VOC Gas Sensors for Industrial AQMS: Photoionization Detectors (PID) vs NDIR vs Metal Oxide Semiconductors (MOS)

Pick the wrong voc gas sensors for your fenceline AQMS and you will either miss a real fugitive leak or drown your control room in false alarms. Neither outcome is acceptable. Volatile Organic Compounds (VOCs) are carbon-based pollutants with high vapor pressures at ambient temperature, and the aromatic subset (benzene, toluene, ethylbenzene, xylene: BTEX) carries acute toxicity, proven carcinogenicity, and drives ground-level ozone plus secondary particulate formation.

Regulations set hard limits on what leaves your property line. Indonesian law (Permen LHK No. 14/2020) makes non-methane hydrocarbons (NMHC) a mandatory parameter inside the national Air Pollutant Standard Index (ISPU). PP No. 22/2021 adds a 3-hour ambient ceiling of 160 µg/m³. On the occupational side, Permenaker No. 5/2018 caps benzene workplace exposure at 0.5 ppm (8-hr TWA) and toluene at 20 ppm.

So the real engineering question is straightforward: which transducer technology do you bolt onto that sampling mast? Three options exist: Photoionization Detectors (PID), Non-Dispersive Infrared (NDIR), and Metal Oxide Semiconductors (MOS). Each one operates on a different physical principle. Spec the wrong one and you are either blind to a benzene plume or chasing phantom readings at 2 a.m.

Transducer Physics: Photoionization, Infrared Absorption, and Semiconductor Redox

Each of these three sensor types converts organic gas molecules into an electrical signal through a different physical mechanism.

Photoionization Detectors (PID)

A PID fires high-energy UV photons from a vacuum-discharge lamp (typically krypton gas, 10.6 eV) at incoming air. Any gas molecule with an Ionization Potential below that photon energy ($IP < 10.6\text{ eV}$) gets stripped of an electron without being destroyed:

$$RH + h\nu \rightarrow RH^+ + e^-$$

An internal electric field sweeps positive ions ($RH^+$) and free electrons ($e^-$) toward collector electrodes. The resulting ion current sits in the pico-ampere ($pA$) range, stays linear from 1 ppb out to several thousand ppm, and responds fast ($t_{90} < 3$ seconds). Here is the practical payoff: nitrogen ($IP = 15.58\text{ eV}$), oxygen ($IP = 12.07\text{ eV}$), and methane ($IP = 12.61\text{ eV}$) all sit well above 10.6 eV. A standard PID lamp physically cannot ionize those background gases, so methane does not interfere with your reading.

Metal Oxide Semiconductors (MOS)

MOS sensors run hot on purpose. A ceramic substrate carries a thin-film metal oxide (usually tin dioxide / $\text{SnO}_2$ or tungsten oxide / $\text{WO}_3$) heated to 200°C through 400°C. At that temperature, atmospheric oxygen grabs free electrons from the semiconductor’s conduction band, creating a high-resistance depletion layer. When a reducing VOC molecule hits that catalytic surface, it reacts with the adsorbed oxygen ions and dumps electrons back into the conduction band. The net effect is a non-linear resistance drop ($\Delta R \propto C^n$). Simple in theory, messy in practice, as we will see in the humidity section below.

Non-Dispersive Infrared (NDIR)

NDIR measures hydrocarbons by infrared absorption along characteristic vibrational bands, following Beer-Lambert:

$$A = \varepsilon \cdot c \cdot L$$

A broadband IR emitter pushes radiation through a gas cell onto a thermopile or pyroelectric detector behind a narrow bandpass filter tuned to the aliphatic $C-H$ stretch at roughly 3.3 to 3.4 µm. NDIR gives you rock-solid long-term stability on straight-chain alkanes. But it absorbs complex aromatics poorly and generally cannot reach the sub-ppm sensitivity needed for ambient trace-level work.

Response Correction Factors (CF) and Compound Selectivity

A PID is a broadband total-VOC (TVOC) instrument. It is calibrated against one reference gas: isobutylene ($C_4H_8$). Every ionizable species in the sample contributes to the same ion current, meaning a PID cannot tell benzene from toluene on its own. For discrete compound identification, you need a GC separation column upstream.

To back-calculate actual concentrations of a known target compound from a raw PID reading, you apply a chemical-specific Correction Factor (CF), referenced to isobutylene ($CF = 1.0$):

$$\text{Actual Target Gas Concentration} = \text{PID Display Reading} \times \text{Correction Factor (CF)}$$

Reference CFs for common industrial VOCs using a standard 10.6 eV UV lamp:

Target VOC Chemical	Molecular Formula	Ionization Potential (IP)	Correction Factor (10.6 eV Lamp)	Occupational Limit (Permenaker 5/2018)
Isobutylene (Span Gas)	$\text{C}_4\text{H}_8$	9.24 eV	1.00 (Calibration Datum)	–
Benzene	$\text{C}_6\text{H}_6$	9.25 eV	0.53	0.5 ppm (8-hour TWA)
Toluene	$\text{C}_7\text{H}_8$	8.82 eV	0.50	20 ppm (8-hour TWA)
o-Xylene	$\text{C}8\text{H}{10}$	8.56 eV	0.43	100 ppm (8-hour TWA)
n-Hexane	$\text{C}6\text{H}{14}$	10.13 eV	4.30	50 ppm (8-hour TWA)
Formaldehyde	$\text{CH}_2\text{O}$	10.88 eV	Not Detected (IP > 10.6 eV)	0.1 ppm (Ceiling Limit)
Methane	$\text{CH}_4$	12.61 eV	Not Detected (IP > 10.6 eV)	Simple Asphyxiant

When CF drops below 1.0, the compound ionizes more readily than isobutylene. Higher CFs mean weaker ionization efficiency. And if the compound’s IP exceeds the lamp’s photon energy, the sensor simply will not see it at all.

Tropical Field Conditions: Humidity Quenching and Baseline Drift

Running continuous gas monitors in the tropics means dealing with relative humidity that routinely sits between 80% and 95% RH. That moisture wreaks havoc on sensor performance in ways temperate-climate test labs rarely simulate.

Humidity Quenching in PID Cells

Heavy water vapor absorbs 10.6 eV UV photons and physically collides with ionized VOC molecules before they reach the collector plates. The result: signal attenuation (humidity quenching) that can suppress real TVOC readings by 30% to 50% during rain or early-morning condensation events.

It gets worse. Water droplets bridging the collector electrodes create leakage currents that show up as false-positive zero-baseline spikes. Industrial-grade PID transmitters fight this with patented fence electrode geometry (diverts surface tracking currents) and built-in micro-heaters that keep the internal cell above the ambient dew point.

Cross-Sensitivity and Severe Drift in MOS Transducers

MOS sensors in the tropics are a headache. The heated $\text{SnO}_2$ crystal lattice competes with water molecules for the same surface adsorption sites meant for target gases. When humidity swings from dry afternoon air to saturated night air, the baseline drifts wildly. In field experience, an uncompensated outdoor MOS unit can register what looks like a fugitive emission spike when nothing is actually leaking. That kind of ghost alarm erodes operator trust fast.

Engineering Comparison Matrix: PID vs NDIR vs MOS

Picking the right transducer means weighing detection limits against response speed, power draw, and maintenance burden:

Metrological Parameter	Photoionization Detector (PID)	Metal Oxide Semiconductor (MOS)	Non-Dispersive Infrared (NDIR)
Detection Physics	Vacuum UV photon ionization	Surface redox film resistance modulation	Infrared absorption spectroscopy
Limit of Detection (LOD)	1 ppb to 10 ppb (High Sensitivity)	50 ppb to 500 ppb (Moderate Sensitivity)	1 ppm to 10 ppm (Coarse Trace Sensitivity)
Measurement Range	1 ppb to 2,000+ ppm	0.1 ppm to 500 ppm	10 ppm to 50,000 ppm (% LEL)
Response Time ($t_{90}$)	Rapid (< 3 seconds)	Moderate (15 to 30 seconds)	Fast (5 to 10 seconds)
Gas Selectivity	Broad-spectrum TVOC ($IP < h\nu$)	Non-selective (Cross-sensitive to $H_2, CO, RH$)	Selective to aliphatic $C-H$ bonds
Humidity Sensitivity	Quenching mitigated by fence electrodes	Extreme baseline drift and noise	Immune to moisture absorption
Electrical Power Draw	Moderate (100 mW to 300 mW)	High (500 mW to 1,000 mW for heater)	Low (50 mW to 200 mW)
Sensor Operating Life	12 to 24 months (Lampu UV & electrode stack)	24 to 36 months (Stable oxide film)	> 5 to 10 years (Durable optical source)
Total Cost of Ownership	Moderate (Requires periodic cleaning & span calibration)	Low initial capex, high validation cost	Low (Minimal field maintenance)
Target Application	Fenceline AQMS & perimeter surveillance	Qualitative indoor screening & trends	Hydrocarbon LEL & explosive gas safety

Decision Framework: Which Sensor for Which Job

The right pick depends on what you are monitoring and what regulations apply:

1. Fenceline AQMS at Petrochemical and Chemical Plants: You need sub-ppm PID transmitters with 10.6 eV lamps, fence electrodes, and active RH compensation. There is no substitute here: only a PID catches fugitive aromatic plumes at the concentrations that matter before they cross the property boundary.
2. Explosion Hazard Perimeters and Pipeline Leak Detection: NDIR is the correct choice. Its optical detection path cannot be poisoned by reactive chemicals, and it delivers stable readings on high-concentration flammable gases up to LEL range.
3. Indoor Screening and Ventilation Trending: Low-cost MOS sensors can work for qualitative trending, but do not submit that data for regulatory compliance. Period.

Fortuna Argatech Industrial AQMS Integration

Fortuna Argatech builds field-ready Air Quality Monitoring Systems (AQMS) around humidity-compensated PID sensors, the GEOVOS 1000 Datalogger, and multi-protocol telemetry feeding centralized cloud dashboards.

The sampling train pulls ambient air through an active aspirating manifold built entirely from inert PTFE (Teflon) or 316 stainless steel. No adsorption on inlet walls. Air passes through a PTFE particulate filter before entering the temperature-stabilized PID cell.

On the data side, raw analog outputs go through multi-point polynomial calibration on the GEOVOS 1000, which cross-references live temperature and humidity readings to strip out quenching artifacts. The system runs multi-tier alarm logic: one threshold for early operational leak warnings, another for hard regulatory exceedance trips. Those alerts feed directly into monitoring sensor alarm threshold configurations and sensor data monitoring dashboard designs that trigger site emergency protocols.

Field QA is straightforward: certified isobutylene span gas and zero-air scrubbers keep your calibration honest. The bottom line is that pairing a precision PID with an inert, climate-compensated sampling station gives you defensible, high-uptime ambient air quality data across the full operating lifecycle of an industrial site.