Tetrahedron Letters No. 17, pp. 1073-1078, 1%3. pergamon Press Ltd. Printed in Great Britain.

AR4DIcAL-cATIGiiInITIAmPOL I?lEEzATIufap8-vINYLcARmmLE

Humy Scott, Olen A. Miller a& Mrtlwr H. Labes

QudatrgMrialolp,The~inInskl~~~boratoriea Phlladolphia 3, Pa.

(Received 29 April 1%3)

Ithe~~~c~~tnit~lkmeaca~ea

utlonic polymariaatioll of N-vinylcubasola at or aMe ita r.p., ti it

is poatnlatad ulatinitlat1on involvxta protontranafer from catalyatto .

monomer’. We have found that a solution polymerinatlon vltb rmendnt

~bilar characteristics can be effected instantly at or below xwm

temperature by ulditlon of tames or mre of collparmds hitherto unknown

to catalyxe additionpo~risation. l%a following collpoand4 weramat

effective: p-chloranll, o-chloxwnil,p-bronail, 2,3-dichloro-5,6-

dicyanobensoqulnone, l,h,?+t0tnchloX'o8n*qUinoIW, 7,7,8,8--t=-

qmqulnod~~ne, tetrag8noethylaae, ahlorhe, bramhe,iti,

nitrogendioxide,and the bra&de salts oftheWtr&er iW of&b- -m

diaminodurene ati 1,6_diminoppeae. Longer inltlctba pQ%d WN

encountered with trichlommx&mitrile, pyronllitic an@dride, p-iOdat&

andp-benzoquinone. Webelieva this tobe tha firatreportad axa@e of

a polyzwieation initiated by a radical-cation.

A 90% yield of poly(N-mlcarbasole) ME recovered w fllterin#

off precipitate formed at room taqarature by l ddiog 3 ml. &stun

reagent grade 1,1,2,2-tetrachloroethane (contains elmhal chlorine) to

1 g. N-vlrqrlcarbazole dissolved in 10 ml. apectroquality aceton.itHle.

The infrared spectrum of thepolywrwas the eamz aa that of thanrlly

polymerized N-tinylcarbazole. Polyvinylcarbazole remind dl.aeolnM whm

preparad in wthylene chloride, chlorofolr, bensene, ethylether,

1073

1074 Polymerization of N-vinylcarbazole No. 17

thiophene, or stymne. Polymsra pwpared at rooa -ntum with 2,3-

dichlom-5;,6-dicyanobenzoquinonrcLnone in acstmitrile, with trichloroaceto-

nitrue In acetmrLtrils, with chlorine (a8 above) in lrl acetonitrile-

metbylenechloride,andwlth chlorine (as abore) inmethylenechlorldeat

-bOV, had molecular weighta293 of WSO, 385'0, h830 and 18100 (18700)

respect&vely. 2,lr Their elementalcoMemtcorresponded to theory ,

Adarkmaterial,preamedtobe charge-tnrmferccmpl I92 ,m*

observe& to folr on the surface of a pinch of powdered p-chlormil or

tetra&noqnjnodiaetae gently dropped into a solution of N-tinyl-

carbaroile. Streams of pawdew polpar were obsemed lx ~9.80 up from such

muface* or those of Wurster salts when ac*tmitrile was u8ed as the

solvent, The polyrsrization ua6 qualitatively unaffected by the premnco

of co*ntively large amount4 of thiophene, a potent retarder of

convent$.onal cationic propagation', l crylonitrile, a radical or

anionic$lly readily polymericed monomer, and styrem, a ummer readily

polymerlsedbyall knommecbanisms7. Polymerization was retarded or

inhibit&d by addition of mall amounts of water, aniline, N,N-dimetbyl-

aniline,? triethylamine, N-methylqulnolinlum chloride or lithium chloride

(dioeolJred in acetonitrile). No polymerication occurred in reagent grade

methano;C, acetone, tetrahydrofuran, or N,N-dlmthylfoxuami.de.

The effects of water and amine*' Indicate that the polymerlsa-

tion c&alyzed by nitroalkanes and the coapouxls we describe involve a

cationic mechanism. It seema unlikely, however, that proton transfer is

imolve$l in the initiation of the polymeri5ation we have observed because

most of the catalysts do not contain wrogen anl they are probably too

electro@ilic to serve as transfer agents for protons from other possible

aource81,

We sagger& that this ia a c&Ionic polymerization initiated bY

No.17 Polymerization of N-vinylcarbazole 1075

a Uurster radical-cation formed by oxidation of N-rinylcarbasole. rha

proposed scheme for a polymerls*tioncatalyeedbya caporud kn0wn to

fom charge-transfer complexas is shharn below.

M(onomsr) + O(xidant) e I¶.O(CT caplex) e I¶'+ + O*-

This schsms is COMi8tantwith the knownforrtion 0fUurster

ions as secondary products fol&ming forvtion of charge-transfer

coqllexes of amine8 nlth quinones8. Manyof the utalyats are known to

form charge-transfer oqlexes with aromatic amines. The halogens and

Wurster salts ue the main cnoeptioas9. Halogens, homevar, are knonn to

react with aromatic amlnes to fiieldkiorster Ions as principalprodu~ts~~.

AWtunterionofanamine otterthan 1oIIQc)r probably initiatis polymerha-

tion by adding to monaar or exchanging an electron with monomr ta forr

the Wnrstar ion of N-vlq&xrbasole.

The initiatlonmd~gationprobabl~do not involve inter-

mediates withindependentradical and ionic fvnctkas suchaa those

forad inpo~isatlons involving reducedmonomrs~' It ir reaao=ble

to aasnme thatthepositlve charge of the WurrWlon of N-~lcax+wole

is concentrated about the nitrogenatomaxl that a aeparation of Ionic

and radical functions would involve a quaternarizatlon of the nitrogen

atomand conseqwntlya deactlntionaf the reactivityof the utlonic

function of the radical-cation (e.g. >&H=cX~+ >&CH-a,.). In

poQmerizatlons of N-wlcati%ole In styrene and acrplonitrile, the

envlrmnt of the radical-oation consists almost entirely of molecules

which are saaceptible to free radical initS.atsd polyrsrization. The

inertness of these sokents indicates the absence of an intemte

with a localieed carbon-xudlcal fuuction. This also suggests that no

separation of radical and Ionic functions of the Wurster ion occurs,

1076 Polymerization of N-vinylcarbazole No.17

l ul it ir cmsltantllitll tllapropel&iar of kncmWtlr5tarllnls12.

Tha aimuaritq of polymeri5ationcl Induced by diffen3ntl.y

constitated catalyuta enggestthatinthe course ofpropagationand

tel¶inatioa no interaction othsr than ionic l soociation occura between

the cation CoMlntling ofthepolymeric addllctofuurstexioI!ldthmo~er

and the anion representing red~~ced cat&at. This in reasonable, em

for casasllkelytoinvol~ radical-anions suchas tbo8e formedby

red~ctla~~ of @nones, since anions of this type are known to be capable

of co-ez&3tingwlth 8Wurater ion for atiae %llchl~rthanthat

reqaimd to complete a polymerl5ation. Addition of chloride ion

probablgrre8nlta in the fomtion of an ion-pair invoWIng this anion

regax&3 of theoxldantueedtoproduce theWnrsterion,and the

retaldii/g effect of ~~thylquinolinium and lithium chloride on the

polymrijsation may be due to coa5on ion suppression of dissociation of

thin ioq-pair at an early stage of polymerization.

The scope and mechanisll of thin polymerization an, being

studied in more detail. One might expect monomera which am notes

readily ;kidized as aromatic amines to be less swceptible to the

powr$eation we have described. N-vinyl-2-pyrrolidone and N-vi@-

phthal.i&ie were polymerized by chlorine and quinones, but not a* readily

as I-vU@carbasole. The application of &dative catalysis does not

appear to be limited to the polymerization of N-viuylated monomrs. A

polymer of kiqlpyridine in the foorm of a charge-transfer coaq~lex with

p-ChloraMlwas obtainedfranthe reactionofmon-rwith this &none.

No.17 Polymerization of N-vinylcarbazole

1

2

3

h

5

6

7

0

9

10

l.l

12

FlEFERENCES

0. F. Solomon, M. Dimonic and T. 1 (1962).

Tomeecu, Hacromlek. Chm. % ='

1077

Single detenninatiom after Soxhlet eXtraction with methmol.

Hechrolab Vapor Presslrre Osmometer. Analytical accuracy decreases

for i$ > 5000.

Chlorine contents were 0.16, 0.37, 0.58 and 0.33% respectively. Lack of quantitative correlation with molecular weight is probably due to experimental errors involved in determining chlorine content by A

standard procedure.

The formation of complex on the surface of solid p-chlozmnil has been identified with l,&diaminodurene in solution. and H. M. Labes, Nature 197, 375 (1963).

See R. Scott, H. Ur,

-

C. 0. Overberger and G. F. hdres, J. Polymer Sci. l6, 263 (1955).

These compounds umre the only solvents paasent. No absorptions other than those of poly(N-vi@carbazole) were detected in the infnzwd spectra of methanol precipitated products.

B. Kainer and A. Uberle, Rer. g, ti7 (1955); W. Liptay, G. Briegleb and K. Sohindler, 2. Elektrochem. g, 331 (1962).

The only other exception is nitrogendiozzldeuhich canbe safely assumedtobe capable ofoxidizingmost aromatic amims to their Wursterions. We have prepared charge transfer complexes of trichlorcecetonitrile with aromntic amine8 in unpublished work.

L. Michaelis and S. Granick, J. Am. Chem. Sot. 65, 17h7 (19h3).

M. Szuarc, M. Levy and Ft. Milkovich, J. Am. Chem. Sot. 78, 2656 (1996).

Although lmown Wurster ions do not contain a polymerizable double bond, it is interesting to note that they are monomeric ruder cozparable conditions, i.e. they are paramzgnetic in solution. To account for the diamagnetism of most solid Wurster salts, Michaelis (ref. 10) postulated that Wurster ions undergo dimerization in the solid state, although he recognized that this could not involve a nonzalcovalentbond. whys. 36, 2393

H. M. McConnell and R. Lynden-Bell, J. Chem. (1962), have recently explained the shzrp decrease

in the magnetic susceptibility of solid Wurster salts in texmz of a spin-pairing phencnana which does not involve the foation of A valence bond.

1078 Polymerization of N-vinylcarbazole No.17

u J.W.~tm,&j&dN udtf. ~~~vIII,J.A~.C~CII&~OG &1339

(1p92); J. W. lhdun, 0. Il. Androe a If. cilh, m a, 1067

(1%2).