Thermal decomposition of potassium bicarbonate - The Journal of

Dávid Pajtás , Krisztina Kónya , Attila Kiss-Szikszai , Petr Džubák , Zoltán Pethő , Zoltán Varga , György Panyi , and Tamás Patonay. The Journal of O...
0 downloads 0 Views 318KB Size
THERMAL DECOMPOSITION OF POTASSIUM BICARBONATE

2875

Thermal Decomposition of Potassium Bicarbonate' by I. C. Hisatsune and T. Ad1 Department of Chemistry, Whitmore Laboratory, The Pennsylvania State Universitg, University Park, Pennsylvania 16802 (Received April 8, 1970)

The thermal decomposition reactions of potassium bicarbonate dispersed in the KBr pressed disk have been studied by observing the changes in the infrared spectrum of the disk with heating. In the temperature range of 140-220°, the principal reaction in a disk containing up to about 2 mg/g of solute was the decomposition of the cyclic bicarbonate dimer into two monomeric anions with a rate constant of 7.2 x 102 exp[- (14 f 2 kcal)/RT] sec-l. Some carbonate ion was also produced during this reaction, and its yield increased with increasing initial concentration of the solute. At higher reaction temperatures, the formate ion was also produced at a rate second order in the bicarbonate monomer. The rate constant was 7.6 x 10'8 exp [ - (49 f 6 kcal)/RT] M-" sec-' for the temperature range 420-500', and the reaction stoichiometry suggested one formate ion produced from each bicarbonate monomer. The rate of carbonate production in the temperature range 450-550' appeared to be second order in the bicarbonate monomer with an Arrhenius activation energy of about 20 kcal/mol, but quantitative kinetic results could not be obtained for this reaction because of interference by the formate reaction,

Introduction When potassium bicarbonate consisting of hydrogenbonded cyclic dimer anions2is dispersed in a KBr pressed disk and heated a t about 500', monomeric bicarbonate ion3 as well as the carbonate ion is trapped in the matrix. Continued heating of the disk causes the monomer ion to decompose, and the formate ion appears as an unexpected reaction p r ~ d u c t . ~I n the present study, these decomposition reactions of the bicarbonate monomer and dimer anions have been examined quantitatively by infrared spectroscopy.

Experimental Section Fisher reagent grade potassium bicarbonate, without further purification, was our solute, and this was mixed with Harshaw Co. optical grade KBr powder to prepare the pressed disks. The technique of fabrication of these disks and their use in thermal decomposition studies have been described before. 8,6 All kinetic data were obtained with a Perkin-Elmer Model 521 grating infrared spectrophotometer.

Results A temperature range of 140-220' and solute concentrations from 0.4 to 2 mg/g of matrix were used to study the decomposition of the bicarbonate dimer into two monomer ions. Under these experimental conditions, the 880-cm-' COa2- and the 1633-cm-I HCOzinfrared bands were usually absent during the reaction, and the optical density of a HCOa- monomer infrared band varied linearly with respect to a dimer infrared band. The 830- and 697-cm-' bands of the dimere were used to follow the kinetics since other fundamental absorption bands overlapped with those of the reaction products. The sharp monomer banda a t 1697 cm-I was not used generally since its optical density became too large as the reaction progressed. However, whether

an infrared band of the dimer or that of the monomer was used, the decomposition of the dimer was observed to be first order in the dimer. The rate constants from these kinetic runs are summarized as an Arrhenius plot in Figure 1, and the reaction activation energy and frequency factor are listed in Table I. Table I : Arrhenius Parameters for Some Reactions of Potassium Bicarbonate in a KBr Matrix Temp range, Reaction

OC

(HCOa-)2 + 2HCOs2HCOs- -+ 2HC020 2 2HCOa- + CO?COz H2O

+

+

+

Frequency factor

Activation energy, kcal/mol

140-220 420-500

7 . 2 X lo2 see-1 7 . 6 X 10'' M-' sec-l

14 f 2 49 f 6

450-550

...

w20

Once the HCOa- ion was trapped in a KBr disk, it was relatively stable until the reaction temperature was increased appreciably. At about 450" , the infrared spectrum of the disk showed a gradual decay of the monomer bands and a concomitant growth of the HC02- bands. During these changes the optical density of the 1633-cm-' formate band varied linearly (1) This work was supported by PHS Grant EC-97 from the Environmental Control Administration and by Grant AFOSR-907-67 from the AFOSR(SR.C)-OAR,USAF. (2) I. Nitta, Y. Tomiie, and C. H. Koo, Acta Crystallogr., 5, 292

(1952). (3) D. L. Bernitt, K. 0. Hartman, and I. C. Hisatsune, J. Chem. Phys., 42, 3553 (1965). (4) I. C. Hisatsune and K. 0. Hartman, Science, 145, 1455 (1964). (6) I. C. Hisatsune, Nippon KaQarCu Zasshd, 89, 1143 (1968). (6) K. Nakamoto, Y. A. Sarma, and H. Ogoshi, J. Chem. Phys., 43, 1177 (1966). The Journal of Physical Chemistry, Vola74,No. 16, 1970

2876

I. C. HISATSUNE AND T. ADL IO

5

1.0

L

E

0,5

1 0 ~ 1 ~(

~

2,4

2.2

2.0

103/r

2.6

~ K T '

Figure 1. First-order rate constants for (HCOs-)a in a KBr matrix.

+ 2HCOa-

with respect to the optical densities of the bicarbonate bands. From nine such optical density correlation plots, it was found that the 1697-cm-1 HCO3- band was 1.65 f 0.36 times more intense than the 1633-cm-' HC02- band whose molar extinction coefficient from independent calibration measurements was 1.81 X lo3 M-' cm-'. Since the most reasonable reaction stoichiometry consistent with the experimental data was to take the mole ratio HCOa-/HC02- as one, the molar extinction coefficient of the 1697-cm-1 bicarbonate band was calculated to be 2.99 X lo3M-' cm-l. Using this value together with the known extinction coefficients for HC02- and C032- infrared bands, we have determined the yield of ionic products in a series of decomposition reactions of the dimer. The results are summarized in Table 11. I n calculating the carbon

Table I1 : Decomposition Stoichiometry of Potassium Bicarbonate in a KBr Matrix Initial mol,a ,-(HCOB-)~ HCOa-

1.6 2.8 5.5 7.7 7.0 7.1 7.1 4.1 12

1.3 2.5 4.9 6.2 5.2 6.3 6.4 8.5 66

Final mola------. HCOaGO@-

0.41 1.3 0.73 0.94 0.24 0.73 0.55 0.41 6.9

0.35 0.76 2.1 1.6 2.7 1.9 2.2 8.0 84

Carbon balanoeb

1.0 1.1 1.1 1.1 1.2 0.92 0.94 1.2 1.4

--HeatingTime, Temp, min "C

9 9 35 5 25 35 35 30 25

488 488 450 488 450 450 450 300 505

Units are in 10-7 mol. * Ratio of final mol of (HCOaHC022C082-) t o twice the initial mol of (HCOa-)z.

+

+

balance in this table, the contribution of CO? to the reaction products was taken as twice the observed conThe Journal of Physical Chemktry, Vol. 74, No. 16, 1970

o ~ j '

Figure 2. Second-order rate constants for 2HCOa- + 2HC0202 in a KBr matrix.

+

centration since C02, the remaining carbon product in the formation of C O F , was only partially trapped in the KBr disks. For the temperature range of 420-500", the formate producing decomposition reaction of bicarbonate was second order in HC03-, and the rate constants obtained from the infrared bands of the formate and the bicarbonate monomer were the same within experimental uncertainties. Figure 2 summarizes the experimental rate constants, and the Arrhenius parameters are given in Table I. A number of kinetic runs on the formate production showed, near the end of the reaction, systematic deviations of the experimental points from the expected second-order behavior. I n such runs, the formate production appeared to cease and the decay of the bicarbonate monomer was about an order of magnitude slower than during the formate growth stage. This slow-decay reaction was observed more often when the initial concentration of the monomer was low. Although it was not possible to differentiate clearly this slow-decay reaction from the faster formate reaction, it appeared that a product of the former reaction was carbonate and that the reaction was also second order in HC03-. The rate constants were of the order of M-I sec-l in the temperature range of 450-500" and showed only a small temperature dependence corresponding to about 2O-kcal/mol activation energy. The decomposition of HC03- was complicated, in addition, when the initial concentration of C03'- was high. In such a disk, the HCOa- decay was generally slower and the COa2- infrared band often showed some decay with reaction time. As summarized in Table 11, a high initial concentration of bicarbonate dimer gave more C032- relative to HCOa-. Prolonged grinding of the solute and matrix salts also led to disks with higher concentrations of C O P . However, freezedrying of the solute did not appear to have much effect

THERMAL DECOMPOSITION O F POTASSIUM BICARBONATE on the HCOa- reaction. One disk prepared by the freeze-dry method with 0.22 mg/g of solute gave a formate yield of 85%, based on the actual initial HCO3concentration, and a rate constant of 0.11 M-' sec-I a t 450". On the other hand, the decomposition a t 488" of a disk prepared by grinding 0.21 mg/g solute gave a 79% yield of formate and a monomer decay rate constant of 0.71 M-lsec-'.

Discussion The following sequence of reactions, which may not necessarily be elementary, appears to account for the bicarbonate decompositions observed in this study. (HC03-)2

2HC03-

(1)

+ 2HCOa- J_ COa2- + COz + HzO 2HCOa-

2HC0.2-

0 2

(2) (3)

Although the experimental activation energy given in Table I for reaction 1 is similar to the hydrogenbond energies in carboxylic acid dimers,' the small frequency factor and essentially the same activation energies observed in recombination reactions of anion free radicals* suggest that the Arrhenius energy for reaction 1 may be the activation energy for the diffusion of HCOa- in a KBr pressed disk. The occurrence of the reverse step in reaction 1 is suggested by an observation3 that the HCO3- infrared bands all decrease and dimer spectrum is partially regenerated when the disk is allowed to stand a t room temperature for a few weeks or when the disk is reground. I n reaction 2 an attempt was made to detect the formation of oxygen by decomposing the bicarbonate dimer in a K I matrix a t 490°, but no iodate was produced. However, the reverse of this reaction was

2877

observed in our earlier studyg of the decomposition of HC02-. I n that decomposition, HC03- was always observed as a minor product, and DC02- gave DC03which soon changed to HCOa- on heating. (This rapid hydrogen exchange reaction prevented us from examining the decomposition kinetics of DC03-.) Also, when a heated formate disk was ground in an oxygen atmosphere and then reheated, slightly more HCO,was formed than from a disk ground in air. Interestingly, the decomposition of the formate ion was also a second-order reaction, and its Arrhenius activation energy of 50.7 3.5 kcal/mol is similar to the activation energy of reaction 2. However, the rate of bicarbonate decomposition is about three orders of magnitude faster than the rate of formate decay. Reaction 3 is a familiar one, and it becomes dominant as the bicarbonate concentration is increased (Table 11). Thus, in the decomposition of pure potassium bicarbonate or of bicarbonate/KBr mixture in a powdered form, only the carbonate ion is detected as the final ionic product. The kinetics of this reaction is uncertain, but our estimated reaction activation energy is comparable to a value reported recently'O for the decomposition of pure potassium bicarbonate. The activation energy for the decomposition of silver carbonate giving COzis also about 20 kcal/mol.l'

(7) G. C. Pimentel and A. L. McClellan, "The Hydrogen Bond," W. H. Freeman, San Francisco, Calif., 1960. (8) I. C. Hisatsune, T. Adl, E. C. Reahm, and R. J. Kempf, submitted for publication to J . Phys. Chem. (9) K. 0. Hartman and I. C. Hisatsune, ibid., 70, 1281 (1966). (10) I. P. Ishkin and E. S. Dubil, Zh. Prikl. Khirn. (Leningrad), 41, 62 (1968). (11) T. Wydeven and M. Leban, Anal. Chem., 40, 363 (1968).

The Journal of Physical Chemistry, Vol. YJ$~No. 16, 1970