An electrochemical (amperometric) gas detector uses a sealed cell with a working electrode, counter electrode, reference electrode and a liquid or gel electrolyte; the target gas diffuses through a capillary or membrane and is oxidised or reduced at the working electrode, producing a current proportional to concentration [S1].
Typical full-scale ranges sit at 0–100 ppm for highly toxic species (HCN, Cl2, NO2) and 0–1000 ppm or 0–2000 ppm for less toxic species (H2S, NH3), with a 24-month nominal sensor life in clean service, and IP65/IP66 housings are standard for both portable and fixed formats [S1][S2].
Detection Principle and 2-Electrode vs 3-Electrode Cell
Amperometric cells fall into two practical categories: 2-electrode (working + counter) designs, which are simpler and cheaper but drift as the reference potential shifts, and 3-electrode (working + counter + reference) designs, where a potentiostat holds the working electrode at a fixed potential versus the reference so the output current is governed only by gas diffusion [S4]. The 3-electrode topology is the de facto standard for fixed toxic-gas transmitters because it yields a flatter baseline and slower drift, while 2-electrode cells survive in low-cost disposable cartridges for single-gas portables [S1][S4]. Electrolyte choice varies by target gas: aqueous sulfuric acid is common for CO and H2S cells, while formamide-based mixtures have been documented for phosgene sensing where water content must be controlled to suppress side reactions [S3].
The working electrode is normally a noble-metal catalyst (Pt, Au, or carbon) printed onto a porous PTFE or hydrophobic membrane; the hydrophobic barrier limits electrolyte loss and defines the diffusion-limited current, which is the physical basis for the linear output of the sensor [S4].
Target Gases, TWA/STEL/IDLH Thresholds and Range Selection
Selection starts with the toxicology of the target gas: TWA (8-hour weighted average), STEL (15-minute short-term exposure), IDLH (immediately dangerous to life or health) and MAC (maximum allowable concentration) define the alarm setpoints and the required sensor resolution [S1]. Common electrochemical targets include CO (TWA 25 ppm, STEL 200 ppm in many jurisdictions), H2S (TWA 10 ppm, STEL 15 ppm), NH3 (TWA 25 ppm, STEL 35 ppm), Cl2 (TWA 0.5 ppm, STEL 1 ppm) and SO2 (TWA 2 ppm, STEL 5 ppm); these figures are typical occupational values and the specifier should always reconcile them against the local regulatory schedule before fixing the instrument range [S1].
For a single-gas confined-space application, a dedicated 0–100 ppm or 0–500 ppm H2S sensor is the standard pick; for mixed atmospheres a multi-gas cartridged with separate EC, LEL and O2 cells is preferred over a single broad-range EC cell, because cross-sensitivity between H2S and CO on a shared electrode is a documented source of false alarms [S1].
Portable vs Fixed Format, IP Rating and Output
Portable units are typically diffusion-sampled, weigh 200–400 g and run on Li-ion or AA cells for 12–24 hours per charge, with PC/TPU housings rated IP64–IP66 and audible/visual alarms at the low and high setpoints [S2]. Fixed units use pumped or diffusion sampling, mount on a 3/4" NPT or M20 conduit entry, output 4–20 mA with optional HART, Modbus RTU over RS-485 or relay contacts, and are specced against IEC 60079 series for hazardous-area use where flammable atmospheres may coexist with the toxic target gas [S1][S2].
For specification against combustible-gas risks, see the combustible gas detector reference for LEL/catalytic-bead and NDIR options that pair with EC cells in a multi-gas head. The same logic applies to fixed gas detector builds where the EC cell is one channel in a 4–8 channel controller.
Comparison: EC vs Catalytic-Bead vs NDIR vs PID vs MOS
EC vs catalytic-bead (pellistor): EC cells are ppm-accurate for specific toxic gases at low concentration but are poisoned by silicone, H2S bursts and sustained high humidity; catalytic beads give %LEL response for most flammables regardless of gas identity but cannot measure below ~1% LEL and burn out in high gas concentrations. EC vs NDIR: NDIR is non-consumptive and stable for years on CO2 and hydrocarbons, while EC cells are 12–24 month consumables; the trade is selectivity versus consumable cost. EC vs PID: PID is the only realistic option for ppb-level VOCs (benzene, isocyanates) but is non-selective and humidity-sensitive, whereas EC is gas-specific at ppm level. EC vs MOS: metal-oxide semiconductor sensors are cheap and respond to a wide range of reducing gases, but baseline drift and humidity cross-sensitivity make them unsuitable for safety-of-life toxic-gas alarms where EC or PID are preferred [S1].
For pricing context on the catalytic-bead and NDIR alternatives, see the catalytic bead and certification cost breakdown and the infrared gas detector pricing benchmarks referenced for the 2026 procurement cycle. Where the application is a multi-gas confined-space entry, the multi-gas detector reference outlines channel count, battery chemistry and bump-test cadence.
Cross-Sensitivity, Interference Gases and Environmental Limits
Every EC cell has a published cross-sensitivity table: a typical H2S cell will respond to NO, NO2 and SO2 at 10–30% of the target-gas signal, while a CO cell will read a positive signal on H2 and on most light hydrocarbons [S1]. Temperature and humidity envelopes are typically 0–50 °C and 15–90% RH non-condensing; below 0 °C the electrolyte viscosity rises and the cell slows, above 50 °C the electrolyte dries and the baseline drifts [S1]. Pressure variation follows a partial-pressure law, so a fixed-gas EC transmitter needs a pressure-compensation input or a controlled-vent housing if the process runs above 1.1 bara.
Mounting orientation and wind velocity also matter: a 0.5 m/s wind on a diffusion-sampled EC head can starve the capillary and under-read by 10–20%, which is why sensor manufacturers recommend a calibration wind shield or a pumped sample line for outdoor fixed installations [S1].
Calibration, Bump Test and Sensor Replacement Cycle
Field calibration uses a span gas of certified concentration (typically 50% of full scale for CO and H2S sensors) and zero air, with a 60–90 second response time T90 for most EC cells and a recommended 30–60 second bump test using a concentration at or above the low alarm setpoint [S1]. Calibration interval is commonly 30–90 days for fixed safety-critical toxic-gas transmitters and 180 days for non-safety process measurement, with a sensor end-of-life flag driven by the on-board electronics once the cell's residual sensitivity falls below ~50% of new-cell output [S1].
Replacement strategy is the biggest lifecycle-cost lever: a 24-month cell at 60 USD per replacement on a 200-instrument fleet is ~24,000 USD/year in consumables, which is why some operators standardise on field-replaceable smart sensors with automatic calibration constants read over HART or Modbus rather than hard-wired analog cells [S1][S2].
Standards, Hazardous Areas and Sourcing Signals
For flammable-target co-monitoring, ATEX 2014/34/EU and IECEx equipment certification mark the build, and hazardous-area classification drives the housing choice between Ex d (flameproof) and Ex i (intrinsically safe) — for a toxic-only application on a non-flammable site, neither is mandatory and a general-purpose IP65 head with a 4–20 mA output is sufficient [S1]. Performance verification against EN 45544 or equivalent local toxic-gas standards is the specifier's responsibility, and the manufacturer's certificate of calibration should be requested per cell serial number at procurement [S1].
Sourcing signals to track: cell-pricing pressure on Chinese suppliers reported in 2026 first-half industrial channels suggests a 5–10% year-on-year softening of replacement-cell list prices, and the major portable-detector vendors are converging on Bluetooth-enabled bump-test printers as standard accessories rather than options [S2]. For a deeper spend analysis on the multi-gas head, see the multi-gas detector cost and certification guide reference compiled from the 2026 supplier base.