Introduction In-situ monitoring of catalytic conversion with only one single (gas) sensor component represents a novel approach in gas sensor technology. The background of this idea is a sensor device that "compares" two gas atmospheres before (upstream) and after (downstream) of a catalyst. The sensor element itself separates two gas compartments from each other, but connects two (identical) electrodes in each of these gas compartments by means of an ion bridge. According to the mixed- potential principle, a half-cell potential is generated at each individual electrode. The potential difference between the two electrodes is evaluated as a sensor signal and, according to the model of mixed-potential formation, should depend directly on the ratio of the concentrations of a specific gas species between the two gas spaces, i.e., in this case directly on the conversion of the gas over the catalyst [1].The sensor device itself is the basis for reliable and reproducible measurements in that field. It is made from oxygen ion-conducting yttria-stabilized zirconia (YSZ) and designed as a full ceramic disc (figure 1a). Furthermore, it comprises an internal heating structure. So, the sensor can be brought to an appropriate measuring temperature (up to 650 °C) locally in the middle of the disc – exactly in that region where the planar electrodes are integrated on both sides and where the ionic conductivity is necessary. In the outer region, the sensor disc remains cold enough to use simple sealing materials (e.g. Viton). More details on the sensor setup can be found in [2]. A special housing (made from PEEK-cells) acts as sensor chamber with gas feed lines connected to the up- and downstream atmospheres of the catalyst (details can be found in [3]).In the present contribution, dynamic tests concerning the catalyst temperature and the inlet concentration are discussed. In these investigations, it should be verified that the sensor signal is independent from the inlet concentration. Experimental Setup Measurements were conducted in the lab. The inlet gas was mixed with mass-flow-controllers (MFCs) from propene in lean (e.g. oxygen containing) gas atmosphere (N2 balance). The gas flows over an oxidation catalyst (Pt as precious metal component on cordierite honeycomb substrate, diameter 1”, located in a quartz tube) inside a horizontal furnace (Carbolite). The gas / catalyst temperature was controlled by a thermocouple (upstream) inside the reactor. During the experiment, the furnace was heated up and – by that – conversion over the catalyst is increased, as long as the catalyst light-off temperature is reached (exothermic conversion is visible by a small temperature increase, measured with a downstream thermocouple inside the reactor, figure 2 first graph). The sensor device (disc inside the housing) was connected to the gas atmospheres up- and downstream the catalyst – each purging one compartment of the sensor housing with contact to each electrode. Simultaneously to the catalyst heating, the inlet (upstream) concentration was changed (see figure 2, second graph), which might be typical for real applications in the automotive field. The downstream HC-concentration was measured by FID. Out of this information, the conversion was calculated and plotted as (1 – conv.) in figure 2 (third graph). Figure 2 (last graph) shows the sensor raw signal that is a voltage signal as a difference of both half-cell potentials at the up- and downstream electrodes (in this case made from zinc chromite) measured at a sensor temperature of 500 °C. Results and Conclusions If one has a look on the raw data first (figure 2), the following results can be found. Regarding the up- and downstream temperature one assumes a light-off of the catalyst at about 150 °C (t = 15000 s). Starting from here, the downstream temperature increases more than the upstream temperature. The gas data indicate a small conversion even before (lower downstream concentration @ t > 7500 s). Although the calculated conversion is still “0”, (i.e. (1 - conv.) = 1), the sensor signal shows a slight increase in that area. In the medium temperature range (15000 – 30000 s, 150 °C < T < 175 °C) the conversion depends on the inlet concentration. Independently from that fact, the sensor signal follows the calculated conversion factor. In the higher temperature range (t > 30000 s, T > 175 °C), conversion is near 100 % (i.e. (1 - conv.) approaches “0”); the sensor shows its maximum signal.Now, in a second data evaluation, the sensor signal is plotted against the calculated conversion data in a half-logarithmic representation (figure 3, thick line including all measured data points in comparison to the thin line that was obtained in stationary experiments). We found an impressing correlation between both data sets, although the experimental sequence changed dynamically two parameters (temperature and inlet concentration) simultaneously and in a wide range. Therefore, the presented novel sensor device seems to be an appropriate candidate for catalyst conversion control.
Read full abstract