Stereochemical Study of CycIopropylcarbinyl Derivatives in Carbonium-lon Reactions Citation Vogel, Martin (1961) Stereochemical Study of CycIopropylcarbinyl Derivatives in Carbonium-lon Reactions. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/7H4C-6W65. https://resolver.caltech.edu/CaltechTHESIS:04132011-075059835 Abstract Cyclopropylmethylcarbinylamine and cyclopropylmethylcarbinol were resolved and their configurations related. The former was deaminated with aqueous nitrous acid to give the latter with 0-4% net inversion of configuration. The solvolyses of several N-methyl-4-alkoxypyridinium salts were investigated. Solvolysis rates for N-methyl-4- (cyclopropylcarbinyloxy)-pyridinium iodide and perchlorate in water and in 80% ethanol and of N-methyl-4-(cyclopropylmethy1carbinyloxy)-pyridinium iodide in water were measured. The first-order rate constant for the hydrolysis of N-methyl-4-(cyclopropylmethylcarbinyloxy)-pyridinium iodide at 30° was calculated to be 6 x 10^3 times that of N-methyl-4-(cyclopropylcarbinyloxy)-pyridinium iodide. The hydrolysis of optically active N-methyl-4-(cyclopropylmethylcarbinyloxy)-pyridinium iodide produced cyclopropylmethylcarbinol with 4.4 ± 1.5% inversion of configuration. The stereochemistry of the deamination of cyclopropylmethylcarbinylamine and of the hydrolysis of N-methyl-4-(cyclopropylroethylcarbinyloxy)-pyridinium iodide and the ratio of rate constants for the cyclopropylroethylcarbinyl and cyclopropylcarbinyl pyridinium salts are discussed with respect to the possible intervention of non-classical carbonium-ion intermediates in these reactions. 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): Roberts, John D. Thesis Committee: Unknown, Unknown Defense Date: 1 January 1961 Record Number: CaltechTHESIS:04132011-075059835 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:04132011-075059835 DOI: 10.7907/7H4C-6W65 Related URLs: URL URL Type Description https://doi.org/10.1021/ja00962a033 DOI Article from thesis Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 6317 Collection: CaltechTHESIS Deposited By: Tony Diaz Deposited On: 13 Apr 2011 18:28 Last Modified: 16 Dec 2025 18:42 Thesis Files Preview PDF - Final Version See Usage Policy. 31MB Repository Staff Only: item control page

A STEREOCHEMICAL STUDY OF CYCLOPROPYLCARBINYL DERIVATIVES IN CARBONIUM-ION RFACTIONS Thesis by Martin Vogel In Partial Fulfillment of the Requirements For the Degree of Doctor of Philosophy California Institute of Teohnology Pasadena, California 1961 ACKN~VLEDGMENTS This thesis would not have been possible without the inspirational guidanoe of Professor John D. Roberts. National Science Foundation Predoctoral Fellowships for 1958-1960 are gratefully acknowledged. D ABSTRACT Cyolopropylmethylcarbinylamine and cyclopropylmethyloarbinol were resolved and their configurations related. The former was deaminated with aqueous nitrous acid to give the latter with 0-4% net inversion of configuration. The solvolyses of several N-methyl-4-alkoxypyridinium salts were investigated. Solvolysis rates for N-methyl-4- (cyclopropylcarbinyloxy)-pyridinium iodide and perchlorate in water and in 80% ethanol and of N~ethyl-4-(cyolopropyl­ methylcarbinyloxy)-pyridinium iodide in water were measured. The first-order rate constant for the hydrolysis of N-methyl-4-(cyclopropylmethyloarbinyloxy)-pyridinium iodide at 30· was caloulated to be 6 x 108 times that of N-methyl4-(cyclopropylcarbinyloxy)-pyridinium iodide. The hydrol- ysis of optically active N-methyl-4-(cyolopropylmethylcarbinyloxy)-pyridinium iodide produoed cyclopropylroethylcarblnol with 4.4 t 1.5% inversion of configuration. The stereoohemistry of the deamination of cyolopropylmethylcarbinylamine and of the hydrolysis of N-methyl-4(cyclopropylmethylcarbinyloxy)-pyridinium iodide and the ratio of rate constants for the cyclopropylmethylcarbinyl and oyclopropylcarbinyl pyridinium salts are discussed with respect to the possible intervention of non-olassical carbonium-ion intermediates in these reactions. TABLE OF CONTENTS Page I. INTRODUCTION. • • • • • • • • • • • • • • • • 1 II. RESULTS AND DISCUSSION. • • • • • • • • • • • 17 III. EX.PRRIMENTAL • • • • • • • • • • • • • • • • • 42 Cyclopropylmethylcarbinylamine • • • • • Resolution of Cyclopropylmethylcarbiny1- 42 8JDille • • • • • • • • • • • • • • • • •• ~ ~-N-(Cyclopropylmethylcarbinyl)-benz&mid~ 46 Cyclopropylmethylcarbinol • • • • • • •• 46 Resolution of Cyclopropylmethylcarbinol 47 A. d,l-Cyclopropylmethylcarbinyl HYdrogen Phthalate • • • • • • •• 47 B. Resolution of Cyclopropy1methylcarbinyl Hydrogen Phthalate ••• •• 47 C. Reduction of d-Cyclopropylmethylcarbinyl Hydrogen Phthalate • • •• 50 -d-Cyclopropylmethy1carbinyl Benzoate •• 61 Preparation and Decomposition of Optioally Active N-Nitroso-N-(cyclopropylmethylcarbinyl)-ben~ide • • • • • • • • • • • • 52 A. Deoomposition in Pentane • • • • • 63 B. Decomposition in Dioxane • • • • • 54 66 TABLE OFCONTENTS--continued Page Racemization of Cyclopropylmethylcarbinol under the Conditions of the Amine-Nitrou8 Acid Reaotion • • • • • • • • • • • • • •• 59 Deamination of Optically Active I-Phenylethylamine with Nitrous Acid •• • • • •• 59 4-Pyridylpyridinium Dichloride • • • • •• 63 4-Hydroxypyrldlne • • • • • • • • • • • •• 64 4-Chloropyridine • • • • • • • • • • • •• 64 4-Benzyloxypyridine • • • • • • • • • • •• 64 N-Methyl-4-benzy1oxypyridinlum Iodide • •• 64 N-Methy1-4-benly1oxypyridinium Perchlorate 64 4-(Cyc1opropy1carbinyloxy)-pyridine • • •• 65 N-Methyl-4-(cyclopropylcarbinyloxy)pyridinium Iodide • • • • • • • • • • • •• 66 N-~ethy1-4-(cyolopropylcarbinyloxy)­ pyridinium Perohlorate •••• • • • • • • 66 4-(Cyclopropylmethylcarbinyloxy)-pyridine. 67 N-Jlethyl-4-( oyclopropylmethylcarbinyloxy)pyridinium Iodide • • • • • • • • • • • • • 68 4-(1-Phenylethoxy)-pyridine • • • • • • •• 69 Optically Active 4-(Cyc1opropylmethylcarblnyloxy)-pyridine ••• • • • • • • • •• 71 Preparation and Hydrolysis of Optically Active N~ethy1-4-(cyc1opropylmethy1car­ binyloxy)-pyridine ••• • • • • • • • •• 71 TABLE OF CONTENTS--oontlnued Page Solvolysis of N-Methyl-4-(cyclopropylcarbinyloxy)-pyridine in 80% Ethanol • • • •• 72 Reaction of N-Methyl-4-benzyloxypyridinium Iodide and Perchlorate with Glacial Acetic Acid • . . • • . • . • • • ••. . . • • . . 73 Reaction of N-Methyl.4-(cyclopropylmethylcarbinyloxy)-pyridinium Iodide with Ethanol 74 Reaction of N-Methyl-4.(cyclopropylmethylcarbinyloxy)-pyridinium Iodide with Glacial Acetic Acid • • • • • • • • • • • • • • • • \:6 l Hydrolysis of N-Methyl-4-(cyclopropylcarblnyloxy)-pyridinium Iodide • • • • • • • • N-Methyl- 't -pyridone IV. • • • • • • • • • • • i \ \ 7~ 78 Kinetics of the Solvolyses of the N-llethyl4.alkoxypyridinium Salts • • • • • • • • • 82 N-Methyl-4-(cyclopropyloarbinyloxy)pyridinium Iodide • • • • • • • • • • • • • 84 N-Methyl-4-(cyclopropylcarbinyloxy)pyridinium Perohlorate • • • • • • • • • • 84 N-Methyl-4-(cyclopropylmethyloarbinyloxy)pyridinium Iodide • • • • • • • • • • • • • 84 Reaction of 80% Ethanol with Hydrogen Iodide • • • • • • • • • • • • • • • • • • 86 • • • • • • • • • • • • • • • • • • 94 PROPOSITIONS • • • • • • • • • • • • • • • • • • • • 100 , I References for Propositions • • • • • • • • • • , 77 \1\ NMR spectra • • • • • • • • • • • • • • • • REFERENCES \ 107 , \ I• INTBODUCfiOli Considerable interest attends the nature ot the intermediates in carbonium ion-type interconversion reactions ot c1Oloprop1lcarbinyl, oyolobutyl and al111carbinyl derivatives. Evidenoe regarding the., intermediates has come prinoipally trom two sources, n&DI811 studies ot solvol18h rates and product distributions in such intereonv.rsion reactions. The purpose ot this thesis was to gain addi. tional information trom the viewpoint of stereochemistry. A wide variety ot carbonium ion-type reactions ot both cyolopropyloarbinyl and cyclobutyl derivatiyes giTes similar relatlve amounts ot products wlth the oyclopropyloarblnyl, C1clobut11 and allyloarblnyl structures, almost independently ot whioh ot the two starting structures is employed. Such reaotions inolude the reaction ot c1Oloprop1lcarbinylamine and 01clobutylamine with nitrous acid (1, 2), the 8olvolJ81s of o101oprop1lcarbinyl and cyclobutyl derivatives (1, 3), and the reactions of c1Olopropyloarbinol and C101obutanol with thionyl ohloride (1, 4) and of c1Olopropy1oarbinol with hydrogen bromide or phosphorou8 tribroaide (1). This behavior suggests that all the above reactions go through common cationic intermediates; small variations in produot -2- composition mar be acoounted by specific effeots ot eaoh reaotion and do not require postulation ot a drastic ohange in Dl8chan18la. The Bolvol1Bes ot crclopropyloarbinyl and cyclobutyl derivatives are unusually fast. For instance, in 50% aqueous ethanol at 50°, cycloprop1lcarbinyl chloride is 40 tLmes aore reactive than ~methylallyl chloride and cyolobutyl chloride 1s 15 times more reactive than isopropyl ohloride (1). Unusually high solvolJais rates are otten indicative of non-claslical cationic Inte~ediate8 (5). The weight ot the evidence is such, theretore, that the CamRon intermediate(s) in the solvol7Ses ot o1Olopropylcarb1nTl and orclobutyl derivative. have been interred to be non-olualoal in nature (6). More detailed information as to the oommon non-cl&8lioal inter.mediate(a) may be obtained fra. isotopic labeling Both cyclopropyloarblnrlam1ne-a.16C (6) and allyloarbinylamlne-a. 16 C (2) haVt betn deaainated with experiment.. aqueous nitrous aoid and the oJOlobutanol and cyclopropyloarbinol produoed by the reactions were degraded in eaoh cue with the results shown below. -3- I 0.7% 36.8% lIONO, [>-'.CHaNB. ..... [:fH ··· _--- --, ··• to._ .. .. ___ .. + 28.1% ·l 36.8% [>+cHaOR I 48.3% I i53.2% I I 31.6% CH.=CR-CHa-~·CHaNH. HOBO» : 0.0% c!r~ ----_ ... --, I • · 36.8% ! 31.6% · [)+cn-OR 70.3% ·129.7% I I + I The isotopic carbon of the o7Clopropyloarblnol 'de rive a fro. the oorresponding amine i. 1188 extensively scrambled . than that of the other products examined. About 16% of thi. oyclopropyloarbinol appears to ari.e from a non-rearranging path as shown by the ratios of croloproPTloarblnol to 07010butanol obtained tra. the deaminatloDI ot cTcloproPTlcarbinylwne and of al1yloarbi1l11wu (1.39 and 1.17, reapeotiTelT (2». If a correction for the non-rearranging path is made, the o7C10propylcarbinol deriTed trom oycloproPTI. carbinylamine~~·C !!! the non-olaa.ical lntermediate(a) conta1ne 42~ of the label in the a-poaltlon and 58% in the ring positions, in line with the results of the other degradations • ( ,' The isotopic label ot the c70lopropylcarbinol and crclobutanol approaohes but doe. not attain a random distribution among the three methyl.ne groups. These results bave been correl.&ted .Ith a loheme (p. 5) having the three non-olalsioal RbioyolobutoDiumR ions (la.c) approaohing but not quite attaiDiag .,uilibriua with one another. Other 8yatelll8 of non-olaaeioal inte1'lMdiates can accotmt for the avallable exper1Mntal data but the ache_ presented here is the most att1'&otive from a theoretlcal viewpoint. A possible intermediate .t&ge tor the Interconversion of iOnl 1a-o is the 8,..etl'ical -tric1clobutonium· ion (II). The plane determined by the oarbon and ~ \~(~~~yH H ~~ C" H I H II the two h7drogen atOJU of each .ethylene group of II is perpendioular to the plane of the three Jl8thylene carbon atou. The ion II cannot be the lole intermediate in thl •• reactions beeaul. II predicts oaaplete scrambling of the isotopio label bet.een the methTl.ne groups of the oyolopropyloarblnol and cyolobutanol !or.med trom 2 r CH 2--CH 14CH I NH2 CH 2--CH C\4CH ,NH, HONO HONO , ,I I" (+) 1 , 1 " 1 (+): , I 1 I Ie CH 2--'CH " , " (T) I , 1 , 1 " 1 CH ,2-----CH2 ,_ I4(:H 2 --'CH /n ~ 2 14CH ---- CH ,2 • 2 " Ie \.. 2 - - CH ~H , C~2---J4CH 14~H20H + ~H20H C~CH I4cH2 14CH 2--CH + ____ ~~;H'OH + "" ' CH 2--CH C\ ~ CH 2 --CHOH 14CH2--~H2 14CH2-- CHOH CH2-~H2 CH 2--CHOH -CH2-14CH2 I I en al171oarbin1l.a1.ne-a.- u C and in the oyclobutanol trom cycloprop1lcarbin,.laaine-a-~·C. The ettect ot methyl substItution at the methylene groupe ot the bio1Olobutonium lou Ia-c has been investigated br Silver (7) by mean. ot a produot study ot the deamtDAtion ot appropriate ..iDl. with aqueous nitrous aoid. Silver's results, su.a.rhed in Tabl. I, tit rather well tnt 0 a IChelll .Wlar to that pre •• nbd tor the UlIIlIthylated aminel with 'OM understandable differences due to the presenoe ot the added methyl group. The methTI-substituted iona IIIa-o an no longer equal 1n ltabilit1 (in contrut to the unaubltltuted ions Ia-c) and their order appearl to be IlIa> IlIc » IIIb. Two or the amlnes, 2.mathylcyolo. butylu1.ne and 2.uthyloycloprop11oarblDTlamine, can torm direotl, either ot two non-elas.ioal iona but only the aore stable ot the possible ions aotua1ly appears to be tor.med. The succe.. or a soheme ot non-olassical ions in accounting tor Silver's result, luggested that the added methyl group bu not drastically altered the nature ot the intermediate, in the deaminationa of the methyl.substituted coapOUDde. Conaequently, w. oho •• to study the stereo. chealltry of the carbonium ion.tJPe reactions or optically activ. oyolopropylmethylcarbiDTl derivatives with a " 74& i~NHa 47 17& CHaVNHa ~ 49& 39 35 0 0 0 ~ OR 3?a 9 13 0 0 0 t>-cHa~H CIla a P1UB iaOMl"ic (allyllc) alcohols !J!. h1dl"ide sh1:tte. 51 ~-NH. eRa 100 100 (rrm. CRa [>-bH-CHa lma OR I [>-CR-CH. 0 0 0 0 0 0 <>-aH eRe 0 5 0 0 0 0 CHaVeH PERCEmGES OF lSOJWUC licoHOLS FROI AMINE-NITROUS ACID KlPE.RD(ENTS (7) TABLE I • • ...;J CHZ~CH CHZ-CHNH Z CHZ-CHNH Z CH 3CH--CH Z I I / CHZ-CH CH31\r2NH2 • CH'L_r ~ ~HZ CHZ-CH / me CHZ-'CH CH3r':!~-r2 1 3 I , , I I II "~, (+). I CH ----CH III:b CH3CH~CH ~ i NHT' ~HCH'/ I',:,~:) I I I mo CH ----CHCH CH Z--CHCH 3 l~r'~ ~H2 Z CHZ-CH CH Z I I CH3~~ I~r' OH I + CH'I~['OH I I co -9- reasonable hope ot obtaining eTid.no. tor a non-olas8lcal intermediate .uoh as IlIa. We .hall now examine the POB.ible oona'qulnce, ot non-ala•• leal intermediate. upon the stereooheal.t17 ot carboniUll ion-tr.pe reaotioJl8 of 0)'010proPTlmethfloarblnrl darlTatiTe. u.1ng water as the solvent. ~)~ I X IV The toraat1on ot the non-olaa81oal ion IlIa (shown here 1n an ata.1c orbital repre.entation) trom an optlcall7 actIve cyclopropylmeth7loarbinrl derivative IV should proceed preterentially in the deploted sense, with the ring bond farthest fram the leaving group (1) breaking to provide £-orb1tal overlap on the rear alde of the developing ~ orb1 tal on the a-carbon atom. 0Dl.7 the preferred contora&tion of the o.-oarbon atom relative to the cyclopropane ring of the aubstrate IV is discussed here. Howver, oonsideration of the other possible confor.mation with the metbTl group nearest the cyclopropane ring does not alter the -10. stereoobaDioal p~ediction8. water should attack ion IlIa predc:a1.nant17 from below, leading to 01010p~op;yl.ll8thyloar­ binol, V. with the same configuration as the sta:rtlng material IV. Thus, if the non.olusical ion lIla is the 801e ionio intermediate, the traaatormatlon of the 8ubst~te IV to the aloohol V should occur with predominant ~etention ot oonfiguration. The &8JDm8t:ric non-olassical lon IlIa ma;y have tlma to efUlllbrate with othe~ ionic aploie., inoluding ita mirror lage. betON reacting with wate:r. However, the i.omerio lona lIIb and 1110 do not appear to tora to a 81gnificant extent in ..te~ becau.e DO products derived tro. thlse iona are ob8e~ed when the 1n1tlall1 tormed specie. le IlIa. 'l'be _thyl-substi tutea tP1oJClobutonium ion VI ¥C ~ \(_~~/H H '0" I H H VI . , app:roaoh equilibrium. with IUa. (Reaction of VI with water could Nuonably giTe onl1 oycloprop;yl.meth;yloarbinol.) ..11- To the extent that this equlllbl'1l11l ooours, the product, 07Cloprowllleth7lcarbinol. will be Noaized since ion VI haa & plau ot a,.etry. VI could alternatively serve al the transit10n atate tor the tao11e oonTers1on ot the as,-tl'io ion IlIa to ita airl'Ol' 1mage, alao leading to raoealHd product. While still further ItructureS oan be visualized for the oarbonlua lon.type reactions of oJOlopropylmethrloarbinrl aeriTatlves, the preceding dl,oUl.ion suftioes to illustrate the general shreocMllical concluaion: That portion ot the reaction proceediag through non-classioal InterMdlate. 1I1ght reaaonab11 lead to products with 1llTwhere fra. nearly coaplete retention ot contiguration to complete raoeaization. The reaotions chosen tor thi •• tereoohemical stud:r wert not the cuatoma1'1 lolvolJtio reactions of elk:r1 halides and lateI'I tor .eTeral reasons. The 01010pl'op:rlmeth7lcarbl~1 haUde. haYe UTeI' been obtained in a pure tora because ot their read,. rearrangement to homerlo structures (7). Like. wlle, attapb to prepare cyclopl'Op:rlmlth11carbi~1 brosylate haTe aot been luccesstul (7). The IOlTOl7sis ot carbor,rlic estel'8 alW&JB lnTolTes the queltlon ot aoyl-oxrgen !!. alk71-0~a tission. In addition, the 801vol78e8 ot halide. and •• tera &re complicated 01 the preble. of lnteP.Dal return fram intimate 10~pa1r. (5). The reaotl011l actually aploJed were the d...bation ot o1OloProp1t..thylcarbinylaaSu with aqueous nltrous acld and the h1drol78h of N-raethrl...4.(o701oproP1wthylcarblDTloz;r)-p11"ldiniUII lodlde, giTing in both cases 07010proPT_thTlo&1"oinol 1.1 the produot. Both ot the 07010- proP11a8th71oarblnyl stArting materials are stable and 81.111,. handled. int,re.t In addltion, both pUNte the lon(a) ot o,. ..las of & neutral le&Tiag group whioh ahould obviate the aaaplication ot intel'Ul. return. Goering (8) baa ahown that the 801v011Bia of s..-thr1.2-01c1ohexeDT1 ~ni tttObeuoate in 80J acetone 18 &Cooapanled b1 interuJ. return but the aoid.cat&1)"Zed .olvo11lls of the 1&118 ,ubstl'at. 11 not. Pl"88U11flbly, the tmMtalyzed 801volraie pro- c•• d.I through An ion-pail', held togetur large1r by eleotrostatlo torces, from which inte1'D&l return ocourl (9). The aold.caW7Zed reaotion, where tht leaTlnggroup is a neutral !,-Dl trob.nzolo &oid .oleoule, Camlot fOM an ionpail" and the leaving group .scapes before lnt.rnal return oan ooov. The first reaotion of tntePiat. the deam1nation ot prima17 aliphatio aminea with nitrous &old, 1a generally ...13.. I.. considered to prooeed through a oarboalua-ion intermediate al though ot the aore detailed upeots ot the reaotion are a matter of aotive oontrov.rar (10). BOIQ ' mm.- RNH-NO - RN=NOH +H· .&0' The .tereocheaistry of the ..tne-nitrous aoid r.aotion, in gtneNJ.. 11 t1Pioal ot SNl proo••••• (11). The deminatlon 1n exoesl aqueous acid of 2-butrl•• 2-oot1l-, and l-pbenrl.thTlam1n8s 1s reported to give rao..iaation aDd m.rlin ot oonfiguration, alani. giT80 ret.ntion aDd pheJ!11gl.rcine give. retention aDd raoea1aation (12). Wiberg round 22% net inversion of configuratlon in the de am 1nat Ion of 2-butrlamSne in exce.s aqueous sulfurio aold and 26% inversion in glacial aoetio aoid (13). The deaminatlon of l-butylaa'na.l.·H in glacial aoetic a01d gives 69% net inTeraion ot configuration (14). OptioallT aotive a.-Mthyla1lr1.. Jna i8 d.aminated in glacial aoetio aoid to giTe a m.1xtuzoe at a.- and 'f .... thTlallTl ao.tatea. The a-aoetate showd 16% Det lnTerslon and neither the ratio of a- to 'f-aoetate. nor the optioal aotivitr of the a-aoetate waa &trected br the addition of acetate ion (16). However, le •• rear~nt to the 1-aoetat.... ob.erved than in thl s1lv.r-oatal7Sed &oetol)"8l1 of ~th1lal1yl ohloride. To acoount tor the smaller 8.JIlOunt ot "&l"1"&llg8JHnt in the dl.hatton reaotion. Young (15) proposed the intermediacy of a Whot- oarbonium ion (without the proper spatial orientation tor ally1io delooalizatloa ot the positiye charge) whioh glve. unrearranged and predom J naDt17 inverted aoetate in oa.petltloa with the normal delooalized 10n which gl...18 . raoudo pl"Oduct with the aame 1I0ur distribution ob'lM'ed in thl lo1yo!1tl0 reaotion. Unfortunately. the aaine-n1troua aold reaction haa onl7 rare17 been used to providl stereoohemical evidenoe tor non-claJ.ioal lonl. An examplt ot ita use Is the deamiD-tloa of e1ther ~- or exo.2-norbo~lam'ne to give ezclus1Te17 products of the !!2. oonfiguration (16-18) whioh hu been oited (18) as evidence in .upport of the non-alas.loal nature of the cationic intermediate. derived fram thl. syat.., this non-cla8sical nature 1. well documented by othtr d&ta (1a..20). It should lHt notea that there are eUllple. in the 11 terature ot a...inat1on reactions. unOOllplioated bT neighboring group or unuaual struotural efteot., c l.ading to produots of retained oonfiguration, apparentlT by an SNi Mob&n10. A particularl, interesting exaaplt 1a thl aeams..tion of optioally ... U .... l-pheD1letql .. hl, VII, 1n oal"8tullJ" drbd acetio uid (21). 11&. The maJ or product the o01"1"8lpondlng acetate, t01".lll4 with 8% net retention ot oontigwoa.tion. Honver, a ....U aJIOunt ot l-pheIl71. ethanol, VIII, was alao tOl'Jll8d ud with 8'7% retention. The toraatlon of VIII 8S explained by an Sri! decompos1 tion of the llltemediate ella..oacid IX (21). rn. C.H....c.NH. I H CRa I JI)!Q~ CeRa·C.N=N.oH I VII H IX < . .. CRa I + N. C.a..C..oH I H VIII The seoond reaction used in this work, the hydr01J811 ot N... thr1.4-(07oloProP7laethy1oarblll7loxy)-pyridiDiua iodide, 1, a new exaaple of the 1011'017811 ot a oationio sub.trat.. S1lD11ar reaotions tb4t have been .tudied pre. viouaJ.r are the aolYol7lea of aulfo1l1Ul1 .alta and of quaterD&l7 a.on1\111 ...lta. Swain baa .h01ll1 that while the deoom- position ot trimethfl. and tribensy1.,ultoDiua salt. cln solution are biaoleoular p~OO.I'.8 (22); the decomposition ot t-butyldimethylaultonlum salts 18 .. uniaolecula~ lolvolylla (23). The decomposition of benzhJdryltrimethylammonium hydroxide is reported to give benzhydrol by a first-order process and methanol by a second-order process (24). The hydrolysis of isobornrltrimethTlammonium hydroxide or iodide gives camphene and trioyclene, undoubtedly through a carboniuro ion intermediate (25). The stereochemistry of the solvolysis of cationic substrates i8 poorly documented. The deoomposition of optically aotive l-pheD7lathyltrimethylammonium iodide in methanol, ethanol and ... n-butanol gives the respective ether• of l-pheD1lethanol with complete racemi!ation (26). The starting material and the products were shown to be optically stable under the reaction oonditions. Although no kinetio evidenoe i8 given, the raoemlzed products are taken as proof of a un1molecular mechanism (26). The reaction of 2-bansa- midoalkyldtmethylsulfonium salts with water gives the products t1Pical of a solvolysis with participation of a neighboring bensamido group,8xceptwhen 8uch participation is sterioally prohibited (27). The second-order reaotion of optioally active l-phenylethyldtmethylsulfonium iodide with azide ion proceeds with complete inversion of configuration (28). -17II. RESULTS AND DISCUSSION c,yo1opropylmethy1carblny1&mine, X, .as resolved by means of its tartrate salt and the maximum rotation obtained was a··D +2l.scr (neat). That the resolution was complete or nearly complete was indicated b,y the oonversion ot the amine X ot maximum rotation to the benlamide XI in 98% yield =5.0). the melting point ot the optically active benzamide XI could not be with [a)··D +33.7· (absolute ethanol, c improved by reorystallization. v, X = 00 CRa X [>-9H X, X = NR. XI, X a:-NRCOC.H. XII, X =-o.CC.H.Co.H XIII, X --a.CC.R. Cyclopropylmethylcarbiny1 hydrogen phthalate, XII, was resolved with brucine. The almost constant melting point and specific rotation of the resolved phthalate XII upon traotional reorystallization apeak for its optical purity. The reduction of optically pure phthalate XII with lithium aluminum hydride gave cyolopropylmethrlcarbinol, V, oontaining impurities but a value ot aD +20.1° (neat) could -18be calculated for pure V. The alkoxide ion oorresponding to the optically pure aloohol V reacted with benzoyl chloride to give the optically pure bensoate XIII 1d th a"D +33.8~ (neat). The transfoI'Dlation of amine X to benzamide XI and the transformation of aloohol V to benzoate XIII do not disturb a~ of the bonds attached to the asymmetrio carbon atom so that amine X and benumlde IX of the same sign of rotation have the same oonfiguration and alcohol V and benzoate XIII of the same sign of rotation have the same oonfiguration. It remains to establish the relative configurations of the alcohol and amine deriTatives. There are at least 13 pairs of compounds with structurel XIV and XV whoae relative configurations have been .stabliahed (29). R' I R-C-NH. I H XIV R' I R-C-OH I H XV For each pair, with no exoeptions, the amine and alcohol of the same configuration have the same sign of rotation. It would not be unreasonable to conolude that oyclopropylI methyloarbinylamine, X, and oyolopropylmethyloarbinol, V, -19of the same sign of rotation also have the same configuration, espeoially because the optical rotations ot X and V ar8 both oomparatively large. Additional information on the relative oonfigurations of amine X and alcohol V was sought from the thermal decomposition of N-nitroso-N-(cyclopropylmethyloarbinyl)bensamide, XVI, to bensoate XIII. In a variety of polar and non-polar solvents, N-(l-arylethyl)-nitrosamides CR. CH. C:>-?-o.CCeRII + N. C>i-N(NO)COC e He H R XVI XIII d.oompol. with predominant retention of configuration (21, 80, 31). White (30) has found that N-nitroao-N(2-butyl)-benz8m1de d.oomposes with retention in polar solvents such as dioxane. On the other hand, a bimolecular inversion reaotion (with benzoic aoid relealed into the solution by an oletin-torming side reaction) compete8 in non-polar solvents such a8 pentane, leading to a slight net inversion of oonfiguration. In general, however, the decomposition of N-alkylnitrosamides ocours with retention ot configuration and this reaction has been used to relate the configuration. of at least one amine and alcohol (82). -20- When n1t~08amide XVI, p~lpared by the nitrosation ot amide (+)-X1, was deoomposed in dioxane and in pentane, benzoate (+)-XI11 was obtained with 21.2% and l6.8~ retention ot optical aotivity, ~espectiY'ly. It appears sat, to oonclude that the transformation ot (+)-Xl to (+)-X11I, eepeoially 1n dioxane, ~epresentl ~etention of configuration and the~efo~e alcohol (+).V and amine (+)-X bave the lame configuration, in agreement with the conclusion ~eached above by conlide~ation of the signs of rotation. how.ve~, Note, that the per cent retention of configuration observed here is small camp~ed to the 40-94% (pola~ 801vent.) observed in other oal8. (21, 30, 31). When optically active amine X was deaminated with aqueous nitrous aoid at high pH, nearly racemic aloohol V was obtained with a mmall (up to 4%) but detinite excell of V of the opposite sign and hence opposite configuration to the starting amine X. The largest oblerved rotation ot the product was a· 8 D +O.2~ (neat), 8ta~lng with amine X with a 8 °D -9.67'. 1% of the amine was in the produot. If the oont~ibutions of amine and aloohol to the ob8e~ved rotation a~e additive, then the alcohol would have aD +O.3~ When tree of amine. Thia cor~.cted rotation oorrespond. to ~ net inv'~lion using the maximum rotations for amine -21and alcohol given above. expe~1ment The p~oduot ot one denmination had the same sign of ~otation as the 8ta~ting amine but a simi1a~co~~ectlon to~ amine in the p~oduot ~eversed the sign. Signifioant amounts ot methyl cyc10- propyl ketone were also tound in the products. The alcohol V was shown to ~acemize only slightly (16%) under the reaction conditions. For one deamination experiment, the major produot was bis-(cyclopropylmethylcarbinyl) etherJ the formation of ethers under mild oonditions is indicative ot a highly stable carbonium-ion inter.medlate (33). The obserTed stereoohemioal ~esult of the deamination of amine X might be due in part to any of the tol10wing efteots: (1) The racemization of the p~oduot V by means of a base-oatalyzed hydride-transfer meohanism with meth1l cyolop~opyl ketone. (2) an "abnormal" SNl deamination mooh- anismJ (3) a ·nor.mal" deamination reaotion p~ooeoding through a diazonium-ion intermediate. can be largely disoounted. ~aoamizes The first effect Optioally active I-phenylethanol to the extent of 60% in the presenoe of potassium and fluorenone in t-butanol solvent at 100· after -17t-butoxlde hours, a base-catalyzed hydride-transfer p~e8umably by meohanism (34). Under the muoh milder conditions (partiou- larly the much le81 basic condi tiona) employed fo~ the -22- deaminatlon of cyolopropylmeth11oarbinylamine, oyolopropylmethyloarbinol should racemise little if aDT due to this meohanism. (Assumptions that corresponding derivatives of the I-phenylethyl and oyclopropylmeth71carbinrl structupes behave similarly will be made often in this disoussion. Justification for this assumption may be tound in the tact that both ebuotUP88 &.J>8 secondary oarbinyl 8;ystema with comparable sterie requirements and both have the same order of 801vol;ytlc reaotivity.) The second etfect. the ~1 mechanism, is a more serious problem and oan be disoussed in terms ot the following description ot the amine-nitrous acid reaotion. The -nor.ma1· deamination, leading to inverted produots, is piotured al +H.O, -Dt ~ •••• OHQ ROH (1) (2) + N. - - ROH an acid-catalyzed deoomposition of the intermediate dialo- - aoid (path 1). However. an uncatal;yzed SNi mechanism, via - . an ion-pair, ma;y also be _visualised (path 2). Evidence for the occurrence of path 2 in glacial acetic aoid has a!l'eady been presented (p. 14) but this path may also be -28- available for deaminations in water, especially under the basio oonditions employed for the deamination of cyclopropybnetbylcarbinylamine which should discoul'age the acidcata17zed path 1. Evidenoe for the occurrence of path 2 in an aqueous deamlnation l'eaetion was obtained in this wOl'k fl'om a study of the d88.D\ina tion of optioally acti ve l~ph~lethylamine to l-phenylethanol under variou8 pH oonditions, the l'esults of the study are summarised in Table II. Examination of Table II shows a definite ohange in the stereochemioal result with pH, with high pH leading to more nearly racemic produot. Expel'i~ent H 1s exceptional in that it gave raoemio aloohol, undoubtedly due to the acid-catalyzed racemisation of the same. The product should be optioally stable under the reaction oonditions of at least experiments Band C because oyolopl'opylmethyloarblnol was shown to be reasonably stable under conditions olosely approximating those of these two experiments. The most reasonable explanation tOI' the mol'S nearly raoemic l-phenrlethanol obtained in experiments B and C (and probably A) i8 the oompetition of an SNi meohanimn (path 2) with the normal mechanism (path 1) at high pH. The stereochemioal result ot path 1 tor the deaminatlon of cyclopropy1methy1carbiny1amine ~a1 be due to contributions 4.3 7.4 .. 5.9 -6.0 1.01 1.20 phosphate buffer phosphate buffe.. C D E F G H 2 hr. at reflux; 12 hr. at room t_perarue 13.6 6.2 -6.2 8.95-4.20 phospba te bufter phosphate buffer " 4 hr. at renux 2 hr. at reflux, S hr. at room t_pera tur. 13.2 6.06-6.01 0 S hr. at pe1'1n oveMlight at Pota tcpePature 4 hr. at 1"81'1UX -~---------~ 20 hr. at rsnux Re&ction Conditions 11.6 -----~ 5.2 -----------..----- 0.90 B %Net Inveraion 5.5 0.80 A pH 8.4 -8.9 Ji[oles of HelO.! );toles of Amine Expt. DEAlrITNATION OF OPTICALLY ACTIVE 1-PHENYLEWYLA1UNE IN WATER TABLE II • • ~ -25- from (1) an SN2 displaoement by water of n1 trogen 1'r(ll1 the intermediate diazonium ion or (2) a Rhot W classical oarbonium ion whioh can gIve product. directly or go over to (3) the more stable ·cold· (solvolytio), presumably non-ola.aical ion(s) suoh a8 discus8ed in the Introduction. Evidenoe for either of the first two possibilities may be found in the non-rearranging path whioh is important in the deamination 01' eyclopropylcarbLnylamine-a.16 C (s8e p. 8). Young (15) found .that 43% of the acetate for.med in the deaminatlan of a.-methylallylamine in aoetio acid appear. to oCllle frcm a non-rearranging mechanism (and the rest frcm a solvolytio oarbonium ion). He conoluded from the calculated stereo- ohemistry of the non-rearranging prooess (40% net inversion if the solvolytio carbonium ion gives raoemic produot) that a Whot- oarbonium ion is most likely to be involved rather than an SNi or SN2 process (see p. 13). Beoause of the unknown relative oontributions of the SNi meohanism and a Whot Wcarbonium-ion intePmediate to the observed stereochem1cal .l'esul t of the deamination of oyclo.. ~ propylmethy1carblD1lamine, it i, imp08sible to dete~in8 the contribution ot the solvolytic intermediate(.). Hawever, it appears reasonable that theshreoohemioal oonsequence of the solvolytic intermediate(.) is not a high c -26- degree (> 50%) of retention of configuration although sou retention of configuration cannot be excluded. The next approach to the general problem was to find a method of generating carboniua iona trom the solvol,.i, ot a positivelT charged oycloproPTlaethylcarbinyl derivative. Thua, the disadvantages of the amine-nitrous acid reaction might be avoided and the advantages retained: The carbonium ion would no longer be tormed in an exothermic step, obviating the problem of -hot- carbonium ions; the formation ot the carbonium ion would be the slow step ot the reaction and consequently accelaible to kinetic investigationJ internal return might Itill be avoided. Desirable properties of the cationic substrate would be ease of preparation and purification, ability to be rendered optically active and to be configurationally related to the 801vol1Bia products, ability to be followed kineticallT during lolvol1B1s, and treedom tram aide reactions during solvol78ia. A clas8 ot compounds was found which generallT met theae qualifications, namely N..eth1'l-4-alkox1PyridiniUII salta (XVII). ·2'7- 00-< • }-eRa XQ ..§L. RS + O~ )-CHe + HI XVII XVlIa, R = CeEOCH.-, X = I MId, R:= C>-CHsa" XVIlb, R D CeRaCH.-, X =CIO. XVIle, R:= [>-CH(CH.), X D I XVIle, R = t>-CH.-, X := I XVII!, R ,X=CIO. =C.H.CH(CH.), X := I I The 8)'11thea1s of XVII was accOlIpl1shed by the following route: ROH + NaR + C1-GN ( CH. ),80,. RO F\r ~. AgClO., CIO.Q• The rielda wen 86-89% for the fit·st step, 68-100% for the second 8tep and 51-72% for the third ,tep. All of the pyrldlnium salts were readily purified solids exoept for XVIle and f which were viscous liquids. The fo~er liquid was purified by low.temperature crlltallisation but the purification of the latter was not attempted. The produots of the 801vol188S of the pyridinium salts in various solvents were determined .alnly by means of .28- vapor-phase chromatography. Thut, the reaotion ot N-methyl4-bensylox1Prridinium iodide, XVIla, with glaoial aoetio acid at 109 0 gave a quantitative yield of benzll aoetate atter 92 houri; the halt-lite ot the reaotion was 2-3 houri. In contrast, the corresponding perohlorate salt XVIIb was compl.tely unreaotiTe under the same conditions, indicating that the anion must be involved in the rate-determining step of the reactlon ot the iodide aalt XVIIa. The 801. Tol,.is of N-meth,l.4-(cyolopropyloarbinyloxy)-prridinium iodlde, XVIlc, in 80% ethanol at 120° gave cyclopropyloarbinol, OTclobutanol and allyloarbinol but predominantly the oorresponding ethyl ethers .. products. The ratio ot the three etherl W&8 cyclopropyloarbinyl:cyolobutyl: a1lylcarbinyl =3.6:1.5:1.0; thi. ratio may be oompared to the ratio oba.rved for the ether products of the solvolysis ot oycloproP71oarbinyl ohloride in 80% ethanol at 97° of 5.7:2.5:1.0 (4). The hydrolysi. ot the cycloproplloarbinyl derlvatlve XVIlc at 98.6° in the presence of lithium oarbonate gave a 99% yield ot cycloproPllcarbinol, cyolobutanol and alllloarbinol; 8% ot the aloohol mixture was alll1oarbinol and the remaining 92% was c7010propylcarbinol and cyclobutanol in the ratio ot 1.2.1.5 to 1. The aqueous deamination ot a11ylcarbinylamine; whare cyoloproP1loarbin01 -29- and o1Olobutanol oan ariae prelu.ably only trom nonclauloal ions, gives a ratio of these alcohols ot 1.17 to 1 (2). Th. ethanolysls of N.methTl-4-(cyclopropylmetbyl- oarbinyloZT)-pyrldini_ iodide, role, at 98.4° gan 01010propyI.ethJloarbinyl ethyl ether with no isomerio .therl; hydrolysis at rOo. temperature 1n the presence ot lithium oarbonate gay' only oyolopropylmethyloarblnol • . On the other hand, the acetolysis ot the .ame substrate at 87° gave a boat ot products, inoludlJlg c1clopropylmethylcarbiD11 aoetat., the produot mixture ob•• rYed at 126° oontained no o1010propylmeth;rlcarbinyl aoetate, undoubtedly due to the ilomerization of this substance under the conditions ot the reaction. Aoetol18is of the I-phenylethyl derivative XVIIt gave I-phenylethyl aoetate. The tate ot the heterooyolic ring during the solvol18il reactions was not rigorously determined. From the sol. vol,.'1 ot the oyclopropylmethyloarbinyl derivative XVIIe in acetio acid and in ethanol and the lolvol18ia ot the I-phen11ethrl derivative XVIIf 1n aoetic aoid were isolated what · appeared to be the same yellow cr18talline solld. The solid could not be obtained in a pure .tate but oontained carbon, hydrogen, nitrogen and lonizable iodine, was very soluble in bJdrox1lio solvents and insoluble in non-polar -30- solvents, and was very acidic. The available evidence indieat" that this material ... the iapure hydroiodide ot N-Jdthfl. "l-p11"ldone, as expeo~ed. When a solution of H.aethyl- ~ -mldon. in aqueous hydrogen iodide was evaporated to dr1ll'''' a yellow soUd remained with a wide melting-point range that appeal'td to be similar to the . sollda obtained trom the solvolytle reaotions. The kinetics of the 80lv017l" of N-methyl-4.(cyclo- proPTloarblnyloxy)-prridinium iodide, XVIIa, and perchlorat., XVIId. in water and in 80% ethanol and of N-lI8thyl-4(oTclopropylmethylcarbinyloxy)-pyrldlnium iodide, XVlle. in _tel' nre measured. The reaoUoH were followed bT titration of the evolved acid and the data are summarized in Table III. The tirst-order rate constants stayed rela. tively oonstant through the first or second half-life ot eaoh reaction but deoreased slgaitleant11 in value thereafter (see Tables IV-IX). hoa experiments 2 and 3, an ~E* =31.2 + 0.9 kcal./mole, and an activation energy, - - activation entropy, ~ g:t: = 3.6 + 2.5 •• u., may be calcu... lated (6) tor the hydrol78is of the o1Olopropyloarbin,1 derivative XYlic. The rate constant for this oompound, extrapolated to 30°, is !.:a. = 2.65 x 10·- hr.-:&'; therefore, the c701oProPylmeth7lcarbinrl d.rivative XVII. 98.6 30.0 80% etlwlol water _ter mId XVIId mle 4 5 6 ~N-CH. xi 0.010 98.4 .ter mlo 3 'D(\J=',' 0.0065 70.7 water IVIIo C>- CRe- ,1:: CIO. ,X:: I 84.5, 84.7, 86.1 90.7 84.0 89.4 90.7, 91.9 89.4 IntlniiL.eTiter (% or or1) XVIIa, R:: [>-CH(CRe ), X :: I mId, R:: [:>-CH.- 0.16 0.19 0.20 XYlIc, R:: 98.6 0.013 2 98.4 80% ethanol role 1 ~ (hr. -J.) SolTent Subatnt8 Expt. 'l'!JRP. , °C. SOLVOLYSIS OF N-KETHIL-4 UKOXYPYRIDIDUll S!ll'S (XVII) TiBli III <) CA) ....• • -32- (!) a 0.16 hr.-~) is 6 x 10· more reactive than the orelopropyloarblnyl derivative XVIIo in water at 30°. The aloohols (or esters) obtained trom the solvol1B1. ot the pyridin1um salta _y arbe b7 meana ot three reasonable meohanisms: (1) A direct solvol7Bia reaction (SNl); (2) an SN2 displacement b7 iodide to give unrearrang.d alkyl iodide which then 8olvol7lea to give the tinal productaa and (3) the displacement otthe alkoxy group trom the ppid1ne ring by an aromatic nuoleophilic substitution mecbanbll. The aolvolyslS ot lI.... thyl.4-(oycloprop11oarbinyloxr)-ppid1niUll iodide in 80% ethanol and in water cannot be SN2 becauBe the rate. are virtually the a&a8 as thole ot the oorresponding perchlorate (aee Table III). The very weak nucleophilicity of perchlorate ion compared to iodide ion (35) would demand a large difterence in rate between the iodide and the perchlorate salts if the SN2 meohanism were operative. Furthermore, the large rate enhancement (6 x 10·) produced b1 a-.methyl substitution on the 01C'lopropylcarbinyl substrate is consistent only with the direot solvol1Bis reaotion (5). The aromatio nucleophilic .ubatitution may oompete successtully in aome of the inveatigated solvol1Bes. However, the bJdrolysis ot the oyolopropylcarbinyl aeriTativI must be prooeeding -33mostl7 by the direct solvol18is reaction because the aleoholio products were a mixture ot 1somers typical (within the wide limits ot experimental error) of carboniua iontype reactions of c1Olopropylcarb1n7l '78tems (sle p. 28) wher... the nucleophilic displao...nt .eohanism predicts no i.omerization. The enhanoed 8olvol7tl0 reactlvit7 of the o70loproP7lmethyloarbl~1 derivative should ooapletelT oTerwhola any tendency to reaot b7 the aromatio substitution meohanism 10 the produot. c7cloproP1lmethylcarbinol, undoubtedly 11 tormed exolusivelT bT a direot solvol1Bi8 meohanisa. The obBerved inf1n1tT titer. of the kinetic experiment. presented in Table III (84-92% ot theory) were di.turbing11 low and the factors responBl~e tor this phenomenon are very 1ikeiT related to those causing the decrease of rate conatanta in tho later stages ot the reactions. 1 pO.lible explanation 1. the formation ot I1omer1o unreaotive pyridini. . .alt. b7 internal return. Streit.ilser (5) has explaiDed the acid-catalysed a111110 rearrangements at 2-hexen-4-o1 to 3.hexen-2-01 (36) and of l-phen,I-2.buten. 1-01 to l-pheBTI-l-buten-3-ol (37) with partial net retention of oonfiguration on the baail ot internal return tro. an -ion-palr- with water playing the role ot the anion part of a true lon-pair. Thus, there 11 aome evidence for -34- internal return with neutral 1eaviag groups. Cyc1opropyl- carb1n71 ohloride shows a remarkable propensity to rearr~ by internal return, the aoetolJals of the ohloride appears to give 70% ot rear~d ohlorides (1) and even the bJdrol78il gives a 10% iaolated Jie1d ot rearranged ohlorides (1). However, internal return should oertainly be muoh leu important with a neutral rather than an anionio leaving group (se. p. 12). In addition, the tact that the oyolopropylGarbinrl derivative XVIlc gave a 99% yield ot a1oohols (admittedly under difterent oonditions than tho.e ot the kinetic experiments) indioates that internal return is un.1mportant tor this substrate. Another possible explaDation is the consumption ot the evolved acid by a side reaotion. When hydrogen iodide was dissolved in 80% ethanol under oonditions approxLmating those ot experiment 1, 15% ot the initial aoid titer di.appeared atter 17 hours, due. probably. to the formation ot ethJl iodide or molecular iod1ne (the spent solution was brown-oolored). ThuI, this taotor 18 undoubtedly ~portant in experiment 1 but its importance in the other kinetio experiments is not known. Still further alde reaotions ma, be occurring. A caretul inve.tigation of the solvolysis produots should shed oonsiderable light on the source ot the low infinity titers and d1a1nlshiDg rate constants. -35. The ratio ot rate constant. tor the solvol7Ses of N-mathJl-4-(o101opropylmethylcarbinTlor,y}-pyridinium iodide and ot '-aethrl-4-(cycloproP11oarbinylor,y)-pyrldinium iodide in water at 30· (6 x 10·) i. the first reliable measurement ot the .rrect ot an a-methyl group upon the 801vo17a18 rate ot a oyolopropylcarbinrl derivative. Silver (7) tound that the aoid-oatal1Zed solvol18is ot oycloproP7lmethyloarbin1l E.-nltroben.soate is 40 times raster than that of o1clopropyloarb1nJl E.-nltrobenzoate but then is good reason to belbve that the alower of the two ocapounda 1s reacting b1 aoyl..oxnen ti88ion. From the .olvolya1e in 50% ethanol ot a Il1nure ot broJlidea that probabl, oontained about 10% ot oycloproP1lmeth7lcarblnyl bromide, Sllver (7) calculated a minimum rat. oonstant for c101oproP1lmethylcarblnyl bromide whioh was about twioe that ot o7010proP11oarbinyl bromide. It the .olvolJSls ot a oyoloproP11carblnyl derivative prooe.a. through the cla.sical oarbonium ion, one would expeot a rate enhanoement due to a-metbyl ,ubstitution of greater than 10'. The formolya1. ot i-butyl bromide 18 about 10' time. taster than that ot isoprop11 bromide (38). Sino. the solvolJSls of isopropyl bromide is far from limiting (39), the tonaolysis ot isopropyl bromide 1, being -36assi.ted to lame extent by nuoleophilic attaok and the rate enhancement due to a-meth11 aub.tItution would be still greater if the lolvol18ia ot i.opropyl bromide were limiting. Availabl. rate data (1. 40) indicate that the 80lvol18i. of o1Clopropylcarblnyl ohlorid. 1. nearly limiting (~Q 0.9) so that one would predict tor thll case a rate enhancement or greater than 10· upon a-meth11 substitution it a classical cation were tOrBed in the transition state. In addi- tion, one might expect the eftect ot methyl substItution upon a primarr carbonium ion (c101oprop11carbinyl) to be great.r than upon a seoondary oarbonium ion (isopropyl) because the po.itive oharge ot the •• condar1 oarbonium ion would be more delocalized over the addItional alk,rl group. The observ.d rate .nhancement due to a-methyl substi- tution of the hldrolrais rate of N.-.thTl-4-(cyolopropyl. carbin1loxy)-pyridinium iodide ot 6 x 10- is oonaiderablT lesa than 10· and doea not .peak tor the classioal intermediate. However, the tactor ot 6 x loa is large enough to raiae the question ot whether the intermediates in the 80lvol1B1. or o1010proP11oarbiD11 and o70lobut11 derivatives are b.tter tonaulated as cla.aical o101opropyloarbinyl and oyclobutyl iona, respectively, rather than a common -bicyolobutonium8 ion. For a claasioal c1cloprop11oarbinyl -37. intel'Dl8diate, one would expeot Tef7 little rate enhanoement due to ~thyl substitution on o1Olopropylcarbinyl ohloride [(l-Mth1'lc70lopropyl}-oarbinyl ohloride] f whereas a factor ot about 10· is aotually obserYed (41). Also, for a olassical o1Olobuty1 intermediate, one would expect a solvolysil rate enhancement due to a-methJ1 lubstitution on cyclobuty1 ohloride (l... thylcyolobutyl ohloride) approaching 10· but the ob,erTed etteot is a faotor of only 10· (41). Thes. results lndicate that the poait1Te oharge ot the intexwdi. atea in the 801Tol1888 of oyclopropy1oarblnyl chloride and cyclobuty1 ohloride is highly de100allzed and are inconsistent with rate-determining formation ot olassical °7010proP71oal"blnyl and oyolobutyl iona. In this connection, it would be usetul to have rate. on some 2.methylcyclobuty1 derivatives. These substances 'hould be quite reaotive (pel"haps 10·.10· taster than c1010butyl) because they should ionise to the most stable of the three methyl-substituted -bioyclobutoaiUl'l- ions (IlIa) III should C101oPl"OPyl.metbrleal"b1n71 aerivatives. a classical inter.ediat. would prediot essentially no rate enhanoement. The rather large ettect (6 x 10-) upon the solvolysi. rate of N~8th1l-4.(cyclo­ propy1oarblDTloxy)-ppidin1um iodide due to a-metbJl substltution ,ugge,ts that the int.rmadiate in the ..38- .ethTl-.ub.tltuted caBe mar be better toraulated as a more claesical (le •• delooalized) struoture luch a8 the -hJperconJugated- 10n XVIII or pOl8ibl1 &8 a true hoaoal1rl1c catioa XIX (42). U J• 1,)Ca:::.cH-CHe 1J. CH. XVIII ~::c=-c~ · XU In the toregoing d1acuulon of the effect of Jl8th1l sub.tltution on solTol78h rates, no mention has been _de at groUDd•• tate tree energies. Tbl. 18 a potentialll dangeHU• .oversight because 80lvol1l1s rates are a retlec- tion ot both the transition.. state (.... ntial1r the oarboniua 10n) and the ground-state stabl11tie. (42). Howevlr, the approaoh taken here should minlmls. this peril because the etfeot of metbTl substitution on ground.state stabilities should be approxlmatel, conataat tor the various substrate. and oonaeguentlT introduoe a canatant factor into the obsened rate enhancements. ThUl_ the relative value. ot the rate enhancements due to .. thJl .ubstltution can be attributed to carbonium-lon stabilities even it their absolute values cannot. -39- Both N-methTl-4-(cyoloproP1loarblD7loxy)-pyridinium iodide and perohlorate are about two time. more reactive in water than in 80% ethanol. Th1. small dependenoe of rate on solvent is surprising when oompared to the 3110 times greater reaotivity of i-butTl ohloride at 25° in water than in 80% ethanol (43). However. Swain (23). in a study ot the SNl solvo11Bis ot !,-butyld1methyllulfoniUlll chloride in a wide range ot solvente and solvent mixtures J tound le.. than a four.fold variation in the first-order rat. constant., in general, the b.tter solvating media gave the slowtat rat.s (the 801volys11 in 80% ethanol was about twice &8 fast &8 that in water). Thi. behavior was attributed to an Laportant role played by the cation-solvating power of the lolvent; the better solvating media stablll.e the .ultonium ion with its localised charge relative to the tranaltlon .tate where the oharge is distributed OYer sulfur, carbon and nine hfdrog8na (28). It is reasonablt, then. that the pyridinium .alts, whert the charge of the substrate il already deloealized oYer the pyridine ring and oxJgtn, should show behaviors intermediate between those of i-butyl ohloride and l-butyldtaethylsulionlum chloride. n -40- When N-methyl-4-(o701oprop,lmethylcarbinylor,y). pyridin1ua iodide wal s1llthes1.ea, lta.rtin« with optioa.ll.r active o1OlopJ-OP7lmethylcarblnol (V). and hydrolrzed in the presenoe ot 11 thlUlll oarbonate. the product. V, was of the oppo.lte oonfiguration and had 4.4 ~ 1.5% of the original aotivit,. We have already ,hown that the c1OloproP11matbTloarblnol 1s undoubtedl, tor.mad exolusive1, b7 an SNl-tne .. ohanlaa. Howyer, the interpretation ot thh stereoohaaical result 18 oomplioated b7 the possible ooapetItion between neighboring o101opropyl and lolYent participation. To illustrate this point we . , consider tbt 101vol1tl1 of I-phanyl.2-proP71 tOl7late which giYe, !!. 85% . t lnveraion in ethanol (") i 29~ net inversion in aoetio aoid (45) and up to 70% retention in lormio acid (45), Theae results are explaiatd 1n teru ot a competition between a 01&881cal 10n Inter.med1ate and a non-olas8ical phenon1\'81l lon with the latter being blportant only In torah acid (45). EYen though the c1Oloprop71 group appears to make a ver,r large oontribution to the hydrol181s of N-aethTl-4-(oJOloproP7lm8th7lcarbinTloX1)-p,rrldin1ua iodide, an appreoiable part ot the reaotion ma, be proceeding through a 01&l110al 10n and the oontributions ot the olas~ioal and non-ola8sioa1 intermediate. to the product oannot -41b. evaluated. The aoetolysis or formolysis of this pyri- dinium salt would minim!!. the possibility of solvent participation. Consequently, the acetolysis of the pyri- dinium salt was investigated but unfortunately no satisfactory method of following the kinetios of the reaction could b. found and considerable amounts of rearranged products appeared to be formed (see p. 29). This approach 18 worthy of further experimentation. It is interesting that the decomposition of N-nitro8oN-(cyclopropylmethylcarbiD1l)-benzamide, the deamination of cyclopropylmethyloarbinylamine and the hydrolysis of N.methyl-4-(cyclopropylmetbylcarbinyloxy)~pyridinium iodide all give unusually rac~ic products for their respective . types of reaction. Taken together, these results suggest the existence of a mechanism for facile racemization of the cammon cationic intermediates in these reactions suoh as the one proposed in the Introduction. -42III. EXPERIMENTAL All melting points and boiling points are uncorreoted. Elemental analyses were performed by Dr. Adalbert Elek, Elek Microanalytical Laboratories, L08 Angeles, California. Infrared absorption speotra were obtained using a PerkinElmer double-beam recording infrared speotrophotometer, Model 21, or a Beckman infrared apectrophotometer, Yodel IR-? All vapor ohromatographs were obtainedowith a Perkin-Elmer Vapor Fractometer, Model l54-C, using diisodecyl phthalate (Column A) as the stationary phase, unlesl otherwise indicated. All optical rotations were obserTed with a Winkel-Zeiss Polarimeter relative to air or the appropriate solvent in the cases of neat liquids and solutions, respectively, the rotations were observed with a l-dm. tube and are aocurate to + O.oZO unless otherwise noted. Appendix I gives detailed descriptions of sane of the fraotional-distillation columns used in this work. ClclopropylmethylcarbiAllamine.--In a 2-1. three-necked flask, equipped with a pressure-equalizing addition funnel, a mechanioal stirrer, a nitrogen inlet and a reflux condenser fitted with a caloium chloride drying tube, were placed 77 g. (2.0 moles) of lithium aluminum hydride and -43760 mI. of anhydrous ether. To the stirred mixture in a nitrogen atmosphere was added dropwise 105 g. (1.06 moles) of methyl oyolopropy1 ketox~e dissolved in 450 mI. of dry ether 80 as to maintain a moderate rate of reflux. The first half of the oxime solution was added over a period of 1.5 hr.1 the flask was cooled with an ioe-water mixture and the remaining oxime was added over a l-hr. period. The mixture was stirred at room temperature for 18 hr. and then under reflux for 34 hr. The condenser was replaced by a Dry-Ice oondenserl the flask was immersed in an icewater bath and the reaction mixture was hydrolyzed with 500 mI. of 10% aqueous sodium hydroxide. and water (250 mI.) were then added. Ether (250 ml.) The flask was equipped for downward distillation, and distillation was conducted until a seoond, aqueous phase formed in the reoeiver. The aqueous phase was saturated with potassium oarbonate and the ethereal phase was removed and dried over potassium carbonate. The solvent was stripped through Column land the residue was distilled through a SO-om. Vigreux column, yielding 70.0 g. (78%) of material, b.p. 91-94° (740 rom.), n·eD 1.4266 n·eD 1.4243 (lit. b.p. 94.2-94.S- (745 mm.), ... (46». -44Resolution of Cyclopropylmetbylearbinzlamine.--From a solution of 108 g. (1.27 moles) ot amine and 190 g. (1.27 moles) of d-tartaric acid in 300 mI. of water was obtained 121 g. (41%) of the amine tartrate salt. Six recrystalli- zationa from 93% ethanol gave 69 g. of salt, m.p. 161.8163.0·. A l5-g. sample of the tartrate and 30 g. of potas. sium hydroxide were added to 20 mI. of water. The organic layer was withdrawn and repeatedly shaken with sodium hydroxide pellets until an aqueous layer no longer formed. The recovered amine was dried over barium oxide and di.tilled through Column 2, yielding 3.6 g. ~th b.p. 88-9Z-. The amine was redistilled through Column 3, giving 1.15 g., b.p. 92-94°, a··D +21.30- (neat). Since analysis by Tapor- phase chromatography (VPC) showed the sample to contain about 5% of impurities, it was purified by preparative VPC. The material which collected had a··D +21.l4~ (neat) and contained about 5% water as determined by VPC. The remaining amine tartrate, 54 g., was crystallized twioe tram 93% ethanol giving 46 g., m.p. 161.0-162.0'. - Anal. N, 6.96. Caled. for C.HuNOes Found: C, 45.96J H, 7.29, C, 46.19, H, 7.23, N, 6.11. A IS-g. sample of the analytically pure tartrate was decamposed as above and the crude amine was purified by -45preparative vpe and dried over barium oxide, yielding material with a··D +21.~ (neat). The sample contained 5% water (vpe). The ramaining tartrate salt (31 g.), after two more reorystallizations, gave 26.3 g., m.p. l62.6-l64.ZO, which was dissolved with 60 g. of potassium hTdroxide in 45 mI. of water. The solution was continuously extracted with ether and the extracts were dried over potassium carbonat •• The ether was removed through Column I and the residue wal distilled from barium oxide through Column 2, giving 8.0 g. (83%) of amine, b.p. 92-9:f, aUD +21.8<f (neat), 99% pur. by vpe. The combined ethanolio mother liquors from all the crystallizations were evaporated to 700 mI., and, on cooling, 30 g. of tartrate was obtained. Decomposition of the salt with aqueous base, continuous extraotion of the resulting 8olution with ether, drying of the extract, over potassium carbonate and then barium oxide, stripping of the solvent and final11 distillation of the residue through Column 3 gave 5.4 g. (50%) of amine, b.p. 93°, aUD +21.45· (neat). The amine tartrate remaining in the original aqueous mother liquor was deoomposed with 150 g. of potassium -46- hydroxide. The solution was saturated with sodium ohloride and the organio phase was rmnoved, dried with potassium hydroxide and then with barium oxide and distilled through Column 2, yielding 38.2 g. of amine, b.p. 88-9gt, a··D -11.14° (neat). d_N_(ClcloproPllmethylcarbinyl)-benlamide.--A mixture of 5.7 g. (0.067 mole, a··D +21.80·) of d-cyclopropylmethJl - - carbinzlamine, 8.3 g. of sodium hydroxide, 12 g. (0.085 mole) of benzoll chloride and 60 mI. of water was stirred magnetioally for 10 hr. The white precipitate which sep- arated was oollected by vacuum filtration and was rec~ltal­ lized from 60% ethanol-water and then from 50% benzenehexane. The yield of purified benzamide '«as 11.0 g. (9~), m.p. 123.4-124.0·, [a)··D +33.7 ~ O.g- (absolute ethanol, ! = 5.0). Further recrystallization failed to raise the melting point of the produot. fhe infrared spectrum of the optically active benzamide was identical to that of a sample of racemic bensamide with m.p. 96.0-97.0· (lit. m.p. 96.8-97.6- (7». CycloproPllmethy1carbinol was prepared in 62% yield by the reduction of methyl oyclopropyl ketone (62.7 g., 0.746 mole) with lithium aluminum hydride {11.9 g., 0.31 - mole} in ether solution and had b.p. 119·, n··D 1.4299 - (lit. b.p. 123.6·, n8 °D 1.4816 (47». -47Resolution of Clelopropllmethyloarbinol. A. d,l- ClcloproPllmethllearbinzl Htdrogen Phthalate.--In a 100mI., round-bottomed flask was placed 75 mI. ot reagent-grade toluene and one-third ~s distilled to remove any moisture that may have been present. The remaining toluene was allowed to cool (protected from moist air with a calcium chloride drying tube) and 10.8 g. (0.126 mole) of d,l-clclo- -- propylmethylcarbinol and 18.5 g. (0.125 mole) of phtha1io anhydride were added. The mixture was stirred magnetically and heated in an oil bath at 86-90· for 12 hr., it was then cooled and poured into a solution of 18 g. of sodium carbonat. in 900 ml. of water. The aqueous phase was tiltered and acidified with oonc. hydrochloric acid. The oil that separated solidified upon standing and was collected and vacuum-dried to give 20.2 g. (69%) of white crystals, m.p. 67.8-69.4". An analytioal sample, m.p. 68.4-69.SO, was obtained bl recrystallization from acetic acid-water. Anal. Galcd. tor C .H .O.: C, 66.66, H, 6.02. Found: C, 66.66. H, 6.94. 1 B. 1 Resolution of Clolopropylmethylcarbiny1 Hydrogen Phtha1ate.--To 800 ml. of reagent-grade acetone were added 81 g. (0.346 mole) of raoemic cyclopropy1methylcarbinyl phthalate and 136 g. (0.345 mole) of anhydrou8 brucine. -48When the mixture was heated to reflux, the reactants dissolved but a precipitate of crude brucine cyclopropylmethylcarblnyl hydrogen phthalate soon formed. The hot acetone solution was vacuum-filtered fram the preoipitate and the flltrate, upon cooling, deposited crystals of the bruoine salt. The yield of crystals was 52 g. The mother liquor was used to dissolve more of the crude bruoine salt. Filtration and cooling of the 8olution gave 88 g. of crystals. Repetition of the process dissolved the remaining crude brucine and afforded 16 g. of orystals. " The three crops of crystallised bruoine salt (156 g.) were dissolved in the minimum amount (360 mI.) of boiling reagent-grade methanol and the solution was refrigerated, affording 81 g. of salt, m.p. 166-170-. Seven additional recrystallizations gave 38 g. of bruoine d-cyolopropyl- - methylcarbinyl phthalate, m.p. 175.0-177.6°. Fr~ prevlou. experiments, it was kno~n that this melting point oould not be raised by repeated recrystallisationl from methanol. Anal. Calcd. for C•• H.oN.O.: C, 68.77, H, 6.41. Found C, 68.18, H, 6.35. I The pure salt was decomposed by dissolving it in the minimum ~ount of boiling 95% ethanol and pouring the resulting solution into 240 mI. of water and 16 mI. of -49oonc. hydrochloric aoid in a separatory funnel. separated at the bottom ot the funnel. An oil It was known fram preliminary experiments that the oil could not be induced to crystallise directly, consequently, it was dissolved in five volumes of glacial acetic acidJ then, water was added to the cloud point. The solution was fro sen solid in a nry-Ice acetone bath and scratched as the solld warmed and melted. At about 15- a finely divided precipitate remained but turned to oil at roam temperature. ~ben the solution was oooled before it reached roam temperature, oiling was prevented. The first crop, A, of optically pupe phthalate was oollected and dried in a vacuum desiocator over phosphorous pentoxlde, 2.3 g., m.p. 57.8-69.8-, [a]··n +33.1 + O.~ (carbon tetraohloride, 0 = 10.2). the infrared spectrum was identical to that of the d,l- -phthalate. Water was added to the mother liquor to the cloud point. Freezing, scratohing while warming, and refrigeration of the solution gave a seoond crop, B, 1.6 g., m.p. 68.2-69.S-, [a]··n +33.7 t 0.2" (0 =10.2). A third orop, C, was obtained in a similar manner, 1.0 g., m.p. 57.4-59.0-, [a)··n +32.9 t 0.2- (~= 10.1). A fourth crop could not be realized. -60Crop B was racrfstalliBed from acetic acid-water by the above procedure into three fractions: B-1, 0.15 g., m.p. 57.8-59.6-1 B-2, 0.50 g., m.p. 57.8-59.~, [a)··D +33.1 :!:. O.~ (! = 9.7>' B-S, 0.3 g., m.p. 57.6-59.~. C. Reduotion ofd-Cyolopropylmethylcarbinll HYdrogen Phthalate.--The combined crops A, B-l,2,3, and C of the - d-phthalate (4.0 g., 0.017 mole, including all the samples used for measuring optioal rotations, which had been evaporated to dryness under an air stream and dried) were dissolved in 75 mI. of dry ether. The solution was added over a 0.5-hr. period to 2.3 g. (0.060 mole) of lithium aluminum hydride suspended in 75 mI. of ether in a 5OO-ml., threenecked flask equipped with a mechanical stirrer, a reflux condenser, a pressure-equalizing addition funnel and a nitrogen inlet. A nitrogen atmosphere and stirring were maintained throughout the reaction. The mixture was held at reflux for 1 hr. and then 7.7 mI. of ethyl acetate was added to decompose excess hydride. The mixture was stirred for an additional 15 min., the oondenser was replaced by a Dry-Ice condenser and 10 mI. or water was added. The mix- ture was then stirred and heated under reflux for 1 hr. and at roam temperature for 4 hr., and was then allowed to stand for 3 hr. until all the grey salts had disappeared, -61~ leaving only a white precipitate. The ether was decanted and the precipitate was washed twioe with 5O~1. portioDS of ether. The combined ethereal solutions were dried over potassium carbonate. The ether was removed through Column 1 and the residue was distilled through Column 2. Low-boiling materials (b.p.<76·) were removed at atmospheric pressure and the rest of the distillation was carried out under reduced pressure. After 0.26 g. of forerun, 1.17 g. of a mixture, containing 82.6% of cyclopropylmethylcarbinol and 17.4% of ethanol and water (lPC), was obtained with b.p. 70-7ZO (96 mm.), aaoD +18.26· (neat). The main fraction was redistilled through the same column giving 0.19 g. of forerun and 0.68 g. of a mixture, oontaining 93.7% of carbinol and 6.3% of ethanol and water with b.p. 70(90 mm.), a··D +19.46° (neat). Linear extrapolation of the two observed rotations to 100% oarbinol gave aD +20.1°. d-Cyolopropylmetbylcarbinyl Benzoate.--In a lOO-ml., round-bottomed flask, equipped with a magnetic stirrer, a reflux condenser and a calcium chloride drying tube were placed 50 mI. of dry ether, 0.30 g. (O.012 mole) of sodium hydride and 0.54 g. (0.006 mole, a··D +19.46°) of d-oyolo- ... propylmethyloarbinol. for 4 hr. The mixture was stirred under reflux The white sodium alcoholate precipitated. The -52mixture was oooled to room temperature and 1.0 ml. (0.009 mole) of benzoyl chloride was added. The mixture was stirred at reflux for 1.5 hr. and at room temperature for 12 hr. After 24 more hr., 5 mI. of water was added to decompose the remaining hydride and to dissolve the sodium lalts. The ethereal pha.e was separated, extraoted twice with 2-ml. portions of water and dried over magnesium lulfate. After removal of the solvent, the residue was dis- tilled through Column 2. After 0.05 g. of forerun, 0.52 g. (44%) of benzoate was obtained with b.p. 47-49- (1 mm.), - a··D +33.8~ (neat), n··D 1.5064. the infrared spectrum was identioal to that of an analytioally pure sample obtained from the deoompositionof N-nitroso-N-(eyclopropylmethyloarbiDll)-benzamide (below). Preparation and Decomposition of Qptically Aotive N-Nitroso-N-(cyclopropylmethylcarbinyl)-bensamide.--For each of the two expertments, the nitrosandde was prepared from a mixture of 100 mI. of a 0.5M 801ution of nitrogen tetroxide (Yatheson) in reagent-grade oarbon tetraohloride, 8.3 g. (0.10 mole) of sodium aoetate and 4.7 g. (0.026 mole) of the above d-N-(cyclopropylmethylcarbiDJl)-benzamide, stirred magnetically for 30 min. at 6·. The mixture was poured onto 100 g. of ioe and the organic layer (~8 removed -63- and extracted with water and with 6% aQueous sodium bicarbonate and then dried OTer sodium sulfate. all operations were carried out at 0-6°. The solution was decanted and evaporated to dryne8s under vacuum at 0·, leaving the orud. nitrosamide as a yellow oil. The per oent I'etention of optical aotivity in the product (oyclopropylmethylcarbinJ1 benzoate) of the decomposition of the nitrosamide was caloulated on the basis that the starting bensamide was optioally pure and the rotation of optically pure produot is a.D 33.8~. A. Deoomposition in Pentane.--The orude nitro~ide was stirred overnight in 125 mI. of reagent-grade pent~e at room temperature. The solution was then stirred with lodium hydroxide pellets to preoipitate the benzoic acid formed during the deoomposition. After filtration and distillation of the solvent, the residue was distilled through Column 2, yielding 2.00 g. (60%) of cyolopropylmethylcarbinJl bensoate, b.p. 55-58° (1 mm.), a-oD +5.26° (neat). A broad absorption band . at 3200 cm.-~ in the infrared spectrum of the produot suggested the pr.sence of benzoio aoid as an impurity. Consequently, the product was dissolved in 50 ml. of ether and the ethereal solution was extraoted with 5Q% aqueous potassium oarbonate and -64dried over sodium sulfate. Removal of the solvent and dis- tillation of the residue through Column 2 gave 1.26 g. of benzoate, b.p. 56-68- (1 mm.), a··D +5.68- (neat, 16.8% retention of optical activity), ... n··D 1.5066. the 3200 am • -1 band had disappeared. The nuolear magnetio resonanoe (~) speotrum (see p. 78) and the infrared speotrum were consistent with the assigned structure. Anal. .......... Found: B. Calcd. for Cl.Hl.0.S C, 75.76. H, 7.42 • C, 76.11. H, 7.39. Decomposition in Dioxane.--The orude nitrosamid. was stirred magnetically in 100 mI. of purified (48) dioxane at room temperature for 18 hr. The solution was then stirred with a sodium hydroxide pellet and decanted. The solvent was .tripped and the residue was distilled through Column 2, yielding 1.40 g. (35%) of liquid with b.p. 67-7ze (1 mm.). Because of the presence of an infrared ab80rption band at 3200 cm.- 1 , the product was purified by dissolving it in 40 mI. of ether and treating the solution &s in Part A, yield 0.71 g., b.p. 66- (1 mm.), a··0+7.18- (neat, 21.2% retention of optical activity), ... neeD 1.&066. the infrared spectrum was now identical to that of the analytically pure sample of Part A. Deamination of Optically Active grclopropylmethylcarbinrlamine with Nitrous Acid.--For all but the last experiment, the reaction vessel was a lOO-ml., round-bottomed flask, equipped with a magnetic stirrer, a reflux oondenser and a gas exit tube. The evolved gases were oolleoted by displaoing water fram an inverted graduate oylinder. The pH measurements were made with a Beokman Glass Electrode pH Meter, Model H, on the entire reaotion mixture cooled to room temperature. The pH of the reaction mixture was deliberately kept high to avoid acid-catalyzed raoemization of the product, oyclopropylmethyloarbinol. The decminatlonl were conduoted at reflux but were still very slow, as indicated by the rate of gas evolution, and virtually oeaaed as the reaction proceeded and the .olution became more basic. Consequently, the reaction was periodioally inter- rupted and more acid was added to lower the pH and speed up the reaotion. The pH of the solution was measured before and after the reaotion and before and after each addition of acid but only the extremes of the pH readings are reported. The total period of time under reflux and the total amount of acid used are given for each experiment. The products were isolated by oontinuous extraotion of the reaction mixture {saturated with sodium ohloride} with -~ethe~. The ethe~eal solution was d~ied over sodium sulfate and deoanted. The ether was removed through Column 1 and the residue was distilled through Column 2. All optical rotations were measured with neat samples. The ~eactants were 50 mI. of wate~, 8.7 g. (0.052 A. mole) of 60% aoueous pe~chloric acid, 4.3 g. (0.050 mole, a-oD +9.1Er) of amine and 10 g. (0.14 mole) of sodium and 870 mI. of gas was evolved arte~ 9.5 hr. of nit~ite The pH ~ange was 8.2-9.2. heating. du~e The isolation p~oc.­ gave 0.25 g. of fo~erun which was mostly methyl oyolo- p~opyl ketone as determined by infra~8d analysis and VPC. The main fraotion, 0.78 g., b.p. 69-72- (99 mm.), a"D +0.04-, had an infrared spectrum identical to that of cyclopropylmeth11ca~binol with the exception ot additional bands at 1600, 1660 and 1696 am.- 1 • The 1696 am.- 1 bana . oorresponded to the carbonyl ablorption band of methyl -1 cyc10p~opyl ketone and the 1600 and 1660 cm. att~ibuted to the N-H deformation mode of cyclopropy1- metby1oarbiny1amine. bands were No satisfaeto~ method of analyzing for small amounts (1-2%) of the amine (or the ketone) in the p~esence of the ca~bino1 by VPC could be developed. Howeve~, the concent~ation of the amine in the p~oduct wal estimated as 2% by oomparing the infrared 8pect~um of tbl -57produot with those of mixtures of amine and aloohol of known conoentrations. B. The reactants were S.6 g. (0.052 mole) of per- chlorio acid, 4.9 g. (0.050 mole, a··O +9.lS-) of amine and 5 g. (0.07 mole) of sodium nitrite in 40 mI. of water and 9S0 ml. of gas was oolleoted over 11 hr. of heatipg. The pH range was 8.3-8.9. The produots isolated were 0.19 g. of forerun, mostly alooho1 with same ketone (VPC), and 0.67 g. of alcohol oontaining 1% of amine (infrared), b.p. 70-72- (98 mm.), a·~D -0.04-. C. The reactants were 50 mI. of water, 8.9 g. (0.058 mole) of perohlorio aoid, 4.3 g. (0.050 mole, aeoO -9.67-) of amine and 10 g. (0.14 mole) of sodium nitrite and 1050 mI. of gas was evolved over 21.5 hr. of heating. range was 7.6-9.4. The pH The usual isolation prooedure (with the exoeption that the ether was removed through a 50-em. Helipak-packed column) gave 0.24 g. of forerun, mostly ketone (VPC), and 1.39 g., b.p. 70· (95.6 mm.). The main fraction was redistilled fram 0.5 g. of adipio acid (to remove amine) through Column 2, giving 0.19 g. of forerun and 1.01 g. of aloohol oontaining 1% of amine (infrared), b.p. 7ct (95 mm.), a··O +0.27-. -58D. The apparatus wae a lOO-ml., three-necked flask equipped with pH electrodes, a magnetic stirrer, a reflux oondenser, a thermometer and a gas-collecting system. To the flask were added 40 mI. of water, 8.0 g. (0.048 mole) of acid, 4.3 g. (0.050 mole, aaon +9.18-) of amine and 5 g. (0.07 mole) of sodium nitrite. The electrodes and the thermometer were adjusted so that they were well below the surface of the solution but did not interfere with the magnetic stirring bar. the reaction. Stirring was maintained throughout At 26-, the pH was 8.0. The solution was heated to 80- with an 011 bath over a 10-min. period, during which time 46 mI. of gas was oollected. The pH meter now read 7.8 without adjusting the temperature compensating oontrol. The solution was held at 80· and additional acid was added periodically, through the condenser, closing the system as 800n as possible thereafter. By this method, the reading of the pH meter was maintained between 6.0 and 7.8. After 10 hr., 1200 mI. of gas had been oollected and the meter read 6.4 (5.6 after cooling the solution to room temperature). Isolation of the products gave 0.50 g. of forerun, 1.61 g., b.p. 72-101- (98-81 mm.) and 0.28 g., b.p. 6tr (11 mm.). The main fraction contained 14% of cyclopropylmethylcarbinol and 75% of an unknown compound, -69A (VPC). The last fraction contained 3% of the aloohol and 91% of A (VPe) and its infrared speotrum showed no prominent absorption in the region of 1600-5000 em. -2, for bands in the C-H stretching region (2800-3100 om. exoept (:. a,. ). On the basis of the h1ffi (fig. 1) and infrared speotra of the last fraction, A was assigned the struoture bls-(oyclopropylmetby1oarblnyl) ether. Raoemization of Cyclopropylmethy1oarblnol under the Conditions of the Amine-Nitrous Aoid Reaotion.--To 50 mI. of water were added 6.4 g. (0.088 mole) of 60% perohlorio (0.11 mole) of sodium nitrite and 1.21 g. of cyclopropyl- acid, 5.0 ml. (0.039 mole) of d,l-l-phenyletbylamine, 7 g. methyloarbinol, aeaD -1.36- (neat). 7.0. The solution was heated under reflux for 6 hr. and evolved 965 mI. of gaB. 8.9. The initial pH was The cooled 8olution was now pH The usual isolation prooedure gave 0.08 g. of forerun and then 0.75 g. of cyolopropylmethylcarbinol, b.p. 70-71(101 mm.), a·eD -1.14- (neat), corresponding to 84% ret.ntion of optical aotivity. The still residue contained 4.6 g. of a red liquid. Deamination of Qptioally Aotive l-Phenylethylamine with Nitrou8 Acid.--The amine was resolved by the method of Theilacker and Winkler (49). The products were isolated by continuous extraotion of the reaction mixture with ether; the ethereal solution was then dried, stripped of solvent through Column 1, and the residue was distilled through Column 2. It was found that 1% of the starting amine in the produot could easily be deteoted by vpC by examining mixtures of amine and aloohol of known composition. Some of the products contained appreoiable amounts of amine whioh could be removed by redistillation fram adipic acid through Column 2. The final samples of l-phenylethanol contained much less than 1% of amine unless otherwise indioated. All pH measurements were made at roam temperature with a Beckman pH Meter. All optical rotations were measured with neat liquid.. The stereochemical results were calculated using a value of a-·D 38.80° for the optically pure amine (49)J «gOD 48.44° was used (50) tor the alcohol. Amine and alcohol ot the s~e sign ot rotation have the same configuration (51). A. To 50 ml. of water were added 5.2 g. (0.031 mole) of 60% perchloric acid, 5.0 IIll.(O.039 mole, a·eD +26.46-) of amine and 7 g. (0.11 mol.) of sodium nitrite, the pH was 8.4. The solution was stirred magnetically at reflux for 20 hr. raising the pH to 8.9. The isolation proc.dure gave 0.13 g. of fo~erun and 2.84 g., b.p_ 68-7~ (4.5 mm.), (, -61- a· 7 D +3.15°. Redistillation fre. 1 g. ot adipic acid gave 0.17 g. ot forerun and 1.11 g. of pure l-phenylethanol, b.p. 63° (2.8 mm.), ~··D 1.5253, a··n _1.57° (5.5% net inver.ion of configuration). The reported values are b.p. 100· (18 mm.), ~.oO 1.5211 (60). B. The reaction mixture waa 35 ml. of water, 5.9 g. (0.035 aole) of perchloric acid and 6.0 mI. (0.039 mole, a·oO -22.18°) ot amine and 7 g. (0.11 mole) of sodium nitrite. Alter 0.07 g. of forerun, 3.43 g., b.p. 77_83° ~ (6 ma.), a··O .1.42°, was obtained. Rediltillation from adipic acid gave 0.04 g. of forerun and 1.75 g. of pure - alcohol, b.p. 61_62° (2.4-2.8 ... ), n·an 1.5250, a·on +1.30° . (5.2% LaYeraion). c. The reaction mixture contained 36 ml. of water, 7.0 g. (0.042 mol.) of acid, 5.0 al. (0.039 mole, a·on -22.180 ) of amine and 7 g. (0.11 mole) of sodium nitrite and ... stirred magnetioally at reflux for 4 hr. Atter 0.24 . g. of forerun, 3.34 g., b.p. 89°(11.5 lIIIl.), a·on +1.04·, was obtained. Redistillation of the main fraction from adipio acid gave 0.04 g. of forerun and 2.42 g. of pure alcohol, b.p. 63° (3.0 ma.), n··D 1.5260, a·oD +1.08° (4.3% Inversion). -62D. To 50 mI. of water were added 7.84 g. (0.047 mole) of aoid, 5.0 ml. (0.039 mole, ~·D +25.46°) of amine and 7 g. (0.11 mol.) of sodiua nitrit.. The solution was stirred tor 15 .In. and allowed to stand overnight. The isolation prooedure gave 0.17 g. ot torerun and 3.20 g. of pure alcohol, b.p. 72_76° (4.5 ma.), a·~O _2.11° (7.4% inversion) • i. In 100 mI. of water were dissolved 3.1 g. (0.026 mole, a··n _22.18°) of amine, 69 g. (0.5 mole) of sodium dihydrogen phosphate monoh,urate and 8.0 g. (0.2 mole) of sodium bJdroxide; the solution had pH 5.9. A. solution ot 6.0 g. (0.087 mole) of sodium nitrite in 15 ml. of water was added. The 801ution had pH 6.0 after 3 hr. at reflux. After 1.07 g. of forerun, 1.44 g. of pure alcohol, b.p. 75-78° (6.5 ma.), a··O +2.92° (11.6% inversion) was obtained. F. To 150 m1. of water were added 104 g. (0.76 .01.) of sodie dibydrogen phosphate monobJdrate, 12.4 g.(0.31 mole) of sodium hydroxide, 5.0 ml. (0.039 mole, a··D -19.240 ) of amine and 7 g. (0.11 mole) of sodium nitrite; the pH was 6.05. The solution was heated to reflux for 2 hr. and after cooling the same over a 3-hr. period, had pH 6.07. Alter 0.15 g. of forerun, 3.33 g. of pure aloohol, b.p. 71-75° (5.5 am.), a-°O +2.87° (13.2% inversion of configuration), was obtained. .63 .. G. The reaction mixture oonsisted of 150 ml. of water, 62 g. (0.38 mole) of 80dlua dlhJdrogen phosphate monohJdrate, 6.0 g. (0.15 mole) of todium hydroxide, 6.0 ml. (0.039 mole, ~·D +25.46°) of amine and 7.0 g. (0.11 mole) of sodium nitrite and had pH 6.2. The solution was heated under reflux tor 2 hr. and allowed to stand at room temperature tor 12 hr.; the pH wa. unchanged. iftel' 0.36 g. ot forerun, 3.33 g., b.p. 73-7r- (5.5 mm.), a·eD -3.88°, was obtained. Redistillation of the main fraotion fro. adipic acid gaTe 0.53 g. ot forerun and 1.80 g. of pure aloohol, b.p. 56° (2.0 ma.), ~··D 1.5256, a·oD _3.91° (13.6% IDYersion). H. The reaotion mixture 1IU the SIllll8 as that of experiment F (exoept that the sodium hydroxide was omitted) and had pH 3.95. After heating the solution under retlux for 4 hr., the pH was 4.20. The isolation routine gaTe 0.S8 g. ot forerun and 2.32 g., b.p. 65_70° (4 ma.), a··D -0.23·, ot aloohol contaiAing 1% of amine (VPC). 4-Pzridllpzridinium Dichloride was prepared by the method of Bowden and Green (52). 1ro. 1000 g. of technioal- grade thioll11 ohloride and 340 ml. of p;yridine was obtained 293 g. of a golden-brown powder, m.p. 160_166° (lit. m.p. 158_160° (52». .64.. 4..!b'droxmridine was prepared b1 the .. thod ot Koenigs and Grebsar (53). From 120 g. ot 4-P11'id11ppidinium diohloride waa obtained 34.3 g. (69%) of tan cr7Stala. 4-ChloroPllidine waa prepared b1 the method of Leia and Curran (64). From 38 g. at phosphorous pentaohloride, 38 g. of phosphorous oxychloride and 33 g. of 4.hydroxypyridi~ was obtained 24 g. (61%) ot produot, ~··D 1.5290 (lit. ~··D 1.S28O (54», which waa .tored at _50 to preTent po I1J18rization. " 4-Beullox1Pyridine was prepared atter the method of Shaw (55) and bad a.p. Sl.6.S4.8· (lit. m.p. 5S_56° (55». N-Mathll-4-beulloxlpyridiDium Iodide.--On a steam bath were heated 0.50 g. of 4..beu11oX1Pyridine and 0.50 g. ot metbTl iodide. The solution loliditied atter a tew minute.. The 801id was or7Stalllsed tram ethanol-etbTl acetate giving 0.70 g. (79%), m.p. 146.8-149.8°. Two reor1ltal1izatioDB from absolute ethanol gave 0.6 g. ot light.tan crJltals, m.p. 149.4.150.8- atter vacuum dr71ng. - Anal. Calod. tor C~.~.INOt I, 38.79. Found: C, 47.72; H, 4.31; C, 47.34J H, 4.49, I, 39.09. N.YetBrl-4-beDSyloxzpzridlgium Perchlorate was prepared troa 88 me. at N-methfl-4-beDS11oX1P11'idinlum iodide dis'OITea in 6 mI. of 80% ethanol to which was added a solution -65of lilTe~ pt~chlorate in ethanol until no more yellow preoipitate would fo~. The solution was filtered from the preoipitate and evaporated to drmees under vacuum. The reaidue was dissolved in 10 ml. ot methylene chloride, filtered, and ether was added to the cloud point. Cooling gave 41 ag. (51%) of white orystal., m.p. 141.4-142.2° after vaouum drying over phosphorous pentoxide. - Anal. Ca1cd. for C~.H~.ClNO.: Cl, 11.83, N, 4.67. Found: C, 62.09; H, 4.71, C, 51.66, H, 4.69; Cl, 11.96, N, 4.89. 4-(Crolopropzlcarbinyloxll-pzrldine.--In a 50-..1., round-bottomed flask, proteoted by a caloium chloride drying tube, .ere placed 2.4 g. (0.032 mole) of clcloproPl1carbinol, 0.76 g. (0.032 mole) of sodium hJdride and 10 lIl. ot purified (56) d1methfl aultoxide. The mixture wu stirred magneticalll for 1 hr., when 3.2 g. (0.028 mole) of 4-chlo~opyrldin. was added, and 8ti~~ing was continued for 8 hr. The red solution was tiltered and the precipi. tate . . .abed with hot benseM, The dimethyl sulfoxide solution and benzene washings .ere combined, the b.nlene was removed under vacuum and the residue was distilled through Column 2 at 1 Mm. After the d~thll sulfoxide (b.p. 31°) was distilled, 3.73 g. (89%) ot a colorless liquid, b.p. 57 0 (1 ma.), was obtained which was redistilled through the .... column giving three tractions, b.p. 85- •• ~ D 1.5268, was ana17zed. (3 _.), n··D 1.5214-1.5260. The third fraction, !pal. Founds Galed. for C.~lHOs C, 72.45; H, 7.43; H, 9.89. C, 72.83; H, 7.82; N, 9.29. The NMR spectrum ot the third traction was consistent with the &aligned structure (se. p. 78). l-Methll-4-(cxclopropxlcarblnrloxr)-pyridin1um Iodide was prepared from 1.5 g. of 4-(cyclopropylearbinrloxy)pyridb.• and 2 ml. ot methyl iodide in a tightly stoppered test tube. The exothermic reaction was moderated by peri- odica1l7 cooling the tube in lce water. The mixture bee... thiok and then crystallized. Reor7ltalllzation trom meth. ylene ohloride-ether gave 2.00 g. (68%) of yellow orystals, m.p. 110.4-111.10 atter vacuum dJt71ng. The mm spectl'Ulll was oonsiatent with the assigned .truoture (see p. 78). C, 41.25; H, 4.85; I, 43.59, H, 4.81. Founds C, 41.40, H, 4.85; I, 44.791 Anal. Calcd. for Clo~.INO: N, 5.04. N~thll-4-(Clc1opropl1carbiDlloXl}-Blrldinium Per- chlorate wa. prepared from 0.40 g. of H-meth71-4-{cycloproP11carblD71ox7)-pyridin1um iodide dissolved in 10 ml. -67of absolute ethanol to which a .olution of silver perchlorate in ethanol was added until no .or. 78110w precipitate would form. The Jdrlure was boiled lAd filt.red and the precipi- tate ... washed with 10 ml. of hot .thanol. The combined eth&nollo filtrates were evaponted to 10 mI. under vacuum whereupon a whit. precipitate .'parated. Subsequent boiling and cooling gaTe 0.28 g. of whit, or,atals, which, alt.r vacuum ar,ing; had m.p. 87.0-87.6- (cloudy). The produot was taken up in 2 ml. of meth11••• ohloride and filter.d fro. a ..all a.ount of grey insoluble aaterial. Addition of ether to the filtrate and cooling cav. 0.26 g. (72%) of material having m.p. 87.6-88.0· alt.r vacuum drying. !!!!. Caled. for C~o~.CllOe: CIt 13.45, N, 6.31. Found: C, 45.55; H, 5.35; C, 45.80, H, 5.52; Cl, 13.50, N, 5.80. 4-(Czcloproprlmethrlcarbi!lloXl)-pzrldine was prepared in a eimilar manner to 4-(c101oproP11oarbinyloxy)-pyridine. From 5.6 g. (0.065 lIlole) of ,cycl,propylmetbyloarbinol, 1.56 g. (0.066 .ole) of sodium hydrid., 20 ale of purified dlmethTl sulfoxide and 7.4 g. (0.OS5 ao1e) of 4-chlorop)Tidlne ... obtained 9.1 g. (86%) of a oolorle •• liquld, b.p. 72° (1 _.). Redistillation gave three fractiona, b.p. •• ~ D 1.5160, ... redistilled fro. barium oxide and analyz.d. 77° (2 am.), n··D 1.5130-1.5160. The third fraction, -68. - Anal. N, 8.58. Caled. for C~OH~8NOl Found: C, 73.59; H. 8.03; C, 73.91; H, 8.02, N, 9.03. The mm lpeotrua ot the produot was consistent with the utigzwd ItruotUI"8 (se. p. 78). N-Me t hll -4.(czolopropzlmetBrloarblnlloXl)-prrldln1ua Iodide.--In a Itoppered flask were placed 2.94 g. of 4-(oroloproP1l.methylcarbiIl11ox1)-p11'ldine and 3 Ill. ot lIeth71 10dlde. The exothermic reaotion wa, moderated b1 cooling the tlask under a stream of water. The exoesl meth71 iodide was remoyed under vaouum leaYing 5.74 g. (104%) ot & y1acou8 onnge oil. It . . known trom pre- r.limina17 experiments that this 011 could not be induoed to crystallize trom methylene chloride-ether; however, its NJ.IR speotl"Ua _I consistent with the assigned structure (see p. 78). The 011 was dissolved in 15 ml. ot reagent. grade acetone and tlltered. C~ta1l18atlon wal induced b1 oooling the solution in a D17 Iee.acetone bath with vigorous scratohing. Just atter the first crystal appeared, the solution wa, cooled at 0°, and or,atalllsation 810wly continued. The white orystals were oolleoted in a 2-ml., ainte1'ed-glalS Beuchner tunnel. The funnel was qulck11 transter1'ed to a stoppered, 15~. centrifuge tube and was oentrifuged tor 5 min. in a centrifuge cooled by placing -69- Dr1 I.e in the guard bowl. The centrlfuged crystals, 1.8 g., .. re quloklr placed in a vao\lUll d.. ic08.tor where the1 began to luse. WAS Betore fusion was ooap1ete, the produot oooled and 1.2 g. of acetoDi W&. added. The mixture was 8ti~d until solution was almost complete and oooled again. The precipitate that tormed ... collected as betore and ... plaoed in a vacuum desiocator where it melted to a sllght11 18llow oil weighing 0.90 g. Two more or1Btal- lizations lrom acetone gave 0.40 g. of a near11 colorles. oil which was dissolved in methrlene chloride and lilterea to remove a te. particles ot toreign material. The methylene ohloride solution was platea in a vaouum desiccator over phosphorous pentoxide tor tour da78, leaving 0.41 g. ot oil whioh waa analyzed. N, Found: C, 43.49; H, 5.58, '1 , Anal. Calod. tor CliHl.INC: C, 43.29; H, 6.29; I, 41.69J 4.59. 41.23, N, 4.52. }(etatbes1. ot samples ot the .thiodide with silver perchlorate and with ailver trltluoroaoetate produoed colorless 011e .hioh could not be induoed to cr1ltallize. 4-(l...Phenxlethop)-Puidine . 8 prepared in a aimilar manner to 4-(c1010proP11carblnTlor,y)-pyridlne. From 5.1 g. (0.042 mole) of I-phen11etbanol, 1.0 g. (0.042 .ole) of sodia. h7dride, 10 ml. ot purified dimethyl sulfoxide and .. 70. 5.1 g. (0.045 mole) ot 4-chloroP1?ld1ne was obtained 7.2 g. (86~) ot a viscous ,.110. liquid, b.p. 120° (1 ma.). The t.' product waa retraotionated and a nearly colorless central cut, 5.0 g., had b.p. 94· (1 ma.), a··D 1.5644. N, 7.03. Found: C, 78.02; H, 6.79; N, 7.04. Anal. Calcd. for C1 .H1 .NO: C, 78.36; H, 6.57; The anal,tlcal sample darkened upon standing unless sealed under Ditrogen. The NMR speotrum of the product was consiltent with the assigned struoture (see p. 78). The reaotion of the alkox1P71'id1ne with meth71 iodide produced a vi.coua orange oil which could not be induced to cryatallize although its NWR lpeotrum (see p. 78) was conai.tent with the structure N... th1l-4-(1-phenr1ethoxT)pyridlnium iod1de. From 1.00 'g. ot the alkox1Pyridine waa obtained 1.73 g. (101%) ot an o~ oil after removing the .%oe" .. thyl 10dide under yaouua. Th. oil was dis- solv.d in 1.90 g. of glacial acetio aoid and heat.d on a steaa bath tor 4 hr., at which tlJu two phases toratd. The lower pbue solidified upon. cooling the system to room temperature. The precipitate wu oollected and waahed with ether and had a.p. 129-187°. ReorJatallization tram aoetone gave 0.14 g. ot yellow crystals, a.p. 125-137°. The liquid phase ot the reaction mixture was dl.tl11ed through a -71- simple head at 1 mm., the reoeiTer being oooled in Dry Ioeaoetone. Examination of the dbt1llate b;r VPC showed acetic aoid and one other peak. The uaknown peak was colleoted on a prepa~tiT' scale giving 182 mg. of a yellow liquid. Repurifioation by preparative VPC failed to remove the yellow 00101'. The unknown material .... assigned the 8truOture l-phenJlethyl acetate on the baai. of it. NMR spect~ (see p. 78). Optically Active 4-(CyoloproPllmetpyloarbinyloXl)pzridlu wal prepared in a sWlar fashion to the l'aceaic compound. From 3.0 g. (0.035 mol., aaYD .3.39°) of C1Olo- propylmeth71carbinol, 0.84 g. (0.035 mole) of lodium hydrid., 15 ml. of purified dlmethTl sulfoxide and 4.0 g. (0.035 mole) of 4-ohlol'opyridine ... obtained 4.1 g. (72%) of pl'oduot, b.p. 65_66° (1 mm.). Redistillation of the product gave 0.57 g. of fOl'erun and 3.40 g., b.p. 106.107(6.5 sm.), ~··D 1.5156, a··D -0.87- (neat). Preparation and Hydrolysis ot Optically Aotive N-Methyl-4-(cyclopl'°Pllmethyloarbl!l10xy)-pyridine.--The reaotion of 2.37 g. ot the above optically aotive 4-(orelopropylmetbylcarbinylox;r}-pyridine with exoe81 methyl iodide gave 4.36 g. (98%) ot a visoou. orange oil, [a)··D +0.3 0 - =6.S), atter removal of the excess methyl (ehlol'otorm, 0 c· -72iodide in a vacuum desiocator OYer phosphorous pentoxide tor flve days_ The infrared splotrus (neat) of the oil was at.ilar to but not superimposable with that of the raoemic methiodide _ The variations .. re minor exoept tor a broad band at 3440 em. -J, which wu such stronger in the speotrum ot the raoemic aethiodlde. The optlcallr active metblodide (3.85 g.) was dissolved in 50 al. of triply distilled water and 2 g. ot reagentgrade lithiu. carbonate was added to bulfer the solution. The mixture was stirred magnetioally at room temperature tor 96 hr. and waa then continuoualy extraoted with ether. The ethereal extraots (75 mI.) were dried over sodium sultate and then magne8ium 8ultate, stripped of solvent through Column 1 and the residue was distilled through Column 2, giving 0.57 g. (53%) ot product with b.p. 70° (98 DB.), -ne·D 1.4373, a·oD +0.15 +- 0.05° (neat). The infrared speotrum ot tM product was identioal to that ot an authen. tic sample of oyc10propylmethyloarbin01 exoept tor the presenoe of additional, weak abaorptlon band8 at 965, 1170, 1240, 1645, and 1715 ClIl. -l,. The latter two bands were broad and highl.f sugge.tin of an amine. Solvolxala of N-Methll-4.(ololopropyloarblnyloxy)pyridlnein 80% Ethanol.--! mixtur. of 51.8 ag. of the methlodide, 0.73 mI. of absolute ethanol and 0.15 sl. of water was .ealed in a glasstube and heated at 120 0 (bath temperature); the solid d1810lTed. Atter the solution was heated for 18 hr. , it was anal)'lled by VPC under the operating conditions used by Graham (4) tor the anal;ysit!l ot the products trom the solvolysis of oyelopropyloarbinyl ohloride. Three major peaks aPP'ared in the chromatograa, oorresponding in retention times to the ethyl ethers ot oyclopropyloarbinol, oyclobutanol and allyloarbinol. The cyolobut11 and oyclopropyloarbin,l ethers were poorly resolved, but oould be separat.d under different operating conditions. The peaks were cut out of the chromatograph and weighed; the ratio ot the weights of the three ethers was oyo1opropyloarbinyl:oyolobutylsa11ylcarblnyl = 3.6:1.5:1.0. Peaks with retention times equal to those of C101obutanol, oyclopropyloarblno1, and allyloarbinol were alao deteotable but were too amall to measure. Reaction of N-Methll-4-benszlo!Jplridinlum Iodide and Peroblo~te with Glacial Acetio Acld ••-In 0.98 g. ot glaoial aoetio aoid was dissolved 96 mg. of the methiodide by warming the mixture on a steam bath. In capillary tubes were sealed approximately 30 ~. samples and the tubes were heated in an oil bath at 109°. Capillaries were -74- periodically withdrawn and their contents analyzed with a F and )( Programmed Temperature au Clutomatograph, Model 202, using a silicone rubber oollDlD.. The appearance of benzyl acetate was measured by comparing the ohromatographs to that of a solution of benzyl aaetat. in acetio aoid ot known oonoentre.tion. By thls uthod the half.life ot the reaotion wa. utimated as 2..3 hr. A 34-ag. lample of the perchlorate was treated in the same way. Little or no benzyl aoetate could be detected after 36 hr. ot heating. lor oompu1son, bens)''! cbloP1de was found to be unreactive under the same conditiona. Reaction or N-Methll-4-(ololopropzlmetezlcarbiDlloxt)pyridin1um Iodide with Ethanol.-Themethiodide (0.G9 g.) was a Imple ot unknown purity that had been recrystallized trom aoetone and dried under vaouua. A solution of the methiodide in 0.69 g. of absolute ethanol was aealed in a gla81 tube and heated for 80 min. at 98.4°. The resulting orange solution deposited a yellow precipitate when cooled to room temperature. The volatil•• were removed under vacuum and aollected in a Dry Ice.aoetone trap. The residue was cr1Stalllsed trom ethanol giving 0.18 g. of yellow cr18tal., m.p. 141.5-149.5°. Recrlltallization trom ethanol -75gave 0.10 g., m.p. 145-149°, which was triturated with boiling acetone in which it was insoluble. The vaouum- dried yellow crystals, m.p. 148.0-149.7°, were soluble in c ethanol and the solution gave an immediate yellow preoipitate with alcoholic silver perchlorate. A solution ot the orysta1s in water, in whioh it was very soluble, turned pH paper red (pH 2). Anal. .......... Found: C, 35.51; H, 3.54 • The volatiles from the original reaction (0.73 g.) were examined by VPC and were found to oontain one product besides ethanol. The product was separated by preparative VPC, giving 66 mg. ot a colorless liquid which was assigned the structure cyclopropylmethylcarbinyl ethyl ether on the basis of its NMR speotrum (fig. 1). Reaction of N-Methyl-4-(cyclopropylmethy1oarbinyloxy). pYridinium Iodide with Glacial Aoetic Aoid.--A sample of the methiodide of unknown purity was dissolved in glaoial aoetic aoid and was heated at 125° (bath temperature) for 8 hr. The solution became dark red. The volatiles were transferred to a Dry-Ice trap under vacuum. The dark-orange crystalline residue was crystallized trom ethanol giving yellow crystals, m.p. 147.4-149.8°. A mixed melting point with the crystalline produot from the reaction ot the methlodlde with ethanol (above) was 147.8-149.7°. -76- Anal. - Found: I, 48.18; N, 7.72. The volatiles were examined by VPC and were found to contain, besides acetic acid, four major products, none of whioh oorresponded in retention time to cyclopropylmethylcarbinyl acetate. The reaction was repeated with 0.28 g. of methiodide and 0.32 g. of glaoial aoetic acid at 87° for 12 hr. The non- volatile portion of the product was 0.15 g. of a dark-red, partially orystalline oil. examined by VPC. The volatiles (0.32 g.) were Three of the four peaks of the ohromato- graph from the 1250 reaotion were present but in different ratios. ! new peak appeared which had the same retention time as cyclopropylmethylcarbinyl acetate. Hydrol181s of N-Methyl-4-(clclopropylcarbinyloxy)pyridinium Iodide.--! solution of 25.0 mg. of the iodide in 83.2 mg. of triply distilled water was sealed in a glass tube and heated in an oil bath at 98.6 0 for two days. The reaotion mixture was examined by VPC and was shown to oontain, besides water, allylcarbinol, cyclobutanol, and cyolopropylcarbinol, although the latter two aloohols were poorly resolved from each other. The peaks of the ohromato- gram were weighed giving a ratio of oyclobutanol and cyclo. propyloarbinol to allyloarbinol of 3.8 to 1. The ratio -77of oyclobutanol to cyolopropyloarbinol was judged, visually, to be 10 to 1. That this ratio of alcohols resulted from rearrangement of the initially produced alcohols was indioated by adding 1 ~. of cyclopropylcarbinol to about one-third of the spent reaction mixture, sealing in glass, and heating tor two days at 98.6°. The ratio of aloohols was essentially unchanged. Consequently, a mixture of 12.2 mg. of iodide, 7.3 mg. of lithium carbonate and 48.0 mg. of water was sealed in glass and heated for two days at 98.6°. The conoentrations of the alooholic produots were estimated by VPC and a solution of the alcohols in water ot the estimated ooncentrations was made. Comparison of the ratios of peak areas of the chromatograms of the reaction products and of the solution ot known concentrations gave a more refined estimate. The total yield of alcohols was 99%; the ratio of oyolobutanol and cyclopropylcarbinol to allylcarbinol was 12.0 to 1; the ratio ot cyclopropylcarbinol to cyc1obutanol was 1.2-1.5 to 1. N-l4ethzl- 't -pyridone was prepaJ"ed by the method of Toomey and Riegel (57). The extremely hygroscopic solid had m.p. 96.2_98.4° (sealed capillary), lit. m.p. 92_94° (58). -78- In an attempt to prepare the bydroiodide, 0.10 g. of the pyridone and 0.69 g. of 48% aqueous hydrogen iodide were mixed and the water and excess hydrogen iodide were removed in a vacuum desiccator OTer phosphorous pentoxide. The yellow crystalline residue had m.p. 131-140°. mm spectra were observed at 60 140. with a Varian Model V4300B spectrometer equipped with a Super Stabilizer. The samples were either neat liquids or solutions in methylene ohloride with tetramethylsilane as an internal standard. The chemioal shift between two prominent peaks in each spectrum was measured by the audio side-band superposition method and other chemioal shifts were obtained by int.rpolation or extrapolation, assuming a linear sweep-rate. The proton resonances of the following compounds were observed: 4-{cyclopropy1carbinyloxy)-pyridine and methio- dide, 4-(oyolopropylmethylcarblnyloxy)-pyridine and methiodide, 4-(1-phe~lethoxy)-pyridine and m.ethiodide, cyclopropylmetbylcarbinyl ethyl ether, bis-(cyclopropylmethyloarbinyl) ether, cyclopropylmethyloarbinyl benzoate, l-phenyletbyl acetate and I-phenylethanol (oontaining one drop of cono. hydrochloric acid). For the two compounds containing the cyclopropylcarblnyloxy structure, the -79seoondary ring protons appeared a. a highly split pattern oenteHd at 9.4 and 9.6 7'*; the te!"tiary ring proton appeared as a quartet, eaoh peak of which was split again, at 8.7 and 8.8~; at 6.3 and 5.7 T appeared a doublet for the methylene group. In the case ot the five cyclopropyl- methylcarbinyloxy struotures, the secondary ring protons were at 9.4-9.7 r ; the tertiary ring multiplet was at 8.8-9.3 r, partially obscured by the methyl doublet at 8.5-8.9 T ; the tertiary proton was a quintet at 5.5-7.1 'T. The four l-phenylethoxy structuHs had phenyl multiplet! at 2.4-2.8 ~, tertiary proton quartets at 3.9-4.7 'T and methyl doublets at 8.1-8.3 r. The a!"omatic protons ot the three alkox1Pyridines appeared as a doublet at 1.6-1.7 ~ and a doublet at 3.2-3.3 7; in two ot the cases, each peak of the doublets was resolvable into two lines. In the case of the three methiodides, the protons of the pyridine ring appeared as doublets at 0.8-0.9 and 2.4-2.5 ~; the N-methyl group was a singlet at 5.5-5.6 'T. Two spectra are worthy of special comment. The tertiary proton ot bi.-(cyclopropylmethylcarbinyl) ether (tig. 1) was observed to be a quintet at 7.0 r with eaoh peak of the quintet split into two linea by about 3 c.p.a. This .. r (in p.p.m.) = 10.0 - 10e( V obs - VYe.Si )/YYe.Si. -80- 6.3 8.6 7.1 8.7 9.4 I 7.0 8.9 9.7 Fig. 1. - Proton nuclear magnetic resonance spectra at 60 Mc. Chemical shifts are in p.p.m. relative to tetramethylsilane = 10.0 as internal standard (not shown). Top: cyclopropylmethylcarbinyl ethyl ether in methylene chloride (not shown). Middle: region of top spectrum near 6.3 ~ at slower sweep. A satellite resonance produced by ~3C in the solvent is indicated. Bottom: bis-tcyclopropylmethylcarbinyl) ether (neat). -81- additional splitting is understandable if the ether is an approximately equal molar mixture of the ~ and ~,l forms and the tertiary protons of the two forms have slightly different chemioal shifts. The ten-line pattern is then a superposition of the five lines of each diastereomer. The methyl protons of the ether, farther removed from the oenter of asymmetry, show the same effect but to a smaller extent; both peaks of the methyl doublet at 8.9 ~ are split by about 1 c.p.s. \¥hile the methyl protons of the ethyl group of cyolopropylmethyloarbinyl ethyl ether (fig. 1) are unexceptional (triplet at 8.7 ~), the methylene group at 6.3 ~ is a very oomplioated pattern whioh can be explained if the two methylene protons are unequivalent due to the proximity of an asymmetrio center (59). The obserTed spectrum agrees well with that oaloulated for an ABC. system with ~B = 9.4 c.p.s., ~C = 7.35 c.p.s., ~C =6.68 o.p.s., and SAB = 9.0 o.p.s. We are indebted to Mr. Donald R. Davis for recording the spectra in fig. 1 and to Dr. Stanley L. Manatt and the Computing Center of the Jet Propulsion Laboratory for IBM 704 oalculations of oomparison speotra to oheck the chemioal-shift and spinooupling parameters for oyclopropylmethylcarbinyl ethyl ether. -82Kinetics of the SOlvolyses ot the N-Methyl-4-alkoxypyridinium Salta.--! stock solution ot 80% ethanol was made up tram 400 ml. ot purified absolute ethanol (kindly supplied by Mr. E. F. Kiefer (eo» boiled distilled water. and 100 mI. of freshly For solvolyses in purely aqueous solvent, triply distilled water was used. The reactions were tollowed titrlmetrically with O.026M aqueous sodium hydroxide, made from the appropriate amount of 50% aqueous sodium hydroxide filtered into freshly boiled distilled water. The base was standardized against potassium aoid phthalate before and after the two-month period during whioh the rate measurements were made. The titer decreased by 1.5% during this period and the base oonoentration at intermediate times was determined by a linear interpolation of concentration with time. All titrations were performed in a rubber-stoppered beaker protected from atmospherio carbon dioxide by an Ascarite tube and fitted with pH electrodes, a magnetic stirrer, and a I-mI. Kooh burette with a capillary tip extending below the surfaoe of the solution being titrated; a Leeds and Northrop pH Indicator was used. The temperatures of the constant-temperature baths were measured with a National Bureau of Standards standardized thermometer. The temperature variation during -83- a single run was no more than !9.04° as observed with a Beckmann thermometer. For the experiments at 71 and 96°, a weighed sample of salt was transferred to a volumetric flask and diluted to the mark with the appropriate solvent. Approximately 5.5-ml. samples were sealed in treshly cleaned and dried glass ampoules. After the ampoules were placed in the bath, a ·zero time" sample was withdrawn when the bath re-attained equilibrium (about 5 min.). Ampoules were periodically withdrawn, quenched in ice water, brought to room temperature and a 5-ml. aliquot of the contents was added to 20 ml. of carbon dioxide-fre. distilled water. The endpoint of the titrations was taken as the pH corresponding to the point of maximum slope in a plot of pH !!. volume of base, determined for a tew ot the points in each run. Although the endpoint ranged trom pH 7.0 to 7.4 during the various runs, the variation during a single run was no more than 0.2 pH units. Solvent "blanks· were treated in the same manner as the kinetio samples; the blank corrections were negligible in all oases. Solutions of the iodide aalts turned noticeably brown during the kinetic runs; the perchlorate salt solutions remained colorless. For all the experiments, the data were treated by oalculating a first-order rate oonstant tor eaoh sample relative to the ·zero time" sample. The ·constants· dete~ned in this tashion tell off seriously after the first or seoond halt-life. Consequently, the samples where this decrease was notioeable were ignored, and the remaining constants were averaged to give the final rate-oonstant. The kinetic data are presented in Tables IV-IX and the individual experiments are discussed below. N-Methyl-4-(clolopropzlcarbinyloxy)-pyridinium Iodide in 80% ethanol at 98.4° gave an infinity titer of 89.4% after 72 hr. at 119_121°; in water at 98.6° gave infinity titers of 90.7 and 91.9% after 49 hr. at 98.6°; in water at 70.7° gave an infinity titer ot 89.4% atter 54 hr. at 100°. N-Kethyl-4-(oyolopropylcarbiPlloxy)-pyridinium Perohlorate in 80% ethanol at 98.4° gave an infinity titer ot 84.0% after tour days at 119_121°; in water at 98.6° gave an infinity titer of 90.7% after 49 hr. at 98.6°. N-Methyl-4-(cyelopropylmethylcarbinyloxy)-prridinium Iodide.--! sample of the analytioally pure iodide was placed in a vacuum desiooator over phosphorous pentoxide tor six days to insure dryness. The sample did not reach a constant -85weight durtng this time but lo.t about 2% of its weight eaoh day. After dissolution ot the sample with thorough mixing in triply distilled water (previously equilibrated in the thermostat) to the mark of a volumetric flask, the ·zero time ft aliquot was withdrawn immediately_ The ali- quots were quenched by pipettlng them lnto 20 ml. of 95% ethanol, preoooled in the titration vessel in an lee-water bath. the time was reoorded when the pipetting was complete. The potentiometric titration was oarried out imme- diately, keeping the titration vessel in an ice-water bath. The effectiYeness of the quenching technique was indicated by the oomplete absenoe of drift in the pH reading at the endpoint. The titration curves were unusual in that the pH rose rapidly at pH 6.1-6.3, leveled off at around pH 7.5 and then rose rapidly again. The point of maximum slope (pH 6.l-6.3) was taken rather arbitrarily as the endpoint. A solvent blank, treated by the quenching tech- nique, had pH 7.6. The infinity titer was determined atter 50-51 hr.; using the quenching technique, the titers were 80.1 and 80.7%, however, the titration method employed for the rest of the kinetic experiments (described above) gave infinity titers of 84.5, 84.7 and 86.1% with an endpoint at pH 7.2. -86Reaction of 80% Ethanol with Hydrogen Iodide.-A 3.1 x 10-~ solution of hydrogen iodide in 80% ethanol was titrated using the technique described for the kinetic experiments in the same solvent. ampoules and heated at 99.2-. Samples were sealed in After 13 hr., titers of 88.2 and 97.5% of the original titer were observedJ after 17 hr., the titer was 84.8%. The sample with the titer of 97.5% was nearly oolorless while the other two were brown-oolored. -87- TABLE IV SOLVOLYSIS OF 6.77 x 10-~ N-METHYL-4-(GYCLOPROPYLCARBINYLOXY)-PYRIDINIUM IODIDE0 IN 80% ETHANOL-WATER AT 98.4 Time (hr.) %Unreacted 0.0 100.0 6.5 92.4 0.0146 11.6 86.5 0.0126 23.1 73.2 0.0136 38.6 62.0 0.0124 49.7 50.8 0.0136 73.8 38.6 0.0129 148.6 19.9 0.0109 196.8 17.0 0.0090 264.6 13.7 0.0076 -88- TABLE V SOLVOLYSIS OF 7.06 x 10-~ N-METHYL-4-(CYCLOPROPYLCARBINYLOXY)-PlRIDINIUM IODIDE IN WATER AT 98.6- - Time (hr.) %Unreacted 0.00 98.0 0.62 88.6 0.198 1.00 79.0 0.216 1.62 71.9 0.204 2.00 64.2 0.211 2.60 69.6 0.199 3.02 63.4 0.201 3.60 49.4 0.196 4.00 43.8 0.201 5.06 36.1 0.198 6.07 30.4 0.193 7.00 26.3 0.188 8.02 22.6 0.184 9.00 19.0 0.182 12.00 12.3 0.173 k,. = 0.20 hr. -1 - k1 (hr. -1 ) -89- TABLE VI SOLVOLYSIS OF 6.94 x 10-·104 N-METHYL-4 ... (CYCLOPROPYLCARBINYLOXY)-PYRIDINIUM IODIDE IN WATER AT 70.7Time (hr.) % Unreacted 0.0 99.7 9.3 92.7 0.00778 23.8 85.1 0.00666 32.9 80.5 0.00663 44.7 74.5 0.00662 56.8 68.7 0.00657 n.3 63.1 0.00643 80.9 60.0 0.00627 kl = 0.0065 hr.- 1 - -90- TABLE VII SOLVOLYSIS OF 6.69 x 10-~ N-METRYL-4-(CYCLOPROPYLCARBINYLOXY) -PYRIDINIUM PERCHLORATE IN 80% ETHANOL-WATER AT 98.4Time (hr.) % Unreacted 0.0 99.0 7.7 90.7 0.0113 18.S 82.5 0.0110 24.2 78.2 0.0097 32.8 70.7 0.0102 43.0 65.7 0.0095 95.9 45.0 0.0082 142.3 37.1 0.0069 k:a. ::& 0.010 hr. -:a. - -91- TABLE VIII SOLVOLYSIS OF 7.14 x 10-~ N~Er:rrYL-4-{CYCLOPROPYLCAR­ BINYLOXY)-PYRIDINIUM PffiCHLORATE IN WATER AT 98.6° Time (hr.) % Unreaoted 0.00 95.0 0.76 84.1 0.161 1.50 71.4 0.190 2.25 62.0 0.189 3.01 53.7 0.190 3.75 45.4 0.197 8.00 22.7 0.179 12.00 14.2 0.158 - kl = 0.19 hr. -1 -92- TABLE IX SOLVOLYSIS OF 2.95 x 10-&Y N-~ETlrrL-4-{CYCLOPROPYLMETHYLCAR­ BINYLOXY) ..PYRIDINIUM IODIDE IN VIA TF..R AT 30.0· Time (hr.) % Unreacted 0.00 89.9 1.03 77.0 0.151 1.30 74.2 0.148 2.08 64.4 0.161 2.50 60.0 0.163 3.20 54.0 0.169 3.67 49.2 0.166 4.24 46.0 0.168 9.52 24.7 0.136 10.61 23.9 0.125 11.53 23.3 0.117 - k1 III 0.16 hr. -1 c' -93- APPENDIX I DESCRIPTION OF FRACTIONATING COLUMNS Column No.1. - A 1.3 x SO-am. column packed with glass helices and equipped with an electrioally heated air jaoket and a total reflux head. Column No.2. - A 0.6 x SO-em. column packed with a tantalum wire coil and carrying a partial reflux head (61). Column No.3. - A O.S x S6-am. spinning-band oolumn equipped with an electrically heated air jacket and a total reflux head. The band was a stainless-steel, wire-gauze spiral and was operated at about 1100 r.p.m. -94IV. 1. REI"FRENCES J. D. Roberts and R. H. Masur, J. Am. Cham. Soo., 73, 2509-2620 (1961). 2. E. Renk and J. D. Roberts, to be published. 3, J. D. Roberts and V. C. Chambers, J. Am. Chem. Soo., 4. -73, 5034-6040 (1961). M. C. Caserio, W. H. Graham and J. D. Roberts, Tetrahedron, 11, 171-182 (1960). - 6. A. Streitwieser, Jr., Chern. Reviews, 66, 671-752, -Cf. (1956). 6. R. H. Mazur, W. N. White, D. A. Semenow, C. C. Lee, M. S. Silver and J. D. Roberts, J. Am. Chem. Soc., 81, 4390-4398 (1969). 7. M. S. SlIver, Ph.D. Thesis, California Institute of Technology, 1959. 8. H. L. Goering and E. H. Silversmith, J. Am. Cham. Soc., zy, 6249-6263 (1955). 9. 169-181 (1958). 10. ----- - Cf. S. ~instein and G. C. Robinson, ibid., 80, ....... (a) A. Streitwieser, Jr., J. Org. Chern., 22, 861-869 (1957)J (b) - D. J. Cram and J. E. McCarty, J. Am. Chem. Soc., 79, 2866-2875 (1957). - -9511. C. K. Ingold, ·Structure and Mechanism in Organio Chemistry,· Cornell University Press, Ithaca, N. Y., 1953, pp. 381-398. 12. P. Brewster, F. Hiron, E. D. Hughes, C. K. Ingold - and P. A. D. S. Rao, Nature, 166, 179-180 (1950). 13. K. B. Wiberg, Ph.D. Thesis, Columbia University, 1950. 14. A. Streitwieser, Jr. and \". D. Sohaeffer, J. Am. Cham. 800., 79, 2888-2893 (1957). 15. - - - D. Semeno'IV, C.-H. Shih and 'W. G. Young, ibid., 80, 5472-5475 (1958). - 16. K. Alder and G. Stein, Ann., 614, 211-227 (1934). 17. G. Komppa and S. Beckmann, ibid., 512, 172-186 (1934). 18. J. D. Roberts, C. C. Lee and W. R. Saunders, Jr., J. Am. Chern. Soo., 76, 4501-4610 (1954). 19. - - S. Winstein and D. Trifan, ibid., 74, 1147-1154 (1962) • - Idem, ibid., 74, 1154-1160 (1952). 21. R. TIuisgen and C. Ruchardt, Ann., 601, 21-39 (1956). 22. C. G. Swain and L. E. Kaiser, Am. Chern. Soe., 20. J. -80, 4089-4092 (1968). 23. - C. G. Swain, L. E. Kaiser and T. E. C. Knee, ibid., 80, 4092-4094 (1958). - -9624. - E. D. Hughes and C. K. Ingold, J. Chern. Soc., 1933, 69-75. - 25. J. McKenna and J. B. Slinger, ibid., 1958, 2759-2768. 26. D. N. Kursanovand S. V. Vitt, Izv8st. Akad. Nauk - - S. S. S. R., Otde1. khim Nauk, 1959, 1445-1452. 27. T. Taguchi and M. Kojima, J. Am. Chern. Soc., 81, - 4318-4322 (1959). 28. S. H. Harvey, P. A. T. Hoye, E. D. Hughes and C. K. Ingold, J. Cham. Soc., 1960, 800-806. 29. - See J. A. Mills and W. Klyne in W. Klyne, "Progress in Stereochemistry," Vol. 1, Butterworths Scientifio Publications, London, 1954, pp. 194-195, for a summary with references. 30. E. H. ~bite, J. Am. Chern. Soc., 77, 6014-6022 (1955). 31. E. H. White and C. A. Aufdermarsh, Jr., ibid., 80, - 2597-2598 (1958). 32. - - C. J. Collins, W. A. Bonner and C. T. Lester, ibid., 81, 466-475 (1959). 33. - !.£., N. Weliky and E. S. Gould, ibid., 79, 2742-2746 (1957). 34. - - - - - W. v. E. Doering and T. C. Aschner, ibid., 71, 838840 (1949). -9735. J. Hine, "Physical Organic Chemistry," McGraw-ITill Book Co., Inc., New York, N. Y., 1956, p. 140. 36. R. S. Airs, M. P. Balfe and J. Kenyon, J. Chem. Soc., 1942, 18-26. 37. J. Kenyon, S. M. Partridge and H. Philips, ibid., 1937, 207-218. 38. L. C. Bateman and E. D. Hugest ibid., 1940, 945-948. - 39. S. Winstein, E. Grunwald and H. W. Jones, J. Am. Chem. Soc., 73, 2700-2707 (1961). 40. H. C. Brown and M. Borkowski, !i, 1894-1902 ~., (1952) • 41. E. F. Cox, Ph.D. Thesis, California Institute of Technology, 1955. 42. Cf. S. Winstein and E. M. Kosower, J. Am. Chem. Soc., ~, 43. 4399-4408 (1959). - A. H. Fainberg and S. Winatein, ibid., 78, 2770-2777 - (1956). 44. H. Phillips, J. Chem. Soc., 1923, 44-59. 45. S. ~'in8tein, M. Brown, K. C. Schreiber and A. H. - Schlesinger, J. Am. Cham. Soc., 1!, 1140-1147 (1962). 46. N. J. Demjanov and S. Pinegin, J. RUBB. Phys. Chern. Soc., ~, 58-61 (1914)1 Cham. Zentr. !£!!, 1999. -9847. R. V. Volkenburgh, K. W. Greenlee, J. M. Darfer and C. E. Boord, J. Am. Chem. Soc., 71, 3696-3697 (1949). 48. L. F. Fieser, "Experiments in Organic Chemistry," 2nd Edition, D. C. Heath and Co., New York, N. Y., 1941, p. 369. 49. (1964) • 60. - W. Thei1aoker and H.-G. Winkler, Ber., 87, 690-691 - R. H. Piokard and J. Kenyon, J. Chern. Soc., 99, 45-72 (1911). 61. P. A. Levene, A. Rothen and ~. Kuna, J. BioI. Ch«m., 62. -K. 120, 777-797 (1937). Bowden and P. N. Green, J. Chern. Soc., 1964, 1796-1798. 53. E. Koenigs and H. Grainer, Ber., §!, 1049-1066 (1931). 64. D. G. Leis and B. C. Curran, J. Am. Cham. Soe., 67, - 79-81 (1946). 55. E. Shaw, ~., 1!, 67-70 (1949). 56. D. J. Cram, J. L. Mateo!, F. Hauok, A. Langemann, K. R. Kopecky, W. D. Nielsen and J. Allinger, ibid., 57. -81, 5774-5784 (1959). - R. F. Toomey and E. R. Riegel, J. Org. Chern., 17, 1492-1493 (1962). - -99c' 58. A. E. Tschitschibabin and R. D. OS88trowa, Ber., 59. -cr. P. M. Hair and 58, 1708-1712 (1925). ~, 60. J. D. Roberts, J. Am. Cham. Soo., 4565-4566 (1957). See E. F. Kiefer, Ph.D. Thesis, California Institute of Technology, 1961, for the purification prooedure. 61. C. V'. Gould, Jr., G. Holzman and C. Niemann, Anal. Chero., 20, 361-363 (1948). - -100- PROPOSITIONS 1. The py~olysie of an este~ and the p~olysis of the co~~esponding amine oxide usually give very similar mixtures of isomeric olefine (1). In contrast, the pyrolysis of l~ethylcyolohexyl acetate yields an olefin mixture containing about 76% of the endooyolio isomer (2) While py~olysis of the cor~esponding dtmethllamine oxide gives almost exolusively the exocyelie isamer (3). Thi, apparent anomaly has been explained (2) in terms of the transition states for the two elimination reactions, that of the aoetate being a six-m~be~ed ring and that ot the amine oxide a five-membered ~ing. The ca~bonyl oxygen of l-methyloyclohexyl acetate can reach one of the gauohe hydrogens in the cyolohexane ring While the ring is in the ohair form and eliminate to form the more stable .......... endoolefin. In oontrast, the oxygen of the co~responding amine oxide cannot reach a gauche hydrogen without oonsiderable strain and consequently eliminates an eclipsed hydrogen in the methyl group to form the exo-olefin. It would be of interest to determine the isomer distribution of the olefins produoed by the pyrolysis of the acetates of 8XO- and ........... endo-2-methylnorborneol and the -101corresponding dimethy1amine oxides. Elimination into the oyolohexane ring would oocur by removal of a proton oonstrained to an eclipsed configuration by the bicyclic structure and consequently any difference in the isamer distributions of the olefina from the acetates and fram the corresponding amine oxides would have to be ascribed to factors other than those believed to be responsible for the behavior of the l-metbyloyolohexyl system. Also, the known instability of the norbornene structure (4) might have an interesting effect on the ratio of exocyclic to endocyclic olefina. 2. - Cis and trans isomers of vinyl lithium compounds (5) and cis and trans isomers of vinyl Grignard reagents - (6) have been prepared. The vinyl lithium oompounds pre- pared from ~- and trans-1-bromo-2-(~-ch1orophenyl)-1,2diphenylethylene are slowly interconverted at elevated temperatures (7). It is proposed to measure the rates of isomerization of ~ and trans vinyl lithiums and of the corresponding Grignard reagents to elucidate the mechanism of the isomerization reaction. 3. Whl tmore (8) has thoroughly investigated the reaction of methylamine with aqueous nitrous acid and although as little as 25% of the starting amine was -102recoverable in some cases, no products (methanol, dimethyl ether, methyl chloride or methyl nitrite) could be detected. This puzzling observation should be reinvestig&ted vdth the aid of recently developed analytical tools such as vapor-phase chromatography and nuclear magnetic resonance (NMR). It is suggested that diazomethane may h&ve been evolved and have esoaped deteotion by Whitmore. If so, this reaction would be an extremely convenient preparation of diazomethane. 4. Divalent carbon intermediates are probably involved in the Arndt-Eistert synthesis (9). RCOCl'" CHeN. - RCOCHN. - RCOCH: ... N. - RCR=C=O In view of the recent successes in generating carbenes by a-elimination of hydrogen halide from alkyl halides (10), it is suggested that a-haloketones may be capable of eliminating the elements of hydrogen halide and rearranging to ketenes. The 10s8 ot halide ion may not occur 0 0 0 " CeI!e-C-CT-J.Cl ... NaH - --- II Q CeHe-C-CHCl - II CeH.-C-CIf: ... Cl CeH.-CH=C=O spontaneously but might take place in the presence of Q -103silver ion. Acylcarbenes may add to olefins, providing a useful synthesis of acyloyclopropanes. 5. The NMR spectra of compounds of structure R~R.R.COCH.CH0 exhibit ABX.-type spectra (11) for the ethyl protons (12). \~ile the non-equivalence of the methylene hydrogens is clearly due to the presence of the asymmetric oenter in the molecule, the detailed manner in whioh this center exerts its influence is debatable. It (a) undoubt- · edly makes the two methylene hydrogens unequivalent in anT one of the possible rotational conformations about the carbon-oxygen bonds and (b) may make the populations of the various oonformations unequal. The relative contribu- tions of effects (a) and (b) to the observed chemical shift between the methylene hydrogens are diffiou1t to asseS8. Examination of the NW~ spectra as a function of temperature of a series of compounds R~R.R.COCH.CH. where R~ = H, R. = CR. or Cl, and R. is variable should allow an evaluation of the relative importance of effects (a) and (b) and also provide information about the enthalpies of the rotational conformations about the carbon-oxygen bonds. 6. Swain (13) has suggested that ~participation of acetic acid as a nucleophilic reagent in the rate-determining step may be the general rule for solvolyses in acetio -104acid." In particular, t-butyldimethylaulfonium perchlorate is believed to behave in this way (14). It is proposed to determine the rho (15) values for the solvolyses of a series of f-substituted phenyldimethylcarbinyldimethylsulfonium perchlorates in different solvents. The rho value for acetolysis should be quite anomalous if Swain's suggestion is oorrect. 7. Dipole moment, infrared, ultraviolet, and optical rotatory dispersion studies have been used to differentiate between an axial halogen and an equatorial halogen in a-halooyclohexanones (16). We propose that NMR may prove effective toward the same end. The NMR spectra of cis- feasibility of the method. In particular the splitting and trans-2-chloro-4-t-butylcyclohexanone should show the pattern of the hydrogen alpha to the halogen atom should be characteristic of an axial or an equatorial conformation. 8. Recent work by Deasy (17) indicates that the Grignard reagent has the formula RaMg.MgX. and not RMgX. The ~&R spectrum of the Grignard reagent prepared from - I-phenyl ethyl bromide may show the presence of a meso and a d,l form which would be compatible with the ReMg.MgX. -- structure but not with the RMgX structure. -1059. To account for the direction of addition of reagents guoh as iodine nonochloride, bromine chloride, hypochlorous aoid, and hypobromous aoid to allylio double bonds, de la Mare has considered it necessary to postulate a rapid equilibrium between more than one carbonium-ion intermediate (18). + products X :.: Cl, Dr, or I Y :.: H, OH, Cl, or Br + produots The same results can be accommodated by one bridged intermediate whioh is distorted one way or the other depending on the nature of X and Y. It is proposed to study the products from hydrolysis of 2-chloro-2-methyl3-butyl brosylate and the isomeric 3-chloro-2-methyl-2butyl brosylate and the produots from addition of hypochlorous acid to 2-methyl-2-butene. All three reactions should prooeed through the same carbonium-ion -106intermediate(s). This would be a very favorable ease in which to detect the presence of more than one carbonium-ion intermediate if there is any validity to the idea. If there is more than one intermediate and equilibration between the intermediates is slow compared to the rate of product formation, different product mixtures should be obtained from the isomeric brosylates. 10. It is proposed to study the stereochemistry of double 1,2-carbonium ion shifts. For instance, the deamina- tion of optically active I could reasonably be expected to give II after two 1,2-hydride shifts and it would be interesting to determine whether or not II has the same C.R. C.R. I HOCH.-CH-CHIINH. + RNO. - II I configuration as T. I CEO-Cn-CHe Also, a study of the migratory apti- tudes of R1 and R. in the deamination of III and IV and the pinaeolic rearrangement of V should illuminate the nature of the carbonium-ion intermediates involved in these reactions. OH I R1 R.CCHCH. NH. III IV -107REFERENCES FOR PROPOSITIONS 1. £!. C. H. DePuy and R. W. King, Chern. Revs., 60, 431-457 (1960). 2. P. H. Froemsdorf, C. H. Collins, G. S. Hammond and C. R. DePuy, J. Am. Chern. Soc., Q!, 643-647 (1959). 3. A. C. Cope, C. L. Bumgardner and E. E. Schweizer, 4. -ibid., -79, 4729-4733 (1957). P. von R. Soh1eyer, ibid., 80, 1700-1704 (1958). - - 5. !.~., D. Y. Curtin and J. W. Crump, ibid., ~, 1922-1926 (1958). 6. R. Normant and P. Maitte, Bull. soc. ohim. France, 7. 1439-1441. -1956, D. Y. Curtin, H. W. Johnson, Jr. and E. G. Steiner, c' J. Am. Cham. Soo., 77, 4566-4570 (1955). 8. - (1941). 9. - F. C. V.~itmore and R. S. Thorpe, ibid., 63, 1118-1123 - cr. W. E. Bachmann and W. S. struve, "Organic Reao-tiona," Vol. I, John Wiley and Sons, Inc., New York, N. Y., 1942, p. 41. 10. !.~., G. L. Closs and L. E. Closs, J. Am. Chem. Soc., -82, 5723-5728 (1960). -10811. J. A. Pople, W. G. ~ohneider and H. J. Bernstein, "High-Resolution Nuclear Magnetic Resonance," MoGrawHill Book Co., Inc., New York, N. Y., p. 98. 12. P. R. Shafer, D. R. Davis, M. Vogel, K. Nagarajan - and J. D. Roberts, Proc. Nat. Aoad. Soi. U. S., 47, 49-51 (1961). 13. C. G. Swain and A. MaoLaohlan, J. Am. Chem. Soc., ~, 6095-6101 (1960). 14. This Thesis, p. 39, and reference 23. 15. L. P. Hammett, "Physical Organic Chemistry," McGrawHill Book Co., Ino., New York, N. Y., 1940, p. 186. 16. !.~., N. L. Allinger, J. Allinger, L. A. Freiberg, R. F. Czaja and N. LeBel, J. Am. Chern. Soo., ~, 5876-5882 (1960), C. Djerassi, R. Reoords and B. Bach, Chemistry and Industry, 258-259 (1961). 17. R. E. De88Y, G. S. Handler, J. H. \l,' otiz and C. A. Hollingsworth, J. Am. Cham. Soo., ~, 3476-3477 - - (1957), R. E. Dessy and G. S. Handler, ibid., 80, 5824-5826 (1958). 18. P. B. D. de la Mare and S. Galandauer, J. Chern. Soo., -1958, 36-43.