Passivation and corrosion of Al current collectors in lithium-ion batteries

Lithium-ion batteries (LIBs) used in portable electronics and electric vehicles must maintain the physico-chemical stability of internal components, including separators, current collectors, electrolytes, and electrodes. High-energy-density applications push toward higher cutoff voltages and stable structures, making internal component stability critical to avoid degradation such as current collector corrosion and electrode material spoilage. Al and Cu are widely used current collectors for cathodes and anodes due to cost, conductivity, and inertness. Al’s apparent inertness arises from a native oxide bilayer (outer crystalline hydrous aluminite Al2O3·H2O with pores and inner dense amorphous Al2O3 barrier). Despite this, Al is attacked by ester-based electrolytes distinct from aqueous corrosion. Corrosive species (e.g., HF from LiPF6 hydrolysis, TFSI/FSI salts) and parasitic interfacial reactions trigger chemical/electrochemical corrosion of current collectors and electrode dissolution. The corrosion/failure of Al current collectors at multiple charge voltages, particularly >4.5 V, remains insufficiently evaluated in relation to battery performance. This study asks: how do passivation, passivity breakdown, and corrosion of Al in commercial LIB electrolytes evolve with potential, and how do they impact electrode integrity and cell performance? We address this by comparing reclaimed Al (RA) after full battery service versus fresh Al (FA), correlating interfacial phenomena with electrochemical performance, and using DFT to elucidate adsorption/oxidation mechanisms.

Prior work indicates that Al passivation in LiPF6-based electrolytes arises mainly from electrochemical reactions rather than solely HF attack; uncoordinated solvent oxidation can generate radical cations initiating corrosion (Yoon et al.). Reused Al current collectors from spent LIBs can degrade high-rate performance of cathodes (e.g., LiNi a& Mn0.2Co0.2O2 on reused Al in LiPF6 EC:DMC+VC). Reviews highlight that Al corrosion depends strongly on electrolyte solvent/salt composition (e.g., imide salts like LiTFSI/LiFSI can aggravate Al dissolution), and strategies such as coatings (graphene, MXene) can mitigate current collector corrosion. However, systematic coupling of electrolyte anions/solvents with Al passivation/corrosion across practical charge voltages and direct linkage to battery performance degradation in commercial-like systems is limited. This work fills that gap by quantifying passivation thickness evolution, voltage-dependent passivity breakdown (notably at 4.4–4.8 V), localized F−/Al3+ transport, and correlating these with Coulombic efficiency loss and capacity fade in LiCoO2||Li cells.

Materials and sample preparation: Fresh Al (FA) foil, LiCoO2, high-purity Li foil, and solvents were obtained from Guangdong Canrd New Energy Technology Co., Ltd. Electrolyte: 1 M LiPF6 in EC/DEC/FEC (3:6:1 v/v/v), prepared in an Ar-filled glovebox (H2O < 0.1 ppm, O2 < 0.1 ppm). Solvent H2O < 20 ppm; LiPF6 water < 50 ppm; free HF < 100 ppm. Reclaimed Al (RA) foils were harvested from waste LiCoO2 batteries (Huizhou Desay Battery Co., Ltd., 1852001-AL, limited charge voltage 4.3 V) after full service and discharge to <50% of rated capacity. RA and FA foils were rinsed with DEC then NMP to remove contaminants in Ar atmosphere. Electrode/cell fabrication: Cathode slurry: LiCoO2:PVDF:acetylene black = 8:1:1 (mass), dispersed in NMP, coated on FA or RA (loading ~2.3 mg cm−2), dried at 80 °C for 12 h, and punched for 2032 coin cells. Electrolyte dosage: 30 μL. For Al||Li half-cells, Al foil disks (1.13 cm2) were assembled against Li. Electrochemical testing: LiCoO2||Li rate/cycling at 0.5, 1, and 5 C (1 C = 200 mAh g−1) at ambient temperature. Al||Li half-cells: CV from 2–6 V at 10 mV s−1 at room temperature to probe anodic processes. EIS of FA/RA||Li in 1 M LiPF6 EC/DEC/FEC: frequency 0.1 Hz–105 Hz, amplitude 5 mV, measured at OCP and 4.6 V before/after 3rd CV; equivalent circuits fitted with ZView 3.1. Three-electrode configuration was implemented using CR2032-type coin half-cells with Li as both reference and counter electrodes; all assembly in Ar glovebox. Chronoamperometry and voltage holds: FA||Li cells were first cycled 3 times between 2–6 V to stabilize passivation, then held for 24 h at each potential: 4.0, 4.2, 4.3, 4.4, 4.5, 4.6, 4.8, 5.0 V. Characterization: After electrochemical tests, RA/FA samples were disassembled, washed with DEC, and dried in Ar. SEM (ZEISS Gemini 300) imaged surface morphologies of current-collector sides facing Li. XRD characterized crystallographic orientations (dominant Al (220), (311)). XPS (Thermo K-Alpha, Al Kα 1486.6 eV; pass energy 50 eV narrow scans; C 1s at 284.8 eV used for calibration) with Ar+ etching probed depth profiles (0, 10, 20 nm); Li anode surfaces also analyzed for Al deposition. AFM with KPFM (Bruker Dimension Icon, tapping) measured topography and surface potential (scan 10 × 10 × 10 μm; vertical resolution 0.04 nm; lateral 0.15 nm); RMS roughness was quantified. DFT calculations: Binding energies of HF and EC molecules on Al surfaces were computed using PBE-GGA. Models: 5×5 supercells of Al(111), (220), (311) slabs with 5 layers (bottom two fixed), 15 Å vacuum. Plane-wave cutoff 520 eV; Monkhorst-Pack grid 4×4×4 for Al lattice optimization; structures relaxed until total energy < 1e−4 eV and forces < 1e−2 eV Å−1. Binding energy defined as Ead = EM/Al − EAl − EM, where M = HF or EC. Data analysis: Corrosion/passivation states inferred from CV onset, chronoamperometric current saturation/oscillation, EIS-derived resistances (Rct, diffusion-related elements), AFM/KPFM roughness and surface potentials, XPS chemical states (Al 2p, O 1s, F 1s, Li 1s) and depth distributions, and SEM evidence of pits/cracks/nanograins.

- Performance impact: In LiCoO2||Li half-cells, using reclaimed Al (RA) as current collector led to capacity loss >20% versus fresh Al (FA) across rates. From rate plots, RA exhibited relative capacity decreases of approximately 24.28% (0.1 C), 21.15% (0.5 C), 23.52% (1 C), 25.83% (2 C), 32.88% (5 C), and 45.60% (10 C) compared with FA. Over long cycling (200–800 cycles at 1 C), distinct capacity fade behaviors were observed linked to Al structure differences (plane FA vs corroded-porous RA). - Coulombic efficiency: CE of RA-based cells was 98.70%, lower than FA’s 99.84%, i.e., 1.14% absolute CE loss, consistent with ongoing irreversible anodic currents from passivity breakdown/localized corrosion. - CV onset and anodic behavior: For Al||Li cells, initial anodic current for FA started at 3.76 V vs Li/Li+, while RA began at 3.89 V with higher anodic current, reflecting differences in native/passivated surface layers. FA anodic dissolution diminished upon further cycling due to concurrent passivation; RA showed dramatic anodic current increase by the 5th cycle, indicating localized corrosion. - Chronoamperometry and voltage dependence: At 4.0–4.3 V, current densities saturated smoothly, indicating stable passivation; at 4.4–5.0 V, currents fluctuated before saturating to ~10–20 μA cm−2, evidencing interfacial inhomogeneity and electrolyte oxidation coupling. Extended 4.0 V holds (100 h) increased surface roughness and passivation thickness, while higher potentials (>4.2 V) triggered cracking/peeling. - Passivation layer evolution: During cycling, a passivation layer >20 nm formed near median voltage. With increasing voltage, newly formed Al–F–O particles induced stress, squeezing/deforming the passivation layer, causing stress corrosion cracks, peeling, and re-generation—leading to continuous Al consumption. Depth-profile XPS showed AlF3/Al–O–F features near the surface initially; at 4.4 and 4.8 V, AlF3 nearly disappeared at ~20 nm depth (area ratio ~0.01), implying faster outward Al3+ diffusion than inward F− diffusion and thinning/stripping of fluoride films. - Transport directions versus voltage: Inward diffusion of F− dominated up to ~4.4 V; above 4.4 V, outward diffusion of Al3+ dominated, accelerating passivity breakdown and localized corrosion. Al 2p signals were detected on Li anodes (notably after 4.8 V treatment), indicating dissolved Al species migrating and reductively depositing, disrupting the SEI and Li+ transport, increasing impedance, and causing additional irreversible Li loss. - Surface morphology and electrical properties: AFM/KPFM revealed RMS roughness increased from 38.11 nm (after 3rd cycle) to ~70.52 nm (4 V, 100 h), with enlarged surface potentials indicating more insulating passivation. SEM/AFM observed nanograins (AlF3/Al2O3), deposits, holes, and cracks at higher voltages (4.2–5.0 V), confirming incompatibility and tensile stress accumulation in the film. - EIS: RA exhibited lower corrosion resistance than FA after cycling (lower charge-transfer and diffusion resistances), consistent with greater surface area and more heterogeneous interphases. Rct evolution at 4.6 V after 3rd cycle indicated reconstructed passivation that transiently protected the Al substrate. - Mechanistic insights and stress/volume effects: Conversion of Al2O3 to AlF3 (via HF and F−) forms Al3+ vacancies, accelerates ion diffusion in crevices, and causes microstructural volume expansion of ~64%, leading to cracking and delamination. Active material loading can create localized electrolyte confinement (wormhole contact) and corrosive media concentration, exacerbating localized corrosion. - DFT insights: EC molecules adsorb more strongly on Al surfaces, especially Al(220), than DEC; EC also coordinates HF more strongly than DEC. Strong EC adsorption at high potentials promotes corrosion bridges/couples, facilitates HF ingress into crevices, accelerates EC oxidation and electrolyte composition changes, and correlates with passivity breakdown at 4.4–4.8 V.

The study connects voltage-dependent interfacial chemistry of Al current collectors to measurable battery performance degradation. At moderate potentials, Al forms a relatively protective passivation layer (>20 nm), but as charging voltage approaches and exceeds 4.4 V, EC-rich electrolytes promote localized adsorption and oxidation, inward F− penetration, and, subsequently, outward Al3+ dissolution that thins and fractures the passivation film. This leads to stress corrosion cracking, peeling, and re-passivation cycles that continuously consume the Al substrate. The heterogeneity of these processes is evident from current fluctuations during voltage holds, increased roughness, and spatially varying surface potentials. Dissolved Al species can migrate and deposit on Li anodes, disturbing SEI and Li+ transport, further lowering CE and increasing impedance. These interfacial failures directly translate to reduced Coulombic efficiency (by ~1.14%) and substantial capacity losses (>20%) across rates in LiCoO2||Li half-cells using reclaimed Al. The EIS trends corroborate increased interfacial reactivity and reduced corrosion resistance of RA. DFT results implicate EC adsorption and HF coordination as key drivers for passivity breakdown at high voltage, highlighting electrolyte-solvent selection’s role. Overall, the findings underscore that high cutoff voltages (>4.4 V) in LIBs critically challenge Al current collector stability; mitigating strategies must address both interphase chemistry and mechanical integrity under electrochemical stress.

This work establishes a detailed mechanistic picture of Al current collector passivation and corrosion in LiPF6-based EC/DEC/FEC electrolytes and quantifies their impacts on LIB performance. Key contributions include: (1) identification of a voltage threshold (~4.4 V) where passivity breakdown and localized corrosion emerge; (2) demonstration that repeated crack–peel–repassivation cycles consume Al and degrade electrode integrity; (3) evidence that dissolved Al can deposit on Li anodes, harming SEI and kinetics; and (4) correlation of these processes with measurable losses in CE (~1.14%) and capacity (>20%). DFT reveals strong EC adsorption (especially on Al(220)) and HF coordination that facilitate corrosion at high voltage. These insights advocate engineering robust, chemically/mechanically resilient passivation layers on Al (e.g., protective coatings, fluoride/oxide-stabilizing interphases) and rational electrolyte/solvent design with weaker adsorption/coordination tendencies to suppress passivity breakdown at high voltages. Future research should explore alternative electrolytes and additives that limit EC-driven adsorption/corrosion, advanced coatings compatible with high-voltage operation, and operando diagnostics to capture localized breakdown dynamics under realistic cell architectures.

- Baseline differences between reclaimed Al (RA) and fresh Al (FA) foils (surface roughness, prior damage) complicate direct comparisons; the authors mitigate this by using pristine FA to study mechanisms but residual differences may influence results. - Electrochemical testing employed coin-type three-electrode configurations and a single electrolyte formulation (1 M LiPF6 in EC/DEC/FEC 3:6:1); findings may vary with other salts/solvents/additives and full-cell environments with different cathodes. - Depth profiling by XPS is limited to ~20 nm etching steps; ultrathin interphase features or deeper gradients may be insufficiently resolved. - Migration and deposition of dissolved Al on Li anodes were inferred from ex situ XPS; quantification of Al deposition amounts and their spatial distribution across separators/electrodes was limited. - Mechanical stress evolution in passivation layers (crack initiation/propagation) was deduced from morphology; in situ stress/strain measurements were not performed.