JP2011203561A - Mems element and method for manufacturing the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent an electric field-shielding base board from colliding with a leaf spring to improve the yields of elements.SOLUTION: An MEMS element includes: a mirror base board 100; an electrode base board 200 which is arranged facing the mirror base board 100; an electric field-shielding base board 300a which is joined to the mirror base board 100; a leaf spring 105a which is a movable electrode formed with its displacement possible to the mirror base board 100; a leaf spring-driving electrode 201a which is a fixed electrode formed on the electrode base board 200; and a shielding electrode 302 which is formed on the surface of the electric field-shielding base board 300a on the opposite side to the mirror base board 100, and is inserted and fitted in the aperture part 101 of the mirror base board 100. A size in the horizontal direction of the electric field-shielding base board 300a is larger than the size in the horizontal direction of the aperture part 101, and the thickness of the shielding electrode 302 is equal to the depth of the aperture part 101 or smaller than the depth of the aperture part 101.

Description

  The present invention relates to an electrostatically driven MEMS element used for a wavelength selective switch for optical communication and the like, and a method for manufacturing the same.

  Optical communication has come to be used to transmit large amounts of information (data). In an optical network, elements that generate, transmit, switch and receive infrared rays are used. Information is transmitted by electronically controlling these elements. An element used for switching light is called an optical switch element, which is a wavelength selective switch (WSS) that demultiplexes wavelength-multiplexed light and switches an optical path for each wavelength. There is a switch such as an optical cross connect switch (OXC) that switches the optical path without demultiplexing the transmitted light. These optical switches are configured using MEMS (Micro Electro Mechanical Systems), liquid crystal, PLC (Planar Waveguide Circuits), or the like.

  When using a PLC, an optical switch is realized by a TO switch that changes an interference state according to temperature and switches an optical path by using an interference device constituted by an optical waveguide and a local heating structure. In this case, since a 1 × 2 switch can be realized, a 1 × 2 switch is connected in multiple stages to form a 1 × N switch. A WSS is configured by combining this optical switch, an element that demultiplexes the wavelength-multiplexed light for each wavelength, and an element that is multiplexed after switching the optical path. In the case of PLC, demultiplexing and multiplexing are realized by AWG (Arrayed Waveguide Grating), so that switching can be realized without taking out signal light from the optical waveguide to the outside.

  On the other hand, WSS using MEMS and liquid crystal takes out the signal light transmitted through the fiber once into the space outside the fiber, and uses wavelength dispersion elements such as gratings to wavelength-multiplex the light. Each one leads to a different position. In the case of WSS using MEMS, MEMS mirrors are arranged at different positions for each wavelength of light. In the case of a switch that handles n-wave light, n MEMS mirrors are arranged. By changing the relative angle between the light applied to the MEMS mirror and the MEMS mirror, control of the direction in which the reflected light propagates, that is, switching, is realized.

When a wavelength dispersion element such as a grating is used for demultiplexing light, light of different wavelengths can be separated, but light of continuous wavelengths is demultiplexed into a continuous space as seen in a rainbow. That is, two lights having a small wavelength difference are demultiplexed into adjacent spaces. As light for communication, light having a center wavelength at a frequency called a grid defined by ITU (International Telecommunication Union) is used. Therefore, in WSS, it is essential to arrange the number of MEMS mirrors corresponding to the number of wavelengths that can be handled in an array in close proximity at intervals according to the wavelength (exactly according to the wavelength and the optical system). In the current MEMS mirror, the distance between the mirrors is about 100 μm.
In the MEMS mirror array, a mirror structure using electromagnetic force or a mirror structure using electrostatic force is used. The mirror structure using any force is designed so that the force acts between the movable structure and the fixed structure, that is, between the separated structures.

In MEMS mirrors used for WSS, in principle, the mirrors must be placed close to each other and the force acting between the distant structures is used. There is a problem that so-called interference occurs. When electrostatic attraction is used, electrical interference occurs. When electromagnetic force is used, magnetic interference occurs.
In any case, in order to suppress interference, it is necessary to use a shielding technique such as an electric field shield or a magnetic shield. However, it is very difficult to place an effective shielding structure between MEMS mirrors close to about 100 μm. It is difficult to.

  As a method of preventing interference between mirrors, as shown in Patent Document 1, a wall-like structure is arranged between adjacent mirror structures, or as shown in Patent Document 2, There are a method of increasing the distance to the mirror or the wiring to be used, a method of using a multilayer structure in which a structure such as a wiring to which a potential is applied is arranged below the planar GND electrode, and the like. However, the MEMS mirror array used in WSS is required to place the mirror structure at the exact position determined by optical design, and the distance between the mirror structure and the distance between the wiring and the mirror structure can be freely increased. However, it is difficult to increase the electrical interference, and there is a problem that electrical interference increases.

  In addition, it is very effective to place a structure such as a wall between adjacent mirror structures, but considering the characteristics of the switch, the distance between adjacent mirrors should be as small as possible. There is a problem that a compromise must be made by lowering the characteristics of the switch. When a mirror is formed by a bulk micromachining technique for joining a mirror chip and an electrode chip, a structure for electric shielding is generally formed on the electrode chip side.

For example, in the case of an element adopting a structure in which the drive electrode structure is formed in three dimensions and the movable electrode enters into the electrode structure, the electric field that wraps around from the upper part of the movable electrode is the movable electrode of the adjacent element May be affected.
In order to shield such an electric field, Non-Patent Document 1 discloses an example in which an electric field shield structure is also formed on the mirror substrate side as shown in FIGS. 8, 9, and 10, for example. 8 is a perspective view of the WSS chip disclosed in Non-Patent Document 1, FIG. 9 is a cross-sectional view taken along line AA ′ of the WSS chip of FIG. 8, and FIG. 10 is a cross-sectional view taken along line CC ′ of the WSS chip of FIG. It is.

  The WSS chip has a structure in which the mirror substrate 100 and the electrode substrate 200 are bonded to each other, and the electric field shield substrate 300 is fitted into the opening 101 having a rectangular shape in plan view formed in the mirror substrate 100. As shown in FIGS. 9 and 10, the mirror substrate 100 includes a support layer 102 made of silicon, a buried oxide (BOX) layer 103 made of a silicon oxide film, an SOI (Silicon on Insulator) layer 104, and the like. Consists of. An opening 101 through which the SOI layer 104 is exposed is formed in the support layer 102 and the BOX layer 103. Leaf springs 105a and 105b, a mirror 106, and connecting portions 107a and 107b are formed on the SOI layer 104 of the opening 101. The mirror 106 can rotate around a first rotation axis that passes through the pair of connecting portions 107 a and 107 b and is parallel to the y axis, and is parallel to the x axis that is orthogonal to the length direction of the mirror 106. It is possible to rotate around 2 rotation axes. The alignment of the leaf spring 105a, the connecting portion 107a, the mirror 106, the connecting portion 107b, and the leaf spring 105b forms a structure on the mirror substrate side of one MEMS mirror element, and such a structure is an x axis perpendicular to the alignment direction. A plurality of elements are arranged along the direction to form a structure on the mirror substrate side of the MEMS mirror array.

On the electrode substrate 200, leaf spring drive electrodes 201a and 201b and a mirror drive electrode 202 are provided for each MEMS mirror element. The leaf spring drive electrodes 201a and 201b have a U-shaped cross section in which grooves are formed along a direction parallel to the y-axis, that is, a direction parallel to the alignment direction of the leaf springs 105a and 105b. Such leaf spring drive electrodes 201a and 201b and mirror drive electrode 202 form a structure on the electrode substrate side of one MEMS mirror element, and a plurality of such structures are arranged along the x-axis direction, and MEMS is provided. A structure on the electrode substrate side of the mirror array is formed.
Since the electric field shield substrate 300 is formed as a chip thinner than the depth of the opening 101 of the mirror substrate 100 (the thickness obtained by adding the support layer 102 and the BOX layer 103), the electric field shield substrate 300 is higher than the height of the surface of the support layer 102. The surface of the electric field shield substrate 300 is positioned slightly lower.

JP 2007-248809 A International Publication WO06 / 073111

M.Usui, S.Uchiyama, E.Hashimoto, K.Hadama, Y.Ishii, N. Matsuura, T.Sakata, N.Shimoyama, Y.Sato, H.Ishii, T.Matsuura, F.Shimokawa, and Y Uenishi, “Electrically Separated Two-axis MEMS Mirror Array Module for Wavelength Selective Switches”, Proc. Of Optical MEMS '2009, Clearwater Beach, Florida, USA, Aug. 17-20, 2009, pp.158-159

  As shown in FIGS. 9 and 10, the electric field shield substrate 300 is disposed so as to be in contact with the SOI layer 104 of the mirror substrate 100, but it is structurally desired to be disposed so as not to contact the leaf springs 105 a and 105 b. It is. However, actually, when the electric field shield substrate 300 is inserted into the opening 101 of the mirror substrate 100, the electric field shield substrate 300 may come into contact with the leaf springs 105a and 105b. If they come into contact with the leaf springs 105a and 105b, there is a greater possibility that the mirror structure such as the leaf springs 105a and 105b will be damaged, and the yield of the elements will be reduced.

  The present invention has been made to solve the above-described problems, and can prevent collision between an electric field shield substrate and a movable electrode such as a leaf spring, and can improve the yield of the device and its manufacture. It aims to provide a method.

The MEMS element of the present invention includes a first substrate, a second substrate disposed to face the first substrate, and a surface of the first substrate that does not face the second substrate. A bonded third substrate, a movable electrode formed to be displaceable on the first substrate, and an opening from the surface of the first substrate not facing the second substrate to the movable electrode And a fixed electrode for driving the movable electrode, disposed on the surface of the second substrate facing the first substrate in a state where a gap is provided between the movable substrate and the movable electrode. And a shield electrode formed on a surface of the third substrate facing the first substrate and inserted into an opening of the first substrate, and the horizontal direction of the third substrate Is larger than the horizontal size of the opening of the first substrate, and the thickness of the shield electrode is It is characterized in thinner than the depth of the opening of the first of the same or the the depth of the opening of the substrate and the first substrate.
Moreover, in one structural example of the MEMS element of the present invention, the shield electrode has a recess formed at a position facing the movable electrode.

The method for manufacturing a MEMS device according to the present invention includes a movable electrode forming step of forming a movable electrode on an SOI layer of a first SOI substrate, a silicon support layer of the first SOI substrate, and a part of a buried oxide film layer. Then, the silicon support layer and the buried oxide film layer existing in the region where the movable electrode is formed are removed by etching until the movable electrode is exposed to form an opening in the first SOI substrate. An opening forming step, a first dicing step of processing the first SOI substrate into a predetermined size, an electrode substrate forming step of forming a fixed electrode on the electrode substrate, the first SOI substrate, and the A first substrate bonding step in which an electrode substrate is disposed and bonded so that the movable electrode and the fixed electrode are spaced apart from each other; and a thickness of the first SOI substrate and the silicon support layer is the same. 2 SOI groups A shield electrode forming step of etching and removing the silicon support layer so that a shield electrode region smaller than the horizontal size of the opening of the first SOI substrate remains, and the second SOI substrate having a predetermined size. A second dicing step to be processed, and a shield electrode of the second SOI substrate is inserted into an opening of the first SOI substrate, so that the first SOI substrate and the second SOI substrate are A second substrate bonding step of bonding, wherein a horizontal size of the second SOI substrate after processing is larger than a horizontal size of the opening of the first SOI substrate. To do.
Further, according to one configuration example of the method for manufacturing a MEMS element of the present invention, a buried oxide film is formed by etching away a buried oxide film layer around a shield electrode of the second SOI substrate before the second dicing step. A film layer removing step is provided.

According to another aspect of the present invention, there is provided a method of manufacturing a MEMS device, comprising: a movable electrode forming step of forming a movable electrode on an SOI layer of an SOI substrate; An opening forming step of forming an opening in the SOI substrate by etching and removing the silicon support layer and the buried oxide film layer existing in the region where the electrode is formed until the movable electrode is exposed; and the SOI substrate A first dicing step of processing the substrate into a predetermined size, an electrode substrate forming step of forming a fixed electrode on the electrode substrate, the SOI substrate and the electrode substrate, the movable electrode and the fixed electrode being separated from each other And a first substrate bonding step for arranging and bonding so as to oppose each other, and an electric field shield substrate with a sheet made of a conductive material smaller than the horizontal size of the opening of the SOI substrate. A shield electrode joining step for joining a shield electrode, a second dicing step for processing the electric field shield substrate into a predetermined size, and inserting the shield electrode into an opening of the SOI substrate, A second substrate bonding step for bonding to the electric field shield substrate, wherein the horizontal size of the electric field shield substrate after processing is larger than the horizontal size of the opening of the SOI substrate, and the shield electrode The thickness of is the same as the depth of the opening of the SOI substrate or is thinner than the depth of the opening of the SOI substrate.
Moreover, one structural example of the manufacturing method of the MEMS element of this invention is further equipped with the recessed part formation process which forms a recessed part in the position facing the said movable electrode of the said shield electrode before the said 2nd dicing process. It is characterized by.

  According to the present invention, the shield electrode is inserted into the opening portion of the first substrate by making the horizontal size of the third substrate larger than the horizontal size of the opening portion of the first substrate. The third substrate comes into contact with the first substrate, and the position of the third substrate in the thickness direction is fixed. At this time, since the thickness of the shield electrode is the same as the depth of the opening of the first substrate or thinner than the depth of the opening of the first substrate, the shield electrode and the movable electrode do not collide. As a result, in the present invention, a third substrate having a shape and structure capable of realizing the effect as an electric field shield substrate for preventing electrical interference can be realized, and the third substrate can be bonded to the first substrate. In addition, it is possible to prevent the movable electrode from being damaged by the collision of the shield electrode and the movable electrode, and the yield can be improved.

  In the present invention, the possibility of collision between the shield electrode and the movable electrode can be further reduced by forming the recess at a position facing the movable electrode of the shield electrode.

  Further, in the present invention, the silicon support layer of the second SOI substrate having the same thickness as the first SOI substrate on which the movable electrode is formed is made larger than the horizontal size of the opening of the first SOI substrate. And the second SOI substrate is processed so that the horizontal size after processing of the second SOI substrate is larger than the horizontal size of the opening of the first SOI substrate. The SOI substrate is processed, the shield electrode of the second SOI substrate is inserted into the opening of the first SOI substrate, and the first SOI substrate and the second SOI substrate are joined. Accordingly, in the present invention, when the shield electrode is fitted into the opening of the first SOI substrate, the second SOI substrate comes into contact with the first SOI substrate, and the thickness direction of the second SOI substrate The position of is fixed. At this time, since the thickness of the shield electrode is the same as the thickness of the support layer of the first SOI substrate, the shield electrode and the movable electrode do not collide. As a result, in the present invention, a second SOI substrate having a shape and structure capable of realizing the effect as an electric field shield substrate for preventing electrical interference can be realized, and the second SOI substrate can be used as the first SOI substrate. When joining, it can prevent that a shield electrode and a movable electrode collide and a movable electrode is damaged, and a yield can be improved.

  In the present invention, the electric field shield substrate is made of a conductive material that is smaller than the horizontal size of the opening of the SOI substrate and is the same as the depth of the opening of the SOI substrate or thinner than the depth of the opening of the SOI substrate. The shield electrode is joined, the electric field shield substrate is processed so that the horizontal size of the electric field shield substrate after processing is larger than the horizontal size of the opening of the SOI substrate, and the shield electrode is opened in the SOI substrate. The SOI substrate and the electric field shield substrate are bonded to each other by being inserted into the part. Thus, in the present invention, when the shield electrode is inserted into the opening of the SOI substrate, the electric field shield substrate comes into contact with the SOI substrate, and the position of the electric field shield substrate in the thickness direction is fixed. At this time, since the thickness of the shield electrode is the same as the depth of the opening of the SOI substrate or thinner than the depth of the opening of the SOI substrate, the shield electrode and the movable electrode do not collide. As a result, in the present invention, an electric field shield substrate having a shape and structure capable of realizing its effect as a substrate for preventing electrical interference can be realized, and when the electric field shield substrate is bonded to the SOI substrate, the shield electrode and the movable electrode can be moved. It is possible to prevent the movable electrode from being damaged by colliding with the electrode, and the yield can be improved.

It is a perspective view of a WSS chip concerning an embodiment of the invention. 1 is a cross-sectional view of a WSS chip according to an embodiment of the present invention. It is sectional drawing of the WSS chip | tip before inserting an electric field shield board | substrate in a mirror board | substrate in embodiment of this invention. 1 is a cross-sectional view of a WSS chip according to an embodiment of the present invention. It is a perspective view which shows the structure of the mirror of the WSS chip | tip which concerns on embodiment of this invention, a leaf | plate spring, and an electrode. It is sectional drawing which shows a mode when a leaf | plate spring is displaced below in embodiment of this invention. It is sectional drawing which shows a mode when a leaf | plate spring is not displaced in embodiment of this invention. It is a perspective view of the conventional WSS chip. It is sectional drawing of the conventional WSS chip | tip. It is sectional drawing of the conventional WSS chip | tip.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a perspective view of a WSS chip which is a MEMS element according to an embodiment of the present invention. The WSS chip has a structure in which the mirror substrate 100 and the electrode substrate 200 are joined, and the electric field shield substrate 300a is inserted into the opening 101 having a rectangular shape in plan view formed in the mirror substrate 100. Since the electric field shield substrate 300a has a protruding portion 301 wider than the opening 101 obtained by etching away the support layer of the mirror substrate 100, the electric field shield substrate 300a is fitted into the opening 101 of the mirror substrate 100. At this time, the protruding portion 301 is bonded so as to be in contact with the surface of the support layer of the mirror substrate 100.

  2 is a cross-sectional view taken along line BB ′ of the WSS chip of FIG. 1, FIG. 3 is a cross-sectional view of the WSS chip before the electric field shield substrate 300a is inserted into the mirror substrate 100, and FIG. FIG. 5 is a perspective view showing a structure of a mirror, a leaf spring, and an electrode of the WSS chip of FIG. In FIG. 5, the electric field shield substrate 300a above the leaf spring is not shown for easy description.

  As shown in FIGS. 2 to 5, the mirror substrate 100 includes a support layer 102 made of silicon, a BOX layer 103 made of a silicon oxide film, and an SOI layer 104. An opening 101 through which the SOI layer 104 is exposed is formed in the support layer 102 and the BOX layer 103. In the SOI layer 104 of the opening 101, leaf springs 105a and 105b, which are movable electrodes, a mirror 106, and connecting portions 107a and 107b are formed. One end of each of the leaf springs 105a and 105b is fixed to a frame portion (not shown) which is the peripheral SOI layer 104. 2 to 5, the leaf spring 105a and the leaf spring 105b are aligned on a line parallel to the y-axis direction. The plate spring 105a and the plate spring 105b each have a cantilever structure in which the other end can be displaced in the normal direction (z-axis direction) of the mirror substrate.

  A mirror 106 is disposed between the leaf spring 105a and the leaf spring 105b by being coupled by a pair of bendable coupling portions 107a and 107b. The mirror 106 is arranged in a row with the leaf spring 105a and the leaf spring 105b, and is rotatably disposed between the leaf spring 105a and the leaf spring 105b. The connecting portion 107a connects the other end of the leaf spring 105a and the mirror 106, and the connecting portion 107b connects the other end of the leaf spring 105b and the mirror 106. 2 to 5, the leaf spring 105a, the connecting portion 107a, the mirror 106, the connecting portion 107b, and the leaf spring 105b are aligned in order on a line parallel to the y-axis direction.

  The mirror 106 can rotate around a first rotation axis that passes through the pair of connecting portions 107 a and 107 b and is parallel to the y axis, and is parallel to the x axis that is orthogonal to the length direction of the mirror 106. It is possible to rotate around 2 rotation axes. A reflective film made of gold or aluminum is formed on the surface of the mirror 106, and can reflect light in the infrared region, for example. The alignment of the leaf spring 105a, the connecting portion 107a, the mirror 106, the connecting portion 107b, and the leaf spring 105b as described above forms a structure on the mirror substrate side of one MEMS mirror element. A plurality of elements are arranged along the vertical x-axis direction to form a structure on the mirror substrate side of the MEMS mirror array.

  On the other hand, leaf spring drive electrodes 201a and 201b and mirror drive electrodes 202a and 202b are provided on the electrode substrate 200 for each MEMS mirror element. When the mirror substrate 100 and the electrode substrate 200 are joined, the leaf spring drive electrodes 201a and 201b are disposed so as to face the leaf springs 105a and 105b, respectively, and the mirror drive electrodes 202a and 202b face the mirror 106. Are arranged as follows. The leaf spring drive electrodes 201a and 201b have a U-shaped cross section in which grooves are formed along a direction parallel to the y-axis, that is, a direction parallel to the alignment direction of the leaf springs 105a and 105b. As shown in FIG. 6, when the leaf springs 105a and 105b are displaced downward, the leaf spring drive electrodes 201a and 201b are driven without touching the opposing leaf spring drive electrodes 201a and 201b. It is set so that it can enter into the grooves of the electrodes 201a and 201b.

  The plate spring drive electrodes 201a and 201b and the mirror drive electrodes 202a and 202b as described above form a structure on the electrode substrate side of one MEMS mirror element, and a plurality of such structures are arranged along the x-axis direction. Thus, the structure on the electrode substrate side of the MEMS mirror array is formed. That is, a pair of leaf springs 105a and 105b, a mirror 106, a pair of connecting portions 107a and 107b, a pair of leaf spring drive electrodes 201a and 201b, and a pair of mirror drive electrodes 202a and 202b constitute one MEMS mirror element. It is configured.

  A drive voltage for driving the leaf springs 105a and 105b is supplied to the leaf spring drive electrodes 201a and 201b via a leaf spring drive electrode wiring (not shown). Further, a drive voltage for driving the mirror 106 is supplied to the mirror drive electrodes 202a and 202b via a mirror drive electrode wiring (not shown). The leaf springs 105a and 105b, the mirror 106, and the connecting portions 107a and 107b are equipotential. Here, the equipotential may be a ground potential, for example.

  Next, the operation of the MEMS mirror element will be described. First, the rotation around the second rotation axis will be described. First, when a predetermined driving voltage is applied to the leaf spring drive electrode 201b to apply a force that attracts the leaf spring 105b to the electrode substrate 200 side by the generated electrostatic attractive force, the leaf spring 105b is applied to the frame portion. The supported end is bent (deformed) as a fulcrum, and the other end of the leaf spring 105b is displaced so as to be drawn toward the electrode substrate 200 side. As a result, the mirror 106 rotates with the connecting portion 107a as a fulcrum so that the connecting portion 107b side is pulled toward the electrode substrate 200 side. This means that the mirror 106 is rotated around the second rotation axis that passes through the center portion of the mirror 106 and is parallel to the arrangement direction (x-axis direction) of the MEMS mirror elements.

  On the other hand, when a predetermined driving voltage is applied to the leaf spring drive electrode 201a to apply a force that attracts the leaf spring 105a toward the electrode substrate 200 by electrostatic attraction, the leaf spring 105a is supported by the frame portion. The other end of the leaf spring 105a is displaced so as to be drawn toward the electrode substrate 200 side. As a result, the mirror 106 rotates with the connecting portion 107b as a fulcrum so that the connecting portion 107a side is pulled toward the electrode substrate 200 side. This rotation operation rotates the mirror 106 in the opposite direction to the case where the drive voltage is applied to the leaf spring drive electrode 201b.

  Next, the rotation around the first rotation axis will be described. By controlling the voltage applied to the mirror drive electrodes 202a and 202b, the mirror 106 can be rotated around the first rotation axis passing through the pair of connecting portions 107a and 107b. For example, when a voltage higher than that of the mirror drive electrode 202b is applied to the mirror drive electrode 202a, the mirror 106 rotates around the first rotation axis so as to be inclined toward the mirror drive electrode 202a. As described above, the mirror 106 rotates about two orthogonal axes.

  In the present embodiment, the leaf spring drive electrodes 201a and 201b are U-shaped in cross section, thereby suppressing crosstalk between adjacent MEMS mirror elements as described below. Will be able to. Since the MEMS mirror element is arranged with a small interval between adjacent MEMS mirror elements, when the plate spring drive electrodes 201a and 201b are simple parallel plate electrodes, the electrostatic attraction force is a plate spring of the MEMS mirror element. In addition to the above, the leaf spring of the adjacent MEMS mirror element is also affected, and its position may be displaced. As a result, electrical interference (crosstalk) may occur between adjacent mirrors 106.

  In contrast, in the present embodiment, the leaf spring drive electrodes 201a and 201b are U-shaped in cross section, so that the electric field for driving the leaf springs can be separated for each MEMS mirror element. And crosstalk can be suppressed. However, as shown in FIG. 6, when the leaf springs 105a and 105b enter the grooves of the leaf spring drive electrodes 201a and 201b, the electric field that wraps around from above the leaf spring drive electrodes 201a and 201b is adjacent to the MEMS. There is a possibility of affecting the leaf springs 105a and 105b of the mirror element. Therefore, in the present embodiment, the electric field shield substrate 300a is inserted into the opening 101 of the mirror substrate 100 in order to prevent such a wraparound of the electric field.

  As shown in FIGS. 2 to 4, the electric field shield substrate 300 a includes a shield electrode 302 made of silicon, a BOX layer 303, and an SOI layer 304. The shield electrode 302 is equipotential (ground potential) with the leaf springs 105a and 105b, the mirror 106, and the connecting portions 107a and 107b.

  When the electric field shield substrate 300a is inserted into the opening 101 of the mirror substrate 100, the shield electrode 302, the BOX layer 303, and the SOI layer 304 are disposed above the leaf springs 105a and 105b, and the mirror 106 and the connecting portion 107a, An opening 305 that penetrates the shield electrode 302, the BOX layer 303, and the SOI layer 304 is formed in the electric field shield substrate 300a so that 107b is exposed. When the element of this embodiment is used as a WSS chip, as shown in FIG. 4, the signal light is reflected by being incident on the mirror 106 through the opening 305 and is emitted to the outside again through the opening 305. Will do. As described above, by controlling the drive voltage applied to the leaf spring drive electrodes 201a and 201b and the mirror drive electrodes 202a and 202b, the rotation angle of the mirror 106 can be controlled, and the incident light is set to a desired angle. It can be reflected.

  The horizontal size (x-axis direction and y-axis direction) of the BOX layer 303 and the SOI layer 304 of the electric field shield substrate 300a is set to be larger than the horizontal size of the opening 101 of the mirror substrate 100. Yes. A BOX layer 303 and an SOI layer 304 that are wider than the opening 101 constitute an overhang portion 301. When the electric field shield substrate 300a is lowered when the electric field shield substrate 300a is inserted into the opening 101 of the mirror substrate 100, the BOX layer 303 of the overhanging portion 301 is formed on the mirror substrate 100 as shown in FIGS. In contact with the support layer 102, the position of the electric field shield substrate 300a in the z-axis direction is fixed. The thickness of the shield electrode 302 of the electric field shield substrate 300 a is set to be the same as or thinner than the support layer 102 of the mirror substrate 100. Therefore, when the electric field shield substrate 300a is fitted into the opening 101 of the mirror substrate 100, the shield electrode 302 does not collide with a mirror structure such as the leaf springs 105a and 105b. The difference between the thickness obtained by adding the support layer 102 and the BOX layer 103 of the mirror substrate 100 and the thickness of the shield electrode 302 is the distance between the shield electrode 302 and the leaf springs 105a and 105b.

  Further, a recess 306 is formed on the surface of the shield electrode 302 facing the leaf springs 105a and 105b. When the electric field shield substrate 300a is inserted into the opening 101 of the mirror substrate 100, the recess 306 faces the leaf springs 105a and 105b and is parallel to the y axis, that is, the alignment direction of the leaf springs 105a and 105b. Are formed so as to extend along a direction parallel to the. As shown in FIG. 7, the size of the recess 306 is set so that it can enter the recess 306 when the leaf springs 105 a and 105 b are not displaced.

  Next, a method for manufacturing the WSS chip according to the present embodiment will be described. First, a method for manufacturing the mirror substrate 100 will be described. The mirror substrate 100 is manufactured using an SOI substrate. The leaf springs 105a and 105b, the mirror 106, and the connecting portions 107a and 107b are formed on the SOI layer 104 by using a known lithography technique using the SOI substrate as a starting substrate.

  First, a photosensitive resist is applied in a desired film thickness on the surface of the SOI layer 104 using a spin coating method. A reticle (mask) having a light-shielding body having a shape corresponding to a pattern is aligned and held on a substrate coated with a photosensitive resist. The reticle is irradiated with light having the photosensitive wavelength of the photosensitive resist, and the shadow of the light shielding body is imaged on the wafer coated with the photosensitive resist. Irradiation is performed after determining the irradiation time so that a desired pattern can be obtained with a desired dimension. The resist with the latent image printed thereon is immersed in a developer, and the time for immersing the resist-coated wafer in the developer is determined so that a desired pattern can be formed with a desired dimension. Is immersed in the developer. Thereafter, the entire wafer is dried to cure the resist. Thus far, a resist pattern can be formed on the SOI substrate.

  Next, using the formed resist as a mask, leaf springs 105a and 105b, a mirror 106, coupling portions 107a and 107b, and the like are applied to the SOI layer 104 using a DRIE (Deep Reactive Ion Etching) etching technique which is a deep silicon processing technique. Form. At that time, an etching intermediate film may be used, or silicon may be directly processed using a resist as an etching mask. The photosensitive resist used for the mask is removed after the silicon etching. After removal, cleaning is performed to expose a clean silicon surface.

  Thereafter, the base substrate (support layer 102) is processed. First, a protective organic film is applied to the SOI surface side. A photosensitive resist may be used, or a film with clear workability such as polyimide may be used. Since this film must be finally removed, it is desirable to use a film having excellent releasability and removability. However, since a resist is used for processing the back surface with the protective film attached, it is essential to select a material that does not cause a problem at that time.

  After a protective film is formed on the surface of the SOI layer 104, a photosensitive resist film is applied to the support layer 102. In the case of the application of the photosensitive resist, the SOI surface side comes into contact with the sample table of the coating apparatus, but since the protective film is deposited, the structure formed on the SOI surface is not damaged.

  After the photosensitive resist is applied, a reticle (mask) having a light-shielding body corresponding to a desired pattern is aligned with the substrate coated with the photosensitive resist, particularly with the structure on the SOI surface. Hold. The reticle is irradiated with light having the photosensitive wavelength of the photosensitive resist, and the shadow of the light shielding body is imaged on the wafer coated with the photosensitive resist. Then, the irradiation is performed after determining the light irradiation time so that a desired pattern can be obtained with a desired dimension. The resist with the latent image printed thereon is immersed in a developer, and the time for immersing the resist-coated wafer in the developer is determined so that a desired pattern can be formed with a desired dimension. Is immersed in the developer. Thereafter, the entire wafer is dried to cure the resist. Thus far, a resist pattern can be formed on the base substrate.

  Next, using the formed resist as a mask, the silicon of the unnecessary support layer 102 in the portion where the mirror 106 and the coupling portions 107a and 107b are formed on the SOI layer 104 is etched using the DRIE etching technique which is a deep silicon processing technique. Remove. At that time, an etching intermediate film such as a silicon oxide film may be used, or silicon may be directly processed using a resist as an etching mask. The photosensitive resist used for the mask is removed after the silicon etching. After removal, cleaning is performed to expose a clean silicon surface.

  Next, the BOX layer 103 is removed by etching. Since the BOX layer 103 is formed of a silicon oxide film, wet etching can be performed with a hydrofluoric acid-based solution such as dilute hydrofluoric acid. Since wet etching of a silicon oxide film using a hydrofluoric acid-based solution is isotropic etching, the silicon oxide film below the support layer 102 is also partially etched. When design or characteristic problems become obvious, dry etching is used instead of wet etching. In any case, the etching of the silicon oxide film does not cause a problem even if a known etching technique is used. Thus, by removing the support layer 102 and the BOX layer 103 by etching, an opening 101 having a rectangular shape in plan view could be formed in the mirror substrate 100.

  Thereafter, the organic film protecting the pattern such as the mirror structure of the SOI layer 104 is removed by etching to form a movable structure. In the case where a plurality of chips are formed on the wafer, the protective film of the SOI layer 104 is removed by etching after a process called dicing, which is a process of cutting the wafer and cutting it into chips.

  Finally, a metal film is deposited on the front and back surfaces of the mirror 106 in order to reflect light having a communication wavelength. Here, for example, a metal such as gold or aluminum is deposited in order to increase the reflectance of infrared rays to be used. When the surface of the structure is silicon, an adhesion improving layer such as titanium or chromium is used in order to improve adhesion of deposited gold or aluminum to silicon. When the flatness of the mirror 106 can be improved, it is not necessary to deposit metal on the back surface of the mirror 106 (the lower surface in FIG. 4). In order to partially deposit the metal film, the metal is deposited using a stencil mask or the like using an evaporation method or a sputtering method. With the above method, the mirror substrate 100 including the rotatable mirror 106 is formed.

  Next, a method for manufacturing the electrode substrate 200 will be described. First, a thermal oxide film is formed on the surface of the base substrate. A metal sputtered film is deposited on the surface of the thermal oxide film and used as the first seed layer in the subsequent plating process. Thereafter, a resist is applied, and a resist pattern necessary for forming the leaf spring driving electrode wiring and the mirror driving electrode wiring is formed using a known lithography technique. Then, using this resist pattern as a mask, a metal such as gold is grown by plating.

  After forming the leaf spring drive electrode wiring and the mirror drive electrode wiring, the resist used as a mask is peeled off, and after cleaning, unnecessary first seed layer portions are removed by etching. An insulating film such as a silicon oxide film is deposited on the base substrate on which the leaf spring driving electrode wiring and the mirror driving electrode wiring are formed by a CVD (Chemical Vapor Deposition) method or the like. When the flatness of the surface of the silicon oxide film is poor, a flattening step is performed after depositing an insulating film. A resist is applied onto the insulating film by a spin coating method, and then a resist pattern for forming a via hole is formed by a known lithography technique. Using this resist as a mask, the insulating film is etched using a known etching technique to form a via hole. After forming the via hole, the resist covering the surface is peeled off, and cleaning is performed so that the surface becomes clean.

  Next, a metal sputtered film is deposited on the surface of the insulating film and used as a second seed layer in the subsequent plating process. A resist is applied, and a resist pattern for forming the leaf spring drive electrodes 201a and 201b and the mirror drive electrodes 202a and 202b is formed using a known lithography technique. Then, the plate spring drive electrodes 201a and 201b and the mirror drive electrodes 202a and 202b are formed by plating using the resist pattern as a mask. Then, after removing the resist, the unnecessary second seed layer is removed by etching, and the surface is cleaned again by cleaning. The leaf spring drive electrodes 201a, 201b and the lower leaf spring drive electrode wiring are connected via via holes formed in the insulating film. Similarly, the mirror drive electrodes 202a and 202b and the mirror drive electrode wiring below the mirror drive electrodes 202a and 202b are connected via via holes.

  Subsequently, on the outside of the leaf spring drive electrodes 201a and 201b and the mirror drive electrodes 202a and 202b on the insulating film, the gap between the leaf springs 105a and 105b and the leaf spring drive electrodes 201a and 201b, and the mirror 106 and the mirror drive electrodes 202a and 202b, A pedestal for forming a gap between 202b is formed. It goes without saying that the height of the pedestal is set to be higher than the leaf spring drive electrodes 201a and 201b. In order to form such a pedestal, a thick resist is applied on the insulating film by a spin coating method, and then a resist pattern is formed by a known lithography technique. Using this resist as a mask, a pedestal is formed on the insulating film by plating, and then the resist covering the surface is peeled off and washed so that the surface becomes clean.

  The electrode substrate 200 is formed by the above method. In the method of forming the electrode substrate 200 described above, the leaf spring driving electrodes 201a and 201b, the mirror driving electrodes 202a and 202b, the leaf spring driving electrode wiring, the mirror driving electrode wiring, and the pedestal are made of gold (Au). The invention is not limited to this. As materials for these electrodes, wirings and pedestals, aluminum (Al), tungsten (W), tantalum (Ta), titanium (Ti), copper (Cu), tin (Sn), silver (Ag), and the like can be used. In addition, compounds of these materials may be used. Further, a semiconductor material such as silicon may be used.

In this embodiment, each structure is formed by using a plating method. However, as long as the conductive film can be deposited with a desired thickness, a known film such as a sputtering method, a CVD method, or an evaporation method can be used. There is no problem using the deposition technique.
When a plurality of chips are formed on the wafer, the electrode substrate 200 is completed after a process called dicing, which is performed by cutting the wafer and cutting the wafer into chips.

  After the mirror substrate 100 and the electrode substrate 200 are separately formed, the mirror substrate 100 and the electrode substrate 200 are aligned, an Ag paste is applied to the upper surface of the pedestal of the joint portion of the electrode substrate 200, and the SOI of the mirror substrate 100 is obtained. In a state where the layer 104 and the pedestal of the electrode substrate 200 are in contact with each other, the Ag paste is solidified by pressing and heating, and the mirror substrate 100 is bonded to the electrode substrate 200. Thus, the mirror substrate 100 and the electrode substrate 200 are arranged in parallel to each other, and the mirror substrate 100 is fixed on the electrode substrate 200 at a predetermined distance.

  Next, a method for manufacturing the electric field shield substrate 300a will be described. A recess 306 is formed on the surface of the support layer to be the shield electrode 302 by using an SOI substrate having the same thickness as that of the SOI substrate used for manufacturing the mirror substrate 100. As shown in FIGS. 2, 3, 6, and 7, the recess 306 may be inclined at the side surface or may be vertical at the side surface.

  In order to form the recess 306, first, a resist is applied to the surface of the support layer of the SOI substrate, and a resist pattern is formed using a known lithography technique using a mask having a desired pattern. Next, the support layer is etched using the resist as an etching mask. In order to form the concave portion 306 whose side surface is inclined, the support layer made of silicon is wet-etched using an aqueous solution of TMAH or KOH. Further, in order to form the concave portion 306 having a vertical side surface, the support layer is etched by dry etching by DRIE. In any case, the recess 306 can be formed using a known etching technique. After the silicon etching, the resist is removed and washed.

  Next, the shield electrode is set so that the sizes of the opening 101 of the mirror substrate 100 in the x-axis direction and the y-axis direction are substantially the same as the sizes of the shield electrode 302 of the electric field shield substrate 300a. A groove for forming the overhanging portion 301 is formed around the support layer while leaving the support layer 302. In order to form this groove, a resist is applied again on the surface of the support layer, and a resist pattern is formed using a mask in which a mouth-shaped pattern to be formed is arranged. Thus, a resist pattern in which a resist was not applied only on the surface of the support layer forming the groove could be formed. Using the formed resist pattern as an etching mask, the support layer is etched using dry etching by DRIE. Here, etching is performed until the BOX layer 303 is exposed. The BOX layer 303 also serves as an etching stopper. By this etching, a shield electrode 302 having a rectangular shape in a plan view and an overhang portion 301 including a BOX layer 303 and an SOI layer 304 extending outside the shield electrode 302 could be formed. Note that, as is apparent from inserting the shield electrode 302 into the opening 101 of the mirror substrate 100, the horizontal size of the shield electrode 302 is actually smaller than the horizontal size of the opening 101. It is preferable.

  Next, in a state after the etching is completed, dicing is performed to the size of the electric field shield substrate 300a to be used. If the thickness of the SOI layer 304 is sufficient in strength, dicing can be performed without any problem. When the thickness of the SOI layer 304 is thin, there is a possibility that the SOI layer 304 portion is damaged by a water flow used during dicing. In that case, a chip for reinforcing rigidity is affixed separately, or a dicing method that does not use a water flow such as stealth dicing is used for dicing. The electric field shield substrate 300a is formed by the above method.

  Next, after joining the mirror substrate 100 and the electrode substrate 200 as described above, the electric field shield substrate 300 a is inserted into the opening 101 of the mirror substrate 100. Since the thickness of the support layer 102 of the mirror substrate 100 and the thickness of the support layer on which the shield electrode 302 of the electric field shield substrate 300a is formed are initially set to the same thickness, the electric field shield substrate 300a is inserted into the opening 101 of the mirror substrate 100. However, the shield electrode 302 does not collide with a mirror structure such as the leaf springs 105a and 105b.

  In order to join the electric field shield substrate 300a and the mirror substrate 100, solder deposited on the surface of the BOX layer 303 of the protruding portion 301 of the electric field shield substrate 300a and solder deposited on the surface of the support layer 102 of the mirror substrate 100 are used. Bonding may be performed by using, or gold deposited on the surface of the BOX layer 303 and gold deposited on the surface of the support layer 102 may be combined to directly bond Au—Au. Moreover, you may join using Ag paste. In any case, a known joining technique can be used. If the mirror substrate 100 to which the electric field shield substrate 300a is bonded is die-bonded to a package, a WSS chip is completed. If necessary, the lid may be bonded to the package and hermetically sealed.

  As described above, in the present embodiment, the shield electrode 302 is formed on the mirror substrate 100 by making the horizontal size of the electric field shield substrate 300 a larger than the horizontal size of the opening 101 of the mirror substrate 100. When inserted into the opening 101, the BOX layer 303 of the electric field shield substrate 300a comes into contact with the support layer 102 of the mirror substrate 100, and the position of the electric field shield substrate 300a in the thickness direction is fixed. At this time, since the thickness of the shield electrode 302 is the same as the thickness of the support layer 102 of the mirror substrate 100, the shield electrode 302 and the leaf springs 105a and 105b do not contact each other. As a result, in this embodiment, the electric field shield substrate 300a having a shape and structure capable of realizing the effect as a substrate for preventing electrical interference can be realized, and the electric field shield substrate 300a can be bonded to the mirror substrate 100. The shield electrode 302 and the leaf springs 105a and 105b can be prevented from colliding with each other and the leaf springs 105a and 105b can be prevented from being damaged, and the yield can be improved.

  In this embodiment, separate substrates are used for the mirror substrate 100 and the electric field shield substrate 300a. However, the mirror substrate 100 and the electric field shield substrate 300a may be formed on the same substrate and cut out simultaneously. In this case, it goes without saying that the thicknesses of the support layer 102 and the shield electrode 302 are in good agreement.

  Further, in this embodiment, the BOX layer 303 around the shield electrode 302 of the electric field shield substrate 300a is left, but the shield electrode 302 is made in order to make the electric potential of the electric field shield substrate 300a the same as the electric potential of the mirror substrate 100. The exposed BOX layer 303 around the substrate may be removed by etching. When the BOX layer 303 is not removed by etching as described above, it is important to connect the electric field shield substrate 300a and the mirror substrate 100 after mounting so that they have the same potential (for example, ground potential).

  The mirror substrate 100 and the electric field shield substrate are arranged so that the shield electrode 302 and the leaf springs 105a and 105b do not come into contact with each other when selecting either the process of removing the BOX layer 303 of the electric field shield substrate 300a by etching or the process of not removing the BOX layer 303. It is important to set the thickness of each BOX layer 103, 303 of 300a. In addition, the process can be selected so that the shield electrode 302 and the leaf springs 105a and 105b do not come into contact with each other based on the thickness of the BOX layer.

  In the present embodiment, when the BOX layer 303 is removed by etching with the BOX layers 103 and 303 having the same thickness, the shield electrode 302 and the BOX layer 303 entering the opening 101 of the mirror substrate 100 are added. Since the thickness is the same as the depth of the opening 101 (the thickness obtained by adding the support layer 102 and the BOX layer 103 of the mirror substrate 100), the shield electrode 302 and the leaf springs 105a and 105b may contact each other. In reality, however, there is a bonding material (solder, Au, Ag, etc.) between the SOI layer 304 of the electric field shield substrate 300a and the support layer 102 of the mirror substrate 100, and the thickness of this bonding material is the same. Since only a gap is generated between the shield electrode 302 and the leaf springs 105a and 105b, the shield electrode 302 and the leaf springs 105a and 105b do not contact each other. Further, since the concave portion 306 is formed on the surface of the shield electrode 302 facing the leaf springs 105a and 105b, the possibility of contact between the shield electrode 302 and the leaf springs 105a and 105b can be further reduced.

In this embodiment, an SOI substrate having a good film thickness accuracy of each layer is used as a material for the mirror substrate 100 and the electric field shield substrate 300a. However, a multilayer substrate using other materials may be used.
In this embodiment, the shield electrode 302 is formed using the silicon support layer of the SOI substrate. However, the present invention is not limited to this, and the shield electrode is made of a conductive material separately from the electric field shield substrate. The shield electrode may be bonded to the electric field shield substrate. Also in this case, the thickness of the shield electrode may be set so as to be the same as the depth of the opening 101 of the mirror substrate 100 or thinner than the depth of the opening 101.

  The present invention can be applied to electrostatically driven MEMS elements such as wavelength selective switches.

  DESCRIPTION OF SYMBOLS 100 ... Mirror board | substrate, 101,305 ... Opening part, 102 ... Support layer, 103, 303 ... BOX layer, 104, 304 ... SOI layer, 105a, 105b ... Leaf spring, 106 ... Mirror, 107a, 107b ... Connection part, 200 ... Electrode substrate, 201a, 201b ... Leaf spring drive electrode, 202a, 202b ... Mirror drive electrode, 300a ... Electric field shield substrate, 301 ... Overhang, 302 ... Shield electrode, 306 ... Recess.

Claims (6)

A first substrate;
A second substrate disposed opposite the first substrate;
A third substrate bonded on a surface of the first substrate that does not face the second substrate;
A movable electrode formed to be displaceable on the first substrate;
An opening from the surface of the first substrate not facing the second substrate to the movable electrode;
A fixed electrode for driving the movable electrode, disposed on the surface of the second substrate facing the first substrate, in a state of providing a space between the movable electrode and the movable electrode;
A shield electrode formed on a surface of the third substrate facing the first substrate, and fitted into an opening of the first substrate;
The horizontal size of the third substrate is larger than the horizontal size of the opening of the first substrate,
The MEMS element according to claim 1, wherein a thickness of the shield electrode is the same as a depth of the opening of the first substrate or thinner than a depth of the opening of the first substrate. The MEMS device according to claim 1, wherein
The MEMS element according to claim 1, wherein the shield electrode has a concave portion formed at a position facing the movable electrode. A movable electrode forming step of forming a movable electrode on the SOI layer of the first SOI substrate;
A portion of the silicon support layer and the buried oxide film layer of the first SOI substrate, wherein the movable electrode exposes the silicon support layer and the buried oxide film layer existing in the region where the movable electrode is formed. An opening forming step of forming an opening in the first SOI substrate by etching away until
A first dicing step of processing the first SOI substrate into a predetermined size;
An electrode substrate forming step of forming a fixed electrode on the electrode substrate;
A first substrate bonding step of arranging and bonding the first SOI substrate and the electrode substrate so that the movable electrode and the fixed electrode are spaced apart from each other;
The silicon support layer of the second SOI substrate having the same thickness as the first SOI substrate and the silicon support layer has a shield electrode region smaller than the horizontal size of the opening of the first SOI substrate. Forming a shield electrode to be etched away,
A second dicing step of processing the second SOI substrate into a predetermined size;
A second substrate bonding step of bonding the first SOI substrate and the second SOI substrate by inserting the shield electrode of the second SOI substrate into the opening of the first SOI substrate; ,
A method for manufacturing a MEMS element, wherein a size in a horizontal direction after processing of the second SOI substrate is larger than a size in a horizontal direction of an opening of the first SOI substrate. In the manufacturing method of the MEMS element of Claim 3,
And a buried oxide film layer removing step of etching and removing a buried oxide film layer around the shield electrode of the second SOI substrate before the second dicing step. . A movable electrode forming step of forming a movable electrode on the SOI layer of the SOI substrate;
Etching the silicon support layer and the buried oxide film layer, which are part of the silicon support layer and the buried oxide film layer of the SOI substrate and exist in the region where the movable electrode is formed, until the movable electrode is exposed. An opening forming step of removing and forming an opening in the SOI substrate;
A first dicing step of processing the SOI substrate into a predetermined size;
An electrode substrate forming step of forming a fixed electrode on the electrode substrate;
A first substrate bonding step in which the SOI substrate and the electrode substrate are arranged and bonded so that the movable electrode and the fixed electrode are spaced apart from each other;
A shield electrode joining step for joining a shield electrode made of a conductive material smaller than the horizontal size of the opening of the SOI substrate to the electric field shield substrate;
A second dicing step of processing the electric field shield substrate into a predetermined size;
A second substrate bonding step of inserting the shield electrode into an opening of the SOI substrate and bonding the SOI substrate and the electric field shield substrate;
The horizontal size of the electric field shield substrate after processing is larger than the horizontal size of the opening of the SOI substrate,
The method of manufacturing a MEMS device, wherein the thickness of the shield electrode is the same as the depth of the opening of the SOI substrate or thinner than the depth of the opening of the SOI substrate. In the manufacturing method of the MEMS element of any one of Claims 3 thru | or 5,
Furthermore, the manufacturing method of the MEMS element characterized by including the recessed part formation process of forming a recessed part in the position facing the said movable electrode of the said shield electrode before the said 2nd dicing process.

Images