Dry Electrode Coating: 8 Defect Modes and Their Signatures

Dry electrode coating, the solvent-free process where PTFE binder is fibrillated under shear into a freestanding film and laminated onto a current collector, introduces eight distinct defect classes not found in conventional wet coating. Each has a specific process parameter origin, a measurable visual or structural signature, and a predictable cell-level consequence. The challenge on a production line is that the defect you see is rarely where the failure started.

What dry electrode coating actually is, and why it fails differently

In conventional wet electrode manufacturing, active material, conductive additive, and a PVDF binder are dissolved in N-methyl-2-pyrrolidone (NMP) to form a slurry, which is slot-die coated onto aluminum or copper foil and passed through a drying oven. The ovens are the bottleneck. They run 30 to 100 meters long, consuming 30 to 50 kWh of energy for every kilowatt-hour of finished cell capacity^[1]. They dominate the factory floor and define the production rate.

Dry electrode coating eliminates the solvent entirely. The process has four discrete steps: dry mixing of active material, conductive additive, and PTFE binder; fibrillation of PTFE under shear force into a nanofibril network; calendering of the powder mixture into a self-supporting film; and thermal lamination of that film onto the current collector^[2]. No oven. No solvent recovery. Factory footprint shrinks by roughly 10× for the electrode manufacturing section, and energy consumption by over 70%.

The mechanical basis of the process is the unique fibrillation property of PTFE. When subjected to shear above its β-transition temperature of 19 °C, PTFE beads begin to elongate into fibrils. Complete fibrillation, however, requires elevated temperature above 80 °C, as Chen and colleagues established^[3]. This fibril network holds active material and conductive additive in a mechanically coherent film without any liquid-phase binder migration. It's structurally very different from a wet-coated electrode, and it fails in structurally very different ways.

Case study: Tesla's 4680 ramp and what it taught the industry

Maxwell Technologies developed the Dry Battery Electrode (DBE) process for supercapacitors beginning in 2008, and by 2017 had demonstrated it for battery electrodes at high areal loadings. Tesla acquired Maxwell in 2019 and, at Battery Day in September 2020, announced the 4680 cylindrical cell using dry electrode technology for both anode and cathode^[2]. The 4680 promises an electrode up to 250 μm thick with an areal capacity of 6 mAh/cm², roughly double conventional loadings.

What followed is the most instructive public record of dry electrode failure modes at production scale. During the Q2 2021 earnings call, Elon Musk disclosed that the primary calendering challenge was physical: high-nickel NMC cathode particles were hard enough to dent the calender rolls themselves^[4]. A dented roll creates a periodic thickness defect that repeats at every revolution. Every centimeter of electrode that passes carries the signature of that damage. Early pilot lines at Kato Road ran at 70–80% yield^[5]. The industry norm for a mature wet electrode line is 94–96%.

Tesla took an anode-first approach. The dry-electrode graphite anode entered production before the cathode. Current Cybertruck and certain Model Y cells built in Texas use Tesla's dry-processed anode paired with a conventionally wet-processed cathode from LG or CATL^[6]. The dry cathode was reportedly finalized at laboratory scale in late 2022 but stalled in mass production after technical misjudgments; a course correction following the departure of engineering lead Drew Baglino in April 2024 eventually led to a reported breakthrough. As of early 2026, Berlin lines are reportedly achieving 92% yield on 4680 production^[7]. The gap between 70% and 92% represents five years of defect investigation.

The 8 defect classes

1. Under-fibrillation voids

Under-fibrillation occurs when PTFE does not form a continuous fibril network. This happens when the shear force is insufficient, the mixing temperature was too low, the mixing time was truncated, or the wrong PTFE grade was used. The only fluoropolymer that fibrillates under mechanical shear is PTFE. FEP, PFA, and ETFE do not^[2].

The visual signature at the electrode surface is a non-uniform appearance: dark spots where active material distribution is inconsistent, and macroscopic regions that appear loosely bound compared to well-fibrillated areas. In SEM cross-section, under-fibrillated electrodes show large randomly distributed voids that persist even after calendering^[8]. The cell-level consequence is mechanical weakness and low initial coulombic efficiency, particularly in anode applications where PTFE is susceptible to reductive decomposition at low potentials.

2. Pinholes

Pinholes are point defects where the electrode film has no material. They are direct through-holes in the active layer. In dry electrode processing, they are the downstream consequence of under-fibrillation: regions where the PTFE network failed to bridge the powder into a coherent film, leaving gaps. They are not self-healing the way wet-process bubbles can be during the drying phase.

Optical inspection systems detect pinholes above approximately 50 μm diameter using strobed photometric stereo methods or line-scan cameras^[9]. Below that threshold, they are invisible to standard inline systems and require X-ray or computed tomography. At cell level, a pinhole is a lithium dendrite nucleation site. Under cycling, dendrites grow preferentially through low-resistance paths; a pinhole in the separator direction is a latent internal short.

3. Edge cracking

Edge cracking appears during lamination when the dry electrode film is pressed onto the current collector. The lamination step requires compressive force to bond the freestanding film to the foil. If the PTFE fibril network has insufficient ductility, from under-fibrillation or from an incompatible film-to-foil elongation mismatch, the film cannot accommodate the deformation stress and fractures along the edges^[8].

The signature is macroscopic cracking visible to the naked eye at the electrode margins, with SEM cross-sections showing fracture lines running parallel to the current collector near the interface. Poorly fibrillated electrodes show severe macro-cracking during lamination; well-fibrillated electrodes laminate cleanly^[8]. The consequence is active material loss at the cell edges, which reduces realized capacity and creates mechanical instability that propagates during winding into a full cell.

4. Calender roll damage and periodic thickness variation

This is the defect Tesla disclosed most directly. High-nickel NMC particles, which are required for high energy density cathodes, have particle hardness sufficient to plastically deform the surface of the calender rolls used to press the dry powder into film. Each dent then stamps a corresponding thickness anomaly into every subsequent meter of electrode web^[4].

The signature is a periodic pattern in laser caliper thickness measurements, with the repeat distance matching the roll circumference. On a 200 mm diameter roll, you'll see a thickness anomaly every 628 mm of web. Unlike a random coating defect, this defect is deterministic and propagates indefinitely until the rolls are reground or replaced. Locally thicker regions have lower porosity; locally thinner regions may have insufficient active loading. Both translate to non-uniform electrochemical performance across the cell.

5. Delamination from the current collector

Delamination is the separation of the electrode film from the foil, either partially or completely. In dry electrode processing it is primarily a lamination-stage failure: insufficient bonding temperature, insufficient lamination pressure, or an elongation mismatch between the film and the foil causes adhesion failure at the interface^[10].

The visual signature ranges from visible peeling at the electrode edge to a subtle gap detectable only in SEM cross-section or by surface and interfacial characterization analysis system (SAICAS) peel strength measurement. Dry electrodes that laminate successfully show no gap at the film-foil interface under SEM; delaminated electrodes show a clear separation layer. At cell level, delamination increases contact resistance, accelerates capacity fade from the first cycle, and in severe cases causes cell failure during winding. The electrode tears rather than winds cleanly.

6. Porosity non-uniformity

Porosity in a dry electrode is set entirely by the calendering step. There is no solvent evaporation to contribute or compensate. If calendering pressure is too high, the pore network collapses and ionic transport is throttled. If it is too low, the electrode is under-densified and the PTFE network is insufficiently consolidated. Horst and colleagues demonstrated that the optimal porosity for NCM electrodes is approximately 32%, at which point charge transfer resistance is minimized and discharge capacity is maximized at medium and high rates^[11].

The signature of porosity non-uniformity is not visible optically. It requires mercury porosimetry or X-ray computed tomography to map pore distribution across the electrode cross-section. Locally under-compressed and over-compressed regions coexist on the same web when roll gap is inconsistent or when powder feed rate fluctuates. At cell level, high-porosity regions show lower rate capability; low-porosity regions show faster capacity fade as lithium-ion transport becomes diffusion-limited.

7. Active material particle cracking

Under excessive calendering pressure, secondary particles of active material fracture. This is a particular problem with high-nickel NMC cathode materials, which have both high hardness and a layered structure that is susceptible to intergranular cracking under compressive stress. Nano-X-ray computed tomography experiments on NMC electrodes before and after calendering show a measurable increase in intragranular cracks within secondary particles as a direct function of applied pressure^[9].

Particle cracking is not visible at the electrode surface. It requires nano-XCT or post-mortem SEM of cross-sectioned particles. The cell-level consequence is severe and delayed: fresh particle surfaces exposed by cracking react with the electrolyte, generating additional SEI formation, transition metal dissolution, and structural degradation that accelerates capacity fade over the first 50–100 cycles. This is the failure mode that can pass initial cell qualification testing and manifest as premature field degradation.

8. Surface chattering and wrinkles

Chattering is a periodic surface undulation, a wave pattern pressed into the electrode film during calendering. It results from stick-slip dynamics in the roll-to-roll process: inconsistent powder feed, mismatched roll speeds, or tension imbalance between the front and back of the web. Analysis of wrinkle formation in calendering shows that the severity increases linearly with web tension difference; reducing the tension difference from 20 N to 5 N cut corrugation tortuosity by 34.2% in controlled trials^[10].

The signature is visible optically as a regular surface wave pattern, typically oriented perpendicular to the machine direction. Minor chattering is often eliminated during lamination to the current collector. Persistent chattering is not. At cell level, chatter marks create localized thickness variation and porosity banding that produces non-uniform electrochemical utilization across the electrode web.

The process parameters that drive these defects

Every defect above traces back to one of four process parameter domains. At the fibrillation stage, temperature is the primary lever: below 80 °C, PTFE forms partial fibrils that lack the mechanical continuity needed for a coherent film, generating under-fibrillation voids and downstream pinholes^[3]. Mixing time and shear rate determine fibril density; shortened mixing produces agglomerates that appear optically as surface spots and translate to local capacity hotspots.

Active material morphology is underappreciated as a process variable. Horst and colleagues showed that graphite's platelet morphology causes PTFE to slide rather than fibrillate, requiring longer mixing times or higher shear. The same mixing recipe that works for spherical LFP will produce an under-fibrillated graphite anode if applied unchanged. NCM, with its high density and faster compaction speed, promotes rapid fibrillation but also requires higher calendering compression stress, increasing the risk of particle cracking^[11]. A process recipe is not transferable across chemistries without re-optimization.

At the calendering stage, pressure and temperature interact. Increasing process temperature enhances fibrillation degree and simultaneously reduces the critical stress at which NCM particles crack. Two benefits from a single adjustment^[12]. Roll surface condition is a one-way degradation: once dented, the roll is a defect source until replaced. At the lamination stage, the temperature and pressure window for adhesion is narrow and chemistry-specific, and it cannot be determined from wet-electrode process knowledge.

Where most dry electrode lines get this wrong

The detection problem in dry electrode manufacturing is fundamentally different from wet coating. In slot-die coating, a process upset, whether a pressure drop or a slurry viscosity change, shows up as a visible defect within seconds of web travel. You catch it fast.

In dry electrode, a temperature drift in the fibrillation mixing zone does not produce a visible defect immediately. The under-fibrillated powder travels through the calender and through lamination before any surface anomaly becomes detectable by an optical camera. By the time the inspection system flags a region, 30 to 90 minutes of web production may have already been compromised. Potentially hundreds of meters of electrode. The defect signature is thermal and mechanical, upstream, not optical and downstream.

Most lines monitor what is easy to measure: thickness at the calender output, optical appearance after lamination. They do not monitor the fibrillation state of the powder before it enters the calender. That's the gap. A process temperature excursion that lasts 15 minutes will produce a defect zone that extends far beyond the flagged region, because the output signal lags the input disturbance by the full transit time of the powder through the process.

What AI defect detection changes for dry electrode lines

A line-scan vision system running on a dry electrode line covers 100% of the electrode web. A human inspector covers approximately 5%. The difference matters for any defect class, but it matters most for periodic defects like calender roll damage, where the defect repeats predictably and can be caught early enough to prevent roll damage from propagating across an entire production run.

Niobia AI addresses the upstream lag problem directly. Rather than monitoring only the optical output of the electrode, the platform combines vision-based defect classification with process-parameter telemetry (fibrillation temperature, mixing torque, calender roll gap, lamination pressure) and learns the relationship between process state and defect signature. When a temperature excursion occurs in the mixing zone, Niobia AI correlates the thermal event with the defect that appears 30 to 90 minutes later on the web, and flags the intervening material for review before it reaches slitting. When a defect type appears that doesn't match any existing pattern, and dry electrode lines encounter defect classes that wet-line inspection libraries have never catalogued, Niobia AI automatically logs it as a new entry in the facility's defect library, including its visual signature, process context, and cell-level outcome linkage.

Root cause analysis on a dry electrode line is particularly time-consuming because the defect-to-cause chain crosses four process stages. Niobia AI reduces the time from defect detection to confirmed root cause by 50×. A manual RCA on a multi-stage dry electrode line typically takes three to five days. The platform surfaces a structured root cause report within minutes.

Summary

Dry electrode coating introduces a failure mode taxonomy that doesn't exist in wet processing: eight defect classes rooted in the mechanics of PTFE fibrillation, calendering dynamics, and lamination physics. Tesla's 4680 ramp, from 70–80% pilot yield to a reported 92% in production, is the most detailed public record of what it takes to understand and control these defect mechanisms at scale. Catching them requires both 100% web coverage and upstream process-parameter correlation. Niobia AI connects vision-based defect detection with fibrillation and calendering telemetry, closing the 30–90 minute lag between process drift and visible defect that conventional inspection systems miss entirely.

About the author

Dr. Gaurav Jha is the Founder of Niobia AI, which builds AI-powered defect detection and process intelligence platforms for battery gigafactories. His PhD focused on fast-charging niobium pentoxide (Nb₂O₅) based nanostructured anodes, with broader research spanning gas sensors, ion sensors, and energy storage materials. At Intel, he worked on wet etch defect reduction in 5nm and 7nm chip fabrication, developing a hands-on instinct for process root cause analysis at scale that translates directly to electrode manufacturing.

He returned to batteries to develop one of the first large-scale lithium-sulfur cathode coatings at Lyten, then moved to Sila Nanotechnology where he worked on silicon anode particles for high energy density and fast-charging applications across consumer electronics and automotive programs. Across these roles, Dr. Jha led manufacturing scaleup from lab to high-volume production, conducted industrial root cause investigations, commercialized key materials products, and developed new electrode chemistries from first principles. He founded Niobia AI to bring that depth of manufacturing and materials science experience into an AI platform built specifically for the production floor.

References

1. Rigaku Corporation. (2026). Tesla's dry electrode manufacturing: From patent to factory reality. Rigaku Industries Blog. https://rigaku.com/industries/environment-and-energy/batteries/blog/tesla-dry-electrode-manufacturing
2. Schumm, B., Schwerdt, B., Günther, T., Kwade, A., & Fleischer, J. (2025). Dry Battery Electrode Technology: From Early Concepts to Industrial Applications. Advanced Energy Materials, 15, 2406011. https://doi.org/10.1002/aenm.202406011
3. Sadan, M. K., Kim, H., Kim, C., & Cho, G. B. (2023). Co, Ni-Free Ultrathick Free-Standing Dry Electrodes for Sustainable Lithium-Ion Batteries. ACS Applied Energy Materials, 6(21), 10773–10782. https://doi.org/10.1021/acsaem.3c02448
4. Musk, E. (2021, July 26). Tesla Q2 2021 Earnings Call. Tesla, Inc. Public earnings call transcript; calender roll challenges disclosed in response to analyst questions on 4680 yield.
5. Alvarez, S. (2021, August 27). Tesla's 4680 battery cell pilot production line hits 70-80% yield: report. Teslarati. https://www.teslarati.com/tesla-4680-battery-production-ramp-challenges-explained/
6. TesmAg Editorial. (2025, November 5). Tesla sprints toward the "complete" 4680 battery: a final gamble after five years. LifePO4 Battery Shop Blog. https://www.lifepo4batteryshop.com/blogs/journey-to-tesla-4680-battery-breakthrough.html
7. TesmAg Editorial. (2026, February 13). The Berlin threshold: Why 92% yield on dry electrode 4680 changes everything for Tesla's European margins. TesmAg. teslaacessories.com
8. Author Group. (2025). Dual-fibrous PTFE structure enabling uniform and thick dry electrodes for high-energy-density and long-lasting batteries. Energy & Environmental Science. https://doi.org/10.1039/D5EE03240G
9. Du Baret De Limé, A., Ganser, M., Braunwarth, W., Soczka-Guth, T., & Elsner, C. (2022). Impact of Electrode Defects on Battery Cell Performance: A Review. Batteries & Supercaps, 5(10), e202200239. https://doi.org/10.1002/batt.202200239
10. Abdollahifar, M., Cavers, H., Scheffler, S., Diener, A., Lippke, M., & Kwade, A. (2023). Insights into Influencing Electrode Calendering on the Battery Performance. Advanced Energy Materials, 13(40), 2300973. https://doi.org/10.1002/aenm.202300973
11. Horst, M., Beverborg, F., Bahlmann, L., Schreiber, S., Gerk, J., Michalowski, P., & Kwade, A. (2025). Effect of active material morphology on PTFE-fibrillation, powder characteristics and electrode properties in dry electrode coating processes. Powder Technology, 451, 120451. https://doi.org/10.1016/j.powtec.2024.120451
12. Huber, K., et al. (2025). Dry battery electrode production enabled by a scalable, continuous powder mixing. Journal of Energy Storage. https://doi.org/10.1016/j.est.2025.015634
13. Choudhary, N., Clever, H., Ludwigs, R., Rath, M., Gannouni, A., Schmetz, A., Hülsmann, T., Sawodny, J., Fischer, L., Kampker, A., Fleischer, J., & Stein, H. S. (2022). Autonomous Visual Detection of Defects from Battery Electrode Manufacturing. Advanced Intelligent Systems, 4(12), 2200142. https://doi.org/10.1002/aisy.202200142