Current statistics, such as those of the US Environmental Protection Agency (EPA), show that people spend almost 90 percent of their time indoors, while the concentrations of some pollutants indoors are often 2 to 5 times higher than typical outdoor concentrations . CO2 concentration is a key indicator of air quality. At this point it is worth noting that about 140 years ago, Max von Pettenkofer laid the foundations for current regulations relating to air quality with his studies on carbon dioxide levels. The higher the CO2 value in a building, the less comfortable it becomes for the people inside. In poorly ventilated rooms, the CO2 concentration increases rapidly. For example, in a space of about four-square meters occupied by only one person, the CO2 value rises onwards from 500 ppm (0.05 percent) to over 1,000 ppm (0.1 percent) in just 45 minutes. At this level, the odorless and colorless gas can cause headaches, drowsiness and poor concentration, often resulting in reduced productivity. From 2,000 ppm onwards (0.2 percent), even the cognitive abilities of humans are influenced, and there is a significant risk to health at higher levels (Figure 1.).
However, there are other health risks connected with the indoor CO2 concentration. If there is a high amount of exhaled CO2 in the air, there is also a high number of aerosols. A high concentration of aerosols increases the risk of infection for everyone else in the room. Especially in times of Covid-19 this becomes crucial in offices, schools, shops and the like. When a person infected with the coronavirus coughs, speaks or sneezes, a spray consisting of droplets and aerosols is generated, which penetrates the room air and then spreads. In the paper “Aerosol transmission of SARS-CoV-2” , several researchers have shown that poorly or non-ventilated indoor spaces can increase the likelihood of aerosol transmission of Covid-19. Insufficient ventilation may lead to a long-range airborne transmission of the virus and opportunistic infection. A study of the TU Berlin  revealed that the indoor climate plays a key role in health protection, as pathogens remain in rooms for hours at typical air exchange rates in residential and office buildings (Figure 2). The sink rate and process of air renewal take a considerable time. As such, any increase in the supply of fresh air is to be recommended. To monitor and control the air quality, innovative CO2 sensors like the new XENSIVTM PAS CO2 (Figure 3.) from Infineon provide precise, cost-effective and space-saving solutions. These can optimize the air quality in rooms for more healthy and productive indoor living and working conditions.
Teaching and office work during the coronavirus pandemic raise concerns about aerosols and the risk of infection. Wherever you have a large number of people in a room, there is a considerable amount of exhaled air which contains CO2 , carbon dioxide. The Federal Environment Agency of Germany and the ASHRAE (American Society of Heating, Refrigeration and Air Conditioning Engineers) published recommendations long before the coronavirus outbreak: In classrooms and offices, the CO2 concentration should not exceed 1,000 ppm – By way of comparison: in the fresh air outside, the CO2 concentration is 400 ppm. In this context, the idea of installing CO2 measuring devices in classrooms and offices, as well as in other indoor public spaces like gyms, bars and restaurants, is about preventing the spread of the virus. It would of course be possible to measure aerosols in the air and sound the alarm if these become excessively high. But such measuring devices are complex and expensive. On the other hand, inexpensive and compact CO2 measuring devices are now available that can warn against high concentrations of CO2 in the air, and thus also against high levels of aerosols. These could be used to indicate a potential increased risk of infection with coronavirus.
Reliable CO2 monitoring is not only important with regard to Covid-19, but also beneficial to overall well-being when spending time indoors and to productivity. CO2 sensors can be used to measure the carbon dioxide concentration and thus the quality of indoor air (Figure 4). But in order to improve the surrounding air quality, and consequently, increase indoor comfort and people productivity, more reliable and affordable CO2 sensors are required. Currently there are two options: Sensors/devices that are accurate but bulky and expensive, or sensors that are small but inaccurate, providing grossly estimated values unsuitable for proper control. The XENSIV™ PAS CO2 sensor, in contrast, is ideal for a broad spectrum of applications, providing precise results in a compact format.
Energy and cost savings through air control
A ventilation system not only benefits human well-being. Effective air control in residential and commercial buildings can save energy, which at the same time reduces the corresponding costs and CO2 emissions. Private households also benefit from an upgrade. In the United States, families spend an average of about US 2,000 Dollars per year on energy. With a suitable upgrade, they could save about US 400 Dollars per year. However, other sectors could also benefit from the use of air quality control based on reliable CO2 measurements. Schools, hospitals, restaurants and shops also have high energy requirements and associated high expenses. In total, countries like the USA spend more than 400 billion Dollars every year on supplying all buildings with energy. They use about 74 percent of the electricity generated in the USA, which accounts for about 40 percent (Figure 5) of the country’s total energy expenditure. With effective building automation that also controls ventilation, known as demandcontrolled ventilation (DCV), the power consumption of US buildings could be reduced by up to 20 percent, saving about 80 billion Dollars a year in energy costs .
If buildings are properly planned, constructed and operated from the outset, for example with demand-controlled ventilation (DCV), the energy efficiency of buildings can be increased by up to 30 percent  of the HVAC (Heating, Ventilation and Air Conditioning) energy bill. This would in turn make it possible to reduce the overall energy demand, which would solve the global problem of energy shortages on the one hand and reduce the threat to the environment on the other. One typical example: A US school with an average size of about 7,000 m²  shows a yearly HVAC energy consumption around 5.6 USD/m² . Assuming 20 percent energy efficiency based on DCV, the saving is around 8,000 USD per year. This translates to savings of 80,000 kWh (assuming 10 cent/kWh), which translates to 35 tons CO2 emission reduction. This is equivalent to the yearly CO2 consumption of a forest with 1,600 trees.
Possible applications for CO2 sensors
The data measured by CO2 sensors can be used in many ways. In a DCV, HVAC systems use the values to automatically adjust the air mixture in the room to that of the outside air according to the targeted application requirements. This keeps indoor carbon dioxide con-centration at a specific value, for instance below 1,000 ppm according to the ASHRAE standard on ventilation and acceptable indoor air quality in residential buildings. Given the benefits of CO2 sensing with regard to health and mitigation of infection risk, one can expect widespread adoption of CO2 sensors in classrooms, offices, gyms and bars, where the sen-sors will detect bad air quality. One example is the so-called CO2 traffic light, which warns the occupants of a high CO2 level and therefore a high concentration of aerosols, which is a clear signal to air the room. These sensors can be organized in a sensor network connected to cloud solutions for data intelligence and remote access.
There are many other potential uses for CO2 sensors. Small CO2 sensors are suitable e.g. for smart home assistants and IoT devices like air purifiers or thermostats. Other applications could follow in the future, such as infant monitoring, food quality control, fitness tracking or agriculture.
Limitations of existing CO2 sensor solutions
NDIR sensors (non-dispersive infrared sensors) are often used in building automation nowadays. They consist of an IR light source, a sample chamber, a spectral filter and reference and absorption IR detectors, which is why they are relatively large and expensive. Although they provide true and accurate CO2 measurements, their form factor makes them difficult to integrate, making them unsuitable for installation in small IoT devices or smart home components.
The various indoor pollutants can also be detected by the so-called eCO2 sensors, but unlike the NDIR sensor, they do not perform real measurements. Instead, they use algorithms to calculate an equivalent CO2 value. These sensors deliver estimated values based on many assumptions, e.g. based on how many people – as the cause of an increasing CO2 load – are present. With this method, the air quality is not always improved at the right moment, which means that the climate control system consumes an unnecessarily large amount of energy. There are currently no comparable solutions available on the market that both provide accurate and true CO2 measurements and are small and cost effective.
CO2 sensor with photoacoustic spectroscopy Thanks to its experience and leading position in MEMS technology, Infineon has succeeded in developing a new CO2 sensor based on photoacoustic spectroscopy (Figure 6), or PAS for short. The PAS method is based on the photoacoustic effect first discovered by Alexander Graham Bell in 1880. Infineon owns a comprehensive and continuously growing patent portfolio on the PAS technology, from sensor design to system implementation. The method uses the fact that gas molecules only absorb light with a specific wavelength. In the case of carbon dioxide, for example, the wavelength is 4.2 µm. In rapid succession, light – i.e. energy – is supplied to the gas in exactly this wavelength via an infrared source with an optical filter. Due to rapid heating and cooling, this leads in turn to thermal expansion and contraction, generating a pressure change that can be recorded by an acoustic detector optimized for low frequencies. The signal is then evaluated and used to draw conclusions about the amount of CO2 . The stronger the signal, the higher the CO2 concentration. . The highly sensitive MEMS acoustic detector, which acts as a pressure sensor, is used as a detector, allowing for significant miniaturization.
Architecture and challenges The Infineon CO2 sensor integrates a photoacoustic transducer with a detector, infrared source and optical filter on a small PCB. The sensor uses a microcontroller for on-board signal processing, sophisticated algorithms and a MOSFET for operating the infrared source.
A major challenge in developing a PAS-based CO2 sensor was to push the performance of the detector to its limits and minimize system noise, i.e. to isolate the MEMS detector from external noise so that only the pressure change originating from the CO2 molecules in the chamber is detected. The absorption chamber is acoustically isolated from external noise to provide accurate CO2 sensing information, otherwise the function of CO2 detection would be significantly disrupted. While developing the solution, Infineon could benefit from its many years of experience in acoustics and related applications. The modelling of the MEMS microphone response, patented acoustic isolation of the diffusion port, as well as a fast prototyping for validating the modelling results enabled an optimal system design.
Advantages of the CO2 sensor
Infineon has leveraged its state-of-the-art capabilities in sensors and MEMS microphones to develop a disruptive environmental sensing technology for CO2 . The XENSIV™ PAS CO2 (Table 1) is a real CO2 sensor based on the photoacoustic spectroscopy (PAS) principle. The sensor uses Infineon´s highly sensitive XENSIV™ MEMS microphone – – which detects the pressure change generated by CO2 molecules within the sensor cavity without picking up external noise. As output it provides CO2 concentration in ppm. The data shows highquality results even with the smallest pressure fluctuations. Accordingly, small amounts of gas are sufficient for an exact determination, which is why the size of the sample chamber could be designed suitably small.
XENSIV™ PAS CO2 offers an exceptionally small form factor that is 4 times smaller (14 mm x 13.8 mm x 7.5 mm) and 3 times lighter (2 g) than the typical NDIR sensor, allowing for more than 75 percent space-saving in customer systems. Furthermore, the majority of commercial NDIR sensors come with connectors which are not compatible with high-volume assembly standards and lead to a time-consuming manufacturing process. The XENSIV™ PAS CO2 , on the other hand, is designed and offered (in tape and reel) with large-volume automatic manufacturing in mind, possessing SMT capabilities for easy assembly and quick integration into customers systems.
In short, the sensor provides high accuracy in a super compact design, which make it the right choice for HVAC control (DCV) applications enabling energy saving and compliance to major smart building standards (e.g. LEED, WELL).
Availability and outlook
All sensor components are developed and designed in-house according to high-quality standards. Infineon will continue the development of PAS technology for further size-shrinking and cost optimization as well as performance adaption to other CO2 sensing appli-cations in industrial andconsumer markets. Other gases could potentially be addressed by the PAS technology platform. Additionally, the Infineon/Cypress eco-system will be leveraged to provide full-system offerings to the market including sensing, processing, actuating and connecting.
Prototypes of the new PAS CO2 sensor have already been tested and validated in key cus-tomer applications. A PAS CO2 evaluation kit is currently available for sampling. A complete suite of product evaluation boards (PAS CO2 evaluation board, Arduino-based Shield2Go board, Adafruit feather-based PAS CO2 wing board based on Infineon/Cypress eco-system) , software libraries as well as comprehensive documentation, including applications notes, will also be available soon to support customers and accelerate the design-to-market of the PAS CO2 sensor. Ultimately, the sensor will lead to significant improvements in indoor air quality and therefore in our health.
 U.S. Environmental Protection Agency. 1989. Report to Congress on indoor air quality: Volume 2. EPA/400/1-89/001C. Washington, DC.
 Li, Yuguo, et al,: Aerosol transmission of SARS-CoV-2; medRxiv; https://www.medrxiv.org/content/10.1101/2020. 04.16.20067728v1.article-metrics
 Hartmann, Anne; Kriegel, Martin: Risk assessment of aerosols loaded with virus based on CO2-concentration; Technische Universität Berlin; https://depositonce.tu-berlin. de/handle/11303/11478.2
 DOE, U. (2015). An assessment of energy technologies and research opportuni-ties. Quadrennial Technology Review. United States Department of Energy.
 Typical energy cost is $1.30/ft² (in 2007).
Figure 1: CO2 matters because levels of above 2,000 ppm significantly impact cognitive function
Figure 2: The increase of the concentration of CO2 (left axis) and aerosols (right axis) in a classroom the duration of two lessons with a break.
Figure 3. The XENSIV™ PAS CO2 sensor measures only 13.8 mm x 14 mm x 7.5 mm
Figure 4: The development of CO2 levels in one of Infineon’s meeting rooms.
Figure 5: Building energy accounts for 40% of total energy consumption in the U.S. and Eu-rope
Figure 6: The photoacoustic spectroscopy principle (PAS) Source: DOE, U. (2015). Chapter 5: Increasing Efficiency of Building Systems and Technolo-gies. Quadrennial Technology Review: An Assessment of Energy Technologies and Research Oppor-tunities, 143-181.
Table 1: Key specifications of the PAS CO2 sensor:
- Operation range: 400ppm to 10,000 ppm
- Accuracy: ± 30ppm +3% of reading between 400ppm and 5,000ppm
- Lifetime: 10 years at 1 measurement/minute
- Operation temperature: 0 – 50°C
- Relative humidity: 0% to 85% (non-condensing)
- Interface and compensation: I2C, UART and PWM
- Supply voltage: 12.0 V for the emitter and 3.3 V for other components
- Average power consumption: 11 mW at 1 measurement/ minute
- Package dimensions: 13.8 mm x 14 mm x 7.5 mm