Semiconductor device and manufacturing method thereof

By incorporating recesses in GaN wafers to manage stress, the accuracy of pre-division tests is maintained, addressing stress variation issues in semiconductor device manufacturing.

JP2024157928AInactive Publication Date: 2024-11-08DENSO CORP +2
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
JP2023072605
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-04-26
Publication Date
2024-11-08
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The challenge in manufacturing semiconductor devices using gallium nitride (GaN) wafers is that the stress applied to semiconductor elements before and after wafer division can vary significantly, affecting the accuracy of pre-division characteristic tests, especially as the device configuration wafer thickness decreases.

Method used

Forming recesses, such as bevel structures or trenches, in the outer peripheral regions of GaN wafers to alleviate stress during division, ensuring consistent stress application before and after wafer separation.

Benefits of technology

The recesses help maintain the accuracy of characteristic tests by reducing stress variations, allowing for precise inspection and division of semiconductor devices.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024157928000001_ABST
    Figure 2024157928000001_ABST
Patent Text Reader

Abstract

To easily maintain the accuracy of characteristic inspection performed before division.SOLUTION: A semiconductor device includes a semiconductor substrate 10 that is made of gallium nitride and has one surface 10a and another surface 10b opposite to the surface 10a, on which a semiconductor element is formed. The semiconductor substrate 10 has a cell region 1 in which the semiconductor element is formed and a peripheral region 2 surrounding the cell region 1, and the peripheral region 2 is configured to form a recess 15 by depressing the surface 10a.SELECTED DRAWING: Figure 1
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a semiconductor device made of gallium nitride (hereinafter also referred to as GaN), and a method for manufacturing the same. [Background technology]

[0002] Conventionally, a method for dividing a processed wafer made of GaN has been proposed (for example, see Patent Document 1). Specifically, in this method, a processed wafer is prepared, and then a degenerated layer is formed inside the processed wafer along the surface direction by irradiating the wafer with a laser beam. Then, in this method, the processed wafer is divided into two wafers using the degenerated layer as the starting point for division. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2019-043808 A Summary of the Invention [Problem to be solved by the invention]

[0004] The present inventors are considering manufacturing a semiconductor device using a processed wafer. Specifically, the present inventors are considering manufacturing a semiconductor device by dividing the processed wafer with a plurality of semiconductor elements formed thereon, and using the side on which the semiconductor elements are formed as device configuration wafers and dividing the device configuration wafer into individual pieces. In this case, in order to simplify the inspection, it is conceivable to perform characteristic inspection such as electrical characteristics of each semiconductor element in the processed wafer state, and then divide the processed wafer into device configuration wafers.

[0005] However, when a processed wafer is divided, the divided surface becomes rough, and stress is easily applied to the device configuration wafer due to the Twyman effect, etc. Therefore, the stress applied to the semiconductor element in the processed wafer state before division is likely to differ from the stress applied to the semiconductor element in the device configuration wafer state after division. In particular, the thinner the device configuration wafer is, the greater the difference between the stress applied to the semiconductor element in the processed wafer state before division and the stress applied to the semiconductor element in the device configuration wafer state after division is likely to be. Therefore, it may be difficult to maintain the accuracy of the characteristic inspection performed before division.

[0006] In view of the above, an object of the present invention is to provide a semiconductor device and a manufacturing method thereof that can easily maintain the accuracy of a characteristic inspection performed before division. [Means for solving the problem]

[0007] Claim 1 for achieving the above object is a semiconductor device comprising a semiconductor substrate (10) made of gallium nitride, on which a semiconductor element is formed and having one surface (10a) and another surface (10b) opposite the one surface, the semiconductor substrate having a cell region (1) in which the semiconductor element is formed and a peripheral region (2) surrounding the cell region, and the peripheral region has recesses (15, 16) formed on one surface.

[0008] According to this, since the recess is formed in the outer peripheral region, when the processed wafer is divided to construct the device configuration wafer, the stress can be relaxed in the recess. In other words, it is possible to suppress a large difference in the stress applied to the semiconductor element before and after dividing the processed wafer. Therefore, it is easy to maintain the accuracy of the characteristic inspection performed before dividing the processed wafer.

[0009] Claim 7 is a manufacturing method for the semiconductor device of claim 1, comprising the steps of: preparing a processed wafer (200) having one surface (200a) and another surface (200b) opposite to the one surface, made of gallium nitride, and having a plurality of device formation regions (RA); forming a semiconductor element on one surface of the processed wafer; inspecting the characteristics of the semiconductor element; forming an altered layer (15) inside the processed wafer along the surface direction of the processed wafer by irradiating the inside of the processed wafer from the other surface of the processed wafer with laser light (L) after the characteristic inspection; dividing the processed wafer at the altered layer as a boundary, thereby dividing the processed wafer into a device configuration wafer (230) on one surface of the processed wafer and a recycled wafer (240) on the other surface of the processed wafer; and singulating the device configuration wafer; and forming a recess on the outer edge of the device formation region by forming the semiconductor element.

[0010] According to this, since the recess is formed on the outer edge of the device formation region, when the processed wafer is divided to form the device configuration wafer, the stress can be relieved by the recess. In other words, it is possible to prevent the stress applied to the semiconductor element before and after dividing the processed wafer from being significantly different. Therefore, it is easy to maintain the accuracy of the characteristic inspection performed before dividing the processed wafer.

[0011] The reference symbols in parentheses attached to each component indicate an example of the correspondence between the component and specific components described in the embodiments described below. [Brief description of the drawings]

[0012] [Figure 1] 1 is a cross-sectional view of a semiconductor device according to a first embodiment. [Diagram 2] FIG. 2 is a plan view of the semiconductor substrate shown in FIG. [Figure 3A] 2A to 2C are cross-sectional views showing a manufacturing process of the semiconductor device shown in FIG. [Figure 3B] 3B is a cross-sectional view showing a manufacturing process of the semiconductor device subsequent to FIG. 3A. [Figure 3C]3C is a cross-sectional view showing a manufacturing process of the semiconductor device subsequent to FIG. 3B. [Figure 3D] 3D is a cross-sectional view showing a manufacturing process of the semiconductor device subsequent to FIG. 3C. [Figure 3E] 3D; FIG. 4 is a cross-sectional view showing a manufacturing process of the semiconductor device subsequent to FIG. 3D; [Figure 3F] 3F is a cross-sectional view showing a manufacturing process of the semiconductor device subsequent to FIG. 3E. [Figure 3G] 3F is a cross-sectional view showing a manufacturing process of the semiconductor device subsequent to FIG. 3F. [Figure 3H] 3C is a cross-sectional view showing a manufacturing process of the semiconductor device subsequent to FIG. 3G. [Figure 3I] 3C is a cross-sectional view showing a manufacturing process of the semiconductor device subsequent to FIG. 3H. [Figure 3J] 3I. FIG. 3I is a cross-sectional view showing a manufacturing process of the semiconductor device. [Figure 3K] 3J is a cross-sectional view showing a manufacturing process of the semiconductor device subsequent to FIG. 3J. [Figure 4] FIG. 3C is a plan view of the processed wafer corresponding to FIG. 3B. [Diagram 5] 3D is an enlarged plan view showing a device formation region and dicing lines of the processed wafer corresponding to FIG. 3C. FIG. [Figure 6] 11A and 11B are diagrams showing the relationship between the drain current and the gate voltage before and after division in this embodiment. [Figure 7] FIG. 13 is a diagram showing the relationship between the drain current and the gate voltage before and after division in the comparative example. [Figure 8] FIG. 11 is a diagram showing threshold voltages before and after division in the present embodiment and the comparative example. [Figure 9] FIG. 13 is a diagram for explaining a bevel structure used in a simulation. [Figure 10] 11A and 11B are diagrams illustrating threshold voltages before and after division and differences in the amounts of fluctuation in threshold voltages in the present embodiment and the comparative example. [Figure 11] FIG. 13 is a diagram showing the relationship between compressive stress and threshold voltage. [Figure 12] FIG. 13 is a diagram showing the relationship between the distance from one surface of a device configuration wafer and stress. [Figure 13]FIG. 11 is a cross-sectional view of a semiconductor device according to a second embodiment. [Figure 14] FIG. 11 is a cross-sectional view of a semiconductor device according to a third embodiment. [Figure 15] 13 is an enlarged plan view showing a device formation region and dicing lines of a processed wafer in another embodiment. FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the following embodiments, parts that are the same or equivalent to each other will be denoted by the same reference numerals.

[0014] (First embodiment) The first embodiment will be described with reference to the drawings. In this embodiment, a semiconductor device in which a vertical n-channel MOSFET (short for metal oxide semiconductor field-effect transistor) is formed will be described as an example. The semiconductor device of this embodiment is preferably applied to a semiconductor device mounted on a vehicle such as an automobile for driving various electronic devices for the vehicle.

[0015] 1 and 2, the semiconductor device S1 has a cell region 1 and an outer peripheral region 2 surrounding the cell region 1. FIG. 1 corresponds to a cross-sectional view taken along line II in FIG. 2. The semiconductor device S1 is configured by forming a MOSFET in the cell region 1. MOSFET is an abbreviation for metal oxide semiconductor field effect transistor. Specifically, the semiconductor device S1 has n +The semiconductor device is provided with a substrate 11 made of n-type GaN. In this embodiment, the substrate 11 constitutes a drain layer. A drift layer 12 made of n-type GaN is formed on the substrate 11. A p-type base layer 13 is formed on the surface portion of the drift layer 12. This base layer 13 constitutes the channel of the MOSFET. The surface portion of the base layer 13 has an n-type + A shaped source region 14 is formed.

[0016] In this embodiment, the semiconductor substrate 10 includes a substrate 11, a drift layer 12, a base layer 13, a source region 14, etc. In the following description, the surface of the semiconductor substrate 10 on the substrate 11 side is referred to as the other surface 10b of the semiconductor substrate 10, and the surface of the semiconductor substrate 10 on the source region 14 side is referred to as the first surface 10a. The semiconductor substrate 10 (i.e., the semiconductor device S1) in this embodiment has a rectangular shape in plan view.

[0017] A bevel structure 15 is formed as a recess in the outer peripheral region 2 on one surface 10a of the semiconductor substrate 10. In this embodiment, the bevel structure 15 is formed in a tapered shape in which the depth gradually deepens from the cell region 1 side toward the opposite side to the cell region 1. The bevel structure 15 is formed along the outline of the planar shape of the semiconductor substrate 10, and is configured to include at least a portion formed along two opposing sides of the outline of the planar shape of the semiconductor substrate 10. In this embodiment, the bevel structure 15 is formed in a frame shape along each outline. That is, the bevel structure 15 is formed so as to surround the cell region 1. The depth of the bevel structure 15 can be changed as appropriate, but is set to be less than the thickness of the semiconductor substrate 10.

[0018] An insulating film 20 made of an oxide film or the like is formed on one surface 10a of the semiconductor substrate 10. In this embodiment, the insulating film 20 is formed at least on a portion located on the base layer 13. A gate electrode 31 is formed on the insulating film 20. Although not shown, a source electrode electrically connected to the base layer 13 and the source region 14 is also disposed on the one surface 10a of the semiconductor substrate 10.

[0019] On the other surface 10b side of the semiconductor substrate 10, a drain electrode 32 electrically connected to the substrate 11 is formed.

[0020] The above is the configuration of the semiconductor device in this embodiment.

[0021] In such a semiconductor device, when a voltage equal to or higher than a predetermined threshold voltage is applied to the gate electrode 31, an inversion layer is formed in a portion of the base layer 13 facing the gate electrode 31. Then, a current flows between the source region 14 and the substrate 11 (i.e., the drain layer) through the inversion layer.

[0022] Next, a method for manufacturing the above semiconductor device will be described with reference to FIGS. 3A to 3K.

[0023] First, as shown in Fig. 3A, a GaN wafer 100 having one surface 100a and another surface 100b and in the form of a bulk wafer is prepared. For example, the GaN wafer 100 is doped with silicon, oxygen, germanium, etc., and has an impurity concentration of 5 x 10 17 ~5×10 19 cm -3 The GaN wafer 100 is prepared with a thickness of about 400 μm. The GaN wafer 100 of this embodiment has one surface 100a as a gallium surface and the other surface 100b as a nitrogen surface. After the manufacturing process of the semiconductor device S1 described below is performed, the GaN wafer 100 is prepared by reusing a recycled wafer 240 shown in FIG. 3K, which will be described later.

[0024] 3B and 4, an epitaxial film 110 made of GaN with a thickness of about 10 to 100 μm is formed on one surface 100a of the GaN wafer 100, and a processed wafer 200 is prepared in which a plurality of device formation regions RA are partitioned by dicing lines DL. + Epitaxial layer of type n -The epitaxial layers of the n-type are deposited in sequence. + The epitaxial film of the mold is doped with, for example, silicon, oxygen, germanium, etc., and the impurity concentration is 5×10 17 ~1×10 18 cm -3 It is said to be about that level. - The epitaxial layer of the mold is doped with silicon, for example, and has an impurity concentration of 1×10 16 ~1×10 17 cm -3 In this embodiment, n + The substrate 11 in the semiconductor device S1 is configured to include an epitaxial layer of n - The epitaxial layer of the mold constitutes a drift layer, etc. In this embodiment, each device formation region RA surrounded by the dicing lines DL has a rectangular shape in plan view.

[0025] In the following, the surface of the processed wafer 200 facing the epitaxial film 110 is referred to as one surface 200a of the processed wafer 200, and the surface of the processed wafer 200 facing the GaN wafer 100 is referred to as the other surface 200b of the processed wafer 200. Each device formation region RA is formed on the one surface 200a side of the processed wafer 200.

[0026] Next, as shown in FIG. 3C, a front side process, which is a process for the one surface 200a side of a general semiconductor manufacturing process, is performed. Specifically, as the front side process, ion implantation, deposition, wet process, etc. are appropriately performed to form the above-mentioned base layer 13, source region 14, insulating film 20, gate electrode 31, source electrode, etc. In addition, in this embodiment, a bevel structure 15 is formed in the outer peripheral region 2 of the device formation region RA. In this embodiment, as shown in FIG. 3C and FIG. 5, the bevel structure 15 is formed so as to surround the cell region 1 of each device formation region RA and reach the dicing line DL between each adjacent device formation region RA from each device formation region RA. Note that FIG. 5 is a plan view of the processed wafer 200 with the insulating film 20 and the like omitted.

[0027] Thereafter, as shown in FIG. 3D, a probe needle PN is brought into contact with the semiconductor element formed in each device formation area RA to perform a characteristic inspection of the electrical characteristics, etc., of each semiconductor element.

[0028] 3E, a holding member 300 is placed on the first surface 20a of the processed wafer 200. Before placing the holding member 300, a surface protection film made of a resist or the like is formed on the first surface 200a of the processed wafer 200, as necessary.

[0029] The holding member 300 is, for example, a dicing tape having a base material 310 and an adhesive 320. The base material 310 is made of a material that is not easily warped during the manufacturing process, such as glass, a silicon substrate, or ceramics. The adhesive 320 is made of a material that can change its adhesive strength, such as a material whose adhesive strength changes depending on temperature or light. In this case, the adhesive 320 is made of, for example, an ultraviolet curing resin, wax, or double-sided tape. However, the adhesive 320 is made of a material that maintains its adhesive strength even when forming the drain electrode 32 of FIG. 3I, which will be described later.

[0030] Next, as shown in Fig. 3F, the processed wafer 200 is irradiated with laser light L from the other surface 200b to form a device-use deteriorated layer 210 on the dicing line DL. In this embodiment, a laser device is prepared that includes a laser light source that oscillates the laser light L, a dichroic mirror arranged to change the optical axis of the laser light, a focusing lens for focusing the laser light, and a displaceable stage. When forming the device-use deteriorated layer 210, the processed wafer 200 is placed on the stage, and the position of the stage is adjusted so that the focusing point of the laser light L is relatively scanned along the dicing line DL.

[0031] As a result, the thermal energy causes gallium and nitrogen to decompose, forming a device-use altered layer 210 on the dicing line DL. More specifically, the device-use altered layer 210 is formed by evaporating nitrogen as a gas and precipitating gallium, by irradiating the laser light L. Note that the device-use altered layer 210 is in a state in which minute voids are formed due to the separation of nitrogen.

[0032] In addition, in this embodiment, when forming the device deteriorated layer 210, the stage or the like is moved appropriately, and the laser light L is irradiated so that the focal point moves to two or more different locations in the thickness direction of the processed wafer 200. In this case, the device deteriorated layer 210 is formed at different locations in the thickness direction of the processed wafer 200, and each device deteriorated layer 210 may be separated from each other or may be connected. In addition, when the focal point is moved to two or more different locations in the thickness direction of the processed wafer 200, the focal point is moved from the one surface 200a side to the other surface 200b side of the processed wafer 200.

[0033] The device-use deteriorated layer 210 of this embodiment is formed so that when the wafer-use deteriorated layer 220 of FIG. 3G is formed, nitrogen generated by forming the wafer-use deteriorated layer 220 can be released to the outside through the pores of the device-use deteriorated layer 210. Although not particularly limited, in this embodiment, when the device-use deteriorated layer 210 is formed, a solid-state laser light, a green laser having a wavelength of 532 nm, is used as the laser light L. The laser light L is irradiated with a processing point output of 2 μJ, a pulse width of 500 ps, ​​and a processing speed of 500 mm / s. However, these conditions are only an example, and the inventors have confirmed that the device-use deteriorated layer 210 is appropriately formed even when the processing point output of the laser light L is lower or the pulse width is shorter. The inventors have also confirmed that the device-use deteriorated layer 210 is appropriately formed even when the processing point output of the laser light L is higher or the pulse width is longer.

[0034] 3G, laser light L is applied from the other surface 200b of the processed wafer 200 to form a wafer-use deteriorated layer 220 along the surface direction of the processed wafer 200 at a position at a predetermined depth D from the one surface 200a of the processed wafer 200. In this embodiment, the wafer-use deteriorated layer 220 is formed using the laser device used for forming the device-use deteriorated layer 210.

[0035] When forming the wafer-use damaged layer 220, the position of a stage or the like is adjusted so that the focal point of the laser light L is scanned relatively along the surface direction of the processed wafer 200. As a result, the wafer-use damaged layer 220 is formed along the surface direction on the processed wafer 200. Note that the wafer-use damaged layer 220 is configured such that nitrogen evaporates as a gas and gallium is precipitated, similar to the device-use damaged layer 210.

[0036] In this case, in this embodiment, the wafer-use deteriorated layer 220 is formed so as to intersect with the device-use deteriorated layer 210 or pass directly below the device-use deteriorated layer 210. This makes it easier for nitrogen to escape through the device-use deteriorated layer 210 when forming the wafer-use deteriorated layer 220, and makes it possible to suppress application of large distortion to each device formation region RA when forming the wafer-use deteriorated layer 220.

[0037] The predetermined depth D when forming the wafer altered layer 220 is set according to the ease of handling and on-resistance of the semiconductor device S1, and is set to, for example, about 50 μm. The location where the wafer altered layer 220 is formed varies according to the thickness of the epitaxial film 110, and is formed either inside the epitaxial film 110, at the boundary between the epitaxial film 110 and the GaN wafer 100, or inside the GaN wafer 100. When the wafer altered layer 220 is formed inside the epitaxial film 110 or at the boundary between the epitaxial film 110 and the GaN wafer 100, the substrate 11 becomes n + When the wafer-use damaged layer 220 is formed inside the GaN wafer 100, the substrate 11 is an n +The epitaxial layer 110 and the GaN wafer 100 are formed as a wafer-use altered layer 220 at the boundary between the epitaxial film 110 and the GaN wafer 100. In this embodiment, an example in which a wafer-use altered layer 220 is formed at the boundary between the epitaxial film 110 and the GaN wafer 100 will be described.

[0038] In the following description, the portion of processed wafer 200 on one side 200a side of wafer-deteriorated layer 220 will be referred to as device configuration wafer 230, and the portion of processed wafer 200 on the other side 200b side of wafer-deteriorated layer 220 will be referred to as recycled wafer 240.

[0039] Next, as shown in FIG. 3H, the auxiliary member 400 is placed on the other surface 10b side of the processed wafer 200. As with the holding member 300, the auxiliary member 400 is, for example, a dicing tape having a base material 410 and an adhesive 420. The base material 410 is, for example, made of glass, a silicon substrate, ceramics, etc. The adhesive 420 is, for example, made of ultraviolet curing resin, wax, double-sided tape, etc. Then, the holding member 300 and the auxiliary member 400 are gripped and a tensile force or the like is applied in the thickness direction of the processed wafer 200, and the wafer is divided into the device configuration wafer 230 and the recycled wafer 240 with the wafer deterioration layer 220 as the boundary (i.e., the starting point of the branching).

[0040] At this time, since the divided surface is rough, the device configuration wafer 230 is likely to warp due to the Twyman effect. However, in this embodiment, the bevel structure 15 is formed, and as will be described later in detail, the bevel structure 15 can relieve the stress. Therefore, it is possible to suppress a large difference in the stress applied to the semiconductor element before and after dividing the processed wafer 200.

[0041] In the following description, the surface of the device configuration wafer 230 on which the source region 14 and the like are formed is referred to as one surface 230a, the divided surface of the device configuration wafer 230 is referred to as the other surface 230b, and the divided surface of the recycled wafer 240 is referred to as the one surface 240a. The substrate 11 in the semiconductor device S1 is formed on the other surface 230b side of the device configuration wafer 230, and the portion on the substrate 11 becomes the drain 12 layer. In addition, in each of the figures from FIG. 3H onwards, the wafer alteration layer 220 and the like remaining on the other surface 230b of the device configuration wafer 230 and the one surface 240a of the recycled wafer 240 are omitted as appropriate.

[0042] Thereafter, as shown in FIG. 3I, a general semiconductor manufacturing process is performed, and as the remaining semiconductor manufacturing process, a back surface process is performed to form the drain electrode 32 and the like on the other surface 230b of the device configuration wafer 230.

[0043] Before performing the backside process of the drain electrode 32, etc., a step of planarizing the other surface 230b of the device configuration wafer 230 by a CMP (short for chemical mechanical polishing) method or the like may be performed as necessary. FIG. 3I shows a diagram of the case where the other surface 230b of the device configuration wafer 230 is planarized. After performing the step of forming the drain electrode 32, a heat treatment such as laser annealing may be performed as necessary to make the drain electrode 32 and the other surface 230b of the device configuration wafer 230 into ohmic contact.

[0044] 3J, the holding member 300 is expanded to divide each device formation region RA with the device altered layer 210 as a boundary (i.e., a branching starting point) as the boundary, thereby dividing the semiconductor device S1 into individual devices.

[0045] Thereafter, the adhesive strength of the adhesive 320 is weakened by heat treatment or light irradiation, and the semiconductor device S1 is picked up. In this way, the semiconductor device S1 of this embodiment is manufactured. Before dividing each device formation region RA, slits or the like may be formed in the drain electrode 32 at the boundaries of each device formation region RA, as necessary, so that the drain electrode 32 for each device formation region RA can be easily divided. In this case, in the step of FIG. 3I, a metal mask may be prepared to cover the divided portions so that the drain electrode 32 is not formed in the divided portions.

[0046] The above is the method for manufacturing the semiconductor device in this embodiment.

[0047] The inventors then investigated the threshold voltage of the semiconductor device S1 manufactured by the above manufacturing method before and after dividing the processed wafer 200, and obtained the results shown in Fig. 6. The inventors also used a semiconductor device manufactured without forming the bevel structure 15 in Fig. 3C as a comparative semiconductor device, and investigated the threshold voltage of this semiconductor device before and after dividing the processed wafer 200, and obtained the results shown in Fig. 7. The results of Figs. 6 and 7 can be summarized as shown in Fig. 8.

[0048] The threshold voltage before division is the threshold voltage measured in the state of processed wafer 200, and the threshold voltage after division is the threshold voltage measured in the state of device configuration wafer 230 after dividing processed wafer 200. The results shown in FIG. 6 are the results when, as shown in FIG. 9, width a of bevel structure 15 is 33.3 μm, width b of the portion of one surface 10a sandwiched by bevel structure 15 is 600 μm, depth c of bevel structure 15 is 3.5 μm, and taper angle θ of bevel structure 15 is 6°. The semiconductor device of the comparative example is a semiconductor device manufactured by the same manufacturing method as above, except that bevel structure 15 is not formed. The threshold voltage is 1×10 -9The error bars in Fig. 8 indicate the average ± standard deviation. Fig. 8 also shows 12 samples whose threshold voltages were measured before division and 9 samples whose threshold voltages were measured after division.

[0049] 6 to 8, when comparing this embodiment with the comparative example, it is confirmed that the amount of change in threshold voltage is smaller in this embodiment than in the comparative example before and after dividing the processed wafer 200. For this reason, it can be said that the bevel structure 15 relieves the stress when dividing the processed wafer 200.

[0050] A comparison of the threshold voltage before and after dividing the processed wafer 200 is also shown in Fig. 10. As shown in Fig. 10, in this embodiment and the comparative example, it is confirmed that the difference in the amount of variation in the threshold voltage before and after dividing the processed wafer 200 is 1.2 V.

[0051] Here, the inventors placed a semiconductor device having a MOSFET formed on a support base, performed a four-point bending test, and investigated the relationship between the stress generated and the threshold voltage, obtaining the results shown in Fig. 11. As shown in Fig. 11, the relationship between the compressive stress generated in the semiconductor device and the threshold voltage is expressed by the following formula 1, where the compressive stress is y and the threshold voltage is x.

[0052] [Number 1] y = -0.0035x + 4.6422... (Formula 1) Therefore, the compressive stress relieved by the bevel structure 15 of this embodiment is estimated to be 1.2 / 0.0035 ≈ 343 (MPa) because the difference in the variation amount of the threshold voltage is +1.2 V. In other words, it is confirmed that the bevel structure 15 can relieve the stress generated when the processed wafer 200 is divided.

[0053] Furthermore, the inventors have studied the stress in this embodiment and the comparative example, and obtained the results shown in Fig. 12. Fig. 12 shows the results of stress evaluation by Raman analysis in the state of the device configuration wafer 230. Fig. 12 shows the results when the thickness of the device configuration wafer 230 is set to 50 μm.

[0054] 12, it is confirmed that the stress on the one surface 230a side of the device configuration wafer 230 can be more relaxed in this embodiment than in the comparative example. Therefore, the stress on the one surface 10a side is relaxed in the semiconductor device S1 manufactured by the above manufacturing method.

[0055] According to the present embodiment described above, the bevel structure 15 is formed as a recess in the outer peripheral region 2. Therefore, when the processed wafer 200 is divided to form the device configuration wafer 230, the bevel structure 15 can relieve stress. In other words, it is possible to prevent the stress applied to the semiconductor element from significantly differing before and after dividing the processed wafer 200. Therefore, it is possible to easily maintain the accuracy of the characteristic inspection performed before dividing the processed wafer 200.

[0056] (1) In this embodiment, the bevel structure 15 is formed so as to surround the cell region 1. Therefore, the stress that may be applied from the outer periphery region 2 to the cell region 1 can be uniformly relaxed.

[0057] Second embodiment A second embodiment will be described. In this embodiment, a trench is formed instead of the bevel structure 15 of the first embodiment. As the rest is similar to the first embodiment, a description thereof will be omitted here.

[0058] 13, in the semiconductor device S1 of this embodiment, a trench 16 is formed as a recess instead of the bevel structure 15. Note that, similar to the first embodiment, the trench 16 is formed in the outer periphery region 2 so as to surround the cell region 1. Also, the depth of the trench 16 is set to be less than the thickness of the semiconductor substrate 10.

[0059] Such a semiconductor device is manufactured by forming a trench 16 instead of the bevel structure 15 in the process of FIG. 3C.

[0060] According to the present embodiment described above, the trench 16 is formed as a recess in the outer circumferential region 2. Therefore, it is possible to obtain the same effects as those of the first embodiment.

[0061] Third embodiment A third embodiment will be described. In this embodiment, a filling material is disposed in the trench 16 in comparison with the second embodiment. As the rest is similar to the second embodiment, a description thereof will be omitted here.

[0062] 14, in the semiconductor device of this embodiment, a trench 16 is filled with a filler 17. For example, an oxide film, a polysilicon, a metal film, or the like is used as the filler 17. When the filler 17 is made of a metal film, an insulating film may be disposed along the wall surface of the trench 16.

[0063] According to the present embodiment described above, the trench 16 is formed as a recess in the outer circumferential region 2. Therefore, it is possible to obtain the same effects as those of the second embodiment.

[0064] (1) In this embodiment, the filler 17 is embedded in the trench 16. This makes it possible to prevent foreign matter from entering the trench 16, and to prevent the foreign matter from causing a change in the characteristics of the semiconductor device S1.

[0065] (Other embodiments) Although the present disclosure has been described based on the embodiment, it is understood that the present disclosure is not limited to the embodiment or structure. The present disclosure also encompasses various modifications and modifications within the equivalent range. In addition, various combinations and forms, as well as other combinations and forms including only one element, more than one element, or less than one element, are also within the scope and concept of the present disclosure.

[0066] For example, in each of the above embodiments, an example has been described in which a vertical n-channel MOS transistor is formed as the semiconductor element. However, the semiconductor element may be a p-channel MOS transistor or may be horizontal. In addition, the semiconductor element may be an optical semiconductor element such as a light-emitting diode, a semiconductor laser, or the like.

[0067] In addition, in each of the above-described embodiments, an example has been described in which the device-use deteriorated layer 210 is formed by irradiating the laser light L. However, for example, instead of the device-use deteriorated layer 210, a groove may be formed by using a dicing blade or the like.

[0068] Furthermore, in each of the above embodiments, an example has been described in which the device-use deteriorated layer 210 is formed before forming the wafer-use deteriorated layer 220. However, for example, the device-use deteriorated layer 210 may not be formed, the wafer-use deteriorated layer 220 may be formed, the processed wafer 200 may be divided into the device configuration wafer 230 and the recycled wafer 240, and the device configuration wafer 230 may be subjected to the remaining steps before being diced into individual pieces using a dicing blade or the like.

[0069] In the above first embodiment, an example has been described in which the bevel structure 15 is formed so as to surround the cell region 1. However, the bevel structure 15 may be formed along one of two opposing sides of the outline of the planar shape of the semiconductor substrate 10 as shown in Fig. 15, for example, and may not be provided in the vicinity of the other of the two opposing sides. [Explanation of symbols]

[0070] 10. Semiconductor Substrate 10a one side 10b Other side 15 Bevel structure (concave) 16 Trench (recess)

Claims

1. A semiconductor device comprising: A semiconductor substrate (10) is provided with a semiconductor element formed thereon, the semiconductor substrate (10) has one surface (10a) and another surface (10b) opposite to the one surface, and is made of gallium nitride; The semiconductor substrate has a cell region (1) in which the semiconductor element is formed and a peripheral region (2) surrounding the cell region, The semiconductor device has a recess (15, 16) formed in the outer periphery region, the one surface being recessed.

2. The semiconductor substrate has a rectangular shape in plan view, The semiconductor device according to claim 1 , wherein the recess is formed along two opposing sides of an outline of the planar shape of the semiconductor substrate.

3. The semiconductor device according to claim 1 , wherein the recess is formed along each side of an outline of the planar shape of the semiconductor substrate.

4. 4. The semiconductor device according to claim 1, wherein the recess has a bevel structure in which an outer periphery of the one surface is recessed.

5. 4. The semiconductor device according to claim 1, wherein the recess is formed by a trench (16).

6. 6. The semiconductor device of claim 5, wherein the trench is filled with a fill material (17).

7. A semiconductor substrate (10) is provided with a semiconductor element formed thereon, the semiconductor substrate (10) has one surface (10a) and another surface (10b) opposite to the one surface, and is made of gallium nitride; The semiconductor substrate has a cell region (1) in which the semiconductor element is formed and a peripheral region (2) surrounding the cell region, A method for manufacturing a semiconductor device, in which a recess (15, 16) is formed in the outer peripheral region by recessing the one surface, comprising: Preparing a processed wafer (200) having one surface (200a) and another surface (200b) opposite to the one surface, made of gallium nitride, and having a plurality of device formation regions (RA); forming the semiconductor element on one surface of the processed wafer; performing a characteristic inspection of the semiconductor device; After the characteristic inspection, a laser beam (L) is irradiated from the other surface side of the processed wafer to the inside of the processed wafer to form an altered layer (220) along a surface direction of the processed wafer inside the processed wafer; Dividing the processed wafer at the deteriorated layer as a boundary, thereby dividing the processed wafer into a device configuration wafer (230) on one side of the processed wafer and a recycled wafer (240) on the other side of the processed wafer; and singulating the device configuration wafer; The method for manufacturing a semiconductor device includes forming the recess on an outer edge of the device formation region, in forming the semiconductor element.

8. 8. The method for manufacturing a semiconductor device according to claim 7, wherein a bevel structure (15) is formed as the recess by forming the semiconductor element.

9. 8. The method for manufacturing a semiconductor device according to claim 7, wherein a trench (16) is formed as the recess by forming the semiconductor element.

Citation Information

Patent Citations

  • Semiconductor device

    JP2017168515A

  • Semiconductor device and method for manufacturing the same

    JP2020013861A

  • Semiconductor chip and manufacturing method thereof

    JP2021170594A

  • Semiconductor device and manufacturing method thereof

    JP2022100379A

  • Semiconductor device

    JP2023037280A