Environ. Sci. Technol. 2004, 38, 5584-5590
Lead Phosphate Minerals: Solubility and Dissolution by Model and Natural Ligands C A R M E N E N I D M A R T IÄ N E Z , * , † ASTRID R. JACOBSON,‡ AND MURRAY B. MCBRIDE‡ Department of Crop and Soil Sciences, The Pennsylvania State University, University Park, Pennsylvania 16802, and Department of Crop and Soil Sciences, Cornell University, Ithaca, New York 14853
Due to their relatively low solubility, lead-phosphate minerals may control Pb solution levels at a low value in natural environments. We report the solubility of Pb from two lead-orthophosphate mineral suspensions (β-Pb9(PO4)6 and PbHPO4) after aging for 3 years. Lead (Pb2+) activity in the aged suspensions was compared to the activity calculated using the Ksp values of various Pb-PO4 minerals reported in the literature. We also determine the timedependent dissolution of the aged lead-phosphate minerals by organic and inorganic ligands containing S-functional groups (cysteine, methionine, and thiosulfate) and by a soil extracted humic acid. We find the activity of Pb2+ in the aged lead-phosphate suspensions to be 1-2 orders of magnitude higher than predicted by the Ksp values reported in the literature. Disagreement between measured and Kspcalculated activities has been reported in other investigations of Pb-PO4 minerals; we compiled some of the data and present them together with our results. Furthermore, the time-dependent dissolution experiments indicate that, in most cases, lead phosphates are partly dissolved in the presence of soluble ligands, i.e., model sulfides and humic acid. The soil-extracted humic acid enhanced the dissolution of Pb from the high pH (7.2) lead-phosphate (β-Pb9(PO4)6) mineral while suppressing Pb dissolution from the low pH (3.8) lead-phosphate (PbHPO4) mineral. While the low molecular weight sulfur-containing ligands enhanced Pb dissolution, their effect was less pronounced. We conclude that (i) nonequilibrium conditions prevail in the mineral suspensions even after 3 years of aging; and (ii) soluble ligands present in soils, sediments, and natural waters can potentially dissolve Pb from lead-phosphate minerals; such ligands, then, may enhance the biological availability and mobility of Pb in the environment.
Introduction The formation of metal-phosphate precipitates in metal contaminated soils and sediments represents a potentially effective remediation strategy (1-3). Based on the fact that metal phosphate minerals have a relatively low solubility, * Corresponding author phone: (814)863-5394; fax: (814)863-7043; e-mail:
[email protected]. Corresponding author address: 116 ASI Building, The Pennsylvania State University, University Park, PA 16802. † The Pennsylvania State University. ‡ Cornell University. 5584
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 21, 2004
the remediation strategy involves the in situ chemical immobilization of metal contaminants after adding a source of phosphate; in particular lead phosphates, which are among the most stable environmental Pb forms. Thus many remediation strategies promote the formation of Pb-PO4 solid phases in Pb contaminated soils and sediments as a desirable end point. Laboratory investigations have shown reduced Pb solubility after reaction of hydroxyapatite (HA: Ca5(PO4)3OH) with a Pb solution and several retention mechanisms have been proposed (4, 5). Using Rietveld analysis of X-ray diffraction (XRD) patterns, Mavropoulos and co-workers (5) demonstrated that a Pb/Ca solid solution (Pb(10-x)Cax(PO4)6OH2) forms which becomes hydroxypyromorphite (HP: Pb(10)(PO4)6OH2) with time and that both precipitation of HP as well as Pb sorption onto the HA surface are involved. In addition, studies using atomic force microscopy (AFM) showed that some of the Pb is sequestered by the homogeneous nucleation (in solution) and precipitation of HP (4). Independent of the mechanism involved, these studies on Pb removal from solution upon addition of HA point out to a time-dependent process and to the complex nature of the reactions that might occur when HA is added to contaminated soils and sediments. Experimental evidence indicates that lead-phosphates can form after addition of HA to Pb-contaminated soils (68). A reduction in dissolved Pb after addition of phosphate rocks to soils has also been reported (9). It was also observed that various Pb-solid phases are partly transformed to chloropyromorphite (CP: Pb5(PO4)3Cl) after reaction with HA and chloride in model systems (10, 11) and in a Pb contaminated soil (12). The concentration of Cl- (NaCl solution) used in these studies, however, was 1 mM (35.5 mg/L) which raises the question of whether the concentration of native Cl- in most contaminated soils would be sufficiently high to promote the formation of CP. Yet, pure metalphosphate minerals are unlikely to form under the environmental conditions prevailing in most soils and sediments. For instance, Ni- and U-rich secondary Al/Fe-phosphate coprecipitates have been identified in soils and sediments contaminated with a variety of metals after HA addition (2). Ma and co-workers (13, 14) also reported that the formation of HP is affected by the presence of high concentrations of other metals (Al, Fe, Cu, Cd) and CO32-. The concentration of Pb measured in lead-phosphate mineral suspensions or in soil solutions after phosphate additions to Pb contaminated soils can differ from theoretical concentrations calculated using tabulated values of the solubility products, Ksp (10, 15-18). In an attempt to understand Pb solubility controls in soils, a synthetic leadphosphate mineral (16) and a synthetic HA (19) were added to soils. Results from both experiments showed that Pb solubility values were not controlled by chemical equilibrium (Ksp) with pure Pb solid phases. In the study by Zhang and co-workers (19) pyromorphite was not formed at pH 6 although a decrease in dissolved Pb was observed. Sauve´ and co-workers (16) hypothesized that at pH >6 soluble organics dissolve Pb from the phosphate minerals. Hence, we can infer that equilibrium was not attained in the soils under investigation and that other factors (i.e., presence of soluble ligands and other metal cations, additional retention sites and mechanisms) influence Pb solubility. The discrepancy in solubility between measured and theoretical values is not limited to Pb-PO4 minerals. After equilibration of calcium phosphate minerals for up to 6 months for example, Jaynes and co-workers (20) found the IAP (ion activity 10.1021/es049617x CCC: $27.50
2004 American Chemical Society Published on Web 09/24/2004
product) of monetite (CaHPO4) and hydroxyapatite (Ca5(PO4)3OH) to be consistent with previously published values, while whirtlockite (β-Ca3(PO4)2) resulted in lower solubility than published values. Although the solubility of the elements involved might often reach steady-state, the IAP does not necessarily match published values. This is particularly the case in natural environments where mineral phases are most likely formed in the presence of a variety of cations and soluble ligands. Low molecular weight organic and inorganic ligands have been shown to increase (21-24), decrease (25), or have no measurable effect (24, 26-27) on metal dissolution from mineral phases. Humic substances have also been shown to inhibit (25, 28) or enhance the dissolution of (hydr)oxide (29, 30) and sulfide (24, 29) minerals and thus the potential for metal mobility. The effect soluble ligands might have on mineral dissolution has important implications for both the mobilization and availability of metals in soils and sediments; especially for neo-formed minerals of relatively small particle size and reduced crystallinity and in soils and sediments with considerable amounts of organic matter. In this investigation we measure the concentration of Pb in solution and estimate Pb2+ activity for two lead-orthophosphate minerals after aging the mineral suspensions for 3 years. We compare the activity of Pb2+ in the aged suspensions to the activity calculated based on published Ksp (solubility product) values. Because dissolved organic matter (DOM) present in soils, sediments, and natural waters can potentially dissolve lead-phosphate minerals, our second objective is to quantify the dissolution of lead-phosphate minerals by soluble ligands. To fulfill this objective, a timedependent experiment was carried out in which the aged lead-phosphate minerals were reacted with sulfur-containing soluble ligands and with a humic acid extracted from soil.
Experimental Section Solubility of Lead-Phosphate Minerals. Two Pb-orthophosphate minerals were synthesized in the laboratory (15, 16) and equilibrated for 3 years before use in the present investigation. Briefly, a secondary Pb orthophosphate (identified as PbHPO4 by XRD analysis) was synthesized by slow mixing of a Na2HPO4 and a Pb(NO3)2 solution at 80 °C. The suspension was then maintained at low heat (approximately 35 °C) for 3 h. A similar procedure was used to synthesize a tertiary Pb orthophosphate (identified as β-Pb9(PO4)6 by XRD analysis), but Pb acetate was used instead of Pb(NO3)2. Once formed, the β-Pb9(PO4)6 suspension was rinsed with distilled-deionized water (dd-H2O) three times. The suspensions were stored for 3 years at room temperature before use in this study. To compare the solubility of Pb in the aged (3 years) Pborthophosphate suspensions to the solubility of Pb predicted by the solubility products of pure lead-phosphate minerals reported in the literature, an aliquot (25 mL) of the suspensions was sampled, the pH was measured, and the suspensions were centrifuged at 15 000 rpm (27 000 rcf) for 15 min. The supernatant was filtered using a 0.2 µm polycarbonate membrane filter, and the filtered solution was analyzed by ICP (Thermo Jarrell Ash IRIS ICP-OES). The solids (pellet) were freeze-dried and analyzed by X-ray diffraction (CuKR1 radiation) and specific surface area (BET N2 adsorption method). The experimental XRD peaks were matched with JCPDS-ICDD standards. The concentration of Pb measured by ICP analysis was used in the speciation program Visual MINTEQ (31) to calculate the activity of Pb2+ in the filtrates. This calculation accounts for the effect on Pb2+ activity of the concentration of the most relevant elements present in the filtered supernatants (Pb, P, Na, K, Ca, S, Mg as determined by ICP analyses), CO2(g) (0.00035 atm), and the pH of the suspensions. For the PbHPO4 system (pH 3.8), 99.3% of the
FIGURE 1. X-ray diffraction (XRD) patterns of Pb-orthophosphate minerals after equilibration for 3 years. total dissolved Pb was calculated to be in the free ionic form (Pb2+) and 91.6% of the total dissolved P as the H2PO4- soluble species. For the β-Pb9(PO4)6 system (pH 7.2), 54.9% of the total dissolved Pb was calculated to be in the free ionic form (Pb2+), while phosphate (17.8% as PbHPO4 (aq)), hydroxyl (16.9% as PbOH+(aq)), and carbonate (7.1% as PbCO3(aq)) aqueous species accounted for a significant fraction of the total dissolved Pb. The predominant phosphorus soluble species in the β-Pb9(PO4)6 system were HPO42- (53.4%) and H2PO4- (40.9%). The concentration of the free ionic Pb2+ species (Pb activity) was plotted for our samples. The total dissolved P was used in constructing the solubility diagrams of Figure 2. Specifically, we assumed a PO43- activity of 0.001 M to determine the solubility lines of pure Pb phosphate minerals. We also determined the solubility lines using 0.003 M PO43- and found they do not change in any significant manner. Dissolution Experiments. A batch-type reaction was used to monitor the time-dependent dissolution of Pb orthophosphates in the presence of organic and inorganic sulfurcontaining soluble ligands and a humic acid. The leadorthophosphate suspensions were centrifuged, and the solid was freeze-dried before use in dissolution experiments. One hundred milliliters of a solution containing the ligand was added to a polyethylene bottle containing 200 mg of solid and 20 (5 mm) glass beads (to ensure good mixing). The bottle air space was purged with N2(g) to retard oxidation of the solutions. Sets of duplicate solid-ligand suspensions were incubated at 25 °C for 1, 2, 4, 7, 14, 21, and 34 days in orbital shaking (175 rpm) laboratory incubators. At the end of each equilibration period, a subsample (15 mL) was centrifuged at 27 000 rcf (15 000 rpm) for 15 min. The pH of the supernatant was measured, and the supernatant was filtered through a 0.2 µm polycarbonate membrane filter. The filtrates were analyzed by axial-view ICP-optical emission spectrometry (ICP-OES: SPECTRO Analytical Instruments - SPECTRO CIROSCCD fitted with a short depth-of-field lens transfer optic). Dissolved organic carbon (DOC) was determined in the humic acid samples (filtered supernatants) by persulfate oxidation/CO2 analysis (OI Analytical Model 1010 Total Carbon Analyzer). Ligand Solutions. Organic and inorganic ligands containing S-functional groups and a humic acid were used in the dissolution experiments. The ligands include cysteine (C2H4O2N-CH2SH), methionine (C2H4O2N-CH2CH2SCH3), thioVOL. 38, NO. 21, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
5585
FIGURE 2. Panel A shows the activity of Pb in the lead-phosphate mineral suspensions (initial conditions and after 3 years of aging), the results from ref 17, and the solubility lines for various leadphosphate minerals. Panel B presents the Pb solubility results from studies by Zhang and Ryan (10) (diamonds), Stanforth and Qiu (17) (squares), and Lang and Kaupenjohann (18) (triangles). The solubility lines for various lead-phosphate minerals calculated from the Ksp values reported in the literature are also shown. sulfate (Na2S2O3), and humic acid. We chose the amino acids cysteine and methionine because they are present in soil solutions and model S-functional groups present in humic acids. We also use thiosulfate (S-containing inorganic ligand) since a silver-thiosulfate soluble complex has been shown to cross green alga membranes via an anion (SO42-) transporter (32), thus providing a mechanism for the import of a toxic metal. Ligand solutions were prepared to contain equimolar concentrations of total sulfur: 0.5 mM cysteine, 0.5 mM methionine, 0.25 mM Na-thiosulfate, and 1.1055 g of humic acid per L (0.5 mM S). All ligand solutions were prepared at pH 6 in KNO3 (1 mM) background electrolyte. A 1 mM KNO3 solution was used as the control. Humic Acid Preparation and Characterization. The humic acid was extracted from a dark brown (10YR2/3) calcareous soil from Foresta Umbra (Italy) covered with Quercus cerris. The Foresta Umbra humic acid contains 50.3% C, 4.8% H, 3.7% N, 2.9% S, and 38.2% O. It has 8.4% water and 1.2% ash content. A complete description of the humic acid extraction procedure and characterization can be found in Jacobson and co-workers (25).
Results and Discussion Lead-Phosphate Minerals. After the initial laboratory synthesis, the secondary Pb orthophosphate was identified as PbHPO4, while the tertiary Pb orthophosphate was identified as β-Pb9(PO4)6 by XRD analyses of the solids (16). We identified 5586
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 21, 2004
our (aged) tertiary Pb-orthophosphate mineral as β-Pb9(PO4)6 since, although very similar to the XRD pattern of hydroxypyromorphyte [Pb5(PO4)3OH], it lacks a diagnostic peak at approximately 18 °2θ. No change in the identity of the leadphosphate minerals was apparent after equilibration for 3 years at room temperature, yet an increase in the degree of crystallinity was observed by conventional XRD analyses (Figure 1). Evidence of increased crystallinity are increases in the intensity of the main peaks (30.5 and 31.7 °2θ for β-Pb9(PO4)6; 13.5 and 27 °2θ for PbHPO4) after 3 years. In addition, a decrease in the width at half-height (w1/2) with aging time from 0.20 °2θ initially to 0.15 °2θ after 3 years was observed for the β-Pb9(PO4)6 mineral, reflecting an increase in degree of structural order. While hydroxypyromorphite (Pb5(PO4)3OH) is expected to form in systems at pH < 8 (7), β-Pb9(PO4)6 is still present after 3 years of aging at room temperature. The delayed transformation, which should presumably have occurred via the aging of metastable structures, indicates that the system is at nonequilibrium even after 3 years of reaction in simple laboratory suspensions. Although an increase in peak intensity was observed in the XRD patterns of PbHPO4, an increase in the width at half-height (w1/2) with aging time (0.19 °2θ initially to 0.26 °2θ after 3 years) was calculated. The aged (3 years) lead-phosphate minerals had surface areas (m2 g-1) characteristic of crystalline phases: 15.1 ( 3.1 for PbHPO4 and 4.8 ( 0.1 for β-Pb9(PO4)6. Solubility of Lead-Phosphate Minerals. The aged PbHPO4 suspension had a pH of 3.8 and a concentration of Pb and P in solution equal to 99.2 µM and 422 µM, respectively. In addition to a higher pH (7.2), the aged β-Pb9(PO4)6 suspension had a lower concentration of total dissolved Pb (2.6 µM) and a higher concentration of total dissolved P (972 µM). The speciation program Visual MINTEQ (31) was used to calculate the activity of Pb2+ in the filtrates (see Experimental Section). The solubility equilibria for various lead-phosphate minerals used in constructing the solubility diagrams of Figure 2 are listed below (33):
PbHPO4(c) + H+ ) Pb2+ + H2PO4Pb3(PO4)2(c) + 4H+ ) 3Pb2+ + 2H2PO4-
Log Ksp ) -4.25
Log Ksp ) -5.26
Pb5(PO4)3OH(c) + 7H+ ) 5Pb2+ + 3H2PO4- + H2O Log Ksp ) -4.14 Pb5(PO4)3Cl(c) + 6H+ ) 5Pb2+ + 3H2PO4- + ClLog Ksp ) -5.06 As shown in Figure 2A, the activity of Pb2+ at initial conditions (Sauve´, personal communication) and after aging the PbHPO4 mineral suspension for 3 years was very similar and in both cases higher than the activity predicted by published Ksp values for PbHPO4(c). In an earlier study (15), the solubility of PbHPO4(c) was measured at pH e 4. Moreover, this study also observed disagreements among Ksp values for PbHPO4(c) reported in the literature between 1930 and 1960. Although the solubility product of β-Pb9(PO4)6 was not found in the literature, the activity of Pb2+ in the β-Pb9(PO4)6 system was underestimated by Ksp values for Pb5(PO4)3OH(c) and Pb3(PO4)2(c). Yet, in contrast to the PbHPO4 mineral suspension, the activity of Pb2+ in the β-Pb9(PO4)6 system increased after 3 years relative to the initial value (Figure 2A). A possible explanation is that the surface layer may have transformed in 3 years from the presumed deprotonated phosphate phase (β-Pb9(PO4)6) to perhaps an HPO42- or an H2PO4- phase. These phases will not be detected by conventional XRD since they may only be a few atoms thick.
Alternatively, a hydrated phase might have formed at the surface of the solid while the core remained as β-Pb9(PO4)6. These hydrated forms would be unstable initially because of the conditions of synthesis (high temperature) used to favor the tertiary phosphate. The conversion to a more stable surface phase would then increase Pb activity because the inherent solubility of Pb(H2PO4)2, PbHPO4, and Pb5(PO4)3OH is higher (see solubility equilibrium equations above and ref 33). This is analogous to the reversion of CuO (formed at elevated temperature in water) to Cu(OH)2 at room temperature, with a concomitant increase in Cu activity (34). While the addition of soluble phosphate was shown to reduce Pb solubility in a heavily Pb contaminated rifle range soil (17), Pb solubility values were about an order of magnitude higher than those predicted by Pb-PO4 model compounds (Figure 2A). Since chloropyromorphite (CP) is considered a highly insoluble form of Pb, several investigators have prompted its formation in both model systems and soils. As reported by Stanforth and Qiu (17), addition of chloride and phosphate reduced the solubility of Pb, however, not to the level predicted by the Ksp value of CP (Figure 2B). Lang and Kaupenjohann (18) also reported disagreement between calculated (Ksp) and experimental Pb solubility values for CP. In Figure 2B, we plotted the Log (Pb2+) value calculated from their reported Pb activity together with the solubility line for CP. We constructed the solubility line for CP using 0.002 M PO43- and the concentration of Cl- (0.002 M) used by the authors (labeled Pb5(PO4)3Cl - high P & Cl-). It is observed that Pb activity is about 1-2 orders of magnitude higher than calculated using the Ksp value of CP. Furthermore, although CP was always formed in their experiments (18), the presence of dissolved organic matter (DOM) during its synthesis seemed to exaggerate the apparent discrepancy. It has also been observed that various solid phases (cerrusite, PbCO3, and hydroxyapatite, Ca5(PO4)3OH) are partly transformed to chloropyromorphite after reaction with phosphate and chloride in model systems (10). We plotted the measured (experimental) concentrations of dissolved Pb at pH 4, 5, 6, and 7, and using the concentration of PO43- (0.000001 M) and Cl- (0.001 M) reported by the authors constructed the solubility line for CP (labeled Pb5(PO4)3Cl - low P & Cl) (Figure 2B). At all pH values, the suspensions were oversaturated with respect to chloropyromorphite after equilibration for 60 min. Although in all cases CP was identified by XRD analysis as comprising a fraction of the mineral composition of the solids, the authors (10) suggest that pH, low PO43levels, and kinetics may be factors limiting the transformation to CP. Oversaturation seems to be a consistent feature of leadphosphate minerals (Figure 2). Although the solubility product (Ksp) is a thermodynamic property that can be used to predict the solubility of a particular mineral phase, our results (lead-phosphate suspensions equilibrated for 3 years) and those of others indicate that Pb activities are not equal to the ones predicted for pure solid phases. Therefore, Ksp values often do not serve as good predictors of solubility controls in natural systems. Among the causes for such deviations are the lack of equilibrium and standard state conditions, but kinetic limitations may contribute as long as the metastable solid persists. Discrepancies between measured and Ksp-calculated activities might also result if the core and surface properties of the precipitate differ, if a mixed lead-phosphate phase is present, or if the suspension consists of aggregates that differ in degree of crystallinity. Moreover, solid phases present in soils and other natural systems can exhibit more defects, structural irregularities, and a smaller particle size than do the crystalline phases to which Ksp refers. The size and degree of crystallinity of neo-formed PbPO4 precipitates may influence other properties of the precipitate; for instance, phase stability, aggregation behavior,
FIGURE 3. Lead-phosphate mineral dissolution in the presence of model ligands and humic acid as a function of dissolution time. Panel A for PbHPO4 and Panel B for β-Pb9(PO4)6. Error bars represent 1 standard deviation from the mean of duplicate values. and crystal growth (35). The potential availability (i.e., solubility and dissolution) and mobility of Pb in the environment also depends on the size of the Pb-PO4 mineral formed as well as on the presence of soluble ligands. While the presence of dissolved organic matter (DOM) during the precipitation of CP resulted in crystallites of smaller particle size than when CP was formed in the absence of DOM, the smaller particles (colloids) showed increased mobility in laboratory column experiments (18). Because soils and contaminated sites are likely to contain other metals in addition to Pb (e.g., Cd2+, Zn2+, Ni2+, Cu2+), substitution for Pb is possible and can result in decreased crystallinity of the lead-phosphate mineral. For example, the lattice energy (calculated using ab initio quantum mechanics) for crystalline chloropyromorphite (-13163.8 kJ mol-1) increased with Cd2+ (-12980.1 kJ mol-1) and Zn2+ (-12875.7 kJ mol-1) substitution, thus decreasing the stability of the crystalline structure (36). Decreased crystallinity can result in increased solubility and enhanced dissolution by soluble ligands. Dissolution of Lead-Phosphate Minerals by Soluble Ligands. The time-dependent dissolution of aged leadphosphate minerals by soluble ligands resulted in variable concentrations of Pb in solution (Figure 3). The initial reaction between the solid phases and the ligand solutions (initially at pH 6) produced H+, as evidenced by a decrease in pH of the mineral suspensions. In the PbHPO4 system (initially at pH 3.8) the pH dropped to 3-3.5 for all ligand solutions. In the β-Pb9(PO4)6 system (initially at pH 7.2) the pH dropped to 4.5-5.5. Although the pH of the mineral suspensions increased toward their original value during the experiment, VOL. 38, NO. 21, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
5587
TABLE 1. Stability Constants for Various Pb-Ligand Complexesa,b Pb-ligand complex
equilibrium Log K
cysteine Pb2+ + (C3H5O2NS)2- ) PbC3H5O2NS Pb2+ + 2 (C3H5O2NS)2- ) Pb(C3H5O2NS)22methionine Pb2+ + (C5H10O2NS)1- ) PbC5H10O2NS1+ Pb2+ + 2 (C5H10O2NS)1- ) Pb(C5H10O2NS)2 thiosulfate Pb2+ + S2O32- ) PbS2O3 Pb2+ + 2 S2O32- ) Pb(S2O3)22nitrate Pb2+ + NO3- ) PbNO31+ Pb2+ + 2 NO3- ) Pb(NO3)2 humic acid Pb2+ + humic acid ) Pb-humic acid (at pH ∼ 7) Pb2+ + humic acid ) Pb-humic acid (at pH < 4.9) histidine Pb2+ + (C2H3O2N-CH2-C3N2H3)1- ) (Pb-C2H3O2N-CH2-C3N2H3) 1+ Pb2+ + 2 (C2H3O2N-CH2-C3N2H3)1- ) Pb-(C2H3O2N-CH2-C3N2H3)2 acetate Pb2+ + CH3COO- ) PbCH3COO1+ Pb2+ + 2 CH3COO- ) Pb(CH3COO)2 citrate (-OOCCH2-C(OH)COO-CH2COO-) Pb2+ + citrate ) Pb(citrate) Pb2+ + 2 citrate ) Pb(citrate)2 oxalate Pb2+ + -OOC-COO- ) Pb(oxalate)
ML/M‚L 12.5 ML2/M‚L2 18.6 ML/M‚L ML2/M‚L2
4.38 8.62
ML/M‚L ML2/M‚L2
2.42 4.86
ML/M‚L ML2/M‚L2
1.17 1.4
ML/M‚L
>8.7
ML/M‚L
8.7
ML/M‚L
5.96
ML2/M‚L2
9.0
ML/M‚L ML2/M‚L2
2.7 4.1
ML/M‚L ML2/M‚L2
5.4 8.1
ML/M‚L
4.2
Constants were compiled from refs 40-42. b The constants were not determined at the same ionic strength (I) or under the same experimental conditions and are therefore not strictly comparable; variations over a Log unit can be expected. a
TABLE 2. Critical Stability Constants (with H+) for Various Amino Acidsa and a Humic Acidb chemical equation cysteine (C2H4O2N-CH2SH)0 (C3H5O2NS)2- + H+ ) (C3H6O2NS)1(C3H6O2NS)1- + H+ ) (C3H7O2NS)0 (C3H7O2NS)0 + H+ ) (C3H8O2NS)1+ methionine (C2H4O2N-CH2CH2SCH3)0 (C5H10O2NS)1- + H+ ) (C5H11O2NS)0 (C5H11O2NS)0 + H+ ) (C5H12O2NS)1+ thiosulfate S2O32- at pH > 1.7 humic acid a
equilibrium
Log K
HL/L‚H H2L/HL‚H H3L/H2L‚H
10.29 8.15 1.88
HL/L‚H H2L/HL‚H
9.05 2.20
Constants were compiled from refs 40 and 41. acid at I ) 0.1 (42).
4.92 b
pKa of a humic
no direct relationship was observed between the concentration of Pb in solution and pH of the system (figure not shown). As shown in Figure 3A, while cysteine and thiosulfate enhanced the dissolution of the PbHPO4 (pH 3.8) solid phase, methionine had no effect, and Pb dissolution was similar to the values found using KNO3. Thiosulfate and cysteine, unlike methionine, contain a free thiolic functional group (-SH) that can form a soluble Pb-ligand complex. In addition, the stability constants for Pb-cysteine and Pb-thiosulfate complexes are generally higher that those for methionine and KNO3, especially at low pH values (Tables 1 and 2). The fact that thiosulfate enhanced Pb solubility may perhaps have important implications for the uptake of Pb by soil organisms. A mechanism similar to the one suggested by Fortin and Campbell (32), namely, the uptake of an Ag-thiosulfate soluble complex via an anion (SO42-) transporter, may provide a 5588
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 21, 2004
mechanism for the import of Pb. Although the bottles air space was purged with N2(g) to retard oxidation of the solutions, dissolved O2(g) was in fact present in the ligand solutions. Variability in the efficacy of cysteine and thiosulfate (ligands more susceptible to oxidation) in dissolving PbHPO4 (Figure 3A) may be the result of unstable conditions brought about by the oxidation of the reduced sulfur in the ligands. Oxidation of the ligand solutions may be accompanied by a decrease in pH since most oxidation reactions produce acidity. At the end of the experiment (34 days), the concentration of Pb brought to solution by all ligands was similar (with the exception of the humic acid) and higher than the 99.2 µM Pb measured in the mineral suspension (Figure 3A). The humic acid suppressed Pb dissolution from the PbHPO4 system (Figure 3A). At low pH the humic acid may sorb onto the PbHPO4 solid thus blocking its surface and depressing its dissolution. Alternatively, the humic acid may form a complex with Pb which can become large enough (>0.2 µm) to be filtered out from solution. Lead dissolution from the β-Pb9(PO4)6 mineral (pH 7.2) by the humic acid solution was very pronounced, with solubility increasing over the period of the experiment until it reached a value of approximately 550 µM after 34 days (Figure 3B). The concentration of Pb in solution after mineral reaction with low molecular weight model sulfides was, however, below approximately 30 µM at all times (Figure 3B insert). The amount of Pb dissolved by model sulfides is often higher than the 2.58 µM Pb concentration measured in the mineral suspension. The results suggest that two different mechanisms might be involved in the dissolution of the β-Pb9(PO4)6 mineral by the various ligands. In the experiments using model sulfides, the concentration of Pb in solution seems to be affected by the pH of the system (protonpromoted dissolution); in the humic-acid system, the concentration of Pb in solution seems affected by a ligand promoted dissolution reaction. Similar results were obtained by Ravichandran and co-workers (24) where the presence of S-containing organic ligands (cysteine and mercaptoacetic acid) dissolved Hg from cinnabar (HgS) to different degrees. The concentration of Hg dissolved by S-containing organic ligands was, however, lower than the amount dissolved by the humic and fulvic acids fraction (24). The effect of soluble ligands on the dissolution of neo-formed (and secondary) phosphate minerals has also been reported for hydroxyapatite treated sediments (2) where Ni and U solubility was higher in sediments with higher concentrations of DOC. To determine how much of the humic acid was in solution throughout the course of the experiment, we measured the concentration of dissolved organic carbon (DOC) in the humic acid reacted lead-phosphate filtrate solutions. We found increased concentrations of DOC as the pH of the systems increased (Figure 4). No clear trend was observed, however, between total dissolved Pb and DOC (figure not shown). It seems, then, that the effect of the humic acid depends on the pH of the system under investigation. Although Pb reacts strongly with dissolved organic matter to promote metal dissolution from adsorption sites (37, 38), the humic acid is evidently very efficient in suppressing Pb dissolution from the PbHPO4 system (Figure 3A). This is probably the result of the relatively low pH of the system and of the tendency high-molecular-weight soluble organics have to sorb onto surfaces at this low pH, thus blocking the surface and preventing further dissolution. These interactions can result in reduced concentrations of Pb in solution. The fact that the humic acid suppressed Pb dissolution from the PbHPO4 system and enhanced Pb dissolution from the β-Pb9(PO4)6 system emphasizes the importance of pH conditions on the behavior of humic acids toward solid phases. At low pH humic acids tend to close their structure as their functional groups are protonated; at higher pH the functional groups
FIGURE 4. Dissolved organic carbon (DOC) as a function of pH for the humic acid-reacted PbHPO4 and β-Pb9(PO4)6 minerals. Error bars represent 1 standard deviation from the mean of duplicate values. present in humic acids tend to deprotonate thus increasing their negative charge and chelating ability (39). In addition, there is less competition by H+ ions at this higher pH value. The time-dependent dissolution experiments demonstrate the ability of soluble ligands to enhance the dissolution of Pb from solid phases considered highly insoluble. It is also clear that increased amounts of soluble humics that occur at high pH can have a significant effect in the dissolution of lead-phosphates. The dissolution effect of humic acids could not be modeled by simple ligands, perhaps due to the greater effect of a polyvalent ligand in weakening the Pb-PO43- bond. Despite cysteine’s larger stability constant (Pb-ligand complex formation) compared to N- and O-containing soluble ligands (Table 1), its dissolution effect was not as pronounced. As reported for the dissolution of cinnabar (24), the presence of cations could possibly diminish this effect. Although soluble complexes may readsorb, by promoting the formation of soluble Pb-organic complexes, DOC could in field situations enhance Pb mobility through the soil. We can therefore conclude that at high pH the solubility, dissolution, and mobility of Pb from lead-phosphate minerals may increase due to the formation of Pb-organic complexes.
Acknowledgments This work was supported in part by the USDA-NRI Competitive Grants Program (Award no. 2003-35107-13650 to C.E.M.). The authors thank Dr. Matteo Spagnuolo (Universita` di Bari, Italy) for providing the soil-extracted humic acid, Dr. Se´bastien Sauve´ (Universite´ de Montre´al, Canada) for providing the initial XRD and Pb solubility data, and Dr. Sridhar Komarneni (The Pennsylvania State University, U.S.A.) for surface area measurements.
Literature Cited (1) McGowen, S. L.; Basta, N. T.; Brown, G. O. Use of diammonium phosphate to reduce heavy metal solubility and transport in smelter-contaminated soil. J. Environ. Qual. 2001, 30, 493500. (2) Seaman, J. C.; Arey, J. S.; Bertsch, P. M. Immobilization of nickel and other metals in contaminated sediments by hydroxyapatite addition. J. Environ. Qual. 2001, 30, 460-469. (3) Knox, A. S.; Kaplan, D. I.; Adriano, D. C.; Hinton, T. G.; Wilson, M. D. Apatite and phillipsite as sequestering agents for metals and radionuclides. J. Environ. Qual. 2003, 32, 515-525. (4) Lower, S. K.; Maurice, P. A.; Traina, S. J. Simultaneous dissolution of hydroxylapatite and precipitation of hydroxypyromorphite: Direct evidence of homogeneous nucleation. Geochim. Cosmochim. Acta 1998, 62, 1773-1780.
(5) Mavropoulos, E.; Rossi, A. M.; Costa, A. M.; Perez, C. A. C.; Moreira, J. C.; Saldanha, M. Studies on the mechanisms of lead immobilization by hydroxyapatite. Environ. Sci. Technol. 2002, 36, 1625-1629. (6) Ma, Q. Y.; Traina, S. J.; Logan, T. J In situ lead immobilization by apatite. Environ. Sci. Technol. 1993, 27, 1803-1810. (7) Laperche, V.; Traina, S. J.; Gaddam, P.; Logan, T. J. Chemical and mineralogical characterizations of Pb in a contaminated soil: Reactions with synthetic apatite. Environ. Sci. Technol. 1996, 30, 3321-3326. (8) Cotter-Howells, J. D.; Caporn, S. Remediation of contaminated land by formation of heavy metal phosphates. Appl. Geochem. 1996, 11, 335-342. (9) Ma, Q. Y.; Logan, T. J.; Traina, S. J. Lead immobilization from aqueous solutions and contaminated soils using phosphate rocks. Environ. Sci. Technol. 1995, 29, 1118-1126. (10) Zhang, P.; Ryan, J. A. Transformation of Pb(II) from cerrusite to chloropyromorphite in the presence of hydroxyapatite under varying conditions of pH. Environ. Sci. Technol. 1999, 33, 625630. (11) Zhang, P.; Ryan, J. A. Formation of chloropyromorphite from galena (PbS) in the presence of hydroxyapatite. Environ. Sci. Technol. 1999, 33, 618-624. (12) Ryan, J. A.; Zhang, P.; Hesterberg, D.; Chou, J.; Sayers, D. Formation of chloropyromorphite in a lead-contaminated soil amended with hydroxyapatite. Environ. Sci. Technol. 2001, 35, 3798-3803. (13) Ma, Q. Y.; Logan, T. J.; Traina, S. J.; Ryan, J. A. Effects of NO3-, Cl-, F-, SO42-, and CO32- on Pb2+ immobilization by hydroxyapatite. Environ. Sci. Technol. 1994, 28, 408-418. (14) Ma, Q. Y.; Traina, S. J.; Logan, T. J.; Ryan, J. A. Effects of aqueous Al, Cd, Cu, Fe(II), Ni, and Zn on Pb immobilization by hydroxyapatite. Environ. Sci. Technol. 1994, 28, 1219-1228. (15) Nriagu, J. O. Lead orthophosphates. 1. Solubility and hydrolysis of secondary lead orthophosphate. Inorg. Chem. 1972, 11, 24992503. (16) Sauve´, S.; McBride, M.; Hendershot, W. Lead phosphate solubility in water and soil suspensions. Environ. Sci. Technol. 1998, 32, 388-393. (17) Stanforth, R.; Qiu, J. Effect of phosphate treatment on the solubility of lead in contaminated soil. Environ. Geol. 2001, 41, 1-10. (18) Lang, F.; Kaupenjohann, M. Effect of dissolved organic matter on the precipitation and mobility of the lead compound chloropyromorphite in solution. Eur. J. Soil Sci. 2003, 54, 139147. (19) Zhang, P.; Ryan, J. A.; Yang, J. In vitro soil Pb solubility in the presence of hydroxyapatite. Environ. Sci. Technol. 1998, 32, 2763-2768. (20) Jaynes, W. F.; Moore, P. A.; Miller, D. M. Solubility and ion activity products of calcium phosphate minerals. J. Environ. Qual. 1999, 28, 530-536. (21) Huang, W. H.; Keller, W. D. Dissolution of clay minerals in dilute organic acids at room temperature. Am. Mineral. 1971, 56, 10821095. (22) Stumm, W.; Furrer, G. The dissolution of oxides and aluminum silicates: Examples of surface-coordination-controlled kinetics. In Aquatic Surface Chemistry; Stumm, Ed.; 1987; pp 197-219. (23) Welch, S. A.; Ullman, W. J. The effect of organic acids on plagioclase dissolution rates and stoichiometry. Geochim. Cosmochim. Acta 1993, 57, 2725-2736. (24) Ravichandran, M.; Aiken, G. R.; Reddy, M. M.; Ryan, J. N. Enhanced dissolution of cinnabar (mercuric sulfide) by dissolved organic matter isolated from the Florida Everglades. Environ. Sci. Technol. 1998, 32, 3305-3311. (25) Jacobson, A. R.; Martı´nez, C. E.; Spagnuolo, M.; McBride, M. B.; Baveye, P. The influence of model sulfides and humic acid on silver solubility from acanthite (β-Ag2S). Environ. Pollut. 2004, in review. (26) Harrison, W. J.; Thyne, G. D. Predictions of diagenetic reactions in the presence of organic acids. Geochim. Cosmochim. Acta 1992, 56, 565-586. (27) Fein, J. B.; Hestrin, J. E. Experimental studies of oxalate complexation at 80 °C: Gibbsite, amorphous silica, and quartz solubilities in oxalate-bearing fluids. Geochim. Cosmochim. Acta 1994, 58, 4817-4829. (28) Ochs, M. Influence of humified and nonhumified natural organic compounds on mineral dissolution. Chem. Geol. 1996, 132, 119124. (29) Baker, W. E. The role of humic acids from Tasmanian podzolic soils in mineral degradation and metal mobilization. Geochim. Cosmochim. Acta 1973, 37, 269. VOL. 38, NO. 21, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
5589
(30) Gamble, D. S.; Schnitzer, M.; Kendorff, H.; Langford, C. H. Multiple metal ion exchange equilibria with humic acid. Geochim. Cosmochim. Acta 1983, 47, 1311-1323. (31) Visual MINTEQ, version 2.14. Gustafsson, J. P., KTH, Department of Land and Water Resources Engineering, Stockholm, Sweden, 2003. (32) Fortin, C.; Campbell, P. G. C. Thiosulfate enhances silver uptake by a green alga: Role of anion transporters in metal uptake. Environ. Sci Technol. 2001, 35, 2214-2218. (33) Lindsay, W. L. Chemical Equilibria in Soils; John Wiley & Sons: 1979. (34) Schindler, P.; Althaus, H.; Hofer, F.; Minder, W. Lo¨slichkeitsprodukte von zinkoxid, kupferhydroxid und kupferoxid in abha¨ngigkeit von teilchengro¨sse und molarer oberfla¨che. Ein beitrag zur thermodynamik von grenzfla¨chen fest-flu ¨ ssig. Helv. Chim. Acta 1965, 48, 1204-1215. (35) Nanoparticles and the Environment; Banfield, J. F.; Navrotsky, A., Eds.; Reviews in Mineralogy and Geochemistry, Vol. 44; Mineralogical Society of America (Ribbe, P. H., Ed.) and Geochemical Society (Rosso, J. J., Ed.), 2001. (36) Shevade, A. V.; Erickson, L.; Pierzynski, G.; Jiang, S. Formation and stability of substituted pyromorphite: a molecular modeling study. J. Hazard. Subst. Res. 2001, 2, 1-12.
5590
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 21, 2004
(37) Schwertmann, U.; Kodama, H.; Fischer, W. R. In Interactions of soil minerals with natural organics and microbes; Huang, P. M., Schnitzer, M., Eds.; Soil Science Society of America: Madison, WI, 1986; p 223. (38) Saar, R. A.; Weber, J. H. Lead(II) complexation by fulvic acid: how it differs from fulvic acid complexation of copper(II) and cadmium(II). Geochim. Cosmochim. Acta 1980, 44, 1381-1384. (39) Stumm, W.; Morgan, J. J. Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters, 3rd ed.; John Wiley & Sons: New York, 1996. (40) Martell, A. E.; Smith, R. M. Critical Stability Constants, Vol. 1: Amino Acids; Plenum Press: New York, 1974; pp 47-50. (41) Kotrly´, S.; Sˇ u ˚ cha, L. Handbook of Chemical Equilibria in Analytical Chemistry; John Wiley & Sons: New York, 1985; pp 116-133. (42) Stevenson, F. J. Stability constants of Cu2+, Pb2+, and Cd2+ complexes with humic acids. Soil Sci. Soc. Am. J. 1976, 40, 665672.
Received for review March 11, 2004. Revised manuscript received August 2, 2004. Accepted August 2, 2004. ES049617X
