Patent Application
20190137426
The present invention relates to filters in gas sensors
Gas sensors have the problem of being poisoned by organo-siloxanes and similar gases. Therefore, filters such as active carbon (Patent Document 1: JP4104,100B), mesoporous silica (Patent Document 2: JP2013-242,269A), colloidal silica (Patent Document 3: JP5841,810B), and so on have been proposed. Further, it is known that sulfo group in the filter is effective for the removal of siloxanes (Patent Document 3). More, it is known to restrict the introduction amount of surrounding air into the filter is effective for using the filter.
Patent Document 1: JP4104,100B
Patent Document 2: JP2013-242,269A
Patent Document 3: JP5841,810B
For improving the long-term stability of gas sensors, it is necessary to increase the amount of adsorbent in the filter. However, the increase in the amount of adsorbent results in a delay in response and makes the size of the gas sensor larger.
The object of the invention is to improve the long-term stability of gas sensors while keeping the delay in response to the gas to be detected within an allowable range and keeping the size of the gas sensors within a practical range.
A gas sensor according to the invention introduces the surrounding atmosphere through a filter to a gas sensing element,
the filter comprises a silica-based adsorbent containing sulfo group,
the gas sensor is provided with a housing accommodating the filter and the gas sensing element,
the housing is provided with an opening through which the surrounding atmosphere is introduced into the filter,
the diameter of the opening is denoted by D in mm, the length of the filter in the direction from the opening toward the gas sensing element is denoted by L in mm, and
0.1≤D≤1.5, 2≤L≤12, L/D2/3≤10, and 5≤L/D.
The silica-based adsorbent containing sulfo group and active carbon conventionally used in gas sensors are different in the following points:
the silica-based adsorbent containing sulfo group has not a high siloxane adsorption capacity; and
siloxanes adsorbed on the silica-based adsorbent containing sulfo group does not desorb but siloxanes adsorbed on active carbon desorb.
When using the silica-based adsorbent containing sulfo group as the filter and when extending the length of the filter (the thickness of the filter from an opening to surrounding atmosphere toward the gas sensing element), utilizing the short detection delay, the period before breakthrough of siloxanes is made longer. The extended length of the filter, however, makes the size of the gas sensor too large. On the contrary, when making the diameter of the opening for introducing the surrounding atmosphere to the filter smaller, then the period before breakthrough is extended without extending the length of the filter.
The inventor has investigated the influence of the length L of the filter and the diameter D of the opening on the detection delay and on the period before breakthrough of siloxanes. Then it is found that the detection delay is determined according to L/D2/3 and the period before breakthrough is according to L/D. The dependency on D is different between the detection delay and the period before breakthrough, and when making D smaller, then the period before breakthrough is made longer while keeping the detection delay short. In consideration with controlling the diameter D of the opening uniform and with keeping the size of the gas sensor within a practical range, 0.1≤D≤1.5, 2≤L≤12, L/D2/3≤10, and 5≤L/D in mm 5≤L/D is the condition for extending the period before breakthrough, and L/D2/3≤10 is the condition for keeping the detection delay within an allowable range. Further, for keeping the length of the filter within a practical range, L is determined such that 2≤L≤12, for satisfying the condition of 5≤L/D, D is made up to 1.5 mm, and for fabricating the uniform openings, D is made down to 0.1 mm.
Here, when 0.3 mm≤D≤1.2 mm, the openings with a uniform diameter are easily fabricated,
when 3 mm≤L≤10 mm, the gas sensor has a size suitable for easy implementation, and when 7≤L/D, the period before breakthrough is made longer. Further, when 0.3 mm≤D≤1.2 mm, 3 mm≤L≤10 mm, 7≤L/D, and L/D2/3≤6.5, the detection delay is made further shorter.
The diameter of the filter is denoted by R in mm in a plane perpendicular to a direction from the opening of the filter toward the gas sensing element (the lengthwise direction of the filter). When making R larger, the influence on the detection delay is small and the period before breakthrough is extended. Therefore preferably, R is set 6 mm≤R≤16 mm.
For applying the invention to a non-circular opening, a virtual diameter D′ of the opening and the area S of the opening are related in such a way that S=π/4·D′2 and D′=(4S/π)1/2, and the virtual diameter D′ is used in place of the diameter D. Similarly, when the cross-section of the filter is not circular, a virtual diameter R′ of the filter and the area S′ of the filter are related in such a way that S′=π/4·R′2 and R′=(4S′/π)1/2, and the virtual diameter R′ is used in place of the diameter R.
The silica-based adsorbent is, for example, silica gel, mesoporous silica, high-silica zeolite and, since it has a large mean pore diameter, the detection delay is short, and the adsorbed siloxanes are polymerized by the sulfo group introduced. The mean pore diameter is, for example, down to 1 nm and up to 20 nm, specifically down to 2 nm and up to 20 nm, preferably down to 3 nm and up to 20 nm, and particularly preferably down to 4 nm and up to 20 nm.
A gas sensor according to the invention introduces the surrounding atmosphere through a filter to a gas sensing element,
the filter comprises a silica-based adsorbent containing sulfo group,
the gas sensor is provided with a housing accommodating the filter and the gas sensing element,
the housing is provided with an opening through which the surrounding atmosphere is introduced into the filter, and
the diameter of the opening is denoted by D in mm, the length of the filter in the direction from the opening toward the gas sensing element is denoted by L in mm, and the area of the opening is denoted by S in mm2, and
0.1≤D≤1.5, 2≤L≤12, 5≤L/(4S/π)1/2, L/(4S/π)1/3≤10.
Preferably, 0.3≤D≤1.2, 3≤L≤10, 7≤L/(4S/π)1/2, and L/(4S/π)1/3≤6.5. The inventor has confirmed that, when the opening is not circular or when plural openings are present, the sum of the areas of the openings is important and that L/(4S/p)1/2 should be used in place of L/D. More, the inventor has confirmed that L/(4S/p)1/3 should be used in place of L/D2/3. For example, when one opening is present and when the diameter of the opening is D, S is given by π/4·D2, and inversely D=L/(4S/π)1/2 holds. Regarding the area S′ of the opening, 6≤(4S′/π)1/2≤16 is preferable, when S′ denotes the area of the filter in mm2 in a cross-section perpendicular to the direction from the opening toward the gas sensing element.
In the present specification, the filter comprising a silica-based adsorbent containing sulfo group means that the silica-based adsorbent is 60% or more by mass ratio in the adsorbent in the filter, preferably 70% or more. For example, active carbon loaded with a precious metal such as Pt may be contained at a concentration of 40% or less by mass ratio, preferably 30% or less, in the adsorbent. Since an adsorbent such as active carbon is a cause of detection delay, when a layer of active carbon or the like is provided within the filter, the thickness of this layer is included in the thickness of the filter.
The best embodiment for carrying out the invention will be described in the following.
Structure of Gas Sensor
The openings 14, 17 are for example circular, and the cap 12 is for example in the shape of a circular tube. The diameter of the opening 14 is denoted by D, the length of the filter 10 along the direction connecting the openings 14-17 is denoted by L, and the diameter of the filter 10 on a plane perpendicular to the above direction is denoted by R. In the present specification, the length and the diameters of the openings are indicated in mm. The diameter of the opening 17 is made, for example, more than twice the diameter of the opening 14.
A gas sensor 2 where the gas sensing membrane 22 was a 30 μm thick SnO2 membrane containing 1.5 mass % Pd was prepared. The material of the filter 10, the diameter D of the opening 14, the length L of the filter, and its diameter R were changed so that their influences on the delay time τ before detection and the periods before siloxane gases penetrated the filters were measured. The gas sensor 2 was a sensor for detecting methane or LPG, was operated with a period of 30 seconds and, during every period, the gas sensing membrane 22 was heated to 450° C. for 0.1 second.
Measurements
The gas sensors 2 were operated for 80 days in an atmosphere containing siloxanes M3, D4, and D5, each at a concentration of 50 ppm, and the resistivities of the gas sensing membrane in a methane 3000 ppm atmosphere were measured for measuring breakthrough periods till siloxane gases penetrating the filter. When the resistivity decreased to or under an initial resistivity in a methane 500 ppm atmosphere, the inventor deemed that a breakthrough of the filter 10 occurred, in other words, the siloxane gas penetrated the filter 10, and the periods till the breakthrough were measured. The test in the atmosphere containing siloxanes M3, D4, and D5, each 50 ppm, is an extremely accelerated one, and the breakthrough period of 9 days corresponds to the durability of one year in actual use.
Further, the heating period of the sensor 2 was shortened to 1 second, and the sensor 2 was made in contact with an atmosphere containing methane 12500 ppm. The lag time before the resistivity of the gas sensing membrane decreased to or under the resistivity corresponding to methane 3000 ppm was measured and was evaluated as the delay time τ in response. Practically, τ should be equal to or under 50 seconds and is preferably equal to or under 30 seconds.
Preparatory Experiment
As the filter material, five commercial species of granular active carbon (active carbons A-E: all without sulfo group) were tested. In addition to them, a silica gel introduced with sulfo group was tested. A raw silica gel having a BET specific surface area of 500 m2/g, a pore volume of 0.8 cm3/g, and a mean pore diameter of 6.4 nm was mixed with a para-toluene sulfonic acid aqua solution (5 mass % concentration) and then dried at a highest temperature of 140° C. so as to prepare the silica gel introduced with sulfo group. In the example, while the silica gel contained 5 mass % of para-toluene sulfonic acid; the para-toluene sulfonic acid content is arbitrary and it may be changed, for example, within a range of 1 mass % to 15 mass %. Instead of para-toluene sulfonic acid, other sulfonic acid compounds such as naphthalene sulfonic acid and bis-phenol sulfonic acid may be usable and the species of the organic compound for introducing sulfo group is arbitrary. Assuming the whole amount of the mixed para-toluene sulfonic acid was loaded in the silica gel, then, the concentration of sulfo group was 2.4 mass % when 5 mass % para-toluene sulfonic acid was contained. The concentration of sulfo group in the silica gel is, for example, not less than 0.4 mass % and not more than 7 mass %.
The caps where the diameter D of the opening was 4 mm and the diameter R of the filter was 8 mm were filled with one of the filter materials at an amount of 100 mg and the detection delay τ (in second) and the periods before the breakthrough of the filter (day) were measured. The results are shown in Table 1.
While the active carbon E showed the longest period before the breakthrough, the detection delay is nearly at the upper limit, and increasing the amount of active carbon E was, therefore, difficult. The silica gel introduced with sulfo group showed a moderate period before the breakthrough but showed the shortest delay τ in detection. Therefore, the inventor noticed the possibility to elongate the period before the breakthrough and to keep the detection delay τ within 30 seconds by increasing this filter material. Upper limits of the period before the breakthrough are shown in Table 1; they are the expected periods when adjusting the amounts of the filter materials so as to make the detection delay τ to the same as that of the active carbon E. The upper limit is calculated as the period before the breakthrough ÷τ×25 and the constant 25 was determined so that the active carbon E has an upper limit of 21 days. The silica gel introduced with sulfo group showed the longest upper limit before the breakthrough of 70 days.
The 100 mg silica gel filter had a length of 4.4 mm. In order to achieve the upper limit of 70 days, if the length is made five times, it becomes 22 mm. This length hinders the implementation of gas sensor 2 on a substrate. Therefore, the inventor considered to restrict the diameter D of the opening in order to restrict the size of the gas sensor within a practical range. The prerequisite for this was that the detection delay τ was within an allowable range (for example 30 seconds or less) and that the period before the breakthrough is long (for example, 30 days or more and 50 days or more, if possible).
From the filters after the siloxane durability test, the active carbon E and the silica gel in Table 1 were extracted, were heated till 200° C. and the desorbing materials were analyzed by GCMS. Peaks for the siloxanes were detected for the active carbon but they were not detected for the silica gel. This indicates that siloxanes polymerized in the silica gel with the interaction between the sulfo group.
Silica gels introduced with sulfo group where 5 mass % of para-toluene sulfonic acid was added and had mean pore diameters of 4.8 nm and 11 nm were subjected to a similar test to the one in Table 1, where the filter amounts were 100 mg. The periods before breakthrough were 12 days (mean pore diameter of 4.8 nm) and 10 days (mean pore diameter of 11 nm) and the detection delays were 10 seconds (mean pore diameter of 4.8 nm) and 6 seconds (mean pore diameter of 11 nm). On the contrary, the mean pore diameters of the active carbon A-E were from 1.8 to 2.5 nm. Based upon these facts, the inventor estimated that the small mean pore diameter of the active carbons caused the detection delay and that the large mean pore diameter of the silica gel made the polymerization of siloxanes in the pores by the sulfo group.
Experimental
While fixing the length L of the filter at 8 mm, the diameter R at 8 mm, the diameter D of the opening was changed from a conventional value of 4 mm to 0.1 mm (specimens 1-7), where the silica gel content was 0.18 g and it was the silica gel with sulfo group used in the preparatory experiment in Table 1. More, while fixing the diameter D of the opening of the filter at 1.0 mm, the diameter R at 8 mm, the length L of the filter was changed within a range of 12 mm to 4 mm (specimens 8-11). Further, specimens where the values of D and L were randomly changed (specimens 12-16) and specimens where the diameter R was changed (specimens 17-20) were prepared. In the preparation of the specimens, the filter material was filled and tapped in order to make the packing density uniform between the specimens. Then, the detection delay τ and the period before breakthrough were measured for these specimens.
The results are shown in Table 2 and
The period before breakthrough was determined by the ratio L/D of the length L and the diameter D of the opening. When L was fixed at 8 mm, when D was fixed at 1.0 mm, and when both L and D were randomly changed, the same value of L/D resulted in the similar periods before breakthrough. Furthermore, the detection delay τ and the period before breakthrough had a different dependency on the diameter D of the opening. Therefore, when L/D2/3 is made up to 10 and L/D down to 5, the detection delay Σ is made up to 30 seconds and the period before breakthrough is made down to 30 days (lifetime of 3 years or more in actual use). When L/D2/3 is made up to 10 and L/D down to 7, the detection delay τ is made up to 30 seconds and the period before breakthrough is made down to 50 days (lifetime of 5 years or more in actual use). And when L/D2/3 is made up to 6.5 and L/D down to 7, the detection delay τ is made up to 20 seconds and the period before breakthrough is made down to 50 days.
The diameter D of the opening has no intrinsic lower limit, but for making the fabrication easier and for preventing the variation of the diameter of the opening, D is made not smaller than 0.1 mm and not larger than 1.5 mm and is preferably not smaller than 0.3 mm and not larger than 1.2 mm
Table 2 also show results for various diameters R of the filter. When L and D are constant, the increase in R significantly increases the period before breakthrough but does not significantly increase the detection delay τ. Namely, a larger value of R is advantageous. Since a suitable range of gas sensor size is present for easy implementation, R is preferably not smaller than 6 mm and not larger than 16 mm
S′=p/·4R′2 R′=(4S′/p)1/2 and the virtual diameter R′ is used. The filter 10′ is set as a layer, and the thickness of the layer is used as the filter length L.
The inventor has noticed the total area S of the opening for the cases where the opening is not circular or plural openings are present and has confirmed experimentally that (4·S/p)1/2 is usable in place of the diameter D of the opening. This indicates the opening area is important and the shape of circular opening nor the number of the openings is not important. Therefore, L/D is replaceable by L/(4S/π)1/2 and L/D2/3 is replaceable by L/(4S/π)1/3. Therefore, L/(4S/π)1/2 should be 5 or more and L/(4S/π)1/3 should be 10 or smaller. Preferably, L/(4S/π)1/2 is 7 or more. No transformation to the length L of the filter is necessary even when the diameter is not circular, and the lower limit of the diameter D of the opening should be determined according to the processability of D. Therefore, D is set not smaller than 0.1 mm and not larger than 1.5 mm, and L is not smaller than 3 mm and not larger than 10 mm. Preferably, D is set not smaller than 0.3 mm and not larger than 1.2 mm, L not smaller than 3 mm and not larger than 10 mm
In the embodiments, metal oxide semiconductor gas sensors have been described, but the invention is applicable to electrochemical gas sensors and contact combustion-type gas sensors. Further, the invention is similarly applicable to metal oxide semiconductor gas sensors without the mems chip 4.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2016-124212 | Jun 2016 | JP | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/JP2017/017820 | 5/11/2017 | WO | 00 |
