Effects of Hexamethyldisilazane Modification on Gas-Sensing Properties of Semiconductor Gas Sensors

Abstract

Introduction Metal-oxide based semiconductor-type gas sensors have been used as gas leak detectors, but their sensing performances are easily influenced by the adsorption of organosilicon compounds containing in spray cans and air refreshers. In this study, the surface of SnO2-based gas sensors was thermally modified by hexamethyldisilazane (HMDS, ((CH3)3Si)2NH)), one of representative organosilicon compounds, and their sensing properties to CH4 and H2 were examined. Based on the obtained results, effects of the adsorption of HMDS on the gas sensing properties were discussed. Experimental A SnO2 powder was prepared according to the following procedure. An appropriate amount of NH4HCO3 aqueous solution was added into SnCl4 aqueous solution. The obtained white precipitate was repeatedly centrifuged and washed with pure water, and then dried at 100°C for overnight. The resultant powder was calcined at 600°C for 3 h in air. In some cases, Pd or Pt nanoparticles were loaded onto the surface of SnO2 powders by using a conventional impregnation method. The resultant solids were heat-treated at 200°C for 2 h in H2, to obtain metallic nanoparticles on the oxide surface. The prepared SnO2-based powders were mixed with α-terpineol, and the obtained paste was screen-printed onto an alumina substrate equipped with a pair of interdigitated Pt electrodes (gap size: ca. 200 μm), followed by drying at 100°C. Then, they were calcined at 500°C for 3 h in air.The obtained gas sensors were mounted in the tubular furnace, and the surface of some sensors was modified with HMDS at 400°C, in dry air saturated with HMDS at 80°C. The obtained SnO2 sensors loaded with and without noble metal (N: Pd or Pt) were denoted as nN/SnO2(t) and SnO2(t) [n: amount of noble metal, 0.5, 1.0 (wt%); t: exposure time to HMDS-saturated air, 0, 15 or 30 (min)], respectively. Gas responses of the obtained sensors to CH4 and H2 (2000 ppm) diluted with dry air were measured at a flow rate of 100 cm3 min−1 at 300–400°C. The magnitude of response was defined as the ratio (R a/R g) of resistance in air (R a) to that in CH4 and H2 balanced with air (R g). Results and Discussion The resistance of the SnO2(0) sensor in air decreased with an increase in the operating temperature, and the thermally HMDS-modified SnO2(t) sensors (t: 15, 30) showed smaller resistance in air than the SnO2(0) sensor over all the temperature range studied. The resistance of the SnO2(t) sensors decreased upon exposure to CH4, and the magnitude of response to CH4 increased with an increase in the operating temperature.Figure 1 shows the variations in response of the SnO2(t) and nN/SnO2(t) sensors to CH4 and H2 at 400°C. The CH4 and H2 responses of the SnO2(0) sensor were increased by the loading of Pt and Pd, except for those of the 1.0Pd/SnO2(0) sensor, and the 1.0Pt/SnO2(0) sensor showed the largest responses to CH4 and H2 among the examined sensors. In addition, the CH4 response of all the sensors decreased with an increase in the HMDS exposure time, while their H2 response increased with an increase in the HMDS exposure time. The response time of the 1.0Pt/SnO2(t) sensor to CH4 increased with an increase in the amount of HMDS exposure time, although its response time to H2 was almost independent of the exposure time.These results can be explained by a decrease in the amount of negatively charged oxygen adsorbates on the SnO2 surface and a restriction of the gas diffusion to the SnO2 surface (especially CH4 and O2, due to their larger molecular size than that of H2), and they were caused by the adsorption of Si-based compounds on the SnO2 surface by the thermal decomposition of the HMDS. The details will be discussed in the presentation. Figure 1