P-Type Semiconducting Hydrocarbon Sensors : Mechanistic Model

Titelangaben

Sahner, Kathy ; Moos, Ralf:
P-Type Semiconducting Hydrocarbon Sensors : Mechanistic Model.
2007
Veranstaltung: The 7th East Asian Conference on Chemical Sensors (EACCS 7) , 3.-5.12.2007 , Singapore.
(Veranstaltungsbeitrag: Kongress/Konferenz/Symposium/Tagung , Sonstige Präsentationstyp)

Weitere URLs

Abstract

The catalytic properties of a semiconductor are closely connected to electronic processes occurring at its surface. This knowledge was applied to explain the sensor effect of conductometric n-type semiconducting sensors. Since conductivity of the near-surface region is correlated to the amount of chemisorbed oxygen q, any process that changes q, for example a catalyzed oxidation of reducing gases, immediately affects the conductance of a sensor device. Although this mechanism has been successfully studied in the literature for ntype semiconductors, it presented several deficiencies when transferred to p-type materials. In the present contribution, a novel mechanistic model for semiconducting hydrocarbon sensors is proposed. The p-type conducting perovskite family SrTi1-xFexO3-d, which was shown to present promising sensor characteristics in the temperature range from 350 °C to 450 °C, is chosen as a model substance. In general, p-type semiconductors at least partially exchange lattice oxygen when catalyzing oxidation of a reducing gas. If the semiconductor can accommodate a large oxygen nonstoichiometry d, the participation of lattice oxygen in a surface reaction may be promoted. Thus, in the case of SrTi1-xFexO3-d, the oxidation reaction very probably is not limited to adsorbed oxygen species, but involves the bulk of the semiconductor. Assuming that a reducing gas Red attacks the semiconducting lattice consuming nearsurface lattice oxygen, this would lead to the consumption of holes and the formation of nearsurface oxygen vacancies. Assuming a fast bulk diffusion, which is in general valid for SrTi1-xFexO3-d, bulk defect concentrations in presence of a reducing gas thus differ from their equilibrium values in pure air. In particular, hole concentration p changes, which immediately affects conductivity s of the material (p ~ s). This leads to a set of defect chemical equations in the bulk, which were then solved for steady-state conditions in order to express hole concentration p as a function of the reducing gas concentration cRed. Subsequently, this function was combined with a conventional diffusion-reaction model proposed for porous sensor devices, which describes the local concentration profile c(Red) as a function of the film penetration depth. In the experimental section of this work, numerical calculations were compared with experimental response of SrTi1-xFexO3-d, samples towards a variety of test gases. For sensor optimization, operating temperature and film thickness had been varied. For the fit procedure, the diffusion coefficient was estimated using microstructure data from SEM images. Kinetic data, i.e., the reaction constant of the heterogeneously catalyzed surface reaction, was determined in a separate measurement series. A very good correlation between the model calculations and the experimental data sets was observed. In addition to describing the basic sensor characteristics, in particular the sensitivity at different operating temperatures, different methods for selectivity enhancement such as the variation of film thickness and the use of nanosized precursor powders were included in the model. Again, the calculated results were validated by experimental data sets.