Abstract

Semiconducting copper oxide (CuO) gas sensing layers show a remarkable conductance behavior if exposed to hydrogen sulfide (H2S) gas at low operating temperature (180°C). At first conductance decreases as expected for a p-type semiconducting metal oxide offering reducing test gas. After a certain exposure time, however, a sudden steep increase in conductance can be observed.In a first approach this behavior is explained by the formation of metallic conducting copper sulfide (CuS, degenerate p-type semiconductor) clusters which eventually form conducting pathways across the sensing layers short-circuiting the remaining CuO phase. In the field of statistical physics such behavior can be described by the so-called percolation theory.Here we present a detailed experimental and theoretical analysis of the observed effect utilizing RF-magnetron-sputtered copper oxide films with different stoichiometry (CuO, Cu4O3 and Cu2O) as model systems. The layers are exposed to H2S for different time spans and analyzed with respect to their morphology (SEM, XRD) and chemical composition (XPS, ToF-SIMS). Analysis of the transient behavior of the conductance by means of a percolation model and comparison of the results to the experimental data allow the identification of different processes. For CuO samples first the formation of different non-CuS copper–sulfur–oxygen phases is observed followed by the percolation regime with the steep conductance increase. Afterwards diffusion processes superimposing the percolation leading to a slower conductance increase and eventually the process is dominated by diffusion of copper ions from the bulk. For oxides with other stoichiometry (Cu2O, Cu4O3) no percolation regime is observed which is attributed to higher diffusion rate of copper ions weakening the percolation effect in these samples.Based on these observations a model for the electronic conductance behavior of copper oxide gas sensors under exposure to hydrogen sulfide (H2S) at temperatures below 200°C is proposed. A better understanding of these systems will enable the preparation of reliable sensors with inherent thresholds.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call