Semiconductor device and method for its manufacture

The self-alignment process for forming n-type and p-type regions in SiC MOSFETs addresses registration errors, resulting in high-performance devices with stable breakdown voltage and low ON resistance.

DE102017212709B4Active Publication Date: 2024-11-28HITACHI LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
DE102017212709
Authority / Receiving Office
DE · DE
Patent Type
Patents
Current Assignee / Owner
Priority Date
2016-07-25
Filing Date
2017-07-25
Publication Date
2024-11-28
Estimated Expiration
2037-07-25

AI Technical Summary

Technical Problem

The manufacturing of silicon carbide (SiC) power metal oxide semiconductor field effect transistors (MOSFETs) faces challenges due to registration errors in lithography steps, leading to fluctuations in device performance and the inability to maintain a predetermined breakdown voltage while achieving low ON resistance.

Method used

A self-alignment process is employed to form n-type source and p-type body regions using a single lithography step, with impurity implantation directions inclined relative to the substrate normal, ensuring precise channel length and reduced ON resistance.

Benefits of technology

This method enables the production of high-performance SiC MOSFETs with both high breakdown voltage and reliability by minimizing registration errors and optimizing channel length without increasing ON resistance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 00000000_0000_ABST
    Figure 00000000_0000_ABST
Patent Text Reader

Abstract

Semiconductor device comprising an active region (1002) in which several unit cells are arranged regularly, each comprising the unit cells: a channel region formed over a front face of a semiconductor substrate and exhibiting a first conductivity type; a source region (211) having a second conductivity type different from the first conductivity type, formed above the front face of the semiconductor substrate so that it is in contact with the channel region; and a JFET region (204) having the second conductivity type and is formed above the front of the semiconductor substrate on the opposite side of the channel region from the source region (211), so that it is in contact with the channel region; wherein the canal area comprises a first canal area and a second canal area with a higher concentration of faults than the first canal area; and wherein the first channel region and the second channel region are arranged along a first and a second direction respectively to surround outer edge sections of the source region (211) of each unit cell in a top view.
Need to check novelty before this filing date? Find Prior Art

Description

Background of the invention 1. Field of the invention

[0001] The present disclosure relates to a semiconductor device and a method for manufacturing the same.

[0002] In particular, the present disclosure relates to a technology effective when applied to a semiconductor device including a power semiconductor switching device using a wide band gap semiconductor and a method for manufacturing the same. 2. Description of the state of the art

[0003] In recent years, in light of problems such as global warming and concerns about fossil fuel supplies, attention has been drawn to power electronics as a technology for promoting the effective use of energy. Power electronics devices are responsible for converting and controlling electrical power, and there is a need to improve the performance of power semiconductor devices, which are a key component of power electronics devices.

[0004] Silicon (Si) substrates have long been used for power semiconductor devices. Improvements have been made to Si power semiconductor devices to achieve lower loss and higher performance. As a result, their device performance has approached theoretical limits determined by the properties of the Si material, making it difficult to achieve further performance improvements in the future.

[0005] Under such circumstances, research into low-loss power semiconductor devices using silicon carbide (SiC) or gallium nitride (GaN) has been actively conducted. Silicon carbide (SiC) and gallium nitride (GaN) exhibit a dielectric breakdown field strength approximately one order of magnitude higher than silicon (Si), allowing for a thinner drift layer. Therefore, power semiconductor devices using the wide-bandgap semiconductor can exhibit lower ON resistance and thus lower loss compared to Si power semiconductors.

[0006] As an example of the conventionally used power metal oxide semiconductor field effect transistor (power MOSFET), a vertical double diffused MOSFET (DMOSFET) is described below.

[0007] Fig. Figure 1 is a plan view of a typical DMOSFET semiconductor chip. A terminal end region 1001 is provided in an outer peripheral portion of the semiconductor chip. An active region 1002 and a gate pad region 1003 are provided on the interior of the terminal end region 1001. A plurality of unit cells constituting the MOSFET structure are regularly arranged in the active region 1002.

[0008] Fig. 2 is a cross-sectional view of a main part of a unit cell of a typical vertical DMOSFET structure. A main surface of an n+ -type SiC substrate 101 is provided on an epitaxial layer 102, also made of SiC, and a gate insulating film 221 and a gate electrode 222 are formed over the epitaxial layer 102. The front and side surfaces of the gate electrode 222 are covered with an interlayer insulating film 231. A source electrode 232 is formed in contact with the front surface of the epitaxial layer 102 in a source contact region 230 opened in the interlayer insulating film 231. A drain electrode 103 is provided on the back surface of the SiC substrate 101.

[0009] A p-type body region 201 is formed near the front surface of the epitaxial layer 102. An n-type source region 211 and a p+-type body contact region 202, which has p-type impurities at a higher concentration than the p-type body region 201, are formed within the p-type body region 201. The source electrode 232 is formed in contact with and in electrical connection with the front surfaces of the n-type source region 211 and the p+-type body contact region 202. Therefore, the n-type source region 211 and the p+-type body contact region 202 are electrically short-circuited by the source electrode 232.

[0010] When a positive voltage is applied to the gate electrode 222, a channel region is formed in this portion of the p-type body region 201, which is contacted by the gate insulating film 221 on the front side, and electrons flow through the channel from the n-type source region 211 to the drain electrode 102. Thus, by applying a voltage to the gate electrode 222, a switching operation is performed. A plurality of unit cells are regularly and densely arranged in the active region while sharing the source electrode 232 and are electrically connected in parallel. By increasing the number of unit cells connected in parallel or the number of unit cells superimposed over the active region and increasing the width of the channel regions arranged in the active region, a reduction in the resistance of the MOSFET chip as a whole can be realized.

[0011] Fig. Figure 3 shows an example of a top view layout of the unit cells arranged in the active region. Typically, the Fig. 3A and Fig. 3B shown layouts are used. In the Fig. 3A and Fig. 3B, a top view of the n-type source regions 211, the channel regions at the front of the p-type body regions 201, the p+ type body contact regions 202, and the source contact regions 230 formed over the front of the epitaxial layer 102 of the DMOSFET structure are shown.

[0012] The unit cell 1011 in Fig. 3A has an elongated ribbon-like shape, and the channel region is formed along the longitudinal direction of the ribbon. The sectional structure of the unit cell 1011 shown in Fig. 2, forms the section along a cutting plane line XX' of Fig. 3A off.

[0013] The unit cell in Fig. 3B has a quadrilateral shape, and the channel region is formed along the four edges of the quadrilateral. The sectional structure of the unit cell shown in Fig. 2, forms the section along a cutting plane line XX' and a cutting plane line YY' of Fig. 3B. In the square cells, the length of the channel regions arranged in the active region can be increased compared to the ribbon-shaped cells, so that the square cell is advantageous from the point of view of reducing the resistance in a conduction state. Summary of the invention

[0014] When manufacturing a SiC DMOSFET, the DMOSFET structure, such as the n-type source region and the p-type body region, is formed by selective ion implantation. In impurity ion implantation, lithography is typically used for the n-type source region and the p-type body region to perform implantation while using separate masks.

[0015] When using separate masks for ion implantation for the n-type source regions and the p-type body regions, mask registration in the two lithography steps poses a manufacturing challenge. When a stepper is used, a registration error of 0.1 to 0.3 µm is typically generated. Additionally, the size (pattern) of the resist mask formed after development may shift due to small variations in ion dose or temperature. Since the channel length is typically less than 1 µm, variations and fluctuations in device performance resulting from registration errors and pattern shift are not negligible.

[0016] Additionally, although a shorter channel length is preferable for obtaining a lower ON resistance, a channel length that is too short would induce punch-through, resulting in a defect that cannot sustain a predetermined breakdown voltage. Therefore, an average channel length should be designed with a sufficient edge clearance to prevent punch-through from occurring even with the shorter channel length due to matching errors. However, providing such an edge clearance would result in an increase in ON resistance and switching loss.

[0017] To address the aforementioned registration error problem, a so-called self-alignment process has been proposed, in which the body regions and source regions are formed through a single lithography step. According to the self-alignment process, the channel length can be determined without being affected by registration errors or pattern shifts that might result from the twice-performed ion irradiation as described above, thus enabling the fabrication of a MOSFET with a short channel and low on-resistance without fluctuations.

[0018] As an example of the self-alignment process, JP 2009-147381 A discloses a method for manufacturing a vertical MOSFET in which ion implantation for forming source regions and ion implantation for forming channel regions are performed using the same mask. Furthermore, US 2015 / 0084066 A1 discloses a silicon carbide MOSFET in which a particularly low on-resistance can be achieved. Specifically, said MOSFET has N+, P-well, and JFET regions extending in one direction (Y-direction), as well as P+ and source contacts extending in an orthogonal direction (X-direction). A polysilicon gate of the device covers the JFET region and is terminated above the P-well region to minimize the electric field at the polysilicon gate edge. However, the manufacturing process described in JP 2009-147381 A has the following problems.

[0019] According to the method of JP 2009-147381 A, using a mask material having a tapered shape, n-type impurities for a source region are implanted in the direction normal to a substrate surface, and further, p-type impurities are implanted in a direction inclined relative to the normal of the substrate to form a channel in the shape of the penetration below the mask. To form the channel by implanting ions deep below the mask, it is efficient to implant the ions in a direction perpendicular to the contour of an opening portion in the mask. This approach is suitable for a ribbon-type unit cell as shown in Fig. 3A, effective. However, for example, in the case of a quadrangular unit cell, as shown in Fig. As shown in Figure 3B, if ions are implanted at the four edges of the quadrilateral in directions perpendicular to the respective edges, the impurities are not necessarily sufficiently implanted into the corner portions of the quadrilateral. For example, if the corner portions have a low impurity concentration, a problem would arise that a predetermined breakdown voltage cannot be maintained.

[0020] Thus, there is a need for a silicon carbide semiconductor device having lower on-resistance and high reliability, and a simpler method for manufacturing the silicon carbide semiconductor device. The above and other objects of the present disclosure will become apparent from the following description of embodiments with reference to the accompanying drawings.

[0021] In one aspect of the present disclosure, a semiconductor device is disclosed having an active region in which a plurality of unit cells are regularly arranged, each of the unit cells including: a channel region formed over a front side of a semiconductor substrate and having a first conductivity type; a source region having a second conductivity type different from the first conductivity type and formed over the front side of the semiconductor substrate so as to be in contact with the channel region; and / or a JFET region having the second conductivity type and formed over the front side of the semiconductor substrate on an opposite side of the channel region from the source region so as to be in contact with the channel region.The channel region includes a first channel region and a second channel region having a higher impurity concentration than the first channel region, the first channel region and the second channel region being arranged along first and second directions, respectively, to surround outer edge portions of the source region of each unit cell in a plan view.

[0022] In another aspect of the present disclosure, a method for manufacturing a semiconductor device is disclosed, including: a step of forming a mask over a semiconductor substrate; a channel implantation step of implanting impurities of a first conductivity type, with the mask as a shielding film; and / or a source implantation step of implanting impurities of a second conductivity type different from the first conductivity type, with the mask as a shielding film. The channel implantation step is performed in a direction inclined relative to a normal of the semiconductor substrate and includes implanting four times with each rotation of the semiconductor substrate by 90 degrees.

[0023] According to the semiconductor device disclosed herein, a high-performance vertical DMOSFET with both a high breakdown voltage and high reliability can be provided. In addition, according to the method of manufacturing a semiconductor device disclosed herein, a high-performance MOSFET in which a source region and a base region are formed on the self-alignment basis can be manufactured by using a mask formed by one-time lithography. Brief description of the drawings Fig. Figure 1 is a schematic top view of a typical SiC power MOSFET chip; Fig. 2 is a cross-sectional view of a main part of a unit cell of the typical SiC power MOSFET; Fig. 3A is a plan view showing an example of a unit cell of a SiC power MOSFET according to the prior art. Fig. 3B is a plan view showing an example of a unit cell of a SiC power MOSFET according to the prior art. Fig. 4 is a plan view showing a layout of unit cells of a SiC power MOSFET according to Embodiment 1 of the present disclosure. Fig. 5A is a sectional view of a unit cell taken along a section plane X1-X1' of the SiC power MOSFET according to Embodiment 1; Fig. 5B is a sectional view of the unit cell taken along a section plane X2-X2' of the SiC power MOSFET according to Embodiment 1; Fig. 5C is a sectional view of the unit cell taken along a section plane X3-X3' of the SiC power MOSFET according to Embodiment 1; Fig. 5D is a sectional view of the unit cell in a section plane X4-X4' of the SiC power MOSFET according to Embodiment 1; Fig. 6A is a sectional view of the unit cell taken along the X1-X1' plane in a step of manufacturing the SiC power MOSFET according to Embodiment 1; Fig. 6B is a sectional view of the unit cell in a section plane X1-X1' in a step following Fig. 6A of the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. Figure 7A is a sectional view of the unit cell in the section plane X1-X1' in a step following Fig. 6B shows the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. Figure 7B is a sectional view of the unit cell in the section plane X2-X2' in a step following Fig. 6B shows the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. Figure 7C is a sectional view of the unit cell in the section plane X3-X3' in a step following Fig. 6B shows the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 7D is a sectional view of the unit cell in the section plane X4-X4' in a step following Fig. 6B shows the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 7E is a plan view showing a layout of the unit cells in a step following Fig. 6B shows the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. Figure 8A is a sectional view of the unit cell in the section plane X1-X1' in a step following Fig. 7A of the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 8B is a plan view showing a layout of the unit cells in a step following Fig. 7E of the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 8C is a plan view showing a layout of the unit cells in a step following Fig. 8B shows the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 8D is a plan view showing a layout of the unit cells in a step following Fig. 8C shows the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 8E is a plan view showing a layout of the unit cells in a step following Fig. 8D shows the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 9A is a sectional view of the unit cell in the section plane X1-X1' in the step following Fig. 8D shows the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 9B is a sectional view of the unit cell in the section plane X2-X2' in the step following Fig. 8D shows the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 9C is a sectional view of the unit cell in the section plane X3-X3' in the step following Fig. 8D shows the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 9D is a sectional view of the unit cell in the section plane X4-X4' in the step following Fig. 8D shows the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 10A is a sectional view of the unit cell in the section plane X1-X1' in a step following Fig. 9A of the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 10B is a sectional view of the unit cell in the section plane X1-X1' in a step following Fig. 10A of the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 10C is a sectional view of the unit cell in the section plane X1-X1' in a step following Fig. 10B shows the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 10D is a sectional view of the unit cell in the section plane X1-X1' in a step following Fig. 10C of the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 10E is a sectional view of the unit cell in the section plane X1-X1' in a step following Fig. 10D shows the manufacturing method of the SiC power MOSFET according to Embodiment 1; Fig. 11 is a plan view showing a layout of unit cells of a SiC power MOSFET according to Embodiment 2 of the present disclosure. Fig. 12 is a plan view showing a layout of unit cells of a SiC power MOSFET according to Embodiment 3 of the present disclosure. Fig. 13 is a plan view showing a layout of unit cells of a SiC power MOSFET according to Embodiment 4 of the present disclosure. Fig. 14 is a plan view showing a layout of unit cells of a SiC power MOSFET according to Embodiment 5 of the present disclosure; and Fig. 15 is a plan view showing the layout of unit cells in a step of a manufacturing method of the SiC power MOSFET according to Embodiment 5. Description of the preferred embodiments

[0024] Some embodiments of the present disclosure will be described in detail below with reference to the drawings. It should be noted that the present disclosure should not be construed as being limited to the contents of the descriptions of the following embodiments. It is readily understood by one skilled in the art that specific configurations of the present disclosure may be modified without departing from the spirit or gist of the present disclosure.

[0025] In the configurations of the present disclosure described below, parts having the same or equivalent functions are generally denoted by the same reference numerals in the different drawings, and overlapping descriptions thereof may be omitted.

[0026] The terms "first," "second," "third," and the like in this specification are used to identify configuration components and do not necessarily limit their number or order. Additionally, numbers are used to identify configuration components based on context, and a number used in one context does not necessarily indicate the same configuration in other contexts. Furthermore, a configuration component identified by one number should not be prevented from also functioning as a configuration component identified by a different number.

[0027] The position, size, shape, area, and the like of each configuration shown in the drawings may not represent an actual position, size, shape, area, and the like for the sake of easy understanding of the present disclosure. Accordingly, the present disclosure is not necessarily limited to the positions, sizes, shapes, areas, and the like disclosed in the drawings.

[0028] A configuration component expressed herein in a singular form includes a plural form unless clearly indicated otherwise in a specific context. Embodiment 1 <<1. Structure of the silicon carbide semiconductor device according to Embodiment 1>>

[0029] The structure of a silicon carbide semiconductor device according to Embodiment 1 of the present disclosure will be described below with reference to Fig. 4 and Fig. 5A to Fig. 5D described.

[0030] Fig. 4 is a plan view of a plurality of unit cells arranged in an active region of a SiC power MOSFET in the present disclosure.

[0031] The Fig. 5A to Fig. 5D are sectional views of a main part of the SiC power MOSFET. Sectional views taken along the cutting plane lines X1-X1', X2-X2', X3-X3', and X4-X4' shown in Fig. 4 are shown, are taken, correspond to the Fig. 5A , Fig. 5B , Fig. 5C or Fig. 5D . In addition, a terminal end region is provided on the outside of an active region over which the unit cells are supported.

[0032] One in Fig. The sectional structure shown in Figure 5A will be described. An epitaxial layer 102 containing n-type SiC having a specific dopant concentration and a specific film thickness is provided on a main surface of an n+ type SiC substrate 101. A gate insulating film 221 is provided over a part of a front surface of the epitaxial layer 102, and a gate electrode 222 is provided on the gate insulating film 221. Furthermore, a source electrode 232 is provided over a part of the front surface of the epitaxial layer 102. In addition, a drain electrode 103 containing metal is provided on a back surface of the SiC substrate 101.

[0033] A p-type channel region 201a is formed near the front surface of the epitaxial layer 102. The n-type source region 211 and a p+-type body contact region 202, which has p-type impurities at a higher concentration than that in a p-type body region 201, are formed within the p-type channel region 201a. The p-type body region 201 is formed on the bottom surface of the n-type source region 211 and the p+-type body contact region 202 at a position spaced from the front surface of the epitaxial layer 102 and is electrically connected to the source electrode 232 via the p+-type body contact region 202, which similarly has a p-type conductivity type. The unit cells 1011 are each essentially square in shape.

[0034] As shown in the top view in Fig. As shown in Figure 4, the n-type source regions 211, the p-type channel regions 201a, 201b, and the p-type body regions 201 of the adjacent unit cells 1011 are not spaced apart from each other, but are connected to each other. The channel regions 201a and 201b are provided in such a manner as to surround edge portions of the n-type source region 211, and the channel region includes the first channel region 201a, which has a low impurity concentration, and the second channel region 201b, which has a higher impurity concentration than the first channel region 201a. In addition, the first channel regions 201a and the second channel regions 201b form a contour portion of a closed region, and the interior of this closed region is a JFET region 204. Fig. 5B , Fig. 5C and Fig. 5D show sectional views taken along section plane lines. Fig. 5B and Fig. 5C contain both the first channel region 201a and the second channel region 201b.

[0035] Here, the JFET region refers to an n-type region tightly sandwiched between the p-type body regions 201 in a vertical MOSFET. Specifically, this is a portion where the n-type epitaxial layer 102 is left on the substrate and where the concentration of n-type impurities in the epitaxial layer 102 may be slightly increased. This serves the purpose of lowering the resistance, given that this region is tightly sandwiched and therefore current is difficult to flow through. On an operating basis, when a positive drain voltage is imposed, a depletion layer spreads to the JFET region side. As shown in Fig. As shown in Figure 4, in this embodiment, the JFET region 204 is surrounded by the channel regions 201a and 201b. Therefore, the channel width can be made long, which is advantageous from the perspective of reducing the resistance in a conduction state. <<2. Method for manufacturing the silicon carbide semiconductor device according to Embodiment 1>>

[0036] A method of manufacturing the silicon carbide semiconductor device according to Embodiment 1 of the present disclosure will be described in the sequence of steps with reference to the Fig. 6A to Fig. 10E. The following description mainly describes the sections corresponding to the section of Fig. 5A, and if necessary, the cuts corresponding to the Fig. 5B , Fig. 5C and Fig. 5D and the top view, the Fig. 4, is also shown. <<2-1. Forming the Body Region>>

[0037] As shown in a sectional view in Fig. As shown in Figure 6A, an epitaxial layer 102 containing SiC having an n- type conductivity is formed on the main surface of the n+ type 4H-SiC substrate 101. N-type impurities are introduced into the n+ type SiC substrate 101. The n-type impurities are, for example, nitrogen (N), and the impurity concentration of these n-type impurities is, for example, 1 × 1018 to 1 × 1021 cm-3.

[0038] The epitaxial layer 102 containing SiC can be formed on the front side (the first main surface) of the SiC substrate 101, for example, by epitaxy. The epitaxial layer 102 has a predetermined thickness and a predetermined dopant concentration determined according to the device specifications. The thickness of the epitaxial layer 102 is, for example, in the range of 3 to 30 µm. In addition, the n-type dopant added to the epitaxial layer 102 is, for example, nitrogen, and the dopant concentration is, for example, in the range of 1 × 1014 to 1 × 1017 cm-3.

[0039] As in Fig. As shown in Figure 6B, a mask material is applied to the front surface of the n-type epitaxial layer 102, and patterning is performed to form a mask 301, after which p-type impurities 601 are implanted so that the p-type body region 201 is formed in such a way that it is embedded in the epitaxial layer 102 at a position spaced from the front surface of the epitaxial layer 102.

[0040] In the case where a photoresist is used as the material for the mask 301, the mask can be formed by applying the photoresist and then patterning the applied photoresist by a known lithography method.

[0041] In the case where SiO2 is used as the material for the mask, SiO2 is deposited, after which a photoresist is further applied, and a resist pattern is formed by a known lithography method. Further, with the resist pattern as an etching mask, SiO2 is etched, for example, by a reactive ion etching method, after which the photoresist is removed, whereby the mask can be obtained. The thickness of the mask 301 is a thickness sufficient to shield the implantation of ions and can be, for example, 1.0 to 5.0 μm.

[0042] As the p-type impurities to be implanted, aluminum (Al) or boron (B), for example, can be used. This allows the p-type body region 201 to be formed in a device formation region of the epitaxial layer 102. The depth of the end on the bottom side of the p-type body region 201 from the front side of the epitaxial layer 102 can be, for example, about 0.5 to 2.0 μm. The depth of the end on the front side of the p-type body region 201 from the front side of the epitaxial layer 102 is, for example, about 0.2 to 0.5 μm, and the concentration of the p-type impurities at the front side is equal to or less than 1 × 1017 cm-3. In addition, the dopant concentration of the p-type body region 201 is, for example, in the range of 1 × 1016 to 1 × 1019 cm-3 . <<2-2. Forming the Source Area>>

[0043] As in Fig. Next, as shown in Figure 7A, after the mask 301 is removed, another mask 311 is further formed, and ions of n-type impurity 701 are implanted to form the n-type source region 211. Nitrogen (N) or phosphorus (P) can be used as the n-type impurity. The impurity concentration of the n-type source region 211 can be, for example, in the range of 1 × 1017 to 1 × 1021 cm-3. The depth of the n-type source region 211 from the front surface of the epitaxial layer 102 can be, for example, about 0.01 to 0.2 μm.

[0044] The Fig. 7B to Fig. 7E show a state at the time of the step of Fig. 7A . The top view is as in Fig. 7E. A portion covered with the mask 311 is seen, and the n-type source region 211 formed is seen because it is not covered with the mask 311. Sectional views taken along the cutting plane lines X1-X1', X2-X2', X3-X3', and X4-X4' shown in Fig. 7E are shown, are taken, the Fig. 7A , Fig. 7B , Fig. 7C or Fig. 7D . Since the p-type body region 201 is formed by being embedded in the epitaxial layer 102 at a position spaced from the front surface of the epitaxial layer 102, the p-type body region 201 is not shown in the plan view. <<2-3. Forming the canal area>>

[0045] As in Fig. 8A(a) in a section X1-X1', while the same mask 311 as in the step of Fig. 7, the p-type impurities 801 are implanted in a direction inclined to X1 from the normal to the substrate surface. Specifically, in the step of Fig. 7 the n-type impurities 701 are implanted perpendicular to a substrate surface 802, as in Fig. 8A(b), on the other hand, in the step of Fig. 8A(a) the p-type impurities 801 are implanted at an acute angle θ with respect to the substrate surface 802 in the section X1-X1', as in Fig. 8A(c). Here, the angle θ is defined as an angle on the side opposite to the mask 311, which is selected from angles formed between the implantation direction of the p-type impurities 801 and the substrate surface 802. This ensures that the p-type impurities 801 penetrate into a part of the portion shielded by the mask 311 to form a p-type channel region (first channel region) 201a. Here, in the channel implantation step, the tilt angle formed between the implantation direction and the normal to the semiconductor substrate is set to 15 to 45 degrees. In other words, θ is equal to 45 to 75 degrees.

[0046] Fig. Figure 8B is a plan view showing the results of impurity implantation of Fig. 8A shows. As in Fig. As shown in Figure 8B, the first channel region 201a is formed in the vertical direction to the paper surface along one of the edges of the n-type source region 211. The first channel region 201a is formed under the mask 311.

[0047] Aluminum (Al) or boron (B), for example, can be used as the p-type impurities to be implanted. The tilt angle of the implantation can be 15 to 45 degrees. Furthermore, to ensure that the impurities reach the epitaxial layer 102 through the mask, the acceleration energy for implantation is desirably 300 to a maximum of 1,500 keV.

[0048] Next, the p-type impurities are implanted in a direction inclined to Y from the normal to the substrate surface at the same tilt angle as above.

[0049] Fig. Figure 8C depicts the results of the p-type impurity implantation. By this implantation, the first channel region 201a is newly formed under the mask 311 in the horizontal direction on the paper surface along another one from the edges of the n-type source region 211, in addition, the impurities are also formed in a part of the first channel region 201a that has already been Fig. 8B is formed in an overlapping manner to form the second channel region 201b having a higher concentration.

[0050] The Fig. 8D and Fig. 8E show changes by the subsequent steps. Following the above, the p-type impurities are implanted in a direction inclined to X1' from the normal to the substrate surface at the same inclination angle as above, resulting in the state of Fig. 8D. Further, the p-type impurities are subsequently implanted in a direction inclined to Y' from the normal to the substrate surface at the same inclination angle as above, resulting in the state of Fig. 8E leads.

[0051] The Fig. 9A to Fig. 9D show sectional views taken along section plane lines of Fig. 8E. The sectional views taken along the section plane lines X1-X1', X2-X2', X3-X3' and X4-X4' shown in Fig. 8E are shown, are taken, respectively the Fig. 9A to Fig. 9D. In the second and subsequent ion implantation steps, some portions are subjected to ion implantation in such a way as to overlap with a portion of the already formed first channel region 201a, so that they become the second channel regions 201b having a higher impurity concentration. As a result, the impurity concentration of the second channel region 201b is approximately 1.6 to 2.5 times the impurity concentration of the first channel region 201a.

[0052] In the four implantation steps, only the direction of the tilt is different, and other conditions such as the tilt angle, the type of implanted ions, the acceleration energy, and the dose can be the same. It should be noted here that, however, in the case where the implantation depth varies depending on the crystal orientation of the substrate surface, the implantation conditions such as the tilt angle can be appropriately changed according to the direction, thereby controlling the implantation results so that the lengths and impurity concentrations of the channels formed in the four directions are uniform. Through the above steps, the source regions and the channel regions are formed above the substrate surface. <<2-4. Forming the body contact area>>

[0053] Next, after the mask 311 is removed, another mask 321 is formed, ions of the p-type impurity 1101 are implanted using the mask 321 as a shielding film to form the p+ type contact region 202.

[0054] Fig. Fig. 10A shows the manner when the p+ type body contact region 202 is formed, in the sectional view along X1-X1' of Fig. 9A. The p+ -type contact region 202 is formed in contact with an inner side surface of the n-type source region 211. Aluminum (Al) or boron (B) can be used as the p-type impurities. The impurity concentration of the p+ -type contact region 202 is, for example, in the range of 1 × 1019 to 1 × 1021 cm-3. The depth of the p+ -type contact region 202 from the front surface of the epitaxial layer 102 is, for example, about 0.1 to 0.4 µm.

[0055] After the mask 321 of Fig. 10A has been removed, as in Fig. As shown in Figure 10B, a heat treatment is performed, thereby activating the ion-implanted impurities. Although omitted from the figure, surface coating films containing carbon (C) having a thickness of, for example, about 0.05 μm may be formed on the front surface of the epitaxial layer 102 and the back surface of the SiC substrate before the activating heat treatment. These surface coating films have the effect of preventing the front surface of the epitaxial layer 102 and the back surface of the SiC substrate from becoming roughened at the time of the activating heat treatment. After the activating heat treatment, the surface coating films are removed, for example, by an oxygen plasma treatment. <<2-5. Forming the electrode and other>>

[0056] With reference to the Fig. 10C to Fig. 10E the following steps are shown in the X1-X1' section view of Fig. 10B described.

[0057] As in Fig. 10C, the gate insulating film 221 and the gate electrode film 222 doped with the n-type impurities are formed over the front surface of the epitaxial layer 102. The gate insulating film 221 includes a silicon dioxide (SiO2) film formed by, for example, a thermal chemical vapor deposition (CVD) method. The thickness of the gate insulating film 221 may be, for example, 0.02 to 0.2 µm. The gate electrode film 222 includes, for example, an n-type polycrystalline silicon film formed by a thermal CVD method. The thickness of the gate electrode film 222 is, for example, about 0.2 to 0.5 µm. In addition, the polycrystalline silicon can be deposited in a polycrystalline state or can be deposited in an amorphous state and then converted to polycrystalline by heat treatment.

[0058] As in Fig. 10D, a mask 321 is formed, and the gate electrode 222 is processed by a dry etching process.

[0059] Subsequently, the mask 321 is removed, after which the interlayer insulating film 231 is formed, for example, by a plasma CVD method, in such a way as to cover the gate electrode 222 and the gate insulating film 221. Further, using a mask 331, the interlayer insulating film 231 and the gate insulating film 221 are processed by dry etching to form a contact hole reaching a part of the front surface of the n-type source region 211 and the front surface of the p+-type body contact region 202.

[0060] Fig. Figure 10E shows the state after the contact hole is formed. Next, although omitted from the drawing, another mask is formed, and the interlayer insulating film 231 is processed by dry etching to form a contact hole reaching the top surface of the gate electrode 222.

[0061] Subsequently, a metallic film, for example, a stacked film including a titanium film (Ti film), a titanium nitride film (TiN film), and an aluminum film (Al film), is built up over the contact hole formed on the surface of a part of the n-type source region 211 and the surface of the p+ type contact region 202 and the interlayer insulating film 231 connected with the opening portion (omitted in the drawing) reaching the gate electrode 222.

[0062] Subsequently, the metallic film is processed to form the source electrode 232, which is electrically connected to the n-type part of the source region 211 and the p+-type body contact region 202, and a gate wiring electrode (omitted from the drawing) electrically connected to the gate electrode 222. After that, external wiring is electrically connected to the source electrode 232 and the gate wiring electrode (omitted from the drawing).

[0063] Although the description is omitted, the drain electrode 103 containing a metal is formed over the back surface of the SiC substrate 101. In addition, for electrical connection between the drain electrode 103 and the back surface of the SiC substrate 101, a region in which n-type impurities are implanted at a high concentration and a silicide layer may be formed over the back surface of the SiC substrate.

[0064] Through the above steps, the structure of the MOSFET can be realized as shown in the Fig. 5A to Fig. 5D. Embodiment 2

[0065] Fig. 11 shows a plan view of unit cells of a SiC MOSFET according to a second embodiment of the present disclosure. Embodiment 2 differs from Embodiment 1 in that the JFET region 204 is formed in a square shape, and a source contact region 230 is formed in the n-type source region 211 formed on the outer side on the four sides of the JFET region 204. In other words, the n-type source regions 211, each including the source contact region 230, are arranged in a checkerboard pattern so that corners of the n-type source regions 211 overlap each other.

[0066] On the outer side of the square JFET region 204, the first channel regions 201a and the second channel regions 201b are formed with a higher impurity concentration than the first channel regions 201a. The first channel regions 201a are formed along the four edges of the square JFET region 204, and the second channel regions 201b are formed in contact with the four vertices of the square JFET region 204.

[0067] In this embodiment, since the source contact region 230 is formed in the n-type source region 211 near the channels, the resistance of the portion in the range from the source contact region 230 to the channel region 201a, 201b can be reduced. On the other hand, the n-type source region 211 including the source contact region 230 is enlarged, and as a result, the length of the JFET region 204 is increased. Therefore, this embodiment is advantageous over Embodiment 1 in a use where a higher breakdown voltage is required. This is because, in the case of a high breakdown voltage, it is necessary to increase the length of the JFET region 204 because the impurity concentration of the JFET region 204 is low and its resistance tends to be high.On the other hand, Embodiment 1 is advantageous in use at a comparatively low breakdown voltage because the electric field impressed on the gate oxide film tends to be high, with the resistance of the JFET region being low.

[0068] To manufacture the MOSFET of this embodiment, the same or similar method to the manufacturing method in Embodiment 1 can be used. Embodiment 3

[0069] Fig. 12 shows a top view of unit cells of a SiC MOSFET according to a third embodiment of the present disclosure. Embodiment 3 differs from Embodiment 2 in that the n-type source region 211 including the source contact region 230 is rectangular rather than square, and the unit cells 1011 are also rectangular. The n-type source regions 211 are arranged in a checkerboard pattern so that corners of the n-type source regions 211 overlap each other, as in Embodiment 2.

[0070] In this embodiment, the JFET region 204 is also rectangular in shape, unlike Embodiment 2. The first channel regions 201a and the second channel regions 201b, which have a higher impurity concentration than the first channel regions 201a, are formed on the outer side of the JFET region 204. The first channel regions 201a are formed along the four edges of the rectangular JFET region 204, and the second channel regions 201b are formed in contact with the four vertices of the rectangular JFET region 204.

[0071] In this embodiment, since the source contact region 230 is formed in the n-type source region 211 near the channels, as in Embodiment 2, the resistance of the portion in the range from the source contact region 230 to the channel can be reduced. Since the JFET region is rectangular, if the length of the shorter edges of the rectangle is shortened, the resistance of the JFET region is thereby increased, but the electric field impressed on the gate oxide film can be reduced, so that higher reliability can be obtained. Although the channel width is smaller than in Embodiment 1, the resistance of the portion extending from the source contact region to the channel can be reduced, so that a lower ON resistance can be realized by controlling the lengths of the longer edges and the shorter edges of the rectangle.

[0072] To manufacture the MOSFET of this embodiment, the same or a similar method to the manufacturing method in Embodiment 1 can be used. Embodiment 4

[0073] Fig. Fig. 13 shows a plan view of unit cells of a SiC MOSFET according to a fourth embodiment of the present disclosure. In Embodiment 4, with respect to the band-shaped unit cells shown in Fig. 3A, the adjacent unit cells are bridged by a source region, and the channel regions are also provided in the portions near the bridging source region in the horizontal direction. Therefore, the JFET region 104 is a closed region having a rectangular shape, and the first channel regions 201a and the second channel regions 201b, which have a higher impurity concentration than the first channel regions 201a, are formed on the outer side of the JFET region 204. The first channel regions 201a are formed along the four edges of the rectangular JFET region 204, while the second channel regions 201b are formed in contact with the four vertices of the rectangular JFET region 204.Although the p+ type contact regions 202 are shown as being discrepantly aligned in the vertical direction in the figure, a p+ type contact region 202 having a ribbon-like shape elongated in the vertical direction may be employed.

[0074] In this embodiment, in the case where the length of the JFET region (Lj in Fig. 13) is greater than the width of the bridging area (Lp in Fig. 13), the channel width is larger than that of the Fig. 3A. Therefore, this embodiment is advantageous in an application requiring a higher breakdown voltage, where it is necessary to increase the length of the JFET region because the impurity concentration of the JFET region is low.

[0075] To manufacture the MOSFET of this embodiment, the same or similar method to the manufacturing method in Embodiment 1 can be used. Embodiment 5

[0076] Fig. 14 shows a plan view of unit cells of a SiC MOSFET according to a fifth embodiment of the present disclosure. In Embodiment 5, the source regions are spaced apart from each other and are octagonal in shape. Note that the octagonal shape is not a regular octagon, but has two kinds of edge lengths, with longer edges and shorter edges alternately connected; in other words, the octagonal shape is an octagonal shape as if obtained by cutting off the four corners (portions near the vertices) of a square. The first channel regions 201a are formed in contact with the longer edges, and the second channel regions 201b having a higher impurity concentration than the first channel regions 201a are formed in contact with the shorter edges. Further, the channel length L1 of the first channel region 201a is longer than the channel length L2 of the second channel region 201b.

[0077] To manufacture the MOSFET of this embodiment, the same or a similar method to the manufacturing method in Embodiment 1 can be used.

[0078] Fig. Figure 15 shows a plan view of a step specifically at the time of forming the channel regions. When a mask having octagonal opening portions is formed and then p-type impurities are implanted in a direction inclined from the normal to the substrate by the same or similar method as that in Embodiment 1, the channel regions are formed as shown in Fig.15. In this case, the channel length L2 of the portion in contact with the shorter edge is smaller than the channel length L1 of the portion in contact with the longer edge, and approximately L1 ≒ L2 × √2. When four implantation steps are performed, the channel regions in contact with the shorter edges are subjected to two implantation steps in an overlapping manner, thereby forming the second channel regions with higher impurity concentration than the first channel regions along the longer edges. As a result of the four implantation steps, for example, the channel length L1 of the first channel regions 201a becomes 1.1 to 1.7 times the channel length L2 of the second channel regions 201b.

[0079] The present disclosure is widely applicable to power semiconductor devices primarily used for high-voltage, high-current applications. According to the present disclosure, a vertical DMOSFET structure with high reliability and high performance, including channels formed on a self-alignment basis, can be manufactured. In particular, the present disclosure is advantageous when applied to a power semiconductor device configured by a short-channel DMOSFET having a channel length equal to or less than 2 µm.

[0080] The present disclosure is not limited to the embodiments described above, but includes various modifications. For example, part of the configuration of one embodiment may be replaced by the configuration of another embodiment, and the configuration of another embodiment may be added to the configuration of one embodiment. Furthermore, part of the configuration of each embodiment may be added, omitted, or replaced by the configurations of other embodiments.

[0081] The functions of the "source" and "drain" of a transistor may be reversed when using a transistor of a different polarity, when the current direction is reversed during circuit operation, or in other cases. For this reason, the terms "source" and "drain" may be used in reverse.

[0082] The terms "electrode" and "wiring" in this specification and the like do not limit the configuration components on a functional basis. For example, an "electrode" may be used as part of a "wiring," and vice versa. Furthermore, the terms "electrode" and "wiring" include cases where multiple "electrodes" and "wirings" are formed in an integral form.

[0083] Features, components, and specific details of the structures of the above-described embodiments may be interchanged or combined to form further embodiments optimized for the particular application. To the extent that these modifications are readily apparent to one skilled in the art, they are intended to be disclosed implicitly by the above description, without expressly specifying every possible combination for the sake of brevity of the present description.

Claims

[1] A semiconductor device comprising an active region (1002) in which a plurality of unit cells are regularly arranged, each of the unit cells comprising: a channel region formed over a front surface of a semiconductor substrate and having a first conductivity type; a source region (211) having a second conductivity type different from the first conductivity type and formed over the front side of the semiconductor substrate so as to be in contact with the channel region; and a JFET region (204) having the second conductivity type and formed over the front side of the semiconductor substrate on an opposite side of the channel region from the source region (211) so as to be in contact with the channel region; wherein the channel region comprises a first channel region and a second channel region having a higher impurity concentration than the first channel region; and wherein the first channel region and the second channel region are arranged along first and second directions, respectively, to surround outer edge portions of the source region (211) of each unit cell in a plan view. [2] The semiconductor device according to claim 1, wherein the impurity concentration of the second channel region is 1.6 to 2.5 times the impurity concentration of the first channel region. [3] The semiconductor device according to claim 1, wherein the unit cell has a rectangular or square shape. [4] A semiconductor device according to claim 1, where the JFET region (204) is a rectangular region, the first channel region is formed along an edge of the JFET region (204), and the second channel region is formed in contact with a tip of the JFET region (204). [5] The semiconductor device of claim 1, wherein the JFET region (204) is surrounded by the channel region. [6] A semiconductor device according to claim 1, wherein the source region (211) constituting the unit cell is formed in such a manner that the source regions (211) of adjacent unit cells are separated from each other, and is of octagonal shape. [7] The semiconductor device according to claim 6, wherein edges forming the octagon have two or three kinds of lengths, the first channel region being formed along a longest edge and the second channel region being formed along a shortest edge. [8] The semiconductor device according to claim 6, wherein a channel length of the first channel region is 1.1 to 1.7 times the channel length of the second channel region. [9] A semiconductor device according to claim 1, comprising: a gate insulating film (221) formed over the channel region and at least a part of the source region (211); a gate electrode (222) formed over at least a part of the gate insulating film (221); a body region (201) having the first conductivity type and formed under the source region (211); a body contact region (202) having the first conductivity type, formed in contact with the body region (201) and having a higher impurity concentration than the body region (201); a source electrode (232) in contact with the body contact region (202); and a drain electrode (203) formed on a surface opposite the front side of the semiconductor substrate. [10] A method of manufacturing a semiconductor device, comprising: a step of forming a mask over a semiconductor substrate; a channel implantation step for implanting impurities of a first conductivity type with the mask as a shielding film (321); and a source implantation step of implanting impurities of a second conductivity type different from the first conductivity type with the mask as a shielding film (321); wherein the channel implantation step is performed in a direction inclined relative to a normal to the semiconductor substrate and includes implanting four times by rotating the semiconductor substrate by 90 degrees each time. [11] A method of manufacturing a semiconductor device according to claim 10, wherein during the four-time implantation constituting the channel implantation step, the second and later times of implantation are each carried out in a manner to further implant in an overlapping manner into the part of a region where the implantation has already been carried out in a previous time or times. [12] A method of manufacturing a semiconductor device according to claim 10, wherein a plan view of the mask has a pattern in which a plurality of unit cells are regularly arranged in a two-dimensional matrix structure such that the unit cell is repeated in an X direction and a Y direction orthogonal to the X direction, and Tilt directions in the channel implantation steps are parallel to the X-direction and the Y-direction. [13] A method of manufacturing a semiconductor device according to claim 12, where the unit cell has a rectangular or square shape and a direction of an edge of the unit cell is parallel to an ion incident direction in the channel implantation step. [14] A method of manufacturing a semiconductor device according to claim 10, wherein the four-time implantation in the channel implantation step is the same in tilt angle, a type of impurity, implantation energy and implantation dose and is different only in the rotation angle. [15] The method of manufacturing a semiconductor device according to claim 10, wherein an inclination angle between the implantation direction and the normal to the semiconductor substrate in the channel implantation step is 15 to 45 degrees.

Citation Information

Patent Citations

  • Method of manufacturing sic vertical mosfet

    JP2009147381A

  • High voltage mosfet devices and methods of making the devices

    US20150084066A1