JP2010286380A - Method and apparatus for measuring micropore - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method and an apparatus for simply and quantitatively measuring a micropore of a porous body, especially of a gas diffusion layer or a catalyst layer of a solid polymer fuel cell. <P>SOLUTION: The micropore measuring method includes the steps of: acquiring an electron microscope image of a surface or a cross section of a porous sample; and separating a micropore area from the other area in the electron microscope image, and is characterized by further including the step of removing luminance unevenness from the electron microscope image, which is caused by a combination of an electron microscope and the sample, or by a composition distribution or a shape of the sample. Thus, the micropore is measured precisely. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a method for measuring pores of a porous material, and can be suitably used particularly for evaluation of the pore structure of a catalyst layer or gas diffusion layer of a polymer electrolyte fuel cell.

  Fuel cells that generate power using fuels such as hydrogen are expected to be highly efficient and clean power sources. Among them, the polymer electrolyte fuel cell is small and light and has a low operating temperature, and thus is expected to be widely applied to automobile energy sources and household cogeneration systems.

  The general structure and function of the polymer electrolyte fuel cell will be described below with reference to the drawings.

  FIG. 2 is a schematic diagram illustrating an example of a cross-sectional structure of a single cell that is a basic structural unit of a solid polymer fuel cell. The single cell has a proton conductive electrolyte membrane 4 as a center, a porous anode catalyst layer 3 on one side, and a porous cathode catalyst layer 5 on the other side. The gas diffusion layers 2 and 6 are arranged on the two sides, respectively, and are further sandwiched between the separators 1 and 7 from both sides.

  A fuel gas such as hydrogen is supplied to the anode catalyst layer 3 through the flow path 1a and the gas diffusion layer 2 carved in the separator 1. On the other hand, on the cathode side, a gas containing oxygen is supplied through the flow path 7 a of the separator 7, passes through the gas diffusion layer 6 and reaches the cathode catalyst layer 5. The oxygen that has reached the cathode catalyst layer 5 reacts with the protons that have passed through the electrolyte membrane 4 from the anode side and the electrons that have passed through the gas diffusion layer 6 on the surface of the catalyst particles contained in the cathode catalyst layer 5, It becomes.

  The catalyst layers 3 and 5 of the polymer electrolyte fuel cell play a role of supporting catalyst particles that promote an electrochemical reaction, and generally, a mixture of carbon particles supporting the catalyst particles and an electrolyte polymer, and in the order of nm to μm. It has a complex porous structure including pores. This is because the protons, electrons, gas and water involved in the reaction are supplied and discharged through different paths (electrolytes, carbon particles, pores), so that all of the protons, electrons, and gas can reach the catalyst particles. The area of the upper three-phase interface is made as large as possible, and the reaction of hydrogen gas separated into protons and electrons on the surface of the anode catalyst particles, and the protons and electrons passing through the electrolyte membrane on the surface of the cathode catalyst particles This is because the reaction to react and produce water proceeds smoothly and the produced water is discharged smoothly.

  In order for the perfluorosulfonic acid polymer mainly used as the material of the electrolyte membrane 4 to exhibit sufficient proton conductivity, it is necessary to contain water. As water for maintaining the water-containing state of the electrolyte membrane 4, water vapor is often added to the gas supplied to the anode and the cathode in addition to the water generated by the reaction in the cathode catalyst layer 5.

If the amount of generated water generated at the cathode and the amount of water supplied by gas humidification at the anode is insufficient, the electrolyte membrane 4 is dried, proton conductivity is lowered, and power generation characteristics are lowered. On the other hand, even when the amount of water is excessive, liquid water accumulates in the catalyst layer 5 and the gas diffusion layer 6 and prevents the supply of gas necessary for the reaction, so that the power generation characteristics deteriorate. This is called a flooding phenomenon. Therefore, in order to improve the power generation characteristics of the polymer electrolyte fuel cell, water management is extremely important. For this purpose, the porous pore structures of the catalyst layers 3 and 5 and the gas diffusion layers 2 and 6 are controlled. It is important to. (For example, see Patent Documents 1 and 2.)

  Conventionally, mercury porosimeters have been mainly used for measuring pores in catalyst layers and gas diffusion layers. (For example, see Non-Patent Document 1.) However, the measurement principle of the mercury porosimeter is to determine the pore volume and pore size distribution assuming a cylindrical pore opening on the sample surface. It is inherently unsuitable for measuring the pore distribution of a porous body having a complicated network structure as found in a gas diffusion layer. Further, in order to measure fine pores of the order of 10 nm with a mercury porosimeter, it is necessary to apply a high pressure of about 100 MPa, and it is impossible to accurately measure a sample that is easily broken at a high pressure or a sample that includes a compressible substrate. In addition, a relatively large amount of sample is required for measurement with a porosimeter, and if the catalyst layer has a non-uniform pore structure in the thickness direction or in-plane direction, only an averaged result can be obtained. .

  Therefore, as a more direct means for evaluating the pore structure, a method of observing the surface or cross section of a catalyst layer or a gas diffusion layer with an electron microscope is performed. Unlike the method using a mercury porosimeter, this method is superior in that a complicated structure of pores can be directly observed, but quantitative measurement is difficult by visual evaluation of images.

  In order to solve this problem, image processing is performed on the electron microscope image of the surface or cross section of the porous sample to separate the pore region and the region other than the pore, which is a problem in conventional electron microscope observation. In other words, it is possible to objectively and quantitatively measure the pores by eliminating the subjective judgment by visual inspection. (For example, see Patent Document 3)

  However, the electron microscopic images of the surface and cross section of the actual porous material, especially the catalyst layer and gas diffusion layer of the polymer electrolyte fuel cell, include uneven brightness due to various causes. However, there is a problem that the accuracy of separation between the pore region and the region other than the pores is lowered. Among the unevenness in luminance, the unevenness in luminance caused by the electron microscope optical system and the image sensor can be removed by shading correction (see, for example, Patent Document 4), but the shape or composition distribution of the sample or the combination of the sample and the electron microscope The luminance unevenness caused by the problem still decreased the measurement accuracy.

JP 2005-235525 A JP 2008-91330 A Japanese Patent Laid-Open No. 2-82374 JP 2000-180346 A

Journal of the Electrochemical Society (USA), 2004, Vol. 151, No. 11, p. A1841-A1846 IEICE Transactions D, 1980, Vol. 63, No. 4, p. 349-356 Pattern Recognition (Netherlands), 1986, Vol. 19, No. 1, p. 41-47

The present invention has been made to solve the above-described conventional problems, and is a method for easily and quantitatively measuring the pores of a porous body, particularly a catalyst layer or a gas diffusion layer of a solid polymer fuel cell. And an apparatus.

  As means for solving the above-described problems in the present invention, a step of obtaining an electron microscope image of the surface or cross section of the porous sample, a step of separating the pore region and the region other than the pores of the electron microscope image, Including a step of removing luminance unevenness due to the shape or composition distribution of the porous sample, or luminance unevenness due to a combination of the porous sample and the electron microscope, from the electron microscope image. This is a method for measuring pores.

  The invention according to claim 2 is the pore measuring method according to claim 1, wherein a spatial frequency filter is used in the step of removing the luminance unevenness.

  The invention according to claim 3 is the pore measuring method according to claim 2, wherein the spatial frequency filter is a spatial frequency filter using fast Fourier transform.

  The invention according to claim 4 is a step of separating the pore region and the region other than the pores of the electron microscope image by binarization processing using the maximum entropy method, the Otsu method, or the Kittler method. It is isolate | separated, It is a pore measuring method as described in any one of Claims 1 thru | or 3 characterized by the above-mentioned.

  In addition, the invention according to claim 5 is a fine combination of means for acquiring an electron microscope image of the surface or cross section of a porous sample, and means for separating a pore region and a region other than the pores of the electron microscope image. The pore measuring apparatus includes means for removing, from the electron microscope image, luminance unevenness due to the shape or composition distribution of the porous sample, or luminance unevenness due to a combination of the porous sample and the electron microscope. It is a pore measuring device.

  The invention described in claim 1 includes a step of removing luminance unevenness caused by the shape or composition distribution of the sample or the combination of the sample and the electron microscope from the electron microscope image. As a result, the brightness unevenness seen when SEM images are taken in the hole dug in the catalyst layer surface by FIB processing, the brightness unevenness due to the composition distribution of the region other than the pores seen on the catalyst layer surface or cross section, the electron Since the influence of luminance unevenness due to charge-up and edge effects that are often included in microscopic images can be removed, the measurement of pores can be performed with high accuracy.

  In the invention according to claim 2, since the spatial frequency filter is used as the step of removing the luminance unevenness, the luminance unevenness changing at a spatial scale different from the representative pore size can be efficiently removed. The pores can be measured with higher accuracy.

  In the invention according to claim 3, by using a spatial frequency filter using fast Fourier transform (FFT) as the spatial frequency filter, it is possible to measure the pores with higher accuracy and higher speed.

  In the invention according to claim 4, by using the binarization processing by the maximum entropy method or the method of Otsu or the method of Kittler as the step of separating the pore region and the region other than the pore of the electron microscope image, Since the binarization threshold can be determined without being affected by the brightness, contrast, noise, and the like of the microscopic image, the pores can be measured with higher accuracy.

  In the invention according to claim 5, any of the above-described effects can be obtained by using the method according to any one of claims 1 to 4.

  As described above, the present invention has an effect that the pores of the porous body, particularly the catalyst layer or gas diffusion layer of the solid polymer fuel cell can be measured easily and quantitatively.

The flowchart which shows the process sequence of this invention Example 2. Cross-sectional schematic diagram of a single polymer electrolyte fuel cell unit cell The figure which shows the SEM image of the catalyst layer surface in this invention Example 1 The figure which shows the 1st binarization result of the catalyst layer surface SEM image in this invention Example 1. The figure which shows the 2nd binarization result of the catalyst layer surface SEM image in this invention Example 1. The figure which shows the result of having removed the brightness nonuniformity from the catalyst layer surface SEM image in this invention Example 1. The figure which shows the binarization result of the image which removed the brightness nonuniformity from the catalyst layer surface SEM image in this invention Example 1. The figure which shows the result of having measured the pore area ratio of the catalyst layer surface using the method of this invention with respect to 4 samples including the sample of this invention Example 1. FIG. The figure which shows the catalyst layer cross-section SEM image in this invention Example 2. The figure which shows the 1st binarization result of the catalyst layer cross section SEM image in this invention Example 2. The figure which shows the binarization result of the image which removed the brightness nonuniformity from the catalyst layer cross-section SEM image in this invention Example 2. The figure which shows the cross-sectional pore area ratio distribution measured from the SEM image of a total of 10 sheets containing the catalyst layer cross-sectional SEM image in this invention Example 2. FIG.

  Various embodiments of the present invention will be described below with reference to the drawings.

  The method of the present invention includes obtaining an electron microscope image of the surface or cross section of a porous sample. Various known methods such as focused ion beam (FIB) processing, microtome processing, or freeze fracture can be used as a method for obtaining a cross section of the sample. FIB processing is particularly suitable due to features such as small disturbance. In FIB processing or microtome processing, cooling with liquid nitrogen or the like may be performed to reduce damage to the cross-sectional pores due to heat. A method called cryo FIB processing or cryomicrotome processing corresponds to this. Further, in order to prevent damage to the cross-sectional pores or to easily distinguish the pore region and the region other than the pores in the cross-sectional microscopic image, pre-processing such as resin embedding may be performed before the cross-sectional view. . As the electron microscope, a scanning electron microscope (SEM), a transmission electron microscope (TEM), a scanning transmission electron microscope (STEM), or the like can be used.

  The method of the present invention may include a shading correction step of removing luminance unevenness caused by an electron microscope optical system or an image sensor from an electron microscope image.

  Various image processing filters can be used as a step of removing luminance unevenness caused by the shape or composition distribution of the sample or the combination of the sample and the electron microscope from the electron microscope image. For example, smoothing filters such as median filters and Gaussian filters are used to remove brightness unevenness that changes on a smaller spatial scale than typical pore sizes, such as charge-up during electron microscope image acquisition and brightness unevenness due to edge effects. Is suitable.

On the other hand, in the electron microscope image, luminance unevenness changing on a spatial scale larger than a typical pore size is often seen. FIG. 9 is an SEM image (magnification is about 10,000 times) of a cross section of the solid polymer fuel cell catalyst layer cut by FIB processing. The pore cross section 12 and the pore cross section 13 seen in the image are cross sections of the pores forming a three-dimensional network structure in the catalyst layer, and the brightness appears low in the SEM image. On the other hand, the solid cut surface 14 in the SEM image of FIG. 9 is a cut surface formed by cutting the solid constituting the catalyst layer with FIB, and the luminance looks higher than the pore cross sections 12 and 13. The solid cut surface 15 is a solid cut surface that is cut by FIB in the same manner as the solid cut surface 14, but the average brightness of the solid cut surface 15 is lower than that of the solid cut surface 14. This is presumably because the local composition (carbon particle / electrolyte ratio, etc.) is different between the solid cut surface 14 and the solid cut surface 15, and the secondary electron emission rate and the ease of charge-up are different. In addition, the SEM image in FIG. 9 has luminance unevenness as a whole, and the average luminance is lower in the lower part of the image, particularly in the lower left part (around the solid cut surface 16), compared to the upper part of the image. This is considered to be due to the influence of the sample geometry in addition to the composition distribution. A spatial frequency filter such as a high-pass filter or a band-pass filter is effective as a process for removing luminance unevenness that changes on a spatial scale larger than the typical pore size, and uses a fast Fourier transform (FFT). The spatial frequency filter used is particularly suitable.

  As a process of separating the pore region and the region other than the pore of the electron microscope image, the percentile method, the mode method, the maximum entropy method, the Otsu method (for example, refer to Non-Patent Document 2), the Kittler method (for example, the non-patent method) Binarization processing using various known threshold determination methods such as a differential histogram method and a Laplacian histogram method can be used. However, it depends on the brightness, contrast, noise, etc. of the original image. Since a stable result that is difficult to be obtained is obtained, a binarization process using the maximum entropy method, the Otsu method, or the Kittler method is preferable. It is also possible to divide the entire image into a plurality of regions and perform binarization processing by applying the above-described threshold value determination method or the like for each region. In addition, as in the method of Patent Document 3, an image is divided into a large number of regions using edges extracted using a differential filter or the like, and attention is paid to the average luminance, the shape of a histogram, the edge strength of the contour, etc. for each region. Thus, it is possible to determine whether the region is a pore region or a region other than the pore.

  Embodiment 1 of the present invention will be described below with reference to the drawings.

  FIG. 3 is an image (magnification is about 5,000 times) of the cathode catalyst layer surface of the polymer electrolyte fuel cell taken from the normal direction of the catalyst layer surface by SEM (manufactured by Hitachi High-Technologies Corporation). This catalyst layer is composed of carbon particles carrying platinum-based catalyst particles and an electrolyte (perfluorosulfonic acid polymer), and is formed with a thickness of about 5 μm on both surfaces of the electrolyte membrane (perfluorosulfonic acid polymer).

  A low-luminance region (such as the pore 8) having a clear outline, which is often seen in the SEM image of FIG. 3, is a pore that appears on the surface of the catalyst layer. On the other hand, the region having a relatively high luminance like the solid surface 9 is the surface of the solid (mixture of carbon particles and electrolyte) constituting the catalyst layer. On the solid surface, luminance unevenness that is considered to be caused by the local composition distribution is observed. For example, the average luminance is low around the solid surface 10, but the contour of the low luminance region is blurred, and there are regions (such as the pores 11) seen as pores in the region. It can be seen that the region is not a pore but a solid surface.

  Focusing on the difference in brightness between the pore area (pore 8 etc.) and the solid surface area (solid surface 9 etc.) in the SEM image of FIG. 3, the result of an attempt to discriminate both areas by binarization processing is shown in FIG. It is. Although the pores 8 and the solid surface 9 can be discriminated almost correctly, the periphery of the solid surface 10 is erroneously discriminated as pores due to uneven brightness.

FIG. 5 shows the result of binarizing the SEM image of FIG. 3 as in FIG.
The threshold value of the binarization process is adjusted so that the periphery of is correctly identified as a solid surface. However, in this case, a considerable portion of the pore region disappears or shrinks, and it is difficult to accurately measure the pore area ratio and the like from this result.

  FIG. 6 is an image obtained by applying fast Fourier transform (FFT) to the SEM image of FIG. 3 and removing the luminance fluctuation component having a period of 0.5 μm or more and then performing inverse transform. By this treatment, the luminance unevenness considered to be caused by the local composition distribution on the solid surface was almost completely removed. Further, FIG. 7 shows a result of performing binarization processing on the image of FIG. 6 using a threshold determined by the method of Otsu after removing noise using a median filter (equivalent to a radius of 0.01 μm). . Through these series of image processing, the pores (pores 8, 11 etc.) and the solid surface (solid surfaces 9, 10, etc.) could be correctly discriminated. FIG. 8 shows the result of applying the same measurement method to a plurality of samples prepared by changing the conditions of the manufacturing process of the catalyst layer and the blending of the materials. The standard sample in FIG. 8 is the sample of Example 1. In this way, we were able to obtain useful knowledge about the influence of the catalyst layer manufacturing process conditions and material composition on the pore area ratio of the catalyst layer surface, and further on the power generation characteristics of the fuel cell, especially the flooding characteristics. .

  Embodiment 2 of the present invention will be described below with reference to the drawings.

  FIG. 1 is a flowchart showing an outline of a pore measurement procedure in Example 2.

  FIG. 9 is a SEM image (magnification is about 10,000 times) of a cross section of the catalyst layer of the polymer electrolyte fuel cell cut by FIB processing as described above. This catalyst layer is composed of carbon particles carrying platinum-based catalyst particles and an electrolyte (perfluorosulfonic acid polymer), and is formed with a thickness of about 5 μm on both surfaces of the electrolyte membrane (perfluorosulfonic acid polymer). A hole (approximately 10 μm square, depth approximately 7 μm) is dug in the surface of the electrolyte membrane with a catalyst layer by FIB processing, a platinum conductive film is deposited on the side wall cross section, and oblique (45 degrees with respect to the normal of the cross section to be observed) 9), the image taken with SEM (manufactured by Hitachi High-Technologies Corp.) was stretched vertically by 1.414 times to correct the tilt, and then the noise was removed using a median filter (equivalent to a radius of 0.01 μm). It is a SEM image of. The upper side of the SEM image corresponds to the surface direction of the catalyst layer, and the lower side of the image corresponds to the bottom direction of the hole. In this SEM image, as described above, luminance unevenness considered to be caused by local composition distribution, charge-up at the time of SEM imaging, geometrical shape of the sample, etc. is observed.

  In the case of an image without such uneven brightness, binarization processing is performed using a threshold value determined by an appropriate method, so that the pore cross section and the solid cut surface are distinguished, and the area ratio of the pore cross section is calculated. The pore distribution can be obtained by obtaining or counting the pores by size.

  FIG. 10 shows the result of binarization processing using the threshold value determined using the maximum entropy method on the SEM image of FIG. The pore cross section 12 and the pore cross section 13 are black because they are determined as pore cross sections, and the solid cut surface 14 is white because it is determined as a solid cut surface. However, due to the luminance unevenness included in the SEM image of FIG. 9, the solid cut surface 15 that should be originally identified as a solid cut surface is almost entirely black and has been misclassified as a pore cross section. Recognize. Similarly, in the SEM image of FIG. 9, the lower part of the image, particularly the lower left part (the periphery of the solid cut surface 16) is almost black, and the pore cross section and the solid cut surface are not correctly discriminated. I understand that.

On the other hand, FIG. 11 is determined using the maximum entropy method on the image obtained by performing fast Fourier transform (FFT) on the SEM image of FIG. 9 and removing the luminance fluctuation component having a period of 1 μm or more and then performing inverse transform. It is the result of performing the binarization process by the threshold value. By using the FFT high-pass filter, luminance unevenness that changes on a spatial scale larger than the typical pore size can be almost eliminated, so that the pore cross-section (pore cross-section 12, 13 etc.) and the solid cut surface (solid The cut surfaces 14, 15, 16, etc.) could be correctly identified.

  A total of ten SEM images obtained by the same method using the same sample as the SEM image of FIG. 9 are binarized by performing the same series of image processing as described for FIG. The result of obtaining the distribution of the cross-sectional pore area ratio by performing the counting process is the graph of FIG. In this way, it is possible to measure the pore distribution from the cross-sectional SEM image of the polymer electrolyte fuel cell catalyst layer, and the effects of the catalyst layer manufacturing process conditions and material composition on the pore structure of the catalyst layer In addition, useful information was obtained about the influence of the material transport in the catalyst layer on the power generation characteristics of the fuel cell.

  The present invention can be generally used for measurement and evaluation of the pore structure of a porous body, but is particularly suitable for measurement and evaluation of the pore structure of a catalyst layer or a gas diffusion layer of a polymer electrolyte fuel cell.

DESCRIPTION OF SYMBOLS 1 ... Separator 1a ... Flow path 2 ... Gas diffusion layer 3 ... Anode catalyst layer 4 ... Electrolyte membrane 5 ... Cathode catalyst layer 6 ... Gas diffusion layer 7 ... Separator 7a ··· Channel 8 ··· Pore 9 ··· Solid surface 10 · · Solid surface 11 · · Pore 12 · · Pore cross section 13 · · Pore cross section 14 · · Solid cut surface 15 · · Solid cut surface 16. Solid cut surface

Claims (5)

  In the pore measurement method comprising the steps of obtaining an electron microscope image of the surface or cross section of the porous sample, and separating the pore region and the non-pore region of the electron microscope image, from the electron microscope image, A method for measuring pores, comprising a step of removing luminance unevenness due to a shape or composition distribution of the porous sample or luminance unevenness due to a combination of the porous sample and an electron microscope.   2. The pore measuring method according to claim 1, wherein a spatial frequency filter is used in the step of removing the luminance unevenness.   The pore measuring method according to claim 2, wherein the spatial frequency filter is a spatial frequency filter using a fast Fourier transform.   The step of separating the pore region and the region other than the pore of the electron microscope image is separated by a binarization process using a maximum entropy method, an Otsu method, or a Kittler method. 4. The pore measuring method according to any one of 3 above.   In the pore measuring apparatus having both means for obtaining an electron microscope image of the surface or cross section of the porous sample and means for separating the pore region of the electron microscope image and the region other than the pore, from the electron microscope image, A pore measuring apparatus comprising means for removing luminance unevenness due to the shape or composition distribution of the porous sample or luminance unevenness due to a combination of the porous sample and an electron microscope.