'etrahedron Letters No. 45, Pp 4495 - 4490, 1973. Per-on Pr888. Printed in hat Britain.

ASIM!?LEIiDVPE'IO~YCI-ILOK.EDI!ZIt@IWG!CJIBANDANDSYSTEW

J.Mwkin,O.MsMkim*aadH.Susscfiitzky

Departrnentofc%eInistryand~lled~stry,

Wlvexsity ofSalfonI,SalfordU5 ~,Lancashire,Ehgland.

(Reoeived in UK 17 September 1973; accepted for publication 28 September 1973)

DuringatMqtstiprepare N-sul#mylanilim derivatives @htWS02) weexaminedthe

lnteracticn of sul@ryl &loride and o_amincdialkylmillms (1).

R c: B C 4 #+2 CY

' ?"

0 : ’ NH2

so*cl> = Cl/

0 I

CIA Cl Cl

a;R=H

b:F-R=(CT&

c; R- R = (cH213

d;R-R=(CH$4

e; R -R= (CX2OUi2)

(1) (2)

Arermrkablytigorous,clesnbutm~ Imic,reactlctlcxclm inthecoldsndinthe

absence of solvent, resulting in chlorhatim and cyclisatim. Inmost cases, agwdyield

of ths tatrachlorckenzirfddls (2) was isolated (Table). !the prcduct (2a) frun o-dimzthyl-

aminoaniline (la) was identical to thatpxzpazdky the four stags synthesis ofNewbold and

(zxm?Aers? mePrmedure is equally applicable to the correqmding

-quinollne (51, and -anthqhns (71, givins the analogOU3 ak&sfA

Cl

(3)

aminawridine (3),

irni.dazoles (41, (6)

4496 No. 45

(7) (8)

and (8). Iimever, the o_aminoFhenylpyrrolidine (lb) gave a mixture of the corres~g

-r a-, snd txi~zle (cf. 2b).

&be&w thatthe reacti~ini.tlsllyfolhis twodistinctpathways (Scfiem),cne

yielding the sulphQlyl.tmlne (11 or/and 12), while the other involves electrqhilic

ChWticnof thehighly activatedsmmticnuclevs. 9he shsequant cyclisaticm of the

ti#nxylamine tithehnzimidazolemaybe envisagedtocxxur eitherbywayofthenitrene

TABLE

Productsderi~frantheadionof~2C12cn~alkylimLinoaryl2s

Startlnq w Produd Yield(%) m Ramarks --

la 2a 50 255' lltl, m.p. 258-61'

lb 2b 131 mixhxe ofmno-,di- andtrichloro -derivatives

lc 2c 60 188

Id 2a 85 232

le 2e 50 143

3 4 40 243

5 6 40 246

7 8 40 269

No. 45 4497

(la) SO7CI7

/I C'&JCl

Cl -so2 (13)

1121

\

P Cw+ Me 6 &so,

Cl ‘Lb/

Cl 6 Iv-

/ Cl Cl

(12a)

Me

Mc.&CH2)

Cl/

b I

g-soy

Cl\ Cl Cl

SCHEME

(13) (knmm to yield hnzinddazoles2) pmduced by loss of S02, or by the bkranulecular

H-abstractlm and cycllsaticn (vlaUa),by analogywithnwmrous relatedcyclisaticns3.

ItaFpearsthatbenzimidazolef~~anisslaJerthanchlorinaticnsinaethemch~~~

benzimidazolesgiw9mlyamlxtureofnmo- anddich lom-&rivatwesalY?aacticnwlth

sulphrnylcfilorids. Ihe fact that the ~lidine &rivative (16) didnotyielda fully

chlorinatedpmduct (2b)wazld~catethatinthiscaeethecycllsaticnetcrpis~

rapia thannuclearchlcninatkn. mshdlre&lysvFpartsthesecmdcyclieatlanlmcilanianl

4498 so. 45

(l2+l2a) sinae~h~f~otherexrmplesin~cfithe~rrolidinederiva~~sare

cycused faster thm the rest.

It is no-y UlattiN~l-hetemq cles (e.g., 13) giw solely vie

mrreqcmdlng sul#wnyl d-&rides (e.g., 14) under the abow azndlticns, while the N-acetyl-

(13)

NHAc

(14) (15)

derivative gave no benzimidazole (cf. 2) but a cuplex mbture of chlorinated proaucts.

I-

l. D. E. Eurtm, A. J. Lmbie, D. W. J. Lane, G. T. Newbold md A. Percival,

J. Clmn. Sot. (C), 1968, 1268.

2. 0. Mzti-C&n, R. K. Smalley and H. SusMtzky, J. Umn. Sot., 1963, 1666

3. 0. Msth-Cchn andH. Suschitzky,Advanc8s in Hetemcyclic Chemistry, 1972,g, 211.