NDIR is an industry term for "nondispersive infrared", and is the most common type of sensor used to measure carbon dioxide, or CO2.
An infrared (IR) lamp directs waves of light through an tube filled with a sample of air toward an optical filter in front of an IR light detector. The IR light detector measures the amount of IR light that passes through the optical filter.
The 4.2 micron band of IR radiation produced by the lamp is very close to the 4.26 micron absorption band of CO2. Because the IR spectrum of CO2 is unique, matching the light source wavelength serves as a signature or "fingerprint" to identify the CO2 molecule.
As the IR light passes through the length of the tube, the CO2 gas molecules absorb the specific band of IR light while letting other wavelengths of light pass through. At the detector end the remaining light hits an optical filter that absorbs every wavelength of light except the 4.2 micron wavelength absorbed by CO2 molecules in the air sample tube.
Finally, an IR detector reads the remaining amount of 4.2 micron light that was not absorbed by the CO2 molecules or the optical filter.
How is the CO2 level calculated?
The difference between the amount of light radiated by the IR lamp and the amount of IR light received by the detector is measured. Since the difference is the result of the light being absorbed by the CO2 molecules in the air inside the tube, it is directly proportional to the number of CO2 molecules in the air sample tube.
All measurements start out as analog micro-voltages. While some sensors output this as an analog voltage or 4-20mA signal, some also include an analog to digital converter on the sensor PCB that converts the voltages into serial or RS-485 output. Serial output is especially useful for using NDIR CO2 sensors with Arduino or Raspberry Pi micro controllers.
Early CO2 measurement devices
Since Joseph Black, a Scottish chemist and physician first identified carbon dioxide in the 1750s, scientists have looked for a way to measure this common gas. One of the earliest CO2 measurement devices was a mercury manometer. Manometers use a U-shaped glass tube filled with mercury to measure gas pressure. If temperature, pressure and volume of a dry gas sample containing CO2 molecules are known, the moles of CO2 can be calculated using the Ideal gas law (PV=nRT).
While mercury manometers can be very accurate, the procedure for measuring CO2 levels in air samples can take hours. That’s why when Charles Keeling was asked by the US Weather Bureau to begin taking hourly atmospheric CO2 measurements on the Mauna Loa volcano in Hawaii he used an early infrared (IR) gas analyzer calibrated against his manometer. The original Applied Physics Corp. Infrared Gas Analyzer operated on Mauna Loa from 1958 until 2006.
Like every IR gas sensor, the analyzer at Mauna Loa used the same basic principle for measuring CO2. It has an infrared light radiation source at one end of a gas sample tube and an IR detector at the other. The 4.2 micron band of infrared radiation is very close to the 4.26 micron absorption band of CO2. As a result, the amount of light radiation that is absorbed by the CO2 molecules is proportional to the amount of carbon dioxide in the gas sample. However, since low levels of CO2 do not absorb much light, a long tube is needed before the effect can be measured.
While the original IR gas analyzer was accurate, it was bulky. The sample tube alone was 40cm (16 inches) long.
The challenge is getting the right balance. Engineers use a longer optical path to measure lower levels of CO2 more accurately - which means a larger gas sample chamber. On the other hand, in ambient air environments - such as schools, offices and homes - the demand is for ever-smaller sensors to fit neatly inside compact devices.
Small NDIR CO2 sensor breakthrough
An engineering breakthrough occurred in 1993 when SenseAir AB patented a design for small-footprint CO2 sensors. These sensors solved the size problem by using folded optics and metallized molded plastic to reflect the light through a curved shape (refered to as a 'waveguide') that was longer than the footprint of the sensor module. The highly reflective coating insured that the CO2 molecules inside the gas sample chamber would absorb the same amount of light as a traditional straight-path design.
By using advanced optics, new waveguide designs allowed for progressively smaller sensors with increased sensitivity. For example, in 2003 SenseAir's K20 CO2 sensor (shown above) used a "banana" waveguide design. This sensor was used for many years in OEM consumer safety products.
Modern NDIR CO2 sensors
Today's newest generation of CO2 sensors have even more optimized waveguides, allowing a longer optical light path to be folded into an even smaller 8 mm x 33 mm x 20 mm footprint.
In addition to smaller size, new mid-range infrared light-emitting diode (LED) light sources have been developed that allow NDIR sensors to operate at much lower power levels. Combined with a photo-diode light detector, these solid-state sensors also provide a much longer life-span.
For example, the CozIR®-LP Miniature 5,000ppm CO2 Sensor shown here consumes 3mW of power, while typical incandescent IR sensors consume 50 to 200mW. This low power consumption makes the new generation of NDIR CO2 sensors capable of running for months on battery or solar power alone.
How NDIR CO2 sensors are made
The fabrication of CO2 sensors resembles that of smart phones or any other state-of-the-art electronics. High-speed robotic assembly takes place in specially designed clean rooms. The modules are then batch-tested and calibrated before shipment. SenseAir AB has a promotional video where you can see sensors being assembled at their factory in Sweden here.
CO2 sensors in the future
What is in store for the future? For years, high-end gas analyzers have been produced based on photo-acoustic spectroscopy, or PAS. PAS is was discovered by Alexander Graham Bell in the 1880’s. He noted that strobed sunlight shown on different materials produced an audible sound. Since PAS systems do not rely on the length of the waveguide, an even smaller CO2 sensor using a pulsed MEMS (Micro Electro Mechanical System) mirror and MEMs microphone are theoretically possible. When they go into production, we'll tell you about them here.
Images courtesy of SenseAir® AB