JP2004012193A - Method for identifying and sensing gas - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain a method for identifying and sensing a gas which easily identifies and senses the gas although a conventional sensor array is used for identifying and sensing the gas relatively difficult to be identified and sensed. <P>SOLUTION: The method for identifying and sensing the gas implements a transient response output capture process for individually obtaining transient response outputs related to the identified gas from a plurality of semiconductor gas sensors and a frequency characteristic derivation process for obtaining frequency characteristics related to the outputs captured from a plurality of the semiconductor gas sensors and identifies the identified gas based on a relationship of the frequency characteristic of the identified gas obtained from each semiconductor gas sensor. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a method for identifying and detecting a plurality of types of gases with a plurality of semiconductor gas sensors.
[0002]
[Prior art]
As a gas identification detection method for identifying and detecting a plurality of types of gas, a sensor array including a plurality of types of gas sensors having different characteristics is provided in a gas detection unit, and these gas sensors are sensitive to a gas to be identified. There is known an output (static information) obtained by performing the identification detection.
[0003]
Usually, in this type of identification detection, the gas to be identified is brought into contact with the sensor array for a certain period of time, and the gas is identified and detected based on the maximum output value of each gas sensor obtained after the lapse of the time or its standardized value. .
[0004]
[Problems to be solved by the invention]
However, when the identification target gas is a mixed gas, which is composed of a relatively large amount of the main component gas and a trace amount of the minor gas (component extract), and the latter component extract is to be detected by identification, the conventional method described above makes the identification detection extremely difficult. There was a problem that was difficult.
[0005]
For example, when the mixed gas is a mixture of ethanol as a main component and a trace amount of a tea component, and this tea component is a component derived from green tea, black tea, or oolong tea, the conventional method is adopted. The maximum output value (normalized value) between the plurality of gas sensors has a pattern as shown in FIG.
[0006]
In FIG. 6, reference numerals S1 to S8 are axes corresponding to different types of gas sensors. The solid line corresponds to green tea, the broken line corresponds to black tea, and the one-dot chain line corresponds to oolong tea. The point on each axis indicates the maximum output value within a fixed time (for example, 120 seconds) for operating the sensor as described above.
[0007]
As can be seen from this figure, the patterns are overlapped between the respective tea components, and the identification cannot be detected.
[0008]
Therefore, the conventional identification method requires a sensor array having gas sensors with extremely high selectivity, but such a sensor array does not exist in reality.
[0009]
An object of the present application is to provide a gas identification detection method that can easily identify and detect a gas that is relatively difficult to detect as described above, using a conventional sensor array.
[0010]
[Means for Solving the Problems]
To achieve this object, the gas detection and identification method according to the present invention is characterized by:
Transient response output of the plurality of semiconductor gas sensors for the identification target gas, a transient response output capturing step of individually acquiring,
In the transient response output capturing step, a frequency characteristic deriving step of obtaining a frequency characteristic of the output captured for each of the plurality of semiconductor gas sensors is performed,
In the frequency characteristic deriving step, the identification target gas is identified based on a relationship between frequency characteristics of the identification target gas obtained for each semiconductor gas sensor. The operation and effect are as follows. .
[0011]
The theoretical basis considered by the inventors who have reached the present invention is as follows.
The adsorption and desorption of gas at the sensor-sensitive part in a semiconductor gas sensor is chemisorption and can be regarded as the formation of a kind of surface compound.
[0012]
On the other hand, the rate of progress of adsorption or desorption on the surface of various gas sensors is closely related to the rate of chemical reaction on the surface of the sensor. come.
[0013]
Therefore, even if the gas to be identified in the gas identification is dominated by the same main component in relation to the comparison object, a difference in the adsorption / desorption speed occurs due to the difference in the reaction speed between the minor components, and the sensor output The transient response will be different.
[0014]
For example, the semiconductor type sensor has a fast adsorption and desorption with respect to a gas having a high reaction rate, and has a slow rate when the gas has a low reaction rate. Therefore, when the reaction rate of the minor component is faster than the reaction rate of the main component, the difference between the minor components appears in the transient response at the time of adsorption or desorption. Therefore, by detecting the frequency characteristic of the transient response waveform, the component can be identified and detected.
[0015]
Therefore, in the method of the present invention, the capture of the transient response output is positively executed, and the waveform output having the difference in the reaction speed is obtained. Then, for example, performing a Fourier transform to obtain a frequency-dependent characteristic, and performing a principal component analysis based on the spectrum intensity data between the sensors to obtain a characteristic in a frequency domain associated with the sensor type and the gas type. Identification detection can be performed satisfactorily.
[0016]
Further, in the above method, as described in claim 2, in the transient response output capturing step, capturing the transient response output of the plurality of semiconductor gas sensors for the identification target gas,
It is preferable to execute a detection temperature changing step of changing a gas detection temperature by the semiconductor gas sensor in a state where the identification target gas is present.
[0017]
The sensitivity of a semiconductor gas sensor to gas depends on the operating temperature of the sensor, and the operating temperatures having the maximum sensitivity to various gases are different.
Therefore, when identifying a gas having a main component, it is possible to change the gas detection temperature by changing the gas detection temperature, for example, so as to have different sensitivities between the minor components. Minor component differences can be detected and discriminated.
[0018]
However, it is difficult to select an appropriate operating temperature in advance, and since the operating temperatures having the maximum sensitivity to the minor component of a plurality of types of gases are different, the gas detection temperature is, in effect, continuously or within a certain range. The measurement is performed so as to obtain an output waveform on which information based on the frequency characteristic based on the difference in the reaction speed is loaded while being changed stepwise.
Then, by obtaining the transient response output accompanying the temperature change and deriving the frequency characteristic, the identification detection of the present application can be made good.
[0019]
Further, in the above method, as described in claim 3,
After passing through the detected temperature changing step,
At an ordinary detection temperature of the semiconductor gas sensor, an air exposure step of exposing the semiconductor gas sensor to air,
It is preferable that either or both of a purge step of exposing the semiconductor gas sensor to air at a purge temperature higher than a normal detection temperature of the semiconductor gas sensor or both are included in the transient response output capturing step.
[0020]
In the above example, in order to obtain a transient response output, an output carrying desired information is obtained by changing the gas detection temperature. However, once the gas is adsorbed, the gas sensor desorbs the gas. Even in such a case, similarly, a response due to a gas type (in other words, a response rate based and dependent on a gas sensor type) can be obtained.
[0021]
This operation may be performed at a so-called normal detection temperature, or may be performed by a purge operation.
When a transient response output is obtained in a state where the gas to be identified is present around the gas sensor, and then a transient response output on the desorption side is obtained, more frequency characteristics are further required. It is possible to obtain a transient response output on which useful information is obtained, and as a result, it is possible to perform effective identification detection.
[0022]
BEST MODE FOR CARRYING OUT THE INVENTION
[Structure of gas detection device]
1 Device Configuration FIG. 1 shows a schematic structure of a gas identification and detection device 1 using the gas identification and detection method of the present invention. The figure is substantially the same as, for example, the configuration of an odor identification device EOS-100 manufactured by New Cosmos Electric Company to which the present inventor belongs. The operation software and analysis software stored in the storage unit 2 are different from those of the related art as described below, and a unique operation and analysis for the purpose of gas identification can be realized.
[0023]
In general, the device 1 is configured such that a gas introduction system 5 for introducing a gas to be identified into a sensor cell serving as a gas detection unit 4 in which a gas sensor array 3 is housed, and gas from the gas detection unit 4 is exhausted. And a gas exhaust system 6 for controlling the operation of the entire apparatus.
[0024]
The gas introduction system 5 includes a gas introduction unit 51 connected to a sampling system 7 having a known structure for individually introducing a gas to be identified into the identification device, an air introduction unit 52 including activated carbon, and introducing air. And a switching valve 53 for guiding the gas to the gas detection unit 4.
By employing this configuration, the gas to be identified from the sampling system 7 or the air from the outside air can be sent to the gas detection unit 4 in accordance with a command from the computer 2. The means for executing this operation is referred to as an operation means 201.
[0025]
The gas exhaust system 6 is provided with a flow control device 61 and a pump 62 below the flow control device 61 so that gas or air to be exhausted from the gas detection unit 4 can be appropriately exhausted.
This operation is also in accordance with a command from the operation means 201 from the computer 2.
[0026]
2 Gas sensor array 3
The configuration of the gas sensor array 3 provided in the gas detection unit 4 will be described in detail below.
The gas sensor array 3 is provided with a plurality of gas sensors 31 on a substrate 32 having a predetermined shape. The outputs of the individual gas sensors are individually taken in through the sensor output circuit 8 as shown in FIG. 2 are sent.
[0027]
As shown in FIG. 2, each gas sensor 31 is configured as a so-called substrate type sensor. This sensor has a heater 33 provided on the back side of a substrate 32 made of alumina or the like, and a pair of comb-shaped electrode films 34 provided on the front side thereof. A tin oxide film layer 35 is provided so as to cover the pair of electrode films 34. Further, the catalyst layer 36 is provided with or without the catalyst layer 36 on the front side.
The tin oxide film layer 35 carries lead oxide, and the gas reaching the tin oxide film layer 35 changes the resistance between the pair of electrode films 34. The change in the resistance can be taken out as a sensor output. It is possible. That is, this part is a sensitive part of the gas sensor 31.
[0028]
The catalyst layer 36 controls the gas reaching the sensitive part of the gas sensor 31.
[0029]
The sensor array 3 of the present application is provided with eight types of gas sensors 31, the configuration of the sensitive parts 35 is common, and the catalyst layer 36 is made of a different composition as described below. The composition of the catalyst layer 36 is shown in the following table.
[0030]
[Table 1]
Sensor 1 (S1) Sensor 2 without catalyst layer (S2) Al2O3
Sensor 3 (S3) ZnO
Sensor 4 (S4) ZnO2
Sensor 5 (S5) TiO2
Sensor 6 (S6) Mo-FeOx
Sensor 7 (S7) V / SiO2
Sensor 8 (S8) SiO2-Al2O3
[0031]
This type of gas sensor 31 is manufactured through the following steps.
A thick tin oxide film 35 is formed on a substrate by a screen printing method, baked at 1200 ° C. for 2 hours, and PbOx is carried on the film by an impregnation method. A catalyst having the above composition is applied on the sensitive film of the sensor, dried, subjected to a secondary sintering process, and completed through aging.
These gas sensors 31 exhibit different sensitivity characteristics to different gases, as is conventionally known.
[0032]
3. Identification Detection Operation In the gas identification detection device 1 of the present invention, the computer 2 appropriately controls the capture of information necessary for gas identification by the operation of the operation means 201, and the sensor output captured in a state unique to the present application. Is to be analyzed. This means on the analysis processing side is referred to as analysis means 202.
[0033]
This operation and its analysis can be summarized as follows.
1 Transient response output capturing step (mainly executed by the operation means 201)
The transient response outputs of the plurality of semiconductor gas sensors 31 to the identification target gas are individually acquired. Specifically, a sensor output as shown in FIG. Here, FIG. 3 shows the sensor output waveform in (a), the temperature change for obtaining this output in (b), and the intermediate type of the atmosphere around the sensor, as shown in detail below. It is a thing.
[0034]
2. Frequency characteristic deriving step (mainly executed by the analyzing means 202)
In the transient response output capturing step, the frequency characteristics of the output captured for each of the plurality of semiconductor gas sensors 31 are obtained. Here, the frequency characteristic means that the transient response output is subjected to frequency analysis (specifically, Fourier transform processing) to obtain a spectral intensity for each frequency (specifically, a spectral intensity of a frequency useful for identification detection). This means that the spectral intensity obtained for each sensor is recognized and examined in a pattern over all sensor types, and principal component analysis is performed.
[0035]
3. Identification detection step In the frequency characteristic derivation step, the identification target gas is identified based on the relationship between the frequency characteristics of the identification target gas obtained for each semiconductor gas sensor.
This identification determination can be performed by the operator, and can be automatically performed in comparison with experience data constructed in the computer 2. In this case, the computer 2 includes a determination unit.
[0036]
The above is the main flow in the case where the apparatus 1 of the present invention performs the identification detection. Hereinafter, the transient response output capturing step will be further described in detail with reference to FIG.
This step includes the following four steps.
[0037]
a Gas adsorption step In this step, the sensor is maintained at the normal detection temperature, which is the normal sensor temperature at which the gas sensor 31 is used, and the gas to be identified is led to the gas detection unit 4 from the blank state where there is no gas and is applied to each gas sensor. This is the step of adsorbing or sensitizing gas.
In this step, a gas sensor output as shown by area a in FIG. 3A is obtained, and its waveform becomes a rising and saturated waveform.
[0038]
b Sensor temperature changing step In this step, the sensor is changed to a temperature different from (normally lower than) the normal detection temperature while guiding the gas to be identified to the gas detection unit. This is the step of obtaining the sensor output. The form of this change may be a step-like change or a continuous change.
In this step, when the temperature is decreased, the sensor output decreases as indicated by the b1 area in FIG. 3A (the reverse change in the temperature occurs when the reverse temperature operation is performed). Shown). With this operation, it is possible to easily capture the sensitivity data between the sensor types having different reaction speeds and detection temperature-sensitivity characteristics.
[0039]
c Air exposure step In this step, the gas sensor 31 in the state of adsorbing the gas as described above is brought into contact with air having no gas to be identified while maintaining the gas sensor 31 at the normal detection temperature, and one of the gases in the adsorbed state is removed. This is the step of releasing the part.
In this step, for example, as shown in FIG. 3A, if the detected temperature is lowered and returned in the sensor temperature changing step, and then this step is performed, as shown in area c, Sensor output decreases.
[0040]
d Purge Step This step is to expose the sensor 31 to the air at a purge temperature higher than the normal detection temperature, and basically intends to purge the whole of the adsorbed gas from the gas sensor 31.
In this step, the sensor output rapidly decreases as indicated by area d in FIG.
Since these steps include a change from a state without gas to a state with gas or vice versa, and a change in sensor temperature in a state with gas, the sensor output change does not reach steady state. It is of a transient response type.
As a result, it contains a great deal of information that can obtain a favorable frequency response characteristic for specifying the characteristic of the gas.
[0041]
Such a process change operation is based on a predetermined operation pattern stored in the storage unit 20 of the computer. As described above, the operation unit 201 acquires the sensor output of each gas sensor 31 in such an operation pattern, and the analysis unit 202 executes a predetermined frequency analysis process and a principal component analysis. It is possible to obtain an analysis result useful for accurate identification detection as described above.
[0042]
[Specific example of gas identification detection]
Hereinafter, the detection status will be described in detail by taking as an example the identification of a 10% ethanol solution of green tea, black tea, and oolong tea extract.
[0043]
1. Transient Response Output Capture Step As described above, FIG. 3 shows the time on the horizontal axis, the sensor output (a), and the sensor set temperature (b) on the vertical axis. In FIG. 3B, the heater voltage corresponds to the temperature, and the higher the voltage, the higher the temperature.
As shown in the figure, in the gas adsorption step, the gas sensor is set to the normal detection temperature of 500 ° C. for 120 seconds. Thereafter, in the temperature changing step, the sensor temperature is continuously changed for 120 seconds. Specifically, after the temperature is once decreased from 500 ° C. to 400 ° C., the temperature is returned to 500 ° C. again. During this time, the gas to be identified is continuously supplied individually at a flow rate of 150 ml.
[0044]
Next, when the sensor temperature returns to the normal detection temperature of 500 ° C., refreshing of the gas sensor 31 is executed for 30 seconds by supplying clean air at a flow rate of 500 ml as an air exposure step. Further, in order to cause the desorption of the gas rapidly, the temperature is raised to 560 ° C. as a purging step, and purging is performed for 30 seconds.
[0045]
The output waveform of each gas sensor 31 at that time is as shown in FIG.
In this figure, each line indicates the sensor output of each of the eight gas sensors. The component extract contained in the gas is specifically green tea in FIG.
Therefore, in the case of this specific example, a figure having the same structure as this figure can be obtained for black tea and oolong tea.
[0046]
2 Frequency characteristic deriving step As described above, the waveform of the output curve obtained for each gas sensor type and for each identification target (green tea, black tea, oolong tea) is individually Fourier transformed, and the spectrum pattern is obtained for each frequency component spectrum. And analyze the pattern by a principal component analysis method. In the example in which the analysis results are shown in FIGS. 5 and 6, almost the entire time domain shown in FIG. 4 was set as a single frequency analysis target domain.
[0047]
FIG. 4 shows the spectral intensity after Fourier transform in a pattern as described above. In the figure, a component three times the fundamental frequency is displayed. The spectral intensity is standardized. Each axis corresponds to the eight types of gas sensors (S1 to S8) described above, and the types of solid lines, broken lines, and dashed lines are the same as those described above with reference to FIG. Compared to FIG. 6, it can be seen that there is variation between line types, and that gas identification and detection is possible at that pattern level.
[0048]
FIG. 5 shows the results of the principal component analysis. The horizontal axis represents the first principal component, and the vertical axis represents the second principal component. It can be seen that the distribution areas for each component are clearly separated according to the type of tea, and can be easily identified.
[0049]
[Another embodiment]
(1) In the above embodiment, the case where the number of sensor types is eight is shown. However, in principle, it is sufficient that the number is two or more, and five or more types are used, and relatively simple identification is performed. Detection becomes possible. Further, in the above embodiment, the substrate type sensor is shown, but a so-called hot wire type sensor can be used.
[0050]
(2) In the above embodiment, different types of tea are discriminated and detected. However, the inventor carried out the method of the present invention also for soy sauce, but could effectively discriminate and detect. That is, the present application is particularly useful for discrimination and detection of a gas containing a small amount of an extract component as a main component.
[0051]
(3) Since the identification and detection method of the present application is capable of identifying and detecting a component contained in a very small amount in the main component, it can be naturally used even when inspecting an identification and detection target in which a general identification difference easily occurs.
[0052]
(4) In the above embodiment, the gas adsorption step, the sensor temperature changing step, the air exposure step, and the purging step are performed sequentially and continuously. The present application can be applied only when the sensor temperature is changed, and only when the sensor temperature is changed thereafter.
Further, the present invention can be applied to a case where a plurality of types of gas sensors in a gas adsorption state are exposed to air and desorbed.
In the transient response output capturing step, the output capturing time can be arbitrarily set according to the purpose of the present application.
[0053]
(5) In the above embodiment, the case where the normal detection temperature is 500 ° C. has been described. However, in the case of a semiconductor gas sensor, the normal detection temperature may be selected in the range of 450 to 550 ° C. In this case, in the temperature changing step, it is preferable to change the temperature within a range of ± 100 ° C. to 200 ° C. from the normal detected temperature.
Further, it is preferable to select the purge temperature in a temperature range of about 560 ° C. to 650 ° C.
[0054]
(6) In the description so far, the gas to be identified is simply referred to as gas, and used as a term for air. However, as this kind of gas, a gas to be subjected to odor identification and a composition type Naturally, it includes those to be identified and any ones.
(7) In the above embodiment, the Fourier transform method was used to obtain the frequency characteristics, but any method such as the DWT (Discrete Wavelet Transforms) method introduced by Tim Edwards of Stanford University and the like was used. Can be adopted.
[Brief description of the drawings]
FIG. 1 is a diagram showing a configuration of a gas identification device used in the present application; FIG. 2 is a diagram showing a configuration of each sensor of a sensor array provided in the gas identification device; FIG. 3 is a sensor output and a detection temperature setting for each gas sensor; FIG. 4 shows a spectrum intensity pattern. FIG. 5 shows a principal component analysis result. FIG. 6 shows a pattern obtained based on the maximum output value from each sensor.
DESCRIPTION OF SYMBOLS 1 Gas identification detection apparatus 2 Computer 3 Gas sensor array 4 Gas detection unit 20 Storage unit 201 Operating unit 202 Analysis unit

Claims (3)

A gas identification detection method for identifying and detecting a plurality of types of gases with a plurality of semiconductor gas sensors,
Transient response output of the plurality of semiconductor gas sensors for the identification target gas, a transient response output capturing step of individually acquiring,
In the transient response output capturing step, for the output captured for each of the plurality of semiconductor gas sensors, performing a frequency characteristic deriving step of obtaining a frequency characteristic,
A gas identification detection method for identifying the identification target gas based on a relationship between frequency characteristics of the identification target gas obtained for each semiconductor type gas sensor in the frequency characteristic derivation step. In the transient response output capturing step, to capture the transient response output of the plurality of semiconductor gas sensors for the identification target gas,
The gas identification detection method according to claim 1, wherein a detection temperature changing step of changing a gas detection temperature by a semiconductor gas sensor is performed in a state where the identification target gas is present. After passing through the detected temperature changing step,
Any of an air exposure step of exposing the semiconductor gas sensor to air at the normal detection temperature of the semiconductor gas sensor, and a purge step of exposing the semiconductor gas sensor to air at a purge temperature higher than the normal detection temperature 3. The method according to claim 2, wherein one or both of them are included in the transient response output capturing step.

Images