BROMINATION OF DEOXYBENZOIN
Sept. 5 , 1963 [COSTRIBUTION FROM
THE JAMES
2591
BRYANTCONANT LABORATORY OF HARVARD UNIVERSITY, CAMBRIDGE 38, MASS.]
Halide Catalysis in the Bromination of Deoxybenzoin BY R O N A L D F. W. CIECIUCH’ A N D F. H . WESTHEIMER RECEIVED MARCH9, 1963 The rate of bromination of deoxybenzoin in aqueous acetic acid solution, in the presence of perchloric acid and at constant formal ionic strength, is zero order in bromine, first order in ho and in the concentrations of deoxybenzoin and water. The bromination in 97.9% acetic acid proceeds 13 times more rapidly with 0.1 M HBr, and 26 times more rapidly with 0.1 M HCI, than with 0.1 11.I perchloric acid, The superiority of the halogen acids decreases markedly in more aqueous solvents. For the portion of the reaction in 97.9%, 91.0yo, and 79.904 acetic acid which is specifically catalyzed by t h e halogen acids, the addition of LiC10, sharply decreases k / h o , but the addition of LiCl increases this ratio. .it constant formal ionic strength, the reaction is zero order in bromine, first order in ho, in the concentration of deoxybenzoin, and in the stoichiometric concentration of halide ions. These d a t a have been discussed in terms of the tentative hypothesis that halide ions function as bases in 79.9, 91.0, and 97.9% acetic acid, where they are only partiallg hydrated. I n this formulation, the ketone is first protonated and a hydrogen ion is then removed from the a-carbon atom by a halide ion (or possibly by a closely related ion-pair).
Introduction T h e halogenation of ketones provides a convenient measure of the rate of enolization.2 In acid solution, the reaction proceeds3 in two steps. T h e enol, formed RCOCHIR
+ H30+
K
RCCH2R (rapid equilibrium)
(1)
I1
*OH k
RCCHzR
it
‘OH
+ B +RC=CHR
(rate-limiting step)
i
(2)
+ BHi,
OH
in the rate-limiting4 step, is then rapidly halogenated. In dilute aqueous solutions of mineral acids, the base which removes the proton is a water molecule. When weak acids are present, the effective bases include corresponding anions; the kinetic equation for reactions 1 and 2 is v = (ketone) [k(H+) ~ H A ( H A ) ]Studies . of the halogenation of ketones in solutions of strong acids have generally been concerned with the appropriate acidity function which should be applied to the reacti0n.j In the present study, deoxybenzoin was brominated in 79.9%, 91.0%, and 97.9y0 acetic acid. in the presence of perchloric, hydrochloric, and hydrobromic acids. The rates with 0.1 31 HBr and HC1, in 97.9% acetic acid solution, were 43 and 26 times as fast as with 0.1 M perchloric acid, despite the greater acid strength of the latter, and therefore the greater h, values of its solutions. Similar and less spectacular findings were obtained in 91.0% acetic acid; the specific catalytic effect of the halogen acids in i9.9yoacetic acid is small. The interpretation of these results is complicated by the incomplete dissociation of the acids into ions; in 97.9y0 acetic acid, ion-pairs predominate. However, the results are best understood on the basis of general acid catalysis by the halogen acids. An understanding of the special circumstances which permit such catalysis is offered in the Discussion. Experimental
+
Materials.-One batch (600 ml.) of d u Pont C.P. Reagent grade acetic acid was purified by adding 6 g. of chromium trioxide to the hot acid and allowing it t o stand overnight. I t was flash-distilled under vacuum and then carefully fractionated with a 26-cm. column packed with glass helices. The main fraction, ( 1 ) National Science Predoctoral Fellow, 1955-1965.
(2) A . Lapworth, J . C h e m . S o c . , 86. 30 (100-4) ( 3 ) For a discussion of the mechanism ot enolization see I < . P. Bell, “The Proton in Chemistry,” Cornell University Press, Ithaca, N Y . , 1959. ( 4 ) For the definition of “rate-limiting,“ see J . Iiocek, F H . U’estheirner, A . Eschenmoser, I,. LMoldoJanyi, and J . Schreiber, Heiv C h i m A c t o , 46, 2554 (1962). iootnote 7 . (5) 1,. Zucker and 1. P Hammett, J A m C h r m . Sac , 61, 2785 (1935); D P iY Satchell, J Chem S a c , 2878 (1957); G Archer and I< P. Bell, i b i d . , 3228 ( 1 5 % , H J . Campbell and J . T . Edward, C a n . J . C h e w . , 38,2105 (1960); C G . Swain and A S . Rosenberg, J . A m . C h e m . Sac., 8 3 , 2134 (1561)
boiling a t 118.0 f 0.2”, had a melting point of 16.59 z t O.O1° (N.B.S. thermometer). This corresponds to 99.97574 acetic acid, assuming t h a t water is the dominant impurity.6 Four other batches of d u Pont C.P. Reagent acetic acid were used without purification; the d a t a were accepted if they corresponded to those from the purified acid. One batch of acetic acid was found t o contain a trace of a n aldehydic impurity (2,A-dinitrophenylhydrazine test) tentatively identified as glyoxal a t perM . The Ho measurements made with this batch haps 3 X of solvent differed by as much as 0.15 unit from those in purified acetic acid, presumably because the glyoxal reacts selectively with the basic form of the indicator. The suspected Hn values were discarded. Rates of bromination determined in this solvent were consistent with those found with other batches. Reagent grade bromine (Merck) was fractionated froin 0 . 5 % of potassium bromide ( t o remove traces of chlorine). Matheson 99.0% (min.) hydrogen bromide was passed tlirough a drying tube packed with Drierite; Matheson 99.0(;/, (inin.) hydrogen chloride was dried with calcium chloride. Merck 70‘,2 perchloric acid was standardized before use. Analytical reagent lithium chloride (Mallinckrodt) was dried for 5 hr. a t 210’; lithium perchlorate (prepared from reagent grade lithium carbonate and reagent grade 70% perchloric acid) was dried to the anhydrous salt by heating to 230’ for 20 hr. Mallinckrodt S . F . grade lithium bromide was recrystallized from water, then dried first over phosphorus pentoxide, and finally a t 100° and 0.2 pressure for 3 hr. Mallinckrodt Reagent grade acetic :inhydride was fractionated through a 46-cm. \.igreux column. Stuart Oxygen Co. deuterium oxide, 99.5[;, was used without purification. Acetic acid-I-d was prepared by refluxing a slight excess of deuterium oxide with acetic anhydride for about 8 hr. Excess deuterium oxide was removed by azeotropic distillation with benzene thrriugh a 35-cni. column packed with glass helices. Fractionatio:; produced acetic acid-I-d, b.p. 117-118.5’. The melting point of this acid, determined 31 months after its preparation, was 15.25 f 0.02’ ( S . B . S . thermometer), lit.’ 15.66 f 0.05’. Deoxybenzoin, after recrystallization from ethanol, melted a t 55.7-56.3’, 54.5-55.7”; desyl bromide, recrystallized from 955i8 ethanol, melted a t 57.1-57.5‘, lit.9 54-55”, Dibromodeoxybenzoin, recrystallized from acetic acid, melted a t 111.8-112.7’, lit.Io 112’. o-Sitroaniline, from ethanol, melted a t 71.7-72.5’; p-nitroaniline, from 50(, ethanol, melted a t 148.1-119.1°. Other chernicals were reagent grade. Solutions.-The “97.9y0,” “ 9 l . O ( ~ , , , ”and “79.9‘);’’ acetic acid were made up accurately by weight” from distilled water and from “glacial” acetic acid of known melting point (and hence of known watcr content). The standard sample of 97.9(,; acetic acid melted a t 13.29 f 0.02’. One batch of 97.9‘70 acid which melted a t 13.13 f 0.02’ presumably contained 0.07-.0.08>~excess water. “97.95;” acetic acid-1-d was prepared so as t o be 1.1620 ’11 in deuterium oxide in acetic acid-1-d, and so correspond to the solvent with ordinary water. The soluticin of DCI i n this s)lvent was mrlde by adding acetyl chloride and DjO to acetic acid-1-d in quantities calculated to yield the 5ippropri:rte acid solution. Stock solutions of hydrogen halides in aqueous acetic acid were prepared by bubbling the dry gas into the appropriate solvent (6) J . S Fritz and G S. Hammond, “Quantitative Organic Analysis,” John Wiley and Sons, I n c , S e w York, N. Y , 1957, p 225 (7) H . 1,inschitz M E Hobbs, and P. 11 Gross, J . A m C h r m . S O L , 63, 3231 (1541). (8) A . C . Cope, P A Trumbull, and E R Trumbull, ibid., 80, 2844 (15.58). ( 5 ) E . Knoevenagel, C h r m Ber., 21, 13.53 (1888). (10) T . Curtius and H . Lang, J . p r n k l Chrm , [ Z j 44, ,547 (1891) (11) “International Critical Tables,” Vol. 111, McCraw-Hill Book C o , Inc , New York, S . Y , 1928, p. 324
RONALD F. W. CIECIUCH AND F. H . WESTHEIMER
2592
and standardizing the resulting solution by potentiometric titration for halide.I2 Solutions of hydrogen bromide and lithium bromide had t o be freshly prepared every few days; they became yellow on prolonged standing. Solutions of bromine were prepared each d a y . Methods.-The acidity function, H o , was determined spectrop h o t ~ m e t r i c a l l >with - ~ ~ a Beckman quartz DU spectrophotometer equipped with a specially constructed thermostated cell compartment, maintained a t 30.15 i O.0lo. Two indicators (0- and p-nitroaniline) were used. The absorbancy of the acid form, IH', of o-nitroaniline was neglected, b u t that of p-nitroaniline a t 375 mp was taken into account as a small correction.I4 The absorbancies of the fully basic forms of the indicators were measured when sodium acetate was added to the solvent. The value 0.99 was assigned13 to the p K of p-nitroaniline, and a value of -0.48 was then found for o-nitroaniline, from measurements with both indicators in 79.9c/c acetic acid. Although the p K of a-nitroaniline in watcrI3 is +0.29, our result is consistent with those of others who worked in glacial or aqueous acetic acid as solvent.l*,15 Experiments with carefully prepared solutions of 97.71 and 98.00i-(i acetic acid showed t h a t the values of Ho generally run about 0.05 unit more negative in the less aqueous solvent. Since sonie of the measurements were made in solvent containing about O,lcd. extra water, all the measurenients were corrected t o a common basis. The corrections were small, and the error in the correction is probably negligible. The detailed data are available e l s e ~ h e r e . ~ ~ The rate of bromiiiation of deoxybenzoin was followed by observing the disappearance of the specific absorbance of bromine and tribromide ion a t their isosbestic point; in the presence of chloride ion, the isosbestic point for bromine and bromochloride ion was used. The needed d a t a are presented in Table I.
1 BROMINE A S D COMPLEX
T.4BLE
ISOSBESTIC Solvent wt % AcOH
97 9 91 0 i9 9
P O I S T S FOR
--
Brj-BrJmi
450 448 448
t
119 122 117
ION5
YBrz-BrK? A. m p
429 429 429
e
152 150 5
1-17
The rate measureas , t s usu l!y were conducted with 0.0200 M deoxybenzoin and 0.002 M bromine. Thus all the bromine was consumed when only 1 0 ~ of o the deoxybenzoin had reacted, and the reaction therefore followed nearly perfect zero-order kinetics. The recorded first-order rate constants are tlie quotient of the observed zero-order rate divided by the concentration of deoxybenzoin. (Only in 97.9':; acetic acid, with perchloric acid as catalyst, was a special situation encountered; there the strong catalytic influence of HBr made itself felt in autocatal?;sis.) Most experiments were carried out in duplicate or triplicate; m e set (used to standardize the solvents) was repeated six times. The spread between duplicates was seldom as great as 5 % ; the average deviation of the sextet was 1:;. Product.-The product of bromination was determined in 97 9','6 acetic acid after ten half-lives in tlie presence of HCI. The solvent was removed by evaporation under vacuum and a weighed sample of the residue dissolved in carbon disulfide. The itifrared spectrum was virtually identical with t h a t obtained with pure desyl bromide; none of the bands characteristic of either deoxybenzoin or a,cu-dibromodeoxybenzoiri appeared in the product, which was a t least 98'; pure. In other experiments, the rate of bromination of desyl bromide was determined; it is only about l/,oo that of deoxybenzoin. Although the brotnination of deosybenzoin is promoted by strong illumination, no appreciable photochemical reaction occurred because of brief illumination in th? spectrophotometer during measurements, or in the diffuse light of the laboratory while the solutioris were prepared. The r:rte with a degassed sample under vacuum was the same as t h a t obtained in ordinary runs.
Results Ho-Values.--The H-values for the acids and acidsalt mixtures used are presented in Table 11. These data are consistent with those found by others for aqueous acetic acid solutions. 1 4 ' l i , l Y ( 1 2 ) I Ilustrovsky and M H a l m a n n , J . Chem. S o c , 506 (1953); B. C a v a n n g l i , z b ~ d, 2207 (19271 ( 1 3 ) M 4 . Paul and F A . I.ong, Chem Rea., 5 7 , 1 (1957). 11-11 J IiuEek, Collection Cerrh C h e m C o m m u n . , 22, 1 (1967) (I;) N P. Hall and F Meyer, J A m ('hem. .Soc , 62, 2493 (1940). (161 Ronald F. W Cieciuch, P h . U Thesis, Harvard University, 1960. (17) K . B Wiberg and R J E v a n s . J A m Chem Soc , 80, 3019 (1Y58). (18) T I, S m i t h and 1.H. Elliott. rbid , 7 6 , 3566 (1953)
FOR
Acid
Mole/l.
VOl. 85
TABLE I1 AQUEOUSACETICACIDSOLUTIONS - Ho in aqucmis acetic acid
Salt
Mole/l.
97 9 7 , a
Y1.0%
7Y.97,
0.100 .. . 1 . 4 4 0 02 - 0 65 ,100 LiClO, 0.200 1.62 20 - 18 ,300 . . . . 2 22 .75 01 HBr ,020 . .. . 0 61 ,020 LiCIOa ,080 .63 ,025 . .. , 73 025 LiCIOI ,075 75 ,100 .. . 1.44 100 LiC104 100 1.47 HCI ,025 0.44 ,025 LiCl ,075 48 025 LiClO, .075 .58 025 LiCIO, ,275 60 100 1 12 0 03 65 ,100 LiClO, ,200 1.35 18 - 51 100 LiCl 200 .16 300 .71 +.01 Some measurements in Q7.9T0 acetic acid corrected by 0.05 unit t o allow for O.l".b error in water content; see Experimental. HClOa
+
Kinetics. (a) Perchloric Acid.-The results of a kinetic run with 0.100 M HClO,, 0.0202 Jf deoxybenzoin, and 0.0020 *If bromine in 97.97, acetic acid are shown in Fig. 1. Autocatalysis is clearly visible. Experiments showing that H B r is a much more powerful catalyst than perchloric acid suggest that the former is responsible for the observed autocatalys,s. When the effect of HBr was taken into account, the corrected zero-order plot of Fig. 1 was obtained. T h e method of correction is given in the -1ppendix. Experiments in the presence of HBr or HC1 gave excellent zero-order curves without correction; further, in 91.0% a n d 79.97, acetic acid, the autocatalysis was negligible, even with perchloric acid as catalyst. T h e rates are first order in the concentration of deoxybenzoin over the range from 0.02 to 0.06 .If, and each run was zero order in bromine. The data for 9i.9yo acetic acid are collected in Table 111, those for 91.0% acetic acid in Table IV, and those for i9.9yo acetic acid in Table V. T h e observed zero-order rates divided by the concentrations of deoxybenzoin are given as k l . (b) HBr and HC1.-Inspection of the data in Table I11 shows that the order of the reaction in HCl and HBr is considerably greater than unity. Thus, a t a constant formal ionic strength of 0.100 X , an increase in the concentration of HBr from 0.020 to 0.050 M increases the first-order rate constant from 7.12 to 33.3; an increase of the concentration of HBr to 0.100 M increases the rate constant to 126. In i9.9% and 91.0% acetic acid, the rates for hydrogen chloride and hydrogen bromide are larg-er than but comparable to that for perchloric acid; the specific effect of the halogen acids must be separated from their catalytic effect as acids. T h e data here reported for the rate of broxnination can be expressed by the equation u =
- (d(Bry)/dt)
=
k,,&(deoxy) = ho(deoxy)[kh
+ kx(R/IX)I
(3)
where k.4 is the rate constant for perchloric acid and kx is the specific catalytic constant for the halides present ( i e . , both HX and LiX). Figure 3 shows the data for 97.9% and W0yoacetic acid; the numerical results for 91.0yo and 79.97, acetic acid are presented in Tables VI and V I I . Solvent Isotope Eff ect.-The reaction catalyzed b y 0.0367 M DC1 in "97.97," DOAc-D20 proceeds about
Sept. 5, 1963
BROMINATION O F
KINETICS FOR Acid
Mole/l
HC104
0 100 100 100 200 300
Mole/]
LiC104 SaC104 TaC104
0 025 050 100 100 100 100 010
2593
TABLE I11 BROMINATION OF DEOXVBENZOIN IN 97.9Yc ACETIC ACID
THE
Salt
0 020 020 025 025 025
HBr
DEOXYBENZOIN
Remarks
0 200 200 100
LiC104
0 080
LiClO4
0 075
LiBr LiC104
0 075 0 050
LiClO4 LiC104
98 0% H 0 . h 98 0c/c HOAC 98 07c H 0 . i ~ 98 0(,; HOXc 0 075 HC104, 98 0 % HOAc 98 0 % H 0 . 1 ~ 98 0 % HOXc 98 070 HOAC 98 07, H 0 . 4 ~ 98 070 HOA4c 0 , 0 9 0 AM HC104, 98 0 % HOA4ca
0 100 0 200
IOjkl,
sec
2 3 3 7 14
64 25 14 56 3
23 7 26 9 37
2 12
29 34 126 114 80 65 15
-1
7 63 3 3 5
9
8 6
ho
105kl/h
27 5 41 7
0 096 0 078
0 086
166
4 4 6 6 31
58 8 03 3 0
5 06 1 48 4 4 1 52 1 20
6 14 31 27 32 41 31
3
4 6 2 43 4 06 4 16 2 47
1 5
7 7
0
1 6 0 53
0 0166 NaC104 0 0834 2 53 8 72 0330 KaClO4 0 0668 0323 16 6 0330 LiC104 0 0670 8 77 0330 LiCl 19 5 0 0670 0999 59 3 0250 LiClOl til 0 07SO 3 8 1 0250 LiCl 0250 0 050 LiClO4 10 5 3 5 3 0250 LiCl 16 5 0750 3 0 5 LiCI 050 050 34 2 5 6 3 100 9 8 . 0 % HOAc 70 5 13 7 5 100 67 9 13 1 5 100 LiClOa 0 200 25 3 22 4 1 0367 DCI in DOAc - D 2 0 57 8 Set of three experiments, with initial concentrations of bromine of 0.00065, 0.0020, arid 0.0086; average deviation 5 % HCl
a
6 0 6 4 13 18 62
TABLE II' KINETICSFOR
DEOXYBENZO~S I N 91.070 ACETICACID
T H E BROMISATION OF
105kl. Acid
Mole/l.
HC104
2.100 100 300 0300 0300 100 100 100 100
HCI
Salt
Molejl.
LiClOI
0.200
LiClOl LiCl
0.070 0 070
LiClOd LiCl
0.200 0.200
Remarks
0.20034 HClOi
309
-1
0.504 il 010 1 87 0 387 0 707 2 78 2.00 4 44 6 70
15 1 TAX%
KINETXS FOR
sec
ho
lOEkl/ho
1.05
1.57 5 60
0 48 .39 .33
1 08 1 51 1 46 5.5
2.59 1.33 3.04 12
5 13
2.05
L'
THE B R O M I S L T 1 3 S CT
DXXYBESZOIS I S 79.97,
ACETICACID Acid
HCIO,
HBr HCI
Yiole/l
0 I00 100 1ii0 300 100 100 100 300
Salt
Mole/1
LiCIO, LiC10,
0 160 0 200
LICIO~
o mn
10.' sec - 1
0 240 265 281 7x3 2 78 333 368 1 64
ho
105ki/ho
0 225
1 07
0 333 1 02 0 226 226 312 1 00
0 0 1 1 1 1
845 765 23 45 15 64
2.9 times as fast as the corresponding reaction (interpolated from the data of Table 111) in HOAc-H20. Salt Effects.-The data in Tables 111, IV, and V show that the rate of the reaction catalyzed by perchloric acid increases with increasing salt concentration, b u t so
Time, min
Fig. 1.-Bromination of deoxybenzoin in 97.9%; acetic acid, catalyzed by perchloric acid. The observed points are shown by the open circles; the d a t a corrected for autocatalysis (see Appendix) are represer,ted by triangles and by t h e solid line.
2394
RONALD F. W. CIECIUCH A N D F. H . WESTHEIMER
Vol. 85
with the activity (rather than the molarity of water) and have offered the approximate equation -ffo
=
log (HT)- n log ~
H
-k~ log O
( . ~ I ~ H - / ~ I H + () 4 )
where n is the average hydration number of the proton and f~ and ~ I H + represent the activity coefficients of the Hammett indicator and its conjugate acid. Our data and those of \%?berg a n d Evans are plotted in Fig. 2 against the logarithm of the activity of waterz0 in solutions of aqueous acetic acid. -1good straight line of slope three is obtained, suggesting that the proton holds three more water molecules than the conjugate acids, IH+, of the indicator bases used in this study. Mechanism of the Bromination Catalyzed by Perchloric Acid.-The rate of the bromination of deoxybenzoin, catalyzed by perchloric acid, is well correlated by the expression L' = log "$ p':
P
Fig. 2.-The dependence of - H o on the activity of water for 0.300 .If HCIO, in acetic acid-water mixtures. The circles are from the data of Wiherg and Evans, the triangles from our data,
TABLE 1.1 PARTIAL RATESI N 9lC)il ACETIC~ C I DFOR BROMINATION CATALYZED SPECIFICALLY BY HYDROGEN CHLORIDE Sybtem
0.100 -1f HCI 100 .IT HC1 200 M LiC10,
lOjhokx(SIX)
2.30
lOjkX(S1X)
IO~kX
2 13 0 93
21 3" 9.3
(5)
kH,ohn(deoxy)(H20)
TABLE VI11 KIXETIC DATA FOR BROMINATIOSCATALYZED BY 0,100 M PERCHLORIC ACID CHzO,
W t . Yo HOAc
97 9 91.0 79 9
105kl, sec
-1
lOjkl/ho
2 64
0.096 0 48 1.07
0,504 0.240
molesll. (in solvent)
1.22
5 30 11 88
IOskI/ (ha)(CH20)
0 079 ,091 090
T h e data agree then with eq. 1 and 2, where B is a water molecule. T h e prior equilibrium should be fast, reversible, and unfavorable to protonation; the most 100 ;M HC1 3.96 2.72 9 1 acid solutions employed in this study had an H,]value 200 dl LiCl of -2, whereas the pK of deoxybenzoin is unlikely t o 100 Jf H C 1 4.88 0.89 8 9 be very different from -6.15, the value for aceto,200 .If HClOg phenone.*l The transition state and the starting ma300 .If HC1 13 4 2 62 8.7 tefials each have one position charge but the somewhat " Formal ionic strength, 0.100 Jf, greater delocalization of the charge of the transition state should lead to a small negative salt effect, as obTABLE 1-11 PARTIALRATES I N 79.9:; ACETIC FOR BROMINATIOS served for k h,. Nature of the Ionic Species in 79.97,-97.9y0 Acetic CATALYZED S P E C I F I C A L L Y BY HYDROGEN CHLORIDE Acid.-Kolthoff and Willrnanz2 concluded from conSystem IO~hokx(XI)o 10SkX(hIX) 105k~ ductivity measurements t h a t perchloric acid is com0 , 1(10 ~lfHCI 0 093 0 41 4 1" pletely ionized to ion-pairs in glacial acetic acid and 100 JI HC1 104 33 3.3 that these ion-pairs are slightly dissociated; hydro200 JI LiCIOI chloric acid is only slightly ionized in the anhydrous 300 121 HCI 875 88 2 9 acid, but ionized and (like perchloric acid) partially Formal ionic strength, 0.100 *If. dissociated in the presence of 1-27, water. More recently, TViberg and Evans" measured the does h,,, so that the quotient k h, decreases slightly as conductivity cf sodium perchlorate in 91% and 9570 salt is added. In sharp contrast, the rates in the presacetic acid; treating their data by Shedlovsky's equaence of hydrochloric a n d hydrobromic acids are subject t i o n Z 3they found dissociation constants for the ionto a large negative salt effect; since salt increases the a n d 4.5 x 1W3, respectively; pairs of 3.23 x value of ho, the salt effect upon the quotient, k.'h,, the dissociation constant for the salt in %yo acetic is even more pronounced. Thus the addition of 0.2 J I acid is large. These are thermodynamic dissociation lithium perchlorate to 1~.100.lI hydrogen chloride in constants; in order to calculate concentrations, apST.SC/b acetic acid decreases the quotient, 105k ho, propriate activity coefficients must be taken into acfrom .i.13 to 1.62; the same change in 91.rIycacid decount. &-iberg and Evans estimated, for exaniple, creases the quotient from 2.39 to 1.33. that 0.0017 LI sodium perchlorate is 73yo dissociated Discussion in 95yo acetic acid as solvent and almost completely 130 Function.-As n'iberg and Evans have pointed dissociated in 91% acid, but of course the extent of the values of H o increase linear!y with the condissociation will be much less a t the concentrations centration of perchloric acid, provided t h a t a constant here used. Their data, as well as those of others,IX2 4 formal ionic strength is maintained. This applies to suggest that even in solutions which are (1,(11-0.3 our data in 79.9yc and 91 .ilycacetic acid, and approxidissociation is extensive in 79.9y0 acetic acid as solvent mately t o our data in 9T.STo acetic acid, as well as (20) R S. H a n s e n , F A . Miller, and S L) Christian, J P h y s Chrm , 69, to their data in 85 91yo,and 95% acetic acid. 391 (1955) Ti-iberg and Evans correlated their data with the (21) R S t e w a r t a n d K. Yates, J A m Chem. S o c , 80, 635-5 (1058) molarity of water in the acetic acid solvent. Bascombe ( 2 2 ) I . hl. Kolthoff a n d A Willman, ibid., 6 6 , 1007 (1934) (23) T Shedlovsky, J . F r a n k l i n I n s l , 225, 739 ( 1 9 3 8 ) , R M Fuoss a n d and have suggested t h a t Ho should be correlated 1 41
( I Y ) K S Bascomhe a n d R . P Bell, Discussioits F Q V U ~ Q SO Y L ,24, 158 I l R Z 7 ) . c f . P A H \Vyatt, ibid , 24, 162 (1957)
T . Shedlovsky, J . A m Chem S O L ,71, 1496 (1949) (24) S I M Jones a n d E . Griswold, i b i d , 7 6 , 3247 ( 1 9 5 4 ) ; S Bruckenstein a n d I A 1 Kolthoff, ibid , 78, 2974 (19.56)
Sept. 5 , 1963
BROMINATION O F
and important in 9176 acid, but the ion-pairs present. in !Xycacetic acid are not extensively dissociated. Mechanism of Bromination in the Presence of HC1 and HBr.-The data for the broniination of deoxybenzoin show a catalytic role for HC1 and H B r in 9i.9yc, 91.0yc, and 79.9% acetic acid solutions, b u t the specific effect of the halogen acids decreases with increasing water content of the solvent. The rates are independent of the concentration of bromine and vary linearly with the concentration of deoxybenzoin, with ho, and with the stoichionietric concentration of halide present in the solution. T h e data correlated by eq. 3 and 5 lead t o the catalytic constants presented in Table I X ; see Fig. 3. TABLE
CATALYTIC COSSTASTS FOR
DEOXYBENZOIN
2593
-HCI i n ,t 97.9% HOAc
-
5.0
HBr i n 98.0% HOAC 4.00
2z
\
J 3.0 -
In
0
Ix
>ACID-CATALYZED
ESOLIZATIOS OF
DEOXURENZOIS 7--
I," .I4
0.1 3 lOjkci1 3 105kBp.1 ~1 Formal ionic strength. l@kH,o
Snivmt, wt.
7c HOAc----
79 9 %
91 0%
0 090 0 068
0.081 0 068
97.97,
0 08 0.067 58 15 46'
4.1 21 3 3.1 9.0 1.7 .. 98.0'; H O h c .
These data are most simply explained by the scheme shown in eq. 1 and 2, where the base B is halide ion. The large negative salt effect is consistent with this formulaticn and with the transition state H
H
/
xFurther, an increase in water content should diminish the basicity of halide ions and therefore their catalytic efficiency, in conformity with our ohervations. T h e finding t h a t chloride and bromide ions pzrticipate a s bases, but only in dry or nearly dry solvents, is consistent with previous knowledge.zi ?R In particular, Cromwell. Kevill, and their collaborators have shown that these ions, in solution in acetonitrile, function as bases in the dehydrohalogenation of 2-benzyl-2-bromo-4,4-dimethyl-1-tetralin, and of other ketones.?' Catalysis by hydrogen bromide for the racemization of optically active ketoneszxin glacial acetic acid has also been observed. (Catalysis by the relatively basic fluoride ionz9 in enolization is of course well known.) Similar increase in the effective activity of cations with decreasin: hydration accounts for the increase in the h,, functionI9 and for the increase in the rate of mercuration of benzene3" which accompanies a decrease in the activity of water (2:) S Winstein. I, G Savedoff. S S m i t h , I I1 I< Stevens, a n d J S G:ill, 7'r!rah~iiroii/.r!!vi.s, No. 9 , 24 (1960) (26) A J P a r k e r , 0uoi.f Rrz ! l . n n < I - n ) . 16, 103 (1962) ( 2 7 ~I) S Kevill a n d 1. H C r u r n w l l . .l Aiii C h f i n S o < , 8 3 , :
