Abstract
Abstract
Arsenic exists ubiquitously in both groundwater and surface water and it is extremely harmful to human health. In this study, basic yttrium carbonate (BYC) was successfully synthesized using the one-step precipitation method and for the first time, we report the removal of arsenate [As(V)] from aqueous solution by BYC through fixed-bed column. Effects of various parameters, including BYC dosage, As(V) concentration, flow rate, solution pH, and coexisting anions, on adsorption efficiency were investigated. Results demonstrate that the adsorption rate and capacity of As(V) increased with decreasing BYC dosage and with an increase of As(V) concentration or flow rate. Removal of As(V) was pH dependent and optimum adsorption efficiency was achieved at pH 6. Presence of NO3−, SO42−, and CO32− had a weak influence on removal of As(V); however, PO43− significantly inhibited As(V) adsorption, suggesting that As(V) was specifically adsorbed by BYC. Under acidic and neutral conditions, ligand exchange between As(V) and hydroxyl groups or carbonate on the surface of BYC was mainly responsible for As(V) removal, while the precipitation of As(V) and Y3+ ions released from BYC partially contributed to removal of As(V). Under alkaline conditions, however, no precipitation reaction was observed. Our results reveal the underlying mechanism of As(V) removal by BYC and provide a valuable reference for practical application.
Introduction
A
Arsenic is a carcinogen for humans, and the WHO recommended the limit of As in drinking water to be 10 μg/L. According to this, more than 100 million people are globally at risk (Singh et al., 2015). Therefore, it is urgent to explore effective techniques for arsenic removal from groundwater. The adsorption method has become one of the most effective methods of arsenic removal due to advantages of convenience, high efficiency, and low cost. There are a number of adsorbents that have been developed for arsenic uptake, such as activated carbon (Chen et al., 2007), activated alumina (Singh and Pant, 2004), zeolites (Xu et al., 2002), titanium dioxide (Ferguson et al., 2005; Xu et al., 2007), and iron oxides (Lafferty and Loeppert, 2005; Jang et al., 2006). In addition, some novel bimetal oxides, such as Fe–Mn binary oxide and Fe–Cu bimetal oxide, have also been reported for arsenic adsorption (Zhang et al., 2007a, 2007b, 2013).
Recently, adsorbents based on rare-earth-metal elements for the removal of anionic contaminants have gained widespread concern (Zhang et al., 2005, 2014; Li et al., 2010; Yu et al., 2015). These rare-earth-metal-containing adsorbents commonly exhibit a higher anion adsorption capacity. Deng and Yu (2012) synthesized a cerium impregnating fibrous protein to adsorb anions through batch experiments, and the maximum adsorption capacities of As(V), phosphate, and fluoride were reported as 172.3, 65.3, and 106.2 mg/g, respectively. A highly ordered mesoporous silica, impregnated with lanthanum, was used for As(V) uptake, whose adsorption capacity (123.7 mg/g) was ∼14 times higher compared with La(III)-modified silica gel (Jang et al., 2004). As the most abundant element of rare-earth-metals, yttrium-based adsorbents for anion removal are anticipated to bear potential. It has been reported that basic yttrium carbonate (BYC) exhibited an extremely high adsorption capacity for the removal of As(III)/As(V) and phosphate and was easily regenerated for reuse (Wasay et al., 1996; Haron et al., 1997). Recently, Y–Mn bimetal composites were successfully synthesized and employed to remove As(V) from aqueous solutions (Yu et al., 2015). Hydroxyl groups (M–OH) of the adsorbent have been suggested to play a primary role in the efficient removal of As(V) (Yu et al., 2015).
However, most of the studies on arsenic adsorption in the literature were conducted in batch experiments, and the removal of arsenic in a fixed-bed column has been rarely reported (Kundu and Gupta, 2007; Maji et al., 2012). In fact, continuous flow columns with prominent advantages over batch experiments are more useful for large-scale wastewater treatment. Based on these advantages of BYC mentioned above, this study focused on evaluating the performance of BYC for As(V) adsorption in a fixed-bed column. The effects of different parameters, including flow rate, BYC dosage, initial As(V) concentration, solution pH, and competing anions, on the adsorption efficiency of As(V) were also investigated. Combining the determination of the effluent pH and CO32− concentration with a characterization of the sorbent, a process and mechanism of As(V) uptake were proposed.
Materials and Methods
Materials
Y(NO3)3 · 6H2O, NaHCO3, and NaNO3 were purchased from Aladdin, Inc., Beijing Chemical Works, and Shanghai Zhenxin Reagent Factory, respectively. Urea, NaOH, and Na2SO4 were obtained from Xilong Chemical Co., Ltd. HCl and NaH2PO4 · 2H2O was bought from Nanjing Chemical Reagent Co., Ltd.
All chemical reagents were of analytical reagent grade and were used without further purification. All solutions were prepared in deionized water. As(V) stock solution with an As(V) concentration of 1000 mg/L was obtained by dissolving the required amount of Na2HAsO4 · 7H2O (Sigma-Aldrich) in deionized water and freshly diluted according to experimental requirements.
Preparation of BYC
BYC was prepared by homogeneous precipitation of Y(NO3)3 · 6H2O and CO(NH2)2 according to the method described by Wasay et al. (1996) and Haron et al. (1997). Briefly, Y(NO3)3 · 6H2O (0.02 mol) and CO(NH2)2 (0.05 mol) were dissolved in 100 mL of deionized water and then the mixture was slowly heated to ∼90°C in the pH range of 6.5–7.0. The obtained precipitate was filtered and washed with deionized water until the filtrate pH was nearly neutral. BYC was finally dried at 60°C in a vacuum drying oven.
Fixed-bed column study
The experimental setup in the fixed-bed column study is shown in Supplementary Fig. S1. The adsorption was conducted in a glass column (internal diameter: 6 mm, length: 350 mm), packed with BYC between two supporting layers of quartz sand, and then cotton was packed to prevent the floating of BYC and quartz sand in the As(V) effluent [our control experiment demonstrated that the effect of cotton on As(V) adsorption can be completely neglected]. The influent As(V) solution was pumped through a peristaltic pump into the column bed in the upflow mode, and the effluent was collected with a fraction collector. To establish optimal conditions for As(V) removal in the column study, the adsorption kinetics of As(V) at different initial solution pH (4, 6, 8, and 10), As(V) concentrations (20, 50, 100, and 150 mg/L), BYC dosages (0.05, 0.1, and 0.2 g), and flow rates (0.5, 1.0, and 1.5 mL/min) were investigated at room temperature. Furthermore, the effect of different oxyanions, including NO3−, SO42−, CO32−, and PO43− on As(V) removal, was also performed at 100 mg/L As(V) solution with a flow rate of 1.0 mL/min at pH 6.
The amount of As(V) adsorbed by BYC [
Where ‘i’ is the number of sample points, ti (min) is the ith time point,
Because the adsorbent (BYC) used in this study was in a small amount (0.1 g), which was mixed with sands, and the adsorption could not reach saturation, breakthrough curves were not the typical ones. We did not obtain some valuable data from these breakthrough curves. Instead, we used the curves plotting adsorption capacity against adsorption time to exhibit our results.
Analytic methods
A PB-10 acidity meter (Sartorius Scientific Instruments Co., Ltd.) was used to measure the pH values in different systems after three-point calibration. The concentration of As(V) was analyzed by colorimetric method based on molybdenum blue (Dhar et al., 2004). The concentration of carbonate in the effluent of the dissolution of BYC was also determined through a Shimadzu TOC-L analyzer. Y3+ concentration was measured according to the method reported by Qin et al. (2016). Fourier transform infrared spectroscopy (FTIR) spectra of BYC before and after As(V) adsorption were carried out using a Nicolet 380 FT-IR spectrophotometer (Thermo Scientific).
Results and Discussion
Effects of various parameters on As(V) adsorption
BYC dosage
Adsorption is usually dependent on the sorbent dosage as it determines the number of active sites available for the adsorption procedure (Bhaumik et al., 2013). Thus, the effect of BYC dosage on As(V) adsorption in the fixed-bed column was investigated by varying the mass of BYC from 0.05 to 0.2 g and at constant As(V) concentration of 100 mg/L at pH 6 with a flow rate of 1.0 mL/min. It is evident from Fig. 1 that the adsorption capacity increased with decreasing BYC dosage. This is due to a decrease of binding sites at a low dosage of BYC and the adsorbent easily reaches the saturation state of adsorption. When the mass of BYC was 0.05, 0.1, and 0.2 g, the adsorption capacity of As(V) within 480 min was 269.59, 279.51, and 168.86 mg/g, respectively. However, the corresponding total As(V) adsorbed was 13.48, 27.95, and 33.77 mg, respectively, which is due to more BYC in the fixed-bed column providing more active sites for As(V) adsorption (Baral et al., 2009).

Effect of BYC dosage on As(V) adsorption in a fixed-bed column. Experimental conditions: solution pH = 6.0, c[As(V)] = 100 mg/L, and flow rate = 1.0 mL/min. BYC, basic yttrium carbonate; As(V), arsenate.
As(V) initial concentration
The effect of initial As(V) concentrations ranging from 20 to 150 mg/L on the performance of BYC for As(V) adsorption in the fixed-bed column was investigated at a flow rate of 1.0 mL/min and at pH 6 with a dosage of BYC (0.1 g). The plots of adsorption capacity of As(V) versus time are illustrated in Fig. 2. It can be seen that the adsorption capacity increased noticeably with rising initial As(V) concentration. In addition, the adsorption capacities were nearly identical by the end of the reaction with As(V) initial concentrations at 100 and 150 mg/L, which may be ascribed to limited binding sites at the dosage of BYC. The main driving force for adsorption originates in the difference of adsorbate concentrations between solution and surface of the sorbent (Aksu and Gönen, 2004). The adsorbate mass transfer was significantly influenced by the solute concentration (Xu et al., 2015), which clearly elucidates the different behaviors of As(V) adsorption that were observed in this study.

Effect of initial As(V) concentration on As(V) adsorption by BYC in a fixed-bed column. Experimental conditions: solution pH = 6.0, BYC dose = 0.1 g, and flow rate = 1.0 mL/min.
Flow rate
Figure 3 shows the effect of flow rate (0.5–1.5 mL/min) on As(V) adsorption by BYC in a fixed-bed column. A sharper sorption curve for As(V) adsorption was obtained at a higher flow rate. According to the results illustrated in Fig. 3, the amount of total As(V) adsorbed by BYC was almost identical at a flow rate of 0.5 and 1.0 mL/min (133.77 and 134.63 mg/g, respectively); however, it decreased with an increasing flow rate when the volume of As(V) solution flowing through the column was 180 mL. The adsorption capacity decreased from 134.63 mg/g at a flow rate of 1.0 mL/min to 119.85 mg/g at a flow rate of 1.5 mL/min. Kundu and Gupta (2007) reported a similar result. This can be explained by the fact that the reduced residence time causes a deficient distribution of the adsorbate in the column at higher flow rates. Hence, the increase of contact time favors As(V) removal.

Effect of flow rate on As(V) adsorption by BYC in a fixed-bed column. Experimental conditions: solution pH = 6.0, c[As(V)] = 100 mg/L, and BYC dose = 0.1 g.
Solution pH
Effect of influent pH (4, 6, 8, or 10) on the adsorption of As(V) was investigated and the results are presented in Fig. 4. Removal of As(V) is highly pH dependent. Optimal removal efficiency of As(V) was realized at pH 4 and 6; however, the adsorption capacity of As(V) was getting lower with increasing solution pH from 8 to 10. The solution pH affects both surface charge of adsorbent and the species of As(V). It has been reported that H2AsO4− and HAsO42− are predominant As(V) species in pH ranging from 4 to 10 (Hu et al., 2012). However, the point of zero charge (pHpzc) of BYC is ∼7.5 (Wasay et al., 1996), it can be inferred that the protonation of the functional group (Y–OH) on the surface of BYC occurs at pH < 7.5, which is favorable for As(V) adsorption by electrostatic attraction between positively charged binding sites of BYC and As(V) anions. In contrast, the surface of BYC becomes negatively charged at pH > 7.5. Here, the adsorption of As(V) will decrease due to electrostatic repulsion. The phenomenon indicates that electrostatic interaction plays an important role in the BYC-mediated adsorption of As(V).

Effect of influent pH on As(V) adsorption by BYC in a fixed-bed column. Experimental conditions: c[As(V)] =100 mg/L, BYC dose = 0.1 g, and flow rate = 1.0 mL/min.
Coexisting anions
There are various oxyanions coexisting in natural water, such as NO3−, SO42−, CO32−, and PO43−, and these may compete with As(V) ions for the binding sites of the sorbent (Su and Puls, 2001; Jia and Demopoulos, 2005; Kanematsu et al., 2011; Cui and Weng, 2013). The effect of competing anions on As(V) adsorption by BYC in a fixed-bed column is displayed in Fig. 5. NO3−, SO42

Effect of competing anions on As(V) adsorption by BYC in a fixed-bed column. Experimental conditions: solution pH = 6.0, c[As(V)] = 100 mg/L, BYC dose = 0.1 g, and flow rate = 1.0 mL/min.
It has been reported that nonspecifically adsorbing nitrate is adsorbed through outer-sphere adsorption, while sulfate is removed by both outer-sphere adsorption and inner-sphere complexation (Jia and Demopoulos, 2005). Lefèvre (2004) reported that sulfate could be predominantly removed by outer-sphere adsorption at pH > 6. In this study, despite the initial pH = 6, the final pH increased (discussed later). Both As(V) and phosphate can form inner-sphere complexes with active functional groups of the adsorbent (Arai et al., 2005). Hence, the As(V) that has been specifically adsorbed by inner-sphere complexes with hydroxyl groups on the surface of BYC is rarely inhibited by nonspecifically adsorbing nitrate and sulfate; however, it is strongly suppressed by specifically adsorbing phosphate. Zhang et al. (2003) observed similar adsorption behaviors of phosphate and As(V), which is ascribed to both H3AsO4 and H3PO4 being triprotic acids with similar ionization constants. Although carbonate can also form inner-sphere complexes with hydroxyl groups on the surface of the adsorbent, the affinity of carbonate for the surface hydroxyl groups is much lower compared with As(V) (Brechbühl et al., 2012). Consequently, it cannot noticeably influence As(V) adsorption by BYC in a fixed-bed column.
Variation of effluent pH and CO32−
To further understand the adsorption process and mechanism, pH and CO32− concentration in the effluent were analyzed. Supplementary Figure S2 exhibits the difference of the effluent pH values between As(V) solution and deionized water at initial pH values of 4 and 10. The effluent pH of both As(V) solution and deionized water was higher than their initial pH 4 throughout the whole reaction process. The reasons are that hydroxyl groups on the surface of BYC were protonated and BYC was partly dissolved in the weak acidic solution. The effluent pH of deionized water (at initial pH = 4) slowly increased from 5.05 at the beginning to 6.30 at the end. The effluent pH of 100 mg/L As(V) solution with an initial pH 4 was approximately equal to the effluent pH of the deionized water at the beginning of the adsorption process (0–50 min), and then abruptly increased to 7.32. Subsequently, it slowly decreased again up to the effluent pH of the deionized water. Based on these results, it can be inferred that H2AsO4− ions are first adsorbed by the protonated hydroxyl groups (Y–OH2+) of the adsorbent through electrostatic attraction, and then H2AsO4− ions form complexes through ligand exchange between H2AsO4- and the protonated hydroxyl groups (−OH2+)/hydroxyl groups (−OH) on the surface of BYC. Furthermore, OH- was released into the solution, resulting in rising of the effluent pH of the As(V) solution. With consumption of the active sites on BYC, the effluent pH began a gradual decline.
In contrast, Supplementary Fig. S2 reveals that the effluent pH of both As(V) solution and deionized water at initial pH 10 were lower than their influent pH, caused by the deprotonation of the hydroxyl groups on the surface of BYC (releasing H+ ions into the effluent). However, the effluent pH of the As(V) solution was much higher compared with deionized water through the whole adsorption process, which can be attributed to the ligand exchange between HAsO42- ions and hydroxyl groups as previously discussed.
Carbonate released into the effluent solution during the As(V) adsorption process was also determined and the results are illustrated in Supplementary Fig. S3. As shown in Supplementary Fig. S3, carbonate ions were hardly released into the effluent solution within 60 min, while the As(V) adsorbed onto BYC increased rapidly to 0.039 mmol. Subsequently, with continuing adsorption of As(V), CO32− was gradually released into the effluent. Nevertheless, only 0.018 mmol CO32− were released into the effluent by the end of the reaction, even though 0.084 mmol As(V) was adsorbed by BYC during the same period. Considering that no CO32− was detected in the effluent when deionized water with an initial pH 10 leached the fixed-bed column instead of As(V) solution (data not shown), it can be inferred that ligand exchange between As(V) and carbonate may also contribute to As(V) adsorption.
Precipitation of As(V) and Y3+
Considering that BYC may partly dissolve at pH below 6.5, releasing Y3+ into the solution (Wasay et al., 1996), the precipitation reaction between As(V) and Y3+, which is beneficial to As(V) removal, was measured. Supplementary Figure S4 revealed that more than 30% of initial As(V) was removed from the solution in the coexistence of As(V) and Y3+ at pH 6 due to precipitation. Nevertheless, As(V) concentration at pH 4 decreased by only 7.7%. A possible reason is that a low pH is not favorable for the formation of the precipitate between As(V) and Y3+. Based on these results, it can be concluded that the precipitation of As(V) and Y3+ will partially contribute to the removal of As(V) by BYC in an acidic solution [no Y3+ was detected in an alkaline BYC solution, which fully agrees with the report by Wasay et al. (1996)].
FTIR analysis
Figure 6 displays the FTIR spectra of BYC before and after As(V) adsorption. In the spectrum of virgin adsorbent, the bands at 3433 cm−1 and 1635 cm−1 are assigned to the stretching vibration of O–H within BYC and the bending vibration of H2O, which is physically adsorbed on the surface of BYC (Li et al., 2012), respectively. The two peaks at 1516 and 1385 cm−1 are ascribed to the vibration of the carbonate groups (Yu et al., 2015) and the band at 1051 cm−1 corresponds to the bending vibration of the hydroxyl groups (Y–OH) on the BYC surface (Zhang et al., 2005). For As(V)-loaded BYC, both peaks of the carbonate groups decreased noticeably, further verifying that the carbonate groups on the surface of BYC were partially replaced by As(V). The intensity of the peak at 1051 cm−1 almost completely disappeared after As(V) adsorption, revealing that hydroxyl groups (Y–OH) are mainly responsible for As(V) adsorption through ligand exchange between hydroxyl groups and As(V). Furthermore, the appearance of two new peaks after As(V) adsorption at 876 and 816 cm−1 should be assigned to the vibration of Y–O–As groups (Zhang et al., 2014; Yu et al., 2015). This further confirms that As(V) mainly bonded as a surface complex.

Fourier transform infrared spectroscopy spectra of
According to these results and discussion, the mechanism of As(V) removal by BYC in a fixed-bed column is proposed: under acidic and neutral conditions (pH < pHpzc of BYC), As(V) is adsorbed by both adsorption and precipitation. However, at pH > pHpzc of BYC, the removal of As(V) only participates in the adsorption process. For clarity, the detailed mechanism of As(V) removal by BYC is depicted in Fig. 7.

Mechanism of As(V) adsorption:
Conclusions
BYC obtained through homogeneous precipitation was utilized as an adsorbent to remove As(V) from aqueous solution in a continuous fixed-bed column. These investigations demonstrated BYC to be a very efficient adsorbent for As(V) with an extremely high adsorption capacity. The removal efficiency of As(V) depended on BYC dosage, initial As(V) concentration, flow rate, solution pH, and competing anions. The optimal removal efficiency of As(V) could be realized at pH 6. The presence of NO3−, SO42−, and CO32− had a limited impact on As(V) removal. However, PO43− strongly competed with As(V) for the active sites on the adsorbent, greatly suppressing As(V) adsorption by BYC. Under acidic and neutral conditions, both adsorption and precipitation contributed to As(V) removal, while only adsorption was responsible for As(V) removal under alkaline conditions. The hydroxyl groups on the BYC surface played an indispensable role in the adsorption of As(V), and the ligand exchange between hydroxyl groups and As(V) ions resulted in a strong adsorption of As(V). Furthermore, ion exchange between As(V) and carbonate on the surface of BYC contributed to the removal of As(V) to a certain degree.
Footnotes
Acknowledgment
This study was supported by the National Natural Science Foundation of China (Grant No. 21637003 and 21607076).
Author Disclosure Statement
No competing financial interests exist.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
