Silicon carbide semiconductor device and power conversion device
By adopting a multi-well region structure and surface density difference control method in the silicon carbide semiconductor device, the problems of increased manufacturing costs and deterioration of characteristics caused by stacking defect expansion are solved, and the efficient suppression effect of the silicon carbide semiconductor device is achieved.
Patent Information
- Application Number
- CN202080104747.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-08-11
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2040-08-11
AI Technical Summary
When the prior art suppresses the expansion of stacking defects in silicon carbide semiconductor devices, it leads to a problem of significantly increasing manufacturing costs or deteriorating characteristics.
The silicon carbide semiconductor device structure is adopted with an active region and an outer peripheral region, wherein a plurality of first well regions and second well regions are arranged in the active region, and the current density and holes are controlled through the surface density difference between the sparse region and the standard region. Concentration distribution to inhibit the expansion of stacking defects.
The significant increase in manufacturing costs is effectively avoided, and the characteristics of the silicon carbide semiconductor device are fully suppressed, thereby improving the reliability of the device.
Smart Images

Figure CN116137935B_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to a silicon carbide semiconductor device and a power conversion device, and in particular to a silicon carbide semiconductor device having an ohmic electrode and a power conversion device having the silicon carbide semiconductor device. Background Art
[0002] In recent years, semiconductor devices using silicon carbide (SiC) have been widely used as power semiconductor devices. Typically, a SiC semiconductor device has a single crystal SiC substrate and a SiC layer formed thereon as an epitaxial layer. In the case where the device includes a structure of a pn diode, when a forward current, i.e., a bipolar current, continues to flow through the pn diode, stacking defects sometimes occur in the crystallization of the SiC layer. It is believed that the reason is that due to the recombination energy when minority carriers injected through the pn junction of the pn diode recombine with majority carriers, stacking defects as surface defects expand in the SiC layer starting from basal plane dislocations (BPD: Basal Plane Dislocation) and the like present in the SiC substrate. The expansion of stacking defects causes characteristic degradation such as an increase in forward voltage (on-state voltage) and a decrease in withstand voltage due to obstruction of the flow of current.
[0003] The above-mentioned characteristic degradation may occur not only when the semiconductor device is a simple pn diode, but also in the case of a vertical MOSFET (Metal Oxide Semiconductor Field Effect Transistor). The reason is that the vertical MOSFET has a body diode (parasitic diode) built in between the source and the drain, which is a pn diode. Specifically, when the body diode functions as a freewheeling diode of the MOSFET in a power electronic system, the above-mentioned characteristic degradation occurs due to the forward current flowing through the body diode.
[0004] For example, as disclosed in International Publication No. 2018 / 155553 (Patent Document 1), SiC-MOSFET and SiC-Schottky Barrier Diode (SBD: Schottky Barrier Diode) use the epitaxial layer grown on the SiC substrate as a withstand voltage holding layer. There are usually a large number of defects in the SiC substrate, and among these defects, there are also defects that become the starting point for the expansion of stacking defects in the epitaxial layer. Therefore, in order to make the epitaxial layer grown on the SiC substrate high quality, attempts to make various defects harmless have been made for many years.
[0005] As a method of suppressing the above-mentioned characteristic degradation caused by the forward current flowing through the body diode, the following three methods have been mainly studied. As the first method, when growing an epitaxial layer on a SiC substrate, a method of preventing stacking defects from extending from a basal plane dislocation by converting the basal plane dislocation from the SiC substrate to the epitaxial layer from a basal plane dislocation to a through-edge dislocation in the initial stage of growth is studied (for example, refer to non-patent document 1). As a second method, a method of preventing stacking defects from occurring from a basal plane dislocation of a SiC substrate by promoting the recombination of holes and electrons using a buffer layer with a high impurity concentration formed on the SiC substrate (for example, refer to non-patent document 2). As a third method, a method of preventing the recombination of holes and electrons near the basal plane dislocation of the SiC substrate by introducing a recombination center into a region where a parasitic pn diode is provided is studied (for example, refer to patent document 2).
[0006] Prior art literature
[0007] Patent Literature
[0008] Patent Document 1: International Publication No. 2018 / 155553
[0009] Patent Document 2: International Publication No. 2015 / 189929
[0010] Non-patent literature
[0011] Non-patent document 1: "Influence of growth conditions on basal plane dislocation in 4H-SiC epitaxial layer", Journal of Crystal Growth271(2004)1-7
[0012] Non-patent document 2: "Short minority carrier lifetimes in highly nitrogen-doped 4H-SiC epilayers for suppression of the stacking fault formation in PiNdiodes", JOURNAL OF APPLIED PHYSICS Vol. 120, 115101 (2016) Summary of the invention
[0013] The inventors have noticed that all the above-mentioned previous methods have not fully studied the differences in bipolar current density in semiconductor chips that are SiC semiconductor devices. Specifically, the inventors have found that the bipolar current density locally increases in a local area near the boundary between the active area and the peripheral area in the chip. Moreover, by further studying the influence of the distribution of such bipolar current density, the inventors have come to the conclusion that by simply applying the above-mentioned previous methods, although the expansion of stacking defects can be suppressed in most areas of the chip, there is a possibility that the expansion of stacking defects cannot be fully suppressed in the above-mentioned local areas. The following is a more specific explanation of this problem.
[0014] In order to effectively apply the first method, it is necessary to form a transition point from a basal plane dislocation to a through edge dislocation as deep as possible from the surface of the epitaxial layer. Therefore, in order to obtain a sufficient effect in the above-mentioned local area by simply applying the first method, the epitaxial layer needs to be further thickened. As a result, the manufacturing cost of the chip increases significantly.
[0015] In order to effectively apply the second method, it is necessary to form a buffer layer thick enough to fully prevent the penetration of minority carriers. Therefore, in order to obtain sufficient effects in the local area by simply applying the second method, the buffer layer needs to be thicker. As a result, the manufacturing cost of the chip increases significantly.
[0016] In order to effectively apply the third method, it is necessary to introduce recombination centers in the area where the pn diode is provided to a degree that can sufficiently reduce the injected holes. Therefore, in order to obtain sufficient effects in the above-mentioned local area by simply applying the third method, it is necessary to introduce more recombination centers in the peripheral area and its vicinity or to introduce more recombination centers in the entire chip. In the former case, the manufacturing cost of the chip increases significantly due to the increase in the number of processes and the increase in the number of masks in lithography. In the latter case, the recombination centers are usually defects such as point defects, so due to the presence of excess recombination centers, the withstand voltage is reduced, the on-state voltage is increased, and the leakage current is increased.
[0017] As described above, when attempting to sufficiently suppress the expansion of stacking defects even in the above-mentioned local regions by simply applying the conventional method, a new factor that significantly increases the manufacturing cost or causes degradation of characteristics arises.
[0018] The present disclosure has been made to solve the above problems, and an object of the present disclosure is to provide a silicon carbide semiconductor device that can avoid a significant increase in manufacturing cost and suppress degradation of characteristics.
[0019] A silicon carbide semiconductor device according to one embodiment of the present disclosure comprises a silicon carbide substrate, a silicon carbide layer, a gate insulating film, a gate electrode, a gate pad, a source electrode, and a plurality of ohmic electrodes. The silicon carbide substrate has a first conductivity type. The silicon carbide layer has a first surface facing the silicon carbide substrate and a second surface opposite to the first surface in the thickness direction, and has an active region and a peripheral region arranged on the periphery of the active region in the in-plane direction perpendicular to the thickness direction. The silicon carbide layer includes a drift layer, a plurality of first well regions, a plurality of source regions, and a second well region. The drift layer is arranged on the silicon carbide substrate and has a first conductivity type. A plurality of first well regions are arranged on the drift layer in the active region and have a second conductivity type different from the first conductivity type, and are separated from each other when viewed in at least one cross section. A plurality of source regions are arranged on a plurality of first well regions and have a first conductivity type. A second well region is arranged on the drift layer in the peripheral region and has a second conductivity type. The gate insulating film faces a plurality of first well regions. The gate electrode has a portion disposed above the active region and facing the plurality of first well regions through a gate insulating film, and a portion disposed above the peripheral region and insulated from the second well region. The gate pad is disposed above the second well region, insulated from the second well region, and connected to the gate electrode. The source electrode is disposed above the second surface of the silicon carbide layer. A plurality of ohmic electrodes are disposed on the second surface of the silicon carbide layer, connected to the source electrode, electrically ohmically connected to the plurality of first well regions, and having a plurality of surface regions in ohmic contact with a portion of the second conductivity type in the second surface of the silicon carbide layer. The active region includes a standard region portion and a sparse region portion between the standard region portion and the peripheral region. The sparse region portion has a lower surface density of the plurality of surface regions when viewed from above than the standard region portion.
[0020] The silicon carbide semiconductor device involved in other embodiments of the present disclosure has a silicon carbide substrate, a silicon carbide layer, a gate insulating film, a gate electrode, a gate pad, a source electrode and a plurality of ohmic electrodes. The silicon carbide substrate has a first conductivity type. The silicon carbide layer has a first surface facing the silicon carbide substrate and a second surface opposite to the first surface in the thickness direction, and has an active region and a peripheral region arranged on the periphery of the active region in the in-plane direction perpendicular to the thickness direction. The silicon carbide layer includes a drift layer, a plurality of first well regions, a plurality of source regions and a second well region. The drift layer is arranged on the silicon carbide substrate and has a first conductivity type. A plurality of first well regions are arranged on the drift layer in the active region and have a second conductivity type different from the first conductivity type, and are separated from each other when viewed in at least one cross section. A plurality of source regions are arranged on a plurality of first well regions and have a first conductivity type. The second well region is arranged on the drift layer in the peripheral region and has a second conductivity type. The gate insulating film faces the plurality of first well regions. The gate electrode has a portion disposed above the active region and facing the plurality of first well regions through a gate insulating film, and a portion disposed above the peripheral region and insulated from the second well region. The gate pad is disposed above the second well region, insulated from the second well region, and connected to the gate electrode. The source electrode is disposed above the second surface of the silicon carbide layer. A plurality of ohmic electrodes are disposed on the second surface of the silicon carbide layer, connected to the source electrode, electrically ohmically connected to the plurality of first well regions and the second well region, and have a plurality of surface regions that are in ohmic contact with a portion of the second conductivity type in the second surface of the silicon carbide layer. The peripheral region includes an outer region portion and a sparse region portion between the outer region portion and the active region. The sparse region portion has a lower surface density of the plurality of surface regions when viewed from above than the active region.
[0021] According to the present disclosure, it is possible to avoid a significant increase in manufacturing cost and suppress degradation in characteristics of a silicon carbide semiconductor device.
[0022] The objects, features, aspects and advantages of the present disclosure will become more apparent from the following detailed description and accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] Figure 1 This is a top view schematically showing the structure of the silicon carbide semiconductor device according to the first embodiment.
[0024] Figure 2 is Figure 1 A dotted line portion II in FIG. 1 schematically shows a partial top view of the structure of a silicon carbide layer included in the silicon carbide semiconductor device.
[0025] Figure 3 It is along Figure 2 The line III-III in FIG. 1 schematically shows a partial cross-sectional view of the structure of the silicon carbide semiconductor device.
[0026] Figure 4 is Figure 2 A partial top view schematically shows the first well region, the second well region, a portion having the second conductivity type in the upper surface of the silicon carbide layer, and multiple surface regions of the ohmic contacts of the multiple ohmic electrodes.
[0027] Figure 5 Yes Figure 4 The field of view is enlarged in the longitudinal direction and shows a partial top view while illustrating the definition of the surface density of the surface area.
[0028] Figure 6 It is shown in Figure 5 A graph of the distribution of the surface density defined in the x-axis direction.
[0029] Figure 7 is a graph showing the current density distribution in the freewheeling operation corresponding to the comparative example ( Figure 8 as well as Fig. 9 ) and hole concentration distribution ( Fig.10 ) is a cross-sectional view of the structure of the pn diode used in the simulation.
[0030] Figure 8 The corresponding comparative example is shown. Figure 7 A distribution diagram of the simulation results of the current density distribution in the pn diode structure shown.
[0031] Fig. 9 The bottom surface of the well region is shown along the bottom surface of the well region, corresponding to the comparative example. Figure 7 FIG. 5 is a graph showing simulation results of current density distribution for two types of substrate thicknesses in the pn diode structure shown.
[0032] Fig.10 The depth of the bottom surface of the drift layer is shown in FIG. Figure 7 FIG. 5 is a graph showing simulation results of hole concentration distributions at two substrate thicknesses in the pn diode structure shown.
[0033] Fig.11 is a graph showing the hole concentration distribution in the freewheeling operation ( Fig.12 ) is a cross-sectional view of the structure of a pn diode used in a simulation in which the surface density distribution is varied.
[0034] Fig.12 is shown at the depth of the bottom surface of the drift layer Fig.11 FIG. 4 is a graph showing simulation results of hole concentration distribution using different surface density distributions in the structure of the pn diode shown.
[0035] Fig.13 It is shown Figure 4 A partial top view of a first variation of the structure shown.
[0036] Fig.14 It is shown Figure 4 A partial top view of a second variation of the structure shown.
[0037] Fig.15 It is shown Figure 4 A partial top view of a third variation of the structure shown.
[0038] Fig.16 It is shown Figure 4 Partial top view of a 4th variation of the structure shown.
[0039] Fig.17 So with Figure 4 The same viewing angle schematically shows a partial top view of the first well region, the second well region, and a portion of the upper surface of the silicon carbide layer having the second conductivity type and multiple surface regions of the ohmic contacts of the multiple ohmic electrodes in the silicon carbide semiconductor device according to the second embodiment.
[0040] Fig.18 This is a block diagram showing the configuration of a power conversion system to which the power conversion device according to the third embodiment is applied.
[0041] (Explanation of symbols)
[0042] 10: SiC substrate (silicon carbide substrate); 20: drift layer; 30: first well region; 31: second well region; 35: contact region; 40: source region; 50: gate insulating film; 60: gate electrode; 70: ohmic electrode; 71: surface region; 80: source electrode; 81: gate pad; 82: gate wiring; 84: drain electrode; 100: active region; 101, 201: sparse region portion; 102: standard region portion; 200: peripheral region; 202: outer region portion; 711~715, 721: MOSFET (silicon carbide semiconductor device); 910: power supply; 920: power conversion device; 921: main conversion circuit; 922: drive circuit; 923: control circuit; 930: load; SL: SiC layer (silicon carbide layer). DETAILED DESCRIPTION
[0043] Hereinafter, embodiments of the present disclosure are described with reference to the accompanying drawings. In addition, in each embodiment described in this specification, an n-channel MOSFET, that is, a MOSFET in which the first conductivity type is n-type and the second conductivity type is p-type, is described as an example of a silicon carbide semiconductor device. The description of the high and low potentials in this description is a description of the case where the first conductivity type is n-type and the second conductivity type is p-type. In the case where the first conductivity type is p-type and the second conductivity type is n-type, the high and low potentials are reversed.
[0044] <1. Implementation Method 1>
[0045] <1-1. Basic structure>
[0046] Figure 1 This is a top view schematically showing the structure of MOSFET 711 (silicon carbide semiconductor device) according to Embodiment 1. MOSFET 711 has a source electrode 80, a gate pad 81, and a gate wiring 82 on its upper surface. The gate wiring 82 extends from the gate pad 81, and the source electrode 80 is arranged away from them.
[0047] Figure 2 is Figure 1 The dotted line portion II in FIG. 1 schematically shows the silicon carbide layer SL ( Figure 3 )'s partial top view. Figure 3 It is along Figure 2 The line III-III in FIG. 1 schematically shows a partial cross-sectional view of the structure of MOSFET 711. MOSFET 711 has a silicon carbide substrate 10 (hereinafter also referred to as SiC substrate 10), a silicon carbide layer SL (hereinafter also referred to as SiC layer SL), a gate insulating film 50, a gate electrode 60, a gate pad 81, a source electrode 80, a plurality of ohmic electrodes 70, and a drain electrode 84.
[0048] SiC substrate 10 has n-type (first conductivity type). SiC layer SL has a lower surface (first surface) facing SiC substrate 10 and an upper surface (second surface opposite to the first surface in the thickness direction). The lower surface of SiC layer SL may be in contact with SiC substrate 10 .
[0049] The SiC layer SL has an active region 100 and a peripheral region 200 arranged on the periphery of the active region 100 in the in-plane direction perpendicular to the thickness direction. Typically, in the in-plane direction (in other words, when viewed from above), the peripheral region 200 surrounds the periphery of the active region 100. The active region 100 directly assumes the main function of the MOSFET 711, that is, the function of switching the current. Therefore, the MOS gate structure of the MOSFET 711 is not arranged in the peripheral region 200 but in the active region 100. The SiC layer SL includes a drift layer 20, a plurality of first well regions 30, a plurality of source regions 40, a second well region 31, a contact region 35, and a JTE (Junction Termination Extension) region 37.
[0050] The drift layer 20 has an n-type. The impurity concentration of the drift layer 20 is lower than the impurity concentration of the SiC substrate 10 . The drift layer 20 is disposed on the SiC substrate 10 .
[0051] The first well region 30 has a p-type (a second conductivity type different from the first conductivity type). The first well region 30 is disposed on the drift layer 20 in the active region 100 and typically forms a portion of the upper surface of the SiC layer SL. The first well region 30 is Figure 3 The drift layer 20 has a JFET region 21 between two adjacent first well regions 30. The impurity concentration of the JFET region 21 as a part of the drift layer 20 may be the same as or different from the impurity concentration of other parts of the drift layer 20.
[0052] The source region 40 has an n-type. The source region 40 is disposed on the plurality of first well regions 30. The source region 40 is separated from the drift layer 20 by the first well regions 30. Figure 3 In the structure shown, on the upper surface of the silicon carbide layer SL, the outer edge of each source region 40 is arranged on the inner side than the outer edge of each first well region 30 , thereby on the upper surface of the silicon carbide layer SL, each source region 40 is separated from the drift layer 20 by the first well region 30 .
[0053] The contact region 35 has a p-type and has an impurity concentration higher than that of the first well region 30. The contact region 35 has an upper end that reaches the upper surface of the SiC layer SL and a lower end that reaches the first well region 30. Figure 2 In the example shown, the contact region 35 is surrounded by the source region 40 in a plan view.
[0054] The second well region 31 has a p-type. Typically, the second well region 31 partially forms the upper surface of the SiC layer SL. Typically, the second well region 31 is larger than each of the first well regions 30. The second well region 31 is arranged on the drift layer 20 in the peripheral region 200. Specifically, the second well region 31 is arranged outside the outermost first well region 30 via the JFET region 21. In the first embodiment, the second well region 31 is separated from the first well region 30, but as a modified example, it may be connected to the first well region 30 (for example, see Fig.13 wait).
[0055] A plurality of ohmic electrodes 70 are provided on the upper surface of the SiC layer SL. The ohmic electrodes 70 in the active region 100 are in ohmic contact with the source region 40 and the contact region 35 reaching the first well region 30. The ohmic electrodes 70 in the peripheral region 200 are in ohmic contact with the contact region 35 reaching the second well region 31. Through the contact with the contact region 35, the plurality of ohmic electrodes 70 are electrically ohmically connected to the plurality of first well regions 30 and the second well region 31.
[0056] As described above, the plurality of ohmic electrodes 70 have a plurality of portions in ohmic contact with the n-type portion of the upper surface of the SiC layer (specifically, the source region 40) and a plurality of portions in ohmic contact with the p-type portion of the upper surface of the SiC layer (specifically, the contact region 35), and in particular, the latter will be referred to as a plurality of surface regions below. According to this definition, each of the plurality of surface regions is a boundary surface between the p-type portion of the upper surface of the SiC layer and the ohmic electrode 70.
[0057] The material of the ohmic electrode 70 is selected so that electrons and holes can be easily exchanged with the source region 40 and the contact region 35 both of which have relatively high impurity concentrations. Typically, the material is a silicide alloy, such as a nickel silicide alloy.
[0058] The gate insulating film 50 has at least a portion facing the plurality of first well regions 30. Figure 3 In the example shown, the gate insulating film 50 has a portion straddling the boundary between the source region 40 and the first well region 30 and a portion covering the first well region 30 between the boundary and the JFET region 21. The gate insulating film 50 may further have a portion facing the second well region 31.
[0059] The gate electrode 60 includes a portion disposed above the active region 100 and facing the plurality of first well regions 30 via the gate insulating film 50, and a portion disposed above the peripheral region 200 and insulated from the second well region 31. The surface layer of the first well region 30 that faces the gate electrode 60 via the gate insulating film 50 functions as a channel region. The gate electrode 60 extends from the inside of the active region 100 to the inside of the peripheral region 200.
[0060] The interlayer insulating film 55 covers the gate electrode 60 . In the peripheral region 200 , the gate electrode 60 is insulated from the second well region 31 by the gate insulating film 50 and the field insulating film 51 , and is connected to the gate wiring 82 via a contact hole 95 of the interlayer insulating film 55 .
[0061] The field insulating film 51 is provided on the peripheral region 200. The thickness of the field insulating film 51 is greater than that of the gate insulating film 50. Figure 3 In the example shown, the field insulating film 51 is disposed only in the peripheral region 200 , but as a modification, it may have a portion extending from the peripheral region 200 into the active region 100 . In this modification, the field insulating film 51 may extend over a portion of the first well region 30 across the JFET region 21 .
[0062] Gate pad 81 ( Figure 1 ) and gate wiring 82 ( Figure 1 ) is arranged in the second well region 31 ( Figure 3) and is insulated from the second well region 31. The gate pad 81 is connected to the gate electrode 60 via the gate wiring 82. The gate pad 81 and the gate wiring 82 are insulated from the second well region 31. Figure 3 In the structure shown, the insulation is ensured by the field insulating film 51. In addition, as a modification, the insulation may be ensured by another insulating film instead of the field insulating film 51 or together with the field insulating film 51, and the other insulating film may also be the gate insulating film 50, for example.
[0063] The source electrode 80 is provided above the upper surface of the SiC layer SL and is connected to each of the plurality of ohmic electrodes 70. Thus, the source electrode 80 is electrically ohmically connected to the source region 40 via the ohmic electrode 70, and is electrically ohmically connected to the first well region 30 and the second well region 31 via the ohmic electrode 70 and the contact region 35. In addition, the source electrode 80 is insulated from the gate electrode 60 by the interlayer insulating film 55.
[0064] exist Figure 3 In the illustrated structural example, a contact hole 90 penetrating the gate insulating film 50 and the interlayer insulating film 55 is formed on the ohmic electrode 70 connected to the first well region 30. The source electrode 80 reaches the ohmic electrode 70 connected to the first well region 30 through the contact hole 90. With this structure, the first well region 30 and the source electrode 80 are electrically ohmically connected.
[0065] Furthermore, in Figure 3 In the structural example shown, a contact hole 91 penetrating the gate insulating film 50 and the interlayer insulating film 55 is formed on the ohmic electrode 70 connected to the second well region 31. Through the contact hole 91, the source electrode 80 reaches the ohmic electrode 70 connected to the second well region 31. Through this structure, the second well region 31 is electrically ohmically connected to the source electrode 80. As a modification, other structures in which the second well region 31 is electrically ohmically connected to the source electrode 80 can also be used. For example, in the case where the second well region 31 and the first well region 30 are connected at any position when viewed from above, the source electrode 80 and the second well region 31 can be electrically ohmically connected even without forming the contact hole 91. As a further modification, a structure in which the source electrode 80 and the second well region 31 are capacitively connected can be used (for example, refer to the above-mentioned International Publication No. 2018 / 155553), in which case the source electrode 80 and the second well region 31 do not need to be electrically ohmically connected. In this case, a displacement current during switching flows between the source electrode 80 and the second well region 31 via the capacitive connection.
[0066] The JTE region 37 is arranged on the outer peripheral side of the second well region 31 (the side opposite to the first well region 30 ) on the drift layer 20 . The JTE region 37 has a p-type and its impurity concentration is lower than that of the second well region 31 .
[0067] The drain electrode 84 is provided on the back surface of the SiC substrate 10. The drain electrode 84 is electrically ohmically connected to the SiC substrate 10. To obtain the ohmic connection, a back ohmic electrode (not shown) may be provided between the drain electrode 84 and the SiC substrate 10.
[0068] <1-2. Details of Layout of Impurity Regions in the In-Plane Direction>
[0069] Reference Figure 2 The active region 100 includes a standard region portion 102 and a sparse region portion 101 between the standard region portion 102 and the peripheral region 200. The sparse region portion 101 is disposed adjacent to the standard region portion 102. The peripheral region 200 is disposed outside the standard region portion 102 via the sparse region portion 101. Typically, the sparse region portion 101 surrounds the standard region portion 102 occupying a large proportion of the active region 100, and the peripheral region 200 surrounds the sparse region portion 101.
[0070] For the sake of convenience in description, the active region 100 is divided into a plurality of periodic regions RP. In other words, for the sake of convenience in description, it is considered that the active region 100 is formed by repeatedly arranging the periodic regions RP periodically in a plurality of directions. Figure 2 In the example shown, each periodic region RP has a square shape with a length LP of one side, and a plurality of periodic regions RP are arranged along the side to form the active region 100. The layout of the impurity region in each periodic region RP will be described below.
[0071] In the standard region portion 102 , the unit cells U0 are arranged in each of the periodic regions RP. Therefore, the standard region portion 102 is formed by periodically arranging the unit cells U0. The unit cell U0 includes a first well region 30, a source region 40, and a contact region 35.
[0072] In the sparse region 101, a unit cell U1 or a unit cell U2 is disposed in each of the plurality of periodic regions RP. The unit cells U1 and the unit cells U2 are mixed, so that the sparse region 101 includes at least one unit cell U1 and at least one unit cell U2. The pattern in which the unit cells U1 and the unit cells U2 are mixed is preferably a periodic pattern, for example, Figure 2 In the figure, unit cells U1 and unit cells U2 are alternately arranged in the x direction (horizontal direction in the figure) and the y direction (vertical direction in the figure). Unit cells U1 and unit cells U2 each have a first well region 30 and a source region 40. Furthermore, unit cell U1 has a contact region 35 similarly to unit cell U0. On the other hand, unit cell U2 does not have a contact region 35. The outer edge pattern of the source region 40 of unit cells U0 to U2 can be common.
[0073] The first well region 30 of the unit cell U2 is directly connected to the first well region 30 of the unit cell U1 or is indirectly connected via the first well region 30 of another unit cell U2. Figure 2 In the example shown, the first well region 30 of the unit cell U2 is directly connected to the first well region 30 of the unit cell U1. This connection prevents the first well region 30 of the unit cell U2 from being in a floating state. In order to obtain this connection, the first well region 30 of at least any unit cell of the unit cell U1 and the unit cell U2 may have an extension portion that extends the shape of the first well region 30 of the unit cell U0, especially Figure 2 In the example shown, the first well region 30 of each unit cell U1 and the unit cell U2 has an extension E1 and an extension E2. The extension E1 and the extension E2 are each connected to the first well region 30 of the adjacent periodic region RP, especially Figure 2 In the example shown, the extension portion E1 and the extension portion E2 are connected to each other.
[0074] In the standard region portion 102, the plurality of first well regions 30 have polygonal shapes that are repeatedly arranged in the in-plane direction. Figure 2 In the example shown, the first well regions 30 have a shape of a polygon repeatedly arranged in the x direction and the y direction. Furthermore, if the extension E1 and the extension E2 are ignored, the first well regions 30 also have a shape of a polygon repeatedly arranged in the in-plane direction in the sparse region 101. In other words, if the extension E1 and the extension E2 are ignored, the first well regions 30 have a shape of a polygon repeatedly arranged in the in-plane direction in the entire active region 100 including the standard region 102 and the sparse region 101, especially in Figure 2 In the example shown, the shapes have squares which are repeatedly arranged along the x direction and the y direction.
[0075] On the other hand, the above description of the division of the regions RP using periodicity does not apply to the outer peripheral region 200. In the outer peripheral region 200, the second well region 31 larger than each first well region 30 of the standard region portion 102 is provided. Figure 2 In the illustrated example, a plurality of contact regions 35 are periodically arranged on the second well region 31 .
[0076] Furthermore, in the above description, the case where the first well region 30 of each standard region portion 102 has a square shape has been described in detail. However, a rectangular shape having long sides and short sides or another polygonal shape may be used instead of the square shape.
[0077] Figure 4 is Figure 2The first well region 30, the second well region 31 and a partial top view of the surface region 71 are schematically shown in the field of view, and the source region 40 ( Figure 2 ) is not shown. Here, the surface region 71 is defined as the SiC layer SL ( Figure 3 ) of the upper surface of the p-type portion and the ohmic electrode 70 ( Figure 3 ) The area of ohmic contact. Figure 4 As shown, the surface density of the surface region 71 of the sparse region portion 101 is lower than that of the standard region portion 102 in a plan view.
[0078] Figure 5 Yes Figure 4 The view in FIG. 1 is magnified in the longitudinal direction (y direction), and the definition of the above surface density is described in detail with reference to this figure. In the figure, the y direction is along the boundary line between the active area 100 and the peripheral area 200, and the x direction is the direction perpendicular to it. Figure 5 In the figure, the rectangular region RR has a length LR in the x direction, a width WR in the y direction, and a center position CR. The area of the surface region 71 in the rectangular region RR is divided by the area of the rectangular region RR, and the value obtained is defined as the surface density at the position CR. The length LR is the length LP of the period of the periodic region RP ( Figure 2 ), specifically, more than 2 times, and is sufficiently smaller than the total length (dimension in the x direction) of the SiC layer SL. In addition, the width WR is the length LP of the period of the periodic region RP ( Figure 2 ), specifically, more than 10 times, and a length that is sufficiently smaller than the full width (dimension in the y direction) of the SiC layer SL. In addition, the above description describes the case where the period of the periodic region RP is common in the x-direction and the y-direction, but they may be different from each other.
[0079] As described above, the surface density of the surface region 71 in the plan view of the sparse region 101 is lower than that of the standard region 102. An example of a method of determining whether the distribution of the surface density satisfies this condition will be described below.
[0080] About Figure 1 For a quadrilateral chip as shown, the surface density is calculated in an area that is sufficiently away from the four corners of the chip. In other words, the center position CR ( Figure 5 ) is set to a position sufficiently away from the four corners of the chip. In addition, the length LR ( Figure 5 ) is set to length LP( Figure 2 ) is 10 times the width WR( Figure 5 ) is set to twice the length LP. Figure 6 The relationship between the x coordinate of the center position CR and the surface density S when the y coordinate of the center position CR is made constant is shown. Figure 6In the above, if the vicinity of the boundary between the standard region 102 and the sparse region 101 is ignored, the surface density S in the sparse region 101 is lower than the surface density S in the standard region 102 (specifically, it becomes half), and the above condition is satisfied. In addition, regardless of whether the vicinity of the boundary is ignored as described above, the average surface density S in the sparse region 101 is lower than the average surface density S in the standard region 102 (specifically, it becomes half), and the above condition is satisfied.
[0081] <1-3. Manufacturing method>
[0082] Next, an example of a method for manufacturing MOSFET 711 is described below.
[0083] Reference Figure 3 First, a SiC substrate 10 composed of polytype, n-type and low-resistance silicon carbide having 4H is prepared. The surface orientation of the upper surface of the SiC substrate 10 has a slight deviation angle from the (0001) plane. Next, on the upper surface of the SiC substrate 10, a SiC layer SL having a thickness of 5 μm to 200 μm and having an n-type is formed by epitaxial growth by chemical vapor deposition (CVD). The n-type impurity concentration (donor concentration) is, for example, 1×10 14 cm -3 Up to 1×10 17 cm -3 The portion of the SiC layer SL formed in this way other than the portion in which the conductivity type is inverted by ion implantation described later becomes the drift layer 20 for the MOSFET 711 .
[0084] Next, an implantation mask is formed on the upper surface of the SiC layer SL by photoresist or the like. Using the implantation mask, Al (aluminum) as a p-type impurity (acceptor) is added to the upper surface of the SiC layer SL by ion implantation. The depth of the ion implantation does not exceed the thickness of the SiC layer SL, and is set to about 0.3 μm to 3 μm, for example. In addition, the doping concentration of Al is higher than the impurity concentration of the drift layer 20, for example, 1×10 17 cm -3 Up to 1×10 19 cm -3 Thereafter, the implantation mask is removed. The region imparted with p-type by adding Al in this step becomes the first well region 30 and the second well region 31.
[0085] In addition, some additional layers may be provided on the upper surface side of the SiC substrate 10 before forming the SiC layer SL, in which case the additional layers are regarded as part of the SiC substrate 10. As an additional layer, for example, an n-type buffer layer having a higher concentration than the drift layer 20 may be formed. The n-type buffer layer enables the minority carriers intruding from the drift layer 20 into the SiC substrate 10 to recombine. Instead of or in addition to the n-type buffer layer, a conversion layer that converts the BPD present on the surface of the SiC substrate 10 into a threading edge dislocation may be formed as an additional layer by epitaxial growth.
[0086] Next, an implantation mask is formed on the upper surface of the SiC layer SL by using a photoresist or the like. Using the implantation mask, Al as a p-type impurity is added to the upper surface of the SiC layer SL by ion implantation. The depth of the ion implantation does not exceed the thickness of the SiC layer SL, and is set to about 0.3 μm to 3 μm, for example. In addition, the doping concentration of Al is higher than the impurity concentration of the drift layer 20 and lower than the impurity concentration of the second well region 31, and is, for example, 1×10 16 cm -3 Up to 5×10 18 cm -3 The implantation mask is then removed. The region imparted with p-type by adding Al in this process becomes the JTE region 37. Similarly, by ion implanting Al at a higher impurity concentration than the impurity concentration of the first well region 30 in a predetermined region on the surface of the first well region 30, a contact region 35 is formed.
[0087] Next, an implantation mask is formed on the upper surface of the SiC layer SL by a photoresist or the like so as to expose a predetermined portion that is further inside the outer edge of the first well region 30. Using this implantation mask, N (nitrogen) as an n-type impurity (donor) is added to the upper surface of the SiC layer SL by ion implantation. The depth of this ion implantation is shallower than the thickness of the first well region 30. In addition, the doping concentration of N exceeds the p-type impurity concentration of the first well region 30, for example, 1×10 18 cm -3 Up to 1×10 21 cm -3 The region imparted with n-type by adding N in this step becomes the source region 40.
[0088] Next, annealing is performed in a heat treatment apparatus in an inert gas atmosphere such as argon (Ar) gas at a temperature of 1300° C. to 1900° C. for 30 seconds to 1 hour. Through this annealing, N and Al added by ion implantation are electrically activated.
[0089] Next, a field insulating film 51 made of silicon oxide with a film thickness of 0.3 μm to 2 μm is formed on a portion of the peripheral region 200 on the SiC layer SL by using a CVD method or a photolithography technique. Next, a silicon oxide film is formed as a gate insulating film 50 by thermally oxidizing the portion of the upper surface of the SiC layer SL that is not covered by the field insulating film 51. Next, a polycrystalline silicon film having conductivity is formed on the gate insulating film 50 and the field insulating film 51 by a reduced pressure CVD method. The gate electrode 60 is formed by patterning the polycrystalline silicon film. Next, an interlayer insulating film 55 made of silicon oxide is formed by a reduced pressure CVD method. Next, a contact hole 90 is formed that penetrates the interlayer insulating film 55 and the gate insulating film 50 and reaches the contact region 35 and the source region 40 in the active region 100. Similarly, a contact hole 91 is formed that reaches the second well region 31.
[0090] Next, a metal film mainly composed of Ni is formed by sputtering or the like. Then, by heat treatment at a temperature of 600° C. to 1100° C., Ni atoms of the metal film and Si atoms of the SiC layer SL react in the contact holes 90 and 91, thereby forming a silicide layer as the ohmic electrode 70 between the SiC layer SL and the metal film. Next, the portion of the metal film that has not been transformed into the silicide layer is removed by wet etching.
[0091] Next, a metal film mainly composed of Ni is formed on the back surface of the SiC substrate 10. The metal film is heat-treated to form a back surface ohmic electrode (not shown).
[0092] Next, by etching using a photoresist or the like as a mask, a contact hole 95 penetrating the interlayer insulating film 55 is formed.
[0093] Next, by film formation using a sputtering method or the like and patterning using a photolithography method or the like, a portion of source electrode 80 is formed on ohmic electrode 70 in contact hole 90 and contact hole 91 .
[0094] Next, on the upper surface side of the SiC substrate 10 , a wiring metal such as Al is deposited by sputtering or vapor deposition and patterned by photolithography or the like to form the remaining portion of the source electrode 80 , the gate pad 81 , and the gate wiring 82 .
[0095] Next, the drain electrode 84 is formed on the surface of the back ohmic electrode (not shown) formed on the back surface of the SiC substrate 10. As described above, the MOSFET 711 is completed.
[0096] <1-4. Action>
[0097] Next, the operation of the MOSFET 711, particularly its freewheeling operation, will be described below, taking as an example the case where the SiC material of the SiC layer SL has a polytype 4H. The diffusion potential of the pn junction of SiC of the polytype 4H is approximately 2V.
[0098] When MOSFET 711 is used as a switching element in a power conversion system in Embodiment 3 described later, for example, the state of the channel region of MOSFET 711 is switched between an on state and an off state. In the off state, there is a period during which the drain voltage (the voltage of the drain electrode 84) becomes lower than the source voltage (the voltage of the source electrode 80), and at this time, the drain voltage becomes, for example, several V to more than ten V. During this period, MOSFET 711 performs a freewheeling operation.
[0099] Specifically, a forward voltage exceeding 2V equivalent to the diffusion potential is applied to the pn junction formed by the first well region 30 and the second well region 31 ohmically connected to the source electrode 80 and the drift layer 20 ohmically connected to the drain electrode 84. That is, a voltage exceeding the forward voltage is applied to the body diode of MOSFET711. As a result, MOSFET711 acts as a freewheeling diode. Therefore, a large bipolar current flows in the direction from the source electrode 80 toward the drain electrode 84. The current density of this current is 100A / cm 2 In the above case, significant conductivity modulation occurs in the drift layer 20. That is, a large number of holes flow into the drift layer 20, holes are accumulated in the drift layer, and electrons having a concentration higher than that in the thermal equilibrium state are electrically attracted to the holes, so that electrons are accumulated in the drift layer 20 in the same manner as holes. As a result, the resistivity of the drift layer 20 is reduced to the same level as the resistance value of the SiC substrate 10, which is a low resistance.
[0100] Note that operations other than the freewheeling operation are substantially the same as those of a normal MOSFET, and thus description thereof will be omitted.
[0101] <1-5. Simulation>
[0102] In the above continuous flow action, at the surface density S( Figure 6 ) is not optimized, the current density increases locally near the boundary between the active region 100 and the peripheral region 200, and on the other hand, as described in this embodiment, the surface density S( Figure 6 ) is optimized, this increase can be suppressed. To illustrate this, a two-dimensional T-CAD simulation modeling a body diode included in a vertical MOSFET is described below.
[0103] Figure 7 is the surface density S( Figure 6 ) in the active area 100( Figure 3) is a cross-sectional view of a comparative example in which the body diode is roughly equal. Figure 3 ) are modeled separately, and the cathode electrode 84D, SiC substrate 10D, SiC layer SLD and anode electrode 80D are set. The SiC layer SLD has a drift layer 20 ( Figure 3 ) and the first well region 30 and the second well region 31 ( Figure 3 ) is used to model the well region 30D. The SiC substrate 10D has a thickness t sub , the drift layer 20D has a thickness t drift , their sum is the thickness t tot . Regarding the in-plane direction ( Figure 7 horizontal in the middle), Figure 7 The left end of the model is connected to the active area 100 ( Figure 3 ) corresponds to the center position. The electrode width LA is the dimension from the center position to the outer edge of the anode electrode 80D, and the electrode width LC is the dimension from the center position to the outer edge of the cathode electrode 84D. Figure 7 In order to make the drawings easier to see, the longitudinal dimensions are exaggerated. In actual dimensions, the electrode width LA and the electrode width LC are much larger than the thickness t tot .
[0104] Figure 8 The anode electrode 80D is shown near the end of the anode electrode 80D ( Figure 7 )'s center (in Figure 7 The distribution diagram of the simulation result of the current density when a current equivalent to 1000A per unit area of the anode electrode 80D is applied (at the left end position in the middle). According to this simulation, the area with extremely high current density appears in an elliptical shape with the end of the anode electrode 80D as the center.
[0105] Fig. 9 is along the well region 30D ( Figure 7 ) is shown by the bottom surface at t sub =t drift ×39 case and t sub =t drift The following is a graph of current density obtained by performing the same simulation as above under the two conditions of the case of ×11. Fig. 9 The current density is the largest at the position x=0 in the figure. sub When the value is large, the peak value and the peak width of the current density are large, which indicates that the thickness t of the SiC substrate 10D is large. sub Affects the state of current concentration.
[0106] Refer again Figure 8 It can be seen that the expansion of the current in the drift layer 20D and the expansion of the current in the SiC substrate 10D are roughly continuous. This is believed to be caused by the resistivity of the drift layer 20D being reduced to the resistivity of the SiC substrate 10D through conductivity modulation. Therefore, it is believed that the thickness t of the SiC substrate 10D is not sub Or the thickness t of the drift layer 20D drift The thickness t of the resistor including both the SiC substrate 10D and the drift layer 20D is not tot The distribution of relative current density is of more direct relevance.
[0107] In the analysis Fig. 9 When the simulation results are shown, regardless of the two thickness conditions, the current density ( Fig. 9 The current density on the left end of the graph is 10% higher than the reference current density, which is from the end of the anode electrode 80D ( Fig. 9 x=0 in the figure) to the inner side ( Fig. 9 Left direction) thickness t tot In addition, the position where the current density is 30% higher than the current density at the center of the anode electrode 80D is the position from the end of the anode electrode 80D to the inner side with a thickness of t tot In addition, the position where the current density is 50% higher than the current density at the center of the anode electrode 80D is the thickness t from the end position of the anode electrode 80D inward. tot According to this analysis, the sparse region portion 101 ( Figure 2 ) is preferably a width (dimension in the x direction) of the SiC substrate 10 ( Figure 3 ) and the thickness of the SiC layer SL( Figure 3 ) is 20% or more of the sum of the thicknesses of the layers 1 and 2, more preferably 30% or more, and even more preferably 50% or more.
[0108] Fig.10 is shown with Fig. 9 The following is a graph of the simulation results of the hole concentration distribution under the same conditions. Regarding the hole concentration distribution, it is also observed that the current density distribution ( Fig. 9 ) The same distribution is obtained. Therefore, it is believed that by adopting a structure that alleviates the concentration of current density, the local increase in hole concentration can also be suppressed. Therefore, it is believed that by suppressing the degradation of the characteristics of the body diode, a highly reliable silicon carbide semiconductor device can be achieved.
[0109] In addition, it is considered that when the sparse region 101 expands to a region where the hole concentration does not substantially increase, a new increase in hole concentration occurs inside the sparse region 101. Therefore, if the width of the sparse region 101 is set in a manner to suppress this new increase, it is presumed that the above-mentioned effect can be obtained more reliably. Fig.10 When the simulation results are given, the hole concentration on the central side of the anode electrode 80D ( Fig.10 The hole concentration on the left end of the graph is taken as the reference. The position where the hole concentration is 1% higher is the position from the end of the anode electrode 80D ( Fig.10 x=0 in the figure) to the inner side ( Fig.10 Left direction) thickness t tot In addition, the position where the hole concentration is 3% higher than the center of the anode electrode 80D is the position from the end of the anode electrode 80D to the inner side with a thickness of t tot Based on this analysis, it is estimated that the sparse region 101 ( Figure 2 ) is preferably a width (dimension in the x direction) of the SiC substrate 10 ( Figure 3 ) and the thickness of the SiC layer SL( Figure 3 ) is less than 110% of the sum of the thicknesses of the layers 1 and 2, more preferably less than 70%.
[0110] Fig.11 Considering the active area 100( Figure 3 ) in the surface density S( Figure 6 ) is a cross-sectional diagram that models the body diode. Figure 7 The difference between the models is that, in order to model the difference in the surface density S in the active region, the contact resistivity is defined at the boundary between the anode electrode 80D and the well region 30D. Specifically, the boundary is at the sparse region portion 101 ( Figure 3 ) corresponds to a width LT with a contact resistivity of ρcc = 1×10 -4 (Ω·cm 2 ), in accordance with the standard area section 102 ( Figure 3 ) has a contact resistivity of ρct = ρcc × TM (Ω·cm 2 )(TM is the coefficient). Fig.11 In the model, the thickness t tot Set to unit size. In other words, t tot = 1. Moreover, t drift =0.08, t sub =0.92, LA=39, LT=0.88.
[0111] Fig.12 The anode electrode 80D is shown near the end of the anode electrode 80D ( Fig.11 )'s center (in Fig.11 The distribution diagram of the simulation result of the hole concentration distribution when a current equivalent to 1000A per unit area of the anode electrode 80D is applied (at the left end position in the middle). In this simulation, in order to investigate the relationship between ρct=ρcc×TM( Fig.11 ) value, set TM = 1.0, 1.2, 1.5, 2.0, 5.0, 10.0, 20.0 and 50.0.
[0112] The simulation results show that by making ρct 1.5 times or more of ρcc, the hole concentration peak near x=0 is significantly suppressed. Figure 4 ), the average contact resistivity in the sparse region 101 is preferably set to be 1.5 times or more of the average contact resistivity in the standard region 102. The contact resistivity of the surface region 71 is considered to be approximately constant, so in order to change the distribution of the average contact resistivity, it is necessary to change the surface density of the surface region 71. In order to obtain the above-mentioned average contact resistivity distribution, the surface density of the surface region 71 of the sparse region 101 is preferably less than 2 / 3 of the surface density of the surface region 71 of the standard region 102.
[0113] On the other hand, when the value of the coefficient TM is greater than 1, at a position closer to the center ( Fig.12 A new peak value of hole concentration appears at the position near x=-1 in the graph of TM. The value of this peak value when the coefficient TM is greater than 20 is roughly close to the value of the peak value near x=0 when TM=1. Therefore, in order to avoid the adverse effects caused by the increase in hole concentration at the position of the new peak, the coefficient TM is preferably less than 10. Correspondingly, the surface density of the surface area 71 of the sparse area portion 101 is preferably greater than 1 / 10 of the surface density of the surface area 71 of the standard area portion 102.
[0114] <1-6. Effect>
[0115] According to the present embodiment, the sparse region portion 101 has a larger surface area 71 ( Figure 4 ) surface density S( Figure 6 ) is lower. Such a surface density distribution can be obtained without particularly complicating the manufacturing method, so the manufacturing cost will not be significantly increased. Moreover, the surface density S of the surface area 71 is lower than that of the standard area 102 by the sparse area portion 101, and the local increase of the bipolar current in the local area near the boundary between the peripheral area 200 and the active area 100 is suppressed. As a result, the expansion of the stacking defect can be fully suppressed in this local area. Therefore, it is possible to fully suppress the occurrence of the SiC layer SL ( Figure 3) The characteristics of the MOSFET 711 are deteriorated due to the expansion of the stacking defect in the MOSFET 711. According to the above, it is possible to avoid a significant increase in manufacturing cost and suppress the characteristics of the MOSFET 711 from being deteriorated.
[0116] If the surface density S of the sparse region 101 is 1 / 10 or more and 2 / 3 or less of the surface density S of the active region 100, the contact resistivity of the sparse region 101 is 1.5 or more and 10 or less of the contact resistivity of the standard region 102. In this case, if the simulation result of the hole concentration distribution is considered ( Fig.12 ), then near the boundary between the sparse area portion 101 and the peripheral area 200 ( Fig.12 ) and the boundary between the sparse area portion 101 and the standard area portion 102 ( Fig.12 In this way, the local peak value of the hole concentration distribution is prevented from becoming too large. Therefore, the local increase of the extension of the stacking defect near each boundary is avoided.
[0117] The width of the sparse area portion 101 ( Figure 2 The lateral dimension in the figure) is the SiC substrate 10 ( Figure 3 ) and the thickness of the SiC layer SL( Figure 3 ) is more than 20% of the sum of the thicknesses, if Fig. 9 According to the simulation results shown in FIG. 1 , the local increase of the bipolar current is suppressed in the substantially entire range where the local increase of the bipolar current is particularly large. In addition, if the percentage reaches about 70%, the effect can be obtained more fully. In connection with the simulation, according to the above reasons, if it reaches about 110%, the effect can be obtained substantially ideally. Therefore, from the viewpoint of avoiding the proportion of the sparse region 101 in the active region 100 from becoming too large, the upper limit of the percentage can be set to 110%, and further can be set to 70%.
[0118] The current density of MOSFET 711 in the on state can be 100A / cm 2 As described above, when the current density is high, a large number of holes flow into the drift layer 20, causing significant conductivity modulation. Therefore, the resistivity of the drift layer 20 is significantly reduced. According to this embodiment, the local increase of the bipolar current generated under such conditions can be effectively suppressed.
[0119] In this embodiment, in the standard region portion 102, a plurality of first well regions 30 ( Figure 4 ) has a repeatedly arranged polygonal shape (specifically, a rectangular shape, more specifically, a square shape). Thus, each of the plurality of first well regions 30 can be provided with a polygonal outer edge in a plan view.
[0120] <1-7. Modification of Implementation Example 1>
[0121] Fig.13 is shown as MOSFET711 ( Figure 4 ) of the first variant of MOSFET712 (silicon carbide semiconductor device). Figure 4 ) is different, in MOSFET712( Fig.13 ), the surface region 71 is arranged away from the second well region 31. In other words, the ohmic electrode 70 ( Figure 3 ) is arranged offset from the second well region 31. In addition, unlike the first embodiment described above, in this variation, the second well region 31 ( Fig.13 ) is connected to the first well region 30 at the boundary between the peripheral region 200 and the active region 100 .
[0122] Fig.14 is shown as MOSFET711 ( Figure 4 ) of the second variant of MOSFET713 (silicon carbide semiconductor device). Figure 4 ) is different, in MOSFET713( Fig.14 ) in which the plurality of first well regions 30 have a stripe shape in the in-plane direction.
[0123] Fig.15 is shown as MOSFET711 ( Figure 4 ) is a partial top view of a MOSFET 714 (silicon carbide semiconductor device) of a third variant. Fig.15 ) and MOSFET712( Fig.13 ) Similarly, the surface region 71 is arranged away from the second well region 31 and is connected to the MOSFET 713 ( Fig.14 ) Similarly, the plurality of first well regions 30 have a stripe shape in the in-plane direction.
[0124] Fig.16 is shown as MOSFET711 ( Figure 4 ) is a partial top view of a MOSFET 715 (silicon carbide semiconductor device) according to a fourth variant. Fig.16 ), with respect to the surface region 71, the surface density of the sparse region portion 101 increases as it moves toward the outside ( Fig.16 Specifically, the sparse area portion 101 is lowered from the inside to the outside (in Fig.16The sparse region section 101 is composed of a plurality of parts (from the inside to the right side), and the surface density of these plurality of parts decreases as it goes to the outside. Therefore, the surface density of the sparse region section 101 has a plurality of levels. More specifically, the sparse region section 101 is composed of a part 101a and a part 101b from the inside to the outside, and the surface density of the part 101b is lower than that of the part 101a. Therefore, the surface density of the sparse region section 101 has a relatively high level in the part 101a and a relatively low level in the part 101b.
[0125] According to the fourth modification, the degree of freedom in designing the distribution of the surface density is increased, so that the width ( Fig.16 The same effect as that of the first embodiment is obtained. Fig.16 The example in FIG. 2 illustrates a case where the number of levels of surface density is 2, but the number of levels of surface density is not limited to 2. As a modified example, a structure can be used in which the number of levels is sufficiently increased so that the surface density of the sparse region portion 101 decreases approximately continuously toward the outside.
[0126] In addition, if Figure 13 to Figure 16 When the top view is assumed to be a cross-sectional view along an inclined straight line (not shown), the second well region 31 (with Figure 3 ) appears separately from the first well region 30. Therefore, in common with the first embodiment and its modified example, the first well region 30 and the second well region 31 are separated from each other when viewed in at least one cross section.
[0127] <2. Implementation Method 2>
[0128] Fig.17 So with Figure 4 (Embodiment 1) The same view schematically shows a partial top view of the structure of MOSFET 721 according to Embodiment 2.
[0129] Unlike the first embodiment, in the second embodiment, the active region 100 does not need to have a sparse region portion 101 ( Figure 4 ), for example, only the standard area unit 102 ( Figure 4 On the other hand, in the second embodiment, the outer peripheral region 200 includes an outer region portion 202 and a sparse region portion 201 between the outer region portion 202 and the active region 100 .
[0130] The sparse region 201 has a lower surface density in the surface region 71 when viewed from above than the active region 100. Furthermore, the standard region 102 ( Figure 4 ) of the surface density and sparse area portion 101 ( Figure 4) of the surface density described in the first embodiment can also be applied to the active region 100 ( Fig.17 ) and the sparse region portion 201 ( Fig.17 ) relationship.
[0131] The configuration other than the above is substantially the same as that of the first embodiment, and thus the description thereof will not be repeated. The same modification as that described in the modification of the first embodiment can also be applied to the second embodiment.
[0132] <3. Implementation Method 3>
[0133] This embodiment is an example of applying any of the silicon carbide semiconductor devices (MOSFETs 711 to 715 and 721) involved in any of the above embodiments or their variations to a power conversion device. The application of these silicon carbide semiconductor devices is not limited to a specific power conversion device. Hereinafter, as Embodiment 3, the case where these silicon carbide semiconductor devices are applied to a three-phase inverter is described.
[0134] Fig.18 9 is a block diagram showing a configuration of a power conversion system according to Embodiment 3. The power conversion system includes a power source 910 , a power conversion device 920 , and a load 930 .
[0135] The power supply 910 is a DC power supply, and supplies DC power to the power conversion device 920. The power supply 910 may be composed of, for example, a DC system, a solar cell, or a storage battery, or may be composed of a rectifier circuit or an AC / DC converter connected to an AC system. Alternatively, the power supply 910 may be composed of a DC / DC converter that converts the DC power output from the DC system into a predetermined power.
[0136] The power conversion device 920 is a three-phase inverter connected between the power supply 910 and the load 930, and converts the DC power supplied from the power supply 910 into AC power, and supplies the AC power to the load 930. The power conversion device 920 includes: a main conversion circuit 921 that converts the DC power into AC power and outputs it; a drive circuit 922 that outputs a drive signal for driving each switching element of the main conversion circuit 921; and a control circuit 923 that outputs a control signal for controlling the drive circuit 922 to the drive circuit 922.
[0137] The load 930 is a three-phase motor driven by the AC power supplied from the power conversion device 920. In addition, the load 930 is not limited to a specific use, for example, it is a motor mounted on various electrical equipment, for example, used as a motor for hybrid vehicles, electric vehicles, railway vehicles, elevators or air conditioning equipment.
[0138] The power conversion device 920 will be described in detail below.
[0139] The main conversion circuit 921 receives input power from the power supply 910, converts the input power into output power, and outputs the output power to the load 930. Specifically, the main conversion circuit 921 has a switching element and a freewheeling diode (not shown), and converts the DC power supplied from the power supply 910 into AC power by switching the switching element, and supplies it to the load 930. There are various examples of the specific circuit structure of the main conversion circuit 921, but the main conversion circuit 921 involved in this embodiment 3 is a 2-level three-phase full-bridge circuit, which can be composed of 6 switching elements and 6 freewheeling diodes connected in anti-parallel to each switching element. As each switching element of the main conversion circuit 921, a silicon carbide semiconductor device involved in any embodiment or its modified example in the above-mentioned embodiment is applied. With respect to the 6 switching elements, each 2 switching elements are connected in series, thereby forming 3 upper and lower branches. Each upper and lower branch constitutes each phase (U phase, V phase, W phase) of the full-bridge circuit. Moreover, the output terminals of each upper and lower branch, that is, the three output terminals of the main conversion circuit 921 are connected to the load 930.
[0140] The drive circuit 922 generates a drive signal for driving the switch element of the main conversion circuit 921, and supplies the drive signal to the control electrode of the switch element of the main conversion circuit 921. Specifically, the drive circuit 922 outputs a drive signal for turning on the switch element and a drive signal for turning off the switch element to the control electrode of each switch element according to a control signal from a control circuit 923 described later. When the switch element is maintained in the on state, the drive signal is a voltage signal (on signal) above the threshold voltage of the switch element, and when the switch element is maintained in the off state, the drive signal is a voltage signal (off signal) less than the threshold voltage of the switch element.
[0141] The control circuit 923 controls the switch elements of the main conversion circuit 921 via the drive circuit 922 in such a manner that the desired power is supplied to the load 930. For this purpose, the control circuit 923 calculates the time (on time) when each switch element of the main conversion circuit 921 should be in the on state according to the power to be supplied to the load 930. For example, the main conversion circuit 921 can be controlled by PWM control that modulates the on time of the switch element according to the voltage to be output. In addition, the control circuit 923 outputs a control signal (control instruction) for controlling the drive circuit 922 to the drive circuit 922. According to the control signal, the drive circuit 922 outputs an on signal or an off signal as a drive signal to the control electrode of each switch element. The control signal of the control circuit 923 is generated in such a manner that the drive circuit 922 outputs an on signal corresponding to the switch element that becomes the on state and outputs an off signal corresponding to the switch element that becomes the off state at each time point.
[0142] In the power conversion device involved in this embodiment, a silicon carbide semiconductor device involved in any embodiment of the above-mentioned embodiment or its modified example is applied as a switching element of the main conversion circuit 921, so that a power conversion device that avoids a significant increase in manufacturing cost and suppresses degradation of characteristics can be realized.
[0143] In the third embodiment, a two-level three-phase inverter is exemplified as the power conversion device, but the power conversion device using the silicon carbide semiconductor device involved in any of the above embodiments or its modified examples is not limited to this. The power conversion device may be, for example, a multi-level power conversion device with three or more levels or a single-phase inverter that supplies power to a single-phase load. In addition, when supplying power to a DC load or the like, the power conversion device may be a DC / DC converter or an AC / DC converter.
[0144] In addition, the power conversion device using the silicon carbide semiconductor device involved in any of the above embodiments or its modified examples can be used without being limited to the case where the above load is a motor. For example, the power conversion device can be used as a power supply device for an electric discharge machine, a laser processing machine, an induction heating cooker, or a contactless power supply system. Alternatively, the power conversion device can be used as a power conditioner for a solar power generation system or a power storage system.
[0145] In addition, each embodiment can be freely combined or each embodiment can be appropriately modified or omitted. Although the present disclosure is described in detail, the above description is illustrative in all aspects and is not limited to this. It should be understood that countless modifications that are not exemplified can be conceived from the present disclosure.
Claims
1. A silicon carbide semiconductor device comprising: A silicon carbide substrate having a first conductivity type; and a silicon carbide layer having a first surface facing the silicon carbide substrate and a second surface opposite to the first surface in a thickness direction, and having an active region and a peripheral region arranged at the periphery of the active region in an in-plane direction perpendicular to the thickness direction, The silicon carbide layer comprises: A drift layer, disposed on the silicon carbide substrate, having the first conductivity type; a plurality of first well regions, arranged on the drift layer in the active region, having a second conductivity type different from the first conductivity type, and separated from each other when viewed in at least one cross section; a plurality of source regions, disposed on the plurality of first well regions, and having the first conductivity type; and a second well region, arranged on the drift layer in the outer peripheral region and having the second conductivity type, The silicon carbide semiconductor device further comprises: a gate insulating film facing the plurality of first well regions; a gate electrode having a portion disposed above the active region and facing the plurality of first well regions via the gate insulating film and a portion disposed above the peripheral region and insulated from the second well region; a gate pad, arranged above the second well region, insulated from the second well region, and connected to the gate electrode; A source electrode disposed above the second surface of the silicon carbide layer; as well as a plurality of ohmic electrodes disposed on the second surface of the silicon carbide layer, connected to the source electrode, electrically ohmically connected to the plurality of first well regions, and having a plurality of surface regions in ohmic contact with a portion of the second conductivity type in the second surface of the silicon carbide layer, The active region includes a standard region portion and a sparse region portion between the standard region portion and the peripheral region, wherein the sparse region portion has a lower surface density of the plurality of surface regions when viewed from above than the standard region portion. A width of the sparse region portion corresponding to a distance between the standard region portion and the outer peripheral region is equal to or greater than 20% of a sum of a thickness of the silicon carbide substrate and a thickness of the silicon carbide layer.
2. The silicon carbide semiconductor device according to claim 1, wherein: The plurality of ohmic electrodes are arranged offset from above the second well region.
3. A silicon carbide semiconductor device comprising: A silicon carbide substrate having a first conductivity type; and a silicon carbide layer having a first surface facing the silicon carbide substrate and a second surface opposite to the first surface in a thickness direction, and having an active region and a peripheral region arranged at the periphery of the active region in an in-plane direction perpendicular to the thickness direction, The silicon carbide layer comprises: A drift layer, disposed on the silicon carbide substrate, having the first conductivity type; a plurality of first well regions, arranged on the drift layer in the active region, having a second conductivity type different from the first conductivity type, and separated from each other when viewed in at least one cross section; a plurality of source regions, disposed on the plurality of first well regions, and having the first conductivity type; and a second well region, arranged on the drift layer in the outer peripheral region and having the second conductivity type, The silicon carbide semiconductor device further comprises: a gate insulating film facing the plurality of first well regions; a gate electrode having a portion disposed above the active region and facing the plurality of first well regions via the gate insulating film and a portion disposed above the peripheral region and insulated from the second well region; a gate pad, arranged above the second well region, insulated from the second well region, and connected to the gate electrode; A source electrode disposed above the second surface of the silicon carbide layer; as well as a plurality of ohmic electrodes disposed on the second surface of the silicon carbide layer, connected to the source electrode, electrically ohmically connected to the plurality of first well regions, and having a plurality of surface regions in ohmic contact with a portion of the second conductivity type in the second surface of the silicon carbide layer, The active region includes a standard region portion and a sparse region portion between the standard region portion and the peripheral region, wherein the sparse region portion has a lower surface density of the plurality of surface regions when viewed from above than the standard region portion. The plurality of ohmic electrodes are arranged offset from above the second well region.
4. The silicon carbide semiconductor device according to any one of claims 1 to 3, characterized in that: In the in-plane direction, the plurality of first well regions have polygonal shapes that are repeatedly arranged in the standard region portion.
5. A silicon carbide semiconductor device comprising: A silicon carbide substrate having a first conductivity type; and a silicon carbide layer having a first surface facing the silicon carbide substrate and a second surface opposite to the first surface in a thickness direction, and having an active region and a peripheral region arranged at the periphery of the active region in an in-plane direction perpendicular to the thickness direction, The silicon carbide layer comprises: A drift layer, disposed on the silicon carbide substrate, having the first conductivity type; a plurality of first well regions, arranged on the drift layer in the active region, having a second conductivity type different from the first conductivity type, and separated from each other when viewed in at least one cross section; a plurality of source regions, disposed on the plurality of first well regions, and having the first conductivity type; and a second well region, arranged on the drift layer in the outer peripheral region and having the second conductivity type, The silicon carbide semiconductor device further comprises: a gate insulating film facing the plurality of first well regions; a gate electrode having a portion disposed above the active region and facing the plurality of first well regions via the gate insulating film and a portion disposed above the peripheral region and insulated from the second well region; a gate pad, arranged above the second well region, insulated from the second well region, and connected to the gate electrode; A source electrode disposed above the second surface of the silicon carbide layer; as well as a plurality of ohmic electrodes disposed on the second surface of the silicon carbide layer, connected to the source electrode, electrically ohmically connected to the plurality of first well regions and the second well regions, and having a plurality of surface regions in partial ohmic contact with the second conductivity type in the second surface of the silicon carbide layer, The peripheral region includes an outer region portion and a sparse region portion between the outer region portion and the active region, wherein the sparse region portion has a lower surface density of the plurality of surface regions when viewed from above than the active region. A width of the sparse region portion corresponding to a distance between the active region and the outer region portion is equal to or greater than 20% of a sum of a thickness of the silicon carbide substrate and a thickness of the silicon carbide layer.
6. The silicon carbide semiconductor device according to claim 5, wherein: The width of the sparse region portion is equal to or less than 110% of the sum of the thickness of the silicon carbide substrate and the thickness of the silicon carbide layer.
7. The silicon carbide semiconductor device according to any one of claims 1 to 6, wherein: The plurality of first well regions have a stripe shape in the in-plane direction.
8. The silicon carbide semiconductor device according to any one of claims 1 to 7, wherein: The surface density of the sparsely populated region is not less than 1 / 10 and not more than 2 / 3 of the surface density of the active region.
9. The silicon carbide semiconductor device according to any one of claims 1 to 8, characterized in that: The surface density of the sparse region portion decreases toward the outside.
10. The silicon carbide semiconductor device according to any one of claims 1 to 9, characterized in that: The surface density of the sparse region portion has a plurality of levels.
11. The silicon carbide semiconductor device according to any one of claims 1 to 10, characterized in that: The current density in the on state is 100A / cm 2 above.
12. A power conversion device comprising: A main conversion circuit, comprising the silicon carbide semiconductor device according to any one of claims 1 to 11, the main conversion circuit receiving input power, converting the input power into output power, and outputting the output power; a driving circuit, outputting a driving signal for driving the silicon carbide semiconductor device to the silicon carbide semiconductor device; as well as The control circuit outputs a control signal for controlling the drive circuit to the drive circuit.
