Slow Adsorption Reaction between Arsenic Species and Goethite (α

Langmuir 2005, 21, 2895-2901

2895

Slow Adsorption Reaction between Arsenic Species and Goethite (r-FeOOH): Diffusion or Heterogeneous Surface Reaction Control Junshe Zhang and Robert Stanforth* Department of Chemical and Biomolecular Engineering, National University of Singapore, Block E5 No. 02-9, 4 Engineering Drive 4, Singapore 117576 Received September 23, 2004 The slow stage of phosphate or arsenate adsorption on hydrous metal oxides frequently follows an Elovich equation. The equation can be derived by assuming kinetic control by either a diffusion process (either interparticle or intraparticle) or a heterogeneous surface reaction. The aim of this study is to determine whether the slow stage of arsenic adsorption on goethite is more consistent with diffusion or heterogeneous surface reaction control. Adsorption kinetics of arsenate and dimethylarsinate (DMA) on goethite (R-FeOOH) were investigated at different pH values and inert electrolyte concentrations. Their adsorption kinetics was described and compared using Elovich (Γ vs ln time) plots. Desorption of arsenate and DMA was studied by increasing the pH of the suspension from pH 4.0 to pH 10.0 or 12.0. The effective particle sizes and ζ-potential of goethite were also determined. Effective particle size increased rapidly as the pH approached pHIEP, both in the absence and presence of arsenic. Inert electrolyte concentrations and pH had no effect on the slow stage of arsenate adsorption on goethite, while the kinetics of DMA adsorption on goethite was influenced by both parameters. The slow stage of arsenate adsorption on goethite follows an Elovich equation. Since effective particle size changes with both pH and inert electrolyte concentrations, and effective particle size influences interparticle diffusion, the arsenate adsorption kinetics indicate that the slow adsorption step is not due to interparticle diffusion. DMA also has complex adsorption kinetics with a slow adsorption stage. DMA desorbed completely and rapidly when the pH was raised, in contrast to the slow adsorption kinetics, indicating that the slow adsorption step is not due to intraparticle diffusion. The slow adsorption is not the result of diffusion, but rather is due either to the heterogeneity of the surface site bonding energy or to other reactions controlling arsenic removal from solution.

Introduction Arsenic is widespread in the aquatic environment, soils, and sediment and is known to be highly toxic. Arsenic exists predominantly in nature as an oxyanion with an oxidation state of either +3 or +5. Organoarsenic compounds, such as monomethylarsonate (MMA, CH3AsO(OH)2) and dimethylarsinate (DMA, (CH3)2AsOOH), are also widely distributed in the environment.1 The interaction between arsenic species and mineral surfaces will affect arsenic’s mobility and potential bioavailability in natural systems. Iron oxides and oxyhydroxides (ferric (hydr)oxide), which are present in soils, sediments, and aquatic environments, have a strong affinity for arsenic species.2-4 Understanding the reactions between ferric (hydr)oxide and the different arsenic species is important for understanding and modeling the movement of arsenic in environmental systems. Both adsorption and desorption and the kinetics of these reactions are important in this regard. Goethite (R-FeOOH) has been widely used as a representative iron oxide in adsorption studies because it is widespread in nature, can be synthesized readily in the laboratory, and has a well-characterized structure.5 The kinetics of arsenic adsorption is important not only for predicting the fate of arsenic and its dynamic interac* Corresponding author. E-mail: [email protected]. (1) Cullen, W. R.; Reimer, K. J. Chem. Rev. 1989, 89, 713. (2) Jain, A.; Raven, K. P.; Loeppert, R. H. Environ. Sci. Technol. 1999, 33, 1179. (3) Arai, Y.; Sparks, D. L.; Davis, J. A. Environ. Sci. Technol. 2004, 38, 817. (4) Sun, X. H.; Doner, H. E. Soil Sci. 1996, 161, 865. (5) Cornell, R. M.; Schwertmann, U. The Iron Oxides: Structure, Properties, Reactions, Occurrence and Uses, 2nd ed.; Wiley-VCH: Weinheim, Germany, 2003.

tion with soils but also for providing insight into the mechanism of adsorption on hydrous metal oxides. Adsorption of arsenate on hydrous iron oxides has been shown to be rapid initially followed by a slow stage.6-9 Two-phase sorption kinetics patterns have also been observed for phosphate as well as many trace elements.5 Two factors have been suggested to account for the slow stage adsorption: diffusion (interparticle or intraparticle),10,11 or surface reactions, including surface precipitation, and surface site bonding energy heterogeneity or other surface reactions.12,13 Diffusion is dependent on the particle size, porosity, or crystal structure of the adsorbent, while the surface reaction heterogeneity is related to variations in bonding energy between the surface site and adsorbate. Fuller et al.6 and Raven et al.8 suggested that the reactions between arsenate/arsenite and ferrihydrite were diffusion-controlled, and used a parabolic diffusion equation to describe the kinetic data. Alternatively, Zhao and Stanforth13 suggested that surface precipitation accounted for the slow stage of phosphate and arsenate sorption on goethite. The Elovich equation has found wide application in studying the kinetics of gas adsorption on solid surfaces.14 (6) Fuller, C. C.; Davis, J. A.; Waychunas, G. A. Geochim. Cosmochim. Acta 1993, 57, 2271. (7) O’Reilly, S. E.; Strawn, D. G.; Sparks, D. L. Soil Sci. Soc. Am. J. 2001, 65, 67. (8) Raven K. P.; Jain, A.; Loepert, R. H. Environ. Sci. Technol. 1998, 32, 344. (9) Waltham, C. A.; Eick, M. J. Soil Sci. Soc. Am. J. 2002, 66, 818. (10) Barrow, N. J. J. Soil. Sci. 1983, 34, 733. (11) Eick, M. J.; Peak, J. D.; Brady, P. V.; Pesek, J. D. Soil Sci. 1999, 164, 28. (12) Grossl, P. R.; Eick, M.; Sparks, D. L.; Goldberg, S.; Ainsworth, C. C. Environ. Sci. Technol. 1997, 31, 321. (13) Zhao, H. S.; Stanforth, R. Environ. Sci. Technol. 2002, 35, 4753.

10.1021/la047636e CCC: $30.25 © 2005 American Chemical Society Published on Web 02/25/2005

2896

Langmuir, Vol. 21, No. 7, 2005

Zhang and Stanforth

The Elovich equation has also been used for aqueous systems to describe isotopic exchange of phosphate on goethite,15 adsorption of phosphate on gibbsite,16 phosphate and arsenate on goethite,13 and kinetics of silicate adsorption and desorption on aluminum hydroxide.17 The Elovich equation is as follows:

Γ)

1 ln(1 + Rβt) β

(1)

where R and β are constants, t is the time, and Γ is the surface coverage at t. When Rβt is much greater than 1, eq 1 can be simplified to

with arsenate (pK1 ) 2.2, pK2 ) 6.97, and pK3 ) 11.53) except that two hydroxyl groups are replaced by methyl groups. At present there are few studies on adsorption or desorption of DMA on goethite.22,23 In this paper, the question of diffusion control versus surface reaction control of the slow step during both arsenate and DMA on goethite is addressed using three approaches: (1) analyzing the adsorption kinetics at different pH values and inert electrolyte concentrations, (2) measuring desorption of arsenate and DMA as pH is increased, and (3) investigating adsorption kinetics on particles of different size, due to differences in solution pH or inert electrolyte concentrations. Materials and Methods

1 1 1 Γ ) ln(Rβt) ) ln(Rβ) + ln(t) β β β

(2)

Kinetics results are linear on a Γ vs ln t plot if the results follow an Elovich equation. Parravano and Boudart18 noted that a number of different processes, including bulk and surface diffusion as well as chemisorption, might be described by an Elovich equation. The Elovich equation can be derived from either a diffusion-controlled process or a reaction-controlled process. If the Elovich equation is based on adsorption on an energetically heterogeneous surface, the parameter β is related to the distribution of activation energies. In the diffusion control model, β is a function of particle and diffusion coefficient. On the basis of the application of the absolute rate theory to adsorption on an energetically heterogeneous surface with a rectangular distribution of activation energies for adsorption, Cerofolini19 suggested that using the Elovich equation to describe adsorption kinetics implies that the adsorption process is at quasi-equilibrium, that the process proceeds in a stepwise fashion, and that activation energy increases linearly with surface coverage. Rudzinski and Panczyk14 showed that the Elovich equation should apply at conditions where the desorption rate can be neglected. Atkinson et al.15 derived an Elovich equation using a continuous distribution of activation energy. In contrast, Pavlatou and Polyzopoulos20 suggested that conformation to the Elovich equation alone might be taken as evidence that the rate-determining step is diffusion in nature. Aharaoni et al.21 also suggested that diffusion accounted for the Elovich kinetics pattern. It is questionable to infer the mechanism for the ratedetermining steps based only on the fit of kinetics data to a particular model. Other studies are necessary to determine what processes control the reaction rate. One method is to use similar but not identical adsorbates and evaluate how slight differences in the adsorbate influence the kinetics. DMA (pK ) 6.27) has a similar structure (14) Rudzinksi, W.; Panczyk, P. In Surfaces of Nanoparticles and Porous Materials; Schwarz, J. A., Contescu, C. I., Eds.; Dekker: New York, 1998; p 355. (15) Atkinson, R. J.; Posner, A. M.; Quirk, J. P. J. Inorg. Nucl. Chem. 1972, 34, 2201. (16) Vanderdeelen, J.; Baert, L. Pedologie 1971, 21 (3), 360. (17) Hingston, F. J. In Adsorption of Inorganics at the Solid-Liquid Interface; Anderson, M. A., Rubin, A. J., Eds.; Ann Arbor Science: Ann Arbor, MI, 1981; p 67. (18) Parravano, G.; Boudart, M. In Advances in Catalysis; Frankenburg, W. G., Komarewsky, V.I., Rideal, E. K., Eds.; Academic Press: New York, 1955; Vol. 7, p 47. (19) Cerofolini, G. F. In Adsorption on New and Modified Inorganic Sorbents; Dabrowksi, A., Tertykh, V. A., Eds.; Elsevier: Amsterdam, 1996; p 435. (20) Pavlatou, A.; Polyzopouls, N. A. J. Soil Sci. 1988, 39, 425. (21) Aharoni, C.; Sparks, D. L.; Levinson, S.; Revina, I. Soil Sci. Soc. Am. J 1991, 55, 1307.

Syntheses and Characterization. Goethite was synthesized using the method of Atkinson et al.24 The pH of a ferric nitrate solution (Fe(NO3)3‚9H2O (Merck)) was raised to pH 12.0 with NaOH (Merck) to precipitate amorphous iron hydroxide (ferrihydrite). The suspension was aged at 60 °C in a plastic bottle for 72 h, with periodic shaking until a change in color from red to orange was observed. The suspension was filtered once and then washed by placing it in a 4 L plastic bottle with distilled deionized water, allowing the precipitate to settle and then decanting the supernatant. Washing was continued until the conductivity of supernatant was less than twice that of deionized water and pH was near 7.0. At least six to seven washings were needed to remove the sodium and nitrate ions, polymeric complexes, and small amorphous particles. The goethite was stored in a plastic bottle in deionized water until use. The goethite was characterized using X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), field emission scanning electron microscopy (FESEM), and transmission electron microscopy (TEM). The specific surface area of synthesized goethite determined from N2-BET adsorption is 27 m2 g-1. During the goethite preparation and adsorption experiments, the samples were exposed to the atmosphere. It is probable that some carbonate adsorption occurred on goethite.25 Experiments in the air and under nitrogen have shown that adsorption isotherms for both arsenate and DMA are not affected by the presence of carbonate at pH values below 8. Adsorption and Desorption Kinetics. A 200 mL aliquot of goethite suspension and 25 mL of arsenic solution (sodium dihydrogen arsenate or sodium dimethylarsinate), both at the target pH of the experiment, were added to a 250 mL Teflon bottle, and the pH was adjusted with 0.1 M HNO3 or NaOH, followed by continuous magnetic stirring or ultrasonication of the suspension to ensure thorough mixing. Aliquots were taken from the sample over a period of 0.25-24 h. The aliquots were filtered through a 0.45 µm filter (Whatman Autovial), acidified, and stored until analysis. For the desorption studies, arsenate/ DMA was adsorbed at pH 4.0 for 14 or 24 h, and then pH was increased from 4.0 to 10.0 or 12.0 to desorb arsenic. Aliquots were taken from the sample at a preset time. All experiments were conducted at room temperature. The samples were analyzed using a Perkin-Elmer Optima 3000 DV ICP-OES. ζ-Potential. A ZetaPlus (Brookhaven Instruments Corp.) was used to measure the ζ-potential of the goethite suspension. The recommended inert electrolyte concentration is 10-3 to 10-2 M for measuring the ζ-potential, and the maximum solids concentration that can be tolerated is about 0.4 g L-1. The ζ-potential was measured at different pH values and arsenic concentrations were at a constant inert electrolyte concentration (0.001 M NaNO3). A 40 mL aliquot of 0.25 g L-1 goethite suspension and appropriate concentrations of arsenate/DMA were placed in 60 mL Teflon bottles and shaken on a rotary shaker for 96 h. A sample of several milliliters was taken from the suspension to measure the ζ-potential of goethite. (22) Brain, P. J.; Miller, W. P. Soil Sci. Soc. Am. J. 2000, 64, 1616. (23) Bowell, R. J. Appl. Geochem. 1994, 9, 279. (24) Atkinson, R. J.; Posner, A. M.; Quirk, J. P. J. Inorg. Nucl. Chem. 1968, 30, 2371. (25) Wijnjia, H.; Schulthess, C. P. Soil Sci. Soc. Am. J. 2001, 65, 324.

Arsenic Species and Goethite Adsorption Kinetics

Langmuir, Vol. 21, No. 7, 2005 2897

Figure 1. FESEM image of goethite crystal. Effective Particle Size. A 90 Plus (Brookhaven) was employed to determine the effective particle size. The goethite suspension (0.25 g L-1) was equilibrated for 96 h with NaNO3 (0.001, 0.01, and 0.1 M) or different initial arsenic concentrations in the presence of 0.001 M NaNO3 at different pH values. The pH of suspensions was readjusted to a given value prior to measurement. In the study of the effect of ultrasonication on effective particle size, the effective particle sizes of goethite suspension were measured immediately after ultrasonication. Error Analysis. Arsenic analysis in the adsorption experiments using ICP-OES has a relative standard deviation of less than 5%. Replicates of the adsorption and desorption experiments were run in order to evaluate the experimental reproducibility. For the adsorption and desorption kinetic runs, the mean percent deviation of the replicate solution concentrations was 4.8% with a standard deviation of 4.5% (n ) 18 pairs) for arsenate and 0.76% with a standard deviation of 0.34% (n ) 15 pairs) for DMA. The mean percent deviations of the replicates of arsenate and DMA adsorption at three pH values in the presence of 0.1 M NaNO3 and at three salt levels at pH 4.00 were respectively 0.91% with a standard deviation of 0.97% (50 pairs) and 1.71% with a standard deviation of 1.37% (38 pairs) for the calculated surface coverage results. The mean percent deviation of the replicates of ζ-potential is 18% with a standard deviation of 25%, and for effective particle size the mean percent deviation of the replicate is 10% with a standard deviation of 5%.

Figure 2. Surface coverage (a) and ζ-potential (b) of arsenate adsorbed on goethite as a function of pH in the presence of 0.001 M NaNO3 with different initial arsenate concentrations. The goethite concentration is 0.25 g L-1.

Results and Discussion Characterization of Goethite. The XRD and FTIR results indicate that the adsorbent is goethite, with no other forms of hydrous iron oxide detected. Viewed under an SEM, the crystals are needle-shaped. A more magnified view using TEM shows that the goethite crystals consist of parallel subunits. FE-SEM analysis shows uniform plates with no obvious porosity, at least at a 10 nm scale (Figure 1). The results are typical for laboratory prepared goethite.5,24 ζ-Potential and Particle Sizes. The pHIEP of goethite in 0.001 M NaNO3 is about 8.5 (Figure 2), which is consistent with the result of Kovacevic et al.26 and Janusz.27 This value is lower than the result of Boily et al.28 and Hiemstra and van Riemsdijk29 for goethite prepared under carbonate-free conditions, possibly due (26) Kovaceive, D.; Cop, A.; Bradetic, A.; Kallay, N.; Pohlmeier, A.; Narres, H. D.; Lewandowski, H. Prog. Colloid Polym. Sci. 2001, 17, 32. (27) Janusz, W. L. In Interfacial Forces and Fields: Theory and Applications; Hsu, J. P., Ed.; Dekker: New York, 1998; p 190. (28) Boily, J.-F.; Lutzenkirchen, J.; Balmes, O.; Beattie, J.; Sjoberg, S. Colloid Surf., A 2000, 179, 11.

Figure 3. Surface coverage (a) and ζ-potential (b) of DMA adsorbed on goethite as a function of pH in the presence of 0.001 M NaNO3 with different initial DMA concentrations. The goethite concentration is 0.25 g L-1.

to different methods of preparation or to the adsorption of carbonate. The effects of arsenate and DMA on the ζ-potential of goethite are given in Figures 2 and 3. The ζ-potential decreased in the presence of arsenate, while DMA adsorption has no significant effect on the ζ-potential. Kreller et al.30 found that dimethyl phosphate ((CH3O)2POOH) had a negligible effect on the colloid ζ-potential, similar to the results presented here for DMA. Adsorption of strongly bound anions such as phosphate and arsenate, (29) Hiemstra, T.; Van Riemsdijk, W. H. J. Colloid Interface Sci. 1996, 179, 488. (30) Kreller, D. L.; Gibson, G.; Van Loon, G. W.; Hortorn, J. H. J. Colloid Interface Sci. 2000, 254, 205.

2898

Langmuir, Vol. 21, No. 7, 2005

Zhang and Stanforth

Figure 4. Effective particle size of goethite with different initial arsenate and DMA concentrations in the presence of 0.001 M NaNO3. Filled symbols are DMA and open symbols are arsenate. The goethite concentration is 0.25 g L-1.

Figure 5. Adsorption and desorption kinetics of arsenate and DMA with two solids concentrations in the presence of 0.1 M NaNO3. The initial concentration of DMA is 130 µM (1.98 g L-1, 9) and 71 µM (0.25 g L-1, b). The initial concentration of arsenate is 120 µM (1.98 g L-1, 0) and 15 µM (0.25 g L-1, O).

Table 1. Effective Particle Sizes (nm) of Pure Goethite Suspension at Different pH Values and NaNO3 Concentrations effective particle sizes of pure goethite pH

0.001 M

0.01 M

0.1 M

4.00 5.16 6.75 10.06

692 743 1245 1578

815 1169 2452 2765

1713 2461 3692 4284

induces a shift in the IEP to lower pH values, and this shift is more pronounced at higher surface coverage.31-34 The isoelectic point (IEP) of goethite moved to near pH 6.5 with an initial concentration of 5.3 µM arsenate. For initial concentrations from 28.6 and 73.7 µM, the pHIEP was around 3.6, and surface coverage of arsenate is almost the same for these two initial concentrations (Figure 3). Manning and Goldberg35 reported that the pHIEP of goethite is 2.8 in the presence of 133 µM arsenate, and this low pHIEP may be due to the high surface coverage of arsenate in their study. Li and Stanforth33 reported that the pHIEP of phosphated goethite with a high surface coverage was the same as the pHIEP for ferric phosphate. Figure 4 shows the effective particle sizes of goethite in the presence and absence of arsenate and DMA at different pH values. Since adsorption of ions affects the pHIEP, the pH is shown with respect to the pHIEP for the effective particle sizes. The effective particle size increased rapidly at the pH approached the pHIEP, both in the absence and presence of arsenic. There is no evidence that the presence of either arsenate or DMA affects effective particle size other than by changing the particle charge. Effective particle sizes are not influenced by DMA, consistent with the effect of DMA on the ζ-potential of goethite. However, the effective particle sizes are influenced by the concentration of inert electrolyte (Table 1), with larger aggregated particles in the higher concentra(31) Beatriz, B. C.; Tejedor-Tejedor, M. I.; Anderson, M. A. Langmuir 1999, 15, 2316. (32) Goldberg, S.; Johnson, C. T. J. Colloid Interface Sci. 2001, 234, 204. (33) Li, L.; Stanforth, R. J. Colloid Interface Sci. 2000, 230, 12. (34) Malotky, D. T.; Anderson, M. A. In Colloid and Interface Science: Proceedings of the International Conference on Colloids and Surfaces; Kerker, M., Ed.; Academic: New York, 1976; Vol. 4, p 281. (35) Manning, B. A.; Goldberg, S. Soil Sci. 1997, 162, 886.

Figure 6. Adsorption kinetics of arsenate on goethite at pH 4.0 in the presence of three inert electrolyte concentrations as described by a plot of Γ vs ln t. The goethite concentration is 1.98 g L-1, and the initial arsenate concentration is 120 µM.

tions of inert electrolytes at a given pH. After ultrasonication, the aggregates are dispersed and thus the effective particle size decreases (Figure 4). Effect of Solids Concentration. Adsorption kinetics and ζ-potential were measured at different solids concentration. A solids concentration of 1.98 g L-1 was used in the kinetics experiments in facilitate adsorption measurement using a loss-from-solution method. A lower solids concentration (0.25 g L-1) was used for ζ-potential measurement due to sample requirements for the instrument. A comparison of the adsorption and desorption kinetics of both arsenate and DMA at the two solids concentrations shows that the difference in solids concentration had no effect on the measured values during either adsorption or desorption (Figure 5). Effect of Inert Electrolyte Concentration on Adsorption Kinetics. Inert electrolyte concentration has little effect on the adsorption rate and surface coverage of arsenate at the pH 4.0 (Figure 6). DMA adsorption at pH 4.0 approached a plateau after 14 h (ln t ) 2.64), with adsorption approaching the plateau more quickly at lower inert electrolyte concentrations (Figure 7). The surface

Arsenic Species and Goethite Adsorption Kinetics

Figure 7. Adsorption kinetics of DMA on goethite at pH 4.0 in the presence of three inert electrolyte concentrations as described by a plot of Γ vs ln t. The lines show the Elovich portions of the results. The goethite concentration is 1.98 g L-1, and the initial DMA concentration is 130 µM.

coverage approaches the same value for the three inert electrolyte concentrations tested at the end of run. This result suggests that both arsenate and DMA form innersphere complexes on goethite, which is consistent with the findings of Fendorf et al.,36 Grossl et al.,12 and Sun and Doner.4 Elovich analysis (Γ vs ln t plots) of the kinetics results for arsenate show that arsenate adsorption closely follows an Elovich equation (Figure 6), consistent with the findings of Zhao and Stanforth.13 The tailing off at the longer time periods is due to the almost complete removal of arsenate from solution (complete removal gives a surface coverage of 0.061 mmol g-1) rather than of a deviation from Elovich equation. In contrast, the DMA has a more complex kinetics pattern (Figure 7). Adsorption is still a two-phase process, with a rapid step during which much of the adsorption occurs, followed by a slower stage lasting throughout the experiment time period. However, the adsorption kinetics are not linear on an Elovich plot throughout the time period measured, although a portion of the results appear to be linear. Effect of pH on Adsorption Kinetics. The effect of pH on the adsorption kinetics of arsenate is presented in Figure 8. The Γ vs ln t plots are linear and parallel at all three pH values (Table 2). The slow stage of arsenate adsorption was independent of pH, while the amount adsorbed during the initial stage (before the first measurement) increased with deceasing pH. These results are consistent with previous findings.13,37 At 6.75 the adsorption of DMA with an initial concentration of 130 µM and in 0.1 M NaNO3 approached steady state in about 5 h (ln t ) 1.6), whereas steady state was not attained at pH 4.0 throughout the run (Figure 9). Rate-Determining Step during Adsorption. At pH 4.0 and an initial concentration of 120 µM, arsenate adsorption on goethite at three inert electrolyte concentrations is initially very rapid, followed by a slow stage. DMA adsorption on goethite with an initial concentration of 130 µM at pH 4.0 also follows a similar pattern. This biphasic kinetics has also been observed for anions such (36) Fendorf, S.; Eick, M. J.; Grossl, P. R.; Sparks, D. L. Environ. Sci. Technol. 1997, 31, 315. (37) Ler, A.; Stanforth, R. Environ. Sci. Technol. 2003, 37, 2694.

Langmuir, Vol. 21, No. 7, 2005 2899

Figure 8. Adsorption kinetics of arsenate on goethite in the presence of 0.1 M NaNO3 as described by a plot of Γ vs ln t. The goethite concentration is 1.98 g L-1, and the initial arsenate concentration is 120 µM. Table 2. Linear Fitting Results of Adsorption Kinetics of Arsenate at Three pH Values in the Presence of 0.1 M NaNO3a pH

intercept (mmol g-1)

slope (µmol g-1/ (ln time (h)))

R2

4.00 7.00 10.0

0.053 0.037 0.022

3.2 3.1 3.1

0.99 0.99 0.99

a

Results are shown in Figure 8.

Figure 9. Adsorption kinetics of DMA on goethite in the presence of 0.1 M NaNO3 as described by a plot of Γ vs ln t. The goethite concentration is 1.98 g L-1, and the initial DMA concentration is 130 µM.

as arsenic, phosphate, and silica as well as trace elements.6-9 The kinetics results presented here can be used to elucidate the type of controlling mechanism for the slow adsorption step. If the interparticle diffusion is the ratecontrolling step, then the adsorption patterns should be sensitive to inert electrolyte concentrations or pH. The effective particle size of goethite increases with an increasing inert electrolyte concentration due to double layer compression and the resulting coagulation of the colloidal particles. At pH 4.0 the effective particle size of 0.25 g L-1 goethite increases from 692 nm, to 815 nm, to

2900

Langmuir, Vol. 21, No. 7, 2005

1713 nm in 0.001 M, 0.01 M, and 0.1 M NaNO3 respectively. As pH approaches pHIEP, the goethite particles aggregate, and this results in an increase of the effective particle sizes (Figure 4). Increasing the effective particle size will increase the diffusion path length from the external surface of the aggregate to reactive sites located on the internal surface of aggregate, so it will take a longer time for the ions to reach these reactive sites from the external solution. In addition, the total external surface area of the aggregate per unit volume deceases with increasing effective particle size, although the total internal surface area of aggregates remains the same or reduces slightly. The inert electrolyte concentration has no significant effect on arsenate adsorption kinetics (Figure 6), indicating that effective particle size does not influence the adsorption kinetics. Likewise, the effective particle sizes are different at the three pH values in the presence of the same initial arsenate concentration of 120 µM, since ζ-potential of goethite changes with pH at the same initial arsenate concentration (Figure 3). However, Γ vs ln t plots for arsenate adsorption are parallel straight lines with different intercepts at three pH values (Figure 8, Table 2), showing that the rate for the slow adsorption step is the same at the different pH values and effective particle sizes. The rate was not influenced by effective particle size or diffusive path lengths and hence cannot be interparticle (or interaggregate) diffusion controlled. Likewise, the kinetics patterns of DMA adsorption in the presence of 0.01 and 0.1 M were not significantly influenced by the inert electrolyte concentrations (Figure 7). At pH 6.75 the effective particle sizes of goethite in the presence of DMA are larger than at pH 4.0 since the pH is close to pHIEP (Figure 4, Table 1). However, adsorption of DMA approached steady state in about 5 h at pH 6.75, while steady state was not attained at pH 4.0 throughout the run (Figure 9). Interparticle diffusion cannot account for the different kinetics patterns for DMA at pH 4.0 and 6.75. On the basis of the above arguments, interparticle or external diffusion is not the ratedetermining step. One sample was agitated using ultrasonic mixing rather than a magnetic stirrer. Ultrasonic mixing provides high agitation energy to the sample and should disperse any loosely formed aggregates in the solution, such as would be expected as the particles coalesce due to a lower surface charge. The effective particle size decreased with ultrasonication, especially near pHIEP (Figure 4). If the interparticle diffusion is the rate-controlling step, then the adsorption rate should be faster using ultrasonication than using magnetic stirring. The adsorption kinetics of arsenate with ultrasonication and magnetic stirring are identical (Figure 10), further demonstrating that interparticle diffusion is not the rate-determining steps during arsenate adsorption. Desorption was used to determine whether the slow adsorption step is due to intraparticle diffusion, i.e., diffusion into pores in the solid or into solid itself. The presence and shape of the pores depends on the crystal properties and does not change with pH or inert electrolyte concentrations. If intraparticle diffusion controlled the slow adsorption step and desorption process, then diffusion of the ions out of the pores or solids during desorption should follow the same two-stage pattern as the diffusion into the pores or solids during adsorption. The desorption results are inconsistent with an intraparticle diffusion controlled process. DMA desorption occurred rapidly after the pH is raised to strong alkaline values (pH 10.0 and 12.0), whereas the adsorption kinetics shows a slow stage

Zhang and Stanforth

Figure 10. Effect of mixing methods (ultrasonication vs magnetic stirring) on the adsorption kinetics of arsenate at pH 4.20 in the presence of 0.001 M NaNO3. The goethite concentration is 1.98 g L-1, and the initial concentration of arsenate is 120 µM.

Figure 11. Adsorption (at pH 4.0) and desorption kinetics of arsenate at pH 12.0 and DMA at pH 10.0 (b) and 12.0 (2) in the presence of 0.1 M NaNO3.The goethite concentration is 1.98 g L-1, and the inert electrolyte concentration is 0.1 M NaNO3. The initial concentration of arsenate is 120 µM, and DMA is 130 µM.

during the first five 5 h adsorption (Figures 7, 11). In contrast with the DMA results, arsenate concentrations increased slowly with time during desorption with much of the arsenate not desorbing (Figure 11) and did not reach equilibrium at the end of the run. If intraparticle diffusion were important, the DMA desorption should show a slow stage, similar to that observed for the arsenate desorption. Since this was not observed, these results demonstrate that diffusion into the crystal is not the rate-controlled step for DMA or arsenate adsorption. The DMA desorption results are also inconsistent with interparticle diffusion. The effective particle size of goethite in the presence of DMA at pH 10.0 and 12.0 is larger than that at pH 4.00. If interparticle diffusion is the rate-determining step during DMA adsorption and desorption, the desorption of DMA at pH 10.0 or 12.0 should be slower than the adsorption at pH 4.00, in contrast with the results observed.

Arsenic Species and Goethite Adsorption Kinetics

Conclusions The slow stage of arsenate adsorption on hydrous metal oxides frequently follows an Elovich equation. This equation can be derived by assuming kinetics control from either diffusion or a heterogeneous surface reaction. This study shows that the slow reaction giving rise to the Elovich portion of arsenate adsorption on goethite is not due to diffusion, and thus is most likely due to the heterogeneity of the surface site bonding energy or to other reactions occurring on the surface. The effective particle size increases with inert electrolyte concentrations and as the surface charge approaches zero (i.e. at pH close to pHIEP). However, the slow step in the arsenate adsorption kinetics is not affected by inert electrolyte concentrations at pH 4.0 or by pH in the presence of 0.1 M NaNO3. These results indicate that the adsorption during the slow stage of arsenate adsorption on goethite in not controlled by effective particle size and, hence, interparticle diffusion.

Langmuir, Vol. 21, No. 7, 2005 2901

Desorption of DMA was very rapid at pH 12.0, in contrast to the slow adsorption process at pH 4.0. If intraparticle diffusion is the controlling step, desorption should follow a pattern similar to the two-stage adsorption pattern since the crystal sizes at two pH values are roughly the same. The rapid desorption therefore indicates that intraparticle diffusion is not the controlling step for DMA adsorption. If both intraparticle and interparticle diffusion do not control the reaction kinetics during the slow adsorption step, then by elimination the reaction must be controlled by surface heterogeneity or by reactions other than surface complexation influencing the removal of arsenic from solution. Acknowledgment. The support of the National University of Singapore for this research is gratefully acknowledged. LA047636E