AROMATICS WITH HYDRATED ELECTRONS
Dec. 20, 1964 [CONTRIBUTION FROM
THE
5633
CHEMISTRY DIVISION,ARGONNE NATIOXAL LABORATORY, ARGONNE, ILLINOIS]
The Reactivity of Aromatic Compounds toward Hydrated Electrons] BY 11,AN BAR^
AND
EDWINJ. HART
RECEIVED JULY 17, 1964 T h e specific rates of bimolecular reactions between aromatic compounds and hydrated electrons (eaQ-) have been determined. It has been shown t h a t the reactivity of aromatic compounds toward hydrated electrons is directly related t o their reactivity toward nucIeophilic reactants as expressed by Hammett’s u-function. The contribution of substituents a t different sites on the aromatic nucleus t o its reactivity has been analyzed. A mechanism has been suggested for the eaq- reactions involving the formation of a carbanion as primary product. T h e hydrated electron may be considered the most elementary nucleophile and by far the most reactive one.
Introduction The recently discovered hydrated electron that has been identified by its absorption spectrum3 is a unique chemical species consisting of unit negative charge chemically unbound to any particular atom. I n its simplicity, this species is unparalleled by any other negative ion. The reaction of the hydrated electron does not involve an atom transfer or sharing of electrons between two species, but must unambiguously be considered as an ideal charge-transfer process. The combination of a hydrated electron with a given molecule requires the incorporation of the electron in one of the molecular orbits of the reactive substrate. Although extremely reactive toward certain reactants, the hydrated electrons may exist in the presence of other compounds for long periods, which are limited only by the eaqeaq- and eaqH bimolecular interactions in alkaline methanol solutions. The reactivity of a given compound toward the hydrated electron is evidently determined by the availability of a molecular orbital of sufficiently low energy to interact with the hydrated electron in solution. Following this line of thought, it is evident that the basic chemical properties of a given compound can be examined by a study of its reactivity with eaq-. Further information on a class of compounds may be gained when certain parameters can be changed a t will, as by substitution of one atom by another atom or group. Benzene and its derivatives provide a chemical system in which reactivity is substantially changed by substitution. A study of the aromatic system lends itself as a model to illustrate the potentialities of rate studies of eaq- reactions. It was, therefore, decided to study the chemical behavior of aromatic compounds from the standpoint of their reactivity with eaq-. In previous it has been shown t h a t aromatic compounds react with hydrated electrons at rates which strongly depend on the substituent on the aromatic nucleus. metu-Directing substituents have been found to activate the aromatic compound in its reaction with hydrated electrons,6 but the effect of ortho-paradirecting groups remained an open question. Another ambiguous point was whether the hydrated electron
+
+
(1) Based o n work performed under t h e auspices of the U. S. Atomic Energy Commission. (2) On sabbatical leave from the Weizmann Institute of Science, Rehovoth, Israel. (3) J. W. Boag and E. J. H a r t , N a t u r e , 197, 45 (1963): J. P . Keene, i b i d . , 197, 47 (1963); E. J. Hart and J. UT. Boag, J . A m . Chem. Soc., 84, 4090 (1962). (4) E. J. Hart, S. Gordon, and J. K. Thomas, J . Phys. Chcm., 68, 1271 (1964). ( 5 ) S . Gordon, E. J. Hart, and J. K. Thomas, i b i d . , 6 8 , 1262 (1964). (6) A. S z u t k a , J. K. Thomas, S. Gordon, and E . J. Hart, ibid., in press.
interacts with the aromatic nucleus or with its conjugated substituent.6 It is the purpose of the present study to examine a large number of aromatic compounds in order to elucidate the factors involved in the reactivity of aromatic compounds and subsequently to postulate a mechanism for these reactions Experimental T h e rates of reaction of hydrated electrons with organic compounds in homogeneous deaerated aqueous solutions were measured by pulse radiolysis. The experimental set-up as well as the experimental procedures have been previously described in detai1.4J All rate measurements were carried out in the p H range 10.5-11.5. The aromatic compounds under study were of highest purity commercially available: benzene, toluene, phenol, fluorobenzene, chlorobenzene, bromobenzene, cyanobenzene, nitrobenzene, benzyl alcohol, o,m,p-toluic acid, o,p-chlorophenol, p-bromophenol, o,m,p-hydroxybenzoic acid, o,m,p-nitrophenol, o,mdichlorobenzene, and o,m-chlorobenzoic acid were Fluka puriss grade. Iodobenzene, benzoic acid, sodium benzenesulfonate, naphthalene, trifluoromethylbenzene, p-chlorotoluene, p-toluenesulfonic acid, p-nitrotoluene, p-chlorobenzoic acid, o,m,p-fluorophenol, m-chlorophenol, o,m-bromophenol, p-iodophenol, p dichlorobenzene, o,m,p-fluorobenzoic acid, and o-iodobenzoic acid were Fluka purum grade. o,m,p-Iodotoluene, m,p-iodobenzoic acid, p-aminobenzoic acid, cy- and 6-naphthol, thiophenol, benzamide, and benzenesulfonamide were Eastman White Label chemicals. o,m,p-Cyanophenol, p-cyanobenzoic acid, 1- and 2-cyanonaphthalein were Aldrich C.P. chemicals; a- and 6-naphthoic acids were Matheson C.P. chemicals. Benzoic acid, p-chlorobenzoic acid, p-aminobenzoic acid, and o,m,pfluorobenzoic acids were recrystallized from a methanol-water solution. All other materials were used without any further purification. Stock solutions (1 M ) of benzene, toluene, fluoro-, chloro-, bromo-, iodo-, cyano-, trifluoromethyl, and dichlorobenzenes were prepared in redistilled methanol. Aliquots of these solutions were injected into the matrix consisting of an aqueous solution of 0.001 M S a O H and 0.001 M metlian01,~thus scavenging the H’ and OH produced by the pulse. The methanol concentration in these experiments did not exceed 0.25 M. Sormally 0.001-0.002 M MeOH was used in most of the solutions studied. It was shown t h a t methanol up t o 3 hf in aqueous solution does not affect either the lifetime of e& or its reactivity. The specific bimolecular rate constants were derived from the first-order rates of disappearance of emq- which ranged from 1 X lo5 to 3 X 106 sec.-l. These values were corrected for t h e spontaneous rate of disappearance of enq- in the same matrix, which ranged from 2 . 5 t o 3 . 5 X l o 4set.-'. The errors involved in the determination of the bimolecular rate constants originate from three major sources: A , gravimetric and volumetric errors in the concentrations of the reactants; B , errors involved in derivation of rate constants from the photographic records; C , wrongly estimated rate constants owing to the presence of reactive impurities. T h e first two sources of error limit the accuracy of the specific rate constants t o a standard deviation of i l O S > . The third source of error is negligible in the case of reactive compounds where k > 108 M-1 sec.-1, but it may cause serious error estimates of bimolecular rate constants for compounds of low reactivity. Low rate constants with k < lo7 M-’ set.-' therefore show a nonsymmetrical distribution of errors and in es-
M. A N B A R
5634
AND
EDWINJ. HART
Vol. 86
TABLE I SPECIFIC RATECONSTANTS A N D RELATIVE RATESOF REACTIOS OF SCBSTITUTED BENZENES, TOLUENES, ASD PHEXOLS
-----
~ _ _ _ _ CnH_&14 x 12 x 4 0 x 6 0 X 5 0 x 4 3 x 12 x 16 X 3 1x 4 0 x 3 0x 1 8 X 1.3 X 4.7 X 1,7x 1.6 X
H CHI OH F
c1 Br I CN
coo SO3 -
NOz CFJ CHzOH
SH CONHz SOsNHz
107 107 106 lo7 108 109 10'0 10'0 109 109
0 00 07 f 0 07 54 f 20 63 f 08 1 55 i 05 2 52 f 08 2 93 f 06 3 06 f 06 2 35 f 05 2 46 f 05 3 33 i 05 2 . 1 1 f .05 0 . 9 7 f .05 0.53 f 05 3 08 f .05 3 . 0 6 f .05
1010
lo9 lo8 107 1010 1O1o
1 2
x
107
0 00
4
4 5
x
108
1 57
2 0 5 0
1 1 3 1 1
3 x 1010 4 x 10'0 6 X log 65 x 109 9 x 10'0
Results and Discussion The specific rate constants of the bimolecular reactions of hydrated electrons, eaq-, with a series of monosubstituted benzene derivatives are given in Table I. The rate of reaction varies over four orders of magnitude, from 4 X lo6 set.-' for phenol to 3 X 1Olo M-l set.-' for nitrobenzene. The specific reaction rates of substituted benzene have been related to that of benzene and expressed in terms of 7-values, where 7 = log [ K C ~ H I X + ~ ~ . - / K C ~ H ~ H + ~ . , - ]Comparable . values of 7) are obtained for substituted toluenes and phenols. 1
T
0
0
0
d
-0 1
Fig. 1.-7
1
I
0
I
--------m - X CS H10 H-------
x
106
0 00
-
treme cases k may be considered only as an upper limit. All these factors were taken into account in estimating the standard deviations of ?-values given in Table I.
-C
~-XCSH~CH~-----
I
7.
3
a s a function of u of Hammett's e q ~ a t i 0 n . I ~ 7 = log [kC8HaX+e,,-/kCsHs+e.,-]
The 7-values which indicate the relative reactivity of an aromatic system carrying a given substituent may be correlated with the effect of the same substituent on
3 3 2 2 3
03 07 48 14 22
x x z7x
108
108 109
4 8 X lo9 1 06 x 109 2 5
x
10'0
1 70 2 10 2 83
3 08 2 42 3 80
the reactivity of these systems toward nucleophilic or electrophilic reagents. A quantitative evaluation of these effects has been achieved by the use of Hammett's equation, log [ k C 6 H a X j k C S H 6 ]= pg.'-12 \i:hen the 7-values are plotted against the uporavalues compiled by Van Bekkum, et aZ.,'oa satisfactory correlation is obtained for all substituents studied with the exception of bromine and iodine (Fig. 1). The u-values of Van Bekkum, et al., are equal within experimental error with the rparavalues given by McDaniel and Brown.'l From the 7-u correlation, a p-value of 4.8 is obtained, which is within the range of p-values obtained for various substitution reactions. ' I t is evident that the reactivity of aromatic compounds toward eaq- is determined by the same parameters which control their reactivity toward electrophiles and nucleophiles. As in aromatic bimolecular nucleophilic substitution, l 3 a negatively charged intermediate is formed as the primary product. The formation of carbanions in the reactions of eaq- with organic molecules has been dernonstrated.5a'4 The lifetime of these carbanions may be limited by their dissociation to a substituted phenyl radical and a stable anion. This pattern of reaction has been demonstrated in the eaq- induced deiodination of the iodobenzoic acids, where iodide ions and biphenic acids are formed. l 5 Alternatively, the lifetime of these carbanions is determined by the rate of their neutralization by the solvent resulting in the addition of an H+.14 The analogy between eaq- reactions and nucleophilic aromatic substitution may be of questionable value since our eaq- rate constants correlate better with "normal" U-values derived from electrophilic substitution than with para U-values obtained from data of nucleophilic reactions. This is not surprising (7) L. P. H a m m e t t , "Physical O r g a n i c C h e m i s t r y , ' ' McGraw-Hill Book Co., Inc., X e w York, S . V . . 1940. Chapter 7 . (8) H. H. JaffP, Chem. R e a . , 5 3 , 191 (1953). (9) R. W . Taft a n d C . Lewis, J . A m Chem. SOL.,81,5343 (1959); R. W. Taft, N. C. Deno, a n d P. S . Skell, A n n . Rei,. P h y s . Chem., 9, 287 (1958). (IO) H. Van B e k k u m , P. E. Verkade, a n d B. 51. Wepster, Rec. Ivav. chim.. 78, 815 (1959). (11) D. H. hlcDaniel and H.C. B r o w n , J . Org. C h e m . , 23, 420 (1958). (12) L. 54. Stock a n d H. C. B r o w n , Adu. Phys. Org. Chem., 1, 36 (196:3). (13) S . D. Ross, Puog. Org. Phys. C h e m . , 1, 31 (1963). (14) S . Arai a n d L. M. Dorfman, J . Chem. Phys., 41, 2190 (1964). (15) l f . Anbar, hI. Bobtelsky, and M.M a n t e l , to be published.
Dec. 20, 1964
AROMATICS WITH HYDRATED ELECTRONS
5635
in view of the fact that with eaq- we are concerned with an interaction of an electron with the n-orbitals of the ring, as in electrophilic substitution, rather than with effects on electron distribution and polarizability of a certain substituent. The analogy between u- and ?-values has certain limitations. This is found when checking on the additivity of q-values in disubstituted benzenes. A series of para derivatives of benzoic acid was examined for their rate of reaction with eaq- (Table 11), and their TABLEI1 SPECIFIC RATECONSTANTS AND RELATIVE RATESOF REACTIOXS OF para SUBSTITUTED BEXZOIC ACIDS kpXCsHrC00
Substituent
H 0-
”* CHc F C1 Br
a
M
-1
sec. -1
3.1 x 4.0 x 2.1 x 3.0 x 3 8X
6.0
7.7 I 9.1 CS 1.0 coo 7.3 Value taken from ref. 6.
x x x
x x
109 108 109 109 log 109 109 109 10’0
log ( k X C s H i C 0 0 -/ kCaHoCOO-,
v
0.00 - .89 - .17 - .01 ,09 ,29 .40 .47
.51 .37
1 0 9 ~
7-values, relative to benzoic acid, were calculated. With the exception of the value for OH, these 7-values are proportional to the 7-values of the benzene series (Fig. 2 ) and the excellent correlation between these two series yields B . ~ ~ C ~ H ~ C O=O ~H c ~ H Consequently, ~ . taking q = pu, one may assign to the benzoic acid series a p-value of 4.8/6.5 = 0.74. The behavior of the benzoic acid series exhibits a principal difference between the application of Hammett’s equation to aromatic substitution and to the eaq- reactions. I n aromatic substitution, the U-values are additive to a first approximation and p is constant for any particular reaction, e . g . , nitration, bromination, hydrolysis etc. U-values of two substituents become nonadditive when there is an electronic interaction between both. However, as long as such an interaction does not affect the reactive center, U-values are The 7-values, on the other hand, express changes in the over-all probability of incorporation of an additional electron into the molecule as a whole. It is evident that 7-values of different substituents are not additive; however, a constant p is obtained for any homologous disubstituted series in which one substituent is kept constant and the other is being changed. The higher the rate constant of the nonsubstituted parent compound the lower the qvalues of the series. This conclusion is confirmed by the ?-values of the phenol series (Table I). A molecule, CBH&, which has a high probability of accepting an electron, will be affected to a smaller extent by an additional substituent. The smaller ?-values observed in the more reactive series may also be anticipated in view of the fact that the rate of these reactions approaches the diffusion-controlled limit, 10’O M-’ set.-', where a further increase in rate cannot occur. A deeper insight into the mechanism of interaction of a hydrated electron with aromatic molecules may be gained by considering the reciprocal action of two substituents. The relative rates of a series of di-
-
0
I
2
3
4
?Crus.
Fig.
- as a function of V C , H ~ . log [kxcmcoo - + e , q - / k ~ a ~ a ~ O O - + e a ~ - l log [kCsHsX+eao-/kcaHs+eao-l
2.-?7C6A,coo
PcgHacoo- = 7CsHs
substituted benzenes presented in Table I11 compares the effects of two substituents when in ortho, meta, or para positions to each other. The compounds may be divided into two groups: A, compounds in which the meta and para derivatives are of comparable reactivity, within experimental error. These include the COO-CH3, C1-0-, Br-0-, NOz-O-, and CI-Cl couples. B, compounds in which kmela> kpara: COO--F, COO-I, COO--0-, CN-0-, and F-0-. At first sight it could be suggested that when the two substituents are conjugated in the para positions they decrease the probability of a direct interaction of eaq- with them. This is, however, rather unlikely as COO-, 0-, or CN groups show a very small tendency to accept an electron into their orbitals. Aliphatic carboxylic acids, alcohols, as well as acetonitrile, show a low reactivity toward eaq-. It may be concluded, therefore, that one of the substituents of the second group increases the n-electron density of the ring by resonance while it withdraws electrons by induction. The meta-para effect may be explained by the changes in electron density a t the different positions on the aromatic ring. This may be illustrated by the examples : Taking fluorophenolj one obtains the following charge distribution for the meta and para isomers (OH a t position 1). Positions 2 and 6 in both isomers are of high electron density owing to the electrondonating power of the OH. Positions 3 and 5 in the para isomer are of relatively lower electron density, but in the meta isomer position 5 is exceptionally electron deficient and comprises an ideal site for the attack of a nucleophile. Fluorobenzoic acid presents a similar situation. Although COO- is an electronwithdrawing group, it manifests a negative inductive effect on positions 2 and 6 leaving again position 5 of the meta isomer as an exceptionally reactive electrophilic center. From this discussion i t may be implied that the eaq- reaction is, a t least in part, a highly localized attack on the aromatic nucleus and eaq- may therefore be considered as a reactive nucleophilic species in the classical sense. l7 In view of the conclusion that eaq- reaction may be rather localized, it is surprising to find such a good (16) S. Gordon, E. J. Hart, M. S. Matheson, J. Rabani, and J. K . Thomas Discussions F a r a d a y s o c . , 36, 193 (1963). (17) J. F. Bunnett, A n n . Reu. Phys. Chcm., 14, 271 (1963).
M. ANBARA N D EDWIN J. HART
5636
Vol. 86
TABLEI11 THE RELATIVE RATESOF REACTION OF DISUBSTITUTED AROMATIC COMPOUNDS --Substituents---
koXCsHrY
X
coo coo coo coo coo F
c1
Br CX KO2 c1
Y
CHa F
c1 I 000000-
c1
M -1 sec. -1 2 7 x 108 3 1 x 109 1 2 x 109 4 6 X lo9 3 2 x 109 3 4 x 108 2 0 x 108 1 9 x 109 8 2 X lo9 2 0 x 10’0 4 7 x 109
kmXCaBdY
M
2 6 5 1 1 2 5 2 4 2 5
-1
knXCmHdY
sec. -1
6 X lo9 7 X lo9 5 x 109 3 x 10’0 1 x 109 0 x 108 0 x 108 7 x 109 8 X log 5 x 10’0 2 x 109
M-1
sec.
3 0 x 3 8X 6 0 X 9 1x 4 0 x 1 2 x 6 4 X 2 9 x 2 0 x 2 5 x
5
ox
log ( k m c t a
log ( k m ( h o /
kporo)
kmcta)
-1.03 -0.30
lo9
-0 06 25 - 04
109 108 108 lo8 109 109 10’0 109
16 42 22 - 10 - 03 38 00 01
- .93 .48
-1
109 log
-
.66 .23
- .40 - .l5 .23 - 10 - .04
correlation with upovavalues. It is likely that the enhancqd electron withdrawal from the. halogen by the eaq- reaction is controlled by the existence of a vacant HOOCC& radical compared with C&. We conorbital wherever made available; thus in monoclude that a direct interaction with the substituent is substituted benzene its reactivity is determined by the likely to contribute to the over-all rate of reaction over-all r-electron density, whereas in the disubstiof aromatic iodine and bromine derivatives. tuted aromatic derivatives the formation of a localized A number of other aromatic compounds has been electron-deficient center becomes of greater imporexamined for their rates of reaction with eaq-. The tance. results are summarized in Table IV. The naphthalene Taking the k o v l h o / k m e r o ratios, one encounters two TABLE IV groups of disubstituted derivatives: A, kortho > kmeta> SPECIFIC RATECONSTANTS OF MISCELLAXEOUS AROMATIC kp,,, including COO--O-, F-0-, CN-0-. I n these COMPOUXDS cases the electron-donating capacity of 0- appears to Compound k , M - 1 sec.-l be partially neutralized by the neighboring electron1-Naphthol 9 . 6 X lo8 withdrawing groups resulting in a smaller effect on the 2-Iiaphthol 1 . 2 x 109 Naphthalene 5 . 4 x 109 electron density of the ring. B, koriho < kmelaincludNaphthalene 5.2 x 1 0 9 ~ ing all the disubstituted benzenes which shows kpara 1-Naphthoate 6 . 1 X lo9 kmela. It is suggested t h a t a mutual inductive effect 2-Naphthoate 9 . 5 x 109 of two electron-withdrawing groups, which is enhanced 1-Cyanonaphthalene 2 . 1 3 X 1Ol0 by the negative charge on one of them, decreases the 2-Cyanonaphthalene 2.0‘7 X 10’O withdrawal of electrons from the n-orbitals of the arop-KHzCsH4SOs4 . 6 X lo8 matic nucleus.1s Thus we encounter an “ortho effect” p-CH3CsH4S03i , 6 5 x 109 which has nothing to do with steric hindrance. This CaHsCHzCl 5 . 1 x 109* inductive effect manifests itself also in the pKa values CaHsCHzC1 5.5 x 109~ 4 . 0 X lo8‘ OHCH9CHzCl of the corresponding acids and phenols. CaHscCl~ 8.3 x 1 0 9 ~ The case of iodo and bromo derivatives requires a 3 . 0 X 10lod CHCls separate discussion. These compounds show higher a S’alue obtained in EtOH.18 b Value obtained in EtOH. ?-values than expected from their effects on the electron d Value taken from c Value taken from ref. 20 for comparison. density of the ring. Comparing iodine and bromine ref. 4 for comparison. with other substituents for their reactivity with eaqderivatives represented in the first group show an in aliphatic systems one finds the iodoaliphatic comanalogous behavior to the benzene derivatives (Table pounds highly reactive toward eaq-, unlike cyano, I ) , with OH acting as a deactivating and COO- and carboxy, fluoro, or even chloro derivative^.^*'^ In CN as activating substituents. It should be noted other words, carbon-bound iodine, and to a certain that naphthalene is much more reactive toward eaqextent bromine and chlorine, are capable of direct than benzene, although i t is also more reactive toward interaction with hydrated electrons. In the case of electrophiles.12 This is a result of the reciprocal aromatig chlorine this tendency is neutralized by elecaction between the two rings-whereas one is more tron withdrawal from the nucleus. Iodine, having a nucleophilic than benzene the other is more electrosmall inductive effect,9,l8 may retain this property, philic. a-Naphthol and a-naphthoate are less reacand an iodo aromatic compound may interact with tive than their P-isomers, a fact probably caused by a eaq- in two fashions-through the nucleus and through stronger interaction with the nonsubstituted ring. I n the iodine atom. A single product of reaction may the case of a-naphthol partial deactivation of the nonstill be formed as the negatively charged molecule substituted ring results, whereas in the case of awill achieve its most stable configuration in a very naphthoate the electrophilic reaction center becomes short time compared with its lifetime. The good agreement both for Br and I in the O C ~ H ~ - O C ~ H ~ C O O H more diffuse. A comparison of the reactivities of p-NH&&S03correlation implies an enhancement also of the “direct” and p-CH3C6H4S03- demonstrates again the proreaction with eaq- induced by the presence of the carnounced deactivating power of electron-donating boxylic group. This may be plausible in view of the groups. (18) R. U’. T a f t , J r . , “Steric Effects i n Organic Chemistry,” M. S. Newm a n , Ed., John Wiley a n d Sons, Inc., New York, N. Y . , 1956, Chapter 13. (19) M. h n b a r and E. J. Hart, J . P h y s . Chem.. in press.
(20) I. A. T a u b , D. A. Harter, M. C. Sauer, a n d L. M. Dorfman, J . Chrm. Phys., 41, 979 (1964).
SYNTHESIS A N D ROTATORY DISPERSION OF DIARYL SULFOXIDES
Dec. 20, 1964
The third group of compounds to be considered in Table IV is benzyl chloride and cy, a,a-trichlorotoluene. The rate of the C&&,CHzC1reaction is faster than expected for the CHzCl group as an aliphatic substituent, when compared with 6-chloroethanol. We conclude, therefore, that the aromatic nucleus is activated toward a nucleophilic attack by electron withdrawal, both by inductive and by resonance effects. The chlorine atom, on the other hand, is also activated by the aromatic nucleus.ly The interiction of eaq- with C6H6CC13,on the other hand, is in greater part a direct interaction with the chlorine atoms.lg In fact, this rate is slower than that with c h l ~ r o f o r m . ~This is attributed to a withdrawal of electrons from the aromatic nucleus resulting in partial deactivation of the -CC13 group.lY This last case is another example of a direct reaction of eaqwith a substituent. The use of our eaq- rate constants as a measure of electron density in the nucleus is therefore limited to substituents which do not readily interact directly with eaq-.
5637
It is interesting to find that comparable rate constants are obtained for the naphthalene and benzyl chloride esol- reactions in water and in ethanol, As the anionic product is a sparingly solvated species, this result corroborates the conclusion t h a t the solvation of the electron in both solvents is comparable.20 In conclusion, i t can be stated that the hydrated electron may be considered as the most elementary nucleophile as well as the simplest reducing agent. The rules that apply to the eaq- reaction are governed by the same parameters that determine aromatic substitution13 and the information gained from the eaq- reactivity may be applied to aromatic chemistry in general. It is gratifying to observe that the Hammett's freeenergy relationship holds up to the limit of diffusioncontrolled processes. Acknowledgment.-We wish to acknowledge the highly valued help of Dr. S. Gordon, the technical assistance of ,Miss P. Walsh, and the cooperation of Messrs. B. E. Clifft and E. R. Backstrom of the Linac Group.
+
THE DEPARTMENT OF CHEMISTRY, UNIVERSITY OF XEW HAMPSHIRE, DURHAM, N WESTERN REGIONAL RESEARCH LABORATORY,' ALBANY10, CALIFORNIA]
[A JOIST CONTRIBUTION FROM
H., A N D
THE
Optically Active Sulfoxides. The Synthesis and Rotatory Dispersion of Some Diary1 Sulfoxides2 B Y KENNETH K.
A ~ N D E R S E NU, ' I~L~L I A M
GAFFIELD,NICHOLAS E. R . I. P E R K I N S 3 a ' C
P A P A N I K O L A O U , 3 a JAMES
W.
FOLEY,3a'b AND
RECEIVED JUNE 19, 1964 Several optically active diaryl sulfoxides have been prepared by treating optically active ( - )-menthyl ( - )arenesulfinates with arylmagnesium bromides. This reaction is shown t o produce optically pure sulfoxides. The optical rotatory dispersion curves and ultraviolet spectra of the sulfoxides and the sulfinate esters were determined and are discussed.
Sulfoxides can be synthesized by treating sulfinate esters with Grignard reagents4 If the sulfinate esters are optically active a t sulfur, optically active sulfoxides are formed. The first example of the use of this method was reported in a preliminary c o m m ~ n i c a t i o n . ~ (+)-Ethyl p-tolyl sulfoxide (2) was prepared by the reaction of (-)-menthyl (-)-p-toluenesulfinate (1) with ethylmagnesium iodide.
0
.
