JP5446236B2 - Physical quantity sensor, manufacturing method thereof, internal pressure control method thereof, and semiconductor device - Google Patents

Description

  The present invention relates to a sealed device that seals a semiconductor substrate on which a movable part is formed, and more particularly to a physical quantity sensor that detects a physical quantity using a capacitive element that can be displaced according to an external force and a method for manufacturing the same.

  In recent years, various electronic devices have been reduced in size, weight, functionality, and functionality, and electronic components to be mounted have been required to have higher density. In response to such demands, an increasing number of electronic components are manufactured as semiconductor devices. For this reason, in addition to the semiconductor device manufactured as a circuit element, a sensor or the like for detecting a physical quantity is also manufactured using the semiconductor device, thereby achieving a reduction in size and weight. For example, in an acceleration sensor or an angular velocity sensor having a small and simple structure using MEMS (Micro Electro Mechanical Systems) technology, a movable portion that is displaced according to an external force is formed on a semiconductor substrate, and the displacement of the movable portion is electrostatically detected. A type of sensor (so-called capacitive sensor) that is detected using a capacitive element has been put into practical use. Such a sensor has a structure in which the semiconductor substrate is sealed with a sealing material (for example, a glass substrate) in order to displace the movable part stably, and the sealed sealing space is vented. This is done to eliminate the factor that hinders the displacement of the movable part. A device having a sealing structure in which such movable parts are sealed is referred to as a sealed device in this document. Sealed devices include SAW (Surface Acoustic Wave) elements and F-BAR (Thin Film Bulk Acoustic Wave Resonators) elements in addition to MEMS elements. In general, a capacitance type sensor has a weight part that can be displaced with a predetermined degree of freedom in a semiconductor substrate sandwiched between a pair of glass substrates, and the weight part can be displaced with acceleration or angular velocity. It is used as a weight part to detect. The displacement is detected based on the capacitance value of the capacitive element. Conventionally, in order to detect a physical quantity of a multi-axis component in a capacitance type sensor, a combination of a plurality of single-axis sensors has been used, but this is a problem in terms of size and cost.

  Therefore, research on a capacitive sensor capable of detecting multi-axis components with a single sensor element is in progress. A sensor that detects a physical quantity of a multi-axis component with such a single sensor element is disclosed (Patent Document 1 and Non-Patent Document 1).

  The principle of operation of the capacitive angular velocity sensor will be described with reference to FIG. 19A and 19B are diagrams showing a schematic configuration and basic operation of a capacitive angular velocity sensor. In FIG. 19A, the angular velocity sensor 700 includes an upper glass substrate (first substrate) 701, a semiconductor substrate 702, and a lower glass substrate (second substrate) 703. A driving electrode 704 is formed on a surface of the upper glass substrate 701 facing the semiconductor substrate 702. The semiconductor substrate 702 is formed with a weight portion 705 and a flexible portion 706 that supports the weight portion 705 so as to be displaceable in the vertical direction in the drawing. The semiconductor substrate 702 is sealed by bonding its upper and lower surfaces to the upper glass substrate 701 and the lower glass substrate 703 and sealing the inside.

  FIG. 19A is a diagram illustrating a state in which 5 V is applied to the electrode 704 as a driving voltage. In this case, the weight portion 705 is pulled upward by the electrostatic attractive force Fc between the electrode 704 and the weight portion 705. FIG. 19B is a diagram illustrating a state in which the drive voltage applied to the electrode 704 is 0V. In this case, the weight portion 705 is returned to the original position by the spring restoring force Fb of the flexible portion 706. In this angular velocity sensor 700, the Coriolis force associated with the angular velocity is detected in a state where the weight portion 705 is vibrated by repeating the operation of applying the drive voltage shown in FIGS. 19 (A) and 19 (B).

  In the angular velocity sensor 700 shown in FIGS. 19A and 19B, the semiconductor substrate 702 is vacuum-sealed in a region surrounded by an upper glass substrate 701 and a lower glass substrate 703. For example, Patent Document 2 proposes a mechanical quantity sensor that increases the degree of vacuum in the region sealed by the upper and lower glass substrates 701 and 703.

In this mechanical quantity sensor, a gas molecule absorbing material made of the same material as that of the detection electrode for detecting the displacement of the weight portion is provided on the inner surfaces of the upper and lower glass substrates that form the sealed chamber (sealed space). The vacuum level is raised. This is to suppress attenuation of the vibration operation due to the gas viscosity of the weight portion 705. That is, it can be seen that the vibration operation of the weight portion 705 depends on the internal pressure in the sealed space. For example, as shown in FIG. 20, the vibration frequency width of the resonance spectrum of the weight portion 705 needs to have a desired resonance frequency width (half-value width in the drawing) depending on the product specifications and the like. FIG. 20 shows a case where the resonance frequency of the resonance spectrum is fz. In this specification, the resonance spectrum is a peak appearing in the frequency spectrum having the maximum value at the resonance frequency, the resonance frequency is the peak frequency of the resonance spectrum, and the resonance frequency width is the resonance spectrum width. It is a half width.
JP 2006-226770 A JP 2007-57469 A Transactions on Sensors and Micromachines, Vol. 126, no. 6,2006 (Journal of the Institute of Electrical Engineers of Japan, Vol. 126, No. 6, 2006)

  However, the angular velocity sensor 700 as shown in FIGS. 19A and 19B is not good when the resonance frequency width is narrow, and corresponds to the sensor characteristics of the angular velocity set according to the specifications. It is desirable to adjust the desired resonance frequency width of the resonance spectrum.

  In view of the above, the present invention provides a sealed device and a capacitance-type physical quantity sensor that can control the internal pressure in the sealed space and adjust the resonance spectrum of the weight portion to a desired resonance frequency width. An object of the present invention is to provide a stationary device, a physical quantity sensor, a manufacturing method thereof, and an internal pressure control method thereof.

A physical quantity sensor according to an embodiment of the present invention includes a frame part, a weight part disposed inside the frame part, and a flexible part that connects the weight part and the frame part. A first substrate bonded to one side of the frame portion; a second substrate bonded to the other side of the frame portion; and a first substrate provided on the first substrate and facing the weight portion. One electrode and a second electrode provided on the second substrate and facing the weight portion, and disposed in a sealed space sealed by joining the first substrate and the second substrate. , a gas absorbing member that absorbs sealing space inside of the gas in the semiconductor substrate by heating the disposed sealing space, to release the gas inside the sealed space by heating Bei and a gas pressure adjusting member for adjusting the pressure inside the sealed space , The gas pressure adjusting member, the first electrode, a portion of the first electrode, the second electrode, or a portion of the second electrode, a first conductive layer comprising a first conductive material And a second conductive layer that is stacked on the first conductive layer and includes a second conductive material having a melting point higher than that of the first conductive material .

A method of manufacturing a physical quantity sensor according to an embodiment of the present invention includes: a semiconductor substrate; a frame portion; a weight portion disposed inside the frame portion; and a flexible portion that connects the weight portion and the frame portion. And forming a first electrode and a first wiring electrically connected to the first electrode on a first substrate bonded to one side of the frame portion, and Forming a second electrode and a second wiring electrically connected to the second electrode on a second substrate bonded to the other side; on the first substrate or the second substrate; a gas absorbing member that absorbs sealing space inside of the gas in the semiconductor substrate by heating, a gas pressure adjusting member for adjusting the pressure by releasing gas within said sealed space by heating The weight portion and the first electrode are opposed to each other, and the first substrate and the frame are formed. One side of the frame portion, the weight portion and the second electrode are opposed to each other, the second substrate and the other side of the frame portion are joined, and the first heat treatment is performed to Gas in the sealed space in the semiconductor substrate sealed by bonding of one substrate and the second substrate is absorbed by the gas absorbing member, and mixed by the second heat treatment when forming the gas pressure adjusting member. Adjusting the pressure inside the sealed space by discharging the gas into the sealed space, and when forming the first electrode, when forming the first wiring, when forming the second electrode, Alternatively, when forming the second wiring, a first conductive layer including a first conductive material and a second conductive layer including a second conductive material having a melting point higher than that of the first conductive material are stacked. Then, the gas pressure adjusting member is formed .

An internal pressure control method for a physical quantity sensor according to an embodiment of the present invention includes: a frame part; a weight part disposed inside the frame part; a flexible part that connects the weight part and the frame part; A semiconductor substrate comprising: a first substrate bonded to one side of the frame portion; a second substrate bonded to the other side of the frame portion; and the weight provided on the first substrate. A first electrode opposed to a portion, a second electrode provided on the second substrate, opposed to the weight portion, disposed on the first substrate or the second substrate, and heated to thereby A gas absorbing member that absorbs gas inside the sealed space in the semiconductor substrate sealed by bonding the first substrate and the second substrate; and disposed in the sealed space and heated to seal the seal. internal pressure regulation to adjust the pressure by releasing gas therein stop space In the physical quantity sensor comprising a member, the gas absorbing member absorbs the gas inside the sealed space in the semiconductor substrate sealed by joining the first substrate and the second substrate by the first heat treatment. , the second heat treatment, said gas which is mixed at the time of formation of the gas pressure adjusting member is emitted into the sealed space comprises adjusting the pressure inside the sealed space, the gas pressure adjusting member, The gas pressure adjusting member is formed by laminating a first conductive layer including a first conductive material and a second conductive layer including a second conductive material having a melting point higher than that of the first conductive material. An area is set at the time of forming, and the pressure inside the sealed space is adjusted by controlling the heating temperature and the heating time of the second heat treatment according to the area.

  According to the present invention, in a sealed device and a capacitance-type physical quantity sensor, sealing that enables adjustment of the resonance spectrum of the weight portion to a desired resonance frequency width by controlling the internal pressure in the sealing space. A mold device, a physical quantity sensor, a manufacturing method thereof, and a pressure control method thereof can be provided.

Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. Note that in this embodiment, an example of a physical quantity sensor as a sealed device will be described.
<Structure of physical quantity sensor>
FIG. 1 is an exploded perspective view showing a state in which the physical quantity sensor 100 is disassembled. In FIG. 1, two axes (X axis and Y axis) perpendicular to the plane of the physical quantity sensor 100 are set, and a direction perpendicular to the two axes is defined as a Z axis. The physical quantity sensor 100 is configured by sandwiching a semiconductor substrate W between a first substrate 140 and a second substrate 150 positioned above and below the semiconductor substrate W. The semiconductor substrate W is configured by laminating a silicon film 110, a BOX layer 120, and a silicon substrate 130 in this order. The semiconductor substrate W is displaced by a frame-like frame (including the frame portion 111 and the frame portion 131) and a flexible portion 113 (113a to 113d) having flexibility in the frame by a manufacturing process described later. The weight joint portions (including the weight joint portion 112 and the weight joint portion 132) that are supported in an integral manner are integrally configured to form a sensor unit that detects a physical quantity. In addition, block upper layer portions 114a to 114j are formed in the silicon film 110 so as to be separated from the frame portion 111, the weight joint portions 112a to 112e, and the flexible portions 113 (113a to 113d).

  The outer periphery of the silicon film 110, the BOX layer 120, the silicon substrate 130, the first substrate 140, and the second substrate 150 has a substantially square shape of, for example, 3 mm × 3 mm, and their heights are 20 μm, 2 μm, 600 μm, and 500 μm, respectively. 500 μm. These external shapes and heights are examples, and are not limited to the above.

  The semiconductor substrate W including the silicon film 110, the BOX layer 120, and the silicon substrate 130 can be manufactured using an SOI (Silicon On Insulator) substrate. The first substrate 140 and the second substrate 150 are made of any one of a glass material, a semiconductor material, a metal material, and an insulating resin material.

  Next, a detailed configuration of the silicon film 110 will be described with reference to FIG. 2A is a plan view of the silicon film 110, FIG. 2B is a cross-sectional view of the semiconductor substrate W as viewed from line AA in FIG. 2A, and FIG. 2C is from line CC in FIG. It is sectional drawing of the semiconductor substrate W which was seen.

  In the silicon film 110 shown in FIG. 2A, a frame portion 111, weight joint portions 112a to 112e, flexible portions 113a to 113e, and block upper layer portions 114a to 114j are formed. The frame portion 111 is a polygonal frame-shaped substrate whose outer periphery is substantially square and whose inner periphery corresponds to the shape of the weight joint portions 112a to 112e. The weight joint portions 112a to 112e have a substantially clover-like shape when FIG. 2A is viewed from the Z direction. The weight joint portions 112a to 112e are joined to the weight portion 132 (the weight portion 132 shown in FIGS. 2B and 2C) having substantially the same shape as the weight joint portions 112a to 112e via the BOX layer 120, and the frame It is displaced integrally with the part 111. The flexible portions 113a to 113d are substantially rectangular substrates, respectively, and connect the frame portion 111 and the weight joint portions 112b to 112e in four directions. The flexible portions 113a to 113d have flexibility because they are thin, and function as beams that can be bent. When the flexible portions 113a to 113d are bent, the weight joint portions 112a to 112e can be displaced with respect to the frame portion 111. The block upper layer portions 114a to 114j are formed apart from the frame portion 111, the weight joint portions 112a to 112e, and the flexible portions 113a to 113e.

  The weight joint portions 112a to 112e and the flexible portions 113a to 113e can be partially displaced with respect to the first substrate 140 formed at a position lower than the surface of the frame portion 111, as shown in FIG. It is.

  The upper surface of the weight junction 112a functions as a drive electrode. The upper surface of the weight junction 112a is connected to the drive electrodes 144a to 144e (described later) installed on the lower surface of the first substrate 140 by a voltage applied to the weight junctions 112a to 112e in the Z axis. Vibrate in the direction. Details of this drive will be described later.

  The upper surfaces of the weight joint portions 112b to 112e function as detection electrodes to be described later that detect displacements in the X-axis and Y-axis directions of the weight joint portions 112b to 112e. The upper surfaces of the weight junctions 112b to 112e are capacitively coupled to detection electrodes 141b to 141e (described later) installed on the lower surface of the first substrate 140, respectively. In addition, the alphabet part (b-e) of the code | symbol attached | subjected to the weight junction parts 112b-112e and the detection electrodes 141b-141e, respectively is attached | subjected in the same order corresponding to mutual positional relationship. Details of this detection will be described later.

  In FIG. 2B, a frame portion 131, weight portions 132 (132a to 132e), and block lower layer portions 134a to 134j are formed on the silicon substrate 130. In the silicon substrate 130, the frame portion 131 and the weight portion 132 (132a to 132e) can be formed by forming an opening by etching the semiconductor substrate W. The height of the weight portion 132 (in the Z-axis direction in FIG. 2B) is made lower than the height of the frame portion 131. This is because a gap corresponding to the measurement range is secured between the weight part 132 and the second substrate 150 and the weight part 132 can be displaced. In addition, the block lower layer parts 134a-134j are formed away from the frame part 131, the weight part 132, and the flexible parts 113a-113e.

  The frame part 131 is a polygonal frame-shaped substrate whose outer periphery is substantially square and whose inner periphery corresponds to the shape of the weight part 132, and has a shape corresponding to the frame part 111 of the silicon film 110. The frame part 131 is joined to the frame part 111 via the BOX layer 120 a and is integrated with the frame part 111.

  The weight part 132 functions as a weight (action body) that receives a force caused by acceleration or a Coriolis force caused by angular velocity. The weight part 132 is divided into substantially rectangular parallelepiped weight parts 132a to 132e. Weight parts 132b to 132e are connected to the weight part 132a arranged at the center from four directions, and can be displaced (moved or rotated) integrally as a whole. That is, the weight part 132a functions as a connection part for connecting the weight parts 132b to 132e. The weight part 132 has a substantially clover-like shape when FIG. 2A is viewed from the vertical direction.

  The weight parts 132a to 132e have substantially square cross-sectional shapes (shapes seen from the XY coordinate plane in FIG. 2A) corresponding to the weight joint parts 112a to 112e, respectively. The weight parts 132a to 132e are joined to the weight joint parts 112a to 112e via the BOX layer 120b. The weight joint 112 is displaced according to the force applied to the weights 132a to 132e, and as a result, the physical quantity can be measured.

  The reason why the weight part 132 is configured as the weight parts 132a to 132 is to achieve both miniaturization and high sensitivity of the physical quantity sensor 100. When the physical quantity sensor 100 is downsized (capacity reduction), the capacity of the weight portion 132 is also reduced and the mass thereof is reduced, so that the sensitivity to the physical quantity is also reduced. The weights 132b to 132e are distributed and arranged so as not to hinder the bending of the flexible parts 113a to 113d, thereby securing the mass of the weight part 132 as a whole. As a result, the physical quantity sensor 100 can be both reduced in size and increased in sensitivity.

  The lower surface of the weight portion 132a (the surface facing the upper surface of the second substrate 150) functions as a driving electrode E described later. The lower surface of the weight portion 132a is configured such that the weight joint portions 112a to 112e are Z-axised by a voltage applied to a driving electrode 151a (described later with reference to FIG. 3B) installed on the upper surface of the second substrate 150. Vibrate in the direction. Details of this drive will be described later.

  The lower surfaces of the weight portions 132b to 132e function as detection electrodes to be described later that detect displacements in the X-axis and Y-axis directions of the weight joint portions 112b to 112e. The detection electrodes on the back surfaces of these weight portions 132b to 132e are capacitively coupled to detection electrodes 151b to 151e (see FIG. 3B), which will be described later, installed on the upper surface of the second substrate 150, respectively. In addition, the alphabet part (b-e) of the code | symbol attached | subjected to the weight parts 132b-132e and the detection electrodes 151b-151e, respectively is attached | subjected in the same order corresponding to the mutual positional relationship. Details of this detection will be described later.

  The BOX layer 120 shown in FIGS. 2B and 2C includes a BOX layer 120a that connects the frame portion 111 and the frame portion 131, and a BOX layer 120b that connects the weight joint portions 112a to 112e and the weight portions 132a to 132e. And the BOX layer 120c connecting the block upper layer portions 114a to 114j and the block lower layer portions 134a to 134j. The BOX layer 120 is not connected to the silicon film 110 and the silicon substrate 130 at portions other than the portions shown in FIGS.

  The block lower layer portions 134a to 134j have substantially square cross-sectional shapes corresponding to the block upper layer portions 114a to 114j, respectively, and are joined to the block upper layer portions 114a to 114j by the BOX layer 120c. The block obtained by joining the block upper layer portions 114a to 114j and the block lower layer portions 134a to 134j supplies power to the drive electrode 141a and the detection electrodes 141b to 141e, and the drive electrode 151a and the detection electrodes 151b to 151e described later, respectively. It is used for wiring purposes.

  2A to 2C, a part of the frame part 111 and the central parts of the block upper layer parts 114a to 114j, a part of the BOX layer 120b, and the central parts of the BOX layer 120c. A conductive portion 160 is formed to connect the first substrate 140 and the silicon substrate 130 at necessary portions. The conducting part 160 conducts the first substrate 140 and the silicon substrate 130, and part of the frame part 111 and block upper layer parts 114a to 114j of the silicon film 110, part of the BOX layer 120b and BOX layer 120c. Is formed. The conducting part 160 is configured by arranging a metal layer in a tapered through hole, for example. These conducting portions 160 are formed in accordance with the formation positions of wiring terminals T1 to T11 formed on the first substrate 140 described later. It is preferable that the conduction part 160 is not formed on the upper surface of the active layer.

  Next, the first and second substrates 140 and 150 will be described with reference to FIG. FIG. 3A is a plan view of the first substrate 140 seen through from the positive Z direction. A drive electrode 141a and detection electrodes 141b to 141e are disposed on the lower surface of the first substrate 140 (the side facing the silicon film 110). The drive electrode 141a is electrically connected to the wiring terminal T1 through the wiring L1. The detection electrodes 141b to e are electrically connected to the wiring terminals T3 to T6 through the wirings L3 to L6. The subscript numbers correspond to each other. The numerical portions (1, 3 to 6) of the reference numerals attached to the wirings L1, L3 to L6 and the wiring terminals T1, T3 to T6 are attached in the same order corresponding to the mutual positional relationship. Yes. In this way, it is possible to take out electrical signals from the drive electrode 141a and the detection electrodes 141b to e.

  FIG. 3B is a plan view of the second substrate 150 viewed from the positive Z direction. A drive electrode 151a and detection electrodes 151b to 151e are disposed on the upper surface of the second substrate 150 (the side facing the silicon substrate 130). The drive electrode 151a is electrically connected to the wiring terminal T2 through the wiring L2. The detection electrodes 151b to 151e are electrically connected to the wiring terminals T7 to T10 through the wirings L7 to L10. Although details are not shown here, the wirings L1 to L10 are in a state of being interposed between the block portion and the first substrate or the second substrate. In this way, it is possible to take out electrical signals from the drive electrode 151a and the detection electrodes 151b to 151e. The numerical portions (2, 7 to 10) of the reference numerals attached to the wirings L2 and L7 to L10 and the wiring terminals T2 and T7 to T10 are attached in the same order corresponding to the mutual positional relationship. Yes.

  With the configuration shown in FIGS. 2 and 3 above, electrical connection to the outside of the physical quantity sensor 100 (CV conversion circuit or the like) and the drive electrodes 141a and 151a and the detection electrodes 141b to 141e and 151b to 151e is established. It is possible.

  For example, the driving electrode, the detection electrode, and all or part of the wiring are, for example, a first conductive layer containing a first conductive material as a lower layer, and a second conductive having a melting point higher than that of the first conductive material. A two-layer structure in which the second conductive layer containing a conductive material is an upper layer can be used.

  Further, on the upper surface side of the second substrate 150 shown in FIG. 3B, a getter 201 (gas absorbing member) containing a gas absorbing substance and a dummy electrode 202 (gas pressure adjusting member) containing a substance for adjusting the gas pressure are provided. And are arranged. For the getter 201, for example, a substance that absorbs water or oxygen, such as zirconium Zr, can be used. This getter 201 chemically absorbs and absorbs oxygen molecules or water molecules remaining in the sealed space of the semiconductor substrate W, the first substrate 140, and the second substrate 150 bonded by heating, and is absorbed in the sealed space. Improve the degree of vacuum.

  Note that the material for forming the getter 201 may be any metal that can absorb and react with oxygen molecules or water molecules by heating, such as titanium Ti, vanadium V, iron Fe, nickel Ni, zirconium. Zr or an alloy containing these may be used. The heating temperature is desirably 100 ° C. or higher, for example, but is higher than the temperature at which the pattern formed by the metal thin film such as the driving electrodes 141a and 151a, the detection electrodes 141b to 141e, and 151b to 154e is damaged. It is desirable that the temperature be lower. For example, when the drive electrodes 141a and 151a and the detection electrodes 141b to 141e and 151b to 154e are formed of aluminum Al, the temperature is desirably 400 ° C. or lower.

  The dummy electrode 202 has the first conductive layer containing the first conductive material as a lower layer, has a higher melting point than the first conductive material, and releases the adsorbed gas at a temperature higher than the annealing temperature (heating temperature). A two-layer structure having a second conductive layer containing a second conductive material as an upper layer is used. The dummy electrode 202 is provided to control the gas pressure in the sealed space, and the gas pressure in the sealed space is adjusted by adjusting the area during formation, the annealing temperature, and the annealing time. It is possible to control. Details of this will be described later.

<Operation of physical quantity sensor>
As described above, in the physical quantity sensor 100, the weight portion in which the weight joint portion 112 and the weight portions 132 (132 a to 132 e) are integrally formed is supported by the flexible portion 113 extending from the frame portion 111, and the first substrate 140. The second substrate 150 and the semiconductor substrate W are configured to be displaceable within a space.

  When the physical quantity sensor 100 is used as an acceleration sensor, it is only necessary to detect the displacement of the weight part 132 caused by the action of acceleration. The acceleration detects the displacement of the weight part (the joined body of the weight joint part 112 and the weight part 132) based on the capacitance change of the capacitive element formed by the weight joint part 112, the weight part 132, and the detection electrode. The acceleration in the X and Y axis directions can be detected by detecting the inclination of the weight portion, and the acceleration in the Z axis direction can be detected by detecting the displacement of the weight portion along the Z axis direction.

  When the physical quantity sensor 100 is used as an angular velocity sensor, the weight portion 132 is caused to vibrate up and down by the drive electrodes 141a and 151a (generally, an alternating voltage is applied to make a single vibration), and the weight portion 132 generated due to the action of the angular velocity. What is necessary is just to detect the displacement of. For example, if the angular velocity ω is applied when the weight part 132 is moving in the Z-axis direction at the speed vz, the Coriolis force F acts on the weight part 132. Specifically, according to the angular velocity ωx in the X-axis direction and the angular velocity ωy in the Y-axis direction, the Coriolis force Fy (= 2 · m · vz · ωx) in the Y-axis direction and the Coriolis force Fx (= 2 · m · vz · ωy) acts on the weight portion 132 (m is the mass of the weight portion 132). When the Coriolis force Fy due to the angular velocity ωx in the X-axis direction is applied, an inclination in the Y direction occurs at the weight joint 112. As described above, the weight joint 112 is inclined (displaced) in the Y direction and the X direction by the Coriolis forces Fy and Fx caused by the angular velocities ωx and ωy. Therefore, by detecting the displacement of the weight portion 132 in each axial direction, the value of the angular velocity of each axial component can be obtained. In the physical quantity sensor 100, the value of the angular velocity of each axial component can be detected by detecting the change in the capacitance of the capacitive element formed between the weight part 132 and each electrode.

  When a voltage is applied between the driving electrodes, the driving electrodes are attracted to each other by the Coulomb force, and the weight portions (the weight joint portion 112 and the weight portion 132) are displaced in the positive direction of the Z axis. Further, when a voltage is applied between the driving electrodes, the driving electrodes are attracted to each other by the Coulomb force, and the weight joint 112 (also the weight 132) is displaced in the Z-axis negative direction. That is, by alternately applying a voltage to the upper and lower drive electrodes, the weight joint 112 (also the weight 132) vibrates in the Z-axis direction. The voltage can be applied using a positive or negative direct current waveform (a pulse waveform when considering non-application), a half-wave waveform, or the like. The period of vibration of the weight joint 112 is determined by the period for switching the voltage. This switching cycle is preferably close to the natural frequency of the weight joint 112 to some extent. The natural frequency of the weight portion is determined by the elastic force of the flexible portion 113, the mass of the weight portion 132, and the like. If the period of vibration applied to the weight part 132 does not correspond to the natural frequency, the energy of vibration applied to the weight part is diverged and energy efficiency is reduced. Note that an AC voltage having a frequency half that of the natural frequency of the weight portion may be applied only to either the driving electrodes or between the driving electrodes.

  In general, the angular velocity signal is several kHz or more, and the acceleration signal has a frequency two or more digits lower than the angular velocity signal, so that each can be identified by an external signal processing circuit. In other words, the acceleration and angular velocity are processed by the filter circuit for the low frequency component (or bias component) and the signal following the vibration frequency by the signal processing circuit provided outside, and each signal after the processing is detected. It is possible to detect the acceleration in the triaxial (X, Y, Z) direction and the angular velocity in the biaxial (X, Y) direction. That is, by using the physical quantity sensor 100 as one sensor element, it is possible to detect the acceleration in the triaxial (X, Y, Z) direction and the angular velocity in the biaxial (X, Y) direction. Further, the physical quantity sensor 100 can be used as a sensor that detects only acceleration / angular velocity. The physical quantity sensor 100 described in the present embodiment can detect the acceleration in the three-axis (X, Y, Z) direction and the angular velocity around the two axes (X, Y). When detecting acceleration in the three-axis direction, the drive electrodes 141a and 151a described above function as detection electrodes for detecting acceleration in the Z-axis direction.

<Method of manufacturing physical quantity sensor>
Hereinafter, the manufacturing method of the physical quantity sensor 100 will be described with reference to FIGS. 4 (A) to (D) and FIGS. 5 (A) to (C). 4 (A) to 4 (D) and FIGS. 5 (A) to 5 (C) show the respective manufacturing steps based on the cross section viewed from the line CC shown in FIG. 2 (A).

(1) Preparation of semiconductor substrate W (see FIG. 4A)
A semiconductor substrate W (SOI substrate) formed by stacking a silicon film 110, a BOX layer 120, and a silicon substrate 130 is prepared. As described above, the silicon film 110 is a layer constituting the frame portion 111, the weight joint portion 112, the flexible portion 113, and the block upper layer portion 114. The BOX layer 120 is a layer that joins the silicon film 110 and the silicon substrate 130 and functions as an etching stopper layer. The silicon substrate 130 is a layer constituting the frame part 131, the weight part 132, and the block lower layer part 134. The semiconductor substrate W is produced by SIMOX or a bonding method.

(2) Processing of silicon film 110 (see FIG. 4B)
By forming a mask for processing the frame portion 111, the weight joint portions 112a to 112e, the flexible portions 113a to 113d, and the block upper layer portions 114a to 114j, and etching the silicon film 110 through the mask, the frame portion 111, and a concave portion 170 is formed at a position where the weight joint portions 112a to 112e and the flexible portions 113a to 113d are formed. As an etching method, a RIE (Reactive Ion Etching) method can be used.

(3) Processing of the silicon film 110 and the BOX layer 120 (see FIG. 4C)
By etching the silicon film 110 and the BOX layer 120 on which a predetermined mask is formed, the processing positions of the weight junction portions 112a to 112e, the flexible portions 113a to 113d, the block upper layer portions 114a to 114j, and the conduction portion 160 are determined. An opening is formed, and weight junction portions 112a to 112e, flexible portions 113a to 113d, block upper layer portions 114a to 114j, and an opening 161 corresponding to the conductive portion 160 are formed in the silicon film 110 and the BOX layer 120. As an etching method, an etching method called DRIE (Deep Reactive Ion Etching) can be used. In DRIE, the etching process that digs while eroding the material layer in the thickness direction and the deposition process that forms a polymer wall on the side of the dug hole can be repeated alternately, allowing erosion to proceed almost only in the thickness direction. Become. In this case, an etching material having etching selectivity between silicon oxide and silicon may be used. For example, a mixed gas of SF ₆ gas and O ₂ gas may be used in the etching stage, and C ₄ F ₈ gas may be used in the deposition stage. Subsequently, for example, Al is deposited by an evaporation method, a sputtering method, or the like in the opening 161 penetrating the silicon film 110 and the BOX layer 120a, thereby forming the conductive portion 160. Unnecessary metal layers deposited on the upper surface of the silicon film 110 (metal layers outside the upper edge (not shown) of the conductive portion 160) are removed by etching. In FIG. 4C, a processed cross section of the frame portion 111 and the weight joint portion 112a is shown.

(4) Joining the first substrate 140 (see FIG. 4D)
The bonding of the first substrate 140 is performed by the following steps 1) to 3).

1) Creation of the 1st board | substrate 140 The 1st board | substrate 140 is comprised from either a glass material, a semiconductor, a metal material, and an insulating resin material. A case where a glass material is used as the first substrate 140 will be described. A glass substrate containing movable ions (for example, Pyrex (registered trademark) glass) is used. The driving electrode 141a, the detection electrodes 141b to 141e, and the wirings L1 and L3 to L6 are made of Al, for example, at positions facing the weight junctions 112a to 112e on the surface of the first substrate 140 facing the silicon film 110, respectively. (See FIG. 3A). Further, eleven wide conical through holes 171 for forming wiring terminals T1 to T11 are formed by etching or sandblasting the first substrate 140 (see FIGS. 4D and 3A). . FIG. 4D shows a cross section in which the wiring terminal T1 is formed.

  As described above, the drive electrode 141a detection electrodes 141b to 141e and the wirings L1, L3 to L6 can have a two-layer structure (first conductive layer and second conductive layer). Further, a configuration in which a part thereof has a two-layer structure instead of a two-layer structure may be adopted. When the driving electrode 141a and the detection electrodes 141b to 141e have a two-layer structure, it is possible to prevent the driving electrode 141a and the detection electrodes 141b to 141e from being deteriorated by an etching gas during etching. These first and second conductive layers can both be formed by sputtering. When gold Au, platinum Pt or the like is used as the first conductive layer, an adhesion film (copper Cu in the case of gold Au, titanium Ti in the case of platinum Pt) between the upper surface of the second substrate 150 is used. It is preferable to add.

2) Bonding of Semiconductor Substrate W and First Substrate 140 The first substrate 140 and the semiconductor substrate W are bonded by anodic bonding (see FIG. 4D). Since the first substrate 140 is anodically bonded before the movable portion is formed, the weight bonded portions 112a to 112e are not affected by the first substrate even if electrostatic attraction occurs during the anodic bonding of the semiconductor substrate W and the first substrate 140. It is not attracted to the 140 side.

3) Formation of wiring terminals T1 to T11 In the upper surface of the first substrate 140 and the conical through hole 171, for example, a metal layer is formed in the order of a Cr layer and an Au layer by vapor deposition or sputtering. Unnecessary metal layers (metal layers outside the top edges of the wiring terminals T1 to T11) are removed by etching to form wiring terminals T1 to T11 (see FIG. 3A). The wiring terminals T1 to T11 may be formed before bonding to the semiconductor substrate W.

(5) Processing of silicon substrate 130 (see FIG. 5A)
Etching the area corresponding to the position where the frame part 131, the weight parts 132a to 132e and the block lower layer parts 134a to 134j are formed and the position where the getter 201 and the dummy electrode 202 are arranged on the surface of the silicon substrate 130 facing the second substrate 150. By doing so, the recess 180 is formed. As an etching method, a RIE (Reactive Ion Etching) method can be used.

(6) Processing of silicon substrate 130 (see FIG. 5B)
A mask for processing the frame part 131, the weight parts 132a to 132e, and the block lower layer parts 134a to 134j is formed, and the silicon substrate 130 is etched through the mask, whereby the frame part 111 and the weight joint parts 112a to 112a. An opening 181 corresponding to 112e and the flexible portions 113a to 113d is formed. As an etching method, an etching method called DRIE (Deep Reactive Ion Etching) can be used. FIG. 5B shows a processed cross section of the frame portion 131 and the weight portion 132a.

(7) Joining the second substrate 150 (see FIG. 5C)
The bonding of the second substrate 150 is performed by the following steps 1) to 2).

1) Creation of the second substrate 150 As the second substrate 150, substantially the same material as the first substrate 140 described above can be used. In this embodiment, the case where a glass substrate is used as the second substrate 150 will be described. The driving electrode 151a, the detection electrodes 151b to 151e, and the wirings L2 and L7 to L10 are formed, for example, in a pattern made of aluminum Al or the like at positions facing the weight portions 132a to 132e of the glass substrate containing movable ions. (See FIG. 3B). Further, the getter 201 is formed by a pattern made of, for example, aluminum Al at the position shown in FIG. 3B, and the dummy electrode 202 is formed by a pattern as a two-layer structure at the position shown in FIG. 3B. To do.

  As described above, the driving electrode 151a and the detection electrodes 151b to 151e can have a two-layer structure (a first conductive layer and a second conductive layer). The wirings L2, L7 to L10 may have a structure using the first conductive layer, and a part of the wirings may have a second conductive layer on the first conductive layer. By making the driving electrode 151a and the detection electrodes 151b to 151e have a two-layer structure, it is possible to suppress the deterioration of the driving electrode 151a and the detection electrodes 151b to 151e due to the etching gas during etching. These first and second conductive layers can both be formed by sputtering. When gold Au, platinum Pt or the like is used as the first conductive layer, an adhesion film (copper Cu in the case of gold Au, titanium Ti in the case of platinum Pt) between the upper surface of the second substrate 150 is used. It is preferable to add.

2) Bonding of the semiconductor substrate W and the second substrate 150 The second substrate 150 and the semiconductor substrate W are bonded by anodic bonding. FIG. 5C shows a state where the semiconductor substrate W and the second substrate 150 are bonded.

(8) Dicing of the semiconductor substrate W, the first substrate 140, and the second substrate 150 For example, after the getter 201 is activated by annealing at 450 ° C. to absorb oxygen molecules and water molecules remaining in the sealed space, The bonded semiconductor substrate W, first substrate 140, and second substrate 150 are cut with a dicing saw or the like and separated into individual physical quantity sensors 100. The physical quantity sensor 100 can be manufactured as described above. When the driving electrode 151a and the detection electrodes 151b to 151e have a two-layer structure, a sputtering gas (for example, Ar gas) adsorbed by the heating temperature during the annealing process is released from these electrode groups. Details of the gas release phenomenon by the laminated electrode will be described later.

  Here, the reason why a two-layer structure (first conductive layer, second conductive layer) is used for the drive electrode 151a and the detection electrodes 151b to 151e will be described.

  By using a two-layer structure (first conductive layer, second conductive layer) for the drive electrode 151a and the detection electrodes 151b to 151e, a gas release phenomenon, which will be described later, occurs in the annealing process after the physical quantity sensor 100 is sealed. Occur. It is possible to adjust the internal pressure in the sealed space by utilizing this gas release phenomenon. Further, by using a two-layer structure (first conductive layer, second conductive layer) for the drive electrode 151a and the detection electrodes 151b to 151e, “sticking”, “hillock”, etc. described below can be suppressed. It becomes possible. When the semiconductor substrate W and the second substrate 150 are anodic bonded, the weight bonding portions 112a to 112e are movably supported by the flexible portions 113a to 113d, so that the weight portions 132a to 132e are caused by electrostatic attraction generated during anodic bonding. There is a possibility of being drawn toward the second substrate 150 side. For example, a state called “sticking” occurs in which the weights 132 a to 132 e are attracted to the second substrate 150 and adhered to the second substrate 150 by electrostatic attraction generated when the semiconductor substrate W and the second substrate 150 are anodically bonded. there's a possibility that. The presence / absence of “sticking” can be confirmed by examining conduction between the weights 132a to 132e and the driving electrode 151a. If “sticking” has occurred, a short (conduction) occurs, and if “sticking” does not occur, it becomes open (insulation).

  In addition, when the semiconductor substrate W and the second substrate 150 are anodically bonded, a “hillock” may occur. The “hillock” is, for example, a hemispherical protrusion generated on the drive electrode, the detection electrode, or the like. The presence / absence of “hillock” causes variations in characteristics between products and causes a short circuit between electrodes. The presence or absence of “hillock” can be confirmed by observing with an optical microscope.

  In order to adjust the internal pressure in the sealing space and prevent “sticking” and “hillock”, the driving electrode 151a and the detection electrodes 151b to 151e include the first conductive material described above. Using a two-layer structure (first conductive layer, second conductive layer) having a conductive layer as a lower layer and a second conductive layer containing a second conductive material having a melting point higher than that of the first conductive material as an upper layer. did.

  Here, FIG. 6 shows an example in which the driving electrode 151a and the detection electrodes 151b and 151d shown in FIG. In FIG. 6, the drive electrode 151a and the detection electrodes 151b and 151d are formed by lower layer electrodes 151a1, 151b1, 151d1 (first conductive layer) and upper layer electrodes 151a2, 151b2, 151d2 (second conductive layer). The As the first conductive material for forming the lower layer electrodes 151a1, 151b1, and 151d1 and the second conductive material for forming the upper layer electrodes 151a2, 151b2, and 151d2, the following combinations are preferable.

(1) The first conductive material of the lower layer electrode: Al, Al—Cu, Al—Nd, Al—Si, Al—Si—Cu, Pt, Au, etc. These first conductive materials are as follows: It is preferable to have various conditions.
-Low resistance is preferable (influence on the sensitivity of the physical quantity sensor 100).
It is preferable that the electrical connection between the wirings L2 and L7 to L10 and the back surfaces of the block lower layer portions 134b and 134g to 134j is good. Specifically, it is preferable to make ohmic connection (or low resistance connection that can be regarded as ohmic connection) to the back surface (semiconductor such as silicon) of the block lower layer part 134b or the like at the temperature of anodic bonding (200 ° C. to 400 ° C.).
-It is preferable that it has the softness | flexibility of the grade collapsed by joining pressure under the temperature of anodic bonding. The wirings L2, L7 to L10 are sandwiched between the back surface of the block lower layer part 134b and the second substrate 150 and connected to the back surface of the block lower layer part 134b and the like. At this time, it is preferable that the wirings L2 and L7 to L10 are deformed so that the entire back surface of the block lower layer part 134b and the like and the entire connection location of the wirings L2 and L7 to L10 are in uniform contact. By deforming the lower layer electrode, the reliability of the electrical connection between the wirings L2, L7 to L10 and the block lower layer parts 134b and 134g to 134j is increased.

(2) Second conductive material of upper layer electrode: Cr, W, TiN, Mo, Ta, etc. These first conductive materials preferably have the following conditions.
-It is desirable that the melting point be higher than that of the first conductive material. It is desirable that it is hard and does not soften at the temperature of anodic bonding (˜500 ° C.). Thereby, fusion with the lower surfaces of the weight portions 132a to 132e is limited.
-It is desirable not to form a compound (for example, silicide) with the material of the lower surface of the weight parts 132a-132e at the temperature of anodic bonding.
・ It is desirable to suppress the “hillock” mentioned above.
-Low resistance is desirable. However, this requirement is low compared to the lower layer electrode.
-It is preferable that adhesiveness with a lower layer electrode is favorable.

(3) Weight parts 132a to 132e: Semiconductor material such as silicon. It is desirable not to make a compound with the material of the upper electrode at the temperature of anodic bonding.

  The first conductive material: Al, Al—Nd, Al—Si, Pt in consideration of the conditions of the first conductive material of the lower electrode and the conditions of the second conductive material of the upper electrode. , Au and the second conductive material: Cr, W, TiN, Mo, Ta were evaluated, and the following results were obtained.

  “Sticking” is less when the lower surface and the upper layer electrode of the weight parts 132a to 132e are made of a conductive material (Cr, W, TiN) having a melting point higher than that of the first conductive material of the semiconductor (Si) and the lower layer electrode. It was. Regarding “hillock”, the “hillock” of the lower layer electrode could be suppressed by covering the lower layer electrode of Al with the upper layer electrode of the metal-containing material having a melting point higher than that of Al. When the lower electrode is Au or Cr, “hillock” does not occur, so the material of the upper electrode does not contribute to the suppression of “hillock”. The electrical connection was generally good except when the lower layer electrode was Cr.

  From the above results, when the two-layer structure (first conductive layer, second conductive layer) is used for the driving electrode 151a and the detection electrodes 151b to 151e, the semiconductor substrate W and the second substrate 150 are anodic bonded. The occurrence of “sticking” and “hillock” can be suppressed.

<Gas release phenomenon of electrode>
In the physical quantity sensor 100 of the present embodiment, the getter 201 is disposed on the upper surface of the second substrate 150 in order to improve the degree of vacuum in the sealed space. The getter 201 is made of a metal such as aluminum Al. After the anodic bonding of the second substrate 150 to the semiconductor substrate W, the getter 201 is activated by an annealing process at a predetermined annealing temperature (for example, 415 ° C. or higher), and oxygen molecules or water molecules remaining in the sealed space It absorbs by chemical reaction and improves the degree of vacuum in the sealed space. A two-layer structure is formed by sputtering using an Al-based material (Al, Al—Nd, Al—Si) as the first conductive material of the lower electrode and Cr as the second conductive material of the upper electrode. It has been found that when the electrode is formed, the gas mixed in the laminated electrode film is released during the sputtering process in the annealing process after anodic bonding.

  The drive electrode 151a and the detection electrodes 151b and 151d shown in FIG. 6 use, for example, an Al-based material (Al, Al—Nd, Al—Si) as the first conductive material of the lower layer electrodes 151a1, 151b1 and 151d1. The upper layer electrodes 151a2, 151b2, 151d2 can be formed by sputtering using Cr as the second conductive material. In this case, a rare gas (for example, Ar gas) used in the sputtering process is mixed in the electrode films of the lower layer electrodes 151a1, 151b1, 151d1 and the upper layer electrodes 151a2, 151b2, 151d2. It has been found that the Ar gas mixed in the electrode film is not released at the temperature at the time of anodic bonding, but is released at an annealing temperature (415 ° C. or higher) at the annealing process for activating the getter after sealing. did.

The gas emission phenomenon described above will be described with reference to a graph shown in FIG. 9 in which an Al—Nd single layer electrode and a Cr / Al—Nd laminated electrode are compared. FIG. 9 is a graph showing the gas release characteristics of the single-layer electrode and the multilayer electrode in correspondence with the heating temperature (° C.) (horizontal axis) and the Ar gas emission intensity (au) (vertical axis). . In this case, the area of each sample of the single layer electrode and the laminated electrode is 1 cm ² . In FIG. 9, the solid line shows the gas release phenomenon of the laminated electrode, and the dotted line shows the gas release phenomenon of the single-layer electrode. In addition, Ar gas used in the sputtering process is mixed in the electrode film, for example, 0.01 to 1 at%. In FIG. 9, the temperature at the time of anodic bonding between the second substrate 150 and the semiconductor substrate W is 360 ° C.

  In FIG. 9, Ar gas mixed into the Al—Nd single layer electrode by the sputtering process is released when the heating temperature reaches about 470 ° C. Since this heating temperature is higher than the above-described annealing temperature (420 ° C.), the gas release phenomenon does not occur in the Al—Nd single layer electrode. In FIG. 9, Ar gas mixed into the Cr / Al—Nd laminated electrode by the sputtering process is released when the heating temperature reaches about 415 ° C. Since this heating temperature is close to the above-described annealing temperature (420 ° C.), a gas release phenomenon occurs in the Cr / Al—Nd laminated electrode. That is, the Cr / Al—Nd multilayer electrode does not generate a gas release phenomenon at the temperature at the time of anodic bonding (360 ° C.), and does not release the gas at the annealing temperature (420 ° C.) at the annealing process for activating the getter 201. It was found to occur. Further, in the gas release characteristics of the Cr / Al—Nd laminated electrode shown in FIG. 9, gas release starts when the heating temperature reaches about 420 ° C., and then the gas release amount becomes substantially constant even when the heating temperature is raised. There was found.

  Due to the above gas release phenomenon, in the physical quantity sensor 100, oxygen molecules or water molecules remaining in the sealing space are absorbed by the chemical reaction of the getter 201 during the annealing process for activating the getter 201 after sealing. It was found that Ar gas released from the Cr / Al—Nd laminated electrode is not absorbed by the getter 201. That is, it has been found that the internal pressure in the sealed space is increased by the Ar gas released from the laminated electrode. The physical quantity sensor 100 of the present embodiment is characterized in that the internal pressure in the sealed space is controlled by utilizing the gas release phenomenon caused by the laminated electrode.

<Internal pressure control method in sealed space>
As described above with reference to FIG. 20, the resonance frequency width of the resonance spectrum of the weight portion of the physical quantity sensor (half-value width shown in FIG. 20) depends on the internal pressure in the sealed space. If it is narrow, the performance is not good, and it is necessary to adjust the resonance frequency width of the resonance spectrum corresponding to the detection range of acceleration and angular velocity set according to the specification. For this reason, in the physical quantity sensor 100 of the present embodiment, the internal pressure in the sealed space is controlled using the gas release characteristics of the Cr / Al—Nd laminated electrode shown in FIG. Is adjusted to a desired resonance frequency width.

Here, Formula (1) for obtaining the internal pressure is shown below.
P = Nk _B T / V = (vSk _B T) t / V (1)
In Equation (1), V: internal volume, S: Ar gas release area, v: Ar gas release rate, t: heat treatment time, T: product use temperature, N = vSt: number of Ar molecules inside, k _B : Boltzmann constant. Nk _B T / V is an ideal gas equation of state.

  In the physical quantity sensor 100 of the present embodiment, the area S of the conductive material (for example, Cr / Al—Nd) forming the laminated electrode, its annealing temperature T, and annealing based on the mathematical formula (1) for obtaining the internal pressure. By adjusting the time t, the amount of Ar gas released into the sealed space can be controlled to control the internal pressure of the sealed space. In particular, considering the gas release characteristics of the Cr / Al—Nd multilayer electrode shown in FIG. 9, the amount of Ar gas released is constant at the annealing temperature (420 ° C.), so the area S and the heat treatment time t are mainly used. By controlling, the internal pressure of the sealed space can be easily set to a desired pressure. The gas discharge temperature of the laminated electrode is desirably close to the annealing temperature set when the material used as the getter 202 is activated, and is higher than the heating temperature (first heating temperature) at the time of anodic bonding ( The second heating temperature is desirable.

  In the example shown in FIG. 6, the driving electrode 151a and the detection electrodes 151b and 151d have a two-layer structure in which Cr / Al—Nd is laminated, and the area S, the annealing temperature T, and the annealing time t are adjusted. The internal pressure in the sealed space can be controlled by controlling the amount of Ar gas released into the sealed space. Further, as a part of the drive electrode 151a and the detection electrodes 151b to 151e, for example, as shown in FIG. 7, only the detection electrode 151d has a two-layer structure in which Cr / Al—Nd is stacked (in the drawing, As the lower layer electrode 151d1 and the upper layer electrode 151d2), the area S, the annealing temperature T, and the annealing time t may be adjusted.

  Also, as shown in FIG. 3B, a dummy electrode 202 is disposed on the upper surface of the second substrate 150, and as shown in FIG. 8, a two-layer structure in which Cr / Al—Nd is laminated (the lower layer in the figure). As the electrode 202a and the upper layer electrode 202b), the area S, the annealing temperature T, and the annealing time t may be adjusted. Further, as shown in FIG. 10, the driving electrode 151a, the detection electrodes 151b and 151d, and the dummy electrode 202 have a two-layer structure in which Cr / Al—Nd is laminated, and these areas S and the annealing temperature T The annealing time t may be adjusted. Further, as shown in FIG. 11, for example, a part of the wiring L10 is formed as a dummy electrode 210 having a two-layer structure in which Cr / Al—Nd is laminated. The area S of the dummy electrode 210 and its annealing temperature T And the annealing time t may be adjusted. Also, as shown in FIG. 12, the dummy electrode and the dummy wiring 211 connected to the wiring L10 are formed by stacking Cr / Al—Nd, the area S of the dummy electrode and the dummy wiring 211, and the annealing temperature T thereof. And the annealing time t may be adjusted.

  In addition, as a method for verifying that Ar gas is released from a laminated electrode having a two-layer structure in which Cr / Al—Nd is laminated, the sheet resistance value of the Cr / Al—Nd laminated electrode is measured before and after annealing. There is a way. FIG. 13 shows an example in which the sheet resistance value of the Cr / Al—Nd laminated electrode was measured before and after the annealing treatment. In this case, the thickness of the Al—Nd layer is set to 300 nm, the thickness of the Cr layer is set to 100 nm, and the annealing temperature is set to 420 ° C. The sheet resistance value before annealing is about 0.4 (Ω / sq), and the sheet resistance value after annealing is about 3.1 (Ω / sq). This change in the sheet resistance value is peculiar to the gas release phenomenon, and is a result different from the conventional annealing process for reducing the sheet resistance value. That is, the characteristics of the gas release phenomenon in the exemplified Cr / Al—Nd laminated electrode are preferably close to the annealing temperature set when activating the material used as the getter 202 as described above, and the heating during anodic bonding It is desirable that the heating temperature (second heating temperature) is higher than the temperature (first heating temperature) and is lower than the annealing temperature setting at the time of annealing treatment for reducing the sheet resistance value.

  Further, for example, even when a single-layer electrode is formed using a single conductive material such as an Al-based material, the gas release phenomenon occurs. However, in such a single layer electrode, the annealing temperature at which the gas release phenomenon occurs is higher than that in the case of the Cr / Al—Nd laminated electrode, for example, 470 ° C. In this case, since the above-mentioned “hillock” is generated on the surface of the single-layer electrode, it is not desirable to use it as an electrode of the physical quantity sensor 100. When the Cr / Al—Nd laminated electrode is used, “hillock” does not occur because the temperature at which the gas release phenomenon occurs is as low as 420 ° C. as compared with the case where the single layer electrode is used. Accordingly, it is optimal for the physical quantity sensor 100 to use a laminated electrode in which two kinds of conductive materials are laminated as an electrode for setting the sealed space to a desired internal pressure. In the above-described embodiment, the Cr / Al—Nd multilayer electrode is exemplified as the multilayer electrode. However, the present invention is not limited to these conductive materials, and the conductive material has a characteristic of generating a gas release phenomenon under the above-described conditions. Any material can be used. As illustrated above, the case where a part of the driving electrode 151a and the detection electrodes 151b and 151d, the dummy electrode 202, or the dummy wiring 211 has a two-layer structure and the gas release phenomenon is generated by the annealing process has been shown. However, as a result of the occurrence of the gas release phenomenon, there is a possibility that the resistance of the portion having the two-layer structure is increased depending on the conductive material used. Therefore, the portions where the resistance increase is not desirable, such as the moving electrode 151a and the detection electrodes 151b and 151d, use TiN / Al-Nd as a conductive material that does not generate a gas release phenomenon in the annealing process and does not increase the resistance. It is preferable to use Cr / Al—Nd as the conductive material for the dummy electrode 202 and the dummy wiring 211 that utilize the gas release phenomenon.

  14 and 15 show metal micrographs of the Cr / Al—Nd laminated electrode surface before and after the annealing treatment. In this case, the annealing temperature is 420 ° C. As shown in FIG. 15, “hillock” does not occur on the surface of the Cr / Al—Nd laminated electrode after the annealing treatment.

  As described above, in the capacitance type physical quantity sensor 100, the driving electrode 151a, the detection electrodes 151b to 151e formed on the surface of the second substrate 150, a part of these electrodes, or the dummy electrode 202 is provided. Was formed as a laminated electrode in which a lower electrode including an Al-based first conductive material and an upper electrode including a Cr-based second conductive material having a melting point higher than that of the first conductive material were stacked. Since this laminated electrode has a feature of releasing a certain amount of gas (Ar gas) mixed during film formation at an annealing temperature higher than the heating temperature at the time of anodic bonding, by controlling the area, annealing temperature and annealing time, It is easy to set the internal pressure of the sealed space in the physical quantity sensor 100 to a desired pressure. For this reason, it becomes easy to arbitrarily set the resonance vibration frequency width of the resonance spectrum of the weight portions 132a to 132e depending on the internal pressure of the sealed space. As a result, it is possible to easily manufacture the physical quantity sensor 100 in which the resonance spectrum of the weight portion is adjusted to a desired resonance frequency width according to specifications such as detection sensitivity.

  The physical quantity sensor 100 according to the embodiment of the present invention is mounted on a circuit board on which an active element such as an IC is mounted, for example, and is a wiring terminal T (T1 to T11) by a known method and material such as wire bonding connection. By connecting an electronic circuit board or an active element such as an IC, the physical quantity sensor and the electronic circuit can be provided as one electronic component. This electronic component can be distributed in the market by being mounted on a mobile terminal such as a game machine or a mobile phone.

  Hereinafter, a processing circuit that processes displacement signals of acceleration and angular velocity detected by the physical quantity sensor 100 will be described.

<Processing circuit>
A configuration example of each processing circuit that processes displacement signals of acceleration and angular velocity detected by the physical quantity sensor 100 will be described with reference to FIG.

FIG. 16 is a diagram illustrating a circuit configuration of a processing circuit 300 that processes displacement signals of acceleration and angular velocity detected by the physical quantity sensor 100. In FIG. 16, the processing circuit 300 includes a CV converter 301, an amplifier circuit (Amp) 302, and a filter circuit 303.

  The CV converter 301 converts each axial displacement signal (capacitance change) output from the physical quantity sensor 100 according to the applied acceleration and angular velocity into a voltage signal and outputs the voltage signal to the amplifier circuit 302. The amplifier circuit 302 amplifies the voltage signal input from the CV converter 301 with a predetermined amplification factor and outputs the amplified signal to the filter circuit 303. The filter circuit 303 has a filter function that allows a signal component of several kHz or more to pass therethrough. The filter circuit 303 passes a signal component of several kHz or more from the voltage signal amplified by the amplifier circuit 302 and outputs it as an angular velocity detection signal in the X-axis direction and the Y-axis direction. The filter circuit 303 outputs a low-frequency signal component as an acceleration detection signal in the X-axis direction, the Y-axis direction, and the Z-axis direction.

  Next, an example of a semiconductor device in which the physical quantity sensor 100 and the processing circuit 300 are mounted will be described. Note that in this specification, a semiconductor device refers to all devices that can function using semiconductor technology, and electronic devices are also included in the scope of semiconductor devices.

  FIG. 17 is a diagram illustrating an example of a sensor module 400 as a semiconductor device on which the physical quantity sensor 100 and the processing circuit 300 are mounted. 17, in the sensor module 400, a signal processing chip 401 including the processing circuit 300, a memory chip 402, and a sensor chip 403 including the physical quantity sensor 100 are mounted on a substrate 404. Each chip 401, 402, 403 is connected by a bonding wire 405. The memory chip 402 is a memory that stores a control program, parameters, and the like for the signal processing chip 401.

  Providing the sensor module 400 as described above facilitates mounting on a mobile terminal such as a game machine or a mobile phone.

  Next, an example in which the sensor module 400 shown in FIG. 17 is implemented as an electronic device in, for example, a mobile terminal will be described.

  FIG. 18 is a diagram illustrating an example of the portable information terminal 500 in which the sensor module 400 is mounted. In FIG. 18, the portable information terminal 500 includes a display unit 501 and a keyboard unit 502. The sensor module 400 is mounted inside the keyboard unit 502. The portable information terminal 500 has a function of storing various programs therein and executing communication processing, information processing, and the like by the various programs. In this portable information terminal 500, by using the acceleration and angular velocity detected by the sensor module 400 in the application program, for example, it is possible to add a function of detecting the acceleration at the time of dropping and turning off the power. become.

  By mounting the sensor module 400 on the mobile terminal as described above, new functions can be realized, and the convenience and reliability of the mobile terminal can be improved.

It is a disassembled perspective view which shows the state which decomposed | disassembled the physical quantity sensor which concerns on one embodiment of this invention. 1A is a plan view showing the upper surface of a silicon film, FIG. 1B is a cross-sectional view of the silicon substrate as viewed from line BB in FIG. 2A, and FIG. It is sectional drawing of the silicon substrate seen from CC line. It is a figure which shows the structure of a board | substrate, (A) is a top view which shows the lower surface of a 1st board | substrate, (B) is a top view which shows the upper surface of a 2nd board | substrate. It is a figure which shows the manufacturing method of a physical quantity sensor, (A) is sectional drawing which shows the semiconductor substrate before a process, (B) forms a recessed part in the area | region corresponding to a frame part, a weight junction part, and a flexible part in a semiconductor substrate. FIG. 6C is a cross-sectional view showing the step of forming the frame portion, the weight joint portion and the conductive portion on the semiconductor substrate, and FIG. 6D is a cross-sectional view showing the step of bonding the first substrate to the semiconductor substrate. It is. It is a figure which shows the manufacturing method of a physical quantity sensor, (A) is sectional drawing which shows the process of forming a recessed part in the area | region corresponding to a weight part etc. in a semiconductor substrate, (B) is the process of forming a weight part etc. in a semiconductor substrate. FIG. 6C is a cross-sectional view showing a step of bonding the second substrate to the semiconductor substrate. It is sectional drawing which shows the example which formed the drive electrode and detection electrode on a 2nd board | substrate as a laminated electrode. It is sectional drawing which shows the example which formed only the electrode for a detection on a 2nd board | substrate as a laminated electrode. It is sectional drawing which shows the example which provided the dummy electrode formed as a laminated electrode on the 2nd board | substrate. It is a graph which shows each gas discharge | release characteristic of an Al-Nd single layer electrode and a Cr / Al-Nd laminated electrode. It is sectional drawing which shows the example which provided the drive electrode, detection electrode, and dummy electrode which were formed as a laminated electrode on the 2nd board | substrate. It is a top view which shows the example which formed a part of wiring on the 2nd board | substrate as a dummy electrode of a laminated electrode. It is a top view which shows the example which formed the dummy wiring and dummy electrode connected to the wiring on a 2nd board | substrate as a laminated electrode. It is a figure which shows an example of the sheet resistance value before and behind the annealing process of the laminated electrode which a gas discharge phenomenon generate | occur | produces. It is a figure which shows the SEM photograph of the Cr / Al-Nd multilayer electrode surface before annealing treatment. It is a figure which shows the SEM photograph of the Cr / Al-Nd multilayer electrode surface after annealing treatment. It is a figure which shows an example of the angular velocity processing circuit which processes the displacement signal of the angular velocity detected by a physical quantity sensor. It is a figure which shows an example of the sensor module which mounted the physical quantity sensor and the processing circuit. It is a figure which shows an example of the mobile terminal which mounted the sensor module. It is a figure which shows the operation principle of the conventional physical quantity sensor, (A) is a figure which shows the state at the time of drive voltage 5V application, (B) is a figure which shows the state at the time of drive voltage 0V application. It is a figure which shows an example of the vibration frequency in the conventional physical quantity sensor.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 ... Physical quantity sensor, 110 ... Silicon film, 111, 131 ... Frame part, 112a-112e ... Weight junction part, 113a-113d ... Flexible part, 120 ... BOX layer, 130 ... Silicon substrate, 132 (132a-132e) ... Weight part, 140 ... first substrate, 1441, 151a ... driving electrode, 141b to 141e, 151b to 151e ... detection electrode, 150 ... second substrate, 151a1, 151b1, 151d1, 202a ... lower layer electrode, 151a2, 151b2, 151d2, 202b ... upper layer electrode, 160 ... conductive portion, 201 ... getter, 202 ... dummy electrode, 300 ... processing circuit, 400 ... sensor module, 500 ... portable information terminal, L1, L2 ... (connected to driving electrodes) Wiring, L3 to L10 (wiring connected to detection electrode).

Claims (21)

A semiconductor substrate comprising a frame portion, a weight portion disposed inside the frame portion, and a flexible portion connecting the weight portion and the frame portion;
A first substrate bonded to one side of the frame portion;
A second substrate bonded to the other side of the frame portion;
A first electrode provided on the first substrate and facing the weight portion;
A second electrode provided on the second substrate and facing the weight portion;
A gas absorbing member that is disposed in a sealed space sealed by bonding of the first substrate and the second substrate and absorbs the gas inside the sealed space in the semiconductor substrate by being heated ;
A gas pressure adjusting member that is disposed in the sealed space and is heated to release gas into the sealed space to adjust the pressure inside the sealed space ;
Equipped with a,
The gas pressure adjusting member includes a first conductive layer containing a first conductive material in the first electrode, a part of the first electrode, the second electrode, or a part of the second electrode; A physical quantity sensor formed by laminating a second conductive layer laminated on the first conductive layer and containing a second conductive material having a melting point higher than that of the first conductive material . The physical quantity sensor according to claim 1, wherein a temperature at which the pressure adjusting member releases gas is higher than a temperature at which the gas absorbing member absorbs gas. The physical quantity sensor according to claim 1, wherein the gas absorbed by the gas absorbing member is a different type of gas from the gas released by the pressure adjusting member. It said gas pressure adjusting member according to claim 1, characterized in that the first electrode or the first conductive layer as a dummy electrode which is independent from the second electrode and the second conductive layer is formed by stacking Physical quantity sensor. A first wiring provided on the first substrate and electrically connected to the first electrode;
A second wiring provided on the second substrate and electrically connected to the second electrode;
It said gas pressure adjusting member, the first wiring, a part of the first wiring, the second wiring, or a part of the second wiring, and the first conductive layer, and the second conductive layer The physical quantity sensor according to claim 1 , wherein the physical quantity sensor is formed by being laminated. The gas pressure adjusting member is formed by laminating the first conductive layer and the second conductive layer as a dummy electrode connected to the first electrode or the second electrode. The physical quantity sensor according to 1. The gas pressure adjusting member includes the first conductive layer and the second conductive layer as a dummy electrode or a dummy wiring connected to the first electrode, the second electrode, the first wiring, or the second wiring. The physical quantity sensor according to claim 5 , wherein the physical quantity sensor is formed by being laminated. The physical quantity sensor according to any one of claims 1 to 7 second conductive material is Cr, W, Mo, or wherein at least either of Ta,. It said first conductive material is Al, according to Al-Cu, Al-Nd, any one of claims 1 to 7, characterized in that at least one of Al-Si or Al-Si-Cu Physical quantity sensor. Physical quantity sensor according to any one of claims 1 to 7 before crisis body absorbing member, characterized in that provided on the first substrate or the second substrate. A semiconductor substrate is formed with a frame portion, a weight portion disposed inside the frame portion, and a flexible portion connecting the weight portion and the frame portion,
Forming a first electrode and a first wiring electrically connected to the first electrode on a first substrate bonded to one side of the frame portion;
Forming a second electrode and a second wiring electrically connected to the second electrode on a second substrate bonded to the other side of the frame portion;
A gas absorbing member that absorbs the gas inside the sealed space in the semiconductor substrate by being heated on the first substrate or the second substrate, and a gas in the sealed space by being heated. release to form a gas pressure adjusting member for adjusting the pressure,
Bonding the first substrate and one side of the frame portion with the weight portion and the first electrode facing each other;
Bonding the second substrate and the other side of the frame portion with the weight portion and the second electrode facing each other;
By the first heat treatment, the gas absorbing member absorbs the gas inside the sealed space in the semiconductor substrate sealed by joining the first substrate and the second substrate,
The second heat treatment includes adjusting the pressure inside the sealed space by releasing the gas mixed during the formation of the gas pressure adjusting member into the sealed space ;
When forming the first electrode, forming the first wiring, forming the second electrode, or forming the second wiring, a first conductive layer containing a first conductive material; A method of manufacturing a physical quantity sensor, wherein the gas pressure adjusting member is formed by laminating a second conductive layer containing a second conductive material having a melting point higher than that of the first conductive material . The method of manufacturing a physical quantity sensor according to claim 11, wherein the second heat treatment has a heating temperature higher than that of the first heat treatment. The method of manufacturing a physical quantity sensor according to claim 11, wherein the gas absorbed by the gas absorbing member is a different type of gas from the gas released by the pressure adjusting member. The first substrate or the second substrate is heated by a first heating temperature and bonded to the frame part,
Set area during the formation of the gas pressure adjusting member to set the heating temperature of the second heat treatment in the above first heating temperature second heating temperature, pressure heat depending on the area and the heating temperature method of manufacturing a physical quantity sensor according to claim 1 1 Symbol mounting, characterized in that to set the time. A semiconductor substrate comprising a frame portion, a weight portion disposed inside the frame portion, and a flexible portion connecting the weight portion and the frame portion;
A first substrate bonded to one side of the frame portion;
A second substrate bonded to the other side of the frame portion;
A first electrode provided on the first substrate and facing the weight portion;
A second electrode provided on the second substrate and facing the weight portion;
The gas inside the sealed space in the semiconductor substrate, which is disposed on the first substrate or the second substrate and is sealed by bonding of the first substrate and the second substrate by being heated, is absorbed. A gas absorbing member;
In a physical quantity sensor comprising: a gas pressure adjusting member that is disposed in the sealed space and is heated to release gas into the sealed space to adjust the pressure.
By the first heat treatment, the gas absorbing member absorbs the gas inside the sealed space in the semiconductor substrate sealed by joining the first substrate and the second substrate,
The second heat treatment includes adjusting the pressure inside the sealed space by releasing the gas mixed during the formation of the gas pressure adjusting member into the sealed space ;
The gas pressure adjusting member is formed by laminating a first conductive layer including a first conductive material and a second conductive layer including a second conductive material having a melting point higher than that of the first conductive material. And
A physical quantity that adjusts the pressure inside the sealed space by setting an area when forming the gas pressure adjusting member and controlling a heating temperature and a heating time of the second heat treatment according to the area. Sensor internal pressure control method. 16. The internal pressure control method for a physical quantity sensor according to claim 15, wherein the temperature at which the pressure adjusting member releases gas is higher than the temperature at which the gas absorbing member absorbs gas . The internal pressure control method for a physical quantity sensor according to claim 15, wherein the gas absorbed by the gas absorbing member is a different type of gas from the gas released by the pressure adjusting member . The gas pressure adjusting member has the first conductive layer and the second conductive layer on the first electrode, a part of the first electrode, the second electrode, or a part of the second electrode. inner pressure controlling method for the physical quantity sensor of claim 1 5, wherein the are laminated is formed. Forming a first wiring electrically connected to the first electrode on the first substrate; forming a second wiring electrically connected to the second electrode on the second substrate; the first wiring, a part of the first wiring, the second wiring, or, with the first conductive layer on a part of the second wiring, the Rukoto forming shape by laminating a second conductive layer inner pressure controlling method for the physical quantity sensor of claim 1 5, wherein. The first substrate or the second substrate is heated by a first heating temperature and bonded to the frame part,
The heating temperature of the second heat treatment, the internal pressure of the physical quantity sensor according to any one of claims 15 to 1 9, characterized in that set in the above first heating temperature second heating temperature Control method. The physical quantity sensor according to any one of claims 1 to 10 ,
A processing circuit for processing a physical quantity detection signal detected by the physical quantity sensor;
A semiconductor device comprising: