The Hydrolysis of Chloromethyl Aryl Sulfides1

found to follow a first-order law and were independent of hydrogen ion concentration. An earlier suggestion that the mechanism forhydrolysis of -halo ...
0 downloads 0 Views 355KB Size
F G . BORDW'ELL. GLENND. COOPER.\ND H. MORIT.~ [CoSTRIBUTION FROM

THE

CHEMISTRV DEPARTMENT OF

SORTHIVESTERN

Vol. 79

CNIVERSIT?;]

The Hydrolysis of Chloromethyl Aryl Sulfides BY F. G. BORDWELL, GLENND. COOPERA N D H. NORITA RECEIVED .AUGUST 21, 1956 T h e hydrolysis rates for eight m and p-substituted chloromethyl aryl sulfides, determined conductometrically in 50% aqueous dioxane, were found to follow a first-order law and were independent of hydrogen ion concentration. .4n earlier suggestion that the mechanism for hydrolysis of a-halo ethers and a-halo sulfides involves the cleavage of a carbon-oxygen (or carbon-sulfur) bond in a complex, such as [ROHCH2C1]+, is discarded in light of the kinetic evidence. Although the variation in hydrolysis rates from p-N02CsH4SCH2C1 to p-CH30C6H4SCH2C1was almost 600-fold, the d a t a fitted the Harninett equation nicely. The lower rate for phenyl as compared to alkyl chloromethyl sulfides and the failure of p - c H 3 0 C s H ~ SCH2C1 to deviate appreciably from the value calculated using the Hammett equation are interpreted as indicating oiily minor resonance contributions involving the aryl groups in the carbonium-sulfonium intermediate, [ ArS-CHz .\rS+=CHz]. This is consistent with other d a t a indicating a reluctance of divalent sulfur to expand its valence shell to ten clectrons. +

An investigation of the hydrolysis rates of a-halo sulfides and of a-halo ethers has been reported previously by BohmeZa and by Bohme, Fisher and Frank.?L These authors followed the hydrolysis in dioxane solutions containing 0.2 to 40% of water by titrating the hydrochloric acid produced with a 0.1 ilr solution of tribenzylamine in acetone. The results obtained from Bohme's initial studyza may be summarized as follows: (1) the reaction is first order in halide (average deviations of about 5y0in first order constants); (2) the reaction is autocatalyzed by the hydrochloric acid produced ; ( 3 ) ClCHzOCHzCHs hydrolyzes in dioxane containing 2y0 water a t a rate approximately 1000 times that of C1CH2SCHzCH3. Biihme?" suggested the following sequence of steps for the hydrolysis of a-chloro ethers and thioethers to account for these experimental results ROCHaCl

+1

H [ROCHICI]

1 ' s

+

+ HzO --+

[CHyC1]+

H [ROC€I?CI] (fast) +

+ [CH2C1] (slow) CH?O + 2H' + C1- (fast) ROH

+

(1) (2) (3)

The autocatalytic effect observed was ascribed to the necessity of formation of the oxonium or sulfonium salt in step 1. The slower rate of hydrolysis of the a-chloro sulfides than a-chloro ethers was attributed to a lesser tendency of the sulfonium ion than the oxonium ion to undergo ionization in step 2. Actually, the proposed mechanism does not satisfy the data. Since step 2 is assumed to be rate determining, the rate will be proportional to the oxonium or sulfoniuni ion concentration, and this depends on the hydrogen ion concentration. €I [ROCH2Cl]

phenyl chloromethyl sulfide, CRHSSCH2C1,was undertaken. The rates of hydrolysis of seven mand @-substituted phenyl chloroinethyl sulfides were measured to determine the electrical effects of the m- and @-substituents. Bohme, Fisher and Frankzb have shown that the over-all reaction involved in the hydrolysis is that given in equation 4. 2.1rSCH2C1 + H 2 0 = (.\rSj2CH2 C H 2 0 4-

+

2H+

Inasmuch as [H+] increases throughout the reaction, the rate expression would not be a simple first-order one but would involve [H+]. In order to clarify the situation with regard to the mechanism, a kinetic study of the hydrolysis of (1) This investigation was supported by t h e American Petroleum Institute under Project 4 8 0 and b y t h e Oflice of Naval Research under Contract S o . N7onr-45007. A preliminary account of t h e work was given a t t h e American Chemical Society Meeting a t Atlantic City, K ,J . , Septembrr, 1552, Ahst. 11. 5->'I. ( 2 ) (a) H. Bijhme, R w , ,74, 218 (1941); ( b ) H. Biihme, H. Fisher and R . I h n k . 4nn., 663, 54 (1940); (c) see also R. Peters and 13. Walker, HioChPnz. J , 17. SGO (1923)

+ 2C1-

(4)

Since the reactions are very rapid and are sensitive to solvent when small quantities of water are present, i t is much more convenient to follow the rates conductonietrically than by b titration method. I n this way the rates could be followed in solutions containing more water (50% aqueous dioxane, by volume) than is possible by Bohme's method, even though the half-lives were as low as 23 seconds. Rate constants were determined graphically from a plot of log ( R / R - R E ) us. t , where R is the resistance at time t and R Eis the equilibrium resistStraight-line relationships were observed in all instances, often out to 90% completion of the reaction. No indication of auto-catalysisZa was evident in any of the runs (see e.g. Fig. 1). The rate constants given in Table I are for the most part the average of two or more runs. A plot of log k vs. l / T gave straight lines for the hydrolysis run at three or more temperatures. TABLE I KINETIC DATA FOR THE HYDROLYSIS O F ARYL CHLOROMETHYL SULFIDES, YCBI14SCH2C1,I N 50% AQEEOCS DIOXANE Em*

Y

Ke,[ROCHzCI] [ H + ]

-

O-CHaO $-CHI m-CHn H p-CI p-Br m-CI P-NO2

--Rate constants X 1 0 5 scc. -I-kcal./ 210 2 5 . 1 ~ 3 o o 34.83O 50' 60.7' mole 20 eno 1900 3000 22.5 200 860 1400 700 iin 190 500 2,730 20 I40 90

AS+.

em. -2.8 3.8

-G c i

49

5.7

33

86

21.5

-10.6

Discussion Since no autocatalytic effect was observed and the hydrolysis rates followed first-order kinetics ( 3 ) A. A Frost and R . G. Pearson, "Kinetics and hlechanism." John Wiley and Sons. Tnc , S c m York, X . \-., 1953.

377

HYDROLYSIS OF CHLOROMETHYL ARYL SULFIDES

Jan 20, 1957

1.2 I

-0.4

I

0

I

I

0.4

0.8

1

U.

0.0 0

100

50

I50

Fig. 2.-Relationship between the rate of hydrolysis of chloromethyl aryl sulfides and Hammett sigma values; determination of rho.

-2.6, showing a somewhat greater effect of substituents than in the hydrolysis of benzyl halides. The fact that the log k/ko values for the and p-CH,O compounds fell near the line when despite the large increase in [ H + ] during the reac- plotted against the “normal” U-values (those based tion, the reaction appears to be better represented on the ionization constants of the benzoic acids5) as an ionization to form a carbonium-sulfonium indicates that resonance interactions of these ion than by the mechanism suggested by B o h n ~ e . ~groups ~ with [ -%CH,] are no greater than their resonance interactions with the carboxyl group of ArS-CH2-Cl --+ [ArS-CHz+ ++ Ar$=CHz] C1- benzoic acid. This is a t Erst sight somewhat 5urThe importance of the participating effect of prising since in other examples of reactions where sulfur in promoting this hydrolysis is made evident unshared electrons are available on an atom a t by the fact that C6H5SCH2C1hydrolyzes 10 times tached t o the ring (reactions of amines, phenols, as rapidly under these conditions as does (CH3)3- thiophenols) the p-NO2 group and similar groups CC1. Substitution of the C6H5S- group for a show large deviations6 (the apparent U-value for hydrogen of CH,CI therefore has a greater effect p-NO2 may be increased by as much as 0.5 unit, than substitution of all three hydrogens by CH3. which would lead to a rate about 10-fold that calThe data of Peters and Walker2‘ indicate that the culated on the basis of the “normal value”). participating effect of sulfur to the release of halide Similarly, a deviation for the p-CH30 group might ion is about five times as great in the a-position be possible, since when a positive charge develops (ClCH2SCH2CI) as in the &position (CICHzCH2- on the atom attached to a benzene ring in the transition state of a reaction (e.g., hydrolysis of ArCH2SCHzCHzCl). The hydrolysis rate of a 0.001 M solution of C1 or ArCH20S02C7H7),the p-CH30 may show a C6H5SCH~C1 was increased only 1.5-fold by the pres- greatly enhanced effect6 I n terms of current structural representations the ence of an equal concentration of hydroxide ion results suggest that neither resonance structures of (1.1-fold by 0.001 i W KCIOI). Since the hydroxide ion concentration is increased by lo4 in this me- the type Ia nor I I a make very large contributions dium, the ionization is virtually unaffected by the to the hybrid structure. presence of alkali. CHZ CHz CHz CHz I I’ /I The greater effect of oxygen than sulfur in pro:s+ :S: :s+ moting the ionization of an a-halogen2 is underl /I I standable since oxygen is known to enter into electron-donor type resonance to a larger extent than does sulfur. For the hydrolysis of benzyl halides in aqueous acetone p of the Hammett e q ~ a t i o n log , ~ k/ko = up, is negative ( p - 1.87) and groups with positive /\ I I -0 0CH3 CHa u values ( m or P-N02,CN, COCH,, S02CH3, etc.) I Ia I1 IIa retard the reaction rate, whereas groups with negative U-values (p-CH,O, p-CH3, m-CH,, etc.) The unimportance of I a may be rationalized on accelerate it. Similarly it was found that p-NO2 the basis of Pauling’s adjacent charge rule.7 The caused a 100-fold retardation in the rate of hydrol- failure of ITa to make a substantial contribution is ysis of C&&,SCH2C1 and p-CH8O caused a 5.4-fold more surprising. However, evidence has been acacceleration. A plot of log k / k o us. the “normal” cumulating to indicate that divalent sulfur is relucu-values5 for the groups gave a reasonably good ( 6 ) See, for example, G. E. K. Branch and hl. Calvin, “Theory of straight line (Fig. 2). The p-value (slope) was Organic Chemistry,” Prentice-Hall,Iac., h-ew York, N. Y . , 1941, p. SECONDS.

Fig. I.-Hydrolysis of chloromethyl phenyl sulfide in 50% aqueous dioxane at 34.85’.

f

+

+

(4) Pertinent data are summarized by F. G. Bordwell and P. J. Boutan, THISJ O U R N A L . 78, 861 (19.56). ( 5 ) I. P Hammett, “Physical Orgrnic Chemistry.” hIcGraw-Hill Book C o , Inc., h’ew York, N. Y . , 1940, Chapter VII.

+

442; C. J. Swain and W. P. Langsdorf, THIS JOURNAL, 73, 2813 (1951); J. K. Kochi and G. S . Hammond, i b i d . , 76, 3445 (1953). (7) L. C. Pauling, “The Nature of the Chemical Bond,” Oxford University Press, London, 1948.

F. G. BORDWEIL.GLENND. COOPERAXD H. M O K I T . ~

378

Vol. i ! )

tant to expand its valence shell to ten Experimental13 which is required in IIa. Evidently the m- and Preparation of Chloromethyl Aryl Sulfides.---The iiietliotl p-substituents in drSCH2Cl do not relay their used was chlorination of the corresponding inethyl aryl sulwith sulfuryl chloride according to the procedure ticelectrical effects primarily through resonance in- fides scribed previously for phenyl and p-riietliox~-phen)-lchlorovolving sulfur, and yet p has a more negative value methyl sulfides.14 The coinpounds were purified by v:icuutii here than for the hydrolysis of XrCHZCl. This distillation just prior to use except for ~-tiitrophciiylchlorofurther bears out the importance of inductive methyl sulfide, which was crystallizcd from a solution of benzene and pentane. These compounds may cause ;L pereffects and field effects.g sistent and annoying dermatitis so t h a t caution ill handling The inability of the positively charged sulfur of is advisable; pnitrophenyl chloromethyl sulfide appears t o ArS+=CH2 to distribute its charge over the be particularly toxic. Table I1 summarizes the 1-ields obphenyl ring by resonance is further emphasized by tained :~iidgives the analyticd clata. the observationzb that the rate of hydrolysis of 'r.11jt.E 11 KSCHzCl (R = Me, E t , Pr, etc.) is about 200 times CIiLOROME'iX1YL i k T L SCLFIDES that of CsH6SCH2Cl in 90% aqueous dioxane. I n Analyses, (,: Chluromethyl B P, Yield, Calcd. Found our more dilute solutions the rate of solvolysis of sulfide ;C. n I m . ';6 C H C Ii CHJSCHzCl was found to be too fast to measure. ~ I - C H ~ C ~ I I I S C H ? C11'5 I 11, There is no question but what the rate is consider- ;p-CIIaC6H4SCII?CI l z % l ? ( > 13 8:i .fi-l .j 2 3 55 !I8 . 5 . 3 ? ably greater than for the phenyl compound, as re- I,I-CIC~H~SCH?CI I l . j - l l 0 3 7 3 1 3 . 3 1 ,'1.1.1 4 : L A . j ; i . : j i I,'< 4;3,7Ij 3 . 2 2 ported.2b The phenyl group is much superior to p-CICaHaSCH~C1 118-1 2 9 1 2 08 4 3 . 2 4 .i. 3 3 -L(J 2 . X 1 3 > . 3 1 2 . 7 I i alkyl groups in stabilizing the positive charge on an P - B ~ C B H I S C H ~ C I 150 15 !I1 4 1 2 8 2 . 9 7 41.38 3 . 0 0 CI~ adjacent carbon atom. Thus, substitution of ~ - N O ~ C B H I S C H ~02-64" Calcd.: C1, 17.48. Fouiid: C1, 17.30. h1.p. phenyl for an whydrogen in ethyl chloride produces an effect comparable to that of two methyl groups Kinetic Runs.--.-\~)proxiriiately 0.0003 mole of the chloro(in SOYo alcohol k for solvolysis of C6H&H(CHa)C1 methyl sulfide was weighed into a flask and dissolved in 25 cc. of purified dioxane a t 34.85'. T o this solution 25 cc. = 0.59hr,-I a t 50"'" and k for (CH,),CCI = 0.37 hr.-l a t 45'"). The superiority of phenyl over of water ( a t 34.85') was added quicklj-, t h e solution swirled vigorously t o ensure complete mixing and a. portion poured alkyl in stabilizing a carbonium ion is also ap- into an open conductivity cell. T h e resistance was measparent from Bohnie's work,2bwhich shows that the ured with a Jones bridge.16 Plots of log R / ( R - R E )V.T. relative hydrolysis rates of CH3SCH2Cl, CH, time have excellent straight line relationships for all eight T h e rate constants were obtained by multiSCH(CH3)Cl and CH3SCH(C6Hj)C1 in 00% diox- compounds. plying the slopes of the lines by 2.3. dctivatioii energies ane a t 25' are 1:500:70,000. The relatively slow were calculated by multiplying the slopes of the lines froin hydrolysis of CcH&3CHzCl us. CHSSCHzCl suggests the log k u s . l ! T plots b y 4.58. Entropies of activation that resonance contributions of structures such as were calculated using the equation I11 to the transition state are of less importance + ==SLCH2 - ; - __ ~ = $ - C..H . C l ,_

~~

~

111

IV

than resonance typified by IV in the ground state. Again the reluctance of divalent sulfur to expand its valence shell is made e ~ i d e n t . ~Examination ,~ of Table I shows that the entropies of activation are not constant in the present series. -1greement with the Hammett relationship is achieved rather by a compensation in the activation energy and entropy terms.I2 (8) See A . RIangini a n d I