In air quality monitoring, building ventilation, fresh air control, and various environmental sensing devices, NDIR and MOS are two of the most frequently compared gas sensing technologies. From a strict engineering perspective, however, they are not two equivalent solutions that can simply replace one another. If the discussion is centered on the direct measurement of actual CO₂ concentration, then the two differ fundamentally in sensing principle, output meaning, long-term stability, and application boundary.
1. NDIR and MOS Do Not Solve Exactly the Same Problem
NDIR (Non-Dispersive Infrared) is based on the absorption of specific infrared wavelengths by the target gas. In the case of CO₂, the sensor detects the attenuation of infrared light after it passes through the sampled gas, and then converts the signal into a corresponding CO₂ concentration value through a calibration model. The core value of NDIR therefore lies in one key fact: it measures CO₂ directly as the target gas itself.
MOS (Metal Oxide Semiconductor), by contrast, operates through a completely different mechanism. It relies on the sensing layer interacting with surrounding gases under heated conditions through adsorption, oxidation, or reduction, which changes the device resistance. These changes are then interpreted by algorithms. For this reason, MOS is better suited to reflecting overall changes in reactive gases or pollutants in the environment, rather than delivering exclusive quantitative measurement of one specific gas.
This is exactly why, when the goal is to determine the actual indoor CO₂ level in ppm, NDIR and MOS are not truly competing on the same level. More precisely, NDIR is closer to measuring CO₂ itself, while MOS is closer to sensing changes in overall air quality condition.
2. Differences in Principle Lead Directly to Differences in Selectivity
For CO₂ detection, selectivity is one of the most critical technical indicators. The reason NDIR has become the mainstream CO₂ solution lies in the fact that it uses CO₂ absorption characteristics at specific infrared wavelengths. When the optical structure, filtering design, and algorithm compensation are properly implemented, the system can map energy changes at the relevant wavelength band to CO₂ concentration changes with strong stability, making the output both more targeted and more interpretable.
MOS, on the other hand, responds to a wide range of VOCs, odor molecules, and reducing or oxidizing gases. Its strength lies in sensitivity, especially for detecting changes in airborne pollutants, odor, and volatile organic compounds. In a strict technical sense, however, it does not operate around the exclusive spectral signature of one specific gas, so its selectivity for single-gas CO₂ detection is generally not as strong as NDIR.
In practical engineering terms, this means: if the task is to determine whether CO₂ in a meeting room, classroom, office building, or greenhouse has risen above a defined threshold, NDIR is typically the more suitable choice. If the objective is to sense kitchen odor, cleaning agent emissions, furniture off-gassing, alcohol, perfume, or broader air pollution changes, MOS may instead offer higher sensitivity.
3. Why Do Many MOS-Based Products Also Display “CO₂”? The Key Is eCO₂
Many MOS-based air quality modules on the market also display a “CO₂” or “CO₂ equivalent” value on the interface. This is one of the most common sources of misunderstanding. In many cases, this value does not represent a direct measurement of CO₂ by the sensor itself. Instead, it is an eCO₂ (equivalent CO₂) value estimated by algorithms based on the empirical relationship between VOC variation and human exhaled air patterns.
From an engineering semantics perspective, eCO₂ should be understood more appropriately as an estimated indicator of environmental condition, rather than being treated as actual CO₂ ppm. For example, when alcohol, perfume, cleaning agents, cooking vapors, or renovation emissions are present, the MOS response may change significantly. That does not necessarily mean real CO₂ concentration has increased at the same time.
Therefore, if the application requires precise ventilation control, indoor CO₂ threshold alarm, classroom or meeting room air exchange judgment, or any other closed-loop system logic that depends on actual CO₂ data, it is essential to distinguish clearly between the two: NDIR measures CO₂ directly, while many MOS outputs are eCO₂ estimates based on VOC response.
4. Long-Term Stability: Why NDIR Is Better Suited to Continuous CO₂ Monitoring
From the perspective of long-term operation, another advantage of NDIR is that it does not rely on ongoing chemical reactions between the sensing layer and the target gas to complete the measurement process. For continuous CO₂ monitoring, this means the system is generally better able to maintain stable and interpretable output logic. Modern NDIR modules are also commonly supported by automatic baseline calibration, forced calibration, temperature and humidity compensation, and pressure correction mechanisms to further improve long-term consistency.
MOS, by contrast, depends more heavily on sensing material condition, micro-heater operating mode, environmental history, and algorithm modeling. It is highly valuable for recognizing air quality trends, but when the requirement is long-term, continuous, traceable quantitative monitoring of real CO₂, the output meaning of MOS is usually less direct than that of NDIR.
This is also why in HVAC systems, fresh air systems, building ventilation control, greenhouse CO₂ management, and people-occupancy monitoring in classrooms or meeting rooms, NDIR is more likely to remain the long-term mainstream solution.
5. MOS Also Has Clear Strengths: Miniaturization, Low Power, and Comprehensive Air Sensing
That does not mean MOS lacks advantages. On the contrary, MOS is highly competitive in many product formats. It typically offers higher integration, smaller size, and lower power consumption, making it easier to integrate into portable devices, IoT modules, smart home terminals, and consumer electronics.
In addition, MOS often provides stronger sensitivity in detecting VOCs, odor, pollution changes, and abnormal air conditions. It is also well suited to algorithm-based interpretation of broader questions such as whether air quality is deteriorating or whether unusual odor conditions have emerged.
From a product definition perspective, MOS should therefore not be viewed as a lower-tier substitute for NDIR. It is better understood as a route more suitable for comprehensive air sensing. The point is simply that when the requirement becomes specifically focused on actual quantitative CO₂ measurement, the technical meaning of NDIR is more precise.
6. How to Choose by Application: First Ask What You Actually Need to Measure
One of the most common selection mistakes is to assume that because a sensor can output a ppm value, it is automatically suitable for CO₂ control. In reality, a truly professional selection logic should begin by answering one key question: Are you trying to measure CO₂ itself, or changes in air pollution condition?
If your application involves building HVAC, fresh air systems, indoor air quality monitoring, greenhouse control, or demand-controlled ventilation in meeting rooms and classrooms, then what you need is actual CO₂ concentration, and such requirements generally prioritize NDIR.
If your application involves air purifiers, smart home devices, portable equipment, odor recognition, VOC trend analysis, or abnormal volatile pollutant detection, MOS often provides stronger advantages.
7. A Practical Judgment Standard That Rarely Leads You Wrong
If you are reviewing product materials, writing technical content, or selecting a solution, and you want a fast way to judge the positioning of the two technologies, the following simplified logic is useful:
When you see NDIR, first associate it with: direct CO₂ measurement, strong selectivity, long-term continuous monitoring, and suitability for ventilation control.
When you see MOS, first associate it with: VOC, odor, and comprehensive air quality sensing, smaller size, lower power, and greater algorithm flexibility, but not necessarily direct CO₂ measurement.
When you see eCO₂, the next question should always be: Is it a direct measurement of actual CO₂, or an equivalent estimate derived from VOC response?
Conclusion
The core difference between NDIR and MOS is not simply that one is more expensive and the other more affordable, or that one is larger while the other is smaller. The more fundamental distinction is this: NDIR performs direct measurement based on the spectral absorption of the target gas, while MOS performs comprehensive sensing based on the chemical response of the sensing layer to multiple gases.
For that reason, when the specific issue is CO₂, the two should not be treated as equivalent solutions that can simply replace each other. If the goal is actual CO₂ concentration, long-term stability, and suitability for closed-loop ventilation control, NDIR is generally the more rigorous choice. If the goal is VOCs, odor, air condition changes, and low-power compact integration, MOS is often the more flexible option.
Professional product selection does not begin with the question of which technology is more advanced. It begins with a much more important question: What is it that you actually want to measure?
