The urgent electrolyte sustainability challenges for electric vehicle batteries

Abstract

At a time when society is moving towards electric transport to replace gasoline-powered cars, lithium-ion battery electrolytes are a black box in terms of their production and the raw materials they are made from. This article puts into perspective the urgent need to find alternatives to conventional electrolytes based on fluorinated salts in carbonate solvents.

As the adoption of electric vehicles continues to grow, the production and raw materials of lithium-ion battery electrolytes deserve further scrutiny. Here, authors share their perspective on the need to find alternatives to traditional carbonate solvent and fluorinated salt-based electrolytes.

Introduction

The current global warming-related disasters and air pollution-related illnesses cannot be overcome without drastically reducing the emissions of greenhouse gases and toxic particles such as nitrogen oxides, hydrocarbons or particulate matter. The ever-growing transport sector is a major source of such emissions, representing nearly one quarter of the global energy-related direct CO₂ emissions, with the passenger car being the main contributor¹. It is in this context that the electric car revolution was born, owing to a technology first commercialized in 1991 – the lithium-ion battery (LIB)². Whereas electric vehicles represent approximately 2% of the global car fleet today, it has been predicted that in 2030, approximately 40% of all new car sales will be electric¹. In a truly exceptional manner, this transition has been encouraged at every scale: international programs, national objectives, local initiatives and pledges from car manufacturers, Supplementary section 1 (Supplementary Figs. S1 and S2). This transformation is reshaping the automotive industry as well as fields related to the production and transportation of petroleum products and electricity. Naturally, the demand for batteries is rapidly expanding and giving rise to a wave of new and expanded battery manufacturing plants, representing substantial economic investments, Supplementary section 2³. LIBs contain an intricate formulation of different materials. At the heart of these batteries is the liquid electrolyte, which has a key role in terms of performance, longevity and safety. Eastern Asia is the largest and fastest growing producer and consumer of electrolytes, with China leading the way, followed by Japan and South Korea^4,5. Key players in electrolyte production are Mitsubishi Chemical Holdings Corp., LG Chem., Ube Industries Ltd., Shenzhen Capchem Technology Co. Ltd, Tinci Materials, 3 M Co., Sanyo Electric and Gotion, Supplementary section 3. To produce more electrolyte materials for the growing number of gigafactories, multiple new electrolyte production plants are being constructed and existing sites are being expanded⁶.

The problematic surrounding electrode raw materials are generally well known to the wider public and discussed with intensive scrutiny⁷. In contrast, the same problematics surrounding electrolytes are rarely discussed in such detail^8,9. Nonetheless, it is in the interest of all parties within the electric vehicle (EV) field to understand the sustainability problematics of each battery component if any long-term, viable and sustainable transport solution is to be attained using Li-ion batteries. Herein, we describe in detail the most prominent industrial production processes used for the current generation of electrolytes composed of ethylene carbonate (EC), ethyl methyl carbonate (EMC) and LiPF₆. Next, we estimate the required quantities of each raw material needed to produce 1 kt of this electrolyte. We then estimate national and global electrolyte and raw material requirements for different levels of EV adoption. Subsequently, we examine the geographical distribution of LiPF₆ raw materials and the criticality of these minerals. Finally, we present an overview of EC, EMC and LiPF₆, alternative sustainable electrolytes and possible strategies towards a greener electrolyte conceptualization.

The current generation of battery electrolytes

Today’s commercial electrolytes are composed of LiPF₆ dissolved in mixtures of cyclic and linear carbonates^10,11. LiPF₆ permits high ionic conductivity, solid electrolyte interphase formation and efficiently passivates Al current collectors¹¹. Although other lithium electrolyte salts exist, the commercial standard remains LiPF₆². Details of other electrolytes can be found in Supplementary section 4 (Supplementary Figs. S3 and S4). EC is a cyclic carbonate present in most Li-ion battery electrolytes. This compound possesses a high dielectric constant and forms low-energy complexes with Li⁺, permitting high ionic conduction. As EC has a high melting point (36.4 ° ts producing liquids with high dielectric constants and low viscosity¹². The typical commercial electrolyte contains EC and EMC (with a 3 to 7 weight ratio of EC to EMC, respectively) and LiPF₆ with a 1 mol.L⁻¹ concentration. This electrolyte will be simply referred to herein as EC/EMC/LiPF₆. In addition to LiPF₆ and carbonates, small amounts of additives can also be used (generally less than 10%). The effect of these additives on the battery’s performance are diverse, including interface stabilization, fire-retardant agents and increased longevity. Fan et al. proposed a comprehensive review of these compounds¹³. At the industrial scale, EC is mostly used as a solvent for lubricants, whereas EMC and LiPF₆ are used primarily for electrolyte purposes. The market values of EC, EMC and LiPF₆ have been estimated to be worth 800 M US dollars, 674 M US dollars and 2.47B US dollars each, respectively, Supplementary section 5^14–16.

The production of EC/EMC/LiPF₆ electrolytes

Having identified the typical composition of Li-ion battery electrolytes (EC/EMC/LiPF₆), the production and raw materials of those compounds are next examined.

LiPF₆ is produced via the reaction of lithium fluoride (LiF) and phosphorus pentafluoride (PF₅), Fig. 1 and Supplementary Fig. S5¹⁷. The production of these two chemicals will be described, beginning with the sources of the elements that compose them: Lithium, Fluorine and Phosphorus.

Fig. 1. Isometric representation of raw material sources, production flows, and key chemical reactions for the synthesis of LiPF₆ and EC/EMC.

The top section illustrates the sources and flow of raw materials used to produce LiPF₆; the middle section shows the production routes of EC and EMC; and the bottom section is annotated with the corresponding chemical reactions.

Today, two different raw materials are used as sources of lithium: mineral ores and brine. Li can be extracted from a variety of minerals however, spodumene (LiAlSi₂O₆) is the most common. This silicate mineral typically contains between 1-2% Li₂O in addition to trace elements of Al, Fe, Ca, Mg and Na¹⁸. Certain lake brines, the second main source of Li, contain Li⁺ at concentrations up to 0.157%wt¹⁹. In order to produce LiPF₆, these lithium-poor raw materials are first processed into Li₂CO₃. On an industrial scale, spodumene undergoes multiple roasting and grinding steps, followed by treatment with sulfuric acid to form Li₂SO₄. Treatment of Li₂SO₄ with Na₂CO₃ yields Li₂CO₃. As for Li brine, the production of Li₂CO₃ involves concentrating the brine by solar evaporation until salt precipitation occurs when the lithium content reaches 6%. Various steps are then required to remove the other elements, at which point carbonation with Na₂CO₃ yields Li₂CO₃²⁰. LiF, the direct precursor of LiPF₆, is formed via a fluorination reaction of Li₂CO₃ with hydrogen fluoride (HF)²¹.

Fluorite (CaF₂), also called fluorspar, is the principal source of fluorine. Raw fluorite minerals have low fluorine content and are often associated with other minerals such as calcium carbonate or quartz²². Raw fluorite requires purification via a process involving grinding and flotation cleansing in order to yield acid-grade purity CaF₂ (97% pure)²². HF is then produced via the reaction of acid-grade fluorite with H₂SO₄²¹.

Phosphate rock, whose major component is fluorapatite (Ca₅(PO₄)₃ F), is the main source of phosphorus²³. Crude phosphate rock is used to produce white phosphorus (P₄) via the Wohler process wherein phosphate rock is mixed with coke and silica in an arc furnace at 1500 °C²³. The chlorination of white phosphorus with Cl₂ generates phosphorus trichloride (PCl₃) whose subsequent chlorination yields phosphorus pentachloride (PCl₅)²⁴. Finally, the fluorination of PCl₅ with HF provides phosphorus pentafluoride(PF₅), the direct precursor of LiPF₆^24.

As for EC and EMC, these organic compounds are produced using different fossil fuels as starting materials, as illustrated in Fig. 1 and Supplementary Fig. S5. The starting material for EC is ethylene. Globally, the majority of ethylene is produced via the steam cracking of naphtha, a fraction of crude oil containing hydrocarbons ranging from C₅ and to C₁₂ and representing 15-30% of the crude oil weight²⁵. Ethylene can also be produced via the steam-cracking of ethane or liquified gas²⁵. The catalytic oxidation of ethylene with oxygen yields ethylene oxide (EO)²⁶. Finally, the cycloaddition of CO₂ to EO provides EC²⁷.

EMC is generally produced via the transesterification reaction of DMC and ethanol⁹. DMC is for the most part produced via the oxidative carbonylation of methanol with carbon monoxide (CO), although other paths exist^9,28,29. Both methanol and CO originate from syngas - a mixture of gases (mainly H₂ and CO) – typically obtained through the steam reforming of methane-rich natural gas³⁰. Although different sources of syngas exist, such as coal and oil, 90% of methanol globally is produced from natural gas feedstocks³¹.

More details on all the above processes and the production pathways of all other required precursors can be found in Supplementary sections 6 to 8 (include Fig. S5). In addition to all the precursors of EC/EMC/LiPF₆ described above, further materials are required for solvation, catalysis and purification purposes. Taking those chemicals and their raw materials into consideration can only increase the already heavy environmental burden of EC/EMC/LiPF₆ electrolytes and critical raw materials requirements. Although the focus of this discussion is sustainability and raw materials, it is important to note that a troubling array of hazardous compounds are required to produce this electrolyte, posing unique risks for manufacturing, transport and storage processes. The health, physical and environmental hazards of each precursor are detailed in Supplementary Table S1 as well as a focused discussion on the hazards of lithium-based salts and ethylene oxide in Supplementary Section.

Estimation of EC/EMC/LiPF₆ raw materials

The amounts of all pure body precursors required for 1000 tons of EC/EMC/LiPF₆ (containing 3:7 weight EC:EMC and 1 mol.L⁻¹ LiPF₆) were calculated stoichiometrically (see Fig. 2). All data and calculations can be found in Supplementary Table S2 and Supplementary Figs. S7, 8.

Fig. 2. Sankey flow diagram to produce 1 kt of EC/EMC/LiPF₆ electrolyte illustrating the required amounts of each precursor.

The color of each branch is indicative of the required weight as represented by the legend. To facilitate reading: The H₂O required to produce Na₂CO₃ and H₂SO₄ is not represented; All the O₂ required to produce H₂SO₄ is represented for SO₂; intermediary NaHCO₃ is not represented for Na₂CO₃. All data and calculations are presented in Supplementary Table S2 and Supplementary Figs. S7, 8.

The basic raw materials that contribute atoms directly to LiPF₆, EC and EMC are lithium salt lake brine/spodumene, phosphate rock, fluorite, crude oil and natural gas. All the raw materials needed for EC/EMC/LiPF₆ - except for O₂, N₂, CO₂ and NaCl - are non-renewable sources derived from fossil fuels, mined from the earth or extracted from salt lakes. Whereas the problematics surrounding the geographical distributions and supply chains of fossil fuels are well known, we propose a focus on the other materials: lithium, Salt Lake brine/spodumene, phosphate rock and fluorite.

Mineral criticality and global distribution

The definition of a critical mineral varies from region to region, however, a mineral is generally declared as critical when it is indispensable for a region’s economic and national security^32,33. These minerals are required for essential activities such as manufacturing, agriculture, defense, communication, transportation or energy sectors. Supply chain disruptions of critical minerals can lead to serious economic damage, vulnerability and shortages³³. When considering the three mineral raw materials for LiPF₆ (phosphate rock, fluorite and lithium carbonate), all are currently considered critical to a worrying number of countries and regions as represented in Fig. 3. Lithium is assumed as critical to the EU, the US, Japan, India, Brazil, South Korea, Australia and Canada^33–36. Phosphate rock, phosphate ores or elemental phosphorus are assumed as critical by the EU, India, Brazil and South Korea^33,35,36. Fluorite or fluorine is assumed as critical by the EU, US, Japan, India and Canada^33–35. These minerals are distributed across the Earth in deposits of varying concentration and accessibility. A “mineral reserve” is defined as the economically mineable part of a measured or indicated mineral resource. Global lithium reserves are estimated at 21 million tons³⁷. These reserves are situated in 2 main regions, South America and Australia (52.8% and 22.4% of the global lithium reserves, respectively)³⁷. China is also a lithium-rich territory, possessing 7.1% of the global lithium reserves³⁷. The global reserves of phosphate rock are estimated at 71,000 million tons³⁷. Morocco has 70.8% of the global phosphate reserves³⁷. The second most important source of phosphate rock reserves is China, possessing 4.5%³⁷. The following countries are estimated to possess at least 1000 Mt phosphate rock reserves: Egypt, Algeria, Syria, Brazil, Saudi Arabia, South Africa, Australia, the US and Finland³⁷. Global fluorite reserves are estimated at 320,000 kt. Mexico possesses the largest share (21.5%), followed by China (13.3%), South Africa (13.0%), Mongolia (7.0%) and Spain (3.2%)³⁷. Significant amounts of fluorine are also present in phosphate rock³⁷. Mineral reserves data can be found in Supplementary Table S3.

Fig. 3. Annual EC/EMC/LiPF₆ requirements considering the total transition of the passenger car market to battery powered vehicles (scenario 1) annotated with the distribution of lithium, phosphate rock and fluorite mineral reserves and countries where those minerals are assumed as “critical” raw materials.

The color of a region represents its EC/EMC/LiPF₆ requirements. Triangular symbols represent the nations where lithium (Li), phosphate rock/phosphorus (P) or fluorite/fluorine (F) are assumed as “critical” raw materials. Circles represent the share of global mineral reserves of lithium (purple), phosphate rock (pink) and fluorite (orange) (area proportional to the mineral reserve).

Global EC/EMC/LiPF₆ electrolyte requirements

To estimate the global and national electrolyte requirements for LIB-powered electric cars, we conceived a model electric car equipped with a battery pack of similar specifications to one of the most sold EV models: the Tesla Model 3 (TM3). Considering that a TM3 battery pack is composed of 2976 cells (21700 format), weighing 69 g each and containing 12 %wt of electrolyte, we can estimate that each TM3 contains 24.64 kg of EC/EMC/LiPF₆. Details on this estimation can be found in Supplementary Fig. S9^. In this work, three different scenarios of EV adoption were explored, as described in Supplementary Section 11. Scenario 1 represents the total electrification of the passenger car market in its current form.

Scenarios 2 and 3 represent predicted EV adoption levels based on stated policies scenario and announced pledges scenarios, respectively ambitions as described by the International Energy agency (IEA) in 2023¹. The electrolyte requirements of each nation for each of these scenarios was estimated.

A total of 64.28 million passenger cars were sold worldwide in 2019. China was the largest passenger car market, followed by the EU and the US, representing 33.4%, 19.8% and 7.3% of total car sales, respectively. If all these passenger cars were LIB powered (scenario 1), it can be estimated that 1584 kt of EC/EMC/LiPF₆ would be required to produce the necessary number of batteries. Details on these calculations can be found in Supplementary Section 10,and car sales data from 2019 can be found in Supplementary Table S4. As to produce that amount of electrolyte, 48.9 kt of pure lithium carbonate, 222.6 kt of pure fluorapatite and 310.2 kt of pure fluorite would be required. For this scenario 1, the electrolyte needs of each nation are presented in Fig. 3, and all the data can be found in Supplementary Table S5. As an example, the US alone would need 116.3 kt per year of EC/EMC/LiPF₆ electrolyte which represents 22.8 kt of pure CaF₂, 16.3 kt of pure Ca₅(PO₄)₃ F and 3.6 kt of pure Li₂CO₃, all critical raw materials.

In 2025, the IEA has projected that the number of battery-powered passenger car sales will reach 16.0 million units (for both scenarios 2 and 3)¹. In order to produce that number of cars, 394.2 kt of EC/EMC/LiPF₆ will be required. In 2030, the IEA has projected that 31.0 million to 33.0 million batteries powered electric cars will be sold (scenario 2 and 3 respectively)¹, representing 763.9 to 813.2 kt of EC/EMC/LiPF₆. Car sales and estimated requirements in EC/EMC/LiPF₆, Li₂CO₃, Ca₅(PO₄)₃ F and CaF₂ data for scenario 2 and 3 using IEA data can be found in Supplementary Table S6 and Supplementary Fig. S10^.

Due to the enormous quantities and highly heterogenous distribution of their raw materials as well as the heavy processes and dangerous chemicals required, EC/EMC/LiPF₆ type electrolytes – in their current form – resemble more a short-term solution than a long-term sustainable solution for LIB powered electric vehicles. The scientific community must accelerate the search for realistic and sustainable electrolytes for large-scale applications. Different approaches to this problematic may offer a solution: production of conventional electrolytes from sustainable sources, development of novel sustainable electrolytes and recycling electrolyte materials.

EC/EMC/LiPF₆ from sustainable sources

Using bioderived chemicals to produce conventional carbonate electrolytes is highly advantageous for controlling and expanding the supply of raw materials whilst reducing the environmental burden of their raw material extraction. One obvious approach relies upon replacing the fossil sourced naphtha used for the manufacture of ethylene: the basic building block of EC. A promising alternative consists in using bio-ethylene which can be obtained via the catalytic dehydration of bio-ethanol, a molecule produced by fermenting saccharide rich biomass³⁸. It is worth to highlight that in 2013, this process was already implemented in commercial facilities across Brazil and India, producing 200 kt of “bio-ethylene” annually in the largest plants³⁸. Moreover, the carbon dioxide emissions related to bio-ethylene production are significantly lower than those of conventional ethylene production³⁹. Additionally, the carbon dioxide released during the fermentation of glucose could be reacted with EO to produce EC thus further offsetting the environmental impact of bio-EC production. Nonetheless, the profitability of this process is highly dependent on raw material feedstock prices, which differ from region to region⁴⁰. It has also been shown that bio-EC can equally be obtained using biomass-derived glycolaldehyde and urea, which avoids the use of toxic and dangerous EO⁴¹. Using bio-EC as a precursor, bio-DMC, bio-EMC and bio-DEC could be produced via transesterification reaction with bio-methanol and/or bio-ethanol, wherein bio-methanol could be derived from the products of biomass gasification⁴². The large-scale production of DMC via the transesterification of EC was licensed by Asahi Kasei, and its profitability was found to be higher than that of conventional EC-production routes^43,44. In addition to the advantageous raw material benefits, these alternative pathways for carbonates present reduced energy consumption and net CO₂ emissions when compared to conventional routes⁴³. Furthermore, ethylene glycol is a valuable by-product of these transesterification reactions, as its subsequent reaction with urea can yield EC again, further reducing the environmental burden of these carbonates. A scheme for this source-friendly production of EC, DMC and EMC is proposed on Fig. 4. It is worth remarking that a comprehensive environmental and techno-economic assessment is crucial for the proposed approach, as it relies on integrating existing industrial pathways that are currently unconnected and unoptimized.

Fig. 4. Sustainable production routes of solvent materials EC, DEC, and EMC via integrated industrial processes.

The main section shows the synthesis of EC from bio-ethylene via catalytic dehydration, oxidation, and CO₂ cycloaddition; a secondary pathway depicts the formation of DMC, EMC through transesterification using bio-EC and bio-alcohols; and the circular pathway highlights the generation of EC from the by-product ethylene glycol.

Regarding LiPF6, whilst less hazardous and cheaper precursors have been proposed for its synthesis, the basic critical raw material demand remains the same due to the inorganic nature of this molecule. As such, LiPF₆ poses a more challenging problem than EC and the linear carbonates, whose organic nature offers more opportunities for alternative synthesis paths. Nonetheless, alternative fluorine, phosphorus and lithium sources are currently being explored but remain mostly at the lab scale^45–47.

Alternative electrolyte materials

Development of novel electrolyte components which totally or partially avoid the use of critical raw materials would present a more optimistic scenario for long-term solutions within LIBs. As previously remarked, special attention should be paid to LiPF₆ since intensive mining is required for the raw materials. Considering that more than half of the weight of LiPF₆ arises from fluorine atoms, the utilization of alternative elements in electrolyte salt compositions could prove beneficial not only from an economical and strategic point of view, but also from a safety perspective with regards to F-derived compounds (i.e., HF, PF₅). Different fluorine-free lithium salts composed of perchlorates, phosphates^, borates and heterocyclic anions have been examined⁴⁸. However, since fluorinated species play a major role in passivating aluminum cathode current collectors at high potentials, fluorine-free salts generally present corrosion issues. One exception worth consideration is lithium bis(oxalato)borate (LiBOB). The production paths and raw materials of LiBOB and LiFSI are respectively proposed in Supplementary Section 12 (Supplementary Figs. S11 and S12). In addition to efficient Al passivation, this borate salt presents numerous advantages: stable SEI on graphite, high thermal stability, safe decomposition products and higher tolerance to water than LiPF₆⁴⁹. Nevertheless, a slightly higher resistivity and lower rate capability limit the use of LiBOB to additive proportions within LiPF₆ based electrolytes⁵⁰. Whilst fluorine-free electrolyte formulations are at early stages of development, other promising alternative lithium salts are more established and have higher chances of large-scale application. Indeed, imide-based salts best exemplified by lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) and lithium bis(fluorosulfonyl)imide (LiFSI) have been claimed as the most suitable replacement for LiPF₆ in commercial devices⁵¹. These salts present remarkable solvation abilities, high ionic conductivities and both thermal and electrochemical stability, favorable attributes for LIBs with extended lifespans. It is also worth remarking on their substantially improved safety features, demonstrating considerably lower (LiFSI) or negligible (LiTFSI) sensitivity toward hydrolysis when compared to LiPF₆⁵¹. The stability of LiFSI is slightly lower than that of LiTFSI due to the lability of the S-F bond, which contributes to the reduction of FSI^- anion on the graphite surface, thus forming a protective SEI layer. Moreover, LiFSI enhances the overall conductivity of LIBs due to its high lithium- ion dissociation ability and relatively low associated viscosity⁵². Imide salts clearly have certain interesting advantages over LiPF₆. In the case of LiTFSI, large-scale adoption will only be possible when lower production costs and efficient recycling strategies are achieved⁵³. In contrast, LiFSI has been widely used as an additive for fast charging LIBs and is attracting further attention from major battery manufacturers since Nippon Shokubai demonstrated large-scale production in 2013. Currently, an increase of production capacity from companies in China, Japan and Korea is expected to further decrease the price of LiFSI, which in 2021 was lower than the current price of LiPF₆⁵⁴. Accordingly, the portion of LiFSI salt in commercial electrolyte configurations gradually increases as the economies of scale push prices down. Regarding critical raw materials, LiFSI contains no phosphorus and three times less fluorine when compared to LiPF₆ and LiTFSI, a substantial improvement. In consequence, an optimistic outlook could be expected if the great efforts towards LiFSI development led to the successful application of this salt for LIBs. Nevertheless, additive amounts of LiPF₆ or LiBOB would still be required since LiFSI alone does not protect aluminum current collectors from corrosion.

Concerning sustainable electrolyte solvents, different lab-scale solutions have been proposed. Replacing conventional organic solvents by water has obvious sustainability and fire safety benefits. However, the high concentration of fluorinated salts required, generally LiTFSI, results in a considerable increase of device costs and environmental impact, whilst also presenting drastically reduced energy densities⁵³. On the other hand, several bio-derived chemicals such as amino acids or polysaccharides have also been validated as suitable precursors for ionic liquid-based electrolytes, a promising category of non-flammable electrolytes capable of achieving higher applied potentials⁵⁵. Polymer-based electrolytes can equally be produced using sustainable bio-based polymers, which present an incredibly rich diversity of chemistries: starch, xanthan gum, cellulose, chitosan, agar, natural rubber, vegetable oils, etc ref. ⁵⁶. Even though green electrolyte solvents do not currently satisfy the high-performance requirements of modern devices, their development and perfection present a pragmatic and promising long-term goal.

Recycling of electrolyte materials

In every predicted scenario, the number of EV sales is expected to grow considerably. However, the average service life span of their LIBs is 8 to 10 years⁵⁷. As a consequence, the number of spent LIBs are expected to skyrocket, resulting in serious environmental and safety concerns if not handled properly. To address this issue, significant efforts have been devoted over the last two decades towards developing suitable recycling techniques. Indeed, extracting their valuable materials is not only beneficial from an economical point of view but will equally alleviate the heavy burden of LIBs on minerals resources. Unfortunately, a large portion of current recycling companies rely in traditional energy-intensive techniques such as pyrometallurgy, wherein all the cell components are pyrolyzed with the aim of recovering valuable metals (i.e., cobalt, copper, nickel)⁵⁸. The electrolyte, anode and separator are totally oxidized as to provide energy for the metallurgical process, whereas lithium is lost in the slag. To date, hydrometallurgy processes are steadily gaining ground thanks to their superior efficiency when recovering cell materials. In hydrometallurgy approaches, metal ions in spent LIBs are leached out and then separated by solvent extraction or precipitation. Alarmingly, there is an apparent tendency within industrial recycling processes to focus on the recovery of cathode materials whilst ignoring the other cell materials. Obviously, this disregard for efficient battery recycling is not sustainable and challenges the viability of battery-powered EVs. Accordingly, international organizations, governments and companies are developing sustainable legislation and strategies towards recycling all battery components with the goal of achieving a viable and closed-loop system⁵⁹.

Due to the high concentration of inorganic salts, the electrolyte contained within an EV’s battery pack (approx. 25 kg for the TM3) should be considered as an important source of critical elements. In that sense, particular attention should be paid to recovering the electrolyte from an economic, strategic and environmental perspective. Indeed, if treated incorrectly, LiPF₆ can lead to serious contamination of the soil and air primarily due to the release of highly toxic fluorine and phosphorus compounds. Electrolyte recovery is, however, no easy task as electrolyte compositions vary from manufacturer to manufacturer and depend on the age and condition of a cell as chemical modifications can take place over time or during harsh cycling conditions (overheating, deep discharge, overcharging, etc.). Among the few companies addressing electrolyte recovery, an array of different procedures are employed, although in some cases the organic solvents are not recovered or LiPF₆ is decomposed⁶⁰.

One approach for recovering spent electrolyte utilizes solvents capable of permeating the different elements of a spent battery. Organic solvents like propylene carbonate have proven to be highly effective for recovering organic compounds in that manner⁶¹. Nevertheless, organic solvent extractions have limitations such as low yields and the presence of impurities, which hinder the reutilization of the recovered materials. In contrast, the use of supercritical CO₂ as an extracting agent avoids hazardous emissions and limits impurities, therefore, enabling an easier separation of the reclaimed products⁶². Unfortunately, low recovery rates of LiPF₆ via this approach means that the combination of both strategies is required as to efficiently recover the electrolyte⁶³. Another important consideration is the sensitivity of LiPF₆ towards moisture, which complicates the separation and purification processes and imposes using anhydrous solvents.

In view of the inherent difficulties involved with recycling conventional electrolytes, a paradigm shift towards bottom-up electrolyte development could present promising opportunities. In this approach, electrolyte formulations would be conceptualized according to sustainable recycling methods and not the opposite, which is generally the case today. On this basis, water-based recycling processes that avoid using complex instrumentation, expensive materials, or toxic and polluting organic solvents should be strongly considered. Such water-based recycling processes could also allow the simultaneous aqueous exfoliation of the electrodes, thus potentially avoiding the use of toxic N-methyl-2-pyrrolidone in electrode exfoliation or preparation⁶⁴. Whereas distillation or water/oil processes can effectively separate organic solvents in recycling processes, the hydrolysis of LiPF₆ in those conditions greatly limits their application and amplifies the need for water-stable salts. In contrast, water-stable salts could potentially be recovered and reused in such processes. One exceptionally water-stable salt is LiTFSI, whose use is generally restricted to lab scales due to its high price, but which could become economically interesting if efficient water-based recycling was developed. Attempts at recycling this relatively mature salt produced great results when considering yield and purity^65,66. Furthermore, steps are being taken to reduce the cost of this salt by optimizing processes and starting materials and by exploring alternative synthesis routes. Similarly, LiFSI has a relatively good resistance to hydrolysis⁶⁷. This salt may open the door to less challenging electrolyte recycling methods that do not require strict anhydrous conditions (unlike for LiPF₆) or to water-based recycling approaches. Surprisingly, although LiFSI is considered a next-generation salt, in-depth examinations of its recyclability are glaringly absent in the literature. It is, however, worth mentioning that producers of LiFSI claim this salt is compatible with hydrometallurgy processes.

In summary, serious efforts should be dedicated towards developing methods to recycle the current generation of organic electrolytes whilst also advancing the next generation of electrolytes whose recovery must be safe, cheap and sustainable. A closed loop lithium cycle will dramatically diminish the heavy reliance on mining to produce lithium-based salts which is crucial for any realistic scenario wherein LIBs are sustainable.

Conclusions and outlook

EC/EMC/LiPF₆ type electrolytes - the main electrolyte systems used in EV LIBs - require many non-renewable resources mainly: lithium, phosphate rock, fluorite, crude oil and natural gas. Enormous amounts of raw materials will be required to satisfy the production of EVs within the context of the electric vehicle transition. Further accounting for the extra materials required for purification, solvation and catalysis adds further stress on the sustainability of EC/EMC/LiPF₆. This increasing demand for LIBs will only amplify the “criticality” of those materials as they are distributed across the globe in a highly heterogeneous manner. These factors place the electrolyte in a precarious position, which could be susceptible to supply chain disruptions. In addition, several highly hazardous intermediaries (ethylene oxide, HF, PF₅, PCl₃, P₄, etc.) are required to produce this electrolyte.

All these factors underline the need to move towards sustainable electrolytes. Whereas carbonate type solvents (EC, EMC, etc.) and other organic materials (ionic-liquids and polymers) can be produced from renewable resources, the inorganic nature of LiPF₆ means that it cannot be made from resource-friendly sources. In addition, the current recycling processes are not adapted for LiPF₆, notably due to its hydrolysis, a disadvantage common to many lithium electrolyte salts. Nevertheless, the steady industrial transition towards the large-scale implementation of LiFSI presents an encouraging opportunity for more sustainable electrolytes owing to their lower reliance on critical raw materials and higher potential for recyclability. Similarly, despite LiTFSI having a heavier reliance on non-renewable resources, a more complex production path and higher costs compared to LiFSI, this salt is also an interesting alternative electrolyte salt due to its unrivaled stability and proven recyclability.

Author contributions

T.-F.B. and O.F. conceived the paper, collected and analyzed the data, designed and edited the figures and wrote and edited the manuscripts. J-L.G.-U. and A.B. collected data, wrote the paper and designed the figures. Y.Z. Edited the figures and collected data.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work.

Data availability

All data (generated and referenced) is available in the Supplementary Information.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-025-60711-7.