Tailoring of field effect gas sensors for sensing of non- hydrogen containing substances from mechanistic studies on model systems Mike Andersson, Anita Lloyd Spetz Dep of Physics, Chemistry, and Biology Linköping University SE-581 83 Linköping, Sweden mikan@ifm.liu.se Abstract— To gain knowledge about the transduction mechanisms involved in the sensitivity of field effect gas sensors towards non-hydrogen containing substances, such as O 2 , NO and CO, the sensor signal characteristics during exposure of some conceptually different model sensors to these as well as hydrogen containing gases have been investigated. The MOS capacitor based model sensors employ different combinations of insulator and contact materials, such as MgO, LaF 3 , IrO 2 etc. The gas composition downstream of the sensor during test gas exposure at various conditions has also been studied by mass spectrometry (MS) and compared for the different model systems. The results have been compared to the characteristics of ordinary SiC/SiO 2 /Pt structures and from the information obtained a tailor made field effect sensor structure for oxygen sensing, to our knowledge for the first time with minimal interference from H 2 , CO, and hydrocarbons, has been tested with promising results. I. INTRODUCTION Gas sensitive field effect devices based on the semiconductor silicon carbide (SiC) or other wide band gap materials like diamond, GaN and Al x Ga 1-x N have shown good possibilities of realizing sensors for high temperature applications such as combustion control or monitoring of car exhaust after-treatment systems [1]. The wide band gap of SiC and related materials permits their use at elevated temperatures (up to 800° C for SiC) and the material is also chemically inert, what makes it suitable for use in aggressive environments. By applying a gas sensitive material (commonly a catalytic metal) as contact to semiconductor/ insulator structures, capacitor-, Schottky diode-, or transistor devices with gas sensing properties and a simple readout can be constructed. The different Metal Insulator Silicon Carbide (MISiC) field effect sensor device designs are shown in Fig. 1 and in respect of a dense, homogeneous, catalytically active gate material (commonly Pt, Pd, and Ir), hydrogen dissociating on the surface rapidly diffuses through the gate metal, presumably creating a layer of polarized hydrogen atoms/ hydroxide groups at the oxide side of the metal/insulator interface [2]. This dipole layer changes the electric field in the semiconductor, affecting the density of mobile carriers in the near surface region, resulting in a shift of the current/ voltage or capacitance/ voltage characteristics. The sensor output is the applied voltage to keep a constant current through or capacitance over the device. Figure 1. Basic structure of the three principally different field effect devices applicable to gas sensing – a) Field Effect Transistor (FET), b) Metal Oxide Semiconductor (MOS) capacitor, and c) Schottky diode Substrate Substrate Epi-layer Epi-layer Insulator Insulator V V D S G Substrate Insulator Epi-layer V (a) (c) (b) 978-1-4244-5335-1/09/$26.00 ©2009 IEEE 2031 IEEE SENSORS 2009 Conference