Power Semiconductor Devices

Abstract

The power semiconductor device (1) comprises a gate commutated thyristor cell (20) including a cathode electrode (2), a cathode region (9) of a first conductivity type, a base layer (8) of a second conductivity type, a drift layer (7) of the first conductivity type, an anode layer (5) of a second conductivity type, an anode electrode (3), and a gate electrode (4). The base layer (8) comprises a cathode base region (81) having a first depth (D1) disposed between the cathode region (9) and the drift layer (7), a gate base region (82) having a second depth (D2) disposed between the gate electrode (4) and the drift layer (7), and an intermediate base region (83) having a third depth (D3) of two different values ​​between the first depth (D1) and the second depth (D2) disposed between the cathode base region (81) and the gate base region (82).

Description

[Technical field]

[0001] The present disclosure relates to power semiconductor devices. The present disclosure relates to the field of power semiconductor devices, and more particularly, but not exclusively, to high voltage power semiconductor devices. The power semiconductor device includes, for example, a four-layer npnp active thyristor structure that can be turned off via a gate. The power semiconductor device can include one gate commutated thyristor cell or multiple gate commutated thyristor cells. [Background technology]

[0002] Document US7816706B2 describes a power semiconductor device. The publication "High Temperature Operation of HPT+IGCTs", M. Arnold et al., proceedings of PCIM Europe, May 17-19, 2011, Germany, concerns integrated gate-commutated thyristors, abbreviated as IGCT.

[0003] Document CN108242465A describes gate commutated thyristors and how they are supplied. Document "The Corrugated P-Base IGCT-a New Benchmark for Large Area SQA Scaling" by TOBIAS WIKSTROM et al., POWER SEMICONDUCTOR DEVICES AND IC'S, 2007. ISPSD'07.19 THINTERNATIONAL SYMPOSIUM ON,IEEE,PI,1 May2007(2007-05-01),pp.29-32,XP031129660,ISBN:978-1-4244-1 095-8 describes an IGCT with exceptional Safe Operating Area (SOA). Document EP2622639B1 describes a reverse conducting power semiconductor device. Document CN104795439B describes a gate commutated thyristor chip. Summary of the Invention [Means for solving the problem]

[0004] SUMMARY OF THE DISCLOSURE Embodiments of the present disclosure are directed to providing power semiconductor devices with improved turn-off performance.

[0005] According to an embodiment, a power semiconductor device includes a gate-commutated thyristor cell, a first main surface, and a second main surface opposite to the first main surface. The gate-commutated thyristor cell includes, in order from the first main surface to the second main surface, a cathode electrode disposed on the first main surface, a cathode region of a first conductivity type, a base layer of a second conductivity type forming a first junction to the cathode region, a drift layer of the first conductivity type forming a second junction with the base layer, an anode layer of a second conductivity type, and an anode electrode disposed on the second main surface. The second conductivity type is different from the first conductivity type. The gate-commutated thyristor cell further includes a gate electrode disposed laterally with respect to the cathode region. The base layer comprises a cathode base region having a first depth disposed between the cathode region and the drift layer, a gate base region having a second depth disposed between the gate electrode and the drift layer, and an intermediate base region having a third depth of at least two different values ​​disposed between the cathode base region and the gate base region, each of the at least two values ​​being deeper than the first depth and shallower than the second depth.

[0006] A gate commutated thyristor cell advantageously includes three or more distinct base regions: the cathode base region can be optimized for best on-state performance, the gate base region can be configured for best turn-off performance, and the intermediate base region provides an extra degree of freedom in optimization, essentially allowing separate optimization of the gate and cathode base regions, so that the gate commutated thyristor cell achieves both a lower on-state voltage and a large turn-off current capability.

[0007] "Depth" is understood herein as the thickness measured in the direction from the first major surface to the second major surface.

[0008] The first major surface and / or the second major surface may be flat within manufacturing tolerance limits.

[0009] The second junction may have the following shapes: Moving laterally above the gate base region, i.e. parallel to the second major surface, the second junction may be located at a constant height relative to the second major surface; Moving laterally above the intermediate base region away from the gate base region towards the cathode base region, the height of the second junction may increase monotonically, e.g. strictly monotonically; Moving above the cathode base region, the height of the second junction may always be higher than in the gate base region and / or intermediate region.

[0010] According to another embodiment, the cathode electrode forms an ohmic contact with the cathode region. The anode electrode forms an ohmic contact with the anode layer. The gate electrode forms an ohmic contact with the base layer. The anode layer is separated from the base layer by a drift layer. The first junction and the second junction are realized as pn junctions. The base layer may include other regions.

[0011] According to another embodiment, the intermediate base region includes a region between the gap between the gate electrode and the cathode region on one side and the second junction on the other side. A third depth of two different values ​​may be present in this region. For example, the third depth of two different values ​​does not overlap the cathode electrode and the gate electrode in the lateral direction from the gate base region to the cathode base region.

[0012] According to another embodiment, the intermediate base region further includes at least one of an adjacent region between the gate electrode and the second junction and an adjacent region between the cathode region and the second junction.

[0013] According to other embodiments, the maximum or average doping concentration of the gate base region is higher than the maximum or average doping concentration of the intermediate base region. The maximum or average doping concentration of the intermediate base region is higher than the maximum or average doping concentration of the cathode base region. "Higher" can mean at least two times or at least ten times.

[0014] According to another embodiment, the doping concentration (average or maximum) of the intermediate base region and / or the gate base region at the first major surface, or the gate base region in the area closest to the first major surface, is higher than the doping concentration (maximum or average) of the cathode base region at the first junction and / or the cathode base region in the area closest to the first major surface.

[0015] According to another embodiment, the first depth is a minimum depth of the cathode base region and the second depth is a maximum depth of the gate base region.

[0016] According to another embodiment, the first depth is the shortest distance of the cathode region to the drift layer, and the second depth is the longest distance of the gate electrode to the drift layer.

[0017] According to another embodiment, each of the third depth values ​​is greater than the first depth and less than the second depth. Advantageously, the intermediate base region provides a smooth transition from the cathode base region to the gate base region.

[0018] According to another embodiment, the portion of the second junction located between the intermediate base region and the drift layer includes a first base transition region forming a step, a second base transition region forming another step, and a plane between the first and second base transition regions, where the plane is, for example, a portion running parallel to the first and / or second major surfaces.

[0019] According to another embodiment, the cathode region has a shape of the group including a circle, a rectangle, an octagon and a hexagon on the first main surface. This shape is visible in a top view of the power semiconductor device. Advantageously, the shape of the cathode region or cathode electrode allows flexible design for gates with low resistance and inductance, including improved performance and more degrees of freedom for design automation.

[0020] According to another embodiment, the gate electrode surrounds the cathode region at the first major surface. According to another embodiment, the anode layer includes a gate anode region and a cathode node region. The maximum doping concentration of the gate anode region is less than the maximum doping concentration of the cathode node region. The gate anode region is adjacent to the gate electrode. The cathode node region is adjacent to the cathode region. Thus, the anode layer is divided into two distinct regions or areas. The cathode node region is vertically disposed below the cathode region and the gate anode region is vertically disposed below the gate electrode. The gate anode region below the gate has a lower doping concentration, e.g. to reduce the PNP transistor gain.

[0021] According to another embodiment, the gate-commutated thyristor cell comprises a buffer layer of a first conductivity type. The buffer layer is arranged between the drift layer and the anode layer. The buffer layer includes, for example, a gate buffer region and a cathode buffer region. The maximum doping concentration of the gate buffer region is higher than the maximum doping concentration of the cathode buffer region. The buffer layer is therefore divided into two distinct regions or areas. The cathode buffer region is vertically arranged below the cathode region and the gate buffer region is vertically arranged below the gate electrode. The gate buffer region below the gate has a higher doping concentration, for example to reduce the PNP transistor gain.

[0022] According to another embodiment, the gate commutated thyristor cell is realized as an integrated gate commutated thyristor cell, abbreviated IGCT cell. The gate commutated thyristor cell can be turned off by a signal provided to the gate electrode. The gate commutated thyristor cell can be realized as a member of a group including an asymmetric integrated gate commutated thyristor cell (abbreviated asymmetric IGCT), a reverse conducting integrated gate commutated thyristor cell (RC-IGCT cell), a reverse blocking integrated gate commutated thyristor cell (RB-IGCT cell) and a bidirectional turn-off thyristor cell (BTT cell). The proposed gate-cathode design concept can thus be applied to different GCT families such as asymmetric IGCT, reverse conducting IGCT, reverse blocking IGCT and BTT. The RB-IGCT cell is configured to block in both forward and reverse directions.

[0023] According to an embodiment, the power semiconductor device comprises: - providing a drift layer of a first conductivity type in a semiconductor substrate; - providing a cathode region of a first conductivity type; - providing an anode layer of a second conductivity type; - providing a cathode electrode on the first major surface and providing a gate electrode on the first major surface laterally relative to the cathode region; - providing an anode electrode on the second major surface; - producing a base layer of a second conductivity type; The present invention is produced by a method comprising the steps of:

[0024] According to another embodiment, the base layer is formed by, after manufacturing the power semiconductor device, a cathode base region having a first depth disposed between the cathode region and the drift layer; a gate base region having a second depth disposed between the gate electrode and the drift layer; an intermediate base region disposed between the cathode base region and the gate base region and having a third depth of at least two different values; The third depth of the two values ​​is greater than the first depth and less than the second depth.

[0025] Advantageously, the intermediate base region allows for decoupling the gate base region from the cathode base region. Advantageously, the power semiconductor device has a large turn-off capability and a small on-state voltage drop. The steps and processes shown above are not performed in the order given above. The method for manufacturing a power semiconductor device is suitable for a power semiconductor device. Thus, features and advantages described in relation to a power semiconductor device can be used for the method and vice versa.

[0026] According to another embodiment, the intermediate base doping process includes depositing a first mask covering at least a portion of the area of ​​the cathode region and performing a first ion implantation process. The gate base doping process includes depositing a second mask covering the area of ​​the cathode region and also covering at least a portion of the area of ​​the intermediate base region and performing a second ion implantation process. The cathode base doping process includes performing a third ion implantation process without using a mask.

[0027] According to another embodiment, the power semiconductor device is realized as a fine-patterned gate commutated thyristor, abbreviated as GCT, which has low on-state losses and large turn-off capability, which is achieved by a fine-patterned cell design. The power semiconductor device belongs, for example, to the field of bipolar silicon devices.

[0028] According to another embodiment, in the proposed GCT design concept, the depth and doping profile of the p-base region are optimized. The base layer can be referred to as a p-base layer. The base layer has three distinct regions: (i) a cathode base region with a relatively shallow depth and maximum doping concentration, (ii) a gate base region with a deeper depth and a higher maximum doping concentration compared to the cathode base region, and (iii) an intermediate base region with an intermediate depth and an intermediate maximum doping concentration compared to the cathode base region and the gate base region. The intermediate base region (also referred to as intermediate p-base region) provides an extra degree of freedom in optimization, essentially allowing separate optimization of the gate base region (also referred to as gate p-base region) and the cathode base region (also referred to as cathode p-base region), with the cathode base region designed for best on-state performance and the gate base region configured for best turn-off performance, resulting in a small on-state voltage drop and a large turn-off current capability. Power semiconductor devices (along with planar GCTs) are enablers of fine pattern GCTs (small cathode and gate feature sizes) and cellular design of the cathode region.

[0029] According to another embodiment, the power semiconductor device optimizes the state of the art trade-off between on-state voltage VT and maximum controllable current (abbreviated as MCC) and allows a much smaller on-state voltage drop and at the same time a larger current controllability. This is achieved by structuring the base layer with at least three different levels of maximum doping concentration and depth. The doping concentration can also be called, for example, doping density.

[0030] According to another embodiment, the gate-commutated thyristor cell is not characterized by only a uniform base depth or two distinct base regions. The gate-commutated thyristor cell allows better handling of carriers generated during dynamic avalanche and better extraction of plasma and holes during turn-off. Local increases in current density therefore have no adverse effect on the latching (retriggering) of the thyristor. The MCC can be achieved without a high on-state voltage. On the other hand, power semiconductor devices achieve high MCC and also low on-state voltage.

[0031] According to another embodiment, the gate commutated thyristor cells are designed with three and more distinct p-base regions (multiple-level corrugations) to achieve lower on-state voltage (conduction losses) and at the same time higher turn-off current capability. The gate commutated thyristor cells have narrow cathode and / or gate regions, e.g., less than 250 μm, combined with multi-level corrugations. In an example, the gate commutated thyristor cells include split buffer regions and / or split anode regions. The gate commutated thyristor cells achieve smaller feature size cell designs with cathode shapes such as hexagonal, octagonal, circular and / or rectangular (striped morphology).

[0032] According to another embodiment, the power semiconductor device obtains less loss and higher durability, which opens the door to higher power density. The method makes available the processing of GCTs with enhanced performance. The smaller the base, the better, if it is shallow and also if the temperature increases. In an example, if the base is shallower than 75 μm, narrow fingers, for example less than 150 μm, will facilitate the improvement of MCC width. The dimensions of the gate contact area are typically equal to the dimensions of the cathode area. Gate commutation thyristor cells can be realized using narrower cathode and gate areas (for example, the width of the cathode area is less than 250 μm).

[0033] The present disclosure includes several aspects of power semiconductor devices and corresponding manufacturing methods. All features described in relation to one of these aspects are also disclosed herein in relation to other aspects, even if the respective feature is not expressly mentioned in the context of a particular aspect.

[0034] The accompanying figures are included to provide further understanding. In the figures, elements of the same structure and / or functionality may be referenced by the same reference numerals. It should be understood that the embodiments shown in the figures are illustrative representations and are not necessarily drawn to scale. [Brief description of the drawings]

[0035] [Figure 1] 1A-1D are cross-sectional views of power semiconductor devices according to different embodiments. [Diagram 2] 1A-1D are cross-sectional views of power semiconductor devices according to different embodiments. [Figure 3A] 1A-1D are cross-sectional views of power semiconductor devices according to different embodiments. [Figure 3B] 1A-1D are cross-sectional views of power semiconductor devices according to different embodiments. [Figure 4] 1A-1D are cross-sectional views of power semiconductor devices according to different embodiments. [Diagram 5] 1A-1D are cross-sectional views of power semiconductor devices according to different embodiments. [Figure 6] 1A-1D are cross-sectional views of power semiconductor devices according to different embodiments. [Figure 7] 1A-1D are perspective views of a power semiconductor device according to different embodiments; [Figure 8] 1A-1D are perspective views of a power semiconductor device according to different embodiments; [Figure 9A] 1A-1D are top views of power semiconductor devices according to different embodiments. [Figure 9B] 1A-1D are top views of power semiconductor devices according to different embodiments. [Figure 9C] 1A-1D are top views of power semiconductor devices according to different embodiments. [Figure 9D] 1A-1D are top views of power semiconductor devices according to different embodiments. [Figure 10] 1 illustrates a cross section of a power semiconductor device and a mask according to an embodiment. [Figure 11] 1 is a diagram and characteristics of a power semiconductor device according to different embodiments; [Figure 12] 1 is a diagram and characteristics of a power semiconductor device according to different embodiments; [Figure 13] 1 is a diagram and characteristics of a power semiconductor device according to different embodiments; [Figure 14] 1 is a diagram and characteristics of a power semiconductor device according to different embodiments; DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0036] FIG. 1 is a cross-sectional view of a power semiconductor device 1 according to an embodiment. The power semiconductor device 1 includes a four-layer npnp active thyristor structure that can be turned off by applying a negative voltage between the gate electrode 4 and the cathode electrode 2. The power semiconductor device 1 includes a gate commutated thyristor cell 20, abbreviated as cell. The power semiconductor device 1 has a first main surface 21 and a second main surface 22. FIG. 1 shows a known gate commutated thyristor cell. The cell has an anode layer 5, a buffer layer 6, a drift layer 7 (which can be called a second base layer), a base layer 8 and a cathode region 9 (which can also be called a cathode layer). The cathode region 9 is formed by an n+ doped region. The cathode region 9 is in contact with a cathode electrode 2, which can also be called a cathode metallization. The base layer 8 is realized as a p-doped base layer and is in contact with the cathode region 9 on the side opposite the cathode electrode 2.

[0037] The drift layer 7 is n-doped and may also be referred to as an n-doped second base layer. The base layer 8 is p-doped. The drift layer 7 is in contact with the base layer 8 on the side opposite the cathode region 9. The buffer layer 6 is n-doped. The buffer layer 6 is in contact with the drift layer 7 on the side opposite the base layer 8. The anode layer 5 or anode region is p-doped. The anode layer 5 of the power semiconductor device 1 is in contact with the buffer layer 6 on the side opposite the base layer 8. The anode electrode 3, which may be referred to as an anode metallization, is in contact with the anode layer 5 on the side opposite the buffer layer 6. The n-doped layer or region provides a first conductivity type. The p-doped layer or region provides a second conductivity type. The second conductivity type is different from the first conductivity type.

[0038] The integrated gate commutated thyristor (abbreviated IGCT) is a high voltage (HV) single wafer semiconductor device made of many gate commutated thyristor cells 20 (GCT cells). Each GCT cell 20 contains a cathode electrode 2, a gate electrode 4 and an anode electrode 3. The GCT cell 20 includes at least one gate electrode 4 in the form of an annular metal area, also called a gate contact, either in the center, center and / or periphery of a power semiconductor device wafer, also called a GCT wafer. The gate contact area is physically continuous with all gate electrodes 4 in the wafer, meaning that the gate electrodes 4 are electrically and thermally connected to each other and, for example, integral with the gate contact area. Each GCT cell 20 includes or consists of four layers with different electrical conductivities, forming a thyristor structure.

[0039] State-of-the-art GCT cells are turned off by applying a voltage to the gate electrode 4. Holes are extracted from the base layer 8 to the gate electrode 4 and very quickly all injection of electrons from the cathode region 9 stops. During the next phase of turn-off, the cathode region 9 remains off until the device is completely turned off. After that, only the doped base layer 8, the doped drift layer 7, the doped buffer layer 6 and the doped anode layer 5 conduct current until the power semiconductor device 1 is completely turned off, which results in a combination of semiconductor layers similar to an open-base PNP transistor. When turning off large currents, IGCTs can fail due to parasitic retriggering of the thyristor during turn-off. Parasitic retriggering of the thyristor is due to the large current densities observed in the vicinity of at least one cathode segment in the wafer device, a phenomenon induced by dynamic avalanches, as well as inherent design asymmetries and electrical impedance imbalances.

[0040] A power semiconductor device 1 having a thyristor internal structure can be turned off via the gate electrode 4 which has a large ability to turn off the current if, during turn-off, all or a reduced number of carriers are kept away from the vicinity of the cathode region 9. On the other hand, a reduction in the on-state losses can be achieved by enhancing the injection of electrons from the cathode, reducing the thickness of the base layer 8 of the power semiconductor device 1 and / or reducing the maximum doping concentration to reduce the electron recombination, which has an opposite effect on the turn-off ability.

[0041] The interfaces between the cathode region 9 and the base layer 8 and between the gate electrode 4 and the base layer 8 are disposed in a single plane. The interfaces between the cathode region 9 and the base layer 8 and between the gate electrode 4 and the base layer 8 are flat and coplanar. The base layer 8 and the cathode region 9 form a first junction 31. The drift layer 7 and the base layer 8 form a second junction 32. The buffer layer 6 and the anode layer 5 form a third junction 33. If the buffer layer is omitted, the drift layer 7 and the anode layer 5 form the third junction 33.

[0042] Figure 2 is a cross-sectional view of a power semiconductor device 1 according to an embodiment that is another development of the embodiment shown in Figure 1. The p-doped base layer 8 of the GCT cell 20 has two distinct regions, namely a cathode base region 81 adjacent to the cathode region 9 and a gate base region 82 adjacent to the gate electrode 4 and located at a distance from the cathode region 9. The gate base region 82 has a higher maximum doping concentration and / or a deeper depth than the cathode base region 81 in order to modulate the field in the blocking state and to defocus the generated holes from the cathode when driven by dynamic avalanche phenomena. The cathode base region 81 can also be called a cathode p-base region. The gate base region 82 can be called a gate p-base region.

[0043] FIG. 3A is a cross-sectional view of a power semiconductor device 1 according to an embodiment that is another development of the embodiment shown in FIG. 1 and FIG. 2. The gate commutation thyristor cell 20 does not have a buffer layer 6. The drift layer 7 and the anode layer 5 form the third junction 33. The base layer 8 includes at least three distinct base regions, namely, the cathode base region 81, the gate base region 82 and the intermediate base region 83. The cathode base region 81 has a relatively shallow depth and a relatively low maximum doping concentration to enable reduced electron recombination, smaller sheet resistance and thin wafer thickness. The gate base region 82 has a relatively deep depth and a high maximum doping concentration to draw plasma from deep inside the drift layer 7 (which can be called the second base layer) during turn-off, to provide a location for high electric field and dynamic avalanche carrier generation at a desired location directly under the gate electrode 4, and to support high voltage. Therefore, during turn-off, the plasma removal process is faster and any carriers generated by dynamic avalanche are removed by the gate electrode 4 without reaching the cathode region 9 or the vicinity of the cathode electrode 2 .

[0044] The cathode base region 81 is disposed vertically below a central portion of the cathode region 9. The gate base region 82 is disposed vertically below a central portion of the gate electrode 9. The intermediate base region 83 is disposed vertically below the gap between the cathode region 9 and the gate electrode 4, but may additionally be disposed below an edge portion of the gate electrode 9 and an edge portion of the cathode region 9.

[0045] The intermediate base region 83 has an intermediate depth and an intermediate maximum doping concentration compared to the gate base region 82 and the cathode base region 81. The depth, width and maximum doping concentration of the intermediate base region 83 determine the exact extent of the two other regions 81, 82 as well as the shape of the transition area. Furthermore, by having a completely different intermediate level, an extra degree of freedom in optimization is achieved, essentially allowing separate optimization of the other two base regions 81, 82, the cathode base region 81 for best on-state performance, and the gate base region 82 for best turn-off performance, preventing a trade-off between turn-off capability and on-state performance. Advantageously, the power semiconductor device 1 achieves precise pinpointing of the high dynamic avalanche region and also achieves complete suppression of high current density from avalanche generated carriers and thyristor re-triggering during turn-off in the open base transistor mode. At the same time, the power semiconductor device 1 allows a shallow cathode base region 81 to be designed to improve on-state performance. Advantageously, the power semiconductor device 1 achieves a large turn-off capability and a small on-state voltage drop. The power semiconductor device 1 provides a large maximum controllable current (MCC) while providing a small on-state voltage drop.

[0046] The cathode base region 81 has a first depth D1. The first depth D1 is the shortest distance between the cathode region 9 and the drift layer 7. The first depth D1 has a value in the range between 5 μm and 110 μm, alternatively between 20 μm and 110 μm, and alternatively between 50 μm and 80 μm. The values ​​used can be selected for the 4.5 kV GCT cell 20, but can also be selected for other GCT cells.

[0047] The gate base region 82 has a second depth D2. The second depth D2 is the maximum distance between the gate electrode 4 and the drift layer 7. The second depth D2 has a value in the range between 70 μm and 160 μm, alternatively between 100 μm and 130 μm.

[0048] The intermediate base region 83 has a third depth D3. The third depth 83 has at least two different values. Typically, the third depth D3 has three or more different values. The value of the third depth D3 ranges between 30 μm and 140 μm, alternatively between 80 μm and 100 μm. The depth of the base layer 8 is, for example, the thickness of the base layer 8. The third depth D3 is, for example, the thickness of the intermediate base region 83. The third depth D3 is, for example, the distance of the surface of the intermediate base region 83 to the drift layer 7.

[0049] The values ​​used are selected such that the first depth D1 is less than the second depth D2. The third depth D3 is between the first depth D1 and the second depth D2. At least two values, i.e. individual values, of the third depth D3 are between the first depth D1 and the second depth D2. Because the third depth has at least two different values, the second junction 32 has, for example, a slope or a step in the intermediate base region 83.

[0050] The value of the first depth D1 is less than the value of the second depth D2 when the value of the first depth D1 is the shortest distance between the cathode region 9 and the drift layer 7 and when the value of the second depth D2 is the longest distance between the gate electrode 4 and the drift layer 7. At least two values, i.e. individual values, of the third depth D3 are less than said value of the second depth D2 and greater than said value of the first depth D1.

[0051] A portion of the second junction 32, i.e., the interface disposed between the cathode base region 81 and the drift layer 7, is at least partially disposed in the first plane. A portion of the second junction 32, i.e., the interface disposed between the gate base region 82 and the drift layer 7, is at least partially disposed in the second plane. A portion of the second junction 32, i.e., the interface disposed between the intermediate base region 83 and the drift layer 7, is disposed between the first plane and the second plane. The first plane and the second plane are parallel to the first main surface 21.

[0052] The width W1 of the cathode region 9 or the width of the cathode electrode 2 ranges between 50 μm and 250 μm, alternatively between 100 μm and 200 μm. If the p-base is shallow, it is advantageous for W1 to be narrow. For example, if the base layer 8 is shallower than 75 μm, narrow fingers, for example less than 150 μm, will promote improved MCC. The dimension of the width W2 of the gate electrode 4 is typically equal to the dimension of the width W1 of the cathode region 9. The proposed design results in a failure mechanism at higher ratings. The width W1 is the distance of one edge of the cathode region 9 to the opposite edge of the cathode region 9 in the cross-sectional view shown in FIG. 4. The width W2 is the distance of one edge of the gate electrode 4 to the opposite edge of the gate electrode 4 in the cross-sectional view shown in FIG. 4.

[0053] The power semiconductor device 1 achieves a good MCC trade-off, which allows a much smaller on-state voltage drop and at the same time a higher current controllability. This is achieved by structuring the base layer 8 with at least three different levels of maximum doping concentration and depth. The power semiconductor device 1 allows a better handling of carriers generated during dynamic avalanches and a better extraction of plasma and holes during turn-off. A local increase in current density therefore has no adverse effect on the latching of the thyristor. This can be achieved without a large on-state voltage drop. In contrast to the embodiment shown in FIG. 3A, the examples shown in FIGS. 1 and 2 feature either a uniform base depth or only two completely different base regions.

[0054] The gate commutated thyristor cell 20 is realized with a double corrugation. The second junction 32 between the base layer 8 and the drift layer 7 is not flat.

[0055] In an alternative embodiment not shown, gate commutated thyristor cell 20 includes four or more distinct p-base regions.

[0056] Figure 3B is a cross-sectional view of a power semiconductor device 1 according to an embodiment that is another development of the embodiment shown in Figures 1 and 2. The interface or junction of the intermediate base region 83 to the drift layer 7 includes a first base transition region 10 and a second base transition region 11. The shapes of the first base transition region 10 and the second base transition region 11 impart a lateral velocity component to holes moving from the anode layer 5 that points towards the gate electrode 4.

[0057] Figure 4 is a cross-sectional view of a power semiconductor device 1 according to an embodiment that is another development of the embodiment shown in Figures 1, 2, 3A and 3B. The power semiconductor device 1 includes M gate commutated thyristor cells 20. In this example, the number M is 2. The power semiconductor device 1 comprises the gate commutated thyristor cells 20 and other gate commutated thyristor cells 23. Typically, the number M of cells is greater than 2.

[0058] The width W1 of the cathode region 9 ranges between 50 μm and 250 μm, alternatively between 100 μm and 200 μm. The width W2 of the gate electrode 4 ranges between 50 μm and 250 μm, alternatively between 100 μm and 200 μm. The difference between the width W1 of the cathode region 9 and the width W2 of the gate electrode 4 ranges from 10 μm to 70 μm. Alternatively, the difference is less than 30 μm, alternatively less than 10 μm, and alternatively less than 5 μm.

[0059] In an example, the distance between the cathode region 9 and the drift layer 7 can be as short as 10 μm to minimize the on-state voltage drop. Furthermore, the smaller the cell size, the larger the MCC and the larger the gate-cathode coupling may be, e.g., the cathode layer width W1 is less than 250 μm. Advantageously, the MCC can be increased or kept unchanged when the cell size is decreased and the gate-cathode coupling is increased, because the thyristor cell 20 latches a larger current density.

[0060] Thus, the distance between the cathode region 9 and the second junction 32 (also referred to as the first depth D1) can be as short as 10 μm, and the width W1 of the cathode region 9 can be less than 200 μm. The power semiconductor device 1 resembles a combination of an extremely shallow cathode base region 81 and a narrow cathode segment width W1. In this case, the cell dimensions are proportionally smaller (e.g., a deviation of 10% is allowed) and an equivalent increase in gate-cathode coupling is achieved, while the total active area of ​​the GCT cell 20 remains nearly unchanged.

[0061] In the example, the power semiconductor device 1 is fabricated on a wafer, for example a 91 mm wafer, which may be a GCT wafer. The power semiconductor device 1 realizes a fine pattern GCT (narrow cathode and gate region with very shallow cathode base region). The wafer is for example a silicon wafer. A single wafer may contain exactly one power semiconductor device 1. The power semiconductor device 1 has the size of one wafer.

[0062] The gate commutated thyristor cell 20 having narrower cathode and gate regions (e.g., a cathode region width W1 of less than 250 μm) results in a fine-patterned GCT with cathode electrodes 2 and / or cathode regions 9 of any shape for flexible design with low gate resistance and inductance, including improved performance and more freedom for design automation.

[0063] Figure 5 is a cross-sectional view of a power semiconductor device 1 according to an embodiment that is another development of the embodiment shown in Figures 1 to 4. In the power semiconductor device 1, the avalanche current is confined to the area under the gate electrode 4 and the first junction 31 (which can also be called the cathode / base layer junction) is kept in reverse bias. Thus, in the event of a failure, the power semiconductor device 1 is an open-base PNP that fails to turn off. To overcome this and to further improve the turn-off capability, the PNP transistor gain can be reduced by a split buffer design as shown in Figure 5 and / or a split anode design as shown in Figure 6.

[0064] In the split buffer design, the buffer 6 comprises a gate buffer region 61 and a cathode buffer region 62. The gate buffer region 61 is located at least directly below the gate electrode 4, and may also be below the gap between the cathode region 9 and the gate electrode 4. The cathode buffer region 62 is directly below the cathode region 9. The gate buffer region 61 has a higher maximum doping concentration compared to the cathode buffer region 62.

[0065] The buffer layer 6 is therefore divided into two distinct areas, one of which is vertically disposed below the cathode region 9 and the other of which is disposed below the gate electrode 4. The high maximum doping concentration of the gate buffer region 61 reduces the PNP transistor gain.

[0066] FIG. 6 is a cross-sectional view of a power semiconductor device 1 according to an embodiment which is another development of the embodiment shown in FIGS. 1 to 5. The anode layer 5 comprises a gate anode region 51 and a cathode node region 52. The gate anode region 51 is at least directly below the gate electrode 4. Furthermore, the gate anode region 51 may also be below the gap between the cathode region 9 and the gate electrode 4. The cathode node region 52 is directly below the cathode region 9. The anode layer 5 is thus divided into two distinct areas or regions, one of which is vertically disposed below the cathode region 9 and the other of which is disposed below the gate electrode 4. The gate anode region 51 has a lower maximum doping concentration compared to the cathode region 52, which reduces, for example, the PNP transistor gain.

[0067] Figure 7 is a perspective view of a power semiconductor device 1 according to an embodiment which is a further development of the embodiment shown in Figures 1 to 6. The gate electrode 4 surrounds the cathode region 9 and therefore also the cathode electrode 2. Correspondingly, the intermediate base region 83 surrounds the cathode base region 81. Furthermore, the gate base region 82 surrounds the intermediate base region 83.

[0068] FIG. 8 is a perspective view of a power semiconductor device 1 according to an embodiment which is a further development of the embodiment shown in FIGS. 1 to 7. The gate-commutated thyristor cell 20 is realized as a planar GCT, i.e. a flat GCT, so that the silicon is not etched, i.e. the first main surface 21 of the silicon is flat. The cathode region 9 is connected to the base layer 8 and thus to the cathode base region 81. The cathode region 9 is fabricated inside the substrate of the power semiconductor device 1. The insulation between the gate and the cathode is achieved via a difference in height of the metallizations of the gate and the cathode. The cathode metallization is thicker in comparison with respect to the gate metallization. The gate-commutated thyristor cell 20 further comprises an additional cathode metallization 2a which is arranged between the cathode electrode 2 and the cathode region 9. The double layer of the cathode electrode 2 and the additional cathode metallization 2a is therefore thicker than the gate electrode 4. Optionally, examples of power semiconductor device 1 shown in other figures may include a cathode region 9 and / or additional cathode metallization 2a as described in Figure 8. Figures 7, 8 and 9A-9D show circular and hexagonal cathode electrode 2 and cathode region 9 designs, as well as planar and mesa cathode design examples. The substrate of power semiconductor device 1 is at least one of silicon and a wide-bandgap semiconductor, such as silicon carbide.

[0069] 9A-9D are top views of a power semiconductor device 1 according to different embodiments, which are further developments of the embodiments shown in FIGS. 1-8. The cathode regions 9 and thus the cathode electrodes 2 can have the form of a circle as shown in FIG. 9A, a hexagonal form as shown in FIG. 9B, a rectangular or striped form as shown in FIG. 9C, and / or an octagonal form as shown in FIG. 9D. Other forms are also suitable. Different shapes of the cathode regions 9 and cathode electrodes 2 can be realized due to the possibility of reducing the cathode segment feature size. The designer can use a hexagonal or circular cell concept (more suitable for techniques using low resolution lithography) together with striped cathode fingers (FIG. 9C), and / or the planar GCT concept shown in FIGS. 7 and 8. The advantage is a further improvement of the electrical parameters (MMC) and the possible processing of the cathode and gate regions narrower than 250 μm as explained above.

[0070] FIG. 10 shows a cross section and masks of a power semiconductor device 1 according to an embodiment that is another development of the embodiment shown in the previous figures. The base layer 8 is manufactured by performing at least a gate base doping process 86, an intermediate base doping process 85 and a cathode base doping process 87. The intermediate base doping process 85 includes depositing a first mask 88 covering at least a part of the area of ​​the cathode region 9 and performing a first ion implantation process. The gate base doping process 86 includes depositing a second mask 89 covering the area of ​​the cathode region 9 and also covering at least a part of the area of ​​the intermediate base region 83 and performing a second ion implantation process. The cathode base doping process 87 includes performing a third ion implantation process without using a mask. In FIG. 10, the bold rectangles indicate areas where ions can be provided to the first main surface 21 by ion implantation. The ions are indicated by arrows.

[0071] The gate base doping process 86, the intermediate base doping process 85 and the cathode base doping process 87 provide doping of the second conductivity type. The gate base doping process 86 is configured to realize an implanted lightly doped p region using a second mask 89 to obtain the gate base region 82. The intermediate base doping process 85 is configured to realize an implanted lightly doped p region using a first mask 88 to obtain the intermediate base region 83. The cathode base doping process 87 is configured to realize an implanted lightly doped p region without using a mask to obtain the cathode base region 81.

[0072] Furthermore, a shallow base doping process 90 is performed without the use of a mask, the shallow base doping process being configured to produce an implanted highly doped region in the base layer 8, i.e., at the interface of the base layer 8 towards the first major surface.

[0073] A gate base doping process 86, an intermediate base doping process 85, a cathode base doping process 87 and a shallow base doping process 90 provide doping of the second conductivity type (p-doping).

[0074] The other processes for manufacturing the power semiconductor device 1 are conventional processes or steps. - providing a drift layer 7 of a first conductivity type in a semiconductor substrate; - providing a cathode region 9 of a first conductivity type; - providing an anode layer 5 of a second conductivity type; - providing a cathode electrode 2 on the first main face 21 and, laterally to the cathode region 9, a gate electrode 4 on the first main face; - providing an anode electrode 3 on the second main face 22; Includes.

[0075] The order of processes during manufacture may differ from the order written above. The method for manufacture may include other processes not mentioned above, for example, providing a passivation layer.

[0076] After fabrication of the power semiconductor device 1, e.g. after three base doping processes 85-87 and after the final annealing step and / or after the final high temperature step, the base layer 8 comprises the intermediate base region 83, the cathode base region 81 and the gate base region 82 described above. An example of the dopant distribution is shown in FIG.

[0077] Optionally, a buffer layer 6 is further provided. 11-14 are diagrams and characteristics of a power semiconductor device 1 according to a different embodiment, which is another development of the embodiment shown above. In Fig. 11-13, a cross-sectional view, a top view and a current / voltage characteristic of the power semiconductor device 1 are shown. As shown in Fig. 11, the gate commutation thyristor cell 20 is realized as an asymmetric integrated gate commutation thyristor cell 100, abbreviated as asymmetric IGCT cell. The gate commutation thyristor cell 20 comprises another cathode electrode 101 and another cathode region 102. In the I / V characteristic, FC stands for forward conduction and FB stands for forward blocking.

[0078] As is evident in Fig. 12, the gate commutated thyristor cell 20 is realized as a reverse conducting integrated gate commutated thyristor cell 104, i.e. an RC-IGCT cell. The power semiconductor device 1 further comprises a diode 105. As shown in the cross-sectional view of Fig. 12, an isolation region 106 separates the gate commutated thyristor cell 20, i.e. the cell, from the diode 105. The diode 105 is realized as a PN junction having a diode anode 107 and a diode cathode 108. The diode anode 107 is disposed on the first main surface 21. The diode cathode 108 is disposed on the second main surface 22. RC stands for reverse conduction.

[0079] 13, the gate commutation thyristor cell 20 is realized as a reverse blocking gate commutation thyristor cell 109, abbreviated as RB-IGCT. In this example, the buffer layer 6 is omitted. RB stands for reverse blocking.

[0080] 14, the power semiconductor device 1 is realized as a bidirectional turn-off thyristor 110, abbreviated as BTT, and includes a gate rectifier thyristor cell 20 and another gate rectifier thyristor cell 115 connected in anti-parallel to the gate rectifier thyristor cell 20. The other gate rectifier thyristor cell 115 includes another cathode electrode 101, another cathode region 102, another gate electrode 111, another base layer 112, another anode electrode 113 and another anode layer 114.

[0081] The base layer 8 in the examples shown in Figures 11 to 14 is realized as shown in Figures 3A, 3B or 4. Thus, the new gate-cathode design concept can be applied to different GCT families such as asymmetric IGCT, reverse conducting IGCT, reverse blocking IGCT and BTT.

[0082] While the present disclosure is amenable to various modifications and alternative forms, details of which have been shown by way of example in the drawings and will be described in detail. It is to be understood, however, that it is not intended to limit the disclosure to the particular embodiments described. On the contrary, it is intended to cover all modifications, equivalents, and alternatives falling within the scope of the present disclosure as defined by the appended claims.

[0083] 1-14 represent exemplary embodiments of the improved power semiconductor device 1 and manufacturing method, and therefore they do not constitute an exhaustive list of all embodiments of the improved power semiconductor device and method. Actual power semiconductor devices and methods may vary from the illustrated embodiments, for example, with respect to layout, devices, layers, regions, design, and processes. [Explanation of symbols]

[0084] 1. Power Semiconductor Devices 2. Cathode Electrode 2a Additional cathode metallization 3 Anode Electrode 4. Gate electrode 5 Anode layer 6 Buffer layer 7 Drift Layer 8 Base Layer 9 Cathode Region 10 First base transition region 11 Second base transition region 13 Cathode Area 20 Gate commutated thyristor cells 21 First main surface 22 Second main surface 31 First Junction 32 Second Junction 33 Third Junction 51 Gate anode region 52 Cathode Node Region 61 Gate Buffer Area 62 Cathode buffer area 81 Cathode base region 82 Gate base region 83 Intermediate Base Area 85 Intermediate Base Doping Process 86 Gate-base doping process 87 Cathode Base Doping Process 90 Shallow base doping process 100 Asymmetric integrated gate commutated thyristor cells 101 Other cathode electrodes 102 Other cathode regions 104 Reverse Conducting Integrated Gate Rectifier Thyristor Cell 105 Diode 106 Separation area 107 Diode Anode 108 Diode Cathode 109 Reverse-blocking integrated gate-commutated thyristor cell 110 Bidirectional turn-off thyristor cell 111 Other gate electrodes 112 Other Base Layers 113 Other anode electrodes 114 Other anode layers

Claims

1. A power semiconductor device (1) comprising a gate commutation thyristor cell (20), a first main surface (21), and a second main surface (22) opposite to the first main surface (21), wherein the gate commutation thyristor cell (20) is arranged in the following order from the first main surface (21) to the second main surface (22): a cathode electrode (2) arranged on said first main face (21); a cathode region (9) of a first conductivity type, a base layer (8) of a second conductivity type different from said first conductivity type, forming a first junction (31) to said cathode region (9); a drift layer (7) of said first conductivity type forming a second junction (32) with said base layer (8); an anode layer (5) of said second conductivity type, an anode electrode (3) arranged on said second main face (22); the gate commutation thyristor cell (20) further comprises a gate electrode (4) arranged laterally with respect to the cathode region (9); The base layer (8) a cathode base region (81) having a first depth (D1) and arranged between said cathode region (9) and said drift layer (7); a gate base region (82) arranged between said gate electrode (4) and said drift layer (7) and having a second depth (D2); an intermediate base region (83) arranged between said cathode base region (81) and said gate base region (82) and having a third depth (D3) of at least two different values; Equipped with The third depth (D3) of the at least two values ​​is deeper than the first depth (D1) and shallower than the second depth (D2); the maximum doping concentration of the gate base region (82) is higher than the maximum doping concentration of the intermediate base region (83); the maximum doping concentration of the intermediate base region (83) is higher than the maximum doping concentration of the cathode base region (81); an intermediate base region (83) comprising the gap between the gate electrode (4) and the cathode region (9) on one side and the second junction (32) on the other side; Power semiconductor device (1).

2. The intermediate base region (83) the adjacent region between said gate electrode (4) and said second junction (32), and the adjacent area between said cathode region (9) and said second junction (32); The power semiconductor device (1) according to claim 1, further comprising at least one of:

3. A power semiconductor device (1) comprising a gate commutation thyristor cell (20), a first main surface (21), and a second main surface (22) opposite to the first main surface (21), wherein the gate commutation thyristor cell (20) is arranged in the following order from the first main surface (21) to the second main surface (22): a cathode electrode (2) arranged on said first main face (21); a cathode region (9) of a first conductivity type, a base layer (8) of a second conductivity type different from said first conductivity type, forming a first junction (31) to said cathode region (9); a drift layer (7) of said first conductivity type forming a second junction (32) with said base layer (8); an anode layer (5) of said second conductivity type, an anode electrode (3) arranged on said second main face (22); the gate commutation thyristor cell (20) further comprises a gate electrode (4) arranged laterally with respect to the cathode region (9); The base layer (8) a cathode base region (81) having a first depth (D1) and arranged between said cathode region (9) and said drift layer (7); a gate base region (82) arranged between said gate electrode (4) and said drift layer (7) and having a second depth (D2); an intermediate base region (83) arranged between said cathode base region (81) and said gate base region (82) and having a third depth (D3) of at least two different values; Equipped with The third depth (D3) of the at least two values ​​is deeper than the first depth (D1) and shallower than the second depth (D2); the maximum doping concentration of the gate base region (82) is higher than the maximum doping concentration of the intermediate base region (83); the maximum doping concentration of the intermediate base region (83) is higher than the maximum doping concentration of the cathode base region (81); a portion of the second junction (32) disposed between the cathode base region (81) and the drift layer (7) is at least partially disposed in a first plane; a portion of the second junction (32) disposed between the gate base region (82) and the drift layer (7) is at least partially disposed in a second plane; a portion of the second junction (32) that is disposed between the intermediate base region (83) and the drift layer (7) is disposed between the first plane and the second plane; Power semiconductor device (1).

4. The portion of the second junction (32) disposed between the intermediate base region (83) and the drift layer (7) is a first base transition area (10) forming a staircase; a second base transition area (11) forming another step; a plane between said first base transition region (10) and said second base transition region (11); Including, A power semiconductor device (1) according to claim 3.

5. a maximum doping concentration of the gate base region (82) in an area closest to the first main surface (21) is higher than a maximum doping concentration of the cathode base region (81) closest to the first main surface (21); A power semiconductor device (1) according to any one of the preceding claims.

6. the first depth (D1) is a minimum depth of the cathode base region (81); the second depth (D2) being the maximum depth of the gate base region (82); the third depth (D3) of the at least two values ​​is greater than the first depth (D1) and less than the second depth (D2); A power semiconductor device (1) according to any one of the preceding claims.

7. the first depth (D1) is the shortest distance of the cathode region (9) to the drift layer (7); the second depth (D2) is the maximum distance of the gate electrode (4) to the drift layer (7); the third depth (D3) of the at least two values ​​is greater than the first depth (D1) and less than the second depth (D2); A power semiconductor device (1) according to any one of the preceding claims.

8. The first depth (D1) has a value in the range between 5 μm and 110 μm; A power semiconductor device (1) according to any one of the preceding claims.

9. The width (W1) of the cathode region (9) is in the range between 50 μm and 250 μm; A power semiconductor device (1) according to any one of the preceding claims.

10. The difference between the width (W1) of the cathode region (9) and the width (W2) of the gate electrode (4) is in the range of 10 μm to 70 μm; A power semiconductor device (1) according to any one of the preceding claims.

11. On said first main surface (21), said cathode region (9) has a shape from the group consisting of a circle, a rectangle, an octagon and a hexagon. A power semiconductor device (1) according to any one of the preceding claims.

12. the anode layer (5) comprises a gate anode region (51) and a cathode node region (52), the maximum doping concentration of the gate anode region (51) being less than the maximum doping concentration of the cathode node region (52); A power semiconductor device (1) according to any one of the preceding claims.

13. the gate commutated thyristor cell (20) comprises a buffer layer (6) of the first conductivity type; the buffer layer (6) is disposed between the drift layer (7) and the anode layer (5) and includes a gate buffer region (61) and a cathode buffer region (62); the maximum doping concentration of the gate buffer region (61) is higher than the maximum doping concentration of the cathode buffer region (62); A power semiconductor device (1) according to any one of the preceding claims.

14. said gate commutated thyristor cell (20) being realized as a member of a group including an asymmetric integrated gate commutated thyristor cell (100) (asymmetric IGCT), a reverse conducting integrated gate commutated thyristor cell (104) (RC-IGCT cell), a reverse blocking integrated gate commutated thyristor cell (109) (RB-IGCT cell), and a bidirectional turn-off thyristor cell (110) (BTT cell); A power semiconductor device (1) according to any one of the preceding claims.

Images