Formation of Pb(IV) Oxides in Chlorinated Water

Recent research has shown that Pb(IV) oxides play a significant geochemical role in drinking water distribution systems. However, most of the guidance for lead control in drinking water is based on the presumption that Pb(II) solids control lead solubility. Therefore, a better understanding of the chemistry of Pb(IV) in water is needed. Long-term lead precipitation

experiments were conducted in chlorinated water (1–3 mg/L Cl2) at pH 6.5, 8, and 10, with ␤ and without sulfate. Results showed that two Pb(IV) dioxide polymorphs—plattnerite ( -PbO2) ␣ and scrutinyite ( -PbO2)—formed over time, as long as a high suspension redox potential BY DARREN A. LYTLE was maintained with free chlorine. Neither mineral formed spontaneously, and the rate of AND MICHAEL R. SCHOCK formation increased with increasing pH. Hydrocerrusite and/or cerrusite initially precipitated

out and over time either disappeared or coexisted with PbO2. Water pH dictated mineralogical presence. High pH favored hydrocerrusite and scrutinyite; low pH favored cerrusite 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 grey (differing with pH) and a decrease in lead solubility. If free chlorine was permitted to dissipate, the aging processes (i.e., mineralogy, color, and solubility) were reversible. Formation of Pb(IV) oxides in chlorinated water

ontrolling plumbosolvency and lead release in drinking water distribution systems from lead pipes, brass fixtures, and lead-based solders is a goal of all water utilities. The US Environmental Protection Agency (USEPA) Lead and Copper Rule (LCR) established an action level for lead C at the consumer’s tap of 0.015 mg/L in a 1-L first-draw sample (the sample must sit for at least 6 h before being measured; USEPA, 1992; 1991a; 1991b). Since passage of the LCR, the understanding of relationships between water quality and the solubility of lead-containing minerals found in drinking water distribution systems has grown extensively. The effects of such factors as pH, dissolved inorganic carbon (DIC), and orthophosphate on the solubility of Pb(II) solids are relatively well known (Schock & Clement, 1998; Schock et al, 1996; Schock & Wagner, 1985; Schock & Gardels, 1983). When adjusted appro- priately, these parameters can be used to control lead levels at the tap. Although most water treatment and water chemistry specialists are familiar with the radically different solubilities and scale formation properties of ferrous iron versus ferric iron or cuprous copper versus cupric copper, the same potential

2005 © American Water Works Association 102 NOVEMBER 2005 | JOURNAL AWWA • 97:11 | PEER-REVIEWED | LYTLE & SCHOCK This series of micrographs shows the color transformation of lead solids during experiment A (dissolved inorganic carbon = 10 mg/L C, temperature = 24oC, pH = 7.75–8.19). Numbers indicate the elapsed time in days. Over the first 82 days of the study, the color of the lead precipitate changed from white to dark orange-red in color, but by the end of the experiment (day 397), the red solid had disappeared.

behavior of lead is less well known. Potential-pH dia- Pb(IV) in drinking water systems. Schock and colleagues grams for the lead system going back many years have a (Schock, 2001; Schock et al, 1996) have published the prominent stability field for the highly insoluble lead only reports of the presence of Pb(IV) in drinking water dioxide (PbO2) solid (Schock, 1999; Schock et al, 1996; distribution systems. USEPA analyses of pipe scales from Schock & Wagner, 1985; Schock, 1981; Schock, 1980; lead pigtails and service lines collected in the late 1980s Pourbaix et al, 1966; Delahay et al, 1951). Thus an anal- and early 1990s verified that one or both of the common ␤ ogy can be drawn between the Pb(IV)–Pb(II) redox cou- polymorphs of PbO2—plattnerite ( -PbO2) and scruti- ␣ ple and the Fe(III)–Fe(II) redox couple. The higher-oxi- nyite ( -PbO2)—were present in varying degrees on the dation-state forms of the iron oxide or oxyhydroxide pipes from several different water systems (Schock et al, pipe-scale phases have much lower solubility than those 1996). In subsequent years, pipe-scale analyses from many of the lower oxidation state. In the case of lead, how- additional water systems have been performed. Of more ever, the oxidation–reduction potential (ORP) required for than 85 lead pipe specimens obtained from 34 water sys- the transformation of Pb(II) to Pb(IV) is much higher tems, at least 16 specimens representing 9 systems have ␣ ␤ than for the ferrous to ferric iron transformation. -PbO2, -PbO2, or both present in clearly identifiable quantities based on x-ray diffraction (XRD) analysis. More samples may have trace amounts that are hard to positively confirm. Usually the PbO2 was found to exist in the form of patches or a thin surficial layer at the water boundary. In the case of lead-service line deposits from Cincinnati, Ohio, the pipe scales were in contact with relatively high disinfectant concentrations for long periods of time, followed by a combination of granular activated carbon and free chlorine treatment (Schock, 2001). Only traces of basic cerussite (PbCO3) were found. It is not known when the PbO2 scale formed in the Cincinnati pipe, because no historical scale analyses were performed during several earlier treatment schemes. The distribution system area where the Cincinnati pipe specimens were obtained has recorded elevated pH (8.5–9.2) since the 1980s for corrosion control. In con- trast, lead service line specimens obtained from Madison, Wis., had a thin layer of continuous PbO2 at the water interface, with PbCO3 mak- ing up the bulk of the underlying scale. Madison represents a groundwater low in total organic carbon but with high alkalinity (~300 mg/L as calcium carbonate [CaCO3]) and a pH near neutral. Because of the existence of the PbCO3 Stereomicrographs collected on filters during experiment A show lead particles phase, lead concentrations were expected to be at 18 elapsed days (A), 64 elapsed days (B), 82 elapsed days (C), 145 elapsed high. Instead, the 90th-percentile lead concen- days (D), and 397 elapsed days (E). (Experimental conditions were: dissolved tration during LCR monitoring in 1992 was inorganic carbon = 10 mg/L C, temperature = 24oC, pH = 7.75–8.19; all images remarkably low (only 0.016 mg/L), which is are at 65× magnification). more than a factor of 10 below the solubility of

2005 © American Water Works Association LYTLE & SCHOCK | PEER-REVIEWED | 97:11 • JOURNAL AWWA | NOVEMBER 2005 103 Micrographs show the color difference in lead solids produced in experiment A at pH 8, day 145 (left), and experiment D at pH 10, day 30 (right).

tain a high concentration of free chlorine to combat biofilm problems or overcome corrosivity toward iron and its pipe wall demand. Questions raised. The previous scale analysis observations raised several important questions. How prevalent PbCO3 (Cantor, 2004). Subsequent targeted lead service is Pb(IV) in drinking water distribution systems? Can it line sampling also showed that approximately 90% of occur in some parts of a water distribution system and not all of those samples were <0.015 mg/L in total lead (Can- others? What is the significance of the presence of Pb(IV) tor, 2004). relative to lead release, and how does it affect lead treat- A significant amount of tetravalent lead scale mater- ment guidance? How quickly and by what pathways are ial was also found in pipe specimens from different parts Pb(IV) solids formed, and how stable are they in response of the Oakwood, Ohio, distribution system. The water has to a variety of water treatment changes? What is the a pH of 7 and a high alkalinity (~350 mg/L as CaCO3); chemical pathway by which PbO2 solids are broken down, greensand filtration is used for iron removal. Taken in and why does it cause periodic increases in lead concen- combination, these observations provide a reasonable trations? Furthermore, these questions draw attention to

Controlling plumbosolvency and lead release in drinking water distribution systems from lead pipes, brass fixtures, and lead-based solders is a goal of all water utilities. hypothesis to explain the apparent anomaly observed in a large gap in the thermochemical data in terms of the sol- many field studies in which high-alkalinity waters did ubility and speciation of tetravalent lead for these min- not tend to produce nearly as high levels of lead release erals, which makes it difficult to precisely and accurately as would be expected based on the knowledge of lead model either characteristic (Schock & Giani, 2004; solubility chemistry and bench-scale tests (Edwards et al, Schock, 2001). 1999; Dodrill, 1995; Dodrill & Edwards, 1995; Dodrill Purpose of this research. This investigation was initiated & Edwards, 1994). in 2002 to investigate the many questions surrounding the PbO2 and ORP. The presence of PbO2 is associated with occurrence of tetravalent lead solids and their relationships water of persistently high ORP, which is only possible to lead release. The research was undertaken with two by maintaining sufficient levels of free chlorine in water principal objectives. First was the initial exploration of the in contact with lead service lines and other lead materi- water quality conditions and pathways that lead to the for- als. Certain water chemistry conditions favor higher and mation of PbO2 in water during long-term precipitation persistent ORP. These conditions include: experiments. Second, the stability of the solids in response • pristine groundwaters low in natural organic matter to changes in ORP was investigated to determine the pos- and with little bulk oxidant demand allowing significant persistence of free chlorine; TABLE 1 Summary of experimental conditions • waters that effectively passivate iron and remove its oxidant demand Elapsed Free Measured Redox Day Chlorine (such as hard waters with high Time Chlorine Potential Adjustment buffering intensity); and Experiment days pH mg/L mV Was Made • waters with low oxidant de- A 397 7.75–8.19 0.1–3.0 707–201 26, 105 mand resulting from oxidative treat- B 329 6.65–7.81 0.1–2.9 850–202 No adjustment ments such as greensand filtration C 310 7.71–8.60 0.1–3.0 725–225 No adjustment (enabling the stability of high ORP). D 30 9.62–9.86 0.8–3.0 500–200 20 An additional example of waters with high ORP are those that con-

2005 © American Water Works Association

104 NOVEMBER 2005 | JOURNAL AWWA • 97:11 | PEER-REVIEWED | LYTLE & SCHOCK sible sensitivity of PbO2 scales to treatment changes that would affect FIGURE 1 Electromotive force–pH diagram for the lead–water– ORP. The physical properties, miner- carbonate system alogy, and solubility of lead solids at several pH values in well-chlorinated 1.50 water (<3 mg/L Cl2) were examined. The effect of sulfate on PbO2 forma- Water oxidiz PbO2 (s) tion was also briefly investigated. 1.00 ed (P PbO – O = 0.21 atm) 3 2 – Pb(OH)3 LITERATURE BACKGROUND 0.50 ␤ PbCO 0 PbO2 has two polymorphs— - Pb2+ 3 ␣ ␤ PbO2 and -PbO2. -PbO2 is the 0.00 Water reduced 2– 2– tetragonal dimorph of lead dioxide Pb(CO3)2 ( Pb(OH)4 PH =1 atm) 2 and in mineral form the dimorph is SHE v versus brownish black to black in color with –0.50 Eh— a dark red streak. ␣-PbO is ortho- 0 2 Pb(OH)2 rhombic in structure and dark red- Pb (s) –1.00 – – – 3 3 3 dish brown in color (Greninger et al, 3 2– CO 1975). The name “scrutinyite” alludes 2 HCO HCO CO –1.50 H to the close scrutiny that is necessary 01234567891011121314 in order to distinguish it from plat- pH tnerite. Greninger and colleagues also H CO —carbonic acid, HCO –—bicarbonate ion, P—pressure, Pb—lead, observed that PbO solids may read- 2 3 3 2 PbO (s)—lead(IV) oxide, (s)—solid, SHE—standard hydrogen electrode ily function as semiconductors. Thus 2 there is likely to be considerable elec- Lead species = 0.015 mg/L, dissolved organic carbon = 10 mg/L C, ionic strength correction = 0 trochemical 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 ORP are rather fast and measurable. trodeposition of PbO2 from 1 M perchloric acid solution To date, the bulk of literature on PbO2 is associated of Pb(II) (0.001–0.2 M Pb) on a gold rotating disk elec- with the battery industry. “Lead–acid” battery storage trode. They concluded that PbO2 formation involves sev- cells consist of two lead electrodes, one of which is cov- eral steps. First, hydroxide species chemisorb on the elec- ered with lead dioxide. Sulfuric acid (~20% solution) is trode. Next these particles interact with lead compounds employed as the electrolyte. The battery electrodes are to form soluble Pb(OH)2+, which is electrochemically reversible and the complete cell reaction can be written as oxidized to PbO2. Pb(III) products were reported to be intermediates in the overall process. o ␣ Pb + PbO2(s) + 2H2SO4(aq) ៞ 2PbSO4(s) + 2H2O Bagshaw and colleagues (1966) chemically prepared - ␤ ␤ PbO2 and -PbO2 for XRD studies. -PbO2 was pre- Because the positive electrode in the lead–acid battery is pared by oxidation of yellow monoxide (using a fused ␣ lead dioxide, some time has been spent on its study. mixture of sodium chlorate and sodium nitrate); -PbO2 ␤ ␣ PbO2 (both and structures) has been produced was prepared by hydrolysis of lead tetra-acetate. The and studied in the laboratory by electrochemical tech- researchers reported that electrodeposited PbO2 showed ␣ niques. The deposition of -PbO2 occurs in both basic and preferred orientation effects, whereas the chemically pre- ␤ acidic solutions at low current densities whereas -PbO2 pared PbO2 did not. can be obtained only in acidic media at high current den- Petersson and co-workers (1998) found that certain sities (Hampson, 1979). A mixture of the two phases was perchlorate salts have an effect on formed PbO2, par- deposited on titanium oxide substrate at the neutral pH ticularly on its discharge properties, self discharge, and range using a photoelectrochemical method (Matsumoto its active surface area. These results were explained by et al, 1999). structure defects caused by different cations; these dif- Mahato (1979) used a linear sweep technique to sim- ferent positively charged ions influence the ratio between ␣ ␤ ulate the discharge cycling of a lead–acid battery. The the two lead dioxide phases, - and -PbO2. Of these ␣ cyclic corrosion process taking place at the lead–lead two phases, -PbO2 has a more compact structure, which dioxide interface was controlled by an inner corrosion allows for a desirable contact between the particles. ␣ ␤ ␣ product layer of -PbO2 and an outer layer of -PbO2. This structure makes the phase harder to discharge Velinchenko and co-workers (1995) studied the elec- (Bode, 1977). However, Ruetschi (1992) found that ␣-

2005 © American Water Works Association LYTLE & SCHOCK | PEER-REVIEWED | 97:11 • JOURNAL AWWA | NOVEMBER 2005 105 Mosseri and colleagues (1990) found evidence that at neutral to FIGURE 2 X-ray diffractogram of lead solids during experiment A slightly alkaline pH, hydroxide rad- icals reacted quickly with divalent lead ions in solution to form a tran- sient Pb(III) form, which then dis- Day 395 proportionated and dehydrated into 7,500 Day 145 Pb(IV) species. Ultimately, after fur- ther dehydration into a lead dioxide Day 82 solid, the transformation ended with 5,000 a semiconductorlike (PbO ) Day 64 2 n. The addition of phosphoric acid Day 18 to the sulfuric acid electrolyte or to counts 2,500 Day 7 the positive active material of a lead–acid battery has been practiced Day 1 for many years, and its effect on

Intensity— Day 0 0 PbO2 has been studied extensively. 76-0564> Plattnerite, synthetic (PbO ) 2 Some inconsistency is apparent in

76-2056> Cerussite (PbCO3) the interpretation of reported results. Phosphate-based compounds have 72-2440> Scrutinyite, synthetic (PbO ) 2 been reported to increase cycle life

75-0991> Hydrocerussite [Pb3(CO3)2(OH)2] by reducing the seeding and irre- versible sulfation of the positive-plate 10 20 30 40 50 60 70 80 90 active material (Bullock, 1979a). Bul- 2 θ—degrees lock and McClelland (1977) also Dissolved inorganic carbon = 10 mg/L C, temperature = 24˚C, pH = 7.75–8.19 proposed that phosphate reversibly absorbs to PbO2 during charge while also modifying crystal growth. PbO2 with absorbed phosphate, however, was not easily reduced to PbSO4. PbO2 had a higher capacity for discharge when materi- Therefore the insulating sulfate layer between the positive als were of equal surface area and under conditions of grid and the active material was not formed during reduc- equal acid access. tion (Bullock, 1979a). Pavlov et al (2004) extensively studied the structure of Using a lead–acid storage battery, Bullock studied the lead sulfate (PbSO4) and PbO2 solids on battery elec- effect of constant potential corrosion on the positive grid. trode surfaces. Precise extrapolation is difficult in Phosphate was incorporated into the PbO2 structure dur- instances of neutral and basic pH values in which diva- ing corrosion, which affected its morphology. The phos- lent lead solubility and aqueous speciation are consid- phate-modified PbO2 discharged to PbSO4 at a slower erably different and background concentrations of sulfate rate than it would have if formed in a pure electrolyte. In are orders of magnitude lower. However, the research an extension of earlier work, Bullock (1979b) consid- identified several important factors in scale structure ered how the phosphate mechanism affects the positive that are likely relevant. First, the PbO2 structure and electrode in the lead–acid battery and found that phos- morphology exert an influence on its electrochemical phate produced an intermediate tertiary lead phosphate activity. Second, the lead oxides react with water to form [Pb3 (PO4)2] solid between the oxidation of Pb to PbO2. hydrated surface layers, in which lead dioxide particles The resulting PbO2 formation resisted reduction to PbSO4 consist of a mixture of crystalline PbO2 and “gel zones,” and caused a battery failure from the formation of PbSO4 described as PbO(OH)2·H2O. The gel zones actively film on the active material interface. This failure was conduct both electrons and ions. Another equilibrium thought to be attributable to the lower surface area and exists between the gel layer and the ions in the water. porosity of the resulting PbO2 (Bullock, 1979b). Alter- An implication of this structure is that the electrochem- natively, previous work has suggested that phosphoric ␣ ␤ ical reactivity—Pb(II)–Pb(IV) interconversion—is strongly acid increases the ratio of - to -PbO2 of corrosion affected by the relative amounts of gel zone versus crys- scales (Bode & Voss, 1962). talline material in the scale layer particles. The researchers Lead dioxide voltammetric electrodes have been used also found that Pb(IV) formation seemed to proceed to measure polyphosphate concentrations in solutions through a step in which Pb(II) ions passed through solu- (Karweik & Huber, 1978; Huber & El-Meligy, 1969). tion before being oxidized. The benefit of phosphate is that it approaches a limited

2005 © American Water Works Association

106 NOVEMBER 2005 | JOURNAL AWWA • 97:11 | PEER-REVIEWED | LYTLE & SCHOCK concentration at 0.8%. Once the PbO2 surface is saturated with phosphate, higher concentrations may FIGURE 3 X-ray diffractogram of lead dioxide (PbO ) collected in be detrimental to battery perfor- 2 experiment A compared with a plattnerite PbO standard mance. Furthermore, increased poly- 2 phosphate may solubilize Pb(IV). The cathodic reduction of Pb(IV) to Pb(II) on electroplated PbO2 elec- 750 trodes has been shown to be Day 82 enhanced by polyphosphates, which was thought to be in response to the stabilization of Pb(II) products 500 and/or adsorption effects (Huber et

al, 1976; Wopschall & Shain, 1967). counts Karweik and Huber (1978) examined the reduction of electroplated 250 ␣ ␣ ␤ Pb(IV) oxides (i.e., -, – mix-, Intensity— ␤-PbO ) in the presence of ortho- 2 PbO standard phosphate and several polyphos- 2 0 ␤ 89-2805> Plattnerite (PbO ) phate formulations. The -PbO2 2 electrode had the largest sensitivity and responded to the lowest con- 72-2440> Scrutinyite, synthetic (PbO2) centration of polyphosphates and was implemented for subsequent 10 20 30 40 50 60 70 80 90 2 θ—degrees work. Current potential curve com- parisons indicate that the relative enhancement to reduce Pb(IV) occurred in the following order: P3O10 > P2O7 > P4O13 > NH3. Overall, polyphosphates and Berger (1972) detected soluble Pb(IV) species in sul- were more selective than orthophosphates. furic acid only when phosphoric acid was present. How- Although the reported experimental systems are not ever, no stability constants (or Gibbs free energies) and close analogs to drinking water, the body of data sug- very few data exist for tetravalent lead hydroxide species, gests that phosphate corrosion inhibitors could have specifically tetravalent lead carbonate and phosphate an observable effect on Pb(IV) solid formation rates complexes. and stability. The ORP at which PbO2 becomes stable is at very MATERIALS AND METHODS high potentials starting near the upper stability boundary A series of long-term precipitation experiments was for water (Figure 1). The high redox potential necessary conducted to evaluate the effects of water chemistry and to achieve PbO2 formation in water can only be met with time on the mineralogy and corresponding solubilities of the use of strong oxidants (e.g., free chlorine, chlorine lead solids in water. During these tests, redox conditions dioxide) and their persistence into the distribution system. were chosen to represent the high end of realistic utility

Since passage of the Lead and Copper Rule, the understanding of relationships between water quality and the solubility of lead-containing minerals found in drinking water distribution systems has grown extensively.

Because there are few data on the aqueous speciation practice and were maintained with the intent to form ␤ ␣ of Pb(IV), it is impossible to confidently predict PbO2 Pb(IV) solids ( -PbO2 and/or -PbO2). High redox poten- solubility in water. Pb(IV) complexes containing high tial of the water was maintained with free chlorine, which phosphate concentrations have been produced chemi- was replenished as needed to maintain a goal concentra- cally and electrochemically (Laitinen & Watkins, 1975; tion of 3 mg/L Cl2 throughout most of the studies. Late Huber & El-Meligy, 1969; Fleischmann et al, 1967; Bode into the test runs, chlorine residuals were permitted to & Voss, 1962). Using polarographic techniques, Amlie dissipate to test the reversibility of mineral-phase devel-

2005 © American Water Works Association LYTLE & SCHOCK | PEER-REVIEWED | 97:11 • JOURNAL AWWA | NOVEMBER 2005 107 changes brought about by the addition of lead and chem- FIGURE 4 Soluble lead measured over time during ical reactions; adjustments were made by adding small experiment A increments of acid or base. The computer software7 recorded pH values and titrant volumes, which were maintained in a data file. 0.16 Appropriate amounts of sodium bicarbonate, sodium sulfate, and sodium hypochlorite were then added to 3 L 0.14 of deionized water. The titration system was programmed to the desired pH level and initiated. After the pH stabi- 0.12 lized, lead chloride (PbCl2) was added to provide an initial lead concentration of 20 mg/L. Grab samples were 0.10 drawn out of the cell with a syringe to collect water for

0.08 DIC analysis and free and total chlorine concentrations. —mg/L Additional sample volume was filtered through a 25-mm

Lead 8 0.06 polypropylene syringe disk filter (0.2 µm) to separate colloidal lead for metal analysis. The remaining experi-

0.04 mental water was used to fill ten 250-mL PTFE bottles with no headspace, which were placed in a tumbler and o 0.02 allowed to age at room temperature (23 C). Bottles were periodically removed from the tumbler and analyzed for 0.00 pH, redox potential, soluble lead concentrations, and 0 100 200 300 400 500 free and total chlorine concentrations. If a low chlorine Elapsed Time—days residual was determined (<1 mg/L Cl2), chlorine stock Dissolved inorganic carbon = 10 mg/L C, temperature = 24˚C, solution was added to the remaining experimental bottles pH = 7.75–8.19 to bring the concentration to 3 mg/L. Polypropelene syringe filters were used to differentiate colloidal lead from soluble lead. The remaining experimental water was filtered through a 0.2-µm polycarbonate membrane filter opment. One run was performed in water containing sul- using a tabletop vacuum pump. Once the precipitation fate to investigate the role of an additional major anion cake formed, the filter was used to conduct solid char- on the development of lead solids. Pb(II) sulfate and acterization analysis. hydroxl-carbonate-sulfate solids have been observed in Analytical methods. The pH was measured with a some lead drinking water pipes. All experiments were benchtop pH/ion-selective electrode meter9 with a com- conducted at room temperature (~23oC). The chemicals, bination pH electrode10 and a portable combination test protocol, and analytical methods used for this study pH/dissolved oxygen (DO) meter11 with temperature cor- are detailed in subsequent sections. rections. The instruments were standardized daily using Chemicals. Unless otherwise specified, all chemicals a two-point calibration with pH 7 and 10 standard solu- used in this study were analytical reagent-grade. The tions.12 DO was measured with a DO meter13 and amount of ultrapure nitric acid1 used to preserve sam- probe.14 These instruments were internally standardized. ples for metals analysis was 1.5 mL/L per sample. Dilute Redox potential was measured with a pH/ion meter15 0.6 M hydrochloric acid2 and 0.5 N sodium hydroxide3 with platinum combination redox electrodes.16 The instru- were used to adjust the pH, and sodium bicarbonate3 ment was standardized with Zobells standard solution. was used to adjust water quality. Sulfate was added to test ORP was converted to EH by adding 241 mV according water as sodium sulfate.3 Sodium hypochlorite (4–6% to the manufacturer’s specifications. NaOCl, purified-grade)3 was used to oxidize Pb(II), which Sulfur (sulfate) was analyzed with a purged induc- was added as lead chloride.4 Double-deionized water was tively coupled argon plasma spectrometer.17 Lead was prepared by passing building demineralized water through analyzed with a graphite furnace atomic absorption spec- a cartridge deionized water system5 with a resistivity Ն trophotometer,18 using method 7421 (USEPA, 1986). 18.2 M⍀·cm. Free and total chlorine were measured with a spec- Test protocol. Lead aging studies were conducted in a trophotometer19 using the DPD method (Standard Meth- 3-L glass beaker. Secured at the top of the beaker were a ods, 1998). Total inorganic carbon was analyzed by a pH electrode and two redox electrodes, dissolved oxy- coulometric procedure on a carbon dioxide (CO2) gen (DO) and temperature probes, a mechanical stirrer, coulometer with an acidification module20 operated under and an injection line for both acid and base. A computer computer control. software-controlled dual titrator system6 was used to Magnified images of the lead precipitate were col- adjust the initial pH and rapidly compensate for pH lected by a stereo microscope21 linked to a camera and

2005 © American Water Works Association 108 NOVEMBER 2005 | JOURNAL AWWA • 97:11 | PEER-REVIEWED | LYTLE & SCHOCK software.22 An image magnification range of 10–63× was used. FIGURE 5 X-ray diffractogram of lead solids during experiment B XRD was used to identify crystalline phases of lead solids. The ␪–␪ 23 5,000 diffractometer used a Cu K␣ radi- Day 95 ation X-ray tube at 35 kV and 40 4,500 4,000 Day 74 mA; scans were typically over the 3,500 Day 61 range of 5 to 90 degrees 2␪, with 3,000 Day 33 0.02-degree step sizes that were held 2,500 Day 21 for 2 s each. Pattern analysis was 2,000 1,500 Day 14 performed using software,24 which counts 1,000 Day 4 generally followed procedures spec- 500 Day 0 ified by the American Society for 0

Intensity— 47-1734> Cerussite, synthetic (PbCO ) Testing and Materials. Reference pat- 3 terns were from a 2002 release from 35-1222> Plattnerite, synthetic (PbO2) the International Center for Diffrac- 13-0131> Hydrocerussite, synthetic [Pb3(CO3)2(OH)2] tion Data, Newtown Square, Pa. 45-1416> Scrutinyite (PbO ) Because of some peak position over- 2 laps in the XRD analyses of the pre- 10 20 30 40 50 60 70 80 90 cipitates and the possibility of for- 2 θ—degrees mation of amorphous Pb(IV) phases, Dissolved inorganic carbon = 10 mg/L C, temperature = 24˚C, pH = 6.65–7.51 results were corroborated using Pb (13035 eV) LIII x-ray absorption near-edge spectroscopy analysis at Sector 20-BM, Pacific Northwest Consortium Collabo- demonstrate conditions and pathways that Pb(IV) oxides rative Access Team of the Advanced Photon Source at form in water as well as corresponding lead solubility Argonne National Laboratory, Argonne, Ill. Filter mate- changes. rials on the mounts used for XRD analyses were inserted Experiment A. Experiment A was conducted initially directly into the sample holders with no additional prepa- at pH 8 in water with a DIC concentration of 10 mg/L ration. Scans were run in triplicate and averaged for each carbon (C). During this run, the final pH was relatively specimen. consistent (pH 7.75–8.19) throughout the entire 397-day test period. Free chlorine was adjusted on days 26 and 105 RESULTS to maintain the redox potential of the water and to regain Table 1 summarizes the lead precipitation test runs lost chlorine. Chlorine was completely consumed during conducted in this study. Typically, these runs lasted more the days before these readjustments. Because no effort than 240 days and were conducted at three initial pH was made to maintain chlorine residual beyond 105 days, values: 6.5, 8, and 10. Because no attempt was made to chlorine levels were completely depleted sometime maintain constant pH levels during each test run, however, between days 105 and 395. The remaining sample at the the final pH values varied to some degree from the initial end of the study had consumed nearly 9 mg/L Cl2. The

Although most water treatment and water chemistry specialists are familiar with the radically different solubilities and scale formation properties of ferrous iron versus ferric iron or cuprous copper versus cupric copper, the same potential behavior of lead is less well known. values (Table 1). The authors recognize that control of pH redox potential of the water was ~942 mV (versus stan- and ORP during the experiments was limited and the fol- dard hydrogen electrode [SHE]) when the maximum free lowing results should be interpreted accordingly. Because chlorine dose of 3 mg/L was achieved and was ~440 mV of the difficulty and sensitivity of the experiments and (versus SHE) at the end of the study when no chlorine was the time necessary to achieve anticipated results, it was present. nearly impossible to maintain initial water quality con- The color of the lead precipitate visible in the water ditions. Nonetheless, at a minimum the findings do sample bottles changed gradually but dramatically from

2005 © American Water Works Association LYTLE & SCHOCK | PEER-REVIEWED | 97:11 • JOURNAL AWWA | NOVEMBER 2005 109 is a dark reddish brown color. However, color is affected by crys- FIGURE 6 X-ray diffractogram of lead solids during experiment C tal size, impurities, and other factors. The XRD scan from day 82

7,000 was plotted against the pattern of Day 93 a ␤-PbO standard (Figure 3) to 6,000 2 Day 70 provide an additional reference and 5,000 evidence for the presence of PbO . Day 49 2 4,000 Although the two XRD patterns Day 23 ␤ 3,000 are identical, the -PbO2 standard Day 14 was black in color and the domi-

counts 2,000 Day 7 nant color of the experimentally 1,000 Day 0 precipitated lead was dark red. 0 11-0549> Scrutinyite (PbO ) After 82 days, the XRD pattern Intensity— 2 scan was still dominated by ␤- 35-1222> Plattnerite (PbO2) ␣ PbO2 and -PbO2. The presence ␣ 13-0131> Hydrocerussite [Pb3(CO3)2(OH)2] of -PbO2 became more prominent over time, suggesting that it may 47-1734> Cerussite (PbCO3) be an aging product of PbO2 and 10 20 30 40 50 60 70 80 90 the more thermodynamically sta- 2 θ—degrees ble phase. Once again, there was

Dissolved inorganic carbon = 10 mg/L C, temperature = 24˚C, pH = 7.71–8.6, 210–218 no evidence for the presence of ␤ mg/L sulfate Pb3(CO3)2(OH)2. On day 145, - ␣ PbO2 and some -PbO2 continued to dominate the mineralogical com- position of the lead solids, and a trace of PbCO3 reappeared at this white to dark orange-red in color over the first 82 days time. This observation coincided with the color changes of the study (see photo on page 103). The dark red color and suggests that the formation of PbO2 is reversible somewhat resembled the recognizable color of Fe(III) (see photo on page 103). At the end of the study (day precipitates and corrosion deposits. The red solid grad- 397), PbCO3 was the only lead solid mineral identified. ually disappeared (days 145 and 397) as chlorine was The XRD pattern of PbCO3 at this time was sharper consumed to the eventual point of depletion. At the com- and more complete (based on the number of pattern pletion of the study, only white lead solids were filtered lines) than the patterns identified earlier. Additionally, all from the water, and no visible red color was observed. traces of PbO2 phases were gone. This established that lead particle transformations were Throughout this aging test run, soluble lead concen- both reversible and redox-dependent. The stereomicro- trations (given a 0.2-µm filter) showed a dramatic change graphs (under low magnification, 65×) show the lead in lead solubility (Figure 4) corresponding directly to the particles that were collected from the filters at five dif- change in lead solid color and mineralogy (Figure 2 and ferent times during the study (see parts A and B of photo photo on page 103). Soluble lead concentration began on page 103). Dark red and almost black particles were at ~0.145 mg/L and dropped to a minimum of 0.002 dispersed throughout the filters, and some were as large mg/L by day 82, coinciding with maximum transition to as 50 µm. PbO2. Lead concentrations were greater after 118 days XRD analysis of the filtered solids showed the and eventually increased to 0.113 mg/L by the end of the dynamic nature of lead mineralogy over the aging period study as PbO2 transformed to PbCO3. (Figure 2). During the first seven days, hydrocerussite Experiment B. Experiment B was conducted at an ini- [Pb3(CO3)2(OH)2] was the dominant lead phase pre- tial pH of 6.65 in water having a DIC of 10 mg/L C. The sent and a trace amount of PbCO3 was also detected. On pH of the water during the run was not as consistent as day 18, the major solid phase shifted to PbCO3, and the pH during experiment A, increasing to as high as only a trace amount of Pb3(CO3)2(OH)2 was detected. 7.56 over the aging period. No chlorine adjustment By day 64, neither cerussite nor Pb3(CO3)2(OH)2 was was made through the first 140 days. Consequently, ␤ present, but -PbO2 was detected, along with a small free chlorine levels dropped from 3 mg/L to only 1.4 ␣ ␤ ␣ amount of -PbO2. -PbO2 and -PbO2 are Pb(IV) lead mg/L during the testing period. At the initial pH and oxide polymorphs believed to be responsible for the red chlorine concentration, the redox potential was 1,099 ␤ color of the lead particles. -PbO2 is reported as brown- mV (versus SHE). As in experiment A, the color of the ␣ ish black to black with a red-brown streak, and -PbO2 precipitated lead visible in the sample bottles changed

2005 © American Water Works Association 110 NOVEMBER 2005 | JOURNAL AWWA • 97:11 | PEER-REVIEWED | LYTLE & SCHOCK over time. The lead solids turned from white to dark red; these changes became particularly FIGURE 7 X-ray diffractogram of lead solids during experiment D apparent by day 63. The red color of the solids appeared to be nearly identical to the solids that formed 2,500 in experiment A. Stereomicrograph images of lead 2,000 Day 29 particles were collected as they were trapped on the filter at various 1,500 times throughout the study. Al- Day 19 counts 1,000 though the particles were similar Day 7 in appearance to those collected 500 during experiment A, these particles

Intensity— Day 0 (especially those that were darker in 0 color) were sparse and distinct. 45-1416> Scrutinyite (PbO2) XRD results (Figure 5) showed that 13-0131> Hydrocerussite, synthetic [Pb (CO ) (OH) ] the lead minerals transformed over 3 3 2 2 time from a mixture of PbCO3 and Pb (CO ) (OH) (first day of ex- 10 20 30 40 50 60 70 80 3 3 2 2 2 θ—degrees periment) to PbCO3 alone (days 4 and 14) and then back to a mixture Dissolved inorganic carbon = 10 mg/L C, temperature = 24˚C, pH = 9.62–9.86 of PbCO3 and Pb3(CO3)2(OH)2 on ␤ days 21 and 33. By day 61, -PbO2 was present along with PbCO3 and ␣ a trace amount of -PbO2. This mixture of solids was present in all runs until the end of present for the remainder of the study. The presence of ␤- ␣ the experiment. Relative changes in XRD pattern intensities PbO2 and -PbO2 increased over time, and these materi- indicated that the prevalence of both PbO2 phases als were also present through the remaining study period. ␤ appeared to increase with time. Day 70 showed the strongest pattern for -PbO2, which Experiment C. Experiment C was conducted to eval- also corresponded with the large number of dark parti- uate the effect of sulfate (210–218 mg/L SO4) on the cles observed simultaneously. aging of lead solids at pH 8 (7.75–8.0) and 10 mg/L Experiment D. Lead aging experiment D was con- C. Free chlorine was not adjusted during the entire ducted at an initial pH of 10 (9.62–9.86) in water study period. At 93 days into the study, nearly 1 mg/L with a DIC concentration of 10 mg/L C at 23oC. Chlo- Cl2 was still measured, and by the end of the study no rine dropped rapidly during this test run and had to be free chlorine was present. The initial redox potential replenished after only 29 days. The initial redox poten- was 967 mV (versus SHE) and dropped to 492 mV (ver- tial when free chlorine was 3 mg/L was 742 mV (versus SHE) by the end of the test run. Again, as in previ- sus SHE). Sample color changed from white to light ous experiments, the color of the precipitated lead brownish red after only seven days. This color con- turned from white to dark red. Visual inspection sug- tinued to become increasingly darker until day 30. gested that the red color may have appeared slightly Micrograph images (see photo on page 104) compare earlier than when sulfate was not present (experiment the color of filtered lead solids at pH 8 (experiment A) A). This red color was strongly apparent 14 days into and pH 10 (experiment D). The lead particles formed the run and did not change significantly throughout at pH 10 have a stronger brownish tint than those the rest of the test period. formed at pH 8. Stereomicrographs of magnified lead As in experiment A, a mixture of small, distinct, dark red particles, aging from 0 to 19 days, showed an increased particles sat on a background of lighter red-colored solids presence of small, dark red particles (not shown). Ini- (not shown). At 71 days, there were a large number of tial XRD diffraction analysis showed the presence only large, dark-colored particles present. XRD results (Figure of hydrocerussite (Figure 7). However, by day 7 ␤- ␣ 6) showed that initially (day 0), Pb3(CO3)2(OH)2 was the PbO2 and -PbO2 joined Pb3(CO3)2(OH)2. After 19 ␤ ␣ only solid identified. PbCO3 joined Pb3(CO3)2(OH)2 as days, -PbO2 disappeared, leaving only -PbO2 and the only lead materials present by day 7. Pb3(CO3)2(OH)2 Pb3(CO3)2(OH)2 in the pattern. Finally, after 29 days ␣ disappeared by day 14, and PbCO3 was the primary solid when Pb3(CO3)2(OH)2 was no longer apparent, - ␤ ␣ identified. Traces of -PbO2 and -PbO2 were also detected. PbO2 was the only identifiable lead mineral phase After this time, PbCO3 continued to be the dominant solid remaining.

2005 © American Water Works Association LYTLE & SCHOCK | PEER-REVIEWED | 97:11 • JOURNAL AWWA | NOVEMBER 2005 111 DISCUSSION 25oC, and no ionic strength corrections. Calculations ␤ Study results indicated that the Pb(IV) oxides -PbO2 were performed using a FORTRAN computer program ␣ and -PbO2 form over time in chlorinated water (<3 adapted for a personal computer (Froning et al, 1976). mg/L Cl2) at pH 6.65–10 (DIC = 10 mg/L C, temperature The redox stability region for PbO2 is seen to widen as o = 23 C). Spontaneous precipitation of PbO2, however, did pH increases. The change in the boundary of the sta- not occur. Even attempts to precipitate crystalline PbO2 bility field is consistent with the experimental observa- in superchlorinated water (~30 mg/L Cl2 at pH 8 and tion that PbO2 formed more readily at a higher pH 10) failed to precipitate PbO2 in short-term experiments (based on chlorine demand, color change, and XRD (<30 min). However, it is possible that amorphous Pb(IV) scans), which could be attributed to a higher level of solids may have spontaneously formed as reported by supersaturation and stronger thermodynamic driving other researchers (Edwards & Dudi, 2004) but could force for the same ORP. not be identified by conventional approaches used in this work. This suggests that the formation of PbO2 in CONCLUSIONS chlorinated water appears to require time (“aging”) to The results showed that the Pb(IV) oxides plattnerite overcome kinetic barriers and may require precursor and scrutinyite (PbO2) will form over time in chlorinated Pb(II) mineral phases. The rate of aging and mineral- water (<3 mg/L Cl2) at pH 6.65–10. These findings raise a number of potential prac- tical implications. The pri- The effect of sulfate needs to be re-examined mary concern is the redox under more controlled conditions before control on lead mineralogy and its corresponding solu- conclusions can be confidently drawn. bility. Distribution systems that have historically main- phase development sequence, however, was strongly pH- tained high chlorine residuals could have PbO2 func- dependent. Results indicate that the rate of PbO2 for- tioning to limit lead release, while they are presuming mation and chlorine consumption increased when pH hydrocerussite or cerussite to be the main components of increased. The sequence of lead mineral-phase develop- the passivating films or other diffusion barriers. As noted ⇒ ⇒ ment at pH 6.5–8 was Pb3(CO3)2(OH)2 PbCO3 previously, the relative ease with which PbO2 can form ␤ ⇒ ␣ -PbO2 -PbO2. at neutral pH in waters with persistent free chlorine Initially, Pb3(CO3)2(OH)2 and PbCO3 coexisted, but residuals suggests that PbO2 likely occurs in many systems over time Pb3(CO3)2(OH)2 converted (i.e., dissolved, with substantial lead piping and where lead concentra- reprecipitated, reordered, dehydrated) to the more insol- tions are unusually low, relative to what would be uble, thermodynamically favored PbCO3 as predicted expected for the divalent lead carbonate or hydroxycar- (Schock et al, 1996). Only PbCO3 was a precursor to or bonate solubility at the ambient pH and alkalinity. coexisted with PbO2. Plattnerite appeared to form earlier Changes in pH, drops in ORP, or both could destabilize than scrutinyite and developed over time. The transfor- these PbO2 films and thus increase plumbosolvency and mation of Pb(II) to Pb(IV) was complete at pH 6–6.5. possibly increase the release of scale particles (Edwards Sulfate may have slightly increased the rate of Pb(II) oxi- & Dudi, 2004; Schock & Giani, 2004). Some water sys- dation to Pb(IV) and affected the completeness of Pb(II) tems have observed erratic lead release from lead ser- transformation [i.e., complete conversion of Pb(II) to vice lines and increases or decreases without clear cor- Pb(IV) was not achieved at pH 8]. The effect of sulfate relations with pH, DIC, and temperature; these needs to be reexamined under more controlled condi- developments may be caused at least in part by effects of tions before conclusions can be confidently drawn. ORP changes. Water systems may have a false sense of The formation of PbO2 was completely reversible to security about their corrosion control strategies because cerussite in a relatively short time period when the redox of a lack of observation and understanding of the pipe potential in the system was not sufficiently maintained. deposit mineralogy. The aging of lead to PbO2 was indicated by a change in Substantial research still needs to be conducted to pre- solid color (white to dark red or brownish red, depend- cisely determine the ORP threshold for PbO2-phase sta- ing on pH) and lead solubility as predicted by Schock bility, its solubility in relation to pH, and the relation- (2001). For example, at a pH of 8, ClO2 and Cl2 at appro- ship to disinfectant type, residual, and oxidant demand in priate dosages are able to achieve the redox potentials the pipe scale. Furthermore, the decomposition pathway sufficient for PbO2 stability, whereas oxygen and mono- of PbO2 scales must be elucidated because it may pose chloramine cannot. serious problems with treatment changes that could result Figure 1 shows an electromotive force–pH diagram in lowering of the redox potential. for lead in water that contains a DIC concentration of This research shows that PbO2 formation can occur 10 mg/L C, lead species of 0.015 mg/L, temperature of in weeks to months and that suspended PbO2 itself can

2005 © American Water Works Association 112 NOVEMBER 2005 | JOURNAL AWWA • 97:11 | PEER-REVIEWED | LYTLE & SCHOCK revert or decompose in a similar timeframe with the mitigate adverse effects on lead concentrations from depletion of oxidant and consequent lowering of solution some treatment changes. ORP. If the scale does act as a semiconductor (as is suggested in the literature), then there is likely a delicate ACKNOWLEDGMENT equilibrium to maintain a stable elevated redox potential The authors thank Barbara Francis, Jeremy Payne, Jessica through the scale, bridging the strongly reducing lead Lewis, and Keith Kelty for water quality and solids analy- metal at the pipe wall contact and the water solution sis and Christy Muhlen of the USEPA for her assistance in interface. The scale can rapidly decompose or reform as operating the experimental system. They thank Kathrine long as it is almost entirely composed of conductive McCullough of the University of Miami (Ohio) for assistance forms of PbO2. Prolonged periods of depleted oxidant at with preparing and editing the manuscript. They also thank the water side, however, would perturb the equilibrium their drinking water research colleagues, especially (and in and leave a strong thermodynamic gradient to promote no specific order) Marc Edwards, Rhodes Trussell, Phil the chemical reduction (electron acceptance) in the scale Singer, Gregory Korshin, and Steve Reiber. Finally, use of the by the effects of the pipe metal itself. Destabilization Advanced Photon Source was supported by the US Depart- carries many risks, given the higher solubility of most ment of Energy, Office of Science, Office of Basic Energy Sci- divalent lead compounds relative to the predicted and ences, under contract W-31-109-Eng-38; the assistance of observed solubility of the tetravalent lead compounds the Pacific Northwest Consortium Collaborative Access in most pH and DIC conditions. Identifying the nature of Team staff at Sector 20-BM is acknowledged. Any opinions the breakdown products and the rates of their reactivity expressed in this article are those of the authors and do with passivating ions (e.g., carbonate, hydroxide, and not necessarily reflect the official positions and policies of orthophosphate) in solution is critical to understanding USEPA. Any mention of products or trade names does not how to either stabilize the tetravalent lead films or to constitute recommendation for use by USEPA.

REFERENCES Water Utility Process Research Solu- lation of Potential-pH Diagrams of tions, Madison, Wis. Metal–Ion–Water Systems by Computer. Amlie, R.F. & Berger, T.A., 1972. Polarographic Corrosion Sci., 16:371. Analysis of Lead(IV) Species in Solutions Delahay, P.; Pourbaix, M.; & Rysselberghe, Containing Sulfuric and Phosphoric P.V., 1951. Potential-pH Diagram of Lead Greninger, D.; Kollonitsch, V.; & Kline, C.H., Acids. Jour. Electroanal. Chem. & Inter- and Its Applications to the Study of Lead 1975. Lead Chemicals. Intl. Lead Zinc facial Electrochem., 36:2:427. Corrosion and to the Lead Storage Bat- Res. Org., New York. tery. Jour. Electrochem. Soc., 98:57. Bagshaw, N.E.; Clarke, R.L.; & Halliwell, B., Hampson, N.A., 1979. Aqueous System 4+ 2+ 1966. Preparation of Lead Dioxide for X- Dodrill, D.M., 1995. Lead and Copper Corro- Pb /Pb : Electrochemical Aspects. The Ray Diffraction Studies. Jour. Appl. sion Control Based on Utility Experience. Electrochemistry of Lead. Academic Chem., 16:180. Master’s thesis, Dept. of Science in Press, New York. Envir. Engrg., Univ. of Colorado, Boulder. Bode, H., 1977. Lead–Acid Batteries. John Huber, C.O.; Berger, T.A.; & Reim, R.E., 1976. Wiley & Sons, New York. Dodrill, D.M. & Edwards, M., 1995. Corrosion Electrocatalysis on Non-metallic Sur- Control on the Basis of Utility Experi- faces. US Natl. Bur. Stand., Spec. Publ., Bode, H. & Voss, E., 1962. Blei(IV)-Phosphate, ence. Jour. AWWA, 87:7:74. 455:183. ein Beitrag zur Kenntnis der Reaktionen Dodrill, D.M. & Edwards, M., 1994. A General Huber, V.F. & El-Meligy, M.S.A., 1969. Uber an der Bleidioxydelektrode. Electrochim. Framework for Corrosion Control. Proc. Blei(IV)-Phosphatverbindungen. Acta, 6:11. 1994 AWWA WQTC, San Francisco. Zeitschrift fur Anogranische und Allege- meine Chemie, 367:154. Bullock, K.R., 1979a. Effect of Phosphoric Edwards, M. & Dudi, A. 2004. Role of Chlorine Acid on the Positive Electrode in the and Chloramine in Corrosion of Lead- Karweik, D.H. & Huber, C.O., 1978. Polyphos- Lead–Acid Battery II. Constant Potential bearing Plumbing Materials. Jour. phate Currents at the Lead Dioxide Elec- Corrosion Studies. Jour. Electrochem. AWWA, 96:10:69. trode for Hydrolysis Rate and Instability Soc., 126:3:360. Edwards, M.; Jacobs, S.; & Dodrill, D., 1999. Constant Measurements. Jour. Anal. Bullock, K.R., 1979b. Effect of Phosphoric Desktop Guidance in Mitigation of Pb Chem., 50:8:1209. Acid on the Positive Electrode in the and Cu Corrosion By-products. Jour. Laitinen, H.A. & Watkins, N.H., 1975. Cathodic Lead–Acid Battery III. Mechanism. Jour. AWWA, 91:5:66. Stripping Coulometry of Lead. Jour. Anal. Electrochem. Soc., 126:11:1848. Fleischmann, M. et al, 1967. Investigation of Chem., 47:1352. Bullock, K.R. & McClelland, D.H., 1977. Effect the Kinetics of Electrode Reactions by Mahato, B.K., 1979. Cyclic Corrosion of the of Phosphoric Acid on the Positive the Application of Repetitive Square Lead–Acid Battery Positive. Jour. Elec- Electrode in the Lead–Acid Battery. Pulses of Potential. Electrochim. Acta, trochem. Soc., 126:3:365. Jour. Electrochem. Soc., 124:10:1478. 12:967. Matsumoto, Y.; Noguchi, M.; & Matsunaga, T., Cantor, A.F., 2004. Monitoring of Lead Service Froning, M.H.; Shanley, M.E.; & Verink, E.D. 1999. Deposition of Metallic Oxides on Line Replacement. Rept. for Madison Jr., 1976. An Improved Method for Calcu- TiO2 Electrode Using the Photoelectro-

LYTLE & SCHOCK | PEER-REVIEWED | 97:11 • JOURNAL AWWA | NOVEMBER 2005 113 ABOUT THE AUTHORS 4Malinckrodt Inc., Paris, Ky. 5Milli-Q Plus© Millipore Corp., Bedford, Mass. Darren A. Lytle (to whom corre- 6Schott Geräte, Mainz, Germany spondence should be addressed) is 7Jenson Systems, Hamburg, Germany 8Whatman Inc., Clifton, N.J. an environmental engineer with the 9Model 50125, EC40, Hach Co., Loveland, Colo. US Environmental Protection 10Model 48600, Hach Co., Loveland, Colo. 11Sension 156, Hach Co., Loveland, Colo. Agency (USEPA), 26 W. Martin 12Whatman, Hillsboro, Ore. Luther King Dr., Cincinnati, OH 13Model DO175, Hach Co., Loveland, Colo. 14Model 50180, Hach Co., Loveland, Colo. 45268; e-mail [email protected]. 15Model 450, Corning, N.Y. A member of the agency’s drinking 16Model 9678BN, Thermo-Orion, Buerly, Mass. 1761E®, Thermo Jarrel Ash, Franklin, Mass. water research group for more than 15 years, he has 18Model SH-4000, Thermo Jarrel Ash, Franklin, Mass. focused on filtration, distribution system issues, and 19DR/2000, Hach Company, Loveland, Colo. 20Model 5011 with Model 50 acidification module, UIC, Joliet, Ill. arsenic removal. Lytle has a PhD in environmental 21SM 7800, Nikon, Tokyo, Japan engineering from the University of Illinois at 22Spot Insight camera and software, Diagnostics Instruments, Sterling Heights, Mich. Urbana– Champaign, an MS in environmental engi- 23XDS-2000, Scintag Inc., Santa Clara, Calif. neering from the University of Cincinnati (Ohio), and 24Jade+ software, version 7, MDI Inc., Livermore, Calif. a BS in civil engineering from the University of Akron (Ohio). Michael R. Schock is a research chemist at USEPA.

FOOTNOTES If you have a comment about this article, 1Ultrex, J.T. Baker Chemical Co., Phillipsburg, N.J. please contact us at [email protected]. 2Malinckrodt Inc., Paris, Ky. 3Fischer Scientific, Fairlawn, N.J.

chemical Epitaxial Growth Technique ity and Treatment: A Handbook of Com- Method 4500-C1G. APHA, AWWA, and (PEEG), and Electrochemical Behavior of munity Water Supplies. McGraw-Hill, WEF, Washington. the PbO2/TiO2 Electrode. Jour. Phys. New York. Velinchenko, A.B.; Girenko, D.V.; & Danilov, Chem. B, 103:34:7190. Schock, M.R., 1981. Response of Lead Solubil- F.I., 1995. Electrodeposition of Lead Diox- Mosseri, S.; Henglein, A.; & Janata, E., 1990. ity to Dissolved Carbonate in Drinking ide at an Au Electrode. Elektrochimiya., Trivalent Lead as an Intermediate in the Water (erratum). Jour. AWWA, 73:3:36. 31:2803. Oxidation of Pb(II) and the Reduction of Schock, M.R., 1980. Response of Lead Solubil- USEPA (US Environmental Protection Pb(IV) Species. Jour. Phys. Chem., ity to Dissolved Carbonate in Drinking Agency), 1986 (3rd ed.). Test Methods for 94:6:2722. Water. Jour. AWWA, 72:12:695. Evaluating Solid Waste, Physical, Chemi- Pavlov, D. et al, 2004. Influence of H SO Con- cal Methods. Publication SW-846. 2 4 Schock, M.R. & Clement, J.A., 1998. Control of centration on the Mechanism of the USEPA, Washington. Processes and on the Electrochemical Lead and Copper With Non-Zinc Activity of the Pb/PbO /PbSO Electrode. Orthophosphate. Jour. NEWWA, 112:1:20. USEPA, 1992. Drinking Water Regulations. 2 4 Maximum Contaminant Level Goals Jour. Power Sources, 137:2:288. Schock, M.R. & Gardels, M.C., 1983. Plumbo- and National Primary Drinking Water solvency Reduction by High pH and Low Petersson, I.; Ahlberg, E.; & Berghult, B., 1998. Regulations for Lead and Copper. Final Carbonate-Solubility Relationships. Jour. Parameters Influencing the Ratio Rule; correcting amendments. CFR 40, AWWA, 75:2:87. Between Electrochemically Formed ␣- Parts 141 and 142, Fed. Reg., ␤ and -PbO2. Jour. Power Sources, Schock, M.R. & Giani, R., 2004. Oxidant/Disin- 57:125:28785. 76:1:98. fectant Chemistry and Impacts on Lead USEPA, 1991a. Drinking Water Regulations. Corrosion. Proc. 2004 AWWA WQTC, San Pourbaix, M. et al, 1966. Lead. Atlas of Elec- Maximum Contaminant Level Goals and Antonio. trochemical Equilibria in Aqueous Solu- National Primary Drinking Water Regu- tions. Natl. Assn. of Corrosion Engrs., Schock, M.R. & Wagner, I., 1985. Corrosion lations for Lead and Copper. Final Rule; Houston. and Solubility of Lead in Drinking Water. correction. CFR 40, Parts 141 and 142. Reutschi, P., 1992. Influence of Crystal Struc- Internal Corrosion of Water Distribution Fed. Reg., 56:135:32112. Systems. AwwaRF/DVGW ture and Interparticle Contact on the USEPA, 1991b. Maximum Contaminant Forschungsstelle, Denver. Capacity of PbO2 Electrodes. Jour. Elec- Level Goals and National Primary trochem. Soc., 139:1437. Schock, M.R.; Wagner, I.; & Oliphant, R., 1996 Drinking Water Regulations for Lead Schock, M.R., 2001. Tetravalent Lead: A Hith- (2nd ed.). Corrosion and Solubility of and Copper. Final Rule. Fed. Reg., erto Unrecognized Control of Tap Water Lead in Drinking Water. Internal Corro- 56:110:26460. sion of Water Distribution Systems. Lead Contamination. Proc. 2001 AWWA Wopschall, R.H. & Shain, I., 1967. Effects of AwwaRF/TZW, Denver. WQTC, Nashville, Tenn. Adsorption of Electroactive Stationary Schock, M.R., 1999 (5th ed.). Internal Corro- Standard Methods for the Examination of Electrode Polargraphy. Anal. Chem., sion and Deposition Control. Water Qual- Water and Wastewater, 1998 (20th ed.). 39:1514.