Reverse conducting power semiconductor device and method of manufacturing the same

Abstract

The invention relates to a reverse conducting power semiconductor device, wherein a plurality of thyristor cells (50) and freewheeling diodes (60) are integrated in a semiconductor wafer. A diode anode layer (32) comprises a plurality of first diode anode layer segments (321) which are strip-shaped in an orthogonal projection onto a plane parallel to a second main side (22) of the semiconductor wafer (20), a longitudinal main axis MA of each first diode anode layer segment (321) extending in a lateral direction away from a lateral center of the semiconductor wafer (20). A first lateral width (w1) of each first diode anode layer segment (321) in a direction perpendicular to its longitudinal main axis (MA) is at least 1000 pm, or at least 1200 pm, at any position along the longitudinal axis. The device further comprises a plurality of strip-shaped local lifetime control regions (91). A diode cathode layer (33) comprises a plurality of strip-shaped diode cathode layer segments (331). In the orthogonal projection, each local lifetime control region (91) and each diode cathode layer segment (331) is arranged within a corresponding first diode anode layer segment (321), respectively.

Description

Technical Field

[0001] The present invention relates to a reverse-conducting power semiconductor device according to the preamble of claim 1, and to a method for manufacturing such a reverse-conducting power semiconductor device. Background Technology

[0002] Integrated gate-commutated thyristors (IGCTs) have been established as the preferred device for many high-power applications, such as medium-voltage drives, STATCOMs, and pumped-storage generators. Today, IGCTs are optimized for current-source inverter (CSI) and voltage-source inverter (VSI) applications using state-of-the-art devices with rated voltages ranging from 4.5kV to 6.5kV, and are available as asymmetrical, symmetrical (reverse blocking), and reverse-conducting (RC) devices. Due to their thyristor-like turn-on behavior and transistor-like turn-off behavior, integrated gate-commutated thyristors (IGCTs) are the ideal device of choice for many high-power electronic applications.

[0003] A reverse-conducting integrated gate-commutated thyristor (RC-IGCT) is a reverse-conducting power semiconductor device that includes an IGCT section and a single built-in freewheeling diode section within a single semiconductor wafer. - The doped drift layer and the n-doped buffer layer are separated from the p-doped anode layer and the n-doped anode layer. + A doped cathode layer. The diode portion is circular and, in a top view, is adjacent to the IGCT portion at the center of the semiconductor wafer. There is an n-type cathode layer between the IGCT portion and the diode portion. – A doped separation region separates the p-doped base layer of the thyristor unit from the p-doped anode layer of the diode unit in the IGCT section. In this RC-IGCT, the diode section of the device is optimized using lifetime control to reduce the peak reverse recovery current, thereby reducing reverse recovery losses and thus protecting the diode from high-power failures. However, the IGCT section of a semiconductor wafer does not utilize any lifetime control. Therefore, when proton (H) is applied... + ) or helium ions (He 2+ When high-energy ion implantation is used to generate recombination centers and thereby form a local lifetime control (LLC) region in the diode section, a metal mask with a thickness of approximately 0.5 mm (depending on the ion energy and mechanical stability of the mask) is used to effectively block heavy ions and prevent the formation of recombination centers in the region of the IGCT section. In both forward and reverse conduction states, only a portion of the semiconductor wafer is used for current. Therefore, losses are relatively high in both forward and reverse conduction states.

[0004] A known reverse conducting power semiconductor device, which was developed for reducing the on-state losses by utilizing the entire semiconductor wafer in forward and reverse conducting state, respectively, is a bi-mode gate commutated thyristor (BGCT) as shown in Figure 1 and Figure 2 . Figure 1 The BGCT is shown in a top view and Figure 2 in a cross-sectional view taken along the line c'-c in Figure 1 The BGCT comprises a plurality of thyristor cells 2 electrically connected in parallel to each other in a single semiconductor wafer 1. In the BGCT shown in Figure 1 and Figure 2 each of the thyristor cells 2 consists of three thyristor cathode electrodes 3 in the form of cathode metallization layers, an n + doped thyristor cathode layer, a p-doped base layer 5, an n – doped drift layer 6, an n-doped buffer layer 7, a p + doped thyristor anode layer 8 and thyristor anode electrodes 9 in the form of anode metallization layers. The thyristor cells 2 each further comprise a gate electrode 10 in the form of a gate metallization layer in contact with the p-doped base layer 5. The gate metallization layer is arranged in a plane which is lower than the plane in which the thyristor cathode electrodes 3 are arranged, such that the gate electrodes 10 are vertically separated from the thyristor cathode electrodes 3. The BGCT comprises a single common gate contact 11 in the form of a ring-shaped metal region in the center of the semiconductor wafer 1. The common gate contact 11 is in direct contact with the gate metallization layers, such that the gate contact 11 and the gate electrodes 10 of all thyristor cells 2 are electrically connected to each other. The BGCT comprises a plurality of diode cells 12 distributed between the thyristor cells 2. The diode cells 12 are electrically connected in parallel to each other and to the thyristor cells 2, although with reversed polarity. Each diode cell 12 comprises a diode anode electrode 17, a p-doped diode anode layer 13, an n + doped diode cathode layer 14 and a diode cathode electrode 16, wherein the p-doped diode anode layer 13 and the n + doped diode cathode layer 14 are separated by the n – doped drift layer 6 and the n-doped buffer layer 7. Adjacent thyristor cells 2 and diode cells 12 are separated by separation regions 15, respectively.

[0005] As described above, a metal mask with a thickness of approximately 0.5 mm is used to selectively form LLC regions (single diode portions integrated with the IGCT portions on the same semiconductor wafer) within the diode portion of a known RC-IGCT. However, for reproducible processes, the structure of the masked portion must be larger than the mask thickness. Using a typical BGCT layout, the width of the diode cell is on the same order of magnitude as or smaller than the required thickness of the metal mask. Moreover, to form the LLC region, heavy ion implantation must be performed at a 7° tilt angle to the surface normal to avoid channeling effects. Therefore, even with perfect alignment, heavy ion implantation deviates from the diode segment, and avoiding implantation into the GCT region becomes even more critical.

[0006] Therefore, due to the fine details of the BGCT layout, excluding LLC is judged to be a prerequisite in BGCT. The concept of an emitter-controlled (EMCON) diode is considered a development of the classic BGCT design. However, it requires the diode section ( Figure 2 p-doped diode anode layer 13 and n - The pn junction between the doped drift layer 6 and the GCT portion ( Figure 2 p-doped base layer 5 and n - Different junction depths between the doped drift layers 6 (pn junctions) are used to maintain a low diode anode-emitter efficiency. On the other hand, if the diode portion and the GCT portion have different blocking junction depths, the diode is less robust.

[0007] According to US 2018 204 913A1, a turn-off power semiconductor device is known, comprising a plurality of thyristor units, each thyristor unit including a cathode region; a base layer; a drift layer; an anode layer; a gate electrode arranged laterally relative to the cathode region and in contact with the base layer; a cathode electrode; and an anode electrode. The interfaces between the cathode regions and cathode electrodes of the plurality of thyristor units, and between the base layer and the gate electrode, are planar and coplanar. Furthermore, the base layer includes a gate well region extending from its contact with the gate electrode to a depth at least half the depth of the cathode region, wherein, for any depth, the minimum doping concentration of the gate well region at this depth is 50% higher than the doping concentration of the base layer at a lateral location between the cathode region and the gate well region at this depth and at a distance of 2 μm from the cathode region in an orthogonal projection onto a plane parallel to the first main side. The base layer includes a compensation region of a second conductivity type, which is arranged directly adjacent to the first main side and between the cathode region and the gate well region, wherein the density of the first conductivity type impurity is at least 0.4 relative to the net doping concentration in the compensation region.

[0008] The above BGCT is disclosed and discussed, for example, in the paper "Concept of Dual Mode Gate Commutated Thyristor" by Umamaheswara Vemulapati and Marco Bellini, Proceedings of the 24th International Symposium on Power Semiconductor Devices and ICs, Bruges, Belgium, June 3-7, 2012, pages 29-32.

[0009] From US 2016 013 302 Al, a reverse conducting power semiconductor device is known, having a wafer with first and second main sides arranged opposite and parallel to each other. The device comprises a plurality of diode cells and a plurality of gate-commutated thyristor (GCT) cells. Each GCT cell comprises layers of a first conductivity type (e.g. n-type) and a second conductivity type (e.g. p-type) between the first and second main sides. The device comprises at least one hybrid section, in which diode anode layers of diode cells alternate with first cathode layers of GCT cells. In each diode cell, a diode buffer layer of the first conductivity type is arranged between the diode anode layer and the drift layer, such that the diode buffer layer covers a lateral side of the diode anode layer to a depth of about 90% of the thickness from the first main side to the diode anode layer.

[0010] From US 5 682 044 A, a reverse conducting (RC) thyristor of a planar gate structure for low to medium power use is known, which is relatively simple in construction due to adoption of a planar structure for each of a thyristor and diode region, allows simultaneous formation of the two regions and has high speed performance; and an RC thyristor of a buried gate or recessed gate structure, which has a higher breakdown voltage by using a buried gate or recessed gate structure, allows simultaneous formation of a thyristor and diode region and high speed, high current switching performance, and the RC thyristor of a planar gate structure has a construction including a static induction (SI) thyristor of a planar gate structure in a thyristor region or a small-sized GTO and an SI diode of a planar structure in a diode region, the diode region has a Schottky contact between an n emitter or diode cathode short region at a cathode side thereof, and the thyristor region has an SI anode short structure formed of a (p+) anode layer, a wave-shaped anode layer or an anode (n+) layer at an anode side thereof; in the case of a high breakdown device, an n buffer layer is added; similarly, the RC thyristor of a buried gate or recessed gate structure has a construction including an SI thyristor of a buried gate or recessed gate structure in a thyristor region and an SI diode of a buried gate or recessed gate structure. SUMMARY

[0011] In view of the above, it is an object of the present application to provide a reverse conducting power semiconductor device which can overcome some or all of the above-described problems in the prior art. In particular, it is an object of the present application to provide a reverse conducting power semiconductor device in which forward conduction losses and switching conduction losses can be kept low at the same time, while ensuring good thermal performance.

[0012] The object of the present application is achieved by a reverse conducting power semiconductor device according to claim 1. Further refinements of the application are specified in the dependent claims.

[0013] The reverse conducting power semiconductor device of the present application comprises a plurality of thyristor cells and a freewheeling diode integrated in a semiconductor wafer having a first main side and a second main side opposite to the first main side. Each of the plurality of thyristor cells comprises, in order from the first main side to the second main side: a thyristor cathode layer of a first conductivity type; a base layer of a second conductivity type different from the first conductivity type, wherein a first p-n junction is formed between the base layer and the thyristor cathode layer; a drift layer of the first conductivity type, which forms a second p-n junction with the base layer; and a thyristor anode layer of the second conductivity type, which is separated from the base layer by the drift layer. Each thyristor cell further comprises: a gate electrode arranged laterally with respect to the thyristor cathode layer and forming an ohmic contact with the base layer; a thyristor cathode electrode arranged on the first main side and forming an ohmic contact with the thyristor cathode layer; and a thyristor anode electrode arranged on the second main side and forming an ohmic contact with the thyristor anode layer. The freewheeling diode comprises: a diode anode layer of the second conductivity type at the first main side, which forms a third p-n junction with the drift layer and is separated from the base layer by the drift layer; a diode anode electrode on the first main side, which is electrically connected to the diode anode layer; a diode cathode layer of the first conductivity type at the second main side, which is electrically connected to the drift layer; and a diode cathode electrode on the second main side, which forms an ohmic contact with the diode cathode layer. The diode anode layer comprises a plurality of first diode anode layer segments which are strip-shaped in an orthogonal projection onto a plane parallel to the second main side, a longitudinal main axis of each first diode anode layer segment extending away from a lateral center of the semiconductor wafer in a lateral direction, wherein a first lateral width of each first diode anode layer segment in a plane parallel to the second main side and in a direction perpendicular to its longitudinal main axis is at least 1000 pm, or at least 1200 pm, at any position along the longitudinal main axis. This means that a minimum lateral width of each first diode anode layer segment is at least 1000 pm, or at least 1200 pm.

[0014] Throughout the specification, lateral refers to a direction parallel to the second main side, and the lateral center of the semiconductor wafer is determined as the center in a plane parallel to the second main side. Also, throughout the specification, the center of a region is understood as the centroid, i.e. the arithmetic mean position of all points in that region. Also, throughout the specification, a bar-shaped element is defined as any element having a width in a predetermined longitudinal direction that is at least twice as large as its width in any other direction, wherein the width in the predetermined longitudinal direction is along any line perpendicular to the predetermined longitudinal direction. The longitudinal principal axis of a bar-shaped element extends along the predetermined longitudinal direction.

[0015] In the reverse conducting power semiconductor device of the present invention, implementing a freewheeling diode with a plurality of bar-shaped first diode anode layer segments ensures good thermal and current spreading within the semiconductor wafer. Further, the segmentation of the freewheeling diode leads to a less snappy behavior of the freewheeling diode, which in turn allows to reduce the thickness of the semiconductor wafer, leading to reduced losses in the reverse and forward conducting state. The lower limit of the first lateral width of the bar-shaped first diode anode layer segments leads to lower forward conduction losses compared to known BGCTs. This can be explained by minimizing the conductivity modulation due to reduced current spreading. Even with a first lateral width of the bar-shaped first diode anode layer segments significantly higher than the higher lower limit in known BGCTs, the bar-shaped thermal spreading is still effective and the reverse conducting power semiconductor device exhibits a good surge current capability.

[0016] The reverse conducting power semiconductor device of the present invention comprises a plurality of local lifetime control regions comprising radiation-induced recombination centers, wherein each local lifetime control region is bar-shaped in an orthogonal projection onto a plane parallel to the second main side and is arranged within a corresponding one of the first diode anode layer segments in the orthogonal projection such that a longitudinal principal axis of each local lifetime control region extends along a longitudinal principal axis of the corresponding one of the first diode anode layer segments, and each local lifetime control region has a second lateral width that is at least 200 pm or at least 300 pm smaller than a first lateral width of the corresponding one of the first diode anode layer segments. Exemplarily, in each vertical cross-section along a plane orthogonal to the second main side and orthogonal to the longitudinal principal axis of the corresponding one of the first diode anode layer segments, the second lateral width is at least 200 pm or at least 300 pm smaller than the first lateral width of the corresponding one of the first diode anode layer segments.

[0017] In the inventive reverse conducting power semiconductor device, the relatively high carrier lifetime at the edges of the drift layer portion in the bar-shaped diode portion (the portion of the semiconductor wafer that overlaps with the bar-shaped first diode anode layer segment in an orthogonal projection onto a plane parallel to the second main side) leads to a much higher injection of charge there than in the central portion of the bar-shaped diode portion. Therein, the central portion refers to the portion that is centered with respect to a direction perpendicular to the longitudinal main axis and parallel to the second main side. As the diode recovers, the p-n junction in the proton implanted region is quickly cleared as there is less charge present there. This will initiate a transfer of voltage from the switch and other circuit elements (such as a choke) to the diode. The appearance of the diode voltage slows down the current change rate and brings the diode reverse recovery peak current Irr to a comparable level as if the diode had been irradiated to the full width of the first diode anode layer segment. The charge hill at the edges of the diode is still there. Electrons travel down in the general direction of the high voltage. When they displace laterally to the n + The diode cathode layer when it comes to the edges of the diode anode layer segment, they have to flow laterally to reach the diode cathode layer. Obviously, this is the situation where the well-known field charge extraction (FCE) effect can come into play. This means that in such exemplary embodiments, the freewheeling diode is a FCE diode enhanced by lifetime segmentation. The charge carrier hill gets the best lateral position and the low lifetime zone ensures a reasonable diode reverse recovery peak current Irr. This leads to a soft recovery of the freewheeling diode.

[0018] In the invention, the diode cathode layer comprises a plurality of diode cathode layer segments, wherein each diode cathode layer segment is bar-shaped in an orthogonal projection onto a plane parallel to the second main side and is arranged within a corresponding one of the bar-shaped first diode anode layer segments in this orthogonal projection such that a longitudinal main axis of each diode cathode layer segment extends along a longitudinal main axis of the corresponding one of the first diode anode layer segments, and wherein each diode cathode layer segment has a third lateral width that is at least 200 pm or at least 300 pm smaller than a second lateral width of a corresponding one of the local lifetime control zones that is arranged within the corresponding one of the first diode anode layer segments in an orthogonal projection onto a plane parallel to the second main side. In such exemplary embodiments, the FCE effect during reverse recovery of the freewheeling diode is more pronounced, leading to a soft recovery and less abrupt behavior of the freewheeling diode. Exemplarily, in each vertical cross section along a plane that is orthogonal to the second main side and orthogonal to the longitudinal main axis of the corresponding one of the first diode anode layer segments, the third lateral width is at least 200 pm or at least 300 pm smaller than the second lateral width of the corresponding one of the local lifetime control zones. In exemplary embodiments, the semiconductor wafer has a circular shape and the longitudinal main axis of each first diode anode layer segment extends along a radial direction of this circular shape.

[0019] In an exemplary embodiment, the first lateral width of each first diode anode layer segment is less than 5000 pm or less than 4000 pm or less than 3000 pm at any position along its longitudinal main axis. This means that the maximum value of the first lateral width is less than 5000 pm or less than 4000 pm or less than 3000 pm.

[0020] In an exemplary embodiment, a circular portion of the freewheeling diode is arranged at the lateral center of the semiconductor wafer in an orthogonal projection on a plane parallel to the second main side. In this arrangement, the available semiconductor wafer area is most effectively utilized. In this exemplary embodiment, each first diode anode layer segment can laterally extend from the circular portion of the freewheeling diode.

[0021] In an exemplary embodiment, the length of each first diode anode layer segment in the direction along its longitudinal main axis is at least 20% or at least 25% of the width of the semiconductor wafer in this direction. With such a relatively long strip-shaped first diode anode layer segment, the thermal performance is improved due to an effective heat spreading of the heat generated in the freewheeling diode.

[0022] In an exemplary embodiment, the diode anode layer comprises strip-shaped second diode anode layer segments extending in a radial direction, which are laterally arranged between two adjacent first diode anode layer segments, respectively, each second diode anode layer segment being at a distance from the lateral center of the semiconductor wafer that is greater than the distance of each first diode anode layer segment from the lateral center of the semiconductor wafer. In such an exemplary embodiment, the variation of the distance between adjacent strip-shaped diode anode layer segments can be reduced and the heat spreading is enhanced.

[0023] In an exemplary embodiment, the minimum value of the first lateral width of each first diode anode layer segment is less than 2000 pm. With such a parameter, the thermal performance is improved and the freewheeling diode exhibits less abrupt behavior than in known RC-IGCTs.

[0024] In an exemplary embodiment, the third lateral width is at least 600 pm or at least 800 pm less than the first lateral width of a corresponding one of the first diode anode layer segments. In such an exemplary embodiment, the FCE effect during the reverse recovery of the freewheeling diode is more pronounced, resulting in a soft recovery and less abrupt behavior of the freewheeling diode. Exemplarily, in each vertical cross-section along a plane orthogonal to the second main side and orthogonal to the longitudinal main axis of the corresponding one of the first diode anode layer segments, the third lateral width is at least 600 pm or at least 800 pm less than the first lateral width of the corresponding one of the first diode anode layer segments.

[0025] In an exemplary embodiment, the depth of each base layer and the depth of the diode anode layer are the same. In such exemplary embodiment, diode robustness is improved compared to known BGCTs having different diode junction depths (i.e. the depth of the diode anode layer) and GCT portions (i.e. the depth of the base layer). Also, in such exemplary embodiment, the base layer and the diode anode layer can be formed simultaneously in the same process step. Thus, manufacturing of the reverse conducting power semiconductor device is facilitated.

[0026] In an exemplary embodiment, the gate electrodes of the plurality of thyristor units are electrically connected to each other, and the device further comprises a common gate contact for contacting the gate electrodes of the plurality of thyristor units, wherein the common gate contact is arranged on a circumferential edge of the semiconductor wafer on the first main side.

[0027] In an exemplary embodiment, the thyristor cathode layer comprises a plurality of separate thyristor cathode layer segments which are at least partially enclosed in a plane parallel to the first main side by the gate metallization layer forming the plurality of gate electrodes and the connections therebetween. In the latter exemplary embodiment, the thyristor cathode layer segments of the plurality of thyristor units can be arranged at the first main side as strips placed in concentric rings around a lateral center of the semiconductor wafer, with the longitudinal main axis of each strip extending along a radial direction which is a direction extending from the lateral center of the semiconductor wafer and being parallel to the first main side. In this exemplary embodiment, abrupt commutation of the conduction current from the cathode to the gate is facilitated.

[0028] The reverse conducting power semiconductor device can be manufactured by the method according to claim 15. BRIEF DESCRIPTION OF DRAWINGS

[0029] In the following, detailed embodiments of the application will be explained with reference to the accompanying drawings, in which:

[0030] Figure 1 A top view of a bi-mode gate commutated thyristor (BGCT), a known turn-off power semiconductor device, is shown;

[0031] Figure 2 A cross-section of the BGCT taken along line c’c in Figure 1 is shown;

[0032] Figure 3 A top view of a reverse conducting power semiconductor device according to an embodiment of the application is shown;

[0033] Figure 4 An enlarged portion A of the top view of Figure 3 is shown;

[0034] Figure 5 A partial vertical cross-section along line B-B’ in Figure 4 is shown;

[0035] Figure 6 It shows Figure 3 The orthogonal projection of the strip freewheeling diode portion of the reverse-conducting power semiconductor device onto the horizontal plane;

[0036] Figure 7 The diagram shows lifetime control regions (proton radiation widths) of different widths. Figure 3 The peak step voltage of a reverse-conducting power semiconductor device;

[0037] Figure 8 This shows diode cathode layer segments of different widths. Figure 3 The peak step voltage of a reverse-conducting power semiconductor device;

[0038] Figure 9 A top view of a reverse-conducting power semiconductor device according to a second embodiment is shown; and

[0039] Figure 10A and Figure 10B Method steps for manufacturing a reverse-conducting power semiconductor device according to an embodiment of the present invention are shown.

[0040] The reference numerals used in the accompanying drawings and their meanings are summarized in the list of reference numerals. Generally, similar elements have the same reference numerals throughout the specification. The described embodiments are intended as examples and should not limit the scope of the invention. Detailed Implementation

[0041] In the following text, refer to Figures 3 to 6 A reverse-conducting power semiconductor device according to a first embodiment of the present invention is described. Figure 3 A top view of a reverse-conducting power semiconductor device is shown; Figure 4 It shows Figure 3 A magnified portion of the top view; Figure 5 It shows along Figure 4 The local vertical cross-section of line B-B' in the middle; and Figure 6 This shows an orthogonal projection of a portion of a reverse-conducting power semiconductor device onto a horizontal plane.

[0042] like Figure 5 As shown, the reverse-conduction power semiconductor device according to the first embodiment includes a semiconductor wafer 20 having a first main side 21 and a second main side 22. A plurality of thyristor units 50 and freewheeling diodes 60 are integrated in the semiconductor wafer 20. Each thyristor unit 50 includes, in order from the first main side 21 to the second main side 22:

[0043] • n-type thyristor cathode layer 51;

[0044] • a p-type base layer 52, wherein a first p-n junction is formed between the base layer 52 and the thyristor cathode layer 51;

[0045] • an n-type drift layer 53, which forms a second p-n junction with the base layer 52;

[0046] • an n-type buffer layer 55; and

[0047] • a p-type thyristor anode layer 54, which is separated from the base layer 52 by the drift layer 53.

[0048] The buffer layer 55 is optional and has a higher doping concentration than the drift layer 53. Each thyristor cell 50 further comprises a gate electrode 55, which is arranged laterally with respect to the thyristor cathode layer 51 and forms an ohmic contact with the base layer 52, a thyristor cathode electrode 56, which is arranged on the first main side 21 and forms an ohmic contact with the thyristor cathode layer 51, and a thyristor anode electrode 57, which is arranged on the second main side 22 and forms an ohmic contact with the thyristor anode layer 54.

[0049] The freewheeling diode 60 integrated in the semiconductor wafer 20 comprises:

[0050] • a p-type diode anode layer 32 at the first main side 21, which forms a third p-n junction with the drift layer 53 and is separated from the base layer 52 by the drift layer 53;

[0051] • a diode anode electrode 31 on the first main side 21, which is electrically connected to the diode anode layer 32;

[0052] • an n-type diode cathode layer 33 at the second main side 22, which is electrically connected to the drift layer 53 by the buffer layer 55 (throughout the specification, if two semiconductor regions of the same conductivity type are described as being electrically connected, it shall mean that the two semiconductor regions are either in direct contact, or connected to each other by one or more semiconductor regions of the same conductivity type, or connected to each other by a metal); and

[0053] • a diode cathode electrode 36 on the second main side 22, which forms an ohmic contact with the diode cathode layer 33.

[0054] The base layer 52 of the thyristor cell 50 is separated from the diode anode layer 32 by a separation region 70 comprising at least a portion of the drift layer 53.

[0055] The diode anode layer 32 comprises a plurality of first diode anode layer segments 321, which are arranged in a first diode anode layer segment row 322 and in a second diode anode layer segment row 323. Figure 3 in a top view and in a cross-sectional view along the line A-A in Figure 4In the partial top view, the strip portion 31a corresponds to the diode anode electrode 31 extending on the first main side 21 of the first diode anode layer segment 321. Figure 3 and Figure 4 In the image, the diode anode electrode 31 is shown in light gray. (See image.) Figure 3 and Figure 4 As shown, each first diode anode segment 321 has a longitudinal principal axis MA extending laterally away from the lateral center of the semiconductor wafer 20. The longitudinal principal axis MA can be defined as the direction in which the first diode anode segment 321 has its maximum extension. In an exemplary embodiment, the longitudinal principal axis MA forms the axis of symmetry of the first diode anode segment 321 about reflection in a top view (i.e., in an orthogonal projection onto a plane parallel to the second principal side 22). (i.e., the projection of the first diode anode segment 321 is mirror-symmetric with respect to the longitudinal principal axis MA).

[0056] In the first embodiment, the semiconductor wafer 20 has a circular shape, and the longitudinal main axis MA of each first diode anode layer segment 321 extends radially from the lateral center of the circular semiconductor wafer 20.

[0057] The strip-shaped first diode anode layer segment 321 has a first lateral width w1 in a vertical cross-section along a plane perpendicular to the second main side 22 and the longitudinal main axis MA of the first diode anode layer segment 321. It should be mentioned that, in Figure 5 In the diagram, only half of the strip-shaped portion of the freewheeling diode 60 corresponding to the anode layer segment 321 of the first diode is shown. Therefore, Figure 5The width 0.5 x w1 of the first diode anode layer segment 321 is shown in the middle. The first lateral width w1 of the first diode anode layer segment 321 can vary along the longitudinal main axis MA of the first diode anode layer segment 321. In the present application, the first lateral width w1 of each first diode anode layer segment 321 in a direction perpendicular to its longitudinal main axis MA is at least 1000 pm or at least 1200 pm at any location along the longitudinal axis MA (i.e. the minimum of the first lateral width w1 is at least 1000 pm or at least 1200 pm). The maximum of the first lateral width w1 of each first diode anode layer segment 321 can be less than 5 times or less than 4 times the minimum lateral width of that first diode anode layer segment 321. For example, the first lateral width w1 of each first diode anode layer segment 321 can be less than 5000 pm or less than 4000 pm or less than 3000 pm at any location along the longitudinal axis (i.e. the maximum of the first lateral width is less than 5000 pm or less than 4000 pm or less than 3000 pm). In exemplary embodiments, the minimum of the first lateral width w1 of each first diode anode layer segment 321 in a direction perpendicular to its longitudinal main axis MA is less than 2000 pm, such that the minimum of the first lateral width w1 is in the range between 1000 pm and 2000 pm or in the range between 1200 pm and 2000 pm.

[0058] As shown in Figure 3 each first diode anode layer segment 321 has a length in the direction along its longitudinal main axis MA which is exemplarily at least 20% or at least 25% of the width or diameter of the semiconductor wafer 20 in this direction.

[0059] The depth of each base layer 52 and the depth of the diode anode layer 32 are exemplarily the same as shown in Figure 4 .

[0060] The reverse conducting power semiconductor device according to the first embodiment further comprises a plurality of local lifetime control regions 91 in a region close to the p-n junction between the first diode anode layer segments 321 and the drift layer 53. Each local lifetime control region 91 comprises radiation induced recombination centers. In an orthogonal projection onto a plane parallel to the second main side 22, each local lifetime control region 91 is strip-shaped and is arranged within a corresponding one of the first diode anode layer segments 321 in such orthogonal projection such that a longitudinal main axis of each local lifetime control region 91 extends along the longitudinal main axis MA of one of the first diode anode layer segments 321 in the orthogonal projection. This means that each local lifetime control region 91 and the corresponding first diode anode layer segment 321 share the same longitudinal main axis MA. Exemplarily, the local lifetime control regions 91 have mirror symmetry with respect to the longitudinal main axis MA as a symmetry axis, similar to the corresponding first diode anode layer segments 321.

[0061] As Figure 5 illustrated, each local lifetime control region 91 has a second lateral width w2 in each vertical cross-section along a plane orthogonal to the second main side 22 and to the longitudinal main axis MA of the corresponding one of the first diode anode layer segments 321, which second lateral width is at least 200 pm or at least 300 pm smaller than the first lateral width wl of the corresponding one of the first diode anode layer segments 321 in this vertical cross-section.

[0062] In addition to the local lifetime control regions 91, (optional) additional second local lifetime control regions 92 can be arranged at a greater depth in the drift layer 53, as Figure 5 illustrated. The lateral width of such additional second local lifetime control regions 92 can be the same as the lateral width of the local lifetime control regions 91.

[0063] Referring to Figure 5 , the diode cathode layer 33 comprises a plurality of diode cathode layer segments 331, wherein each diode cathode layer segment 331 is bar-shaped in an orthogonal projection onto a plane parallel to the second main side 22 and is arranged within the corresponding one of the bar-shaped first diode anode layer segments 321 in this orthogonal projection such that a longitudinal main axis of each diode cathode layer segment 331 extends along the longitudinal main axis MA of the corresponding one of the first diode anode layer segments 321. This means that each diode cathode layer segment 331 and the corresponding first diode anode layer segment 321 share the same longitudinal main axis MA. Exemplarily, each diode cathode layer segment 331 has mirror symmetry with the longitudinal main axis MA as an axis of symmetry, similar to the mirror symmetry of the corresponding first diode anode layer segment 321. Each diode cathode layer segment 331 has a third lateral width w3 in each vertical cross-section along a plane orthogonal to the second main side 22 and to the longitudinal main axis MA of the corresponding one of the first diode anode layer segments 321, which third lateral width is at least 200 pm or at least 300 pm smaller than the second lateral width w2 of the corresponding one of the local lifetime control regions 91 in this cross-section, wherein the corresponding one of the local lifetime control regions 91 is arranged within the corresponding one of the first diode anode layer segments 321 in an orthogonal projection onto a plane parallel to the second main side 22.

[0064] This means that the following relation (1) is fulfilled:

[0065] wl > w2 > w3 (1)

[0066] This relation is illustrated in Figure 6The figure shows a top view (orthogonal projection onto a plane parallel to the second main side 22) of the cross-section of the first diode anode layer segment 321, the corresponding cross-section of the corresponding lifetime control region 91, and the corresponding cross-section of the diode cathode layer segment 331, which overlap each other in this orthogonal projection. Figure 6 In the cross-section shown, the first lateral width w1, the second lateral width w2, and the third lateral width w3 are shown as constant. However, as explained above, these lateral widths w1, w2, and w3 can vary along the longitudinal principal axis MA within the limitations described above. Each first diode anode layer segment 321, the corresponding lifetime control region 91, the corresponding diode cathode layer segment 331, and the finger portion 31a of the diode anode electrode 31 correspond to the strip portion of the freewheeling diode 60.

[0067] Alternatively, or in addition to the relationship indicated above between w3 and w2, each diode cathode segment 331 may have a third lateral width w3 in each vertical cross-section along a plane orthogonal to the longitudinal principal axis MA of the corresponding one of the first diode anode segments 321, the third lateral width being at least 600 μm or at least 800 μm smaller than the first lateral width w1 of the corresponding one of the first diode anode segments 321 in this cross-section.

[0068] As from Figure 3 As can be further seen in the top view, in addition to the first finger portion 31a corresponding to the first diode anode layer segment 321, the diode anode electrode 31 also includes a second finger portion 31b and a third finger portion 31c. In the first embodiment, the longitudinal main axis of all these finger portions 31a, 31b and 31c extends radially from the lateral center of the semiconductor wafer 20.

[0069] In the lateral center of the reverse conducting power semiconductor device, the diode anode electrode 31 comprises a circular portion 31d from which first finger portions 31a extend in a radial direction. Second finger portions 31b extend in the radial direction between two adjacent first finger portions 31a, respectively. The second finger portions 31b are shorter than the first finger portions 31a and are separated from the circular portion 31d. Third finger portions 31c extend in the radial direction between a first finger portion 31a and the nearest second finger portion 31b. The third finger portions 31c are shorter than the first finger portions 31a and also shorter than the second finger portions 31b. The third finger portions 31c are separated from the circular portion 31d by a greater distance than the second finger portions 31b. Finally, fourth finger portions 31e are arranged between each pair of directly adjacent first to third finger portions 31a to 31c at the circumferential edge of the semiconductor wafer 20. As discussed above, each first finger portion 31a corresponds to a first diode anode layer segment 321 and each has a longitudinal main axis which, in an orthogonal projection onto a plane parallel to the second main side 22, extends along the longitudinal main axis MA of the corresponding one of the first diode anode layer segments 321. Likewise, the second, third and fourth finger portions 31b, 31c and 31e correspond to the bar-shaped diode anode layer segments 321, the bar-shaped lifetime control regions 91 and the bar-shaped diode cathode layer segments 331 on which they extend. In a vertical cross-section, each finger portion of the freewheel diode 60 corresponding to each of the second, third and fourth finger portions 31b, 31c and 31e of the diode anode electrode 31 has the same structure as described above with reference to the finger portions of the diode anode electrode 31. Figure 5 The same structure is described for the finger portions of the freewheel diode 60 corresponding to the first finger portions 31a. The bar-shaped diode anode layer segments 321 corresponding to the second finger portions 31b are referred to as second diode anode layer segments 321.

[0070] The thyristor cathode layer 51 of the plurality of thyristor units 50 comprises a plurality of thyristor cathode layer segments 511 which are bar-shaped and separated from each other. On the thyristor cathode layer segments 511, corresponding finger portions of the thyristor cathode electrode 56 are formed. In a top view, the finger portions of the thyristor cathode electrode 56 have the same or a corresponding pattern and shape as the thyristor cathode layer segments 511. As can best be seen from Figure 3 In a top view of the reverse conducting power semiconductor device, the thyristor cathode layer segments 511 are arranged at the first main side 21 in concentrically placed bars around the lateral center of the semiconductor wafer 20, wherein the longitudinal main axis of each bar extends along a radial direction which is a direction extending from the lateral center of the semiconductor wafer 20 and parallel to the first main side 21. In each ring, the thyristor cathode layer segments 511 have the same distance to the lateral center of the semiconductor wafer 20. In the radial direction, the thyristor cathode layer segments 511 are separated from each other by a distance which is greater than the width of the thyristor cathode layer segments 511. Figure 3In the exemplary embodiment shown in Fig. 1, the thyristor cathode layer segments 511 are arranged in twelve concentric rings. However, the number of concentric rings can be any other number. Exemplarily, a number of concentric rings is provided. In an orthogonal projection onto a plane parallel to the second main side 22, in each of the concentric rings, a group of the thyristor cathode layer segments 511 alternates with the finger portions 31a to 31c, 31e. In Figure 3 In the example shown, in the first three inner concentric rings in which the thyristor cathode layer segments 511 are arranged, the groups of directly adjacent thyristor cathode layer segments 511 alternate with the first finger portion 31a along each ring, respectively. In the next three concentric rings, the groups of directly adjacent thyristor cathode layer segments 511 alternate with the first and second finger portions 31a and 31b along each ring, respectively. Finally, in the next five concentric rings in which the thyristor cathode layer segments 511 are arranged, the groups of directly adjacent thyristor cathode layer segments 511 alternate with the first, second and third finger portions 31a, 31b and 31c along each ring. In the outermost ring in which the thyristor cathode layer segments 511 are arranged, two groups of directly adjacent thyristor cathode layer segments 511 alternate with the first to fourth finger portions 31a, 31b, 31c and 31e along this outermost ring, respectively. In Figure 3 and Figure 4 In a top view of

[0071] As can be best seen in Figure 3 and Figure 4 , in the first embodiment, the gate electrodes 55 of the plurality of thyristor cells 50 are electrically connected to each other, and the reverse conducting power semiconductor device further comprises a common gate contact 40 for contacting the gate electrodes 55 of the plurality of thyristor cells 50, wherein the common gate contact 40 is arranged on the circumferential edge of the semiconductor wafer 20 on the first main side 21. The first diode anode layer segments 321 extend in radial direction through all concentric rings in which the thyristor cathode layer segments 511 are arranged. In such exemplary embodiment, the flow of gate current from the thyristor cells 50 to the common gate contact 40 at the circumferential edge of the semiconductor wafer 20 (or from the common gate contact 40 at the circumferential edge of the semiconductor wafer 20 to the thyristor cells 50) is not blocked by the finger portions of the freewheel diodes corresponding to the bar-shaped first diode anode layer segments 321. This is particularly advantageous for current commutation during turn-off.

[0072] In Figure 9 , a top view of a reverse conducting power semiconductor according to a second embodiment of the present application is shown. In view of many similarities, only the differences to the first embodiment will be described in the following. With respect to the remaining features, reference is made to the above description of the first embodiment. In Figure 9In the middle, the region in light gray color corresponds to the region of the diode anode electrode 31. Four first finger portions 301a and four second finger portions 301b of the diode anode electrode 31 extend in radial direction from the lateral center of the semiconductor wafer 20. An additional third finger portion 301c extends in lateral direction from each of the four finger portions 301a away from the lateral center of the semiconductor wafer 20. The first and second finger portions 301a and 301b are arranged rotationally symmetrically. Each of the finger portions 301a to 301c corresponds to a bar portion of the freewheeling diode 60 and has a length of 200 pm, a width of 1200 pm and a height of 200 pm, as indicated by the dimensions in the cross-sectional view in the middle of Fig. 6. Figure 5 In the middle, the region in light gray color corresponds to the region of the diode anode electrode 31. Four first finger portions 301a and four second finger portions 301b of the diode anode electrode 31 extend in radial direction from the lateral center of the semiconductor wafer 20. An additional third finger portion 301c extends in lateral direction from each of the four finger portions 301a away from the lateral center of the semiconductor wafer 20. The first and second finger portions 301a and 301b are arranged rotationally symmetrically. Each of the finger portions 301a to 301c corresponds to a bar portion of the freewheeling diode 60 and has a length of 200 pm, a width of 1200 pm and a height of 200 pm, as indicated by the dimensions in the cross-sectional view in the middle of Fig. 6. Figure 5 The thyristor unit 50 in the reverse conducting power semiconductor device according to the second embodiment differs from the thyristor unit 50 in the first embodiment only in the different arrangement in top view. In the second embodiment, the cathode layer segments 511, which are shown as black bars in the middle, are arranged along the longitudinal main axes of the bars extending laterally between directly adjacent finger portions 301b, 301c, respectively. This means that the cathode layer segments 511 extend in a direction which is inclined with respect to the radial direction. The structure of the thyristor unit 50 of the reverse conducting power semiconductor device according to the second embodiment is basically the same as described above with reference to Figure 9 the cross-sectional view in the middle and above. Figure 5 the cross-sectional view in the middle and above. Figure 5 the cross-sectional view in the middle and above.

[0073] Figure 7 A graph illustrating the dependence of the step voltage VRMsn of the bar-shaped freewheeling diode 60 on the second lateral width w2 of the lifetime control region 91 is shown for a freewheeling diode 60 having a first lateral width w1 of 1200 pm and a third lateral width w3 of 200 pm. Different symbols, i.e. triangle, solid circle and open circle, correspond to different proton irradiation doses, which are shown in the middle as 1 E+13, 2 E+13 and 4 E+14, respectively. At a second lateral width w2 of the lifetime control region of about w2 = 800 pm, the step voltage VRMsn has its minimum value and starts to increase again with increasing second lateral width w2. Figure 7 A graph illustrating the dependence of the step voltage VRMsn of the bar-shaped freewheeling diode 60 on the second lateral width w2 of the lifetime control region 91 is shown for a freewheeling diode 60 having a first lateral width w1 of 1200 pm and a third lateral width w3 of 200 pm. Different symbols, i.e. triangle, solid circle and open circle, correspond to different proton irradiation doses, which are shown in the middle as 1 E+13, 2 E+13 and 4 E+14, respectively. At a second lateral width w2 of the lifetime control region of about w2 = 800 pm, the step voltage VRMsn has its minimum value and starts to increase again with increasing second lateral width w2. Figure 8 A graph illustrating the dependence of the step voltage VRMsn of the bar-shaped freewheeling diode 60 on the second lateral width w2 of the lifetime control region 91 is shown for a freewheeling diode 60 having a first lateral width w1 of 1200 pm and a third lateral width w3 of 200 pm. Different symbols, i.e. triangle, solid circle and open circle, correspond to different proton irradiation doses, which are shown in the middle as 1 E+13, 2 E+13 and 4 E+14, respectively. At a second lateral width w2 of the lifetime control region of about w2 = 800 pm, the step voltage VRMsn has its minimum value and starts to increase again with increasing second lateral width w2. Figure 8 A graph illustrating the dependence of the step voltage VRMsn of the bar-shaped freewheeling diode 60 on the second lateral width w2 of the lifetime control region 91 is shown for a freewheeling diode 60 having a first lateral width w1 of 1200 pm and a third lateral width w3 of 200 pm. Different symbols, i.e. triangle, solid circle and open circle, correspond to different proton irradiation doses, which are shown in the middle as 1 E+13, 2 E+13 and 4 E+14, respectively. At a second lateral width w2 of the lifetime control region of about w2 = 800 pm, the step voltage VRMsn has its minimum value and starts to increase again with increasing second lateral width w2.

[0074] In the following, a method for manufacturing a reverse conducting power semiconductor device according to any one of the above-described embodiments is shown with reference to Figure 10A and Figure 10B The method comprises the following steps:

[0075] (a) providing an n-type semiconductor layer 200, the semiconductor layer 200 having a first side 201 and having a second side 202 opposite the first side 201, and wherein the doping concentration of the semiconductor layer 200 corresponds to the doping concentration of the drift layer 53 in the final reverse conducting power semiconductor device;

[0076] (b) providing a first mask layer 210 on the first side 201 of the semiconductor layer 200, the first mask layer 210 comprising first openings 211;

[0077] (c) selectively applying p-type dopants 220 to the first side 201 of the semiconductor layer 200 through the first openings 211 of the first mask layer 210 for forming the base layer 52 and the diode anode layer 32 of the final reverse conducting semiconductor device;

[0078] (e) forming a second mask layer 240 on the first side 201, the second mask layer 240 having second openings 241;

[0079] (f) selectively applying radiation 250, such as electron or heavy ion radiation, to the first side 201 through the second openings 241 in the second mask layer 240 for forming a local lifetime control region 91 comprising radiation induced recombination centers in the semiconductor layer 200;

[0080] (g) forming the thyristor cathode layer 51 on the first side 201;

[0081] (h) forming the thyristor anode layer 54 and the diode cathode electrode 36 on the second side 202; and

[0082] (i) forming the diode cathode layer 33 on the second side 202;

[0083] The second openings 241 comprise strip-shaped openings which are laterally aligned with regions of the semiconductor layer 200 in which first diode anode layer segments 321 are to be formed, such that in an orthogonal projection onto a plane parallel to the first side 201 each second opening 241 is arranged within the projected area of a corresponding one of the first diode anode layer segments 321, the strip-shaped openings 241 having a lateral width w4 which is at least 200 pm or at least 300 pm smaller than the lateral width wl of the corresponding one of the first diode anode layer segments 321 at all positions along the longitudinal main axis MA of the corresponding one of the first diode anode layer segments 321 in the final reverse conducting power semiconductor device.

[0084] It will be readily apparent to a person skilled in the art that modifications of the above-described embodiments are possible without departing from the idea of the application as defined by the appended claims.

[0085] The above described embodiments are explained with a specific conductivity type. The conductivity type of the semiconductor layers in the above described embodiments can be switched such that all layers described as p-type layers are n-type layers and all layers described as n-type layers are p-type layers.

[0086] The above described embodiments of the reverse conducting power semiconductor device are described with a circular semiconductor wafer 20. However, the semiconductor wafer can have any other shape, such as a rectangular or a polygonal shape.

[0087] According to the method of the above described embodiments with reference to Figure 10A and Figure 10B different method steps are described in a certain order. However, the method steps can be performed in a different order.

[0088] It should be noted that the term "comprising" does not exclude other elements or steps and the use of "a" or "an" does not exclude a plurality. Also individual elements described in association with different embodiments can be combined.

[0089] List of reference signs

[0090] 1, 20 semiconductor wafer

[0091] 2, 50 thyristor cell

[0092] 3, 56 thyristor cathode electrode

[0093] 4, 511 thyristor cathode layer section

[0094] 5, 52 base layer

[0095] 6, 53 drift layer

[0096] 7, 59 buffer layer

[0097] 8, 54 thyristor anode layer

[0098] 9, 57 thyristor anode electrode

[0099] 10, 55 gate electrode

[0100] 11, 40 common gate contact

[0101] 12 diode cell

[0102] 13, 32 diode anode layer

[0103] 14, 33 diode cathode layer

[0104] 15, 70 separation region

[0105] 16, 36 diode cathode electrode

[0106] 17, 31 diode anode electrode

[0107] 21 first main side

[0108] 22 second main side

[0109] 200 semiconductor layer

[0110] 201 first side

[0111] 202 second side

[0112] 210 first mask layer

[0113] 211 first opening

[0114] 220 p-type dopant

[0115] 240 second mask layer

[0116] 241 second opening

[0117] 31a, 31b, 31c, 31e;

[0118] 301a, 301b, 301c finger portion of diode anode electrode 31

[0119] 31d center portion of diode anode electrode 31

[0120] 321 diode anode layer segment

[0121] 331 diode cathode layer segment

[0122] 51 thyristor cathode layer

[0123] 60 freewheeling diode

[0124] 91 local lifetime control (LLC) region

[0125] 92 second local lifetime control region

[0126] w1 first lateral width (of 321)

[0127] w2 second lateral width (of 91)

[0128] w3 third lateral width (of 331)

[0129] w4 lateral width (of 241)

Claims

1. A reverse-conducting power semiconductor device, comprising: Semiconductor wafer (20) having a first main side (21) and a second main side (22) opposite to the first main side (21); Multiple thyristor units (50); and Freewheeling diode (60) Each of the plurality of thyristor units (50) comprises, in order from the first main side (21) to the second main side (22): First conductivity type of thyristor cathode layer (51); A base layer (52) of a second conductivity type, different from the first conductivity type, wherein a first pn junction is formed between the base layer (52) and the thyristor cathode layer (51); A drift layer (53) of the first conductivity type, wherein the drift layer and the base layer (52) form a second pn junction; and The second conductivity type of thyristor anode layer (54) is separated from the base layer (52) by the drift layer (53). Each thyristor unit (50) further includes: a gate electrode (55) arranged laterally relative to the thyristor cathode layer (51) and forming an ohmic contact with the base layer (52); a thyristor cathode electrode (56) arranged on the first main side (21) and forming an ohmic contact with the thyristor cathode layer (51); and a thyristor anode electrode (57) arranged on the second main side (22) and forming an ohmic contact with the thyristor anode layer (54). The freewheeling diode (60) mentioned above includes: A diode anode layer (32) of the second conductivity type is located at the first main side (21), the diode anode layer and the drift layer (53) form a third pn junction, and are separated from the base layer (52) through the drift layer (53); A diode anode electrode (31) is located on the first main side (21), and the diode anode electrode is electrically connected to the diode anode layer (32). A diode cathode layer (33) of a first conductivity type is located on the second main side (22), the diode cathode layer being electrically connected to the drift layer (53); and A diode cathode electrode (36) is located on the second main side (22), and the diode cathode electrode forms an ohmic contact with the diode cathode layer (33). The diode anode layer (32) comprises a plurality of first diode anode layer segments (321), which are strip-shaped in orthogonal projection onto a plane parallel to the second main side (22). The longitudinal principal axis (MA) of each first diode anode layer segment (321) extends laterally away from the lateral center of the semiconductor wafer (20), wherein the lateral direction relates to a direction parallel to the second main side (22). Its features The first lateral width (w1) of each first diode anode layer segment (321) in a plane parallel to the second main side (22) and in a direction perpendicular to its longitudinal main axis (MA) is at least 1000 µm or at least 1200 µm at any position along the longitudinal axis. The reverse-conducting power semiconductor device includes multiple local lifetime control regions (91) containing radiation-induced recombination centers. Each local lifetime control region (91) is strip-shaped in an orthogonal projection onto a plane parallel to the second main side (22), and is arranged in the orthogonal projection within a corresponding one of the first diode anode layer segments (321), such that the longitudinal principal axis of each local lifetime control region (91) extends along the longitudinal principal axis (MA) of the corresponding one of the first diode anode layer segments (321). Each local lifetime control region (91) has a second lateral width (w2) that is at least 200 µm or at least 300 µm smaller than the first lateral width (w1) of the corresponding one in the first diode anode layer segment (321). The diode cathode layer (33) includes multiple diode cathode layer segments (331). Each diode cathode segment (331) is strip-shaped in the orthogonal projection onto the plane parallel to the second main side (22), and is arranged in the orthogonal projection within a corresponding one of the strip-shaped first diode anode segments (321), such that the longitudinal principal axis of each diode cathode segment (331) extends along the longitudinal principal axis (MA) of the corresponding one of the first diode anode segments (321), and Each diode cathode segment (331) has a third lateral width (w3) that is at least 200 µm or at least 300 µm smaller than the second lateral width (w2) of the corresponding one of the local lifetime control regions (91), wherein the corresponding one of the local lifetime control regions (91) is arranged in the orthogonal projection on the plane parallel to the second main side (22) within the corresponding one of the first diode anode segments (321).

2. The reverse-conduction power semiconductor device according to claim 1, wherein the semiconductor wafer (20) has a circular shape, and the longitudinal main axis (MA) of each first diode anode layer segment (321) extends along the radial direction of the circular shape.

3. The reverse-conducting power semiconductor device according to claim 1, wherein the first lateral width (w1) of each first diode anode segment (321) is less than 5000 µm, less than 4000 µm, or less than 3000 µm at any location along the longitudinal main axis (MA).

4. The reverse-conducting power semiconductor device according to any one of claims 1 to 3, wherein in the orthogonal projection on a plane parallel to the second main side (22), a circular portion of the freewheeling diode (60) is arranged in the lateral center of the semiconductor wafer (20), and each first diode anode layer segment (321) extends laterally from the circular portion.

5. The reverse-conducting power semiconductor device according to any one of claims 1 to 3, wherein the length of each first diode anode segment (321) in the direction along its longitudinal main axis (MA) is at least 20% or at least 25% of the width of the semiconductor wafer (20) in that direction.

6. The reverse-conducting power semiconductor device according to any one of claims 1 to 3, wherein the diode anode layer (32) comprises a strip-shaped second diode anode layer segment (321) extending in a radial direction and laterally arranged between two adjacent first diode anode layer segments (321), wherein the distance between each second diode anode layer segment (321) and the lateral center of the semiconductor wafer (20) is greater than the distance between each first diode anode layer segment (321) and the lateral center of the semiconductor wafer (20).

7. The reverse-conducting power semiconductor device according to any one of claims 1 to 3, wherein the minimum value of the first lateral width (w1) of each first diode anode segment (321) is less than 2000 µm.

8. The reverse-conducting power semiconductor device according to any one of claims 1 to 3, wherein the third lateral width (w3) is at least 600 µm or at least 800 µm smaller than the first lateral width (w1) of the corresponding one of the first diode anode layer segments (321).

9. The reverse-conducting power semiconductor device according to any one of claims 1 to 3, wherein the depth of each base layer (52) is the same as the depth of the diode anode layer (32).

10. The reverse-conduction power semiconductor device according to any one of claims 1 to 3, wherein the gate electrodes (55) of the plurality of thyristor units (50) are electrically connected to each other, and the device further includes a common gate contact (40) for contacting the gate electrodes (55) of the plurality of thyristor units (50), wherein the common gate contact (40) is disposed on the circumferential edge of the semiconductor wafer (20) on the first main side (21).

11. The reverse-conducting power semiconductor device according to any one of claims 1 to 3, wherein the thyristor cathode layer (51) comprises a plurality of discretely shaped thyristor cathode layer segments (511), the thyristor cathode layer segments being at least partially surrounded by a gate metallization layer in a plane parallel to the first main side (21), the gate metallization layer forming the plurality of gate electrodes (55) and the connections therebetween.

12. The reverse-conduction power semiconductor device according to claim 11, wherein the thyristor cathode segments (511) of the plurality of thyristor units (50) are arranged on the first main side (21) as strips placed in a concentric ring around the lateral center of the semiconductor wafer (20), the longitudinal main axis of each strip extending in a radial direction, the radial direction being a direction extending from the lateral center of the semiconductor wafer (20) and parallel to the first main side (21).

13. A method for manufacturing a reverse-conducting power semiconductor device, the semiconductor device comprising: Semiconductor wafer (20) having a first main side (21) and a second main side (22) opposite to the first main side (21); Multiple thyristor units (50); and Freewheeling diode (60) Each of the plurality of thyristor units (50) comprises, in order from the first main side (21) to the second main side (22): First conductivity type of thyristor cathode layer (51); A base layer (52) of a second conductivity type, different from the first conductivity type, wherein a first pn junction is formed between the base layer (52) and the thyristor cathode layer (51); A drift layer (53) of the first conductivity type, wherein the drift layer and the base layer (52) form a second pn junction; and The second conductivity type of thyristor anode layer (54) is separated from the base layer (52) by the drift layer (53). Each thyristor unit (50) further includes: a gate electrode (55) arranged laterally relative to the thyristor cathode layer (51) and forming an ohmic contact with the base layer (52); a thyristor cathode electrode (56) arranged on the first main side (21) and forming an ohmic contact with the thyristor cathode layer (51); and a thyristor anode electrode (57) arranged on the second main side (22) and forming an ohmic contact with the thyristor anode layer (54). The freewheeling diode (60) mentioned above includes: A diode anode layer (32) of the second conductivity type is located at the first main side (21), the diode anode layer and the drift layer (53) form a third pn junction, and are separated from the base layer (52) through the drift layer (53); A diode anode electrode (31) is located on the first main side (21), and the diode anode electrode is electrically connected to the diode anode layer (32). A diode cathode layer (33) of a first conductivity type is located on the second main side (22), the diode cathode layer being electrically connected to the drift layer (53); and A diode cathode electrode (36) is located on the second main side (22), and the diode cathode electrode forms an ohmic contact with the diode cathode layer (33). The diode anode layer (32) comprises a plurality of first diode anode layer segments (321), which are strip-shaped in orthogonal projection onto a plane parallel to the second main side (22). The longitudinal principal axis (MA) of each first diode anode layer segment (321) extends laterally away from the lateral center of the semiconductor wafer (20), wherein the lateral direction relates to a direction parallel to the second main side (22). The first lateral width (w1) of each first diode anode layer segment (321) in a plane parallel to the second main side (22) and in a direction perpendicular to its longitudinal main axis (MA) is at least 1000 µm or at least 1200 µm at any position along the longitudinal axis. The reverse-conducting power semiconductor device includes multiple local lifetime control regions (91) containing radiation-induced recombination centers. Each local lifetime control region (91) is strip-shaped in an orthogonal projection onto a plane parallel to the second main side (22), and is arranged in the orthogonal projection within a corresponding one of the first diode anode layer segments (321), such that the longitudinal principal axis of each local lifetime control region (91) extends along the longitudinal principal axis (MA) of the corresponding one of the first diode anode layer segments (321). Each local lifetime control region (91) has a second lateral width (w2) that is at least 200 µm or at least 300 µm smaller than the first lateral width (w1) of the corresponding one in the first diode anode layer segment (321). The diode cathode layer (33) includes multiple diode cathode layer segments (331). Each diode cathode segment (331) is strip-shaped in the orthogonal projection onto the plane parallel to the second main side (22), and is arranged in the orthogonal projection within a corresponding one of the strip-shaped first diode anode segments (321), such that the longitudinal principal axis of each diode cathode segment (331) extends along the longitudinal principal axis (MA) of the corresponding one of the first diode anode segments (321), and Each diode cathode segment (331) has a third lateral width (w3) that is at least 200 µm or at least 300 µm smaller than the second lateral width (w2) of the corresponding one of the local lifetime control regions (91), wherein the corresponding one of the local lifetime control regions (91) is arranged within the corresponding one of the first diode anode segments (321) in the orthogonal projection onto a plane parallel to the second main side (22). The method includes the following steps: (a) A semiconductor layer (200) of the first conductivity type is provided, the semiconductor layer (200) having a first side (201) and a second side (202) opposite to the first side (201), and wherein the doping concentration of the semiconductor layer (200) corresponds to the doping concentration of the drift layer (53) in the reverse conduction power semiconductor device; (b) A first mask layer (210) is provided on the first side (201) of the semiconductor layer (200), the first mask layer (210) including a first opening (211); (c) A dopant of the second conductivity type is selectively applied to the first side (201) of the semiconductor layer (200) through the first opening (211) of the first mask layer (210) to form the base layer (52) and the diode anode layer (32). (e) A second mask layer (240) is formed on the first side (201), the second mask layer (240) having a second opening (241); (f) Radiation (250) is selectively applied to the first side (201) through the second opening (241) in the second mask layer (240) to form a local lifetime control region (91) including radiation-induced recombination centers in the semiconductor layer (200). (g) The thyristor cathode layer (51) is formed on the first side (201); (h) The thyristor anode layer (54) and the diode cathode electrode (36) are formed on the second side (202); and (i) The diode cathode layer (33) is formed on the second side (202); The second opening (241) includes a strip opening that is laterally aligned with the region of the semiconductor layer (200) where the first diode anode segment (321) is formed, such that in an orthogonal projection onto a plane parallel to the first side (201), each second opening (241) is arranged within the projection area of ​​a corresponding one of the first diode anode segments (321), the strip opening having a fourth lateral width (w4) that is at least 200 µm or at least 300 µm smaller than the first lateral width (w1) of the corresponding one of the first diode anode segments (321) at all locations along the longitudinal principal axis (MA) of the corresponding one of the first diode anode segments (321).

Images