Semiconductor device and method for manufacturing the same

By alternately aligning the structures of n-type and p-type cylinders in the SiC power device to avoid the formation of hollows, the problem of increased leakage current of SiC power devices in the prior art is solved, and the cost-effectiveness is improved.

CN116057712BActive Publication Date: 2024-11-05MITSUBISHI ELECTRIC CORP
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
CN202080102138.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2020-06-24
Publication Date
2024-11-05
Estimated Expiration
2040-06-24

AI Technical Summary

Technical Problem

In the superstructure formation of SiC power devices, the prior art, voids are prone to occur, which lead to an increase in leakage current, and the process of reducing leakage current increases the cost of the semiconductor device.

Method used

A plurality of n-type and p-type cylinders arranged alternately on the semiconductor substrate are used to avoid the occurrence of hollows when forming the cylinder through epitaxial growth, thereby reducing the invalid area and reducing leakage current.

Benefits of technology

It effectively reduces the invalid area of ​​the semiconductor device, reduces leakage current, and reduces manufacturing costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116057712B_ABST
    Figure CN116057712B_ABST
Patent Text Reader

Abstract

The present disclosure relates to a semiconductor device using a super junction structure, comprising: a semiconductor substrate of a first conductive type; a column portion having a plurality of first columns of the first conductive type and a plurality of second columns of the second conductive type, which are arranged on the semiconductor substrate in a manner protruding in the thickness direction of the semiconductor substrate; a column peripheral portion of the first conductive type or the second conductive type, which is arranged around the column portion; and a semiconductor element, wherein the column portion is set as an active region, and the plurality of first and second columns have a stripe-like shape when viewed from above, and are alternately arranged in a parallel manner in a column width direction orthogonal to their respective length directions.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to a semiconductor device, and more particularly to a semiconductor device using a super junction structure. Background Art

[0002] Silicon carbide (SiC) used in power devices has a higher withstand voltage than silicon (Si) and can reduce resistance, and SiC-SBD (Schottky Barrier Diode) and SiC-MOSFET (MOS Field Effect Transistor) have been commercialized.

[0003] As a structure that exceeds the theoretical limit of unipolar devices and realizes further high withstand voltage and low resistance of SiC power devices, the super junction (SJ) structure can be cited. The SJ structure refers to a structure in which p-type impurity layers (p-type pillars) and n-type impurity layers (n-type pillars) are alternately arranged in a direction orthogonal to the direction of the main current flowing in the semiconductor layer.

[0004] Among the methods for forming the SJ structure of SiC power devices, there are multiple epitaxy methods that repeat ion implantation and epitaxial growth, and buried epitaxy methods that form trenches and perform buried epitaxy. In high-voltage devices above 3.3kV, drift resistance is dominant in on-resistance, and the advantages of using the SJ structure are great. In addition, in high-voltage devices above 3.3kV, a thick SJ structure needs to be formed, so when considering productivity, the buried epitaxy method is advantageous.

[0005] In the case of forming an SJ structure of a SiC power device by buried epitaxy, as in Patent Document 1, Fig.11 As shown in FIG. 1 , a cavity is formed at the end of the trench. When a cavity is formed, the leakage current of the silicon carbide semiconductor device increases. Therefore, as described in Patent Document 1 Fig. 22 As shown in the figure, a groove for scribing is formed, and then as in Patent Document 1 Fig.24 The structure obtained by such a process is as shown in Patent Document 1. Fig.31 The leakage current of the silicon carbide semiconductor device is reduced.

[0006] Prior art literature

[0007] Patent Literature

[0008] Patent Document 1: International Publication No. 2019 / 160086 Summary of the invention

[0009] The manufacturing method of Patent Document 1 generates an ineffective region with a very wide void formed therein, and requires an additional step and structure for reducing leakage current, which results in an increase in the cost of the semiconductor device.

[0010] The present disclosure has been made to solve the above-mentioned problems, and an object of the present disclosure is to provide a semiconductor device that reduces an ineffective area of ​​the semiconductor device and reduces leakage current.

[0011] The semiconductor device involved in the present disclosure comprises: a semiconductor substrate of the first conductive type; a column portion, having a plurality of first columns of the first conductive type and a plurality of second columns of the second conductive type, which are arranged on the semiconductor substrate in a manner protruding in the thickness direction of the semiconductor substrate; a column peripheral portion of the first conductive type or the second conductive type, which is arranged around the column portion; and a semiconductor element, wherein the column portion is set as an active area, and the plurality of first columns and the plurality of second columns have a stripe-like shape when viewed from above, and are alternately arranged in a parallel manner in a column width direction orthogonal to their respective length directions.

[0012] According to the semiconductor device of the present disclosure, voids are less likely to be formed when the first and second pillars are formed by epitaxial growth, and the ineffective area of ​​the semiconductor device can be reduced to reduce leakage current. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 FIG. 1 is a perspective view schematically showing the structure of a column portion of the silicon carbide semiconductor device according to the first embodiment.

[0014] Figure 2 : is a cross-sectional view showing the structure of a column portion of the silicon carbide semiconductor device according to the first embodiment.

[0015] Figure 3 : is a cross-sectional view showing the structure of a column portion of the silicon carbide semiconductor device according to the first embodiment.

[0016] Figure 4 It is a diagram showing the sizes of n-type pillars and p-type pillars.

[0017] Figure 5 This is a perspective view schematically showing the silicon carbide semiconductor device according to the first embodiment.

[0018] Figure 6 : is a figure which shows the structure of the MOSFET cell area.

[0019] Figure 7 : is a figure which shows the structure of the MOSFET cell area.

[0020] Figure 8This is a cross-sectional view showing the structure of a silicon carbide semiconductor device according to Modification 1 of Embodiment 1.

[0021] Fig. 9 This is a cross-sectional view showing the structure of a silicon carbide semiconductor device according to a second modification of the first embodiment.

[0022] Fig.10 This is a cross-sectional view showing the structure of a silicon carbide semiconductor device according to Modification 3 of Embodiment 1.

[0023] Fig.11 This is a cross-sectional view showing the structure of a silicon carbide semiconductor device according to Modification 3 of Embodiment 1.

[0024] Fig.12 This is a diagram schematically showing the flow of a main current in a silicon carbide semiconductor device according to Modification 3 of Embodiment 1.

[0025] Fig.13 This is a perspective view schematically showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the second embodiment.

[0026] Fig.14 This is a perspective view schematically showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the second embodiment.

[0027] Fig.15 This is a perspective view schematically showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the second embodiment.

[0028] Fig.16 This is a perspective view schematically showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the second embodiment.

[0029] Fig.17 This is a perspective view schematically showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the second embodiment.

[0030] Fig.18 FIG. 1 is a perspective view schematically showing the structure of a column portion of a silicon carbide semiconductor device according to the second embodiment.

[0031] Fig.19 1 is a cross-sectional view showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the second embodiment.

[0032] Fig. 20 1 is a cross-sectional view showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the second embodiment.

[0033] Fig.21 1 is a cross-sectional view showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the second embodiment.

[0034] Fig. 22 1 is a cross-sectional view showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the second embodiment.

[0035] Fig.23 1 is a cross-sectional view showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the second embodiment.

[0036] Fig.24 1 is a cross-sectional view showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the second embodiment.

[0037] Fig.25 1 is a cross-sectional view showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the second embodiment.

[0038] Fig.26 This is a cross-sectional view showing a first modification of the method for manufacturing the column portion of the silicon carbide semiconductor device according to the second embodiment.

[0039] Fig. 27 This is a cross-sectional view showing a first modification of the method for manufacturing the column portion of the silicon carbide semiconductor device according to the second embodiment.

[0040] Fig.28 This is a cross-sectional view showing a first modification of the method for manufacturing the column portion of the silicon carbide semiconductor device according to the second embodiment.

[0041] Fig.29 This is a cross-sectional view showing a second modification of the method for manufacturing the column portion of the silicon carbide semiconductor device according to the second embodiment.

[0042] Fig.30 This is a cross-sectional view showing a second modification of the method for manufacturing the column portion of the silicon carbide semiconductor device according to the second embodiment.

[0043] Fig.31 This is a perspective view schematically showing a silicon carbide semiconductor device according to the second embodiment.

[0044] Fig.32 It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0045] Fig.33 It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0046] Fig.34 It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0047] Fig.35It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0048] Fig.36 It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0049] Fig.37 It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0050] Fig.38 It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0051] Fig.39 It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0052] Fig.40 This is a cross-sectional view showing the structure of a modification of the silicon carbide semiconductor device according to the second embodiment.

[0053] Fig.41 This is a cross-sectional view showing the structure of a modification of the silicon carbide semiconductor device according to the second embodiment.

[0054] Fig.42 This is a cross-sectional view showing Modification 1 of the method for manufacturing a silicon carbide semiconductor device according to the second embodiment.

[0055] Fig.43 This is a cross-sectional view showing Modification 1 of the method for manufacturing a silicon carbide semiconductor device according to the second embodiment.

[0056] Fig.44 This is a cross-sectional view showing a second modification of the method for manufacturing a silicon carbide semiconductor device according to the second embodiment.

[0057] Fig.45 This is a cross-sectional view showing a second modification of the method for manufacturing a silicon carbide semiconductor device according to the second embodiment.

[0058] Fig.46 This is a cross-sectional view showing a third modification of the method for manufacturing a silicon carbide semiconductor device according to the second embodiment.

[0059] Fig.47 This is a cross-sectional view showing a third modification of the method for manufacturing a silicon carbide semiconductor device according to the second embodiment.

[0060] Fig.48 This is a cross-sectional view showing a fourth modification of the method for manufacturing a silicon carbide semiconductor device according to the second embodiment.

[0061] Fig.49 This is a cross-sectional view showing a fifth modification of the method for manufacturing a silicon carbide semiconductor device according to the second embodiment.

[0062] Fig.50 This is a cross-sectional view showing a fifth modification of the method for manufacturing a silicon carbide semiconductor device according to the second embodiment.

[0063] Fig.51 It is a cross-sectional view for explaining a dicing step of two adjacent silicon carbide semiconductor devices.

[0064] Fig.52 It is a cross-sectional view for explaining a dicing step of two adjacent silicon carbide semiconductor devices.

[0065] Fig.53 This is a cross-sectional view showing a method for manufacturing two adjacent silicon carbide semiconductor devices.

[0066] Fig.54 This is a cross-sectional view showing a method for manufacturing two adjacent silicon carbide semiconductor devices.

[0067] Fig.55 This is a perspective view schematically showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the third embodiment.

[0068] Fig.56 This is a perspective view schematically showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the third embodiment.

[0069] Fig.57 This is a perspective view schematically showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the third embodiment.

[0070] Fig.58 This is a perspective view schematically showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the third embodiment.

[0071] Fig.59 This is a perspective view schematically showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the third embodiment.

[0072] Fig.60 FIG. 1 is a perspective view schematically showing the structure of a column portion of a silicon carbide semiconductor device according to the third embodiment.

[0073] Fig.61 1 is a cross-sectional view showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the third embodiment.

[0074] Fig.62 1 is a cross-sectional view showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the third embodiment.

[0075] Fig.63 1 is a cross-sectional view showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the third embodiment.

[0076] Fig.64 1 is a cross-sectional view showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the third embodiment.

[0077] Fig.65 1 is a cross-sectional view showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the third embodiment.

[0078] Fig.66 1 is a cross-sectional view showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the third embodiment.

[0079] Fig.67 1 is a cross-sectional view showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the third embodiment.

[0080] Fig.68 1 is a cross-sectional view showing a method for manufacturing a column portion of a silicon carbide semiconductor device according to the third embodiment.

[0081] Fig.69 This is a perspective view schematically showing a silicon carbide semiconductor device according to the third embodiment.

[0082] Fig.70 It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0083] Fig.71 It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0084] Fig.72 It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0085] Fig.73 It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0086] Fig.74 It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0087] Fig.75 It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0088] Fig.76 It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0089] Fig.77 It is a cross-sectional view showing the manufacturing process of the MOSFET cell in the MOSFET cell region.

[0090] Fig.78 This is a cross-sectional view showing the structure of a modification of the silicon carbide semiconductor device according to the third embodiment.

[0091] Fig.79 This is a cross-sectional view showing Modification 1 of the method for manufacturing a silicon carbide semiconductor device according to the third embodiment.

[0092] Fig.80 This is a cross-sectional view showing Modification 1 of the method for manufacturing a silicon carbide semiconductor device according to the third embodiment.

[0093] Fig.81 This is a cross-sectional view showing a second modification of the method for manufacturing a silicon carbide semiconductor device according to the third embodiment.

[0094] Fig.82 This is a cross-sectional view showing a second modification of the method for manufacturing a silicon carbide semiconductor device according to the third embodiment.

[0095] Fig.83 This is a cross-sectional view showing a third modification of the method for manufacturing a silicon carbide semiconductor device according to the third embodiment.

[0096] Fig.84 This is a cross-sectional view showing a third modification of the method for manufacturing a silicon carbide semiconductor device according to the third embodiment.

[0097] Fig.85 This is a cross-sectional view showing a fourth modification of the method for manufacturing a silicon carbide semiconductor device according to the third embodiment.

[0098] Fig.86 This is a cross-sectional view showing a fifth modification of the method for manufacturing a silicon carbide semiconductor device according to the third embodiment.

[0099] Fig.87 This is a cross-sectional view showing a fifth modification of the method for manufacturing a silicon carbide semiconductor device according to the third embodiment.

[0100] Fig.88 It is a cross-sectional view for explaining a dicing step of two adjacent silicon carbide semiconductor devices.

[0101] Fig.89 It is a cross-sectional view for explaining a dicing step of two adjacent silicon carbide semiconductor devices.

[0102] Fig.90 This is a cross-sectional view showing a method for manufacturing two adjacent silicon carbide semiconductor devices.

[0103] Fig.91 This is a cross-sectional view showing a method for manufacturing two adjacent silicon carbide semiconductor devices.

[0104] Fig.92 FIG. 1 is a diagram showing an example of the structure of a MOSFET cell region of the silicon carbide semiconductor device according to the first to third embodiments.

[0105] Fig.93 FIG. 1 is a diagram showing an example of the structure of a MOSFET cell region of the silicon carbide semiconductor device according to the first to third embodiments.

[0106] Fig.94 FIG. 1 is a diagram showing an example of the structure of a MOSFET cell region of the silicon carbide semiconductor device according to the first to third embodiments.

[0107] Fig.95 FIG. 1 is a diagram showing an example of a structure of a MOSFET terminal region of the silicon carbide semiconductor device according to the first to third embodiments.

[0108] Fig.96 FIG. 1 is a diagram showing an example of a structure of a MOSFET terminal region of the silicon carbide semiconductor device according to the first to third embodiments.

[0109] Fig.97 This is a diagram showing an example of a structure in which an SBD region and an SBD terminal region are provided instead of a MOSFET cell region in the silicon carbide semiconductor device according to the first to third embodiments.

[0110] Fig.98 This is a diagram showing an example of a structure in which an SBD region and an SBD terminal region are provided instead of a MOSFET cell region in the silicon carbide semiconductor device according to the first to third embodiments.

[0111] Fig.99 This is a diagram showing an example of a structure in which an SBD region and an SBD terminal region are provided instead of a MOSFET cell region in the silicon carbide semiconductor device according to the first to third embodiments.

[0112] (Explanation of symbols)

[0113] 3: semiconductor substrate; 4: n-type SiC layer; 6: p-type pillar peripheral portion; 7: pillar portion; 7n: n-type pillar; 7p: p-type pillar; 5: p-type chip peripheral portion; 60: p-type SiC layer; 70: n-type SiC layer; 71: n-type chip peripheral portion; 72: convex portion; 73: n-type pillar peripheral portion. DETAILED DESCRIPTION

[0114] <Foreword>

[0115] The semiconductor device in each embodiment described below refers to a semiconductor chip obtained by separating a semiconductor device manufactured by a wafer process into chip units by a dicing process, and the chip peripheral portion refers to the outer peripheral portion of the semiconductor chip.

[0116] In the following, “outside” refers to a direction toward the outer periphery of the semiconductor chip, and “inside” refers to a direction opposite to “outside”.

[0117] In the following description, regarding the conductivity type of impurities, generally, n-type is defined as the "first conductivity type" and p-type, which is the conductivity type opposite to n-type, is defined as the "second conductivity type", but the definition may be reversed. - Type indicates that the impurity concentration is lower than that of n-type. + Type indicates that the impurity concentration is higher than that of n-type. Similarly, p - Type means that the impurity concentration is lower than that of p-type. + Type indicates a concentration where the impurity concentration is higher than that of the p-type.

[0118] In addition, the accompanying drawings are schematically shown, and the size and position of the images shown in different drawings may not be correctly recorded, but may be appropriately changed. In addition, in the following description, the same components are illustrated with the same symbols, and their names and functions are also the same. Therefore, sometimes a detailed description about them is omitted. In addition, in this specification, when it is called "on" and "covering" it does not prevent the presence of inclusions between the components. For example, when it is recorded as "B arranged on A" or "A covers B", it can mean that other components C are arranged between A and B, or it can mean that other components C are not arranged. In addition, in the following description, sometimes "up", "down", "side", "bottom", "table" or "back" are used to mean specific positions and directions, but these terms are used appropriately to make the contents of the implementation method easy to understand, and have nothing to do with the direction when actually implemented.

[0119] In addition, the term "MOS" was previously used for the junction structure of metal-oxide-semiconductor, using the initials of Metal-Oxide-Semiconductor. However, in particular, in field effect transistors (MOSFETs) having a MOS structure, the materials of the gate insulating film and the gate electrode have been improved from the perspectives of integration and improvement of manufacturing processes in recent years.

[0120] For example, in MOSFET, polysilicon is used as the material of the gate electrode instead of metal from the viewpoint of forming the source / drain in a self-matching manner. In addition, from the viewpoint of improving electrical characteristics, a material with a high dielectric constant is used as the material of the gate insulating film, but the material is not necessarily limited to oxide.

[0121] Therefore, the term "MOS" is not necessarily limited to the stacked structure of metal-oxide-semiconductor, and this specification does not take such a limitation as a premise. That is, in view of common technical knowledge, "MOS" here has not only the abbreviation derived from its etymology, but also has a broad meaning including the stacked structure of conductor-insulator-semiconductor.

[0122] <Implementation Method 1>

[0123] <Device Structure>

[0124] Figure 1 This is a perspective view schematically showing the structure of a column portion of a silicon carbide semiconductor device 100 having an SJ structure as a semiconductor device according to Embodiment 1. The silicon carbide semiconductor device 100 represents a vertical MOSFET in which a main current flows in a direction perpendicular to a main surface of a semiconductor substrate, but for ease of explanation, main electrodes and unit cells of the MOSFET are omitted from the illustration.

[0125] The column part 7 is as follows Figure 1 As shown, a plurality of n-type pillars 7n (first pillars) and a plurality of p-type pillars 7p (second pillars) having a stripe shape when viewed from above are alternately arranged in parallel in a width direction perpendicular to the respective longitudinal directions on the semiconductor substrate 3. The arrangement direction thereof is perpendicular to the direction in which the main current of the silicon carbide semiconductor device 100 flows. A p-type pillar peripheral portion 6 containing p-type impurities is provided so as to surround the pillar portion 7, and the outer side of the p-type pillar peripheral portion 6 becomes a p-type chip peripheral portion 5 containing p-type impurities.

[0126] The semiconductor substrate 3 is n + The n-type SiC substrate is a commercially available 4H-SiC substrate having an off angle of 4 degrees in the [11-20] direction, a thickness of 300 to 400 μm, and an n-type impurity concentration of 5×10 18 ~1×10 20 cm -3 .

[0127] An n-type SiC layer 4 (first semiconductor layer) containing n-type impurities is provided between the semiconductor substrate 3 and the p-type chip peripheral portion 5. The thickness of the n-type SiC layer 4 is 0.5 to 10 μm, and the concentration of the n-type impurities is 1×10 14 ~5×10 19 cm-3 In addition, the semiconductor substrate 3 and the n-type SiC layer 4 are sometimes collectively referred to as a semiconductor base.

[0128] Figure 2 Shown along Figure 1 The cross-sectional view in the direction indicated by the arrow of line AA (a line parallel to the Y axis) in FIG. Figure 3 The cross-sectional view along the BB line (a line parallel to the X-axis) is shown in the direction indicated by the arrow. Figure 2 as well as Figure 3 The n-type pillar 7n and the p-type pillar 7p are arranged to protrude from the n-type SiC layer 4 in the height direction (Z-axis direction), the p-type pillar peripheral portion 6 surrounds the pillar portion 7, and the p-type chip peripheral portion 5 covers the outer periphery of the n-type SiC layer 4. In addition, the end surface of the n-type SiC layer 4 is exposed on the side surface of the semiconductor chip.

[0129] Figure 4 7nW of the n-type pillar 7n and the p-type pillar 7p are equal, for example, they are formed to have a length of 0.5 to 5 μm. In addition, the pillar height 7nH of the n-type pillar 7n and the pillar width 7pH of the p-type pillar 7p are equal, for example, they are formed to have a height of 5 to 100 μm. Moreover, the product of the impurity concentration and the width of the n-type pillar 7n is formed to be substantially the same as the product of the impurity concentration and the width of the p-type pillar 7p. The concentrations of the n-type impurities and the p-type impurities are, for example, 5×10 15 ~1×10 18 cm -3 With such a structure, when the silicon carbide semiconductor device 100 is in the off state, the depletion layer spreads over the entire n-type pillar 7n and the p-type pillar 7p.

[0130] In order to relax the electric field in the joint with the p-type pillar 7p, the impurity concentration of the n-type SiC layer 4 is preferably the same as or lower than the impurity concentration of the n-type pillar 7n, but can also be set to a higher concentration than the impurity concentration of the n-type pillar 7n. In addition, it is also possible not to provide the n-type SiC layer 4, and the n-type pillar 7n and the p-type pillar 7p can be directly bonded to the semiconductor substrate 3. In the case where the impurity concentration of the n-type SiC layer 4 is higher than the impurity concentration of the n-type pillar 7n and the n-type SiC layer 4 is not provided, the effect is smaller than the case where the impurity concentration of the n-type SiC layer 4 is the same as or lower than the impurity concentration of the n-type pillar 7n, but by providing Figure 1 The pillar portion 7 shown can realize a silicon carbide semiconductor device having a higher withstand voltage and lower resistance than conventional structures.

[0131] In addition, Figure 1 to Figure 4, a structure is shown in which three p-type pillars 7p and four n-type pillars 7n are provided in the pillar portion 7, but in reality, the number of p-type pillars 7p and n-type pillars 7n corresponding to the size of the carbide semiconductor device is formed. As previously described, the pillar width is 0.5 to 5 μm, and the overall Y-direction width of the MOSFET cell region in which the unit cell of the MOSFET is provided is, for example, 1 to 10 mm, so the number of p-type pillars 7p and n-type pillars 7n accommodated in at least the MOSFET cell region is formed.

[0132] In addition, if Figure 3 as well as Figure 4 As shown, by making the width 6W of the p-type pillar peripheral portion 6 thicker than the width 7pW of the p-type pillar 7p, when the impurity concentrations in the p-type pillar peripheral portion 6 and the p-type pillar 7p are the same, the product of the impurity concentration and the width of the p-type pillar peripheral portion 6 is greater than the product of the impurity concentration and the width of the p-type pillar 7p. By setting such a structure, the depletion layer does not extend to the entire p-type pillar peripheral portion 6, the electric field intensity of the p-type pillar peripheral portion 6 is suppressed to be low, and the increase in leakage current and the risk of discharge are suppressed. In this case, the width of the p-type pillar peripheral portion 6 can be set to be 1.2 times or more of the width of the p-type pillar 7p.

[0133] In addition, even when the width 6W of the p-type column peripheral portion 6 and the width 7pW of the p-type column 7p are the same, by making the impurity concentration of the p-type column peripheral portion 6 more than 1.2 times the impurity concentration of the p-type column 7p, the depletion layer does not extend to the entire p-type column peripheral portion 6, the electric field strength of the p-type column peripheral portion 6 is suppressed to a low level, and the increase in leakage current and discharge risk are suppressed.

[0134] Figure 5 1 is a perspective view schematically showing a silicon carbide semiconductor device 100 in which a MOSFET unit cell is provided in a pillar portion 7 and a MOSFET termination region MTR provided on the periphery thereof.

[0135] use Figure 6 as well as Figure 7 , illustrating the structure of the MOSFET unit region MCR. Figure 6 is along Figure 5 The cross-sectional view in the direction indicated by the arrow of the CC line (a line parallel to the Y axis) in FIG. Figure 7 Show Figure 6 Magnified view of area A in FIG.

[0136] like Figure 7The MOSFET shown in the figure has an n-type SiC layer 4 provided on one main surface of a semiconductor substrate 3, and a plurality of p-type pillars 7p and a plurality of n-type pillars 7n are alternately provided on the n-type SiC layer 4. Furthermore, a plurality of p-type well regions 8 are selectively provided from the upper portion of the p-type pillars 7p to the upper portion of the n-type pillars 7n, and a p-type contact region 10 is provided in each well region 8 in a manner penetrating the well region 8.

[0137] Furthermore, an n-type source region 9 is provided in contact with both side surfaces of the contact region 10 in the upper layer of the well region 8. In addition, the source region 9 is configured to be thinner than the well region 8, so that the thickness of the contact region 10 is the same as that of the well region 8 or one side of the contact region 10 is slightly deeper, and the contact region 10 is electrically connected to the p-type pillar 7p.

[0138] A gate insulating film 11 is selectively formed on the source region 9 of the adjacent well region 8 so as to cross, and a gate electrode 12 is formed on the gate insulating film 11. That is, the gate insulating film 11 is provided between adjacent source regions 9 so as to reach from a portion of the upper portion of the source region 9 from the well region 8 and the n-type pillar 7n to a portion of the upper portion of the source region 9 of the adjacent well region 8, and the gate electrode 12 is provided on the gate insulating film 11.

[0139] Furthermore, an interlayer insulating film 13 is formed so as to cover the gate insulating film 11 and the gate electrode 12, and a source electrode 14 is formed so as to cover the interlayer insulating film 13. In the interlayer insulating film 13, a contact hole is provided in a region other than the region covering the gate electrode 12, which penetrates the interlayer insulating film 13 in the thickness direction and reaches a part of the source region 9 and the entire surface of the contact region 10. Furthermore, the source electrode 14 is filled in the contact hole, and the source electrode 14 is connected to the source region 9 and the contact region 10.

[0140] A drain electrode 15 is provided on the other main surface (back surface) of the semiconductor substrate 3 opposite to the side provided with the source electrode 14. An example of the structure of the MOSFET cell region MCR and the MOSFET terminal region MTR will be described later.

[0141] As described above, by providing the column portion 7 in which a plurality of n-type columns 7n and a plurality of p-type columns 7p having a stripe shape when viewed from above are alternately arranged on the semiconductor substrate 3, voids are not formed when the p-type columns 7p are formed by epitaxial growth, and the ineffective area can be reduced. Therefore, processing for separating voids becomes unnecessary, and the manufacturing cost can be reduced.

[0142] In addition, Figure 7, a structure in which two unit cells of MOSFET are provided is shown, but in reality, the number of unit cells corresponding to the size of the carbide semiconductor device is formed.

[0143] <Modification 1>

[0144] Figure 8 is a cross-sectional view showing the structure of a silicon carbide semiconductor device 100A according to a first modification of the first embodiment, and is similar to Figure 7 The corresponding cross-sectional diagram. Figure 8 As shown, the silicon carbide semiconductor device 100A includes an n-type SiC layer 40 (second n-type SiC layer) between the well regions 8 instead of the n-type SiC layer 4 .

[0145] <Modification 2>

[0146] Fig. 9 is a cross-sectional view showing the structure of a silicon carbide semiconductor device 100B according to a second variation of the first embodiment, and is similar to Figure 2 The corresponding cross-sectional diagram. Fig. 9 As shown, the silicon carbide semiconductor device 100B has a structure in which the p-type chip peripheral portion 5 is not provided on the n-type SiC layer 4 .

[0147] Such a structure has the characteristic that it is easy to manufacture when the p-type pillar 7p is formed by buried epitaxy or when the p-type pillar 7p is formed by full-surface etching. Fig. 9 In the embodiment, n-type SiC layer 4 has a uniform thickness, but the thickness of n-type SiC layer 4 in the region excluding p-type chip peripheral portion 5 can be made smaller than the thickness of p-type pillar peripheral portion 6 and the lower portion of p-type pillar 7p.

[0148] <Variation 3>

[0149] Fig.10 as well as Fig.11 is a cross-sectional view showing the structure of a silicon carbide semiconductor device 100C according to a third modification of the first embodiment, and Figure 2 as well as Figure 3 The corresponding cross-sectional diagram. Fig.10 as well as Fig.11 As shown, in the silicon carbide semiconductor device 100B, a height 4H2 of the pillar periphery of the n-type SiC layer 4 is lower than a height 4H1 of the n-type SiC layer 4 at the pillar portion 7 .

[0150] Fig.12 2 is a diagram schematically showing the flow of the main current in the MOSFET cell region MCR when such a structure is adopted using arrows. Fig.12The main current in the MOSFET cell region MCR shown flows in the MOSFET cell region MCR, and it is difficult for the current to flow in the chip periphery. The reason is that the thickness of the column periphery of the n-type SiC layer 4 becomes thinner and the current path becomes narrower, and the extension width of the main current to the chip periphery becomes narrower, and the main current reaches the semiconductor substrate 3 before extending to the chip periphery. In addition, the ratio of the height 4H1 and the height 4H2 can be set to 2 to 1, for example.

[0151] The chip periphery has defects during cutting, etc., which easily cause malfunctions when the main current flows. By making the height 4H2 of the pillar periphery of the n-type SiC layer 4 lower than the pillar height 4H1 of the n-type SiC layer 4 at the pillar portion 7, it is difficult for the main current to flow in the chip periphery, and the current tolerance when a large current flows in the MOSFET as the main current becomes larger.

[0152] A silicon carbide semiconductor device using silicon carbide as a semiconductor has excellent voltage resistance and a higher allowable current density than a semiconductor device using silicon as a semiconductor, and can obtain a semiconductor device that has excellent heat resistance and can operate at high temperatures.

[0153] <Implementation Method 2>

[0154] <Method for Manufacturing the Column>

[0155] Next, as a second embodiment, a method for manufacturing a column portion of a silicon carbide semiconductor device 200 will be described. Fig.13 In the process shown, for example, a commercially available 4H-SiC n-type substrate is prepared with an off angle of 4 degrees in the [11-20] direction, a thickness of 300 to 400 μm, and an n-type impurity concentration of 5×10 18 ~1×10 20 cm -3 The semiconductor substrate 3.

[0156] Next, in Fig.14 In the process shown, an n-type SiC layer 70 (first semiconductor layer) containing n-type impurities is formed by epitaxial growth on one main surface of the semiconductor substrate 3. The n-type SiC layer 70 can be made, for example, to have a thickness of 5 to 100 μm and an impurity concentration of 1×10 15 ~1×10 18 cm -3 In addition, the semiconductor substrate 3 and the n-type SiC layer 70 are sometimes collectively referred to as a semiconductor base.

[0157] Next, in Fig.15In the process shown, the n-type SiC layer 70 is etched by dry etching to form the convex portion 72 of the n-type SiC layer 70 and the n-type chip peripheral portion 71 of the n-type SiC layer 70. When the semiconductor substrate 3 has an off angle of 4 degrees in the [11-20] direction, the convex portion 72 is formed to have a stripe shape extending in a direction parallel to the X-axis ([11-20] direction) when viewed from above, and a plurality of convex portions 72 are arranged at intervals in the direction along the Y-axis ([1-100] direction). The n-type SiC layer 70 has a symmetrical shape in a direction perpendicular to the [11-20] direction, so when the p-type SiC layer is formed between the convex portions 72 by epitaxial growth, it is easy to control the shape of the p-type SiC layer. In addition, the convex portion 72 is formed to have a width in the Y-axis direction of, for example, 0.5 to 5 μm and a height in the Z-axis (

[0001] direction) direction of, for example, 5 to 100 μm.

[0158] On the other hand, when a substrate without an off angle is used as the semiconductor substrate 3, the stripe-shaped protrusions 72 can be formed in an orientation rotated 90 degrees around the

[0001] axis from the [11-20] direction. Fig.15 The [11-20] axis and the [1-100] axis are swapped to form a convex portion 72, and the p-type SiC layer formed between the convex portions 72 by epitaxial growth has a symmetrical shape, so it is easy to control the shape of the p-type SiC layer.

[0159] Next, in Fig.16 In the process shown, the p-type SiC layer 60 (second semiconductor layer) is formed by epitaxial growth so as to cover the n-type chip peripheral portion 71 and the convex portion 72 of the n-type SiC layer 70. Fig.16 In the figure, the surface of the p-type SiC layer 60 of the convex portion 72 of the n-type SiC layer 70 is flat, but may have concavities and convexities.

[0160] like Fig.15 As shown, the p-type SiC layer 60 is epitaxially grown while the surrounding of the convex portion 72 of the n-type SiC layer 70 is dug down as the n-type chip peripheral portion 71, so that no voids are formed due to the difference in crystal growth rate caused by the difference in crystal plane orientation, and the ineffective area can be reduced. Therefore, processing for separating voids becomes unnecessary, which can reduce manufacturing costs.

[0161] The concentration of the p-type impurity of the p-type SiC layer 60 may be set so that the product of the impurity concentration and the width of the n-type pillar 7n is substantially equal to the product of the impurity concentration and the width of the p-type pillar 7p.

[0162] Next, in Fig.17In the process shown, the p-type SiC layer 60 on the convex portion 72 of the n-type SiC layer 70 is removed by grinding or dry etching, so that the upper surface of the convex portion 72 is exposed, and the p-type SiC layer 60 is left between the convex portions 72. The convex portion 72 becomes an n-type pillar 7n, and the p-type SiC layer 60 between the convex portions 72 becomes a p-type pillar 7p, thereby constituting the pillar portion 7. In addition, a p-type pillar peripheral portion 6 having a predetermined width is formed so as to surround the pillar portion 7. In addition, a portion of the upper portion of the convex portion 72 of the n-type SiC layer 70 can be removed. In addition, a portion of the upper portion of the p-type chip peripheral portion 5 on the n-type chip peripheral portion 71 of the n-type SiC layer 70 can be removed.

[0163] Fig.18 FIG. 2 is a perspective view showing a MOSFET cell region MCR and a MOSFET terminal region MTR of a silicon carbide semiconductor device 200 formed by the above method before being formed. Fig.18 As shown in FIG. 2 , in the silicon carbide semiconductor device 200, except that the end surface of the n-type chip peripheral portion 71 of the n-type SiC layer 70 is exposed on the side surface of the semiconductor chip, Figure 1 The silicon carbide semiconductor devices 100 shown are the same.

[0164] In addition, Figure 14 to Figure 18 , a structure is shown in which three p-type pillars 7p and four n-type pillars 7n are provided in the pillar portion 7, but in reality, the number of p-type pillars 7p and n-type pillars 7n corresponding to the size of the carbide semiconductor device is formed.

[0165] Next, use the Fig.18 The cross-sectional view in the direction indicated by the arrow along the AA line (line parallel to the Y axis) and the cross-sectional view in the direction indicated by the arrow along the BB line (line parallel to the X axis) illustrate a modified example of the method for manufacturing the column portion of the silicon carbide semiconductor device 200.

[0166] first, Figure 19 to Figure 25 Show Figure 14 to Figure 17 The cross-sectional views along the lines AA and BB in the process shown. Fig.19 yes Fig.14 The cross-sectional view along line AA in the process shown in FIG. Fig.19 same. Fig. 20 yes Fig.15 The cross-sectional view along line AA in the process shown, Fig.21 It is a cross-sectional view along line BB. Fig. 22 yes Fig.16 The cross-sectional view along line AA in the process shown, Fig.23 It is a cross-sectional view along line BB. Fig.24 yes Fig.17 The cross-sectional view along line AA in the process shown, Fig.25 It is a cross-sectional view along line BB.

[0167] <Variation 1 of the Method for Manufacturing the Column>

[0168] right Fig.19 The structure shown, such as Fig.26 As shown in FIG. 1 , an n-type SiC layer 4 containing n-type impurities can be formed by epitaxial growth on one main surface of the semiconductor substrate 3, and an n-type SiC layer 70 containing n-type impurities can be formed by epitaxial growth thereon, thereby forming a two-layer structure. In this case, the n-type SiC layer 4 serving as the first layer can be made, for example, to have a thickness of 0.5 to 10 μm and a concentration of n-type impurities of 1×10 14 ~1×10 19 cm -3 The n-type SiC layer 70 serving as the second layer can be made, for example, thicker than 5 μm and have an n-type impurity concentration of 1×10 15 ~1×10 18 cm -3 In addition, both the n-type SiC layer 4 and the n-type SiC layer 70 can be referred to as a first n-type SiC layer.

[0169] By making the first n-type SiC layer a two-layer structure, it is possible to alleviate the distortion caused by the difference in the lattice constant of the crystal due to the difference in impurity concentration between the semiconductor substrate 3 and the epitaxial layer. In addition, the reason for changing the thickness and impurity concentration in the first layer and the second layer is that the thickness and impurity concentration required to alleviate the distortion are different in the first layer and the second layer.

[0170] When the first n-type SiC layer is formed into a two-layer structure, Fig. 20 The structure shown, such as Fig. 27 As shown in FIG. 1 , the peripheral portion of the first n-type SiC layer is made into only one layer of the n-type SiC layer 4, and the thickness of the n-type SiC layer 4 which is only one layer can be made thinner. In this way, by making the thickness of the first n-type SiC layer in the peripheral portion of the chip thinner, such as using Fig.12 This indicates that it is difficult for the main current of the MOSFET to flow through the peripheral portion of the chip, and when a large current flows through the MOSFET as the main current, the current carrying capacity increases.

[0171] In addition, it is also possible to Fig.28 As shown, a portion of the second n-type SiC layer 70 is left at the periphery of the first n-type SiC. With such a structure, the bottom of the pillar contacts the epitaxial layer of the same impurity concentration, so it is easy to design to ensure the electric field balance at the contact portion, and it is easy to maintain the withstand voltage.

[0172] <Variation 2 of the Method for Manufacturing the Column>

[0173] like Fig.19 In the structure shown in FIG. 1 , when the first n-type SiC layer is made into a single-layer structure, as shown in FIG. Fig.29 As shown, when the n-type SiC layer 70 is etched by dry etching, the semiconductor substrate 3 can be partially removed by overetching, forming projections and depressions on the semiconductor substrate 3. With such a structure, the bottom of the pillar contacts the semiconductor substrate 3 with a high impurity concentration, so that the resistance can be reduced.

[0174] In addition, if Fig.30 As shown in FIG. 1 , when the n-type SiC layer 70 is etched by dry etching, the etching is performed in a manner that does not cause over-etching, so that the first n-type SiC, that is, the n-type SiC layer 70, does not remain under the convex portion 72. With such a structure, the bottom of the pillar contacts the semiconductor substrate 3 with a high impurity concentration, so that the resistance can be reduced, and compared with Fig.29 The structure shown makes it easy to design a balanced electric field.

[0175] <Method for Manufacturing Semiconductor Device>

[0176] Fig.31 1 is a perspective view schematically showing a silicon carbide semiconductor device 200 in which a MOSFET unit cell is provided in a pillar portion 7 and a MOSFET termination region MTR provided on the periphery thereof.

[0177] use Figure 32 to Figure 35 , describing the manufacturing process of the MOSFET cell in the MOSFET cell region MCR. Fig.32 is with Fig.24 The corresponding cross-sectional diagram, Fig.33 Show Fig.32 A magnified view of region B in FIG. Fig.33 About and use Figure 7 The same structures as described above are assigned the same reference numerals, and duplicate descriptions are omitted.

[0178] like Fig.33As shown, an n-type SiC layer 70 is provided on one main surface of a semiconductor substrate 3, and a plurality of p-type pillars 7p and a plurality of n-type pillars 7n are alternately provided on the n-type SiC layer 70. Furthermore, a plurality of p-type well regions 8 are selectively formed by ion implantation of p-type impurities from the upper layer of the p-type pillars 7p to the upper layer of the n-type pillars 7n. In addition, in each well region 8, a p-type contact region 10 is formed by ion implantation of p-type impurities in a manner penetrating the well region 8. In addition, in the upper layer of the well region 8, an n-type source region 9 is formed by ion implantation of n-type impurities in a manner contacting both side surfaces of the contact region 10. In order to restore the crystal defects formed by the ion implantation, activation annealing is performed to activate the implanted impurities.

[0179] The thickness of the p-type well region 8 can be, for example, 0.2 to 1.5 μm, and the concentration of the p-type impurity can be, for example, 1×10 14 ~1×10 19 cm -3 The thickness of the source region 9 can be, for example, 0.1 to 0.5 μm, and the concentration of the n-type impurity can be, for example, 1×10 17 ~1×10 21 cm -3 The thickness of the contact region 10 can be, for example, 0.2 to 1.5 μm, and the concentration of the p-type impurity can be, for example, 1×10 18 ~1×10 21 m -3 In addition, in the drawings, the well region 8 and the contact region 10 are depicted as having the same thickness, but the thicknesses of the well region 8 and the contact region 10 are not limited to being the same.

[0180] Afterwards, in Fig.34 In the process shown, an insulating film of a material to be the gate insulating film 11, such as a silicon oxide film 111, is formed on the pillar portion 7, and further, a conductive film, such as a polysilicon film, is formed on the silicon oxide film 111 to be the gate electrode 12. The polysilicon film is patterned to form the gate electrode 12 so as to extend over the edge of the adjacent source region 9. Thereafter, an insulating film of a material to be the interlayer insulating film 13, such as a silicon oxide film 131, is formed so as to cover the gate electrode 12 and the silicon oxide film 111.

[0181] Afterwards, in Fig.35In the process shown, the silicon oxide films 111 and 131 are patterned to form an interlayer insulating film 13 covering the gate insulating film 11 and the gate electrode 12. In this patterning, a contact hole is formed that penetrates the interlayer insulating film 13 in the thickness direction and reaches a part of the source region 9 and the entire surface of the contact region 10. After that, a conductor film is formed in a manner that is buried in the contact hole and covers the interlayer insulating film 13 to form the source electrode 14, and a drain electrode 15 is formed on the other main surface (back surface) of the semiconductor substrate 3 on the side opposite to the side where the source electrode 14 is provided, thereby completing the silicon carbide semiconductor device 200. In addition, an example of a method for manufacturing the MOSFET cell region MCR and the MOSFET terminal region MTR will be further described later.

[0182] exist Fig.35 In the silicon carbide semiconductor device 200 shown in the figure, a plurality of p-type well regions 8 are selectively provided from the upper portion of the p-type pillar 7p to the upper portion of the n-type pillar 7n, and a p-type contact region 10 and an n-type source region 9 are provided in each well region 8. However, it is possible to have a structure in which, instead of providing the p-type well region 8 in the upper portion of the n-type pillar 7n and the p-type pillar 7p, an n-type SiC layer 40 (third semiconductor layer) is provided on the n-type pillar 7n and the p-type pillar 7p, and a p-type well region 8 is provided in the n-type SiC layer 40. As another example of a method for manufacturing a semiconductor device, the manufacturing process thereof is described below. Figures 36 to 40 Provide explanation.

[0183] In use Fig.17 After the process described, Fig.36 In the process shown, an n-type SiC layer 40 is formed by epitaxial growth in a manner covering the area from the column portion 7 to the p-type chip peripheral portion 5. The n-type SiC layer 40 can be formed on the entire surface of the semiconductor chip, can be formed only in the MOSFET cell region where the MOSFET cell is formed, and can also be formed in the MOSFET cell region and the MOSFET terminal region.

[0184] Fig.37 It is shown Fig.36 Figure 1 shows an enlarged area of ​​the Fig.38 Show Fig.37 Magnified view of area C in . In addition, Fig.38 About and use Figure 7 The same structures as described above are assigned the same reference numerals, and duplicate descriptions are omitted.

[0185] like Fig.38As shown, a plurality of p-type well regions 8 are selectively formed in the n-type SiC layer 40 from above the p-type pillar 7p to above the n-type pillar 7n by ion implantation of p-type impurities. In addition, in each well region 8, a p-type contact region 10 is formed by ion implantation of p-type impurities in a manner penetrating the well region 8. In addition, in the upper layer of the well region 8, an n-type source region 9 is formed by ion implantation of n-type impurities in a manner contacting both side surfaces of the contact region 10. In order to restore the crystal defects formed by the ion implantation, activation annealing is performed to activate the implanted impurities.

[0186] Afterwards, in Fig.39 In the process shown, an insulating film of a material to be the gate insulating film 11, such as a silicon oxide film 111, is formed on the n-type SiC layer 40, and further, a conductive film, such as a polysilicon film, to be the gate electrode 12 is formed on the silicon oxide film 111. Furthermore, the polysilicon film is patterned to form the gate electrode 12 so as to straddle over the edge of the adjacent source region 9. Thereafter, an insulating film of a material to be the interlayer insulating film 13, such as a silicon oxide film 131, is formed so as to cover the gate electrode 12 and the silicon oxide film 111.

[0187] Afterwards, in Fig.40 In the process shown, the silicon oxide films 111 and 131 are patterned to form an interlayer insulating film 13 covering the gate insulating film 11 and the gate electrode 12. In this patterning, a contact hole is formed that penetrates the interlayer insulating film 13 in the thickness direction and reaches a part of the source region 9 and the entire surface of the contact region 10. Thereafter, a conductor film is formed in a manner that is buried in the contact hole and covers the interlayer insulating film 13 to form a source electrode 14, and a drain electrode 15 is formed on the other main surface (back surface) of the semiconductor substrate 3 on the side opposite to the side where the source electrode 14 is provided, thereby completing the silicon carbide semiconductor device 200A.

[0188] In addition, Fig.40 In the silicon carbide semiconductor device 200A shown in FIG. 1 , the thickness of the p-type well region 8 is made equal to the thickness of the n-type SiC layer 40. Fig.41 As shown, the p-type well region 8 is formed deeper than the thickness of the n-type SiC layer 40. With such a structure, the corners of the p-type well region 8 do not contact the high-concentration n-type SiC layer 40, and when a high voltage is applied to the MOSFET, the electric field strength at the corners of the p-type well region 8 can be reduced to maintain the withstand voltage. In addition, the depth of the p-type well region 8 is about 0.1 to 1 μm deeper than the thickness of the n-type SiC layer 40.

[0189] In the silicon carbide semiconductor devices 200 and 200A according to the second embodiment described above, voids are not formed when the p-type pillars 7p are formed by epitaxial growth, and the ineffective region can be reduced. Therefore, processing for isolating voids is not required, and manufacturing costs can be reduced.

[0190] In addition, Fig.35 as well as Fig.40 , a structure in which two unit cells of MOSFET are provided is shown, but in reality, the number of unit cells corresponding to the size of the carbide semiconductor device is formed.

[0191] <Variation 1 of the Method for Manufacturing a Semiconductor Device>

[0192] use Fig.42 as well as Fig.43 , Modification 1 of the method for manufacturing a silicon carbide semiconductor device according to Embodiment 2 is described. In addition, Fig.42 as well as Fig.43 are respectively Fig. 22 as well as Fig.23 The corresponding figure.

[0193] If used Fig.15 as well as Fig. 20 In the embodiment, after forming the convex portion 72 of the n-type SiC layer 70 and the n-type chip peripheral portion 71 of the n-type SiC layer 70, when forming the p-type SiC layer 60 by epitaxial growth so as to cover the n-type chip peripheral portion 71 of the n-type SiC layer 70 and the convex portion 72, as shown in FIG. Fig.42 as well as Fig.43 As shown, even after the p-type SiC layer 60 is embedded between the protrusions 72 of the n-type SiC layer 70 , epitaxial growth continues, so that the width 6W of the p-type pillar peripheral portion 6 is formed to be significantly larger than the width 7pW of the p-type pillar 7p .

[0194] As a result, when the impurity concentrations in the p-type pillar peripheral portion 6 and the p-type pillar 7p are the same, the product of the impurity concentration and the width of the p-type pillar peripheral portion 6 is significantly greater than the product of the impurity concentration and the width of the p-type pillar 7p. With such a structure, the depletion layer does not extend to the entire p-type pillar peripheral portion 6, the electric field strength of the p-type pillar peripheral portion 6 is suppressed to be low, and the increase in leakage current and the risk of discharge are suppressed.

[0195] <Variation 2 of the method for manufacturing a semiconductor device>

[0196] use Fig.44 as well as Fig.45 , Modification 2 of the method for manufacturing a silicon carbide semiconductor device of Embodiment 2 is described. In addition, Fig.44 as well as Fig.45 are respectively Fig. 22 as well as Fig.23The corresponding figure.

[0197] If used Fig.15 as well as Fig. 20 In the embodiment, after forming the convex portion 72 of the n-type SiC layer 70 and the n-type chip peripheral portion 71 of the n-type SiC layer 70, when forming the p-type SiC layer 60 by epitaxial growth so as to cover the n-type chip peripheral portion 71 of the n-type SiC layer 70 and the convex portion 72, as shown in FIG. Fig.44 as well as Fig.45 As shown, the p-type impurity concentration of the p-type pillar peripheral portion 6 is formed to be higher than the p-type impurity concentration of the p-type pillar 7p. For this reason, the epitaxial growth conditions of the p-type SiC layer 60 are adjusted so that the p-type pillar peripheral portion 6 is more likely to take in p-type impurities than between the convex portions 72 of the n-type SiC layer 70.

[0198] That is, by adjusting the raw material gas ratio, temperature, and pressure during epitaxial growth, the amount of impurities taken in based on the crystal surface changes. The p-type pillar peripheral portion 6 is almost limited to crystal growth from the bottom surface of the groove, but because the crystal grows from multiple surfaces of the bottom surface of the groove and the side wall of the groove between the convex portion 72, the amount of impurities taken in is different. In addition, the ease of supply of raw material gas and impurity gas between the p-type pillar peripheral portion 6 dug with a wide area and the convex portion 72 dug with a narrow width also has an effect, so these factors are considered to adjust the epitaxial conditions of the p-type SiC layer 60.

[0199] As a result, even when the width 6W of the p-type pillar peripheral portion 6 and the width 7pW of the p-type pillar 7p are the same, the product of the impurity concentration and the width of the p-type pillar peripheral portion 6 is greater than the product of the impurity concentration and the width of the p-type pillar 7p. With such a structure, the depletion layer does not extend to the entire p-type pillar peripheral portion 6, the electric field intensity of the p-type pillar peripheral portion 6 is suppressed to be low, and the increase in leakage current and the risk of discharge are suppressed.

[0200] In addition, by combining Figure 42 to Figure 45 By forming the p-type SiC layer 60 with a structure of , the product of the impurity concentration and the width of the p-type pillar peripheral portion 6 can be made much larger than the product of the impurity concentration and the width of the p-type pillar 7p. With such a structure, the depletion layer does not extend to the entire p-type pillar peripheral portion 6, the electric field intensity of the p-type pillar peripheral portion 6 is suppressed to a low level, and the increase in leakage current and the risk of discharge are suppressed.

[0201] <Variation 3 of the method for manufacturing a semiconductor device>

[0202] use Fig.46 as well as Fig.47 , Modification 3 of the method for manufacturing a silicon carbide semiconductor device according to the second embodiment is described. In addition, Fig.46 as well as Fig.47 are respectively Fig. 22as well as Fig.23 The corresponding figure.

[0203] If used Fig.15 as well as Fig. 20 In the embodiment, after forming the convex portion 72 of the n-type SiC layer 70 and the n-type chip peripheral portion 71 of the n-type SiC layer 70, after forming the p-type SiC layer 60 by epitaxial growth so as to cover the n-type chip peripheral portion 71 of the n-type SiC layer 70 and the convex portion 72, as shown in FIG. Fig.46 as well as Fig.47 As shown, ion implantation of p-type impurities is performed in the p-type pillar peripheral portion 6 , so that the concentration of the p-type impurities in the p-type pillar peripheral portion 6 is higher than the concentration of the p-type impurities in the p-type pillar 7 p .

[0204] As a result, the product of the impurity concentration and the width of the p-type pillar peripheral portion 6 is greater than the product of the impurity concentration and the width of the p-type pillar 7p. With such a structure, the depletion layer does not extend to the entire p-type pillar peripheral portion 6, the electric field strength of the p-type pillar peripheral portion 6 is suppressed low, and the increase in leakage current and the risk of discharge are suppressed.

[0205] In addition, ion implantation into the p-type pillar peripheral portion 6 can be performed only at necessary locations. In the case of SiC, the epitaxial growth rate and epitaxial concentration may differ depending on the crystal orientation. As a result, when the p-type SiC layer 60 is epitaxially grown, the product of the impurity concentration and the width of the p-type pillar peripheral portion 6 may be unevenly formed, and a location may be formed where the product of the impurity concentration and the width of the p-type pillar peripheral portion 6 is smaller than the product of the impurity concentration and the width of the p-type pillar 7p. In this case, ion implantation is performed only in the portion where the product of the concentration and the width of the p-type pillar peripheral portion 6 is smaller than the product of the impurity concentration and the width of the p-type pillar peripheral portion 6, so that the product of the concentration and the width of the p-type pillar peripheral portion 6 can be made greater than the product of the impurity concentration and the width of the p-type pillar peripheral portion 6.

[0206] In addition, if the semiconductor device has the same structure and size, the place where the product of the impurity concentration and the width of the p-type pillar peripheral portion 6 is smaller than the product of the impurity concentration and the width of the p-type pillar 7p always occurs in the same place, so it can be determined by making samples and decomposing and verifying them.

[0207] In addition, as a reason why the product of the impurity concentration and the width of the p-type column peripheral portion 6 is smaller than the product of the impurity concentration and the width of the p-type column 7p, it is considered that depending on the epitaxial conditions, the width of the p-type column peripheral portion 6 becomes narrower or the amount of p-type impurities introduced based on the crystal surface is small and the concentration of the p-type impurities becomes lower.

[0208] <Variation 4 of the method for manufacturing a semiconductor device>

[0209] use Fig.48, Modification 4 of the method for manufacturing a silicon carbide semiconductor device of Embodiment 2 is described. In addition, Fig.48 is with Fig.24 The corresponding figure.

[0210] If used Fig.17 as well as Fig.24 In the embodiment, when the p-type SiC layer 60 on the convex portion 72 of the n-type SiC layer 70 is removed by grinding or dry etching to expose the upper surface of the convex portion 72, as shown in FIG. Fig.48 As shown, the p-type chip peripheral portion 5 outside the p-type column peripheral portion 6 is removed.

[0211] With such a structure, the p-type chip peripheral portion 5 can be removed simultaneously with the p-type SiC layer 60 on the convex portion 72 of the p-type SiC layer 60 by full-surface etching, thereby simplifying the manufacturing process. At this time, a part of the n-type chip peripheral portion 71 of the n-type SiC layer 70 can also be removed.

[0212] <Variation 5 of the method for manufacturing a semiconductor device>

[0213] use Fig.49 as well as Fig.50 , Modification 5 of the method for manufacturing a silicon carbide semiconductor device according to Embodiment 2 is described. In addition, Fig.49 as well as Fig.50 are respectively Fig. 20 as well as Fig.21 The corresponding figure.

[0214] If used Fig. 27 In this way, when the peripheral portion of the first n-type SiC layer is made into only one layer of the n-type SiC layer 4, the thickness of the n-type SiC layer 4 that is only one layer can also be made thinner. As a result, the height 4H2 of the column peripheral portion of the n-type SiC layer 4 is lower than the height 4H1 of the n-type SiC layer 4 at the column portion 7. In this way, by making the thickness of the first n-type SiC layer in the peripheral portion of the chip thinner, as in the case of using Fig.12 This indicates that it is difficult for the main current of the MOSFET to flow through the peripheral portion of the chip, and when a large current flows through the MOSFET as the main current, the current carrying capacity increases.

[0215] <Cutting process>

[0216] use Fig.51 as well as Fig.52 , describing the dicing process of separating semiconductor devices manufactured through wafer processing into chip units. Fig.51 as well as Fig.52 This corresponds to a cross-sectional view showing adjacent silicon carbide semiconductor devices 200 in a wafer state. Fig.51 is along Fig.18 The cross-sectional view in the direction indicated by the arrow of line AA, Fig.52 Along with Fig.18 The cross-sectional view along the arrow line BB of FIG. 200 is equivalent to FIG. Fig.35 The state shown in FIG. 1 is different from that shown in FIG. 1 , but the source electrode 14 and the drain electrode 15 are omitted.

[0217] exist Fig.51 as well as Fig.52 In the embodiment, the two silicon carbide semiconductor devices 200 are cut at the position indicated by the arrow between them and separated into chip units. In order to perform such cutting, the interval between the protrusions 72 of the n-type SiC layer 70 in the adjacent silicon carbide semiconductor devices 200 is made 10 times or more than the column width of the n-type column 7n and the p-type column 7p of 0.5 to 5 μm, for example, 50 μm or more.

[0218] exist Fig.53 as well as Fig.54 In Fig. 20 The convex portion 72 of the n-type SiC layer 70 in the step shown, Fig.53 as well as Fig.54 Respectively Fig.51 as well as Fig.52 Corresponding. Fig.53 In FIG. 1 , the interval between the protrusions 72 in the adjacent silicon carbide semiconductor devices is represented as d1. Fig.54 In FIG. 1 , the interval between the protrusions 72 in adjacent silicon carbide semiconductor devices is represented as d2.

[0219] By widening the intervals between the convex portions 72 of the n-type SiC layer 70 in the adjacent silicon carbide semiconductor devices, dicing can be performed, and the formation of voids when the p-type SiC layer 60 is formed by epitaxial growth can be suppressed. Fig.54 The spacing d2 of the protrusions 72 in the direction shown is greater than Fig.53 The wide interval d1 between the protrusions 72 in the direction shown can more effectively suppress the formation of voids.

[0220] <Implementation Method 3>

[0221] <Method for Manufacturing the Column>

[0222] Next, as Embodiment 3, a method for manufacturing the pillar portion of the silicon carbide semiconductor device 300 will be described. Fig.13 For example, a commercially available 4H-SiC n-type substrate is prepared with an off angle of 4 degrees in the [11-20] direction, a thickness of 300 to 400 μm, and an n-type impurity concentration of 5×10 18 ~1×10 20 cm -3 The semiconductor substrate 3.

[0223] Next, in Fig.55 In the process shown, an n-type SiC layer 4 (semiconductor layer) containing n-type impurities is formed by epitaxial growth on one main surface of the semiconductor substrate 3. The n-type SiC layer 4 can be made, for example, to have a thickness of 0.5 to 10 μm and an impurity concentration of 1×10 14 ~1×10 19 cm -3 Furthermore, the formation of the n-type SiC layer 4 can be omitted.

[0224] Next, in Fig.56 In the illustrated step, a p-type SiC layer 60 (first semiconductor layer) containing a p-type impurity is formed on the n-type SiC layer 4 by epitaxial growth.

[0225] Next, in Fig.57 In the process shown, the p-type SiC layer 60 is etched by dry etching to form a convex portion 62 of the p-type SiC layer 60, and the p-type SiC layer 60 around the convex portion 62 is removed to expose the surface of the peripheral portion of the n-type SiC layer 4. At this time, a part of the n-type SiC layer 4 can be removed to form a concave-convex surface on the surface of the n-type SiC layer 4.

[0226] When the semiconductor substrate 3 has an off angle of 4 degrees in the [11-20] direction, the convex portion 62 is formed to have a stripe shape extending in a direction parallel to the X-axis ([11-20] direction) when viewed from above, and a plurality of convex portions 62 are arranged at intervals in the direction along the Y-axis ([1-100] direction). The p-type SiC layer 60 has a symmetrical shape in a direction perpendicular to the [11-20] direction, so when an n-type SiC layer is formed between the convex portions 62 by epitaxial growth, it is easy to control the shape of the n-type SiC layer. In addition, the convex portion 62 is formed to have a width in the Y-axis direction of, for example, 0.5 to 5 μm, and a height in the Z-axis (

[0001] direction) direction of, for example, 5 to 100 μm.

[0227] On the other hand, when a substrate without an off angle is used as the semiconductor substrate 3, the stripe-shaped protrusions 62 can be formed in an orientation rotated 90 degrees around the

[0001] axis from the [11-20] direction. Fig.57 The [11-20] axis and the [1-100] axis are inverted to form the projections 62. Since the n-type SiC layer formed between the projections 62 by epitaxial growth has a symmetrical shape, it is easy to control the shape of the n-type SiC layer.

[0228] Next, in Fig.58In the process shown, an n-type SiC layer 70 (second semiconductor layer) is formed by epitaxial growth so as to cover the convex portion 62 of the p-type SiC layer 60 and the peripheral portion of the n-type SiC layer 4 around the convex portion 62. Fig.57 In FIG. 5 , the surface of the n-type SiC layer 70 of the convex portion 62 of the p-type SiC layer 60 is flat, but may have concavities and convexities.

[0229] like Fig.57 As shown, by epitaxially growing the n-type SiC layer 70 until the surface of the n-type SiC layer 4 exposes the periphery of the convex portion 62 of the p-type SiC layer 60, voids are not formed due to the difference in crystal growth rate caused by the difference in crystal plane orientation, and the ineffective area can be reduced. Therefore, processing for separating voids becomes unnecessary, and the manufacturing cost can be reduced.

[0230] The concentration of the n-type impurity of the n-type SiC layer 70 can be set so that the product of the impurity concentration and the width of the p-type pillar 7 p is substantially equal to the product of the impurity concentration and the width of the n-type pillar 7 n.

[0231] Next, in Fig.59 In the process shown, the n-type SiC layer 70 on the convex portion 62 of the p-type SiC layer 60 is removed by grinding or dry etching to expose the upper surface of the convex portion 62. At this time, the upper portion of the convex portion 62 of the p-type SiC layer 60 can be partially removed. In addition, the upper portion of the n-type chip peripheral portion 71 of the n-type SiC layer 70 can be partially removed.

[0232] Fig.60 FIG. 2 is a perspective view showing a MOSFET cell region MCR and a MOSFET terminal region MTR of a silicon carbide semiconductor device 300 formed by the above method before being formed. Fig.60 The columnar portion 7 shown is a region where a plurality of n-type columns 7n and a plurality of p-type columns 7p are alternately arranged in a stripe shape when viewed from above on the semiconductor substrate 3, and their arrangement direction is a direction orthogonal to the direction in which the main current of the silicon carbide semiconductor device 300 flows. An n-type column peripheral portion 73 containing n-type impurities is provided to surround the columnar portion 7, and the outer side of the n-type column peripheral portion 73 becomes an n-type chip peripheral portion 71 containing n-type impurities.

[0233] In addition, Figures 55 to 60 , a structure is shown in which three p-type pillars 7p and four n-type pillars 7n are provided in the pillar portion 7, but in reality, the number of p-type pillars 7p and n-type pillars 7n corresponding to the size of the carbide semiconductor device is formed.

[0234] Next, Figure 61 to Figure 66 Shown along Fig.60A cross-sectional view in the direction indicated by the arrow along line AA (a line parallel to the Y axis) and a cross-sectional view in the direction indicated by the arrow along line BB (a line parallel to the X axis). Fig.61 yes Fig.55 The cross-sectional view along line AA in the process shown in FIG. Fig.55 same. Fig.62 yes Fig.56 The cross-sectional view along line AA in the process shown in FIG. Fig.56 same. Fig.63 yes Fig.57 The cross-sectional view along line AA in the process shown, Fig.64 It is a cross-sectional view along line BB. Fig.65 yes Fig.58 The cross-sectional view along line AA in the process shown, Fig.66 It is a cross-sectional view along line BB. Fig.67 yes Fig.59 The cross-sectional view along line AA in the process shown, Fig.68 is a cross-sectional view along line BB. In addition, Fig.67 As shown, by making the width 73W of the n-type pillar peripheral portion 73 thicker than the width 7nW of the n-type pillar 7n, when the impurity concentrations in the n-type pillar peripheral portion 73 and the n-type pillar 7n are the same, the product of the impurity concentration and the width of the n-type pillar peripheral portion 73 is greater than the product of the impurity concentration and the width of the n-type pillar 7n.

[0235] With such a structure, the depletion layer does not extend to the entire n-type pillar peripheral portion 73, the electric field strength of the n-type pillar peripheral portion 73 is suppressed to be low, and the increase in leakage current and the risk of discharge are suppressed. In this case, the width of the n-type pillar peripheral portion 73 can be 1.2 times or more of the width of the n-type pillar 7n.

[0236] <Method for Manufacturing Semiconductor Device>

[0237] Fig.69 1 is a perspective view schematically showing a silicon carbide semiconductor device 300 in which a MOSFET unit cell is provided in a pillar portion 7 and a MOSFET termination region MTR provided on the periphery thereof.

[0238] use Figure 70 to Figure 73 , describing the manufacturing process of the MOSFET cell in the MOSFET cell region MCR. Fig.72 is with Fig.67 The corresponding cross-sectional diagram, Fig.71 Show Fig.70 A magnified view of region D in FIG. Fig.71 About and use Figure 7The same structures as described above are assigned the same reference numerals, and duplicate descriptions are omitted.

[0239] like Fig.71 As shown, an n-type SiC layer 4 is provided on one main surface of a semiconductor substrate 3, and a plurality of p-type pillars 7p and a plurality of n-type pillars 7n are alternately provided on the n-type SiC layer 4. Furthermore, a plurality of p-type well regions 8 are selectively formed by ion implantation of p-type impurities from the upper portion of the p-type pillars 7p to the upper portion of the n-type pillars 7n. In addition, in each well region 8, a p-type contact region 10 is formed by ion implantation of p-type impurities in a manner penetrating the well region 8. In addition, in the upper portion of the well region 8, an n-type source region 9 is formed by ion implantation of n-type impurities in a manner contacting both side surfaces of the contact region 10. In order to restore the crystal defects formed by the ion implantation, activation annealing is performed to activate the implanted impurities.

[0240] The thickness of the p-type well region 8 can be, for example, 0.2 to 1.5 μm, and the concentration of the p-type impurity can be, for example, 1×10 14 ~1×10 19 cm -3 The thickness of the source region 9 can be, for example, 0.1 to 0.5 μm, and the concentration of the n-type impurity can be, for example, 1×10 17 ~1×10 21 cm -3 The thickness of the contact region 10 can be, for example, 0.2 to 1.5 μm, and the concentration of the p-type impurity can be, for example, 1×10 18 ~1×10 21 m -3 .

[0241] Afterwards, in Fig.72 In the process shown, an insulating film of a material to be the gate insulating film 11, such as a silicon oxide film 111, is formed on the pillar portion 7, and further, a conductive film, such as a polysilicon film, is formed on the silicon oxide film 111 to be the gate electrode 12. The polysilicon film is patterned to form the gate electrode 12 so as to extend over the edge of the adjacent source region 9. Thereafter, an insulating film of a material to be the interlayer insulating film 13, such as a silicon oxide film 131, is formed so as to cover the gate electrode 12 and the silicon oxide film 111.

[0242] Afterwards, in Fig.73In the process shown, the silicon oxide films 111 and 131 are patterned to form an interlayer insulating film 13 covering the gate insulating film 11 and the gate electrode 12. In this patterning, a contact hole is formed that penetrates the interlayer insulating film 13 in the thickness direction and reaches a part of the source region 9 and the entire surface of the contact region 10. Thereafter, a conductor film is formed in a manner that is buried in the contact hole and covers the interlayer insulating film 13 to form a source electrode 14, and a drain electrode 15 is formed on the other main surface (back surface) of the semiconductor substrate 3 on the side opposite to the side where the source electrode 14 is provided, thereby completing the silicon carbide semiconductor device 300.

[0243] exist Fig.73 In the silicon carbide semiconductor device 300 shown, a plurality of p-type well regions 8 are selectively provided from the upper portion of the p-type pillar 7p to the upper portion of the n-type pillar 7n, and a p-type contact region 10 and an n-type source region 9 are provided in each well region 8. However, it is possible to have a structure in which, instead of providing the p-type well region 8 in the upper portion of the n-type pillar 7n and the p-type pillar 7p, an n-type SiC layer 40 (third n-type SiC layer) is provided on the n-type pillar 7n and the p-type pillar 7p, and a p-type well region 8 is provided in the n-type SiC layer 40. As another example of a method for manufacturing a semiconductor device, the manufacturing process thereof is described below. Figure 74 to Figure 78 Provide explanation.

[0244] In use Fig.59 After the process described, Fig.74 In the process shown, the n-type SiC layer 40 is formed by epitaxial growth in a manner covering the area from the column portion 7 to the n-type chip peripheral portion 71 of the n-type SiC layer 70. The n-type SiC layer 40 can be formed on the entire surface of the semiconductor chip, can be formed only in the MOSFET cell region where the MOSFET cell is formed, and can also be formed in the MOSFET cell region and the MOSFET terminal region.

[0245] Fig.75 It is shown Fig.74 Figure 1 shows an enlarged area of ​​the Fig.76 Show Fig.75 Magnified view of region E in FIG. Fig.76 About and use Figure 7 The same structures as described above are assigned the same reference numerals, and duplicate descriptions are omitted.

[0246] like Fig.76As shown, a plurality of p-type well regions 8 are selectively formed in the n-type SiC layer 40 from above the p-type pillar 7p to above the n-type pillar 7n by ion implantation of p-type impurities. In addition, in each well region 8, a p-type contact region 10 is formed by ion implantation of p-type impurities in a manner penetrating the well region 8. In addition, in the upper layer of the well region 8, an n-type source region 9 is formed by ion implantation of n-type impurities in a manner contacting both side surfaces of the contact region 10. In order to restore the crystal defects formed by the ion implantation, activation annealing is performed to activate the implanted impurities.

[0247] Afterwards, in Fig.77 In the process shown, an insulating film of a material to be the gate insulating film 11, such as a silicon oxide film 111, is formed on the n-type SiC layer 40, and further, a conductive film, such as a polysilicon film, is formed on the silicon oxide film 111 to be the gate electrode 12. The polysilicon film is patterned to form the gate electrode 12 so as to straddle the edge of the adjacent source region 9. Thereafter, an insulating film of a material to be the interlayer insulating film 13, such as a silicon oxide film 131, is formed so as to cover the gate electrode 12 and the silicon oxide film 111.

[0248] Afterwards, in Fig.78 In the process shown, the silicon oxide films 111 and 131 are patterned to form an interlayer insulating film 13 covering the gate insulating film 11 and the gate electrode 12. In this patterning, a contact hole is formed that penetrates the interlayer insulating film 13 in the thickness direction and reaches a part of the source region 9 and the entire surface of the contact region 10. Thereafter, a conductor film is formed in a manner that is buried in the contact hole and covers the interlayer insulating film 13 to form a source electrode 14, and a drain electrode 15 is formed on the other main surface (back surface) of the semiconductor substrate 3 on the side opposite to the side where the source electrode 14 is provided, thereby completing the silicon carbide semiconductor device 300A.

[0249] In the silicon carbide semiconductor devices 300 and 300A of the second embodiment described above, voids are not formed when the n-type pillars 7n are formed by epitaxial growth, and the ineffective region can be reduced. Therefore, processing for isolating voids is not required, and the manufacturing cost can be reduced.

[0250] In addition, Fig.73 as well as Fig.78 , a structure in which two unit cells of MOSFET are provided is shown, but in reality, the number of unit cells corresponding to the size of the carbide semiconductor device is formed.

[0251] <Variation 1 of the Method for Manufacturing a Semiconductor Device>

[0252] use Fig.79 as well as Fig.80, Modification 1 of the method for manufacturing a silicon carbide semiconductor device according to Embodiment 3 is described. In addition, Fig.79 as well as Fig.80 are respectively Fig.65 as well as Fig.66 The corresponding figure.

[0253] If used Fig.57 as well as Fig.63 In the embodiment, after the convex portion 62 of the p-type SiC layer 60 is formed and the surface of the peripheral portion of the n-type SiC layer 4 is exposed, when the n-type SiC layer 70 is formed by epitaxial growth so as to cover the convex portion 62 of the p-type SiC layer 60 and the peripheral portion of the n-type SiC layer 4 around the convex portion 62, as shown in FIG. Fig.79 as well as Fig.80 As shown, even after n-type SiC layer 70 is embedded between protrusions 62 of p-type SiC layer 60 , epitaxial growth continues, so that width 73W of n-type pillar peripheral portion 73 is formed to be significantly larger than width 7nW of n-type pillar 7n .

[0254] As a result, when the impurity concentrations in the n-type pillar peripheral portion 73 and the n-type pillar 7n are the same, the product of the impurity concentration and the width of the n-type pillar peripheral portion 73 is significantly greater than the product of the impurity concentration and the width of the n-type pillar 7n. With such a structure, the depletion layer does not extend to the entire n-type pillar peripheral portion 73, the electric field strength of the n-type pillar peripheral portion 73 is suppressed to be low, and the increase in leakage current and the risk of discharge are suppressed.

[0255] <Variation 2 of the method for manufacturing a semiconductor device>

[0256] use Fig.81 as well as Fig.82 , Modification 2 of the method for manufacturing a silicon carbide semiconductor device according to Embodiment 3 is described. In addition, Fig.81 as well as Fig.82 are respectively Fig.65 as well as Fig.66 The corresponding figure.

[0257] If used Fig.57 as well as Fig.63 In the embodiment, after the convex portion 62 of the p-type SiC layer 60 is formed and the surface of the peripheral portion of the n-type SiC layer 4 is exposed, when the n-type SiC layer 70 is formed by epitaxial growth so as to cover the convex portion 62 of the p-type SiC layer 60 and the peripheral portion of the n-type SiC layer 4 around the convex portion 62, as shown in FIG. Fig.81 as well as Fig.82As shown, the n-type impurity concentration of the n-type pillar peripheral portion 73 is higher than that of the n-type pillar 7n. For this reason, the epitaxial growth conditions of the n-type SiC layer 70 are adjusted so that the n-type pillar peripheral portion 73 is more likely to take in n-type impurities than between the convex portions 62 of the p-type SiC layer 60.

[0258] That is, by adjusting the raw material gas ratio, temperature, and pressure during epitaxial growth, the amount of impurities taken in based on the crystal surface changes. The n-type pillar peripheral portion 73 is almost limited to crystal growth from the bottom surface of the trench, but because the crystal grows from multiple surfaces of the bottom surface of the trench and the side wall of the trench between the convex portion 62, the amount of impurities taken in is different. In addition, the difference in the ease of supply of raw material gas and impurity gas between the n-type pillar peripheral portion 73 dug with a wide area and the convex portion 62 dug with a narrow width also affects it, so these factors are considered to adjust the epitaxial growth conditions of the n-type SiC layer 70.

[0259] As a result, even when the width 73W of the n-type pillar peripheral portion 73 and the width 7nW of the n-type pillar 7n are the same, the product of the impurity concentration and the width of the n-type pillar peripheral portion 73 is larger than the product of the impurity concentration and the width of the n-type pillar 7n. With such a structure, the depletion layer does not extend to the entire n-type pillar peripheral portion 73, the electric field strength of the n-type pillar peripheral portion 73 is suppressed to be low, and the increase in leakage current and the risk of discharge are suppressed.

[0260] In addition, by combining Figures 79 to 82 By forming the n-type SiC layer 70 with a structure of , the product of the impurity concentration and the width of the n-type pillar peripheral portion 73 can be made much larger than the product of the impurity concentration and the width of the n-type pillar 7n. With such a structure, the depletion layer does not extend to the entire n-type pillar peripheral portion 73, the electric field intensity of the n-type pillar peripheral portion 73 is suppressed to a low level, and the increase in leakage current and the risk of discharge are suppressed.

[0261] <Variation 3 of the method for manufacturing a semiconductor device>

[0262] use Fig.83 as well as Fig.84 , Modification 3 of the method for manufacturing a silicon carbide semiconductor device of Embodiment 3 is described. In addition, Fig.83 as well as Fig.84 are respectively Fig.65 as well as Fig.66 The corresponding figure.

[0263] If used Fig.57 as well as Fig.63In the embodiment, after the convex portion 62 of the p-type SiC layer 60 is formed and the surface of the peripheral portion of the n-type SiC layer 4 is exposed, when the n-type SiC layer 70 is formed by epitaxial growth so as to cover the convex portion 62 of the p-type SiC layer 60 and the peripheral portion of the n-type SiC layer 4 around the convex portion 62, as shown in FIG. Fig.83 as well as Fig.84 As shown, ion implantation of n-type impurities is performed in the n-type pillar peripheral portion 73 , so that the concentration of n-type impurities in the n-type pillar peripheral portion 73 is higher than the concentration of n-type impurities in the n-type pillar 7n.

[0264] As a result, the product of the impurity concentration and the width of the n-type pillar 73 is greater than the product of the impurity concentration and the width of the n-type pillar 7n. With such a structure, the depletion layer does not extend to the entire n-type pillar 73, the electric field strength of the n-type pillar 73 is suppressed low, and the increase of leakage current and the risk of discharge are suppressed.

[0265] In addition, ion implantation into the n-type pillar peripheral portion 73 can be performed only at a necessary location. In the case of SiC, the epitaxial growth rate and epitaxial concentration may differ depending on the crystal orientation. As a result, when the n-type SiC layer 70 is epitaxially grown, the product of the impurity concentration and the width of the n-type pillar peripheral portion 736 may be formed unevenly, and a location may be formed where the product of the impurity concentration and the width of the n-type pillar peripheral portion 73 is smaller than the product of the impurity concentration and the width of the n-type pillar 7n. In this case, ion implantation is performed only in a portion where the product of the concentration and the width of the n-type pillar peripheral portion 73 is smaller than the product of the impurity concentration and the width of the n-type pillar peripheral portion 73, so that the product of the concentration and the width of the n-type pillar peripheral portion 73 can be made larger than the product of the impurity concentration and the width of the n-type pillar peripheral portion 73.

[0266] Furthermore, if the semiconductor device has the same structure and size, the location where the product of the impurity concentration and the width of the n-type pillar peripheral portion 73 is smaller than the product of the impurity concentration and the width of the n-type pillar 7n always occurs at the same location, so it can be determined by making a sample and then dissecting and verifying it.

[0267] In addition, as a reason why the product of the impurity concentration and the width of the n-type column peripheral portion 73 is smaller than the product of the impurity concentration and the width of the n-type column 7n, it is considered that depending on the epitaxial conditions, the width of the n-type column peripheral portion 73 becomes narrower or the amount of n-type impurities introduced based on the crystal surface is small and the concentration of the n-type impurities becomes lower.

[0268] <Variation 4 of the method for manufacturing a semiconductor device>

[0269] use Fig.85 , Modification 4 of the method for manufacturing a silicon carbide semiconductor device according to Embodiment 3 is described. In addition, Fig.85 is with Fig.67 The corresponding figure.

[0270] If used Fig.59 as well as Fig.67 In the embodiment, when the n-type SiC layer 70 on the convex portion 62 of the p-type SiC layer 600 is removed by grinding or dry etching to expose the upper surface of the convex portion 62, as shown in FIG. Fig.85 As shown, the n-type chip peripheral portion 71 outside the n-type pillar peripheral portion 73 is removed.

[0271] With such a structure, n-type chip peripheral portion 71 can be removed simultaneously with n-type SiC layer 70 on convex portion 62 by full-surface etching, thereby simplifying the manufacturing process. At this time, a portion of the peripheral portion of n-type SiC layer 4 can also be removed.

[0272] <Variation 5 of the method for manufacturing a semiconductor device>

[0273] use Fig.86 as well as Fig.87 , Modification 5 of the method for manufacturing a silicon carbide semiconductor device of Embodiment 3 is described. In addition, Fig.86 as well as Fig.87 are respectively Fig.63 as well as Fig.64 The corresponding figure.

[0274] If used Fig.57 It is explained that when the p-type SiC layer 60 is etched by dry etching to form a convex portion 62 of the p-type SiC layer 60, and the p-type SiC layer 60 around the convex portion 62 is removed to expose the surface of the peripheral portion of the n-type SiC layer 4, a portion of the n-type SiC layer 4 is removed to form a concave-convex surface on the surface of the n-type SiC layer 4, so that the height 4H2 of the peripheral portion of the column of the n-type SiC layer 4 is lower than the height 4H1 of the n-type SiC layer 4 at the column portion 7.

[0275] Thus, by reducing the thickness of the n-type SiC layer 4 at the chip periphery, as in the case of using Fig.12 This indicates that it is difficult for the main current of the MOSFET to flow through the peripheral portion of the chip, and when a large current flows through the MOSFET as the main current, the current carrying capacity increases.

[0276] <Cutting process>

[0277] use Fig.88 as well as Fig.89 , describing the dicing process of separating semiconductor devices manufactured through wafer processing into chip units. Fig.88 as well as Fig.89 This corresponds to a cross-sectional view showing adjacent silicon carbide semiconductor devices 300 in a wafer state. Fig.88 is along Fig.69 The cross-sectional view in the direction indicated by the arrow of line AA, Fig.89 Along with Fig.69 The cross-sectional view along the arrow line BB of FIG. 300 is equivalent to FIG. Fig.73 The state shown in FIG. 1 is different from that shown in FIG. 1 , but the source electrode 14 and the drain electrode 15 are omitted.

[0278] exist Fig.88 as well as Fig.89 In the embodiment, the two silicon carbide semiconductor devices 300 are cut at the position indicated by the arrow between them and separated into chip units. In order to perform such cutting, the interval between the protrusions 62 of the p-type SiC layer 60 in the adjacent silicon carbide semiconductor devices 300 is made 10 times or more than the column width of the n-type column 7n and the p-type column 7p of 0.5 to 5 μm, for example, 50 μm or more.

[0279] exist Fig.90 as well as Fig.91 In Fig.63 The convex portion 62 of the p-type SiC layer 60 in the step shown, Fig.90 as well as Fig.91 Respectively Fig.88 as well as Fig.89 Corresponding. Fig.90 In FIG. 1 , the interval between the protrusions 62 in the adjacent silicon carbide semiconductor devices is represented as d1. Fig.91 In FIG. 1 , the interval between the protrusions 62 in adjacent silicon carbide semiconductor devices is represented as d2.

[0280] By widening the intervals between the convex portions 62 of the p-type SiC layer 60 in the adjacent silicon carbide semiconductor devices, dicing can be performed, and the formation of voids when the n-type SiC layer 70 is formed by epitaxial growth can be suppressed. Fig.91 The spacing d2 of the protrusions 62 in the direction shown is greater than Fig.90 The wide interval d1 between the protrusions 62 in the direction shown can more effectively suppress the formation of voids.

[0281] <Other Application Examples>

[0282] In the semiconductor devices of Embodiments 1 to 3 described above, a silicon carbide semiconductor device in which a MOSFET unit region and a MOSFET terminal region are formed in the columnar portion 7 is shown. However, the application of the present disclosure is not limited to MOSFETs. It is also possible to provide a silicon carbide semiconductor device in which an SBD region and an SBD terminal region are formed in the columnar portion 7, and it is also possible to provide a silicon carbide semiconductor device in which an IGBT (Insulated Gate Bipolar Transistor) unit region and an IGBT terminal region are formed in the columnar portion 7. In addition, the present disclosure is not limited to transistors, and it is also possible to provide a silicon carbide semiconductor device in which a pn diode region and a pn diode terminal region are formed in the columnar portion 7. As long as it is a vertical power device, the same effect can be obtained. In addition, in the semiconductor devices of Embodiments 1 to 3, the application to a silicon carbide semiconductor device is exemplified, but the present disclosure can also be applied to a silicon semiconductor device.

[0283] <An example of the MOSFET unit region>

[0284] Use Figure 92 to Figure 94 , to describe an example of the structure of the MOSFET unit region MCR in the semiconductor devices of Embodiments 1 to 3 described above.

[0285] Fig.92 is a cross-sectional perspective view schematically showing the unit cell structure of the MOSFET 101. Fig.93 is Fig.92 an enlarged view of a portion where the illustration of the source electrode 31 in Fig.94 is Fig.93 a view in which the illustration of the structure near the gate electrode 29 in

[0286] As Fig.92 shown, the MOSFET 101 has an n-type semiconductor substrate 21, a drain electrode 32, a super junction layer 90, a plurality of p-type well regions 25a, a plurality of n-type source regions 26a, a plurality of p-type well regions 25b, a plurality of n-type source regions 26b, a gate electrode 29, and a source electrode 31. The MOSFET 101 has a gate insulating film 28, a gate electrode 29, and an interlayer insulating film 30 in order to form a MOS structure. The MOSFET 101 has an epitaxial layer 22. The MOSFET 101 has a p-type contact region 27a and a p-type contact region 27b.

[0287] The semiconductor substrate 21 has a lower surface S1 and an upper surface S2 opposite to the lower surface S1. In addition, the XYZ coordinate system shown in the figure is configured so that the XY plane is parallel to the upper surface S2, and the Z axis is parallel to the thickness direction of the semiconductor substrate 21. The current path of the MOSFET 101 is formed to connect the lower surface S1 and the upper surface S2. Therefore, the MOSFET 101 is a so-called vertical switching device.

[0288] The epitaxial layer 22 is a layer formed by epitaxial growth on the upper surface S2 of the semiconductor substrate 21 . The epitaxial layer 22 has an n-type. Typically, the impurity concentration of the epitaxial layer 22 is lower than the impurity concentration of the semiconductor substrate 21 .

[0289] The super junction layer 90 is disposed on the upper surface S2 of the semiconductor substrate 21 via the epitaxial layer 22. The super junction layer 90 has a plurality of n-type pillars 23 and a plurality of p-type pillars 24 alternately arranged in the in-plane direction (XY in-plane direction) of the upper surface S2. Specifically, the n-type pillars 23 and the p-type pillars 24 are alternately arranged in the X direction of the in-plane direction (XY in-plane direction), and the n-type pillars 23 and the p-type pillars 24 are each arranged along the in-plane direction ( Figure 1 That is, in a layout parallel to the upper surface S2 of the semiconductor substrate 21, the n-type pillars 23 and the p-type pillars 24 are arranged in stripes.

[0290] The semiconductor substrate 21 , the epitaxial layer 22 , and the super junction layer 90 are made of SiC.

[0291] The p-type well region 25 a is provided in the upper layer portion of each of the p-type pillars 24 . The well region 25 a extends on the superjunction layer 90 to reach the n-type pillars 23 .

[0292] The n-type source region 26 a is provided in the upper layer portion of each well region 25 a and is separated from the n-type pillar 23 by the well region 25 a .

[0293] The p-type well region 25b is provided in the upper layer portion of each of the n-type pillars 23. The well region 25b is arranged away from the p-type pillars 24.

[0294] The n-type source region 26 b is provided in the upper layer portion of each well region 25 b and is separated from the n-type pillar 23 by the well region 25 b .

[0295] like Fig.94 As shown, in a layout parallel to the upper surface S2 of the semiconductor substrate 21, the well region 25a and the well region 25b are arranged in a stripe shape. In addition, the width of each well region 25b is smaller than the width of each well region 25a. In addition, the widths of each well region 25a may be the same, and the widths of each well region 25b may be the same.

[0296] The source electrode 31 is disposed on the upper surface S2 side of the semiconductor substrate 21 and is respectively joined to the well region 25a, the well region 25b, the source region 26a, and the source region 26b.

[0297] The gate electrode 29 is opposed to the well region 25a between the n-type column 23 and the source region 26a and is opposed to the well region 25b between the n-type column 23 and the source region 26b with the gate insulating film 28 therebetween. As shown in Fig.93 the figure, the gate electrode 29 has a striped planar layout. The interlayer insulating film 30 insulates between the gate electrode 29 and the source electrode 31.

[0298] <An example of the MOSFET cell region and the MOSFET terminal region>

[0299] Using Fig.95 and Fig.96 , an example of the structure of the MOSFET cell region MCR and the MOSFET terminal region MTR in the semiconductor device of the above-described Embodiments 1 to 3 will be described.

[0300] As shown in Fig.95 the figure, in the MOSFET 102, an n-type epitaxial layer 54 is provided on one main surface of an n-type semiconductor substrate 53 made of silicon carbide. In the upper portion of the epitaxial layer 54, a plurality of p-type well regions 57 are selectively provided. In each well region 57, p-type contact regions 60a are provided so as to penetrate the well region 57.

[0301] Moreover, n-type source regions 58 are provided in contact with both side surfaces of the upper portion of the well region 57 and the contact regions 60a. In addition, the source regions 58 are provided such that the thickness of the source regions 58 is thinner than the thickness of the well region 57, and the thickness of the contact regions 60a is made the same as the thickness of the well region 57 or the contact regions 60a are made slightly deeper.

[0302] A gate insulating film 61 is selectively formed on the epitaxial layer 54, and a gate electrode 63 is formed on the gate insulating film 61. That is, the gate insulating film 61 is provided between adjacent source regions 58, from a part of the upper portion of the source region 58, from above the well region 57, and from above the epitaxial layer 54 to a part of the upper portion of the source region 58 of the adjacent well region 57, and the gate electrode 63 is provided so as to cover the gate insulating film 61.

[0303] Furthermore, an interlayer insulating film 64 is formed so as to cover the gate insulating film 61 and the gate electrode 63, and a source electrode 65 is formed so as to cover the interlayer insulating film 64. In the interlayer insulating film 64, a contact hole SC is provided in a region other than the region covering the gate electrode 63, which penetrates the interlayer insulating film 64 in the thickness direction and reaches a part of the source region 58 and the entire surface of the contact region 60a. Furthermore, the source electrode 65 is filled in the contact hole SC, and the source electrode 65 is connected to the source region 58 and the contact region 60a.

[0304] In this way, a plurality of MOSFETs composed of the source region 58 and the like are arranged in a direction horizontal to the main surface of the semiconductor substrate 53, and are connected in parallel to form an element group. The region where the element group is provided is referred to as the element region (active region) ER, and a terminal region TR is provided at the periphery of the element region ER to realize the withstand voltage of the MOSFET 102. In addition, the element region ER corresponds to the MOSFET cell region MCR, and the terminal region TR corresponds to the MOSFET terminal region MTR.

[0305] A p-type contact region 60b is provided in the upper layer portion of the epitaxial layer 54 in the termination region TR so as to define the outer edge of the element region ER. The contact region 60b is provided to have the same thickness as the contact region 60a but is wider than the contact region 60a.

[0306] Furthermore, outside the contact region 60 b , a p-type resurf region 69 is provided to have the same thickness as that of the contact region 60 b .

[0307] In the epitaxial layer 54, a plurality of n-type pillar layers 55a and p-type pillar layers 56a are alternately arranged in the element region ER so that the respective numbers are equal, and a plurality of n-type pillar layers 55b and p-type pillar layers 56b are alternately arranged in the terminal region TR in the formation region of the contact region 60b and the resurf region 69. In addition, a plurality of n-type pillar layers 55a and p-type pillar layers 56a are alternately arranged in the terminal region TR outside the arrangement region of the n-type pillar layers 55b and the p-type pillar layers 56b.

[0308] All pillar layers are provided to extend from the outermost surface of the epitaxial layer 54 toward the semiconductor substrate 53 in the depth direction of the epitaxial layer 54 , and the deepest part thereof is set to be shallower than the thickness of the epitaxial layer 54 .

[0309] In addition, the n-type pillar layer 55a and the p-type pillar layer 56a have the same width, and the total value of the two is set as the pillar spacing W1. In addition, the n-type pillar layer 55b and the p-type pillar layer 56b have the same width, but each width is set to be wider than the width of the n-type pillar layer 55a and the p-type pillar layer 56a, and the pillar spacing W2, which is the total value of the width of the n-type pillar layer 55b and the p-type pillar layer 56b, is larger than the pillar spacing W1.

[0310] Furthermore, in the termination region TR, a field insulating film 81 is provided on the epitaxial layer 54 , and an interlayer insulating film 64 is provided on the field insulating film 81 .

[0311] The source electrode 65 is provided on the stacked film of the field insulating film 81 and the interlayer insulating film 64 extending from the element region ER to the terminal region TR. In addition, in the stacked film of the field insulating film 81 and the interlayer insulating film 64, a contact hole TC penetrating the stacked film in the thickness direction and reaching the contact region 60b is provided in the region corresponding to the upper portion of the contact region 60b. In addition, the source electrode 65 is filled in the contact hole TC, and the source electrode 65 is connected to the contact region 60b.

[0312] In addition, a stacked film of a field insulating film 81 and an interlayer insulating film 64 is arranged to cover a portion of an upper portion of the MOSFET at the outermost periphery of the element region ER, and a passivation film 87 is arranged in a manner to cover a portion of an upper portion of the source electrode 65 and an upper portion of the stacked film of the field insulating film 81 and the interlayer insulating film 64.

[0313] In addition, a drain electrode 86 is provided on the other main surface (back surface) of the semiconductor substrate 53 on the side opposite to the side on which the source electrode 65 is provided.

[0314] Next, use Fig.96 The manufacturing process of the impurity layer after the pillar layer is formed is described below. Fig.96 As shown, after forming the n-type pillar layers 55a and 55b and the p-type pillar layers 56a and 56b, that is, after using Figure 13 to Figure 17 After the manufacturing process described, impurity ions are implanted using a resist mask (not shown) patterned by photolithography to selectively form a well region 57, a source region 58, a resurf region 59, and contact regions 60a and 60b in the upper portion of the epitaxial layer 54.

[0315] Here, p-type impurities are introduced into the well region 57, the resurf region 59, and the contact regions 60a and 60b, and n-type impurities are introduced into the source region 58. The contact regions 60a and 60b can be ion-implanted using the same resist mask, and the impurity concentration can be 1×10 18 ~1×10 21cm -3 range.

[0316] The impurity concentration of the well region 57 and the resurf region 59 can be 1×10 15 ~1×10 19 cm -3 The impurity concentration of the source region 58 can be in a range exceeding the impurity concentration of the well region 57, for example, 1×10 18 ~1×10 21 cm -3 In addition, the depth of the source region 58 does not exceed the depth of the well region 57.

[0317] The contact regions 60a and 60b can be ion implanted using the same resist mask, and the impurity concentration can be 1×10 18 ~1×10 21 cm -3 In addition, ion implantation can be performed at substrate temperatures above 200°C.

[0318] The contact regions 60 a and 60 b are regions provided to achieve good metal contact with the well region 57 and the resurf region 59 , and the semiconductor device can operate even without providing the contact regions 60 a and 60 b .

[0319] Although not shown in the figure, after the impurities are introduced, heat treatment is performed in an inert gas such as argon or nitrogen or in a vacuum, for example, at a temperature of 1500 to 2200° C. for 0.5 to 60 minutes. Thus, the implanted impurities are electrically activated. Thereafter, an oxide film is formed by sacrificial oxidation of the epitaxial layer 54, and then the oxide film is removed by using hydrofluoric acid to remove the surface altered layer of the epitaxial layer 54 and obtain a clean surface.

[0320] like Fig.96 As shown, in the element region ER, the well region 57 is formed to cover the upper layer of the p-type pillar layer 56a, and the p-type pillar layer 56a is connected to the source electrode 55 ( Fig.95) Electrically connected. Additionally, in the terminal region TR, the contact region 60b and the surface electric field reducing region 59 are formed so as to cover the upper portions of a plurality of p-type pillar layers 56b respectively across the upper portions, and the plurality of p-type pillar layers 56b covered by the contact region 60b become at the same potential via the contact region 60b, and the plurality of p-type pillar layers 56b covered by the surface electric field reducing region 59 become at the same potential via the surface electric field reducing region 59. Moreover, since the contact region 60b and the surface electric field reducing region 59 are arranged such that their sides are in contact with each other, the plurality of p-type pillar layers 56b covered by the contact region 60b are electrically connected to the source electrode 55 via the contact region 60b ( Fig.95 ) Electrically connected.

[0321] <An example of the SBD region and the SBD terminal region>

[0322] Use Figures 97 to 99 , an example of the structure in which an SBD region and an SBD terminal region are provided instead of the MOSFET cell region MCR in the semiconductor device of Embodiments 1 to 3 described above will be described.

[0323] Fig.97 FIG. is a plan view showing the surface structure of the semiconductor substrate 44 of the semiconductor device 103, Fig.98 is a cross-sectional view taken along the direction of the arrow of line A1 - A2 in Fig.97 , and Fig.99 is a cross-sectional view taken along the direction of the arrow of line B1 - B2 in Fig.97 .

[0324] As Fig.98 and Fig.99 show, the semiconductor device 103 is formed using a semiconductor substrate 44 made of n-type SiC having an off-angle. On the semiconductor substrate 44, an n-type drift layer 41 is formed by epitaxial growth.

[0325] In the drift layer 41, a plurality of p-type pillar regions 42 are formed. As shown in Fig.97 , each of the p-type pillar regions 42 has a striped shape in plan view. Since a plurality of p-type pillar regions 42 are provided in the drift layer 41, the drift layer 41 sandwiched by the p-type pillar regions 42 becomes an n-type pillar region.

[0326] As Fig.98 and Fig.99 show, on the drift layer 41 including the p-type pillar region 42, a surface electrode 45 serving as the anode electrode of the SBD is formed. Additionally, on the lower surface of the semiconductor substrate 44, a back electrode 46 serving as the cathode electrode of the SBD is formed. The surface electrode 45 is Schottky-connected to the drift layer 41 and the p-type pillar region 42, and the back electrode 46 is ohmically connected to the semiconductor substrate 44.

[0327] In the upper layer of the drift layer 41 including the p-type pillar region 42, a plurality of frame-shaped withstand voltage retaining structures 43 as p-type semiconductor regions are concentrically formed so as to surround the surface electrode 45. The region surrounded by the withstand voltage retaining structures 43 becomes the active region of the semiconductor device 103, and the region where the withstand voltage retaining structures 43 are formed and the outside thereof become the terminal region of the semiconductor device 103. In addition, there is a case where the withstand voltage retaining structure 43 is included and the outer side than the active region is set as the terminal region.

[0328] like Figure 7 As shown, each of the withstand voltage maintaining structures 43 includes a side extending parallel to the p-type pillar region 42 and a side orthogonal to the p-type pillar region 42 when viewed from above. The chip shape of the semiconductor device 103 when viewed from above is rectangular. Therefore, in the vicinity of the side parallel to the extension direction of the p-type pillar region 42 of the semiconductor device 103, each withstand voltage maintaining structure 43 extends parallel to the p-type pillar region 42, and in the vicinity of the side perpendicular to the extension direction of the p-type pillar region 42, each withstand voltage maintaining structure 43 extends in a manner orthogonal to the p-type pillar region 42.

[0329] At least one of the plurality of pressure-resistant retaining structures 43 is formed so as to overlap with a portion of the surface electrode 45 in a plan view. More specifically, Fig.98 as well as Fig.99 As shown, the innermost pressure-resistant holding structure 43 is formed so as to overlap with the end of the surface electrode 45 .

[0330] Although the present disclosure has been described in detail, the above description is in all aspects illustrative and the present disclosure is not limited thereto. Numerous modifications not shown in the examples are conceivable without departing from the scope of the present disclosure.

[0331] In addition, the present disclosure can freely combine the various embodiments or appropriately modify or omit the various embodiments within the scope of the disclosure.

Claims

1. A semiconductor device comprising: A semiconductor substrate of the first conductivity type; a column portion including a plurality of first columns of a first conductivity type and a plurality of second columns of a second conductivity type provided on the semiconductor substrate in a manner protruding in a thickness direction of the semiconductor substrate; A column peripheral portion of the first conductivity type or the second conductivity type, disposed around the column portion; and The semiconductor element is configured such that the column portion is set as an active region. The plurality of first pillars and the plurality of second pillars have a striped shape in a plan view and are alternately arranged in parallel in a pillar width direction orthogonal to the respective longitudinal directions. The column peripheral portion is provided so as to have a width to surround the entire column portion protruding from the semiconductor substrate and to expose the upper surfaces of the plurality of first columns and the plurality of second columns. Regarding the product of the concentration of impurities in the peripheral portion of the column and the width of the peripheral portion of the column, when the conductivity type of the peripheral portion of the column is the second conductivity type, it is set to be larger than the product of the concentration of impurities of the second conductivity type of the second column and the width of the second column; when the conductivity type of the peripheral portion of the column is the first conductivity type, it is set to be larger than the product of the concentration of impurities of the first conductivity type of the first column and the width of the first column.

2. The semiconductor device according to claim 1, wherein The conductivity type of the peripheral portion of the column is the second conductivity type, The concentration of the second conductivity type impurity in the periphery of the pillar is the same as the concentration of the second conductivity type impurity in the second pillar. The width of the pillar peripheral portion is set larger than the width of the second pillar.

3. The semiconductor device according to claim 1, wherein The conductivity type of the peripheral portion of the column is the first conductivity type, The concentration of the first conductivity type impurity in the periphery of the pillar is the same as the concentration of the first conductivity type impurity in the first pillar. The width of the pillar peripheral portion is set larger than the width of the first pillar.

4. A semiconductor device comprising: A semiconductor substrate of the first conductivity type; a column portion including a plurality of first columns of a first conductivity type and a plurality of second columns of a second conductivity type provided on the semiconductor substrate in a manner protruding in a thickness direction of the semiconductor substrate; A column peripheral portion of the first conductivity type or the second conductivity type, disposed around the column portion; as well as The semiconductor element is configured such that the column portion is set as an active region. The plurality of first pillars and the plurality of second pillars have a stripe-like shape in a plan view and are alternately arranged in parallel in a width direction orthogonal to their respective longitudinal directions. The semiconductor substrate has: A semiconductor substrate of a first conductivity type; as well as A semiconductor layer of the first conductivity type is disposed on the semiconductor substrate, The height of the semiconductor layer disposed under the pillar peripheral portion is lower than the height of the semiconductor layer disposed under the pillar portion.

5. The semiconductor device according to any one of claims 1 to 4, wherein: The semiconductor substrate further includes a peripheral portion which is provided on the outer side of the pillar peripheral portion and has the same conductivity type as that of the pillar peripheral portion.

6. The semiconductor device according to claim 1, wherein The semiconductor substrate uses silicon carbide as a semiconductor material.

7. A method for manufacturing a semiconductor device, comprising: (a) forming a first semiconductor layer of a first conductivity type on a semiconductor substrate of a first conductivity type by epitaxial growth; (b) etching the first semiconductor layer to form a plurality of projections protruding in the thickness direction of the semiconductor substrate at intervals, and forming peripheral portions of the plurality of projections; (c) forming a second semiconductor layer of a second conductivity type between the plurality of protrusions, on the upper portions of the plurality of protrusions, and on the upper portion of the peripheral portion by epitaxial growth; (d) a process of removing at least the second semiconductor layer on the upper portions of the plurality of protrusions to expose the upper surfaces of the plurality of protrusions as a plurality of first pillars of the first conductivity type, using the second semiconductor layer between the plurality of protrusions as a plurality of second pillars of the second conductivity type to form a pillar portion, and using the second semiconductor layer on the upper portion of the peripheral portion as a pillar peripheral portion of the second conductivity type surrounding the pillar portion; and (e) forming a semiconductor element using the pillar portion as an active region, The step (b) includes etching so that each of the plurality of convex portions has a stripe-like shape in plan view and is arranged in parallel with a gap in a width direction orthogonal to the respective longitudinal directions. The plurality of first pillars and the plurality of second pillars have a stripe-like shape in plan view, and are alternately arranged in parallel in the width direction orthogonal to the respective longitudinal directions.

8. The method for manufacturing a semiconductor device according to claim 7, wherein: In the step (c), the second semiconductor layer is epitaxially grown in a manner so as to surround the entire column portion protruding from the semiconductor substrate with a width and make the product of the concentration of the second conductive type impurity and the width greater than the product of the concentration of the second conductive type impurity of the second column and the width of the second column.

9. The method for manufacturing a semiconductor device according to claim 8, wherein: In the step (c), the second semiconductor layer is epitaxially grown in such a manner that the concentration of the second conductive type impurities in the peripheral portion of the pillar is the same as the concentration of the second conductive type impurities in the second pillar and the width of the peripheral portion of the pillar is larger than the width of the second pillar.

10. The method for manufacturing a semiconductor device according to claim 7, wherein: In the step (c), the second semiconductor layer is epitaxially grown so as to have a width surrounding the entire columnar portion protruding on the semiconductor substrate. Between step (c) and step (d), there is also a step of performing ion implantation on the second semiconductor layer at the peripheral portion of the plurality of protrusions, so that the product of the concentration of the second conductive type impurities at the peripheral portion of the column and the width is greater than the product of the concentration of the second conductive type impurities of the second column and the width of the second column.

11. The method for manufacturing a semiconductor device according to claim 7, wherein: In the step (d), the second semiconductor layer on the upper portions of the plurality of protrusions is removed by full-surface etching, and a peripheral portion of the same conductivity type as the pillar peripheral portion and provided outside the pillar peripheral portion is removed.

12. The method for manufacturing a semiconductor device according to claim 7, wherein: In the step (b), when forming the plurality of protrusions and the peripheral portions of the plurality of protrusions, the first semiconductor layer is etched in such a manner that a height of the first semiconductor layer below the pillar peripheral portion of the first semiconductor layer is lower than a height of the first semiconductor layer below the pillar portion.

13. The method for manufacturing a semiconductor device according to claim 7, wherein: The semiconductor substrate uses silicon carbide as a semiconductor material.

14. A method for manufacturing a semiconductor device, comprising: (a) forming a first semiconductor layer of a second conductivity type on a semiconductor substrate of a first conductivity type by epitaxial growth; (b) etching the first semiconductor layer to form a plurality of projections protruding in the thickness direction of the semiconductor substrate at intervals, and exposing the semiconductor substrate around the plurality of projections as a peripheral portion; (c) forming a second semiconductor layer of the first conductivity type between the plurality of protrusions, on the upper portions of the plurality of protrusions, and on the upper portions of the peripheral portions of the plurality of protrusions by epitaxial growth; (d) a process of removing at least the first semiconductor layer on the upper portions of the plurality of protrusions to expose the upper surfaces of the plurality of protrusions as a plurality of second pillars of the second conductivity type, using the first semiconductor layer between the plurality of protrusions as a plurality of first pillars of the first conductivity type to form a pillar portion, and using the first semiconductor layer on the upper portion of the peripheral portion as a pillar peripheral portion of the first conductivity type surrounding the pillar portion; and (e) forming a semiconductor element using the pillar portion as an active region, The step (b) includes etching so that each of the plurality of convex portions has a stripe-like shape in plan view and is arranged in parallel with a gap in a width direction orthogonal to the respective longitudinal directions. The plurality of first pillars and the plurality of second pillars have a stripe-like shape in plan view, and are alternately arranged in parallel in the width direction orthogonal to the respective longitudinal directions.

15. The method for manufacturing a semiconductor device according to claim 14, wherein: The semiconductor base has a semiconductor layer of the first conductivity type formed by epitaxial growth on a semiconductor substrate of the first conductivity type. The first semiconductor layer is formed on the semiconductor layer.

16. The method for manufacturing a semiconductor device according to claim 14, wherein: In the step (c), the first semiconductor layer is epitaxially grown in a manner so as to have a width that surrounds the entire column portion protruding from the semiconductor substrate and makes the product of the concentration of the first conductive type impurity and the width greater than the product of the concentration of the first conductive type impurity of the first column and the width of the first column.

17. The method for manufacturing a semiconductor device according to claim 14, wherein: In the step (c), the second semiconductor layer is epitaxially grown so as to have a width surrounding the entire columnar portion protruding from the semiconductor substrate. Between step (c) and step (d), there is also a step of performing ion implantation on the first semiconductor layer at the peripheral portion of the plurality of protrusions, so that the product of the concentration of the first conductive type impurity at the peripheral portion of the column and the width is greater than the product of the concentration of the first conductive type impurity of the first column and the width of the first column.

18. The method for manufacturing a semiconductor device according to claim 14, wherein: In the step (d), the first semiconductor layer on the upper portions of the plurality of protrusions is removed by full-surface etching, and a peripheral portion of the same conductivity type as the pillar peripheral portion and provided outside the pillar peripheral portion is removed.

19. The method for manufacturing a semiconductor device according to claim 14, wherein: In the step (b), when forming the plurality of protrusions and the peripheral portions of the plurality of protrusions, the semiconductor substrate is etched in such a manner that the height of the semiconductor substrate below the pillar peripheral portion of the semiconductor substrate is lower than the height of the semiconductor substrate below the pillar portion.

20. The method for manufacturing a semiconductor device according to claim 14, wherein: The semiconductor substrate uses silicon carbide as a semiconductor material.

Citation Information

Patent Citations

  • Semiconductor device

    WO2019160086A1

  • Semiconductor device

    CN101399286A

  • Semiconductor device and method for manufacturing the same

    CN102074581A

  • Semiconductor device

    CN111133586A

  • Semiconductor device and method of manufacturing the same

    CN1790714A