Deep Copper Removal from High-Arsenic, Low-Copper Spent Copper Electrolyte by Gas–Liquid Sulfidation

Abstract

The separation of copper and arsenic from spent copper electrolyte plays a pivotal role in electrolyte recirculation and arsenic-bearing solid hazardous waste minimization. In this study, the deep copper removal process in high arsenic and low copper spent copper electrolyte by gas–liquid sulfidation is studied. Thermodynamic analysis indicates that under strongly acidic conditions, regulating the oxidation-reduction potential enables the selective precipitation of Cu²⁺ as CuS while inhibiting the formation of As₂S₃. The influence of hydrogen sulfide excess coefficient and gas–liquid sulfidation temperature on copper and arsenic co-precipitation behavior is investigated. Under the optimal gas–liquid sulfidation conditions with the sulfide excess coefficient of 47 and gas–liquid sulfidation for 60 min at 328.15 K, the copper concentration can be reduced from 0.312 g/L to 1.25 mg/L, while arsenic co-precipitation can be effectively suppressed. The copper gas–liquid sulfidation process is chemical reaction and diffusion mix controlled with an activation energy of 33.47 kJ/mol, while arsenic sulfidation is chemical reaction controlled with an activation energy of 51.22 kJ/mol. The copper–arsenic co-precipitated sludge predominantly consists of As₂S₃, CuS, and Cu₂S. Arsenic precipitation involves a multi-step process: As(V) is first reduced to As(III) and subsequently sulfurized. However, the majority of cupric ions are directly precipitated as sulfides, whereas a minor fraction is firstly reduced by hydrogen sulfide and subsequently precipitated. The present study clarifies the intrinsic mechanism and external regulatory factors for the gas–liquid sulfidation deep copper removal process, providing a theoretical basis for optimizing sulfidation processes to synergistically achieve valuable metal recovery and arsenic pollution control.

1. Introduction

Large volumes of waste electrolyte generated from the copper electrorefining process are classified as typical hazardous waste due to their high concentrations of arsenic and copper as well as high acidity [1,2,3]. This copper–arsenic polymetallic acidic wastewater not only contains hazardous arsenic but is also rich in valuable metallic elements such as copper, offering potential for resource recovery [4,5]. The treatment of such wastewater focuses on achieving high-precision separation of valuable metals from arsenic, thereby simultaneously mitigating environmental pollution and promoting resource utilization [6,7]. Conventional neutralization–precipitation methods tend to yield arsenic-bearing hazardous residues and are often inadequate for efficient recovery of valuable metals [8]. In contrast, the sulfide precipitation process, which enables selective stepwise recovery of metal ions based on differences in sulfide solubility products, offers advantages such as fast reaction kinetics, low residue yield, high precipitate stability, and the possibility of direct recycling of precipitates to the smelting system. As a result, it has attracted significant attention in recent industrial research [9,10,11,12].

Traditional sulfide precipitation processes are typically conducted under a single set of conditions, leading to the simultaneous precipitation of all targeted metal ions [13]. However, in waste copper electrolytes where multiple metal ions coexist, significant differences exist in the optimal dosage of sulfurizing agents and the precipitation kinetics for their respective sulfides [14,15]. Consequently, the conventional two-stage sulfide precipitation method often results in co-precipitation of various metals, generating a mixed-composition sulfide residue [16,17]. During subsequent resource recovery processes, impurities such as arsenic in this mixed residue can re-enter the production circuit, causing cyclic enrichment and pollution. Moreover, the presence of arsenic severely compromises the quality of the recovered copper product [18,19]. Jiang et al. reported that controlling the dosage of hydrogen sulfide (H₂S) achieved a relatively satisfactory separation between copper and arsenic [20]. Nonetheless, as the copper concentration decreases, the increasing concentration gradient between copper and As(V) intensifies their competition for H₂S. Once As(V) is reduced to As(III), its rapid reaction with H₂S leads to the inevitable formation of As₂S₃ during copper recovery, thereby increasing the arsenic content in the copper sulfide residue [21]. In contrast, the three-stage sulfide precipitation process has emerged as one of the most promising technological approaches, owing to its potential for achieving stepwise separation and separate enrichment of valuable metals. In the three-stage sulfidation process, the first stage is designed to achieve selective sulfide precipitation of copper with minimal co-precipitation of arsenic, enabling the sulfide precipitate to be directly returned to the smelting circuit. The second stage aims to reduce the copper ion concentration in solution to below 0.1 g/L while suppressing arsenic precipitation to the greatest extent. The third stage serves to completely precipitate the residual arsenic as sulfide, which is subsequently processed to produce metallic arsenic.

The secondary-stage sulfidation process within the three-stage sulfidation technique presents unique and complex challenges [22,23]. Firstly, its feed consists of the pregnant solution from the first-stage treatment, in which the speciation of metal ions and the solution chemistry have already been altered. The influence of gas–liquid sulfidation conditions on the selectivity of copper–arsenic separation in the secondary stage remains poorly understood. Secondly, the sulfide precipitation regions of copper and arsenic overlap thermodynamically [24], making the process control window required to achieve near-complete separation under practical kinetic competition extremely narrow and sensitive [16,25]. Furthermore, the secondary-stage sulfidation must not only accomplish efficient separation, but the properties of its products must also meet the requirements of downstream processing steps.

The sulfidation method, particularly using hydrogen sulfide as the precipitant, is regarded as an efficient and economically viable technology for treating copper–arsenic acidic wastewater [26,27,28,29]. However, its practical application in achieving highly selective separation of copper and arsenic still faces significant challenges, mainly stemming from the similarity and complexity of their chemical behaviors. Firstly, although the solubility product (K_sp ≈ 10⁻³⁶) of CuS is much lower than that of As₂S₃ (K_sp ≈ 10⁻²²) [30], which thermodynamically favors the preferential precipitation of copper, the dynamic addition of H₂S allows As(III) to compete with Cu²⁺ for limited S²⁻, leading to severe co-precipitation of arsenic [31,32,33]. Moreover, arsenic in actual wastewater often exists in the As(V) form, which must be reduced to As(III) before it can react with S²⁻. This multi-step reduction process is kinetically slow and coupled with the copper precipitation pathway, making the reaction difficult to control [34,35]. At the same time, subtle variations in process parameters, such as sulfide excess coefficient, reaction time, and temperature, can significantly affect the precipitate composition. Excessive H₂S, while ensuring complete copper precipitation, exacerbates arsenic co-precipitation. Whereas insufficient dosage fails to lower the copper ion concentration to meet the feed standard for the subsequent third-stage process. Therefore, achieving efficient and selective precipitation requires precise control over the reaction process.

Beyond the scope of single-stage sulfidation, recent advances in process integration and synergistic treatment strategies have provided new insights for complex systems. For instance, a novel reagent mixture was designed to simultaneously inhibit corrosion and salt precipitation in the petroleum industry [36], highlighting the potential of multifunctional reagent design for multiphase reaction control. Likewise, an ambient-temperature synthesis of amine-rich demulsifiers revealed hydrogen bond-driven demulsification mechanisms [37], demonstrating the importance of molecular-level interaction regulation. These integration concepts are increasingly relevant to the treatment of copper–arsenic acidic wastewater, where emerging techniques such as nanomaterial-assisted sulfide removal and advanced gas–liquid reaction engineering are being explored for selective metal recovery. Inspired by these strategies, the present study aims to achieve deep copper removal with minimal arsenic co-precipitation via precisely controlled gas–liquid sulfidation in the secondary stage.

This study focuses on the reaction mechanisms and key influencing factors of the secondary-stage sulfidation process within a three-stage sulfidation system. The core objective of this stage is to achieve highly efficient and selective separation of copper and arsenic: ensuring minimum arsenic precipitation while deeply purifying the copper ion concentration in the solution to below 0.1 g/L, thereby meeting the feed requirements for the subsequent third-stage sulfidation. The research employed the first-stage sulfidation solution (an arsenic-rich solution) as the treatment target, and systematically investigated the effects of sulfide excess coefficient, reaction temperature, and time on the precipitation behaviors of copper and arsenic. The study first established a theoretical framework through thermodynamic analysis, calculating the E-pH diagrams and species distribution diagrams for the specific system based on the actual ionic composition of the solution. This provided a theoretical verification of the feasibility of selective precipitation. Building on this foundation, sulfidation kinetic experiments at different temperatures, combined with monitoring of solution concentrations at various time points, were conducted to analyze in depth the kinetic differences in copper/arsenic sulfide precipitation. The study systematically elucidates the intrinsic mechanisms and external controlling factors of selective copper/arsenic separation during secondary-stage sulfidation, aiming to provide a solid theoretical basis and process design guidance for an efficient and controllable deep copper removal technology.

2. Materials and Methods

2.1. Chemical Agents and Materials

Analytical reagent-grade sulfuric acid and sodium hydrosulfide were employed to generate hydrogen sulfide gas. The arsenic-rich spent copper electrolyte solution was supplied by a local copper electrorefining plant. All reagents and materials were used as received without further purification.

2.2. Sample Preparation

All sulfide precipitation experiments were conducted using the arsenic-rich solution, the composition and concentration of which are listed in

Table 1. Chemical composition of the arsenic-rich solution.

Content	Cu (g/L)	As (g/L)	Sb (g/L)	Bi (mg/L)	Ni (g/L)	H+ (mol/L)
	0.312	7.2	0.129	0.235	8.5	5.07

. Hydrogen sulfide (H₂S) gas, used as the sulfur source, was generated in situ by reacting sodium hydrosulfide (NaHS) with dilute sulfuric acid, as shown in Equation (1). The required mass of NaHS was determined based on the S/Cu molar ratio, calculated according to Equation (2).

2NaHS + H₂SO₄ = Na₂SO₄ + 2H₂S

(1)

$$M_{\mathrm{NaHS}}={\displaystyle \frac{{\mathsf{\rho }}_{\mathrm{Cu}}\mathrm{V}{R}_{\mathrm{S}/\mathrm{C}\mathrm{u}}}{M_{\mathrm{Cu}}}}$$

(2)

where M_NaHS (mol) is the amount of NaHS used; M_Cu (g/mol) is the molar mass of copper; ρ_Cu (g/L) is the concentration of copper ions in the electrolyte; R_S/Cu is the excess coefficient of H₂S (S and Cu molar ratio); V (mL) is the volume of arsenic-rich solution used.

The experimental setup consisted of four main systems: a sulfurizing agent preparation system, a circulation system, an oxidation-reduction potential (ORP) detection system, and a sulfide precipitation system, as illustrated in Figure 1. A total of 200 mL of the arsenic-rich solution was placed in a 400 mL sealed glass reactor equipped with a water bath and a magnetic stirrer. The ORP during the experiment was monitored by the ORP detection system. Subsequently, varying doses of H₂SO₄ and NaHS were added to the sulfurizing agent preparation system to generate hydrogen sulfide (H₂S). Using a micro air pump, the reaction was conducted for 60 min under different S/Cu molar ratios and temperatures, with a gas flow rate of 800 mL/min. Electrolyte samples were collected at specified time intervals. After standing for 8 h, the samples were filtered under vacuum to separate the liquid and solid phases. The collected solids were dried in a constant-temperature oven at 338.15 K for 12 h and subsequently subjected to analysis. Concentrations of copper and arsenic were measured by inductively coupled plasma (ICP) spectrometry. The sulfur content in the solids was determined by thermogravimetric analysis (DSC-TG).

In this study, precipitation experiments were conducted on the arsenic-rich solution under varying dosages of sulfurizing agent, different temperatures, and different reaction times. Subsequently, to further investigate the reaction mechanism of the sulfidation process, kinetic calculations were performed based on the precipitation rates of copper and arsenic at different temperatures and time intervals. The precipitation rates of Cu and As were calculated using the following equations:

$$\eta ={\displaystyle \frac{C_0-C_t}{C_0}}$$

(3)

where η (%) is the sulfidation precipitation rate of Cu or As, C₀ (g/L) is the initial concentration of Cu or As, and C_t (g/L) is the concentration of Cu or As at time.

3. Results and Discussion

3.1. Thermodynamic Analysis

Based on thermodynamic analysis of the potential–pH diagrams for the As-S-H₂O and Cu-S-H₂O systems at 298.5 K to 328.5 K, as shown in Figure 2 and Figure 3, this study identifies the key potential window for achieving selective precipitation of copper while suppressing co-precipitation of arsenic. Under the strongly acidic conditions (pH < 2), the stable forms of arsenic are highly dependent on the redox potential: in the higher redox potential region (E > 0.4 V vs. SHE), arsenic exists predominantly as soluble arsenic acid (H₃AsO₄) and its ionic species, whereas solid As₂S₃ is thermodynamically stable only in the lower reduction potential region (E < 0.2 V). In contrast, cupric sulfide remains thermodynamically stable over a wide redox potential range, including the region where arsenic is stable at higher potentials. Therefore, a clear selective precipitation window exists. by precisely controlling the system redox potential within this window, Cu²⁺ can be thermodynamically driven to precipitate selectively as CuS, while arsenic is retained in solution as oxyanionic species, thereby fundamentally inhibiting the formation of As₂S₃. Although increasing temperature slightly shifts the phase boundaries, this selective window persists, providing a theoretical basis for conducting kinetic studies at multiple temperatures while maintaining selective control. This analysis establishes a solid theoretical foundation for the innovative experimental strategy employed in this study, in which ORP is used as the core control parameter to systematically investigate the effects of H₂S excess coefficient and temperature on precipitation behavior and kinetics.

Arsenic primarily exists in the oxidation states of –3, 0, +3, and +5. In most compounds, arsenic exhibits positive oxidation states, with +3 and +5 being the most stable and common, widely found in oxides, oxyacids, and their corresponding salts. During the copper electrorefining process, arsenic impurities from the anode dissolve into the electrolyte as trivalent ions (As(III)). Under the action of dissolved oxygen, As(III) is gradually oxidized to As(V). Therefore, arsenic in the electrolyte exists in both As(III) and As(V) forms, with As(V) being the predominant species. The relevant chemical equilibrium reactions and their equilibrium constants for the As(V)-H₂O solution system are listed in

Table 2. Equilibrium constants of As(V)-related reactions at different temperatures.

No.	Ionization Reaction	Ionization Equilibrium Constant	lg k
			298 K	308 K	318 K	328 K
(4)	${\mathrm{H}}_3{\mathrm{AsO}}_4\rightleftharpoons {\mathrm{H}}_2{\mathrm{AsO}}_4^{-}+{\mathrm{H}}^{+}$	${\mathrm{k}}_1={\displaystyle \frac{\left[{\mathrm{H}}^{+}\right]\left[{\mathrm{H}}_2{\mathrm{AsO}}_4^{-}\right]}{\left[{\mathrm{H}}_3{\mathrm{AsO}}_4\right]}}$	−2.258	−2.301	−2.347	−2.395
(5)	${\mathrm{H}}_2{\mathrm{AsO}}_4^{-}\rightleftharpoons {\mathrm{HAsO}}_4^{2-}+{\mathrm{H}}^{+}$	${\mathrm{k}}_2={\displaystyle \frac{\left[{\mathrm{H}}^{+}\right]\left[\mathrm{H}{\mathrm{AsO}}_4^{2-}\right]}{\left[{\mathrm{H}}_2{\mathrm{AsO}}_4^{-}\right]}}$	−6.772	−6.759	−6.756	−6.763
(6)	$\mathrm{H}{\mathrm{AsO}}_4^{2-}\rightleftharpoons {\mathrm{AsO}}_4^{3-}+{\mathrm{H}}^{+}$	${\mathrm{k}}_3={\displaystyle \frac{\left[{\mathrm{H}}^{+}\right]\left[{\mathrm{AsO}}_4^{3-}\right]}{\left[\mathrm{H}{\mathrm{AsO}}_4^{2-}\right]}}$	−11.604	−11.508	−11.432	−11.374

.

In the As(V)-H₂O system, the aqueous solution contains H₂AsO₄, ${\mathrm{H}}_2{\mathrm{AsO}}_4^{-}$, $\mathrm{H}{\mathrm{AsO}}_4^{2-}$, and ${\mathrm{AsO}}_4^{3-}$. Let α₀, α₁, α₂, and α₃ represent the distribution coefficients of these four species in the aqueous solution, respectively, i.e.,:

$${\mathsf{\alpha }}_0={\displaystyle \frac{\left[{\mathrm{H}}_3{\mathrm{AsO}}_4\right]}{\left[{\mathrm{As}}_{\mathrm{T}}\right]}}$$

(7)

$${\mathsf{\alpha }}_1={\displaystyle \frac{\left[{\mathrm{H}}_2{\mathrm{AsO}}_4^{-}\right]}{\left[{\mathrm{As}}_{\mathrm{T}}\right]}}={\mathrm{k}}_1{\displaystyle \frac{\left[{\mathrm{H}}_3{\mathrm{AsO}}_4\right]}{\left[{\mathrm{H}}^{+}\right]\left[{\mathrm{As}}_{\mathrm{T}}\right]}}={\mathsf{\alpha }}_0{\mathrm{k}}_1{10}^{\mathrm{pH}}$$

(8)

$${\mathsf{\alpha }}_2={\displaystyle \frac{\left[\mathrm{H}{\mathrm{AsO}}_4^{2-}\right]}{\left[{\mathrm{As}}_{\mathrm{T}}\right]}}={\mathrm{k}}_2{\displaystyle \frac{\left[{\mathrm{H}}_2{\mathrm{AsO}}_4^{-}\right]}{\left[{\mathrm{H}}^{+}\right]\left[{\mathrm{As}}_{\mathrm{T}}\right]}}={\mathsf{\alpha }}_0{\mathrm{k}}_1{{\mathrm{k}}_{2}10}^{2\mathrm{pH}}$$

(9)

$${\mathsf{\alpha }}_3={\displaystyle \frac{\left[{\mathrm{AsO}}_4^{3-}\right]}{\left[{\mathrm{As}}_{\mathrm{T}}\right]}}={\mathrm{k}}_3{\displaystyle \frac{\left[{\mathrm{HAsO}}_4^{2-}\right]}{\left[{\mathrm{H}}^{+}\right]\left[{\mathrm{As}}_{\mathrm{T}}\right]}}={\mathsf{\alpha }}_0{\mathrm{k}}_1{{\mathrm{k}}_2{\mathrm{k}}_{3}10}^{3\mathrm{pH}}$$

(10)

At the same time,

$$\left[{\mathrm{As}}_{\mathrm{T}}\right]=\left[{\mathrm{H}}_3{\mathrm{AsO}}_4\right]+\left[{\mathrm{H}}_2{\mathrm{AsO}}_4^{-}\right]+\left[\mathrm{H}{\mathrm{AsO}}_4^{2-}\right]+\left[{\mathrm{AsO}}_4^{3-}\right]$$

(11)

α₀ + α₁ + α₂ + α₃ = 1

(12)

Substituting Equations (4) and (5) into Equation (9), the relationships between α₀, α₁, α₂, α₃ and the solution pH are obtained as follows:

$${\mathsf{\alpha }}_0={\displaystyle \frac{1}{1+{\mathrm{k}}_1{10}^{\mathrm{pH}}+{\mathrm{k}}_1{\mathrm{k}}_2{10}^{2\mathrm{pH}}+{\mathrm{k}}_1{\mathrm{k}}_2{\mathrm{k}}_3{10}^{3\mathrm{pH}}}}$$

(13)

$${\mathsf{\alpha }}_1={\displaystyle \frac{{\mathrm{k}}_1{10}^{\mathrm{pH}}}{1+{\mathrm{k}}_1{10}^{\mathrm{pH}}+{\mathrm{k}}_1{\mathrm{k}}_2{10}^{2\mathrm{pH}}+{\mathrm{k}}_1{\mathrm{k}}_2{\mathrm{k}}_3{10}^{3\mathrm{pH}}}}$$

(14)

$${\mathsf{\alpha }}_2={\displaystyle \frac{{\mathrm{k}}_1{\mathrm{k}}_2{10}^{2\mathrm{pH}}}{1+{\mathrm{k}}_1{10}^{\mathrm{pH}}+{\mathrm{k}}_1{\mathrm{k}}_2{10}^{2\mathrm{pH}}+{\mathrm{k}}_1{\mathrm{k}}_2{\mathrm{k}}_3{10}^{3\mathrm{pH}}}}$$

(15)

$${\mathsf{\alpha }}_3={\displaystyle \frac{{\mathrm{k}}_1{\mathrm{k}}_2{\mathrm{k}}_3{10}^{3\mathrm{pH}}}{1+{\mathrm{k}}_1{10}^{\mathrm{pH}}+{\mathrm{k}}_1{\mathrm{k}}_2{10}^{2\mathrm{pH}}+{\mathrm{k}}_1{\mathrm{k}}_2{\mathrm{k}}_3{10}^{3\mathrm{pH}}}}$$

(16)

Therefore, using the values of k₁, k₂, and k₃ at different temperatures, the distribution coefficients α₀, α₁, α₂, and α₃ for H₃AsO₄, ${\mathrm{H}}_2{\mathrm{AsO}}_4^{-}$, $\mathrm{H}{\mathrm{AsO}}_4^{2-}$, and ${\mathrm{AsO}}_4^{3-}$ at the corresponding temperatures can be calculated. Based on these data, the species distribution diagram for the As(V)-H₂O system within the temperature range of 298.5 to 328.5 K is plotted and shown in Figure 4.

Based on the species distribution diagram of the As-H₂O system, the regulatory effect of temperature on the speciation distribution of arsenic acid is clarified. As the temperature increases from 298.15 to 328.15 K, the stepwise dissociation equilibria of H₃AsO₄ shift systematically. Specifically, the predominant regions of H₃AsO₄ and ${\mathrm{H}}_2{\mathrm{AsO}}_4^{-}$ shift noticeably toward higher pH, whereas the dominant regions of $\mathrm{H}{\mathrm{AsO}}_4^{2-}$ and ${\mathrm{AsO}}_4^{3-}$ expand significantly toward lower pH. This trend indicates that rising temperature promotes the dissociation equilibrium of arsenic acid toward the formation of more highly charged anions, i.e., thermodynamically, increasing temperature favors the transformation of arsenic acid into more dissociated species. Within the typical acidic pH range involved in the secondary-stage sulfidation process, although H₃AsO₄ and ${\mathrm{H}}_2{\mathrm{AsO}}_4^{-}$ remain the predominant species, the temperature rise still leads to a slight increase in the relative proportion of ${\mathrm{H}}_2{\mathrm{AsO}}_4^{-}$. This temperature-dependent evolution of the initial arsenic speciation may influence the subsequent reaction pathway and kinetics of the arsenate reduction–sulfide precipitation process by altering the interaction energy between the reacting species and the reducing agent.

3.2. Effect of Gas–Liquid Sulfidation Conditions on Copper and Arsenic Precipitation

3.2.1. Hydrogen Sulfide Excess Coefficients

To investigate the effect of precipitant dosage on selective separation efficiency, the precipitation process was systematically studied at 298.15 K under different hydrogen sulfide excess coefficients (R_S/Cu = 35, 39, 43, 47). The variations in Cu and As concentrations in the solution, as well as the oxidation-reduction potential (ORP), are shown in Figure 5.

As shown in Figure 5a,b, arsenic removal efficiency and its final concentration exhibit a significant dependence on R_S/Cu [33]. As R_S/Cu increased from 35 to 47, the arsenic removal ratio after 60 min of reaction markedly improved from approximately 28% to over 40%, with the corresponding arsenic concentration in the solution decreasing from 5.32 g/L to 4.05 g/L. Notably, when R_S/Cu exceeds 43, the incremental gain in arsenic removal efficiency diminishes, indicating the existence of an optimal excess sulfur coefficient range. Beyond this range, the economic benefit from additional H₂S consumption may not be proportional to the improvement in arsenic removal.

As shown in Figure 5d,e, copper removal exhibits a similar trend. Higher R_S/Cu significantly accelerates copper precipitation, effectively reducing the final copper concentration from 0.312 g/L to 0.105 g/L. This indicates that the sulfidation precipitation process of Cu²⁺ is highly sensitive to the concentration of sulfide ions.

The difference in precipitation behavior between arsenic and copper points to fundamentally distinct rate-controlling mechanisms. The precipitation of copper is a direct and thermodynamically favorable anion-exchange process, primarily controlled by mass transfer and sulfide ion activity. In contrast, the precipitation of arsenic in acidic sulfate media is far more complex. Under these conditions, arsenic exists predominantly as arsenate (As(V), e.g., H₃AsO₄), which must first be reduced to arsenite (As(III), e.g., H₃AsO₃) before it can precipitate as As₂S₃, as shown in Equations (17) and (18). Since the reduction of arsenate to arsenite is governed by slow electron-transfer kinetics, this step is widely recognized as the rate-controlling step for the overall process, a mechanism strongly supported by the observed data. The removal rate of copper is initially rapid and shows a consistent response to increases in R_S/Cu. In contrast, arsenic removal is overall slower, and the observation that its removal efficiency improves with higher R_S/Cu suggests that excess H₂S not only supplies the necessary sulfide ions for precipitation but, more importantly, alters the reaction environment, thereby promoting the critical As(V) reduction step [22].

H₃AsO₄ + 2H⁺ + 2e⁻→H₃AsO₃ + H₂O

(17)

2H₃AsO₃ + 3H₂S→As₂S₃ + 6H₂O

(18)

Real-time monitoring of the oxidation-reduction potential (ORP) provides key in situ evidence linking macroscopic trends to underlying electrochemical mechanisms. As shown in Figure 6c, a continuous negative shift in ORP over time was observed in all experiments, confirming the progressively established reducing environment by H₂S. Furthermore, compared with experiments conducted at lower R_S/Cu, those with higher R_S/Cu consistently maintained higher ORP values throughout the reaction. This elevated ORP can be attributed to the formation of intermediate sulfur species, such as polysulfides (S_x²⁻) and elemental sulfur (S⁰) [32]. In an acidic medium containing excess H₂S and oxidizing metal ions (e.g., Cu²⁺, As(V)), H₂S can be partially oxidized, generating these sulfur intermediates that possess a higher electrochemical potential than the H₂S/S²⁻ couple. The presence of these polysulfide/sulfur intermediates creates a localized high-potential environment, which electrochemically favors the As(V) reduction step. This also explains why a higher R_S/Cu leads to a faster arsenic removal rate.

3.2.2. Gas–Liquid Sulfidation Temperature

Based on the optimal hydrogen sulfide excess coefficient (R_S/Cu = 47) determined, the effect of temperature on the precipitation behavior of copper and arsenic was further investigated within a temperature range of 298.15 to 328.15 K, with the aim of optimizing reaction kinetics and elucidating the underlying mechanism, as shown in Figure 6.

As shown in Figure 6, temperature exhibits distinctly different effects on the removal behavior of copper and arsenic. For copper, increasing temperature significantly enhances both its precipitation rate and final removal efficiency. As illustrated in Figure 6d,e, the copper concentration in the solution drops to an extremely low level after approximately 30 min of reaction at 328.15 K, and the removal rate remains essentially stable thereafter with prolonged time. In contrast, arsenic removal shows an optimal temperature window. According to Figure 6a,b, although raising the temperature accelerates arsenic precipitation during the initial stage, the final arsenic removal efficiency does not increase monotonically with temperature in the middle to later stages of the reaction. Instead, it reaches an optimum at 308.15 K, while the removal effect at 328.15 K is somewhat reduced. The decline in arsenic removal at excessively high temperatures. As a result, under the optimal gas–liquid sulfidation conditions with the sulfide excess coefficient of 47 and gas–liquid sulfidation for 60 min at 328.15 K, the copper concentration can be reduced from 0.312 g/L to 1.25 mg/L, while arsenic co-precipitation can be effectively suppressed.

This discrepancy is closely associated with the redox state of the system at elevated temperatures. The ORP monitored experimentally, as shown in Figure 6c, reveals that under identical sulfurizing agent feed conditions, the steady-state oxidation-reduction potential of the higher temperature system is significantly lower than that of the lower temperature system. The decrease in ORP can be attributed to two primary factors. From the perspective of electrochemical thermodynamics, the electrode potential itself exhibits a dependence on temperature. The standard potentials of redox systems involving H⁺ and dissolved H₂S species undergo a systematic shift as temperature increases. From the standpoint of solution chemistry behavior, elevated temperature reduces the solubility of H₂S and alters its dissociation constant (pK_a), thereby affecting the activity of the effective reducing species (HS⁻/S²⁻) in the solution. Both effects contribute to the markedly lower measured ORP at higher temperatures.

Elevated temperatures enhance the reaction kinetics of arsenic precipitation while simultaneously reducing the ORP of the system. Under the combined effects of high temperature and low ORP, the solubility of H₂S is reduced, the activity of effective reducing species (HS⁻/S²⁻) in solution is affected, and side reactions such as competitive disproportionation of sulfur species are triggered. This results in the consumption of effective sulfur sources, which in turn suppresses the final removal of arsenic. This provides a reasonable explanation for why arsenic removal efficiency decreases rather than increases at high temperatures.

3.3. Kinetic Analysis

In this study, the gas–liquid mass transfer coefficient (kLa) was not experimentally determined. The experiments were conducted in a laboratory reactor with a constant gas flow rate (800 mL/min), fixed stirring speed, and identical reactor geometry across all runs. Under these conditions, kLa is assumed to remain essentially constant, allowing direct comparison of apparent kinetic parameters derived from concentration–time data at different temperatures and sulfide excess coefficients.

The separation efficiency of copper and arsenic significantly improved as the reaction time increased. This phenomenon is primarily attributed to the competition between copper and arsenic for the limited sulfurizing agent during the sulfidation process, ultimately resulting in the formation of copper–arsenic sulfide co-precipitation, thereby establishing a co-precipitation mechanism. To further elucidate the kinetic differences between copper and arsenic during sulfidation, this study conducted a kinetic analysis using the Avrami model and calculated the apparent activation energy of the reaction based on the Arrhenius equation, as shown in Figure 7.

Based on precipitation rate data at different temperatures, the Avrami model (Equation (19)) was applied for linear fitting. The obtained linear relationships showed good agreement, indicating that the Avrami model is suitable for describing the precipitation process in this system. The apparent rate constant (k) was calculated from the slopes and intercepts of the fitting lines [12,38]. Subsequently, according to the Arrhenius equation (Equation (20)), the natural logarithm of k (ln k) was plotted against 1000/T, and linear fitting was performed to determine the apparent activation energy (E_a).

ln[−ln(1−x)] = lnk + nln(t)

(19)

lnk = lnA + Ea/RT

(20)

where n is the Avrami exponent, x is the sulfidation precipitation fraction of Cu or As, k is the apparent rate constant, t is the reaction time (s), T is the absolute temperature, R is the molar gas constant, and A is the pre-exponential factor.

Based on the Arrhenius equation, the calculated apparent activation energy for arsenic precipitation is 51.22 kJ/mol, while that for copper precipitation is 33.47 kJ/mol. According to the principles of chemical reaction engineering, the magnitude of activation energy can be used to determine the rate-controlling step of a process: typically, an activation energy below 20 kJ/mol indicates a diffusion-controlled process, above 40 kJ/mol indicates an interface chemical reaction-controlled process, and between 20 kJ/mol and 40 kJ/mol suggests a mixed control mechanism. Based on this analysis, the activation energy for copper precipitation in this study is 33.47 kJ/mol. This indicates that during the deep removal stage of secondary-stage sulfidation, the precipitation rate is not solely determined by a single diffusion step. Instead, it is jointly constrained by the mass transfer of ions to the solid surface and the surface chemical reaction step. Increasing the temperature strengthens mass transfer by reducing solution viscosity and increasing ion diffusion coefficients, while also directly accelerating the chemical reaction for CuS formation at the interface. Consequently, efficient copper removal is achieved through this dual effect. In contrast, the activation energy for arsenic precipitation is as high as 51.22 kJ/mol, indicating that its precipitation process requires overcoming a high chemical energy barrier of the reduction of As(V) to As(III), which is characteristic of a chemically reaction-controlled process.

3.4. Copper–Arsenic Co-Precipitation Residue Characterization

3.4.1. DSC-TG Analysis

To evaluate the thermal stability of the precipitate residue obtained from the secondary-stage sulfidation process, differential scanning calorimetry-thermogravimetric (DSC-TG) analysis was performed, and the results are shown in Figure 8.

The DSC-TG curve clearly reveals that the precipitate residue undergoes two main thermal events during programmed heating. The first significant thermal event occurs near 673.15 K. The TG curve shows the first rapid weight-loss step in this range, while the corresponding DSC curve exhibits a sharp, strong endothermic peak. The temperature range, peak shape, and associated weight-loss behavior of this endothermic peak align well with the characteristics of melting and vigorous volatilization of elemental sulfur. Elemental sulfur melts at 388.15 K; before reaching its boiling point of 718.15 K, its vapor pressure increases significantly, leading to substantial sublimation of sulfur molecules. This process is strongly endothermic (physical) and directly results in sample mass loss. This finding corroborates the earlier detection of elemental sulfur in the precipitate residue by X-ray photoelectron spectroscopy, indicating the presence of a certain amount of free sulfur. Its volatilization is the primary cause of the weight loss observed in the material between 620.15 K and 739.15 K.

Following the endothermic peak caused by elemental sulfur volatilization, the DSC curve exhibits a broad exothermic plateau at higher temperatures above approximately 739.15 K, accompanied by a continuous, gradual mass loss shown in the TG curve. The exothermic behavior and mass loss in this stage can be attributed to the oxidative decomposition of the main sulfide minerals present in the precipitate residue. Specifically, the major component, arsenic trisulfide (As₂S₃), can be oxidized by atmospheric oxygen within this temperature range, generating gaseous sulfur dioxide (SO₂) and arsenic oxides. This oxidation reaction is strongly exothermic. Simultaneously, the coexisting copper sulfide (CuS) may also be further oxidized to copper oxide (CuO) with the release of SO₂, collectively contributing to the exothermic effect and mass reduction observed in this stage.

Comprehensive analysis indicates that the secondary-stage sulfide precipitate residue exhibits distinct stepwise thermal decomposition characteristics. Upon heating, it first undergoes low-temperature mass loss and endothermic effects due to the volatilization of elemental sulfur; subsequently, at higher temperatures, the oxidative decomposition of the main sulfide phases becomes dominant, manifested as an exothermic reaction accompanied by further mass loss. This thermal behavior clearly demonstrates that the material becomes thermally unstable above 620.15 K, leading to the release of sulfur- and arsenic-containing gaseous species along with significant thermal effects.

3.4.2. XPS

The surface chemical status of the co-precipitation residue was analyzed using X-ray photoelectron spectroscopy (XPS), and the results are shown in Figure 9. The XPS survey scan revealed the presence of elements such as Cu, As, S, and O, as shown in Figure 9, using the C 1s peak at a binding energy of 284.80 eV as reference.

The high-resolution S 2p spectrum (Figure 9c) indicates the presence of sulfur in multiple chemical states. The fitted peaks at binding energies of 161.9 eV, 162.7 eV, 163.4 eV, and 164.5 eV correspond to Cu₂S, CuS, elemental S, and As₂S₃, respectively. The formation of elemental S (163.4 eV) consumes the effective sulfur source, which is consistent with the aforementioned phenomena of increasing ORP with higher sulfurizing agent dosage and the decrease in arsenic removal at elevated temperatures. The presence of sulfur in higher oxidation states can raise the local potential in the environment, while a reduced concentration of effective S²⁻ directly impacts the precipitation of As₂S₃. The As 3d high-resolution spectrum (Figure 9d) shows only one peak for As 3d_5/2 at a binding energy of 43.1 eV, which aligns perfectly with the As 3d peak position for the As(III)-S bond (i.e., As₂S₃). No characteristic peak for As(V) was detected. This result clearly confirms that under the optimized secondary-stage sulfidation conditions, arsenic in the solution is ultimately immobilized in the form of trivalent sulfide (As₂S₃). Under acidic conditions, As(V) primarily exists as H₃AsO₄ or ${\mathrm{H}}_2{\mathrm{AsO}}_4^{-}$. Due to its high reduction potential, As(V) cannot react directly with hydrogen sulfide; it must first be reduced to As(III) before precipitating with H₂S. The Cu LM2 spectrum (Figure 9b) exhibits two peaks at binding energies of 916.5 eV and 917.5 eV, corresponding to Cu⁺ and Cu²⁺, respectively. Combined with the analysis of the Cu 2p spectrum (Figure 9e), which shows peaks at 932.8 eV (Cu 2p_3/2) and 952.7 eV (Cu 2p_1/2), along with the distinct satellite features in the 942–962 eV range characteristic of Cu²⁺, the copper is confirmed to exist in both Cu⁺ and Cu²⁺ states. The fitted peaks at binding energies of 932.8 eV and 932.1 eV correspond to CuS and Cu₂S, respectively. Combined with the S 2p spectrum, the primary chemical forms of copper are identified as CuS and Cu₂S. The spectrum in Figure 9f shows no discernible peaks related to nickel, indicating its absence.

3.4.3. XRD and SEM

As shown in Figure 9a, the X-ray diffraction (XRD) pattern of the residue obtained after secondary-stage sulfidation treatment of waste copper electrolyte indicates that its phase composition is dominated by well-crystallized arsenic and copper sulfide compounds, and also some of the phases with a low degree of crystallinity. The main phases are identified as Cu₃AsS₄, As₂S₃, CuS, CuS₂, and metallic Cu. Characteristic peaks observed at diffraction angles of 18.5°, 22.4°, and 31.2°, among others, show a high degree of agreement with the standard diffraction data for As₂S₃. The characteristic diffraction peaks located at approximately 28.7°, 29.8°, and 48.0° correspond to the standard card of Cu₃AsS₄, indicating significant interactions among copper, arsenic, and sulfur in the reaction system, leading to the formation of a thermodynamically more stable ternary compound. Furthermore, the diffraction peaks present at 37.4°, 46°, and 48.6° correspond to CuS₂, while those at 31.8° and 32.8° correspond to CuS. The coexistence of these copper sulfides with different valence states reflects the non-uniformity of local sulfur chemical potential and redox potential during the reaction. A relatively weak diffraction peak observed at 43.3° in the pattern corresponds to the (111) crystal plane of metallic copper, which may originate from a slight reduction during the reaction. XRD detects this trace crystalline phase in the bulk, while XPS fails to detect it due to its surface-sensitive nature and the dilution of the signal by the surrounding matrix. Overall, the precipitate residue is primarily composed of As₂S₃.

Scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS) further revealed the micro-morphology and elemental distribution of the residue. As shown in Figure 10b,c, the residue after secondary-stage sulfidation exhibits a typical composite morphology consisting of irregular blocks and layered aggregates. The layered stacking structure of CuS results from the layer-like growth tendency of its crystal lattice, while the irregular blocks correspond to aggregated As₂S₃ particles. The corresponding EDS elemental mapping clearly shows that the signals of arsenic and sulfur are highly overlapped and continuous, visually confirming that the two elements have combined to form stable compounds, which is consistent with the As₂S₃ and Cu₃AsS₄ phases detected by XRD. Copper signals appear as discrete punctate distributions within this As-S matrix, and their enriched regions show spatial correlation with the As and S signals, indicating that copper mainly exists in the form of copper sulfides or Cu₃AsS₄. Quantitative analysis indicates that the atomic percentages in this micro-region are O 45.45%, S 34.18%, As 18.18%, and Cu 2.19%. The relatively high As/S atomic ratio (≈1:1.9) supports the analysis of the coexistence of multiple phases (As₂S₃, Cu₃AsS₄ and S), while the significant oxygen signal suggests a certain degree of surface oxidation or the adsorption of oxygen-containing species.

4. Conclusions

This study elucidates the underlying mechanism for achieving efficient copper–arsenic separation during the secondary stage of the three-stage sulfidation treatment process through multi-perspective analysis integrating thermodynamics, kinetics, and co-precipitation product characterization. The key lies in utilizing the thermodynamic potential difference in sulfide precipitation between copper and arsenic, and by adjusting process parameters (S/Cu ratio, temperature) to regulate the system ORP and reaction environment. This enables the kinetic differentiation and control of the mixed-controlled copper precipitation and chemical reaction-controlled arsenic precipitation processes. The research provides critical theoretical insights for optimizing the three-stage sulfidation process and explores the relationship between copper/arsenic sulfide precipitation. The following key conclusions are drawn through theoretical analysis and experimental results:

(1)

Thermodynamic analysis indicates that by examining the E-pH diagrams of the Cu-S-H₂O and As-S-H₂O systems, a thermodynamic window (E > 0.4 V vs. SHE) exists under strongly acidic conditions (pH < 2). By controlling the system’s redox potential, Cu²⁺ can be selectively precipitated as CuS while arsenic remains in solution as soluble arsenate species. This thermodynamically verifies the feasibility of copper–arsenic separation.

(2)

The optimal process parameters have been determined: the hydrogen sulfide excess coefficient (S/Cu) and the reaction temperature are key factors controlling the separation efficiency. At 298.15 K, increasing the S/Cu ratio promotes copper precipitation but intensifies arsenic co-precipitation, with diminishing returns observed when S/Cu > 43. To achieve deep copper removal while suppressing arsenic co-precipitation, the optimal conditions are identified as S/Cu = 47, temperature = 328.15 K, and reaction time = 60 min. Under these conditions, the copper concentration can be reduced from 0.312 g/L to 1.25 mg/L, and arsenic co-precipitation is effectively suppressed.

(3)

The kinetic mechanisms of copper and arsenic precipitation are fundamentally different. Kinetic analysis based on the Avrami model shows that the apparent activation energy for copper precipitation is 33.47 kJ/mol, indicating a mixed control process, while that for arsenic precipitation is 51.22 kJ/mol, corresponding to an interfacial chemical reaction-controlled mechanism.

(4)

The reaction mechanism and product characteristics have been elucidated. XPS analysis confirms that arsenic in the precipitate exists in the form of As(III)–S bonds, demonstrating that As(V) must first be reduced before forming As₂S₃. Excess H₂S not only provides a sulfur source but also generates polysulfides upon oxidation, which can increase the local redox potential, thereby promoting the reduction of As(V). This also explains why arsenic removal efficiency increases with a higher S/Cu ratio. However, excessively high temperatures (e.g., 328.15 K) significantly lower the redox potential, triggering competitive side reactions of sulfur species that consume the effective sulfur source and consequently inhibit arsenic removal.