Battery Failure Analysis with SEM-EDS

Why Your Lithium-Ion Batteries Keep Failing: Finding the Root Cause with SEM-EDS-Raman

January 2026 • 6 min read

That thermal runaway incident? The capacity fade that’s killing your cycle life? The answer is probably visible at 10,000x magnification.

Battery development is a race measured in charge cycles and energy density. When cells fail during testing—or worse, in the field—finding the root cause isn’t optional. It’s the difference between a design fix and a recall.

Scanning electron microscopy with energy dispersive spectroscopy (SEM-EDS) has become the go-to technique for battery failure analysis because it answers the two questions that matter most: What does the failure look like? and What’s it made of?

73%
of battery failures involve electrode or separator issues visible with SEM
$150M+
cost of major EV battery recalls in recent years
4 hrs
typical time from cell teardown to root cause identification

What Battery Failures Look Like Under SEM

Most lithium-ion battery failures fall into recognizable patterns when you know what to look for:

Lithium Dendrite Formation

The nightmare scenario. Metallic lithium deposits grow from the anode surface during charging, eventually piercing the separator and causing internal short circuits. Under SEM, dendrites appear as needle-like or mossy structures on the anode surface. EDS confirms they’re pure lithium (though lithium is too light for standard EDS—you identify it by eliminating other elements and observing the morphology).

What you’ll see: Needle-like projections 1-50 microns long growing perpendicular to the electrode surface. In severe cases, dendrites penetrate through the separator.

Electrode Particle Cracking

Cathode particles (NMC, LFP, NCA) expand and contract during cycling. Over hundreds of cycles, this mechanical stress causes cracks that isolate portions of the active material, reducing capacity.

What you’ll see: Fractured particles with clear crack patterns. Cross-sections show internal cracking not visible from the surface. EDS mapping reveals whether cracks correlate with compositional variations.

SEI Layer Problems

The solid electrolyte interphase (SEI) on the anode should be thin, uniform, and stable. Excessive SEI growth consumes lithium and increases impedance. Uneven SEI causes non-uniform current distribution.

What you’ll see: Surface films of varying thickness. EDS shows elevated fluorine, oxygen, and carbon from decomposition products. Compare failed cells to good cells to establish baseline.

Separator Damage

Thermal events, mechanical abuse, or dendrite penetration can damage the separator. Even microscopic holes or thin spots can lead to internal shorts.

What you’ll see: Holes, tears, or melted regions in the polymer separator. EDS may reveal metal contamination at damage sites (evidence of internal shorting).

Metal Contamination

Foreign metal particles from manufacturing can cause localized shorts. Even particles in the 10-50 micron range can nucleate dendrite growth or directly pierce the separator.

What you’ll see: Bright particles in BSE imaging (metals are heavy, appear bright). EDS immediately identifies the contaminant—iron, copper, aluminum, or other metals point to specific manufacturing process issues.

Adding Raman: The Chemical Fingerprint

While SEM-EDS tells you what elements are present, Raman spectroscopy reveals the molecular structure—critical for understanding battery degradation mechanisms. Together, they provide a complete picture of what’s happening in your cells.

Cathode Phase Identification

Different cathode phases (NMC, LFP, LCO, NCA) have distinct Raman signatures. More importantly, phase changes during cycling—which directly impact capacity and stability—are immediately visible in Raman spectra.

What you’ll see: Peak shifts indicating structural changes. New peaks appearing from phase transformations. Broadening that indicates disorder or degradation.

Graphite Anode Degradation

The D/G band ratio in Raman is a direct measure of graphite disorder. Fresh anodes show a strong G band (ordered graphite). As cycling damage accumulates, the D band grows, indicating structural breakdown.

What you’ll see: D band intensity increasing with cycle count. Changes correlate directly with capacity fade. Compare cycled vs. fresh electrodes to quantify degradation.

SEI Layer Composition

The SEI layer contains organic and inorganic compounds from electrolyte decomposition. Raman identifies specific components—carbonates, fluorides, organic polymers—that EDS can’t distinguish.

What you’ll see: Organic carbonate peaks, Li2CO3 signatures, polymer decomposition products. Different electrolyte additives produce different SEI compositions.

Electrolyte Decomposition

When electrolyte breaks down on electrode surfaces, Raman identifies the decomposition products. This helps diagnose whether failures are related to electrolyte instability, operating outside the voltage window, or thermal events.

SEM + Raman Workflow: Use SEM imaging to locate the defect, then position the Raman laser on that exact spot for chemical identification. The combination answers both “where is the problem?” and “what exactly happened?”

Real Case: The Mysterious Capacity Fade

A battery startup was seeing 15% capacity loss after just 200 cycles—far below their 500-cycle target. Visual inspection showed nothing. SEM revealed cathode particle cracking concentrated at particle centers. EDS mapping showed these particles had higher manganese content than spec. The problem was traced to a supplier batch variation. Root cause identified in one afternoon.

The Analysis Workflow

Here’s how battery labs typically approach failure analysis with SEM-EDS:

  1. Cell teardown: Disassemble the cell in a glovebox or dry room to preserve moisture-sensitive components
  2. Sample preparation: Cut electrode samples, mount separators, prepare cross-sections if needed. Minimize air exposure for anode samples.
  3. Initial survey (SE imaging): Low magnification scan of electrode surfaces. Look for obvious damage, contamination, or morphology changes.
  4. Detailed imaging: High magnification of suspect areas. Document particle morphology, surface films, damage patterns.
  5. BSE imaging: Switch to backscatter mode to identify compositional variations. Heavy contaminants appear bright.
  6. EDS analysis: Point analysis on specific features. Element mapping across larger areas to show distribution patterns.
  7. Compare to baseline: Image good cells with the same parameters. Differences highlight failure mechanisms.
Sample Prep Tip: Anode samples with lithium are highly reactive. Transfer under argon and minimize SEM chamber exposure. Some labs use a glovebox-attached sample transfer system.

What You Need

Recommended Configuration for Battery Analysis

  • SNE-Alpha Desktop SEM: 5nm resolution for fine particle and dendrite imaging. Fast 90-second vacuum for reactive sample handling.
  • Bruker XFlash EDS: Elemental identification and mapping. Detect contaminants, verify compositions, map element distributions across electrodes.
  • Raman Spectroscopy: Chemical fingerprinting for phase ID, graphite degradation analysis, and SEI layer composition. Correlate with SEM imaging for site-specific analysis.
  • Cross-section prep: Ion milling or focused ion beam (FIB) for internal structure analysis. (Can be outsourced if not available in-house.)

ROI: Why In-House SEM Pays Off

Most battery companies start by outsourcing SEM analysis. At $200-500 per sample with 1-2 week turnaround, this adds up fast—and slows development cycles.

The math: If you’re analyzing 10 samples per week at $300 each, that’s $156,000/year in service lab fees. A desktop SEM-EDS system pays for itself in 12-18 months while giving you same-day results.

More importantly: when a cell fails on Friday afternoon, you can have answers Monday morning instead of waiting two weeks for a service lab report.

Getting Started

If your battery development program doesn’t have in-house SEM capability, you’re making decisions with incomplete information. Every cell that fails contains data—but only if you can see it.

See Your Battery Failures in Detail

Send us your failed cells or electrode samples. We’ll image them and show you exactly what SEM-EDS reveals.

Request Sample Analysis

Additional Resources

NanoImages Assistant

Online

Hi! I'm the NanoImages AI assistant. Ask me anything about our SNE-Alpha desktop SEM, applications, sample prep, or scheduling a demo.