JPH07283419A - Semiconductor sensor and its manufacturing method - Google Patents

Abstract

PURPOSE:To nearly match the central thick part of a first substrate to the center of a plumb bob by forming a first substrate and a machining mark used for separating a second substrate on the second surface of the second substrate after joining the second substrate. CONSTITUTION:The area of a glass layer 5 in contact with a central thick part 14 is made larger than that of the central thick part and the area of the glass layer 5 in contact with a peripheral thick part 15 is made smaller than the peripheral thick part 15, thus joining first and second substrates 1 and 2. Then, a groove part 7 is formed by etching the silicon of the second surface of the second substrate 2 and the outer periphery of the groove part 7 is used as a machining mark for determining a line (dicing line) for separating a plumb bob 22 and a support 21. The outer periphery of the groove part 7 is matched to a dicing blade and a wafer is moved by 100mum and is subjected to separation machining at that position. The dicing line is provided at a position overlapping with the peripheral thick part 15 of the first substrate 1 and the move of the plumb bob 22 is limited by the peripheral thick part, thus easily matching the center of the plumb bob 22 to the center of the central thick part 14.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a semiconductor sensor for detecting physical quantities such as acceleration and force.

[0002]

2. Description of the Related Art Recently, in order to detect the movement of an object,
A sensor that detects acceleration is drawing attention. JP-A-3-
As disclosed in Japanese Patent No. 2535, the conventional acceleration sensor has a shape as shown in FIG. FIG. 6 is a schematic sectional view showing a conventional acceleration sensor. This acceleration sensor is composed of three substrates, and the first substrate 1 is
The central thick portion 14, the peripheral thick portion 15, and the thin portion 13 on which the resistive element is formed on the front surface (the first surface of the first substrate) are formed. The second substrate 2 is a part of the second substrate 2. A weight body 22 that is connected to the central thick-walled portion 14 and a support body 21 that is connected to the peripheral thick-walled portion 15 and that supports the peripheral thick-walled portion 15. Substrate 3 is weight 2
It is a pedestal for limiting the movement of 2.

In the acceleration sensor shown in FIG. 6, the movement of the weight body 22 is restricted so that the thin portion 13 and the like will not be destroyed even if an excessive acceleration acts. Weight 22
Is restricted between the second surface of the peripheral thick portion 15 of the first substrate and the first surface of the weight body 22 and between the second surface of the weight body 22 and the first surface of the third substrate. Has been done. Further, in order to provide such a space, the first surface of the second substrate (the first of the weight body 22
Surface) and the first surface of the third substrate are partially cut away.

This acceleration sensor, as shown in FIG.
When a lateral acceleration acts on the sensor, the weight 2
2 performs a rotational movement with the center of the upper surface of the central thick portion 14 of the first substrate 1 as a reference, the thin portion 13 is deformed accordingly, and the resistance change of the resistance element 12 formed in the thin portion 13 Detect acceleration. When the vertical acceleration is applied, the weight body 22 moves up and down as shown in FIG.
Is deformed, and the acceleration can be detected in the same manner. When excessive acceleration acts, the second surface of the first substrate 1 (the peripheral thick portion 1
5) Or, when the pedestal 3 hits the edge or the second surface of the weight body 22 (that is, the strong ones hit each other), an excessive acceleration that may destroy the thin portion 13 is applied to the acceleration sensor. However, the thin portion 13 is prevented from being excessively deformed (the sensor is prevented from being broken).

As described above, in order to restrict the movement of the weight body 22, between the second surface of the peripheral thick portion 15 and the first surface of the weight body 22, and the second surface of the weight body 22. (Or edge) and third
A gap is required between the substrate 3 and the first surface of the pedestal 3, and conventionally, in order to form this gap, the first surface of the weight body 22 and the third substrate 3 of the pedestal 3 are formed. A gap is secured by processing one surface.

When glass is used as the material of the second substrate as disclosed in Japanese Patent Laid-Open No. 3-2535,
In order to facilitate the weight separation (separation of the weight from the support) and the chip separation (separation of the substrate for each individual sensor), a second separation groove is provided in advance.
It must be formed on the first surface of the substrate. This is because glass is a hard material and positioning is required for separating the weight body.

After forming the separation groove, the first
The second surface of the substrate and the first surface of the second substrate are joined together, and then the groove formed on the first surface of the second substrate is viewed from the second surface (the glass is a transparent material. The groove can be detected even from the above), and dicing (separation) is performed along the groove to perform weight separation or chip separation.
Therefore, the size of the weight body is determined by the groove for separating the weight body, so that the first substrate and the second substrate must be joined in an extremely accurate positional relationship. this is,
When the weight body deviates from the correct position (the position where the center of the weight body and the center of the central thick portion coincide), a crosstalk component occurs in the sensor output (accurate acceleration cannot be detected) This is because the characteristics of the sensor are significantly deteriorated.

[0008]

In the conventional acceleration sensor, it was extremely difficult to align the first substrate and the second substrate when joining them. For this reason, if the second substrate is processed and then bonded to the first substrate, the positional error between the weight body and the central thick portion deviates from the correct position due to an alignment error at the time of bonding. There was a problem that it got worse.

[0009] Therefore, an object of the present invention is to provide a semiconductor sensor in which the center of the weight body and the center of the central thick portion can be made to substantially coincide with each other and the physical quantity detection accuracy is high, and a method for manufacturing the same.

[0010]

That is, according to the present invention, a central thick portion located substantially in the center of a first substrate made of a semiconductor substrate, and a thin thin portion located around the central thick portion, A peripheral thick portion located around the thin portion, a weight body formed by separating the second substrate from a part of the second substrate and joined to the central thick portion, and another one of the second substrate. In a semiconductor sensor including a support body formed from a portion and joined to a peripheral thick portion, a second substrate is formed after the second substrate and the first substrate are joined.
By processing the second surface of the substrate, a processing mark used when separating the second substrate is formed on the second surface of the second substrate (claim 1).

In this case (claim 1), preferably, the second surface of the weight body is cut so that the second surface of the weight body is lower than the second surface of the support body (claim 2). In this case (claim 1 or 2), preferably, a bonding layer for bonding the first substrate and the second substrate is provided between the first substrate and the second substrate (claim 3). Further, according to the present invention, in a method for manufacturing a semiconductor sensor, a first step of defining a central thick portion region on a first substrate made of a semiconductor substrate, a thin portion region around the central thick portion region, and a peripheral thick portion region on the periphery thereof A second step of forming a resistive element in the thin-walled region, a third step of partially removing the second surface of the first substrate to give flexibility to the thin-walled region, and a second step of the first substrate. A fourth step of joining the first surface of the second substrate to the first surface, and a fifth step of forming a processing mark used when separating the second substrate by processing the second surface of the second substrate, By separating the second substrate using the processing mark, the weight body which is joined to the central thick portion region and which is composed of a part of the second substrate is joined to the peripheral thick portion region. A sixth substrate forming a support composed of the other part of the substrate Tsu has a flop, the (claim 4).

Further, according to the present invention, in a method of manufacturing a semiconductor sensor, a central thick portion region is defined on a first substrate made of a semiconductor substrate, a thin portion region is defined around the central thick portion region, and a peripheral thick portion region is defined around the central thick portion region. 1 step, 2nd step of forming a resistance element in the thin portion region, 3rd step of partially removing the 2nd surface of the 1st substrate in order to give flexibility to the thin portion region, 1st substrate A fourth step of joining the first surface of the second substrate to the second surface of the second substrate, and a processing mark and at least a weight body used when separating the second substrate by processing the second surface of the second substrate. The fifth step of forming a groove on the second surface of the region to be formed, and the second substrate is separated by using the processing mark so as to be bonded to the central thick portion region.
A sixth step of forming a weight body composed of a part of the substrate and a support body bonded to the peripheral thick portion region and composed of the other part of the second substrate. 5).

In this case (claim 5 or 6), preferably, before the fourth step, there is a step of forming a bonding layer for bonding the first substrate and the second substrate (claim 6).

[0014]

According to the present invention, after the first substrate and the second substrate are bonded to each other, a processing mark used for separating the second substrate is formed on the second surface of the second substrate, so that the center thickness of the first substrate is increased. The center of the meat portion and the center of the weight body can be made to substantially coincide with each other, and the detection accuracy of the physical quantity in the semiconductor sensor can be improved.

[0015]

The present invention will be described in more detail with reference to the following examples, but the present invention is not limited thereto. Hereinafter, a method of manufacturing the semiconductor sensor according to the first embodiment of the present invention will be described with reference to FIG.

In the first embodiment, a silicon substrate having a (100) plane is used for the first substrate and the second substrate, and a p-type substrate is used for the first substrate. The first substrate has a thickness of about 250 μm, and the second substrate has a thickness of about 1.4 mm.
Is. First, LP- is formed on the first and second surfaces of the first substrate 1.
The nitride film 4 is formed by the CVD method or the like, and then the nitride film on the first surface of the first substrate 1 is removed. Next, the n-type epitaxial growth layer 11 is formed to a thickness of 10 μm on the first surface of the first substrate 1 from which the nitride film has been removed. Further, a silicon oxide film having a thickness of 0.6 μm is formed on the epitaxial growth layer 11 (not shown). Next, on the first surface (on the n-type epitaxial growth layer 11) of the region where the thin portion is formed, the resistance element 12 for detecting acceleration is formed by the well-known IC planar process, and then the resistance element 12 is connected. Metal wiring (not shown) is formed. At this time, the metal wiring is insulated from the n-type epitaxial growth layer 11 by the silicon oxide film. Next, a resist pattern is formed by photolithography so as to partially remove the nitride film 4 formed on the second surface of the first substrate 1 (not shown).
The photolithography at this time uses a double-sided aligner, and the resistance element 12 formed on the first surface of the first substrate 1 is used.
An accurate positional relationship between the pattern of the detection part such as the above and the pattern of the resist formed on the second surface of the first substrate 1 is maintained. Further, when the double-sided aligner is used, the alignment mark on the first surface of the first substrate 1 is used. The alignment mark used when forming the resistance element 12 and the like is used. Then, the nitride film 4 is partially etched using the resist as a mask. [FIG. 2 (a)].

Next, using the nitride film 4 partially left on the second surface of the first substrate 1 as a mask, the first substrate 1 is etched with a strong alkaline etching solution such as an aqueous solution of potassium hydroxide to a desired depth. Then, the etching is stopped to form the thin portion 13. At this time, since the (111) plane of silicon is hardly etched, the silicon is etched at a constant angle as shown in FIG. Further, the etching is performed by using electrochemical etching,
It can be stopped at the surface of the type epitaxial growth layer 11 [FIG. 2 (b)].

Next, a nitride film 24 having a thickness of 0.1 μm is formed on both surfaces of the second substrate 2. After that, a glass layer 5 (bonding layer) containing mobile ions and having a thickness of 5 to 10 μm is formed on the first surface of the second substrate 2. The glass layer 5 (bonding layer) is formed by a method such as sputtering or anodic bonding a thin glass plate. Next, when joining the first substrate 1 and the second substrate 2, the peripheral thick portion 15 and the central thick portion 1 of the first substrate 1
After forming a pattern with a resist by photolithography so as to leave the glass layer 5 (bonding layer) in contact with the glass layer 4, the glass layer 5 (bonding layer) is removed by etching using a hydrofluoric acid aqueous solution. At this time, when the glass layer 5 (bonding layer) is bonded to the first substrate 1 and the second substrate 2,
It must be formed so as to contact the central thick portion 14 and the peripheral thick portion 15 of the substrate 1. Therefore, as shown in FIG. 2C, the area of the glass layer 5 (bonding layer) contacting the central thick portion 14 is made larger than that of the central thick portion 14, and the glass layer 5 (contacting the peripheral thick portion 15 ( Making the area of the bonding layer smaller than that of the peripheral thick portion 15 sufficiently acts as the bonding layer even if there is some alignment error, and when separating the second substrate 2 described later, the peripheral thick portion The glass layer 5 (bonding layer) in contact with the portion 15 does not overlap the separation line (dicing line), which is preferable. After that, the second surface of the first substrate 1 and the first surface of the second substrate 2 are bonded by the anodic bonding method via the glass layer 5 (bonding layer) [FIG.
(C)].

Thereafter, photolithography and dry etching using a reactive gas are performed to partially remove the nitride film 24 in at least the region where the weight body is formed on the second surface of the second substrate 2, and the etching pattern (first 2) A pattern for etching the second surface of the substrate is formed. In the photolithography at this time, a double-sided aligner is used as in the case of etching the second surface of the first substrate 1,
Further, by using the alignment mark formed on the first surface of the first substrate 1, the pattern of the detection unit such as the resistance element 12 formed on the first surface of the first substrate 1 and the pattern of the resist are formed. Try to keep the positional relationship accurate. Next, the silicon is etched with a strong alkaline aqueous solution to form the groove portion 7 having a depth of 5 to 10 μm. Groove 7 at this time
The shape is as shown in FIG. 4 (a) [FIG. 2 (d)]. Further, the outer circumference of the groove portion 7 formed on the second surface of the second substrate 2 is
It is used as a processing mark for determining a line (dicing line) for separating the weight body and the support body in the weight body separation processing performed next [FIG. 4 (a)].

After that, the groove portion 7 formed on the second surface of the second substrate is separated from the outer periphery by a distance of 100 μm from the outer periphery [a line shown by a dotted line in FIG. 4 (a)]. First, while observing the microscope of the dicing saw, the outer periphery of the groove portion 7 is aligned with the dicing blade (the blade that cuts the second substrate 2), and then the wafer (the first substrate 1 and the second substrate 2 are bonded together) ) Is moved by 100 μm, and separation processing is performed at that position [FIG. 2 (e)]. At this time, the accuracy when moving the wafer is about 1 μm, and the alignment accuracy of the double-sided aligner is about 5 μm.
The accuracy of separation is determined by the accuracy of the double-sided aligner.

The groove portion 7 formed on the second surface of the second substrate
It is not necessary to match the outer periphery of the substrate with the line separating the two substrates (dicing line), but at least the dicing line is positioned so as not to overlap the glass layer 5 (bonding layer) bonded to the peripheral thick portion 15. There is a need. Further, in order to limit the movement of the weight body 22 at the peripheral thick portion 15, the dicing line needs to be positioned so as to overlap the peripheral thick portion 15 of the first substrate 1.

After that, the second surface of the support 21 is fixed to the third substrate or the package. When silicon is used for the third substrate, it is preferable that a glass layer is formed on the third substrate in advance and the second substrate 2 and the third substrate are joined by an anodic bonding method. When the glass layer is formed on the third substrate, it is preferable to adjust the depth of the groove portion 7 according to the thickness of the glass layer.

As described above, a semiconductor sensor in which the center of the weight body 22 and the center of the central thick portion 14 can be easily aligned can be manufactured. Further, in the semiconductor sensor manufactured according to the first embodiment, the movement of the weight body 22 can be limited by the gap due to the thickness of the glass layer 5 (bonding layer) and the depth of the groove portion 7. It should be noted that instead of the glass layer 5 (bonding layer), the peripheral thick portion 15 of the first substrate 1
The movement of the weight body 22 may be limited by providing a gap by etching the portion of the second surface where the weight body contacts. Further, in the first embodiment, it is not necessary to process the third substrate in order to provide the gap for limiting the movement of the weight body 22. Therefore, it is not necessary to use the third substrate by directly joining the second surface of the second substrate to a place where the physical quantity is desired to be detected (a substantially flat place), for example, a package or the like.

Although silicon is used for the second substrate in the first embodiment, a substrate made of another material may be used. In this case, it is preferable to use a material having good workability such as silicon as the material of the second substrate. Also, the first
Although glass is used as the bonding layer for bonding the substrate 1 and the second substrate 2, other materials may be used. Further, in the first embodiment, the weight body 22 has a square shape as shown in FIG. 4A, but may have another shape (for example, a circular shape or a regular octagonal shape). In that case, The shape of the groove may be a shape corresponding to the shape.

Further, in the first embodiment, the gap for limiting the movement of the weight body 22 and the processing mark used when separating the second substrate are also used as the groove portion 7, but only the processing mark is used as the second processing mark. The present invention can also be implemented by forming it on the second surface of the substrate. Therefore, an example in which only the processed mark is formed will be described below as a second example. Hereinafter, a method for manufacturing a semiconductor sensor according to the second embodiment of the present invention will be described with reference to FIG. The same parts as those in the first embodiment are designated by the same reference numerals.

In the second embodiment, as in the first embodiment, silicon substrates having a (100) plane are used for the first and second substrates, and the first substrate is a p-type substrate. There is. The first substrate has a thickness of about 250 μm, and the second substrate has a thickness of about 1.4 mm. First, the first substrate 1
The nitride film 4 is formed on the first and second surfaces of the first substrate 1 by the LP-CVD method or the like, and then the nitride film on the first surface of the first substrate 1 is removed. Next, the n-type epitaxial growth layer 11 is formed to a thickness of 10 μm on the first surface of the first substrate 1 from which the nitride film has been removed.
Further, a silicon oxide film having a thickness of 0.6 μm is formed on the epitaxial growth layer 11 (not shown). Next, the first surface (on the n-type epitaxial growth layer 11) of the region where the thin portion is formed
Then, a resistance element 12 for detecting acceleration is formed by a well-known IC planar process, and then a metal wiring (not shown) for connecting the resistance element 12 is formed. At this time, the metal wiring is insulated from the n-type epitaxial growth layer 11 by the silicon oxide film. Next, the first substrate 1
A resist pattern is formed by photolithography so as to partially remove the nitride film 4 formed on the second surface of (not shown). In the photolithography at this time, a double-sided aligner is used, and the pattern of the detection portion such as the resistance element 12 formed on the first surface of the first substrate 1 and the pattern of the resist formed on the second surface of the first substrate 1. Maintain an accurate positional relationship with. Further, when the double-sided aligner is used, the alignment mark on the first surface of the first substrate 1 is used. The alignment mark used when forming the resistance element 12 and the like is used. Then, the nitride film 4 is partially etched using the resist as a mask.
[FIG. 3 (a)].

Next, the first substrate 1 is etched with a strong alkali etching solution such as an aqueous solution of potassium hydroxide using the nitride film 4 partially left on the second surface of the first substrate 1 as a mask. Then, the etching is stopped to form the thin portion 13. At this time, since the (111) plane of silicon is hardly etched, the silicon is etched at a constant angle as shown in FIG. Further, the etching is performed by using electrochemical etching,
It can be stopped at the surface of the type epitaxial growth layer 11 [FIG. 3 (b)].

Next, the nitride films 24 are formed on both surfaces of the second substrate 2.
μm is formed. After that, a glass layer 5 (bonding layer) containing mobile ions and having a thickness of 5 to 10 μm is formed on the first surface of the second substrate 2. The glass layer 5 (bonding layer) is formed by a method such as sputtering or anodic bonding a thin glass plate. Next, when joining the first substrate 1 and the second substrate 2,
A pattern is formed with a resist by photolithography so as to leave the glass layer 5 (bonding layer) in a portion in contact with the peripheral thick portion 15 and the central thick portion 14 of the first substrate 1, and then glass is formed using an aqueous solution of hydrofluoric acid. The layer 5 (bonding layer) is removed by etching. At this time, the glass layer 5 (bonding layer)
Needs to be formed so as to contact the central thick portion 14 and the peripheral thick portion 15 of the first substrate 1 when joining the first substrate 1 and the second substrate 2. Therefore, as shown in FIG. 3C, the area of the glass layer 5 (bonding layer) in contact with the central thick portion 14 is made larger than that of the central thick portion 14, and the peripheral thick portion 15
The area of the glass layer 5 (bonding layer) in contact with the peripheral thick portion 1
If it is smaller than 5, it works sufficiently as a bonding layer even if there is some alignment error, and the second
When the substrate 2 is separated, the glass layer 5 (bonding layer) contacting the peripheral thick portion 15 does not overlap the separation line (dicing line), which is preferable. After that, the second surface of the first substrate 1 and the first surface of the second substrate 2 are bonded by the anodic bonding method via the glass layer 5 (bonding layer) [FIG. 3 (c)].

After that, photolithography and dry etching using a reactive gas are performed to partially remove the nitride film 24 on the second surface of the second substrate 2 to form the weight body separating groove 6 (processing mark). Form. Groove 6 for weight separation at this time
As shown in FIG. 4B, the (working mark) is formed by a straight line having a width of about several tens of μm like the line (dicing line) separating the second substrate, and has a square shape. In the photolithography at this time, a double-sided aligner is used as in the case of etching the second surface of the first substrate 1, and the alignment mark formed on the first surface of the first substrate 1 is used. Then, the positional relationship between the pattern of the detection portion such as the resistance element 12 formed on the first surface of the first substrate 1 and the pattern of the resist is accurately maintained [FIG. 3 (d)].

Thereafter, the weight body and the support body 21 are separated along the weight body separating groove 6 (working mark) formed on the second surface of the second substrate in the same manner as in the first embodiment. It is processed [Fig. 3 (e)]. After that, the second surface of the supporting body 21 is fixed to the third substrate 3 or the package, but since the second surfaces of the peripheral supporting body 21 and the weight body 22 are at the same height, the third substrate is directly attached. Or it cannot be fixed to the package. Therefore, a bonding layer such as a glass layer is provided to bond the second surface of the second substrate and the third substrate 3 or the package. Further, by controlling the thickness of the bonding layer, it is possible to control the movement of the weight body 22 [FIG.
(F)].

In the second embodiment, the nitride film 24 on the entire portion of the dicing line is removed, but the nitride film 24 may be removed partially so that the dicing line can be seen. That is, the processed mark 51 or the like as shown in FIGS. 5A and 5B may be used. Further, it is clear from the first embodiment that the processed mark does not necessarily have to be formed on the dicing line and may be formed at another place.

Although not described in the first and second embodiments, the present invention is also suitable for chip separation, which is performed in a manufacturing method for mass-producing a plurality of sensors. In that case, similarly to the first and second embodiments, a processing mark corresponding to a line (dicing line) for chip separation for each sensor is formed on the second surface of the second substrate 2. You can leave it.

[0033]

As described above, according to the present invention, the positioning of the weight body is performed after the first substrate and the second substrate are bonded to each other, whereby the semiconductor sensor can be easily manufactured, and The accuracy of the semiconductor sensor could be improved.

[Brief description of drawings]

FIG. 1 is a side sectional view showing the structure of an acceleration sensor according to the present invention.

FIG. 2 is a schematic view showing a process of the semiconductor sensor according to the first embodiment of the present invention.

FIG. 3 is a schematic view showing a process of the semiconductor sensor according to the second embodiment of the present invention.

FIG. 4 is a diagram showing a pattern formed on a second surface of a second substrate, respectively.

FIG. 5 is a view showing a processing mark formed on the second surface of the second substrate of the semiconductor sensor according to the second embodiment of the present invention.

FIG. 6 is a side sectional view showing a conventional acceleration sensor.

FIG. 7 is an explanatory diagram showing the operation of the acceleration sensor when lateral acceleration is applied.

FIG. 8 is an explanatory diagram showing the operation of the acceleration sensor when vertical acceleration is applied.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... 1st substrate 2 ... 2nd substrate 3 ... 3rd substrate, pedestal 4 ... Nitride film 5 ... Glass layer 6 ... Weight separation groove (processing mark) 7 ... ..Groove 11 ... N-type epitaxial growth layer 12 ... Resistor element 13 ... Thin portion 14 ... Central thick portion 15 ... Peripheral thick portion 21 ... Support 22 ... Heavy Weight 51 ... Machining mark

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Jun Iwasaki 3 2-3 Marunouchi, Chiyoda-ku, Tokyo Inside Nikon Corporation

Claims (6)

[Claims] 1. A central thick portion located substantially in the center of a first substrate made of a semiconductor substrate, a thin thin portion located around the central thick portion, and a periphery located around the thin portion. A weight part formed from a part of the second substrate by separating the thick part and the second substrate and joined to the central thick part; and a weight part formed from another part of the second substrate. In a semiconductor sensor including a support joined to a peripheral thick portion, the second substrate is formed by processing the second surface of the second substrate after joining the second substrate and the first substrate. A semiconductor sensor, wherein a processing mark used for separation is formed on the second surface of the second substrate. 2. The semiconductor sensor according to claim 1, wherein the second surface of the weight body is made lower than the second surface of the support body by grinding the second surface of the weight body. 3. The semiconductor according to claim 1, wherein a bonding layer for bonding the first substrate and the second substrate is provided between the first substrate and the second substrate. Sensor. 4. A first step of defining a central thick portion region on a first substrate made of a semiconductor substrate, a thin portion region around the central thick portion region, and a peripheral thick portion region around the first thick portion region, and a resistance element in the thin portion region. A second step of partially removing the second surface of the first substrate to provide flexibility in the thin portion region, and a second step of forming a second surface of the first substrate on the second surface. A fourth step of joining the first surface of the substrate; a fifth step of forming a processing mark to be used when separating the second substrate by processing the second surface of the second substrate; By separating the second substrate using a mark, a weight body which is joined to the central thick portion region and which is composed of a part of the second substrate is joined to the peripheral thick portion region. And a support composed of another part of the second substrate. The method of manufacturing a semiconductor sensor characterized by having a sixth step of forming a. 5. A first step of defining a central thick portion region on a first substrate made of a semiconductor substrate, a thin portion region around the central thick portion region, and a peripheral thick portion region on the periphery thereof, and a resistance element in the thin portion region. A second step of partially removing the second surface of the first substrate to provide flexibility in the thin portion region, and a second step of forming a second surface of the first substrate on the second surface. A fourth step of joining the first surface of the substrate; a processing mark used when separating the second substrate by processing the second surface of the second substrate; and a region for forming at least a weight body. A fifth step of forming a groove on the second surface, and a step of separating the second substrate by using the processing mark so that the second substrate is joined to the central thick portion region and is composed of a part of the second substrate. By connecting it to the surrounding thick-walled area. The method of manufacturing a semiconductor sensor comprising: the sixth step, the forming and formed support from other parts of the second substrate. 6. The semiconductor sensor according to claim 4, further comprising a step of forming a bonding layer for bonding the first substrate and the second substrate before the fourth step. Manufacturing method.