No abstract

NDIR CO 2 gas sensors using a 10-cm-long gas channel and CMOS-compatible 12% doped ScAlN pyroelectric detector have previously demonstrated detection limits down to 25 ppm and fast response time of ∼2 s. Here, we increase the doping concentration of Sc to 20% in our ScAlN-based pyroelectric detector and miniaturize the gas channel by ∼65× volume with length reduction from 10 to 4 cm and diameter reduction from 5 to 1 mm. The CMOS-compatible 20% ScAlN-based pyroelectric detectors are fabricated over 8-in. wafers, allowing cost reduction leveraging on semiconductor manufacturing. Cross-sectional TEM images show the presence of abnormally oriented grains in the 20% ScAlN sensing layer in the pyroelectric detector stack. Optically, the absorption spectrum of the pyroelectric detector stack across the mid-infrared wavelength region shows ∼50% absorption at the CO 2 absorption wavelength of 4.26 μm. The pyroelectric coefficient of these 20% ScAlN with abnormally oriented grains shows, in general, a higher value compared to that for 12% ScAlN. While keeping the temperature variation constant at 2 °C, we note that the pyroelectric coefficient seems to increase with background temperature. CO 2 gas responses are measured for 20% ScAlN-based pyroelectric detectors in both 10-cm-long and 4-cm-long gas channels, respectively. The results show that for the miniaturized CO 2 gas sensor, we are able to measure the gas response from 5000 ppm down to 100 ppm of CO 2 gas concentration with CO 2 gas response time of ∼5 s, sufficient for practical applications as the average outdoor CO 2 level is ∼400 ppm. The selectivity of this miniaturized CO 2 gas sensor is also tested by mixing CO 2 with nitrogen and 49% sulfur hexafluoride, respectively. The results show high selectivity to CO 2 with nitrogen and 49% sulfur hexafluoride each causing a minimum ∼0.39% and ∼0.36% signal voltage change, respectively. These results bring promise to compact and miniature low cost CO 2 gas sensors based on pyroelectric detectors, which could possibly be integrated with consumer electronics for real-time air quality monitoring.

No abstract

No abstract