1. Field of the Invention
The present invention relates to packaged micro movable devices such as acceleration sensors and angular velocity sensors. The present invention also relates to methods for making such micro movable devices.
2. Description of the Related Art
In recent years, very small devices produced by micromachining technology have been finding wide application in various technical fields. Those devices, having tiny movable or swingable parts, serve as sensing devices such as angular velocity sensors and acceleration sensors. These sensing devices are used in fields of camera shake control technology for video cameras or mobile telephones equipped with cameras, for example. In addition, the sensing devices can be used for car navigation systems, airbag release timing control systems, and attitude control systems in automobiles and robots.
Miniaturized sensing devices include, for example, a swingable or movable part, a stationary part, a connecting part for connecting the movable part and the stationary part, a driver-electrode pair for driving the movable part, and a detection-electrode pair for detecting the operation or the amount of displacement of the movable part. Such a sensing device has a problem of deteriorated operation performance due to electrode contamination by foreign matters or dirt, or damage caused to the electrode. In order to avoid such electrode contamination or damage, packaging is often performed at a wafer level in manufacturing process of the sensing device. Techniques covering such a packaging process are disclosed in the following patent documents 1 and 2 for example.
a) shows a device wafer 300′, in which a plurality of sensing devices 300 are already formed after a predetermined process. Each sensing device 300 includes a movable part 301 which is swingable, a stationary part 302, a connecting part (not illustrated) for connecting the movable part 301 and the stationary part 302, a driver-electrode pair (not illustrated) for driving the movable part 301, and a detection-electrode pair (not illustrated) for detecting the operation or the amount of displacement of the movable part 301. The movable part 301 is thinner than the stationary part 302. To the device wafer 300′, packaging wafers 303′, 304′ are bonded. In other words, the device wafer 300′ or each of the sensing devices 300 is packaged at a wafer level. The device wafer 300′ which is packaged at a wafer level is then separated as shown in
In the sensing device 300, a sufficient amount of gaps G, G′ are provided between the movable part 301 and the packaging members 302, 304 so that the movable part 301, which makes a swinging movement when the device is driven will not make contact with the packaging members 303, 304. Each of the parts in each sensing device 300, e.g. the movable part 301 and the stationary part 302, is formed within the device wafer 300′ through etching and other predetermined processes performed to the device wafer 300′ which has a uniform thickness originally. In order to create the gaps G, G′, the device wafer 300′ is etched so as to make the movable part 301 thinner than the stationary part 302 which is bonded to the packaging members 303, 304.
However, according to such a conventional method as the above, relatively large nonuniformity is unavoidable in the thickness of a plurality of movable parts 301 formed in the device wafer 300′. As a result, relatively large nonuniformity is unavoidable among those movable parts 301 obtained from the same single devicewafer 300′, in terms of the inertia of those movable parts 301 in their swinging movement when the devices are in operation. Inertial nonuniformity in the movable part 301 of the sensing device 300 or micro movable device can be a cause of nonuniformity in operational characteristic of the movable part 301, and therefore should be as small as possible. Likewise, inertial nonuniformity in the movable part 301 in the sensing device 300 can be a cause of nonuniformity in the detecting characteristic regarding detection of movement or the amount of displacement of the movable part 301, and therefore should be as small as possible.
The present invention has been proposed under the above-described circumstances, and it is therefore an object of the present invention to provide a packaged micro-device that is suitable for reducing operational characteristic nonuniformity of the movable part in the micro movable device.
A first aspect of the present invention provides a method for making a packaged micro-device which includes a micro movable device having a movable part, a first packaging member having a first recess located correspondingly to the movable part, and a second packaging member having a second recess located correspondingly to the movable part. The method includes a first bonding step, a second bonding step and a dicing step. In the first bonding step, a first packaging wafer is bonded to a first surface side of a device wafer from which a plurality of the micro movable elements each having the movable part are to be formed. The device wafer has the first surface and a second surface which faces away from the first surface. The first packaging wafer is provided with a plurality of the first recesses. In the second bonding step, a second packaging wafer is bonded to the second surface side of the device wafer. The second packaging wafer is provided with a plurality of the second recesses. In the dicing step, a laminate assembly which includes the device wafer, the first packaging wafer and the second packaging wafer is cut.
In the first bonding step of the present method described above, the device wafer and the first packaging wafer are bonded together in such a way that each of the first recesses faces or will face the movable part in one of the micro movable devices which are already formed or to be formed in the device wafer. In the second bonding step, the device wafer and the second packaging wafer are bonded together in such a way that each of the second recesses faces the movable part in one of the micro movable devices which are already formed in the device wafer. Through such a first and a second bonding steps as described above, packaging at a wafer level is accomplished. Thereafter, the dicing step is performed to obtain individual pieces, i.e. individual micro movable elements each being in a packaged state.
In the present method, the first packaging wafer which is already formed with the first recesses each providing operation space for the movable part is bonded to the device wafer in the first bonding step, and further, the second packaging wafer which is already formed with the second recesses each providing operation space for the movable part is bonded to the device wafer in the second bonding step. Therefore, there is no need for etching the movable part thereby making the movable part thinner than the stationary part in order to provide a sufficient gap between the movable part and the two packaging members to prevent the movable part from contacting the first or the second packaging members as it makes swinging movement when the device is in operation. If the movable part is made thinner than the stationary part by etching, relatively large nonuniformity is unavoidable in the thickness of the movable part in a plurality of micro movable devices as has been described earlier in relation with a conventional method. The nonuniformity causes inertial nonuniformity in the movable parts, and the inertial nonuniformity can cause undesirable operational characteristic nonuniformity of the movable part. The present method which requires no thinning of the movable part is suitable for reducing inertial nonuniformity in the movable part. The present method which is suitable for reducing inertial nonuniformity in the movable part that swings when the device is in operation is suitable for reducing operational characteristic nonuniformity of the movable part.
Further, the present method does not require any step of thinning the movable part that is formed within a device wafer having a uniform thickness. Thus, the present method is suitable for manufacturing packaged micro-devices at a high yield rate.
In addition, the present method enables packaging at a wafer level during manufacturing process of the micro-device, and therefore is suitable to reduce operation performance deterioration of the movable part caused by dirt attached to the micro-device, or by damage incurred thereto.
The micro movable element of the present invention preferably includes, in addition to the movable part, a stationary part and a connecting part for connecting the stationary part and the movable part. The movable part is swingable. The micro movable element (and hence the micro-device) serves as a sensing device such as an angular velocity sensor or an acceleration sensor. Inertial nonuniformity in the movable part as a part of the sensing device is a cause for detection characteristic nonuniformity which affects detection of movement or the amount of displacement of the movable part. The present method which is suitable for reducing inertial nonuniformity in the movable part is suitable for reducing operational characteristic nonuniformity of the movable part, and in addition, suitable for reducing detection characteristic nonuniformity regarding detection of movement or the amount of displacement of the movable part.
In the method according to the first aspect of the present invention, preferably, the device wafer has a laminate structure including: a first layer having the first surface; a second layer having the second surface; and an intermediate layer between the first and the second layers. With this arrangement, a step of etching the first layer using a predetermined mask pattern as a mask may be performed before the first bonding step. In this case, the etching of the second layer using a predetermined mask pattern as a mask may be performed after the first bonding step and before the second bonding step.
Preferably, the stationary part of the micro movable device includes a terminal portion for external connection. The first packaging member includes an electroconductive portion extending through the first packaging member to be connected with the terminal portion. Such an arrangement as the above allows for electric wires which are electrically connected with the micro movable device and extended out of the package appropriately. With this arrangement, the electroconductive portion penetrating the first packaging member may be formed before the first bonding step. Alternatively, it is possible to make the electroconductive portion after the first bonding step.
Preferably, the first bonding step or the second bonding step or the both may be performed by one of methods such as an anodic bonding method, a direct bonding method, a room-temperature bonding method or an eutectic bonding method.
Preferably, the border between the device wafer and the first packaging wafer and the border between the device wafer and the second packaging wafer may be provided by an insulation film. Such an arrangement as described above prevents undue electric connection between the device wafer or each micro movable device and the first packaging wafer, or between the device wafer or each micro movable device and the second packaging wafer.
Preferably, the first recesses and/or the second recesses may be formed by DRIE, anisotropic wet etching or isotropic wet etching. These methods enable one to make the first recesses and the second recesses properly.
According to a second aspect of the present invention, there is provided a packaged micro-device that includes: a micro movable element having a movable part; a first packaging member including a first recess corresponding in position to the movable part; and a second packaging member including a second recess corresponding in position to the movable part. The micro-device configured in this manner can be appropriately made by the method according to the present invention. The micro movable element may further include, in addition to the movable part, a stationary part, and a connecting part for connecting the movable part to the stationary part so that the movable part is swingable relative to the stationary part. The micro movable element (hence the micro-device) may serve as a sensing device such as an angular velocity sensor and an acceleration sensor.
The packaged device X includes a sensing device Y, a packaging member 81 (not illustrated in
The sensing device Y includes a land 10, an inner frame 20, an outer frame 30, a pair of connecting parts 40, a pair of connecting parts 50, and comb-teeth electrodes 61, 62, 63, 64, 71, 72, 73, 74, and serves as an angular velocity sensor or an acceleration sensor. Also, the sensing device Y is made by means of bulk micromachining technology such as MEMS technology, from an SOI (Silicon On Insulator) substrate wafer. The wafer has a laminate structure including e.g. a first and a second silicon layer, and an insulation layer between the silicon layers. Each silicon layer is doped with impurity and has a predetermined level of electric conductivity. In
As shown in
As shown in
As shown in
The pair of connecting parts 40 which connect the land 10 and the inner frame 20 are formed from the first silicon layer. Each connecting part 40 is provided by two torsion bars 41. As shown in
The pair of connecting parts 50 which connect the inner frame 20 and the outer frame 30 are formed from the first silicon layer. Each connecting part 50 is provided by three torsion bars 51, 52, 53. As shown in
The comb-teeth electrode 61 is formed from the first silicon layer, and is provided by a plurality of electrode teeth 61a extending from the first layer portion 11 of the land 10. As shown in
The comb-teeth electrode 62 is formed from the first silicon layer, and is provided by a plurality of electrode teeth 62a extending from the first layer portion 11 of the land 10 away from the electrode teeth 61a of the comb-teeth electrodes 61. The electrode teeth 62a are parallel to each other.
The comb-teeth electrode 63 is formed from the first silicon layer, opposed to the comb-teeth electrode 61, and provided by a plurality of electrode teeth 63a which extends from the portion 21b of the first layer portion 21 in the inner frame 20. As shown in
The comb-teeth electrode 64 is formed from the first silicon layer, opposed to the comb-teeth electrode 62, and provided by a plurality of electrode teeth 64a which extends from the portion 21e of the first layer portion 21 in the inner frame 20. The electrode teeth 64a are parallel to each other, as well as to the electrode teeth 62a of the comb-teeth electrode 62. The comb-teeth electrode 64 and the comb-teeth electrode 61 as described above constitute a detection-electrode pair in the sensing device Y.
The comb-teeth electrode 71 is formed from the first silicon layer, and provided by a plurality of electrode teeth 71a which extends from the portion 21c of the first layer portion 21 in the inner frame 20. As shown in
The comb-teeth electrode 72 is formed from the first silicon layer, and provided by a plurality of electrode teeth 72a which extends from the portion 21f of the first layer portion 21 in the inner frame 20. The electrode teeth 72a are parallel to each other.
The comb-teeth electrode 73 is formed from the first silicon layer, opposed to the comb-teeth electrodes 71, and provided by a plurality of electrode teeth 73a which extends from the portion 31g of the first layer portion 31 in the outer frame 30. As shown in
The comb-teeth electrode 74 is formed from the first silicon layer, opposed to the comb-teeth electrodes 72, and provided by a plurality of electrode teeth 74a which extends from the portion 31h of the first layer portion 31 in the outer frame 30. The electrode teeth 74a are parallel to each other, as well as to the electrode teeth 72a of the comb-teeth electrode 72. The comb-teeth electrode 74 and the comb-teeth electrodes 72 as described above constitute a driver-electrode pair in the sensing device Y.
The packaging member 81 is bonded to the first layer portion 31 side of the outer frame 30 in the sensing device Y, and has a recess 81a correspondingly to the movable part of the sensing device Y. As shown in
In the packaged device X, a sufficient amount of gap is provided between the movable part and the packaging members 81, 82 so that the movable part (the land 10 and the inner frame 20) will not make contact with the packaging members 81, 82 during its swinging operation when the device is in operation.
When the sensing device Y is in operation, the movable part (the land 10 and the inner frame 20) makes a swinging operation around the axis A2 at a predetermined frequency or period. The swinging operation is accomplished by repeating a cycle of voltage application to between the comb-teeth electrodes 71, 73 and voltage application to between the comb-teeth electrode 72, 74. The voltage application to the comb-teeth electrodes 71 can be achieved through the conductor plug P3, the portion 31c in the outer frame 30, the torsion bar 53 of one of the connecting parts 50 and the portion 21c in the inner frame 20. The voltage application to the comb-teeth electrodes 72 can be achieved through the conductor plug P6, the portion 31f in the outer frame 30, the torsion bar 53 of the other of the connecting parts 50 and the portion 21f in the inner frame 20. The voltage application to the comb-teeth electrode 73 can be achieved through the conductor plug P7 and the portion 31g in the outer frame 30. The voltage application to the comb-teeth electrode 74 can be achieved through the conductor plug P8 and the portion 31h in the outer frame 30. In the present embodiment, the swinging operation of the movable part can be accomplished by grounding the comb-teeth electrodes 71, 72, and then repeating a cycle of sequential application of a predetermined electric potential to the comb-teeth electrode 73 and a predetermined electric potential to the comb-teeth electrode 74.
Now, assume that a certain amount of angular velocity or acceleration acts on the sensing device Y or the land 10 while the movable part is being swung as described above. This causes the land 10 to rotate about the axis A1 to a predetermined extent, thereby making a displacement to change the electrostatic capacity between the comb-teeth electrodes 61, 63 and as well as between the comb-teeth electrodes 62, 64. Based on the electrostatic change between the comb-teeth electrodes 61, 63, and the electrostatic change between the comb-teeth electrodes 62, 64, it is possible to detect the amount of rotational displacement of the land 10. Based on the detection result, it is possible to calculate the amount of angular velocity or acceleration acting on the sensing device Y or the land 10.
In the manufacture of the packaged device X, first, a device wafer 100 as shown in
Next, as shown in
Next, as shown in
Next, as shown in
The packaging wafer 201 can be made as follows for example: Specifically, first, dry etching (anisotropic dry etching) by means of DRIE is performed to a silicon wafer, using a predetermined mask, to form recesses 81a. Next, through-holes which penetrate the silicon wafer are formed through dry etching by means of DRIE performed to the silicon wafer, using a predetermined mask. Then, the through-holes are filled with electrically conductive material to form the conductor plugs Px.
Continuing with the manufacture of the packaged device X, next, as shown in
Next, exposed portions of the insulation layer 103 are removed by predetermined etching, and thereafter, as shown in
The packaging wafer 202 can be made as follows for example: Specifically, dry etching (anisotropic dry etching) by means of DRIE is performed to a silicon wafer, using a predetermined mask, to form recesses 82a. Additionally, if the sealing of the packaged device X need not be air-tight, through-holes may be formed so that inside of the recess 82a in the packaging wafer 202 will communicate with the outside.
Next, as shown in
In the first bonding step described with reference to
According to the present method, the packaging wafer 201, which is formed with recesses 81a each providing operation space for the movable part (the land 10 and the inner frame 20) in corresponding one of the sensing devices Y is bonded to the device wafer 100 in the first bonding step, and further, the packaging wafer 202, which is formed with recesses 82a each providing operation space for the movable part (the land 10 and the inner frame 20) in corresponding one of the sensing device Y is bonded to the device wafer 100 in the second bonding step. Therefore, according to the present method, there is no need for etching the movable part thereby making the movable part thinner than the stationary part (the outer frame 30) in order to provide a sufficient gap between the movable part and the packaging members 81, 82 to prevent the movable part from contacting the packaging members 81, 82 when it swings during operation of the device. If the movable part is made thinner than the stationary part (the outer frame 30) by etching, relatively large nonuniformity is unavoidable in the thickness of the movable parts in a plurality of micro movable devices as has been described earlier in relation with a conventional method. The nonuniformity causes inertial nonuniformity in the movable parts, and the inertial nonuniformity can cause undesirable operational nonuniformity of the movable parts. The present method, which requires no thinning of the movable part, is suitable for reducing inertial nonuniformity in the movable part. The present method which is suitable for reducing inertial nonuniformity in the movable part that swings when the device is in operation, is suitable for reducing operational nonuniformity of the movable part. Also, inertial nonuniformity of the movable part as a part of the sensing device is a cause for detection characteristic nonuniformity which affects detection of movement or the amount of displacement of the movable part. The present method which is suitable for reducing inertial nonuniformity in the movable part is suitable for reducing operational nonuniformity of the movable part, and in addition, for reducing detection characteristic nonuniformity which relates to detection of movement or the amount of displacement of the movable part.
Further, the present method does not require any step of thinning the movable part (the land 10 and the inner frame 20) which is formed within a device wafer 100 that has a uniform thickness originally. Thus, the present method is suitable for manufacturing sensing devices Y at a high yield rate.
In addition, the present method enables packaging at a wafer level, and therefore is suitable to reduce operation performance deterioration of the movable part caused by dirt attached to the micro movable device or sensing device Y, or by damage incurred thereto.
a) shows a packaging member 83 as a variation of the packaging member 81. The packaging member 83 includes a recess 83a. The recess 83a is formed in the silicon wafer by means of anisotropic wet etching performed with the use of a predetermined mask. The etchant for this wet etching may be KOH or TMAH. In the first bonding step which was described with reference to
b) shows a packaging member 84 as a variation of the packaging member 81. The packaging member 84 includes a recess 84a. The recess 84a is formed in the silicon wafer by means of isotropic wet etching performed with the use of a predetermined mask. The etchant for this wet etching may be a solution mix containing HF, HNO3, and CH3COOH. In the first bonding step which was described with reference to
a) shows a packaging member 85 as a variation of the packaging member 82. The packaging member 85 includes a recess 85a. The recess 85a is formed in the silicon wafer by means of anisotropic wet etching performed with the use of a predetermined mask. The etchant for this wet etching may be KOH or TMAH. If the sealing of the packaged device X need not be air-tight, through-holes may be formed so that inside of the recess 85a will communicate with the outside. In the second bonding step which was described with reference to
b) shows a packaging member 86 as a variation of the packaging member 82. The packaging member 86 includes a recess 86a. The recess 86a is formed in the silicon wafer by means of isotropic wet etching performed with the use of a predetermined mask. The etchant for this wet etching may be a solution mix containing HF, HNO3, and CH3COOH. If the sealing of the packaged device X need not be air-tight, through-holes may be formed so that inside of the recess 86a will communicate with the outside. In the second bonding step which was described with reference to
The first bonding step in the above-described manufacturing method may be performed by eutectic bonding method. If eutectic bonding method is used in the first bonding step, a eutectic metal pattern 91 as shown in
In the first bonding step of the above-described manufacturing method, a packaging wafer 201 as shown in
Number | Date | Country | Kind |
---|---|---|---|
2007-048806 | Feb 2007 | JP | national |