Ex Situ Sediment Remediation Using the Electrokinetic (EK) Two-Anode Technique (TAT) Supported by Mathematical Modeling

Abstract

Heavy metals are non-biodegradable environmental pollutants, and if present in sludge/sediment in elevated concentrations, they can cause serious problems. In this paper, the possibility of applying two-anode electrokinetic treatment was investigated for the ex situ remediation of copper (Cu) and nickel (Ni)-contaminated sediments. The influence of the following parameters on the treatment efficiency was investigated: applied electric field, physicochemical changes in the system, and the characteristics of the pollution (concentration and forms of metal occurrence). Additionally, based on the results of the sequential extraction procedure, a risk assessment of sediment before and after treatment was performed. Also, we developed a mathematical model that allows us to define the time required to reduce nickel and copper to non-hazardous levels from contaminated sediment via electrokinetic treatment. The results obtained indicate that changes in the pseudo-total content and changes in Cu and Ni availability along the electrokinetic cell are consistent with the physicochemical changes in the sediment. The amount of applied electric field does not notably affect the treatment efficiency in most cases. Based on the results, the majority of samples of treated sediment can be dislocated without special protection measures. The most acceptable treatment for ex situ remediation is the one with solar panels, as it is considered economically and environmentally most appropriate. For this treatment, according to risk assessment code, the risk was found to be low (Cu) to moderately low (Ni). Since more than 50% of Cu and Ni content is related to the organic and residual fraction, and based on the physicochemical conditions and high percentage of clay, we can assume that there are no environmental hazards. This work serves as a starting point for the developed mathematical model that has proven to be very promising for prediction of the time necessary for sediment metal remediation.

1. Introduction

In aquatic ecosystems, sediments act as long-term sinks for pollutants originating from industrial activities, urban runoff, municipal wastewater discharges, and the often unregulated use of agrochemicals on agricultural land near riverbanks [1]. Over time, these inputs result in the accumulation of heavy metals and other toxic compounds, particularly in industrialized and densely populated regions, posing significant environmental and ecological risks. Heavy metals such as copper (Cu) and nickel (Ni) are of particular concern due to their persistence, non-biodegradability, and tendency to accumulate in sediments and aquatic food webs. At elevated concentrations, Cu can disrupt enzymatic activity, impair photosynthesis in aquatic plants, and induce oxidative stress in invertebrates and fish, while Ni exposure has been associated with growth inhibition, reproductive toxicity, and genotoxic effects in aquatic organisms. Both metals can be remobilized from contaminated sediments under changing redox and pH conditions, increasing their bioavailability and facilitating transfer across trophic levels. Chronic exposure to Cu- and Ni-contaminated sediments therefore poses risks not only to benthic ecosystems but also to terrestrial wildlife and human health through bioaccumulation and biomagnification in the food chain [2,3,4,5].

Dredging is widely applied as a conventional response to manage contaminated sediments and maintain waterway functionality. However, dredging does not eliminate contamination; instead, it transfers it from the aquatic environment to land. The dredged material is transformed into sludge—a semi-solid matrix with higher water content, increased contaminant concentration, and enhanced pollutant mobility—requiring additional treatment prior to disposal or reuse.

A variety of remediation technologies have been explored for contaminated sediments and dredged sludge, including physical, thermochemical, biological, and combined approaches. Physical and thermochemical methods, such as thermal treatment and microwave-assisted processes, can effectively reduce contaminant concentrations but are often limited by high energy demand, elevated operational costs, and the risk of secondary emissions. Biological methods, including aerobic and anaerobic digestion, offer environmentally benign alternatives; however, their effectiveness is frequently constrained by long treatment times, sensitivity to toxic metals, and limited removal efficiency for inorganic contaminants. Chemical stabilization and washing techniques may achieve rapid metal immobilization or extraction but often generate secondary waste streams and require substantial reagent consumption [6]. Consequently, these limitations underscore the need for alternative remediation strategies that are both effective and economically viable for fine-grained, heterogeneous sediment matrices.

Electrokinetic (EK) remediation has been recognized as an effective ex situ technique for treating metal-contaminated sediments and sludge matrices [7,8,9]. The process involves the application of a direct-current (DC) electric field, inducing contaminant transport through electroosmosis, electromigration, and electrophoresis, coupled with electrochemical and geochemical reactions [10,11,12,13,14,15]. Despite its advantages, EK remediation is constrained by two major bottlenecks. The first is cathodic alkalization resulting from water electrolysis, which leads to the precipitation of metal hydroxides, carbonates, and other secondary phases near the cathode. This accumulation hinders metal electromigration and represents one of the primary limitations of EK treatment efficiency [11,16,17,18]. To overcome this limitation, the two-anode technique (TAT) has been proposed, in which a secondary anode placed near the cathode generates H⁺ ions that migrate toward the cathodic region, stabilizing the pH profile, preventing premature metal precipitation, and enhancing metal desorption and transport without the addition of chemical reagents [19]. The second major limitation of EK remediation is its electricity demand, which typically accounts for 10–15% of total operational costs [7,20,21]. This energy requirement restricts large-scale and field applications, particularly in remote areas with limited grid access. The integration of renewable energy sources, especially solar power, has therefore attracted increasing attention due to its economic and environmental advantages. Previous studies have demonstrated that EK remediation powered by solar energy or operated under intermittent DC supply conditions does not significantly compromise contaminant removal efficiency, despite fluctuations caused by weather conditions and diurnal cycles [7,21,22,23]. Consequently, combining the TAT configuration with solar-powered operation represents a promising strategy to simultaneously address the key technical and economic limitations of EK remediation.

Applying mathematical models could become a very useful tool in predicting the results of remediation [13,24]. Mathematical modeling has become an essential tool for understanding and optimizing electrokinetic (EK) remediation processes, as it enables quantitative interpretation of coupled transport, electrochemical and geochemical mechanisms. Previous studies have shown that transport-based models can successfully describe metal migration and accumulation patterns under complex electrode configurations, such as hexagonal or multi-anode systems, thereby supporting experimental observations and system design [25]. More advanced approaches combining the Nernst–Planck–Poisson framework with geochemical speciation models, such as PHREEQC coupling, have further demonstrated that accurate prediction of EK performance requires simultaneous consideration of electromigration, diffusion, electroosmosis, and biogeochemical reactions [26]. Recent critical reviews emphasize that such predictive models are not only valuable for mechanistic interpretation but also represent a key step toward process scaling, energy optimization, and reduction of experimental effort in EK remediation of contaminated soils and sediments [27]. We used a modified mathematical model for metal transport simulation and prediction of time needed for sediment metal remediation to concentration levels that have no risk to the environment.

In this study, we investigate the ex situ application of the two-anode electrokinetic (TAT) configuration for the remediation of Cu- and Ni-contaminated dredged sediment, and we evaluate its performance under three operational regimes: continuous DC supply, interrupted power supply (night shutdown), and solar-powered operation. Unlike previous electrokinetic studies mainly focused on laboratory soils and single-metal systems, this work addresses real multi-metal sediment and integrates geochemical profiling (pH, ORP, EC) with metal speciation and risk-based interpretation of removal efficiency. Particular emphasis is placed on the role of the secondary anode in stabilizing the pH profile along the transport pathway and mitigating cathodic alkalization, thereby enhancing metal electromigration and reducing undesired accumulation. Furthermore, a mathematical transport model is developed and validated against experimental data to predict the remediation time required to reduce Cu and Ni concentrations to the regulatory non-hazardous (target) levels, providing a decision-support tool for process scaling and energy-efficient operation.

2. Materials and Methods

2.1. Sediment Sample

In the Republic of Serbia, the management of contaminated sediments obtained through dredging poses a serious environmental and regulatory challenge, as large quantities of sediment must be removed from waterways and, in accordance with national legislation, remediated whenever contamination is confirmed. For this reason, research into effective post-dredging treatment technologies is of particular practical importance.

The sediment used in the present ex situ electrokinetic experiments was obtained by dredging from the Veliki Bački Canal (Vojvodina, Serbia), one of the most critically polluted canal systems in the region. The canal has been exposed to long-term pollution since the second half of the 20th century, due to the continuous discharge of untreated and partially treated industrial and municipal wastewater. The most intense contamination occurs along a 6 km hydraulically stagnant section near Vrbas, where approximately 400,000 m³ of fine-grained sediment has accumulated in the channel bed over several decades as a result of emissions from the food, meat, and metallurgical industries, as well as municipal wastewater. This area is officially classified as one of the three most polluted water bodies in Serbia and represents a constant secondary source of metals and organic pollutants.

Sediment samples were collected in accordance with the SRPS ISO 5667-12:2019 [28] standard. Sampling was conducted across the entire cross-section of the channel at three lateral positions: left bank (45°35′2.39″ N, 19°37′31.26″ E), center of the channel (45°35′2.04″ N, 19°37′30.83″ E), and right bank (45°35′1.60″ N, 19°37′30.56″ E). Deep, undisturbed sediment cores were collected along the full sediment depth profile (100–160 cm) using an Eikelkamp Becker core sampler. Three vertically collected cores were combined and mechanically homogenized in a polyethylene container to obtain a representative composite sediment sample. The sediment was processed with its initial moisture content and natural Ni and Cu concentrations.

2.2. Electrokinetic Device

The electrokinetic device used as well as experiment details (solar panels characteristics, electrodes, etc) are described by Krčmar et al. [7]. For zero (conventional) treatment, two graphite electrodes were used as primary anodes, two iron electrodes as cathodes, and for TAT, another perforated graphite electrode was used as the secondary anode. The electrokinetic device for TAT treatment with electrode position and locations of sampling is shown in Figure 1.

Conventional treatment (e0) and three experiments of TAT were performed: (a) e1—with continuous application of electric current, (b) e2—with lack of electricity during the night (from 8 pm to 6 am), and (c) e3—with the use of solar panels. All experiments were conducted in triplicate to ensure data reproducibility and quality control, and the results presented in this paper correspond to the mean values of three parallel measurements.

All treatments lasted for 20 days. Samples were taken from 6 different places (Figure 1b) at certain distances from the anode (z/L-z = distance from the anode, L = length of the sample compartment) (Figure 1b) before the start of the experiment and after 10 and 20 days of experimentation. pH, oxidation reduction potential (ORP), electroconductivity (Ep), and pseudo-total content of Cu and Ni, were measured, and a sequential extraction procedure was performed on the samples.

2.3. Chemicals and Analytical Methods

All the chemicals used in this work were at least analytical-reagent grade. The methods used for sediment characterization are listed: preparation of sediment [29], sediment particle size [30], sediment dry matter [31], sediment organic matter content [32], sediment pH [33], sediment-specific electrical conductivity and sediment CEC [34], and pseudo-total metal content [35].

The pH and oxidation–reduction potential (ORP) were measured using a multiparameter meter (340i, WTW, Weilheim, Germany). The pH was determined with a SenTix^® pH electrode, whereas ORP values were measured using a SenTix^® ORP electrode via direct immersion in the tested medium. The determination of the specific electrical conductivity of the sediment was performed with conductor Cond 3210 and TetraCon electrode (WTW, Weilheim, Germany).

Structural analysis of sediment was performed using N2 adsorption isotherms at 77 K using an Autosorb iQ Surface Area Analyzer (Quantachrome Instruments, Boynton Beach, FL, USA). The Brunauer, Emmett, and Teller method (abbreviated BET) was used to characterize the sediment. Sediment X-ray diffraction (Philips PW automated X-ray powder diffractometer, Philips, Eindhoven, The Netherlands) was used to identify the mineral composition. The analysis was performed at room temperature, on a pre-dusted sample. Measurements were made in the 2θ range of 10–90°, with a step of 0.02 and 10 s exposure.

The results of pseudo-total metal concentrations (pseudo-total metal content refers to the concentration obtained after strong acid digestion, representing the environmentally relevant fraction of metals excluding those structurally bound within silicate minerals) and the corresponding sediment pollution classes are interpreted in relation to the Serbian regulation, which is harmonized with the Dutch sediment quality standards [36]. The assessment methodology is described in detail by Varga et al. (2017) [22] and Krčmar et al. (2018) [7]. Pollution with priority substances was classified on a 0–4 scale (from clean to heavily polluted), where Class 0 corresponds to background (non-hazardous) levels, representing the target concentrations prescribed by the regulation. Because the regulatory limits refer to a “standard sediment” containing 10% organic matter and 25% clay, the measured concentrations were first normalized to this reference composition using correction formulas based on the actual organic matter and clay content of the samples. Sediments classified as Classes 3–4 are considered to be of unacceptable quality and require dredging and remediation measures.

The speciation of Cu and Ni in sediment was analyzed using a sequential extraction scheme [36]. This sequential extraction schemes has been validated through inter-laboratory comparison experiments and using certified reference materials (CRM BCR 701). The leachate was obtained through centrifugation under laboratory conditions. It was filtered (0.45 µm), acidified with nitric acid to pH < 1, and stored in polyethylene containers at 40 °C. Analyses of metals were carried out using flame atomic absorption spectrophotometry (PerkinElmer; AAnalyst 700, Waltham, MA, USA) in accordance with U.S. EPA Method 7000 b [37]. Metals not detected using flame spectrometry were determined using the graphite method according to EPA 7010 [38]. For sediment samples, the limits of detection were 2.2 mg/kg for Cu and 0.11 mg/kg for Ni. For leachate (water phase), the limits of detection were 0.90 μg/L for Cu and 2.2 μg/L for Ni. The relative standard deviations (%RSD, n = 3) were below 5%, confirming good analytical precision and repeatability of the measurements.

2.4. Mathematical Model and Simulation Procedure

We modeled the transportation phenomena of Cu and Ni ions as a one-dimensional problem. Namely, the experimental configuration of the electrodes in the electrokinetic cell (Figure 1a) indicates that the metal separation process is carried out in one direction. Using the fact that the general movement of ions is due to advection, dispersion, sorption, and chemical reactions [39] in one direction, the partial differential equation for the mass distribution is as follows [40]:

$${\displaystyle \frac{\mathsf{\partial }}{\mathsf{\partial }\mathit{t}}}\left(\mathit{\theta }\mathit{C}\right)={\displaystyle \frac{\mathsf{\partial }}{\mathsf{\partial }\mathit{x}}}\left(\mathit{\theta }\mathit{D}{\displaystyle \frac{\mathsf{\partial }\mathit{C}}{\mathsf{\partial }\mathit{x}}}\right)-{\displaystyle \frac{\mathsf{\partial }}{\mathsf{\partial }\mathit{x}}}\left(\mathit{\theta }\mathit{v}\mathit{C}\right)+{\mathit{q}}_{\mathit{s}}{\mathit{C}}_{\mathit{s}}+\sum {\mathit{R}}_{\mathit{u}}$$

(1)

where θ denotes the total porosity, C is the particle concentration, t is time, and x is the distance along the coordinate axis. The term D represents the hydrodynamic dispersion coefficient, v is the groundwater (pore-water) velocity, and q_s is the volumetric flow rate per unit volume of the aquifer, with positive values indicating sources and negative values indicating sinks. C_s denotes the concentration associated with the source or sink flux, while ∑R_u represents the chemical reaction terms [41]. The left-hand side term in Equation (1) can therefore be rewritten as follows:

$${\displaystyle \frac{\mathsf{\partial }}{\mathsf{\partial }\mathit{t}}}\left(\mathit{\theta }\mathit{C}\right)=\mathit{C}{\displaystyle \frac{\mathsf{\partial }\mathit{\theta }}{\mathsf{\partial }\mathit{t}}}+\mathit{\theta }{\displaystyle \frac{\mathsf{\partial }\mathit{C}}{\mathsf{\partial }\mathit{t}}}=\mathit{C}{\mathit{q}}_{\mathit{s}}^{\mathbf{\prime }}+\mathit{\theta }{\displaystyle \frac{\mathsf{\partial }\mathit{C}}{\mathsf{\partial }\mathit{t}}}$$

(2)

and using the mathematical formulation of the reaction terms,

$$\sum {\mathbf{R}}_{\mathbf{u}}=-{\mathsf{\rho }}_{\mathbf{b}}{\displaystyle \frac{\mathsf{\partial }\overline{\mathbf{C}}}{\mathsf{\partial }\mathbf{t}}}-\mathsf{\lambda }{\mathsf{\rho }}_{\mathbf{b}}\overline{\mathbf{C}}$$

(3)

we obtain the transport phenomena model in the following form:

$${\mathsf{\rho }}_{\mathbf{b}}{\displaystyle \frac{\mathsf{\partial }\overline{\mathbf{C}}}{\mathsf{\partial }\mathbf{t}}}+\mathsf{\theta }{\displaystyle \frac{\mathsf{\partial }\mathbf{C}}{\mathsf{\partial }\mathbf{t}}}={\displaystyle \frac{\mathsf{\partial }}{\mathsf{\partial }\mathbf{x}}}\left(\mathsf{\theta }\mathbf{D}{\displaystyle \frac{\mathsf{\partial }\mathbf{C}}{\mathsf{\partial }\mathbf{x}}}\right)-{\displaystyle \frac{\mathsf{\partial }}{\mathsf{\partial }\mathbf{x}}}\left(\mathsf{\theta }\mathbf{v}\mathbf{C}\right)+{\mathbf{q}}_{\mathbf{s}}{\mathbf{C}}_{\mathbf{s}}-\mathbf{C}{\mathbf{q}}_{\mathbf{s}}^{\mathbf{\prime }}-\mathsf{\lambda }{\mathsf{\rho }}_{\mathbf{b}}\overline{\mathbf{C}}$$

(4)

where $q_s^{\mathbf{\prime }}={\displaystyle \frac{\partial \theta }{\partial t}}$ denotes the accumulation gradient, ${\rho }_b$is the bulk density of the porous medium, λ is the first-order reaction rate constant for the dissolved–sorbed phase, and $\overline{C}$ is the concentration of ions sorbed onto the electrodes. By introducing the retardation factor [42],

$$R=1+{\displaystyle \frac{{\rho }_b}{\theta }}{\displaystyle \frac{\partial \overline{C}}{\partial C}}$$

(5)

Equation (4) thus becomes as follows:

$$\mathit{R}\mathit{\theta }{\displaystyle \frac{\mathsf{\partial }\mathit{C}}{\mathsf{\partial }\mathit{t}}}={\displaystyle \frac{\mathsf{\partial }}{\mathsf{\partial }\mathit{x}}}\left(\mathit{\theta }\mathit{D}{\displaystyle \frac{\mathsf{\partial }\mathit{C}}{\mathsf{\partial }\mathit{x}}}\right)-{\displaystyle \frac{\mathsf{\partial }}{\mathsf{\partial }\mathit{x}}}\left(\mathit{\theta }\mathit{v}\mathit{C}\right)+{\mathit{q}}_{\mathit{s}}{\mathit{C}}_{\mathit{s}}-\mathit{C}{\mathit{q}}_{\mathit{s}}^{\mathbf{\prime }}-\mathsf{\lambda }{\mathit{\rho }}_{\mathit{b}}\overline{\mathit{C}}$$

(6)

For the special case in which the porosity $\mathit{\theta }$, velocity $\mathit{v}$, and dispersion coefficient $\mathit{D}$ are constant, Equation (6) reduces to the following:

$$\mathit{R}\mathit{\theta }{\displaystyle \frac{\mathsf{\partial }\mathit{C}}{\mathsf{\partial }\mathit{t}}}=\mathit{\theta }\mathit{D}{\displaystyle \frac{{\mathsf{\partial }}^2\mathit{C}}{\mathsf{\partial }{\mathit{x}}^2}}-\mathit{\theta }\mathit{v}{\displaystyle \frac{\mathsf{\partial }\mathit{C}}{\mathsf{\partial }\mathit{x}}}+{\mathit{q}}_{\mathit{s}}{\mathit{C}}_{\mathit{s}}-\mathsf{\lambda }{\mathit{\rho }}_{\mathit{b}}{\mathit{K}}_{\mathit{d}}\mathit{C}$$

(7)

since the accumulation gradient is equal to zero, i.e., ${\mathit{q}}_{\mathit{s}}^{\mathbf{\prime }}\mathbf{=}{\displaystyle \frac{\mathsf{\partial }\mathit{\theta }}{\mathsf{\partial }\mathit{t}}}\mathbf{=}\mathbf{0}.$ Finally, the transport equation describing the migration of Ni and Cu ions is given as follows:

$$\mathit{R}{\displaystyle \frac{\mathsf{\partial }\mathit{C}}{\mathsf{\partial }\mathit{t}}}=\mathit{D}{\displaystyle \frac{{\mathsf{\partial }}^2\mathit{C}}{\mathsf{\partial }{\mathit{x}}^2}}-\mathit{v}{\displaystyle \frac{\mathsf{\partial }\mathit{C}}{\mathsf{\partial }\mathit{x}}}+{\displaystyle \frac{{\mathit{q}}_{\mathit{s}}{\mathit{C}}_{\mathit{s}}}{\mathit{\theta }}}-\mathit{\lambda }(\mathit{R}-\mathbf{1})\mathit{C}$$

(8)

It is a second-order partial differential equation, whose solution is assumed in the following form:

$$\mathit{C}=\mathit{b}-\mathit{a}\mathit{e}\mathit{x}\mathit{p}(\mathit{p}\mathit{t}-\mathit{s}\mathit{x})$$

(9)

where $\mathit{a}$, $\mathit{b}$, $\mathit{p}$, and $\mathit{s}$ are unknown constants. The expression in Equation (9) is used as the fitting model for the problem. The four unknown constants are determined using a standard numerical curve-fitting procedure. By substituting the experimentally obtained data into Equation (9), a system of four algebraic equations is formed, from which the parameters $\mathit{a}$, $\mathit{b}$, $\mathit{p}$, and $\mathit{s}$ are calculated.

3. Results

3.1. Initial Sediment Properties

The sediment characteristics are shown in

Table 1. Properties of the sediment used in the electrokinetic experiments.

Parameter			Value
Water content (%)			61.2 ± 2.8
Organic matter (%)			16.8 ± 0.5
Clay (%)			48.1 ± 1.8
CEC, meq/100 g			20.6 ± 0.9
BET (m2g−1)			13.0 ± 0.4
Metal (mg/kg)		Corrected value (mg/kg)	Class
Cd	2.55 ± 0.09	1.84	1
Cr	111 ± 5.32	76.0	0
Cu	146 ± 5.87	97.4	3
Pb	97.3 ± 4.23	72.0	0
Ni	93.5 ± 3.71	56.3	3
Zn	288 ± 9.65	183.7	1
Sediment fractions		Cu (mg/kg)	Cu (%)
Weak acid soluble		28.6 ± 0.94	2.22
Reducible		59.3 ± 2.61	56.1
Oxidizable		34.4 ± 1.47	20.5
Residual		34.8 ± 1.52	21.2
Sediment fractions		Ni (mg/kg)	Ni (%)
Weak acid soluble		21.2 ± 0.88	24.7
Reducible		27.1 ± 1.12	37.8
Oxidizable		14.8 ± 0.63	10.5
Residual		22.2 ± 0.96	27.0

. The sediment is characterized by a high content of organic matter, which is also responsible for the high CEC value. An approximate value for CEC was published by Rajić et al. [43]. The high content of clay in the sediment explains the obtained (high) value of the specific surface. The BET value plays an important role in adsorption elucidation—larger specific particle surface results in a stronger adsorption capacity [44].

The determined and corrected concentrations of heavy metals in the sediment for comparison with the limit values for assessing the status and trend of sediment quality [45], as well as the results of sequential extraction, are shown in Table 1. The obtained values indicate that the sediment is contaminated by Cu and Ni, as they correspond to class 3. Class 3 means the sediment is contaminated. It is not allowed to be dislocated without special protection measures. The largest amounts of Cu and Ni are bound to iron and manganese oxyhydroxides. It is a reducing fraction of sediment, and if the oxide reductive potential and oxygen levels change, metal desorption and secondary pollution can occur [46]. According to the Risk Assessment Code—RAC [47], Cu has low risk to aquatic systems, and Ni has moderate risk.

It is well known that geochemical characteristics affect the availability of heavy metals. The geochemical status and potential mobility of metals can be determined based on mineralogical composition directly by applying X-ray diffraction (XRD) analysis to quantify mineral components [1]. The most abundant mineral of the Veliki Bački Canal sediment is ilite (Figure 2). Illite belongs to the group of 2:1 phyllosilicates, which are considered to be concentrators of environmental cationic pollutants due to excess negative charge in the structure, a larger number of edge sites, and a larger internal volume, i.e., interlayer space [48].

Based on the physicochemical characteristics of calcite (5%) and dolomite (3%) that react when in contact with HCl and biotite (6.1%), which is considered highly unstable and easily susceptible to chemical degradation, we can conclude that metals bound to these minerals can be easily desorbed and become available. This conclusion was proved by Ryan et al. [49] because calcite and dolomite completely disintegrated after the addition of CH₃COOH applied to extract the first sediment fraction using a sequential extraction procedure. Clinochlorine (4%) belongs to the chlorite group [50], which is why we can assume that it behaves similarly. Ryan et al. [49] demonstrated that chlorite (6.1%) was partially extracted from carbonates, oxides, sulfides, and organic matter after the addition of the appropriate chemicals. Such observations indicate that when the oxide-reducing conditions change, metals can be desorbed from these minerals and become available in aquatic systems.

Quartz (22.2%) and muscovite (16.2%) are significantly present in the Veliki Bački Canal sediment. Metals are thought to be tightly bound to these minerals, unavailable to the environment because they remain unchanged after the sequential extraction procedure has been applied [49]. Clay minerals have been proven in many papers to dictate the adsorption capacity of sediment [44]. In line with the above comments, we can conclude that in this sediment, more than 60% of the minerals present are susceptible to sorption or desorption of metals due to the physicochemical changes in the sediment that are expected during the treatment of electrokinetic remediation.

3.2. Profiles of pH, Redox Potential (ORP), and Electrical Conductivity (EC) After All EK Experiments

pH is a key parameter governing the migration and retention of heavy metals, as it controls adsorption–desorption reactions and ion-exchange processes involving soluble organic matter, carbonates, sulfides, and Fe-(oxy)hydroxides. Consequently, variations in pH directly affect the metal content and mobility within the sediment [51]. In the conducted experiments, the pH value of the anode region (z/L 0.1) decreased to ~6 during all the treatments performed. In the cathode region (z/L 0.9 and 1), the pH value increased to ~10 during e0, whereas it was slightly lower (~9) during the two-anode treatment (e1, e2, e3). Such a pH distribution is particularly important because an increase in pH promotes the precipitation and accumulation of metals in the form of hydroxides and carbonates, thereby limiting their mobility and transport [52]. Accordingly, it can be concluded that the secondary anode in the e1–e3 configuration effectively mitigates the advancement of the alkaline front, providing more favorable conditions for electromigration of metals along the transport pathway. Moreover, the absence of pronounced pH variations in our system indicates that metal removal occurred under chemically stable conditions, without compromising the geochemical integrity of the sediment [25].

Interactions between the electrochemical processes and the sediment itself influence the sediment redox potential. ORP of sediments is considered as a parameter controlling the mobility of heavy metals [53] since hydrogen and oxygen produced can dictate degradation of the pollutants. Reducing conditions (−343 mV) were in the initial sediment due to the presence of naturally occurring organic matter. It can be seen from Figure 3b that after all treatments (e0, e1, e2, e3), ORP increased over time at z/L 0.1, 0.3, 0.5 (e0), and at z/L 0.1, 0.3, 0.5, 0.7 (e1, e2, e3) due to the presence of a secondary electrode. The increasing value of ORP in sediment enables oxidation of sulfides and degradation of organic compounds, thus accelerating the release of adsorbed/complexed heavy metals [53,54]. In contrast, the ORP had a decreasing trend at z/L of 0.7, 0.9, and 1 (e0), and at z/L of 0.9 and 1 (e1, e2, e3), indicating that at these distances from the anode, reducing conditions take over, resulting in the formation of insoluble sulfides. Similar observations were made by Hahladakis et al. [55,56] and Rajic et al. [57].

The conductivity determines the amount of current flowing through the treated medium [58,59]. It also influences other properties, including moisture content, dissolved salts, particle size, temperature, texture, and ionic capacity changes [58]. In the e0 treatment, at z/L 0.1, 0.3 there was an increase in conductivity, and we can see that the EC decreased from the anode to the cathode. On the other hand, after e1, e2, e3, TAT EC decreased at all distances from the anode relative to the initial value, and a significant decrease was observed at z/L 0.9 and 1 (Figure 3c). The obtained values in e1, e2, e3 can be explained by a decrease in the amount of free ions over time in both the anodic and cathodic regions. The decrease in the conductivity of the anode region indicates that the free ions present in the initial sediment migrated; that is, with the decrease in the ion migration rate, the conductivity of the sediment also decreases [59]. On the other hand, under the influence of the increased pH in the cathode region, the free ions precipitated as insoluble hydroxides, causing the conductivity in these parts to decrease significantly. This declining trend in conductivity is consistent with Hahladakis et al. [55,56].

3.3. Current Changes and Energy Consumption

Figure 4 illustrates the temporal variation in the electric current during the electrokinetic treatments. In all experiments, the current reached its maximum at the beginning of the process, which is attributed to the initially high concentration of dissolved ions available for electromigration in the pore water. As the treatment progressed and the concentration of dissolved ions gradually decreased, the electrical current declined accordingly. This behavior is consistent with the pH evolution and ion-transport mechanisms discussed in Section 3.2, and it is in agreement with previous studies that reported a similar decreasing trend in current intensity during EK treatment (Lei et al., [60]; Zhang et al., [53]; Varga et al., [22]).

The power generated by solar cells depends on both the time of day and weather conditions; therefore, the effectiveness of the treatment is also influenced by these factors [21,61,62]. In treatment e3, several pronounced drops in current intensity were observed at specific time intervals, corresponding to periods of cloudy weather and rainfall (days 2, 3, 5, 6, 7, 12, 13, and 17). Since the power output of the photovoltaic source depends on meteorological conditions, the available electrical input and thus the EK treatment efficiency fluctuated during these periods, as also reported in previous studies on solar-powered EK systems.

In treatments e1 and e2, the initial current was approximately 0.5 A under a constant applied voltage of 80 V. In contrast, during e3, where solar energy was used, the initial current was slightly higher (~0.65 A), while the applied voltage varied as a function of weather conditions, consistent with the observations reported by Krčmar et al. [7]. The total electrical charge supplied to the EK reactor during e0, e1, e2, and e3 was approximately 99 Ah, 103 Ah, 63 Ah, and 69 Ah, respectively.

3.4. Distribution of Cu and Ni

The pseudo-total copper and nickel contents in the sediment after electrokinetic treatments (e0, e1, e2, and e3), together with the corresponding sediment classification, are presented in

Table 2. Pseudo-total metal contents in the sediment after electrokinetic treatment and corresponding sediment classification.

Treatment	Metal	Cu (mg/kg)-10 day			Cu (mg/kg)-20 day			Ni (mg/kg)-10 day			Ni (mg/kg)-20 day
	z/L	MV	CV	Class	MV	CV	Class	MV	CV	Class	MV	CV	Class
e0	0.1	96.5 ±4.2	64.1	2	68.1 ±2.9	45.5	2	70.5 ± 3.2	42.5	2	59.1 ± 2.6	35.6	2
	0.3	102 ± 4.4	68.1	2	84.9 ±3.5	56.7	2	66.9 ± 2.9	40.3	2	66.2 ± 2.8	39.9	2
	0.5	120 ± 5.1	80.1	2	100 ± 4.7	66.7	2	76.5 ± 3.3	46.1	3	61.6 ± 1.9	37.1	2
	0.7	151 ± 5.9	100.8	3	176 ± 7.9	117	3	86.2 ± 3.9	51.9	3	104 ± 4.8	62.7	3
	0.9	164 ± 6.3	109.5	3	228 ± 9.3	152	3	106 ± 4.8	63.9	3	122 ± 5.7	73.5	3
	1	158 ± 5.3	105.5	3	207 ± 8.8	138	3	100 ± 4.2	60.2	3	124 ± 5.9	74.7	3
e1	0.1	78.3 ±3.1	52.3	2	52.4 ±2.2	35	1	64.4 ± 2.6	38.8	2	51.8 ± 2.2	31.2	0
	0.3	79.4 ±2.9	53.0	2	47.4 ± 1.9	31.6	1	64.9 ± 2.5	39.1	2	50.4 ± 2.1	30.4	0
	0.5	77.8 ±2.8	51.9	2	54.1 ± 2.3	36.1	2	65.4 ± 2.7	39.4	2	54.1 ± 2.4	32.6	0
	0.7	81.4 ±3.2	54.3	2	56.4 ±2.6	37.6	2	70.0 ± 3.1	42.2	2	53.3 ± 2.4	32.1	0
	0.9	133 ± 6.4	89.0	2	111 ± 5.1	74.4	2	82.7 ± 3.6	49.8	3	72.6 ± 3.3	43.7	2
	1	153 ± 6.9	102.1	3	165 ± 7.8	110	3	96.7 ± 4.5	58.3	3	105 ± 4.7	63.1	3
e2	0.1	86 ± 3.9	57.4	2	55.9 ± 2.2	37.4	2	62.6 ± 2.1	37.7	2	49.3 ± 2.1	29.7	0
	0.3	85.1 ±3.5	56.8	2	59.7 ± 2.4	39.8	2	60.9 ± 2.3	36.7	2	53.2 ± 2.3	32.0	0
	0.5	89.7 ±3.7	59.9	2	59.7 ± 2.8	39.9	2	65.2 ± 2.7	39.2	2	48.5 ± 2.0	29.2	0
	0.7	93.7 ±4.2	62.5	2	64.8 ± 2.9	43.2	2	67.3 ± 2.9	40.5	2	49.1 ± 2.2	29.5	0
	0.9	135 ± 6.5	90.2	3	110 ± 5.1	73.2	2	81.7 ± 3.5	49.2	3	72.4 ± 3.3	43.6	2
	1	153 ± 7.1	102.2	3	178 ± 8.4	119	3	97.0 ± 4.6	58.4	3	103 ± 4.8	61.9	3
e3	0.1	90.7 ±3.5	60.5	2	61.3 ± 2.8	40.9	2	56.5 ± 1.9	34.0	2	46.0 ± 1.8	27.7	0
	0.3	94.5 ±4.1	63.1	2	56.1 ± 2.3	37.4	2	62.4 ± 2.9	37.6	2	52.6 ± 2.2	31.7	0
	0.5	90.3 ±3.6	60.3	2	61.4 ± 2.6	41.0	2	62.5 ± 2.9	37.7	2	49.3 ± 2.1	29.7	0
	0.7	97.5 ±4.4	65.1	2	56.7 ± 2.1	37.8	2	67.3 ± 3.1	40.5	2	50.1 ± 2.2	30.2	0
	0.9	136 ± 5.8	91	3	118 ± 4.2	78.6	2	83.9 ± 3.7	50.5	3	74.0 ± 3.5	44.6	2
	1	162 ± 7.7	107.9	3	169 ± 7.7	113	3	102 ± 4.9	61.3	3	106 ± 4.4	64	3

MV, measured value (mg kg^–1); CV, corrected value (mg kg^–1) [45].

. During the e0 treatment, the highest removal efficiency for both metals is achieved at z/L 0.1 (Cu—53%, Ni—36%). With distance from the anode, the efficiency in e0 decreases, even at z/L 0.7, 0.9 and 1. In contrast, after the e1, e2, e3 treatments, significant decreases in Cu and Ni concentrations are observed at z/L 0.1, 0.3, 0.5, 0.7, ranging from 56% to 68% for Cu and from 41% to 50% for Ni. Also, z/L 0.9 showed decreases in the metals of 19% to 25% for Cu and 10 to 20% for Ni. Based on the above experimental results, we can conclude that higher efficiency is achieved in the first 10 days of all experiments. The faster migration of available metals in the first half of the treatment is primarily due to the higher electrical current as well as the higher sediment conductivity [63]. These observations are consistent with the changes presented in 3.2. and 3.3. It is also observed that the percentage of concentration reduction on z/L of 0.1, 0.3, 0.5, 0.7 Cu and Ni during the TAT treatments is approximately equal to the sum of the carbonate and reducing fractions (

Table 3. Removal efficiency after electrokinetic treatment experiment.

Experiment Assignation	Removal Efficiency (%)
	Cu	Ni
Fe0	10	11
e1	56	39
e2	52	41
e3	52	41

). This leads to the conclusion that the removal efficiencies have been achieved in accordance with the percentage content of fractions considered available and subject to electrokinetic processes. A similar conclusion was reported by Varga et al. [22].

As mentioned, after e0 at z/L 0.7, 0.9, and 1, both metals were accumulated in comparison with the initial values (30–36% of Ni and 20–56% of Cu). At the end of e1, e2, e3, a higher concentration was detected only at z/L 1 of 13% to 22% for Cu and ~15% for Ni. This localized accumulation is attributed to physicochemical changes in the cathodic region, primarily the increase in pH and associated redox conditions, and it represents a well-known side effect of electrokinetic remediation rather than permanent removal of metals [27,64]. The observed accumulation of metals is responsible for the change in pH at the indicated distances from the anode (Figure 3a). Based on this, we can assume the formation of hydroxide (Cu(OH)₂) [65], oxide (CuO, Cu₂O), sulfides (CuS, Cu₂S), or complexes with organic ligands (humic and fluvic acids) of copper [66,67], or that Ni precipitated in the form of Ni(OH)₂ hydroxide, or, in turn, it wsa converted to the form of Ni₃O₄, Ni₂O₃, and peroxide oxides [68]. Consistent with recent review studies, these transformations indicate metal immobilization via precipitation in the cathodic zone, which should be interpreted as redistribution or temporary immobilization rather than true remediation [27,64].

The overall efficiencies of the treatments performed except for z/L 1 are shown in Table 2. In this study, the overall treatment efficacy (%) was calculated according to the following expression:

% = [(m_i − m_k)/m_i] × 100

where m_i represents the initial amount of Cu or Ni in the sediment, and m_k is the amount remaining after treatment. Importantly, this calculation excludes the cathodic accumulation zone (z/L = 1.0) and therefore reflects the net removal efficiency within the bulk sediment profile rather than local pollutant transfer or immobilization.

Based on the efficiencies shown, it is clear that the use of a secondary electrode greatly improves EK treatment. The absence of electricity during the night, changes in current during solar panels, or differences in the overall amount of applied electric field during the treatment do not significantly affect the treatment efficiency (e1, e2, e3). The efficiencies obtained also indicate that the sediment after e1, e2, e3 with z/L 0.1, 0.3, 0.5, 0.7, 0.9 could be dislocated without special protection measures [45] since the sediment is unpolluted (Cu 2 class, Ni 0 or 2 class). The sediment at z/L 1 is considered polluted (Cu 3 class, Ni 3 class), highlighting that cathodic accumulation represents an undesirable side effect of electrokinetic remediation rather than permanent removal, as emphasized in recent review studies on EK treatment of soils and sediments [27,64]. From both an economic and environmental perspective, treatment e3, based on solar-powered electrokinetic remediation, represents the most sustainable option, as it achieves comparable removal efficiencies while reducing energy consumption and operational costs.

3.5. Speciation of Cu and Ni in the EK Cell

Sequential extraction was performed in order to better understand the achieved migration of the tested metals, but also to identify changes in the availability of metals during electrokinetic remediation. The percentages of Cu and Ni in different sediment fractions after 10 and 20 days of e0, e1, e2, e3 treatments are presented in the Supplementary Materials as Figures S1 and S2.

Cu according to the results of sequential extraction during e0 (z/L0.1, 0.3, 0.5), e1, e2, e3 (z/L 0.1, 0.3, 0.5, 0.7) in the first 10 days, there was a decrease in the II and III fractions. Fraction I shows an increase of 8.5% to 13% in e0, e1, and 6.5% to 8% in e2, e3. After 20 days of treatment, the available amount of Cu was <5% at the indicated distances from the anode. The percentage of Cu bound to Fe and Mn oxides was reduced at all distances from the anode (from 19% to 35%). The results obtained confirm the depletion of mobile Cu fractions that undergo electrokinetic processes. On the other hand, upon completion of the conducted treatments, a significant increase in the organic fraction (~20%) was observed to z/L 0.7, 0.9, and 1 in e0, and to z/L 0.9 and 1 in e1, e2 and e3. In general, Cu is known to have a high affinity to building highly stable complexes with organic matter [67,69]. Also, the residual fraction was increased by about 10% along the electrokinetic cell after e0, e1, e2, e3, which can be explained by the fact that Cu is readily chemisorbed and incorporated into clay minerals [70,71].

Ni. Generally, during the treatments (e0, e1, e2, e3), there was a decrease in the percentage of Ni content in the reducing fraction at all distances from the anode. After 10 days of treatment, the available carbonate fraction increased by about 6% to 10% at z/L 0.1, 0.3, 0.5 (e0) and at z/L 0.1, 0.3, 0.5, 0.7, (e1, e2, e3). After 20 days, the Ni content decreased by 10% to 15%, with the increasing content in the oxidizable fraction. This trend is correlated with the physicochemical changes, as well as with the achieved efficiency of the conducted treatments (indicating that Ni accumulation occurred in the cathode region). Thus, in the second half of the electrokinetic cell, the exchangeable/carbonate and reductive fraction is converted into an oxidizable fraction, which increased by 20 to 25% in that region (z/L 0.7, 0.9, and 1-e0, respectively, z/L 0.9 and 1-e1, e2, e3). Based on the obtained results, we can conclude that there is a redistribution of Ni between the carbonate and oxidizable fractions. Similar conclusions were reached by Gao et al. [72]. The percentage of the residual fraction did not change significantly.

Sediment risk assessment according to RAC was performed only for the cathode region (z/L 1) of e3 treatment in which solar panels are applied, as we have assessed that this treatment is best suited for real field application. The obtained results indicate that the risk for aquatic systems is low in terms of Cu and moderately low in terms of Ni. Considering the high content of clay (Table 2), environmental conditions in this region (pH, ORP, EC), and the fact that more than 50% Cu and Ni contents are found in the organic and residual fractions, we can assume that there are no environmental hazards.

3.6. Leached Water

The dissolved metals, when applied to the horizontal electric field, as in the case of these experiments, move downstream of the diffusion and dispersion processes and reach deeper layers of sediment and groundwater [42]. Therefore, in order to evaluate the quality of the water that would reach underground during deposition of this sediment at the selected site, as well as to examine the possibility of returning this water to the watercourse if there was a waterproofing layer, the contents of Cu and Ni in the leachate were analyzed.

The detected concentrations of Cu (e0—70 ± 2.6 µg/L, e1—64.4 ± 1.5 µg/L; e2—48.2 ± 1.8 µg /L; e3—61.8 ± 2.6 µg /L) and Ni (e0—63 ± 2.1 µg/L, e1—24,5 ± 0.4 µg /L; e2—39.7 ± 2.1 µg /L; e3—26,3 ± 0.6 µg/L), according to the regulatory threshold values for Cu and Ni [30], indicate that the quality of the leachate corresponds to a moderate ecological status. Surface waters belonging to this class provide acceptable living conditions based on the prescribed quality limits and may be used for the following purposes: drinking water supply (after treatment with coagulation, flocculation, filtration, and disinfection), bathing and recreation, irrigation, and industrial use (process and cooling water). The exception is the leachate from treatment e0, which corresponds to a poor ecological status with respect to Ni. Surface waters belonging to this class cannot be used for any of the above purposes.

3.7. Mathematical Model—Numerical Simulation of the Transport Phenomena

The mathematical model or the numerical simulation of the transport phenomenon was applied to experiments that showed satisfactory removal efficiency—e1, e2, e3.

The target concentrations (corresponding to the background levels, non-hazardous levels) for Ni and Cu were calculated in accordance with the regulation [45]. The target values for VBKs were 30 mg/kg. For the experimental results obtained in 3.4. the analytically obtained equations were applied [7]. The concentration changes in Cu and Ni as a function of time for selected distances from the anode (e1, e2, e3; z/L = 0.1, 0.3, 0.5, 0.7, 0.9) are graphically presented in the Supplementary Materials as Figures S3 and S4. The experimentally obtained concentrations are shown as data points on the fitted c–t curves; the dashed lines represent the target values of Cu and Ni, and the solid lines indicate the corresponding corrected concentrations according to the regulations.

Numerical simulation indicates that the time required for reaching the Ni target values increases with distance from the anode. Target concentrations according to numerical simulation in e1 and e3 are reached after 32 days, and in e2, after 29 days for Ni. The results obtained for Cu in e1 and e2 indicate that the target values will be reached in 34 days, and in e3, that the sediment with respect to Cu will be purified in 40 days at z/L 0.1, 0.3, 0.5, 0.7, while more than 40 days are required to reach the z/L 0.9 anode distance target.

Figure 5 directly compares the experimentally observed and model-predicted times required to reach the target Cu and Ni concentrations, where the experimental values represent mean concentrations at z/L = 0.1–0.9 corrected to standard sediment. Target values of Ni are achieved at all distances from the anode (e1, e2, e3) during the conducted treatments (class 0). Class 0 means that the concentrations of pollutants in the sediment are at the level of the natural background. Sediments can be dislocated without special protection measures. The figure further clearly demonstrates that there is no significant difference between the experimentally observed treatment duration required to reach the target concentrations and the time predicted by the numerical model. Cu was present at slightly higher concentrations (Class 2), indicating a slightly polluted sediment. According to the regulation, sediments belonging to this class may be disposed of without special protection measures within a zone up to 20 m from the watercourse. Due to the difference between the experimentally achieved Class 2 concentrations for Cu and the Class 0 target concentrations predicted by the model, a more pronounced difference in the time required to reach the target values was observed for Cu. Nevertheless, the model predictions showed good agreement with the experimental results, which is consistent with the findings reported by Krčmar et al. [7]. Using the model, we can calculate the time required to reduce the concentration of metals to the target values or non-hazardous levels without conducting an experiment. In this way, we can reduce the number of analyses, the energy consumed, and therefore, the total cost of treatment.

4. Conclusions

This paper presents the ex situ remediation of sediment by TAT treatment under three different conditions: continuous application of electric current, no current at night, and solar panels. All treatments were applied to Cu- and Ni-contaminated Veliki Bački Canal sediment. Based on the results obtained, we can conclude that the behavior of Cu and Ni during EK treatment (in terms of changes in the pseudo-total content and changes in the availability of Cu and Ni along the EK cell) is consistent with physicochemical changes (pH, Ep, ORP). The overall efficacy of the treatments performed, except for z/L 1, indicates that the use of a secondary electrode enhances EK treatment. Differences in the total amount of applied electric field during treatment do not significantly affect the treatment efficiency (e1, e2, e3). After TAT treatment, most of the samples could be dislocated without special protection measures. Considering all the factors (efficacy, cost, environmental compatibility), we can conclude that the treatment in which solar panels are applied is the most appropriate.

Also, in this paper, a mathematical model was applied to the obtained results in order to predict the time necessary for sediment remediation. The obtained results indicate that the developed mathematical model can be successfully applied as a useful tool for electrokinetic remediation treatments.