The Decomposition of Aqueous Sodium Bromite C. L. LEE AND M. W. LISTER Department of Chemistry, University of Toronto, Toronto 181, Ontario Received March 24, 1971 An investigation of the rate of decomposition of aqueous sodium bromite is described. The main reactions are: 2BrOZ- + BrO- + Br03- and Br- + Br0,- + 2Br0-, followed by BrO- + BrOz- + Br- + Br03-. The rate constants and activation parameters of the decomposition reactions are reported. As the second reaction is the reverse of that previously studied in the decomposition of sodium hypobromite, its equilibrium constant (at various temperatures) can be obtained, and hence the thermodynamic properties of aqueous bromite ions. These are briefly compared with the properties of similar ions containing bromine or chlorine. On dkcrit les recherches entreprises en vue de determiner la vitesse de decomposition d'une solution aqueuse de bromite de sodium. Les reactions principales sont: 2Br02- + BrO- + Br03- et Br- + Br02- +2Br0- puis,bro- + Br0,- + Br- + Br03-. On rapporte les constantes de vitesse et les parametres d'activation pour ces reactions de dccomposition. Puisque la deuxieme reaction est I'inverse de celle ktudike preckdemment pour la dkcomposition de l'hypobromite de sodium, sa constante d'equilibre (B diffkrente temperature) peut Btre obtenue et par consequent les proprietks thermodynamique des ions bromites en solution aqueuse. Celles-ci sont brievement comparkes avec les proprietes d'ions similaires contenant du brome ou du chlore. Canadian Journal of Chemistry, 49, 2822 (1971) This paper is concerned with the kinetic and, to a lesser extent, thermodynamic stability of bromite ions in aqueous solution. The results permit some comparisons to be made amongst the series of bromine oxvions. and these in turn are briefly compared witi the gnalogous chlorine compounds. Bromite ions have been relatively little investigated, but a study of their kinetic stability was made by Engel, Oplatka, and Perlmutter-Hayman (I) as part of their research on hypobromite ions. We have examined their decomposition under somewhat different conditions and over a range of temperature, and a comparison with their results will be made later in this paper. A fairly extensive investigation of the conditions of formation of bromite from hypobromite ions, and of the preparation of sodium bromite, has been made by Breiss (2). The present work was done at suficiently high ph, so that protonated species (particularly HOBr) would be present only at very low concentrations. Acidified sodium bromite decomposes rapidly, so presumably other reactions, not investigated in the present work, are then important. Experimental Sodium Bronzite Several sources were used for the reagent. Firstly, it was prepared by the method of Breiss (2), which gave crystals of the trihydrate, NaBrOz.3HZ0. Secondly, crystals of the same composition were made available by La SocietB d'~tudes Chimiques pour 1'Industrie et I'Agriculture, of Argenteuil, France, to whom we are much indebted. Thirdly, a solution of the compound was obtained from the Olin Mathieson Chemical Corporation of New York, and this solution was evaporated and the sodium bromite recrystallized. Most of the experiments used material from the first two sources. Since the solid showed a tendency to decompose, the crystals were dissolved in dilute sodium hydroxide solution to give a stock solution that was reasonably stable, and which was diluted as required for the experiments. Sodium hydroxide and bromine, used in the preparation of sodium bromite, were of analytical reagent grade, as were the sodium bromide and sodium perchlorate which were added to the solutions in the kinetic experiments. Apparatlrs In the kinetic runs, samples of the appropriate solution were sealed in Pyrex glass ampoules, and immersed in a thermostat for known lengths of time. The thermostat contained silicone oil, controlled to f 0.02 "C. The solutions contained sodium bromite, sodium. hydroxide, sodium bromide (not added in some runs), and sodium perchlorate to adjust the ionic strength to a value of 0.50. At suitable times the ampoules were removed, rapidly cooled, and the contents analyzed. Analysis of Sol~rtions The solutions in general contained sodium bromide, hypobromite, bromite, and bromate, sodium hydroxide and possibly a little carbonate. The sodium perchlorate present was assumed not to react, so its concentration was known from the amount added initially. The sodium bromide concentration was also known from the amount put in initially, plus the amount formed in the reaction (see below). The sodiilm bromite and hypobromite were estimated
LEE AND LISTER: DECOMPOSlTlON OF AQUEOUS SODIUM BROMITE 2823 spectrophotometrically with a Beckmann DU Spectrophotometer. Their extinction coefficients were measured from the absorbances of freshly prepared alkaline solutions; in the case of sodium hypobromite this was made in the usual way by adding bromine to cold sodium hydroxide solution. It was checked by polarography that these solutions did not contain appreciable amounts of the other ion (e.g. hypobromite in the bromite solution). The polarographic experiments would have detected 0.5% of the other ion as impurity. The extinction coefficients found were as follows: Extinction coefficient Wavelength (nm) NaBrO NaBrO, The solutions used to obtain these extinction coefficients were standardized by titration with arsenious oxide at a ph regulated by excess sodium bicarbonate (3). The arsenious oxide reacts with both sodium hypobromite and bromite. In the kinetic runs measurements were made at 296 and 380 mp, and it was checked that the other species present did not contribute appreciably to the absorbances at these wavelengths. It was found that the total oxidizing normality, measured by the iodine liberated on adding excess potassium iodide and sulfuric acid, did not change during a run, so presumably oxygen is not evolved. Consequently changes in bromide and bromate concentrations could be calculated from the mass balance of bromine and oxygen atoms. The absence of oxygen evolution was confirmed (within moderate limits) by the fact that the ampoules did not develop increased pressure during a run. The concentration of sodium hydroxide was found by adding excess hydrogen peroxide to a sample, boiling to destroy the excess, followed by titration with standard acid. The hydrogen peroxide reacts with both sodium bromite and hypobromite. Mechanism and Calculation of Rate Constants Ellgel, Oplatka, and Perlmutter-Hayman (1) proposed a mechanism for the decomposition of sodium bromite. Our results fit their mechanism, though with some difference in the relative contributions of the various reactions. They found a dependence of the rate on ph,while under the conditions of the present experiments (0.06 to 0.18 M sodium hydroxide), no change in rate constant was found with varying [OH-]. They interpreted their results as showing that at the lower ph of most of their experiments protonated species are reacting. Presumably these species are present in such low concentrations under the conditions of our experiments as to have no effect. However, it should be mentioned that in the somewhat analogous decomposition of hyperbromite ions, the rate increases with [OH-], and the rate in fact passes through a minimum at about the ph values we have used (3, 4). It is possible, therefore, that a similar minimum occurs for bromite ions, and that the apparent lack of ph dependence arises because our experiments fall in the region of this minimum. We can therefore re-write their mechanism, and propose that the following reactions occur: Reaction 4 is the reverse of reaction 3, and is the initial step in the decomposition of hypobromite ions. Engel, Oplatka, and Perlmutter-Hayman found that, at 25 "C and at the concentrations they used, reaction 3 was more important than reaction 1. However, in our experiments, both reaction paths made appreciable contributions, with reaction 1 predominating. The mechanism above gives rate equations which could not be integrated, but since k2 is relatively high, the steady state approximation was made that d[bro-]/dt = 0. From this, putting [Br-] = x, [BrO-] = y, and [BrO,-] = z, we get Values of k4 are known (1, 4, 5), and also of k2 (5), and these combined with the values of k, and k, obtained below show that the term 2k4y2 is small enough to be neglected. Hence y = (klz + 2k3x)/k2. With this approximation, the rate equations can be integrated to give (with P = xolzo)
2824 CANADIAN JOURNAL OF ZHEMISTRY. VOL. 49, 1971 A plot of l/z against t should give an exponential curve of increasing slope, but it was found that our experimental results were such that the curvature was very slight during the decomposition of the first 20-25% of the sodium bromite. The slope d(l/z)/dt is and the initial slope is 3(k3j3 + k,). In practice, we took an average slope, (l/z - l/zo)/t, over the first part of the reaction. This average slope can be shown to be So (ect - l)/ct,where Sois the initial slope, and c = k3(3p + l)z,. The average slope was therefore corrected to give the initial slope by this expression. This involves a knowledge of k,, which was obtained as described below. The process of correction was repeated until the value of k, used to obtain So was the same as the final value of k,. Finally So was plotted against P, which gave a linear plot, though with some scatter. The best line through the points was obtained by least mean squares, and k, and k, were obtained from this line, since So = 3(k,P + k,). The corrections from the average to the initial slopes were mostly under 574, and in only one run above 10%. After the rate constants had been obtained, the validity of the steady state approximation was checked as follows. Firstly, the concentrations of hypobromite ion, which are also obtainable from the measured absorbances, were compared with the steady state values given by (k,z + 2k,x)/k,. However, [BrO-] was always so low, that the most that can be said is that the values were of the right order of magnitude. In a typical run [BrO,-] = 0.05 M, and the calculated [BrO-] = 0.0004 M. As the extinction coefficients are not very different, it follows that a very small error in absorbance gives a large error in [BrO-1. Secondly, the differential equations were integrated numerically, with the assumption that initially [BrO-] = 0. It was found that the steady state concentration of hypobromite had developed rapidly compared with the rate of decomposition of the bromite. 95% of the steady state [BrO-] was reached during the first 1 to 23% of the bromite decomposition, the exact amount depending chiefly on the concentration of bromide ions. At the same time the slope, which starts low if [BrO-] is initially zero, had risen to 97 to 98% of its steady state value. The first sample was taken after a time that was long enough to ensure firstly that temperature equilibrium had been reached and secondly that steady state conditions had been virtually established. Results Table 1 gives the results for the main series of runs at 76.6 "C. The slopes are corrected on the assumption that k, = 3.925 x M-l s-', and the least mean squares line drawn through So plotted against P gives k, = 3.917 x 10-7~-1 s-'. The best value for k, is therefore 3.92 x M- ' s-'. The value derived for k, is 3.04 x M-' sc1. Similar data for 86.0 "C are given in Table 2. Here the slopes are corrected on the assumption that k, = 6.90 x M-' s-l, and the resulting plot of So against j3 makes k, = 6.896 x 10-7 M-l s-l. Hence the best value for k, is 6.90 lo-' M-' s-,, and the value for k, is found to be 8.10 x M-' s-'. Table 2 also contains data for 96.7 "C. The slopes are corrected with k, = 12.08 x M-' s-', and these in turn give k, = 12.06 x M-' s-'. The best value for k, is taken to be 12.07 x M-' s-', and the value of k, is found to be 24.79 x M-' s-l. A plot of log k, or log k3 against 1/Tis reasonably linear in both cases. For k,, the activation energy is 26.8, kcal/mol. If we assume that AH* isindependent of temperature, then AH* = 26.2, kcal/mol (at 25 "C) and AS* = -7.3 cal/deg mol; log k, extrapolated to 25 "C is -8.43. For k,, the activation energy is 13.7 kcal/mol. From this, AH* = 13.1 kcal/mol (at 25 "C) and AS* is - 47 cal/deg mol; log k, extrapolated to 25 "C is -7.96. The slope, So, and hence the rate constants do not appear to depend on the concentration of hydroxide ions present. Neither k, nor k, gave any systematic trend with [OH-] over the range 0.06 to 0.18 M. Discussion The results of the present work differ somewhat from those of Engel, Oplatka, and Perlmutter- Hayman (I), but the differences can be attributed to different experimental conditions. Firstly, these workers find the rate decreases with increasing ph, by a factor of 3 for every ph unit. However, their results are only for ph values of 8.5, 9.7, and 13.0, and they did not investigate whether there is any change around ph 13. Our
LEE AND LISTER: DECOMPOSlTlON OF AQUEOUS SODlUM BROMlTE 2825 TABLE 1. Decomposition of sodium bromite at 76.6 "C [BrOz-lo x lo2 [Br-Io x lo2 [OH-] so x lo5 Run (M) (M) (M) (M-I s-i) P TABLE 2. Decomposition of sodium bromite at 86.0 "C Temperature [Br02-lo x [Br-lo x [OH- 1 so x lo5 Run c'c) (M) (M) (M) (M-I s-l) P results indicate that at high ph, the rate does not change with ph; this is presumably because protonated species (such as HOBr), which could offer alternative routes for the reaction, are effectively absent in 0.1 M sodium hydroxide. Hence we have written the reactions as occurring between ions. Secondly, the earlier workers found that at 25 "C the reaction Br- + Br02- -, 2Br0- predominated under their experimental conditions. However, the rate constants they quote are not very different: in our nomenclature, they give k, = 1.3 x M-' s - ', and k, = 1.43 x lop4 M-' s-'. At high temperatures, we find k1 greater than k,, but the activation parameters are such that this order reverses at low temperatures, and the rate constants extrapolated to 25 "C are k, = 3.7 x lo-', k, = 1.1 x lo-' M-' s-'. This is a long extrapolation, but at least it agrees with Engel, Oplatka, and Perlmutter- Hayman's result that k, is more important at low temperatures. Their rate constants quoted above are for a ph of 8.5, which is presumably why they are much larger than ours. However, the discrepancy is larger than would be expected from their observations on the effect of ph, and cannot at present be explained. The results allow some calculations to be made on the thermodynamic properties of bromite ions. Reactions 3 and 4 above lead to an equilibrium k3 Br- + BrO2-=2Br0- k4 with the activation parameters: reaction 3 has AH* = 13.1 kcal/mol, AS* = -47 cal/deg mol;
2.9 2826 CANADIAN JOURNAL OF CHEMISTRY. VOL. 49, 1971 TABLE 3. Conventional thermodynamic quantities for aqueous ions at 25 "C Ion AHOr AGO, So - Ion AHor AGOr So C1- -40.0-31.35 13.2 Br- -28.9-24.6 19.3 C10- -25.4-8.9 10.0 BrO- -21.9-8.0 12.0 CIOz - -16.6 2.7 24.1 Br02- -8.9 6.5 30.5 C10,- -25.15-2.15 38.8 Br03- -16.2, 4.3 38.6 c104- -31.0-2.1 + 43.5 BrOo; - 28.9 44.7 clzozz -* -26.3 10.1 8.7 BrzOz -25.0 9.1 3.0 and reaction 4 has AH* = 18.8 kcal/mol, AS* = -21 cal/deg mol (5). Hence AH for the reaction in the k, direction is - 5.7 kcal/mol, and AS is - 26 cal/deg mol. Thermodynamic data on dilute aqueous solutions are for sodium bromide : AH0, = -86.2, AGO, = - 87.2 kcal/mol, SO = 33.7 cal/deg mol; and for sodium hypobromite: AH0, = -79.1, AGO, = -70.6 kcal/mol, So = 26.4 cal/deg mol (6, 7). Hence calculated values for sodium bromite (aq) are: AH,' = -66.3 kcal/mol, and So = 45 cal/deg mol. These results, combined with data for the elements in sodium bromite, make the value for its AGO, = -56.1 kcal/mol. Some confirmation of this is provided by the value for AHO, of - 66.0 kcallmol from calorimetric measurements on the reaction of aqueous sodium bromite and sodium sulfite (8). Table 3 summarizes values for chlorine and bromine oxyions, in order to show that the values deduced above for sodium bromite (aq) are reasonable. The data come from refs. 6,7, and 9. Values are included for the activated species X2OZ2- * (for C12022- *, see ref. 10). The values in Table 3 are conventional values, relative to H+ (aq) as zero. The general trends of these quantities are similar, apart from the well-known low stability of perbromates. As regards bromite ions, their AH0,, relative to the other bromine ions, is somewhat more positive compared to the chlorine series. This is somewhat offset by their higher entropy, so that there is less contrast inagof. Thus the equilibrium constant for X- + X0,- =$ 2x0- is calculated to be 1.4 x for the bromine compounds, and 1.4 x lo-'' for chlorine. The activated species X2OZ2- * are also not very different in the two series, although here the entropy of the bromine species is somewhat lower. At present, it does not seem to be possible to explain these differences in detail. In a recent paper (ll), Massagli, Indelli, and Pergola report results of the decomposition of bromite at ph of 8.5 and below. Their results follow a rate equation Rate = [HBr02][Br-]{kl + k, [H']} which is evidently a different mechanism from ours, the difference being presumably attributable the much lower ph of their experiments. 1. P. ENGEL, A. OPLATKA, and B. PERLMUTTER-HAYMAN. J. Am. Chem. Soc. 76, 2010 (1954). 2. J. BREISS. Ph.D. Thesis. L'universitC de Strasbourg, Strasbourg, France. 1959. 3. R. M. CHAPIN. J. Am. Chem. Soc. 56, 2211 (1934). 4. B. PERLMUTTER-HAYMAN and G. STEIN. J. Phys. Chem. 63, 734 (1959). 5. M. W. LISTER and P. E. MCLEOD. Can. J. Chem. In publication. 6. W. M. LATIMER. Oxidation potentials. 2nd ed. Prentice-Hall, Inc., New York. 1952. 7. National Bureau of Standards Circular 500, U.S. Government Printing Office, Washington. 8. M. B. KENNEDY and M. W. LISTER. Unpublished result. 9. G. K. JOHNSON, P. N. SMITH, E. H. APPELMAN, and W. N. HUBBARD. Inorg. Chem. 9, 119 (1970). 10. M. W. LISTER. Can. J. Chem. 34, 465 (1956). 11. A. MASSAGLI, A. INDELLI, and F. PERGOLA. Inorg. Chim. Acta, 4, 593 (1970).