics between the rearrangements of the homo- tropylium cation and the bicyclo(3.1.0)hexenyl cation have been ration- alized in terms of stabilization of the ground-state of homotropylium 18 cations through aromaticity . Although the proposed mechanism in Scheme 7 is still doubtful, it does provide a good explanation for the formation of 4 or _8_. Structure 4.0r 2 could be tested by labeling experiments. However, the product 22f derived from 227 showed no deuterium labeling. One rationale is that the deuterons of the CD3 group were exchanged during the workup steps. Thus structure 4, which would have the deuterated methyl group in an enolizable position, is more favored to be the structure of compound 22. EXPERIMENTAL 1. General Procedures Except where otherwise noted, all NMR spectra were measured in CDCl3 or CCl4 solutions using TMS as an internal standard. The 60 MHz spectra were recorded on a Varian T-60 spectrometer. The small numbers placed next to protons in the structures in the results and discussion section are the NMR chemical shifts of those protons. The numbers beside the chemical shifts in parentheses are the normalized euronium shift numbers. These were obtained by adding small increments of tris- (6,6,7,7,8,8,8-heptafluoro-2,2—dimethyl-3,5-octanedione)Eu(III) to the CDCl3 or CCl4 solution of the compound being investigated. After each addition the NMR spectrum was scanned and the new frequency of each absorption was recorded. The shift for each absorption is the differ- ence between the frequency of the shifted absorption and the original one. The normalized shift numbers are ratios obtained by dividing the shift of each signal in the spectrum by the shift of the least shifted signal. Infrared spectra were taken on a Perkin Elmer 237 grating spectro- photometer and were calibrated against a polystyrene film. Ultraviolet spectra were obtained with a Unicam SP-800 in methanol unless otherwise noted. Mass spectra were obtained from a Hitachi-Perkin Elmer RMU-6 operated by Mrs. Ralph Guile. Melting points were determined with a Thomas-Hoover Melting Point Apparatus and are uncorrected. Varian 35 36 Aerograph gas chromatographs were used. Analyses were performed by Spang Microanalytical Laboratories, Ann Arbor, Michigan. 2. Epoxidation of Hexamathylbicyclot3.2.0]hepta-3,6-dien-2-one (1) with Alkaline Hydrogen Peroxide To a solution containing 2.0 g (10.5 mnol) of1 and 3.6 g (31.5 mmol) of 30% aqueous hydrogen peroxide in 10 ml of methanol was added, at 0°C, 1 ml (6 mmol) of 6N aqueous sodium hydroxide. After being stirred at room temperature overnight, the reaction mixture was diluted with water and extracted with ether. The ether extracts were washed with saturated salt solution and dried (M9504). Evaporation of the solvent seft 2.01 g (9.76 mmol, 93%) of a colorless oil considered to be hexamethylbicyclo[3.2.01hepta-3,4-epoxy-6-en-2-one (12), which on standing at room temperature for 1 hr gave colorless crystals, mp 40-41°C; NMR (CCl4) 8 1.02 (s, 3H), 1.12 (s, 3H), 1.27 (s, 3H), 1.34 (s, 3H), 1.43 (q, 3H, J = 1 Hz), 1.58 (q, 3H, J = 1 Hz): ir (neat) 3000 (s), 2900 (m), 1720 (s), 1455 (s), 1380 (s), 1315 (w), 1280 (w), 1250 (w), 1175 (m), 1110 (m), 1070 (m), 1060 (m), 1050 (m), 1020 (s), 860 (s), 740 (w) cm'l; uv (MeOH) Am 215 nm (21,950), 237 (1,340), 315 (320); mass spectrum (70 ev) m/e (rel intensity) 206 (33), 178 (8), 164 (9), 163 (8), 149 (15), 136 (53), 135 (100), 121 (32), 119 (35), 107 (20), 105 (26), 93 (16), 91 (23), 79 (12). ' '42211 Calcd. for C13H1802: C, 75.69; H, 8.80. Found: C, 75.49; H, 8.82. 37 3. Preparation of 4-Trideuterohexamethylbicyclo(3.2.0]hepta-3,6-dien- 2-one (1*) A solution of 500 mg of 1_in 10 ml of CH300 containing 100 mg of NaOCH3 was allowed to stir at room temperature overnight. The reaction mixture was then poured into ice-water and extracted with ether. The ether layer was washed with water and dried (M9804). Removal of the solvent gave quantitative yield of_1f. 4. Epoxidation of 1* The procedure and workup were as described for the epoxidation of 1: The product 12: had an NMR Spectrum identical with that of 12 except that the signal at 8 1.34 was absent. 5. Epoxidation of Hexamethylbicyclo(3.2.0lhepta-3,4-epoxy-6-en-2-one (13) with m-Chloroperbenzoic Acid To a solution of 206 mg (1.0 mmol) of 12 in 2 ml of methylene chloride was added, at 0°C, a solution of 250 mg (1.23 mmol) of m-chloroperbenzoic acid in 3 ml of methylene chloride. The mixture was stirred at 0°C for 5 hr. After the solvent was evaporated under reduced pressure, petroleum ether (bp 30-600) was added to the residue. The precipitated m-chlorobenzoic acid was removed by filtration. The filtrate was washed successively with aqueous sodium bicarbonate, saturated salt solution and dried (MgS04). Evaporation of the solvent left 200 mg of a colorless oil. The crude material was considered to be a mixture of two stereoisomeric diepoxides hexamethylbicyclo(3.2.0}- hepta-3,4;6,7-diepoxy-2-one (12 and 14) in a 9:1 ratio shown by the 38 NMR spectrum. Preparative vpc (6510.25 in column, 15% SE-30 on chromo- sorb w, Aw-oncs 60/80, 165°C, 60 ml/min) gave 14_(ret time 4 min), which on standing at room temperature for hours gave white crystals, mp 86-87°c. NMR (0014) S 0.90 (s, 3H), 1.04 (s, 3H), 1.18 (s, 3H), 1.37 (s, 3H), 1.38 (s, 3H), 1.45 (s, 3H); ir (neat) 2970 (m), 2940 (m), 2880 (w), 1740 (s), 1450 (m), 1375 (m), 1265 (w), 1175 (w), 1160 (m), m. 1110 (w), 1085 (w), 1065 (w), 1050 (m), 850 (w), 820 (w), 760 (w) cm'lz ' uv (MeOH) Am” 215 nm (5. 1,110), 310 (60): mass spectrum (70 ev) m/e (rel intensity) 222(1), 164 (8), 151~(36), 149 (8), 137 (22), 136 (9), 125 (11), 124 (100), 123 (30), 109 (41), 91 (10), 81 (11). 1; ‘42211 Calcd. for C13H1803: c, 70.24; H, 8.16. Found: C, 70.22; H, 8.16. 12(ret time 4.5 min): NMR (CCl4)8 0.94 (s, 3H), 1.12 (s, 3H), 1.28 (s, 3H), 1.32 (5, 3H), 1.34 (s, 3H), 1.37 (s, 3H): ir (neat) 2975 (m), 2940 (m), 2880 (w), 1740 (s), 1470 (m), 1450 (m), 1385 (m), 1375 (m), 1310 (w), 1255 (w), 1215 (m), 1185 (w), 1105 (m), 1080 (m). 1010 (m), 890 (w), 845 (m), 720 (w) cm'l: uv (MeOH).Ahax 218 nm (2 830), 308 (50); mass spectrum (70 ev) m/e (rel intensity) 222(2), 164 (10), 151 (39), 149 (10), 137 (23), 136 (10), 125 (11), 124 (100), 123 (33), 109 (42), 91 (11), 81 (12). .1931; Calcd. for c H 0 . c, 70.24; H, 8.16. 13 18 3‘ Found: C, 70.22, H, 8.16. 6. Epoxidation of 13* with m-Chloroperbenzoic Acid The procedure and workup were as described for the epoxidation of 12, The product 12* had an NMR spectrum identical with that of 12 39 except that the signal at 8 1.34 was absent. The product 14* had an NMR spectrum identical with that of 14 except that the signal at 5 1.45 was absent. 7. Epoxidation of Hexamethylbicyclo[3.2.0)hepta-6,7-epoxy-3-en-2-one (10) with Alkaline Hydrogen Peroxide To a solution containing 103 mg (0.5 mmol) of 12 and 170 mg (1.5 mmol) of 30% aqueous hydrogen peroxide in 0.5 ml of methanol was added, at 0°C, 0.1 ml (0.6 mmol) of 6N aqueous sodium hydroxide. After being . stirred at room temperature overnight, the reaction mixture was diluted with water and extracted with ether. The ether extracts were washed with saturated salt solution and dried (M9504). Evaporation of the solvent gave a quantitative yield of the diepoxide 14 as the only product. 8. Epoxidation of 10 with m-Chloroperbenzoic Acid To a solution of 103 mg (0.5 mmol) of 1Q_in 1 ml of methylene chloride was added, at 0°C, a solution of 500 mg (2.5 mmol) of m-chloro- perbenzoic acid in 3 ml of methylene chloride. The mixture was stirred at room temperature for 72 hr. The solvent was evaporated and petroleum ether (bp 30-600) was added to the residue. The m-chlorobenzoic acid was removed by filtration. Evaporation of the solvent from the filtrate gave a quantitative yield of the diepoxide 14 as the only product. 4O 9. Acid-Catalyzed Rearrangement of Hexamethylbicyclo[3.2.0)hepta-3,4- epoxy-6-en-2-one (13) A solution of 200 mg (0.97 mmol) of 12 in 3 ml of ice-cold tri- fluoroacetic acid was stirred at 0°C for one hour. The reaction was quenched by adding the mixture dropwise to a precooled saturated sodium carbonate solution at 0°C. The mixture was extracted with methylene chloride. The combined methylene chloride layers were washed succes- sively with aqueous sodium bicarbonate, saturated aqueous sodium chloride and dried (M9504). Evaporation of the solvent left 220 mg of light brown crystals. An NMR spectrum of the crude material showed it to be 95% pure. The crude product was recrystallized from methylene chloride to give 198 mg (0.62 mmol, 64%) of white crystals which were considered to be a 10:9 mixture of two isomeric compounds 12_and'1z. The mixture had the following properties: mp 140-150°c; NMR (CCl4) 8 1.20 (s, 12 H), 1.67 (s, 6H); NMR (acetone-d6):8 1.12 (s, 3H), 1.15 (s, 2.7H), 1.20 (s, 2.7a), 1.33 (s, 3H), 1.70 (s, 5.7M): 19F NMR (acetone-d6) two singlets at 8 84.4, 85.9 (upfield from CCl3F) in a ratio of 10:9; ir (KBr) 3400(5), 3000 (m), 1755 (s), 1470 (m), 1450 (m), 1430 (m), 1400 (m), 1255 (m), 1220 (s), 1190 (s), 1115 (s), 1080 (s), 1060 (s), 980 (w), 930 (m), 750 (m) cm' : uv (MeOH) 214 nm (8 1,030), 232 (590); mass spectrum (70 ev) m/e (rel intensity) 320 (3), 206 (13), 165 (14), 164 (100), 163 (18), 149 (49), 147 (10), 137 (23), 136 (26), 135 (15), 121 (12), 119 (11), 105 (14), 91 (15), 77 (13), 69 (11). .4221; Calcd. for C H 0 F - C, 56.24; H, 5.97 15 19 4 3’ Found: C, 56.28: H, 5.55 41 10. Acid-Catalyzed Rearrangement of 13* The procedure and workup were as described for the acid-catalyzed rearrangement of 12, The product mixture 12? and 12? had an NMR spectrum (CCl4) identical with that of 12 and 12 except that the signal at.8 1.20 was reduced by one-quarter in area. 11. Hydrolysis of 16 and 17 L7 A solution of 46 mg of a 10:9 mixture of.12 and 12 in 2 ml of 7% 6 solution of K2C03 in aqueous methanol (2:5 v./v.) was stirred at room L 1.3 temperature for 4 hr. The reaction mixture was then poured into water and extracted with ether. Removal of the solvent from the combined ether layer left white crystals which were identified as 12 and 12 by NMR and ir spectra. 12. Photolysis of Hexamethylbicyclo(3.2.0)hepta-3,4-epoxy-6-en-2- one (13) A solution of 75 mg (0.36 mmol) of 12 in 25 ml of anhydrous ether in a pyrex test tube sealed with a septum was deoxygenated with a nitrogen stream for 15 min. This solution was irradiated through a pyrex filter with a 450w Hanovia lamp. The photolysis was followed by analytical vpc (5ix0.125 in column, 5% FFAP on chromosorb w, AH-DMCS 80/100, 145°, 60 ml/min). As the reaction proceeded, the peak corres- ponding to 12 (ret time 3.9 min) decreased in area. Three peaks corres- ponding to 22_(ret time 3.7 min), 12 (20.6 min) and 24 (27 min) appeared along with several other small peaks. The reaction was com- plete in 18 hr. An NMR spectrum (CCl4) of the crude reaction mixture 41 10. Acid-Catalyzed Rearrangement of 13* The procedure and workup were as described for the acid-catalyzed rearrangement of 12, The product mixture 12f and 12: had an NMR spectrum (CCl4) identical with that of 12 and 11 except that the signal at.5 1.20 was reduced by one-quarter in area. 11. Hydrolysis of 16 and 17 A solution of 46 mg of a 10:9 mixture of_12 and 12_in 2 ml of 7% solution of K2C03 in aqueous methanol (2:5 v./v.) was stirred at room temperature for 4 hr. The reaction mixture was then poured into water and extracted with ether. Removal of the solvent from the combined ether layer left white crystals which were identified as 12 and 12 by NMR and ir spectra. 1 12. Photolysis of Hexamethylbicyclo[3.2.0)hepta-3,4-epoxy-6-en-2- one (13) A solution of 75 mg (0.36 mmol) of 12 in 25 ml of anhydrous ether in a pyrex test tube sealed with a septum was deoxygenated with a nitrogen stream for 15 min. This solution was irradiated through a pyrex filter with a 450w Hanovia lamp. The photolysis was followed by analytical vpc (55<0.125 in column, 5% FFAP on chromosorb w, AH-DMCS 80/100, 145°, 60 ml/min). As the reaction proceeded, the peak corres- ponding to 12_(ret time 3.9 min) decreased in area. Three peaks corres- ponding to 22 (ret time 3.7 min), 12 (20.6 min) and 24_(27 min) appeared along with several other small peaks. The reaction was com- plete in 18 hr. An NMR spectrum (CCl4) of the crude reaction mixture 42 showed four peaks in the region offiS 1.00-1.26 corresponding to 22 (70%): 1.00 (s, 6H), 1.03 (s, 3H), 1.08 (s, 3H), 1.26 (s, 6H); europium shift reagent resolved signals of 22 to six equal singlets with relative slopes ( in the order of the above signals): 1.74, 2.11, 1.06, 2.15, 1.00, 1.34; and there were three peaks in the region of 8 1.97-2.12 corresponding to 12 (8%) and 24 (5%). An ir spectrum of the crude reaction mixture showed bands at 3500 (w), 3000 (s), 2960 (s), 2900 (m), 1760 (S), 1640 (w), 1450 (s), 1380 (s), 1170 (s), 1080 (m), 1065 (m), 865 (w), 835 (w), 780 (m) cm’l. The mixture was subjected to prepara- tive vpe (5510.25 in column, 10% FFAP on chromosorb w, 80/100). Three major components 22, 12 and 24 were collected. Compound 22 (60%) was identified as l-acetyl-pentamethylcyclopentadiene by comparing its spectral data with those of an authentic sample: NMR (CCl4) S 1.10 (s, 3H), 1.45 (s, 3H), 1.65 (q, 6H, J = 1 Hz), 1.80 (q, 6H, J ' 1 Hz): ir (neat) 2950 (w), 1695 (s), 1650 (w), 1450 (s), 1355 (s), 1200 (s), 1090 (s), 1070 (m), 970 (m), 875 (w), 765 (w) cm'lt uv (MeOH))«max 217 nm (5 4,800), 255 (6, 400): mass spectrum (70 ev) m/e (rel inten- sity) 178 (51), 163 (7), 136 (100), 135 (63), 121 (72), 120 (20), 119 (42), 107 (21), 105 (32), 93 (17), 91 (27). COmpound‘24 was iden- tified as pentamethylphenyl acetate by comparing its spectral data with those of an authentic sample: NMR (CCl4) 8 1.97 (s, 6H), 2.12 (s, 9H), 2.20 (s, 3H); ir (KBr) 2950 (m), 1740 (s), 1460 (m), 1380 (m), 1230 (s), 1090 (m) cm'l; uv (MeOH) Am 227 nm (2 3,000); mass spectrum (70 ev) m/e (rel intensity) 206 (17), 165 (13), 164 (100), 163 (11), 149 (63), 105 (12), 91 (13). Compound 12 was identified as pentamethyl- phenol by comparing its NMR data with those of an authentic sample: 43 2.12 (5): mp 127-128°c. 13. Photolysis of 13* The procedure and workup were as described for the photolysis of 12. The crude product mixture had an NMR spectrum identical with that of photolysis of 12 except that the signal at S 1.26 was reduced in area to half (europium shifted spectrum showed that the signal with shift slope 1.34 was absent) and the signal at.8 2.20 was absent. 22? had an NMR spectrum identical with that of 22 except that the signal at 5 1.45 was reduced 50% in area. 14. Addition of Dichlorocarbene to Hexamethylbicyclo[3.2.0]hepta- 3,6-dien-2-one (1) To a solution of 3.8 g (20 mmol) of hexamethylbicyclo(3.2.0)- hepta-3,6-dien-2-one in 80 ml of CHCl3 containing 500 mg of triethyl- benzylammonium chloride at 0°C was added dropwise with stirring 54 ml of 50% aqueous NaOH solution, and the mixture was stirred at room temperature for 72 hr. The reaction mixture was then diluted with 400 ml of water and extracted with CHZClZ. The combined CHZCl2 layer was washed successively with water, saturated NaCl solution and dried (M9504). After removal of the solvent under reduced pressure, the residue was chromatographed on silica gel (EM60, finer than 230 mesh) using methylene chloride as eluent. The first fraction gave 3.2 g of light yellow crystals, which were recrystallized from petroleum ether (bp 30-60°) to give 2.5 g (9 mmol, 45%) of 29 as white crystals, mp 91-92°c; NMR (CCl4) s 1.05 (s, 3H), 1.22 (s, 3H), 1.28 (s, 3H), 1.30 44 (s, 3H), 1.43 (q, 3H, J= 1H2), 1.65 (q, 3H, J = 1 Hz); The six signals were shifted by europium shift reagent to give the following slopes: 3.18, 2.23, 1.29, 1.46, 1.70, 1.00. Ir (KBr) 3000 (m), 1715 (s), 1645 (m), 1445 (m), 1385 (m), 1295 (m), 1160 (m), 1095 (w), 1065 (w), 1015 (s), 905 (w), 895 (m), 815 (m), 765 (w) cm‘lz uv (MeOH) "max 230 nm (2 2,210), 305 (440); mass spectrum (70 ev) m/e (rel intensity) 274 (0.8), 272 (0.8), 239 (24), 237 (72), 221 (10), 173 (18), 158 (15), 139 (65), 137 (100), 135 (95), 120 (22), 119 (37), 107 (24), 105 (22), 93 (20), 91 (28). 14224. Calcd. for C14H1800l2: C, 61.54; H, 6.64. Found: C, 61.54; H, 6.53. The second fraction gave 300 mg of light yellow crystals which ‘were recrystallized from petroleum ether (bp 30-600) to give 240 mg (0.75 nmol, 3.8%) of 21 as white crystals, mp 112-113°c; NMR (CCl4) 8 1.18 (s, 3H), 1.33 (s, 3H), 1.60 (q, 3H, J = 1 Hz), 1.70 (s, 3H), 1.90 (q, 3H, J = 1 Hz), 1.92 (d, 1H, J = 8 Hz), 2.16 (d, 1H, J = 8 Hz); ir (KBr) 3000 (m), 2950 (w), 1690 (s), 1640 (s), 1450 (s), 1390 (s), 1325 (m), 1260 (w), 1235 (w), 1080 (w), 1050 (w), 960 (m), 775 (m), 755 (w), 700 (w) oh'l; uv (MeOH) am 242 nm (6 8,700), 332 (270): mass spectrum (70 ev) m/e (rel intensity) 322 (2), 320 (6), 318 (6), 224 (18), 222 (54), 207 (15), 187 (100), 159 (32), 127 (13), 128 (15), 115 (9), 105 (7), 91 (11). 1991, Calcd. for c H on 15 17 3 Found: C, 56.33; H, 5.32. : C, 56.36; H, 5.36. The third fraction gave 275 mg of light yellow crystals, which were recrystallized from petroleum ether (bp 30-600) to obtain 210 mg 45 (0.66 mmol, 3.3%) of 22 as white crystals, mp 144°C; NMR (CCl4) S 1.00 (s, 3H), 1.08 (s, 3H), 1.57 (d, 1H, J = 8 Hz), 1.69 (q, 3H, J = 1 Hz), 1.93 (s, 3H), 1.95 (q, 3H, 0 a 1 Hz), 1.97 (d, 1H, J 2'8 Hz); ir (KBr) 3000 (w), 2950 (w), 1700 (s), 1650 (m), 1450 (s), 1385 (s), 1330 (m), 1275 (w), 1235 (w), 1085 (m), 1020 (s), 965 (m), 880 (w), 810 (w), 775 (s), 735 (w) cm'l; uv (MeOH))s.max 244 nm (8.6.230); mass spectrum (70 ev) m/e (rel intensity) 322 (2), 320 (6), 318 (6), 287 (11), 285 (65), 283 (100), 268 (11), 255 (6), 247 (8), 235 (9), 219 (21), 205 (9), 185 (22), 153 (14), 128 (16), 115 (17). .5221; Calcd. for C15H170Cl : C, 56.36; H, 5.36. 3 Found: C, 56.29; H, 5.25. 15. Addition of Dichlorocarbene to 4-Trideuterohexamethylbicyclo- (3.2.0)hepta-3,6-dien-2-one (1*) The procedure and workup were as described for the addition of dichlorocarbene to 1, The product 24* had an NMR spectrum identical with that of 29 except that the singlet at S 1.28 disappeared: 21? was identical with that of 21 except that the quartet at 8 1.90 dis- appeared and the quartet at 8 1.60 collapsed to a singlet; 22? was identical with that of 22 except that the quartet atis 1.95 disap- peared and the quartet at 8 1.69 collapsed to a singlet. 16. Silver Perchlorate Promoted Ring Expansion of 50 To a solution of 273 mg (1 mmol) of 29 in 1 ml of 90% aqueous acetone was added a solution of 1.04 g (5 mmol) of silver perchlorate in 20 ml of 90% aqueous acetone. The mixture was stirred at room 46 temperature for 24 hr. After the silver chloride precipitate was. removed by filtration, 10 ml of water was added. The solution was extracted with ether and the combined ether layers were washed with water and dried (M9504). Evaporation of the solvent left 248 mg of a light yellow oil which was shown by NMR to consist of 40% of unreact- ed §Q_and 60% of a product considered to be 4-chloro-5-hydroxy-1,3,5, 6,7,8-hexamethylbicyclot4.2.0]octa-3,7-dien-2-one (55). The mixture was r" chromatographed on a thin layer silica gel plate using CHZCl2 as eluent to obtain 90 mg of 55 as light yellow solids. It was further purified by preparative vpc (65:0.25 in column, 15% SE-30 on chromo- sorb w, 60/80, 210°, 60 ml/min) to give white solids. NMR (CCl4) 6 1.06 (s, 3H), 1.20 (s, 3H), 1.23 (s, 3H), 1.40 (q, 3H, J = 1 Hz), 1.62 (q, 3H, J = 1 Hz), 1.80 (s, 3H), 2.03 (broad singlet, 1H); ir (KBr) 3460 (s), 2980 (m), 2940 (m), 1660 (s), 1610 (m), 1440 (m), 1380 (m), 1300 (s), 1190 (m), 1170 (m), 1060 (m), 965 (m), 930 (m), 800 (w), 750 (m) cm‘l; uv (MeOH) "max 237 nm (2 2,700), 248 (4,550), 285 (1,230); mass spectrum (70ev) m/e (rel intensity) 256 (34), 254 (100), 241 (13), 239 (34), 219 (67), 211 (27), 204 (29), 203 (68), 175 (45), 173 (44), 147 (33), 135 (28), 133 (28), 119 (29), 105 (33), 91 (51), 83 (49). lflyiL; Calcd. fOr C14H1902Cl: C, 66.00; H, 7.52. Found: C, 66.01; H, 7.44. 17. Acid—Catalyzed Rearrangement of 50 in Sulfuric Acid To a solution of 273 mg (1 mmol) of 59 in 1 ml of CHZCl2 and 5 ml of MeOH was added dropwise, at 0°C, 3 ml of concentrated sulfuric acid. The mixture was stirred at 0°C for 10 minutes. The reaction was then 47 quenched by adding 20 ml of water and the mixture was extracted with methylene chloride. The combined methylene chloride layers were washed successively with saturated NaHCO3 solution, water, saturated NaCl' solution and dried (M9504). Evaporation of the solvent left 245 mg of a yellow-brown oil. The crude mixture was chromatographed on silica gel (EH 60, finer than 230 mesh) using methylene chloride as solvent. The first fraction gave 23.5 mg (0.1 mmol, 10%) of 56. NMR (CCl4) 5 1.20 (s, 6H), 1.57 (q, 3H, J = 1H2), 1.70 (q, 3H, J = 1 Hz), 1.87 (s, 3H), 4.80 (s, 1H), 4.88 (s, 1H): ir (neat) 2980 (m), 2940 (m), 1775 (s), 1510 (w), 1450 (m), 1380 (m), 985 (m), 885 (m) cm‘l; uv (MeOH))smax 253 nm (a 11,800), 217 (8,570); mass spectrum (70 ev) m/e (rel intensity) 238 (15), 236 (40), 223 (12), 221 (32), 208 (24), 201 (15), 195 (13), 193 (36), 173 (100), 158 (40), 157 (24), 154 (22), 143 (21), 141 (19), 128 (21), 119 (28), 115 (20), 91 (23). Due to the compound's thermal instability, no attempt was made to obtain its elemental analysis. The second fraction gave 170 mg (0.72 mmol, 72%) of 5]. NMR (CCl4) 5 1.42 (s, 3H), 1.67 (q, 6H, J = 1 Hz), 1.87 (q, 3H, J = 1 Hz), 2.00 (q, 3H, J = 1 Hz), 4.83 (s, 1H), 4.87 (s, 1H); ir (neat) 3000 (m), 2950 (m), 1705 (s), 1640 (s), 1450 (s), 1385 (s), 1325 (m), 1020 (m), 915 (s), 900 (m), 855 (m), 755 (m), 720 (s) cm'l; uv (MeOH))»max 237 nm (5 14,100), 260 (shoulder, 7,880), 339 (980); mass spectrum (70 ev) m/e (rel intensity) 238 (26), 236 (76), 223 (10), 221 (30), 202 (35), 201 (84), 187 (54), 186 (21), 185 (20), 173 (100), 159 (30), 158 (35), 157 (21), 143 (22), 142 (15), 141 (18), 128 (22), 119 (19), 115 (18), 105 (14), 91 (22). Due to the compound's thermal instability, 48 no attempt was made to obtain its elemental analysis. 18. Acid-Catalyzed Rearrangement of 50* The procedure and workup were as described for the acid-catalyzed rearrangement of 59, The product 56f had an NMR spectrum identical with that of 56 except that the singlet at.5 1.87 was absent; 57f was identical with that 51 and showed no deuterium labeling. 19. Treatment of 57 in Trifluoroacetic Acid-d1 A solution of 15 mg of 51 in 0.5 ml of trifluoroacetic acid-d1 was stirred at room temperature for 10 min. The mixture was then poured into 020. After workup, the product showed identical NMR spectrum with that of 51 except that the signals at,8 1.87, 4.83 and 4.87 were disappeared. Europium-shifted spectrum showed that the signal at 5 1.67 (with Eu-shift slope 1.00) sharpened to a singlet. 20. Treatment of 57 in CH30Na/CH300 solution A solution of 15 mg of 51 in 1 ml of CH 00 and 10 mg of CH 0Na 3 3 was stirred at room temperature for 0.5 hr, then poured into ice- water and extracted with methylene chloride. NMR spectrum showed that §Z_had been converted into a different compound due to its instability under basic condition. BIBLIOGRAPHY 1 2. 3. 4 6. 7. R. K. Lustgarten, M. Brookhart, and S. Hinstein, J. Am. Chem. Soc., BIBLIOGRAPHY H. Hart and M. Nitta, Tetrahedron Lett., 2109 (1974) H. Hart and M. Nitta, Tetrahedron Lett., 2113 (1974) Sun-Mao Chen, Ph.0. theses, Michigan State University, 1975 (a) L. A. Paquette, “Modern Heterocyclic Chemistry", 4th ed, H.A. Benjamin, 1976, Chapter 1 (b) House "Modern Synthetic Reactions", 2nd ed, H.A. Benjamin, Menlo Park, Calif., 1972 . For an example, see 4(b), p. 320 H. Hart and 1. Huang, J. Org. Chem., 39, 1005 (1974) "‘ _8_9, 6350 (1967) 8. Unpublished result of H. Hart and R. Gupta. 9. (a) A. Padwa, in Organic Photochemistry (0. L. Chapman, ed.), vol. 1, Dekker, New York, 1967, p. 91 (b) A. Padwa, Acc. Chem. 33,4, 48 (1971) (c) N. R. Bertoniere and G. N. Griffin, in Organic Photochemistry (O. L. Chapman, ed.), vol. 3, Dekker, New York, 1973, p. 138 10. (a) S. S. Hixson, P. S. Mariano, and H. E. Zimmerman, Chem. Rev., 13, 531 (1973) (b) H. G. Dauben, G. Lodder and J. Ipaktschi, Topics in Current Chemistgy,,§fl, 73 (1973) (c) K. N. Houk, Chem. Rev., 76, 1 (1976) 11. J. Ehrenfreund, Y. Gaoni and 0. Jeger, Helv. Chim. Acta., 52, 2704 (1974) 12. M. Mokosza and M. Hawrzyniewicz, Tetrahedron Lett., 4659 (1969) 13. T. A. Antkowiak, D. C. Sanders, G. B. Trimitsis, J. 8. Press, and H. Shechter, J. Am. Chem. Soc., 24, 5366 (1972) 49 191.4: ”III 14. 15. 16. 17. 18 50 (a) R. F.)Childs and S. Hinstein, J. Am. Chem. Soc., 99, 7146, 1968 (b) H. Hart, T. R. Rodgers, and J. Griffiths, ibid., 21, 754 1969 (c) V. A. Koptyug, L. I. Kuzubora, I. S. Isaev, and V. I. Mamatyuk, Chem. Commun., 389 (1969) D. H. Swatton and H. Hart, J. Am. Chem. Soc., 82, 5075 (1967) J. A. Benson and J. A. Jenkins, J. Am. Chem. Soc., 24, 8907 (1972) R. F. Childs and C. V. Rogerson, J. Am. Chem. Soc., 28, 6391 (1976) H. J. Hehre, J. Am Chem..Soc., 26, 5207 (1974) 5 5 5 5 5 III 5 5 5 III 5 '5 all 5 I ll 5 III 5 5 1293 03196 5274 3 'IHIHIHIHHIIWIW I l