Most buyers start from a gas list and a model number. The sensing principle is usually already fixed by the time anyone reads the datasheet, because it was implied by the first quote received. That ordering is backwards, and it is the most common source of detectors that read plausibly and report the wrong number.
Four principles cover the large majority of industrial applications. Each one solves a different problem, and each one fails in a characteristic way.
Electrochemical
Gas diffuses through a membrane to a working electrode, where it is oxidised or reduced. The resulting current is proportional to concentration. The cell consumes a tiny amount of the target gas, which is why the reading is selective and why the sensor has a finite life.
Use it for toxic gases at ppm and ppb levels, and for oxygen. Hydrogen sulphide, carbon monoxide, sulphur dioxide, chlorine, ammonia and ozone all fall here.
Limits worth knowing: an electrochemical cell exposed to a concentration far above its range can be permanently damaged, so range selection protects the sensor as much as it protects the measurement. Performance degrades in very dry or very cold conditions. Expect two to three years of life, shorter for oxygen cells.
Catalytic bead
The gas burns on a heated catalytic bead, and the temperature rise changes the bead resistance. It measures how much combustible gas is present, expressed as a percentage of the lower explosive limit.
Use it for %LEL monitoring of methane, propane, hydrogen and solvent vapours. It is inexpensive, well understood, and directly tied to the quantity that matters for explosion risk.
Limits worth knowing: the bead needs oxygen to work, so it cannot measure combustibles in an inert or oxygen-displaced atmosphere. It responds to almost any combustible gas without distinguishing between them, which is acceptable for a safety alarm and unacceptable if you need to know which gas leaked. Sulphur compounds, silicones and halogens poison the catalyst permanently, and the sensor will keep responding to its calibration gas while reading low on the real hazard.
NDIR
The gas absorbs infrared light at wavelengths characteristic of its molecular bonds. A detector compares absorption at the measurement wavelength against a reference wavelength.
Use it for carbon dioxide, methane, and other hydrocarbons where a stable long-life sensor matters. It also reads correctly in oxygen-deficient atmospheres, which is where catalytic beads cannot go.
Limits worth knowing: it is the more expensive option, resolution is coarser than electrochemical for low ppm work, and it only works for gases with a suitable infrared absorption band. Carbon monoxide and hydrogen are poor candidates.
PID
A UV lamp ionises gas molecules, and the resulting current indicates total concentration. A 10.6 eV lamp covers most volatile organic compounds.
Use it for total VOC measurement: solvent vapours in a paint shop, fugitive emissions around a chemical plant, leak surveys where the specific compound is unknown.
Limits worth knowing: a PID reports a total, not a species. It cannot tell you whether the reading comes from toluene or from the solvent next to it. Response varies by compound, so converting a reading into a concentration for a specific gas requires a response factor. The lamp is a consumable, typically replaced every six to twelve months.
Selection matrix
| What you need to measure | Principle | Why |
|---|---|---|
| Combustible gas, %LEL, normal air | Catalytic bead | Directly measures explosion risk, lowest cost |
| Combustible gas in inert atmosphere | NDIR | Catalytic bead needs oxygen to burn |
| Toxic gas at ppm or ppb | Electrochemical | Selective and sensitive at low concentration |
| Oxygen deficiency or enrichment | Electrochemical, dedicated cell | A partial-pressure measurement, not a toxicity one |
| Carbon dioxide | NDIR | No practical electrochemical option |
| Total VOC, unknown species | PID | Broad response covers compounds you did not list |
| Specific VOC in a known process | Electrochemical or NDIR | PID cannot distinguish one VOC from another |
Where cross-sensitivity bites
No sensor is perfectly selective, and the datasheet cross-sensitivity table is usually a footnote.
A catalytic bead responds to everything combustible. If a solvent vapour and methane are both present, the %LEL reading is a weighted sum of the two, which is correct for safety purposes and wrong if you are trying to attribute the reading.
Electrochemical cells have known interferences. A hydrogen sulphide cell responds to sulphur dioxide. A carbon monoxide cell can respond to hydrogen. In a coke oven or a landfill, where several gases coexist, this matters.
PID readings are the easiest to misread. The instrument shows a total in ppm, and that number is not the concentration of any single compound unless a response factor has been applied.
If your process has more than one gas present, the specification should name the interference and state the expected magnitude. A supplier who cannot answer that question has not tested the combination.
What this means for a module specification
When a multi-channel module is configured, the sensing principle is chosen per channel, not per module. A four-gas confined-space instrument typically carries one catalytic bead for %LEL, one electrochemical cell for oxygen, and two electrochemical cells for hydrogen sulphide and carbon monoxide. A semiconductor process tool monitoring might carry NDIR for carbon dioxide, electrochemical for an acid vapour, and PID for solvent.
Modules that hold six channels can mix principles freely, which is what makes a single board practical across different applications. The constraint is power and warm-up time, not the principle itself.
Sensor selection depends on the specific gas mix, the concentration range and the environment. Confirm against the datasheet for the configured module before design freeze; this page explains the trade-offs, not a specific product configuration.