Carbon monoxide gas sensor made of stabilized zirconia

Carbon monoxide gas sensor made of stabilized zirconia

Solid State lonics 1 (1980) 319-326 0 North-Holland Publishing Company CARBON MONOXIDE GAS SENSOR MADE OF STABILIZED ZIRCONIA H. OKAMOTO, H. OBAYAS...

427KB Sizes 0 Downloads 17 Views

Solid State lonics 1 (1980) 319-326 0 North-Holland Publishing Company



H. OKAMOTO, H. OBAYASHI and T. KUDO Central Research Laboratory, Hitachi Ltd., I-280, Higashi-Koigakubo, Kokubunji, Tokyo 185, Japan Received

11 April 1980

The emf of zirconia galvanic cells, 02 (1) + CO, Pt/stabilized 2102 /Pt, 02 (II), becomes anomalously higher than that calculated from Nernst’s equation when PCO/P02 (1) < 2 at lower temperatures such as 350°C. This anomalous emf is considered to be caused by a mixed electrode potential. This potential results from the electrochemical reactions of O*- in the solid electrolyte with the CO and oxygen both adsorbed on Pt during the CO oxidation on Pt. A new CO gas sensor which can detect small amounts of CO in air is proposed and examined. This new sensor is based on the anomalous emf and has two electrodes on both sides of a stabilized Zr02 ceramic pellet. One electrode is a Pt electrode exposed directly to a sample gas, while the other is a Pt pseudoair electrode which is covered with a CO oxidation catalyst. The sensor generates a response emf corresponding to the CO concentration in the sample gas without a reference 02 gas. It provides an emf higher than 30 mV for 100 ppm of CO in air in the temperature range between 260 and 350°C.

1. Introduction Stabilized Zr02 is an oxide ion conductive material [ 1,2] and has been used as a solid electrolyte in fuel cells [3], 0, pumps 141, and 02 gauges [5,6]. The 02 gauges measure the 0, partial pressure in a sample gas with the aid of a reference 0, gas. They consist of a galvanic cell: 02(Ph2),






where PG, is the 0, partial pressure of the sample gas and P& is that of the reference 0, gas, e.g., air. The galvanic cell (A) shows an emf, E, gken by Nernst’s,equation: E = (R T/4F) ln(Pb2 /P& ) ,


where R is the molar gas constant, T is the’absolute temperature, F is Faraday’s constant. Pg, can be calculated using eq. (l), when E is measured under a known P&, . When the sample gas contains combustible gases such as CO, the galvanic cell now can be described as: 02(Po2)

+ CO(Pco), Pt/stabilized





H. Okamoto et al. /Carbon

Fig. 1. emf of an 02 gas senior


at 350°C.


gas sensor made of stabilized zirconia

+ CO(PCO),

Here the P& to be measured is the equilibrium tion at the anode: 2coto2=2co*.




0, partial pressure of the reac-


The equilibrium constant of this reaction is 2.0 X 103* atm-l (2.0 X 1O33 Pa-l) at 350°C [7]. Consequently, an emf gap appears at the stoichiometric ratio of reaction (2) depicted by the broken curve in fig. 1. This is why zirconia 0, gauges can be used as 0, gas sensors for automotive exhaust gases to detect whether the air-fuel ratio is stoichiometric or not [8]. Conventional 0, gas sensors give a curve close to the ideal one at temperatures above 500°C. At lower temperatures around 350°C the emf becomes anomalously high as depicted by the solid curve in fig. 1 when Pco/Po2< 2 (anomalous emf). This anomalous emf has been accounted for by competition between 0, and CO gases for triple point adsorption sites on the anode electrode [8]. However the mechanism involved has not been clearly explained in detail yet. This paper presents a new CO gas sensor which utilizes this anomalous emf. A pseudoair electrode, i.e., a Pt electrode covered with a CO oxidation catalyst, serves as a reference gas electrode. Consequently this sensor can detect CO in an ambient gas without any particular reference 0, gas.

2. Experimental Sample preparation. Yttria stabilized ZrO, was used as a solid electrolyte. Eight mol% ofY203 (99,.9%, Kojundo Kagaku Co.) and 92 mol% of ZrOz (99.2%, Daiichi Kido Kagaku Co.) were mixed with 1 wt% of SO2 (Kojundo Kagaku Co.) and the mixture was calcined at 1500°C for 1 h. Next, the calcined specimen was pulverized. Then three grams of the powder were pressed into a pellet and sintered at

H. Okamoto et al. / Carbon monoxide gas sensor made of stabilizedzirconia


1600°C for 3 h. The sintered pellet was about 1 mm thick and 20 mm in diameter and had an apparent density of 94-95% of the theoretical one calculated from X-ray diffraction. Platinum electrodes were deposited by electron beam evaporation on both sides of the pellet at 300°C. Typical electrodes were 0.3 pm thick and 12 mm in diameter. CO oxidation catalysts were prepared by impregnating r-Al203 (Alcoa Co. Fl) with H,PtCl, (Mitsuwa Kagaku Co.) aqueous solution and then reduction with NaBH4 (Wake Junyaku Co.) aqueous solution. Before mounting the catalyst, an electric lead made of colloidal Au (Tokuriki Kagaku Co., No. 8556) was formed on one of the Pt electrodes for later connection to an external lead wire. The catalyst powder was then mounted over that Pt electrode with a few drops of 1 wt% methyl cellulose aqueous solution and dried. Only one electrode was covered with the catalyst. Experimental set-up. The apparatus employed to measure the anomalous emf is shown in fig. 2a. Springs were used to hold the pellet specimen between the glass tubes and make the system air-tight. All gases were supplied from commercially available gas cylinders without any special purification. Air was fed onto the upper surface of the specimen (air electrode), and a mixed gas of CO, 0, and N2 was passed over the other surface of the specimen (sample gas electrode). The velocity of each gas was 1.7 m s-l (at 23°C). Air or a gas mixture flowed through a guide glass tube over the specimen surface and then out into the outer glass tube through four narrow openings in the guide glass tube near the specimen surface. The specimen temperature was monitored by a chromel-alumel thermocouple arranged to contact the sample gas electrode. The emf was recorded via an impedance converter




guide glass t&e


Pt electrode



R electt7xk



samp!e gas >







(a> Fig. 2. Experimental gas sensor.

set-up employed

for emf measurement

of (a) an O2 gas sensor and (b) a CO


H. Okamoto et al. / Carbon monoxide

Ras sensor made of stabilized zirconia

(Nikko Keisoku Co. type IC-2, nominal impedance; lOlo 5G?)by a recorder (Hitachi Ltd. type 056). The apparatus employed to study the behavior of the new CO gas sensor consisted of one glass tube containing a sample holder as shown in fig. 2b. A specimen was placed horizontally on the sample holder and the Pt electrode connected to a Pt net lead. At the same time, the catalyst covered electrode was connected to another Pt net lead via the Au lead. As an ambient gas, air flowed at a velocity of 0.07 m s-l (23°C) and could be mixed with a small amount of CO when desired. Specimen temperature and emf were measured in the same way as that for measuring anomalous emf.

3. Results and discussion The behavior of a galvanic cell (B) near stoichiometry at lower temperatures has already been shown by the solid curve in fig. 1. The result for emf measurement when the Pco/Po2ratio is smaller than 2 at 350°C is shown in fig. 3. The observed values are quite different from the ones calculated using Nernst’s equation (1). The anomalous emf is about 100 mV when Pco/Po2< _ 0.1[region (I)], while it is more than 700 mV for PcolPo,> _ 0.7[region (III)]. In the region 0.1 5 Pco/Po25 0.7 [region (II)], the emf increases sharply with oscillating values, providing the emf range shown in fig. 3. The mechanism responsible for the appearance of anomalous emf is considered to be as follows. The microscopic states of the adsorption of gases near a triple contact Pt-stabilized Zr02-gas are shown schematically in fig. 4. At the air electrode, only oxygen is adsorbed on Pt leading to the reaction: O(a) + 2eU =+02-



where O(a) represents oxygen adsorbed on Pt; e- is an electron in Pt; and 02-

Fig. 3. Anomalous

emf of an O2 gas sensor when Pco/Po2

< 2


H. Okamoto et al. /Carbon




air electrode I$. 4. Schematic representation contact at lower temperatures.


gas sensor made of stabilized zircorzia

sample gas ektrode

of adsorbed

states ofgases

near a Pt-stabilized

ZrOz ~-gas triple

an oxide ion in stabilized ZrO,. Reaction (3) gives the equilibrium electrode potential. On the other hand, both oxygen and CO are adsorbed on Pt at lower temperatures at the sample gas electrode. The reactions: CO(a) + 02-

-+ CO,(g) + 2e- ,


O(a) + 2e- + 02-


occur where C02(g) is gaseous and hardly adsorbed on Pt [9]. Reactions (4) and (5) give a mixed electrode potential. Thus the observed emf at lower temperatures is the difference between this mixed electrode potential and the equilibrium electrode potential. The relation between the emf and the Pco/Po2 ratio at lower temperatures (fig. 3) is quite complicated. However it seems to indicate a surface adsorption state during the CO oxidation on Pt. The CO oxidation on Pt proceeds via either the Eley-Rideal mechanism: O(a) + CO(g) + CO2 (8) a or the Langmuir-Hinshelwood O(a) + CO(a) + CO,(g) .

(6) mechanism

[IO] : (7)

In region (I) in fig. 3, the main species adsorbed on Pt is anticipated to be oxygen, so reaction (6) is expected to dominate. In region (III), the Pt surface will be largely covered with CO, and reaction (7) is expected to be predominant. However in region (II), both reactions (6) and (7) occur and the amount of CO adsorbed on Pt is not constant [IO] leading to the oscillation of the emf. In this way, the anomalous emf of an 0, gas sensor is considered to have a direct relation to the amounts of the CO and oxygen adsorbed on Pt during the CO oxidation on Pt. A CO gas sensor detects very small amounts of CO in air and thus is related to the anomalous emf in region (I). Therefore, region (I) should be examined in more detail to clarify the possibility of the CO gas sensor.


H. Okamoto et al. / Carbon monoxide


gas sensor made of stabilized zirconia







Fig. 5. emf change

for a small amount

of CO in air.

The emf response behavior in region (I) was obtained by adding 160 ppm of CO to air (pCo/Poz = 8 X 10p4) at 300°C using the galvanic cell (B) as shown in fig. 5. Approximately five minutes after CO introduction, the emf reached 90% of the saturated emf. The difference between the original and the saturated emf is called the response emf. When the CO supply was stopped, the emf returned to its original level within about ten minutes. If the equilibrium of reaction (2) should be established, the calculated emf would be on the order of 0.1 mV. The observed response emf, however, greatly exceeds this value and it is as high as 50 mV as shown in fig. 5. AC0 gas sensor, in making use of these results, requires a reference 02 gas. However the new CO gas sensor shown in fig. 6 does not require a reference 02 gas. One of the two electrodes is in direct contact with the sample gas, while the other has a CO oxidation catalyst over it and does not directly contact the sample gas. The uncovered electrode gives an anomalous electrode potential. The covered one gives a pseudoair electrode potential. This is because a small amount of CO reacts with 0, in air while diffusing through the porous CO oxidation catalyst and only 02 and a small amount of CO, are present at the Pt electrode surface. Thus the CO gas sensor in fig. 6 gives an emf corresponding to the partial pressure of CO without any special reference 0, gas. The activity of the catalyst, then, is the key to this sensor’s performance. Aluminasupported Pt was examined as the CO oxidation catalyst and its effect on the anoma-

CO oxhtlon Pt ( pseulo



( CO m air)

I Fig, 6. Structure


Pt([email protected]

catalyst air electrode) Zr02 gas electrode)

of the new CO gas sensor operable

in air.

H. Okamoto et al. / Carbon monoxide

Table 1 Effect of Al2 OS-supported PCOlPO,

Pt catalyst

gas sensor made of stabilized zirconia

on the anomalous

emf at 350°C. Catalyst


weight = 50 mg,

= 1 ___~ Catalyst


none 0 wt%Pt 0.05 0.2 (talc.

0.5 0.4 0.1 0.06 0.04)


_.._~ _

emf at Pco/Po = 1, at 350°C is shown in table 1. The amount of the catalyst was fixed at 50 mg.\t is seen that the electrode covered with a catalyst containing more than 0.2 wt% of Pt gives a good approximation of the equilibrium electrode potential. Next, the behavior of this new type CO gas sensor was studied. The relation between the response emf and the CO concentration in air at 300°C was studied and the results are shown in fig. 7. This relation is analogous to the curve in region (I) in fig. 3. The emf increases and its slope becomes small as the CO concentration is increased. The temperature dependence of the emf at 100 ppm of CO in air is shown in fig. 8. Between 260 and 35O”C, the response emf is more than 30 mV. Above 35O”C, it decreases sharply with increasing temperature and approaches 0 mV, which corresponds to the ideal behavior of the sample gas electrode as an 0, gas sensor electrode. Around 250°C the emf was difficult to measure, probably because the velocities of electrochemical reactions (4) (5) became very slow. lous

100 80

\ E 0


300 93

20 $

60 0 I: 0


4CO 600



ICOl/pprn Fig. 7. CO concentration


of the response

emf of the CO gas sensor.


H. Okamoto

et al. /Carbon

gas scnsar made of‘ stabilized



IO 0

.-L-_--I. 250






Fig. 8. Temperature


of the response

emf of the CO gas sensor.

4. Summary A CO gas sensor has been developed which has Pt electrodes on both sides of a stabilized ZrO2 pellet, one electrode in direct contact with the sample gas and the other one covered with a CO oxidation catalyst. This new device has proved to be a convenient gas sensor detecting CO on the 100 ppm order in air at around 300°C The response emf of the sensor is considered to result from a mixed electrode potential caused by the electrochemical reactions of 02- with the CO and oxygen adsorbed on Pt during the CO oxidation reaction on Pt. The response emf increases with the CO concentration in air, but its slope becomes small. Between 260 and 35‘0°C the response emf is more than 30 mV at 100 ppm of CO in air.

Acknowledgement The authors wish to thank Dr. M. Seki for valuable suggestions.

References [l] [2] [3] [4] [5] [6] [7]

W.D. Kingery, J. Pappis, M.E. Doty and D.C. Mill, J. Am. Ceram. Sot. 42 (1959) 394. D.T. Bray and U. Merton, J. Electrochem. Sot. 111 (1964)447. J. Weissbart and R. Ruka, J. Hectrochem. Sot. 109 (1962) 723. D. Yuan and F.A. Kroger, J. Electrochem. Sot. 116 (1969) 594. J. Weissbart and R. Ruka, Rev. Sci. Instr. 32 (1961) 593. C.J. Mogab, J. Vacuum Sci. Technol. 10 (1973) 852. J.P. Elliott and M. Gleiser, Thermochemistry for steel-making, Vol. 1 (Addison-Wesley, Reading, 1960) p. 170. [8] W.J. Fleming, J. Electrochem. Sot. 124 (1977) 21. [9] P.R. Norton, Surface Sci. 44 (1974) 624. [lo] E. McCarthy, J. Zahradnik, G.C. Kuczynski and J.J. Carberry, J. Catalysis 39 (1975) 29.