Skip to content
AkerminMeasure. Monitor. Improve. Compare Products
Learning Center

NDIR vs Metal Oxide Sensors: Which CO₂ Monitors Actually Measure CO₂

Two sensor technologies share the label "CO₂ monitor". One measures the gas; the other guesses from unrelated chemicals. Here is how to tell them apart on a product page.

NDIR sensors measure CO₂ directly by infrared absorption. Metal-oxide (MOX) sensors measure volatile organic compounds and, in some products, convert that signal into an “equivalent CO₂” (eCO₂) estimate. The estimate assumes people are the only VOC source, so hand sanitizer, cooking or new furniture makes it read thousands of ppm high. For ventilation decisions, only NDIR (including photoacoustic NDIR) counts as a measurement.

How each works

NDIR

Infrared light at 4.26 µm passes through a chamber; CO₂ absorbs it in proportion to concentration; a detector measures what gets through. The result is specific to CO₂ because almost nothing else in room air absorbs that wavelength. See how CO₂ monitors work.

Metal oxide

A heated tin-oxide surface changes resistance when reducing gases (hundreds of organic compounds, hydrogen, carbon monoxide) react on it. The sensor cannot tell ethanol from acetone from breath. Firmware turns the resistance into a VOC index and, in “eCO₂” products, into a fictional CO₂ number based on the assumption that VOC and CO₂ rise together when people are the source.

Why eCO₂ fails

Event Real CO₂ Typical eCO₂ reading
Applying hand sanitizer nearby Unchanged Jumps to 2,000 to 5,000 ppm for minutes
Cooking with oil Slight rise Large rise, slow to fall
New furniture or paint Unchanged Elevated for days
Full meeting room, no VOC sources 1,500 ppm Often reads lower than reality; the sensor drifts to call the current air “baseline”

The last row is the dangerous one: MOX algorithms continuously reset their baseline, so a room that is always stuffy gets reported as fine.

Spotting the technology on a product page

  • Says NDIR or photoacoustic, or names a module (Senseair, Sensirion SCD30/40/41, Cubic, Winsen MH-Z19): measures CO₂.
  • Says eCO₂, CO₂eq, “CO₂ equivalent”, TVOC/CO₂: estimates.
  • Ten sensors in a tiny cheap box: the CO₂ channel is almost certainly estimated; a real NDIR module alone costs more than such devices sell for.
  • No mention of the sensor at all: assume estimated until proven otherwise.

Where MOX sensors are useful

As a VOC trend indicator, MOX sensors are genuinely informative. A rising VOC index when nobody is cooking points to cleaning products, off-gassing or a spill, and combination monitors such as the Airthings View Plus include one alongside a real NDIR CO₂ sensor for exactly that purpose.

Bottom line

Buy NDIR for CO₂. If you want VOC awareness too, buy a monitor that has both sensors and labels them honestly. The best CO₂ monitors guide only includes NDIR and photoacoustic models.

Frequently Asked Questions

How can I tell if a monitor uses NDIR?

The listing or datasheet will say NDIR, photoacoustic, or name a module such as Senseair S8, Sensirion SCD30/SCD40/SCD41 or Cubic CM1106. If it says eCO₂, CO₂eq, TVOC-based, or does not mention the sensor, assume it is an estimator.

Are metal-oxide sensors useless?

No. They are decent at showing that VOC levels changed, which is useful for spotting cleaning products, cooking and off-gassing. They are simply the wrong tool for CO₂.

Are all NDIR sensors equally good?

Dual-channel NDIR sensors with a reference wavelength hold calibration better than single-channel ones. Photoacoustic sensors such as the SCD4x are compact and accurate but rely on automatic baseline correction, which needs the room to reach outdoor levels regularly.