Environ. Sci. Technol. 2008, 42, 3786–3790
Arsenate Removal by Nanostructured ZrO2 Spheres K I R I L D . H R I S T O V S K I , * ,† PAUL K. WESTERHOFF,‡ JOHN C. CRITTENDEN,‡ AND LARRY W. OLSON§ Environmental Technology Laboratory, Arizona State UniversitysPolytechnic Campus, 6075 S. WMS Campus Loop W, Mesa, Arizona 85212, Department of Civil and Environmental Engineering, Arizona State University, Box 5306, Tempe, Arizona 85287-5306, and Environmental Technology Management, Arizona State UniversitysPolytechnic Campus, 6075 S. WMS Campus Loop W, Mesa, Arizona 85212
Received November 26, 2007. Revised manuscript received February 7, 2008. Accepted February 11, 2008.
A new zirconium oxide-based media for arsenate removal from water was fabricated and evaluated in batch and continuous flow experiments. Highly porous (P ≈ 0.9) nanostructured zirconium oxide spheres were fabricated by the impregnation of macroporous ion-exchange media (CalRes 2103, Calgon) with zirconium salt; the media was then ashed at T > 750 ( 50 °C to remove the organic polymer resin and obtain ZrO2 spheres. The spheres generally ranged from 200 to 800 µm in diameter, and consisted of ZrO2 nanoastructures generally ranging between 20 and 100 nm. They also exhibited monoclinic and tetragonal crystalline structures, and had an isoelectric point of 5.6. Equilibrium batch experiments were conducted in 10 mM NaHCO3 buffered nanopure water at three pH values (6.4, 7.3, and 8.3) with 120 µg/L As(V). Data were fit with the Freundlich isotherm equation (qe ) K × CE1⁄n ), resulting in an intensity parameter (1/n) of ∼ 0.33 and capacity parameters (K) ranging from 115 to 400 (µg As(V) g-1 dry media)(L µg-1)1/n. The pore diffusion coefficient and toruosity were estimated to be 6.4 × 10-6 cm2 s-1 and 1.3, respectively. For a packed bed adsorbent operating at a loading rate of 11.5 m3 m-2 hr-1 in a realistic continuousflowexperiment,theexternalmasstransportcoefficient was estimated to be kf ≈ 6.3 × 10-3 cm s-1. The pore diffusion coefficient and the external mass transport coefficient were used with the pore surface diffusion model (PSDM) to predict the arsenate breakthrough curve. A short bed adsorbent (SBA) test was conducted under the same conditions to validate the model. In this study, surface diffusion was ignored because the particles have a very high porosity. The validated model was used to predict arsenate breakthrough in a simulated full-scale system. The overall combined use of modeling, material characterization, equilibria, and kinetics tests determined the suitability of the media for arsenate treatment cheaper, easier, faster, and with less media than a long * Corresponding author phone: 480 727-1132; fax: 480 727-1684; e-mail:
[email protected]. † Environmental Technology Laboratory, Arizona State UniversitysPolytechnic Campus. ‡ Arizona State University. § Environmental Technology Management, Arizona State UniversitysPolytechnic Campus. 3786
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 10, 2008
duration pilot test would have. Although the fabricated zirconium oxide spheres exhibited adsorption capacity comparable to some commercially available media such as iron based (hydr)oxides, the high cost of fabrication may render the media not feasible for wide use in commercial applications. However, the very high porosity of this media provides for improved pore diffusion and faster overall mass transport, which may be critical for applications where mass transport is the limiting factor.
Introduction Arsenate, a ubiquitous metalloid that occurs naturally in surface and groundwater, has raised potential health and regulatory concerns as a result of its toxic and carcinogenic properties. Arsenate adsorbs onto metal (hydr)oxide surfaces by forming innersphere bidentate ligands (1–4). This property makes metal (hydr)oxides good candidates for use as absorbents in treatments that remove arsenate from water. Many reports examine the use of metal (hydr)oxides for arsenate treatment, but the majority of them focus on iron and titanium based (hydr)oxides (5). Several reports pertaining to zirconium (hydr)oxides for arsenate treatment focus on their deposition onto support materials such as granulated activated carbon (GAC) and polymer resin. However, a recent study by Hristovski et al. (6) suggests that zirconium-based nanomaterials exhibit high enough affinity for arsenate to be considered as media in water treatment for arsenate. Because of unique characteristics such as large surface area and specific functionality, nanomaterials may provide an excellent alternative to the conventional adsorbent metal (hydr)oxide materials used in the removal of pollutants from water. However, use of nanoparticles in a fixed bed adsorber can be very challenging because of their size and potential release into the finished water. Nanomaterials can be aggregated using binding agents, which would prevent their release into the water. However, aggregated nanomaterials can become intraparticle diffusion limited, which would limit the overall removal performance of the treatment media in a packed bed system. As such, addressing the issue of contaminant mass transport is critical in assessing the viability of nanomaterial adsorbents in full-scale systems. One approach to minimizing intraparticle diffusion rates is to develop adsorbents that have highly porous nanostructures but still maintain a particle size large enough to allow their use in a fixed bed adsorber. The goal of this study is to develop zirconium oxide nanostructured adsorbent media and evaluate their suitability for arsenate removal from water. To achieve this goal, these steps were undertaken: (1) fabricate the zirconium oxide nanostructured media; (2) characterize the media by X-ray diffraction (XRD), energy dispersive X-ray microanalysis (EDX), focused ion beam/scanning electron microscopy (FIB/SEM) and focused X-ray microscopy (FXS), and surface charge and surface area analysis; (3) evaluate the adsorption capacity by conducting equilibrium tests; (4) quantify the mass transport processes that control the rate of arsenate adsorption in a fixed bed column; and (5) predict the performance of a full scale system using the pore surface diffusion model (PSDM).
Experimental Approach Media Fabrication. Nanostructured zirconium oxide spheres were fabricated by modifying a synthesis method developed for the preparation of a porous resin loaded with crystalline hydrous ZrO2 (7). Briefly, 50 mL of macroporous ion-exchange 10.1021/es702952p CCC: $40.75
2008 American Chemical Society
Published on Web 04/02/2008
resin CalRes 2103 by Calgon (polystyrene cross-linked with divinylbenzene) was soaked in ultrapure water ( 500; 0.6 e Sc e 104; 1 e Re < 100; 0.26 < < 0.935; where kf is the external mass transport coefficient (calculated kf ≈ 6.3 × 10-3 m s-1); Re is the Reynolds number (unitless); Sc is the Schmidt number (unitless); dp is the adsorbent particle diameter (dp ) 0.400 × 10-3 m); Dl is the free liquid diffusivity for arsenate (Dl ) 9.05 × 10-10 m2 s-1) (10); is the bed void fraction ( ) 0.376); µl is the dynamic viscosity of water at 20 °C (1.002 × 10-3 N s m-2); Fl is the density of water at 20 °C (Fl ) 998.2 kg m-3); Φ is the sphericity of the particle (Φ ) 1); and vl is the liquid superficial velocity (vl ≈ 0.00319 m s-1). Considering that the material was very porous (the particle porosity ≈ 0.92), the impact of surface diffusion was assumed to be negligible. As suggested by Sontheimer et al. (8), the pore diffusion coefficient was estimated using eq 6: DP )
P × Dl τ
(6)
The tortuosity was estimated using the correlation suggested by Mackie and Meares (eq 7) for electrolyte solutions (11): τ)
(2 - P)2 P
(7)
where τ is the toruosity factor and P is the particle porosity (P ≈ 0.92). The estimated toruosity value was τ ≈ 1.3. The estimated value for the pore diffusion coefficient was DP ≈ 6.4 × 10-6 cm2 s-1. The pore and surface diffusion model (PSDM) was used to predict the arsenate breakthrough curve (12–14). Pore surface diffusion model is a dynamic fixed bed model that incorporates a set of assumptions and governing partial differential equations that describe the adsorber dynamics in a fixed bed setup. Pore surface diffusion model simulations were conducted using AdDesignS software (Michigan Technological University) (15). To validate the calculated kf and DP values, a short bed adsorber (SBA) test was conducted. Short bed adsorber tests are continuous flow column experiments with a packed bed long enough to describe the dissolved pollutant mass transfer zone (16, 17). The initial model estimates predicted that a SBA column with a diameter of 1.1 cm and a bed depth of 1 cm should be sufficiently long enough to describe the dissolved pollutant mass transfer zone at a hydraulic loading rate of 3.2 L m-2 s-1, an initial arsenate concentration C0 of 110 µg L-1 and a temperature of 20 °C. Such loading rates are typical for full scale fixed bed column absorbers (6, 18, 19). In the short bed adsorber test, a 1 cm deep adsorbent media bed (mass of 0.415 g) was packed atop a support composed of quartz sand and a metal screen in a glass column with diameter (dcolumn) of 1.1 cm (Ace Glass). Glass beads were placed above and below to provide evenly distributed flow. As illustrated in Figure 1a, the size of the adsorbent media particles generally ranged between 200 µmand 800 µm (geometric mean of 400 µm) and provided a dcolumn/dp ratio of ∼27. According to Benenati and Brosilow (20) and Chu and Ng (21), the wall effect on mass transfer can be neglected for dcolumn/dp ratios >20. Effluent from the short bed adsorber test was collected in 30 mL sample aliquots (the minimum needed to conduct the necessary analyses), which represented an average of 30 bed volumes. The relative importance of internal (pore) and external (film) mass transport resistances was evaluated by estimating the pore (BiP) Biot number using the relationships given by Sontheimer et al. (8): BiP )
k f × dP 2 × DP
VOL. 42, NO. 10, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
(8)
9
3787
FIGURE 1. ZrO2 spheres observed at different magnifications using a focused X-ray microscope (a) and a focused ion beam/ scanning electron microscope (b, c, d). The maximum number of bed volumes that could be treated with the media was also estimated using a relationship given by Sontheimer et al. (8): BVMAX )
q0 × FBED × 1000 C0
(9)
where FBED is the bed density of the media in the packed bed (g cm-3) and the multiplication factor results from unit conversion. Full-Scale System Modeling Using Validated Pore Surface Diffusion Model. The pore surface diffusion model was used to model the performance of full scale fix bed systems operating at the same loading rate as the short bed adsorber column tests and empty bed contact times (EBCTs) of 2.5, 5, and 10 min. To maintain the same loading rate, the length of the packed bed was changed to achieve the desired EBCTs. The modeling was conducted with a realistic value of C0 ) 30 µg L-1. The water chemistry, pH, and bed porosity were assumed to be the same as those used in the SBA test. Arsenic Analysis. Arsenate was analyzed using a graphite furnace atomic absorption spectrophotometer (GF-AAS) Varian Zeeman Spectra 400 with the GTA 96 system (22).
Results and Discussion Media Characterization. FXS and SEM/FIB analyses of the media revealed the presence of highly porous and spherical particles with sizes generally ranging between 200 µm and 800 µm (Figures 1a and 1b). As illustrated in Figure 1c, analysis of the material at higher magnification (∼ 50 000×) suggests that the microscopic features of this media are actually composed of ZrO2 nanostructures ranging between 20 and 100 nm and providing a sponge like structure. Although the surfaces of the particles contain large pores (>5 µm), Figure 3788
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 10, 2008
FIGURE 2. X-ray diffraction (XRD) and energy dispersive X-ray (EDX) analysis of the ZrO2 spheres. 1d suggests relatively uniform distribution of the pores throughout the media. EDX analysis of the spheres suggested the elemental composition of ZrO2, while the XRD analysis suggested the presence of monoclinic and tetragonal crystalline structures (Figure 2). The distinctive peak at 2θ angles of ∼ 30° is typical of tetragonal zirconia structures, while the peak at ∼28° is representative of monoclinic structures (23, 24). The specific surface area of the material was estimated to be ≈ 32 m2 g-1, and the material density was estimated as FM ≈ 5.65 g cm-3. The particle density is FP ≈ 0.41 g cm-3, and the particle porosity is P ≈ 0.92. From Equation 1, the
FIGURE 3. Arsenate adsorption isotherms for ZrO2 spheres in 10 mM NaHCO3 buffered nanopure water after a contact time of 3 days (C0-As(V) ≈ 120 µg L-1). average pore diameter is dpore ≈ 270 nm, indicating that the spheres are macroporous according to the IUPAC classification (9). The estimated isoelectric point for the spheres is pHZPC ≈ 5.6, which is similar to values reported in the literature for ZrO2 (6, 25). The zeta potential (ζ) was ζ ≈ -7 mV at pH 6.4 ( 0.1, and ζ ≈ -13 mV at pH 7.3 ( 0.1. The zeta potential was ζ ≈ -22 mV at pH 8.3 ( 0.1; this implies that arsenate adsorption may be much lower than that for pH values of 6.4 and 7.3. Equilibrium Adsorption Experiments. Figure 3 presents arsenate adsorption isotherms. Table 1 summarizes the Freundlich isotherm adsorption parameters for the equilibrium experimental conditions and the [H2AsO4-]/[HAsO42-] ratios at each pH as modeled using MINEQL+ (26). Arsenate adsorption is highest at pH 6.4 ( 0.1 and lowest at pH 8.3 ( 0.1. Since arsenate has pKa2-Arsenate ) 6.8 and pKa3-Arsenate ) 11.6, H2AsO4- and HAsO42- species will be dominant in waters with pH between 6 and 9, such as those typically found in the environment (27, 28). At pH 8.3, the HAsO42-/H2AsO4ratio ≈ 46, implying that almost all of the arsenate will be present in the more negative form, which will tend to adsorb less onto a negatively charged ZrO2 surface. In contrast, at pH 6.4, the [HAsO42-]/[H2AsO4-]ratio ≈ 0.65, implying a greater presence of H2AsO4- ions and thus better adsorption onto the negatively charged surface of the ZrO2 spheres. The adsorption capacities at pH 6.4 ( 0.1 and pH ) 7.3 ( 0.1 were only slightly different, as illustrated in Figure 3. This is not surprising considering the small difference in ζ at the given pH values, and the pKa values for arsenate. The Freundlich adsorption intensity parameter 1/n ≈ 0.33; this corresponds to favorable adsorption, i.e., 1/n < 1. The Freundlich adsorption capacity parameters (K) ranged from 115 to 400 µg As(V)/g dry media. Comparable 1/n and K values have been reported for ZrO2 nanopowder and other media such as iron (hydr)oxides (5, 6, 29). Short Bed Adsorber Test and Pore Surface Diffusion Modeling. Figure 4 presents the data and prediction for arsenate breakthrough in SBA tests conducted at a loading rate of 3.2 L m-2 s-1 (4.7 gal min-1 ft-2; Re × Sc ≈ 2100) and an initial arsenate concentration of C0(As) ≈ 110 µg L-1. As illustrated, rapid breakthrough occurred initially due to the
FIGURE 4. PSDM prediction and experimental data from the SBA tests for ZrO2 spheres. short bed depth, reaching 50% (C/C0 ≈ 0.5) at approximately 100 bed volumes. A complete breakthrough (C/C0 ≈ 0.95) occurred much later, at approximately 33 500 bed volumes. The rapid breakthrough is expected because the adsorption sites located on the outermost surfaces of the particles quickly become occupied with arsenate, and the only available sites are located inside the particles. The gradual breakthrough following the rapid increase is also expected as the intraparticle mass transport becomes more limiting due to the longer time needed for arsenate to diffuse inside the media particle. Based upon the calculated values of kf ≈ 6.3 × 10-3 cm s-1 and DP ≈ 6.4 × 10-6 cm2 s-1 from eqs 3 and 6, the pore surface diffusion model provided a good prediction (line in Figure 4) of the arsenate breakthrough (R2 ≈ 0.97). The estimated pore Biot number (BiP) was ≈ 20. A Biot number g20 implies that intraparticle diffusion controls the overall mass transport of the system (8). Performance of Full-Scale Fixed Bed Systems. Figure 5 illustrates the use of validated pore surface diffusion model to predict arsenate breakthrough at commonly used empty bed contact times (EBCTs). The number of bed volumes that can be treated until the maximum contaminant level (MCL) of 10 µg/L As(V) (C/C0 ) 0.33) is reached increases with increasing EBCT. The pore surface diffusion model predicts that approximately 14 870 bed volumes can be treated at an EBCT of 2.5 min; this number increases to approximately 15 110 bed volumes by doubling the EBCT to 5 min. At an EBCT of 10 min, the number of bed volumes that can be treated is approximately 15 130, which approaches the maximum number of bed volumes that can be treated, estimated as BVMAX ≈ 15 200 BV. Figure 5b illustrates the same arsenate breakthrough predictions presented in Figure 5a, but expressed as liters of treated water per g dry media; this can be used to compare the performance of the ZrO2 spheres with other commercially available media that have been evaluated using column tests under similar conditions (pH 7.6 ( 0.3 and C0-As ≈ 30 µg/L). As illustrated in Figure 5b, approximately 35 L can be treated per g of dry media until maximum contaminant level of 10 µg/L As(V). This value is comparable to the values reported by Barduzzaman et al. (18), Badruzzaman (30) and Westerhoff et al. (19) for commercially available iron (hydr)oxides.
TABLE 1. Zeta Potentials, Freundlich Isotherm Adsorption Parameters and [HAsO42-]/[H2AsO4-] Ratios at pH 6.4 ± 0.1, pH 7.3 ± 0.1, and pH 8.3 ± 0.1 pH
zeta potential ζ (mV)
K(µgAs(V)/m2/ (µg/L)1/n)
K(µgAs(V)/g/ (µg/L)1/n)
K(mgAs(V)/g/ (mg/L)1/n)
1/n (unitless)
[HAsO42-] ⁄ [H2AsO4-]
6.4 ( 0.1 7.3 ( 0.1 8.3 ( 0.1
-7 -13 -22
12.50 10.22 3.59
399.81 327.09 114.75
3.19 3.45 1.42
0.30 0.34 0.36
∼0.65 ∼8 ∼46
VOL. 42, NO. 10, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3789
(7)
(8)
(9) (10) (11)
(12)
(13)
(14)
(15)
FIGURE 5. PSDM prediction of a full scale system packed with ZrO2 spheres as the media expressed as (a) bed volumes treated, and (b) liters treated per gram media. Considering that ZrO2 spheres would cost more to fabricate on a large scale than iron based (hydr)oxides, the media may not be feasible for wide use in commercial applications for arsenate removal and may be limited to specific applications where iron based (hydr)oxides cannot be used. In contrast, the very high porosity of this media provides for improved pore diffusion and faster overall mass transport, which may be critical for applications where mass transport is the limiting factor. However, further investigations are needed that may broaden the palette of applications of this novel ZrO2 media for uses beyond an adsorbent.
(16)
(17)
(18)
(19)
(20) (21) (22)
Acknowledgments We acknowledge Daniel Wilson, Ph.D., for help with the statistics, Grant Baumgardner for help with the microscopy, and Thomas Groy, Ph.D., for help with the X-ray diffraction analysis.
Literature Cited (1) Goldberg, S.; Johnston, C. T. Mechanisms of arsenic adsorption on amorphous oxides evaluated using macroscopic measurements, vibrational spectroscopy, and surface complexation modeling. J. Colloid Interface Sci. 2001, 234 (1), 204–216. (2) Sherman, D. M.; Randall, S. R. Surface complexation of arsenic(V) to iron(III) (hydr)oxides: Structural mechanism from ab initio molecular geometries and EXAFS spectroscopy. Geochim. Cosmochim. Acta 2003, 67 (22), 4223–4230. (3) Weerasooriya, R.; Tobschall, H. J.; Wijesekara, H. K. D.; Bandara, A. Macroscopic and vibration spectroscopic evidence for specific bonding of arsenate on gibbsite. Chemosphere 2004, 55 (9), 1259–1270. (4) Ona-Nguema, G.; Morin, G.; Juillot, F.; Calas, G.; Brown, G. E. EXAFS analysis of arsenite adsorption onto two-line ferrihydrite, hematite, goethite, and lepidocrocite. Environ. Sci. Technol. 2005, 39 (23), 9147–9155. (5) Mohan, D.; Pittman, C. U. Arsenic removal from water/ wastewater using adsorbents - A critical review. J. Hazard. Mater. 2007, 142, 1–53. (6) Hristovski, K.; Baumgardner, A.; Westerhoff, P. Selecting metal oxide nanomaterials for arsenic removal in fixed bed columns:
3790
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 10, 2008
(23)
(24) (25) (26)
(27)
(28)
(29)
(30)
from nanopowders to aggregated nanoparticle media. J. Hazard. Mater. 2007, 147, 265–274. Suzuki, T. M.; Bomani, J. O.; Matsunaga, H.; Yokoyama, T. Preparation of porous resin loaded with crystalline hydrous zirconium oxide and its application to the removal of arsenic. React. Funct. Polym. 2000, 43 (1–2), 165–172. Sontheimer, H.; Crittenden, J.; Summers, S. Activated Carbon for Water Treatment, 2nd ed.; DVGW-Forschungsstelle, EnglerBunte Institut, Universitat Karlsruhe: Karlsruhe, Germany, 1988. Crittenden, J. C., Trussell, R. R., Hand, D. W., Howe, K. J., Tchobanoglous, G., Eds. Water Treatment: Principles and Design, 2nd ed.; Wiley & Sons, Inc.: Hoboken, NJ, 2005. CRC Handbook of Chemistry and Physics, 87th ed.; Lide, D., Ed.; Taylor and Francis Group: Boca Raton, FL, 2006. LeVan, D. M.; Carta, G.; Yon, C. M. Chapter 16, Adsorption and Ion Exchange, In Perry’s Chemical Engineers’ Handbook, 7th ed.; Perry, R. D., Green D. W., Eds.; McGraw-Hill: New York, 1997. FriedmanG. Mathematical Modeling of Multicomponent Adsorption in Batch and Fixed-Bed Reactors. Master’s Thesis, Michigan Technological University, Houghton, MI, 1984. Crittenden, J. C.; Hutzler, N. J.; Geyer, D. G.; Oravitz, J. L.; Friedman, G. Transport of organic compounds with saturated groundwater flow: Model development and parameter sensitivity. Water Resour. Res. 1986, 22, 271–284. Hand, D. W.; Crittenden, J. C.; Hokanson, D. R.; Bulloch, J. L. Predicting the performance of fixed-bed granular activated carbon adsorbers. Water Sci. Technol. 1997, 35, 235–241. Mertz, K. A.; Gobin, F.; Hand, D. W.; Hokanson, D. R.; Crittenden, J. C. Manual: Adsorption Design Software for Windows (AdDesignS); Michigan Technological University: Houghton, MI, 1999. Weber, W. J.; Smith, E. H. Simulation and design models for adsorption processes. Environ. Sci. Technol. 1987, 21 (11), 1040– 1050. Smith, E. H.; Weber, W. J. Modeling activated carbon adsorption of target organic-compounds from leachate-contaminated groundwaters. Environ. Sci. Technol. 1988, 22 (3), 313–321. Badruzzaman, M.; Westerhoff, P.; Knappe, D. R. U. Intraparticle diffusion and adsorption of arsenate onto granular ferric hydroxide (GFH). Water Res. 2004, 38 (18), 4002–4012. Westerhoff, P.; Highfield, D.; Badruzzaman, M.; Yoon, Y. Rapid small-scale column tests for arsenate removal in iron oxide packed bed columns. J. Environ. Eng-Asce 2005, 131 (2), 262– 271. Benenati, R. F.; Brosilow, C. B. Void fraction distribution in bed of spheres. AIChE J. 1962, 8 (3), 351–361. Chu, C. F.; Ng, K. M. Flow in packed tubes with a small tube to particle diameter ratio. AIChE J. 1989, 35 (1), 148–158. Standard Methods for the Examination of Water and Wastewater, 19th ed.; Franson, M. A. H., Eaton, A. D., Clesceri, L. S.; Greenberg, A. E., Eds.;American Public Health Association: Washington DC, 1995. Ksapabutr, B.; Gulari, E.; Wongkasemjit, S. Preparation of zirconia powders by sol-gel route of sodium glycozirconate complex. Powder Technol. 2004, 148, 11–14. Guo, G. Y.; Chen, Y. L. Unusual structural phase transition in nanocrystaline zirconia. Appl. Phys. 2006, A84, 431–437. Greenwood, R.; Kendall, K. Selection of suitable dispersants for aqueous suspensions of zirconia and titania powders using acustophoresis. J. Eur. Ceram. Soc. 1999, 19, 479–488. Schecher, W. D.; McAvoy, D. C.; William, D. MINEQL+ a Chemical Equilibrium Modeling System: Version 4.0 for Windows User’s Manual; Environmental Research Software: Hallowell, ME, 1988. Smedley, P. L.; Kinniburgh, D. G. A review of the source, behaviour and distribution of arsenic in natural waters. Appl. Geochem. 2002, 17 (5), 517–568. Dutta, P. K.; Ray, A. K.; Sharma, V. K.; Millero, F. J. Adsorption of arsenate and arsenite on titanium dioxide suspensions. J. Colloid Interface Sci. 2004, 278 (2), 270–275. Sperlich, A.; Werner, A.; Genz, A.; Amy, G.; Worch, E.; Jekel, M. Breakthrough behavior of granular ferric hydroxide (GFH) fixedbed adsorption filters: Modeling and experimental approaches. Water Res. 2005, 39 (6), 1190–1198. Badruzzaman, M. Mass Transport Scaling and the Role of Silica on Arsenic Adsorption onto Porous Iron Oxide (Hydroxide). Ph.D. Dissertation, Tempe, AZ, 2005.
ES702952P
