Environ Geochem Health (2011) 33:133141 DOI 10.

1007/s10653-010-9349-z

ORIGINAL PAPER

Removal of arsenate from water by adsorbents: a comparative case study

Sunbaek Bang Maria E. Pena Manish Patel Lee Lippincott Xiaoguang Meng Kyoung-Woong Kim

Received: 18 January 2010 / Accepted: 3 August 2010 / Published online: 3 November 2010 Springer Science+Business Media B.V. 2010

Abstract Laboratory and eld ltration experiments were conducted to study the effectiveness of As(V) removal for ve types of adsorbent media. The media included activated alumina (AA), modied activated alumina (MAA), granular ferric hydroxide (GFH), granular ferric oxide (GFO), and granular titanium dioxide (TiO2). In laboratory batch and column experiments, the synthetic challenge water was used to evaluate the effectiveness for ve adsorbents. The results of the batch experiments showed that the As(V) adsorption decreased as follows at pH 6.5: TiO2 [ GFO [ GFH [ MAA [ AA. At pH 8.5, however, As(V) removal decreased in the following order: GFO = TiO2 [ GFH [ MAA [ AA. In column experiments, at pH 6.5, the adsorbed As(V) for
S. Bang (&) International Environmental Research Center (IERC), Gwangju Institute of Science and Technology (GIST), Gwangju 500-712, South Korea e-mail: sbang@gist.ac.kr M. E. Pena X. Meng Center for Environmental Systems, Stevens Institute of Technology, Hoboken, NJ 07030, USA M. Patel L. Lippincott New Jersey Department of Environmental Protection, P.O. Box 409, Trenton, NJ 08625, USA K.-W. Kim School of Environmental Science and Engineering, Gwangju Institute of Science and Technology (GIST), Gwangju 500-712, South Korea

adsorbents followed the order: TiO2 [ GFO [ GFH, whereas at pH 8.5 the order became: GFO = TiO2 [ GFH when the challenge water containing 50 lg/L of As(V) was used. Field ltration experiments were carried out in parallel at a wellhead in New Jersey. Before the efuent arsenic concentration increased to 10 lg/L, approximately 58,000 and 41,500 bed volumes of groundwater containing an average of 47 lg/L of As(V) were treated by the lter system packed with GFO and TiO2, respectively. The As(V) adsorption decreased in the following sequence: GFO [ TiO2 [ GFH [ MAA [ AA. Filtration results demonstrated that GFO and TiO2 adsorbents could be used as media in small community ltration systems for As(V) removal. Keywords Filtration Arsenate Adsorption Adsorbent

Introduction Naturally occurring arsenic is found in groundwater in many regions of the world, including countries such as Bangladesh, India, China, Chile, Argentina, Mexico, Taiwan, Vietnam, and the United States (Smedley and Kinniburgh 2002). It is estimated that naturally occurring arsenic in drinking water supplies has affected more than 100 million people worldwide (Bang et al. 2005a). Arsenic concentrations in groundwater are usually higher than in surface water.

123

134

Environ Geochem Health (2011) 33:133141

It is well known that the symptoms of chronic arsenic poisoning have occurred in Bangladesh since large populations have been exposed to high arsenic concentrations over a long period of time (Das et al. 1996). In order to reduce the health risk, the United States Environmental Protection Agency (USEPA) decreased the arsenic maximum contaminant level (MCL) from 50 to 10 lg/L in 2001 (USEPA 2001). Numerous technologies have been used for the treatment of water containing arsenic. The most widely applied technique for arsenic removal is conventional coagulation/ltration using ferric salt or alum (Cheng et al. 1994; Scott et al. 1995). The conventional coagulation technique has been demonstrated to remove arsenic effectively. However, this technique is not suitable for use in small systems such as point-of-use (POU) and point-of-entry (POE) systems. Anion exchange and reverse osmosis techniques can be used in POU systems for the removal of arsenic. However, the presence of high sulfate and total dissolved solids (TDS) can signicantly reduce the removal effectiveness of the ion exchange resins for As(V) (USEPA 2000a). In addition, reverse osmosis (RO) has a low recovery of product water and its performance is adversely affected by the presence of turbidity, iron, manganese, and silica (USEPA 2000b). Adsorption using simple ltration packed with adsorbent media is a promising technology for small communities because the operation of the system and maintenance is simple and easy. Many adsorbents have been evaluated for their effectiveness to remove arsenic. Adsorbents such as activated alumina (AA), modied AA (MAA), and granular ferric hydroxide (GFH) have been used in POU systems to remove arsenic as well as in a full-scale water treatment facility (Khandaker et al. 2006). Novel materials such as granular ferric oxide (GFO) and granular titanium dioxide (TiO2) have been developed and used for the removal of arsenic from groundwater (Bang et al. 2005b). In the present study, the As(V) adsorption rate and adsorbed amount of ve adsorbent media were investigated and compared using synthetic challenge water (simulated groundwater) and groundwater. Batch experiments were used to study the rate of As(V) adsorption and the effect of pH on As(V) adsorption onto the surface of adsorbents. Column experiments were conducted to determine the amount

of As(V) adsorption for three effective adsorbents (GFH, GFO, and TiO2) at pH 6.5 and 8.5 because the effectiveness for AA and MAA on As(V) removal was low. Field ltration tests were operated at the wellhead to evaluate the performance of lters packed with the ve adsorbents for As(V) removal.

Materials and methods Materials All chemicals were analytical grade, and all stock solutions were prepared with deionized water (DI) by a Milli-Q water system (Millipore, Billerica, MA). The As(V) stock solutions were prepared from Na2HAsO47H2O (Fisher Scientic, Fair Lawn, NJ). Na2SiO35H2O (Fisher Scientic, Fair Lawn, NJ), NaH2PO4H2O (Fisher), NaHCO3 (Fisher Scientic, Fair Lawn, NJ), MgSO47H2O (Fisher Scientic, Fair Lawn, NJ), NaNO3 (Aldrich, Milwaukee, WI), NaF (Fisher Scientic, Fair Lawn, NJ), and CaCl2 (Fisher Scientic, Fair Lawn, NJ) were used. Five commercially available adsorbents for the removal of As(V) were investigated to determine their As(V) adsorption amounts and to select suitable adsorbents for further experiments. The adsorbents include activated alumina (AA, Alcan Inc., Cleveland, OH), modied (ferric-coated) activated alumina (MAA, Alcan Inc., Cleveland, OH), granular ferric hydroxide (GFH, US lter, Warrendale, PA), granular ferric oxide (GFO, Apyron Technologies, Atlanta, GA), and granular titanium dioxide (TiO2, Hydroglobe, Hoboken, NJ). All adsorbents were used as received. The BrunauerEmmettTeller (BET) surface area analysis for the ve adsorbents is shown in Table 1. The surface area for GFH was measured after drying using N2 gas because the GFH had more than 20% moisture content. Laboratory batch and column experiments were conducted using the challenge water prepared daily with DI water in accordance with NSF International Standards 53 (NSF/ANSI 53-2001). The chemical compositions of the challenge water are listed in Table 2. Na2SiO35H2O was added to the DI water to make the challenge water. After adjusting the solution pH using HCl or NaOH to 6.5 0.25 for the low pH test and to 8.5 0.25 for the high pH test, chemicals such as NaHCO3, MgSO47H2O, NaNO3,

123

Environ Geochem Health (2011) 33:133141 Table 1 Surface areas and particle sizes of the ve adsorbents investigated AA Surface area (m2/g) Particle size (mm) 296 0.350.65 MAA 243.9 0.350.65 GFH 300.7 0.250.85 GFO 137.7 0.250.60 TiO2 250.7

135

0.150.60

Table 2 Characteristics of challenge water used in laboratory experiments Ca (mg/L) 40 Mg (mg/L) 12 SiO2 (mg/L) 20 PO4 (mg/L) 0.04 HCO3 (mg/L) 181.5 NO3 (mg/L) 8.8 F (mg/L) 1 SO4 (mg/L) 50

NaF, NaH2PO4H2O, CaCl2, and NaClO were added into the solution. The mixed solution was spiked with Na2HAsO47H2O, and the nal pH was measured and readjusted using HCl or NaOH. Batch experiments The rate of As(V) adsorption on adsorbents was determined in batch experiments. In the batch tests, 0.1 g (air dry weight) of adsorbent media was added to 100-mL polypropylene bottles lled with challenge water containing 10 mg/L As(V). The solutions were ltered through a 0.4-lm lter membrane after 0.5, 1, 2, 4, 8, 12, 24, and 48 h of mixing. No pH adjustment was performed during the experiments. The pH of the ltered solution was determined with a pH meter (Orion). All samples for arsenic analysis were acidied with nitric acid. Column experiments Laboratory experiments were conducted to investigate the effectiveness of the adsorbents for As(V) removal from the challenge water in accordance with an arsenic reduction protocol (NSF/ANSI Standard 53-2001) developed for testing the POU drinking water treatment devices at pH 6.5 and 8.5. The system was operated on a 50% on/50%-off cycle basis with 30-min cycles and was run 16 h/day followed by an 8-h rest. The challenge water containing 50 or 300 lg/L of As(V) was passed through a glass column under downward mode with a peristaltic pump. The glass column had an inside diameter of 2.54 cm and bed height of 7 cm and was lled with 35 mL of adsorbent medium (25.9 g of TiO2, 35.2 g of GFH, or 17.2 g of GFO). The ow rate was 140 mL/min (empty bed contact

time = 15 s (EBCT)). Efuent water samples were collected and acidied with nitric acid before arsenic analysis. Efuent pH was measured from collected samples. Field ltration experiment The ltration system containing water ow totalizer (for measuring treated volume), pressure gauge, and lter house was used in the eld experiments. The 3.785-L (1 gallon) lter had an inside diameter of about 11.4 cm and was lled with 3 L of adsorbent medium with a bed height of about 29 cm. Arseniccontaminated groundwater was passed through lters containing different adsorbents in a downward ow mode. Five ltration units were installed and operated in parallel at the wellhead in central NJ, USA. The characteristics of the arsenic-containing groundwater from the site are listed in Table 3. The water ow rate was set to 1 L/min (1,440 L/day) with continuous ow, resulting in an EBCT of 3 min. Turbidity, pH, ow rate, and pressure were measured at the site. All samples for arsenic analysis were acidied with nitric acid. Analytical methods A Graphite Furnace Atomic Absorption Spectrometer (GFAAS, Varian) was used to determine the arsenic concentration. During GFAAS analysis, 20 lL water samples and 3 lL palladium-citric acid modier were injected into a partitioned graphite tube. Spike recovery and National Institute of Standards and Technology (NIST) Standard Reference Material (SRM) 1640 solution were used to check the accuracy of the arsenic analysis. All samples were measured at least two times in order to assess the repeatability of

123

136 Table 3 Characteristics of groundwater used in eld experiments As (lg/L) 3952 Ca (mg/L) 41 Mg (mg/L) 26 SiO2 (mg/L) 21 PO4 (mg/L) \0.01

Environ Geochem Health (2011) 33:133141

NO3 (mg/L) 3

F (mg/L) \0.1

SO4 (mg/L) 65

the measurement. Samples were reanalyzed if the error of the SRM sample exceeded 10% or the relative standard deviation of the measurement exceeded 5%.

Results and discussion Batch experiments Batch experiments were conducted to determine the rate of As(V) removal from challenge water by ve adsorbents (Fig. 1). The initial As(V) concentration was 10 mg/L, and the initial pH of the solution was adjusted to 6.5 and 8.5. After 48 h of reaction, approximately 87 and 65% of the As(V) was removed by AA at pH 6.5 and 8.5, respectively. More than 89 and 74% of the As(V) was adsorbed on the surface of the MAA at pH 6.5 and 8.5, respectively. At pH 6.5 and 8.5, the GFH adsorbent achieved an As(V) removal efciency of approximately 95 and 92%, respectively. The adsorption of As(V) was not different between pH 6.5 and 8.5 with GFO and TiO2 adsorbents after 48 h of reaction. At pH 6.5 and 8.5, more than 97% of the As(V) was eliminated by the GFO and TiO2 adsorbents. When the initial pH of the solution was adjusted to 6.5, As(V) adsorption was enhanced for all adsorbents. These results indicate that As(V) adsorption is pH dependent and related to typical anion adsorption. Similar adsorption trends for As(V) have also been observed for alumina (Arai et al. 2001; Lin and Wu 2001), iron (hydr)oxides (Driehaus et al. 1998; Meng et al. 2000; Holm 2002), and titanium dioxide (Dutta et al. 2004; Bang et al. 2005b; Pena et al. 2005). Arai et al. (2001) found that the adsorption of As(V) decreased from 90 to 25% when the pH was increased from 5 to 10. The adsorption of As(V) onto hydrous ferric oxide increased from 20 to 100% when the solution pH decreased from 10 to 7 (Holm 2002). The removal of As(V) decreased from 99 to 70% in the 1.0 g/L TiO2 suspension when the solution pH was increased from 7.3 to 10.3 (Bang et al. 2005b).

The As(V) adsorption rate decreased in the following order at pH 6.5 (TiO2 [ GFO [ GFH [ MAA [ AA) and at pH 8.5 (GFO = TiO2 [ GFH [ MAA [ AA), indicating that iron-based and titanium-based adsorbents have more effectiveness for As(V) adsorption when compared to aluminum-based adsorbents. The adsorption amount of GFH for As(V) is higher than that of AA (Wingrich and Wolf 2002; Thirunavukkarasu et al. 2003). At an initial As(V) concentration of 29 mg/L and pH 7, approximately 18.8 mg/g As was adsorbed on the surface of the iron-coated alumina adsorbent prepared by in situ precipitation of iron hydroxide on the surface of AA (Hlavay and Polyak 2005). The adsorption density of the GFH was 75 mg/g Fe at a residual As(V) of 10 lg/L (Driehaus et al. 1998). The acid digestion analysis showed that 1 g of GFH contained 269 mg of Fe (Thirunavukkarasu et al. 2003). The calculation based on the total iron content in GFH showed that the As(V) adsorbed on the surface of the GFH was 20.2 mg/g. Approximately, 22 mg of As(V) was adsorbed on 1 g of GFO and the adsorption for As(V) on TiO2 was 40 mg/g at an equilibrium As(V) concentration of 45 mg/L and pH 7 (Pena et al. 2005). Bang et al. (2005b) calculated a conditional maximum As(V) adsorption amount on TiO2 from the Langmuir equation. The values of Cmax from adsorption data using groundwater and challenge water were 41.4 and 40.0 mg/g, respectively. Column experiments Laboratory column experiments were conducted to compare the adsorption effectiveness of three adsorbents (GFH, GFO, TiO2) for the removal of As(V) from challenge water containing 50 lg/L As(V). The concentrations of As(V) in the column efuent as a function of bed volumes at inuent pH 6.5 and 8.5 are shown in Figs. 2 and 3, respectively. The pH of ltered water ranged between 6.3 and 6.7 at inuent pH 6.5 and from 8.3 to 8.7 at inuent pH 8.5. The results in Fig. 2 showed that GFH treated

123

Environ Geochem Health (2011) 33:133141 Fig. 1 Rate of As(V) adsorption from challenge water using different adsorbents. a AA, b MAA, c GFH, d GFO, e TiO2. An As(V) concentration of 10 mg/L was mixed with 1.0 g/L adsorbent medium

137

Fig. 2 Efuent arsenic concentrations as a function of bed volumes. The inuent contained 50 lg/L As(V) in the challenge water at pH 6.5. The ow rate and the EBCT in the column experiment were 140 mL/min and 15 s, respectively

Fig. 3 Efuent arsenic concentrations as a function of bed volumes. The inuent contained 50 lg/L As(V) in the challenge water at pH 8.5. The ow rate and the EBCT in the column experiment were 140 mL/min and 15 s, respectively

123

138

Environ Geochem Health (2011) 33:133141

Fig. 4 Efuent arsenic concentrations as a function of bed volumes. The inuent contained 300 lg/L As(V) in the challenge water at pH 6.5. The ow rate and the EBCT in the column experiment were 140 mL/min and 15 s, respectively

Fig. 5 Efuent arsenic concentrations as a function of bed volumes. The inuent contained 300 lg/L As(V) in challenge water at pH 8.5. The ow rate and the EBCT in the column experiment were 140 mL/min and 15 s, respectively

approximately 8,700 bed volumes of challenge water before the efuent arsenic concentration exceeded the drinking water standard for arsenic of 10 lg/L. The arsenic breakthrough for GFO and TiO2 occurred at approximately 16,000 and 30,000 bed volumes, respectively. Approximately, 2,900 and 7,600 bed volumes were treated using GFH and GFO, respectively (Fig. 3). The gradual breakthrough began at about 2,400 bed volumes for TiO2, and the nal breakthrough was observed at approximately 7,700 bed volumes. The TiO2 adsorbent was about 3.4 and 2.7 times more effective than GFH at pH 6.5 and 8.5, respectively. On the other hand, As(V) adsorption of TiO2 and GFO was not much different at pH 8.5, whereas the adsorption efciency on TiO2 was about 1.9 times greater than that of GFO at pH 6.5. The adsorption efciency of As(V) for the three adsorbents were compared using challenge water containing 300 lg/L of As(V) (Figs. 4 and 5). Efuent arsenic concentrations showed the breakthrough of GFH, GFO, and TiO2 at 2,500, 4,300, and 8,000 bed volumes, respectively (Fig. 4). Efuent pHs were within 0.2 pH units of inuent pHs. At pH 8.5, the efuent concentrations of arsenic showed the breakthrough of GFH, GFO, and TiO2 at 230, 500, and 1,700, respectively (Fig. 5). Rapid breakthroughs were observed at pH 8.5. When a high As(V) concentration was used in the inuent solution, treated bed volumes were dramatically decreased. The treatment efciency of TiO2, GFH, and GFO for As(V) is summarized in Table 4.

Table 4 Treatment efciency for GFH, GFO, and TiO2 on As(V) removal from challenge water As(V) (lg/L) Adsorbent pH 50 300 6.5 8.5 6.5 8.5 Unit = bed volume GFH 8,700 2,900 2,500 230 GFO 16,000 7,600 4,300 500 TiO2 30,000 7,700 8,000 1,700

The adsorbed amount of As(V) for the adsorbents followed the order: TiO2 [ GFO [ GFH, indicating that titanium-based adsorbents have more effectiveness for As(V) adsorption when compared to ironbased adsorbents. Fluctuations of efuent arsenic concentration were observed at pH 6.5 (Figs. 2 and 4). After a shutdown period of 8 h, efuent arsenic concentrations were signicantly reduced for all adsorbents. Westerhoff et al. (2005) conducted a column experiment with onoff pump cycles using GFH. They also observed similar trends with the on off ltration mode. This can be attributed to the diffusion of arsenic into the adsorbents during the shutdown period, resulting in the reduction of arsenic on the stagnant layer of adsorbent (Badruzzaman et al. 2004; Westerhoff et al. 2005). Because of a concentration gradient between the liquid and surface arsenic concentrations, efuent arsenic

123

Environ Geochem Health (2011) 33:133141

139 Table 5 Treated efciency for AA, MAA, GFH, GFO, and TiO2 on As(V) removal from groundwater AA MAA 7,300 GFH 20,900 GFO 58,000 TiO2 41,500

concentrations decreased when the ow of column was resumed after the shutdown period. Field experiments

2,000

Field experiments were conducted to study the effectiveness of different adsorbents for As(V) removal from groundwater in New Jersey. Five adsorbents were compared under the same ltration conditions. The results of the ltration experiments are shown in Fig. 6. The arsenic concentrations of the raw groundwater varied from 39 to 52 lg/L. As speciation cartridge was used to separate As(III) species from total arsenic but no As(III) was detected in the groundwater. Approximately 2,000 and 7,300 of bed volumes were treated by AA and MAA, respectively. The AA adsorbent was saturated within approximately 5 days of ltration and a breakthrough occurred in the MAA lter after 15 days. Efuent turbidity and pH were in the range of 0.070.12 (NTU) and 7.88.3, respectively. A breakthrough was observed at about 20,900 bed volumes with GFH. The GFH was about 2.8 and 2 times less effective than GFO and TiO2, respectively. The efuent pH was between 7.8 and 8.1, and the average efuent turbidity was 0.09 NTU. Gradual breakthrough began at about 43,000 bed volumes, and a breakthrough was observed at approximately 58,000 bed volumes with GFO. The turbidity of the ltered water decreased to 0.030.10 NTU, and the ltered water pH varied between 7.48 and 8.08. The breakthrough of TiO2 occurred at approximately 41,500 bed volumes

Unit = bed volume

before the efuent arsenic concentration was over 10 lg/L. The turbidity of ltered water decreased to 0.060.12 NTU, and the ltered water pH varied from 7.67 to 8.36. The treatment effectiveness of each adsorbent medium for As(V) removal from groundwater is summarized in Table 5. According to the ltration results, the bed volumes on As(V) removal decreased in the following order GFO [ TiO2 [ GFH [ MAA [ AA at an empty bed contact time (EBCT) of 3 min before the efuent arsenic concentration increased to 10 lg/L. Bang et al. (2005b) reported that approximately 45,000 bed volumes of groundwater had been treated by a TiO2 lter at an inuent As(V) concentration of 3544 lg/ L under the same conditions. The results of the eld experiments are not in agreement with those of the laboratory column tests in the case of GFO and TiO2. The GFO adsorbent medium had better results for As(V) removal in the eld experiments than in the laboratory column experiments. The different operating conditions such as reactor size, contact time, operation mode (continuous or on/off mode) between laboratory column and eld experiments might also be responsible for the different ltration behaviors of the two adsorbents. The results of the eld experiments suggested that either GFO or TiO2 could be effectively and economically applied to remove arsenic-contaminated groundwater. The leachability of arsenic from spent adsorbents The Toxicity Characteristic Leaching Procedure (TCLP, 0.1 M acetate solution) was used to evaluate the arsenic leachability in spent adsorbents used in the eld experiments. The leachability tests of spent adsorbents generated in the eld experiments were performed (Jing et al. 2005). Arsenic concentrations were below 0.18 mg/L in the TCLP extraction solutions for all spent adsorbents. These concentrations were much lower than the USEPA regulation level for arsenic (5.0 mg/L).

Fig. 6 Efuent arsenic concentration as a function of bed volumes. The inuent contained 3952 lg/L As(V) in the groundwater. The ow rate and the EBCT in the eld experiment were 3 L/min and 3 min, respectively

123

140

Environ Geochem Health (2011) 33:133141 Cheng, R. C., Liang, S., Wang, H. C., & Beuhler, M. D. (1994). Enhanced coagulation for arsenic removal. Journal of American Water Work Association, 86(9), 7990. Das, D., Samanta, G., Mandal, B. K., Chowdhury, T. R., Chanda, C. R., Chowdhury, P. P., et al. (1996). Arsenic in ground water in six districts of West Bengal, India. Environmental Geochemistry and Health, 18, 515. Driehaus, W., Jakel, M., & Hildebrandt, U. (1998). Granular ferric hydroxide-a new adsorbent for the removal of arsenic from natural water. Journal of Water Supply: Research and Technology-Aqua, 47, 3035. Dutta, P. K., Ray, A. K., Sharma, V. K., & Millero, F. J. (2004). Adsorption of arsenate and arsenite on titanium dioxide suspensions. Journal of Colloid and Interface Science, 278, 270275. Hlavay, J., & Polyak, K. (2005). Determination of surface properties of iron hydroxide-coated alumina adsorbent prepared for removal of arsenic from drinking water. Journal of Colloid and Interface Science, 284, 7177. Holm, T. R. (2002). Effects of CO32-/bicarbonate, Si, and PO43- on arsenic sorption to HFO. Journal of American Water Work Association, 94(4), 174181. Jing, C., Liu, S., Patel, M., & Meng, X. (2005). Arsenic leachability in water treatment adsorbents. Environmental Science and Technology, 39, 54815487. Khandaker, N. R., Krumhansl, J., Neidel, L., & Siegel, M. (2006). Performance evaluation of AlCAN-AASF50-ferric coated activated alumina and granular ferric hydroxide (GFH) for arsenic removal in the presence of competitive lons in an active well: Kirtland eld trialinitial studies. SAND2005-7693, Sandia National Laboratories, Albuquerque, New Mexico and Livermore, California, USA. Lin, T., & Wu, J. (2001). Adsorption of arsenite and arsenate within activated alumina grains: equilibrium and kinetics. Water Research, 35, 20492057. Meng, X., Bang, S., & Koratis, G. P. (2000). Effects of silicate, sulfate, and carbonate on arsenic removal by ferric chloride. Water Research, 34, 12551261. Pena, M. E., Koratis, G. P., Patel, M., Lippincott, L., & Meng, X. (2005). Adsorption of As(V) and As(III) by nanocrystalline titanium dioxide. Water Research, 39, 23272337. Scott, K. N., Green, J. F., Do, H. D., & Mclean, S. J. (1995). Arsenic removal by coagulation. Journal of American Water Work Association, 87(4), 114126. Smedley, P. L., & Kinniburgh, D. G. (2002). A review of the source, behaviour and distribution of arsenic in natural waters. Applied Geochemistry, 17, 517568. The US Environmental Protection Agency. (2000a). Arsenic removal from drinking water by ion exchange and activated alumina plants. EPA 600-R-00-088, USEPA, Cincinnati, Ohio, USA. The US Environmental Protection Agency. (2000b). Technologies and costs for removal of arsenic from drinking water. EPA 815-R-00-028, USEPA, Washington DC, USA. The US Environmental Protection Agency. (2001). National primary drinking water regulations: arsenic and clarications to compliance and new source contaminant monitoring; nal rule. Federal Register, 66(14), 69757066.

Conclusions The comparisons of the efciency of the adsorbents (AA, MAA, GFH, GFO, TiO2) on As(V) removal were made under the various conditions. The rate of As(V) adsorption at pH 6.5 was faster than at pH 8.5. The batch experimental results suggested that aluminum-based adsorbents (AA, MAA) had a lower afnity for As(V) than iron-based (GFH, GFO) and titanium-based adsorbents (TiO2). In the laboratory column experiments, the titanium-based adsorbent (TiO2) had more effectiveness on As(V) removal at pH 6.5 than iron-based adsorbents (GFH, GFO), whereas the effectiveness of TiO2 and GFO were not much different at 8.5. In the eld ltration experiments, the breakthrough of GFO was observed at approximately 58,000 bed volumes while a breakthrough occurred at 41,500 bed volumes for TiO2, indicating that GFO had a much higher effectiveness for As(V) in the groundwater than in the challenge water containing phosphate because the As(V) competed with phosphate on the surface of the GFO. Experimental results demonstrated that GFO and TiO2 were better adsorbents for As(V) removal than the GFH, MAA, and AA. In addition, the results of the TCLP leaching tests indicated that the spent GFO and TiO2 used in the eld experiments were considered as nonhazardous wastes. These results showed that GFO and TiO2 could be used in POE and POU units for the effective removal of As(V).
Acknowledgments The authors would like to thank the New Jersey Department of Environmental Protection for the support of this research. We also would thank Christine Chin Choy for her assistance with the BET analysis.

References
Arai, Y., Elzinga, E. J., & Sparks, D. L. (2001). X-ray absorption spectroscopic investigation of arsenite and arsenate adsorption at the aluminum oxide-water interface. Journal of Colloid and Interface Science, 235, 8088. Badruzzaman, M., Westerhoff, R., & Knappe, D. R. U. (2004). Interparticle diffusion and adsorption of arsenate onto granular ferric hydroxide (GFH). Water Research, 38, 40024012. Bang, S., Koratis, G. P., & Meng, X. (2005a). Removal of arsenic from water by zero-valent iron. Journal of Hazardous Materials, 121, 6167. Bang, S., Patel, M., Lippincott, L., & Meng, X. (2005b). Removal of arsenic from groundwater by granular titanium dioxide adsorbent. Chemosphere, 60, 389397.

123

Environ Geochem Health (2011) 33:133141 Thirunavukkarasu, O. S., Viraraghavan, T., & Subramanian, K. S. (2003). Arsenic removal from drinking water using granular ferric hydroxide. Water South African, 29, 161170. Westerhoff, P., Higheld, D., Badruzzaman, M., & Yoon, Y. (2005). Rapid small-scale column tests for arsenate removal in iron oxide packed bed columns. Journal of Environmental Engineering, 131, 262271.

141 Wingrich, H., & Wolf, C. (2002). Research on the removal of arsenic from drinking water by occulation with iron solutions and adsorption with granulated iron hydroxides (GEH). Journal of Water Supply: Research and Technology-Aqua, 51, 273281.

123