Removal of Cu and Pb in Contaminated Loess by Electrokinetic Remediation Using Novel Hydrogel Electrodes Coupled with Focusing Position Adjustment and Exchange Electrode

Abstract

Electrokinetic (EK) remediation is a promising approach for the removal of heavy metals from fine-grained soils; however, its efficiency is often hindered by electrode polarization, pH imbalance, and ion accumulation. In this study, we developed a novel hydrogel-based electrode (NH electrode), composed of sodium alginate and multilayer graphene oxide (GO), to enhance the electrokinetic removal of Cu²⁺ and Pb²⁺ from loess. The electrode was systematically characterized by atomic force microscopy (AFM), cyclic voltammetry (CV), and electrochemical impedance spectroscopy (EIS), confirming its structural integrity, electrochemical activity, and interfacial conductivity. The NH electrode exhibited a smooth layered graphene structure with abundant oxygen-containing functional groups (AFM), negligible electrochemical polarization (CV), and low internal resistance with high conductivity (EIS), enabling efficient ion transport and adsorption. Electrokinetic tests revealed that the NH electrode outperformed conventional graphene (Gr) and electrokinetic graphite (EKG) electrodes. Single regulation strategies, including focusing position adjustment and electrode exchange, improved local removal efficiency by mitigating ion accumulation in targeted regions. The combined regulation strategy, integrating both measures, achieved the most uniform Cu²⁺ and Pb²⁺ removal, significantly suppressing hydroxide precipitation in cathodic zones and enhancing ion migration in the mid-section. Compared with literature-reported systems under similar or even more favorable conditions, the NH electrode and combined regulation approach achieved superior performance, with Cu²⁺ and Pb²⁺ removal efficiencies reaching 47.25% and 16.93%, respectively. These findings demonstrate that coupling electrode material innovation with spatial–temporal pH/flow field regulation can overcome key bottlenecks in EK remediation of heavy-metal-contaminated loess.

1. Introduction

Rapid industrialization and urban expansion in north-western (NW) China have led to increasingly severe environmental pollution, posing substantial risks to the integrity of geotechnical infrastructure, subsurface environmental quality, and public health [1,2]. Among the principal contaminants, heavy metals such as copper and lead, which primarily originate from mining and smelting activities, migrate and accumulate within soil systems [3,4]. This has resulted in widespread contamination of loess deposits, which are predominant in NW China and particularly vulnerable due to their unsaturated nature and structurally fragile fabric [5,6]. Nationwide, an estimated 20 percent of arable land is affected by heavy metal pollution. The loess region, with its distinctive geotechnical properties and extensive spatial coverage, is therefore at heightened environmental and ecological risk. Loess, characterized by its high porosity, loose structure, and poor water retention capacity, exhibits a markedly greater propensity for heavy metal accumulation and migration compared to other soil types [7]. These properties not only accelerate the transport and buildup of contaminants within the soil matrix but also complicate subsequent remediation efforts. Studies have shown that heavy metal pollution in loess regions is typically associated with strong toxicity, high concealment, and long-term persistence [8], posing sustained threats to soil ecosystem stability and regional geotechnical safety. As a result, it has drawn considerable attention in both environmental and engineering disciplines [1,5,6]. In particular, remediation under co-contamination by copper and lead remains technically challenging, with limited progress in developing mature and reliable treatment strategies. There is thus an urgent need for efficient and adaptable soil remediation technologies to improve both the environmental quality and engineering performance of heavy metal-contaminated loess. Such advancements are essential to support sustainable land reuse and effective risk management in vulnerable loess regions.

To date, various methods have been employed to remediate heavy metal-contaminated soil, including phytoremediation [9,10,11,12], thermal desorption [13,14,15], soil replacement [16], stabilization/solidification [17,18,19], chemical leaching [20,21], microbial remediation [22,23], and electrokinetic (EK) remediation technology. Notably, EK remediation technology has recently gained attention as a promising and innovative approach, particularly for low-permeability soils, by inserting electrodes into the soil and applying a low direct-current (DC) voltage gradient [24]. In recent years, significant progress has been made in the field of EK remediation, both in theoretical understanding and practical applications. Researchers have highlighted the key advantages and limitations of this technique. As an effective in situ remediation method, EK remediation is widely recognized for its low cost, ease of operation, high removal efficiency, and minimal secondary pollution. The removal of heavy metal contaminants is primarily achieved through mechanisms such as electroosmosis, electromigration, and electrophoresis. Despite these advantages, two major challenges remain [25,26]. First, delayed electrochemical reactions can result in asymmetric electrode charging and discharging behavior, causing deviations in electrode potential from the theoretical values. This phenomenon, known as electrochemical polarization, negatively impacts ion migration efficiency [25,26]. Second, the electrolysis of water at the cathode generates OH⁻ ions, which migrate toward the anode and react with metal cations migrating in the opposite direction. This leads to the formation of metal hydroxide precipitates near the cathode, a phenomenon referred to as the focusing effect [27,28,29,30]. These precipitates can clog soil pores, obstruct ion transport, and ultimately reduce removal efficiency. However, most previous studies have not fully addressed or resolved the issues of electrochemical polarization and the focusing effect, which remain critical barriers to the widespread and effective application of EK remediation.

Through a systematic review of current research [31,32,33,34], we have developed a comprehensive understanding of the application status of EK remediation technologies both domestically and internationally. It is well established that prior research has focused on improving the electrochemical activity and conductivity of electrode materials used in EK remediation. To mitigate the loss of electrochemical activity caused by the formation of oxide films on electrode surfaces during the electrokinetic process, Mendez et al. [35] and Jeon et al. [36] investigated the use of metal oxides such as SnO₂–SnO₃, IrO₂, Ta₂O₅, and RuO₂ as alternative electrode materials. These oxide coatings were intended to suppress electrode polarization and enhance electrochemical stability. Suzuki et al. [37] compared electrodes made from various materials and reported that titanium electrodes exhibited superior performance in the electrokinetic removal of lead from kaolinite. Although titanium and other composite electrodes have shown promise in alleviating electrochemical polarization compared to conventional graphite electrodes, their high cost and complex fabrication procedures remain major barriers to large-scale engineering applications [38]. Several studies have also aimed to mitigate the ‘focusing effect’ caused by hydroxide ion accumulation near the cathode. Lee et al. [39] and Zhang et al. [40] controlled the pH of the cathodic chamber using sulfuric acid to maintain a mildly acidic environment, thereby suppressing hydroxide precipitation and enhancing the remediation of manganese-contaminated sludge from drinking water treatment plants. Similarly, Wu et al. [41] applied acidic neutralization strategies to reduce local OH⁻ generation near the cathode, effectively mitigating pore clogging and improving ion migration efficiency. In addition, Hsu et al. [42] evaluated the influence of fine-grained soil properties and surface potentials on contaminant interactions, along with the role of EDTA under various operational conditions in enhancing heavy metal removal. Wang et al. [43] and Cuevas et al. [44] demonstrated that modifying electrode configurations into triangular, square, and hexagonal layouts can significantly improve removal efficiency at contaminated sites. Kim et al. [45] applied EK remediation at Cu- and Pb-contaminated sites using vertical electric fields powered by solar energy, with EDTA and citric acid as complexing agents to enhance metal ion extraction. In a field-scale application in Changshou, Jiangsu Province, Liu et al. [46] achieved in situ EK remediation of copper-contaminated electroplating soils, reducing heavy metal concentrations in wastewater to meet national standards and achieving a copper removal efficiency of up to 81%.

Based on the review of existing domestic and international research, it is evident that substantial efforts have been made to improve electric current density and removal efficiency in EK remediation, primarily through the development of advanced electrode materials. However, the persistent issue of electrochemical polarization during the EK process remains unresolved. Although many studies [25,26,47,48] have attempted to alleviate the focusing effect by adjusting the pH of the anolyte and catholyte or introducing chelating agents and acidic additives, these strategies have yielded only partial success. The challenge of hydroxide precipitation near the cathode, which blocks pore spaces and impedes ion migration, continues to limit the overall efficiency and consistency of EK remediation. To address these gaps, this study proposes the use of a novel hydrogel electrode for EK remediation of heavy metal-contaminated soils [25,26]. Field application of this electrode system requires a comprehensive setup, including a power supply unit, electrolyte pumping module, electroosmotic flow collection system, and an automated monitoring platform. Trenches are excavated on both sides of the contaminated site to accommodate the installation of hydrogel electrodes and electrolyte reservoirs, which are isolated from the soil matrix using geomembranes. The hydrogel-based EK remediation system can also be integrated with other remediation techniques to enhance the removal of Cu, Pb, Mn, and Zn, and is particularly effective for sites with high contamination levels (e.g., >500 mg/kg).

Despite recent advancements in electrokinetic remediation, key challenges remain unresolved, including electrode polarization, hydroxide-induced pore clogging, and limited ion transport efficiency [31,32,33,34]. These issues are particularly pronounced in loess contaminated with multiple heavy metals, where complex soil structure and chemical interactions further constrain treatment outcomes. In addition, the performance and mechanistic behavior of novel electrode systems, such as composite hydrogel electrodes, have yet to be systematically investigated under relevant geotechnical and environmental conditions. In light of this, the objectives of this study are (1) to establish a control group using conventional graphite (Gr) electrodes and electrokinetic Geosynthetics electrode (EKG) electrodes applied to Cu- and Pb-contaminated loess without any enhancement measures, and to compare EK parameters and removal efficiencies with those of the novel hydrogel (NH) electrode system; (2) design and evaluate three enhanced electrokinetic configurations, including focusing position adjustment, electrode exchange, and their combined implementation, in order to improve ion transport dynamics and spatial distribution of precipitation; (3) elucidate the underlying physicochemical mechanisms governing copper and lead migration in loess during electrokinetic treatment, and to propose a refined remediation strategy that addresses the coupled limitations of electrode polarization and the focusing effect.

2. Materials and Methods

2.1. Sampling and Specimen Preparation

In NW China, the formation of loess is primarily attributed to long-term aeolian dust deposition on the Loess Plateau, accompanied by leaching processes, sedimentary evolution, and wind-blown accumulation. Geomorphologically, the region features three typical landform units: loess ridges, loess hills, and bedrock-incised gullies, forming a landscape of considerable complexity. Borehole data indicate that the loess cover in this area is mainly composed of paleosol layers and overlying Malan loess (Q₃), both of which rest on sandstone bedrock. Their thickness ranges from 22 to 31 m for the paleosol and 2 to 5 m for the Malan loess [1,49]. The loess used in this study was collected from Lantian County, Shaanxi Province, approximately 22 km southeast of Xi’an. Samples were taken from a depth of 3.0 to 4.5 m. According to the Chinese Standard for Geotechnical Testing Methods [50], the undisturbed samples were tested for basic physicochemical properties (see

Table 1. Physicochemical properties of the loess.

Property	Loess
Sand (%)	3.3
Silt (%)	87.4
Clay (%)	9.3
Void ratio, e	0.898
Bulk unit weight, γ (kN/m3)	16.2
Specific gravity, Gs	2.69
Water content, ωn (%)	16.5
Liquid limit, ωL (%)	31.6
Plastic limit, ωP (%)	19.5
USCS symbol	CL
Permeability (m s−1)	2.55 × 10−6
Organic matter (mg g−1)	4.1
pH	7.8
Electrical conductivity (μs cm−1)	244
BET specific surface area (m2 g−1)	24.1
Composition of ions
Ca2+ (mg/kg)	126
Mg2+ (mg/kg)	40
Na+ (mg/kg)	103
K+ (mg/kg)	4.6

). All parameters of natural loess listed in Table 1 (including electrical conductivity, permeability, and organic matter content) were determined through laboratory measurements. Each value represents the mean of three independent tests, with the measurement error in all cases being less than 10%. The results show a specific gravity of 2.69 and a natural moisture content of 16.5%. Particle size distribution was determined using a laser particle size analyzer (WJL-602 model, Shanghai Jingke Technology Co., Ltd., Shanghai, China), revealing that the silt fraction accounts for 87.40% of the sample, with clay and sand contents of 9.30% and 3.30%, respectively (see Figure 1). These results indicate typical silty loess characteristics. The plasticity index was measured at 12.1%, calculated from the difference between the liquid limit and plastic limit. Additionally, a major ion analysis was performed using an IA-300 ion analyzer (Toa DKK, Tokyo, Japan) to identify soluble ionic species within the loess matrix. The results confirmed the absence of heavy metal ions such as copper (Cu) and lead (Pb), indicating that the sample remains uncontaminated. This supports its suitability as a model soil for artificially contaminated loess in subsequent remediation experiments. The sample provides a reliable basis for investigating the electrochemical migration and distribution behavior of Cu²⁺ and Pb²⁺ within loess pore media.

To prepare simulated heavy metal-contaminated loess samples, air-dried loess was uniformly mixed with aqueous solutions of copper nitrate (Cu(NO₃)₂) and lead nitrate (Pb(NO₃)₂), such that the initial concentrations of Cu²⁺ and Pb²⁺ in the soil were both controlled at 500 mg/kg. This level was intentionally chosen to exceed the control values defined in the soil environmental quality: risk control standard for soil contamination of agricultural land (200 mg/kg for Cu and 240 mg/kg for Pb) [51], so as to better represent high-risk contamination scenarios. The use of such elevated concentrations allowed us to investigate the remediation mechanisms and electrokinetic behaviors under more severe contamination conditions. All contaminated specimens were prepared following the same procedure to ensure reproducibility. This approach enables precise regulation of contamination levels, providing a controlled framework for systematically investigating heavy metal migration and removal mechanisms under laboratory conditions, and facilitating a deeper understanding of key influencing factors and optimization strategies in remediation processes. To ensure homogeneous distribution of the contaminants and consistent moisture content, the prepared Cu- and Pb-contaminated loess samples were sealed and cured in a humidity-controlled chamber for 72 h, with the final moisture content adjusted to 15%. Following curing, the samples were compacted into the test reactors at a dry density corresponding to a compaction degree of 0.80, with care taken to maintain consistent moisture content during the process. Prior to testing, all contaminated samples were divided into five equal portions and compacted in layers using a wooden block to ensure uniform density throughout, thereby meeting the requirements of the subsequent electrokinetic remediation experiments.

2.2. Materials

The chemical reagents used in the experiments included analytical-grade or chemically pure acetic acid (CH₃COOH), hydrochloric acid (HCl), and sodium hydroxide (NaOH). The heavy metal sources, Cu(NO₃)₂ and Pb(NO₃)₂, were also of analytical grade. All reagents were purchased from Macklin Biochemical Co., Ltd. (Shanghai, China) and Sinopharm Chemical Reagent Co., Ltd. (Xi’an, China), ensuring consistency and reproducibility of the experimental data.

2.3. NH Electrode Preparation

In this study, a novel composite hydrogel electrode was developed in-house, consisting of sodium alginate, graphene, and calcium chloride (CaCl₂). The detailed formulation is provided in Table S1. Sodium alginate, rich in carboxyl and hydroxyl functional groups along its polymer chains, provides abundant reactive sites for the coordination-based adsorption of heavy metal ions such as Cu²⁺ and Pb²⁺. Graphene enhances both the structural stability and mechanical integrity of the hydrogel through hydrogen bonding and electrostatic interactions with sodium alginate and Ca²⁺. Furthermore, its large surface area and abundant surface functional groups contribute to the effective adsorption and immobilization of heavy metals, offering excellent environmental responsiveness and application potential in engineering remediation. The hydrogel electrode was prepared using the following procedure (see Figure S1). First, a 2 wt% sodium alginate solution was prepared by dissolving sodium alginate in deionized water under continuous stirring at 3000 rpm for 60 min using a magnetic stirrer, resulting in a uniform and transparent solution. Subsequently, 7 wt% graphene powder was gradually added to the solution and stirred at the same speed for another 60 min to ensure homogeneous dispersion and integration with the alginate matrix. The resulting graphene–sodium alginate composite solution was then injected into an integrated polymethyl methacrylate (PMMA) mold (150 mm long × 7 mm wide × 150 mm high) for hydrogel formation. To initiate crosslinking, the filled mold was immersed in a 2 wt% CaCl₂ solution and left undisturbed for 12 h. During this process, ionic crosslinking between Ca²⁺ and the carboxyl groups on the alginate chains generated a stable three-dimensional polymeric network. Simultaneously, graphene nanosheets were embedded within the gel structure via hydrogen bonding and electrostatic interactions, further enhancing its mechanical properties and structural cohesion. It is important to note that the concentration of CaCl₂ is pivotal for achieving optimal crosslinking: both insufficient and excessive levels compromise network formation. After crosslinking, the hydrogel solidified into a monolithic structure conforming to the mold dimensions, which was then demolded and directly deployed as a functional electrode in subsequent EK remediation experiments [25,26]. The resulting NH electrode displayed robust mechanical integrity and precise dimensional conformity, ensuring reproducibility across specimens. Owing to its dual functionality, it acted not only as an active electrode but also as a sorptive medium, facilitating the removal of Cu²⁺ and Pb²⁺ during electrokinetic treatment. Importantly, the concentration of crosslinking agents was found to be pivotal: both insufficient and excessive levels impaired the structural stability and electrochemical performance of the electrode [25].

In addition to the experimental details described above, it is worth noting that our previous research has systematically investigated electrode design and the associated electrochemical mechanisms. In particular, we developed a graphene oxide–alginate composite hydrogel electrode, which was shown to enhance ion transport efficiency and improve electrode stability during electrokinetic remediation of Cu(II)-contaminated loess [25]. Furthermore, we proposed a nanocomposite hydrogel electrode capable of mitigating electrochemical polarization and alleviating the focusing effect in Cu- and Pb-contaminated loess, thereby improving overall remediation performance [26]. Building on these prior studies, the present work further explores the preparation and application of NH electrodes, aiming to validate their effectiveness under more complex contamination scenarios and to provide deeper insights into the design–mechanism relationship of advanced functional electrodes.

2.4. Electrokinetic Reactor and Experimental Design

EK remediation systems generally consist of two electrode chambers and a central soil compartment, where electrolysis produces H⁺ at the anode and OH⁻ at the cathode, inducing electroosmotic flow that drives contaminant migration [3,9]. Numerous laboratory-scale EK remediation devices have been reported with soil volumes typically <1000 cm³ [52,53,54], suitable for mechanistic studies but limited in scalability. In contrast, larger soil specimens (>1000 cm³) have been shown to better reflect in situ conditions, reproduce realistic electric field distributions, and yield more representative contaminant transport patterns [54,55,56]. They also improve the accuracy of monitoring current, electroosmotic flow, and temperature, thereby enhancing the engineering relevance of experimental results [57,58,59]. Based on these insights, this study designed and constructed a large-scale EKR apparatus to extend conventional laboratory systems toward pilot-scale applications and provide a stronger foundation for field-scale remediation of heavy metal-contaminated soils.

The EK remediation reactor system used in this study consists of the following main components: an integrated polymethyl methacrylate (PMMA) soil chamber, two electrode compartments, a pair of electrodes, two electrolyte reservoirs, two peristaltic pumps, a DC power supply, and a data acquisition system. A schematic of the apparatus is shown in Figure 2. The internal dimensions of the soil chamber are 500 mm in length, 150 mm in width, and 150 mm in height. Electrode compartments, each measuring 100 mm × 150 mm × 150 mm, are attached to both ends of the soil chamber. The soil specimen length was 30 cm, with an electrode spacing of 40 cm, resulting in an applied potential difference of 60 V corresponding to an electric field strength of 1.5 V/cm. These compartments are separated from the soil by composite barriers comprising perforated plates and geotextile fabric, which allow for efficient ionic exchange while preventing soil particle intrusion. The soil chamber is further divided into six equal monitoring sections, labeled S1 through S6, to track contaminant migration and changes in physicochemical parameters during the EK process. To maintain stable electrolyte levels in the electrode chambers, an overflow control system is incorporated. Each electrode compartment contains one electrode, and a constant voltage is applied using a regulated DC power supply to establish the electric field. The electrode chambers are connected to two electrolyte reservoirs (each with a diameter of 150 mm and a height of 250 mm), and continuous circulation of the electrolyte is maintained using peristaltic pumps. Throughout the experimental operation, the data acquisition system enables real-time monitoring of electric current, electrical conductivity (EC), pH, and electroosmotic flow (EOF) across all zones (S1–S6), allowing for the capture of dynamic responses during remediation. After each EK trial, the residual concentrations of Cu²⁺ and Pb²⁺ in each section are determined, and the removal efficiency is calculated accordingly to compare remediation performance under different experimental conditions. In addition, the NH electrode was connected to the power supply through a wire, with the wire attached to the power source using a binding post. At the electrode side, a crocodile clip was used to secure a small graphite piece (2 cm × 1 cm × 0.5 cm), which was inserted into the NH electrode to ensure stable electrical contact.

As summarized in

Table 2. Experimental design of the enhanced electrokinetic remediation of copper and lead-contaminated loess.

Test	Electrode Type	Pollutant	Voltage Gradient/V cm−1	Enhancement Methods	Duration Time/h
EK1	Gr 0.125	Cu + Pb	1.5	/	72
EK2	EKG	Cu + Pb	1.5	/	72
EK3	NH	Cu + Pb	1.5	/	72
EK4	NH	Cu + Pb	1.5	focusing position adjustment	72
EK5	NH	Cu + Pb	1.5	exchange electrode	72
EK6	NH	Cu + Pb	1.5	focusing position adjustment + exchange electrode	72

Note: ‘/’ indicates that the corresponding test was not used the treatment during EK remediation time.

, six EK remediation tests (EK1–EK6) were conducted to systematically evaluate the influence of electrode type and regulation strategies on contaminant migration behavior and removal efficiency. In the EK1 to EK3 groups, three different electrode configurations were employed: Gr electrodes (EK1), EKG electrodes (EK2), and NH electrodes (EK3). A constant electric field intensity of 1.5 V/cm was applied across all tests to compare the effectiveness of different electrode materials in mitigating electrochemical polarization and enhancing overall remediation performance. The results demonstrated that the novel hydrogel electrodes offered notable advantages in reducing polarization effects, improving electrode stability, and enhancing ion migration efficiency.

Building on this foundation, three additional test groups (EK4 to EK6) were designed based on the hydrogel electrode system, incorporating progressively enhanced regulation strategies targeting the focusing effect. In the EK4 experiment, a “focusing position adjustment” strategy was implemented. Specifically, a controlled amount of tartaric acid solution was introduced at two-hour intervals into the central section of the soil chamber (zones S3–S4). This approach aimed to artificially establish a pH buffering zone along the contaminant migration pathway, thereby inducing electroosmotic flow and contaminant accumulation toward the center. The resulting pH gradient was intended to enhance the central focusing effect and suppress ionic diffusion toward the chamber boundaries. In the EK5 experiment, an “electrode exchange” strategy was applied by periodically reversing the polarity of the DC power supply every 12 h, effectively switching the positions of the anode and cathode. This method was designed to alleviate electrode polarization and extend the operational stability of the system. Moreover, the periodic reversal of the electric field facilitated the back-and-forth migration of heavy metal ions, improving both their transport efficiency and spatial removal uniformity. In the EK6 experiment, both the focusing position adjustment and electrode exchange strategies were simultaneously applied based on the hydrogel electrode system. This configuration was intended to investigate the synergistic effects of the combined strategies on Cu²⁺ and Pb²⁺ migration pathways, spatial redistribution, and overall removal performance. Deionized water was used in both the anodic and cathodic reservoirs. In the EK5–6 electrode-exchange tests, the electrolyte was not replaced; instead, the original anode reservoir was reused to collect the electroosmotic flow after electrode exchange. Through comparative analysis of these configurations, their effects on migration pathway optimization, electric field uniformity, and current utilization efficiency were assessed. The findings aim to elucidate the enhancement mechanisms and applicable conditions of each strategy, providing a theoretical basis and technical support for optimizing EK remediation systems.

2.5. Analytical Procedure

All six EK remediation tests were conducted at a constant electric field intensity of 1.5 V/cm for a duration of 72 h. During each test, electrokinetic parameters were continuously monitored, and upon completion, the residual concentrations of Cu²⁺ and Pb²⁺ were quantified to determine the corresponding removal efficiencies (experimental design is summarized in Table 2). After each test, soil samples were collected from the six monitoring sections (S1–S6). The samples were oven-dried, ground, and sieved, after which 3.00 g of each sample was placed into a 50 mL centrifuge tube. Subsequently, 27 mL of 1.00 mol/L HCl solution was added, and the tubes were agitated on a shaker at 200 rpm for 6 h. The suspensions were then centrifuged at 4000 rpm for 15 min, and the supernatants were filtered through 0.22 μm membranes. The filtrates were diluted to appropriate concentrations and analyzed for residual heavy metal ion content using an atomic absorption spectrophotometer. The resulting solutions from the acid digestion were analyzed using an atomic absorption spectrophotometer (Hitachi, Tokyo, Japan). The removal efficiency of heavy metal ions through EK is calculated using the following equation:

Removal efficiency(%) = (C₀ − C_f)/C₀ × 100%

where C₀ is initial heavy metal concentration (mg/kg) of loess, and C_f is the final concentration of heavy metals (mg/kg) after EK treatment in loess.

Based on the EK1 and EK2 configurations, which served as baseline systems, EK3–EK6 were designed as enhanced electrokinetic remediation tests incorporating NH electrodes and targeted focusing effect mitigation strategies. EK3 employed the NH electrodes alone; EK4 built upon EK3 by introducing a focusing position adjustment strategy; EK5 incorporated an electrode exchange strategy into EK3; and EK6 combined both focusing position adjustment and electrode exchange strategies on the NH electrode system used in EK3.

3. Results and Discussion

3.1. Effect of Electrode Tapes

Figure 3 illustrates the temporal variations in current, electroosmotic flow, and soil temperature during electrokinetic remediation with different electrode types. The NH electrode significantly reduced electrode electrochemical polarization, resulting in higher current density and consequently a more pronounced electroosmotic flow. Under all three electrode configurations, the soil temperature exhibited an increasing trend over time, primarily attributable to Joule heating. For the NH electrode, a decrease in temperature was observed in the later stage of the test, mainly due to the reduction in current and the subsequent dissipation of heat through thermal conduction. The observed current decline after 30–40 h in the NH electrode test corresponded to decreased ion mobility and thermal dissipation, which explains the non-uniform heating trend (Figure 3a). Notably, the soil temperature in the NH electrode tests remained consistently higher than that in the Gr and EKG electrode tests. This difference is explained by the pronounced electrochemical polarization of the latter electrodes, which led to lower current levels and, consequently, insufficient Joule heat generation. The temperature increase observed across different soil sections during EK remediation is driven by ohmic losses arising from the ionic resistance of the soil. In the final stage of testing, the sharp decline in current reduced heat generation, and combined with heat dissipation through environmental exchange, this led to the gradual decrease in soil temperature.

Figure 4a compares the post-treatment pH profiles of loess under three electrode configurations. In all cases, the pH increased progressively from the anode to the cathode. This gradient is attributed to water electrolysis during the EK process, where the anode generates protons (H⁺) and the cathode produces hydroxyl ions (OH⁻). Driven by the applied electric field, H⁺ migrates toward the cathode, inducing acidic conditions near the anode, whereas OH⁻ migrates in the opposite direction, rendering the cathode-adjacent regions alkaline. In the EK3 test, which employed the NH electrode, higher current levels led to greater H⁺ and OH⁻ generation compared with the other two groups. Consequently, pH values near the anode (S1) were markedly lower, while those near the cathode (S6) were substantially higher. Notably, although the pH at S1 in EK3 was slightly lower than in EK2 and EK1, the pH values in S2–S6 were significantly higher, underscoring the strong acid-buffering capacity of loess, which mitigates acidification in the migration pathway. By contrast, EK2 and EK1 exhibited relatively uniform pH distributions from S1 to S6, reflecting their comparable current magnitudes. The slightly higher current in EK2 produced marginally lower pH values near the anode and slightly higher pH values near the cathode compared with EK1. Across all tests, the S4–S6 regions in EK3 exceeded the initial pH of loess, while S6 in EK2 and EK1 also reached elevated alkalinity. Such high-pH conditions promote the precipitation of Cu²⁺ and Pb²⁺ as hydroxides, which can hinder heavy metal removal near the cathode zone.

Figure 4b presents the post-treatment EC profiles of loess under the three electrode configurations. Soil EC reflects the concentration of water-soluble salts and thus the bulk electrical conductivity of the medium. Across all tests, EC exhibited a similar spatial trend: it increased from the anode side, reached a maximum in the mid-section, and subsequently declined toward the cathode. This pattern is governed by ion migration under the applied electric field. Near the anode, water electrolysis generates abundant H⁺, which promotes the desorption of metal cations from soil particles and enhances their electromigration toward the cathode. This leads to a progressive increase in EC from S1 to the peak region. The observed higher EC values in the NH electrode tests can be attributed to the enhanced electrochemical activity of the NH electrode, which sustained higher current levels during operation. As a result, the generation and migration of H⁺ and OH⁻ ions were more pronounced, leading to increased ionic strength and, consequently, higher EC values in the soil pore water. In contrast, the Gr electrode produced lower current densities due to stronger polarization, which limited ion release and maintained lower soil EC. However, approaching the cathode, the elevated pH favors hydroxide precipitation of migrated metal ions, a phenomenon consistent with the “focusing effect.” The accumulation of these precipitates not only reduces soluble ion concentrations but also clogs pore channels, impeding both electromigration and electroosmotic flow. Consequently, EC declines markedly from S3 toward the cathode. Among the tested configurations, the NH electrode (EK3) yielded the highest EC values, attributed to its higher current density, which intensified H⁺ generation at the anode and thereby enhanced metal desorption and ion transport. In comparison, EK1 and EK2 produced lower and more similar EC values, with EK2 slightly exceeding EK1 due to marginally higher current, resulting in correspondingly improved desorption and transport efficiency.

Figure 5 compares the removal efficiencies of Cu and Pb across the three electrode configurations. Overall, Cu exhibited consistently higher removal rates than Pb, primarily due to the stronger adsorption affinity of loess for Pb²⁺, which hinders its desorption and subsequent electromigration. This result is strongly related to the intrinsic characteristics of loess: it has a relatively high clay content and contains abundant mineral phases with strong sorption affinity for heavy metals, especially Pb²⁺. These properties make the desorption and migration of Pb²⁺ particularly difficult, thereby lowering the removal efficiency. In EK3, the use of the NH electrode generated the highest current density, which enhanced H⁺ production at the anode and promoted ion migration, thereby improving overall removal performance. However, the elevated current also increased OH⁻ generation at the cathode, facilitating precipitation of Cu²⁺ and Pb²⁺ in the cathodic region. This effect was particularly pronounced for Pb, leading to significant accumulation in S6, where the residual concentration exceeded pre-treatment levels—indicating localized re-pollution. In contrast, EK1 and EK2 achieved slightly lower overall removal efficiencies but maintained more uniform performance across S1–S6, without the excessive Pb accumulation observed in EK3. This spatial uniformity reflects better operational stability despite lower current levels. These results highlight a trade-off for NH electrodes: while capable of enhancing removal through increased current and ion transport, they also exacerbate cathodic precipitation under uncontrolled alkaline conditions. Effective pH regulation in the cathode zone is therefore a critical requirement for realizing the full potential of NH electrodes in EK remediation. The negative values of remediation efficiency observed in the S6 soil section (Figure 5b) indicate local accumulation of Pb²⁺ rather than net removal. This effect is mainly attributed to the focusing phenomenon that occurs near the cathode. During migration, Pb²⁺ ions move toward the cathode, but the elevated pH in the cathodic region induces precipitation of Pb compounds (e.g., Pb(OH)₂ or PbCO₃), which accumulate in the adjacent soil layers instead of transferring completely into the cathodic reservoir. Consequently, this results in apparent negative efficiency values in that section. The subsequent control strategies proposed in this work were designed specifically to overcome this limitation by mitigating precipitation and focusing effect near the cathode.

3.2. Effect of Regulation Strategies

As shown in Figure 6a, EK4–EK6 exhibited the relationships of electric current and accumulated electroosmosis versus time under different regulation strategies. In EK4, the focusing position adjustment of tartaric acid into the central zone caused transient increases in local ionic concentration, resulting in short-term current surges following each injection cycle. However, the average current level remained slightly lower than EK3 due to partial ion consumption in acid–base neutralization reactions. EK5, incorporating electrode polarity reversal, maintained a more stable current output over the 72 h operation. The polarity switching mitigated electrode polarization and slowed electrode passivation, reducing the steep current decay observed in EK3. EK6 demonstrated the most consistent current density across the test period, as the combined strategies simultaneously enhanced ionic availability in the central zone and prevented polarization at the electrodes, ensuring effective electric field utilization throughout the operation. EOF trends (Figure 6b) closely followed current variations. Figure 6b clearly shows that after 72 h, EOF values reached approximately 1300 mL for EK3, 1500 mL for EK4 and EK5, and 1900 mL for EK6, confirming the enhancement achieved with tartaric acid addition and electrode exchange. In EK4, the focusing position adjustment produced localized enhancement of electroosmotic velocity in S3–S4, while EOF in distal sections (S1–S2, S5–S6) showed moderate suppression due to altered hydraulic gradients. EK5 displayed a bidirectional EOF pattern in response to periodic polarity reversal, which redistributed pore water movement across the soil profile and minimized extreme moisture depletion near the anode or cathode. EK6 yielded the highest cumulative EOF volume, as the synergistic effect of focusing and polarity reversal both intensified central zone flow and prevented water stagnation at the electrodes.

Post-treatment pH sections (see Figure 7) revealed that EK4’s acid injection effectively created a central pH buffer (S3–S4 maintained near neutral), reducing the extent of acidic and alkaline extremes in the soil section. The beneficial effect of tartaric acid in EK4 is primarily attributed to its complexation/chelation ability rather than simple pH buffering. The deprotonated carboxyl groups of tartaric acid readily form complexes with Pb²⁺ and Cu²⁺ ions. As a result, the OH⁻ generated near the cathode is less likely to induce hydroxide precipitation, thereby improving ion migration and enhancing removal efficiency. EK5 achieved a flatter pH gradient from S1 to S6 due to OH⁻ and H⁺ alternately migrating under reversed polarity, which prevented excessive accumulation at either end. EK6 combined these advantages, showing both moderated electrode pH extremes and a central zone maintained in the optimal range for metal desorption and transport. EC distributions (see Figure 7) were consistent with ion migration dynamics. EK4’s central injection caused EC peaks in S3–S4, reflecting concentrated ionic flux. EK5’s polarity reversal flattened the EC curve, indicating uniform ion distribution. EK6 maintained elevated EC in the central zone while preventing severe ionic depletion or precipitation near electrodes, reflecting balanced migration and minimal pore blockage.

In the EK6 test group, the synergistic application of the “focusing position adjustment” and “exchange electrode” strategies effectively resolved the issue of metal ion accumulation in the central S3–S4 region observed when using the exchange electrode strategy alone (EK5) (see Figure 8). Specifically, the periodic introduction of tartaric acid into the S3–S4 region every 2 h not only established a localized acidic environment but also promoted the complexation of tartrate anions with metal ions, thereby markedly enhancing desorption and migration in this zone. As a result, the removal efficiency in the central section of EK6 was substantially higher than that of EK5. Moreover, compared with the focusing position adjustment strategy alone (EK4), EK6 achieved a particularly pronounced improvement in removal efficiency within the cathodic region (S5–S6). This enhancement is primarily attributed to the periodic reversal of polarity during electrode exchange, which alternately generates H⁺ at both electrode compartments, creating cyclical acidification in the soils at both ends (S1–S2 and S5–S6). This cyclic acidification persistently weakens the binding between metal ions and soil particles, thereby facilitating desorption and subsequent migration. Collectively, EK6 not only improved focused migration efficiency in the central region but also substantially enhanced removal performance in the cathodic zone, highlighting the synergistic advantage of combining these two control strategies in suppressing metal ion accumulation and boosting overall remediation efficiency.

3.3. Disscussion

As shown in Figure 6, the overall current in the EK4 group (focusing position adjustment) was slightly higher than that of the unregulated NH electrode group (EK3) (see Figure 3). This increase is primarily attributed to the localized acidification in the central section following tartaric acid injection, which markedly enhanced metal ion desorption and migration, thereby improving system conductivity. Each acid addition produced a modest current peak, followed by a gradual decline due to ion depletion and polarization effects. In the EK5 group (exchange electrode), the current exhibited a periodic oscillation pattern. After each polarity reversal, the metal precipitates (mainly hydroxides) that had accumulated at the former cathode were dissolved in the newly acidic environment, releasing metal ions and causing a sharp, transient increase in electrical conductivity, which resulted in pronounced current peaks. The EK6 group (focusing position adjustment + exchange electrode) combined both mechanisms. During each polarity reversal, not only were precipitates released, but the concurrent acidification in the central region further promoted desorption, yielding the largest current fluctuations and peak values among all groups. Notably, EK6 maintained relatively high current levels even in the later stages of operation, reflecting its sustained capacity to enhance ion migration. For electroosmotic flow, EK4 exhibited a short-term increase at each acid addition, driven by an enhanced osmotic pressure gradient in the pore fluid of the central zone, which promoted water convergence toward this region. EK5 displayed a flow reversal after each polarity exchange, followed by a brief surge to high flow rates. This reciprocating flow facilitated precipitate flushing and improved ion distribution uniformity. In EK6, the combined effects of polarity reversal and central acidification produced the highest flow peaks and the most frequent oscillations, indicating stronger hydraulic field perturbations that effectively disrupted ion migration bottlenecks and suppressed localized accumulation.

Regarding pH distribution, EK4 suppressed the alkalinization trend in the central S3–S4 zone through localized acidification; however, the cathodic region (S5–S6) still maintained relatively high pH values. In EK5, polarity reversal induced alternating generation and migration of H⁺ and OH⁻ at opposite ends, creating a cyclic acidification–alkalinization environment that mitigated the accumulation of extreme pH on a single side. EK6, integrating both measures, maintained a stable acidic environment in the central zone while simultaneously lowering the high pH values in the cathodic region via polarity alternation, thereby markedly reducing the likelihood of OH⁻ combining with Cu²⁺ and Pb²⁺ to form precipitates. In terms of EC distribution, EK4’s central acidification markedly increased the electrical conductivity in S3–S4. In EK5, each polarity reversal dissolved precipitates and released ions, leading to a rapid EC increase in the former cathodic region. EK6, benefiting from the superposition of both mechanisms, not only sustained high EC in the central zone but also elevated EC peaks at both ends, resulting in a more spatially uniform EC section.

For Cu removal, EK4’s focusing position adjustment significantly enhanced migration efficiency in S3–S4 but yielded limited improvement at the cathodic end. EK5’s polarity reversal improved removal at both ends (S1–S2 and S5–S6) but posed a risk of central accumulation. EK6, through the synergistic effects of acidification and polarity reversal, simultaneously increased removal efficiency in both the central and terminal zones, achieving the highest overall Cu removal rate with the most uniform spatial distribution. For Pb removal, the generally stronger adsorption of Pb in loess resulted in consistently lower removal efficiencies compared to Cu. EK4 moderately improved Pb removal in the central zone but had limited effect on suppressing cathodic accumulation. EK5 markedly reduced Pb accumulation at the cathodic end but did not significantly improve central removal. EK6, leveraging dual action from central acidification–desorption and polarity reversal, substantially decreased Pb deposition in both the central and cathodic zones, greatly enhancing overall removal efficiency—particularly in S5–S6, where Pb migration and removal were most effectively achieved. Taken together, the electrokinetic response, hydraulic field disturbance, pH/EC regulation, and heavy metal removal performance indicate that EK6 outperformed the other strategies. It not only resolved the localized accumulation issues inherent in single strategies but also delivered simultaneous improvements in both Cu and Pb removal, highlighting the significant advantages of strong coupled regulation in EK remediation. The enhanced removal observed in EK4 and EK6 tests can be attributed primarily to the chelation of Pb²⁺ and Cu²⁺ by tartaric acid rather than to its buffering capacity. At soil pH values of 7–8, tartaric acid is largely deprotonated, and its –COO⁻ groups effectively coordinate with Pb²⁺ and Cu²⁺ ions to form stable complexes. The stability and transport behavior of these chelated species are further influenced by the pH gradients that develop across different soil zones during electrokinetic remediation, which govern their speciation and mobility. It should be noted that during the EK5–6 electrode-exchange tests, there is a possibility that certain metal ions (e.g., Pb²⁺ or Cu²⁺) may be partially retained within the NH electrode due to intercalation, chelation, or other binding processes. Upon polarity reversal, such ions might not be completely released and could even undergo redox transformation (e.g., Pb²⁺ → Pb⁴⁺) inside the composite. While this potential migration and transformation mechanism was beyond the scope of the present study, it represents an intriguing direction for future research.

A comparison of treatment time, electric field intensity, and removal efficiency with relevant domestic and international studies is presented in

Table 3. Comparison of studies with the research about soil type and remediation efficiency.

Electrode	Contaminants	Initial Concentration C0 (mg/kg)	Time t (h)	Intensity (V/cm)	Soil Type	Removal Efficiency (%)	References
Aluminum electrode	Cr, Ni	186, 132	72	0.6–1.0	Sand	3.1–30.1, 27.3–48.5	[42]
Hydroel electrode	Cr	200	0.5	4.8	Sand	70	[59]
Composite electrode	Cu	200	60	1.25	kaolin	17.5	[56]
Graphite	Pb	120	168	1.08	Clay	14.15	[57]
Graphite	Cu, Pb	327.8 240.8	705	1.0	Black soil	94.84, 95.85	[52]
/	Cu, Cd	248.4, 82	240	1.0	sludge	4.59, 30.65	[61]
Graphite	Cu	1000	144	1.0	Kaolinite soil	27	[62]
NH electrode	Cu, Pb	500,500	72	1.5	loess	47.25, 16.93	This study

. Under identical initial concentrations and treatment durations, the use of a single regulation strategy in this study resulted in Cu and Pb removal efficiencies superior to those obtained with NH electrodes, as well as graphene (Gr) and electrokinetic graphite (EKG) electrodes [60,61,62]. Moreover, despite the higher initial concentrations and longer treatment times reported by Ouhadi et al. [63], the NH electrode system employed in this work demonstrated a distinct performance advantage. Similarly, Behrouzinia et al. [64] conducted a 60 h electrokinetic remediation of low-concentration Cu-contaminated clay; however, their removal efficiency was only approximately one-third of that achieved in this study. Overall, although the NH electrode system proposed herein exhibits notable potential and superior performance in removing Cu and Pb compared with current literature, the removal efficiency remains suboptimal. The application of combined regulation measures (EK6) further enhanced remediation performance, achieving Cu and Pb removal efficiencies of 47.25% and 16.93%, respectively. This represents a substantial improvement over existing domestic studies (Table 3), underscoring the efficacy of synergistic regulation in overcoming the limitations of single-strategy EK systems.

4. Underlying Mechanisms

4.1. Atomic Force Microscopy (AFM) Characterization

Figure 9 presents the Atomic Force Microscopy (AFM) analysis of the NH electrode. As shown in Figure 9a, the AFM image of monolayer graphene reveals a highly ordered lamellar structure with a smooth and uniform surface, curled edges, and a thickness of approximately 1 nm. Figure 9b displays the AFM image of multilayer graphene oxide (GO), which exhibits a thin, veil-like morphology characterized by a wrinkled and rough surface with slight folding. The average thickness is about 3.2 nm, corresponding to roughly three stacked GO layers. The three-dimensional AFM topography further indicates that the GO surface morphology consists of grooves and pores, which can be attributed to electrostatic repulsion among the functionalized oxygen-containing groups introduced during oxidation. The incorporation of GO particles into the NH electrode increases both the average pore size and the pore volume, thereby facilitating the adsorption of Cu²⁺ and Pb²⁺. Moreover, multilayer GO possesses a larger specific surface area, providing abundant active adsorption sites, and contains a high density of oxygen-containing functional groups—hydroxyl (–OH), carboxyl (–COOH), and epoxy (C–O–C) groups. These functional groups actively participate in the adsorption of Cu²⁺ and Pb²⁺ during electrokinetic remediation, effectively preventing the re-release of these heavy metals into the surrounding environment. Figure S2 shows the Fourier-transform infrared (FTIR) spectra of the NH electrode before and after electrokinetic remediation. Characteristic absorption bands above 1000 cm⁻¹ correspond to C–O–C stretching, while those at 1415–1631 cm⁻¹ arise from –COOH vibrations. A broad band at 3360–3394 cm⁻¹ is attributed to –OH stretching. These functional groups provide active coordination sites, enabling the adsorption and complexation of Cu²⁺ and Pb²⁺ during remediation.

4.2. Cyclic Voltammetry (CV) Characterization

Cyclic voltammetry (CV) is widely employed to investigate electrode properties and electrochemical parameters. In this study, CV testing was conducted to evaluate the electrochemical polarization behavior of the NH electrode. The electrode samples were trimmed to dimensions of 7.5 cm in diameter and 2.0 cm in height, and the measurements were performed using an electrochemical workstation (CHI660E, Shanghai Chenhua, China).

Figure 10 shows the CV curves of the NH electrode. The results indicate that the reversibility of electroactive species on the electrode surface, such as dissolved oxygen and hydrogen, is relatively poor, as evidenced by the noticeable difference between the oxidation and reduction branches within a single scan cycle. Furthermore, the CV curve deviates from an ideal elliptical shape, suggesting that the NH electrode does not exhibit ideal capacitive behavior. Only weak oxidation and reduction peaks are observed, differing substantially from the “duck head” shape typically recorded for conventional electrodes. This behavior can be attributed to the electrode’s non-Faradaic characteristics, implying that no significant charge transfer occurs at the NH hydrogel electrode–electrolyte interface. Consequently, the electrochemical polarization of the NH electrode is minimal or nearly negligible. These findings confirm that the NH electrode can effectively suppress electrochemical polarization effects during electrokinetic remediation, thereby improving the stability and efficiency of contaminant removal.

4.3. Electrochemical Impedance Spectroscopy (EIS) Analysis

Electrochemical impedance spectroscopy (EIS) is a widely used electrochemical characterization method that evaluates electrode properties by measuring the variation in internal resistance with respect to an applied sinusoidal frequency. In this study, EIS was employed to assess the electrochemical resistance characteristics of the NH electrode. Samples were trimmed to dimensions of 3.0 cm × 1.0 cm × 1.5 cm and tested using an electrochemical impedance analyzer (CHI660E, Shanghai Chenhua, China).

As shown in Figure 11, the high-frequency region of the Nyquist plot corresponds to the electrode resistance (R_p), the mid-to-high frequency region represents the electrolyte resistance (R∞), and the low-frequency region reflects the charge transfer resistance (Rct) and double-layer capacitance (Rmt). In the absence of significant electrochemical polarization, the high-frequency semicircle almost disappears, and the distance from the Y-axis to the semicircle onset remains below 7 Ω, indicating that the NH electrode exhibits extremely low R_p. The presence of R∞ in the mid-to-high frequency region further confirms that electrolyte resistance during electrokinetic remediation cannot be neglected. To improve clarity, Figure 11 now presents both the original experimental data (symbols) and the fitted curves (dotted lines), with the equivalent circuit model displayed in the inset. This revised presentation allows the distinction between raw spectra and model fitting, ensuring that the analysis is transparent. Notably, the NH electrode shows no 45° Warburg diffusion line in the low-frequency region, implying the absence of ideal capacitive behavior-a conclusion consistent with the non-elliptical cyclic voltammetry results. The relatively low Rct and Rmt values indicate high electrical conductivity, attributed to the sodium alginate binder forming a hydrogen-bonded crosslinked network with graphene oxide. This structure not only enhances the electrode’s mechanical integrity but also increases the density of active adsorption sites. Overall, the EIS results highlight the NH electrode’s minimal electrochemical polarization, low resistance, and high conductivity, underscoring its advantages in physical stability, mechanical strength, electrochemical performance, and adsorption capacity. These features collectively enhance its potential for stable and efficient electrokinetic remediation of heavy metal-contaminated soils.

5. Conclusions

This study evaluated the performance of a NH electrode and three regulation strategies—focusing position adjustment (EK4), electrode exchange (EK5), and their combination (EK6)—for the EK remediation of Cu- and Pb-contaminated loess. The results integrated electrochemical characterization (AFM, CV, EIS) with EK performance metrics (current, electro-osmotic flow, pH/EC distribution, and removal efficiency) and were benchmarked against existing domestic and international studies. The main conclusions from the results and discussion are as follows:

(a)

The NH electrode exhibited highly ordered layered graphene structures, abundant oxygen-containing functional groups, negligible electrochemical polarization, low internal resistance, and high conductivity. These properties enhanced ion transport and adsorption, enabling higher Cu²⁺ and Pb²⁺ removal efficiencies compared to conventional Gr and EKG electrodes.

(b)

Focusing position adjustment (EK4) effectively mitigated alkalization and ion accumulation in the mid-section (S3–S4), improving Cu²⁺ removal in that zone. Electrode exchange (EK5) alternated acidic and alkaline conditions at the soil ends, reducing metal precipitation in cathodic zones (S5–S6) and enhancing removal there. However, each single strategy improved only specific regions, leaving other zones suboptimal.

(c)

The EK6 strategy integrated focusing position adjustment with electrode exchange, maintaining a stable acidic environment in the mid-section while lowering cathodic pH and sustaining high current and electro-osmotic flow. This approach eliminated localized Cu²⁺ and Pb²⁺ accumulation, improved removal uniformity, and achieved the highest overall efficiencies (47.25% for Cu²⁺ and 16.93% for Pb²⁺), outperforming both single-strategy groups and the literature-reported benchmarks under comparable conditions.