Abstract
Hydrogen gas (H2) sensors based on ultrathin (~10 nm) indium nitride (InN) epilayers with and without a thin catalytic platinum (Pt) layer atop have been fabricated and demonstrated. The bare In-polar InN sensor exhibits a detectable response to hydrogen at temperature of >=150 C and a maximum response at 225 C in both air and N2 ambiences. The maximum response of resistance variation under 1000 ppm H2 exposure in air (N2) ambience at 225 C is 16.2 ohm (4.6 ohm), corresponding to a variation ratio of 6.1% (2.5%), with a response time of 182 s (620 s). The plot of log (DI) vs. log (concentration of H2) reveals a linear relationship with a slope of 0.69 at a wide H2 concentration range from 50 ppm to 10000 ppm (1%) in air ambience. The hydrogen adsorption activation energy for the In-polar InN surface in N2 ambience at the temperature range from 150 C to 250 C is derived to be 0.91 eV. Moreover, the bare N-polar InN sensor exhibits a very similar response to its In-polar counterpart at various temperatures (150-250 C) and at different H2 concentrations (50-10000 ppm) in both air and N2 ambiences. The similar response to hydrogen for bare InN sensors with different InN polarities was explained as a result of the formation of an In-adlayer at InN reconstructed surfaces regardless of their polarities, which dominates the hydrogen sensing behavior. On the other hand, the Pt-coated In-polar InN sensor exhibits a much higher response to hydrogen even at lower temperatures (<=150 C). At room temperature, the current variation ratio of 64% is observed while exposed to 1000 ppm H2/air. At 150 C, the sensor shows a high current variation ratio of 177% and a short response/recovery time of 134 s/358 s under the same gas exposure environment. In comparison with a bare In-polar InN sensor tested in the same conditions, the Pt catalytic layer enhances the current variation by ~635 times. The detection limit of the sensor is experimentally found to be less than 5 ppm. The linear relationship of log (DI) vs. log (concentration of H2) displays a slope of 0.74 at a H2 concentration range from 5 ppm to 2000 ppm. In addition, the activation energy of the sensor at the temperature range from 25 to 150 C is derived to be 0.31 eV. For the Pt-coated N-polar InN sensor, it exhibits a similar response to its Pt-coated In-polar counterpart upon 1000 ppm H2/air exposure at various temperatures (250-150 C). However, the response to different H2 concentration at 150 C is rather different. The slope of log (DI) vs. log (concentration of H2) is as high as 1.34 at a H2 concentration range from 50 ppm to 2000 ppm due to the lower response at lower H2 concentration (<=250 ppm). The response is much higher than those of catalytic metal-gated HEMT-based H2 sensors operated in a normally-on mode with an unbiased gate under the similar gas exposure conditions. The higher response of hydrogen detection for the Pt-coated InN sensors might be associated with the incorporation of the catalytically active hydrogen atoms into the near-surface region of InN, which will act as donors and thus enhance the surface conduction current. In N2 ambience, due to the absence of oxygen to consume the catalytic hydrogen atoms, the Pt-coated In-polar InN sensor exhibits a much higher and faster response to hydrogen than that tested in air ambience. At 150 C, the response time is 14 s and the current variation ratio is 170% upon 100 ppm H2 exposure, which is close to the saturated response. In the meantime, a slow recovery rate is found because the hydrogen atoms would remain in the Pt bulk for a long time due to the same mechanism (the absence of oxygen to consume the catalytic hydrogen atoms).