PART I THE MECHAMSM OF THE LEQUED PHASE PYROLYSIS QF DIALKYL QXALATES PAR? {é NUCLEAR MAGNETiC EESQNANCE STUDIES OF N-RfiTE-WLPHENYLHYDRAZQNES Thosts 50:0 “to Degree (:5 pk. D. MICHIGAN STATE UKEVERSITY Karl L. Krumei 1965 mm sun cum ‘ LIBRARY III! ’ 'Wtain TH Lb. MOCN'BAN ' Mich' to 13:11 Sta 5916 Universaty > O (3,) ABSTRACT PART I THE MECHANISM OF THE LIQUID PHASE PYROLYSIS OF DIALKYL OXALATES PART II_ NUCLEAR MAGNETIC RESONANCE STUDIES OF N-METHYLPHENYLHYDRAZONES by Karl L. Krumel The purpose of this investigation was to study in greater detail the mechanism of the liquid phase pyrolysis of dialkyl oxalates that was originally suggested to proceed through an ion pair inter— mediate (l). The oxalates of tertiary alcohols decomposed smoothly at 140— 1700 affording excellent yields of alkenes while those of primary or secondary alcohols decomposed only after prolonged heating at 300—3600. Systems prone to carbonium ion rearrangements, 3,3. bis—(2- cyclohexyl-Z-propyl) oxalate and bis—(2,3,3-trimethyl-2—buty1— 1,1,1-d3) oxalate, underwent approximately 10% of 1,2-hydride or 1,2-a1kyl shift. Bis-(erythro-l,2-diphenylethyl-2-d1) oxalate and bis—(Egggg- 1,2-dipheny1ethy1-2-d1) oxalate decomposed stereospecifically to give 90% of trans-stilbene- secondary > primary (1), is in accord with the suggested mechanism. The thermal decomposition of alkyl halides has also been interpreted as proceeding through a concerted cyclic mechanism in which the products are the thermodynamically more stable alkenes arising from Saytzeff elimination (1,11). R2? _x R29 """ it The relative rates of elimination, tertiary::>’ secondary ::> primary, are much larger than those encountered for the acetate pyrolysis and may be rationalized by suggesting more charge separation at the transition state. An isotope effect of kH/kD = 2.2 at 5000 strongly suggests that cleavage of the carbon- 5.—hydrogen bond is extensive in the transition state (12). The predominant Saytzeff elimination may be explained by isomerization of the initial alkenes by the hydrogen halide formed in the homo— geneous vapor phase reaction. Several other thermal elimination reactions yield alkenes whose formation is better explained by invoking a carbonium ion mechanism. The pyrolysis of borate esters has been shown to give excellent yields of alkenes at temperatures of 250-2900 (13,14). Chapman investigated the structural integrity of the reaction by studying systems which are prone to carbonium ion rearrangements. The pyrolysis of the borates of 3,3-dimethyl—2-butanol and cyclohexylcarbinol are examples of what is observed. 3““ (CH3)3C-CH=CH2 (CH3)2CH-C=CH2 (CH3)2C=C(CH3)2 Borate r OH Pyrolysis 0.3% 26% 73.7% I Acetate k Pyrolysis 100% CH CH Borate \ CHZOH PyrolysiSTT' 6% 82% 12% [:::]/ Acetate k Pyrolysisgr_ 100% Chapman concluded that the reaction proceeds by ionization of the borate ester to either an ion pair or to more completely separated ions. The olefins are then derived from carbonium ions sufficiently free to rearrange. The dehydration of tertiary aliphatic alcohols in dimethyl sulfoxide at 160-1850 provides alkenes in good yield (15). The isomeric distribution arises from predominantly Saytzeff elimination and a carbonium ion mechanism is suggested. The actual mechanism is not clear since alcohols treated under the same conditions in the absence of dimethyl sulfoxide do not react. The elimination may be due to acid impurities in the dimethyl sulfoxide. DMSO CH3CH2CCCH3)20H ZS: >— CH3CH=C(CH3)2 + CH3CH2C(CH3)=CH2 59% 41% The first apparent study of the liquid phase pyrolysis of dialkyl oxalates was performed by Karabatsos and co-workers (16). They studied several oxalate esters and made the following observa- tions: (1) esters derived from tertiary aliphatic alcohols decompose smoothly at 140-1700 affording good yields of alkenes in which the thermodynamically more stable isomers resulting from Saytzeff elimination predominate; (2) esters derived from primary and secondary acyclic alcohols do not decompose upon heating to 3250; (3) esters derived from secondary alicyclic alcohols decompose at 250—3000 yielding the expected alkenes and the corresponding formate esters; (4) the composition of the alkene mixture is not affected by decomposing the ester in the presence of one equivalent of oxalic acid; and (5) no carbon skeletal rearrangements were observed in the decomposition of tertiary dialkyl oxalates. They concluded that due to the similarity of the olefin composition to those observed from acid catalyzed dehydration and from the dehydration in dimethyl sulfoxide of alcohols rather than to those observed from the pyrolysis of acetates, the mechanism of the elimination involves the formation of an ion pair with the oxalate anion acting as the nucleophile rather than proceeding through a cyclic concerted mechanism. The purpose of this investigation was to elucidate more clearly the mechanism of the liquid phase pyrolysis of dialkyl oxalate esters by analyzing the resulting alkene mixtures and by studying the effects of the structure of the alkyl portion on the ease of decomposition. Of secondary interest was an evaluation of the reaction as a tool for the synthesis of alkenes. RESULTS AND DISCUSSION The dialkyl oxalates were prepared by the addition of oxalyl chloride to an ether solution of pyridine and the appropriate alcohol. The purity of the esters was established by infrared (i.r.) and nuclear magnetic resonance (n.m.r.) analysis. The i.r. spectra showed two strong carbonyl stretching frequencies at about 5.7/x and 5.8/i in agreement with the data reported by Corbett (3). The absence of the hydroxyl stretching absorption at about 3.0/x. was considered adequate proof that the esters were sufficiently pure. Table I summarizes the results of the thermal decomposition of dialkyl oxalates. In all subsequent illustrations the formula R-OX refers to the dialkyl oxalate ester. Pyrolysis of oxalate esters was effected by heating the sample in a small flask equipped with a four-inch Vigreux column and distillation head until a visible reaction accompanied by distillation took place. The distillate was collected in a receiver cooled with an isopropyl alcohol-dry ice bath and immediately analyzed by vapor phase chromatography (v.p.c.) and by n.m.r. Benzylic Systems Since tertiary aliphatic oxalates decompose smoothly at 140— 1700, it was anticipated from normal orders of reactivity that the benzylic system should behave similarly. Bis-(l-phenylethyl) I on em am as I o m :a mo I 003 I a o ea as ow an NH I am me I 0:: mm I N I I ma fiam m mm seas 003 s om> mzz came» a mHoz Camachoauuuo: ocmcuonuoc ocoaohuwupuoc mcmcuonuo: ocmnumfikucmmoHohu ocoxmnofiozo mamucmmoHozoahnumana ocmucoaoHozomcmamcuoa mcmxoaofio>oahnumala mamXCSOHozomcoahnuos CamuUOIN CCCDCOIH ostUOIN ocmgooua mcmnawumumwo ocmnawumlmcwoa mcmo>um cospom mmcoxfid muoumm mummeo mo cowuwmomaoooo Hmaumne mmmummm AHACCOCCOCINIoczovlmHm mmmlmdm Aazcuonuoclmuoxovzmflm omMIomm maznumfimucmmofiomolmvumflm oomuomm AHmcflnumoflzucmmoHozovanm oomlomm AfikcfinumofikxmnoHokovlmHm ommnomm Aakuooumvlmflm oomIomm AflmuooIHVImHm n AaanumfiacmnaaoI~.Hvaam okHIomH hasnomaacmnAIavaam .maofi mumamxo .maouoa .H magma 10 $w: I om H mconawumImamou I on CIUOIocmnHHumImcmou n mahnumamconmfloIm.HIoounuvafim $om I OH mamnawumImcmuu I mom HeIVOImcmnHaumImcmto n AHaeamHacmeaaeIN.HIOCAENCvamHm fimm m I ocmocomoHohoamnumslmIocoHAnumEIH ma I memuamaoaoaoasnsmaaeIm.~ msfiIosH Aesocma um I ocoucomofiohofihnuosflva.H IoHomoahnuosHvIN.HImcmpuvlmwm $Nm o I mcmucmaoaohoazzumaINImcoaznumaIH ma I odoucmmo~okuthmeHva.N OdaImmH Aahocmmofloho mm I acmucmaoHoaoaaeumaaeIm.H IaaeumsaeI~.HImauCImam smk :3 meIs.3.:ImcmuanIHIHanumaauoIm.m.~ o Hm Newa.HImcousnIHIHssumaaqum.m.N moHIosH AmeIH.H.HIHaa=nIN mm memugnIHIHssomaaeIm.MIAmeIHsnsmava IaseswaatuIm.m.Nvaam $om so On ocmucmmoHomomchHHSQOHQOmM am on mamucomoHohofimaoumomHIH NH OH mcmpcomoHomloammonQOmH mmHImda AHumoumINIHkucmaoHohunmvImHm $mm NH ma CGNXC£OHo>omcflvwHmaoumomfi 0H ocmanoflozoahmoua0mHIH mmw Nu mcmanoHoonhcomouaomw ooHIomH AHamouaINIHSRCAOHCSUINvaHm I um I mcmcuoncmucmnbhbhmenV I ,,aI x . ON I mamauauaeouo: omMIomm Assess m I ococuonuoc AamcmucomoHuhuImnwvINvaHm I on I Camachowuuuoc new I gemstoneoc QNMIomm AmeIm.m.«IaaeponnocINIoecvamam s om> mzz _ emguom mmemxa< .memp .I mumfimon CHCHN fi mHoz .mEoomQ 11 fimm $00 $ vfimww .mwmkamcm mwnu on HOMCA vo>oamu was muonfiaumImwo och .xmom CHmCHm m omzozm .u.a.>m .cssaoo coowawm $o~ a co copmummom won some mocsomaoo 03» echo .uoamH ummmsomwo on Hafiz Cam mum: wouo: no: mm ucmucoo aswuousmu echo .coflumumoucH oumusoom cm no“ mmcmxam map mumummmm hfiucowofiuusm uo: c~=o3 casaou command gem mafia .AONMN .a.nv ocfifiocflsv mcHXSHumu aw vomoaaoomc mos vcsoaaoo mafia .oscwmou on“ Ca omcMmEou poahaoa msomm ooH mamaemusnIm.HIHaeumEEeIm.~ o o ocmxocofiumoamnpmsIHIaxcmmouQOmeH do so mcmxonoaozoazusnnuuouIH NH 0H mcmxonoaoonzusnluuoulalmcoahnuoala mm om ocmanoHexcazusnluumuIelamnumaIH m I mcmxmnoHohothsnlunmpIdImcmthumaIH Ha I CCCXCnoHomoahusnquouIdIHznuoEIH I mm meIm.m.m.H.3ImcoaouaHRAumaIu I as meIm.n.mIocmaoeaAmeIHanumava om> mzz vmsuom mmcmxa< & 0H0: Conlomm ooHIOmH omalmma deIOdH ONH .maoH .meooma ocowvmcmxmnoHomoIm.m Iaanumsmsumqu.o.m.mImxoaoIa.a AHaxmeoaoaofiaosnIapmuIHvaam AHSXCnoHozoazusn qumuIalfiznumEIHImamuuvImam AahxmnoHozoakusn IuomnIsIHmnumsIHImHovaam AeoI~.N.N.H.H.H IfihmoumlmlflhsumaINuImHm oumfimxo 12 oxalate decomposed at 150—1700 to yield styrene as the sole product which was identified by its i.r. spectrum and from the melting point of the dibromide; m.p. 73—740 (reported (17) m.p. 73o). Bis-(1,2-diphenylethyl) oxalate was decomposed in refluxing quinoline to yield approximately 95% of trans—stilbene and 5% of gis—stilbene. The trans-stilbene was purified by recrystallization from ethanol and its identity was confirmed by i.r. and no mixed melting point depression with an authentic sample. Primary and Secondary Acyclic Systems Corbett observed no decomposition when oxalates derived from primary or secondary alcohols were heated to 3250 for one hour. Bis-(l-pentyl) oxalate and bis-(3-pentyl) oxalate were reheated and distilled smoothly at 300-3300 with no decomposition as evidenced by the fact that the i.r. spectra of the unheated sample, the distillate, and the pot residue were identical. Bis-(l-octyl) oxalate and bis-(Z-octyl) oxalate were prepared and pyrolyzed. Their boiling points were sufficiently high to allow heating to 3600 at which temperature they slowly decomposed. After heating for four hours enough distillate was collected for n.m.r. and v.p.c. analysis. The alkenes from bis-(l—octyl) oxalate were 98% of l-octene and 2% of 2—octene. In addition to the alkenes, the distillate from both esters contained several higher boiling compounds that were not characterized. The alkenes from, bis—(2—octyl) oxalate were 40% of l-octene and 60% of 2-octene. The identity of the alkenes was proven by comparison with authentic samples. 13 Bis-(Cyclohexylcarbinyl) Oxalate As stated previously, Chapman interpreted the mechanism of the pyrolysis of borate esters as proceeding either by an ion pair or by a more completely separated carbonium ion that is sufficiently free to rearrange. Bis-(cyclohexylcarbinyl) oxalate decomposed slowly at 330—3600 yielding a mixture of alkenes and higher boiling compounds. Only the alkenes were identified and were shown to be 88% of methylene— cyclohexane and 12% of l-methylcyClohexene. Significantly, no cycloheptene was formed. 1,0X- fast , ld le I slow 4. _ CHZOX CH'2"0X ycnz A x L slow '7 fast ” ///j (1) la lb 1f fast CH3 " ‘ >~ fast lg 14 These data are consistent with an ion pair mechanism as evidenced by the 1,2-hydride shift to the more stable tertiary carbonium ion. Apparently elimination to form methylenecyclohexane is favored over expansion to give the cycloheptyl cation. Energetically there is little to be gained in going from a cyclohexyl to a cycloheptyl system. Bis-(Cyclopentylcarbinyl) Oxalate Bis-(cyclopentylcarbinyl) oxalate was heated to 3600 for four hours and one ml. of distillate was collected. Analysis showed it to be a mixture of alkenes and higher boiling compounds. Only the alkenes were identified and shown to be 5% of cyclohexene, 79% of methylenecyclopentane, and 16% of l-methylcyclopentene. These results again indicate that a carbonium ion intermediate is involved and they show an example of ring expansion from a less stable five-membered—ring to a more stable six-membered-ring. Felkin and LeNy reported that ring expansion occurred in the acetolysis of cyclopentylcarbinyl brosylate but not in the acetolysis of cyclohexylcarbinyl brosylate (l8). Bartlett and co-workers reported that ring expansion occurs to the extent of 80-92% in the acetolysis of cyclopentylcarbinyl p-nitrobenzenesulfonate (19). The fact that only 5% of cyclohexene was formed is consistent with the argument that a tight ion pair, not free to undergo normal carbonium ion rearrangements, is involved. However, the energy gained in going from the less stable five—membered-ring to the more 15 stable six-membered-ring is sufficient to permit the rearrange- ment to occur in competition with elimination. *,1DC ”“““Ias£”'“w*“ 2d 2s A slow (2) —>= —~u - —>— slow L__ fast a 2b 2f .1 Cvlf- fast Y ¢ CH CH moxé --.. ...--- “>— 3 ‘fast , 2C 2g Bis-(2-Cyclopentylethyl) Oxalate Bis-(2-cyclopentylethyl) oxalate decomposed upon heating at 350—3600 for four hours yielding cyclopentylethene as the only detectable alkene from v.p.c. and n.m.r. analysis. The distillate also contained several high boiling compounds that were not characterized. 16 Norbornyl Systems The variety of carbonium ion reactions involving bicyclic systems in which the norbornyl derivatives have been studied extensively (20), prompted an examination of the pyrolysis of bis-(gxg—Z-norbornyl) oxalate and bis-(egggg2-norbornyl) oxalate. Bis—(gxg-Z—norbornyl) oxalate decomposed at 245-2550 yielding a three component mixture (v.p.c.) of which the first two peaks, 60%, were identified as 95% of norbornene and 5% of nortricyclene by comparison with authentic samples. The third peak, which amounted to 40% of the total mixture, was identified as gxg-2-norbornyl formate. Bis—(gndg—Z—norbornyl) oxalate decomposed at-275—285° to also yield a three component mixture of which the formate ester Comprised 28%. The remaining 72% was 76% of nortricyclene and 24% of nor- bornene. The fact that the 252 isomer decomposed 300 below the £222 isomer is consistent with current views of anchimeric assistance by the 1,6—bond in ago but not in 2292 compounds (21). The strik— ing feature of these eliminations is the difference in the product ratio of norbornene to nortricyclene obtained from the 352 isomer- as compared to the 3232 isomer. Whereas bis-(g§§272-norborny1) oxalate forms mainly nortricyclene, bis-(gig—Z-norbornyl) oxalate almost exclusively forms norbornene. This difference may be rationalized using two reasonable assumptions with regard to the mechanism: (a) the ionization to the tight ion pair occurs in a slow step and (b) the elimination of the hydrogen to form the 17 products occurs in a very fast step that competes favorably with carbonium ion rearrangements and ion pair equilibration. For example, in bis-(gxg-Z-norbornyl) oxalate, the oxalate anion is situated in a position favorable for removal of a proton from carbon—3 or carbon-7 to form norbornene, and unfavorable for the removal of a proton from carbon-6 to form nortricyclene (equation 3). On the other hand, the intermediate from bis-(gngg—Z-norbornyl) oxalate has the oxalate anion in a position most favorable for abstraction of a proton from carbon-6 to yield nortricyclene (equa- tion 4). Abstraction of the BESS proton of carbon—3 would yield norbornene. 18 4a 4b 4C (4) 4e 4f In order to ascertain the extent of rearrangement of the norbornyl cation prior to elimination, bis-(gndg—Z-norbornyl- 2,3,3-d3) oxalate was pyrolyzed. The ester decomposed at 290— 3200 yielding products whose isomer distribution was identical with that reported for the unlabeled bis—(endg—Z-norbornyl) oxalate. The product mixture was analyzed by n.m.r. in order to determine the deuterium content on the vinyl positions of norbornene. This was accomplished by integrating the vinyl proton region versus the bridgehead proton region of the norbornene which numerically was 0.43 to 1.00. 19 The 2-norbornyl-2,3,3-d3 cation (for simplicity it is presented as the classical ion) may rearrange by three paths as in equation 5. Assuming that any isomerization of the carbonium ion derived from the ion pair of bis-(endo-Z-norbornyl-Z,3 3-d ) _,.._ ' 3 oxalate proceeds by the Wagner-Meerwein path, the extent of rearrangement may be estimated from the above integration. 2,3-deuteride shift \ V \ /D +/ ,/D D wagner-Meerwein shift ' /D *‘7— + \ (S) Z:__;::>/ D \D .D I“ D D \ H 2,6—hydride shift 5 ,. __,___ -__._._ _. I ...... 1+ Accounting for the fact that the deuterium content at carbon-3 of the oxalate was about 85%, and assuming an isotope effect of kH/kD=l.8 (the determination of the isotope effect will be discussed later), the calculations show that less than 32% of the norbornene arose from rearrangement. This lends support to the suggestion that the hydrogen abstraction from the slowly formed ion pair is a fast process that competes with rearrangement. In reactions involving more complete carbonium ion formation in norbornyl 20 derivatives,the products resulting from rearrangements within the norbornyl cation have exceeded 60% (22). Bis-(2—(ZS3—Cyclopentenyl)ethyl) Oxalate In view of the interesting solvolytic behavior of 2—(133- cyclopentenyl)ethyl tosylate (23) and p-nitrobenzenesulfonate (24) to yield only er—Z-norbornyl derivatives, and in order to test further the previous conclusions regarding the importance of the position of the anion, bis-(2-(133-cyclopentenyl)ethyl) oxalate was synthesized and pyrolyzed, It decomposed upon heating at 320-3500 for four hours to yield a mixture of alkenes and higher boiling compounds. The alkenes were analyzed by v.p.c. and n.m.r. and shown to be 3% of norbornene, 70% of nortricyclene, and 27% of [33-cyclopentenylethene. The higher boiling compounds were not characterized. These data suggest that the double bond plays an important role in product formation. Since bis-(2-(Z33- cyclopentenyl)ethyl) oxalate and bis-(2-cyclopentylethyl) oxalate required nearly identical conditions and times for pyrolysis, the double bond apparently does not influence the rate of ioniza— tion to the ion pair. The solvolysis of 2-(133-cyclopentenyl)ethyl p-nitrobenzene- sulfonate in acetic acid and sodium acetate yields only 252-2- norbornyl acetate that may arise from a tight ion pair (6b), a solvent separated ion pair (6c), or from the free ions (6d) (equation 6). 21 ONs 6a (6) OAc 6e 66 In the case of bis—(Z—(ZS3-cyclopentenyl)ethyl) oxalate the product ratios can be rationalized if the rate of hydrogen abstrac- tion is assumed to be faster than the rate of ion pair equilibration. ¥\\ HHT’ A \‘\ k A‘\ H‘\/ \ H \ _. x” I ox 0X H 7a 7b (7) /¢ 7 d / ./ / m The fact that only 3% of norbornene is formed suggests that the equilibrium between 7c and 7d is slow compared to the rate of hydrogen abstraction (equation 7). 1,2-Hydride Shifts and 1,2-A1ky1 Shifts To test further the ion pair mechanism,several esters were studied whose decomposition should lead to carbonium ions that can undergo facile 1,2—hydride or 1,2—alkyl shifts. CHQ‘FH C(CH ) OX \\C i 3 2 i —I_,. 72% CH\ 9H, C(CH3)20x c -—-—-—>— 12% q. 18% 10% 23 ox- 9X 2+ (CH ) C—C—CD -———————+>- (CH3) C-C-CD ——————+>— 3 3 ' 3 3 I 3 CH CH 3 3 (10) CD CH CH C-3 (CH ) C C-CD (CH ) C C-CH (CH3)3C— -CH2 4 3 3 - - 2 * 3 2." ‘ 2 CD3 55% 31% 14% CH3 ox C(CH ) CH C CH 3 3 3 1( 3)3 ——-——>- + (1.1) 6% 94% Bis—(2-cyclohexyl-2—propyl) oxalate (equation 8) and bis- (2—cyclopentyl-2-propyl) oxalate (equation 9) decomposed at 150- 1600 with appreciable 1,2-hydride shifts as evidenced by the formation of l-isopropylcyclohexene and 1-isopropylcyclopentene respectively. Bis-(2,3,3-trimethyl-2-butyl—l,l,1-d3) oxalate (equation 10) and bis—(l—EEEE-butylcyclohexyl) oxalate (equation 11) decomposed at 150-1700 with 14% and 6% methyl rearrangement respectively. In the two examples involving methyl shifts (equations 10 and 11), it should be noted that the carbonium ion is the very labile neg-pentyl ion that has been shown to rearrange extensively even under very mild conditions (25). 2a Stereospecificity The stereospecific integrity of the pyrolysis of dialkyl oxalates was evaluated from a study of the decomposition products from bis-(gig-l,2-dimethy1cyclopenty1) oxalate, bis-(Eggns-l,2- dimethylcyclopentyl) oxalate, bis-(erythro-l,2-diphenylethyl-2-d1) oxalate, and bis-(thggg-l,2-diphenylethyl-2-d1) oxalate. The results are compared with those obtained from the corresponding acetates in Table II. If the mechanism of the oxalate pyrolysis were a concerted stereospecific gig elimination, very little 1,2—dimethylcyclopentene should be formed from bis-(Egang-l,2-dimethylcyclopentyl) oxalate. Since 22% of 1,2-dimethylcyclopentene was formed from the trans ester, an ion pair is suggested which, upon elimination, can form the thermodynamically more stable alkene. In view of the signifi- cantly different amounts of 1,2-dimethylcyclopentene and 2,3-dimethyl- cyclopentene obtained from the BER and trans esters, the elimination of the 63-hydrogen must be governed by three factors: (1) the thermodynamic stability of the olefin being formed; (2) the availa- bility of a 63-hydrogen for abstraction; and (3) the position of the anion with respect to the cation. It is apparent that olefin formation is faster than equilibration between 12a and 12b. (12) l 25 .cosuou mQCQDHHHmlmCQHH Qfiuw EH COHHCOflmH EDHHOHDOG HGGU .HGQ 05H EOHM UNHGHDUHNU Gums» mhwfladfi MWQJH .NN 00cmhwwmmo .moaom cacmwuo mo mocomoma map aw CNHCCEOmH mfifimauonu new ooom m>onm omoaaoumc mmcmnawam .NN mocoummmm n .000: um haunwwam muammaomw mocmxam mnH .om mocmpmmmmm mcmnawumImcmmu onoIoconHHumImcmuu H ozonHHumImcmmu cIKuImcwnHHumImcmpu H mcmucmao~ozofimnumalwnocoahnuoaIa mcmucmmoHohoflhnumchIm.m ocoucoaoHoaoamnuosHoIm.H mcmucmmoHaheaznumslmumcoaxnumaIH mcmucmaoHomoamnuoaHva.m mcmucoaofluzoahnuoaHCIm.H HeINIHaaumaaamaaaeIm.HIomanH HeINIaaaamaaamaaHeIm.anaaaNam HmucmmoHozoaznuoEHoIN.HumcmuH HaacmaoHoaoHaaamaaeIm.aImao o.n&:n $00 $0M goa o.n$mm $oo fi dIH $m fimmImm gmm m$oHIm $NN fiomnom $0 fiomImN gma m$omI03 $mm mamNHouNm mwmhaoumm owmuoo< mumamxo a mac: weapon mocmxdd .HvaIHocmnumahamnmwclm.HIomunu Cam I acauaom meH< ounuNum cam HocmucmmoHuhofihnumsHD IN.HIm:muu cam Imau mo mmumumu< cam mmumamxo mfiu scum mauscoum mflmhaouzm asp mo acmwpmaaoo < .HH manmh 26 The importance of the position of the anion with respect to the cation is demonstrated further by the results from the pyrolysis of bis—(erythro-l,2-diphenylethy1-2-d1) oxalate and bis-(£2532: 1,2-diphenylethyl—2—d1) oxalate. The oxalate esters were decomposed in refluxing quinoline so as to preclude any adverse reactions of oxalic acid with the stilbenes at the elevated temperatures. The trans-stilbene was isolated, purified by recrystallization from' ethanol, and analyzed by n.m.r. Figures I and II illustrate the observed n.m.r. spectra of the trans-stilbenes obtained from the erythro and Ehrgg esters. The spectrum of the stilbene from the erythro ester shows in the vinyl region (7’=2.9l) a triplet of almost equal peak intensity as expected for the vinyl proton split by the deuterium. The spectrum of the stilbene obtained from the three ester shows a singlet (7”=2.88) for the vinyl protons. This spectrum was almost identical with that recorded for authentic trans-stilbene. The per cent of labeled and unlabeled trgng-stilbene in each sample was determined by integration of the phenyl region versus the vinyl region. The ratios were found to be 9.2 to 1.0 for the erythro ester pyrolysis product (Figure I), and 5.3 to 1.0 for the £2533 ester pyrolysis product (Figure II) for the phenyl region and the vinyl region respectively in each case. 27 Figure I. N.M.R. spectrum of the Product from Bis-(erythro- l,2-diphenylethyl-2-d1) oxalate. ._ .... ...“ _ _,~_.» ___~_'..-q—.__.r...-‘_—._..fl Figure 11. N.M.R. Spectrum of the Product from Bis-(threo— 1,2—diphenylethyl-2-d1) oxalate. 28 The results are mechanistically interpreted in equations l3 and 14. OX ?X g\/D ¢ ,/’H ¢\\ ’ ,/H C ”\xk 510:, H a fast , LI (13) l H O D D erythro iTSlow ox' g /H ¢ ox D a ' D \C (/ slow i // fast , II (l4) /\ A 1‘1 g C H O / \ H 9 H H threo Isotope Effect The extent of carbon-(3-hydrogen bond breaking in the ion pair may be estimated by studying the isotope effect when the CB—hydrogen is replaced with deuterium. The results are summarized in Table III. Bis-(2-methyl—2-propyl-1,1,l,2,2,2-d6) oxalate and bis—(2,3,3-trimethyl-2-butyl—l,l,l-d3) oxalate decomposed at 160-1700 yielding the olefins shown in Table III. The alkene mixture was analyzed by n.m.r. (Table III and Figures III-VI) and the isotope effect was calculated by the internal competition method as described by DePuy and co-workers (8). The corresponding unlabeled compounds pyrolyzed at 140-1500. The experimental kH/kD was 1.8. Wiberg has shown that at 1700 the theoretical kH/kD is 3.8. 29 m A oIH.H.H o.a o.H m.o H> m.a :H moms.:.:-ocmasnuauasnumsauulm.m.m okauooa Hm eta.Humcmusnuauaszuoaaaplm.m.~ nasusnnmuassume mm mamasnnanasnumaauum.muameuasspmao-~ nauu-m.m.mv-mam o.m o.m o.m > ooH mamasnuanasnumaauunm.m.~ omauosa Aamusnumuasaama -auu-m.m.mvumam o.H N.H . >H m.a mm mcum.m.m.H.a-m:mmouaasnumaum a: moxm.m.m-mcmaopaA c-H>Aumavum oaa Ace -N.~.~.H.H.H-H»aoaa -N-Hsauma-~v-mam o.H o.m . HHH ooH mamasusn0ma omanoaa Aazusnuuumuvumam Hmcw> Hmmuoa Hmusnuu mumwwm mzz cospom mocmxH< .maoH oumamxo mOMHmm Qx\mx R .anoma coflumumoHCH mzz .mccsoaaoo voaonmflcn mowvcoammuuoo on“ cam .mumamxo AvaH.H.HIH%p:n umnaanumaauoum.m.~vumam .mumamxo Ac c-~.~.m.a.a.Huasaoua-muasnoma-mv-mam mo mamsaousm may .HHH mHan 30 FigureIII. N.M.R. Spectrum of the Pyrolysis Product from Bis—(tert-butyl) oxalate. r_-*__ -.FM -.___fi~____i_w_ _. 7H=8.27 TMS 74:10.0 Figure IV. N.M.R. Spectrum of the Pyrolysis Product Bis-(2-methyl-2-propy1-1,1,l,2,2,2-d6) oxalate. s from i I Bis-(2,3,3-trimethyl-2—buty1) oxalate. y 1 Figure V. N.M.R. Spectrum of the Pyrolysis Product from i Ho /__ ____, f L 7 =5.3 l i i Figure VI. N.M.R. Spectrum of the Pyrolysis Products from Bis-(2,3,3—trimethyl-2—butyl-1,1,1-d3) oxalate. 7’=10.0 32 At #000, kH/kD observed for the pyrolysis of acetate esters is 1.6—1.9 (8,9) (theoretical kH/kD = 2.3 (10)). This large isotope effect has been interpreted in terms of extensive carbon-g-hydrogen bond breaking in the transition state in accord with the gig con— certed elimination mechanism previously discussed. The smaller isotope effect observed in the pyrolysis of oxalate esters is consistent with the ion pair mechanism, although by itself is insufficient evidence to distinguish between an ion pair mechanism and a concerted mechanism involving small carbon—f3-hydrogen bond breaking. Mechanism In summary, the mechanism of the liquid phase pyrolysis of dialkyl oxalates proceeds first by ionization of the carbon—oxygen bond in a slow step to form the tight ion pair. The oxalate anion then abstracts a §»-hydrogen in a fast step before the ion pair may completely equilibrate and at a rate that competes favorably with carbonium ion rearrangements. That oxalate and acetate esters decompose by different mechanisms is not surprising in view of the fact that the decompositions of the oxalates are performed in the liquid phase and oxalic acid is a stronger acid than acetic acid. Conformational Effects of Cyclohexyl Ring Systems Karabatsos and co-workers observed (16) that whereas bis- (l-methylcyclohexyl) oxalate and bis-(l-ethylcyclohexyl) oxalate 33 decomposed at 160°, bis-(l-isopropylcyclohexyl) oxalate decomposed 0 at 250 . This difference in the rates of decomposition could be attributed to differences in the relative populations of 15a and 15b (equation 15). For example, when R is methyl or ethyl the popu— ox R =2 (15) 15a 15b lation of 15a should be significant. When R is isopropyl, 15b should be the main conformer. Consideration of ground state and transition state energy differences between 15a and 15b and between their ion pairs leads to the conclusion that 15a should decompose faster. First of all, because of more severe 1,3—alky1-hydrogen interactions in 15a than in 15b, 15a should be the less stable conformer. Secondly, these interactions should be relieved in the transition state leading to the ion pair from 15a but not from 15b (equations 15a and 15b) This explanation was tested by synthesizing and pyrolyzing bis-(gig—l~methy1-4-tg££-butylcyclohexyl) oxalate (equatorial oxalate) and bis-(Eggng-l-methyl-4-£E££-butylcyclohexyl) oxalate (axial oxalate). The use of the tert-butyl group as a conformational 34 anchor for the cyclohexane ring is an established procedure (29). The finding that the cis isomer decomposes 300 below the trans isomer (TableIV) SUpports the proposed explanation. R H R H ,1 H H +-' 0X ‘\‘OX- ,V —+ use» 15a H ex H ox‘ H H ‘x‘ //R \ + ,V ._R_—.+ M (15b) 15b The fact that bis-(l-tert—butylcyclohexyl) oxalate, which should exist exclusively in 15a, decomposes at lower temperatures, 150-1600, is not surprising in view of the severe crowding of the groups in the ground state of this molecule. l,9-Dioxa-5,S,6,6-tetramethylcyclohexane-2,§-dione 1,4-Dioxa-S,S,6,6-tetramethylcyclohexane-Z,3—dione, prepared from pinacol and oxalyl chloride, was pyrolyzed to see if it would decompose by a cyclic mechanism as shown in equation 16. Pyrolysis CH3 0 0 CH CH CH3EC/V7\C// 3 \C/ 3 | {I ————»- || 4 2002 (16) CH3-C‘l /C\ /C\ / o \0 CH3 CH3 CH 35 Table IV. Pyrolysis of Bis—(l—methylcyclohexyl) oxalate, Bis- (cis—l-methyl-4-tert—butylcyclohexyl) oxalate, and Bis-(trans- l-methyl-h-tert-buty1cyclohexyl) oxalate. Decomp. Exo- Endo- Oxalate Temp. Olefin Olefin Bis-(cis-l-methyl-u-tert-butyl- cyclohexyl) 140—150 9% 91% Bis-(trans-l-methyl-Q-tert- butylcyclohexyl) 175-180 10% 90% Bis—(l-methylcyclohexyl)a 160 6% 94% aData from G. J. Karabatsos and co-workers (16). 36 was effected by heating to 350—3600 for four hours. The distillate contained 2,3—dimethyl-l,3-butadiene as the only alkene and several higher boiling materials that were not characterized. The 2,3- dimethyl-l,3—butadiene was identified by its n.m.r. spectrum and by comparison with an authentic sample. The decomposition mechanism is apparently the same as the one discussed above for non-cyclic dialkyl oxalates. Evaluation of the Reaction as a Synthetic Tool The liquid phase pyrolysis of oxalates derived from tertiary alcohols provides a convenient method for the synthesis of alkenes in view of the simplicity of the apparatus required, the mild conditions under which decomposition is effected, and the good yields that are obtained. The fact that more than one alkene is obtained is an obvious disadvantage. EXPERIMENTAL Preparation of Cyclohexylcarbinol Thirty—five g. (0.225 moles) of ethyl cyclohexanecarboxylate was reduced with 6.0 g. (0.158 moles) of lithium aluminum hydride in anhydrous ether. After stirring at room temperature for 24 hours, the reaction was hydrolyzed with 50 ml. of water and the reaction contents were poured onto 250 m1. of 10% sulfuric acid and ice. The ether layer was separated, the aqueous layer was eXtracted~with 100 ml. of ether, the ether layers were combined, dried over hag— mesium sulfate,and fractionated, yielding 20.3 g. (79%) of cyclohexyl- carbinol, b.p. 84—85715 mm. (reported (30) 820/11 mm.). Preparation of Cyclopentylcarbinol Ten g. (0.07 moles) of ethyl cyclopentanecarboxylate was reduced with 1.9 g. (0.05 moles) of lithium aluminum hydride by the above procedure yielding 5.4 g. (77%) of cyclopentylcarbinol, b.p. 82922 mm., ngs l.u562 (reported (31) b.p. 68—700/15 mm., n35 1.u570). Preparation of 2¥Cyclopenty1ethanol Cyclopentylacetic acid was prepared by a typical malonic ester synthesis from bromocyclopentane and diethyl malonate. The procedure was adapted from one described by Adams and Johnson (32). The cyclopentylacetic acid was reduced with lithium 37 38 aluminum hydride to yield 2-cyclopenty1ethanol, b.p. 85-900/40 mm. (reported (33) 70-730/22 mm.). Preparation of endo-2-norborneol To a suspension of 1.14 g. (0.08 moles) of lithium aluminum hydride and 100 ml. of anhydrous ether was added in the period of one hour 10 g. (0.106 moles) of 2—norbornanone dissolved in 50 ml. of ether. The solution was then stirred and refluxed for five hours. After cooling,the reaction was hydrolyzed with 25 m1. of water. The ether layer was decanted, the inorganic complex was washed with four 100 ml. portions of ether, the ether layers were combined, dried over magnesium sulfate, and the ether was removed by distillation. The crude gpgp—Z—norborneol was recrystallized from low boiling petroleum ether yielding 6.98 g. (68%) of pure gpgp-Z-norborneol, m.p. 147-149o (reported (34) m.p. 149°), its 3,5-dinitrobenzoate melted at 123—1240 (reported (34) m.p. 1230). Preparation of 2-norbornanone—3,3-d2 Twenty—five g. (0.23 moles) of 2—norbornanone was refluxed with 50 ml. of about 80% deuterium oxide and a trace of sodium methoxide for 24 hours. The solution was cooled and the 2-nor— bornanone was extracted with 50 ml. of anhydrous ether. The ether layer was dried over calcium chloride and the ether was removed by distillation. The recovered 2-norbornanomawas treated with a fresh portion of 80% deuterium oxide by the same procedure as above. 39 An n.m.r. analysis of the resulting 2-norbornanone showed 60% deuterium exchange at carbon—3. This procedure was then repeated three more times using 25 ml. portions of pure (99.77%) deuterium oxide. The resulting 2-norbornanone showed approximately 90% deuterium exchange at carbon-3. A more complete exchange could not be attained utilizing this procedure. Preparation of endo-2-norborneol-2,3,3-d3 2-Norbornanone-3,3-d2, 12.3 g. (0.11 moles), was reduced with 2.0 g. (0.047 moles) of lithium aluminum deuteride by the same procedure as described for the unlabeled 2-norbornanone. Six g. (47%) of epdp—Z-norborneol-Z,3,3-d3, m.p. 142—145°, was obtained. An i.r., n.m.r., and v.p.c. analysis showed that the product was contaminated with a trace of unreacted 2-norbornanone and less than 2% of the exp isomer. The n.m.r. analysis showed 98% deuterium content at carbon—2, and 85% deuterium content at carbon-3. Preparation of 4-bromoeyclopentene Cyclopentadiene was treated with bromine at -300 by the method described by Bartlett and Rice (35) to give 3,5-dibromocyclopentene which was immediately reduced with lithium aluminum hydride yielding 5% of 4-bromocyclopentene, b.p. 47—490/40 mm., n35 1.4987 (reported (35) b.p. u3°/35 mm., n35 1.u992). 40 Preparation of diethyl A3-cyclopentenyl malonate Diethyl malonate, 27.7 g. (0.175 moles), was added to a solution of sodium ethoxide and ethanol prepared by the addition of 3.65 g. (0.16 moles) of freshly cut sodium to 100 ml. of absolute ethanol. 4-Bromocyclopentene, 23.42 g. (0.159 moles), was added during five minutes and the stirred solution was refluxed for seven hours. The ethanol was removed by distillation, 250 ml. of benzene was added to dissolve the organic materials, and the inorganic materials were removed by washing with 100 ml. of 2% hydrochloric acid solution. The benzene layer was separated, dried over magnesium sulfate, and fractionated, yielding 18.43 g. (51%) of diethyl [13-cyclopentenyl malonate, b.p. 1050/3 mm. Preparation of AQB-cyclopentenyl malonic acid To a solution of 18 g. of potassium hydroxide and 20 ml. of water was added 18.43 g. (0.08 moles) of diethyl [33-cyclopentenyl malonate. A vigorous reaction took place almost immediately during which time most of the ethanol boiled off. The reaction was then refluxed for two hours. Fifty ml. of water was added and the solution was distilled until 50 ml. of distillate was collected. The residual liquid was cooled and acidified with 6N hydrochloric acid. The acid solution was extracted with four 125 m1. portions of ether, the ether layer was dried over magnesium sulfate, and the ether was removed by distillation. The resulting crude [SB-cyclo- pentenyl malonic acid, m.p. 156—1600 (reported (36) m.p. 149-1500), was not purified further. 41 Preparation of £13-cyclopentenyl acetic acid The crude ‘33-cyclopentenyl malonic acid was refluxed for 20 hours with 50 ml. of pyridine. After cooling, the reaction mixture was dissolved in 200 ml. of ether and washed with three 100 ml. portions of 5% hydrochloric acid. The ether layer was dried over magnesium sulfate, and the [SB-cyclopentenyl acetic acid was reduced immediately without isolation. Preparation of 2-(1A3-cyclopentenyl) ethanol To a mixture of 3.1 g. (0.081 moles) of lithium aluminum hydride and 100 ml. of ether was added the ether solution of Zl3-cyclopentenyl acetic acid. After refluxing for 18 hours, the mixture was hydrolyzed with 30 ml. of water and the ether layer was decanted. The complex was washed with three 50 ml. portions of ether, the ether layers were combined, dried over magnesium sulfate, and fractionated, yielding 2.87 g. (29%-- based on diethyl 233fcyclopentenyl malonate) of 2-(133-cyclo— 25 pentenyl) ethanol, b.p. 84-850/20 mm. nD 1.4678 (reported (36)txp. 180-1820, ngs 1.4691) Preparation of 2-cyclohexyl—2—propanol Seventy—eight g. (0.5 moles) of ethyl cyclohexane carboxylate was dissolved in an equal volume of ether and slowly added over a period of two hours to an ether solution of methyl magnesium bromide prepared from 24.3 g. (1.0 mole) of magnesium and excess methyl bromide. The reaction was stirred at room temperature 42 overnight. The solution was hydrolyzed with 250 m1. of saturated ammonium chloride solution, the ether layer was separated, dried over anhydrous magnesium sulfate, and fractionated, yielding 50.1 g. (72%) of 2-cyclohexyl-2-propanol, b.p. 82-830/12 mm., n30 1.4683 20 D 1.4688). (reported (37) b.p. 100-102°/25 mm., n Preparation of 2-cyclopenpyl-25propanol Ethyl cyclopentane carboxylate was reacted with methyl magnesium bromide using the same molar quantities and experimental technique as described above. The resulting alcohol was contaminated with unreacted ester which could not be separated by distillation so the product mixture was refluxed with 250 ml. of 8% sodium hydroxide for 12 hours. The organic layer was separated, dried over anhydrous calcium chloride, and fractionated, yielding 23.6 g. (37%) of pure 2-cyclopenty1-2-propanol, b.p. 67-690/13 mm. Preparation of 3,3-dime.thyl—2—butanone-l,1,l-d3 A mixture of 20 g. (0.20 moles) of 3,3-dimethyl—2-butanone, 50 m1. of about 80% deuterium oxide, and a trace of anhydrous potassium carbonate was refluxed for 24 hours. The organic layer was separated and dried over calcium chloride. This procedure was repeated three more times using 25 ml. portions of pure (99.77%) deuterium oxide. An n.m.r. analysis of the recovered 3,3-dimethyl- 2-butanone showed greater than 99% deuterium exchange at carbon-l. 43 Preparation of 2,3,3—trimethyl-2-butanol-l,1,l-d3 3,3-Dimethyl—2-butanone-l,1,l—d3 was added to an ether solution of methyl magnesium iodide affording a 56% yield of 2,3,3—trimethyl— 2-butanol, b.p. 126—1310. An n.m.r. analysis show greater than 99% deuterium content at carbon-l. Preparation of 2,3,3-trimethyl-2-butanol 3,3-Dimethyl-2—butanone was added to an ether solution of methyl magnesium iodide affording 51% of 2,3,3-trimethy1-2-butanol, b.p. 127.129O (reported (38) 131—1320). Preparation of cis and trans l,2—dimethylcyclopentanol The pig and ppgpg l,2-dimethylcyclopentanols were prepared and separated by the method described by Hammond and Collins (39). Treatment of 2—methylcyclopentanone with methyl magnesium iodide yielded a mixture of the SEE and Epaps alcohols. The two isomers were separated by fractional distillation using an l8-inch spinning band column with a reflux to take-off ratio of 20:1. Tpaps—l,2-dimethylcyclopentanol boiled at 51-540/18 mm. (reported (39) b.p. 50-51.20/17 mm.) and gig-l,2—dimethylcyclopentanol boiled at 63—63.50/18 mm. (reported (39) b.p. 59.5-610/16 mm.) Vapor phase chromatography, using a six-foot Carbowax 20M column, showed that the alcohols were pure. Preparation of trans-stilbene oxide Thirty g. (0.167 moles) of trans-stilbene was oxidized with 47.3 g. (0.25 moles) of 40% peracetic acid by the method described 44 in Organic Syntheses (40). Twenty-three g. (72%) of trans-stilbene oxide, m.p. 68—70O (reported (40) m.p. 68-690), was obtained. Preparation of cis-stilbene oxide Thirty—three g. (0.184 moles) of gig-stilbene was oxidized with 52.3 g. (0.275 moles) of 40% peracetic acid by the same method as described in Organic Syntheses (40) for the synthesis of £3223- stilbene oxide. The crude gig-stilbene oxide was obtained as an oil after the removal of the solvent. The oil was dissolved in 100 ml. of hot 70% ethanol. The oil that was obtained on cooling solidified when placed overnight in a refrigerator. The solid was filtered, dried, and recrystallized from hexane yielding 18.1 g. (51%) of gig-stilbene oxide, m.p. 38.5-39.5O (reported (27) uo_42°>. Preparation of erythro-l,2-diphenylethanol-2-d1 Eleven g. (0.056 moles) of Epapg—stilbene oxide was reduced with 3.41 g. (0.081 moles) of lithium aluminum deuteride by the ' procedure described by Curtin and Kellom (27) except that the reaction mixture was refluxed for 11 hours. Recrystallization from ethanol yielded 8.78 g. (79%) of pure erythro—l,2-diphenyl- ethanol-Z—dl, m.p. 65-66O (reported (27) 64.4-65.40). Preparation of threo-l,2-diphenylethanol-2-d1 cis-Stilbene oxide, 9.8 g. (0.05 moles), was reduced with 1.60 g. (0.038 moles) of lithium aluminum deuteride by the same 45 procedure as described above, affording 8.43 g. (85%) of threo- 1,2—diphenylethanol-2-d1, m.p. 66-66.50 (reported (27) m.p. 64.5—65.50). Preparation of 2-methy1-2-propanol-l,1,l,2,2,2-d5 Ten g. (0.156 moles) of acetone-d6 was added to an ether solution of methyl magnesium iodide prepared from 25.4 g. (0.18 moles) of methyl iodide and 4.37 g. (0.18 moles) of magnesium. The reaction was refluxed and stirred for eight hours. Sixty m1. of saturated ammonium chloride solution was added, the ether layer was separated, the aqueous layer was extracted with 50 ml. of ether: and the ether layers were combined. The aqueous layer was dis- tilled until the temperature of the distillate remained constant at 1000 for five minutes. The distillate was saturated with sodium chloride and extracted with 50 ml. of ether. The ether layers were combined, dried over magnesium sulfate, and fractionated, yielding 6.50 g. (52%) of 2-methyl-2—propanol—l,1,l,2,2,2—d6, b.p. 80—810. An n.m.r. analysis showed the deuterium content at carbon-l and carbon-2 to be greater than 99%. Preparation of cis and trans-l-methy174-tert-butylcyclohexanol The pig and pgépg—l-methyl-4:£e£E-butylcyclohexanols were prepared and separated by the method described by DePuy and King (29). 4—Tepp-butylcyclohexanone was treated with methyl magnesium iodide and the resulting alcohol mixture was chromatographed on alumina. The progress of the separation was followed by v.p.c. he utilizing a six-foot Carbowax 20M column. The crude ppgpp isomer was sublimed at 600/6 mm. The pure EEEEE-l-methyl—4- pegp—butylcyclohexanol melted at 66-680 (reported (29) m.p. 70.5-710). The crude pip isomer was recrystallized from hexane. The pure pip-l-methyl-4-Ee£p-buty1cyclohexanol melted at 93.5- 950 (reported (29) m.p. 97.5—980). A v.p.c. analysis showed each alcohol to be isomerically pure. Preparation of l-tert-bupyleyclohexanol An ether solution of EEEE-butyl lithium was prepared from 6.9 g. (1.0 mole) of lithium sand and 92 g. (1.0 mole) of pepp- butyl chloride by the method described by Bartlett and Lefferts (41). To this solution, maintained at —40° throughout the addition, was added 73.5 g. (0.75 moles) of cyclohexanone dissolved in an equal volume of ether. The addition required two hours. After the addition was complete, the solution was stirred at —450 for one hour. The reaction was hydrolyzed by the careful addition of 250 ml. of water while the reaction temperature was maintained below ~10° until all of the excess lithium metal was destroyed. The mixture was then allowed to warm to room temperature, the ether layer was separated, the aqueous layer was extracted with 100 m1. of ether, and the ether layers were combined. After drying over magnesium sulfate, the ether solution was fractionated yielding 14.4 g. (12%) of I‘EEEET butylcyclohexanol, b.p. 73-760/10 mm., m.p. 48-50O (reported (42) b.p. 800/13 mm., m.p. 49—500). 47 Preparation of lig-dioxa—S,5,6,6-tetramethy1cyclohexane-2,3-dione To a solution of 5.9 g. (0.05 moles) of 2,3-dimethyl-2,3- butanediol, 4.19 g. (0.053 moles) of pyridine, and 100 ml. of ether was added over a 40-minute period, 6.34 g. (0.05 moles) of oxalyl chloride dissolved in 25 m1. of ether. After stirring, at room temperature for 18 hours, the reaction was hydrolyzed with 25 m1. of water. The ether layer was washed with three 20 m1. portions of 10% sulfuric acid, three 20 ml. portions of saturated sodium bicarbonate solution, and once with 20 ml. of water. After drying over magnesium sulfate, the ether was removed by distillation. The crude product was recrystallized from hot ethanol yielding 4.2 g. (50%) of 1,4—dioxa—5,5,6,6-tetramethy1- cyclohexane—2,3-dione, m.p. 110-1110. Other Alcohols Used l-Phenylethanol, 1,2—dipheny1ethanol, 1-octanol, 2-octanol, and exo-2—norborneol were available as commercial samples of high purity and were used without further purification. 48 General Procedure for the Preparation of Dialkyl Oxalates In a 300_ml., three-necked flask equipped with a stirrer, condenser, dropping funnel, and drying tube were placed 0.10 moles of the alcohol, 0.11 moles of pyridine, and 100 ml. of anhydrous ether. Five-hundredfim moles of oxalyl chloride dissolved in 25 ml. of anhydrous ether was added over a period of 30—40 minutes. The reaction was then stirred at room temperature for at least 12 hours. The ether layer that was separated after hydrolysis with 30 m1. of water was washed with three 30 ml. portions of 10% sulfuric acid, three 30 ml. portions of saturated sodium bicarbonate solution, and once with 30 ml. of water. The ether layer was dried over magnesium sulfate and the ether was removed by distillation. If the oxalate ester was obtained as an oil, it was not purified further. If the oxalate ester was obtained as a solid, it was recrystallized from hot ethanol. N.m.r. and i.r. spectra were taken for each ester to establish its purity and authenticity. Table V lists the reaction time, yield, and melting point of each ester. General Procedure for the Thermal Decomposition of Dialkyl Oxalates Three g. of the ester were placed in a 10-ml., pear-shaped flask equipped with a four-inch Vigreux column and a distillation head. The receiver was cooled with a dry ice-isopropyl alcohol bath. If the temperature of decomposition was anticipated to be 0 below 200 , the flask was heated with an oil bath. For temperatures 49 Table V. Reaction Time, Yield, and Melting Point of the Oxalates. Oxalate Bis-(l—phenylethyl) Bis-(1,2-diphenylethyl) Bis—(l-octyl) Bis-(2-octyl) Bis-(cyclohexylcarbinyl) Bis—(cyclopentylcarbinyl) Bis-(2-cyclopentylethyl) Bis-(egg-Z-norbornyl) Bis-(epgg-Z-norbornyl) Bis-(epgg-2-norbornyl-2,3,3—d3) Bis-(2-( A 3—cyclopentenyl )ethyl) Bis—(2-cyclopentyl-2-propyl) Bis—(2-cyclohexyl-2-propyl) Bis-(2,3,3-trimethyl—2-butyl— 1,1,1-d3) Bis-(cis-l,2-dimethylcyclopentyl) Bis-(trans-l,2-dimethy1cyclopentyl) Bis-(erythro-l,Z-diphenylethyl-Z-dl) Bis-(threo-l,2-diphenylethyl-2-d1) Bis-(2-methyl-2—propyl-l,1,l,2,2,2-d6) Reaction Time (hrs.) 14 14 20 20 12 18 18 12 14 18 16 16 14 18 20 15 14 14 12 % Yield 81 82 86 77 80 65 57 68 72 85 7O 86 84 46 77 78 73 62 Melting Point oil 67—71o oil oil 48-500 oil oil oil ' oil oil oil oil 54—560 143—1440 oil oil 67—720 69—750 0 71-73 Table V. Oxalate Bis-(cis-l-methyl-4-tert-butyl- cyclohexyl) Bis-(trans-l-methyl-4-tert-butyl- cyclohexyl) Bis-(l—tert—butylcyclohexyl) 50 (continued) Reaction % Time (hrs.) Yield 12 42 19 82 105 41 Melting Point oil oil oil 51 above 2000 a sand bath was used. The flask was heated slowly until a vigorous reaction accompanied by distillation of the alkene mixture occurred. The temperature of decomposition was noted by placing a thermometer close to the heated pear-shaped flask. Esters derived from tertiary or benzylic alcohols decomposed easily at 150-1700 requiring five or ten minutes for a complete reaction. Complete reaction of esters derived from secondary alicyclic alcohols required heating at 250-3000 for 60—90 minutes. Esters derived from primary aliphatic alcohols required Prolonged heating at temperatures of 330-3600 in order to obtain enough distillate for analysis. The technique employed was to heat the flask with a sand bath until the ester was refluxing smoothly approximately halfway up the Vigreux column, and to maintain this temperature until approximately one m1. of distillate was collected. This usually required around four hours. The alkenes were immediately analyzed without further purification by i.r., n.m.r., and v.p.c. Thermal Decomposition of bis-(1,2-dipheny1ethyl) oxalateirbis- (erythro—l,2-diphenylethyl-2—dl) oxalate, and bis—(threo—1,2- diphenylethyl—Z—dl) oxalate One g. of the ester was refluxed in 1.0 ml. of quinoline, b.p. 237? for one hour. The reaction mixture was cooled and dis— solved in 25 ml. of ether. The ether solution was washed with three 20 ml. portions of 10% sulfuric acid, once with 20 ml. of water, once with 20 ml. of saturated sodium bicarbonate solution, 52 and once with 20 m1. of water. The ether layer was decolorized with decolorizing carbon, dried over magnesium sulfate, and the ether was removed by distillation. The crude material was analyzed by n.m.r. to obtain the ratio of pig-stilbene to Eggpg-stilbene. The solid material was then recrystallized from hot ethanol for further n.m.r. analysis of the purified trans-stilbene. Infrared Spectra (I.R.) All i.r. spectra were recorded using a Beckman I.R.-5 double beam spectrophotometer. The liquids were analyzed as thin films between salt plates. The solids were analyzed as 20% solutions in carbon tetrachloride using 0.1 mm. spaced solution cells. Vapor Phase Chromatogrgphy (V.P.§,) All v.p.c. data were obtained using an Aerograph A-90—P equipped with a thermal conductivity detector and employing helium as the carrier gas. The alcohols were analyzed using a six-foot by 0.25— inch column packed with 20% Carbowax 20M on Chromosorb W. The alkene mixtures were analyzed using a six-foot by 0.25-inch column packed with 20% Apiezon L on Chromosorb W. Peak areas were deter— mined using a planimeter. Nuclear Magnetic Resonance Spectra (N.M.R.) All n.m.r. spectra were recorded using a Varian Associates Model A—60 Analytical N.M.R. Spectrometer. The samples were placed 53 in thin—walled tubes using tetramethylsilane as an internal reference. All liquid samples were analyZed neat. The solid materials were analyzed as approximately 20% solutions in an appropriate solvent. Peak areas were determined by electronic integration. PART II NUCLEAR MAGNETIC RESONANCE STUDIES OF N-METHYLPHENYLHYDRAZONES 54 INTRODUCTION Nuclear magnetic resonance spectroscopy (n.m.r.) has been successfully applied to the study of configurational isomerism resulting from hindered rotation about a carbon—nitrogen double bond in phenylhydrazones (1), 2,4-dinitrophenylhydrazones (2), semicarbazones and thiosemicarbazones (3), methyl ether imines, methyl imines, and hydrazones (4), N-methylhydrazones and N,N- dimethylhydrazones (5), nitrosamines (6), alkyl nitrites (7), and oximes (8,9). The purpose of this investigation was to study various N-methylphenylhydrazones with the following objectives in mind: (a) to measure the effect of solvent on chemical shifts; (b) to obtain several proton-proton and proton—130 spin-spin coupling constants; and (c) to establish the relative stabilities of the various configurations and conformations of these compounds. 55 RESULTS AND DISCUSSION Chemical Shifts To facilitate the discussion of the results, the following notations will be used: (a) The syn isomer of I will have the N—methylanilino group SEE to the smaller R group. This notation will also be used to distinguish between various protons of the R groups: ‘i.e., protons will be referred to as being exp or appi with respect to the N-methylanilino group. In those cases where both 312 and 2231 isomers exist, if a particular proton exp to the anisotropic group resonates at a higher magnetic field than the corresponding appi proton, it will be termed "shielded." Conversely, if it resonates at a lower magnetic field, it will be termed "deshielded." Table I summarizes the chemical shifts of aldehyde N-methyl— phenylhydrazones in neat, 10% in carbon tetrachloride, and 10% in benzene. Table II summarizes those of the ketone N-methylphenyl- hydrazones and the s n-appi isomer ratios. Figures I-VI show the spectra of the N-methylphenylhydrazones of acetaldehyde, propion- aldehyde, acetone, 2-butanone, 3-methyl-2—butanone, and pinacolone. 56 57 o o 1 mm.m Am.A Am.e as o N m o oA.N ON.m Nm.o sm.e Aoo - mo- m o I 1 I I umoz . e o 1 ms m om.A Ao.o NA.A s: o m m . mN.N sm.m sm.o No.s mA.A Hoo -Nmo-o A moo AA.N ms.m ON.N mo.o om.A sooz I Nm.m NN.A Ao.o . om.N . ”moo N N m . Nm.N - No.o No.m - oA.A Aoo . mo1mo A moo oA.N as.m sA.A oA.o ss.A sw.A .aooz - mm.m mN.N mo.o om.w mA.A “moo N N m . mN.m mm.o No.o as.m eo.A Aoo - A moo mo AN.N ms.m oA.A mo.e Am.m om.A aooz- - mm.m NN.N mo.A omeo N m me.N oo.m mA.o mm.N .saoo . no mo em.N Nm.m NA.N NN.N aooz -1 ms.m mN.N sa.m ”zoo m mo.N HA.N mN.o me.A Aoo - mo mm.N as.m NA.N AN.m aooz : apes : saw mmoo-z : mmouz «zoom mono omuun wrong mo-x.Nmo-h.nmo-xoaso>Aom m e :ooasooasz .E/ \m mmooAmmovz-m .moeosoooaeaaeoroassaozuz No Aeosae>Jiwv mamarm Nooaeoeo .A oases 58 .xoflmaoo ouwsv ma Cowman ahxaHom um Am oa use d «51m £81 d 530 N530 mmodo onOmH m Am mmeoAmzov21m .moeosoooesaasoeoAarao21z No Amosao>11oo mamasm Aooasoso .HH oaooe 60 .Aumocv ocoumuvhcfimcofiaahnuoalz ovhzovflwpood mo Esauoomw .m.z.z .H ousmwh WEB 61 .Apmocv ozonmwoknahconaamnuoalz ow>nowAmconoum mo sapwooam .m.2.z .HH madman . a a - - o D11- <. 3 i111! 11111111. .11- \P 1: \(§<§§¢ “flit: 62 OH WEB . will W .Aumocv ocoumuc%nahcmnmaznuoauz o:oumo< mo asuuoomm .m.z.z 1\l\ 1.)! .\:l\ \l‘ J \1 1 ..II.\(I\.Iv .11); i...» ll ((1 .‘Alu.’o‘...\ II 4‘! J. (\5 Sl,I1(/\ \1‘12( (3 1‘ I; x < .HHN ooswam M H Lo 63 .Aumocv ocowmuohthconmfiknuoEIz ococmpsmlm mo Esauoomm .m.z.z .>H madman +1115 1 is. 11)., ‘1}, ‘3‘; 1.§V-|}li I}I\ 1.111,.1, o m s mush; L _ 64 .Aaooev oeosoaeasassoeoassaoa1z odosoaso1N1aaeao21m mo asaaooam .m.z.z .> Gunman OH +111 L; ‘Ch j 65 I a . . ouzwam .m.z z H> . vmnamconmahfiuoa z ocofioomCHm mo Esuuoo m w .Aumocv ocowm E m. L . li‘l‘llrl‘of”; 66 The chemical shifts are accurate to i 0.01 p.p.m. and the differences in chemical shifts of the E22 and appi protons are accurate to I 0.10 c.p.s. The s n-appi ratios, accurate to about 5%, were calcu— lated from integration of the appropriate peak areas. Table III summarizes the differences in chemical shift of exp and appi protons. Structural assignments, exp and a233, were based on the assumption (9) that, when R1 is methyl (I), the equilibrium ratio 3 n-appi increases as R2 changes from ethyl to isopropyl to Eegp-butyl. The chemical shifts for any given series of hydrogens, o<-methyl, Q5-methylene, etc., are internally consistent as their variation in a particular solvent is usually less than a 0.10 p.p.m. As shown in Table III, whereas .emoo 5 m ..ooEovo... E 1:8 .3 $5183 .HE meow: ---1 .. . .. - .. 11.1. .- . 11.1. - .111 . . fl om AK om on 917 ('m'd'd) 69 omoo Ca mmulcmml H> 0cm .omcol :4 .3 m HOD CH Qo mo1aaso HA . m I l mo1aaso > .omeo ea mmo11z >H sAoo ea mmo11z H .auarm Aooaaoeo so soaasaao No ocoumuvhnahcocmHhfipmzlz odouood «0 fi mHoE -.I! 11 . on 1 T 1111.11.11] p.111. om Om .sAoo ea seesaw m .HHH> dogmas OH mo1sAm HAN 1mo.+ 10A.+ mH.+ (°m°d°d) 3 V 70 These results can be adequately explained by assuming stereo— specific association between benzene and substrate as illustrated in II and III. Such association has been proposed for amides (10) 1‘ C 96H5 111/ 1 .1 ‘ 4‘ // 3 N/// \"CH3 N I) CH3 H CH3 CH3 II III and alkyl nitrites, nitrosamines, and oximes (8,9). The peculiar behavior of the N-methyl protons can be rationalized by assuming that the conformation of the N-methylanilino group of acetaldehyde N-methylphenylhydrazone is different from that of acetone N-methyl- phenylhydrazone. Thus, whereas the N-methyl protons of the acetalde- hyde derivative are in the shielding region of benzene (II), those of the acetone derivative are in the deshielding region (III). Further support of this conformational argument will be given later. Proton—13C Spin—Spin Coupling Constants In general, J13C-H is a function of the p character of the carbon bonding orbital (11). The limitations of the correlation have been recently discussed (12). The J13C-H coupling constants for a number of the aldehyde N—methylphenylhydrazones were calculated from 13C natural abundance 71 measurements. The values are well within the limits anticipated for an ap2 hybridized carbon and are summarized in Table IV. Long Range Proton-Proton Coupling Constants An interesting feature of the n.m.r. spectra of aldehyde N-methylphenylhydrazones is the appearance of five and six bond coupling constants between HN and H1 and between HO; and HN respectively (IV). Figure 1x shows some pertinent spectra. N /C6H5 H \CH (N) C 3 These long-range couplings are probably less than 0.1 c.p.s. in the ketone N-methylphenylhydrazones as demonstrated by the fact that the N-methyl and the OC-hydrogen resonance peaks have half widths of 0.5-0.6 c.p.s. (half width of tetramethylsilane was 0.5-0.6 c.p.s.). Table V and Table VI summarize the long-range coupling constants for the aldehyde derivatives and the line half widths for the ketone derivatives respectively. This anomaly in long-range coupling can be again ascribed to conformational effects (II versus III). This point will now be considered in detail. ‘When Z is NHY, both exp (V) and appi (VI) isomers of aldehyde derivatives are present in solution, the isomer ratios being functions Table IV. 72 13 . . Proton- C Spin-Spin Coupling Constants. Methyl Ethyl n—Propyl i—Propyl i—Butyl 2-Butyl 3-Pentyl Cyclohexyl Nl-N(CH3)C H 65 R/ \H J 13C—H (c.p.s.) 162 154 157 156 156 153 155 158 73 Table V. Long Range Spin—Spin Coupling Constants (c.p.s.). N_N/C(SHS Hot. H \CH3 (N) x\C