JP2008155314A - Method for manufacturing MEMS device - Google Patents

Abstract

A MEMS device manufacturing method that does not require a temporary bridge is provided.
An oxide film forming step in which a silicon wafer is heated to form an oxide film on both surfaces of the silicon wafer, and a first oxidation in which the oxide film formed on one surface of the silicon wafer is removed. A film removing step, a molding step of forming a plurality of upper layer substrates 2 on a silicon wafer from which an oxide film on one surface has been removed, a silicon wafer having a plurality of upper layer substrates formed thereon, and an insulating wafer having a plurality of lower layer substrates formed thereon And a second oxide film removing step of removing an oxide film formed on the other surface of the silicon wafer of the laminate, and a stacking step of forming a plurality of laminates in which an upper layer substrate and a lower layer substrate are laminated, And a dicing step of separating a plurality of stacked bodies formed in the stacking step to form a plurality of MEMS devices.
[Selection] Figure 2

Description

  The present invention relates to a method for manufacturing a MEMS device capable of simultaneously manufacturing a plurality of MEMS devices.

  2. Description of the Related Art In recent years, MEMS devices manufactured by MEMS (Micro Electro Mechanical Systems) technology have been widely used in various apparatuses constituting image forming apparatuses such as printers, copiers, and projectors using laser light. The MEMS technology is a technology for realizing miniaturization of various machine elements by applying technology in a semiconductor manufacturing process such as silicon. As a MEMS device used in the image forming apparatus, a MEMS mirror scanner disclosed in Patent Document 1 can be cited. Such a MEMS mirror scanner includes a swingable mirror, and can be used as a laser light scanning device that scans printing paper, a screen, or the like by reflecting laser light emitted from a light source with the mirror.

  This MEMS mirror scanner includes an upper layer substrate and a lower layer substrate. The upper substrate includes a movable part and a fixed part. The movable portion includes the aforementioned mirror and a movable electrode formed integrally with the mirror, and the fixed portion includes a fixed electrode on which an electrostatic force acts between the movable electrode and the movable portion. The movable part and the fixed part are made of a single silicon wafer, but are physically separated from each other in order to keep the movable electrode and the fixed electrode in an insulated state. The movable portion is supported by a fixing pad of the lower layer substrate, and the fixed portion is supported by the peripheral wall portion of the lower layer substrate. The lower layer substrate that supports the movable portion and the fixed portion is formed of an insulating material such as glass, and is therefore supported by the movable portion of the upper layer substrate supported by the fixing pad of the lower layer substrate and the peripheral wall portion of the lower layer substrate. The upper layer substrate is supported in an insulated state from the movable portion. When an AC voltage is applied between such a movable electrode and a fixed electrode, an electrostatic force that varies according to the cycle of the AC voltage is generated between the movable electrode and the fixed electrode. The mirror of the part swings.

In this MEMS mirror scanner, a plurality of upper layer substrates are fabricated on a silicon wafer, a plurality of lower layer substrates are fabricated on an insulating wafer formed of an insulating material, and then the wafers are stacked to form a plurality of upper layer substrates and a plurality of lower layer substrates. Are manufactured at the same time. The movable portion and the fixed portion of the upper substrate are physically separated as described above, but are connected by a temporary bridge until the upper substrate is stacked on the lower substrate. The temporary bridge is a portion of the silicon wafer that is the material of the upper layer substrate that is left without removing a portion between the movable portion and the fixed portion in order to connect the movable portion and the fixed portion. Such a temporary bridge is cut by laser processing or the like after a silicon wafer and an insulating wafer are stacked and an upper layer substrate and a lower layer substrate are stacked.
JP 2006-201520 A

  As described above, a plurality of MEMS mirror scanners are manufactured simultaneously by laminating a plurality of upper layer substrates and a plurality of lower layer substrates. Therefore, it is necessary to cut the temporary bridges of the plurality of MEMS mirror scanners when manufacturing the MEMS mirror scanner. For this reason, in the above-described manufacturing method, there are many temporary bridges to be cut, and the time and labor required for cutting the temporary bridge are large. Therefore, there is a problem that the time and labor required for manufacturing the entire MEMS mirror scanner are also increased. Further, when the temporary bridge is cut, there is a problem that the fragments of the temporary bridge adhere to the mirror or the like.

  The present invention has been made to solve the problems of the prior art, and an object of the present invention is to provide a MEMS device manufacturing method that does not require a temporary bridge.

  According to a first aspect of the present invention, there is provided a MEMS device comprising: an upper layer substrate having a movable portion and a fixed portion separated from the movable portion; and a lower layer substrate laminated on the upper layer substrate. In the manufacturing method, the silicon wafer is heated to form an oxide film on both sides of the silicon wafer, and one side of the silicon wafer among the oxide films formed on both sides of the silicon wafer A first oxide film removing step for removing the oxide film formed on the substrate, a molding step for forming the plurality of upper layer substrates on the silicon wafer from which the oxide film on one surface is removed, and a plurality of the upper layer substrates are formed. A lamination step of laminating the silicon wafer and an insulating wafer on which a plurality of the lower layer substrates are formed to form a plurality of laminated bodies in which the upper layer substrate and the lower layer substrate are laminated; A second oxide film removing step for removing the oxide film formed on the other surface of the silicon wafer, and a dicing step for separating the plurality of stacked bodies formed in the stacking step to form a plurality of MEMS devices. A method for manufacturing a MEMS device is provided.

  In the MEMS device manufacturing method according to the present invention, first, in the oxide film forming step, the silicon wafer is heated to form oxide films on both surfaces of the silicon wafer. Thus, by heating the silicon wafer, a thick oxide film is formed on both sides of the silicon wafer as compared with the natural oxide film. Next, in the first oxide film removing step, the oxide film formed on one surface of the silicon wafer is removed from the oxide films formed on both surfaces of the silicon wafer. As a result, the oxide film on one surface is removed, and the silicon wafer can be processed from the surface from which the oxide film has been removed using the MEMS technology or the like. Next, in the forming step, a plurality of upper layer substrates are formed on the silicon wafer from which the oxide film on one surface has been removed. At this stage, the oxide film formed on the other surface of the silicon layer is not removed. Therefore, even when an upper substrate having a movable part and a fixed part separated from the movable part is formed on the silicon wafer, the movable part and the fixed part are fixed to the oxide film formed on the other surface, The movable part and the fixed part are connected via the membrane. That is, this oxide film performs the same function as the temporary bridge. Since this oxide film is thicker than the natural oxide film, it is difficult to tear, and therefore there is little risk of the connection between the movable part and the fixed part being released. Next, in the stacking step, a silicon wafer in which a plurality of upper layer substrates are formed and an insulating wafer in which a plurality of lower layer substrates are formed are stacked to form a plurality of stacked bodies in which the upper layer substrate and the lower layer substrate are stacked. Next, in the second oxide film removing step, the oxide film formed on the other surface of the silicon wafer of the stacked body is removed. When the oxide film formed on the other surface of the silicon wafer is removed, the movable part and the fixed part of the upper substrate connected by this oxide film are separated. Even if the movable part and the fixed part are separated, since the upper layer substrate is already laminated on the lower layer substrate, the movable part and the fixed part are connected via the lower layer substrate and will not be separated. . Next, in the dicing process, the plurality of stacked bodies formed in the stacking process are separated to form a plurality of MEMS devices. Thereby, a plurality of MEMS devices are manufactured. As described above, according to the MEMS device of the present invention, the movable portion and the fixed portion of the upper layer substrate are formed by the oxide film formed on the other surface of the silicon wafer until the upper layer substrate is stacked on the lower layer substrate. Are connected to each other, so a temporary bridge is unnecessary.

  Preferably, the second oxide film removing step is performed by etching an oxide film formed on the other surface of the silicon wafer. Features.

  In such a preferable manufacturing method, the oxide film formed on the other surface of the silicon wafer connecting the movable part and the fixed part of the plurality of upper layer substrates is removed by etching, so that the movable part and the fixed parts of the plurality of upper layer substrates are fixed. The connection with the part can be released simultaneously in one step. Therefore, since it does not take time to remove the oxide film that performs the same function as the temporary bridge, according to such a preferable manufacturing method, it does not take time to manufacture the MEMS device.

  According to the present invention, a MEMS device can be manufactured without the need for a temporary bridge. Therefore, a MEMS device can be manufactured without taking time and effort, and fragments of the cut temporary bridge adhere to the MEMS device. Such a problem can be avoided.

  Hereinafter, a method for manufacturing a MEMS device according to an embodiment of the present invention will be described with reference to the accompanying drawings. FIG. 1 is an exploded perspective view showing a schematic configuration of a MEMS device 1 manufactured by a method for manufacturing a MEMS device according to an embodiment of the present invention. Such a MEMS device 1 can be used as a laser beam scanning apparatus.

  As shown in FIG. 1, the MEMS device 1 includes an upper substrate 2 formed by processing one silicon wafer and a lower substrate 3 formed by processing an insulating wafer (for example, a glass wafer) vertically. It has a laminated structure (FIG. 1 shows a state in which the upper layer substrate 2 and the lower layer substrate 3 are separated, but both are actually laminated).

  The upper layer substrate 2 includes a movable part 2A and a fixed part 2B separated from the movable part 2A.

  The movable portion 2A includes a swing member 11, a connection portion 12, suspension beams 13A and 13B, adhesive pads 14A to 14D, hinges 15A to 15D, swing comb teeth 18A and 18B, and an electrode pad 20. An elliptical oscillating member 11 is formed at the center of the oscillating axis (X axis shown in FIG. 1) of the movable portion 2A. Although not shown in FIG. 1, an aluminum film for reflecting laser light is formed on the surface of the swing member 11. Suspension beams 13A and 13B are formed at both ends of the swing member 11 in the swing axis direction via connection portions 12, respectively. The swing member 11 is supported by the suspension beams 13A and 13B so as to be swingable around the swing axis. The ends of the suspension beams 13A and 13B in the swing axis direction (ends opposite to the side connected to the swing member 11) are respectively attached to the adhesive pads 14A and 14B serving as anchors via hinges 15A and 15D. It is connected. The suspension beam 13A is provided with an adhesive pad 14C and a hinge 15B, and the hinge 15B is connected to the adhesive pad 14C. Similarly, an adhesive pad 14D and a hinge 15C are formed on the suspension beam 13B, and the hinge 15C is connected to the adhesive pad 14D. Further, swing comb teeth 18A and 18B extending in the Y-axis direction are formed as electrodes at ends of the suspension beams 13A and 13B in the axis direction (Y-axis shown in FIG. 1) orthogonal to the swing axis. The electrode pad 20 is a portion of the movable portion 2A other than the swing member 11, the connection portion 12, the suspension beams 13A and 13B, the adhesive pads 14A to 14D, the hinges 15A to 15D, and the swing comb teeth 18A and 18B. Is formed. Since the material of the movable portion 2A is silicon and has conductivity, the swinging comb teeth 18A and 18B and the electrode pad 20 are electrically connected.

  The fixed portion 2B includes fixed comb teeth 19A and 19B that are formed as electrodes on the outside of the swinging comb teeth 18A and 18B and extend in the Y-axis direction, and an electrode pad 21. The fixed comb teeth 19A and 19B are alternately arranged along the swing axis direction with the swing comb teeth 18A and 18B. The electrode pad 20 is formed in a portion other than the fixed comb teeth 19A and 19B in the fixed portion 2B. Further, since the material of the fixing portion 2B is silicon and has conductivity, the fixing comb teeth 19A and 19B and the electrode pad 21 are electrically connected. A space is formed between the fixed portion 2B and the movable portion 2A, and the movable portion 2A and the fixed portion 2B are separated by this space. Since the movable portion 2A and the fixed portion 2B are separated, the movable portion 2A and the fixed portion 2B are kept in an insulated state.

  Next, the lower substrate 3 is substantially positioned at a position corresponding to the swing member 11 so that the swing member 11 and the suspension beams 13A and 13B provided on the upper substrate 2 can swing about the swing axis. An elliptical digging region 16 is formed, and substantially rectangular digging regions 16A and 16B are formed at positions corresponding to the suspension beams 13A and 13B. Further, fixing pads 17A and 17B are formed on the lower substrate 3 so as to fix the back surfaces of the adhesive pads 14C and 14D provided on the upper substrate 2 in a state where the upper substrate 2 and the lower substrate 3 are laminated. Has been. Further, in a state where the upper layer substrate 2 and the lower layer substrate 3 are laminated, the back surfaces of the adhesive pads 14A and 14B and the back surface of the fixing portion 2B are fixed to the upper surface of the peripheral wall portion 17C of the lower layer substrate 3.

  In the MEMS device 1 described above, the electrode pad 20 provided in the movable part 2A of the upper substrate 2 and the electrode pad 21 provided in the fixed part 2B are connected to an AC power source, and the electrode pads 20 and 21 are connected to each other. If an alternating voltage is applied between the swinging comb teeth 18A and 18B and the fixed comb teeth 19A and 19B, an electrostatic force acts between the swinging comb teeth 18A and 18B and the fixed comb teeth 19A and 19B. Due to this electrostatic force, the suspension beams 13A and 13B have the bonding pads 14A to 14D as fixed ends and swing around the swing axis while resisting the elastic force of the hinges 15A to 15D, and the suspension beams 13A and 13B swing. By moving, the swing member 11 also swings around the swing axis.

  In the MEMS device 1 according to this embodiment, the swinging comb teeth 18A and 18B and the fixed comb teeth 19A and 19B are alternately arranged along the swing axis direction, and the swinging comb teeth 18A and 18B and the fixed comb teeth are arranged. The positions of 19A and 19B are shifted in the swing axis direction. Therefore, even if the suspension beams 13A and 13B swing around the swing axis, the swing comb teeth 18A and 18B and the fixed comb teeth 19A and 19B do not come into contact with each other, so that the swing comb teeth 18A and 18B are fixed. The comb teeth 19A and 19B can be arranged close to each other. The swinging comb teeth 18A and 18B and the fixed comb teeth 19A and 19B can be disposed close to each other through a space of 5 μm to 30 μm, for example.

  When the oscillating member 11 that oscillates in this manner is irradiated with laser light, the reflection direction of the laser light varies due to the oscillation of the oscillating member 11. By changing the reflection direction of the laser light, it is possible to scan the printing paper or the screen with the reflected light.

  Next, a method for manufacturing the MEMS device 1 having the above-described configuration will be described. 2 and 3 are manufacturing flowcharts showing the manufacturing procedure of the MEMS device 1. In FIG. 3, the thickness of the silicon wafer is drawn thicker than that in FIG. 2 for ease of explanation.

  As shown in FIG. 2A, first, in the oxide film forming step, the silicon wafer 4 is heated to form oxide films 41 on both surfaces of the silicon wafer 4. The oxide film 41 can be formed, for example, by heating the silicon wafer 4 at 900 to 1100 ° C., for example, for 5 to 24 hours in the air or oxygen. As described above, when the silicon wafer 4 is heated, the oxide film 41 (1 μm to 5 μm) thicker than the natural oxide film (generally 0.001 μm to 0.010 μm) may be formed on both surfaces of the silicon wafer 4. it can. The thickness of the oxide film 41 is not particularly limited, and can be, for example, about 1.0 μm to 1.5 μm. Further, the thickness of the silicon wafer 4 is not particularly limited, and can be, for example, about 100 μm to 300 μm. As an example of the oxide film formation step, when the silicon wafer 4 is heated at 1100 ° C. for 5 hours in the air or oxygen, an oxide film 41 having a thickness of 1.5 μm is formed.

  Next, in the first oxide film removing step, the oxide film 41 formed on one surface of the silicon wafer 4 is removed from the oxide film 41 formed on both surfaces of the silicon wafer 4. The removal of the oxide film 41 is performed by applying an acid resistant resist 42 to the oxide film 41 formed on the other surface of the silicon wafer 4 as shown in FIG. Then, the etching can be performed by etching the silicon wafer 4 in which the resist 42 is applied to the oxide film 41. For the resist 42 used in this step, for example, a positive resist novolac resin can be used. This etching can be performed by immersing the silicon wafer 4 having the resist 42 applied to the oxide film 41 in an aqueous solution in which hydrofluoric acid and ammonium fluoride are mixed for 10 to 20 minutes. As shown in FIG. 2C, by removing the oxide film 41 on one surface, the silicon wafer 4 can be processed from the surface from which the oxide film 41 has been removed using MEMS technology or the like. . At the end of the first oxide film removal step, as shown in FIG. 2D, the silicon wafer 4 is immersed in a stripping solution to remove the resist 42. For the stripping solution, an organic solvent such as acetone or NMP (n-methylpyrrolidone) can be used.

  Next, in the molding step, a plurality of upper layer substrates 2 are formed on the silicon wafer 4 from which the oxide film 41 on one surface has been removed. First, in this molding step, as shown in FIG. 2E, a photoresist 43 is applied to one surface of the silicon wafer 4 from which the oxide film 41 has been removed, using spin coating or the like. For this photoresist 43, for example, a positive resist novolac resin can be used.

  Next, the photoresist 43 is exposed with a contact exposure machine or a stepper through a mask in which a plurality of patterns (for example, 20 to 500) of the upper substrate 2 are formed, and the photoresist 43 is developed with a developer. A plurality of patterns 2P of the upper substrate 2 as shown in FIG. As this developer, a developer using a strong alkaline aqueous solution such as potassium hydroxide or sodium hydroxide as a raw material can be used. Next, as shown in FIG. 2G, the heat-resistant glass 45 is bonded to the oxide film 41 formed on the other surface of the silicon wafer 4 through the temporary fixing agent 44. For the temporary fixing agent 44, for example, a paraffin-based or glycol-based thermoplastic adhesive can be used. In this way, the heat-resistant glass 45 is bonded to the oxide film 41 formed on the other surface of the silicon wafer 4 because the silicon wafer 4, the oxide film 41, and the photoresist 43 are applied to an ICP plasma etching apparatus used in the next process. This is because it is easy to move the composite 40 made of and to take out the composite 40 from the ICP plasma etching apparatus after etching with the ICP plasma etching apparatus.

  Next, as shown in FIG. 2H, the silicon wafer 4 is etched to form a plurality of upper layer substrates 2 on the silicon wafer 4. This etching can be performed using an ICP (Inductively Coupled Plasma) plasma etching apparatus. Specifically, this etching can be performed using, for example, Deep RIE (Reactive Ion Etching) processing. As an example of such DeepRIE processing, an etching process is performed to dig down the silicon wafer 4 in SF6 (sulfur hexafluoride) gas, and a side wall portion of the silicon wafer 4 formed by digging is used with C4F8 (octafluorocyclobutane) gas. An example is a Bosch method in which the step of forming the protective film is repeated every few seconds to several tens of seconds. The etching rate in this Bosch method can be set to 2 to 10 um / min. This etching does not proceed on the oxide film 41 formed on the other surface of the silicon wafer 4 but is performed only on the silicon wafer 4. As described above, since the upper layer substrate 2 includes the movable portion 2A and the fixed portion 2B separated from the movable portion 2A, the silicon wafer 4 is a portion corresponding to the movable portion 2A of the upper layer substrate 2 by this etching. And a portion corresponding to the fixed portion 2B. However, the oxide film 41 formed on the other surface of the silicon wafer 4 is not etched and remains in a single piece, so that the oxide film 41 is fixed to a portion corresponding to the movable portion 2A of the upper substrate 2 of the silicon wafer 4. The portion corresponding to the portion 2B is fixed to the oxide film 41 and connected via the oxide film 41.

  At the end of the molding step, as shown in FIG. 2I, the photoresist 43 is removed, and the temporary fixing agent 44 and the heat-resistant glass 45 are removed from the oxide film 41. The removal of the photoresist 43 can be performed with acetone or NMP.

Next, as shown in FIG. 3 (j), in the stacking step, a composite comprising a silicon wafer 4 on which a plurality of upper layer substrates 2 are formed and an oxide film 41 formed on the other surface of the silicon wafer 4. 40 is laminated on the insulating wafer 5 with the silicon wafer 4 facing the insulating wafer 5 side, and a plurality of laminated bodies 7 in which the upper substrate 2 and the lower substrate 3 are laminated are formed. On this insulating wafer 5, a plurality of lower layer substrates 3 (the same number as the upper layer substrate 2 formed on the silicon wafer 4) are formed. The formation of the lower layer substrate 3 on the insulating wafer 5 is not limited as long as it is before the lamination step. The insulating wafer 5 is made of an insulating material, and in this embodiment is made of glass. Lamination of the silicon wafer 4 and the insulating wafer 5 can be performed by bonding the silicon wafer 4 and the insulating wafer 5 by anodic bonding. In this anodic bonding method, for example, a heat-resistant glass 45 is overlaid on an oxide film 41 formed on the other surface of the silicon wafer 4, and the overlaid silicon wafer 4 and the heat-resistant glass 45 are placed in a temperature atmosphere of 200 to 500 ° C. It can be performed by applying a voltage of 500 to 1500 V to the silicon wafer 4 and the heat-resistant glass 45 until no current flows. It is desirable that the heat resistant glass 45 has a small difference in thermal expansion coefficient from that of the silicon wafer 4. The thermal expansion coefficient of silicon at room temperature is 26 × 10 ⁻⁷ / ° C., and the thermal expansion coefficient of silicon at 300 ° C .: 38 × 10 ⁻⁷ / ° C. Examples of the glass that can be used for the heat-resistant glass 45 include Corning Pyrex (registered trademark) (borosilicate glass / # 7740, etc.) and Schott Tempax (registered trademark) (borosilicate glass / # 8330, etc.). be able to. The thermal expansion coefficients of Pyrex (registered trademark) and Tempax (registered trademark) are (32 to 32.5) × 10 ⁻⁷ / ° C. Note that a resin material such as PEEK (polyetheretherketone) or a ceramic material such as alumina can be used as an insulating material as a material of the insulating wafer 5. When the insulating wafer 5 formed of an insulating material such as a resin material or a ceramic material is used, the silicon wafer 4 and the insulating wafer 5 can be laminated using a thermosetting adhesive such as epoxy. When the composite 40 composed of the silicon wafer 4 and the oxide film 41 is laminated on the insulating wafer 5, each upper layer substrate 2 formed on the silicon wafer 4 is replaced with each lower layer substrate 3 formed on the insulating wafer 5. The composite 40 composed of the silicon wafer 4 and the oxide film 41 and the insulating wafer 5 are aligned so as to be stacked on the upper surface. Specifically, as shown in FIG. 1, the back surfaces of the adhesive pads 14C and 14D of the upper substrate 2 are on the upper surfaces of the fixing pads 17A and 17B of the lower substrate 3, and the back surfaces of the adhesive pads 14A and 14B of the upper substrate 2 In addition, the composite 40 composed of the silicon wafer 4 and the oxide film 41 and the insulating wafer 5 are aligned so that the back surface of the fixed portion 2B is laminated on the upper surface of the peripheral wall portion 17C of the lower substrate 3.

  As described above, since the portion corresponding to the movable portion 2A and the portion corresponding to the fixed portion 2B of the upper substrate 2 of the silicon wafer 4 are fixed to the oxide film 41, the silicon wafer 4 and the oxide film 41 are formed. When stacking the composite 40 on the insulating wafer 5, the portion corresponding to the movable portion 2A of the upper substrate 2 of the silicon wafer 4 swings relative to the portion corresponding to the fixed portion 2B, or corresponds to the movable portion 2A. There is no fear that the portion and the portion corresponding to the fixed portion 2B are separated. Further, as described above, the oxide film 41 is thicker than the natural oxide film and has a relatively high strength. Therefore, when the composite 40 composed of the silicon wafer 4 and the oxide film 41 is stacked on the insulating wafer 5. In addition, there is little risk of damage. The possibility that the oxide film 41 is damaged is smaller as the space formed between the movable portion 2A and the fixed portion 2B is narrower. As described above, the MEMS device 1 according to the present embodiment is fixed to the swinging comb teeth 18A and 18B because the swinging comb teeth 18A and 18B and the fixed comb teeth 19A and 19B are displaced in the swing axis direction. The comb teeth 19A and 19B are arranged close to each other. Therefore, in the MEMS device 1 of the present embodiment, the space formed between the movable portion 2A and the fixed portion 2B is narrow, and there is almost no possibility that the oxide film 41 is damaged.

  Next, as shown in FIG. 3K, the oxide film 41 formed on the other surface of the silicon wafer 4 is removed in the second oxide film removing step. This second oxide film removing step can be performed by etching the oxide film 41 formed on the other surface of the silicon wafer 4. This etching can be performed by, for example, wet etching using a hydrogen fluoride aqueous solution or the like. However, in wet etching, a phenomenon called sticking that is dry and adheres while the electrodes are in contact with each other during drying tends to occur. As an etching method in which sticking is unlikely to occur, dry etching performed by flowing a hydrofluoric acid gas or the like in a corrosive gas atmosphere of about 30 to 80 ° C. for 10 to 20 minutes can be given. For example, methanol can be used as the corrosive gas. If the oxide film 41 is removed by etching, the oxide film 41 connecting the movable portion 2A and the fixed portion 2B of the plurality of upper layer substrates 2 can be simultaneously removed in one step. Therefore, since it does not take time to remove the oxide film 41 that performs the same function as the plurality of temporary bridges connecting the plurality of movable parts 2A and the fixed part 2B, the labor for manufacturing the MEMS device 1 can be reduced. Can do. As described above, when the oxide film 41 formed on the other surface of the silicon wafer 4 is removed, the movable portion 2A and the fixed portion 2B of the upper substrate 2 connected by the oxide film 41 are separated. Even if the movable portion 2A and the fixed portion 2B are separated, since the upper layer substrate 2 is already laminated on the lower layer substrate 3, the movable portion 2A and the fixed portion 2B are connected via the lower layer substrate 3. And will not separate.

  Next, as shown in FIG. 3L, in the vapor deposition step, an aluminum film 11a is formed by aluminum vapor deposition on a portion corresponding to the swing member 11 of the upper substrate 2 of the silicon wafer 4. Although not shown, in this vapor deposition step, electrode pads 20 and 21 are formed on the portions corresponding to the movable portion 2A and the fixed portion 2B of the silicon wafer 4 by aluminum vapor deposition.

  Next, as shown in FIG. 3 (m), in the dicing process, a dicing tape 6 is attached to the lower surface of the insulating wafer 5, and a plurality of laminated bodies having the upper layer substrate 2 and the lower layer substrate 3 formed in the laminating step. 7 is cut off. Then, by removing the dicing tape 6 from each lower layer substrate 3, a plurality of MEMS devices 1 shown in FIG.

  As described above, according to the MEMS device manufacturing method according to the present embodiment, the oxide film 41 formed on the other surface of the silicon wafer 4 is used until the upper substrate 2 is stacked on the lower substrate 3. Since the movable part 2A and the fixed part 2B of the upper substrate 2 are connected, a temporary bridge is unnecessary. Therefore, the MEMS device 1 can be manufactured without taking time and effort, and the problem that the fragments of the temporary bridge adhere to the MEMS device 1 can be avoided.

Using the following examples and comparative examples, it will be described that the manufacturing time of the MEMS device 1 is shortened by using the manufacturing method of the MEMS device 1 described in the above embodiment. Table 1 is an operation table showing the operation time in each manufacturing process in Examples and Comparative Examples.

In the examples and comparative examples, 50 upper substrates are formed on a silicon wafer, and 50 lower substrates are formed on an insulating wafer, and the silicon wafer and the insulating wafer are stacked to form 50 MEMS devices at a time. Manufacture. The work time shown in Table 1 is the work time required in each process for manufacturing MEMS devices for one wafer (50 wafers).

(Production method in Examples)
First, in a receiving inspection process, a receiving inspection of a silicon wafer is performed. Next, in the oxide film forming step, the silicon wafer is heated using a high temperature furnace to form oxide films on both sides of the silicon wafer.

  Next, in the first oxide film removing step, the oxide film formed on one surface of the silicon wafer is removed from the oxide films formed on both surfaces of the silicon wafer. In this first oxide film removal step, first, a resist is applied to the oxide film formed on the other surface of the silicon wafer using a spin coater, and the silicon wafer having the resist applied to the oxide film is baked in an oven. The organic solvent contained in the resist is evaporated. Next, the baked silicon wafer is etched using an acid draft, and the oxide film formed on one surface is removed. Further, using an organic draft, the silicon wafer from which the oxide film on one surface has been removed is immersed in a stripping solution to remove the resist.

  Next, in the forming step, a plurality of upper layer substrates are formed on the silicon wafer from which the oxide film on one surface has been removed. In this molding process, first, a photoresist is applied to one surface of the silicon wafer from which the oxide film has been removed using a spin coater, and the silicon wafer coated with the photoresist is baked in an oven and included in the photoresist. Evaporate the organic solvent. Next, the photoresist is exposed using an exposure machine, and then the photoresist is developed with an alkaline solution or the like using an organic draft to form 50 upper substrate patterns on the photoresist. Thereafter, using an optical microscope, it is inspected whether the pattern of the upper substrate is formed on the photoresist. Next, using a spin coater, a temporary fixing agent is applied to the heat-resistant glass bonded to the oxide film formed on the other surface of the silicon wafer, and the heat-resistant glass is applied to the other surface of the silicon wafer using a hot plate. Bonding with the formed oxide film. Next, the silicon wafer is etched to form a plurality of upper layer substrates on the silicon wafer. This etching is performed by applying a deep RIE process to the silicon wafer using an ICP etching apparatus. Next, using an organic draft, the photoresist is removed, and the temporary fixing agent and the heat-resistant glass are removed from the oxide film.

  Next, in the laminating process, a composite consisting of a silicon wafer on which a plurality of upper layer substrates are formed and an oxide film formed on the other surface of the silicon wafer is laminated on the insulating wafer with the silicon wafer facing the insulating wafer. Then, a plurality of stacked bodies in which the upper layer substrate and the lower layer substrate are stacked are formed. On the insulating wafer 5, the same number of lower layer substrates 3 as the upper layer substrate 2 previously formed on the silicon wafer 4 are formed by blasting or the like. The insulating wafer 5 is made of glass. Next, using an optical microscope, the positional deviation between the upper layer substrate and the lower layer substrate forming the laminate is inspected.

  Next, in the second oxide film removing step, the oxide film formed on the other surface of the silicon wafer is removed. This oxide film removing step is performed by performing dry etching on the oxide film formed on the other surface of the silicon wafer. Such dry etching is performed by flowing a hydrofluoric acid gas or the like in a corrosive gas atmosphere of about 30 to 80 ° C. for 10 to 20 minutes.

  Next, in the vapor deposition step, an aluminum film is formed by aluminum vapor deposition on a portion corresponding to the swing member of the upper substrate of the silicon wafer, and electrode pads are formed on the portions corresponding to the movable portion and the fixed portion by aluminum vapor deposition. These vapor depositions can be performed using a vacuum vapor deposition apparatus. Next, using the optical microscope, the formation position of the aluminum film and the electrode pad and the thickness inspection of the aluminum film and the electrode pad are performed.

  Next, in the dicing step, a dicing tape is attached to the lower surface of the insulating wafer, and a plurality of laminated bodies having the upper layer substrate and the lower layer substrate formed in the laminating step are separated using a dicer. A plurality of MEMS devices 1 are formed by removing the dicing tape from each lower layer substrate 3.

  Finally, in the chip characteristic evaluation step, electric characteristics are evaluated for each MEMS device. The electrical characteristics are evaluated by, for example, applying an AC voltage of 150 V +/− 50 V between the swinging comb teeth and the fixed comb teeth of each MEMS device at a frequency close to the mechanical resonance frequency unique to each MEMS device, This is based on whether or not the movable part normally swings.

(Production method in comparative example)
First, in a receiving inspection process, a receiving inspection of a silicon wafer is performed. Next, in the natural oxide film removing step, the natural oxide film acid formed on both surfaces of the silicon wafer is removed. The removal of the natural oxide film is performed using, for example, the same method as the removal of the oxide film performed in the first oxide film removal step of the embodiment.

  Next, a molding process is performed. This molding process is performed in the same manner as the molding process of the example except that a photoresist is applied to one surface of the silicon wafer and then a temporary bridge is formed. As shown in FIG. 4, the temporary bridge 22 connects the movable portion 2A and the fixed portion 2B of the upper substrate 2 to each other. The temporary bridge 22 is formed by using a photomask having the temporary bridge 22 as a photomask used when the photoresist applied to the silicon wafer is exposed by an exposure machine, so that the upper substrate 2 having the temporary bridge 22 is formed. A pattern is formed on the photoresist. Then, the etching is performed on the silicon wafer in which the pattern of the upper substrate 2 having the temporary bridge 22 is formed in the photoresist.

  Next, a lamination process is performed. This lamination process is performed by the same method as the lamination process of the embodiment.

  When the lamination process is finished, a vapor deposition process is performed next. This vapor deposition step is also performed in the same manner as the vapor deposition step of the example.

  When the vapor deposition process is completed, a temporary bridge cutting process is performed next. In this temporary bridge cutting step, the temporary bridge 22 shown in FIG. 4 is laser-cut one by one with a laser processing machine. In this comparative example, 50 upper layer substrates are formed on a silicon wafer at a time, and therefore, in this comparative example, four temporary bridges 22 are formed on one upper layer substrate 2 as shown in FIG. Therefore, there are 50 temporary bridges to be cut in this temporary bridge cutting step. Therefore, as shown in Table 1, the temporary bridge cutting process takes 120 minutes.

  When the temporary bridge cutting process is completed, each process is performed in the order of a dicing process and a chip characteristic evaluation process. The dicing process and the chip characteristic evaluation process are performed in the same manner as the dicing process and the chip characteristic evaluation process of the embodiment.

(Comparison result of production time in Example and Comparative Example)
As shown in Table 1, in the example, the total work time for manufacturing the MEMS device of 50 wafers per wafer was 347 minutes. On the other hand, in the comparative example, the total work time for manufacturing one wafer (50 pieces) of MEMS devices was 425 minutes. The working time of the comparative example is long because the time required for cutting the temporary bridge is long. Therefore, according to the manufacturing method according to the embodiment, the manufacturing of the MEMS device for one silicon wafer (50 pieces) can be shortened by 78 minutes.

FIG. 1 is an exploded perspective view showing a schematic configuration of a MEMS device manufactured by a method for manufacturing a MEMS device according to an embodiment of the present invention. FIG. 2 is a manufacturing flowchart showing the manufacturing procedure of the MEMS device according to the present embodiment. FIG. 3 is a manufacturing flowchart showing the manufacturing procedure of the MEMS device according to the present embodiment. FIG. 4 is a plan view of the upper substrate on which the temporary bridge is formed.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... MEMS device, 2 ... Upper layer board | substrate, 2A ... Movable part, 2B ... Fixed part, 3 ... Lower layer board | substrate

Claims (2)

In a method for manufacturing a MEMS device, comprising: an upper layer substrate including a movable portion and a fixed portion separated from the movable portion; and a lower layer substrate stacked on the upper layer substrate.
An oxide film forming step of heating the silicon wafer to form an oxide film on both sides of the silicon wafer;
A first oxide film removing step of removing an oxide film formed on one surface of the silicon wafer out of oxide films formed on both surfaces of the silicon wafer;
A molding step of forming a plurality of the upper substrate on the silicon wafer from which the oxide film on one surface has been removed;
A stacking step of stacking the silicon wafer on which the plurality of upper layer substrates are formed and the insulating wafer on which the plurality of lower layer substrates are formed to form a plurality of stacked bodies in which the upper layer substrate and the lower layer substrate are stacked; ,
A second oxide film removing step for removing an oxide film formed on the other surface of the silicon wafer of the laminate;
And a dicing step of forming a plurality of MEMS devices by separating the plurality of stacked bodies formed in the stacking step.   The MEMS device manufacturing method according to claim 1, wherein the second oxide film removing step is performed by etching an oxide film formed on the other surface of the silicon wafer.