5659
REACTION OF HYDROGEN PEROXIDE AND BENZENEBORONIC ACID
ORGANIC AND BIOLOGICAL CHEMISTRY [CONTRIBUTION FROM THE
DEPARTMENT OF CHEMISTRY, UNIVERSITY
O F N E W HAMPSHIRE]
Electrophilic Displacement Reactions. IX. Effects of Substituents on Rates of Reactions between Hydrogen Peroxide and Benzeneboronic BY HENRYG. KUIVILA AND ALBERTG. ARMOUR RECEIVED APRIL 8, 1957 The effects of meta and para substituents on the rates of reactions between benzeneboronic acid and hydrogen peroxide have been measured. Ten substituents and five different reaction paths are included in the study. The Hammett equation correlates the rates for each reaction path suggesting that the mechanisms involved do not have in detail the characteristics of aromatic electrophilic substitution in the slow steps.
The reaction between hydrogen peroxide and benzeneboronic acid in aqueous solutions produces quantitative yields of phenol and boric acid. Formally this would appear to be either a homolytic or an electrophilic substitution in the benzene ring, in which the boronic acid group rather than the usual hydrogen is replaced. Homolytic reactions involving hydrogen peroxide and organic compounds are usually c ~ m p l e x . ~Electrophilic hydroxylation of arenes is unknown. These considerations led to a series of investigations on the kinetics of the reaction between hydrogen peroxide and benzeneboronic acid5 which have revealed that the reaction proceeds by several mechanisms each of which can be shown to predominate under appropriate conditions. Those which were studied in the present investigation are depicted belowk in postulated forms which are consistent with the kinetic information now availab1e6"sb though not necessarily unique in this regard. Ph-B-OH
I
+ HOOH
OH I OH
H,O
+ Ph-B-OH
Hf
OOH I I1 I11
+
-
[ ] I
Ph-B-OH
(1)
OOH I11
+ H' +HzO + ku
kl
111 +HOI11
+ Ph-B-OH
HzO
I
+
OH
kz
+IV
acid
I1
+ -0-B-Ph
catalysis
v
I
(4)
OH (5)
The rates of all of these reactions depend on the concentrations of both peroxide and boronic acid. It is assumed that the equilibria (1) are maintained a t all times. Reaction 2 represents a pH-independent reaction leading to monophenyl borate which is presumed to hydrolyze rapidly to phenol and boric acid. This is the predominating reaction in the p H range 1-3. Reaction 3 has as its distinguishing characteristic a first power dependence of rate on hydroxide ion concentration" Reaction 4 shows the same p H dependence but is complicated by the fact that its rate depends on the square of the boronic acid concentration. Since it bas the same stoichiometry as the other reactions, one molecule of boronic acid must function as a catalyst. The acid-catalyzed reaction5 shows different characteristics in perchloric, sulfuric and phosphoric acid@ and is discussed separately below.
Results and Discussion
IV OH I Ph-0-B-OH V
(2)
( I ) Preceding paper in this series, H. G . Kuivila and I,. E. Benjamin, THIS J O U R N A L , 77,4834 (1955). (2) Substantial support of this work by the Office of Naval Research is gratefully acknowledged. (3) Taken in part from the senior thesis of A. G. Armour, June, 1955. Presented in part a t the 130th Meeting of the American Chemical Society, Atlantic City, N. J., September, 1956. (4) For example the iron salt catalyzed reaction between benzene and hydrogen peroxide produces a complex mixture of products including phenol and other products of f u r t h e r oxidation (A. Kraus, Rocsniki Chem., 27, 9 (1953); C. A , , 48, 5624 (1954)). (5) (a) H . G. Kuivila, THISJ O U R N A L76, , 870 (1954); (b) H. G. Kuivila, ibid., 7 7 , 4014 (1955); H. G . Kuivila and R . A. Wiles, i b i d . , 77, 4830 (1955); (c) H . G. Kuivila, paper presented a t the 130th Meeting of the American Chemical Society, Atlantic City, N. J., September 20, 1956, Abstracts p. 58-0.
In addition to benzeneboronic acid, four para and five meta substituted acids were included in this investigation. The experimental methods used have been described earlier.ea*b The pH-dependent Reactions.-A minimum of four runs with varying initial boronic acid concentration were carried out in 25% ethanol with each (6) T h e structure I11 for t h e intermediate is preferred over a conVIII). T h e p a t h ventional conjugate base of I1 (e.g., Ph-B-0-,
- -
I
OOH
IV V appears t o be a more probable one in aqueous solution t h a n one proceeding from VI11 t o PhOB=O with loss of a hydroxide ion. Further evidence for 111 comes from the effects of ortho chloro a n d methyl groups on the acidity of benzeneboronic acid. T h e acidweakening effect of these groups has been interpreted as evidence for a quadricovalent anionic species P h B -(OH)a rather t h a n t h e tercovalent species P h B ( 0 H ) O - ; D. H. McDaniel and H. C. Brown, THIS JOURNAL, 77, 3756 (1955).
I11
I I
0
thus obtained versus initial boronic acid coiiceiitration was then tnatle. The intercept obtained by the method of least squares provided k l , the rate constant for the reaction first order in boronic acid, and the slope provided the value of kz, the rate constant for the reaction second ordcr in boronic acid. These plots are shown in Figs. 1 and 2, and the rate constants are listed in Table I. I t remains to demonstrate that the uncatalyzed reaction makes no significant contribution to the rate constants obtained in this way. To do so two runs were made with each boronic acid a t pH 5.90 with different initial boronic acid concentrations (usually 0.015 and 0.030 molar). I t was shown iii an earlier paper5" that the rate constants kl and k? increase linearly with hydroxide ion activity. .Is a result the over-all rate constant, [ k l k 2 ('h-13JC)H)2)]U, at pH 5.90 can bc coiiiputed froin tlic values obtained a t $11 5.4%. When this was done the observed and computed values agrced within 5yo. If the reaction characterized by k , m r e making a significant contribution to the rate a t pH 5.42, the absolute magnitude of this contribution would be the same a t pH 5.90, but this would constitute a smaller fraction of the total rate constant. As a result the observed rate constant would be smaller than the calculated one. Since this was not the case it can be concluded that only the two @H dependent reactions were important in this range. (The values of k , obtained in this work cannot be used in this case because they were obtained in aqueous solution, whereas kl and f i r were obtained in 25% aqueous ethanol. In this solvent rates are about a tenth as fast as in water a t the same @H.) The values oE hi and k2 show very little variation. Hamniett plots as shown in Fig. 2 have 1 sriuarcs slopes (p-values) of 0.071 for kl and (I.% for k,. This lack of dependence of over-all ratc on the electron density a t the carlmi bearing the boron is undoubtedly an indication of conil)lcsity iii the reaction. The concentration of I V a t :t givcn PH obviously will increase as the elcetron density on the boron decreases (positive p ) . 0 1 1 the other hand, ionic clcavage of the peroxide bond with departure of a hydroxide ion as well as migration of the aryl group to oxygen should be favored by high electron density (negative p ) . I t cecn1s clear that the formation ot TI1 and its subsequent transformation to products have essentially c q u d I J U t opposite electronic demands. I n recent years it has beconie apparent that tile Hammett equation does not correlate the rates of reactions involving electrophilic substitution iii the benzene ring when the original U-values are used.' The greatest deviation observed in the broiiiiiiolysis of benzeneboronic acids is due to Lhe p-inethoxy group.7c In the present data this s u b stituent appears to be "iioriiial." Therefore it seems quite probable that carbon--oxygen bond fur-
+
/
2 .01
.02
.03
.04
.R5
R B tOH12.
7
6
103k.
5
4
3
Figs. 1 ;tiid 2.---Plots of second-ortlcr rate coiistaiits 1's. initial benzenebororiic acid conceiitmtions. A-umbers correspond t o those in Table I.
boronic acid in an acetate buffer a t @H5.45 with ionic strength adjuste(1 to 0.02 with sodium perchlorate. .I plot C J ~the .ipparent rate constants
( 7 ) (a) J. D . Roberts, J . K . Sanford, 1'. L, 1,Sixma, €1. C e r f o n l a i n R.Zagt, THISJ O C R N A I . , 76, 4525 (1954). (b) C . \T. McGar?, Jr., Y. Okamoto a n d H. C. Brown, ibid., 77, 3037 (1055). (c) €1. G. Kuibila a n d A . R . Ilendrickson, i b i d . , 7 4 , BO68 (1952); H. G. Kuivila and I, E . B e n j a m i n , i b i L , 77, 4S34 (1952). ( ( I ) See H. C. B r o w n and Y. Okamoto, ibi,/,,79, 191'3 ( 1 9 5 7 ) . concerning sigma conitants and
fear n r i i l n a t i c eli~ctr
