The Structural Elucidation of a Proposed Intermediate in the Stereoselective Synthesis of dl-Desoxypodocarpic Acid Citation Grand, Paul Sheldon (1969) The Structural Elucidation of a Proposed Intermediate in the Stereoselective Synthesis of dl-Desoxypodocarpic Acid. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/YBB4-6J75. https://resolver.caltech.edu/CaltechTHESIS:02222016-105427770 Abstract The hydroxyketone C-3, an intermediate in the stereo-selective total synthesis of dl-Desoxypodocarpic acid (ii), has been shown by both degradative and synthetic pathways to rearrange in the presence of base to diosphenol E-1 (5-isoabietic acid series). The exact spatial arrangements of the systems represented by formulas C-3 and E-1 have been investigated (as the p-bromobenzoates) by single-crystal X-ray diffraction analyses. The hydroxyketone F-1, the proposed intermediate in the rearrangement, has been synthesized. Its conversion to diosphenol E-1 has been studied, and a single-crystal analysis of the p-bromobenzoate derivative has been performed. The initially desired diosphenol C-6 has been prepared, and has been shown to be stable to the potassium t-butoxide rearrangement conditions. Oxidative cleavage of diosphenol E-1 and subsequent cyclization with the aid of polyphosphoric acid has been shown to lead to keto acid I-2 (benzobicyclo [3.3.1] nonane series) rather than keto acid H-2 (5-isoabietic acid series). Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Chemistry) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Ireland, Robert E. Thesis Committee: Unknown, Unknown Defense Date: 16 September 1968 Funders: Funding Agency Grant Number NIH UNSPECIFIED NSF UNSPECIFIED Caltech UNSPECIFIED Record Number: CaltechTHESIS:02222016-105427770 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:02222016-105427770 DOI: 10.7907/YBB4-6J75 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 9573 Collection: CaltechTHESIS Deposited By: INVALID USER Deposited On: 22 Feb 2016 21:16 Last Modified: 29 Apr 2024 21:05 Thesis Files Preview PDF - Final Version See Usage Policy. 5MB Repository Staff Only: item control page

THE STRUCTURAL ELUCIDATION OF A PROPOSED INTERMEDIATE IN THE STEREOSELECTNE SYNTHESIS OF dl-DESOXYPODOCARPIC ACID Thesis by Paul Sheldon Grand In Partial Fulfillment of the Requirements For the Degree of· Doctor of Philosophy California Institute of Technology Pasadena, California 1969 (Submitted September 16, 1968) To the three beautiful women in my life- my wife, Deborah; my daughter, Lisa Brooke; and my dog, Pogo. ii ACKNOWLEDGMENTS I wish to express my appreciation to Professor R. E. Ireland for suggesting this problem and for his guidance during the period of this research. Thanks also go to my colleagues, especially Dr. D.A. Evans, Mr. S. W. Baldwin, and Dr. G. M. Rubottom for their many illumin.ating and interesting discussions. I am indebted to Dr. R. Marsh, Dr. R. Stanford, and especially Dr. J. Bordner and Professor R. E. Dickerson for their aid in the X-ray crystallographic study performed as part of this work. I would like to thank Miss Nicki Wilson and Mrs. Edi Bierce for typing this manuscript. For financial support during this research, I would like to thank the National Institutes of Health, the National Science Foundation, and the California Institute of Technology. iii ABSTRACT The hydroxyketone ,...,,..,,.. C-3, an intermediate in the stereoselective total synthesis of dl-Desoxypodocarpic acid (ii), ,....,... has been shown by both degradative and synthetic pathways to rearrange in -- the presence of base to diosphenol E-1 (5-isoabietic acid series). The exact spatial arrangements of the systems represented by formulas ~ and!.::! have been ·investigated (as the .e_-bromobenzoates) by single-crystal X-ray diffraction analyses. The hydroxyketone F-1, the proposed intermediate in the rearrangement, has been """"""' synthesized. Its conversion to diosphenol ......,._ E-1 has been studied, and a single-crystal analysis of the p-bromobenzoate derivative has been -- - performed. The initially desired diosphenol C-6 has been prepared, and has been shown to be, stable to the potassium t-butoxide rearrangement conditions. Oxidative cleavage of diosphenol .!:! and subsequent cyclization with the aid of polyphosphoric acid has been shown to lead to keto acid I-2 (benzobicyclo [ 3. 3 .1] nonane series) rather than keto """" acid H-2 (5-isoabietic acid series). """"' . iv TABLE OF CONTENTS Title Page . ........................................... . i Acknow ledgn1ents ...................................... . ii Abstract . ................................ ·............. . iii Table of Contents . ..................................... . iv Historical Introduction . .... ~ ........................... . 1 Discussion . ........................................... . Experimental . ......................................... . 10 44 7[3, 9[3, Dimethyl-2-hydroxy-7 a-phenylhexahydro-8[3H1, 2-indene-3-one (E-1) a) Potassium t-omoxide rearrangement from hydroxyketone C-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . b) Potassium t-buWxide rearrangement from hydroxyketone F-1. . . . . . . . . . . . . . . . . . . . . . . . • . . . . . c) Sodium hydroxlcle'rearrangement from hydroxyketone C-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . d) Sodium hydroxlcle'rearrangement from hydroxyketone ,,...,...,... F-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 7[3, 9{3-Dimethyl-2-methoxy-7a-phenylhexahydro-8j3H1, 2-indene-3-one (D-2) ...................•.. ;...... 47 "'"'"""" 44 45 46 7[3~ 9[3-Dimethyl-~-h¥droxy-7a-phenylhexahydro-8[3H- 1ndane-2-one (D-4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 ~ 7[3, 9[3-Dimethyl-~acetoxy-7a-phenylhexahydro-8[3Hindane-2-one (D-5) .................................. 50 ~ 7{3, 9{3-Dimethyl-7 a-phenylhexahydro-8{3H-indan-2-one ,. (D-6) ,,.,.,....,... ............................ ~ ................. 51 7[3, 9[3-Dimethyl-7a-phenylhexahydro-8a.H-indan-2- one (XVI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . """""" 52 7j3, 9a-Dimethyl-7a-phenylhexahydro-8{3H-indan-2. one (IV) .•.•...•..•..•••· •...••..•.••••.••••••••••.. · 53 """"" v I I 6-Methyl-2-(3-methallyloxy-l)methylene-6-phenylc ye lohexanone (,,,....,...,,. G-2) .............................. . 54 1 6-Methyl-2-(3 -methallyl)-6-phenylcyclohexanone ( ,,..,,,....,... G-4) ........................................... . 55 2{3, 6{3-Dimethyl-2a-(3'-methallyl)-6a-phenylcyclohexanone ( G-5) .................................. . 56 • 2a, 6{3-Dimethyl-2j3-(3-methallyl)-6a-phenylcyclohexanone (XIII) . ~ ................................ . 57 2{3, 6[3-Dimethyl-2a-acetonyl-6a-phenylcyclohexanone ( G-6) . ..................................... . 58 7[3, 9{3-Dimethyl-7 a-phenylhexahydro-1, 8-indene-2one ( G-7) . ...................................... . 59 7{3, 9p-Dimethyl-7 a-phenylhexahydro-8{3H-indane-2one ( F-2) ........................................ . 60 7[3, 9a-Dimethyl-2-hydroxy-7 a-phenylhexahydro8aH-1, 2-indene-3-one (C-6) .......•...........•.. 62 7{3, 9a -Dimethyl-2 -methoxy-7 a -phenylhexahydro8aH-1, 2-indene-3-one (XVII) ................••.•. 65 7[3, 9{3-Dimethyl-7a-phenylhexahydro-1, 8-indene2, 3 -dione (~ ............................... . 66 7p, 9[3-Dimethyl-3a-hydroxy-7 a-phenylhexahydro1, 8-indene-2-one (F-1) .,...,,.,.... .......................•.. 66 7[3, 9a-Dimethyl-3.a-(E_-bromobenzoyl)-7 a-phenylhexahydro-1, 8-mdene-2-one (XIX) ............•••. 67 7{3, 9j3-Dimethyl-2-(Q-b.romobenzoyl)-7a-phenylhexahydro-8{3H-1 ;2 -mdene-3 -one (XXIJ ......•.••.. 68 7f3, 9{3-Dimethyl-39!-(J2.-bromobe12~-~yl)-7a-phenyl­ hexahydro-1, 8-indene-2-one (~ ...........•.•..• 70 2a-Carboxymethy.l-1J3,. 3j3-dimet.hyl-3a-phenylcyclohexanecarboxyllc acid anhydride (H-1) ....•...•...• 71 syn-9-(1, 5-Dimethyl-4-oxo-2, 3-benzobicyclo[ 3. 3 .1]nonane)acetic acid (I-2) ......•....•....• ~ •.•.••... 72 """""" ~ """"" """"' """"' """"' """""' """"' """"' """"' """" vi syn-9-(1, 5-Dimethyl-2, 3-benzobicyclo( 3. 3 .1]nonane)acetic acid (I-3)..... . . . . . . . . . . . . . . . . . . . . . 73 syn-9-(1,. 5-Dimethyl-2, 3-benzobicyclo( 3. 3 .1]nonane)-2' -ethyl alcohol (J-1).................... 74 syn-9-(1 5-Dimethyl-2 3-benzobicyclo( 3. 3 .1]nonane)acetaldehyde (J-2)........ . . . . . . . . . . . . . . • . 75 · syn-9-Ethyl-l, 5-dimethyl-2, 3-benzobicyclo[ 3. 3 .1]nonane (J-3).................................... 76 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Appendix.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Proposition I. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Proposition II... . . . . • . . . . . . . . . . . . . . . . . . . . . . . . . . . . . • . . 104 ill...................................... 118 Proposition IV. . . . . . . . . . . . . . . . . . . . • . . . . . • . . . . . . . . . • • 13 6 Proposition V............... . . . . . . . . . . . . . . . . . . . . . . . . 150 ~ ~ """""' """"" Proposition -1HISTORICAL INTRODUCTION Podocarpic acid (j) was first isolated from Podocarpus cup res sina var. imbricata by Oudemans la, 1b. Campbell and Todd2 , in 1941, elegantly and conclusively proved the structure of Podocarpic acid as ,,... i. OH 12 2 3 4 6 Synthetic efforts, directed toward the attainment of both podocarpic acid ,,...i and desoxypodocarpic acid ,,...,,... ii and reported prior to 4 1967, are reviewed in a variety of publications 3 ' ' 5 ' 5 , 7 . In general, the syntheses involved two methods of attack: (a) ring B formation by attachment through the potential 9, 10 carbons of the intact A and C rings, or (b) manipulation of the preformed tricyclic ring system to introduce the correct stereochemistry at carbon atoms 4 and/ or 5. These synthetic approaches lacked stereochemical control and high yields. An interesting synthetic approach to the podocarpic acid skeletal system recently reported by Carman, Deeth, Marty, Mori, 8 and Matsui , is shown in Chart A. The scheme involves elaboration of ring Con the already preformed AB bicyclic system. Their ini- -2tial reaction, which involved ozonolysis of dimethyl agathate (A-1), """""" Chart A Cifs 1 ..... CH., 2 ..... 3 ..... R 4 5 a, R = CH(CH.,) 2 ""b, R = H §.a, R = CH(Clfs)2 b, R = H · -3- yielded a mixture of diketone esters consisting principally of the compound A-2. Hydroxyketone ,,...,...,,... A-3 crystallized out of a solution of ,,...,...,... A-2 in methanolic sodium methoxide. ,,...,...,,... Distillation of the alcohol A-3 ~ at 0.1 mm gave principally the unsaturated ketone A-4, together with . ""'"'""' some of the 8(9) isomer. The crude ketoesters ,...,...,.... A-4, treated with . isopropyl magnesium bromide, yielded the mixture of esters A-5a. ~ Dehydrogenation over palladium on charcoal at 230-240° afforded methyl 4-epidehydroabietate (A-6a). Methyl desoxypodocarpate (A-6b) """""" .......,.,,,_ was obtained as a by-product from the dehydrogenation. This approach would also seem to be applicable to the direct synthesis of methyl desoxypodocarpate (~). The recently reported synthesis of podocarpic acid (!) by Spencer 9, which also incorporates the AB--7ABC approach, is described in Chart B. The approach had as its key step introduction of the axial carboxylate function by stereoelectronically controlled carbonation of enolate anion ,,...,...,,... B-2, generated from a, J3-unsaturated ketone ,,...,..,.... B-1. The axial carbomethoxylation product ,,...,...,,... B-3 was not detected in the product mixture. The major product (ca. 30% yield) was the hydroxyketone resulting from the reduction of the double bond B-1. The saturated diol was isolated on occasion in of structure ,,...,...,,... yields up to 7%. The on:y monocarbomethoxylation product ever obtained was the epimeric 4-a ester. One of the key steps in the synthesis of methyl deisopropyldehydroabietate reported by Spencer 10 involved stereoselective /3 methylation of ketone B-4 to afford .methyl"""" -4Chart B 0-D 0 + 3 ""' 0 0 H HOH 3 6 7 ""' ""' 9 ""' 10 OH 0 11 """"" -5ated ketone ~ after methanolysis. However, there was also formed approximately 10% of the a methylation epimer ,..,..,,,... B-3. The yield of oily ,,...,,....,.. B-3 of reasonable purity isolated was roughly 5% from """"""" B-4. Reduction of ketone B-3 to desketo B-6 was effected by Raney Ni reduc- · """"""" """"""" tion of the thioketal derivative. Oxidation of alcohol ,,...,,,..,.. B-6 afforded the desired ketone ~. To facilitate annelation of ketone ~, it was converted to its hydroxymethylene derivative B-8. Reaction with 1"""" -- diethylaminobutanone-3-methiodide afforded Michael adduct B-9, which was then cyclized to tricyclic ketone B-10. Conversion of ke"""'""" tone B-10 to methyl dl-podocarpate (B-11) was accomplished by treat""""" ~ ment with N-bromosuccinimide. The lack of stereochemical control and low yields reported in the synthetic pathways cited and discussed resulted in an investigation of approaches directed toward an efficient stereoselective synthesis of dl-Desoxypodocarpic acid <.!&in these laboratories. This investigation p_roved successful. The synthesis 7 , as shown in Chart C, involved the selenium dioxide oxidation of ketone C-7 to diketone C-4. Diketone ,,.....,_ C-4 was reduced with sodium borohydride to hydroxyketone C-3 . Catalytic reduction of the unsaturated hydroxyketone """"" C-3 ,,.....,_ yielded the corresponding cis-fused hydroxyketone C-2. Hydroxy~- """"' ketone ,,...,,....,... C-2 was converted to diacid ,..,..,,,... C-1 by chromic acid--glacial acetic acid oxidative cleavage. Dehydrative cyclization of diacid .......,..,.. C-1 to the tricyclic keto acid ,,....,_ C-5 was effected with the aid of polyphosphoric acid. Hydrogenolysis of keto acid ,,....,_ C-5 yielded Desoxypodo- -6- carpic acid (ill .. Chart C Clfs 1 3 ,,.... ... .. 1 ,,.... 4 1 . C!Jso ,, ~ I IfsC ~c OH 0 5 ,,.... 0 6 0 7 An alternative route through the diosphenol ,...,...,.... C-6 and subsequent cleavage to diacid ,,...,...,,... C-1 would serve as a more efficient scheme for the synthesis of dl"".Desoxypodocarpic acid (ii). A mechanistic path""""' way which would result in the desired transformation is indicated on the following page. Examination of the mechanism of the transformation indicates that reversible basic abstraction of the C-3 methinyl hydrogen of hydroxyketone ,...,...,.... C-3 would produce the enolate species ..... I. -7- This anion could either revert to the hydroxyketone ........,._ C-3 by the addition of a proton or be further transformed to the enolic diketone carbanion II. Even if the equilibrium resided far towards the enolate """"' species L protonation of the tertiary carbanion would displace the equilibrium and result in overall reduction of the C-8 center and oxidation of the C-3 center. C-6 ........,._ C-3 """' ~c ~c 0 I II 11 A recent rep;rt by Yoshida and Ku~ta , noting the successful execution of a similar·rearrangement in the androstane series, lends support to the proposed rearrangement. Their transformation, as indicated below, was effected with the aid of an excess of sodium hydroxide. 0 H 73% yield when R = H 70% yield when R = COC2 Ifs -8- The rearrangement hypothesis was tested by treatment of the hydroxy indenone C-3 with two equivalents of potassium -t-butoxide ~ in.!_-butyl alcohol. Upon mixture of these components, the solution developed a blood red color which remained until the reaction mixture was quenched with hydrochloric acid. A product, different from the starting hydroxyketone, was obtained in a yield of 49% when the reaction components were stirred in a nitrogen atmosphere at room temperature for one hour. The assignment of the structure of the product as diosphenol C-6 is aided by the interpretation of the spectral data. ......,,...,,,. Its infrared spectrum exhibited absorption bands at 3500 and 3350 (shoulder) cm- 1 (enolic-OH), and 1700 and 1650 cm- 1 (enolic dione). A maximum was exhibited at 262 mµ ( E 9, 038) in the ultraviolet. The nuclear magnetic resonance spectrum exhibited signals at o1.32 (s, 3, C;..9 C!Jg), 1.48 ...... (s, 3, C-7 C!!s), 2. 85(d,1, J = 3 hz, C-8 _g), 5.65 12 (d, 1, J = 3 Hz, C-1 vinyl ...... H), and 6.22 (broads, 1, -OH). ...... The rearrangement product melted at 162-164°. The compound exhibited a purple color 13 upon the addition of ferric chloride test solution. The course of the base-catalyzed rearrangement is dependent upon the presence or absence of oxygen. The diosphenolic product is isolated in the absence of oxygen. The presence of a trace amount of oxygen results in the formation of a number of oxidation products. The mixture consisted mainly of diketone ............... C-4 and the anhydride formed from 1, 8-dehydrodiacid .......,._ C-1 . -9Oxidation of the rearrangement product did not produce a diacid identical with the one derived from oxidation of hydroxyketone £.:1 (i.e., diacid £::1). Oxidation of the rearrangement product with hydrogen peroxide--aqueous sodium hydroxide and subsequent reaction with acetic anhydride afforded an anhydride 14 . Comparison of both melting point and infrared spectral data indicates that the tricyclic keto acid, obtained by conversion of the anhydride using polyphosphoric acid, was not identical with 7 -ketodesoxypodocarpic acid ( C-5). """""' This indicates that the diosphenolic product does not have the structure indicated by formula C-6. A similar comparison of compounds in the """""" 5-isopodocarpic acid series with the corresponding compounds in the rearrangement product series indicates that the rearrangement product cannot be the corresponding trans-fused diosphenol (i.e., compound C-6 with the C-8 hydrogen f3). -10DISCUSSION The current research program has been concerned with the determination of the structure of the rearrangement product and a study of the mechanism of the rearrangement. A reinvestigation of the potassium t-butoxide initiated rearrangement indicated that the conversion to the diosphenol was quantitative when the reaction components were stirred at 45 ° for one hour. Quantitative conversion was also effected upon heating the hydroxyketone C-3 in aqueous methanolic sodium hydroxide for thirty minutes """"' under carefully de oxygenated conditions. Confirmation that the diosphenol was neither compound ,,...,,,._ C-6 nor its epimer with the C-8 hydrogen /3 was selected as the initial goal. This confirmation would be accomplished by transformation of the rearrangement product, now assigned structure D-1 only for the """"""' ease of visual presentation, to hydrindanone D-6 by a series of reac"""' tions that would neither affect the stereochemistry of center 8 nor that of 9 (Chart D). A lack of identity between both ketone ,,_ m and .......... IV and ketone D-6 would prove that the rearrangement product was neither _~c···· ~c ~c 0 0 IV """' H CHzC02 H CHzC02 H v,,... -11diosphenol C-6 nor its C-8 epimer. The solid cis-fused ketone m - """""" ............ and the liquid trans-fused ketone !Y. ar.e formed during lithium-ammonia reduction of unsaturated ketone £;:J (in a 2:3 ratio) 7, and consequently are available for comparative study. Chart D le OH ~... ~ I ----t ~c OH 4 0 "" OAC 5 0 0 6 "" "" The rearrangement product D-1 was converted to the acetoxy~ ketone~ in 79% overall yield. The diosphenol !2,::1 was first quantitatively converted to the corresponding keto enol ether D-2 with 15 the aid of dimethyl sulfate--potassium carbonate . The keto enol ~ ether was reduced by lithium aluminum hydride to the hydroxy enol ether D-3 which was hydrolyzed to the hydroxyketone D-4. A com~ ~ -12parison of the melting point and spectral data of the hydroxyketone D-4 of the rearrangement series with the catalytic reduction product ,,..,,..,... C-2 of the podocarpic acid series as presented in Table I below indi- """" cates that these two compounds are not the same. mp (°C) ir (cm -l) v CHCls max nmr (ppm) 6 CDCls HO C- 2 0 '""'"' ,.... 173-174. OH D-4 O -""""' 122-124.5 3600 (m) 3450 ~w) 1735 s) 3550 3450 w) 1748 s) 0.67 ~s~ 1.14 s 1.43 (s~ and 1.45 (s t) TMS C-9 methyl C-7 methyl The hydroxyketone D-4 was esterified with acetic anhydride ~ and pyridine in 83% yield, and the resultant acetoxyketone D-5 was 16 reduced with lithium--ammonia . The resultant crude reaction ~ mixture of ketone D-6 and alcohol reduction product was oxidized with chromic acid 17 to the hydrindanone D-6. ~ ·~ A comparison of both the spectral and melting-point data of the degradation ketone D-6 of the rearrangement series with the ~ trans-fused ketone .......,... IV and the -cis -fused ketone .......,... ill is presented in -13Table II below. The stereochemistry of ketones III and IV was ascer"""" """"' tained through independent synthesis of the trans-locked isomer from the diacid V of known 18 configuration in 98% yield. The trans-ketone """ IV, obtained by pyrolysis of the lead sau19 of this diacid, proved to """"" Table II HgC HgC 0 0 IV III """"" ketone mp ( ° C) oil ketone mmp { ° C) """"""" 100.5-101 100.5-101 III + D-6 meltedOetween 70 and 102° 221-223 semicarb. mmp (°C) IV + D-6 meltecfOe'tween 208° and 243° -1 v CHC1s max D-6 """"" semicarb. mp ( ° C) ir (cm ) 0 241-243 220. 5-222 1735 1738 1738 0.33 ~s~ 1.28 s 0. 53 ~s~ 1. 06 s 1. 33 ~s~ 1.40 s nmr(ppm) o CC14 · TMS C-9 methyl C-7 methyl be identical to the liquid ketone obtained on lithium- -ammonia re due- -14- tion of the enone C-7. This was accomplished through direct spectral """"" comparisons as well as mixture melting point determination of their semicarbazones 7 . This conclusion meant that the crystalline saturated ketone from the lithium--ammonia reduction was the cis-locked ketone III. This cis-ketone was selectively obtained by catalytic hydrogenation 7 of the enone C-7. The C-9 methyl protons of both cis-fused """' """"" ketone III and trans-fused ketone IV are highly shielded. This is easily """"' """' recognized in the conformational study shown below. The ketone ,D-6 ,_ is neither the cis-fused ketone ,,...,.... ID nor the trans-fused ketone .-... IV. The 0 ~c 0 III ~CH.,01 ,, ::-..... '' .................. ~c 0 IV ,,...,.... H -15rearrangement product D-1 is therefore neither the cis-fused - """"' diosphenol 0 nor the C-8 epimeric trans-fused diosphenol, since the transformation of the rearrangement product D-1 to the ketone """"' D-6 involved isomerization of neither the C-8 nor C-9 assymetric ~ centers. Our attention was drawn to the possibility that the rearrangement product D-1 was a stereoisomeric diosphenol. Evidence for this ~ belief included the spectral properties of the rearrangement product D-1, its response to the ferric chloride test, and the analogy of the """" rearrang~ment in the androstane series already cited11 • Interest Chart E ~a c() ~ ) ~ < IfsC B IfsC C-8H is /3 2, C-8H is a. 1, ,,.... 11 ""' cc ,,,, ~ B c;,0 ~ ~ r-- f-- IfsC 0 C-3 """"""' """"""' -16became centered on the possibility that the rearrangement product was the diosphenol ~. Isomerization of hydroxyketone £.::1, as indicated in the alternative mechanistic routes shown above in Chart E, would afford either diosphenol E-1 or epimeric diosphenol E-2. """""' """""' Cis-fused hydrindones being more stable than their trans-fused epimers20, the formation of diosphenol E-1 would seem to be preferred """""' to that of E -2 . Diosphenol E-1 can be for med as indicated by an initial """"' """"" retro-aldolization performed on hydroxyketone C-3. Subsequent re"""' cyclization would yield the C-9 epimeric methylated hydroxyketone F-1. This hydroxyketone would then rearrange to 'diosphenol E-1 by """" ~ the mechanism originally postulated. Considerations of steric effects in a study of molecular models shown ·ta· Chart F indicate that the cis-fused hydroxyketone F-1 is equally as stable as the cis-fused -- -- ""'""" hydroxyketone C-3. The relative stability of hydroxyketone C-3 and """""' """"' F-1 would not therefore present a barrier to the formation of hydroxy- """"""' ketone F-1 by the retro-aldolization reaction. Diosphenol E-1 can """""' ~ also be formed by an alternative route which would afford the initially desired diosphenol C-6 as a reactive intermediate. Re""""""' arrangement of diosphenol C-6 to a dienic 9-membered ring compound """'"' and subsequent cyclization would afford diosphenol """"""" E-1. . If diosphenol .......,_ E-1 is the rearrangement product, then its degradation by the route already described would result in the formation of hydrindanone ~. Hydrindanone B has been synthe- -17Chart F 0 ...... 0 OH C-3 Clfs 0 0 OH 1 2 sized by the route shown in Chart G on the following page. Comparison of ketone F-2 with that of ........,_ D-6 would firmly establish whether ~ diosphenol kl is indeed the rearrangement product. A number of observations pertaining to thermal rearrangements, stereoselectivity of alkylation, and lithium--ammonia reductions either noted in this laboratory or cited in the literature indicated that this synthetic approach was feasible. Reports by both Claisen and Burgstahler offered support for the postulated thermal -- -- isomerization of allyl vinyl ether G-2 to the aldehydic ketone G-3. -18- Chart G '/"" c~ I ) 0 llf CHOH 1 OHC _ " c;.O ------) .....-. c;. 0 ycH, c;O ~ ~ 0 ~c 4,,,.... .....-. c;O c;O 5 ~c ~~c~ 0 6 0 F-2 ~ 0 .....-. ) 7 ,,,.... 0 c~ c~ c~ 3 2 .....-. .....-. -19Claisen in his original note 21 disclosed the rearrangement of 3-allyloxymethylene camphor VI to the C-allyl derivative VII. """"""" ,,......... Burgstahler later expanded this synthetic technique by incorporating o:HOC~CH _: VI VII """' """"""' 22 the participating allyl group into the ring . It was felt that ketone G-4 could be stereoselectively alkylated """"""' in a manner to afford the j3 axially oriented alkylating agent residue. Ireland and Kierstead found that 2a, 6j3-dimethyl-6a-(E,-isopropylphenyl)-cyclohexanone (VIII) upon alkylation with 1, 3-dichlorobutene"""' 2 afforded the axially oriented 3-chloro-2-butenyl substituted cyclo18 hexanone ,,........ IX in 69% yield . When they treated 2a-(3, 3-ethylenedioxybutyl)-6a -(p-isopropylphenyl)-6{3-methylcyclohexanone (X) """ with methyl iodide, the axial 2j3-methylated ketone ,,......... XI was produced in 89% yield. Ireland has also found the same results with analogous compounds in which the phenyl group does not contain an isopropyl substituent. An earlier stage in the synthesis of dl-Desoxypodocarpic acid (ill in our laboratories 7 involved the alkylation of 2, 6-dimethyl6-phenylcyclohexanone (XII) with methallyl chloride. The alkylated """""""' -20- c~ Cl c~~I •' ........ ,................... 0 liaC x XI """"' ""'""' product~ was again shown to be the one derived from axial attack. liaC CH.a HaC c~ XII ,,...,,_ XIII ......,_ XN ~ -21The lithium-ammonia reduction of enone ,,....,....,.. G-7 was expected to lead to both ketone .,.,,...,... F-2 and the trans-fused epimer (with the C-8 - hydrogen a). This expectation is based on the observation that ketone C-7 yields a mixture of both hydrindanones ,,...,.._ ill and ,,....,... IV under similar .,.,,...,... conditions 7 . Both ketones would therefore be available for comparison with the degradation ketone D-6. """""" The synthesis was effected as follows. Selective 0-alkylation of the formyl ketone G-1, prepared in the manner reported by Ireland and Kierstead18 , with methallyl alcohol and Q_-toluenesulfonic acid ~ afforded the allyl vinyl ether .......,.,.,... G-2 in 81 % yield. Claisen rearrangement of the enol ether G-2 yielded the methallyl aldehydic ketone G-3, """"" ~ which was converted to the methallyl ketone G-4 upon mild base treat"""""' ment in 80% overall yield. Alkylation of methallyl ketone G-4 in 95% """""' yield was effected with methyl iodide. Potassium t-butoxide was employed as the base. The major alkylation product was the expected axial p-methylated ketone ........._ G-5. The minor product was determined to be the amethylated epimer ........,._ XIII as follows. The minor product was eluted from a silicic acid column immediately prior to ketone 0, but quantitative separation was difficult to achieve. The nmr spectral analysis of the crude product indicated that the ratio of j3-methylated material .......,.,.,... G-5 to the minor product was 5: 1 (e.g., absorption of C-2 methyl singlets at 1.10 ppm vs. 0. 68 ppm respectively). The attempt to separate what was assumed to be the epimeric methylated product -22through silicic acid chromatography resulted in an early cut whose ratio of trans-dimethyl XIII to cis-dimethyl G-5 was 9:1. This material, oxidized with osmium tetroxide and £-periodic acid23 , ~ -- ~ afforded a compound which was shown to be identical with authentic diketone (obtained from F. F. Giarrusso 7) by infrared and nmr m- spectral analyses and mixture melting point. A reinvestigation of the quantitative alkylation of ketone XII """"" with methallyl chloride indicated the presence of both epimeric ketones G-5 and XIII. The ratio, as determined by nmr spectroscopy, was """"' ~ 9: 1 in favor of the trans-dimethylated ketone XIII. The C-2 methyl """"" group absorption of ketone ~ occurs at 0. 68 ppm and the corresponding absorption of ketone G-5 occurs at 1.10 ppm. This ratio was """"" confirmed by gas chromatographic analysis which indicated that the cis-dimethylated ketone ,,....,....,... G-5 is eluted from the SE-30 column after 675 sec. and that the trans-dimethylated ketone XIII is eluted from the """""' column after 73 5 sec. ( 152 ° oven temperature) . The factors controlling the direction of alkylation of an enolate anion are difficult to quantify, and differing viewpoints and observations have consequently been reported. House, in a recently reported 24 study concerning the stereochemistry of alkylation of 4-t-butylcyclohexanone, concludes "that there is no inherent factor which strongly favors the alkylation of a cyclohexanone enolate anion from that direction which will form a product with an axial alkyl substituent." But a high degree of stereoselectivity during alkylation of a number of cyclohexanone compounds in these laboratories has been observed. -23Ireland concludes 18 "that during the transition from enolate to product, the lower energy pathway is that which maintains the maximum orbital overlap between the carbonyl 1T system and the negative charge on the a carbon. The population of that conformation wherein the phenyl group is more equitorially oriented is probably very great. Therefore, the maintenance of overlap between the orbitals of the carbanion and the carbonyl group dictates the formation of only one product and is more important in determining the lower energy pathway than steric consideration in the enolate. In cases where the conformation of the enolate is not nearly so fixed, this stereoelectronic requirement may be satisfied in more than one conformation, and then other considerations will become more important in determining product stereochemistry." The synthesis of ketone ,,...,..._. F-2 was completed by first formation of the diketone ,,...,..._. G-6 in 85% yield; the oxidation of the methallylated ketone .2.::§ being effected with osmium tetroxide and E_-periodic acid23 . The indenone ,,...,..._. G-7 was then prepared in quantitative yield through an internal aldolization reaction of the dione with the aid of potassium t-butoxide. The reduction of indenone G-7 and subsequent chromic acid """"' 17 oxidation of the crude product afforded an isomerically pure white crystalline solid in 90% yield. It is to be recalled that lithium-ammonia reduction of the indenone C-7 did not selectively lead to the """""' thermodynamically more stable cis-fused indanone III, but to ·a mix- - """" -24- ture of both cis-fused """' III and trans-fused ..........-. IV indanones in the ratio of 25 2:3 7 . Stork observes "the relative energies of the stereoelectronically allowed transition states, rather than those of the reduction products, will determine the stereochemistry of the products." In other words, if conformational interactions do not forbid overlap of the anion with the enolate system, the reduced product will reflect that j3-carbanion which can maintain most efficient overlap with the enolate. An examination of molecular models of the carbanions generated by lithium-:-ammonia reduction of indenone C-7 """"' indicates that both {3-carbanions can effectively over lap with the enolate. The formation of an isomeric mixture is consequently reasonable to expect. A molecular model study of the carbanions generated by the lithium--ammonia reduction of the indenone G-7 also """"" indicates that a mixture of both the trans -fused ketone XVI and the """"" desired cis-fused ketone F-2 could be expected. The results would ~- """""" seem to indicate that consideration of subtle factors, whose effects are difficult to weigh, play an important role in metal- -ammonia reductions. Since the energy difference between the two possible {3-carbanions in each reduction is quite small, factors such as relative molecular solvation and steric interactions in the transition state probably control the product mix. The hydrindanone product was assumed to be either ketone F-2 or XVI. Authentic trans-fused ketone XVI was obtained """"'"""'""""""' ~ 19 by pyrolysis of the lead sau of diacid ~ of known1 8 ·configuration. -25- xv XVI "'"'"""' """" A comparison of both spectral and melting-point data of the transfused ketone XVI and the reduction product shown in Table III on the """"-"' following page indicates that they are not identical. A comparison of both the spectral and melting-point data of the indanone formed in 90% yield by lithium- -ammonia reduction and the degradation ketone D-6 indicate that they are identical - both being hydrindanone """""' F-2. This proves that the compound prepared by potassium t-butoxide - """"""' rearrangement is indeed the diosphenol E-1. """"' The relative merits of the alternative mechanisms proposed for the formation of diosphenol E-1 (shown in Chart E) could be investi""""" gated by the determination of the stability of the originally desired diosphenol 0 to the potassium t-butoxide rearrangement conditions. The conversion of diosphenol C-6 to diosphenol E-1 with the aid of """""' """""' potassium t-butoxide would strengthen the possibility that diosphenol C-6 is an intermediate in the conversion of hydroxyketone .......,_ C-3- to """"' -26- Table Ill c() 0 XVI """'""" 79.5-80.5 ketone mp ( ° C) ketone mmp ( ° C) -1 F-2 0 100.5-101 ~ XVI+ F-2 ffielte cfbe'twe en 79 and 93° ir (cm ) v CHC1s max nmr (ppm) o CC14 1735 1738 1. 09 ~s~ 1.39 s 1. 33 (s) 1. 40 (s) TMS C-9 methyl C-7 methyl C-6 was prepared in 56% yield by chromic diosphenol """""' E -1. Diosphenol .......,_ acid 17 oxidation of hydroxyketone ,...,...,,... C-2 at 0°. Diosphenol ,...,...,,... C-6 was shown to be stable to the potassium !_-butoxide conditions. This fact eliminates the possibility that diosphenol ,,..,,._ C-6 is an intermediate in the _rearrangement, and consequently strengthens the mechanism involving the retro-aldolization. Diosphenol E-1 was also shown to be stable to ~ potassium t-butoxide. The sensitive nature of the hydroxyketone system C-2 with """"" respect to oxidative cleavage is indicated by the following observations. -27- Mainly starting material was recovered when the oxidation was performed at -10°, and the major products when the reaction was performed at +10° seemed to be a mixture of the diacid C-1 and the ,___ ~ corresponding anhydride. A compound, isolated in 14% yield from the product mixture when the reaction was performed at 0°, was shown to be the diacid C-1 by both infrared spectral comparison and mixture """""" C-2 """""""' ~ 0 0 OC~ OH C-6 """""" keto · ~ enol ~ XVII ............... melting point. Another portion of the product mixture was thought to be the corresponding anhydride. The spectral and chemical properties of the oxidation product aided in its assignment as diosphenol C-6. The infrared spectrum of """"""' diosphenol .......,.,..... C-6, which exhibited a purple color with ferric chloride test solution 13 , displayed absorption bands at 3480 and 3240 cm -i (enolic -OH), and 1695 and 1652 cm -i (enolic dione). A maxirnum -28was exhibited at 259 mµ (E 9, 620) in the ultraviolet. The nuclear magnetic resonance spectrum exhibited signals at 50. 93 (s, 1, C-9 -C!b); 1.20(s,1, C-7 -C!b); 3.22, 3.24, 3.28, and 3.30(q,1, the secondary splitting being caused by W coupling, J 1 _ 8 = 3 .4 Hz, C-8 H); and 6.68(d,1, J = 3.4 Hz, vinyl H). The white crystalline "' "' a-cis-fused diosphenol 0, which melted at 121-123°, started melting at 105° when mixed with the ,B-cis-fused diosphenol E-1. The -- """"' formation of the enol ether XVII, quantitatively prepared from a-cis-~ diosphenol ,,.._,....,.... C-6 with dimethyl sulfate employed as the alkylating agent, is a confirmation of the presence of the diosphenolic grouping. Smaller scale chromic acid oxidations, worked up by extracting the chromic salts into water, afforded a yellow crystalline solid instead of the white diosphenol ,,.._,....,.... C-6. Preparation of diosphenol "'"""" C-6 involved the technique of filtering off the chromic salts and concentrating the filtrate in the presence of sodium bicarbonate. The yellow color is characteristic of a -diketones. A solution of the yellow compound in chloroform did not impart any color upon the addition of ferric chloride test solution. A drop of pyridine was then added to the solution, which was heated on the steam bath for thirty seconds; the system now turned purple. The infrared spectrum of this compound in a chloroform solution exhibited a carbonyl absorption band at 1748 -1 -1 cm , with a shoulder observed at 1758 cm ; the absence of hydroxylic absorption was noted. The nuclear magnetic resonance spectrum displayed signals at o0. 89 and 1. 09 (two s, 6, C-9 and C-7 -Cf!J) ,.... and -292.52, 2.60, 2.61, and 2.72 (d of d,2, C-1 -C!b-); vinylic absorption was absent. The yellow crystalline solid, which can be converted to diosphenol C-6 upon stirring with sodium bicarbonate, most likely is "'""""" the corresponding non -enolized a -diketone. The conversion of hydroxyketone F-1, the intermediate in the "'""""" retro-aldolization mechanism, to diosphenol E-1 upon the addition """"' of potassium t-butoxide would offer additional support for this mechanism. Hydroxyketone F-1 was consequently synthesized, as shown """""' below. Selenous acid effected the quantitative oxidative conversion of IfsC 0 XVIII F-1 ~ ~ 0 OH E-1 ~ the unsaturated ketone G-7 to the dione XVIII. The reduction of the """"' ~ dione to the unsaturated hydroxyketone ,,...,._ F-1 in 90% yield was effected by sodium borohydride. This hydroxyketone was quantitatively converted to the diosphenol E-1 upon being caused to reflux with an """"' -30- -- excess of sodium hydroxide. Hydroxyketone F-1 was also converted to diosphenol E-1 in the presence of potassium t-butoxide approximate- - """" ly five times as quickly as the starting material hydroxyketone _...,... C-3 was converted to diosphenol E-1. Aliquots, extracted with a syringe """'"""' at suitable intervals, were quenched with hydrochloric acid and worked up as usual. The rate of the rearrangement was determined -- by vapor phase chromatography at 275°--diosphenol E-1 eluted after 40 sec. and hydroxyketone F-1 eluted after 46 sec. """""' The proposed structure of diosphenol E-1 was confirmed by """'"""' the heavy-atom single-crystal x -ray crystallographic technique. This technique was also applied to hydroxyketone C-3 26 and hydroxyketone ,,...,....,... LJ. All three substances were studied as their :e_-bromobenzoate derivatives. Inspection of the conformational drawings below indicates the presence of a 1, 3-methyl-phenyl interaction in the :e_-bromobenzoate derivative ~ of hydroxyketone ~ . The non-b<;mded distance from the carbon atom of the methyl group to the closest carbon atom of 0 the benzene ring is 3. 35 A. Inspection of the E,-bromobenzoate derivative XX F-1 indicates the presence of a 1, 3............ of hydroxyketone ,,....,...,... methyl-methyl interaction. The non-bonded distance from the carbon atom of one methyl group of derivative ............ XX to the carbon atom of the . 0 other methyl group is 3. 33 A; the corresponding closest H-H inter0 action is 2 .14 A. A similar methyl-methyl interaction is also present in the .E_-bromobenzoate derivative ~of diosphenol !.::..! . The nonbonded distance from the carbon atom of one methyl group of deriva- -310 tive ~to the carbon atom of the other methyl group is 3 .43 A; 0 the corresponding closest H-H interaction is 2. 03 A. The non-bonded distance cited above refers to the static crystalline state in which the plane of the benzene ring is "frozen" such that it is spatially as far removed from the methyl group as possible. The freely rotating benzene ring in solution would impose a more serious steric interaction than is indicated by the non-bonded crystalline interactions. A possible driving force for the initial retro-aldolization and subsequent recyclization is the conversion of hydroxyketone C-3 containing the ~ 1, 3-steric interaction of the methyl and phenyl groups to the hydroxyketone ~ which contains a less sterically encumbered 1, 3-methylmethyl interaction. Since the rearrangement product has been shown to be 7~, 9/3-Dimethyl-2-hydroxy-7 a-phenylhexahydro-8j3H-1, 2-indene-3-one (E-1), the previously described tricyclic keto acid formed by oxida""""' tion of the rearrangement product and subsequent cyclization would therefore seem to be 7-keto-5-isodesisopropyldehydroabietic acid (~). Inversion of the configuration of the C-5 hydrogen of acid~' possibly through palladium catalysis 27 , would allow entry into the natural abietic acid series. Hydroxyketone C-3 could consequently """"" be employed as the common starting material for entry into two different terpenoid series, the podocarpic and abietic acid series, if the assignment of the tricyclic keto acid as ......,.,...... H-2 is correct. It was therefore decided to prove that the tricyclic keto acid l -33is indeed H-2. The conversion of diosphenol E-1 to the known tri~ ~ cyclic acid H-3, as shown in Chart H, would confirm the assignment ~ of the tricyclic keto acid as 7-keto-5-isodesisopropyldehydroabietic acid (H-2). The anhydride H-1 was formed in 92% yield from the """""' ~ Chart H E-1-t """"' 0 0 1 IfsC IfsC 2 3 diosphenol E-1 by initial oxidative cleavage with the aid of hydrogen peroxide--sodium hydroxide 28 , and subsequent cyclization with acetic ~ anhydride. Intramolecular acylation was effected by cyclization of the anhydride H-1 in polyphosphoric acid2 9a, 2 9b to the tricyclic """""' . keto acid ,,...,,...,,... H-2 in 86% yield. ·The spectral and physical properties of keto acid H-2 are indicated in Table IV. Hydrogenolysis resulted in """""' the quantitative conversion of tricyclic keto acid H-2 to tricyclic """"" acid H-3. The spectral and physical properties of tricyclic acid H-3 """"' """"' -34are also indicated in Table IV. Melting points of 146-147° 3 oa and Table IV · PPA Series melting point ( ° C) ir (cm -i) 162-164 3560-2224 3540-2320 1750 (m) 1710 (sl 1675 (s 1599 (s 1750 (m) 1710 (s) v CHCls max acid hydroxyl monomer acid carbonyl shoulder dimer acid carbonyl ketone carbonyl aromatic double bond UV (mµ) 172-173.5 ft. C:fisOH 1600 (w) 253 ( E 10, 750) max nmr (ppm) o CDCls TMS 1.23 (s) 1. 01 (s) 1. 49 (s) 1. 39 (s) 2.25 (d, J = 1.5 Hz) 2 .22 (m) 2.66(swith w coupling) 145-147° 3 0b have been reported in the literature for 5-isodesisopropyldehydroabietic acid. The melting point of the hydrogenolysis product was determined to be 162-164°. Examination of the nmr spectra confirms that the polyphosphoric acid cyclization product is not 7-keto-5-isodesisopropyldehydroabietic acid (H-2) and that the """"' hydrogenolysis product is also not 5-isodesisopropyldehydroabietic acid (H-3). The chemical shift of the two C-6 methylene hydrogens """""' -35of keto acid H-2 would be expected to occur at approximately 3. 0 ppm. """"" The methyl absorption of acetophenone occurs at 2. 59 ppm31 ; the C-6 methylene hydrogens of 7-ketodesoxypodocarpi~ acid (C-5) occurs """"' 14 at 3 .10 ppm . No such adsorption is present in the spectrum of the· cyclization product. The absorptions at 2. 25 ppm in the cyclization product and 2. 22 ppm in the hydrogenolysis product would not be expected in 7-keto-5-isodesisopropyldehydroabietic acid and 5-isodesisopropyldehydroabietic acid respectively. A comparison of reported melting points indicates that the hydrogenolysis product also cannot be any of the three other C-4 and/or C-5 isomeric acids. The melting point of desoxypodocarpic acid (ill has been reported as· 232-233° 32 and 233-235° 7 . The melting point of 5-isodesoxypodocarpic acid (XXII) has been reported as 206-207° 32 and 210-211° 18 . ~ The melting point of desisopropyldehydroabietic acid (~ has been reported as 177-179° 3 0b and 174. 5-175° 18 . ~c XXII ............,_ XXIII ~ -36An alternative mode of cyclization of anhydride H-1 is illus~ trated in Chart I. The chart employs diacid ......,..... I-1 (corresponding to anhydride H-1) for descriptive purposes. An examination of ~ molecular models of diacid ......,..... I-1 indicates that the conformation necessary for cyclization to form keto acid I-2 results in a diaxial ""' carboxyl-aromatic ring interaction. A diaxial dimethyl interaction and an axial -CH2 C02 H grQuping al;'e noted in the diacid conformation required for cyclization to form keto acid ,,...,....,... H-2. The relative energies of the two conformations are difficult to assign. Inspection of the nuclear magnetic resonance spectrum (re Table IV) of the tricyclic keto acid indicates that acylation involved the C-4 carboxyl grouping. Absorption of the keto acid at 2 .25 ppm can be ascribed to the methylene grouping adjacent to the carboxyl group. The methyl group of acetic acid exhibits a signal at 2 . 10 ppm 31 . A corresponding methylene grouping would be expected at approximately 2. 50 ppm, but a molecular model examination of keto acid ~2 indicates that the -C~C02 H grouping is somewhat shielded by the aromatic ring. The maintenance of this signal (at 2. 22 ppm), and the addition of a signal at 2. 66 ppm (assigned to the two benzylic hydrogens), are additional confirmation of the assignment of the tricyclic acid as ......,.... I-3 . Spectral observations of compounds obtained in the degradative scheme described in Chart J (page 38) corroborated the assign-. ment of the tricyclic acid as structure I-3. The conditions for the """" -37- Chart I Clfs 0 H-2 """"" li HsC H 1 ....... 8 6 H IfsC 3 ....... -38- Chart J I-3 1 """ "" ~ :EfsC C~CHO 2 ....... 3 "" degradative sequence employed were initially applied and worked out on the model series shown in Chart K. The starting material for the sequence was podocarpic acid 33 (i). The overall yield was .,.... 91%. Reduction of podocarpic acid over a four-day period at room temperature affords podocarpinol in only 56% yield34 . The acid -39- Chart K . oc~ OH oc~ D1m c~ HOfl:?C 2 ...... 1 ......i oc~ 1) HuangMinlon 2) DMS ~c OHC 4 3 ...... was therefore esterified prior to lithium aluminum hydride reduction. Oxidation of the alcohol K-2 to the aldehyde ,,...,,...,,... K-3 was accomplished with the recently reported solid chromic oxide-dipyridine complex35 . ~ Huang-Minlon reduction 36 and subsequent remethylation resulted in the formation of 0-methylpodocarpane (K-4). A comparison of the .,..,,...,.., -40- spectral properties of the reduction product with that of authentic 0-methylpodocarpane (K-4) reported in the literature37 indicates .,....,...,.... that they are the same. The in~rared spectrum of the reduction product taken in chloroform exhibits absorption bands at 1605 cm - 1 (m) and 1570 cm - 1 (w). The assymetric methyl bending of the gem1 dimethyl group is split into a doublet38 at 1372 and 1367 cm- • The absorption bands reported in the literature (spectrum taken in chloroform) occur at 1607 cm -l (m) and 1566 cm -i (w). The ultraviolet spectrum of the reduction product exhibited A. ~~ C2IfsOH = 280 mµ. (E 1,490), A.shoulder= 286 mµ.(E 1,320). The ultraviolet spectrum reported in the literature exhibited A. ~~ C2IfsOH = 279 mµ. ( E 1, 500), A. shoulder = 284 mµ. ( E 1, 3 90). The reduction product was an oil. 0-methylpodocarpane, reported in the literature, occurs as an oil which crystallizes (mp 31-32°) upon standing. The tricyclic acid I-3 was reduced to the tricyclic alcohol """" J-1 in 83% yield with the aid of lithium aluminum hydride. The ~ nuclear magnetic resonance spectrum ( CDC1s) of the alcohol exhibited a triplet at 3. 60 ppm (J = 8 Hz). The triplet was assigned to the two hydrogens attached to the carbon atom containing the hydroxyl group. The corresponding signal would be expected to be a singlet if the alcohol was in the isoabietic acid series. The tricyclic alcohol J-1 was oxidized to the tricyclic aldehyde """"""' J-2 in 97% yield with the aid of the solid chromic oxide-dipyridine 35 . reagent . The nuclear magnetic resonance spectrum (CDC1s) of the ~ -41aldehyde exhibited two singlets at 0. 94 ppm and 1. 32 ppm assigned to the two methyl groups, and a triplet at 9. 83 ppm (J = 1 Hz) assigned to the aldehydic hydrogen. The corresponding aldehydic signal would _be expected to be a singlet if the aldehyde was in the isoabietic acid series. The Huang-Minlon procedure 36 was employed to reduce the tricyclic aldehyde -"'-"'""' J -2 to the tricyclic hydrocarbon ,,...,....,... J -3 in 94% yield. The infrared spectrum (CHC1s) of the tricyclic hydrocarbon exhibited 1 a singlet absorption band at 1373 cm- (m). The corresponding signal would be expected to be a doublet (indicative of the gem-dimethyl grouping38 ) if the hydrocarbon was in the isoabietic acid series. The 60 Mc nuclear magnetic resonance spectrum (CDC1s) of the hydrocarbon exhibited two single peaks at 2. 54 ppm and 2. 72 ppm assigned to the two benzylic hydrogens. Two much smaller peaks at 2.28 ppm and 3.01 ppm (J = 18 Hz) were assumed to be the remainder of the Ab d of d pattern. This section of the spectrum resolved itself into a more recognizable AB pattern (2. 90, 2. 72, 2. 54, 2. 36 ppm) when the spectrum was retaken on the Varian Associates HA-100 spectrometer. The corresponding benzylic signal would be expected to be a complicated multiplet if the hydrocarbon was in the isoabietic acid series. The nitrile ~ 39 exhibits a multiplet assigned to the benzylic hydrogens centered at 2. 83 ppm. 0-methylpodocarpane ~ exhibits a multiplet assigned to the benzylic hydrogens centered at 2. 82 ppm. Methyl absorptions in the 60 Mc spectrum of the hydro- -42- CR,, XXIV ~ carbon J-3 occurred at 1. 00 ppm and 1. 40 ppm. Integration indicated """"'""" that two methyl groups were present at 1. 00 ppm. The methyl group of the ethyl chain would be expected to exhibit a triplet signal. This signal was not sufficiently resolved. A survey of the nuclear magnetic resonance spectra14 of the polyphosphoric acid cyclization product L-2 of the diacid L-1 in the """""" """"' pimaric acid series, and subsequent hydrogenolysis product L-3 40 , """"" indicates that the ~ carboxyl group participated in this cyclization. Diacids ..........._. L-1 and ,,..,,... I-1 have the necessary -cis-1, 3-carboxyl-phenyl reChart L Clis 1 ...... 2 ...... ~ CR,, Pd/C ) 3 ....... -43- lationship required for this "abnormal" cyclization. The diacid £.:.! and its epimer at the -C~C02 H center do not have this relationship. Models indicate that the signals which correspond to the methylene group adjacent to the carboxyl group in both L-2 and L-3 """""' """"' would be expected to be downfield relative to the corresponding methylene absorption in J-1 and J-2. This is because only the methyl~ ~ ene groups of J -1 and J -2 are shielded by the aromatic ring. This """""' ~ has been found to be the case. An indication of the purity of keto acid L-2 and acid L-3 was not given 14 . A signal at approximately ~ ~ 3. 0 ppm, characteristic of the c6 hydrogens Ci. to the carbonyl group of the keto acid which would be formed if cyclization had occurred in the "normal" manner, was absent. A signal, ascribable to the methylene group adjacent to the ester gouping of keto ester L-2, was """""'" centered at 2. 52 ppm. It consisted of an unsymmetrical doublet (J = 2. 5 Hz) and four smaller peaks which integrated for a total of approximately three protons. The signal, ascribable to the same methylene grouping of ester L-3, was centered at 2. 40 ppm. It ""'""""' consisted of a singlet peak and four smaller peaks which integrated for a total of approximately three protons. The benzylic hydrogens exhibited two clean single peaks at 2. 70 ppm and 2. 83 ppm. -44- EXPERIMENTAL Melting points, unless otherwise noted, were taken on a Kofler Hot Stage and are uncorrected. Boiling points are also uncorrected. Infrared spectra were recorded on a Perkin-Elmer Infrared Spectrometer Model 237B; and ultraviolet spectra were recorded on a Cary Recording Spectrometer Model llM. Nuclear magnetic resonance spectra were recorded on a Varian Associates Model A-60A Nuclear Magnetic Resonance Spectrometer. All gas chromatographic analyses were taken on an F and M Model 810 Gas Chromatograph using a 6 ft silicone gum rubber (SE-30) column. The pressures of the gases employed during the vapor phase chromatographic analyses were helium, 50 psi; hydrogen, 22 psi; and compressed air, 24 psi. A Buerger precession camera and General Electric-Datex diffractometer were employed in the X-ray analyses. Petroleum ether, unless otherwise noted, refers to the fraction boiling in the range 30-60°. Microanalyses were performed by Spang Microanalytical Laboratories, Ann Arbor, Michigan; and Elek Microanalytical Laboratories, Torrance, California. 2indene-3-one(E-1) .(a) Potassium t-butoxide rearran ement from ~ ~ To a stirred solution of 1. 31g(11.7 mmoles) of potassium J.-butoxide in 60 ml of J.-butyl alcohol (freshly distilled from calcium hydride) contained in a nitrogen-protected flame-dried flask was dropwise added a solution of 1. 50 g (5. 85 mmoles) of unsaturated -45hydroxyketone C-3 in 60 ml oft-butyl alcohol {dried as above). - The ~ blood-red colored solution was stirred for an additional 1 hr at approximately 45 °, acidified with iced concentrated hydrochloric acid, and diluted with 100 ml of water. The mixture was thrice extracted with 100 ml portions of an ether--benzene solution (4:1). The com- bined organic layers were washed three times with 40 ml portions of water, twice with 25 ml portions of a saturated salt solution and then dried (N~S04 ). Removal of the solvent at reduced pressure afforded 1. 50 g of solid diosphenol E-1. Gas chromatographic analysis of the ~ crude product at 285 ° indicated the presence of a single peak with a retention time of 25 sec accounting for 95% of the material and two minor impurities. Chromatography on silicic acid afforded 1. 37 g (91%) of diosphenol E-1 ,,..,,,...,.... on elution with 3 1. of a 10% ether--petroleum ether solution. The analytical sample, obtained as colorless plates melting at 162-164°, was prepared by two crystallizations of a portion of this material from ether--petroleum ether. Anal (Spang) calc'd for 7 C17H20 0 2 : C, 79.65; H, 7.86%. Found: C, 79.61; H, 7.84%. ir (CHC13 ) 3500, 3350 (OH), 1700, 1650 cm -i (enolic dione); uv max (MeOH) 262 mµ (E 9, 038); nmr (CDC13 ) 6 1. 32 (s, 3, C-9 C~ 3 ), 1. 48 (s, 3, C-7 C~ 3 ), 2. 85 (d, 1, J = 3 Hz, C-8 ~), 5. 65 (d, 1, J = 3 Hz, C-1 vinyl H), 6. 22(s,1, OH). "' ~ By the same procedure as described above 0.100 (0.195 mmole) of the unsaturated hydroxyketone £.:1 underwent rearrangement to the diosphenol E-1 in a mixture of 0.088 g (0.390 mmole) of potassium t"""" . - butoxide and 8 ml t-butyl alcohol upon being stirred in a nitrogen atmos- -46phere for 1 hr at room temperature. There resulted 0.100 g of a pale yellow crystalline solid which was shown by vpc analysis at 300° to consist of equal amounts of hydroxyketone C-3 (18 sec) and diosphenol """"' E-1 (14 sec). -"'-"""' (b) Potassium t-butoxide rearran ement from hy:droxy:ketone F-1: To a stirred solution of 0.106 g (0. 934 mmole) of potassium t-butoxide in 6 ml of t-butyl alcohol (freshly distilled from calcium hydride) contained in a nitrogen-protected flame dried flask was dropwise added a solution of 0.120 g (0.467 mmole) of unsaturated hydroxyketone F-1 ~ in 6 ml oft-butyl alcohol (dried as above). The blood-red colored solution was stirred at room temperature. Four equal aliquots, removed and quenched with iced concentrated hydrochloric acid, were worked up as described above. The progress of the reaction was followed by vapor phase chromatography--the retention time of diosphenol E-1 was 40 sec and the retention time of hydroxyketone F-1 """"" ~ was 46 sec sec at an oven temperature of 2 75 °. The diosph~nol/hy­ droxyketone ratio of the 0. 03 0 g oil isolated after 10 min was 1/4. The ratio of the partially crystalline oil was 3. 5/5. 0 after 20 min, 4/1 after 43 min, and 6. 5/1. 0 after 80 min. (c) and (d) Sodium h droxide rearran ements from h C-3 and F-1: To a stirred solution of 2.4 ml of 40% aqueous sodium ~ hydroxide solution in 10 ml of methanol in a nitrogen atmosphere, repeatedly degassed at 0. 03 mm with the aid of liquid nitrogen, was added dropwise 0.400 g (1. 58 mmoles) of hydroxyketone ~ in 10 ml of methanol. The red colored solution, which turned pale yellow in 15 min, was heated under reflux for 3 0 min, cooled, quenched with -47iced concentrated hydrochloric acid, and then diluted with 100 ml of water. The mixture was thrice extracted with 2 5 ml portions of ether--benzene (4:1), and the combined ethereal extracts were washed three times with 10 ml portions of a saturated brine solution and dried (N~S04 ). Removal of the solvent at reduced pressure afforded 0.400 g of a white solid- -exhibiting a single peak ( 100% yield) with retention time of 23 sec by gas chromatographic analysis at 285°. The infrared spectrum, melting point, and mixture melting point with diosphenol E-1, prepared from the potassium t-butoxide rearrangement described """""" above, indicated that the isolated product and diosphenol ~ were identical. By the same procedure as described above 0. 040 g (0.158 mmole) of the unsaturated hydroxyketone F-1 ,,...,...,... underwent rearrangement to the diosphenol E-1 in a degassed solution of O. 25 ml of 40% ~ aqueous sodium hydroxide in 2 ml of methanol. There resulted 0. 040 g of an off-white crystalline solid which produced a single peak on vpc analysis at 270° with a retention time of 33 sec. The infrared spectrum, melting point, and mixture melting point with authentic diosphenol E-1 indicated that the isolated product was identical with """"" diosphenol E-1. '"'"'""" ~: A stirred mixture of 1. 40 g (5. 46 mmoles) of diosphenol ~' 5. 96 ml (62. 2 mmoles) of dimethyl sulfate 15 , 19. 95 g (144 mmoles) of anhydrous potassium carbonate, and 100 ml" of anhydrous acetone was heated under reflux in a nitrogen atmosphere -48for 22 hr. The two phase system was cooled, and the mixture was then concentrated on the rotary evaporator under reduced pressure until approximately 20 ml of acetone remained. The cooled mixture, diluted with 400 ml of water, was placed in the refrigerator for 2 hr. The pale yellow solid, separated from the aqueous solution by filtration, was washed four times with 20 ml portions of water. The residue was heated at 56° in a vacuum (O. 1 mm) until a constant weight of 1. 470 g (100% yield) was achieved and maintained. Gas chromatographic analysis of the crude product, (melting at 109-111°) at 28 5 ° indicated a single peak with retention time of 28 sec. The analytical sample, prepared by two recrystallizations of a portion of this material from ether--petroleum ether (60-75°), consisted of thick colorless platelets melting at 110. 5-111.5°. Anal (Spang) calc'd for 14 C18 H22 0 2 : C, 79. 96; H, 8. 20%. Found: C, 80. 03; H, 8. 28% . ir (CHC13 ) 1710 (C=O) 1628 (double bond linked to oxygen) 1250, 1075 cm - i (vinyl ether); uv max (MeOH) 257 mµ (E 8, 410); nmr (CDC13 ) o 1. 34 (s, 3, C-9 C~ 3 ), 1. 50 (s, 3, C-7 C~ 3 ), 2. 97 (d, 1, J = 3 Hz, C-8 !:!), 3. 46 (s, 3, OC~ 3 ), 5. 38 (d, 1, J = 3 Hz, vinyl~). ~: To a suspension of 0. 577 g (15. 2 mmoles) of lithium aluminum hydride in 3 5 ml of tetrahydrofuran (freshly distilled from lithium aluminum hydride) was added 0. 412 g (1. 52 mmoles) of unsaturated ketone D-2 in 35 ml of tetrahydrofuran (dried as above), ""'"""'"" and the reaction was heated under reflux in a nitrogen atmosphere for 1. 5 hr. After treatment of the cooled solution with 0. 15 ml of 10% -49- aqueous sodium hydroxide solution, the precipitated salts were removed by filtration, and the filtrate was diluted with 100 ml of ether. The ethereal solution was washed three times with 15 ml portions of water, two times with 10 ml quantities of saturated salt solution, dried (Na2 S04 ), and concentrated at reduced pressure. Gas chromatographic analysis of the· 0. 407 g oily residue indicated the presence of a single peak with the retention time of 24 sec accounting for 98° of the material. ir (CHC1 3 ) 3594, 3470 {OH), 1650 cm -l (double bond linked to oxygen). A solution of 1. 00 g (11. 1 mmoles) of oxalic acid in 10 ml of water was added to the 0. 407 g residue dis solved in 60 ml of acetone. ature for 48 hr. The mixture was stirred at room temper- The solution was concentrated on the rotary evapo- rator under reduced pressure until approximately 10 ml of acetone remained and 100 ml of 9:1 ether--benzene was added. The system was washed three times with 10 ml portions of 10% aqueous sodium bicarbonate, twice with 10 ml quantities of water, twice with 10 ml quantities of saturated salt solution and dried (Na2 S04 ). Removal of the solvent at reduced pressure yielded O. 390 g {95% overall yield) of an off-white colored solid, exhibiting a single peak with retention time of 26 sec by gas chromatographic analysis at 278°, which melted at 120-124°. The analytical sample, obtained after four recrystallizations of a portion from ether --petroleum ether, consisted of small colorless plates melting at 122-124 °. Anal (Spang) calc'd for C11 H22 0 2 : C, 79. 03; H, 8. 58%. Found: C, 78. 85; H, 8. 44% .. ir (CHC1 3 ) 3550, 3450 (OH), 1748 cm -1 (C=O); nmr {CDC13 ) o 1. 43, -501. 46 (two s, 6, C-9 and C-7 C~ 3 ), 3. 15 (' s, 1, Ol.I), 3. 83 ( s, 1, C-3 H). "" 7~ ~: A solution of 0. 412 g (1. 60 mmoles) of hydroxyketone ~ in 8. 6 ml (107 mm oles) of anhydrous pyridine was treated with 10 ml (107 mmoles) of acetic anhydride and stirred at room temperature for 15 hr. A nitrogen stream was employed to blow off the major portion of the solution. The residue, taken up in 100 ml of ether--benzene solution (4:1), was successively washed three times with 8 ml portions of 5 N sulfuric acid, three times with 10 ml portions of water, three times with 10 ml quantities of aqueous sodium bicarbonate, two times with 10 ml portions of water, two times with 10 ml portions of saturated salt solution, and dried (N~S04 ). Removal of the sol vent at reduced pressure yielded 0. 446 g of a pale yellow solid. Crystal- lization from acetone--hexane afforded 0. 319 g of a white crystalline solid, mp 172-176°. Chromatography of the mother liquor on 15 g florisil (60-100 mesh) afforded an additional 0. 055 g (83% combined yield) of crystalline material, mp 174-177°, on elution with 350 ml of ether. The analytical sample, a colorless hexagonal tubular solid, was prepared by three additional recrystallizations of a portion of the once recrystallized material from acetone--hexane and melted at 175176. 5°. Anal (Spang) calc'd for C19 H24 0 3 : C, 75. 97; H, 8. 05%. Found: C, 75. 98; H, 8. 13%. ir (CHC13 ) 1763 (ester C=O), 1738 (C=O), 1215 cm -i (ester); nmr (CDC1 3 ) o 1. 38, 1. 46 (two s, 6,. C-9 and C-7 C~ 3 ), 2.13 (s, 3, OCOC~ 3 ), 5.13 (s, 1, C-3 ~). -517{3-9{3-Dimeth l-7a-phenylhexahydro-8f3H-indan-2-one (D-6): After 0. 155 g (22. 3 atoms) of lithium wire had been allowed to dissolve in ca 100 ml of liquid ammonia (dried over sodium immediately preceding the reaction), a solution of 0. 270 g (1. 100 mmole) of the acetoxy ketone D-5 in 10 ml of tetrahydrofuran (dried over lithium . ~ aluminum hydride immediately preceding the reaction) was added dropwise. The reaction mixture was stirred for 1 hr. Solid ammonium chloride was then added to discharge the blue color, and the ammonia was allowed to evaporate. The resulting solid white residue was taken up in 100 ml of ether--benzene (4:1). The ethereal solution was washed with 10 ml of 3 N hydrochloric acid, thrice with 10 ml portions of water, twice with 10 ml portions of saturated salt solution and dried (Na2 S04 ). Removal of the solvent at reduced pressure afforded 0. 232 g of a yellow oil, whose infrared spectrum (e.g., carbonyl absorption at 1735 cm - 1 and hydroxy absorption at 3590 and 3420 cm -l in chloroform) indicated partial reduction of ketone D-6 to the corresponding alcohol. ~ Consequently an ice-cooled mixture of.the 0. 232 g crude product in 5 ml of acetone was oxidized with 0. 23 ml of 8 N aqueous chromic acid solution 17 . After the mixture had been stirred in the cold for 15 min, isopropyl alcohol was added to destroy the excess oxidant, and the mixture was diluted with 100 ml of water and thrice extracted with 60 ml portions of ether--benzene (4:1). The combined organic layers were washed two times with 5 ml quantities of 10% aqueous sodium bicarbonate,. two times with 10 ml portions of water, two times with 10 ml portions of -52brine and dried (Na2 S04 ). Removal of the solvent at reduced pressure yielded 0. 205 g of an orange oil which crystallized overnight. The infrared spectrum indicated the absence of hydroxylic absorption. The crude product was purified by preparative thin layer chromatography on a 2 mm silicic acid plate (20 x 20 cm), on elution with 30% ether--petroleum ether. The 0. 129 g (59% yield) white crystalline solid, which exhibited a single peak with retention time of 24 sec by vapor phase chromatography at 300°, was isolated from a band whose Rf was 0. 60. A sample of analytical purity, obtained as thin plates after two recrystallizations of a portion from hexane, melted at 100. 5-101°. A mixture of this material and the -cis-fused ketone III, ,,...,... mp 100. 5-101°, melted between 70 and 102°. The semicarbazone, formed by the procedure of Fieser 41 , afforded a yellow crystalline solid of melting point 220. 5-222° after two recrystallizations from a mixture of methanol and ethanol. A mixture of this material and the semicarbazone of the trans-fused ketone IV, mp 221-223°, melted between 208 and 243 °. Anal (Elek) calc 'd for C18 H25 N3 0: C, 72. 20; H, 8.42; N, 14.04%. Found: C, 72.18; H, 8.56; N, 13.94%. 7j:3-9j:3-Dimeth l-7a- hen lhexah~dro-8ciH-indan-2-one (XVI : An intimate mixture of 0. 200 g (0. 658 mmole) of diacid E (sample obtained from R. E. Ireland) and 0. 214 g (0. 798 mmole) of lead carbonate 19 , contained at the sealed end of a 7 mm pyrex tube was heated at 28 5 ° (at 5 mm pressure) for 2 hr, the ketone immediately distilling as it was formed. The distillate was taken up in 100 ml of ether--petroleum ether (4:1), and the solution was washed three times with 5 ml portions of 5% potassium hydroxide solution, two -53times with 5 ml portions of saturated salt solution and then dried (Na2 S04 ). The solution, concentrated under reduced pressure, afforded 0. 133 g of an off-white colored solid residue. The crude product, placed on a 2 mm thick silicic acid plate (20 x 20 cm) and developed with 40% ether- -petroleum ether, yielded 0. 107 g {67% yield) of a white colored crystalline solid (exhibiting a single peak with retention time of 97 sec by vapor phase chromatography at 256°) which was extracted from the major band, Rf 0.44. The analytical sample, obtained from two recrystallizations of a portion from hexane, melted at 79. 5-8 0. 5 ° . The trans -fused ketone XVI and ketone ~ F-2 exhibited the same vapor phase chromatograph retention time and """""' could therefore not be distinguished by this technique. The thin layer (i.e., silicic acid analytical plate developed with 50% ether--petroleum ether) chromatographic Rf value of the trans -fused ketone ~ was 0.51 and that of ketone LJ was 0.49. Anal (Elek) calc'd for C 17 ~ 2 0: C, 84.25; H, 9.15%. Found: C, 84.25; H, 9.22%. ir (CHCis) 1710 cm(C=O); nmr (CC~) o 1.09 (s,3, C-9 C_!b), 1.39 (s,3, C-7C!b). An intimate mixture of 0. 060 g (0. 197 mmole) of diacid V (sample "' obtained from R. E. Ireland) and 0. 064 g (0. 240 mmole) of lead carbonate 19 , contained at the sealed end of a 7 mm pyrex tube was heated at 281 ° (at 5 mm pressure) for 15 min, the ketone distilling as it was formed. The glassy white oily pyrolysate, which resisted attempts at crystallization, was formed in a yield of 98% (O. 047 g). The combustion analysis of the semicarbazone derivative, mp 1 -54221-223°, was correct 7 . ir (CHC13 ) 1743 cm-1 (C=O); nmr (CC14 ) 6 0.33 (s,3, C-9 C!1 3 ), 1.28 (s,3, C-7 C!13 ). A solution of 21. 6 g (100 mmoles) of formyl ketone Q:!, prepared in the manner reported by Ireland and Kierstead 18 , 8. 30 g ~ (115 mmoles) of methallyl alcohol, and a trace of .E_-toluene sulfonic acid were dissolved in 100 ml of benzene. The solution was heated under reflux in a nitrogen atmosphere for 10 hr, cooled, and then diluted with 100 ml of ether. The organic solution was then washed four times with 10 ml portions of 10% aqueous potassium hydroxide solution, four times with 10 ml portions of saturated salt solution and dried (Na2 S04 ). Removal of the solvent at reduced pressure yielded 2 5. 0 g of a yellow oil. Chromatography of the crude product on silicic acid afforded 21. 9 g (81%) of the desired ally! vinyl ether .,...,..,.,.... G-2, a pale yellow oil which crystallizes upon standing, on elution with 15 l. of a 12. 5% ether--petroleum ether solution. Ally! vinyl ether ........,_ G-2 decomposes on the column during vpc analysis. The analytical sample, obtained as platelets, was prepared by two recrystallizations of a portion of the material from ether--isopentarie and melted at 66. 5-67. 5°. Anal (performed by H. King, U. C. L.A. Microanalytical Laboratories) calc'd for C18 H22 0 2 : C, 79. 96; H, 8. 20%. Found: C, 79. 69; H, 8. 24%. ir (neat) 1675 (C=O), 1582 (double bond linked to oxygen), 1078 cm - 1 (vinyl ether); uv max (MeOH) 278 mµ (E 11, 960); nmr (CDC13 ) 6 1. 43 (s, 1, C-2 C,!1 3 ), 1. 70 (m, 3, vinylic C,!1 3 ), 4. 37 ' =CHz), 7.4_1 (m, 1, (s, 2 -0-Cliz), 4.96 (m, 2 ....-C -55:::c=CH-0-). -""" of 15. 1 g (55. 8 mmoles) of ally!' vinyl ether _...,... G-2 under nitrogen in an oil bath at 190° for 1 hr resulted in the formation of 14. 9 g of a yellow-orange oily product. ir (neat) 1731 and 1697 cm -l (C=O) indicates that the material is the product of Claisen rearrangement 21 , 2-formyl-2-methallyl-6-methyl-6-phenylcyclohexanone (G-3). A "'"""""'"'" solution of 30. 0 g (217 mmoles) of potassium carbonate in 30 ml of water was added to the crude product dissolved in 450 ml of methanol. The mixture was stirred for 9 hr at room temperature, and then concentrated under reduced pressure to approximately 150 ml of solution. The reaction mixture, diluted with 400 ml of water, was thrice extracted with 10 ml portions of ether - -benzene (1: 1). The combined ethereal layers were washed three times with 25 ml portions of saturated salt solution and then dried (N~S04 ). Removal of the solvent at reduced pressure yielded 13. 5 g of a yellow-orange oil. Chromatography of the residue on silicic acid afforded 10. 8 g (80%) of the desired deformylated ketone G-4, a clear colorless oil, ~ on elution with 3 1. of 10% ether--petroleum ether. The oil exhibited a single peak with retention time of 66 sec at 237° during vpc analysis. The analytical sample, obtained by evaporative distillation of a portion of the material at 0. 05 mm pressure (15 cm path length) boiled at 54-57°. Anal (Spang) calc'd for C17H22 0: C, 84. 25; H, 9. 15%. 889 cm Found: C, 84. 35; H, 9. 25%. ir (neat) 1718 (C=O), 1_651, -1 (C=CH2 ); nmr (CDC13 ) o 1. 23 (s, 3, C-2 C.!i3 ), 1. 56 (s, 3, -56- To a suspension of 23. 6 g (207 mmoles) of potassium .!_-butoxide in 700 ml of dry benzene in a nitrogen atmosphere was added with stirring at room temperature a solution of 10. 0 g (41. 3 mmoles) of methallyl ketone _....,.... G-4 in 300 ml of dry benzene. After the reaction mixture had stirred for 15 min (generating a pale yellow colored enolate), it was cooled in an ice-water bath for 3 min and then 25. 0 ml (56. 8 g, 400 mmoles) of methyl iodide was added all at once. The reaction mixture was then allowed to stir overnight as the cooling bath came to room temperature. An additional 75 ml of methyl iodide was then added, and the reaction mixture was stirred and maintained at reflux for 2. 5 hr. The cooled mixture was concentrated under reduced pressure to approximately 30% of its original volume, and 200 ml of water and 100 ml of ether were added. The aqueous layer was twice washed with 70 ml portions of ether--benzene (4:1). The combined ethereal layers were washed two times with 40 ml portions of a saturated salt solution and dried (Na2 S04 ). The 238° vapor phase chromatographic analysis of the 10. 5 g (100%) pale yellow oily residue remaining after the solvent was removed under reduced pressure indicated the presence of a single peak (75 sec). The peak was assigned to the methylated product and accounted for 95% of the material. The a-methylated isomer XIII is eluted from a silicic acid ~ column immediately prior to the {3-methylated epimer ~ upo~ increasing the polarity of the eluent from 70-99% benzene--petroleum -57- ether, but quantitative separation is quite difficult to achieve. The nmr spectral analysis of the crude product indicates that the ratio of ,a-methylated material ,,...,...,... G-5 to a-methylated epimer ,,...,...,... XIII (e.g., C-2 methyl absorption at 1. 10 ppm vs. 0. 68 ppm, respectively) is 5:1. The analytical sample, obtained by distillation of a portion of the material at 0. 05 mm pressure, boiled at 84-87°. Anal (Spang) calc'd for C18 H24 0: C, 84. 32; H, 9. 44%. - ir (neat) 1698 (C=O), 1643, 890 cm Found: C, 84. 57; H, 9. 66%. -1 (C=CH2 ); nmr (CDC13 ) o 1. 10 (s, 3, C-2 C~ 3 ), 1. 33 (s, 3, C-6 ·C~ 3 ), 1. 54 (d, 3, J= 1 Hz, vinylic C!! 3 ), 4. 48, 4. 77, (two m., 2, C= C~ 2 ). 1 2a, 6,8-Dimeth 1-2,8- 3 -methall l}-6aTo a rapidly stirred nitrogen protected mixture of 0. 300 g (1_. 48 mmoles) of methylated ketone XII 42 , 0. 252 g (2. 24 mmoles) of """""' potassium _!-butoxide, and 4 ml of dry benzene was added 0. 435 ml (0. 402 g, 4. 34 mmoles) of methallyl chloride (distilled between 70. 571° immediately preceding the reaction). The pale yellow colored mixture was stirred for an additional 10 hr at room temperature, and then heated under reflux for an additional 5 hr. The cooled solution was diluted with 40 ml of water and thrice extracted with 30 ml portions of ether--benzene (4:1). The combined ethereal layers were washed three times with 7 ml portions of water, one time with 10 ml of saturated salt solution and dried (Na2 S04 ). Removal of the solvents at reduced pressure afforded 0. 381 g (100% yield) of a pale yellow oil. Gas chromatographic analysis of the crude product indicated quantitative methylative conversion. The ratio of the two epimeric -58- methylated ketones, as determined by nmr spectroscopy, was 9:1 in favor of the trans -dimethylated ketone .,..,,,....,... XIII. This ratio was confirmed by the gas chromatographic analysis which indicated that the cis dimethylated ketone G-5 is eluted from the column in 675 sec, and the ~ trans-dimethylated ketone ,,....,,...,... XIII is eluted from the column in 735 sec at the following machine settings-oven at 152°, injection port at 317°, and detector at 302°. ir (neat) 1698 (C=O), 1643, 890 cm -i (C=CH 2 ); · nmr (CDC1 3 ) o 0. 68 (s, 3, C-2 C!! 3 ), 1. 31 (s, 3, C-6 C!! 3 ), 1. 71 (d, 3, C= CH .,..... 2 ) • A solution of 6. 45 g (25. 2 mmoles) of the terminal methylene olefin G-5 (nmr analysis indicates the presence of 94% ,,....,,...,... G-5 and 6% epimeric .,....,....,... XIII) in 650 ml of dioxane (distilled immediately preceding the reaction ~ from lithium aluminum hydride), 64. 5 ml of glacial acetic acid, and 129 ml of water was treated with 0. 065 g of solid osmium tetroxide 23 , and the resulting solution was allowed to stir 10 min, during which time it became dark brown. There was then added in small portions, over a period of 25 min,22. 9 g (101 mmoles) of crystalline £_-periodic acid. After stirring for 24 hr, the reaction mixture was diluted with 100 m-1 of water and the system was extracted four times with 500 ml portions of chloroform. The combined organic layers were then successively washed three times with 300 ml portions of water, three times with 70 ml portions of 10% aqueous potassium hydroxide solution, 200 ml of water, saturated salt solution and then dried (Na2 S04 ). The solution, concentrated to dryness under reduced -59pressure, yielded 6. 50 g of a yellow oil. Chromatography of the crude product on silicic acid afforded 5. 52 g (85% yield) of the desired diketone .,...,...,.... G-6, a pale yellow crystalline solid, on elution with 12 1. of 30% ether--petroleum ether. Two successive crystallizations of a portion from petroleum ether afforded colorless crystalline analytically pure platelets melting at 88. 5-89. 5° (Thomas Hoover capillary silicone oil bath). Anal (Spang) calc'd for C17 H22 0 2 : C, 79. 03; H, 8. 58%. Found: C, 79. 13; H, 8. 60%. ir (CHC13 ) 1712, 1692 cm -l (C=O); nmr (CDC13 ) 6 1. 22 (s, 3, C-2 C!! 3 ), 1. 44 (s, 3, C-6 C!! 3 ), 1. 90 (s, 3, vinylic C!! 3 ), 2. 42, 2. 58 (two s, 2, -C!!2 -COCH3 ). The assumption that the minor product of the methylation reaction was the epimeric trans-dimethylated cyclohexanone XIII was confirmed as follows. ~ The attempt to separate what was assumed to be epimeric methylated product through the aid of silicic acid chromatography resulted in an early cut whose ratio of trans-dimethylated ketone XIII to -cis-di~ methylated ketone ,,....,....,.. G-5 was 9:1. The material, weighing 1. 013 g, oxidized with osmium tetroxide and .E_-periodic acid and worked up as above yielded a crystalline solid (mp 92-93. 5°). The melting point of authentic 2{3-acetonyl-2a, 6{3 -dimethyl-6a-phenylcyclohexanone ,,....,...,... XIII is 92-93". 5°, and the mixture melting point is 92-94°. Their and nmr spectra of both compounds were. identical. ir (CHC13 ) 1715, 1690 cm -l (C=O); nmr (CDC13 ) 6 0. 63 (s, 3, C-2 C!!3 ), 1. 29 (s, 3, C-6 C!! 3 ), 2. 06 (s, 3, -CO-CH ,,.... 3 ). A solution of 2. 00 g (7. 75 mmoles) of dione ,,....,...,... G-6 in 17 ml of . -60- azeotropically dried benzene was added dropwise to a stirred,nitrogen protected, incomplete solution of 2. 62 g (23. 2 mmoles) of potassium .!_-butoxide in 25 ml of .!_-butyl alcohol (freshly distilled from calcium hydride), and the reaction mixture was stirred at room temperature for 24 hr in a nitrogen atmosphere. The orange colored solution was acidified with 2. 5 ml of 3 N hydrochloric acid and most of the .!.-butyl alcohol was removed at reduced pressure. The residue was then partitioned between 100 ml of ether--benzene (4:1) and 20 ml of water. The organic layer was washed three times with 10 ml portions of water, once with 10 ml of saturated salt solution and then dried (N~S04 ). The solution, concentrated under reduced pressure, left a 0. 190 g (100% yield) white crystalline solid residue shown to be pure by both vapor phase (i. e. , exhibiting a single peak with retention time of 90 sec at 27 5 °) and thin layer chromatography (i. e. , exhibiting a single spot with Rf value of 0. 42 upon being developed in 50% ether-petroleum ether on a silicic acid plate). The analytical sample, obtained as colorless thick platelets after three recrystallizations of a portion from ether--hexane, melted at 90-91 °. Anal (Spang) calc'd for C17 H20 0: C, 84. 96; H, 8. 39%. Found: C, 84. 95; H, 8. 49%. ir (CHC13 ) 1690 (C=O), 1593 (C=C); nmr (CDC13 ) o 1. 47 (s, 3, C-9 C!! 3 ), 1. 71 (s, 3, c:...7 C!i 3 ), 2. 33 (s, 2, -C~2 -CO-), 5.13 (s, 1, C=CH-CO-) . ....... After 0. 852 g (12. 2 g atoms) of lithium wire had been allowed to dissolve in ca 250 ml of liquid ammonia (dried by distillation from -61- sodium immediately preceding the reaction), a solution of 0. 600 g (2. 50 mmoles) of a, (3-unsaturated ketone ,,...,..,.... G-7 in 90 ml of dry ether (distilled from lithium aluminum hydride immediately preceding the reaction) was added dropwise. The reaction mixture was stirred for 1 hr. Solid ammonium chloride was then added to discharge the blue color, and the ammonia was allowed to evaporate. The resulting solid residue was partitioned between 200 ml of ether--benzene {4:1) and 40 ml of water. The organic layer was successively washed three times with 20 ml portions of water, two times with 10 ml portions of a saturated salt solution and dried (Na2 S04 ). Removal of the solid at reduced pressure afforded 0. 588 g of an off-white colored solid. Both hydroxylic and carbonyl absorptions were noted in their spectrum. An ice -cooled mixture of the ketone - -alcohol mixture in 15 ml of · acetone was consequently oxidized with 1. 5 ml of 8 N aqueous chromic acid solution 17 . After the mixture had been stirred in the cold for 10 min, isopropyl alcohol was added to destroy the excess oxidant, and then the mixture was diluted with 100 ml of water and extracted four times with 40 ml portions of ether--benzene (4:1). The combined ethereal layers were washed two times with 7 ml portions of 10% aqueous sodium bicarbonate, two times with 10 ml portions of water, two times with 10 ml quantities of saturated salt solution and then dried (N~S04 ). Removal of the solvent at reduced pressure yielded 0. 567 g of an off-white colored crystalline solid. Analysis of the vapor phase chromatogram of the crude oxidation product taken with an oven temperature of 264 ° indicated complete absence of the alcohol, 97% -62ketone F-2 exhibiting a retention time of 82 sec, and 3% unreacted ~ unsaturated ketone G-7 exhibiting a retention time of 91 sec. Sepa""""' ration of the mixture was achieved on three 2 mm silica gel plates (20 x 20 cm) on elution with 40% ether--petroleum ether. The 0. 526 g (90% yield) white crystalline solid extracted from the Rf 0. 56 bands was shown to be pure by gas chromatographic analysis. The analytical sample, obtained as thin platelets after two recrystallizations of a portion from hexane, melted at 99. 5-100°. (Thomas Hoover capillary silicone oil bath). The melting point of ketone ,,...,...,... D-6, obtained by degradation from diosphenol .,...,...,.... B-1 was 99-100° (Hoover), the mixture melting point (Hoover) exhibited no depression. Both the infrared and nmr spectra of both compounds are identical. The thin layer chromatogram of both these compounds on silicic acid, employing 40% ether- -petroleum ether as developer, exhibited the identical Rf value of 0. 56; the gas chromatographic peak enhancement study also indicated that these compounds are identical. The mixture melting point with the trans-fused ketone m mp 81. 5-82. 5 (Hoover), occurred between 79 and 93°. These facts indicate that the potassium..!,-butoxide rearrangement product has the structure E-1. ,,...,...,... Anal (Spang) calc'd for C17 H22 0: C, 84. 25; H, 9. 15%. Found: C, 84. 33; H, 9. 21%. ir (CHC13 ) 1738 cm -i (C=O); nmr (CDC13 ) o 1. 33, 1. 40 (two s, 6, C-7 and C-9 CH 3 ). "' ~: An ice-cooled mixture of 0. 980 g (3. 78 m.moles) of hydroxyketone ,,...,...,... C-2 in 100 ml of acetone was oxidized with 4 ml of -638 N aqueous chromic acid solution 17 After the mixture had been stirred at 0° in a nitrogen atmosphere for 45 min, isopropyl alcohol was added to destroy the excess oxidant. To the filtered reaction mixture was added 5 ml of saturated sodium bicarbonate solution and the resultant mixture was concentrated on the rotary evaporator at 40° at reduced pressure until 20 ml of liquid remained. The filtered chromic salts were twice washed with 250 ml portions of water and the aqueous washings were extracted three times with 50 ml portions of ether--benzene (4:1). The combined ethereal washings were added to the concentrated rotary evaporator mixture and the resultant system was washed three times with 15 ml portions of saturated salt solution and dried (N~S04 ). Removal of the solvent at reduced pressure yielded 0. 860 g of an off-white colored oily solid. Chroma- tography on silicic acid afforded 0. 545 g (56%) of diosphenol C-6 on """'"""" elution with 900 ml of 12% ether--petroleum ether. The analytical sample, obtained as platelets after two recrystallizations of a portion from ether--petroleum ether melted at 121-123°. The compound exhibited a purple color upon the addition of ferric chloride test solution. The mixture melting point with diosphenol E-1, obtained """'"""" from potassiumJ.-butoxide rearrangement (mp 162-164°), exhibited a depression - the mixture starting to melt at 105°. Separation of both dios phenols was obtained on the gas chromatograph with the oven temperature at 268°-cis-a-diosphenol C-6 eluting after 112. 5 sec and """"""' the cis-,B-diosphenol .!f_:l after 121. 5 sec. Anal (Spang) calc'd for C17 H20 0 2 : C, 79. 65; H, 7. 86%. Found: C, 79. 77; H, 7. 86%. -64- ir (CHC13 ) 3480, 3240 (OH), 1695, 1652 cm -l (enolic dione); uv max (MeOH) 259 mµ. (E 9, 620); nmr (CDC13 ) o 0. 93 (s, 3, C-9 C!! 3 ), 1. 20 (s, 3, C-7 C!! 3 ), 3.26 (d of d, 1, ~ = 3.4H -H Hz, 1 8 J = 1. 5 w coupling Hz, C-8 !!), 6. 68 (d, 1, J = 3. 4 Hz, vinyl!!). An off-white colored oily solid which weighed 0. 150 g and crystallized to an off-white colored solid upon being seeded with authentic diacid ---.. C-1 was isolated from the above aqueous layer. Its infrared spectrum exhibits peaks both at 3540-2270 (acid OH) and 1700 cm -l (C=O) which are characteristic of diacid C-1. The compound, obtained after """""" three recrystallizations of a portion from ether--petroleum ether, melte-d at 206-208 °. The mixture melting point with authentic diacid C-1, 207-209°, was 205-208° .. smaller scale chromic acid oxidations ~ worked up by extracting the chromic salts into water (as opposed to filtering off the chromic salts and concentrating the filtrate with sodium bicarbonate as performed above) yielded a yellow crystalline solid. ir (CHC13 ) 1758 (shoulder) and 1748 cm -l (carbonyls), hydroxylic absorption absent; nmr (CDC13 ) o 0. 89, 1. 09 (two s, 6, C-9 and C-7 C!! 3 ), 2. 52, 2. 60, 2. 61, 2. 72 (d of d, 2, C-1 C!!2 ). This compound did not impart a color to ferric chloride test solution 13 , but exhibited a purple color when the chloroform solution of the compound containing a drop of pyridine is heated on the steam bath for 3 0 sec before the addition of ferric chloride. A white solid is obtained after 0. 030 g of yellow crystalline solid, probably the non-enolized adiketone _...,.... C-6, is stirred at room temperature in 8 ml of acetone with 1 ml aqueous saturated sodium bicarbonate solution for 2 hr in a . -65- nitrogen atmosphere. The purified white solid, obtained after two recrystallizations of a portion from ether--petroleum ether, melted at 120-122°. The mixture melting point with authentic diosphenol U, whose melting point is 121-123 °, is 120-122°. 2indene-3-one (XVII): A stirred mixture of 0. 205 g (O. 798 mmole) of diosphenol C-6, 0. 75 ml of dimethyl sulfate 15 , 2. 5 g of anhydrous ~ ,,....,....,... potassium carbonate, and 50 ml of anhydrous acetone was heated under reflux in a nitrogen atmosphere for 14 hr. The mixture was concentrated to near dryness under reduced pressure and the residue, diluted with 120 ml of ether--petroleum ether (4:1), was washed three times with 15 ml portions of a saturat~d salt solution and then dried (Na2 S04 ). The solution, concentrated to dryness on the rotary evaporator, afforded 0. 217 g (100% yield) of a pale yellow oil which crystallized upon standing. The ether exhibited a single peak with retention time of 118 sec by vapor phase chromatography at 300°. The analytical sample, obtained as white platelets melting at 73-75 °, was prepared by one recrystallization of a portion of this material from ether--petroleum ether. Anal {Elek) calc'd for C18 H22 0 2 : C, 79. 96; H, 8. 20%. Found: C, 79. 87; H, 8. 13%. ir {CHC13 ) 1710 {C=O), 1628 {double bond linked to oxygen), 1258, 1080 cm -i {vinyl ether); uv max (MeOH) 256 mµ (E 8, 110); nmr (CDC13 ) 6 0. 83 (s, 3, C-9 C~ 3 ), 1. 09 (s, 3, C-7 C!.! 3 ), 3.16 (d, 1, J = 3 Hz, C-8 !,!), 3. 68 (s, 3, OC~ 3 ), 6. 36 {d, 1, J = 3 Hz, vinyl!!). -66- ~: A solution of 0. 800 g (3. 36 mmoles) of unsaturated ketone G-7, 1. 16 g (10. 4 mmoles) of selenium dioxide and 0. 188 ml (10. 4 """"" mmoles) of water in 40 ml of glacial acetic acid was heated under reflux for 4 hr. Filtration of the cooled reaction mixture and removal of the acetic acid from the filtrate at reduced pressure left an orangecolored oil. The oily residue, diluted with 200 ml of ether--benzene (4:1) was washed five times with 20 ml portion of water, three times with 20 ml portion of 5% aqueous sodium bicarbonate, three times with 15 ml portions of saturated salt solution and then dried (N~S04 ). Removal of the solvent at reduced pressure afforded 0. 843 g (100% yield) of a yellow crystalline solid - exhibiting a single spot with an Rf value of 0. 46 on a silicic acid thin layer chromatogram upon being developed with 35% ether--petroleum ether. Diketone XVIII decom~ poses in the vapor phase chromatograph when the oven temperature is 274°. Two successive crystallizations of a portion from ether-- petroleum ether afforded analytically pure thin yellow platelets which melted at 111. 5-112. 5°. Anal (Spang) calc'd for C17H18 0 2 : C, 80. 25; Found: C, 80. 38; H, 7. 25%. ir (CHC13 ) 1765 (C=O) 1717 H, 7. 13%. (conjugated C=O), 1622 cm -i (C=C); uv max (MeOH) 278 mµ (E 5, 720); nmr (CDC1 3 ) o 1. 53, 1. 80 (twos, 6, C-9 and C-7 C!! 3 ), 5. 97 (s, 1, C=CH-CO) . .,.... 7 ~: An ice-water cooled solution of 0. 889 g (3. 49 mmoles) of diketone XVIII in 25 ml of methanol was treated with 0. 033 g ~ -67- (0. 873 mmole) of sodium borohydride in 4 ml of water and the mixture was stirred at room temperature in a nitrogen atmosphere for 3. 5 hr. The reaction mixture was diluted with 150 ml of water and extracted with three 50 ml portions of ether--benzene (4:1). The combined organic extracts were washed with three 15 ml portions of a saturated salt solution, dried (Na2 S04 ) and evaporated to dryness under reduced pressure. The 0. 887 g pale yellow solid residue, eluted from 50 g of silicic acid (60-200 mesh) with 1300 ml of 50% ether--petroleum ether, afforded 0. 823 g (90% yield) of a white crystalline solid exhibiting a single spot whose Rf value is O. 82 on silicic acid thin layer chromatography when developed with ether. The hydroxyketone F-1 decomposes in the gas chromatographic column when the oven is ,,....,,...,.., maintained at 246°. The analytical sample, obtained as fine thin platelets after one crystallization of a portion from ether--petroleum ether, melted at 116. 5-118. 5°. Anal (Spang) calc'd for C17 H20 0 2 : C, 79. 65; H, 7. 86%. Found: C, 79. 69; H, 7. 87%. ir (CHC13 ) 3 514, 3370 (OH), 1710 (C=O), 1585 cm -i (C=C); uv max (MeOH) 235 mµ. (E 10, 790); nmr (CDC13 ) o 1. 37, 1. 75 (two s, 6, C-9 and C-7 C~ 3 ), 3.16 (.s, 1, O~), 4. 04 (d, 1, ;r = 1 Hz, C-3 !_I), 5.29 (s, 1, vinyl~). 8~: A solution of 0. 618 g (2. 81 mmoles) of _E- bromobenzoyl chloride in 20 ml of dry pyridine was added to 0. 600 g (2. 31 mmoles) of hydroxyketone C-3 dissolved in 30 ml of dry """"' pyridine - the system being maintained at 0° during addition, and then -68- stirred at room temperature for 18 hr in a nitrogen atmosphere. A nitrogen stream and reduced pressure were employed to remove the major portion of pyridine. The residue, taken up in 200 ml of ether-- petroleum ether (4:1), was successively washed three times with 10 ml portions of 5 N sulfuric acid, three times with 10 ml portions of water, twice with 10 ml portions of saturated salt solution and then dried (N~S04 ). Removal of the solvent at reduced pressure yielded 1. 02 g of an off-white colored solid. The crude product, placed on four 2 mm silicic acid plates (20 x 20 cm) and developed with 45% ether--petroleum ether, yielded 0. 712 g (70% yield) of a white crystalline solid exhibiting a single peak with retention time of 340 sec by vapor phase chromatographic analysis at 303° which was extracted from the major band, Rf 0. 60. The analytical sample, obtained as tiny thin plates after one recrystallization of a portion from ether-methanol and a second recrystallization from acetone - -petroleum ether, melted at 177.5-179°. Anal (Spang) calc'dfor C24H23 Br03 : C, 65. 61; H, 5. 28; Br, 18. 19%. Found: C, 65. 66; H, 5. 21; Br, 18. 07%. ir (CHC13 ) 1720-1715 (C=O), 1592 (aromatic ring C=C), 1270 cm -i (ester); uv max (MeOH) 246 mµ (E 31, 100); nmr (CDC13 ) o 0. 62 (s, 3, C-9 C~ 3 ), 1. 44 (s, 3, C-7 C~ 3 ), 5. 37 and 6. 39 (twos, 2, C-3 ~and vinyl~), 7. 75 (AzEi, 4, J = 23. 5 Hz, J = 8. 5 Hz, E.-bromobenzoate ring). -bromobenzoyl)-7a-phenylhexahydro-8{3H1 2-indene-3-one (XXI : A solution of 1. 26 g (4. 50 mmoles) of .E_bromobenzoyl chloride in 40 ml of dry pyridine was added to 0. 400 g -69- (1. 56 mmoles) of diosphenol ,,...,.....,.. E-1 dissolved in 20 ml of dry pyridine - the system being maintained at 0° during addition and then stirred at room temperature for 15 hr in a nitrogen atmosphere. A nitrogen stream and reduced pressure were employed to remove the major - portion of pyridine. The residue, taken up in ether--benzene (4:1), was successively washed with 5 ·N sulfuric acid, water, saturated salt solution and then dried (Na2 S04 ). Removal of the solvent at reduced pressure yielded 0. 8 50 g of a yellow oily crystalline solid. The crude product, eluted from 50 g of silicic acid (60-200 mesh) with .600 ml of 15% ether--petroleum ether yielded 0. 533 g of a white oily solid shown · by vapor phase chromatography at 305° to be composed of 90% of the desired bromobenzoate ,,...,.....,.. XXI (with a retention time of 145 sec) and a single impurity present in 10% yield with a retention time of 3 sec, and 0. 189 g of pure white crystalline ,,...,.....,.. XXI. Washing the 0. 533 g oily white solid with cold petroleum ether resulted in the isolation of 0. 457 g [yielding 0. 646 g (94% yield) of combined pure solid] of additional bromobenzoate XX!. ~ The analytical sample, obtained as needles after one recrystallization of a portion from acetone - -petroleum ether, melted at 139. 5-140. 5°. Anal (Spang) calc'd for C24H23 Br03 : C, 65. 61; H, 5. 28; Br, 18.19%. 18. 25%. Found: C, 65. 69; H, 5. 35; Br, ir (CHC13 ) 1743, 1723 (C=O), 1643 (enol ester C=C), 1590 (aromatic ring C=C), 1262 cm -i (ester); uv max (MeOH) 238 mµ (E 34, 850); nmr (CDC13 ) o 1. 42, 1. 53 (two s, 6, C-9 and C-7 CH 3 ), "' 3. 02 (d, 1, J = 2. 4 Hz, C-8 H), 6. 48 (d, 1, J = 2. 4 Hz, vinyl H), 7. 75 - (~B2 , "" - 4, J = 21.5 Hz,~= 8.8 Hz, _Q-bromobenzoate ring). ""· -70- 7{3, 9{3-Dimethyl-3a-(e_-bromobenzoyl)-7a-phenylhexahydro-1, 8indene-2-one (XX): A solution of 0. 230 g (1. 05 mmoles) of E_-bromo~ benzoyl chloride in 5 ml of dry pyridine was added to 0. 150 g (0. 585 mmole) of hydroxyketone F-1 dissolved in 5 ml of dry pyridine - the ~ system being maintained at 0° during addition and then stirred at room temperature for 17 hr in a nitrogen atmosphere. A nitrogen stream and reduced pressure were employed to remove the major portion of pyridine. The residue, taken up in ether--benzene (4:1), was successively washed with 5 N sulfuric acid, water, saturated salt solution and dried (N~S04 ). Removal of the sol vent at reduced pressure yielded 0. 237 g of an off-white colored solid. The crude product, placed on two 2 mm silicic ac_id plates (20 x 20 cm) and developed with 30% ether--petroleum ether yielded 0. 166 g (67%) of a white crystalline solid. The analytical sample, obtained as a needle matte after two recrystallizations of a portion from ether--methanol, melted at 107. 5-108. 5°. The sample was heated at 143° at 0. 03 mm pressure for 1. 5 hr before the melting point was taken because the bromobenzoate XX occludes ether upon crystallization. Anal (Spang) """"" calc'd for C24H23 Br03 : C, 65. 61; H, 5. 28; Br, 18. 19%. Found: C, 65. 81; H, 5. 33; Br, 18.15%. ir (CHC13 ) 1725-1715 (C=O), 1592 (aromatic ring C=C), 1270 cm -i (ester); uv max (MeOH) 246 mµ (E 29, 620); nmr (CDC13 ) o 1. 41, 1. 77 (two s, 6, C-9 and C-7 C!!3 ), 5. 36, 5. 47 (twos, 2, C-3 !! and vinyl~), 7. 81 (~fl;i, 4, ~ = 23 .4 Hz, ~ = 8. 7 Hz, E_-bromobenzoate ring). -712a-Carbo carbox lie acid anh dride (H-1): To a stirred solution of 0. 942 g (3. 66 mmoles) diosphenol ,,..,,....,... E-1 in 40 ml of methyl alcohol was separately added 1. 5 ml of 10% aqueous sodium hydroxide solution and 4 ml of 30% hydrogen peroxide. The solution was heated at reflux for 1 hr, cooled, and 1. 5 ml of 10% sodium hydroxide and 4 ml of 30% hydrogen peroxide were again added. The solution was heated at reflux for 1 hr, cooled and an additional 4 ml of 30% hydrogen peroxide was again added. The solution was heated for an additional 1 hr. To the cooled solution was added 100 ml of 4:1 ether--benzene. The ethereal solution was twice extracted with 50 ml portions of 3% aqueous sodium hydroxide solution. The combined aqueous layers were.. acidified with iced concentrated hydrochloric acid, and then thrice extracted with 30 ml portions of 4:1 ether--benzene. The combined ethereal layers were washed four times with 10 ml portions of saturated salt solution, and then dried (Na2 S04 ). Removal of the solvent at reduced pressure afforded 1. 10 g of a pale yellow crystalline solid. ir (CHC1 3 ) 35302326 (acid OH), 1775 (acid C=O monomer), 1725-1700 cm -i (acid C=O dimer). Gas chromatographic analysis of the crude product at 300° indicated the presence of a single peak with a retention time of 106 sec accounting for 94% of the material. The crude product was combined with 30 ml of acetic anhydride and refluxed under nitrogen for 2 hr. The solution was concentrated under reduced pressure to 1. 030 g of an orange oily residue which readily crystallized upon being triturated with ether. The crude product, recrystallized from -72- ether--petroleum ether, afforded 0. 337 g of a white crystalline product, mp 122-124 ° ._ The concentrated mother liquor afforded 0. 338 g of a white crystalline solid, mp 121-124 °. Gas chromatographic analysis of the combined crystalline solids at 300° indicated the presence of a· single peak with a retention time of 66 sec accounting for 99% of the material. Gas chromatographic analysis of the mother liquor, a pale yellow oily solid, at 300° indicated that this mixture contained 70% anhydride ,,...,..,.. H-1 (combined yield of 92%). A single crystallization of a portion of the solid material from ether - -petroleum ether afforded a white crystalline solid melting at 124-125° (reported14 mp 12-4-125°). Anal (Spang) calc'd for C17H20 0 3 : C, 74. 97; H, 7. 40%. Found: C, 75. 01; H, 7. 26%. ir .(CHC13 ) 1804, 1759 cm -i (anhydride C=O); nmr 14 (CDC1 3 ) o 1. 42, 1. 53 (two s, 6, C-1 and C-3 C~ 3 ). §_yQ-9-(1, 5-Dimethyl-4-oxo-2, 3-benzobicyclo [ 3. 3. 1] nonane)acetic acid (I-2): To 38. 4 ml of 85% phosphoric acid under nitrogen . ~ was added 48 g of phosphorous pentoxide. After stirring for 1 hr, 0. 320 g of anhydride ,,,..,....,.._, H-1 was added and the mixture was heated with stirring at 90° for 1 hr. While still warm, the brown colored reaction mixture was poured over approximately 100 g of crushed ice, and the precipitate was extracted into ether (upon being thrice washed with 50 ml portions of 4:1 ether--benzene). The combined ethereal layer was twice washed with 15 ml portions of water, washed four times with a saturated salt solution, and then dried (N~S04 ). Removal of the solvent at reduced pressure afforded 0. 310 g of a yellow solid. Keto acid ,,...,....,.,.. I-2 did not elute from the SE 30 vpc column. Crystallization -73of the crude product from ethyl acetate--hexane afforded 0. 275 g (86% yield) of a white crystalline solid, mp 171-174°. The analytical sample, obtained as white platelets melting at 172-173. 5°, was prepared by two further crystallizations from ethyl acetate--hexane. Anal (Elek) calc'd for C17 H20 0 3 : C, 74. 97; H, 7. 40%. Found: C, 74. 91; H, 7. 28%. ir (CHC13 ) 3560-2224 (acid OH), 1710 (acid C=O), 1675 (ketone C=O), 1599 cm -i (C=C); uv max (MeOH) 253 mµ (E 10, 750), 292 mµ (E 1, 620); nmr (CDC1 3 ) o 1. 23 (s, 3, C-5 C!!3 ), 1. 49 (s, 3, C-1 C!! 3 ), 2. 25 (d, 2, J = 1. 5 Hz, -C!!2 -C02 H), 10. 65 H) . (s, 1, C02 ....... s n-9-(1, 5-Dimethyl-2, 3-benzobicyclo [ 3. 3. 1] nonane)acetic ~: To a suspension of 0.1gof10% palladium on carbon in 5 ml of glacial acetic acid was added a solution of 0. 197 g (O. 722 mmole) of keto acid l-2 and 4 drops of 60% aqueous perchloric acid """""'" in 10 ml of glacial acetic acid, and the mixture was stirred in a hydrogen atmosphere until the uptake of hydrogen gas ceased (2 hr). During this period 36. 5 ml of hydrogen was absorbed. The catalyst was removed by filtration and washed with 50 ml of benzene. The filtrate was diluted with 100 ml of water, and the benzene layer was separated. The aqueous layer was thrice extracted with 30 ml portions of 1 :1 ether--benzene. The combined organic layers were thrice washed with 15 ml portions of water, thrice washed with 15 ml portions of a saturated brine solution, and then dried (N~S04 ). Removal of the solvent at reduced pressure afforded 0. 189 g (100% yield) of an off-white colored crystalline solid, mp 159-163°. The -74- analytical sample, obtained as white platelets melting at 162-164°, was prepared by two further crystallizations from ethyl acetate-hexane. Anal (Elek) calc'd for C17 H22 0 2 : C, 79. 03; H, 8. 58%. Found: C, 78. 94; H, 8. 48%. ir (CHC13 ) 3540-2320 (acid OH), 1710 cm -i (C=O); nmr (CDC13 ) 6 1. 01 (s, 3, C-5 C!! 3 ), 1. 39 (s, 3, C-1 C!!3 ), 2 .. 22 (m, 2, -C!!2 -C02 H), 2. 56 (s, 2, w coupling causes pattern to look liked of d, Jw coupling= 0.6 Hz, benzylic C~), 11.23(s,1, C02 1I). I ~-9-(1, 5-Dimethyl-2, 3-benzobicyclo[ 3. 3. 1] nonane)-2-ethyl alcohol (J-1): A solution of 0. 300 g (1. 16 mmoles) of acid .,...,....,... I-3 in 20 ml ~ of tetrahydrofuran (freshly distilled from lithium aluminum hydride) was dropwise added to 0. 228 g (6. 00 mmoles) of lithium aluminum hydride in 80 ml of tetrahydrofuran (dried as above). The stirred mixture was heated at reflux in a nitrogen atmosphere for 3 hr. To the cooled mixture was cautiously added 5 ml of 10% aqueous sodium hydroxide, and the resultant granular sodium aluminate salts were separated by filtration. The filtrate was concentrated at reduced pressure to near dryness. The residue was taken up in 100 ml of 1 :1 ether--benzene, twice washed with 10 ml portions of water, thrice washed with 10 ml portions of a saturated salt solution, and then dried (Na2S04 ). Removal of the solvent at reduced pressure afforded 0. 260 g of a yellow oil. Gas chromatographic analysis of the crude product at 260° indicated the presence of a single peak with a retention time of 58 sec accounting for 90% of the material. A 0. 060 g portion of the crude product, placed on a 2 mm thick silicic acid plate (20 x 20 cm) and developed with 50% ether--petroleum ether, afforded 0. 055 g -75- (83% yield) of a clear colorless oil which was extracted from a band centered at Rf 0. 65. Gas chromatographic analysis indicated that alcohol J -1 was pure. ~ The analytical sample, obtained as a clear colorless oil by evaporative distillation over a 10 cm path length, - exhibited bp 100-102° (0. 20 mm). Anal (Elek) calc'd for C17H24 0: C, 83. 55; H, 9. 90%. Found: C, 83. 64%; H, 9. 98%. ir (CHC13 ) 3606 and 3420 cm - 1 (shoulder) (OH); nmr (CDC13 ) o 0. 99 (s, 3, C-5 C!t3 ), 1. 18 (" s, 1, O~), 1. 39 (s, 3, C-1 C~), 3. 60 (t, 2, J = 8 Hz, -C!t2 0H). s n-9-(1, 5-Dimethyl-2, 3-benzobicyclo [ 3. 3. 1] nonane)~: To 0. 207 g (90% pure, 0. 760 mmole) of alcohol J-1 dissolved in 24 ml of methylene chloride was added 1. 31 g (5. 06 .,...,,...,... mmoles) of solid chromic oxide--dipyridine complex 35 . The dark brown colored mixture was stirred at room temperature for 15 min, and then filtered through 25 g of Merck acid-washed alumina with 100 ml of methylene chloride. The clear colorless eluent was concen- trated at reduced pressure to almost dryness and taken up in 150 ml of 1 :1 ether--benzene. The ethereal solution was successively washed two times with 5 ml portions of 1 N hydrochloric acid, two times with 10 ml portions of water, 3 times with 10 ml portions of a saturated salt solution and then dried (N~S04 ). Removal of the solvent at reduced pressure afforded 0. 197 g of a yellow oil. Gas chromatographic analysis of the crude product at 270° indicated the presence of a single peak with a retention time of 41 sec accounting for ~8% of the material (100% yield). ir (CHC13 ) 1722 cm -l (C=O); nmr (CDC13 ) -76- o 0. 94 (s, 3, C-5 C!! 3 ), 1. 32 (s, 3, C-1 C!! 3 ), 9. 83 (t, 1, J = 1 Hz, -CHO). The aldehyde was converted into its semicarbazone derivative ,,.... by the procedure of Fieser 41 . An analytical sample of the semicarbazone prepared by two crystallizations from methanol--ether melted at 159-160°. Anal (Elek) calc'd for C18 H25 N3 0: C, 72. 21; H, 8. 42; N, 14. 03%. Found: C, 72. 05; H, 8. 54; N, 13. 58%. ir (CHC1 3 ) 3530, 3474, 3402, 3350 (NH), 1690, 1635, 1567 cm -i (amide bands). ~-9-Ethyl-l, 5-dimethyl-2, 3-benzobicyclo[ 3. 3.1 ]nonane ~ To 0. 090 g (88% pure, 0. 326 mmole) of aldehyde .:[:1 was added 0. 224 g (3. 40 mmoles) of 85% potassium hydroxide dissolved in 10 ml of triethylene glycol, and 0. 192 g (5. 70 m.moles) of 95% hydrazine was added to the resultant clear pale yellow solution. The stirred solution was first heated at 105° for 2. 5 hr in a nitrogen atmosphere, and then at 205° for 4 hr. A Dean-Stark trap was employed to remove the excess hydrazine and water. The solution was cooled and 100 ml of 4:1 petroleum ether--ether was added. The ethereal mixture was thrice extracted with 50 ml portions of water. The combined aqueous layers were back-extracted with three 30 ml portions of 1 :1 ether-benzene. The combined ethereal layers were washed with three 10 ml portions of water, one 10 ml portion of a saturated salt solution, and dried (N~S04 ). Removal of the solvent at reduced pressure afforded 0. 085 g of a yellow oil. Gas chromatographic analysis of the crude product at 280° indicated the presence of a single peak with a retention time of 26 sec accounting for 80% of the material. The crude product -77- was purified by preparative thin layer chromatography on a 2 mm silicic acid plate (20 x 20 cm) by elution with 10% benzene--petroleum ether. The 0. 058 g (94% yield) clear colorless oil, which exhibited a single peak during vapor phase chromatography, was isolated from a band whose Rf was 0. 69. Evaporative distillation (over a path length of 15 cm) afforded a clear colorless oil of analytical sample purity, bp 55-57° (0. 06 mm). Anal (Elek) calc'd for C17 H24 : C, 89. 41; H, 10. 59%. Found: C, 89. 37; H, 10. 68%. ir (CHC13 ) 1373 cm -i (singlet, CH3 ); nmr60 Mc (CDCls) o 1. 00 (s +smaller spikes, 6, two CB3 ), 1.40 (s, 3, C~), 2. 63 (d of d, 2, J = 26 Hz, ~ = 18 Hz, · benzylic C~); nmr 100 Mc (CDCis) o 2.63 (d of d, 2, J = 36 Hz, J = 18 Hz, benzylic C~). -78- REFERENCES 1. (~) 2. W. P. Campbell and D. Todd, J. Am. Chem. Soc., 64, 928 """ (1942). 3. J. L. Simonsen and D. H. R. Barton, "The Terpenes", Volume ill, Cambridge University Press, Cambridge, .1952. 4. P. DeMayo, "The Higher Terpenoids", Interscience Publishers, Ltd., London, 1959. 5. A. R. Pindar, "The Chemistry of the Terpenes", Chapman and Hall, Ltd. , London, 1959. · 6. N. A. J. Rogers and J. A. Barltrop, Quart. Rev., (London), 16, 117 (1962). A. C. Oudemans jun.'· Ann., 170, 213 (1873). ( ) A. C. Oudemans jun., Ber., §;'1122 (1873). """" 7. F. F. Giarrusso, Thesis, University of Michigan, 1966. 8. R. M. Carman, H. C. Deeth, R. A. Marty, K. Mori, and M. Matsui, Tetrahedron Lett., 3359 (1968). 9. T. A. Spencer, R. J. Friary, W. W. Schmiegel, J. F. Simeone, and D. S. Watt, J. Org. Chem., ~' 719 (1968). 10. T. A. Spencer, T. D. Weaver, R. M. Villarica, R. J. Friary, J. Posler, and M. A. Schwartz, J. Org. Chem., ~' 712 (1968). 11. K. Yoshida and T. Kubota, Tetrahedron, 21, 759 (1965). 12. The observation that the vinylic proton is found at somewhat higher field than the corresponding resonance in simple a, {3unsaturated ketones was of great utility in the structural elucidation of the diosphenol-containing cucurbitacins. (a) C. R. Noller, A. Melera, M. Gut, J. N. Shoolery, and L. F. Johnson, Tetrahedron Lett., 15 (1960). (b) D. Lavie, Y. Shvo, 0. R. Gottlieb, and E. Glatter, J. Org. Chem., 27, 4546 (1962). """" '('"" 13. R. L. Shriner, R. C. Fuson, and D. Y. Curtin, "The Systematic Identification of Organic Compounds", Fourth Edition, John Wiley and Sons, Inc., New York, 1965, p. 113. 14. F. F. Giarrusso, unpublished results. -79- . 15. E. M. Bickoff, R. L. Lyman, A. L. Livington, and A. N. Booth, J. Am. Chem. Soc., 80, 3969 (1958). "'""' 16. J. H. Chapman, J. Elks, G. H. Phillipps, and L. J. Wyman, J. Chem. Soc., 4344 (1956). 17. A. Bowers, T. G. Halsall, E. R. H. Jones, and A. J. Lemin, J. Chem. Soc., 2548 (1953). 18. R. E. Ireland and R. C. Kierstead, J. Org. Chem., 31, 2543 (1966). """'" 19. W. S. Johnson, D. K. Banarjee, W. P. Schneider, C. D. Gutsche, W. E. Shelberg, and L. J. Chinn, J. Am. Chem. Soc., 74, 2832 (1952). "'""' 20. L. Velluz, J. Valls~ and G. Nomine, Angew. Chem. Intern. Ed. Engl. , 1, 181 (1965J. 21. L. Claisen, Ber., 45, 3157 (1912). 22. A. W. Burgstahler and I. C. Nordin, J. Am. Chem. Soc., ,,_ 83, 198 (1961}. . 23. R. Pappo, D. S. Allen, Jr., R. U. Lemieux, and W. S. Johnson, J. Org. Chem., ll' 478 (1956). 24. H. 0. House, B. A. Tefertiller, and H. 0. Olmstead, J. Org. Chem., 91_, 935 (1968). 25. G. Stork and S. D. Darling, J. Am. Chem. Soc., 86, 1761 """' (1964). 26. The data for the .E_-bromobenzoate derivative of hydroxyketone C -3 is reported in Appendix I. The remaining data is being prepared for publication by Dr. J. Bordner of the California Institute of Technology. 27. C. T. Mathew, G. Sen Gupta, and P. C. Datta, Proc. Chem. Soc., 336 (1964). 28. E. Kyburz, B. Riniker, H. R. Schenk, H. Reusser, and 0. Jeger, Helv. Chim. Acta, 36, 1891 (1953). -- "'""' "'""' 29. (a) R. Scholl and K. Meyer, Ber., 65, 902 (1932). ,,..,,.,. (b) A. Koebner and R. Robinson, J. Chem. Soc., 1994 (1938). -80- 30. (a) N. N. Sana, B. K. Ganguly, and P. C. Dutta, J. Am. Chem. Soc., 81, 3670 (1959). (b) S. N. Mahapatra and R. M. Dodson, Chem. Ind. (London), 253 (1963). 31. Varian Associates, NMR Spectral Catalog, Volume I, Palo Alto, California, 1962. 32. U. R. Ghatak, D. K. Datta, and S. C. Ray, J. Am. Chem. Soc., 82, 1728 (1960). """' 33. Obtained from Timber Processing Company Limited, Auckland, New Zealand. 34. L. F. Fieser and M. Fieser, "Reagents for Organic Synthesis", John Wiley and Sons, Inc., New York, 1967, p. 585. 35. J. C. Collins, W. W. Hess, and F. J. Frank, Tetrahedron Lett., 3363 (1968). 36. Huang-Minlon, J. Am. Chem. Soc., ,,...,... 68, 2487 (1946). 37. E. Wenkert, V. I. Stenberg, and P. Beak, J. Am. Chem. Soc., 83, 2320 (1961). """"" 38. K. Nakanishi, "Infrared Absorption Spectroscopy--Practical", Holden-Day, Inc., San Francisco, 1962, p. 22. 39. Sample obtained from Professor R. E. Ireland. 40. The experiments were performed by I. Thomas, H. P. Gribi and F. F. Giarrusso in these laboratories. 41. L. F. Fieser, "Experiments in Organic Chemistry", Third Edition, D. C. Heath and Co., Boston, 1955, p. 85. 42. Sample obtained from Dr. F. F. Giarrusso. -81- APPENDIX Data Employed in the Determination of the Structure of 7f3, 9a-Dimethyl-3a-(£-bromobenzoyl)-7 aphenylhexahydro-1, 8 -indene -2-one ~ by X-ray Analysis Crystals suitable for an x-ray analysis were grown utilizing the slow evaporative technique employing a mixture of ether and methyl alcohol. The resulting crystals were surveyed on a precession 0 . camera. The unit cell dimensions (A) were: a= 24. 80, b = 12. 62, 0 c = 13. 09. The corresponding bond angles ( ) were: a= 90. 00, b = 93. 64, y = 90. 00. The systematic extinctions were: hkl, h+k odd; hol, 1 odd. The space group was c 2;c· There are 8 molecules in the unit cell. The calculated density is 1. 426 g/cm3 , and the observed 3 density is 1. 43 g/cm • The number of reflections recorded were 2146, of which 2071 were non·-zero reflections. One angstrom intensity data was collected for the compound on the General Electric-Datex diffractometer using nickel filtered copper radiation. A e -2(} scan technique was employed, background was counted for 10 sec at each end of the scan, and the scan rate was 2° per min in 28. A single check reflection was monitored every 30 reflections. The check reflection indicated no crystal damage and was reproducible well within counting statistics. The diffractometer output was processed using subprograms of -82- the CRYRM crystallographic computer system 1. The processing included corrections for background and for Lorentz and polarization 2 effects. It also included calculation of the F value and its standard deviation for each of the reflections. The standard deviations were assigned on the basis of the following equation: a2(I) = S + (B1 +~)cl + (dS) 2 Where Sis the scan count, B1 and~ are the background counts, dis an empirical constant equal to 0. 02, and a = n/2 mt where n = scan range, m = scanning speed, and t =time for background count in seconds. Finally, the data were placed on an absolute scale by means of Wilson statistics. Structure Determination and Refinements The trial structure was derived by the usual Patterson and Fourier techniques in three dimensions. Full matrix least-squares refinement of coordinates, isotropic temperature factors, and scale factor reduced the R -index to 11. 5% after the refinement. A difference Fourier indicated no misplaced or missing Br, C or 0 atoms. The difference Fourier was also utilized to locate the hydrogen atoms. The addition of the hydrogen atoms to the structure factor calculation and the application of anisotropic temperature factors to the refinement reduced the R-index to its final value, 7. 2%. The observed and calculated structure factors are listed in Table I. The final refined atomic parameters and their standard deviations are presented in -83- Table II. The bond lengths and angles of the structure are given in Tables III and IV. The standard deviations in the coordinates, derived from the residuals and the diagonalized elements of the inverse matrix of the final least-squares cycle, correspond to positional uncertainties 0 of about 0. 0009A for the C and 0 atoms. The uncertainties for the 0 Br, C, and 0 bonds are about 0. 007 A. Uncertainties in the bond 0 angles involving the C, O, and Br atoms are approximately 0. 5 . Since the hydrogen positions were not refined, no error estimates will be made for distances and angles involving these atoms. The coordi- nates of the hydrogen atoms are listed in Table V. The three structures obtained were stereographically plotted using the ORTEP computer progr~m of C. K. Johnson 2. H42 XIX '""""" Table I The Observed and Calculated Structure Factors. Within each group are values of 1, 10F0 , lOFc. Reflections indicated with asterisk were assigned zero weights in the final least squares cycle . ...... -!!!!~~:;;~~~~ ...... ....... - • .,. ..... o.!.! ,:.;_; :r ... :;;:;; ........ c:...!J.!o;;:';; ; ;';:::;;;,.,,. ..... c .!. ! ! ;_ ~;:'; ;::;;::: i;:;; ;;;:;:; ..... ..... ~:.. ! .:. !~:::;; ~~ :::::;;:;';;::; .. .,. .... o:...!! O:,~;:;:;;:;;::;:::i;: ~::::~;;;;:'; ~O'"'.,. ...... o.!. ! .:. ! :;;::;; ;~:::~ ..... o ......... o,. ... ,,. ... o % :; :! ;~~~~~: ~ff:~;~~~ E~i~e!: :~ ~=~;~~~g~E~~~ ~ ~ ~~ ~: ;~ l ~~ ~ l ~:h~~~m~:e~~: rn" ~?g!e ~gH£s;~~ ~~m: :: <:!; ~ ~~~§~~ fi~:~~ "iim :~: =~ h ~ ~~ ~ !o~~ l: ~ +I .:..!. ,:. !. ~:: ;:; .; ; : ;;:::;:: ............. -! !.'. !. ~ ~;;; : ; ~ ::: ; :: :; : :: "' ...... "' - ! ! ,:. !. ! : ; ~ ~; ~ :: ; ::; :; ::;; ;:: ..... "' ...... !! .!. !, ! : : ~:;.; ~ ~ ::::::;:::;: ;;: .......... -!.!..'.!.!::;:~~~ :::~:;::;:;;;: .... .,. . . _,!. .!. ,:. ! ~ ::~.;;::~ ::~:;::;:::: ...... .'.!. ~:;;:;::;; ~~ ....... .. .... - - =:~;?E~:!Ei ~:: :: l :z:;: :g;: ~~§ f ?~~! :::: E~! ::::::~~§:.!!: l §~;~~!:: ! ~:~~:~:::~:!!:::~~~:: :. . :~~ !;" "~ ;:g ~ :~ m ! ~~;:: 'm: :::: g::~:: ~ ~ ~!~ E:;m~ ;: ;; :: ~:: L~;~~!!~E :! : :z::~.:: !;_ :.J~~::;~~~~z~ · ::!~~ g~!f ~~~~~~~~~~~~ ... : ~ ~ f ~: i;;~~~~~~ ~ ~~: ~li .. ~§! E~~~~~;~;§:~~E~ ~ ~~ :~"' S!.~ .. ! ~; ~~=~~~~;~~i ~ !~~i~. ! :i; ! ~;~~ ~~~ ~~~~~~~z~ 1i~ . . ~~~ ~ !~~ ~ §; ;~!~ ! ; ~ ~§!~~ ~~" :~~;;:;E:::!~!~- !~:~~::E~l!;:~ . . .,..,_,!.,!. .. w ... ~~:;;=~:;.~~:i ;;*~ :::~:;;::;o ..... o.!:.:.~5~;;;~i;Z :::~:;:#;;o .... ,.,.,,!n. .:.~~.;;;~6~ ~:::~;:;;;;o ........ o.!l.l..!~~;;;*6~ ~:::~;:;;::;c;.,u .. ., .!.! ............ .!.!!!!~ : .............. .!.!!!!=: ::: .... .,. ....... .!.!!!!==~~ :;:: .............. .!.!.!!,!:'::;~ ::;::~ ... =~::: ::!::e~~~~§§:i~~~~:;:!~~: ~:!~!e:!~~~~~:~~:~~~!~!~~= :!~~s::§~~~;~~~~E~~~!~:~~= ~~~~E::~::i:g: !3: ~~t;~~~=:~: ~~;;~;!~~~~~::: ~t:!~~::!g~::!~::: :;~;~~~:~!~~~~::~: !!;!~~ :i:~. :~~~~~~~~;~~~~ f ~ :~i~~! ~f~ §: :ii~ii ~g~ ~~~~~:: ~= !~~: ~:~~ ~~E ~~~~ ~~=:~§g5 i~ ~; ~r· ~ ;;.~i~=~~~~~~ .. s~ = ~~!! ~~!~!!:;; .! ~: t: ~~~:eSE~ = : t~~: !~~:~~~~~±~:; i ~~ !~:; ~g§i5 ~ i~;:~~., ~~ 1~ i ~ r· •::;;; :. . . . . . .:i ==~:: ~ ! .:. J..:. ! ~; ~ ~~ ~ :: :; :; :;; ::: =•..... ..,_!.:..:. .!. ! : =~ 6; :!! :: • :;:; ::: : ...... "'_!.:. J. ! ! : : ~ ::; :: ~ :~!::::: ::::::m~m:~;:~:;:~~= [!~::zm~mm;~~;::~: [!::;:!~~!~~~:!~==~~£~;~: ~=~~:§ .. ~!~~=~; i~~~! ~ §~~~i~~ ~.. ! ~~ ~~~~~~~~~;~!~;~ i:::~ . . ~~~1~~:~ ii~~! i~~~ ~~! ~;:" ~H ••••• ~H~i;~ :;; ••••• ~ ;.!;t; ;:::o .. I0- ..... 0.:.!.:.!.;~~; ;-;=:~ ............ .:.!.:.!.;~~;; ;;;;:::o .. .,. .... .,.:.!.:.!;:::;;:;;;:: ;;-;:::;:; ........ .:.!.:.!;;::::;;;~ :<i;;::::;:; .. ,,. :::: :?~~=:~:::;: ~=:~~~!: =~~E~: : i~ !:~=~~:~ :~e:E: E: l :'!~~~:~~: ;:~:~ l:: l :~: ~!~~::::~~~!E; :::~: =~~~:~=~~;e~~~!! ::!~: ~~=~~;~~E : i.:.:: ~:~;~: :~ ~ ~~ = E~::~~:~~~~; !~t:; : !i5:!~E~i~~~~~E. E~i :~!i:~~:::: S~:~~ .. ~~ei~~ :i ~~~ S~~ !i~~:~ .. : : ~~E::~~~ !~~~! ~ ~!~~ .. ~!::;~~~~~~ ~;i:2:~;~:;~ .. ~!~ i~~ ~~~=~ ~= ~~~;~ ~~ .. ~· ~~~~ ~ ~:~; Table I (Cont'd) The Observed and Calculated Structure Factors. Within each group are values of 1, lOFo, lOFc· ,Reflections indicated with asterisk were assigned zero weights in the final least squares cycle. :;:;::;: .......... -! .:..:. !,l,:::;::;::• =:;::;:::; =.... ,......... ! .:. .:. !. ,:, f =;:; ~ =:;:; ::;: "' . . "'::! .:. ! ! ,:, =:::;: ~ ~ ~:;; :;: ;: :; : ..... "' . . -! .:. ..~ ! ,:,-~: ;:: .; ~ ................ !.!. .!. !.! : : ::-: :; ::;::; : ......... - ! .:. .:. !. ,:, : : ; :: ~ ~ :: :.-: ~ :~ 5! ~ ~ ! ~ :! E~ ! ~: ~::: ! ~ ~ ~ ~ ~ ~ ~ ~ ~ ~: ~ ~ ~~:: ~~~!!!Ee:~~~~~!~~~~~::: := :~ ! ~ ~; ! ~ ~:: ! ~ !~ ~ e~ ~~ . ::;:::;:'!:.:':::'t~::'.!~8 .. ~':.:e:!~~!: :; ;~!!!E~~~?; ~~ §~ !:!~E :! "' ~!~:~~~g~: ;::::;: E; ::~;"' ~~!;~~~!~!!_. ............................................................... ...... .;.!;:: .. ! : ... "' ~:!:~!! ~~ ~~:E~:;i . . ~ . . ;~i:~~~~~~~~:H~:~. E1 !Ei ~~ i: ~~§~:!i ~~E i~ .. ~:::.;:~c;~;::;:;~=~:;:~ -;.'; '::i ~ ~::;: i~ ~ ~~~~;~~;~~ ~ ~~ ~ ~!:; ~ ~ ... =~!~~~g§~~~ ~ ~ ;~~~~~~~ ~ ... : :t~ ~!: ~: ~ .. ! :~ ~: ~: ~ ~ ~=; !~ ;g-:.:.~~~.::..::::':::2 ,,,,, __ -,,,,,,.. __ ;;;::;c;,..,u.. o!H!;~;;;~ ~;;;::o~ ........ o!!;.!~~;;;~ ~;;;:::;; .. .,.,..,,.o!J..:.!;~;;;~ ~;;;;:;o ............ !J.;.~~~±;;~ ~;;;::;;::; .......... ., .!. ,,............................................... -... ................. ............... ... .......... ..,_ ... .,..,. ...... -"'"" --·····---: : .... "' ... -! .:. .:. ! .:.: : ; : ~ ~:;;::::;:;;:: ::;:;~~ -.o- ... .,. ... - .... .,.,.._ - ...,_,_ ..... .,...., « .................... ,.. ,,. ... ~,. ... "' t I I I I I ... If It I• I I t • I ... I I I I - - ...,,, ::! ~:::~~=~~~::~~~ t: !~§~E~!~::~~!:!:::~!~:: ~~::::;~~~!!;:::~~:~!: !~;~~!g:~~~!;:~?~!::~: :~~~~:!!~:~: !~!~~~!~~ . . ~~~;;:;~:;~i~'" :~~~;~~~~~:!~~: . . ;~=~~~~:;;~!;:£:::: . . :~ I ' ,, ,, ... . . (P . ............................ ' - .............. ' .. i;~ Eti;~~~s~~~iE:i:; · ~;:~~s~:~~;~!~~~~;E;~- 5§~g~~~::~i~§;~~~~i~~! .. ~!~~~!~~~E:e~:~~~~~~~- ~:~:s~~~§~~-- :: :: !~~~~~!~~:~~~g~d-. ~~ c..n !:!: ' ;;;:~;'.:::.~;:;!:;<; .:;:;::: ........... .!..:..:..:..:.:::::;;:;; :;:;::::: .............. .!..!.j..!..:.:::::;:;; :;:;:;:::::.............. ............ !!!!;~ ... ............ ::;o ............. !!!!o:::: o;;:;;; .......... o!!J.!;::;;; ---- I I I 1 ...... ,........ ..... ..,, .. ~,. ........ .., .. ,.,.c..... .,.,. ~::;::;~:::~:::=";:.:.;~;: ............ I I ;;;;:;; ........... .:. ;.;. ! ; ::;; ;;;:; I ... ,..,..,..,.. .,.,,...,na.••"'o"'••"'<>"'•""' t : i::::ii:E!~~=mm~~:: ~::E~~:m~~:~m~~: :b~~s::i~~: ~;m;~ : l~~:!~!~~~ . . =:~:~~~~~:~~~: l· ::::~::!~~~ l~;:m: ~! !:!~~~S ~~:; :~:~ ! l ~ l ~ ;:~ !f Z~~~ ~:!:6~!~ . . :;:::~ ::! : ::: : : ~ i ::; ~ ~ ~~~ . . ! ... ! '_ ... ' ! ....... :;;~~i5~g~~.:~~~. i~ i ~ ~!~~~: i~ ~~~~ .. ~~s~~~~~ ~!~~:= ~;~ ~ . ..... ___ ,,.,,...,_.,..,.,. ... ,,. .. ; ~ ~;~ ~; ~ i ;~ ~~ ~ ~!; t ~~ .. :s:~~='.:~~:Ei~~:!~~;f' ~1:~~~!E~~e;:~:;~~~=- ~!~~==~~~iin i;;;~~~~It I t I I - I I ... - .. ~.c- .... ... - .. ' ........ ~::::g:;;:-;:":<:::.::;:';' & -.~- ......... -.-«--.,, .......... -.., .. «,, ............. !!j..!.~= ::: ............. !.!.J.!~== ::: ........... -.!.!J.!.:.::~ :;:;: ............ -.!.!J.!~=~~~ :;~:: ........... -!!J.!.:.::;: ---;;::;;;;o ........ e.:. ! ~ .d:~;; ............ .................................... ·····-----. . • ~:;::: .... ,.. ... - ! .:. .:. .!. .:. : : ;; ~ ~ •. :;:~:m:::1:~~::m: ::~;:::1:E: ::::fo1~: m:::mm: ii::~~m::bm;: :::~:rn:~m~~~- ;::; .. .. I;.,,:;::::::::~;:;:~;:::; ... "' ......................... ,.. ......... . . g: :~~~~! !!~:~! l f:;? . . ~~~=:~:~ ~~s:~:~:~';' i . . :: ?!'!==~=~! ::; ~:!~ ~~ . . ~ :I: ~ f ,;i;;;,g~~:!:~if;~~::" =;:;: .......... -!!!!!:::: It 11 I Cl ::;!!.>!:!::=~ ......... -...... :i:i~ .. ~:1- o.!.! ~- :~~~:Ei~~i~ .. ~~~S:~~:~:~~~:( ~:~:;~i~z=~~~;(' ;[ :~ ~z~!;; ~~~:~;i . ~! !:f ~ ~: ~~~~~~ .. : :; ........ !!!! :; ............. ~!!!;:;::; .. ::;o ............ !!.:..:.;~± ;::;~ .... ,..,,o!!~!;~;; ;:::::o.~,. ... o!!!.:.;~;; r· ·- .!__,!.,,.! ..... .:. ..... !.!.! C:';' !.! ~ ~::: ~:;:; ! .:::~; ! : ::;;: ;;;;;:; .......... !!!!;~~;; ~~;::i~!~~!i!2!~~~~i . . ;;-;:;:;;.,.,. ,. ... o!! 0:.; ;~.;; ;:;::;;;,..,..,..,o,!.!;.!;;;;; ... :~;~: .... ~:!::;t;~ .. ~~'::;~ : ! ~ :: ~: :: ! E~: !~ !! ~: :::~:g~~:g: ! 1~: ~:~:~~!: .. =~~~~~=;;::~g 5E: ~ ~ ~ ~ ~; ~: E~~. ~E ~= ~: ~ ~ ~ ~ ! ~ ~::: .. :: ::;:;~~~~~:;:::! :~: :~:B~~~~;:~:~: 1::! . . .......... ,.,,. ...................... . . ~~ ?:: =~~: :: ~~ ! : !! l? ,!. .:. ... '.!..!. .... !!.!. .. - ........ ;:t:~:::;: ~::.::':! ~ ;~::. ....... o.!.J.i:.!:t~ !,,._j.!.., .. !o ,!._,!.' ... ..,! .. ..,.!.,!...,._ .... .. ! .. '.!. ... ! .. ' ... .!..!.'" .!..!. .... !! ... ! ! ... .!..!.! ... '"' !~;~;~~~:::~i- ~..:::-:.<:.";:;~';:':;:;!~!;.:!~ ~ i :~ ~~;~~~~~~~~~. ... ::::~: .. ;::~<;;:;~;::.:: :~!~~i~~~;i~~~::~~ . . ';...... :C!':: '! ~=-:. ~'! ';ci ~';.",.:';: ~ g ~~:!.~~~~i~~~~~~: ~; :.::.!~o;;:;~e~ -.::o ....... o.!.!!!o::;; ~ o;::;:; ........ !!!!;:;::;; ...... .,............. ... ...... ............. !,!.j.!,! :;: .............. !.!..:.!l:;~ : ............. .!..:..:.!.:.:::;; :;:;;: ............ !!j.!!:;~ --,,,,, __ _ ......................................... .. ............................... ... -·····----"' ... .... ::=~~=:1:~;:; ::::::!!~~~~:~:~ ~::g~~~~~~tet:~~~ ::~~::~::;; ::! . . ~:::!~n:;~ .. ~~e;E~=~~=~ . :~:=~~!g!~~!!. =:;:~::!;~~~:e: . . ~~:::::!:=::!::~~1 . . ~:g~:~:::::E:e:~. ::=~!.~~~::1:=~~~~;~ ! '.!.! .......... ,!,!..,! ... .!."' ;:~~5~:~:~~:-· ~:~;~~:~::::S~~~~ .. ~:::;.:.~:E~.:;~( =~~~;::~: ~::( t::~~~S~~- .. ~!:~E!~!~~~ ... ::z~~;~=~~~~! :-:~~~==~~:;:;;;:~:;: .. ~ i~i::~i~~~;g~~i:; ~~ .. :; : i~:d ~ ~~~ ~~ ~!~ i h:: ~~ ==; ! ~s~ ~? . I' • ... .... .!l ....... ., ............. .. ........ !:: .......................... llltl 'I : .......... -!.!.!!~::: ' I • . "' ................... ,,,,, .... ...... w ........ .:.!!! .. ,. ................ ., ; ........... !!!!;::: ::::; ............ .!..!! !;:: .. o :.o:::::;~ .:;:.~~·- I.: e I:.:~ I= ;::;;~~~:!:g: :~;:~~m:: l; ~~:~;l~~:!~!: ::::~::: E~~!~:~:;; :~~~e:~£: t~ ::;:;;~!~; l =~ ~~;;;;~~~;:~:~~ ........... .,. .... _ ................... ... .... =~ :~:=~~ :::;~::::: ~:=~~!~:: ....:::lo;:;~ '' ......' ... ..... . :=:=!:!~':::: !;;;:i:~~ .. ::::.!::.:::::::~.:.: ~:::Z~:-2~~ ... ~ . . :~:::§~~!~~~:· o~ ' ... - '! ... !"' .. ~SEn ~~~:~ . ~~:~~i .:.i~ . . . .::;:~~ .... :~~~ !~~~~~:~~~:: ---"'oe"''""'"'D-"'""' r ::di" :i~::~~ . . ~ ~..,~ ~ ··~- 0 ............ 0 ..... ~. . . . . . . . . . " ......... .... Table II The Final Refined C, O, Br Parameters and Their Standard Deviations. Values for bromine atom have been multiplied by 105 , for the other atoms, by 104 • 2 The temperature factor is in the form T = exp { -(b11 h + b22 k2 + b3312 + b12 hk + b13hl + b23 kl)}. Atom x y z b11 b22 bg3 b12 b13 b23 Br (1) c (2) c (3) c (4) c (5) c (6) c (7) c (8) c (9) c (10) 37703(3) 230(2) 307(2) 789(2) 1285(2) 1247(2) 736(2) 870(2) 1754(2) 2118(2) 1765(2) 2601(1) 1270(2) 2032(1) 2340(2) 3277(7) 1634(4) 1347(5) 1916(4) 1705(4) 1853(4) 1323(4) 1414(5) 1250(4) 1242(4) 1405(4) 1125(3) 3028(4) 1715(3) 1045(5) -4767(6) 4168(4) 5303(5) 5875(4) 5262(4) 4094(4) 3626(4) 6951(5) 3831(4) 4809(4) 5636(4) 4841(3) 3777(4) 3006(3) 2487(4) 204(2) 11(1) 15(1) 14(1) 11(1) 11(1) 16(1) 20(1) 14(1) 12(1) 12(1) 14(1) 19(1) 18(1) 19(1) 1496(10) 92(5) 95(5) 71(5) 57(5) 72(5) 72(5) 95(5) 55(5) 50(5) 75(5) 99(4) 62(5) 52(3) 52(5) 822(6) 69(5) 65(5) 42(4) 45(4) 33(4) 54(5) 57(5) 46(4) 67(4) 48(4) 72(3) 48(4) 47(3) 46(4) -279(6) - 7(4) - 18(4) - 13(4) - 7(4) - 0(4) - 6(4) - 4(4) 1(4) 3(4) - 8(4) 8(3) 6(4) 5(3) 6(4) 252(4) - 5(4) 13(4) 2(4) 0(3) 2(3) - 5(4) 16(4) 6(4) 0(4) 2(4) 0(3) 8(4) 20(2) 5(4) 385(13) - 17(8) 9(8) 29(8) - 6(7) 8(7) - 12(8) 37(9) 7(7) 20(7) 6(7) 2(6) 30(8) 11(5) 16(9) c (11) 0 (12) c (13) 0 (14) c (15) I CX> o:i I Table II {Cont'd) Atom x y z bll b22 b33 b12 b13 h23 c (15) 2340(2) 2340(2) 2671(2) 3038(2) 3379(2) 3318(2) 2950(2) 2627(2) 701(2) 1144(2) 1071(3) 188(2) 129(3) 573(3) 1045(5) 107(3) 1606(4) 1027(5) 1532(5) 2602(4) 3181 (4) 2689(4) 3125(4) 3760(5) 4834(5) 3547(5) 4625(6) 5243(5) 2487(4) 2600(3) 1770(4) 1251(4) 597(4) 429(4) 910(4) 1596(4) 6025(4) 6262(4) 6481(5) 6021 (5) 6229(5) 6458(5) 19(1) 36(1) 14(1) 23(1) 19(1) 14(1) 15(1) 14(1) 17(1) 20(1) 29(2) 17(1) 32(2) 43(2) 52(5) 52(3) 46(5) 60(5) 97(6) 74(5) 55(5) 51(5) 82(5) 70(6) 89(7) 102(7) 134(8) 83(7) 46(4) 70(3) 40(4) 60(5) 51(5) 46(4) 55(4) 53(4) 29(4) 62{5) 74(6) 71(5) 91{6) 69(6) 6(4) 5(3) 1(4) 15(4) 11(4) - 14(4) - 11(4) 0(4) 5(4) 3(4) - 2(5) 19(5) 83(6) 20(6) 5(4) 49(3) 2(3) 13(4) 16(4) 4(4) 4(4) 9(4) 5(4) 5(4) 9(5) 6(4) - 9(6) -11(6) 16(9) 12(6) - 4(7) - 1(9) - 20(9) 2(8) - 2(8) - 20(8) 2(8) - 13(9) - 26(10) - 24(10) - 54(12) - 43(10) 0 (16) c (17) c {18) c {19) c {20) c (21) c (22) c {23) c {24) c {25) c {26) c {27) c {28) I 00 -::r I -88- Table III Atomic Distances (The numbering scheme represents a convenient crystallographic one and is not the correct organic nomenclature) 0 0 Atom Atm Distance (A) Atom Atom Distance (A) Br (1) c (2) c (2) c (3) c (4) c (4) c (4) c (5) c (5) c (6) c (6) c (6) c (9) c (9) c (10) c (10) 0 (14) c (15) c (15) c (17) c (17) c (18) c (19) c (20) c (21) c (23) c (23) c (20) c (3) c (7) c (4) c (5) c (8) c (23) c (6) c (11) c (7) c (9) c (13) c (10) 1. 89 1. 53 1. 53 1. 55 1. 53 1. 55 1. 56 1. 54 1. 31 1. 53 1. 53 1. 54 1. 52 1. 44 1. 45 1. 20 1. 35 1.19 1. 47 1. 38 1. 39 1. 39 1. 38 1. 35 1. 39 1. 38 1. 38 c (24) c (25) c (26) c (27) c (2) c (2) c (3) c (3) c (7) c (7) c (8) c (8) c (8) c (9) c (11) c (13) c (13) c (13) c (18) c (19) c (21) c (22) c (24) c (25) c (26) c (27) c (28) c (25) c (28) c (27) c (28) 1. 40 1. 34 1. 40 1. 37 1. 06 1.04 1. 08 1. 08 1. 08 1.05 1. 06 1.04 1. 08 1. 03 1. 07 1. 09 1. 03 1. 04 1.02 1. 03 1. 00 1. 00 1. 00 1. 02 1. 01 1. 08 1. 05 0 (14) c (11) 0 (12) c (15) 0 (16) c (17) c (18) c (22) c (19) c (20) c (21) c (22) c (24) c (26) H (50) H (51) H (45) H (46) H (47) H (48) H (32) H (33) H (34) H (49) H (44) H (29) H (30) H (31) H (36) H (37) H (38) H (35) H (39) H (40) H (43) H (42) H (41) -89- Table IV Bond Angles Atom Atom Atom Angle ( 0 ) Atom Atom Atom Angle ( 0 ) Br (1) Br (1) c (3) c (4) c (5) c (5) c (5) c (8) c (8) c (23) c (6) c (6) c (11) c (7) c (7) c (7) c (9) c (9) c (13) c (2) c (10) c (10) 0 (14) c (11) c (20) c (20) c (2) c (3) c (4) c (4) c· (4) c (4) c (4) c (4) c (5) c (5) c (5) c (6) c (6) c (6) c (6) c (6) c (6) c (7) c (9) c (9) c (9) c (10) c (19) c (21) c (7) c (2) c (8) c (23) c (3) c (23) c (3) c (3) c (11) c (4) c (4) c (9) c (13) c (5) c (13) c (5) c (5) c (6) c (11) c (10) c (10) c (11) c (9) c (9) c (10) c (9) c (17) 0 (14) c (6) c (6) 0 (12) 119 120 109 114 110 111 107 107 108 114 114 120 126 112 111 110 112 99 113 113 110 105 114 130 0 (12) c (5) c (15) 0 (14) 0 (16) c (15) 0 (16) c (15) c (17) c (15) c (18) c (17) c (18) c (17) c (22) c (17) c (19) . c (18) c (20) c (21) c (22) c (17) c (24) c (24) c (26) c (25) c (28) c (27) . c (28) c (25) c (19) c (20) c (21) c (22) c (23) c (23) c (23) c (24) c (25) c (26) c (27) c (28) 106 124 110 116 124 0 (14) 124 0 (14} 112 c (22) 119 c (15) 118 c (15) 123 c (17) 120 c (18) 119 c (19) 122 c (20) 120 c (21) 120 c (26) 120 c (4) 119 c (4) 121 c (23) 120 c (24) 120 c (23) 119 c (26) . 120 c (27) 121 -90- Table V Hydrogen Coordinates Values for hydrogen atoms have been multiplied by 104 • Isotropic temperature factor for all hydrogens equals 4. 0. Atom x y z H (29) H (30) H (31) H (32) H (33) H (34) H (35) H (36) H (37) H (38) H (39) H (40) H (41) H (42) H (43) H (44) H (45) H (46) H (47) H (48) H (49) H (50) H (51) 1619 1330 920 979 1166 495 2348 3096 3657 2910 1505 1414 540 261 160 1932 388 61 780 681 1654 241 140 3396 3095 3425 610 1827 1455 3086 235 1086. 4184 3008 3963 6868 7380 7323 1961 1381 234 811 6254 6636 6619 6220 5839 6406 5336 5685 3640 2844 3610 4171 3797 3967 3416 5263 6055 5015 3161 1321 507 1446 471 1497 490 2477 1434 -91- --------- ------------------------- -92- REFERENCES 1. D. J. Duchamp, paper B-14, Amer. Cryst. Assoc. Meeting, Bozeman, Montana, 1964, p. 29. 2. C. K. Johnson, ORTEP ORNL-3794, Oak Ridge National Laboratories, Oak Ridge, Tennessee, 1963. -93PROPOSITION I It is proposed that the use of negative ion mass spectrometry as a mechanistic tool in the study of elimination reactions be studied. The ring opening of 3 -carboxy and 3 -unsubstituted pyrazoles, each substituted with different electron-attracting substituents in the 4-position, has been studied by heating the acids in boiling quinoline for 90 minutes and by heating the 3-unsubstituted pyrazoles with potassium t-butoxide in t-butyl alcohol for several hours 1 . The authors found that the relative ratio of j3-anilino-acrylo-nitrile to decarboxylated 4-substituted pyrazole increased as the electronattracting capability of the 4-substituent increased.· These experiments support the reaction mechanism hypothesis of the pyrazole ring cleavage described in Chart A. Chart A The electron-attracting substituents in the 4-position were acetyl, benzoyl, cyano, nitro, and hydrogen. The substituents in the 5position were methyl, phenyl, and hy~rogen. -94- Many of the arguments used to rationalize and predict the course of organic reactions in solution have been applied to the study of unimolecular decomposition of ions in the gas phase. It is proposed to study the above system in an attempt to determine whether a corrolation exists between electron-impact decarboxylation and ring-opening, and solution chemistry decarboxylation and cleavage by comparing the ratio of the abundance of the p-45 fragment to the abundance of the p-1 fragment peak heights as the electron-attracting 4-substituent is varied. The corrolation should exist because the probability of occurrence of a particular gas phase electron impact initiated reaction is increased if it produces a more stable ion. A positive corrolation would have the result of greatly broadening the application of negative ion mass spectrometry by employing it as an easily applied non-solution chemistry mechanistic tool for the study of certain elimination reactions. Recently it has also been reported that 3-unsubstituted-1alkyl-indazoles are cleaved to the corresponding N-alkyl-anthranilonitrile by reaction with sodium amide and caustic soda in boiling xylene 2 as shown in Chart B. Chart B 6a4 s H I f~N I ~ 11 2 s I n-Bu 0 II c,~ OC NH I n-Bu 30% + ~C=N ~NH I n-Bu 34% -95A correlation between both solution and electron impact decarboxylation and ring opening of either 3-unsubstituted or 3-carboxyl indazoles substituted in either the 6- or 8-position with various electron-attracting substituents should exist. The sodium salts of 5-alkyl isoxazole-3-carboxylic acids on mild thermolysis readily undergo decarboxylation. followed by ring 3 4 cleavage ' as shown in Chart C. Chart C 5-alkyl isoxazole-3-carboxylic acids substituted with various electronegative substituents in the 4-position consequently lend themselves to a similar study. The same type of decarboxylative ring cleavage has recently been reported upon distillation of 2-phenyl-l, 2, 3-triazole4-carboxylic acid from barium·hydroxide 5 as shown in Chart D. Chart D 0 II H 4 5 1( H-0-C ~ Js- - ~ 1 ~ I ~~ en( --> Jli N HN/ I ~~ . H _ 191\\( ___,.) N H 1 N '-NH I ~~ -96Two recent positive ion mass spectral studies have demonstrated the mass spectrometers ability as a mechanistic elucidation tool. A linear relationship between the intensities of the fragments corresponding to the benzoyl and acetyl ions in the positive ion mass spectra of substituted benzophenones and acetophenones has been found 6 . Much of the data can be corrolated by the common relationships of solution chemistry. The ratio of the intensities of the 2norbornyl radical carbonium ion fragment peak to the two exo- and endo-2-norbornyl bromide molecular ion peaks determined from electron -impact studies in the gas phase has been found to be comparable to the corresponding ratio obtained from solvolytic studies 7 . Negative ion mass spectra have neither been tabulated to any large degree nor applied to analytical problems. The apparent lack of extensive fragmentation has negated its use as a structural elucidation tool. It has been found that the formation of positive ions is about four orders of magnitude more probable than the formation of negative ions when using 70 ev ionizing electronsSa, Sb, the standard energy beam. Negative ions are often formed with excess kinetic energy which greatly reduces the collection efficiency for them by mass spectrometers. Ion-molecule reactions sometimes result, yielding substances whose mass is greater than the molecular ion. Another frequently introduced objection is that the negative ions are produced by four independent non-concurrent modes of monoenergetic electrons 8b. -97In addition to the fact that negative ion mass spectrometry possesses unique advantages for certain mechanistic studies, many of the disadvantages cited are easily corrected. If the elimination pathway is energetically favorable then the lack of extensive fragmentation results in a faster assignment and easier study. Considering the instrumental conditions to be used for the decarboxylative ring-opening study, it is felt that the only two major peaks will be the p-1 and p-45 peaks. Exotic molecular rearrangements frequently found in positive ion mass spectrometry have not been noted in negative ion spectral studies. High detector sensitivity is achieved by the use of a 10-20 stage electron multiplier which can usually increase the detection sensitivity for both positive and nega4 tive ions by a factor of 10 or more over conventional electrometer techniques. Both positive and negative ions are produced upon impact with the electron beam in the ionization chamber. Spectra of negative ions may be obtained principally in the same way as positive ions with a normal mass spectrometer after reversing the potentials in both the ion source and analyzer tube. A small negative potential between the back wall of the ion source and the first accelerator plate pushes the negative ions toward the accelerating region, attracting at the same time the positive ions which are then discharged at the repeller plate. The study will probably be performed at electron energies lower than the appearance potential of the positive molecular ion and no positive ions will therefore be generated. The four modes of negative ion formation occur at different energies (i.e., experimental conditions), and each of the mechanisms can therefore be -98studied separately. Ropp 9a, 9b has applied .the arguments used to predict and rationalize the ionization of formic, acetic, and propionic acids in solution to the negative ion mass spectrometric study of these acids. The cracking of deuterated formic acid yields a positive· ion at mass 45 by loss of either hydrogen attached to oxygen or carbon with equal facility. The negative ion generated by loss of a single hydrogen from formic acid is formed almost entirely by loss of hydrogen from ' oxygen and not carbon. The negative ion observation is expected based on the resonance stabilization of the carboxylate anion and the inductive effect of the oxygen atoms. The ratio of the loss of the acidic proton to yield the base peak ion RC02 - (R = H, CHs, C2 H5 ) from both acetic acid and proprionic acid compared to the proton lost from formic acid was 4/10. This is in accordance with the inductive effect of methyl and ethyl versus hydrogen in carboxylic acids. Acetic acid would be expected to form the carboxylate anion more readily than proprionic acid, although the difference should be much smaller than the difference between formic acid and acetic acid. Decarboxylation can occur by both the pyrolytic and electron impact routes. The decarboxylation should be controlled so that a single route is being studied throughout the experiment. Melton found in his study of formic acid that the temperature of the ionization chamber had to be reduced to 120°C since the formic acid sample decomposed into hydrogen and carbon dioxide at higher temperatures 9a, 9b. M. von Ardenne, in his negative ion mass spectrometric -99study of the stable high-melting fatty acids, noted no thermal decarboxylation10. M. von Ardenne also noted the presence of low relative abundance peaks which he ascribes to the loss of acetic acid from acetoxy steroids and water from sterols 11 . He feels that these peaks result from pyrolytic decomposition due to the high vaporization temperature required. Vaporization of the cyclic compounds to be studied should not present the problem that was encountered in the high molecular weight fatty acids and steroids. Thermal decarboxylation can be achieved by injecting the sample into the normal sample inlet system--maintained at a temperature high enough for decarboxylation. Electron impact decarboxylation can be achieved by injecting the sample into the normal sample inlet system--maintained at a temperature low enough to preclude decarboxylation; or preferentially, injection of the sample directly into the ion source. The temperature of the ion source compartment at which no thermal decarboxylation and ring opening occurs can be determined by observing the relative ratio of the p-45 to the p-1 peak as the temperature is both increased and decreased in small increments. Partial thermal decomposition is indicated by slight increases in the product ion intensity ratios for several substituted - compounds relative to the parent compounds as the temperature is raised. Thermal decomposition is absent if the ratios do not decrease as the temperature is lowered. Negative ions are known to be formed in the gas phase by four different mechanisms, three of which can be initiated directly- by electron impact. -100- The electron-impact-initiated reactions are: a. dissociative resonance capture, AB + e --> A+Bb. resonance capture AB+ e ---) AB- c. ion pair production AB+ e Tabulations of negative ion mass spectra carried out at the usual operating pressures of less than 10 -6 mm and the usual electron energies of 50 to 75 ev would include only negative ions produced by pair production, mechanism (c). Conversely, negative ion mass spectra produced by 2 to 10 ev electrons 8b would be concerned mainly with ions produced by dissociative resonance capture, mechanism (a), and the relative abundance of each ion would be critically dependent upon the electron energy in this range. The appearance potential of B - (the appearance potential being the minimum electron energy necessary to produce a given ion from its neutral precursor) is equal to the dissociation energy of the AB bond minus the electron affinity of the B atom plus the excess energy (e.g., the sum of both the kinetic energy and the electronic excitation of the fragments). Since the dissociation energies of most chemical bonds lie in the range of 3 to 5 ev and the electron affinities of most atoms lie in the range of 1. 5 to 3. 5 ev, it follows that the values of the appearance potential of B- will be expected to lie in the range of 1. 5 to 4 ev 12 . By the same process we see that the appearance potentials of negative ions arising by an ion pair process will generally lie in the range of 6 to 13 ev. At electron energies less than approximately 2 ev, only negative ions produced by mechanism (b), resonance capture,· would -101be detected. The decarboxylative ring opening study, an example of the dissociative resonance mechanism, will be studied over the ionizing range of 2 to 6 ev. The intensity of the bombarding electron beam would be adjusted to a value which results in a spectrum sufficiently intense to display peaks for all the fragments of interest. The intensity has to be kept constant over the whole scan of the spectrum if reproducible spectra are to be obtained. The study of carboxylic acids, which have a low vapor pressure, and the use of a low electron energy, two factors which result in a much lower ionization efficiency, may require ionizing currents of 50 to 100 µamp in order to obtain a high ion yield. But the injection of the sample directly into the ionization chamber and the use of the electron multiplier detector probably will allow the use of the normal 10 µamp ionizing current. -102REFERENCES 1. R. Fusco, V. Rosnati, and G. Pagani, Tetrahedron Letters, 1739 (1966). 2. B.K. Martsokha and A.M. Simonov, Zh. Obshch. Khim., ..._._ 133 (11), 3777 (1963). 3. A~ Quilico in (R.H. Wiley, ed.) "Heterocyclic Compounds", Interscience Publishers, New York, 1962, p. 44-48. 4. A. Quilico and Freri, Gazz. chim. ital., 62, 436 (1932). 5. R.M. Carman, D.J. Brecknell, and H.C. Deeth, Tetrahedron """"' Letters, 4387 (1966). 6. M.A. Bursey and F.W. McLafferty, J. Am. Chem. Soc., 88, """' 529 (1966). 7. D.C. DeJongh and S.R. Shrader, J. Am. Chem. Soc., 88, _.... 3881 (1966). 8.a) C.E. Melton and P.S. Ruldolph, J. Chem. Phys., ll_, 1485 (1959). b) C.E. Melton in (F. W. McLafferty, ed.) "Mass Spectrometry of Organic Ions", Academic Press, New York, 1963, ch. 4. 9.a) G.A. Ropp, C.J. Danby, and D.A. Dominey, J. Am. Chem. Soc., 79, 4944 (1957) . ............. b) G.A. Ropp and C.E. Melton, J. Am. Chem. Soc., .....,.... 80, 3509, (1958) . 10. M. von Ardenne, K. Steinfelder, and R. Tummler, Z. Physik. Chem. ,....,_ 220, 105 (1962). 11. M. von Ardenne, K. Steinfelder, R. Tummler, and K. Schreiber, Experentia, J&., 178 (1963). -103- 12. C.A. McDowell in (C.A. McDowell, ed.) "Mass Spectrometry", McGraw Hill Book Co., Inc., New York, 1963, p. 527. -104- PROPOSITION II The existence, stability, and synthetic usefulness and versatility of phosphorous ylids have resulted in a search for other molecular systems containing the same general features. The investigation has led to the synthesis of many sulfur, nitrogen, arsenic, and antimony ylids of varying stability 1 . Relatively little is known about the physical properties of sulfonium ylids because many of them are noticeably unstable, and have therefore been generated and used in situ without any attempt to isolate and characterize them. The decreased reactivity of fluorenylidenedimethylsulfurane (1) 2 a, 2b, ""' the first stabilized sulfonium ylid, and the less highly reactive 22 nitro-2c and 2, 7-dinitro- c derivatives is a result of the decrease in basicity and nucleophilicity due to the delocalization of the negative charge by the aromatic system. The preparation and study of the chemical reactivity of a number of stabilized sulfonium ylids have recently been reported. These include phenacylidenemethylphenylsulfurane (2a) 3a and phenacyclidenedi,,_ methylsulfurane (~) b, 3c, a cyclic bis ester ylid ~ 3d, 3-dimethyl3 sulfuranylidene-2, 4, 5-pyrollidinetrione (4) 3e, dicyanomethylenedi"' -105me thy lsulfurane (§) 3f and cyclopentadienylidenedimethylsulfurane , (6)3g. ,,.... R1 ~0 / e 0 11 S-CH-C-C6 H5 ~ 4 ....... 3 .,..,. ~ , R 1 = c~, ~ = c6 Rs ~ , R 1 = CH3 , ~ = C~ \? ~'\CH HsC 5 ...... 3 6 ...... The various methods of y lid stabilization (e. g. , carbony 1 and cyanide conjugation, generation of aromatic character) produce a decrease in nucleophilicity and basicity. Sulfonium ylids are, in the simplest sense, really nothing more than carbanions, and would be expected to undergo normal carbanionic reactions. It is proposed to make use of this fact, the difference in stability of reactive and stabilized sulfonium ylids towards carbonyl compounds, the relative acidity of various ylid protons, and the excellence of dimethylsulfide as a leaving group, in the study of a new condensation technique for the construction of bicyclic, bridged- -106bicyclic, and spiro systems; thereby making use of sulfonium ylids as a synthetic tool. The technique involves the introduction of a four carbon keto-carbene precursor in a very simple manner. Chart A delineates one aspect of the proposed study. Chart A 1- - I I 'Z:E:Is/c :E:Is C:E:Ig CH3 's/ @ o:+Jo S@ .&0~0 0 I I~ ) I .I- - 7 "" I I I I I R R --' 8 ....... . l 9 "" hv 0 R 10 """" R 11 """" The base-catalyzed Michael condensation of substituted cyclohexanones with the sulfonium salt ...... 7 (or the corresponding sulfide which would later be converted to the sulfonium salt by r~action with methyl iodide) will yield the ylid ...... 8. A literature survey indicates -107that the synthesis of neither the sulfonium salt 1 nor corresponding sulfide has been reported. The synthesis of these compounds is indicated in Chart B. Chart B AlCls ClCR:aCOCl + C2 ~ - 79% 1) C6 H5 N(Et)2 5 ClCR:aCR:aCOCR:aCl 0% > 2) C~SH k SC~ or C~sw- SC~ C~I 0 ~Cl c~sc~ AlCls BrCR:aCOBr + C2 If.i --~ BrCR:a CR:aCOCR:aBr 60% Chloromethyl vinyl ketone has been prepared from the dichloroketone 4a, 4 b .' It is proposed to convert it to the monosulfide with methyl mercaptol as shown. Two equivalents of mercaptol will insure the formation of the disulfide if the mercaptide initially adds to the vinyl ketone in the Michael manner. The disulfonium salt can be converted to the monosulfonium salt through the base-catalyzed -108- Hofmann elimination of dimethyl sulfide 5 . Cyclopentadienylidenedimethylsulfurane (6) was synthesized3 g by elimination of dimethyl ,.... sulfide from the disulfonium salt. The dibromoketone has also been synthesized by the Friedel-Crafts reaotion4 b, and can be substituted for the dichloride in the above sequence. Epoxides have been prepared by the reaction of non-stabilized sulfonium ylids with carbonyl compounds. Corey has reacted dimethylsulfoniummethylide 6a, ethylide, and butylide 6b with a variety of carbonyl compounds including benzaldehyde, benzophenone, carvone, and cyclohexanone. Johnson has reacted benzylidenediphenylsulfurane2b with benzaldehyde, benzaldehyde carrying substituents ranging in electronic effect from nitro to methoxyl, and aliphatic aldehydes. The reaction with cyclohexanone was not successful. Aromatic and carbonyl stabilization of sulfonium ylids produce a decrease in reactivity because of the ability of these electronattracting groups to efficiently disperse the electron density of the ylid carbanion. Fluorenylidenedimethylsulfurane (1) ,.... reacted only with benzaldehydes carrying electron-withdrawing substituents 2b. It did not react with cyclohexanone. The stability of the recently reported phenacylidenephenylmethylsulfurane (2a) 3 a was such that it could not """ be thermally decomposed. It was also claimed not to react with E_nitrobenzaldehyde. This is contrary to the behavior of phenacylidenedimethylsulfurane3b which formed an oxirane, although in low yield, upon reaction with E_-nitrobenzaldehyde. It has been suggested 7 that stable sulfonium ylids are of limited synthetic utility since nucleo- -109philic reactivity is weak and reaction with carbonyl compounds sluggish or nonexistent. Future discussion in this proposition will disprove this observation. The reactivity of ylid ...... 8 towards epoxide formation cannot be predicted with any great degree of certainty. But it is felt that based on the analogies described above it will be a fairly stable ylid, resistant to reactivity even upon being heated for a short length of time. The stereochemistry of the ketol intermediate in the Robinson annelation of 2-substituted cyclohexanones with methyl vinyl ketone has been determinedBa, Bb. Extrapolation to the analogous sulfonium ylid situation results in the prediction that the epoxide will be transfused when the R-substituent is hydrogen and cis-fused when it is methyl or another larger group. The formation of substituted cyclopropanes and alkenes from ylid decomposition has been ascribed to the generation of an intermediate carbene 1 ' 2b, 3 a. The formation of the phenacyl carbene (12), .......... generated by the photolysis of phenacylidenedimethylsulfurane (2b), .......... 3 has recently been reported c. The photolytic decomposition of diazoacetophenone 9 produced the same products in very similar ratios as hv > 2b 12 the ylid decomposition. This observation indicates that the same -110- intermediate carbene 12 is formed in both cases. The butene """" trapping experiments of the carbene indicated that it exists in the triplet state. If, as expected, the sulfonium ylid ...... 8 is unreactive with respect to carbonyl attack, the photolysis of the ylid would yield the keto-carbene which would add to the carbonyl double bond to form the epoxide ...... 9. This method would therefore be an excellent technique for the generation of a four carbon keto-carbene precursor. The use of cyclic and acyclic amines and alcoholate anions instead of enolates would greatly generalize the approach. The formation of cyclopropyl ketone 11 from ylid 8 would """" ..... expand and generalize the usefulness of the ylid. The keto-ylid !!, would be converted to the exomethylene compound by the use of methylenetriphenylphosphorane. Photolysis of the resulting ylid would yield the cyclopropyl ketone ........... 11. It is felt that initial attack of the phosphonium ylid will occur at the carbonyl carbon and not the sulfur atom. If initial attack does occur at the sulfur atom then the generated betaine would either reverse to reform starting material {which would eventually form the desired methylene ketone with irreversible elimination of triphenylphosphine oxide) or "ylid-interchange" would occur yielding methylenedimethylsulfurane and a new keto phosphonium ylid (e.g., a new ylid generation technique). It was found that phosphonium ylids which were not reactive enough to be cleaved by oxygen would react with ozone. Thus phenacylidenetriphenylphosphorane 13 reacted with ozone at -70° C in """" methylene chloride solution to afford phenylglyoxal (14) .......... and triphenyl- -111- phosphine oxide in high yieldlO. It would be interesting to explore the analogy between ketophosphonium ylids and ketosulfonium ylids, C6 IfsCOCH =P(C6 H5 ) 3 13 Os> C6 H5 COC;!J-P(C 6 lis)3 -7C6 IfsCOCHO 6_c6-oe + @ (C6 Ifs) 3 PO + ~0 11 · 2 ~EC6 H5 )3 PO~ as the applicability of reactions involving the more broadly investigated phosphonium ylids to sulfonium ylids such as~ would broaden the synthetic utility of these compounds. Vinyldimethylsulfonium sal(s have been shown to undergo typical Michael addition reactions with a variety of nucleophiles including hydroxide, ethoxide, and sodiomalonic ester 11 . It is proposed to study the alkylation of 2-substituted cyclohexanones with the vinyldimethylsulfonium salt 15 as shown on the following page in Chart C. """" The ylid 16, not being stabilized by the carbonyl group, is """" quite_ reactive. It can hypothetically react through any of the intermediates shown below to yield any of the described products. The protons alpha to the carbonyl are more acidic than those on the alkyl chain alpha to the sulfur atom. This is shown by the fact that very strong bases such as the dimethyl sulfoxide anion are required to generate the ylid from the trimethylsulfonium sau 12 . It l.s therefore felt that a prototropic shift will occur in the ylid 16 forming the """" enolate-salt 11 or]& at the expense of the formation of the 6, 4-epoxide. -112Chart C no ~s~ . ~o;:s® ('r-o ;s~ rrro (>s~ ~--)~or~ or~ c~ R Clig 0 . CH3 /Clig e Clig Clfs . R 0 Qj d R CH3 CH3 l 18 """" l 0 R It is difficult to predict the equilibrium concentration of enolate anions as many subtle factors must be taken into consideration. These factors include the relative acidity of the involved alpha protons, the effectiveness of the leaving group, the relative rates of ring closure, the nature of the solvent and base, and the tendency for C- versus -113O-alkylation 13 a, b, c. House found that there was little preference for the initial removal of the axial proton at C-2 rather than the axial proton at C-6 in 2-methylcyclohexanone. In general, the relative rates of closure for rings of varying size follow the order three-, five-, and six-membered rings> seven-membered rings> fourand larger than seven-membered rings, and noncyclic compounds. The choice of reaction conditions may be of prime importance in deter mining the ratio of 0-alkylated to C-alkylated products obtained. Nonhydroxylic polar solvents appear to favor 0-alkylation. Alkylation at the more electronegative atom of the ambient anion is also favored by the use of a very reactive alkylation agent (i.e., one that contains a good leaving group such as dimethylsulfide, tosylate, iodide). Since the energy differences between many of the enolates studies are small (1 Kcal/mole or less), appreciable changes in the position of the enolate equilibrium are observed by changing the cation or solvent. The observations that changing from a lithium to a potassium or sodium enolate or changing the solvent from 1, 2dimethoxyethane to a more polar dimethyl sulfoxide shifts the equilibrium toward the more highly substituted enolate exemplify this point. The study of the Michael reaction with 2-methylcyclohexanone 13b seems relevant to the control of the product distribution of the intramolecular alkylation of the anionic salt. The equilibrium concentrations for the enolate anions of 2-methylcyclohexanone are summarized in Chart D. -114- Chart D 0 ee GG> OK OK OCH, HOCH, with (C 6 H5 )3 C - K+, DME withJ.-BuO - K+, t-BuOH ca. 35% ca. 65% ca. 20% ca. 80% with t-BuO - K+, DME ca. 22% ca. 78% 6CH, + It is seen that the positions of the relevant enolate anion equilibrium favors the more highly substituted isomer. This is borne out by the alkylation product ratio as shown in Chart E. Chart E 0 CI:i.? = CHC02 Clfs CH3 o-CH,CH,C0 CH, 2 t-BuO-K+ a) 93% (t-BuOH) b) 88% (DME) CH,6CH,CH,C0 CH, 0 2 a) 7% (t·BuOH) b) 12% "{!>ME) + -115- The total yield of monoalkylated products was 61 %. It is suggested that the greater proportion of the approximately 15% higher molecular weight products is formed by further alkylation of the 2, 6-dialkylated ketone. It would seem reasonable to conclude that reaction conditions could be controlled to lead to selective alkylation. The use of a non-polar solvent and a sodium or potassium base should lead to the spiro-cyclopropyl compound (when R = H)--cyclopropyl ring formation being kinetically favored and elimination of dimethylsulfide rendering the reaction irreversible. The intramolecular alkylation / of bromoketone 19 yielded mainly the enol ether shown in Chart F, . """"' Chart F Q. + 19 """"' 8% 78% the reaction conditions being potassium t-butoxide and benzene 14 . Although the reaction conditions favored C-alkylation (e.g .. , spirocyclobutane formation), the formation of a four-membered ring is not favored. Formation of the bridged (3. 3 .1.) ketone involves a six - membered transition state, but the axial conformation of the side chain is required and considerable nonbonded interactions are built up in the transition state. -116- REFERENCES 1. A. W. Johnson, "Ylid Chemistry~' Academic Press, New York, 1966. 2. a) C.K. Ingold and J.A. Jessop, J. Chem. Soc., 713 (1930). b) A.W. Johnson, V.J. Hruby, and J.L. Williams, J. Am. Chem. 86, 918 (1964). -Soc., """' c)E.D. HughesandK.I. Kuriyan, J. Chem. Soc., 1609(1935). 3. a) H. Nozaki, K. Kendo, and M. Takaku, Tetrahedron Letters, 251, (1965). b) A. Wm. Johnson and R. T. Az::el, Tetrahedron Letters, 819, (1966). 88, 1587 (1966). c) B.M. Trost, J. Am. Chem. Soc., ,,_ d) A. Hodrainer and F. Wessely, Tetrahedron Letters, 721 (1965). e) E.G. Howard, A. Kotch, R. V. Lindsay, Jr., and R. E. Putnam, J. Am. Chem. Soc., ,,_ 80, 3924 (1958). f) W. J. Middleton, E. L. Buhle, J. G. Mc Nally, Jr., and M. Zanger, J. Org. Chem., ~' 2384 (1965). g) H. Beringer and F. Scheid!, Tetrahedron Letters, 1757 (1965). 4. a) Yu. A. Arbuzov and B. L. Dyatkin, Doklady. Akad. Nauk. SSSR, 111, 1249 (1956). """""" b) J.R. Catch, D.F. Elliot, D.H. Hey, and E.R.H. Jones, J. Chem. Soc., 278 (1948). 5. V. Franzen and C. Mertz, -Ber., ........... 93, 2819 (1960). 6. a) E.J. Corey and M. Chaykovsky, J. Am. Chem. Soc., ........... 84, 3782 (1962). 86, 1899 (1964) • b) E.J. Corey and W. Oppolzer, J. Am. Chem. Soc., ........... -117- 7. K. W. Ratts and A. N. Yao, J. Org. Chem., g, 1689 (1966). 8. a) T.A. Spencer, K.K. Schmiegel, and K.L. Williamson, J. Am. Chem. Soc., 85, 3785 (1963). """' b) J.A. Marshall and W.I. Fauta, J. Org. Chem., ~' 2501 (1964). 9. D.D. Cowan, M.M. Couch, K.R. Kopecky, and G.S. Hammond, J. Org. Chem., ~' 1922 (1964). 10. F. Ramirez, R.B. Mitra, and N.B. Desai, J. Am. Chem. Soc., 82, 5763 (1960) . .--. 11. W. von E. Doering and K.C. Schreiber, J. Am. Chem. Soc., 77 """" 514 (1955). 12. E.J. Corey and M. Chaykovsky, J. Am. Chem. Soc., 87, 1353 """" (1965). 13. a) H.O. House, "Modern Synthetic Reactions", W.A. Benjamin, Inc., New York, 1965, ch. 7 and references therein. b) H.O. House and V. Kramar, J. Org. Chem., ~' 3362 (1963). c) H.O. House, W.L. Roelofs, and B.M. Trost, J. Org. Chem., 31, 646 (1966). """' 14. S.J. Etheredge, J. Org. Chem., g, 1990 (1966). -118- PROPOSITION III Widespread interest in the che~ical and physical properties of highly strained molecules has long been evident. The instability of these molecules is due to the extreme bond angle deformations. These compounds lend themselves nicely to a study of the effects of ring substituents on the properties of carbon-carbon bonds, on carbonhydrogen bonds, and on the various orbitals. Chemical reactions involving strained systems have proven to be quite interesting. Strained small ring systems also serve as interesting models for testing the prediction of spectral properties. ·The system with the most strain per carbon atom thus far synthesized is bicyclobutanela, lb. The synthesis of a new highly strained system, the bicyclobutanone system, is proposed. The bicyclobutanone system can be converted to bicyclobutanes substituted in the methylene position. These compounds are quite difficult to prepare. It is proposed that the bicyclobutanone system ..... 2 be synthesized by the following reaction. Projected syntheses of the diacid 1 are 0 ~ 2 ..... -119indicated in Chart A. Chart A a) ~OHC 1 ..... OHC c~ b) 0 owo u o~~o 1 ..... 4 ...... c) 0 <> 0 0 ~ Br-<>-Br ~ NC-<>-CN ) 1 ..... -120- 5-Norbornen-2-one 2 (3) can be converted by selenous acid "" to the a-dione, which is then converted to the diazo ketone (through the tosylhydrazone). Irradiative ring contI'.action yielding the acid (aqueous solvent) or ester (alcoholic solvent) is brought about by the Wolff rearrangement. Involvement of the double bond in the carbenic rearrangement would necessitate the protection of the bond, possibly as the acetonide or the thionocarbanate. The acid or ester can be reduced to the alcohol with lithium aluminum hydride and then dehydrated to yield the exo-meth)Tlene compound. The exo-methylene compound can also be synthesized by converting the acid to the substituted amide through the acid ohloride, then to the amine through lithium aluminum hydride reduction, then to the amine oxide which is thermally eliminated to yield the terminal methylene compound. The terminal methylene is then oxidized to the ketone through the use of ozone or potassium periodate--potassium permanganate--potassium carbonate. This sequence is discussed and has been used by both 3 Meinwald a and Wiberg3b on similar systems. The dialdehyde can be oxidized to the diacid 1 with the aid of silver oxide. "" The alkylative conversion of glutaric anhydride (4) to the cis" -- 2, 4-cyclobutanonecarboxylic acid anhydride would be accomplished by the use of diethyl carbonate, ethyl chloroformate, or phosgene (employing a base such as sodium hydride). Cis-1, 2-cyclobutanecarboxylic acid anhydride is a known compound. The anhydride is readily hydrolyzed in water to the diacid. A literature search failed to turn up the corresponding ~-1, 3-anhydride. A molecular model -121study indicates that the cis-1, 3-anhydride is readily capable of existence. It should undergo hydrolysis much more readily than the cis-1, 2-anhydride to yield the diacid 1· It is certainly conceivable that the second bond-forming reaction will reverse itself. The formation of a bridged cyclobutanone, bicyclo (3 .1.1) heptanone, by alkyla2 . 4a 4b 4c tive SN displacement has been reported by three groups ' ' 4 Wiberg d though, in investigating this reaction with a variety of bases, found that the product is a mixture of two ketones, formed in about equal amounts. The possibility of decarboxylation must also be + protected against. Another possible route to the diacid ,,.., 1 involves the formation of the 2, 4-dibromide of cyclobutanone through the employment of two equivalents of N-bromosuccinimide, and the subsequent conversion to the diacid (through the intermediacy of the dicyanide). The conversion of diacid ,,.., 1 to bicyclobutanone (2) ....... would be accomplished by Kolbe anodic oxidative electrolysis 5a, 5b. When trans, trans, trans, 1, 3-decarboxy-2, 4-dicarbomethoxycylobutane (~) was electrolyzed under Kolbe conditions 2, 4-dicarbomethoxybi- cyclobutane (§), assumed to be cis, was the major product isolated (17% after gas chromatography) 6 . -122- 5 The usual procedure for carrying out reactions of the Kolbe type is to dissolve the starting acid in methanol containing sufficient sodium methoxide to neutralize asmall portion (about 2%) of the acid, and then to electrolyze the solution between two platinum foil electrodes until it becomes slightly alkaline. As the electrolysis proceeds, the carboxylate ions are converted at the anode into products. The sodium liberated at the cathode reacts with the solvent, and the resulting alkali neutralizes some of the excess acid. The process continues in this way until all the starting material has been consumed. This procedure suffers from the fact that the nucleophilic methoxide ion formed could be expected to attack the generated 7 bicyclobutanone system. An alternate procedure , which eliminates this potential problem, is to use sufficient base to neutralize all of the starting material and to electrolyze the resulting methanolic solution between a platinum anode and a mercury cathode. As long as a potential difference is applied across the electrodes the electro- -123- lyte does not turn alkaline. The excess sodium liberated at the mercury cathode is held there as the amalgam. Owing to the initial high carboxylate ion concentration, large currents can be used with a consequent saving in reaction time. The current drops as the carboxylate ion concentration is reduced by electrolysis, and the termination of the reaction is therefore readily detected. Many successful Kolbe anodic oxidations of carboxylic acids have been reported to proceed without the use of any base. Acetonitrile (neat) 8a, acting as an ionizing agent producing carboxylate anions, and acetonitrile--methanol mixture 8 b, Sc are frequently employed. Acrylonitrile and triethyl amine 8d, triethyl amine and / dimethyl formamide 8 e, and methanol and pyridine 8b have also been cited as suitable solvent systems. The maintenance of the integrity of the carbonyl function during the electrolysis of diacid 1 is a moot point. A literature "' review did not yield any pertinent data for electrolysis of .B-keto acids, but successful electrolytic decarboxylative dimerizations of acyclic y.- and E-keto acids have been reported9a, 9b, 9c. Alphaalkylated keto acids decrease the efficiency of the electrolysis 5a, 5b. Alkyl groups further removed from the carboxyl group than the a.position seem, in. general, to exert little if any adverse influence. Phenyl groups in the a-position have been noted not to hinder the . coup 1mg reac t·ion2a, 2d, 2e Bicyclobutanes, formed in good yields by the electrolytic cyclization of 1, 3 -dihalocyclobutanes, have recently been reported. -124The electrolysis of 1-chloro-3-bromocyclobutane 1Oa in dimethylformamide saturated with lithium bromide afforded a mixture of bicyclobutane (60%), cyclobutene (20%) and cyclobutane (10%). Similarly 1, 3-dibromo-1, 3-dimethylcyclobutane was electrolyzed to give 1, 3-dimethylbicyclobutane in from 55 to 94% yieldlOb, lOc. The electrolytic cyclization of variously substituted 2, 4-dihalo cyclobutanones would therefore seem to offer an effective alternative route to the bicyclobutanone system. The syntheses and successful 0 bromination of variously substituted cyclobutanones are described in Chart B. Chart B 0 a) is commerciall¥ available from Aldrich (C-9600/ 1967-68 catalog). b}lla,16 0 c:~1) -80°C 75% <> 0 Br2 CHC1s 85% ) OBr -125- c) llc lld ' C~Br I c~ 1 C~Br 0 <(02C2Ifs + EtOOC C02Et oco2:yrolysivo 1 · H20 c~ NaOEt of corr~s . I -----io HCi ponding C02C2Ifs A acetate, . 2 xanthate, or · amine oxide ·, and subsq. oxid. 0 0 0-c~ <>c~ <> HsC CHs 0 CHs Clis c~ have also been prepared 0 0 )~/CHs 0 Bz )" /Clis ~Br +~Br 65% 35% -126- 0 0 <>· 0 0 <>Br <>Br+<)<:; Br-<>-Br Br2 cc~> Clfs Clfs Clfs Clfs Clfs trans 28% cis 18% 18% T5"% d)lle O <> ?COC!fs C~=C-Clfs OOCOC~S· cc~ ------- 600 37% 41% e)llf, llg, llh o <> <> c,H,;Li c.XH »V 0 0 CH 6 5 0 Br2 . cc~ -~ 80% -127- f)lli <?Cl )<> F C6H5 -CH =CH2 + CI:aC ==CF2 120-130° F 0 Cl C6Hs CH2 0 C02 H <> ~ C5Hs Clis The chemistry of the bicyclobutanone system should prove to be quite interesting. Many chemical reactions of bicyclobutanones can be extrapolated from the reactivity of other similar systems. Cyclopropanone reacts smoothly with methanol at -78° to afford the hemiketal, and with diazomethane to yield the cyclobutanone 12 a. The lithium aluminum hydride reduction of tetramethylcyclopropanone yields the corresponding alcohol 12b. Bicyclobutanones would be expected to react similarly. For instance, bicyclobutanone would react with diazomethane to yield the bicyclo (2 .1. O) pentanone system. Bicyclobutanes substituted in the methylene position are difficult to prepare. The active carbonyl functionality of bicyclobutanones could be modified to generate such bicyclobutanes. -128- 1, 3-dimethyl bicyclo (1.1. 0) butane reacts with methylene, ' generated photolytically from diazomethane, to give 2, 4-dimethylpenta-1, 4-diene as well as the eXpected methylene insertion into the C-H methyl bond13 . Bicyclobutanone would correspondingly be expected to yield penta-1, 4-diene-3-one. C-H insertion would likewise also be exp€cted with 2 -me thy1-bicyclobutanone. Thermally or photochemically induced decarbonylation of the bicyclobutanone system would be expected to yield substituted cyclopropenes. Pyrolysis of methylcyclopropenone at 205° produces propyne 14 . The photolysis of tetramethyl-1, 3-cyclobutanedione yields tetramethylethylene (tetramethylcyclopropanone being an intermediate product) 12 b. Reaction with nucleophilic bases would be expected to yield the cyclopropylcarboxylic acid derivative. Cyclopropene has been noted to be a reactive dienophile in DielsAlder reactions 15 . A literature search concerning the use of bicyclobutane as a Diels-Alder dienophile did not prove fruitful. The employment of the active central bond of bicyclobutanone as a dienophile would seem to be a possibility worthy of investigation--the conversion of the bicyclobutanone system to the bridged cyclobutanone system being driven by the release of strain. The ability of the bicyclobutane system to conjugate with a double bond is indicative of its unsaturated character. A comparison of the ultraviolet spectra of methyl 3-methylbicyclo (1.l.O)butane-1- -129- carboxylate with methyl cyclopropane carboxylate is quite revealing in this respect la. Whereas the latter has only a very weak absorption, the bicyclobutane ester has A. max 215 mu (E = 12, 500). Although all the carbon-carbon bonds of bicyclobutane contain 1T character, the 1, 3-bond exhibits the greatest degree of unsaturation 16 a, 16 b. Bicyclobutanones may in this context be viewed as substituted cyclopropenones. Good evidence for strong polarization of diphenylcyclopropenone. is its dipole moment of 5 .14 D. (benzophenone = 3. 0 D.) Cyclopropene was found to have a dipole moment of 0.455 ± 0. 01 D. It would be interesting to compM'e the dipole moment of the bicyclobutanone system with that of the cyclopropanone system. Hydrogenation of bicyclobutanes leads to a variety of products when different catalysts are used. In one case both monocyclic and 13 acyclic products are obtained . Since it was shown that all the y --"'4> CHaC~C~9HCHa . + Clia 63% When X is CHa, and Y is H, yield is 32% When X is H, and Y is CHa, yield is 5 % possible inonocyclic products are stable to the hydrogenation conditions used, the two molecules of hydrogen must be added without desorption -130- of the bicyclobutane from the catalyst surface. Reduction of bicyclobutanone with one mole of hydrogen would be expected to yield cyclobutanone. A literature search of the catalytic reduction of cyclobutanone was unsuccessful. Tetramethylcyclobutane-1, 3-dione is catalytically reduced to 2, 2, 4, 4-tetramethyl-3-hydroxy-cyclobutanone, and it would therefore seem reasonable that cyclobutanone (the reduction product of bicyclobutane) would be further reduced to cyclobutanol. Reduction of bicyclobutanone to either 2-butanone or butyraldehyde, corresponding to the reduction of bicyclobutane with two moles of hydrogen, cannot be predicted. The reaction of bicyclobutane with halogens can take two different courses depending on the halogen employedlb. Iodination of bicyclo- ~~ x yield x XH2 C_ + a A x a b Cl 100 0 x Br mixture b I 0 100 butanone would be expected to yield 2, 4-diiodocyclobutanone. A possible initial chlorination product might be either ketone 7 or 8. Dehy"' ""' drochlorination of ketone 1 would-yield ketone Q, or less likely, ketone ..._ 10 • COCl Cl 0 0 Av A C~Cl 7 "" Cl CH2 8 ""' 0 Cl A CHs 10 ..._ Cl -131- Bicyclobutane and most of its derivatives readily undergo hydration in dilute acid solution lb. The products of the hydration of bicyclobutane using 0. 0001 N ~S04 were found to be cyclobutanol (55%) and cyclopropylcarbinol ( 45%). The reaction of bicyclobutane1-methanol with dilute acid (£.H 4) occurs almost instantaneously and leads to 1-hydroxycyclobutanemethanol. As might be expected the addition of methanol across the central bond in 3-methylbicyclobutane1-carboxylate gives one of the two isomeric 3-methyl-3-methoxycyclobutane-1-carboxylates. The type of products formed in these solvolyses reactions and the extraordinary !ability toward acids suggests that the addition proceeds via a bicyclobutonium ion t 7a. The exact composition of the mixture obtained depends on the attacking reagent--water showing less preference than methanol in addition. l ,, G> I \ I I ' \ 0 lla """"' llb """'"" llc . -"'- -132- The exact structure of the bicyclobutonium ion would be expected to differ according to the substitution pattern of the substrate. Thus substituents which would stabilize one resonance form favor the products which are derived from that form. The solvolysis of 1-methyl cyclobutyl chloride yields only the corresponding cyclobutano1 17 b, 17 c. The tertiary carbonium ion is sufficiently good as a classical ion that little extra stability would be derived from bicyclobutonium ion formation. The cyclobutanone carbonium ion llb and methyl vinyl """"" ketone carbonium ion llc are destabili.Zed relative to the cyclopropyl""'""' carbinyl carbonium ion lli because of the contribution of ill and fu respectively. This would favor products derived from the cyclopropanone- e,b 0 lld """""' carbinyl carbonium ion lli· Products derived from the cyclobutanone carbonium ion would be more signifioant if the a-position was substituted with either a methyl or phenyl group. -133- REFERENCES 1. a) K.B. Wiberg, Rec. Chem. Prag., ~' 147 (1965). b) K.B. Wiberg, G.M. Lampman, R.P. Ciula, D.S. Connor, P. Schertler, and J. Lavanish, Tetrahedron, 21, 2749 (1965). """' 2. P.D. Bartlett and B.E. Tate, J. Am. Chem. Soc., 78, 2473 """"' (1956). 3. a) J. Meinwald and P.G. Gassman, J. Am. Chem. Soc., 82, """" 2857 (1960}. b) K.B. Wiberg, B.R. Lowry, and T.H. Colby, J. Am. Chem. Soc., _§1, 3998 (1961). 4. a) W.J. Brown, Ph.D. Thesis, Columbia University, 1958. b) E. Wenkert and D.P. Strike, J. Org. Chem., U, 1883 (1962). c) F. Nerdel, D. Frank, and H. Marschall, Angew. Chem. Int. Ed., 1, 457 (1962). - ""' d} K. B. Wiberg and G. W. Klein, Tetrahedron Letters, 1043 (1963}. 5. a}B.C.L. Weedon, Advan. Org. Chem. Methods Results, 1, 1(1960}. b) G. W. Thiessen, Record Chem. Progr., g, 243 (1960). 6. 7. 8. A.F. Vellturo and G.W. Griffin, J. Org. Chem., 31, 2241 {1966). """"' N. Dinh-Nguyen, Acta Chem. Scand., 12, 585 (1958). """' a)J.K.K. Wilshire, AustralianJ. Chem.,~' 432(1963). b) A.I. Levin, O.N. Chechina, and S.V. Sokotov, Zh. Prild. Khim., 38, 1778 (1965). """"' c) 0. N. Chechina, A.I. Levin, and S. V. Sokotov, Zh. Prild. Khim., 39, 2120 (1966). """"' -134- d) K. Koyama, K. Yoshida, and S. Tsutsami, Bull. Chem. Soc. - - Japan, ~' 516 (1966). e) L. Rand and A.F. Hohar, J. Org. Chem.,~' 3885 (1965). 9. a) H. Hofer, Ber., 33, 655 (1900). b) F. Fichter and S. Lurie, Helv. Chim Acta, 16, 885 (1933). """" c)E. BuchtaandF. Burger, Ann., 580, 125(1953). -- """"" 10. a) K. B. Wiberg and G. M. Lampman, Tetrahedron Letters, 2173 (1963). b) K. Griesbaum, W. Naegle, and G. G. Wanless, J. Am. Chem. 87, 3151 (1965). -Soc., .......,..., c) M.R. Rifi, J. Am. Chem. Soc., 89, 4442 (1967). """" 11. a) J.M. Conia and P. Leriverend, Compt rend., ~' 1078 ( 1960). b) J.M. Conia and J. L. Ripoll, Compt. rend., ~.b 1071 (1961). c) J.M. Conia, P. Lerireend, and J. L. Ripoll, Compt rend., ill' 1803 (1961). d) J.M. Conia, J. Gore, J. Salaun, and L. Ripoll, Bull. Soc. Chim. France, 1976 (1964). e) I. V. Machinskaya, G. P . Smirnova, and V. A. Borkhosh, Zh. Obshch. Khim., U, 2563 (1961). f) A. Burger and R. Bennett, J. Med. and Pharm. Chem., ...... 2, 687 (1960). g) C. Beard and A. Burger, J. Org. Chem.,~' 2335 (1961). h)A. BurgerandH.H. Org, J. Org. Chem.,~ 2588(1964). -135- i) E. F. Silversmith, Y. Kitahara, M. C. Caserio, and J.D. Roberts, J. Am. Chem. Soc., 80, 5840 (1958). """"' 12. a) N.J. Turro and W.B. Hammond, J. Am. Chem. Soc., 88, """" 3672 (1966). b) N.J. Turro, W.B. Hammond, and P.A. Leermaker, J. Am. Chem. Soc., ll, 2774 (1965). 13. W. von E. Doering and J. F. Coburn, Jr., Tetrahedron Letters, 991 (1965). 14. R. Breslow and L. Altman, J. Am. Chem. Soc., 88,- 504 (1966). 15. G.L. ClossandD.K. Krantz, J. Org. Chem., g, 638(1966). """"' 16. a) M. Pomerantz and E.W. Abrahamson, J. Am. Chem. Soc., ,_ 88, 3970 (1966). b) M. Pomerantz, J. Am. Chem. Soc., 88, 5349 (1966). ""' 17. a)J.D. Roberts, et.~ J. Am. Chem. Soc., fil.., 4390(1959). b) H.C. Brown and M. Barkowski, J. Am. Chem. Soc., 74, 1894 """"' (1952). c) E.F. Cox, M.C. Caserio, M.S. Silver, and J.D. Roberts, J. Am. Chem. Soc., 83, 2719 (1961). """"' -136PROPOSITION IV The chroman ring system1a ...... 1,- shown in Chart A, is found in a large group of plant products. Many of the plant and flower pigments, as well as some of the colored components produced by molds, are derived from this parent ring system. Coqmarinslb ~ and chromones 1c 3a are the most important divisions of this system • .......... Flavones le (i.e., 2-phenyl chromones) 3b, and isoflavones le (i.e., """" 3-phenyl chromones) ..3c, are widely distributed in plants as glycosides. A number of physiological, highly active substances possessing the a-pyrone 1d nucleus 4 (a-pyrones can be viewed as nor-benzo"' coumarins) have been isolated.- Substituted y-pyrones 1d .§._, which can be viewed as nor-benzochromones, are widespread throughout nature. The pyrans (a-pyran 1d 6 and y-pyran 1d 7) also offer a synthetic "' "' challenge, being enol ethers. Dihydropyran1d 8 has been employed "' as a model in carbohydrate chemistry. Chart A 5 0 4 o:x~· 0 1 ...... 1 Rl ~~, R 1 =~=H 2 ~' R 1 = C6 H5 , ~ = H 0 0 0 8 .£, R1 =H, ~=CG~ a () 0 ·O 0 4 0 0 5 6 "' 7 ....... -137- It is proposed to synthesize representatives of these classes of comp.ounds by an enamine technique 2a, 2b, 2 c as shown below in Chart B. Chart B + \/ \/ N N )- -i I \I b) various com- - - pounds in the series N N 111 + cI R1 ) I \I \/ I N ) 6 I I •.- N R1 ) 0 R1 I _, # = ~ 0 0 R1 c) 0 )~ R1 ~ + > -0.:--> ~R1 N/ \ I ~ ! various comseries > pou~ds in the N.....\ . -138- d) I n AN: AA N Ill + cI R1 I I ~---): I I N\ R1 ,_ I e) 0 ~x C( ~ QY N Ill I (( X =Cl 0 II Y = CRs,H OY + I I i _l v . --1 Q > -139- The reactions to be cited, exhibiting high yields under mild conditions and incorporating nucleophilic enamine attack on various carbonyl and nitrile groupings, 1, 4-cycloadditions, 0-alkylation attack, amine elimination, and ease of aromatization, offer justification for the proposed routes delineated in Chart B by both flow and static methods. Danishefsky 3a, aware that derivatives of j3-vinyl acrylates 3 are well known to undergo 1, 6 addition preferably b, studied the reaction of the A -pyrrolidinocyclohexene (9) with ~thyl trans-2, 41 "' pentadienoate (10) ,,_ as shown below in Chart C. In refluxing tetrahydrofuran a 1:1 molar ratio of (9) and (10) after 17 hours gave in . "' ,,_ Chart C 0 C02C2Hs 0 +) 9 "' 10 """' C02C2lis ) 11 ""' 12 ""'""' l 13 """' C02C2H5 -140- good yield a mixture composed of amino ester 11 and dienic ester 12. """""' """" Under more vigorous conditions cycloaddition, retro-Michael elimination, and aromatization were realized, thus affording a one-step synthesis of aromatic ester 13 in 58% average yield. The authors note """""' that the yellow color observed on admixture of the reactants is reminiscent of many types of Diels-Alder reactions. Berchtold3c, who performed the study with a number of similar enamines, found that the enamines reacted with methyl trans-2, 4-pentadienoate to give intermediate 1: 1 adducts arising from 1, 4-cycloaddition. These adducts were generally found to readily lose the elements of pyrrolidine or morpholine under the reaction conditions affording 3, 4 dihydrobenzoate derivatives. The reaction has been shown to be general for pyrrolidine enamines of both cyclic ketones and simple aldehydes. Paquette has reported 4 an interesting and germane zanthone synthesis which involves an enamine-initiated attack on an aldehydic group to form the alkoxide anion and 0-alkylation of the resultant iminium salt as shown in Chart D on the following page. When equimolar quantities of 2-hydroxy-1-napthaldehyde and 1-morphilinocyclohexene were allowed to stand in benzene solution overnight at room temperature the crystalline adduct 14 was obtained in 91. 5% """""' yield. The facile cyclization of enamines containing a nitrile group -141- Chart D 0 \I CH p ~1 o:c (N) 0 oe I c ) OHh C) OH ~p-~ e C) 0 ~ Pd/C OJ() -1-0~-%~ 14. ,,..,,,... Cr03 /pyr. 75% ) -142in the presence of magnesium has been reported to produce amino5 5 dihydropyridinium salts in good yield a, b. N Ill c:D 1. M~ > xe f R The 1, 4-cycloaddition of l-diethylamino-1, 3-butadiene with electrophilic olefins to produce 4-substituted-3-diethylamino cyclohexenes6 is described in Chart E. Yields are excellent and the experimental conditions quite mild. Chart E N(C2lfs}2 >- ()x yield(%) > ax deamination 94 (cis) not reported 93 no deamination noted 42 +some despontaneously deamin.amination product ates upon standing CHO 84 deaminates quite. easily -143- Hamer, in his instructive book "1, 4-Cycloadditions- -The Diels-Alder Reaction in Heterocyclic Synthesis117 , notes that the mechanism of these 1, 4-cycloaddition reactions remain a matter of debate. A number of interesting observations have been reported concerning the cycloaddition reaction of butadiene and substituted cyanides. The reader is referred to the chapter written by G. J. Janz 7 who also performed most of the experiments reported therein. The cycloaddition products obtained in the gas flow system are invariably the completely aromatized ring system (i.e., pyridines) rather than the primary diene addition products (i.e., dihydro- pyridines). Experimental conditions and yields for a variety of cyanides are reported in Chart F on the following page. These results indicate that although cyanides do undergo cycloaddition reactions with butadiene, only cyanides which contain polar groups adjacent to the dienophile able to enhance the polarizability .of the reaction site are efficiently cyclized. The previous studies cited, relevant to the proposal described in Chart B, resulted in good yields under mild reaction conditions upon reaction of a variety of activated and non-activated carbonyl and nitrile groupings with various nucleophilic monoenes and dienes (i.e., the enamines). These studies suggest that the employment of electron-rich l-diethylamino-1, 3-butadiene will be effectuated under mild conditions (i.e., low temperature, short reaction, nonactivated dienophiles, etc.) in a static system. Repeating both the catalyzed and non-catalyzed type of gas flow system experiments -144- Chart F ( + +~ RI c () If I N N catalyst R temp. ( °C) contact time (hr.) yield of pyr. pdt. (%) 1.0 R system H homogen. gas phase, rxn none heterogen. gas not noted phase rx n 404 15 x 10~ 404 4.2 x 10- 1.01 none CN homogeneous heterogeneous not noted 480 400 2 x 10-3 6.3 x 10-2 34 6.7 395 408 12 x 10:: 6 x 10 0.1 3.6 c~ none 2 homogeneous heterogeneous A~03 C6 H5 homogeneous heterogeneous A~03 393 400 7. 5 x 10~ 6. 5 x 10 1.4 27 CF3 homogeneous none 400 44 x 10-2 99 none mentioned above with the various reactions shown in Chart B should also prove quite profitable. The dienamide technique could also be profitably applied with nitroso and azo dienophiles. The formation of 4-membered rings is not competitive with the . formation of 6-membered rings, especially if the reaction is considered to be initiated by nucleophilic attack and therefore full~ -145- reversible. The corresponding 1, 2-cycloaddition reactions, yielding an oxetane 8 by scheme Ba), substituted rare 1- or 2-azetines 8 by scheme Bb) and Bd), or oxeta~e 8 by scheme Be), would not there- \/" J=l N scheme Ba) - R1 ~ an oxetane Iii scheme.Eb} R1 N< T-) 1-azetine azete lT N 2-azetine scheme Be) R , . J = f . Ri N<___ ~ P ~ an oxete -146- ·scheme Bd) dN<_~ 11 J f N' fore seem to be competitive with the formation of six-membered rings obtained by 1, 4-cycloadditions. Appropriate R1 and~ groupings in the nitrile and carbonyl starting materials shown in Chart B would seem to be H, Clfs, C6 H5 , and CN. Based on the observations already discussed it is felt that strong electron attracting groupings would not be necessary for satisfactory conversions. The presence of the cyanide group offers a good "handle" in the sequence as it can easily be converted to various functional groupings. Employment of various substituted crotonaldehydes also increase the generality of this synthetic approach. The use of phenyl methyl ketone and phenyl acetaldehyde as the enamine precursor in scheme Be) would result in the formation of the flavone and isoflavone systems, respectively. The initial products of scheme Bb) and Bd) would be the di.hydropyridines shown. Thermodynamic considerations indicate that spontaneous dehydrogenation of the dihydro adduct to the pyridine -147- adduct would be expected and this has been shown to be the case in the flow system studies 7 at 400°. Isolation of the amino dihydropyridines and subsequent hydrolysis would yield the trihydro9 and dihydropyridones shown in Chart B. Conversion of the enamines formed in schemes Ba) and Be) into the corresponding iminium salts and subsequent nucleophilic attack2b, 2 c by hydride, cyanide, and grignard reagents would result in functionalization of the ring carbon atom attached to the nitrogen atom. The a- and y-pyrones ,,.... 4 and ,,.... 5 can be prepared from the enamine products of schemes Ba) and Be) by first hydrolysis and then subsequent N-bromo succinimide 11 brominations and dehydrobrominations. Hydrolysis of the enamine formed in scheme Ba) would yield the lactone whose reduction and subsequent dehydration would form the dihydropyran 8. Hydrolysis could also yield a hydroxy .dihydropy-,, ""' ran 11 . Hydrolysis of the imine prepared in scheme Be) would yield the desired chromone. The cyanide grouping (i.e., R 1 or Ra in scheme Ba) or Be)) is easily removed by first hydrolysis and then lead tetraacetate oxidative decarboxylation 12 , thereby also introducing an additional double bond. -148REFERENCES 1. a) S. Wawzonek in (R. C. Elderfield, ed.) "Heterocyclic Compounds", ~' J. Wiley and Sons, Inc., London, 1951, p. 393. b) S. Wawzonek in (R. C. Elderfield, ed.) "Heterocyclic Compounds'', ~' J. Wiley and Sons, Inc. , London, 19 51, p. 173. c) S. Wawzonek in (R. C. Elderfield, ed.) "Heterocyclic Compounds", ....... 2, J. Wiley and Sons, Inc., London, 1951, p. 229. d) S. Fried in (R. C. Elderfield, ed.) "Heterocyclic Compounds", 1, J. Wiley and Sons, Inc., London, 1950, p. 343 . ....... 2. a) G. Stork, A. Brizzolara, H. Landesman, J. Szmuszkovicz, and R. Terrell, J. Am. Chem. Soc., ............. 85, 207 (1963) . b) S. Szmuszkovicz in (R.A. Raphael, E.C. Taylor, H. Wynberg, eds.) "Advan. in Org. Chem. Methods and Results", 4, Inter"" science Publishers, New York, 1963, p. 1. c) K. Blaha and Otaker Cervinkajn in (A. R. Katritzky and A. J. Boulton, eds.) "Advan. in Heterocycl. Chem.", ....... 6, Academic Press, New York, 1966, p. 147. 3. a) S. Danishefsky and R. Cunningham, J. Org. Chem., ~' 3676 (1965). 48, b) E. P. Kohler and F. R. Butler, J. Am. Chem. Soc., .....,.. 1041 (1926). -149- c) G.A. Berchtold, J. Ciabattoni, and A.A. Tunick, J. Org. Chem., 30, 3679 (1965). """" 4. 5. L.A. Paquette, Tetrahedron Letters, .1291 (1965). a) A.I. Meyers, A.C. Kovelesky, and S. Singh, J. Heterocycl. Chem., ....... 4, 461 (1967) . b) A. I. Meyers, A. C. Kovelesky, and S. Singh, Abstract, 155th National Meeting of the Amer. Chem. Soc., San Francisco, 1968. 6. S. Htinig and H. Kahanek, Chem. Ber., 90, 238 (1957). 7. J. Hamer, "1, 4-Cycloadditions--The Diels-Alder Reaction in """" Heterocyclic Synthesis", Academic Press, New York, 1967. 8. Y. Etienne, N. Fischer, J.A. Moore, and S. Searles, Jr., in (A. Weissberger, ed.) "Heterocyclic Compounds with Three- and Four-Membered Rings--Part II", Interscience Publishers, New York, 1964. 9. E. Klingsberg in (R.A. Barnes, ed.) "Pyridine and Its Derivatives--Part r', Interscience Publishers Inc., New York, 1960. 10. L. Horner and E.H. Winkelmann in (W. Foerst, ed.) "Newer Methods of Preparative Organic Chemistry", ,,.... 3, Academic Press, New York, 1964, p. 1. 11. M. von Strandtmann, M. P. Cohen, and J. Shavel, Jr., Tetrahedron Letters, 3103 (1965). 12. R. Criegee in (K. B. Wiberg, ed.) "Oxidation in Organic Chemistry--Part A", Academic ~ress, New York, 1965, p. 293. -150- PROPOSITION V It is proposed to generate the benzyne-tricarbonyl-chromium complex (2) from the suitably substituted cyclopentadienyl benzene"" tricarbonyl-chromium(O) complex -"" 1 as shown below. The complex would be expected to exhibit a greater degree of stability than the benzyne molecule which is generated as a reactive intermediate by other techniques. base 0 c Cr /I' c c 0 ,.......er, C I C 0 c 0 0 0 1 2 Salts of tropylium-tricarbonyl-chromium(O) (3) are prepared by -"" hydride abstraction from cycloheptatriene-tricarbonyl-chromium(0) 1 ' 2a. This method was previously employed for the preparation of tropyliumtricarbonyl-molybdenum(O) salts 3 . These salts undergo two types of reactions with nucleophiles depending on the character of the latter 2 a, 2b. With nucleophiles such as methoxide or phenyl anion, the products are -151substituted (probably exo ) cycloheptatriene-tricarbonyl-chromium(O) complexes. When the salt is added to excess sodium cyclopentadienide or sodiomalonic ester, a remarkable rearrangement occurs which leads to benzene-tricarbonyl-chromium{O) (4). The most favorable reaction "" conditions result in a 46% yield of the complex 4. "' + Q I ) I 0 c 0 c ,.....er, I c c 0 0 3 4 "' Munro and Pauson2 a performed a number of experiments in order to establish whether the benzene ring attached to the chromium in the product was derived entirely from the seven-membered ring or in part from the cyclopentadiene ring. A decision in favor of the former view was reached qualitatively by repeating the reaction first with methylcyclopentadienylsodium, whereupon benzene-tricarbonylchromium(O) (4) was the only chromium complex isolated. Then, with ...... -- methyltropylium-tricarbonyl-chromium, toluene-tricarbonyl-chromium(O) -152- was obtained. The conclusion that it is the seven-membered ring which produces the benzene complex by ring contraction (and thus, presumably, the five-membered ring which yields the free benzene molecule by ring expansion) was corroborated by labeling the tropyllum complex ...... 3 with tritium. The derived product ...... 4 was shown to contain the expected tritium count. The same ring contraction was again observed when the tropylium complex ...... 3 was treated with an excess of diethylsodiomalonate . The simultaneous isolation of the malonate adduct ...... 5 (R = H) suggested . that the "normal" anion addition is an intermediate step in these rearrangements. This is confirmed by showing that ester ......5 (R = H) is transformed into benzene complex 4 by the further action of sodiomal"" onate or by other bases such as sodium methoxide. The corresponding intermediate ...... 6 (R = H) in the cyclopentadienide reaction was easily isolated when an excess of tropylium complex ...... 3 was used and was I I ' 0 c ,.......er, I c Cr c / I 'c C 0 0 c 0 0 5 6 ...... "" 0 -153- similarly transformed into benzene complex 4 by methoxide ion. "" Complex 6 (R = H) had a six degree melting range and was assumed to be a mixture of stereo-isomers. That the methylmalonate complex ~ (R = Cfls) fails to rearrange indicates that ionization of the acidic hydrogen (i.e., R in 5 and 6) induced by the basic catalyst is an essen"" "' tial step in the ring contraction. Munro and Pauson do not comment upon the fact that the rearrangement proceeds all the way to benzene complex 4 with sodium methylcyclopentadienide. "" The above observations lead Munro and Pauson to propose that the electronic shifts indicated below represent "a possible mechanistic picture for this ring contraction". On this basis the "C6 lfa" fragment formed in the formation of complex 4 would be fulvene. "' Fulvene has not been isolated in this reaction and would not be expected to survive · > ) I . . . . er C I 'c 0 c 0 0 I 0 Cr c "I 'c c 0 7 "' 0 i H + 'c=c~ 0 -154the reaction conditions. Munro and Pauson note in their 1961 article 2 a that attempts are in progress to find a situation where the correspondjng product may be isolated. A survey of the subsequent literature indicates that a report of these attempts has not yet been published. That formulation ...... 7 is at best a first approximation of the actual mechanism is evident since it neglects the function of the tricarbonyl chromium grouping. Its importance to the reaction is, however, evident as cyclopentadienylcycloheptatriene is unaffected by sodium methoxide under the conditions employed for rearrangement of its complex ...... 6 (R = H) . The synthesis of complex ...... 1, X being a suitable leaving group, can be effected in various ways. An efficient approach would involve the reaction of ortho- and alpha-disubstituted toluenes with an equivalent amount (the equivalency determined either by salt isolation or employment of a standard solution) of sodium cyclopentadienide as indicated below. <£) ) + Na y 8 ...... 9 ...... Disubstituted toluenes with X being halide, tosylate, or acetate and Y being halide or tosylate would be suitable substrates. Representative examples of innumerable syntheses of these compounds present in the -155- literature are shown in Chart I. Chart I cc X = F, y =Cl 4a y 8 .....-. Br Br4b Cl Br4c Cl OH4d Br OH4d The complex 1 can be formed by either causing the cyclopenta" dienyl-benzene grouping to selectively react at the triene site with chromium hexacarbonyl or by preforming the complex with the reaction of chromium hexacarbonyl and the disubstituted toluene. Many monoand di-substituted benzene complexes, such as chlorobenzene-tricarbonylchromium( 0) 5a, phenol-tricarbonyl-chromium( O) 5b, benzyl alcoholtricarbonyl-chromium( 0) 5a, toluene-tricarbonyl-chromium(0} 5a, pheny!acetate -tricarbony1-chro mium( O) 5c, ortho-toluidine-tricar bony lchromium( 0) 5a, ortho-xylene-tricarbonyl-chromium(0} 5a, and metamethoxybenzoic acid-tricarbonyl-chromium(0} 5d, have been prepared in yields from 2 0 to 90% by causing the substituted benzene to .react with chromium hexacarbonyl. Chromium tricarbonyl, requiring the donation of six electrons to fill the valence shell of chromium, will preferentially complex with -156- benzene rather than cyclopentadiene. Cyclo-octa-1, 5-diene affords the yellow complex [ Cr(C0) 4 (C 8 H12 )] in only 2% yield after being refluxed for two day.s with. a two-~f old excess of chromium hexacarbonyl 6a. Cais isolated 1-(phenyl-chromium tricarbonyl)-4-phenylbutadiene (60% yield) and a small amount of 1, 4-di-(phenyl-chromium tricarbonyl} butadiene when 1, 4-diphenylbutadiene was caused to react with chromium hexacarbony1 6b. The mono-complex could be converted to the dicomplex in over 50% yield, but butadiene-complex formation was not noted. Iron pentacarbonyl, on the other hand, was shown to effectively complex with the butadiene moiety. No literature report of the reaction . of chromium hexacarbonyl with cyclopentadiene to yield the dienecarbonyl complex was found. Cyclopentadiene will react with chromium hexacarbonyl under forcing conditions (i.e., 280-350° hot tube vapor phase experiment) to form bis-cyclopentadienylchromium(II) in 30% . ld6c . y1e Abstraction of the acidic cyclopentadienic proton in complex ....... 1 by base would be expected to readily occur based on the cycloheptatrienic analogy. Subsequent elimination of the leaving group X would result in the formation of the desired benzyne complex ....... 2 and fulvene . The most reasonable proposals for 1T-type bonding between a benzenoid compound and a transition metal involve donation of electrons to the metal - - the donated electrons allowing the metal to achieve the rare gas configuration. This direction of electronic transfer creates an electron deficienty on the arene an_d leaves it less prone to electrophilic attack than the free arene. These arene complexes, on the other hand, -157- undergo nucleophilic substitution much more readily than the corresponding arene. The facile reaction of chlorobenzene-tricarbonylchromium(O) with sodium methoxide to yield anisole-tricarbonylchromium5a demonstrates the enhanced reactivity of the tr-bonded benzenoid nucleus towards nucleophiles. This enhanced reactivity should aid the proposed nucleophilic-type reaction required for the ejection of the leaving group. The substitution of the malonate moiety for the cyclopentadienyl portion of complex ,.., 1 would also result in the formation of an effective precursor for the formation of benzyne complex 2. "' It would be quite interesting to investigate the possibility of employing an anthranilic acid-tricarbonyl-chromium type compound as the benzyne complex precursor. Diazotization of the anthranilic acidtricarbonyl-chromium complex would present a problem as arenetricarbonyl-metal(O) complexes are, in general, unstable to strong acid. Benzenediazonium-2 -carboxy late-tricarbony1-chromium, prepared by the reaction of preformed benzenediazonium-2-carboxylate with chromium hexacarbonyl, would serve as a suitable substrate. Benzothiadiazole-1, 1-dioxide, which has been employed to generate benzyne quite efficiently under very mild conditions 7, would also serve as an interesting substrate for benzyne complex formation. Huisgen has shown with the aid of a series of Diels-Alder reactions 8 that the benzyne liberated by metals from benzene halides (which may exist as either the free alkyne or as a complex with a metal cation or anionic halide) and the intermediate generated by -158- thermolysis of benzenediazonium-2-carboxylate are identical. The benzyne complex 2 could therefore be expected to generate the same ""' intermediate. It would be quite difficult to predict the relative stability of benzyne complex 2. Benzyne 9a, 9b, although easy to generate, is quite ""' unstable and has therefore been characterized only by the adducts it for ms with other compounds and through labeling experiments. The only direct physical measurement of benzyne is the transcient absorption spectrum obtained by, flash photolysis of the diazonium 10 salt . Numerous spectral investigations of arene-tricarbonyl-chromium complexes have been reported. A number of pronounced changes were found to occur in the infrared region upon complexinglla. These include shifting of the 1600 and 1500 bands, generally attributed to skeletal C-C stretching modes, from their normal position by 25-100 cm - 1 and 30-40 cm -l respectively toward lower energy. These changes 1 also include the presence of two intense bands at 1990 and 1920 cm- - indicative of the tricarbonyl moiety. , The carbonyl frequency is dependent on the extent of back-donation from the metal, carbonyl frequencies shifting to lower values when the electron density on the aromatic nucleus is increased. In general, three ultraviolet absorptions are quotedllb in the regions 218-221, 251-264 (shoulder), and 315-324 mµ. with log E of 4.4, 3.8, and 3.2 respectively. No efforts have been made to interpret these transitions, although from analogies to other carbonyl-metal complexes it appears that the absorptions below 300 mµ. are due to the -159carbonyl groupings. The ring protons in the nuclear magnetic resonance spectrumllc appear ~t higher field strength than in free benzenoid compounds. This is because the aromatic ring current is perturbed by 7T-bond formation to the metal. The benzyne complex 2, if relatively unstable, could be trapped "' in the normal manner by compounds such as furan, anthracene, or carboxylic acids 9a, 9b, or nucleophiles such as thiophenoxide, anisole, anilide, triphenylmethide, or n-butide (from n-butyl lithium) 9a, 9b. - - Heating of arene complexes with an excess of mesitylene results in arene exchange (i.e., formation of free arene and the corresponding 2 mesitylene complex) a. Distillation of toluene-tricarbonyl-chromium(O) 5 at 200° resulted in the isolation of toluene in 76% yield a. These observations would seem to indicate that the benzyne complex ..... 1, if reasonably stable, could be converted to benzyne upon gentle heating in a relative stable solvent such as benzene 13 or the alkylated benzenes. The bonding system of benzyne, which shows no reactions characteristic of radicals, is not well understood. But it is felt that the aromatic character of the cyclic system is reasonably undisturbed9b. Models of the electronic structure show that the ring strain is not excessively high and can be compared with that of the cyclopropene ring9b. It has been generally accepted9a that the two non-hydrogen 2 carbon atoms maintain their sp hybridization as in benzene; and that the electron pair of interest is resident in the two sp2 orbitals ordinarily used for bonding with hydrogen or substituents. These orbitals are thought to have their axes in the plane of the ring and overlap sideways -160to some extent, creating a weak bond. Recently published calculations and observations by Hoffmann 12 also indicates that there is some overlap between the two half-occupied lone pairs. He considers these lobes as somewhat delocalized and quite different from sp2 hybridization. Additional stabilization of the benzyne complex 2 might be achieved ~ by tilting of the benzyne plane such that carbon atoms 1 and 2 would be brought closer to the chromium atom than atoms 4 and 5. This would result in a slight destabilization due to the decreased electron donation of the C3 -C4 and C5 -C6 double bonds to chromium, but a greater stabilization of the "triple bond" (i.e., the unstable portion of the molecule) due to chromium back donation. Tilting the molecule such that carbon atoms 2 and 3 will be closer to the chromium atom will result in only a single double bond, the C5 -C6 bond, being further removed from chromium. The degree of extra stabilization would probably not be such that a single carbonyl group would be lost from complex 2. ~ The question of self condensation of complex 2 involving "triple bond" 7T-bonding with elimination of a single molecule of carbon monoxide is an intriguing one to consider. Stable complexes of formula l Arene Cr( C0) 2 RC =cR] have been obtained by irradiation of solutions of hexamethylbenzene-tricarbonyl-chromium(O) or mesitylene-tricarbonyl-chromium(O) and diethylacetylenedicarboxylate or diphenylacetylene14. The complex prepared by irradiation of a solution of benzene-tricarbonyl-chromium(O) and diphenyl acetylene was also obtained. The yields varied from 30-55%. The performance of this -161- reaction Arene Cr(C0) 3 + R-C=C-R~ Arene Cr(C0) 2 R-C=C-R +CO by a thermal route has not been reported. It is therefore possible but unlikely that the benzyne complex ....... 2 would. be converted wholly or in part to polymeric complex """ 10 under the experimental conditions em. ployed. It would be of interest to attemp.t to prepare complex .............. 11 by reaction of complex ....... 2 with chromium hexacarbonyl. The "triple bond" would then, of course, not be involved in stabilization with the tricarbonyl chromium grouping in complex ll· Benzyne should also be o. I " .... .... Q co .... \ Cr(C0) 5 I Cr Cr C/ I 'C 0 c C / I 'c 0 0 ... ...... 0 co ·cr I .,_ co . . . . . . . . , fi·, I c 0 .... 0 ~ ~ 10 .............. 11 .............. 0 -162- generated from either benzenediazonium-2-carboxylate or benzothiadiazole-1, 1-dioxide in the presence of hexacarbonyl chromium--two different possibly competitive complexing sites, the benzene ring and the "triple bond", being available. -163- REFERENCES 1. H. Zeiss, P.J. Wheatley, and H.J.S. Winkler, "Benzenoid Metal Complexes", Ronald Press, New York, 1966. 2. a) J.D. Munro and P.L. Pauson, J. Chem. Soc., 3475 and 3479 (1961). b) J.D. Munro and P.L. Pauson, Proc. Chem. Soc., 267 (1959). 3. H.J. Dauben and L.R. Hohnen, J. Am. Chem. Soc., 80, 5570 """"" (1958). 4. a)S. SchiemannandH.G. Baumgarten, Ber., 70, 1416(1937). - """" b) J. Kennerand and J. Wilson, J. Chem. Soc., 1168 (1927). c) M. Haring, Helv. Chim. Acta, 43, 104 (1960). """" d)J.B. ShoesmithandR.H. Slater, J. Chem. Soc., 214(1926). 5. a) B. Nicholls and M.C. Whiting, J. Chem. Soc., 551 (1959). b) G. Natta, F. Calderazzo, and E. Santambrogio, Chim. Ind. (Milan), 1Q, 1003 (1958). c) B. Nicholls and M.C. Whiting, Proc. Chem. Soc., 152 (1952). d) A. Mandelbaum, Z. Neuwith, and M. Cais, Inorg. Chem., ~' 902 (1963). 6. a) E.O. Fischer and W. Frohlich, Chem. Ber., 92, 2995 (1959). """ b) M. Cais and M. Feldkimel, Tetrahedron Letters, 444 (1961). c) G. Wilkinson, J. Am. Chem. Soc., 76, 209 (1954). ""' 7. G. WittigandR.W. Hoffmann, Chem. Ber., 95, 2718(1962). 8. R. Huisgen and R. Knorr, Tetrahedron Letters, 1017 (1963). 9. """"' a) J. Bunnett, J. Chem. Ed., 38, 278 (1961). """"" -164- b) H. Heaney, Chem. Rev., .......,.,, 62, 81 (1962) . 10. R.S. Berry, G.N. Spokes, andM. Stiles, J. Am. Chem. Soc., 84, 3570 (1962). """' 11. a) R.E. Humphrey, Spectrochim. Acta, D., 93 (1961). b) R. T. Lindquist and M. Cais, J. Org. Chem., ll' 1167 (1962). c) W. Strohmeier and H. Hellmann, Chem. Ber., 97, 1877 (1964). """' 12. R. Hoffmann, A. Imamura, and W.J. Hehre, J. Am. Chem. Soc., 90, ...._ 1499 (1968) . 13. R. G. Miller and M. Stiles, J. Am. Chem. Soc., ~' 1798 (1963). 14. W. Strohmeier and H. Hellmann, Chem. Ber., ........... 98, 1598 (1965) •