Research Effects of pH on Heavy Metal Sorption on Mineral Apatite X I A O B I N G C H E N , * ,† J U D I T H V . W R I G H T , ‡ JAMES L. CONCA,‡ AND LONI M. PEURRUNG† Pacific Northwest National Laboratory, P.O. Box 999, Battelle Boulevard, Richland, Washington 99352, and UFA Ventures, 2000 Logston Boulevard, Richland, Washington 99352
The sorption of aqueous Pb, Cd, and Zn onto a mineral apatite from North Carolina was investigated in relation to a wide range of pH. The effects of pH on solid-phase precipitation were particularly emphasized. The heavy metals were applied as single or multiple species. Solution pH greatly affected metal sorption mechanism by apatite and metal-apatite reaction products. The sorption of aqueous Pb was primarily through a process of the dissolution of apatite followed by the precipitation of variable pyromorphitetype minerals under acidic condition or of hydrocerussite [Pb3(CO3)2(OH)2] and lead oxide fluoride (Pb2OF2) under alkaline condition. Otavite (CdCO3), cadmium hydroxide [Cd(OH)2], and zincite (ZnO) were formed in the Cd or Zn system, especially under alkaline condition; while hopeite [Zn3(PO4)2‚4H2O] might precipitate only under very acidic condition. Alternative sorption mechanisms other than precipitation of the crystalline phases were important in reducing Cd and Zn concentrations by the apatite and might include ion exchange, adsorption, absorption, complexation, coprecipitation, or precipitation of amorphous phases. Removal of Cd and Zn by the apatite was pH dependent, whereas removal of Pb was not. The removals were about 0.729 mmol of Pb, 0.489-1.317 mmol of Cd, and 0.5962.187 mmol of Zn/g of apatite, representing removal of 99.9%, 37.0-99.9%, and 27.0-99.9% of heavy metals from solution, respectively.
Introduction Many studies have been conducted on the crystal structure and chemistry of synthetic and mineral apatites (1, 2). Apatites occur commonly in soils and sediments and are the principal mineral component of phosphate rock. Phosphate rock serves as the raw material in all phosphate fertilizers and can also be directly applied to the field. Geochemical studies have demonstrated that a variety of trace elements are enriched in apatites (3, 4). Thus, apatites could regulate the concentrations of calcium, phosphate, lead, cadmium, zinc, and other heavy metals (5) in natural environments and may provide a cost-effective technology for remediating metal-contaminated soils and water.
Sedimentary phosphorite is one of the primary sources of phosphate rock (>80% worldwide) and can be obtained readily and inexpensively. The principal phosphate mineral in sedimentary phosphorites is carbonate fluorapatite (6), which is microcrystalline and differs in composition from pure synthetic apatites because of extensive and complex ion substitutions in apatite structure (1, 2, 7-9). The substitutions result in a tremendous variation in the chemical reactivity and stability of carbonate fluorapatites. In general, their solubility increases with increasing carbonate substitution (8, 10). The interaction of apatites with heavy metals may form relatively insoluble metal phosphates and/or result in the adsorption of heavy metals on apatites, thus significantly reducing aqueous metal concentrations. For instance, studies have suggested that the reaction of synthetic hydroxyapatite with aqueous Pb can result in the formation of pyromorphites (or lead apatites), which could be hydroxypyromorphite [Pb10(PO4)6(OH)2], chloropyromorphite [Pb10(PO4)6Cl2], or fluoropyromorphite [Pb10(PO4)6F2], depending on the presence or absence of Cl and F in solution (11-13). However, the mechanisms of these interactions are not agreed upon in the literature. Earlier work proposed that an ion exchange reaction is responsible for the formation of the solids (11, 14-16), whereas recent studies concluded that the dissolution of hydroxyapatite and the precipitation of pyromorphites was the mechanism responsible for reducing Pb concentration to below regulatory limits (12, 13, 17). Previous work (18-22) indicated that the solubilities of pyromorphites are extremely low. In addition, the interactions of apatites with other heavy metals have also been investigated, including Cd (23, 24) and Zn (24, 25), but their reaction mechanisms are also not well understood. Because of the limited knowledge on the mechanisms for removing heavy metals from solution by apatites, the term sorption (26) is herein used loosely as a general term to describe attachment of heavy metals from a solution to its coexisting apatite surfaces. The studies discussed above were mainly concerned with the interaction of synthetic hydroxyapatite with heavy metals. The competition and interference of cations and anions on the sorption processes were only studied occasionally and are poorly understood (11, 16, 17, 27, 28). Furthermore, the reaction of mineral apatite with heavy metals has rarely been studied and is more complicated because of the coexistence of PO43-, CO32-, F-, OH-, and other cations in the structure of mineral apatites, and thus of their possible presence in solution. Recently, the reaction of mineral apatites with aqueous Pb was reported (29); however, the effects of pH on the interaction and reaction products were not considered. The influence of pH on the sorption of heavy metals on calcite, oxides, clay minerals, and soils has been extensively studied (30-34). Therefore, the objectives of this study were to (1) investigate the pH effects on the sorption behaviors of Pb, Cd, and Zn in the presence of mineral apatite and (2) examine the solid residues from the interactions and the mechanisms possibly involved in the interaction under variable pH.
Experimental Section * Corresponding author fax: 509-372-6328; e-mail:
[email protected]. † Pacific Northwest National Laboratory. ‡ UFA Ventures.
624
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 31, NO. 3, 1997
Materials. The mineral apatite used in this investigation was from a sedimentary phosphate rock deposit situated in North Carolina and was supplied by Texasgulf as a raw mining
S0013-936X(95)00882-0 CCC: $14.00
1997 American Chemical Society
product. The pelletal grains of the apatite were ground to a fine powder to pass through a 400-mesh (38 µm) standard sieve using an 8000 Spex Mixer/Mill equipped with a tungsten carbide sample holder. The sample is made of carbonate fluorapatite as indicated by X-ray diffraction (XRD), having a composition of Ca9.53Na0.34Mg0.13(PO4)4.77(CO3)1.23F2.49 (35) and a surface area of 27 m2/g by the EGME method. Deionized distilled water (DD-H2O) was used to prepare all solutions and suspensions. Individual metal concentrations of Pb, Cd, and Zn were 2.50 × 10-2, 4.50 × 10-2, and 7.50 × 10-2 M, respectively, prepared from their nitrate salts. For each metal, a set of 11 solutions with pH 1-12 were prepared. The pH was adjusted with concentrated nitric acid or sodium hydroxide to the desired values. Some metal hydroxide phases started to form at very basic pH before the solutions were contacted with the apatite. The metal solutions were equilibrated with the apatite separately or in combination. Most of the experiments were run in one replicate and some in triplicate as control tests. Single-Species Sorption Tests (SSST). The apatite (1.20 g) was equilibrated with each of the pH-adjusted metal solutions (35 mL) in acid-washed 50-mL Nalgene polycarbonate centrifuge tubes. The apatite suspensions and appropriate blanks were then shaken at ambient temperature for 24 h. Multiple-Species Sorption Tests (MSST). The individually prepared metal solutions were mixed in equal volumes to form a multiple-species mixture. The mixed solutions were then adjusted to the desired pH accordingly. Similar procedure as the SSST experiment was used. Analytical Methods. After a 24-h equilibration period, the apatite suspensions were centrifuged for 15 min, and the supernatants were filtered through 0.2-µm Nalgene syringe filters with cellulose acetate membranes. Solution pH was measured with a Beckman Φ12 pH/ISE meter, using a glass electrode paired with an internal calomel reference system and a thermistor for automatic temperature compensation. The pH calibration was done with buffers of known pH at 4, 7, and 10. The metal concentrations were analyzed with inductively coupled plasma-mass spectroscopy (ICP-MS). The solid residues were examined with an X-ray diffractometer. X-ray diffraction was performed on a Siemens D-5000 X-ray diffractometer operated at 35 kV and 30 mA using CuKR radiation. Measurements were made using a step scanning mode with a fixed 0.02° 2θ step and a count time of 2 s/step. All XRD analyses were performed with back-filled, randomly oriented cavity mounts, which were prepared using a powderpress technique modified from Gibbs (36). Selected solid residues were also observed under a Hitachi S570 scanning electron microscope equipped with X-ray energy dispersive spectroscopy (SEM-EDX). The samples were mounted on stainless steel stubs using double-stick tape and were then coated with either gold or carbon. The SEM was operated at 25 kV for secondary electron (SE) imaging and 15 kV for EDX analysis.
Results and Discussion SSST of Heavy Metals with Apatite. The concentrations of Pb, Cd, and Zn after sorption by the apatite in relation to final suspension pH are shown in Figure 1. Overall, the sorption behaviors of Cd and Zn are similar to each other, with differences in the amounts sorbed. They both are, however, dramatically different from that of Pb sorption. The effects of the initial pH on the final pH of apatite suspensions are summarized in Figure 2. Lead. The influence of pH on the sorption of aqueous Pb onto the apatite is substantially different than onto clays and soils. In general, the amount of Pb sorbed by clays and soils increases with pH from 2 to 5 (31, 34, 37). However, the effects of pH on Pb sorption by the apatite was not significant (Figure 1), with a 99.9% reduction for pH 3-10.5 and 95.5%
FIGURE 1. Heavy metal concentrations after sorption with North Carolina apatite with respect to pH in SSST.
FIGURE 2. Relationship of initial and final pH after solutions reacted with North Carolina apatite. reduction at pH 12.0. The Pb removals by the apatite under most of the pH conditions were about 0.729 mmol of Pb/g of apatite. Ma et al. (29) studied the attenuation of aqueous Pb by mineral apatites from different localities and concluded that the primary mechanism of Pb removal was through the dissolution of mineral apatites and the precipitation of carbonate fluoropyromorphite [Pb10(PO4)3(CO3)3FOH] and/ or hydrocerussite [Pb3(CO3)2(OH)2]. Our study suggests that solid reaction products of aqueous Pb with the apatite was pH-dependent (Figure 3 and Table 1). The peaks of the pyromorphite-type compounds became weaker and broader as the solution pH increased, suggesting that the quantity and crystallinity of the pyromorphites decreased with increasing pH, whereas those of hydrocerussite became stronger with increasing pH, with the strongest peaks at an initial pH of 7-8. The identification of different pyromorphites was based on a comparison of the shifting of their XRD maxima with the standard XRD patterns of pyromorphites given by the Joint Committee on Powder Diffraction Standards (JCPDS) (Table 2). The XRD maxima of synthetic fluoropyromorphite and hydroxypyromorphite are 2.92 and 2.97 Å, respectively. Thus, the incorporation of OH- into the structure of fluoropyromorphite will cause the shifting of its XRD maximum to larger d-spacing. The pyromorphites formed at an initial solution
VOL. 31, NO. 3, 1997 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
625
suming Ca10(PO4)6-x(CO3)xF2+x as the general formula for carbonate fluorapatite, the chemical reactions involved can be summarized as
Ca10(PO4)6-x(CO3)xF2+x(c) + 12H+ f 10Ca2+ + (6 - x)H2PO4- + xH2CO30 + (2 + x)F- (1) 10Pb2+ + 6H2PO4- + 2F- f Pb10(PO4)6F2(c) + 12H+
(2)
As the initial pH of Pb solution increased to 2.7-5.1, the equilibrium pH increased to neutral range (6.6-6.8). This higher pH resulted in an increase of the activity of OH- and the formation of hydroxyl fluoropyromorphite:
Ca10(PO4)6-x(CO3)xF2+x(c) + (12 - x)H+ f 10Ca2+ + (6 - x)H2PO4- + xHCO3- + (2 + x)F- (3) 10Pb2+ + 6H2PO4- + 2(F-,OH-) f Pb10(PO4)6(F,OH)2(c) + 12H+ (4)
FIGURE 3. XRD patterns of North Carolina apatite before (A) and after interaction with aqueous Pb at initial/final pH of 1.1/3.1 (B), 2.7/6.6 (C), 6.0/7.1 (D), 8.4/10.6 (E), and 12.1/11.9 (F). Peak labels: A, apatite; FP, fluoropyromorphite; HF, hydroxyl fluoropyromorphite; CHF, carbonate hydroxyl fluoropyromorphite; HP, hydroxypyromorphite; HC, hydrocerussite; LO, lead oxide fluoride.
TABLE 1. Summary of New Solid Phases Formed in Interaction of Apatite with Aqueous Pb with Respect to pH initial pH
final pH
new solid phase
1.1-2.0 2.7-5.1 6.0-8.4
3.1-6.2 6.6-6.8 7.1-10.6
6.0-12.1 10.7-12.1
7.1-11.9 10.7-11.9
fluoropyromorphite hydroxyl fluoropyromorphite carbonate hydroxyl fluoropyromorphite hydrocerussite hydroxypyromorphite lead oxide fluoride
pH of 1.1-2.0 had an XRD maximum equal to 2.93 Å and was identified as fluoropyromorphite. However, at an initial pH of 2.7-5.1, the new precipitates had XRD maxima of 2.95 Å, indicating that a certain amount of OH- was incorporated into the structure of pyromorphites. In addition, the XRD data suggest that carbonate ions did not exist in the crystal structure of the compound because the incorporation of carbonate into the structure of apatites would have caused their XRD maxima to shift to smaller d-spacings (Table 2). Thus, the pyromorphite was classified as hydroxyl fluoropyromorphite. At an initial pH of 6.0 and 7.2, the compounds formed had relatively smaller d-spacing (Table 2) and were identified as carbonate hydroxyl fluoropyromorphite. The phosphate compound formed at very alkaline conditions was recognized as hydroxypyromorphite because of the XRD maximum of about 2.97 Å. Although the formation of the new solid phases described above involved apatite dissolution followed by the precipitation of the solids, which was also suggested in previous studies (12, 13), the detailed processes depended on reaction pH. When the initial metal solution pH was from 1.1 to 2.0, the final pH was from 3.1 to 6.2 (Figure 2). At this final pH range, the carbonates dissolved from the apatite existed mainly as H2CO30, and thus, the activities of carbonate and OH- ions were extremely low (38, 39), resulting in the difficulty of incorporating them into the pyromorphite structure. As-
626
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 31, NO. 3, 1997
Rickard and Nriagu (40) stated that cerussite is formed in an aqueous environment of high dissolved Pb, high carbonate ions, and relatively low pH, while hydrocerussite is common under conditions of high dissolved Pb, high pH, and low carbonate. The requirements for the precipitation of hydrocerussite were met when the initial pH of Pb solution was above 7.0, which resulted in the final pH above 10.0. Thus, the solid reaction product was dominated by hydrocerussite, with minor amounts of carbonate hydroxyl fluoropyromorphite (Figure 3):
Ca10(PO4)6-x(CO3)xF2+x(c) + 6H+ f 10Ca2+ + (6 - x)HPO42- + xHCO3- + (2 + x)F- (5) 3PbOH+ + 2HCO3- + OH- f Pb3(CO3)2(OH)2(c) + 2H2O (6) 10Pb2+ + 6(HPO42-,HCO3-) + 2(F-,OH-) f Pb10(PO4,CO3)6(F,OH)2(c) + 6H+ (7) At a pH slightly above 7.0, both Pb2+ and PbOH+ contribute significantly to total lead in solution. Therefore, for simplicity, the species of dissolved Pb are presented generically as shown in eqs 6 and 7. At the final pH of above 10, the activity of carbonate ions in eq 7 was higher than that in eqs 2 and 4, and the driving force for carbonate incorporation into pyromorphite sites increased, resulting in the precipitation of carbonate hydroxyl fluoropyromorphite. When the pH of Pb solutions was between 10.7 and 12.1, the reactions involved in the formation of hydrocerussite, hydroxypyromorphite, and lead oxide fluoride were shown as
Ca10(PO4)6-x(CO3)xF2+x(c) + (6 - x)H+ f 10Ca2+ + (6 - x)HPO42- + xCO32- + (2 + x)F- (8) 3Pb(OH)3- + 2CO32- f Pb3(CO3)2(OH)2(c) + 7OH- (9) 10Pb(OH)3- + 6HPO42- f Pb10(PO4)6(OH)2(c) + 6H2O + 22OH- (10) 2Pb(OH)3- + 2F- f Pb2OF2(c) + H2O + 4OH- (11) The solubility of apatites is highly pH dependent with lower solubility at higher pH (41, 42), which resulted in a drop of dissolved phosphate, carbonate, and fluoride in the aqueous
TABLE 2. X-ray Diffraction Maxima of Pyromorphite-Type Compounds and a Comparison of Effect of Carbonate Substitution into Apatite Structure on d-Spacing d-spacing (112, 211), Å initial/final pH of aqueous Pb solution pyromorphite-type compd
JCPDS
1.1/ 3.1
2.0/ 6.2
Pb10(PO4)6F2 Pb10(PO4)6(F,OH)2 Pb10(PO4)6(OH)2 Pb10(PO4,CO3)6(F,OH)2
2.92
2.93
2.93
Ca10(PO4)6(OH)2 Ca10(PO4)6F2 Ca10(PO4)3(CO3)3(OH)2
2.81 2.80 2.78
2.7/ 6.6
3.6/ 6.8
4.4/ 6.8
5.1/ 6.8
2.95
2.95
2.95
2.95
6.0/ 7.1
7.2/ 10.2
2.92
2.91
2.97
10.7/ 10.7
12.1/ 11.9
2.97
2.97
Synthetic Apatite
system. This suggested that the anions might be required to individually combine with dissolved Pb as shown in eqs 9-11 because Pb was oversupplied. In alkaline solutions, hydrocerussite is less soluble than fluoropyromorphite, but slightly more soluble than hydroxypyromorphite (5). In addition, the activity of OH- was much higher than that of F-, resulting in the formation of hydroxypyromorphite instead of fluoropyromorphite. Although the solubility of hydrocerussite is slightly higher than that of hydroxypyromorphite, the shortage of dissolved phosphate ions and the oversupply of aqueous Pb resulted in the precipitation of hydrocerussite as well as lead oxide fluoride. Because lead oxide is thermodynamically very soluble (5), it may be eventually converted into hydroxypyromorphite or hydrocerussite, providing the continuous supply of phosphate and carbonate. In addition to the XRD results shown in Figure 3, the formation of the new precipitates was supported by the SEM examination of selected solid residues. Scattered needle- or rod-shaped crystals, presumably fluoropyromorphite, were observed in the sample obtained from the reaction at very acidic conditions (Figure 4). Similar crystals were reported previously (29). However, no obvious differences were detected under SEM between the unreacted apatite and those reacted at higher pH conditions (data not shown), although the XRD data in Figure 3 indicated the presence of pyromorphites in the samples. This observation further suggests that the crystallinity of the pyromorphites decreased with increasing pH. From the above discussion, it appears that the precipitation of variable solid phases in the interaction of aqueous Pb with the apatite was consistent with thermodynamic predictions (5). The thermodynamic calculation suggested that pyromorphites should be the most insoluble phases under acidic to neutral conditions while Pb carbonates and oxides should be less soluble than fluoropyromorphite under alkaline conditions. Such predictive relationships are what we observed in our study. Thus, the removal of aqueous Pb from solution primarily resulted from the formation of variable Pb compounds, depending on solution pH and available anions. Although the precipitation of variable Pb solid phases was the primary mechanism in the removal of aqueous Pb from solution, other sorption processes cannot be excluded as possible attenuation mechanisms for aqueous Pb with apatite, in particular, at neutral to alkaline conditions. Previous studies suggested that ion exchange and adsorption were responsible for sorption of aqueous Pb with synthetic hydroxyapatite (11, 14-16, 43). Additionally, formation of poorly crystalline or noncrystalline phases has also been attributed to Pb removal from solution by apatite (29). However, it is unknown to what extent these mechanisms may be involved. Cadmium. The effects of pH on the sorption of aqueous Cd onto the apatite is substantially different from that of
aqueous Pb (Figure 1), but was very similar to Cd sorption onto calcite and clay minerals (31, 33, 34). The Cd sorption increased with increasing pH. The Cd removal by the apatite varied substantially, ranging from 0.489 to 1.317 mmol of Cd/g of apatite and representing 37% to nearly 100% removal of Cd from solution. At final pH 3.4-6.2, the Cd sorption increased slowly with pH (Figure 1). However, as the suspension pH increased to above 6.2, the Cd sorption increased abruptly to nearly 100%. Selected XRD patterns of the reaction products from the interaction of the apatite with aqueous Cd are shown in Figure 5. No crystalline cadmium phosphates were detected as reported in other studies (24, 28). Nevertheless, otavite (CdCO3) and cadmium hydroxide [Cd(OH)2] were formed in this study:
Cd2+ + HCO3- f CdCO3(c) + H+
(12)
Cd2+ + 2H2O f Cd(OH)2(c) + 2H+
(13)
The sorption of aqueous Cd on the apatite (Figure 1) can be explained by XRD data (Figure 5). Otavite formed in the reactions at all pH except for initial pH 1.1 (Figure 5). At this low pH, otavite is too soluble (5), resulting in a lower removal of Cd from solution (Figure 1). At initial pH > 1.1 (final pH > 3.1), otavite precipitated. But the intensity of its XRD peak was low and remained the same at final pH below 6.2, suggesting that only a small amount of otavite was formed. As final pH increased to above 6.2, a significant increase of XRD intensity of otavite occurred. This, coupled with the occurrence of Cd(OH)2 at very alkaline solutions (with its strongest peaks at initial pH 12.2), could account for the abrupt increase of Cd sorption. The dissolution of the apatite may supply the carbonate required for the precipitation of otavite. When the final pH was below 6.2, the dissolved carbonate mainly occurred as H2CO30 as described in eq 1, resulting in a low activity of carbonate ion. This may be one of the factors for the limited amount of otavite formed in the reactions at final pH below 6.2. Another factor regulating the amount of otavite precipitated was its relatively high solubility at low pH. Theoretically, cadmium phosphate [Cd3(PO4)2] should be less soluble than otavite under acidic condition (5), and such relationship reverses under alkaline condition. However, Cd3(PO4)2 was not identified in our XRD analysis. In our experiments, otavite and crystalline Cd(OH)2 precipitated and dictated the concentrations of aqueous Cd in solution at very low levels under alkaline conditions. The thermodynamic modeling and experimental studies by Rai et al. (44, 45) also suggest that otavite and crystalline Cd(OH)2 have very low solubility under alkaline conditions. In addition to precipitation of solid phases, other sorption mechanisms were also involved in the removal of Cd from
VOL. 31, NO. 3, 1997 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
627
FIGURE 5. XRD patterns of North Carolina apatite before (A) and after interaction with aqueous Cd at initial/final pH of 6.8/6.2 (B), 8.0/7.3 (C), and 10.5/9.7 (D). Peak labels: A, apatite; O, otavite; CH, cadmium hydroxide.
FIGURE 6. XRD patterns of North Carolina apatite before (A) and after interaction with aqueous Zn at initial/final pH of 1.1/3.3 (B), 2.0/5.5 (C), and 8.1/9.2 (D). Peak labels: A, apatite; H, hopeite; Z, zincite.
FIGURE 4. SEM micrographs of North Carolina apatite before (A) and after interaction with aqueous Pb at initial/final pH of 1.1/3.1 (B). solution by the apatite because the formation of solid phases cannot account for all the Cd removed. Jeanjean et al. (46) studied the structural modification of the products from the interaction of synthetic hydroxyapatite with Cd solution. They concluded that Cd ions were sorbed by exchange with Ca and Na ions in the lattice of the apatite. The Cd ions that exchanged with Ca preferentially occupy Ca(2) sites while the concentration of Ca(1) remains unchanged. In an investigation of calcium-cadmium hydroxyapatite coprecipitation, however, Nounah et al. (47) suggested that Cd are mainly located in Ca(2) sites but are also found in Ca(1) sites.
628
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 31, NO. 3, 1997
Many other studies also preferred the ion exchange process as the main mechanism in the sorption of Cd into apatite (15, 16, 23, 48). However, Xu et al. (24) suggested that surface complexation and calcium-cadmium hydroxyapatite coprecipitation were the primary processes in the uptake of Cd by synthetic apatite while ion exchange and solid diffusion might be secondary processes. Dalas and Koutsoukos (49) also argued that the removal of Cd cannot be explained by simple adsorption mechanism and assumed that the coprecipitation of a surface Ca-Cd phase and surface diffusion may be involved. Zinc. The overall sorption behavior of aqueous Zn by the apatite was similar to that of aqueous Cd, with minor differences (Figure 1). One of the differences was observed at the final pH range of 3.3-6.5. In the Cd tests, the sorption increased with pH, while in the Zn tests, the sorption decreased from 0.7953 (37% sorbed) to 0.5964 mmol (27% sorbed) of Zn/g of apatite with the increase of pH. An abrupt increase of sorption took place when the final pH was above 6.5, resulting in nearly all of the aqueous Zn (99.9%) being removed from solution. Figure 6 shows selected XRD patterns of the solid residues from the interaction of the apatite with aqueous Zn at variable pH. Hopeite [Zn3(PO4)2‚4H2O] was identified in both XRD and SEM (Figure 7) examinations. In a study of the interaction
FIGURE 7. SEM micrograph of North Carolina apatite after interaction with aqueous Zn at initial/final pH of 1.1/3.3.
FIGURE 8. XRD patterns of North Carolina apatite before (A) and after interaction with mixed heavy metal solutions at initial/final pH of 1.1/3.3 (B), 3.0/6.0 (C), 6.4/6.2 (D), 10.3/10.9 (E), and 12.2/12.0 (F). See Figures 3, 5, and 6 for peak labels.
of aqueous Zn with hydroxyapatite, Misra and Bowen (25) also reported the formation of hopeite. The dissolution of the apatite (eq 1) supplied dissolved phosphate to the aqueous Zn solution, which was followed by the formation of hopeite:
TABLE 3. Summary of New Solid Phases Formed in Interaction of Mineral Apatite with Mixed Heavy Metal Solutions with Respect to pH
3Zn2+ + 2H2PO4- + 4H2O f Zn3(PO4)2‚4H2O(c) + 4H+ (14) The XRD peaks of hopeite decreased with pH increase. At the initial pH range of 3.1-7.1, no crystalline compound besides the original apatite could be identified. The decrease of Zn removal from solutions as final pH increased from 3.3 to 6.5 may partially be interpreted by the absence of or substantially lower occurrence of hopeite at higher pH. When the initial pH was increased to above 8.0, zincite (ZnO) was detected in the XRD patterns (Figure 6):
Zn2+ + H2O f ZnO + 2H+
(15)
Theoretically, hopeite should be much less soluble than zincite and zinc carbonate (smithsonite, ZnCO3) at acidic condition while zincite and smithsonite should be less soluble than hopeite in an apatite system at alkaline condition (5). Our experimental data confirmed this thermodynamic prediction. In a study of the interaction of aqueous metals with calcite, Zachara et al. (33) concluded that Zn carbonate was not kinetically as easily precipitated as cadmium carbonate. This may explain the absence of smithsonite in our study. Like the interaction of apatites with aqueous Cd, ion exchange has also been proposed as an important mechanism in the removal of Zn from solution (48, 50). Additionally, Xu et al. (24) suggested that the dominant sorption processes involving Zn interaction with hydroxyapatite was surface complexation with hydroxyapatite surface functional groups such as tPOH and tCaOH and coprecipitation of Zn with Ca into the apatite structure. MSST of Heavy Metals with Apatite. Selected XRD patterns of the solid residues obtained from the MSST are shown in Figure 8. The solid phases identified from the patterns are summarized in Table 3, with respect to pH range.
initial pH
final pH
new solid phase
1.1-2.1 3.0-6.4
3.3-5.8 6.0-6.2
8.4-12.2
8.8-12.0
fluoropyromorphite hydroxyl fluoropyromorphite carbonate hydroxyl fluoropyromorphite otavite hopeite hydrocerussite otavite hydroxypyromorphite zincite
Many of the solids formed in the SSST were also found in the MSST, and no new solid phase (for instance, coprecipitate) was identified. Pyromorphite-type solids, hydrocerussite, otavite, and zincite were formed as expected. Hopeite was not formed at initial pH 1-2 as in the SSST (Figure 8). The absence of hopeite may result from the competition of Pb for phosphate ions. Unexpectedly, hopeite occurred in the solid residue from the test at initial pH 6.4 (Figure 8), which cannot be explained at present. The oxides and hydroxide of Pb and Cd that were recognized in the SSST could not be identified in the MSST. In the MSST, competitive sorption among the aqueous heavy metals is expected. Figure 9 shows the concentrations of Pb, Cd, and Zn after sorption onto the apatite with respect to pH. Comparing the sorption in the MSST (Figure 9) and those in the SSST (Figure 1) suggests that the sorption behaviors of aqueous Pb are basically identical for both SSST and MSST. They both resulted in nearly 100% Pb removal for most of the pH conditions, suggesting that the sorption of Pb was not affected by the presence of aqueous Cd and Zn. The overall sorption behaviors of aqueous Cd and Zn in the MSST were similar to those in the SSST, but with a minor difference. For easy comparison, their sorption data are re-
VOL. 31, NO. 3, 1997 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
629
From the XRD data and sorption information, it seems that the mechanisms responsible for Pb removal from solution by the apatite remained unchanged in the presence of aqueous Cd and Zn. However, the competitive effects among the heavy metals may alter the sorption behaviors of Cd and Zn to some extent. Here, the competitive effects may include competition for adsorption sites as well as competition for precipitation onto apatites. When the heavy metals were applied individually, they could precipitate as metal phosphates, carbonates, hydroxides, and oxides at their respective precipitation pH with competition solely between ions of the same metal (internal competition) and they were in competition only with H+ for adsorption sites. However, when the heavy metals were applied in their combination, there was internal competition, competition with H+, and competition with each other for precipitation and for adsorption sites. In the MSST, the absence of hopeite at final pH 3.3 might be the result of competition between Pb and Zn for precipitation.
Acknowledgments FIGURE 9. Heavy metal concentrations after reaction with North Carolina apatite with respect to pH in MSST.
We are grateful to Tim Moody, Paul Didzerekis, T. Joseph Mockler, Scott Cornelius, Charles Knaack, and Chris Davit for their excellent assistance. Reviews by James Amonette, Wayne Martin, and three anonymous reviewers led to significant improvements. Financial support was provided by the U.S. Department of Defense through the Strategic Environmental Research and Development Program. Pacific Northwest National Laboratory is operated for the U.S. Department of Energy by Battelle under Contract DE-AC0676RLO 1830.
Literature Cited
FIGURE 10. Comparison of sorption of aqueous Cd (A) and Zn (B) in North Carolina apatite with respect to pH in SSST and MSST. graphed in Figure 10. The primary difference in the removal of Cd and Zn between SSST and MSST occurred at acidic to neutral pH. With an increase of suspension pH from 3.3 to 6.5, the amount of Cd sorbed increased in both SSST and MSST (Figure 10A) while the amount of Zn sorbed increased in the MSST, but decreased in the SSST (Figure 10B). This can be explained by the absence of hopeite precipitation at initial pH 1.1 (final pH 3.3) in the MSST, but it is present in the SSST.
630
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 31, NO. 3, 1997
(1) McConnell, D. Apatites: Its Crystal Chemistry, Mineralogy, Utilization, and Geologic and Biologic Occurrences; SpringerVerlag: Berlin, 1973. (2) Wright, J. In Skeletal Biomineralization: Patterns, Processes and Evolutionary Trends. Volume I; Carter, J. G., Ed.; Van Nostrand Reinhold: New York, 1990; pp 445-459. (3) Altschuler, Z. S. Soc. Econ. Paleontol. Mineral. Spec. Publ. 1980, 29, 19-30. (4) Wright, J.; Schrader, H.; Holser, W. T. Geochim. Cosmochim. Acta 1987, 51, 631-644. (5) Lindsay, W. L. Chemical Equilibria in Soils; John Wiley & Sons: New York, 1979. (6) McClellan, G. H. J. Geol. Soc. London 1980, 137, 675-681. (7) McClellan, G. H.; Lehr, J. R. Am. Mineral. 1969, 54, 1379-1391. (8) LeGeros, R. Z. Prog. Crystal Growth Charact. 1981, 4, 1-45. (9) LeGeros, R. Z.; LeGeros, J. P. In Phosphate Minerals; Nriagu, J. O., Moore, P. B., Eds.; Springer-Verlag: Berlin, 1984; pp 351385. (10) Chien, S. H.; Black, C. A. Soil Sci. Soc. Am. J. 1976, 40, 234-239. (11) Suzuki, T.; Ishigaki, K.; Miyake, M. J. Chem. Soc., Faraday Trans. 1 1984, 80, 3157-3165. (12) Ma, Q. Y.; Traina, S. J.; Logan, T. J.; Ryan, J. A. Environ. Sci. Technol. 1993, 27, 1803-1810. (13) Xu, Y.; Schwartz, F. W. J. Contam. Hydrol. 1994, 15, 187-206. (14) Miyake, M.; Ishigaki, K.; Suzuki, T. J. Solid State Chem. 1986, 61, 230-235. (15) Takeuchi, Y.; Suzuki, T.; Arai, H. J. Chem. Eng. Jpn. 1988, 21, 98-100. (16) Takeuchi, Y.; Arai, H. J. Chem. Eng. Jpn. 1990, 23, 75-80. (17) Ma, Q. Y.; Logan, T. J.; Traina, S. J.; Ryan, J. A. Environ. Sci. Technol. 1994, 28, 408-418. (18) Nriagu, J. O. Inorg. Chem. 1972, 11, 2499-2503. (19) Nriagu, J. O. Geochim. Cosmochim. Acta 1973, 37, 367-377. (20) Nriagu, J. O. Geochim. Cosmochim. Acta 1973, 37, 1735-1743. (21) Nriagu, J. O. Geochim. Cosmochim. Acta 1974, 38, 887-898. (22) Nriagu, J. O. In Phosphate Minerals; Nriagu, J. O., Moore, P. B., Eds.; Springer-Verlag: Berlin, 1984; pp 318-329. (23) Middelburg, J. J.; Comans, R. N. J. Chem. Geol. 1991, 90, 45-53. (24) Xu, Y.; Schwartz, F. W.; Traina, S. J. Environ. Sci. Technol. 1994, 28, 1472-1480. (25) Misra, D. N.; Bowen, R. L. In Adsorption from Aqueous Solutions; Tewari, P. H., Ed.; Plenum: New York, 1981; pp 179-192. (26) Sposito, G. In Geochemical Processes at Mineral Surfaces; Davis, J. A., Hayes, K. F., Eds.; American Chemical Society Symposium
(27) (28) (29) (30) (31) (32) (33) (34) (35) (36) (37) (38) (39) (40)
Series 323; American Chemical Society: Washington, DC, 1986; pp 217-228. Suzuki, T.; Ishigaki, K.; Ayuzawa, N. Chem. Eng. Commun. 1985, 34, 143-151. Ma, Q. Y.; Traina, S. J.; Logan, T. J.; Ryan, J. A. Environ. Sci. Technol. 1994, 28, 1219-1228. Ma, Q. Y.; Logan, T. J.; Traina, S. J. Environ. Sci. Technol. 1995, 29, 1118-1126. Harter, R. D. Soil Sci. Soc. Am. J. 1983, 47, 47-51. Elliott, H. A.; Liberati, M. R.; Huang, C. P. J. Environ. Qual. 1986, 15, 214-219. Davis, J. A.; Fuller, C. C.; Cook, A. D. Geochim. Cosmochim. Acta 1987, 51, 1477-1490. Zachara, J. M.; Cowan, C. E.; Resch, C. T. Geochim. Cosmochim. Acta 1991, 55, 1549-1562. Yong, R. N.; Phadungchewit, Y. Can. Geotech. J. 1993, 30, 821-833. Chien, S. H.; Clayton, W. R.; McClellan, G. H. Soil Sci. Soc. Am. J. 1980, 44, 260-264. Gibbs, R. J. In Procedures in Sedimentary Petrology; Carver, R. E., Ed.; Wiley-Interscience: New York, 1971; pp 531-539. Farrah, H.; Pickering, W. F. Chem. Geol. 1979, 25, 317-326. Stumm, W.; Morgan, J. J. Aquatic Chemistry, 3rd ed.; John Wiley & Sons: New York, 1996. Drever, J. I. The Geochemistry of Natural Waters; Prentice Hall: Englewood Cliffs, NJ, 1988. Rickard, D. T.; Nriagu, J. O. In The Biogeochemistry of Lead in the Environment; Nriagu, J. O., Ed.; Elsevier/North-Holland Biomedical Press: Amsterdam, 1978; pp 219-284.
(41) Christoffersen, J.; Christoffersen, M. R. J. Crystal Growth 1982, 57, 21-26. (42) Kanabo, I. A. K.; Gilkes, R. J. Fert. Res. 1987, 12, 165-174. (43) Suzuki, T.; Hatsushika, T.; Miyake, M. J. Chem. Soc., Faraday Trans. 1 1982, 78, 3605-3611. (44) Rai, D.; Felmy, A. R.; Szelmeczka, R. W. J. Solution Chem. 1991, 20, 375-390. (45) Rai, D.; Felmy, A. R.; Moore, D. A. J. Solution Chem. 1991, 20, 1169-1187. (46) Jeanjean, J.; Vincent, U.; Fedoroft, M. J. Solid State Chem. 1994, 108, 68-72. (47) Nounah, A.; Lacout, J. L.; Savariault, J. M. J. Alloys Compd. 1992, 188, 141-146. (48) Suzuki, T.; Hatsushika, T.; Hayakawa, Y. J. Chem. Soc., Faraday Trans. 1 1981, 77, 1059-1062. (49) Dalas, E.; Koutsoukos, P. G. J. Chem. Soc., Faraday Trans 1 1989, 85, 3159-3164. (50) Ingram, G. S.; Horay, C. P.; Stead, W. J. Caries Res. 1992, 26, 248-253.
Received for review November 22, 1995. Revised manuscript received October 14, 1996. Accepted October 22, 1996.X ES950882F X
Abstract published in Advance ACS Abstracts, January 15, 1997.
VOL. 31, NO. 3, 1997 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
631
