8/3/2019 4.3 f

1/6

s o v 3. 19t i4 KE,\CTIVITE' O F HALIDES I N s N 2 REACTIONS 4645O( to sulfur. rather than 3 to sulfur (where it would becloser to th e halogen a to m) .The trends in chemical shifts noted for 14 and 15 arecontinued in the sp ectrum of th e dibromo derivative16. The CHB r2 singlet is a t 1 . 3 7 p.p.m. and thequartet for th e single ring hydrogen atom is at abo ut; j , 7 p.p.rn. Peak B is shifted to 5 . 3 3 , whereas peaks Cand D remain essentially fixed [doublet centered a t5 , S S (I = 7 . 0 c .P . s . ) ] . Peaks E and F are now re-solved. as in the parent sultone 13, but are shifteddownfield to 5.SS nd 3.99 p.p.m. They are separatedby a chemical shift (6 = 0.11 p.p. m.) comparable inmagnitude to th at observed in 13 ( 6 = 0 .07 p . p . m . ) .Sultone 17 was prepared by th e sulfonation of 1-t-butylcyclohexene? Th e structure assigned on th ebasis o f the expected course of is th at shown.This assignment is supported by the p.m.r . spectrum,

which can be compared with tha t of 13. Structurally,17 differs from 13 by the fact that the cyclohexanering has replaced one methyl group on each of th ecarbon atoms cy an d p to oxygen. The structuralfeatures held in common (gem-dimethyl group a tooxygen. methyl group /3 to oxygen, and single hydrogenatom y to oxygen) give rise to closely similar p.m.r.signals (Table I ) . The chemical shifts observed i nchanging from 17 to 18 should be, and are, comparableto those in changing from 13 to 15.

Acknowledgment.-This work was generously sup-ported by the Office of Ordnance Research under O O RGr an t D,1-*4RO(D)-31-12 4-G19. IVe wish to than kDr. K. It-.Bartz, Hu mble Oil and Refining C o. , Bay-town, Texas, for checking some of the spec tra arid forseveral helpful discussions.

I C U S T R I B L T T I O S F R O M THE I)EPARTMENT OF CHEMISTRY, SORTHII'ESTERS C'NIVERSITY, EVANSTOS,LLISOIS]The Effect of the Carbonyl and Related Groupson the Reactivity of Halides in S N ~eactions

BY F. G. BORDWELLND \y. T. BRANNEN, JR .RECEIVEDUS E 25 , 1961

Th e rates of reaction of potas sium iodide in acetone solution with several series of co mpoun ds of th e typeYiCH?),,CI n = 1 to *5 ) have been measured. Th e strong deactivating effects of F,C and C6H,SOn groupswhen attached to the a-carbon atom are attributed to steric and field effects, their inductive effects being as -sumed to be tnildly rate enhancing. Seit herthese groups nor related group s, such as CaH,S, C6HjC0, or C S , exhibit much effect of a ny kind from 0 r moreremote positions The mild activating effect of the B-CsHjCO group is the one euccption. Th e rnethailolysisrates for CFH,S(CHB),~CIn = 1 to 5 ) har e been measured.

The CaHjSO group was found to have a mild retarding effect.

Conant, Kirner, and Hussey first called attentionto the profound influence of Y groups on the reactivityof \-CH2Cl type halides toward potassium iodide inacetone. ' Th e powerful activati ng effect of th e car-bonyl, cyano, and related groups in this and otherS N ~eactions has since been the subject of muchst ud y arid speculation.? Our interest in this effect wasfirst aroused several years ago by the observation thatthe activating effect of the carbonyl group, not onlyis not shared , but is actually reversed, by the sulfonylgroup. This is surprising since the carbonyl and sul-fonyl groups exert the same ty pe of inductive and reso-nance effects as judged by their con st ants.^ I t wassuggested tha t th e retar ding effect of sulfonyl group wasof steric origin.'e" ly e decided to extend the s tudy t othe sulfinyl and trifluoromethyl groups, since thesegroups exhibit closely similar inductive and resonance

( 1 , J. B Conan t , \V. I i . Kirner, an d K . E. Hussey. J . Am . C h r m . SOL., 7,18 8 ( 1 9 2 , j J .( 2 ) la ) J. W . aker , J . ( ' h e i n . So c , 1148 (1932), (b ) E. D. Hughes,

l ' r i z n s . Fnr . uday SOC. .37, 0.1 (1911D (C J J. W. Bakei-, 5 2 L 37, 49 (1941);id ) &I . J S Dewar. "The Electronic Theory of Organic Chemistry." Ox-f o l d L-ni\-enity Press London, 1!44

8/3/2019 4.3 f

2/6

4646 F. G. BORIIWELLND W . T. B R A N N E N ,R. lro1. 86runs were made at nearly the same potassium iodideconcentration (near 0.04 M ) n order to minimize effectscaused by differing reactiviti es of ion-pair aggregates.A s a prelude to th e studies of t he S N ~eactions,the relat ive rat es of methanolysis of t he chloro sulfides,Cd&S(CH2),C1, were dete rmined ( n = 1 to 5 ) . Al-though measurements of th e rates for one or moremembers in this series have been made in severaldifferent solvent^,^ the complete series does not seemto have been run previously in a single medium.The relative rates for methanolysis were found to be:C&SCH2C1, 3.3 X lo4; C&S(CH2)2C1, 1.5 X l o 2 ;CfiHsS(CH2)&1, 4.3. Since CG H $ ( C H ~ ) ~ C ~as beenfound to solvolyze in aqueous dioxane a t a ra te roughlycomparable to tha t of n-hexyl ~h lo ri de ,~ 'he presentresults indicate that the a-bond conjugative effect ofthe a-CgH.23 group causes an approximate 33,000-fold increase in methanolysis rate, and that partici-pation of the sulfur atom leading to u-bond formationcauses abo ut a 100-fold increase in rate when a three-or five-membered ring is formed (B - or 6-C6HsSgroup).The effect of the t - C & , S group is more modest, butis still significant. I t was anticipated that thesevalues for participation might be useful for comparisonwi th th e effect s of CsHbS, C,&SO, C6HbS02, and FaCgroups on S N ~eactivity in Y(CH2),Cl type series,since the reported ac tiva ting effects of C&f$O groupsattached to @- and y-carbon atoms also seemed likelyto have their basis in participation. At least, the p-an d r-C&SO groups were expected to show activatingeffects. Actually, as i t turned out, no appreciableeffect on s N 2 reac tivi ty was observed for any of thesegroups beyond the a-position. Th e CfiHbS.CfiHsS02,and F3C groups were mildly deac tivating from @-positions ; th e CfiH5S0 group was mildly acti vating(Table I ). All of the groups, except F3C,were mildlyactivating from more remote positions. I t seemedadvisab le, in view of these results, to check t he originalreported carbonyl activation . Repetition of the earlierwork failed to show any appreciable activating effectfor the y-C&CO group. Th e supposed activatingeffect also fails to materialize for y-CH3C0, y C N ,or r-EtOzC groups. @- C& ,C O nd p-CH3C0 groupswere found to have activating effects, although not aslarge as expected. The relative rate data are summa-rized in Tab le I .

Since attac hmen t of a n alkyl substituent at eitheran a-carbon atom or a 6-carbon atom is known toretard S N ~eactions, the assumption has been madethat an electron-releasing inductive effect is rate re-tarding, and that inductive electron withdrawal is rateaccelerating.2b On the other hand, i n view of themild deactivati ng effect produced by attaching a halo-gen, IiO, or R S grouping a t a @-carbon tom , inductiveelectron withdrawal has been assumed to be rateretarding. 2h,"' Th e latter conclusion is weakened bythe fact that C6H5C0,CH3C0. C&SO , Et02C. andCN groups are mildly rate enhancing when attachedto a &carbon atom (Table I) , despite their consider-ably larger electron-attracting inductive effects, ascompared to X, R O . or RS groups6 Furthermore,

C ~ H ~ S ( C H Z ) ~ C I ,.0; CsHsS(CHz)4C1, 1.3 X IO2;

:9 ) ( a ) G. 31. Bennet t and A I-. H o c k , J . C h r m S o c , -177 (1927) ; (b )H . Bh h me a n d K . Sel l , Be i . , 81 , 12 3 ( 1948 ) , (c ) F . C Bo r d se l l , G. D .Coopel . and H Xlol-ita, J An i . C h e m S n c . 79 , :376 ( 1 9Z7 )

1101 Se e t h e discussion on p 1 5 of ref. 2g

TABLERELATIVEATESOF R E A C T I O N IT H POTASSIUMODIDE

IS ~ C E T O N E T 75"--Sei-ies

CH3CHnCH1CH2CIUCsHsS( CH2!,,CICsHsSO( CHZ),,C1CsHsSOz( CHz),,ClFjC(CHa),,OTshFIG( CHZ ,LBrcCH1CO(CH*),,ClCeHjCO( CHs),,CIN-C( CH*),CIEtOnC( CHZ),C1

~-Values fo r ,,------.- ~ _ ~1 2 x 4 51 0 1 . 0 1 0 1 0 1 . 0

540 0 .79 3 . 1 1 8 1 10 2 5 2 . 7 2 6 3 3 2 6< 02 0 . 3 8 3 . 6 3 8 2 7

.00016 20 . . . . . . . .32,000 44 0 2 1 1 . 7 , . ,3,O0Od 2 8 3 7 2 . 8 . .1,700d 1 6' 1 6' 1 . 4 e , ,

00007 .2 2 0 79 . . .~35,000" 7 0 1 . 4 . . . . .

The rates determined by amperometric titration were0.244 + 0.018 . Jf hr. -l a t 00" and 0.783 + 0,029 1. M- ' hr . la t 75'. These rates are about 2.5-fold faster than tha t reportedby Conant, et nl. ' Relative to the rate of PrOTs at 40'.Relative to the rate of PrBr a t 35'; dat a of E. T. McRee,I< . D. Battershell, and H. . Braendlin, J . .4wz. Chenz. S o r . , 84,3157 (1962) Dat a of J . R . Conant, W. K . Kirner, and K. .Hussey, ref. 1 Dat a of J . B. Conant and W. K.Kirner, ih id . ,46, 232 (1924).electron-withdrawing substituent s in ,4rCOCH2Cl andArSOCH2C1 are rate enhancing (Tables 111 an d V ) . I 1Finally, the reactivity of UCH=CHCH2Cl compoundstoward potassium iodide in acetone has been observedto be mildly enhanced (10- to 100-fold) when U isC N , 2 e CfiHjS02,2er F3C.'? These data suggest thatthe r-inductiv e effectI3 (and by inference the u-in-ductive effect) of these groups is rat e enhancing. Th efact that the C6H6S02nd F3C groups exhibit an effectcomparable to th at of CN when operating through theu-bonds and a-bonds of a Y-CH=CH-CH2Cl system,bu t show an opposite effect in a 17-CH2C1 system, sug-gests tha t proximity effects other t han inductive effectsare causing the large rate retardations and rate enhance-ments observed in the latter system. Our view isthat the inductive effect for electron-withdrawinggroups is probably mildly rate enhancing, but that itis completely overshadowed in a lmost eve ry Y-CH2Clsystem by th e presence of large steric effects, fieldeffects, an dl or conjugative effects.Ingold14 has classified th e str uctu ral kinetic effectsoperating in simple S N ~ubstitution reactions intothree groups : (1) polar effects, depending on charge;(2 ) steric effects, depending on bulk: and (3) ponderaleffects, depending on mass. When the atom attach edto the a-carbon atom is unsaturated, or otherwisecapable of entering into conjugation, we must add afour th effect-(4) a conjugative effect, depending oneffective orbi tal overlap. Of these facto rs, stericand polar effects appear to be most likely to be re-sponsible for the large retarding effects of FBC andCEHsS02groups. Since we have assumed th at th einductive effe ct plays a relatively minor role, which isprobably rate enhancing, this suggests that a sizablerate-retarding field effect is combining with a steric

(11) An activating effect of electron~w ithdrawing dbstituents has alsobeen observed in the reaction of ArCOCH2Bt- with pyridine in acetonesolution [ J . W . Baker-. J . Chem. S O C . , 44 5 (1948)l, fo r the reaction of Ar-COCH?Br wi th 2.6-dimethyl-4-thiopyrone in benzene solution [F . J . Ozog,V , omte, and L. C King , J Am . Chem. S oc , 74 , 622: ( 1 9 5 2 ) ] , and fo r theIreactions of ArCHK HzCI, AI-SCH?CH?Cl, and ArSO?L'HzZHsCI w i t hpotassium iodide in acetone (ref. %g , p( 1 2 ) J . .4. Pegolotti and W. G. Young. J . A m Chrnz . So c . . 8 3 , 3258(1961)

( 1 3 ) R f . J S . Dewar and P . J . GI-isdale, rbid , 84 , 3549 (19623(141 C. K . Ingold, @ua i , f . Rei. (London) , 11, 1 ( lQ 57) .

17: .

8/3/2019 4.3 f

3/6

Nov. 5, 1964 REACTIVITYF HALIDESN sN2 REACTIONS 4647effect to give the F3C and CsHjS02groups the ir powerfuldeactivating effects. It is no doubt significant th at th emost probable p ath for the incoming nucleophile to theF3CC HzOT s and C6HjSOzCHzC1 molecules is en-cumbered by a n atom of high electron density, insharp contrast to the situation with the C6H5COCH~C1molecule (compareX r B with C) .

OTs c1

I1- 1-A B

1-C

Fur thermore, if we consider possible transit ion sta tesfor the reactions (A') ', and C') we see that the re-pulsions between the entering and leaving groups withth e highly electronegative atoms on the F3C and CeH5-SO2 groups are probably large, certainly much largerthan for the CfiHjCO group. (These repulsions are,of course, balanced to some extent b y the positivecharge on t he P-atom.)

A' B' C'The favorable effect that the carbonyl and related(ch', Co 2R , etc.) groups have on sN2 reactivity

probably stems in part from the absence of r ate-retarding ste ric and field effects, coupled with a mildlyrate-enhancing inductive effect. The very strongaccelerating effect of t hese groups (compared toBuC1) must , however, include some sort of an addi -tional conjugative factor, as first suggested by Dewar.2dThis view is strongly supported by the observation ofBartlett and TrachtenbergZh ha t in a system incapa-ble of achieving transition sta te C ' the enhancing effectof the carbonyl group is lost.'j

Th e natur e and precise description of t he conjugativeeffect is still open to question. One possibility is toutilize the electrons in the parti ally formed I-C andpartial ly broken C-Cl bonds in a a-bond overlap withthe p-o rbital of t he carbon atom of the carbonyl groupacting as an electron pair acceptor (D or D'). Thisrequires that bond formation leads over bond breaking

c1 Cl'+mC-C-0 OR C=C-O-mI 16+

D D'in the transition sta te. Besides the unattractiveness ofstructures such as D', there are two arguments againstthis representation. Firs t, there is good reason tobelieve that the activating effect of C=C, C=C, RO,

( 1 5 ) Approach of the reagent as in C leading to transition stat e C' no tonly offers h e best opportunity for conjugation, but also appears to be bestfrom an electrostatic view.*' I t would appear, then , tha t electrostaticeffects and conjugative effects act in consort to enhance S N ~eactivi ty , andtha t the re is no reason to separate the two.

and ArS groups in S N ~eactions (note, for example,the 540-fold acceleration listed in Table I for the CY-C&jS group) is caused by a conjuga tive effect of es-sentially the opposite type, where these groups act aselectron-pair donors in a a-bond overlap (E an d E').'&

CGHSS-C +-+sH6+$=-=CC16- c1-

i s- 1-E E'This view is strengthened by the observation in thepresent work that electron-withdrawing groups havemild retarding effects on the rate for ArSCHzCl (TableV I ) . I 7 While it is conceivable tha t th e transition sta temight have carbanion character in one type of S N ~reaction and carbonium ion character in another, itwould be preferable to avoid this assumption.A second argument against representations such asD and D' is that one would expect on this basis thatCsHjSO2, C6H jS0 , and F3C groups would also act ivat e,since they too are capable of acting as electron accepto rsin conjugative type interaction^.^,^ Their activatingeffects would not be expected to be as large in magni-tude, but should at least be in the same direction.This is clearly not the case (Table I ) .I t is possible to get around these difficulties by a dop t-ing the suggestion of Streitwieserlsb th a t the conjuga -tion responsible for carbonyl activa tion is caused by anoverlap of the p-orbital of th e carbon atom of t he car-bonyl group with orbitals on the entering and leavinggroups (as in F' and F"), rather than a a-bond typeoverlap.

cis- c1 c1\c-c=o ++ c-c-0- ++ c-c-o-i s - I IIF F' F"

/

By adopting this view, the reasonable assumptioncan be made t ha t t he transition s tates for reactions ofall types of YCHzCl compounds with iodide ion inacetone are similar in natu re. Activation by CY-C=C,-RO, -ArS, and like groups can be att ributed primarilyto their ability to provide electrons for a a-bond over-lap. Activation by a-carbonyl and like groups can beattributed primarily to their ability to provide anelectron-deficient p -orbita l with which electrons on th eentering and leaving groups can inte ract . Deactiva-tion by a-CeHjS02, -F3C, and like groups can be at-tributed to steric and field repulsions between t he highlyelectronegative atoms on these groups and the enteringand leaving groups.

The energy and entropy activation parameters forthe S N ~eactions discussed above are summarized inTable 11.Most of th e rate measurements are accurate t owithin f %, as judged by probable error. This meansthat the Ea values are good to roughly i l cal.Jmoleand the S* values to roughly =k3 e.u. Deviations ashigh as 1 2 kcal./mole or 6 e.u. are possible, however.

(16) (a ) P. Ballinger, P. B. D. de la Mare, G. K ohns tam , and B . M J .Presst, J . C h e m . SOL. , 611 (19.55); (b ) ref. Zg, pp . 2 6 - 2 8 .(17) Xote, however , that t he p-NO3 gt-oup has only a 2-fold deacti\,atingeffectas compared to about a 100-fold deactivat ing effect n t he methanolysisof ArSCHZCl (Table VII ) . This indicates that carbonium io n characterin the Sh-2 reaction is probably not very highly developed.

8/3/2019 4.3 f

4/6

4648 F. G . BORDWELL N D W . T. B R A N N E N ,R. Vol. 86TABLE11

KISETIC DATA OR TH E REACTIONF ARYLCHLOROMETHYLKETONES IT H POTASSIUMODIDEN ACETONET 0'

Substituent Rel. rates E, , kcal., mole S *,e .u .p-?;0, 10 .8 15 - 2p-Br 2.64 17 - 6p-c1 2 .44 16 - 9P-CHaO 1 21 I5 - 11P-H 1 . 0 17 - 8

TABLEVARYLKETONES IT H POTASSIUMODIDEX A C E T O N ESOLUTIONT 60"

KINETICDATA FO R TH E REACTIONF P-CHLOROETHYL

Substituent Kel. rates E,, kcal. mole S* . e uP- H I O 19" - 4p-Br 0 .89 17 " -21p-CHaS02 19 22 - 0p-c1 .10 20" - 8P-CHaO 065 18" - 3P-CHaS 042 19 - 3

Activation energies were calculated by multiplying theslopes of th e lines from log k us , 1/ T plots by 4.58.TABLE-

KINETICDATA OR THE REACTIONF ;\RYL CHLOROMETHYLSULFOXIDESIT H POTASSIUXODIDEN A C E T O N EAT 75'P- H 1 . 0 19 - 4p-Br 1 9 22 - 7p-h-02 3 . 2 2 0 - 0

Subst i tuen t Rel . ra tes E, , kcal. 'mole . S * , e.u.

TABLEIKINETIC ATA OR THE REACTIONF CHLOROMETHYLARYLSULFIDESITH POTASSIUMODIDEN XCETOSE

E,, kcalCompound T , OC. ki, 1. .Vi-' hr . - l mole S* , e. u

CsHjSCH2CI 25 4 .93f 27p-BrC6H,SCH2C1 25 4 . 6 7 i. .35

35 13 5 f . 9 18 -1235 12 .1 i 8 17 - 635 5 30 i. 27 16 -21P-SOrCsHaSCH2Cl 25 2 18+ 06

TABLEI1KISETICDATA OR TH E METHASOLYSISF CHLOROALKYL

ARYLSULFIDESE,, kcal.1

Compound T . ' C . k , 1. h-1 r , - ' mole S' , e. uc6H5sCI~2c1 25 8.64 xp-BrC6H5SCH2C1 25 1 .5 7 Xp-S02C6H4SCHpC1 25 8 72 X

35 2 .55 x 10-2 20 -1235 4 53 x 10-3 19 -1735 2 .85 x 22 -15

GHjSCHzCH2CI 40 2 .51 x 10-450 6 12 x 18 -30Experimental18

Kinetic Procedure .-Reagent grade potassium iodate and silvernitrate were dried and used as such. Commercial acetone wasdried for several days over a large quan ti ty of calcium chlorideor Drierite; the mixture was then filtered and distilled.Tubes containing suitable quantities of the organic halide werethermostated at temperatures from -16 to 40" for 15 min., andto these were then added similarly thermostated acetone solu-tions of st and ard potassium iodide. The time of mixing w a staken as zero time. -4t temperatures above 40" the halide andiodide solutions were mixed at room temperature in tubes andcooled to Dry Ice temperature. The tubes were removed for

TABLE1ACTIVATION PARAMETERS FO R S N 2 REACTIONSIT H

POTASSIUMODIDEN ACETONECompound E, , kcal./mole S* , e u

181720"181819171618201916201618181718181521"2219"20"182220

- 6- 8- 4- 0-21- 4- 8- 9- 2- 9- 4- 0- 6- 9- 2- 2- 9- 4- 6- 3- 6- 5- 5- 8- 2- 7- 8

a The activation energy was calculated by multiplying theRateE. T. McBee, R. C . Battershell,slope of t he line from t he log k v s . 1/ T plot by 4.58.relative to BuCl a t 75 " is 3.8.and H. P. raendlin, J . Am. Chem. Soc . , 84 , 3157 (1962).

N o significance can be at tached to small differencesin Ea and S*, therefore. I t is significant, however, tha tthe large rate enhancement caused by the (Y-C&COgroup is accompanied by a very sharp increase in acti-vation entropy (see also Table 111) when compared ton-bu tyl chloride, CeHsSOCHzC1, or chlorides whereC&fj CO, C6HjS0, or CeH6SO2 are in more remotepositions (Table 11). On the other hand, the largerate retardation caused by the a-F 3C group is ac-companied by an equally sharp decrease in activationentropy in the tosylate series, and a lesser, but stillsignificant, decrease in the bromide series (Table 11).Although it is difficult to predict the manner in whichactivation parameters will change with changingstructure in a n S N ~eaction,2g t is possible to accom-modate a n increase in activation entropy in terms oflowered steric requirements and a decrease in activa-tion ent ropy in te rms of increased steric restrictionin the transition state.'g'I4I t is of interes t tha t substi tution of an electron-withdrawing group at a position more remote than (Ygenerally has but little effect on either the s N 2 rate(Table I) or the activation parameters (Table 11).The effects appear to be too small to warrant muchspeculation, except perhaps for the mild activatingeffect of t he P-C6Hj C0group. I t is tempting to postu-late some kind of conjugative effect here too. I t wouldappear t o be different in type from th at of the(Y-CCH~COroup, however, since electron-withdrawinggroups cause rate retardation rather than rate accelera-tion (compare Tables 111and I V ) .

(18)Microanalyses were by Xiss H. Beck, Melting points and boilingpoints are uncorrected. We wish to thank Prof. R C Bowers for his as -sistance with the analytical pi-ocedures.

8/3/2019 4.3 f

5/6

xov . 5 , 1963 R E A C T I V I T Y O F H A L I D E SN s N 2 R E A C T I O N S 4649sealing and placed immediately in the constant temperaturebat h. Initial time readings were recorded after 10 t o 20 min. inth e 50 o 100' ra nge; blanks showed th at less than 3T0 reactionhad occurred durin g the time of mixing.Runs were made using a fivefold, or greater, r atio of halide toiodide.' Kate constants were calculated using the equation

2.303 . w - 2

where w = a / b = [ R C l ] / [ K I ] , = x, ' [KI] , and x = amountof I- reacting in time t .The reactions gave good second-order constants in the rangeof concentrations 0.05 o 0.25 Af in halide.The amount of unreacted iodide ion was determined usingeither (or both) an amperometric or a potentiometric titration.The amperometric titrations were performed, using a rotatingplatinum electrode and a coiled silver wire as the nonpolarizedelectrode, at a con stant potential of 0.2 ., in a 0.01 dlammoni umhydroxide sol~t ion. '~The potentiometric titrations were made using a PhotovoltpH meter as the potentiometer, a coiled platinum wire as theindicating electrode, and a saturated calomel electrode as thereference electrode. Th e solution was placed in a beaker (helda t 0 " ) containing 25 ml . of wate r, 25 ml . of concentrated hydro-chloric acid, and 25 g. of crushed ice and titra ted with stan dar d0.005 A potassium iodate.

'KIOI+ 2KI + 6HCl+ IC1 + 3HzO8+ 3KClThe amperometric titration was preferred since it can be car-ried out more rapidly (,3to 5 2's. 10 o 15 min. ). It is unsuitable,however, for reactive halides. The CY- and p-chloro ketones,or-chloro nitrile s, and the 8 -phen ylsulf onyle thyl chloride were,therefore, titr ated potentiometrically. Both titrations wereused for 6-trifluoromethylethyl tosylate, P-phenylsulfinylethylchloride, 7-phenylsulfonylpropyl chloride, and yph enylsulfi nyl-propyl chloride.The relative rates and activation parameters are summarizedin Tables I-Y.Since halomethyl phenyl sulfides hydrolyze rapidly, th e acetonesolutions for these were poured directly into a separator y funnelcontaining 50 ml . of water and 100 ml. of chloro form. Th echloroform layer was removed and subjected to one rapid wash( < 2 min .). The combined aqueous layers were then treatedwith 6 ml. of concentrated ammonia a nd ti trated amperometri-cally. The rat e data (uncorrected for solvolysis) are summarizedin Table 1.1 (see also Tab le I ) . Blank runs indicate th at theserates should be increased by a bou t 30yc to correct for solvolysis.The solvolysis rates for the chloroalkyl aryl sulfides werecarried out as described previouslyQc sing methanol as a solvent.The kinetic data ar e summarized in Table VI I.Activation energies for reactions run a t only two temperatu reswere calculated from the equati on

The agreement was good (within 5yc).

Activation entropies were calculated using t he equation-S * = 4.606[9.681 + log T - og 12 -

(E , - RT)/2.303RT]Chloroalkyl phenyl sulfones were prepared by methods re-ported in the literature.20 Phenyl 5-chloropentyl sulfone waspurified by chromatog raphy on silica gel, eluting with 5yc ether-hexane; n Z o D1.5360.An a l . Calcd . for CI ~H~ ~C IO ZS :, 53.53; H , 6.14. Found:C, 53.70; H , 6.45.Chloroalkyl Phenyl Sulfoxides.-The chloro sulfoxides wereprepared by passing a stream of ozone through a solution con-taining 0.05 mole of chloro sulfide in 175 ml . of dry methylenechloride a t - 78 ' . When a blue color developed in the solutionthe ozonation was stopped an d t he solution was allowed to warmto room temperatur e. The solvent was removed under reducedpressure, and the residual oil chromatographed on a 2 X ,50 cm .silica gel colu mn. Elution with lOYc ther-hexane gave starting(19) H . A . Lai t inen , W. P. Jennings, and T. D. Parkes, I n d Eng Chem ,

A n a l . Ed. , 13, 355 (1946).( 2 0 ) Details as to t he methods used and th e physical constants of th e com-

pounds pi-epared are contained in t he P h.D . Dissel-tation of W T . Brannen ,J r , submitted to Sol-thwestern University, August, 1962.

chloro sulfide in th e first liter and tra ces of chloro sulfone in thenext 1.5 . On increasing th e ether content (u p to a maximum of20%;) small amou nts of unidentified oils appeared in the first 3 1 .of el uent , followed by t he sulfoxide (str ong infrar ed maximumnear 1050 ern.-' and no sulfone ban ds) . The yields and proper-ties of the chloro sulfoxides are summarized in T able 1'111.T A B L E I11PROPERTIESXD AKAL YSE S F W-CHLOROALKYLHENYLSULFOXIDES

,-Calcd., %- -Found, 70-Compound To pi-operties bon gen bon genYield, Physical Car - Hydro- Car- Hydro-

CsHsSOCHsCI 74 h1.p. 28-29' 4 8 . 1 3 4 . 0 4 4 7 . 9 7 3 . 9 5C6HaSO(CH2)2Cl 67 M . p . 31-32' 50.91 4 .81 5 1 33 4 . 6 8CsHsSO(CH%'zCI 71 11"D 1 .5710 5 3 . 31 5 . 4 8 5 2 . 7 9 5 . 2 6CaHaSO(CHn)nCI 55 nZGD1 . X 00 5 5 . 4 1 6 .0 6 35.88 5 . 9 8CaHsSO(CH2)jCI ;18 )zZoo . 5 3 7 8 5 7 . 2 4 6 . 56 56 .4Y 7 .10

a,a,a-Trifl uoroethyl p-Toluenesu1fonate.-To a solution of 5g. (0.05 mole) of trifluoroethanol in 70 ml. of dry benzene wasadded 1.2 . (0.05 .-a tom) of freshly cut sodium metal. Afterthe initial reaction had subsided, the resulting mixture was re-fluxed until all the sodium had disappeared. The solution wascooled and 9.5 g. (0.05mole) of p-toluenesulfonyl chloride wasadded arid the mixture refluxed for 12 hr. The reaction mixturewas hen filtered free of sodium chloride, the solvent w a s removed,and the residue was recrystallized several times from hexane toyield 7.2 g. ( 5 6 4% ) f a white crystalline solid melting at 39-41'(li t .21m.p. 41').p,p,p-Trifluoropropyl p-To1uenesulfonate.--A solution of 20g. (0.089mole) of 3,3,3-trifluoropropyl iodide2?an d 30 g. (0.11mole) of silver p-toluenesulfonate in 150 ml. of dr y aceton itrilewas refluxed for 48 hr. The mixture was filtered and t he filtratepoured into 300 ml. of water. Eth er extraction and vacuumevaporation gave a residual oil which was vacuum distilled togive 15.1 g. (6 3% ) of a colorless liquid, b. p. 120-122' (1 mm.),

An a l . Calcd. for C10HllF303S: C , 44.77; H , 4.14. Found:C, 45.56; H , 4.11.-,,y,r-Trifluorobutyl p-Toluenesu1fonate.-To a solution of 4g. (0.03mule) of 4,4,4-trifluorob~tanoI*~n 75 ml . of dry benzenewas added, in one portion, 1.4 g. (0.03mole) of sodium hydride(50% in mineral nil) and, subsequently, 6 g. (0.031 mole) of p-toluenesulfonyl chloride. Th e solut ion was refluxed for 2.5 hr. ,filtered free of sodium chl oride, and th e solvent removed und erreduced pressure. Th e residual oil was crystallized from ether-hexane, giving 5.1 g. (6Oyc)f a whi te solid melting at 15-16",z Z k 1.4690.A d . Calcd. for CllH13F303S: C, 46.80; H , 4.65. Found:C, 46.38; H, 4.59.p-Chloroethyl Meth yl Ketone .--Using th e proce dure of Blaiseand but substitu ting cadmium chloride for zinc chloride,gave 46% of a colorless liquid , b. p. 57-60' (15 mm.) (

8/3/2019 4.3 f

6/6

4630 F. G. BORDWELLND W . T. BRANNEN,R. L'ol. M jTABLES

Starting chloride time, hr. Iodo compound h l .p , oc. Calcd. Found Calcd. FoundCarbon , 7c--- ---Hydroyen, L;----eflux

p-SOyC,HjSOCH*Cl 18 C7HeISOaS 158-159 27 02 27 41 1 95 1 92CsH,SOCH>CHzCI 36 CsHgIOSC6H2SO,CH?CHiC1 141 CaHgI02-SC&H,SO?( Hz )iCI 36 CioHirI O3

chloride. Th e mixtur e was refluxed 1 hr. and poured cautiouslyinto 1 kg. of ice and 150 ml. of c oncent rated hydrochloric aci d.This mixt ure was extrac ted with two 100-m1. portin ns of e the r,dried over anhydrous sodium sulfate, and the ether removedunder reduced pressure. Th e residual oil was crystallized fromether-hexane, giving 39.5 g. (8OC;) of a white sulid melting atAnal. Calcd. for CSHsClBrO: C, 43.66; H , 3.26. Found:C, 43.93; H, 3.18.p-Chloroethyl p-Meth ylthio pheny l Ketone.-To a suspensioncomposed of 12.4 g. (0 .1 mole) of methyl phenyl sulfide and 30g. (0 .23 mole) of anhydrou s aluminum chloride in 100 m l . of s ym -tetrachloroethane was added with stirring 12.7 g. (0.1 mole) of6-chloropropionyl chlori de. Upon completion of the addition

the mixture was heated on a steam bath for 1 hr. and pouredcautio usly onto 1 kg. of ice an d 175 ml . of c oncent rated hydr o-chloric acid. The organic layer was separated and the aqueouslayer extracted with three 100-ml. portio ns of ethe r. Th eorganic layer and ethereal solutions were combined and driedover anhydrou s sodium sulfate. The solvent was removed underreduced pressure (t he pot tempe rature mas kept below 40' toprev ent decomposition of pro du ct ). Th e residual oil solidified:several crystallizations from acetic acid gave 16.2 g. ( 7 5 5 4 ) of ared crystalline solid, m.p. 111-113'.Ana l . Calcd. for CloHllClSO: C, 55.93; H, 5.17, Fou nd:C, 55.64; H , 4.61.6-Chloroethyl p-Methy lsulfo nylph enyl Keto ne -To 250 g.(1.3 moles) of 40yc peracetic acid at 0" was added with rapidstirring 36 g. (0.17 mole) of 0-chloroethyl p-methylthiophenylketone. The solution was stirred at 0" for 0.5 hr. and at roomtemperature for 1.5 hr . after which it was poured into 600 ml.of water and extracted with three 100-ml. portions of ether. Theethereal solutions were combined and dried over anhydroussodium sulfate. The ether was removed under reduced pressureand, th e oil crystallized from ether-hexane, yielding 14.6 g.( 3 5 5 ) of a white c rystal line solid melting a t 85-87".

Ana l . Calc d. for CloH11C10,S: C, 48.67; H , 4.50. Fo un d:C, 49.09; H, 4.47.a-Chloroethyl p-Methoxyphenyl Ketone.-To a suspensionmade up of 54 g. (0 .5 mole) of anisole and 150 g. (1. 1 moles)of a nhyd rou s aluminum chloride in 200 ml. of d ry carb on disulfidewas added with stirring 63.5 g. (0.5 mole) of 0-chloropropionylchloride. Cpo n completion of the addition the,mixtur e was re-fluxed 1 hr. and poured very cautiously onto 1. 5 kg. of ice and300 ml . of concentrated hydrochloric acid. The resulting mixturewas extracted with three 100-ml. portions of ethe r, the combinedether ex tracts were washed with 107% odium hydroxide and water,and dried over anhydrous sodium sulfate, and the ether was re-moved under reduced pressure. The resulting oil was crystallizedfrom ether-hexane, giving 47.6 g. f48yo) of a white crystallinesolid melting at 66-68".An d . Calcd. for CLOH1lCIOz: C, 60.45; H , 5.59. Fo un d:

y-Chlo robut yroph enone --Since th e compound prepared bythe method described previouslyz6 id not give the reported rateconsta nt, an alternative preparative route was used. r-Chloro-butyryl chloride (14.2 g., 0.1 mole) was added with stirring to asuspe nsion of 30 g. (0 .2 3 mole) of an hydr ous alumin um chloridein 7.8 g. (0.1 mole) of d ry benzene and 100 ml. of dr y carbon di-

59-61'.

( 2 6 ) J B. Conan t , J. B Segur, and W. . Kirner , J. A m . Ch em . Soc . , 461882 i lS2- i )

63-64 34 29 34 93 3 24 8 0985-86 32 44 33 26 3 07 3 1671-73 37 05 37 49 4 05 4 (16

sulfide. ;\iter 1 h r . of reflux the mixture was cooled and pouredslowly onto 400 g. of ice-water mixt ure. Th e oil obta ined byether extraction mas dissolved in methanol and the solutiorl tle-colorized twice with activated charcoal. af te r removal of th emethan ol the oil was crystallized from ether-hexane to give 10.2g. (567;) of solid, m. p. 19-20' (l it. 26 1.p. 19-20"). T h e twosamples were identical.

6-Chlorobutyl Phen yl Ketone .-To a Grigna rd prepar ed from15.7 g. (0. 1 mole) of brornobenzene and 2.6 g. (0.1 g.- atom ) ofmagnesium in 100 ml. of dry e ther was added with stirring 10.6g . (0.09 mole) of 6-chlorovaleronitrile in 100 ml. of dry e the r.Th e mixture was refluxed for 1 hr ., hydrolyzed with 100 ml. ofice-water followed by 100 ml . of 12 S ulfuric acid, the etherlayer separated, and the aqueous layer further extracted with tw o100-ml. portions of ether. The ether extracts were combined,washed with 5 0-ml. portions of 10yc sodium carb onate and water,dried over anhydrous sodium sulfate, and the ether removedunder reduced pressure. Th e oil obtain ed was crystallized fromether-hexane giving 13.9 g. (79:;) of a white solid melting at48-50" (lit.27m.p. 51').Isolation of Produc ts.-Th e iodo compou nds were isolated i na number of instan ces by refluxing the chlorides with a three-fourfold excess of po tassium iodide in aceto ne solution for ;c nextended period of tim e. A4 5% yield of phenacyl iodide, m. p.23~-24' (lit . B O z 5 ) was obtained after 1hr. ; a 72'3 yield of iodo-propiophenone, m. p. 61-62" ( l i t . 2 961") was obtained.The other iodo compounds obtained do n o t appear to llavebeen prepared previously. Analyses for carbon were generallyhigh, indicating tha t they were not entirely free of the st artingchloro cornpound. Their properties are summarized in Tabl e I S .l-Benzenesulfonyl-4-chloro-l-butene.-Toa solut irin of 20 g.(0 .1 mole) of 6-chlorobutyl phenyl sulfide in 300 ml. of dr y hexaneat 0" was added slowly with stirring 14.9 g. (0 .11 mole) of freshlydistilled sulfuryl chloride. The mixture was allowed to comeslowly to room temperature and , upon disappearance of the chlo-rosulfonium salt, was refluxed for 1. 5 hr. The solvent was re-moved under reduced pressure and the residual oil fractionatedunder a vac uum giving 15.1 g. (76 7,) of a colorless liquid boilingat 112-115' 10.35 rnm.), n P o ~.5840. r\n infrared spectrumindicat ed the presence of a do uble bond. .I solution o f 7 g.(0.035 mole) of the above material and 11.3 g. (0. 1 mole) of 30%hydrogen peroxide in 100 ml. of acetic acid was refluxed for 1 8 hr .It was then poured into water, extracted with two 100-ml.portio ns of chlor oform , and th e chloroform removed unde r re-duced pressur e. The oil obtained was dissolved in 10 inl. ofchloroform and the solution chromatographed on a 2 X 30 cm.silica gel column , eluting with 10Tc ether-hexane. .After theelution of 2..i I . , which contained only small amoun ts of unidenti-fiable oils, a sharp band containing a colorless liquid resultedweighing 3. 4 g. (41y c), n z 0 ~.5603. The n.m.r . spectrum andcarbon and hydrogen analyses were consistent with the structureassigned.A n a l . Calcd. for CloH1IC102S: C, 52.05; H , 4.81. Fou nd:C, 51.86; H, 4.77.

Acknowledgment.-This work was supported in partby the American Petroleum Institute under Project4XB, and in part by th e Office of Ordnance Researchunder Grant No . Dh-XRO(D)-31-124-Gl9.

( 2 i l H. So r m a n t , Com p l . v e n d . , 231, 09 (1950).( 2 8 ) A . Lucas, B e ? , 32, 600 (1899).(29) 1%'. J Hale an d E . C. Britton, J . A m C h e m . .So< , 4 1 , 84 1 ( I LI lY) .