Analyzing Organic Electrolyte Solvents from Spent Lithium-Ion Batteries as a Basis for Distillative Value Component Recovery

Abstract

The rapid expansion of lithium-ion batteries (LIBs), largely driven by the rising demand for electric vehicles, will lead to a significant increase in end-of-life (EOL) batteries, necessitating efficient recycling processes, which must be accompanied by equally efficient purification steps. This study addresses the challenge of reusing organic electrolyte solvents from spent LIBs, a key component often overlooked in existing recycling strategies. To address this issue, we developed a gas chromatography (GC) method. A variety of spent electrolyte samples of different origin, including mechanical-thermal pretreatment or direct cell recovery, were analyzed by quantification of common solvents and identified organic impurities. Results demonstrated that the composition of the recovered electrolytes was highly variable, with concentrations fluctuating. Impurities were identified, which may originate from various sources throughout the lifespan of an LIB and have the potential to reduce the performance of second-life LIBs by reusing the electrolyte without any purification. The findings highlight the necessity for advanced purification methods like a distillation process to remove these impurities and ensure the viability of recycled electrolytes in maintaining the performance and safety standards required for LIBs. This research contributes to the broader goal of enhancing the sustainability and reuse of battery materials.

1. Introduction

As a consequence of the significant increase in the production of secondary batteries such as lithium-ion batteries (LIBs) and sodium-ion batteries, the global production capacity of over 1 TWh was reached for the first time in 2023 [1]. Based on the currently announced production facilities, a further increase in production is projected to reach approximately 9 TWh by 2030, with the majority of the market expected to be for battery electric vehicles (BEVs) [2]. Consequently, an increase in end-of-life (EOL) batteries can be expected with a time delay due to the lifespan of a LIB. Based on calculations from Fraunhofer ISI [3], it is estimated that approximately 420 kt of EOL batteries will be incurred in 2030, rising to approximately 2100 kt in 2040. It is anticipated that a large proportion of the battery waste will continue to originate from cell production. Assuming that the current forecasts materialize, the proportion of EOL LIBs from BEVs will increase from around 20% in 2035 to around 50% in 2040 [3]. Once a battery has reached EOL, it can either be downgraded for a second, less demanding use, e.g., as stationary storage, or recycled in order to recover the materials. The recycling of LIBs is regulated by the European Union, which has set a target recycling quota of 70% by average weight by the end of 2030 [4]. In order to achieve this recycling quota and close the material cycle, the recovery and processing of the electrolyte is becoming increasingly important. The electrolyte accounts for approximately 8–16 wt-% [5] and is therefore a substantial waste product on a kiloton scale.

Electrolytes used in LIBs usually consist of a lithium salt dissolved in an organic solvent mixture comprising a minimum of two components. Additionally, they are frequently combined with additives [6]. The electrolyte solvents are typically a combination of a cyclic carbonate, such as ethylene carbonate (EC) or propylene carbonate (PC), and at least one linear carbonate, such as dimethyl carbonate (DMC), ethyl methyl carbonate (EMC), or diethyl carbonate (DEC) [7,8,9]. Furthermore, other solvents, such as fluorobenzene (FB), are commonly used in commercial LIBs [10,11]. The advantages of FB include a positive influence on the formation of a stable solid electrolyte interface (SEI) between the anode and the electrolyte enriched in lithium fluoride (LiF) [12,13].

Additives are added to the electrolyte in order to enhance the characteristics of the LIB. Film-forming additives, such as vinylene carbonate (VC) or fluoroethylene carbonate (FEC), are utilized to optimize the interfaces between the electrodes and the electrolyte, thereby enhancing the lifespan and performance [8,14,15]. Cyclohexylbenzene (CHB) is a commonly employed overcharge protection additive [16].

Processes for recovering the solvents already exist. One option is a mechanical-thermal pre-treatment of the LIB. This process involves shredding the LIB, with the organic volatile electrolyte solvents subsequently separated from the solids through vacuum-drying in an inert gas atmosphere. This allows for the recovery of part of the electrolyte as condensate [17,18]. However, these solvents have not yet been considered for further processing; instead, they are thermally utilized. An alternative approach for the recovery of the electrolyte is sub- and supercritical CO₂ extraction [19,20,21]. In either case, the recovered electrolytes contain impurities that can be attributed to various sources. On the one hand, decomposition reactions within the cell can result in the liberation of impurities. These reactions are primarily observed during electrochemical ageing. Initially, this occurs during the formation and growth of the SEI within the cell [22,23,24,25] and subsequently during calendar ageing [19,22,23,24,26,27,28,29,30]. However, this phenomenon is also evident in thermal ageing, see, e.g., [22,23,27,28,29,31,32,33,34]. On the other hand, inadequate storage, particularly over an extended period of time in an atmosphere with a high water content or an excessively high temperature, can also result in the formation of contaminants. These include a reaction between the organic carbonates and water, as well as the transesterification of the organic carbonates to form impurities such as ethanol (EtOH) or methanol (MeOH), among others [24,35,36]. The recycling process itself may also be a potential source of contamination of the electrolyte. The degradation products of the electrolyte are analyzed during ageing and/or after removal from the electrolyte using suitable analytical methods, including gas chromatography-flame ionization detection (GC-FID), gas chromatography-mass spectrometry (GC-MS), high-performance liquid chromatography (HPLC), Fourier transform infrared spectroscopy (FT-IR), nuclear magnetic resonance spectroscopy (NMR), and others. However, the composition of the utilized electrolyte and cell material was clearly defined, and the experimental parameters are well established and contingent upon the specific field of investigation. Nevertheless, this is not the case in the context of industrial recycling of LIBs, where a significant number of different LIBs are processed, which differ in terms of their cell chemistry with various formulations for the solvent mixtures initially used. The recycling of LIBs from different manufacturers and battery generations with different life histories, recycled all together, can result in notable variation in the composition of the recovered electrolyte. Consequently, a diverse range of substances with varying compositions from all LIBs tend to accumulate in the recovered electrolyte.

Further studies have demonstrated that, in addition to the recovery of the initial electrolyte solvents, the removal of decomposition products and impurities, such as water, hydrofluoric acid (HF), and others, is crucial for the reuse of the recovered electrolyte. Liu et al. [37] recovered electrolytes through CO₂ extraction, and the extract was purified by anion exchange and molecular sieves to reduce the HF and water content. No further processing of the recovered electrolyte was carried out. In comparison to commercial electrolytes, the recovered electrolyte showed a reduced performance. It was postulated that additional purification steps for reclaimed electrolytes are necessary [37]. Xu et al. [38] pursued an approach for the reuse of spent electrolyte solvents with treatment by vacuum distillation. Compared to commercial electrolytes, the electrochemical performance of the recovered electrolyte was satisfactory in both charging and discharging, which substantiates the efficacy of the purification strategy for the recovered electrolyte, for instance, through distillation.

In order to illustrate the issue of recycling waste and to investigate the purification of recovered electrolyte solvents in more detail, a GC method was developed for analyzing organic electrolyte solvents from spent LIBs. So far, no standard GC method is available that allows for the detailed analysis of these complex mixtures. Due to the absence of a diluent, which is a rather unconventional technique for the analysis by GC, the newly developed method underwent a validation procedure. Samples were obtained directly from LIB cells or recycling processes at both laboratory and pilot plant scales. The aim was to quantify the frequently occurring components using GC-FID and to identify other substances, to the greatest possible extent, using GC-MS. The GC-MS analysis approach was used to generate a dataset of potential substances present in the recovered electrolyte solvents with the objective of aiding the design of future distillation approaches. These compounds are discussed and categorized in the context of their origin. This represents a pragmatic approach, focusing on the identification of potential impurities.

2. Materials and Methods

2.1. Chemicals and Reagents

DMC (≥99.8%), DEC (≥99.8%), and PC (≥99.7%) were supplied by Carl Roth (Karlsruhe, Germany). MeOH (≥ 99.8%), FB (99%), EC (99%), CHB (98%), chlorobenzene (CB, >99%), and acetonitrile (AcN, ≥99.8%) were supplied by Fisher Scientific (Schwerte, Germany). FEC (battery grade) and VC (battery grade) were supplied by BASF (Florham Park, NJ, USA). EMC (≥99.95%), EtOH (≥99.8%), acetone (≥99.8%), and sodium hydrogen carbonate (NaHCO₃, ≥99.7% Merck) were supplied by E-Lyte (Münster, Germany), Sigma-Aldrich (Taufkirchen, Germany), VWR International (Darmstadt, Germany), and Merck (Darmstadt, Germany). Commercial battery electrolyte (LP57) containing 1M lithium hexafluorophosphate (LiPF₆) dissolved in a 7:3 mass ratio of EMC:EC supplied by E-Lyte (Münster, Germany).

2.2. Spent Electrolytes

The recovered electrolyte samples from used LIBs were obtained by third parties and various battery sources, all in the context of commercial LIB recycling. Hence, detailed information on lifetime and life cycles was not always available. Presumably, Samples A and B originate from test cells with only a few cycles and Sample C_O from a mixture of EOL LIBs, test cells, and production scrap. The recovered samples were stored at 4 °C under atmospheric conditions.

Sample A, see Figure A1, was obtained from a prismatic cell with a nickel-manganese-cobalt (NMC) chemistry on the cathode side and a graphite chemistry on the anode side. Approximately 5 mL of surplus electrolyte directly accessible after opening the cell by cutting the housing under inert conditions in a glovebox was removed to obtain this sample.

Two electrolyte samples were obtained from the mechanical-thermal pre-treatment of the recycling route: Sample B, see Figure A1, was obtained from a pouch cell (NMC/graphite chemistry), which was shredded first and subsequently subjected to vacuum-drying in a laboratory scale under an inert atmosphere. The volatile compounds were recovered by means of condensation (approximately 10 mL).

A total of 5 L of Sample C_O, see Figure A1, was obtained from a pilot LIB recycling facility, and the cell chemistry was largely unknown; however, it can be assumed that most of the material originated from modules that had an NMC/graphite chemistry. All modules were shredded and treated in a similar manner to Sample B. Due to the vacuum-drying in an inert gas atmosphere followed by condensation, the recovered samples contained only volatile organic compounds, as the conduction salts remain in the solid fraction in this process.

Due to the absence of conduction salt and the large volume of accessible samples, approximately 400 mL of Sample C_O was distilled under atmospheric conditions in a laboratory setup located within a fume hood. This setup comprised a heating plate, a flat bottom flask containing the sample, a magnetic stirring bar, a condenser operated with a water cooler, and a flat bottom flask to collect the condensate. See Figure A2 for a visual representation of this setup. After distillation, approximately 390 mL were collected as distillate (Sample C_D, see Figure A1) and 8 mL as residue (Sample C_R, see Figure A1). All samples were prepared as outlined in Section 2.5 and analyzed by GC-FID and GC-MS (Section 2.3 and Section 2.4, respectively).

2.3. GC-Method

In order to perform a quantitative analysis of the specified components, the following GC method was developed. GC SCION 45 (SCION Instruments, Goes, The Netherlands) equipped with an FID, combined with the autosampler Combi PAL-XT (CTC Analytics, Zwingen, Switzerland), was utilized via the SCION Instrument CDS software, version 3.0.1.16. A Rxi-624Sil MS Capillary column (30 m, 0.25 mm ID, 1.4 μm), provided by Restek (Bad Homburg, Germany), was used for this study. Prior to and after each injection, the 1.2 µL syringe was rinsed with acetone in order to prevent cross-contamination between samples. The 1177-type injector was set at a temperature of 280 °C using a split-liner (borosilicate) HP 78.5 × 6.3 × 4 mm and a split ratio of 1:200. The injection volume was 0.2 μL. Nitrogen was employed as the carrier gas with a flow rate of 0.63 mL·min⁻¹. The temperature program was 60 °C (1 min)—20 °C·min⁻¹—200 °C (0 min)—40 °C·min⁻¹—260 °C (5 min). FID temperature was maintained at 350 °C, while gases were set at 300 mL·min⁻¹, 30 mL·min⁻¹, and 25 mL·min⁻¹ for synthetic air, hydrogen, and nitrogen (makeup gas), respectively.

2.4. GC-MS Identification of LIBs Compounds

For the identification of unknown components, GC-MS measurements were conducted using an Agilent Technologies 6890 N (Agilent Technologies Deutschland GmbH, Böblingen, Germany) gas chromatograph coupled to an Agilent Technologies 5975B mass spectrometer (Agilent Technologies Deutschland GmbH, Böblingen, Germany). GC-column and temperature program was identical to the GC-FID method (Section 2.3), but helium was used as the carrier gas with a flow rate of 1.0 mL·min⁻¹. Split ratio was set at 1:20, and the injection volume was 1 µL. The MS temperatures of the analyzer and source were set at 150 °C and 230 °C, respectively, and the mass range was recorded from 29 to 450 m/z. The identification of separated components was achieved by comparing spectra using the NIST23 library program and spectra database.

2.5. Sample Preparation

For validation of the GC method (see Section 2.6), samples were prepared with synthetic substances in accordance with Section 2.1. To initiate the validation process, each pure substance (except for EC) was individually measured without further preparation by GC-FID. Due to the normal melting point of EC at 36.4 °C, EC was previously dissolved in EMC. Subsequently, a standard mixture comprising all components was prepared for the purpose of testing the selectivity of the method. This mixture was then combined with the internal standard (InSt) AcN in a 2:1 solvent-to-acetonitrile volumetric ratio and analyzed by GC-FID. Additionally, LP57 was also prepared and analyzed under atmospheric conditions in the same way to investigate whether the LiPF₆ contained therein had any influence on the analytic procedure and results.

In order to conduct calibration and linearity experiments without the use of a diluent, stock solutions (StSo) of the binary mixtures MeOH/EtOH, FB/DEC, and DMC/EMC were prepared in the range of 0.1–95 g_i/g_tot,StSo. Moreover, stock solutions of VC/FEC/EC/PC in the range of 0.1–50 g_i/g_tot,StSo were prepared and, if necessary, filled with DMC. Additionally, stock solutions for CB/CHB in the range of CHB = 0.1–50 g_CHB/g_tot,StSo were prepared. CB, which exhibited no reaction with the other compounds mentioned and is a well-known component in the field of distillation, was incorporated into the quantification process as a planned component for future processing steps in a distillation procedure. Subsequently, the stock solutions were prepared with the InSt AcN at a 2:1 solvent-to-acetonitrile volumetric ratio per vial. To extend the range of the calibration, pure MeOH, EtOH, CB, DMC, EMC, DEC, and FB were also prepared in a volumetric ratio of 2:1, 2.1:1, and 2.2:1, respectively, thus achieving a total range of these components in the vials of 0.067–0.6875 g_i/g_tot.

The masses of all stock solutions and samples in the vials were determined gravimetrically using a PRACTUM 224-1S balance (Sartorius Lab Instruments, Göttingen, Germany) and were prepared using a HandyStep touch repetitive pipettes (Brand, Wertheim, Germany).

As the recovered samples derived from spent LIBs may contain acids with the potential to cause damage to the GC system, the acids were neutralized prior to analysis by the addition of NaHCO₃. To ensure the neutralization with NaHCO₃, a series of preliminary tests were carried out. For the neutralization, a spatula tip of NaHCO₃ was added to approximately 5 mL of each sample, mentioned in Section 2.2. If a gas release were observed, this procedure was repeated until no more gas was released. This resulted in the formation of carbon dioxide, a sodium salt of the acid as a solid, and water as a reaction product. Excess sodium hydroxide carbonate and potential sodium salts were separated by centrifugation. The prepared samples were mixed in triplicates with the internal standard acetonitrile in a volumetric ratio of 2:1 and measured by GC-FID. In order to investigate the influence of neutralization on the analysis, a second measurement of the standard mixture was conducted. This was prepared in the same way by adding NaHCO₃ and then analyzed by GC-FID in comparison with the initial measurement.

Prior to switching to no diluent, GC-MS measurements were conducted. Consequently, the preparation of samples for GC-MS measurements differed from the previously described procedure. Instead, 100 μL of the prepared electrolyte samples were diluted in 1250 μL methanol by adding 150 μL of an InSt, which had a concentration of 30 mg/mL acetonitrile diluted in MeOH.

2.6. GC-Validation Procedure

The method was validated in accordance with ICH R2(Q2) [39] and FDA [40] guidelines. Therefore, selectivity, linearity, range, limit of quantification (LOQ), limit of detection (LOD), accuracy, and precision were determined.

Selectivity of the method was evaluated in conjunction with its system suitability. In order to achieve this, blank samples of pure acetone, pure AcN, as well as the samples of each substance and the sample containing all substances, see Section 2.5, were injected and measured in triplicate. The clear separation and unambiguous assignment of the peaks were assessed based on the resolution RS, with a satisfactory resolution defined as RS > 1.5 [41]. Additionally, the asymmetry factor and number of theoretical plates were determined according to USP 621 [42].

The linear dependence of the signal (ratio of analyte peak area to the peak area of InSt A_i·A_AcN⁻¹) in relation to the ratio of the mass of the analytes to the mass of the InSt m_i·m_AcN⁻¹ was evaluated. For the purpose of analysis, different mass ratios were determined by quintuplicate measurements, and Equation (1) was applied to these values.

A_i · A_AcN⁻¹ = a · m_i · m_AcN⁻¹ + b(1)

The slope a and y-intercept b are the regression parameters. In order to ascertain the presence of linearity, the coefficient of determination, R², is employed.

The range is defined by the upper and lower bounds of the calibration, which are determined by the linearity of the set mass ratios. Due to the wide concentration ranges, the calibration was divided into three categories: low, medium, and high. Each category was assigned its own calibration curve.

LOD and LOQ were calculated in accordance with the Guidance Document on the Estimation of LOD and LOQ for Measurements in the Field of Contaminants in Feed and Food [43]. The limits were established using a sample in which all solvents were dissolved in pure CB, with a mass lower than the minimum mass required for the calibration range. The stock solution was prepared once, and the samples for LOD determination were prepared 10 times individually, with each sample measured once. LOD and LOQ were calculated using Equations (2) and (3), with the slope a from the calibration and the standard deviation s_y,b from these measurements.

(2)$LOD=3.9\cdot {\displaystyle \frac{{\mathrm{s}}_{\mathrm{y},\mathrm{b}}}{b}}$

(3)$LOQ=3.3\cdot LOD$

Accuracy and precision of the method were evaluated using the recovery percentage %R_i, defined as the ratio of the measured value m_i,meas to the theoretical value m_i,th., Equation (4), and the relative standard deviation (RSD). %R_i and RSD were determined for nine separate solutions in which the components were mixed together in different mass ratios and combinations. Additionally, the sample, which contains all the aforementioned components, was used for this purpose. From these solutions, five separately applied samples were each measured.

(4)$\%R_i={\displaystyle \frac{m_{i,meas.} }{m_{i,th.}}}$

3. Results and Discussion

3.1. Method Development

In the course of method development, a number of different solvents was tested for sample dilution, including methanol, ethanol, chlorobenzene, toluene, and others. The GC-separation of most of these solvents would have been suitable because there was no chromatographic overlap between the solvent and the target analytes. However, most of the solvents led to highly asymmetric peaks of the target components. Only the use of methanol or ethanol as a solvent, which eluted earlier than the target compounds, resulted in symmetrical and reproducible target peaks.

However, it was noticed that methanol-diluted samples with a high EMC and low DMC content (stored for extended periods of time) contained ethanol. This observation can be explained by the transesterification of EMC with MeOH to form DMC and EtOH, Equation (5) [35].

C₄H₈O₃ + CH₃OH ⇌ C₃H₆O₃ + C₂H₅OH(5)

Pilot experiments confirmed that samples that had been dissolved in methanol showed an ethanol content of <1.0 g_EtOH/g_tot, whereas the samples that had not been dissolved in a solvent exhibited no or significantly lower DMC and EtOH. To avoid such solvent-derived alterations of the samples, the method was revised and changed to direct injection of undiluted samples. Since GC-FID is an analytical technique for trace analysis, direct injection of undiluted samples is rather unusual. However, by decreasing the injection volume to 0.2 µL and increasing the split ratio to 1:200, peak shape and peak size could be obtained in the usual dimensions. Furthermore, the only sample preparation step was the addition of the InSt AcN. Utilization of nitrogen as a mobile phase keeps costs within a manageable range. The method has undergone a validation procedure due to the absence of a solvent, which is a rather unconventional technique for GC analysis.

3.2. Method Validation

3.2.1. Selectivity and System Suitability

The chromatogram, see Figure A3, of the sample, which was composed entirely of all synthetic components, indicated the absence of degradation products and demonstrated that the resolution of all components was at least RS > 3 and that the number of theoretical plates (NTP) was >170,000. No effect of NaHCO₃ addition was observed (Figure A3). During the linearity measurements, the asymmetry factors for all components were found to range between 0.73 and 1.52.

As shown in Figure A4, the chromatogram of LP57 revealed the presence of multiple peaks, of which 99.5% of the peak area is attributed to AcN, EMC, and EC. The presence of trace peaks has also been observed, and it has been theorized that these are generated during preparation under atmospheric conditions or in the GC itself due to decomposition reactions involving LiPF₆. As this method only leads to a small proportion of decomposition products, as can be seen from the peak area fraction of the known components, it is also capable of analyzing samples containing conduction salt. However, if decomposition reactions are to be avoided completely, it is necessary to handle the samples in an inert atmosphere and remove the dissolved conductive salt prior to analysis.

3.2.2. Linearity, Range, LOQ, and LOD

Given the large calibration range, the process was subdivided into two or more calibration curves, with each calibration line comprising a minimum of five calibration points. Table A1 presents the range of each calibration, the calibration parameters, the coefficient of determination R², and the LOQ and LOD. For each individual calibration, R² exceeds the recommended threshold value of 0.99, as suggested by [41]. Moreover, Figure 1 illustrates the calibration results as example for the solvents EMC and DMC. The highest LOD were identified for DMC with a value of 0.4189 mg_DMC·g_AcN⁻¹.

3.2.3. Accuracy and Precision

The accuracy and precision for the low and medium calibration curves were checked at least once for all components. This resulted in 54 measurement results, which are presented in Table A2. Of these measurements, which also include the sample of the standard mixture prepared with NaHCO₃, 52 show a %R in the range 90–110, of which 44 are in the range 95–105. Only two measurements showed a %R > 110. The lowest recovery rate was observed for low calibrations of FB, with a value of %R = 90.39. In contrast, the highest recovery rate was observed for the medium calibration of EC, with an approximate value of %R = 116. The RSD for all components was ≤6.01%, except for two measurements for EC and CHB, where the RSD was <6.70%.

3.3. Composition and Impurities of Spent Electrolytes

A total of 106 peaks were identified across all samples, of which 42 could be unambiguously assigned based on the MS spectra and library spectra (NIST23) comparison using a match factor >800 as the criterion for identification. In general, it should be noted that in 14 cases, the assignment of certain peaks did not meet this level of certainty. However, these peaks could be adequately proposed based on literature data. A total of 50 peaks could only be characterized insufficiently and were registered as unknowns. These unknowns were only observed as trace components. If it turns out that one or more of these components play an important role in the recycling process, they would have to be specifically enriched in order to elucidate the structure of the component. Currently, there is no evidence for this.

The analysis of the samples revealed similarities in their composition and the identified substances. Upon neutralization of the samples with NaHCO₃, no visible gas formation was observed in any sample, which leads to the conclusion that the samples have no or very low acid content. It is evident that the procedure described above does not allow for the ascertainment of acidity levels. Consequently, any attempt to determine acidity levels would have to be conducted in a supplementary manner using a titration system, for example. Furthermore, all samples were found to contain EtOH, MeOH, DMC, EMC, DEC, and EC. The presence of carbonates can be attributed to their utilization as a solvent for the conducting salt. It seems feasible that the low concentration of some solvents, particularly DMC and DEC, may not have originated from their initial use as solvents but rather as a decomposition product within the cell during the recycling process or as a consequence of storage of the recovered electrolyte under ambient conditions. Matadi et al. [44] observed the presence of DMC and DEC during calendar ageing where initially the electrolyte solvent EMC/EC was used. This phenomenon can be explained by the reduction of EMC on lithiated graphite surface [44,45]. The aforementioned reactions (Equations (6) and (7)) were published by Kim et al. [45], which demonstrated that EMC decomposes to an alkoxide like (MeO⁻ or EtO⁻) and an alkyl formate radical. This can lead to a reaction between the alkoxides and EMC to form DMC or DEC and the other alkoxide (EtO⁻ or MeO⁻).

C₄H₈O₃ + e⁻ → CH₃O⁻ + C₃O₂H₅(6)

C₄H₈O₃ + e⁻ → C₂H₅O⁻ + C₂O₂H₃(7)

This provides an explanation for the low content of DMC and DEC, as well as the presence of MeOH and EtOH in the samples. Another route can be the disproportionation of EMC to DMC and DEC or the transesterification of EMC and MeOH or EtOH to DMC and EtOH or DEC and MeOH. See Equations (5), (8), and (9) [35]. Moreover, water, for instance, from the ambient environment can react with EMC due to inadequate storage conditions, resulting in the contamination with degradation products, including DMC, DEC, MeOH, and EtOH [46].

2 C₄H₈O₃ ⇌ C₃H₆O₃ + C₅H₁₀O₃(8)

C₄H₈O₃ + C₂H₅OH ⇌ C₅H₁₀O₃ + CH₃OH(9)

The two-phase formation in Sample B indicates the inclusion of water. The carbonates exhibit a mixing gap with water [47]. The aqueous droplet, which measured approximately 5 mm in diameter, was too small to be analyzed. To determine the water content in each sample, Karl Fischer titration (KFT) could be performed.

The calibrated components PC, FEC, and VC were suspected but not identified in any of the samples. This can be attributed to the fact that they were not initially utilized for the electrolyte or due to their decomposition during SEI formation. The former option is more plausible in the case of PC, given that PC has the potential to cause the exfoliation of graphene on the anode side. In contrast, EC has been demonstrated to form a more stable SEI protective film on graphitic anodes in comparison to PC [6,7]. Consequently, the detection of EC in all samples provides compelling evidence that no additional PC was utilized. In the case of VC and FEC, which are used as SEI building additives, it is more probable that if they were initially used, these additives were fully decomposed during the formation process.

3.3.1. Sample A—Electrolyte Directly Out of Prismatic Cell

The electrolyte sample exhibited a brown hue (see Figure A1) and displayed a total of 25 peaks combined in the GC-MS and GC-FID measurements, as illustrated in Figure 2 in partial scale as well as in Figure A5 in full scale and listed in Table A3. Of these, 11 substances could not be further identified. It should be noted that the chromatograms of the GC-MS and GC-FID measurements can be aligned, allowing for the clear assignment of individual peaks in the unaltered (undiluted) GC-FID run using the identification data from the GC-MS analysis (Figure 2). Nevertheless, there are isolated cases in which peaks are observed in only one of the two analytical approaches. To illustrate, substances with a retention time of less than 3.4 min in the FID are undetectable in the MS because of the solvent delay feature used in GC-MS to protect the filament and the detector from damage due to the excessive solvent content. Additionally, acetone can only be detected in the FID since it was only employed for GC-FID as a rinsing solvent. Moreover, Sample A displays Peak 7 in the GC-FID measurement, whereas Peaks 15 and 24 are only present in the MS measurement, neither of which are observed in the other chromatograms. As a result, those remain unassigned.

The GC-FID measurements demonstrate a peak area fraction of the quantifiable components (knowns) of 97.8% (including AcN). This finding, at least qualitatively, suggests that the quantity of non-quantifiable components is, for the moment, negligible, with a peak area fraction of 2.2%. The analysis of Sample A, conducted using the calibrated GC-FID method, revealed the presence of DMC (375.57 ± 1.78 mg·mL⁻¹), EMC (312.64 ± 2.60 mg·mL⁻¹), DEC (9.05 ± 0.10 mg·mL⁻¹), EC (322.11 ± 11.07 mg·mL⁻¹), and CHB (39.35 ± 0.86 mg·mL⁻¹). The elevated concentration of DMC, EMC, and EC indicates that a solvent mixture of these three components was utilized within the cell, with a probable ratio of DMC:EMC:EC of 1:1:1. It seems reasonable to postulate that CHB was added as an additive and that the low proportion of DEC was the result of the aforementioned reaction of DMC or EMC during the ageing of the cell. This also provides an explanation for the low concentration of MeOH.

In the case of Sample A, the presence of a dissolved conducting salt, which is expected to remain in the sample due to the nature of the recovery process, suggests that reactions during storage and handling under atmospheric conditions or in the GC with the setup method may influence the chromatogram. This is evident in the comparison with the chromatograph of LP57 (Figure A4). The occurrence of Peaks 1, 2, 7, 11, 14, 18, and 20 in Sample A is consistent with the peaks observed in LP57, suggesting they may be attributed to decomposition products resulting from sample handling or GC-analysis. While the treatment presented here of samples containing conduction salt leads to an increase in conversion components, these only contribute a small proportion to the total peak area, as already explained in Section 3.2.1. Thus, it can be deduced that the decomposition products occurring in traces only have a minor influence on the concentrations. However, in instances where a comprehensive characterization of a sample containing conduction salts is required, the aforementioned changes in sample preparation and treatment (see Section 3.2.1) and the use of ion chromatography (IC) are nevertheless valuable.

Furthermore, analysis identified the presence of trimethyl phosphate (TMP) and an R-dimethyl phosphate. The most probable match in the database for the residue was identified as an R=ethyl or butyl. TMP is employed in LIB as a flame-retardant additive [48,49] and also occurs as an ageing product. This is demonstrated by Weber et al. [34] in the ageing of LIB cells under defined conditions and thermal treatment at 95 °C. However, as TMP has the same retention time as a peak in the LP57 analysis, it is possible that this peak is also TMP and the thermal treatment in the GC could lead to the formation of TMP and others. The formation of TMP and other phosphates, such as R-dimethyl phosphate with two methyl groups and one residue, can be explained by successive reactions of phosphorus fluoridates with organic carbonates. Alternatively, TMP may be formed in the reaction of phosphorus fluoridates with MeOLi, which results in the release of LiF and the subsequent formation of TMP [22,23].

Similarly, Lemordant et al. [50] identified the presence of oligoethylene glycols in the analysis of the electrolyte following its formation. One of the mass spectra presented in the study is analogous to the spectrum of Peak 21 and was also attributed to dimethyl diglycol carbonate. However, Lemordant et al. posit that this spectrum is more consistent with oligocarbonates. Nevertheless, the occurrence of diglycolcarbonates cannot be discounted. This is because diglycolcarbonates are well documented in the context of electrolyte ageing, as evidenced by the studies of Takeda et al. [29] and Gachot et al. [23]. Takeda et al. [29] identified the presence of diethyl diglycolcarbonate in the cell ageing of a DEC/EC electrolyte. In contrast, Gachot et al. [23] identified dimethyl diglycol carbonate in an EC/DMC electrolyte derived with a reaction scheme based on oligocarbonates, including diethyl-2,5-dioxahexane carboxylate (DEDOHC). It should be noted that oligocarbonates, including dimethyl-2,5-dioxahexane carboxylate (DMDOHC), ethylmethyl-2,5-dioxahexane carboxylate (EMDOHC), and DEDOHC, have been identified as ageing products in LIB electrolytes [19,23,24,25,27,29,51]. This hypothesis is also supported by the occurrence of Peak 22, DEDOHC. This is due to the fact that DEDOHC can be formed from dioxahexane carboxylate compounds in conjunction with the action of lithium alkoxides and the presence of linear organic carbonates [23]. Additionally, a nucleophilic attack of a lithium alkoxide on EC and a linear carbonate may occur [51]. In a recent study, Hofmann et al. (2023) [24] examined the formation of dicarbonates, including DMDOHC, DEDOHC, and EMDOHC. The researchers investigated the formation of these dicarbonate products both with and without the presence of conductive salts, such as LiPF₆. Their findings suggest that the formation of these dicarbonates is enhanced in the presence of conductive salts.

A comparison with the database for Peak 25 did not permit a clear assignment to one substance while indicating a match factor >900 for two possible compounds. The comparison with 2-vinylnaphthalenes, as well as with biphylene (BP), demonstrated a high degree of spectral agreement. Nevertheless, a further occurrence of 2-vinylnaphthalenes from spent LIBs is not documented in the existing literature. Conversely, the utilization of BP as a flame-retardant additive in LIBs is well known. Consequently, the impact of BP has been subjected to comprehensive examination within the existing literature [44,52,53,54,55]. For example, Lee et al. [52] demonstrate that BP can be employed as a stand-alone agent or in conjunction with other additives, such as CHB. The combination of BP and CHB has been demonstrated to offer enhanced overcharge protection compared to the use of either compound alone. Consequently, BP is frequently employed in LIBs and is regarded as a more plausible candidate for Peak 25. Therefore, it can be assumed that a combination of BP and CHB was initially utilized in this cell.

Additionally, octadecamethylcyclononasiloxane (Peak 24) was identified, yet the match (793) is insufficient for a definitive assignment, and this substance is not documented in the literature in relation to LIBs. Nevertheless, it appears to be a siloxane, which was also identified in Sample B and is discussed in greater detail there.

3.3.2. Sample B—Condensate from Mechanical-Thermal Recycling Route in Laboratory Scale

The sample resulting from the mechanical-thermal drying step during the recycling of a LIB on a laboratory scale exhibits 13 peaks, displayed in Figure 3 in partial scale as well as in Figure A6 in full scale and Table A4. Nine of these can be identified. These include DMC (2.67 ± 0.03 mg·mL⁻¹), EMC (964.62 ± 12.67 mg·mL⁻¹), DEC (10.75 ± 0.19 mg·mL⁻¹), and EC (10.40 ± 7.77 mg·mL⁻¹). In this instance, the peak area fraction of quantifiable components was found to be 99.97%, a value that can be attributed to three factors. Firstly, there was a low number of non-quantifiable peaks. Secondly, there was the method of recovery, where the conduction salt and components with low saturation vapor pressure and low concentration remain in the solid fraction of the shredded LIB. The absence of conduction salt has been identified as a factor that reduces the likelihood of decomposition products. Additionally, the utilization of the LIB may be restricted to a test level, with a reduced number of cycles, as outlined in Section 2.2, thereby leading to a reduced generation of decomposition products.

In comparison to Sample A, the EC content is notably reduced. This is due to the fact that the recovery via drying enables low-boiling components, in particular, and heavier-boiling components, such as EC or CHB, to remain in the solid fraction of the crushed LIB [18]. Based on the available information, it can be concluded that a solvent mixture comprising EMC and EC was employed for the cell. The low proportions of DMC and DEC suggest that the mentioned transesterification of EMC occurred.

Moreover, ethylene glycol (MEG) was identified in the MS sample. This can be attributed to inadequate storage of the sample, resulting in water retention and subsequent hydrolysis of EC [36,46,56], or to the addition of methanol and the transformation of EC to EG and DMC [57,58].

In addition, octamethylcyclotetrasiloxane has also been identified in the MS sample. In general, siloxanes can be added to the electrolyte to facilitate the formation of interfaces between the electrolyte and the cathode and anode surfaces, among other functions [59]. Research is also being conducted on octamethylcyclotetrasiloxanes, which Wang et al. [60] added as an additive in an EC:EMC:DMC electrolyte. In addition to their use in additives, siloxanes such as dodecamethylcyclohexasiloxanes have also been detected in investigations of thermal runaways on LIBs in particulate recoils. It is assumed that this is due to the used instruments. In this instance, the identified siloxanes may also manifest in the chromatogram through column bleeding. Both are possible explanations for the occurrence of siloxanes in Sample A.

3.3.3. Sample C—Condensate from Mechanical-Thermal Recycling Route in Pilot Scale

A total of 68 peaks were identified in the chromatograms of the three samples C_O, C_D, and C_R, as illustrated in Figure 4 in partial scale as well as in Figure A7 in full scale. Of these, 25 could be assigned using the NIST database with good agreement; 8 showed moderate agreement with the database. However, the proposed structures could be justified with additional literature data, and 33 peaks could not be assigned (see Table A5).

The original Sample C_O exhibited 23 peaks, in addition to the known quantifiable components MeOH (1.23 ± 0.02 mg·mL⁻¹), EtOH (1.03 ± 0.01 mg·mL⁻¹), and DMC (52.48 ± 0.23 mg·mL⁻¹), FB (48.08 ± 0.22 mg·mL⁻¹), EMC (831.45 ± 6.06 mg·mL⁻¹), DEC (21. 58 ± 0.21 mg·mL⁻¹), EC (10.87 ± 0.09 mg·mL⁻¹), and CHB (1.07 ± 0.01 mg·mL⁻¹). The analysis also encompassed 3 unknowns and 12 other substances.

In comparison to Sample B, a reduced concentration of EMC and a comparable EC concentration can be recorded. The higher concentrations of MeOH, EtOH, DMC, and DEC in Sample C_O can be attributed to two reasons. Firstly, the use of a different initial electrolyte is likely due to the appearance of FB. Secondly, the recycling of multiple LIBs, which may differ in initially used electrolytes as well in lifetime and life cycles, can also explain the higher number of impurities. However, this sample probably represents a more realistic scenario, as it will be observed in the future for the recycling of LIBs of different lifetimes and manufacturing technologies.

After distillation of the original Sample C_O leading to the distillate, Sample C_D, and the residue, Sample C_R, alterations were observed in both samples. These are demonstrated in the material balance presented in Figure 5. The distillate exhibited an elevated concentration of low-boiling components (e.g., EMC, reaching 866.72 ± 9.61 mg·mL⁻¹), along with a complete transfer of light-boiling components (e.g., FB, benzene, or DMC) to the distillate. Conversely, the residue C_R demonstrated an enrichment of high-boiling components (e.g., EC or CHB). This also resulted in a significant number of components that were not previously detected because they were below the LOD in Sample C_O but became quantifiable after distillation due to the concentration increase in Sample C_R. While the proportion of quantifiable components in Sample C_O is 99.7%, there was a slight increase to 99.8% for Sample C_D and a decrease to 90.6% for the residue, Sample C_R, suggesting an accumulation of impurities in the residue, as previously mentioned. Consequently, 63 peaks were identified in the residue, Sample C_R, and peaks that had previously been unclassifiable in Sample C_O could be assigned to individual substances in Sample C_R due to the presence of clear mass spectra and identical retention times, as depicted with the chromatograms displayed one above the other in Figure 4 in partial scale as well as in Figure A7 in full scale.

The elevated number of peaks observed in Sample C_R is attributable to the fact that this sample originated from a pilot plant where a substantial quantity of spent LIBs were recycled in batch, and the impurities were subsequently concentrated through distillation. It can be assumed that the additional heat input during distillation led to further decomposition of the solvents and an increase in decomposition products. To avoid the accumulation of decomposition products during the electrolyte purification process, a more thermally gentle treatment, such as distillation or evaporation under vacuum conditions, would be preferable.

Examination of the identified components reveals the presence of trimethoxyborane (TMB) and tert-butylbenzene, which both can be added to the electrolyte as additives. TMB has been shown to enhance the stability of the interface between the electrode and electrolyte at high voltages, while the formation of a protective layer on the cathode surface inhibits the decomposition of the electrolyte and prevents the structural degradation of the electrode [61,62]. In general, tert-alkylbenzenes can be employed in combination with biphenyl, resulting in enhanced overcharge inhibition [63]. This is consistent with the observation of a peak in Sample C_R, which could be attributed to both biphenyl and 2-vinylnaphthalene, as discussed in Section 3.3.1.

In addition to tert-alkylbenzene, further substances containing benzene rings were identified in the samples, including benzene, fluorostyrene, BP, 2-vinylnaphthalene, and 5-methylquinoline. Benzene can be introduced either through the addition of an electrolyte [10] or as a result of a decomposition reaction in the electrolytes themselves [64]. Benzene was identified in the investigation of thermal decomposition by Chen et al. [64]. The occurrence of benzene can be explained by a decomposition reaction from EC, PC, and DEC to ethene and methyl-ethene via ethyne and methyl-ethyne, which then undergoes further decomposition to form benzene and methyl-benzene. An alternative hypothesis is that FB releases fluorine through cell ageing reactions during SEI formation, resulting in the incorporation of LiF into the SEI and the formation of benzene in the electrolyte. The occurrence of fluorostyrene is likely to be derived from a reaction based on FB, although this has not yet been documented in the literature, as little information on the influence of FB is available in the public domain.

Additionally, the presence of dioxanes and dioxalanes was identified by mass spectrometry. The presence of 1,4-dioxanes is identified in two distinct contexts in literature: during the thermal ageing of EMC/EC (1 mol LiPF₆) electrolytes [32] and on the anode and cathode sides, where the electrolyte was to be washed out in both cases [30]. Furthermore, Gachot et al. [22] identified the presence of 2-methyl-1,3-dioxalane in addition to 1,4-dioxanes, even after the cycling of cells. Although the latter was not detected in the present study, it only shows a longer alkyl residue up to 2-ethyl-1,3-dioxalane. This indicates that comparable reaction mechanisms for the formation of dioxalanes, as well as dioxanes such as 2,5-dimethyl-1,4-dioxanes, may occur within the cell.

The electrolyte samples were found to contain several other ester compounds, including methyl octanoate, ethyl octanoate, an isomer of octyl cyclohexane, 1-methoxy-2-propyl acetate, 2-hydroxyethyl octanoate, and butyl caprylate. The group of esters is well documented in the context of aged electrolytes in LIBs, with the majority of research focusing on short-chain ester compounds. A reaction of linear carbonates, such as DMC, with a lithium cation can result in the formation of radicals, including the methyl radical and the methyl carboxyl radical [22]. Similarly, it is conceivable that ethyl radicals and ethyl carboxy radicals are formed by the living reaction mechanism with EMC or DEC. Consequently, alkyl radicals, hydrogen ions (hydronium ions), and carboxy radicals may undergo a reaction to form acetate or (methyl) formate. It can be assumed that ethene, which can be formed during the reaction of EC with Li+ [23,26], is attacked by the radicals. This could result in a chain elongation of the carboxy radical, which can be terminated by the reaction with an alkyl radical or H+. This would explain the occurrence of the majority of the listed esters [65]. In addition, long-chain ester compounds were identified, which were attributed to a similar rationale and a reaction with ethene.

Although there is considerable overlap between the components listed here and those identified in the literature, the origin of certain components, such as 1,1-dimethoxypropane, 1-butanol, 1,1-dimethoxy-2-methylpropane, 1-butoxy-2-propanol, ethyl butyl carbonate, and 4,4-dimethyl-2-oxazolidinone, remains unexplained. Despite an extensive literature search, no evidence of corresponding ageing processes could be found.

The analysis of Sample C_O illustrates the difficulties associated with the preparation of recycled electrolytes. The impact of individual impurities remains unknown; therefore, the purification of target solvents, such as linear carbonates or FB, requires further investigation.

4. Conclusions

This contribution presents a new GC-FID and GC-MS method capable of identifying the organic components of recovered electrolytes originating from spent LIBs.

The method was applied to three samples of spent electrolytes, of which Sample A was recovered directly from the cell and Sample B and C_O were recovered from shredded LIB scrap through vacuum-drying in an inert gas atmosphere. In addition, Sample C_O from a recycling plant in pilot scale was distilled, resulting in a distillate (Sample C_D) and a residue fraction (Sample C_R).

The analyses enabled the formulation of conclusions regarding the solvents and additives employed. Quantification and measurement of some of the most-used solvents for LIB electrolytes, EMC, DEC, and DMC and EC were conducted in all samples. The analysis revealed that EMC was present in the highest concentrations in almost all original samples. Although DMC and DEC were detected in the samples, this can be partly attributed to a decomposition reaction due to their low concentrations in Samples B and C_O. The presence of MeOH and EtOH, which are also present in traces in all samples, can be explained in a similar manner. In summary, conclusions can be drawn about the solvents used, but these are also strongly influenced by the recovery method of the electrolyte samples. For example, the comparatively low EC content in Samples B and C_O can be explained by the fact that, due to the low saturation vapor pressure of EC, the drying process during vacuum-drying is insufficient; therefore, less EC is present in the recovered electrolyte sample. The presence of common additives such as CHB or BP, along with the occurrence of decomposition products, which are formed during cell ageing, including methyl octanoate, ethyl octanoate, and DEDOHC, has been proven. Notably, the presence of FB and benzene-ring-containing substances in Sample C_O, which are absent in all other samples, is an important finding.

Additionally, some components remain unidentified, and their provenance may be traced back to formed components inside the LIB, alternatively attributed to the handling of the samples and the conditions in the GC, or perhaps have been formed during the distillation process. While the analytics employed offer insight into the composition of the recovered electrolyte samples and draw attention to potential processing challenges, it should be noted that further analyses and, if necessary, further analysis techniques are necessary for a comprehensive characterization of samples from spent LIB. These may include IC to determine the salt content, KFT to determine the water content, or an enrichment for specific, as yet unknown, contaminations to elucidate the structure of the components. These investigations are an important step to learn how complex these recovered electrolyte mixtures from spent LIBs are, elucidate the structures of the main components, and determine what processes might be involved in the formation of degradation products and trace components. This was particularly evident for Sample C_D obtained by distillation. The sample exhibited a distinct chromatogram, indicating that the process of distillation or rectification may facilitate the recovery and reuse of electrolyte solvents. Moreover, the retention times of certain components, including high-boiling DEC, EC, and CHB, provide further support for this conclusion. To reuse the target solvents, it is necessary to separate alcohols such MEOH and EtOH from decomposition reactions in addition to other impurities. The distillation of Sample C_O enabled the identification of some of these components, with most of these components being primarily impurities and consistent with the findings of existing literature.

Finally, a comparison of all samples revealed considerable variations in their composition and the components present, which provides an outlook of what can be expected from future large-scale industrial LIB recycling processes. Studies like those presented here will help to find out whether a single general process can be carried out to enable sustainable reuse of electrolyte materials in new battery cells. The analytical techniques presented here will not only help in the development and optimization of such processes but will also be essential in ensuring the quality of the recycled materials.

Author Contributions

Conceptualization, M.W., K.J., T.B. and S.S.; Data curation, M.W.; Formal analysis, M.W. and K.S.; Funding acquisition, K.J. and S.S.; Investigation, M.W., K.S., K.A., P.M. and T.B.; Methodology, M.W., K.S., K.A. and P.M.; Project administration, K.J.; Resources, M.W.; Supervision, K.J. and S.S.; Validation, M.W., K.S. and P.M.; Visualization, M.W.; Writing—original draft, M.W.; Writing—review & editing, M.W., K.J., T.B. and S.S. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

The original contributions presented in the study are included in the article; further inquiries can be directed to the corresponding author.

Acknowledgments

Special thanks to Steffen Fischer of the TU Braunschweig, Institute of Particle Technology, for providing the samples.

Conflicts of Interest

The authors declare no conflicts of interest.

Nomenclature

Symbol	Description
Ai	peak area from component i
a	slope
b	y-intercept
mi	mass of component i
mi,meas	measured value (mass) of component i
mi,th	theoretical value (mass) of component i
R²	coefficient of determination
RS	resolution
sy,b	standard deviation
%Ri	recovery percentage of component i
Abbreviations	Description
AcN	acetonitrile
BEVs	battery electric vehicles
BP	biphylene
CB	chlorobenzene
CHB	cyclohexylbenzene
DEC	diethyl carbonate
DEDOHC	diethyl-2,5-dioxahexane carboxylate
DMC	dimethyl carbonate
DMDOHC	dimethyl-2,5-dioxahexane carboxylate
EC	ethylene carbonate
EMC	ethyl methyl carbonate
EMDOHC	ethylmethyl-2,5-dioxahexane carboxylate
EOL	end-of-life
EtOH	ethanol
FB	fluorobenzene
FEC	fluoroethylene carbonate
FID	flame ionization detection
GC	gas chromatography
HF	hydrofluoric acid
IC	ion chromatography
InSt	internal standard
KFT	Karl Fischer titration
LIB	lithium-ion battery
LiF	lithium fluoride
LiPF6	lithium hexafluorophosphate
LOD	limit of detection
LOQ	limit of quantification
MEG	ethylene glycol
MeOH	methanol
MS	mass spectrometry
NaHCO3	sodium hydrogen carbonate
NMC	nickel-manganese-cobalt
PC	propylene carbonate
RS	resolution
RSD	relative standard deviation
SEI	solid electrolyte interface
StSo	stock solution
TMB	trimethoxyborane
TMP	trimethyl phosphate
VC	vinylene carbonate

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Tables

Figure 1. Calibration points and linear regressions for (a) EMC and (b) DMC.

Figure 2. Comparison of gas chromatograms detected with GC-MS (top) and GC-FID (bottom) for Sample A.

Figure 3. Chromatogram detected by GC-FID from Sample B recovered through mechanical disclosure combined with thermal drying.

Figure 4. GC-FID chromatograms of Sample CO (origin, center), Sample CD (distillate, top), and Sample CR (residue, bottom).

Figure 5. Mass concentrations of Samples CO, CD, and CR.

Appendix A

Appendix A.1. Samples and Setup

Figure A1. Photo of each sample. (a) Sample A after sample preparation. (b) Sample B after addition of NaHCO3 (right) and after sample preparation (left). (c) Sample CO (right) and after sample preparation (left). (d) Sample CD after addition of NaHCO3 (right) after sample preparation (left). (e) Sample CR (right) and after sample preparation (left).

Figure A2. Experimental setup for the distillation of sample CO.

Appendix A.2. Additional Information for Method Validation

Figure A3. GC-FID chromatograms of standard mixture comprising all components for validating the selectivity of the method (a) without and (b) with NaHCO3 preparation.

Figure A4. GC-FID chromatograms of LP57 in (a) full scale and (b) partial scale.

Table A1

Range, slope a, intercept b, and coefficient of determination R² for each calibration curve and LOD and LOQ for each component.

Substance	Upper Limit	Lower Limit	a	b	R²	LOD	LOQ
	${\displaystyle \frac{g_i}{g_{AcN}}}$	${\displaystyle \frac{g_i}{g_{AcN}}}$	${\displaystyle \frac{\mathsf{\mu }{\mathrm{V}\mathrm{m}\mathrm{i}\mathrm{n}}_i\cdot g_{AcN}}{\mathsf{\mu }{\mathrm{V}\mathrm{m}\mathrm{i}\mathrm{n}}_{AcN}\cdot g_i}}$	${\displaystyle \frac{\mathsf{\mu }{\mathrm{V}\mathrm{m}\mathrm{i}\mathrm{n}}_i}{\mathsf{\mu }{\mathrm{V}\mathrm{m}\mathrm{i}\mathrm{n}}_{AcN}}}$	-	${\displaystyle \frac{m{g}_i}{g_{AcN}}}$	${\displaystyle \frac{m{g}_i}{g_{AcN}}}$
MeOH	0.2131	0.0031	0.7419	−1.1167 × 10−3	0.9994	0.0877	0.2893
	1.8247	0.2131	0.7519	2.2946 × 10−3	0.9998
	2.2128	1.8247	0.7735	−3.8319 × 10−2	0.9987
EtOH	0.2028	0.0031	1.0587	1.5370 × 10−5	0.9995	0.0917	0.3027
	1.8136	0.2028	1.1243	−1.5579 × 10−2	0.9999
	2.2143	1.8136	0.9866	2.4923 × 10−1	0.9968
FB	0.2486	0.0037	2.2373	−1.3387 × 10−3	0.9993	0.0948	0.3129
	2.3244	0.2486	2.0201	2.7882 × 10−2	0.9999
	2.8624	2.3244	2.1570	−2.7103 × 10−1	0.9948
DMC	0.2613	0.0047	0.3897	1.0010 × 10−4	0.9998	0.4189	1.3823
	2.4424	0.2613	0.4135	−8.8866 × 10−3	0.9999
	2.9646	2.4424	0.3965	3.4019 × 10−2	0.9974
EMC	0.2823	0.0043	0.7326	−8.5645 × 10−4	0.9998	0.1101	0.3635
	2.3473	0.2823	0.7412	−1.3123 × 10−3	0.9997
	2.8326	2.3473	0.7296	2.1402 × 10−2	0.9991
DEC	0.2731	0.0040	0.8916	−8.3882 × 10−4	0.9968	0.0460	0.1518
	2.2376	0.2731	0.8911	1.6395 × 10−2	0.9997
	2.7289	2.2376	0.8777	3.1273 × 10−2	0.9951
VC	0.2554	0.0065	0.3563	2.8799 × 10−4	1.0000	0.3263	1.0767
	1.7008	0.2554	0.3585	−2.7006 × 10−4	0.9995
FEC	0.2461	0.0178	0.2448	−4.8183 × 10−4	1.0000	0.2618	0.8639
	1.7090	0.2461	0.2203	−2.0051 × 10−3	0.9943
PC	0.2577	0.0069	0.6279	−2.9934 × 10−4	0.9995	0.1526	0.5037
	1.7050	0.2577	0.5700	−1.4452 × 10−3	0.9965
EC	0.2415	0.0032	0.3564	−2.1831 × 10−4	0.9999	0.0999	0.3296
	1.6764	0.2415	0.3148	1.8939 × 10−3	0.9976
CHB	0.2597	0.0061	2.4046	−1.1336 × 10−3	1.0000	0.2924	0.9649
	1.3052	0.2597	2.5992	−5.7071 × 10−2	0.9994
CB	2.5251	3.1126	1.6368	2.5157 × 10−1	0.9908	-	-
	3.1126	2.5251	2.0577	−8.9418 × 10−1	0.9908

Table A2

%R and RSD of all measurements for checking accuracy and precision for all components.

mi,th	mi,meas.	%R	RSD	mi,th	mi,meas.	%R	RSD
mg	mg	%	%	mg	mg	%	%
MeOH				EC
82.47	80.54	97.65	1.93	53.21	51.68	97.13	3.9
358.38	360.41	100.57	0.87	538.85	626.93	116.35	6.5
31.31	29.35	93.74	5.22	54.44	57.1	104.88	3.53
31.42	28.76	91.53	5.23	54.63	55.63	101.83	5.34
EtOH				FB
85.45	85.55	100.12	0.78	804.28	813.56	101.15	1.37
396.57	393.14	99.14	0.42	39.95	36.11	90.39	1.7
33.33	31.97	95.93	1.98	42.28	38.48	91.02	3.12
33.45	32.17	96.19	1.39	42.43	39.51	93.12	2.36
DEC				FEC
845.01	852.75	100.92	1.88	50.56	47.78	94.5	2.39
91.57	88.44	96.58	1.05	567.94	562.48	99.04	1.8
42.62	41.6	97.61	5.98	61.17	63.43	103.7	4.24
856.04	860.31	100.5	2.13	83.51	84.76	101.51	2.6
42.76	42.13	98.51	3.15	61.39	62.54	101.89	4.6
PC				CHB
52.11	55.59	106.69	3.13	51.52	53.96	104.73	2.68
629.39	639.28	101.57	2.52	41.18	42.31	102.74	6.43
53.47	55.58	103.96	3.98	53.98	54.07	100.18	6.69
78.76	87.06	110.54	3.61	211.66	210.41	99.41	3.64
53.65	54.55	101.68	5.12	41.32	42.59	103.07	4.94
DMC				VC
480.45	480.85	100.08	1.69	53.63	54.03	100.75	4.12
117.08	121.72	103.96	2.60	539.12	576.81	106.99	3.94
46.04	45.22	98.23	3.49	59.84	59.85	100.02	4.25
46.2	46.2	100.00	2.12	60.04	59.25	98.69	3.63
EMC				CB
477.21	483.47	101.31	2.92	953.32	960.49	100.75	3.59
96.35	97.29	100.97	3.11	627.96	623.17	99.24	2.82
43.49	42.18	97.00	5.03	584.84	582.65	99.63	5.02
79.51	77.69	97.71	3.04	586.87	590.3	100.58	2.76
209.08	207.23	99.11	2.43
43.64	42.93	98.39	2.73

Appendix A.3. Full-Scale Chromatograms and Overview of Appeared Peaks

Figure A5. Chromatogram of Sample A in full-scale for (a) GC-FID and (b) GC-MS.

Table A3

List of peaks and the assignment for Sample A.

Peak-ID	Substance	Match Factor	Concentration
			mg·mL−1
1 FID	Unknown		Tr
2 FID	Unknown		Tr
3 FID,p	MeOH		0.80 ± 0.09
4	Unknown	701	Tr
5 FID	Acetone		Tr
6	AcN	912	Tr
7 FID	Unknown		Tr
8	Unknown	557	Tr
9	DMC	894	375.57 ± 1.78
10	EMC	806	312.64 ± 2.59
11	Unknown	736	Tr
12	DEC	817	9.05 ± 0.10
13	Unknown	677	Tr
14	Trimethylphosphate	903	Tr
15 MS	Unknown	710	Tr
16 t,p	R-Dimethylphosphat with R: Ethyl or Butyl	757–780	Tr
17	EC	888	322.11 ± 11.07
18	Unknown	552	Tr
19	Unknown	582	Tr
20	Unknown	687	Tr
21 p	Dimethyl diglycolcarbonate	613	Tr
22 p	Diethyl-2,5-dioxahexane carboxylate	733	Tr
23	CHB	944	39.35 ± 0.86
24 MS,p	Octadecamethylcyclononasiloxane	793	Tr
25 t,p	2-Vinylnaphthalene	945	Tr
25 t,p	Biphenyl	941	Tr

^MS only detected via MS, ^FID only detected via FID, ^p putative, ^t tentatively.

Figure A6. GC-FID chromatogram of Sample B in full scale.

Table A4

List of peaks and the assignment for Sample B.

Peak-ID	Substance	Match Factor	Concentration
			mg·mL−1
1 FID	MeOH		0.73 ± 0.01
2	EtOH	807	0.22 ± 0.01
3	AcN	876	Tr
4 MS	Unknown	785	Tr
5	DMC	877	2.67 ± 0.03
6	EMC	820	964.62 ± 12.67
7 MS	MEG	813	Tr
8	DEC	857	10.75 ± 0.19
9	Unknown	507	Tr
10 MS	Octamethylcyclotetrasiloxane	811	Tr
11 MS	Unknown	596	Tr
12 p	EC	764	10.40 ± 7.77
13	Unknown		Tr

^MS only detected via MS, ^FID only detected via FID, ^p putative, ^t tentatively.

Figure A7. Full-scale GC-FID chromatograms of Sample CO (origin, (center)), Sample CD (distillate, (top)), and Sample CR (residue, (bottom)).

Table A5

List of peaks and the assignment for sample C_O, C_D, and C_R.

Peak-ID	Substance	Concentration in Sample mg·mL−1
		CO	CD	CR
1 FID	MeOH	1.23 ± 0.02	1.17 ± 0.02	0.87 ± 0.07
2	EtOH	1.03 ± 0.01	1.14 ± 0.01	0.70 ± 0.081
3 FID	Acetone		Tr	Tr
4	can	Tr	Tr	Tr
5 MS	Trimethoxyborane			Tr
6	Unknown	Tr	Tr
7	DMC	52.48 ± 0.23	52.93 ± 0.45
8 p	Benzene	Tr	Tr
9	FB	48.08 ± 0.22	47.50 ± 0.39
10	1,1-dimethoxypropane			Tr
11	1-Butanol			Tr
12	EMC	831.45 ± 6.06	866.72 ± 9.61	266.86 ± 4.78
13 p	1,4-Dioxane			Tr
14	1,1-dimethoxy-2-methylpropane			Tr
15	2-ethyl-1,3-dioxolane		Tr	Tr
16 p	MEG	Tr		Tr
17 MS	Unknown		Tr
18	DEC	21.58 ± 0.21	23.89 ± 0.27	78.51 ± 1.62
19	Unknown			Tr
20 p	2,5-Dimethyl-1,4-dioxane	Tr		Tr
21	Unknown			Tr
22	Unknown			Tr
23	Unknown	Tr	Tr	Tr
24 FID	Unknown			Tr
25	1-Methoxy-2-propyl acetate	Tr		Tr
26 FID	Unknown			Tr
27	Unknown			Tr
28 p	Fluorostyrene			Tr
29	Unknown			Tr
30	1-Butoxy-2-propanol	Tr	Tr	Tr
31 p	Ethyl butyl carbonate	Tr	Tr	Tr
32 MS	Unknown	Tr		Tr
33	tert-Butylbenzene	Tr	Tr	Tr
34	Unknown			Tr
35	Unknown			Tr
36	Unknown			Tr
37	EC	10.87 ± 0.09	0.84 ± 0.02	577.65 ± 18.11
38	Methyl octanoate	Tr		Tr
39	Unknown			Tr
40	Unknown			Tr
41	Unknown			Tr
42	Ethyl octanoate	Tr		Tr
43	Unknown			Tr
44 t	Isomer of octyl crylate	Tr		Tr
45 MS	Unknown			Tr
46	Unknown			Tr
47	Unknown			Tr
48 MS	Unknown			Tr
49	4,4-Dimethyl-2-oxazolidinone	Tr		Tr
50	diethyl-2,5-dioxahexane dicarboxylate			Tr
51	CHB	1.07 ± 0.01	0.34 ± 0.01	39.09 ± 1.08
52 MS	Butyl caprylate			Tr
53 MS	5-Methylquinoxaline			Tr
54 MS	Unknown			Tr
55 t,p	Biphenyl			Tr
55 t,p	2-Vinylnaphthalene			Tr
56 MS	2-Hydroxyethyl octanoate			Tr
57 FID	Unknown			Tr
58 FID	Unknown			Tr
59 FID	Unknown			Tr
60 FID	Unknown			Tr
61 FID	Unknown			Tr
62 FID	Unknown			Tr
63 FID	Unknown			Tr
64 FID	Unknown			Tr
65 FID	Unknown			Tr
66 FID	Unknown			Tr
67 FID	Unknown			Tr
68 FID	Unknown			Tr

^MS only detected via MS, ^FID only detected via FID, ^p putative, ^t tentatively.

References

1. Hettesheimer, T.; Neef, C.; Rosellón Inclán, I.; Link, S.; Schmaltz, T.; Schuckert, F.; Stephan, A.; Stephan, M.; Thielmann, A.; Weymann, L. et al. Lithium-Ion Battery Roadmap—Industrialization Perspectives Toward 2030; Fraunhofer ISI: Karlsruhe, Germany, 2023; [DOI: https://dx.doi.org/10.24406/publica-2153]

2. Kampker, A.; Heimes, H.; Dorn, B. Battery Monitor 2023. Available online: https://content.rolandberger.com/hubfs/07_presse/Batteriemonitor_2023_digital_final.pdf (accessed on 29 November 2024).

3. Schmaltz, T. Recycling of Lithium-Ion Batteries Will Increase Strongly in Europe. Available online: https://www.isi.fraunhofer.de/en/blog/themen/batterie-update/recycling-lithium-ionen-batterien-europa-starke-zunahme-2030-2040.html (accessed on 9 August 2024).

4. European Union. Regulation (EU) 2023/1542 of the European Parliament and of the Council of 12 July 2023 concerning batteries and waste batteries, amending Directive 2008/98/EC and Regulation (EU) 2019/1020 and repealing Directive 2006/66/EC. Off. J. Eur. Union; 2023; Available online: https://eur-lex.europa.eu/eli/reg/2023/1542/oj/eng (accessed on 24 August 2024).

5. Arnberger, A.; Coskun, E.; Rutrecht, B. Recycling von Lithium-Ionen-Batterien. Recycling und Rohstoffe; Thomé-Kozmiensky Verlag GmbH: Neuruppin, Germany, 2018; Volume 11, pp. 583-599.

6. Xu, K. Nonaqueous Liquid Electrolytes for Lithium-Based Rechargeable Batteries. ChemInform; 2004; 104, pp. 4303-4418. [DOI: https://dx.doi.org/10.1002/chin.200450271]

7. Ehrlich, G.M. LITHIUM-ION BATTERIES. Handbook of Batteries; McGraw-Hill handbooks; McGraw-Hill: New York, NY, USA, 2002; ISBN 978-0-07-135978-8

8. Hartnig, C.; Schmidt, M. Elektrolyte und Leitsalze. Handbuch Lithium-Ionen-Batterien; Springer: Berlin/Heidelberg, Germany, 2013; pp. 61-77. ISBN 978-3-642-30652-5

9. Julien, C.; Mauger, A.; Vijh, A.; Zaghib, K. Lithium Batteries: Science and Technology; Springer International Publishing: Cham, Switzerland, 2016; ISBN 978-3-319-19107-2

10. Yew, K.-H.; Song,; Jung, C.-S.; Lee, Y.-B. Elektrolyt für Eine Lithiumbatterie und Lithiumbatterie Denselben Enthaltend. DE602006000279T2, 27 November 2008.

11. Pražanová, A.; Kočí, J.; Míka, M.H.; Pilnaj, D.; Plachý, Z.; Knap, V. Pre-Recycling Material Analysis of NMC Lithium-Ion Battery Cells from Electric Vehicles. Crystals; 2023; 13, 214. [DOI: https://dx.doi.org/10.3390/cryst13020214]

12. Jiang, Z.; Zeng, Z.; Liang, X.; Yang, L.; Hu, W.; Zhang, C.; Han, Z.; Feng, J.; Xie, J. Fluorobenzene, A Low-Density, Economical, and Bifunctional Hydrocarbon Cosolvent for Practical Lithium Metal Batteries. Adv. Funct. Mater.; 2021; 31, 2005991. [DOI: https://dx.doi.org/10.1002/adfm.202005991]

13. Jiang, Z.; Zeng, Z.; Zhai, B.; Li, X.; Hu, W.; Zhang, H.; Cheng, S.; Xie, J. Fluorobenzene-Based Diluted Highly Concentrated Carbonate Electrolyte for Practical High-Voltage Lithium Metal Batteries. J. Power Sources; 2021; 506, 230086. [DOI: https://dx.doi.org/10.1016/j.jpowsour.2021.230086]

14. Markevich, E.; Salitra, G.; Aurbach, D. Fluoroethylene Carbonate as an Important Component for the Formation of an Effective Solid Electrolyte Interphase on Anodes and Cathodes for Advanced Li-Ion Batteries. ACS Energy Lett.; 2017; 2, pp. 1337-1345. [DOI: https://dx.doi.org/10.1021/acsenergylett.7b00163]

15. Wang, D.Y.; Sinha, N.N.; Burns, J.C.; Aiken, C.P.; Petibon, R.; Dahn, J.R. A Comparative Study of Vinylene Carbonate and Fluoroethylene Carbonate Additives for LiCoO2/Graphite Pouch Cells. J. Electrochem. Soc.; 2014; 161, pp. A467-A472. [DOI: https://dx.doi.org/10.1149/2.001404jes]

16. Tian, X.; Yi, Y.; Fang, B.; Yang, P.; Wang, T.; Liu, P.; Qu, L.; Li, M.; Zhang, S. Design Strategies of Safe Electrolytes for Preventing Thermal Runaway in Lithium Ion Batteries. Chem. Mater.; 2020; 32, pp. 9821-9848. [DOI: https://dx.doi.org/10.1021/acs.chemmater.0c02428]

17. Hanisch, C.; Westphal, B.; Haselrieder, W.; Schoenitz, M. Method for the Treatment of Used Batteries, in Particular Rechargeable Batteries, and Battery Processing Installation. US11050097B2, 29 June 2021.

18. Stehmann, F.; Bradtmöller, C.; Scholl, S. Separation of the Electrolyte—Thermal Drying. Recycling of Lithium-Ion Batteries; Sustainable Production, Life Cycle Engineering and Management Springer International Publishing: Cham, Switzerland, 2018; pp. 139-153. ISBN 978-3-319-70571-2

19. Grützke, M.; Mönnighoff, X.; Horsthemke, F.; Kraft, V.; Winter, M.; Nowak, S. Extraction of Lithium-Ion Battery Electrolytes with Liquid and Supercritical Carbon Dioxide and Additional Solvents. RSC Adv.; 2015; 5, pp. 43209-43217. [DOI: https://dx.doi.org/10.1039/C5RA04451K]

20. Nowak, S.; Winter, M. The Role of Sub- and Supercritical CO₂ as “Processing Solvent” for the Recycling and Sample Preparation of Lithium Ion Battery Electrolytes. Molecules; 2017; 22, 403. [DOI: https://dx.doi.org/10.3390/molecules22030403]

21. Zachmann, N.; Fox, R.V.; Petranikova, M.; Ebin, B. Implementation of a Sub-and Supercritical Carbon Dioxide Process for the Selective Recycling of the Electrolyte from Spent Li-Ion Battery. J. CO2 Util.; 2024; 81, 102703. [DOI: https://dx.doi.org/10.1016/j.jcou.2024.102703]

22. Gachot, G.; Ribière, P.; Mathiron, D.; Grugeon, S.; Armand, M.; Leriche, J.-B.; Pilard, S.; Laruelle, S. Gas Chromatography/Mass Spectrometry As a Suitable Tool for the Li-Ion Battery Electrolyte Degradation Mechanisms Study. Anal. Chem.; 2011; 83, pp. 478-485. [DOI: https://dx.doi.org/10.1021/ac101948u]

23. Gachot, G.; Grugeon, S.; Armand, M.; Pilard, S.; Guenot, P.; Tarascon, J.-M.; Laruelle, S. Deciphering the Multi-Step Degradation Mechanisms of Carbonate-Based Electrolyte in Li Batteries. J. Power Sources; 2008; 178, pp. 409-421. [DOI: https://dx.doi.org/10.1016/j.jpowsour.2007.11.110]

24. Hofmann, A.; Müller, F.; Schöner, S.; Jeschull, F. Revealing the Formation of Dialkyl Dioxahexane Dioate Products from Ethylene Carbonate Based Electrolytes on Lithium and Potassium Surfaces. Batter. Supercaps; 2023; 6, e202300325. [DOI: https://dx.doi.org/10.1002/batt.202300325]

25. Yoshida, H.; Fukunaga, T.; Hazama, T.; Terasaki, M.; Mizutani, M.; Yamachi, M. Degradation Mechanism of Alkyl Carbonate Solvents Used in Lithium-Ion Cells during Initial Charging. J. Power Sources; 1997; 68, pp. 311-315. [DOI: https://dx.doi.org/10.1016/S0378-7753(97)02635-9]

26. Laruelle, S.; Pilard, S.; Guenot, P.; Grugeon, S.; Tarascon, J.-M. Identification of Li-Based Electrolyte Degradation Products Through DEI and ESI High-Resolution Mass Spectrometry. J. Electrochem. Soc.; 2004; 151, pp. A1202-A1209. [DOI: https://dx.doi.org/10.1149/1.1760992]

27. Schultz, C.; Vedder, S.; Streipert, B.; Winter, M.; Nowak, S. Quantitative Investigation of the Decomposition of Organic Lithium Ion Battery Electrolytes with LC-MS/MS. RSC Adv.; 2017; 7, pp. 27853-27862. [DOI: https://dx.doi.org/10.1039/C7RA03839A]

28. Schultz, C.; Kraft, V.; Pyschik, M.; Weber, S.; Schappacher, F.; Winter, M.; Nowak, S. Separation and Quantification of Organic Electrolyte Components in Lithium-Ion Batteries via a Developed HPLC Method. J. Electrochem. Soc.; 2015; 162, pp. A629-A634. [DOI: https://dx.doi.org/10.1149/2.0401504jes]

29. Takeda, S.; Morimura, W.; Liu, Y.; Sakai, T.; Saito, Y. Identification and Formation Mechanism of Individual Degradation Products in Lithium-ion Batteries Studied by Liquid Chromatography/Electrospray Ionization Mass Spectrometry and Atmospheric Solid Analysis Probe Mass Spectrometry. Rapid Commun. Mass Spectrom.; 2016; 30, pp. 1754-1762. [DOI: https://dx.doi.org/10.1002/rcm.7652]

30. Terborg, L.; Weber, S.; Passerini, S.; Winter, M.; Karst, U.; Nowak, S. Development of Gas Chromatographic Methods for the Analyses of Organic Carbonate-Based Electrolytes. J. Power Sources; 2014; 245, pp. 836-840. [DOI: https://dx.doi.org/10.1016/j.jpowsour.2013.07.030]

31. Campion, C.L.; Li, W.; Euler, W.B.; Lucht, B.L.; Ravdel, B.; DiCarlo, J.F.; Gitzendanner, R.; Abraham, K.M. Suppression of Toxic Compounds Produced in the Decomposition of Lithium-Ion Battery Electrolytes. Electrochem. Solid-State Lett.; 2004; 7, A194. [DOI: https://dx.doi.org/10.1149/1.1738551]

32. Grützke, M.; Weber, W.; Winter, M.; Nowak, S. Structure Determination of Organic Aging Products in Lithium-Ion Battery Electrolytes with Gas Chromatography Chemical Ionization Mass Spectrometry (GC-CI-MS). RSC Adv.; 2016; 6, pp. 57253-57260. [DOI: https://dx.doi.org/10.1039/C6RA09323J]

33. Kraft, V.; Weber, W.; Grützke, M.; Winter, M.; Nowak, S. Study of Decomposition Products by Gas Chromatography-Mass Spectrometry and Ion Chromatography-Electrospray Ionization-Mass Spectrometry in Thermally Decomposed Lithium Hexafluorophosphate-Based Lithium Ion Battery Electrolytes. RSC Adv.; 2015; 5, pp. 80150-80157. [DOI: https://dx.doi.org/10.1039/C5RA16679A]

34. Weber, W.; Kraft, V.; Grützke, M.; Wagner, R.; Winter, M.; Nowak, S. Identification of Alkylated Phosphates by Gas Chromatography–Mass Spectrometric Investigations with Different Ionization Principles of a Thermally Aged Commercial Lithium Ion Battery Electrolyte. J. Chromatogr. A; 2015; 1394, pp. 128-136. [DOI: https://dx.doi.org/10.1016/j.chroma.2015.03.048] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25846260]

35. Keller, T.; Holtbruegge, J.; Niesbach, A.; Górak, A. Transesterification of Dimethyl Carbonate with Ethanol To Form Ethyl Methyl Carbonate and Diethyl Carbonate: A Comprehensive Study on Chemical Equilibrium and Reaction Kinetics. Ind. Eng. Chem. Res.; 2011; 50, pp. 11073-11086. [DOI: https://dx.doi.org/10.1021/ie2014982]

36. Metzger, M.; Strehle, B.; Solchenbach, S.; Gasteiger, H.A. Hydrolysis of Ethylene Carbonate with Water and Hydroxide under Battery Operating Conditions. J. Electrochem. Soc.; 2016; 163, pp. A1219-A1225. [DOI: https://dx.doi.org/10.1149/2.0411607jes]

37. Liu, Y.; Mu, D.; Li, R.; Ma, Q.; Zheng, R.; Dai, C. Purification and Characterization of Reclaimed Electrolytes from Spent Lithium-Ion Batteries. J. Phys. Chem. C; 2017; 121, pp. 4181-4187. [DOI: https://dx.doi.org/10.1021/acs.jpcc.6b12970]

38. Xu, R.; Lei, S.; Wang, T.; Yi, C.; Sun, W.; Yang, Y. Lithium Recovery and Solvent Reuse from Electrolyte of Spent Lithium-Ion Battery. Waste Manag.; 2023; 167, pp. 135-140. [DOI: https://dx.doi.org/10.1016/j.wasman.2023.05.034]

39. Q2(R2) Validation of Analytical Procedures; Center for Drug Evaluation and Research, Office of Regulatory Policy: Silver Spring, MD, USA, 2024.

40. Analytical Procedures and Methods Validation for Drugs and Biologics; FDA: Silver Spring, MD, USA, 2015.

41. Kromidas, S. Handbuch Validierung in der Analytik; 1st ed. Wiley: Hoboken, NJ, USA, 2000; ISBN 978-3-527-29811-2

42. <621> Chromatography. Available online: https://www.usp.org/harmonization-standards/pdg/excipients/chromatography (accessed on 29 August 2024).

43. Wenzl, T.; Haedrich, J.; Schaechtele, A.; Robouch, P.; Stroka, J. Guidance Document on the Estimation of LOD and LOQ for Measurements in the Field of Contaminants in Feed and Food; Publications Office: Luxembourg, 2016; ISBN 978-92-79-61768-3

44. Matadi, B.P.; Geniès, S.; Delaille, A.; Waldmann, T.; Kasper, M.; Wohlfahrt-Mehrens, M.; Aguesse, F.; Bekaert, E.; Jiménez-Gordon, I.; Daniel, L. et al. Effects of Biphenyl Polymerization on Lithium Deposition in Commercial Graphite/NMC Lithium-Ion Pouch-Cells during Calendar Aging at High Temperature. J. Electrochem. Soc.; 2017; 164, pp. A1089-A1097. [DOI: https://dx.doi.org/10.1149/2.0631706jes]

45. Kim, H.; Grugeon, S.; Gachot, G.; Armand, M.; Sannier, L.; Laruelle, S. Ethylene Bis-Carbonates as Telltales of SEI and Electrolyte Health, Role of Carbonate Type and New Additives. Electrochim. Acta; 2014; 136, pp. 157-165. [DOI: https://dx.doi.org/10.1016/j.electacta.2014.05.072]

46. Heider, U.; Oesten, R.; Jungnitz, M. Challenge in Manufacturing Electrolyte Solutions for Lithium and Lithium Ion Batteries Quality Control and Minimizing Contamination Level. J. Power Sources; 1999; 81–82, pp. 119-122. [DOI: https://dx.doi.org/10.1016/S0378-7753(99)00142-1]

47. Cheng, K.; Feng, W.; Wang, L.; Chen, Y. Phase Equilibria for Quaternary Systems of Water + Methanol or Ethanol + Ethyl Methyl Carbonate + Octane at T = 298.15 K. J Solut. Chem; 2017; 46, pp. 1337-1348. [DOI: https://dx.doi.org/10.1007/s10953-017-0641-8]

48. Wang, X.; Yasukawa, E.; Kasuya, S. Nonflammable Trimethyl Phosphate Solvent-Containing Electrolytes for Lithium-Ion Batteries: I. Fundamental Properties. J. Electrochem. Soc.; 2001; 148, pp. A1058-A1065. [DOI: https://dx.doi.org/10.1149/1.1397773]

49. Xu, F.; Cai, X.; Zhang, J.; Zhang, L.; Wang, J.; Zhang, N.; Li, S. Improving Compatibility between Trimethyl Phosphate and Graphite Anodes by Preconstructing a Stable Solid Electrolyte Interphase Film. ACS Appl. Energy Mater.; 2022; 5, pp. 11370-11378. [DOI: https://dx.doi.org/10.1021/acsaem.2c01862]

50. Lemordant, D.; Zhang, W.; Ghamouss, F.; Farhat, D.; Darwiche, A.; Monconduit, L.; Dedryvère, R.; Martinez, H.; Cadra, S.; Lestriez, B. Artificial SEI for Lithium-Ion Battery Anodes. Advanced Fluoride-Based Materials for Energy Conversion; Elsevier: Amsterdam, The Netherlands, 2015; pp. 173-202. ISBN 978-0-12-800679-5

51. Sasaki, T.; Abe, T.; Iriyama, Y.; Inaba, M.; Ogumi, Z. Formation Mechanism of Alkyl Dicarbonates in Li-Ion Cells. J. Power Sources; 2005; 150, pp. 208-215. [DOI: https://dx.doi.org/10.1016/j.jpowsour.2005.02.021]

52. Lee, H.; Lee, J.H.; Ahn, S.; Kim, H.-J.; Cho, J.-J. Co-Use of Cyclohexyl Benzene and Biphenyl for Overcharge Protection of Lithium-Ion Batteries. Electrochem. Solid-State Lett.; 2006; 9, pp. A307-A310. [DOI: https://dx.doi.org/10.1149/1.2193072]

53. Shim, E.-G.; Nam, T.-H.; Kim, J.-G.; Kim, H.-S.; Moon, S.-I. Effect of Vinyl Acetate plus Vinylene Carbonate and Vinyl Ethylene Carbonate plus Biphenyl as Electrolyte Additives on the Electrochemical Performance of Li-Ion Batteries. Electrochim. Acta; 2007; 53, pp. 650-656. [DOI: https://dx.doi.org/10.1016/j.electacta.2007.07.026]

54. Tobishima, S.; Ogino, Y.; Watanabe, Y. Influence of Electrolyte Additives on Safety and Cycle Life of Rechargeable Lithium Cells. J. Appl. Electrochem.; 2003; 33, pp. 143-150. [DOI: https://dx.doi.org/10.1023/A:1024039302020]

55. Xiao, L.; Ai, X.; Cao, Y.; Yang, H. Electrochemical Behavior of Biphenyl as Polymerizable Additive for Overcharge Protection of Lithium Ion Batteries. Electrochim. Acta; 2004; 49, pp. 4189-4196. [DOI: https://dx.doi.org/10.1016/j.electacta.2004.04.013]

56. Yang, Y.-L.; Ramaswamy, S.G.; Jakoby, W.B. Enzymatic Hydrolysis of Organic Cyclic Carbonates. J. Biol. Chem.; 1998; 273, pp. 7814-7817. [DOI: https://dx.doi.org/10.1074/jbc.273.14.7814]

57. Bhanage, B.M.; Fujita, S.; Ikushima, Y.; Arai, M. Transesterification of Urea and Ethylene Glycol to Ethylene Carbonate as an Important Step for Urea Based Dimethyl Carbonate Synthesis. Green Chem.; 2003; 5, pp. 429-432. [DOI: https://dx.doi.org/10.1039/b304182d]

58. Fang, Y.-J.; Xiao, W.-D. Experimental and Modeling Studies on a Homogeneous Reactive Distillation System for Dimethyl Carbonate Synthesis by Transesterification. Sep. Purif. Technol.; 2004; 34, pp. 255-263. [DOI: https://dx.doi.org/10.1016/S1383-5866(03)00198-9]

59. Wang, H.; Chen, S.; Li, Y.; Liu, Y.; Jing, Q.; Liu, X.; Liu, Z.; Zhang, X. Organosilicon-Based Functional Electrolytes for High-Performance Lithium Batteries. Adv. Energy Mater.; 2021; 11, 2101057. [DOI: https://dx.doi.org/10.1002/aenm.202101057]

60. Wang, H.; Sun, D.; Li, X.; Ge, W.; Deng, B.; Qu, M.; Peng, G. Alternative Multifunctional Cyclic Organosilicon as an Efficient Electrolyte Additive for High Performance Lithium-Ion Batteries. Electrochim. Acta; 2017; 254, pp. 112-122. [DOI: https://dx.doi.org/10.1016/j.electacta.2017.09.111]

61. Ding, J.; Wen, Y.; Lan, X.; Hu, R. Roles of Trimethyl Borate in Constructing an Interphase on Li Anode: Angel or Demon?. ACS Appl. Mater. Interfaces; 2023; 15, pp. 6768-6776. [DOI: https://dx.doi.org/10.1021/acsami.2c19417]

62. Liu, Q.; Yang, G.; Liu, S.; Han, M.; Wang, Z.; Chen, L. Trimethyl Borate as Film-Forming Electrolyte Additive To Improve High-Voltage Performances. ACS Appl. Mater. Interfaces; 2019; 11, pp. 17435-17443. [DOI: https://dx.doi.org/10.1021/acsami.9b03417] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31021075]

63. ABE, K.; Matsumori, Y.; Ueki, A. Nonaqueouselectrolytic Solution and Lithium Secondary Batteries. US 2008/0050658 A1, 28 February 2008.

64. Chen, S.; Wang, Z.; Wang, J.; Tong, X.; Yan, W. Lower Explosion Limit of the Vented Gases from Li-Ion Batteries Thermal Runaway in High Temperature Condition. J. Loss Prev. Process Ind.; 2020; 63, 103992. [DOI: https://dx.doi.org/10.1016/j.jlp.2019.103992]

65. Chen, S.; Wang, Z.; Yan, W. Identification and Characteristic Analysis of Powder Ejected from a Lithium Ion Battery during Thermal Runaway at Elevated Temperatures. J. Hazard. Mater; 2020; 400, 123169. [DOI: https://dx.doi.org/10.1016/j.jhazmat.2020.123169]