Diagnosing Transition-Metal Crossover

An ICP-OES/XPS Root-Cause Workflow

Investigation of Catastrophic Battery Failures walks through one forensic case study of field failures in e-bikes and e-scooters. This page covers a narrower question that sits underneath a lot of those investigations: when a cell has aged badly or failed and cathode transition-metal dissolution is on the suspect list, how do you actually confirm it happened, and how far did it get? The workflow below pairs ICP-OES (or ICP-MS) elemental analysis with XPS surface analysis on the same failed cell to answer both questions.

Why Transition-Metal Crossover Matters

NMC and NCA cathodes hold their manganese, nickel and cobalt in an oxide lattice, but that lattice isn’t perfectly stable. Trace HF in the electrolyte — itself a hydrolysis product of the LiPF₆ salt reacting with any moisture that made it into the cell — slowly etches the cathode surface and pulls transition-metal ions into solution [1]. Those ions don’t stay in the electrolyte. They migrate to the anode and plate out there, where they do two damaging things: they catalyse further electrolyte breakdown at the graphite surface, and they get incorporated into the SEI in a way that consumes cyclable lithium and leaves the SEI thicker and less protective than it should be [1]. This is one of the degradation pathways summarised on our Degradation Mechanisms page — this piece is about how you’d actually go and prove it happened on a specific returned or failed cell, rather than infer it from a capacity-fade curve alone.

Disassembly and Sample Preparation

The cell is opened in an inert-atmosphere glovebox to stop ambient moisture reacting with whatever electrolyte is left on the electrode surfaces, which would contaminate exactly the chemistry you’re trying to measure. Anode and cathode are separated, and each is given a short soak in dimethyl carbonate to rinse off residual electrolyte salts and solvent without disturbing the SEI or any deposited surface species [2]. From there the anode sample gets split in two: part of it goes forward for acid digestion ahead of ICP analysis, and part is kept intact and dried under vacuum for XPS, since XPS needs an undamaged surface rather than a dissolved one.

ICP-OES/ICP-MS: How Much Crossed Over

ICP-OES (or the more sensitive ICP-MS) answers the “how much” question. A representative portion of the anode — a punched disc or a scraped-off sample of the electrode coating — is fully digested in a mix of concentrated acids (a nitric/hydrochloric/hydrofluoric combination is typical) so that every trace of manganese, nickel and cobalt ends up in solution as free ions rather than locked in a solid deposit [2]. The resulting solution is diluted and run through the ICP, which reports each element in parts per million against the digested mass. The number only means something next to a baseline: an anode pulled from a fresh, never-cycled cell of the same design, put through the identical digestion and measurement. Transition-metal levels meaningfully above that baseline are the first hard evidence of crossover, and the ratio between manganese, nickel and cobalt in the deposit can also point back toward which mechanism dominated, since not all three dissolve at the same rate under the same conditions [3]. Where the deposit needs to be mapped rather than just totalled — checking whether crossover is uniform across the electrode or concentrated near an edge or a defect — laser-ablation ICP-MS can image the distribution directly on the intact electrode instead of averaging it away in a bulk digestion [4].

XPS: Confirming Identity and Oxidation State

ICP tells you a quantity but not a chemical story: it can’t distinguish metal that’s genuinely bound into the SEI from metal that arrived as loose surface contamination during handling, and it says nothing about what oxidation state the metal is in once it gets to the anode. XPS on the intact (non-digested) anode sample fills that gap. A narrow-scan around the Mn 2p or Ni 2p binding energy region resolves the spin-orbit doublet for each element, and the exact peak position within that doublet is diagnostic of oxidation state — manganese’s Mn 2p3/2 peak shifts across roughly 640–643 eV moving from Mn(II) through Mn(III) to Mn(IV), for example, while nickel’s Ni 2p3/2 doublet separates into components consistent with Ni²⁺ and Ni³⁺ [5]. Depth-profiling the same spot with a mild sputter etch (or comparing survey scans before and after a light etch) shows whether the metal sits at the very outer surface, consistent with recent or ongoing deposition, or is buried within the SEI, consistent with metal that has been there for a while and been built over by subsequent SEI growth.

Reading the Combined Result

Neither technique on its own makes the case. ICP-OES/MS establishes that a meaningful mass of transition metal left the cathode and reached the anode, above what a clean baseline cell shows. XPS establishes what chemical form that metal is in and roughly how it’s distributed through the SEI, which is what separates “this is disrupting the SEI and consuming lithium” from “this is a light dusting of surface contamination that doesn’t explain the fade.” Put together with the cell’s own capacity-fade and impedance-rise history — see our pages on Cell Cycle Lifetime and internal resistance — that combination is what turns “transition-metal crossover” from a plausible suspect into a documented root cause.

Practical note: always digest and measure a matched baseline sample alongside the suspect cell, prepared and handled identically. Without it, a raw ppm figure on its own is hard to interpret — ambient handling and even glovebox atmosphere quality can introduce a small amount of background metal contamination, and it’s the delta above a proper baseline that carries the diagnostic weight, not the absolute number.

References

Waldmann, T. et al., An Unpredictable Hazard in Lithium-ion Batteries from Transition Metal Ions: Dissolution from Cathodes, Deposition on Anodes and Elimination Strategies, Journal of The Electrochemical Society.
Post-Mortem Analysis of Aged Lithium-Ion Batteries: Disassembly Methodology and Physico-Chemical Analysis Techniques, Journal of The Electrochemical Society.
Battery Analysis Using ICP-OES and ICP-MS, Separation Science.
Analysis of Deposition Patterns and Influencing Factors of Lithium and Transition Metals Deposited on Lithium Ion Battery Graphitic Anodes by LA-ICP-MS, Spectroscopy Online.
Sim, R. et al., Delineating the Impact of Transition-Metal Crossover on Solid-Electrolyte Interphase Formation with Ion Mass Spectrometry, Advanced Materials, 2024.