Alternative Method for the Treatment of Hydrometallurgical Arsenic-Calcium Residues: The Immobilization of Arsenic as Scorodite.

Arsenic-calcium residue (ACR) is one of the major hazardous solid wastes produced by the metallurgical industry that poses a serious threat to the environment. However, a suitable method for the effective treatment of ACR is still lacking. In this study, an alternative treatment method for ACRs via the immobilization of As as scorodite was proposed with the use of two types of ACRs (ACRreal directly collected from a Pb refinery and ACRlab precipitated from waste sulfuric acid in the lab). The treatment of ACR included preparing the As-enriched solution via H2SO4 dissolution-neutralization of ACR at pH < 2, As(III) was oxidized by H2O2, and As(V) was immobilized as scorodite. The results showed that gypsum produced from ACRlab in the dissolution-neutralization process contained 68 mg/kg of As, far below the Chinese national standard for hazardous solid wastes (<0.1 wt %, GB5085.62007). The gypsum produced from ACRreal contained 5400 mg/kg of As due to the presence of original high-As gypsum (1.6 wt %) in ACRreal. These results showed that the preliminary removal of SO4 2- from waste sulfuric acid by lime neutralization-precipitation at pH ∼ 2 could produce pure-phase gypsum by avoiding the HAsO4 2- isomorphic substitution for SO4 2-. The scorodite produced from both ACRs displayed good As stability at pH 4.95 (0.9 and 0.5 mg/L) via the toxicity characteristic leaching procedure (TCLP) method and at pH 3-7 (0.4-3.0 mg/L) via a 15 day short-term stability test.

Introduction

Arsenic (As) is one of the most toxic elements widely present in contaminated water, soil, and sediments. It occurs most commonly in association with other minerals, such as sulfides and arsenides.[1] Arsenic can be liberated into waste sulfuric acid during the metallurgical processing of base and precious metals (e.g., Pb, Zn, Cu, and Au) by the oxidation and acid dissolution of As-containing minerals.[2] Tens of thousands of tons of As are liberated every year in the nonferrous metal industry around the world.[3] During the past 20 years, the commercial use of As in herbicides, insecticides, and wood preservatives has declined or been banned due to its potential threat to living organisms and the natural environment.[4] Because of its toxicity and limited marketability, most of the As in waste sulfuric acid must be removed and immobilized as a stable solid to prevent contamination to surrounding soils and waters.

The lime neutralization–precipitation method for the removal of As from waste sulfuric acid has been widely used due to its low operating costs and simple technological processes.[5−8] In this method, As in waste sulfuric acid can be removed efficiently via the formation of insoluble calcium arsenate/arsenite precipitates and the aqueous As concentration can be less than several mg/L at high pH values (>12).[9−11] However, this method generates a large volume of arsenic–calcium residue (ACR), which is classified as a hazardous solid waste due to its high As content and poor solution stability. In particular, ACR may release a very large amount of As to the surrounding environment under acidic to circumneutral conditions.[12−15] For example, Martínez-Villegas et al.[15] reported that the dissolution of ACR in an inactive smelter in Santa María de la Paz, Mexico, caused high levels of As pollution in an adjacent down gradient 6 km perched aquifer, and the dissolved As concentration reached 158 mg/L. Furthermore, ACR may react with carbon dioxide/carbonates (CO2/CO32–) and then form calcite (CaCO3), thus leading to the strong release of As into solution when the ACR is exposed to air.[1,16] To reduce the environmental risks associated with ACR, various types of technologies, such as cement stabilization/solidification (S/S),[17−21] polymeric encapsulation,[22,23] and hydrothermal precipitation transformation,[14] have been developed. The cement S/S and polymeric encapsulation methods are based on the idea that an inert material and a binder are used to prevent physical contact between the ACR and the surrounding environment. However, these methods would largely increase the volume of solid waste because of the high demand for inert materials and binders. Furthermore, the As stability of this kind of cement S/S-treated product under mildly acidic conditions is still an environmental concern.[19,21] Viñals et al.[14] proposed that ACR transforms into crystalline arsenical natroalunite (∼(Na, Ca)(Al, Fe)3((S, As, P)O4)2(OH)6) when AsO43– is structurally substituted for SO42– under strongly acidic conditions at 180–200 °C. However, this process may not be suitable for the treatment of ACR owing to its low As precipitation yield (<65%) and the low fixed As content in arsenical natroalunite (<4 wt %). On the other hand, the high-temperature (>180 °C) hydrothermal process leads to high operating costs and high consumption of energy.

Scorodite (crystalline FeAsO4·2H2O) has been advocated as an attractive As(V) carrier for As immobilization because of its high As content (As ∼ 33 wt %) and high stability. The formation of scorodite has been widely investigated over the past 20 years, including methods involving hydrothermal precipitation from acidic solutions using high-temperature and high-pressure autoclaves[24] and atmospheric pressure precipitation at elevated temperatures.[25−34] Among these technologies, the atmospheric process is more promising because of its low operating costs.

Herein, we propose an alternative method for the treatment of the hazardous ACR via leaching As using sulfuric acid followed by immobilization of As as scorodite. This method, to our best knowledge, has not been reported previously. The objectives of the present study are (1) to transform ACR into a stable As-containing mineral, (2) to investigate the reaction mechanism during the dissolution–neutralization of ACR, and (3) to present an alternative treatment method for ACR.

Results and Discussion

Analysis of Initial ACRreal and ACRlab

The initial chemical compositions of ACRreal and ACRlab are summarized in Table . The results show that the ACRreal contains 3.8 wt % As(T), 1.7 wt % As(III), 28 wt % Ca, 7.4 wt % S, 7.1 g/kg Cu, 1.1 g/kg Zn, 1.1 g/kg Pb, and 0.6 g/kg Cd. The morphology of ACRreal solids mainly appeared as loose cottonlike particles with a considerable amount of rodlike particles (Figures b and S2). The elemental composition analysis using energy-dispersive X-ray spectrometry (EDX) indicated that the ACRreal contains trace amounts of Cu, Zn, Pb, and Cd (Figure b). The main X-ray diffraction (XRD) peaks of the ACRreal were located at the same positions as those of the standard XRD patterns of gypsum (CaSO4·2H2O PDF #6-46) and calcite (CaCO3 PDF #05-0586), thus indicating that the dominant crystalline phases in ACRreal were gypsum and calcite, which were also responsible for the high Ca and S contents. This result also suggests that the rodlike particles in ACRreal were gypsum crystals, in line with the chemical composition and EDX results. The presence of CaCO3 was ascribed to the excessive neutralizing reagent of CaCO3 during the ACRreal precipitation process. No identifiable diffraction lines of the calcium arsenate/arsenite phase were observed attributable to the major components of gypsum and calcite in the ACRreal. However, the raised baseline of the ACRreal XRD patterns between 25 and 35° 2θ may be ascribed to the loose cottonlike phase, suggesting that ACRreal contained amorphous Ca–As phases. The analysis results show that the ACRlab contains 17 wt % As(T), 16.4 wt % As(III), 13.8 wt % Ca, 0.4 wt % S, 5 g/kg Cu, 1.3 g/kg Zn, 3.7 g/kg Pb, and 56 g/kg Cd. The EDX images also indicated that the ACRlab contains considerable amounts of As, Ca, and Cd with trace amounts of S, Cu, Zn, and Pb (Figure d). The scanning electron microscopy (SEM) image indicated that the ACRlab showed a loose cottonlike morphology. The main diffraction peaks of ACRlab (Figure c) were located at the same positions as those of the standard XRD patterns of calcium arsenite (PDF #1-828). These results suggested that the ACRlab mainly exists in a poorly crystalline/amorphous form and that As predominantly exists as As(III).

Table 1

Contents of As, Ca, S, and Trace Metals in the Dried ACRreal and ACRlab

element	As(T) (wt %)	As(III) (wt %)	Ca (wt %)	S (wt %)	Cu (g/kg)	Zn (g/kg)	Pb (g/kg)	Cd (g/kg)
ACRreal	3.8	1.7	28.2	7.4	7.1	1.1	1.1	0.6
ACRlab	17	16.4	13.8	0.4	5.0	1.3	3.7	56

Figure 1

XRD patterns and SEM-EDX images of ACRreal (a, b) and ACRlab (c, d). The vertical bars represent the standard XRD patterns of calcium arsenite (PDF #1-828), calcium arsenate (PDF #39-10), calcite (PDF #05-0586), and gypsum (PDF #6-46).

The original gypsum in ACRreal was separated by HCl dissolution–neutralization at pH ∼ 2 to remove the acid-soluble phases (i.e., calcite, calcium arsenate/arsenite, etc.). The results showed that the content of the original gypsum in ACRreal reached up to 45 wt % (Table S1). The SEM image of the original gypsum in ACRreal showed that the particles were stubby rodlike in the size of tens of micrometers (Figure a). The cross sections of the single crystal particle of the original gypsum in ACRreal and the EDX elemental composition analysis are shown in Figure b. The results showed that the original gypsum dominantly consists of Ca, S, and O and trace amounts of As, thus indicating that acid-soluble phases in ACRreal (i.e., calcite, amorphous Ca–As phase, metal-salt, etc.) were removed completely after five times acid treatment. The chemical composition analysis showed that the As content in the original gypsum was up to 1.6 wt % (Table ). This result suggests that the one-step lime/limestone neutralization–precipitation of waste sulfuric acid to alkaline pH could incorporate a considerable amount of As into the structure of gypsum. The mechanism of incorporation of As(V) into gypsum was the isomorphic substitution of HAsO42– for SO42– due to HAsO42– being the dominating As(V) species in the pH range 7–10.[3,35]

Figure 2

SEM image of the original gypsum in ACRreal (a) and cross-sectional and EDX images of a single crystal particle of the original gypsum (b).

Table 2

Contents of As, Ca, S, and Trace Metals in the Original Gypsum in ACRreal

As (wt %)	Ca (wt %)	S (wt %)	Cu (g/kg)	Pb (g/kg)	Zn (g/kg)	Cd (g/kg)
1.6	23.7	17.3	1.1	0.9	0.4	0.1

Dissolution–Neutralization of ACRreal and ACRlab

Solid-Phase Analysis

Gypsum precipitated during the dissolution–neutralization of ACRreal and ACRlab in the H2SO4, as described by eqs and 2. The XRD patterns of the gypsum precipitated matched well with the standard XRD patterns of gypsum (PDF #6-46), indicating that gypsum was the dominant crystalline phase (Figure a,c). The SEM images showed that the morphology of the gypsum appeared as rodlike particles in the size of tens of micrometers (Figure b,d). The elemental composition analysis of EDX suggests that the gypsum from ACRlab (gypsum-ACRlab) is pure gypsum, whereas the gypsum-ACRreal contained trace amounts of metal cations.

Figure 3

XRD patterns and SEM-EDX images of the gypsum formed via the H2SO4 dissolution–neutralization process from ACRreal (a, b) and ACRlab (c, d).

Chemical composition analysis (Table ) showed that the As content in gypsum-ACRreal (5400 mg/kg) was considerably higher than that in gypsum-ACRlab (68 mg/kg). This could be ascribed to the high As content in the original gypsum in ACRreal (Table ). This high-As gypsum got mixed with the newly formed gypsum during the H2SO4 dissolution–neutralization process, thus leading to the high As content in the gypsum produced from ACRreal. The toxicity characteristic leaching procedure (TCLP) results showed As leachability was 0.1 and 99 mg/L for gypsum-ACRlab and gypsum-ACRreal, respectively. The high-As gypsum-ACRreal is still classified as hazardous solid waste and therefore could be of environmental concern if directly disposed into the environment. In the present study, the two-step lime neutralization of waste sulfuric acid could be used to avoid this kind of high-As gypsum. The results indicated that the As content in the gypsum precipitated in the first step was only 46 mg/kg (Table S2) and hence is far below the Chinese national standard for hazardous solid wastes (As <0.1 wt % = 1 g/kg, GB5085.62007).

Table 3

Contents of As, Ca, S, and Trace Metals in Gypsum Obtained from ACRreal and ACRlab via the Dissolution–Neutralization Processa

element	As (mg/kg)	Ca (wt %)	S (wt %)	Cu (g/kg)	Pb (g/kg)	Zn (g/kg)	Cd (mg/kg)
gypsum-ACRreal	5400	23.6	17.8	0.2	0.1	0.1	18
gypsum-ACRlab	68	23.5	17.3	UD	0.2	UD	89

UD represents undetectable.

According to a previous study by Fujita et al.,[36] the fine-particle morphology of the gypsum obtained via the addition of H2SO4 solution into calcium arsenate sludge suggests bad crystallinity. However, in our study, the good crystallinity and bigger particle size of the gypsum crystals may be attributed to the lower impurity concentrations (As, Cu2+, Zn2+, Pb2+, and Cd2+) that are favorable for gypsum growth during the slow continuous addition of ACR sludge into the H2SO4 solution.[3] The major reaction mechanism for the formation of gypsum crystals under our experimental conditions can be summarized as follows4The Ca–As phase was decomposed into Ca2+, H3AsIIIO3, and HAsVO4(3– (eqs and 2) after the ACR sludge was added dropwise into the H2SO4 solution (pH < 2). Then, the gypsum was formed via the reaction of dissolved Ca2+ with aqueous SO42– (eq ), which then disrupted the dissolution equilibrium of eq . The degree of supersaturation declined during CaSO4·2H2O formation because of the slow release of SO42– from H2SO4 (eq ). The incorporation of As and trace metal impurities into the gypsum structure can alter the growth of the crystal in certain orientations, while the impurities uptake is dependent on the level of supersaturation. Therefore, the supersaturation was controlled by the simultaneous addition of ACR sludge and H2SO4 solutions, with pH control playing a critical role in producing well-grown and phase-pure gypsum crystals.[37−39] Based on the Ostwald ripening theory of crystal nucleation and growth,[40] the early precipitated gypsum could also play an important role as a seed for the growth of gypsum, thus inducing the formation of larger gypsum particles.

During nonferrous metallurgical processes, good crystallinity offers efficient settling properties, solid–liquid separation, and facilitates washing. As discussed above, the H2SO4 dissolution–neutralization of ACR in the present study via the continuous slow addition of ACR sludge into the H2SO4 solution is feasible for industrial applications. Notably, the trace metal content found included Cu 0.2 g/kg, Zn 0.1 g/kg, Pb 0.6 g/kg, and Cd 18 mg/kg in gypsum-ACRreal as well as Pb 0.2 g/kg and Cd 89 mg/kg in gypsum-ACRlab after washing five times with an acidic gypsum-saturated solution (Table ). This result suggested that trace metal cations (i.e., Cu2+, Zn2+, Cd2+) could be incorporated into the gypsum during the precipitation of Ca2+ with SO42– in the presence of trace metal cations.[41]

In a previous study, Fujita et al.[36] proposed a method of preparation of an As(V) solution from As- and Cu-bearing byproducts for scorodite synthesis. In their work, the aqueous As(V) was extracted by the precipitation of Ca5(AsO4)3(OH) via the addition of hydrated lime (Ca(OH)2) under alkaline conditions. Then, an As(V)-enriched solution was prepared via the addition of H2SO4 to the Ca–As(V) sludge, which liberated As(V) into solution while simultaneously capturing Ca2+ as gypsum. However, the As content in the precipitated gypsum reached 433 mg/kg after five times washing with 1 mol/L H2SO4 solution and thus could be of environmental concern if this kind of As gypsum is disposed of in the environment. In our work, we proposed adding the ACR slurry to the H2SO4 solution, and the As content in the precipitated gypsum was as low as 68 mg/kg. The lower As content in gypsum-ACRlab may be ascribed to the slow release of As to the aqueous phase during the dropwise addition of ACR sludge into the H2SO4 solution. The released AsO33– and/or AsO43– ions will rapidly complex with H+ ions and then exist in the form of H3AsIIIO30, H2AsVO4–, and H3AsVO4, thus having less similarity to SO42– during the dissolution–neutralization process in acidic solutions (Figure S3) and hence preventing As from taking place of an isomorphic substitution with SO42– at an acidic pH (∼2).[3,42]

Liquid-Phase Analysis

The concentrations of As, Ca, and trace metal cations in the filtrate after the H2SO4 dissolution–neutralization of ACRreal and ACRlab are summarized in Table . The results indicated that the total As (As(T)) in the filtrate reached 10.1 and 15.8 g/L and As(III) reached 6.5 and 12.8 g/L in the ACRreal and ACRlab systems, respectively. The redistribution of As among the solid and liquid phases indicated that almost all of the As in ACRreal and ACRlab was leached out after the dissolution–neutralization process. This was in agreement with the chemical composition analysis of the precipitated gypsum. The results also suggested that trace metal cations (Zn2+, Pb2+, Cu2+, and Cd2+) were also leached out after the dissolution–neutralization process.

Table 4

Concentrations of As, Ca, and Trace Metals in the As-Containing Solution after H2SO4 Dissolution–Neutralization of ACRreal and ACRlaba

element	As(T)	As(III)	Ca	Cu	Zn	Pb	Cd
ACRreal	10.1	6.5	3.2	0.6	0.2	0.3	0.3
ACRlab	15.8	12.8	3.5	20 × 10–2	0.1	0.1	0.5

The unit of concentration is g/L.

Oxidation of As(III) to As(V)

Hydrogen peroxide (H2O2) is the environmentally preferred reagent for the oxidation of As(III) to As(V) due to its high oxidation potential and the decomposition product, which is water that is harmless.[28,43] The concentrations of As(III) and As(T) and the changes in the pH and Eh of solution during the As(III) oxidation process in the ACRreal and ACRlab systems are shown in Figure . The results indicated that As(III) was oxidized to As(V) with an oxidation efficiency of up to 98.6%. For instance, during the As(III) oxidation process in the ACRreal system, the concentration of As(III) decreased from 6.5 to 0.1 g/L, while the concentration of As(V) increased from 3.6 to 10.1 g/L. During the oxidation process in the ACRlab system, the concentration of As(III) decreased from 12.8 to 0.3 g/L, while the concentration of As(V) increased from 3.1 to 15.8 g/L. The pH value decreased from 1.8 to 1.3 and 1.5 to 1.2 in the ACRreal and ACRlab systems, respectively.

Figure 4

Concentrations of aqueous As(T) and As(III), pH, and Eh during the As(III) oxidation process in the ACRreal (a, b) and ACRlab (c, d) systems. (0.5 and 1 mL of 20% H2O2/As(III) at a molar ratio of ∼1/10 were used in the ACRreal and ACRlab systems, respectively).

The major reactions during As(III) oxidation in our study can be summarized as in eq . After As(III) was oxidized to As(V), more H+ would be released to the solution (eqs and 7, Figure S3). In particular, in our experimental conditions (pH ≤ 2), As(III) existed only as H3AsO3, whereas H3AsO4 and H2AsO4– could be formed after As(III) was oxidized to As(V); thus, the pH of the solution decreased because of the released H+ (Figure ). This reaction mechanism indicated that the pH will reach a constant value after As(III) was completely oxidized to As(V) by H2O2 (the As(III)/H2O2 molar ratio was approximately 1). Thus, the pH of the As-enriched solution could also be used as an indicator of the endpoint of the As(III) oxidation process.Furthermore, the oxidation–reduction potential (ORP) could be used as an indicator of the residual concentration of As(III) by measuring the highest and the lowest ORPs after H2O2 is added.[41] As can be seen, the Eh value increased rapidly after the addition of H2O2 and then decreased after the consumption of H2O2 by As(III). After most of the As(III) was oxidized to As(V), the Eh value increased sharply when excess H2O2 was added. For instance, the Eh increased significantly after adding 9 mL of 20% H2O2 to the ACRlab system when As(III) was oxidized completely (Figure ). This suggested that the pH and ORP could be dual indicators of the endpoint of the As(III) oxidation process.

As(V) Immobilization via Scorodite and Leachability of Produced Scorodite

Liquid- and Solid-Phase Analyses

Table summarizes the chemical composition of the filtrate obtained before and after the As(V) precipitation of scorodite. The results suggested that almost all of the trace metal cations remained in the postreaction solution. A previous study indicated that stable and well-crystallized scorodite was synthesized in the presence of 43 g/L Zn and 37 g/L Cu,[42] thus suggesting that the liberated Zn and Cu ions from the As(V)-enriched solution likely did not play a significant role in scorodite formation. In our work, an As removal efficiency of 96.5% was achieved via the formation of scorodite in the presence of trace metal impurities (Pd, Zn, Cd, and Cu). The removal mechanism of As(V) could be summarized as follows:The aqueous As(V) speciation as a function of pH calculated using Visual MINTEQ is shown in Figure S3. The calculation results indicated that the As(V) solution contained approximately 10% H2AsO4– and 90% H3AsO4 at pH ∼1.3 (our experimental conditions). After the formation of scorodite, the protonated H+ from H2AsO4– and H3AsO4 was released into the solution (eqs and 9), which was consistent with the decreased pH value. The contents of Pb2+ and Zn2+ in the aqueous phase decreased significantly after scorodite precipitation. We estimated the solubility of the Pb- and Zn-arsenate solids using Visual MINTEQ, and the results are shown in Figure S4. The results suggested that Pb- and Zn-arsenate solid phases could not be formed during the scorodite precipitation process under our experimental conditions, thus indicating that Pb2+ and Zn2+ may be captured by the precipitated scorodite via incorporation, as suggested by Fujita et al.[44] The SEM images suggested that the morphology of the produced scorodite mainly appeared as agglomerated spherical particles with good crystallinity (Figure ). The chemical composition and EDX analysis results showed that both ScoroditeACRreal and ScoroditeACRreal dominantly consist of Fe, As, and trace amounts of S, Ca, Pb, Zn, and Cd (Table and Figure b,d). In comparison, the As-enriched solution was also synthesized by HCl dissolution–neutralization of ACRlab at pH < 2. The results showed that ACRlab could be completely dissolved in the HCl solution, and the dissolved Ca2+ and As(V) finally formed a mixture of scorodite and gypsum after the addition of Fe(SO4)1.5 (Figure S5). Besides, the HCl dissolution–neutralization of ACR showed its economic infeasibility because of the high cost of HCl.

Table 5

Analytical Data for the Filtrate before and after Scorodite Synthesisa

element		As(V)	Ca	Cu	Zn	Pb	Cd
ACRreal	before	10.1	3.2	0.6	0.2	0.3	0.3
	after	0.7	0.9	0.5	0.2	4 × 10–2	3 × 10–2
ACRlab	before	15.8	3.5	2 × 10–2	0.1	0.1	0.5
	after	0.6	1.2	2 × 10–2	UD	UD	0.3

The unit of concentration is g/L.

Figure 5

XRD patterns and SEM-EDX images of the produced scorodite from ACRreal (a, b) and ACRlab (c, d). The vertical bars represent the standard XRD patterns of scorodite (PDF #37-468) and gypsum (PDF #6-46).

Table 6

Contents of Fe, As, Ca, S, and Trace Metals in the Precipitated Scoroditea

element	Fe	As	Ca	S	Cu	Zn	Pb	Cd
scoroditeACRreal	22	28	0.3	1.1	0.3	0.2	1.0	0.1
scoroditeACRlab	21	29	0.5	0.6	0.1	0.3	0.8	0.5

The unit of concentration is wt %.

In the previous study, Fujita et al.[32,33] proposed a method of scorodite synthesis at atmospheric pressure and temperatures below 100 °C conducted in a Fe(II)–As(V)–H2O system. In this method, oxygen (O2) was injected to oxidize Fe(II) to Fe(III), which then reacted with As(V) to precipitate as scorodite. Scorodite was formed near the precipitation boundary by controlling the degree of supersaturation via controlling the oxidation rate of Fe(II). However, it should be noted that oxygen injection in practical industrial applications may cause significant heat loss because a large amount of steam is released. In our work, we controlled the degree of supersaturation via the dropwise addition of Fe(III) to the As(V) solution and avoided the potential large consumption of energy.

Stability Tests on Scorodite

The TCLP results of the precipitated scorodite are shown in Table . The concentrations of the leached elements of concern were 0.9 mg/L As, 0.4 mg/L Pb, and 0.4 mg/L Zn for scoroditeACRreal, as well as 0.5 mg/L As, 0.1 mg/L Pb, 0.1 mg/L Zn, and 0.2 mg/L Cd for scoroditeACRlab and both are below the identification standard for hazardous solid wastes (As, 5.0 mg/L; Pb, 5.0 mg/L; and Cd, 1.0 mg/L). The short-term stability test showed that the concentrations of As after 15 days of leaching were 1.7, 1.3, and 2.6 mg/L for scoroditeACRreal, while these reached 0.7, 0.4, and 3 mg/L for scoroditeACRlab at pH values of 3, 5, and 7, respectively (Figure ). Thus, the scorodite synthesized in this study had excellent short-term stability at various pH values. However, to further improve the stability and reduce the risk of scorodite, the precipitated scorodite could be microencapsulated by an inert material such as aluminum phosphate and aluminum silicate.[45−48]

Table 7

Concentrations of As, Ca, and Trace Metals in TCLP Testsa

element	As	Cu	Zn	Pb	Cd
TCLPreal	0.9	UD	0.4	0.4	UD
TCLPlab	0.5	UD	0.1	0.1	0.2

The unit of concentration is mg/L.

Figure 6

Short-term stability test of the produced scorodite from ACRreal (a) and ACRlab (b) at pH values of 3, 5, and 7.

Recovery of Trace Metal Cations as Metal Sulfides

The postreaction solution after scorodite synthesis contained a considerable amount of metal cations Cu2+, Zn2+, and Cd2+ (Table ) and as such needs further treatment. The metal cations can be removed via the formation of metal sulfides because of their lower solubilities.[49] The dissolution equilibrium equation and log(solubility product) (log Ksp) of the metal sulfides are summarized by eqs –13. After the addition of S2– into the postreaction solution, almost all Cu2+, Zn2+, and Cd2+ can be recovered as metal sulfides such as CuS, ZnS, and CdS (data not shown).

Industrial Application Problems

Treatment of Gypsum, Washing Liquor, and Postreaction Solution

Figure shows the water usage and As balance for the production of scorodite from ACRreal and ACRlab, and how As ions from the starting ACR were distributed among the solid and aqueous phases (e.g., gypsum, scorodite, washing liquor, and postreaction solution). The As content in gypsum-ACRreal (5400 mg/kg) was much higher than the standard value for the hazardous industrial solid waste of 0.1 wt % (Table ), and hence it is classified as a hazardous solid waste (GB5085.62007, China). This kind of As-bearing gypsum could be recycled as applicable gypsum and/or anhydrite via the hydrothermal recrystallization in acid solutions.[2] The washing liquor and the postreaction solution are also a concern due to their high contents of As and trace metal cations (Figure S1 and Table ). The washing liquor can be reused in the dispersion of ACR solids or the treatment of smelter off-gasses in the Cu refining processes. The H2SO4-containing postreaction solution from scorodite synthesis could be used in the ACR dissolution–neutralization process. Notably, the water usage in the present study does not represent the practical industrial applications because all of the output solids were washed five times for solid characterization.

Figure 7

Simplified flowsheet and mass balance of As and water usage for the treatment of ACRreal and ACRlab.

Cost of Chemical Reagents

To verify the economic feasibility in practical industrial applications, the cost of chemical reagents in the treatment of ACRreal by adopting this process was evaluated (Table ), where H2SO4, H2O2, and Fe2(SO4)3 were used for the disposal of ACRreal. The cost was calculated according to the contents of As(T), Ca, and S in ACRreal (3.8, 28, and 7.4 wt %, respectively). The calculated cost for chemical reagents was nearly 28.5 US$/t (about 199 Chinese RMB/t) for the treatment of ACRreal. If waste sulfuric acid was used as the initial H2SO4 for dissolution–neutralization of the ACR, the total cost could be reduced to 10.5 US$/t (about 74 Chinese RMB/t).

Table 8

Economic Evaluation of Industrial Scale for per ton ACRreala

process chemical reagents	unit price (US $/t)	doses (kg/t)	cost (US $/t)
dissolution–neutralization 98% H2SO4	115	156	18
As(III) oxidation 28% H2O2	143	21.5	3.1
precipitation–crystallization Fe2(SO4)3	185	40	7.4
total cost			28.5

The unit price of chemical reagents was from China suppliers.

Industrial Applications

Based on the proposed ACR treatment method and the hydrometallurgical conversion of As(V) to scorodite at ambient pressure conditions,[43] we suggest the cotreatment of waste sulfuric acid and ACR to reduce the reagent costs and economize the water resources. Figure presents an example of a process flow that incorporates the hydrometallurgical treatment of ACR and the scorodite process. The integration of the processes was based on the use of waste sulfuric acid as the starting H2SO4 solution for the dissolution–neutralization of ACR. The use of waste sulfuric acid for the dissolution–neutralization of ACR will offer multiple advantages including the increase of As concentration for scorodite crystallization and optimal use of wastewater. As can be seen, in the integrated treatment process, the As in waste sulfuric acid and ACR can be immobilized as scorodite and the SO42– and trace metals Cu2+, Zn2+, Pb2+, and Cd2+ can be recycled as gypsum and metal sulfides.

Figure 8

Flow diagram integrating the hydrometallurgical treatment of ACR and the scorodite process.

Conclusions

In the present work, an alternative treatment method for arsenic–calcium residue (ACR) via immobilization of As as scorodite was proposed. The major contributions of this work include the following: (1) this is the first attempt to leach As from ACR and then immobilize As as stable scorodite; (2) this work presents an alternative method for the treatment of ACR; (3) the As content in the precipitated gypsum from ACRlab was only 68 mg/kg, which is below the Chinese national standard for hazardous solid wastes (<0.1 wt %, GB5085.62007); (4) gypsum with large particles could be formed under acidic conditions; (5) the ORP and pH value could be indicators of the endpoint of As(III) oxidation process; and (6) the final scorodite have high stability for safe disposal.

The immobilization of As from waste sulfuric acid as ACR via the lime neutralization–precipitation method has been widely used due to its low operating cost and simple process in some countries, such as China. ACR is a critical As-containing hazardous solid waste in some trailing ponds and abandoned mining sites that could cause serious As pollution in surrounding soil and water systems. Scorodite is an ideal As carrier because of its high As content and high stability. The present study proposed an alternative method for the treatment of ACR via the immobilization of As as scorodite. The proposed treatment method for ACR in the present study may be suitable for the treatment of arsenic–calcium residue and reduce its environmental risks.

Experimental Section

Materials

All chemicals were of analytical grade, purchased from Sigma-Aldrich Company Ltd., and used without further purification. Deionized (DI) water was used in all experiments. All glassware were cleaned by soaking in 5% HNO3 for at least 12 h and rinsed three times with DI water before use. Two types of ACRs were studied in the present work. The real ACR (defined as ACRreal) was directly collected from a Pb refinery, which was produced in a one-step neutralization–precipitation process of waste sulfuric acid to pH 12 ± 0.5 using CaO/CaCO3 as neutralizing reagents. For comparison with the ACRreal, another type of ACR was precipitated in the lab (defined as ACRlab) via a two-step lime neutralization of waste sulfuric acid obtained from the Pb refinery. The chemical composition of waste sulfuric acid is presented in Table S2. Briefly, the waste sulfuric acid was neutralized to the desired pH (∼2) by adding slacked lime (2 mol/L Ca(OH)2) for the precipitation of H2SO4 as CaSO4·2H2O (gypsum), followed by solid/liquid separation. Then, the solids were filtered and washed five times with an acidified saturated pure gypsum solution (HCl 0.018 mol/L, Ca2+ 0.02 mol/L, SO42– 0.02 mol/L, pH ∼ 2) to remove the interparticle-entrained residual solution in solids. The chemical composition of the precipitated gypsum is presented in Table S3. The concentrations of As and trace metals Cu, Zn, Pb, and Cd in the washing liquor were monitored, and it was found that five times washing can fully remove the residual solution in solids (Figure S1). Then, ACRlab was obtained by further neutralization of the remaining filtrate after the first-step neutralization to pH 12 ± 0.1 using slacked lime (2 mol/L Ca(OH)2) with the slurry stirred vigorously (300 rpm) and then solid/liquid separation. Both ACRreal (63.6 dry wt %) and ACRlab (31.2 dry wt %) were used for the treatment without drying due to which the metal hydroxides such as Cu(OH)2 can easily transform into metal oxides such as CuO during the drying process.

Dissolution–Neutralization of the ACRs

The ACRreal or ACRlab sludge was dissolved by slowly adding (5 mL/min) to the sulfuric acid solution (1 mol/L H2SO4). The system was maintained at pH < 2 by the continuous addition of 2 mol/L H2SO4 solution with the slurry stirred vigorously (300 rpm). After reaction for 2 h, the slurry was filtered using a 0.22 μm membrane. The solids were washed five times with the acidified saturated-gypsum solution at a solid/liquid ratio of 1 g/5 mL to remove the interparticle-entrained residual solution in solids. Then, the obtained solids (gypsum) were vacuum-dried at 40 °C for 24 h. The As-enriched solution was retained for arsenic immobilization treatment.

Oxidation of As(III)

Hydrogen peroxide (20% H2O2) was added to the above-mentioned As-enriched solution (the concentrations of As(III) were 6.5 and 12.8 g/L for ACRreal and ACRlab systems, respectively) in a dropwise mode using a peristaltic pump at a rate of 1 mL/min. The ACRreal and ACRlab systems were stabilized for 20 min after every addition of 1 or 0.5 mL (20%) of H2O2 (the molar ratios of H2O2/As(III) ∼ 1/10 for each time). The oxidation–reduction potential (ORP) was measured before and after every addition of 1 or 0.5 mL of H2O2 during the oxidation process. The addition of H2O2 was terminated when the ORP did not decrease significantly. The detailed ORP data are shown in the Results and Discussion section (Figure ).

As(V) Immobilization in the Form of Scorodite

The above-mentioned As(V)-containing solution (10.1 and 15.8 g/L As(V) for ACRreal and ACRlab systems, respectively) was heated to the desired temperature (95 °C). Ferric sulfate solution (from Fe2(SO4)3·9H2O, 100 mL of 30 g/L Fe(III) and 100 mL of 63 g/L Fe(III) used for ACRreal and ACRlab systems, respectively) was then added to the As(V)-containing solution in a dropwise mode with a peristaltic pump at a rate of 25 mL/h in 4 h to reach the target Fe/As molar ratio of approximately 1. Then, the mixture was further stirred for 4 h. The solids and liquids (pH 0.95) were then separated by pressure filtration through a 0.22 μm membrane. The supernatants were analyzed for the concentrations of As, Cu2+, Zn2+, Pb2+, and Cd2+. The solids were washed three times with HCl solution (pH ∼1) and then vacuum-dried at 40 °C for 24 h.

Recovery of Metal Cations as Metal Sulfides

The slaked lime (2 mol/L Ca(OH)2) was slowly added to the above-mentioned supernatants, and the pH value of the slurry was increased to approximately 2. The solids and liquids were then separated by pressure filtration through a 0.22 μm membrane. A 0.01 mol/L Na2S solution was added to the filtrate slowly with a peristaltic pump at a rate of 5 mL/min until the pH value increased to approximately 6.5. Stirring was continued for 1 h, and the solid was separated by pressure filtration and vacuum-dried at 40 °C for 24 h. The supernatants were analyzed for the concentrations of Cu2+, Zn2+, Pb2+, and Cd2+.

Determination of the Concentrations of As, Ca, SO4, and Trace Metals in the Liquid/Solid Phase

A known amount of solids was digested in 6 mol/L HCl for the analysis of the concentrations of As, trace metals, Ca, and SO4. An atomic fluorescence spectrometer coupled with a hydride generator (HG-AFS, Haiguang, China) was used to determine the concentrations of total As (As(T)) and As(III). The detection limit for As was 0.01 μg/L with an uncertainty of ±5%. For As(T) detection, the testing solution was pretreated with a mixed thiourea/ascorbic acid agent (5%) and diluted with 5% HCl before HG-AFS measurement. For As(III) detection, a pH 5.0 disodium citrate buffer (0.5 mol/L) was used instead of thiourea/ascorbic acid agent and 5% HCl during the HG-AFS analysis. The concentration of As(V) was calculated as the difference between As(T) and As(III). Atomic absorption spectroscopy (AAS, Varian) was used to determine the concentrations of Pb, Cd, Zn, Cu, and Ca. The detection limits for Pb, Cd, Zn, Cu, and Ca were approximately 10 μg/L with an uncertainty of ±5%. The concentration of SO42– was determined using a nephelometric method on a UV-2550 visible spectrophotometer (UV, Shimadzu, Japan) at a wavelength of 420 nm. The detection limit for SO42– was 10 μg/L with an uncertainty of ±5%. The redox potentials during the oxidation of As(III) were measured by an electrometric method using an INESA 501 ORP platinum electrode. For Eh, the measured data collected with platinum and saturated calomel electrodes (Ag/AgCl) were converted to the standard hydrogen electrode reference. The pH values during the oxidation process were measured using a pH meter (PB-10, Sartorius, Germany).

Solid-Phase Characterization

The mineralogy of the output solids was characterized on a Rigaku D/max 2400 (XRD) X-ray diffractometer (Rigaku Corporation, Japan) equipped with a copper target (Cu Kα1 radiation, λ = 1.5418), a crystal graphite monochromator, and a scintillation detector. The morphologies of the output solids were analyzed on a scanning electron microscope combined with an energy-dispersive X-ray spectrometer (SEM-EDX, S-3400N, Hitachi, Japan). The samples were mounted on pin stubs with the use of double-sided carbon tape and sputter-coated with gold. All images were collected at 25 kV and a magnification factor of 5000. Besides, the gypsum particles in ACRreal were mounted in the cold setting epoxy resin. The solidified epoxy resin was polished to create a cross section of the particle. Then, the cross section of the gypsum particle was analyzed by EDX to observe the internal elemental composition.

Stability Evaluation of the Produced Gypsum and Scorodite

The toxicity characteristic leaching procedure (TCLP) proposed by the United States Environmental Protection Agency (US EPA) was employed to determine the stability of the produced gypsum and final scorodite. This procedure consists of leaching a solid sample for 18 h in an acetic acid–sodium acetate buffer solution of pH 4.95 at a liquid/solid proportion of 20/1 with agitation at 50 rpm and a temperature of 22 °C. The identification standards of hazardous solid wastes for As, Pb, and Cd leaching concentration are regulated as 5.0, 5.0, and 1.0 mg/L, respectively.[50]

Short-term stability tests were performed under atmospheric conditions by adding 1 g of scorodite to 100 mL of HCl solution at pH values of 3, 5, and 7. During the stability test, the pH of the slurry was maintained constant with 0.01/0.1 mol/L NaOH and HCl solutions. Before stability testing, the solids were subjected to a surface cleaning procedure by stirring a 3% solid suspension at pH 1 and room temperature (22 °C) for 24 h to ensure that any amorphous arsenate phase was removed.[51]