SYNTHETIC AND STRUCTURAL STUDIES ON THE 1,3 - DIPHENYLCYCLOBUTENYL CANON; THE PROBLEM OF 1,3-0RBITAL OVERLAP Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY ABDOUALTL MOSTASHARI 1969 TH (its This is to certify that the thesis entitled SYNTHETIC AND STRUCTURAL STUDIES ON THE 1,3-DIPHEHYL- CYCLOBUTENYL CATION THE PROBLEM OF l,3-0RBITAL OVERLAP presented by AbdoTjalil Mostashari has been accepted towards fulfillment of the requirements for Ph.D. degree in Organic Chemistry Major professor Date December , 1969 0-169 ' u mum! IIND ABSTRACT SYNTHETIC AHD STRUCTURAL STUDIES CH TFE l,3-DIPHEHYLCYCLOBUTEHYL CATIOH. THE PROBLEH CF l,3-0RBITAL OVERLAP. Ev AbdoTjaTi] Hostashari In order to examine the nossibiiity of 1,3—orbita1 overiap in the cycTobutenyT cation system‘, the garafproton chemical shift2 of the 1,3-dipheny1cyclobuteny1 cation was determined. This cation was synthesized from available precursors: a—truxiTTic acid 1 and B-nitrostyrene dimer g were converted to trgngfi,3-dipheny1cyciobutanc g. coon CCgtBU Pl Ph Ph ——-————;> Ph coon C03tBu E“\\\\\\\\§Si \ H0 3 'b (369 t . \ 2 I T.(CH3)3 \,:\\\\\ ’3 Ph Ph h '"___'_E’ h EDT: CH3)3 3—: N0 2 'b The hydrocarbon g was dihrominated to both nossibie isomers g and g. PO Br Br 84:- 80: Both of the dibromo derivatives were converted in good yields to the corresponding cyclobutenyl derivatives, Q and z, by a novel methodidnch used sodium iodide as displacement catalyst in dimethyl sulfoxide as solvent. Br OCH3 3 Ph .‘laI Ph \ P" CH3oH 7 Br Dnso. P Q and 5 g 3" OH Br “'Ph T'aI gm] L," 7 L h orzso Ph In order to interpret the results, the l,3-diphenylcyclopentenyl and l,3-diphenylcyclohexenyl cations were prepared as model compounds from their correspondino dienes: g3 and 2“. me. Q Ph Ph R 8‘0 The para-protons of the three cations lQ’ ll and lg appeared at r 2.l4, 2.10 and l.98 respectively in their nmr Spectra. ph’\ Ph Ph’fiah Ph‘.‘Ph 12 h, 1 SN The increase of participation of the phenyl groups in stabilization of the cations from cyclohexenyl lg to cyclobutenyl lg was taken as evidence against predominance of l,3-orbital overlap contribution in the stabilization of the latter ion. Instead, an electrostatic charge repulsion between positive charges on C1 and C3, which increases with the decrease in the distance between C] and (3, was soonested to he resnonsitle for the trend in the deshielding of the para-protons. As an ancillary synthesis, the reaction of l,3-dicarboxylic acids with lead tetraacetate was studied. y-Lactones were found to he the only products. a-Truxillic acid l yielded two lactones lg and lg. coon . Ph Ph'm LTA + on L CWT“ H l h The reaction was found to be applicable to the available l,3-diacids TS and 11 which yielded y-lactones lg and 12 resoectively. LTA Hooc-(cnz)3—ceon ’ —;> Z:;:>E=h 1i 15 on. ”DOC COCH LTp Ph Ph 0 17 1g 'Vlu SYNTHETIC AND STRUCTURAL STUDIES ON THE l,3-DIPHEHYLCYCLOBUTEVYL CATICH. THE PROBLEM OF l,3-CRBITAL DVEPLAP. By Abdoljalil Hostashari A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1969 (guyzl 74—70 To Enci without whose love this work would not have been possible ACKNOWLEDGMENT The author wishes to express his deepest gratitude to Professor Donald G. Farnum, for his guidance and encouragement throughout these studies. Thanks are also extended to individual members of the organic faculty of Michigan State Universitv for their advice and assistance during the author's graduate studies in this Department. The author acknowledges the financial assistance of the people of Iran, as administered by the Ninistry of Sciences and Higher Educations, during the course of these studies. Financial assistance bv Michigan State University, in the ferm of teaching assistantshins from Fall, l964 to Fall, 1966, and by the National Science Foundation in the form of research assistantships from winter, lDC7 to Fall, TOE? is gratefully acknowledged. ii TABLE OF CONTENTS Page INTRODUCTION .......................... 1 RESULTS AND DISCUSSION ..................... 4 I. l,3-Diphenylcyclopentenyl System ........... 4 a) Preparation of 3-Phenylcyclopent-Z-ene-l-one(v. . . 4 L‘I b) Preparation of l,4-Diphenyl-l,3-cyclonentadiene(§). II. l,3-Diphenylcyclohexenyl System ............ 6 Preparation of 3-Phenylcyclohex-2-ene-l-one(Z) . . . . C Preparation of l,3-Diphenyl-l ,3-cyclohexadiene Q2). . . C III. Preparation of Diphenylcyclobutenyl Cation Precursor . 8 Photodimerization trans Cinnamic Acid and e-Nitrostyrene .................... 0 Preparation of trans l,3-Diphenylcyclobutane(lg) . . . l0 Photodimerization of B-Hitrostyrene .......... l3 Reduction and Alkylation of Dinitrodiphenylcyclo- butane Q/fil’) ...................... l3 Preparation of l,3-Diphenyl-2,4-bismethiodide(lg). . . l5 Reductive Deamination of lg ............. l5 Preparation of l,3-Dibromo-l,3-diphenylcyclobutane(l2) l6 Attempts at Conversion of Dibromocyclobutane lg to Cyclobutenyl Derivatives ............... 17 Preparation of l,3-Diphenylcyclobut-Z-enyl Cyanide(ggh l7 Attempted Conversion of the Cyanide gg to the Corresponding Amide or Acid .............. l8 iii TABLE OF CONTENTS (Continued) Page Preparation of l,3—Diphenylcyclobut-Z-ene Carboxami de (a) .................... 19 Attempted Hydrolysis of the Amide 21 ......... 20 Preparation of l-Hethoxy-l,3-diphenylcyclotut-2- ene (23) ........................ 20 Preparation of l,3-Diphenylcyclobut—Z—ene-l-ol(2g) . . 21 Preparation of l,3-Diphenylcyclohexenyl Cation(2gl . . 23 Preparation of l,3-Diphenylcyclopentenyl Cationczl). . 23 Preparation of l,3-Diphenylcyclobutenyl Cation (3/2) . . 24 Comparison of Cations 22, 31, and 32 ......... 26 EXPERIHENTAL .......................... 20 Preparation of 3-Phenylcyclopent-Z-ene-l-one(l). . . . 29 Preparation of l,4-Diohenyl-l,3-cyclopentadiene (9. . . 30 Preparation of 3-Phenylcyclohex-2-enone(Z) ...... 30 Preparation of l,3-Diphenyl-l ,3-cyclohexadieneQQ. . . 3l Photodimerization of tran§_Cinnamic Acid ....... 32 Preparation of Truxilloyl Chloride .......... 33 Preparation of di-t-Butylpertruxillate(lg) ...... 33 Titration of di-t-Butylpertruxillate ......... 33 Pyrolysis of di-t-Butylpertruxillate ......... 34 Preparation of l,3-Dibromo-l,3-diphenylcyclobutane(lg) 35 Preparation of B-Nitrostyrene ............. 36 Photodimerization of B-Nitrostyrene .......... 36 Reductive Alkylation of l,3-Diphenyl-2,4- dinitrocyclobutane(l§l ................ 3( iv TABLE OF CONTENTS (Continued) Page Preparation of l,3-Diphenyl-2,4-bi§; (dimethylamino)cyclobutane ............... 37 Preparation of l,3-Diphenyl-2,4-Lis— (dinethylamino)cyclobutane DimethTEdide ........ 38 Reductive Deamination of fi§;methiodide lg ....... 38 Reaction of l,3-Dibromocyclobutane lg with Methanol in the Presence of Sodium Iodide ,,,,,,,,,,,, 39 Treatment of Dibromocyclobutane lg with Sodium Bicarbonate ...................... 39 Attempted Preparation of l,3-Diphenvlcyclobutenyl Cation. . . . ..................... 39 Reaction of Dibromocvclobutane with Sodium Hethoxide . . 40 Preparation of l,3-Diphenylcyclobut-Z-enyl Cyanide (a). an Preparation of l,3-Diphenylcyclotut-2-ene-l- carboxamide (g2) .................... 4l Preparation of 3-Methoxy-l,3-diphenylcyclobutene (23). . 41 Preparation of l,3-Diphenylcyclobut-Z-ene-l-ol(2Q). . . 42 Preparation of trans-Z-PhenylcyclonrOpylcarboxylic Acid .......................... 42 Preparation of trans-l-Benzoyl~2-phenylcyclopropane . . 43 Preparation of l-Benzovl-2-phenylcvcloprooanetosyl- hydrazone (fl/l) ...................... 4f Preparation of gi§_and trans-a-Phenylcinnamic Achdcya .46 Preparation of a~Phenylcinnamonitrile (lg) ...... 47 Preparation of l,2-Diphenylcyclopropyl Cyanide(ll). . . 47 Hydrolysis of l,2-Diphenylcyclopr0pvl Cyanide ..... 47 ‘J Preparation of Hethyl cis-a-Phenylcinnamate(lg) . . . . 4 In V Preparation of l,2-Diphenylcyclopropylcarboxylic Acid Q and B ...................... 48 TABLE OF CONTENTS (Continued) Paoe Reduction of l,2-Diphenylcarboxylic Acid l ...... 49 Reduction of l,2-Diphenylcarboxylic Acid g ...... 50 Treatment of Truxillic Acid with Lead Tetraacetate . . 50 Treatment of Truxillic Acid with Lead Tetraacetate in Various Solvent Systems .............. Sl Attempted Isomerization of Lactone lg to Lactone ll. . 52 Treatment of a-Phenylglutaric Acid with Lead Tetraacetate ..................... 52 IV. Ancillary Synthesis. Discussion of Results ...... 54 Reaction of Lead Tetraacetate with l,3-Dicarboxylic Acids ......................... 55 Reaction of a-Truxillic Acid with Lead Tetraacetate. . 56 REFERENCES ........................... 59 vi LIST OF TABLES TABLE Page 1. Chemical Shifts of Cations 22, ll, and £2 ........ 26 II. Treatment of a-Truxillic Acid with Lead Tetraacetate in Various Solvent Systems ................ 52 vii FIGURE 1. C0 LIST OF FIGURES Nmr Spectrum of l,4-Diphenyl-l,3-Cyclooentadiene(6) in CDCl 3 OOOOOOOOOOOOOOOOOOOOOOOOO Nmr Spectrum of l,3-Dinhenylcyclopentenyl Cation in HSO3F (THA Standard) at Room Temperature ......... Nmr Spectrum of l,3-Diphenylcyclohexenyl Cation in HSO3F (TMA Standard) at Room Temperature ......... Nmr Spectrum of l,3-Diphenvlcyclohutenyl Cation at .-70° in HSO3F (TNA Standard) ............... Nmr Spectrum of l ,3-Diphenylcylcobutenyl Cation at 0° in HSO3F (TNA Standard) ................. Nmr Spectrum of l,3-Diphenylcyclobutenyl Cation at Room Temperature in HSO3F (TMA Standard) ........... Nmr Spectrum of 3-Methoxy-l,3-diphenylcyclobutene (25) . . Nmr Spectrum of Lactone lg ................ viii INTRODUCTION The possibility of 1,3-n overlap in cyclobutenyl cation has been explored by Kiefer and Robertsl. They suggest that simple N0 calculations predict a delocalization energy of 4.1-10.9 kcal/mole, when the 1,3 resonance integrals (8 1,3) are assigned values of 0.2- 0.5 8 based on the distance between C1 and C3. Their results imply that the structure of the cation could be better presented as (11) I II rather than (I). Based on this prediction, they investigated the solvolysis of 1-methy1cyclobut-2-enyl bromide. They found that the relative rate of this compound compared with that of an open chain allylic model compound (a, y, y, trimethylallyl bromide) was l/l7,000. The authors suggested that in the absence of l,3-n overlap, this ratio should have been much smaller in view of the tremendous strain introduced in the already strained cyclobutene system in the transition state by addition of one more sp2 center in the molecule. Based on corrective factors the authors introduce, they attribute an accelerating factor of 103-105 (corresponding to 4.7-7.0 kcal/mole) in the rate of solvolysis of the cyclobutenyl bromide. -2- In another investigation2 acid catalyzed solvolysis of the p;nitrobenzoate of 3-pheny1-2-cyclobutenol was found to be only 7.8 times slower (corresponding to 1.1 kcal/mole) than that of methyl— cinnamyl alcohol pfnitrobenzoate. This type of solvolysis (AAL I) which goes through a carbonium ion type transition state in a cyclo- butene system should result in a difference of at least 5 kcal/mole of Ea rather than 1.1 kcal/mole. Roberts and coworkers therefore suggested that this difference is due to l,3-n overlap in the intermediate cation as in cation (IV) which is presumably much more stable than III. III IV Katz3 has also interpreted the nmr spectra of some polymethyl cyclobutenium cations as supporting l,3-n overlap. When he treated different l,2,3,4-tetramethyl-3,4-dihalocyclobutenes with excess silver hexafluoroantimonate in methylene chloride, he only isolated one equivalent of silver halide and the nmr spectrum of the cation formed showed for instance for the dichloro compound: three signals at r 7.09, 7.57 and 7.72 with relative intensities 2:1 :1 assigned aswfollows. Based on 7.72 / "" \ 7.57 7.09 V VI a comparison with polymethylated cyclopentenyl cations he concluded that the allylic methyls at 1 7.09 were shifted upfield, while the central methyl at r 7.57 was shifted downfield. He took this as evidence for a -3- contribution from bicyclobutyl cation VI. However,we believe Katz has ignored some important effects on the chenncal shifts. Thus, the hybridization in the strained cyclobutenyl cation will be different from that in the less strained cyclopentenyl cation. This will have an indeterminate effect on the directly attached methyls. Also, since the geometry is different in the two cations, the anisotropic effect of the 'allyl system on the attached methyls will be different - most probably resulting in an upfield shift of the allylic methyls on the cyclobutenyl cation. Therefore, we have some reservations about the reliability of Katz's conclusion of l,3-over1ap in the cyclobutenyl cation. The studies reported in this thesis were undertaken in order to determine the possibility of the use of the pa:a_proton downfield shift in the nmr spectrum of 1,3-diphenyl substituted cyclobutenyl system as a measure of l,3—n overlap. This criterion has been explored by D. G. Farnum“. It was reasoned that if there is any l,3-n overlap as a major contributor to stability, then the pgra_protons of the phenyls would feel less positive charge. Therefore, by comparison of their chemical shift with those of model compounds such as the l,3-diphenyl- cyclohexenyl and l,3—diphenylcyc10pentenyl cations, the extent of the effect of l,3-n interaction could be determined. It should be noted that from the approximate chemical shift-delocalization energy relationship suggested by Farnum“, the delocalization energy of 4.1-10.9 kcal/mole expected from Roberts calculations1 should be detectable by this method. As a corollary part of these studies it was necessarv to explore the synthesis of a number of small ring compounds as potential precursors to the l,3—diphenylcyclobutenyl cation. RESULTS AND DISCUSSION 1. WWWe - A natural precursor of 1,3 diphenylcyclopentenyl cation was thought to be l,3-diphenylcyclopent- 2-ene-l-ol or the dehydrated analogue l,3-diphenyl-l,3-cyclopentadiene. A literature survey showed no reported synthesis of the former, but the latter had been reported by Borsche and Henzs, through a low yield preparation of 3-phenylcyc10pent-2-en-1-one l and subsequent treatment of this ketone with phenyl magnesium bromide. a) Prgaaration of 3-phenylcyclopent-2-ene-l-one (l). In this preparation the method of Borsche was employed. Condensation of the sodium salt of ethylacetoacetate with phenacyl bromide afforded the substituted acetoacetate ater 2. Saponification and decarboxylation 0 0 0 CO Et Ph-U-CHzBr + CH3-UfEH-U0Et -——-;> Ph-U-CHz-CHE—C-CH3 3 of 2 was carried out cccording to the modification of Wilds and JohnsonG, using very dilute aqueous potassium hydroxide (0.8%). o o l \ - - - - g / Ph-U CH2 c112 c! CH3 it Use of a more concentrated solution of base leads to the undesired cleavage reaction: -5- \ i- _ - - g / Ph CH2 cw2 COZEt + CH3 C00 The diketone 2 was characterized by its spectroscopic preperties. The two carbonyl stretching vibrations at 1720 and 1694 cm'] are as expected for a simple aliphatic ketone andzabenzoyl group respectively. The nmr spectrum showsa benzoyl pattern at T 2.2 and 2.68 with relatively intensities of 2:3, the two methylene groups show an A282 pattern centered at r 7.18 and the methyl ketone at r 7.95. Ring closure was accomplished by treating diketone 2 with 10% KOH. The product l shows a doublet at 1705 (shoulder) and 1695 cm'1 in its ir spectrum which is apparently normal with cyclopentenones due to fermi resonance7. The nmr spectrum of l exhibits a 5 proton multiplet signal at r 2.6 due to the phenyl group. The vinylic proton (HA) is represented by a peak at r 3.6 as a triplet 3...: JAB = 1.5 cps. TheA282 pattern centered at 8.35 is assigned to the two methylene groups, the lower field portion of the A282 at t 8.05 (H3) 15 more complicated by further coupling to HA. b) Preparation of supposedly 1,3-diphenyl-l,3-cyclopentadiene. The cyclopentadiene was prepared by treatment of the cyclopentenone l with phenyl magnesium bromide. The nmr spectrum of the cyclopentadiene product showed,besides the phenyl absorption, two triplets at T 3.12 and 6.27 with relative intensities of 1:1. The two triplets with a coupling ~6- constant of l cps suggest a more symmetrical structure than 1,3 diphenyl- 1,3-cyclopentadiene l which should show two vinyl hydrogens with different coupling constants with the methylene group. The Spectrum rather suggests a structure of the type 2 or Q. t t 32 Structure 2 seems unlikely considering the synthetic route and small coupling constant involved. Structure 2 seemed to be best in accord with the nmr. A literature survey revealed that Drake and Adams8 have reported a one step synthesis of 1,4-diphenyl-l,3-cyclopentadiene whose properties are identical with those of the described l,3-diphenyl- 1,3-cyclopentadiene. This compound later has been reported9 to form Diels-Alder adducts of the 1,4-dipheny1 substituted bicyclo[2.2.l]hexan structure. 11- WWMO - The We compound was prepared from treatment of 3-phenylcyclohex-Z-ene-l-one 2 with phenyl magnesium bromide. Ketone l was in turn prepared by Walker's method“). Treatment of ethylbenzoylacetate with methylvinylketone in the presence of benzyhnfinethylammonium hydroxide (triton B) as base yielded a mixture which was triturated with ether to give a solid. Walker failed to assign a structure to this solid but found the molecular formula to be 015H1804. The ir spectrum of this compound was also reported - — -l by him to exhibit absorptions at 3448 cm 1, 1694c:m1 and 1709 cm . The -7- nmr spectrum of this compound was found to be more helpful in structure elucidation. It shows a five proton multiplet at 1 2.72, a one proton broad singlet at r 5.79, a two proton quartet at r 6.05, a one proton multiplet at r 6.5, a two proton singlet at r 7.4, a four proton multiplet at r 7.55and a three proton triplet at r 9.02 which were assigned to structure 2 as follows: 7.55 'm f\/ o s q t H 6.5 6.05 9.02 7.4 / co CH -CH 4?”””’T” H 2 2 3 Ph 5.79 2 (r values) Assignment of the signal at 5.79 to the -0H proton was based on the fact that when the sample was shaken with 020, this signal disappeared. The keto ester'2 and the oil obtained from evaporation of the ethereal solution, to which Walker has assigned the structure 2, were both converted to ketone l on treatment with 10% KOH. o o o o o 1! 1! '8 ¢- -cn2- -0Et + CH2=CH- -CH3 -———;> + [H OzEt C02Et Ph Ph t 2 O 2 and 2 :> Ph -3- Ketone 2 shows a carbonyl stretching absorption at 1675 cm']. This ketone has been reported by several other investigatorsll,12.13. l,3- Diphenylcyclohexadiene(lg)was prepared by treatment of ketone 2 with phenylmagnesium bromidell. 7.4 _ T ‘fTT HC JAB - 1 cos (JHC Ph JAC = 0.5 cps JBC] = 4.5 cps T 3.32 B A quintet Assignment of the low field vinylic absorption to HA is based on the expected anisotropic effect of the two neighboring phenyl rings and the coupling constants. m. Wemmm- - Preparation of the potential precursors of 1,3-diphenylcyc10butenyl cation with the general structure ll could have been persued via two routes. a) Preparation of l,3—dipheny1cyclobutene Q2 and function - alization of C3 by halogenation at benzylic and allylic C3; Xph 3h 1 14 Ph ll Ph 1% There are two reported syntheses of l2 in the 1iterature1”’15, neither of which is on a preparative scale. The Spectroscopic properties reported by the two authors are not in agreement. Furthermore l2 has been reported to be very unstable and decomposes rapidlyls. Based on these difficulties this route was not pursued. -9- b) Preparation of l,3-diphenyl cyclobutane l2 and conversion of this compound to the precursor ll, such as to avoid the intermediate cyclobutene l2. Although simple 1,3-diphenylcyclobutane has not been reported in the literature, substituted derivatives are very well known. Systems such as truxillic acid15(Ll) l,3-diphenyl-2,4-dinitrocyclobutane Q2) and l,3-dipheny1-2,4-diacetylcyclobutane18 QR) have been reported to be photodimers of B substituted styrenes. 3H3 COOH NO2 CC Ph Ph Ph Ph h h Lt 12 IA Ready availability of these compounds and available synthetic methods for decarboxylation of the acid or removal of the nitro group led us to believe that these compounds would be useful as synthetic intermediates in the preparation of l,3-diphenylcyclobutane ([2). l-Photodimerization offgragsficinnamic acid and_§-nitrostxreney General: Solid state photodimerization of trag§_cinnamic acid has been studied extensively and most of the possible isomers of the dinww-have been isolated19. Practically, however, solid state photolysis of trag§_cinnamic acid could only be carried out with small amounts. Stirring of a solid which would allow all the crystals to be exposed to light is a problem in irradiation of large amounts. Suspension of the crystals in water on the other hand had been reported to proceed in low yields (LS-9.5%)”,21 over long periods of time. The problem was thought to be that -10- the crystals are not fully exposed to light. When the source of light was placed inside a suspension of commercial cinnamic acid in water and the suspension was stirred vigorously a good yield (20-60%) of dimer was obtained. Commercial cinnamic acid,which is presumably deposited quickly at the time of preparation, has been reported to give only the undesired isomer B-truxinic acid22(ll) when irradiated in the solid state. COOH COOH -' AZ Ph However, in our study, a-truxillic acid (w seemed to be the only product. This is perhaps due to the solution of the metastable 8 form of trans cinnamic acid in water and recrystallization as the stable 0 form. This conversion has elegantly been demonstrated by Cohen and Schmidt23. Preparation of trans-1,3-dipheny1cyclobutane (l2). Decarboxylation of simple secondary carboxylic acids doesn't seem to be an easy task. Conventional methods such as ouinoline and copper or basic or acidic decarboxylations are applicable only in the case of unsaturated acids or B keto acids and so on. Wibergz” gt_gl.had reported decarboxylation of bicyclo[2.l.l]hexane-5-carboxylic acid by conversion of the acid to the t-butylperester and subsequent decomposition of the perester in p-cymene. This method has been successfully used by Eaton25 in the conversion of cubane dicarboxylic acid to the parent hydrocarbon in moderate yield. -1]- Although Eaton has applied this reaction to cubane dicarboxylic acid, it was suspected that our case would be much more like the reaction of vicinal diacids in which the carbon radicals could combinezs. Therefore, we expected complications from this method. The diperester of truxillic acid, l2, was prepared by the reaction of the bi§_acid chloride with t-butyl hydroneroxide. The perester, a seemingly stable solid, which was titrated to be 94-95 percent active, showed a carbonyl 1 which has been reported for other stretching vibration at 1745 cm' perester527. When the diperester was heated in p-cymene or p-diisopropyl- benzene, it decomposed to give tragsyl,3-diphenylcyclobutane(lg)in moderate yield. Isolation of the desired product from the product mixture proved to be difficult. Extensive column chromatography was necessary to separate ll from the dimers of the solvents. The most convenient method of separation was found to he preparative vapor phase chromatography which is accompanied by approximately 30% loss, but yields pure hydrocarbon. Ph A COZOtBu - 0+ @c-c—o— / l I Pn COZOtBuPh 13 Vb The nmr spectrum of l2 shows a quintet at r 6.4 and a triplet at r 7.48 J = 6.8 cps with relative intensities of 1:2. The signal at lower field was assigned to the benzylic protons coupled to four a protons, and the triplet was assigned to the two methylene groups. -12- An alternative synthesis of l,3-diphenylcvclobutane (l2). The difficulty in isolation of cyclobutane in the diperester decomposition method made us look for an alternative synthesis of l,3-diphenyl cyclo- butane. Photodimerization of B-nitro styrene has been reported to proceed in moderately good yield to trans, trans, trans-l,3-diphenyl-2,4-dinitro- cyclobutane23(l§) This compound could be reduced to the corresponding diamine which could be deaminated by one of several methods reported for reductive deamination of aliphatic amines. Nickon29 treated amine sulfonamides with hydroxyl amine-O-sulfonic acid in alkaline solution: NH x R-NH-SOZAr jg—é ArSOZG + N2 + R-H o X = OSO3H or C1 This method, however, is tedious and the yields are usually low to moderate. Another method reported by Grovenstein and Rogers30 at the same time is much simpler and deals with cleavage of quarternary ammonium salt with sodium in liquid ammonia. Reduction of dinitro compound ll to the diamino analog and conversion of the amine to the quarternary ammonium salt should not be difficult. Therefore an alternate route to the cyclo- butane l2 was devised as shown below: /=/N02 “" > PI 7 Ph Lt -13- N(CH3)2 IGEN(CH3)3 P'h > Pb“ Lg N(CH3)2 (n+3)3 e ‘9 L10? Na \ p, 12 {1113(1) 7 Ph 1,3 Photodimerization of e—nitrostyrene. The apparatus used for this reaction was very simple: an 18 inch crystallization dish and a 550 H Hanovia lamp jacketed by a Pyrex cooling well. The suspension method gave very good yields of the dimer ll. The dimerization had been reported in the solid state to give yields of 25-40% over a period of three weeksze. The suspension method gave yields of 35-50% within 24 hrs. When the crystallization dish was cooled to 18—20° in a water bath the dimerization yield increased to 50-70% in a 24 hrs period of irradiation. Separation of dimer was accomplished by soxhlet extraction of the mixture of dimer and monomer. This gave almost pure dimer which was further purified by recrystallization from benzene. The nmr of ll shows an AA'BB' system centered at r 6.6 with 08 - VA = 53 cps at 60 Mc. Reduction and alkylation of dinitrodiphenylcyclobutane(ll) Reduction by zinc and hydrochloric acid of ll has been reportedza. No yield has been reported for this reaction since the resultant diamine was too unstable to be isolated. 0n the other hand, this amine has been obtained by Curtius degradation of a-truxillic acid21, and subsequently -14- converted to the bjgjdinnthyl)analog 1% in 66% yield31 by using the Eschweiler-Clarke32 modification of the Leuckart reaction,which consists Vof reductive amination of formaldehyde by formic acid. It occurred to us that combination of the reduction of the dinitro compound l3 and alkylation of the diamine therefrom with formaldehyde and formic acid is possible by using formic acid both as an acid in combination with the zinc and as reducing agent for conversion of imines obtained from the reaction of formaldehyde with the produced amine. Thus when the dinitro compound 13 was treated with a large amount of formic acid and zinc dust in the presence of formaldehyde, a good yield (50-80%) of l,3-diphenyl-2,4-§i§;(dimethylamino)cyclobutane(l§)was obtained which in most cases didn't need purification. The reaction of zinc dust and farmic acid is very vigorous and formic acid has to be added dr0pwise with initial heating to start the reaction. In one instance there was seemingly no reaction until all of the acid had been added. Then,5-l0 minutes after the addition was finished.a vigorous reaction took place which can be called an explosion. The purity and yield of lg was proportional to the amount of zinc dust and especially of formic acid For an 80 g run of 13, 600-800 ml of formic acid is necessarv for the best results. The impurities are usually unnamethyl, bjsfimonomethyl), and 2-dimethylamino-4-monomethylamino compounds. In cases when purification was found necessary the best method was fOund to be chromatography of the mixture on alumina rather than sublimation asxwwwkush/ reported31. l§ is eluted with hexane and crystallizes as it comes out of the column. It could be used without further purification (mp ll9-122°). One recrystallization from hexane gives pure lg. Attempted reductive -15- alkylation of dinitro compound 13 by hydrogenation in the presence of formaldehyde failed33, and only with Raney Nickel catalyst a yield of l% of lg was obtained. l§ shows an interesting doublet at 2800 cm’] 1 which is due to C-H stretching of methyl groups attached and 2850 cm- to nitrogen3“. Preparation of bis—finethiodideXlg) White and Dunathan prepared this compound from the reaction of lg with excess methyl iodide in methanol for 7 days in 77% yield . It was found that reaction goes much faster and cleaner in dry acetone with initial brief heating on a steam bath to give an 85-90% yield in 20-30 minutes. The methiodide obtained has a much higher melting point and recrystallization is not necessary. Reductive Deamination. Cleavage of'bis-hmthiodide)lg was accomplished by lithium in liquid ammonia with isopropanol as a proton source”. A mixture of l,3-diphenylcyclobutane (w and some other hydrocarbons was obtained in which 1% is recognized by its characteristic peaks in the nmr spectrum of the mixture. The total yield of hydrocarbon is 50-55%. Isolation of 1% by conventional methods was not possible and the mixture was successfully used in subsequent syntheses as a source of cyclobutane lg. The hydrocarbon impurities are perhaps the product of side reactions of intermediates lland lg such as those illustrated. -16- + N(CH3)3 + N(CH3§ + l(CH3)3 h 'h 'h Ph + ’ - H I 2e>'——>l—-> 1% Ph i Ph Ph @1013)3 Ph 11 T ,(D J/NHZG \L N"? Ph Ph Ph Ph N(CH ) Ph P. 1 (+3 3 3 j‘ dimers and Ph 7 Ph / . polymers <5 ;/ ‘E_J Preparation of l,3-dibromo-l,3-diphenylcyclobutane (1,25). In order to avoid preparation of the unstable l,3-diphenylcyclobutene, it was decided to convert cyclobutane lg to dibromide lg, and then either selectively dehydrohalogenate one of the bromines, or alternatively substitute one of the bromines with another functional group (such as methoxide) and then dehydrohalogenate the other bromine. Dibromination of pure 1% yielded two different isomers. Analysis of both indicated lefmlir2 and both behaved similarly in subsequent reactions. The isomer (a) mp l45-l45.5° was first to crystallize out of an ether solution of the mixture in the form of needles. It shows a simple nmr pattern: a ten proton multiplet at 1 2.6 and a four proton singlet at r 5.85. The other isomer (b) could not be obtained pure but was always contaminated with a bit of the isomer (a) and was analyzed as such. It shows the same pattern in the nmr, only the phenyl multiplet is shown at T 2.7 and the methylene protons at T 6.0. Since the bromo compounds are decomposed on -17- heating and because subsequent reactions showed that both react to give the same product in different reactions, no further attempts at the separation yam made and the products were assumed to have structures 123 and 12b. Ph Ph Ph Br Ph 3 l9 (a and b) The nmr is unexpected because it suggests that both of the protons of the methylene groups in both isomers are magnetically equivalent. It could only be said that the chemical shifts are fortuitously the same. Both isomers are unstable and decompose in air and on being heated in solution. Attempts at conversion of dibromocyclobutane lg to cyclobutenyl derivatives. Selective substitution or dehydrohalogenation of one of the bromines in 423 or the mixture of £28 and $35 was unsuccessfully tried by reacting them with sodium methoxide in methanol. sodium methoxide in THF, sodium bicarbonate in THF, lithium chloride in DMF, sodium iodide in THF, sodium iodide in methanol or potassium carbonate in DMSO. In all cases the disubstimted compound or starting material or intractable polymeric material was obtained. Preparation of l,3-diphenylcyclobut-Z-enylcyanide (2Q) In a review of the attempted reactions it was decided that all of the reagent used are either strong bases and poor nucleophiles or weak bases. -13- Consequently they are either unable to effect a reaction or they dehydrohalogenate both of the bromines or displace both of them. It seemed necessary to find a reagent which is a good nucleophile and a moderate base. Sodium cyanide is a good nucleophile but in aqueous solvent it is a strong base due to hydrolysis, on the other hand, it does not dissolve or ionize in most organic solvents. Dimethyl sulfoxide seemed to be the solvent of choice since it has a high ionization power and it was found that it also dissolves sodium cyanide. Reaction of the dibromide 123 or the mixture of 123 and 19p with sodium cyanide in dimethyl sulfoxide gave an 80-90% yield of the title compound (20) which was easily purified by chromatography over Fluorisil. 20 is eluted with hexane as a yellow oil. Its ir spectrum shows a band at 2220 cm"1 characteristic of cyanides (-CsN), and its nmr shows a ten proton signal at r 2.70 as a multiplet, a sharp one proton singlet assigned to HA at r 3.72 and an AB quartet worth two protons centered at 6.79, JAB = l3 cps assigned to H and H . C B cu Ph HC 32 Attempted converSion of the cyanide 20 to the corresponding amide or acid. Hydrolysis of cyanide 20 by alkaline hydrogen peroxide36 at room temperature was unsuccessful and a mixture of starting material and intractable material was obtained. A reported method for hydrolysis of nitriles with manganese dioxide in methylene chloride37 at room temperature fer 60 hrs failed to convert the cyanide into the corresponding -19- amide, but resulted in substantial decomposition to intractable materials. Basic hydrolysis with sodium hydroxide or acidic hydrolysis with hydrobromic acid were found unsuccessful. Preparation of l,3-diphenylcyclobut-Z-ene carboxamide(gD. A literature survey showed that at least in one case a quarternary nitrile had been converted to the corresponding imido-ester when treated with sodium ethoxide38. The method was used to convert the cyanide 20 to\amide 21. The cyanide 29 was added to a solution of sodium methoxide in dimethylsulfoxide. After 48 hrs, aqueous workup and extraction with ethergave the amide 21in good yield as a solid: 1 is characteristic of amides. This is substantiated by a doublet at 3360 and 3500 cm']. mp l43-l45% carbonyl absorption at l685 cm- The nmr spectrum of 21 is similar to that of 20, it shows a ten proton multiplet at r 2.7, a one proton singlet at 1 3.3l, a very broad two proton peak at r 4.25 and a two proton AB quartet centered at 1 6.68, JAB = l3.l cps. The broad peak at 4.25 was assigned to the protons of the -NH2 group. Later we learned that conversion of cyanides to amides has been reported by the same method39. C“ T 6.67 CONHZ T 4.25 Ph [ J‘Ph CH30Na :> DMSO ______ p Ph 1 3.31 -20- Attempted hydrolysis of the amide. An attempt was made to hydrolyze the amide 21 with dry Na0H in dimethylsulfoxide39. Acidification of the products gave an acid which was not the one expected (22) as judged by spectroscopic pr0perties. All other attempts to convert the amide to the corresponding acid or methyl ester failed. Theirattempts included reaction of the amide with nitrous acid in fluorosulfonic acid at -78° and treatment of the product with sodium methoxide in methanol, whereby the amide was recovered along with polymeric material. Reaction of the amide with isgfamylnitrite in the presence of sodium methoxide failed. ONHZ 200:1 ~ Ph Ph Ph a. 27% Preparation of l-methoxy-l,3-diphenylcyclobut-Z-ene 25. The reaction of sodium cyanide in DHSC with the dibromide made us think that thereius a possibility that dehydrohalogenation of one bromine could be combined with substitution of the other. Therefore the di- bromide and methanol were mixed in DMSO. This reaction was not successful since after 5 hrs, only a mixutre of dimethoxycyclobutane and starting material was obtained. Sodium iodide has been used successfully as a catalyst for displacementbo, and in a steroid system a l,3-dibromo compound 23 was converted to gg“l. -21- CH Br L\\ NaI “I Br H éfi Sodium iodide and small amounts of methanol were dissolved in dimethylsulfoxide and the bromide 12 (mixture of both isomers) was added to the solution. After 8 hrs, aqueous workup and extraction with ether gave an oil obtained in almost quantitative yield. The oil showed an nmr spectrum consisting of a ten proton multiplet at r 2.75, one proton singlet at r 3.35, a sharp three proton singlet at r 6.82, and a two proton AB quartet (JAB = 13 cps) at r 7.02 which was consistent with the structure 25. The ether peak in the ir spectrum is shown at lll7 cm". The mass spectrum shows a parent peak at 236 with fragments at 22l and l9l. The l9l peak was observed in the mass spectrum of almost all cyclobutene derivatives and could be due to diphenylcyclopropenium ion. The product of the reaction is pure and no further purification was needed. Br OCH 3'“) CH30H .h NaI A); I wso / Ph [Hi 8r g§ 12 Preparation of l,3-diphenylcyclobut-Z-ene-l-ol 26. When it was observed that the reaction of methanol and dibromide in the presence of sodium iodide in dimethyl sulfoxide yielded the methoxy -22- cyclobutene 25, it was reasoned that water could be used in the same manner to yield the cyclobutenol 26. Thus a mixture of the isomers l2 Bl“ .I 0” L20 . P“ Hal \ / Ph Dhso IVA/Q Ph Ph 32 (a and b) was treated with water in dimethylsulfoxide in the presence of sodium iodide. Workup yielded the solid cyclobutenol 26 in good yield. Since the compound decomposes rapidly no analysis could be obtained. The nmr spectrum of the compound consists of a ten proton multiplet at r 2.65, a one proton singlet at r 3.4, a two proton singlet at r 6.8, and a one proton peak at r 7.45 which disappeared when the sample was shaken with 020. The ir spectrum shows a hydroxyl stretching vibration at 3450 cm']. Structure 23 was assigned based on the Spectral evidence, and the nmr spectrum of the cation produced when the alcohol is treated with fluorosulfonic acid, which is identical with that produced from methoxycyclobutene 25. As stated, alcohols of this type (21, 28) are known to be very unstable and decompose to Open chain ketones“2. Ph h Ph Ph OHC H H0 31 P. Ph [I '” CH h \V Ph (3 -23- Preparation of l,3-diphenylcyclohexenyl cation (22). This cation was prepared by direct addition of 1,3-dipheny1-1,3-cyclohexadiene lg to fluorosulfonic acid at -78° under a nitrogen atmosphere in a concentration range of 15-20%. The nmr spectrum of the cation was determined at a temperature range of -60° to +25°. The nmr signals at lower temperature were broad due to viscosity of the solution. They started to show the expected pattern only at -20° and higher temperatures. Ph .0 a A / Ph 10 29 «A. M. At room temperature the phenyl protons of cation 29 appeared at r 1.77, 2.41 and 2.14 assigned“3 to ortho, meta and para protons respectively (Fig. 2). The rest of the spectrum is summarized in Table I. Preparation of l,3-diphenylcyclooentenyl cation (31). when 1,4-diphenyl-l,3-cyclopentadiene 6 was dissolved in fluorosulfonic acid cation (31) was formed. Protonation of 6 should give rise to cation (32). .0. > t9 ‘0 6 31 x- 1/ - 3;“ -24- Conversion of 22 to 28,cou1d occur by two 1,4 hydride shifts, 1,4 1:‘ shift 1,4 H-Shift —_> ,__._. Q 6"— p. 9 h Ph Ph Ph Ph .32 21 The nmr spectrum of this cation was determined over a temperature range of -60° to room temperature. The problem of broad signals at lower temperature was encountered in this case also. The cation exhibited the expected pattern for a l,3-diphenylallylic cation at temperatures above -20° (Table I) (Fig. 3). Cations 22 and 22 had identical phenyl absorption patterns with that reported for diphenylcyclopropenium cation“3. Preparation of l,3-dipheny1cyclobutenyl cation (22). This cation was prepared by addition of a Freon 114 8, solution of methoxycyclobutene 22 to fluorosulfonic acid at -78°. It was alternatively prepared by addition of a methylene chloride solution of cyclobutenol 22 to fluoro- sulfonic acid at -78°. The cation 22 prepared from either precursor gave OCH3 EWPh Ph/ E; \ Ph +1 Ph 5A -25- the same spectrum, except that the cation prepared from methyl ether 25 exhibited a singlet due to methanol produced in ionization of the methyl ether. The cation is stable at temperatures of -70 to +35°. At -70° the nmr showed a broad two proton doublet at r 1.46 and a nine proton multiplet at r 1.96, a singlet at 1 1.74 is merged with the signal at T 1.96 (Fig. 4). 0n warming, the lower field signal collapsed, but the rest of the.spectrum including the singlet at 1.74 sharpened up (Fig. 5). At about room temperature the lower field signal also sharpened up to a doublet of doublets at higher field than the original signal. The singlet previously observed at r 1.74 appears as a shoulder on this doublet of doublets. At room temperature (Fig. 6) the Spectrum consisted of a 5 proton doublet of doublets at T 1.74 with area ratio of ga_ 1.5:3.5. The singlet one proton signal at r 1.87,which appears as a shoulder in the second portion of the doublet of doublets, is responsible for increasing the area of the higher field doublet. Therefore the four proton signal at r 1.75 was assigned to the grghg_protons and the singlet at 1.87 was assigned to the vinyl proton HA. The pgr§_and meta_protons appear as a six proton multiplet, the signal centered at r 1.98 was assigned to the parg_protons and the signal centered at 1 2.18 was assigned to the meta protons on the basis of the near identity of the spectrum with that calculated for diphenylcyclopropenium ion“3. Collapse of the two proton doublet at r 1.46 (at -70°) on warming and resharpening of the signal as a four proton doublet of doublets was assumed to be due to the phenyl groups not being allowed to rotate freely at low temperature, which would make the two grthg_protons of each ring magnetically different. The two in the proximity of the allylic system probably appear at lower field. Warming would allow rotation to increase to a frequency -26- comparable to that of the nmr transition. At room temperature the two ortho protons in each ring have the same average magnetic field because of free rotation and the position of the peak is averaged out as shown by the signal at higher field at r 1.75. Comparison of cations 22, 22, and 22. Comparison of the nmr spectrum of the three cations shows a small, but perhaps significant downfield shift of the phenyl protons in going from system 22 to 22 while the methine proton of the allylic system shows a gradual upfield shift (with the exception of the special case of the cyclopropenyl cation) (Table I). Table I: Chemical Shifts for l,3-diphenylcycloa1kenium ions (T) Cation/H ortho para meta methine Other protons cyclohexenyl 1.77 2.14 2.41 1.46 6.42 (t,J=6.5 cps,4H) 7.64 (q,J=6.5 cps,CH) cyclopentenyl 1.76 2.10 2.35 1.51 6.04 (s, 4H) cyclobutenyl 1.75 1.98 2.18 1.87 5.82 (s, 2H) cycloprOpenyl“3 1.55 1.92 2.10 -0.45 If the bicyclobutyl cation 20 were a major resonance contributor the reverse trend would have been expected. Namely the closer C1 and C3 get the morel,3-over1ap would be expected. This would lead to less participation of the phenyl in the stabilization of the cation and -27- concentration of more positive charge on C2. As observed this was not the case. An alternative is dismissal of the presence of such contribution and distribution of the charge as expected for the 1,3-diphenylallylic system 22. This, however, doesn't explain the low field shift trend from 22 to 22 and upfield shift of the methine proton from r 1.46 to 1 1.87. Although the shifts are small, nevertheless, they exist. The fact that the trend is correlated with a decrease in ring size suggests the possibility that there is an electrostatic repulsion between the positive charges which tends to disgerse them as far away from each other as possible. This, however, doesn't exclude the possibility of the presence presence of 1,3-n overlap. It only implies that if there is such overlap it is not the predominant factor in determining the chemical shifts. The upfield shift of the methine protons could be the result of the altered geometry in the smaller ring accompanied by less deshielding of the methine proton by the phenyls. The 1,3-dipheny1 is not the best system to study the possibility of 1,3-n overlap. Thus the predicted delocalization energy gained tv this process (4.1 - 10 kcal/molel) is perhaps not sufficient to balance any -28- loss of the delocalization energy contributed by the phenyl rings. A better system perhaps would be 1,3-dipheny1-2-p;methoxypheny1cyc1obutenyl cation 22 where the p-methoxyphenyl could stabilize the positive charge placed on C2 by 1,3-n overlap. .. Q .. O OCH3 The fluorine nuclear magnetic resonance deshielding of Effluorophenyl carbonium ions has been shown to be related linearly to the stabilization energy of para-substituted triphenylnethyi cations““. A series of unpublished results from this laboratory have confirmed p;f1uorine is a more sensitive probe for determination of charge distribution in the phenyl ring as compared with the g-proton”5. Thus, in order to avoid a probable uncertainty caused by the small change in chemical shift of the p-protons, we suggest that the p-fluorophenyl analogues of these cations be used in further studies. EXPEPIHEHTAL mmmwmmmmm (a) Londensation- Ethyl acetoacetate (26 g, 0.2 mole) in dry benzene (250 ml) was added to sodium hydride (9.6 g of 50% dispersion in oil) in benzene and the mixture was stirred under a nitrogen atmosohere. After the completion of the reaction (indicated by cessation of bubbling of hydrogen), phenacyl bromide (40 g, .20 mole) was added at once and the mixture was refluxed for six hrs. The reaction mixture was then washed with cold water twice (total of 800 ml). The benzene solution was dried over magnesium sulfate and evaporated to give an oil the nmr spectrum of which exhibited peaks characteristic of ethylbenzoylpropionate and l-pheny1-3-carbethoxypentane- 1,4-dione 2. b) Saponification and decarboxylation. The product of the above reaction was placed in a three liter flask, an 0.8% solution of potassium hydroxide (2.5 1) was added, and the mixture was refluxed under a nitrogen atmOSphere for 2.5 hrs. The reaction mixture was then cooled and extracted with ether. The ether solution was washed with water and dried over magnesium sulfate. Evaporation of the ether gave a brownish oil which was chromatographed on an alumina column. The diketone 3 (l-phenylpentane—l,4-dione)(7g) was eluted first with ethyl acetate-benzene (20:80). A mixture of the diketone and the cyclopentenone (7.5 g) followed by the cyclooentenone (4.5 g) was eluted with ethyl acetate. The diketone was separated from the mixture 1 l by distillation: (bp 104° at 0.35 mm); ir 1720 cm- , 1(95 cm- : nmr 2C -30- (CC14) 2.2 (m, 2H), 2.68 (m, 3H), 7.16 (A282, q, 4H), 7.97 (s, 3H). 3-Phenylcyclopentenone (2 g) was obtained from the residue by crystallization from hexane-ether. c) Ring closure. The combined diketone crop (12 g) was refluxed in aqueous potassium hydroxide (250 ml of a 10% solution) under a nitrogen atmosphere for 2 hrs to give 9.5 g of a light brownish phenylcyclopentenone isolated by chromatography as above, giving a total overall yield of 5 %. The pure ketone 1.15 obtained by sublimation of the crude product: mp 83.5 - 84.5°; ir (cc14) 1705-1695 cm“; nmr (coc13) 2.6 (m, 5H), 3.6 (t, J = 1.5 cps, 1H), 7.05 (m, 2H), 7.55 (ng 2H). WWWWL Ketone .1. (3.16 g, 0.02 mole) in ether (20 ml) was added to phenyl Grignard prepared from bromobenzene (3.3 g, 0.021 mole) and magnesium turnings(50 mg) in ether under an argon atmosphere. After 3 hrs the reaction was worked up by adding concentrated ammonium chloride solution. The ether solution was washed with water and evaporated to give an oil, which was dissolved in warm methanol. When the methanolic solution was chilled,2.9 g (35%) of solid was obtained: mp 155.5 - 157°(lit8 155-157); nmr (CDC13) r 2.6 (m, 10H), 3.12 (t, J = 1 cps,2H), 6.27 (t, J = l cpg,2H). meme/ewe (a) w___chae1 Condensation. This reaction was done according to Walkerlo. A 40% methanol solution of Triton B (60 ml) was added to a stirred solution of ethyl benzoylacetate (56.75 g, 0.295 mole) and methylvinyl- ketone (20.75 g, 0.296 mole) in t-butyl alcohol (75 ml) at 0°. The solution became brownish and then green. After 5 hrs the reaction was worked up according to Walker to give a mixture of crystalline material -31- and oil. The mixture was triturated with ether and filtered to give 39.5 g of a crystalline compound 2, mp 122-125°, which, when recrystallized from methanol, had mp 125-127° (lit.10 128-130°); ir (Nujol) 3400 cm'], 1707 cm", 1693 cm"; nmr (CDC13) r 2.7 (m, 10), 5.79 (b.s., 1H), 6.05 (q, 2H), 7.4 (s, 2H), 7.55 (m, 4H) and 9.02 (t, 3H). The filtrate when evaporated yielded 29.7 g of a yellow brown oil, 2. b) Decarboxylation of the condensation products. The solid obtained from above (31.5 g) was treated with an 8% aqueous potassium hydroxide solution (250 m1) and refluxed for 12 hrs. The mixture was then cooled and extracted with ether, which was washed with water and dried over magnesium sulfate. Evaporation of the solvent resulted in 19.1 g of the product which was recrystallized from hexane-ether (80:20) to give 17.5 g of the ketone: mp 64.5-65.5°(1it. 64-65°); ir 1665 cm“: nmr (CDC13) T 2.7 (m, 5H), 3.64 (t, J = 1.5 cps, 1H), 7.13-8.l (complex m, 6H). The brown oil (29.1 9) obtained in (a) was refluxed in 8% potassium hydroxide (250 ml) to give the desired ketone (7.5 g) as a brownish solid. This crude product when recrystallized from hexane-ether (80:20) gave 6.1 g of the pure ketone 2. The total yield of pure ketone was 23.6 g (46.5%). WMWMWMW) B-Phenw-Z- cyclohexenone (6 g) in dry tetrahydrofuran was added to a Grignard Reagent prepared from bromobenzene (18.3 g) and magnesium turnings (2.8 g) in dry ether (150 ml) under an argon atmosphere. The reaction was allowed to go on overnight and then worked up by addition of a concentrated solution of ammonium chloride. The organic layer, after drying over magnesium sulfate, was evaporated to give 5.7 g of a thick yellow oil which was taken up in -32- 95% ethanol and the solution chilled to give white platelets mp 89-93°. Recrystallization from 95% ethanol resulted in 4.9 g (66%) of the same type of crystals: mp 95-97° (lit.ll 98-99°); nmr (CDC13), 2.67 (m, 10H), 3.33 (sextet, J = 1.5 cps, 1H), 3.9 (dt, J = 1.5 cps, J = 4.5 cps, 1H), 7.4 (m, 4H). t i ' t' - ' ' ° . Practical cinnamic acid (145 g) was placed in a 5 liter two neck round bottom flask which was used as the photolysis apparatus. The two joints are 5 55/60 and 24/40 size. Water (500 ml) was added to the cinnamic acid, and the suspension was stirred mechanically to obtain a more thorough suspension. A Pyrex photolysis insert was introduced into the wider joint. Water (4 liter more) was added through the side joint and a mechanical stirrer fitted with a Teflon paddle (1 inch long) was inserted in such a way as to avoid contact with the jacket or the walls of the flask during stirring. A football shape magnetic bar was used to stir the suspension from the bottom. The suspension was irradiated with a 550 watt Hanovia lamp. The cinnamic acid gathered on the top at first, but a very good suspension was fermed gradually as the reaction proceeded. After 5 days the reaction was stopped. The suspension was filterd and air dried on the Buchner funnel. The dried solid was washed several times with ether (total 1200 ml) to remove the unreacted cinnamic acid and dried in an oven at 60° for one hr. The solid was placed in a one liter beaker, triturated with boiling ether, and filtered to give 81.5 g of a-truxilic acid 22 (56.2%) mp 280-285° (lit.l 280-281.5°). The crude material gives 72 g of pure acid, mp 290-29l°. -33- 222 ti t ill 1 h ' . This compound was prepared according to E. H. White21. Treatment of truxillic acid (29.6 g, 0.1 mole) with thionyl chloride (300 ml) yielded crystalline acid chloride. Recrystallization from benzene-pet ether (60:40) gave white prisms (26.7 g,81%): mp 121-123°; ir (Nujol) 1785 cm“. 222222222822222222222221eertruéillets ”2422). Truxilloyl chloride (22 g) was mixed with t-butylhydroperoxide (40 ml) in dry benzene, under a nitrogen atmosphere. The mixture was cooled in an ice-salt bath to 0° and dry pyridine"7 (25.0 ml) was added dropwise. The reaction mixture was allowed to stir for 12 hrs. The contents of the flask were then washed with 5% hydrochloric acid in ice water to remove the excess of pyridine. The unreacted acid and traces of hydrochloric acid were removed by washing with potassium hydroxide solution (150 ml of a 5% solution). The organic layer was dried over magnesium sulfate and solvent evaporated under reduced pressure at room temperature. The solid residue (25.5 g) was recrystallized from ether-hexane (1:2) (150 ml). The resulting needles weighed 23.7 g (81.5%): mp 142-142.5°: ir (Nujol) 1750 cm“. i t' i- - ' t . Two samples of perester (200 mg each) were dissolved in chloroform (15 ml for each). Ferric chloride solution in acetic acid (20 m1 of a 0.002% solution) and sodium iodide (2 g) was added to each“8. Both samples were kept in the dark for 10 minutes and then immediately titrated with 0.10 solution of sodium thiosulfate to the end point as indicated by starch indicator. A blank solution containing the same ingredients except for the perester was run. The two samples consumed 18.5 and 18.4 ml of standard thiosulfate and 1.1 m1 of thiosulfate was necessary for the blank. mole perester —-——5> 212 —-——;> 4 nmle Na2503 17719- x 10' = 4.35x10‘ 4 4 mole of perester 4.35 x 440 x 10- 0.1914 g of pure perester in 0.2 g of crude perester Qfil§%fl- X 100 = 95.7% active perester Wmam The perester (22.5 g) was su5pended in p-cymene (600 m1 of solvent, distilled at 207.5-208.5°) and the suspension was heated under a stream of nitrogen with stirring. The perester dissolved at around 55°. At 130° the solution turned yellowish and effervescence began which became vigorous at around 150°. After 1.5 hrs the bubbling stopped, but heating was continued for 12 more hrs. The yellowish solution was then attached to a 36” vigreaux column and the balk of the g-cymene was distilled (at 35°, 0.5 mm). The yellow pasty residue was taken in ethanol and chilled. The p;cymene dimer which crystallized was filtered off and the mother liquor was stripped of sovlent and distilled through a 4 inch Vigreaux column. g-Cymene was distilled first. The fraction which distilled at 125-140°, 1.00 mm was rich in the desired product (70-80% as judged by nmr). The residue of distillation also contained 20-30% of the product. The above mentioned fraction and the residue were combined and redistilled to give 5.8 g of an oil (110-123° at 0.25 mm) containing 90% of the cyclobutane 22. -35.. Column chromatography over silicic acid yielded 4.5 g (42%) of pure hydrocarbon as a thick oil. Recrystallization from methanol yielded needles: mp 40-41°; ir (film) 3030, 2980, 1610, 1500, 1450 cm"; nmr (CC14) 1 2.8 (m, 100), 6.4 (q, J = 7 cps, 2H), 7.48 (t, J = 7 cps, 4H). Anal. Calcd. for C C, 92.26; H, 7.74; 16H16‘ Found: C, 92.07; H, 7.60. RttB8t8Alenmeiglpéafliktzmaaleéa81RASAXAEXElRRAARARAAXAR)- 1,3-Dipheny1cyclobutane (2.2 g, 105 mmole) was dissolved in carbon tetrachloride (50 ml) and H-bromosuccinimide (4.3 g, 2.31 nnnle) and benzoylperoxide (0.2 g) were added. The mixture was brought to reflux by means of an oil bath. When all of the n-bromosuccinimide was consumed the succinnimide was filtered and the solvent was removed at reduced pressure. The residue was taken up in ether. The ethereal solution yielded needles (800 mg): mp 145-145.5° (dec); nmr (CDC13) 1 2.6 (m, 10H), 5.85 (s, 4H). Anal. Calcd. for C]6H]4Br2: C, 52.46; H, 3.83; Br, 43.71: Found: C, 52.62; H, 3.94: Br, 43.6. Addition of hexane to ethereal mother and cooling gave a solid (1.9 9) (total yield 69%) which was a mixture of the solid previously -obtained and another substance which showed nmr (C0013) 1 2.7 (10H), 6 (s, 4H). Anal. Calcd. for C16H14Br2: C, 52.46: H, 3.83; Br, 43.71; Found: C, 52.62; H, 3.82: Br, 43.62. Both isomers decompose when allowed to stay in air. -35- 22222222282222222222222222222. The title compound was prepared according to the organic synthesis preparation“€ In this preparation benzaldehyde (530 g, 5 moles), nitromethane (305 g, 5 moles) and sodium hydroxide (500 ml of a 50% aqueous solution) were used. The yield of crude material was 580 g (77.8%). Recrystallization from hexane yielded long needles: mp 56.5-57.5°(1it."7 57-58°). 22222222222222222222222222222222222. B-Nitrostyrene (76.5 g) was placed in an 18 inch open crystallization dish and water (800 ml) was added. A water bath cooled to 18°C by passage of tap water through copper coils installed in it was used to cool the suspension. A mechanical stirrer fitted with a 3 inch Teflon paddle was used to stir the suSpension. A 550 watt Hanovia lamp in a Pyrex water cooled jacket placed horizontally on the tap of the crystallization dish was used to irradiate the suSpension. After 24 hrs the irradiation was stopped and the suspension was filtered and washed by soxhlet extraction with hexane to yield 59 g (77%) of crude 1,3-dinitro-2,4-diphenylcyclobutane: mp 175-181°. Recrystallization from benzene gave 41.3 g (54%) of pure dimer 22: mp 183-185° (lit.28 181-182°); ir (Nujol) nmr (CDC13) 1 2.68 (m, 10H), 6.6 (A282 system, 4H). The dinitro compound 22 (45 g, 1.51 mole) was placed in a three liter round bottom flask and zinc dust (300 g) and formaldehyde (250 ml of a 37% aqueous solution) were added. The suspension was warmed on a steam bath and flormic acid (550 ml of a 88% aqueous solution) was added dropwise over a perjod of 4 hrs with vigorous stirring. As soon as the reaction -37- started the external heat was discontinued since heat of reaction was sufficient to maintain the temperature at about 80-90°. After the addition was completed the reaction was heated over a steam bath again for 6 hrs more, and formic acid (200 ml in 2 portions) was added to ensure completion of the reaction. The insoluble material was then filtered and the filtrate was concentrated under reduced pressure at 60°. The concentrated solution was extracted with benzene and the aqueous phase was made basic with sodium hydroxide solution. A white precipitate formed which was filtered and the solid was washed with ether. The filtrate was also extracted with ether. The combined ethereal solution was washed with water, and dried over magnesium sulfate. Evaporation of ether yielded 38.5 g of a white solid, which was recrystallized from ether-hexane (20:80) to give 35.5 g (79.9%) of pure 22: mp 125-127° with spectroscopic preperties identical to those reported below. Truxilloyl chloride (26.7 g, 0.08 mole) was treated with activated sodium azide (38.0 g, 0.58 mole), in dry benzene (200 m1) according to White31 To the di-isocyanate (ir Nujol 2280) obtained in this manner formic acid (400 ml of 85% solution) and formaldehyde (120 ml of 35% aqueous solution) were added. The mixture was heated at 60° for 24 hrs. The solution was then cooled and made basic with sodium carbonate solution and was extracted with ether. After removal of ether a solid material was obtained which was sublimed. Recrystallization of the sublimed material from ether hexane (20:80) gave large crystals: (10.6 g, 45.8%) mp 124-125° (lit.3l 122-123°); ir (CHC13) 2990, 2950, 2850 and 2800 cm']; nmr (CDC13) 1 2.75 (m, 10H), 6.7 (A282 q, 4H), 8.25 (s, 12H). Truxillic acid (2 g) was recovered from acidification of the basic aqueous solution. Preparation of l,3-diphenyl-2,4-bj§;(dimethylamino)cyclobutane- WWWWW 22W) 2,4-Dipheny1-1,3-_bj_s_(dimethylamino)cyclobutane (6.7 g) was dissolved in anhydrous acetone (50 ml) with a little warming. Methyl iodide (10 ml) was added and the mixture was warmed for 2 minutes. The flask was stappered and allowed to stand for half an hour. The dimethiodide was filtered and proved to be pure without any need for recrystallization: mp ZZZ-225° (lit.31 215-217°). '5333553ve deamination of bis-methiggigsvlg. bis-Methiodide‘29 (14 g, 0.024 mole) was added to distilled liquid ammonia (400 m1) at -78°, isopropanol (3.6 ml) was added and lithium ribbon (washed with pet ether to remove the oil) was cut and added in pieces until the blue color persisted for 2 minutes. The mixture was then allowed to stir at -78° for 30 more minutes.and then ammonium chloride (10 g) was added. Ammonia was then allowed to evaporate and the residue was taken in ether, and washed with water and then hydrochloric acid solution. The ether solution was dried and evaporated to give a yellow fluorescent oil. The oil was distilled (91° at 0.15 mm) to give 3 g of a clear liquid, the nmr spectrum of which showed peaks characteristic of 1,3-diphenylcyclobutane plus some other material. The mixture could not be separated by gas liquid chromatography, column chromatography or fractional distillation, and was successfully used in the preparation of dibromide 22. -39- 2222222222222222222222222. The dibromo cyclobutane (A) (150 mg) was dissolved in methanol (100 ml) and sodium iodide (l g) was added. The solution was stirred under a nitrogen atmosphere. After 8 hrs the mixture was poured inuicold water (500 ml) and the aqueous solution was extracted with ether. The ethereal solution was washed with water and dried over magnesium sulfate. Evaporation of the ether afforded a yellow oil (110 mg). The product showed spectral properties similar to‘UKBe of the product of the reaction of dibromocyclobutane with sodium methoxide (see below). The dibromocyclobutane (A) (200 mg) was dissolved in dry tetrahydrofuran (50 ml) and sodium bicarbonate (300 mg) was added. The mixture was stirred under a nitrogen atmosphere and maintained at 40-50° by means of a water bath. Filtration of the mixture and evaporation of the solvent yielded starting material (180 mg) which had spectral properties identical with thus of an authentic sample. mmmmmmmwwwmm The dibromocyclobutane (22) (100 mg) in methylene chloride (5 ml) was added to a mixture of antimony pentafluoride (0.2 ml) and fluorosulfuric acid (1.1 ml) at -78° under an argon atmosphere. The nmr Spectrum of this solution over a range of temperature (-70° to -10°) shows a multiplet l.7-2.6 ppm upfield from fluorosulfuric acid signal, and a singlet at 7 ppm upfield from the same signal in an area ratio of 15:2. -40- WWWMMW The isomer (A) (150 mg) was dissolved in dry methanol (25 ml) and powdered sodium methoxide (350 mg) was added. The mixture was stirred under nitrogen for 5 hrs at room temperature. The reaction mixture was then added to water (200 m1) and extracted with ether (250 ml). The ethereal solution was washed with water and dried over magnesium sulfate. Evaporation of the ether gave a yellowish oil whose nmr spectrum included signals characteristic of l,3-dimethoxy-l,3-diphenylcyclobutane. . E I. C 1 3‘1. I J i 1 l (‘6' J .184211 1,3-Dibromo-l,3-diphenylcyclobutane (2 g, 5 mmole) (A) was dissolved in acetonitrile (150 ml) and dimethylsulfoxide (60 m1) under a nitrogen atmosphere. Sodium cyanide (2 g) was added to the solution and the mixture was stirred at room temperature for 10 hrs. The milky yellow solution was then poured into one liter of water and extracted with ether (twice 300 ml). The ethereal solution was washed with water several times and then with saturated sodium chloride solution and was dried over magnesium sulfate. Evaporation of the solvent yielded a brownish oil which was chromatographed on Fluorisil. The desired product was eluted with hexane as a yellow oil (1.03 g, 81.66): ir (neat) 2220 cm"; nmr (CC14) 2.70 (m, 10H), 3.72 (s, 1H), 6.79 (A282 q, J = 13 cps, 2H). In another run 4 gof a mixture of isomers A and B of the dibromocyclobutane gave 2.15 g, of the product. -41- Sodium methoxide (commercial powder 1.2 g) was dissolved in dimethyl- sulfoxide (50 ml). Compound 22 (1.1 g, 4.7 mmole) in dimethylsulfoxide (5 ml) was added. The solution was stirred at room temperature under a nitrogen atmOSphere. After 48 hrs, the solution was poured into 800 m1 ice water and extracted with ether. The ether solution was washed with water several times and dried over magnesium sulfate. “19111219 e‘ulersolution was concentrated to 15 m1 and hexane was added a white precipitate formed which was filtered and recrystallized from hexane-ether to give 0.872 g (73%) of a white solid: mp l43-145°;ir (CHCl3) 3500, 3360, 1685 cm"; nmr (CDC13) T 2.7 (m, 10H), 3.31 (s, 1H), 4.25 (broad, 2H), 6.68 (A282 q, J = 13.1 cps, 2H). WWmthWmZ/h god.- um iodide (3 g) was dissolved in dinethylsulfOxide (50 m1) and absolute methanol (1.5 ml) was added to the solution. l,3-dibromo-l,3-dipheny1- cyclobutane (1.5 g of the mixture of isomers) was dissolved in 5 m1 of dimethylsulfoxide and added to the solution. After a while a yellow color developed which gradually turned red-brown (iodine). After 12 hrs the solution was added to ice water (800 m1), and the cloudy solution was extracted with ether. The ether solution was washed with a dilute sodium sulfite solution to remove the iodine and washed several times with water to remove the dimethylsulfoxide. The ether solution was dried and evaporated to give 0.953 g (98.5%) of an oil: ir (CHC13) 1117 em“; nmr (cc14) T 2.75 (m, 10H), 3.35 (s, 1H), 6.82 (s, 3H), 7.02 (A282 q, J = 13 cps, 2H); m/e 236. -42- Cyclobutanedibromide (lg) (0.500 g) was dissolved in a solution of sodium iodide (1 g) and water (1 ml) in dimethylsulfoxide (50 ml) at room temperature. After 8 hrs the solution was poured into water (800 ml) and extracted with ether. The ether solution was washed with water and dried over magnesium sulfate. Evaporation of the ether yielded a solid with a broad mp range: ir (CHC13) 3600 cm']; nmr (CDC13) r 2.65 (m, 10H), 3.4 (s, 1P), 6.8 (s, 2H), 7.45 (b5, 1H); mass Spectrum: intense peak at 205 correSponding to 1,3-diphenyl- cyclobutenium ion. WWW:- (a) Preparation of y-phenyl-y-butyrolactone Q4) This compound was prepared according to Julia“8 from s-benzoylpropionic acid (252 9). After workup a colorless oil (207.4 g, 83%) was obtained: bp l42-3°, 2.5 mm; ir (neat) 1780 cm']; nmr (CC14) r 2.8 (s, 5H), 4.7 (dd, J = 8 cps, J a 7 cos, 1H), and 7.5-8.5 (complex m, 4H). (b) Preparation of ethyl-y-chloro-y-phenyl butyrate(3§) The lactone from part (a) (32 g) was treated with thionyl chloride (73 g) in dry benzene and the y-chloro acid chloride obtained was treated with acidic ethanol according to Julia. The yield of distilled product was 38 g (62%): ir (neat 1720 cm']; nmr (CC14) r 2.7 (m, 5H), 5.05 (m, 1H), 6.0 (q, 2H), 7.8 (m, 4H), 8.8 (t, 3H). (c) Preparation of ethyl -t_r;a_r§~2-phenylcyc10propylcarboxylate (36). The chloro ester obtained in (b) (26 g, 0.11 mole) was cyclized by treatment with a sodium t-amylate solution. Workup resulted in 12.3 g (56%) of -43- the product as a colorless liquid which solidified in the refrigerator to give crystals: mp 31-33°; ir (neat) 1719 cm']; nmr (0014) r 2.9 (m, 5H), 5.92 (q, 2H), 7.56 (octet, J1 = 7.5 cps, J2 = 6 cps, J3 = 4.5 cps, 1H), 8.03-8.8 (m, 3H), 8.78 (t, 3H). This ester was converted quantitatively to the corresponding acid by refluxing in methanolic potassium hydroxide for 5 hrs. The crude acid mp 84-86° on recrystallization from hexane gave colorless light crystals: mp 90-91°: ir (Nujol) 1700 cm"; nmr (c0013) 1 2.8 (m, 5H), 7.4 (octet J] = 10 cps, J = 7 cps, J3 = 4 ops, 1H), 7.95-8.85 (m, 3H). 2 (a) trans-Z-Phenylcyclopropylcarboxylchloride(38) The acid (20 g) was treated with thionyl chloride (40 ml) in refluxing dry benzene (100 ml) for 4 hrs. The thionyl chloride and benzene were then removed under reduced pressure and the residue distilled (104° at 1.5 mm) to give 18.05 g (88%) of the acid chloride as a yellow liquid (1). (b) Diphenylgggmium reaction. Dried cadmium chloride (39 g, 0.21 mole) was added to a Grignard Reagent prepared from bromobenzene (64.8 g, 0.42 mole) and magnesium turnings (9.8 g,0.48 m) in dry ether (180 ml). The mixture was refluxed for one hour with stirring. The ether was then distilled off, dry benzene (220 nfl) was added to the solid residue, and the mixture was heated to the boiling point of benzene. This mixture was cooled to 10° and tran§;2-phenylcycloprooylcarboxyl chloride (18.05 g, 0.109 mole) in dry benzene (15 ml) was added drapwise with vigorous stirring. The reaction mixture was then refluxed for 3 hrs. -44- The thick suspension was then poured on ice-dilute (10%) hydrochloric acid solution and extracted with ether. The ether solution was washed with dilute sodiunlhydroxide solution (to remove the acid) and washed with water and dried over magnesium sulfate. Evaporation of the ether yielded 15.2 g (66%) of the ketone 32 in the form of a liquid. An attempt to distill the product caused decomposition of the ketone. The liquid shows the following Spectrosc0pic properties: ir (neat) 1665 cm’]; nmr (C014) 1 2.15 (m, 2H), 2.7 (m, 8H), 7.3 (m, 2H), 8.2 (octet, J = 3 cps, J2 = 6 cps, J3 = 11 cps, 1H), 8.7 (sextet, Av = 3 cps, 1H). Crystallization of the ketone from hexane ether (80:20) yielded crystals mp 30-36° which had the same spectrosc0pic pr0perties. WWWWQl The title compound was prepared according to Corey“9. BenzalacetOphenone (31.2 g, 0.15 mole) in dry dimethylsulfoxide (75 ml) was added to dimethylsulfoxonium methylide prepared from trimethylsulfoxonium iodide (35.25 g, 0.16 mole) and sodium hydride (6.1 g of a 55% dispersion in oil) in dimethylsulfoxide (175 m1). workup according to Corey“7 yielded a yellowish oil which still contained the oil from sodium hydride. The product was taken up in acetonitrile and washed with hexane to remove the oil. Evaporation of the acetonitrile yielded an oil which solidified in the refrigerator. Recrystallization from hexane afforded a solid (29 g, 87%): mp 40-58°; ir (Nujol) 1670 cm"1 nmr (C014) 1 2.15 (m, 2H), 2.7 (m, 8H), 7.3 (m, 2H), 8.2 (m, 1H), 8.7 (m, 1H) 3 g of a yellow oil was obtained from the mother liquor with the same spectrosconic properties. 45 -46- Crude l-benzoyl-2-phenylcyc10prOpane (15 g, mp 39-48°) was added to p-toluenesulfonylhydrazide (27.9 g, 0.15 mole) in 95% ethanol (250 ml), and p-toluenesulfonic acid (1 g) was added to the mixture. The mixture was refluxed for 24 hrs. Concentration of the produced solution and chilling yielded 24 g of white solid mp l39-l46°. When this solid was recrystallized from 95% ethanol, it gave light needles, mp 152-155° (18.5 g, 68%). Anal. Calcd. for C23H22N2025: C, 70.76; H, 5.64; N, 7.18; S, 8.20; Found: C, 70.79; H, 5.74; N, 7.48: S, 8.56. Preparation of gig; and trans-a-Phenylcinnamic acid. The WWW/WWW mixture of gi§_and trans acid was prepared according to the organic synthesis51 preparation. A recipe of phenylacetic acid (109.2 g) benzaldehyde (81 m1) and triethylamine (80 ml) in acetic anhydride (160 ml) yielded the mixture of acids. The product was dissolved in an appropriate amount of sodium hydroxide and the basic solution acidified with acetic acid. The resulting precipitate was filtered, washed with water, and recrystallized from 95% ethanol to give the cis-a-phenylcinnamic acid 41 mp 169-173° (103 g) (lit.51 l72-l73°). Hydrochloric acid was added to the filtrate to lower the pH of the solution to 3 (as indicated by congo red paper). A precipitate formed which was filtered, washed with water, and recrystallized from hexane-ether (50-50) to give the tran§_acid a; (20 9) mp l35-137° (lit.51 1364137°) (total yield 68%). -47- e ti - l i it il 4 The title compound was prepared according to Organic Synthesissz. Benzaldehyde (106 g) and benzyl cyanide (117 g) were used as starting materials. The yield was 150 g of solid (73%) mp 86-87° (lit.52 88°). WWW/€280 "methyl sulfoxonium iodide (66 g, 0.3 mole) was dissolved under a nitrogen atmosphere in dry dimethylsulfoxide (400 ml) wiUi a little warming. Sodium hydride (15.5 g of a 50% diSpersion in oil, 0.3 mole) was added gradually with stirring. when the evolution of hydrogen ceased, a-phenyl- cinnamonitrile (48.2 g, 0.24 mole) in dry dimethylsulfoxide (150 ml) was added dropwise. After the addition of nitrile was over the temperature of the reaction was raised to 50-55° by means of a water bath 53. The reaction was continued for 2 more hrs. The reaction mixture was then poured on ice (200 g), to give a precipitate which was filtered and triturated with hot methanol. The methanolic solution was diluted with water and extracted with chloroform. When the solution was dried and the solvent evaporated, an oily solid was obtained which was recrystallized from hexane-ether (50-50) to give 34.2 g (65%) of solid mp 57-59°. Recrystallization from methanol raised the mp to 60-61° (lit. 58-60°51). The material which was not dissolved by trituration weighed 11.3 g and had mp > 300°. The purified product had: ir (nujol) 2230 cm‘]: nmr (CC14) r 2.72 (broad, s, 10H), 7.48 (dd, J1 = 7 cps, J2 = 9 cps, 1H) 8.09 (m, 2H). WWW The nitme (5.26 g, 0.024 mole) in diethylene glycol (5 ml) and potassium hydroxide solution (1.4 g in 10 m1 glycol) were mixed and refluxed under nitrogen -43- for 72 hrss“. The dark brown solution was then diluted with water and extracted with ether to remove the unreacted nitrile. Acidification and extraction with chloroform gave brownish crystals (1.4 g) (24.7%) of acid 45 mp 218-222°. Recrystallization from 95% ethanol gave colorless crystals mp 224-227° (lit. 220-222°2). In another run 17 g of the cyanide yield 5.8 g of the same acid. WWWMIEW). A solution of diazo methane in ether (300 ml) prepared from EXR101 (Dupont)55 was added to gl§;a-phenylcinnamic acid (30 g) in tetrahydrofuran (150 ml). After one-half hour the solution was washed with sodium carbonate solution and then with water and dried over magnesium sulfate. Evaporation of the solvent and recrystallization of the residual solid from hexane yielded 29.2 g (91.5%) of needles: mp 76-77°; ir (c014) 1705 cm"; nmr (cc14) r 2.33 (s, 1H), 2.86 (m, 5H), 3.05 (s, 5H), 4.37 (s, 3H). (8). Sodium hydride (6.54 g of a 56% dispersion in oil) was added to trimethylsulfoxonium iodide (33 g, 0.15 mole) in dry dimethylsulfoxide (150 ml) with vigorous stirring under nitrogen until hydrogen evolution stopped. A solution of methyl §j§;a-phenylcinnamate (28.56 g, 0.12 mole) in dimethylsulfOXide (100 ml) was added dr0pwise to the methylide over a period of one hour. The viscous solution was stirred for one hour more and poured on ice water. Extraction with ether and drying and evaporation of the ether yielded a yellow oil. The product still contained the oil from the sodium hydride which was removed as before by extraction of an acetonitrile solution of the product with hexane. Evaporation of -49- the acetonitrile gave the mixture of the methyl esters (20.5 g, 67.7%). Saponification of this mixture in methanolic potassium hydroxide (30 g, potassium hydroxide in 400 ml, 75% aqueous methanol) for 15 hrs gave after acidification a mixture of the acids which upon recrystallization from ethanol gave 5.81 g of a soliciQQ (acid A) mp 224-227°; ir (CHCl3) 3500-2500, 1700 cm']. Concentration of the mother liquor gave 8.65 of the acid $1 (8), mp 148-151° (lit.51 145-150°) which on recrystallization from ether gave colorless crystals: mp 153-155°; ir (CHCl3) 1708 cm'1: nmr (CDC13) r 3.3 (m, 1H), 7.4 (t, J = 8 cps, 1H), 8.05 (dd, J1 = 8 cps, J2 = 5 cps, 1H), 8.7 (dd, J1 = 5 cps, J2 = 8 cps, 1H). MWWWWW The acid (3.2 g) was dissolved in dry ether (150 m1) and added to a solution of lithium aluminum hydride (0.74 g) in 50 ml of ether. The mixture was refluxed with stirring for 12 hrs. After this time one m1 of a saturated solution of sodium sulfate in water was added. The precipitated aluminum salts were then filtered and the clear ethereal solution washed with water and dried over magnesium sulfate. Evaporation of the ether yielded the solid alcohol 48 (A) (2.8 g, 93.8%) which was “recrystallized from hexane to give feathery light crystals: mp 65-67.5°; ir (CHC13) 3650 cm" and 3500 cm" and 1050 cm"; nmr (00c13) T 3.05-3.25 (m, 10H), 6.5 (A8 quartet, J = 12 cps, 2 m), 7.4 (broad singlets, 1H), 7.75 (dd, J].‘ 6 cps, J2 = 8 cps, 1H), 8.6 (s, 1.5H), 8.72 (d, l/2H, Av = 2 cps). The signal at 7.4 disappeared when the sample was treated with deuterium oxide. Anal. Calcd. for cl6Hl6O: C, 85.68; H, 7.19; Found: C, 85.75; H, 7.31. -50- WWW Acid (8) (6 g) in dry ether (50 ml) was added tola suspenSion of lithium aluminum hydride (2 g) in dry ether (100 ml). The reaction was refluxed overnight. A saturated solution of sodium potassium tartrate was added (3.5 m1) and the mixture stirred for one half hour. The suspension was removed by filtration and the ether solution was washed with waters dried, and evaporated. An oily solid was obtained (4.4 g, 78.6%) which was crystallized from hexane to give cotton candy crystals (needles): mp l l 79-30°; ir (CHC13) 3500 cm" , 3450 cm", 1050 cm“ ; nmr (c0c1 3) r 2.7 (m, 10H), 6.7 (s, 2H), 7.58 (broad triplet, J = 7.5 cps, 1H), 8.6 (s, 1H), 8.68 (s, 1.5H), 8.76 (doublet, Av = 0.5 cps, 0.5H). The signal at 8.6 disappeared when the sample was shaken with deuterium oxide. Anal. Calcd. for c16H160: C, 85.68: H, 7.19; Found: C, 85.63: H, 7.27. '. 0‘ . ‘ ' . '. 1° -.. - .. - . -. Truxillic acid (10 g, 0.033 mole) and lead tetraacetate (46.6 g, 0.1 mole) were suspended in dry benzene (250 ml) under a nitrogen atmosphere. Pyridine (15 ml) was added and the mixture was placed in a 70-80° oil bath. A light yellow clear solution was formed. After ten minutes the solution started to become turbid and a precipitate was formed. After 1.5 hrs ethyl acetate (50 ml) was added and the reaction mixture was filtered. The filtrate was treated with a solution of sodium mgtg_bisulfite (200 ml of 3% solution). The white precipitate of lead oxide was filtered. The organic layer of the filtrate was separated and treated with sodium -5]- hydroxide solution (250 ml of a 5% solution) to remove the unreacted truxillic acid. The organic layer was dried and evaporated to give 8.6 g of a yellow oily solid. The solid was taken in ether-—hexane (90:10) from which 4.8 g (57%) of a solid depositedQQ} mp l73-174°; ir (Nujol) 1780 cm"; nmr (cpc13) 2.8 (s, 50), 2.98 (m, 5H), 4.81 (dd, J] = 1.5 cps, J2 = 7 cps, 1H), 5.76 (m, 1H), 5.92 (bs, 1H), 6.38 (dd, J1 = 7 cps, J = 3 cps, 1H); m/e 250. 2 Anal. Calcd. for C 0 C, 81.59; H, 5.65; 17”14 2‘ Found: C, 81.50; H, 5.96; Addition of hexane to the filtrate caused deposition of a solid 51 (2.1 g, 24.8%) which was recrystallized from hexane-ether (50:50) to give needles: mp 132-133° (lit.55’S7 1330; ir (Nujol) 1770 cm']; nmr (CDC13) r 2.7 (bs, 10H), 5.26 (bs, 1H), 7.4 (m, 3H). Anal. Calcd. For C17H1402: C, 81.59: H, 5.65; Found: C, 81.75: H, 5.68. Truxillic acid was treated with lead tetraacetate in acidic and neutral media. The following table summarizes the conditions and resfilts of such experiments. -52- TABLE II I tme t f T 1111 Acid megcmft/e RUN Truxillic Lead Tetra- Recovered Total Patio Time Solvent # acid, 9 acetate, 0 acid, 0 Yield 50/51 Hours 1 8.88 71 3.7 32% 1.15/2.45 5 ACOH 300 ml 2 10 50 5.5 16% 2.5/1 5 Benzene 600 ml 3 10 60 7.5 0% - 5 ACOH H20 4 5 / 32a 9 459 4/1 5 Benzene recrystallized 400 ml 5 10 46 3.7 20% 1/2 10 130 ml Acor 66011an0 L \J aLead tetraacetate recrystallized from acetic acid and dried in the dark. (22922898222292L928222282292822228922928222 9992999 50 (500 mg) was taken in benzene (50 m1) and acetic acid (1 ml) was added and the solution was refluxed overnight. After 12 hrs the mixture was washed with water and sodium bicarbonate solution and dried. Evaporation of benzene yielded a solid (400 me) which showed an nmr spectrum identical with that of starting material. agPhenylglutaric acid (7 g, 0.033 mole) was dissolved in benzene (150 m1) and lead tetraacetate (23.3 g, 0.05 mole) was added to the solution. -53- Pyridine (6 ml, 0.075 mole) was added and the mixture was heated in an oil bath to 75-80°. A yellow clear solution was formed which after a few minutes started to become cloudy and yielded a precipitate. Workup yielded 5.4 g ( 35%) of a yellow oil which was recrystallized (from hexane to give 5.1 g of crystals, mp 34-36°, mich had identical spectrosc0pic properties with y-phenyl—y-butyrolactone prepared before" . AKCILLARY SYNTHESIS P DISCUSSIRN 0F WESULTS -55- WWRWWWM Lead tetraacetate has been used for oxidative decarboxylation of aliphatic carboxylic acids, and 1,2-diacid558’59, to obtain the corresponding acetate or olefins. The earlier work suggested a direct oxidation to a carbonium ion intermediate, based on the isolated productseo. Later work by Kochi51552953, and Davis and Naring5“’55 showed that the primary product is a radical, although in certain cases, such as bicyclo[2.2.1]- hexane derivatives, the primary product is oxidized further to the corresponding carbonium ion and no product from a free radical intermediate is isolated55’57. In our studies of various methods for decarboxylation of a-truxillic acid 14, we treated it with lead tetraacetate in the hope of converting it to a diacetate 52 or the cyclobutenyl acetate 53. But the reaction instead yield lactones 50 and 51. P L.T.A. ; 0 + h 0011 0 P 1' 50 ‘ 51 ’VD 'VV 1? \/ CH $H3 c=e C02 Ph 2 ph Ph Ph 0 CH 52 3 22 -5(- Hhen this reaction was carried out on glutaric acid and a-phenyl glutaric acid, the corre5ponding lactones were obtained in good yields. He therefore suggest that this could be a general reaction for conversion of 1,3-diacids to the corresponding y-lactones. 299922222822222222229222822222228292222992222- The reaction was carried out in benzene solution in the presence of pyridine catalyst. The two products 51 and 52 were isolated by fractional crystallization. The product having mp 173-174°, gave an analysis consistent with C]7H1402. The parent peak in the mass spectrum appeared at 250. It showed a very high frequency carbonyl absorption at 1816 cm'] and the nmr spectrum exhibited a five proton singlet at r 2.8, a five proton multiplet at r 2.98, a one proton doublet of doublets at r 4.81, J1 = 1.5 cps, J2 = 7 cps, a one proton multiplet at r 5.76, a one proton broad singlet at T 5.92, and a one proton doublet of doublets at r 6.38, J1 = 7 cps, J2 = 3 cps. The ir spectrum is consistent with a B lactone or a very strained y lactone. All possible 8 lactones are unacceptable based on the chemical shifts and coupling constants of the saturated protons and the expected reactions of cyclobutane-l,3-dicarboxylic acid Lg. The data are well explained, however, by structure 51, which has an additionally strained y-lactone in a bicyclo[2.l.1]system, The trans_ nature of the phenyls is consistent with the origin of the compound and required by the four different types of aliphatic protons, and two different types of phenyl protons in the nmr. The nmr peaks were assigned to various hydrogens as follows: -57- JAB = l cps JAD = 7 cps £2 JBD = 3 cps JAC and JBC exist but are small. The lowest field aliphatic signal at r 4.81, was assigned to HA’ the bridgehead hydrogen alpha_to the oxygen. The second low field signal at r 5.76 was assigned to “8’ the bridgehead proton alpha to the carbonyl. Assignment of the other two signals to H8 and HC was more difficult, because the magnetic anisotropic effect of the lactone bridge and the phenyls could not be determined with confidence. However, a decision could be made by consideration of coupling constants with reference to a molecular model. Thus, it was clear that HA should be coupled more strongly to the proton anti_to the lactone than to the §y9_ proton. The anti proton must therefore be HD (JAD = 7 cos). The smaller coupling of HB to HD (3.cps) must reflect a distortion in the system which makes it unsymmetrical. This distortion might arise from the smaller C-O-C bond angle in the lactone or from steric repulsion of the syg_phenyl, or both. -58- The other product of the reaction is reported in the literature. It was obtained by electrolytic oxidation of truxillic acid57, and also by nitrosyl bromide deamination of y-truxinic amino acid 5455. Ph COOH Ph H 0 NH NOBr " Ph Ph /F—' I 22 H 22 H The ir spectrum (-C=0 at 1770 cm']) is in accordance with the 5-membered ring lactone. Our work provides no further stereochemical evidence for the suggested syp structure. The one proton singlet at r 5.26 was assigned to the proton a to the oxygen and the three proton multiplet at 7.4 was assigned to the cyc10pr0pane hydrogens. When the reaction of a-truxillic acid was carried out in different solvent systems, it appeared that the ratio of 52/51 was related to the acidity or'basicity of the solvent system. when the lead tetraacetate was freed from acetic acid, the ratio of 51/50 was found to be 4/1 while in acetic acid the ratio was 1/2.5. In order to find out whether acetic acid is responsible for rearrangement of 50 to 51, lactone 50 was refluxed in benzene in the presence of acetic acid. The starting material was recovered after 24 hrs almost quantitatively and lactone 51 could not be detected in the product. Formation of lactone 50 from a-truxillic acid gives a novel and unique method for preparation of bicyclic[2.l.l]lactones. To our knowledge this is the first example of such a class of lactone to have been prepared. l\) 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. REFEREHCES E. F. Kiefer and J. D. Roberts, J. Am. Chem. Soc., 84, 784 (1962). J. D. Roberts, S. L. Manat, H. Vogel and D. Knutson, Ibid., 8., S 2645 (1964). T. J. Katz and E. H. Gold, Ibid,, 86, 1600 (1964). D. G. Farnum, Ibid,, 82, 5379 (1967). w. Borsche and w. Henz, 8g:,, 41, 194 (1908). A. L. Wilds and Th. L. Johnson, J. Am. Chem. Soc., 61, 286 (1945). P. Yates and L. L. Williams, Ibid., 82, 5896 (1958). N. L. Drake and J. R. Adams, Ibid., 61, 1326 (1939). ’\/\a Kaarina Leppanen, Ann. Acad. Sci. Fennicac, Ser. AII, No. 131, 50 pp ’VVVVVVM (1965); Chem. Abstracts, 64, l4112h (1965). 0. Walker, J. Am. Chem. Soc., 71, 3666 (1955). G. Forrest woods and I. 8. Tucker, Ibid., 72, 2176 (1948). F. C. NoVello, H. E. Christy and J. N. Sprague, Ibid., 72, 1330 (1953). S. H. Abdollah, J. Ind. Chem. Soc., lg, 62 (1935). S. Hasamune, Tetrahedron Letters, 945 (19(5). G. H. Badger, P. Cheuychite and N. H. Sasse, Aust. J. Chem., 17, 353 (1964). R. Stoermer and G. Foester, Chem. Ber., 52, 1263 (1919). J. Meinsenheimer and F. Heim, Annal. Chem., §§§9 260 (1937). N. Sugiyama and K. Yamada, Chem. Abstracts, 65, 151889 (1966). R. Stoermer, J. Neumeaker and R. Schmidt, Chem. Abstract, 1341 (1926). 1:. Stobbe and A. Lehfeldt, 835., 8232’ 2415 (1925). -59- 21. 22. 23. 24. 25. 26. 27. 28. 30. 31. 32. 33. 34. 36. 37. 38. 39. -50- E. H. White and H. C. Dunathan, J. Am. Chem. Soc.,,ag, 6055 (1956). H. Bernstein and H. C. Quimby, Ibid., 55, 1815 (1943). H. D. Cohen and G. H. Schmidt, J. Chem. Soc., 1996 (1964). K. B. Hiberg, D. R. Lowry and Th. H. Colby, J. Am. Chem. Soc., 83, 2998 (1961). P. E. Eaton and Th. H. Cole, Jr., 1310,, 86, 962, 3157 (1964). E. H. Cain, R. Vukov and S. fiasamune, Chem. Comm., 98 (1969). L. J. Bellamy, The Infra-RedASpectra of Complex Molecules, John Wiley and Sons, New York, 1964, p R. D. Campbell and R. F. Dfstead, Proc. Iowa Acad. Sci., 7;, 197 (1964). A. Nickon and A. Hill, J. Am. Chem. Soc., 56, 1152 (1964). E. Grovenstein, Jr. and L. C. Rogers, Ibid., 86, 855 (1964). E. H. White and H. C. Dunathan, J. Am. Chem. Soc., 86, 453 (1964). H. Krauch and H. Kunz, Organic Hame Reactions, John Hiley and Sons, N. Y. (1964), p. 153. H. . Emerson, 0rganic Reactions, Vol. IV, John Wiley and Sons, S H. Y. (1948), p. 174. J . Bellamy, Advances in Infrared Group Frequencies, Nethuen and Co., Ltd., London, (1968), p. 4. H. Hinn and H. E. Zaug, J. Org. Chem., 33, 3779 (1968). K. B. Hiberg, J. Am. Chem. Soc., 17, 2519 (1955). H. J. Cook, E. J. Forbes and G. M. Khan, Chem. Comm., 121 (1966). G. Stork, H. S. Horral and J. J. Papas, J. Am. Chem. Soc., 82, 4315 (1960). w. Roberts and H. C. Whiting, J. Chem. Soc., 1290 (1965). 40. 41. 42. 43. 44. 45. 46. 47. -61- K. Rorig, J. D. Johnson, R. H. Hamilton and T. J. Telinski, Organic Synthesis, Coll. Vol. IV, John Hiley and Sons, R. Y. (1963) p. 576. F. Sondheimer, G. Rosenkanz, C. Hancera and C. Djerassi, J. Am. Chem. Soc., 75, 2601 (1953). R. Breslow, J. Lockhart and A. Small, Ibid., 2793 (1962). D. G. Farnum and C. F. Hilcox, J. Am. Chem. Soc., 89, 5379 (1967). R. H. Taft, L. D. HcKeever, J. Am. Chem. Soc., 87, 2489 (1965). D. G. Farnum and D. S. Patton, unpublished results. Paul B. Bartlet and R. Hiat, Ibid., 89, 1398 (1958). L. Silbert and D. Swern, A381. Chem., 30, 385 (1958). Org nic Synthesis, Coll. Vol. I, John Hiley and Sons, N. Y. _ (1941) D. 413. H. Julia, S. Julia and B. Bemont, Bull. Chem. Soc. France, 394 (1960). E. J. Corey and H. Chaykovsky, J. Am. Chem. Soc., 87, 1357 (1965). Organic Synthesis, 6011. Vol. IV, John Wiley and Sons, H. Y. (1963) p. 777. L. F. Fieser, Experiments in Organic Chemistry, 3rd Edition, 0. C. Heath and 60., Boston, p. 182. Organic Synthesis, Coll. Vol. III, John Hiley and Sons, N. Y. (1965) p. 715. A. Patel, Acta. Chem. Scand., 22, 1424 (1966). D. D. Roberts, J. Org. Chem., 2712 (1968). Organic Synthesis, Vol. XLI, John Wiley and Sons, N. Y. (1961) p. 16. H. I. Bernstein and I. S. Goldstein, J. Am. Chem. Soc., 66, 760 (1944). G. H. Griffin and A. F. Vellturo, J. Org. Chem,3/1, 2241 (1966). 66. 67. H. C- O f'... 0. G. -62- .. Hosher and C. L. Kehr, J. Am. Chem. Soc., 75, 317? (1953). . Crab and A. Heiss, Helv. Chim. Acta, 43, 1340 (1060). . Kochi, J. Am. Chem. Soc., 87, 1811 (1965). . Kochi, Ibid., 219 2500 (1965). . Kochi, Ibid., 87, 3609 (1965). . Davis and C. Haring, Chem. Comm., 263 (1965). . Davis and C. Haring, J. Chem. Soc., 1639 (1967). . Corey and J. Casanova, J. Am. Chem. Soc., 85, 165 (1963). . Gream and D. Hege, Tetrahedron Letters., 593 (1963). 63 F mapuHu c n v ..OF 0 w h _ 14 - 1 ‘ 1 1 - d ‘ ‘ 1‘1 — I I < d1 I- 1 1 4 ‘ -o 8. 8a b Ll) H r P r h b P 5 Li P b r 1. h D r h D _1 D D 5 ’ - b D (D b — b D D F - P D b D _ D h b b P hi? 5 b H ’ r b h 1- L (P I ’ ”L + 0.. . o.« 0.0 ‘- 1- oi 41114444“_dd141‘d1“¢41_l14J4d421411—d1+111‘1‘l4‘fi‘ _ 64 a .N4 N was: s c m e m N dam—«Human..q—qfiqw.fiqfium_.udfi—wq42qmdq8—jiqq-q.q—qa‘dA-fi-u-qqd_-ll- a 65 m muDLHu _ n - o n v n m J24-#-_-48 -q-_4q-_-__afi_-d-_aH-8:--_j:--___:__-_-__a-_d- ‘1'— ‘.y s ’3"?' .g—‘I—I- 21‘ ’1‘ .9 \ p on I __-F_:_:e-:-2L_2_-_-__-___-__-r_-LL_-r-_-FHF_-8_-__-FL__-.Vpr___ a 8m 8.. on...» 88 com 8%; 028- . H _ n .: emu 8m 8m 88 om 028 66 8. v mgpuHL .:.ze_ — 8... 8V 9 is 67 m ungmu 68 ‘ 69 n maawfiu —_ “1111111111111112111221111111111111“