pubs.acs.org/Langmuir © 2009 American Chemical Society
Organic Arsenic Adsorption onto a Magnetic Sorbent Soh-Fong Lim,†,‡ Yu-Ming Zheng,† and J. Paul Chen*,† †
Division of Environmental Science and Engineering, and ‡Department of Chemical and Biomolecular Engineering, National University of Singapore, 10 Kent Ridge Crescent, Singapore 119260 Received September 10, 2008. Revised Manuscript Received February 15, 2009
The adsorption of organic arsenate, monomethylarsonate (MMA), onto a calcium alginate encapsulated magnetic sorbent is studied in this paper. A novel alginate encalsulated magnetic sorbent was used in the experiments on adsorption isotherm, kinetics, and pH effect. It was found that the equilibrium sorption can be attained within 25 h. Solution pH plays a key role in the removal of MMA from the solution. A greater adsorption can be achieved at pH 4 and below. The maximum sorption capacity of MMA was 8.57 mg As/g, which is slightly higher than the reported adsorbents. The interaction characteristics between the organic arsenate and magnetic sorbent were elucidated by applying FT-IR and XPS analyses. It is shown that the -COOH and Fe-O groups in the sorbent are involved in the adsorption process. The appearance of As-CH3 and alkane C-H groups in the FT-IR spectrum reveals the binding of the organic arsenate to the sorbent. The XPS analysis indicates that reduction of organic arsenate to organic arsenite on the sorbent’s surface happens through solid state redox reaction via charge transport from Fe(II) and C-O species in the sorbent. The XPS results also show the disappearance of C-OH and formation of As-O. It is deduced from the spectral results that mechanisms of organic arsenate adsorption involve C-OH, As-O, and Fe-O groups with the solid state redox process.
Introduction Arsenic contamination has become one of the global environmental issues.1-4 It has posed a series of severe health problems. As a result, the US Environmental Protection Agency (USEPA) and World Health Organization (WHO) implemented the reduction of arsenic maximum contaminant level (MCL) in the drinking water from 50 to 10 μg/L. The implementation of the new MCL arsenic standard in drinking water has initiated numerous studies throughout the world in order to meet the more stringent requirements. Depending on solution chemistry (e.g., pH, competitive factors, and redox potential), arsenic in natural waters and soils is in the forms of arsenic oxyanions such as arsenite and arsenate and the methylated arsenate species of monomethylarsonate (MMA) and dimethylarsinate (DMA). The inorganic arsenic species are predominant; however, the importance of organic arsenic (e.g., MMA) cannot be ignored. The methylation of arsenic can be carried out by a variety of organisms ranging from bacteria to fungi to mammals;5 anthropogenic input of organic arsenic compounds arose since the 1970s due to their increasing usage as pesticides and herbicides worldwide.6 US statistics shows that between 2 and 4 million pounds of the sodium salt of MMA were used in the US by industrial, commercial, and government sectors in 1999.7 High concentrations of organic arsenic have been *Corresponding author: e-mail
[email protected], jchen.enve97@ gtalumni.org; Fax +65-6872-5483, +1-831-303-8636. (1) Mahuli, S.; Agnihotri, R.; Chauk, S.; Ghosh-Dastidar, A.; Fan, L. S. Environ. Sci. Technol. 1997, 31(11), 3226–3231. (2) Kanel, S. R.; Manning, B.; Charlet, L.; Choi, H. Environ. Sci. Technol. 2005, 39(5), 1291–1298. (3) Zhang, G. S.; Qu, J. H.; Liu, H. J.; Liu, R. P.; Li, G. T. Environ. Sci. Technol. 2007, 41(13), 4613–4619. (4) Sarkar, S.; Blaney, L. M.; Gupta, A.; Ghosh, D.; SenGupta, A. K. Environ. Sci. Technol. 2008, 42(12), 4268–4273. (5) Cullen, W. R.; Reimer, K. J. Chem. Rev. 1989, 89, 713–764. (6) Solo-Gabriele, H.; Sakura-Lemessy, D.; Townsend, T.; Dubey, B.; Jambeck, J. Quantities of Arsenic Within the State of Florida; Florida Center for Solid and Hazardous Waste Management: Gainesville, FL, 2003. (7) Xu, Z. H.; Jing, C. Y.; Li, F. S.; Meng, X. G. Environ. Sci. Technol. 2008, 42, 2349–2354.
Langmuir 2009, 25(9), 4973–4978
detected in groundwater and surface water. Anderson and Bruland reported that methylated arsenic species, predominantly DMA, were on an average of 24% of the total dissolved arsenic in the lakes surveyed in California.8 The organic arsenic concentration was 10% of the total arsenic in the groundwater in the northeastern United States.9 Like inorganic arsenic, organoarsenic is highly carcinogenic. Kenyon and Hughes found that it has multiorgan tumor promoting activities.10 Lafferty and Loeppert reported that some methyl arsenic species can be more toxic than inorganic species.11 There are numerous studies on the removal of inorganic arsenic from aqueous environments. However, not much emphasis has been given to the decontamination of organic arsenic species.12 Thirunavukkarasu et al. used few sorbents to remove DMA.13 An arsenic removal capacity of 8 μg/g was reportedly achieved when an iron oxide-coated sand was used. A bed volume of 585 BV was reported when a Fe3+-based ion-exchange resin was employed. Xu et al. used nanocrystalline titanium dioxide for photocatalytical degradation of MMA and DMA and subsequent removal of the converted inorganic As(V).7 A two-level seven-factor fractional factorial design analysis was conducted by Pokhrel and Viraraghavan to study seven parameters influencing DMA uptake by an iron oxide-coated A. niger biomass.14 The important factors on DMA removal include the presence of calcium ions in solution, the DMA concentration, and the temperature. The MMA and DMA adsorption onto the hydrous ferric oxide and activated alumina decreased with an increasing pH while it was weakly dependent on ionic strength.15 (8) Anderson, L. C. D.; Bruland, K. D. Environ. Sci. Technol. 1991, 25(3), 420– 429. (9) Banerjee, K.; Helwick, R. P.; Gupta, S. Environ. Prog. 1999, 18(4), 280–284. (10) Kenyon, E. M.; Hughes, M. F. Toxicology 2001, 160(1-3), 227–236. (11) Lafferty, B. J.; Loeppert, R. G. Environ. Sci. Technol. 2005, 39, 2120–2127. (12) Zhang, J. S.; Stanforth, R. S.; Pehkonen, S. O. J. Colloid Interface Sci. 2007, 306(1), 16–21. (13) Thirunavukkarasu, O. S.; Viraraghavan, T.; Subramanian, K. S.; Tanjore, S. Urban Water 2002, 4, 415–421. (14) Pokhrel, D.; Viraraghavan, T. Chem. Eng. J. 2008, 140(1-3), 165–172. (15) Cox, C. D.; Ghosh, M. M. Water Res. 1994, 28(5), 1181–1188.
Published on Web 3/26/2009
DOI: 10.1021/la802974x
4973
Article
Lim et al.
Figure 1. Adsorption of MMA onto the alginate encapsulated magnetic sorbent: (a) kinetics; (b) isotherm. Experimental conditions: pH = 3-4, m = 0.5 g/L, T = 293 K.
Surface complex formation model (SCFM) was used to elucidate the mechanisms of organic arsenic sorption. Cox and Ghosh studied the adsorption of MMA and DMA on hydrous ferric oxide and activated alumina by the triple-layer SCFM.15 The model well described the effect of pH on the adsorption. Jing et al. studied the adsorption mechanisms of MMA and DMA on the nanocrystalline titanium oxide.16 A satisfactory interpretation of the experimental data was reportedly achieved by the charge distribution multisite complexation model. In this study, we investigated the adsorption of organic arsenate by the alginate encapsulated magnetic sorbent recently developed in our laboratory.17 As this sorbent has both magnetite and alginate, it demonstrates multifunctional adsorption properties for both cationic and anionic inorganic contaminants. The investigation was carried out through a series of batch sorption experiments to better understand the adsorption behavior. Fourier transform infrared spectroscopy (FT-IR) and X-ray photoelectron spectroscopy (XPS) were employed to probe the interactions of the sorbent and organic arsenate. The objective of the study was to obtain understanding of the adsorption mechanism for the organic arsenic uptake.
Materials and Methods Materials. Disodium methyl arsenate (CH3AsNa2O3), one of the monomethylarsonate species, was purchased from Chem Service (USA); its purity is 98%. Disodium hydrogen arsenate heptahydrate was purchased from Fluka (Switzerland). Nitric (16) Jing, C. Y.; Meng, X. G.; Liu, S. Q.; Baidas, S.; Patraju, R.; Christodoulatos, C.; Korfiatis, G. P. J. Colloid Interface Sci. 2005, 290(1), 14–21. (17) Lim, S. F.; Chen, J. P. Appl. Surf. Sci. 2006, 253, 5772-5775.
4974
DOI: 10.1021/la802974x
Figure 2. (a) MMA adsorption per unit mass sorbent as a function of pH for two different amounts of initial arsenic concentrations. (b) Effect of MMA sorption using calcium alginate magnetic sorbent on solution pH. Experimental conditions: m = 0.5 g/L, T = 293 K. acid and sodium hydroxide from Sigma-Aldrich (Singapore) were used to adjust the pH of the solutions. Sodium alginate and calcium chloride dehydrate from Fluka (Switzerland) were used in the preparation of alginate pellet. All chemicals used are of analytical grades. The magnetic sorbent in this study is composed of magnetite particles encapsulated by calcium alginate. The magnetite was first synthesized by a coprecipitation approach. The sodium alginate solution with the magnetite dripped through a needle connected to a high-voltage power generator. The size of the droplets can be controlled by the voltage. The droplets with desired size dripped into a calcium chloride solution, and the beads can then immediately be formed. The solidified beads were washed, dried by a freeze-dryer, and stored. Calcium alginate pellet was also prepared by the same approach, except that the magnetite was not added. Adsorption Experiments. Batch sorption experiments were conducted by adding a dosage of 0.5 g/L calcium alginate magnetic sorbent in a conical flask with MMA solution of a known initial concentration. In the isotherm experiment, the solution pH was controlled at 3-4. The mixtures in the flasks were shaken on a rotary shaker for 48 h at 293 K to reach the equilibrium. The samples were taken at the end of the experiments for analysis of the arsenic concentration by an inductively coupled plasma emission spectrometer (ICP-OES; Perkin-Elmer Optima 3000). Calcium alginate pellet was tested for its uptake capacity for the MMA. The procedure is the same as the above. In order to compare the uptake of organic arsenic with that of inorganic arsenic, an adsorption isotherm experiment was conducted. The procedure was the same as the above with an exception that a disodium hydrogen arsenate heptahydrate was used. Langmuir 2009, 25(9), 4973–4978
Lim et al.
Article
Figure 5. As 3d XPS spectra of organic As loaded magnetic Figure 3. FT-IR spectra for (a) fresh alginate encapsulated mag-
sorbent.
netic sorbent and (b) organic arsenate loaded alginate encapsulated magnetic sorbent.
Figure 6. Fe 2p XPS spectra of (a) the fresh magnetic sorbent and (b) organic arsenate loaded magnetic sorbent at pH 3.
Figure 4. XPS wide scan spectra of the (a) fresh sorbent and (b) organic arsenate loaded sorbent.
In the pH effect study, HNO3 or NaOH was added to adjust the initial pH. The magnetic sorbent was added into MMA solution with a known initial concentration. The mixtures were shaken at 293 K for 48 h. The samples were collected and measured by the ICP-OES. In the kinetics experiments, the MMA solution with a known initial concentration and a fixed pH was prepared. The magnetic sorbent was then added to the solution. The mixture solution was shaken at 293 K. The samples were extracted at appropriate time intervals and analyzed for arsenic concentration by the ICP-OES. Fourier Transform Infrared Spectroscopy. FT-IR spectroscopy was used to determine the vibration frequency changes in the sorbents. Virgin and MMA loaded sorbents were analyzed. Each sample was mixed with pure potassium bromide which acts as background and then grounded in an agate mortar. The resulting mixture was pressed at 10 tons for 5 min to form pellet. Langmuir 2009, 25(9), 4973–4978
The pellets were characterized with infrared transmission spectra using a FTS-135 (Bio-Rad) spectrometer. The spectra were collected within the range of 400 and 4000 cm-1. All the spectra were recorded and plotted in the same scale on the transmittance axis. X-ray Photoelectron Spectroscopy. The chemical analyses on the virgin and MMA loaded sorbents were conducted by XPS (Kratos XPS System-AXIS His-165 Ultra, Shimadzu, Japan). All the samples were dried for 12 h at 50 °C prior to the XPS measurement. The XPS spectra were obtained with monochromatized Al KR X-ray source (1486.71 eV) working at 150 W, 15 kV, and 10 mA and base pressure of 3 10-8 Torr in the analytical chamber. For wide scan spectra, an energy range of 0-1100 eV was used with pass energy of 80 eV and step size of 1 eV. The high-resolution scans were conducted according to the peak being examined with pass energy of 40 eV and step size of 0.05 eV. The carbon 1s electron binding energy corresponding to graphitic carbon at 284.5 eV was used as reference for calibration purposes. The XPS results were collected in binding energy forms and fitted using a nonlinear least-squares curve-fitting program DOI: 10.1021/la802974x
4975
Article
Lim et al.
Figure 7. C 1s XPS spectra of (a) the fresh magnetic sorbent and (b) organic arsenate loaded magnetic sorbent at pH 3. (XPSPEAK41 software). The XPS spectra were deconvoluted with the subtraction of linear or Shirley background. Linear type background was used for nonmetal elements, while Shirley type background was applied for transitions metals. The line width (full width at half-maximum) of those spectra was fixed constant between 1 and 2. The percentage of Lorentzian-Gaussian was set at 0 for all the deconvoluted spectra. Surface compositions for each element were determined from the deconvoluted peaks area ratios.
Results and Discussion Adsorption Study. In the batch kinetics experiments, two experiments with different initial MMA concentrations under the same physical conditions have been carried out. Figure 1a shows a plot of sorption capacity (q) vs time. It can be found that most of MMA uptake rapidly occurs in the first 9 h, followed by a relatively slow process. The adsorption equilibrium can be reached in 25 h. The adsorption isotherm was investigated over a range from 0 to 45 mg MMA/L using 0.5 g/L dose of the sorbent, with the result given in Figure 1b. Langmuir and Freundlich isotherms are commonly used in description of adsorption performance. Both isotherms were used to model the experimental data. It was found that the data can better be fit by the Langmuir equation q bCeq given as follows q ¼ 1max þbCeq , where qmax is the maximum adsorption capacity, Ceq is the equilibrium concentration in the solution, and b is the adsorption affinity constant related to the binding energy of adsorption. The maximum adsorption capacity of 8.57 mg As/g and the adsorption affinity constant of 0.44 L/mg are determined. The equation satisfactorily fits the isotherm data with a correlation coefficient of 0.99. This indicates that monolayer adsorption exists in the uptake of arsenic. The organic arsenic uptake capacity is higher than other sorbents reported in the literatures 4976
DOI: 10.1021/la802974x
Figure 8. O 1s spectra of (a) the fresh magnetic sorbent and (b) organic arsenate loaded sorbent magnetic sorbent at pH 3.
(e.g., refs 13 and 15). This demonstrates that the sorbent has a greater affinity for the MMA. The maximum adsorption capacity of 11 mg As/g for inorganic As(V) was found in our adsorption isotherm experiment (shown in Figure 1b). This indicates that the magnetic sorbent can remove more inorganic arsenic than organic arsenic. Adsorption equilibrium study by calcium alginate pellet was conducted in order to determine the importance of magnetite and alginate. It was found that the adsorption by the calcium alginate was less than 5% of that by the magnetic sorbent. This indicates that the arsenic adsorption onto the magnetic sorbent is dominant by the magnetite. The pH effect experiments were carried out with two different initial concentrations of MMA: 1 and 3 mg As/L. Figure 2a demonstrates that the better MMA uptake can be achieved at the final pH < 4. The uptake becomes very low at pH > 5. Thus, the magnetic sorbent should be applied at pH < 4 to achieve a greater removal. As shown in Figure 2b, the final pH increases slightly at the initial pH is increased from 2 to 5, and it does not change in the initial pH 5-7; it decreases in the initial pH 7-10. These observations might be due to some buffering during the MMA uptake process. A similar trend was observed for inorganic arsenic (arsenate) sorption on the same sorbent.17 Weak acid type organic functional groups (from the alginate) such as the carboxyl group in the sorbent may provide the buffering to the adsorption system. Fourier Transform Infrared Spectroscopy Analysis. Figure 3 reveals the changes in absorption bands of the surface functional groups of the magnetic sorbent due to the uptake of organic arsenate. The spectrum before sorption (sample a) shows dual adsorption bonds at 1600 and 1420 cm-1 for CdO and C-O in carboxyl group (-COOH), respectively. The bond Langmuir 2009, 25(9), 4973–4978
Lim et al.
Article Table 1. Binding Energies and Relative Contents of As O, and Fe in Adsorbentsa
valence state
sample
assignment
binding energy (eV)
intensity (counts s-1)
relative content (%)
As(III) 45.2 139.5 As(V) 46.7 25.6 O 1s MS metal oxide 529.4 512.7 CdO 531.0 1533.9 C-OH, C-O-C 532.1 416.6 metal oxides 529.4 642.2 MS-AsORG CdO 531.1 1682.7 C-OH, C-O-C 532.1 145.4 C 1s MS C-C 284.5 543.6 C-OH, C-O-C 286.2 523.3 CdO 287.9 301.0 C-C 284.5 675.2 MS-AsORG C-OH, C-O-C 286.2 285.0 CdO 287.9 427.2 Fe 2p MS Fe(II) oct 709.4 534.6 Fe(III) oct 711.3 480.4 Fe(III) tet 713.6 161.5 Fe(II) oct 709.7 442.0 MS-AsORG Fe(III) oct 711.6 224.9 Fe(III) tet 714.4 603.0 a MS: magnetic sorbent; MS-AsORG: organic arsenate loaded magnetic sorbent. Conditions: m = 0.5 g L-1, pH = 3, T = 293 K. As 3d
MS-AsORG
at 1600 cm-1 is sharper than that after the sorption (wavenumber of 1612 cm-1 in sample b), and the shape of the peak at 1420 cm-1 (sample a) is also different from that after organic arsenate adsorption (wavenumber of 1386 cm-1 in sample b). Another obvious feature is that the spectra before and after arsenic sorption experience minor shifts in the frequencies of the absorption band of both CdO and C-O in carboxyl group (-COOH). Both bands observed shift to different extents. The C-O bond shifts to the lower frequency (from 1420 to 1386 cm-1), which could be attributed to the associations of the hydroxyl group with arsenic ions. The CdO bond shifts to the higher frequency (from 1600 to 1612 cm-1); this may result from high electron density induced by the sorption of arsenic onto the adjacent hydroxyl group. These changes indicate the carboxyl groups in the sorbent are involved in the arsenic uptake. Arsenic in monomethylarsonate (MMA) contains three different chemical bonds: As-CH3, AsdO, and As-O. The appearance of an additional peak at 1258 cm-1 (sample b) shows an evidence of sorption of As-CH3 functional group onto the sorbent.18 The spectrum after sorption (sample b) shows a Fe-O broad peak at around 580 cm-1 is sharper than that before sorption (sample a). The significant change could be due to the complexes of organic arsenate with the sorbent’s surface sites. The carbonoxygen (C-O) stretching vibration peak at about 1030 cm-1 remains the same after adsorption. X-ray Photoelectron Spectroscopy Analysis. XPS wide scan spectra of the virgin and organic arsenate loaded sorbents are illustrated in Figure 4. Three major peaks at binding energies of 284, 346, and 531 eV, designated for the C 1s, Ca 2p, and O 1s, respectively, are observed for the virgin sorbent (Figure 4a).19 Significant changes can be found in Figure 4b after organic arsenate adsorption; the peak at binding energy of 346 eV for Ca 2p disappears while a new weak peak at binding energy of about 45 eV for As 3d appears. High-resolution spectra of As 3d, Fe 2p, C 1s, and O 1s regions are shown in Figures 5-8. The As 3d spectrum of the organic arsenate loaded sorbent can be deconvoluted into two (18) Clothup, N. B.; Daly, L. H.; Wiberley, S. E. Introduction to Infrared and Raman Spectroscopy, 3rd ed.; Academic Press: London, 1990. (19) Wagner, C. D.; Riggs, W. M.; Davis, L. E.; Moulder, J. F.; Muilenberg, G. E. Handbook of X-ray Photoelectron Spectroscopy; Physical Electronics Division, Perkin-Elmer Corp.: Eden Prairie, MN, 1979.
Langmuir 2009, 25(9), 4973–4978
84.5 15.5 20.8 62.3 16.9 26.0 68.1 5.9 39.7 38.3 22.0 48.7 20.5 30.8 45.5 40.8 13.7 34.8 17.7 47.5
individual component peaks, which come from the different valent arsenic atom and overlap on each other. As shown in Figure 5, the peaks at binding energies of 45.2 and 46.7 eV can be assigned to the arsenite (As(III)) and arsenate (As(V)) atom, respectively. These two assignments reflect the different chemical valence of organic arsenic on the sorbent. The quantitative analysis shows that 84.5% of arsenic existing as As(III) is detected on the sorbent’s surface as demonstrated in Table 1. In addition, no arsenite was detected in the aqueous solution. This result indicates solid state reduction of arsenate to arsenite on the surface of the sorbent. Furthermore, it can be observed in Figure 6 that the deconvolution of the Fe 2p spectra of the samples produces octahedral ferrous (Fe(II)oct), octahedral ferric (Fe(III)oct), and tetrahedral ferric (Fe(III)tet) peaks with the binding energies of 709.4, 711.3, and 713.6 eV (for virgin sorbent) and 709.7, 711.6, and 714.4 eV (for organic arsenate loaded sorbent). These values are close to the values previously reported for magnetite.20,21 Changes in the relative abundance of Fe(II) and Fe(III) upon organic arsenate sorption process are quantitatively elucidated as indicated in Table 1. It shows that relative content of the Fe(II)oct decreases from 45.5 to 34.8% while that of Fe(III)tet increases significantly from 13.7 to 47.5%. These changes indicate the oxidation of Fe (II) to Fe(III). As shown in Figure 7, the C 1s spectra can be deconvoluted into three peaks representing three functional groups of C-C, C-O, and CdO at binding energies of 284.5, 286.2, and 287.9 eV, respectively. Table 1 demonstrates that the C-O content (C-OH and C-O-C) decreases from 38.3 to 20.5% while that of CdO increases from 22.0 to 30.8% due to the arsenic uptake, indicating the oxidation of C-O to CdO. This change reveals involvement of alginate as a reducing agent in the reduction of organic arsenate. The deconvolution of the O 1s spectra of the sample yields three individual component peaks demonstrated in Figure 8, which represent different groups. The peak at binding energy of 529.4 eV (for virgin sorbent) can be assigned to Fe-O (lattice oxygen in magnetite). The peaks with binding energies of 531 and 532.1 eV (for virgin sorbent) as well as 531.1 and 532.1 eV (20) Fujii, T.; Groot, F. M. F.; Sawatzky, G. A. Phys. Rev. B 1999, 59(4), 3195– 3202. (21) Scott, T. B.; Allen, G. C.; Heard, P. J.; Randell, M. G. Geochim. Cosmochim. Acta 2005, 69(24), 5639–5646.
DOI: 10.1021/la802974x
4977
Article
Lim et al.
Figure 9. Schematic diagram of organic arsenate adsorption and reduction process by magnetic sorbent.
(for organic arsenate loaded sorbent) can be assign to CdO (carboxyl and/or quinone groups) and C-O (ether and/or alcohol), respectively, which reflect the functional groups of alginate in the sorbent. Alginate is a polysaccharide-like polymer and mainly contains three different functional groups: COOH (carboxyl), -C-O-C- (ether), and -OH (hydroxyl).22 Compared with the virgin sorbent (Figure 8a), metal oxide spectrum of the organic arsenate loaded sorbent (Figure 8b) can be further deconvoluted into two individual component peaks: Fe-O and As-O. The spectrum after the adsorption (Figure 8b) shows that the Fe-O peak changes slightly. Table 1 shows that the metal oxide content increases from 20.8 to 26.0%, indicating the formation of As-O, which is due to the binding of organic arsenate onto the oxygen atom in the sorbent. Furthermore, the oxygen in form of C-OH/C-O-C group decreases significantly from 16.9 to 5.9% after the adsorption. Apparently, these indicate the binding of organic arsenate to the sorbent, which results in the disappearance of C-OH and formation of As-O. Thus, the XPS results suggest that the arsenate sorption requires the involvement of C-OH, As-O, and Fe-O groups. On the basis of the FT-IR and XPS analysis, the mechanism of organic arsenate removal via magnetic sorbent is proposed in Figure 9. The removal of organic arsenate by the sorbent is a complicated process, in which adsorption and redox processes are involved.
Monomethylarsonic acid (H2AsO3CH3) has pKa values of 4.19 and 8.77.15 The equilibrium pH in this study ranged from 1.8 to 7.7, as shown in Figure 2. Thus, the arsenic mainly existed in the form of HAsO3CH3 . First, the functional groups on the sorbent are protonated in order to form surface complexes with the anionic arsenic. The magnetite behaves as a weak acid.23 It plays a key role in the sorption. The surface complex formation shown in Figure 9 explains the pH effect on the adsorption (Figure 2a). When pH is lower, the functional groups are greater protonated, which is more beneficial to the adsorption. Second, redox on the solid state of sorbent occurs during the uptake of organic arsenate, namely As(V) being reduced while Fe(II) being oxidized. The reduction of the As(V) might happen through the arsenate sorption to the sorbent’s surface followed by reduction of the surface bound arsenate species to arsenite via charge transport from Fe(II) species in the sorbent, which is similar to the case of uranium reduction observed by Scott and co-workers.21 Some of the electrons from carbons may also contribute to the reduction of arsenic.
(22) Chen, J. P.; Lie, D.; Wang, L.; Wu, S. N.; Zhang, B. P. J. Chem. Technol. Biotechnol. 2002, 77, 657–662.
(23) Sun, Z. X.; Su, F. W.; Forsling, W.; Samskog, P. O. J. Colloid Interface Sci. 1998, 197, 151–159.
4978
DOI: 10.1021/la802974x
Acknowledgment. This research was funded by Academic Research Funds (R-288-000-023-112 and R-288-000-017-133) from the National University of Singapore. The PhD scholarship to S.F.L. is awarded by the Tropical Marine Science Institute, Singapore.
Langmuir 2009, 25(9), 4973–4978
