Environ. Sci. Technol. 2008, 42, 436–442
Carbonate and Magnesium Interactive Effect on Calcium Phosphate Precipitation XINDE CAO AND WILLIE HARRIS* Soil and Water Science Department, University of Florida, Gainesville, Florida 32611
Received July 6, 2007. Revised manuscript received October 16, 2007. Accepted November 6, 2007.
Precipitation of Ca phosphates, an important process in controlling P stability and activity in P-fertilized soils and P-rich wastewater, is often affected by other components. The purpose of this study was to document interactive effects of CO32- and Mg2+ on Ca phosphate precipitation under conditions simulating (i) dairy manure-amended soil leachate (system I; pH 7.1) and (ii) P recovery from flushed dairy manure wastewater (system II; pH 9.2). Hydroxyapatite (HAP) and more soluble amorphous Ca phosphate (ACP) were formed in the control solutions of system I and system II, respectively. Carbonate only slightly affected the crystallinity of the precipitate, but significantly reduced the precipitation rate via CO32- competition for PO43- (system I) or preemptive CaCO3 precipitation (system II). Magnesium severely inhibited both precipitate crystallinity and precipitation rate, allowing formation of ACP in both systems, presumably due to Mg2+ incorporation into the Ca phosphate structure to form Mg2+-substituted structure that crystallized to whitlockite upon heating. Coexistence of CO32and Mg2+ in system I showed a synergistic inhibitory effect, compared to their individual presence, probably because both CO32- and Mg2+ were incorporated into the precipitate. However, in system II, the individual inhibitory effect of CO32or Mg2+ was eliminated when both were present. The likely mechanism involves formation of aqueous MgCO3 (aq) which reduces free CO32- and Mg2+ activities, resulting in less preemptive CaCO3 formation and enhanced Ca phosphate precipitation.
Introduction Calcium phosphate minerals play an important role in controlling P stability and activity in the P-fertilized soils and P-rich wastewater (1, 2). The most stable Ca phosphate mineral is hydroxyapatite (Ca5(PO4)3OH; HAP), but other Ca phosphates reported in the environment include octacalcium phosphate (Ca4H(PO4)3 · 2.5H2O; OCP), dicalcium phosphate dihydrate (CaHPO4 · 2H2O; DCPD), and amorphous Ca phosphate (Ca3(PO4)2 · nH2O; ACP) (1). Formation of more soluble Ca phosphates can be favored kinetically over more stable phases because of faster nucleation. A number of ions can inhibit Ca-P precipitation by forming a surface complex on the newly forming surfaces that disrupt nucleation (1). Inhibitors include organic acids (1, 3), CO32- and Mg2+. The influence of CO32- or Mg2+ on Ca phosphate precipitation has been discussed by some * Corresponding author phone: (352) 392-1951, ext. 251; fax: (352) 392-3902; e-mail:
[email protected]. 436
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 2, 2008
authors (1, 4-8). In a seeded precipitation experiment, Salimi et al. (4) showed that Mg2+ has a marked inhibiting effect on HAP growth, despite a lesser effect on OCP, and almost no influence on the growth of DCPD. They attributed the inhibitory effect of Mg2+ to its adsorption at active growth sites. Results of spontaneous precipitation experiments suggested that Mg2+ kinetically hinders nucleation and subsequent growth of HAP by competing for structural sites with chemically similar but larger Ca2+ (5). Therefore, it is possible for HAP to incorporate a small percentage of Mg2+ into its structure, but this causes structural changes and has an inhibitory effect on longrange order and hence HAP formation. Coprecipitation of Mg2+ with Ca phosphate promotes formation of ACP rather than HAP (6). The presence of CO32- may also affect precipitation of HAP, either by blocking phosphate nucleation sites or by inducing CaCO3 precipitation instead (7). Apfelbaum et al. (8) found that >3.5% CO32- substitution in apatite markedly reduced the crystallinity and solubility of apatite crystals. Nelson (9) observed an increase in disordered regions between apatite crystal domain structures with increased carbonate content. The individual effect of CO32- and Mg2+ on Ca phosphate precipitation has been extensively investigated. However, their interactive effect has never to our knowledge been tested under environmentally relevant conditions. A previous study testing the efficacy of P recovery from flushed dairy manure wastewater using a fluidized bed reactor documented that Mg2+, added originally to encourage struvite (MgNH4PO4) formation, actually enabled P recovery in the form of calcium phosphate at elevated pH (10). The manure wastewater had high dissolved CO32-, raising a hypothesis that Mg2+ was forming an aqueous MgCO3 (aq) at high pH to prevent CaCO3 precipitation, a reaction otherwise preemptive of Ca phosphate precipitation. The general purpose of the present study was to document the interactive effects of CO32- and Mg2+ on Ca phosphate formation and crystallinity under chemical conditions that could occur for (i) dairy manure-amended soils and (ii) phosphate recovery or wastewater treatment scenarios. Implications for effects of NH4+ and the prospect of struvite as an alternative P recovery pathway are also addressed.
Materials and Methods Precipitation Experiment. The pH and solution concentrations of PO43-, CO32-, Ca2+, and Mg2+ were selected to represent two systems (I and II; Table 1 and Table S1 of the Supporting Information (SI)): manure-amended soil leachate solutions (11) and flushed dairy manure wastewater (10), respectively. Each system included one control and three treatments with three replicates: control (CK), carbonate (CO3), magnesium (Mg), and carbonate + magnesium (CO3+Mg). Other ions (e.g., Na+, NH4+) that commonly occur in real solutions and that would affect ionic strength and charge balance were approximately represented by 20 mM KCl (Table 1). Recovering P as Ca phosphates precipitate from wastewater requires elevation of pH between 8.5 and 9.5 (10), so a pH of 9.2 was used in this study for all treatments in system II (Table 1). A chemical speciation model (Visual MINTEQ, ver. 2.32) was used to calculate solution species and identify phases that could precipitate (12) (SI Table S2). Modeling results indicated that the two controls (CK) were supersaturated with respect to ACP, DCPD, OCP, and HAP, and the system II solutions had higher saturation than that in system I (SI Table S2). 10.1021/es0716709 CCC: $40.75
2008 American Chemical Society
Published on Web 12/14/2007
elucidate the possible mechanism responsible for the inhibitory effect, a portion of the as-precipitated solids were heated at 550 °C for 3 h and cooled down at room temperature. The phases of the as-precipitated and heat-treated solids were identified using X-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FTIR) (Supporting Information). Statistic Analysis. All results were expressed as an average of three replicates with standard deviation. Treatment effects were determined by analysis of variance according to the general linear model procedure of the Statistical Analysis System (SAS Institute Inc.). Differences among means were distinguished by least significant difference at p e 0.05.
TABLE 1. Initial Chemical Composition and Concentrations of the Solutions Used in This Study (25 ± 2 °C, 20 mM KCl) PO43-T
a
Ca2+T
CO32-T
system
treatment
pH
I
CK (no CO3 and Mg) b CO3 Mg CO3+Mg CK (no CO3 and Mg) CO3 Mg CO3 + Mg
7.1
3
5
7.1 7.1 7.1 9.2
3 3 3 1
5 5 5 3
10
9.2 9.2 9.2
1 1 1
3 3 3
15
II
Mg2+T
10
15
10 10
15 15
Results and Discussion Precipitation of Ca phosphates in absence of CO32- and Mg2+. Phosphorus in both control systems rapidly decreased within the first 10 min due to precipitation, followed by a slow decline until 60 min and leveling off afterward (SI Figure S1a and b), agreeing with the observation of previous work (13). However, P declined faster in system II than in system I. By the end of the experiment (6 h), as much as 85% P was lost in system I, with the precipitation rate constant of ∼31 M-1 · min-1; while in system II, 96% P disappeared, with a significantly higher precipitation rate constant (345 M-1 · min-1) (Table 2). Hydroxyapatite formed in system I although it was poorly crystalline as indicated by three broad HAP-derived main peaks at 2θ ) 25.8°, 28.3°, and 31.9° (Figure 1a) . The molar ratio of Ca/P in the solid was 1.65 (Table 3), close to the stoichiometric value of 1.67 for HAP (SI Table S4). Heating the HAP powder improved resolution of the XRD peaks (Figure 1a′), especially the peak at 2θ ) 34°, due to increased crystallization of HAP at elevated temperature (6). System II was supersaturated with respect to the most stable HAP (SI Table S2), but only soluble ACP precipitated as shown by a typical broad peak of the ACP between 2θ ) 25° and 35° (16) (Figure 2a), probably due to high pH (9.2) and high saturation (SI ) 18) (SI Table S2). High alkalinity and high supersaturation favors formation of thermodynamically unstable ACP because energy savings in the precipitation greatly outweigh those of lattice ordering into HAP (7, 16). The molar ratio of Ca/P for ACP was 1.47 (Table 3), which is on the high end of the range of values given in the literature (16). Heat treatment induced crystallization of ACP, transforming it into crystalline HAP (Figure 2a′). Carbonate Effect. The effect of carbonate in the absence of Mg2+ on precipitation of Ca phosphates varied for two systems. In system I, CO32- slightly reduced crystallization of HAP precipitate, as evidenced by a lower XRD peak resolution (disappearance of HAP-derived peak at 2θ ) 28.3°) compared to the control precipitate HAP (Figure 1a and b). However, CO32- significantly reduced the precipitation rate
a
Total ion concentration (mM). All chemical solutions were prepared using reagent-grade chemicals (Fisher Scientific Co.). Stock solutions of 0.5 M Ca2+, 0.3 M PO43-, 1.0 M Mg2+, and 0.5 M CO32- were made by dissolving CaCl2 · H2O, KH2PO4, MgSO4 · 6H2O, and NaHCO3 in double deionized water prepurged with N2. b Treatment abbreviation: CK, control in absence of CO32- and Mg2+; CO3, treatment with presence of CO32-; Mg, treatment with presence of Mg2+; and CO3 + Mg, treatment with presence of both CO32- and Mg2+.
Experiments were conducted in a 250 mL PYREX flask containing 20 mM KCl as electrolyte solution at 25 ( 2 °C. Chemicals were introduced sequentially to avoid possible preemptive chemical reactions (SI Table S3). When addition of chemicals was completed, the solution pH was immediately brought to and maintained at pH 7.1 for system I and 9.2 for system II with 0.1 M KOH or 0.1 M HCl. The reaction mixtures were stirred continuously and covered with Parafilm during precipitation to minimize atmospheric CO2 entering the solution. The precipitation was performed for 6 h which was determined to be enough for the complete precipitation in a previous study (13). Five milliliters of the liquid were collected at various time intervals and filtered (0.22 µm). The filtrate was immediately acidified to pH < 2 with concentrated HNO3 prior to chemical analysis. Solution Ca and Mg concentrations were determined by atomic absorption spectroscopy and P concentration by colorimetry using the molybdate/ascorbic acid method (14). Precipitated solids (referred to as “as-precipitated solids”) at the end of the experiment (6 h) were collected and washed free of P, Ca, and Mg. The “clean” solids were air-dried for the subsequent physical and chemical characterization. Precipitate Solid Characterization. Chemical analysis of the precipitate solids was performed by first digesting the solids following U.S. Environmental Protection Agency method 3052 (15) and measuring element concentrations in the solutions. It has been found that elevated temperature enhances crystallinity of precipitation phases (6). Thus, to
TABLE 2. Final Solution Chemical Concentration and Precipitation Rate Constant (25 ± 2 °C, 20 mM KCl) system I treatment CK CO3 Mg CO3 + Mg
PO43-Ta c
0.46c 0.66c 2.26b 2.64a
Ca2+T 1.06c 1.11c 3.29b 4.07a
Mg2+T
system II CO32-T
d
–
9.10a 9.21a 8.65a
9.17a
kb
PO43-Ta
Ca2+T
30.6a 19.0b 1.72c 0.76 cd
0.04b 0.22a 0.28a 0.05b
1.63a 0.78b 0.82b 1.69a
Mg2+T
CO32-T 11.2b
13.5a 14.7a
14.5a
k 345a 47.9b 40.7b 293a
a
Total ion concentration (mM). b Precipitation rate constant was calculated by measuring decrease of P concentration in solutions during the first 60 min using the equation: k ) 1⁄t × (1⁄Ct - 1⁄C0) , described in our previous work (13), where k is a precipitation rate constant with a unit of M-1 min-1; t is time (min); C0 is the initial P concentration (3 × 10-3 M in system I and 1 × 10-3 M in system II); and Ct is the solution P concentration at time t. c All data are presented as mean of three replicates (n ) 3). Data with the same letter in the same column are not significantly different (p < 0.05). d Below detection limit.
VOL. 42, NO. 2, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
437
FIGURE 1. XRD patterns of the as-precipitated solids from system I with (a) CK; (b) CO3; (c) Mg, and (d) CO3 + Mg at 25 °C; and the precipitates heated at 550 °C for 3 h (a′) CK; (b′) CO3; (c′) Mg, and (d′) CO3 + Mg. The minerals with peak labeled: A, amorphous Ca phosphate; H, hydroxyapatite; W, whitlockite.
TABLE 3. Element Percentage and Ca/P Molar Ratios in the Precipitated Solid Determined by Chemical Analysis system I treatment CK CO3 Mg CO3 + Mg
Pa
Ca c
17.9 17.0 16.3 16.0
Na d
39.8 29.7 23.8 22.6
a Element percentage (%). detection limit.
– 0.23 0.08 b
system II K
0.02 0.12 0.03 0.05
Ca/P molar ratio.
c
Mg
Ca/Pb
P
Ca
0.39 0.49
1.65 1.46 1.13 1.02
16.2 8.71 15.6 11.5
30.8 38.6 22.7 20.5
9
0.11 0.03
K 0.02 0.05 0.03 0.10
Mg
Ca/P
0.51 0.12
1.47 3.44 1.13 1.37
All data are presented as mean of three replicates (n ) 3).
constant by ∼38%, maintaining higher Ca and P in solution (Table 2). This precipitation suppression resulted from CO32substitution for PO43- into the HAP structure, as confirmed by IR analysis (Figure 3a and b) . The control precipitate showed a typical IR spectrum of HAP (6), with presence of an OH--derived band at 630 cm-1 and PO43--derived bands at 1090, 1050, 603, and 569 cm-1 (Figure 3a). The peak at 1652 cm-1 is associated with adsorbed H2O. Typical peaks of CO32- at 1455 and 1420 cm-1 (17) were significantly increased by CO32- treatment (Figure 3b) compared to the control, which is consistent with some CO32--for-PO43substitution in the HAP structure (6, 17, 18). The presence of small amounts of CO32- in the control was probably due to some atmospheric CO2 despite efforts to minimize that input to solutions via prepurging with N2 and covering with Parafilm. Disappearance of resolution between peaks at1090 and 1050 cm-1 in the presence of CO32- (Figure 3b) may result from CO32- substitution. Although a peak at 872 cm-1 is shared by HPO42- and CO32, emergence of the peak only in the CO32- treatment solid (Figure 3b) is also evidence of CO32- substitution. Disappearance of an OH--derived band at 630 cm-1 in the presence of CO32- (Figure 3b) indicated a possible CO32--for-OH- substitution. A similar tendency was observed with Ca-deficient carbonate apatites, in which CO32- replaced OH-, resulting in a low OH- content (8). The absence of a CO32- band in the region 720–700 ruled out the presence of CaCO3 as a separate phase (16). Precipitation of calcite is less likely at pH ) 7.1 (system I) where CO32represents less than 0.1% of carbonate species (SI Figure S2). 438
Na
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 2, 2008
d
Below
XRD analysis of heat-treated precipitate provided further evidence of CO32- substitution; HAP-derived XRD peaks at 2θ ) 28 o and 34°, which were present for the control, were not present for CO3 treatment (Figure 1a′ and b′). Incorporation of CO32- should favor an increase in the Ca/P molar ratio. However, a lower Ca/P (1.46) was observed (Table 3), possibly because the CO32- substitution effect is offset by substitution of other cations (e.g., K+, Na+) for Ca2+ in the structure (19). The CO32--treated precipitate contained 0.2% Na and 0.1% K, much higher than those in the control (Table 3). In system II, high pH (9.2) resulted in as much as 15% CO32- species (SI Figure S2), which is favorable for CO32precipitated as calcite, and calcite accompanied ACP formation as confirmed by XRD (Figure 2b). FTIR spectra showed calcite-derived peaks at 720 and 1799 cm-1 in addition to some CO32- substitution (Figure 3c and d). Heat treatment resulted in crystallization of ACP into HAP (Figure 2b′). As a result of calcite precipitation, the Ca-P precipitation rate constant was reduced by as much as 86% (Table 2), while the ratio of Ca/P increased up to 3.44 due to high Ca from both calcite and ACP (Table 3). Comparing the effect of CO32- between the two systems, two different mechanisms contributed to inhibition of Ca phosphate precipitation. In system I, lower pH (7.1) favored substitution of CO32- for PO43- while higher pH (9.2) in system II allowed more CO32- species to form calcite,
FIGURE 2. XRD patterns of the as-precipitated solids from system II with (a) CK; (b) CO3; (c) Mg, and (d) CO3 + Mg at 25 °C; and the precipitates heated at 550 °C for 3 h (a′) CK; (b′) CO3; (c′) Mg, and (d′) CO3 + Mg. The minerals with peak labeled: A, amorphous Ca phosphate; C, calcite; H, hydroxyapatite; W, whitlockite.
FIGURE 3. IR spectra of the precipitates in system I with (a) absence and (b) presence of CO32- and in system II with (c) absence and (d) presence of CO32-. agreeing with the finding of Ferguson and McCarty (5) who showed the absence of CaCO3 at low pH and the presence at pH 8–11. Magnesium Effect. The presence of Mg2+ without CO32addition consistently inhibited Ca phosphate precipitation in both systems (Table 2). After 6 h, as much as 75% P was still in system I solution, much higher than the 15% P for the control. Precipitation rates were significantly reduced from ∼31 M-1 · min-1 to ∼1.7 M-1 · min-1 (95% reduction) (Table 2). In system II, P concentration was 7 times that in the control. The precipitation rate constant was reduced by ∼88% (Table 2).
XRD analysis revealed that precipitates were soluble ACP in both systems in presence of Mg2+ (Figures 1c and 2c), with the net effects of minimizing Ca phosphate precipitation and maintaining higher P concentrations in solution (Table 2). This inhibitory effect may be attributed to incorporation of Mg2+ and consequent reduction of long-range order due to the smaller size of Mg2+ and its greater tendency to bond covalently (5). The presence of Mg2+ has been shown to inhibit conversion from ACP to HAP (6). Incorporation of Mg2+ into the precipitated Ca phosphate phase was supported by XRD analysis of heat-treated solids which verified formation of Mg2+-containing whitlockite (Ca, Mg)3(PO4)2 (Figures 1c′ and VOL. 42, NO. 2, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
439
FIGURE 4. MgCO3(aq) percentage of solution carbonate species in the two systems as a function of Mg2+/CO32- molar ratio. 2c′). As a result of the incorporation, Mg2+ concentrations in system I and II solutions were reduced by 8 and 15%, respectively, (Table 2), and the solids contained ∼0.4 and ∼0.5% Mg, respectively, corresponding to a decrease in the molar ratios of Ca/P to ∼1.1 (Table 3). Carbonate and Magnesium Interactive Effect. Solutions containing both CO32- and Mg2+ (CO3+Mg treatment) behaved differently from solutions containing each individually and there was a contrast between the two systems. In system I, the CO3+Mg solution maintained higher P than those in the control and in solutions containing each
individually (Table 2). The precipitation rate constant was reduced to as low as 0.76 M-1 · min-1 (98% reduction). In effect, coexistence of CO32- and Mg2+ resulted in synergistic inhibition of Ca phosphate precipitation. CO32- and Mg2+ were incorporated into the Ca phosphate structure, resulting in formation of ACP (Figure 1d). XRD analysis of the heattreated precipitate confirmed formation of Mg-substituted whiltlockite (Figure 1d′). Up to 0.5% Ca was replaced by Mg in the precipitate. As a result of CO32- and Mg2+ incorporation, solution CO32- and Mg2+ concentrations were reduced by 8 and 14%, respectively (Table 2), and the molar ratio of Ca/P decreased to ∼1 (Table 3). In system II, CO32- and Mg2+ coexistence exhibited an antagonistic effect. Solution P concentrations were reduced to below those in the presence of individual CO32- or Mg2+, down close to that in the control. The precipitation rate was 293 M-1 · min-1, comparable to the 345 M-1 · min-1 for the control. Results confirm that at pH 9.2 the inhibition of P precipitation caused by CO32- or Mg2+ was mitigated by presence of the other component. The likely mechanism is that CO32- and Mg2+ formed the stable aqueous ion pair MgCO3 (aq). This hypothesis is consistent with the speciation model showing that up to 12% MgCO3 (aq) would form in system II but that system I solutions would contain only 0.1% of that species (aq) (Figure 4). Since CO32- was associated with Mg2+ under system II conditions, free CO32- and Mg2+ activities were reduced, making calcite precipitation (a competitive reaction) and Mg2+ substitution less likely. There was no detectable loss of Mg2+ from solutions (Table 2). Heat treatment of the precipitated phase promoted transformation of ACP to crystalline HAP; no calcite or Mg2+-substituted whitlockite was detected (Figure 2d′). The ability of Mg2+ to inhibit calcite formation was further evidenced by varying Mg2+/CO32- ratios. Increasing Mg2+/CO32- from 0.1 to 2 gradually suppressed calcite formation with increasing formation of ACP (Figure 5a-d). XRD analysis of the heat-
FIGURE 5. XRD patterns of the as-precipitated solids from the CO3 + Mg treatment of the system II with (a) Mg2+/CO32- ) 2; (b) Mg2+/CO32- ) 1; (c) Mg2+/CO32- ) 0.1; and (d) Mg2+/CO32- ) 0 at 25 °C and the precipitates heated at 550 °C for 3 h (a′) Mg2+/CO32) 2; (b′) Mg2+/CO32- ) 1; (c′) Mg2+/CO32- ) 0.1; and (d′) Mg2+/CO32- ) 0. The mineral peaks are labeled C, calcite; A, amorphous Ca phosphate; and H, hydroxyapatite. 440
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 2, 2008
treated precipitate confirmed a decrease of calcite, as shown by gradual decline of calcite peaks with increasing Mg2+/ CO32- ratio (Figure 5a′-d′). Interestingly, even at Mg2+/CO32ratio as high as 2, no Mg-substituted whitlockite was formed (Figure 5a′). This may be because more Mg2+ associates with CO32- to form a stable aqueous MgCO3 at higher ratios of Mg2+/CO32-. Modeling shows that MgCO3 (aq) is positively correlated to the Mg2+/CO32- molar ratio, and its formation is much more sensitive to that ratio in system II than in system I (Figure 4). For example, as high as 18% of CO32- was formed as MgCO3 (aq) in system II when Mg2+/CO32- was increased to 2, whereas system I contained only 0.15% CO32as MgCO3 (aq) (Figure 4). Possible Mitigation by NH4+-Induced Struvite Formation. The presence of NH4+ raises the prospect of struvite formation as an alternative to calcium phosphate (20). Nuisance struvite precipitation in animal waste conveyances has been reported as a common occurrence, and researchers have documented the efficacy of struvite recovery from swine manure wastewater (21–23). However, the greater solubility of struvite compared to hydroxyapatite thermodynamically favors the latter in some wastewaters where the range of Ca2+ activity is relatively high (1, 10, 24). Poorly crystalline apatite can form fast enough to be recovered in a fluidized bed reactor (10) or in brief ( 0 in two systems with four treatments each; Sequence of chemicals addition in each treatment; a list of several common calcium phosphate minerals; change of solution P concentration with time during Ca phosphate
precipitation; percentage distribution of dissolved carbonate species in the CO32--H2O system as a function of pH; modeling of ion activity products for struvite and hydroxyapatite as related to their solubility products. This information is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Valsami-Jones, E. Mineralogical controls on phosphorus recovery from wastewaters. Mineral. Mag. 2001, 65, 611–620. (2) Hu, C.; Zhang, T. C.; Huang, Y. H.; Dahab, M. F.; Surampalli, R. Effects of long-term wastewater application on chemical properties and phosphorus adsorption capacity in soils of a wastewater land treatment system. Environ. Sci. Technol. 2005, 39, 7240–7245. (3) Grossl, P. R.; Inskeep, W. P. Precipitation of dicalcium phosphate dihydrate in the presence of organic acids. Soil Sci. Soc. Am. J. 1991, 55, 670–675. (4) Salimi, M. H.; Heughebaert, J. C.; Nancollas, G. H. Crystal growth of calcium phosphates in the presence of magnesium ions. Langmuir 1985, 1, 119–122. (5) Ferguson, J. F.; McCarty, P. L. Effects of carbonate and magnesium on calcium phosphate precipitation. Environ. Sci. Technol. 1971, 5, 534–540. (6) Suchanek, W. J.; Byrappa, K.; Shuk, P.; Riman, R. E.; Janas, V. F.; Tenhuisen, K. S. Mechanochemical-hydrothermal sysnthesis of calcium phosphate powders with coupled magnesium and carbonate substitution. J. Solid State Chem. 2004, 17, 793–799. (7) van der Houwen, J. A. M.; Valsami-Jones, E. The application of calcium phosphate precipitation chemistry to phosphorus recovery: The influence of organic ligands. Environ. Technol. 2001, 22, 1325–1335. (8) Apfelbaum, F.; Mayer, I.; Rey, C.; Lebugle, A. Magnesium in maturing synthetic apatite: a Fourier transform infrared analysis. J. Cryst. Growth. 1994, 144, 304–310. (9) Nelson, D. G. A. The influence of carbonate on the atomic structure and reactivity of hydroxyapatite. J. Dent. Res. 1981, 60, 1621–1629. (10) Harris, W. G.; Wilkie, A. C.; Cao, X.; Sirengo, R. Bench-scale recovery of phosphorus from flushed dairy manure wastewater. Bioresour. Technol. 2007, doi:10.1016/j.biortech.2007.06.065 (in press). (11) Josan, M. S.; Nair, V. D.; Harris, W. G.; Herrera, D. Associated release of magnesium and phosphorus from active and abandoned dairy soils. J. Environ. Qual. 2005, 34, 184–191. (12) Gustafsson, J. P. Visual MINTEQ (version 2.32); Department of Land and Water Resources Engineering; The Royal Institute of Technology: Stockholm, Sweden, 2005; http://www.lwr.kth.se/ english/OurSoftWare/Vminteq/. (13) Cao, X.; Harris, W. G.; Josan, M. S.; Nair, V. D. Inhibition of calcium phosphate precipitation under environmentallyrelevant conditions. Sci. Total Environ. 2007, 383, 205–215. (14) Olsen, S. R.; Sommers, L. E. Phosphorus. In Methods of Soil Analysis. Part 2: Chemical and Microbiological Properties No. 9, 2nd Edition; Page, A. L., Miller, R. H., Keeney, D. R., Eds.; ASA: Madison, WI, 1982. (15) U.S. Environmental Protection Agency (USEPA). Test Methods for Evaluating Solid Waste, Physical/Chemical Methods, SW846 3rd ed. Method 3052; USEPA: Washington, DC, 1995. (16) Alvarez, R.; Evans, L. A.; Milham, P. J.; Wilson, M. A. Effects of humic material on the precipitation of calcium phosphate. Geoderma 2004, 118, 245–260. (17) Kapolos, J.; Koutsoukos, P. G. Formation of calcium phosphates in aqueous solutions in the presence of carbonate ions. Langmuir 1999, 15, 6557–6562. (18) van der Houwen, J. A. M.; Cressey, G.; Cressey, B. A.; ValsamiJones, E. The effect of organic ligands on the crystallinity of calcium phosphate. J. Cryst. Growth 2003, 249, 572–583. (19) Nriagu, J. O.; Moore, P. B. Phosphate Minerals; Springer-Verlag Berlin Heidelberg: New York, Tokyo. 1984. (20) Morse, G. K.; Brett, S. W.; Guy, J. A.; Lester, J. N. Review: Phosphorus removal and recovery technologies. Sci. Total Environ. 1998, 212, 69–81. (21) Burns, R. T.; Moody, L. B.; Celen, I.; Buchanan, J. R. Optimization of phosphorus precipitation from swine manure slurries to enhance recovery. Water Sci. Technol. 2003, 48, 139–146. (22) Bowers, K. E.; Westerman, P. W. Performance of cone-shaped fluidized-bed struvite crystallizers in removing phosphorus from wastewater. Trans. ASAE. 2005, 48, 1227–1234. VOL. 42, NO. 2, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
441
(23) Suzuki, K.; Tanaka, Y.; Kuroda, K.; Hanajima, D.; Fukumoto, Y. Recovery of phosphorus from swine wastewater through crystallization. Bioresour. Technol. 2005, 96, 1544–1550. (24) Seckler, M. M.; Bruinsma, O. S. L.; VanRosmalen, G. M. Calcium phosphate precipitation in a fluidized bed in relation to process conditions: A black box approach. Water Res. 1996, 30, 1677–1685.
442
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 2, 2008
(25) Parent, G.; Belanger, G.; Ziadi, N.; Deland, J.; Laperriere, J. Precipitation of liquid swine manure phosphates using magnesium smelting by-products. J. Environ. Qual. 2007, 36, 557–567.
ES0716709
