B. PERLMUTTER-HAYMAN AND A. PERSKY
276
Because of its unfavorable charge distribution VI seems unlikely as a reaction intermediate. Yet TV cannot be formed unless VI also is formed, and evidence is presented in the Experimental section that biacetyl can be produced directly by ozonization of hexamethylbenzene (either in carbon tetrachloride or acetic acid) without hydrolysis of the reaction products. Products V and VI may be 0
111
20s
0
0
Ill1
I!
+R-C-C-R
IV
0-0-
R-C-C-R
I +
v
o-o-
-0-0
I
R-C-C-R
1
+ + 1-1
presumed to undergo polymerization to form polymeric peroxides or, in acetic acid, to form hydroperoxides by reaction with the solvent. On hydrolysis under reducing conditions these compounds should break down to diketone IV. I n one variation of the suggestions concerning the reaction mechanism intermediate I11 may be pictured as undergoing polymerization or re-
[CONTRIBUTION FROM THE
Vol. 82
action with hydroxylic solvent before its double bonds are attacked by ozone. Product V is very likely susceptible to rearmngeIt is ment to the isomeric anhydride,?YIV. therefore reasonable that, as is the case, acetic acid should be obtained by hydrolysis of the ozonization products of hexamethylbenzene and mesitylene in carbon tetrachloride and that propionic acid should be obtained similarly from hexaethylbenzene. It is noteworthy, however, that acetic acid has been detected, by gas-liquid chromatography, as an immediate product of reaction of mesitylene and ozone in carbon tetrachloride. I t is conceivable that this acid was formed by thermal decomposition of a reaction product a t the of the relatively high temperature (-100’) chromatograph column. Acknowledgment.-The authors are indebted to the National Science Foundation for a grant in support of this research. (28) R. Criegee and hl. Lederer, A n n . , 6 8 3 , 29 (1953).
DAVIS,CALIFORNIA
DEPARTMENT OF PHYSICAL
CHEMISTRY,
HEBREW UNIVERSITY]
The Kinetics of the Oxidation of D-Glucose by Bromine and by Hypobromous Acid BY B. PERLMUTTER-HAYMAN AND A. PERSKY RECEIVED JULY 6, 1959 The kinetics of the oxidation of D-glucose by aqueous bromine solutions has been reinvestigated. From the dependence of the rate of reaction on bromide concentration i t was concluded t h a t molecular bromine is the oxidizing agent, tribromide and hypobromous acid making only negligible contributions. This was confirmed by t h e slowness of the reaction of hypobromous acid with glucose. T h e dependence of the reaction rate on pH is consistent with t h e assumption t h a t t h e anionic form of glucose is oxidized much faster than glucose itself.
The oxidation of D-glucose by bromine has been the subject of several investigations. 1-4 Bunzel and Mathews2 studied the kinetics in dilute acid solutions and interpreted their results on the assumption that OH- takes part in the rate-determining step. Isbell and Pigmans carried out an extensive study at the pH of a BaCOa-CO2 mixture. The dependence of the rate on bromide concentration pointed to free bromine as the only oxidizing agent. The primary product was the &lactone of gluconic acid. Furthermore, P-D-glucose was found to react about 35 times faster than the a-form. More recently4 i t was concluded from the strong increase of the rate with increasing pH that hypobromous acid is the oxidizing agent in bromine water. I n view of these discrepancies and as a continuation of our p r ~ g r a m ~to- ~investigate the specific oxidizing properties of bromine and of hypobro(1) J. W. Green, Advances i n Carbohydrate C h c m . , 3, 129 (1918). ( 2 ) H. H. Bunzel and A. P. Mathews, THISJOURNAL, 81, 461 (1909). (3) H. S. Isbell and W. Pigman, J . Research N a t l . Bur. Sfandnrds, 10, 337 (1933); 18, 141 (1937). (4) K. C . Grover and R. C. Mehrotra, Z. phyrik. Chem. ( F r a n k f u r t ) , 14, 345 (1958). ( 5 ) L. Farkas, B . Perlmutter and 0. Schschtar, 1’1iI3 JOURNAL, 71, 2829 (1949). (6) Y.Knoller and B. Perlmutter-Hayman, i b i d . . 7 7 , 3212 (1955). (7) L. Binoun and B. Perlmutter-Havman, Bnli. Reseorch C o u ~ i l Israel, AS, 52 (1955).
mous acid i t seemed interesting to reinvestigate the oxidation of D-glucose by bromine water, measuring both the dependence on bromide concentration and on pH. A direct measurement of the rate of reaction between hypobromous acid and glucose also seemed desirable. Experimental T h e experiments were carried out at O ” , in glass-stoppered erlenmeyer flasks immersed in a Dewar bottle containing ice and water. T h e progress of the reaction was followed by iodometric titration, using 0.01 iV thiosulfate solution. The PH was measured at the beginning and a t the end of each reaction, usiiig a Beckman Model G pH-meter. Unless otherwise stated, it changed by less than 0.1 unit. The buffers used were SazHP04-NaH2POa, sodium acetateacetic acid, SaHnPOb-phosphoric acid, and appropriate concentrations of sulfuric acid, thus covering the desired p H range. Hypobromous acid solutions were prepared as usual .6 All the reagents were “Ana1ar”-grade. The a-o-glucosc was Hopkin and Williams, and the p-n-glucose Nutritional Biochemical Corporation. The concentration of the oxidizing agent was between ( 5 to 10) X M , and that of D-glucose was always in ercess and was varied between 0.025 and 0.500 A f according to conditions, in order to get convenient reaction rates. When experiments were repeated, their rate constants agreed within less than =+=37,. Rate constants are expressed in mole-’ 1. Inin.-’.
Results 1. The Oxidation of a- and of P-D-Glucose,
and the Influence of Anomerization.-At
pH 4.95,
OXIDATION OF D-GLUCOSE BY BROMINE AND HYPOBROMOUS ACID
Jan. 20, 1960
277
a freshly prepared solution of a-D-glucose, and a freshly prepared solution of /3-D-glucose were allowed to react with bromine. A plot of log (Z, - .)/(a - x ) ( b and a are the initial molarities of glucose and of oxidizing agent, respectively, and x that of the product) against time yielded a curve whose slope increased with time in the first case, but decreased in the second. This confirms the greater reactivity of the @-form, and a t the same time shows that the rate of anomerization is comparable to that of oxidation of @-D-glucose. We can assume that within experimental accuracy the rate equation may be written dxldt = k8[fl-~-glucose](a- x )
(1)
It would be difficult, however, to calculate the value of [P-D-glucose] a t every stage of a given run, since oxidation and anomerization follow different orders, and both are affected by a change in pH, but in different ways.8a Instead, when starting with an anomerically equilibrated solution (containing the CY- and @-forms3in the ratio 0.37 to 0.63), we considered the two extremes (a) No transformation from a to p-form takes place while the oxidation is measured and equation 1 becomes dx/dt = kB(0.63b
- X ) ( U - X)
(2)
(b) Anomerization is fast in comparison with oxidation, when equation 1becomes dx/dt = 0.63kB(b
-
X)(U
-X)
(3)
According to conditions (PH, nature of oxidant) we used equation (2) or (3). When in doubt, we calculated our results according to both. Since the ratio of the two constants is ( b - 1.6x)/(b x) and b is always in large excess, the difference between the results obtained from the two assumptions is a t most lo%, and usually much smaller, and in no case was found to obscure the conclusions which can be drawn from our results. 2. Dependence of the Rate of Reaction on Bromide Concentration.-If we consider the three oxidizing species to be Br2, Bra- and HOBr, the meaning of k S in equation 1becomes
-
+
+
+
kp = ~B,&K,/(K, [Br-I) k B r r [Br-I/(K3 IBr-1) kHoBr K3Khl(K3 [Br-l)[Br-l IH+l (4)
+
+
where K3 is the dissociation constant of tribromide, Kh the hydrolysis constant of bromine, and use is made of the fact that under our experimental conditions [HOBrl
