Why conventional TEM fails — and how cryo-EM solves it

Conventional electron microscopy cannot reliably image the solid electrolyte interphase because SEI components — lithium metal, lithium carbonate, lithium fluoride, and related organics — decompose under standard electron beam conditions before a usable image can be acquired. The result is not merely blurred data but systematic mischaracterisation: phases that appear stable under conventional TEM may be beam-damage artifacts, while genuinely present nanocrystallites are destroyed before detection. Cryo-EM addresses this by combining cryogenic sample preservation with ultralow electron-dose imaging, allowing native structural information to be captured without introducing the artifacts that have historically distorted SEI models.

As documented in the comprehensive workflow review from the Institute of Physics, Chinese Academy of Sciences (2021), the cryo-EM protocol for battery materials encompasses four critical steps: sample preparation under inert atmosphere, cryogenic transfer to prevent air or moisture exposure, low-dose imaging protocols, and rigorous data interpretation strategies. Each step is necessary; a failure at any stage — for example, air exposure during transfer — irreversibly alters the SEI chemistry before the electron beam is even engaged. According to Nature-published reviews of battery characterisation, the field has long recognised that ex situ techniques introduce surface contamination and phase transformation artifacts, making in-place preservation techniques such as cryo-EM indispensable for ground-truth structural data.

The US Army Research Laboratory’s 2021 cryo-TEM study made this concrete: atomic imaging of SEI phase components and their spatial arrangement was achievable only using ultralow-dosage cryogenic transmission electron microscopy. That same study also uncovered that a significant fraction of deposited lithium metal possesses amorphous atomic structure — a finding attributable to carbon and oxygen impurities — which would have been obscured entirely under conventional imaging conditions. The implication is that prior literature built on conventional TEM may systematically underreport amorphous lithium fractions and overstate the crystallinity of SEI phases.

What cryo-EM has revealed about SEI phase composition and stability

Cryo-EM has produced a series of findings about SEI chemical phases and stability that contradict assumptions built on conventional characterisation. The most consequential is the thermodynamic instability of lithium carbonate: the US Army Research Laboratory’s 2021 cryo-TEM study showed that crystalline Li₂CO₃ is not stable when in direct contact with lithium metal, continuously reacting with the electrolyte to produce gas and forming a dynamically evolving, porous SEI. This overturned a long-standing assumption in battery research and has direct consequences for electrolyte additive design, since strategies aimed at promoting Li₂CO₃ formation as a protective phase are working against the thermodynamics of the system.

The UC Berkeley study from 2019 — the foundational cryo-TEM investigation of carbonaceous anodes — tracked SEI evolution on carbon black negative electrodes during cycling at atomic resolution. The authors found that a thin, primarily amorphous SEI nucleates during the first cycle and subsequently evolves into one of two distinct morphologies: a compact SEI rich in inorganic components that effectively passivates the electrode surface, or a porous morphology associated with accelerated degradation. This morphological bifurcation, resolved only under cryo conditions, provides direct mechanistic grounding for why nominally identical cells exhibit divergent cycle life — a question that had remained unanswered by bulk electrochemical measurements alone.

“The SEI on carbon black evolves into either a compact, inorganic-rich passivating layer or an alternative porous morphology associated with accelerated degradation — a bifurcation visible only at atomic resolution under cryo conditions.”

The Communications Chemistry commentary (2021) corroborated these instability findings, explicitly noting that low-dosage cryo-TEM visualises how Li₂CO₃ decomposes and how electrolyte additives can stabilise the interface against this degradation pathway. This positions cryo-EM not merely as a characterisation tool but as a direct screening platform for electrolyte engineering: additive effects on SEI chemistry that would be invisible to bulk techniques become directly observable at nanometre resolution. Standards bodies including IEC are increasingly recognising atomic-scale interface characterisation as relevant to battery safety certification, underscoring the practical importance of these findings.

Three-dimensional SEI architecture and its link to Coulombic efficiency

The three-dimensional geometry of the SEI — not merely its chemical composition — determines whether lithium can be efficiently replated in subsequent cycles. Two landmark cryo-EM studies from the Southern University of Science and Technology (2021) established this quantitatively. The first, using low-dose cryo-EM tomography with expectation-maximisation reconstruction, confirmed that lithium metal deposits are spherical and that the SEI forms a conformal shell around each deposit — a structural feature that is inaccessible by any two-dimensional imaging approach and was previously inferred only indirectly from electrochemical data.

The companion study from the same institution resolved the composition of this conformal SEI skin layer: it consists largely of nanocrystalline LiF and Li₂O embedded in an amorphous polymeric matrix. When lithium is fully stripped, this framework buckles and wrinkles rather than maintaining its geometry. The mechanical flexibility and resilience of the SEI skin layer is therefore the property that determines whether the SEI network remains intact and can host newly plated lithium in subsequent cycles — a conclusion that shifts the design target from chemical purity toward mechanical durability. Corroborating this, cells subjected to controlled uniaxial pressure showed improved Coulombic efficiency in the same study, directly linking the cryo-EM-revealed 3D SEI structure to a macroscopic electrochemical outcome.

These findings are consistent with the broader framework established by OECD energy storage roadmaps, which identify mechanical stability of electrode-electrolyte interfaces as a primary barrier to achieving the cycle life targets required for grid-scale and electric vehicle applications. The cryo-EM data from Southern University of Science and Technology provides the first direct atomic-scale evidence for why interface mechanics matter as much as chemistry in this context.

Cryo-EM beyond lithium-ion: sodium batteries and cathode interphases

Cryo-EM’s utility extends beyond lithium-ion systems to sodium-metal batteries and cathode-side interphases, broadening its relevance across the full landscape of next-generation cell chemistries. Peking University Shenzhen’s 2021 cryo-TEM study of sodium-metal electrodes showed that without the fluoroethylene carbonate (FEC) additive, a NaPF₆-containing carbonate-based electrolyte produced an unstable SEI rich in Na₂CO₃ and Na₃PO₄ that continuously consumed the sodium reservoir. The FEC additive modified the SEI chemistry in ways detectable only at nanometre resolution, establishing cryo-TEM as a platform for screening additive effects on SEI chemistry across alkali metal systems — not just lithium.

On the cathode side, HKUST’s 2021 cryo-EM study of a flower-shaped Na₃V₂(PO₄)₃/C cathode in ether-based electrolyte for sodium-ion batteries resolved a cathode electrolyte interphase (CEI) layer only 1.1 nm thick. This sub-nanometre thickness placed the CEI below the detection threshold of most surface analytical techniques, yet cryo-EM resolved its structure and chemistry. The thin, well-defined CEI correlated with a capacity retention exceeding 88% after 1,600 cycles — a direct demonstration that cryo-EM can connect nanometre-scale interface structure to long-term macroscopic performance. As WIPO patent trend data confirms, sodium-ion battery filings have accelerated substantially since 2020, making the extension of cryo-EM characterisation to these systems timely and strategically significant.

Integrating cryo-EM with operando and surface characterisation methods

Cryo-EM delivers unmatched spatial resolution of quenched SEI structures, but its limitations — principally the inability to image under active electrochemical conditions and a restricted field of view — motivate its integration with complementary techniques rather than its use in isolation. The consensus across the dataset is that no single technique provides a complete picture of SEI formation, evolution, and degradation, and that cryo-EM occupies a unique and irreplaceable niche for artifact-free atomic-resolution imaging within a multimodal characterisation strategy.

UCLA’s operando STEM-EELS study (2023) demonstrated the most direct complement to cryo-EM: real-time chemical fingerprinting of SEI constituents including Li and LiH dendrites during charge-discharge cycles using ultrathin liquid cells. This approach provides time-resolved chemical mapping during actual battery operation — the dynamic information that cryo-EM’s structural snapshots cannot supply independently. In situ liquid cell TEM, reviewed at the University of Sheffield (2018), offers similar advantages but faces challenges from beam-induced electrolyte damage and the difficulty of replicating realistic cell environments. Cornell University’s 2014 demonstration of nanoscale lithium ion distribution imaging using valence EELS in a liquid flow cell — enabling determination of lithiation states in individual LiFePO₄ particles — established the technical feasibility of this approach over a decade ago, and the field has since advanced substantially in dose management and cell design.

Surface-sensitive and mechanical characterisations of the SEI provide complementary data at accessible length scales. In situ AFM studies of SEI nucleation and growth on carbonaceous electrodes were reported by the Skolkovo Institute of Science and Technology (2020), while Fujian Normal University (2023) linked electrochemical AFM with 3D nano-rheology to probe in situ SEI formation from the electrical double layer through to full nanostructured SEI development. Multimodal approaches using in situ TEM, SIMS, XPS, and atom probe tomography were reviewed by Pacific Northwest National Laboratory (2016), establishing that cryo-EM, operando EELS, AFM, and synchrotron techniques are complementary rather than competitive. The NIST reference framework for battery interface characterisation similarly emphasises that atomic-resolution structural data from cryo-EM must be integrated with chemical mapping and dynamic measurements to support predictive models of SEI behaviour.

The institutional landscape of this multimodal research reflects a clear geographic division: Chinese academic centres — the Chinese Academy of Sciences, Southern University of Science and Technology, Peking University, and HKUST — dominate recent cryo-EM battery publications, while US national laboratories and universities lead in operando TEM and multimodal characterisation. The trend across 2019–2023 is toward lower electron doses, 3D tomographic reconstruction, extension to non-lithium alkali metal systems, and integration of cryo-EM structural data with electrochemical performance metrics. For R&D teams and patent strategists tracking this space, understanding which institutions own the key cryo-EM methodologies — and how those methods are being extended — is essential for competitive intelligence. PatSnap’s R&D intelligence platform and patent analytics tools provide structured access to this institutional and IP landscape.