JPH09321317A - Force converting element and manufacture thereof - Google Patents

Abstract

PROBLEM TO BE SOLVED: To eliminate troublesome works applied to a force transmitting part wafer to avoid contacting electrodes to the wafer in a production process of a force converting element. SOLUTION: Protrusions 31b for forming gauges, etc., and those 31c for forming electrodes 29 are formed on the surface of a Si wafer 31 by dry etching. The protrusions 31c with the electrodes 29 are lower than those 31b with a step difference. This serves for contacting a glass wafer 40 with the tops of the protrusions 31b, without striking against the electrodes 29. After bonding both wafers 31, 40, the surface of the wafer 40 is diced to form matrix grooves 40a, a less adhesive tape is pasted and peeled to remove undesired parts of the wafer 40, and the Si wafer 31 is cut in elements.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is used as a force sensor or a pressure sensor, for example, and is a force converter for converting mechanical energy such as added compression force into electric energy and outputting an electric signal according to the compression force. The present invention relates to a method for manufacturing an element and a force conversion element.

[0002]

2. Description of the Related Art A force conversion element of this type is disclosed, for example, in Japanese Patent Publication No.
-82847, Japanese Patent Publication No. 7-14069, and Japanese Patent Application Laid-Open No. 6-34456. The force conversion element includes a base body made of single crystal silicon having a piezo effect, a force transmission part made of, for example, a glass block bonded on the base body, and electrodes serving as input / output terminals formed on the base body. . The resistance value of the gauge portion, which is the joint portion of the force transmission portion, changes according to the force (pressure) applied to the force transmission portion by the piezo effect. A constant current flows through the two input terminals, and a potential difference according to the gauge resistance value that changes according to the compressive force applied is generated between the two output terminals, and from the signal value (voltage value) The magnitude of the applied compression force is detected.

For example, in Japanese Examined Patent Publication No. 7-14069 and Japanese Patent Laid-Open No. 6-34456, a large number of force conversion elements are formed on a silicon wafer, and each element is individually cut to separate the force conversion elements. A method of manufacturing a force conversion element intended for mass production is disclosed. According to this manufacturing method, electrodes for a plurality of force conversion elements are formed on the surface of a silicon wafer, a glass wafer for a force transmitting portion is bonded to the surface of the silicon wafer, and the bonded bodies of both wafers are individually separated. By separating each of the force conversion elements, a large number of force conversion elements can be mass-produced at once.

The steps subsequent to the bonding of the glass wafers were performed according to the procedure shown in FIG. As shown in FIG. 3A, in the silicon wafer 71 that has been subjected to the electrode forming step, the electrode 7 has a larger diameter than the gauge portion 72 to which the force transmitting portion is joined.
3 is in a state of protruding upward by the thickness thereof. Therefore, the escape portion 74a is formed on the glass wafer 74 at a portion facing the electrode 73 so that the glass wafer 74 contacts the gauge portion 72. Further, in order to easily remove an unnecessary portion of the glass wafer 74 (a portion which is not used as the force transmitting portion 75) at the time of separation after joining, the glass wafer 7
The notch (matrix groove) 74b was formed in No. 4.
Then, the glass wafer 74 is positioned on the silicon wafer 71 with the pattern of the electrodes 73 and the escape portions 74a and the matrix grooves 74b, and then the two wafers 71 and 74 are bonded by an anodic bonding method or the like.

If the glass wafer 74 is, for example, crystallized glass, it is white and opaque, so that it is difficult to align the silicon wafer 71 and the glass wafer 74. Therefore,
Positioning is performed by abutting both wafers 71 and 74 at predetermined positions on the jig using a ceramic jig.
The silicon wafer 71 and the glass wafer 74 have been previously subjected to the dicing process for vertical alignment and positioning.

After the glass wafer 74 is bonded, as shown in FIG. 9B, unnecessary portions of the glass wafer 74 are cut off by dicing. At this time, the force transmitting portion 75 is formed by removing the unnecessary portion along the matrix groove 74b. Then, as shown in FIG. 9C, the silicon wafer 71 is then diced into individual substrates 76 and separated into force conversion elements 77.

[0007]

However, the relief portion 74a formed on the glass wafer 74 by dicing or the like.
It took a lot of man-hours to produce. That is, the dicing process of the glass wafer 74 is chipped (broken).
It is carried out at a very low speed to prevent it. The cutting speed is, for example, 10 mm / sec. Or more for silicon, while it is about 2 mm / sec. Or less for glass, which is 1/5 or less. Therefore, for example, a 3 inch glass wafer 74
It took about half a day to process by dicing.

[0008] Further, by abutting against a jig made of ceramic, 2
According to the positioning method for positioning the wafers 71 and 74, a relatively large alignment error (for example, about 50 μm) occurs due to variations in abutting against the jig.

Furthermore, when the unnecessary portion of the glass wafer 74 is removed by dicing, the broken pieces of the glass that have been cut and dropped or scattered may hit the electrode 73 and damage the electrode 73. If the electrode 73 is damaged, there is a possibility that wire bond failure may occur. Therefore, it has been a cause of lowering the yield of the force conversion element.

The present invention has been made to solve the above problems, and a first object thereof is that it is not necessary to perform troublesome processing for avoiding contact with electrodes on a wafer for force transmitting portion. It is to provide a method of manufacturing a force conversion element and a force conversion element. A second object is to eliminate the fear of damaging the electrodes during the work of removing the unnecessary portion of the force transmitting portion wafer.

[0011]

In order to solve the above problems, according to the invention of claim 1, electrodes for a plurality of force conversion elements are formed on a semiconductor wafer, and a force transmitting portion is formed on the semiconductor wafer. In the method for manufacturing a force transducing element in which individual force transducing elements are separated after joining the wafers for use in the semiconductor wafer, the electrodes are not protruded from a gauge portion which is a joint surface of the force transmitting wafer. A step forming step of forming a step portion for forming an electrode, an electrode forming step of forming an electrode on the step portion formed in the step forming step,
A wafer bonding step of bonding the force transmitting portion wafer to the semiconductor wafer after the electrode formation on the surface of the gauge portion side;
And a wafer removing step of removing an unnecessary portion of the force transmitting portion wafer located above the electrode.

According to a second aspect of the invention, in the method of manufacturing the force conversion element according to the first aspect, in the wafer removing step, the unnecessary portion of the force transmitting wafer is cut and then the unnecessary portion is removed. The portion was adapted to be detackified by an adhering means.

According to a third aspect of the present invention, the force transducing element is manufactured by the method for producing a force transducing element according to the first or second aspect, and an electrode is provided on the stepped portion formed on the substrate. By being formed, the electrode does not protrude from the gauge portion.

(Operation) According to the invention described in claim 1,
In the step forming step, a step portion is formed on the semiconductor wafer for forming the electrode so as not to protrude from the gauge portion, and in the electrode forming step, the electrode is formed on the step portion formed in the step forming step. Then, in the wafer joining step, the force transmitting portion wafer is joined to the semiconductor wafer after the electrodes are formed on the surface of the gauge portion side. At this time, since the electrode does not protrude from the gauge portion, the force transmitting wafer is in contact with the gauge portion without being hindered by the electrode. Therefore, it is not necessary to subject the wafer for force transmitting portion to extra processing such as a relief portion. In addition, since it is not necessary to align the processed portion such as the relief portion with a predetermined portion such as an electrode on the semiconductor wafer side, a certain degree of error is allowed in the alignment of the force transmitting portion wafer with respect to the semiconductor wafer. After joining the two wafers, an unnecessary portion above the electrodes of the force transmitting wafer is removed in the wafer removing step.

According to the second aspect of the present invention, in the wafer removing step, after the unnecessary portion of the force transmission wafer is cut, the unnecessary portion is adhesively removed by the adhesive means. Therefore, there is no need to worry that the unnecessary portion may hit the electrode and damage the electrode during removal.

In the invention according to claim 3, since the electrode formed in the step portion formed on the base of the force conversion element does not protrude from the gauge portion, the invention according to claim 1 is produced at the time of manufacturing the electrode. The same effect as can be obtained. Further, if the manufacturing method in which the unnecessary portion is cut after the wafer for the force transmitting portion is bonded and the unnecessary portion is adhesively removed by the adhesive means, the same effect as that of the invention according to claim 2 can be obtained.

[0017]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to FIGS. FIG. 1 is a partially cutaway plan view of a force conversion element 11 according to this embodiment. The base 12 constituting the force conversion element 11 includes a pair of input electrode portions 13 and 14 and output electrode portions 15 and 16, four gauge portions 17 to 20, a protective portion 21 to 24, and a stress adjusting portion 25. , 26 are each formed in a convex shape. At the center of the surface of the base body 2, a glass block 2 serving as a force transmitting portion is provided.
7 is joined while being supported by the respective parts 17 to 26.

The substrate 12 is made of N-type <110> plane-oriented single crystal silicon. The upward direction in FIG. 1 corresponds to the crystal axis <11 bar 0> direction of silicon, and the rightward direction is the same as <001> direction. It is equivalent. P-type diffusion layers 28 (illustrated in FIG. 2) in which boron (B) is implanted in the upper layers of the respective portions 13 to 26 which are shaded in FIG. 1 (however, the portions covered by the glass block 27 are omitted). Are formed.

As shown in FIGS. 1 and 2, the gauge portions 17 to 20 projecting from the surface of the substrate 12 in a substantially square ring shape have electrode portions 13 to 16 at the four corner portions of the substantially square ring shape. It is connected with. An electrode 29 made of aluminum is bonded onto each of the electrode portions 13 to 16, and each electrode 29 and each of the gauge portions 17 to 20 are in a connected state via the P-type diffusion layer 28.

The protective portions 21 to 24 and the stress adjusting portions 25 and 26 are the gauge portions 17 to 20 and the electrode portions 13.
16 to 16 are divided and protruded, and the P-type diffusion layer 28 formed on the upper portion is also divided to the gauge portions 17 to 20 and each electrode 29. The surface of the base 12 is covered with an insulating oxide film 30 except for the electrodes 29.

The force conversion element 11 includes the gauge portions 17 to 20.
The gauge resistance of 1 forms a Wheatstone bridge circuit as shown in FIG. The rate of change of the gauge resistance according to the compressive force applied to each of the gauge portions 17 to 20 is determined by the piezoresistance coefficient of the P-type diffusion layer 28 of each of the gauge portions 17 to 20. The piezoresistive coefficient has anisotropy due to the crystal orientation, and the gauge portion 1 extending in the <001> orientation is used.
Piezoresistance coefficients πa and πc of 7 and 19 are the same,
Piezoresistive coefficients πb and πd of the gauge portions 18 and 20 extending in the <11 bar 0> direction are both equal (πa = πc ≠
πb = πd).

When a compressive force is applied to each of the gauge parts 17 to 20 through the glass block 27 in the state where a constant current I is applied between the input electrode parts 13 and 14, the <001> direction according to the compressive force. Gage resistance values (R) of the gage parts 17 and 19 of <11 bar 0> direction,
The gauge resistance value (R + Δr) of 20 has a different value according to each piezo resistance coefficient. Therefore, a potential difference occurs between the output electrode portions 15 and 16 according to the compression force, and the magnitude of the compression force is detected by measuring this potential difference.

As shown in FIG. 2, each electrode portion 13-16
Are projected one step lower than the upper surfaces of the gauge portions 17 to 20, and the electrodes 29 formed on the electrode portions 13 to 16 do not protrude from the upper surfaces of the respective portions 17 to 26 which are the bonding surfaces of the glass block 27. It is located low on. In the present embodiment, the thickness of the electrode 29 is taken into consideration so that a distance of about 3 μm is ensured between the upper surface of the electrode 29 and the lower surface of the glass block 27 (that is, the upper surface of each portion 17 to 26). The height of the parts 13 to 16 is set. The P-type diffusion layer 28 has a thickness of about 1 μm, and the insulating oxide film 30 has a thickness of about 1 μm.
The thickness of the glass block 27 is 00 nm and the thickness of the glass block 27 is about 1 mm.

Next, a method of manufacturing the force conversion element 11 will be described with reference to FIGS. The cross section of the silicon wafer 31 as a semiconductor wafer shown in FIGS.
5 corresponds to the cross section of the force conversion element 11 of FIG.
16 is the same cross section cut so as not to pass through 16.

The silicon wafer 31 shown in FIG.
The wafer is, for example, a 5-inch wafer made of N-type <110> plane-oriented single crystal silicon and has a thickness of about 600 μm. As shown in FIG. 3B, first, the silicon wafer 3
A silicon oxide film (SiO2 film) 32 is formed on the surface (upper surface) of 1 by CVD (Chemical Vapor Deposition), and then a resist 33 is applied (spin coating) on the surface of the silicon wafer 31 for patterning. The resist 33 is patterned so as to cover the entire area to be covered with the glass block 27.

Next, as shown in FIG. 3C, the surface of the silicon wafer 31 is dry-etched to form a pattern of the resist 33. First, SiO2 film 3
2 is removed by dry etching into a patterned shape, the resist 33 is removed by ashing (ashing), and then the silicon wafer 31 is etched to a predetermined depth (about 3 to 4 .mu.m) using the remaining SiO2 film 32 as a mask. By this dry etching, a step portion 31a is formed on the silicon wafer 31. Dry etching was performed in two steps because the resist 33 has a silicon of about 3 μm.
This is because it cannot withstand some degree of dry etching.

Next, the SiO2 film 32 is once removed, and then, as shown in FIG.
As shown in (d), the silicon wafer 31 is subjected to a thermal oxidation treatment to perform a pretreatment for forming a thermal oxide film (SiO2 film) 34 on the surface thereof. After that, an implant process is performed to implant boron (B) ions into the surface of the silicon wafer 31, and a heat treatment for diffusion is performed to electrically activate the boron. Thus, the P type diffusion layer 28 is formed.

Next, after the thermal oxide film 34 is once removed, CV is formed on the surface of the silicon wafer 31 as shown in FIG.
A new silicon oxide film (SiO2) 35 is formed by D. Then, after applying (spin coating) the resist 36, patterning for forming the gauge portions 17 to 20, the protecting portions 21 to 24, the stress adjusting portions 25 and 26, and the input / output electrode portions 13 to 16 is performed.

Then, as shown in FIG. 4B, the surface of the silicon wafer 31 is dug by dry etching.
First, the patterned pattern of the resist 36 is formed on the SiO2 film 35.
Is dry-etched and the resist 36 is once removed by ashing, and then the silicon wafer 31 is etched to a predetermined depth (for example, about 3 .mu.m) using the remaining SiO2 film 35 as a mask. By this dry etching, the gauge parts 17 to 20, the protection parts 21 to 24, and the stress adjusting part 2
5, 31 and the convex portion 31b and the input / output electrode portions 13 to 16
And a convex portion 31 c that will be formed on the silicon wafer 31.

Next, the SiO2 film 35 is once removed, and then, as shown in FIG.
As shown in (c), C is formed on the entire surface of the silicon wafer 31.
A new insulating oxide film (SiO2 film) 37 is formed by VD. Then, a part of the insulating oxide film 37 on the convex portion 31c is removed by wet etching to form a contact portion 38 for making contact with the electrode 29. After forming the contact portion 38, the resist used for wet etching is removed by ashing.

Subsequently, as shown in FIG. 4D, an electrode 29 made of aluminum is bonded to the upper surface of the convex portion 31c after the contact portion 38 is formed. First, silicon wafer 3
An aluminum film (Al) 39 is formed on the surface of 1 by a sputtering method or a vacuum deposition method, and then a resist (not shown) is applied to perform patterning for forming an electrode 29. After that, the silicon wafer 31 is wet-etched to remove the unnecessary aluminum film 39 to remove the electrode 29.
Is formed. Then, after removing the resist by ashing, a sintering process is performed in order to surely make a contact between the aluminum film 39 (electrode 29) and the silicon wafer 31.

When the process of bonding the electrodes 29 is completed, the process on the side of the silicon wafer 31 is completed, and then the process of bonding the glass wafer 40 as the force transmitting portion wafer which becomes the glass block 27 is performed. Here, the upper surfaces of the convex portions 31b such as the gauge portions 17 to 20 which are the bonding surfaces with the glass wafer 40 protrude upward from the upper surfaces of the electrodes 29, and a distance of about 3 μm is secured between them. There is.

Next, as shown in FIG. 5A, a glass wafer 40 of 3 inches is bonded to the surface of the silicon wafer 31 (5 inches). The glass wafer 40 is made of crystallized glass having a thermal expansion coefficient matched to that of the silicon wafer 31. First, the 3-inch glass wafer 40 is placed on the 5-inch silicon wafer 31 so as to be superposed. The glass wafer 40 is placed on the silicon wafer 31 so as to cover the entire pattern formed on the surface of the silicon wafer 31 in a state where the glass wafer 40 does not hit the electrodes 29 but abuts on the convex portions 31b forming the respective portions 17 to 26 such as the gauge portion. Placed. In this state, a space of about 3 μm is secured between the lower surface of the glass wafer 40 and the electrode 29.

Then, the silicon wafer 31 and the glass wafer 40 are bonded by the anodic bonding method. At the time of this anodic bonding, a voltage of about 600 to 1000 V is applied between the two wafers 31 and 40, but since a space of about 3 μm is secured between the glass wafer 40 and the electrode 29, the electrode at the time of this bonding. There is no leakage current through 29. The glass wafer 40 is bonded on the upper surface of the convex portion 31b through the insulating oxide film 37 (FIG. 5A).

Next, as shown in FIG. 5B, a half-cut corresponding to the size and shape of the glass block 27 is formed by dicing on the surface (upper surface) of the glass wafer 40 to form a matrix groove 40a. The alignment of the dicing blade when forming the matrix groove 40a is performed according to the pattern while actually observing the pattern of the silicon wafer 31 which is projected from the 3-inch glass wafer 40, and therefore the matrix groove 40a is formed by silicon. It is formed with high positional accuracy with respect to the pattern on the wafer 31 side.

Next, after attaching a tape as an adhesive means having a relatively weak adhesive force to the upper surface of the glass wafer 40, the tape is peeled off to remove unnecessary portions of the glass wafer 40 by the adhesive force of the tape. Yes (Fig. 5
(C)). At this time, if necessary, the adhered tape is pressed by a roller to break the unnecessary portion. Glass wafer 40
Since the unnecessary portion of is adhered to the tape, the glass fragments to be removed do not hit the electrode 29 and damage it. Further, the unnecessary portion of the glass wafer 40 is broken and removed along the matrix groove 40a formed with high positional accuracy, so that the glass block 27 is correctly arranged at a predetermined position with respect to the pattern on the silicon wafer 31 side. As the tape for removing the adhesive, it is preferable to use a re-peelable dicing tape having a relatively weak adhesive force or a UV curable dicing tape whose adhesive force is reduced by ultraviolet irradiation (about 10 to 1/100). .

Through the above steps, a large number of force conversion elements 11 are manufactured in a matrix on the silicon wafer 31.
Then, as shown in FIG. 5D, the silicon wafer 31 is separated into individual elements by dicing,
A large number of force conversion elements 11 are obtained.

As described above in detail, according to the present embodiment,
The effects listed below can be obtained. (A) A step portion 31 a is formed on the silicon wafer 31, and a convex portion 31 b such as the gauge portions 17 to 20 joined to the glass wafer 40 is formed so as to protrude above the electrode 29.
Therefore, in order to avoid contact with the electrode 29 during bonding to the silicon wafer 31, it is possible to eliminate troublesome processing such as a relief portion provided on the glass wafer 40. here,
Although the number of steps for forming the step portion 31a is increased, the number of steps is significantly reduced as compared with the step for forming the relief portion.

(B) Matrix grooves 4 on the glass wafer 40
0a is formed in advance, and the unnecessary portion of the glass wafer 40 is adhesively removed by using a tape having a relatively weak adhesive force, so that the unnecessary portion of the glass wafer 40 is prevented from being dropped or scattered and damaging the electrode 29. be able to. Therefore, it contributes to the improvement of the yield of the force conversion element 11, and eventually to the reduction of wire bond defects.

(C) The alignment of the dicing blade when forming the matrix groove 40a is performed by actually seeing the pattern on the silicon wafer 31, and if it is aligned therewith, the matrix groove 40a which determines the cutting position of the glass block 27 is formed. Can be formed with high positional accuracy, which in turn can increase the positional accuracy of the glass block 27 with respect to the base 12.

On the other hand, in the prior art, it is necessary to align the processed portion such as a relief portion formed in advance on the glass wafer so as to face the electrode, and about 50 μm caused by the alignment. The error of was the error of the glass block as it was. However, in the present embodiment, it is possible to make only the alignment error of the dicing blade (about 5 to 1/10 of the conventional one).
Since the detection accuracy of the force conversion element 11 is affected by the position accuracy of the glass block 27, the present embodiment can provide the force conversion element 11 with higher detection accuracy.

(D) The positional accuracy of the glass block 27 is
Since it is determined by the dicing accuracy for removing the unnecessary portion of the glass wafer 40 after joining the two wafers 31 and 40,
Even if there is an error in joining the glass wafer 40 to the silicon wafer 31, the error can be absorbed in the dicing process. Therefore, the alignment when bonding the glass wafer 40 to the silicon wafer 31 can be easily completed.

(E) When the glass wafer 40 is placed on the silicon wafer 31, the electrode portions 13 to 13 are arranged so that a gap of about 3 μm is opened between the glass wafer 40 and the electrode 29.
Since the dry etching depth for forming 16 (step portion 31a) is determined, it is possible to prevent the generation of leak current through the electrode 29 when a high voltage is applied by the anodic bonding method. Therefore, the occurrence of defective bonding can be reduced, which contributes to an improvement in the yield of the force conversion element 11.

(F) The depth of the step portion 31a is set to about 3 to 4 μm.
Since it is relatively small, even when the resist is spin-coated in the resist coating step after forming the step portion 31a (for example, FIG. 4A), the resist is surely coated on the entire surface of the silicon wafer without being obstructed by the step. can do.

The present invention is not limited to the above embodiment, but may be configured as follows, for example, without departing from the spirit of the invention. (1) Although the present invention is applied to the force conversion element 11 having the mesa step structure, the manufacturing method of the present invention can also be applied to the force conversion element 51 having no mesa step structure as shown in FIG. 7.
As shown in FIG. 7, the force conversion element 51 includes a base 52 made of, for example, P-type silicon single crystal, and a gauge portion 53 of the base 52.
, A pair of input electrodes 55a and 55b, and a pair of output electrodes 56 joined together by
a and 56b.

FIG. 8 shows a method of manufacturing the force conversion element 51.
This figure corresponds to a cross section taken along line III-III of the force conversion element 51 of FIG. 7. As shown in FIG. 8A, a step portion 61a is formed on the surface of a silicon wafer 61 made of P-type silicon single crystal by dry etching, and an electrode 62 (corresponding to the electrodes 55a to 56b) is formed on the upper surface of the step portion 61a. To form.
The electrode 62 is a gauge portion 53 that is joined to the glass wafer 63.
It is located lower. A glass wafer 63 is placed on the surface of this silicon wafer 61, and both wafers 61 and 63 are bonded by an anodic bonding method. At this time, by ensuring an appropriate gap between the glass wafer 63 and the electrode 62, a leak current does not occur during anodic bonding.

Next, as shown in FIG. 8B, cuts (matrix grooves) 63a are formed on the upper surface of the glass wafer 63 by dicing. Then, the unnecessary portion of the glass wafer 63 is adhesively removed by the adhesive tape, and the glass block 54 is formed as shown in FIG. 8C. And FIG.
As shown in (d), the silicon wafer 61 is cut into each of the bases 52 by dicing, and then cut into the force conversion elements 51.

By adopting this manufacturing method, the same effect as that of the above embodiment can be obtained even in the force conversion element 51 having no mesa step structure. That is, (a) it is not necessary to subject the glass wafer 63 before bonding to a troublesome machining of the escape portion, (b) no positioning is required when bonding the glass wafer 61, and (c) the positional accuracy of the glass block 54 is high. Since it is determined by the dicing accuracy of the glass wafer 61, the positional accuracy of the glass block 54 can be improved as compared with the conventional method.
(D) The effect that the detection accuracy of the force conversion element 51 can be improved by (c), and (e) the unnecessary portion of the glass wafer 63 is adhesively removed, so that there is no fear of damaging the electrode 62 and the yield is improved. Is obtained.

Further, in the conventional manufacturing method, the area error of the bonding area between the silicon wafer and the glass wafer (that is, the pressure receiving area) is determined by the dicing accuracy when the matrix groove is formed. Is determined by the dry etching accuracy when forming the step portion having a sufficiently higher accuracy than the dicing accuracy, so that this area error can be reduced as compared with the conventional method.

(2) As a method for forming the step portion, an appropriate method can be adopted. For example, the step portion may be formed by wet etching with an alkali solution or the like instead of dry etching. A method other than etching may be used as long as the required dimensional accuracy is obtained.

(3) The method of removing the unnecessary portion from the glass wafer in which the matrix groove is formed is not limited to the adhesive removal by the tape. For example, it may be removed by suction. Even if removed by suction, there is no concern that the glass fragments will damage the electrode. Further, the method for removing the adhesive is not limited to the method using the tape, and is not limited to the tape as long as the glass fragments can be removed by the adhesive.

(4) Instead of the method of forming the matrix groove in the glass wafer and adhesively removing the unnecessary portion, the unnecessary portion of the glass wafer may be directly cut off by dicing. If a certain distance is secured between the glass wafer and the electrode, the dicing blade will not hit the electrode. With this method as well, the positional accuracy of the glass block (force transmitting portion) can be improved, and the yield of the force conversion element can be improved.

(5) The wafer forming the substrate 12 (52) of the force conversion element is not limited to the silicon wafer, and other semiconductor wafers can be used. Further, the wafer for the force transmitting portion is not limited to the glass wafer, and other materials having high compressive strength can be used. For example, it may be an uncrystallized glass wafer or a ceramic wafer such as Al2 O3.

(6) The force conversion element to which the present invention is applied is not limited to those shown in the above-mentioned embodiment and the embodiment of (1). The present invention can be widely applied to a force conversion element manufactured by a manufacturing method in which a wafer for a force transmitting portion is bonded to a semiconductor wafer after electrodes are formed, and then separated into individual elements for mass production. In the present invention, the portion on the semiconductor wafer that is joined to the force transmitting portion is the gauge portion.

The invention (or invention) which is grasped from the above-mentioned embodiment and is not described in the scope of claims is described below together with its effect. (A) In the invention according to any one of claims 1 to 3, the distance between the electrode and the wafer for force transmitting portion is set to a value that does not cause a leak when a voltage is applied for bonding both wafers. There is. According to this configuration, when the wafer for force transmitting portion is bonded to the semiconductor wafer by a method such as the anodic bonding method, since an appropriate space is maintained between the electrode and the wafer for force transmitting portion, It is possible to prevent the occurrence of leak. Therefore, defective bonding can be reduced.

(B) In the invention according to any one of claims 1 to 3, the step portion is set to a height that does not hinder the spin coating of the resist. According to this configuration,
When a resist is spin-coated on the surface of the semiconductor wafer after forming the step portion, the resist can be surely coated on the entire surface of the semiconductor wafer without being obstructed by the step.

[0057]

As described above in detail, according to the present invention, the step portion for forming the electrode is formed on the semiconductor wafer, and the electrode formed on the step portion is prevented from protruding from the gauge portion. Therefore, it is not necessary to perform troublesome processing such as a relief portion for avoiding the electrodes on the wafer for force transmitting portion to be bonded to the semiconductor wafer.

According to the second aspect of the present invention, in the wafer removing step, after the unnecessary portion of the force transmitting wafer is cut, the unnecessary portion is adhesively removed by the adhesive means. It is possible to eliminate a fear that an unnecessary portion which should be contacted with the electrode and damages it.

According to the invention described in claim 3, in the force conversion element, since the electrode is not projected from the gauge portion because the electrode is formed on the stepped portion formed on the substrate, The same effect as that of the invention described in Item 1 can be obtained. Further, by adopting a manufacturing method in which the unnecessary portion is cut after the wafer for force transmitting portion is bonded and the unnecessary portion is adhesively removed by an adhesive means, the same effect as that of the invention according to claim 2 can be obtained. .

[Brief description of drawings]

FIG. 1 is a partially cutaway plan view of a force conversion element according to an embodiment.

FIG. 2 is a sectional view taken along line II-II of FIG.

FIG. 3 is a schematic cross-sectional view showing the manufacturing process of the force conversion element.

FIG. 4 is a schematic sectional view of the same.

FIG. 5 is a schematic sectional view of the same.

FIG. 6 is an electric circuit diagram of the force conversion element.

FIG. 7 is a plan view of a force conversion element according to another example.

FIG. 8 is a schematic cross-sectional view showing the manufacturing process of the force conversion element of another example.

FIG. 9 is a schematic cross-sectional view showing a manufacturing process of a force conversion element according to a conventional technique.

[Explanation of symbols]

11, 51 ... Force conversion element, 12, 52 ... Base, 16-2
0, 72 ... Gauge portion, 31, 61 ... Silicon wafer as a semiconductor wafer, 31a, 61a ... Step portion, 40, 6
3 ... Glass wafer as wafer for force transmitting portion, 29, 6
2 ... Electrode.

Claims (3)

[Claims] 1. A method of manufacturing a force conversion element in which electrodes for a plurality of force conversion elements are formed on a semiconductor wafer, a wafer for force transmission section is bonded to the semiconductor wafer, and then the individual force conversion elements are separated. In the method, a step forming step of forming a step portion for forming the electrode on the semiconductor wafer so that the electrode does not protrude from a gauge portion serving as a bonding surface of the force transmitting portion wafer; and the step forming step. An electrode forming step of forming an electrode on the step formed by the step, a wafer bonding step of bonding the force transmitting portion wafer to the gauge portion side surface of the electrode-formed semiconductor wafer, and a wafer bonding step above the electrode. And a wafer removing step of removing an unnecessary portion of the force transmitting portion wafer. 2. The force conversion according to claim 1, wherein, in the wafer removing step, after cutting an unnecessary portion of the force transmitting wafer, the unnecessary portion is adhesively removed by an adhesive means. Device manufacturing method. 3. The force converting element manufacturing method according to claim 1 or 2, wherein an electrode is formed on the stepped portion formed on the base body of the force converting element, so that the electrode is the gauge portion. A force conversion element characterized in that it does not project further.