Environ. Sci. Technol. 2009, 43, 6624–6631
The Inhibition of Pb(IV) Oxide Formation in Chlorinated Water by Orthophosphate DARREN A. LYTLE,* MICHAEL R. SCHOCK, AND KIRK SCHECKEL U.S. Environmental Protection Agency, ORD, NRMRL, 26 W. Martin Luther King Dr., Cincinnati, Ohio 45268
Received March 25, 2009. Revised manuscript received July 14, 2009. Accepted July 15, 2009.
Historically, understanding lead solubility and its control in drinking water has been based on Pb(II) chemistry. Unfortunately, there is very little information available regarding the nature of Pb(IV) oxides in finished drinking water and water distribution systems, and the conditions under which they persist. The objective of this research was to explore the impact of orthophosphate on the realistic pathways that lead to the formation of Pb(IV) oxides in chlorinated water. The results of XRD and XANES analysis showed that, in the absence of orthophosphate (DIC ) 10 mg C/L, 24°C, pH 7.75-8.1, 3 mg Cl2/L goal), Pb(IV) oxides formed with time following a transformation from the Pb(II) mineral hydrocerussite. Under the same experimental conditions, orthophosphate dosing inhibited the formation of Pb(IV) oxides. The Pb(II) mineral hydroxypyromorphite, Pb5(PO4)3OH, was the only mineral phase identified during the entire study of over 600 days, although the presence of some chloropyromorphite, Pb5(PO4)3Cl, could not be ruled out. The conclusions were further supported by SEM, TEM, and XANES analysis of lead colloids, and lead precipitation experiments conducted in the absence of free chlorine. The findings provide an important explanation for the absence of Pb(IV) oxides in some water systems that have used, or currently use, orthophosphate for corrosion control when otherwise, based on disinfection practices and water quality, its presence would be anticipated, as well as why the conversion from free chlorine to chloramines was not observed to increase lead release.
Introduction Controlling plumbosolvency and lead release in drinking water distribution systems from plumbing materials is a goal of all water utilities. The U.S. Environmental Protection Agency’s (USEPA) Lead and Copper Rule established an action level for lead at the consumer’s tap of 0.015 mg/L in a 1 L first draw sample (1-3). The impacts of factors such as pH, dissolved inorganic carbon (DIC), and orthophosphate on the solubility of Pb(II) solids are relatively well-known (4-7). Until recently, that knowledge has served as the basis for reducing lead levels in drinking water and, therefore, the risk of public exposure to lead. The way in which the water industry approaches lead control is being reconsidered as a result of the identification * Corresponding author phone: (513) 569-7172; e-mail: lytle.darren@ epa.gov. 6624 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 17, 2009
of Pb(IV) minerals in many drinking water distribution systems (DWDSs). The significance of considerable lead dioxide, PbO2, in pipe deposits is threefold. First, PbO2 is far less soluble than Pb(II) carbonate or hydroxycarbonate. Second, PbO2 can readily form at much lower pHs than optimally protective Pb(II) compounds. Third, the relationships of PbO2 stability to major water chemistry factors do not always parallel Pb(II). Schock et al. (6, 9, 10) have published reports of the presence of Pb(IV) in DWDSs, following the discovery of PbO2 formation in experimental pipe loop rig systems (4). In subsequent years, more pipe scale analyses have been performed on samples obtained from many additional water systems. Of the more than 215 lead pipe specimens obtained by USEPA researchers from 42 domestic water systems, one or more specimens from 14 of those systems have either of two Pb(IV) oxides: plattnerite (β-PbO2) or scrutinyite (R-PbO2)sor bothspresent in clearly identifiable quantities based on X-ray diffraction (XRD) analysis. Plattnerite is the tetragonal dimorph of lead dioxide, and in mineral form is brownish-black to black, with a dark red streak. Scrutinyite is orthorhombic in structure and darkreddish brown (11). Greninger et al. (11) observed that PbO2 solids may readily function as semiconductors. Thus, in DWDSs, there is likely to be considerable electrochemical reversibility, and ease of electron transport between the water and the underlying lead metal of the pipe. Changes in the scale makeup, in response to oxidation-reduction potential (ORP), are rather fast and measurable. Potential-pH diagrams for the lead system going back many years have a prominent stability field for the highly insoluble lead(IV) dioxide solid, PbO2 (5, 6, 12-16). The bulk of the literature on PbO2 is associated with the battery industry (17-22). Until recently, little information was available on the occurrence conditions of Pb(IV) in drinking water. What is known is that the presence of PbO2 in drinking water distribution systems is associated with waters of persistently high ORP, starting near the upper stability boundary for water (Supporting Information (SI) Figure S1). The high ORP necessary to achieve PbO2 formation in water can only be met with the consistent use of strong oxidants such as free chlorine and chlorine dioxide in waters with a relatively low oxidant demand. Lead(IV) oxides have been shown to form in water under “realistic” drinking water conditions, as has the importance of redox conditions on the stability and solubility of Pb(II)/ Pb(IV) (4, 23-25). Lytle and Schock (23) conducted longterm lead precipitation experiments in chlorinated (1-3 mg Cl2/L) water at pH 6.5, 8, and 10, both with and without sulfate. Results showed that both plattnerite and scrutinyite formed over time as long as high suspension ORP was maintained with free chlorine. Depending on pH and/or DIC levels, hydrocerussite (Pb3(CO3)2(OH)2) and/or cerussite (PbCO3) initially precipitated out and, with time, either disappeared or coexisted with PbO2. Water pH dictated mineralogical presence, where a high pH favored hydrocerussite and scrutinyite and low pH favored cerussite and plattnerite. Along with a transformation of Pb(II) to Pb(IV) came a change in particle color from white to a dark shade of red to dark gray (differs with pH), and a decrease in lead solubility. Liu et al. (26) studied reactions of lead(II) solid phases (hydrocerussite, cerussite) with chlorine in accord with the Lytle and Schock (23) study and prior theories of Pb(II) solid evolution (6). Chlorine consumption profiles for these solids exhibited a lag phase, during which little consumption of chlorine occurred, and an ensuing rapid 10.1021/es900399m
Not subject to U.S. Copyright. Publ. 2009 Am. Chem. Soc.
Published on Web 07/31/2009
FIGURE 1. X-ray diffractogram of lead solids during experiment I (DIC ) 10 mg C/L, 24°C, pH 7.88-8.01). reaction phase. The durations of these phases were affected by the pH, carbonate, and chlorine concentrations. The authors observed that hydrocerussite started to be transformed into cerussite during the lag phase, and the transformation was confirmed by SEM and XRD. Understanding factors that influence the oxidation state of lead is very important to avoiding and controlling lead release events in DWDSs. Concerns about predicting the consequences of changes in treatment that result in substantial change in ORP also pertain to drinking water systems that have previously applied phosphate-based corrosion inhibitors, or systems contemplating phosphate use in the future. Orthophosphate can promote the formation of relatively insoluble compounds such as hydroxypyromorphite (Pb5(PO4)3(OH)) and tertiary lead orthophosphate (Pb3(PO4)2) in DWDSs (27). Prior observations on the general modes of occurrence of Pb(IV) deposits showed that major PbO2 deposition was rare in phosphate-treated systems (10). No specific information on thermodynamics of mineralogical Pb(IV) orthophosphate or carbonate-containing solids has been uncovered so far, other than some studies done in the battery industry (28-33). Phosphate compounds are known to kinetically inhibit the growth of many divalent metal carbonate, oxide, and hydroxycarbonate minerals in natural, industrial, and environmental systems (34-40). Because of the many questions surrounding the occurrence of tetravalent lead solids and their relationships to lead release, USEPA researchers initiated a research program in 2002 to (a) explore the conditions and pathways that lead to the formation of PbO2 in water during long-term precipitation experiments and (b) determine the possible sensitivity of PbO2 scales to treatment changes that would affect ORP by analyzing the stability of the solids in response to changes in ORP. In this phase of the research program, the impact of orthophosphate on the formation of PbO2 and the properties of preformed PbO2 in water were investigated.
Materials and Methods A series of long-term lead precipitation experiments were conducted to evaluate the effects of water chemistry and time on the mineralogy and corresponding solubility of lead solids in water. During these tests, redox conditions were chosen to represent the high range of realistic utility practice, and were maintained with the intent to support Pb(IV) solids
(plattnerite and/or scrutinyite). High ORP of the water was maintained with free chlorine, which was maintained at a concentration goal of 3 mg Cl2/L throughout the study. Chlorine residuals were permitted to dissipate in some tests to examine the stability of Pb(IV). All experiments were conducted at room temperature (∼23 °C). Chemicals. Unless otherwise specified, all chemicals used in this study were Analytical Reagent (AR) grade. The amount of ultrapure nitric acid, HNO3 (Ultrex, J.T. Baker Chemical Company, Phillipsburg, NJ), used to preserve samples for metals analysis was 0.15% by volume. Dilute 0.6 M HCl (Mallinckrodt, Inc., Paris, KY) and 0.5 N NaOH (Fischer Scientific, Fairlawn, NJ) were used to finely adjust the pH, and sodium bicarbonate (NaHCO3) (Fisher Scientific, Fairlawn, NJ) was used to control DIC level. Orthophosphate was adjusted with Na3PO4 · 12H2O (Fisher Scientific, Fairlawn, NJ). Sodium hypochlorite (4-6% NaOCl, purified grade) (Fisher Scientific, Fairlawn, NJ) was used to oxidize lead(II), which was added as lead chloride (PbCl2) (Mallinckrodt, Inc., Paris, KY). Double deionized (DDI) water was prepared by passing building demineralized water through a Milli-Q Plus cartridge deionized water system (Millipore Corp., Bedford, MA), having a resistivity g18.2MΩ · cm. Test Protocol. Lead precipitation studies were conducted in a 3 L glass beaker. Secured at the top of the beaker were pH and two redox electrodes, dissolved oxygen (DO) and temperature probes, a mechanical stirrer, and an injection line for both acid and base. A computer software-controlled dual titrator system (Schott Gera¨te, Germany) was used to adjust and maintain the pH by adding small increments of acid or base. The computer software (Jenson Systems, Hamburg, Germany) recorded pH values and titrant volumes to a data file. Appropriate amounts of sodium bicarbonate, phosphate, and hypochlorite were then added to 3.0 L of DDI water. The titration system was programmed to the desired pH level and initiated. After the pH stabilized, lead chloride (PbCl2) was added to give an initial lead concentration of 20 mg/L (and chloride concentration of approximately 6.8 mg/L). Samples were drawn out of the cell with a syringe to collect water for DIC analysis and free and total chlorine concentrations prior to lead addition. After lead addition, a water sample was drawn for total lead analysis, and an additional sample volume was filtered through a 25 mm polypropylene syringe disk filter (0.2 µm) (Whatman Inc., Clifton, NJ) to separate colloidal lead for metal analysis. An additional VOL. 43, NO. 17, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6625
FIGURE 3. Soluble lead measured over time during experiment I (DIC ) 10 mg C/L, pH 7.75-8.19).
FIGURE 2. TEM and SEM micrographs of lead colloids collected during experiment I after (a) 0 (TEM), (b) 8 days (pH 7.97-8) (TEM), and (c) 8 days (pH 7.97-8) (SEM). volume (200 mL) was filtered through a 0.2 µm polycarbonate membrane filter using a table top vacuum pump. Once the precipitation cake formed, the filter was used to conduct solid characterization analysis. The remaining experimental water was used to fill ten 250 mL Teflon PTFE bottles with no head space, which were placed in a tumbler and allowed to age at room temperature (23°C). Bottles were periodically removed from the tumbler over 657 days, and the contents were analyzed for pH, ORP, total and soluble lead and orthophosphate concentrations, and free chlorine concentration. When a low chlorine residual was observed (
