In May 2018, the World Health Organization (WHO) released a startling and sobering update on global air pollution. The new data reveals that approximately 4.2 million deaths each year can be attributed to poor outdoor (ambient) air quality, and an additional 3.8 million result from exposure to household pollutants such as dirty cookstoves and fuels. In fact, 91 percent of the world’s population are regularly exposed to levels of airborne pollutants that exceed WHO guidelines.
Airborne particulate matter of 2.5 µm or less in diameter (PM2.5) from pollutants such as sulfate, nitrates and black carbon can penetrate deeply into the cardiovascular and respiratory systems, reducing lung function and causing a range of health issues including heart disease, stroke and cancer. As part of the U.S. Environmental Protection Agency’s 1990 Clean Air Act, standards were established for both PM2.5 and 10 µm sized particles (PM10). The maximum exposure concentration over a 24-hour period is 35 µm/m3 for PM2.5 and 150 µm/m3 for PM10.
One of the first steps in addressing the air quality problem is measuring it locally—inside homes, buildings and outdoors. The good news is that there are low-cost, embeddable particle and gas sensor technologies available which can directly quantify these and other key air quality factors. Integrating air quality sensors into IoT and embedded products has the potential to add entirely new dimensions of functionality and value—improving our quality of life. And, when data from many points are networked, uploaded to the cloud and correlated geographically, emerging problem areas and pollution sources may be identified and addressed earlier than was previously possible.
Airborne particulates are generally measured using light-scattering techniques. A narrow beam of light is focused on air flowing through a chamber, with a detector placed at angle (perhaps 90°) from the beam Airborne particulates are generally measured using light-scattering techniques. A narrow beam of light is focused on air flowing through a chamber, with a detector placed at angle (perhaps 90°) from the beam such that it is shielded from the light. When particles are present in the air sample, the light hitting the particles will be scattered and some will enter the detector. The light source is often an LED or laser diode, as in Figure 2. One such airborne particle sensor is the Telaire SM-UART-01L+ from Amphenol Advanced Sensors. It provides a digital readout of PM2.5 concentration over a range of 1 to 999 µm/m3. All airborne “dust sensors” must have a means for drawing sample air into their measurement chambers. Some incorporate small fans; others use a heat from a resistor at the bottom of the chamber to pull air up by thermal convection.
In addition to fine particulates, carbon dioxide can also significantly impact air quality and health, particularly in offices, schools and other enclosed spaces. The most widely accepted guidelines for CO2 concentration are from the American Society of Heating, Refrigeration and Air Conditioning Engineers (ASHRAE). They recommend that the CO2 level in a typical office be no more than about 700 ppm above the local outdoor value. Outdoor CO2 concentrations vary based on a number of factors. On average in the U.S. it is about 300-500 ppm. So, if the local outside value is 380 ppm, the target indoor value would be 380 + 700 = 1,080 ppm or less.
The maximum allowable CO2 concentration for industrial workplaces is substantially higher (5,000 ppm), allowing for CO2 generation from machinery and industrial equipment. The primary source of CO2 overall is burning fossil fuel. In industry it may be a byproduct of chemical production and more “benign” processes, like beer fermentation. It is also not uncommon for commercial greenhouses to raise CO2 levels to accelerate photosynthesis and increase productivity and yield.
There are a number of methods to measure CO2. The most widely used is called NDIR (non-dispersive infrared). NDIR sensors leverage the fact that CO2 molecules have a characteristic resonance mode which corresponds to 4.26 µm mid-infrared light. When the molecules resonate, they absorb energy at that wavelength. Internally, an NDIR sensor is typically arranged as a small tube with a broad-spectrum IR emitter at one end, and an IR detector with a 4.26 µm bandpass filter at the other end. As air passes through the tube, the CO2 molecules in the sample will resonate at 4.26 µm and absorb a portion of the energy. This results in less energy reaching the detector. NDIR sensors have the benefit of being very selective with low drift. To compensate for changes in the IR emitter over time, NDIR sensors will usually include a second channel with an unfiltered detector. Figure 3 illustrates a typical two-channel NDIR CO2 sensor used in remote air quality monitoring systems. Like many NDIR sensors, the broad-spectrum emitter is actually a small, low-current incandescent bulb which turns on only during measurements. For “continuous” sampling, the bulb is switched on very briefly every few seconds. When power consumption is critical, it can operate in “on-demand” mode as well.
Enabled by advanced sensor technology, air quality measurement may well be the next “big thing” in remotely-deployed IoT nodes, residential HVAC management and even consumer wearables. Smart phones, for example, have rapidly evolved to satisfy consumer demand for health and fitness data. Today, many have MEMS accelerometers used to count steps and other outdoor activities, and some have UV sensors to help users avoid excess sun exposure. Why not add air quality monitoring to the feature set?
Well, if we consider integrating “classic” NDIR sensors into a smart phone, cost, size and power consumption may be major obstacles. However, there is another technology on the horizon which may fill the bill quite nicely. It’s known as photoacoustic spectroscopy, or PAS. PAS is based on an interesting discovery made by Alexander Graham Bell in the 1880’s. While experimenting on his “photophone”, he noted that strobed sunlight could produce an audible sound on certain materials.
A PAS sensor has four primary elements: a strobed light source, a sample chamber, a small sealed chamber containing gas (the same gas that we are sensing), and a microphone. Three of these can be fabricated using MEMS technology, potentially yielding a very small, very low-cost CO2 sensor.
To measure CO2 concentration using PAS, the sealed chamber would be filled with CO2, as illustrated in the diagram above. One side of the chamber has an IR-transmissive window, another side has a microphone attached to it. As noted earlier, CO2 molecules resonate at 4.26 µm. A small amount of heat is released in the process. Because the chamber is sealed and contains a fixed volume of gas, the added heat causes its pressure to rise (remember , the ideal gas law, from chemistry class). Thus when short pulses of 4.26 µm IR pass into the sealed chamber through the window, the temporary thermal expansion causes a vibration which is picked up by the microphone. In effect, the combination of a CO2-filled sealed chamber and microphone makes a very selective 4.26 µm bandpass filter.
To build the complete system, a sample chamber area has a strobed light source (such as an LED) on one end and the sealed chamber filled with CO2 on the other When CO2 molecules enter the sample area, they will resonate, absorbing a portion of the 4.26 µm energy before it reaches the sealed chamber. As a result, the sealed chamber will undergo a proportionally lower pressure change during each strobe, and the signal from the microphone will decrease.
It’s hard to predict which types of sensors will enable the next waves of industrial, embedded, IoT-enabled and consumer devices. We can be sure, however, that as the “last mile” to our physical world, sensors will play an increasingly important role in product design, and may potentially fuel advances into entirely new markets.