Silicon carbide device with trench gate

Abstract

A silicon carbide component (500) includes a strip-shaped trench gate structure (150) extending from a first surface (101) into a silicon carbide body (100), wherein the gate structure (150) is delimited along a lateral second direction (292) by a first gate sidewall (151, 151') of the gate structure (150) and a second gate sidewall (152, 152') of the gate structure (150). The trench gate structure (150) comprises a first section (150-1) and a second section (150-2) that are laterally offset from one another along the lateral first direction (291). In the first section (150-1), the first gate sidewall (151') extends to a first depth (t1) and in the second section (150-2), the first gate sidewall (151) extends to a second depth (t2*) along the vertical direction (104) into the silicon carbide body (100), wherein the second depth (t2*) is greater than the first depth (t1).

Description

TECHNICAL FIELD

[0001] The present disclosure relates to a silicon carbide device, in particular to a silicon carbide switch with transistor cells. BACKGROUND

[0002] Electrical circuits for converting electrical energy, for example, in DC / AC converters, AC / AC converters, or AC / DC converters, and in electrical circuits that drive inductive loads, e.g., in motor driver circuits, can include power semiconductor components as switches. Switching large inductive loads can trigger LC oscillations. On the other hand, the dielectric breakdown field strength of silicon carbide (SiC) is high compared to silicon. SiC devices can be significantly thinner than equivalent silicon devices for the same nominal blocking voltage capability, and consequently, the on-resistance of SiC devices can be significantly lower.

[0003] SiC devices can be very sensitive to variations in different capacitances within the device, for example, a gate-drain capacitance (Cgd) and a gate-source capacitance (Cgs). Capacitance, especially their relationship, can affect the switching behavior of the device and oscillations during switching. Increasing Cgs relative to Cds has proven effective in some applications.

[0004] Therefore, there is a need to increase Cgs relative to Cgd within the device. SUMMARY

[0005] One embodiment of the present disclosure relates to a silicon carbide device. The silicon carbide device includes a stripe-shaped trench gate structure extending from a first surface into a silicon carbide body, the gate structure having a gate length along a lateral first direction. The stripe-shaped trench gate structure is bounded along a lateral second direction by a first gate sidewall of the gate structure and a second gate sidewall of the gate structure, the first gate sidewall and the second gate sidewall being connected via a bottom surface of the gate structure, and the lateral second direction being perpendicular to the lateral first direction. The trench gate structure comprises a first portion and a second portion laterally offset from each other along the lateral first direction.In the first portion of the trench gate structure, the first gate sidewall extends to a first depth from the first surface along a vertical direction into the silicon carbide body, the vertical direction being perpendicular to both the lateral first direction and the lateral second direction. In the second portion of the trench gate structure, the first gate sidewall extends to a second depth from the first surface along the vertical direction into the silicon carbide body. The second depth is greater than the first depth.

[0006] One embodiment of the present disclosure relates to a silicon carbide device. The silicon carbide device includes a stripe-shaped trench gate structure extending from a first surface into a silicon carbide body, the gate structure having a gate length along a lateral first direction. The stripe-shaped trench gate structure is bounded along a lateral second direction by a first gate sidewall of the gate structure and a second gate sidewall of the gate structure, the first gate sidewall and the second gate sidewall being connected via a bottom surface of the gate structure, and the lateral second direction being perpendicular to the lateral first direction. The trench gate structure comprises a first portion and a second portion laterally offset from each other along the lateral first direction.In the first section of the trench gate structure, the first gate sidewall extends to a first depth from the first surface along a vertical direction into the silicon carbide body, the vertical direction being perpendicular to both the lateral first direction and the lateral second direction. In the first section of the trench gate structure, the second gate sidewall extends to a third depth from the first surface along the vertical direction into the silicon carbide body. The third depth is greater than the first depth. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] The accompanying drawings are included to provide a further understanding of the embodiments and are incorporated in and constitute a part of this specification. The drawings illustrate embodiments of a silicon carbide device and, together with the description, serve to explain principles of the embodiments. Additional embodiments are described in the following detailed description and claims. Fig. 1A-1D illustrate schematic top view and vertical cross-sectional views of a portion of a silicon carbide device having source regions and shield regions, according to embodiments. Fig. 2A-2D illustrate schematic top view and vertical cross-sectional views of a portion of a silicon carbide device according to an embodiment having a shield region including deep subsections arranged in a matrix. Fig. 3A-3D illustrate schematic top view and vertical cross-sectional views of a portion of a silicon carbide device according to an embodiment having a shielding region including stripe-shaped deep sections. Fig. 4A-4D illustrate schematic top view and vertical cross-sectional views of a portion of a silicon carbide device according to an embodiment having a shielding region including a grid-shaped deep section. Fig. 5A-5B illustrate schematic perspective views of a portion of a silicon carbide device according to another embodiment having a shield region including laterally separated deep subsections. Fig. 6A-6C illustrate schematic horizontal cross-sectional views of portions of silicon carbide devices according to further embodiments. Fig. 7A-7G illustrate schematic top view and vertical cross-sectional views of a portion of a silicon carbide device according to an embodiment having a shield region including deep subsections arranged in a matrix. DETAILED DESCRIPTION

[0008] In the following detailed description, reference is made to the accompanying drawings, which form a part hereof, and in which is shown by way of illustration specific embodiments in which a silicon carbide device may be practiced. It should be understood that other embodiments may be utilized and structural or logical changes may be made without departing from the scope of the present disclosure. For example, features illustrated or described for one embodiment may be used on or in conjunction with other embodiments to yield yet another embodiment. It is intended that the present disclosure include such modifications and variations. The examples are described using specific language that should not be construed as limiting the scope of the appended claims.The drawings are not to scale and are for illustrative purposes only. Corresponding elements are designated by the same reference numerals throughout the various drawings unless otherwise indicated.

[0009] The terms "comprising," "containing," "include," "comprise," and the like are open-ended, and the terms indicate the presence of the specified structures, elements, or features, but do not preclude the presence of additional elements or features. The articles "a," "an," and "the" are intended to include both the plural and the singular, unless the context clearly indicates otherwise.

[0010] The term "electrically connected" describes a permanent, low-resistance connection between electrically connected elements, for example, a direct contact between the elements in question or a low-resistance connection via a metal and / or heavily doped semiconductor material. The term "electrically coupled" implies that one or more intermediate elements designed for signal and / or power transmission can be connected between the electrically coupled elements, for example, elements that are controllable to temporarily provide a low-resistance connection in a first state and a high-resistance electrical decoupling in a second state. An ohmic contact is a non-rectifying electrical junction with a linear or nearly linear current-voltage characteristic.

[0011] The figures illustrate relative doping concentrations by indicating "-" or "+" next to the doping type "n" or "p." For example, "n" indicates a doping concentration lower than the doping concentration of an "n" doping region, while an "n+" doping region has a higher doping concentration than an "n" doping region. Doping regions with the same relative doping concentration do not necessarily have the same absolute doping concentration. For example, two different "n" doping regions may have the same or different absolute doping concentrations.

[0012] Two adjacent doping regions with the same conductivity type and different dopant concentrations form a unipolar junction, e.g., an n / n+ or p / p+ junction along an interface between the two doping regions. At the unipolar junction, a dopant concentration profile orthogonal to the unipolar junction may exhibit a step or inflection point where the dopant concentration profile changes from concave to convex or vice versa.

[0013] Ranges specified for physical dimensions include the limiting values. For example, a range for a parameter y from a to b is read as a ≤ y ≤ b. The same applies to ranges with a limit such as "at most" and "at least."

[0014] The main components of a layer or structure made of a chemical compound or alloy are those elements whose atoms form the chemical compound or alloy. For example, nickel and silicon are the main components of a nickel silicide layer, and copper and aluminum are the main components of a copper-aluminum alloy.

[0015] The term "over" should not be interpreted to mean "directly upon." Rather, if an element is positioned "over" another element (e.g., a layer is "over" another layer or "over" a substrate), another component (e.g., another layer) may be positioned between the two elements (e.g., another layer may be positioned between a layer and a substrate if the layer is "over" the substrate).

[0016] With respect to structures and doped regions formed in a silicon carbide body, a second region is located "below" a first region when a minimum distance between the second region and a first surface at the front of the silicon carbide body is greater than a maximum distance between the first region and the first surface. The second region is located "directly below" the first region where the vertical projections of the first and second regions overlap into the first surface. The vertical projection is a projection orthogonal to the first surface.

[0017] Regions and / or structures can be laterally separated from each other within the same horizontal layer. Laterally separated regions and / or structures can also be vertically separated (i.e., positioned in different horizontal layers). In the latter case, orthogonal projections of the separated regions and / or structures into a horizontal projection plane are laterally separated. Regions and / or structures overlap laterally if orthogonal projections of the respective regions and / or structures into a horizontal projection plane overlap laterally.

[0018] The term “power semiconductor device” refers to semiconductor devices with a high voltage blocking capability of at least 30 V, for example 100 V, 600 V, 3.3 kV or more and with a nominal forward current or forward current of at least 1 A, for example 10 A or more.

[0019] According to one embodiment, a silicon carbide device may include a stripe-shaped trench gate structure extending from a first surface into a silicon carbide body. A lateral first direction is perpendicular to a longitudinal direction or main extension direction of the gate structure. A lateral second direction is perpendicular to the lateral first direction. The lateral first direction and the lateral second direction may both be aligned parallel to the two main surfaces of the silicon carbide body.

[0020] The silicon carbide body may have two substantially parallel major surfaces of the same shape and size and a lateral surface area connecting the edges of the two major surfaces. For example, the silicon carbide body may be a polygonal (e.g., rectangular or hexagonal) prism with or without rounded edges, or a cylinder. The silicon carbide body may have a surface extension along the two lateral directions and may have a thickness along a vertical direction perpendicular to the horizontal directions. The horizontal directions are also referred to as lateral directions. In other words, the vertical direction is perpendicular to the lateral first direction and the lateral second direction.

[0021] The material of the silicon carbide body can be, for example, 15R-SiC (silicon carbide of the 15R polytype) or a silicon carbide with a hexagonal polytype such as 2H-SiC, 4H-SiC, or 6H-SiC. In addition to the main components of silicon and carbon, the silicon carbide body can contain dopant atoms, for example, nitrogen (N), phosphorus (P), beryllium (Be), boron (B), aluminum (Al), and / or gallium (Ga). Furthermore, the silicon carbide body can contain undesirable impurities, for example, hydrogen, fluorine, and / or oxygen.

[0022] The stripe-shaped trench gate structure may extend from a first surface on a front side of the silicon carbide body into the silicon carbide body. The gate structure has a gate length along the lateral first direction and a gate width along the lateral second direction, which is orthogonal to the first direction.

[0023] The gate structure may include a conductive gate electrode. The gate structure may further include a gate dielectric formed between the gate electrode and the silicon carbide body. Two opposing first and second gate sidewalls may be vertical or may be slightly inclined to the vertical direction. The first and second gate sidewalls may be tapered or may be parallel.

[0024] In general, at least the first gate sidewall may extend substantially along a crystal plane of the silicon carbide body in which charge carrier mobility is high (e.g., one of the {11-20} or {1-100} crystal planes). The first gate sidewall may be an active sidewall, meaning a transistor channel may extend along the first gate sidewall. In some embodiments, the second gate sidewall may also be an active sidewall (e.g., in the case of parallel first and second gate sidewalls such as a vertical trench gate structure). In other embodiments (e.g., in the case of a tapered trench gate structure), the second gate sidewall may be an inactive sidewall. Viewed from the front side of the silicon carbide body, the first gate sidewall is located on a first side of the gate structure, and the second gate sidewall is located on an opposite second side of the gate structure.

[0025] A bottom surface at the bottom of the gate structure connects the first and second gate sidewalls via first and second bottom edges. The bottom surface may include a horizontal portion. The first gate sidewall may include a straight portion. The first bottom edge may connect a horizontal portion of the bottom surface and a straight portion of the first gate sidewall.

[0026] Accordingly, the second lower edge may connect a horizontal portion of the lower surface and a straight portion of the second gate sidewall. The first lower edge between the lower surface and the first gate sidewall may be sharp-edged or may be rounded and / or beveled (e.g., with an obtuse angle). The second lower edge between the lower surface and the second gate sidewall may be sharp-edged or may be rounded and / or beveled (e.g., with an obtuse angle).

[0027] The silicon carbide device may further include at least one source region. The at least one source region may be a doped region having a first conductivity type. The source region may be in contact with the first gate sidewall of the gate structure or may be in contact with a first gate sidewall of another gate structure. In other words, no source region, a single source region, or a plurality of source regions may be formed along the first gate sidewall of the gate structure. If no source region is formed along the first gate sidewall of the gate structure, at least one source region may be formed along another gate structure. Source regions formed along the same gate structure may be separated from each other along the first direction. A length of each source region along the first direction may be at least 500 nm, e.g., at least 1 µm.

[0028] The silicon carbide device may further include a shielding region. The shielding region may be a doped region having a second conductivity type. The first conductivity type and the second conductivity type are complementary conductivity types. The first conductivity type may be n-type, and the second conductivity type may be p-type. Alternatively, the first conductivity type may be p-type, and the second conductivity type may be n-type.

[0029] The shielding region is in contact with the first lower edge over at least 20% of the gate length of the gate structure. For example, the shielding region may extend along the entire length of the gate structure. The shielding region may extend at least partially along the first gate sidewall. According to other examples, the shielding region extends over at least 20% of the first gate sidewall. Along the first lower edge, the shielding region may be absent at least in portions vertically below the source regions. When the shielding region is in contact with the first lower edge, the shielding region may extend vertically along the first gate sidewall from the first surface to the first lower edge.

[0030] According to another example, the shielding region may extend over at least 30% of the distance between adjacent source regions.

[0031] The shielding region may further be in contact with the second gate sidewall, with the second lower edge, and / or the lower surface of the gate structure. The shielding region may be in contact with the second gate sidewall and the second lower edge over the entire gate length of the gate structure. The shielding region may be in contact with a fully shielded portion of the lower surface along the second lower edge over the entire gate length of the gate structure. The shielding region may be in contact with a partially shielded portion of the lower surface along the first lower edge in portions between adjacent source regions.

[0032] When the shielding region is in contact with a significant portion of the first lower edge, a significant portion of the gate structure may be fully embedded in the shielding region. Since the shielding region may shield the gate structure from a potential applied to a backside potential, e.g., the drain potential, a raised portion of the shielding region along the first gate sidewall may have a gate-to-drain capacitance C GD reduce. The shielding region may be electrically connected to a front-side potential, e.g., the source potential. In this case, a raised portion of the shielding portion relative to the source regions may have a gate-to-source capacitance C GS Increased C GS and reduced C GD significantly reduce the shutdown oscillation tendency.

[0033] The larger section of the shield region along the first surface can provide a larger contact area between the shield region and a front-side electrode formed on the first surface. The larger contact area can reduce the ohmic contact resistance between the front-side electrode and the shield region. In addition, the larger surface section of the shield region along the first surface can further facilitate the formation of reliable, low-resistance ohmic contacts between the shield region and the front-side electrode. As a result, the surge current robustness of a body diode forming the shield region with a drift structure can be significantly improved.The larger contact area and reduced ohmic resistance of the contact between the front electrode and the shielding section can also help reduce current overshoot during turn-on, reduce body diode losses, and / or reduce the turn-off oscillation tendency.

[0034] Forming the shielding region along the first gate sidewall can reduce the surface area of ​​the source regions and consequently the overall transistor channel width. The reduced transistor channel width, in combination with the improved shielding of the transistor channel by those portions of the shielding region formed between the source regions along the first direction, can contribute to reducing the transistor saturation current and thus improve short-circuit robustness. Additionally, the fully shielded portion of the bottom area is increased, and the incompletely shielded portion of the bottom area can be effectively shielded from all four lateral sides. Both effects can further contribute to increasing the reliability of the gate dielectric.

[0035] Since in high-voltage devices, for example, in devices with a voltage blocking capability of at least 600 V, for example, at least 3 kV, the resistance of a voltage-sustaining layer dominates the on-state losses, it is possible that a possibly slightly increased on-state resistance of the transistor channel may be negligible. On the other hand, the formation of the shielding region along a significant portion of the first gate sidewall may significantly mitigate the switching behavior, may improve the body-diode characteristics, and / or may increase short-circuit robustness. In particular, high-voltage devices with a voltage blocking capability of at least 600 V, for example, at least 3 kV, may benefit from a shielding region extending over a significant portion of the first gate sidewall.

[0036] According to one embodiment, the shielding region may be in contact with the first lower edge over at least 30%, for example at least 50%, of the gate length. The larger portion of the shielding region may cover the C GD further reduce, the C GS further increase and / or can further improve the reliability of the component.

[0037] According to one embodiment, the shielding region may include an upper shield portion and a deep shield portion. The upper shield portion is located between the first surface and the deep shield portion. The upper shield portion may be adjacent to (e.g., directly adjacent to) the first surface. A vertical extent of the upper shield portion may be greater than a vertical extent of the gate structure. The upper shield portion may, for example, be in contact with the first lower edge at least in places.

[0038] The upper shielding portion may be in contact with the second lower edge of the gate structure along the entire length of the gate structure. The deep shielding portion may be formed in a layer of the silicon carbide body between the lower surface of the gate structure and a second surface on the backside of the silicon carbide body.

[0039] A horizontal cross-sectional area of ​​the deep shield section may be the same or substantially the same as the horizontal cross-sectional area of ​​the upper shield section, and the same implantation mask may be used to form the deep shield section and the upper shield section. Alternatively, the horizontal cross-sectional areas of the upper shield section and the deep shield section may be significantly different. In the latter case, different implantation masks may define the deep shield section and the upper shield section.

[0040] The upper shielding section and the deep shielding section may be directly connected to each other along the vertical direction. The upper shielding section and the deep shielding section may overlap with each other in the sense that one or more peaks at the end of the range of implantations defining the deep shielding section may be located within the upper shielding section. The deep shielding section may be continuous along the vertical direction.

[0041] The deep shielding section can improve the shielding effect on the transistor channel and on those portions of the gate dielectric not directly embedded in the shielding region. Improved shielding of the transistor channel can reduce DIBL (drain-induced barrier depression).

[0042] The improved lateral shielding effect can enable sufficient shielding even with a comparatively small vertical extension of the shielding region, e.g., the deep shielding section. For example, the improved lateral shielding can at least partially compensate for a reduction in the vertical extension of the deep shielding section by omitting implantation(s) with implantation energy of more than 1.3 MeV. For example, a vertical distance between the gate bottom surface and a lower edge of the deep shielding section can be reduced to at least 50 nm, e.g., at least 300 nm.

[0043] According to one embodiment, a first distance between the upper shielding sections and the first gate sidewall may be smaller than a second distance between the deep shielding sections and the first gate sidewall. For example, a surface portion of the upper shielding section may directly adjoin the source region. The deep shielding section may have a lateral distance from the source regions along at least one lateral direction and / or may laterally overlap with the source region along at least one lateral direction.

[0044] According to one embodiment, the upper shielding section may include separation sections. The separation sections may be in contact with the first gate sidewall. The separation sections may extend downward from the first surface to the first lower edge. The separation sections may laterally separate source regions formed along the first direction along the gate structure. In this case, the upper shielding section may shield each transistor channel from all four lateral sides.

[0045] According to one embodiment, the upper shielding section may include separation sections. The separation sections may be located between the source regions. For example, the separation sections may be in contact with the first gate sidewall. Along the first surface, the separation sections and the source regions may cover a continuous portion of the first gate sidewall of the gate structure along the first direction. The separation sections and the source regions may completely cover the first gate sidewall along the first surface.

[0046] The separation sections and the source regions may have the same width along the second direction. Along the first surface, the separation sections of the upper shielding section and the source regions may complement each other to form a first contiguous region without gaps. The absence of further doped regions along the first gate sidewall or in the vicinity of the first gate sidewall may enable the formation of the upper shielding section and the source regions using comparatively simple photomasks.

[0047] According to one embodiment, the silicon carbide device may include a first gate structure and an adjacent second gate structure. Each of the first gate structure and the second gate structure may be embodied as the aforementioned gate structure. The first gate structure or the second gate structure may even correspond to the aforementioned gate structure.

[0048] The upper shield portion and the source regions associated with the first gate structure may be arranged between the first gate sidewall of the first gate structure and the second gate sidewall of the second gate structure.

[0049] In particular, it is possible that no further doped region having the conductivity type of the source region and being electrically connected to the voltage maintenance layer by an ohmic path is adjacent to the first surface in the region between the first gate structure and the second gate structure.

[0050] At the first surface, a region between the first gate sidewall and the second gate sidewall may be filled with the upper shielding portion and the source regions. In particular, the region between the first gate structure and the second gate structure may be completely filled with the exposed surfaces of the shielding portion and the source region.

[0051] In other words, between the first gate sidewall of the first gate structure and the second gate sidewall of the second gate structure, the upper shielding section and the source regions along the first surface may complement each other to form a second contiguous region. The second contiguous region includes the first contiguous region and a further strip-shaped section of the upper surface of the upper shielding section in the first surface. The absence of further doped regions between adjacent gate structures may further simplify the formation of the upper shielding section and the source regions.

[0052] According to one embodiment, the trench gate structure comprises a first section and a second section that are laterally offset from each other along the lateral first direction. A plurality of first sections and a plurality of second sections of the trench gate structure may be arranged in an alternating manner along the longitudinal or main extension direction of the gate structure, or in other words, the first lateral direction. For simplicity, the first and second sections are described in the singular below.

[0053] According to one embodiment, the first gate sidewall is in contact with the at least one source region within the first portion of the trench gate structure. An overlap between the first gate sidewall and the at least one source region along the lateral first direction may define the first portion of the trench gate structure. The first portion of the trench gate structure may be defined as the portion of the trench gate structure that laterally overlaps the at least one source region. The second portion of the trench gate structure may be defined as the portion of the trench gate structure that does not laterally overlap the at least one source region. The second portion of the trench gate structure may be arranged between the source regions along the lateral first direction.

[0054] According to one embodiment, in the first portion of the trench gate structure, the first gate sidewall extends to a first depth from the first surface along the vertical direction into the silicon carbide body, and in the second portion of the trench gate structure, the first gate sidewall extends to a second depth from the first surface along the vertical direction into the silicon carbide body, and the second depth is greater than the first depth.

[0055] According to one embodiment, in the first section of the trench gate structure, the first gate sidewall extends to a first depth from the first surface along the vertical direction into the silicon carbide body, and in the first section of the trench gate structure, the second gate sidewall extends to a third depth from the first surface along the vertical direction into the silicon carbide body, and the third depth is greater than the first depth.

[0056] In yet another embodiment, the first gate sidewall extends to a first depth from the first surface along the vertical direction into the silicon carbide body, and in the first portion of the trench gate structure, the second gate sidewall extends to a third depth from the first surface along the vertical direction into the silicon carbide body. Furthermore, in the second portion of the trench gate structure, the first gate sidewall extends to a second depth from the first surface along the vertical direction into the silicon carbide body. The second depth is greater than the first depth, and the third depth is greater than the first depth. The second depth may be substantially equal to or greater than the third depth.

[0057] The mentioned depths are defined as the distance from the first main surface. If the sidewalls merge into the bottom surface of the trench gate structure, the depth of the respective sidewall also defines the depth of a respective portion of the bottom surface of the trench gate structure adjacent to the respective sidewall. Therefore, different portions of the bottom surface have different depths. A first portion of the bottom surface, adjacent to the first gate sidewall in the first portion of the trench gate structure, may have a depth equal to the first depth. A second portion of the bottom surface, adjacent to the first gate sidewall in the second portion of the trench gate structure, may have a depth equal to the second depth. A third portion of the bottom surface, adjacent to the second gate sidewall in the first portion of the trench gate structure, may have a depth equal to the third depth.

[0058] In the second portion of the trench gate structure, the second gate sidewall may extend to the second depth from the first surface along the vertical direction into the silicon carbide body. Alternatively, in the second portion of the trench gate structure, the second gate sidewall may extend to a fourth depth from the first surface along the vertical direction into the silicon carbide body. The fourth depth may be greater than the second depth. A fourth portion of the bottom surface adjacent to the second gate sidewall in the second portion of the trench gate structure may have a fourth depth. The fourth depth may be equal to or greater than the second depth.

[0059] According to one embodiment, a lateral dopant profile along the first direction may include a plateau section through a transition between one of the isolation sections and one of the source regions. The plateau section may reflect that, along the first direction, the openings in an implantation mask for the source region implant may be narrower than the length of the mask columns of an implantation mask for the upper shield section implant. The length of the plateau sections corresponds to the difference between the extension of the mask columns for the upper shield section implant and the extension of the mask openings for the source region implant along the first direction. The length of each plateau section may be at least 50 nm, 200 nm, or even 500 nm.For example, in the plateau section, the dopant concentration may not change by more than an order of magnitude over a distance of 50 nm, 200 nm, or even 500 nm. In the plateau section, the conductivity type may be the conductivity type of the isolation section, the conductivity type of the source region, or it may be intrinsic. The plateau section may simplify the alignment of implantation masks for forming the source regions and / or isolation sections.

[0060] According to one embodiment, the silicon carbide device may include a first gate structure and an adjacent second gate structure. The first gate structure and the second gate structure may be embodied as the aforementioned gate structure. The first gate structure or the second gate structure may even correspond to the aforementioned gate structure.

[0061] The deep shielding portion may include a deep portion. Along the second direction, the deep portion may be laterally separated from the first gate sidewall of the first gate structure. Furthermore, each deep portion may laterally overlap with the second gate sidewall of the second gate structure.

[0062] For example, the silicon carbide device may include a plurality of gate structures. The deep shield portion may include a plurality of deep portions separated from each other at least along the second direction. Along the second direction, each deep portion may be laterally separated from the first gate sidewall of a first of two adjacent gate structures and may laterally overlap with the second gate sidewall of a second of the two adjacent gate structures.

[0063] According to one embodiment, the deep section may form a continuous strip with a longitudinal axis parallel to the first direction. The deep section may extend over at least 90% of the gate length or over the entire gate length.

[0064] For example, the silicon carbide device may include a plurality of gate structures and a plurality of deep sections. The deep sections may be laterally separated along the second direction. The deep sections may form contiguous stripes with longitudinal axes parallel to the first direction. For example, each deep section may extend over at least 90% of the gate length or over the entire gate length. With striped deep sections, it may be possible to form the deep sections without fine-tuning a photomask along the first direction. Each striped deep section may contribute to shielding transistor channels formed on opposite longitudinal sides of the striped deep sections.

[0065] According to another example, the deep section may include a plurality of deep subsections. The deep subsections may be laterally separated along the first direction. The deep subsections may be arranged in a matrix-like manner in rows and lines, wherein the lines may be orthogonal to the lines. Two of the deep subsections are formed on opposite sides of each source region along the second direction. In this way, each transistor channel may be shielded by at least two deep subsections formed on opposite lateral sides.

[0066] If the source regions associated with adjacent gate structures are offset from each other, e.g., by half the center-to-center distance between adjacent source regions, each transistor channel can be shielded by four deep subsections formed on four opposite lateral sides. The deep subsections can be omitted in regions where a shielding region has only a low or negligible shielding effect. With matrix-arranged deep subsections, it may be possible to maintain a high lateral spread of the forward current. Matrix-arranged deep subsections can minimize potential adverse effects of the deep subsections on other electrical properties of the silicon carbide device, e.g., the on-resistance.

[0067] According to another embodiment, a horizontal cross-section of the deep shielding section may include a grid with grid openings. Each grid opening may laterally enclose at least a portion of a source region. For example, each grid opening may laterally enclose an entire source region. According to another example, each grid opening may enclose a channel sidewall portion. For example, a horizontal cross-section of the deep shielding section may be more or less identical to the horizontal cross-section of the upper shielding section under a plurality of gate trenches and may be formed using the same implantation mask. The grid-shaped deep shielding section may enable a high shielding effect.

[0068] According to a further embodiment, the silicon carbide device may include a first gate structure and an adjacent second gate structure. The first gate structure and the second gate structure may be embodied as the aforementioned gate structure. The first gate structure or the second gate structure may even correspond to the aforementioned gate structure.

[0069] The shielding region may be in contact with the first lower edge along the gate length of the first gate structure. It is possible that alignment of the shielding region along the first direction may not be required. The manufacturing process can therefore be very effective.

[0070] According to one embodiment, the source region may extend along the gate length of the second gate structure. It is possible that alignment of the source region along the first direction may not be required. The manufacturing process can therefore be very efficient.

[0071] According to another embodiment, the silicon carbide device may include a body region having the second conductivity type and a current spreading region having the first conductivity type. The body region and the current spreading region may be formed in the silicon carbide body. The body region may separate the source region and the current spreading region.

[0072] In the on-state of the silicon carbide device, a transistor channel may be formed in the body region. The transistor channel may be an inversion layer formed along the gate structure. The inversion layer enables a unipolar charge carrier flow between the source region and the current spreading region in the on-state.

[0073] According to another embodiment, the silicon carbide device may include a drift structure between the gate structures and the second surface of the silicon carbide body. The drift structure may include a voltage maintenance structure. The voltage maintenance structure may include a lightly doped drift zone with a vertical extension of at least 4 µm, 12 µm, 20 µm, or even at least 100 µm. The minimum vertical extension may depend on the desired blocking capability of the silicon carbide device. For example, for a desired blocking capability of 650 V (or 1.2 kV or 1.7 V or 3.3 kV or 6.5 kV), the vertical extension of the drift zone may be at least 4 µm (or at least 8 µm or at least 12 µm or at least 20 µm or at least 40 µm, respectively). Alternatively or additionally, the stress maintenance structure may include a compensation structure, e.g., a superjunction structure.

[0074] The shield region and the drift structure may form a pn junction. The pn junction may be effective as a body diode or as part of a body diode that is in forward conduction mode when the silicon carbide device is reverse biased. The large contact area for the shield region along the first surface may enable low-resistance contact between the shield region and a front-side electrode and may improve the electrical properties of the body diode.

[0075] Fig. 1A-1D show a portion of a silicon carbide device 500 having a silicon carbide body 100. The silicon carbide device 500 may be or include, for example, an IGFET (Insulated Gate Field Effect Transistor), such as a MOSFET (Metal Oxide Semiconductor FET), or an MCD (MOS Controlled Diode), or an IGBT (Insulated Gate Bipolar Transistor).

[0076] The silicon carbide body 100 may include or consist of single-crystal silicon carbide, e.g., a silicon carbide crystal containing the primary components of silicon and carbon. The silicon carbide crystal may include undesirable impurities such as hydrogen and / or oxygen and / or intentional impurities, e.g., dopant atoms. The polytype of the silicon carbide crystal may be 15R or may be hexagonal, e.g., 2H, 6H, or 4H. The silicon carbide body 100 may include or consist of a silicon carbide layer grown by epitaxy.

[0077] A first surface 101 on a front side of the silicon carbide body 100 can be planar or ribbed. A mean plane of the first surface 101 extends along horizontal directions. The mean plane of a planar first surface 101 is identical to the planar first surface 101. In the case of a non-planar first surface 101, for example, in the case of a ribbed first surface 101, the mean plane can be a planar least-squares plane. The position and orientation of the planar least-squares plane are defined such that the sum of the squares of the deviations of surface points of the ribbed first surface 101 from the planar least-squares plane is minimum. A vertical direction 104 is orthogonal to the horizontal directions, e.g., parallel to the surface normal to the mean plane. The horizontal directions are also referred to below as lateral directions.

[0078] The vertical direction 104 may coincide with a main lattice direction or may be inclined to a main lattice direction by an off-axis angle, wherein the off-axis angle may be in a range of 2° to 8°, in particular 4°. At the rear side of the silicon carbide body 100, a second surface may extend parallel to a planar first surface 101 or parallel to the least-squares plane of a ribbed first surface 101.

[0079] Transistor cells TC are formed on the front side of the silicon carbide body 100. A drift structure 130 extends laterally through the silicon carbide body 100 between the transistor cells TC and the second surface. The drift structure 130 may include a voltage maintenance structure, e.g., a lightly doped drift zone, and / or a compensation structure, e.g., a superjunction structure.

[0080] Each transistor cell TC includes a source region 110 of a first conductivity type and a body region 120 of a second conductivity type. The body region 120 and the drift structure 130 form a first pn junction pn1. The body region 120 and the source region 110 form a second pn junction pn2. A vertical extension of the body region 120 corresponds to a channel length of the transistor cells TC and can be in a range from 0.2 µm to 1.5 µm.

[0081] Stripe-shaped trench gate structures 150 extend along a lateral first direction 291. At least one gate structure 150 is in contact with the source regions 110 and the body regions 120 of the transistor cells TC. The gate structures 150 include a conductive gate electrode 155, which may include or consist of a heavily doped polycrystalline silicon layer and / or a metal-containing layer. A gate dielectric 159 separates the gate electrode 155 from the silicon carbide body 100 along at least one side of the gate structure 150. The gate dielectric 159 may include or consist of thermally grown or deposited silicon oxide, silicon nitride, silicon oxynitride, another deposited dielectric material, or any combination thereof. A thickness of the gate dielectric 159 can be selected to obtain transistor cells TC with a threshold voltage in a range of 1.0 V to 8 V.The gate structures 150 may include only the gate electrode 155 and the gate dielectric 159 or may include further conductive and / or dielectric structures in addition to the gate electrode 155 and the gate dielectric 159.

[0082] The gate structures 150 may be evenly spaced and / or may have a uniform gate width w0. A center-to-center distance between adjacent gate structures 150 may be in a range from 0.5 µm to 10 µm, e.g., from 1 µm to 5 µm. A gate length L0 of the gate structures 150 may be up to several millimeters. A vertical gate extension v0 of the gate structures 150 may be in a range from 0.3 µm to 5 µm, e.g., in a range from 0.5 µm to 2 µm.

[0083] Opposite first and second gate sidewalls 151, 152 of each of the gate structures 150 may extend substantially along the vertical direction 104 or may be inclined by an inclination angle with respect to the vertical direction 104. In the latter case, the gate structures 150 may have parallel first and second gate sidewalls 151, 152, or the gate structures 150 may taper with increasing distance from the first surface 101. The inclination angle between the gate sidewalls 151, 152 and the vertical direction 104 may be selected according to the orientation of the crystal axes and / or according to the off-axis angle.

[0084] For example, the absolute value of the tilt angle between the first gate sidewall 151 and the vertical direction 104 may deviate from the absolute value of the off-axis angle by no more than ±1° (e.g., in the case of 4H-SiC, the tilt angle may range from at least 3° to at most 5°). However, the tilt angle may deviate from the off-axis angle in orientation. The tilt angle between the second gate sidewall 152, which is opposite the first gate sidewall 151, and the vertical direction 104 may be oriented equal to or opposite to the tilt angle of the first gate sidewall 151. The larger the tilt angle, the narrower a tapered gate structure 150 becomes starting from the first surface 101.

[0085] In general, at least the first gate sidewall 151 may extend substantially along a crystal plane of the silicon carbide body 100 in which charge carrier mobility is high (e.g., one of the {11-20} or {1-100} crystal planes). The first gate sidewall 151 may be an active sidewall, meaning that the transistor channel may extend along the first gate sidewall 151. In some embodiments, the second gate sidewall 152 may also be an active sidewall (e.g., in the case of a vertical trench gate structure 150). In other embodiments (e.g., in the case of a tapered trench gate structure 150), the second gate sidewall 152 may be an inactive sidewall.

[0086] A bottom surface 158 at the bottom of the gate structures 150 connects the first and second gate sidewalls 151, 152 or forms at least part of a connection between the first gate sidewall 151 and the second gate sidewall 152. The bottom surface 158 may include a horizontal section. The bottom surface 158 and the first gate sidewall 151 of each gate structure 151 may be connected via a first bottom edge 156. The bottom surface 158 and the second gate sidewall 152 of each gate structure 150 may be connected via a second bottom edge 157. The first bottom edge 156 may be sharp-edged or may be rounded and / or beveled. The second bottom edge 157 may be sharp-edged or may be rounded and / or beveled.

[0087] In Fig. 1A, a plurality of isolated (separated from each other) source regions 110 are formed in a portion of the silicon carbide body 100 between two adjacent gate structures 150. A lateral length L1 of the source regions 110 along the first direction 291 may be at least 500 nm, e.g., at least 1 µm.

[0088] Separation sections 161 of the shielding regions 160 may laterally separate adjacent source regions 110 at the first surface 101 along the first direction 291. A lateral width w2 of the separation sections 161 and a lateral width w1 of the source regions 110 may be the same. The separation sections 161 and the source regions 110 formed between the same two gate structures 150 may complement each other to form a first contiguous region 410 in the plane of the first surface 101. The first contiguous region 410 does not include a gap.

[0089] Fig. 1B shows a single source region 110 formed between two adjacent gate structures 150. The source region 110 may extend along the entire gate length L0 without any interruption. In other words, the source region 110 may extend from one longitudinal end of the gate structure 150 to the opposite longitudinal end. Along another gate structure 150 without a source region 110, the shield region 160 may extend along the entire gate length L0 without any interruption.

[0090] The shielding region 160 may extend along one or more further gate structures 150 from one longitudinal end of the respective gate structure 150 to the opposite longitudinal end.

[0091] Channel sidewall portions 153 of the first gate sidewalls 151 extend downward from the source regions 110 to the first lower edges 156.

[0092] The shielding region 160 extends from the first surface 101 into the silicon carbide body 100, as shown in Fig. 1C and Fig. 1D. The shielding region 160 may be in direct contact with the inactive second gate sidewalls 152 over the gate length L0 and over the vertical gate extension v0. Along the second gate sidewalls 152, the shielding region 160 extends downward from the first surface 101 to below the gate structure 150. A first vertical extension v1 of the shielding regions 160 is greater than the vertical gate extension v0. For example, a vertical distance v3 between the gate bottom surface 158 and a lower edge of the shielding region 160 may be at least 50 nm, e.g., at least 300 nm.

[0093] The shielding region 160 includes sections between the body regions 120 and the inactive second gate sidewalls 152. The body regions 120 and the shielding region 160 may form unipolar junctions. The shielding region 160 may include the separation sections 161 that separate adjacent source regions 120 along the first direction 291. The separation sections 161 are in direct contact with the first gate sidewalls 151 outside the channel sidewall sections 153.

[0094] A maximum dopant concentration in the shielding region 160 may be higher than a maximum dopant concentration in the body region 120. A vertical dopant concentration profile in the shielding region 160 may have a local maximum at a position below the gate structure 150. Along the second gate sidewall 152, a dopant concentration in the shielding region 160 may be higher, ie, at least ten times higher than a dopant concentration in the body region 120 along the first gate sidewalls 151.

[0095] In Fig. 2A-2D, the shielding region 160 includes an upper shielding portion 168 and a deep shielding portion 169. The deep shielding portion 169 and the upper shielding portion 168 are connected to each other along the vertical direction 104.

[0096] A vertical extension v2 of the upper shielding section 168 is greater than the vertical gate extension v0. The deep shielding section 169 includes a plurality of deep subsections 164 arranged along parallel rows and parallel rows, wherein the rows are orthogonal to the rows. Along the horizontal second direction 292, each deep subsection 164 is formed between two adjacent source regions 110. The deep subsections 164 may be laterally separated from the source regions 110. Along the horizontal first direction 291, adjacent deep subsections 164 of the same row are laterally separated. A horizontal length of the deep subsections 164 along the first direction 291 may be equal to, less than, or greater than a length of the source regions 110 along the first direction 291.

[0097] For example, the upper shielding portion 168 may enhance the lateral shielding of the gate dielectric and the transistor channel to such an extent that it is possible to symmetrically retract the deep subportions 164 along the first direction 291 with respect to the edges of the source regions 110, as illustrated. In this case, the horizontal length of the deep subportions 164 is smaller than the length of the source regions 110.

[0098] A first distance d1 between the upper shield portions and the first gate sidewalls 151 may be smaller than a second distance d2 between the deep shield portions 168 and the first gate sidewalls 151.

[0099] The drift structure 130 may include a drift zone 131 of the first conductivity type. The drift zone 131 forms a voltage-sustaining structure, wherein a vertical extension and a dopant concentration in the drift zone 131 may be selected such that the silicon carbide device 500 provides a nominal blocking voltage capability in a blocking state of the silicon carbide device 500. The drift zone 131 may be formed in a layer grown by epitaxy. An average net dopant concentration in the drift zone 131 may, for example, be in a range of 1E15 cm -3 up to 5E16 cm -3 According to another example, the drift structure 130 may include a compensation structure, e.g., a superjunction structure.

[0100] A heavily doped contact section 139 may be formed between the drift structure 130 and a backside electrode directly adjacent to the second surface 102 of the silicon carbide body 100. The heavily doped contact section 139 and the backside electrode form a low-resistance ohmic contact. The contact section 139 may have the same conductivity type as the drift zone 131, the opposite conductivity type, or may include zones of both conductivity types.

[0101] The drift structure 130 may include current spreading regions 137. The current spreading regions 137 may be formed between the body regions 120 and the voltage sustaining structure, e.g., the drift zone 131. The current spreading regions 137 may be in contact with the body regions 120. The current spreading regions 137 may laterally separate adjacent deep subsections 164. Portions of the current spreading regions 137 may be formed directly beneath the deep shielding sections 169. The current spreading regions 137 have a higher average dopant concentration than the drift zone 131 and may enable better lateral spreading of the forward current.

[0102] Implantation masks defining the deep subsections 164 and / or the source regions 110 may include rounded openings, e.g., oval or circular openings. Accordingly, the horizontal cross-sectional areas of the deep subsections 164 and / or the source regions 110 may include circular segments, oval segments, circles, and / or ovals. An implantation mask defining the upper shielding section 168 may include rounded columns, e.g., oval or circular columns. Accordingly, the horizontal cross-sectional areas of openings in the upper shielding section 168 may be circular segments, oval segments, circles, and / or ovals.

[0103] A first load electrode 310 on the front side of the silicon carbide body 100 is electrically connected to the source regions 110, the body regions 120, and the shield regions 160. The gate electrode 155 may be electrically connected to a gate metallization on the front side of the silicon carbide body 100. The gate metallization forms a gate terminal or is electrically connected or coupled to a gate terminal.

[0104] Portions of an interlayer dielectric 210 separate the first load electrode 310 and the gate electrode 155 in the gate structures 150. The first load electrode 310 may form a first load terminal, which may be an anode terminal of an MCD or a source terminal of a MOSFET, or may be electrically connected or coupled to a first load terminal.

[0105] A second load electrode 320 forms a low-resistance ohmic contact with the contact portion 139. The second load electrode 320 may form a second load terminal, which may be a cathode terminal of an MCD or a drain terminal of a MOSFET, or may be electrically connected or coupled to a second load terminal.

[0106] The illustrated silicon carbide device 500 is an n-channel SiC TMOSFET, wherein the first load electrode 310 forms a source terminal S or is electrically connected or coupled to a source terminal S, and wherein the second load electrode 320 forms a drain terminal D or is electrically connected or coupled to a drain terminal D. The transistor cells TC may be electrically connected in parallel.

[0107] In Fig. 3A-3D, deep sections 163 of the deep shielding section 169 form continuous strips with longitudinal axes parallel to the horizontal first direction 291.

[0108] In Fig. 4A-4D, a horizontal cross-sectional area of ​​the deep shield section 169 forms a grid with grid openings 167. Each grid opening 167 is formed around a source region 110.

[0109] Fig. 5A-5B show perspective views of a silicon carbide device 500 with source regions 110 formed along channel sidewall portions 153 of the first gate sidewalls 151. The deep shield portions 169 include point-shaped deep sub-portions 164. The shield regions 160 include a heavily doped contact portion 162 formed along the first surface 101.

[0110] Fig. 6A and Fig. 6B refer to silicon carbide devices 500 having a plurality of gate structures 150. A plurality of source regions 110 are formed along the first gate sidewalls 151 of the gate structures 150. Along the first surface 101, upper shield portions 168 of shield regions and the source regions 110 complement each other to form a second contiguous region 400 between two adjacent gate structures 150.

[0111] In Fig. 6A, the source regions 110 are arranged in a matrix-like manner in rows and lines, with the lines extending orthogonally to the lines. In other words, the source regions 110 are formed in the black and white squares of a checkerboard.

[0112] In Fig. 6B, the source regions 110 of adjacent gate structures 150 are shifted from each other by half the center-to-center distance between adjacent source regions 110 along the first direction 291. In other words, the source regions 110 may only be formed in the "white" squares of a checkerboard.

[0113] In Fig. 6C, each source region 110 extends along the entire gate length of gate structures 150 of a first type and is completely absent along the first gate sidewalls 151 of gate structures 150 of a second type. Gate structures 150 of the first type and the second type may form a regular pattern. For example, one, two, three, or more gate structures 150 of the first type (in contact with source regions 110) may be formed between each pair of gate structures 150 of the second type (without contact with source regions 110).

[0114] For illustrative purposes, various scenarios were described with respect to a silicon carbide device. Similar techniques can be implemented in semiconductor devices based on other types and types of compound semiconductor material for a semiconductor body, e.g., gallium nitride (GaN) or gallium arsenide (GaAs), etc.

[0115] Now, with reference to Fig. 7A to Fig. 7G describes a varying depth of the trench gate structure 150. Fig. Figure 7A shows a horizontal projection of a silicon carbide device 500. Fig. 7B to Fig. 7G show vertical sections along the Fig. 7A indicated cutting lines. Fig. 7C is a section along CC through a first portion 150-1 of the trench gate structure 150. Fig. 7D is a section along DD through a second portion 150-2 of the trench gate structure 150. Both sections CC and DD are parallel to the lateral second direction 292. Fig. 7E is a section along EE parallel to the trench gate structure 150 (and the first lateral direction 291) adjacent to the first sidewall 151, 151'. Fig. 7F and G are sections along FF and F'F', respectively, parallel to the trench gate structure 150 (and the first lateral direction 291) adjacent to the second sidewall 152, 152'. Fig. 7G is an alternative embodiment slightly different from Fig. 7B to Fig. 7F deviates.

[0116] The trench gate structure 150 includes first sections 150-1 and second sections 150-2 that are laterally offset from each other along the lateral first direction 291. The first sections 150-1 and second sections of the trench gate structure 150-2 are arranged in an alternating manner along the longitudinal direction of the trench gate structure 150.

[0117] A portion of the first gate sidewall 151' is in contact with the source regions 110 within the first portions 150-1 of the trench gate structure 150. An overlap between the first gate sidewall 151' and the source regions 110 along the lateral first direction 291 defines the first portion 150-1 of the trench gate structure 150. The first portion 150-1 of the trench gate structure 150 is only the portion of the trench gate structure 150 that laterally overlaps one of the source regions 110. The second portion 150-2 of the trench gate structure 150 is defined as the portion of the trench gate structure 150-2 that is not in lateral overlap with an adjacent source region 110. The second sections 150-2 are arranged between the source regions 110 along the lateral first direction 291.

[0118] In the first section 150-1 of the trench gate structure 150, the first gate sidewall 151' extends to a first depth t1 from the first surface 101 along the vertical direction 104 into the silicon carbide body 100. Also in the first section 150-1, the second gate sidewall 152' extends to a third depth t2 from the first surface 101 along the vertical direction 104 into the silicon carbide body 100. In the second section 150-2, the first gate sidewall 151 extends to a second depth t2* from the first surface 101 along the vertical direction 104 into the silicon carbide body 100. In the second section 150-2, the second gate sidewall 152 also extends to the second depth t2* from the first surface 101 along the vertical direction 104. into the silicon carbide body 100.

[0119] The second depth t2* is greater than the first depth t1, and the third depth t2 is greater than the first depth t1. The second depth t2* is greater than the third depth t2. Each of the depths t1, t2, t2* is defined as the shortest distance from the first main surface 101.

[0120] The sidewalls 151, 151', 152, 152' merge into the bottom surface 158, 158' of the trench gate structure 150 at the respective depths. Therefore, different sections of the bottom surface 158, 158' have different depths. A first portion of the bottom surface 158', adjacent to the first gate sidewall 151' within the first section 150-1, is at the first depth t1 from the first main surface 101. A second portion of the bottom surface 158, adjacent to the first gate sidewall 151 in the second section 150-2, is at the second depth t2* from the first main surface 101. A third portion of the bottom surface 158', adjacent to the second gate sidewall 152' in the first section 150-1, is at the third depth t2 from the first main surface 101. A fourth portion of the bottom surface 158, adjacent to the second gate sidewall 152 in the second section 150-2, is at a fourth depth t2** from the first main surface 101.The fourth depth t2** may be equal to the second depth t2* or greater than the second depth t2*. In the second section 150-2, the depth of the trench gate structure 150 may be at the second depth t2* (cf. ). Fig. 7D) be uniform if the fourth depth t2** is equal to the second depth t2*.

[0121] In the first section 150-1, the depth of the trench gate structure 150 goes from the first depth t1 to the third depth t2 (cf. Fig. 7C) between the first section of the floor surface 158' and the third section of the floor surface 158'. This results in a step between the first section of the floor surface 158' and the third section of the floor surface 158' (cf. Fig. 7C).

[0122] Beside the first gate sidewall 151, 151', the depth of the trench gate structure 150 extends from the first depth t1 to the second depth t2* at the boundary between the first section 150-1 and the second section 150-2 of the gate trench structure 150 (cf. Fig. 7E). This results in a step in the depth of the gate trench structure 150 between the first section 150-1 and the second section 150-2 adjacent to the first gate sidewall 151, 151' (cf. Fig. 7E). Within the first section 150-1, the deeper section (third section) of the bottom surface 158' as well as the deeper second sidewall 158' contribute to Cgs due to their proximity to the deep subsection 164 and the shielding section 168. The shallower section (first section) of the bottom surface 158' as well as the shallower second sidewall 158' contribute to Cgd. The step in the first section 150-1 also contributes to Cgs. Therefore, the shape of the gate trench structure in the first section 150-1 increases the Cgs to Cgd ratio compared to the embodiment of Fig. 3.

[0123] Beside the second gate sidewall 152, 152', the depth of the trench gate structure 150 extends from the second depth t2* to the third depth t2 at the boundary between the first section 150-1 and the second section 150-2 of the gate trench structure 150 (cf. Fig. 7F). This results in a step in the depth of the gate trench structure 150 between the first section 150-1 and the second section 150-2 adjacent to the first gate sidewall 152, 152' (cf. Fig. 7E). The step adjacent to the first gate sidewall 152, 152' is smaller than the step adjacent to the first gate sidewall 151, 151'. Within the second section 150-2, both sidewalls 151, 152 and the bottom surface 158 contribute to Cgs.

[0124] In an alternative embodiment, which is only available in Fig. As shown in Figure 7G, the second depth t2* and the third depth t2 are essentially the same. In this case, there is no step.

[0125] For illustrative purposes, various scenarios have been described with respect to a silicon carbide device without an integrated Schottky diode. Some embodiments may be combined with Schottky contacts between the front-side electrode and diode regions having the first conductivity type. The diode regions may extend between adjacent gate structures from the first surface to the drift structure. Alternatively or additionally, isolation sections may laterally separate Schottky contacts from source regions. For example, p-doped isolation sections may laterally separate n-doped diode regions from n-doped source regions.

Claims

[1] Silicon carbide device (500), comprising a strip-shaped trench-gate structure (150) extending from a first surface (101) into a silicon carbide body (100), wherein the gate structure (150) has a gate length (L0) along a lateral first direction (291), wherein the strip-shaped trench-gate structure (150) is bounded along a lateral second direction (292) by a first gate side wall (151, 151') of the gate structure (150) and a second gate side wall (152, 152') of the gate structure (150), wherein the first gate side wall (151, 151') and the second gate side wall (152) are connected via a lower surface (158) of the gate structure (150), and wherein the lateral second direction (292) is perpendicular to the lateral first direction (291), wherein the trench-gate structure (150) comprises a first section (150-1) and a second section (150-2) which are laterally offset from each other along the lateral first direction (291), wherein in the first section (150-1) the first gate sidewall (151') extends to a first depth (t1) from the first surface (101) along a vertical direction (104) into the silicon carbide body (100), wherein the vertical direction (104) is perpendicular to both the lateral first direction (291) and the lateral second direction (292), wherein in the second section (150-2) the first gate sidewall (151) extends to a second depth (t2*) from the first surface (101) along the vertical direction (104) into the silicon carbide body (100), and where the second depth (t2*) is greater than the first depth (t1). [2] Silicon carbide device according to the preceding claim, wherein in the first section (150-1) the second gate sidewall (152) extends to a third depth (t2) from the first surface (101) along the vertical direction (104) into the silicon carbide body (100), wherein the third depth (t2) is greater than the first depth (t1). [3] Silicon carbide device (500), comprising a strip-shaped trench-gate structure (150) extending from a first surface (101) into a silicon carbide body (100), wherein the gate structure (150) has a gate length (L0) along a lateral first direction (291), wherein the strip-shaped trench-gate structure (150) is bounded along a lateral second direction (292) by a first gate side wall (151, 151') of the trench-gate structure (150) and a second gate side wall (152, 152') of the trench-gate structure (150), wherein the first gate side wall (151, 151') and the second gate side wall (152) are connected via a lower surface (158) of the gate structure (150), and wherein the lateral second direction (292) is perpendicular to the lateral first direction (291), wherein the trench-gate structure (150) comprises a first section (150-1) and a second section (150-2) which are laterally offset from each other along the lateral first direction (291), wherein in the first section (150-1) the first gate sidewall (151') extends to a first depth (t1) from the first surface (101) along a vertical direction (104) into the silicon carbide body (100), wherein the vertical direction (104) is perpendicular to both the lateral first direction (291) and the lateral second direction (292), wherein in the first section (150-1) the second gate sidewall (152') extends to a third depth (t2) from the first surface (101) along the vertical direction (104) into the silicon carbide body (100), and where the third depth (t2) is greater than the first depth (t1). [4] Silicon carbide device according to one of the preceding claims 1 or 2, wherein the second depth (t2*) is greater than the third depth (t2). [5] Silicon carbide device according to one of the preceding claims 1 or 2, wherein the second depth (t2*) is equal to the third depth (t2). [6] Silicon carbide device according to one of the preceding claims, further comprising at least one source region (110) of a first conductivity type, wherein the first gate side wall (151') is in contact with the at least one source region (110) within the first section (150-1) of the trench-gate structure (150). [7] Silicon carbide device according to the preceding claim, wherein an overlap between the first gate sidewall (151') and the at least one source region (110) along the lateral first direction (291) defines the first section (150-1) of the trench-gate structure (150). [8] Silicon carbide device according to one of the preceding claims, further comprising a shielding area (160) of a second conductivity type, wherein the first gate side wall (151) is in contact with the shielding area (160) within the second section (150-2) of the trench-gate structure (150). [9] Silicon carbide device according to the preceding claim, wherein an overlap between the first gate side wall (151') and the shielding area (160) along the lateral first direction (291) defines the second section (150-2) of the trench-gate structure (150). [10] Silicon carbide device according to one of the preceding claims, wherein in a conducting state of the silicon carbide device a part of the load current of the silicon carbide device is conducted along the first gate side wall (151') only within the first section (150-1) of the trench-gate structure (150). [11] Silicon carbide device according to one of the preceding claims, further comprising a gate electrode (155) arranged between the first gate side wall (151, 151') and the second gate side wall (152, 152'). (L0).

Images