(Invited) Effects of Noble-Metal Loading Onto Metal-Oxide Electrodes on CO-Sensing Properties and Mechanism of Potentiometric Gas Sensors Utilizing an Anion-Conducting Polymer Electrolyte

Abstract

Introduction Some kinds of electrochemical CO sensors using a cation-conducting polymer (e.g., Nafion®) as an electrolyte and carbon-based electrodes loaded with noble metal (mainly Pt) have been presently commercialized as a sensor operated at room temperature, and they practically show attractive CO-sensing properties. On the other hand, many kinds of metal oxides are basically limited to be used as the sensing-electrode materials of the sensors, due to the lack of chemical stability under the strong acidic environment. Recently, anion-conducting polymers (ACP) have been a focus of constant attention, in the field of various electrochemical applications, such as fuel cells and water hydrolysis, because of the large hydroxide-anion conductivity and the relatively long-term stability. We have so far developed various kinds of electrochemical gas sensors utilizing the ACP electrolyte, and we firstly demonstrated that Pt- or Pd-loaded carbon black (CB) was quite promising as a sensing-electrode material to detect H2 [1], CO [2], and CO2 [3]. However, their gas selectivity was relatively poor, and thus we have investigated various metal oxides loaded with noble-metal nanoparticles as a CO-sensing electrode material [4–7]. In this presentation, we will discuss the fundamental CO-sensing properties of their sensors and the CO-sensing mechanism. Experimental n wt% noble-metal (N) nanoparticles were loaded onto metal-oxide (MO) powder by precipitation-deposition (only for Au) or general impregnation (for other noble metals) technique, respectively, and they were heat-treated at elevated temperatures in air or H2 for 1 h. The obtained powder (nN/MO) was mixed with an appropriate amount of ACP solution (AS-4, Tokuyama Corp.) and then the paste obtained was applied on the surface of both sides of an ACP membrane (A201, Tokuyama Corp.) as sensing and counter electrodes by blade coating, and then EC(nN/MO(Tm)) sensors (T: annealing temperature (°C), m: annealing atmosphere (air or H2) were obtained. The sensor was sandwiched with Au meshes (100 mesh) as a current collector and was set up in a gas-sensing measurement system with two electrode compartments. Electromotive force (E) of all sensors to 500 ppm CO or H2 balanced with synthetic air (O2: 20%, N2: 80%), which flowed over the sensing electrode, was measured at 30°C, while synthetic air was flowed over the counter electrode. Relative humidity (RH) of these flowing gases was controlled to be 0–80% (generally, 57%) at 30°C by bubbling the gases into distilled water. The magnitude of response was defined as a change in E value induced by a sample gas (ΔE SG, SG (sample gas): CO or H2). CO selectivity against H2 was defined as a ratio of CO response to H2 response (ΔE CO/ΔE H 2). Results and Discussion Different combinations of various N (Au, Ir, Ru, Pd, Pt, etc.) and MO (Bi2O3, CeO2, In2O3, TiO2, SnO2, V2O5, etc.) have been estimated as a CO-sensing electrode material. Among them, we confirmed that only three kinds of nN/MO, nAu/In2O3, nAu/SnO2, and nPt/SnO2, were promising candidates as the sensing-electrode material of highly sensitive CO sensors, together with relatively excellent operating stability. Response transients of representative EC(2Au/In2O3(Tm)), EC(2Au/SnO2(Tm)), and EC(2Pt/SnO2(Tm)) sensors to 500 ppm CO and H2 are shown in Fig. 1. Both the EC(2Au/In2O3(400air)) and EC(2Au/SnO2(400air)) sensors showed larger CO response than H2 response, but the CO selectivity against H2 (2.6 for EC(2Au/In2O3(400air)), 2.3 for EC(2Au/SnO2(400air))) was smaller than we expected. The annealing of the 2Au/In2O3 and 2Au/SnO2 at 250°C in H2 decreased the H2 response and thus enhanced the CO selectivity against H2 (28 for EC(2Au/In2O3(250H2)), 35 for EC(2Au/SnO2(250H2))). However, the small H2 response gradually increased and thus the CO selectivity against H2 decreased with their operating time, even though the CO response remained almost unchanged during the same period. On the other hand, the EC(2Pt/SnO2(500air)) sensor showed the largest CO selectivity against H2 (8.3) among all the EC(2N/MO(Tair)) sensors, together with the excellent long-term stability. On the other hand, the heat treatment of 2Pt/SnO2 powder at 250°C in H2 enhanced the responses to both CO and H2 and consequently reduced the CO selectivity against H2 (1.9 for EC(2Pt/SnO2(250H2))). The reasons for such variations in sensing property and the gas-sensing mechanism on the N/MO surface will be discussed in our presentation.