A Green Electroslag Technology for Cadmium Recovery from Spent Ni-Cd Batteries Under Protective Flux with Electromagnetic Stirring by Electrovortex Flows

Abstract

The recycling of nickel–cadmium batteries poses a significant environmental challenge due to cadmium’s high biotoxicity. This study proposes a green method for recovering cadmium from cadmium oxide (CdO) using carbon (coal) in the presence of a molten binary flux (KCl:NaCl = 0.507:0.493, melting point 667 °C). The flux’s relatively low density and conductivity enable cadmium reduction beneath and through the flux layer. Brown coal (5–25 mm) served as the reductant. The reduction of cadmium from cadmium oxide with carbon (brown coal) took place in the temperature range from 667 °C to 700 °C. To enhance the process, electrovortex flows (EVF) were employed—generated by the interaction between non-uniform AC electric currents and their self-induced magnetic fields resembling conditions in a fluidised bed reactor. The graphite crucible acted as both one of the electrodes, with a graphite rod as the second electrode. As Cd and CdO are denser than both the flux and coal, the reduction proceeded below the flux layer. The flux facilitated CdO transport to the reductant, speeding up the reaction. X-ray diffraction (XRD) and scanning electron microscopy (SEM) confirmed the formation of metallic cadmium beneath and within the flux layer. This method demonstrates the feasibility of flux-assisted cadmium recovery without prior mixing and offers a foundation for further optimisation of sustainable battery recycling.

1. Introduction

Modern technologies depend on critical raw materials such as cobalt, lithium, rare earth elements, and others. However, the supply of these materials is often concentrated in just a few countries, creating geopolitical risks for the rest of the world. For example, cobalt, lithium, and neodymium are essential for batteries and electronics, yet their mining is largely limited to regions like China, Russia, and the Democratic Republic of Congo [1]. Such concentration makes supply chains vulnerable to trade restrictions or political instability. In response, many nations are seeking to “urban mine” their own waste streams—recovering valuable metals from used products—to reduce dependence on imports and improve resource security [1].

Recycling not only enhances supply security but also offers significant energy and environmental benefits. Extracting metals from secondary sources (scrap and waste) generally consumes far less energy than producing them from virgin ore. For instance, recovering nickel and cadmium from spent Ni-Cd batteries requires 46% and 75% less energy, respectively, compared to mining and refining those metals from primary ores [2]. This means recycling can dramatically cut energy use and greenhouse emissions while also diverting hazardous materials from landfills. Indeed, the reuse and recycling of materials are core strategies of the circular economy—an economic model aimed at keeping resources in use for as long as possible. By treating waste as a resource, societies can reduce pollution and relieve pressure on natural mineral reserves [3]. In the face of rising raw material prices and geopolitical uncertainties, recycling is becoming an increasingly competitive and attractive alternative to traditional mining [1]. Overall, these factors have created a strong impetus for developing efficient recycling systems—effectively turning our cities and waste into urban mines for critical materials.

To meet this demand, researchers have developed novel techniques that go beyond conventional pyrometallurgical and hydrometallurgical methods, integrating mechanical, electrochemical, and cavitation-driven processes. These innovations not only increase metal recovery rates but also reduce the environmental footprint of recycling operations, making them more compatible with circular economy principles. For example, applying alternating current (AC) in hydrometallurgy has enabled efficient leaching of precious and base metals from e-waste, achieving over 80% Au and Cu recovery without the need for cyanide or intensive preprocessing [4,5,6]. To enhance such processes, mechanical pretreatment can be employed—Shishkin’s team developed a selective disintegration–milling method that yields metal-rich fractions (up to ~95% metal content) from printed circuit boards, greatly improving downstream recovery efficiency [7,8]. Beyond solid wastes, Shishkin et al. have also pioneered sustainable treatment of metal-bearing solutions: a cavitation–dispersion cementation method for industrial wastewater was shown to rapidly remove and recover copper using elemental iron, illustrating the breadth of innovative techniques for closing metal loops [9]. These advancements collectively demonstrate how targeted process innovations—from flux shielding to AC-stimulated leaching and hybrid mechanical–chemical approaches—can maximize metal yield while minimizing hazardous emissions.

Spent Ni-Cd batteries pose a complex recycling challenge due to their toxic cadmium content and multi-component makeup. Conventional recovery methods rely on high-temperature carbothermic reduction or vacuum distillation, which can capture ~99.9% of Cd but only by operating at ~800–900 °C. Such processes risk cadmium vapor emissions and incur high energy costs, underscoring the need for greener alternatives. Recent studies by Shishkin and colleagues have advanced innovative metal recovery strategies aligned with circular economy principles. For example, Blumbergs et al. demonstrated a molten-salt electroslag technique for CdO reduction that avoided Cd volatilization by using a protective chloride flux layer at only ~650 °C [1]. In a comprehensive review, ref. [2] highlighted that no existing Ni-Cd recycling technology fully meets environmental and economic requirements, reinforcing the importance of new approaches like flux-assisted electroslag remelting. Building on this foundation, the present work introduces a green electroslag technology for cadmium recovery from spent Ni-Cd batteries that integrates a low-melting KCl–NaCl flux and electromagnetic stirring by electrovortex flows. This approach seeks to achieve high-purity Cd metal extraction without external mixing or excessive heating, thereby transforming toxic battery waste into reusable material in an environmentally sound manner, in full alignment with circular economy objectives.

Recent advances in cadmium recovery from spent Ni-Cd batteries have focused on environmentally friendly and efficient processes, with most research emphasizing hydrometallurgical and electrochemical methods rather than electroslag technology under protective flux. Hydrometallurgical approaches, such as acid leaching [10,11] followed by electrowinning or solvent extraction [12,13], have demonstrated high cadmium recovery rates (up to 99.5%) and the ability to produce high-purity metals, while minimizing secondary pollution and enabling process scalability [14,15,16]. Electrochemical techniques, including galvanostatic electrodeposition and single-cell electroassisted leaching, offer effective cadmium separation and recovery, with process efficiency influenced by current density and solution composition [12,17,18]. Thermal and biohydrometallurgical methods have also been explored, with thermal separation processes enabling cadmium collection from flue gas and bioleaching achieving complete cadmium extraction using bacteria or ferric iron solutions [18,19,20]. While these methods are well-established, there is a notable gap in the literature regarding the application of electroslag technology—especially under protective flux and with electromagnetic stirring—for cadmium recovery from batteries. This suggests a promising area for future research, as electroslag processes could potentially offer advantages in metal separation and environmental control if adapted for battery recycling [2].

Table 1. Green technologies and methods for cadmium recovery from spent Ni-Cd batteries.

Paper Title	Year	Summary	Ref.
A Sustainable Approach for Cadmium Recovery from Oxide Using Molten Salt Slag	2020	Describes a green method for cadmium recovery from CdO using a ternary chloride slag, minimizing Cd vapor release.	[4]
Cadmium Recovery from Spent Ni-Cd Batteries: A Brief Review	2021	Reviews recent progress and commercial methods for cadmium recovery from spent Ni-Cd batteries.	[2]
Electrochemical recovery of cadmium from spent Ni-Cd batteries	2005	Investigates cadmium recovery from spent Ni-Cd batteries using galvanostatic electrodeposition from acidic solutions.	[17]
Hydrometallurgical recovery of cadmium and nickel from spent Ni-Cd batteries	2007	Proposes a hydrometallurgical process for selective recovery of cadmium and nickel with high purity and efficiency.	[10]
Direct recovery of cadmium and nickel from Ni-Cd spent batteries by electroassisted leaching and electrodeposition in a single-cell process	2016	Develops a single-cell electrochemical process for direct recovery of cadmium and nickel from spent Ni-Cd batteries.	[11]

provides several works on green technologies and methods for cadmium recovery from spent Ni-Cd batteries, including approaches using molten salt slags and electrochemical processes. The literature misses the combination of “electroslag technology” + “protective flux” and “electromagnetic stirring by electrovortex flows”, which indicates the present research novelty; several papers address sustainable and innovative cadmium recovery methods.

Alkaline nickel–cadmium (Ni-Cd) batteries are commonly used as autonomous sources of industrial and domestic electric power due to their favourable combination of technical, economic, and electrical properties [21].

This paper explores the technical aspects of a future cadmium battery recycling technology. Currently, metallurgical processing can be performed in three different ways: pyrometallurgy, hydrometallurgy, or hybrid processes that combine pyro- and hydrometallurgy techniques for the production of metals or their compounds [22]. A comprehensive review of the cadmium battery market and existing recycling technologies is available in previous publications [2]. Current environmental and economic demands for battery disposal and recycling have also been discussed previously [23].

However, there is still no environmentally friendly and economically viable technology for processing spent batteries to produce quality materials. Thus, ongoing research in this area remains highly relevant.

The cumulative charge/discharge reactions in Ni-Cd batteries are described by the following Equation (1) [24]:

$${{\mathrm{C}\mathrm{d}\left(\mathrm{O}\mathrm{H}\right)}_2+2{\mathrm{N}\mathrm{i}\left(\mathrm{O}\mathrm{H}\right)}_{2 }}_{\overrightarrow{charge}}^{\stackrel{discharge}{\leftarrow }} \mathrm{C}\mathrm{d}+2\mathrm{N}\mathrm{i}\mathrm{O}\mathrm{O}\mathrm{H}+2{\mathrm{H}}_2\mathrm{O}$$

(1)

For Cd recovery from Cd(OH)₂, the proposed method uses thermal decomposition at 400 °C. A monoclinic γ-Cd(OH)₂ phase and a cubic CdO phase form at 300 °C. A pure cubic CdO phase is obtained at 400 °C. At 700 °C, both CdO and CdO₂ phases appear [25] Since cadmium evaporation becomes significant above 765 °C, the reduction process is limited to 700 °C to avoid CdO₂ formation and reduce Cd loss [26].

Cd(OH)₂ → CdO + H₂O

(2)

The overall reduction reaction was ensured by the reduction of Cd according to Volynsky et al. [18] by a cumulative reaction from 650–1100 °C (3):

2CdO + C → 2Cd + CO₂

(3)

Further reduction can also proceed through intermediate reactions (4) and (5):

CdO + C → Cd + CO

(4)

CdO + CO → Cd + CO₂

(5)

In this study, an excess molar ratio of CdO:C = 1:50, which significantly exceeds the minimum ratio of 1:0.5 required for reaction (3) to proceed, was used to ensure complete reduction via reaction (4) and partial conversion through reaction (5). The surplus carbon prevents the reoxidation of the produced Cd on the slag surface.

Carbon-based reductants such as anthracite [27] and carbon black [18] have been investigated in pyrometallurgical processes for Cd recovery.

Slags are proposed in Ni-MH battery recycling for extracting rare earth oxides [28,29]. In this case, as a result of high-temperature treatment, a Ni-Co alloy and a heterooxide material are obtained. In these processes, slag systems of the CaO-CaF₂, CaO-SiO₂-MgO, and SiO₂-Al₂O₃ types are used.

Electroslag remelting (ESR) is a secondary refining process where a consumable electrode is remelted under a protective slag. Slag functions include heating (via Joule heating), dissolving non-metallic inclusions, refining the metal, and shielding it from the atmosphere [30,31]. The metallothermic reduction of TiCl₄ by liquid Mg in an ESR setup using a non-consumable electrode was demonstrated by Platacis et al. [30].

To reduce energy and reagent consumption and to minimize or eliminate Cd evaporation, a sustainable production approach must be implemented.

Sustainable practices aim to reduce energy and reagent consumption and eliminate cadmium evaporation, as the experiments were conducted in a temperature range from 667 °C to 700 °C, which is below the boiling point of cadmium of 765 °C. The goal of this study is to avoid Cd loss from the reaction volume during high-temperature reduction by limiting the temperature of cadmium oxide reduction by carbon to the upper limit of the reduction reaction of 700 °C, the lower limit is determined by the flux liquidus temperature of 667 °C. A flux was selected with a melting point below the upper temperature limit of 700 °C to create a liquid seal in the event of an emergency increase in temperature above the boiling point of cadmium of 765 °C. The flux is also selected based on the following conditions: the flux density in the liquid state should be less than the density of the carbon-containing reducing agent (brown coal, anthracite, metallurgical coke, petroleum coke), so that the reaction of cadmium oxide reduction by carbon occurs under a flux layer playing the role of a liquid seal, which prevents the cadmium reduction reaction from occurring on the surface and plays the role of a seal. The flux is selected to be electrically conductive in a liquid state, which will allow, in the presence of a graphite crucible and a graphite electrode, to use the effect of electro-vortex flows, which allows transporting cadmium oxide to carbon with a liquid conductive flux to carry out the reduction reaction and to combine microparticles of reduced cadmium into a single mass in the process of electrovortex flow. This method enables Cd pre-separation from spent batteries without heating the entire mass excessively. Cd has a melting point of 321 °C and a boiling point of 765 °C [26]. Reduction is carried out using crushed coal (5–25 mm) within molten chloride salts at temperatures below Cd’s boiling point and without inert gas protection. The flux is critical for implementing the electroslag process effectively.

This study qualitatively evaluates the influence of salt flux on the efficiency of CdO reduction in and beneath the flux layer, and during CdO transport. An alternating current heats the chloride flux mixture in a graphite crucible using a graphite electrode. Electrovortex flows stir the molten flux, CdO, and reduced Cd, enhancing interaction with the static crushed carbon bed.

These flows result from the interaction of a non-uniform electric current and its self-induced magnetic field [32]. Such flows are common in electroslag welding, arc remelting, and ESR processes and significantly influence product quality [33,34,35,36,37,38].

This study has proven the possibility of carrying out the reaction of cadmium reduction from cadmium oxide with carbon under a flux layer with the reduction reaction temperature limited to 700 °C, with the flux acting as a gate preventing the reduction reaction on the reactor surface and evaporation of cadmium from the surface, and also transporting cadmium oxide to carbon and combining microparticles of reduced cadmium into a single mass in the process of electrovortex flow. This study has opened up the possibility of industrial processing of cadmium-containing waste and storage batteries by electroslag reduction without the risk of contamination of the territory with cadmium evaporated during the reduction process, given its high biological toxicity.

2. Materials and Methods

2.1. Justification of Method of Conducting Experiments

To prevent Cd evaporation, a binary KCl-NaCl flux was chosen with a molar ratio of 0.507:0.493, corresponding to the eutectic point with the lowest melting temperature (667 °C) [39], as shown in the liquidus projection diagram of the binary KCl-NaCl (KN) shown in Figure 1. This approach prevents the evaporation of the formed Cd, which occurs at temperatures above 765 °C [26].

To create the slag coating, chemically pure (>99.0%) mixtures of reagents KCl (Firma CHEMPUR, Piekary Śląskie, Poland) and NaCl (Sigma-Aldrich, Merck KGaA, Darmstadt, Germany) were used.

The weight of all components was measured on a laboratory balance (KERN 440-35A, Balingen-Frommern, Germany, accuracy 0.01 g) and then ground using a kitchen mill (Clatronic KM 3350 Kitchen Machine, Clatronic International GmbH, Kempen, Germany) equipped with a plastic vessel for 3 min. The total weight of the mixture added to the mill was 100 g for each grinding process.

Figure 2 shows the dependence of the specific electrical conductivity of the melt of the binary system KCl-NaCl on the molar ratio of the components and the dependence on the temperature range. The specified temperature range corresponds to 800–900 °C and was chosen due to the melting point of KCl—776 °C and NaCl—801 °C, taking into account the component of the binary flux with a higher melting point. This does not refute the electrical conductivity of the binary system KCl-NaCl with the molar ratio of KCl-NaCl chosen by us in the temperature range from 667 °C to 700 °C, in which the experiments were conducted. The temperature of 700 °C was chosen as the upper limit.

Figure 3 shows the interaction of electromagnetic forces, which leads to the formation of electrovortex flows. The alternating current through the molten flux interacts with the alternating magnetic field, creating electromagnetic forces inside the melt. These forces have a constant and pulsating component. The constant component of the electromagnetic force is highly non-uniform due to the geometry of the device, which leads to rotation of the melt.

Numerical modelling of the electro-vortex flow was performed in the electrodynamic (non-inductive) approximation and was based on the solution of the Navier–Stokes Equation (6), which describes the hydrodynamics of the electro-vortex flow with the electromagnetic force Fci = J × B as a source as follows:

$$\mathrm{\rho }\left({\displaystyle \frac{\partial \mathrm{U}}{\partial \mathrm{t}}}+\left(\mathrm{U}\nabla \right)\mathrm{U}\right)=-\nabla \mathrm{p}+\mathrm{\rho }\mathrm{v}∆\mathrm{U}+{\mathrm{F}}_{\mathrm{e}\mathrm{l}}$$

(6)

where U—the velocity of the liquid flux, ρ—the flux density, ∇—the nabla operator, v—the kinematic viscosity coefficient, p—the pressure, and F_el—the electromagnetic force.

At the same time:

F = J × B

(7)

where J is the electric current density, and B is the magnetic field.

2.2. Sample Preparation and Experimental Parameters

CdO (>99% purity, 5–20 µm) (Sigma-Aldrich, Merck KGaA, Germany) was used as a Cd source. Brown coal—carbon (C) with a particle size of 5 mm to 25 mm was used as a reducing agent. The experiments were carried out in a 2 L graphite crucible with a 30 mm diameter graphite electrode (Figure 4a,b). Figure 5 shows the components of the two-component flux KCl-NaCl and the stainless steel crucible in which they were mixed before heating to 850 °C. Figure 4b shows a crucible filled with brown coal.

A graphite crucible was filled with 600 g of carbon (50 mol) and 128 g of CdO (1 mol) and heated to 300 °C using a gas burner. The molten flux (2 kg) was preheated to 850 °C and poured into the crucible, which was preheated to 300 °C and under alternating voltage (70 V, 19 A, 50 Hz). The electroslag process began immediately (Figure 6a). Figure 6b shows a crucible filled with flux KN, C, and Cd after electroslag remelting.

The reduction process continued for 60 min, followed by 24 h of cooling. Six experiments were performed.

Table 2. Value of experimental parameters.

Experimental Parameter	Value
Crucible temperature before pouring flux	300 °C
Temperature of pouring flux into crucible	850 °C
KN flux weight	2000 g
Weight of brown coal fraction 5–25 mm	600 g
Weight of cadmium oxide	128 g
Electric current, current strength	19 А
Electric current, voltage	70 V
Frequency of electric current	50 Hz

provides value of experimental parameters.

2.3. Sample Extraction and Preparation for Analysis

After 24 h of cooling, the crucible contents were divided into three zones: lower (carbon and flux), middle (flux), and upper (flux with traces of carbon). The samples were leached with distilled water, filtered, dried for 3 h at 120 °C, and ground. The powder was fractionated using a vibrating sieve shaker (Retsch AS 200 digit, Retsch GmbH, Haan, Germany) equipped with sieves with mesh sizes of 0.1 mm, 0.15 mm, 0.2 mm, and 0.25 mm.

2.4. Microscopy and XRD Analysis

Surface morphology and elemental composition were examined by SEM (SEM, Tescan Lira, TESCAN GmbH, Dortmund, Germany).

Optical images were taken using a VHX-2000 microscope (VHX-2000, Keyence Corporation, Osaka, Japan) with VH-Z20R/W objectives.

X-ray diffraction (XRD) measurements were carried out using a MiniFlex 600 RIGAKU X-ray diffractometer (Rigaku Corporation, Tokyo, Japan), operating at a voltage of 40 kV and a current of 15 mA, using Cu Kα radiation. The international Centre of Diffraction Data (ICDD) database was used for comparison of the results (Cd: PDF 00-005-0674; CdO: PDF 00-005-0640; CdO2: PDF 01-073-6494; Carbon: PDF 00-006-0675; Graphite: 00-056-0159). The measurements were performed in the 5–90° range with a step size of 5°.

3. Results

3.1. Morphological Analysis

All six experiments showed a consistent three-layer structure. The top contained slag with carbon inclusions, the middle had clean flux, and the bottom had carbon and flux. Optical microscopy confirmed these layers. After 1 h of electroslag reduction, all six experiments consistently exhibited the same structure. Three distinct zones were observed pronounced three zones of the electroslag process (Figure 7):

• The top zone (1) consisted of a slag crust with carbon inclusions (Figure 8a);
• The middle zone (2) contained only flux (Figure 8b);
• The bottom zone (3) contained both flux and coal (Figure 8c).

All samples were sectioned and leached separately with distilled water to remove water-soluble components KCl and NaCl, following the procedure outlined in the Section 2.

Figure 9 shows representative optical images of the upper (Figure 9a), middle (Figure 9b), and lower (Figure 9c) sections from experiment CD6 at 50× magnification.

3.2. SEM and XRD Analysis

3.2.1. Results of the Study of the Bottom Layer

Figure 10 and Figure 11 present the SEM and XRD analysis of the bottom layer (sample CD5-3 and CD6-3, respectively). The presence of metallic cadmium (Cd), CdO, and CdO₂ was confirmed. Total cadmium content reached 43.975%. The presence of CdO₂ indicates that atomic oxygen in the melt reacted with metallic Cd at temperatures exceeding 700 °C.

Carbon was found in very high concentrations (49.120%). The presence of a two-component potassium–sodium flux is indicated by the percentage of chlorine (1.375%) and potassium (1.404%). Figure 12 shows the spectra of cadmium (Cd), cadmium oxide (CdO), and cadmium oxide (CdO₂), as well as the spectrum of sodium, but in an amount that is not detectable by SEM.

Table 3. SEM analysis results of the bottom layer, sample CD5-3.

Element	Extracted Spectrum
	Line Type	Intensity	Net Counts	Weight %	Atom %	Atom % Err	Norm. wt.%	Chemical Formula	Compound %	Norm. Compound %
C K	K	432.974	201…	49.120	87.667	0.381	49.120	C	49.120	49.120
Cd L	L	379.994	175…	43.975	8.387	0.038	43.975	Cd	43.975	43.975
Ca K	K	16.766	78,015	1.404	0.751	0.017	1.404	Ca	1.404	1.404
Cl K	K	24.303	113…	1.375	0.832	0.013	1.375	Cl	1.375	1.375
Fe K	K	6.668	31,026	1.181	0.453	0.009	1.181	Fe	1.181	1.181
Si K	K	16.076	74,802	0.998	0.762	0.022	0.998	Si	0.998	0.998
Al K	K	9.872	45,937	0.708	0.562	0.034	0.708	Al	0.708	0.708
Ti K	K	3.197	14,878	0.382	0.171	0.011	0.382	Ti	0.382	0.382
Cu K	K	1.172	5455	0.338	0.114	0.011	0.338	Cu	0.338	0.338
S K	K	5.442	25,228	0.268	0.179	0.012	0.268	S	0.268	0.268
Mn K	K	0.657	3059	0.112	0.044	0.008	0.112	Mn	0.112	0.112
Cr K	K	0.674	3135	0.094	0.039	0.006	0.094	Cr	0.094	0.094
Mg K	K	0.469	2183	0.045	0.040	0.32	0.045	Mg	0.045	0.045
				100.000	100.000		100.000		100.000	100.000

shows the results SEM analysis results of the bottom layer, sample CD5-3.

Figure 11 represents the chemical composition of the mixture (bottom layer) by the following components: Cd, CdO, CdO₂, NaCl, KCl, and C (graphite, carbon).

3.2.2. Results of the Middle Layer Study

Figure 12 and Figure 13 show the results of scanning electron microscopy (SEM) and X-ray diffraction (XRD) analysis of the middle layer (samples CD5-2 and CD6-2, respectively). Scanning electron microscopy (SEM) (Figure 12) confirmed the absence of metallic cadmium (Cd), cadmium oxide (CdO), and cadmium oxide (CdO₂). Carbon was detected in very low concentrations (0.045%). The presence of two-component potassium–sodium flux is indicated by the percentage of chlorine 54.338%, potassium 25.496%, and sodium 18.854%. However, Figure 13 shows spectra of cadmium (Cd), cadmium oxide (CdO), and cadmium oxide (CdO₂), but in amounts that are not detectable by SEM.

Figure 13 represents the chemical composition of the mixture (middle layer) by the following components: Cd, CdO, CdO₂, NaCl, KCl, and C (graphite, carbon).

Table 4. SEM analysis results of the middle layer, sample CD5-2.

Element	Extracted Spectrum
	Line Type	Intensity	Net Counts	Weight %	Atom %	Atom % Err	Norm. wt.%	Chemical Formula	Compound %	Norm. Compound %
Cl K	K	2051.682	774…	54.338	50.285	0.148	54.338	Cl	54.338	54.338
K K	K	630.639	237…	25.496	21.394	0.081	25.496	K	25.496	25.496
Na K	K	451.336	170…	18.854	26.907	0.109	18.854	Na	18.854	18.854
Al K	K	27.301	102…	0.840	1.022	0.013	0.840	Al	0.840	0.840
Fe K	K	2.649	9993	0.227	0.134	0.015	0.227	Fe	0.227	0.227
Ti K	K	3.421	12,907	0.198	0.136	0.010	0.198	Ti	0.198	0.198
C K	K	0.114	429	0.045	0.123	0.007	0.045	C	0.045	0.045
O K	K	0.000	0	0.000	0.000	-	0.000	O	0.000	0.000
				100.000	100.000		100.000		100.000	100.000

shows SEM analysis results of the middle layer, sample CD5-2.

3.2.3. Results of the Study of the Upper Layer

Figure 14 and Figure 15 show the results of the scanning electron microscopy (SEM) and X-ray diffraction (XRD) analysis of the top layer (samples CD5-1 and CD6-1, respectively). Scanning electron microscopy (SEM) (Figure 14) confirmed the absence of metallic cadmium (Cd), cadmium oxide (CdO), and cadmium oxide (CdO₂). Carbon was detected in very low concentrations (0.029%). The presence of two-component potassium–sodium flux is indicated by the percentage of chlorine 53.304%, potassium 26.454%, and sodium 19.433%. However, Figure 15 shows spectra of cadmium (Cd), cadmium oxide (CdO), and cadmium oxide (CdO₂), but in amounts that are not detectable by SEM.

Table 5. SEM analysis results of the upper layer, sample CD5-1.

Element	Extracted Spectrum
	Line Type	Intensity	Net Counts	Weight %	Atom %	Atom % Err	Norm. wt.%	Chemical Formula	Compound %	Norm. Compound %
Cl K	K	2331.917	965…	53.304	49.254	0.144	53.304	Cl	53.304	53.304
K K	K	761.487	315…	26.454	22.164	0.083	26.454	K	26.454	26.454
Na K	K	540.033	223…	19.433	27.691	0.111	19.433	Na	19.433	19.433
Al K	K	19.761	81,835	0.529	0.642	0.012	0.529	Al	0.529	0.529
Ti K	K	5.024	20,807	0.252	0.172	0.010	0.252	Ti	0.252	0.252
C K	K	0.085	350	0.029	0.078	0.005	0.029	C	0.029	0.029
				100.000	100.000		100.000		100.000	100.000

shows SEM analysis results of the upper layer, sample CD5-1.

Figure 15 represents the chemical composition of the mixture (upper layer) by the following components: Cd, CdO, CdO₂, NaCl, KCl, and C (graphite, carbon).

4. Discussion

The SEM and XRD analyses presented in Figure 10, Figure 11, Figure 12, Figure 13, Figure 14 and Figure 15 confirm the successful application of the electroslag reduction method for cadmium (Cd) recovery from cadmium oxide (CdO) using molten KCl-NaCl flux and brown coal as a reductant. In the bottom layer (Figure 10 and Figure 11), Figure 11a,c reveal the clear presence of metallic cadmium alongside residual CdO and CdO₂, as well as graphite and elemental carbon. The XRD pattern in 11a demonstrates strong peaks corresponding to elemental Cd (PDF 00-005-0674), confirming that the reduction proceeded efficiently in this zone. The high carbon content (49.12%) and Cd species (43.98%) support the conclusion that the reduction is most effective in the lower zone, where CdO interacts directly with carbon particles.

The detection of CdO₂, a higher-valent oxide (Fire 11a), indicates partial reoxidation of metallic Cd likely caused by atomic oxygen presence at elevated local temperatures above 700 °C. This suggests the need for further improvement of the process by using vacuum pre-treatment of the flux to degas it under an inert argon gas atmosphere, consistent with prior findings in electroslag systems for titanium recovery, where inert atmospheres minimized oxidation side reactions [30].

In the middle layer (Figure 12 and Figure 13), Figure 13a confirms the coexistence of Cd, CdO, and CdO₂, although with significantly lower carbon content (0.045 wt.%, see SEM map in Figure 12). XRD Figure 13b shows persistent flux peaks (NaCl and KCl), indicating that this zone is dominated by the molten salt phase with only minimal carbon ingress. These results suggest that reduction in the middle layer is diffusion-limited and driven by electrovortex transport of CdO particles from the top and carbon from the bottom. Similar effects were observed in zinc and lead slag cleaning systems, where convective mass transfer occurs [43].

In the upper layer (Figure 14 and Figure 15), the reduction is clearly least effective. SEM mapping (Figure 14) and XRD (Figure 15a) still indicate traces of metallic Cd and its oxides, but carbon presence is minimal (0.029 wt.%), as seen in Figure 15c. The persistence of Cd species in this zone—despite the lack of reducing agent—underscores the role of convection and electrohydrodynamic flow in redistributing reaction products upward. However, the yield is lower, likely due to reduced carbon contact. Our current study continues the logic of the studies on the application of electroslag processes in metal reduction, where the accumulation of the product depends on the flux density. This study is consistent with our previous studies and articles on the reduction of titanium from TiCl4, where electroslag remelting and electroslag separation were used. Briefly, the essence of the process: in previous studies on the electroslag reduction of titanium from titanium tetrachloride with magnesium, a liquid flux was selected with a density lower than the density of the titanium sponge obtained in the process of reducing titanium tetrachloride with magnesium and higher than the density of the molten magnesium reducing metal. The reduction reaction took place above and on the surface of the selected flux. The reduced titanium sank in the molten flux and collected at the bottom of the reactor [30].

Importantly, these findings validate the key hypothesis of this study: that effective Cd reduction can be achieved without pre-mixing carbon and oxide, relying instead on density separation and electromagnetic flow. Yet, the incomplete reduction in middle and upper zones indicates room for process optimization. Enhancements may include the use of denser or more reactive carbon (e.g., metallurgical coke), finer control over flux composition, and improved thermal management to avoid Cd reoxidation.

The use of simple natural components (brown coal—carbon) as a reducing agent in electroslag reduction does not require additional processes, unlike hydrometallurgy, where the production of reagents necessary for the hydrometallurgical process creates a large carbon footprint. Unlike the pyrometallurgical method, the electroslag reduction process has a lower temperature of the reduction process, and does not require energy-intensive vacuuming for the precipitation of reduced cadmium, which creates a significant carbon footprint.

In conclusion, this study contributes to the broader field of green metallurgy by demonstrating a viable process for cadmium recovery that avoids cadmium volatilization, minimizes environmental impact, and offers potential scalability—comparable to molten-salt-based recovery systems for Zn, Pb, and REEs cited in the literature (e.g., Binnemans et al., Journal of Sustainable Metallurgy, 2015) [44].

5. Conclusions

This study demonstrates the viability of a green, electroslag-based method for the reduction of cadmium oxide (CdO) to metallic cadmium (Cd) using a molten KCl-NaCl flux and carbon reductant in a graphite crucible. The process leverages electrovortex-driven electromagnetic stirring to enable efficient transport of reactants within the molten flux, thereby eliminating the need for prior mechanical mixing of CdO and carbon. The process of cadmium reduction from cadmium oxide with carbon was evidenced by the bubbling of the flux around the graphite electrode, which indicates the release of CO₂ formed during the chemical reaction. The eutectic composition of KCl:NaCl (0.507:0.493 molar ratio) ensured a low melting point (667 °C) and high electrical conductivity, supporting stable operation below cadmium’s boiling point and minimizing Cd vaporization.

SEM and XRD analyses confirmed that the majority of metallic Cd formed in the lower part of the crucible, where direct contact between CdO and carbon was maintained. Partial reduction also occurred in the middle and upper zones, facilitated by convective flows, although with reduced efficiency due to limited carbon availability. The presence of CdO₂ in some regions suggests local oxidative conditions, emphasizing the need for inert gas protection or vacuum operation to fully suppress reoxidation.

To avoid CdO₂ formation in future experiments, the flux melting process will be conducted under vacuum followed by argon backfilling. The carbon content of 49.120% supports the assumption that coal, being slightly denser than the molten KCl-NaCl flux, remains at the crucible’s bottom and is not carried upwards by electrovortex flows.

The main conclusion: reduction of CdO does not require prior mixing with carbon, as confirmed by SEM and XRD data.

The process showed promising performance within a short reaction time (60 min), validating its potential as a scalable and environmentally safer alternative to conventional pyrometallurgical methods. Future optimization will focus on increasing Cd yield through improved flux control, carbon type selection, and enhanced atmosphere management. The findings position this method as a valuable addition to the portfolio of sustainable technologies for the recycling of toxic cadmium-containing waste, aligning with circular economy principles and current environmental directives.