Environ. Sci. Technol. 2009, 43, 2537–2543
Enhanced Arsenic Removal by Hydrothermally Treated Nanocrystalline Mg/Al Layered Double Hydroxide with Nitrate Intercalation K O K - H U I G O H , † T E I K - T H Y E L I M , * ,† A N D ZHILI DONG‡ School of Civil and Environmental Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Republic of Singapore, and School of Materials Science and Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Republic of Singapore
Received October 3, 2008. Revised manuscript received January 29, 2009. Accepted February 2, 2009.
A nanocrystalline Mg/Al layered double hydroxide (FCHTLDH) adsorbent was developed and investigated through stoichiometric calculations, nitrate displacement investigation, comprehensive sorption/desorption experiments, and analyses with XPS, XRD, FTIR, CHNS/O, and EDX for better understanding of the predominant nature of arsenate (As(V)) interaction with FCHT-LDH. FCHT-LDH demonstrated a higher sorption capacity and a faster sorption rate compared to the layered double hydroxides (LDHs) prepared by conventional methods, due to its higher surface area, better porosity characteristics, and nanocrystalline property. These results also indicated the important role of hydrothermal treatment during the synthesis process for enhanced As(V) removal. The observed nitrate-arsenate molar displacement ratio, increased interlayer spacing, and decreased nitrogen content in the interlayer region revealed the predominance of anion exchange mechanism in As(V) sorption by FCHT-LDH. However, a slight pH increase during As(V) sorption equalization and the presence of ca. 25% irreversibly sorbed As(V) signified the occurrence of ligand exchange process as the secondary sorption mechanism. This specific sorption process that possibly involved formation of inner-sphere As(V) complexes with a monodentate mononuclear configuration at the aluminum center, rendered the FCHT-LDH a high affinity for As(V) over nitrate but induced hysteretic sorption/desorption characteristic that limited its regenerated sorption capacity.
Introduction Sorption technology is regarded as one of the most promising technologies for arsenic (As) removal from water because it is cost-effective, versatile, and simple to operate (1). Recently, layered double hydroxides (LDHs) have been investigated for effective oxyanions removal (2-7) owing to their large surface area, high anion exchange capacity, good thermal * Corresponding author tel: +65-6790-6933; fax: +65-6791-0676; e-mail:
[email protected]. † School of Civil and Environmental Engineering. ‡ School of Materials Science and Engineering. 10.1021/es802811n CCC: $40.75
Published on Web 02/25/2009
2009 American Chemical Society
stability, and ability to reconstruct the calcined LDH to its original layered structure through its “memory effect” (8, 9). LDHs, with a general formula [M2+1-xM3+x(OH)2]x+(An-)x/n · mH2O, are mixed metal hydroxides containing positively charged brucite-like sheets with intercalated anions (An-) in the hydrated interlayer regions (8). Among the intercalated anions, nitrate appears to be a suitable candidate according to the sequence of anion exchange capacity (10). LDHs have a relatively weak interlayer bonding and consequently exhibit an excellent ability to exchange their interlayer anion for As in the bulk solution. Most of the LDHs investigated for oxyanions sorption were prepared from conventional coprecipitation methods (2-7) which produced unstable large LDH particles with aggregated crystallites (11). Nanocrystalline materials demonstrate several unique characteristics such as large surface area and high specificity that make them excellent functional materials for water treatment applications. A recent study (12) suggested that nanocrystalline TiO2 exhibited high affinity for As. Therefore, it is of worth to engineer a new breed of nanocrystalline LDH for the removal of As and other oxyanions. Most previous works have suggested that As was removed by LDHs via anion exchange mechanism, but remained hypothetical at best (2, 4, 7). A limitation of these studies was that insufficient experiments were conducted to illustrate the ion exchange stoichiometry for LDH and to provide direct evidence of the occurrence of anion exchange mechanism. It was therefore rather difficult to quantitatively assess the impact of anion exchange property of LDH on As removal. Wang et al. (5) indicated that arsenate (As(V)) sorption by Mg/Al LDH could occur through surface adsorption and interlayer anion exchange depending on the interlayer nitrate orientation. Nonetheless, the predominant mechanism was not clearly identified. Moreover, the quantification of the contributions of the different mechanisms (e.g., anion exchange and adsorption) to As(V) removal will greatly promote the thorough understanding of As interaction with LDH. An extended X-ray absorption fine structure spectroscopic (EXAFS) study performed by Liu et al. (6) reported that As(V) was removed by Li/Al-Cl LDH by reacting with Li cations located in the octahedral holes within Al(OH)3 and Al in the edges of Al(OH)3 layers. They showed As(V) adsorption on the surface of Li/Al-Cl LDH with the formation of inner-sphere bidentate complexes. Conversely, Wang et al. (5) suggested that As(V) was adsorbed on Mg/Al LDH by forming outer-sphere complex. These limited but significant reports indicated that As(V)-LDH surface complex might vary with LDH with different structural aspects and revealed the need for more information of adsorbed complexes via other promising techniques such as X-ray photoelectron spectroscopy (XPS) which has been used to study the effect of surface properties on As(V) adsorption (13, 14). The specific objectives of this study were to (1) assess As(V) binding by nanocrystalline Mg/Al LDH with nitrate intercalation through anion exchange mechanism using stoichiometric calculations, nitrate displacement investigation, and analyses of XPS, XRD, FTIR, CHNS/O, and EDX, (2) unravel the predominant nature of the interactions between the LDH and As(V) and subsequently quantify the contributions of the different mechanisms to As(V) removal using sorption/desorption experiments, and (3) understand the possible dominant As(V)-LDH surface complex based on the XPS analysis. Other objectives included comparison of the sorption behaviors of the nanocrystalline LDH with that of the LDHs prepared by conventional methods, investigation VOL. 43, NO. 7, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2537
TABLE 1. Chemical Composition, Textural Properties, and Lattice Parameters of the LDHs chemical composition 2+
type of LDH IE-LDH COP-LDH CAL-LDH FCHT-LDH FCHT-LDH (As(V)-loaded)
3+
Mg /Al C (wt %) molar ratio 3.1/1 2.1/1 2.4/1 2.0/1 n/a
N (wt %)
9.2 1.4 1.8 7.5 1.2 5.7 0.7-2.5 2.6-3.6 0.7-2.6 0.3-0.6
textural property a
b
lattice parameter
c
As S V D particle sizes crystal ad (Å) ce (Å) (mass %) (m2/g) (cm3/g) (nm) [average size] (nm) sizef (nm) n/a n/a n/a 0 14g
56 22 90 127 n/a
0.04 0.01 0.06 0.31 n/a
3.0 1.4 1.4 8.0 n/a
22-564 [200] [>1000] [>1000] 38-396 [122] n/a
3.04 3.04 n/a 3.04 3.04
22.67 22.91 n/a 23.53 23.90
21.5 12.1 n/a 5.9 9.1
a BET surface area. b BJH pore volume. c BJH pore size. d a ) average metal-metal distance inside the brucite-like sheets ) 2 × d(110). e c ) interlayer distance regulated by the size and charge of the anion placed between the brucite-like sheets ) 3 × d(003). f Average value calculated from fwhm of peak (003) and (006) using Scherrer equation. g Value determined from SEM-EDX analysis; *n/a denotes “data not available”.
of the thermodynamics of As(V) sorption by the LDH, and examination of the reusability of the regenerated LDH for subsequent cycles of As(V) sorption.
Experimental Section Synthesis and Characterization. All the chemicals used are described in the Supporting Information (S1). A nanocrystalline Mg/Al LDH intercalated with nitrate (denoted as FCHT-LDH) was prepared by fast coprecipitation followed by hydrothermal treatment method modified from Xu et al. (15) (described in (S2)). To prepare LDH by conventional coprecipitation method, mixed metal salts dissolved in decarbonated ultrapure water (DW) were added into NaOH solution. The slurry was aged, washed, and finally dried (details in S3). The material obtained was denoted as COPLDH. A part of the COP-LDH was calcined at 500 °C for 5 h (denoted as CAL-LDH). The LDH produced by conventional ion exchange method, Mg6Al2(CO3)(OH)16 · 4H2O (denoted as IE-LDH) was purchased from Sinopharm Chemical. The elemental analysis of the LDHs was carried out using ICP-OES of Perkin-Elmer Optima 2000 after microwaveassisted acid digestion. The CHNS/O analyzer (Perkin-Elmer 2400 Series II) and the scanning electron microscope with energy dispersive X-ray spectroscopy (SEM-EDX, JEOL JED2300) were used for determining carbon and nitrogen, and arsenic content, respectively. The BET surface area and porosity characteristics were determined using a QuantaChrome Autosorb-1 Analyzer. Photon correlation spectroscopy (PCS, Malvern Zetasizer Nano ZS) was used to analyze the particle size distribution of various LDH samples. The phase purity of the LDH samples was analyzed by Bruker AXS (D8 Advance) X-ray diffractometer (XRD) under a condition of 40 kV, 40 mA, and Cu KR radiation of λ ) 1.54 Å. Fourier transform infrared spectroscopy (FTIR) spectra were collected on a Perkin-Elmer 2000 FTIR spectrometer. Powdered samples deposited onto carbon-coated copper grids were analyzed using TEM performed at 200 kV on a JEOL JSM2010F microscope. Field limiting apertures used for selected area electron diffraction (SAED) were 5-60 µm in diameter whereas high-resolution TEM image (HRTEM) was collected using a high-contrast objective aperture of 20 µm. XPS analyses were performed on a Kratos AXIS Ultra spectrometer with the monochromatic Al KR X-ray radiation at 1486.71 eV. The energy scale of the XPS spectra was calibrated with the binding energy of the C 1s peak due to the surface contamination. Nitrate Displacement Experiments. The displacement of interlayer nitrate by As(V) was investigated as a function of As(V) loading and reaction time. A predetermined dose of FCHT-LDH was contacted with As(V) solutions of concentrations ranging from 0.013 to 0.667 mM for the former experiment whereas FCHT-LDH was added into 1 L of 0.2 mM As(V) solutions and stirred for 6 h with intermittent sampling at selected time intervals in the latter. The 2538
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 7, 2009
concentrations of As(V) were determined by ICP-OES whereas the concentrations of the nitrate were analyzed by ion chromatograph (IC, Dionex ICS-1000). For measurements of As(V) concentrations at µg/L level, a graphite furnace atomic absorption spectrophotometer (GFAAS, Perkin-Elmer SIMMA 6000) was used. Arsenic speciation analysis using the method modified from Lim and Goh (16) showed insignificant transformation of As(V) to As(III) up to 24 h. Sorption Isotherms, Kinetics, pH Effect, and Regeneration. All the experiments were performed using the batch sorption technique at 25 °C and pH 9.5 (near the natural pH of the LDH suspension), unless otherwise stated. At pH 9.5, HAsO42- is the dominant As(V) species. NaNO3 was used as the background electrolyte. Triplicate measurements were performed for most experiments. The sorption isotherms were determined by contacting various LDHs with 0.0130.667 mM As(V) solutions. The LDH and As(V) solutions were reacted in the centrifuge tubes for 24 h to allow complete equilibrium. For the experiment investigating temperature effect on As(V) sorption by FCHT-LDH, the suspensions were reacted at temperature-controlled shaker at different temperatures. The suspension in each tube was centrifuged and filtered, and the concentrations of As(V) were determined by ICP-OES or GFAAS. Kinetic experiments were performed by adding LDH samples into 0.2 mM As(V) solutions and stirring for 6 h, and at selected time intervals 4 mL samples were extracted, filtered, and analyzed. To investigate the pH effect on As(V) sorption by FCHT-LDH, 20 mL of FCHT-LDH suspension and 20 mL of 0.46 mM As(V) solution, both with preadjusted pH (in the range of 5-11), were mixed and stirred for 4 h with intermittent pH adjustment. The regeneration of FCHT-LDH was evaluated by reacting As-loaded FCHTLDH in the regenerating solution for several hours. The amount of As(V) desorbed was determined through analysis of As(V) released into the solution whereas the amount of As(V) uptake by the regenerated FCHT-LDH was determined through analysis of As(V) removed from the solution. For the best-performing regeneration method, four cycles of As(V) sorption-desorption were performed.
Results and Discussion Characterizations. The Mg2+/Al3+ molar ratios in all the LDHs were very close to that of their starting salts (Table 1). The CHNS/O analyses showed the presence of N and C in all the LDHs, implying the coexistence of nitrate and carbonate in the interlayer spaces of these LDHs. The higher C content in IE-LDH agreed with its composition while the presence of C in COP-LDH, CAL-LDH and FCHT-LDH was possibly due to adsorption of the atmospheric CO2. The pore sizes of all the LDHs fell in the meso size range (2-50 nm). The surface area, pore volume, and pore size of FCHT-LDH were 127 m2/g, 0.31 cm3/g, and 8 nm, respectively, the highest among the investigated LDHs. This was probably due to the effective prevention of LDH particle aggregation during the hydro-
Keq )
γAs(V)CAs(V)(NNO3-)2
(2)
(γNO3-)2CNO3-(NAs(V))
where γAs(V) and γNO3- are the solution-phase activity coefficients of the two oxyanions calculated using the extended Debye-Hu ¨ ckel equation (20), CAs(V) and CNO3- are the equilibrium concentrations (M) of As(V) and nitrate, respectively, and NNO3- and NAs(V) are the LDH-phase activity in terms of mole fraction (NNO3- ) {NO3- }/({NO3- } +{As(V)}) and NAs(V) ) {As(V)}/({As(V)} +{NO3- })), where braces indicate the LDH phase composition in mol/kg). Kex is then computed from the following equation: ln Kex )
FIGURE 1. HRTEM image of FCHT-LDH with the elliptic regions highlighting the diffraction pattern of a number of randomly oriented crystal grains (the inset on the top left is its SAED pattern whereas the inset on the bottom left is the identified interplanar distance (0.26 nm) that is close to that of d(009)). thermal treatment of FCHT-LDH. The PCS analysis demonstrated that FCHT-LDH had a narrow particle size distribution with an equivalent hydrodynamic diameter of 122 nm and all particles fell within 38-396 nm. A single hydrotalcite-like phase (ICSD 81963) can be identified for all the LDHs, except for CAL-LDH (Figure S1). FCHT-LDH showed the smallest crystal size and highest interlayer spacing (indicated by c value) among the LDHs (Table 1). The peak positions of the FTIR spectra for all the LDHs were close despite very small shifts (Figure S2). The FTIR spectra of COP-LDH, CAL-LDH, and FCHT-LDH showed a peak at ∼1384 cm-1 whereas the spectra of IE-LDH exhibited a slightly strong peak at 1360 cm-1, indicating the significant presence of nitrate and carbonate, respectively, and these results were consistent with that of CHNS/O analysis. Figure 1 depicts the HRTEM image of the fresh FCHT-LDH. The elliptic regions show the randomly oriented LDH crystallites consisting of nanocrystalline grains which ranged from 5.3 to 5.8 nm. The identified interplanar distance (0.26 nm) was close to that of (009) plane. The selected area diffraction (SAED) patterns confirmed the presence of LDH (as identified by the diffraction rings corresponding to (003) and (006) planes). Interactions of As(V) with Nanocrystalline LDH. Figure 2a shows that the amount of sorbed As(V) increased linearly with As(V) loading up to 114 cmol/kg, which corresponded to the anion exchange capacity estimated from the FCHTLDH stoichiometry (Mg2Al(OH)6(NO3)0.49(CO3)0.25 · mH2O) and the anion exchange stoichiometry (eq 1), assuming that interlayer carbonate was not exchangeable (17).
∫ ln K 1
0
eqd(LDH)As(V)
(3)
where (LDH)As(V) is the equivalent fraction of As(V) on the LDH. All the parameters obtained in the computation process of Kex are shown in Table S1. The calculated value of Kex is 30.6. The value obtained was within the range reported by Miyata (21) (Kex ) 24.5-69.2) in similar heterovalent exchange system (e.g., SO42- exchanged for NO3-) but smaller than the one demonstrated by Israe¨li et al. (22) for a homovalent exchange system (e.g., OH- exchanged for Cl-). The obtained Kex confirmed the stoichiometric and reversible electrostatic interaction between the As(V) in the solution and the nitrate in the interlayer region of FCHT-LDH. Figure 2b shows that nitrate was released steadily from FCHT-LDH with simultaneous As(V) sorption. A molar ratio ranging from 2 to 3 for the displacement of nitrate by As(V) on FCHT-LDH throughout the reaction time supported the occurrence of anion exchange process. The slight difference between the obtained molar ratio and the stoichiometry ratio of 2 was possibly due to the leaching of residual nitrate (from the precursor used) remaining in the FCHT-LDH product. According to Table 1, the (003) reflection peak moved to a
HAsO24 (aq) + 2NO3 - LDH(s) T 2NO3(aq) + HAsO4 LDH(s)(1)
To assess the extent of As(V) binding by anion exchange, the equilibrium exchange constant, Kex was determined. This parameter is useful for determining the state of ionic equilibrium at different ion concentrations. Kex was calculated by referring to the methods of Vanselow (18) and Gaines and Thomas (19). According to these methods, Keq (apparent equilibrium exchange constant) is written as:
FIGURE 2. (a) As(V) sorbed as a function of the As(V) loading (no background electrolyte added) and (b) As(V) remained in solution and nitrate displaced as a function of reaction time (number in bracket indicates the molar ratio of nitrate displaced to As(V) sorbed). VOL. 43, NO. 7, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2539
FIGURE 4. Sorption results for the reaction of As(V) and As(III) with FCHT-LDH as a function of pH.
FIGURE 3. (a) Equilibrium isotherms of As(V) on various LDHs and (b) equilibrium isotherms of As(V) on FCHT-LDH at different temperatures (pH ) 9.5, time ) 24 h, dose ) 0.4 g/L, background ) 0.01 M NaNO3) and the inset is the van’t Hoff plot of ln Kc against reciprocal temperature. Experimental data are reported as points and model data (Langmuir Isotherm) are reported by curves. lower 2θ after FCHT-LDH reacted with As(V), indicating the exchange of As(V) (molecular diameter ∼0.5 nm) for nitrate (molecular diameter ∼0.4 nm). The decrease of the intensity of (003) reflection peak also suggested that As(V) entered the interlayer of FCHT-LDH. Decrement of N with simultaneous increment of As in the exhausted FCHT-LDH was observed in the XPS, CHNS/O, and EDX analyses (Table S2 and Table 1). These results confirmed that anion exchange was one of the important mechanisms for the As(V) removal by FCHTLDH and subsequently As(V) might bind on the internal surfaces by forming outer-sphere complex (5). After FCHTLDH reacted with As(V), the FTIR band assigned to interlayer nitrate (1384 cm-1) became weaker, while a new band, corresponding to As-O stretching vibration, was detected at 830 cm-1, further suggesting the exchange of nitrate with As(V). The concurrent monitoring of the solution pH showed a discernible pH increment of ca. 0.4-0.6 during As(V) sorption equalization, implying the occurrence of ligand exchange of the surface OH groups with As(V) under this experimental condition. The exchange of As(V) with OH- was possible since both ligands are hard bases which preferentially react with hard acids such as Al3+ center, according to hard and soft acids and bases (HASB) theory. The peak shift to 3599 cm-1 from 3480 cm-1 after As(V) sorption also suggested the possible involvement of OH group in the sorption process. Therefore, it could be deduced that while anion exchange was the predominant mechanism for As(V) sorption by FCHTLDH, and ligand exchange might occur concurrently as the secondary sorption mechanism. These mechanisms are further discussed in a later section. Sorption Isotherms, Thermodynamics, and Kinetics. The isotherms for As(V) sorption on FCHT-LDH were compared with those for LDHs prepared with the conventional methods (Figure 3a). Langmuir isotherm was found more suitable than Freundlich isotherm for describing the As(V) sorption on all 2540
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 7, 2009
FIGURE 5. XPS spectra of (a) Mg 2p and As 3d, (b) Al 2p, and (c) O 1s for fresh and As(V)-loaded FCHT-LDHs. the LDHs on the basis of the linear coefficient of determination (R2) and nonlinear Chi-square (χ2) statistical test (Table S3). The sorption maxima for As(V) on FCHT-LDH was 114
TABLE 2. O 1s Peak Parameters for Fresh and As(V)-Loaded FCHT-LDHs type of FCHT-LDH FCHT-LDH
FCHT-LDH-50 cmol As(V)/kg
FCHT-LDH-334 cmol As(V)/kg
speciesa
binding energy (eV)
fwhm (eV)
percentb (%)
O2OHH2O
530.65 531.32 531.80
1.6 1.6 1.6
18.1 33.2 48.7
O2OHH2O
530.52 531.42 531.80
1.7 1.7 1.7
25.7 56.1 18.2
O2OHH2O
530.39 531.50 532.36
1.5 1.5 1.5
25.7 59.1 15.2
a O2-: oxygen bonded to metal; OH-: hydroxyl bonded to metal; H2O: surface and/or interlayer water. b Percentage of the atomic concentration of each species to the total oxygen atoms.
cmol/kg or 85 mg/g and it was much higher than those of the other LDHs. The higher As(V) sorption capacity of FCHTLDH could be due to its low carbonate content, higher surface area, porous nature, small crystallite size, unique morphology, and nanocrystalline property. The highly porous nanocrystalline FCHT-LDH with high surface area has allowed for an enhanced surface chemical reactivity and imparted a very large area of contact between As(V) and FCHT-LDH particles, resulting in its good sorption ability. Figure 3b depicts the equilibrium sorption isotherms of As(V) at 25, 35, 45, and 55 °C. The equilibrium capacity increased with increasing temperature, suggesting that the interaction between As(V) and FCHT-LDH was endothermic in nature. Thermodynamic parameters such as Gibbs free energy (∆G °), standard enthalpy change (∆H°), and standard entropy change (∆S °) were determined using the following equation and van’t Hoff plot: ∆Go ) -RTln Kc ) ∆Ho - T∆So
(4)
where R is universal gas constant (8.314 J/mol.K), T is temperature (K), and Kc is the thermodynamic equilibrium constant that can be defined as follows: Kc )
as υs C s ) ae υe Ce
(5)
where as is the activity of sorbed As(V), ae is the activity of the As(V) in solution at equilibrium, Cs is the surface concentration of As(V) (cmol/kg), Ce is the concentration of As(V) at equilibrium (cmol/L), υs is the activity coefficient of the sorbed As(V), and υe is the activity coefficient of the As(V) in solution. As the concentration of the As(V) in the solution approaches zero, the activity coefficient approaches unity, and eq 5 can be simplified to: as Cs ) ) Kc Ce ae
(6)
Kc values were obtained by plotting ln (Cs/Ce) vs Cs and extrapolating Cs to zero. The values of ∆H° and ∆S ° were derived respectively from the slope and intercept of the van’t Hoff plot. The negative ∆G ° value (-6.43 kJ/mol) indicated that the As(V) removal by FCHT-LDH was a spontaneous process, implying that the exchange force was strong enough to break the potential and drive the reaction to displace nitrate with As(V) in the FCHT-LDH. The positive value of ∆H° (3.73
kJ/mol) confirmed the endothermic nature of anion exchange process. The positive value of ∆S ° (0.03 kJ/mol.K) revealed the increased randomness at the solid/solution interface for the anion exchange between As(V) and interlayer nitrate. Four kinetic models were used to fit the experimental data, including the commonly used pseudo first order and pseudo second order models, modified multiplex model (23), and Elovich model. Elovich was considered the most suitable in describing the sorption kinetics of As(V) on the LDHs in terms of the statistical parameters (R2 > 0.990, χ2 < 0.95, and ∆q < 5) (Table S4). Elovich equation, which considers the heterogeneous sorptive sites (24), also matched the sorption properties of FCHT-LDH since it was postulated that there were interlayer and surface sites of FCHT-LDH for the As(V) sorption. The activation energy (Ea) is usually less than 30 kJ/mol for anion exchange-controlled processes (25). The Ea value calculated from the gradient (-Ea/R) of ln k2 vs 1/T using the data for As(V) sorption at 5-65 °C in this study was 24.7 kJ/mol (Figure S3), indicating the predominantly anion exchange process for the As(V) sorption by the FCHT-LDH. pH Effect. The pH influence on As(V) sorption on FCHTLDH is illustrated in Figure 4, along with the results of a similar study performed on arsenite (As(III)). The results showed that As(V) sorption decreased gradually with increasing pH in the range of pH 5.3-9.5. For the As(V) loading of 115 cmol/kg, more than 80% of As(V) (equivalent to ∼92 cmol/kg) was removed by FCHT-LDH in this pH range. The As(V) sorption, however, decreased remarkably at pH >9.5. The decrease of As(V) sorption with increasing pH was due to the decreasing Coulombic attraction resulting from the decreasing positive charge of FCHT-LDH. This sorption characteristic was different from that of As(III). As(III) removal increased from 12% to 52% when the pH was increased from 5.3 to 10.1. The maximum sorption was observed at pH value slightly higher than the first dissociation constant of H3AsO3 (pK1 ) 9.2). The higher sorption of As(V) than As(III) by FCHTLDH was due to the presence of monovalent H2AsO4- and divalent HAsO42- which were more favorable for the anion exchange process compared to the uncharged H3AsO3 and monovalent H2AsO3-. Regeneration. Highest As(V) desorption rate (∼76%) was achieved by desorbing As-loaded FCHT-LDH in Na2CO3 with Mg and Al dissolutions of less than 0.1% of their initial masses (Table S5). Both desorption and regeneration rates decreased gradually with number of cycle (Figure S4). Dissolution of Mg was relatively high in the initial cycles but progressively reduced to 0.13% in the forth cycle whereas dissolution of Al remained constant at
