WEB Towards understanding response generation in conductometric gas sensorsFriday (07.08.2020) 20:41 - 20:41 Poster Room Part of:
Conductometric gas sensors (CGS) translate the change in composition of surrounding atmosphere into a measurable electrical response. Their construction is simple: a sensitive layer connecting two electrodes, which makes them inexpensive and easy to fabricate, and therefore one of the most popular gas sensing solutions. Although SnO2, a common sensitive material, shows excellent chemical stability and is currently used for detection of numerous gases, CGS still suffer from limited sensitivity and selectivity, leaving room for research-driven improvement.
The precise mechanism of response generation in these devices largely remains a mystery. Based on phenomenological studies, a model called “the oxygen chemisorption mechanism” was developed, which explains the change in sensor resistance by means of charged oxygen adsorbates altering the Fermi level in the subsurface region of the sensitive layer. Although this model was successfully used to explain some observations, no such oxygen adsorbates have ever been spectroscopically observed under working conditions, including in our own extensive operando studies.
In this work, a summary of multiple experiments showing the lack of direct spectroscopic evidence in support of an oxygen chemisorption mechanism is presented, along with a demonstrated correlation between sensor response and oxygen vacancy formation. NAP-XPS experiments were performed on a working SnO2-based CGS with in situ resistance measurements in different O2 pressures and temperatures. Downward binding energy shifts in the Sn 3d and O 1s emissions accompanied by increased resistance were observed in every instance, indicating both spectroscopic and phenomenological response of the sensor to the increasing pressure of O2. However, the O 1s peak shape, which could indicate formation of oxygen adsorbates, remains unchanged regardless of oxygen pressure, therefore making emergence of such adsorbates unlikely. On the other hand, analysis of the area of Sn and O emissions suggests that the resistance change correlates not only with the binding energy shift, but also the ratio of O to Sn in the analysed subsurface volume, and possibly in the bulk. A description of conductivity as a function of oxygen vacancy concentration, which in turn is a function of surrounding oxygen partial pressure, could revolutionise the way we perceive sensing mechanisms of CGS and result in a more complete picture, eventually leading to the design of new materials.