Carbonate Radical Formation in Radiolysis of Sodium Carbonate and

Jan 15, 2010 - Carbonate Radical Formation in Radiolysis of Sodium Carbonate and Bicarbonate. Solutions up to 250 °C and the Mechanism of its Second ...
8 downloads 0 Views 263KB Size
2142

J. Phys. Chem. A 2010, 114, 2142–2150

Carbonate Radical Formation in Radiolysis of Sodium Carbonate and Bicarbonate Solutions up to 250 °C and the Mechanism of its Second Order Decay Kyle S. Haygarth, Timothy W. Marin, Ireneusz Janik, Kotchaphan Kanjana, Christopher M. Stanisky, and David M. Bartels* Radiation Laboratory, UniVersity of Notre Dame, Notre Dame, Indiana 46556 ReceiVed: NoVember 3, 2009; ReVised Manuscript ReceiVed: December 15, 2009

Pulse radiolysis experiments published several years ago (J. Phys. Chem. A, 2002, 106, 2430) raised the possibility that the carbonate radical formed from reaction of · OH radicals with either HCO3- or CO32- might actually exist predominantly as a dimer form, for example, · (CO3)23-. In this work we re-examine the data upon which this suggestion was based and find that the original data analysis is flawed. A major omission of the original analysis is the recombination reaction · OH + · CO3- f HOOCO2-. Upon reanalysis of the published data for sodium bicarbonate solutions and analysis of new transient absorption data we are able to establish the rate constant for this reaction up to 250 °C. The mechanism for the second-order self-recombination of the carbonate radical has never been convincingly demonstrated. From a combination of literature data and new transient absorption experiments in the 1-400 ms regime, we are able to show that the mechanism involves pre-equilibrium formation of a C2O62- dimer, which dissociates to CO2 and peroxymonocarbonate anion:

· CO3- + · CO3- T C2O62- f CO2 + O2COO2· CO3- reacts with the product peroxymonocarbonate anion, producing a peroxymonocarbonate radical · O2COO-, which can also recombine with the carbonate radical:

· CO3- + CO42- f · CO4- + CO32· CO3- + · CO4- f C2O72I. Introduction -

The carbonate anion radical · CO3 is one of the most extensively studied inorganic radicals due to its ubiquitous nature in the environment, its relatively long lifetime, and its accessible absorption at 600 nm.1-15 Typically the species is generated by · OH radical reaction with either CO32- or HCO3- ion as in reactions 1 and 2:

· OH + CO32- f · CO3- + OH-

(1)

· OH + HCO3- f · CO3- + H2O

(2)

Since direct measurement of the weak UV absorption of · OH is difficult, the carbonate ion is particularly useful as a scavenger for · OH radicals in laboratory studies. A study of the carbonate radical yield in pulse radiolysis of both carbonate and bicarbonate solutions at elevated temperature15 suggested several years ago that the radical actually exists in a dimeric form, for example, · (CO3)23-. In this work we re-examine the evidence cited for this claim and show it to be based on a flawed analysis. However, we also confirm * Corresponding author. Phone: +1 574 631 5561; Fax: +1 574 631 8068; E-mail: [email protected].

the observation15 that decay of the radical deviates significantly from simple second-order kinetics when the radical is generated in bicarbonate solutions. We review the evidence available for the recombination reaction and suggest a mechanism that can explain the data. For such a well-known radical, the structure, acid/base properties, and recombination mechanism of · CO3- have remained controversial for a very long time. The first comprehensive study of the carbonate radical using pulse radiolysis was accomplished by Weeks and Rabani,1 who recorded the transient spectrum and investigated the decay kinetics as a function of pH. They found the decay of the carbonate radical in alkaline solution was second order over 80% of the decay, but deviated from second-order kinetics at longer time. A shift in the nature of the deviation could be seen between pH 11.5 and 13. The authors provided no explanation for the deviation from second order. Two sets of reaction products were postulated as consistent with the data:

· CO3- + · CO3- f CO2 + CO42f 2CO2 + H2O2 + 2OH-

(3a) (3b)

Behar et al. generated the carbonate radical in oxygenated solutions both by radiolysis and by photolysis of hydrogen

10.1021/jp9105162  2010 American Chemical Society Published on Web 01/15/2010

Carbonate Radical Formation

J. Phys. Chem. A, Vol. 114, No. 5, 2010 2143

peroxide solutions.2 They measured the reaction rates of carbonate radical with both H2O2 and HO2-, finding a much faster rate for HO2-:

· CO3- + H2O2 f · HO2 + HCO3-

(4)

· CO3- + HO2- f · O2- + HCO3-

(5)

They explained the deviation from second-order kinetics observed by Weeks and Rabani1 as the reaction of carbonate radical with the peroxide generated as a product of the recombination, according to reaction 3b. The pH dependence around pH ) 12 results from the acid-base equilibrium of H2O2 and HO2-, and the very different rates of reactions 4 and 5.2 Behar et al. discounted the possibility that the carbonate radical itself might be present in an acid-base equilibrium 6 in the pH range studied:

reached an incorrect conclusion, and reiterated their earlier conclusion that pKa for reaction 6 is less than 0. It would seem that the work of Lymar et al.14 and the Raman study of Bisby et al.10 had settled the issue of · CO3- radical structure and pKa. However, the U. Tokyo group15 then published a survey of carbonate radical yield and second-order decay in a wide concentration range of carbonate and bicarbonate solutions up to 350 °C. To explain apparent anomalies in the yield and decay, Wu et al.15 proposed that the carbonate radical is normally present in the form of a dimer anion, either · (CO3)23- or · H(CO3)22- formed via the following reaction processes:

· CO3- + HCO3- T · H(CO3)22- (in NaHCO3 solution)

(7) · CO3- + CO32- T · (CO3)23- (in Na2CO3 solution)

(8)

which are converted between the acid and base forms via: -

+

· HCO3 T H + · CO3

-

(6)

Lilie et al.4 applied simultaneous transient absorption and conductivity detection to the study of · CO3- reactions. On the basis of the conductivity transients, these authors ruled out the second-order recombination mechanism 3b in favor of 3a. They provided no explanation for the pH-dependent deviation from second-order kinetics seen by other workers. Chen, Cope, and Hoffman3 also found pH-dependent deviation from second-order decay kinetics in alkaline solution. From kinetic measurements in the presence of added solutes the authors concluded that the pKa of · HCO3 in reaction 6 is 9.6 ( 0.3. However, their measured rate constants of recombination at high pH were systematically higher than those reported in other studies. Eriksen et al. used the · CO3- radical as a one-electron oxidant in studies of cyclic hydrazines in the pH region reported by Chen et al.,3 but found no indication of a possible protonation of the · CO3- radical.5,6 In further investigations16 they reported a range of 7.0-8.2 for the pKa based on an increase in the initial absorbance of the radical with pH. This study also first reported a negative temperature coefficient for the recombination reaction 3, although no data was presented. It was proposed that the reaction involves a pre-equilibrium with the short-lived intermediate C2O62- .The negative temperature coefficient was later confirmed by Ferry and Fox12 who found the rate constant increases again dramatically above 300 °C. Bisby, et al.10 investigated the structure of the carbonate radical using time-resolved resonance Raman spectroscopy. They found no change of the radical’s Raman spectrum between pH 7.5 and 13.5, and concluded that the pKa for reaction 6 must be less than 7.5. Using a stop-flow/pulse radiolysis technique, Czapski et al.11 made measurements of the (relatively slow) reaction of · OH radical with carbonic acid in acidic solution. They found the same spectrum of carbonate radical with 600 nm maximum that is found in neutral and alkaline solutions and concluded that the radical · HCO3 is a strong acid with pKa < 0. At roughly the same time, Zuo et al.13 investigated the pH dependence of reactivity of the (bi)carbonate radical in mixed carbonate/SCN solutions and concluded the · HCO3 radical must have a pKa of 9.6. Lymar, Schwarz, and Czapski14 published a rebuttal of the pKa claims made in Zuo et al.,13 Chen et al.,3 and Eriksen et al.,16 showing how each of these studies had

· H(CO3)2•2- T H+ + · (CO3)2•3-

(9)

Using this model and equilibrium constants determined from their data, Wu et al.15 calculated the pKa of the dimer in reaction 9 to be 9.30 ( 0.15. The authors speculate that the pKa of 9.5-9.6 previously determined for the assumed · HCO3 at room temperature is actually the value for · H(CO3)22- and that the dimer model is quite reasonable by analogy with other radicals having the · X2- formula. Upon considering the Tokyo group’s evidence and kinetic model, we found two serious flaws. The first of these is the omission, in considering the acid-base equilibria of the carbonate/bicarbonate/carbonic acid system, of the additional equilibrium 12 to form aqueous CO2 from carbonic acid:

HCO3- T H+ + CO32-

(10)

H2CO3 T H+ + HCO3-

(11)

H2CO3 T CO2 + H2O

(12)

At elevated temperature this is a serious flaw: at 250 °C, less than half of the expected scavenger concentration is actually present in the form of bicarbonate ion (see below). Second, for low concentrations of the bicarbonate or carbonate scavenger for · OH radicals, the · OH radical persists for a relatively long time in the presence of · CO3- radicals. In this situation the cross-recombination of · CO3- radicals with · OH cannot be ignored as it was by Wu et al.15 In the following sections we describe and evaluate a kinetic model using the ideas laid out in the last paragraph and fit the Tokyo group’s data to estimate the · CO3- + · OH reaction rate. We use the model to fit pulse radiolysis/transient absorption experiments and explain carbonate radical yields, without the need to postulate a dimer radical form. Then, we consider the curious observations of second-order decay rate constant of the carbonate radical in NaHCO3 and Na2CO3 solutions. The mechanism for the second order recombination has never been fully and convincingly worked out. We propose a mechanism that is fully consistent with the data.

2144

J. Phys. Chem. A, Vol. 114, No. 5, 2010

II. Experimental Section Experiments Performed at Argonne National Laboratory. Pulse radiolysis/transient absorption experiments were carried out using 4-40 ns pulses from the Argonne Chemistry Division’s 20-MeV electron linac, where longer pulse width corresponds to higher applied dose. The differing pulse widths used have negligible impact on the observed kinetics, which occur on a time scale of several microseconds. The hightemperature/pressure sample cell, flow system, and basic experimental setup and characteristics were described in previous publications.17-24 Normal temperature and pressure stabilities were (0.2 °C and (0.1 bar, respectively. Analyzing light from a pulsed 75 W xenon lamp (Photon Technology International) was selected using a 40 nm bandwidth interference filter (Andover Corporation) centered at 600 nm, the wavelength corresponding to the maximum absorption of · CO3- (max at 600 nm ) 1970 M-1 cm-1).14 The same filter was used for all data collection, as the maximum absorption is known not to shift with temperature changes.15 A silicon photodiode (FFD100, EG&G) was used for detection. Kinetics were measured at 100, 200, 225, and 250 °C. Sodium bicarbonate (NaHCO3) solutions were prepared at a concentration of 0.0200 M (sodium hydrogen carbonate, Aldrich, 99.99% + used as received) in deionized water (18.2 MΩ-cm, Barnstead Nanopure cartridge system). Water was degassed with dry nitrogen gas for ∼30 min prior to adding sodium bicarbonate to avoid contamination by carbonate ions arising from possible carbon dioxide absorption. For experiments, the NaHCO3 concentration was controlled by mixing with a separate deionized water sample using two separate HPLC pumps (Alltech 301). The water sample was also previously degassed with nitrogen. Immediately before performing kinetic experiments, both the NaHCO3 solution and deionized water samples were purged with N2O gas for ∼30 min, giving an N2O concentration of 0.024 m. Samples were then kept under an N2O atmosphere throughout experiments. Two or three concentrations of NaHCO3 were used at each temperature, as diluted by mixing with deionized water. At 100 °C, concentrations were 0.00188, 0.00313, and 0.00625 m (molal); at 200 °C, concentrations were 0.00177, 0.00313, and 0.00625 m; and at 225 and 250 °C, concentrations were 0.00313 and 0.00625 m. Total system flow rate was 1.6 mL/min, and the experimental pressure was 250 bar. Multiple doses from the linac were used at each temperature in order to extract rate information due to influence of the second-order chemistry. Four to six doses were applied at each temperature, producing final · CO3- concentrations between ∼ 0.7 and 5 × 10-6 m. Experiments Performed at Notre Dame Radiation Laboratory. Pulse radiolysis/transient absorption experiments were carried out using 4-50 ns pulses from the Notre Dame Radiation Laboratory’s 8 MeV electron linac, where longer pulse width corresponds to higher applied dose. An OLIS RSM-2000 timeresolved spectrometer was used to record the kinetics, with dual beam fixed wavelength detection at 600 nm. Sodium bicarbonate (NaHCO3) solutions were prepared at a concentration of 0.0400 M (sodium hydrogen carbonate, Aldrich, 99.998% + used as received) in purified deionized water (resistivity 18 MΩ-cm, total organic carbon