D.J. Triggle and B. Belleau- Studies on the Chemical Basis for Cholinomimetic and Cholinolytic...

8/3/2019 D.J. Triggle and B. Belleau- Studies on the Chemical Basis for Cholinomimetic and Cholinolytic Activity Part 1

1/15

STUDIES ON THE CHEMICAL BASIS FOR CHOLINOMIMETICAND CHOLINOLYTIC ACTIVITY

PA R T I . T H E SY N T H E SI S A N D C O N FI G U R A T IO N O F Q U A T E R N AR Y SA L T SIN THE 1.3-DIOXOLANE AND OXAZOLINE SERIES

D. J. TRIGGLE'ND B. BELLEAUDepartment of Chemistry, University of Ollawa, Otlawa, Canada

Received February 19, 1962ABSTRACT

Structure activity relationships in t he muscarine and the qu aternary l,3-dioxolane (Fourneauseries) series are briefly discussed. Th e most active member of th e lat ter series (2-niethyl-4-dimethylaminomethyl-1,3-dioxolane ethiodide, (I X) , F2268) was shown to consist of amixture of 60% cis and 40% trans isomers. The same was found to appl y to all syntheti cintermediat es in that series. Unequivocal assignment s of con fig ~~r ati onere made by relatingvarious intermediates leading to ( I S) and i ts analogs to D-cis-l,3-dimethyl-l,3-dioxolanetself,obtained b y degradation of 1,6-anhydrogalactose. Atte mpte d separation of cis-trans isomersin the l,3-dioxo lane series was not successful. However, a mixtur e of cis ,trans-2-t richloro-methyl-4-hydroxyn1ethyl-1,3-dioxolaneXVI ) could be fractionated by crystallization of t hecorresponding tosylates. Catalytic hydrogenolysis converted the pure cis- and lrans-trichloro-methyl derivatives ( SV I I) and ( XVI II) to pure cis- and trans-2-n~ethyl-4-hydroxymethyl-1,3-dioxolane tosylates ( SI X ) and ( S S ) , which eventually afforded for the first time pure cis-F2268 and trans-F2268 ( XXI I) and (XX III ).Optically active members in the l,3 -dioxolane series were prepared. NIembers of t heD(-)-series were conveniently obtained from D-isopropylidene glycerol. Members of theL(+)- seri es could be obtained in opt ically impure forms by resolution of dl-tert iary basessuch as (X XXV I) with D- and L-dibenzoyltartaric acid. T he best preparations had a n opticalpuri ty not exceeding 32%. The resolution of t he cis base (X ) was unsuccessful.Th e synthesis of a n oxazoline analog, ( XLI V), of F2268 was accomplished. The reactionsequence involves solvolysis of N-acetyl-2,3-dibromo-fl-propylamine (XLI) to give the5-bromomethyl-2-methyloxazoline (XLII). This unstable intermediate was reacted withdimethylamine to give the tert iary base (X LI II ), which was quaternized with methyl iodidewhereupon the quaternary base (XLIV) was formed in good yield. The structure of thelatt er was established by an independent synthes is of th e hydrolysis product (XLV) .Preliminary pharmacological data are reported for the various new quaternary salts.The compounds were assayed for cholinomimetic activity. It is concluded from these studiesthat quaternary l,3-dioxolanes display structure-activity relationships analogous to themuscarones. The use of triethylammonium analogs has revealed a large degree of preferenceof cholinergic receptors for the presence of a cis configuration in 2,4-disubstituted-l,3-dioso-lanes. I t was also noted tha t t he oxazoline derivative (XL IV) ranks amongst the most activecholinomimetics thus far reported. Relationships between configuration and activity arebriefly discussed.INTRODUCTION

The recent structure elucidation and synthesis of muscarine ( l (a ) , l(b )) has createdrenewed interest in the relationship between the structure and act ivity of drugs actingon the muscarinic cholinergic receptors. A variety of muscarine analogs have beensynthesized and their cholinomi~~~eticctivity determined. This subject has been recentlyreviewed (2), and since then, Waser (3) has examined a number of addit ional structuralanalogs. Mention should also be made of the report of Friess, Witkop, and co-workers(4) on the inhibitory properties of a nuinber of muscarine-related conlpounds towardsacetylcholiilesterase. It einerges from all these iimportant studies that the optiinalstructural requirements for nluscarinic activity are as follows: (1) The muscarinicreceptor displays an aliuost absolute optical specificity, the non-natural enantiomorph (11)

IHolder o a National Research Council of Canada Postdoctoral fellowship, 1959-1961.Canadian Journal of Chemistry.Volume 40 (1962)

1201


2/15


3/15

(IX) (Chart 2 ) , which was reported to match muscari~ien potenc~.. i l l of tlie structura Ianalogs of (I X) appear to be less active. The st ructu ral relationship of ([X)with inuscarinc

CHART 2

id its ;111;~logss obvious a ~ i dt was of i~ lt eres t o examine the relation betweell stereo-cheniistry zuncl acti vity in th e Fourne au series. Th e effect of struc tural vari ation s aboutthe 2-position ancl the cationic head of (I X) has been studied by van Ro ss u~ n nd Ariens(17), who confrniecl t lie operat ion of the concept of affir~it);and irltrinsic activity in thisseries. However, the stereochemistry of (I X) in relation t o activ ity ha s not been investi-gatecl. Because activity in the ~nu sca rin e eries is highIy depend ent on configurat ion,relative and absolute, similar dependency would be expected in the Fourneau compound(IX). I t is a rathe r curious phenomenon th at the configuration of (I X) has never beencstablisliecl, arid of th e two possible stereoisomers, only one appears to have ever beenobtai~ ied r prepared. I t is not 1;nowll as yet whether the high act ivi ty of ( IX ) is associatedwith a trans or cis configuration about the 2- and 4-positions, a situation which, however,has been clarified in the muscarine series. Th e optical fornls of any qua ternary-1 ,3-tlioxolanes have also never been repor ted or test ed. Th e possible inhib ito ry effect of thesedrugs on acetylcholinesterase has also not been investigated.

In thi s first paper of a series, the synthesis and th e configuration of (I X) and a n ~u nb erof analogs will be described. Subsequ eilt publ ications will deal with the ir effect 011acetyl-cholinesterase and with generalizations regarding the probable n atu re of the muscariniccholinergic receptors.


4/15

Relative C'on.ig~~rationf ( I X )'The method most generally employecl for the s?-nthesis oi 2-methyl-4-trimetli5.1-

:~mmoniummetli~~l-I:3-dioxolane iodide ( IX ) (F2268) involves tlie reaction of epichloro-hg~drinwith acetalclehyde in the presence of stannic chloride whereupon the intermediate2-methyl-4-cl1loro111et1~yl-1:i-dioxolane (XV) is produced and thence reactecl withdimetliylamine to give ( X) , which is finally quaternized using methyl iodide (16). Th equaternary salt (IX ) so obtained has been repeatedly reported to melt sharply a t 140-141"and we have confrmetl this. van Rossum (17) has also showed it to be app arentl y homo-geneous as judged from results of paper c l~ ro ma to gr ap li ~~ . otal of :i8 liomologousst]-uctures n this series were prepared b y thi s author and all were reported to melt sharplyand to give rise to single spot s when cliromatograp l~edon paper.

Since tlie relative configuration of (I X) must be the same as that of the st art ingdioxolane (XV ), we sought to establisll the confg uratio n of tlie latt er by relating it to a2,4-cli1iietIiyl-l,3-dioxolai1ef proven confgurat ion . The reaction of propylene glycolwith acetaldehyde has been reported to give a mixture of cis- and tra~zs-2,4-dimet11yl- t -clioxolane, (X II I) ancl (X IV ), which has been claimed t o be separab le by fractional dis-tillation (18). In our hai~ds his separation provecl t o be f ar from complete (see Experi-mental ); the fraction of lower boiling poilit has been assigiied tlie cis configuration. Becauseof our neecl for a much more rigorous proof of configuration we sought to synthesize thecis isomer 11). unaml~iguous methocls. Cleavage oi I ,6-aiil1~~clrogalactoseo u-cis-1,::-dioxolane-%,4-clicarboxaldehyde as been 1-eported in the li terature (I!)). Iiecluction ofthe latter with sodium borohydricle gave the cis diol (X II ) (Chart 2 ), which gave acr):stalline cli-p-nitrobenzoate and a crystalline ditosplate. Attempts to prepare a inono-tosylate were fruitless, a bicyclic ether being presumably the only product. Hyclrogenolysisof the ditosylate witli lithium aluminum hydride afforded ~-c is-2 ,i-d ime thy l-,3-dioxolaiie(X II I) , which gave rise to a single sha rp peak in the vapor phase c1~romat ogr:~phyv.p.c.)instrument.

Th e dioxola i~e VI) resulting from the conclensation of glycerol I-monobenzyl ethe r andacetaldehyde was liydrogenolyzecl and the resulting 2-metliyl-S-h~~droxymet11yl-1,3dioxolane (VI I ) convertetl to n crystalline tosylate, (VI I I) , which ev en tu nl l~ ~ffordeclF2268 (I X) after reaction with dirne tl~ylamine ollowecl by qua terniz atio i~witli methylioclide. Th e identi ty of thi s material was ascertainecl by direct comparison (infrared and1i.m.r.) witli a sample obtained b y t he general litera ture method outlined earlier. Hyd ro-genolysis of t he intermediate tosyla te ( VI II) with lit l~i um luminum hydride or catalyticI~ydro~enolysisi the 4-clilorometliyl intermediate (X V) afforded 2,4-dimetliyl-1,3-dioxolane, wliicli was ~ 1 1 0 ~ ~ ~ 1 1y v.p.c. to consist of a mixture of cis-trans isomers in arat io of GO :40 (authe ntic ~-ci,s-3i,4-dimethyl-l,3-dioxolaneXI 11) being used as a stanclard).In order to establish th at crystallization of tlie quatern ary base (I X) did not change theisomer ratio, it was subiiiittecl to N-dealkplation by treatment with litliium aluminumliydride (20) and the tertiary base (X ) analyzed by v.p.c. The cis-trans isomer ratioproved to be B1:39, th us conclusively establisliing th at F2268 consists of a mixtul-e ofcis and trans isomers in a rat io closely approximating 60:40. Thi s conclusion was confirmedby n.1n.r. spectroscopy, which sliowecl clearly the presence of two split methyl groups ina ratio of G0:SO (see Experimen tal). i\/Ioreover, identical v.p.c. pat terns were obtaiileclwith the product resulting froin tlie reaction of propylene glycol and acetaldehyde.Therefore th e synthesis of ( IX ) proceeds non-stereospecificaIly and all t he in termediatesconsist of a mixture of 60% of cis and 40% of trans isomers. We have subsequently foundthis t o apply to virtually all solid derivatives in this series. I t is clear tha t cis- and t ~ a n s2,4-disubstituted-l,3-dioxola11esrystallize as molecular colilpounds and display very


5/15

TRIGGLE A N D BELLEAU: QUATERNARY SALTS 1205

great similarities in their physical properties. Since the time F2268 (IX) was first synthe-sized al l the pharmacological properties refiorted for it a re , therefore, those of a 6O:dO mixtureof the cis and trans isomers. Th e same would appear to apply to all other homologs, asjudged from additioilal observations recorded below. I n view of the s truc tural analogy of(I X) with inuscarine, none of the previous pharmacological da ta can be relied upon forpurposes of correlatioil between stereochemistry and activity .

Nun ~er ous ttempt s to achieve separation of the isomers of (IX) or its precursors on apractical scale by a conlbiilation of various techniques were fruitless. However, both thecis and the trans isomers of (I X) could eventually be obtained in pure forin by an indirectsynthetic route which is described below (Chart 3).

CHART 3

1 OH38% t r a n s - ( a 1

CCI, CH

0,CH,OTs ,CHN Et,

I t was found that the reaction product of glycerol and chloral (XVI) (Chart 3) gave acrystalline tosylate melting over a wide range, thus indicating t ha t molecular compoundformation was apparen tly suppressed in this type of 1,3-dioxolane. Th is proved to bethe case, since fractional crystallizatioil of the tosylate led to the isolatioil of two sharp-melting isomers, isomer A (XV II), 111.p. 133-134', an d isomer B (XVIII), m.p. 95-96'.Th e trichloromethyl group of each isomer was coilverted to methyl by catalytic hydro-genolysis. The resulting tosylate (XIX) from A had 111.p. 66-68' and when submitted to


6/15

1206 CANADIAN JOURNAL OF CHEMISTRY. VOL 40, 1963aqueous alkaline hydrolysis gave pure trans-2-1nethyl-4-l~ydroxymethyl-l,3-dioxola(VII ), as determined by v.p.c. Similar ly, the tosylate (XX ) fro111 B had n1.p. 64-66" andgave, after alkaline hydrolysis, pure cis-2-1neth~~l-4-l~ydroxymethyl-1,3-dioxolaneVII),as established by v.p.c. Reaction of the pure t ran s tosylate ( XI X) with dimethylarninefollowed by quaternization with methyl iodide afforded trans-2-methyl-4-trin1et11yl-anzmoniummetl~yl-l,3-dioxolaneXX I) (trans-F2268), m.p. 131-132'. Th e homogeneityof the product was confirmed by n.m.r. spectroscopy (single methyl doublet). Similartrea tmen t of t he pure cis tosylate ( XX) gave cis-2-methyl-4-trimethylainn101~iui~~1net1,3-dioxolane (X XI I) (cis-F2268), m.p. 143-144". Th e n.m.r. spectrunl of this compoundalso proved it to be homogeneous. This is the first time that the powerji~l holinomimeticF2268 becomes available in its @ire stereoisonzeric jornzs.

In view of the fact th at t he experimental basis for the concept of affinity and intrinsicactivity a s elaborated by Ariens and van Rossum lies principally on the inversion ofagonistic activity when a tri~nethyla~n~noniurnationic head is changed for a triethyl-a~nrnoniurn ne, it was of interest t o prepare the pure cis- and trans-triethylamnloniumanalogs (X XIV ) and ( XX II I) of cis- and trans-F2268. I t should be emphasized th atAriens and van Rossunl used what we now recognize to be a GO:40 ~nix tu re f cis-transisomers, thus preventing any definite conclusion regarding the effect of stereocheinistryon the cholinolytic activity of tr iet hyl anl n~o niu ~nnalogs.Synthesis in the Optically Active Series

Because of the dramatic difference in chol i~ ~o rn i~ ne ti cctivity between optical isonlersin the ~nuscarine eries, it was of interest to malte available a nuinber of optical isomersin the 1,3-dioxolane series. Quaternary sa lts of the D-series could be prepared in opticallypure forms from optically pure D-isopropylidene glycerol (XXV) (Chart 4). The lattercould be converted by way of the D-tosylate (SXVI) to ~(-)-'2,2-dimetl1y1-4-tri1~~etl~am1noniu1~~meth~~1-1,3-dioxolaneodide (X XVI I) . Th e preparation of t he 2,2-bisnor

CHART 4


7/15

TRIGGLE AN D BELLEAU: QUATERNARY SALTS 1207analog (XXX ) in optically pure forin was best accomplished by mild acid hydrolysis ofD-isopropylidene glycerol tosylate (XXVI) followed by condensation with formaldehydewhereupon the tosylate ( XX IX ) was obtained. Conversion of the la tte r to D(-)-4-trimetl~ylan~moniun~~nethyl-1,3-dioxola1eodide (XXX) was accomplished in theco~lventionalmanner described above. In order to gain access to the L-series, the resolutionof the corresponding racemic tert iary bases was investigated (Chart 5). Ra ce~nic

CHART 5

4-dimet l1yla rn inomet l~yl-1~3-dio~eXXXI) gave crystalline D- and L-dibenzoyl-tartrates (XX XII ) and (X XX II I), which were recrystallized to constant rotation. Thebases were regenerated and converted to the methiodides (X XXIV) and (XXXV). Thelevorotatory salt is therefore of the D-series. The rotat ions showed them to be only 7y0optically pure. This difficulty in achieving resolution must again reflect the tendencyfor diastereoisomers of the 2,4-disubstituted-1,3-dioxo1aneeries to crystallize as molec-ular compounds. Because L-dibenzoyltartaric acid gives rise t o (+)-dioxolanes of theL-series, the 2,2-dimethyl tertiary base (XX XVI) (Chart 5) was converted to a crystallineL-dibenzoyltart rate which was recrystallized to constall t rotation. Regeneration of t hebase followed by quaternization with methyl iodide gave ~(+)-2,2-dimetl1yl-4-trimeth~lammonium~nethyl-1,3-dioxolaneodide (X XS VI I) , which was 33% optically pure.

Application of these procedures to racemic cis- and trans- (X) was only partial lysuccessful. With cis-(X) it was not possible to obtain crystalline dibenzoyltartrate salts.However, trans-(XXXVII I) (Chart 6) gave a crystalline D-dibenzoyltartrate which wasrecrystallized to constant rotat ion. Regeneration of the base and quaternization in theusual way gave D(-)-trans-F2268 (XXX IX). The optical purity of the lat ter was estab-lished as follo\vs: I t was hydrolyzed with dilute hydrochloric acid and the rotation of theresulting quaternary diol (XX VIII ) determined. Comparison of this rotation with tha tof an optically pure sample (Chart 4) derived from optically pure D(-)-(XXV II) byacid hydrolysis under identical conditions showed the D(- )-t rans-(XXXIX) to be 32%

dl- (m) dl- ( x m )1

Resolution

D(-,~XXXII)Resolution;CH,I

L (+,~xEm) v


8/15

CANADIAN JOURNAL OF CHEMISTRY. VOL 40, 1962CHART 6

optically pure. Using again trans-(XXXVIII) but L-dibenzoyl tartaric acid as theresolviilg agent, L(+)-trans-F2268 (XL) was ultimately obtained with an optical purityof 32% (Chart 6) .

With these isomers of variable optical purity it should be possible to establish whetheractivity will depend on absolute configuration, as is t he case in the muscarine series. Ifthe analogy between muscarine and the 2,4-disubstituted-l,3-dioxolanes real, it wouldbe expected that the isomers of the L(+)-series should silllilarly be lnore active cholino-mimetics. I t will also be of interest to determine the anti-acetylcholinesterase inhibitoryproperty of the various isomeric forms of the above quate rnary dioxolanes. This aspectof our work will be reported separately.Synthesis of 5-Trimethylammoniz~mmetlzyl-2-methyl-oxazoleromide (XLIV)

Perhaps one of the most striking aspects of s tructure-activity relationships in themuscarine series is the observation that racemic 4,s-dehydromuscarone (111) surpassesL(+)-muscarine in potency (Chart 1). Because D( -)-muscarine displays a very low orderof activity , it is surprising that the absence of an asymmetric center a t position 5 ofil,5-dehydromuscarol1e should lead t o such a highly active cholinomirnetic. Moreover,and in sharp contrast to the muscarine series, the relative configuration of the methylgroup in the muscarone series is relatively unimportant, racemic allomuscarone (trans-(IV)) being only slightly less active than racemic inuscarone (cis-(IV)). This result alsosuggests th at destruction of the asymmetric center a t position 2 through the introductionof a double bond between carbons 2 and 3 (if th is were feasible) might no t alter activityin the muscarone series. While this problem is being investigated, it appeared of moreimmediate interest to t ranspose this reasoning to the 1,3-dioxolane series and i t is obviousthat in order to produce a trigonal carboll a t position 2 of 1,3-dioxolanes, it is essentialth at the oxygen atoll1 a t position 1 be substituted by a trivalent heteroatom such as


9/15

TRIGGLE AND BELLEAU: QUATERNARY SALTS 1309nitrogen. I t should be noted t ha t no oxazoline analog in the Fourileau series of cholino-mimetics has as yet been reported. Several of the most obvious methods for the synthesisof the desired analog (XLIV) (C hart 7) of (IX) proved unsuccessful. A convenient and

CHART 7

0CH2-C H CH,NMe, CH CHCH,NMe3I 1 w 1Ac NH OH AcN H OH

CH 2 BrCH,CH CH,I I IA cN H Br Br INq

expedient method of synthes is was ultimately discovered which requires the readilyavailable starting material N-ally1 acetamide. This was converted to the dibroinide(XLI) (Chart 7) as described elsewhere and then solvolyzed in acetoilitrile in the presenceof silver carbonate. The participatioil of neighboring arnido groups in such solvolyses haspreviously been deinoilstrated (21) and allows the prediction that the dibromide (XLI)illust be coilverted largely to 5-broinomethyl-2-inethyl-oxazoline (XLII). Attempts toisolate this intermediate were unsuccessful, rapid polymerization of the coinpoui~d akiilgplace upon removal of the solvent. However, if excess anhydrous dimethylamine wasadded to the reaction mixture as soon as maximum precipitation of silver broinide hadoccurred, a good yield of 2-methyl-5-dii~~ethylai~1ii~0i~1etl~yloxazolieXLIII) wasobtained. Quaternization of the latter with methyl iodide finally gave the crystalline sa lt(XLIV) in high yield. I t was desirable to establish the structure of the la tter by un-an~biguousmeans, since the course of the solvolysis can only be inferred. Th e quaternarybase (XLIV) was hydrolyzed with hot water to give a quant ita tive yield of l -acetamido-3-trimethylan~monium-2-propanol odide (XLV). An authent ic specimen of the lat terwas secured by reacting 1-phthalimido-2,3-epoxypropane XLVI) with dimethylamine,whereupon ( XLVII) was produced. Hydrolysis of t he lat ter followed by acetylation andquaternization with methyl iodide afforded (XLV), which proved to be identical withthe sample obtained by hydrolysis of the oxazoline (XLIV). I t is clear tha t thi s inethodof synthesis of (XLIV) could be extended to the preparation of a variety of trialkyl-alnmoniuin analogs.

dcH2N- 0CH3(EL) ~rn) (rn)


10/15

1210 C A N A D I A N JOURNAL OF CHEMISTRY. VOL. 40. 1962P R E L I M I N A R Y P H A R M A C O L O G IC A L R E S U L T S*

The various quaternary dioxolanes and the oxazoline derivative (XLIV) were assayedfor nzuscarinic activity using the guinea pig ileuin as the test organ. The minimum con-centration of drug necessary to elicit a response is recorded in Table I . Of immediate

T A B L E I

Acetyl choline{IBi,( X X X )D ( - ) - ( X X V I I ) *L ( + ) - ( X X X V )L ( + ) - ( X X X V I 1 ) t

Co n c e n t r a t i o n ,mg/ml , guineaCo ~ n p o u n d p i g i le u mnumber (Cx lo-7)

*Optical purity of this i somer is 7% .*Optical purity is 32%.

Concent ra t ion ,mg / ml , g u i n e aCompoulld pig i leumnumber (cX

interest is the fact that quaternary compounds of th e L(+)-series are unifor~n lymuchmore active than their D(-) counterparts. Hence, the same absolute configurationalrequirement as in the ~nus carine eries is operative. I t is also interesting th at cis-F2268(X XI I) surpasses trans-F2268 (XX I) in potency by a factor of 5. Thi s also parallels theeffect of epimerization abou t th e Cz-methyl group of muscarine, racemic alloinuscarinebeing very much less active. However, t he magnitude of the differences in potencybetween cis-(XXII) and trans-(XXI) on the one hand and (&)-rnuscarine and (&)-allo-muscarine on the other is such as to indicate that the asyillilletric center a t C2 of (XXI)and (XX II) is not nearly a s important as it is in the ~nuscarine eries. In this respect, theratio of t he potencies of cis- and trans-F2268 suggests that t hey resemble the inuscaronesizzuch inore than t he muscarines. I t should be llleiltioned t ha t Barlow had speculatedth at cis- and trans-F2268 would not differ grea tly in potency because of a loose structuralanalogy with th e inuscarones (IV) (22). I t is striliing, however, th at the L(+)-trans-F2268 (XL) , which is only 32% optically active, should surpass acetylcl~oline n potencyby a factor of 20. Consideration of th e fact that a 32% enrichment in the L(+) isomerincreases the potency by a factor of 100 over the pure raceinic mixture ( XXI) and th atthe cis raceinate (XXII) is five times more active than the latter, suggests that the pureL(+) cis isomer of (X XI ) should possess extraordina ry potency surpassiilg all knowncholinomimetic drugs. Th e synthesis of this L( +) cis isomer is under way.

Confirmation tha t the configuration of th e as y~ n~ ll et ri center a t Cz of F2268 may notbe a critical factor (although it has an influence) for activity is supplied by the highcl~olinomimetic ctivity of the quate rna ry oxazoline (XLIV) in which C2 s now trigonal.We are fully aware, however, that in this lat ter case, the presence of a basic nit rogen inthe ring introduces an additional factor which ma y alter the mechanism of interactionwith the receptors. We are presently attempting to establish whether the oxazoline(XLIV) interacts with the receptors in the protonated form or as the free base.

Of great interest finally is the marlied difference in the re lative potencies of the trie thyl -a~nilloniu~nnalogs (XXIV) and (X XI II ) of cis- and trans-(F2268). In contradictionwith the conclusions regarding the significance of the asymmetr ic center a t C2 of F2268

*Reszrlts of ozir work a nd that of Dr. 111. Pindell and his staff, Bristol Laboratories, Syraczise, IV.Y.


11/15

TRIG G LE A N D BELLE.iU: QUATERNARY S.4LTS 1211(IX), it appears that in the t r ie thy la?nmoni~~nzeries ( X X I I I ) a nd ( X X I V ) h e c on j5 g~ ~ ra ti onof the C2 metlzyl group is of critical importance. An interpretation of these results will begiven in a forthcoming publication.

cis,trutzs-2-illethyl-4-ltydroynzctltyl-I 3-dioxolane ( V I I )This was prepared accorcli~lg o the method of Brimacornbe, Foster , and Maines (23). The pure alcoholwas analyzed by \..p.c. a nd shown to consist of a mixtur e of cis-trans isomers in a ratio of 64:36. (sceTable 11, entry 1). Thc t os yl at e ( V I I I ) was prepared by treatment \\.ith p-tol~~cnesulphonyIhloride an dTABLE I1

Isomer composition of 2,4-dis~ibstituted-l,3-dioxola1~esColumnEnt ry Compound So . 2-Subst. 4-Subst. yocis- O j, trans- paclring*

*Obtained from Perkin-Elmer and used as such. The v.p.c. app arat us was a Perkin-Elmer instrum ent. Percentag e compositionsmere determined by taking the ratio of peak areas i n th e usr~alway.?Obtained from (VII) by way of (V III) .fobtained from (XV), entry 4.$Obtained from (XV), en try 5.Obtalned from cis.lrans-(XVI) by hydroge~~olysis.4Lower-boiling fraction.**Higher-boiling fraction.pyridine a t 0". \\,hereupon a n oily tosylate was obtained which crystallized from hexane as colorIess needles,m.p. 49-53". Yield: 'i4yO. Calc. for Ci?HIoSOQ: C, 52.9; H, .9. Found: C, 52.9; H, 5.9%.cis,tra?zs-2,4-Di~i~ethyl-1,3-diosola?taX I I I ) and ( X I V )A solution of t he tosylat c of cis,tru~1s-2-methyl-4-hydrox~~1~~eth~~l-l,3-dioxola1eVII I) (13.5 g, 0.05 mole)in diglyrne \vas added drop\\,ise to a stir red solution of lithium aluminc~rn yd ride (1.9 g, 0.05 mole) in diglylneat 100". The mixture was kept a t this temperature for 2 hours, cooled, and treat ed with 10yo sodium hydrox-ide and distilled (bath tenlpcr ature 150'). ?'he dis tillat e was dried ( Na ?S 04 ) nd redistilled to give cis,trans-2,4-dimetl1yl-1,3-dioxolane (4.4 g, 86%) , b.p. 90-95. Th e results of the vapor phase chromatographicanalysis are given in Table I1 (critry 2).Attempted Separation of cis and tran s Isonzers of 2,4-Dimethyl-1,S-dio.-colane X I I I ) and ( X I V )Two hundred grams of the cis-trans ~ni xt urewas prepared according to the meth od of Lucas and Guth rie(18). Repeated fractional distillation and vapor phase chromatographic examination of t he fractions revealedan increase in the cis cont ent of the lower-boiling fraction an d a n increase in the tra ns coilteilt of t he higher-boiling fraction, bu t a s th e results ill Table I1 (entries 10 and 11 respectively) show, the separation was fa rfrom complete.cis,trans-2-Methyl-4-ditnetl~yLat1zinomet7ty1-1S-dioxolane iWetlziodide ( I X )cis,tra~ts-2-Methyl-4-tolue1~esulphonyloxy1ethyl-l,-dioxola1eVI I I ) (13.5 g, 0.05 mole) was dissolvedin 100 1111 of a 3 0% solution of an hydrous dimethylamine in be n~ en e nd kept a t 100 for 12 hours; themixture was cooled, filtered, and the excess dimethylamine removed by heating on the steain bath. Additionof an excess of met hyl iodide gave a ne ar quanti tati ve yield of t he quat erna ry iodide; crystallizatioil fromisopropanol gave colorless needles ( 82%) , n1.p. 141" (unchanged on fur ther c rystalli zation). T he n.m.r.spectrum (in pyridine) showed two lnethyl doub lets with an intensity rati o of 3:2.

' A l l ntelting points were determined nticroscopically o n a Kofler hot stage and are uncorrected. Tlte boilingpoints are also uncorrected. Infraved spectra were determined using a Perki?z-Eltner Iflfracord instrzrment. The1t.m.r. spectra were recorded with a Va ria n instru men t operating at 60 dlc . -4 l icroanalyses by Miss E. B u s k ,Chemistry Department, Uttirersity of Ottawa.


12/15

1212 CANADIAN JOURNAL OF CHEMISTRY. VOL. 40. 1962c i s , t r a n s - 2 - i l ~ e t l t y l - 4 - d i n t e t I ~ y l a n z i n o ~ o l a n eS)The finely powdered preceding methiodide (1x1 (7.2 g, 0.025 mole) was suspended in diglyme (20 ml)with lithium aluminum hydride (0.025 mole) and the reaction mixture heated to 150' for 6 hours and thendecomposed with 10% s odi~ un ydroxide (10 ml), filtered, saturated with salt, and extracted several tilneswith ether t o give cis,trans-2-methyl-4-di1nethylaminometh~l-l,3-dioxolane2.6 g, 71.2y0). Th e results of th evapor phase chromatographic analysis are given in Tabl e I1 (entry 3).cis,tra~zs-2-~1.ietltyl-4-chloro-or bronto-) methyl-1,s-dioxolane (X V)These were prepared from glycerol a-monochlorhydrin or monobromhydrin and acetaldehyde by reactionin boiling benzene, the water being continuously removed with a Dean-Stark tr ap . The results of thevapor phase chromatography experiment s are given in Table I1 (entries 4 and 5). Th e configurations wereestablished by hydrogenolysis to the corresponding 2,4-dimethyl-1,3-dioxolanes s follows: cis,trans-2-methyl-4-halomethyl-1,3-dioxolane0.1 mole) in absolute ethanol in which was suspended sodium bicar-bona te (0.5 mole) and 10% P d/ C (0.5 g) was reduced at 3 atm pressure of hydrogen for 3 hours. Th e reactionmixture was filtered and a sample of the filtrate analyzed by vapor phase chronlatography to give theresults shown in Ta ble I1 (entries 6 an d 7). These results (Table 11) confirm th e configurations suggested byvapor phase chromatographic allalysis of the halomethyl derivatives themselves (see above).D-cis-2,4-(Dihydroxymet/t~~~)-~,~~-dioxo~aneXII)Crude 1,3-dioxolane-cis-2,4-dicarboxaldehyde19) (0.1 mole) was dissolved in absolute methanol andcooled in ice. A solutioil of sodium borohydride (6.0 g, 0.15 mole) in water was added dropwise with st irringa t 10-15". Stirring a t roorn temper atur e was continued for a further 3 hours, 25 ml of 10% sodium hydroxidewas added, and carbon dioxide bubbled it1 until the solution was saturated. Th e solution was evaporated t odryness and the residue extracted with ethyl alcohol to give D-cis-2,4-bis(hydroxymethy1)-1,3-dioxolane(X II ) as a viscous oil (12.1 g, 90.3%). Calc. for CzH1004: C, 44.8; H , 7.5. Found: C, 44.3; H , 7.1%. Wit houtfurther purification, this D-ci~-2,4-(dihydroxymethyl)-1,3-dioxolane6.7 g, 0.05 mole) was dissolved indr y pyridine and th e solution added dropwise to a stirred an d cooled solution of p-nitrobenzoyl chloride(20.5 g, 0.11 mole) in dr y pyridine ; aft er 12 hours th e reaction mixture was poured over ice, filtered, washedwith ice-cold water, dried, and recrystallized from benzene to give the cis-bis(2,4-dinitrobenzoate) s verypale yellow needles (15.0 g, 71.2%), m.p. 94". [U]D~?= -18.5' (C= 1.2, acetone). Calc. for C19H1~N2010C, 52.8; H , 3.7. F ound: C, 52.6; H, 3.3%.Th e bis(dinitrobenzoale) (10.8 g, 0.025 mole) was dissolved in aqueous ethan ol (100 ml) containing sodiumhydroxide (6 g) and heated under reflux for 5 hours. The solution was cooled, saturat ed with carbon dioxide,evaporated to dryness, and extracted with absolute ethanol to give ~-ci~-2,4-(dihydrox~rnethyl)-ldioxolane (X II) (2.5 g, 74.6%) as a viscous oil. Calc. fo r CjH1004: C, 44.8; H , 7.5. Fo~i ild : , 44.3; H, 7.2%.The pure ~-cis-(2,~-di~tydroxymethyl)-1,~-dioxo~ane2.0 g, 0.015 mole) so obtained was dissolved in dr ypyridille a t 0, and P-toluenesulphonyl chloride (6 g, 0.033 mole) was adde d. After 24 hours a t room tempera-ture a large excess of ether was added and the solut ion was extracted with 10% hydrochloric acid until freefrom pyridine. Th e ethereal solution was dried ( Na? S04 ), iltered, and eva porated to give the ditosylate ascolorless needles (4.4 g, 66%), m.p. 84-86". [U]D = -25.5 (c = 2, acetone). Calc. for CleH??S208: C, 51.6;H, 5.0. Found: C, 52.1; H , 5.1%.D-cis-I,4-Di?~iethyl-l,S-di~x~lafieXI I I )Thi s was prepared by li thium aluminum hydride reduction of the preceding ditosylate by the proceduredescribed abov e in the case of t he preparation of cis,trans-2,4-di1lleth~l-1,3-dioxolaneXIII) + (XIV). Theyield of crude ~-cis-2,4-dimethyl-1,3-dioxolaneas ollly 23% and the quantity was not sufficient for puri-fication and empirical analysis. However, vapor phase chromatographic analysis was una~nbiguous s itshowed that the compound could only be pure cis-2,4-dimethyl-1,3-dioxolaneXIII) (Table 11, entry 8).cis,trans-l-Trichlorontethyl-4-l~ydrox~~nzetl~yl-l3-dio.vola?ze X VI)This was prepared according t o the inethod of H ibbert (24). I t was found th at by increasing the reactiontemperature and time t o 90' and 24 hours respectively the yield was increased to 65-70%.The dioxolane so obtained (22.0 g, 0.1 mole) was dissolved in 80% aqueo us methail01 (100 ml) containingsodium bicarbonate (33.6 g, 0.4 mole) and 10% Pd /C (0.5 g). The mixture was reduced at 5 atm pressureof hydrogen for 5 hours. T he solution was filtered and distilled t o give cis,trans-2-methyl-4-hydroxymethyl1,3-dioxolane (VII) (7.5 g,63.5%), b.p. 70-75' a t 8 mm . The isomer ratio was 62338 (Table 11, entry 9).Calc. for CjH1003: C, 50.9; H, 8.5. Fou~ld:C, 51.1; H , 8.6%.Separation of cis,trans-2-Tricltloro~netl~yl-4-l~ydrozymethyl-l,S-dioxolaneX VI)The above trichloromethyl dioxolane (892 g, 4.0 moles) was dissolved in dry pyridine (1.5 liters) and thesolution cooled to 0" and tre ated with p-toluenesulphonyl chloride (840 g, 4.4 moles) over a period of 12hours. Th e reaction mixture was then allowed t o stand for 24 hours at room temperature a nd then pouredinto 5 liters of ice water. Th e precipitate was filtered, washed with water, and drie d in air t o give thep-toluenesulphonate (1495 g, 99%), white needles, m.p. 110-12O0, which upon repeated recrystallizationfrom methanol yielded t he t ran s tosylate (S VI I) , n1.p. 133-134" (250 g). Calc. for Cl?HlaSOjCla: C, 38.3;


13/15

TRIGGLE AND BELLEAU: QUATERNARY SALTS 1213H, 3.5. F ound : C, 38.1; H, 3.4%. Th e mother liquors from the first crystallization yielded, on repeatedcrystallization from methanol, the cis tosylate (X VII I) , m.p. 95-96" (130 g). Fou nd: C, 38.4; H, 3.4%.cis- and trans-2-Methyl-4-hydroxy~netI~y1-1,s-dioxolaescis-VII) and (trans-VII)

The preceding tran s tosylate ( XVII ) (19.0 g, 0.05 mole) was added to 200 ml of aqu eous meth anol(1:1, v/v) together with sodium bicarbonate (14.8 g, 0.2 mole) and hydrogenated a t 3-5 atm pressure with10% Pd /C catalys t (0.5 g) for 24 hours to give the tr ans tosylate (X IX ), 1n.p. 66-68"; yield 10.1 g. It wasrecrystallized from pentane. Calc. for CIGHIGSOB:, 52.9; H, 5.9. Found: C , 52.7; H, 5.9%. By the sameprocedure, the cis tosylate (XVIII) gave the cis tosylate (XX),1n.p. 64-66"; yield 9.8 g. I t was recrystallizedfrom pentane. Calc. for Cl?H16S05:C, 52.9; H, 5.9. Found : C, 52.9; H, 5.6%.Th e trans tosylate ( XIX ) (5.4 g, 0.02 mole) was heated under reflux for 4 hours with a 10% excess ofsodium hydroxide in 70% aqueo us ethanol. The solution was then saturate d with carbon dioxide, evapor atedto dryness, and the residue extracted with absolute ethyl alcohol to give chroinatographically pure (v.p.c.)trans-2-methyl-4-hydroxymethyl-1,3-dioxolanetrans- VII), b.p. 76-80' a t 9 lnin (2.1 g, 70.0%). Sin ~ila rly,the cis tosylate (X X) yielded chromatog raphically pure (v.p.c.)cis-2-methyl-4-hydroxymethyl-l,3-dioxolane(cis-VII), b.p. 78-80" a t 10 mm (2.0 g, 67.0%).Lithium aluminum hydride reduction of the tra ns tosylate (X IX) and the cis tosylate (XX) by the p rc-cedure described above in the case of (VI II) gave th e corresponding pure cis- and trans-2,4-dimethyl-1,3-dioxolanes, as evidenced by v.p.c.cis- and trans-2-A~etltyl-~-di~?tethylaminornethyl-1,S-dioxolaneethiodides (XXII) and (XXI)

cis- and trans-2-Methyl-4-p-toluenesulphonyloxymethy1-13-dioxoaes XX) and (XIX) (5.4 g, 0.02mole) were converted into the corresponding 4-dimethylaminon~ethyl erivatives by the method describedabove for the mixture of cis-trans isomers. Tre atm ent with inethyl iodide and crystallization from ethy lacetate - sopropanol gave pure cis-(XXII) and pure trans-2-methyl-4-dimethylan1inomethyl-l,3-dioxola~1emethiodides (XXI) , m.p. 143-144" and 131-132' respectively. Calc. for: C B H ~ B N ~ O P I :trans): C, 33.5;H , 6.3. Found : (cis): C, 33.8; H , 6.1; (tr ans): C, 33.6; H, 6.2%. Th e n.m.r. spectra (in pyridine) of bot hisomers showed a single methyl doublet.cis- and trans-2-Metltyl-Q-dietItylaminontetItyl-1,S-dioxolanethiodides (XXIV) and (XXIII)

These were prepared in a way similar to that used for the corresponding 4-din~ethylaminometh~lcompounds. Pure trans-2-n1ethyl-4-diethyla111i1101nethyl-l,3-dioxolanethiodide (XXIII) had m.p. 129-131'(frorn isopro pano l-ethy l acetate ) but the cis isomer (XXIV) failed to crystallize. Calc. for CII H2 rN 02 1:C, 40.2; H, 7.3. F ound: (tran s): C, 40.4; H , 7.0; (cis): C, 40.5; H, 7.1%.Otl~erQuaternary Salts of 1,3-D~osolanesThese were the previously known 4-dirnethylaminoinethyl- an d 2 , 2 - d i m e th y l - 4 - d i ~ n e th ~ l a ~ n i n om e t h ~ l1,s-dioxolane methiodides (dl-XS X) and (dl- SXV II), and the 4-dieth>.larninomethyl- and 2,2-dimethyl-4-diethyla~nino~nethyl-1,3-dioxolanethiodidcs (IL) and (L). Their physical constants agreed with theliterature values (16).Resolz~tion f 4-DintethylamtnoiizetI~yl-1,s-dioxolaneXXXI)The amine (X XXI ) (0.15 mole) was dissolved in d ry eth er, and a solution of D-diben zoyltartaric acid(0.15 mole) in dry eth er added an d the mixture allowed to st and overnigh t. Th e ether was decanted an dthe viscous precipitate re cry stal li~e d o con stant rotat ion from ethano l to give 13.0 g of colorless crystals,m.p. 131-132O, [ol]~"= -DOo (c = 2, methanol). Calc. for C?,H?jNOlo: C, 59.1; H, 5.2. Foun d: C, 59.0;H, 5.7%.It was not possible to obtain the (+)-salt from the inother liquors. The above procedure was thereforerepeated using L-dibenzoyltartaric acid, whereupon th e ( +)-s alt was obtained in the same sta te of pu rityas the (-)-salt.Th e free bases were regenerated from the (+)- and (-)-salts by treatment in dry ethanol with anequiinolar amou nt of sodium ethoxide. The precipitated di so di ~i ~nibenz oyltar trates were removed byfiltration after dilution with ethe r. The liltrate was treate d with excess methyl iodide and the p ure methiodidescollected after 24 hours. Both had 1n.p. 157-158'. Th e D(-)-qz~aternary salt (XX XIV ) had [o l] ~~ o-2.2(c = 2, MzO). 'The optical purit y was 7% (see below). Calc. for C7HloOrSI: C, 30.8; H, 5.9. Found :C, 30.9; H, 5.6%. The L(+)-qz~aternary salt (X SS V) had [ o l ] ~ ~ ~2.2' (c = 2, H1O). The optical puritywas 7%. Calc. for C~HIGO?NI:, 30.8; H, 5.9. Fo und: C, 30.9; EI, 5.8%.Optically Pz~re(-)-4-Dimethylanzino-1,s-dioxolane l4ethiodide (XXX)D-Isopropylidene glycerol (25) was converted to th e tosylate ( S S V I ) by th e general procedure outlinedabov e. It crystallized from hexane as colorless needles, m.p. 24-27'. Calc. for CI3H l8o5 S:C, 54.5; H, 6.3.Found: C, 54.7; H, 6.2%.Th e tosylate was hydrolyzed by hea ting to 50' in excess 2N hydrochloric acid for 4 hours. The solutionwas talcen to dryness in vaczlo and the residue reacted in benzene with paraformaldehyde in the presence ofsome p-toluenesulphonic acid. The water was continuously removed with a Dean-Stark tra p and th e produ ct


14/15

1214 CANADIAN JOURNAL OF CHEMISTRY. VOL. 40. 1909isolated in the usual manner. I t crystallized from benzene-pentane, m.p. 35.7". Thi s optically pure D(-)-4-t~s~loxymethyl-1,3-dioxolalleXX IX ) had [a]nZ0= -10.0' (C= 2, isopropanol). Calc. for ClIHIIOLS:C, 50.5; H, 5.5. Found: C, 50.8; H , 5.4%.This tosylate (XXIX) was converted to ~(-)-4-din1ethylarninon1eth~l-1,3-dioxolaneethiodide (XXX )by t he general procedure outlined above. T he methiodide had 1n.p. 147-148" (isopropanol); [alDZ0 -32.5'(c = 2, H?O). Calc. for C~HIGOZNI:, 30.8; H , 5.9. F ound: C, 31.3; H, 5.8%.D(-)-2,2-Dinzetlzyl-4-dintethylaminomctl~yl-1S-dioxolane Alethiodidc ( XX VII)The preceding tosylate of D-isopropylidene glycerol (S XV I) was reacted with dimethylamine an d t heresulting base quaternized with methyl iodide by the procedures described above. The quaternary iodide(XXVII) had rn.p. 215-216" (ethanol); [a]nZ0= -12.0' (c = 2, H?O). Calc. for C,H?oO?NI: C, 35.9; H, 6.7.Found: C, 36.1; H, 6.4%.Of this quate rnary sal t, 200 mg was heated a t 90' in excess 2 N hydrochloric acid for 5 hours. Th e solutionwas taken t o dryness in vacrro and the residue (XXVI II) made up t o 5 ml with water. T he solution hada D ? ~= -1.38". Thi s solution was used as a standa rd for the determination of the opti cal purity of otherdioxolanes described below.~(+)-2,2-Dimethyl-4-dimethylamino1izethyl-l3-dioxolane filethiodidc ( XXX VII )The corresponding tertiary base dl-2,2-dimethyl-4-dimethylaminomethyl-l,3-dioxolaneSXXVI) wasresolved with L-dibenzoyltartaric acid in the same manner as described above in the case of (XXX I) . Th esal t was recrystallized to constan t rotation. I t had n1.p. 133-136"; [a]nZo= 70" (c = 2, methanol). Calc.for C?Gl-13101~N:, 60.35; H, .05. Found: C, 60.5; H, 6.2%.This L-dibenzoyltartrate salt was decomposed as described above in th e case of ( X S S I ) an d th e regene-rate d base reacted with methyl iodide whereupon the methiodide (X SS VI I ) was obtained, m.p. 204-208"(isopropanol). It had [a]nZ0= 4.0' (C= 2, H?O). Since the rotation of the optically pure enantiomorph(XXVII) is - 2.0, the optical purity of this L(+)-isomer (X XS VI I) is 33%. Calc. for CoH2 002 NI: , 35.9;H, 6.7. Found: C, 36.1; H, 6.9%.~(-)-~rans-2-Methyl-~-dimethylam~it101~zethyl-l,3-dioxolanelletlziodide (X XX IX )The tert iary base ( SS XV II I) was resolved with D-dibenzoyltartaric acid by the same procedure outlinedabove in the case of (XXXI). The salt, obtained in 25% yield, was recrystallized to constant rotation,m.p. 131-132', [CY]D~O-84' (C= 2, methanol ). Calc. for C?jH?;OloN: C, 60.0; H , 5.4. Found: C, 60.3;H,-5.5%.Regeneration of th e base followed by quaternization with methyl iodide in the usual manne r afforded th eD( )-methiodide (X XX IX ), n1.p. 100-110" (isopropanol), [a]n20= -8.25 (c = 2, H?O).Calc. for C813180NIC, 33.5; H, 6.3. Fou nd: C, 33.8; H, 6.1%.A 200-mg portion of the methiodide was hydrolyzed in 2 iV hydrochloric acid as described above in thecase of (XX VII ). The resulting quate rnary diol (SX VI II ) had [~]D?O -0.44". Since the pure D(- )quaternary diol had [a]n20= -1.38" under the same conditions, the optical purity of the D(-)-trafzs-~nethiodid e X S S I S ) obtained by resolution is 32%.L (+ ) - t r a ~~ s - 2 -A l e t l z y l - ~ n z e t l t y l a ~~~ . i ? z o ~ l ~ e t l ~ y l - l3-dioxolane Jlethiodide (XL)The same procedure just described above in the case of ( S S X I S ) was applied throughout except t hatL-dibenz~~l ta r ta r iccid was used. Th e L-dibenzoy ltartrate sal t of ( S S S V I I ) had 1n.p. 131-132", [aID?O=85" (c = 2, methanol). Calc. for CZ~I-I?~O~ON:, 60.0; M, .4. Found: C, 60.1; H, 5.8%.Th e L(+) -methiod ide ( S L) had 111.p. 100-11O0, [a]n20= 8.3' (c = 2, 1320).The quaternary diol (XS VI II ) obtained by hydrolysis had [ ~ J D ~ O0.44'. Th e optical purit y of theL(+)-trans- methiodide (XL) obtained by resolution is therefore 32%.\Vith cis-2-methyl-4-dimethylaminomethyl-l,3-dioxolane (cis -S) no crystalli ne salts with D- or L-diben~o~l ta r ta r iccid could be obtained.~-i~l~t~yl-5-dimetl~ylanzi~zo~iz~thyE-o.azoneXLIII) and Its dlethiodide (XLI T)Afte r several trial runs, t he following procedure proved t o be the most c onvenient : to a solution of 30 gof N-acetyl-2,3-dibrom1o-l-propylamine (IXL ) (prepared by bromination of N-ally1 acet a~u ide ccordingto t he literature) (26) in 130 1111 of d ry acetonitr ile was added, with st irring, 16 g of d ry and freshly precipi-tated silver carbonate. When all the silver carbonate appeared to have been converted to silver bromide(about 1/ 2 to 1 hour), i t was quicldy filtered a nd 100 1111of a 20% solution of dr y climethylainine in benzenewas added. Th e mixture was stirred and allowed to st and overn ight. Another portion of 16 g of d ry silvercarboilate was added and the precipitated silver bromide filtered off. The filtrate was evaporated i n vacr~oa t 40" and th e residue distilled in uacuo; a t 57-58' a t 5 mm , 8.5 g of colorless licluid ( SL II I) was obtained.I t gave a st rong band a t 1675 cnl-I in the infrared. Calc. for C;Hl,ON?: C, 511.15; 13, 9.85. Found : C , 59.30;H, 10.0%.The nzetltiodide (SLIV) was obtai~ledn high yield by reacting the base with 20% less than the theoreticalamount of methyl iodide in acetone (a purer product is obtained in this manner). The yield of colorlesscrystals, m.p. 134.5-135", was quant itat ive (based on methyl i odide). Th e product can be recrystallizedfro m methanol-acetone. Calc. for CsHliON2I: C , 33.8; H , 5.98. Fou nd: C , 33.7; M, .88%.


15/15

TRIGGLE AND BELLEAU: QUATERNARY SALTS 1215N-Acetyl-N'-dimethyl-1,s-dianzino-2-propanolethiodide (VL) y Hydrolysis of (XLIV)A solution of 100 mg of t he qua ter nary iodide (XLIV) was heated to 100' in distilled wate r for 60 hours.The water was evapora ted in vaczro and the residue crystallized from methanol-acetone, m.p. 119-120,unchanged by recrystallization. I t proved identical with a n authentic sample of (VL) synthesized as describedbelow.I-Phthalimido-3-dimethylamino-2-propano (IIIL)A solution of 1-phthalimido-2,3-epoxypropane40 g) in 250 ml of 20% dimethylamine in benzene washeated to 100' fo r 4 hours. Evaporat ion of the solvent gave (I II L) as a viscous oil. Calc. for C13Hle03N1:C, 62.8; H, 6.5; N, 11.29. Found: C, 62.3; H, 6.9; N, 11.01%.N-Acetyl- N'-dimethyl-I ,3-diamino-2-propanol Methiodide (VL)The preceding phthalimido derivative (I IIL ) was hydrolyzed by heating under reflux with concentratedhydrochloric acid for 10 hours. The phthalic acid was filtered from the cooled solution and the filtrate con-centrated in vact~o o give the dihydrochloride sal t as a n uncrystallizable oil. A crystalline dip icrate, n1.p.207-209" (isopropanol), was prepared. Calc. for CI ~ H ? O O I ~ N ~ :, 35.4; H, 3.5. Found: C, 35.4; H, 3.5%.The N,N-dimethyl-1,3-diamino-2-propanolihydrochloride (0.04 mole) was converted to the N'-acetylderivative by reaction with a n equirnolar amount of acetic anhydri de in some water containing 4 g of sodiumacetate. The reaction mixture stood for 2 hours, was neutralized with sodium hydroxide, and evaporatedto dryness. Th e residue was extracted with alcohol-ether and th e extract dist illed; a t 130' a t 0.01 mm theacetyl derivative (IIL) was obtained as a vicous oil which could not be induced to crystallize. Calc. forC7HlcO?N?:C, 52.5; H, 10.1. Found: C, 52.65; H, 10.1%.

The methiodide was obtained in the usual manner ancl crystallized from methanol-acetone, m.p. 119-120'.No depression of the melting point was observed when it was admixed with a sample secured by hydrolysisof (XLIV) as described above. The infrared spect ra of t he two compounds (Nujol m~111) ere superimposable.Calc. for C8 HI9O 2N?I : , 31.8; H, 6.3. F ound : C, 31.9; 11, 6.0%.

ACI

Date post:	06-Apr-2018
Category:	Documents
Upload:	smokeysam
View:	221 times
Download:	0 times

D.J. Triggle and B. Belleau- Studies on the Chemical Basis for Cholinomimetic and Cholinolytic...

Documents