In-line electrode coating thickness measurement in lithium-ion cell production is implemented with non-contact optical (laser triangulation, confocal chromatic) and nucleonic (X-ray, beta) gauges, delivering real-time wet- and dry-coat profiles across the full web width at line speeds of 30–80 m/min [S1][S3].
These measurements feed closed-loop signals back to the slot-die coater, controlling either the die gap or the web speed to hold coating weight (g/m²) within process windows, a critical step because even small mass-loading deviations propagate into capacity variance, rate capability loss, and accelerated cell ageing [S4][S8].
Why in-line measurement, not lab off-line
Off-line micrometer and weigh-strip checks sample less than 0.1% of the electrode area and detect a defect 5–20 minutes after it is coated, when several hundred meters of foil have already been rejected as suspect material [S2].
In-line gauges scan the web in both machine-direction (MD) and cross-direction (CD) to detect coating non-uniformity that would otherwise compromise performance, reliability, and safety [S3][S7].
Main sensor technologies compared
Four non-contact families dominate current production lines: laser triangulation, confocal chromatic, beta transmission, and X-ray fluorescence/reflection, each with a distinct operating envelope that dictates where it sits in the process train [S1][S3][S4].
Laser triangulation (Micro-Epsilon optoNCDT 1420/1900 series and similar) is a one-sided, low-cost sensor that resolves single-sided coating thickness to ±1–2 µm and is widely mounted directly after the slot-die for wet-film control [S1]. Confocal chromatic sensors measure a few hundred micrometres standoff with sub-micrometre resolution on shiny copper and aluminium foils, fitting both wet and dry positions [S1].
Beta gauges (typically using Kr-85 or Pm-147 sources) give absolute areal density in g/m² regardless of chemistry, with edge resolution of a few millimetres, and remain the workhorse for dry, post-drying coating weight control [S3]. X-ray gauges read coating weight on each side of the foil independently, enabling first-coat/second-coat math, but at higher capital and radiation-shielding cost [S3][S7].
Closed-loop control strategy
The closed-loop architecture has three layers: (1) local slot-die bolt control for CD profile, (2) gap or pump-rate adjustment for MD mean, and (3) dryer set-point trim to compensate for solvent-loading drift [S4][S7].
CD profile control typically uses 20–60 individually actuated die bolts spaced 10–25 mm apart, with the gauge profile as the setpoint input; the die correction profile is updated every 1–5 seconds, and stable edge exclusion is 5–10 mm on each side of the web [S7]. For two-sided coatings, a total-thickness station after the second coater plus a first-side station after the first coater allow the second-coat thickness to be derived by subtraction [S7].
Process windows and spec targets
Typical coating-weight tolerances on a modern Li-ion line are ±1.5–2.0% (1σ) of nominal loading, with single-coat cathode loadings of 15–25 mg/cm² on each side of the foil and anode loadings of 8–12 mg/cm²; a 1% mass-loading drift on the cathode translates into roughly 1% capacity drift at the cell level [S3][S8].
Wet-film gauges (laser or confocal) must resolve the wet state, where the coating contains 30–50% solvent by mass, so the post-dry thickness is mathematically inferred from the wet reading, the solids fraction, and the substrate base reading taken at the unwinder [S1][S4].
Where the sensors physically sit
On a typical double-coated cathode line, a first-side gauge sits within 200–500 mm downstream of the first slot-die, a second-side gauge follows the second coater, and a beta or X-ray dry-coat gauge sits after the dryer but before the calender; a final calendered-electrode gauge closes the loop on the compaction press [S4][S7].
Gap to the foil is 5–30 mm for triangulation and confocal sensors, and a few millimetres for the optical head of X-ray fluorescence gauges; standoff is set to the smallest value the line web-tension control can hold without inducing flutter, because vertical web oscillation is the dominant noise source and is normally limited to ±0.5 mm peak-to-peak in well-tuned lines [S1][S9].
Limitations and failure modes
Optical one-sided gauges cannot distinguish coating-thickness changes from substrate-thickness changes, so any copper or aluminium foil caliper drift (typically ±1–2 µm on a 10–15 µm foil) appears as a coating-thickness error unless a reference substrate reading is subtracted at the unwinder [S1][S4].
Beta gauges require radioactive-source licensing, periodic reference-wedge re-calibration, and are insensitive to local defects smaller than the source/collimator footprint (typically a few millimetres); laser triangulation on the other hand is a low-cost, fast-responding technique that is fooled by foil flutter, surface gloss, and transparent binder migration patterns that shift the apparent optical surface [S1][S3][S9]. A typical high-yield line therefore uses at least two technologies in series, exploiting their different noise and failure-mode profiles.
Industrial coating context beyond batteries
Although the in-line architectures above are framed for electrode foil, the underlying control logic is a direct descendant of continuous-web coating lines in paper, photographic film, and metal packaging, and the same thickness gauge principles apply across those industries. [S3]
Where the application shifts to non-battery substrates (architectural aluminium, anti-corrosion primers, roll-to-roll flexible electronics), the sensor choice widens to include eddy-current and capacitive probes; an overview of the broader gauge landscape is given in the coating thickness gauge reference and the industrial coating application pages, while ultrasonic thickness gauge methods remain dominant for opaque, non-metallic film substrates such as pipeline waterproof coating inspection, where one-sided access is required and the substrate is incompatible with optical reflection.
Selection criteria and decision matrix
For a new Li-ion electrode line in 2026, a defensible spec is: laser triangulation or confocal at the wet station for fast gap/speed control, beta or XRF for absolute g/m² after drying, and XRF or X-ray transmission at the calender exit for coating-only thickness on the compacted electrode; this combination holds dry-coat mass loading to ±1.5% (1σ) on a 30–60 m/min line [S1][S3][S4][S7].
Cross-checking the choice against four criteria, laser triangulation wins on cost, sub-second response, and ease of integration; confocal chromatic wins on shiny-foil accuracy and small standoff; beta wins on absolute areal density without chemistry assumptions; XRF wins on side-resolved measurement and traceability, but loses on capex and shielding overhead [S1][S3][S7].
Buyers should also look at integration load: laser triangulation drives via 0–10 V or IO-Link and slots into existing PLC/SCADA stacks with minimal code change, while XRF and beta platforms are more often delivered as turnkey scanner frames with their own controllers and analytics, raising both commissioning time and dependence on the vendor's service network [S3][S5].
Sourcing and standards
OEM platforms active in this space as of the past six months include Micro-Epsilon (optoNCDT series, thicknessSENSOR) [S1], Thermo Fisher Scientific (in-line web gauging platforms) [S3], Fraunhofer ITWM (inline layer-thickness measurement for battery foils) [S4], Erhardt+Leimer (EL-THICKNESS non-contact system) [S5], and Hammer-IMS (battery electrode coating thickness systems) [S6].
No single IEC or ISO standard fully governs in-line electrode thickness control; the closest applicable reference documents are vendor-specific calibration procedures and ISO/IEC 17025-aligned gauge re-calibration intervals (typically 12 months for nucleonic sources, 6 months for optical heads) [S3][S4]. Battery makers frequently tie their internal control limits to cell-level quality clauses rather than to a published sensor standard, and these clauses typically translate into a ±1.5–2.0% (1σ) coating-weight band at the in-line gauge.
Verifiable next nodes to watch: (1) release of higher-speed XRF line scanners aimed at >80 m/min battery foils, where present throughput is gated by source strength and detector count-rate limits; (2) tighter closed-loop latency below 200 ms between gauge read and slot-die bolt actuation, which would let line builders push coating-weight bands toward ±1% (1σ) on dry cathodes [S3][S4].
This topic is covered further in Sealing Washer Total Cost of Ownership: A Spec-Driven Breakdown.