As(III) Removal via Combined Addition of Mg- and Ca-Based Adsorbents and Comparison to As(V) Removal via Those Mechanisms

Abstract

Damage to human health caused by As-contaminated water can be prevented using proper As-removal techniques, such as employing excellent arsenic adsorbents. In this study, the combined addition of Mg- and Ca-based adsorbents was investigated for the efficient removal of As from contaminated water. Following a previous study on As(V), As-removal tests targeting As(III) and several additional tests, including X-ray diffraction analysis, were conducted to clarify the mechanism of the improved performance of the combined-addition As removal. Similarly as for As(V), the combined additions of both MgCO₃ + CaO and MgCO₃ + Ca(OH)₂ improved As(III)-removal performance while inhibiting the leaching of base material components; however, they did not remove As(III) as effectively as As(V). The differences in the removal ratios of As(V) and As(III) in these combined additions were concluded to be primarily due to the different As-removal mechanisms. Mg(OH)₂ and CaCO₃ were generated, and As(III) was incorporated into the generated precipitate of Mg(OH)₂ but not into that of CaCO₃. Conversely, As(V) was incorporated into both Mg(OH)₂ and CaCO₃. Additionally, MgCO₃ + Ca(OH)₂ was evaluated as a more efficient combined-addition method because MgCO₃ + Ca(OH)₂ exhibited a higher As-removal ratio value than MgO + CaO. Proposals have been made to remove As(III) using activated carbon modified with heavy metals or transition elements, or concrete waste grafted with polymers, but these methods are complicated to prepare, costly, and involve the risk of leaching of harmful components. Adsorbents that use general Mg and Ca components as their base material do not contain such harmful components. The Mg- and Ca-based adsorbents are readily available and low-cost, and, best of all, there is no concern that they will leach harmful components. Therefore, widespread use of Mg- and Ca-based adsorbents as a measure against arsenic contamination could greatly contribute to a sustainable society.

1. Introduction

The contamination of As in groundwater and well water has become a global problem, and several researchers have reported on the state of its contamination. In 2020, Kobya et al. [1] presented a tabulated review of As concentrations in groundwater in Asia (Bangladesh, Pakistan, China, India, Nepal, Vietnam, Taiwan, and Thailand), America (USA, Canada, Mexico, Argentina, and Chile), Africa (South Africa and Ghana), Europe (Hungary, Serbia, Croatia, and Greece), and Oceania (Australia). Moreover, in 2019, Abiye and Bhattachaya [2] reported maximum As concentrations of 6150 μg/L west of Johannesburg, South Africa. These extremely high levels of As contamination are probably due to the release of As from minerals such as arsenopyrite, arsenic oxide, sulfharsenide, arsenopyritic reefs, leucopyrite, löllingite, and scorodite contained in aquifer rocks [2]. Numerous studies have reported on the adverse effects of exposure to As-contaminated water on human health [3]. For example, ingesting inorganic As causes various skin lesions, skin cancers, and other internal cancers, such as lung, bladder, liver, and kidney cancers [3]. To address these harmful effects, the World Health Organization (WHO) guidelines for drinking water quality have set a provisional As guideline value of 0.01 mg/L [4]. In Japan, the environmental and effluent standards for As are 0.01 and 0.1 mg/L, respectively [5,6].

Most health damage caused by As can be prevented by properly treating As-contaminated water, prompting several researchers to propose and review various As-removal techniques [7,8,9,10]. These methods can be divided into ion exchange, adsorption, filtration, oxidation, and electrocoagulation methods. Among them, adsorption method is generally easy and has low operating costs, making it the method most easily adopted in economically disadvantaged developing countries. Even quite recently, there have been many reports of research on As(III) removal using new adsorbents. Chuc et al. (2024) studied the As(III) removal using zirconia (ZrO₂)-activated carbon (AC) composite synthesized from AC and ZrO₂ nanoparticles by a hydrothermal process [11]. The adsorption of arsenite onto a ZrO₂-AC surface was studied by X-ray photoelectron spectroscopy (XPS) spectra. It was reported that the results confirmed the simultaneous adsorption of both As(V) and As(III) on the surface of ZrO₂-AC materials. They explained that this was because some of the As(III) was oxidized to As(V) during the experiment, and As(III) and As(V) were adsorbed simultaneously. They also reported that with increasing solution pH, the adsorption efficiency of arsenite onto ZrO₂-AC was maximal at a pH of about 5 and then decreased at higher pH. They explained that at pH > 5, the negative charges of ZrO₂-AC and the fraction of arsenite oxoanions increased, resulting in an enhanced electrostatic repulsion between arsenite and the ZrO₂-AC. Furthermore, they reported that the adsorption data fit well to either the Langmuir or Freundlich isotherm, and the maximum monolayer adsorption capacity for arsenite on ZrO₂-AC is 64 mg/g [11]. Yadegari et al. (2024) developed an adsorbent grafted with allyl alcohol and vinyl acetate polymers to concrete waste for As(III) removal [12]. They named their adsorbent CW@MPTMS@poly(AA-co-VyAc)@PABA. In their adsorption test with pH as a parameter, the removal efficiency of As(III) improved with an increase in the pH of the solution from 5 to 7. In their adsorption test with temperature as a parameter, the removal efficiency improved from 55 to 75% with a decrease in temperature from 45 to 25 °C. Therefore, they suggested that the As(III) sorption reaction was exothermic by CW@MPTMS@poly(AA-co-VyAc)@PABA. The Freundlich isotherm model with their data provided a better fit to experimental results than the Langmuir, Temkin, and Dubinin–Radushkevich. The maximum As(III) sorption capacity of CW@MPTMS@poly(AA-co-VyAc)@PABA was assessed from the non-linear Langmuir isotherm model by them, which comprised 6.818 mg/g. Furthermore, the evaluation with the Dubinin-Radushkevich model indicated that the As(III) sorption onto CW@MPTMS@poly(AA-co-VyAc)@PABA was physical adsorption. They inferred that hydrogen bonds were mainly generated between O–H functional groups on the As(III) ion and NH₂ and O–H groups of CW@MPTMS@poly(AA-co-VyAc)@PABA, which contributed to the adsorption process and enabled the high removal efficiency [12]. Choudhury et al. (2025) studied As(III) removal using bagasse-manganese-aluminum (B-Mn-Al), a sugarcane bagasse-derived biochar impregnated with Mn and Al [13]. They revealed that 34.70 wt% of the B-Mn-Al was metal, of which 6.12% was Al and 28.58% was Mn by energy-dispersive X-ray (EDS) analysis. Additionally, analysis of the X-ray diffraction (XRD) spectrum indicated the presence of Mn and Al as aluminum manganese (Al₆Mn). It also verified the successful adsorption of As as manganese arsenide (Mn₂As) on the surface. They reported that by maintaining a contact time of 65 min, the adsorption percentage of As(III) jumped from 44.96% at an adsorbent dosage of 0.25 g/L to 91.43% at a dosage of 1.375 g/L adsorbent. As the solution pH increased from 2 to 9, the redox reaction between As(III) and manganese (Mn) oxides was suppressed, leading to a noticeable decrease in As(III) removal. They considered that the adsorption of As(III) on B-Mn-Al adsorbent surface occurred in two routes; one was electrostatic adsorption, and the other one was through ion or ligand exchange between the adsorbate arsenate ion and the active functional groups on the adsorbent surface. The data analysis showed that the correlation coefficient for the Langmuir isotherm was higher than that for the Freundlich isotherm. As(III) adsorption capacity of the adsorbent calculated by them was 55 mg/g [13].

Both Mg- and Ca-based adsorbents [14,15,16,17,18,19,20,21], which are safe for human health, are recommended as environmentally friendly and sustainable adsorbents [22]. A previous study examined the As(V)-removal performance of various combinations of Mg- and Ca-based adsorbents [22]. Both MgCO₃ + CaO and MgCO₃ + Ca(OH)₂ combined additions were the most effective for removing As(V) among the combinations tested. Additionally, these combinations significantly inhibited the leaching of Mg and Ca from the adsorbent base material. However, whether the same improvements are achieved for As(III) has not yet been confirmed. In this study, As(III) removal tests were performed corresponding to the As(V) removal tests performed in a previous study [22]. The As(III)-removal performances of various combinations of Mg- and Ca-based adsorbents were investigated, and the mechanisms of As-removal improvement and Mg- and Ca-leaching inhibition by the MgCO₃ + CaO and MgCO₃ + Ca(OH)₂ combined additions were experimentally investigated. Additionally, X-ray diffraction (XRD) analysis was carried out to verify the generation of Mg(OH)₂ and CaCO₃ in the combined addition test of MgCO₃-CaO or MgCO₃-Ca(OH)₂ that was inferred in a previous study [22]. The solid samples for XRD analysis were prepared in the absence of As, in the presence of As(V) or As(III). Furthermore, to verify the incorporation of As into the produced Mg(OH)₂ or CaCO₃, ‘Mg(NO₃)-addition tests’ and ‘Ca(NO₃)₂- and Na₂CO₃-addition tests’ were performed using alkaline As(III) and As(V) solutions. Similarly to a previous study on As(V) removal [22], investigating how the combination of different types of adsorbents affects As(III) removal will also provide important guidance for designing new high-performance treatments.

2. Materials and Methods

The reagents used in this study were purchased from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan), unless specified otherwise.

2.1. Mg- and Ca-Based Adsorbents

Following a method presented in a previous study [22], MgO, Mg(OH)₂, and MgCO₃ were employed as the Mg-based adsorbents, while CaO, Ca(OH)₂, and CaCO₃ were used as the Ca-based adsorbents. The specific characteristics of each adsorbent (Mg and Ca contents, reagent purity, median particle diameter, and Brunauer–Emmett–Teller surface area) are described by Sugita et al. [22]. The median particle size D_p50 (μm), the Brunauer–Emmett–Teller (BET) surface area S_BET (m²/g), the measured Mg content α_Mg (%), the measured Ca contents α_Ca (%), and the reagent purity (obtained from α_Mg and α_Ca) P (%) are shown in

Table 1. Characteristics of Mg-based and Ca-based adsorbents used in this study.

Adsorbent	Dp50 (μm)	SBET (m2/g)	αMg (%)	αCa (%)	P (%)
MgO	1.54	4.3	59.1	-	98.0
Mg(OH)2	4.13	22.0	40.6	-	97.3
MgCO3	15.0	26.0	24.8	-	86.1
CaO	19.6	2.7	-	71.2	99.6
Ca(OH)2	41.7	14.3	-	53.5	98.9
CaCO3	15.4	0.8	-	39.7	99.1

Data from Sugita et al. (2023) [22].

. The data in the table were obtained from a previous study [22].

2.2. As(III) Removal Tests Using Mg- and Ca-Based Adsorbents

2.2.1. Preparation of Synthetic As(III)-Contaminated Water

Powdered sodium arsenite (NaAsO₂, 90%) was dissolved in deionized water, and stock solutions of As(III) (2000 mg As/L) were prepared. The synthetic As(III)-contaminated water was prepared by diluting the As(III) stock solution with deionized water. The initial As concentrations, C_AS0, were 1 and 10 mg/L. The pH of the solution was adjusted to approximately 7 using aqueous HNO₃ and NaOH. The pH immediately after pH adjustment is referred to as the initial pH, pH₀. Incidentally, the reason why the value of pH₀ was set to 7 in this study is because groundwater (well water), which usually has a pH close to neutral, is used for drinking and farming.

2.2.2. Experiment Procedure

The experimental procedure for the As(III)-removal test was the same as that used for the As(V)-removal test in a previous study [22], except that the synthetically contaminated water used was different. One or two types of adsorbents were weighed into a 50 mL polypropylene centrifuge tube. The amount of each adsorbent was set as the adsorbent concentration when 50 mL of the liquid was added. The centrifuge tube containing the adsorbent was immediately sealed after adding 50 mL of the fabricated synthetic As(III)-contaminated water and shaken in a constant-temperature shaker (approximately 180 rpm at 20–25 °C). Subsequently, the tubes were centrifuged for 24 h, and the supernatant was filtered through a syringe filter with a pore diameter of 0.45 μm and collected in a polypropylene vessel. The pH of the filtrate (treated water) was immediately measured and referred to as the final pH (pH_f). As, Mg, and Ca in the treated water were quantified using inductively coupled plasma mass spectrometry (ICP-MS; Agilent 7700X, Agilent Technologies, Inc., Santa Clara, CA, USA) and inductively coupled plasma atomic emission spectroscopy (ICP-AES; SII SPS3500DD, Seiko Instruments Inc., Chiba, Japan), and were denoted as C_AS, C_Mg, and C_Ca, respectively. In the single-addition tests, the two adsorbent concentrations (W_Ad/V) were set at 0.2 and 0.4 g/L. In the combined-addition tests with two types of adsorbents, the W_Ad/V of each adsorbent was 0.2 g/L, with a total adsorbent concentration (ΣW_Ad/V) of 0.4 g/L.

2.3. Experimental Reproducibility

To confirm the reproducibility of the experiments, As(III)-removal tests were conducted thrice under specific conditions for each adsorbent. The experimental conditions were C_AS0 = 1.007 mg/L, pH₀ = 7.27, and W_Ad/V = 0.2 g/L. The measured values and standard errors are listed in

Table 2. Mean values and standard errors obtained from three replicated experiments.

Adsorbent	WAd/V (g/L)	CAS (mg/L)	CMg (mg/L)	CCa (mg/L)	pHf
MgO	0.203 ± 0.001	0.632 ± 0.065	8.07 ± 0.71	-	10.84 ± 0.03
Mg(OH)2	0.201 ± 0.001	0.561 ± 0.025	4.84 ± 0.05	-	10.47 ± 0.01
MgCO3	0.201 ± 0.001	1.001 ± 0.003	24.3 ± 0.0	-	10.64 ± 0.00
CaO	0.207 ± 0.002	0.980 ± 0.009	-	125 ± 0	11.77 ± 0.00
Ca(OH)2	0.200 ± 0.000	0.955 ± 0.014	-	92.8 ± 1.1	11.72 ± 0.01
CaCO3	0.204 ± 0.002	1.003 ± 0.003	-	4.87 ± 0.01	9.64 ± 0.05

, which indicates that the experimental reproducibility of this study was reasonably good.

2.4. Preparation of Samples for X-Ray Diffraction (XRD) Analysis

To clarify the mechanisms of As removal in the combined additions of MgCO₃ + CaO and MgCO₃ + Ca(OH)₂, the following samples were prepared for XRD analysis.

(a)

Sample mixed with equal amounts of unused MgCO₃ and CaO;

(b)

Sample mixed with equal amounts of unused MgCO₃ and Ca(OH)₂;

(c)

Solid sample collected after adding MgCO₃ and CaO (each 0.4 g/L) to deionized water;

(d)

Solid sample collected after adding MgCO₃ and Ca(OH)₂ (each 0.4 g/L) to deionized water;

(e)

Solid sample collected after adding MgCO₃ and CaO (each 0.4 g/L) to the As(III) solution (C_AS0 = 10 mg/L and pH₀ = 7);

(f)

Solid sample collected after adding MgCO₃ and Ca(OH)₂ (each 0.4 g/L) to the As(III) solution (C_AS0 = 10 mg/L and pH₀ = 7);

(g)

Solid sample collected after adding MgCO₃ and CaO (each 0.4 g/L) to the As(V) solution (C_AS0 = 10 mg/L and pH₀ = 7);

(h)

Solid sample collected after adding MgCO₃ and Ca(OH)₂ (each 0.4 g/L) to the As(V) solution (C_AS0 = 10 mg/L and pH₀ = 7).

The As(III) solutions used for samples (e) and (f) were prepared as described in Section 2.2.1. The As(V) solution used for samples (g) and (h) was prepared by dissolving the Na₂HAsO₄·7H₂O powder reagent in deionized water and adjusting the solution pH with aqueous HNO₃ and NaOH. To prepare samples (c)–(h), 250 mL of the prepared solution (deionized water, As(III) solution, or As(V) solution) was placed into four different TPX beakers. Subsequently, 0.1 g of MgCO₃ and 0.1 g of CaO or Ca(OH)₂ were added to each beaker, followed by immediate stirring using a magnetic stirrer for 24 h. The samples were then subjected to solid–liquid separation via suction filtration (filter pore size: 0.45 μm). The collected solutions were combined into single containers, and the concentrations of As, Mg, and Ca in the collected solutions of (e)–(h) were determined using ICP-MS and ICP-AES. The collected solid samples were dried overnight at 40 °C, and the crystalline phases of the prepared samples were identified using powder XRD (RINT-2500, Rigaku Co., Akishima, Japan) at the GSJ-Lab (AIST).

2.5. Confirmation Tests for As Removal by Generated Mg(OH)₂ and CaCO₃

As described in Section 3.4, As(III) removal via the combined additions of MgCO₃ and CaO or MgCO₃ + Ca(OH)₂ was presumed to be attributed to the generated Mg(OH)₂ and/or CaCO₃. Therefore, several tests were performed to confirm whether the primary mechanism of As(III) and/or As(V) removal was the generation of Mg(OH)₂ or CaCO₃.

2.5.1. Mg(NO₃)₂-Addition Tests with Alkaline As Solutions

An As(III) solution (2100 mL; C_AS0 = 10 mg/L) with a pH₀ = 12 (adjusted using an aqueous NaOH solution) was prepared. Subsequently, 250 mL of the As(III) solution was placed into five separate TPX beakers, and 1 g of Mg(NO₃)₂·6H₂O powdered reagent (Wako 1st grade) was added to each beaker, followed by immediate stirring with a magnetic stirrer. Because Mg(NO₃)₂·6H₂O is a readily soluble substance, it quickly dissolved; however, new precipitates formed during stirring. After stirring for 24 h, the samples were subjected to solid–liquid separation via suction filtration (filter pore size: 0.45 μm), and the collected solutions were combined in a single container. The As and Mg contents in the collected solutions were determined using ICP-MS and ICP-AES, respectively. The precipitate was removed via filtration, dried overnight at 40 °C in a dryer, and collected in a container. The dried samples were analyzed using a powder XRD apparatus. The W_Ad/V value of Mg(NO₃)₂·6H₂O in the tests was 4 g/L, which corresponded to an Mg concentration of 0.38 g/L, assuming a reagent purity of 100%. However, the actual reagent purity was <100% because Mg(NO₃)₂·6H₂O is a deliquescent substance. Therefore, the value of W_Ad/V was determined by considering the following factors: collecting the amount of precipitate required for XRD analysis, assuming that Mg(NO₃)₂·6H₂O is converted to Mg(OH)₂, and correctly determining whether As has been removed or not. For As(V) removal, the tests were performed under the same conditions using an As(V) solution (C_AS0 = 10 mg/L and pH₀ = 12) instead of the As(III) solution.

2.5.2. Ca(NO₃)₂- and Na₂CO₃-Addition Tests with Alkaline As Solutions

In this series of experiments, the As(III) and As(V) solutions were prepared as described in Section 2.5.1. Briefly, 250 mL of the prepared As(III) solution was placed into three separate TPX beakers, and 0.67 and 0.30 g of Ca(NO₃)₂·4H₂O and Na₂CO₃ powder reagents (Wako 1st grade), respectively, were added to each beaker, followed by immediate stirring with a magnetic stirrer. Similarly to Mg(NO₃)₂·6H₂O, both Ca(NO₃)₂·4H₂O and Na₂CO₃ were readily soluble substances that quickly dissolved; however, new precipitates formed during stirring. After stirring for 24 h, the samples were subjected to solid–liquid separation via suction filtration (filter pore size: 0.45 μm). The collected solutions were combined in single containers. The As and Ca contents in the collected solutions were determined using ICP-MS and ICP-AES, respectively. The precipitate removed by the filter was dried overnight at 40 °C in a dryer and collected in a container. The dried samples were analyzed using a powder XRD apparatus. The W_Ad/V of Ca(NO₃)₂·4H₂O was 2.7 g/L, which corresponded to a Ca concentration of 0.46 g/L, assuming a reagent purity of 100%. However, similarly to Mg(NO₃)₂·6H₂O, the actual reagent purity was <100% because Ca(NO₃)₂·4H₂O is a deliquescent substance. The W_Ad/V of Na₂CO₃ was 1.2 g/L, which was an appropriate amount in terms of molar ratio for producing CaCO₃ relative to the amount of Ca added. The tests were performed under the same conditions using the As(V) solution instead of the As(III) solution.

The types of experiments performed in this study and the valences of As tested in them are outlined in

Table 3. Correspondence table between experiment types and As valences.

Experiment Type	As Valence Tested in This Study
As-removal tests using Mg- and Ca-based adsorbents	As(III)
Samples prepared for XRD analysis	As(III), As(V)
Mg(NO3)2-addition tests with alkaline As solutions	As(III), As(V)
Ca(NO3)2- and Na2CO3-addition tests with alkaline As solutions	As(III), As(V)

The data for As-removal tests using Mg- and Ca-based adsorbents on As(V) are provided from a previous study [22].

.

3. Results

3.1. As-Removal Ratio

3.1.1. As-Removal Ratio for Single Addition of One Type of Adsorbent

The values of C_AS0 and C_AS for the single-addition tests are listed in

Table 4. Values of C_AS0 and C_AS for the single-addition tests.

CAS0	1 mg/L		1 mg/L		10 mg/L		10 mg/L
WAd/V	0.2 g/L		0.4 g/L		0.2 g/L		0.4 g/L
	CAS0	CAS	CAS0	CAS	CAS0	CAS	CAS0	CAS
MgO	1.01	0.632	1.07	0.193	10.8	3.36	10.7	1.28
Mg(OH)2	1.01	0.561	1.06	0.406	10.8	9.35	10.7	7.95
MgCO3	1.01	1.00	1.05	1.04	10.8	10.1	10.6	10.3
CaO	1.01	0.980	1.07	1.07	10.8	10.8	10.7	5.25
Ca(OH)2	1.01	0.955	1.07	1.07	10.8	10.7	10.7	8.81
CaCO3	1.01	1.00	1.05	1.01	10.8	10.8	10.6	10.2

Unit [mg/L].

. The values for C_AS0 = 1 mg/L and W_Ad/V = 0.2 g/L were obtained from Table 2.

The values of C_AS (mg/L) were converted to the As-removal ratio R_AS (%) using Equation (1) to facilitate a comparison because the As(III)-removal tests were performed with two different initial As concentrations.

R_AS = (C_AS0 − C_AS)/C_AS0 × 100.

(1)

Figure 1 shows R_AS calculated using data on Table 4 for the single-addition As(III)-removal tests. Figure 1a and Figure 1b show the R_AS values for C_AS0 = 1 and 10 mg/L with W_Ad/V = 0.2 g/L, respectively. Figure 1c and Figure 1d show the R_AS values for C_AS0 = 1 and 10 mg/L with W_Ad/V = 0.4 g/L, respectively.

As shown in Figure 1, MgO and Mg(OH)₂ exhibited As(III)-removal abilities, and their R_AS values were higher when W_Ad/V was higher; however, MgCO₃ minimally removed As(III) under the test conditions. The Ca-based adsorbents could not remove As(III) under most conditions, as shown in Figure 1a–c. However, when C_AS0 = 10 mg/L and W_Ad/V = 0.4 g/L (Figure 1d), they exhibited a significant increase in R_AS values, particularly when the R_AS of CaO was >50%. The R_AS of each adsorbent varied depending on the combination of C_AS0 and W_Ad/V; however, MgO exhibited the highest As(III)-removal performance.

3.1.2. As-Removal Ratio for Combined Addition of Two Types of Adsorbents

The C_AS values for the combined-addition tests at C_AS0 = 1 and 10 mg/L are listed in Table 3 and Table 4, respectively. The measured values of C_AS0 for the combined-addition tests at C_AS0 = 1 and 10 mg/L were 1.07–1.11 and 10.7–10.8 mg/L, respectively. Figure 2 shows the R_AS values calculated using the data in

Table 5. Values of C_AS for the combined-addition tests at C_AS0 = 1 mg/L.

Combination	MgO	Mg(OH)2	MgCO3	CaO	Ca(OH)2	CaCO3
MgO	-	0.244	0.618	0.954	0.953	0.672
Mg(OH)2	0.244	-	0.814	0.911	0.937	0.585
MgCO3	0.618	0.814	-	0.744	0.613	1.03
CaO	0.954	0.911	0.744	-	1.06	1.04
Ca(OH)2	0.953	0.937	0.613	1.06	-	1.09
CaCO3	0.672	0.585	1.03	1.04	1.09	-

Unit [mg/L].

and

Table 6. Values of C_AS for the combined-addition tests at C_AS0 = 10 mg/L.

Combination	MgO	Mg(OH)2	MgCO3	CaO	Ca(OH)2	CaCO3
MgO	-	3.06	4.48	8.67	8.98	4.66
Mg(OH)2	3.06	-	9.58	9.74	9.69	9.44
MgCO3	4.48	9.58	-	5.68	4.33	10.8
CaO	8.67	9.74	5.68	-	8.48	10.8
Ca(OH)2	8.98	9.69	4.33	8.48	-	10.4
CaCO3	4.66	9.44	10.8	10.8	10.4	-

Unit [mg/L].

. In particular, Figure 2a and Figure 2b show the R_AS values for C_AS0 = 1 and 10 mg/L with ΣW_Ad/V = 0.4 g/L, respectively.

The R_AS value was the highest for MgO + Mg(OH)₂ among the combined additions; however, it was higher for the single addition of MgO with W_Ad/V = 0.4 g/L, as shown in Figure 1c,d. For the combinations containing Mg(OH)₂ and for MgCO₃ + CaCO₃, the R_AS values were lower at C_AS0 = 10 mg/L than at C_AS0 = 1 mg/L. For the other combinations, the R_AS was higher at C_AS0 = 10 mg/L than at C_AS0 = 1 mg/L.

A previous study on As(V) removal [22] reported that MgCO₃ + CaO and MgCO₃ + Ca(OH)₂ could remove almost 100% of As(V) at C_AS0 = 1 mg/L and >60% at C_AS0 =10 mg/L. However, for these two combinations, the R_AS values of As(III) attained <45% and <60% at C_AS0 = 1 and 10 mg/L, respectively, in this study. Although the single addition of 0.2 g/L of MgCO₃, CaO, or Ca(OH)₂ minimally removed As(III), as shown in Figure 1a,b, the combined addition of 0.2 g/L of MgCO₃ and 0.2 g/L of CaO or Ca(OH)₂ exhibited reasonable As(III)-removal performance.

3.2. Mg- and Ca-Residual Ratios

3.2.1. Mg- and Ca-Residual Ratios for Single Addition of One Type of Adsorbent

The values of C_Mg and C_Ca for the single-addition tests are listed in

Table 7. Values of C_Mg and C_Ca for the single-addition tests.

CAS0	1 mg/L		1 mg/L		10 mg/L		10 mg/L
WAd/V	0.2 g/L		0.4 g/L		0.2 g/L		0.4 g/L
	CMg	CCa	CMg	CCa	CMg	CCa	CMg	CCa
MgO	8.07	-	6.71	-	8.24	-	8.07	-
Mg(OH)2	4.84	-	4.91	-	6.07	-	5.94	-
MgCO3	24.3	-	26.9	-	26.9	-	27.9	-
CaO	-	125	-	195	-	112	-	189
Ca(OH)2	-	92.8	-	197	-	99.1	-	198
CaCO3	-	4.87	-	4.64	-	6.41	-	6.37

Unit [mg/L].

. The C_Mg and C_Ca values for C_AS0 = 1 mg/L and W_Ad/V = 0.2 g/L were obtained from Table 2.

Because the Mg and Ca components were leached out from the Mg- and Ca-based adsorbents, the amounts of Mg and Ca remaining as solids in the liquid were calculated as the Mg-residual ratio R_Mg [%] and Ca-residual ratio R_Ca [%], respectively, using the following equations:

R_Mg = 100 − (C_Mg × 1000)/[W_Ad/V × (α_Mg/100)] × 100

(2)

R_Ca = 100 − (C_Ca × 1000)/[W_Ad/V × (α_Ca/100)] × 100

(3)

where α_Mg and α_Ca are the Mg and Ca contents in the adsorbent, respectively. These values were described in a previous study [22].

Figure 3 shows the R_Mg and R_Ca values calculated using the data in Table 7 for single-addition As(III)-removal tests. Figure 3a and Figure 3b show the R_Mg and R_Ca values for C_AS0 = 1 and 10 mg/L with W_Ad/V = 0.2 g/L, respectively. Figure 3c and Figure 3d show the R_Mg and R_Ca values for C_AS0 = 1 and 10 mg/L with W_Ad/V = 0.4 g/L, respectively. No clear difference was observed in the R_Mg and R_Ca values owing to C_AS0; however, they were slightly higher when W_Ad/V was higher. The R_Mg values of MgO and Mg(OH)₂ were similar, but that of MgCO₃ was lower. The R_Ca value of CaCO₃ was >90% under all conditions in this study, which was significantly higher than those of the other Ca-based adsorbents. Notably, the R_Ca value of Ca(OH)₂ was the lowest.

3.2.2. Mg- and Ca-Residual Ratios for Combined Addition of Two Types of Adsorbents

The values of C_Mg and C_Ca for the combined-addition tests are presented in

Table 8. Values of C_Mg for the combined-addition tests.

CAS0	Combination	MgO	Mg(OH)2	MgCO3	CaO	Ca(OH)2	CaCO3
1 mg/L	MgO	-	5.73	23.7	0.10	0.13	8.87
	Mg(OH)2	5.73	-	23.1	0.04	0.04	4.68
	MgCO3	23.7	23.1	-	0.07	0.21	24.3
10 mg/L	MgO	-	6.80	23.8	0.29	0.29	9.12
	Mg(OH)2	6.80	-	23.4	0.04	0.03	6.34
	MgCO3	23.8	23.4	-	0.14	0.52	25.0

Unit [mg/L].

and

Table 9. Values of C_Ca for the combined-addition tests.

CAS0	Combination	MgO	Mg(OH)2	MgCO3	CaO	Ca(OH)2	CaCO3
1 mg/L	CaO	97.0	93.6	64.2	-	193	134
	Ca(OH)2	94.3	94.4	37.5	193	-	102
	CaCO3	2.97	2.55	1.12	134	102	-
10 mg/L	CaO	91.2	93.9	70.1	-	193	128
	Ca(OH)2	97.8	97.0	34.5	193	-	103
	CaCO3	3.67	2.13	1.44	128	103	-

Unit [mg/L].

, respectively. Figure 4 shows the R_Mg values calculated using data in Table 8 for the combined-addition tests. Figure 4a and Figure 4b show the R_Mg values for C_AS0 = 1 and 10 mg/L with ΣW_Ad/V = 0.4 g/L, respectively.

As shown in Figure 4, no clear difference was observed in the R_Mg values between the combined- and single-addition tests owing to C_AS0. For all the Mg-based adsorbents, R_Mg was close to 100% when they were combined with CaO or Ca(OH)₂. The values of R_Mg when Mg-based adsorbents were combined with CaCO₃ were similar to the values for the single-addition tests with W_Ad/V = 0.2 g/L, and MgCO₃ + CaCO₃ exhibited the lowest value (approximately 50%). Regarding only the Mg-based adsorbents, the R_Mg value was the highest for MgO + Mg(OH)₂ (approximately 97%) and lowest for Mg(OH)₂ + MgCO₃ (approximately 83%).

Figure 5 shows the R_Ca values calculated using the data in Table 9 for the combined-addition As(III)-removal tests. Figure 5a and Figure 5b show the R_Ca values for C_AS0 = 1 and 10 mg/L with ΣW_Ad/V = 0.4 g/L, respectively.

As shown in Figure 5, no clear difference was observed in the R_Ca values between the combined- and single-addition tests due to C_AS0. For all the Ca-based adsorbents, R_Ca tended to increase when combined with any Mg-based adsorbent. Particularly, for CaO and Ca(OH)₂, the R_Ca value was significantly increased when they were combined with MgCO₃ compared to their single additions (11–23% to 54–70%). Moreover, the R_Ca value of CaCO₃ when combined with any Mg-based adsorbent was 95% (92–94% for the single addition of 0.2 g/L). Regarding only the Ca-based adsorbents, the R_Ca value was the highest for Ca(OH)₂ + CaCO₃ (approximately 48%) and lowest for CaO + Ca(OH)₂ (approximately 25%).

3.3. pH of Treated Water

3.3.1. pH of Treated Water for Single Addition of One Type of Adsorbent

Figure 6 shows the pH_f values for the single-addition As(III)-removal tests. Figure 6a and Figure 6b show the pH_f values for C_AS0 = 1 and 10 mg/L with W_Ad/V = 0.2 g/L, respectively. Figure 6c and Figure 6d show the pH_f values for C_AS0 = 1 and 10 mg/L with W_Ad/V = 0.4 g/L, respectively. The measured pH₀ values for the single-addition tests at C_AS0 = 1 and 10 mg/L were 7.27–7.41 and 6.80–7.38, respectively. Regardless of the adsorbent added, the pH shifted towards alkalinity. The pH_f values under the test conditions were in the following order: CaCO₃ < Mg(OH)₂ < MgO, MgCO₃ < CaO, Ca(OH)₂. No clear effect of C_AS0 on pH_f was observed for any of the adsorbents. Additionally, for the Mg-based adsorbents and CaCO₃, no clear effect on pH_f was observed when W_Ad/V increased from 0.2 to 0.4 g/L. However, for CaO and Ca(OH)₂, the pH_f increased with increasing W_Ad/V.

3.3.2. pH of Treated Water for Combined Addition of Two Types of Adsorbents

Figure 7 shows the pH_f values for the combined-addition As(III)-removal tests. Figure 7a and Figure 7b show the pH_f values for C_AS0 = 1 and 10 mg/L with ΣW_Ad/V = 0.4 g/L, respectively. The measured values of pH₀ for the combined-addition tests at C_AS0 = 1 and 10 mg/L were 7.01–7.27 and 6.80–7.05, respectively. Similarly as with the single addition, the pH shifted to alkaline in the combined addition, and no clear difference in pH_f was observed owing to C_AS0. Among the combined additions, the pH_f value was the highest in CaO + Ca(OH)₂ and lowest in Mg(OH)₂ + MgCO₃. In the combined addition of Mg-based adsorbents with CaO or Ca(OH)₂, the pH_f was slightly lower than that in the single addition of CaO or Ca(OH)₂. With the combined addition of CaCO₃ with CaO or Ca(OH)₂, the pH_f value was similar to that with the single addition of CaO or Ca(OH)₂. With the combined addition of Mg-based adsorbents and CaCO₃, the pH_f value was lower than the value for each addition of the Mg-based adsorbent.

3.4. XRD Patterns for MgCO₃ + CaO, and MgCO₃ + Ca(OH)₂

The XRD patterns of the MgCO₃ + CaO and MgCO₃ + Ca(OH)₂ mixtures are shown in Figure 8a and Figure 8b, respectively. The XRD patterns of the ‘solids collected after stirring MgCO₃ + CaO in deionized water’ and ‘solids collected after stirring MgCO₃ + Ca(OH)₂ in deionized water’ are shown in Figure 8c and Figure 8d, respectively. The XRD patterns of the ‘solids collected after stirring MgCO₃ + CaO in an As(III) solution’ and ‘solids collected after stirring MgCO₃ + Ca(OH)₂ in an As(III) solution’ are shown in Figure 8e and Figure 8f, respectively. The XRD patterns of the ‘solids collected after stirring MgCO₃ + CaO in an As(V) solution’ and ‘solids collected after stirring MgCO₃ + Ca(OH)₂ in an As(V) solution’ are shown in Figure 8g and Figure 8h, respectively.

As shown in Figure 8a, weak peaks of Ca(OH)₂ (portlandite) were observed in addition to the peaks assigned to Mg₅(CO₃)₄(OH)₂·4H₂O (hydromagnesite) and CaO (lime), corresponding to the composition of the reagents. Although no peak was attributable to Ca(OH)₂ in the CaO reagent, XRD analysis of the CaO and MgCO₃ mixture was performed approximately one day after preparation. We assumed that some of the CaO reacted with the moisture contained in the MgCO₃ reagent to generate Ca(OH)₂. Figure 8b shows only the peaks of Mg₅(CO₃)₄(OH)₂·4H₂O (hydromagnesite) and Ca(OH)₂ (portlandite), and Figure 8c–h show the peaks of Mg(OH)₂ (brucite) and CaCO₃ (primarily calcite, but traces of vaterite are also present in Figure 8g). Therefore, in deionized water and the As(III) solution, MgCO₃ was transformed into Mg(OH)₂, and CaO and Ca(OH)₂ were transformed into CaCO₃. Moreover, no peaks indicated the generation of arsenite and/or arsenate precipitates (Figure 8e–h). These results suggested that As(III) and As(V) in the solutions were not solidified and removed from the liquid phase as precipitates of arsenite and/or arsenate species, but were instead incorporated into the generated precipitates of Mg(OH)₂ and/or CaCO₃.

Table 10. Test conditions and pH and composition of the filtrate for XRD analysis.

No.	As Valence	WMgCO3/V (g/L)	WCaO/V (g/L)	WCa(OH)2/V (g/L)	pH0	pHf	CAS0 (mg/L)	CAS (mg/L)	CMg (mg/L)	CCa (mg/L)	RAS (%)
(e)	As(III)	0.404	0.402	-	7.43	11.91	10.51	4.43	0.04	116	57.8
(f)	As(III)	0.403	-	0.404	7.43	11.57	10.51	2.87	0.17	49.1	72.7
(g)	As(V)	0.404	0.402	-	7.47	11.87	10.04	0.350	0.10	120	96.5
(h)	As(V)	0.403	-	0.400	7.47	11.55	10.04	0.208	0.39	52.6	97.9

shows the test conditions for preparing samples (e)–(h) for the XRD analysis and the pH and composition of the filtrate following solid–liquid separation and R_AS. W_MgCO3/V, W_CaO/V, and W_Ca(OH)2/V in Table 10 were the additive concentrations of MgCO₃, CaO, and Ca(OH)₂, respectively.

For both As(III) and As(V) (Figure 8), the R_AS value of the combined addition of MgCO₃ with Ca(OH)₂ was higher than that with CaO. In both the combined additions of MgCO₃ with CaO and Ca(OH)₂, the R_AS value was higher for As(V) than for As(III).

3.5. Results of Mg(NO₃)₂-Addition Tests with Alkaline As Solutions

The XRD patterns of the precipitates collected during the tests are shown in Figure 9. In particular, Figure 9a and Figure 9b show the precipitates in the As(III) and As(V) solutions, respectively.

Although differences were attributed to fairly broad peaks (Figure 9a) and sharper peaks (Figure 9b), peaks indicating the generation of Mg(OH)₂ (brucite) were observed in both Figure 9a,b. Therefore, Mg²⁺ was released by the dissolution of Mg(NO₃)₂ and combined with OH⁻ to generate Mg(OH)₂ in alkaline solutions.

The test conditions and results are listed in

Table 11. Values for Mg(NO₃)₂-addition tests with alkaline As solutions.

As Valence	WMg(NO3)2/V (g/L)	pH0	pHf	CAS0 (mg/L)	CAS (mg/L)	CMg (mg/L)	RAS (%)
As(III)	4.01	11.99	9.82	10.06	0.489	251	95.1
As(V)	4.01	11.98	9.93	10.04	0.080	225	99.2

. W_Mg(NO3)2/V in the table is the additive concentration of Mg(NO₃)₂·6H₂O in solution.

Although the R_AS value for As(III) was lower than that for As(V), both R_AS values exceeded 95%, indicating that As was effectively removed from the liquid phase. Combined with the XRD results, As(III) and As(V) were effectively incorporated into the generated precipitate of Mg(OH)₂.

3.6. Results of Ca(NO₃)₂- and Na₂CO₃-Addition Tests with Alkaline As Solutions

The XRD patterns of the precipitates collected during the tests are shown in Figure 10. In particular, Figure 10a and Figure 10b show the precipitates in the As(III) and As(V) solutions, respectively.

In both Figure 10a,b, strong peaks indicating the generation of CaCO₃ (calcite) were observed. These results confirmed that the Ca²⁺ released by the dissolution of Ca(NO₃)₂ combined with CO₃²⁻ released by the dissolution of Na₂CO₃ to generate CaCO₃ in the alkaline solutions.

The test conditions and results are listed in

Table 12. Values for Ca(NO₃)₂- and Na₂CO₃-addition tests with alkaline As solutions.

As Valence	WCa(NO3)2/V (g/L)	WNa2CO3/V (g/L)	pH0	pHf	CAS0 (mg/L)	CAS (mg/L)	CCa (mg/L)	RAS (%)
As(III)	2.70	1.21	11.99	11.94	10.06	10.01	251	0.3
As(V)	2.71	1.21	11.98	11.88	10.04	1.12	225	88.9

. W_Ca(NO3)2/V and W_Na2CO3/V in Table 12 are the additive concentrations of Ca(NO₃)₂·4H₂O and Na₂CO₃, respectively.

The R_AS value for As(V) was high (approximately 89%), while that for As(III) was only 0.3%, indicating that As(III) was minimally removed. Combined with the XRD results, the amount of As incorporated into the generated precipitate of CaCO₃ depended on the As valence, with As(V) being more readily incorporated than As(III).

4. Discussion

4.1. Effects of Combined Addition

4.1.1. Effects of Combined Addition on As-Removal Performance

When multiple adsorbents are combined, the R_AS value is typically unequal to the sum of the individual adsorbent R_AS values. This apparent combined-adsorbent synergistic effect was evaluated using a method similar to that used in a previous study [17]. The highest of the two single-addition R_AS values for each adsorbent in the combined-addition test was subtracted from the R_AS value in the combined-addition test. The obtained difference was denoted as ΔR_AS.

Each ΔR_AS value of the combined additions is shown in Figure 11. Figure 11a and Figure 11b correspond to C_AS0 = 1 and 10 mg/L, respectively, and the results from the single-addition tests with W_Ad/V = 0.4 g/L are included.

Figure 11 shows that the ΔR_AS value significantly varied with adsorbent combination and C_AS0. At C_AS0 = 1 mg/L, the ΔR_AS values for the MgO single addition and MgO + Mg(OH)_2, MgCO₃ + CaO, and MgCO₃ + Ca(OH)₂ combined additions were positive. However, the ΔR_AS value for the combined addition of MgO or Mg(OH)₂ with CaO or Ca(OH)₂ was negative. At C_AS0 = 10 mg/L, the ΔR_AS values for the CaO single addition and MgCO₃ + CaO, and MgCO₃ + Ca(OH)₂ combined additions were positive. Moreover, the ΔR_AS value for the combined addition of MgO with CaO was negative. Therefore, from the viewpoint of ΔR_AS, when ΣW_Ad/V increased from 0.2 to 0.4 g/L, the MgO single addition at C_AS0 = 1 mg/L and CaO single addition at C_AS0 = 10 mg/L exhibited the most improved As(III)-removal performances. However, both MgCO₃ + CaO and MgCO₃ + Ca(OH)₂ had extremely high ΔR_AS values at both C_AS0 = 1 and 10 g/L, indicating that combined addition had an extremely high synergistic effect on the removal performance of As(III). This result was similar to that for As(V) reported in a previous study [22].

4.1.2. Effects of Combined Addition on Mg- and Ca-Leaching Behaviors

Even if the Mg or Ca residual ratio was the same, the leached amount of Mg or Ca differed depending on the type of Mg- or Ca-based adsorbent used. Therefore, to evaluate the effects of combined addition on Mg- or Ca-leaching behavior, we compared the amount of change in C_Mg or C_Ca. Thus, the higher C_Mg or C_Ca value in the single-addition test (W_Ad/V = 0.2 g/L) of the two types of adsorbents used in the combined-addition test (ΣW_Ad/V = 0.4 g/L) was subtracted from the C_Mg or C_Ca value in the combined-addition test. The obtained difference was denoted as ΔC_Mg or ΔC_Ca. The lower the value of ΔC_Mg or ΔC_Ca, the higher the effects of inhibiting Mg- or Ca-leaching.

Each ΔC_Mg value of the combined additions is shown in Figure 12. Figure 12a and Figure 12b correspond to C_AS0 = 1 and 10 mg/L, respectively. Similarly to the ΔC_Mg value, each ΔC_Ca value of the combined additions is shown in Figure 13. Figure 13a and Figure 13b correspond to C_AS0 = 1 and 10 mg/L, respectively.

Unlike ΔR_AS, no clear influence of C_AS0 was observed for ΔC_Mg or ΔC_Ca. In both the combined addition among the Mg-based adsorbents and with CaCO₃, the ΔC_Mg value was within ±5 mg/L. In the combined addition of any one of the Mg-based adsorbents with CaO or Ca(OH)₂, the ΔC_Mg value was a large negative. Therefore, the combined addition of CaO or Ca(OH)₂ inhibited the leaching of Mg from Mg-based adsorbents. Although no clear difference was observed between CaO and Ca(OH)₂ in the Mg-leaching inhibition effects, the order was as follows: MgCO₃ > MgO > Mg(OH)₂.

In the combined addition among Ca-based adsorbents, except CaCO₃, the ΔC_Ca values were large positive values. The amount of Ca leaching from CaCO₃ was originally low, and the ΔC_Ca value was −5 to 16 mg/L even in the combined addition of CaCO₃ with other Ca-based adsorbents. Minimal effects on the Ca-leaching behavior from other Ca-based adsorbents were observed. Conversely, in the combined addition of CaO with any one of the Mg-based adsorbents, the ΔC_Ca value was significantly negative. Additionally, for Ca(OH)₂, the ΔC_Ca value was significantly negative only in the combined addition with MgCO₃. Thus, MgCO₃ exerted Ca-leaching inhibitory effects on both CaO and Ca(OH)₂.

4.2. As-Removal Mechanisms in the Combined Additions of MgCO₃ + CaO and MgCO₃ + Ca(OH)₂

4.2.1. Incorporation of As into the Generated Precipitates of Mg(OH)₂ + CaCO₃

Table 13. Estimated values related to the solid samples collected from the As solutions with added MgCO₃ + CaO and MgCO₃ + Ca(OH)₂.

No.	As Valence	Combination	RMg (%)	RCa (%)	WR-Mg(OH)2/V(g/L)	WR-CaCO3/V (g/L)	QAS (mg/g)
(e)	As(III)	MgCO3 + CaO	100	59.3	0.240	0.714	6.37
(f)	As(III)	MgCO3 + Ca(OH)2	99.8	77.3	0.239	0.539	9.82
(g)	As(V)	MgCO3 + CaO	99.9	58.0	0.240	0.714	10.2
(h)	As(V)	MgCO3 + Ca(OH)2	99.6	75.4	0.239	0.533	12.7

shows the various estimated values relating to the solid samples collected from the As solutions with added MgCO₃ + CaO and MgCO₃ + Ca(OH)₂, which were obtained using the data in Table 10.

W_R-Mg(OH)2/V or W_R-CaCO3/V in Table 13 is the concentration of solid Mg(OH)₂ or CaCO₃, respectively, in solution, assuming that all the Mg and Ca remaining in the solid state in the solution was in the form of Mg(OH)₂ or CaCO₃. In addition, when calculating W_R-Mg(OH)2 and W_R-CaCO3 from R_Mg and R_Ca, the molecular weights of Mg(OH)₂ and CaCO₃ were 58.3 and 100.1, respectively.

Furthermore, the Q_AS value in Table 13 was calculated using the following equation:

Q_AS = (C_AS0 − C_AS)/(W_R-Mg(OH)2/V + W_R-CaCO3/V).

(4)

As shown in Table 11, the Q_AS value was higher in MgCO₃ + Ca(OH)₂ than in MgCO₃ + CaO and higher in As(V) than in As(III). Therefore, the condition for obtaining the highest As-removal efficiency was applying the combined addition of MgCO₃ + Ca(OH)₂ to As(V).

4.2.2. Incorporation of As into the Generated Precipitate of Mg(OH)₂

Table 14. Values of R_Mg, W_R-Mg(OH)2/V, and Q_AS for Mg(NO₃)₂-addition tests with alkaline As solutions.

As Valence	RMg (%)	WR-Mg(OH)2/V (g/L)	QAS (mg/g)
As(III)	34.0	0.310	30.9
As(V)	40.7	0.371	26.8

shows several estimated values relating to the precipitates collected from the alkaline As solutions with added Mg(NO₃)₂·6H₂O, which were obtained using the data in Table 11.

The R_Mg and W_R-Mg(OH)2/V values were calculated assuming that the molecular weight of Mg(NO₃)_2·6H₂O was 256.4 and the reagent purity was 100%. However, as described in Section 2.5.1, the actual reagent purity was <100% because Mg(NO₃)₂·6H₂O is a deliquescent substance.

The As-removal mechanism in the tests with Mg(NO₃)₂ was coprecipitation accompanied by the generation of Mg(OH)₂. The Q_AS values for both As(III) and As(V) were high, as shown in Table 14. The As-removal efficiency was essentially similar because the difference in Q_AS between As(III) and As(V) was not significantly large. Therefore, As incorporation occurred with the generation of Mg(OH)₂ and was independent of the As valence.

4.2.3. Incorporation of As into the Generated Precipitate of CaCO₃

Table 15. Values of R_Ca, W_R-CaCO3/V, and Q_AS for Ca(NO₃)₂ and Na₂CO₃-addition tests with alkaline As solutions.

As Valence	RCa (%)	WR-CaCO3/V (g/L)	QAS (mg/g)
As(III)	99.7	1.14	0.04
As(V)	99.6	1.14	7.81

shows several estimated values relating to the precipitates collected from the alkaline As solutions with added Ca(NO₃)₂·4H₂O, which were obtained using the data in Table 12.

The R_Ca and W_R-CaCO3 values were calculated assuming the molecular weight of Ca(NO₃)₂·4H₂O was 236.2 and the reagent purity was 100%. However, as described in Section 2.5.2, the actual reagent purity was <100% because Ca(NO₃)₂·4H₂O is a deliquescent substance.

The As-removal mechanism in the combined-addition tests with Ca(NO₃)₂ and Na₂CO₃ was coprecipitation accompanied by the generation of CaCO₃. The Q_AS value for As(III) was 0.04 mg/g, indicating that almost no As(III) was removed. The Q_AS of As(V) was approximately 7.8 mg/g, demonstrating that As(V) was removed by the generation of CaCO₃. Therefore, the incorporation of As by the generated CaCO₃ depended on the valence state of As.

Scanning electron microscope (SEM) image observation and elemental mapping by an EDS system for As adsorption on Mg- and Ca-based adsorbents have been carried out by other researchers [23,24]. Lin et al. (2023) studied As(V) removal using a monodispersed porous pinecone-like Mg(OH)₂, and reported that As was abundantly and uniformly distributed in the adsorption products by an elemental As mapping diagram [23]. Meanwhile, Lee et al. (2024) studied As(V) removal using a wood bottom ash containing CaCO₃ and Ca(OH)₂ as main components [24]. Their elemental As mapping diagram also showed that As is uniformly adsorbed on the surface of the Ca-based adsorbent. From these results, it is considered that As is essentially uniformly adsorbed on the surface of the Mg- and Ca-based adsorbents used in this study, while, in the combined additions of MgCO₃-CaO or MgCO₃-Ca(OH)₂, As may be concentrated in the generated Mg(OH)₂ and CaCO₃. Therefore, future tasks for this research include additional research such as SEM observation and elemental mapping with EDS.

4.3. Comprehensive Evaluation

The maximum As(III)-adsorption capacities of the three new adsorbents introduced in the Introduction section were 64 mg/g on ZrO₂-AC, 6.8 mg/g on CW@MPTMS@poly(AA-co-VyAc)@PABA, and 55 mg/g on B-Mn-Al [11,12,13]. The adsorption mechanism of As(III) was mainly physical adsorption for CW@MPTMS@poly(AA-co-VyAc)@PABA, and essentially chemical adsorption for ZrO₂-AC and B-Mn-Al [11,12,13]. A comparison of their maximum adsorption capacities would suggest that chemisorption is more advantageous for As(III) removal than physical adsorption. In this study, the maximum adsorption capacity for each combination of the Mg- and Ca-based adsorbents is unknown because there are not enough data to calculate the maximum adsorption amount from the Langmuir model. However, the Mg- and Ca-based adsorbents are also considered to have high adsorption capacities because their adsorption mechanism for As(III) is also chemical adsorption. Additionally, the Mg(NO₃)₂-addition tests showed that the combined addition of MgCO₃-(OH)₂ or MgCO₃-CaO could generate Mg(OH)₂ with high As(III) adsorption capacity (<30 mg/g). From the above, the combined addition of MgCO₃-Ca(OH)₂ or MgCO₃-CaO is expected to be a valuable As(III)-removal method.

From the perspective of R_AS, the most effective strategy for As(III) removal was the single addition of MgO or the combined addition of MgO with Mg(OH)₂. Notably, the addition of MgCO₃ or Ca-based adsorbents alone did not remove As(III); however, the combined addition of MgCO₃ with CaO or Ca(OH)₂ improved As(III)-removal performance. The single addition of 0.2 g/L CaO minimally removed As(III); however, that of 0.4 g/L CaO increased the R_AS value. Thus, MgCO₃, CaO, or Ca(OH)₂ may be selected as less expensive adsorbents when the cost of MgO is higher than that of other adsorbents. Moreover, the combined addition of MgCO₃ with CaO or Ca(OH)₂ significantly inhibited the leaching of Mg and Ca, which are components of the base material.

The combined addition of MgCO₃ with either CaO or Ca(OH)₂ did not remove As(III) to the same extent as As(V). As described in Section 4.2.2 and Section 4.2.3, the Mg(OH)₂ generated in these combined additions incorporated both As(V) and As(III). Conversely, CaCO₃ incorporated As(V) but minimally incorporated As(III). Therefore, introducing a step for oxidizing As(III) to As(V) before applying the combined additions might be an effective strategy. A previous study [25] reported that adsorbed As(III) was more likely to re-leach than As(V). Therefore, As oxidation treatment is recommended to prevent secondary contamination due to the re-leaching of As from spent adsorbents.

When treating As in industrial wastewater, which has a more diverse composition of dissolved components than groundwater, it is extremely useful to have numerous options to propose appropriate adsorbent types and application methods that consider the characteristics of the wastewater.

5. Conclusions

In a previous study [22], the combined addition of Mg- and Ca-based adsorbents was investigated to improve As(V)-removal performance and inhibit the leaching of the base material components from the adsorbents (i.e., improving the environmental stability of the adsorbents). In this study, systematic tests for As(III) removal via the combined addition of Mg- and Ca-based adsorbents were conducted based on the above As(V) study. Additionally, X-ray diffraction (XRD) analysis was carried out to verify the generation of Mg(OH)₂ and CaCO₃ in the combined addition test of MgCO₃-CaO or MgCO₃-Ca(OH)₂ that was inferred in a previous study [22]. The solid samples for XRD analysis were prepared in the absence of As, in the presence of As(V) or As(III). Furthermore, to verify the incorporation of As into the produced Mg(OH)₂ or CaCO₃, Mg(NO₃)-addition tests and Ca(NO₃)₂- and Na₂CO₃-addition tests were performed using alkaline As(III) and As(V) solutions.

As a result of the single- and combined-addition tests for As(III) removal, the R_AS value was the highest for the single addition of MgO and the combined addition of MgO with Mg(OH)₂. From the viewpoint of improving the R_AS value and inhibiting the leaching of the base material components, the most significant synergistic effect was observed for MgCO₃ + CaO and MgCO₃ + Ca(OH)₂. Notably, these two combined-addition methods achieved the highest R_AS values in a previous study for As(V) removal [22]. Following As adsorption, the XRD analysis of the solid samples obtained from these two combined-addition tests confirmed that Mg(OH)₂ and CaCO₃ were produced. Subsequently, the results of the As-removal tests in which Mg(NO₃)₂·6H₂O was added to an alkaline solution containing As(III) or As(V) demonstrated that As was removed from the liquid phase via incorporation by the generated Mg(OH)₂, regardless of the As valence state. Moreover, the addition of both Ca(NO₃)₂·4H₂O and Na₂CO₃ demonstrated that As(V) was incorporated during the generation of CaCO₃; however, As(III) was not incorporated and remained in the solution.

These investigations elucidated the As-removal mechanism expressed by the combined addition of MgCO₃ + CaO and MgCO₃ + Ca(OH)₂. As(V) was removed via incorporation into both Mg(OH)₂ and CaCO₃, whereas As(III) was removed from the liquid phase via incorporation into only Mg(OH)₂. Moreover, MgCO₃ + Ca(OH)₂ was evaluated as a more efficient combined-addition method because MgCO₃ + Ca(OH)₂ exhibited a higher Q_AS value than MgO + CaO. The R_AS value of As(III) was inferior to that of As(V) owing to the differences in the As-removal mechanisms. Therefore, it is recommended to convert As(III) to As(V) via oxidation before applying the combined-addition method.

Processing techniques such as doping and synthesis of various components on activated carbons and concrete wastes are complex and relatively expensive [11,12]. Fe-based adsorbents are also relatively expensive [18]. Additionally, adsorbents doped with heavy metal or transition elements such as Mn or cobalt have problems with their leaching [13,26], which may lead to health risks. Meanwhile, the only components that may leach from Mg- and Ca-based adsorbents are Mg and Ca components, which are harmless to humans and animals. Both Mg-based and Ca-based adsorbents are common and easily available, and they have the major advantages of having high As adsorption efficiency and almost no risk of components leaching from the base material causing adverse effects on human health or the ecosystem. As one measure against As contamination, widespread use of adsorbents based on Mg and Ca components, which are abundant on Earth and do not pose environmental pollution problems, could make a great contribution to a sustainable society.

This study reconfirmed the effectiveness of the combined addition of MgCO₃-Ca(OH)₂ or MgCO₃-CaO for As removal. To increase the As-removal efficiency by this combined addition, it is considered necessary to investigate the optimal combined addition ratio. Additionally, it may be possible to promote the As-incorporation reaction by adjusting the pH appropriately. In the future, it is expected that this research will be further developed by focusing on these issues.