3680
Langmuir 2007, 23, 3680-3689
Geometry, Charge Distribution, and Surface Speciation of Phosphate on Goethite Rasoul Rahnemaie,† Tjisse Hiemstra,* and Willem H. van Riemsdijk Department of Soil Quality, Wageningen UniVersity, P.O. Box 8005, 6700 EC Wageningen, The Netherlands ReceiVed October 9, 2006. In Final Form: December 21, 2006 The surface speciation of phosphate has been evaluated with surface complexation modeling using an interfacial charge distribution (CD) approach based on ion adsorption and ordering of interfacial water. In the CD model, the charge of adsorbed ions is distributed over two electrostatic potentials in the double-layer profile. The CD is related to the structure of the surface complex. A new approach is followed in which the CD values of the various surface complexes have been calculated theoretically from the geometries of the surface complexes. Molecular orbital calculations based on density functional theory (MO/DFT) have been used to optimize the structure of a series of hydrated surface complexes of phosphate. These theoretical CD values are corrected for dipole orientation effects. Data analysis of the PO4 adsorption, applying the independently derived CD coefficients, resolves the presence of two dominant surface species. A nonprotonated bidentate (B) complex is dominant over a broad range of pH values at low loading (e1.5 µmol/m2). For low pH and high loading, a strong contribution of a singly protonated monodentate (MH or MH-Na) complex is found, which differs from earlier interpretations. For the conditions studied, the doubly protonated bidentate (BH2) and monodentate (MH2) surface complexes and the nonprotonated monodentate (M) complex are not significant contributors. These findings are discussed qualitatively and quantitatively in relation to published experimental in-situ CIR-FTIR data and theoretical MO/DFT-IR information. The relative variation in the peak intensities as a function of pH and loading approximately agrees with the surface speciation calculated with the CD model. The model correctly predicts the proton co-adsorption of phosphate binding on goethite and the shift of the IEP at low phosphate loading (e1.5 µmol/m2). At higher loading, it deviates.
1. Introduction Ion adsorption experiments, in-situ spectroscopy, and mechanistic modeling are valuable tools that provide information that may improve our understanding of the surface chemistry of colloids. Model (hydr)oxides such as goethite (R-FeOOH) can be used as a proxy for such an important challenge. The interaction of cations and anions at surfaces can be described with surface complexation models (SCM). The treatment of the electrostatic interactions is a key factor in SCM. Ion adsorption phenomena are particularly sensitive to the location of the ionic charge in the electrostatic-double-layer profile. The electrostatic gradients are very large near the surface. Therefore, Hiemstra and van Riemsdijk1 proposed a charge distribution (CD) model in which the charge of an ion is allowed to be distributed over different interface positions. Application of the CD model to a series of oxyanions with a binding modus, known from spectroscopy, has demonstrated that the adsorption behavior is related to the relative number of oxygen ligands common to the surface2 (i.e., the structure of the surface complex is involved). For inner-sphere surface complexes, usually one or two ligands are common to the metal ion(s) of the solid. The corresponding bond valence charge is electrostatically located on the surface in contrast to the charge of the other ligands that interact with the solution. These ligands have less interaction with the proton charge at the surface. These different degrees of interaction of ligands illustrate the relation between the location of charge and the structure of the surface complex. From this perspective, * Corresponding author. E-mail:
[email protected]. † Present address: Department of Soil Science, Tarbiat Modares University, P.O. Box 14115-336, Tehran, Iran. (1) Hiemstra, T.; Van Riemsdijk, W. H. J. Colloid Interface Sci. 1996, 179, 488-508. (2) Rietra, R. P. J. J.; Hiemstra, T.; Van Riemsdijk, W. H. Geochim. Cosmochim. Acta 1999, 63, 3009-3015.
structural information can be used to constrain the modeling of adsorption data. Phosphorus is omnipresent in the natural environment. Phosphate may interact with iron oxide surfaces via Fe-O-P bonds. Such bonds can also be formed upon the interaction of natural organic phosphorus compounds such as DNA with metal oxide surfaces and is relevant to processes such as bacterial adhesion.3 The surface complexation of phosphate on goethite as function of pH and loading has been studied with in-situ IR spectroscopy.4 A nonprotonated bidentate complex (tFe2O2PO2) is considered to be the main surface species in the neutral pH range. A similar species has been suggested for TiO2.5 Spectral changes at low pH point to the presence of an additional species. The dominance of this species is enhanced by loading. The original band assignments with respect to the presence of a singly protonated bidentate complex (tFe2O2POOH) have been questioned.6,7 Recently, Kwon and Kubicki8 have used molecular orbital (MO) calculations to assess the band assignments. Calculated frequencies were correlated with a preselected set of experimental frequencies. As a matter of fact, the large in-situ data set of Tejedor-Tejedor and Anderson4 was not considered. The formation of a doubly protonated bidentate complex was proposed (tFe2O2P(OH)2), but the presence of a singly protonated monodentate complex (tFeOPO2OH) could not be excluded. In the present contribution, we want to evaluate the surface speciation from the perspective of surface complexation modeling and test it quantitatively against the surface speciation proposed by IR spectroscopy studies. (3) Parikh, S. J.; Chorover, J. Langmuir 2006, 22, 8492-8500. (4) Tejedor-Tejedor, M. I.; Anderson, M. A. Langmuir 1990, 6, 602-611. (5) Connor, P. A.; McQuillan, A. J. Langmuir 1999, 15, 2916-2921. (6) Persson, P.; Nilsson, N.; Sjo¨berg, S. J. Colloid Interface Sci. 1995, 177, 263-275. (7) Arai, Y.; Sparks, D. L. J. Colloid Interface Sci. 2001, 241, 317-326. (8) Kwon, K. D.; Kubicki, J. D. Langmuir 2004, 20, 9249-9254.
10.1021/la062965n CCC: $37.00 © 2007 American Chemical Society Published on Web 02/23/2007
Surface Speciation of PO4 on Goethite
In this article, we follow a new approach in surface complexation modeling. It may be expected that the charge distribution in a surface complex is related to the relative bond length in the surface complex9,10 (i.e., the geometry is involved). In this article, the geometry of a large number of hydrated phosphate surface complexes will be optimized with molecular orbital calculations using density functional theory (MO/DFT). The geometry will be used to derive ionic CD values using the Brown bond valence concept.11 These ionic CD values will be corrected for changes in the dipole orientation10 of interfacial water molecules. Such changes in dipole orientation, induced by an electrostatic field, have been observed with sum frequency spectroscopy.12-15 In our modeling approach, we will also use a new doublelayer profile10 that accounts for the presence of ordering of water molecules near the surface. Recent X-ray reflectivity16-18 has shown the presence of one to three layers of aligned water molecules in contact with the surface. The layers may prevent the ordering of electrolyte ions near the surface in a diffuse pattern as is expected for ions in the diffuse double layer (DDL).19 In contrast, electrolyte ion pairs may penetrate into such layers and reach the minimum distance of approach. This may result in charge separation between the head end of the DDL and the location of ion pairs, as defined in the extended stern (ES) layer model10 that is applied here. In the present contribution, the surface speciation will be evaluated by modeling a new set of accurate adsorption data that cover large ranges of pH values, equilibrium concentrations, and ionic strengths. The results will be critically compared with the spectroscopic information from in-situ FTIR experiments4 and theory.8 The modeling results will also be tested on additional macroscopic information such as the proton co-adsorption and the change in the isoelectric point (IEP) as a function of phosphate loading.4,20 2. Materials and Methods 2.1. Preparation and Characterization of Goethite. The goethite suspension was prepared on the basis of the method of Atkinson et al.,21 as described in detail by Hiemstra et al.22 Freshly prepared 0.5 M Fe(NO3)3 was slowly titrated with 2.5 M NaOH to pH 12. The suspension was aged for 4 days at 60 °C and subsequently dialyzed in ultrapure water. Before using it in the experiments, the goethite suspension was acidified (pH ∼5) to desorb and remove the (bi)carbonate by continuously purging it with N2 for at least 1 day. The BET(N2) specific surface area of goethite was determined to be 85 m2 g-1. (9) Hiemstra, T.; Van Riemsdijk, W. H. On the Relationship between Surface Structure and Ion Complexation of Oxide-Solution Interfaces. In Encyclopedia of Surface and Colloid Science; Hubbard, A. T., Ed.; Marcel Dekker: New York, 2002; pp 3773-3799. (10) Hiemstra, T.; Van Riemsdijk, W. H. J. Colloid Interface Sci. 2006, 301, 1-18. (11) Brown, I. D.; Altermatt, D. Acta Crystallogr. 1985, B41, 244-247. (12) Yeganeh, M. S.; Dougal, S. M.; Pink, H. S. Phys. ReV. Lett. 1999, 83, 1179-1182. (13) Kataoka, S.; Gurau, M. C.; Albertorio, F.; Holden, M. A.; Lim, S. M.; Yang, R. D.; Cremer, P. S. Langmuir 2004, 20, 1662-1666. (14) Ostroverkhov, V.; Waychunas, G. A.; Shen, Y. R. Phys. ReV. Lett. 2005, 94, 046102. (15) Shen, Y. R.; Ostroverkhov, V. Chem. ReV. 2006, 106, 1140-1154. (16) Toney, M. F.; Howard, J. N.; Richer, J.; Borges, G. L.; Gordon, J. G.; Melroy, O. R.; Wiesler, D. G.; Yee, D.; Sorensen, L. B. Surf. Sci. 1995, 335, 326-332. (17) Fenter, P.; Sturchio, N. C. Prog. Surf. Sci. 2004, 77, 171-258. (18) Catalano, J. G.; Park, C.; Zhang, Z.; Fenter, P. Langmuir 2006, 22, 46684673. (19) Pashley, R. M.; Israelachvili, J. N. J. Colloid Interface Sci. 1984, 101, 511-523. (20) Antelo, J.; Avena, M.; Fiol, S.; Lopez, R.; Arce, F. J. Colloid Interface Sci. 2005, 285, 476-486. (21) Atkinson, R. J.; Posner, A. M.; Quirk, J. P. J. Phys. Chem. 1967, 71, 550-558. (22) Hiemstra, T.; De Wit, J. C. M.; Van Riemsdijk, W. H. J. Colloid Interface Sci. 1989, 133, 105-117.
Langmuir, Vol. 23, No. 7, 2007 3681 Table 1. Phosphate Speciation Reactions and Their Aqueous Equilibrium Constants in the NaNO3 Medium (I ) 0) species
reaction
log K
HPO4-2 H2PO4H3PO4 NaHPO4NaPO4-2 H2O NaNO3
PO4-3 + H+ S HPO4-2 PO4-3 + 2H+ S H2PO4PO4-3 + 3H+ S H3PO4 PO4-3 + Na+ + H+ S NaHPO4PO4-3 + Na+ S NaPO4-2 H+ + OH- S H2O Na+ + NO3- S NaNO3
12.35a 19.55a 21.70a 13.40b,c 2.05c 14.00a -0.60d
a
Lindsay.24
b
Turner et al.25
c
This study. d Smith et al.27
The surface charge of goethite was measured in NaNO3 solutions. A sample of goethite was titrated forward and backward by adding base and acid within the pH range of ∼4 to 10.5. The temperature was fixed at 20 ( 0.1 °C using a thermostated reaction cell. The details of the experimental setup have been given elsewhere.23 2.2. Preparation of Reagents. To avoid (bi)carbonate contamination, all solutions (Merck p.a.) were made under a purified N2 atmosphere and stored for a short time in polyethylene bottles to keep them free of silica. The acid solutions were stored in glass bottles because it was found that the acid solutions become polluted by releasing organic materials in polyethylene bottles. Ultrapure water (∼0.018 dS/m) was used throughout the experiments. It had been pre-boiled to remove dissolved CO2 before using it in the experiments. The experiments were carried out in a constanttemperature room (22 ( 1 °C). NaOH solution was prepared CO2-free from highly concentrated 1:1 NaOH/H2O. The mixture was centrifuged to remove any solid Na2CO3. A subsample of supernatant was pipetted into ultrapure water and stored in a desiccator equipped with a CO2-absorbing column. 2.3. Phosphate Speciation in Solution. The solution speciation of phosphate was measured by the titration of solutions with 0, 0.1, 1, and 10 mM NaH2PO4 in 0.05, 0.1, and 0.5 M NaNO3 and using 0.1 M NaH2PO4 in 0.5 M NaNO3. The solutions were prepared under a N2 atmosphere. The experiments were carried out in a 100 mL thermostated reaction vessel at 20 ( 0.1 °C under a N2 atmosphere. The titration was carried out in forward and backward pH cycles at a fixed salt concentration. Before the titration, the solution was kept at pH ∼3 for about 1 h while purging it with N2 to remove CO2. In the interpretation of the data (Table 1), the affinities of protonated phosphate species (HxPO4-3+x) were fixed to values given by Lindsay.24 For NaHPO4-1, our fitted affinity constant was equal to the log K value reported by Turner et al.25 However, the fitted log K value for NaPO4-2 (logK ) 2.05) was higher than that previously reported (log K ) 1.6) by Millero and Schreiber.26 The activity corrections were made using the Davies equation with C ) 0.2. 2.4. Phosphate Adsorption Edges. Adsorption experiments were carried out in individual gastight 23.6 mL low-density polyethylene bottles with fixed amounts of salt, goethite, and phosphate. A certain amount of HNO3 or NaOH was added to the vessels in order to obtain final pH values within the relevant pH range. A certain amount of NaNO3 solution was used to fix the salt concentration of the suspensions. The final volume of the suspensions was 20 mL. The order of addition of solutions may have an important effect on the adsorption process. To avoid the irreversibility of phosphate adsorption due to changes in the ionic strength and pH of the suspension, phosphate was always added to the suspension as the last solution. The bottles were equilibrated for 24 h in an endover-end shaker. After centrifugation, a sample of the supernatant was taken for phosphate analysis. The phosphate concentration was (23) Rahnemaie, R.; Hiemstra, T.; Van Riemsdijk, W. H. J. Colloid Interface Sci. 2006, 293, 312-321. (24) Lindsay, W. L. Chemical Equilibria in Soils; Wiley-Interscience: New York, 1979; p 347. (25) Turner, D. R.; Whitfiled, M.; Dickson, A. G. Geochim. Cosmochim. Acta 1981, 45, 855-881. (26) Millero, F. J.; Schreiber, D. R. Am. J. Sci. 1982, 282, 1508-1540.
3682 Langmuir, Vol. 23, No. 7, 2007 determined using a molybdenum blue method adapted to measure relatively low P concentrations. The pH of the suspensions was measured in the remixed suspensions under a N2 atmosphere. For each data point, the total concentrations of components of the system were calculated on the basis of accounting for the volumes and concentrations of the added solutions. The amount of adsorbed phosphate was calculated from the difference between the total initial phosphate concentration and the final equilibrium concentration. 2.5. Proton Co-Adsorption. Proton co-adsorption upon the addition of NaH2PO4 was measured using the method described by Rietra et al.2 Accurate pH-stat titrations were carried out in a thermostated Metrohm vessel (20.0 ( 0.1 °C) present in a constanttemperature room. The vessel was continuously purged with moist, purified N2 gas. A double-junction reference electrode was used. The outer junction was filled with a 0.125 M NaNO3 and 0.875 M KNO3 solution to minimize the diffusion potential. A certain volume (18 mL) of a CO2-free goethite stock suspension (22.63 g/L) was diluted by pipetting ultrapure water and electrolyte solution to a final volume of 40 mL. The suspension was acidified to pH ∼3.5 and purged with moist purified N2 gas for 2 h to remove adsorbed CO2. The experiments were carried out at constant pH values of 3.98 and 4.80 in 0.1 M NaNO3. The suspensions were titrated in steps (0.40 mL) with a solution of 0.01 mM NaH2PO4. In the experiments, the pH was kept constant at the chosen values by the addition of base from a commercial 0.100 M NaOH solution. 2.6. MO/DTF Computations. The geometry of surface complexes of phosphate was optimized by molecular orbital (MO) calculations using Spartan ’04 software.28 The geometry optimizations were made using density functional theory (DFT). Pseudopotentials, defined in Spartan ’04 as LACVP+** (Los Alamos core valence potentials), were used. This set comprises the 6-31+G** basis set for maingroup elements H-Ar. For nonhydrated structures, the final geometry was calculated with different models (unrestricted BP86, B3LYP, BLYP, EDF1, and a local (SVWN) model29). Hydrated structures were optimized only with the Becke Perdew BP86 model. The latter calculations are very time-consuming. The calculated geometry has been interpreted with the Brown bond valence approach11 in order to obtain the charge distribution value of PO4 surface complexes.
3. Results and Discussion 3.1. Double-Layer Picture. Rahnemaie et al.23 have studied the charging behavior of goethite for a wide range of electrolytes. The positions of the individual types of electrolyte ions (Li+, Na+, K+, Cs+, Cl-, and NO3- and ClO4-) in the double-layer profile were traced with a CD approach for outer-sphere complexation in a free fit of the data. A critical re-evaluation of the data10 shows that most electrolyte ion pairs locate their charge approximately at the same position in the compact part of the double layer, which allows simplification to a double-layer picture with charge location of the ion pairs in one electrostatic plane. Importantly, for a simultaneous description of the whole data set, it is essential that the location of the ion pairs does not coincide with the position of the head end of the DDL. An additional charge-free layer is present at the goethite-water interface. This result points to the ordering of interfacial water, allowing only stepwise changes in the location of electrolyte ions near the surface.10 This picture not only agrees with force measurements of overlapping double layers19,30 but also agrees (27) Smith, R. M.; Martell, A. E. Critical Stability Constants; Plenum: New York, 1981; Vol. 4. (28) Spartan ’04; Wavefunction, Inc.: Irvine CA, 2004. (29) Kong, J.; White, C. A.; Krylov, A. I.; Sherrill, C. D.; Adamson, R. D.; Furlani, T. R.; Lee, M. S.; Lee, A. M.; Gwaltney, S. R.; Adams, T. R.; Ochsenfeld, C.; Gilbert, A. T. B.; Kedziora, G. S.; Rassolov, V. A.; Maurice, D. R.; Nair, N.; Shao, Y.; Besley, N. A.; Maslen, P. E.; Dombroski, J. P.; Daschel, H.; Zhang, W.; Korambath, P. P.; Baker, J.; Byrd, E. F. C.; Van Voorhis, T.; Oumi, M.; Hirata, S.; Hsu, C.-P.; Ishikawa, N.; Florian, J.; Warshel, A.; Johnson, B. G.; Gill, P. M. W.; Head-Gordon, M.; Pople, J. A. J. Comput. Chem. 2000, 21, 15321548. (30) Israelachvili, J. N.; Wennerstrom, H. Nature 1996, 379, 219-225.
Rahnemaie et al. Table 2. Charge Allocation (∆z) and Affinity Constants of Interaction (log K) of Monovalent Cations and Anions Interacting with Oppositely Charged Surface Groups of Goethite as Derived from Modeling of the Goethite Titration Data with the Extended Stern Layer Modela ions
∆z0
∆z1
∆z2
log K
H+ Li+ Na+ K+ Cs+ NO3-1 Cl-1
1 0 0 0 0 0 0
0 +1 +1 +1 +1 -1 -1
0 0 0 0 0 0 0
+9.0 +0.11 ( 0.02 -0.61 ( 0.03 -1.74 ( 0.16 b
-0.70 ( 0.03 -0.44 ( 0.03
a All electrolyte ions are placed on the 1-plane. The fitted capacitance for the first and second layer is C1 ≡ C2 ) 0.92 ( 0.01 F/m2. The standard deviation is given for the fitted parameters. R2 ) 0.998, N ) 320 data points. b The affinity of the adsorption of Cs+ ions is negligible.
Figure 1. Logarithm of the equilibrium concentration in a phosphate-goethite system (3 g/L) with variable ionic strength. The Initial PO4 concentration is 0.4 mM. The model lines are calculated using the CD values derived from MO/DFT-optimized geometries after correction for dipole orientation effects (Table 8).
with recent X-ray reflectivity data.16-18 At the surface, about one to three layers of aligned water are present, which prevents electrolyte ions from being present in a diffuse pattern. Some electrolyte ions may penetrate into the ordered water layer(s), forming hydrated ion pairs with surface groups. The resulting double-layer profile with two Stern layers can be classified as an extended Stern (ES) layer model.31 The basic charging behavior of the present goethite can be described with the parameter set of Table 2. As shown in the Table, no charge (∆z2 ) 0) is located in the 2-plane where the head end of the DDL starts. The capacitance of the second Stern layer C2 is uncertain,10 but the value is the same order of magnitude as the value of the inner Stern layer (C1). As a simplification, we may use arbitrary C1tC2. This was also recently done by Sverjensky.32 3.2. Phosphate Adsorption Data. The pH-dependent phosphate adsorption behavior has been studied as a function of the ionic strength (Figure 1) and phosphate loading (Figure 2). In Figure 1, the variation in the equilibrium concentration (log scale) in a phosphate-goethite system with variable pH and ionic strength is shown. In this system, the initial P concentration of 0.4 mmol L-1 is equivalent to 1.57 µmol m-2. The experimental data show a common intersection point that represents the isoelectric point (IEP) of the system. Because of the introduction of the negative charge of adsorbed PO4-3, the IEP of the particles has shifted from a pristine value (PZC ∼9) to a value of IEP ∼5. (31) Westall, J.; Hohl, H. AdV. Colloid Interface Sci. 1980, 12, 265-294. (32) Sverjensky, D. A. Geochim. Cosmochim. Acta 2005, 69, 225-257.
Surface Speciation of PO4 on Goethite
Figure 2. Logarithm of the equilibrium concentration in phosphategoethite systems in 0.5 M NaNO3 having a range of initial P concentrations (0.4-0.8 mM) and goethite concentrations (3-9 g/L). The model lines are calculated using the CD values derived from MO/DFT-optimized geometries after correction for dipole orientation effects (Table 8).
As shown in Figure 1, the salt effect is opposite on either side of the IEP. This effect of electrolyte ions on ion adsorption can be explained by an increase in the screening of surface charge, lowering the potential.33 Below the IEP, an increase in the salt concentration lowers the electrostatic attraction of phosphate, and above the IEP, a lower repulsion of phosphate is found, resulting in, respectively, higher and lower equilibrium concentrations. The ionic strength of the very low electrolyte level (
