Influence of Co-Occurring Heavy Metals on Cobalt Removal and Recovery from Wastewater by Continuous Flow In-Liquid Plasma Discharge Process

TL;DR

This study explores a new method using plasma discharge to remove and recover cobalt from wastewater, but finds that other metals can significantly reduce its effectiveness.

Contribution

The study introduces a continuous flow in-liquid plasma discharge system for cobalt removal and recovery, and quantifies the inhibitory effects of co-occurring metals.

Findings

Multi-metal systems reduced cobalt removal efficiency by up to 91%.

Copper had the strongest inhibitory effect on cobalt removal at equimolar levels.

Recovered solids formed cobalt–metal oxide composites suitable for reuse.

Abstract

Cobalt, a toxic heavy metal frequently present in wastewater, poses serious environmental and health risks but also represents a valuable resource for recovery. This study investigates a novel, environmentally friendly continuous flow in-liquid plasma discharge (CFILPD) system for simultaneous removal of cobalt, zinc, copper, and lead ions from aqueous solutions. The reactor contains two conductive channels where a stable plasma discharge forms between dielectric plates separating opposing electrodes, generating energetic electrons and hydroxyl radicals that react with dissolved metal ions. The effects of varying concentrations (5, 10, 50, and 100 ppm) of zinc, copper, and lead ions on the removal efficiency of 100 ppm cobalt ions were examined under constant conditions: 0.2 L/min argon flow rate, 200 W input power, and 50 mL/min liquid flow rate for 30 min. Results showed that increasing concentrations of co-existing metals significantly inhibited cobalt removal, with the largest reduction (91%) observed in multi-metal systems. Among individual metals at equimolar levels with cobalt, copper showed the strongest inhibitory effect (85% reduction), followed by zinc (53%) and lead (52%). Characterization of the recovered solids revealed cobalt–metal oxide composites (2.5–3 µm), suggesting their potential reuse in technological applications.

Figures

Full text

1. Introduction

Cobalt pollution in aquatic environments has become an increasing health concern, particularly in areas adjacent to industrial operations and mining zones. Due to the exceptional thermal stability, wear and corrosion resistance, and magnetic properties, cobalt (Co) has been extensively used in various industrial sectors, including in the production of lithium-ion batteries, aerospace components, machinery manufacturing, and as a catalyst in petroleum and chemical industries [1,2], all of which pose a significant threat of polluting water. Elevated cobalt concentrations have been reported in water bodies surrounding major mining sites, such as the Idaho Cobalt Belt (ICB) in the United States and the Katanga Copperbelt (KC) in the Democratic Republic of Congo [3]. Chronic exposure to high concentrations of cobalt can cause adverse health effects, with documented impacts on the blood, lungs, and skin. Therefore, its detection and efficient removal from contaminated water sources is of critical importance [4,5]. Various conventional technologies, such as adsorption, ion exchange, chemical precipitation, membrane filtration, and reverse osmosis, have been employed for cobalt remediation. Unfortunately, these approaches are often limited by low removal efficiency, high operational costs, the generation of secondary pollutants, and the need for additional chemicals, rendering them unsustainable for long-term wastewater treatment applications [6,7,8]. Therefore, alternative techniques to remove cobalt from water need to be sought. Non-thermal plasma technology (NTP) has emerged as a promising and environmentally sustainable alternative for water treatment, particularly as an advanced oxidation process. NTP processes have the capability of degrading complex and non-biodegradable pollutants through the generation of highly reactive species, including energetic electrons, free radicals (e.g., OH^•^, O^•^), positive and negative ions, and metastable molecules [9,10]. Based on the discharge medium, NTP can be categorized into gas phase or in-liquid (liquid-phase) discharge. Liquid-phase discharge provides unique advantages for water treatment, as plasma is generated directly within the liquid, allowing efficient interaction with dissolved contaminants [11,12]. The interaction of energetic electrons with water molecules during discharge events results in ionization, excitation, dissociation, and rotational or vibrational excitation, ultimately producing a spectrum of reactive species, among which hydroxyl radicals (OH^•^) are particularly important due to their highest oxidation potential (2.8 eV), enabling rapid oxidation of metal ions such as cobalt [13].

In this study, continuous flow in-liquid plasma discharge (CFILPD) refers to a process in which an electrical discharge plasma is generated within a liquid medium under continuous-flow conditions, with or without the assistance of working gas. This configuration enables in situ formation of reactive species in immediate proximity to the surrounding liquid, thereby reducing gas–liquid mass transfer limitations. Our early investigations have demonstrated the efficacy of the CFILPD process in both removing and recovering cobalt from aqueous systems, while also optimizing the key operational parameters to enhance its overall performance [14,15]. These studies provided essential mechanistic insights into the reactive species generated within the plasma–liquid interface. In particular, the combined evidence from scavenger experiments and optical emission spectroscopy (OES) analyses confirmed that hydroxyl radicals (OH^•^) are among the dominant reactive species produced in the discharge zone (Figure S2). The presence and activity of these radicals were shown to play a decisive role in the oxidative transformation and subsequent removal of cobalt ions from solutions. Moreover, the formation rate and intensity of hydroxyl radicals were strongly dependent on the applied electrical power, which directly influences plasma discharge energy and radical generation efficiency. Collectively, these findings verified that cobalt removal in the CFILPD system primarily proceeds through hydroxyl radical-mediated oxidation pathways [14,15]:

$$[eqn]$$ $$[eqn]$$ $$[eqn]$$

However, in real-world scenarios, water sources are typically contaminated with a mixture of heavy metal ions. The presence of co-occurring heavy metals can influence cobalt removal efficiency by altering speciation, forming metal complexes, or competing for reactive species and binding sites [16,17,18]. For instance, studies have shown that the recovery efficiency of the Pb^2+^ and Cd^2+^ is reduced in multi-metal systems due to competitive interactions at the electrode surface [19]. Likewise, heavy metals such as zinc, copper, lead, cadmium, and nickel may inhibit cobalt oxidation and removal depending on their redox potentials, ion concentrations, and the pH of the water matrix [20,21]. For example, increasing Zn^2+^ concentration can suppress cobalt removal by competing for oxidants, thereby reducing cobalt oxidation or precipitation [22]. Additionally, metals with higher oxidation potential may preferentially consume reactive species, decreasing the effectiveness of cobalt oxidation and subsequent cobalt removal.

Despite these interactions being recognized in other remediation contexts, there is a notable lack of research investigating how co-occurring metals affect cobalt removal in NTP-based processes. To the best of the authors’ knowledge, no prior studies have addressed this issue within the framework of liquid-phase plasma processes. Accordingly, the main objective of this work is to fill this knowledge gap by evaluating the effects of selected co-existing metals, specifically zinc (Zn^2+^), copper (Cu^2+^), and lead (Pb^2+^) ions, on cobalt removal efficiency using the CFILPD system. Experiments were conducted to assess cobalt removal when each metal was introduced individually and in combination, across varying concentrations. The removal efficiencies of zinc, copper, and lead ions were also quantified under these conditions. In addition, the recovered solid particles from CFILPD treatment were characterized to determine their crystallinity, elemental composition, and purity.

2. Results and Discussion

2.1. Effects of Zinc Ion Concentrations on Cobalt Removal Efficiency

The addition of different concentrations of zinc ions to the cobalt solution showed a significant impact on the removal efficiency of cobalt ions for the CFILPD process (Figure 1). When the zinc ion concentration increased, the efficiency of cobalt ion removal decreased. The cobalt ion removal was decreased by 53% at 30 min treatment after the addition of 100 ppm concentration of zinc ions to the cobalt solution, compared to that of zero addition of zinc. However, the addition of lower concentrations, such as 5 or 10 ppm of zinc ions, showed no significant effect on the cobalt ion removal process.

In addition to the cobalt ion removal process, the zinc ion removal efficiency by the CFILPD process was also determined. Compared to the cobalt ion removal, a significant removal efficiency of zinc ions was obtained when the initial concentration of zinc ions was 5 ppm, even with the very high concentration of cobalt ions in the solution. The highest removal rate of zinc ions achieved was 98.7% at 30 min treatment time. But with the increase in zinc ion concentration, the zinc ion removal efficiency decreased. When the concentrations of both metals were equal (Co^2+^:Zn^2+^ = 1:1), the removal efficiency of zinc ions was higher than that of cobalt ions after 30 min total treatment time. The reason for the faster removal of zinc ions over cobalt ions could be that zinc has lower hydration energy and forms more stable, insoluble precipitates, specifically zinc oxide (ZnO). While both ions are removed through similar precipitation mechanisms, differences in their chemical properties cause zinc to react and settle out of the solution more readily. A study reported that a deposit obtained after treatment of zinc ions by oxygen plasma contained oxides of zinc, ZnO, and Zn(OH)2, and ZnO tended to deposit with lower pH, while Zn(OH)2 deposited with higher pH [23]. Another study also confirmed the production of Zn(OH)2 and ZnO from Zn^2+^ ions in the solution with the use of a plasma-solution system [24]. In this work, the generated reactive species, such as hydroxyl ions/radicals from water dissociation, reacted with zinc ions and produced zinc oxide and hydroxide deposits according to the following reactions:

$$[eqn]$$ $$[eqn]$$ $$[eqn]$$ $$[eqn]$$ $$[eqn]$$

2.2. Effects of Copper Ion Concentrations on Cobalt Removal Efficiency

Figure 2 displays the removal efficiencies of cobalt and copper ions in the co-contaminated solution of 100 ppm of cobalt with different copper ion concentrations. In comparison with zero addition of copper, the removal efficiency of cobalt ions was decreased with the increase of copper ion concentrations in the solution. The lowest removal efficiency (13.8%) after 30 min of CFILPD treatment was obtained after the addition of 100 ppm copper ions. When compared with the effect after the addition of the same concentration of zinc to the cobalt solution, the effect of copper addition was significant on the cobalt removal efficiency. One past study also found that cobalt removal efficiency was decreased in the presence of copper ions in high concentrations compared to single-contaminated solutions of cobalt and copper. The reasons for this could be the competition between the cobalt and copper ions to interact with hydroxide ions generated by plasma discharge, affecting precipitation efficiency [25]. Although copper ion was the most dominant species at pH 6, its relative dominance decreased quickly as pH increased [26]. Given the solution pH range of 6 to 7 during the treatment time in this study, the copper ions tended to form Cu(OH)2 by increasing their removal efficiency:

$$[eqn]$$

Similarly, Collins and Kinsela [27] described the formation of metal complexes of cobalt and copper ions based on their characteristic rate constant of water exchange (k_M-H2O_) for a metal ion, which explained the phenomenon occurring for both cobalt and copper precipitants in the system, with Cu^2+^ (k_M-H2O_  =  1  ×  10^9^ s^− 1^) forming precipitates faster than Co^2+^ (k_M-H2O_  =  2  ×  10^6^ s^− 1^).

In contrast to these results, some studies reported that the addition of copper ions promoted the removal of efficiency of pollutants. For instance, Wang et al. [28] found that the addition of copper ions enhanced the polyvinyl-alcohol degradation efficiency in wastewater by hybrid gas–liquid pulse discharge plasma due to more OH^•^ radical formation from H_2_O_2_ catalytic decomposition by copper ions.

However, with the addition of lower concentrations of copper to the cobalt solution, Co:Cu ratios of 20:1 and 10:1 did not show a significant drop in cobalt removal efficiency. The removal efficiency of copper ions in the solution was increased when the initial copper ion concentration decreased despite the presence of a high concentration of cobalt ions in the solution. In this study, 100% removal efficiency of copper ions was achieved when the initial concentration of copper ions was 5 ppm. At the 100 ppm initial concentration of copper ions, the removal efficiency of copper after 30 min treatment time was 85%, and it increased to 95% and 98% when the initial concentration of copper ions in the solution decreased to 50 ppm and 10 ppm respectively. Similar results were found by Liu et al. [29], in which when the initial concentration of Cu-EDTA increased from 0.2 mM to 0.5 mM, the Cu-EDTA removal efficiency was reduced by 29.8% for the NTP process.

2.3. Effects of Lead Ion Concentrations on Cobalt Removal Efficiency

The addition of lead ions in different concentrations to the cobalt solution could hinder the cobalt ions’ removal efficiency for the CFILPD process. As shown in Figure 3a, the increased lead ion concentration (except 100 ppm) led to lower removal efficiencies of cobalt due to the competition between two metal ions to react with the produced reactive species. When the lead ion concentrations were lower, such as 5 ppm and 10 ppm, the competition was lower, and the cobalt ion removal process was not strongly inhibited by the lower lead ion concentration as compared to their high concentrations due to the lower total metal ion load. It was observed that the cobalt removal efficiency of 65% was obtained when the solution contained a 1:1 ratio of cobalt to lead ions in the solution.

Figure 3b represented the removal efficiency of lead ions by CFILPD with different initial concentrations of lead with 100 ppm cobalt ions in the solution. In general, the lead ion removal efficiency was greater than the cobalt removal efficiency except for the 50 ppm initial concentration of lead ions. Basically, lead (Pb^2+^) has a higher reduction potential (+0.13 V) than cobalt (Co^2+^, −0.28 V), and it is more likely to be reduced in the non-thermal plasma (NTP) treatment, while cobalt is more likely to be oxidized. This has several implications for their removal mechanisms involving high-energy species, including aqueous electrons, hydroxyl radicals, and atomic oxygen species produced by the CFILPD system, and these species drive oxidation and reduction reactions, influencing metal removal. For instance, reduction in Pb^2+^ to Pb (0) (metallic lead deposition) is possible due to the higher reduction potential of lead. And oxidation of Pb^2+^ to lead oxides (PbO, PbO_2_) or hydroxides (Pb(OH)2) only occurs when oxygen-based radicals dominate. Also, lead ion removal is enhanced by precipitation as Pb(OH)2 when plasma increases the solution’s pH [30].

According to Figure 3b, the lead ion removal rapidly increased for all the different initial concentrations to 10 min of treatment, and then decreased for the next 10 min and started to increase again in small percentages. A previous study reported a similar trend of lead ion removal by dielectric barrier discharge (DBD) plasma, and they achieved 13% removal of lead ions in the first 10 min. But from the 10 min to 30 min ozone irradiation period, the removal rate of lead ion decreased. From 60 to 120 min, the removal rate of the lead ion again increased and showed the same removal rate of deposits similar to the 10 min treatment, indicating that continuous changes in chemical reactions occurred during the plasma reaction [31]. Furthermore, lead ions could be reduced by aqueous electrons very rapidly (k = 3.9 × 10^10^ L mol^−1^ s^−1^) by plasma treatment. It was reported that hydrated electrons and 1-hydroxy alkyl radicals could reduce Pb^2+^ to form Pb^+^, which then underwent a disproportionation reaction with a rate constant of 1.7 × 10^8^ L mol^−1^ s^−1^ [32]. These disproportionation reactions of lead could be the reason for the changes in the removal efficiency of lead ions in the water by the CFILPD process:

$$[eqn]$$ $$[eqn]$$

2.4. Effect of Multiple Metal Ions Present in Water on Cobalt Removal Efficiency

The influence of various heavy metals, including zinc (Zn^2+^), copper (Cu^2+^), and lead (Pb^2+^), on cobalt (Co^2+^) ion removal was examined by introducing these metal ions into cobalt-containing solutions. Two different multi-metal solutions were prepared, with each metal ion initially present at either 25 ppm or 100 ppm concentrations. An initial concentration of 25 ppm was chosen to ensure comparable solution conductivity to that of a 100 ppm cobalt ion solution when individual ions were present. These solutions were subjected to treatment using the CFILPD system under optimized operational conditions, i.e., 200 W input power, 0.2 L/min argon gas flow rate, and 50 mL/min liquid flow rate. Metal ion concentrations were measured after 30 min of plasma treatment.

As illustrated in Figure 4, the removal efficiencies of each metal ion were evaluated for both concentration levels. All in all, the removal efficiency declined as the initial metal ion concentration increased. Specifically, cobalt ions exhibited the lowest removal efficiency among the tested metal ions, achieving only 8% removal at the higher concentration (100 ppm per metal ion). In contrast, cobalt ion removal improved significantly, reaching up to 70% in the solution containing lower initial concentrations (25 ppm each).

These results suggested that the presence of multiple heavy metals at high concentrations significantly hindered cobalt ion removal. This reduction could be attributed primarily to competition among metal ions for reactive species generated during plasma treatment, which were essential for redox reactions. Additionally, the elevated electrical conductivity of the multi-metal solution likely contributed to plasma process interference. The initial conductivity of the solution with all selected metals was measured at 951 µS/cm, which had a two times higher conductivity compared to the solution that had only 100 ppm cobalt, and its initial conductivity was 500 µS/cm. Solution conductivity strongly influences plasma ignition, discharge stability, and reactive species production in the CFILPD system. Higher conductivity can lower breakdown voltage but may also reduce localized electric field strength due to increased bulk current flow, thereby weakening plasma intensity and reactive species generation. These changes directly impact metal removal efficiency by altering electron-driven reactions and plasma–liquid interfacial processes.

High conductivity has been shown to adversely affect plasma characteristics, altering discharge behavior and reducing the energy available for effective cobalt oxidation or removal. For example, one study observed decreased dye removal efficiency with increasing the solution conductivity during DBD plasma treatment [33]. It was noted that higher conductivity led to spark discharge and increasing plasma intensity and discharge current, particularly in argon-based systems. Another study [34] reported that elevated sample conductivity reduced reactive species formation, correlating with a rise in discharge current and plasma intensity. These findings support the conclusion that reduced reactive species formation, caused by increased conductivity, may account for the diminished cobalt removal observed in this study.

In general, it was seen in this study that the presence of multiple metals in a single solution negatively impacted their individual removal efficiency, especially at higher concentrations. For instance, after 30 min of treatment, the copper removal efficiency was 94% and 43% at 25 ppm and 100 ppm, respectively. Similarly, zinc showed 93% removal at 25 ppm but dropped to 25% at 100 ppm. For lead, the maximum removal at 25 ppm was 9% after 5 min, followed by a decline likely due to particle deformation over time. At 100 ppm, the lead ion removal rate peaked at 75% after 10 min, then gradually decreased with continued treatment.

These results highlight the critical influence of initial metal concentrations on removal efficiency during the CFILPD treatment. The presence of multiple heavy metals at high concentrations not only introduces competitive reactions but also alters solution conductivity, both of which impair plasma performance and the overall remediation effectiveness.

The first-order rate constant for cobalt removal using the CFILPD system with argon as the working gas was determined to be 0.0964 min^−1^ [15]. However, the presence of co-existing metal ions significantly inhibited cobalt removal kinetics (Figure 5). Upon addition of 100 ppm of individual metal ions, the rate constant decreased to 0.0237 min^−1^, 0.0057 min^−1^, and 0.0380 min^−1^ in the presence of Zn^2+^, Cu^2+^, and Pb^2+^, respectively. When all three metal ions were simultaneously present at 100 ppm each, the rate constant further decreased to 0.0043 min^−1^. These results indicate that competitive interactions and/or scavenging effects associated with co-existing metal ions substantially suppress cobalt removal efficiency in the CFILPD system.

2.5. Comparison of Power and Energy Efficiencies for Cobalt Removal Process with Co-Occurring Metals

The evaluation of energy efficiency for cobalt removal using the CFILPD process is crucial, as it determines the feasibility, cost-effectiveness, and sustainability of water treatment applications. The concentration of each co-existing metal ion in solution was maintained at 100 ppm. The discharge power for the cobalt removal process in the presence of co-occurring metals was calculated using the following equation:

$$[eqn]$$

where P is discharge power (W), V is voltage (V), and I is current (A).

Figure 6 presents the voltage and current profiles recorded during plasma discharge for cobalt treatment in the presence of different metal ions. The temporal variations in voltage and current were similar when individual metal ions were added to the cobalt solution. However, when all metals were introduced simultaneously, the peak-to-peak voltage and current were significantly reduced, resulting in a lower plasma discharge intensity.

The discharge powers for cobalt removal without additional metals and in the presence of 100 ppm Zn^2+^, Cu^2+^, Pb^2+^, and the mixture of all three metals were 38.5%, 35.5%, 35.5%, 38%, and 32.5%, respectively, relative to the applied input power of 200 W. The reduced effective discharge power indicates energy losses within the system. These losses may arise from partial dissipation of input energy as heat due to electrode resistance (Ohmic heating) and other inefficiencies in energy transfer to the plasma channel.

Although the addition of individual metal ions did not cause a substantial change in discharge power, the simultaneous presence of all three metals resulted in an approximately 6% reduction in discharge power. This effect is attributed to the increased conductivity of the solution. The initial conductivity of the solution containing cobalt and all three additional metals was approximately 1 mS/cm, which is nearly twice that of the solution containing only 100 ppm cobalt (≈500 µS/cm). The higher solution conductivity can promote a transition toward unstable or arc-type discharge behavior, thereby reducing plasma stability and effective discharge power. It has been reported that elevated water conductivity can hinder stable plasma formation because dielectric breakdown in highly conductive water becomes more difficult [35].

The energy efficiencies for the cobalt removal process in each case can be calculated using the following equation:

$$[eqn]$$

where C_0_ is the initial cobalt concentration (g/L), V is the solution volume (L), ƞ is the removal efficiency (%), P is the applied power (kW), and t is the treatment time (h).

Based on the cobalt removal efficiencies obtained after 20 min of treatment, the corresponding energy efficiencies were calculated. The energy efficiencies for cobalt removal without additional metals and in the presence of Zn^2+^, Cu^2+^, Pb^2+^, and a mixture of all three metals were 0.38, 0.20, 0.05, 0.29, and 0.04 g kW^−1^ h^−1^, respectively.

The results clearly demonstrate that the energy efficiency of the cobalt removal process decreased significantly in the presence of competing metal ions. The coexistence of high concentrations of additional metals not only reduced cobalt removal efficiency but also substantially lowered the overall energy efficiency of the CFILPD system.

2.6. Characterization of Particles Recovered from the Solution of Cobalt with Co-Occurring Metals

The particles recovered from the cobalt removal process via CFILPD in the presence of additional metal ions such as zinc, copper, and lead ions (metal:cobalt = 1:1) were analyzed to determine their characteristics and composition. According to Figure 7, XRD analysis confirmed that the recovered particles contained metal oxides corresponding to the metal ions present in each solution.

In the cobalt–zinc ions mixture, the resulting particles were composed of cobalt oxide (Co_3_O_4_) and zinc oxide (ZnO). The diffraction peaks at 29.47°, 31.50°, 35.47°, 36.50°, 43.30°, and 47.20° indicated the coexistence of Co_3_O_4_ and ZnO with noticeable changes in peak intensities and slight shifts in peak positions due to the interactions between the two oxides [36]. Similarly, the mixture containing cobalt and copper ions produced particles identified as a combination of Co_3_O_4_ and CuO, as indicated by the peaks at 29.47°, 35.47°, 38.84°, 43.30°, and 48.65° in the XRD spectrum [37]. For the lead and cobalt-containing samples, the presence of lead oxide (PbO) along with cobalt oxide was confirmed by the peaks at 28.62°, 29.40°, and 35.99°. Notably, these peaks showed slight shifts compared to those of pure PbO, which had characteristic peaks at 28.6°, 31.82°, and 35.72°, corresponding to the (111), (020), and (002) planes, respectively [38]. Overall, the XRD results for the particles from the metal mixture by the CFILPD process revealed the peaks corresponding to Co_3_O_4_, ZnO, CuO, and PbO, confirming that the recovered particles were composed of a mixture of metal oxides.

The elemental composition of the recovered particles is summarized in Table 1, highlighting the major elements present in the compound. Additionally, trace amounts of Cl, Ca, and K were detected, likely originating from tap water. In Table 1, atomic percentage represents the proportion of a specific element relative to the total number of atoms in the sample, while the compound percentage refers to the mass or weight of an element relative to the total mass or weight of the sample. Overall, the atomic percentage of oxygen was higher than that of individual metals, as both metals primarily formed oxides. Additionally, the compound percentage of each metal was lower in particles recovered from the mixed-metal solutions compared to those obtained from individual metal–cobalt solutions. This reduction was likely due to the lower removal efficiency of each metal in the mixed-metal system.

Figure 8 presents the SEM images of recovered particles post-calcination. Across all samples, the particles exhibited irregular shapes and agglomeration, a morphology influenced by the plasma environment during the CFILPD process. The rapid plasma-induced nucleation could lead to uneven and uncontrolled particle growth, while the shorter reaction time caused particles to aggregate before they could fully crystallize into uniform structures. This agglomeration could be mitigated by introducing surfactants or a dispersing agent, which helps prevent clustering [39]. In addition, the SEM images of metal oxides revealed a wide range of particle sizes, from fine nanoscale aggregates to larger micro-sized structures, suggesting heterogeneous nucleation and growth mechanisms during formation.

Figure 9 presents the particle distribution of samples recovered from individual metal–cobalt solutions and mixed-metal solutions. The average particle sizes of the Co-Zn oxide, Co-Cu oxide, Co-Pb oxide, and mixed-metal oxide were 2.98, 2.60, 2.99, and 3.0 µm, respectively. The similarity in particle size across samples suggested that the consistent plasma conditions used for each solution resulted in the formation of particles with comparable dimensions. Furthermore, some particles appeared rough and porous, which might enhance their reactivity and surface area, which is potentially beneficial for applications such as catalysis or adsorption. For instance, a mixture of cobalt and zinc oxide was reported to be used in various applications, including catalysis, sensing, photodynamic therapy, antibacterial agents, and UV protection, due to their unique properties [40,41]. Similarly, cobalt and copper oxide mixed particles have potential applications in gas sensing, catalysis, energy storage, and photocatalysis due to their unique properties and synergistic effects when combined.

This study has several limitations that should be considered. Experiments were conducted using synthetic aqueous solutions, which do not fully replicate the complexity of real wastewater, including the presence of natural organic matter, variable ionic strength, and diverse co-existing ions. Detailed plasma diagnostics, such as measurements of electron density, temperature, and radical fluxes, were not performed, limiting quantitative understanding of reactive species generation and their direct role in metal removal. Additionally, the CFILPD system was operated on a laboratory scale, and scale-up challenges, including energy efficiency, electrode stability, and flow optimization, remain to be addressed. Future work will focus on testing the CFILPD system with real industrial or municipal wastewater, integrating plasma diagnostics to better understand reactive species dynamics, and optimizing reactor design for higher throughput and energy efficiency. Addressing these aspects will be critical for translating CFILPD into a practical, selective, and scalable technology for metal recovery and wastewater remediation.

3. Materials and Methods

3.1. Reagents and Material Preparation

The cobalt (II) concentration of a 100 mg/L solution was prepared by dissolving 504 mg of 98% purity of Co(NO_3_)2·6H_2_O (Thermo Fisher Scientific, Waltham, MA, USA) in tap water. Different concentrations of selected heavy metals were prepared by dissolving their nitrates, such as zinc nitrate hexahydrate, 99% (Thermo Scientific, Mumbai, India), copper (II) nitrate trihydrate, 99% (Thermo Scientific, Bourgoin-Jallieu, France), and lead (II) nitrate, 99% (Sigma Aldric, Bengaluru, India) in the cobalt solution. The tap water was used to maintain the conductivity level, which was required to obtain a proper and stable discharge in the plasma.

3.2. Experimental Setup

The setup of a CFILPD reactor was illustrated in Figure 10 [15]. The reactor consisted of a polycarbonate housing divided into three compartments, with a central chamber containing the high-voltage electrode and two outer chambers equipped with ground electrodes. Hollow stainless-steel ring electrodes (12.7 mm outer diameter, 6.05 mm wall thickness) were installed within each compartment of the polycarbonate body. Quartz dielectric plates with a thickness of 3.21 mm were placed at the interfaces separating the central chamber from the two ground electrode chambers. Each quartz plate had a precision-machined orifice with a diameter of 1 mm. Upon application of high voltage, plasma discharge was initiated within these narrow channels, resulting in two localized discharge regions at the interfaces between the high-voltage chamber and each grounded chamber.

Each experiment was conducted using 300 mL of metal-contaminated water, which was circulated through the CFILPD reactor at a flow rate of 50 mL/min from bottom to top using a motor pump (Cole-Parmer, Model 7553-70, Vernon Hills, IL, USA). The solution was recirculated back into the flask, enabling continuous treatment within the discharge zone. Plasma treatment was initiated using high-purity argon gas (99.99%), introduced into the discharge zone via a Venturi injector at a 0.2 L/min flow rate, regulated by a mass flow controller (C100L, Sierra Instruments, Monterey, CA, USA). The high-voltage electrode was powered by a Variac transformer (VEVOR, TDGC2-2KM, Shanghai, China) operating at 60 Hz, with 200 W applied power. Plasma discharge was generated at the aperture, producing energetic electrons and reactive species within the plasma zone. As the solution continuously flowed through this region, metal ions reacted with the plasma-generated species, resulting in the formation of metal oxide particles. The reactive species generated in the plasma zones using argon gas were confirmed by optical emission spectroscopy, as reported previously. It was confirmed that hydroxyl radicals, ions, and oxygen reactive species were dominant in the plasma zone and played a major role in the oxidation of cobalt ions [15].

Each treatment lasted 30 min. Voltage and current across the electrodes were monitored using a high-voltage probe (Tektronix P6015A, Beaverton, OR, USA) and a current probe (Tektronix P6021, Beaverton, OR, USA) connected to a digital storage oscilloscope (Tektronix TBS1052B, Beaverton, OR, USA).

3.3. Experimental Design

In order to determine the effects of different heavy metals on cobalt removal by the CFILPD process, three different heavy metal ions of zinc, copper, and lead were selected. First, the selected metal ions were individually treated with 100 ppm of cobalt solution, and then a mixture was prepared by adding all the selected metal ions to the cobalt solution. To evaluate the effects of different concentrations of each foreign metal ion on cobalt ion removal, various concentrations, such as 100 ppm, 50 ppm, 10 ppm, and 5 ppm of metal, were prepared individually with 100 ppm of cobalt ion solution. All the solutions were treated by the CFILPD process at 200 W applied power, 0.2 L/min argon gas flow rate, 50 mL/min liquid flow rate, and 60 Hz frequency. The volume and treatment time for all treatments were 300 mL and 30 min. The samples were collected at 5 min intervals for each treatment.

The mixture of all the selected heavy metals was added to the cobalt solution, and the individual concentration of each metal was 100 ppm. The above-mentioned treatment conditions were used to treat the mixture with the CFILPD system, and the removal efficiency of each element was evaluated to compare the effects of metals on cobalt removal. The pH and conductivity of each sample were recorded using a pH and conductivity meter (PC850 Portable pH/Conductivity Meter, Apera Instruments, Columbus, OH, USA). All the experiments were replicated.

3.4. Sample Analysis and Data Processing

To determine the metal concentration after the treatment, the collected samples during the CFILPD process were first centrifuged at 3500 rpm for 10 min and then filtered through 0.45 μm polyethersulfone (PES) membrane filters. The filtered samples were acidified by adding 2–3 drops of trace metal-grade nitric acid (67–70%, Fisher chemical), and Co, Zn, Cu, and Pb ion concentrations in the samples were determined using ICP-OES (Agilent 5110, Agilent Technologies, Inc., Santa Clara, CA, USA).

After the plasma treatments of solutions having a 1:1 ratio of metals, the solution containing all the metals together was left undisturbed overnight to settle the generated particles. Then, the supernatant was removed and dried in the oven at 80 °C for 5 h to remove the water content. Next, the dried particles were ground using mortar and pestle to make a homogeneous sample, and then calcined using a furnace at 550 °C for 5 h before being characterized using X-ray diffraction (XRD, Siemens D5000, Princeton, NJ, USA), scanning electron microscopy (SEM, model#: Zeiss Supra 35, Zeiss, Oberkochen, Germany), and energy-dispersive spectroscopy (EDS) (X-max 80T, Oxford, UK) techniques to identify their surface morphology and chemical composition in each particle with different metal oxides. Particle size distribution of each metal oxide was analyzed using ImageJ software (Version 1.54g).

4. Conclusions

The removal of cobalt ions from wastewater using the CFILPD process was found to be significantly influenced by the presence of co-occurring metal ions. In this study, zinc (Zn), copper (Cu), and lead (Pb), commonly found alongside cobalt ions in industrial wastewater, were selected to investigate their effect on cobalt removal efficiency under varying concentrations. Overall, the presence of these foreign metals in the cobalt solution hindered the oxidation of cobalt, resulting in a reduction in its removal efficiency. The extent of inhibition increased with the concentration of the interfering metal. When cobalt and copper were present in a 1:1 ratio, a drop of cobalt removal efficiency by approximately 85% after 30 min of treatment was observed, compared to the control without foreign metals. Similarly, reductions of 53%, 52%, and 91% were observed in the presence of zinc, lead, and a mixture of all three metals, respectively. Among the individual metals tested, copper exhibited the most pronounced negative impact on cobalt removal. The reduction in efficiency may be attributed to competition among metal ions for reactive species, differences in oxidation potentials, variations in solution pH, and other physicochemical properties of the metals.

Despite their inhibitory effects on cobalt removal, the CFILPD process demonstrated high removal efficiencies (>90%) for zinc, copper, and lead when they were present at relatively low initial concentrations (e.g., 5–10 ppm) alongside 100 ppm of cobalt. Particle characterization revealed that the recovered solid particles after treatment were composed of mixed-metal oxides, including cobalt oxide. These particles were generally agglomerated with irregular morphologies, and their average particle size ranged from 2.5 to 3 µm, regardless of the specific foreign metal ion present. Among the mixed-metal oxides, cobalt–zinc oxide and cobalt–copper oxide particles are of particular interest due to their relevance in a variety of industrial and technological applications.