Battery Safety Rules Are Creating a Market for Mid-IR Bandpass Filters

Gas Sensing and Environmental Monitoring · 2026-09-29 · 4 min read

The fastest-growing gas-sensing application of the decade is not in a factory or a laboratory. It is inside a battery pack. Thermal runaway detection — catching a failing cell before it becomes a fire — has moved from an engineering aspiration to a regulatory requirement in several markets, and the optical sensing techniques that make it possible depend on infrared filters and windows that were previously niche components.

Battery Safety Rules Are Creating a Market for Mid-IR Bandpass Filters

The volumes are small compared with consumer optics, but the specification discipline is closer to medical devices than to industrial instrumentation. That combination is unfamiliar to many filter suppliers.

What the sensors are actually measuring

A failing lithium-ion cell vents a characteristic mixture of gases before visible smoke appears. Depending on chemistry, that mixture includes carbon monoxide, carbon dioxide, hydrogen, and volatile organic compounds, sometimes with hydrogen fluoride in small quantities. Detection strategies vary:

  • NDIR absorption. A mid-IR source is filtered to a narrow band centred on an absorption line, passed through a sample cell, and detected. A second reference channel compensates for source drift and contamination.
  • Photoacoustic detection. Modulated light absorbed by the gas generates a pressure wave, which a microphone detects. Optical requirements are similar but the tolerance to stray light is much tighter.
  • Multi-gas arrays. Several bandpass filters on a single detector package, allowing one sensor to distinguish gas species rather than simply detect "something."

All three need bandpass filters with well-defined centre wavelength, narrow bandwidth and deep out-of-band blocking, plus a window that passes the sensing wavelengths and survives the environment inside a pack.

Why the automotive environment is hostile to filters

A filter that works in a benchtop analyser meets a different world inside a vehicle.

Temperature range is the first issue. Bandpass filters shift with temperature, and the shift is not negligible over an automotive range. The filter design, the substrate choice and the detector's own temperature compensation have to be considered together; specifying a centre wavelength at 25 °C and ignoring the rest is how sensors develop false alarms in winter.

Humidity and contamination are the second. A pack is not a cleanroom. Condensation on a window, dust ingress or outgassing from nearby polymers all change the optical baseline. Hydrophobic coatings, hermetic sealing and a reference channel are the usual mitigations, but each adds cost and each needs to be designed in from the start.

Vibration and shock are the third. Thin filters in small packages are mechanically fragile, and a filter that cracks in a crash event is worse than useless. Mounting and, where appropriate, bonding to a robust substrate matter as much as the coating itself.

The specification conversation that actually matters

Buyers coming from industrial gas sensing often specify three numbers: centre wavelength, FWHM and peak transmission. Those are necessary and insufficient.

The numbers that determine whether a sensor works in the field are:

  • Blocking across the full detector response range, not just adjacent to the passband. A mid-IR detector with a broad response will happily see a leak two octaves away.
  • Angle behaviour. NDIR cells use divergent sources, so light arrives across a range of angles. A filter specified only at normal incidence will have a broader and shifted passband in the real cell.
  • Thermal coefficient of the centre wavelength, stated as a number and verified, so the electronics can compensate.
  • Environmental durability of the coated surface, including adhesion under damp heat — the standard automotive test regime, not a generic optical one.

Suppliers who can quote all four, with data, are a much shorter list than suppliers who can quote the first three.

Where the volume will come from

The near-term volume is in stationary storage and in commercial vehicles, where pack-level monitoring is already mandated or strongly encouraged. Passenger cars follow, but the cost target is lower and the packaging constraint is tighter, which pushes designs toward integrated sensor modules with the filter and detector in one package.

That integration trend favours filter suppliers who can work at wafer or panel level and deliver diced, coated, optionally mounted parts rather than loose optics. It also favours suppliers who can support a multi-year programme without changing a recipe silently — the same change-control discipline that medical and automotive programmes already demand.

Battery safety is not a fashion. It is a compliance requirement with a long runway, and the optical content inside it is small, precise and hard to substitute once designed in.

← All 410 articles

Need a Quote or Engineering Review?

Send your drawings, specifications or coating requirements — we reply within 24 hours.

Request a Quote

Related Articles

Request a Quote