Detoxification and Targeted Conversion of Waste Lithium Battery Electrolyte to Light Hydrocarbons via In Situ Catalytic Pyrolysis: Roles of Li, Ni, Co, and Mn Elements

Abstract

Spent lithium-ion battery electrolytes contain fluorine-, sulfur-, and phosphorus-bearing toxins, necessitating deep detoxification and directional conversion into C₁–C₆ light hydrocarbons. To elucidate the specific catalytic roles and sequential activation of cathode metals (Li, Ni, Co, Mn), this work systematically deconvolutes their mono- and multi-metallic migration mechanisms over a CaO-ZSM-5* catalyst during vacuum catalytic pyrolysis (530 °C, 100 Pa). Results reveal that Li⁺ and Ni²⁺ dominate C–O bond cleavage in carbonates and CaO-ZSM-5*-assisted decarboxylation and oxygen fixation, significantly increasing the relative hydrocarbon content. Conversely, Co^2/3+ and Mn⁴⁺ release reactive oxygen species, causing deep oxidation of hydrocarbons into CO₂ and antagonizing the targeted conversion. In multi-metallic systems, forming composite metal oxides (M_xN_yO_z) increases the energy barrier for releasing active catalytic ions, hindering carbonate cleavage and leaving unreacted carbonate feedstocks. For detoxification, F and P are effectively immobilized as CaF₂ and Ca₂P₂O₇. The relative content of detected gas-phase nitriles is minimized to <2% due to the strong antagonistic effect of Ni²⁺ on Li⁺-promoted hexanedinitrile cleavage, while sulfur species derived from 1,3-propane sultone are converted to SO₂ and ultimately mineralized as calcium and metal-sulfur salts. Mechanistically, product distributions and crystallographic properties suggest a hypothesized sequential activation model—Li⁺ → Ni²⁺ → Mn⁴⁺—governing reactivity, whereas Co^2/3+ does not participate in the synergistic detoxification and selective upgrading process. This migration–reaction coupling framework provides critical insights for cathode-assisted in situ catalytic pyrolysis and closed-loop electrolyte recycling.

1. Introduction

The rapid expansion of new energy technologies, driven by global commitments to carbon neutrality and energy security, has led to an unprecedented surge in the production and deployment of lithium-ion batteries (LIBs) [1]. As the core component of electric vehicles and energy storage systems, global LIB production is projected to exceed 1 TWh by 2025, while their limited-service life (8–10 years) will result in a massive wave of retired batteries, reaching millions of tons annually by 2030 [2,3]. This transition is shifting the focus of LIB recycling from simple metal recovery toward the holistic management of complex, multi-component waste streams.

Among these components, the electrolyte represents a particularly challenging yet underexplored fraction. Although accounting for only ~10–15 wt.% of LIBs, spent electrolytes constitute the primary source of organic and heteroatom-containing pollutants during recycling [4]. Typical electrolyte systems consist of carbonate solvents, including ethylene carbonate (EC), diethyl carbonate (DEC), and ethyl methyl carbonate (EMC), lithium salts such as lithium hexafluorophosphate (LiPF₆), and functional additives such as fluoroethylene carbonate (FEC), 1,3-propane sultone (PS), and hexanedinitrile (ADN), which collectively exhibit high volatility, chemical instability, and strong toxicity [4,5]. Upon exposure to air or thermal treatment, LiPF₆ readily decomposes into HF and organophosphorus species, while sulfur- and nitrogen-containing additives generate SO₂ and NO_x during conventional incineration, posing severe environmental and health risks [6,7]. These characteristics render electrolyte detoxification both technically difficult and environmentally critical.

Current treatment strategies, including solvent extraction, incineration, and hydrometallurgical oxidation [6,8,9,10], are largely limited by incomplete detoxification, secondary pollution, or poor resource valorization. In particular, these approaches typically treat electrolytes as “waste to be removed”, rather than as reactive feedstocks with potential for value-added conversion. This paradigm neglects a key opportunity: the intrinsic chemical reactivity and catalytic potential embedded within other battery components.

Notably, spent cathode materials—especially layered ternary oxides, such as Li(Ni_xCo_yMn_z)O₂ cathode materials (NCM)—are not inert residues but chemically active materials under thermal and reductive conditions [11,12,13]. Transition metals such as Ni, Co, and Mn have been widely reported to catalyze cracking, dehydrogenation, and rearrangement reactions of organic molecules, while Li species can participate in nucleophilic and organometallic reaction pathways [14,15,16]. These properties suggest that cathode materials can act as in situ catalytic agents, enabling a “waste-to-waste synergy” in which toxic electrolyte components are simultaneously decomposed and upgraded. However, despite emerging evidence of catalytic activity, the element-specific roles [17,18], synergistic/antagonistic interactions, and migration-controlled reaction pathways of Li, Ni, Co, and Mn remain poorly understood, particularly under vacuum catalytic conditions. Moreover, the interaction between cathode-derived metal species and acid–base bifunctional catalysts (e.g., CaO-modified ZSM-5) [19,20,21] has not been systematically elucidated. This knowledge gap limits the rational design of integrated processes for simultaneous detoxification and fuel production.

This study proposes a cathode–zeolite synergistic catalytic framework for the in situ directional conversion of waste LIB electrolytes into light hydrocarbons. By coupling a CaO-modified ZSM-5 catalyst (CaO-ZSM-5*) with model metal species (LiOH, NiO, Co₃O₄, MnO₂) and representative NCM cathode materials under vacuum catalytic pyrolysis conditions, we aim to: (i) elucidate the element-specific catalytic roles of cathode metals (Li, Ni, Co, and Mn) in promoting C–O bond cleavage in carbonates, decarboxylation and oxygen fixation, and the directional conversion into light hydrocarbons; (ii) uncover the synergistic and antagonistic effects of multi-metal interactions (e.g., the formation of composite metal oxides, M_xN_yO_z) on the energy barrier for active ion release; and (iii) clarify the mechanisms of deep detoxification, specifically the mineralization of F, P, and S, and the suppression of gas-phase nitriles.

We demonstrate that Li⁺ and Ni²⁺ dominate the synergistic detoxification and selective upgrading process by promoting C–O bond cleavage and directional conversion. Conversely, Co^2/3+ primarily contributes via lattice oxygen release that causes deep oxidation of hydrocarbons, and does not participate in the synergistic detoxification and selective upgrading process. A cathode-dictated sequential activation (Li⁺ → Ni²⁺ → Mn⁴⁺) governs the reaction pathways, with Mn⁴⁺ engaging in the later stage for complete sulfur mineralization, revealing a structure–reactivity relationship intrinsic to cathode materials. Importantly, fluorine, phosphorus, and sulfur species are effectively immobilized as stable inorganic salts (e.g., CaF₂, Ca₂P₂O₇, as well as calcium and metal-sulfur salts). Overall, this work establishes a migration–reaction coupling framework and a cathode-assisted self-catalytic paradigm, providing a new strategy for reconciling deep detoxification with directional conversion into light hydrocarbons, thereby enabling closed-loop recycling of lithium-ion batteries.

While our previous work demonstrated the macroscopic feasibility of utilizing spent NCM cathodes for electrolyte defluorination and directional upgrading into light fuels [22], the individual catalytic contributions, multi-metallic interactions, and activation pathways of the constituent metals (Li, Ni, Co, Mn) within the spent cathode materials were not separately deconvoluted. This study systematically investigates these element-specific migration–reaction mechanisms by employing model metal compounds across mono- and multi-metallic configurations over a CaO-ZSM-5* catalyst. The investigation clearly resolves the divergent roles of the metals, identifying that while Li⁺ and Ni²⁺ dominate the targeted hydrocarbon conversion, Co^2/3+ and Mn⁴⁺ release reactive oxygen species causing deep oxidation. Furthermore, we uncover the suppressive effects of composite metal oxides (M_xN_yO_z) on active ion release, the Ni-driven suppression of toxic nitriles, and a cathode-dictated sequential activation order (Li⁺ → Ni²⁺ → Mn⁴⁺). These findings establish a comprehensive migration–reaction coupling framework, providing distinct complementary insights for closed-loop battery recycling. To explicitly delineate these distinctions, a systematic comparison between our prior work [22] and the present study is provided in

Table 1. Comparison of experimental parameters and key outputs between our prior work [22] and the current study.

Parameter	Prior Work	This Manuscript
Feedstock (Waste Stream)	Real spent NCM cathode + PVDF binder + electrolyte	Simulated spent electrolyte
Catalyst Components	CaO + ZSM-5 molecular sieve	Metal Additives (MA: LiOH, NiO, Co3O4, MnO2) + Ca(OH)2 + ZSM-5 molecular sieve
Catalyst Preparation	Energy-intensive multi-step process: (1) Wet mixing in deionized water at 80 °C; (2) Vacuum drying at 110 °C for 12 h; (3) Pre-calcination at 500 °C for 4 h under vacuum.	Simplified two-step physical mixing: (1) Dry mixing ZSM-5 and Ca(OH)2 (2:1 wt.); (2) Blending MA with the base catalyst (1:3 wt.).
Mass Ratios	Electrolyte: CaO-ZSM-5: Cathode = 2:2:1	Metal Additive: Electrolyte: CaO-ZSM-5* = 1:2:3
Process Conditions	Temperature: 530 °C Initial pressure: 100 Pa Pyrolysis time: 20 min	Temperature: 530 °C Initial pressure: 100 Pa Pyrolysis time: 20 min
Research Objective	Macro-level feasibility of electrolyte defluorination and directional upgrading into light fuels	Element-specific mechanistic deconvolution of individual cathode metals and their multi-metallic interactions
Key Outputs/Conclusions	(1) Proof of concept for Li-promoted carbonate cracking (2) Macroscopic defluorination via CaF2 formation	(1) Divergent roles: Li/Ni promote hydrocarbons; Co/Mn cause deep oxidation via reactive oxygen species. (2) Phase interactions: MxNyOz formation hinders active ion release. (3) Nitrile distribution: Ni strongly antagonizes ADN cleavage. (4) Activation sequence: Cathode-dictated sequential activation (Li+ → Ni2+ → Mn4+).

.

2. Materials and Methods

2.1. Chemicals and Materials

Spent LIBs containing LiNi_1/3Co_1/3Mn_1/3O₂ (NCM) cathode materials were sourced from Shanghai Electric Group Corporation (Shanghai, China). To ensure operational safety, the spent LIBs were discharged in a saturated NaCl solution for 24 h, followed by manual dismantling to retrieve the electrode sheets. Since the direct recovery of spent electrolytes is constrained by high volatility and limited mass fraction, a matrix-assisted diffusion method [22] was adopted. Immediately after disassembly, the recovered electrode sheets were immersed in a commercial ternary electrolyte (KLD-1230C; Canrd, Dongguan, China) to extract the residual components, thereby yielding the simulated spent electrolyte (which intrinsically contains trace stabilizing additives, such as aliphatic amines serving as HF scavengers, standard in commercial formulations).

Gas chromatography-mass spectrometry (GC-MS) analysis of this solution, as characterized in our previous study [22] (Table S1), identified the solvent matrix as ethyl methyl carbonate (EMC, 41.47%), diethyl carbonate (DEC, 22.70%), and ethylene carbonate (EC, 21.04%), alongside 1.0 M LiPF₆ (2.29%). Toxic additives, including hexanedinitrile (ADN, 7.37%), fluoroethylene carbonate (FEC, 1.86%), 1,3-propane sultone (PS, 1.75%), and vinylene carbonate (VC, 1.52%), were also quantified, ensuring a rigorous replication of the hazardous waste stream. To systematically decouple the catalytic roles of individual cathode metals, lithium hydroxide (LiOH, 99.95% metals basis; Macklin, Shanghai, China), nickel(II) oxide (NiO, 99.9%; Aladdin, Shanghai, China), cobalt(II,III) oxide (Co₃O₄, 99.5%; Aladdin, Shanghai, China), and manganese(IV) oxide (MnO₂, 99.5%; Aladdin, Shanghai, China) were employed as model compounds to simulate the lithium species and transition metal oxides in the cathode. ZSM-5 molecular sieves (Si/Al ratio = 38; Aladdin, Shanghai, China) and calcium hydroxide (Ca(OH)₂, 95.0%; Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) were utilized as catalyst precursors. Carbon tetrachloride (CCl₄, 98%; Rhawn, Shanghai, China), high-purity nitrogen (N₂, 99.999%), and helium (He, 99.999%; Tianyi Gas, Shanghai, China) were also used.

2.2. Experimental Design

Preparation of composite catalysts. A mechanical mixing strategy was adopted to prepare a series of composite catalysts, aiming to elucidate the catalytic mechanism and decouple the specific roles of the metal species inherent in spent NCM cathodes (Li, Ni, Co, and Mn). The base catalyst, designated as CaO-ZSM-5*, was prepared by physically mixing ZSM-5 molecular sieves with Ca(OH)₂ in a mortar at a mass ratio of 2:1. Ca(OH)₂ was selected to substitute CaO in this study, considering that it not only constitutes the predominant calcium species on the ZSM-5 support but also undergoes in situ thermal conversion to CaO during the pyrolysis process. High-purity metal oxides and hydroxides (LiOH, NiO, Co₃O₄, MnO₂) were selected as model compounds to simulate the characteristic metal species of NCM cathodes, constituting the metal additives (denoted as MA). LiOH was specifically selected because it mimics the residual alkaline layer commonly found on spent cathodes; additionally, any trace water released from its decomposition is kinetically buffered by the massive in situ dehydration of the Ca(OH)₂ base matrix. These components were introduced, either individually or in specific combinations, into the catalyst system. For multi-metallic systems (i.e., those containing two or more metal species), the constituent model compounds were homogenized in equal mass fractions prior to integration. The final composite catalysts, collectively denoted as MA-CaO-ZSM-5*, were formulated by blending MA with CaO-ZSM-5* at a fixed mass ratio of 1:3.

To systematically identify potential synergistic or antagonistic interactions, the experimental framework was stratified into five distinct categories: a control group (CaO-ZSM-5* only), and mono-, bi-, tri-, and tetrametallic systems. The control group was designated as E-0, while the metal-catalyzed groups were designated by the prefix ‘E-’ followed by the initials of the incorporated metal elements (e.g., E-L, E-LN, up to E-LNCM), with comprehensive specifications detailed in Table S2. Additionally, experimental data for the spent NCM cathode-catalyzed system (designated as Group E) were referenced from our previous work [22] for comparative analysis.

Procedure of synergistic catalytic pyrolysis. The schematic flow of the catalytic pyrolysis process is illustrated in Figure S1. Experiments were conducted in a vacuum tube furnace equipped with a digital pressure gauge. To prevent the premature volatilization of the simulated spent electrolyte, a layered loading strategy was employed. In this arrangement, the electrolyte was placed at the bottom of a quartz boat and immediately covered with a uniform layer of the MA-CaO-ZSM-5* catalyst. The loaded boat was then rapidly transferred into the furnace to minimize exposure. Considering the complexity of the vacuum catalytic pyrolysis process, each experimental group was conducted as a rigorously controlled single experiment. In all MA-catalyzed experiments, the mass ratio of MA, the electrolyte, and CaO-ZSM-5* was fixed at 1:2:3, as detailed in Table S2, and the total mass of the loaded materials in the quartz boat was approximately 6.0 g. For the blank control group (E-0), no metal additive was introduced, and the electrolyte/CaO-ZSM-5* loading followed the corresponding 0:2:3 ratio in Table S2. This standardized loading strategy was implemented to ensure comparable filling levels, heat transfer, and reaction conditions across all experiments. First, the reactor was evacuated to an initial pressure of 100 Pa and then sealed. The temperature was then increased from ambient to 530 °C at a heating rate of 10 °C/min and maintained for 20 min. During this closed pyrolysis process, the reactor operated as a sealed static-vacuum system rather than a continuous-flow or constant-pressure system. Given the relatively small electrolyte loading compared with the large free volume of the quartz tube, the generated pyrolysis gas was not quantified using a calibrated gas-volume measurement device. Subsequently, high-purity nitrogen was introduced to purge the pyrolysis gases into a gas collection bag. The system was immediately re-evacuated to vacuum to expel the remaining carrier gas and prevent secondary reactions. Finally, after the system had cooled to room temperature, the pyrolysis residues were collected, and the pyrolysis oil condensed on the inner wall of the quartz tube was thoroughly recovered using CCl₄ for subsequent characterization. CCl₄ was specifically selected because its lack of C–H and C–O bonds effectively prevents solvent peaks from overlapping with both the unreacted oxygenated electrolyte components and the newly formed pyrolysis products during gas chromatography-mass spectrometry (GC-MS) analysis. All resulting CCl₄-containing waste liquids were strictly collected as halogenated organic waste and disposed of according to standard laboratory safety protocols.

2.3. Analysis Methods

The mass-loss profiles during the catalytic pyrolysis of the simulated spent electrolyte mixed with Li-, Ni-, Co-, and Mn-based composite catalysts (CaO-ZSM-5* combined with individual metal additives: LiOH, NiO, Co₃O₄, or MnO₂) were determined via thermogravimetric analysis (TGA; TG 209 F1 Libra, NETZSCH, Selb, Germany). An Al₂O₃ crucible was charged with 5–10 mg of the formulated mixture for each test. The thermal programming was set to increase from 30 to 800 °C at a heating rate of 20 °C/min, maintained under a continuous purge of high-purity N₂ (30 mL/min).

The gaseous and liquid products derived from the catalytic pyrolysis were collected separately and subjected to qualitative and semi-quantitative analyses using a gas chromatography-mass spectrometry system (GC-MS; GCMS-QP2020 NX, Shimadzu, Kyoto, Japan). Different chromatographic columns and methods were selected for the pyrolysis gas and oil. For the pyrolysis gas analysis, a DB-PLOT Q capillary column (30 m × 0.32 mm × 10 μm) was utilized. The injection, ion source, and interface temperatures were uniformly maintained at 200 °C, while the column oven was initially held at 40 °C for 2 min, ramped to 200 °C at a rate of 6 °C/min, and subsequently maintained for 15 min. High-purity helium served as the carrier gas at a flow rate of 1.60 mL/min with a split ratio of 10:1. Mass spectra were acquired over a mass-to-charge ratio (m/z) range of 10–300, and a 1.0 min solvent cut time was applied to exclude the N₂ purge gas. For the characterization of the pyrolysis oil, an Rtx-5MS capillary column (30 m × 0.25 mm × 0.25 μm) was employed. The temperatures for the injection port, ion source, and interface were set to 280 °C, 230 °C, and 280 °C, respectively. Helium was used as the carrier gas at a flow rate of 1.23 mL/min with a split ratio of 20:1. The column oven followed a multi-stage program: initially held at 40 °C for 2 min, increased to 200 °C at 6 °C/min (held for 2 min), and further raised to 280 °C at 8 °C/min (held for 10 min). The mass scanning range was set to m/z 33–650, with a 3.0 min solvent delay to protect the detector. Data acquisition and processing were performed using GCMSsolution software (version 4.45, Shimadzu Corporation, Kyoto, Japan), and chemical components were identified by cross-referencing with the National Institute of Standards and Technology (NIST) 2020 mass spectral library, with a similarity index threshold of >70 applied to ensure identification reliability. To ensure analytical accuracy, the N₂ purge gas peak was strictly excluded from the pyrolysis gas data. For the pyrolysis oil, background and impurity peaks identified from a separately analyzed CCl₄ solvent blank were completely removed during data integration. For the semi-quantitative analysis, an area normalization method was employed. This calculation was based solely on the total integrated area of all successfully identified product peaks. These reliably identified components accounted for more than 90% of the total integrated peak area, while unidentified trace peaks and background noise were excluded from this normalization calculation. It should be explicitly noted that these results solely represent the relative contents of the detected products to reflect the compositional changes across different experimental conditions, rather than absolute mass yields or strict elemental balances. In this study, light hydrocarbons refer to C₁–C₆ non-aromatic hydrocarbons detected in the pyrolysis gas.

The phase composition and surface chemical states of the pyrolysis residues were analyzed using X-ray diffraction (XRD; D/MAX 2500 V/PC, Rigaku, Tokyo, Japan; scanning rate, 5°/min; 2θ range, 5–80°) and X-ray photoelectron spectroscopy (XPS; AXIS Ultra DLD, Kratos, Kawasaki, Japan), respectively. Furthermore, the surface morphology and localized elemental distribution of the residues were characterized by field emission scanning electron microscopy coupled with energy-dispersive X-ray spectroscopy (FESEM-EDS; GeminiSEM 300, Carl Zeiss, Oberkochen, Germany).

3. Results and Discussion

3.1. Thermal Behavior and Reaction Window of Electrolyte Catalytic Pyrolysis

To systematically establish the reaction window and investigate the element-specific thermal degradation behaviors, thermogravimetric and derivative thermogravimetric (TG-DTG) analyses of the simulated spent electrolyte mixed with Li-, Ni-, Co-, and Mn-based composite catalysts were conducted. As depicted in Figure 1, the catalytic pyrolysis process of the electrolyte in these catalyst-containing mixtures generally exhibits intense decomposition below 300 °C, followed by distinct mid-to-high temperature degradation pathways depending on the metal catalyst.

Low-temperature decomposition (Stages I and II): Across all four catalytic systems, the primary mass loss occurs in the first two stages. Stage I occurs below 100 °C, featuring sharp DTG peaks at 48.2 °C (Li), 55.0 °C (Ni), 52.8 °C (Co), and 52.8 °C (Mn). This initial stage corresponds to the volatilization of low-boiling-point solvents (e.g., EMC) and initial mild decomposition, accounting for mass losses ranging from 10.09% to 16.66%. Subsequently, Stage II represents the most drastic thermal degradation phase, spanning approximately 100 °C to 300 °C. The prominent DTG peaks in this region are observed at 129.8 °C (Li), 134.8 °C (Ni), and 144.8 °C (Co and Mn). This major weight loss—constituting 15.44% to 23.15% of the total mass—is primarily attributed to the intense cleavage of bulk carbonate solvents (DEC, EC), the thermal decomposition of LiPF₆, and the breakdown of functional additives.

Mid-to-high temperature degradation (Stages III–V): In the higher temperature regions (>300 °C), the thermal behaviors diverge significantly, reflecting the distinct catalytic activities of the different metal species. The Li-catalyzed system (Figure 1a) exhibits the most prolonged stepwise decomposition, with distinct secondary DTG peaks at 342.4 °C (Stage III) and 499.8 °C (Stage IV), indicating the continuous capability of Li⁺ to crack heavier intermediate fragments. In contrast, the Ni and Mn systems (Figure 1b–d) exhibit only minor weight loss peaks around 387.3 °C and 389.8 °C, respectively.

Total mass loss and determination of the reaction window: Ultimately, the total mass loss reaches 46.24% for Li, 43.18% for Co (Figure 1c), and 44.19% for Mn (Figure 1d). Notably, the Ni-catalyzed system (Figure 1b) records a significantly lower total mass loss of only 29.63%. This pronounced difference suggests that Ni promotes the formation of more stable solid carbonaceous residues or heavy inorganic salts within the pyrolysis residue, effectively suppressing the excessive volatilization of mass into the gas phase. Crucially, the TG curves across all four catalytic systems demonstrate that the mass loss processes begin to plateau as the temperature approaches 500 °C. By 530 °C, the derivative mass change (DTG) for all samples approaches zero, indicating that the thermal decomposition reactions of the spent electrolyte are essentially complete. Therefore, to ensure the thorough degradation of the spent electrolyte and provide sufficient reaction severity for its directional conversion under vacuum conditions, 530 °C was determined to be the optimal and sufficient reaction temperature for the subsequent catalytic pyrolysis experiments.

3.2. Role of Individual Metal Elements During In Situ Catalytic Pyrolysis of Electrolyte

3.2.1. Product Comparison Between Monometallic-Catalyzed and Control Pyrolysis

To investigate the specific catalytic roles of the individual metal elements (Li, Ni, Co, and Mn) typical of ternary lithium batteries during the pyrolysis of spent electrolytes, experiments employing CaO-ZSM-5* mixed with individual additives of LiOH, NiO, Co₃O₄, or MnO₂ (monometallic catalytic systems, designated as groups E-L, E-N, E-C, and E-M, respectively) were conducted. These experiments were performed alongside the CaO-ZSM-5* control group (E-0) at 530 °C and an initial pressure of 100 Pa, with a consistent mass ratio of metal additive to spent electrolyte to CaO-ZSM-5* maintained at 1:2:3. The detailed composition and distribution of the generated pyrolysis gas and oil were analyzed by GC–MS, as presented in Figure 2a and Tables S3 and S4.

Products of control group. For the control group E-0, the pyrolysis gas was dominated by hydrocarbons (33.75%), alcohols (10.57%), esters (14.09%, originating from incomplete pyrolysis of EMC/DEC and carbonates), toxic organics (15.55%), other oxygenates (22.42%), and CO₂ (3.62%). Detailed composition analysis showed that the toxic organics were composed of aldehydes (6.14% acetaldehyde), aromatics (0.93% benzene), nitriles (0.25% 2-propenenitrile), as well as significant amounts of sulfides (5.05%) and fluorides (3.18%). Based on the analysis of elemental composition and molecular structure, the substantial content of esters (14.09%) and remaining oxygenated organics (39.13%, comprising alcohols, aldehydes, and other oxygenates) directly exposes the limitations of the CaO-ZSM-5* catalyst in promoting C–O bond cleavage in carbonates and the decarboxylation/oxygen-fixation of carboxyl-containing compounds, preventing the directional upgrading of spent electrolytes into light hydrocarbons. Meanwhile, the combined content of sulfides and fluorides in the pyrolysis gas reached 8.23%, and small amounts of toxic products such as dimethyl disulfide and 2-(diisopropylamino)ethyl fluoride were detected in the pyrolysis oil of E-0, confirming that CaO-ZSM-5* alone is insufficient for deep detoxification. It is worth noting that the absence of P-containing compounds in both oil and gas phases suggests effective P immobilization within the pyrolysis residues. Additionally, a disproportionately high relative content of tributylamine (87.43%) was detected in the E-0 pyrolysis oil (Table S4). This is not a synthesized product but an intrinsic trace heavy additive from the commercial electrolyte, which was artificially enriched in the relative area normalization due to the extremely low liquid yield and the lack of metallic cracking catalysts in the blank group.

Products of Li-catalyzed electrolyte pyrolysis. Compared with the control group, Table S3 shows that the addition of LiOH significantly increased the hydrocarbon content in the pyrolysis gas of the E-L group to 91.08%. This significant increase is attributed to the fact that the addition of LiOH promoted the formation of 1-butene, (E)-2-butene, and (Z)-2-butene, with their contents increasing from 7.42%, 3.03%, and 1.06% to 10.69%, 14.19%, and 25.27%, respectively, accompanied by the formation of a large number of newly formed hydrocarbon compounds. Simultaneously, the content of esters and toxic organics decreased to 1.44% and 5.59%, respectively, and alcohols and other oxygenates completely disappeared. Analysis shows that the 12.65% decrease in ester content indicates that Li⁺ effectively promoted the breaking of C-O bonds in carbonates, while the 38.55% decrease in the remaining oxygenated organics indicates that Li⁺ can also enhance the decarboxylation and oxygen fixation process of carboxyl-containing products catalyzed by CaO-ZSM-5*, ultimately leading to a 57.33% increase in hydrocarbon content in the E-L group. Furthermore, fluorides were not detected in the pyrolysis gas. Regarding sulfur detoxification, the addition of LiOH only reduced the sulfide content in the pyrolysis gas by 4.38%, leaving a 0.67% residue, indicating that Li⁺ cannot completely remove sulfur from the pyrolysis gas. However, for the pyrolysis oil (Table S4), no S-containing or fluorinated compounds were detected for the E-L group. There was no significant difference in the content of aromatics and nitriles in the pyrolysis gas between the E-L and E-0 groups, both showing low proportions.

Products of Ni, Co or Mn-catalyzed electrolyte pyrolysis. Compared to group E-0, Table S3 shows that the addition of NiO, Co₃O₄, and MnO₂ also reduced the total ester content in the pyrolysis gas of groups E-N, E-C, and E-M by 11.42%, 7.43%, and 10.16%, respectively; meanwhile, the total content of alcohols, aldehydes, ketones, and other oxygenates decreased by 38.39%, 32.64%, and 36.70%, respectively. Ultimately, this led to increases in hydrocarbon content by 45.02%, 13.42%, and 33.65%, respectively. Based on this analysis, it is believed that Ni²⁺, Co^2/3+, and Mn⁴⁺, similar to the previously analyzed Li⁺, can attack various C–O bond sites within carbonates at high temperatures, cleaving them and synergizing with CaO-ZSM-5* to catalyze the decarboxylation and oxygen fixation of carboxyl-containing products, thereby promoting the targeted conversion of carbonates to light hydrocarbons. No sulfides or fluorides were detected in the pyrolysis gas from the NiO, Co₃O₄, and MnO₂ catalytic systems, indicating that the presence of Ni²⁺, Co^2/3+, and Mn⁴⁺ inhibits the formation of gaseous sulfides and fluorides, which is highly beneficial for the deep detoxification of the spent electrolyte pyrolysis gas phase. Table S4 shows that no fluorinated compounds were detected in the pyrolysis oils of the E-N, E-C, and E-M groups, whereas these metal oxides exhibited different sulfur detoxification capabilities. Co₃O₄ completely removed S from the pyrolysis oil, whereas NiO and MnO₂ were less effective in sulfur removal, with S-containing compounds detected in the pyrolysis oils of the E-N and E-M groups at 6.42% and 15.14%, respectively. Table S3 shows that the addition of NiO, Co₃O₄, and MnO₂ all increased the content of aromatics in the pyrolysis gas by more than 10%. Regarding nitriles, the presence of Co^2/3+ and Mn⁴⁺ exhibited a particularly significant impact on their release characteristics in the pyrolysis gas phase. The content of nitriles in the pyrolysis gas of the E-C and E-M groups reached 12.93% and 5.31%, respectively. Compared with the nitrile content of 0.25% in the E-0 group, Co^2/3+ and Mn⁴⁺ significantly promoted the generation of gaseous nitriles.

In summary, (1) CaO-ZSM-5* can achieve the complete immobilization of P in the pyrolysis residue; (2) compared with the control group E-0, the analysis results of the monometallic catalytic pyrolysis systems show that: (i) at high temperatures, Li⁺, Ni²⁺, Co^2/3+, and Mn⁴⁺ can promote the cleavage of C–O bonds at different sites in carbonates and synergize with CaO-ZSM-5* to achieve the decarboxylation and oxygen fixation of carboxyl-containing products, thereby promoting the directional conversion of carbonates to light hydrocarbons; (ii) for nitriles, the presence of Li⁺ and Ni²⁺ has no obvious effect on their gas-phase formation, maintaining a low nitrile content in the pyrolysis gas, whereas Co^2/3+ and Mn⁴⁺ significantly promote their gas-phase generation; (iii) for aromatics, the addition of Li⁺ does not promote the production of benzene in the gas phase, but Ni²⁺, Co^2/3+, and Mn⁴⁺ all significantly promote the gas-phase generation of aromatics; and (iv) for the gas-phase detoxification of F and S, Li⁺, Ni²⁺, Co^2/3+, and Mn⁴⁺ all inhibit the formation of gaseous sulfides and fluorides.

3.2.2. Differences in Catalytic Pyrolysis Products Among Li, Ni, Co, and Mn

Based on the above understanding of monometallic catalytic properties, to explore the differences in the effects of Li⁺, Ni²⁺, Co^2/3+, and Mn⁴⁺ in the spent electrolyte vacuum catalytic pyrolysis system, a detailed comparison was conducted on the composition and distribution of the pyrolysis gas and oil, as well as the substances in the pyrolysis residues, among the E-L, E-N, E-C, and E-M groups.

Formation of hydrocarbons and oxygenated organics. As presented in Table S3, the hydrocarbon content in the pyrolysis gas of the E-L, E-N, E-M, and E-C groups decreased sequentially (91.08%, 78.77%, 67.40%, and 47.17%, respectively), while the content of oxygenated organics (comprising esters, aldehydes, and ketones, with no alcohols or other oxygenates detected) increased sequentially (2.02%, 3.41%, 6.36%, and 13.15%, respectively). Additionally, the total content of carbonate raw materials in the pyrolysis oils (Table S4) of the E-L (3.16%), E-N (6.16%), E-M (37.77%), and E-C (57.31%) groups also exhibited an increasing trend. This indicates that the activities of Li⁺, Ni²⁺, Mn⁴⁺, and Co^2/3+ in promoting the cleavage of C–O bonds at different carbonate sites under vacuum conditions, and synergizing with CaO-ZSM-5* to catalyze the decarboxylation and oxygen fixation of carboxyl-containing products, decreased sequentially. However, the low hydrocarbon content of 47.17% in the E-C group is attributed not only to the low activity of Co^2/3+ in promoting C–O bond cleavage but also to the oxidation of hydrocarbons by reactive oxygen species generated from the high-temperature decomposition of Co₃O₄. Similarly, the reduced hydrocarbon content in the E-M group compared to the E-L/E-N groups is attributed to the aforementioned factors. The generation of reactive oxygen species ultimately led to CO₂ contents reaching 14.28% and 6.86% in the E-C and E-M groups, respectively. This oxidation process is supported by the subsequent XRD and XPS analyses, which indicate the thermal reduction in the initial Co₃O₄ and MnO₂ precursors into lower-valence states (metallic Co⁰ and Mn²⁺), corresponding to the increased relative content of CO₂.

Formation of aromatics and nitriles. Comparisons of aromatics and nitriles were also conducted among the toxic organics. The aromatics content in the pyrolysis gas of the E-L group (3.86%) was significantly lower than that of the E-N (14.21%), E-C (12.47%), and E-M (14.07%) groups, indicating that Li⁺ can inhibit the gas-phase formation of aromatics. This further demonstrates that Li⁺ effectively promotes the targeted conversion of spent electrolytes into light hydrocarbons. Regarding nitriles, their content in the E-C and E-M groups increased to 12.93% and 5.31%, respectively, compared to 0.48% and 0.71% in the E-L and E-N groups, indicating that Li⁺ and Ni²⁺ inhibit the gas-phase formation of nitriles more effectively than Co^2/3+ and Mn⁴⁺. Furthermore, the effects of these metal ions on the gas-phase release of nitriles differed significantly. Analysis of the pyrolysis oil (Table S4) and pyrolysis residues (Figure 2b) revealed that among all monometallic catalytic systems, only the E-N group’s pyrolysis oil contained the electrolyte raw material hexanedinitrile (ADN, 15.09%). This indicates that the activity of Ni²⁺ in promoting ADN pyrolysis is significantly lower than that of the other elements; furthermore, Ni²⁺ promotes the immobilization of a small fraction of ADN pyrolysis products in the pyrolysis residues as Ca(N₃)₂, thereby suppressing their gas-phase formation. Although the nitrile content in the pyrolysis gas of the E-L and E-N systems was similar, Li⁺ promoted the complete decomposition of ADN and caused the vast majority of the pyrolysis products to couple in the pyrolysis oil and pyrolysis residues, fundamentally suppressing nitrile generation in the gas phase. For the Co^2/3+ and Mn⁴⁺ systems, although their pyrolysis oils contained higher N-containing compound contents than the E-L system, and N-containing compounds were detected in the residues of both E-C (Ca(N₃)₂, CaCN₂) and E-M (Mn₃N₂, C₃N₄) groups, their ability to couple and immobilize ADN pyrolysis products in the pyrolysis residues was significantly weaker than that of Li⁺, resulting in a high proportion of nitriles in the pyrolysis gas.

Evolution and immobilization of S- and F-containing species. Li⁺, Ni²⁺, Co^2/3+, and Mn⁴⁺ also exhibited distinct characteristics regarding sulfur detoxification. No sulfides were detected in the pyrolysis gas and no S-containing compounds were detected in the pyrolysis oil of the E-C group, while only 0.67% of sulfides were found in the pyrolysis gas of the E-L group. As indicated by Figure 2b, this is attributed to the effective mineralization and immobilization of S in the pyrolysis residues by Li⁺ (forming CaSO₄) and Co^2/3+ (forming CaSO₄ and Co₉S₈), thereby suppressing the gas-phase formation of sulfides. Similarly, no sulfides were found in the pyrolysis gas of the E-N and E-M systems; however, the content of S-containing organics in their pyrolysis oils reached 6.42% (comprising ethanethiol and PS) and 15.14% (comprising ethanethiol), respectively. This suggests that Ni²⁺ and Mn⁴⁺ promote the coupling of S-containing gaseous fragments with hydrocarbon fragments in the pyrolysis oil, thereby inhibiting the formation of gaseous sulfides. Naturally, Figure 2b also demonstrates that the mineralization and immobilization of S in the pyrolysis residues of the E-N (as CaSO₄ and elemental S) and E-M (as CaSO₄ and MnS₂) groups effectively suppressed the gas-phase release of sulfides. Regarding fluorine detoxification, as detailed above, no fluorides were detected in the pyrolysis gas and no fluorinated compounds were detected in the pyrolysis oil for the Li⁺-, Ni²⁺-, Co^2/3+-, and Mn⁴⁺-catalyzed systems, with XRD analysis supporting the mineralization and immobilization of F as CaF₂ across these monometallic systems.

In summary, a comprehensive comparison of the pyrolysis products among the E-L, E-N, E-C, and E-M groups reveals the following: (1) The activities of Li⁺, Ni²⁺, Mn⁴⁺, and Co^2/3+ in promoting the cleavage of C–O bonds in carbonates, and synergizing with CaO-ZSM-5* to catalyze the decarboxylation and oxygen fixation of carboxyl-containing products, decrease sequentially. (2) Compared to Li⁺ and Ni²⁺, the oxygen atoms bound to Co^2/3+ and Mn⁴⁺ are more prone to be released from the lattice structure under vacuum conditions, generating reactive oxygen species that lead to the oxidation of hydrocarbons. (3) Regarding aromatics, Li⁺ significantly inhibits their gas-phase formation compared to Ni²⁺, Co^2/3+, and Mn⁴⁺. (4) Regarding nitriles, Li⁺ suppresses their gas-phase formation by promoting the coupling of ADN pyrolysis products in the pyrolysis oil and residues. The suppression by Ni²⁺ is primarily attributed to its low catalytic activity for ADN cleavage, whereas Co^2/3+ and Mn⁴⁺ significantly promote the gas-phase formation of nitriles. (5) Li⁺ and Co^2/3+ inhibit the generation of gaseous sulfides by promoting the immobilization of PS pyrolysis products in the pyrolysis residues, whereas Ni²⁺ and Mn⁴⁺ inhibit their gas-phase formation by promoting the coupling of PS pyrolysis products in the pyrolysis oil and residues. (6) Li⁺, Ni²⁺, Co^2/3+, and Mn⁴⁺ all showed favorable fluorine detoxification performance, with no fluorides detected in the pyrolysis gas and no fluorinated compounds detected in the pyrolysis oil.

3.3. Synergistic/Antagonistic Effect of Li, Ni, Co, and Mn Multi-Metallic Combination During In Situ Catalytic Pyrolysis of Electrolyte

3.3.1. Bimetallic Antagonism in Hydrocarbon Upgrading

The preceding analysis of the monometallic catalytic systems revealed distinct catalytic advantages and limitations. To explore potential synergistic and antagonistic interactions among these elements, experiments were conducted by mixing CaO-ZSM-5* with pair combinations of the metal additives (bimetallic catalytic systems, designated as groups E-LN, E-LC, E-LM, E-NC, E-NM, and E-CM). Under the same reaction conditions (530 °C, 100 Pa) and mass ratios (1:2:3) as those used in the monometallic catalytic systems, the composition and distribution of the pyrolysis gas and oil were analyzed by GC–MS (Figure 3a and Tables S5 and S6), while the crystal structures of the residues were characterized by XRD (Figure 3b).

Regarding the hydrocarbon content in the pyrolysis gas (Figure 3a and Table S5), the bimetallic catalytic systems (E-LN (64.84%), E-LC (35.42%), E-LM (55.10%), E-NC (28.44%), E-NM (48.42%), and E-CM (44.20%)) exhibited significantly lower hydrocarbon contents compared to the monometallic catalytic systems. For instance, in all Li⁺-catalyzed bimetallic systems, the introduction of other metal oxides resulted in a 26.24–55.66 percentage-point decrease in hydrocarbon content for the E-LN, E-LC, and E-LM groups compared to the E-L group. Furthermore, the addition of Co^2/3+ and Mn⁴⁺ also led to a sharp decline in hydrocarbon content in the Ni²⁺-catalyzed systems, and similarly, the introduction of Co^2/3+ decreased the hydrocarbon content in the Mn⁴⁺-catalyzed system. Notably, the hydrocarbon content in the pyrolysis gas of the E-NC group was even lower than that of the control group (E-0). These results suggest an antagonistic trend, where the introduction of metal elements with lower catalytic activities for C–O bond cleavage and decarboxylation/oxygen-fixation was associated with a decrease in the hydrocarbon content of the pyrolysis gas.

XRD results (Figure 3b) reveal that the metal compounds in all bimetallic catalytic systems fused at high temperatures to form composite metal oxides (M_xN_yO_z, where M and N represent different elements among Li, Ni, Co, and Mn). The resulting M_xN_yO_z inherently increases the difficulty of releasing active catalytic ions, thereby inhibiting the highly active metals from promoting C–O bond cleavage in carbonates and synergizing with CaO-ZSM-5* to catalyze the decarboxylation and oxygen fixation of carboxyl-containing products.

This deduction is further supported by the high contents of unreacted carbonate raw materials (DEC and EC: 43.65%, 57.51%, 45.15%, 75.61%, 71.42%, and 92.23%) in the pyrolysis oils of the E-LN, E-LC, E-LM, E-NC, E-NM, and E-CM groups (Table S4), as well as the overall increase in the proportions of oxygenated organics in the pyrolysis gas (10.37%, 27.83%, 13.49%, 19.43%, 8.90%, and 10.76%) compared to most monometallic catalytic systems. In terms of structure and composition, the formation of M_xN_yO_z similarly promotes the release of lattice oxygen from NiO, Co₃O₄, and MnO₂, resulting in the deep oxidation of target hydrocarbon products into CO₂ during vacuum catalytic pyrolysis. Consequently, the CO₂ contents in the respective groups reached 8.13%, 21.48%, 12.21%, 26.88%, 10.91%, and 16.55%, further contributing to the reduced hydrocarbon contents.

The gas-phase formation of nitriles and aromatics was also evaluated. Compared to the monometallic catalytic systems, the bimetallic combinations significantly inhibited the generation of naphthalene and o-xylene in the pyrolysis oil, but exhibited a significant promotion effect on the gas-phase formation of aromatics and nitriles. As a result, their combined content in the pyrolysis gas of the E-LN, E-LC, E-LM, E-NC, E-NM, and E-CM groups increased to 16.66%, 15.27%, 19.20%, 25.25%, 31.77%, and 28.49%, respectively.

Furthermore, the consistent detection of unreacted ADN in the pyrolysis oils of all Ni²⁺- containing bimetallic catalytic systems suggests an antagonistic interaction, where Ni²⁺ may partially inhibit the catalytic activities of Li⁺, Co^2/3+, and Mn⁴⁺ towards ADN pyrolysis. Notably, Ni²⁺ exhibited the most pronounced inhibitory effect on the catalytic activity of Li⁺, as evidenced by the substantial retention of unreacted ADN (39.38%) in the E-LN pyrolysis oil, which consequently resulted in a mere 1.72% of nitriles and only 3.89% of other N-containing compounds in the pyrolysis gas and oil, respectively.

Regarding the detoxification of F and S, all bimetallic catalytic systems showed favorable immobilization of these elements within the pyrolysis residues, with no sulfides or fluorides detected in the pyrolysis gas and no S-containing or fluorinated compounds detected in the pyrolysis oil, suggesting a synergistic trend in heteroatom immobilization among the paired combinations of Li⁺, Ni²⁺, Co^2/3+, and Mn⁴⁺ species.

In summary, (1) the formation of composite metal oxides M_xN_yO_z in the bimetallic catalytic combinations: (i) increases the difficulty of releasing active catalytic ions, thereby inhibiting the active metal elements from promoting C–O bond cleavage and synergizing with CaO-ZSM-5* for decarboxylation and oxygen fixation; and (ii) aggravates the release of lattice oxygen from NiO, Co₃O₄, and MnO₂, resulting in the oxidation of hydrocarbon target products. (2) Compared with the monometallic catalytic systems, the bimetallic combinations: (i) inhibit the formation of naphthalene and o-xylene in the pyrolysis oil, but promote the production of aromatics in the pyrolysis gas; (ii) promote the gas-phase formation of nitriles to varying degrees, although Ni²⁺ strongly antagonizes the catalytic activities of Li⁺, Co^2/3+, and Mn⁴⁺ for ADN pyrolysis (with the Ni²⁺/Li⁺ antagonism being the strongest, suppressing gas-phase nitriles to below 2% in the E-LN group); and (iii) show a favorable synergistic trend in the detoxification and immobilization of F and S across all bimetallic combinations, as no sulfides or fluorides were detected in the pyrolysis gas, and no S-containing or fluorinated compounds were detected in the pyrolysis oil.

3.3.2. Progressive Suppression in Trimetallic Catalytic Systems

As demonstrated in the bimetallic experiments, the pairing of different metal elements induced distinct interactive catalytic effects. To further dissect the combined catalytic characteristics among these elements in more complex systems, experiments employing CaO-ZSM-5* mixed with ternary combinations of the metal additives (trimetallic catalytic systems, designated as groups E-LNC, E-LNM, E-LCM, and E-NCM, respectively) were conducted. Under the consistent reaction parameters (530 °C, 100 Pa) and mass ratios (1:2:3) maintained across all experimental groups, the detailed composition and distribution of the trimetallic catalytic pyrolysis gas and oil were analyzed by GC–MS (Figure 4a and Tables S7 and S8), while the crystal structures of the residues were characterized by XRD (Figure 4b).

Figure 4a and Table S7 show that the formation of M_xN_yO_z in the trimetallic catalytic pyrolysis systems significantly inhibits the release of active catalytic ions compared to the bimetallic catalytic pyrolysis systems. This hinders the metal elements’ ability to promote C–O bond cleavage in carbonates and the synergistic decarboxylation and oxygen fixation of carboxyl-containing products catalyzed by CaO-ZSM-5*, thus impacting the targeted conversion of carbonates to light hydrocarbons. Overall, compared with most of their bimetallic counterparts, the trimetallic catalytic systems showed a further upward trend in the content of oxygenated organics in the pyrolysis gas, reaching 18.67%, 21.82%, 20.05%, and 20.40% for the E-LNC, E-LNM, E-LCM, and E-NCM groups, respectively. Correspondingly, the hydrocarbon content in the pyrolysis gas phase showed an overall downward trend, reaching 48.50%, 39.03%, 26.08%, and 24.50%, respectively. Simultaneously, the total content of carbonate feedstock (EC and DEC) in the pyrolysis oils of the E-LNC, E-LNM, E-LCM, and E-NCM groups (Table S8) reached 57.11%, 66.59%, 79.25%, and 73.35%, respectively. This further confirms that the formation of M_xN_yO_z restricts the conversion of carbonate feedstocks during catalytic pyrolysis.

As for aromatics, similar to the bimetallic catalytic systems, the trimetallic combinations strongly inhibited their formation in the pyrolysis oil, whereas their content in the pyrolysis gas reached 13.66%, 17.94%, 23.59%, and 34.78%, respectively. Compared to the bimetallic catalytic pyrolysis systems, the formation of M_xN_yO_z did not further exacerbate the release of lattice oxygen. Consequently, the CO₂ content in the pyrolysis gas from the trimetallic catalytic systems was comparable to that from the bimetallic catalytic systems containing the same metal elements.

Regarding nitriles, their contents in the pyrolysis gas were similar to those in the bimetallic catalytic systems, and the trimetallic catalytic combinations did not exhibit an inhibitory effect on their formation. Specifically, these compounds still accounted for 4.95%, 7.86%, 6.97%, and 6.42% of the pyrolysis gas in the E-LNC, E-LNM, E-LCM, and E-NCM groups, respectively. Furthermore, all trimetallic catalytic combinations showed a favorable synergistic trend in the detoxification of F and S. No sulfides or fluorides were detected in the pyrolysis gas, and no S-containing or fluorinated compounds were detected in the pyrolysis oil, suggesting that Li⁺, Ni²⁺, Co^2/3+, and Mn⁴⁺ synergistically promoted the immobilization of these toxic elements within the pyrolysis residues (Figure 4b). A large amount of unreacted ADN was still found in the pyrolysis oils of all Ni²⁺-containing trimetallic catalytic systems, which further supports the inference that Ni²⁺ can inhibit the catalytic activities of Li⁺, Co^2/3+, and Mn⁴⁺ towards ADN pyrolysis.

In summary, (1) compared to the bimetallic catalytic systems, the formation of the composite metal oxides M_xN_yO_z in the trimetallic catalytic pyrolysis systems (i) further increases the difficulty of releasing active catalytic ions, thereby more significantly inhibiting the metal elements from promoting C–O bond cleavage in carbonates and synergizing with CaO-ZSM-5* for the decarboxylation and oxygen fixation of carboxyl-containing products; and (ii) does not have a significant impact on the release of lattice oxygen from NiO, Co₃O₄, and MnO₂. (2) Similar to the bimetallic catalytic systems, the trimetallic combinations still: (i) maintain a strong inhibition on the formation of aromatics in the pyrolysis oil, but promote their enrichment in the pyrolysis gas phase; (ii) fail to exhibit an inhibitory effect on the gas-phase formation of nitriles, although the inhibitory effect of Ni²⁺ on the catalytic activities of Li⁺, Co^2/3+, and Mn⁴⁺ towards ADN pyrolysis remains evident; and (iii) showed a favorable synergistic trend in the detoxification and immobilization of F and S across all trimetallic combinations, as no sulfides or fluorides were detected in the pyrolysis gas, and no S-containing or fluorinated compounds were detected in the pyrolysis oil.

3.3.3. Analysis of Tetrametallic Catalytic Pyrolysis Products

Composition of pyrolysis gas and oil. To fully evaluate the combined catalytic effects of the cathode metal elements in ternary lithium batteries during vacuum catalytic pyrolysis, an experiment employing CaO-ZSM-5* mixed with the quaternary combination of the metal additives (tetrametallic catalytic system, designated as group E-LNCM) was conducted. Under the consistent reaction parameters (530 °C, 100 Pa) and mass ratio (1:2:3), the detailed composition and distribution of the resulting pyrolysis gas and oil were analyzed by GC-MS (Figure 4a and Tables S7 and S8).

Table S7 shows that the pyrolysis gas of the E-LNCM group is primarily composed of hydrocarbons, alcohols, esters, toxic organics, other oxygenates, and CO₂, with proportions of 13.80%, 11.86%, 10.76%, 20.15%, 3.44%, and 39.99%, respectively. The toxic organics include 7.97% acetonitrile, 5.57% acetaldehyde, 2.55% acetone, and 4.06% benzene. Compared to the mono-, bi-, and trimetallic catalytic pyrolysis systems, the high proportion (34.18%) of oxygenated organics in the pyrolysis gas indicates that the formation of M_xN_yO_z in the tetrametallic catalytic system significantly inhibits the active catalytic ions from promoting C–O bond cleavage in carbonates and synergizing with CaO-ZSM-5* for the decarboxylation and oxygen fixation of carboxyl-containing products. Furthermore, the 39.99% CO₂ content indicates that the formation of M_xN_yO_z in the tetrametallic catalytic system further promotes the release of lattice oxygen from NiO, Co₃O₄, and MnO₂, leading to a significant reduction in the hydrocarbon content in the pyrolysis gas.

Detailed analyses of the toxic organics were also performed. Regarding aromatics, although benzene accounted for only 4.06% of the pyrolysis gas, analysis based on the high CO₂ content suggests that this reduction is attributed to its deep oxidation by reactive oxygen species. As for nitriles, similar to the bimetallic and trimetallic catalytic pyrolysis systems, the tetrametallic catalytic system failed to exhibit an inhibitory effect on their gas-phase formation, and unreacted ADN was still detected in the pyrolysis oil (Table S8). This demonstrates that Ni²⁺ can still inhibit the catalytic activities of Li⁺, Co^2/3+, and Mn⁴⁺ towards ADN pyrolysis. Analysis of gas-phase sulfides and fluorides, together with oil-phase S-containing and fluorinated compounds, reveals that the tetrametallic catalytic system exhibits characteristics consistent with the bimetallic and trimetallic catalytic systems, suggesting favorable immobilization of F and S heteroatoms within the pyrolysis residue. To provide a clearer overview of the extensive GC–MS data, the lumped relative contents of key product categories in the pyrolysis gas across the blank control group and metal-catalyzed groups are summarized in

Table 2. Lumped relative contents of key product categories in the pyrolysis gas derived from GC–MS analysis.

Groups	Total Light Hydrocarbons (C1–C6)/%	Total Oxygenates/%	Aromatics/%	Nitriles/%	Sulfides/%	Fluorides/%	CO2/%
E-0	33.75	53.22	0.93	0.25	5.05	3.18	3.62
E-L	91.08	2.02	3.86	0.48	0.67	0	1.89
E-N	78.77	3.41	14.21	0.71	0	0	2.90
E-C	47.17	13.15	12.47	12.93	0	0	14.28
E-M	67.40	6.36	14.07	5.31	0	0	6.86
E-LN	64.84	10.37	14.94	1.72	0	0	8.13
E-LC	35.42	27.83	7.76	7.51	0	0	21.48
E-LM	55.10	13.49	12.12	7.08	0	0	12.21
E-NC	28.44	19.43	13.49	11.76	0	0	26.88
E-NM	48.42	8.90	23.46	8.31	0	0	10.91
E-CM	44.20	10.76	16.13	12.36	0	0	16.55
E-LNC	48.50	18.67	13.66	4.95	0	0	14.22
E-LNM	39.03	21.82	17.94	7.86	0	0	13.35
E-LCM	26.08	20.05	23.59	6.97	0	0	23.31
E-NCM	24.50	20.40	34.78	6.42	0	0	13.90
E-LNCM	13.80	34.18	4.06	7.97	0	0	39.99

Note: Total light hydrocarbons correspond to the hydrocarbons category in Tables S3, S5 and S7, which exclude aromatics; total oxygenates exclude CO₂.

.

Characterization of pyrolysis residue. To preliminarily analyze the pyrolysis residue of the tetrametallic catalytic system, SEM-EDS was used to observe the micromorphology of the E-LNCM sample (Figure 4c) and to examine the surface distribution of F, P, S, and other elements (Ca, Ni, Co, and Mn). The micromorphological characterization revealed that CaO-ZSM-5* and the metal compounds aggregated during the catalytic pyrolysis process, resulting in significant agglomeration. The EDS elemental mapping revealed a strong spatial correlation between the distributions of F, P, and S, and that of Ca. It is preliminarily inferred that, similar to the mono-, bi-, and trimetallic catalytic systems, the F, P, and S in the E-LNCM group’s electrolyte bind to Ca and become immobilized via mineralization after catalytic pyrolysis. However, the distributions of F, P, and S did not exhibit obvious spatial overlap with those of Ni, Co, and Mn.

To further determine the chemical forms of the removed F, P, and S, XRD analysis was performed on the pyrolysis residue of the E-LNCM group (Figure 4b). The XRD patterns clearly exhibited distinct diffraction peaks for CaF₂, Ca₂P₂O₇, and CaSO₄, confirming the inference that F, P, and S bind to Ca and are immobilized via mineralization as calcium salts during the catalytic pyrolysis process. Additionally, weak diffraction peaks corresponding to NiSO₄, MnSO₃, MnS₂, and Co₉S₈ were detected, indicating that a minor fraction of the sulfur was also mineralized by bonding with the cathode metals. Simultaneously, the XRD results revealed that the metal oxides (NiO, Co₃O₄, and MnO₂) underwent significant phase transformations during the catalytic pyrolysis, as evidenced by the emergence of distinct peaks for metallic Ni, Co, and MnO. Furthermore, the appearance of diffraction peaks for Li₀.₅₁Ni_1.16O₂ confirms the formation of the composite metal oxide M_xN_yO_z in the tetrametallic catalytic system.

To further clarify the specific changes in metallic elements during the catalytic pyrolysis process, XPS analysis of the surface chemistry of the E-LNCM sample was performed, as shown in Figure 4d–f. The Ni 2p_3/2 spectrum can be deconvoluted into two main peaks, located at approximately 852.72 and 855.51 eV, which are assigned to metallic Ni⁰ and oxidized Ni^x+ species, respectively. This is consistent with the detection of metallic Ni and Li_0.51Ni_1.16O₂ in the XRD results. For Co, the binding energy of the main peak of Co 2p_3/2 is 778.59 eV, indicating that Co exists primarily as metallic Co⁰. Although Co₉S₈ is present, its corresponding valence state was not detected by XPS due to its low content, as evidenced by the weak diffraction peaks in the XRD pattern. In addition, the main peaks of Mn 2p_3/2 and Mn 2p_1/2 are located at approximately 641.75 and 646.95 eV, respectively, demonstrating that Mn exists as Mn²⁺. The decrease in the valence states of these metal elements can be attributed to the reducing action of the pyrolytic carbon and reductive gases generated in the pyrolysis system.

3.4. Mechanism of In Situ Catalytic Pyrolysis for Light Hydrocarbon Conversion and Electrolyte Detoxification by Li, Ni, Co, and Mn Metals in Cathode Materials

Based on the aforementioned analyses, a plausible mechanism for the deep detoxification and directional conversion of spent electrolytes into light hydrocarbons during vacuum catalytic pyrolysis, synergistically driven by the cathode metals and CaO-ZSM-5*, is proposed. To elucidate this synergy, it is essential to first explicitly distinguish the respective contributions of the acid–base catalysis provided by the basic matrix and the specific metal-driven reactions:

Role of Acid–Base Catalysis (CaO-ZSM-5*): The bifunctional CaO-ZSM-5* matrix governs the acid-base neutralization and secondary shape-selective upgrading. Specifically, the strong basic sites provided by CaO act as chemical traps, rapidly neutralizing and mineralizing acidic primary pyrolysis fragments (e.g., capturing HF and H_xP_yO_z to form stable CaF₂ and Ca₂P₂O₇). Meanwhile, the acidic sites and the unique microporous structure of the ZSM-5 framework facilitate the secondary dehydration of intermediate alcohols and assist in the decarboxylation of carboxyl-containing fragments. However, as evidenced by the control group (E-0), acid-base catalysis alone is insufficient to initiate the complete cleavage of stable carbonate solvents.

Role of Metal-Driven Reactions (Li, Ni, Co, Mn): The metal species bridge this reactivity gap via strong coordination and redox mechanisms. Metal cations with high Lewis acidity (such as Li⁺ and Ni²⁺) directly polarize and attack the robust C–O and C=O bonds within the carbonate molecules, driving the primary initial cleavage of the electrolyte solvents. Furthermore, specific transition metals govern redox-driven pathways; for instance, Co^2/3+ and Mn⁴⁺ facilitate the release of reactive lattice oxygen, leading to deep oxidation, while Ni²⁺ exhibits unique coordination behaviors that antagonize the primary cleavage of ADN.

Building upon these decoupled baseline roles, the detailed migration-reaction coupling mechanism is elucidated as follows:

(1) 

Li⁺ and Ni²⁺, in synergy with CaO-ZSM-5*, dominate the process of the directional conversion of spent electrolytes into light hydrocarbons during vacuum catalytic pyrolysis.

The specific analysis is as follows: (i) Based on the characteristics of the monometallic catalytic pyrolysis, the activities of Li⁺, Ni²⁺, Mn⁴⁺, and Co^2/3+ in synergizing with CaO-ZSM-5* to promote the directional conversion of electrolytes into light hydrocarbons decrease sequentially. Furthermore, under the conditions of 530 °C and 100 Pa, the hydrocarbon content in the pyrolysis gases of the E-L and E-N groups is highly comparable to that of the cathode material-catalyzed group (Group E, 87.19%) from our previous study. Thus, it can be preliminarily deduced that Li⁺ and Ni²⁺ in the cathode material primarily drive the C–O bond cleavage in carbonates and dominate the synergistic decarboxylation and oxygen fixation of carboxyl-containing intermediates catalyzed by CaO-ZSM-5*, ultimately promoting the conversion of the spent electrolyte into light hydrocarbons. (ii) Compared across catalytic systems with the same number of metal components, the pyrolysis gases from the E-L and E-N groups (among the monometallic systems) and the Li⁺/Ni²⁺ co-catalyzed groups (among the multi-metallic systems) consistently exhibited the highest hydrocarbon content. This further corroborates the dominant roles of Li⁺ and Ni²⁺ in synergizing with CaO-ZSM-5* for the directional conversion of spent electrolytes into light hydrocarbons. (iii) The comparison of the aromatics content in the pyrolysis gas of the monometallic catalytic systems reveals that Li⁺ significantly minimizes the final emission of aromatics into the pyrolysis gas compared to Ni²⁺, Mn⁴⁺, and Co^2/3+, which aligns with the low aromatics content observed in Group E (0.84%). Across all catalytic systems, Ni²⁺ consistently inhibits the activities of Li⁺, Co^2/3+, and Mn⁴⁺ towards ADN cleavage, resulting in the retention of unreacted ADN in the pyrolysis oils of all Ni²⁺-containing systems. This corresponds to the massive amount of unreacted ADN feedstock (42.66%) detected in the pyrolysis oil of Group E. Regarding the gaseous nitriles, despite this antagonistic effect, Ni²⁺ fails to mitigate their abundant emission into the pyrolysis gas in the Co^2/3+- and Mn⁴⁺-containing systems, leading to the detection of abundant nitriles in their pyrolysis gases. This significantly contrasts with the low nitrile content (1.78%) in the pyrolysis gas of Group E, implying that Co^2/3+ and Mn⁴⁺ contribute negligibly to the synergistic conversion process. Conversely, the E-LN bimetallic system exhibited the lowest nitrile content and retained a massive amount of ADN in its pyrolysis oil, sharing multiple highly consistent characteristics with the pyrolysis products of Group E. This comprehensive deduction supports the inference that Li⁺ and Ni²⁺ dominate the directional conversion of electrolytes into light hydrocarbons in synergy with CaO-ZSM-5*.

(2) 

The structure of the cathode material itself suggests a sequential activation model for the metal elements.

The specific analysis is as follows: (i) The formation of composite metal oxides (M_xN_yO_z) in the multi-metallic catalytic pyrolysis system increases the energy barrier for ion release. Furthermore, under high-temperature vacuum conditions, the inherent properties of the metal ions inevitably cause them to migrate at distinct rates within the metal oxide lattice. Therefore, it is deduced that the Li⁺, Ni²⁺, Co^2/3+, and Mn⁴⁺ in the NCM cathode material (Li(Ni_1/3Co_1/3Mn_1/3)O₂) lattice are activated sequentially to exert their catalytic effects. It should be acknowledged that owing to the technical challenges of rapid thermal quenching and ex situ re-oxidation under the current vacuum setup, this sequential activation is proposed as a deduced model rather than a directly monitored temporal process. (ii) Based on the above analysis, Li⁺ and Ni²⁺, in synergy with CaO-ZSM-5*, dominate the directional conversion of spent electrolytes into light hydrocarbons, whereas Co^2/3+ and Mn⁴⁺ contribute negligibly. Consequently, the hydrocarbon content of Group E is theoretically expected to be similar to that of the E-LN group; however, the actual content in the E-LN group is significantly lower. This deviation can be elucidated by the specific relationship observed among the hydrocarbon contents: Group E-N < Group E < Group E-L. When this trend is evaluated alongside the analysis in (i), it suggests that Li⁺ initially participates in the catalytic pyrolysis process, and Ni²⁺ subsequently begins to take effect before reaching the critical temperature point for Li⁺ to promote ADN cleavage. This sequential activation mechanism plausibly accounts for the intermediate hydrocarbon content of Group E among these three groups. (iii) From the perspectives of ionic interactions and ionic radii, both the strength of the ionic bonds formed between the cations within the NCM cathode material lattice and O^2-, and the ionic radii of these cations, increase sequentially in the order of Li⁺, Ni²⁺, Mn⁴⁺, and Co^2/3+. Consequently, under high-temperature conditions, the migration rates of Li⁺, Ni²⁺, Mn⁴⁺, and Co^2/3+ within the NCM lattice decrease sequentially. This further supports the inference that Li⁺ in the cathode material initially participates in the catalytic pyrolysis process, followed by the subsequent activation of Ni²⁺.

(3) 

Li⁺, Ni²⁺, and Mn⁴⁺, in synergy with CaO-ZSM-5*, dominate the deep gas-phase detoxification of the spent electrolyte during vacuum catalytic pyrolysis.

The specific analysis is as follows: During the deep gas-phase detoxification of the spent electrolyte, catalyzed by the cathode metals in synergy with CaO-ZSM-5*: (i) Gas-phase detoxification of F. As detailed in our previous study [22], the removal of F from LiPF₆ is entirely attributed to the previously established thermal decomposition and hydrolysis pathway. Meanwhile, FEC is converted to VC at high temperatures, and the HF eliminated during this process subsequently reacts with the CaO supported on ZSM-5 to form CaF₂. The specific reaction is expressed as follows:

$${\mathrm{C}}_3{\mathrm{H}}_3\mathrm{F}{\mathrm{O}}_3 \stackrel{\mathrm{h}\mathrm{i}\mathrm{g}\mathrm{h} \mathrm{t}\mathrm{e}\mathrm{m}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{e},\mathrm{C}\mathrm{a}\mathrm{O}}{\to } {\mathrm{C}}_3{\mathrm{H}}_2{\mathrm{O}}_3+\mathrm{C}\mathrm{a}{\mathrm{F}}_2+{\mathrm{H}}_2\mathrm{O}$$

Additionally, the pyrolysis gas of the E-0 group contained fluoroethane and fluoroethene, which are structurally inferred to result from the complete cleavage of the Cα-O bonds in FEC. However, these fluorinated gases were not detected in any metal-catalyzed systems. This contrast implies that the removal of F from FEC is driven not solely by high temperatures, but also by these metal cations, which promote the cleavage of C–F bonds in some FEC to generate HF, which is subsequently captured by CaO, leading to the immobilization of F as CaF₂ in the pyrolysis residue. (ii) Gas-phase detoxification of S. Results from the metal-catalyzed pyrolysis experiments indicate that Ni²⁺, Co^2/3+, and Mn⁴⁺ can immobilize S via mineralization as salts in the pyrolysis residue. However, only NiSO₄, NiS, MnSO₃, and MnS₂ were detected in Group E, with no sulfur-containing cobalt salts observed. This indicates that Co^2/3+ does not participate in the detoxification of S during the catalytic pyrolysis of the spent cathode material. Based on the residue analyses across different groups, the detoxification pathway for S in Group E is hypothesized as follows: As established in our previous study, PS is attacked by Li⁺ to generate SO₂. Upon its formation, a portion of the SO₂ directly combines with MnO to form MnSO₃, while another fraction is reduced by pyrolytic carbon and hydrocarbons to elemental sulfur (S), which undergoes a redox reaction with metallic Ni to yield NiS. Finally, reactive oxygen species generated from the decomposition of the NCM ternary material oxidize the remaining SO₂ to SO₃, which then bonds with Ni on the NCM lattice surface to form NiSO₄. (iii) Gas-phase detoxification of P. Results from the E-0 group indicate that the CaO-ZSM-5* molecular sieve can effectively remove phosphorus-containing compounds from the pyrolysis gas and oil, immobilizing P via mineralization as Ca₂P₂O₇ in the pyrolysis residue. This identical removal and mineralization behavior is maintained in Group E and all metal-catalyzed systems. The absence of any discernible difference in P removal across these groups demonstrates that the gas-phase detoxification of P from the spent electrolyte is governed by the pathway established in our previous study [22]. In this process, the thermal decomposition of LiPF₆ generates PF₅, which undergoes successive hydrolysis to generate POF₃ and subsequent acidic intermediates (e.g., H₃PO₄ and H₄P₂O₇) that are captured by the CaO support to form Ca₂P₂O₇, with no participation from the cathode metals (Li, Ni, Co, and Mn). (iv) Gas-phase detoxification of nitriles. Results from the multi-metallic catalytic systems indicate that Ni²⁺ consistently inhibits the ADN cleavage activities of Li⁺, Co^2/3+, and Mn⁴⁺. Furthermore, the low nitrile content in the pyrolysis gas and the high retention of unreacted ADN in the pyrolysis oil of the E-LN group demonstrate that Ni²⁺ almost completely suppresses the Li⁺-promoted ADN bond cleavage. Consequently, during the synergistic catalysis dominated by Li⁺ and Ni²⁺, the gas-phase detoxification of nitriles is primarily attributed to this inhibitory effect of Ni²⁺. Although a minor fraction of ADN still undergoes the thermal cleavage reactions shown below, results from the E-L and E-N groups reveal that Li⁺ and Ni²⁺ further promote the coupling of most nitrogen-containing gaseous fragments within the pyrolysis oil and residue. These combined effects account for the exceptionally low nitrile content (1.78%, comprising HCN, C₃H₃N, and C₃H₅N) in the pyrolysis gas of Group E in our previous study [22].

$$\mathrm{N}\mathrm{C}-{\mathrm{C}}_4{\mathrm{H}}_8-\mathrm{C}\mathrm{N}\stackrel{\mathrm{h}\mathrm{i}\mathrm{g}\mathrm{h} \mathrm{t}\mathrm{e}\mathrm{m}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{e},{\mathrm{L}\mathrm{i}}^{+},{\mathrm{C}\mathrm{o}}^{2/3+},{\mathrm{M}\mathrm{n}}^{4+}}{\to } {\mathrm{C}}_3{\mathrm{H}}_3\mathrm{N}+\mathrm{H}\mathrm{C}\mathrm{N}+{\mathrm{C}}_2{\mathrm{H}}_4$$

$$\mathrm{N}\mathrm{C}-{\mathrm{C}}_4{\mathrm{H}}_8-\mathrm{C}\mathrm{N}\stackrel{\mathrm{h}\mathrm{i}\mathrm{g}\mathrm{h} \mathrm{t}\mathrm{e}\mathrm{m}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{e},{\mathrm{L}\mathrm{i}}^{+},{\mathrm{C}\mathrm{o}}^{2/3+},{\mathrm{M}\mathrm{n}}^{4+}}{\to } {\mathrm{C}}_3{\mathrm{H}}_5\mathrm{N}+\mathrm{H}\mathrm{C}\mathrm{N}+{\mathrm{C}}_2{\mathrm{H}}_2$$

(4) 

The structure of the cathode material itself suppresses the release of lattice oxygen at high temperatures.

The specific analysis is as follows: Results from the multi-metallic catalytic systems indicate that NiO, Co₃O₄, and MnO₂ fuse to form composite metal oxides (M_xN_yO_z), which intensifies the release of lattice oxygen, thereby causing the deep oxidation of target hydrocarbon products. This phenomenon is attributed to the relatively weak bonding between metal cations and O²⁻ within these simple metal oxide lattices. In contrast to NiO, Co₃O₄, and MnO₂, the NCM cathode material possesses a more stable lattice arrangement governed by a bonding network comprising metal-oxygen ionic bonds and O–O covalent bonds. This robust intrinsic structure is highly conducive to maintaining lattice integrity at high temperatures, thereby suppressing the detachment of lattice oxygen.

In summary, building upon the analyses in our previous study, the mechanism by which cathode metals synergize with CaO-ZSM-5* for the deep detoxification and directional conversion of spent electrolytes into light hydrocarbons during vacuum catalytic pyrolysis is detailed as follows. In the initial stage of the reaction, LiPF₆ undergoes thermal decomposition and hydrolysis to generate HF and phosphate compounds. Most of these products react with the CaO supported on the ZSM-5 molecular sieve and are immobilized as CaF₂ and Ca₂P₂O₇, respectively. Meanwhile, the remaining portion corrodes the cathode material, releasing a small amount of free Li⁺ (Figure 5a), which synergizes with the hydrolysis intermediate POF₃ to dominate the decomposition of linear carbonates in the low-temperature range. In the middle stage of the reaction, as the temperature increases, a large amount of Li⁺ in the cathode material lattice overcomes electrostatic attraction and migrates to the crystal surface (Figure 5b). The migrated Li⁺ attacks C–O bonds within linear and cyclic carbonates, inducing the random cleavage of these linkages, with PS decomposing to generate SO₂. As heating progresses, prior to reaching the critical temperature for Li⁺-promoted ADN cleavage, Ni²⁺ also migrates to the surface, co-promoting the breaking of carbonate C–O bonds while simultaneously suppressing the catalytic activity of Li⁺ toward ADN. Simultaneously, the C–F bonds in FEC undergo cleavage under the combined action of Li⁺, Ni²⁺, and high temperature, yielding HF, which subsequently reacts with CaO, immobilizing fluorine within the pyrolysis residue. While Li⁺ and Ni²⁺ cleave the carbonates, nickel-containing compounds and the hydrolysis-generated LiOH work with CaO-ZSM-5* to promote the decarboxylation and oxygen fixation of carboxyl-containing products, as well as the targeted transformation of hydrocarbon and alkoxy radicals into light fractions. At this point, the entire vacuum catalytic pyrolysis process, synergistically driven by cathode metals and CaO-ZSM-5* for the directional conversion of the spent electrolyte into light hydrocarbons, is complete. During this selective upgrading stage, the majority of Ni²⁺ in the pyrolysis system is reduced to metallic Ni by pyrolytic carbon and hydrocarbons. In the later stage of the reaction, upon further heating, Mn⁴⁺ also overcomes electrostatic attraction, migrates to the crystal surface, and engages in the catalytic system. Similar to Ni²⁺, most of the Mn⁴⁺ is reduced to MnO by the same reducing agents. Concurrently, the SO₂ in the pyrolysis gas undergoes a series of reactions with pyrolytic carbon, hydrocarbons, reactive oxygen species derived from the decomposition of the NCM material, metallic Ni, and MnO. These reactions transform SO₂ into elemental S, NiSO₄, NiS, MnSO₃, and MnS₂, thereby supporting the effective immobilization of sulfur within the pyrolysis residue. By this stage, the catalytic detoxification of the spent electrolyte, driven by cathode metals in synergy with the CaO-ZSM-5* molecular sieve, is accomplished. Notably, the chemical state evolution of Co^2/3+ is solely attributed to its redox reactions with species such as pyrolytic carbon, and this ion does not participate in the synergistic vacuum catalytic pyrolysis process with CaO-ZSM-5* for the detoxification and selective upgrading of the electrolyte.

4. Conclusions

This study systematically investigates the element-specific, synergistic, and antagonistic effects of cathode metals (Li, Ni, Co, Mn) on the deep detoxification and directional conversion of spent electrolytes into light hydrocarbons over a CaO-ZSM-5* catalyst. Based on batch experiments of vacuum catalytic pyrolysis, a comprehensive migration–reaction coupling mechanism is elucidated. The main conclusions are drawn as follows:

Element-specific catalytic characteristics: At high temperatures, Li⁺, Ni²⁺, Mn⁴⁺, and Co^2/3+ promote the cleavage of C–O bonds in carbonates and synergize with CaO-ZSM-5* to achieve the decarboxylation and oxygen fixation of carboxyl-containing products, thereby promoting the directional conversion of carbonates into light hydrocarbons. However, their catalytic activities decrease sequentially in the aforementioned order. Regarding toxic organics, Li+ significantly minimizes the final emissions of aromatics into the pyrolysis gas and reduces nitriles in the pyrolysis gas by promoting the coupling of ADN pyrolysis products in the pyrolysis oil and residues. The minimized emission of nitriles into the pyrolysis gas by Ni²⁺ is primarily attributed to its low catalytic activity for ADN cleavage, whereas Co^2/3+ and Mn⁴⁺ significantly promote nitrile release into the pyrolysis gas. For S and F detoxification, Li⁺ and Co^2/3+ immobilize sulfur in the pyrolysis residues, while Ni²⁺ and Mn⁴⁺ mitigate final sulfur emissions into the pyrolysis gas by coupling PS fragments in the non-gaseous phases. Ultimately, all these metals showed favorable fluorine detoxification performance, with no fluorides detected in the pyrolysis gas and no fluorinated compounds detected in the pyrolysis oil. Additionally, CaO-ZSM-5* alone can effectively immobilize P as Ca₂P₂O₇. Notably, compared to Li⁺ and Ni²⁺, the lattice oxygen in Co₃O₄ and MnO₂ is more prone to be released under high-temperature vacuum conditions.

Synergistic and antagonistic effects in multi-metallic systems: The formation of composite metal oxides (M_xN_yO_z) in multi-metallic systems increases the energy barrier for releasing active catalytic ions, thereby hindering the metal-promoted C–O bond cleavage and the synergistic decarboxylation/oxygen fixation. Furthermore, the high-temperature formation of M_xN_yO_z intensifies the release of lattice oxygen from NiO, Co₃O₄, and MnO₂, causing deep oxidation of hydrocarbon target products. While multi-metallic systems strongly inhibit aromatics in the pyrolysis oil, notable antagonism occurs regarding nitriles: Ni²⁺ significantly inhibits the ADN cleavage activities of Li⁺, Mn⁴⁺, and Co^2/3+, with the strongest antagonism observed against Li⁺ (suppressing gas-phase nitriles to <2% in the E-LN group while leaving abundant unreacted ADN). However, all multi-metallic combinations showed a favorable synergistic trend in heteroatom immobilization, with no sulfides or fluorides detected in the final gas products and no S-containing or fluorinated compounds detected in the final oil products under the present GC–MS analysis.

Mechanism of migration-reaction coupling in cathode materials: During the early stage of reaction, LiPF₆ decomposes, with F and P immobilized in the residues. The hydrolysis intermediate POF₃, synergizing with trace Li⁺ released from corrosion, dominates linear carbonate cleavage at low temperatures. As the temperature rises, guided by the structural constraints of the cathode lattice, a sequential activation of metals occurs. A large amount of Li⁺ first migrates to the lattice surface, initiating the cleavage of carbonate C–O bonds and attacking PS to produce SO₂. Prior to reaching the critical temperature for Li⁺-promoted ADN cleavage, Ni²⁺ also migrates to the surface, co-promoting carbonate cleavage while strongly antagonizing Li⁺’s activity toward ADN. Simultaneously, C–F bonds in FEC are cleaved, immobilizing F in the residue. The Li and Ni species then synergize with CaO-ZSM-5* to catalyze decarboxylation, oxygen fixation, and the targeted conversion of radical fragments into light hydrocarbons, accomplishing the selective upgrading process. In the later stages, Mn⁴⁺ engages in the reactions. SO₂ undergoes complex redox reactions with pyrolytic carbon, hydrocarbons, metallic Ni, MnO, and reactive oxygen species, thereby promoting the mineralization of S. Notably, Co^2/3+ does not participate in the synergistic detoxification and selective upgrading process.