Around here ozone (O3) is a major outdoor (and sometimes indoor) pollutant, but most monitors don’t seem to include it. PM2.5 might be very low, and O3 very high, or vise-versa. Maybe a thought to add to the outdoor sensor?
So some searching yields O3 sensors that read in the 0-150 ppb range for about $60 in smaller quantities. Their life is about 2 years. O3 pollution is one of the worst out there. O3 Article Maybe worth considering. Yesterday here it was 129 ppb, which is getting bad.
I did a little searching as well, since our most recent AQ warning was about ozone, not PM2.5. Here’s a quick summary:
-
SPEC 110-406 (pinned), 0-5 ppm range, 20ppb resolution, 5-10 year life, cross-sensitivity to NO2 and Cl2, $20
- Also available with a current-to-voltage breakout board (968-046), +$30
-
SPEC 110-407 (SMD), 0-20 ppm range, 20ppb resolution, 5-10 year life, cross-sensitivity to NO2 and Cl2, $20
-
Winsen MQ131 low-concentration sensor, 0-1.0 ppm range, unknown resolution, 5-year life, unknown unit price, retail availability:
- Open-source breakout board with Winsen MQ131 sensor, $18
- Open-source I2C board with Winsen MQ131 sensor, $21
-
Winsen ZE27-O3, 0-10ppm range, 10ppb resolution, built-in UART, 2-year life, unknown unit price, retail availability:
- DFRobot Gravity SEN0321, I2C interface to Winsen ZE27-O3 , $49
-
Winsen MQ131 high-concentration sensor, 10-1000ppm, 300 ppm resolution, 5-year life, unknown unit price, retail availability:
- MikroElectronika MIKROE-2767, SPI interface to Winsen MQ131 sensor, $54
- Also, the open-source boards above list the low-concentration sensor’s specs but seem to be pictured with the MQ131 high-concentration sensor.
-
Amphenol PS1-O3-5, 0-5 ppm range, 10ppb resolution, possibly not cross-sensitive to NO2 or Cl2, 2-year life, $82 for 1, $60 in bulk
Of these, the SPEC sensors and the Winsen MQ131 look the most interesting at $20:
-
The 10-year SPEC are rated for a 20pbb resolution; I’m guessing related to the noise floor, as they’re analog sensors?
-
The 5-year MQ131 has a lower range, so I would imagine it has better resolution, but I don’t know how to compute ppb from their data sheets.
I may have to tinker with those, whenever I get some time…
Wow, thank you very much for this thorough research. We really appreaciate your support!
In fact, we have now started exploring options to add O3 (ozone) and NO2 (nitrogen dioxide) sensors to our air quality monitors. Since several universities have also expressed interest in measuring these gases for their research projects, our priority is to select sensors that are suitable for scientific purposes. Specifically, we are looking for sensors with:
-
High precision at atmospheric concentrations (maximum 10 ppb resolution) -
Minimal impact from temperature and humidity -
Low cross-sensitivity to other gases
These features are, of course, also beneficial for all other customers 
The SPEC sensors offer a great price but we are not happy with their resolution and cross-sensitivities. Their performance in the AQ-SPEC tests (conducted by the South Coast Air Quality Management District) was disappointing. We reached out to SPEC for comments on the test results, but unfortunately, we did not receive a response.
We have similar concerns with the Winsen sensors, particularly regarding cross-sensitivity and a drift of up to 10% per year.
Then we found another company called EC-Sense which seemed to be a promising alternative. We have already tested their NO2 sensors against a reference instrument but, unfortunately, they are also affected by cross-sensitivity.
We are now considering Alphasense electrochemical sensors. Although they are slightly more expensive, they appear to offer the best overall performance. I have used Alphasense sensors during my PhD research and was pleased with their accuracy and reliability. Other experts in the field have also recommended them, provided they are properly calibrated.
At the moment, we are focusing on the technical aspects of integrating Alphasense sensors into our monitors.
Out of curiosity, are you thinking we would be able to add new modules (such as O3) to existing AirGradient models or would a new kit need to be developed to accomodate?
This would most likely be a new model as the sensors we are looking at are quite expensive and also due to its size needs some redesign of the enclosure.