Machine-learning-potential simulations show that local Li content sets Mn oxidation states at LiMn2O4 surfaces and that the O-H stretch band exposes a dual-site acid-attack vulnerability

Local Li content determines the oxidation states of surface Mn at LMO, and these surface Mn oxidation states play a decisive role in interfacial acid-base chemistry. The mixed-valence electronic structure of LMO was already known to be governed by Li content, but the atomistic link between lithiation state and interfacial degradation remained elusive; this work connects lithiation state, surface Mn oxidation states, and interfacial acid-base behavior. Atomistic simulations using an ab initio-quality machine learning potential across aqueous LMO interfaces at varying lithiation states; the abstract does not report system sizes, simulation durations, or statistics.

The O-H stretching band of interfacial species acts as a sensitive spectroscopic probe of the surface electronic structure, with distinct shifts depending on Mn oxidation state. It elevates computational vibrational spectroscopy from a characterization tool to a probe that reads out surface electronic structure, building a correspondence between experimental spectra and interfacial electronic structure. Based on computational analysis of vibrational features of interfacial species, reporting oxidation-state-dependent shifts; the abstract gives no specific wavenumbers or band ranges.

The mixed-valence spinel surface has an intrinsic vulnerability: Lewis-acidic centers coexist with neighboring oxygen sites that are susceptible to electrophilic attack. It extends the conventional acid-centric explanation of Mn dissolution toward a dual-site mechanism driven by LMO's mixed-valent nature. Inferred from simulated surface structures and vibrational features, an argument at the atomistic mechanism level; the abstract provides no direct experimental validation data.

This work is aimed at researchers studying the stability and degradation mechanisms of aqueous LMO solid-liquid interfaces, as well as materials and electrochemistry communities interested in Li-ion battery cathodes and water-splitting electrocatalysts. Its conclusions apply to the setting of aqueous LMO interfaces at varying lithiation states simulated with an ab initio-quality machine learning potential, used to explain how local Li content affects interfacial acid-base chemistry through surface Mn oxidation states, and to suggest the O-H stretching band as a spectroscopic probe of surface electronic structure. For readers seeking an atomistic understanding of mixed-valence oxide interfacial reactions or an entry point for spectroscopic characterization such as in situ vibrational spectroscopy, this framework offers actionable ideas.

The loaded text is incomplete, containing only the abstract and references, so it is not possible to verify the specific simulation system sizes, the lithiation states sampled, how the machine learning potential was trained and validated, the details of the vibrational spectra calculations, or whether experimental comparisons are included. The specific wavenumbers of the O-H stretch shifts, the quantitative mapping between oxidation state and shift, and the energy barriers of each step in the dual-site mechanism do not appear in the text; these are open questions for readers of the full article. In addition, the dual-site mechanism is currently proposed from computational inference, and its applicability and observability under real aqueous interfacial conditions remain to be tested by follow-up work.