Method for manufacturing curved surface support structure

Abstract

The present invention provides a curved surface support structure more excellent in accuracy while avoiding increased complexity of steps.　According to the present invention, there is provided a method for manufacturing a curved surface support structure comprising a step (a) for obtaining a laminate having an annular band-shaped cavity inside by joining a first wafer provided with a first electrode pattern including a first electrode and a second wafer provided with a second electrode pattern including a plurality of second electrodes, a step (b) for increasing or decreasing the volume of the cavity by deforming the surface of at least one of the first wafer and the second wafer into a curved surface by a pressure difference between the cavity and the outside of the laminate by heating, and a step (c) for removing a part of the first wafer along a circumference concentric with the annular band shape and forming a resonator in which a central part is supported by the second wafer from the first wafer, the first electrode and the second electrode in the laminate facing each other with an interval therebetween in the step (a), and the step (a) including a step for electrically connecting the first electrode to the second electrode pattern of the second wafer.

Description

Method for manufacturing a curved support structure

[0001] The present invention relates to a method for manufacturing a curved support structure.

[0002] Various types of gyroscopes are widely used to detect changes in the attitude of an object. Among these, vibratory gyroscopes (VSGs, also known as Coriolis vibratory gyroscopes (CVGs)) have recently attracted attention due to their ability to achieve high accuracy while being robust and compact. For example, Patent Document 1 listed below discloses a hemispherical resonator gyroscope (HRG) having a resonator with a raised shape and a generally plate-shaped support portion that supports the resonator.

[0003] The hemispherical resonator gyroscope of Patent Document 1 has a structure in which a plurality of first electrodes on the support side and a plurality of second electrodes on the resonator side are opposed to each other with a predetermined gap therebetween. In Patent Document 1, a sacrificial layer is disposed between the first electrodes provided on the lower wafer and the second electrodes provided on the upper wafer, and after the upper and lower wafers are processed into the desired shape, the sacrificial layer is selectively removed by etching. The above-mentioned inter-electrode gap is formed by etching the sacrificial layer.

[0004] International Publication No. 2022 / 224704

[0005] However, depending on the choice of wafer and sacrificial layer materials, etching can leave damage to the wafer.There is room for improvement in realizing relatively complex three-dimensional shapes such as those found in hemispherical resonator gyroscopes.

[0006] According to the present invention, the following inventions are provided: [1] A method for manufacturing a curved support structure, comprising: step (a) of bonding a first wafer provided with a first electrode pattern including a first electrode and a second wafer provided with a second electrode pattern including a plurality of second electrodes to obtain a laminate having an annular cavity therein; step (b) of deforming the surface of at least one of the first wafer and the second wafer into a curved shape by heating due to a pressure difference between the cavity and the outside of the laminate, thereby increasing or decreasing the volume of the cavity; and step (c) of removing a portion of the first wafer along a circumference concentric with the annular shape to form a resonator from the first wafer, the central portion of which is supported by the second wafer, wherein in step (a), the first electrode and the second electrode in the laminate face each other with a gap therebetween, and step (a) includes a step of electrically connecting the first electrode to the second electrode pattern of the second wafer. [2] The manufacturing method according to [1], wherein step (c) includes removing the portion of the first wafer by laser ablation. [3] The manufacturing method according to [1], wherein step (c) includes removing the portion of the first wafer by etching. [4] The manufacturing method according to [3], wherein, between steps (b) and (c), a step of forming a protective film on the entire upper surface of the first wafer and on a portion of the lower surface of the second wafer is further included. [5] The manufacturing method according to [2], wherein, between steps (b) and (c), a step (d) of forming a through-via in the second wafer that reaches the second electrode pattern is further included, and, after step (d), a step (e) of bonding a support substrate on which a third electrode pattern is provided to the lower surface of the second wafer to electrically connect the through-via to the third electrode pattern. [6] The manufacturing method according to any one of [1] to [5], wherein step (a) or step (b) is performed under reduced pressure.[7] The manufacturing method according to any one of [1] to [6], further comprising, before the step (a), a step (a1) of preparing a wafer having an annular recess on one of its main surfaces, and, between the step (a1) and the step (a), a step (a2) of forming the second electrode pattern on the main surface side to obtain the second wafer. [8] The manufacturing method according to any one of [1] to [7], wherein the step (c) includes a step of removing the portion of the first wafer to make the cavity communicate with the outside of the laminate. [9] The manufacturing method according to any one of [1] to [8], further comprising, between the step (a) and the step (b), a step (e) of placing a jig having a hole on the main surface of the first wafer on the side where the first electrode pattern is not provided, and the step (b) includes a step of deforming a portion of at least one of the first wafer and the second wafer that is inside the hole of the jig.

[10] A method for manufacturing a curved support structure, comprising: a step (A) of bonding a support substrate having holes to a first wafer provided with a first electrode pattern including first electrodes; a step (B) of separating the first wafer into an outer wafer and an inner wafer that is a disk-shaped wafer surrounded by the outer wafer and includes the first electrodes by removing a portion of the first wafer that overlaps the support substrate in a plan view; and a step (C) of bonding a second wafer provided with a second electrode pattern including a plurality of second electrodes to the inner wafer and the outer wafer.

[11] A manufacturing method comprising: a step (C) of obtaining a laminate having an annular cavity therein; a step (D) of deforming the surface of at least one of the inner wafer and the second wafer into a curved shape by heating due to a pressure difference between the cavity and the outside of the laminate, thereby increasing or decreasing the volume of the cavity; and a step (E) of removing the support substrate and forming a resonator from the inner wafer, the central portion of which is supported by the second wafer, wherein in the step (C), the first electrode and the second electrode in the laminate face each other with a gap therebetween.

[11] A manufacturing method according to

[10] , wherein the step (C) or the step (D) is carried out under reduced pressure.

[12] The manufacturing method according to

[10] or

[11] , further comprising, before step (C), step (C1) of preparing a wafer having an annular recess on one main surface, and step (C2) between step (C1) and step (C) of forming the second electrode pattern on the main surface to obtain the second wafer.

[13] The manufacturing method according to any one of

[10] to

[12] , wherein step (C) includes a step of electrically connecting the first electrode to the second electrode pattern of the second wafer.

[14] The manufacturing method according to any one of

[10] to

[13] , further comprising, between step (C) and step (D), step (F) of placing a jig having a hole on the support substrate, wherein step (D) includes a step of deforming a portion of at least one of the inner wafer and the second wafer that is inside the hole of the jig.

[15] The manufacturing method according to any one of [1] to

[14] , wherein the resonator has a columnar portion extending upward from the second wafer and a rim arranged in a ring shape around the columnar portion.

[0007] According to at least any one of the embodiments of the present invention, it is possible to provide a curved support structure (for example, a hemispherical resonator gyroscope) with improved precision while avoiding complicated processes.

[0008] 1 is a flowchart for outlining an exemplary manufacturing method of a hemispherical resonator gyroscope according to an embodiment of the present invention. FIG. 1 is a schematic vertical cross-sectional view of an upper wafer having a first electrode pattern. FIG. 2 is a schematic plan view showing an example of a first electrode pattern. FIG. 3 is a schematic vertical cross-sectional view of a bare wafer having a recess formed on one main surface. FIG. 4 is a schematic plan view showing a typical shape of a recess formed in a bare wafer. FIG. 5 is a schematic vertical cross-sectional view of a bare wafer 20 after the formation of a second electrode pattern. FIG. 6 is a schematic plan view showing an example of a second electrode pattern. FIG. 7 is a schematic plan view showing another example of the shape of the second electrode pattern PB. FIG. 8 is a flowchart showing an example of a manufacturing method according to another embodiment of the present invention. FIG. 9 is a diagram schematically showing a vertical cross-section of a laminate 30A obtained by bonding an upper wafer 10WA and a lower wafer 20WA. FIG. 10 is a schematic view showing an example of overlapping between the first electrode pattern PA and the second electrode pattern PB in the laminate 30A. FIG. 11 is a schematic view showing another example of overlapping between the first electrode pattern PA and the second electrode pattern PB in the laminate 30A. 1. A cross-sectional view schematically showing a stacked body 30A in which a jig is placed on an upper wafer 10WA. 1. A schematic plan view of the stacked body 30A and the jig shown in FIG. 13 when viewed in the normal direction to the main surface 10a of the upper wafer 10WA. 2. A schematic vertical cross-sectional view for explaining a method of deforming the upper wafer 10WA using a jig 50. 3. A schematic vertical cross-sectional view of the stacked body 30A after deformation of the upper wafer 10WA. 4. A schematic vertical cross-sectional view showing another example of the shape of a jig having a hole. 5. A schematic vertical cross-sectional view showing yet another example of the shape of a jig having a hole. 6. A schematic vertical cross-sectional view for explaining a method of deforming the upper wafer 10WA using a metal film 52 instead of the jig 50. 7. A schematic vertical cross-sectional view for explaining a modified example in which both the upper wafer 10WA and the lower wafer 20WA are deformed. 8. A schematic vertical cross-sectional view for explaining partial removal of the upper wafer 10WA by laser ablation. 9. A schematic vertical cross-sectional view showing an example of a gyroscope as a curved support structure obtained by the manufacturing method shown in FIG. 10A and 10B are schematic vertical cross-sectional views for explaining partial removal of the upper wafer 10WA by etching, and a schematic vertical cross-sectional view for explaining the progress of etching of the upper wafer 10WA and the lower wafer 20WA.1 is a schematic vertical cross-sectional view illustrating an example of forming a protective film 64 in advance on the surface of the upper wafer 10WA and the surface of the lower wafer 20WA. FIG. 2 is a flowchart illustrating an example of a method for manufacturing a hemispherical resonator gyroscope according to another embodiment of the present invention. FIG. 3 is a plan view showing an example of a support substrate 70 having a hole 70h. FIG. 4 is a schematic vertical cross-sectional view illustrating bonding of the support substrate 70 to the upper wafer 10WA. FIG. 5 is a schematic vertical cross-sectional view illustrating a region on the main surface 10b of the upper wafer 10WA where a protective film 74 is to be disposed. FIG. 6 is a schematic vertical cross-sectional view illustrating division into an inner wafer 10Wx and an outer wafer 10Wy. FIG. 7 is a schematic vertical cross-sectional view illustrating a stacked body 30B obtained by bonding a set of the inner wafer 10Wx and the outer wafer 10Wy to the lower wafer 20WA. FIG. 8 is a cross-sectional view illustrating a stacked body 30B with a jig 50 placed on the support substrate 70. FIG. 9 is a schematic vertical cross-sectional view illustrating a method for deforming the inner wafer 10Wx using the jig 50. 41 is a schematic vertical cross-sectional view showing an example of a curved support structure obtained by separating the inner wafer 10Wx from the outer wafer 10Wy by removing the support substrate 70. FIG. 42 is a schematic vertical cross-sectional view showing an example of forming through holes 20t in the lower wafer 20WA. FIG. 43 is a schematic vertical cross-sectional view for illustrating the formation of through vias 26 in the lower wafer 20WA. FIG. 44 is a schematic vertical cross-sectional view showing a gyroscope 1B having a resonator 10A and a support portion 20B. FIG. 45 is a schematic vertical cross-sectional view for illustrating bonding of the gyroscope 1B to a support substrate 30 having a third electrode pattern PC. FIG. 46 is a schematic vertical cross-sectional view showing an example of a curved support structure obtained by a manufacturing method according to yet another embodiment of the present invention. FIG. 47 is a schematic perspective view showing the appearance of an example hemispherical resonator gyroscope. FIG. 48 is a schematic view showing a ZX cross-section when the gyroscope shown in FIG. 40 is cut near the center. FIG. 49 is a schematic perspective view showing another example of the shape of the resonator. FIG. 49 is a schematic view showing a ZX cross-section when the gyroscope shown in FIG. 42 is cut near the center.

[0009] The following describes embodiments of the present invention. The various features shown in the following embodiments can be combined with each other. Furthermore, each feature can be an independent invention.

[0010] 1. Hemispherical Resonator Gyroscope This application proposes a novel method for manufacturing a curved support structure. Below, a hemispherical resonator gyroscope is used as an example of a curved support structure. To facilitate understanding of the present invention, the basic configuration of a hemispherical resonator gyroscope will first be described.

[0011] Fig. 40 shows the appearance of an exemplary hemispherical resonator gyroscope. Fig. 41 schematically shows a vertical cross section of the gyroscope shown in Fig. 40 taken near the center. For ease of explanation, three arrows indicating the directions of the mutually orthogonal X-axis, Y-axis, and Z-axis are depicted in Figs. 40 and 41. The X-axis, Y-axis, and Z-axis shown here do not indicate a coordinate system fixed in space, but are used to indicate a relative coordinate system fixed to the gyroscope depicted in the drawings. Hereinafter, arrows indicating these three directions may also be depicted in other drawings.

[0012] A gyroscope 900A shown in Fig. 40 includes a resonator 80A and a support portion 90A that supports the resonator 80A. In the configuration illustrated in Fig. 40, the support portion 90A is generally plate-shaped, and here, a bottom surface 90b of the support portion 90A is rectangular and parallel to the XY plane of the figure. The length of one side of the rectangular shape of the bottom surface 90b is, for example, in the range of 10 mm to 20 mm.

[0013] The resonator of a hemispherical resonator gyroscope is also called a "shell," and in the example shown in Figures 40 and 41, the resonator 80A of the gyroscope 900A has a generally hemispherical shape that protrudes on the side opposite the support portion 90A. However, the center of the outer surface of the resonator 80A is recessed toward the support portion 90A (in the negative direction of the Z axis), and a pillar portion 82 extending toward the support portion 90A is formed in the center of the inside of the resonator 80A. The pillar portion 82 connects the resonator 80A to and supports the support portion 90A.

[0014] Resonators in hemispherical resonator gyroscopes generally have various shapes symmetric about a central axis, such as a bell, a wine glass, or a birdbath. The shape of the resonator 80A shown in FIGS. 40 and 41 is also symmetric in the XY plane with respect to the columnar portion 82. The resonator 80A includes an upwardly bulging hemispherical dome portion 84A. Here, an annular recess 90g is provided on the upper surface 90a of the support portion 90A, and a torus-shaped cavity 90t is formed inside the gyroscope 900A by the surface of this recess 90g and the inner surface of the upwardly bulging dome portion 84A.

[0015] Figures 42 and 43 show other examples of resonator shapes. Of these figures, Figure 43 schematically shows a vertical cross section of the gyroscope shown in Figure 42 taken near the center. A gyroscope 900B shown in Figures 42 and 43 includes a resonator 80B and a support portion 90B. While the support portion 90B has a configuration substantially similar to that of the support portion 90A of the gyroscope 900A described above, the resonator 80B differs from the resonator 80A shown in Figures 40 and 41 in that it includes a bowl portion 84B that is recessed annularly from the pillar portion 82 toward the support portion 90B. Similar to the resonator 80A, the resonator 80B is connected to and supported by the pillar portion 82.

[0016] In this way, the shape of the resonator is not limited to a shape that protrudes upward. As shown in Fig. 43, the gyroscope 900B has an internal cavity 90u that has a smaller volume than the cavity 90t of the gyroscope 900A described above.

[0017] 40 and 41 again. The dome portion 84A has a rim 840 arranged annularly around the columnar portion 82, and is configured axially symmetrically with respect to the columnar portion 82 as a whole. A lower surface 840b of the rim 840 is generally flat, and in this example, the lower surface 840b is generally parallel to the upper surface 90a of the support portion 90A. The rim 840 may also be referred to as a "flange."

[0018] The resonator 80A further has one or more first electrodes 81 arranged in a ring shape on the lower surface 840b of the rim 840. On the other hand, the support part 90A has a plurality of second electrodes 92 arranged on the upper surface 90a so as to face the first electrodes 81 of the resonator 80A. In this example, the second electrodes 92 are arranged in a ring shape outside the recess 90g.

[0019] The second electrodes 92 of the support portion 90A include two types of electrodes: a drive electrode and a detection electrode. The drive electrode applies vibrations of a predetermined frequency to the dome portion 84A when the gyroscope 900A is in operation. On the other hand, the detection electrode has the function of detecting changes in capacitance between the dome portion 84A and the support portion 90A. The change in capacitance can be used to determine changes in angular velocity accompanying the rotational motion of the entire gyroscope 900A. As shown in FIG. 41 , terminals 922 for connection to an external circuit can be drawn out from each of the second electrodes 92.

[0020] As can be understood from this detection principle, the shape of the dome portion (or bowl portion) and the size of the gap between the first electrode 81 and the second electrode 92 can have a direct effect on the performance of the gyroscope. Therefore, more precise control of the shape of the dome portion (or bowl portion) and the size of the gap between the electrodes contributes to improving the performance of the hemispherical resonator gyroscope.

[0021] As described in Patent Document 1, in the manufacture of hemispherical resonator gyroscopes, materials with relatively small thermoelastic coefficients, such as quartz and synthetic glass, are used as resonator materials to improve the Q value of the resonator. In the technology described in Patent Document 1, the inter-electrode gap is formed by etching a sacrificial layer made of TEOS or polysilazane. However, the etchant used to remove such a sacrificial layer can also cause considerable damage to the resonator material. In other words, etching the sacrificial layer is one of the factors that affect the shape of the dome portion (or bowl portion) and the size of the inter-electrode gap.

[0022] As described in detail below, the present invention provides a method for forming a curved support structure without requiring a sacrificial layer formation step. More specifically, it provides a method for forming an inter-electrode gap in a curved support structure, such as a hemispherical resonator gyroscope, without requiring a sacrificial layer formation and subsequent removal. According to at least one of the embodiments of the present invention described below, the sacrificial layer formation step and the sacrificial layer removal step can be avoided, and a curved support structure with excellent shape precision can be provided. For example, it is possible to provide a hemispherical resonator gyroscope having a dome portion or bowl portion with excellent shape precision.

[0023] 1 shows an outline of an exemplary method for manufacturing a hemispherical resonator gyroscope according to an embodiment of the present invention. As shown in Fig. 1, the method for manufacturing a hemispherical resonator gyroscope includes, for example, step S1 of bonding a first wafer and a second wafer to obtain a laminate having a cavity therein, step S2 of deforming the first wafer and / or the second wafer by heating to increase or decrease the volume of the cavity, and step S3 of removing a portion of the first wafer to form a resonator whose center is supported by the second wafer.

[0024] As will be understood from the following description, each step of the manufacturing method according to the embodiment of the present invention can be realized by a known semiconductor manufacturing process or a MEMS manufacturing process, such as that described in Patent Document 1. The entire disclosure of WO 2022 / 224704 is incorporated herein by reference.

[0025] (2-1. Step S1 of forming a laminate) As will be described in detail below, in a typical embodiment of the present invention, two wafers, an upper wafer as a first wafer and a lower wafer as a second wafer, are stacked and bonded together, and then one or both of these wafers are deformed to obtain a resonator shape such as those shown in Figures 40 to 43.

[0026] First, an upper wafer having a first electrode pattern is prepared. The upper wafer may be purchased or may be prepared by forming the first electrode pattern on one main surface of a bare wafer of an appropriate shape. Examples of the bare wafer include quartz wafers (fused silica wafers and synthetic quartz wafers), glass wafers (borosilicate glass wafers, low-expansion glass wafers), and silicon wafers.

[0027] After preparing a bare wafer, a first electrode pattern is formed on one main surface of the bare wafer. Figures 2 and 3 schematically show an example of an upper wafer having the first electrode pattern. Figure 2 schematically shows a vertical cross section of the upper wafer shown in Figure 3 cut near the center.

[0028] 2 and 3 , a first electrode pattern PA is formed on one main surface 10b of the bare wafer 10. The first electrode pattern PA includes at least one first electrode 11A. By forming the first electrode pattern PA on the bare wafer 10, an upper wafer 10WA is obtained from the bare wafer 10.

[0029] In this example, a disk-shaped first electrode pattern PA is formed on the main surface 10b of the bare wafer 10. In the configuration illustrated in Fig. 3, the outer edge of the first electrode pattern PA roughly coincides with the outer diameter of the resonator (which may be referred to as the outer diameter of the dome portion). The outer diameter of the resonator of the gyroscope obtained according to the embodiment of the present invention may be, for example, approximately 10 mm.

[0030] Here, a part of the first electrode pattern PA functions as the first electrode 11 A. In the example shown in Fig. 3, an annular portion of the first electrode pattern PA near the outer edge (the portion schematically shown by dark shading in Fig. 3) functions as the first electrode 11A that faces a second electrode described below.

[0031] In a typical embodiment of the present invention, when forming the first electrode pattern PA, as shown in FIGS. 2 and 3 , not only the first electrode pattern PA but also a metal layer 11x having a shape that continuously surrounds the first electrode pattern PA is formed on the main surface 10b. As described below, this metal layer 11x can be used as a bonding layer for bonding the upper wafer 10WA to the lower wafer. In the configuration illustrated in FIG. 3 , the metal layer 11x surrounds the first electrode pattern PA in a rectangular shape. However, this is merely an example, and the shape of the metal layer 11x in a plan view normal to the main surface 10b of the bare wafer 10 may be other shapes, such as a circle, as long as it continuously surrounds the first electrode pattern PA.

[0032] The first electrode pattern PA and the metal layer 11x can be formed from high-melting-point metals such as W, Ta, Nb, V, Ru, Pt, Cr, and Ti. The first electrode pattern PA and the metal layer 11x can be obtained by depositing one or more of these materials. The method for forming the first electrode pattern PA and the metal layer 11x is not particularly limited, and examples of methods that can be used include vacuum deposition, sputtering, ion plating, chemical vapor deposition (CVD), and physical vapor deposition (PVD). Two or more of these methods may be combined to form the first electrode pattern PA having a layered structure. The thickness of the first electrode 11A of the first electrode pattern PA is, for example, in the range of 25 nm to 500 nm.

[0033] The lower wafer, like the upper wafer 10WA, can be prepared by purchasing it or by forming a second electrode pattern having a plurality of second electrodes on one main surface of a bare wafer of an appropriate shape. As with the formation of the upper wafer 10WA, a quartz wafer, a glass wafer, a silicon wafer, or the like can be used as the bare wafer for forming the lower wafer.

[0034] Here, prior to the formation of the second electrode pattern, a recess is formed on one main surface of the bare wafer by etching. Figures 4 and 5 schematically show an example of the bare wafer after etching. Figure 4 schematically shows a vertical cross section of the wafer shown in Figure 5 cut near the center. In this example, an annular recess 20g is formed on one main surface 20a of the bare wafer 20. The recess 20g is sized to accommodate the first electrode pattern PA described above and a second electrode pattern described later.

[0035] 4 and 5, an island-shaped portion 20h that remains unetched is formed in the annular central portion of the recess 20g as a result of selective etching from the main surface 20a side of the bare wafer 20. Note that, instead of the bare wafer 20, a similar recess may be formed in the bare wafer 10 that will form the upper wafer 10WA, or recesses may be formed in both the bare wafer 20 and the bare wafer 10.

[0036] Next, a second electrode pattern is formed on the main surface 20a of the bare wafer 20. Figures 6 and 7 schematically show the bare wafer 20 after the second electrode pattern has been formed. Figure 7 shows an example of the appearance of the bare wafer 20 having the second electrode pattern as viewed from the normal direction of the main surface 20a, while Figure 6 schematically shows a vertical cross section of the wafer shown in Figure 7 cut near the center.

[0037] 7 , a second electrode pattern PB including a plurality of second electrodes 22A, each of which has a generally arc-shaped configuration, is formed on the main surface 20a of the bare wafer 20, thereby obtaining a lower wafer 20WA from the bare wafer 20. As with the material for the first electrode pattern PA, the material for forming the second electrode pattern PB can be one or more selected from the group consisting of W, Ta, Nb, V, Ru, Pt, Cr, and Ti. The same film formation method as used to form the first electrode pattern PA can be used to form the second electrode pattern PB. The thickness of the second electrode pattern PB can be, like the thickness of the first electrode pattern PA, in the range of, for example, 25 nm to 500 nm.

[0038] In the configuration illustrated in FIG. 7 , the second electrode 22A of the lower wafer 20WA includes multiple pairs of drive electrodes 12c and detection electrodes 12d arranged in a ring shape. Each electrode included in the second electrode 22A is connected to a drive circuit for applying a predetermined voltage to the drive electrode 12c or a detection circuit for applying a predetermined voltage to the detection electrode 12d during gyroscope operation. As illustrated in FIG. 7 , in this example, the drive electrodes 12c and detection electrodes 12d are alternately arranged inside the recess 20g so as to surround the island-shaped portion 20h. From another perspective, the island-shaped portion 20h can be considered to be located at the center D of an imaginary circle formed by the entire set of drive electrodes 12c and detection electrodes 12d.

[0039] In the configuration illustrated in FIG. 7 , the second electrode pattern PB includes, in addition to the second electrode 22A (i.e., the drive electrode 12c and the detection electrode 12d), a plurality of first electrical paths 23A extending radially from the center D and a second electrical path 24A annularly surrounding the second electrode 22A. As illustrated in FIG. 7 , each of the plurality of first electrical paths 23A passes between the drive electrode 12c and the detection electrode 12d and is connected to the second electrical path 24A. The second electrical path 24A is connected to a power source (ground here) not shown in FIG. 7 . Note that, as long as the potential of the first electrical path 23A extending from the island portion 20h of the second electrode pattern PB is fixed to ground or the like during operation of the gyroscope, it is not essential for the second electrode pattern PB to include the second electrical path 24A in this embodiment of the present invention. Here, the second electrode 22A, the first electrical path 23A, and the second electrical path 24A are all disposed inside the recess 20g.

[0040] FIG. 8 shows another example of the shape of the second electrode pattern PB. In the configuration illustrated in FIG. 8 , the second electrode pattern PB includes a second electrode 22A arranged in a ring shape, a first electrical path 23A, and a second electrical path 24A, similar to the example illustrated in FIG. 7 . In the example illustrated in FIG. 8 , terminals 122 extend radially from each of the drive electrodes 12c and detection electrodes 12d in the second electrode 22A. Having terminals 122 on each of the second electrodes 22A facilitates connection of a drive circuit or a detection circuit to the second electrodes 22A. It is not essential that each of the second electrodes 22A has a terminal 122. The shape of each of the second electrodes 22A itself is arbitrary as long as connection to a drive circuit or a detection circuit is possible via wire bonding.

[0041] 7 and 8, the second electrode pattern PB has a total of 16 second electrodes 22A arranged at equal intervals around the circumference. The number of second electrodes 22A in the second electrode pattern PB is basically a multiple of four. If the number of second electrodes 22A is, for example, eight, it becomes possible to determine whether the detected angular velocity is the result of erroneous detection due to manufacturing errors or the like. Furthermore, if the number of second electrodes 22A is increased to, for example, 16, it becomes possible not only to detect the angular velocity but also to obtain an output in the form of an angle through an integration calculation.

[0042] As such, a manufacturing method according to an embodiment of the present invention may further include an additional step before step S1 shown in Fig. 1. Fig. 9 shows an example of a manufacturing method according to another embodiment of the present invention. The manufacturing method illustrated in Fig. 9 includes, before step S1 shown in Fig. 1, step S01 of preparing a wafer having an annular recess on one main surface, and further includes step S02 of forming a second electrode pattern on the wafer provided with the annular recess to obtain a second wafer.

[0043] In this case, in parallel with the formation of the second electrode pattern PB, a metal layer 12x having a shape that continuously surrounds the second electrode pattern PB, similar to the metal layer 11x on the upper wafer 10WA, is formed on the main surface 20a. The metal layer 12x is spatially separated from the second electrode 22A of the second electrode pattern PB and is not electrically connected to the second electrode pattern PB.

[0044] After the upper wafer 10WA and the lower wafer 20WA are obtained, they are stacked and bonded together to form a laminate having a cavity therein. Fig. 10 shows a schematic vertical cross section of a laminate 30A obtained by bonding the upper wafer 10WA and the lower wafer 20WA.

[0045] 10 , first, the upper wafer 10WA and the lower wafer 20WA are overlapped with each other in a state where the first electrode 11A (here, the first electrode 11A is a part of the disk-shaped first electrode pattern PA) of the upper wafer 10WA faces the second electrode 22A of the lower wafer 20WA. At this time, the first electrode 11A and the second electrode 22A face each other across a gap whose size is determined by the depth of the recess 20g.

[0046] An appropriate alignment device such as a double-sided mask aligner can be used to align the upper wafer 10WA and the lower wafer 20WA, and the upper wafer 10WA and the lower wafer 20WA can be bonded with the alignment accuracy of the alignment device. For example, by using a double-sided mask aligner, it is possible to keep the order of alignment error between the first electrode 11A and the second electrode 22A within the error range of the semiconductor manufacturing process (e.g., 5 μm or less).

[0047] 11 and 12 show the first electrode pattern PA and the second electrode pattern PB in the laminate 30A. The first electrode 11A and the second electrode 22 in the laminate 30A face each other with a gap therebetween. In FIGS. 11 and 12, the portion of the first electrode pattern PA that overlaps with the second electrode 22A of the second electrode pattern PB is indicated by hatching. In these examples, the first electrode 11A is the portion of the first electrode pattern PA that overlaps with the second electrode 22A of the second electrode pattern PB in a plan view. In the example shown in FIG. 12, the first electrode 11A of the first electrode pattern PA entirely covers the second electrode 22A of the second electrode pattern PB, but this is not a limitation. As shown in FIG. 11, the first electrode 11A may only partially cover the second electrode 22A.

[0048] The upper wafer 10WA and the lower wafer 20WA can be bonded using a general bonding device and substrate bonding techniques such as diffusion bonding and anodic bonding. The method for bonding the upper wafer 10WA and the lower wafer 20WA is not limited to a specific method and may be selected appropriately depending on the materials of the bare wafers 10 and 20.

[0049] The bonding is not limited to direct bonding between wafers, and bonding via a material other than the material constituting the wafer (e.g., a metal layer) may also be applied. Here, as shown in FIG. 10 , when the upper wafer 10WA is placed on the lower wafer 20WA, the metal layer 11x of the upper wafer 10WA and the metal layer 12x of the lower wafer 20WA come into contact with each other. By applying heat and pressure in this state, the metal layer 11x and the metal layer 12x function as bonding layers, and the upper wafer 10WA and the lower wafer 20WA can be bonded by diffusion bonding.

[0050] As described with reference to FIGS. 3 and 7 , the metal layer 11x has a shape that continuously surrounds the first electrode pattern PA, and similarly, the metal layer 12x has a shape that continuously surrounds the second electrode pattern PB. Furthermore, as can be seen from FIGS. 3 and 7 , when the upper wafer 10WA and the lower wafer 20WA are stacked together with the first electrode 11A of the upper wafer 10WA facing the second electrode 22A of the lower wafer 20WA, in this example, the continuously formed metal layer 11x faces the continuously formed metal layer 12x. Therefore, by bonding the upper wafer 10WA and the lower wafer 20WA, an annular, airtight cavity 30ca is formed within the laminate 30A, corresponding to the annular recess 20g on the main surface 20a of the lower wafer 20WA. As will be described later, bonding the upper wafer 10WA and the lower wafer 20WA may be performed under reduced pressure.

[0051] 10 , the portion of the second electrode pattern PB of the lower wafer 20WA that is located on the top surface of the island-shaped portion 20h contacts the first electrode pattern PA of the upper wafer 10WA when the upper wafer 10WA is placed on the lower wafer 20WA. Therefore, by applying, for example, heat and pressure to the interface between these electrode patterns, the first electrode pattern PA and the first electrical path 23A of the second electrode pattern PB are electrically connected at the position of the island-shaped portion 20h in the process of bonding the upper wafer 10WA and the lower wafer 20WA. In this way, the process of bonding the upper wafer 10WA and the lower wafer 20WA includes the step of electrically connecting the first electrode 11A to the first electrical path 23A of the second electrode pattern PB.

[0052] As can be seen from FIG. 10 , after the upper wafer 10WA and the lower wafer 20WA are bonded, the first electrode pattern PA and the second electrode pattern PB are positioned inside the cavity 30ca of the laminate 30A. To accommodate the first electrode pattern PA and the second electrode pattern PB inside the cavity 30ca, for example, the depth of the recess 20g formed in the bare wafer 20 is set to be greater than the total thickness of the first electrode 11A and the second electrode 22A when stacked. The depth of the recess 20g determines the size of the gap between the first electrode 11A on the resonator 10A side and the second electrode 22A on the support portion 20A side. The depth of the recess 20g need only be large enough to accommodate at least the first electrode pattern PA and the second electrode pattern PB. A depth of 5 μm or less is advantageous because it allows the voltage applied to the drive electrode 12c to be several volts or less (e.g., 10 volts or less).

[0053] (2-2. Step S2 of deforming the first wafer and / or the second wafer) Next, the laminate 30A is heated to deform one or both of the upper wafer 10WA and the lower wafer 20WA, thereby increasing or decreasing the volume of the cavity 30ca. Prior to heating the laminate 30A, a jig of a predetermined shape is first placed on the other main surface 10a of the upper wafer 10WA, in other words, on the main surface 10a of the upper wafer 10WA on the side where the first electrode pattern PA is not provided.

[0054] FIG. 13 schematically illustrates a stack 30A with a jig placed on an upper wafer 10WA, and FIG. 14 schematically illustrates an example of the appearance of the stack 30A and jig shown in FIG. 13 when viewed in the normal direction of the main surface 10a of the upper wafer 10WA (here, when viewed in the negative direction of the Z axis). As shown in FIGS. 13 and 14 , an example is described in which a jig 50 having a hole 50h is used. In the configuration illustrated in FIGS. 13 and 14 , the hole 50h is a cylindrical through-hole extending perpendicular to the lower surface 50b of the jig 50. This example is not limiting, and the hole 50h may be a bottomed hole.

[0055] The center of the hole 50h in the jig 50 is aligned with the center C of the disk-shaped first electrode pattern PA (see FIG. 3) and the center D of the virtual circle (see FIG. 7). As shown in FIG. 14, a portion of the main surface 10a of the upper wafer 10WA is exposed inside the hole 50h in the jig 50, and other areas of the main surface 10a are covered by the jig 50. As schematically shown in FIG. 14, the first electrode 11A of the first electrode pattern PA is located outside the opening of the hole 50h in the jig 50. As will be apparent from the following description, the diameter of the circular shape of the hole 50h defines the approximate inner diameter of the rim of the dome portion.

[0056] After placing the jig 50 on the upper wafer 10WA, the stack 30A with the jig 50 placed thereon is placed in a heating furnace and heated to a temperature at which the materials of the upper wafer 10WA and the lower wafer 20WA soften. If, for example, quartz is selected as the material for the bare wafers 10 and 20, the temperature in the heating furnace is raised to 1200°C to 1600°C, and if borosilicate glass is selected, the temperature in the heating furnace is raised to 500°C to 900°C.

[0057] In a typical embodiment of the present invention, the stack 30A is heated under a different environmental pressure than that used when bonding the upper wafer 10WA and the lower wafer 20WA. For example, by evacuating the interior of a heating furnace to lower the ambient pressure therein below atmospheric pressure (referring to standard atmospheric pressure of 101,325 Pa in this specification), the upper wafer 10WA can be deformed by utilizing the pressure difference between the interior of the cavity 30ca of the stack 30A. When the ambient pressure within the heating furnace is lowered below the pressure within the cavity 30ca, the base material (e.g., quartz) of the upper wafer 10WA softens as it is heated. This, combined with the expansion of the gas within the cavity 30ca, causes the portion of the upper wafer 10WA located inside the hole 50h of the jig 50 to expand toward the outside of the stack 30A (see FIG. 15 ). In other words, the volume of the cavity 30ca increases, forming a larger cavity 30t within the stack 30A.

[0058] Due to the deformation of the upper wafer 10WA, the portion of the main surface 10a of the upper wafer 10WA located inside the hole 50h of the jig 50 is deformed into a curved shape. In this way, by heating the laminate 30A, it is possible to obtain a hemispherical structure such as the dome portion 84A shown in Fig. 40 by utilizing the air pressure difference between the inside of the cavity 30ca and the outside of the laminate 30A. In the example shown in Fig. 15, the increase in the volume of the cavity 30ca causes the main surface 10a to bulge in an annular band shape, forming a surface shaped like a torus cut in half (annular convex curved surface).

[0059] On the other hand, the shape of the portion of the upper wafer 10WA that contacts the lower surface 50b of the jig 50 remains plate-like. Here, as shown in FIG. 14 , the opening of the hole 50h in the jig 50 substantially coincides with the inner edge of the first electrode 11A of the first electrode pattern PA. Therefore, the portion of the main surface 10b of the upper wafer 10WA (which may also be referred to as the lower surface of the upper wafer 10WA) where the first electrode 11A is formed remains substantially flat. That is, here, the size of the gap formed between the first electrode 11A on the upper wafer 10WA side and the second electrode 22A on the lower wafer 20WA side remains substantially the same before and after heating the stack 30A.

[0060] In this way, it is possible to selectively deform a portion of the upper wafer 10WA by partially pressing the surface of the upper wafer 10WA with the jig 50. As will be apparent from the following description, the rim of the dome portion can be formed from the portion of the upper wafer 10WA whose deformation is suppressed by the jig 50.

[0061] As described above, the portion of the first electrode pattern PA that overlaps the island-shaped portion 20h of the lower wafer 20WA is bonded to the second electrode pattern PB of the lower wafer 20WA. Therefore, the portion of the upper wafer 10WA located inside the hole 50h of the jig 50 near the center C of the first electrode pattern PA remains near the island-shaped portion 20h without being lifted upward. As a result, as schematically shown in FIG. 15 , the columnar portion 12 of the resonator 10A is formed from a portion of the upper wafer 10WA.

[0062] In this way, the jig 50 has the function of suppressing deformation of a portion of the upper wafer 10WA. After deforming the portion of the upper wafer 10WA located inside the hole 50h of the jig 50, the shape (here, a torus shape) of the main surface 10a of the upper wafer 10WA is fixed by slow cooling. After the torus shape of the upper wafer 10WA is obtained, the jig 50 is removed from the stack 30A, thereby obtaining a structure such as that shown in FIG. 16. The material of the jig 50 may be appropriately selected from the viewpoint of avoiding adhesion of the jig 50 to the upper wafer 10WA. A typical example of the material of the jig 50 is glassy carbon.

[0063] FIG. 17 shows another example of the shape of a jig having a hole. The jig 53 shown in FIG. 17 has a protrusion 53cp at the center of the hole 53h. In this example, the protrusion 53cp has a rod shape extending toward the opening of the hole 53h, and the top surface (e.g., a flat surface) of the rod shape contacts the main surface 10a of the upper wafer 10WA. In this way, by using a jig 53 with a protrusion 53cp provided inside the hole 53h, the central portion of the upper wafer 10WA can be pressed from above by the protrusion 53cp. As a result, during the heating process, deformation of the portion of the upper wafer 10WA that overlaps the island-shaped portion 20h of the lower wafer 20WA is suppressed, thereby more reliably establishing electrical connection between the first electrode pattern PA and the second electrode pattern PB.

[0064] FIG. 18 shows yet another example of the shape of a jig having a hole. The jig 51 shown in FIG. 18 has a circular, bottomed hole 51h. In this example, the main surface 10a of the upper wafer 10WA is deformed into a shape that conforms to the inner surface that defines the bottomed hole 51h. In this way, the jig may function as a mold that determines the shape of the dome portion of the resonator. By limiting the deformation of the wafer during the heating process using the shape of the jig hole, excessive deformation of the wafer surface (e.g., excessive expansion) can be avoided, thereby achieving the desired resonator shape.

[0065] Instead of placing jigs 50, 51, or 53 on the main surface 10a of the upper wafer 10WA, a metal film 52 may be formed on a portion of the main surface 10a of the upper wafer 10WA, as shown in FIG. 19 . In this case, the metal film 52 suppresses the expansion of the upper wafer 10WA due to heating, thereby suppressing deformation of the portion of the upper wafer 10WA covered by the metal film 52 while allowing deformation of the portion of the upper wafer 10WA where the metal film 52 is not disposed. Similar to the material used to form the first electrode pattern PA or the second electrode pattern PB, high-melting-point metals such as W, Ta, Nb, V, Ru, Pt, Cr, and Ti can be used as the material for the metal film 52. The metal film 52 can be formed by vacuum deposition, sputtering, ion plating, CVD, PVD, electroless plating, or electrolytic plating. The metal film 52 typically has a thickness on the order of several micrometers.

[0066] After obtaining the annular shape of the upper wafer 10WA, the metal film 52 is removed by etching, thereby obtaining the shape of the upper wafer 10WA similar to that shown in Fig. 16. In this way, partial deformation of the upper wafer 10WA by heating can also be achieved by using the metal film 52 as a jig.

[0067] In this example, the pressure difference between the cavity 30ca and the outside of the stack 30A is used to selectively deform the upper wafer 10WA of the upper wafer 10WA and the lower wafer 20WA that make up the stack 30A. However, this is not limited to this, and each of the upper wafer 10WA and the lower wafer 20WA may be partially deformed. For example, as shown in FIG. 20 , by placing the stack 30A on a jig 55 having a hole 55h similar to the jig 50 and then further placing the jig 50 on the stack 30A, it is possible to partially deform not only the upper wafer 10WA but also the lower wafer 20WA by heating the stack 30A. In this case, as schematically shown in FIG. 20 , the cavity 30ca can be enlarged approximately symmetrically with respect to the horizontal plane.

[0068] In this way, between step S1 of forming the stack and step S2 of deforming the first wafer and / or the second wafer, a step of placing a jig on one of the main surfaces of the wafer may be additionally performed. By using a jig with a hole, it is possible to selectively deform the portion of one or both of the upper wafer 10WA and the lower wafer 20WA that is inside the hole of the jig.

[0069] 13 to 20, the step of heating the stack 30A is performed under reduced pressure, thereby increasing the volume of the cavity 30ca inside the stack 30A. Conversely, the step of heating the stack 30A may be configured to decrease the volume of the cavity 30ca. For example, the upper wafer 10WA and the lower wafer 20WA may be bonded under reduced pressure.

[0070] By bonding the upper wafer 10WA and the lower wafer 20WA under reduced pressure, the pressure in the cavity 30ca inside the laminate 30A can be made lower than atmospheric pressure. When the resulting laminate 30A is heated, for example, under atmospheric pressure, the base materials of the upper wafer 10WA and the lower wafer 20WA soften, causing the upper wafer 10WA and the lower wafer 20WA to deform so that the lower surface of the upper wafer 10WA and the upper surface of the lower wafer 20WA approach each other. For example, the outer surface of the upper wafer 10WA is recessed in an annular shape toward the lower wafer 20WA. This deformation of the upper wafer 10WA allows a resonator having an annular concave surface, such as the resonator 80B shown in FIGS. 42 and 43 , to be obtained from the upper wafer 10WA.

[0071] To prevent the entire upper wafer 10WA from being distorted by heating the laminate 30A, the jig 50 described above may be fixed to the main surface 10a of the upper wafer 10WA (which may also be called the top surface of the upper wafer 10WA), and only the portion of the upper wafer 10WA located inside the hole 50h of the jig 50 may be selectively deformed. The entire main surface of one of the upper wafer 10WA and the lower wafer 20WA may be fixed to a rigid, heat-resistant plate material (such as a metal plate), thereby selectively deforming only the other of the upper wafer 10WA and the lower wafer 20WA.

[0072] (2-3. Step S3 of Forming Resonators) Next, resonators are formed from the upper wafer 10WA. In this embodiment, the shape of the resonators is obtained from the upper wafer 10WA by removing a portion of the upper wafer 10WA from the structure after annealing.

[0073] To partially remove the upper wafer 10WA, either a physical process such as mechanical processing (e.g., ultrasonic processing) or a chemical process such as etching can be applied. For example, as shown schematically in FIG. 21 , laser ablation using irradiation of a laser beam LB can be used to remove a portion of the upper wafer 10WA in an annular band shape centered on the position of the columnar portion 12. In other words, the upper wafer 10WA is cut by irradiation of a laser beam along a circumference centered on the island-shaped portion 20h of the lower wafer 20WA, for example, along the outer edge of the first electrode pattern PA. This defines the outer edge of the rib of the resonator. For laser ablation, for example, carbon dioxide gas (CO ２ ) Laser can be applied.

[0074] By removing a portion of the upper wafer 10WA around the center of the annular cavity 30t in this manner, the cavity 30t is connected to the outside of the laminate 30A, and the portion of the upper wafer 10WA that corresponds to the dome portion of the resonator is separated from the rest of the upper wafer 10WA. As a result, as shown in Figure 21, a resonator 10A having a dome portion 14A can be obtained from the upper wafer 10WA.

[0075] 21, the resonator 10A has a columnar portion 12 extending upward from the lower wafer 20WA and a dome portion 14A including a portion that bulges out in a circular ring shape. The dome portion 14A is axisymmetric with respect to the columnar portion 12, similar to the dome portion 84A shown in FIGS. 40 and 41, and includes a rim 140 that is annularly disposed around the columnar portion 12.

[0076] The lower wafer 20WA may be cut into a predetermined shape as needed. By removing unnecessary portions, the support portion 20A is formed from the lower wafer 20WA, and as shown in FIG. 22 , a gyroscope 1A is obtained that includes the support portion 20A having the second electrode pattern PB and the resonator 10A whose central portion is supported by the support portion 20A.

[0077] By applying laser processing to partially remove the upper wafer 10WA, the shape of the dome portion 14A can be obtained with the mechanical precision of a laser processing device. When determining the laser irradiation position, for example, the outline of the second electrode 22A can be used as an alignment mark. Dummy electrodes or wiring may be formed on the upper wafer 10WA or the lower wafer 20WA and used as fiducial marks.

[0078] In laser ablation, by appropriately adjusting parameters such as the laser wavelength, focal length, irradiation power density, pulse width, and repetition frequency, it is possible to selectively process the upper wafer 10WA of the laminate 30A while avoiding damage to the lower wafer 20WA. Alternatively, by adjusting these parameters, the upper wafer 10WA and the lower wafer 20WA may be cut at the same time.

[0079] Instead of laser ablation, the resonators 10A can also be formed from the upper wafer 10WA by etching. When etching is used to form the resonators 10A, a laminate 30A having an internal torus-shaped cavity 30t is prepared, for example, by a procedure similar to that described with reference to FIGS. 1 to 20 . Next, as shown schematically in FIG. 23 , a protective film 62 is formed on the entire main surface 10a, which is the upper surface of the upper wafer 10WA, and on a portion of the main surface 20b of the lower wafer 20WA. The main surface 20b of the lower wafer 20WA is the surface opposite the main surface 20a and corresponds to the lower surface of the lower wafer 20WA.

[0080] The protective film 62 functions as an etching resist in a subsequent process of forming a resonator. When hydrofluoric acid (HF) or buffered hydrofluoric acid (BHF) is used as an etchant for the upper wafer 10WA, an Au film or a Cr film can be used as the protective film 62.

[0081] Next, wet etching is performed on the laminate 30A to remove a portion of the upper wafer 10WA, leaving only the portion of the upper wafer 10WA that will become the resonator 10A. The etchant can be selected appropriately depending on the material of the protective film 62. When the laminate 30A is immersed in the etchant (e.g., hydrofluoric acid), the portion of the lower wafer 20WA not covered by the protective film 62 is removed, and the etchant penetrates into the cavity 30t of the laminate 30A, as schematically shown by the thick dashed arrow Et in FIG. 24 . As the etchant penetrates into the cavity 30t, it contacts the area of ​​the main surface 10b of the upper wafer 10WA that is exposed from the first electrode pattern PA. As a result, the portion of the upper wafer 10WA that will become the resonator 10A is left, and the other portions of the upper wafer 10WA can be selectively removed.

[0082] After forming the resonator 10A from the upper wafer 10WA, the protective film 62 on the laminate 30A is removed. If the protective film 62 is, for example, a Cr film, the Cr film can be removed using, for example, a cerium ammonium nitrate solution. By removing the protective film 62, a three-dimensional structure similar to the gyroscope 1A shown in FIG. 22 is obtained. By using etching, the shape of the resonator 10A can be realized with the same precision as lithography.

[0083] By forming the protective film 62 from a material that is resistant to the etchant used to etch the bare wafers 10 and 20, etching damage during the process of forming the resonators 10A by partially removing the upper wafer 10WA can be reduced. A high-melting-point metal that is resistant to the etchant used to etch the wafer can be selected as the material for the protective film 62. It is advantageous to use a metal (e.g., W or Cr) that can be selectively removed after the formation of the resonators 10A.

[0084] 24, since the etchant comes into contact with the first electrode pattern PA and the second electrode pattern PB, as well as with the areas of the surface inside the cavity 30t that are exposed by these electrode patterns, etching damage may occur to the wafer material and the electrode patterns. Therefore, after obtaining the upper wafer 10WA and the lower wafer 20WA and before obtaining the laminate 30A from them, a protective film 64 may be formed on the surface of the upper wafer 10WA and the surface of the lower wafer 20WA, as shown in FIG.

[0085] 3. Other Embodiments of the Manufacturing Method In each of the above-described examples, after the laminate 30A is obtained from the upper wafer 10WA and the lower wafer 20WA, a portion of the upper wafer 10WA is removed to define the outer edge of the resonator 10A. However, this is not limiting, and as described below, the shape of the resonator 10A may be obtained by first hollowing out a portion of the upper wafer 10WA into a disk shape and then deforming this disk-shaped portion.

[0086] 26 shows an outline of an exemplary method for manufacturing a hemispherical resonator gyroscope according to another embodiment of the present invention. The manufacturing method shown in FIG. 26 includes the steps of: step T1 of bonding a support substrate having a hole to a first wafer; step T2 of separating the first wafer into an outer wafer and an inner wafer by removing a portion of the first wafer; step T3 of bonding a second wafer to the inner and outer wafers to obtain a laminate having a cavity therein; step T4 of deforming the inner and / or second wafer by heating to increase or decrease the volume of the cavity; and step T5 of removing the support substrate to form a resonator whose center is supported by the second wafer. Each step shown in FIG. 26 will be described in detail below.

[0087] (3-1. Step T1 of Bonding Support Substrate to First Wafer) In this example, first, a support substrate 70 having a hole 70h as shown in FIG. 27 is prepared. The support substrate 70 may be prepared by purchasing it, or by forming a hole in a substrate made of an appropriate material by etching or the like. Here, the hole 70h is a through-hole with a circular opening. The material constituting the support substrate 70 may be any material as long as it can ensure the necessary rigidity and heat resistance. For example, a silicon substrate, a glass substrate, a quartz substrate, or the like can be used as the material of the support substrate 70.

[0088] Next, as shown in Fig. 28 , a support substrate 70 is bonded to the upper wafer 10WA as a first wafer on which the first electrode pattern PA is provided. As in the example described above, in the configuration illustrated in Fig. 28 , the first electrode pattern PA includes at least one first electrode 11A as a part thereof.

[0089] The method for bonding the upper wafer 10WA and the support substrate 70 is not limited to a specific method, as long as it can temporarily fix the support substrate 70 to the upper wafer 10WA so that the upper wafer 10WA does not fall off the support substrate 70. Here, the support substrate 70 is fixed to the main surface 10a of the upper wafer 10WA via an adhesive layer 72. The adhesive layer 72 may be a metal film such as a tungsten (W) film.

[0090] The support substrate 70 is bonded to the upper wafer 10WA in such a manner that the center of the hole 70h coincides with the center C (see FIG. 3) of the first electrode pattern PA. The opening of the hole 70h is typically located inside the annularly arranged first electrodes 11A in plan view. As will be apparent from the following description, the diameter of the hole 70h is approximately the same as the inner diameter of the rim 140 of the resonator 10A.

[0091] (3-2. Step T2 of Separating the First Wafer into an Outer Wafer and an Inner Wafer) Next, the upper wafer 10WA is divided into two parts by etching or the like. Prior to etching, as shown in FIG. 29, a protective film 74 is formed on the main surface 10b of the upper wafer 10WA as an etching resist. When hydrofluoric acid is used as the etchant, a metal film such as an Au film or a Cr film can be used as the protective film 74.

[0092] For ease of understanding, in FIG. 29 , the region of the main surface 10b on which the protective film 74 is formed is shaded. Here, the protective film 74 is formed on the main surface 10b of the upper wafer 10WA except for the annular region R3 concentric with the first electrode pattern PA. In other words, the protective film 74 includes a disk-shaped first portion Q1 concentric with the first electrode pattern PA and a second portion Q2 surrounding the first portion Q1. The first portion Q1 covers the portion of the upper wafer 10WA that will become the resonator 10A in a later process. On the other hand, the second portion Q2 covers at least the metal layer 11x located at the edge of the upper wafer 10WA.

[0093] Next, etching is performed until the upper wafer 10WA is completely penetrated, removing a portion of the upper wafer 10WA. After etching is completed, the protective film 74 is removed from the upper wafer 10WA. As a result of this etching, the upper wafer 10WA is divided into two portions: an inner wafer 10Wx and an outer wafer 10Wy, as shown in FIG. 30 . The inner wafer 10Wx is the portion of the upper wafer 10WA that was covered by the first portion Q1 of the protective film 74 and has a disk shape. The first electrode pattern PA remains on the inner wafer 10Wx. On the other hand, the outer wafer 10Wy is the portion of the upper wafer 10WA that was covered by the second portion Q2 of the protective film 74 and has a shape that surrounds the inner wafer 10Wx.

[0094] 30 , the region of the upper wafer 10WA that is removed by etching, in other words, the region of the upper wafer 10WA that is the perpendicular projection of the region R3 onto the main surface of the upper wafer 10WA, overlaps the support substrate 70 in a plan view. Therefore, the inner wafer 10Wx and the outer wafer 10Wy are held together by the support substrate 70 even after etching is completed.

[0095] (3-3. Step T3 of Obtaining a Stack Having a Cavity Inside) Next, a lower wafer 20WA is prepared as a second wafer, and a set of the inner wafer 10Wx and the outer wafer 10Wy supported by a support substrate 70 is bonded to the lower wafer 20WA. By bonding these wafers, the stack 30B shown in FIG. 31 is obtained. The lower wafer 20WA can be produced in the same manner as the example described with reference to FIGS. 4 to 8. That is, before step T3 of forming the stack 30B described here, step S01 shown in FIG. 9 and step S02 shown in FIG. 9 may be performed.

[0096] As in the example described with reference to FIGS. 10 to 12 , the wafers are bonded such that the first electrode 11A on the inner wafer 10Wx faces the second electrode 22A on the lower wafer 20WA. Here, the first electrode 11A is part of the first electrode pattern PA on the main surface 10b of the inner wafer 10Wx. As in the example described above, the first electrode 11A in the first electrode pattern PA faces the second electrode 22A on the lower wafer 20WA across a gap. As in the example described above, bonding the set of the inner wafer 10Wx and the outer wafer 10Wy to the lower wafer 20WA mechanically and electrically connects the first electrode pattern PA to the portion of the second electrode pattern PB on the lower wafer 20WA that is located on the island-shaped portion 20h (see FIG. 31 ).

[0097] The stack 30B has an annular cavity 30cb therein, similar to the stack 30A shown in FIG. Bonding between the set of inner wafer 10Wx and outer wafer 10Wy and the bottom wafer 20WA may be performed under reduced pressure. By performing bonding between the wafers under a pressure lower than atmospheric pressure, for example, in a subsequent heating step, the pressure difference between the inside of the cavity 30cb and the outside of the stack 30B can be used to deform the surfaces of the inner wafer 10Wx and / or the bottom wafer 20WA into a curved shape (e.g., an annular concave curved shape) so as to reduce the volume of the cavity 30cb.

[0098] (3-4. Step T4 of deforming the inner wafer and / or second wafer) After obtaining the stacked body 30B, the surface of at least one of the inner wafer 10Wx and the bottom wafer 20WA is deformed into a curved shape by heating the stacked body 30B. At this time, the deformed shape of the wafer may be controlled using a jig with a hole, as in the example described with reference to Figures 13 to 15. Here, as shown in Figure 32, a jig 50 with a hole 50h is placed on a support substrate 70.

[0099] The stack 30B with the jig 50 placed thereon is then moved into a heating furnace, where the stack 30B is heated while a vacuum is drawn. As the inner wafer 10Wx is heated, it softens, and the pressure inside the cavity 30cb is higher than the pressure outside. As a result, the inner wafer 10Wx bulges upward inside the hole 50h of the jig 50, as shown in FIG. 33 . Here, the upper surface of the inner wafer 10Wx bulges, except for the vicinity of the island-shaped portion 20h, resulting in an annular convex curved surface. The deformation of the inner wafer 10Wx allows the formation of a cavity 30t with a volume larger than the cavity 30cb.

[0100] In this manner, similar to the example of the embodiment described above, the inner wafer 10Wx can be deformed by heating the stack 30B using the air pressure difference between the cavity 30cb and the outside of the stack 30B. At this time, by using a jig with a hole, such as the jig 50, the portion of the inner wafer 10Wx that is inside the hole of the jig can be selectively deformed.

[0101] In the heating process of the stacked body 30B, deformation of not only the inner wafer 10Wx but also the bottom wafer 20WA may be permitted. As in the example described with reference to FIG. 20 , for example, by placing a jig with a hole on the main surface 20b side of the bottom wafer 20WA, the portion of the bottom wafer 20WA located inside the jig hole can be caused to bulge outward. In this way, deformation of the inner wafer 10Wx and / or the bottom wafer 20WA can partially deform the surfaces of these wafers into a curved shape.

[0102] (3-5. Step T5 of forming resonator) After obtaining the cavity 30t from the cavity 30cb, the jig 50 on the support substrate 70 is removed, and then the support substrate 70 is removed from the laminate 30B. By removing the support substrate 70, the deformed inner wafer 10Wx supported by the island-shaped portion 20h of the lower wafer 20WA is completely separated from the outer wafer 10Wy, and the resonator 10A with a curved support structure can be obtained from the inner wafer 10Wx as shown in FIG.

[0103] The method for separating the support substrate 70 from the laminate 30B is not particularly limited, and may be appropriately selected depending on the method for fixing the support substrate 70. For example, if the support substrate 70 is bonded to the inner wafer 10Wx and the outer wafer 10Wy via a tungsten film as the adhesive layer 72, hydrogen peroxide (H ２ O ２ The tungsten film can be selectively removed by etching using an aqueous solution of tungsten fluoride (THF) as an etchant. Depending on the method for fixing the support substrate 70, the support substrate 70 may be separated from the stack 30B by a peeling technique called laser lift-off, or the support substrate 70 itself may be removed by etching. When removing the support substrate 70 itself by etching, a protective film may be formed in advance on the portion of the resonator formed from the inner wafer 10Wx, and then the support substrate 70 may be etched.

[0104] After removing the support substrate 70, the lower wafer 20WA is cut into a desired shape by laser processing, etching, or the like, as needed, to form the support portion 20A from the lower wafer 20WA, thereby obtaining a gyroscope 1A similar to that according to the above-described embodiment (see FIG. 22). This embodiment has the advantage that the outer shape of the dome portion 14A can be defined before the inner wafer 10Wx is deformed.

[0105] 7 and 8, to drive the gyroscope 1A, it is necessary to connect a drive circuit or the like to the second electrode pattern PB of the support portion 20A and supply a predetermined amount of power. When the gyroscope 1A is mounted, for example, wiring connected to the drive circuit is connected to a wiring portion or terminal located on the upper surface 20a of the support portion 20A. It would be beneficial to be able to mount a curved support structure (e.g., a hemispherical resonator gyroscope) on a circuit board or the like using a simpler method.

[0106] An exemplary method for manufacturing a hemispherical resonator gyroscope according to yet another embodiment of the present invention will now be described in detail. In this embodiment, first, a laminate having a torus-shaped cavity 30t therein is fabricated according to the procedure described above, as shown in Fig. 16 or 33. Here, the description will continue using the laminate 30A shown in Fig. 16 as an example.

[0107] After the upper wafer 10WA is deformed to obtain a desired surface shape for the resonator 10A (e.g., hemispherical, annular convex, or annular concave), one or more through-holes 20t are formed in the lower wafer 20WA by, for example, etching from the main surface 20b (lower surface) of the lower wafer 20WA, as shown in FIG. 35 . The through-holes 20t are positioned, for example, to overlap the second electrode 22A in a plan view normal to the main surface 20b. The through-holes 20t may also be positioned to overlap the second electrical path 24A. There are no particular limitations on the method for forming the through-holes 20t; in addition to etching, any of mechanical processes such as laser processing, ultrasonic processing, and drilling can be used. The horizontal cross-sectional shape of each through-hole 20t may also be arbitrary.

[0108] Next, each through-hole 20t is filled with a conductive material. For example, the through-holes 20t are filled with a conductive paste by screen printing or the like, and the conductive material inside the through-holes 20t is hardened by heating, drying, or the like. By disposing the conductive member thus obtained inside the through-holes 20t, for example, through-vias 26 reaching the second electrode patterns PB can be formed in the lower wafer 20WA, as shown in FIG. 36 . Instead of filling with a conductive paste, the through-vias 26 may be formed inside the lower wafer 20WA by plating or by depositing tungsten using CVD.

[0109] After the through vias 26 are formed, a portion of the upper wafer 10WA is removed by, for example, laser ablation, to form the resonator 10A from the upper wafer 10WA. Here, the upper wafer 10WA and the lower wafer 20WA are cut together into a circular shape, and unnecessary portions are removed. By cutting the upper wafer 10WA, the resonator 10A is obtained from the upper wafer 10WA. At this time, by cutting the lower wafer 20WA together with the upper wafer 10WA, a support portion 20B that supports the central portion of the upper wafer 10WA can be formed from the lower wafer 20WA. As a result, as shown in FIG. 37 , a gyroscope 1B having the resonator 10A and the support portion 20B is obtained. In this example, since the outer shape of the resonator 10A is circular, the planar shape of the support portion 20B is also circular and concentric with the resonator 10A. Of course, it is not essential that the planar shape of the support portion 20B matches the planar shape of the resonator 10A; for example, the upper wafer 10WA and the lower wafer 20WA may be cut sequentially to obtain a support portion 20B having a planar shape different from that of the resonator 10A.

[0110] Next, a support substrate 30 having a third electrode pattern PC is prepared, and as shown in FIGS. 38 and 39 , the support substrate 30 is bonded to the lower surface 20b of the support portion 20B. The support substrate 30 is obtained by patterning electrodes and wiring on one main surface (here, the upper surface 30a) of an appropriate substrate (e.g., an epoxy substrate) or wafer (e.g., a glass wafer). Common wiring materials, such as Au, Pt, Cu, Al, or Ti, can be used to form the third electrode pattern PC. The third electrode pattern PC may also be formed by combining two or more of these wiring materials. The method for forming the third electrode pattern PC is not limited to a specific method, and known methods such as vacuum deposition, sputtering, ion plating, CVD, and PVD can be used.

[0111] The third electrode pattern PC typically includes a plurality of wirings provided in correspondence with the second electrodes 22 A. As can be seen from Fig. 39 , these wirings have portions that overlap with the through vias 26 connected to the second electrodes 22 A, and are drawn out to a region of the upper surface 30 a of the support substrate 30 that does not overlap with the support portion 20 B.

[0112] The support substrate 30 is bonded to the support portion 20B with its upper surface 30a, on which the third electrode pattern PC is provided, facing the lower surface 20b of the support portion 20B. The method for fixing the support substrate 30 to the support portion 20B is not particularly limited. For example, the support portion 20B may be bonded to the support substrate 30 via an Au bump or a conductive adhesive disposed between the through via 26 and the third electrode pattern PC. The conductive material filled in the through hole 20t may be used as the adhesive.

[0113] An electrical connection can be formed between the through vias 26 and the third electrode pattern PC by joining the support portion 20B and the support substrate 30. Since the wiring included in the third electrode pattern PC extends to the outside of the support portion 20B, a drive circuit or the like can be easily connected to the wiring of the third electrode pattern PC.

[0114] As described above, according to embodiments of the present invention, it is possible to accurately form an air gap between electrodes of, for example, a hemispherical resonator gyroscope, while eliminating the need for processes of forming and selectively removing a sacrificial layer. Furthermore, since damage to the material constituting the resonator of the hemispherical resonator gyroscope, which would be caused by etching the sacrificial layer, can be avoided, a resonator with reduced loss of symmetry in terms of shape can be obtained, and improved gyroscope performance (e.g., improved Q value) can be expected. According to embodiments of the present invention, since the process of mechanically aligning electrodes after obtaining a three-dimensional structure is not required, phenomena that lead to performance degradation of the hemispherical resonator gyroscope, such as quadrature signal drift caused by misalignment between electrodes, can be advantageously avoided.

[0115] According to an embodiment of the present invention, the shape of the shell of a hemispherical resonator gyroscope can be controlled without forming a sacrificial layer by combining existing, uncomplicated, and inexpensive processing processes. In particular, since an electrode pattern is formed on a two-dimensional structure such as a quartz wafer and then a curved support structure is obtained by deforming the wafer, the process of subsequently forming electrodes on a three-dimensional structure is unnecessary. This manufacturing method of an embodiment of the present invention is advantageous not only in terms of manufacturing cost but also in terms of mass productivity. For example, in addition to forming a curved support structure on each wafer, it is also possible to simultaneously form multiple (e.g., tens to hundreds) curved support structures on a single wafer.

[0116] 1A, 1B: gyroscope, 10, 20: bare wafer, 10A: resonator, 10WA: upper wafer, 10Wx: inner wafer, 10Wy: outer wafer, 10a, 10b: main surface (of bare wafer), 11A: first electrode, 12c: drive electrode, 12d: detection electrode, 11x, 12x: metal layer, 12: columnar portion, 14A: dome portion, 20A, 20 B: supporting portion, 20WA: lower wafer, 20a, 20b: main surface, 20g: recess, 20h: island-shaped portion, 20t: through-hole, 22A: second electrode, 23A: first electrical path, 24A: second electrical path, 26: through via, 30: supporting substrate, 30A: laminate, 30B: laminate, 30a: upper surface (of supporting substrate), 30ca, 30cb: cavity, 30t: cavity Tee, 50, 53, 55: jig, 50b: lower surface, 50h, 53h, 55h: hole, 52: metal film, 53cp: protrusion, 62, 64: protective film, 70: support substrate, 70h: hole, 72: adhesive layer, 74: protective film, 80A, 80B: resonator, 81: first electrode, 82: columnar portion, 84A: dome portion, 84B: bowl portion, 90A, 90B: support portion, 9 0a: top surface, 90b: bottom surface, 90g: recess, 90t, 90u: cavity, 92: second electrode, 122: terminal, 140: rim, 840: rim, 840b: bottom surface, 900A, 900B: gyroscope, 922: terminal, PA: first electrode pattern, PB: second electrode pattern, PC: third electrode pattern, Q1: first portion, Q2: second portion

Claims

1. A method for manufacturing a curved support structure, comprising: (a) a step of bonding a first wafer provided with a first electrode pattern including a first electrode and a second wafer provided with a second electrode pattern including a plurality of second electrodes to obtain a laminate having an annular cavity therein; (b) a step of heating the first wafer and the second wafer to deform the surface into a curved shape by the air pressure difference between the cavity and the outside of the laminate, thereby increasing or decreasing the volume of the cavity; and (c) a step of removing a portion of the first wafer along a circumference concentric with the annular shape to form a resonator from the first wafer, the central portion of which is supported by the second wafer; wherein in the step (a), the first electrode and the second electrode in the laminate face each other with a gap between them, and the step (a) includes a step of electrically connecting the first electrode to the second electrode pattern of the second wafer.

2. A manufacturing method according to claim 1, wherein step (c) includes removing the portion of the first wafer by laser ablation.

3. A manufacturing method according to claim 1, wherein step (c) includes the step of removing the portion of the first wafer by etching.

4. The manufacturing method according to claim 3, further comprising, between step (b) and step (c), a step of forming a protective film on the entire upper surface of the first wafer and on a portion of the lower surface of the second wafer.

5. A manufacturing method according to claim 2, further comprising, between step (b) and step (c), step (d) of forming a through via in the second wafer that reaches the second electrode pattern, and after step (d), step (e) of bonding a support substrate on which a third electrode pattern is provided to the underside of the second wafer to electrically connect the through via to the third electrode pattern.

6. A manufacturing method according to any one of claims 1 to 5, wherein the step (a) or the step (b) is carried out under reduced pressure.

7. A manufacturing method according to any one of claims 1 to 5, further comprising, before step (a), step (a1) of preparing a wafer having an annular recess on one of its main surfaces, and, between step (a1) and step (a), step (a2) of forming the second electrode pattern on the main surface to obtain the second wafer.

8. A manufacturing method according to any one of claims 1 to 5, wherein step (c) includes a step of removing the portion of the first wafer to connect the cavity to the outside of the laminate.

9. A manufacturing method according to any one of claims 1 to 5, further comprising, between step (a) and step (b), step (e) of placing a jig having a hole on one of the main surfaces of the first wafer on which the first electrode pattern is not provided, and step (b) includes a step of deforming a portion of at least one of the first wafer and the second wafer that is inside the hole of the jig.

10. A method for manufacturing a curved support structure, comprising: (A) a step of bonding a support substrate having a hole to a first wafer provided with a first electrode pattern including a first electrode; (B) a step of separating the first wafer into an outer wafer and a disk-shaped inner wafer surrounded by the outer wafer, the inner wafer including the first electrode, by removing a portion of the first wafer that overlaps the support substrate in a plan view; (C) a step of bonding a second wafer provided with a second electrode pattern including a plurality of second electrodes to the inner wafer and the outer wafer, to obtain a laminate having an annular cavity therein; (D) a step of deforming the surface of at least one of the inner wafer and the second wafer into a curved shape by heating due to a pressure difference between the cavity and the outside of the laminate, thereby increasing or decreasing the volume of the cavity; and (E) a step of removing the support substrate and forming a resonator from the inner wafer, the center of which is supported by the second wafer. In the step (C), the first electrode and the second electrode in the laminate face each other with a gap therebetween.

11. The manufacturing method according to claim 10, wherein the step (C) or the step (D) is carried out under reduced pressure.

12. A manufacturing method according to claim 10 or 11, further comprising, before step (C), a step (C1) of preparing a wafer having an annular recess on one of its main surfaces, and, between steps (C1) and (C), a step (C2) of forming the second electrode pattern on the main surface to obtain the second wafer.

13. A manufacturing method according to claim 10 or 11, wherein step (C) includes a step of electrically connecting the first electrode to the second electrode pattern of the second wafer.

14. A manufacturing method according to claim 10 or 11, further comprising, between step (C) and step (D), step (F) of placing a jig having a hole on the support substrate, wherein step (D) includes a step of deforming a portion of at least one of the inner wafer and the second wafer that is inside the hole of the jig.

15. A manufacturing method according to claim 1 or claim 10, wherein the resonator has a columnar portion extending upward from the second wafer, and a rim arranged in an annular shape around the columnar portion.

Images