Membrane-based extraction of critical materials from spent batteries and other mineral wastes

1 Introduction

The world of technology has experienced a steadily growing demand for crucial materials such as cobalt, lithium, and nickel for battery production (Khabas et al., 2024; Maisel et al., 2023). The high demand for battery production can be attributed to economic growth (e.g., grid stability, innovations), modern technologies (e.g., electric vehicles, consumer electronics), and sustainable energy (e.g., energy storage for renewables, energy security), with electric vehicles (EVs) and the energy decarbonization sector as the primary drivers (Bridge and Faigen, 2022). All these efforts aim to reduce dependence on greenhouse gas (GHG) emissions, which have significantly contributed to global warming. The United States and China have been the largest contributors to global warming (Author anonymous, 2021; Wu et al., 2022), with China, the leading country in the EV market (Li et al., 2022; Xue et al., 2021), pledging to reduce GHG emissions by 60%–65% by 2030. Consequently, China consumes approximately 50% of the world’s lithium (Hao et al., 2017; Li et al., 2024).

With this global surge in demand, the increased mining of these valuable materials generates substantial mine tailings, posing significant economic and environmental challenges. These mine tailings contain heavy metals and chemicals, and their improper disposal leads to water contamination, biodiversity loss, and soil degradation, with water contamination, particularly acid mine drainage, being the primary issue, as it severely affects the water quality across large areas (Pei et al., 2023; Araujo et al., 2022; Okewale and Grobler, 2023; Karaca et al., 2017). The economic challenges of mineral waste disposal include high recycling costs and a lack of infrastructure. To address these issues, research and development of sustainable extraction technologies for recovering residual metals from mine tailings could be economically viable, reduce reliance on raw materials, and promote a circular economy. These advancements would lower costs over time while minimising environmental damage.

This review is critically evaluating the application of membrane processes in the extraction of valuable materials from spent batteries and secondary mineral waste streams. To ensure robust literature review, the scope of the review is discussing interconnected domains, including battery chemistries, leachate types, waste streams, membrane process families, and quantitative performance metrics. Membrane technology is preferred over conventional techniques, including hydrometallurgy, solvent extraction, and pyrometallurgy, due to its numerous advantages, such as high selectivity and efficiency for specific ions, simultaneous recovery of water and metals, low chemical and energy consumption, and its compact system design (Goh et al., 2022; Chowdhury et al., 2021; Naim et al., 2021; Bera et al., 2022). The conventional techniques, on the other hand, pose considerable environmental threats, long processes, low recovery efficiency, high energy consumption, and storage (Dar et al., 2025; Liu X. et al., 2025). The leachate types considered leach solutions generated from hydrometallurgical routes, including ammonia leaching, acid leaching, with attention to feed chemistry parameters like pH, anionic strength, competing ion presence, and target metal concentrations. The membrane technologies discussed include pressure-driven processes (microfiltration (MF), ultrafiltration (UF), nanofiltration (NF), and reverse osmosis (RO)), electrically driven membranes (electrodialysis (ED), selective electrodialysis (SED), bipolar membrane ED (BMED)), concentration-driven membranes (FO, membrane distillation), as well as emerging hybrid systems such as supported liquid membranes. Briefly, membrane-based extraction is any separation process in which semi-permeable membrane acts as selective barrier for promote the recovery, concentration, or purification of valuable metals from hydrometallurgical leachates. For quantitative metrics, we standardized evaluation according to key performance indicators including selectivity and separation factors, rejection rate (%), permeate flux, antifouling as indicated by flux recovery ratio (FRR), and concentration factor. This review also discusses mechanisms of membranes selectivity, relevant case studies, current challenges, and future advancements.

2 Overview of battery and mineral wastes

The composition of battery and mineral waste is complex and varies depending on the battery chemistry and the materials used in battery production. For instance, lithium-ion batteries (LiBs) contain materials such as lithium iron phosphate, lithium cobalt, lithium nickel cobalt aluminium oxide, nickel manganese cobalt oxide, copper, aluminium, plastics, and electronics (Daniel et al., 2003; Huang et al., 2011; Dobó et al., 2023; Kaya, 2022). These materials can be recovered not only from end-of-life batteries but also from industrial by-products and mine tailings. However, the complexity and toxicity of these wastes make recovery challenging. Major challenges to recovering these materials include low concentrations of target metals, complex tailings composition, the need for advanced and adequate technologies, environmental hazards, and high costs (Luo et al., 2025; Kumar et al., 2025; Kasri et al., 2024). Efficient recovery of the critical materials would reduce reliance on pure metals from mining while mitigating the production and environmental impacts of mine tailings disposal. In addition, extended exposure to metals such as Co and Ni has negative effects on human beings; these include damage central nervous system and lung cancer (Mitra et al., 2022).

3 Taxonomy of aqueous streams for critical material recovery

With a vast difference in the origin, content and challenge of the target aqueous streams, a comprehensive taxonomy of these streams is necessary for better selection and a logical design of membrane-based separation methods for crucial materials. Aqueous resources inclusive of acid mine drainage (AMD), battery leachates, salt-lake brines, and industrial effluents are the four main (see Figure 1) and mostly studied groups into which the aqueous streams filled with the critical materials can be classified into. A concentrated complex cocktail of high-value metals (Li, Co, Mn, and Ni) in a low pH, high salinity matrix is the hallmark of battery leachates as they present highly variable chemistries and complex separations that complicate selective recovery at scale. In addition to the metals, these leachates mostly contain large quantities of organic contaminants from binders and electrolytes which then increase the complexity in their treatment and the recovery of the critical materials. The overall high ionic strength of the leachate makes it difficult to extract essential elements like Li and B from salt-lake brines and other naturally occurring geogenic brines and nearly neutral hypersaline solutions dominated by Na⁺, K⁺, Mg²⁺, and Cl⁻ (Pathak et al., 2023; Aguiar et al., 2016).

AMDs on the other hand possesses an uncontrolled extremely high acidic (pH below 3) nature. This group of waste stream is produced by the oxidation of sulfide minerals. It contains a partially diluted mixture of heavy metals (Al, Fe, Cu, and Zn), sulfate and rare earth minerals. This composition is environmentally dangerous, extremely variable and unstable. Lastly, industrial effluents containing a vast range of process water (from processes such as electroplating, electronics manufacturing and catalyst recycling) often contains toxic chemical matrices made of organics, complexing agents, and other process additives (Zhu et al., 2025; Jiao et al., 2023). Although all these taxonomies are sources of essential materials, their stark variations in properties such as their acidity, ionic strength, metal concentration ratios, and competing impurity profiles necessitate highly customized membrane extraction techniques that go beyond one-size-fits-all approaches.

With each of these taxonomies having unique compositional profiles and extraction difficulties, battery leachates (produced when used batteries are hydrometallurgically processed) are rich in lithium and transition metals and are complicated by organic residues and acidic conditions necessitating the use of highly selective membranes; brines requiring membrane that are scalable and ion selective as they are mostly dominated by NaCl and contains diluted amounts of lithium, boron, or magnesium; the extremely acidic, high sulfate, and dissolved metals containing AMDs which results from the oxidation of the sulfide mineral; and the diverse range of industrial effluents, from textile streams containing dyes and surfactants to electroplating wastewaters enriched in Cr and Cd, all necessitates flexible membrane chemistries and thorough pretreatment for effective water treatment and valuable material recovery (Azmi et al., 2025). Table 1 indicates the ways membrane technology can be used for making all four streams secondary sources of important materials.

TABLE 1. Categorizes wastewater streams (AMD, battery leachates, brines, industrial effluents) by origin, composition, challenges, membrane opportunities, and example membranes for selective ion recover.

4 Membrane integration in battery recycling flows sheets

When Li-ion batteries are recycled, membrane processes are integrated into the hydrometallurgical route, which has considerably high recovery rates and reduced energy use compared to pyrometallurgical techniques. The overall flowsheet generally includes collection and sorting of spent batteries, dismantling and shredding to produce black mass, leaching with an acid to dissolve metals into a leachate, purification and separation of the leachate, and recovery of metal salts and hydroxides through precipitation or other methods. Membrane processes are commonly used in purification steps, where they selectively separation ions, concentrate valuable materials, recover water or acids, and remove impurities. This positioning enables the membrane to handle complex leachates produced upstream. For instance, the pressure-driven membranes are used after mild acid digestion or leaching of disassembled battery components. Ultrafiltration process usually removes large colloids or particulates, while NF separates monovalent ions from multivalent ions, and RO concentrates solutions and recovers water and acids. Electro-driven membranes are used post leaching for ion-specific separations. Additionally, membrane-supported solvent extraction processes are integrated into hybrid hydrometallurgical processes. Thus, the membrane is used to promote solvent-based extraction of metals such as Mn, Co from leachates, thereby enhancing selectivity (Davis and Demopoulos, 2023; Gao et al., 2024; Mishra et al., 2025).

However, it is critical to focus on separations that are realistically achieved in acidic sulfate/chloride leachates. Acidic leachates from LiB recycling are usually produced from sulfate-based or chloride-based at pH less than 2, which contains a mixture of monovalent ions and multivalent ions. Due to their different sizes, mobility, and charge, membranes can effectively separate target materials. Membrane performance is facilitated by leachate chemistry (Nguyen and Lee, 2020; Aaltonen et al., 2017; Zhao Z. et al., 2025). Table 2 summarizes the main realistic separations reported.

TABLE 2. A summary of the main realistic separations.

Leaching routes in hydrometallurgical recovery of valuable materials from LiBs remarkably affect downstream membrane performance and this is due to variations in solution chemistry. Through these routes, stability, impurity profiles, and metal flow, the membrane choice and pretreatment needs are impacted (Zhang et al., 2024; Qadir et al., 2025). Alkaline leaching routes (NaOH, NH₄OH) are less common for Li-ion batteries but generate amphoteric species (Al, Zn) that complicate downstream separations and can foul membranes through precipitation. The presence of oxidants (H₂O₂, persulfates, hypochlorite) accelerates polymer degradation, this necessitate either ceramic membranes or robust pretreatment to quench oxidants. Acidic sulfate leaching (using H₂SO₄, often with reductants such as H₂O₂ or Fe²⁺) produces highly ionic solutions dominated by SO₄ ^2- complexes of Co, Ni, and Mn, with Li⁺ remaining largely as free ions. This favours negatively charged NF membranes for lithium enrichment, but precipitation risks (e.g., Fe³⁺ forming hydroxysulfates) demand pretreatment to avoid scaling. Chloride leaching (HCl, sometimes with oxidants such as Cl₂ or NaClO) yields metal chloride complexes, which are more soluble but can pass through NF membranes more easily, reducing selectivity; here, positively charged membranes or ED units are better suited for acid recovery. Speciation also dictates precipitation risks, for instance, Mn²⁺ and Fe²⁺ oxidize to insoluble oxides/hydroxides, while Al³⁺ readily forms gels, all of which can block membrane pores. Thus, the leaching chemistry sets the pH, ionic strength, and fouling potential, which in turn determines whether membranes can be applied directly or require upstream conditioning (neutralization, oxidation control, solid removal). By explicitly linking leachate chemistry to membrane choice and pretreatment, the flowsheet highlights that membranes are not standalone solutions but must be tailored to the chemical environment established at the very first step.

5 Overview of membrane technologies for extraction

Charged membranes affect ion transport via electrostatic interactions and the trend depends on whether the membrane surface is negatively or positively charged (Xu et al., 2025; Liu K. et al., 2025). A positively charged membrane repels cations such as Ni²⁺, Mn²⁺, Co²⁺ while favouring anions such as sulfate and chlorides. In chloride leachates, a positively charged NF membrane would allow Cl⁻ to permeate while retaining transition metal ions, which can be exploited for recovery of acid. However, the realistic pattern is that co-ions with the same charge as membrane surface are excluded, while counter-ions are enriched. As a result, selectivity diminishes at high ionic strength. This could be the reason lithium recovered from multivalent-rich leachates is feasible with negatively charged NF membrane. Nonetheless, acid recovery is effective with positively charged anion-exchange membranes (Yun and Kwak, 2020; Liu et al., 2022). In contrast, a negatively charged UF or NF membrane repels co-ions while cations pass through. For instance, in acidic sulfate leachates, a negatively charged NF membrane strongly retains sulfates, and this is attributed to Dannon exclusion, and reject divalent cations due to electrostatic interactions and size exclusion, while allowing partial permeation of Li⁺ (Zeng et al., 2025; Tanninen et al., 2004).

Table 3 summarizes the important characteristics of membrane separation technologies for extracting valuable components from an aqueous environment. These membrane technologies’ various separation techniques result in a variety of technological features and application areas. Below is a more thorough description of the use and significance of these membrane separation technologies.

TABLE 3. Membrane-based technologies for valuable materials recovery from aqueous environment.

5.1 Pressure-driven membranes

Ultrafiltration and nanofiltration are mostly used for the separation of macromolecules and colloidal particles. NF membranes are capable of rejecting divalent and multivalent ions and small organic molecules. NF coupled with UF is used often for the recovery of metals like Li and Co. In metal recovery, NF is mostly valuable for its ability to separate heavy metals like Li⁺ ions from magnesium (Mg²⁺) and calcium (Ca²⁺) impacted brines or battery leachates. It is also able to selectively remove Li⁺ from a mixture of salt cations. UF often serves as a pretreatment to remove suspended solids and prevent fouling in subsequent membrane processes. Advantages of using these two are that they require less energy for operation when compared to thermal processes and also have higher selectivity. However, challenges such as membrane fouling, limited monovalent ion rejection in NF, and chemical sensitivity in UF persist (Xu et al., 2025; Zhang L. et al., 2025).

Gherasim et al. used a polyamide NF membrane, AFC 40, for the recovery of Co(II) ions from aqueous solutions. To assess the membrane performance, the ion rejection and flux were analysed. The results showed that the membrane had a relatively high permeability, and there was a high adsorption or retention of Co ions on the membrane surface as the net charge of the membrane increased after the filtration process.The increase in net charge is attributed to the increased zeta potential values when there was an increase in the solution concentration for the same pH value. This suggested that cobalt adsorption occurred at increased concentrations because specific adsorption was enhanced by an increase in the ionic strength. Also, at low pH, the high number of protons in the solution can also contribute to the positive membrane charge. The maximum metal ion rejection recorded was 97% when a 100 mg Co/L feed solution was used (Gherasim et al., 2015). Using both commercially available and experimental NF membranes, Tanninen et al. studied the retention efficiency of copper from an acidic aqueous feed. Several NF membranes, NF 270, Desal-5 DK, Desal KH, BPT-NF-1, and BPT-NF-2, were tested for their separation efficiency and stability when a solution containing 25 g/L CuSO₄ was filtered for months. Commercial NF membranes showed good selectivity, retaining most of the copper sulphate. The Desal KH membrane showed better copper retention values of 92%–95% (Tanninen et al., 2004). To effectively recover Li ions from hydrometallurgical leachates, Roa et al. used NF processes. They used four commercial polymeric NF membranes, Desal 5DL, AMS 3012, Desal KH, and AMS 3014. From their study, Desal DL showed an outstanding performance, reaching 77% Li in Mg(OH)₂ (Roa et al., 2025).

Reverse Osmosis is mostly used for the separation of dissolved salts and ions from water. This has resulted in its application for the production of ultrapure water and the concentration of metal solutions. In metal recovery, RO is often integrated with NF or ED to concentrate metals such as Co and Li that are dissolved in water. The increasing reliance on LiBs increases the need for efficient remediation of metal-impacted wastewater (WW) for the recovery of Li ions and water (Zhu et al., 2025). Liu et al. studied the use of polyamide polymeric membranes for the recovery of both Li ions and water. They observed that despite the increasing application of RO membrane filtration for this purpose, pristine PA membranes showed lower water flux and high fouling susceptibility, which then lessened their treatment efficiency. They developed RO membranes, namely PA-M1, PA-M2, PA-M3, and PA-M4, from the pristine PA membrane (PA-M0). The modified PA membranes were able to maintain Li⁺ retention above 94.0% (Liu K. et al., 2025). Similarly, Rodríguez-Alegre et al. treated acid mine drainage (AMD) impacted with dissolved metals using a treatment system packed with RO for water reclamation and metals concentration. It was observed that 97% of the water was recovered, and around 60% of the targeted metals (Al, Zn, Cu, Mn, Mg, and Ca) were selectively recovered (Rodríguez-Alegre et al., 2024).

5.2 Electrically-driven membranes

Electrodialysis is an electrically driven separation process using alternating cation- and anion-exchange membranes (Figure 2). When a voltage is applied, ions migrate toward oppositely charged electrodes, passing through selective membranes to form concentrated and dilute streams. ED is especially effective for selective ion recovery, such as the extraction of Li from battery leachates. It can distinguish between monovalent and multivalent ions by using selective membranes. The process is energy-efficient for low-salinity solutions and allows for selective recovery. However, its efficiency decreases at higher ionic strengths, and the membranes are prone to fouling by organic matter or particulates (Arana Juve et al., 2022).

This membrane technique relies on the movement of charged ions through membranes that are charged membranes. In this technique, the charged ions move toward oppositely charged electrodes when a voltage is applied, forming concentrated and diluted streams as they pass across selective membranes. ED is used widely for selective ion recovery for the extraction of ions like Li from battery leachates. The membranes packed in the ED system selectively differentiate monovalent from multivalent ions. With increased industrialization, critical metals such as Mn and cobalt Co are found in WW. This has led to the development of sustainable and effective solutions to recover these elements in reusable forms (Ruya et al., 2025). To recover over 66% Fe ions, Chekioua et al. used a wide range of different polymer-based membranes at different experimental parameters for removal in acid-impacted streams (Chekioua and Delimi, 2015). Benvenuti et al. reported that over 95% of Ni was extracted and concentrated using ED. From their experiments, both Ni ions and purified water were obtained. With the wide application of ED for valuable metal recovery, there are few challenges for its application in lithium recovery from industrial spent LiBs leachate (Benvenuti et al., 2014). This is due to the complexity of the LiBs leachate. To achieve improved Li recovery using ED, Xing and Srinivasan successfully separated Li ions from industrial spent LiBs leachate using an electrodialyzer equipped with a bipolar membrane module. From this study, the highest recovery rate and metal purity under optimum conditions reached up to 99.4% and 64%, respectively (Xing and Srinivasan, 2023).

5.3 Other (MD, liquid membranes, hybrids)

Membrane distillation is a membrane process that is thermally driven. This technique uses a hydrophobic membrane to effectively separate non-volatile solutes from water. The vapour that is generated permeates through the membrane, driven by the temperature pressure difference, and condenses to water droplets on the cooler side of the MD system. MD is used widely for concentrating metal-containing WW, especially in zero-liquid discharge (ZLD), and it promotes water reuse (Yun et al., 2023). It has been reported that MD is used mostly for leachate treatment to promote the concentration of metals in water streams and result in the recovery of valuable metal ions. Rasoulnia et al. used MD to obtain a volume concentration factor of 45 in LiCl solution (Rasoulnia et al., 2025) and several other metal ions, such as rare earth metals (Ge et al., 2023). Shaheen et al. developed a photothermal air–gap MD system to remove heavy metal ions from WW. From this study, over 97% of zinc ions were recovered. Furthermore, the best-performing membrane at optimum conditions was used for the treatment of synthetic heavy metal water, and the highest rejection rates for most of the tested heavy metals, inclusive of Zn²⁺, Ni²⁺, Cu²⁺, Fe³⁺, and Cr^3+, were observed (Shaheen et al., 2022).

Supported Liquid Membranes (SLMs) is a technique that involves the utilization of a selective liquid extractant supported onto a semi-porous membrane support. The selective extractant binds target metal ions selectively, and these ions then permeate through the membrane substrate and are collected in the receiving phase. SLMs have received attention for application in selective extraction of metals such as Co, Ni, Li, or rare earth elements from dilute or complex solutions (Szczepański et al., 2021; Imdad and Dohare, 2022; Zhao et al., 2024). This technique offers high selectivity owing to the specific tailoring of the extractants. To obtain extraction and stripping percentages of 85.7% and 83.2% for Ni²⁺ ions was achieved by Srirachat et al. from an alkaline solution (Srirachat et al., 2021).

5.4 Target separations (Li/Mg, Li/Co-Ni-Mn, Co/Ni, removal of Al/Fe/Cu impurities, sulfate/chloride management)

This section focuses on the separation of key targets: lithium from magnesium in brines, lithium from Co-Ni-Mn in battery leachate, and cobalt from nickel from different streams, the removal of impurities such as Al/Fe/Cu, and the management of sulfate and chloride anions to improve process efficiency. Membranes offer versatile approaches for these complex chemical mixtures, as they require precise functionalities, selectivity, and stability tailored to specific ion sizes, complexation affinities, and surface charges. The separation of these materials, however, comes with drawbacks, including fouling, high energy demands, and decreased selectivity at high concentrations. Some reported membrane processes that are effective for these target separations include electrodialysis, NF, and supported liquid membranes (Nowik-Zając and Sabadash, 2025; Abdi et al., 2024). For the separation of lithium from magnesium, NF process and ion-selective tailored membranes offer better efficiency by exploiting their differences in charge densities, coordination chemistry, and hydration energy, as it is particularly difficult to separate these elements due to their similar ionic radii. The governing separation mechanisms are charge exclusion and size-based sieving. Typically, NF and ion-selective membranes possess negatively charged surfaces, which generate electrostatic repulsion that hinders the passage of divalent ions (Mg²⁺) while allowing monovalent ions (Li⁺) to permeate (see Figure 3). This behaviour arises from stronger interactions between water molecules and Mg²⁺, which reduce its passage through the membrane. On the other hand, Li⁺ has a weaker hydration layer and lower charge density, making it permeation easy (Rosentreter et al., 2025; Liu X. et al., 2025).

For Li separation from transition metals such as Co, Ni, and Mn, commonly found in cathode leachates, the same mechanisms generally apply. Transition-metal ions tend to interact more strongly with modified membranes, promoting the selective permeation of Li while retaining divalent ions. This enables high efficiency recovery of lithium in Li-enriched effluents. The separation of Co and Ni is more challenging due to their similar hydrated radii and divalent nature; therefore, simple charge exclusion is insufficient. Supported liquid membranes are preferred for this process, as they incorporate selective barriers capable of preferentially binding one metal ion to form a reversible metal-carrier complex (Low and Wang, 2021; Zhang N. et al., 2025; Zhao Y. et al., 2025). Removing impurities from the leachate before separating valuable materials is of importance because impurities affect selectivity, product quality, process efficiency and causes membrane fouling. Multivalent impurities such as Al³⁺, Fe³⁺, and Cu²⁺ are removed through surface complexation and charged-based exclusion, driven by their strong hydration shells and high charge density. Fe³⁺ and Al³⁺ readily hydrolyse to form hydroxo complexes, which are easily rejected through electrostatic repulsion and size exclusion, as shown in Figure 4. Adsorptive membranes are particularly effective for Cu²⁺ rejection. These processes collectively enable the effective and maximal recovery of lithium (Peng et al., 2019; Klaehn et al., 2023; Ayora et al., 2019). Additionally, managing sulfate and chloride ions presents in leachates from lithium-containing brines and battery recycling streams is also critical, as these anions can corrode equipment and increase osmotic pressure. Proper management in membrane processes improves Li recovery and reduces membrane fouling. Different membrane processes handle these anions differently. For instance, NF can effectively reject multivalent sulfate ions through charge exclusion, while monovalent chloride ions can pass through. RO integrated with ED can also be used, though it can be costly. Therefore, pretreatment steps are often implemented to improve Li recovery and overall process efficiency (Li et al., 2019; Aprilianto et al., 2024; Tian et al., 2024; Xu et al., 2020).

6 Applications in battery and mineral waste treatment

While numerous studies report high rejection or recovery efficiencies for valuable materials, the transition of these results to industrial feasible systems depends not only on initial separation performance but also on fouling resistance, chemical stability, selective durability, and downstream product purification necessities. The following case studies are critically discussed with respect to what is generalizable, what is system-specific, and what constitutes the true rate- or cost-limiting step.

Kumar et al. (2022) developed a membrane-integrated hybrid system combining pretreatment, hollow fiber (HF) UF, HF NF, and Li precipitation using K₂CO₃ for the recovery of Li, Co, Mn, and Ni from cathode-derived leach liquor. The recovery of Li achieved at 70 °C using K₂CO₃ was 88.2%, while the purity was 99.5 wt%. Their recovery system showed resistance to chemical changes of the leach liquor, and this was ascribed to the mixed cathode material of the spent LiBs. In addition, the NF membrane (VNF2) exhibited rejection of 94.6, 92.5, and 95.8% for Co²⁺, Ni²⁺, and Mn²⁺, respectively. The rapid recovery was attributed to the enrichment of bivalent metal ions. The study demonstrates a practical integration strategy, i.e., membrane process followed by chemical precipitation. This hybrid approach reflects a realistic industrial pathway where membranes serve as selective concentration steps rather than final purification units. NF-based enrichment of monovalent Li⁺ through preferable retention of divalent ions is widely applicable to Li containing acidic leachates. One of the limitations encountered was frequent energy pumping and membrane cleaning.

Wang et al. (2022) reported polymer inclusion membrane electrodialysis (PIMED) using PVDF-HFP-based polymer containing selective carriers for Li(I) and Co(II) transport. Strong carrier-polymer interactions (C-F bonding) improved membrane stability and promoted high recovery (Li(I): 99.1%, Co(II): 99.9%) with relatively low energy consumption and reduced GHG emissions (3.1 kg CO₂e kg^-1 Li). The study demonstrates a shift from charge or size exclusion to facilitated transport, which enhanced selectivity beyond Donnan-based NF membrane. The use of ED promotes reduced hydraulic energy demand compared to high pressure NF/RO processes. Additionally, carrier-mediated transport offers a robust route for high selectivity in multi-ion systems. However, these systems often suffer from selectivity collapse over time due to carrier depletion.

Gao et al. (2024) evaluated commercial NF270 and DK membranes under harsh acidic conditions (HCl solution, pH 1 for 0, 7, 14, and 30 days). While both membrane processes exhibited strong retention of divalent ions (96.5%), NF270 showed substantial performance degradation (retention declined from 99.1% to 84.3%). This study addressed an overlooked issue of long-term durability under realistic leachate conditions, as most studies focus on initial rejection. While polyamide NF membranes are vulnerable to acid hydrolysis and systematic rejection decline, particularly at pH <2, the exact degree of degradation is not generalizable, as it depends on temperature, acid type, ionic strength. Thus, the short-term lab tests cannot accurately predict multiple month industrial performance.

Feng et al. (2025) incorporated sulfonated covalent organic framework (COF-SO₃H) nanosheets into an NF membrane to improved acid resistance and Li selectivity. The purpose of the study was to address acid instability through nanofiller incorporation and charge tuning. Initially, Li rejection was moderate (43.7%), while divalent metal ions (Co²⁺, Ni²⁺, and Mn²⁺) were exceeding 90%. However, after 7–14 days in acid, retention performance declined drastically. The incorporated nanofiller could tune diffusion pathways and charge density, and it also improve the initial acid resistance. However, because interfacial compatibility is system-dependent, acid exposure can degrade both the polymer matrix and the critical filler–matrix interfaces. Table 4 shows the studies of the recovery of crucial metals from spent batteries using membrane processes.

TABLE 4. Recovery of crucial metals from spent batteries and other sources.

PA, Polyamide; SED, Selective electrodialysis; COF, Covalent organic framework; TAD-TBMB, 1, 4, 7, 10-Tetraazacyclododecane−1, 3, 5-Tris (bromomethyl)benzene; TFCM, Thin film composite membranes.

Although many studies achieve high rejection or recovery values of over 90%–99% (Table 5), these metrics alone do not demonstrate industrial readiness. Most studies rely on simulated leachates and short-term experiments, whereas real battery-derived streams are highly acidic and rich in competing ions and scaling precursors. Under such conditions, the primary constraints shift from intrinsic selectivity to long term chemical stability, energy consumption, and antifouling. The pressure-driven systems often suffer from polyamide hydrolysis and selectivity decay at low pH, while carrier-mediated systems introduce durability concerns. Electrochemical and hybrid systems improve integration but encounter challenges like stack resistance growth, electrode fouling, and increased electrical demand. Additionally, few studies report membrane lifetime, flux decline, cleaning frequency, or energy normalized per kilogram of product, parameters that ultimately determine techno-economic feasibility. Downstream purification requirements are also frequently overlooked, as membrane enrichment does not automatically yield battery-grade materials. Therefore, the principal bottleneck in membrane-based recovery is not initial separation performance, but sustained stability, fouling control, and process integration under realistic leachate conditions.

TABLE 5. Comparative table showing chemistry-driven failure modes, preventive and cleaning measures.

7 Challenges: failure modes and mitigation

In the extraction of critical materials from spent batteries and mineral wastes using membrane-based processes, failure modes are tightly coupled to the chemistry of the feed streams. These failure mechanisms are interconnected with solution chemistry and include precipitation, complexation, redox changes, colloidal interactions, organic fouling, silica scaling, gypsum formation, and polymer degradation. This highlights the importance of understanding the chemistry-membrane interactions for effective process optimization and membrane selection. Precipitation is regarded as the most common failure mode in membrane systems treating battery leachates. During membrane treatment, changes in temperature, pH, or ionic strength occur, inducing supersaturation of soluble salts like carbonates, sulphates, Al(OH)₃, and Fe(OH)₃. The precipitates that form accumulate on the membrane surface and within pores, causing severe pore blocking and scaling, as shown in Figure 5A. This phenomenon is observed in lithium recovery, where CaCO₃ and Mg(OH)₂ precipitate under alkaline conditions, and CaSO₄ · 2H₂O (gypsum) precipitation occurs in sulfate-rich leachates. These effects result in flux decline and irreversible fouling, requiring harsh chemicals for cleaning (Satyannarayana et al., 2023; Abdel-Karim et al., 2025; Foo and Lienhard, 2025; Botelho Junior et al., 2023).

In complexation reactions, the charge, size, and mobility of target and impurity ions influence membrane separation efficiency. Transition metal ions (Ni²⁺, Co²⁺) form strong complexes with sulphates or chloride in battery leachates, enabling the complexes to permeate the membrane easily or adsorb strongly onto functionalized membrane. This leads to reduced selectivity and increased fouling (Figure 5B). Additionally, changes in the oxidation state of multivalent ions in the leachate or at the membrane interface can introduce failure modes. For example, in electro-driven membranes (ED), such changes change hydrolysis behaviour, membrane affinity, and solubility. When Fe²⁺ oxidises to Fe³⁺, hydrolysis and precipitation occur, causing rapid fouling (Xin et al., 2025; Chernyaev et al., 2023; Zhong et al., 2025; Shan et al., 2025), as shown in Figure 5C. Table 5 provides further detail on other failure modes, the dominant chemistry involved, preventive measures and cleaning approaches employed.

Based on Table 5, membrane failure is fundamentally a surface chemistry problem when extracting valuable materials from complex battery leachates. Therefore, preventive measures are more effective and economically sustainable than post-fouling cleaning, especially for gypsum and silica scaling, as these are often irreversible. In battery or mineral waste leachates, it is critical that membrane processes are supported by pretreatments in order to maintain efficient selectivity and overall membrane performance. ED processes are vulnerable to redox-driven degradation, while RO/NF processes are more prone to fouling. In contrast, SLP needs strict feed chemistry control to avoid carrier saturation and organic phase instability.

7.1 Quantitative metrics for performance

In addition to the wastewater produced from the highly important industrial sectors such as battery manufacturing, mining, pulp and paper, pharmaceuticals, textile, petrochemical and food processing industries, brine produced naturally have shown an increased produce significant environmental threat. The effluent from these processes negatively impacts the water bodies, air and soil quality posing an even greater thread on human and plants health, as well as the integrity of the ecosystem as a whole (Azmi et al., 2025). As a result, membrane separation methods have become widely used because of their demonstrated ability to recover and reuse valuable resources in addition to treating complex industrial effluents.

The increased application of membranes across all sectors for valuable material recovery and water reclamation brings significant economic advantages and environmental benefits. Membrane separation introduces a reduced use of raw material and highly toxic chemicals, potential to making money from the recovered materials, as well as the decreased expenses for waste disposal and compliance (Staszak and Wieszczycka, 2023). The applicability of membrane technology is measured on quantitative metrics such as permeance, rejection, recover, purity, energy per kg product as well as the membrane overall lifespan. Studies has shown excellency in application for valuable material recover.

To assess the applicability of the commercially available NF membranes, NF99, GE and DK nanofiltration membranes were used by Al-zoubi for the treatment of acid mine drainage from a copper mine. The performance of these membranes was observed through the assessment of heavy metals such as Ca, Cu, Fe, Al, SO₄ and Mg. High permeance and rejection of over 98% for all metal ions using all three membranes was observed. The concentrations of the permeate collected met World Health Organisation (WHO) standards for potable water (Al-Zoubi et al., 2010). Yong-Jian Tang et.al. used spent LIBs leaching solution to investigate the viability of NF membranes to selectively recover lithium ions. In comparison with traditional methods, the use of membrane separation uniquely sidesteps the significant metal losses. In this study, commercially available NF270 and DK membranes were used. DK membranes showed high flux, superior high-valence metal ions rejection capabilities with rejection percentages over 99.0%, whereas allowing permeation of over 40% lithium. With 60% Li rejection, the less hydrophilic DK membrane showed an enhanced acid resistance when compared to NF270 (Gao et al., 2024).

To evaluate the impact of membrane filtration operational conditions such as pH, pressure, temperature, flow rates as well as the concentration on metal ions recovery, Algureiri and Abdulmajeed studied the removal of three metal ions (Ni²⁺, Cu²⁺, and Pb²⁺) using simulated wastewater using RO polymeric membranes. It was reported 96%, 98% and 97.5% of Cu²⁺, Ni²⁺ and Pb²⁺ was rejected. With the increase in metal ions concentration, the permeation flux reduced as a result of the formed negatively charged shielding on the membrane surface resulting in higher affinity between the membrane surface and the positively charged ions (Algureiri and Abdulmajeed, 2016). The same authors reported the impact of the operation factors on the removal of Cu²⁺, Ni²⁺, and Pb²⁺ ions using NF systems. This study was conducted using well tested NF membranes that showed excellent performance in terms of production (effluent quality), permeation, as well as rejection efficiencies. It was observed that the NF system removing more than 66%, 85%, and 78% of the Cu²⁺, Ni²⁺, and Pb²⁺ ions respectively. The reported observation showed that the NF membrane possesses best capabilities for treating industrial wastewater and recovering metal ions with any concentration of heavy metals ions. The NF membranes showed efficient heavy metal removal with high productivity and very low operation pressure up to 1 bar. The low pressure implies that the NF system promises very little costing operation (Algureiri and Abdulmajeed, 2017). Cséfalvay et al. observed similar results wherein Cu²⁺, Ni²⁺, and Pb²⁺ ions were recovered with rejection rates at different concentrations showed over 97% average rejection percentages for heavy metals (Cséfalvay et al., 2009).

Using both NF (NF99) and reverse osmosis (RO 98pHt) membranes operated at laboratory scale, Rieger et al. studied the valuable materials rejections. The reported promising observations and AMD treatment applicability using both NF and RO treatment. Over 95% of the polyvalent ions such as dissolved metals and sulphate was observed using both membranes. The NF membrane, NF99 showed high sodium rejection that is a typical behaviour of nanofiltration membranes. Very acidic and concentrated AMD was treated using the membranes and they showed an overall rejection of about 80.3% and 91% for NF 99 as well as a 91% and 95% rejection for RO 98pHt. The NF processes were operated at transmembrane pressures between 10 and 30 bar resulting in lower scaling on the membrane surface and less energy consumption. These makes NF more feasible than RO when used for the treatment of the highly concentrated AMD mine wastewater (Rieger et al., 2009). The same membranes were used by Ambiado et al. in a pilot study setup for the treatment of AMD collected from gold mines. In this study, the core focus was on the removal and recovery of sulphate. Operational pressures of between 4 and 50 bar for RO98pHt and 4–40 bar NF 99 were used at room temperatures. At the optimum operation pressure of 15 bar for both membranes, the optimum ion rejection for NF and RO systems was recorded to be 92% and 98%, respectively. The overall sulphate rejection was found to be 97% for NF and 99% for RO (Ambiado et al., 2017).

To understand the performance applicability comparison between polymeric membranes and ceramic membranes, Vidic and Wadekar used the two for treatment of AMD from a coal mine. From their findings, polymeric membrane showed a higher metal ions rejection abilities than the ceramic membrane. They monitored the rejection of multivalent ions also and observed over 30% difference between the two membranes with polymeric membranes at 96% and ceramic membrane achieving rejections between 55% and 67%. In addition to the observations, both membranes showed less efficiency in the removal of arsenic. This was improved through the addition of antiscalant to the AMD feed water. This resulted in the increased removal from 114% to 200%. The added antiscalant resulted in about 40% decrease in permeability for both NF and RO membranes as the antiscalant resulted in the formation of a more complex and gel-like scale (gypsum) blocking the pore after 75% permeate recovery. From this study, it was proven that AMD wastewater can be treated to drinking water standard especially when the membrane filtration is coupled with the addition of an antiscalant to the feed water. This did not only result in reduced scaling but also led to a prolonged membrane life span. The use of antiscalant together with membrane filtration will lead to a slight increase in the operation cost of the whole system for AMD treatment (Wadekar and Vidic, 2018).

Other factors that determine the long-term applicability of membranes in valuable material recovery are the separation factor and purity of the recovered material. To assess this, Tau He et.al. used NF membranes to recover Li⁺ and Mg²⁺ and compare the MgCl₂ and LiCl purity at Zero-Recovery experiment. It was observed that with a slight change in Mg²⁺ rejection there was a significant increase in a dramatic increase in the separation factor. These NF membranes showed higher Mg²⁺ rejection with even higher LiCl purity of over 97%. It was reported that is not possible to achieve purity above 99% for LiCl through one-pass NF (He et al., 2025).

To assess the cost effectiveness of using NF membrane treatment in valuable material recovery, lithium was recovered from both salt-lake brine and seawater. It was observed that the production cost from seawater was around USD 80 kg^-1 which was higher than production cost from salt-lake brine USD 2 to 3 kg^-1 (Pfeiffer, 2014). Yaksic et al. (Satyannarayana et al., 2023) investigated the cost of lithium carbonate production and the energy consumption when using NF recovered from geothermal and salt-lake brines. It was reported that around USD 5 to 7 kg^-1 and 35 to 48 kWh kg^-1 were used respectively. They reported the use of membranes as an attractive effective technology due to the high market price of lithium carbonate products of approximately USD 24 kg^-1 (Wang et al., 2018).

Amongst most conventional methods used, the use of both ceramic and polymeric membranes like UF and RO, which are widely used in industry, these membranes have a high durability and efficiency. They have longer lifespan of 5–8 years and 15–20 years for polymeric membranes and ceramic membranes respectively. This translates to significant savings over time due to less frequent replacements and maintenance. Modification of polymeric membranes offers even longer lifespan and improved performance (Azmi et al., 2025). Despite the slightly higher operational costs, the potential for valuable material recovery and the produced effluent quality makes membrane filtration systems economically viable for long term applicability.

8 Techno-economic analysis and realistic performance limits

Membrane processes have seen a surge as preferable alternative or complementary technology for recovery of critical materials. This is due to energy and techno-economic profiles differences between membrane processes and conventional hydrometallurgical methods, and these factors also depend on their application (Botelho Junior et al., 2023; Gao et al., 2025). Membrane processes offer sustainable and energy efficient alternatives to conventional extraction methods, particularly in treatment of dilute effluents, battery recycling, and low-grade ore. Another advantage of membranes is high recovery in low concentrations and selection separation of target materials (Zhu et al., 2025; Li et al., 2025b).

Membrane process like ED and NF generally operate at an energy intensity of 0.5-5 kWh/m³ and this is substantially lower than traditional routes that depend on thermal separation or heavy metal precipitation, which often exceeds 10 kWh/m³ when full water-intensive mining process is accounted for. In terms of cost per volume treated, membrane systems range from about $0.5 to $6 per cubic meter. The costly is associated with high-salinity zero liquid discharge applications. When this is normalized to cost per kilogram of metal recovered, membranes are increasingly competitive for high-value metals such as Pd, Pt, In and Li, where the economics are justified not only by the value of the product but also by reduced solvent losses and lower waste disposal fees. Specific energy use per kilogram of metal recovered further highlights the benefits of membrane-hybrid processes, which often demonstrate a 50%–84% reduction in total energy demand compared to conventional solvent extraction. Due to all these positive attributes, membrane technology is more attractive in terms of water recovery, recovery of valuable metals from dilute streams, and sustainability (Abdi et al., 2024; Carvalho et al., 2025; Scholz and Lucas, 2003). Table 6 depicts the differences between membrane-based process and conventional hydrometallurgy processes for critical material recovery. Realistic performance limits for membranes in battery recycling are shaped by the chemistry of acidic sulfate and chloride leachates. Nanofiltration membranes can achieve lithium selectivity in the range of 70%–90%, with transition metal rejections typically above 90%, but separation of chemically similar divalent cations (Ni²⁺ vs. Co²⁺) remains limited. Acid-resistant membranes are essential, since conventional polyamide NF suffers hydrolysis at low pH; newer pH-stable thin-film composites have demonstrated sustained flux and selectivity under strongly acidic conditions. Surface-modified NF membranes have shown improved lithium recovery efficiency, but osmotic pressure and fouling reduce throughput, meaning fluxes are often constrained to <10 L/m²·h in concentrated leachates. In terms of costs, membranes themselves are relatively inexpensive compared to solvent extraction reagents, but the need for pretreatment, fouling control, and periodic replacement raises operating expenses. Economic analyses suggest membranes are most cost-effective when used for lithium concentration and acid recovery, while solvent extraction or precipitation remains necessary for transition metal separation. Overall, membranes realistically function as complementary steps in hybrid flowsheets, offering sustainability and acid recovery benefits, but not yet delivering complete metal separations at industrial scale.

TABLE 6. Comparative table for membrane processes and conventional hydrometallurgy in terms of techno-economic and energy benchmarks.

9 Challenges and future considerations

This section discusses challenges and future considerations for membrane technology in recovering precious metal ions from spent LiBs leachate. Like any technology, membrane technology has limitations in WWT and metal recovery. Surface fouling and selectivity are major challenges. Fouling significantly reduces membrane performance, necessitating frequent cleaning that can compromise the structural integrity of membranes, leading to a shorter lifespan and the need for replacement. The issue of low selectivity is influenced by the complex composition of leachate and low concentrations of target metals in spent LiBs. For instance, recovering Li in the presence of other metal ions can be challenging, particularly with the traditional membrane processes.

Despite these challenges, membrane technology still offers significant benefits. To improve their effectiveness, several steps need to be taken. Advancing the selectivity of the membranes would be required while integrating this technology with others for hybrid processes. The pretreatment of complex tailings can offer higher selectivity, as they can remove impurities and other contaminants, which also delays the surface fouling. Additionally, optimisation of process parameters such as pH, flow rate, and pressure can improve the membranes’ performance.

10 Conclusion

With the gradually increasing technology and industrialization, a need for recycling and repurposing valuable materials such as zinc, nickel, lithium, cobalt, etc., has received considerable attention. As a result, membrane technology occupies a critical niche in flowsheets for complex wastewaters and secondary resource recovery. Membranes are excellent in selective separations and fouling mitigation when properly modified. However, the issue of chemical stability and scale-up challenges are a cause of limited widespread of the technology. Therefore, future progress pivots on targeted material innovations, thorough long-term stability testing, as well as integration into holistic pretreatments and recovery trains. By aligning future research with industrial stability and scalability requirements, membrane technology can shift from promising lab-scale systems to indispensable components of sustainable resource recovery systems.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Marek Bryjak, Wrocław University of Technology, Poland

Reviewed by: Natarajan Elangovan, Saveetha University, India

Shahla Imteyaz, Aligarh Muslim University, India