BIPOLAR TRANSISTOR ELEMENT WITH AN EMISSOR THAT FEATURES TWO TYPES OF EMISSOR AREAS

Abstract

Bipolar semiconductor device that features: a semiconductor body (100) having a first surface (101), and a base region (10) of a first doping type and a first emitter region (20) in the semiconductor body (100), wherein the first emitter region (20) adjoins the first surface (101) and comprises: a plurality of emitter regions (21) of a first type of a second doping type complementary to the first doping type, a plurality of emitter regions (22) of a second type of the second doping type, a plurality of emitter regions (23) of a third type of the first doping type, and a recombination region (24) comprising recombination centers, wherein the emitter regions (21) of the first type and the emitter regions of the second type (22) extend from the first surface (101) into the semiconductor body (100), wherein the emitter regions (21) of the first type have a higher doping concentration and extend from the first surface (101) deeper into the semiconductor body (100) than the emitter regions (22) of the second type, where the emitter areas (23) of the third type border the emitter areas (21) of the first type and the emitter areas (22) of the second type, wherein the recombination area (24) is located at least in the first-type emitter areas (21) and the third-type emitter areas (23), and wherein the ratio between the charge carrier lifetime in the recombination area (24) in the emitter areas (21) of the first type and the charge carrier lifetime in the recombination area (24) in the emitter areas (23) of the third type is between 2 and 4.

Description

[0001] This disclosure generally relates to a bipolar transistor device such as an IGBT (Insulated Gate Bipolar Transistor).

[0002] IGBTs (Insulated Gate Bipolar Transistors) are frequently used to switch various electrical loads. For example, IGBTs can be used in power conversion applications, electric drive applications, or lighting applications, to name just a few.

[0003] An IGBT is a voltage-controlled MOS transistor device that has a collector region (often called the drain region) and an emitter region (often called the source region) with different doping types (conductivity types). An IGBT also contains a gate electrode, which is dielectrically insulated from the body region by a gate dielectric adjacent to the body region and extends from the emitter region to a base region (drift region). The base region is located between the body region and the collector region. When the IGBT is switched on, the gate electrode creates a conductive channel in the body region between the emitter region and the base region, allowing the emitter region to inject charge carriers of a first conductivity type into the drift region.Simultaneously, the collector region injects charge carriers of a second conductivity type into the base region, with the charge carriers of the first and second conductivity types forming a charge carrier plasma in the base region. This charge carrier plasma leads to comparatively low conduction losses of the IGBT.

[0004] Relevant operating parameters of an IGBT are the saturation voltage (often denoted as Vs). CEsatThe switching losses (often referred to as Eoff) and the current withstand capability (which can also be called short-circuit capability) are important factors. The latter can be defined either by the current level of the maximum current that the IGBT can conduct for a specific, predetermined duration (e.g., 10 µs) without being destroyed, or by the current level required to destroy the IGBT. In a conventional IGBT design, there is a trade-off between the saturation voltage and the switching losses, such that the switching losses increase as the saturation voltage decreases, and vice versa. There is also a trade-off between the switching losses and the current withstand capability, such that the current withstand capability increases as the switching losses increase.

[0005] DE 196 30 341 B4 describes an IGBT with a p-emitter having highly doped and less doped emitter regions, each connected to an emitter electrode. DE 10 2007 040 587 A1 describes an IGBT with a basal region in which a recombination region is arranged.

[0006] German patent DE 10 2006 023 171 A1 describes a power thyristor with at least one MOS field-effect transistor structure integrated in the edge region of an anode emitter. The MOS field-effect transistor structure comprises a p-doped region adjacent to a base region and having a more heavily and a less heavily doped section, an n-doped region embedded in the more heavily doped section, and a control electrode. The control electrode is located above the p-doped and n-doped regions and is insulated from them by a dielectric.

[0007] The JP S61-15 370 A describes a transistor with a p-doped emitter located on the back side of a semiconductor body. This emitter has alternating sections that extend to different depths into the semiconductor body.

[0008] US patent 2015 / 0021657A1 describes an IGBT with transistor cells arranged in the front region of a semiconductor body, an n-doped drift region, a p-doped probe in the back region of the semiconductor body, and an n-doped buffer region between the drift region and the probe. The probe has adjacent p-doped regions with different doping concentrations.

[0009] The objective of the present invention is to provide a bipolar semiconductor device, such as an IGBT, with increased current capability. This objective is achieved by a bipolar semiconductor device according to claim 1.

[0010] The bipolar semiconductor device comprises a semiconductor body with a first surface and a base region of a first doping type, and a first emitter region within the semiconductor body. The first emitter region borders the first surface and contains a plurality of emitter regions of a first type with a second doping type complementary to the first, a plurality of emitter regions of a second type with a second doping type, a plurality of emitter regions of a third type with a first doping type, and a recombination region containing recombination centers. The emitter regions of the first type and the emitter regions of the second type extend from the first surface into the semiconductor body.The emitter regions of the first type have a higher doping concentration and extend from the first surface deeper into the semiconductor body than the emitter regions of the second type, and the emitter regions of the third type border the emitter regions of the first type and the emitter regions of the second type, and the recombination region is located at least in the emitter regions of the first type and the emitter regions of the third type.

[0011] Examples are explained below with reference to the drawings. The drawings serve to illustrate certain principles, so only the aspects necessary for understanding these principles are shown. The drawings are not to scale. In the drawings, the same reference symbols denote the same features. Fig. Figure 1 shows a vertical cross-sectional view of a section of a semiconductor body having a first emitter region of a bipolar semiconductor device, according to a first embodiment; Fig. Figures 2A-2B illustrate doping concentrations of the in Fig. 1 semiconductor areas shown; Fig. Figure 3 shows a vertical cross-sectional view of a section of a semiconductor body having a first emitter region and a field-stop region, according to an embodiment; Fig. Figures 4A-4B illustrate doping concentrations of the in Fig. 3 semiconductor regions shown; Fig. Figure 5 shows elongated emitter regions of a first type and elongated emitter regions of a second type of the first emitter region in a horizontal plane of the semiconductor body; Fig. Figure 6 shows ring-shaped emitter regions of the first type and ring-shaped emitter regions of the second type of the first emitter region in a horizontal plane of the semiconductor body; Fig. Figure 7 shows grid-like emitter regions of the first type and rectangular emitter regions of the second type of the first emitter region in a horizontal plane of the semiconductor body; Fig. Figure 8 shows rectangular emitter regions of the first type and rectangular emitter regions of the second type of the first emitter region in a horizontal plane of the semiconductor body; Fig. Figure 9 shows a vertical cross-sectional view of a bipolar semiconductor device implemented as a diode; Fig. Figure 10 shows a vertical cross-sectional view of a bipolar semiconductor device implemented as an IGBT; Fig. Figure 11 shows an inner region and a boundary region of the semiconductor body; Fig. Figure 12 shows switching losses compared to the saturation voltages of different IGBTs implemented with different first emitter regions; Fig. Figure 13 shows the current withstand capability (destructive currents) against the saturation voltages of various IGBTs implemented with different first emitter regions; Fig. 14A-14B show an example of a method for generating emitter regions of the first type and emitter regions of the second type; Fig. 15 shows an example of a method for creating a field stop zone; Fig. 16 shows the generation of at least one emitter region of the third type; Fig. Figure 17 shows an embodiment of the emitter region of the third type; and Fig. Figure 18 shows further examples of the at least one third emitter region.

[0012] Fig. Figure 1 shows a vertical cross-sectional view of a section of a bipolar semiconductor device. Referring to Fig. 1 The semiconductor device contains a semiconductor body 100 with a first surface 101. The semiconductor body 100 also contains a second surface opposite the first surface 101. However, this second surface is located outside the view shown in Figure 1. Fig. 1. Fig. Figure 1 shows a vertical cross-sectional view of the semiconductor body 100, that is, a view in a section plane perpendicular to the first surface 101. The semiconductor body 100 may comprise a conventional semiconductor material such as silicon (Si), silicon carbide (SiC), gallium arsenide (GaAs), gallium nitride (GaN), or the like. The following examples of doping concentrations and dopant doses refer to a semiconductor body 100 comprising silicon. However, these doping concentrations and doses can easily be adapted to a semiconductor body 100 comprising a material other than silicon. For example, the doping concentrations and doses mentioned below can easily be adapted for use in a semiconductor body containing SiC by multiplying the doping concentrations by 10 (1E1).Similarly, the dimensions mentioned below refer to a semiconductor device that has a silicon-based semiconductor body. If these dimensions refer to dimensions in a current-flow direction of the device, they can easily be adapted for use in a SiC-based semiconductor body by dividing the dimensions by 10 (1E1).

[0013] The semiconductor device has a base region 10 (which can also be called a drift region) of a first doping type (conductivity type) and a first emitter region 20. The first emitter region 20 borders the first surface 101 and has a plurality of emitter regions 21 of a first type and a plurality of emitter regions 22 of a second type. The emitter regions 21 of the first type and the emitter regions 22 of the second type are doped regions of a second doping type complementary to the first doping type. The first emitter region 20 also contains a plurality of emitter regions 23 of a third type of the first doping type, and a recombination region 24, which contains recombination centers. Each of the plurality of emitter regions 21 of the first type and each of the plurality of emitter regions 22 of the second type extends from the first surface 101 into the semiconductor body 100.The semiconductor device can further comprise a first electrode 31 on the first surface 101. This first electrode 31 contacts the emitter regions 21 of the first type and the emitter regions 22 of the second type. According to one embodiment, the doping concentrations of the emitter regions 21 of the first type and the emitter regions 22 of the second type on the first surface 101 are such that an ohmic contact exists between the first electrode 31 and these regions 21, 22.

[0014] The emitter regions 21 of the first type extend deeper into the semiconductor body 100 than the emitter regions 22 of the second type. In Fig. In 1, d1 denotes a dimension of the emitter regions 21 of the first type in the vertical direction of the semiconductor body 100. The "vertical direction" of the semiconductor body 100 is a direction perpendicular to the first surface 101. Similarly, d2 denotes a dimension of the emitter regions 22 of the second type in the vertical direction of the semiconductor body 100. According to one embodiment, d1 is between 100 nanometers (nm) and 2 micrometers (µm), in particular between 300 nm and 1 µm, and d2 is between 50 nm and 1 µm, in particular between 60 nm and 300 nm. A ratio d1 / d2 between these dimensions of the emitter regions 21 of the first type and the emitter regions 22 of the second type is, for example, between 1.5 and 5, in particular between 2 and 4.Furthermore, an electrically active doping concentration of the first-type emitter regions 21 is higher than an electrically active doping concentration of the second-type emitter regions 22, or an electrically active doping dose of the first-type emitter regions 21 is higher than an electrically active doping dose of the second-type emitter regions 22. The "electrically active doping concentration" indicates the number of activated dopant atoms per volume (usually per cm³). 3 ) and the "dose of electrically active dopants" refers to the total number of activated dopant atoms that are deposited into a given surface area (usually 1 cm²). 2) of the first surface 101 of the emitter region of the first or second type in question. “Activated dopants” are electrically active dopants introduced into the crystal lattice of the semiconductor body 100 such that these dopants can contribute to electrical conductivity (that is, in the case of “n” type dopants, they can provide electrons, and in the case of “p” type dopants, they can accept electrons). Unless otherwise specified, “doper concentration,” as used herein, means electrically active doping concentration.

[0015] According to one embodiment, the maximum doping concentration in the emitter region 21 of the first type is between 1.5 x 16 cm³. -3 and 1.5E20 cm -3 , especially between 1E17 cm -3 and 1E19 cm -3 , or between 2E17 cm -3 and 2E18 cm -3The maximum doping concentration in the second-type emitter region 22 is between 1E16 cm³ -3 and 1E19 cm -3 , especially between 3E16 cm -3 and 1E18 cm -3 , or between 1E17 cm -3 and 1E18 cm -3 Let N21 / N22 be a ratio between the doping concentration in the first-type emitter region 21 and the doping concentration in the second-type emitter region 22. According to one embodiment, this ratio N21 / N22 is between 1.5 and 1E4, between 2 and 100, or between 2.5 and 10.

[0016] Referring to Fig. 1. The emitter regions 23 of the third type border the emitter regions 21 of the first type and the emitter regions 22 of the second type. In particular, each of the plurality of emitter regions 23 of the third type borders one of the emitter regions 22 of the second type in the vertical direction of the semiconductor body 100 and is arranged between two adjacent emitter regions 21 of the first type. The recombination region 24 is arranged in the emitter regions 21 of the first type and the emitter regions 23 of the third type. Optionally, sections of the recombination region 24 are also arranged in the emitter regions 22 of the second type. That is, the recombination region 24 can extend into the emitter regions of the second type. However, this is in Fig. 1 not shown. In the one in Fig. In the embodiment shown in Figure 1, the recombination area 24 is located in the emitter areas 21 of the first type, however, it is not located in those sections of the base area 10 (or the field stop area 41, see Figure 1). Fig. 3), which border the emitter regions 21 of the first type in the direction of current flow. The “direction of current flow” is in the Fig. In the embodiment shown in 1, the vertical direction of the semiconductor body.

[0017] Fig. Figure 2A schematically illustrates the doping concentration of one of the emitter regions 21 of the first type and the base region 10 along a line running in the vertical direction of the semiconductor body 100. Fig. Figure 2B schematically illustrates the doping concentration of one of the multiple emitter regions 22 of the second type and the base region 10 along a line running in the vertical direction of the semiconductor body 100. Referring to the Fig. 2A and Fig. 2B, the doping concentration of the first-type emitter regions 21 is higher than the doping concentration of the second-type emitter regions 22, and the doping concentration of the second-type emitter regions 22 is higher than the doping concentration of the base region 10. For example, the doping concentration is selected from a range between 5E12 cm -3 and 5E14 cm -3 .

[0018] Referring to Fig. Figure 3, which also shows a vertical cross-sectional view of the bipolar semiconductor device, shows that the bipolar semiconductor device can also have a field-stopping region 41. The field-stopping region 41 is located between the base region 10 and the first emitter region 20 and has the same doping type as the base region 10. The field-stopping region 41 has a higher doping concentration than the base region 10. According to one embodiment, the field-stopping region 41 is implemented such that a peak doping concentration is selected from a range of 5 x 14 cm⁻¹. -3 and 5E15 cm -3 . A doping dose of the field stop area 41 is, for example, between 0.5 x 12 cm³ -2 and 1E12 cm -2 .

[0019] A doping profile of the field-stop region 41 in the vertical direction of the semiconductor body 100 can be selected from several different doping profiles. According to one embodiment, the doping concentration of the field-stop region 41 in the vertical direction between the base region 10 and the emitter regions 21 of the first type is essentially constant. This is shown schematically in Fig. Figure 4A illustrates the doping concentration in one of the emitter regions 21 of the first type and the adjacent field stop region 41. According to a further embodiment (in Fig. (As illustrated in Figure 4A by dashed lines), the doping profile of the field stop region 41 has two or more doping maxima. The maximum doping concentration in each of these maxima is higher than the doping concentration of the base region 10. A minimum doping concentration between these two maxima can be higher than, substantially equal to, or lower than the doping concentration of the base region 10. According to one embodiment, these maxima have different peak concentrations, whereas the peak concentration decreases with increasing distance of the respective maximum from the first surface 101.

[0020] Referring to Fig. Figure 4B, which illustrates the doping concentration along a vertical line passing through one of the second-type emitter regions 22, the third-type emitter region 23, the field stop region 41, and the base region 10, shows that the doping concentration in the emitter regions 23 can decrease in the direction of the second-type emitter regions 22. That is, the third-type emitter regions 23 can have their maximum doping concentration in a region where they border the field stop region 41 and their minimum doping concentration in a region where they border the second-type emitter regions 22. According to another embodiment, the field stop region 41 extends into the third-type emitter region 23. For example, the field stop region 41 can have its maximum doping concentration in a region where the field stop region 41 borders the second-type emitter region 22.

[0021] In the horizontal plane of the semiconductor body 100, there are various possibilities for designing the emitter regions 21 of the first type and the emitter regions 22 of the second type. Referring to the following Fig. Figures 5-8 explain some examples. Each of these figures shows a horizontal cross-sectional view of a section of the semiconductor body 100. In particular, of the Fig. 5-8 each define a section of the semiconductor body 100 in a horizontal section plane AA that passes through the emitter regions 21 of the first type and the emitter regions 22 of the second type. The position of the horizontal section plane AA in the semiconductor body 100 is shown in the Fig. 1 and Fig. 3 shown.

[0022] Referring to Fig. 5. Both the emitter regions 21 of the first type and the emitter regions 22 of the second type can be elongated semiconductor regions. In Fig. Figure 5 shows several embodiments of how elongated emitter regions 21 of the first type can be implemented. According to one embodiment, the emitter regions 21 of the first type are connected in their longitudinal direction. According to another embodiment, an elongated region 21 contains two or more elongated subregions 211, 212, 213, which are spaced apart in their longitudinal direction. According to one embodiment, the elongated emitter regions of the first type have a substantially constant width along their longitudinal direction. According to another embodiment, the width varies along the longitudinal direction between a minimum width and a maximum width. According to one embodiment, the maximum width is twice the minimum width. More generally, the ratio between the maximum width and the minimum width is between 1.5 and 100, particularly between 1.8 and 10.

[0023] The emitter regions 21 of the first type in a component can be implemented with the same form, for example with one of the in Fig. 5 forms shown. However, it is also possible to implement the emitter regions of the first type with different shapes in a single component.

[0024] According to another, in Fig. In the embodiment shown in 6, the emitter regions 21 of the first type and the emitter regions 22 of the second type are implemented as concentric rings. In the embodiment shown in Fig. In the embodiment shown in Figure 6, the center is formed by one of the emitter regions 22 of the second type. However, this is merely an example. It is equally possible that an emitter region 21 of the first type is located in the center of the plurality of concentric rings. The embodiments shown in Figure 6 are for illustrative purposes only. Fig. The six concentric rings shown are arranged around rectangular rings. It should be noted that any other type of concentric ring, such as circular rings, elliptical rings, polygonal rings, or the like, can be used just as well.

[0025] At the in Fig. In the embodiment shown in Figure 7, the plurality of emitter regions 21 of the first type forms a lattice-shaped semiconductor region that surrounds the plurality of emitter regions 22 of the second type. In this embodiment, the emitter regions 22 of the second type have a rectangular shape. However, this is merely an example. The emitter regions 22 of the second type could just as easily be implemented with an elliptical shape, a circular shape, a polygonal shape, or the like. It is even possible to have different shapes of emitter regions 21 of the first type and emitter regions 22 of the second type in a single semiconductor device.

[0026] According to another (not shown) embodiment, the plurality of emitter regions 22 of the second type forms a lattice-shaped semiconductor region surrounding the plurality of emitter regions 21 of the first type.

[0027] According to yet another one, in Fig. In the embodiment shown in Figure 8, each of the plurality of emitter regions 21 of the first type and each of the plurality of emitter regions 22 of the second type has a rectangular shape, so that the overall arrangement of the emitter regions 21 of the first type and the emitter regions 22 of the second type resembles a chessboard.

[0028] According to one embodiment, the emitter regions 21 of the first type and the emitter regions 22 of the second type are implemented such that in the first surface 101 there is a ratio between the total area of ​​the emitter regions 21 of the first type and a total area of ​​the emitter regions 22 of the second type between 0.05 and 5, in particular between 0.1 and 1.

[0029] The topology described above, with the first emitter region 20, the base region 10 and the optional field-stop region 21, can be implemented in any type of bipolar semiconductor device such as a diode, an IGBT, a BJT (bipolar transistor) or a thyristor.

[0030] Fig. Figure 9 schematically illustrates a vertical cross-sectional view of a bipolar diode implemented with a first emitter region 20 of the type described above. Fig. 9 (as well as in the following explanation) Fig. 10) The first emitter region 20 is shown only schematically. For details of the first emitter region 20, reference is made to the explanations provided above. Fig. 1-8 taken. Besides the first emitter region 20 and the base region 10, the area in Fig. Figure 9 shows a second emitter region 51 of the first doping type. The second emitter region 51 borders the base region 10 and has a higher doping concentration than the base region 10. According to one embodiment, the doping concentration of the base region 10 (for silicon devices) is between 5 x 12 cm⁻¹. -3 and 5E14 cm -3 , while the doping concentration of the second emitter region 51 is between 1E19 cm -3 and 1E21 cm -3The second emitter region 51 is located in the area of ​​the second surface 102 of the semiconductor body 100. A second electrode 32 is arranged on the second surface 102 and makes ohmic contact with the second emitter region 51. In the diode, the optional field-stopping region 41 is arranged between the second emitter region 51 and the base region 10.

[0031] According to one embodiment, the first doping type (the doping type of the base region 10, the second emitter region 51, and the optional field-stopping region 41) is of type “n”, and the second type (the doping type of the emitter regions 21 of the first type and the emitter regions 22 of the second type) is of type “p”. In this embodiment, the first electrode 31 forms an anode A of the diode, and the second electrode 32 forms a cathode K of the diode.

[0032] Fig. Figure 10 illustrates a vertical cross-sectional view of an IGBT. Besides the first emitter region 20, the base region 10, and the optional field-stop region 41, the IGBT contains at least one transistor cell with a second emitter region 62 of the first doping type, a body region 61 of the second doping type, and a gate electrode 63. The gate electrode 63 is located adjacent to the body region 61 and is dielectrically isolated from the body region 61 by a gate dielectric 64. Fig. Figure 10 shows several such transistor cells. The entire IGBT can have several thousand, several tens of thousands, several hundred thousand, or more transistor cells. The individual transistor cells are connected in parallel by connecting their second emitter regions 62 (which can also be called source regions) to the second electrode 32. The second electrode 32 forms an emitter node (emitter terminal) E of the IGBT. The first electrode 31 forms a collector C of the IGBT. The gate electrodes 63 are connected to a common gate node G. This connection of the gate electrodes 61 to the gate node G is shown in Fig. Figure 10 is shown only schematically. Although the gate electrodes 63 are drawn, it should be noted that these gate electrodes 63 may represent sections of a continuous gate electrode, which may have a lattice shape in the horizontal plane of the semiconductor body 100. According to one embodiment, in the Fig. In the 10 IGBTs shown, the first doping type (the doping type of the base region 10, the optional field-stop region 41, and the second emitter regions 62) is of type “n”, and the second doping type (the doping type of the emitter regions 21 of the first type, the emitter regions 22 of the second type, and the body regions 61) is of type “p”. The field-stop region can be part of the third emitter region 23 of type “n”.

[0033] According to one embodiment, the IGBT is implemented as a reverse-conducting IGBT (RC-IGBT). In this case, the emitter region contains 20 semiconductor regions (emitter short circuits) 26 (in Fig. 10 shown by dotted lines) of the first doping type, which extend from the first surface 101 through the emitter region 20 to or into the base region 10 or field stop region.

[0034] Fig. Figure 10 shows only sections of the IGBT. Those areas of the IGBT in which transistor cells are located can be referred to as the active device region. According to one embodiment, the first emitter 20 with the emitter regions 21 of the first type, the emitter regions 22 of the second type, and the emitter regions 23 of the third type are located only below the active device region. This is illustrated by reference to Fig. 11 explained. Fig. Figure 11 shows a complete cross-sectional view of a semiconductor body 100 into which the bipolar semiconductor device is integrated. Referring to Fig. Figure 11 states that the semiconductor body 100 contains an inner region 110 (active region) and a boundary region 120. The boundary region 120 can be located (as shown) between an edge of the semiconductor body 100 and the inner region 110. Alternatively, the boundary region is located between the active region 110 of the bipolar semiconductor and the active region of another semiconductor device (not shown) integrated into the same semiconductor body. The "edge" of the semiconductor body 100 is a horizontal surface that bounds the semiconductor body 100 in horizontal directions.

[0035] As in Fig. As shown schematically in Figure 11, both type 1 emitter regions 21 and type 2 emitter regions are found only in the inner region. In the outer region 120, at least the type 1 emitter regions 21 are omitted. According to one embodiment, the first emitter 20 is omitted in the outer region 120. In this case, the doping concentration in the outer region 120 can correspond to the doping concentration in the base region 10. According to another embodiment, the first emitter 20 in the outer region contains a type 22 emitter region, a type 23 emitter region, and the recombination region 24. Both of these alternatives result in a lower carrier density in the outer region 120 compared to the inner region 110, so that in the event of the device being switched off, the risk of a dynamic avalanche breakthrough is reduced and the switch-off resistance is improved.According to a further embodiment, the boundary region 120 contains a first emitter 20 with emitter regions 21 of the first type and emitter regions 22 of the second type, whereas these regions 21, 22 are implemented such that in the boundary region the ratio between the total area of ​​the emitter regions 21 of the first type and the emitter regions 22 of the second type in the boundary region 120 is smaller than in the inner region 110. In a component such as is found in the . Fig. 1 and Fig. As shown in Figure 2, this can be achieved, for example, by increasing w2 in the boundary region 120 compared to the inner region 110 and / or by decreasing w1 in the boundary region 120 compared to the inner region. Optionally, these design measures, which reduce the current density in the boundary region 120, can also be implemented in a transition region between the inner region 110 and the boundary region 120, such that this transition region extends into the inner region 110 to three times the diffusion length of the minority charge carriers, or to 1.5 times the diffusion length of the minority charge carriers. For example, in an IGBT, the diffusion length of the minority charge carriers can be essentially equal to the length (thickness) of the drift region 10 in the current flow direction.

[0036] For the purpose of explanation, it is assumed that the first doping type is of type "n" and the second doping type is of type "p". The in Fig. The diode shown in Figure 9 can be operated in forward and reverse modes. In forward mode, a voltage is applied between the anode A and the cathode K, biasing a pn junction between the base region 10 and the emitter regions 21, 22 of the first and second dopants of the first emitter 20 (this voltage is positive if the first dopant type is of type "n"). In this mode, the first emitter 20 injects second-type charge carriers (holes) into the base region 10, and the second emitter 51 injects first-type charge carriers (electrons) into the base region 10. The first-type and second-type charge carriers injected into the base region 10 form a charge carrier plasma, which ensures low power losses of the diode.In reverse operation, a (negative) voltage is applied between the anode A and the cathode K, so that a pn junction between the emitter regions 21, 22 of the first and second types and the base region 10 (or the emitter regions 23 of the third type) is biased in the reverse direction. In this case, a depletion region (space charge region) spreads in the base region 10, thus preventing current flow through the diode.

[0037] In the context of the IGBT, only forward operation is explained in detail. The IGBT is in forward operation when a positive voltage is applied between the collector and the emitter E. In forward operation, the IGBT can be operated in an on-state and an off-state. In the on-state, the gate electrodes 63 are driven by receiving a suitable drive potential via the gate electrode G, such that they generate conductive channels in the body regions 61 between the second emitter regions 62 (source regions) and the base region 10. Via these conductive channels, the second emitter regions 62 inject first-type charge carriers (electrons) into the base region 10, while the first emitter region 20 injects second-type charge carriers (holes) into the base region 10.In the off state, the gate electrodes 63 are controlled such that conducting channels in the body regions 61 between the second emitter regions 62 and the base region 10 are interrupted. In this case, due to the positive voltage applied between the collector C and the emitter E, a depletion region (space charge region) spreads in the base region 10, starting at a pn junction between the base region 10 and the body regions 61. [The text abruptly ends here, so the translation stops here.] Fig. 10. In a manner not illustrated in detail, the body regions 61 and the second emitter regions 62 are connected to the second electrode 32 or the emitter E.

[0038] The functionality of the first emitter 20 in the switched-on state of the IGBT (which corresponds to the forward-biased state of the diode) is explained in more detail below. In this context, reference is made to the Fig. 1 and Fig. Reference is made to Figure 3, in which the first emitter 20 and the adjacent device regions are described in detail. It should be noted that the following explanation is based on a simplified model and focuses on the most relevant aspects. In the switched-on state, the first-type emitter regions 21 inject second-type charge carriers directly into the base region 10 (see Figure 3). Fig. 1), or through the optional field stop area 41 into the base area 10 (see Fig. 3) The second-type emitter regions 22 inject second-type charge carriers through the third-type emitter regions 23 into the base region 10 or through the third-type emitter regions 23 and the field stop region 41 into the base region 10.

[0039] When the IGBT is switched on, a current level can be defined by a load (not shown) connected in series with the collector-emitter path CE of the IGBT, with the series connection of the load and the IGBT being connected to a supply voltage source. Depending on the operating mode of the load, the current level can vary between relatively low levels, such as a few hundred milliamperes (mA), and high levels, such as a few tens of amperes (A). For example, high current levels (i.e., current levels higher than several times the nominal current) can occur if a short circuit occurs in the load. Hereinafter, a current level that the IGBT can withstand without being destroyed is referred to as the maximum current level.Furthermore, an operating mode in which the current level is below a predetermined current threshold is referred to as normal operation, and an operating mode in which the current level is between the predetermined current level and the maximum level is referred to as high-current operation.

[0040] As explained below, the first emitter 20 helps to achieve low switching losses during normal operation and to achieve a high maximum current level during high-current operation. During normal operation, the second-type emitter regions 22 primarily inject charge carriers into the base region 10, while the first-type emitter regions 21 inject fewer charge carriers than the second-type emitter regions. The reason for this is explained below.

[0041] Between the first-type emitter regions 21 and the base region 10 or the field stop region 41, there are first pn transitions; and between the second-type emitter regions 22 and the third-type emitter regions 23, there are second pn transitions. Each of these pn transitions has a built-in voltage V. bi .

[0042] This built-in voltage of the pn junction is given by Vbi=kTqlnNANDni2 (see SM Sze: “Semiconductor Devices, Physics and Technology”, page 73, Jon Wiley & Sons, 1985, ISBN 0-471-87424-8), where k is the Boltzmann constant, T the absolute temperature, q the elementary charge, In the natural logarithm, N A the doping concentration of the (acceptor) layer of type “p”, N adjacent to the pn junction D the doping concentration of the (donor) layer of type “n” adjacent to the pn junction, and n iThe intrinsic doping of the semiconductor material used to implement the pn junction. For example, n is i in silicon at room temperature approximately 1.45 x 10 cm -3 Charge carriers can flow across the pn junction if a positive voltage is applied between the p-type region and the n-type region, and if the voltage level of this voltage is higher than the built-in voltage.

[0043] Based on equation (1), a built-in voltage V bi1 the first pn transitions by doping concentrations N A1 and N D1 those regions that form the first pn transitions. For the purpose of explanation, it is assumed that the doping type of emitter regions 21 is of the first type and that of emitter regions 22 is of the second type, of type “p”. In this case, N denotes A1 the doping concentration of the emitter regions 21 of the first type, and N D1denotes the doping concentration of the base region 10 (if no field stop region is present) or of those sections of the field stop region 41 that border the emitter regions 21 of the first type (if a field stop region is present). Accordingly, a built-in voltage V bi2 the second pn transitions are determined by a doping concentration N A2 the emitter regions 22 of the second type and a doping concentration N D2 of the emitter regions 23 of the third type. Referring to the explanation above, the emitter regions 21 of the first type have a higher doping concentration than the emitter regions 22 of the second type, that is to say, NA1>NA2

[0044] Furthermore, the doping concentration of those areas of the third-type emitter regions 23 adjacent to the second-type emitter regions 22 can be equal to the doping concentration of the base region 10 (if no field stop region is present), it can be equal to the doping concentration of those sections of the field stop region 41 adjacent to the first-type emitter regions 21, it can be lower or even higher than the doping concentration of those sections of the field stop region 41 adjacent to the first-type emitter regions 21. That is to say, in each case, ND1≥ND2 or ND1 <ND2

[0045] Referring to equations (1)-(2) and (3a), the built-in voltage V can be bi1 The voltage at the first pn junctions must be higher than the built-in voltage V. bi2 the second pn transitions, that is, Vbi1>Vbi2 If, referring to equation (3b), ND1 < N D2 is, can N A1 be chosen so that it is relative to N A2 is high enough that equation (4) is satisfied.

[0046] The current driven by the load through the IGBT is associated with a voltage across the first pn junctions and the second pn junctions. A current begins to flow through the IGBT when the voltage across the second pn junctions exceeds the second built-in voltage VL. bi2 This is reached. At this time, the second-type emitter regions begin to inject second-type charge carriers, while essentially no second-type charge carriers are injected by the first-type emitter regions. As a current driven by the IGBT increases, the voltage across the first pn junctions rises, and the first-type emitter regions begin to inject charge carriers when the voltage across the first pn junctions reaches the first built-in voltage Vpn. bi1This is achieved. In this way, the injection of charge carriers by the first-type emitter regions 21 and the second-type emitter regions depends on the current level (or current density level) of the current through the IGBT. When the current level is below a current threshold, the second-type emitter regions primarily inject charge carriers, which is the case when the IGBT is in its normal operating state. The first-type emitter regions 21 and the second-type emitter regions 22 inject charge carriers when the current level is above the current threshold, which is the case when the IGBT is operating at high current. For example, the IGBT is operating at high current when it is switched on and there is a short circuit in a load connected to the load path (collector-emitter path) of the IGBT.

[0047] In high-current operation, it is desirable for both the first-type emitter regions 21 and the second-type emitter regions 22 to prevent the so-called Egawa or Kirk effect, which in turn increases the current handling capability. Allowing both the first-type emitter regions 21 and the second-type emitter regions 22 to inject charge carriers leads to high emitter efficiency of the first emitter 20 in high-current operation. Generally, switching losses (turn-off losses) in an IGBT increase when emitter efficiency increases. However, in high-current operation, the focus is on increasing the current handling capability by preventing extreme peaks in electric field strength near the first emitter 20.

[0048] In normal operation, the emitter efficiency of the first emitter 20 is lower than in high-current operation, since only the second type of emitter regions 22 inject charge carriers. This results in low switching losses (turn-off losses) in normal operation. Furthermore, the emitter efficiency increases in normal operation as the current level increases. This is explained below.

[0049] Under normal operating conditions, the emitter efficiency is determined by the doping concentration of the second-type emitter regions and by the presence of the recombination region 24. For a given doping concentration of the second-type emitter regions 22, the emitter efficiency of the IGBT with recombination region 24 is lower than the emitter efficiency of a comparable emitter region without recombination region 24. In the recombination region 24, some of the charge carriers injected from the second-type emitter region 22 into the third-type emitter regions 23 recombine, so that fewer charge carriers than those injected from the second-type emitter regions 22 flow through the third-type emitter regions 23 into the base region 10. However, the recombination rate in the recombination region 24 depends on the current density of the current flowing through the recombination region, whereas the recombination rate decreases as the current density increases.Thus, the charge carrier lifetime in the recombination region 24 increases as the current density increases. In normal operation, this increase in charge carrier lifetime at higher current densities corresponds to an increasing emitter efficiency at an increasing current density.

[0050] The recombination region 24 in the third-type regions 23 of the first doping type is more efficient at recombinating second-type charge carriers than the recombination region 24 in the first-type emitter regions 21 of the second doping type. Therefore, the recombination region 24 also reduces the emitter efficiency of the portion of the first emitter region 20 formed by the first-type emitter regions 21, but it reduces the efficiency of these first-type emitter regions 21 less than the efficiency of the portion of the first emitter 20 formed by the second-type emitter regions 22.

[0051] Measurements have shown that the first emitter 20 with the emitter regions 21, 22, 23 of the first type, the second type and the third type helps to increase the current handling capability of the IGBT without deteriorating other performance parameters of the IGBT such as the saturation voltage and the switching losses.

[0052] Fig. Figure 12 shows the saturation voltage Vce,sat versus the turn-off losses Eoff of several IGBTs implemented with different first emitter regions.

[0053] In Fig. The symbols “triangle”, “X”, “star”, “circle” and “plus” represent the saturation voltage and the corresponding turn-off switching losses of IGBTs, as shown in the Fig. 1 and Fig. Figure 3 shows IGBTs implemented with a first emitter region 20. The "diamond" symbols represent the saturation voltage Vce,sat and the corresponding switching losses Eoff of IGBTs implemented with a conventional (homogeneous) first emitter. The IGBTs represented by the "diamonds" were fabricated with different dopant doses of their first emitters, whereas in these conventional IGBTs, the saturation voltage increases and the switching losses decrease as the doping dose of the first emitter decreases. Fig. 12 denotes REF a conventional IGBT that exhibits a similar saturation voltage Vce,sat and similar switching losses as the first-emitter IGBTs 20 according to the Fig. 1 or Fig. 3. This IGBT is subsequently referred to as the reference device REF. The IGBTs represented by the "triangle", the "X", the "star", the "circle", and the "plus" have slightly different saturation voltages. This is due to the fact that these IGBTs were manufactured with different widths w2 of their emitter regions of the second type, while the width w1 of the emitter region 21 of the first type was essentially the same for each of these IGBTs.

[0054] Fig. Figure 13 shows the saturation voltage Vce,sat versus the destruction current Ice,dest with reference to Fig. 12 IGBTs were explained. The "diamond" symbol represents IGBTs with a conventional first emitter region, and the "triangle", the "X", the "star", the "circle", and the "plus" represent IGBTs with a first emitter region as described in one of the Fig. 1 and Fig. 3 is shown. From Fig. Figure 13 shows that the destruction current increases in a conventional first emitter region, whereby, referring to the explanation above, a lower saturation voltage Vce,sat results from a higher doping dose of the first emitter region. The IGBTs with emitter regions 21 of the first type and emitter regions 22 of the second type, even though they are comparable to the reference device REF, which has a conventional first emitter region, with respect to saturation voltage and switching losses, are superior with respect to destruction current. In this specific embodiment, the destruction currents of the IGBT with the non-conventional first emitter region 20 are between approximately two and three times the destruction current of the reference device REF.The specific gain in the destruction current Ice,dest compared to the reference device REF depends on the specific design of the first emitter, in particular on the dimensions of the emitter regions 21 of the first type and the emitter regions 22 of the second type in the horizontal and vertical direction of the semiconductor body 100.

[0055] The Fig. 14A and Fig. 14B illustrates an example of a method for producing a first emitter region 20. Referring to Fig. 14A describes the method of generating the first-type emitter regions 21 by implanting second-type dopant atoms through the first surface 101 into the semiconductor body 100 and by activating the implanted dopant atoms. The implantation of the dopant atoms includes the use of an implantation mask 200 such that dopant atoms are implanted only into those regions of the first surface 101 that are not covered by the implantation mask 200. The implantation dose is, for example, between 1E14 cm⁻¹. -2 and 3E15 cm -2 , especially between 6E14 cm -2 and 9E14 cm -2The implantation energy is, for example, between 10 keV and 200 keV, particularly between 15 keV and 70 keV. Optionally, the implantation of the dopant atoms includes a further implantation process with higher implantation energies. For example, the implantation dose in this additional implantation process is between 1 × 10⁻¹² cm⁻¹. -2 and 1E13 cm -2 , especially between 3E12 cm -2 and 7E12 cm -2The implantation energy is between 150 keV and 190 keV. Activating the implanted dopant atoms can include a healing process in which at least those regions of the semiconductor body 100 into which the dopant atoms were implanted are healed. For example, the healing process is chosen such that essentially 100% of the implanted dopant atoms are activated. Such a healing process can include a laser healing process that melts regions of the semiconductor body 100 near the surface 101. That is, in this laser healing process, a temperature is selected such that the region near the first surface 101 melts. The implantation mask 200 can be removed before the laser healing process.

[0056] The laser annealing process is chosen such that the depth of the molten region essentially corresponds to the desired depth d1 of the first-type emitter regions 21. After the laser annealing process, that is, when the semiconductor body 100 cools, the molten semiconductor region recrystallizes and the implanted dopant atoms are incorporated into the crystal lattice of the recrystallized semiconductor region. In the molten semiconductor region, the doped dopant atoms diffuse vertically (redistribute themselves), so that the doping concentration in the first-type emitter regions 21 is essentially homogeneous after recrystallization. The implanted dopant atoms also diffuse laterally. However, the width w1 of the first-type emitter region 21 is significantly greater than the depth d1, so that diffusion in the lateral direction is negligible.This means that the healing process does not lead to a significant expansion of the emitter regions 21 of the first type. For example, the w1 / d1 ratio is in the range between 2 and 100, particularly between 5 and 50. d1 can be in the range between 0.3 micrometers (µm) and 1 µm, and w1 can be in the range between 0.5 µm and 50 µm, particularly between 1 µm and 35 µm, or between 5 µm and 15 µm.

[0057] For example, the laser healing process is such that the energy is selected from a range between 1 J / cm². 2 and 10 J / cm 2 , especially between 1.5 J / cm² 2 and 4.5 J / cm 2 .

[0058] In another example, the healing process is designed so that less than 100% of the implanted ions are activated. This can be achieved by adjusting the healing depth in relation to the implantation depth.

[0059] Referring to Fig. In section 14B, the procedure also includes the formation of second-type emitter regions 22 by implanting second dopants and partially activating the implanted dopants. The implantation of the second-type dopants can involve unmasked implantation, so that dopants are implanted into those regions of the first surface 101 that were covered by the implantation mask 200 in the first implantation process described above, but also into the first-type emitter regions 21. In this second implantation process, the implantation dose is, for example, between 1E12 cm⁻¹. -2 and 1E15 cm -2 , especially between 2E12 cm -2 and 1E14 cm -2 , or between 3E12 cm -2 and 1E13 cm -2Suitable dopant atoms for both the first and second implantation processes are boron atoms, aluminum atoms, indium atoms, and gallium atoms, if the second doping type is of type “p”.

[0060] The activation of the second-type dopant atoms implanted during the second implantation process may involve only partial activation of the implanted dopant atoms. This means that only a portion of the implanted dopant atoms are activated. This partial activation of the implanted dopant atoms may involve a curing process at temperatures between 300°C and 500°C, particularly between 350°C and 420°C, and a duration of between 0.5 and 5 hours, particularly between 1 and 4 hours.

[0061] The implantation of dopant atoms during the second implantation process creates crystal defects in the semiconductor body 100. During the annealing process at the comparatively low temperatures described above, these crystal defects are not annealed but diffuse deeper into the semiconductor body 100 and form the recombination region 24. In this recombination region 24, the crystal defects form recombination centers. For example, the temperature and duration of the second annealing process are chosen such that the recombination region 24 forms within the emitter regions 21 of the first type, but within and outside the emitter regions 22 of the second type, whereas the recombination efficiency in the recombination region 24 outside the emitter regions 21, 22 of the first and second types is higher than in the recombination region 24 within these regions 21, 22.

[0062] According to one example, the recombination region 24 is generated such that the concentration of recombination centers in the recombination region 24 is such that the carrier lifetime in those parts of the recombination region 24 located outside the emitter regions 21 and 22 of the first and second types is between 100 nanoseconds (ns) and 50 microseconds (µs), 500 nanoseconds and 30 microseconds, or 1 microsecond and 20 microseconds. According to another example, the ratio between the carrier lifetime in the recombination region 24 in the emitter region of the first type and the carrier lifetime in the recombination region 24 in the emitter region 23 of the third type is between 2 and 4.

[0063] The in Fig. The first emitter region 20 shown in 14B corresponds to the one in Fig. 1 shown first emitter area. One in Fig. The first emitter region 20 shown in Figure 3 is obtained by establishing the field stop region 41 before the production of the first-type emitter regions 21 and the second-type emitter regions 22, respectively. Referring to Fig. 15. The generation of the field-stop region 41 can involve the implantation of initial dopant atoms through the first surface 101 into the semiconductor body 100. Alternatively, these dopant atoms can be implanted through the second surface. The implantation of the initial dopant atoms can involve multiple implantation processes at different implantation energies to obtain a field-stop region 41 with two or more spaced-apart doping maxima. For example, the implanted atoms contain at least one of the following: selenium (Se), phosphorus (P), arsenic (As), and antimony (Sb). Alternatively, hydrogen (H) atoms can be implanted, which can form donor-like complexes with radiation-induced damage (e.g., vacancies). The formation of the field-stop region 41 can also include a healing step to at least partially activate the implanted dopant atoms. The generation of a field-stop region 41 with reference to Fig. The type described in section 3 is known, so no further explanation is required in this respect.

[0064] Referring to Fig. 16 The method may further comprise the generation of at least one emitter region 25 of a fourth type of the second doping type by implanting dopant atoms through the first surface 101 using an implantation mask 210 and the activation of the implanted dopant atoms. The at least one emitter region 25 of the fourth type may be generated prior to the generation of the emitter regions 21, 22 of the first and second types, or it may (as in Fig. (16 shown) after the generation of the first-type emitter region and the second-type emitter regions 21, 22. The at least one third-type emitter region is generated such that it has a doping concentration that is higher than a doping concentration of the second-type emitter regions 22, and lower than a doping concentration of the first-type emitter regions 21.

[0065] Referring to Fig. 16. The implantation mask can be chosen such that parts of at least one emitter region 25 of the fourth type are generated in the emitter region 21 of the first type and the emitter region of the second type. However, since the doping concentration is lower than the doping concentration of the emitter regions 21 of the first type, the emitter region 25 of the fourth type is only effective in those areas where it is generated in the emitter regions 22 of the second type. In the vertical direction, the emitter region of the third type can extend into the base region or the field-stop region. The at least one emitter region of the fourth type, in addition to the emitter regions 21 of the first type, helps to counteract the Egawa effect when the device is switched off.

[0066] The shape of the at least one emitter region of the fourth type is independent of the shape of the emitter regions of the first and second types. Several examples of how the at least one emitter region 25 of the fourth type can be implemented are given below with reference to the Fig. 17 and Fig. 18 explained. In these figures, which show a top view of the first surface, the first and second emitter regions 21, 22 (represented by dotted lines) are implemented as stripes (that is, with an elongated shape). However, this is only an example; any other shape explained above can be used as well.

[0067] At the in Fig. In the embodiment shown in Figure 17, at least one emitter region 25 of the fourth type is ring-shaped. In this embodiment, only one emitter region 25 of the fourth type is shown. However, the component can be implemented with two or more emitter regions of the fourth type, which can be implemented as concentric rings.

[0068] Referring to Fig. Figure 18, which shows different embodiments of how the at least one emitter region 25 of the fourth type can be implemented, the emitter region 25 of the fourth type can be elongated or columnar (for example with a circular, elliptical or rectangular cross-section).

[0069] The width of the fourth-type semiconductor region 25 is at least twice, five times, or even more than ten times the width w1 of the first-type emitter region 21. The "width" of the fourth-type emitter region 25 is the smallest lateral dimension of the fourth-type emitter region. For example, the width in the case of the Fig. In the ring-shaped region 25 shown in Figure 17, the width of the elongated area forming the ring is given. For example, an implantation dose for producing the emitter region 25 of the fourth type is selected from a region between 1 x 10 cm. -2 and 1E14 cm -2 , from an area between 3E12 cm -2 and 5E13 cm -2 , or from an area between 5 and 12 cm -2 and 3E13 cm -2 The healing process may include a laser healing procedure in which at least those areas into which dopant atoms have been introduced are melted.

Claims

A bipolar semiconductor device comprising: a semiconductor body (100) having a first surface (101), a base region (10) of a first doping type, and a first emitter region (20) in the semiconductor body (100), wherein the first emitter region (20) adjoins the first surface (101) and comprises: a plurality of emitter regions (21) of a first type of a second doping type complementary to the first doping type, a plurality of emitter regions (22) of a second type of the second doping type, a plurality of emitter regions (23) of a third type of the first doping type, and a recombination region (24) having recombination centers, wherein the emitter regions (21) of the first type and the emitter regions of the second type (22) extend from the first surface (101) into the semiconductor body (100).wherein the emitter regions (21) of the first type have a higher doping concentration and extend from the first surface (101) deeper into the semiconductor body (100) than the emitter regions (22) of the second type, wherein the emitter regions (23) of the third type are adjacent to the emitter regions (21) of the first type and the emitter regions (22) of the second type, wherein the recombination region (24) is located at least in the emitter regions (21) of the first type and the emitter regions (23) of the third type, and wherein the ratio between the charge carrier lifetime in the recombination region (24) in the emitter regions (21) of the first type and the charge carrier lifetime in the recombination region (24) in the emitter regions (23) of the third type is between 2 and 4. Bipolar semiconductor device according to claim 1, further comprising: a field stop region (41) of the first doping type between the base region (10) and the first emitter region (21), wherein a doping concentration of the field stop region (41) is higher than the doping concentration of the base region (10). Bipolar semiconductor device according to claim 2, wherein the ratio between a minimum doping concentration of the field stop region (41) and a maximum doping concentration of the base region (10) is between 2 and 4. Bipolar semiconductor device according to claim 1, wherein a charge carrier lifetime in the recombination region (24) outside the emitter regions (21, 22) of the first and second type is one of the following: between 100 nanoseconds and 500 microseconds, between 500 nanoseconds and 30 microseconds, and between 1 microsecond and 20 microseconds. Bipolar semiconductor device according to claim 1, wherein the emitter regions (23) of the third type are adjacent to the emitter regions (22) of the second type in a vertical direction of the semiconductor body (100) and to the emitter regions (21) of the first type in a horizontal direction of the semiconductor body (100). Bipolar semiconductor device according to claim 1, wherein each of the emitter regions (21) of the first type and the emitter regions (22) of the second type in a horizontal plane of the semiconductor body (100) has a shape selected from the group consisting of: elongated; annular; and rectangular. Semiconductor device according to claim 1, wherein the plurality of emitter regions (21) of the first type and the emitter regions (22) of the second type have the form of a lattice in a horizontal plane of the semiconductor body (100). Bipolar semiconductor device according to claim 1, further comprising: a second emitter region (51; 62) of the second doping type, spaced apart from the first emitter region (20), wherein the base region (10) is arranged between the first emitter region (20) and the second emitter region (51; 62). Bipolar semiconductor device according to claim 8, further comprising: a body region (61) of the first doping type between the second emitter region (62) and the base region (10); and a gate electrode (63) adjacent to the body region (61) and dielectrically insulated from the body region (61) by a gate dielectric (64). Bipolar semiconductor device according to claim 9, further comprising: emitter short circuits (26) of the first doping type extending from the first surface (101) through the first emitter region (20). Bipolar semiconductor device according to claim 1, wherein a depth of the emitter regions (21) of the first type in a vertical direction of the semiconductor body (100) is selected from a range between 0.3 micrometers and 1 micrometer; and wherein a width of the emitter regions (21) of the first type in a lateral direction of the semiconductor body (100) is selected from a range between 0.5 micrometers and 50 micrometers.

Images