Patent Application
20190355819
The present disclosure relates to silicon carbide semiconductor devices with high voltage blocking capability.
Power semiconductor devices are typically used for transforming electric energy, for example in DC/AC converters, AC/AC converters or AC/DC converters, and for driving heavy inductive loads. In vertical power semiconductor devices a load current flows between a first load electrode at a front side and a second load electrode on the backside of a semiconductor die. The high dielectric breakdown field strength of silicon carbide compared to silicon results in silicon carbide semiconductor switches being significantly thinner than equivalent silicon devices for the same nominal blocking voltage. Accordingly, the on-state resistance of a silicon carbide semiconductor device can be significantly lower than the on-state resistance of a silicon device with the same voltage blocking capability.
There is a need for further improving the on-state resistance of silicon carbide devices of a given size.
An embodiment of the present disclosure relates to a semiconductor device that includes a gate electrode and a gate dielectric. The gate electrode extends from a first surface of a silicon carbide body into the silicon carbide body. The gate dielectric is positioned between the gate electrode and the silicon carbide body. The gate electrode includes a metal structure and a semiconductor layer between the metal structure and the gate dielectric.
Another embodiment of the present disclosure relates to a method of manufacturing a semiconductor device. The method includes providing a silicon carbide substrate, wherein a gate trench extends from a main surface of the silicon carbide substrate into the silicon carbide substrate. A gate dielectric is formed on at least one sidewall of the gate trench. A gate electrode is formed in the gate trench. The gate electrode includes a metal structure and a semiconductor layer between the metal structure and the gate dielectric.
The accompanying drawings are included to provide a further understanding the embodiments and are incorporated in and constitute a part of this specification. The drawings illustrate the embodiments of the semiconductor device and the method of manufacturing a semiconductor device and together with the description serve to explain principles of the embodiments. Further embodiments are described in the following detailed description and the claims.
In the following detailed description, reference is made to the accompanying drawings, which form a part hereof and in which are shown by way of illustrations specific embodiments in which the embodiments may be practiced. It is to 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 can be used on or in conjunction with other embodiments to yield yet a further embodiment. It is intended that the present disclosure includes such modifications and variations. The examples are described using specific language, which should not be construed as limiting the scope of the appending claims. The drawings are not scaled and are for illustrative purposes only. Corresponding elements are designated by the same reference signs in the different drawings if not stated otherwise.
The terms “having”, “containing”, “including”, “comprising” and the like are open, and the terms indicate the presence of stated structures, elements or features but do not preclude additional elements or features. The articles “a”, “an” and “the” are intended to include the plural as well as the singular, unless the context clearly indicates otherwise.
The term “electrically connected” describes a permanent low-resistive connection between electrically connected elements, for example a direct contact between the concerned elements or a low-resistive connection via a metal and/or heavily doped semiconductor material. The term “electrically coupled” includes that one or more intervening element(s) adapted for signal and/or power transmission may be between the electrically coupled elements, for example, elements that are controllable to temporarily provide a low-resistive connection in a first state and a high-resistive electric decoupling in a second state.
The Figures illustrate relative doping concentrations by indicating “−” or “+” next to the doping type “n” or “p”. For example, “n−” means a doping concentration which is 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 of 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.
Ranges given for physical dimensions include the boundary values. For example, a range for a parameter y from a to b reads as a≤y≤b. The same holds for ranges with one boundary value like “at most” and “at least”.
Main constituents of a layer or a structure from a chemical compound or alloy are such elements which atoms form the chemical compound or alloy. For example, nickel and silicon are the main constituents of a nickel silicide layer and copper and aluminum are the main constituents of a copper aluminum alloy. Main constituents of a layer or structure from a mixture of substances are those elements that shall have a defined ratio in the mixture. For example, the main constituents of a layer obtained by co-sputtering are the elements contained in the targets used for the co-sputtering. In addition to the main constituents, layers and structures may include unintentional impurities due to process imperfections as well as intentional additives such as dopants.
The terms metal and metallic as used in the context of this application do not include semi-metals. In particular, the terms metal and metallic do not include the elements boron, silicon, germanium, arsenic, antimony and/or tellurium. For example, the term metal includes transition metals and post-transition metals. Post-transition metals may include the chemical elements with the atomic numbers 13, 30 to 31, 48 to 50, and 80 to 84. Transition metals may include at least the chemical elements with the atomic numbers 21 to 29, 39 to 47, and 72 to 79. A metal structure includes at least a metallic portion and may include, in addition to the metallic portion, a further portion or further portions from non-metallic material or from semi-metallic material.
A conformal layer has a thickness that is substantially the same along an interface to a base on which the conformal layer is formed. A conformal layer may exhibit marginal thickness variations along edges, steps or other elements of the base but is nevertheless considered a conformal layer if the magnitude of the thickness variations is low compared to a mean thickness of the conformal layer. A conformal layer may be formed by thin-film deposition methods such as CVD (chemical vapor deposition), plating, or ALD (atomic layer deposition).
An embodiment relates to a semiconductor device with a gate electrode that may extend from a first surface of a silicon carbide body into the silicon carbide body. A gate dielectric may be positioned between the gate electrode and the silicon carbide body. The gate electrode may include a metal structure and a semiconductor layer, wherein the semiconductor layer is between the metal structure and the gate dielectric.
The first surface of the silicon carbide body may be a main surface of the silicon carbide body, along which the silicon carbide body extends. The first surface may be spanned by lateral directions. Perpendicular to the first surface, in a vertical direction, the silicon carbide body has a thickness which is small compared to the extension of the silicon carbide body in the lateral directions.
The metal structure may reduce the line resistance of the gate electrode. The metal structure may also improve uniformity of the switching behavior of the semiconductor device along stripe-shaped gate electrodes and distributes the on-state current more uniformly across a lateral extension of the semiconductor device.
The semiconductor layer may separate the metal structure from the gate dielectric such that a threshold voltage of the semiconductor device can be decoupled from the work function of the metal structure. The presence of the semiconductor layer may prevent a reduction of a dopant concentration in the body region, which in turn typically has to be compensated by a longer channel length to avoid drain induced barrier lowering (DIBL). It is further possible for the semiconductor layer to facilitate the use of a metal gate without inherently increasing the channel resistance as a consequence of a longer channel.
The semiconductor layer may further cover and protect the gate dielectric during forming of the metal structure. Said forming may include the application of aggressive, e.g. chlorine-based, precursors that may lead to defects in the gate dielectric.
According to an embodiment, the semiconductor layer may include at least one of doped polycrystalline silicon, e.g., n-doped polycrystalline silicon, and intrinsic polycrystalline silicon. The effective work function of the gate electrode and the doping in the body region contribute in defining the threshold voltage of the semiconductor device. Providing the semiconductor layer at least partly from heavily doped polycrystalline silicon may facilitate a comparatively high dopant concentration in the body region, wherein the high dopant concentration may result in comparatively low channel resistance. A doping concentration within the semiconductor layer, e.g. in the case of polycrystalline silicon, may be at least 1018 cm−3 and at most 2×1021 cm−3.
The semiconductor layer may include phosphorus atoms as unwanted impurities or as intended impurities, wherein a concentration of the phosphorus atoms in the semiconductor layer is at most 1019 cm−3 or at most 5×1018 cm−3 or at most 1018 cm−3. Such a low phosphorus content may be sufficient to allow for good conductivity of the semiconductor layer in the gate electrode.
For example, the semiconductor layer may be a comparatively heavily doped n-type polycrystalline silicon layer with a mean dopant concentration of at least 1020 cm−3, wherein the semiconductor layer contains arsenic atoms, sulfur atoms and/or selenium atoms. For example, the semiconductor layer contains no phosphorus atoms or only a small portion of phosphorus atoms at a phosphor concentration of at most 1019 cm−3 or at most 5×1018 cm−3 or at most 1018 cm−3.
According to another example, the semiconductor layer may be a comparatively weakly doped n-type polycrystalline silicon layer, wherein a total donor concentration in the semiconductor layer is at most 5×1019 cm−3 or at most 1019 cm−3 or at most 5×1018 cm−3. For example, a phosphor concentration in a weakly doped n-type polycrystalline silicon layer may be at most 5×1018 cm−3 or at most 1018 cm−3.
According to an alternative embodiment, the semiconductor layer may be intrinsic or approximately intrinsic with a net dopant concentration lower than the intrinsic carrier concentration.
According to an embodiment, the semiconductor layer may be or may include a p-type crystalline silicon layer. In particular, the semiconductor layer may be or may include a comparatively heavily doped p-type polycrystalline silicon layer with a mean dopant concentration of at least 1020 cm−3. In this case, it may be possible that the semiconductor layer exclusively contains acceptor atoms, e.g. boron atoms. Alternatively, the semiconductor layer may further contain donator atoms, e.g., phosphorus atoms. In the case of phosphorus atoms, a concentration of phosphorus atoms may then be at most 5×1018 cm−3 or at most 1018 cm−3. Agate electrode with p-doped semiconductor layer may be provided in semiconductor devices with n-channel transistor cells, in particular in semiconductor devices with n-doped source region and p-doped body region.
The metal structure may include a first layer which may be in direct contact with a semiconductor layer and which includes at least a first transition metal.
The first transition metal may be molybdenum, titanium or tantalum. In addition to the first transition metal, the first layer may include nitrogen. A first layer from, e.g. Ti, TiN, Mo, MoN, TaN may form bonds with the semiconductor layer and may provide sufficient adhesion of the metal structure on the semiconductor layer.
The metal structure may further include a fill structure formed on the first layer, wherein the first layer may separate the fill structure from the semiconductor layer. The fill structure may include a second transition metal. The second transition metal may differ from the first layer in at least one main constituent. For example, the second transition metal may be tungsten.
According to an embodiment, the semiconductor device may include a contact layer in contact with the silicon carbide body. The contact layer may be formed directly on the first surface. The material of the contact layer and the material of the first layer may be selected in accordance with an appropriate process that ensures integrity of the gate dielectric, the first layer, the contact layer and of an interlayer dielectric that separates the gate electrode from a load electrode.
According to an embodiment, the contact layer may include titanium, titanium nitride, tantalum and/or tantalum nitride and the first layer of the metal structure includes titanium, titanium nitride, tantalum and/or tantalum nitride. The contact layer and the first layer may be formed from the same material(s) or from different material(s). Both the contact layer and the first layer may be formed at a stage at which both contact regions of the silicon carbide body and the semiconductor layer are exposed. For example, the first layer and the contact layer may be formed contemporaneously without that the formation of the first layer invokes any restrictions for the formation of the contact layer and without that formation of the contact layer invokes any restrictions for the formation of the first layer.
According to another embodiment, the contact layer may include nickel aluminum (NiAl) and the first layer may include molybdenum and/or molybdenum nitride. The first layer may be formed and covered with an interlayer dielectric prior to the formation of the contact layer. Since the formation temperature of molybdenum silicide (MoSi) is higher than that typically applied for the formation of NiAl, the formation of a NiAl contact layer does not damage the previously interlayer dielectric, because the interlayer dielectric does not cover a metal that induces mechanical strain into the interlayer dielectric by silicidation.
According to an embodiment, a second layer may be formed between the first layer and the fill structure, wherein the second layer includes a third transition metal and wherein the second layer differs from the first layer in at least one main constituent.
According to embodiments, the silicon carbide body includes a drift structure, a source region and a body region. The drift structure includes at least a drift region. The drift region forms a voltage sustaining layer, wherein the vertical extension and the dopant concentration in the drift region are selected such that the semiconductor device reaches its nominal blocking voltage capability. The source region is positioned between the body region and the first surface. The body region forms a first pn junction with the drift structure and a second pn junction with the source region. The body region is in contact with the gate dielectric.
According to at least one embodiment, the semiconductor device includes a gate structure on the first surface of the silicon carbide body. The gate structure and the gate electrode may be electrically connected via at most two contact areas. The contact areas are spaced apart from each other along a first direction. The first direction may run parallel to the first surface. For example, the gate structure may form at most two contact areas with a stripe-shaped gate electrode. In this case, the contact areas of the stripe-shaped gate electrode are spaced from each other along the first direction, which may run perpendicular to a main extension direction of the stripe-shaped gate electrode. The gate structure may include a metal.
With the low-resistive metal gate electrode the semiconductor device may not require any additional contacts to the gate electrode between two opposite ends of a stripe-shaped gate electrode. Gate fingers that are typically formed above the first surface and that contact the stripe-shaped gate structures in a distance to both longitudinal ends of the gate structures may be omitted. With the omission of gate fingers, the active area of the semiconductor device and the area efficiency of the semiconductor device are increased.
In at least one embodiment, the semiconductor device comprises a first load electrode on and/or at the first surface of the silicon carbide body. The gate structure may be formed between the first load electrode and a lateral surface of the silicon carbide body. The lateral surface of the silicon carbide body may run along the vertical direction of the silicon carbide body.
According to at least one embodiment, the gate structure may include a gate pad and at least one gate conducting line connected (e.g. electrically connected) with the gate pad. The gate conducting line may form a closed frame or an incomplete frame around a transistor cell field. The gate conducting lines are in contact with each single one of a plurality of stripe-shaped gate electrodes. Each of the gate electrodes may have a longitudinal extension along a first direction. The gate electrodes are spaced from each other orthogonally to the first direction.
In at least one embodiment, the semiconductor device comprises a plurality of gate electrodes in the silicon carbide body. Each of the gate electrodes may have a longitudinal extension along the first direction. Further, the gate electrodes are spaced apart from each other orthogonally to the first direction. For example, the gate electrodes may be equally distributed along a second direction that runs perpendicular to the first direction.
Embodiments of the present disclosure further refer to a method of manufacturing a semiconductor device. At least some embodiments of the semiconductor device described herein may be manufactured with the method described herein. That is to say, features disclosed with respect to the semiconductor device are also disclosed for the method and vice versa.
In at least one embodiment of the method, a silicon carbide substrate is provided. A trench extends from a main surface of the silicon carbide substrate into the silicon carbide substrate. A gate dielectric is formed on at least one sidewall of the trench. A gate electrode in the trench includes a metal structure and a semiconductor layer between the metal structure and the gate dielectric.
According to at least one embodiment, formation of the gate electrode includes depositing a first metal layer on the semiconductor layer and on the main surface. From sections of the first metal layer a contact layer on the main surface and a first layer of the metal structure are formed. The first metal layer and the contact layer may be formed contemporaneously such that the formation of the first metal layer does not adversely affect a previously formed contact layer and formation of the contact layer cannot adversely affect a previously formed first metal layer. For example, the first metal layer and the contact layer may be formed from at least one of Ti, TiN, Ta, TaN.
According to at least one embodiment, the contact layer may be formed on a main surface and then (i.e., subsequently) a first layer of the metal structure is formed on the semiconductor layer. The contact layer can include a metal silicide and/or nickel aluminum. Since the contact layer is formed prior to the first layer, formation of the contact layer does not adversely affect a previously formed first layer. For example, the contact layer may include nickel and aluminum and the first layer may include at least one of Ti, TiN, Ta or TaN as main constituent(s).
According to at least one embodiment, a first layer of the metal structure is formed on the semiconductor layer and then (i.e., subsequently) a contact layer is formed on the main surface. The first layer may include a transition metal or a transition metal nitride with a silicidation temperature above the process temperature for a nickel aluminum contact layer. For example, the first layer may include molybdenum or molybdenum nitride and the contact layer may include nickel and aluminum. Forming the contact layer from molybdenum or molybdenum nitride facilitates a later formation of nickel aluminum as ohmic contact to doped regions in the silicon carbide body.
The semiconductor device 500 shown in
The semiconductor device 500 includes a silicon carbide body 100 which is based on a silicon carbide crystal including the main constituents silicon and carbon. The silicon carbide crystal may include unwanted impurities like hydrogen and oxygen and/or intended impurities, e.g., dopant atoms. The polytype of the silicon carbide crystal may be 2H, 6H, 15R or 4H, by way of example.
A first surface 101 at a front side of the silicon carbide body 100 may be planar or ripped. A surface normal 104 orthogonal to a planar first surface 101 or orthogonal to a mean plane of a ripped first surface 101 defines a vertical direction. Directions orthogonal to the surface normal 104 are horizontal and lateral directions.
The semiconductor device 500 may include a transistor cell TC with a trench gate structure 150 that extends from the first surface 101 into the silicon carbide body 100. The trench gate structure 150 includes a gate dielectric 490 and a conductive gate electrode 400.
The gate electrode 400 is electrically separated from the silicon carbide body 100. For example, the gate dielectric 490 may completely separate the gate electrode 400 from the silicon carbide body 100. According to other embodiments, one or more further dielectric structures with a material configuration different from the gate dielectric 490 and/or thicker than the gate dielectric 490 may be formed between the gate electrode 400 and the silicon carbide body 100.
The gate electrode 400 includes a metal structure 450 and a semiconductor layer 420, wherein the semiconductor layer 420 is positioned between the metal structure 450 and the gate dielectric 490. The semiconductor layer 420 may separate the metal structure 450 from the gate dielectric 490.
The semiconductor layer 420 may be or may include heavily doped polycrystalline silicon. For example, the semiconductor layer 420 includes n-doped or p-doped polycrystalline silicon. A dopant concentration in the semiconductor layer 420 may be at least 1019 cm−3. The metal structure 450 may be a homogeneous structure or may include one or more sublayers of different materials.
The source region 110 and the body region 120 directly adjoin a first sidewall of the trench gate structure 150. The source region 110 is between the body region 120 and the first surface 101. The body region 120 separates the source region 110 from a drift structure 130. The drift structure 130 is formed between the body region 120 and a second surface of the silicon carbide body 100, wherein the second surface is opposite to the first surface 101. 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 trench gate structure 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. Sidewalls of the trench gate structures 150 may be vertical or may tapper with increasing distance to the first surface 101. A width w1 of the trench gate structure 150 in the plane of the first surface 101 may be in a range from 500 nm to 5 μm, e.g., in a range from 1 μm to 3 μm.
The trench gate structure 150 includes a gate dielectric 490 that separates the gate electrode 400 at least from the body region 120. The gate dielectric 490 may include or consist of a semiconductor dielectric, for example, thermally grown or deposited semiconductor oxide, e.g., silicon oxide, a semiconductor nitride, e.g., deposited or thermally grown silicon nitride, and/or a semiconductor oxynitride, e.g., silicon oxynitride. Thickness and material configuration of the gate dielectric 490 may be selected to achieve a suitable threshold voltage for the transistor cell TC. According to an embodiment concerning n-channel transistor cells of the enhancement type, thickness and material configuration of the gate dielectric 490 may be selected to achieve a suitable threshold voltage in a range from 1.0 V to 8 V.
The gate electrode 400 may include a semiconductor layer 420 in contact with the gate dielectric 490. The semiconductor layer 420 may be a conformal layer and may have a thickness th1 in a range from 50 nm to 500 nm, for example from 100 nm to 300 nm.
The metal structure 450 may include a first layer 451 in contact with the semiconductor layer 420, a second layer 452 in contact with the first layer 451 and a fill structure 455. At least one of the first layer 451 and the second layer 452 may be a conformal layer. A total thickness of the first layer 451 and the second layer 452 may be in a range from 50 nm to 500 nm, for example in a range from 100 nm to 300 nm.
The first layer 451 may include a first transition metal, such as titanium, tantalum or molybdenum. For example, the first layer 451 may include a transition metal as the only main constituent. According to an embodiment the first layer 451 may include nitrogen as a second main constituent. The first load electrode 310 may include a contact layer 311 in direct contact with the source region 110 and the body region 120. The contact layer 311 may have the same material configuration as the first layer 451 or another material configuration. According to an embodiment the contact layer 311 may include nickel aluminum.
Gate electrodes 400 extend from the front side into the silicon carbide body 100. The gate electrodes 400 may have a horizontal longitudinal extension 11 along a horizontal first direction 191 and a width w1 along a horizontal second direction 192, wherein the length 11 may be at least 10 times the width w1. The gate electrodes 400 may extend at least from one side of the central region 610 to the opposite side and may extend into the edge region 690. A plurality of stripe-shaped gate electrodes 400 may be arranged along the second direction 192.
At the front side of the silicon carbide body 100 a gate structure 330 is in contact with the gate electrodes 400. The gate structure 330 may be electrically connected to each gate electrode 400 along one single contact area or along two contact areas, wherein the two contact areas are spaced from each other along the first direction 191.
In the embodiment of
The gate structure 330 may further include a gate pad 334. The gate conducting line 336 may be a comparatively thin metal line. The gate pad 334 may include a comparatively thick metallization. The gate pad 334 may be directly connected with the gate conducting line 336 or through a resistor, e.g. a polysilicon structure with an electric resistance of some few ohms.
A first load electrode 310 may include a comparatively thick metallization. The metallization for the gate pad 334 and the first load electrode 310 may be based on the same materials (e.g., may comprise the same materials or consist of the same materials) and may have the same thickness. The first load electrode 310 may vertically overlap with the gate conducting line 336, wherein an interlayer dielectric may electrically separate the first load electrode 310 and the gate conducting line 336. The first load electrode 310 and the gate pad 334 may form landing pads for bond wires or metal clips connecting the first load electrode 310 and the gate pad 334 with terminals of the semiconductor device 500.
According to an embodiment, the first surface 101 may be serrated and includes first and second surface sections. The first surface sections may be shifted to each other and may be tilted to a horizontal plane by the off-axis angle α. The second surface sections are tilted to the first surface sections and connect the first surface sections such that a cross-sectional line of the serrated first surface 101 approximates a saw-tooth line.
On the backside of the silicon carbide body 100 an opposite second surface 102 may extend parallel to the first surface 101. A surface normal 104 to the first surface 101 defines a vertical direction and directions parallel to the first surface 101 are horizontal directions.
The transistor cells TC are formed at the front side along the first surface 101. A drift structure 130 separates the transistor cells TC from a second surface 102 at the backside. The drift structure 130 may include a heavily doped contact structure 139 directly adjoining the second surface 102 and a lightly doped drift zone 131 between the transistor cells TC and the heavily doped contact structure 139.
The heavily doped contact structure 139 may be or may include a substrate portion obtained from a crystalline ingot and forms an ohmic contact with a second load electrode 320 that directly adjoins the second surface 102. A mean dopant concentration in the contact structure 139 is sufficiently high to ensure an ohmic contact with the second load electrode 320. In case the semiconductor device 500 is or includes an MCD or an IGFET, the contact structure 139 has the same conductivity type as the drift zone 131. In case the semiconductor device 500 is an IGBT, the contact structure 139 has the complementary conductivity type of the drift zone 131 or includes zones of both conductivity types.
The drift zone 131 may be formed in a layer grown by epitaxy on the contact structure 139. A mean net dopant concentration in the drift zone 131 may be in the range from 1E15 cm−3 to 5E16 cm−3. A vertical extension of the drift zone 131 is related to a nominal blocking capability of the semiconductor device 500. The vertical extension of the drift zone 131 may be in the range of about 1 μm to several 10 μm. The drift structure 130 may include further doped regions, for example field stop zones, barrier zones and/or current spread zones of the conductivity type of the drift zone 131 or counter-doped regions.
The drift zone 131 may directly adjoin the contact structure 139 or a buffer layer forming a unipolar homojunction with the drift zone 131 may be sandwiched between the drift zone 131 and the contact structure 139, wherein a vertical extension of the buffer layer may be approximately in a range from 1 μm to 10 μm and a mean dopant concentration in the buffer layer may be in a range from 3E17 cm−3 to 1E18 cm−3, by way of example. The buffer layer may relax mechanical stress in the silicon carbide body 100 and/or may contribute to shaping the electric field in the drift structure 130.
The transistor cells TC are formed along trench gate structures 150 that extend from the first surface 101 into the silicon carbide body 100. The silicon carbide body 100 includes semiconductor mesas 170 between neighboring trench gate structures 150.
A longitudinal extension of the trench gate structures 150 along a horizontal first direction 191 is greater than a width along a horizontal second direction 192 orthogonal to the first direction. The trench gate structures 150 may be long stripes extending from one side of a transistor cell region to an opposite side, wherein the length of the trench gate structures 150 may be up to several millimeters. According to other embodiments, a plurality of separated trench gate structures 150 may be formed along a line extending from one side of the transistor cell region to the opposite side, or the trench gate structures 150 may form a grid with the semiconductor mesas 170 formed in the meshes of the grid.
At the bottom, the trench gate structures 150 may be rounded, wherein a radius of the curvature may be at least twice the thickness of a gate dielectric 490 described in the following.
The trench gate structures 150 may be equally spaced, may have equal width, and/or may form a regular pattern. A center-to-center distance (also called pitch) of the trench gate structures 150 may be in a range from 1 μm to 10 μm, e.g., from 2 μm to 5 μm. A vertical extension of the trench 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.
The trench gate structures 150 may taper with increasing distance to the first surface 101. For example, a taper angle of the trench gate structures 150 with respect to the vertical direction may be equal to the off-axis angle α or may deviate from the off-axis angle α by not more than ±1 degree such that at least a first mesa sidewall 171 of two opposite longitudinal mesa sidewalls 171, 172 is formed by a main crystal plane with high charge carrier mobility, e.g., a {11-20} crystal plane. A second mesa sidewall 172 opposite to the first mesa sidewall 171 may be tilted to a main crystal plane by twice the off-axis angle α, e.g., by 4 degree or more, for example, by about 8 degrees. The first and second mesa sidewalls 171, 172 are on opposite longitudinal sides of the intermediate semiconductor mesa 170 and directly adjoin two different, neighboring trench gate structures 150.
The trench gate structure 150 includes a conductive gate electrode 400 and a gate dielectric 490 as described herein. In the example shown in
The gate dielectric 490, the semiconductor layer 420, the first layer 451, and/or (if applicable) the second layer 452 may entirely cover the trench sidewalls and the trench bottom.
The semiconductor mesas 170 include source regions 110 that are positioned at the front side. The source regions 110 may directly adjoin the first surface 101 and may directly adjoin both longitudinal first and second mesa sidewalls 171, 172 of the respective semiconductor mesa 170.
The semiconductor mesas 170 further include body regions 120 that separate the source regions 110 from the drift structure 130. The body regions 120 form first pn junctions pn1 with the drift structure 130 and second pn junctions pn2 with the source regions 110. The body regions 120 directly adjoin the first mesa sidewall 171. A vertical extension of the body regions 120 corresponds to a channel length of the transistor cells TC and may be in a range from 0.2 μm to 1.5 μm. The source regions 110 and the body regions 120 are electrically connected to a first load electrode 310 at the front side.
The first load electrode 310 may form a first load terminal L1 or may be electrically connected or coupled to a first load terminal L1, which may be an anode terminal of an MCD, a source terminal of an IGFET or an emitter terminal of an IGBT. The second load electrode 320 may form a second load terminal L2 or may be electrically connected or coupled to a second load terminal L2, which may be a cathode terminal of an MCD, a drain terminal of an IGFET or a collector terminal of an IGBT.
Shielding regions 160 may separate the body regions 120 and the second mesa sidewalls 172. A dopant concentration in the shielding regions 160 along the second mesa sidewalls 172 may be higher, e.g., at least ten times higher than a dopant concentration in the body regions 120 along the first mesa sidewalls 171.
According to an embodiment, the transistor cells TC are n-channel FET cells with p-doped body regions 120, n-doped source regions 110 and n-doped drift zone 131. According to another embodiment, the transistor cells TC are p-channel FET cells with n-doped body regions 120, p-doped source regions 110 and p-doped drift zone 131.
The gate dielectric 490 capacitively couples portions of the body regions 120 to the gate electrode 400. When a potential at the gate electrode 400 exceeds or falls below a threshold voltage of the semiconductor device 500, the electric field effects that minority charge carriers in the body regions 120 form inversion channels along the gate dielectric 490, wherein the inversion channels connect the source regions 110 with the drift structure 130, thereby turning on the semiconductor device 500. In the on-state, a load current flows through the silicon carbide body 100 approximately along the first mesa sidewalls 171 between the first and second load electrodes 310, 320. At the same time the higher dopant concentration in the shielding regions 160 may suppress the formation of inversion channels along the second mesa sidewalls 172.
The drift structure 130 may include current spread regions 137 that directly adjoin the body regions 120 and that may extend between neighboring shielding regions 160.
The first load electrode 310 may include a contact layer 311 and a main metal structure 315. The contact layer 311 may include a transition metal or a transition metal nitride, for example Ti, TiN, Ta, TaN, Mo, MoN or a metal aluminum compound, for example NiAl with or without silicided portion.
The semiconductor device 500 illustrated in
A first load electrode 310 may include trench contacts 316 that may extend between neighboring trench gate structures 150 from the first load electrode 310 through openings of an interlayer dielectric 210 into the semiconductor mesas. The first load electrode 310 is electrically connected to the source regions 110 and to the body regions 120.
The transistor cells TC may include shielding regions 160 having the conductivity type of the body regions 120. The shielding regions 160 may have a higher dopant concentration than the body regions 120 and may be laterally spaced from the trench gate structures 150. Two adjoining shielding regions 160 may shield the trench gate structure 150 formed between the two shielding regions 160 against a potential applied at a rear side electrode (which may correspond to the second load electrode 320). A vertical extension of the shielding regions 160 may be greater than a vertical extension of the trench gate structures 150. For example, a local dopant concentration maximum in the shielding regions 160 may have a greater distance to the first surface 101 than a bottom of the trench gate structures 150.
The trench gate structures 150 may include a gate electrode 400 as described above with reference to
The semiconductor device 500 of
Semiconductor mesas 170 between neighboring trench gate structures 150 include source regions 110 formed along the first surface 101 and body regions 120 that separate the source regions 110 from a drift structure 130. The body regions 120 may include heavily doped contact zones 149, wherein a mean dopant concentration in the contact zones 149 is significantly higher than in a main zone 121 of the body region 120 outside the contact zone 149. Sidewalls of the semiconductor mesas 170 may be, for example (0-33-8) crystal planes. As regards details of the trench gate structures 150 and the first load electrode 310 reference is made to the description in
In
Gate trenches 750 are formed in a silicon carbide substrate 700. For example, a trench etch mask formed on a main surface 701 of the silicon carbide substrate 700 may mask an anisotropic etch process such as reactive ion etching. The etch process may selectively remove portions of the silicon carbide substrate 700 that are exposed at the main surface 701, thus forming gate trenches 750. The trench etch mask may be removed after the etch process.
Before or after removal of the trench etch mask, a gate dielectric layer 491 may be formed on exposed portions of the silicon carbide substrate 700 and/or in the gate trench 750. Formation of the gate dielectric layer 491 may include a thermal treatment in an atmosphere containing oxygen and/or nitrogen. Alternatively or in addition, formation of the gate dielectric layer 491 may include deposition of a dielectric material, for example, by CVD (chemical vapor deposition).
The gate trenches 750 may be stripe-shaped with a horizontal longitudinal extension orthogonal to the cross-sectional plane. The gate trenches 750 may run parallel to each other and may be equally spaced.
The gate dielectric layer 491 may include at least one of thermally grown silicon oxide, thermally grown silicon nitride, thermally grown siliconoxinitride, and deposited silicon oxide, e.g., CVD silicon oxide, by way of example. The gate dielectric layer 491 may line the gate trenches 750 at approximately uniform thickness and may cover the main surface 701.
A deposition process, such as, for example, CVD, may be used to deposit n-doped and/or p-doped polycrystalline semiconductor material 425 that fills the gate trenches 750 (see
A first resist layer may be deposited and patterned by photolithography to form a first resist mask 611 with first resist mask openings 615 in the vertical projection of central portions of the gate trenches 750. A time-controlled anisotropic etch process may remove portions of the first interlayer dielectric portion 202 and of the polycrystalline semiconductor material 425 in the gate trenches 750 in the vertical projection of the first resist mask openings 615.
According to an embodiment, the formation of the auxiliary trenches 755 may be combined with the process for exposing the contact area 702. For example, one single resist mask may be formed that includes first openings in the vertical projection of the central portions of the gate trenches 750 and second openings in the vertical projection of the contact area 702.
The first resist mask 611 may be removed. A first metal layer 810 may be formed by depositing metal-containing material, e.g., by sputtering and/or vapor deposition. A metal fill layer 815 may be deposited on the first metal layer 810.
According to
A portion of the metal fill layer 815 deposited above the first interlayer dielectric portion 202 may be removed, e.g., by a recess such as a time-controlled and/or endpoint-controlled wet etching, wherein the recess may stop with or after exposure of the first metal layer 810 on the first interlayer dielectric portion 202.
In the method step shown in
In the method step shown in
In the method step shown in
The method of
The method according to the embodiment of
The method described with respect to
Gate trenches 750 and a gate dielectric layer 491 may be formed as described with reference to
Along an exposed surface of the semiconductor layer 420 (e.g., a surface that is opposite the gate dielectric 490), the semiconductor layer 420 may be oxidized to form a sacrificial oxide layer 471 on the semiconductor layer 420. A first sacrificial material 472 may be deposited that fills the gate trenches 750. Portions of the first sacrificial material 472 outside the gate trenches 750 may be removed, for example, by wet etching. A second sacrificial material 475 may be deposited and patterned by photolithography to expose the contact area 702.
Pre-silicidation of the exposed contact area 702 may include, for example, a heating treatment at a temperature of at least 700° C. The remaining portions of the second sacrificial material 475 may be removed to expose the first sacrificial material 472 in the gate trenches 750.
In a further method step, a first metal layer and a metal fill layer may be deposited and recessed (not shown in
The gate trenches 750 in
An exposed portion of the semiconductor layer 420 between the contact area 702 and the metal structures 450 may be partly removed, for example, by a time-controlled wet etch process.
As shown in the exemplary embodiment of
Alternatively, the contact metal layer 830 may be deposited prior to removal of a resist mask that patterns the interlayer dielectric layer 205.
The formation of the contact layer 311 according to
The following methods use nickel aluminum for forming a contact layer between a load electrode and the silicon carbide substrate. Nickel aluminum provides low-ohmic and reliable contacts to n-type and p-type regions in a silicon carbide body. Formation of nickel aluminum contacts typically involves high temperature processes.
According to
The method of
The method steps shown in
Accordingly, in
A photoresist layer may be deposited on the interlayer dielectric layer of
The third resist mask 630 is removed and lifts off the second portion of the nickel aluminum layer 840. A heating treatment may silicide the remaining first portion of the nickel aluminum layer 840 in the contact area 702 of the silicon carbide substrate 700. Non-reacted constituents of the nickel aluminum layer 840 may be removed and a second thermal treatment may be performed, for example a second RTP with a minimum temperature of at least 750° C. and at most 1100° C. A main metal layer may be deposited and patterned as described above.
The temperatures applied for the formation of the contact layer 311 are below the formation temperature of molybdenum silicide. Therefore formation of the contact layer 311 does not involve a silicidation in the trench gate structure 150 that may damage the interlayer dielectric 210.
The trench gate structure 150 includes a gate electrode 400 and a gate dielectric 490. In addition to the gate electrode 400 and the gate dielectric 490, the trench gate structure 150 may include further insulating and further conductive structures, wherein the further conductive structures may be electrically separated from the gate electrode 400.
The gate electrode 400 may include a semiconductor layer 420, which is in contact with the gate dielectric 490. The semiconductor layer 420 may be a conformal layer with a thickness in a range from 50 nm to 500 nm, for example from 100 nm to 300 nm or may completely fill the space between two portions of the gate dielectric 490 which are formed on opposite sidewalls of the trench gate structure 150. The trench gate structure 150 and the gate electrode 400 may be devoid of metallic structures.
The semiconductor layer 420 may include phosphorus atoms as unwanted impurities or as intended impurities, wherein a concentration of the phosphorus atoms in the semiconductor layer 420 is at most 1×1019 cm−3 or at most 5×1018 cm−3 or at most 1018 cm−3. A lower phosphorus content in the semiconductor layer 420 may mitigate negative effects induced by the presence of phosphorus.
For example, the semiconductor layer 420 may be a comparatively heavily doped n-type polycrystalline silicon layer with a mean dopant concentration of at least 1020 cm−3, wherein the semiconductor layer 420 contains arsenic atoms, sulfur atoms and/or selenium atoms. The semiconductor layer 420 may contain no phosphorus atoms or only a small portion of phosphorus atoms at a phosphor concentration of at most 5×1018 cm−3 or at most 1018 cm−3.
According to an embodiment, the semiconductor layer 420 may be a comparatively heavily doped p-type polycrystalline silicon layer with a mean dopant concentration of at least 1020 cm−3, wherein the semiconductor layer 420 may exclusively contain acceptor atoms, e.g. boron atoms. Alternatively, the semiconductor layer 420 may further contain donator atoms, e.g., phosphorus atoms, wherein a concentration of the phosphorus atoms is at most 5×1018 cm−3 or at most 1018 cm−3.
The gate electrode 400 with p-doped semiconductor layer 420 may be combined with an n-channel transistor cell TC with n-doped source region 110 and p-doped body region 120. Although specific embodiments have been illustrated and described herein, it will be appreciated by those of ordinary skill in the art that a variety of alternate and/or equivalent implementations may be substituted for the specific embodiments shown and described without departing from the scope of the present disclosure. This application is intended to cover any adaptations or variations of the specific embodiments discussed herein. Therefore, it is intended that this disclosure be limited only by the claims and the equivalents thereof.
| Number | Date | Country | Kind |
|---|---|---|---|
| 102018111653.5 | May 2018 | DE | national |
| 102019109368.6 | Apr 2019 | DE | national |
