ABSTRACT PART I SYNTHESIS AND PHOTOCHEMISTRY OF CYCLOHEXADIENONE EPOXIDES PART II ACID-CATALYZED REARRANGEMENTS OF CYCLOHEXADIENONE EPOXIDES PART III DIELS-ALDER REACTIONS OF A DIHYDROBENZOPENTALENE AND THE SYNTHESIS OF HIGHLY-STRAINED QUADRICYCLANES PART IV MISCELLANEOUS BY Cheng-Tai Peng The photochemistry of cyclohexadienone epoxides, $3, gé ahd $2! is described in Part I of this thesis. The irradiation of 4,S-epoxy-2,4,5,6,6-pentamethy1-2-cyclo- hexenone (3%) in ether through a Pyrex filter geve l,3,3,4,5- pentamethy1-7-oxabicyclo[2.2.1]hept-5-ene-2-one (QZ' 25%), 4-acety1-2,4,5,S-tetramethyl-2-cyclopentenone (33, 44%), and 3.3.4.6,7-pentamethy1-2(3H)-oxepinone (33, 30%). Cheng-Tai Peng O H o pyrex 3% 31(25%) %g(45%) 32(30%) A new epoxyketone photorearrangement leading from $3 to g3 involving a cyclopropanone-aldehyde intermediate was proposed. Separated irradiation of $1 and 33 did not give any 32, eliminating 31 and 3% as reaction intermediates. Deuterium-labeling experiments support the proposed mechan- ism for the photorearrangement of 3%. The irradiation of 4,5-epoxy-2,3,4,6,6-pentamethyl- 2-cyclohexenone (3%) in ether through a Pyrex filter afforded 322351,3,3,5,6-pentamethylbicyclo[3.l.O]hexan- 2,4-dione (33, 62%), gynf1,3,3,5,6-pentamethy1bicyclo- [3.1.0]hexan-2,4-dione (§8, 11%) and lactone (él, 27%). pyrex éé 42(62%) §Q<11%) ék‘27%) Deuterium-labeling experiments were consistent with the proposed mechanisms for the photorearrangement of gg. The irradiation of 4,S-epoxy-6,G-dimethyl-Z-cyclo- hexenone (gé) in ether through pyrex gave 3,3-dimethyl-2(3H)- Cheng-Tai Peng oxepinone (éér 86%) and 6,6-dimethyl-2-cyclohexen-l,S-dione (31, 14%). Each of these photoproducts underwent further photoisomerization through a Corex filter. Compound ég rearranged to a single photoisomer, 2,2-dimethy1-4—oxa- bicyclo[3.2.0]hept-6-en-3-one (ég) and compound éz re- arranged to 3,3-dimethyl-bicyclo[3.1.0]hexan-2,4-dione (ég) which was also photolabile and rearranged further to 2-9357 (l'-hydroxy-2'-methyl-1'-propenyl)-cyclopropanecarboxylic acid-yélactone (Qg). O hv ____.’ + Pyrex \ / O O 23% élé hv l Corex O hv Corex Cheng-Tai Peng The photoisomerization of 36 could be sensitized (aceto- phenone, benzophenone) and quenched (Eggngfl,3,5-hexa- triene, but not piperylene). The formation of 56 from 36 is presumed to be the same as that proposed for the forma- tion of 32 from éé which involves a cyclopropanone-aldehyde intermediate. In an attempt to trap the cyclopropanone, compound 36 was irradiated in methanol at room temperature and at -78°, only 56 and 23 were formed, in about the same ratio as in ether. However, after irradiation of 36 in tetrahydrofuran at -105°, a weak band at 1815 cm"1 was observed. When the solution was warmed to room tempera- ture, the band at 1815 cm'l, which may be attributed to the cyclopropanone carbonyl, gradually disappeared. The acid-catalyzed rearrangement of cyclohexadienone epoxides, £4, £3 and 36, was studied in Part II of this thesis. In trace acid, epoxide 3% rearranged quantita- tively to 5-hydroxy-4-methylene-2,5,6,6-tetramethyl-2- cyclohexenone (8%). In neat trifluoroacetic acid (TFA), compound 34 rearranged to 4-methylene-2,S-dimethyl-Z-cyclo- pentenone (85, 4%), 5-isopropeny1-4-methy1ene-2,5-dimethy1- 2-cyclopentenone (86, 20%), 2,4,4,6,6-pentamethyl-2-cyclo— hexen-1,5-dione (41, 10%), 4-acety1-2,4,5,5-tetramethyl- 2-cyclopentenone (38, 8%), 4-trifluoroacetoxymethyl-2,5,6,6- tetramethyl-2,4-cyclohexadienone (8;, 52%) and 4-hydroxy- methyl-2,5,6,6-tetramethyl-2,4-cyclohexadienone (88, 6%). When 84 was treated with TFA, the same products were ob- tained and the product ratios were almost the same. On Cheng-Tai Peng longer treatment with TFA, 86 was dealkylated to 85 and acetone. Saponification of gz gave 88 in quantitative yield. Deuterium-labeling experiments support the pro- posed mechanism. ? O H-I- 0 (trace)’ 0 0H 11 Q + H .24 \fi/ 4% O H o ’4, iii + o m m 0 j. 0 CH2 81 R: OCOCF3 an. R m R= 0H Rearrangement of epoxyketone 35 in TFA at room tem- perature gave 2,3,4,6,6-pentamethyl-2-cyclohexen-l,5-dione (54, 60%) and 4-methylene-5-hydroxy-2,3,6,6-tetramethyl-2- cyclohexenone (32, 40%). 3 8 801 F3 '11 13’ 5° .I + g 0 Cheng-Tai Peng In contrast to compound 81, 21 underwent no further re- arrangement on treatment with TFA and this was attributed to preferential protonation at the carbonyl oxygen to give the highly delocalized cation 1. on 1.96 * 1'24 1.32 H 4.46 2.23 : OH 5.66 5.72 Treatment of the epoxyenone 16 in TFA gave the tri- fluoroacetate 98 (95%). ’VD TFA as 0 NW OH OCOCF3 \V Prolonged treatment of 16 with TFA did not bring about any carbocation skeletal rearrangement. In Part III of this thesis, Diels-Alder reactions of a dihydrobenzopentalene 191, in particular with acetyl- enic dienophiles, and the synthesis of highly-strained quadricyclanes were studied. The highly substituted benzodihydropentalene 181 gave Dials-Alder adducts with dimethylacetylene dicarboxylate, diethyl azodicarboxylate, 3,6-dimethylbenzyne and 2-butyne. Irradiation of 192 and 111 in ether through vycor gave a good conversion to quadricyclane derivatives of 116 and 111, respectively. NU ~‘S ‘L 1th +, ‘\ ‘I‘i Cheng-Tai Peng ‘ 0“ ROH HCl 93% x O.— CH2N2 95% CHBOOC COOCH3 HOOC CQQH $82 kké A hv hv A 100% 94% 92% 100% I 0‘1 , “$222 0‘ kké k%1 Some miscellaneous results are described in Part IV of this thesis. In Part IV (1), a new alkylation reagent - high surface sodium (HSS) was described to generate enolates which can then be alkylated. In the cases of hexamethyl-2,4-cyclo- hexadienone, alkylation occurred gig a cross - rather than a fully conjugated to give a single product. Methylation gig the addition of 2-tetralone to HSS proceeded smoothly to yield solely the l-alkylated derivative. In Part IV (2), the synthesis of the diepoxy ketone 111 is described. tra; a\ 9’. a 2.01 f; Cheng-Tai Peng kéé Attempts to rearrange 188 by photolysis and by acid are also described. In Part IV (3), the synthesis of bicyclo[3.2.1]octa- dienyl carbocation 188 is described. The Diels-Alder O O +cn caccn 21° 3 3 H 3 days 3% 181 13% H trace H+ FSOBH/5027C1F aq. acetone:> / _100 H $82 $48 addition of 2-butyne to pentamethyldienone 81 gave 1,2,3,- 5,6,8,8-heptamethyl-bicyclo[2.2.2]octa-2,5-dien-7-one (181, 93%). Reduction of 181 with lithium aluminum hydride gave 1,3,3,5,6,7,8-heptamethyl-bicyclo[2.2.2]octa-5,7-dien- 2-ol (188, 90%). Treatment of 188 with a trace of acid in C W C\ X ‘5‘ u ‘Hfllwv i. It +L 4V... .- ‘II‘ n Cheng-Tai Peng acetone caused it to dehydrate with rearrangement to give 2-methy1ene-3,4,6,7,8,8-hexamethylbicyclo[3.2.1]octa-3,6- diene (188, 41%). When 188 was treated at -78° with FSO3H/ S02C1F, the stable carbocation, heptamethylbicyclo[3.2.l]- octa-3,6-dien-2-yl cation (118) was obtained. In Part IV (4), alkylation studies with 4-methylene- 2,3,S-trimethyl-2-cyclopentenone (81) are described. Alkylation of 81 with methyl iodide gave 4-methylene-2,3,5,5- tetramethyl-2-cyclopentenone (181, 92%) and alkylation of 0 1‘ [(CH3)3Si]2N‘Li+ Kn > -——————9 -- CH3: c331 g1 181(92%) 188(90%) [(CH3)3Si]2N-Li* 0 o , + + o <— / / 147(55%) l48(15%) 149(18%) ’VVV 'V'Vb ’Vb’b 181 gave 4-methy1ene73-ethyl—2,5,5-trimethyl-2-cyclopentenone (118, 90%). Further alkylation of 118 gave 111 (55%), 118 (15%) and 118 (18%). Alkylation of 81 with allyl bromide gave 188 (66%) and 181 (20%). Cheng-Tai Peng [(CH3)3Si]2 N Li+ %f\ ‘ \ Q; CH2=CHCHZB 188(66%) 181(202) Treatment of 81 with methyllithium followed by dehydration gave the triene 188 in 40% yield. kéé %é%(4°%) PART I SYNTHESIS AND PHOTOCHEMISTRY OF CYCLOHEXADIENONE EPOXIDES PART II ACID-CATALYZED REARRANGEMENTS OF CYCLOHEXADIENONE EPOXIDES PART III DIELS-ALDER REACTIONS OF A DIHYDROBENZOPENTALENE AND THE SYNTHESIS OF HIGHLY-STRAINED QUADRICYCLANES PART IV MISCELLANEOUS BY Cheng-Tai Peng A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1977 ACKNOWLEDGMENTS I am deeply indebted to Professor Harold Hart for his guidance, encouragement and leadership throughout the course of this study. I gratefully acknowledge financial support from the National Science Foundation, National Institute of Health and Michigan State University in the form of research and teaching assistantships. ii TABLE OF CONTENTS PART I SYNTHESIS AND PHOTOCHEMISTRY OF CYCLOHEXADIENONE EPOXIDES (1) 4,5-EPOXY-2,4,5,6,6-PENTAMETHYL-2-CYCLOHEXENONE (2) 4,5-EPOXY-2,3,4,6,6-PENTAMETHYL-2-CYCLOHEXENONE (3) 4,S-EPOXY-G,6-DIMETHYL-2-CYCLOHEXENONE Page INTRODUCTION 0 O O O O O O O O O O O O O O O O C O 2 RESULTS AND DISCUSSION 0 O O C O O O O O O O O O O 15 l. Photochemistry of 4,5—Epoxy-2,4,5,6,6- pentamethyl-2-cyclohexenone (88) . . . . . 15 2. Photochemistry of 4,5-Epoxy-2,3,4,6,6- pentamethyl-2-cyclohexenone (88) . . . . . 26 3. Photochemistry of 4,5-Epoxy-6,6-dimethy1- 2-cyclohexenone (18) . . . . . . . . . . . 35 EXPERIMENTAL . . . . . . . . . . . . . . . . . . . 49 General Procedure. . . . . . . . . . . . . . . 49 General Photolysis Procedure . . . . . . . . . 50 1. Synthesis of 4,5eEpoxy-2,4,5,6,6- pentamethyl—Z-cyclohexenone (81) . . . . . 51 2. Synthesis of 2,5-Trideuteromethy1-4,S- epoxy-2,4,6,6,-tetramethy1-2-cyclohexenone (11*). . . . . . . . . . . . . . . . . . . 52 3. Irradiation of 81. . . . . . . . . . . . . 53 4. Irradiation of 81* . . . . . . . . . . . . 55 5. Irradiation of 4-Acetyl-2,4,5,5— tetramethyl-2-cyclopentenone (88). . . . . 56 6. Irradiation of l,3,3,4,5-Pentamethy1- 7-oxabicyclo[2.2.1]hept-5-en-2-one (81). . 57 7. Synthesis of 4,5-Epoxy-2,3,4,6,6-pentamethyl- 2-cyclohexenone (88) . . . . . . . . . . . 57 iii Page 8. Synthesis of 3-Trideuteromethyl- 4,5-epoxy-2,4,6,6-tetramethyl-2-cyclo- hexenone (88*) . . . . . . . . . . . . . . . 58 9. Irradiation of 88. . . . . . . . . . . . . . 59 10. Irradiation of 88* . . . . . . . . . . . . . 61 11. Irradiation of 2,3,4,6,6-Pentamethy1- 2-cyclohexen-l,5-dione (88). . . . . . . . . 61 12. Irradiation of 88* . . . . . . . . . . . . . 62 13. Synthesis of 4,5-Epoxy-6,6-dimethy1- 2-cyclohexenone (88) . . . . . . . . . . . . 62 14. Irradiation of 88. . . . . . . . . . . . . . 63 15. Irradiation of 88 using a Corex Filter or 254 nm Light. . . . . . . . . . . . . . . 65 16. Irradiation of 88 at Low Temperature . . . . 66 17. Irradiation of 3,3-Dimethyl-2(3H)- oxepinone (88) . . . . . . . . . . . . . . . 68 18. Irradiation of 6,6-Dimethyl-2-cyclo- hexen-l,5-dione (81) . . . . . . . . . . . . 69 19. Irradiation of 3,3-Dimethy1-bicyclo- [3.1.0]hexane-2,4-dione (88) . . . . . . . . 69 20. Sensitization and Quenching Studies. . . . . . . . . . . . . . . . . . . 70 PART II ACID-CATALYZED REARRANGEMENTS OF CYCLOHEXADIENONE EPOXIDES (1) 4,5-EPOXY—2,4,5,6,6-PENTAMETHYL-2-CYCLOHEXENONE (2) 4,5-EPOXY-2,3,4,6,6-PENTAMETHYL-2-CYCLOHEXENONE (3) 4,5-EPOXY-6,6-DIMETHYL-Z-CYCLOHEXENONE INTRODUCTION . . . . . . . . . . . . . . . . . . . . 75 RESULTS AND DISCUSSION . . . . . . . . . . . . . . . 84 l. Acid-catalyzed Rearrangement of 4,5- Epoxy-2,4,5,6,6-pentamethy1-2- cyclohexenone (88) . . . . . . . . . . . . . 84 1V 3. Page Acid-catalyzed Rearrangement of 4,5- Epoxy-2,3,4,6,6-pentamethy1-2-cyclo- hexenone (88). . . . . . . . . . . . . . . . 96 Acid-catalyzed Rearrangement of 4,5- Epoxy-G,6-dimethy1-2-cyclohexenone (88)... .103 EXPERIMENTAII O O O O O O O O O O O O I C O O O O O O 1 o 8 l. 5. 6. 10. Acid-catalyzed Rearrangement of 4,5- Epoxy—2,4,5,6,6-pentamethyl-2-cyclo- hexenone (88). . . . . . . . . . . . . . . .108 Acid-catalyzed Dealkylation of 5- Isopropenyl-4—methylene-2,5-dimethyl- 2—cyclopentenone (88). . . . . . . . . . . .111 Saponification of 4-Trif1uoroacetoxy- methyl-2,5,6,6-tetramethy1-2,4-cyclo- hexadienone (81) . . . . . . . . . . . . . .112 Synthesis of 5-Hydroxy-4-methylene- 2,5,6,6-tetramethyl-2-cyclohexenone (88) . . . . . . . . . . . . . . . . . . . .113 Acid-catalyzed Rearrangement of 88 . . . . .114 Acid-catalyzed Rearrangement of 5- Trideuteromethyl-4,5-epoxy-2,4,6,6- tetramethyl—2-cyclohexenone (88*). . . . . .114 Acid-catalyzed Rearrangement of 88*. . . . .115 Acid-catalyzed Rearrangement of 4,5- Epoxy-2,3,4,6,6-pentamethyl-2-cyclo- hexenone (88). . J . . . . . . . . . . . . .115 Acid-catalyzed Rearrangement of 3- Trideuteromethy1-4,5-epoxy-2,4,6,6- tetramethyl-2-cyclohexenone (88*). . . . . .117 Acid-catalyzed Rearrangement of 4— Methylene-S-hydroxy-Z,3,6,6-tetra- methyl—Z-cyclohexenone (88). . . . . . . . .117 Page 11. Acid-catalyzed Rearrangement of 88*. . . . .119 12. Saponification of Trifluoroacetyl derivative 88. . . . . . . . . . . . . . . .120 13. Saponification of 88*. . . . . . . . . . . .120 14. Acid-catalyzed Rearrangement of 4,5- Epoxy-6,6-dimethy1-2-cyclohexenone (88) in Hydrochloric Acid. . . . . . . . . .120 15. Acid-catalyzed Rearrangement of 88 in Hydrogen Chloride . . . . . . . . . . . .121 16. Acid-catalyzed Rearrangement of 88 in Trifluoroacetic Acid. . . . . . . . . . .122 PART III DIELS-ALDER REACTIONS OF A DIHYDROBENZOPENTALENE AND THE SYNTHESIS OF HIGHLY-STRAINED QUADRICYCLANES INTRODUCTION . . . . . . . . . . . . . . . . . . . .125 RESULTS AND DISCUSSION . . . . . . . . . . . . . . .128 1. Structures of the Diels-Alder Adducts. . . . . . . . . . . . . . . . . . .128 2. Quadricyclane Synthesis. . . . . . . . . . .131 EXPERIMENTAL . . . . . . . . . . . . . . . . . . . .137 1. Diels-Alder Adducts of the Dihydro- benzopentalene (888) . . . . . . . . . . . .137 2. Saponification of the Diels-Alder Adduct 888 . . . . . . . . . . . . . . . . .140 3. Photoisomerization of Diene Diacid 888 to Quadricyclane Diacid 888. . . . . . .141 4. Photoisomerization of Diene Diester 888 to Quadricyclane Diester 888 . . . . . .142 5. Isomerization of Quadricyclane Diacid 888 to Diene Diacid 888. . . . . . . . . . .143 vi T“ $.\ I" u ‘1‘ Page 6. Isomerization of Quadricyclane Diester 11g to Diene Diester 1Qg . . . . . . . . . .143 7. Esterification of Quadricyclane Diacid 111 to Quadricyclane Diester 11g . . . . . .143 8. Esterification of Diene Diacid 11g to Diene Diester 193 . . . . . . . . . . . .144 PART IV MISCELLANEOUS (l) A NEW ALKYLATION REAGENT - HIGH SURFACE SODIUM INTRODUCT ION o o o o o o o o o o o o o o o o o o o o 1 4 6 RESULTS AND DISCUSSION . . . . . . . . . . . . . . .150 (2) SYNTHESIS OF POTENTIAL CARCINOGENIC DIEPOXIDES INTRODUCTION . . . . . . . . . . . . . . . . . . . .154 EXPERIMENTAL . . . . . . . . . . . . . . . . . . . .158 1. Synthesis of 2,3:4,5-Diepoxy- 2,3,4,5,6,6-hexamethy1-2,4- cyclohexadienone (13%) . . . . . . . . . . .158 2. Synthesis of 2,3:4,5-Diepoxy-3-tri- deutercmethy1-2,4,5,6,6—hexamethyl- 2,4—cyclohexadienone (1§g*). . . . . . . . .159 3. Synthesis of 2,3:4,5-Diepoxy-3,5-bis- trideuteromethy1-2,4,6,6-tetramethy1- 2,4-cyclohexadienone (13g**) . . . . . . . .159 4. Irradiation of 13% . . . . . . . . . . . . .160 5. Acid-catalyzed Rearrangement of 13%. . . . .160 vii Page (3) THE EFFECT OF METHYL GROUPS AT A BRIDGE-HEAD POSITION ON THE COMPETING CARBONIUM ION RE- ARRANGEMENT OF THE BICYCLO[3.2.l]OCTA-3,6-DIEN- 2-YL SYSTEM INTRODUCTION . . . . . . . . . . . . . . . . . . . .162 RESULTS AND DISCUSSION . . . . . . . . . . . . . . .165 EXPERIMENTAL . . . . . . . . . . . . . . . . . . . .169 1. ‘1,2,3,5,6,8,8-Heptamethy1bicyclo- [2.2.2]octa-2,5—dien-7-one (131) . . . . . .169 2. 1,3,3,5,6,7,8-Heptamethy1bicyclo- [2.2.2]octa-5,7-dien-2-ol (13fi) . . 3. 2-Methy1ene-3,4,6,7,8,8-hexamethy1- bicyclo[3.2.1]octa-3,6-diene (133) . . . . .171 4. HeptamethylbicycloE3.2.1]octa-3,6- dien-Z-yl Cation (1&8) . . . . . . . . . . .172 .170 (4) ALKYLATION STUDIES WITH 4-METHYLENE- 2,3,5-TRIMETHYL-2-CYCLOPENTENONE INTRODUCTION . . . . . . . . . . . . . . . . . . . .173 RESULTS AND DISCUSSION . . . . . . . . . . . . . . .174 EXPERIMENTAL . . . . . . . . . . . . . . . . . . . .180 1. Alkylation of 4-Methy1ene-2,3,5-trimethy1-2- cyclopentenone (g1). . . . . . . . . . . . .180 2. Alkylation of 4-Methy1ene-2,3,5,5- tetramethyl-Z~cyclopentenone (14$) . . . . .181 3. Alkylation of 4-Methy1ene-3-ethy1- 2,5,S-trimethyl-2-cyclopentenone (13$) . . .182 4. Alkylation of £1 with Allyl Bromide. . . . .183 5. Treatment of fik with Methyllithium . . . . .185 BIBLIOGRAPHY....................186 viii PART I SYNTHESIS AND PHOTOCHEMISTRY OF CYCLOHEXADIENONE EPOXIDES (1) 4,5-EPOXY-2,4,5,6,6-PENTAMETHYL-2-CYCLOHEXENONE (2) 4,5-EPOXY-2,3,4,6,6-PENTAMETHYL-2-CYCLOHEXENONE (3) 4,5-EPOXY-6,6-DIMETHYL-2-CYCLOHEXENONE INTRODUCTION The preparation of 7.6-epoxyenones from conjugated dienones can be carried out by a number of methods.1 The most common peracids used to convert dienones to epoxides have been perbenzoic acid,2 monoperphthalic acid3 and m-chloroperbenzoic acid.4 m-Chloroperbenzoic acid is the most convenient oxidizing agent. It is commercially available, and it reacts at a somewhat faster rate than perbenzoic acid, and is ideally suited for epoxidations of the cyclohexadienone system.5 0 O R MCPBA R1 R R4 R2 R R4 2 R 3 3 R1 R2 R3 R4 Ref. Me Me Me Me 5 H Me Me H 23 Me H Me Me This work Me Me Me H This work H H H H This work Preferential oxidation of the y,6-doub1e bond over the a,B-doub1e bond in conjugated dienones is a reflection of the electrophilic nature of organic peracids. Epoxida- tion reactions proceed by electrophilic attack of the peracid upon the double bond; thus the rate of epoxidation is very sensitive to the electron density at the olefinic site.6 In 2,4-cyclohexadienones, the y,6-doub1e bond has a greater electron density than the a,B-doub1e bond and selectivity is easily achieved. Among the synthetic uses of the epoxidation products, the photochemistry and the acid-catalyzed rearrangements 5,7 are of special interest. They provide an unusual array of structures from a single type of starting material. Moreover, those compounds will routinely be screened for antibiotic activities (for example, the antibiotics Magnamycin A, cirramycin A, chalcomycin, and neutramycin contain mainly the y,6-epoxy enone functional group). The first part of my thesis deals mainly with the photochemistry of y,6-epoxyenones. As we know, the un- usual ground state reactivity of oxiranes toward nucleo- philic attack has long been recognized and exploited. Only in recent years, however, has the excited-state chemistry of these inherently strained substrates been eXtensively explored and the potential synthetic utility 0f the photoreactions received attention. Here a brief survey of the photochemistry of oxiranesipertinent to the Present study will be presented. The photochemical reactions of oxiranes lacking an aromatic chromOphore have been studied. Irradiation of 8,9 this type of compound results in decomposition, iso- 10 11 The mechanisms of merization and carbene formation. photomolecular decomposition and isomerization are simple, but the carbene formation is more complicated. For ex- ample, irradiation of a benzene solution of the oxaspiro- pentane derivative 1 affords the cumulene %, an allenic alcohol 3 and allenic oxetane g. The oxetane 6 was shown to be a secondary product resulting from cycloaddition of g to acetone. ——~ >=.=-=< + >=~ Jo % + >: 84:. The cyclopropyl carbene S was proposed to rationalize the formation of 2. Aryloxiranes have been shown to undergo photofrag- mentation in solution to give arylcarbenes and carbonyl compounds.12 For example, photolysis of tetraphenyl- oxirane13 in methanol/benzene gave benzhydryl methyl ether in quantitative yield. Benzophenone, tetraphenylethylene (a primary photoproduct), and diphenyl carbene are observ- able upon irradiation of tetraphenyloxirane in a rigid glass at 77°K. hv ,}<\_ l/B\\ Pthgh —'—*' Ph Ph + Ph Ph Ph Ph Several other phenyl-substituted oxiranes have been found to add methanol photochemically.14 The photochemistry of epoxyketones has been reviewed.15 It is generally accepted that the n,n*-excited state is the chemically significant excited state involved in the photochemical transformations of this class of ketones. In this introduction, a brief review of the photo- chemistry of a,8- and B,y-epoxy ketones and a,B-unsaturated y,6-epoxyketones are presented in order to set the various possibilities'for the photochemical behavior of 34, gg, and gé in perspective. A. The Photochemistry of a,B¥Epoxyketones. The photoisomerization of an a,B-epoxy ketones usually proceeds via a 1,2-alky1 or hydrogen migration to give a B-diketone g. O o /\ m 9 *8 3 n ; RC - ca - CBRB RC -Ia (EB-RB Ra Ra RB Q Q R, Ra, R8 = H, alkyl or aryl The most rational mechanism for this reaction is that they proceed gig homolysis of the Ca-O bond. The carbonyl group can then stabilize the radical at Ca as shown in structure 1. Concurrently or at a later stage an R group migrates from C to Ca' forming a carbonyl group at C 8 8' o o ~o R8 0 R c/\ l» Vic ' -———-» RC ' H R RC - -—-C - R R‘d a—(fB RB "(fa-CS B '0: [B B A {Q R R Ra RB a B a The usual migratory order for different R groups at C 15(a-c),26 B is H > alkyl > aryl. B. The Photochemistry of 8,1rEpoxyketones. Irradiation of a 8,y-epoxy cyclic ketone 2 initially leads to Norrish type I bond cleavage with formation of a diradical which undergoes subsequent epoxide ring open- ing to give an acyl-alkoxy radical 1Q. This then under- goes ring closure to give a lactone 11 and/or hydrogen transfer to afford an aldehyde 1%. If the formation of both of these products is structurally precluded, then decarbonylation occurs to give a biradical (1é) which undergoes disproportion and/or ring closure to afford stable products. These are illustrated in Scheme 1.17 Scheme'l O O O a h\) .' .l ==== a=== -——+ B / O. 00 1 11 11 -CO 0 m~w~w O 11 11 11 11 If the hydrogen at the Y—carbon in the B,Y-epoxyketone moiety of g is replaced by an alkyl substituent, then the formation of aldehyde 1% is precluded. For example, ir- radiation of epoxy ketone 1Q gives lactone 11 as the only major reaction product.17 CH3 H 11 11 C. The Photochemistry of a,B-Unsaturated y,6-Epoxyketones. A mechanistic scheme which rationalizes all previous results on the photochemistry of a,B-unsaturated y,6-epoxy- ketones was proposed by Hart. 18 Scheme 2 {EZY B a R -C-C—C=C-C=O 1 \l l h» R2 CY-O | CY-C6 Rl—C—C-C—C=C—O cleavage E; cleavage 1* Jr 5‘2 ‘12 R -C—C=C—C=C—O R —C. C=C—C=C-o 1 . 1 \ / O- 0 products L rearrangement cyclization fragmentation, carbene formation C6+C O-C y a 1’ V R R -—' R I2 1 CO '2 R1_c_c_c=c_c=o R1_C + :C—C=C-C=O I 10 The more common reaction path involves CY-O bond cleavage at initial step. Three different types of sub- sequent reactions have been observed. These are migration 7,19(a-C) of a group from C6 to CY cyclization at Ca' or fragmentation to a carbene.21 For example, photocycliza- tion at Ca and C6 + CY hydrogen migration in the irradia- 20 tion of eucarvone 18 gave 19 and 20, respectively. ’Vb ’V'b ’Vb 22 The irradiation of trans-B-ionone 2% was interpreted in terms of a carbene intermediate formed by fragmentation of the C7-C6 bond following Cy-O bond cleavage. .A A... _-...-. . 11 C::jé;§p/Lk) _23_+ tifj::§»JR\ ___+ ::;:T’=§~JR\ 11 A 11 11 Very recently, the photochemistry of hexamethy1-2,4- cyclohexadienone-4,5-epoxide (2%) was reported.7 Irradia- tion of 2% gave 2,3,4,5,S-pentamethyl-4-acety1-2-cyclo— pentenone (2%) which then photoisomerized to endo-S-acetyl- l,3,3,4,S-pentamethylbicyclo[2.1.0]pentan-2-one (%§)° ether 0 Pyrex 11 11 11 Finally, the remarkable formation of 28, 23 and 3Q from 4,5-epoxy-3,4,6,6-tetramethy1-2-cyclohexenone 21 was recently reported.18'23 12 s~sst 11 11 11 11 The proposed intermediate leading to products 28 and 22 was not isolated and the new fragmentation mechanism lead- ing to product 3Q was open to some question. Therefore a systematic study of epoxy enones was undertaken as part one of my thesis to further investigate these new photo- reactions. It is clear from the results on 2% and 11 that the replacement of H for CH3 in certain positions can pro- foundly affect the photochemical results. Consequently, it would be desirable to study all possible substitutions systematically. One hydrogen 0 0 1 u H H . . H 11 11 13 Two hydrogens :16; H 55;; Eli: 951 if; Three hydrogens i: i. Four hydrogens 36 mm 14 Naturally not all of these compounds are readily accessible synthetically. In this thesis the results obtained with two compounds containing only one hydrogen (3%) and (gé) are described. The compound with all four hydrogens (3g) was also studied. These examples were selected for two reasons; the precursors for their synthesisle'30 were known and the latter compound particularly with all hydro- gens in the positions of interest could be compared with the already studied fully methylated compound. t‘ -_._A RESULTS AND DISCUSSION 1. Photochemistry of 4,5—Epoxy-2,4,5,§,6-pentamethyl-2- cyclohexanone (3%) The epoxyenone 3% has not been previously described. It was synthesized in good yield from the corresponding dienone 3116 and m-chloroperbenzoic acid. 0 1.73 O 1.32(2.32) MCPBA J=2 Hz d(2.23) 1.06(1.87) "'""'—’ 6.46 1.40(1.o7) 1.52(1.oo) 11 11 Compound 3% was assigned the structure shown on the basis of the following spectral data and chemical trans- formations. The molecular formula C11H1602 was confirmed by the mass spectrum (parent peak m/e 180) and elemental analysis. The ir spectrum showed a strong absorption at 1680 cm.1 for a conjugated C=O, and the uv spectrum was also consistent with conjugation, having a Amax at 250 nm. The nmr spectrum is indicated on the structure (NOTE: Figures in parentheses are the relative shifts with Eu shift reagent). All of the data show that 3% has a double bond in the apB-position, and that the epoxide ring is in the y,6-position. The europium shift data 15 s——-.—- g.» 16 were also consistent with this structure. A. Product Structures The irradiation of 3% (0.01 M in ether) through a pyrex filter was followed by vapor phase chromatography (vpc). As the peak due to 3% decreased in intensity, three new peaks appeared which were assigned to compounds with the structures 31, 38 and 32. After one hour, the peak due 0 O O 1:14:er Pyrex H O H 11 11 (25%) 1g (44%) ;g (30%) to 38 diminished in area in favor of a new peak which was assigned to the structure 38. After two hours, the peak due to 38 had disappeared, and the peak due to £9 was fully developed. Compound 31 was assigned the structure shown, on the basis of the following spectral data. 17 o 1.32 1.12(3.50) (2.98) @ 1.14(3.37) (1.93)5.56 H 1.3o(2.77) q' J=2 Hz 1.72(1.00) d, J=2 Hz 11 The molecular formula C11H1602 was confirmed by the com- pound's mass spectrum (parent peak m/e 180) and elemental analysis. The ir spectrum of this clear oil showed a strong carbonyl absorption at 1740 cm-1 (five-numbered ring ketone) and its ultraviolet spectrum had maxima at 207 nm (e 950) in methanol. The nmr spectrum is sum- marized on the structure. Compound 38 was assigned the structure shown. 0 0.90(4.80) 1.75 l.03(3.40) J=1 Hz d(3.00) 6.92 H Q(2.80) .2 . (1.03)1.95(2.50) 11 The molecular formula C11H1602 was confirmed by the mass spectrum (parent peak m/e 180) and elemental analysis. The nmr signals (see structure) were nearly identical to those of the permethyl diketone 2%. The shift reagent appeared 18 to coordinate primarily at the cyclopentenone carbonyl group. The base peak in the mass spectrum of 88 appeared O 0.90 1.75 ‘SI'r 1.03 1.95 1'9 1.13 0 11 at M-42 (loss of CH2=C=O) and the next most intense peak (rel. intensity 62) was at M-57 (loss of CH =C=0 and CH3). 2 All spectroscopic data were nearly identical to those of the known compound 88.5 Compound 88 was assigned the structure shown. C) 1.23(3.76) (1.00)l.89 ~\ // 1.83(2.03) 1.67 H (1.05) 5.60(1.92) 11 The molecular formula C 1H was confirmed by the mass 1 1602 spectrum (parent peak m/e 180) and elemental analysis. The presence of two equivalent gem-dimethyl signals as revealed by its nmr spectrum suggested that 88 possessed a plane of symmetry or readily passed through such a l9 conformation. The gem-dimethyl signal could not be separated even by using shift reagent. Besides this, the nmr spectrum showed two homoallylically coupled methyl groups at 61.67 and 1.89, and one vinyl methyl group as a doublet at 61.83. The vinyl proton appeared as a multiplet at 65.60. Decoupling at the vinyl proton using a 100 MHz nmr instrument caused the methyl signal at 61.83 to become a sharp singlet. The uv spectrum of 33 showed a maximum at 248 nm (a 10,570), similar to the uv maximum at 248 nm (log 8 3.87) reported for 1,3-cycloheptadiene by Pesch 24 and Friess. The ir spectrum of 32 had an intense carbonyl 1 which showed that there was no absorption at 1750 cm- conjugation between the carbonyl group and the diene moiety. All of the spectroscopic data were similar to those of the known compounds 8023 and 56.31 1.25 1.89 H H 5.13 1,30 1.80 02 Compound g0 was assigned the structure shown. 20 (5.47)1.68 1035(2030) Id (1.62)1.86 1.86(1.50) J = 6 Hz 82 The molecular formula C11H1602 was confirmed by its mass spectrum (parent peak m/e 180) and elemental analysis. Compound 40 showed infrared absorption bands at 1760 (vc=o) and 1662 (vc=c) cm-1 and uv maxima at 222 nm (e 7640) and at 263 nm (e 6250) which indicated that 49 is a A'-butenolide with extended conjugation. In spite of the extended conjugation, compound 49 exhibited a normal uv maximum. The reason for this is attributed to the non—planarity between the butenolide group and the exo- cyclic double bond.7'25 The nmr spectrum of 39 consisted of four vinyl methyl group signals at. 61.66, 1.86 (which showed homoallylic coupling between them) and 1.68, one methyl signal as doublet at. 61.35 and one proton at 64.85 as quartet. Decoupling the proton at. 64.85 caused the peak at «51.35 to sharpen to a singlet. The shift reagent data were also consistent with the structure. Also the spectroscopic data were similar to those of the closely related permethylated butenolide.7 21 B. Mechanism The primary products from the photolysis of 3% through a pyrex filter were 31, 38 and 32. The photoisomerization may occur by the initial cleavage of the Cy-O bond followed by Ca-0 cyclization to give 37, or followed by alkyl re- arrangement from C6 + C to give 38. Preferred ring con- Y traction over methyl migration (to give 41) is consistent with the usual order of migratory aptitudes observed in the photorearrangement of an a,B-epoxyketone to a B-dike- 15(a-c),26 tone. Scheme 3 o 0' hv B 6 ’ Y o H éé CY-O cleavage o. CG-C ° 4 Y -—--—-+ \ / H 0 H O. O H d 41 *0 3.9, 1 O O H i O H 37 38 22 * Consistent with this scheme, irradiation of labeled 3% * * (*=CD3 in place of CH3) gave éz and 8% ° 11 0t 0; 1;} 34* 81* 38* 82* The mechanisms leading from 34* to 37* or from 34* to 38* are fairly_obvious, as this type of photorearrangement has been reported in literature.7'20 Irradiation of 34* also gave 33* and the mechanism leading from 34* to 32* is unusual. There are at least three plausible mechanisms to account for the formation of 32 as shown in Scheme 4. Mechanism 3 involves the initial formation of 37 (an observed reaction product). Further irradiation of 32 might cause a-cleavage followed by bond reorganization as shown, to give 32. This mechanism was eliminated on two grounds. Beginning with 34* one should obtain 32 labeled at C-6 (42) whereas in fact the label appeared at C-3 (32*). Furthermore, separate irradiation of 31 under the same conditions used to transform 3% to 33 (in part) did not produce any 32. In mechanism b_the initial intermediate e.undergoes 23 C6-Ce bond cleavage (as in the mechanism for the formation of 32) but instead of a l,2-shift,a 1,4-shift is proposed, leading to the cyclopropanone-aldehyde intermediate 2. A six-electron electrocyclic reaction of 2 would give 32. This scheme is entirely consistent with the observed labeling results (see Scheme 4). A third alternative mechanism 13's in which 32 is pro- duced from 32 via the same cyclopropanone-aldehyde inter- mediate proposed in mechanism 2. Although the labeling result is consistent with this mechanism, it was ruled out by a control experiment. Direct irradiation of 32 under the same conditions used to convert 33 to 32 (in part) gave exclusively 32 (and 32* from 32*). Consequently mechanism 2 is also eliminated, and the only one of the proposed three which fits the data is b. The quantitative photoisomerization of 32 to 32 can be regarded as proceeding via the initial oxa-di-n-methane 37 which is thermolabile photorearrangement product 23, (Scheme 5). Homolysis of bonds 3 and b'of compound 33 can give ketene 2 as an intermediate, from which 32 can be formed by ring closure with a hydrogen migration. Vinyl migration might also occur, leading to 22; this possibility cannot be excluded, since trace amounts of minor products could not be isolated. A similar mechanism in closely related photoreactions has been reported by 27 28,29 Davis, Burkinshaw and Matsuura. For example, the 25 Scheme 5 O O H O b H a 38 ’VD C; O . hydrogen Vinyl shift H shift O: 'O\ , \ A 82 ‘V 4% spirodiketone 23 rearranged photochemically to an inter- mediate 22 which rearranged directly to 22 at room tem- perature within 2 hours. the intermediate 32 was isolated. No vinyl migration product from 26 2. Photochemistry of 4,5-Epoxy-2,3,4,6,6:pentamethyl-2- cyclohexenone (32) The epoxyenone 32 has not been previously described. It was synthesized in good yield from the corresponding dienone 3316 and m-chloroperbenzoic acid. Its infrared and ultraviolet spectra showed that the carbonyl group was still conjugated with a double bond [vc=o 1657 cm-1, Amax (methanol) 255 nm (e 8460), 325 nm (e 270)] and the nmr spectrum with europium shift data was also consistent with epoxidation having occurred solely at the y,6 double bond. 27 o (2.91)1.66 1.22(2.73) MCPBA _______, 1.00(2.24) H (1.13)1.95 0 H 2-36(1-94) (1.00)1.42 3% 33 A. Product Structures Irradiation of 32 (6°01.fl in ether) through pyrex for 2 hr gave three photOproducts 3% &2(62%) 29(110) 33(270) Compound 22 was assigned the structure shown on the basis of the following spectral data. The molecular formula l.04(1.17) H 1.21(1.00) (1.72)0.97 (1.60)1.03 1.32(1.14) 1.32(1.14) $2 28 C11H1602 was confirmed by the mass spectrum (parent peak m/e 180) and elemental analysis. The infrared spectrum had carbonyl peaks at 1700 and 1738 cm‘l. These two ir absorption bands may originate from coupling between the two carbonyl groups. This kind of coupling is well 7'17'18 The uv spectrum possessed only end documented. absorption and the ir spectrum lacked any olefinic hydro- gen absorption. In the nmr spectrum, all methyl signals appeared at or above 61.32. These spectroscopic data indicate that 32 must be a saturated compound. The struc- tural and stereochemical assignments are based on the 100 MHz nmr spectrum (CC14): 60.97 (s, 3H), 1.03 (s, 3H), 1.04 (d, 3H, J = 5 Hz), 1.17-1.25 (q, 1H, J = 5 Hz), 1.32 (S' 6H). Decoupling by irradiation at 61.21 caused the peak at 61.04 to sharpen to a singlet. The europium shift data showed that the two methyls at 61.32 were equivalent, as required by the plane of symmetry. Also, in compari- son with the shift data for 22 (vide infra) the stereo- chemistry at C-6 is clear. The mass spectrum of 32 showed a base peak at m/e 110 corresponding to the loss of a (CH3)2C=C=O moiety. The spectral data of 32 were similar to those of the closely related compounds 32, 22 and 3218,23,36 29 O H O .7 ./ 0 H I a 33 £2 Compound 22 was assigned the structure shown. 0 H l.46(2.75) (2.89)0.9 1.13(1.00) (2.56)1.os .f1.20(1.e7) O l.20(1.87) ER The molecular formula C11H1602 was confirmed by the mass spectrum (parent peak m/e 180) and the elemental analysis. The mass spectrum showed a base peak at m/e 110 correspond- ing to the loss of (CH3)2C=C=O moiety. The ir spectrum showed two coupled carbonyl absorption peaks at 1741 and 1701 cm-1. The uv spectrum again showed only end absorp- tion. These data strongly suggest that 22 is a stereo- isomer of 22. The 100 MHz nmr (CC14) showed peaks at 60.95 (s, 3H), 1.05 (s, 3H), 1.13 (d, 3H, J = 5 Hz), 1.20 (s, 6H), 1.30-1.60 (q, 1H, J = 5 Hz). Decoupling by ir- radiating the multiplet at 61.46 caused the methyl group 30 at 61.13 to sharpen to a singlet. The europium shift data are consistent with the assignment of exo geometry to the C-6 methyl group. Compound 33 was assigned the structure shown primarily by its spectral properties. The molecular formula was (5.80)1.80 H 5.40(4.28) 1.23 (2.57)1.79 (3.00 (1.30) 2% confirmed by the compound's mass spectrum (parent peak m/e 180) and elemental analysis. Strong carbonyl absorp- 1 tion at 1741 cm- and a uv Ama 225 nm (in methanol) were x consistent with the presence of a A'-butenolide moiety. These data are similar to those reported for 22 and 38.7,24,25 0 " ir(C=O): 1743 cm"1 ir(C=0): 1745 cm"1 H uv(EtOH): A ax 218 nm uv(MeOH): Amax 230 nm H (s 7250) (s 3100) H NMR: 'CH=CH'- 65.65 to 65.68 2% éé 31 The nmr spectrum of 23 (CDC13) showed a singlet at 61.23 (3H), four vinyl methyl signals from 61.75-1.80 and one vinyl proton signal at 65.40 as a multiplet. Decoupling by irradiation at the vinyl methyl signals caused the vinyl proton signal at 65.40 to sharpen to a singlet. This showed that the vinyl proton is coupled to one of the methyl groups. The europium shift data are consistent with the assigned structure. B. Mechanism It is clear that 22 and 32 do not arise directly from the irradiation of 32, but are formed from the photoisomer- ization of primary photoproduct 22. Irradiation of epoxy- ketones generally leads to B-diketones. The formation of 32 occurs by cleavage of the C -0 bond and hydrogen Y 00 H O 35 ' m” ‘ CY-O cleavage 0 0' 2‘C6 + CY C6 + CY ‘ O H H O. 32 migration over ring constraction is consistent with the usual migratory aptitudes observed in the photochemistry of a,8-’-epoxyketones.15(a'c)'26 The photoisomerization of 22 to 22 and 22 is an oxa- 37 di-w-methane rearrangement. O Though no evidence could be obtained for the formation of 22 during the course of the irradiation of 32, there is strong evidence that it is an intermediate in the process. Compound 22 was isolated from the acid—catalyzed rearrange- ment of 32. Irradiation of 22 under the same reaction conditions used to irradiate 32 gave 22 and 22 in the same ratio. 0 H+ hv H ""'_"" ——"$r‘€+§Q O H .35, 2% 33 Concerning the mechanism for the formation of El, Murray found17 in a study of closely related compounds that irradiation of epoxyketone lg in ether gave lactone hv 0 w ether 2% R lz as an only reaction product (see Introduction). A similar mechanism could operate with éé O hv ' . \ fir»: -—) ——» H cleavage H " H éé r'ék This suggests that there is a competition between CY-O Cleavage and Norrish-Type I cleavage in the irradiation at .35. 34 Labeling experiments were done with trideuterioepoxy- ketone aé to test these mechanisms (Scheme 6). Scheme 6 0 l l * 0 * H H O. bu hv l O H 0 H 0 \ O H 0/ * 0/ * 0 * ' * 49* 50* 51* 35 . . * * * The mechanisms leading from £3 to 32 and ég are * straightforward but the mechanism leading from éé to * * él is new. The possibility that éé was an intermediate in the formation of El* was eliminated by direct irradia- * * * tion of ég . Only 32 and £9 were formed in the ratio of 85:15. The direct formation of 3% from 33 represents a new, previously unobserved mechanism in the photochemistry of 3,3-unsaturated y,6-epoxy ketones. 3. The Photoisomerization of 4,5-Epoxy-6,6-dimethyl—2- cyclohexenone (%£l° The epoxyenone 38 has not been previously described. It was obtained as colorless liquid in high yield from the 30 corresponding dienone 3%. The structure of 33 was es- tablished by its spectral properties and chemical trans- formations. O 0 1.26(3.03) MCPBA ‘ 5.80(3.94) 1.09(1.19) 5,91(1.00) 3.16(2.12) 3.16(2.12) IR éé Its infrared and ultraviolet spectra showed that the carbonyl group was still conjugated with a double bond [v 1680 cm-1 c=o 36 Amax (methanol) 235 nm], and the nmr spectrum was also consistent epoxidation having occurred at 7,6 double bond. Irradiation of 36 (0.01 §_in ether) through pyrex gave two photoproducts to which the structures 66 and 6% were assigned. 0 O O hv / + Pyrexwr \ 36 56(86%) 57(l4%) mm mm mm Each of these photoproducts underwent further photoiso- merization slowly through pyrex but more rapidly when irradiated through a corex filter. Compound éé rearranged to fig and compound é; rearranged to 38 which was also photolabile and rearranged further to 68 (Scheme 7). 37 Scheme 7 '. ..___.) + Pyrex G ‘0 22. éé 31 hv 1 Corex th’ Corex 0% hv Corex 562% A. Product Structures The nmr, ir, uv and mass spectral data of §6 were 31-34 identical to those reported in the literature. In 38 o .30 6.39 \\ )/ 5.47 5.63 6.02 éé the literature, enol lactone 66 was obtained by both direct and the dye sensitized photooxygenation of 6,6- dimethylfulvene. The linearly conjugated diene moiety in §6 becomes clear from the product obtained by further irradiation of éé. Irradiation of 66 (0.03 §_in ether, corex) gave a single photoisomer to which the structure £3 was assigned.35 0 (3.33) . hv O l.l6(3.10) fig ' ether, corex: 4.98 \ 3.13(2.00) (2.23) __ 6.30 6.30 (1.00) (1.20) éfi The carbonyl band at 1770 cm"1 showed that ég was a Y-lactone. The europium shift data of compound ég showed that the gem-methyl groups were not identical. The nmr sPectrum of ,‘ég showed only two vinyl protons the methine hYdrogens are easily identified since the one adjacent to 39 the oxygen atom is deshielded relative to the other. The structure of 61 was assigned on the basis of its spectral data and chemical transformations. The molecular o 6.00(3.10) 1.23(2.03) 6.57-7.03 (1.00)- 3.17 (2.06) £1 formula C8H1002 was confirmed by the mass spectrum (parent peak m/e 138) and elemental analysis. The diketone showed two carbonyl absorptions at 1720 and 1675 cm.1 , and a uv maximum absorption (methanol) at 228 nm indicating the presence of one conjugated and one non-conjugated carbonyl group in a six-membered ring. The nmr spectrum of 67 showed a sharp singlet for the gem-methyl groups at 61.23, which indicated that é; possessed a plane of sym- metry. The nmr spectrum also showed two vinyl protons coupled to one another at C—2 and C-3 and two methylene protons at 63.17. Eur0pium shift data were consistent with the assigned structure. Irradiation of £1 (0.03 §_in ether, corex) gave the known photoisomer §2. The compound assigned structure ég showed two coupled carbonyl bands, at 1750(w) and 1710(5) cm-l. The uv spectrum showed only one absorption. 40 hv I ether, corex V 5,2 82 The mass spectrum had an M+ peak at m/e 138. The nmr spectrum was consistent with the literature report.36 Irradiation of compound 52 (0.02 §.in ether, corex; 36 or in benzene, pyrex) gave the known lactone 68. The ir spectrum of 62 showed a strong carbonyl band at 1795 cm"1 which is typical of enol lactone absorption. The uv spectrum showed only end absorption. The nmr spectrum showed two methyl group signals at 61.70 and 1.74, two methylene protons at 60.80-1.50 and two methine protons as multiplets at 62.23 and 2.65. O O ./ h" COI‘ ex .I u) 82. £9. 41 B. Mechanism The formation of 57 from 36 undoubtedly occurred by clea- vage of the CY-O bond to give intermediate 2 and followed by hydrogen migration from C6 to Cy. Scheme 8 o . o CHO as I) H 61, él The preferential hydrogen migration to give 51 rather the alkyl group migration to give 61 is consistent with the usual order of migratory aptitude observed for the photo- chemical rearrangement of a,B-epoxy ketones to B-dike- tones.15(a'c)’26 The formation of compound 52 proceeds via an oxa-di- fl-methane rearrangement of 5137. 42 ‘7 / 8m 8x) 0 8m 8o '0. The photochemical rearrangement of non-enolizable 36 B-diketones is known to give enol lactones with an exo- cyclic double bond. hv u! corex 3% £8 The mechanism for the formation of the enol lactone is as follows: ~ was —+ 22 E 88 43 Photoexcitation of 66 leads to Type I cleavage to give di- radical 6, which recyclizes to the enol lactone 66. The mechanism for the formation of compound 66 from 66 is presumed to be the same as that proposed (Scheme 4) for the formation of 66 from 66. It is outlined again in Scheme 9. Scheme 9 a» am an o / o’ __ *— x? H 88 E Compound 66 is formed from intermediate 6 through frag- mentation of a C-C bond and the formation of cyclopropanone intermediate 6. Subsequent electrocyclic rearrangement of 6 would lead directly to 66. Since the yield of 66 from 66 was high, an effort was made to detect the formation of cyclopropanone intermediate 6. The low temperature irradiation of 66 was carried out using acetone-d6 as solvent at -78°, and the reaction was 44 monitored by nmr. After 4 hr irradiation, besides the strong product peaks [66: 61.30(s), 5.47-6.39(m); 66: 61.23(s), 3.17(q), 5.80-7.03(m)], an a,B-unsaturated aldehyde peak corresponding to intermediate 6 appeared at 69.7 as multiplet. When the solution was allowed to remain at the same temperature for a longer time, the aldehyde proton signal gradually decreased in intensity. Also, if the solution was warmed up to room temperature, the signal due to the aldehyde proton slowly disappeared. Low temperature infrared spectroscopy is another use- ful technique for studying molecules which are unstable at a higher temperature. 41 Murray used 2-methyltetrahydrofuran as a matrix for the low temperature detection of dimethylketene, and Hart 42 and Love reported the observation of a ketene in a pen- tane matrix at -l96°. Chapman43 33 31. have reported the observation of ketene at'low temperatures in an ether- methanol glass and in an EPA (ether-pentane-alcohol) glass. A low temperature study of the photolysis of 16 was carried out in an attempt to observe the intermediate cyclopropanone derivative. A solution of 16 in tetra- hydrofuran was irradiated for short intervals at -105°C in a sodium chloride cavity cell. The infrared spectrum showed a carbonyl absorption at 1815 cm'l. The intensity of the carbonyl absorption of 16 decreased after the irradia- 1 tion was stopped. The absorption at 1815 cm' is attributed to the formation of cyclopropanone 6. Cyclopr0panone 45 itself and its derivatives have now been prepared and show infrared bands in the region 1813-1850 cm-1 depending on the substituents on the ring.38 The low temperature nmr and ir observations in this study are consistent with the mechanism in which cyclo- propanone intermediate 6 was proposed, during the conversion 26 to 0:- It was reported that cyclopropanonesBB-40 can be E5 “54“.“. g trapped by reaction with methanol or ethanol to form a hemiketal. However, when the irradiation of compound 66 was carried out using methanol-d4 as solvent at -78° (the reaction was monitored by nmr and GC) identical results were obtained as in other solvents. The inability to trap the cyclopropanone intermediate 6 with methanol in- dicates that the intramolecular rearrangement to 66 must be very fast with respect to the intermolecular addition of methanol. Some experiments were performed to determine whether the Conversion of 66 to 66 and 61 was a singlet or triplet 19c noted in 1968 that a,B- state reaction. Jeger gt 31. unsaturated-y,6-epoxyketone 66 isomerized almost exclusively to the diketone 66 upon triplet sensitization (using aceto- phenone in benzene). 46 (triplet energy:ca. SS Kcal7mole) 2% 6% 19a Later, Jeger and Schaffner reported that irradiation of 66 using triplet sensitization gave products 66, 66 and 66. £8 88 £1 The photoisomerization of 66 can be sensitized by acetophenone and benzophenone and the triplet conversion is not affected by the addition of piperylene (which often acts efficiently as a triplet quencher). But, the reaction could be efficiently quenched by trans-1,3,5-hexatriene 47 (ET = 47 Kcal/mole). The results suggest that the photo- isomerization of Qé occur via a triplet with a triplet energy of about 50-60 Kcal/mole. In summary, a mechanistic scheme which rationalizes all previous results on the photochemistry of cyclohexa- dienone epoxides is proposed. Following initial excitation (Scheme 10), either Cy-O or a-bond cleavage occurs, the former path being by far the more common. The latter path leads to a diradical which undergoes subsequent epoxide ring opening to give an acyl-alkoxy radical; this then undergoes ring closure to give a lactone. Four different types of subse- quent reactions have been observed for the intermediate produced by CY-O cleavage. These are migration of a group from C6 to C , ring contraction, cyclization at Ca' Y and fragmentation of an R group from C6“ It is clear in the comparison of the results from $4 with 34, or Qé with $1, respectively, that the replace- ment of H for CH3 in the C-3 or C-2 and C-3 positions of permethylated dienone epoxide £4 has only a modest effect on the photochemical results. From the results on g4, éé, gz and 3Q, it is logical to conclude that the replacement of H for CH3 in the C-2 and C-5 positions, especially in C-S position, profoundly effects the photochemical re- sults. Rearrangement Scheme 10 type I C (rare)' Y leavage (I) ( common) 48 Ca-O Cyclization Contraction Radical Fragmentation de/yield g3 ---- ---- 90% 3% ---- 25% 45% éé 73%a ---- ---- g1 78%b ---- ---- gé 14% ---- ---- a. Intermediate éfi forms products 32 (62%) and §9 (11%). b. Intermediate forms products 3g (34%) and c. Yield of (II). 32 (44%) 30% 27%c 22% 86% EXPERIMENTAL General Procedures (These apply to all parts of the thesis). Analytical gas chromatography (vpc) was carried out on a Varian Aerograph Model 1400 (flame ionization detector), and preparative Vpc was performed with a Varian Aerograph Auto Model 700 instrument (thermal conductivity detector). The nmr spectra were obtained on a Varian Associates T-60 spectrometer, usually in CCl4 using tetramethylsilane (TMS) as an internal standard. Decoupling experiments were done on an HA-lOO spectrometer. Carbocation spectra were obtained on an A56-60 spectrometer equipped with a variable temperature probe. The solvent was FSO3H/SOZC1F (23.1:4), sometimes with CDZClz; either (CH3)4NBF4 (63.13) or CH2C12 (65.30) was used as an internal standard. The temperature control was calibrated with a methanol standard sample and is accurate to :0.5°. Low temperature nmr spectra were obtained on an A56-60 spectrometer. The nmr spectra were recorded in units of delta. Numbers placed next to pro- tons in structures in the discussion section refer to chemical shifts of these protons. Numbers in brackets beside chemical shifts in the discussion section are ”europium shift numbers" obtained by adding small incre- ments of tris(l,l,l,2,2,3,3-heptaf1uoro-7,7-dimethyl-4,6- octanedione)Eu(III) to the CCl4 solution being investi- gated. After each addition the nmr spectrum was scanned 49 50 and the new frequency of each absorption was recorded. The shift for each absorption is the difference between the frequency of the shifted absorption and the original one. Shift numbers are the ratios obtained by dividing the shift of each signal in the spectrum by the shift of the least shifted signal. Infrared spectra were obtained on a Unicam SP-200 spectrometer except for the low temperature study in which a Perkin-Elmer 237 grating spectrophotometer was used. Ultraviolet-visible spectra were obtained with a Uni- cam-800 spectrometer. Mass spectra were obtained from a Hitachi-Perkin Elmer RMU-6 operated by Mrs. Lorraine Guile. Melting points were determined with a Thomas-Hoover melting point apparatus and are uncorrected. Elemental analyses were performed by Spang Microanalytical Labora- tories, Ann Arbor, Michigan, and by Clark Microanalytical Laboratories, Urbana, Illinois. General Photolysis Procedures Solutions of the compounds to be irradiated were placed in septum-capped Pyrex tubes or nmr tubes and purged of oxygen by bubbling dry, oxygen-free nitrogen through them for 30 minutes prior to photolysis. Irradiations were carried out with a 450 watt Hanovia Type L medium pressure mercury vapor lamp with the appropriate filter. The tubes 51 were fastened to an immersion well apparatus which was immersed in water at ambient temperature. Alternatively, a Rayonet Photochemical Chamber Reactor or Type RS Prepara- tive Photochemical Reactor was used. Photolyses were monitored by withdrawing small (< 1 pl) aliquots and inject— ing them into a Varian Aerograph Series 1400 analytical gas chromatograph. 1. Synthesis of 4,5-§poxy52,4,5,6,Gjpentamethyl-2-cyclo- hexenone (3%) To a solution of 2.10 g (12.8 mmole) of 2,4,5,6,6- pentamethyl-Z,4-cyclodienone (31)16 in 40 ml of methylene chloride was added, at 0°C, a solution of 2.20 g (12.8 mmol) of m-chlorOperbenzoic acid in 60 m1 of methylene chloride. The mixture was stirred at room temperature for 2 hours (nmr monitoring showed complete reaction). During this time, a white precipitate formed and the precipitate was removed by filtration. The solvent was removed by rotary evaporation, petroleum ether (bp 30-60°C) was added, the filtrate was washed three times with 15% aqueous sodium sulfite, water and saturated sodium chloride solu- tion, dried (MgSO4) and evaporated to give 2.24 g (97.4%) of light yellow oil. The crude product was chromatographed on florisil (mesh 60-200) using ether: hexane (1 : 10) as eluent, to give colorless epoxide 34. Compound 34 was subjected to preparative Vpc (10' x 0.25 in column, 20% 52 SE-30 on chromosorb W, AW—DMCS 80/100 mesh, 180°, 60 ml/ min He, retention time 18 min), 60% of compound 34 remained and 40% of 3% was converted to alcohol due to thermal oxi- rane ring opening. Compound 3% had ir (neat) 3000 (s), 1680 (s), 1550 (w), 1480 (s), 1400 (s), 1380 (m), 1278 (m). 1200 (w), 1130 (m), 1100 (s), 1080 (m), 1060 (m), 1003 (m), 978 (m), 900 (s), 848 (m), 780 (w), 740 (m) cm'l; uv (methanol) xmax 208 nm (e 3200), 250 (9560) 320 (500): nmr (CC14) see structure, 61.06 (s, 3H), 1.33 (s, 3 H), 1.40 (s, 3 H), 1.52 (s, 3 H), 1.73 (d, 3 H, J = 2 Hz), 6.46 (q, l H, J = 2 Hz); mass spectrum (70 eV) m/e (rel intensity) 180 (9), 165 (18), 137 (100), 112 (40), 110 (20), 109 (18), 97 (10), 69 (35), 67 (36). A231. Calcd. for C11H1602: C, 73.30; H, 8.95 Found: C, 73.33; H, 9.02 2. Synthesis of 5-Trideuteromethyl-4,5-epoxy-2,4,6,6- tetramethyl-2-cyclohexenone (€631. One gram of 2,4,5,6,6-pentamethylcyclohexa-2,4-dienone (31) was added to a solution of 0.30 g of potassium t- 48 6. deep red immediately and remained so. The mixture was butoxide in dimethyl sulfoxide-d The solution became stirred at room temperature for 5 hr (nmr monitoring showed complete reaction). The red-brown solution was poured into 300 ml of methylene chloride and washed with ice water (three 50-ml portions). After being dried, the 53 solution was evaported to an oil, which was purified by distillation, then further purified by vpc (S’x 0.25 in. column, 20% FFAP on chromosorb W, AW-DMCS 80/100, 148°, 60 ml/min He, retention time 2 min). The nmr spectrum was consistent with the structure of S-trideuteromethyl-Z,4,6,6- tetramethyl-Z,4-cyclohexadienone. It consisted of three signals at 61.12, 61.81 and 66.60 with relative areas 6:6:1, assigned respectively to the gem-dimethyls, the allylic methyls at C-2, C-4, and the C-3 vinyl proton. To a solution containing 120 mg (0.73 mmole) of 2,4,6,6- tetramethyl-S-methyl-d3-2,4-cyclohexadienone in 2 m1 of methylene chloride was added a solution of 148 mg (0.86 mmol) of m-chloroperbenzoic acid in 2 m1 of methylene chloride. The mixture was stirred at room temperature for 2 hours, and workup was as described for the prepara- tion of 34. The epoxidation product 34* had an nmr spectrum identical with that of 34 except that the signal at 51.40 was absent. 3. Irradiation of 4,5—epoxy-2,4,5,6,6fipentamethyl-2- cyclohexenone (ggl‘ A degassed solution of 100 mg (0.55 mmole) of 34 in 50 m1 of anhydrous ether was irradiated through pyrex with a 450 W Hanovia lamp. The photolysis was followed by vpc, and was complete in about 1 hr. Analytical Vpc (5' x 0.125 in column, 20% FFAP on chromosorb W, AW-DMCS 4,1 1:.., .4 54 80/100, 178°C, 30 ml/min N2) showed three components: 31 (retention time 1 min), 38 (7 min) and 39 (9 min). Preparative vpc (10' x 0.25 in column, 20% SE-30 on chromo- sorb W, AW-DMCS 80/100, 120°, 60 ml/min He) gave pure 31 (25%; retention time 8 min), 38 (45%; 25 min), and 32 (30%; 40 min). For l,3,3,4,5-pentamethy1-7-oxabicyclo[2.2.1]hept-5- ene-Z-one (31): ir (neat) 3000 (s), 1740 (s), 1640 (w), 1470 (m), 1450 (m), 1392 (m), 1300 (w), 1220 (m), 1200 (m), 1180 (m), 1100 (s), 1020 (w), 1000 (s). 900 (w), 860 (w). 810 (s) cm'l; uv (MeOH) Amax 207 nm (e 950) with a shoulder) 280 (e 100); nmr (CC14) 61.12 (s, 3 H), 1.14 (s, 3 H), 1.30 (s, 3 H), 1.32 (s, 3 H). 1.72 (d, 3 H, J = 2 Hz). 5.56 (q, 1 H, J = 2 Hz); mass spectrum (70 eV) m/e (rel intensity) 180 (14), 165 (5), 152 (5), 140 (100). 139 (42). 138 (35), 137 (75), 124 (20), 122 (76), 121 (83), 111 (11), 110 (84), 108 (15), 95 (40), 94 (62), 92 (15), 82 (64), 77 (77). 69 (12). 67 (25). 59 (60). 55 (20). 54 (68), 53 (31). 52 (10). 51 (15). Anal. Calcd. for C H O 11 16 2‘ Found: C, 73.34; H, 8.96 C, 73.30; H, 8.95 For 4-acety1-2,4,5,5-tetramethy1-2-cyclopentenone (38): ir (CC14) 3000 (s), 1710 (s), 1660 (W), 1480 (w), 1460 (W). 1400 (W). 1373 (m). 1305 (W). 1300 (W). 1192 (w), 1160 (m), 1050 (m), 990 (m), 890 (m) cm'l; uv (cyclo- hexane) Ama 265 (s 7000); nmr (CC14), 60.90 (s, 3 H), x 1.03 (s, 3 H), 1.25 (s, 3 H), 1.75 (d, 3 H, J = 1.0 Hz), 55 1.95 (s, 3 H), 6.92 (q, 1 H, J = 1 Hz); mass spectrum (70 eV) m/e (rel intensity) 180 (6), 162 (5). 144 (5), 139 (26, 138 (100, 137 (35). 123 (62), 109 (43), 93 (7), 91 (7). 81 (8). 71 (10). 77 (9). 69 (8). 67 (60), 55 (15). Anal. Calcd. for C11H1602: C, 73.30; H, 8.95 Found:‘ C, 73.14; H, 8.91 For 3,3,4,6,7-pentamethy1-2(3H)-oxepinone (39): ir (neat) 3000 (s). 1750 (s). 1660 (m), 1460 (m). 1400 (m). 1340 (w), 1280 (w), 1260 (w), 1200 (m), 1150 (m), 1120 (w), 1100 (w), 1040 (w), 950 (w), 880 (w) cm‘l; uv (MeOH) Amax 212 nm (e 2700), 250 (10,570); nmr (100 MHz) (CC14) 61.23 (s, 6 H), 1.67 (broad singlet, 3 H), 1.83 (doublet, 3 H, J = 2 Hz), 1.89 (broad singlet, 3 H), 5.60 (m, l H), decoupling at 65.60 caused the doublet at 61.83 to become a singlet; mass spectrum (70 eV) m/e (rel intensity) 180 (61), 165 (7), 138 (30), 137 (98), 109 (100). 108 (71). 93 (75). 91 (32, 97 (34), 67 (50). 65 (20). 55 (15), 53 (20). Anal. Ca1cd.for C11H1602: C, 73.30; H, 8.95 Found: C, 73.21; H, 9.09 4. Irradiation of 34: The conditions and workup procedure were as for the unlabeled material. From 34* the resulting 31* had an nmr spectrum identical with that of 31 except the signal at 61.30 was absent. The spectrum of the resulting 38* was 56 identical with that of 38 except that the singlet at 61.95 was absent. The resulting 39* was identical with that of 39 except that the signal at 61.89 was absent and the peak at 61.67 sharpened to a singlet. 5. Irradiation of 4-Acetyl-2,4,5,5-tetramethyl-2-cyclo- pentenone (38) A degassed solution of 50 mg (0.28 mmole) of 38 in 25 ml of anhydrous ether was irradiated through pyrex with a 450 W Hanovia lamp at room temperature. The photolysis was followed by vpc, and was complete in about 40 minutes. Analytical vpc (5' x 0.125 in column, 10% FFAP on chromo- sorb W, AW—DMCS 80/100, 150°C, 30 ml/min N2) showed two components with retention times of 3.5 and 19 min respec- tively, in a ratio of 1:6. Preparative vpc (10' x 0.25 in column, 20% FFAP on chromosorb W, AW-DMCS 80/100, 120°, 60 ml/min He) allowed collection of the major product 40 with a retention time of 52.5 min. For 40: ir (neat) 3000 (m). 1760 (s). 1662 (m). 1460 (m). 1396 (m). 1360 (w), 1330 (m), 1200 (W). 1120 (m), 1080 (m), 1040 (w), 940 (w) cm'l; uv (MeOH) lmax 222 (e 7640), 263 (6250); nmr (CC14) 61.35 (d, 3 H, J = 6 Hz), 1.66 (s, 3 H), 1.68 (s, 3 H), 1.86 (homoallylic coupling, 6 H), 4.85 (q, 1 H, J = 6 Hz); mass spectrum (70 eV) m/e (rel intensity) 180 (60), 165 (4), 137 (100), 138 (30), 123 (10), 110 (17). 109 (85), 108 (50). 93 (50). 91 (20), 77 (20), 67 (40) S7 55 (10). 53 (14). Anal. Calcd. for C H160 C, 73.30; H, 8.95 2: Found: C, 73.33; H, 8.86 11 6. Irradiation of 1,3,3,4,5-Pentamethyl-7-oxabicyclo- L2.2.1]hept-5-en-2-one (32)_ A degassed solution of 33 (25 mg, 0.14 mmole) in 10 m1_ of anhydrous ether was irradiated through pyrex with a 450 W Hanovia Type L lamp the reaction was followed by nmr. After 1 hour, the nmr spectrum showed no change and com- pound 33 was recovered. 7. Synthesis of 4,5-Epoxy-2,3,4,6,6-pentamethy1-2-cyclo- hexenone (33) To a solution of 1.20 g (7.32 mmole) of 2,3,4,6,6- pentamethyl-Z,4-cyclohexadienone 3316 in 20 m1 of methylene chloride was added, at 0°C, a solution of 1.42 g (8.61 mmole) of m-chloroperbenzoic acid in 20 ml of methylene chloride. The mixture was stirred at room temperature for 3 hours (nmr monitoring showed complete reaction at this time). m-Chlorobenzoic acid was removed by filtra- tion, and the solvent was removed by rotary evaporation. Petroleum ether (bp 30-60°C) was added, the filtrate was washed with aqueous sodium bicarbonate and saturated sodium chloride solution, dried (M9804) and evaporated to give 1.20 g (91%) of 33 as a light oil. The crude product 58 was chromatographed on florisil (mesh 60-200) using ether:hexane (1:5) as eluent, to give pure epoxide 33. Ir (neat) 3000 (s), 1674 (s), 1616 (m), 1480 (m), 1390 (m). 1320 (m). 1260 (m). 1090 (m), 1050 (m). 918 (m); uv (MeOH) Amax 210 nm (6 2970). 255 (8460), 325 (270); nmr (CC14) see structure, the peaks at 61.66 and 1.95 were broadened, but all other methyl groups were sharp singlets at 61.00, 1.22 and 1.42; one proton appeared as singlet at 62.86; mass spectrum (70 eV) m/e (rel intensity) 180 (50), 165 (35), 164 (15), 151 (26), 137 (100), 135 (52). 123 (34), 121 (31), 119 (25), 112 (35), 110 (55). 95 (20). 91 (20). 83 (15), 81 (30). 69 (24). 67 (50). 55 (26), 53 (30). Anal. Calcd. for C11H1602: C, 73.30; H, 8.95 Found: C, 73.24; H, 8.92 8. Synthesis of 3-Trideuteromethyl-4,5-epoxy-2L4,6,6- tetramethyl-2-cyclohexenone (33:) To a solution containing 500 mg (2.77 mmole) of 33 in 10 m1 of dimethyl sulfoxide-d6 was added with stirring and under N2, 310 mg (2.77 mmole) of potassium t-butoxide.23 The mixture was stirred at room temperature for 1 hour, 'then quenched with ice-water and extracted with ether. Organic layers were dried (M9804) and the solution was eVEiporated to give a nearly quantitative yield of 33*. Thea nmr spectrum was identical to that of the starting 59 material, except that the signal at 61.95 had disappeared. 9. Irradiation of 4,5-Epoxy-2,3,4,6,6-pentamethyl-2- cyclohexenone (33L A degassed solution of 100 mg (0.55 mmole) of 33 in 30 ml of anhydrous ether was irradiated through pyrex with a 450 W Hanovia lamp at room temperature for 2 hours. The photolysis was followed by vpc. Analytical vpc (5' x 0.125 in column, 10% FFAP on chromosorb W, AW-DMCS 80/100, 160°C, 30 ml/min N2) showed three components: 33 (62% retention time 2.5 min), 33 (11%,5 min) and 33 (27%, 25 min). Preparative Vpc (10' x 0.25 in column, 20% FFAP on chromosorb W, AW-DMCS 80/100, 140°, 60 ml/min He) gave pure 33 (retention time 18 min), 33 (31 min) and 33 (over 1 hr). Further purification of 33 with vpc (5' x 0.25 in column, 20% FFAP on chromosorb W, AW-DMCS 80/100, 180°, 60 ml/min He) gave pure 33 (retention time 12 min). For 33 (Eggirl,3,3,5,6-pentamethy1bicyclo[3.l.0]hexan- 2,4-dione): ir (KBr) 3000 (m), 1700 (s), 1462 (m). 1380 (m), 1360 (w), 1300 (m), 1205 (m), 1110 (w). 1090 (w). 1040 (m), 845 (m) cm'l; uv (MeOH) Amax 215 nm (e 2610); nmr (CC14) 60.97 (s, 3 H), 1.03 (s, 3 H), 1.04 (d, 3 H, J = 5 Hz), 1.21 (q, l H, J = 5 Hz), 1.32 (s, 6 H); decoupling at 61.06-1.27 caused the doublet at 61.04 to sharpen to a singlet; mass spectrum (70 eV) m/e (rel intensity) 180 (42), 165 (20), 163 (10), 137 (60), 120 (11), 110 (100). 109 (22). 60 105 (15), 95 (25), 82 (22), 79 (13), 67 (80). Anal. Calcd. for C11H1602: C, 73.30; H, 8.95 Found: C, 73.29; H, 8.97 For 33 (syn-1,3,3,5,6-pentamethylbicyclo[3.1.0]hexan- 2,4-dione): ir (neat) 3000 (m), 1742 (w), 1701 (s), 1470 (m), 1398 (m), 1295 (m), 1101 (w), 1080 (m), 1040 (w), 845 (w) cm‘l; uv (MeOH) xma 215 nm (e 1870); nmr (cc14) x 60.95 (s, 3 H), 1.05 (s, 3 H), 1.13 (d, 3 H, J = 5 Hz), 1320(sq 6 H), 1.46 (q, 1 H, J = 5 Hz); decoupling at 61.46 caused the methyl signal at 61.13 to sharpen to a singlet; mass spectrum (70 eV) m/e (rel intensity) 180 (53), 165 (20), 162 (9), 138 (20), 137 (60), 121 (13), 120 (25). 110 (100), 109 (27), 105 (10), 95 (25), 82 (25), 81 (15), 67 (88), 55 (7), 54 (7). Anal. Calcd. for C H 602: C, 73.30; H, 8.95 11 1 Found: C, 73.11; H, 8.98 For 33: ir (neat) 2985 (m), 1741 (s), 1680 (m), 1448 (m), 1382 (m), 1325 (m), 1281 (m), 1180 (w), 1140 (w), 1100 (m), 1005 (m), 760 (m) cm‘l; uv (MeOH) Amax 225 nm (e 7150); nmr (CDC13) 61.23 (s, 3 H), 1.74 (broad s, 3 H), 1.76 (s, 3 H), 1.79 (s, 3 H), 1.80 (s, 3 H), 5.40 (m, l H), decoupling at 61.74-1.80 caused the vinyl proton at 65.40 to sharpen to a singlet; mass spectrum (70 eV) m/e (rel intensity) 180 (60), 165 (92), 137 (30), 135 (80), 125 (20), 119 (25), 112 (21), 110 (35), 105 (20), 97 (100), 91 (19), 61 69 (65), 68 (25), 55 (40), 54 (40), 53 (30). Anal. Calcd. for C11H1602: C, 73.30; H, 8.95 Found: C, 73.11; H, 9.02 10. Irradiation of 3-Trideuteromethy1-4,5-epoxy:2,4,6,6- tetramethyl-2-cyclohexenone (33:) The conditions and workup procedure were as described for the unlabeled material. From 33* the resulting 33* had an nmr spectrum identical with that of 33 except that the signal at 61.32 (s, 6 H) was reduced to half its area (s, 3 H). The spectrum of the resulting 33* was identical with that of 33 except that the signal at 51.20 (s, 6 H) was reduced to half its area (5, 3 H). The spectrum of the resulting 33* was identical with that of 33 except that signal at 61.79 was absent. ll. Irradiation of 2,3,4,6,6-Pentamethy1-2-cyclohexen- 1,5-dione (33) A degassed solution of 33 (50 mg, 0.27 mmole) in 25 m1 of anhydrous ether was irradiated through pyrex with a 450 W Hanovia lamp. The photolysis, followed by Vpc and nmr was complete in about 1 hour. Analytical vpc (5' x 0.125 in column, 10% FFAP on chromosorb W, AW-DMCS 80/100, 160°, 30 ml/min N2) showed two components: 33 (85%, reten- tion time 2.5 min) and 33 (15%, 5 min). Preparative Vpc (10' x 0.25 in column, 20% FFAP on chromosorb W, AW-DMCS 62 80/100, 140° 33 ml/min He) gave pure 33 and 33 in the same ratio. 12. Irradiation of 33: The conditions and workup procedure were as described for the unlabeled material. From 33* (labeled with CD3 at C-3) the resulting 33fand 33'were identical with the products obtained from the irradiation of labeled 33*. The product ratio was nearly the same. 13. Synthesis of 4,5-Epoxy-6,6-dimethy1-2-cyclohexenone £29.). To a solution of 2.5 g (0.02 mole) of 6,6-dimethy1-2,4- cyclohexadienone3OC33)in 20 ml of methylene chloride was added, at 0°C, a solution of 3.54 g (0.02 mole) of m-chloro- perbenzoic acid in 20 ml of methylene chloride. The mix- ture was stirred at room temperature for 8 hr, precipitated m—chlorobenzoic acid was removed by filtration, and the solvent was removed by rotary evaporation. Petroleum ether (bp 30-60°) was added, the filtrate was washed with aqueous sodium bicarbonate and saturated sodium chloride solution, dried (MgSO4), and evaporated to give 2.37 g of a light yellow oil (86%). The crude product was chroma- tographed on florisil (mesh 60-200) using ether:hexane (1:10) as eluent, to give 33. Analytical vpc (5' x 0.125 in column, 20% SE-30 on chromosorb W, AW-DMCS 80/100, 63 130°, 30 ml/min N2) showed retention time 3 min; prepara- tive vpc (10' x 0.25 in column, 20% SE-30 on chromosorb W, AW-DMCS 80-100, 100°, 60 ml/min He, retention time 22 min) gave 4,5-epoxy-6,6-dimethyl-2-cyclohexenone 33: ir (neat) 3000 (s),1680 (s), 1480 (m), 1385 (m), 1380 (w), 1295 (m), 1250 (m), 1225 (w), 1180 (m), 1118 (s), 1065 (m), 940 (m), 860 (m), 830 (s) cm‘l; uv (MeOH) Amax 235 nm (5 35,370) 280 (4080); nmr (CC14) 61.09 (s, 3 H), 1.26 (s, 3 H), 3.20-3.39 (m, 2 H), 5.66-5.92 (d, I H), 6.75- 7.06 (m, 1 H); mass spectrum (70 eV) m/e (rel intensity) 138 (7), 123 (10), 122 (29), 109 (45), 95 (44), 82 (100), 79 (60), 77 (30), 70 (20), 67 (20), 55 (40), 54 (25). 5333. Calcd. for C8H1002: C, 69.54; H, 7.30. Found: C, 69.54; H, 7.26. 14. Irradiation of 4,5-Epoxy:6,6-dimethyl-2-cyclohex- enone (33L, Typical procedure: A degassed solution of vpc collected compound 33 (100 mg, 0.73 mmole) in 50 m1 of anhydrous ether was irradiated through pyrex with a 450 W Hanovia lamp at room temperature. The photolysis, followed by Vpc and nmr, was complete in 7 hr. Analytical vpc (5' x 0.125 in column, 10% Carbowax on chromosorb W, AW-DMCS 80/100, 109° 30 ml/min N2) showed two components 33 (85%, retention time 17 min) and 33 (14%, 24 min). Preparative Vpc (6' x 0.25 in column, 10% FFAP on chromosorb W, AW- DMCS 80/100, 105°, 60 ml/min He) gave the pure compounds 64 éé and 21- Compound 56 (3,3-dimethy1-2(3H)-oxepinone): ir (neat) 1740 cm"1 (Vc=o)' 1640 and 1603 cm'1 (vc=c); uv (ethanol) Amax 243 nm (e 6460); nmr (CC14) (60 MHz) 61.30 (s, 6 H) and multiplets between 65.47 and 6.39. The nmr spectrum (in CC14) (100 MHz) showed four sets of vinyl protons at 65.47, 5.63, 6.02 and 6.39 (J = 6.7 Hz, = 6.2 Hz, 1,2 J2,3 J3,4 = 10.2 Hz, respectively). The mass spectrum (70 eV) m/e (rel intensity) 138 (12) 109 (1.5), 95 (100), 81 (5.2), 79 (6), 77 (4.2), 68 (1.9), 67 (19), 66 (1.4), 65 (4.5), 55 (3), 52 (1.4), 50 (2), 43 (1.5), 42 (1.6), 40 (45), 39 (30), 38 (2.5). All spectral data were identical to the literature report.”-35 Compound 31 (6,6-dimethy1-2-cyclohexen-1,5-dione): ir (neat) 2990 (m), 1720 (s), 1675 (s), 1640 (w), 1530 (m), 1385 (m), 1340 (w), 1300 (m), 1170 (w), 830 (m) cm‘l; uv (MeOH) xma 228 nm (6 7,000); nmr (CC14) 61.23 (S, 6 H), x 3.17 (br, 2 H), 5.80-6.20 (m, 1 H) and 6.57-7.03 (m, 1 H): mass spectrum (70 eV) m/e (rel intensity) 138 (42), 123 (1), 110 (12), 95 (22), 77 (4), 70 (100), 68 (55). Anal. Calcd. for C8H1002: C, 69.54; H, 7.30 Found: C, 69.60: H, 7.39 Identical results were obtained as when carbon tetra- chloride, benzene, methanol, t—butyl alcohol or acetone were used as the solvent for irradiation of compound 36 through a pyrex filter. 65 Irradiation of compound 36 using a uranyl glass filter gave an almost quantitative yield of 56; no 57 was iso- lated. 15. Irradiation of compound 36 using a corex filter or 254 nm light (Rayonet Merry:go-round Model MGR-100, 115 volt, 50/60 Hz AC 5 RPM) The irradiation of an 0.01 M solution of 36 in ether. under these conditions, monitored by vpc, gave compounds 56 and 57 in a ratio of 7:1 after 2 hr of irradiation. Further irradiation for 7 hr, followed by Vpc (5' x 0.125 in column, 10% FFAP on chromosorb W, AW-DMCS 80/100, 125°, 30 ml/min N2) gave four products (% retention time): 56 (38, 6 min), 58 (46, 9.5 min), 53 (12, 17 min), and fig (4, 30 min). For known compound 58 (2,2-dimethyl-4-oxa-bicyclo [3.2.0]-hept-6-en-3-one)35: ir (neat) 2980 (m), 2940 (w), 1770 (s), 1540 (m), 1390 (w), 1360 (m), 1350 (w), 1300 (w), 1270 (m), 1170 (m), 1140 (s), 1095 (s), 1015 (s), 970 (m), 930 (w), 910 (w), 860 (m), 800 (s) cm'l; uv (MeOH) Amax 220 nm (e 235); nmr (CC14) 61.16 (s, 6 H, eurOpium shift showed that the two methyl groups are not identical), 3.13 (m, 1 H), 4.98 (m, 1 H), 6.30 (m, 2 H); mass spectrum (70 eV) m/e (rel intensity) 138 (1.5), 123 (2), 110 (15): 109 (36), 95 (100), 93 (10), 91 (11), 83 (12), 81 (33), 79 (66), 77 (35), 67 (28), 53 (22), 51 (11). 66 For known compound 52 (3,3-dimethyl-bicyclo[3.1.0]- hexane-2,4-dione): ir (neat) 3001 (w), 1750 (w), 1710 (s), 1480 (m), 1400 (w), 1307 (w), 1290 (m), 1241 (w), 1190 (m), 1160 (w), 1010 (m), 900 (m) cm'l: uv (ethanol) Amax 215 nm (e 550), 280 (190): nmr (CC14) 61.04 (s, 6 H), 1.07 (m, 1 H), 1.55 (m, 1 H), 2.50 (m, 2 H): mass spectrum (70 eV) m/e (rel intensity) 138 (47), 123 (9), 110 (21), 109 (16), 97 (17), 95 (70), 93 (6), 91 (7), 83 (6), 82 (10), 81 (15), 79 (24), 77 (8), 70 (100), 69 (19), 68 (69), 67 (55). A11 spectral data were identical to the litera- ture report.36 For known compound £8 (2-cis-(1'-hydroxy-2'-methy1-l'- propenyl)-cyclopr0panecarboxy1ic acid-y-lactone): ir (neat) 3100 (m), 1795 (s), 1720 (s), 1680 (w), 1480 (w), 1380 (w), 1238 (w), 1210 (m), 1150 (m), 1050 (m), 1100 (m) cm-l; uv (methanol) showed end absorption; nmr (CC14) 50.85-1.50 (m, 2 H), 1.70 (s, 3 H), 1.74 (s, 3 H), 2.23 (m, 1 H), 2.65 (m, 1 H): mass spectrum (70 eV) m/e (rel intensity) 138 (50), 123 (18), 110 (15), 109 (20), 96 (18), 95 (80), 79 (35), 77 (25), 70 (100), 68 (80), 67 (62). A11 spectral data were identical to the litera- ture report.36 ,lg. Irradiation of 36 atllow temperature A. A degassed solution of compound 36 (25 mg) in 0.5 xnl of acetone-d6 was irradiated through pyrex with a 450 67 W Hanovia lamp at -78°. The photolysis was followed by nmr (Varian associate A56-60) at that temperature and by vpc. After 4 hr irradiation, besides the strong product signals of 56 and 57 [56: 61.30 (s), 5.47-6.39 (m) and 5g: 61.23 (s), 3.17 (q), 6.0-7.03 (m)], an aldehyde proton peak at 69.7 was observed. When the solution stood at the same temperature for 3 hr, the intensity of the al- dehyde proton signal gradually decreased. Following a similar separate irradiation, the solution was warmed slowly (8 hr) to room temperature, and the signal due to the aldehyde proton disappeared. Replacement of the acetone by methanol-d4 gave identical results. B. The Technology of Low Temperature Infrared Spec- troscopy The apparatus consisted of a lead block (designed to hold liquid nitrogen) into which a sodium chloride cavity cell (0.2 mm path length) could be set gig a hole suitably drilled in the block. The metal block was insulated by a close fitting styrofoam box. When the reservoir was filled with liquid nitrogen, the cell was cooled to an equilibrium temperature of ~100°210°C. Icing of the cavity cell windows was minimized by passing a vigorous stream of very dry nitrogen over the cell windows throughout the experiment. A solution of 20 mg of 36 in 0.1 m1 of tetrahydrofuran 68 was sealed in the cavity inside the styrofoam block and cooled to -105° with liquid nitrogen. A parallel beam of radiation from a 1000-watt Hanovia mercury lamp was filtered through pyrex and diverted into the cavity cell. Examination of the ir spectrum after irradiation fOr 10 min indicated a decrease in the intensity of the carbonyl absorption at 1680 cm.1 due to 55 and the appearance of 1 a sharp intense absorption at 1740 and 1720 cm- , which were attributed to the formation of photoproducts. Most 1 important, a weak peak appeared at 1815 cm- which was attributed to a cyclopropanone. When the solution was 1 warmed to room temperature, the band at 1815 cm- gradually disappeared. l7. Irradiation of 3,3-Dimethyl-2(3H)-oxepinone (55)35 A degassed solution containing 60 mg of 55 in 15 m1 of anhydrous ether was irradiated through Corex with a 450 W Hanovia lamp. The photolysis was followed by analytical Vpc (5' x 0.125 in column, 10% Carbowax on chromosorb W, AW-DMCS 80/100, 100° 30 ml/min N2). As the reaction proceeded, the peak with a retention time 12 min (corresponding to 55) decreased in intensity; and the peak (ret. time 15 min) due to 55 appeared. -After 20 hr irradiation, preparative Vpc (6' x 0.25 in column, 10% FFAP on chromosorb W, AW-DMCS 80/100, 105°, 60 ml/ min He) allowed collection of the single photoproduct 69 55. The nmr (CC14) of 55 showed 61.16 (s), 3.13 (m), 4.98 (m), 6.30 (m). 18. Irradiation of 6,6-Dimethyl-2-cyclohexen-l,5-dione (55) A degassed solution containing 40 mg of compound 55 in 10 m1 of anhydrous ether was irradiated through Corex with a 450—W Hanovia lamp, the photolysis was followed by analytical vpc (5' x 0.125 in column, 10% Carbowax on chromosorb W, AM-DMCS 80/100, 100°, 30 ml/min N2). As the reaction proceeded, the peak with a retention time of 29 min (corresponding to 55) decreased in intensity and the peak due to the product 55 appeared at 39 min. After the complete reaction (2 hr), preparative vpc (5' x 0.25 in column, 10% FFAP on chromosorb W, AW-DMCS 80/100, 125°, 60 ml/min He) allowed collection of the single photo- product 55 in 95%. The nmr spectrum of 55 showed 61.04 (s, 6 H), 1.07 (m, 1 H), 1.55 (s, l H), 2.50 (m, 2 H) and the ir spectrum showed a strong carbonyl absorption at 1710 cm‘l. 19. Irradiation of 3,3-Dimethyl-bigyclo[3.l.0]-hexane- A degassed solution containing 35 mg of compound 55 in 10 ml of anhydrous ether was irradiated through Corex or 30 mg of compound 55 in 10 m1 benzene was irradiated through pyrex with a 450-W Hanovia lamp. The photolysis 70 was followed by nmr (benzene-d6) and analytical Vpc (5' x 0.125 in column, 10% carbowax on chromosorb W, AW-DMCS 80/100, 153°, 30 ml/min N2). The reaction was complete within 2 hr with the pyrex filter. As the reaction pro- ceeded, the peak with a retention time 18 min (correspond- ing to 55) decreased in intensity and the product peak 55 with a retention time 23 min appeared. The preparative Vpc (6' x 0.25 in column, 10% FFAP on chromosorb W, AW- DMCS 80/100, 105°, 60 ml/min He) allowed collection of the single photoproduct in 90%. The ir spectrum of 55 1 and showed enol lactone absorption at 1720 and 1795 cm- the nmr (CC14) showed 60.85-l.50 (m, 2 H), 1.70 (s, 3 H), 1.74 (s, 3 H), 2.23 (m, l H), 2.65 (m, 1 H). 20. Sensitization and Quenching Studies A. Solvent Purification a. Benzene: Analytical grade benzene was puri- fied by stirring it over concentrated sulfuric acid several times for several days until the acid no longer turned yellow. It was then washed with 10% sodium hydroxide solu- tion, water, and sodium chloride solution, respectively, and dried over anhydrous calcium hydride, after drying, the benzene was distilled from potassium, and the middle fraction was retained. b. Methanol. Reagent grade methanol was distilled from magnesium turnings and only the middle fraction was 71 retained. c. Acetophenone (ACP). AcetOphenone was distilled under reduced pressure and Vpc analysis showed no appre- ciable impurities. d. Benzophenone. Reagent grade benzophenone was recrystallized twice from ethanol. e. Trans-1.3.5-hexatriene. Aldrich Chemical Company trans-1,3,5—hexatriene was used as purchased. f. Piperylene. Aldrich Chemical Company piperylene was used as purchased. h. Hexamethylbenzene (HMB) was purified by re- crystallization from ethanol. B. Photolysis The following solutions were prepared: 322_l: 0.02 g solution of epoxyenone 55 was prepared by adding 110 mg of epoxyenone 55 and 64 mg of hexamethyl- benzene (HMB) in 40 ml of benzene. l. Sensitization solution (3 g): 3.6 g of acetophenone was dissolved in 10 m1 of 0.02 §_epoxyenone (55) solution. 2. Quenching solution (2.6 g): 1 g of trans-1,3,5- hexatriene was dissolved in 5 m1 of 0.02 g epoxyenone (55) solution. 3. Blank solution: 0.02 !.°f epoxyenone (55) solution. 52g_3: 0.02 g solution of epoxyenone 55 was prepared by adding 110 mg of 55 and 64 mg of HMB in 40 ml of methanol. 1. Sensitization solution (2 g): 3.5 g of benzophenone 72 was dissolved in 10 m1 of 0.02 giof epoxyenone (55) solu- tion. 2. Quenching solution (5 g): 3.4 g of piperylene was dissolved in 10 ml of 0.02 §,of epoxyenone (55) solution. 3. Blank solution: 0.02 g of epoxyenone (55) solu- tion. Aliquots (2.8 m1) of each above solution were placed in 13 x 100 mm pyrex tubes using a 5 ml syringe with a six inch needle: The tubes had previously been constricted about 2 cm from the top, so that they could be easily sealed after being degassed by five freeze-thaw cycles (p<0.005 torr), by using liquid nitrogen. The samples were irradiated in parallel using a water bath immersed merry-go-round apparatus to insure that all the samples received the same amount of incident light and that the temperature remained constant. A 450 W Hanovia lamp was used as light source, and Pyrex was used as filter. After 7 hr irradiation, the products were identified by careful comparison of the vpc retention time with those of authentic samples in each case. For Run 1: In aceto- phenone solution, the analytical Vpc (5' x 0.125 in column, 10% carbowax on chromosorb W, Aw-DMCS 80/100, 155°, 30 ml/ min N2) showed a peak with a retention time 8 min (corres— ponding to 55). In trans-1,3,5-hexatriene solution, the analytical vpc (5‘ x 0.125 in column, 10% carbowax on chromosorb, AM-DMCS 80/100, 150°, 30 ml/min N2) showed two peaks with retention time of 8.5 min (corresponding to 55, 4‘l 73 10%) and 10.5 min (corresponding to 55, in 90%). In blank solution, the analytical vpc in the same condition as described above except the column temperature at 156° showed a peak with a retention time of 8.5 min (correspond- ing to the compound 55). For Run 2: Analytical vpc (5' x 0.125 in column, 5% SE-30 on chromosorb W, Aw-DMCS 80/100, 131°, 30 ml/min N2) showed a retention time 2 min (cor- responding to the product 55) in sensitization, quenching and blank solutions. PART II ACID-CATALYZED REARRANGEMENTS OF CYCLOHEXADIENONE EPOXIDES (1) 4,5-EPOXY-2,4,5,6,6-PENTAMETHYL-Z-CYCLOHEXENONE (2) 4,5-EPOXY-2,3,4,6,6-PENTAMETHYL-2-CYCLOHEXENONE (3) 4,5-EPOXY-6,6-DIMETHYL-2-CYCLOHEXENONE 74 INTRODUCTION The acid-catalyzed rearrangements of epoxides are of special interest since they provide a simple means of 6e,44 The major converting olefins to carbonyl compounds. product formed from the rearrangement of an epoxide is governed by two main factors. These are the direction of ring opening and the relative migratory aptitude of different substituents. The direction of ring opening in acid-catalyzed re- arrangement is governed by the stability of the carboca- tion which would be formed. The expected sequence is aryl, allyl > tertiary > secondary > primary. Thus in the presence of aluminum trichloride, trimethylethylene oxide rearranges to methyl isoPropyl ketone, isobutylene oxide gives mainly isobutyraldehyde,45 46 and indene oxide gives 2- indanone. Once the ring opens, the structure of the product depends on which group migrates. The relative migratory aptitude of different substituents is, in general, vinyl > 1.5'44 5 it was acyl > H > ethyl > methy In a recent study shown that acyl and methyl migration may compete on an approximately equal basis. Thus on treatment with tri- fluoroacetic acid (TFA), 55 rearranged (Scheme 11) to nearly equal amounts of 55 and (55 + 55). Following protonation, ring opening of 55 occurs in such a manner as to place the positive charge remote from the carbonyl group, giving 75 76 Scheme 11 O O O / i O O ggCHCH3 moiety. The nmr chemical shifts and europium shift data were consistent with structure 85. The product assigned structure 86 corresponded in analysis to loss of water from the epoxyenone 5%. The ir spectrum was very similar to that of 85 (vc 1710, =0 87 ' -1 Vc=CH2 915 cm ), as was the uv spectrum showed a Amax 273 nm (e 9800). The nmr spectrum showed five vinyl pro- tons, and two allylic methyl groups as well as one sharp aliphatic methyl singlet. The nmr chemical shifts and 1.56 0 4°95 (2.23) (2.93)1.8' 1.46 (2.60) l.21(3.33) 4.95(l.56) (2.00)7.00 H 5.09(l.00) europium shift data were consistent with structure 86. Compound 3% was assigned the structure shown, based on (3.12)1.80 l.20(3.70) O 1.20 (2.00) (l.00)6.37 H 4% its spectral properties. The molecular formula C11H1602 was confirmed by the mass spectrum (parent peak m/e 180) and elemental analysis. In the ir spectrum, the diketone $l showed two carbonyl absorptions at 1720 and 1682 cm'l. The uv spectrum showed a Amax at 253 nm, and these data 88 indicate that there is one conjugated and another non- conjugated carbonyl group in a six-membered ring. The nmr spectrum showed an aliphatic singlet for 12 H which is due to two sets of gem-methyl groups. Homoallylic coupling was observed between the allylic methyl and a vinyl proton. Europium shift reagent showed that the two sets of gem-methyl groups were not identical. Compound 81 was assigned its structure based on the O (2'23)1'84 1.20(2.so) (1.38)6.66 H l.96(1.25) CH24.86(1.00) OCOCF3 81 following data, and on its subsequent reaction. The molecular formula C13815<03F3 was confirmed by the mass spectrum (parent peak m/e 276). The ir spectrum showed two strong carbonyl absorptions at 1780 cm'1 (VG-g-CF3) and 1660 cm.1 (vc=o'dienone). The uv spectrum showed a Amax 309 nm (6 4,890) consistent with a dienone chromo- phore. The nmr spectrum showed a gem-dimethyl group, two allylic methyl groups, a methylene group and a vinyl pro- ton. The nmr and europium shift data were consistent with the structure. An alternative structure which was considered 89 and discarded on the basis of labeling results will be described later. The structure of compound 88 is based on its spectral (1-23)1°94 1.23(1.13) (1.00)6.83 H 1.96(l.00) sz 4.06(2.46) on 1.66 8% properties. That it was isomeric with 3% was shown by the mass spectrum (parent peak m/e 180). The ir spectrum showed a broad hydroxyl band at 3100-3700 cm"1 and a strong carbonyl absorption at 1658 cm'l. The uv spectrum showed a Amax at 310 nm (a 4,040). The ir and uv speCtra were consistent with a dienone chromophore. B. Mechanism Alcohol 84 is undoubtedly formed from 33 by proton loss from the intermediate cation k (Scheme 17). The alternative epoxide ring opening mode to give M would lead to structure 82, which is also reasonably consistent with the observed nmr spectrum. However, this structure was conclusively eliminated by a labeling experiment. 90 Scheme 17 o o -——-) 1.20 / H H on on 0 £1 33/3, 12% 4 H + . o o H+ —-—-) H H H on 8 33.2 Preparation of 84 from 34, containing a CD3 group at C-5, gave 84-d3, lacking methyl signal at 61.20. Had the hydroxy ketone been 82, the product would have lacked the vinyl proton signals. Scheme 18 gave plausible mechanisms for the forma- tion of 38 and $1 from 34. 91 Scheme is C) H+ 1,2-alkyl ’25 shift a a + OH R _ H+ 1,2-methyl shift V v . 0 O H O H O 38 mm 4% Scheme 19 gives a plausible mechanism for the forma- tion of 85 from 86. When the reaction was carried out in an nmr tube, the appearance of the sharp singlet due to acetone was observed. _S_cheme 19 92 8m am {I 88 CHBCCH . u N CHBCCH3 + (CF3CO)20 93 The possible routes to 88 and 88 are shown in Scheme 20. Scheme 20 O O + H+ a _H+ Q Hi —.p i -H O / H OH H H 2 C) 3% ._—. H H + $8 TFA 1) 1, 2- -acyl shift 0 m 2) -H+ H IHZ OCOCF3 Scheme 19 fig Hydrolysis 8.5 £3 CH20H 88 mm 94 Independent treatment of 88 with a solution of potassium carbonate in methanol gave a quantitative yield of 88. Scheme 20 also represents the mechanisms by which the various products are obtained from alcohol 88. The expected labeling results (based on Scheme 17-20) are as follows. fié 81*R: OCOCF3 88* R : OH O :0 a 0841 at w-r), The actual results were consistent with these expectations (see Experimental for details of the nmr spectra of the labeled products). 95 Scheme 21 c) C) -H +H' Z- —-> -—-. ——-» H H + 3 H H OCOCF3 8! Q 38 An alternative reaction path from intermediate 8 to give 8 could lead to a product with structure 88 (Scheme 21). Although the nmr, ir and uv spectra for the product assigned structure 81 are.also reasonably consistent with structure 88, this structure (and the mechanism in Scheme 21) is definitely excluded by the labeling experiment. I A product with structure 88, which could be formed from nucleophilic attack at the carbon alpha to the car- bonyl group followed by ring contraction (corresponding to the formation of 88 from 88), was not observed. This may be due to the higher activation energy associated with the formation of a secondary carbonium ion (note the H-atom at C-3). 96 Scheme 22 O O O O O u 34 TFA CF3-C-O TEA mm -+ —//—+ -—-> H H OH H H C 2. AAcid-catalyzed Rearrangement of 4,5-Epoxy-2,3,4,6,6- pentamethyl-2-cyclohexenone (881 Although the epoxy enone 88 could be purified by chromatography over florisil, it was sensitive to heat or small amounts of acid. Chromatography on alumina, treat- ment with a little aqueous acid or subjection to gas chromatography resulted in nearly quantitative rearrange- ment to a hydroxy ketone assigned structure 88. The com- 1 pound 88 shows a v at 1660 cm- consistent with a con- c=o jugated carbonyl group in a six-membered ring. The pres- ence of a hydroxyl group was clear from the vo-H at 3500 cm-1. 97 o (2 16)l 80 0'81(2'64) Trace acid . . 1.00(2.36) ‘ ’ ‘1 4.03(2.60) H (1.00)2.03 OH 1.68 5.32(l.63) 5.35(1.05) 88 2% The nmr spectrum of 88 showed two vinyl protons (55.32, 5.35), two gem-dimethyls as singlets (60.81, 1.00) and two allylic methyl groups (mutually coupled at 61.80, 2.03). Moreover, the ir spectrum showed a terminal methylene group (960, 930 cm'l) and the uv maximum at 275 nm (a 14,430) suggested that both double bonds were linearly conjugated with the carbonyl group. Compound 88 is undoubtedly from 88 by proton loss from the intermediate cation 8 (Scheme 23). The alternative ring-Opening to give 8, possibly followed by a 1,2-acyl shift, would lead to structure 88 which is inconsistent with the observed spectra. Ion 8 is preferred over 8 because it is tertiary and allylic. The formation of 88 is analogous to the formation of 88 from 88 via 8 (Scheme 17). Treatment of 88 with neat TFA at room temperature for 40 minutes gave two isomers of the starting epoxy 98 Scheme 23 o o H+ : 1f; —» n; H H + on éé ~H+ enone. The structure of one product was 2%, already identified. The other product structure SQ was based “1‘ H 0H 554‘, 2% 99 primarily on its spectral properties. The diketone 5% showed two carbonyl absorptions in the ir spectrum at 1720 and 1665 cm-1, and strong uv absorption at 243 nm, indi- cating one conjugated and another non-conjugated carbonyl group in a six-membered ring. The nmr spectrum (in CDC13) showed two mutually coupled vinyl methyl groups (61.98, 2.15), a gem-methyl group (with non-equivalent methyls, as shown by Eu-shift reagent) and a >CH(CH3) group. 0 1.46(3.68) 2. 2 l. 8 ( 3 ) 9 l.46(3.64) (l.00)2.15 O (d)(1.89)1.52 H | 3.32(4.27)(q) J=8 Hz :24 A labeling experiment was performed to establish un- equivocally the nmr assignments of 2% and 5%. Treatment of £5 with DMSO-d6 and potassium t-butoxide gave 35* whose nmr spectrum was identical with that of 35 except that the peak at 51.95 was absent (labeled at C-3). Treat- ment of 35* with neat TFA at room temperature for 40 minutes gave 23* whose nmr spectrum lacked the signals at 52.03 and 53* whose nmr spectrum lacked the signal at 62.15. These results showed that there are no skeletal rearrangements occurring during the acid rearrangement 100 of 35*. Possible mechanisms for the formation of 5% are shown in Scheme 24. Scheme 24 V L The intermediate cation Q can lose a proton to give 25 which, however, was not isolated. Its rapid keto-enol tautomerism accounts for the formation of 2%. An alterna- tive process (preferred reaction pathway) involves a hydride shift and then deprotonation to give product 5%. In the ir spectrum of 5% one could see, in addition to the carbonyl bands, other bands due to Vo-H 3500 (w), and 101 1640 (broad) due to the enol form. Presumably inter- mediate 25 and product 5% are in equilibrium, and the keto form seems the more stable. Obviously in TFA, 25, 5 and product 5% should be in equilibrium In contrast to 12 and 8%, which on protonation re- arranged further, compound 92 underwent no further re- arrangement on treatment with neat TFA under similar conditions. Scheme 25 0 C) a b 5 711' H [I H II OH OH 8 2% fi ‘V 102 There are three possible positions for the protonation of 9% (Scheme 25). Pathway a can be eliminated on the basis that if the allylic cation 8 is formed, it should give 9% and 5% on quenching. But 9% did not rearrange to 5% on treatment with acid. The reaction was repeated on 32*. No trace of 53*was found. Pathway chan be eliminated on the basis of the nmr spectrum of 3% in acid. The nmr spectrum of 2% in neat TFA showed gem-methyls at 51.24 and 1.32 where each ap- peared as a singlet, two allylically coupled methyl groups at 51.96 and 2.23, two vinyl protons at 55.66 and 5.72, and one proton at 54.46 (C-5). The nmr of 22* in neat TFA showed exactly the same chemical shifts except that the signal at 52.23 had disappeared. If carbocation S were formed, it would be difficult to understand why the proton at C-5 should not be strongly deshielded. Examina- tion of the nmr spectrum of 2% in 80% H2504 (54,52), D 504 2 2SO4 (65.33) showed that the proton (64.52) or even 98% D at C-5 is not strongly deshielded. Also, if é were formed, why would it not undergo a 1,2-acyl shift similar to that of compound 12? It is concluded that protonation of 2% occurs via path 9 to give the cation T. This highly stabilized cation may simply be incapable of rearranging under the prescribed experimental conditions, and is converted back to starting material 9% on quenching. Solutions of 32 in TFA or in 80% 82804, D2804 or concentrated D2804 aqueous solution, 103 on quenching, gave back the starting material. o TFA d (2.54)l.89 l.09(3.26) trio (5 ayil l.12(3.18) H ch03 in aq. MeOH H 5.43(2.39) OH (l.37)2.04 COCF3 5.39(1.04) 2% 28 Prolonged treatment of 22 in TFA (over 5 days) con- verted it in part (60%) to 96, through esterification, and 40% of unreacted 2% was recovered. The experiment was repeated with 22*, and showed no skeletal change in the recovered product. The facile removal of the tri- fluoroacetyl group was demonstrated by converting 26 to 2% in a 7% solution of potassium carbonate in aqueous methanol at room temperature for 4 hr. A new and previously unobserved protonation path, namely at the carbonyl oxygen, appears to operate on the acid treatment of 2%. This pathway gives the highly de- localized cation T. 3. Acid-catalyzed Rearrangement of 4,5:Epoxy-6,6-epoxy-6,6- dimethyl-cyclohexen-2:9ney(36) Treatment of enone epoxide £6 with dilute aqueous hydrochloric acid or with hydrogen chloride gas (in 104 ether) leads to the formation of 4-chloro-5-hydroxy-6,6- dimethylcyclohexen-Z-one (21) in 98% yield. HCl V C1 22 22 The halohydrin structure was established primarily from its spectral properties. The molecular formula C8H1102C1 was confirmed by the mass spectrum (parent peak m/e 176) and elemental analysis. The ir spectrum showed a broad 1 hydroxyl absorption at 3100-3800 cm- and a carbonyl ab- 1 sorption at 1685 cm- typical of a conjugated enone. The uv spectrum showed a Amax at 240 nm which substantiated the presence of the conjugated enone moiety. The mass spectrum showed a monochloro fragmentation pattern, showing that the hydroxyl group had not been replaced by a second chlorine atom. The nmr spectrum showed two methyl groups at 61.00 and 1.20 as singlets, two vinyl protons at 65.80 and 6.60 and two methine protons at 63.65 and 4.57. The large coupling constant for the methine protons (J = 8 Hz) implies a trans geometry and hence a trans relationship between the chloro and hydroxyl groups. A plausible mechanistic route to 97 is shown in Scheme 26. Chloride ion displaces at the allylic and also less 105 Scheme 26 O 36 HCl g; c1" mm 7 ' c1'~z’cg crowded carbon to give the trans-chlorohydrin. Treatment of the epoxyenone 36 with TFA gave the tri- fluoroacetate 28.51 The structure of 98 was established o TFA 0 on ococr?3 22 22 primarily by its spectral properties. The molecular formula C10H1104F3 was confirmed by the mass spectrum (parent peak m/e 252). The ir spectrum showed a strong broad hydroxyl group absorption at 3010-3800 cm'1 and a strong carbonyl absorptions at 1780 cm.1 for the trifluoro ester moiety and at 1680 cm-1 for the conjugated enone carbonyl. The uv spectrum had an absorption maximum at 240 nm confirming the conjugated enone moiety. The nmr spectrum showed two methyl groups as a singlet at 61.03, two vinyl protons 65.40-6.04 and 6.40-6.60 as multiplets 106 and two methine protons at 63.74-3.92 and 4.40-5.20 with a coupling constant 8 Hz. The large coupling constant indicates that the hydroxyl and trifluoroacetate groups are trans to each other. Prolonged treatment of 36 with TFA did not bring about any carbocation skeletal rearrangement. In summary, in neat trifluoroacetic acid, the vinyl- logs of a,B-epoxyketones rearrange initially to allylic cations. The intermediate cations may rearrange by a 1,2-hydride shift, a 1,2-methyl shift, a 1,2-alkyl shift (ring contraction), or may be attacked by solvent. In the comparison of the results from 35 with 27, it is obvious that the replacement of H for CH3 in the C-2 position has no effect on the acid-catalyzed rearrangement result. However, the replacement of H for CH3 in the C-3 and C-4 positions significantly effects the acid-catalyzed rearrangements. 107 Scheme 27 R H+ R1 ——’ _—_—. R3 R3 + R1 —% Compound/yield R 2 3a, 24 ea 32 u :34 -- 10% 60% 53% -- deprotonation Rl a b 7 R4 (79)(84)40% 47% -- R2 01‘! 67% 82% C +C 4' 5 : R1 H , R1 27% 8%-- -- -- R2 R R2 301-] R30 nucleophilic CF COOH 6% __ __ __ _- CF coo ‘TL‘2 R1 at Co to 3 R 4 2 carbonyl 2. R 'R . group 3 3 nucleophilic R1 attack - -_ _- _- -- 95% at C to carbony R rou R 4 9 P 2 CFgBOC a. 12 forms fig (97%). b. gg‘forms 85 (4%), QQ(20%) 22(1°%" gg (8%), g1(52%) QQ(6%). EXPERIMENTAL 1. Acid-catalyzed Rearrangement of 4,5-Epoxy-2,4,5,6,6- pentamethyl-2-cyclohexenone (34) in Trifluoroacetic acid A solution of 400 mg (2.22 mmol) of 34 in 4 m1 of ice- cold trifluoroacetic acid was stirred at 0° for 25 min, then at room temperature for 30 min. The reaction was then quenched by pouring the mixture into ice and saturat- ed sodium bicarbonate solution. The products were extracted with ether, and the combined ether layers were washed successively with saturated sodium bicarbonate solution, water, saturated sodium chloride solution and dried (M9804). Evaporation of the solvent left a light yellow liquid, which was analyzed by vpc (10' x 0.125 in column, 20% SE-30 on chromosorb W, AW-DMCS 80/100, 143°, 30 ml/min N2). Products (%, retention time) were observed: 85 (4%, 8 min), 86 (20%, 18 min), 41 (10%, 22 min), 33 (8%, 30 min), 81 (52%, 58 min), 88 (6%, 67 min). The experi- ment was repeated on'a larger scale to isolate all the products. For 4-methy1ene-2,S-dimethyl-Z-cyclopentenone (85): ir (neat) 3010 (m), 2960 (w), 2925 (w), 1710 (s), 1640 (m), 1615 (w), 1460 (m), 1395 (m), 1340 (w), 1300 (w), 1200 (w), 1180 (w), 1160 (w), 1030 (w), 990 (m), 915 (s) 1 cm' ; uv (cyclohexane) Am 275 nm (6 10,800); nmr (CCl4) ax 108 109 51.17 (a, 3 H, J = 7.5 Hz), 1.83 (broad, 3 H), 2.50 (q, l H, J = 7.5 Hz), 4.94 (s, 1 H), 5.03 (s, 1 H), 7.17 (m, 1 H): mass spectrum (70 eV) m/e (rel intensity) 122 (4), 121 (18), 100 (56), 107 (15), 94 (100), 93 (40), 80 (20), 79 (100), 78 (50), 68 (25), 67 (14), 66 (22). Anal. Calcd for C H O: C, 78.65; H, 8.25 8 10 Found: C, 78.63; H, 8.39 For 5-isopropeny1-4-methy1ene-2,S-dimethyl-Z-cyclo- pentenone (86), this product was obtained as white plates which melt at room temperature: ir (neat) 3000 (s), 2965 (w), 1710 (s), 1645 (m), 1610 (w), 1455 (s), 1385 (m), 1340 (w), 1295 (w), 1200 (w), 1150 (w), 1105 (w), 1030 (w), 935 (w), 950 (w), 915 (s) cm'l; uv (cyclohexane) Amax 225 nm (e 9800), 273 (4450): nmr (CC14) see structure; the band at 61.46 was a doublet, J = 1 Hz, that at 51.89 was broadened, that at 57.26 was a multiplet, that at 64.95 was a multiplet (3 vinyl protons) and that at 65.09 was a broaden singlet, and the peak at 61.21 was a sharp singlet: mass spectrum (70 eV) m/e (rel intensity) 162 (65), 147 (35), 134 (89), 133 (21), 119 (100), 118 (25), 105 (32), 91 (70), 77 (35), 65 (20). A231. Calcd for C11H140: C, 81.44; H, 8.70 Found: C, 81.26; H, 8.71 For 2,4,4,6,6-pentamethyl-2-cyclohexen-1,5-dione (41): ir (neat) 3006 (s), 1720 (s), 1682 (s), 1490 (m), 1480 (w), 1400 (m), 1380 (w), 1310 (m), 1260 (w), 1180 (w), 110 1070 (m), 910 (w), 890 (m) cm-l; uv (cyclohexane) Amax 253 nm (e 7200): nmr (CC14) 61.20 (s, 12 H), 1.80 (d, 3 H, J = 0.8 Hz), 6.37 (q, 1 H, J = 0.8 Hz): mass spectrum (70 eV) m/e (rel intensity) 180 (60), 165 (e), 163 (4), 137 (45), 79 (10), 67 (30). Anal. Calcd for C11H1602: C, 73.30; H, 8.95 For 4-trifluoroacetoxymethy1-2,5,6,6-tetramethy1-2,4- cyclohexadienone (81): ir (CC14) 3000 (s), 2900 (s), 1780 (s), 1660 (s), 1600 (m), 1480 (m), 1460 (m), 1395 (m), 1385 (m), 1230 (s), 1180 (m), 1145 (m), 1242 (w), 1010 (w), 905 (w), 820 (s) cm-l; uv (cyclohexane) Amax 223 nm (e 1700), 309 (4900) and UV (CC14) showed Amax at 317 nm; nmr (CC14) 61.20 (s, 6 H), 1.84 (d, 3 H, J = 0.5 Hz), 1.96 (s, 3 H), 4.86 (s, 2 H), 6.66 (m, 1 H): mass spectrum (70 eV) m/e (rel intensity) 276 (20), 185 (1), 162 (77), 147 (58), 134 (28), 119 (100), 105 (25), 91 (52), 79 (20), 77 (30), 69 (42). The mass spectrum of same sample compound changed after a few hours. The parent peak at m/e 276 diminished in intensity and the peak at m/e 180 increased suddenly. At the same time, the sample changed color from yellowish to dark. The mass spectral pattern of the later sample was similar to that of 87. The nmr spectrum also changed; a new set of peaks appeared at 61.23, 1.93, 4.06 and 6.83 corresponding to 81. These data show that 81 was not 111 stable, and no attempt was made to do the elemental analysis. For 4-hydroxymethyl-2,5,6,6-tetramethy1-2,4-cyclo- hexadienone (88): ir (neat) 3510 (br., m), 3500 (w), 3000 (s), 2910 (m), 1658 (s), 1600 (w), 1583 (w), 1460 (w), 1380 (m), 1305 (m), 1280 (w), 1270 (m), 1210 (w), 1180 (m), 1140 (w), 1010 (m), 1000 (m), 890 (m), 820 (s), 815 l (s) cm- ; uv (cyclohexane) Ama 223 (e 2300), 310 (4100) X and uv (CC14) A shifted to 315 nm; nmr (CC14) 61.23 (s, max 6 H), 1.96 (S, 3 H), 1.94 (d, 3 H, J = 1 Hz), 4.06 (s, 2 H), 6.83 (m, l H); mass spectrum (70 eV) m/e (rel intensity) 180 (52), 165 (38), 149 (20), 147 (49), 137 (44), 136 (12): 135 (80), 123 (20), 122 (98), 119 (100), 109 (20), 107 (43), 105 (40), 93 (24), 91 (62), 79 (40), 77 (43), 65 (23). Anal. Calcd for C11H1602: C, 73.30; H, 8.95 Found: C, 73.72; H, 8.93 2. Acid-catalyzed Rearrangement of 5-isopropenyl-4- methylene-2,5-dimethyl-2-cyclopentenone (86) in Tri- fluoroacetic acid A solution of 86 (80 mg, 0.5 mmol) in 1 m1 of trifluoro- acetic acid was allowed to stand at room temperature for 4 hrs, the reaction being monitored by analytical vpc (5' x 0.125 in column, 20% FFAP on chromosorb W, AW-DMCS 80/100, 100°, 30 ml/min N2). The product, compound 85, showed 112 a retention time of 7 min, whereas the starting compound 88 showed a retention time of 16 min. During the reaction, a sharp singlet appeared at 62.33 due to acetone in tri- fluoroacetic acid. In the Vpc, the peak due to 88 dimin- ished and the peak due to 88 increased. The reaction was quenched by pouring it into ice and saturated sodium bi- carbonate solution. The mixture was extracted with ether, and the ether extract was worked up as in the rearrange- ment of 88 to give 75 mg (98%) of 88. 3. Saponification of 4-Trifluoroacetoxymethyl-Z,5,6,6- tetramethyl-Z,4-cyclohexadienone (88) A solution of 88 (80 mg, 0.29 mmol) in 7% potassium carbonate in aqueous methanol (volume ratio of water to methanol was 2 to 7) was stirred at room temperature for 4 hrs. The mixture was poured into ice-water and extracted with ether, the combined ether extracts were washed with water, saturated sodium chloride solution, and dried (M9804). Evaporation of the solvent gave 50 mg of 88 (96%). The residue, analyzed by vpc (5' x 0.125 in column, 20% FFAP on chromosorb W, AW—DMCS 80/100, 160°, 30 ml/min N2), showed that all of 88 was consumed, the sole product was 88. The nmr spectrum also showed that the transfor- mation of 88 to 88 was complete. 113 4. Synthesis of 5-Hydroxy-4-methy1ene-2,5,6,6-tetramethy1- 2-cyclohexenone (88) Compound 88 (0.5 g, 2.7 mmol) was chromatographed on aluminum oxide (Brockmann Activity, Grade 1) using methanol as the eluent, to give 0.5 g (100%) of 88 as a colorless oil. The residue was subjected to analytical vpc (5' x 0.125 in column, 20% FFAP on chromosorb W, AW-DMCS 80/100, 170°, 30 ml/min N2). There was only one peak, correspond- ing to 88 (retention time 6.3 min). Preparative vpc (10' x 0.25 in column, 20% SE-30 on chromosorb W, AW-DMCS 80/ 100, 143°, 60 ml/min He, ret. time 36 min) gave the pure hydroxy ketone 88: ir (neat) 3500 (s), 2995 (s), 1670 (s), 1458 (m), 1382 (w), 1305 (w), 1260 (w), 1200 (m), 1140 (m), 1050 (m), 1005 (m), 950 (m), 930 (m), 805 (w), 780 (m) cm'l; uv (methanol) lma 223 nm (e 4200), 270 x (16,400); mass spectrum (70 eV) m/e (rel intensity) 180 (59), 165 (100), 147 (18), 137 (90), 123 (45), 119 (45), 109 (23), 107 (19), 95 (37), 91 (46), 77 (35), 67 (40), 66 (37); nmr (CC14) 61.00 (s, 3 H), 1.08 (s, 3 H), 1.64 (s, 1 H), 1.72 (d, 3 H, J = 1 Hz), 5.14 (s, 1 H), 5.28 (s, 1 H), 6.96 (q, l H, J = 1 Hz). 5231. Calcd for C11H1602: C, 73.30; H, 8.95 Found: C, 73.14; H, 8.92 Compound 88 could be also obtained by treating 88 with TFA for a shorter period of time (less than 30 min). Also, 114 subjecting to Vpc (10' x 0.25 in column, 20% SE-30 on chromosorb W, AW-DMCS 80/100, 120°, 60 ml/min He) gave 80% of the hydroxy ketone 88. 5. Acid-catalyzed Rearrangement of 5-Hydroxy-4-methy1ene- 2,5,6,6-tetramethy1-2-cyglohexenone (88) in Trifluoro- acetic acid A solution of 88 (400 mg, 2.22 mmol) in 4 m1 of ice cold trifluoroacetic acid was stirred at 0° for 25 min, then at room temperature for 30 min, and then was quenched and worked up as described for the rearrangement of 88 in TFA. The crude product was analyzed by vpc (10' x 0.125 in column, 20% SE-30 on chromosorb W, AW-DMCS 80/100, 143°, 30 ml/min N2) and gave compounds 88, 88, 88, 88, 88, and 88 in approximately the same ratio as from 88. 6. Acid-catalyzed Rearrangement of 5-Trideuteromethyl-, 4,5-epoxy-2,4,6,6-tetramethyl-2-cyclohexenone ( 8:1 The experimental and workup procedures were as described for the treatment of 88 with trifluoroacetic acid. The nmr spectra of the rearrangement products: 88* had an nmr spectrum identical with that of 88 except that the sig- nal at 61.17 disappeared; 88* had an nmr spectrum identical with that of 88 except that signal at 61.21 disappeared; 88* had an nmr spectrum identical with that of 88 except that signal at 61.95 disappeared; 88* had an nmr spectrum identical with that of 88 except that the intensity of the 115 peak at 61.20 was reduced from 12 H to 9 H; 88* had an nmr spectrum identical with that of 88 except that the signal at 61.96 disappeared; 88* had an nmr spectrum identical with that of 88 except that the signal at 61.96 disappeared; 88* had an nmr spectrum identical with that of 88 except that the signal at 61.20 disappeared. 7. Acid-catalyzed Rearrangement of 88: The procedure and workup were as described for the treatment of 88 with TFA, and the same products were ob- tained as in the rearrangement of 88* in TFA. 8. Acid-catalyzed Rearrangement of 4,5-Epoxy-2,3,4,6,6- pentamethy1-2-cyclohexenone (88) A solution of 88 (300 mg, 1.66 mmol) in 3 ml of TFA, prepared at 0°, was allowed to stir for 40 min (monitoring by nmr) at room temperature, then poured into a cold sodium bicarbonate solution and extracted several times with ether. The combined ether extracts were washed succes- sively with saturated sodium bicarbonate solution or 30% potassium hydroxide solution, water and saturated sodium chloride solution, and dried (MgSO4). Solvent removal left 291 mg of a light yellow oil. The residue was sub- jected to analytical vpc (5' x 0.125 in column, 20% FFAP on chromosorb W, AW-DMCS 80/100, 150°, 30 ml/min N2), and showed two peaks corresponding to 88 (60%, retention time 116 8.5 min) and 88 (40%, 16 min). Preparative vpc (10' x 0.25 in column, 20% FFAP on chromosorb W, AW-DMCS 80/100, 180°, 60 ml/min He) gave 2,3,4,6,6-pentamethy1-2-cyclo- hexen-1,5-dione (88): ir (neat) 3500 (w), 3010 (s), 2980 (m), 1720 (s), 1665 (s), 1640 (m), 1480 (m), 1400 (m), 1; uv (methanol) 1 1345 (m), 1180 (m), 1040 (m) cm- max 210 nm (e 3370, shoulder), 243 (5620); nmr (CDC13) 61.46 (s, 6 H), 1.52 (d, 3 H, J = 8 Hz), 1.98 (br., 3 H), 2.15 (br., 3 H), 3.32 (q, 1 H, J = 8 Hz); mass spectrum (70 eV) m/e (rel intensity) 180 (55), 165 (10), 137 (18), 110 (100), 109 (20), 95 (20), 82 (25), 67 (76). 523;. Calcd for C11H1602: C, 73.30; H, 8.95 Found: C, 73.33; H, 9.01 For 4-methy1ene-5-hydroxy-2,3,6,6-tetramethyl-2-cyclo- hexenone (88): ir (neat) 3502 (s), 3000 (s), 2960 (s), 2900 (s), 1660 (s), 1600 (m), 1480 (m), 1400 (s), 1300 (m), 1220 (m), 1125 (m), 1115 (m), 1070 (s), 923 (s) curl; uv (methanol) Ama 208 nm (e 3680), 275 (14,430); nmr (CC14) x 60.81 (s, 3 H), 1.00 (s, 3 H), 1.80 (br., 3 H), 2.03 (br., 3 H), 4.03 (s,.1 H), 5.32 (s, 1 H), 5.35 (s, 1 H): mass spectrum (70 eV) m/e (rel intensity) 180 (63), 175 (23), 151 (21), 135 (100), 122 (30), 121 (30), 120 (21), 119 (45), 110 (10), 109 (25), 91 (24), 80 (35), 79 (34), 78 (28), 67 (20). 523;. Calcd for C11H1602= C, 73.30; H, 8.95 Found: C, 73.23; H, 8.96 117 9. Acid-catalyzed Rearrangement of88: The procedure and workup procedure were as described for the treatment of 88 with TFA. The rearrangement product 88* had an nmr spectrum identical with that of 88 except that the signal at 62.03 disappeared. The other product, 88*, had an nmr spectrum identical with that of 88 except that the signal 62.15 was absent. 10. Acid-catalyzed Rearrangement of 4-Methylene-5-hydroxy- 2,3,6,6-tetramethyl-2-cyclohexenone (88) A. With TFA for 5 hr A solution of 88 (0.25 g, 1.38 mmol) in 3 m1 of ice- cold TFA was stirred at room temperature for 5 hr. The reaction was monitored by nmr. The nmr spectrum of 88, which remained constant with time, in neat TFA (methylene chloride as internal standard) showed peaks at 61.24 (s, 3 H), 1.32 (s, 3 H), 1.96 (s, 3 H), 2.23 (s, 3 H), 4.46 (s, l H), 5.66 (s, l H), 5.72 (s, 1 H). The reaction mix- ture was quenched by pouring it into ice and saturated sodium bicarbonate solution. The product was extracted with ether, and combined ether layers were washed suc- cessively with saturated sodium bicarbonate solution, water and saturated sodium chloride solution, and dried (M9804). Evaporation of the solvent left 0.25 g of un- rearranged compound 88 (by nmr). 118 B. With TFA for 5 days A solution of 2% (250 mg, 1.39 mmol) in 3 ml of TFA was stirred at room temperature for 5 days and worked up as described in (A) to leave 0.31 g of yellowish liquid. One product was the starting material (40%), and the other product, which was analyzed by preparative vpc (10' x 0.25 in column, 20% FFAP on chromosorb W, AW-DMCS 80/100, 190°, 60 ml/min He), gave trifluoroacetyl derivative 26 (ret time 23 min): ir (CC14) 3000 (m), 1780 (s), 1679 (s), 1600 (w), 1480 (w), 1400 (w), 1240 (s), 1180 (s), 1160 (s), 1040 (w) cm‘l; uv (methanol) Ama 210 nm (e 7200), x 272 (16,200); nmr (CC14) see structure; the peaks at 61.89 (3 H) and 2.04 (3 H) showed homoallylic coupling, whereas the other methyl groups were singlets at 61.09 and 1.12. There were also three vinyl hydrogens at 65.39, 5.34 and 5.43; mass spectrum (70 eV) m/e (rel intensity) 276 (38), 171 (25), 163 (100), 147 (40), 135 (59), 134 (34), 128 (30), 119 (78), 112 (30), 107 (28), 105 (45), 91 (55). Anal. Calcd for C13H1503F3: C, 56.65; H, 5.47 Found: C, 56.61; H, 5.56 c. With 80% H,SOA or 0,30, About 0.5 m1 of CCl4 containing 40 mg of 3% was placed in an nmr tube. Under a nitrogen atmosphere, 1 m1 of 80% D2SO4 (or H2804) was added. The contents were then mixed using a "supermixer" (Matheson Scientific, No. 601-0005) 119 to give a brownish yellow solution of 2% and then the CCl4 was removed by a pipette. In another experiment, neat 2% (40 mg) was placed in an nmr tube, and 1 ml of 80% D2804 was added under a nitrogen atmosphere. The contents were mixed. In each case, the nmr spectrum (for 80% H2804 using CH2C12 as internal standard and 80% D2SO4 using (CH3)4N+BFZ as internal standard) showed peaks at 61.32 (s, 6 H), 2.00 (s, 3 H), 2.32 (s, 3 H), 4.52 (s, 1 H), 5.76 (s, 1 H), 5.91 (s, 1 H). The mixture was poured into ice cold saturated sodium bicarbonate solution and extract- ed with ether. Workup as in (A) gave 40 mg of unreacted 2%- D. With Concentrated D go 2 4 The procedure and workup were as described in (C). The nmr spectrum of carbocation [(CH 4N+BF4 as internal 3) standard] showed 61.74 (s, 6 H), 2.37 (s, 3 H), 5.33 (s, 1 H), 7.40 (s, 1 H), 7.63 (s, l H). The mixture was quenched by pouring into ice water and the starting material 2% was recovered. ll. Acid-catalyzed Rearrangement of 2%: The procedure and workup were as described for the treatment of 3% with TFA. The nmr spectrum of the carboca- tion 22* was identical with that of 22 except that the signal at 62.03 had disappeared. On prolonged treatment 120 of 22* (over 5 days) the ester 26* was formed; it had an nmr spectrum identical with that of 96 except that the signal at 62.04 had disappeared. 12. Saponification of Trifluoroacetyl Derivative 96 Compound 96 (380 mg, 1.38 mmol) was hydrolyzed by stirring it with 14 ml of a 7% solution of potassium carbonate in methanol for 4 hr. The reaction mixture was poured into water and the organic product was extracted with ether. The combined ether layers were washed suc- cessively with water, saturated sodium chloride solution and then dried (MgSO4). Evaporation ofsolvent left 240 mg of a light yellow liquid whose nmr and ir spectra were identical with those of 2%. 13. Saponification of 26: The procedure and workup were as described for treat- ment of 26 with aqueous methanol. The product had an nmr spectrum identical with that compound 22*. 14. Acid-catalyzed Rearrangement of 4,5-Epoxy-6,6-dimethyl- cyclohexen-Z-one (36) in Hydrochloric Acid A solution of 100 mg of 36 in 5 m1 of ether was shaken with 5 ml of 3% hydrochloric acid, and left at room tem- perature for 4 hr. The ether layer was then washed with 121 aqueous sodium bicarbonate, dried (MgSO4), and evaporated the solvent to give 123 mg (98%) of 21. Analytical vpc (5' x 0.125 in column, 5% FFAP on chromosorb W, AW-DMCS 80/100, 156°, 30 ml/min N2) showed one peak with a reten- tion time of 3.5 min. Preparative Vpc (10' x 0.25 in column, 20% SE-30 on chromosorb W, AW-DMCS 80/100, 150°, 60 ml/ min He, ret time 20 min) gave only one pure product, trans- 4-chloro-5-hydroxy-6,6-dimethy1-cyclohexen-Z-one (21): ir (neat) 3100—3800 (br.), 3010 (m), 2950 (m), 1685 (s), 1480 (m), 1380 (w), 1360 (w), 1300 (m), 1260 (w), 1220 (w), 1185 (w), 1140 (m), 1100 (s), 1070 (s), 1010 (w), 1000 (w), 875 (s), 840 (s), 800 (s) cm-l; uv (methanol) Amax 240 nm (e 1500); nmr (CC14) 61.00 (s, 3 H), 1.20 (s, 3 H), 3.65 (d, l H, J = 8 Hz), 4.57 (m, l H), 5.80 (m, l H), 6.60 (m, l H); mass spectrum (70 eV) m/e (rel intensity) 176 (6), 175 (2), 174 (18), 139 (7), 122 (11), 104 (29), 102 (85), 95 (10), 91 (10), 79 (20), 77 (18), 72 (100), 70 (15). Anal. Calcd for C H 0 Cl: C, 55.17; H, 6.32 8 11 2 Found: C, 55.06; H, 6.32 15. Acid-catalyzed Rearrangement ofgé in Hydrogen Chloride Compound 36 (100 mg, 0.72 mmol) in 3 m1 ether was placed in a 25-ml flask and cooled to 0° in an ice bath. Hydrogen chloride gas was passed slowly into the ether solution for 30 min. The solution was washed with aqueous sodium 122 bicarbonate and dried (MgSO4), and evaporated to give 120 mg (95%) of gz. 16. Acid-catalyzed Rearrangement ofg§6 in Trifluoroacetic Acid A solution containing 100 mg of 36 in 2 ml of carbon tetrachloride was added to 8 ml of TFA under nitrogen. The mixture was stirred at room temperature and the reac- tion was monitored by nmr. It was complete in a half hour. The mixture was quenched with cold saturated sodium bicarbonate solution and extracted several times with ether. The combined ether extracts were washed succes- sively with saturated sodium bicarbonate solution, water and saturated sodium chloride, and dried (M9804). After solvent removal there remained 120 mg (95%) of a light yellow oil. The crude product was subjected to Vpc (5' x 0.125 in column, 5% FFAP chromosorb W, AW-DMCS 80/100, 155°, 30 ml/min N2). There was one peak, with a retention time of 3.5 min. Preparative Vpc (10' x 0.25 in column, 20% SE-30 on chromosorb W, AW-DMCS 80/100, 150°, 60 ml/min He, ret time 11 min) gave pure hydroxy trifluoroacetate 28: ir (neat) 3010-3800 (br.), 3000 (w), 2950 (w), 1780 (s), 1680 (s), 1475 (w), 1400 (m), 1380 (m), 1230 (m), 1160 (m), 1080 (w), 960 (m) cm’l; uv (methanol) )max 240 nm (e 2900); nmr (CC14) 61.03 (s, 6 H), 1.66 (br. 1 H), 2.61 (s, l H), 3.83 (d, 1 H, J = 8 Hz), 4.40-5.20 (m, 1 H), 5.40-6.04 (m, 1 H), 6.40-6.60 (m, l H); mass spectrum 123 (70 eV) m/e (rel intensity) 252 (4), 225 (2), 196 (2), 180 (9), 156 (6), 138 (31), 122 (9), 109 (21), 95 (19), 93 (12), 91 (10), 86 (23), 85 (34), 84 (100), 83 (33), 82 (30), 72 (90), 68 (57). Anal. Calcd for C10H1104F3: C, 47.63; H, 4.40 Found: C, 47.82; H, 4.36 PART III DIELS-ALDER REACTIONS OF A DIHYDROBENZOPENTALENE AND THE SYNTHESIS OF HIGHLY-STRAINED QUADRICYCLANES 124 INTRODUCTION An unexpected, profound, but facile rearrangement of a highly methyl-substituted bicyclo[2.2.2]-2-octy1 system was reported by Hart. Either of the epimeric alcohols 29 or $88 rearranged on dehydration with strong acid to the benzodihydropentalene lge.52'53 of?” 0:13; o 22 688 49% Zilfiialo 88% $88 fl A plausible mechanism analogous to those postulated for the rearrangement of other bicyclo[3.2.l]octadienyl cations is shown in Scheme 28. The key step is a cyclopropyl- carbinyl-cyclopropylcarbinyl rearrangement.54-57 125 126 Scheme 28 H + + _ 99 +3 1,2-aryl 1.2 aryl """—-) , ; . 3 or 41200 shlft shift 888 888 888 \ q + O —-* O —+ ’ —+ H H 888 cyclopropyl- carbinyl- I . 0 —» cyclopropyl- O H carbinyl rearrangement 888 + 1,2-methyl -H . shift ; O 888 888 127 The structure of 89% was deduced from its spectral properties and chemical transformations. Compound 88% is hydrogenated preferentially at the central double bond, and forms Diels-Alder adducts l0; and leg with tetracyano- ethylene and N-phenylmaleimide, respectively. 1-11 . 0.60 1.42 1.-1 1.04 NC N CN CN 888 It was of interest to further explore the reactions of 888' and in particular to study its reactions with acetylenic dienophiles to see if strained norbornadienes could be formed in this way (bridged from C-l to C-2 by three carbon atoms), and if so, whether they could be converted to quadricyclanes on irradiation. Results of this study form Part III of the thesis. RESULTS AND DISCUSSION 1. Structures of the Diels-Alder Adducts The highly substituted benzodihydropentalene lg; gave Diels-Alder adducts with dimethyl acetylenedicarboxylate,58 diethyl azodicarboxylate,59 3,6-dimethylbenzene,60 and 2- butyne.54 CH lOOCCECCOOCH 3 7‘ O. CH3OOC COOCH 882 (90%) 3 CZHSOOC-N=N-COOC2H 0‘ dioxane , A ,N—N\ CZHSOOC COOCZH5 00‘ fl 1588 (86%) ‘7 A 00 88% (68%) 001 6.0 80 a» v CHBCECCH3 sealed tube .) 888 (72%) 128 129 Compound 89% was assigned the structure shown: 0.67 1.21 1.78 1.23 CHBOOC COOCH3 3.45 3.70 888 The molecular formula C24H2804 was confirmed by the mass spectrum (parent peak m/e 380) and elemental analysis. The ir spectrum of 892 showed carbonyl absorption at 1703 cm'l, consistent with the presence of a,B-unsaturated ester groups. Compound 822 had a simple benzenoid uv spectrum. The nmr spectrum showed four aliphatic methyl singlets, one allylic methyl singlet, two methoxyl singlets and four aromatic protons. The nmr assignment is based on model compound $98 and chemical transformations. Compound llg was assigned the structure shown: 1.06 .20 1.08 1.18 1.80 1.20 7.09-7.29 N..N 4.00 1.18 C! ‘\ CH3-CH200 COO-CHZ-CH3 1.18 4.00 888 130 The molecular formula C24H32N204 was confirmed by the mass spectrum (parent peak m/e 412) and elemental analysis. The ir spectrum of llg showed a broad, strong carbonyl 1 and the uv spectrum absorption peak between 1690-1740 cm- showed only a benzenoid chromophore. The nmr spectrum showed four aliphatic methyl singlets, one allylic methyl singlet, four aromatic protons as multiplet as well as methyl triplets, and methylene quartets for the ethyl groups. ~The nmr assignment is based on model compound 888- Compound lll was assigned the structure shown: 888 The mass spectrum showed a parent peak at m/e 342. The uv spectrum showed a benzenoid chromophore. The nmr spectrum showed five aliphatic methyl singlets and three allylic or aromatic methyl singlets. The nmr assignment is based on model compound 898' Compound ll; was assigned the structure shown. 131 888 The molecular formula C22H28 was confirmed by the mass spectrum (parent peak m/e 292) and elemental analysis. The uv spectrum showed a simple benzenoid chromophore. The nmr spectrum showed five aliphatic methyl singlets, two allylic methyl singlets and four aromatic protons. The nmr assignment was also based on the model compound 888 - 2. Quadricyclane Synthesis61 The well-known photoisomerization of norbornadienes to quadricyclanes was originally reported by Cristol and Snell.62 This intramolecular cycloaddition reaction {%$3 + 1&6) has been studied in detail. The conversion 11; + 88% can be accomplished either by direct or sensi- tized excitation. The photoproduct llg is surprisingly stable thermally (t = 14 hr at 140°), the 20 + 2n 1/2 opening llm + llg being "forbidden" by the Woodward-Hoff- mann rules. The rate of reaction llm + ll; is dramatically 132 increased (tl/2 = 45 min at -26°) by coordination with a metal, making the isomerization an "allowed" process. The bicycloheptadiene carboxylic acid62 has been converted to the quadricyclane isomer without a sensitizer. OOH COOH ‘7 COOH OOH Although numerous publications have been devoted to this one-step synthesis of quadricyclane derivatives, relatively little is known about the synthesis of related highly strained and substituted systems. 133 KOH HCl 93% \ O. V CHZNZ 95% HOOC COOH 888 \ A hv hv A 100% 94% 92% 100% \ cn3ooc COOCHB HOOC COOH 888 888 The synthesis of 881 originated with 892' Saponifica- tion of 822 yielded the di-acid llé. The structure of llé was substantiated in several ways. Esterification of llé with diazomethane proceeded smoothly to give back 892' Photoisomerization of llé gave a quadricyclane diacid 888- Compound llé was assigned the structure shown. The molecular formula C22H24O4 was confirmed by the mass spectrum (parent peak m/e 352) and elemental analysis. The ir spectrum of ll§ showed carbonyl absorptions at 134 888 1690 and 1682 cm.1 which are consistent with the presence of a,B-unsaturated acid groups. The uv spectrum showed a simple benzenoid chromophore. The nmr spectrum showed five aliphatic methyl singlets, one allylic methyl singlet, four aromatic protons as a multiplet, and two carboxylic acid protons. Irradiation of an ether solution of compound 882 through vycor gave a good conversion to 888' The pho- tolysis product was more stable thermally than similar previously reported quadricyclane derivatives. Although 888 is stable at room temperature, it slowly reverts to the starting diene 892 at a higher temperature. At 170°, the half-life of the isomerization is 30 min. The quadricyclane could be recrystallized from cyclohexane without thermal isomerization. Compound 888 was assigned the structure shown. The molecular formula C24H2804 was confirmed by the mass spectrum and elemental analysis. The ir spectrum showed a carbonyl absorption at 1700 cm-1. The uv spectrum showed the benzenoid chromophore. The nmr 135 6.80-7.43 ‘ 888 spectrum showed six aliphatic methyl singlets and two methoxy singlets. Irradiation of an ether solution of 888 through vycor filter gave a good conversion to 881' The photolysis product was less stable thermally than ester 888' Com- pound 888 slowly reverts to the starting diene llé at room temperature, the half-life of the isomerization being about one month. At 89°, the half-life of the isomerization is 30 min. The structure of 888 was proved by spectral properties and chemical transformation. Upon treatment with diazo- methane, compound 888 was converted to 888 in nearly quantitative yield. The molecular formula C22H2404 was confirmed by its mass spectrum (parent peak m/e 352) and elemental analysis. The ir spectrum of ll; showed carbonyl 1 absorption at 1700 cm” . The uv spectrum showed a simple benzenoid chromophore. The nmr spectrum showed six 136 aliphatic methyl singlets. The cycle of photochemical and thermal isomerizations clearly establishes the structural relationships between the four compounds. In summary, highly strained dihydropentalene 898 under- went Diels-Alder reactions with tetracyanoethylene, N- phenylmaleimide, dimethyl acetylenedicarboxylate, diethyl azodicarboxylate, 1,4-dimethylbenzene and 2-butyne easily. Structural distortions of the basic skeleton caused by substitution have practically no influence on the results of the photochemical reaction. Highly strained norbornadiene derivatives substituted by several large groups such as 883 and llé are still very effectively converted to the tetracyclic systems. EXPERIMENTAL 1. Diels-Alder Adducts of the Dihydrobenzopentalene £888.). A. With Dimethyl Acetylenedicarboxylate To a solution of 2.4 g (0.01 mol) of dihydropentalene 88% in 50 ml of xylene was added 1.4 g (0.01 mol) of di- methyl acetylenedicarboxylate. The mixture was allowed to stand for 10 hr, then cooled and evaporated. The crystal- line solid was filtered and recrystallized from ethanol to give 3.39 g (90%) of 892 as a white solid, mp 106-107°; ir (Nujol) 2900 (s), 1703 (br, s), 1603 (s), 1406 (s), 1380 (s), 1340 (w), 1300 (w), 1280 (w), 1230 (m), 1140 (m), 1080 (m), 1050 (w), 1000 (w), 950 (w), 900 (w), 850 1 (m) cm' ; uv (ethanol) Am 228 nm (6 15,850), 250 (5900); ax nmr (CDC13) 60.67 (s, 3 H), 1.21 (s, 6 H), 1.23 (s, 3 H), 1.46 (s, 3 H), 1.78 (s, 3 H). 3.45 (s, 3 H), 3.70 (s, 3 H), 7.15 (m, 4 H); mass spectrum (70 eV) m/e (rel intensity) 380 (4), 365 (43), 349 (4), 333 (100), 307 (39), 289 (56), 274 (14), 259 (18), 231 (6), 215 (4), 202 (4), 191 (4), 178 (6), 159 (6), 73 (18). Anal. Calcd for C24H2804: C, 75.76; H, 7.42 Found: C, 75.63; H, 7.27 137 138 B. With Diethyl Azodicarboxylate The dihydrobenzopentalene 898 (0.5 g) in 10 m1 of dioxane was refluxed with 0.34 g of diethyl azodicarboxy- late for 35 hr. The solution was evaporated to an oil which crystallized on chilling. The crude product was recrystallized from cyclohexane to yield 0.66 g (85%) of 888 as a pale yellowish solid, mp 106-107°. The adduct had the following characteristics: ir (Nujol) 3300 (s), 2950 (8), 1690-1740 (br, s), 1610 (m), 1595 (w), 1520 (m), 1480 (s), 1380 (s), 1320 (s), 1220 (s), 1180 (w), 1110 (m), 1080 (s), 1002 (w), 890 (s), 880 (s) cm'l; uv (ethanol) Amax 210 nm (8 10,730), 220 (10,000), 228 (11,000), 236 (13,200), 244 (11,800), 310 (18,200); nmr (CDC14) 61.06 (s, 3 H), 1.08 (s, 3 H), 1.18 (t, 6 H), 1.20 (s, 6 H), 1.80 (s, 6 H), 4.00 (q, 4 H), 7.09-7.29 (m, 4 H); mass spectrum (70 eV) m/e (rel intensity) 412 (52), 397 (3), 367 (2), 339 (5), 324 (7), 309 (5), 293 (7), 279 (5), 253 (9), 236 (100), 221 (82), 208 (27), 207 (69), 195 (22), 194 (23), 193 (22), 192 (17), 178 (20), 165 (22), 128 (10), 118 (13), 116 (15), 104 (8), 86 (12), 71 (10). Anal. Calcd for C24H32N204: C, 69.88; H, 7.82; N, 6.79 Found: C, 69.88; H, 7.69; N, 6.77 C. With BIG-Dimethylbenzyne 3,6-Dimethylbenzyne was generated by Friedman's method. A mixture of 0.734 g (3.5 mmol) of 3,6-dimethylbenzene- 139 diazonium carboxylate hydrochloride 0.83 g (3.5 mmol) of dihydropentalene 898 and 0.8 m1 of propylene oxide which was added last, in 10 m1 of ethylene chloride was warmed up gradually while stirring was continued. As gas evolu- tion started, the temperature was controlled so that no vigorous foaming took place. Ten minutes after gas forma- tion started, the solution became clear and it was heated at reflux for one hr. Removal of solvent from the re- action mixture left a brown liquid which was redissolved in ether and the ethereal solution was washed with dilute aqueous sodium hydroxide three times then with water two times and was dried (M9804). Evaporation of the solvent yielded a yellow oil lll (68%). Purification was effected by preparative vpc (5' x 0.25 in column, 3% SE-30 on chromo- sorb W, AW-DMCS 80/100, 195°, 60 ml/min He, ret time 7 min). Ir (CC14) 2890 (s), 1480 (w), 1450 (s), 1380 (m), 1360 (m), 1305 (w), 1290 (w), 1120 (m), 1095 (m), 1030 (w), 1020 (w), 940 (w), 800 (m), 760 (m), 680 (m) cm‘l: uv (cyclohexane) Amax 225 nm (e 1800); nmr (CDC13), see structure; mass spectrum (70 eV) m/e (rel intensity) 342 (31), 328 (10), 327 (39), 312 (10), 300 (30), 299 (100), 298 (15), 285 (24), 284 (22), 282 (12), 271 (30), 270 (15), 267 (20), 257 (22), 253 (18), 252 (16), 243 (17), 239 (23), 228 (11), 171 (35), 164 (20), 156 (32), 141 (34), 133 (50), 126 (42), 114 (15), 101 (10), 89 (10). 140 D. With 2-Butyne A solution of 0.5 g of dihydrOpentalene 898 in three times its volume of 2-butyne was heated in a sealed tube at 200° for four and a half days. Evaporation of the vola- tiles left a residue which was purified through preparative Vpc (5' x 0.25 in column, 20% FFAP on chromosorb W, AW— DMCS 80/100, 150°, 60 ml/min He, ret time 25 min) to give pure 88% (72%) as a colorless liquid: ir (neat) 2992 (s), 1620 (w), 1598 (w), 1434 (s), 1380 (s), 1360 (m), 1280 (m), 1260 (w), 1120 (w), 1080 (w), 1029 (m), 1000 (w), 970 (w), 880 (m), 675 (s) cm'l; uv (ethanol) xmax 290 nm (e 3240); nmr (CC14), see structure; mass spectrum (70 eV) m/e (rel intensity) 292 (100), 277 (63), 262 (10), 249 (30), 235 (50), 221 (44), 207 (40), 193 (25), 171 (20), 165 (20), 157 (20), 141 (20), 129 (29), 121 (20), 115 (20). Anal. Calcd for C22H28: C, 90.35; H, 9.65 Found: C, 90.57; H, 9.74 2. Saponification of the Diels-Alder Adduct 882 Typically, the ester 89% (0.8 g, 2.2 mmol) was saponi- fied by refluxing for 8 hr with 47 ml of 8% ethanolic potassium hydroxide solution. The mixture was then poured into 50 ml of water to dissolve the salt residue. Upon acidification of the alkaline solution with 25% dilute hydrochloric acid, 0.245 g (94%) of the acid ll; was obtained. Recrystallization from cyclohexane gave product 141 melting at 242-243°; ir (Nujol) 2900 (s), 2650 (w), 2550 (w), 1690 (s), 1682 (s), 1620 (w), 1480 (s), 1420 (m), 1380 (m), 1300 (m), 1220 (w), 1160 (m), 1120 (w), 1020 (w), 940 (m), 780 (m) cm'l; uv (ethanol) Ama 215 nm (e x 14,150), 250 (4580); nmr (CDC13), see structure; mass spectrum (70 eV) m/e (rel intensity) 352 (1), 337 (19), 334 (48), 319 (100), 293 (92), 275 (41), 260 (10), 247 (17), 232 (12), 215 (24), 202 (24), 189 (21), 178 (20), 165 (18), 99 (30), 84 (41). Anal, Calcd for C C, 74.97; H, 6.86 2H2804’ Found: C, 75.12; H, 7.19 2 3. Photoisomerization of Diene Diacidllé to Quadricyclane Diacid 888 A solution containing 0.354 g (1 mmol) of llé in 130 ml of ether was degassed and irradiated through a vycor filter (Hanovia 450-W lamp). After one and one-half hr, the ether was removed in vacuo to leave yellow crystals (92%) which nmr showed to be mainly one product. The solid was recrystallized from a mixture of carbon tetrachloride- cyclohexane. Ir (Nujol) 2890 (s), 2630 (w), 2580 (w), 1700 (s) 1680 (s), 1600 (m), 1450 (s), 1382 (s), 1300 (m), 1150 (m): 1005 (w), 920 (w), 750 (m) cm'l; uv (ethanol) Amax 217 nm (a 18,300), 249 (9800): nmr (CDC13), see structure; mass spectrum (70 eV) m/e (rel intensity) 352 (1), 337 (5), 334 (20), 319 (100), 293 (99), 260 (45), 247 (54), 232 142 (62), 215 (25), 202 (24), 189 (15), 178 (20), 165 (16). Anal. Calcd for C22H24O4: C, 74.97; H, 6.86 Found: C, 75.12; H, 7.10 The quadricyclane diacid 888 did not have a definite vmelting point. The solid begins to soften at 86°, the exact temperature depending upon the rate of heating. This mp behavior is in striking contrast with the diene diacid 883 which melts at 242-243°. 4. Photoisomerization of Diene Diester 892 to Quadricyclane Dicarboxylate llé A solution containing 0.358 g (1 mmol) of diester 882 in 130 m1 of ether was degassed and irradiated through a vycor filter. After two and one-half hr, the ether was removed in vacuo to leave yellow crystals (94%). An nmr spectrum showed only one product and the solid was re- crystallized from cyclohexane and acetone. Ir (Nujol) 2900 (s), 1720 (s), 1700 (s), 1460 (s), 1380 (s), 1300 (w), 1200 (w), 1180 (w), 1120 (w), 1040 (m), 940 (m) cm‘l; uv (ethanol) Am 234 nm (E 19,700), 268 (5830), 275 (4850); ax nmr (CDC13), see structure; mass spectrum (70 eV) m/e (rel intensity) 380 (6), 365 (7), 349 (6), 333 (67), 321 (57), 307 (100), 289 (58), 274 (47), 261 (30), 247 (59), 231 (50), 233 (25), 215 (45), 202 (38), 189 (37), 178 (43), 165 (26). Anal. Calcd for C24H2804: C, 75.76; H, 7.42 Found: C, 75.74; H, 7.43 143 5. Isomerization of Quadricyclane Diacid 881 to Diene Diacid llé Twenty-five milligrams of ll; dissolved in 1 ml of carbon tetrachloride was heated at reflux in a sealed nmr tube in an oil bath at 89°. The reaction was followed by nmr. As peaks due to 888 decreased in intensity, a new set of peaks corresPonding to llé appeared. After 1 hr, the peaks of llé were fully developed. The nmr and ir spectra of residue were identical with those of an authentic sample prepared by saponification of 882' 6. Isomerization of Quadricyclane Diesterllg to Diene Diesterlgg Twenty-five milligrams of llé dissolved in 1 m1 of carbon tetrachloride was heated at reflux in a sealed nmr tube in an oil bath at 170°. The reaction was followed by nmr. After 1 hr, the reaction was complete. The nmr and ir spectra of residue were identical with those of authentic sample 882‘ 7. Esterification oflgpadricyclane Diacid 882 to Quadri- cyclane Diester 888 Treatment of the quadricyclane diacid ll; with diazo— methane gave the dimethyl ester in excellent yield. A typical preparation is described. To a 100-ml Erlenmeyer 144 flask cooled in an ice bath and containing 1.76 g (5 mmol) of ll; dissolved in 45 m1 of absolute methyl alcohol was added an ethereal solution of diazomethane in small por- tions until gas evolution ceased and the solution acquired a pale yellow color. The solvent and excess reagent were evaporated and the residue was recrystallized from cyclo- hexane and acetone to give 1.80 g (94%) of white crystals. The nmr and ir spectra were identical with those of authen- tic sample of llé. 8. Esterification of Diene Diacid llé to Diene Diester 888 The same procedure used to convert llé to 881 was followed for llé, to give lgg in 95% yield. The nmr and ir spectra were identical with those of authentic sample 892' (1) (2) (3) (4) PART IV MISCELLANEOUS A NEW ALKYLATION REAGENT - HIGH SURFACE SODIUM SYNTHESIS OF POTENTIAL CARLINOGENIC DIEPOXIDES THE EFFECT OF METHYL GROUPS AT A BRIDGE-HEAD POSITION ON THE COMPETING CARBONIUM ION REARRANGEMENT OF THE BICYCLO[3.2.1]OCTA-3,6-DIEN-2-YL SYSTEM ALKYLATION STUDIES WITH 4-METHYLENE-2,3,5-TRIMETHYL -2-CYCLOPENTENONE 145 (1) A NEW ALKYLATION REAGENT - HIGH SURFACE SODIUM INTRODUCTION The term High Surface Sodium (HSS)63 is applied to films of sodium approaching colloidal dimensions, spread over inert solids of high surface area. Sodium in this form is advantageous for the preparation of finely divided metals, for the purification of hydrocarbons and ethers, and for the preparation of inorganic and organic sodium derivatives. In this part of the thesis, the use of HSS to generate ketone enolates which can then be alkylated is described. The ease and simplicity of generating HSS make it desirable to prepare the sodium in this form at the point of use. Although it seems likely that conditions could be developed for storing HSS without appreciable loss of activity, this has not yet been done. The preparation of HSS is accomplished simply by mixing molten sodium with suitable inert solid materials having very large surface areas. Many substances may be used as sodium carriers, e.g., salt, carbon, ceramic materials like aluminum oxide. At a temperature above its melting point (97.5°), sodium spontaneously coats the solid ma- terials. The effective surface area of the solid carrier de- termines the amount of sodium which can be absorbed. Salt 146 147 may carry 2 to 10% sodium; soda ash, 10% sodium; alumina, 20 to 25% sodium; and activated or colloidal carbons over 30% sodium. Within these concentrations a free flowing solid is obtained, whereas above these concentrations the mixture becomes a pasty mass. This free flowing charac- teristic can be maintained at a temperature up to the boiling point of sodium (883°C) depending, of course, on the temperature stability of the carrier. This important feature permits HSS to be handled in fluidized and other moving solid systems. Typical Applications of High Surface Sodium 1. Reduction of Metal Salts and Oxides to Colloidal Metals. Many metal salts and oxides react with sodium adsorbed on an inert carrier smoothly, rapidly and at relatively low temperatures. The metals produced are of colloidal dimen- sions, hence, extremely reactive. They are usually pyro- phoric when exposed to the air. Metal compounds which have been reduced by HSS systems include nickel chloride, copper chloride, lead chloride, chloroplatinic acid, silicon tetrachloride, titanium tetrachloride, zirconium tetra- chloride, iron oxide, zinc oxide and metal soaps such as nickel oleate. 148 2. Preparation of Catalysts for Hydrogenation. Nickel and platinum catalysts prepared by HSS tech- niques are comparable with Raney Nickel and Adams platinum catalyst in the hardening of olive oil. The result shows that finely divided nickel prepared by HSS reduction of nickel oleate produced faster hardening of olive oil than either Raney Nickel or nickel catalyst produced by thermal decomposition of the nickel formate. 3. Reduction of Metal Oxides. Zinc oxide, iron oxide, c0pper oxide, nickel oxide are reducible by sodium. 4. Hydrocarbon Refining The reduction of sulfur in certain petroleum naphthas to less than 0.001% were made by H88. 5. Ether Refining. HSS is especially effective in removing water, alcohols, acids, aldehydes, ketones, and peroxides quantitatively from ethers. Diethyl ether, 1,2-diethoxyethane, and tetra- hydrofuran have been purified in the vapor phase over HSS on soda ash. 149 6. Formation of Sodium Hydride. Sodium hydride with a particle size below 10 microns can be conveniently prepared in situ from H88 and hydrogen. 7. Preparation of Organosodium Derivatives. Many important sodium derivatives, difficult to prepare by classic methods, have been made commercially feasible through the use of sodium dispersions. For example, sodium in this form has proven useful in metalation reactions, Claisen condensation, Wurtz type reaction, reductions, catalytical polymerization and the preparation of alcohol- free aldoxides. Compounds more specifically illustrating the versatility of sodium dispersions are phenylsodium, amylsodium, sodium anilide, sodioacetoacetic ester and sodio-malonic ester. 8. Isomerization High surface sodium is effective for isomerization of alkenes. In contrast to acid-catalyzed isomerizations, no skeletal rearrangements occur. The results indicate some stereo-selectivity. Thus l-butene is isomerized initially to about equal amounts of alae and ananaf2-butene; .eventually the thermodynamic equilibrium mixture rich in the Enanafisomer is obtained. The catalyst was used to effect almost quantitative conversion of methylenecyclo- butane to l-methylcyclobutene. RESULTS AND DISCUSS ION The application of H88 in organic synthesis was limited to the alkylation of ketones. Such alkylations, when carried out in solution using conventional strong bases, often suffer from low to modest yields, lack of regiospecificity, considerable dialkylation and other problems which clearly limit synthetic utility. It was the hope that the reaction would be regiospecific and limited to monoalkylation using HSS, where the reactions might occur on a surface instead of in solution. HSS can be prepared66 readily in glass equipment employing a rugged sweep-type stirrer. This apparatus has proved satisfactory for a wide variety of prepara- tions. A typical procedure for the preparation of 20% sodium on charcoal is illustrated below: 1. Place 2.8 g of dry charcoal in the reaction flask and displace the air with dry nitrogen. Adjust the stirrer speed to 100-300 RPM for expanding the charcoal to approximately twice its settled volumn, and heat the flask contents to about 125°. 2. Add 0.69 g of sodium in 5 to 50 mg pieces through the charging port. As soon as the sodium has melted, stir somewhat more rapidly for about 5 min, then cool to room temperature. 3. Check the reactivity of the HSS by exposing a small sample to air. HSS on charcoal become 150 151 heated to redness due to the oxidation of both sodium and charcoal. Subsequent reactions of HSS may now be started. Add 50 ml of a previously dried solvent (hexane, cyclohexane, THF, benzene) to the flask. A solu- tion containing 25 mmoles of the appropriate ketone in 10 m1 of benzene was added with stirring. The reaction mixture, after being heated for 5 min at reflux, was cooled to room temperature and 3.55 g (25 mmol) of methyl iodide was added. The re- action mixture was refluxed for 2 hr, then allowed to come to room temperature. Workup consisted of addition to 95% ethanol, extraction with ether, and drying over MgSO4. In the case of hexamethyl-2,4-cyclohexadienone, alkyla- its ir and nmr spectra with those of an authentic sample. tion occurred via a cross - rather than a fully - conju- gated enolated anion, to give a single product 888' Un- reacted hexamethyl-Z,4-cyclohexadienone (50%) was also recovered. The product was identified by comparison of 67 HSS CH 888 H ’,/”I’ 123 r 'Vb’b B H +H+I alc. h; 888 Monoalkylation of B-tetralone is a key step in the synthesis of natural products such as (i) callitrisic acid,72 dehydroabietic acid,73 and resin acid.74 76 A rather long route has been reported75' to accomplish the mono- alkylation of B-tetralones in 80% yield. The conventional use of alkoxide and methyl iodide on 2-tetralone gave largely the dimethyltetralone,7o though monoalkylation product had been previously reported to form low yield (58%).71 Methylation zla the addition of 2-tetralone to HSS proceeded smoothly to yield solely the 1-alkylated derivatives. Preparative vpc (10' x 0.25 in column, 20% FFAP on chromosorb W, AW-DMCS 80/100, 148°, 60 ml/min He) gave 89% of the monoalkylation product (l23,ret time 30 min) and 10% of the dialkylation product (llg, 41 min). l-Methyl-Z-tetralone was identified by the comparison of its ir (neat, v 71,75,76 -1 c=o 1710 cm ) with that of the literature report. The mass spectrum showed a parent peak at m/e 160, and the nmr spectrum 153 (CC14) showed a doublet at 61.36 (J = 7 Hz) for the 1- methyl substituent. 1,1-Dimethyl-2-tetralone was identi- fied by the comparison of its mass, nmr, and ir spectra with those of an authentic sample. H58 0 B-Tetralone CH3I , + lgg(89%) lgg(10%) Research on the application of high surface sodium to the alkylation reaction is at an initial stage. It might be interesting to use sodium chloride (40-80 mesh) as the solid carrier in order to increase the sodium ef- ficiency for forming enolate and reduce the possibility of chemical reduction of the ketone (using only 2-10% sodium). This support showed also reduced sample loss in the workup process. Also, it might be interesting to compare the alkylation result with that of high dispersion 65 77 sodium or alkali metalgraphite intercalation compounds. (2) SYNTHESIS OF POTENTIAL CARCINOGENIC DIEPOXIDES A number of epoxides have been shown to be carcino- 78 genic. Simple examples include 127, 128, and 129. ’VVM "b’V‘b 'VVU A 888 888 888 A series of arene oxides such as 130 and 131, and other ’VVM ’VVM oxidized arene metabolites are also carcinogenic.79 888 888 In this part of the thesis, I describe the synthesis of the diepoxy ketone 132. The monoepoxide (24) had been ’V'h'b 'Vb previously prepared. Attempts to rearrange 83% by photolysis and by acid are also described. 154 155 888 RESULTS AND DISCUSSION The reaction of a,B-unsaturated ketones with per- acids6e usually does not lead to epoxidation of the double bond. Although exceptions to this generality are known, they are not common.80 Treatment of l3; with m-chloroperbenzoic acid in benzene afforded the monoepoxide 2%. Further treatment of 24 with m—chloroperbenzoic acid gave a single diepoxide 888' Compound 88% could also be obtained directly from MCPBA MCPBA v 0 O A. 888 88 888 l Excess MCPBA T 156 l3; by using excess oxidant. The gross structure of 888 was established by its spectral properties. The molecular formula C12H1803 was confirmed by the compound's mass spectrum (parent peak 0 (2.45) (2.3l)l.5 1.22 o (1.93) (1.00)l.52 1'30 (1.12)l.38 888 m/e 210) and microanalysis. The ir spectrum showed a carbonyl absorption at 1710 cm-1. The uv spectrum showed only end absorption. The nmr spectrum showed that all methyl signals appeared at or above 61.52. These spectros- copic data indicate that 888 must be a saturated compound. The stereochemical assignment is not completely certain yet. However, it should be pointed out that the epoxida- tion of 888 gives only the cis-diepoxide.81 I 888 88 157 Diepoxide prepared from l33* in which the C-3 methyl was replaced by a CD3 group lacked the singlet at 61.52. Diepoxide prepared from l33** in which both the C-3 and C-5 methyls were replaced by CD3 groups lacked the singlets at 61.52 and 1.30. Thus the nmr assignments shown on the structure are correct. Irradiation of l3; under a variety of conditions led only to recovered starting material. Treatment of 88% with TFA resulted in an interesting molecular rearrange- ment to form two isomeric products. However, additional work is necessary before the structures can be unambiguously characterized. The method described here is probably generally appli- cable to the synthesis of diepoxides with different sub- stituents in the cyclohexadienone system. EXPERIMENTAL 1. Synthesis of 2,3,4,5-Diepoxy-2,3,4,5,6,6-hexamethyl- 2,4-cyclohexadienone (8881 An ice-cold solution of 9.6 g (0.05 mol) of m-chloro- perbenzoic acid in 150 ml of carbon tetrachloride was added slowly to a solution of the monoepoxide 2% (4.85 g, 0.025 mol) in 50 m1 of carbon tetrachloride. Reaction was allowed to warm to room temperature and stirred for 1 hr. The nmr monitoring showed complete reaction at this time. The solvent was removed by rotary evaporation, petroleum ether (b.p. 30-60°) was added, and the m-chloro- benzoic acid was removed by filtration. The filtrate was washed with aqueous sodium bicarbonate, and saturated sodium chloride solution, dried (MgSO4), and evaporated to give 4.90 g of diepoxide 888 in 96% yield. The same product could be obtained directly from dienone using an excess (three or four fold) of oxidant. The crude product can be purified by fractional distillation (b.p. 82°/10'3 mm Hg) or by vpc (10! x 0.25 in column, 10% FFAP on chromosorb W, 80/100). Ir (neat) 3001 (m), 1710 (s), 1480 (m), 1400 (m), 1145 (w), 1122 (w), 1103 (m), 1080 (m), 940 (w), 890 (w), 870 (m), 830 (w), 800 (w), 780 (w) cm-l; uv (cyclohexane) showed only end absorption; nmr (CC14) 61.22 (s, 6 H), 1.30 (s, 3 H), 1.38 (s, 3 H), 1.50 (s, 3 H), 1.52 (s, 3 H); mass spectrum (70 eV) m/e 158 159 (rel intensity) 210 (5), 195 (3), 182 (5), 167 (45), 153 (85), 139 (47), 111 (24), 99 (41), 97 (40), 86 (20), 81 (25), 71 (ll), 69 (40), 55 (24). Anal. Calcd for C12H180 : C, 68.54; H, 8.63 3 Found: C, 68.48; H, 8.54 2. Synthesis of 2,3:4,5-Diep9xy-3-trideuteromethyl-2,4,5,6,6- hexamethyl-Z,4-cyclohexadienone (132:) An ice-cold solution of 1.92 g (0.01 mol) of MCPBA in 30 ml of carbon tetrachloride was added slowly to a solution of 3-trideuteromethyl-2,4,5,6,6-pentamethyl-2,4- cyclohexadienone (0.97 g, 0.005 mol) in 10 ml of carbon tetrachloride and stirred at room temperature for 10 hr, and worked up as described for the corresponding compound l32.48 The epoxidation product l32* had an nmr spectrum identical with that of 888 except that the signal at 61.30 had disappeared.48 3. Synthesis of 2,3:4,S-Diepoxy-B,5-bis-trideuteromethyl- 2,4,6,6-tetramethyl72,4—cyclohexadienone (l32**) The procedure was as described for the corresponding l32.48 The product had an nmr spectrum identical with that of 888 except that the signals at 61.30 and 1.52 had disappeared. 160 4. Irradiation of Diepox1delgg Irradiation of a 0.01 M_solution of 888 in ether through pyrex or quartz for 10-15 hr with a 450 W Hanovia lamp, gave only unchanged starting material. 5. Acid-catalyzed Rearrangement of Diepoxide leg A solution of the diepoxide 88% (100 mg, 0.48 mmol) in 3 ml of TFA, prepared at 0°, was allowed to stir for 15 min at room temperature (monitoring by nmr) then poured into a cold sodium bicarbonate solution and extracted several times with ether. The combined ether extracts were washed successively with saturated sodium bicarbonate solution, water, and saturated sodium chloride solution, and dried (M9804). After solvent removal there remained 95 mg of a light yellow oil. The residue, when subjected to preparative vpc (5' x 0.25 in column, 20% FFAP on chromosorb W, AW-DMCS 80/100, 60 ml/min He, 150°), gave two compounds with retention times of 4 min (40%), 8888’ and 8 min (60%) 8888' respectively. For 8888: ir (neat) 3500 (w), 3000 (s), 2980 (s), 2905 (m), 1738 (s), 1718 (s), 1685 (m), 1460 (m), 1395 (s), 1365 (s), 1305 (m), 1270 (m), 1203 (m), 1160 (m), 1135 (m), 1100 (w), 1058 1. (m) cm‘ , uv (methanol) Am 215 nm (e 2100), 250 (1050); ax nmr (CC14) 61.06 (s, 3 H), 1.09 (s, 3 H), 1.20 (s, 3 H), 1.32 (s, 3 H), 1.83 (s, 3 H), 2.03 (s, 3 H); mass spectrum (70 eV) m/e (rel intensity) 210 (5), 195 (5), 167 (32), 161 154 (10), 153 (90), 139 (10), 125 (20), 109 (4), 97 (7): 81 (15), 69 (10). For 888%: ir (neat) 3350 (m), 2960 (s), 2905 (m), 1705 (S), 1660 (m), 1650 (S), 1400 (S), 1350 (m), 1180 (m), 1100 (m), 1030 (m), 790 (s), 750 (m) cm“; uv (methanol) Amax 223 nm (E 1610), 276 (2030); nmr (CC14)4 61.06 (s, 3 H), 1.09 (s, 3 H), 1.29 (s, 3 H), 1.60 (s, 3 H), 2.00 (s, 3 H), 2.17 (s, 3 H); mass spectrum (70 eV) m/e (rel intensity) 210 (10), 195 (l), 167 (52), 137 (12), 125 (50), 113 (10), 99 (40), 97 (38), 69 (28). Neither product was fully characterized. (3) THE EFFECT OF METHYL GROUPS AT A BRIDGE-HEAD POSITION ON THE COMPETING CARBONIUM ION REARRANGEMENT OF THE BICYCLOE3.2.1]OCTA-3,6-DIEN-2-YL SYSTEM 54-57 observed degenerate Recently, Kuzuya and Hart rearrangements of the nonamethyl bicyclo[3.2.l]octa-3,6- dien-2-y1 cation 888' In strong acid, cation 888 under- goes three distinct types of rearrangements. They are listed in order of increasing activation energy. (a) Circumambulation 8 7 7 / = e 3 6 5 + 5 3 4 6 r 4 8 2 l = v— 7 +4 6) 8 l / 2 l 7 5 6 4 162 163 (b) 1,2-Bridge Shift (c) 1,3-Cyclopropy1carbinyl Shift 8 8 8 1. - 1 7 I 8 5 4 3 + 5 3 6 4 7 5 2 4 6 4 2 3 The fastest of these (k-80° 31.1 sec-1 . / , 65* 10.2 kcal/ mole, ASg -14.5 eu/mole, AHI 7.4 kcal/mole) was detected by changes in the nmr spectrum between -100° and -50° in FSOBH/SOZC1F. It equilibrates the two methyls at C-8 and methyls at C-2,3,4,6, and 7 but leaves the methyls at C-1 and 5 unique. A Circumambulation mechanism accounts for the results. The slower degenerate rearrangement of 888 (AH? > 7.4, but <17.8 kcal/mole) was concealed from nmr detection but was established by deuterium labeling experi- ments. It equilibrates the two methyls at C-8 and all the methyls at C-1 through C-7. A 1,2-bridge shift accounts 164 for the results. In the slowest process, carbocation lag (which in FSOBH/SOZClF undergoes rapid degenerate rearrange- ments below -60°) rearranges irreversibly above -60° to the nonamethylbicyclo[3.3.0]octa-3,6-dien-2-y1 cation. A 1,3-cyclopropylcarbinyl shift accounts for the results. Kuzuya and Hart also found that ion lag rearranged to ion lgéa below -50° by Circumambulation mechanism, and not I \‘t" RI? 30% RR Rék by successive 1,2-bridge shift, as proved by deuterium labeling experiments. Only at higher temperatures do the 1,2-bridge shifts (observed by deuterium scrambling) occur. Again, the slowest reaction was a 1,3-cyclopropylcarbinyl shift. Similarly, Circumambulation (nmr observed) of leek occurred faster than 1,2-bridge shifts and the slowest reaction was a 1,3-cyclopropylcarbinyl shift. Even though C-1 and C-5 remain unique in the circum- ambulation mechanism, positive charge develops at these carbons in various steps of the mechanism. This was clearly observed by Hart and Kuzuya in their study of ion lggb. It was therefore desirable to prepare and study the rearrangements of the ion with hydrogens at both 165 bridgehead positions, and methyl groups at every other position. A preliminary study of this system is described here. RESULTS AND DISCUSSION The scheme used to synthesize ion $39 is summarized below. 0 H H + CH c=CCH 210° LiAlH 3‘ 3-———)/ —_4.’/ 3 days 90% H H 12% R1 Ré + H H trace H FSOHz/SOZCIF / aq. acetone' I -lO° H H 173% MR The Diels-Alder addition of 2-butyne to the penta- methyldienone g2 gave adduct lax in 93% yield. Compound lg; was assigned the structure shown on the basis of its spectral properties and chemical transformations. The molecular formula C H 0 was confirmed by the mass 15 22 spectrum (parent peak m/e 218) and elemental analysis. 166 O.90(2.60) O 1.21(3.13) 1.60(1.19) 1 50(1 19) / 1.75(1.00) 2.§5 1.75(1.oo) (1.72) RI The ir spectrum of 1;; had a strong band at 1690 cm-1. and the uv spectrum showed a Ama at 218 nm (e 1700) con- x sistent with bicyclo[2.2.2]octadienone system. The nmr spectrum is consistent with the Cs symmetry and showed a gem-dimethyl group at 60.90, a bridgehead methyl group at 61.21, four allylic methyl groups at 61.60 and 1.75, and a single bridgehead proton at 62.75. The europium shift data were also consistent with the structure. Reduction of leg with lithium aluminum hydride gave alcohol 138. The structure of 188 is based on spectral properties and chemical transformations. The ir Spectrum showed a broad hydroxyl group absorption at 3200-3700 cm'l. The nmr spectrum showed that the three highest field methyl proton peaks were sharp singlets, as were the two protons at 62.30 and 2.80. The signals at 61.52, 1.60, 1.80, and 1.83 showed homoallylic coupling and the europium shift data clearly delineated the correct nmr assignments. 167 2.80(10.00) H OH .26(5.60) 1.60(2.30) 0.72(3.85) 0.86(2.70) 1.52(1.54) 1.33(1.00) H 1.80(l.31) 2.30(3.00) $38 Treatment of 158 with a trace of acid in acetone caused it to dehydrate with rearrangement to give triene 152. The structure of leg follows from its spectral properties. The two highest field gem-dimethyl peaks H 2.35 (br) 4.62(m) 1.68 1&2 64% were sharp singlets and the peaks at 61.50, 1.58, and 1.78 showed homoallylic coupling; other peaks were singlets or were split as shown on the structure. The nmr spectrum of 131 is shown for comparison, and the close agreement in chemical shifts is apparent. When 133 was treated at -78° with FSO3H/SOZC1F, the 168 stable carbocation 138 was obtained. The nmr spectrum is consistent with the Cs symmetry. It had two 6—proton singlets, one 2-proton singlet, and three 3-proton singlets, assigned as shown on the structure. The only arbitrary feature of the assignment is at C-6, where we can not say with certainty whether the lowest field methyl is syn: or anti to the positively charged bridge. The nmr spectrum of carbocations 135 and lgéb are shown for comparison. H 3.53 Ré . 13819 Carbocation 139 does not undergo either Circumambula- tory rearrangement or 1,2-bridge shifts at -78°. Indeed, the nmr spectrum did not change until the temperature was raised to -lO°. Above -10°, the changes occurred which appeared to be irreversible; however, insufficient material was on hand to investigate the reaction further. EXPERIMENTAL 1. Synthesis of 1,2,3,5,6,8,8—Heptamethyl-bicycloL2.2.21 octa-2,5-dien-7-one (%QZL A mixture of 1 g of 2,4,5,6,6-pentamethy1-2,4-cyclo- hexadienone (4.58 mmol) and an equal volume of 2-butyne were heated in a thick-walled sealed glass tube at 203° for 3 days. Excess 2-butyne was allowed to evaporate from the cooled reaction mixture, and the viscous residue was chromatographed over silica gel (<230 mesh) using methylene chloride as the eluent, to give 1.03 g of 137 (93% yield) as a pale yellow oil. Analytical Vpc (5' x 0.125 in column, 20% FFAP on chromosorb W, AW-DMCS 80/100, 162°, 30 ml/min N2)showed a retention time of 5 min. Preparative vpc gave pure 131: ir (neat) 2905 (s), 2895 (s), 1690 (s), 1440 (s), 1370 (s), 1350 (w), 1300 (w), 1240 (w), 1200 (w). 1145 (m), 1060 (m), 1020 (m), 1003 (s), 870 (m) cm'l; uv (ethanol) Amax 218 nm (e 1630), 245 (shoulder, 390); nmr (CC14) 60.90 (s, 6 H), 1.21 (s, 3 H), 1.60 (broad s, homoallylic coupling, 6 H), 1.75 (broad s, 6 H), 2.75 (s, 1 H); mass spectrum (70 eV) m/e (rel intensity) 218 (10), 203 (13), 175 (11), 160 (3), 149 (53), 148 (100), 134 (45), 133 (85), 119 (20), 117 (20), 115 (18), 105 (32), 91 (42). 77 (27). Anni. Calcd for C H O: C, 82.15; H, 10.16 15 22 Found: C, 82.16; H, 10.18 169 170 2. Synthesis of l,3,3,5,6,7,8-Heptamethyl-bicyclo[2.2.2] octa-5,7-dien-2-ol (138) To a suspension of 0.57 g of LiAlH4 in 25 ml of an- hydrous ether, there was added 1 g of 137 in 25 ml of an- hydrous ether. The mixture was stirred for 8 hr at room temperature. Excess hydride was destroyed by adding water, the ether layer and extracts were washed with saturated sodiumchloride solution, dried (M9804). The solvent was evaporated to give 138 as a pale yellow oil in virtually quantitative yield. The crude product was chromotographed over silica gel using pentanezether (5:1) as eluent to give a colorless oil (138). Analytical vpc (5' x 0.125 in column, 20% FFAP on chromosorb W, 150°, 30 ml/min He) gave two peaks. The retention time of 138 was 7 min. Preparative vpc (10' x 0.25 in column, 20% SE-30 on chromo- sorb W, AW-DMCS 80/100, 140°, 60 ml/min He) showed the retention time of 138 was 20 min. Another peak, retention time 30 min, was identified as due to pentamethylbenzene (82% yield) due to a thermal fragmentation reaction. For 138: ir 3500 (s), 3000 (s), 2950 (s), 1490 (w), 1460 (s), 1400 (w), 1390 (w), 1130 (w), 1250 (s), 1100 1 (w) cm- ; uv (ethanol) Am 215 nm (E 2860); nmr (CC14) ax 60.72 (s, 3 H), 0.80 (s, 3 H), 1.26 (s, 3 H), 1.52 (broad singlet, 3 H), 1.57 (broad singlet, 3 H), 1.60 (broad singlet, 3 H), 1.66 (broad singlet, 3 H), 1.77 (broad singlet, 3 H), 2.29 (s, l H), 2.77 (s, l H), 2.80 (broad, 171 1 H, exchanged with D20); mass spectrum (70 eV) m/e (rel intensity) 220 (l), 202 (10), 187 (23), 172 (8), 149 (20), 148 (90), 133 (100), 119 (9), 107 (8), 105 (10), 91 (15), 77 (8). Since purification through chromatography was accom- panied by the formation of pentamethylbenzene, no attempt was made to perform an elemental analysis. 3. Synthesis of 2-Methy1ene-3l4,6,7,8,8-hexametnylbicyglo [3.2.1]octa-3,6-diene (139) A solution of 138 (50 mg, 17.85 mmole) in 0.5 ml of 0.1% hydrochloric acid diluted with 1 ml of acetone was allowed to stand for 15 hr at room temperature. The mix- ture was diluted with water and extracted with methylene chloride. The combined extracts were washed with 5% sodium bicarbonate, water and dried (Nazso4). Preparative vpc (10' x 0.25 in column, 20% FFAP on chromosorb W, 160°, 60 ml/min He, ret time 1.5 min) gave 132 (41%) as a color- less oil: ir (neat) 3025 (w), 2900 (s), 1638 (s), 1605 (s), 1485 (m), 1468 (s), 1420 (w), 1398 (s), 1390 (s): 1342 (w), 1320 (m), 1310 (w), 1276 (m), 1260 (w), 1178 (w), 1158 (m), 1100 (w), 980 (w), 898 (s), 798 (m) cm’l; uv (ethanol) Ama 212 nm (e 3960), 246 (8680); nmr (CC14) x 60.80 (s, 3 H), 1.00 (s, 3 H), 1.50 (broad singlet, 3 H), 1.58 (broad singlet, 6 H), 1.73 (broad singlet, 3 H), 1.80 (s, 1 H), 2.26 (s, l H): mass spectrum (70 eV) m/e (rel 172 intensity) 202 (57), 187 (100), 172 (32), 157 (15), 145 (30), 133 (15), 119 (10), 105 (10), 91 (12), 77 (10), 44 (40). Anal. Calcd for C15H22: C, 89.04; H, 10.96 Found: C, 88.90; H, 11.03 4. Synthesis of Heptamethylbicyclo[3.2.llocta-3,6-dien- ijl cation_(1égl Approximately 50 pi of FSO3H was placed in a 5 mm- diameter nmr tube, cooled to -78°, and approximately 200 ul of $0201F was condensed above the acid. A CD2C12 solution containing 30 mg of the precursor of compound 132 was placed in the tube above the SOZC1F layer and allowed to stand at -78° for 5 min. The contents of the nmr tube were then mixed using a "supermixer" (Matheson Scientific, Cat. No. 60100-05) to give a purple solution of cation 148. For its nmr spectrum see structure. Carbocation ion spectra were obtained on a Varian Associate A50-60 equipped with a variable temperature probe. The temperature control was calibrated with a methanol standard sample. The carboca- tion 140 was held at-78° for several hours with no change in the spectrum. When the solution was warmed to -10°, the nmr spectra still remained the same, but above -10°, the nmr spectrum became complicated immediately. The solu- tion was cooled to -78° and quenched at that temperature with excess sodium bicarbonate in methanol. There was not enough sample for analysis of the products. (4) ALKYLATION STUDIES WITH 4-METHYLENE- 2,3,S-TRIMETHYL-2-CYCLOPENTENONE The acid-catalyzed rearrangement of epoxyketone 24 can be useful for the synthesis of cyclopentenones, some of which might serve as precursors to certain antibiotic 82 and prostaglandins. Therefore, it was of compounds considerable interest to study the reactions of the dienone 81. O O O H D 0 .0 0 con £4 8% 14% This compound could form two different enolates: 6438 6638 Each of these enolates has extended conjugation, and it is not clear a priori which would be formed preferentially, nor is it readily predictable when each of these enolates would react with an alkylating agent. This final portion of the thesis presents a preliminary 173 'lI-l 1‘ I'll. l . 174 study of alkylation reactions in this system. RESULTS AND DISCUSS ION A tetrahydrofuran (THF) solution of 81 was added at 0° to base prepared by adding hexamethyldisilazane to a solu- tion of n-butyllithium. After 1 hr, excess methyl iodide was added, and the mixture was stirred for 8 hr at room temperature, then worked up to give a 92% yield of 144. Alkylation of 14% following the same procedure or using potassium hydride as the base gave methylation product 136 in approximately 90%. Expected alkylation at a-position 0 6‘6? leading to exocyclic diene 145 was not observed. The unexpected but facile alkylation at the y-position may be a consequence of steric control. Further methylation of 146 by the same procedure gave 142 as the major product, and 148 and 139 as minor products. ll! I} I’llll .II 175 ' . - + 0 0 ' IIKCH3)esl]2N Li KH o *M—8 CH3I CH3I 3% 844(92%) 14g<903) + [(CH3)3Si]2N Li CH3: 0 / 141(55%) 148(15%) 149(18%) Alkylation of 81 at the position also occurred with allyl bromide. Two products, 158 and 151 were obtained (in 66% and 20% yield respectively) on treatment of 81 with lithium hexamethyldisilazane and an excess of allyl bromide. ‘7 [(CH3)3Silzn'Li+ 8 0‘4. 0 CH2=CHCHzBr $§Q(GG%) 151(208) llllll'l' 176 Treatment of 81 with methyllithium followed by de- hydration gave the triene 153 in 40% yield. There was no evidence for the formation of tetramethylfulvene in this reaction. HO CH3 H CH I H H o 3 . 8 ,0 #9 18% kéé“°%’ Compound 14% was assigned the structure shown: 0 (br,3H)1.80 ‘ 1.10(S,6H) 5.10(s,1H) (brl3H)2010 5.02(S,1H) 144 The molecular formula C10H140 was confirmed by the mass spectrum (parent peak m/e 150) and elemental analysis. The ir spectrum of 144 showed a strong carbonyl absorp- tion at 1701 cm‘1 and its uv spectrum had a maxima at 275 nm (a 15,000) in ethanol. The nmr spectrum is sum- marized on the structure. Compound 146 was assigned the structure shown. The molecular weight was confirmed by the mass spectrum (parent peak m/e 164). The ir spectrum of 136 showed a strong 177 o (s,3H)l.80 O 1.09(S.6H) (q,2H)2.49 4.92(s,1H) J=7 Hz [(t,3H)l.15 5-09(s.1H) 146 ’Vb'b carbonyl absorption at 1701 cm”1 (cyclopentenone) and its uv spectrum had maxima at 275 nm (8 15,200) in ethanol. The nmr spectrum is summarized on the structure. Compound 14% was assigned the structure shown 0 (s,3H)l.83 . 1.00(s,6H) 2.89-3.32 H J=7 Hz [ (q,lH) 4.80(s,1H) (d,6H)l.26 S.l7(s,lH) «1,4,1 The molecular formula C12H180 was confirmed by the mass spectrum (parent peak m/e 178) and elemental analysis. The ir spectrum of 141 showed a strong carbonyl absorption at 1701 cm-1 (cyclopentenone) and its uv spectrum had maxima at 275 nm (8 19,200) in ethanol. The nmr spectrum is summarized on the structure. Compound 148 was assigned the structure shown. The molecular weight was confirmed by the mass spectrum (parent peak m/e 178). The nmr spectrum is summarized on the 178 O (s,6H)1.11 . 1.o9(s.6H) 5.32-5.60 / .08(s,lH) J=7 Hz [ (q,1H) 5.12(s,1H) l.87(d,3H) structure. The stereochemical assignment is based on the nmr magnetically anisotropy effect. Compound 149 was assigned the structure shown 0 (s,6H)1.20 . 1.06(S,6H) 5066-6023 I 4060(8’1H) J=7 Hz [ (q,1H) 5.15(s,1H) 1.75(d,3H) «1,42 The molecular weight was confirmed by the mass spectrum (parent peak m/e 178). The nmr spectrum is summarized on the structure. The stereochemical assignment is based on the nmr magnetically anisotrOpy effect. Compound 150 was assigned the structure shown 0 2.20(m,2H) 4.60-4.95(m,3H) (br,3H)1.75 \ 4095 SIIH (br,3H)2-03 5.15:8,1H; RR 179 The molecular formula C12H16O was confirmed by the mass spectrum (parent peak m/e 176) and elemental analysis. The ir spectrum of 150 showed a strong carbonyl absorption at 1693 cm"1 (cyclopentenone) and its uv spectrum had maxima at 278 nm (8 21,000) in ethanol. The nmr spectrum is summarized on the structure. Compound 151 was assigned the structure shown 0 1J71-2.16(m,2H) \ 3.83-5.05(m,3H) 3.83-S.05(m,2H) 1.71-2.16(m,4H) (s,3H)l.69 151 mmm The molecular formula C15H200 was confirmed by the mass spectrum (parent peak m/e 216) and elemental analysis. The ir spectrum of 151 showed a strong carbonyl absorption at 1698 cm"1 (cyclopentenone) and its uv spectrum showed a maxima 278 nm (a 16,800) in ethanol. The nmr spectrum is summarized on the structure. Compound 15% was assigned the structure shown J=7 Hz fl (q,1H)3.20 H1.06(d,3H) 4.52(br,2H) 4.72(br,2H) (s,6H)l.80 68% 180 The molecular weight was confirmed by the mass spectrum (parent peak m/e 134). The ir spectrum of 888 showed a strong carbon-carbon double bond absorption at 1620 cm-1 and its uv spectrum had maxima at 277 nm (e 6000) in ethanol. The nmr spectrum is summarized on the structure. EXPERIMENTAL l. Alkylation of 4-Methy1ene-2,3,5-trimethy1-2-cyclo- pentenone (88) A tetrahydrofuran (THF) solution of 88 (400'mg, 2.94 mmol) was added at 0° to base prepared by adding hexa— methyldisilozane (550 mg, 3.4 mmol) to a solution of n- butyllithium (2 m1 of 2.06 §.in hexane + 10 m1 of THF). After a half hour, methyl iodide (450 mg, 3.1 mmol in 2 ml THF) was added, and the mixture was stirred for 7 hr at room temperature. The reaction mixture was poured onto cracked ice and extracted with ether. Evaporation of the solvent left 4.07 mg (92%) of a light yellow oil which, when subjected to analytical vpc (5' x 0.125 in column, 20% SE-30 on chromosorb W, AW—DMCS 80/100, 30 ml/min N2, 150°) showed one peak corresponding to 888 (retention time 2 min). Preparative vpc (10' x 0.25 in column, 10% FFAP on chromosorb W, AM-DMCS 80/100, 60 ml/min He, 140°, ret time 4 min) gave pure 4-methylene-2,3,5,5-tetramethy1-2- cyclopentenone (888): ir (neat) 2900 (s), 2820 (s), 2720 (m), 1701 (s), 1620 (s), 1500 (m), 1380 (m), 1300 (m), 181 1180 (s), 1100 (s), 880 (s), 860 (s) cm'l; uv (ethanol) Amax 208 nm (e 7050), 217 (15,000); nmr (CC14) 61.10 (s, 6 H), 1.80 (homoallylic coupling, 3 H), 2.10 (homoallylic coupling, 3 H), 5.02 (s, 1 H), 5.10 (s, 1 H); mass spectrum (70 eV) m/e (rel intensity) 150 (42), 135 (15), 122 (27), 107 (100), 105 (14), 91 (45), 79 (30), 65 (17). Anal. Calcd for C10H14o: C, 79.95; H, 9.39 Found: C, 80.00; H, 9.36 The same methylation procedure, but with an excess of methyl iodide (600 mg) being used, gave a mixture of compounds 888 (70%) and 888 (30%). 2. Alkylation of 4-Methylene-2,3,5,5-tetramethyl-2-cyclo- pentenone (8881 A. The same methylation procedure was followed for the alkylation of 888. Preparative vpc (10' x 0.25 in column, 10% FFAP on chromosorb W, AW-DMCS 80/100, 140°, 60 ml/min He, ret time 13 min) gave 4-methy1ene-3-ethyl- 2,5,S-trimethyl-Z-cyclopentenone (888): ir (CC14) 2950 (s), 2859 (m), 1701.(s), 1640 (m), 1420 (m), 1350 (m), 1295 (w), 1250 (m), 1005 (m), 900 (s) cm’l; uv (ethanol) Amax 208 nm (e 7100), 275 (15,200): nmr (CC14) 61.09 (s, 6 H), 1.15 (t, 3 H, J = 7 Hz), 2.49 (q, 2 H, J = 7 Hz), 1.80 (s, 3 H); mass spectrum (70 eV) m/e (rel intensity) 164 (100), 149 (42), 136 (35), 135 (63), 121 (74), 107 (60), 105 (20), 93 (52), 79 (40), 78 (10), 77 (38). 182 B. A weighed sample of potassium hydride dispersion (4 g, 0.02 mol, 20% excess) was placed in a flask equipped with a magnetic stirring bar, condenser, and injection port capped with a rubber sleeve stopper. The apparatus was purged with nitrogen and connected through traps to a gas-measuring device. The oil was removed with pentane. To the dry potassium hydride was added 20 ml of THF and a solution of 888 (2.46 g) in 10 ml of THF at 0°. After a half hour, methyl iodide (2.5 g in 2 m1 of THF) was added and the mixture was stirred for 7 hr at room temperature. Worked up as described for Experiment 1 gave 2.4 g (90%) of 888- C. The same methylation procedures, but with an excess of methyl iodide (600 mg) being used, gave a mixture of 146 and 147. mmm mmm 3. Alkylation of 4-Methylene-3-ethyl-2,5,5-trimethyl-2- cyclopentenoneAj888l The same procedure as described in Experiment 1 was followed for the alkylation of 888. Preparative vpc (10' x 0.25 in column, 20% FFAP on chromosorb W, AW-DMCS 80/100, 140°, 60 ml/min He), gave 888 (15%), 888 (18%), and 888 (55%) with retention times 7, 10, and 16, min, respectively. For 888: ir (neat) 3000 (m), 2900 (m), 1701 (s), 1640 (w), 1605 (m), 1475 (s), 1385 (s), 1260 (m), 1210 (w). 1090 (m), 1060 (s), 1010 (w), 900 (w), 800 (w) cm'l; uv 183 (ethanol) Am 275 nm (6 19,200); nmr (CC14) 61.00 (s, 6 ax H), 1.26 (d, 6 H, J = 7 Hz), 1.83 (s, 3 H), 2.89-3.32 (q, 1 H, J = 7 Hz), 4.80 (s, l H), 5.17 (s, l H); mass spectrum (70 eV) m/e (rel intensity) 178 (100), 164 (27), 163 (77), 149 (20), 148 (12), 136 (30), 135 (9), 121 (27), 119 (12), 105 (52), 103 (39), 93 (25), 91 (31), 81 (25), 79 (20), 77 (35), 67 (17). Anal. Calcd. for C O: C, 80.85; H, 10.18 12H18 Found: C, 80.61; H, 10.27 The small amounts of 888 and 888 isolated were only sufficient for mass and nmr spectra. For 888: nmr (CC14) 61.09 (s, 6 H), 1.11 (s, 6 H), 1.87 (d, 3 H, J = 7 Hz), 5.08 (s, 1 H), 5.12 (s, 1 H), 5.32-5.60 (q, 1 H, J = 7 Hz); mass spectrum (70 eV) m/e (rel intensity) 178 (64), 163 (8), 135 (100), 121 (41), 119 (19), 107 (62), 93 (35), 79 (30), 77 (33). For 888: nmr (CC14) 61.06 (s, 6 H), 1.20 (s, 6 H), 1.75 (d, 3 H, J = 7 Hz), 4.60 (s, l H), 5.15 (s, 1 H), 5.66-6.23 (q, 1 H, J = 7 Hz); mass spectrum (70 eV) m/e (rel intensity) 178 (62), 150 (11), 135 (100), 121 (43), 119 (15), 108 (20), 107 (59), 105 (15), 93 (31), 91 (25), 78 (28). ' 4. Alkylation of88 with Allyl Bromide The same procedure as described in Experiment 1, except that allyl bromide was the alkylating reagent, was followed. Preparative vpc (10' x 0.25 in column, 20% FFAP on chromosorb 184 W, AM-DMCS 80/100, 150°, 60 ml/min He) gave pure 888 (66%) and 888 (20%). For 888: ir (neat) 2903 (m), 1693 (s), 1613 (m), 1606 (m), 1450 (w), 1433 (W), 1397 (m), 1321 (W), 1285 (W): 1211 (w), 1022 (m), 1003 (w), 910 (s), 903 (m) cm'l; uv (95% ethanol) Am 208 nm (6 420), 278 (21,000); nmr (CC14) ax 61.01 (s, 3 H), 1.75 (s, homoallylic coupling, 3 H), 2.03 (s, homoallylic coupling, 3 H), 2.20 (m, 2 H), 4.60-4.95 (m, 4 H), 5.15 (s, 2 H); mass spectrum (70 eV) m/e (rel intensity) 176 (100), 119 (23), 107 (65), 105 (43), 91 (74), 79 (38), 77 (30), 65 (20), 53 (29). nnnl. Calcd for C12H16O: C, 81.77; H, 9.15 Found: C, 81.54; H, 9.08 For compound 888: ir (neat) 3000 cm"1 (m), 1698 (s), 1630 (w), 1618 (m), 1460 (w), 1440 (w), 1400 (w), 1380 (w), 1320 (w), 1260 (w), 1105 (w), 920 (s), 900 (m) cm'l; uv (ethanol) xma 213 nm (e 6000), 278 (16,800); nmr (CC14) x 61.01 (s, 3 H), 1.69 (s, 3 H), 1.71-2.16 (m, 6 H), 3.83- 5.05 (m, 6 H); mass spectrum (70 eV) m/e (rel intensity) 216 (27), 201 (32), 188 (22), 176 (23), 175 (100), 174 (45), 173 (32), 161 (32), 159 (33), 147 (90), 131 (40), 119 (60), 105 (25), 91 (63), 77 (32). nnnl. Calcd for ClSHZOO: C, 83.28; H, 9.32 Found: C, 83.41; H, 9.39 185 5. Treatment of Compound 81 with Methyl Lithium To a 50-ml three necked flask equipped with a condenser, serum cap, magnetic stirrer and nitrogen gas inlet was added 1.5 g (0.011 mmol) of compound 81 in 20 m1 of an- hydrous ether. The solution was cooled to approximately 0°. There was added, by means of a syringe, 8 m1 (excess amount) 1.50 g solution of methyllithium in ether. The reaction mixture was stirred under nitrogen at room tem- perature for 2 hr., then quenched with saturated NH4c1 solution and extracted with ether. The ether extracts were combined, washed with water and saturated NaCl solu- tion and dried over anhydrous MgSO4. After evaporation of the solvent, the residue was subjected to preparative vpc (10' x 0.25 in column, 20% Carbowax on chromosorb W, AW- DMCS 80/100, 140°, 60 ml/min He, ret time 14 min) to give pure 16% in 40% yield. For 18%: ir (neat) 3400 (w), 2900 (s), 1620 (s), 1450 (s). 1380 (s), 1020 (w), 960 (s); uv (ethanol) Xma 277 nm (e 6000); nmr (CC14) 61.06 (d, x 3 H, J = 7 Hz), 1.80 (s, 6 H), 2.70-3.20 (q. 1 H, J = 7 Hz), 4.52 (br s, 2 H), 4.72 (br s, 2 H); mass spectrum (70 eV) m/e (rel intensity) 134 (50), 120 (10), 119 (100), 103 (6), 91 (25), 74 (12). II 3." 1. It I BIBLIOGRAPHY 1 I I'll! 'Ill. (11: 10. 11. 12. REFERENCES (a) H. Hart and P. B. Lavrik, J. Org. Chem., 9 , 1798 (1974); (b) H. Laurent, G.“SchuIz, and R. iechert, Chem. Ber., 199, 2570 (1969). Y. R. Naves, O. Schwarzkopf and A. D. Lewis, Helv. Chim. Acta., 99, 880 (1947). (a) H. Sturzinger and P. Karrer, Helv. Chim. Acta., 99, 1829 (1946); (b) Y. Suhara and—T: Minami, Bull. Chem. Soc. Jap., 99, 168 (1966); (c) S. Akagi and K. Tsuda, Chem. Pharm. Bull. (Tokyo), 9, 464 (1961); (d) L. H. Knox, J. A. Zderic, J. P. Ruelas, C. Djerassi and H. J. Ringold, J. Amer. Chem. Soc., 99, 1230 (1960). (a) P. A. Diassi, S. D. Levine and R. M. Palmere, J. Med. Chem., 19, 551 (1967); (b) E. Farkas and J. M. Owen, J. Med. Chem., 9, 510 (1966). H. Hart, 1. Huang and P. B. Lavrik, J. Org. Chem., 99, 999 (1974). (a) D. Swern, Chem. Rev., 99, l (1949); (b) D. Swern, Org. Reactions, 2, 378 (1953); (c) D. Swern in D. Swern, Ed.,“Organic Peroxides", Vol. 2, Wiley-Interscience, New York, NY, 1971, Chapter 5; (d) H. 0. House, "Modern Synthetic Reactions", 2nd Ed., W. A. Benjamin, Menlo Park, CA., 1972, Chapter 6; (e) M. S. Malinovskii, "Epoxides and Their Derivatives", Daniel Davey, New York, NY, 1965. H. Hart and Eng-mu Shih, J. Org. Chem., 99, 3377 (1976). R. J. Gritter and E. C. Sabatino, J. Org, Chem., 99, 1965 (1964). (a) R. Gomer and W. A. Noyes, Jr., J. Amer, Chem. Soc., 99, 101 (1950); (b) M. K. Phibbs, B. de B. Darwent and E. W. R. Steacie, J2 ChemI Phys., 19, 39 (1948); (c) R. J. Cvetanovic, Can. J, Chem., 33, 1684 (1955). E. E. van Tamelen and D. Carty, J.yAmer. Chem. Soc., 99, 3922 (1967). J. K. Crandall and D. R. Paulson, Tetrahedron Letters, 2751 (1969). W. Kirmse, "Carbene Chemistry", Academic, New York, NY, 1964, p. 87. 186 III‘illll-lll) Ill-l IIJJ‘I (1‘. I‘ll I 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 187 (a) H. Kristinsson and G. W. Griffin, Angew. Chem. Intl. Ed. Engl., 9, 868 (1965); (b) R. S. Becker, J. Kolc, R. O. Bost, N. R. Bertoniere, R. L. Smith and G. W. Griffin, J. Amer. Chem. Soc., 9, 1302 (1970); (c) A. M. Trozzolo, W. A. Yager, G. . Griffin, H. Kristinsson and I. Sarkar, J. Amer. Chem. Soc., 99, 3357 (1967). K. Tokumaru, Bull. Chem. Soc. Jap., 99, 242 (1967). (a) R. Bertoniere and G. W. Griffin, Org. Photochem., , 115 (1973); (b) A. Padwa, Org. Photochem., 9, 91 1967); (c) O. Jerger, K. Schaffner and H. Wehrli, Pure Appl. Chem., 9, 555 (1964). H. Hart and P. M. Collins, J. Chem. Soc., (C), 895 (1967). R. K. Murray, Jr., T. K. Morgan, Jr., J. A. Polley, C. A. Andruskiewicz, Jr., and D. L. Goff, J. Amer. Chem. Soc., 99, 938 (1975). H. Hart and Eng-mu Shih, Org. 212 172nd ACS National Meeting, San Francisco, California, Aug. 29-September 3, 1976. (a) M. Debonon, R. M. Malloy, D. Bauer, T. Iizuka, K. 9chaffner and O. Jeger, J. Amer. Chem. Soc., 99, 420 1970). (b) D. Bauer, T. Iizuka, K. Schaffner and O. Jeger, Helv. Chim. Acta., 99, 852 (1972). (c) J. A. Saboz, T. Iizuka, H. Wehrli, K. Schaffner and O. Jeger, Helv. Chim. Acta., 99, 1362 (1968). A. P. Alder and H. R. Wolf, Helv. Chim. Acta., 99, 1048 (1975). M. Debono, R. M. Malloy, D. Bauer, T. Iizuka, K. Schaffner and O. Jeger, Helv. Chim. Acta., 99, 1948 (1973). B. R. von Wartburg, H. R. Wolf and O. Jeger, Helv. Chim. Acta., 99, 852 (1972). E. M. Shih, Ph.D. Thesis, Michigan State University, 1975. E. Pesch and S. L. Friess, J. Amer. Chem. Soc., 99, 5756 (1950). T. Matsuura and K. Ogura, J. Amer. Chem. Soc., 99, 3850 (1967). 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 188 C. S. Markos and W. Reusch, J. Amer. Chem. Soc., 99, 3363 (1967). G. F. Barkinshaw, B. R. Davis and P. D. Woodgate, g. Org. Chem., 99, 1607 (1970). T. Matsurra and K. Ogura, Tetrahedron, 99, 6167 (1968). T. Matsurra and K. Ogura, Qhem. Comm., 1247 (1967); J. Amer. Chem. Soc., 99, 3850 (1967). (a) H. Pines, V. N. Ipatieff, J. Amer. Chem. Soc., 99, 1076 (1939); (b) K. Alder, F. H. Flock and H. essenich, Chem. Ber., 99, 1709 (1957). W. Skorianetz, K. H. Schulte-Elte, and G. Ohloff, Helv. Chim. Acta., 99, 1913 (1971). N. Harada, S. Suzuki, H. Uda, and H. Ueno, J. Amer. Chem. Soc., 99, 1777 (1972). N. Harada, S. Suzuki, H. Uda and H. Ueno, Chem. Lett., 803 (1972). (a) N. Harada, H. Uda, H. Ueno, and S. Utsumi, Chem. Lett., 1173 (1973); (b) H. Stetter and H. J. Sandhagen, Chem. Ber., 999, 2837 (1967). K. Sato, H. Hagiwara, H. Uda, M. Sato, N. Harada, g. Amer. Chem. Soc., 99, 8281 (1976). T. Okada, K. Kamogawa, M. Kawanisi and H. Nozaki, Bull. Chem. Soc. Jap., 99, 2908 (1970). For review see (a) S. S. Hixson, P. S. Mariano and H. E. Zimmerman, Chem. Rev., 99, 531 (1973); (b) K. N. Houk, ibid, 99, l (1976). N. J. Turro, Account. Chem. Res., 25 (1969). N. J. Turro and W. B. Hammond, Tetrahedron, 99, 6029 (1968). H. G. Richey, Jr., J. M. Richey, and D. C. Clagett, J. Amer. Chem. Soc., 99, 3906 (1964). R. K. Murray, Jr., Ph.D. Thesis, Michigan State Univer- sity, 1968. H. Hart and G. M. Love, J. Amer. Chemt_Soc., 99, 6266 (1971). 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. S6. 57. 58. 59. 60. 189 L. Barber, 0. L. Chapman and J. D. Lassila, J. Amer. Chem. Soc., 99, 5933 (1968). (a) R. B. Parker and N. S. Isaac, Chem. Rev., 99, 737 (1959); (b) S. Winstein and R. B. Henderson in . C. Elderfield, Ed., "Heterocyclic Compounds", Vol. 1, Wiley-Interscience, New York, NY, 1950, pp 1-60. E. L. Eliel and D. W. Delmonte, J. Amer. Chem. Soc., B. Tchoubar Compt. Rend., 999, 117 (1942). H. Hart and E. M. Shih, J. 0:9. Chem., 99, 1128 (1975). H. Hart, P. M. Collins and A. J. Waring. J. Amer. Chem. Soc., 99, 1005 (1966). W. Skorianetz and G. Ohloff, Helv. Chim. Acta., 99, 2439 (1974). K. L. Stevens and R. Lundin, Tetrahedron, 99, 2749 (1975). W. D. Emmons, A. S. Pagano and J. P. Freeman, J. Amer. Chem. Soc., 99, 3472 (1954). A. C. G. Gray, T. Kakihana, P. M. Collins, and H. Hart, J. Amer. Chem. Soc., 99, 4556 (1967). A. C. G. Gray and H. Hart, J. Amer. Chem. Soc., 99, 2569 (1968). H. Hart and M. Kuzuya, J. Amer. Chem. Soc., 99, 2450 (1975). H. Hart and M. Kuzuya, J. Amer. Chem. Soc., 99, 2459 (1975). H. Hart and M. Kuzuya, J. Amer. Chem. Soc., 99, 1545 (1976). H. Hart and M. Kuzuya, J. Amer. Chem. Soc., 99, 1551 (1976). E. T. McBee, D. K. Smith and H. E. Ungnade, J. Amer. Chem. Soc., 99, 387 (1955). S. G. Cohen, R. Zand and C. Steel, J. Amer. Chem. Soc., 99, 2895 (1961). A. Oku, T. Kakihana and H. Hart, J. Amer. Chem. Soc., 99, 4554 (1967). l‘tl‘ll.( 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 190 H. Prinzbach, Pure Appl. Chem., 99, 17 (1968). (a) S. J. Cristol and R. L. Snell, J. émgr. Chem. Soc., , 5000 (1954): (b) S. J. Cristol and R. L. Snell, . Amer. Chem. Soc., 9, 1950 (1958); (c) E. R. H. Jones, G. H. Mangfiel and M. C. Whiting, J. Chem. §gg., 4073 (1956). National Distillers Chemical Co., Ashtabula, Ohio, Technical Bulletin, "High Surface Sodium", Available from U.S. Industrial. (a) H. Pines and W. O. Haag, J. Org. Chem., 99, 328 (1958): (b) W. O. Haag and H.’Pines, J. Amer. Chem. §gg., 99, 387 (1960); (c) J. Shabtai and E. Gil-Av, J. Org. Chem., 99, 2893 (1963). National Distillers Chemical Co., Ashtabula, Ohio, Technical Bulletin, "Dispersion Sodium", available from U.S. Industrial Chemicals. This work was initiated by Mr. Bing-lin Chen, who showed that the reaction with simple cyclic ketones is feasible. A detailed account will appear in his Ph.D. Thesis. H. Hart, G. M. Love, and I. Huang, Tetrahedron Lett., 99, 1377 (1973). M. D. Soffer, R. A. Stewart, J. C. Cavagnol, H. E. Gellerson and E. A. Bowler, J. Amer. Chem. Soc., 99, 3704 (1950). J. C. Cornforth, R. H. Cornforth and S. R. Robinson, J. Chem. Soc., 689 (1942). I. Fleming, "Selected Organic Syntheses", John Wiley & Sons, New York, NY, 1973, p. 75. J. A. Barltrop, J. Chem. Soc., 958 (1946). S. C. Welch and J. H. Kim, Synthetic Communications, 9999, 27 (1976). G. Stork and J. W. Schulenberg, J. Amer. Chem. Soc., 99, 284 (1962). E. Wenkert, A. Afonso, J. B-son Bredenberg, C. Kaneko and A. Tahara, J. Amer. Chem. Soc., 99, 2038 (1964). G. Stork, A. Brizzolara, H. Landesman, J. Szmuszkovicz and R. Terrell, J. Amer. Chem. Soc., 99, 214 (1963). 76. 77. 78. 79. 80. 81. 82. 191 G. Stork, R. Terrell and J. Szmuszkovicz, J. Amer. Chem Soc., 19, 2029 (1954). M. A. M. Boersma, Cat. Rev. Sci. Eng., ( , 243 (1974). 1gmgl B. L. vanDuuren, N. Nelson, L. Orris, E. D. Palmes and F. L. Schmitt, J. Natl. Cancer Inst., 99, 41 (1963). (a) F. A. Beland and R. G. Harvey, g. Amer. Chem. Soc., 99, 4963 (1976); (b) R. G. Harvey, S. H. GOh and C. ortez, J. Amer. Chem. Soc., 99, 3468 (1975). D. D. Keane, W. I. O'Sullivan, E. M. Philbin, R. M. Simona, and P. C. Teague, Tetrahedron, 99, 2533 (1970). H. Hart, M. Verma, and I. Wang, J. Org. Chem., 99, , 3418 (1973). ' T. Haneishi, A. Terahara, M. Arai, T. Hata, and C. Tamura, J. Antibiotics, 99, 393 (1974). Ill“ 1'! II