DETAILED DESCRIPTION OF THE INVENTION

[0080] Embodiments in which the present invention is applied to a surface conduction type electron-emitting apparatus (to be referred to as an SED hereinafter) as a flat panel display apparatus will be described in detail.

[0081] As shown in FIGS. 1 to 3 , this SED has a rear plate 10 and face plate 12 respectively formed of a rectangular glass members. These plates are arranged to oppose each other at a gap of about 1.5 mm to 3.0 mm. The rear plate 10 is formed with a size slightly larger than that of the face plate 12 . The rear plate 10 and face plate 12 are joined at their peripheries through a rectangular frame-like side wall 14 made of glass, thus constituting a flat rectangular vacuum envelope 15 . The side wall 14 is adhered with frit glass, a low-melting metal or alloy such as indium or indium alloy. For example, the internal space of the vacuum envelope is maintained at a high vacuum of about 10 ⁻⁸ Torr.

[0082] A phosphor screen 16 is formed on the inner surface of the face plate 12 . The phosphor screen 16 is formed by lining red, blue, and green phosphor layers and black-colored layers. These phosphor layers are formed in stripes or dots. A metal back 17 made of aluminum or the like is formed on the phosphor screen 16 . A transparent conductive film made of, e.g., ITO, or a color film may be arranged between the face plate 12 and the phosphor screen.

[0083] A number of electron-emitting elements 18 for respectively emitting electron beams are formed on the inner surface of the rear plate 10 to serve electron-emitting sources for exciting the phosphor layers. The electron-emitting elements 18 are arranged in a plurality of columns and a plurality of rows to correspond to respective pixels. Each electron-emitting element 18 is comprised of an electron-emitting portion (not shown), a pair of element electrodes for applying a voltage to the electron-emitting portion, and the like. A number of interconnections (not shown) for applying a voltage to the electron-emitting elements 18 are formed on the rear plate 10 in a matrix.

[0084] The side wall 14 serving as the joining member is adhered to the peripheries of the rear plate 10 and frit glass 20 through frit glass 20 made of, e.g., low-melting glass, to join the face plate and rear plate to each other.

[0085] As shown in FIGS. 2 and 3 , the SED has a spacer assembly 22 arranged between the rear plate 10 and face plate 12 . In this embodiment, the spacer assembly 22 has a plate-like grid 24 and a plurality of columnar spacers standing on the two surfaces of the grid to be integral with them.

[0086] More specifically, the grid 24 has a first surface 24 a opposing the inner surface of the face plate 12 , and a second surface 24 b opposing the inner surface of the rear plate 10 , and is arranged parallel to these plates. A number of apertures 26 and a plurality of spacer holes 28 are formed in the grid 24 by etching or the like. The apertures 26 are disposed to oppose the electron-emitting elements 18 , and the spacer holes 28 are located between the apertures and disposed at a predetermined pitch.

[0087] The grid 24 is formed from a metal plate of, e.g., an iron-nickel-based metal, to a thickness of 0.1 mm to 0.25 mm, and an oxide film made of elements constituting the metal plate, e.g., an oxide film made of Fe ₃ O ₄ or NiFe ₃ O ₄ , is formed on its surface. Each aperture 26 has a rectangular shape of 0.15 mm to 0.25 mm×0.20 mm to 0.40 mm, and each spacer hole 28 has a diameter of about 100 μm to 200 μm.

[0088] First spacers 30 a integrally stand on the first surface 24 a of the grid 24 to overlap the respective spacer holes 28 , and their extending ends abut against the inner surface of the face plate 12 through the metal back 17 and the black-colored layers of the phosphor screen 16 . Second spacers 30 b integrally stand on the second surface 24 b of the grid 24 to overlap the respective spacer holes 28 , and their extending ends abut against the inner surface of the rear plate 10 . The spacer holes 28 and the first and second spacers 30 a and 30 b are aligned with each other, and the first and second spacers are integrally connected to each other through the corresponding spacer holes 28 .

[0089] Each of the first and second spacers 30 a and 30 b integrally has a plurality of steps stacked from the grid 24 toward the extending end and with gradually decreasing diameters. Each step is formed in a tapered manner to be thinner from the grid toward the extending end. In other words, each of the first and second spacers 30 a and 30 b is formed into a stepped tapered shape or stepped truncated-conical shape.

[0090] For example, each first spacer 30 a forms a stepped tapered shape with four steps such that its end on the grid 24 side has a diameter of about 400 μm, its end on the extending end side has a diameter of about 230 μm, and its height is about 1 mm to 1.2 mm, resulting in an aspect ratio (height/grid-side end diameter) of 2.5 to 3.0. Each second spacer 30 b forms a stepped tapered shape with three steps such that its end on the grid 24 side has a diameter of about 400 μm, its end on the extending end side has a diameter of about 280 μm, and its height is about 0.3 mm to 0.75 mm, resulting in an aspect ratio (height/grid-side end diameter) of 0.75 to 1.6.

[0091] As described above, the diameter of each spacer hole 28 is about 100 μm to 200 μm, which is sufficiently smaller than the grid-side end diameters of the first and second spacers 30 a and 30 b. When the first and second spacers 30 a and 30 b are integrally formed to be coaxially aligned with the spacer holes 28 , the first and second spacers are connected to each other through the spacer holes, and are integrally formed with the grid 24 to sandwich it from its two surfaces.

[0092] A predetermined voltage is applied from a power supply (not shown) to the grid 24 of the spacer assembly 22 with the arrangement as described above so as to prevent crosstalk, and to converge the electron beams emitted from the corresponding electron-emitting elements 18 onto desired phosphor layers with the apertures 26 . The first and second spacers 30 a and 30 b abut against the inner surfaces of the face plate 12 and rear plate 10 to support the load of the atmospheric pressure acting on these plates, and maintain the gap between the plates at a predetermined value.

[0093] A method of manufacturing the spacer assembly 22 with the above arrangement, and the SED with the spacer assembly 22 will be described.

[0094] When the spacer assembly 22 is to be manufactured, first, as shown in FIG. 4 , the grid 24 with a predetermined size, and rectangular plate-like first and second molds 32 and 33 with sizes almost equal to that of the grid are prepared. Apertures 26 and spacer holes 28 are formed in the grid 24 in advance, and the entire outer surface of the grid 24 is covered with a blackened film or an oxide film made of granular oxides.

[0095] The first and second molds 32 and 33 have a plurality of through holes 34 corresponding to the spacer holes 28 of the grid 24 . As shown in FIG. 5 , the first mold 32 is formed by stacking a plurality of metal thin plates, e.g., 4 metal thin plates 32 a, 32 b, 32 c, and 32 d.

[0096] This will be described in detail. Each metal thin plate is formed of an iron-based metal plate with a thickness of 0.25 mm to 0.3 mm, and has a plurality of tapered through holes. The through holes formed in each of the metal thin plates 32 a, 32 b, 32 c, and 32 d have diameters different from those of the through holes formed in other metal thin plates. For example, the metal thin plate 32 a has tapered through holes 34 a with a maximum diameter of 400 μm, the metal thin plate 32 b has tapered through holes 34 b with a maximum diameter of 350 μm, the metal thin plate 32 c has tapered through holes 34 c with a maximum diameter of 295 μm, and the metal thin plate 32 d has tapered through holes 34 d with a maximum diameter of 240 μm. These through holes 34 a to 34 d are formed by etching or laser radiation.

[0097] The four metal thin plates 32 a, 32 b, 32 c, and 32 d are stacked such that their through holes 34 a, 34 b, 34 c, and 34 d are aligned almost coaxially and sequentially from the ones with larger diameters, and are diffusion-bonded to each other in vacuum or a reducing atmosphere. Hence, the first mold 32 with a thickness of 1.0 mm to 1.2 mm as a whole is formed. Each through hole 34 is defined by aligning the four through holes 34 a, 34 b, 34 c, and 34 d, and has a stepped, tapered inner surface.

[0098] The second mold 33 is also formed by stacking, e.g., three metal thin plates in the same manner as the first mold 32 . Each through hole 34 is defined by three tapered through holes and has a stepped, tapered inner surface.

[0099] The outer surfaces of the first and second molds 32 and 33 , including the inner surfaces of the respective through holes 34 , are covered with surface layers. Each surface layer has releasability with respect to a spacer forming material (to be described later) and resistance to oxygen, and is formed by, e.g., eutectic plating of Ni—P with fine particles of Teflon, oxide, nitride, or carbide, or eutectic plating of Ni—P with a high-melting metal such as W, Mo, or Re.

[0100] In the steps of manufacturing the spacer assembly, as shown in FIG. 6 A, the first mold 32 is brought into tight contact with the first surface 24 a of the grid so the large-diameter sides of the through holes 34 are located on the grid 24 side, and is positioned such that the through holes are aligned with the spacer holes 28 of the grid. Similarly, the second mold 33 is brought into tight contact with the second surface 24 b of the grid so the large-diameter sides of the through holes 34 are located on the grid 24 side, and is positioned such that the through holes are aligned with the spacer holes 28 of the grid. The first mold 32 , grid 24 , and second mold 33 are fixed to each other by using a damper (not shown) or the like.

[0101] Subsequently, as shown in FIG. 6B, a spacer forming material 40 in the form of paste is supplied from, e.g., the outer surface of the first mold 32 , by using a squeegee 36 , to fill the through holes 34 of the first mold 32 , the spacer holes 28 of the grid 24 , and the through holes 34 of the second mold 33 . An excessive portion of the spacer forming agent 40 leaking to the outer surface of the second mold 33 is scraped off by using a squeegee 38 .

[0102] As the spacer forming material 40 , a glass paste containing at least an ultraviolet-curing binder (organic component) and a glass filler is used.

[0103] Subsequently, as shown in FIG. 6 C, the filling spacer forming material 40 is irradiated with ultraviolet rays (UV) as a radiation from the outer surfaces of the first and second molds 32 and 33 , so it is UV-cured. When the spacer forming material 40 is UV-cured in this manner, adhesion of the spacer forming material with respect to the grid 24 is increased to be higher than that of the spacer forming material with respect to the first and second molds 32 and 33 .

[0104] As shown in FIG. 7 A, while the first and second molds 32 and 33 are in tight contact with the grid 24 , they are heat-treated in a heating furnace to evaporate the binder from the spacer forming material 40 , and then the spacer forming material is properly calcined at about 500° C. to 550° C. for 30 minutes to 1 hour. Hence, the first and second spacers 30 a and 30 b integral with the grid 24 are formed.

[0105] After that, the first and second molds 32 and 33 and the grid 24 are cooled to a predetermined temperature, and the first and second molds 32 and 33 are released from the grid 24 , as shown in FIG. 7B . Hence, the spacer assembly 22 is completed.

[0106] When the SED is to be manufactured by using the spacer assembly 22 manufactured in the above manner, the rear plate 10 having the electron-emitting elements 18 and bonded with the side wall 14 , and the face plate 12 having the phosphor screen 16 and metal back 17 are prepared in advance. The spacer assembly 22 is positioned on the rear plate 10 , and the rear plate and the face plate 12 are arranged in a vacuum chamber. The interior of the vacuum chamber is evacuated, and the face plate 12 is bonded to the rear plate 10 through the side wall 14 . Hence, an SED with the spacer assembly 22 is manufactured.

[0107] According to the spacer assembly 22 with the above arrangement and the SED having the spacer assembly 22 , each spacer integrally has a plurality of steps stacked toward the extending end and with gradually decreasing diameters. Each step is formed in a tapered manner to be thinner toward the extending end, thus resulting in a stepped tapered shape as a whole, i.e., a substantial stepped truncated-conical shape. Therefore, a plurality of spacers can be integrally built on the grid by mold forming, and an easily manufacturable spacer assembly and SED can be obtained.

[0108] According to the spacer assembly manufacturing method described above, after the spacer forming material is arranged on the grid by using a mold, it is calcined, so that a plurality of spacers can be built at predetermined positions on the grid at once. Therefore, a spacer assembly with a plurality of small spacers can be manufactured easily, thus achieving a reduction in manufacturing cost and an increase in manufacturing efficiency.

[0109] Since the spacer forming material is calcined as it fills the through holes of the mold, the spacer forming material will not be squeezed and does not spread during calcination, and spacers each with a sufficiently large height and a high aspect ratio can be formed easily.

[0110] According to this embodiment, a glass paste containing an ultrasonic-curing binder and glass filler is used as the space forming material. Prior to calcination, the spacer forming material is cured by irradiation with ultraviolet rays. Also, the grid covered with an oxide film and molds covered with oxygen-resistant surface layers are used, so adhesion of the spacer forming material with respect to the grid can be increased to be higher than that with respect to the molds. Therefore, in the following calcining and releasing steps, the formed spacers are prevented from attaching to the molds, and spacers integral with the grid can be formed reliably.

[0111] According to this embodiment, each mold is formed by stacking a plurality of metal thin plates each having through holes. Usually, it is very difficult to form spacer-forming small through holes with a diameter of several 100 μm in a metal plate with a thickness of about 1 mm or more. If the metal thin plate has a thickness of about 0.1 mm to 0.3 mm, small through holes can be formed in it comparatively easily by etching, laser radiation, or the like. Therefore, when a plurality of metal thin plates having through holes are stacked as in this embodiment, a mold having through holes with a desired height can be obtained easily.

[0112] In this mold, the through holes formed in each metal thin plate are tapered, and their diameters differ from one metal thin plate to another. Hence, when stacking these plurality of metal thin plates, even if they are slightly misaligned, the through holes of the respective metal thin plates can communicate with each other reliably, so a mold having desired through holes can be obtained. Furthermore, since the mold is covered with a surface layer having releasability with respect to the spacer forming material, the spacer forming material will not easily attach to the interiors of the through holes of the mold. Thus, the mold can be repeatedly used for the manufacture of the spacer assembly.

[0113] An SED having a spacer assembly according to a second embodiment of the present invention, and a method of manufacturing this spacer assembly will be described.

[0114] As shown in FIG. 8 , according to the second embodiment, a grid 24 of a spacer assembly 22 has no spacer holes, and first and second spacers 30 a and 30 b are integrally formed with the grid 24 to be independent of each other.

[0115] More specifically, the plurality of first spacers 30 a stand vertically on a first surface 24 a of the grid 24 between apertures 26 , and abut against the inner surface of a face plate 12 through a metal back 17 and the black-colored layers of a phosphor screen 16 . The plurality of second spacers 30 b stand on a second surface 24 b of the grid 24 between the apertures 26 , abut against the inner surface of a rear plate 10 , and are arranged to be aligned with the first spacers 30 a. Other arrangements are the same as those of the SED according to the first embodiment described above. The same portions are denoted by the same reference numerals, and a detailed description thereof will be omitted.

[0116] To manufacture a spacer assembly 22 with the above arrangement, first, as shown in FIG. 9A, a first mold 32 is brought into tight contact with the first surface 24 a of the grid such that the large-diameter sides of through holes 34 are located on the grid 24 side, and is positioned such that the respective through holes are located between the apertures 26 of the grid. Subsequently, a spacer forming material 40 in the form of paste is supplied from the outer surface of the first mold 32 by using a squeegee 36 , to fill the through holes 34 of the first mold 32 . As the spacer forming material 40 and first mold 32 , those that are identical to those of the embodiment described above are used.

[0117] Subsequently, as shown in FIG. 9 B, the spacer forming material 40 which fills the through holes 34 is irradiated with ultraviolet rays (UV) from the outer surface of the first mold 32 , so it is UV-cured. Thus, adhesion of the spacer forming material 40 with respect to the grid 24 is increased to be higher than that of the spacer forming material with respect to the first mold 32 .

[0118] After that, as shown in FIG. 10 A, while the grid 24 and first mold 32 are held in tight contact with each other, a second mold 33 is brought into tight contact with the second surface 24 b of the grid such that the large-diameter sides of through holes 34 are located on the grid 24 side, and is positioned such that the respective through holes are located between the apertures 26 of the grid. Then, the first mold 32 , grid 24 , and second mold 33 are fixed to each other by using a clamper (not shown) or the like.

[0119] Subsequently, the spacer forming material 40 in the form of paste is supplied from the outer surface of the first mold 32 by using the squeegee 36 , to fill the through holes 34 of the second mold 32 . As the second mold 33 , one which is identical to that of the embodiment described above is used.

[0120] After that, as shown in FIG. 10 B, the spacer forming material 40 which fills the through holes 34 is irradiated with ultraviolet rays from the outer surface of the second mold 33 , so it is UV-cured. Thus, adhesion of the spacer forming material 40 with respect to the grid 24 is increased to be higher than that of the spacer forming material with respect to the second mold 33 .

[0121] As shown in FIG. 10 C, while the first and second molds 32 and 33 are in tight contact with the grid 24 , they are heat-treated in the heating furnace to evaporate the binder from the spacer forming material 40 , and then the spacer forming material is properly calcined at about 500° C. to 550° C. for 30 minutes to 1 hour. Hence, first and second spacers 30 a and 30 b integral with the grid 24 are formed.

[0122] Then, the first and second molds 32 and 33 and the grid 24 are cooled to a predetermined temperature, and the first and second molds 32 and 33 are released from the grid 24 , thereby completing a spacer assembly 22 . An SED having the spacer assembly 22 with the above arrangement is manufactured in accordance with steps similar to those of the embodiment described above.

[0123] In the second embodiment with the above embodiment, the same function and effect as those of the embodiment described above can also be obtained.

[0124] In the first and second embodiments described above, the spacer assembly integrally has the first and second spacers on the two surfaces of the grid 24 . Alternatively, spacers may be formed on only one surface of the grid, as in the third embodiment shown in FIG. 11 .

[0125] More specifically, according to the third embodiment, a spacer assembly 22 has a grid 24 and a plurality of first spacers 30 a integrally standing vertically on a first surface 24 a of the grid. The first spacers 30 a stand between apertures 26 , and abut against the inner surface of a face plate 12 through a metal back 17 and black-colored layers of a phosphor screen 16 .

[0126] A plurality of second spacers 30 b integrally stand on the inner surface of the face plate 12 , are positioned to be aligned with the first spacers 30 a, and abut against a second surface 24 b of the grid 24 .

[0127] The first and second spacers 30 a and 30 b are formed into a stepped tapered shape to be thinner toward the extending end, i.e., into a stepped truncated-conical shape, in the same manner as in the embodiment described above. Other arrangements of the SED are the same as those of the embodiment described above. The same portions are denoted by the same reference numerals, and a detailed description thereof will be omitted.

[0128] In the third embodiment, the spacer assembly 22 is manufactured by the same method as that of the second embodiment described above. Note that the manufacturing steps for the second spacers will be omitted.

[0129] In the third embodiment, the plurality of second spacers 30 b make up another spacer assembly 22 b together with a glass substrate which forms the face plate 12 . The spacer assembly 22 b is also manufactured by the same method as that of the second embodiment.

[0130] More specifically, in place of the grid, the second mold 33 described above is brought into tight contact with the surface of the rear plate 10 formed of a glass substrate such that the large-diameter sides of through holes 34 are located on the rear plate side, and is positioned at a predetermined position. Subsequently, a spacer forming material in the form of paste is supplied from the outer surface of the second mold 33 to fill the through holes 34 of the second mold 32 . As the spacer forming material, one which is identical to that of the embodiment described above is used.

[0131] Subsequently, the spacer forming material which fills the through holes is irradiated with ultraviolet rays, so it is UV-cured. Thus, adhesion of the spacer forming material with respect to the rear plate 10 is increased to be higher than that of the spacer forming material with respect to the second mold 33 . The glass substrate that forms the rear plate 10 need not have an oxide film on its outer surface, since it is an oxide itself.

[0132] While the second mold 33 is in tight contact with the rear plate 10 , this structure is heat-treated in a heating furnace to evaporate the binder from the spacer forming material, and then the spacer forming material is properly calcined. Hence, second spacers 30 a integral with the rear plate 10 are formed.

[0133] Then, the second mold 33 and rear plate 10 are cooled to a predetermined temperature, and the second mold 33 is released, thereby completing a spacer assembly 22 b integrally having the rear plate 10 and second spacers 30 b.

[0134] In the third embodiment with the above arrangement as well, the same function and effect as those of other embodiments described above can be obtained.

[0135] An SED according to a fourth embodiment of the present invention will be described. As shown in FIG. 12, a spacer assembly 22 (to be described later) is disposed between a face plate 12 and rear plate 10 which constitute a vacuum envelope 15 , and has an electrode plate connected to a predetermined potential in order to prevent abnormal discharge between these plates. The face plate 12 and rear plate 10 are supported by the spacer assembly 22 against the atmospheric pressure, and the predetermined gap between the plates is maintained at, e.g., 1.6 mm.

[0136] The face plate 12 has an insulating substrate made of non-alkali glass, and a phosphor screen 16 formed on the inner surface of the insulating substrate. The phosphor screen 16 has striped phosphor layers 13 having red (R), blue (B), and green (G) light-emitting characteristics and arranged at a 0.6-mm pitch, and band-like light-shielding layers 11 arranged between the phosphor layers 13 to improve the contrast ratio.

[0137] A conductive thin film layer 19 made of aluminum or an aluminum alloy is formed on the phosphor screen 16 , and a deposited getter layer 21 made of barium (Ba) is formed on the conductive thin film layer 19 . This conductive thin film layer 19 of the face plate 12 serves as an anode electrode. The deposited getter layer 21 is formed by vapor-depositing a getter material in a vacuum chamber before adhering the face plate 12 and rear plate 10 in the vacuum chamber. When the series of steps from vapor deposition to sealing of the getter material are performed in a vacuum state without being exposed to the atmosphere, a deposited getter layer 21 with a high performance can be obtained.

[0138] As shown in FIGS. 12 to 14 , the rear plate 10 has an insulating substrate made of non-alkali glass. A plurality of scanning electrodes 23 and signal electrodes 25 run on the inner surface of the insulating substrate in the form of a matrix. Gate electrodes 27 and emitter electrodes 29 respectively extending from the scanning electrodes and signal electrodes are formed in the vicinities of the intersections of the scanning electrodes 23 and signal electrodes 25 .

[0139] The gate electrodes 27 and emitter electrodes 29 are arranged to oppose each other at a predetermined gap, e.g., 50 μm. Although not shown, for example, a graphite film is arranged between the electrodes 27 and 29 to oppose them at a gap of 5 μm, thereby making up a surface conduction type electron-emitting element 18 . A protection film 31 is formed on each scanning electrode 23 .

[0140] The spacer assembly 22 is provided between the face plate 12 and rear plate 10 with the above arrangement, to support the face plate and rear plate against the atmospheric pressure. The spacer assembly 22 will be described in detail hereinafter.

[0141] As shown in FIGS. 12 and 14 , the spacer assembly 22 has an electrode plate 42 arranged for preventing abnormal discharge between the face plate 12 and rear plate 10 . The electrode plate 42 is made of an iron-nickel alloy to have a thickness of 0.1 mm, and its surface is oxidized. Although depending on its size, the electrode plate 42 preferably has a thickness of about 0.1 mm to 0.25 mm, if it is to match the effective display region with a diagonal size of 20 inches or more, so that a desired strength is ensured.

[0142] To ensure easy manufacture, the electrode plate 42 may be divided into a plurality of portions. Then, however, the boundaries of the divisional portions adversely affect the display performance. Hence, a large electrode plate with a size matching an effective display region 3 is preferably used where possible.

[0143] The electrode plate 42 is arranged between the face plate 12 and rear plate 10 to be parallel to them. The electrode plate 42 has a plurality of rectangular holes 26 , each with a size of 250 μm×180 μm and allowing to pass electron beams emitted from the surface conduction type electron-emitting elements 18 , to oppose the surface conduction type electron-emitting elements 18 . The electrode plate 42 also has a plurality of circular spacer holes 28 for connecting the first and second spacers (to be described later).

[0144] The electrode plate 42 has a first surface opposing the face plate 12 and a second surface opposing the rear plate 12 . A plurality of first spacers 30 a are formed on the first surface to be integral with the electrode plate 42 , and a plurality of second spacers 30 b are formed on the second surface to be integral with the electrode plate 42 . The first and second spacers 30 a and 30 b are connected to each other through connecting portions 52 arranged in the spacer holes 28 formed in the electrode plate 42 . In this embodiment, one second spacer 30 b is connected to two first spacers 30 a through the connecting portions 52 .

[0145] As shown in FIGS. 12 to 14 , the second spacers 30 b are arranged on the scanning electrodes 23 through the protection films 31 to correspond to the surface conduction type electron-emitting elements 18 , and extend along the interconnection. Each second spacer 30 b has an elongated elliptic section, and is formed such that its length L 1 in the direction of interconnection on the electrode plate 42 side is 0.4 mm, its length L 2 in a direction perpendicular to the direction of interconnection is 500 μm, its length L 1 ′ in the direction of interconnection on the rear plate 10 side is 0.35 mm, its length L 2 ′ in a direction perpendicular to the direction of interconnection is 400 μm, and its height h 1 is 0.5 mm.

[0146] Two first spacers 30 a are arranged for one second spacer 30 b. Each first spacer 30 a has a columnar shape with some taper, and is formed such that its end on the electrode plate 42 side has a diameter Φ 1 of 320 μm, its end on the electrode plate 42 side has a diameter Φ 2 of 230 μm, and its height h 2 is 1.0 mm.

[0147] More specifically, in the fourth embodiment, the first spacers 30 a are formed to have a sufficiently large aspect ratio (the ratio of the height to the length of the end on the electrode plate 42 side in the direction of the major axis of the section) when compar