Environ. Sci. Technol. 2008, 42, 3241–3247
Investigation of the Kinetics and Mechanisms of the Oxidation of Cerussite and Hydrocerussite by Chlorine HAIZHOU LIU, GREGORY V. KORSHIN,* AND JOHN F. FERGUSON Department of Civil and Environmental Engineering, University of Washington, Box 352700, Seattle, Washington 98195-2700
Received October 3, 2007. Revised manuscript received February 4, 2008. Accepted February 5, 2008.
an important role during the oxidation of Pb(II) by hydroxyl radical OH• and PbO2 reduction (16, 17). The oxidation of Pb(II) solutes by OH• was hypothesized to proceed via the generation of Pb(OH)2+ that disproportionated and formed amorphous Pb(IV) oxyhydroxide PbO(OH)2 · H2O (16). EC deposition of PbO2 also involves Pb(III) species such as Pb(OH)2+ (18). Despite the extent of research of PbO2 formation, little has been learned about the mechanism and kinetics of oxidation of lead (II) solids by chlorine. Specifically, effects of pH, concentrations of chlorine, and carbonate on these processes remain unquantified. The goal of this study is to address these issues using representative solids (cerussite and hydrocerussite) that are ubiquitous in water distribution systems.
Materials and Methods Reactions of representative lead (II) solid phases (hydrocerussite, cerussite) with chlorine were examined in this study. Chlorine consumption profiles for these solids exhibited a lag phase, during which little consumption of chlorine occurred, and an ensuing rapid reaction phase. The durations of these phases were affected by the pH, carbonate, and chlorine concentrations. SEM and XRD data showed that hydrocerussite started to be transformed into cerussite during the lag phase. Kinetic analysis indicated that only the protonated form of HClO drives the autocatalytic oxidation step, which is mediated by dispersed PbO2 crystals. The rate of the noncatalytic oxidation decreased with the increase of carbonate due to the formation of unreactive surface carbonate Pb(II) complexes.
Introduction In 2002, a dramatic increase of water-borne lead was detected in Washington DC.drinking water distribution system. This event was later shown to be caused by the introduction of chloramine instead of chlorine as the residual disinfectant (1) and ensuing destabilization of lead dioxide that had been formed on the surface of lead service line ubiquitously present in the Washington, DC distribution system while chlorine was the residual disinfectant (2–5). Analyses of corrosion scales found in distribution systems have shown a wide occurrence of PbO2. These studies showed that Pb(II) and Pb(IV) phases tend to coexist (6–9). In some cases, PbO2 was formed at the scale/war interface (9) while in others PbO2 particles were dispersed within corrosion scales that, in the absence of phosphate inhibitors, are predominated by hydrocerussite Pb3(CO3)2(OH)2, cerussite PbCO3, and sometimes litharge PbO (8, 10, 11). Since PbO2 is practically insoluble at circumneutral pHs and for that reason beneficial for plumbosolvency control (1, 3, 9, 12), it is important to determine how water chemistry affects its generation and stability. It has also been recognized that PbO2 particles can be released from corroding surfaces and contribute to an underestimated component of lead exposure (13). Formation of PbO2 in conditions unrelated to drinking water treatment and distribution has well been studied, with attention paid to the electrochemical (EC) deposition of scrutinyite (R-PbO2) and plattnerite (β-PbO2) in highly acidic or basic media (14, 15). Pb(III) intermediates appear to play * Corresponding author phone: (206) 543-2394; fax: (206) 6859185; e-mail:
[email protected]. 10.1021/es7024406 CCC: $40.75
Published on Web 03/20/2008
2008 American Chemical Society
The oxidation of hydrocerussite and cerussite was carried out in the range of pH and total carbonate concentrations (TOTCO3) in which these solids are thermodynamically stable. Calculations to determine relevant stability boundaries were carried out with MINEQL+ 4.5 software (Environmental Research Software, Hallowell, ME). Analytical grade hydrocerussite and cerussite used in the experiments were purchased from Aldrich. Prior to oxidation, requisite amounts of the solids were suspended in 250 mL flasks containing solutions with target pH and TOTCO3 values for at least 10 h. X-ray diffraction (XRD) studies showed that, in agreement with the MINEQL+ calculations, the solids were structurally stable during that time. Soluble lead concentrations were also stable at the end of equilibration. For instance, for hydrocerussite equilibrated at TOTCO3 0.001 M and pH 7.3, lead concentrations passing through 0.025 µm and 0.45 µm filters were 70 and 200 µg/L, respectively. Following the equilibration step, 2 mg/L (this chlorine concentration is typical for drinking water) to 50 mg/L of chlorine was injected. Stock solution of chlorine was prepared by using 5% NaOCl solution. pH values were maintained at a constant level ((0.05 pH units from the target) using a Eutech Instrument Alpha pH200 controller. 0.05 M HClO4 or NaOH solutions were added to the vessel for pH adjustment. TOTCO3 was varied by adding NaHCO3. Ten mM of NaClO4 was added to maintain a nearly constant level of the ionic strength, which changed by no more than 10% during the oxidation. The ambient temperature was 22 °C. The reaction vessel was sealed with Parafilm. Samples were periodically taken from the system and filtered through 0.45 µm Millipore filters. The concentration of chlorine in the filtrate was measured using the standard DPD method. Selected samples of filter-retained particles were treated with HClO4 at pH 2.0 to isolate Pb(IV) solids formed during the reaction. Solids retained on the filters were analyzed by XRD and scanning electron microscopy (SEM). XRD measurements were performed with a Philips 1820 diffractometer. Jade+ software (version 6) was used for XRD pattern identification. Reference patterns were from the 1995 version of International Center for Diffraction Data (ICDD). SEM examination was performed with a JEOL JSM-7400F microscope. BET surface areas of hydrocerussite and cerussite were measured with a Micromeritics Flowsorb II 2300 BET system. Specific surface areas were determined to be 0.81 and 0.43 m2/g for hydrocerussite and cerussite. Fitting of chlorine consumption data to the model developed in this study was done using OptQuest 2.1 option of Crystal Ball 7.0 software. VOL. 42, NO. 9, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3241
FIGURE 1. Oxidation of hydrocerussite. Normalized chlorine consumption profiles generated for different initial chlorine concentrations. Pb3(CO3)2(OH)2 ) 1.5 g/L, pH 7.5, TOTCO3 ) 0.001 mol/L.
FIGURE 2. Oxidation of hydrocerussite. Effects of total carbonate concentrations on chlorine consumption profiles. Pb3(CO3)2(OH)2 ) 1.5 g/L, initial chlorine concentration 50 mg Cl2/L, pH 7.3.
Results and Discussion Kinetics of hydrocerussite oxidation was ascertained based on chlorine consumption profiles. These profiles exhibited two stages, as demonstrated in Figures 1, 2, and 3 for experiments performed at varying chlorine doses, carbonate concentrations, and pH values, respectively. These stages included (1) an initial phase of incipient oxidation, or a “lag” phase, during which the concentration of chlorine decreased only slightly; and (2) an ensuing “rapid oxidation” phase associated with a pronounced decrease of the oxidant’s concentration. Chlorine concentration (Figure 1) strongly affected on the duration of these phases. For instance, for 50 mg/L chlorine dose, pH 7.5 and carbonate concentration 0.001 M the lag phase lasted for 5 h, while for the same pH and carbonate but chlorine concentration of 2 mg/L its duration 3242
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 9, 2008
was >40 h. The rate of chlorine consumption during the rapid oxidation phase was also higher at higher chlorine doses. The concentration of carbonate impacted the duration of the lag phase, but the rate of chlorine consumption during the rapid oxidation phase was nearly constant for varying TOTCO3 values (Figure 2). For instance, at pH 7.3 and a 50 mg/L initial chlorine dose, the duration of the lag phase was 2 h for a 0.0001 M TOTCO3 concentration and increased to 5 h for a 0.001 M. Effects of pH on the behavior of chlorine during the lag and rapid oxidation phases were distinct from those of the carbonate concentration (Figure 3). Increases of pH from 7.3 to 8.1 did not affect the duration of the lag phase. However, the rate of chlorine consumption during the rapid oxidation phase decreased as the pH increased.
FIGURE 3. Oxidation of hydrocerussite. Effects of pH on chlorine consumption profiles. Pb3(CO3)2(OH)2 ) 1.5 g/L, initial chlorine concentration 50 mg Cl2/L, TOTCO3 ) 0.001 mol/L. To probe the transformations of hydrocerussite, samples were withdrawn from the system for structural examination at the beginning of the reaction, middle, and end of the lag phase, middle of the rapid oxidation phase and the reaction’s end. Prior to oxidation, the XRD pattern of hydrocerussite had a characteristic peak at 2θ of 34° (Figure S1 in the Supporting Information). The peak’s intensity did not change significantly until the end of the lag phase, but during the rapid oxidation phase the peak weakened and ultimately disappeared. Since Pb (II) was present in excess relative to the amount of chlorine and could not be completely oxidized to Pb(IV), the disappearance of the XRD signature of hydrocerussite indicated that it had been transformed into other Pb(II) solids. Indeed, the intensity of XRD peaks located at 2θ values to close to 25° and 26° and corresponding to cerussite started to increase at the end of the lag phase, signifying a conversion of the nonoxidized fraction of hydrocerussite to cerussite. A series of characteristic peaks of R-PbO2 phase (scrutinyite) with the main peak located at 2θ of 29° also appeared in the XRD patterns during the initial part of the rapid oxidation phase and gained in intensity thereafter. XRD and SEM of the samples withdrawn from the system and washed with HClO4 showed that only R-PbO2 phase identical to that described in (5) remained in them (Supporting Information Figure S2). This confirmed that R-PbO2 was the final product of the oxidation of hydrocerussite. SEM data (Figure 4) confirmed these observations. Hydrocerussite was initially predominated by hexagonal platy particles (Figure 4A). During the lag phase, small prismatic crystals of cerussite appeared (Figure 4B). From the beginning of the rapid oxidation phase, larger cerussite crystals were detected, together with small particles of PbO2 (sizes
