:

II

1 5 1

Journa l o f Organometa l l ic Chemist ry , 1 .34 (1977 ) 151-163O Elsev ier Sequoia S.A. , Lausanne - Pr in ted in The Nether lands

THB STEREOCHEMISTRY OF THE PETERS- RHgX)

REACTION (RSO'HgX

LAWRENCE C. COSTA, G. BRENT YOUNG and GEORGE N,{ . WHITESIDES '

Depar tment o f Chemist ry , Massachuset ts Ins t i tu te o f Technology, Cambr idge, Massachuset ts02139 (U .S .A . )

( Received September 2nd, 1976)

Summary

Thermal decomposition of either exo- or endo-2-norbornylsulfinatomercuricchlor ide (exo-A or endo-A) in DMF at 100'C yields only exo-2-norbornyl-mercur ic chlor ide. Under s imi lar condi t ions, erythro-3,3-dimethylbutyl-1-sulfinatomercuric chloride-7,2-d2 yields a Ll7 mixture of erythro- and threo-3,3-dimethylbutyl-1-mercuric chloride-1 ,2-dr. endo-2-Norbornylmercuric chlorideis not epimirized under the reaction conditions, and endo-A is recovered frompartial decompositions without detectable epimerization. These results suggestthat the product-forming step proceeds via a stereochemically labile intermediate.Although several qualitative experiments provide evidence against a free radicalchain mechanism, the detailed pathway of the reaction has not been established.

Introduction

A number of methods are available for generating carbon-mercury bonds[1]. One of the less frequently employed routes is the Peters reaction [2], inwhich organomercury(II) chlorides are obtained on heating a mixture of mer-curic chloride and a sulfinic acid or its sodium salt. A variety of arylmercuric

RSO2H(Na) + HgCI, 5 RHgCI + SO, + (Na)HCl ( 1 )

Achlorides n"": tl#il#;'using this reaction, but onty a rew alkylmercuricchlorides have been prepared. Although yields for the synthesis of alkylmercurichalides by the Peters reaction are modest ((50%\, it seemed possible thatthisreaction might have useful stereoselectivity. There are presently no generallyapplicable procedures for the preparation of organomercury compounds withdefined stereochemistry. Although tris(3,3-dimethylbutyl)boron is converted

t52

by mercuric acetate to the corresponding organomercury compound with in-version of stereochemistry [3], secondary alkylboranes are not cleaved by mer-curic acetate [4], and the methods available for the conversion of secondaryalkylboranes to organomercury compounds are not stereospecific [5,6]. Othermethods available for the preparation of diastereomerically and enantiomerical-ly pure organomercury compounds [7,8] are laborious, specialtzed, and do notalways yield stereochemically pure product.

Results and discussion

Preparation of starting materials. The required sulfinic acid salts were synthe-sized by two methods (free alkylsulfinic acids disproportionate rapidly [9]). Theachiral salts employed as model compounds were obtained by reaction of theappropriate Grignard reagent with sulfur dioxide [9,10]. The sodium salts usedfor stereochemical studie s, e x o -2-norbornylsulfinate (e x o -I), e n do -2-n orborny l-sul f inate (endo-I) , and erythro-S,3-dimethylbutane-1-sul f inate-I ,2-d, ( I I ) , wereobtained by adaptat ion of the method of Truce and Roberts (Scheme 1) t111.

SCHEME 1. Synthes is o f sod ium a lky l su l f ina te .

r q - _ u i L t \

R S Hc o t C H

M C P B A , C H C I 3

No S Ca ts^R S C r r j o < --

c 2 H 5 o H

o

R S C H . C H ^ C Nl l "t lo

)

IH

-\ / - ,

S C 2 N o

e n d o - I

S C 2 N o

e x o - f

S O r \ l o

ll

Attempts to oxidize thiols directly to sulfinic acids using ne-chloroperbenzoicacid (MCPBA) t72l were unsuccessful.

Both exo-2-norbornanethiol (exo-III) and erythro-3,3-dimethylbutane-1-thiol-1,2-d2 were prepared by base-catalyzed transesterification of the correspondingthioacetates [13]. The thioacetates in turn were obtained by unambiguousroutes: exo-Z-norbornylthioacetate was prepared directly from norbornene andthiolacetic acid [14];erythro-S,3-dimethylbutyl- l- thioacetate-1 ,2-dr, a part icular-ly convenient substrate for NMR determinations of stereochemistry [3,15],was prepared by S*2 displacement of tlne threo-p-bromobenzenesulfonate bythioacetate in {l/-dimethylformamide (DMF). The preparation of endo-2-norbor-

| - - t t

5 3

nanethiol (endo-I l I ) was not so straightforward. A var iety of sul fur-containingnucleophi les ( th ioacetate, th iocyanate, and thiourea) were found to be unreac-t ive torvard exo-2-norbornyl bromide and/or the corresponding p-bromobenzene-strifonate and p-toluenesulfonate. In fact, reaction of exo-2-norbornyl p-bromo-benzenesulfonate with the sodium salt of benzyl mercaptan in DNlF resultedonly in displacement of bromide from the aromatic ring. The reaction of exo-2-norbornyl bromide with the sodium salt of benzyl mercaptan in HMPA proceed-

1,,{ a

, , t / l \ , . r -__} PnCHzSi L /i \,\ :]-->>

?' -,-r-, - :1' t)- ".

DM F

lt \_/H C

-\. o

i- /.-!_.1A>sc!2c6H-l - - l l \ - , '

za '

\

/>p) ' -1)a , + NcSCtzCoHs

i

ff^-)/"^- ''i'-N

i\ - a ! / !

H M P A

e n d c - N

- - - . ( 1 ) N o / N t r ,

fr/ --^\,+' \

/ - t )

''-/'-r/-

"

S H

e n d o - I I L

ed readily, however, and gave endo-2-norbornyl benzyl suifide (encto-IY) in goodyield. The benzvl group was cleaved with sodium in l iquid ammonia [16] andyielded endo-2-norbornanethiol (endo-Il\.

The NMR spectra of endo-III and exo-III were distinct; in particuiar, the C(2)methine proton resonance of endo-III appeared as a six-component multipletcentered at 3.2 ppm, whi le the C(2) methine proton resonance of exo-I I I appearedas a four-component multiplet centered at 2.8 ppm. Verif ication of the stereo-chemical assignment for endo-IY was obtained similarly by comparison of theNMR spectra of endo-IV and exo-IY . The latter was prepared by the addition ofbenzyl mercaptan to norbornene t14].

Authentic samples of alkylmercuric chlorides were prepared by reaction of theappropriate Grignard reagent with mercuric chloride. endo-2-Norbornylmercuricclrloride (endo-Y) was obtained by reaction of endo-2-norbornylmagnesiumbromide t17] wi th mercur ic chlor ide. ero-2-Norbornylmercur ic chlor ide (exo-Y)was obtained by adaptation of the preparative procedure for exo-2-norbornyl-mercur ic bromide t181.

l-54

tlodel reactions. The Peters reactior-r has normally been effected by mixingaqueous or alcoholic solutions of mercuric chloride and the sulfinic acid or itssalt, and heating the resr-rlt ing insoluble organosulfinatomercuric chloride [191as a suspension in the solvent employed at 80-100'C for several hours [2] . Theintermediate S-mercury(II) sulfinato complexes can be identif ied by characteris-t i c s t rong IR absorp t ions a t 1000-1100 cm- l and 1100-1250 cm- ' [19 ] .

When this procedure was applied to sodium cyclohexane sulfinate in absoluteethanol, the intractable solid that was obtained showed weak IR absorption banclsdue to cyclohexylmercur ic chlor ide, but the product could not be isolated.Simi lar resul ts were obtained from a mixture of endo-I and exo-L These di f f i -cul t ies could be avoided by running the Peters react ion in the polar, aprot icsolvents pyr id ine, d imethylsul foxide (DN{SO), DMF, and l / -methylpyrrol idoner(NMP), in which the alkylsul f inatomercur ic chlor ides were soluble. When thesesolut ions were heated to ca. 100'C for several minutes, they rapidly becameopaque. The desired alkylmercuric chlorides precipitated on quenching withwater, and could be isolated in y ie lds of 30-607o.The highest y ie lds were ob-tained in DMF, and this solvent was employed for further studies. Applicationof this procedure to representative sodium alkyl sulfinates gave alkylmercuriccirlorides in the yields summarized in Table 1.

Thermal decomposition of n-hexylsulfinatomercuric chloride in the solidstate also produced alkylmercury compound. When a port ion of the formercompound was heated in a subl imator at 150'C and 0.01 Torr , crystals of n-hexyl-mercuric chloride slowly grew on the cold finger. The IR spectrum of thismaterial also corltained weak sulfur-oxygen stretching bands of the startingmaterial. This procedure was not preparatively useful.

Stereochemical resul ts. T?eatment of aqueous solut ions of exo-I , endo-I andII with an aqueous solution of a 50-100% excess of mercr,rric chloride resultedin the precipitation of the corresponding alkylsulfinatomercuric chlorides: exo-2-norbornylsulfinatomercuric chlon de (exo-YI), endo-2-norbornylsulfinato-mercuric chlorid e (e ndo -Y I), and ery t hro - 3, 3-d imethylbutyl- 1 -sulf inatomercu ricchlor ide-1,2-d2 (VII) . The stereochemical assignments for exo-YL, endo-YI andVII were based on NMR analyses analogous to those described above for exo-I I I , endo-I I I , and e ry thro - 3, 3 -dimethylbutyl- 1 - th io acetate-1,2-d r .

When solut ions of e i ther exo-Yl or endo-Vl in DMF were heated to 100'C(5-30 min) and the resulting crude mercurial precipitated with water, onlyexo-Z-norbornylmercuric chloride (exo-Y) could be detected by IR. The stereo-chemical assignment was based on comparison of the IR spectnim of the productwith the IR spectra of mixtures of exo-Y and endo-V of known composi t ion.

TABLE 1

A L K Y L M E R C U R I C C H L O R I D E S P R E P A R E D B Y T H E P E T E R S R E A C T I O N

Product Iso la ted v ie ld (qo)

2 ,2-D imethy lp ropy lmercur ic ch lo r ide

n-Hexy lmercur ic ch lo r ide

Cyc lohexy lmercu r i c ch lo r ide

3, 3 -D imethy lbu ty lmercur ic ch lo r ideexo- 2 -N orborny lmercur ic ch lo r ide

3 5584 54 83 9

1 5 5

An epimeric mixture contair"r ing only 207o of the minor component could haveeasily been identif ied. Attempts to confirm these assignments by conversionof V to norbornyl bromides by reaction with bromine gave erratic results U,201.A report that endo- and exo-2-norbornyl(methyl)mercury(II) can be separatedby GLPC appeared too late to be used in th is work Lzf l .Using simi lar react ioncondi t ions, a solut ion of VI I in DMF gave products ident i f ied on the basis ofcoupl ing constants to be a 1,17 mixture ( t57c) of erythro- and threo-3,3-dimethyl-butyl-1-mercur ic chlor idc ' - I ,2-dz. These resul ts indicate that , under the condi-t ions employed, the product of the Peters react ion is completely epimerized( S c h e m e 2 ) .

ScHEME 2. The Peters reac t ion proceec ls w i th Ioss o f s te reochenr is t rv .

\l , - \

. , 4 \

, , i ) D M F , i c o " ci l r - n

' ' Q f ) L n r -r *

J v 2 , , Y !

i

H

exo-U

D M F- -l>

1 0 0 " c

S O r H g C l

( ! t ;

H +

'#-'),H D

"lr]",1,

hS O r H g C i

e n d o - Y T

The Peters reaction might follow a concerted or a nonconcerted mechanism.If the reaction is concerted, it should be stereospecific. The observed loss ofstereochemistry should then occur prior to or subsequent to reaction. If thereaction is not concerted, it might involve a stereochemically labile intermediate:vrz. ,a carbonium ion, a carbanion, or a f ree radical . A r igorous mechanist icstudy of the reaction has not been performed, and the questions of where andhow the stereochemistry is lost cannot be answered precisely. Several experi-ments have, however, been performed which provide qualitative evidenceagainst some of the various possibil i t ies.

When the Peters reaction was run in the presence of endo-Y using cyclohexyl-sr"r l f inatomercur ic chlor ide as the reactant (DMF, 100"C, 30 min), the productmixture consisted only of cyclohexylmercuric chloride and endo-Y. No ero-Vwas detected by IR. In addition, when the Peters reaction was run in the presenceof 5-hexenylmercuric chloride with n-hexylsulfinatomercuric chloride as thereac tan t (DMF,100 'C,30 min) , the produc t mix tu re cons is ted on ly o f n -hexy l -mercuric chloride and 4-hexenylmercuric chloride. No cyclopentylmethylmer-curic chloride was detected by IR. These results preclude the intervention ofcarbonium ions t23l or free radicals [24,25] derived from the product mercurials

t - B u

! n e l' J

- '

1 5 6

as the agents responsible for their loss of stereochemistry x.When a solut ion of endo-Vl in DMF was heated brief ly at 100"C (3-4 min)

and then rapidly quenched by pouring into ice/water, the recovered startingmaterial was found not to have undergone any detectable epimerization byNMR. The observation that both the starting sulfinatomercuric chloride andthe product alkylmercuric chloride are configurationally stable under thereaction conditions indicates that loss of stereochemistry occurs during thePeters reaction itself.

Carbanions do not seem likely as intermediates, since the reaction can becarried out successfully in protic solvents and in DMF. A free radical chainmechanism, or a mechanism involving free radical intermediates in other capaci-ties, is also unlikely. The qualitative rate of reaction and the yield of productwere insensitive to the presence of radical scavengers such as oxygen 1261,bromotrichloromethane, and hydroquinone. Reaction of VII in the presence of\ mol% of hydroquinone gave a Il7 mixture of erythro- and threo-3,3-dimethyl-butyl-1 -mercuric chloride-L,2-d 2. Reaction of n-hexylsulfinatomercuric chloridein the presence of 1 equiv of hydroquinone gave n-hexylmercuric chloride in ca.407o yield. The same reaction run in the absence of hydroquinone gave ca. 45%yield. The difference is not significant. No difference could be discerned if thereaction was run in anhydrous, distil led DMF under nitrogen, or in reagent DMFtaken directly from the bottle under air. Addition of 7-27o water by volume tothese DMF solutions also had no effect on the reaction.

Experimental evidence that would definitely rule out carbonium ions orsolvent-caged radicals as intermediates has not been obtained. The facility withwhich primary alkylsulfinatomercuric chlorides undergo the reaction would,however, argue against carbonium ion formation. Alkylsulfonyl radicals, generatedby radiolysis of alkylsulfonyl chlorides in cyclohexane, do not undergo signifi-cant decomposition to sulfur dioxide and alkyl radicals t271. If alkylsulfonylradicals are formed during the Peters reaction, their behavior under these condi-tions need not necessarily parallel that which was observed in cyclohexane.

It is thus not clear what the detailed mechanism of the Peters reaction is, butit appears that the reaction is not applicable to the stereoselective synthesis ofalkylmercuric halides. Since the starting materials are difficult to obtain, andsince the reaction does not proceed in high yield, it seems unlikely that thePeters reaction will find application in routine synthesis of alkylmercury com-pounds, although it may prove useful in specialized instances in which otherpolar functionality in the molecule precludes the use of the more conventionalGrignard- or organolithium-based procedures.

Experimental

General methods. All reactions involving organometallic compounds wereconducted under nitrogen using standard techniques for handling air-sensitive

* An effort to check this conclusion by studying the loss of sulfur dioxide from b-hexenylzulf inato-mercuric chloride was abandoned when i t was establ ished that reaction of sodium-5-hexenylsulf inatewi th mercur ic ch lo r ide y ie lded on ly a compound ten ta t i ve ly ident i f ied as 2-ch lo romercur imethy l -th iane-1 .1-d iox ide .

757

compounds [28]. Diethyl ether and tetrahydrofuran were disti l led undernitrogen from solutions of sodium benzophenone dianion. Pyridine, l/,N-di-methylformamide, and.A/-methylpyrrolidone were disti l led under nitrogen fromcalcium hydride. Melting points were recorded on a Thomas-Hoover capil larymelting point apparatus and are uncorrected. Boil ing points are uncorrected.Proton NMR spectra were obtained at 60 MHz on a Varian T-60 instrument.Chemical shifts are reported in ppm downfield from internal tetramethylsilane.Deuterium-decoupled proton NMR spectra were recorded on a Varian HA-100instrument which was modified for heteronuciear decoupling. Infrared spectrawere obtained on a Perkin-Elmer Model 237,2378, or 567 instrument. L iquidsamples were run as thin fi lms between potassium bromide plates; solid sampleswere run in potassium bromide pellets (37owlw). Elemental analyses werecarried out by Robertson Laboratory, Florham Park, N.J.

A sample of threo-3,3-dimethylbutan- l -o l -1 ,2-d, was provided by Dr. paulL. Bock t15l . exo-2-l\orbornyl bromide l,2gl, endo-2-norbornylmagnesiumbromide [17], and exo-2-norbornyl th ioacetate [14] were prepared using l i tera-ture procedures. threo-\,3-Dimethylbutan-1- oI-1,2-d 2 was converted to thecorresponding p-bromobenzene sulfonate by a l iterature procedure for thenon-deuterated material 1221. Commercial samples of p-bromobenzenesulfonylchloride and p-toluenesulfonyl chloride were recrystall ized from light petroleumether. AII other reagents were from commercial sources and were used withoutfurther purif ication.

Sodium alky lsulfinates from Grignard reagents. General procedure [ 10 ] . TheGrignard reagent solution was cooled with a Dry Ice/acetone bath so that theinternal temperature of the solution was -30'C. Anhydrous sulfur dioxide waspassed into the mechanically stirred solution through a 15-gauge hypodermicneedle, and the internal temperature was maintained below -20"C. A whitesolid formed immediately, and sulfur dioxide acldit ion was continued unti l nomore solid precipitated. The mixture was diluted with an equal volume of ether,and then sufficient aqueous,707o HCI was cautiously added to dissolve thesolids and bring the aqueous phase to pH 1. The yellow ether layer was removedand washed with small portions (10-20 ml) of saturated, aqueous sodiumcarbonate solution unti l the pH of the collected washings was 7-8. If the wash-ings had a pH ) 8, the product sulfinate was contaminated with sodium car-bonate, and reaction of these contaminated sulfinates with mercuric chlorideproduced alkylsulfinatomercuric chlorides which were contaminated withmercuric carbonate. Evaporation of the water at room temperature underreduced pressure left the sodium alkylsulfinate as a white powder which wasnot purif ied further. The following sodium alkylsulfinates were prepared inthis manner (alkyl group, y ie ld) : n-hexyl (7G%), cyclohexyl (68%), 2-butyl (b\To),2,2-drmethylpropyl ( 8 3%), 3,3-dimethylbutyl ( 8 1 To),, b-hexenyl (42To) .

Alleylmercuric chlorides from Grignard reagents. General proced.ure. TheGrignard reagent solution (typically 15 mmol) was cooled to 0'C and treatedwith 1 equiv of mercuric chloride. The mixture was stirred at 0'C for 4 h, andthen quenched with 50 ml of water. The crude product was collected by fi l tra-tion, dried, and recrystall ized from ethanol/water (217). The following compoundswere prepared by this procedure endo-2-norbornylmercuric chloride (endo-V) (3570, f f i .P. 119-120'C), 2-norbornylmercur ic chlor ide (ca. 1/1 mixture of

1 5 8

ep imers , 657n, m.p . 169-171 'C) , n -hexy lmercur ic ch lo r ide (77%, m.p . 119-I20"C, l i t . [4 ] , m.p . 127-122"C) , cyc lohexy lmercur ic ch lo r ide (64%, i l .p .161-1 62" C, l i t . [ 30 ] , m.p . 163-164"C) , 5 -hexeny lmercur ic ch lo r id e (68%,

m.p . 1 00-101 'C) , cyc lopenty lmethy lmercur ic c rh lo r ide ( 48%, m.p . 57- -58 'C.l i t . [31 ] m.p . 55"C) , 3 ,3 -d imethy lbu ty lmercur ic ch lo r ide (73%, , m.p . 732-134"C.l i t . [22] m.p. 732-734" C), 2,2-dtmethylpropylmercur ic chlor ide ( 68%, In.p.113-114"C, , I i t . [32 ] m.p . 117- -718 'C) . E lementa l ana l5 rses fo r new a lky lmercnr icchlor ides are given in Table 2.

exo-2- l {orbornylmercur ic chlor ide (exo-V). This compound had a m.p.193-194"C and was prepared by the procedure for exo-2-norbom5' lmercur icb r o m i d e [ 1 8 1 .

Potassium thioacetate. A non-aqueous preparat ion which is strper ior to thel i terature procednre [33] was developed. To a sLlspensior-r of 18.0 g (130 mmol)of ar-rhydrous potassium carbonate tn 25 ml of absolute methanol was added5.0 g (65 mmol) of th ioacet ic acid i r - r smal l port ions. Wher-r carbon dioxideevolut ion had ceased, the mixture was st i r red an addi t ior"ral 10 min and f i l teretc l .The unreacted potassium carbonate was tr i turated with 10 ml of methanol ,and the combined methanol ic solut ions were t ransferred to a 1- l i terr F)r le 'nmeyerf lask. Ether was slowly added to th is solut ion unt i l i t became c: loucly, ar-rd themixture was cooled slowly to -7 8"C. The white needles of potassium thioacetatethat crystal l ized were col lected by suct ion f i l t rat ion, washed with 100 ml ofether, and dr ied under reduced pressure. The yield was 6.85 g (93%).

ery th ro-S.S-Dimethy ' lbu ty l -1 - th ioaceta te- l ,2 -dr . To a s t i r red so lu t ion o f 8 .00 g(70 mmol) of f reshly prepared potassium thioacetate in 75 ml of DMF wasadded a so lu t ion o f 20 .3 g (63 mmol ) o f th reo-S,3-d imethy lbu ty l -1 -p-bromo-benzenesul fonate in 50 ml of DMF. Within a few minutes, the mixture hadbecome warm and a white solid had precipitated. The mixture was stirred over-

T A B L I ] 2

E L E I \ l E N T A L C O M P O S I T I O N S

C o m p o u n d

c r r d o ' 2 - N o r b o r n l ' l m e r c u r i c c h l o r i d e

cr o - 2 -N orb ornv Ime ' rcur ic ch lo r ic le

5-Hexeny lmercur ic ch lo r ide

n- I Iexv lsu f ina tomercur ic

ch lo r id t :

C v c l o h e x v l s u l f i r r e r t o n t t ' r c u r i c

c h l o r i d e

cro - 2 -N orb orn) ' l su l f ina to-

mercur ic ch lo r id t '

2 , 2 -D im e th I ' Ip ro r ry ls t r l f i na to-

mercur ic ch lo r ide

cr1 ' l / r ro -3 ,3 -D imethy lb t t t v I -1 -

su l f iua tornercu r i c ch lo r ide-

7 , 2 4 2

F 'ornru la

C 7 H 1 1 C l H g

C 7 H 1 l C l l l g

C 6 I I 1 1 C l H e

C 6 H l 1 5 O 2 C l H e

C 6 I I 1 1 S O 2 C l I { g

C 7 I l 1 1 S O 2 C l H g

C 5 I I 1 1 S O 2 H g

C 6 H 1 1 D 2 S O 2 C l H g

A n a l y s i s c a l c d . ( f o u n d ) ( % )

U

2 5 . 3 8( 2 5 . r e )2 5 . 3 8

( 2 5 . ( i l )2 2 . 5 7

( 2 2 . 2 3 )1 8 . ? o

( 1 8 . 7 1 1 )1 8 . 8 0

( 1 8 . 7 4 )21.27

( 2 1 . 3 7 )1 6 . 1 7

( 1 5 . 6 7 )1 8 . 7 0

( 1 8 . 7 5 )

I I

: t . 3 5( 3 . 3 0 )

J . . 1 C

( 3 . 3 6 ):1 .17

( 3 . 2 7 )3 . 4 0

( 3 . 3 4 )2 . 8 9

( 2 . 8 8 )2 . 8 1

( 2 . 6 9 )2 . 9 9

( 2 . 8 5 )3 . 4 0

( 3 . 3 4 )

1 5 9

night at room temperature, and then poured into 600 ml of water. The aqueousmixture was extracted with four 50-m1 portior-rs of ether. The combined etherealextracts were washed with two 50-ml port ions of water, dr ied (NIgSOo), and theether was removed at reduced pressure. The remaining yellow oil was disti l ledthrough a 10-cm Vigreux column, and yielded 7.90 g (787o) of . erythro-3,S-dimethylbutyl-1-thioacetate-7,2-d r , , b.p. 83-84' ( 21 Torr) ; NMR ( 1 00 MHz,deuter ium-decoup led , CCl4) 2 .87 (1H, doub le t , J 12 .3 Hz) 2 .30 (3H, s ing le t ) ,I . 4 2 ( 1 H , d o u b l e t ) , 0 . 9 3 ( 9 H , s i n g l e t ) ; I R 1 6 8 5 c m ' ( r ) .

exo-2-I 'Jorbornanet l t io l (exo-I I I ) , A mixture of 9.32 (55 mmol) of exo-2-norbornyl th ioacetate and 0.30 g (0.5 mmol) of sodium methoxide in 25 mlof absolute methanol was refluxed under nitrogen. After 6 h, the mixture wasallowed to cool, and was poured into 100 ml of 57o aeueous HCl. This aqueousmixture was extracted with three 50-ml portions of ether. The combinedethereal extracts were washed with two 50-ml port ions of water, dr ied (MgSOa),and the ether was removed under reduced pressure. The remaining clear, foul-smel l ing oi l was dist i l led through a 10-cm Vigreux column, and yielded 5.90 g(83%) of exo-2-norbornanethiol (exo-I I I ) , b.p. 67-68'C/18 Torr ; NMR( C C l 4 ) 7 . 0 - 2 . 2 ( 1 I i , c o m p l e x ) , 2 . 8 p p m ( 1 H , m u l t i p l e t ) .

e ry th ro-3 ,3-D imethy lbu tane-1- th io l -1 ,2 -dr .Th is compound was prepared in afashion analogous to the preparat ion of exo-I I I in 88%,crude yield; NMR (CCl4)2 . 4 ( 7 H , m u l t i p l e t ) , 1 . 5 ( 1 H , m u l t i p l e t ) , 1 . 1 ( 1 H , d o u b l e t ) , 0 . 9 p p m ( 9 H , s i n g l e t ) .This sample was not purif ied further. The b.p. of a non-deuterated sample was45-48'C 120 Torr .

endo-2- l {orbornyl benzyl sul f ide (endo-IV). To a solut ion of 14.6 g (100mmol) of sodium benzyl mercapt ide (prepared from benzyl mercaptan andsodium hydr ide in ether) in 75 ml of HMPA was added 16.0 g (91 mmol) ofexo-2-norbornyl bromide. Within a few minutes, the mixture had becomewarm, and sodium bromide had begun to precipitate. After 2 h, the mixturewas poured into 500 ml of water, and extracted with three 50-ml portions ofether. The combined organic phases were washed with three 50-ml portions ofwater, dried (MgSOq), and the ether was removed at reduced pressure to leave18.1 g (97%) of crude endo-2-norbornyl benzyl sul f ide as a l ight yel low oi l ;N M R ( C C l 4 ) 7 . 2 ( 5 H , s i n g l e t ) , 3 . 6 ( 2 H , s i n g l e t ) , 2 . 8 ( 1 H , m u l t i p l e t ) , , 0 . 8 - 2 . 2ppm (10H, complex). This mater ia l was homogeneous by NMR and TLC, andwas not purif ied further.

endo-2-I , lorbornanethiol (endo-I I I ) . Ammonia (100 ml) was condensed into a250-pl three-necked flask equipped with a Kel-F coated stirring bar and a DryIce condenser. The f lask was then charged with 18.1 g (83 mmol) of crude endo-2-norbornyl benzyl sul f ide. Diethyl ether (75 ml) was added as a cosolventand gave a c lear, two-phase system. Sodium (0.46 g,200 mg-atom) was thenadded in smal l p ieces to the wel l -st i r red mixture. The blue color was rapidlydischarged at f i rst , and then more slowly as a sol id precipi tated. The color ofthe mixture changed from blue to yellow to green to brown. When all t l-resodium had been consumed, the ammonia was al lowed to evaporate, whi lepowdered ammonium chloride was added to the mixture to destroy any un-reacted sodium and sodium amide. The residue was triturated with 50 ml ofI07o aQueous HCI and extracted with two 100-ml port ions of ether. The com-bined ethereal extracts were washed once with 50 ml of water, dr ied (MgSOa),

1 6 0

and the ether was removed at reduced pressure. The remaining oil was disti l ledthrough a 10-cm Vigreux column to y ie ld 4.75 g e0%\ of endo-2-norbornal t -e th io l (endo- I I I ) , b .p . 73-74"C120 Tor r ; NMR (CCi4) 3 .2 (1H, mu l t , ip le t ) ,0 . 8 - 2 . 2 p p m ( 1 1 H , c o m p l e x ) .

Reactior"ts of t lt iols tt it l t acrylortitr i le. General procedure. To 1 equiv of thecooled (0"C), neat th io l containing three drops of Tr i ton B (A 40% solut ionof benzyl t r imethylammonium hydroxide in methanol) was added slowly twoequiv of acryloni t r i le. The mixture was st i r red under ni t rogen at 0"C for t h,and then allor,ved to warm to room temperature and stirred overnight. Themixture was treated rvith three drops of aqueous \0% HCI and diluted rn'ith anequal volume of ether. Insoluble material was removed by fi l tration, the fi l tratewas dried (MgSOo), and the volati le materials were removed at reduced pressureto y ie ld the crude alkyl 2-cyanoethyl sul f ide. The fol lowirrg compounds wererso obtained erythro-\ , ,3-dimethylbutyl-1 , ,2-dr 2-cyanoethyl sul f ic le (83' /c) ;NN{R (CCI4) 2 .7 (5H, mu l t ip le t ) ,7 .4 (1H, mu l t ip le t ) , 0 .9 pprn (9H, s ing le t ) .This material was not purif ied further, but a non-deuterated sample had lt.p.72-1 3 'C/0.10 Torr . Also, exo-2-norbornyl 2-cyanoethyl sul f ide (827o) wasobta ined as a g lass , f f i .p . ca .25"C; NMR (CCI4) 1 .7 -2 .3 (10H, complex \ ,2 .8ppm (5H, mult ip let) ;endo-2-norbornvl 2-cyanoethyl sul f ide was obtained as av i s c o u s s y r u p ( 6 6 % ) ; N M R ( C C l 4 ) 0 . 8 - 2 . 3 ( 1 0 H , c o m p l e x ) , 2 . 6 ( 4 H , m u l t i p l e t ) ,3 ' .2 ppm (1FI, mult ip let) . These lat ter two compounds were homogeneous byNMR and TLC, and were not purif ied further.

Oxidat ion of a lk l t l 2-cyanoethyl sul f ides. General procedure. To a cooledsolut ion of 2 equiv of MCPBA, (85%) in chloroform was added slorn ' ly a solut ionof 1 equiv of the alkyl 2-cyanoethyi sulfide in chloroform. The oxidatiorr is veryexothermic. The mixture was stirred overnight at room temperature, and theinsoluble m-chlorobenzoic acid was removed by fi l tration. The fi l trate waswashed once with a 50-ml portion of saturated, aqueous sodium bisulfite solu-tiot-t, and then repeatedly with small portions (20-40 ml) of saturated, aqueollssodium bicarbonate unti l carbon dioxide evolution ceased. The mixture waswashed once with 100 ml of water, dr ied (MgSOo), and the chloroform wasremoved at reduced pressure. The remaining oil was takelt up in a minimumamount of hot chioroform, and the solut ion was treated with hexane unt i l c loudy.Sufficient chloroform was added to clarify the solution, and it was allowed tocool . The product alkyl 2-cyanoethyl sul fone crystal l ized on standing. Thefol lowing compounds were obtained in th is manner erythro-3,3-dimeth5r lbuty l -I ,2 -d2-2-cyanoethy l su l fone (90%\ , f f i .p . 88-89 'C; NMR (CDCI . , ) 3 .1 (5H,m u l t i p l e t ) , 7 . 7 ( 1 H , m u l t i p l e t ) , 1 . 0 p p m ( 9 H , s i n g l e t ) ; I R 2 2 4 0 ( m ) , 1 2 5 0 ( s ) ,1100 cm-' (s) ; exo-2-norbornyl 2-cyanoethyl sul fone (947o, v iscous oi l ) ; NMR( C D C I 3 ) 7 . 7 - 2 . 4 ( I O H , c o m p l e x ) , 2 . 8 - 3 . 3 p p m ( 5 H , m u l t i p l e t ) ; I R 2 2 3 5 ( m ) ,1300 (s ) , 1725 cm' ( r ) ; endo-2-norborny l 2 -cyanoethy l su l fone (83%, v iscouss y r u p ) ; N M R ( C D C 1 3 \ 0 . 9 - 2 . 4 ( 1 0 H , c o m p l e x ) , 2 . 8 ( 4 H , m u l t i p l e t ) , 3 . 4 ( 1 H ,n r t r l t i p l e t ) ; I n 2 2 4 0 ( - ) , 1 3 1 0 ( s ) , 1 1 1 0 c m - r ( s ) .

Conuersion of a lkyl 2-cyanoethyl st t l fones to sodium al lzSt lsul f inates. Generalprocedure. To a soiut ion of 1 equiv of the alkyl 2-cyanoethyl sul fone in 50 mlof absolute ethanol was added 1 equiv of sodium thiophenoxide (preparedfrom thiophenol and sodium hydr ide in THF), and the mixture was ref luxedfor 4-6 h under nitroqen. The ethanol was removed at reduced pressure. and

1 6 1

the remaining sodium alkylsulfinate was triturated with 50 ml of ether, collectedby suction fi l tration, washed with 100 ml of ether, and allowed to dry. The fol-Iowing sodium alkylsulfinates were prepared in this manner erythro-3,S-dt-methylbutyl-1 -sulfinate-1,2-d, ( II ) ( 700o/o) ; e x o -2-norbornylsulfinate (e x o-I)(88%) e n d o -2 -norb ornylsulfinate (e n d o -I) (9 5%) .

Preparation of alkylsulfinatomercuric chlorides. General procedure. A solu-tion of 1 equiv of the sodium alkylsulfinate in water was slowly added to asolution of mercuric chloride (2 equiv) in water. A white precipitate formedimmediately. The mixture was stirred at room temperature for 10 min, and theproduct was collected by suction fi l tration and dried. The following compoundswere prepared by th is method: n-hexylsul f inatomercur ic chlor ide (71%);IR1180 (s ) , 1060 cm-r (s ) ; cyc lohexy lsu l f ina tomercur ic ch lo r ide (597a) ; IR 1185(s), 1050 cm-r (s) ; endo-2-norbornylsul f inatomercur ic chlor ide (endo-YI) (50%);N M R ( p y r ) 0 . 8 - 2 . 4 ( 1 0 H , c o m p l e x ) , , 2 . 9 ( 1 I 1 , m u l t i p l e t ) ; I R 1 1 8 0 ( s ) , 1 0 3 0cm-' (s) ; exo-2-norbornylsul f inatomercur ic chlor ide (exo-Vl) (56%\; NMR (pyr)7 . 0 - 2 . 3 ( 1 0 H , c o m p l e x ) , 3 . 0 ( 1 H , m u l t i p l e t ) ; I R 1 1 7 0 ( s ) , 1 0 3 0 c m - t 1 s 1 ; 2 , , 2 -dimethylpropylsul f inatomercur ic chior ide (60%) ; NMR (DMSo-du) 1.1 (gH,s ing le t ) ,3 .0 (2H, s ing le t ) ; IR 1200 and 7 I75 (doub le t , s ) , 1060 and 1000 cm '

( doub let,, s) ; ery thro - 3,3 -dimethylbutyi- 1 -sulf inato mercuric chloride- !,,2 -d 2(v I I ) (627" ) ;NMR (pyr , 100 MHz, deuter ium-decoup led) 0 .?g (gH, s ing le t ) ,1 . 8 0 ( 1 H , d o u b l e t , J 7 2 . 8 H z ) , 2 . 9 6 ( 1 H , d o u b l e t ) ; I R 1 1 8 0 ( s ) , 1 0 b 0 c m - ' ( s ) .Cyclic alkylsulfinato mercurials did not melt but decolorized steadily from ca.150'C to brown-grey residues; acyci ic compounds melted with decomposi t ionand release of gas : n-hexyl, 140-\42" c; 2,2-drmethylpropyl, 1 2g-r30'c,erythro-3,3-dimethylbutyl-1 , ,2-dr, t42-743"C. Elemental analyses for thesematerials are l isted in Table 2. The analysis of the sample of endo-VII wasapproximately 2% low in carbon; this sample was used in further experimentswithout purif ication. Where necessary analytically pure samples of alkylsulfinato-mercurials could be obtained by dissolution in DMF and precipitation byaddition of water at 0'C.

General procedure for the Peters reaction. The alkyisulfinatomercuric chloride(1 mmol) was dissolved in DMF (10 ml) . The solut ion was purged with ni t rogenand the flask was placed into an oil bath heated to 100"C. Within a few minutes,the clear solution rapidly became milky-white. Heating was continued for thedesired length of t ime (most react ions were complete in 5-10 min), and themixture was fi l tered while hot through Celite to remove small amounts of in-soluble material and elemental mercury. The fi l trate was treated with 100 mlof water, and the crude alkyimercuric chloride was collected by fi l tration.Purif ied product was obtained after recrystall ization from ethanol/water (217).The yields of crude product were consistently 40--6 07o, whtle the yields ofpur i f ied product were 30-50%.

Infrored analyses. IR spectra were obtained in potassium bromide pellets(3%wlw). The IR spectrum of exo-Z-norbornylmercur ic chlor ide (exo-Y\shows character ist ic absorpt ion bands at 1100, 995, 949, and 86b cm-r; endo-2-norbornylmercur ic chlor ide (endo-V) has character ist ic bands at 1118, 959,890, and 754 cm-' (see also ref . 18). Cyclopentylmethylmercur ic chlor ide hascharacter ist ic IR absorpt ion bands at 1150,931, and 735 cm-r, and 5-hexenyl-mercuric chloride has characteristic band.s at 990 and 910 cm-r. The IR spectra

762

were recorded using either samples of crude or recrystall ized mercltrials isolated

from the Peters reaction as described in the previous section with indistinguish-

able results. Control experiments established that the isolation procedure did

not resul t i r r epimerizat ion of endo-V, and recrystal l izat ion of a 1/1 mixture of

exo-Y and endo -V as described in the previous sectior-r led to no detec:table

fractionation of the mixture.

Acknowledgments

A sample of t l t reo-3.3-dimethylhutan-1-o1-1 ,2-d, was kindly provided by Dr.

Paul L. Bock. Special thar-rks go to Dr. Bock and Mr. J im Simms for assistal lce

in obtaining 100 MHz deuter ium-decoupled proton NMR spectra. Tl-r is researcl- t

was supported by the National Science Foundation, Grant MPS 7 4-20946. The

arnard of a NATO Postdoctoral Fel lowship (1975-77 ) by the (Br i t ish) Science

Research Counci l to G.B. Young is grateful ly.

References

1 R e . , ' i e r v s : ( a ) l - . N { u e l l e r ( F l c l . ) . l l o u b e n - W e v l . N l e t h o d e n c l e r O r g a n i s c h e n C h e r l i e . 1 3 / 2 b . ( i e o r g

T h i e m e V e r l a g , S t u t t g a r t , 7 9 7 4 , 1 ' t . 1 7 - 2 7 6 ; ( b ) L . G . N t a k a r < ; r ' a a n d A . N . N e s m e v a n o r ' , N l e t h o d s o f

E l e m e n t o - O r g a n i < r C h e n . r i s t r v . V o t . 4 . N o t h - H o l l a n d P u b l . C o . , A m s t e r d a r n , 1 9 7 6 , p . l - 3 3 6 .

2 W . P e t e r s , B e r . , 3 8 ( 1 9 0 5 ) 2 5 6 7 ' , R e f . 1 a , p . 1 2 3 - 1 2 9 : R e f . 1 b . p . 2 5 4 - 2 5 7 .

3 N { . G i e l c n a n d R . F o s t v , B u l l . S o c . C h i m . R t ' l g . , 8 l J ( 1 9 7 ' 1 ) 3 3 3 .

4 R . C . L a r o c k a n d H . C . B r o r v n , . l . A m e r . C h e m . S o c . , 9 2 ( 1 9 7 O ) 2 4 6 7 .

5 R . C . L a r o c k a n d F { . C . B r o r v n , . I . O r g a n o r n e t a l . C h e m . , 2 6 ( 1 9 7 1 ) l l 5 ; R . C . L a r o c k , i b i c l . , 6 7 ( 1 9 7 4 )

3 5 3 : 7 2 ( 1 9 7 4 ) 3 5 .

6 D . S . M a t t e s o n a n d R . A . B o w i e . J . A m e r . C h e r n . S o c . , 8 7 ( 1 9 6 5 ) 2 5 8 1 .

7 l ' .R . . Iensen and B. R ickborn . E lec t r r tph i , l i c Subs t i tu t ion o f Organomcrcur ia ls . l \ {cGran ' -H i l l , Nerv

Y c r r k . 1 9 6 8 . p . 3 2 " 3 5 .

8 D . I . l . B e r g b r c i t e r a n c l G . N I . W h i t e s i c l e s . J . A m e r . C h c m . S o c . , 9 6 ( 1 9 7 4 ) 4 9 3 7 .

9 F l . W r . I l i s c h . ! 1 . G i p s t e i n a n c l O . . I . S r v e e t i n g , J . O r g . C h e n r . , 2 7 ( 1 9 6 2 ) 1 8 1 0 : W . E . T n ' c e a n d A . N { .

X { u r p h r ' . C h e r n . I t e r ' . . 4 8 ( 1 9 5 1 ) 6 9 .

1 0 P . A l l e n . J r . , , J . O r s . C h c n r . , 7 ( 1 9 4 2 ) 2 3 .

1 1 W . E . T r u c e a n c l F ' . ! , . R o b e r t s , J . O r g . C h t ' n r . , 2 U ( 1 9 6 3 ) 5 9 3 .

1 2 W . ( ; . ! ' i l b v . K . G u n t h e r a n c l I l . D . P e n z h o r n , , i . O r g . C h c m . , 3 8 ( 1 9 7 3 ) 4 O 7 O .

1 3 E . ! 1 . R e i c l , O r g a n i c C h e m i s t r v o f B i v a l e n t S u l f u r , V o l . 1 , C h e m i c a l P u b l i s h i n g C o . . N e r v Y o r k , 1 9 5 8 ,

p . i l O .

1 4 D . l . D a r . ' i e s . L . T . P a r f i t t . c . K . A l d e n a n d J . A . c l a i s s c , J . c h e m . S o c . c , ( 1 9 6 9 ) 1 5 8 5 .

I l - ) G . N I . W h i t e s i c l e s a n d D . , J . B o s c h e t t o . J . A m c r . C h e n r . S o c . , 9 1 ( 1 9 6 9 ) 4 3 1 3 : 9 3 ( 1 9 7 1 ) 1 5 2 9 ; P . L .

B c r c k , D . J . I l o s c h e t t o , J . R . R a s m u s s e n , J . P . D e m e r s a n c l G . N I . W h i t e s i d e s , i b i d . , 9 6 ( 1 9 7 4 ) 2 8 1 ' 1 :

P . L . t s 6 c k a p c l G . N { . \ \ ' h i t e s i d e s . i b i c l . , 9 6 ( 1 9 ? . 1 ) 2 8 2 6 t I . L . F r i t z . . l . I I . I ' i s p e n s o r . r , D . A . \ \ ' i l l i a r n s a n d

G . A . N ' l o l a n d e r . i b i d . , 9 6 ( 1 9 ? 4 ) 2 3 7 8 D . E . B e r g b r e i t e r a n d f ) . P . R a i n r i l l e , , I . O r g . C h e m . , ' + 1 ( 1 9 7 6 )

3 0 3 1 .

1 6 J . F ' . W . N 1 < : ( ) r n i e . i n A C ' , ' a n c e s i n O r g a n i c C h e m i s t r t , V o l . 3 , W i l e - v , N e u ' Y o r k , 1 9 6 3 , p . 2 5 2 .

1 ? l ' . R . J c n s e n a n c l K . L . N a k a r n a v e . J . A m e r . C h c m . S o c . , 8 8 ( 1 9 6 6 ) 3 4 3 7 .

1 8 G . 1 \ ' 1 . \ \ ' h i t c s i d e s a n d . l . S a n l ' i l i p p o . J . A r n e r . C h c r n . S o c . , 9 2 ( 1 9 7 O ) 6 6 1 1 .

1 9 , I . R . B n s [ . [ ' . ( i . C < l o k s o n a n c l ( i . B . D e a c o n , J . O r g a n o m e t a l . C h e m . , 3 4 ( 1 9 7 2 ) C 1 l C i . V i t z h u m a n d

F l . L i n d e r , A t r g t ' r v . C h c n r . I n t e r n a t . I t d . , 1 O ( 1 9 7 1 ) 3 1 5 .

2 0 C . P . C a s t ' r ' . I ' i r . I ) - T h e s i s . I \ ' l a s s a c h u s c t t s I n s t i t u t e o f T e c h n o l o g t ' , 1 9 6 7 . p . 2 O 2 - 2 O 3 .

2 1 \ \ ' . A . N u g e r r t , N l . \ \ ' u , T . P . I " t ' h l n e r a t r d . J . K . K o c l . r i , C l ' r e r n . C o m m u n . , ( 1 9 7 6 ) 4 5 6 .

2 2 G . N I . W h i t t ' s i c l e s . . J . P . S e v e n a i r a n c l R . \ \ ' . G o e t z . J . A n . r c r . C h e r n . S o c . , 8 9 ( 1 9 6 7 ) 1 1 3 5 .

2 3 G . D . S a r g c l t . i n G . A . ( ) l a h a n d P . r ' . R . S c h l e y e r ( F l c l s . ) , C a r b o n i u t . t r I o n s , V o l . I I l , W i l e l ' - I n t e r -

s e i c n c e . N t ' r v Y o r k . 1 9 7 1 . p . 1 0 9 9 f f .

2 4 P . D . B a r t l e t t , G . N . F ' i c k e s , F ' . C . H a u p t a n d R . I l e l g e s o n , A c c o u n t s C h e m . R e s . , 3 ( 1 9 7 0 ) 1 7 7 .

2 5 D . L a l . D . G r i l l e r . S . l l u s b a r . r c l a n c l K . U . I n g o l c l , J . A r n e r . C h e m . S o c . , 9 6 ( 1 9 7 4 ) 6 3 5 5 '

2 6 C . L . H i l l a n c i G . N l . \ \ r h i t e s i d c s , J . A m c r . C h c n r . S o c . , 9 6 ( 1 9 7 4 ) 8 7 O .

2 7 A . H o r o u , i t z a n c l L . A . R a j b e n b a c h , , l . A m e r . C h e m . S o c . , f ) 7 ( 1 9 7 5 ) i 0 -

1 6 3

2 8 D . I ' - . S h r i v c r , T h e N l a n i p u l a t i o r L o f A i r - s c n s i t i v t ' C o n ' r p o r - r n c l s . \ l c G r a r v - H i l l . N e r i ' Y o r l i . 1 9 t j 9 : I { . C .Brorvn , G. \V . Kramcr , A .B. Levy ar . rc i 1 .1 . \1 . I l i d land. (J rgan ic Svnthes is r . ' i a Boranes . \ \ ' i l e r ' . Ncrv Yor }< '1 9 7 5 .

2 9 , 1 . D . I l . o b t - ' r t s . F l . R . ' f n r m b u l l , J r . , N . B e n n e t t a n d R . , , \ r m s t r o n g , J . A n r c r . C h e n r . S o c . . 7 2 ( 1 f , l 1 ( t )3 i 1 6 -

3 0 G . G r i . i t t n e r , C h e m . l l e r . , . 1 7 ( 1 9 1 . 1 ) 1 6 5 1 .: 1 1 Y r . r . A . O l ' d t , k o p , N . A . N I a i e r . \ ' u . I ) . I t r , r t ' k o a n d i l { . S . l \ l i n d e l . Z h . O b s c h . K h i m . . . l 1 f i 9 Z l ) l 0 6 ( j .: 1 2 I " . C . W h i t n r o r e . F l . L . W i t t l e a n c i B . R . I l a r r i r n a n . . J . A m e r . C t h e m . S o c . . 6 l ( 1 g 3 9 ) l b 8 i r .: 1 3 C . t l l r i c h . A n r . r . . l O g ( 1 8 5 9 ) 2 7 2 .