HI 1 ¥ S WWNHWIllllliUWill“’IIHLHHHWI IfHESlC This is to certify that the thesis entitled AN EXAMPLE OF A 1,4-CARBENE ADDITION presented by JEFFREY N. RAGGON has been accepted towards fulfillment of the requirements for M . S . degree in CHEMISTRY $42 AJ Major professor /"‘ Date ‘1‘“ l0? lag-.3 0.7639 MS U is an Affirmative Action/Equal Opportunity Institution ' ow. ' " ch91. I‘ \K .I . q l . . r... (I e .I 5 4. 9 z". a) -" . 34‘ k -. «353:1: 1 # ”iv r—‘llfl Mismmsb-M AN EXAMPLE OF A 1,4-CARBENE ADDITION BY Jeffrey W. Raggon A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemistry 1982 G/aaa/i ABSTRACT AN EXAMPLE OF A 1,4-CARBENE ADDITION BY Jeffrey W. Raggon In this thesis, a reinvestigation of the addition of dichlorocarbene to tetracyclones was performed. In the original work,30 29 dichlorocarbene, prepared by the Makosza method, was allowed to react with 2,3,4,5-tetraphenyl- cyclopentadien-l-one (38). The sole product reported for this reaction was the 1,2-dichlorocarbene adduct 32- Upon reinvestigation, another product, 1,1-dichloro- 2,3,4,5-tetraphenylcyclopentadiene 3Q was isolated. It ap- pears to arise from a 1,4-carbene addition. The possibility that $2 might be the result of a secondary rearrangement of the 1,2-dichlorocarbene adduct was examined. Solutions of 32, when subjected to acidic or basic conditions gave only recovered 32. Heating solutions of £2 in mesitylene or o-dichloro- benzene, on the other hand, resulted in the formation of two products. The major product, a tetraphenylchlorophenol Jeffrey W. Raggon. (48), was obtained in 40% yield. The minor product, ob- tained in 18% yield, was the cyclopentadiene g8. Since the temperature of the reaction for generating the di-~ chlorocarbene by the Makosza method never exceeds 140°C (the temperature in which $0 is formed thermally), 3% ap- pears to be a primary product. The proposed mechanism for the formation of 42 involves the addition of dichlorocarbene in a 1,4-fashion to gg fol- lowed by the elimination of carbon monoxide. The addition of dibromocarbene to gg, under the Makosza conditions, was also investigated. In this case, however, the l,4-addition was completely suppressed by the classic 1,2-addition. Finally, the excited state chemistry of 32, or its dibromo-analog 31, was examined. The primary interest of this investigation was to see if the excited singlet states of 32 and/or 41 would undergo rearrangements similar to, or different from, those observed in the ground state ther- molysis reactions. It was also of interest to us to see if the type of halogen on the cyclopropane ring would have any effect on these excited state rearrangements. Irradiation of the nn* transition of 32 or 42 in chloroform, warm dioxane, or carbon tetrachloride for 3.75- 4.5 h through a pyrex glass filter or 5—6 h through a uran- ium glass filter, with a 450-W high—pressure mercury-vapor Hanovia lamp, gave ca. 50% of the ring-expanded 3-chloro- or Jeffrey W. Raggon 3—bromo-2,4,5,6-tetraphenylphenol %Q or SQ. The proposed mechanism for the formation of ég or g0 involves cleavage of the cycloprOpane bond common to both rings followed by halogen migration, loss of halonium ion, aromatization and protonation of the resulting phenoxide anion. The possibility of an ion pair as an intermediate in the formation of 3g or S0 is briefly discussed. ACKNOWLEDGMENTS I wish to thank Professor Harold Hart for his enthusiasm and advice throughout the course of this study. Appreciation is extended to the National Science Founda- tion and Michigan State University for financial support in the form of research and teaching assistantships. I would also like to thank my parents for their en- couragement and enthusiasm during the last two years. Finally, I wish to express my appreciation to Laurel Lucas, Dr. Raymond Giguere and Dr. Lon-Tang Lin for their invaluable comments concerning the preparation of this thesis.' ii TABLE OF CONTENTS Chapter LIST OF FIGURES . . . . . . . . . . . . . . INTRODUCTION 0 O O O O O O O O O O O O O O 0 RESULTS AND DISCUSSION. . . . . . . . . . . .—~—--— «- -—————.—--.——-.- —- -—-- Molecular-Orbital Discussion . . . . . fiéfiEJEEEE'TTT. . . . . . . . . . . . Addition of Dichlorocarbene to 2,3,4,5-Tetraphenylcyclopentadien- l—one (3g) . . . . . . : . . . . . . . Addition of Dibromocarbene to 2,3,4,5- Tetraphenylcyclopentadien-1-one (QS) . Thermolysis of the 1,2-Dichlorocarbene Adduct of Tetracyclone (g2). . . . . . Addition of Dichlorocarbene to 2,5- Diphenyl-3,4-di-p-tolylcyclopenta- dien-l-one ($4). . . . . . . . . . . . Preparation of 4,4'-Dimethylbenzoin (4%) . . . . . . . . . . . . . . . . . Oxidation of 4,4'-Dimethylbenzoin (4g) Page 14 27 29 30 3O 31 32 34 35 36 Chapter Page Condensation of 4,4'-Dimethylbenzil (43) and Dibenzylketone. . . . . . . . . . . 36 Irradiation of the 1,2-Dichlorocarbene Adduct of Tetracyclone (32). . . . . . . . . 37 Irradiation of the 1,2-Dibromocarbene Adduct of Tetracyclone (41). . . . . . . . . 38 REFERENCES. . . . . . . . . . . . . . . . . . . . 39 iv Figure LIST OF FIGURES Page Orbital depiction of the singlet and triplet state of a carbene. . . . . . 1 Molecular orbitals for 1,2- and 1,4- additions. . . . . . . . . . . . . . l3 Symmetry of frontier orbitals of subStituted cyclopentadiene and carbene . . . . . . . . . . . . . . . . . 28 INTRODUCTI ON One characteristic reaction of carbenes is their ad- dition to double bonds to yield cyclOpropanes. This re- action, first reported by Doering and Hoffman in 1954,1 has become an attractive method for synthesizing three- membered rings and has been further applied to other un- 3 saturated substrates, such as allenes,2 acetylenes, and aromatic compounds.4 A reasonable orbital depiction of the lowest energy singlet state of a carbene is an sp2 hybrid with an empty p-orbital (Figure la). ,R (a) (b) Figure l. Orbital depiction of the singlet and triplet state of a carbene. The triplet state of a carbene is envisioned as a linear sp hybrid, with two half-filled p-orbitals (Figure lb).5 Triplet carbenes may also be viewed as biradicals,6 though the two unpaired electrons are located on the same carbon. Carbenes are electron-deficient species which in the singlet state appear to have an orbital structure similar to carbonium ions. On the other hand, they also possess a carbanion-like non-bonding pair of electrons. As a result, the electrophilic or nucleOphilic character of singlet carbene depends strongly on the ability of substituents on the electron-deficient carbon to withdraw electrons from, or supply electrons to, the divalent center.6 Generally, singlet dihalocarbenes add in a 1,2-fashion across a double bond. There are, however, a few isolated examples where l,4-attack occurs and has been observed to unequivocally arise from a one-step process. Unambiguous one-step mechanisms seem to be of five types: intra- molecular 1,4-addition yielding benzvalene,7 homo-1,4- additiOn of various halocarbenes to norbornadiene,8-14 addition of triplet dicyanocarbene to cyclooctatetraene 15 (COT), addition of the nucleophilic carbene cyclohepta- 16,17 trienylidene to tetracyclones, and recently, the one— step intermolecular 1,4-addition of dichlorocarbene to 1,2-bismethylene cycloheptane.18 The first of these examples was observed by Burger, Gandillon and Mareda:7 They found that base-induced Scheme I a-elimination of hydrogen chloride from S-chloromethyl-S- methylcyclopenta-l,3-diene (1) produces l-methyltricyclo- [3.1.0.02’6]hexene-3—(l-methylbenzvalene) (3), together ‘with toluene and. spiro[4.2]hepta-2,4-diene (5). A common inter- mediate S-methylcyclopenta-l,3-dien-5-y1-carbene (2) was presumably responsible for the 1,4-carbene addition, 1,2- carbon shift and CH-insertion, respectively (Scheme I). The authors were also able to synthesize an authentic sample of 2-methylbenzvalene (6) by another method outlined in Scheme II. Contrasting the spectral characteristics of the authentic 2—methylbenzvalene (6) with the spectral characteristics of the three products formed via the pre- sumed intermediate 3 (see Scheme I), allowed the authors to demonstrate that the intramolecular cyclopropanation was completely suppressed by the three alternative pro- cesses (i.e., intramolecular l,4-addition, 1,2-carbon shift and CH-insertion). W3 23:5: A A 6 Li" 8 6 7 4 Scheme II -...— A second example of a l,4-carbene addition involves the addition of various halocarbenes to norbornadiene 8’9 Jefford and co-workers reasoned that the least systems. electrophilic of the dihalocarbenes (342;, difluorocarbene)6 should possess a greater likelihood of reacting with a suitable diene in l,4-addition fashion. This notion was based on the assumption that the l,2- and 1,4-addition modes are in formal competition. Since both theoretical and ex- perimental studies attest to the electrophilic behavior of halocarbenes in the l,2-addition mode, it was reasoned by Jefford that any feature in the carbene which would disfavor the l,2-addition mode could alternatively, favor l,4-addi- tion. Thus, a carbene, less electrophilic with respect to l,2-addition, (e.g., difluorocarbene) should exhibit a greater potential for l,4baddition. They have been + i _____,/.f +fl + KF, 25°c _ F 9 ‘ 1065-6296) 11(4-11%) l 9‘ F F 12(34%) Scheme III successful in showing that singlet difluorocarbene does indeed add in a l,4-fashion to norbornadiene‘g, and that the product does not arise from a secondary thermal rearrange- ment of the l,2-addition product (Scheme III). Other l,4-carbene additions to various norbornadienes Bil-0'11 8,10,11- have also been established for CF2, CFCl, ,9’l2'13 and CBr2.9 A third instance of a l,4-carbene addition involves the CCl2 addition of triplet dicyanocarbene, generated thermally (g3. 80°C), to cyclooctatetraene (13) (Scheme IV). 16 14 Scheme IV Initially, the product of the reaction of singlet di- cyanocarbene with COT (ll) was thought to be the dihydro- 19 indene lg. However, in a reinvestigation carried out by 15 it was discovered that on conducting "variable the authors dilution" experiments there was a significant parallel between the magnitude of the ll to lg ratio and the pre- viously observed19 loss of stereochemistry in the reaction of dicyanocarbene with either cis- or trans- 2-butene. This loss of stereochemistry was shown to arise from the increase in the amount of triplet carbene with increasing dilution. Hence, the authors concluded that the pronounced dilution effect coupled with the knowledge that the two adducts (ll and lg) do not interconvert at the reaction temperature, (ll does not rearrange to lg at the reaction temperature (80°C); it does so slowly; however, at 160°)15 requires that ll and lg are formed from singlet and trip- let dicyanocarbene, respectively. Another unambiguous one-step 1,4-carbene addition to a series of substituted cyclopentadienones has been re- 16'17 In this particular example, the nucleophilic ported. .carbene cycloheptatrienylidene ll was generated in the presence of tetracyclones (1:1.5 M) by pyrolysis (di- glyme; 100°C; 5 h) of the sodium salt of tropone toluene- p-sulfonylhydrazone. Work-up followed by chromatography on alumina gave a mixture of benzocycloheptatrienes lg and lg in a ratio of 33. 1:2 in a total yield of 30-40%. The benzocycloheptatrienes lg and lg presumably arise from intermediates 3% and ll, believed to be equilibrating through the bis-norcaradiene ll (Scheme V). A final example of a conjugate l,4-addition was recently 18 They found that di- 29 discovered by Turkenburg, gt gl. chlorocarbene prepared by the Makosza method reacted with l,2-bismethylenecycloheptane lg yielding, besides the l,2-addition product 32, the 1,4-addition product 30 in a ratio of 99:1, respectively (Scheme VI). It was determined that lg rearranged to 38 only at elevated temperatures (285°C). Since the thermal rearrange- ment was not observed below 200°C, the l,4-adduct 30 must be a primary product. The authors also make reference to an analogous re- 20 action described by Landheer in his Ph.D. thesis. He proposed that the formation of éé from the 1,2-bismethylene- cyclopentane ll occurred gig the intermediate lg (Scheme VII). There have been other claims of achieving l,4-carbene additions which, however, are dubious. These examples either involve gas-phase reactions having a high prob- 21-23 ability of "hot" molecule isomerization, catalysis 24,25 with copper, or conditions so harsh as to raise the possibility of thermal rearrangement of the initially formed product.26 From a qualitative viewpoint, the key interaction in A A r [-6 Ar M H Ar H Ar 26 Scheme V 10 'ttfissnfl t -BuOK ./ t-BuOK > 32 33 Scheme VI 11 “SBn; Br Br ' 36 37 Br 1. 5»:an is 2. t-BuOK / 35 Scheme VII 1,4- and l,2-addition is between the p-orbital (LUMO) of the carbene and the n2 (HOMO) of the 1,3-diene (Figure 2a,b). The n1 bonding orbital of the 1,3-diene is sym- metry forbidden from reacting with the LUMO of the carbene. Consequently, much of the favorable p-nz interaction be- tween the HOMO of the 1,3-diene and the LUMO of the carbene is offset by the repulsive interaction between the filled 'o-orbital of the carbene and the nl-orbital of the diene (Figure 2c). On the other hand, no such difficulty is 12 present in the l,2-addition mode, since both WI and n2 orbitals can donate electron density to the LUMO of the 27 carbene (Figure 2b,d)- These qualitative arguments outlined above have been recently confirmed by semi-empirical MO calculations.28 In this thesis, a new, previously unobserved example of a 1,4-carbene addition to a diene is described. The addition of dichlorocarbene to a substituted cyclo- pentadienone, tetracyclone, under phase transfer conditions was previously reported.30 Narasimhan, Gurumurthy and Balasubaramanian obtained the l,2-dichlorocarbene adduct of tetracyclone in 60% yield as a white crystalline solid. However, in a reinvestigation of that experiment we were able to uncover new results which will be discussed in detail in the Results and Discussion section of this thesis. 13 (a) 1,4-addition mode: p—orbital (LUMO) of carbene interacting With‘fi.(HOMO) of diene. (c) 1,4-addition mode; symmetry-forbidden interaction oqu of diene with p-orbital (LUMO) of carbene. Figure 2. Molecular Orbitals additions. (b) 1,2-addition mode; orbital interactions same as (a). (d) 1,2-addition mode; p-orbital of carbene interacting with.n of diene. (Mo‘s) for l,2- and 1,4- Ph RESULTS AND DISCUSSION Addition of dichlorocarbene under phase transfer condi- 29 tions to tetracyclone lg afforded two products in addi- tion tp some recovered starting material (Scheme VIII). Cl Ph Ph 0 50 /o NaOH,(:I-ICI§_> + - 'TEIUA - ph I 0 Cl ' 38 ' 39 40 Scheme VIII The expected dichlorocarbene adduct 32 was obtained by flash column chromatography of the crude reaction product on silica gel. After recrystallization from methylene chloride and hexane a 30% yield of 32 was obtained as a white crystalline solid. The 1,2-dichlorocarbene adduct's structure was confirmed by comparing its infrared, mass, 14 15 1 30 and H NMR spectra to those reported in the literature. An unexpected second product was also eluted from the same column. After recrystallization from methylene chloride and hexane, a 25% yield of 38 was obtained as bright yel- low plate-like crystals. The product was identified as 1,1-dichloro-2,3,4,5-tetrapheny1cyclopentadiene (40) by infrared, proton and carbon-13 NMR, and mass spectroscopy, as well as by chemical transformations. The absence of a sharp carbonyl absorbance at 1720 cm-1 in the infrared spectrum of 39 indicated that this func- tional group was no longer present in the molecule. More- over, the failure to detect the loss of carbon monoxide (M+-28) in the mass spectrum of 30 also confirmed this fact. 13 In addition, the C NMR spectrum of 30 showed no more than eleven peaks in the aromatic region. This indicated a highly symmetrical molecule consistent with a structure such as 30. A single peak at 690.59 ppm also suggests the 3 presence of an sp carbon which could reasonably be the carbon bearing the two chlorines in 40. Acid hydrolysis of 40 with sulfuric acid in refluxing benzene gave the 13 tetracyclone lg. This seems to substantiate the C NMR result that C-1 of lg bears two chlorines. Finally, unequivocal evidence, which points to a highly symmetrical structure such as 38, was obtained by examin- 13 1 ing the C and H NMR spectra of the cyclopentadiene 16 1 prepared under Makosza conditions similar derivative 3&3 to those used to obtain 39 (See Scheme IX). If the struc- ture of 48 is correct, then the two methyls of both the p- tolyl groups of éé should appear in the proton and carbon- 13 NMR spectra as singlets. And indeed, this is what was observed. Therefore, we can conclude that the original structural assignment of lg (i;g;, substituted cyclopenta- diene where C-l in quarternary center being two chlorines) is correct. The generality of the 1,4-cycloaddition under Makosza conditions was also tested using other dihalocarbenes. Thus, the addition of dibromocarbene to tetracyclone 38 using bromoform, 50% sodium hydroxide, and a catalytic amount of triethylbenzylammonium chloride (TEBA) gave 41% of the 1,2-dibromocarbene adduct of tetracyclone ll (Scheme X). A.pure product was obtained through recrystallization from methylene chloride and hexane. Unfortunately the l,4-addition product (343;, 1,1- dibromo-Z,3,4,5-tetrapheny1cyclopentadiene) did not form. From a qualitative vieWpoint, it appears as though the bulky bromine atoms lack the ability to increase the electrophilicity of the divalent carbon to the extent that permits it to add in a l,4-fashion to an electron-deficient 1,3-diene, such as tetracyclone 38. Since the 1,4-addition mode was not observed when di- bromocarbene was allowed to react with 38, attention was 17 .41 KCN ?"' CH‘.'CHO + CHH'CCj HNO EtOH-H o 3 H; 1——> . 42 ? 1 Ar; Ar; . g .. 207. KOH/ETOH Ar Ar TEBA I 43 . Ar '=Dhenyl All Arz +_ . AI" Ar I 45 46 S cheme IX 18 50% NaOl-l , (:HBr3 TEBA Scheme X again turned to the formation of gg. It was originally thought that gg might result from a secondary rearrange- ment of the 1,2-dichlorocarbene adduct gg. To test this hypothesis, gg was subjected to acid and base. Reacting gg with the two-phase system of 50% sodium hydroxide and chloroform in the presence of a catalytic amount of TEBA resulted in the quantitative recovery of gg. This result proved that gg was not derived from gg under the basic conditions characteristic of the Makosza method. Similar results were obtained when gg was allowed to react under 19 acidic conditions with glacial acetic acid-acetic anhydride with a catalytic amount of 70% perchloric acid or p-toluene- sulfonic acid (PTS) in refluxing benzene or trifluoro- acetic acid in chloroform.‘ In each case gg was recovered unchanged. Thus, it appears that gg in a primary product and not the result of an aCid- or base-catalyzed rearrange- ment of the initially formed dichlorocarbene adduct gg. These results do not, however, rule out the possibility that gg results from a secondary thermal rearrangement of gg. To test this possibility, solutions of gg in various aromatic solvents were heated. Refluxing solutions of gg in benzene and toluene gave only recovered gg. 0n the other hand refluxing solutions of gg in mesitylenecnro-dichloro- benzene, induced a facile rearrangement of gg at 150°C within reasonable reaction times (3-5 h). Heating gg under ‘ these conditions afforded 33% of recovered starting material and two products (gg and gg), all isolated by preparative thin layer chromatography (Scheme XI).. The least polar thermolysis product was obtained at an optimum yield of 18%. Identification of this product as gg was ascertained in two ways. First, by comparing its spectral data to that of an authentic sample of gg prepared previously under the Makosza conditions, and secondly, by hydrolyzing the product in the presence of acid to tetracyclone gg. Recalling that gg was obtained previously at reaction temperatures below 60°C, it does 20 MES or 0008 150°C, 3-5h Scheme XI 21 not appear that gg is formed under Makosza conditions as a result of a secondary thermal rearrangement of gg fol- lowing the initial l,2-addition of dichlorocarbene to gg. Since gg was not formed from gg in appreciable amounts below temperatures of 140°C, one can safely assume that cyclopentadiene lg is a bonafide l,4-carbene adduct of tetracyclone gg. As to the mechanism, it.seems that the l,4-carbene addition to gg is followed by the elimination of carbon monoxide yielding cyclopentadiene gg. The major thermolysis product of gg (§;§;, gg) was ob- tained in 40% yield. When the thermolysis of gg was fol- lowed by infrared spectroscopy, a band began to appear at 1 1680 cm- after 22- 0.5 h. This absorbance, which decreased as the reaction continued, may have been due to an inter- mediate unsaturated dienone resulting from the initial cleavage of the carbon-carbon bond common to both rings in gg (see Scheme XI). Exposure of this unsaturated dienone to silica gel during work-up causes a facile rearrange- ment with aromatization to the 3-chloro-2,4,5,6-tetrapheny1- phenol gg. The structure of gg was determined by examining 13 the infrared, 1H and C NMR, and mass spectra. The infrared spectrum of gg revealed that the sharp 1 carbonyl band at 1720 cm- , characteristic of enone gg, had disappeared. A new absorbance at 3540 cm.1 in the 1 infrared spectrum and a singlet at 65.08 ppm in the H 22 NMR spectrum of gg indicated the presence of a hydroxyl group. Moreover, the electron-impact mass spectrum of gg showed that it had only one chlorine. With these facts in mind, the mechanism in Scheme XI can be proposed. Thermally induced Opening of the cyclo- prOpane bond common to both rings in gg yields the inter- mediate dienone. This intermediate dienone eliminates Cl+ on silica gel from C-6 followed by aromatization of the six- membered ring dienone and subsequent protonation of the resulting phenoxide anion yielding gg. Even though the spectral data and analysis showed that the product gg is a tetraphenyl chlorophenol, the exact location of substituents had to be determined. It was reasoned that removal of the chlorine from gg followed by quenching with a protic solvent would yield a tetra- phenylphenol. The latter could then be utilized by com- paring its spectral and physical characteristics (i;g;, melting point) to those of authentic tetraphenylphenols 32 previously prepared by Yates and Hyre. Since the 2,3,4,5- and 2,3,4,6-tetraphenylphenols melt at 183-184°C and 244- 245°C, respectively,32 one could pinpoint the exact loca- tion of the substituents in gg. It should be noted that the 4-chloro-2,3,5,6-tetraphenylphenol was not considered 1 to be a candidate for structure gg as neither the H nor 13C NMR spectra were consistent with such a symmetric structure. 23 Regrettably, all attempts to dehalogenate gg or its methyl ether gg were unsuccessful. Thus, the o-methyl- 3-chloro-2,4,5,6-tetraphenylphenol gg failed to give a de- halogenated product on treatment with magnesium dust in 33,34 refluxing tetrahydrofuran or sodium formate in di- methylformamide at 100°C in the presence of a catalytic 35 amount of palladium [0]. Moreover, gg was similarly reluctant to undergo dehalogenation when treatedwith a variety of reagents, such as titanium tetrachloride- 36 molecular hydrogen on 10% 37 lithium aluminum hydride, palladium/charcoal in tetrahydrofuran, two equivalents of n-butyllithium,38 and two equivalents of methyllithium both followed by quenching with methanol. Thus, the structural determination of gg rests on the interpretation of the obtained spectral data and a close examination of the most reasonable mechanistic pathways available to gg (iLeL, loss of C1+ from C-6 and subsequent aromatization rather than loss of Cl+ from C-S, see Scheme XI). Knowing that gg undergoes reaction thermally yielding the ring-expanded phenol gg and the cyclopentadiene gg, it was of interest to us to examine the excited state chem- istry of gg and its dibromocarbene analog gg. We were particularly interested in whether the singlet excited state of gg would undergo rearrangements similar to that observed in the thermolysis of gg. Of secondary interest, 38 24 was whether substituting two bromine atoms on the cyclo- propane ring would have any effect on these previously observed rearrangements. Irradiation of the nn* transition (Amax = 310 nm) of gg or gg in chloroform, warm dioxane, or carbon tetra- chloride for 3.75-4.5 h through a pyrex glass filter or 5-6 h through a uranium glass filter, with a 450-W high— pressure mercury-vapor Hanovia lamp gave 22° 50% of the ring-expanded 3-chloro- or 3-bromo-2,4,5,6-tetrapheny1- phenol gg or gg, respectively (see Scheme XII). x 0 ” x I“! ¢ ( >290nm,3.75-4.5h o x 0 a. ' > ¢ . a) >340nm,5-6h ¢ (0 ° 250% 0H X=C|(39) x: on (43) Br(47) Br (50) Scheme XII Filtration through silica gel followed by two re- crystallizations from benzene and high boiling petroleum ether yielded products with sharp melting points. Posi- tive identification of both photolysis products from the 25 two separate irradiations was obtained by comparing the spectral data, either directly (in the case of irradiating gg) or through analogy (in the case of irradiating gg), to that of previously accumulated spectral data on gg as ob- tained from the thermolysis of gg. As to the mechanism for the photolysis of gg and gg, it seems that the irradiation of the nn* chromOphore of the enone results in the cleavage of the cyclopropane bond common to both rings, followed by the migration of one of the halogen atoms (chlorine or bromine) to the carbon bearing the positive charge of intermediate gl or gg.39 WOrkup of the ring expanded dienone gg or gg resulted in workup Scheme XIII 26 loss of halonium.ion, subsequent aromatization, and proton- ation of the resulting phenoxide anion yielding gg or gg Scheme XIII). The possibility also exists that an ion pair could be an intermediate in this process (Scheme XIV).40 hv 39 or 47 + [excited state] 0 Silica Gel 4 4 50 a (trace H20) 8 or Scheme XIV .27 MolecularéOrbital Discussion The formation of the l,l-dichloro-2,3,4,5-tetrapheny1- cyclopentadiene gg from tetracyclone gg and the foregoing discussions led to the assertion that a 1,4-carbene addition has occurred. 28 suggests that a concerted linear cheletropic Theory l,4-addition of singlet carbene to a cisoid 1,3-diene is a symmetry allowed process. Nevertheless, examples of inter- molecular l,4-carbene additions are rare. One factor that accounts for the rarity of the 1,4- addition is of an entropic nature. This shortcoming can best be alleviated by choosing a diene possessing rigidity and/or planarity. This was first demonstrated by Jefford and coworkers in a series of papaer dealing with the homo- l,4-addition of various halocarbenes to norbornadiene.9-15 ‘ Tetracyclone-gg provides another example of a rigid cisoid diene that possesses a high probability of observing a l,4-addition. In addition to this entropic factor, the ideal symmetry of the frontier orbitals demands that both pairs of interacting orbitals (i.e., LUMocarbene-HOMOdiene and LUMOdiene-HOMocarbene) have the correct symmetry for l,4-ring closure (Figure 3). Once we admit that these two factors (iiéir planarity and/or rigidity of diene and correct symmetry of the fron- tier orbitals are present in the intermolecular addition of dichlorocarbene to gg, then the formation of gg can best be explained by a legitimate l,4-carbene addition. 28 Figure 3. Symmetry of frontier orbitals. Left: diene, right: carbene in t0p view. EXPERIMENTAL General Methods Melting points were determined on a Thomas Hoover Unimelt apparatus and are uncorrected. 1H NMR spectra were recorded on a Varian T-60 or on a Bruker WM-250 spectrometer with chemical shifts reported in a-units relative to tetramethylsilane. 13C NMR spectra were recorded on a Varian CFT-20 spectrometer or on a Bruker WM9250 spectrometer. IR spectra were obtained on a Perkin- Elmer 167 Grating Spectrophotometer. Low-resolution mass spectra were determined by a Finnigan Model 4000 spectrom- eter. 29 30 Addition of Dichlorocarbene to 2,3,4,5-Tetraphegylgyclo- pentadien—l-one (ggl .,A solution of tetracyclone (100 g, 26 mmol) in chloro- form (400 mL) was stirred at room temperature with 50% sodium hydroxide (100 mL) and benzyltriethylammonium chloride (1.0 g) until the reaction was complete (color changes from deep purple to yellowish-orange and sodium chloride precipitates). An equal volume of water was added and the aqueous phase was extracted three times with chloroform. The combined organic layers were washed with 10% aqueous hydrochloric acid, dried over magnesium sulfate ‘ and concentrated under reduced pressure to yield 14.7 g of crude reaction product. Purification by column chroma- tography on 400 g of silica gel starting with 1:10 methylene- chloride to hexane and increasing the polarity of suc- cessive elutions by 10% in methylene chloride until 100% methylene chloride is reached afforded 3.7 g (30.5%) of the dichlorocarbene adduct of tetracyclone gg, 2.9 g (25.5%) of 1,1-dichloro-2,3,4,5-tetraphenylcyclopentadiene gg and 1.4 g (13.8%) of the starting material. gg: mp 184-186°C; IR (KBr) 3100-3000, 1715, 1450, 1350, 730, 695 cm-1; 1H NMR (CDC13) 67.0-7.4 (m); mass spectrum, 31 m/e (relative intensity) 469 (13), 467 (58, M+), 466 (61), 439 (89), 438 (100), 432 (71), 431 (68), 403 (30), 396 (44), 367 (68), 289 (89), 189 (32), 77 (52). gg: mp 171-172°C; IR (KBr) 3100-3000, 1950, 1885, 1810, 1765, 1490, 1465, 1450, 1030, 920 810, 700 cm‘l; 1H NMR (CDC13) 67.5-6.6 (m); 13 C NMR (CDZClz) 6143.69, 141.29, 133.50, 132.94, 131.05, 130.15, 128.37, 128.22, 128.05, 90.59; mass spectrum, m/e (relative intensities) 442 (8), 440 (70), 438 (83, M+), 403 (22), 367 (39), 325 (26), 289 (100), 189 (28), 182 (36), 176 (42), 145 (60), 125 (41), 91 (33), 78 (71), 77 (64), 51 (51), 36 (43). Anal. Calcd. for c29320c12: C, 78.80; H, 4.55; C1, 16.14. C, 79.27; H, 4.59; C1, 16.14. Found: Addition of Dibromocarbene to 2,3,4,5-Tetr§phenylcyclo- pentadien-l-one_(gg) A solution of tetracyclone (5.0 g, 13 mmol) in bromo- form (100 mL) was stirred at 5-7°C while 50 mL of 50% sodium hydroxide and 100 mg of benzyltriethylammonium chloride were added. The ice bath was removed and the two phases were allowed to stir vigorously for 3.5 h. An equal volume of water was added and the aqueous phase was ex- tracted twice with chloroform. The combined organic layers were washed with 10% aqueous hydrochloric acid and dried over magnesium sulfate. Removal of the bromoform and chloroform under reduced pressure gave 10.0 g of a dark 32 brown tarry substance. Purification by column chroma- tography on 300 g of silica gel, eluting with methylene chloride/hexane (1/1) gave 3.0 g (41%) of the l,2-dibromo- carbene adduct of tetracyclone (gl) as light brown needle- like crystals with a melting point of 168-170°C. Re- crystallization of (gg) gave a nearly quantitative yield of pale yellow crystals with a melting point of 169-170°C; IR (KBr) 3100-3000, 1720, 1640, 1620, 1500, 1455, 1355, 1170, 750, 740, 700 cm'l; 13c NMR (c0c13) 6169.58, 132.96, 132.78, 131.94, 131.72, 130.65, 130.33, 129.93, 128.95, 128.56, 128.03, 127.53, 54.56; mass spectrum, m/e (rela- tive intensities) 555 (0, M+), 478 (66), 476 (100), 396 (53), 318 (5), 289 (49), 189 (52), 181 (45), 159 (63), 144 (41), 79 (37), 77 (31), 44 (34); Anal. Calcd. for C - 30 H OBrz: C, 64.77; H, 3.62; Br, 28.72. Found: (3, 64.87; 20 H, 3.59; Br, 28.81. Thermolysis of the 1L2-Dichlorocarbene Adduct of Tetra- gyclone (lg) Into a 25-mL pear-shaped flask was placed 300 mg (0.64 mmol) of gg and 10 mL Of freshly distilled 1,3,5- .— trimethylbenzene. The solution was stirred at 150°C for . “. _._..._. H -------_--_.- --.. - —. -.-.- _- -- 5 h. Two compounds in addition to starting material were separated from the reaction mixture using preparative thin layer chromatography. One compound, closest to the solvent ._..__.-- _-_ 33 and obtained in 18% yield as a yellow oil, was identified as 1,1-dichloro-2,3,4,5-tetraphenylcyclopentadiene gg: IR (thin film) 3100-3000, 2950, 2900, 2840, 1260, 750, 725, 680 cm-1; mass spectrum, m/e (relative intensities) 442 (10), 440 (60), 438 (100, M+), 403 (11), 367 (43), 289 (75), 182 (32), 176 (50), 149 (39), 133 (95), 105 (40), 73 (58), 57 (82), 43 (56). Unequivocal structural deter- mination of éQ was obtained by hydrolyzing it to tetra- cyclone gg with a few drops of concentrated sulfuric acid in refluxing benzene (60%). The other thermolysis product was obtained in 40% yield (110 mg). Its white needle-like crystals were identi- fied as 3-chloro-2,4,5,6-tetraphenylphenol (gg). mp 264-265°C; IR (KBr) 3520, 3100-3000, 1600, 1450, 1415, 1275, 1160, 950, 760, 740, 710 cm'l; 1H NMR (c0c13) 7.50 (m, 5H), 7.14 (m, 10H), 6.84 (m, 5H), 5.08 (s, in); 13c NMR (CDC13) 6150.084, 142.321, 138.969, 135.264, 134.970, 133.147, 132.558, 130.912, 130.794, 130.382, 128.136, 128.206, 127.266, 126.913, 126.736, 126.383, 125.795; mass Spectrum, m/e (relative intensities) 434 (30), 432 (100, M+), 396 (6), 379 (4), 367 (4), 289 (18), 189 (11), 182 (15), 145 (9), 44 (14). A similar thermolysis reaction run in a dichloroben- zene at 150°C for 3 h gave the same 1,1-dichlorosubstituted cyclopentadiene gg and ring-expanded phenol gg in 11% and 36% yield, respectively, in addition to 120 mg (40%) of recovered starting material. 34 Addition of Dichlorocarbene to 2,5-Diphenyl-3,4-di:p- tolylgyclopentadien-l-one (Q4) The dichlorocarbene addition was performed according to Makosza conditions as follows: A solution of gg (3.1 g, 7.52 mmol) in chloroform (80 mL) was stirred vigorously at room temperature with 50% sodium hydroxide (35 mL) and 100 mg of benzyltriethylammonium chloride (BTEA or TEBA) for 45 min. At the end of 20 min the deep purple solution had turned to a brownish-orange and a white precipitate had appeared (NaCl). An equal volume of water was added and the aqueous phase was extracted twice with chloroform. The combined chloroform layers were washed with 10% aqueous hydrochloric acid, dried over magnesium sulfate, and the solvent evaporated under reduced pressure afford- ing 5.0 g of crude reaction product. Purification by column chromatography on 135 g of silica gel starting elutions with methylene chloride/hexane (1/10) and increas- ing the polarity of successive elutions by 10% in methylene chloride until 100% methylene chloride is reached afforded two products and recovered starting material. The least polar compound was 1,1-dichloro-2,5-dipheny1-3,4-di-p- tolylcyclopentadiene gg (1.6 g, 45%). Recrystallization from methylene chloride and hexane gave 1.1 g (31%) of a yellow crystalline material melting at l39-141°C; IR (KBr) 3100-2860,.1495, 1450, 1030, 830, 800, 780, 700 cm‘l; 1H NMR (CDZC12) 67.6-7.0 (complex), 6.8 (complex 3), 35 2.2 (s); 13 C NMR (CD2C12) 6148.02, 145.26, 139.43, 138.37, 137.84, 133.32, 132.38, 130.32, 129.85, 129.61, 128.43, 128.14, 126.44, 80.15, 21.17; mass spectrum, m/e (relative intensities) 468 (76), 466 (100, M+), 431 (21), 39s (21), 339 (9), 303 (14), 182 (20). Anal. Calcd. for C31H24C12: C, 79.65; H, 5.18; C1, 15.17. Found: C, 79.48; H, 5.15; C1, 15.27. The other compound eluted from the column was the 1,2- dichlorocarbene adduct of gg (45, 2.0 g, 54%). Recrystal- lization from methylene chloride and hexane gave 1.6 g (43%) of an off-white crystalline material melting at 159-161°C; IR (KBr) 3100-2860, 1710, 1615, 1500, 1450, 1350, 1165, 750, 730, 695 cm‘l; mass spectrum, m/e (rela- tive intensities) 498 (7), 496 (55), 494 (64, M+), 468 (57), 466 (73), 460 (59), 459 (100), 431 (33), 424 (87), 395 (64), 303 (28), 189 (23), 182 (44). Anal. Calcd. for C OClz: C, 77.57; H, 4.88; C1, 32H24 14.31. Found; C, 77.40: H, 4.82; C1, 14.25. Preparation of 4L4'-Dimethy1benzoin (6%)41 In a 500-mL round-bottomed flask fitted with a reflux condenser were placed 66 mL of 95% ethanol, 53 mL of water, 50 mL (0.424 mol) of freshly distilled p-tolualdehyde gl and 5.1 g (0.078 mol) of potassium cyanide. The mixture was heated with vigorous stirring for 5.5 h. At the end of this period the solution was cooled, filtered through a 36 Buchner funnel, and washed with a little water. The yield of crude 4,4'-dimethy1benzoin gg was 36.6 g (72%). This crude product was used without further purification in the 42 next step; mp 70-74°C (lit. mp. 88-89°C); IR (Nujol) 3480, 3080-3000, 1680, 1610, 1100, 780 cm'l. Oxidation of 4,4'-Dimethy1benzoin(gg)43 In a 250-mL round-bottomed flask fitted with a reflux condenser were placed, 12 g (50 mmol) of the crude 4,4'- dimethylbenzoin gg, 60 mL of glacial acetic acid, and 30 mL of concentrated nitric acid. The solution was heated at 90-95°C for 1.5 h. The addition of 200 mL of ice and water resulted in 10.8 g of the crude benzil. Extended cooling gave another 500 mg of product resulting in a total yield of 1.3 g (93%) of the crude 4,4'-dimethyl- benzil gg. Recrystallization from methanol afforded pale yellow crystals with a melting point of 101-103°C (Lit.44 mp. 104-105°C); IR (KBr) 3100-2840, 1660, 1600, 1575, 1225, 1175, 890, 830, 745 cm'l. Condensation of 4,4'-Dimethy1benzil (gg) and Dibenzyl- 45,46 ketone In a 250-mL round-bottomed flask fitted with a reflux .condenser was added, 6.0 g (25.2 mmol) of 4,4'-dimethyl- benzil gg, 5.3 g (25.2 mmol) of dibenzylketone, and 50 mL 37 of absolute ethanol. The temperature of the mixture was raised until refluxing began at which time 4.3 mL of 20% potassium hydroxide in ethanol was added slowly from an addition funnel. When the frothing subsided, the solution was refluxed for about 30 min and then cooled to 0°C. The solid was filtered with suction and washed with a little 95% ethanol affording 6.5 g (63%) of deep-purple crystals of 2,5-dipheny1-3,4-di-p-tolylcyclopentadien-1-one gg; mp. 189-191°C. Recrystallization from benzene and petro- leum ether gave deep purple needle-like crystals with a 46 melting point of 218-220°C. (Lit. mp. 218-219); IR (KBr) 3100-2860, 1710, 1610, 1495, 1450, 1350, 700 cm'l; 1H NMR (CDC13) 57.3-6.7 (m), 2.3 (S). Irradiation of the l,2-Dichlorocarbene Adduct of Tetra- cyclone (ggl A solution of gg (467 mg, 1.0 mmol) in 130-mL of chloroform or carbon tetrachloride was irradiated through a pyrex glass filter with a 450-W high-pressure mercury- vapor Hanovia lamp for 3 h. Evaporation of the solvent under reduced pressure left a reddish-brown residue. Puri- fication by column chromatography on 15 g of silica gel, eluting with methylene chloride/hexane (1/2), gave 220 mg (50%) of gg as a white powdery solid. A purer product was obtained by recrystallization from benzene and petroleum ether. Mp 263-265°C; IR (KBr) 3520, 3100-3000, 1440, 1420, 38 1275, 1150, 750, 700 cm‘l; la NMR (c0c13) 67.506 (m, 5H), 7.128 (m, 10H), 6.828 (m, 58), 5.08 (s, 1H); 13 C NMR (CDC13) 6142.5, 139.20, 135.20, 131 .0, 130.9, 130.50, 129.0, 128.40, 127.50, 127.0, 126.70, 126.0; mass spectrum (m/e) 434 (14.61%), 432 (48.0%) 396, 289, 189, 182, 145, 77, 44, 40. Anal. Calcd. for C30H210C1: C, 83.22; H, 4.89; Cl, 8.19. Found: C, 83.27; H, 4.94; Cl, 8.38. Irradiation of the 1,2-Dibromocarbene Adduct of Tetra- cyclone (g1) A solution of 500 mg (0.900 mmol) of gg in 130 mL of carbon tetrachloride was degassed for 0.5 h and irradiated for 3.75 h through a pyrex glass filter with a high-pres- sure mercury-vapor 450-W Hanovia lamp. Evaporation of the carbon tetrachloride under reduced pressure left a brown residue which was filtered through silica gel and washed with diethyl ether, freeing 33. 200 mg (47%) of the crude ring-opened phenol gg from polymers. Recrystallization from benzene and petroleum ether gave pure gg as a white crystal- line solid melting at 263-264°C; IR (KBr) 3530, 3100-3000, 1450, 1400, 1270, 1150, 750, 700 cm'l; 1H NMR (c0013) 67.49 (m, 5H), 7.15 (m, 10H), 6.86 (m, 5H), 5.04 (br s, 1H); 13 C NMR (CDC13) 131.1, 130.5, 128.8, 128.2, 127.4, 127.0, 126.6, 126.0; mass spectrum (m/e) 478 (86.2%), 476 (100%), 396, 319, 302, 289, 189, 182, 145, 78. Anal. Calcd. for C30H210Br: C, 75.48; H, 4.45; Br, 16.74. Found: c, 75.38; H, 4.44; Br, 16.58. REFERENCES REFERENCES W. von E. Doering and K. Hoffman, J. Am. Chem. Soc., lg, 6162 (1954). _ W. R. Moore and H. R. Ward, J. Org. Chem., g, 2073 (1960); L. Skattebol, Tetgehedren Lett., 16 (1961); Ibid., 2361 (1970); Ibid., 2175 (1965); L. Skattebol, Acte. Qhem, Seenfi., lg, 1683 (1963); K. G. Untch, D. J. Martin, and N. T.Caste11ucci, Q, Qrg. Chem., gg, 3572 (1965); H. Nozaki, T. Aratani and R. Noyori, T t- ;eheggeg, 3645 (1967); H. Nozaki, T. Aratani and R. Noyori, Te;;ehe§;ep_§e§;., 2087 (1968); M. Bertrand, M. Santelli, and R. Mawrin, Bull. See. Chim. Fr., 998 (1967); W. R. Moore and T. M. Ozretich, Tetrahedron Lett., 3205 (1967); W. J. Ball, S. R. Landor and N. Punja, J. Chem. Soc., 194 (1967); M. Muhlstddt and J. Graefe, Chem , ger., lQQ, 223 (1967); E. V. Dehmlow and G. C. Ezimora, Tetrahedron Lett., 4047 (1970). H. M. Cohen and A. H. Keough, J. Org. Chem., gl, 3428 (1966); R. Breslow and R. Peterson, J, Am, Chem. Sec., g2, 4426 (1960); R. Breslow and L. J. Altman, J. Am. Chem. Sec., fig, 504 (1966); D. Seyferth and R. Damrauer, J. dOrg. Chem., gl, 1660 (1966); E. V. Dehmlow, Tetra- hed gen Le;;. 4003 (1965); Chemi_§e£.- lQQ, 3829 (1967); R. Breslow and H. W. Chang, J. Am. Chem. Soc., gg, 2367 (1961); R. Breslow, J. Lockhart, and H. W. Chang, J. Am, Chem. Soc., g}, 2375 (1961). R. W. Murray, Tetrahedron Lett., 27 (1960); G. L. Closs and L. E. Closs, g;_Amer‘_Qhem‘__ee,, gg, 599 (1961); W. E. Parham, D. A. Bolon and E. E. Schweizer, J. Am. ghem, See., fig, 603 (1961); G. Blume and P. Weyerstahl, Tetrahedron Lett. 3669 (1970); S. D. Saraf, Can. J, Qhem. 4%, 1169 (1969); M. K. Saxena and M. M. Bokadia, J. Indian Chem. Sec. fig, 855 (1969); M. Rabinovitz and H. Bregman, Ier, Q, Qhem,, g, 933 (1968); M. E. Volpin, D. N._Kursanov, and V. G. Dulova, Tetrahegron, g, 33 (1960); G. L. Closs and L. E. Closs, Tegrahedron Lett., 38 (1960). M. Jones and R. A. Moss, eds., "Reactive Intermediates", John Wiley and Sons, Vol. I, p. 69, 1978. J. F. Harrison, "Carbene Chemistry", W. Kirmse, Ed., Academic Press, New York, NY, 1971, Ch. 5, 39. U. Burger, G. Gardillon, and J. Mareda, Helv. Chim. Acta., Q4, 844 (1981). ' C. W. Jefford, J. Mareda, J. C. E. Gehret, nT. Kaben- gele, W. D. Graham, and U. Burger, J. Am. Chem. Soc., gg, 2585 (1976), and references cited therein. an 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 40 P. M. Kwantes and G. W. Klumpp, Tetrahedron Lett., 707 (1976). C. W. Jefford, nT. Kabengele, J. Kovacs and U. Burger, Helv. Chim. Acta., gg, 104 (1974). C. W. Jefford, nT. Kabengele, J. Kovacs, and U. Burger, Tetrahedron Lett., 257 (1974). C. W. Jefford, V. de los Heros, and U. Burger, Tetra- hedron Lett., 703 (1976). C. W. Jefford, W. W. Graham, and U. Burger, Tetrahedron Lett., 4717 (1975). C. W. Jefford, A. Delay, T. W. Wallace, and U. Burger, Helv. Chim. Acta., gg, 2355 (1976). A. G. Anastassiou, R. P. Cellura, and E. Ciganek, Tetrahedron Lett., 5267 (1970). K. Saito, Y. Yamashita, and T. Mukai, J. Chem. Soc. Chem. Comm., 58 (1974). T. Mitsuhashi and W. M. Jones, J. Chem. Soc. Chem. Comm., 103 (1974). L. A. M. Turkenburg, W. H. de WOlf, and F. Bickel- haupt, Tetrahedron Lett., 769 (1982). E. Ciganek, J. Am. Chem. Soc., gg, 1979 (1966). P. J. Landheer, Thesis, Vrije Universiteit, 1976, p. 7. . Franzen, Chem. Ber., gg, 571 (1961). . M. Frey, Trans. Faraday_Soc., gg, 516 (1962). Chem. Soc., 4402 (1961); 3826 (1962). U H B. Grybowska, J. H. Knox, and A. F. Trotman-Dickenson, J. J. A. Benson and E. S. Hand, J. Am. Chem. Soc., gg, 19 8 (1964). T. V. Domareva-Mandel'shtam and I. A. D'yakmov, J. H. Nozaki, M. Yamabe, and R. Noyori, Tetrahedron, gl, 1657 (1965). S. L. Regan, J. Am. Chem. Soc., gg, 5956 (1975). 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 41 H. Fujimata and R. Hoffman, J. Phys. Chem., lg, 1167 (1974). M. Makosza and M. Wawrzyniewicz, Tetrahedron Lett., 4659 (1969). R. Gurumurthy, P. Balasubramanian, and K. Narasimhan Synth. Comm., lg, 833 (1980). The substituted cyclopentadienone 4g was prepared in the following manner: Condensation of two moles of p-tolualdehyde 4 in water-ethanol using potassium cyanide as cata yst gave the benzoin 4 which was oxidized to the 4,4'-dimethylbenzi1 g using glacial acetic acid and concentrated nitric acid. Condensa- tion of gg, following a procedure in Org. Syn. Coll. Vol. III, p. 806, 1955, with dibenzylketone in ethanol using 20% potassium hydroxide afforded the substituted cyclopentadienone gg. Addition of dichlorocarbene under phase transfer con- dition329 gave, after recrystallization, 31% gg and 43% gg. P. Yates and J. E. Hyre,.J. Org. Chem., lg, 4101 (1962). R. D. Rieke and P. M. Hudnall, J;_Am. Chem. 893,, g4, 7178 (1972); R. D. Rieke and S. E. Bales, J. Chem. Soc. Chem. Comm., 879 (1973); R. D. Rieke and S. E. Bales, J. Am. Chem. Soc., gg, 1775 (1974). It is interesting to note that the model system, 4- chlorodurenol, upon treatment with activated magnesium followed by quenching with methanol, gave the dehalo- genated product in good yield. P. Helquist, Tetrahedron Lett., 1913 (1978). T. Mukaiyama, M. Hayashi, and K. Narasaka, Chem. Lett., 291 (1973). M. Busch and H. Stave, Chem.L Ber., gg, 1063 (1916); E. Ott and K. Kramer, Chem. Ber., gg, 1655 (1935); R. Baltzy and A. P. Phillips, J. Am. Chem. Soc., gg, 261 (1946); H. Kamerer, L. Horner, and H. Beck, Chem. Ber., , 1376 (1958); D. A. Denton, F. J. McQuiIIin, and P. . Simpson, J. Chem. Soc. 5535 (1964). Org. React. Vol. VI, pp. 339-409; Gilman, Langham, and Jacoby, J. Am. Chem: Soc., gl, 106 (1939); H. Gilman and A. L. Jacoby, J. Org. Chem., g, 108 (1938); G. Wittig, U. Pockels, andiH. Drdge, Chem. Ber., ll, 1903 (1938); H. Gilman and A. L. Jones, J. Am. Chem. Soc., gg, 1441 (1941). 39. 40. 41. 42. 43. 44. 45. 46. 42 M. Weiner and H. Hart, Tetrahedron Lett., 3115 (1981). Similar allylic cations conjugated with an enone seem to be preferred intermediates in the acid catalyzed rearrangements of epoxyenones; e.g., see H. Hart, 1. Huang, and P. Lavrik, J. Org. Chem,, gg, 999 (1974); H. Hart and E. M. Shih, J;i0rg. Chem., gg, 1128 (1975); H. Hart, S.-M. Chen, and S. Lee, J. Org. Chem., gg, 2906 (1980). R. Adams and C. S. Marvel, Org. Syn. Coll. Vol. I, John Wiley and Sons, Inc., New York, 1932, p. 94. R. Stierlin,Chem. Ber., 3g, 380 (1889). Mohrig and Neckers, "Laboratory Experiments in Organic Chemistry", D. van Nostrand Co., New York, 1979, pp. 370-2. R. Stierlin, Chem. Ber., lg, 381 (1889). J. R. Johnson and O. Grummitt, Org. Syn. Coll. Vol. III, John Wiley and Sons, Inc., New York, 1955, p. W. Dilthey and F. Quint, J. ffir prakt. Chem., lgl, 340 (1934). N T (lillllll) ’lllllii’ll("WWII”) WW) 31293 03175 7382