JPS63121742A - Detection of carbon monoxide - Google Patents

Abstract

PURPOSE:To improve detection sensitivity and to suppress influence on the detected values of NOx and humidity by heating a gas sensor which utilizes the change in the resistance value of a metal oxide semiconductor by a heater to a high temp. to make heat cleaning of the sensor, then shifting the sensor to a low temp. region and detecting carbon monoxide from the peak output to the carbon monoxide. CONSTITUTION:A voltage is impressed to the heater 6 by using a switch 22 for the first 10sec and the heater power is kept zero by opening the switch 22 for the next 10sec. A switch 24 is closed just before lapse of 10sec after the heater 6 is turned off in the low temp. region and the sensor output at this point of the time is converted by an AD converter and is stored. A display means 28 is then operated.

Description

[Detailed description of the invention] [Field of application of the invention] The present invention relates to a method for detecting carbon monoxide.

[Prior art] It is known that a gas sensor that uses changes in the resistance value of a metal oxide semiconductor is heated alternately to a high temperature range and a low temperature range, and carbon monoxide is detected from the sensor output in the low temperature range. .

Japanese Patent Publication No. 53-43,320 shows the basic structure of this technology, and Japanese Patent Publication No. 61-17,943 discloses that the gas that interferes with carbon monoxide detection is NOx, which can be removed by an activated carbon filter. It shows. In addition, JP-A-60
No.-231,150 indicates that the heater power in a low temperature range may be set to O.

The inventors have discovered that a transient peak in the sensor output for carbon monoxide occurs in a low temperature range. This peak is based on the temperature characteristics of the sensor, and occurs more prominently as the heater power is reduced in a low temperature range. The output peak particularly occurs when the heater power is set to O in the low temperature range. At this peak point, the sensitivity to carbon monoxide, that is, the ratio of the output in carbon monoxide to the output in clean air, reaches its maximum value, and if detection is performed at this point, carbon monoxide can be detected with high sensitivity. Can be detected.

Other problems with carbon monoxide detection include reduced detection accuracy due to NOx and humidity dependence of the sensor. Experiments have revealed that the influence of NOx increases with time after the transition to the low temperature range, and that the humidity dependence of the sensor also gradually increases with time after the transition to the low temperature range. Therefore, detection using the peak of the sensor output can minimize the effects of NOx and humidity.

[Problems to be solved by the invention] The objects of this invention are: (1) improving the detection sensitivity to carbon monoxide; and (2) suppressing the influence of NOx and humidity on the detected values.

[Structure of the Invention] In the present invention, a gas sensor that utilizes a change in the resistance value of a metal oxide semiconductor is heated to a high temperature range by a heater to perform heat cleaning of the sensor. Next, the sensor is moved to a low temperature range, and carbon monoxide is detected from the sensor output at a time point corresponding to the peak of the sensor output for carbon monoxide.

In this way, the detection sensitivity to carbon monoxide can be increased, and the influence of NOx and humidity can be suppressed. Furthermore, the time it takes for a peak to occur is generally short, and the dead time for detection can be shortened.

[Example] Fig. 7 and Fig. 8 show gas sensors used for detection. 5n
C1t is hydrolyzed with ammonia, and the precipitate obtained is 6
It was fired in air at 00° C. for 1 hour to give a density of 5 nOt. This 5nOt was impregnated with a solution of baladirum chloride, and 550
It was thermally decomposed at ℃ for 30 minutes to support 0.3% by weight of PdO in terms of metal Pd. 5nQ1. is In2O3 or ZnO
, or may be replaced with any metal oxide semiconductor such as NiO. Further, Pd may be replaced by other catalysts such as Rh, RuO2, P2, etc.

This SnO2 was molded into a shape with a pair of coiled electrodes embedded therein, and sintered at 750°C for 5 minutes to form a sensor. In the figure, (2) is a sintered body of SnO2, (4) and (6) are coiled electrodes, and here the Pd-
An Ir alloy electrode was used, and the inner diameter of the coil was 0.6 mm. Coil (4) out of 5nOz.

The area facing (6) is responsible for the electrical conductivity of the sensor,
This becomes the detection area (8). In other words, the electrical conductivity of this portion becomes the output of the sensor. Also coil (4).

The area outside (6) is the NOx reduction catalyst area (10)
In this part, NOx reacts with carbon monoxide and NOx is removed. NOx is generated along with carbon monoxide from combustion equipment, increasing the resistance of 5102 and interfering with carbon monoxide detection. The NOX reduction reaction is expressed by the following reaction formula.

NO+CO→1/2N2+C02 N02+2CO→1/2N2+2COz In principle, the thickness F of the reduction catalyst region (10) is 0.3 mm in the example.
, the distance H between the coils (4) and (6) was set to 0.5 mm.

Next, this sensor was connected to the detection circuit shown in FIG.

In the figure, (11) is the detection power supply and its output and detection voltage Vc is IV in principle, (12) is the heater power supply and its output heater voltage vh is 1.2V, (14) is the load resistance,
(16) is a negative characteristic thermistor for temperature compensation, and (18) is a differential amplifier. (20) is a control circuit with a built-in oscillation circuit, whose configuration itself is well known, and which applies voltage to the heater (6) using the switch (22) for the first 10 seconds.
For the next 10 seconds, the switch (22) is opened to set the heater power to 0. Note that the heater power on the low temperature side does not need to be O, but may be within a range where the output to carbon monoxide peaks in the low temperature range. After turning off the heater (6), close the switch (24) just before 10 seconds have elapsed, convert the sensor output at that point into AD converter (26), store it in memory, and display it on display means (28) such as an LED or meter. ) to work. The display means (28) may be a warning means such as a buzzer.

Figure 1 does not show the behavior of the sensor output after transition to the low temperature range. The sensor is maintained at 300° C. until time 0, and at time 0, the heater (6) is turned off and the heater power is set to O to cool the sensor. The atmosphere was 20° C. and relative humidity 65%, and the figure shows the geometric average of the outputs of five sensors. The vertical axis represents the electrical conductivity σ of the sensor, and the horizontal axis represents time. Note that the scale of the horizontal axis is changed between up to 1 minute and after that time.

The detection target for carbon monoxide is usually about 100 to 200 ppm, which is represented by 100 ppm carbon monoxide. At this concentration, the sensor output peaks about 10 seconds after the transition to the low temperature side. The peak temperature is about 100°C in the case of 5na2. At 1000 ppm carbon monoxide, a gentle peak occurs around the 30th second mark, and in clean air, a trough in the output occurs around the 20th second mark. The positions of the output peaks and valleys vary depending on the sensor material, but S
In the case where 0.1 to 0.5% by weight of noble metal catalyst was added to nO2, the positions were almost the same and the above values were obtained. Therefore, if this peak is used, for example, if the sensor output from 8 to 40 seconds is used, preferably the sensor output from 8 to 20 seconds is used, carbon monoxide can be detected with high sensitivity. Note that the time at which the peak occurred was almost unrelated to the humidity of the atmosphere.

The carbon monoxide peak arises because the temperature at which maximum sensitivity to CO is obtained is between the heat cleaning temperature and room temperature.

Although the peak temperature differs depending on the metal oxide semiconductor, the occurrence of the peak itself is common.

Using 50 ppm No as a representative of NOx,
The effect was investigated by mixing it with 1100 pp of CO. No
The change in sensor output due to the presence or absence of NO increased with the passage of time from the peak, and by using the peak, the influence of NO could be kept to a minimum. This is because the temperature exhibiting maximum sensitivity to NOx is lower than the temperature exhibiting maximum sensitivity to CO.

Note that the temperature at which the maximum sensitivity to CO and NOX occurs differs depending on the metal oxide semiconductor, but it is common that the temperature at which the maximum sensitivity occurs is lower for NOx, and after the peak has passed, NOx
There is no change in the fact that the impact of

Next, as a representative of other interfering gases, 300 pp
The behavior of m towards hydrogen is shown. The sensor output shows a sharp peak immediately after entering the low temperature range, and then rapidly decreases. However, when 10 seconds have elapsed corresponding to the peak, 1100pp of C
The output to O is higher than the output to hydrogen at 300 ppm;
Erroneous detection due to hydrogen can be avoided.

Another important phenomenon is that the output in air or in 1.00 ppm carbon monoxide gradually increases after the peak. This is related to surface conduction due to water vapor adsorption,
Increases with absolute humidity. This phenomenon is also caused by high concentrations of CO
It is small in medium CO, for example 11000 pp. Surface conduction due to adsorption of water vapor, possibly due to the combined effect of adsorbed surface water and impurities such as chlorine ions in the semiconductor, causes detection errors due to humidity.

The peak output to carbon monoxide is related to heater power at low temperatures. This peak appeared more prominently as the heater power was reduced in the low temperature range. The room temperature was 20°C, the heating temperature in the high temperature range was fixed at 300°C, and the voltage applied to the heater (6) in the low temperature range was varied in the range of 0 to 0.4V.

When the heater voltage in the low temperature range was set to 0.4 V (the sensor temperature when left in the low temperature range for a long time was 80° C., hereinafter this temperature will be referred to as the steady heating temperature), no output peak occurred. Next, a faint peak occurred at 0.3 V (steady heating temperature 60° C.), and a clear peak occurred at 0.2 V (steady heating temperature 40° C.). Therefore, for this sensor,
The heater voltage in the low temperature range is preferably 0.25V or less, more preferably 0.2V or less. By using the output peak in the low temperature range, the time required to detect carbon monoxide can be shortened. As a comparative example, Fig. 2 shows a voltage of 1.2V (300V) to the heater (6) attached to the same sensor.
℃) and 0.4V (80℃). At time 0, change the heater voltage from 1.2V to 0°4
When changing to v, the output to carbon monoxide increases gradually and no peak occurs. Further, the output to hydrogen shows a sharp peak and then gradually decreases, but the change is slower than that in FIG. 1 where the heater power is set to 0 on the low temperature side. Focusing on the relative sensitivity between carbon monoxide and hydrogen, the time required to obtain sufficient relative sensitivity is long in FIG. 2 and short in FIG. 1.

Therefore, it can be seen that the detection dead time can be shortened by producing a peak in the output to carbon monoxide.

In FIG. 3, the heater voltage is changed to 1.
The sensor temperature is shown when changing between 2V and OV, and between 1.6V and Ov. Furthermore, the chain line shows the characteristics when the cooling time on the low temperature side was extended for 10 seconds.

Furthermore, the characteristics when the heater voltage on the low temperature side is set to 0.4V are also shown. The temperature at the start of cooling is not very important, and the sensor temperature does not change much after about 10 seconds. Also the first
The peak of carbon monoxide at 10 seconds in the figure corresponds to a little less than 100°C. The temperature at which the maximum sensitivity to carbon monoxide is obtained is approximately around this range.
It can also be seen that when V is used, cooling is slow and detection of carbon monoxide is delayed.

Figure 4 shows -10°C RH 100% and 20°C RH 65%.
The sensor behavior is shown under three conditions: and 508CRH60%. Humidity affects the sensor more than temperature, and absolute humidity is especially important. -10℃ is a dry condition,
20°C corresponds to standard conditions and 50°C to humid conditions. The time scale on the horizontal axis is the same as in Figure 1, 1 after the transition to the low temperature side.
It is changed before and after minutes have passed.

The peak at 100 ppm carbon monoxide around the 10th second, and the peak at 20 to 30 seconds at 11000 ppm carbon monoxide occur at any humidity.

Next, after a long period of time after the transition to the low temperature side, surface conduction occurs due to accumulation of H2O. For example, the power output in clean air at 50° C. increases considerably over time. On the other hand, the output to clean air at -10°C was extremely low and could not be illustrated. Also, even when the output is 100 ppm carbon monoxide, it gradually increases at 20°C and gradually decreases at -10°C. It was also clear from the fact that the sensor resistance depended on the detection voltage that these problems were caused by water adsorbed on the sensor. That is, the sensor resistance is a function of the detected voltage, and as the detected voltage increases, the apparent sensor resistance decreases. Furthermore, the dependence of the sensor resistance on the detected voltage increased with the progress of cooling. In addition, the dependence of the sensor resistance on the detection voltage or humidity changed depending on the shape of the sensor, and increased when the electrode spacing was shortened, and when the contact area between the electrode and the semiconductor was decreased.

These strongly suggest the existence of surface conduction due to water adsorbed on the sensor, or a combined effect of adsorbed water and impurities such as residual chlorine ions in the sensor.

The circuit conditions with the IV detection voltage used in the example, and the sensor shape in which the two coiled electrodes (4) and (6) are opposed with an interval of 15 mm suppress surface conduction and improve humidity dependence. This is to do so. As the detection voltage increases, the humidity dependence becomes more pronounced, and in a sensor with a linear center electrode placed at the center of a coiled electrode, the contact area between the center electrode and the semiconductor is small, and the electric field is concentrated near the center electrode. Therefore, humidity dependence increases. The reason why each electrode is coiled is to increase the contact area between the electrode and the semiconductor and to avoid concentration of electric field.

Looking at the change in sensor output due to differences in humidity conditions in FIG. 4, it is small near the peak of the CO specific output, and then increases gradually. Therefore, detecting carbon monoxide at a time point corresponding to the peak has the significance of reducing humidity dependence.

Figure 5 summarizes the effects of NOx and humidity.

These data are organized from the results shown in Figures 1 and 4. At 20°C and 65% relative humidity, IOOppm
The output to carbon monoxide is the standard, and the results are shown as a ratio to this. Both the influence of NOx and the influence of humidity are the least at the time corresponding to the peak of the output to carbon monoxide at the 10-second mark.

Table 1 shows the dependence of the sensor resistance on the detected voltage. 0
0100ppm, 20℃, 65% atmosphere as standard, -
Shows the output at 10°C 100% and 50°C 60%.

In the sensor of the example, 2■ or less, more preferably 1.5V
The following detection voltage is good, and humidity dependence can be suppressed by setting the detection voltage below the voltage at which surface conduction of adsorbed water occurs.

Table 1 Detection voltage dependence Vc σ/σ. (10 second weight σ/σ. (60
Second 1) Detection voltage -10°C 5 ships! -10℃ 1 boat!
0.5V O,51,60,+5 2.7IV
O,51,60,132,92V O,4+,
8 0.1 3.25V O,42,20,+
410V O,32,50,0851)σ
. is the sensor output for 100 ppm CO at 20°C and 65%, and shows the change in sensor output due to changes in humidity conditions.

The influence of NOx is caused by NOx reduction in the NOx reduction catalyst region (10).
It is relieved by reacting x with CO and removing it. Here, the NOx reduction catalyst was of the same type as the detection semiconductor, but a noble metal catalyst such as Rh, RuO2) Pt, Pd, or a transition metal oxide catalyst such as Mn2O3, Fe2O3, Cr, 03, etc. was used. Any NOx reduction catalyst can be used. When using a noble metal catalyst, the carrier is optional. When using a transition metal oxide catalyst, a noble metal catalyst may be supported and used in combination. Table 2 shows the effects of the thickness and material of the NOx reduction catalyst.

0.1mm 0.40
, 3 mm 0 ,
70, 5 mm 0
, 80.1mm + Mn catalyst 0.2mm 0
．． 80.1mm+RuO2 catalyst 0.2mm 0.
81) The thickness of the catalyst indicates the thickness of the SnO2+0.3Pd catalyst, and the thickness of the Mn catalyst and RuO2 catalyst is 0 and 2 mm.
So, for the Mn catalyst, 360 parts by weight of γ-A120 is used as a carrier,
Supported 20337 parts by weight of Mn to which 3 parts by weight of Rh was added;
u carrying 25 parts by weight of O, 2) σ is 100 ppm C
The sensor output when O and 50 ppm NOx coexist is σ.

] Shows the sensor output in OOppm CO, atmosphere is 20°C 65%.

In the case of a NOX reduction catalyst in which about 5 nOt is added with about 0.1 to 0.5 wt% of a noble metal catalyst, the preferred thickness of the catalyst layer is 0.25 mm or more.

[Effect of the invention] This invention improves the detection sensitivity to carbon monoxide and improves the detection sensitivity to NO.
The effects of X and humidity can be suppressed.

[Brief explanation of the drawing]

Figure 1 is a characteristic diagram of the carbon monoxide detection method of the embodiment, Figure 2
The figure is a characteristic diagram of a conventional carbon monoxide detection method, Figures 3 to 5 are characteristic diagrams of an embodiment, Figure 6 is a circuit diagram of a detection circuit, and Figure 7 is a diagram of a gas sensor used in an embodiment. FIG. 8 is a perspective view with a cutaway portion, and FIG. 8 is a sectional view thereof. In the figure, (2) metal oxide semiconductor, (4), (6)
Coiled electrode, (8) detection area, (10) NOx reduction catalyst area, (20) control circuit, (22), (24) switch,
(26) AD converter.

Claims (4)

[Claims] (1) A method in which a gas sensor that utilizes changes in the resistance value of a metal oxide semiconductor is heated alternately to a high temperature range and a low temperature range using a heater, and carbon monoxide is detected from the gas sensor output in the low temperature range. A method for detecting carbon monoxide, comprising detecting carbon monoxide from a gas sensor output at a time point corresponding to a peak of the gas sensor output to carbon monoxide after transition. (2) The method for detecting carbon monoxide according to claim 1, characterized in that the heater power is set to 0 in a low temperature range. (3) In the method for detecting carbon monoxide according to claim 2, the gas sensor uses SnO_2 as a metal oxide semiconductor, and the time point corresponding to the peak is 8 hours after the transition to the low temperature range. After ~40 seconds, the sensor temperature at that point is 12
A method for detecting carbon monoxide, characterized in that the temperature is 0°C to room temperature. (4) In the method for detecting carbon monoxide according to claim 3, the time point corresponding to the peak is 8 to 2 seconds after the transition to the low temperature range.
A method for detecting carbon monoxide, characterized in that the detection time is 0 seconds.