XANES Investigation of Phosphate Sorption in Single and Binary

JAMES D. MARTIN ‡. Department of Soil Science, Box 7619, North Carolina State. University, Raleigh, North Carolina 27695-7619, and. Department of ...
0 downloads 0 Views 220KB Size
Download PDF
Environ. Sci. Technol. 2005, 39, 2152-2160

XANES Investigation of Phosphate Sorption in Single and Binary Systems of Iron and Aluminum Oxide Minerals N I D H I K H A R E , †,§ D E A N H E S T E R B E R G , * ,† A N D JAMES D. MARTIN‡ Department of Soil Science, Box 7619, North Carolina State University, Raleigh, North Carolina 27695-7619, and Department of Chemistry, Box 8204, North Carolina State University, Raleigh, North Carolina 27695-8204

Phosphate sorption on Fe- and Al-oxide minerals helps regulate the solubility and mobility of P in the environment. The objective of this study was to characterize phosphate adsorption and precipitation in single and binary systems of Fe- and Al-oxide minerals. Varying concentrations of phosphate were reacted for 42 h in aqueous suspensions containing goethite, ferrihydrite, boehmite, or noncrystalline (non-xl) Al-hydroxide, and in 1:1 (by mass) mixedmineral suspensions of goethite/boehmite and ferrihydrite/ non-xl Al-hydroxide at pH 6 and 22 °C. X-ray absorption near edge structure (XANES) spectroscopy was used to detect precipitated phosphate and distinguish PO4 associated with Fe(III) versus Al(III) in mixed-mineral systems. Changes in the full width at half-maximum height (fwhm) in the white-line peak in P K-XANES spectra provided evidence for precipitation in Al-oxide single-mineral systems, but not in goethite or ferrihydrite systems. Similarly, adsorption isotherms and XANES data showed evidence for precipitation in goethite/boehmite mixtures, suggesting that mineral interactive effects on PO4 sorption were minimal. However, sorption in ferrihydrite/non-xl Al-hydroxide systems and a lack of XANES evidence for precipitation indicated that mineral interactions inhibited precipitation in these binary mixtures.

Introduction Phosphate sorption on Fe- and Al-oxide minerals has been studied for purposes of soil fertility and water quality because these minerals are considered the main P sorbents in acid soils (1-4). Sorption studies suggested that adsorption (surface complexation) or precipitation (e.g., surface precipitation) may occur (5-15). Phosphate surface precipitation may involve formation of a new surface phase, multilayer adsorption of phosphate and metal ions (Fe or Al), or formation of a solid solution (5, 16). Because some of the phosphate is buried under the particle surface and isolated from the exchange solution, occurrence of surface precipitation can inhibit phosphate desorption, resulting in desorption * Corresponding author phone: 919-513-3035; fax: 919-515-2176; e-mail: [email protected]. † Department of Soil Science. ‡ Department of Chemistry. § Current address: Department of Geology and Geophysics, University of Wyoming, Laramie, WY 82071. 2152

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 7, 2005

hysteresis (7). Furthermore, surface precipitation may increase the apparent sorption capacity of a mineral beyond that predicted by the density of surface binding sites (14, 15). Hence, it is important to distinguish adsorption from surface precipitation as the phosphate retention mechanism in Feand Al-oxide systems. Common approaches to studying phosphate surface precipitation on single-mineral systems of Fe- and Al-oxides include macroscopic measurements and molecular scale spectroscopic measurements. As examples of macroscopic studies, changes in zeta potential of goethite at varying pH and phosphate concentrations indicated the formation of an Fe-phosphate surface precipitate (7). Spectroscopic techniques such as nuclear magnetic resonance (NMR) spectroscopy provided more direct (molecular-scale) evidence for surface precipitation of Al-PO4. For example, phosphate sorption in noncrystalline (non-xl) Al-hydroxide minerals investigated using NMR showed the formation of an amorphous octahedral aluminum phosphate (9). In addition, sorption of phosphate on γ-Al2O3 studied using solid-state NMR showed evidence for outer- and inner-sphere complexes and surface precipitation (6). While zeta potential measurements are an indirect technique for determining surface precipitation, solid-state NMR spectroscopy (although direct) entails sample drying, which can alter the structure of surface species. For example, Bleam et al. (12) found evidence for Al-PO4 crystallites on the surface of boehmite, but discounted its significance because of possible artifacts of drying the samples. X-ray absorption fine structure spectroscopy, particularly X-ray absorption near edge structure spectroscopy (XANES), can be done on moist samples and hence is a non-invasive method for characterizing phosphate sorption in-situ. XANES has recently been used to study phosphate sorbed on single minerals and in binary mixtures of ferrihydrite and boehmite (17). The features in XANES spectra for PO4 associated with Fe- or Al-oxide minerals have also been assigned to electronic transitions of the core P electron (1s) to bound states (18, 19). Because XANES spectral features were sensitive to the geometric configuration of phosphate associated with oxide minerals (18, 19), this technique has the potential to distinguish adsorption from precipitation. Although the transition from adsorption to surface precipitation has been characterized in single-mineral systems using NMR and zeta potential measurements, such studies have not been conducted in binary mixtures of Feand Al-oxides. Determination of surface precipitation in binary mixtures independently of characterizing surface precipitation in single-mineral systems is important because it can provide insight into mineral-interaction effects on phosphate sorption. Hence, the objective of this study was to determine the mechanisms of phosphate bonding in single and binary systems of Fe- and Al-oxide minerals as a function of sorbed phosphate concentration using a direct spectroscopic (XANES) approach. Our preliminary experiments (data not shown) indicated that samples with 1:1 mass ratios gave better physical mixing and more consistent results on P distribution between minerals in mixtures than samples prepared to give equal PO4 sorption (17). Single and binary mixtures of goethite/boehmite and ferrihydrite/non-xl Alhydroxide were selected for this purpose. Goethite is ubiquitous in soils and sediments, while ferrihydrite, boehmite, and non-xl Al-hydroxide represent poorly crystalline and non-xl analogues of Fe- and Al-hydr(oxide) minerals in soils. 10.1021/es049237b CCC: $30.25

 2005 American Chemical Society Published on Web 02/22/2005

Methods Mineral Synthesis and Characterization. Two-line ferrihydrite was synthesized by hydrolyzing Fe(III) with KOH, and goethite was synthesized by hydrolyzing FeCl2‚4H2O with 1 M NaHCO3 and oxidizing for 48 h according to methods of Schwertmann and Cornell (20). Ferrihydrite was aged for 6 months and goethite was aged for 18 months before use. Poorly crystalline boehmite was purchased from Reheis Co. (Berkeley Heights, NJ) in gel form under the trade name Rehydragel HPA and was used without aging. Diffuse reflectance FT-IR study by Laiti et al. (16) indicates that boehmite does not undergo any surface transformation with time. Non-xl Al-hydroxide was synthesized by titrating a 0.25 M Al2(SO4)3 solution with 1.5 M KOH solution (21) and aged for 2 years before use. Ferrihydrite, goethite, boehmite, and non-xl Al-hydroxide were characterized using X-ray powder diffraction. The X-ray diffraction pattern for non-xl Alhydroxide showed no peaks, confirming that it was X-ray amorphous. The X-ray diffraction pattern for ferrihydrite showed two broad peaks at 0.24 and 0.15 nm that are characteristic of 2-line ferrihydrite (20). Boehmite and goethite patterns showed no impurities. Boehmite, ferrihydrite, and non-xl Al-hydroxide minerals were stable with respect to PO4 sorption in the time frame of experiments as determined by consistencies in maximum PO4 sorption capacities measured 1 week after synthesis and again 9 months after use. Ferrihydrite, goethite, and non-xl Al-hydroxide were washed thrice with 1 M KCl and further washed with 0.01 M KCl to bring to a 0.01 M KCl background electrolyte. Boehmite was purchased as a gel in deionized water and was also brought into a 0.01 M KCl background by adding 1 M KCl solution. All minerals were stored as stock aqueous suspensions of known (measured) solids concentration in 0.01 M KCl (17) containing 40.2 g of ferrihydrite kg-1, 14.1 g of boehmite kg-1, 30.2 g of goethite kg-1, and 53.7 g of non-xl Al-hydroxide kg-1. The mean crystallite dimensions of poorly crystalline boehmite purchased from Reheis were previously determined to be 4.5, 2.2, and 10 nm along the crystallographic a, b, and c axes, respectively (22). Goethite synthesized by the above procedure is reported to produce acicular 100 nm long needles with a BET surface area of 80 m2 g-1 (20). Surface areas of poorly crystalline minerals measured by conventional BET methods have been shown to underestimate the true reactive surface area due to dehydration structural changes when heated to 110 °C (a precursor treatment for BET surface area measurement) (22). Wang et al. (22) developed a special infrared spectroscopic technique to analyze surface area based on adsorption of a monolayer of water and found that the boehmite from the same source as ours had a surface area of 514 m2/g. We would expect that our non-xl Al-hydroxide would have at least that same surface area, but our measured BET surface area was 39 m2/g. Therefore, we did not measure BET surface areas for ferrihydrite or boehmite as these measurements may be misleading and a poor basis for designing mineral mixtures. Sorption Isotherms. Sorption isotherm experiments for ferrihydrite, goethite, boehmite, non-xl Al-hydroxide, goethite/boehmite, and ferrihydrite/non-xl Al-hydroxide mixtures were conducted at pH 6.0 in 250-mL polycarbonate centrifuge bottles following the procedure described in Khare et al. (17). To avoid excess dissolved Al before PO4 addition and prevent precipitation of Al-phosphate, our goethite and non-xl Alhydroxide suspensions were always kept at pH 5.6, the boehmite suspension was kept at pH 6.0, and the ferrihydrite suspension was kept at pH 7.0. The goethite/boehmite and ferrihydrite/non-xl Al-hydroxide mixtures were prepared by mixing stock suspensions of individual components in 1:1 ratio by mass and equilibrating for 20 min. For all mineral

mixtures reported here, we consistently used 1:1 mass ratios rather than 1:1 sorption capacity ratios. Our approach of comparing the P distribution between minerals in mixtures to a no-preference line based on maximum P sorption capacity of each mineral normalizes the results (17). Samples had a suspended solids concentration of 1.50 g kg-1, constant ionic background of 0.01 M KCl, and total sample mass of 133.34 ( 0.01 g. All aqueous solutions for sorption experiments (0.01 M KCl, 0.01 M HCl, 0.01 M KOH, and 0.01 M KH2PO4) were prepared using analytical grade reagents, and degassed (heated and N2 purged), deionized water. Phosphate was added as 0.01 M KH2PO4 in random chronological order using a micropipet. The average pH after PO4 addition was 6.2 ( 0.7, and acid was added within 5 min after PO4 addition to adjust samples to pH 6.0. During equilibration, the pH rose to 6.3 ( 0.2 and was again adjusted back to pH 6.0. By maintaining pH between 5.6 and 7.0, Al-oxide dissolution should be minimal. All samples were shaken on a reciprocating water bath shaker at 1 s-1 for 42 h at 22 °C. Additional isotherm data for the single and mixedmineral systems were obtained on scaled down samples of 30 g total in 50 mL polycarbonate centrifuge tubes under identical experimental constraints, and following an analogous procedure as outlined above and in Khare et al. (17). Isotherm data from both procedures were integrated. XANES Sample Preparation. For XANES analysis, each sedimented mineral sample from the 250-mL centrifuge bottles used for concurrent isotherm experiments was rinsed into 50-mL polycarbonate centrifuge tubes using a portion of its supernatant solution (to avoid changes in sorbed P), and centrifuged at ∼20 000g for 15 min. A portion (∼0.2 g) of the sedimented mineral samples was mounted in acrylic holders and covered with 5 µm PP X-ray film (Spex Industries, Metuchen, NJ) to avoid desiccation as described in Khare et al. (17). Experiments were timed so that sample preparation was completed a maximum of 5 days in advance of XANES data collection. XANES Data Collection. Phosphorus K-XANES data acquisition was done at Beamline X-19A at the National Synchrotron Light Source, Brookhaven National Laboratory in Upton, NY. The electron beam energy was 2.6 GeV, and the maximum beam current was 300 mA. The synchrotron radiation was monochromatized by a germanium [Ge(111)] monochromator. The monochromator energy was calibrated to 2149 eV at the maximum peak in the first-derivative spectrum of a variscite sample diluted to 400 mmol kg-1 in boron nitride (17, 22). Spectra were collected in fluorescence mode using a PIPS (Passivated Implanted Planar Silicon) detector mounted into a He-filled sample chamber. XANES spectra were taken at photon energies between 2079 and 2248 eV with a minimum step size of 0.2 eV between 2099 and 2174 eV. Self-absorption effects were not considered to be important in the concentration range of our samples as discussed in Khare et al. (17). Samples were analyzed in random chronological order to avoid any systematic trends in spectra due to beamline optics or decaying ring current between fills. XANES Data Normalization. The photon energy scale for all samples and standards was normalized to a relative energy scale by subtracting 2149 eV (23). All individual scans were baseline corrected using a linear regression between -40 and -10 eV relative energy (24). To quantitatively analyze the pre-edge region of the spectra, the baseline for each individual scan was further refined by adjusting all spectra to a common fluorescence yield value at -7 eV. After baseline correction and refinement, multiple scans for a given sample were ensemble averaged. To remove P concentration and detector response effects on the edge step, a single-point background normalization (24) was done using the fluorescence yield at the energy of the maximum peak between 10 VOL. 39, NO. 7, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2153

FIGURE 1. Sorption isotherms and Freundlich models for phosphate in single- and 1:1 (by mass) mixed-mineral systems at pH 6.0 ( 0.1. (a) Goethite/boehmite systems; (b) ferrihydrite/non-xl Al-hydroxide systems. Dashed lines represent isotherm models for mixed systems derived as a weighted combination of Freundlich models from single-mineral systems [0.5(qi + qj)]. Model parameters qi and ci represent fitted sorbed and dissolved PO4 concentrations, respectively, for boehmite (B), goethite (G), non-xl Al-hydroxide (A), ferrihydrite (F), and mineral mixtures (M). Ferrihydrite and boehmite isotherms are from Khare et al. (17). and 18 eV in the first derivative XANES spectrum (17). The fluorescence yields over the entire spectrum were divided by the fluorescence yield at the given normalization energy. Phosphate adsorption was distinguished from precipitation by trends in the white-line peak intensity (peak broadening) in XANES spectra for PO4 sorbed on Fe- and Al-oxide minerals as a function of sorbed PO4 concentration. White-line peak broadening was quantified by measuring the full width at half-maximum height (fwhm) of the peak 2154

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 7, 2005

defined as the width at half of its maximum height relative to the computed spectral baseline. In a two-dimensional adsorbed phase, atomic orbitals of metal, oxygen, and phosphorus overlap to form discrete molecular orbitals resulting in sharper, more intense features (for, e.g., the whiteline peak). For a three-dimensional metal phosphate precipitate, however, atomic orbitals broaden into a band because a typical solid contains ∼1022 atoms/cm3, resulting in a nearly continuous distribution of energies. Therefore,

FIGURE 2. Normalized, stacked phosphorus K-XANES spectra for phosphate sorbed on goethite, boehmite [Khare et al. (17)], or mixedmineral systems at pH 6.0 ( 0.1 at selected concentrations. Numbers in the legend denote sorbed PO4 in mmol kg-1. A pre-edge feature between -5 and -1 eV relative energy is apparent in the goethite and mixed-mineral spectra. for P K-XANES data normalized to a per-atom basis (24), broader (and less intense) features (for, e.g., the white-line peak) are predicted for a precipitate phase. The energy distribution of the white-line peak computed using an extended Hu ¨ckel calculation for a two-dimensional adsorbed phase modeled as a monodentate complex (Al-O-PO3H2) and a three-dimensional precipitate modeled as variscite further illustrated this broadening and is included in the Supporting Information. Extended-Hu ¨ ckel computed energies project qualitative trends in XANES spectral features based on the relative distributions of the various antibonding molecular orbitals (19, 28). The white-line peak in XANES spectra for PO4 associated with Fe(III) or Al(III) was generally assigned to an electronic transition from P 1s core shell into a P(3p)-O(2p) antibonding molecular orbital by Franke and Hormes (18). Khare et al. (19) provided a more specific whiteline peak assignment for PO4 associated with Al(III) (an electronic transition of a P 1s core shell into a Al(3p)-O(2p)P(3p) antibonding molecular orbital), which was used to calculate the energy distribution of the white-line peak. The onset of PO4 surface precipitation in mineral-mixtures was investigated by comparing trends in fwhm of white-line peaks in XANES spectra for mixtures with trends in their respective single-mineral components. Relative distribution of PO4 in mixed-mineral systems was determined by fitting XANES spectra for mixtures in the pre-edge region (-6 to -1 eV) with single-mineral standards of PO4 on the respective Fe-oxide or Al-oxide according to the procedure described in Khare et al. (17).

Results and Discussion Phosphate Sorption in Fe- and Al-Oxide Mixtures. Sorption isotherms for goethite, boehmite, ferrihydrite, non-xl Alhydroxide, and 1:1 mixtures of goethite/boehmite and ferrihydrite/non-xl Al-hydroxide (Figure 1a and b) were L-curves that could be fit with Freundlich models (25). The maximum levels of sorption observed were 580 (goethite), 850 (boehmite), 1860 (ferrihydrite), 3400 (non-xl Al-hydroxide), 710 (goethite/boehmite), and 2500 (ferrihydrite/non-xl Al-hydroxide) mmol kg-1. The isotherms essentially reached a plateau (asymptote) for only goethite over the concentration ranges studied. As compared to the other systems, the isotherm for PO4 on non-xl Al-hydroxide was more steeply sloping at the maximum observed level of sorbed PO4. Similar to ferrihydrite/boehmite mixtures (17), the phosphate sorption isotherms for goethite/boehmite and ferrihydrite/nonxl Al-hydroxide mixtures were intermediate between those of the respective single-mineral systems (Figure 1a and b). However, the isotherm model derived as a linear combination of Freundlich models for ferrihydrite and non-xl Al-hydroxide systems [see Khare et al. (17)] predicted up to 16% more sorption for the mixture of these minerals than was observed (Figure 1b). For the goethite/boehmite mixtures, the combination model varied by FeO-Al-PO4). Adsorption versus Precipitation in Single-Mineral Systems. Figures 6 and 7 show fwhm of white-line peaks of XANES spectra for various mineral systems as a function of sorbed phosphate. Because phosphate minerals exhibit greater fwhm than adsorbed phases, increasing fwhm with increasing sorbed phosphate in these figures indicates precipitation of Al-phosphate. Data for boehmite, goethiteboehmite mixtures, and non-xl Al-hydroxide all gave evidence for Al-phosphate precipitation, possibly surface precipitation. In agreement with our results for non-xl Al-hydroxide systems, Lookman et al. (26) found evidence from NMR studies that Al-phosphate precipitated at greater adsorbed PO4 concentrations. Results from a diffuse reflectance IR study by Nanzyo (27) also showed that bulk Al-phosphate formed when non-xl Al-hydroxide was reacted with phosphate. In contrast to Al-oxide systems, the white-line peak in XANES spectra for phosphate sorbed on Fe-oxides usually did not broaden or reduce in intensity (Figures 2, 3, 6, and VOL. 39, NO. 7, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2157

FIGURE 6. The full width at half-maximum height (fwhm) of the white-line peak in XANES spectra for phosphate sorbed on goethite, boehmite, and goethite/boehmite mixtures as a function of sorbed PO4 concentration.

FIGURE 5. Concentration of PO4 sorbed on the Fe-oxide component of Fe- and Al-oxide mixtures (as calculated from XANES fitting analysis) plotted as a function of total sorbed phosphate concentration for mixtures including regression fits: (a) goethite/boehmite mixtures; (b) ferrihydrite/non-xl Al-hydroxide mixtures. The apparent no-preference line was determined on the basis of the maximum observed sorption capacities of the single minerals in Figure 1, assuming no precipitation [see Khare et al. (17)].

TABLE 2. Phosphate Associated with Fe and Al in 1:1 (Mass Basis) Mixtures of Ferrihydrite and Non-xl Al-Hydroxide Calculated Using Linear Combination Fitting of Normalized Phosphorus K-XANES Spectraa total level of sorbed PO4 (mmol kg-1)

% associated with Fe

% associated with Al

χ2

100 200 300 400 600 1230

59 ( 6 22 ( 2 30 ( 1 18 ( 1 14 ( 1 20 ( 0

41 ( 14 78 ( 3 70 ( 2 82 ( 3 86 ( 2 80 ( 1

0.0067 0.0004 0.0003 0.0004 0.0002 0.0001

a Standard errors shown were calculated by the fitting program (Kaliedagraph, Synergy Software Co., Reading, PA).

7). No trend between fwhm and sorbed PO4 was found for goethite or ferrihydrite as indicated by low r2 (0.004 and 0.062) values for linear regressions (Figures 6 and 7). Also, XANES pre-edge features for these systems were less intense than the pre-edge feature observed in XANES spectra for strengite (17), and the pre-edge intensity showed no trend with sorbed phosphate concentration (Figures 2 and 3). If the fwhm of the white-line peak in XANES spectra and the intensity of the pre-edge feature are indicative of a three-dimensional cluster, then our data show no evidence for clustering or precipitation in the ferrihydrite and goethite systems. Our 2158

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 7, 2005

FIGURE 7. The full width at half-maximum height (fwhm) of the white-line peak in XANES spectra for phosphate sorbed on ferrihydrite, noncrystalline Al-hydroxide, and ferrihydrite/non-xl Alhydroxide mixtures as a function of sorbed PO4 concentration. ferrihydrite results are consistent with those of Willett et al. (12), who found no evidence for Fe-phosphate precipitation in ferrihydrite systems based on a combination of infrared spectrometry and electron microprobe analyses. Adsorption versus Precipitation in Mineral Mixtures. The fwhm of the white-line peak in XANES spectra for PO4 sorbed in goethite/boehmite mixtures increased with increasing total sorbed PO4 concentration (Figure 6). These

mixtures showed trends in fwhm and white-line peak shift similar to the trends in single-mineral systems of boehmite but unlike in goethite. Thus, XANES results indicated that phosphate sorption in the 1:1 goethite/boehmite mixtures behaved essentially like a linear combination of sorption to each component. Also, the predicted sorption for the 1:1 goethite/boehmite mixtures based on a linear combination of Freundlich models for goethite and boehmite deviated by 6 µmol L-1 from the Freundlich model fit to isotherm data (Figure 1b). This trend makes sense if precipitation of Al-phosphate occurred in the single-mineral systems of non-xl Al-hydroxide (causing a greater level of PO4 sorption beyond just adsorption), but not in the ferrihydrite/non-xl Al-hydroxide mixtures. Hence, our results suggested that sorption in ferrihydrite/non-xl Alhydroxide mixtures was not a linear combination of sorption to their respective single-mineral components, consistent with XANES findings. In summary, our results indicated that (1) phosphate was bound via an adsorption mechanism as an inner-sphere surface complex on either goethite or ferrihydrite; (2) the trends in fwhm of the white-line peak of XANES spectra for phosphate sorbed on boehmite or non-xl Al-hydroxide showed evidence for precipitation; and (3) phosphate sorption in goethite/boehmite mixtures behaved as a linear combination of the individual minerals; in contrast, mineral interactions affected PO4 sorption in ferrihydrite/non-xl Alhydroxide mixtures.

Acknowledgments We thank Ms. Kimberly Hutchison for assistance with lab work, Dr. Wolfgang Caliebe and Dr. Suzanne Beauchemin for assistance with XANES data collection, and Dr. Dale E. Sayers for suggestions on XANES data normalization. Funding was provided by USDA-NRI Grant No. 2001-35107-10179 and by the Agricultural Foundation of the College of Agriculture and Life Science (CALS) at NC State. This research was carried out (in part) at the National Synchrotron Light Source (NSLS), Brookhaven National Laboratory, which is supported by the U.S. Department of Energy, Division of Materials Sciences and Division of Chemical Sciences.

Supporting Information Available Projected density of states (PDOS) and corresponding antibonding molecular orbitals (MO) for monodentate cluster (Al-O-PO3H2) and variscite. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Pierzynski, G. M.; Logan, T. J.; Traina, S. J.; Bigham, J. M. Phosphorus chemistry and mineralogy in excessively fertilized soils: quantitative analysis of phosphorus-rich particles. Soil Sci. Soc. Am. J. 1990a, 54, 1576-1583. (2) Pierzynski, G. M.; Logan, T. J.; Traina, S. J.; Bigham, J. M. Phosphorus chemistry and mineralogy in excessively fertilized soils: descriptions of phosphorus-rich particles. Soil Sci. Soc. Am. J. 1990b, 54, 1583-1589. (3) Sallade, Y. E.; Sims, J. T. Phosphorus transformations in the sediments of Delaware’s agricultural drainageways: II. Effect of reducing conditions on phosphorus release. J. Environ. Qual. 1997, 26, 1579-1588. (4) Beauchemin, S.; Simard, R. R. Soil phosphorus saturation degree: review of some indices and their suitability for P management in Quebec, Canada. Can. J. Soil Sci. 1999, 79, 615625. (5) Ler, A.; Stanforth, R. Evidence for surface precipitation of phosphate on goethite. Environ. Sci. Technol. 2003, 37, 26942700. (6) Johnson, B. B.; Ivanov, A. V.; Antzutkin, O. N.; Forsling, W. 31P Nuclear magnetic resonance study of the adsorption of phosphate and phenyl phosphates on γ-Al2O3. Langmuir 2002, 18, 1104-1111. (7) Li, L.; Stanforth, R. Distinguishing adsorption and surface precipitation of phosphate on goethite (R-FeOOH). J. Colloid Interface Sci. 2000, 230, 12-21. (8) Nooney, M. G.; Campbell, A.; Murrell, T. S.; Lin, X. F.; Hossner, L. R.; Chusuei, C. C.; Goodman, D. W. Nucleation and growth of phosphate on metal oxide thin films. Langmuir 1998, 14, 2750-2755. (9) Lookman, R.; Gorbet, P.; Merckx, R.; Van Riemsdijk, W. H. Application of 31P and 27Al MAS NMR for phosphate speciation studies in soil and aluminum hydroxide: promises and constraints. Geoderma 1997, 80, 369-388. (10) Bleam, W.; Pfeffer, P.; Goldberg, S.; Taylor, R. W.; Dudley, R. A 31P solid-state nuclear magnetic resonance study of phosphate adsorption at the boehmite/aqueous solution interface. Langmuir 1991, 7, 1702-1712. (11) Martin, R. R.; Smart, R. ST. S.; Tazaki, K. Direct observation of phosphate precipitation in the goethite/phosphate system. Soil Sci. Soc. Am. J. 1988, 52, 1492-1500. (12) Willett, I. R.; Chartres, C. J.; Nguyen, T. T. Migration of phosphate into aggregated particles of ferrihydrite. J. Soil Sci. 1988, 39, 275-282. (13) Bolan, N. S.; Barrow, N. J.; Posner, A. M. Describing the effect of time on sorption of phosphate by iron and aluminum hydroxides. J. Soil Sci. 1985, 36, 187-197. (14) Van Riemsdijk, W. H.; Lyklema, J. The reaction of phosphate with aluminium hydroxide in relation with phosphate bonding in soils. Colloids Surf. 1979, 1, 33-44. (15) Laiti, E.; Persson, P.; Ohman, L.-O. Surface complexation and precipitation at the H+ orthophosphate aged γ-Al2O3 interface. Langmuir 1996, 12, 2969-2975. (16) Laiti, E.; Persson, P.; Ohman, L.-O. Balance between surface complexation and surface phase transformation at the H+ alumina/water interface. Langmuir 1998, 14, 825-831. (17) Khare, N.; Hesterberg, D.; Beauchemin, S.; Wang, S. L. XANES determination of adsorbed phosphate distribution between ferrihydrite and boehmite in mixtures. Soil Sci. Soc. Am. J. 2004, 68, 460-469. (18) Franke, R.; Hormes, J. The P K-near edge absorption spectra of phosphates. Physica B 1995, 216, 85-95. (19) Khare, N.; Martin, J. D.; Hesterberg, D. L. Mode of phosphate bonding on Fe(III) systems in relation to electronic origin of Phosphorus K-XANES features. Geochim. Cosmochim. Acta, in review. (20) Schwertmann, U.; Cornell, R. M. Iron oxides in the laboratory: preparation and characterization; VCH Publ. Co.: Weinheim, Germany, 1991. (21) Prodromou, K. P.; Pavlatou-Ve, A. S. Formation of aluminum hydroxides as influenced by aluminum salts and bases. Clays Clay Miner. 1995, 43, 111-115. VOL. 39, NO. 7, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2159

(22) Wang, S. L.; Johnston, C. T.; Bish, D. L.; White, J. L.; Hem, S. L. Water-vapor adsorption and surface area measurement of poorly crystalline boehmite. J. Colloid Interface Sci. 2003, 260, 26-35. (23) Hesterberg, D. L.; Zhou, W.; Hutchison, K. J.; Beauchemin, S.; Sayers, D. E. XAFS study of adsorbed and mineral forms of phosphate. J. Synchrotron Radiat. 1999, 6, 636-638. (24) Sayers, D. E.; Bunker, B. A. In X-ray absorption: principles, applications, techniques of EXAFS, SEXAFS, and XANES; Koningsberger, D. C., Prins, R., Eds.; John Wiley & Sons: New York, 1988; pp 211-253. (25) Sposito, G. The surface chemistry of soils; Oxford University Press: New York, 1984.

2160

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 7, 2005

(26) Lookman, R.; Gorbet, P.; Merckx, R.; Vlassak, K. Phosphate sorption by synthetic amorphous aluminum hydroxides: a 27Al and 31P solid-state MAS NMR spectroscopic study. Eur. J. Soil Sci. 1994, 45, 37-44. (27) Nanzyo, M. Diffuse reflectance infrared spectra of phosphate sorbed on alumina gel. J. Soil Sci. 1984, 35, 63-69. (28) Whangbo, M. H. Perspective on “An extended Huckel theory. I. Hydrocarbons” (Hoffmann, R. 1963). J. Chem. Phys. 2000, 39, 1397-1412.

Received for review May 23, 2004. Revised manuscript received December 17, 2004. Accepted December 22, 2004. ES049237B