(i) Gas sensitivity.
Plot o f the resistance variation with CO concentration and relative humidity in air vs time for the Ti^SnO2-0.2%Sb sample is given in Fig 6.10. The measurements were performed at room temperature. On initial exposure o f the dry sample to CO in dry air, following decomposition, no resistance response was recorded. On exposure to 50% relative humidity in air a large resistance decrease was observed; such behaviour is a well known characteristic o f n-type tin dioxide [44], as was mentioned earlier (Fig 4.8, Chapter 4). Similarly, as can be seen in Fig 6,10, the resistance did not completely recover in dry air, following the exposure to moisture o f the Ti^SnO2-0.2%Sb sample. Note, however, that following the exposure to moisture (where no CO resistance response was recorded) the sample showed a modest p-type resistance response to CO in dry air: on exposure o f the “wet sample” to CO in dry air a resistance increase was recorded
<— 100000 10000 ---> 800 640 480 320 160 0 0.5 .s 0.3 d o 0-2 g U g u O u 0.1 0.0 T im e/m in
Figure 6.10 Time profile of the room-temperature resistance of Ti modified SnO2-0.2%Sb pellet as a function of water vapor pressure and exposure to ppm concentrations of CO in dry and wet (50% relative humidity) air.
The effect, however, was lost again on subsequent exposure to 50% relative humidity in air. Fig 6.11. shows the continuation o f the experiment for the subsequent dry air - moist air cycles. It can be seen that the p-type CO resistance response can be recovered every time in dry air, having been lost in moist air.
10000 9000 <---- 8000 7000 Ph 6000 5000 4000 - > 3500 2000 2500 3000 1500 0 500 1000 0,3 0.2 0.0 T i m e / m i n
Figure 6.11 Time profile of the room-temperature resistance on the subsequent dry air-wet air cycles
of the Ti modified SnO2-0.2%Sb pellet (as per fig 6.10) as a function of exposure to ppm
concentrations of CO in dry and wet (50% relative humidity) air.
The variation o f the resistance with CO concentration and relative humidity at room temperature is shown in Fig 6.12 for the Ti modified TiO2-10%Nb sample. N ote the large resistance decrease in response to 50% relative humidity in air, which recovered when dry air was reapplied. N ote also that for this system a CO resistance response could be registered in both dry and wet (50% relative humidity) air and it appears that the sensitivity to CO was not affected by the water vapor: a resistance decrease o f the same order o f magnitude in response to the presence o f traces o f CO was recorded. Table 6.4 summarizes the values for the resistance response, S, to lOOOppm CO in dry air at room temperature for Ti^SnO2-0.2%Sb and Ti^TiO2-10%Nb
Table 6.4 Resistance response, S, to lOOOppm CO in air at room temperature for Ti^SnO]- 0.2%Sb and Ti^TiO2-10%Nb. Sample SlOOOppmCO Ti^SnO2-0.2%Sb 0.12 Ti'^TiO2-10%Nb 0.25 o c
Ï
01
o «>I
o kO u 3 oo a K 0 730 1095 1460 1825 2190 2920 3650 T im e /m iaFigure 6.12 Time profile of the room-temperature resistance of a Ti modified TiO2-10%Nb pellet as a function of water vapor pressure and exposure to ppm concentrations of CO in dry and wet (50% relative humidity) air.
6.4 Discussion.
As was seen to be the case with platinized SnO2-0.2%Sb, decoration o f the
BaSno.9 7Sbo.0 3O3 surface with Pt resulted in a room temperature gas sensitive electrical conductivity. The effect was also observed for the 10%Pt/TiO2- 10%Nb, which is a non-tin containing oxide. In the case o f 1 0%Pt/BaSno.9 7Sbo.o3 0 3 surface electronic states o f Pt origin were identified in the band gap above the valence band edge, causing increase in the band bending o f
ca
0.3 eV. It is believed that, similarly to platinized SnO2-0.2%Sb, the room temperature CO resistance response o f the material is induced due to electronic interaction between the Pt additives and the oxide support.Based on the Pt loading and the chemical state o f the surface Pt, the electrical behaviour o f the platinized BaFe0 3.x could be expected to be similar to that o f the other semiconducting oxides studied. However, it was demonstrated that this sample was not sensitive to CO at room temperature. The valence band spectrum o f the ferrate was characterized by a broad peak at 5eV due to Fe 3d, which dominated the electronic structure o f the material in the energy region typically characteristic for the Pt surface state energy. It is apparent that electronic effects due to charge transfer between the supported P t and the oxide would be diminished in this case, resulting in the electrical conductivity being insensitive to interaction o f gases with the supported Pt at room temperature.
W e have demonstrated that the role o f the Pt additives, finely dispersed over the semiconducting oxide surface, is to induce the room-temperature CO resistance response o f the material. The effect, however, can only be observed in the case where the support response mechanism is dominated by electronic properties o f the Pt, so that variations in the Pt chemical state modulates sensitively the semiconducting properties o f the oxide.
Surface grafting o f Ti onto SnO2-0.2%Sb resulted in a modified gas sensing behaviour o f the material. A p-type resistance response to CO was recorded at room temperature, which was only observed in dry air following exposure to moisture. A particularly interesting observation is the actual p-type response in the Ti modified SnOi, since both materials are generally n-type semiconductors [33] and in an unmodified form show no gas sensitive electrical conductivity at room temperature.
The possible mechanism might be substitution o f the adsorbed molecular w ater for CO at the Ti site. Since the effect is a resistance increase, it was dominated by a large resistance decrease in moist air - characteristic for SnO2-0.2%Sb support - and no CO response could be observed in moist air. Maschmeyer
et al \\\9 ^
showed, that Ti"*^ centres, grafted onto mesoporous silica (MCM-41) were stable on exposure to ambient air. On contact with moisture, Ti'^^ centres were shown to coordinate the molecular water, but this did not effect the catalytic activity. The gas sensing behaviour o f the Ti^SnO2-0.2%Sb system is rather different from that observed for the Ru^SnO2-0.2%Sb. Both systems, however, showed room temperature gas sensitive conductivity from surface grafting o f the reactive centres onto oxide support.We have also shown that the effect was not specific to SnO2-0.2%Sb support, since surface grafting o f Ti onto TiO2-10%Nb also resulted in the room temperature gas sensitivity o f the device. The interesting point here is that the surface-grafted Ti was evidently in a different chemical ^environment to the Ti o f the oxide itself, at the surface. In this case, for the first time, the material was actually unaffected by the water vapor. The resistance response to CO was recorded in both moist and dry air. This system, so far, has proved to be particularly promising.