Semiconductor equipment

Abstract

To obtain an RFC diode that is less prone to secondary breakdown. [Solution] On the back side, an n-type cathode layer 12 and a p-type cathode layer 13 are formed alternately in a plan view, and the n-type cathode layer 12 and the p-type cathode layer 13 are connected to the cathode electrode 21. The surface semiconductor region comprises a first p-type anode layer 15 formed to a substantially uniform depth from the surface, and a plurality of second p-type anode layers 16 formed dispersed within the first p-type anode layer 15 in a plan view, and formed deeper than the first p-type anode layer 15. In a plan view, the n-type cathode layer 12 is characterized by being formed directly below each second p-type anode layer 16, wider than each second p-type anode layer 16 so as to include each second p-type anode layer 16, and corresponding to each second p-type anode layer 16.

Description

[Technical Field]

[0001] This disclosure relates to a semiconductor device that functions as a diode. [Background technology]

[0002] In diodes utilizing a pn junction, the RFC (Relaxed Field of Cathode) diode is known as a structure that suppresses ringing during the reverse recovery period (the transition period when switching from on to off) and quickly interrupts the current.

[0003] In a typical diode, a p-type anode layer is formed on the front side of an n-type drift layer, and an n-type cathode layer is formed on the back side. When the diode is ON, current flows between the anode and cathode layers. In contrast, in an RFC-type diode, the n-layer that forms the cathode layer on the back side is divided, and a p-layer is inserted in between. During reverse recovery, holes are injected into the drift layer from this p-layer, suppressing ringing and improving the reverse recovery characteristics.

[0004] In the diode described above, the forward voltage VF is small, and the reverse recovery energy E in the reverse recovery characteristics described above is small. REC It is desirable that the (or reverse recovery charge amount Qrr) be small. However, generally, VF and E REC Because these factors are in a trade-off relationship, it is difficult to reduce both of them simultaneously. Taking this into consideration, a structure that further improves upon such an RFC-type diode is described in Patent Document 1.

[0005] Figure 14 is a simplified cross-sectional view showing the structure of the diode (semiconductor device 9) described in Patent Document 1. Here, the back side (lower side in the figure) of the n-type semiconductor layer (n-layer 91), which forms the drift region, is the cathode, and the front side (upper side in the figure) is the anode. On the cathode side, a high-density n-type layer (n-type cathode layer 92) is provided, as in the conventional design, and on the anode side, a p-type anode layer 93 is provided. In addition, a cathode electrode 94 is formed on the back side, and an anode electrode 95 is formed on the front side.

[0006] Here, as described above, on the cathode side, the n-type cathode layer 92 is formed by division, and a p-type cathode layer 96, which is inversely conductive, is provided between adjacent n-type cathode layers 92. As described above, the presence of the p-type cathode layer 96 suppresses ringing and improves the reverse recovery characteristics.

[0007] Furthermore, similar to the cathode side, the p-type anode layer 93 is composed of a deeply formed second p-type anode layer 93B and a shallower first p-type anode layer 93A, which are alternately formed in the in-plane direction. Here, for convenience in this specification, the terms "first" and "second" are reversed compared to the description in Patent Document 1. The second p-type anode layer 93B has a higher impurity concentration than the first p-type anode layer 93A and is formed to protrude locally toward the back surface.

[0008] In this configuration, VF and E are determined by setting the ratio of the area (length in the left-right direction in the figure) of the n-type cathode layer 92 and the p-type cathode layer 96 on the cathode side, the depth, and the impurity concentration ratio, and by setting the ratio of the area of ​​the first p-type anode layer 93A and the second p-type anode layer 93B on the anode side. REC The relationship between VF and E can be adjusted. REC This allows for optimization. [Prior art documents] [Patent Documents]

[0009] [Patent Document 1] Patent No. 6750668 [Overview of the project] [Problems that the invention aims to solve]

[0010] As described above, when a structure is provided that protrudes locally from the back side beyond the p-type anode layer 93A, breakdown (first-order breakdown) is more likely to occur between the edge of the deep portion of the p-type anode layer (the edge of the second p-type anode layer 93B) and the n-type cathode layer 92 directly beneath it, as shown by region X in Figure 14. In particular, when the second p-type anode layer 93B is made wide, electrons can more easily reach the vicinity of the p-type cathode layer 96 on the back side, which makes it easier for holes to be injected from the p-type cathode layer 96 into the drift region (n-layer 91). This large amount of holes causes conductivity modulation, which locally lowers the resistance of the n-layer 91, causing an even larger current to flow locally through the n-layer 91, that is, making second-order breakdown more likely, and thus making the device more susceptible to destruction. For this reason, an RFC diode that is less prone to such second-order breakdown was desired.

[0011] This disclosure has been made in view of the aforementioned problems and aims to provide a semiconductor device that solves the above-mentioned problems. [Means for solving the problem]

[0012] This disclosure has the following structure in order to solve the above-mentioned problems. The semiconductor device of this disclosure is a semiconductor device in which current flows between a first main electrode provided on the surface side of a semiconductor substrate and a second main electrode provided on the back side, wherein the semiconductor substrate comprises, on the surface side, a drift region of a first conductivity type and a surface-side semiconductor region of a second conductivity type opposite to the first conductivity type formed on the surface side of the drift region and connected to the first main electrode, and on the back side, a back-side first semiconductor region of the first conductivity type having a higher impurity concentration than the drift region and a back-side second semiconductor region of the second conductivity type, in a plan view The semiconductor regions are formed alternately, and the first semiconductor region on the back side and the second semiconductor region on the back side are connected to the second main electrode. The semiconductor region on the front side comprises a first semiconductor region and a second semiconductor region formed in contact with the first semiconductor region and formed deeper than the first semiconductor region. In a plan view, the first semiconductor region on the back side is formed directly below each second semiconductor region, wider than the second semiconductor region so as to include the second semiconductor region, and corresponding to each second semiconductor region. The plurality of the second surface-side semiconductor regions are formed in an arrangement perpendicular to the longitudinal direction in a plan view, and the width of the outermost second surface-side semiconductor region along the said direction may be wider than the width of the other second surface-side semiconductor regions. In a plan view, each of the second surface-side semiconductor regions may be formed in a dot pattern. In a plan view, each of the second surface-side semiconductor regions may include a portion formed in a stripe shape with a common longitudinal direction. The impurity concentration in the second surface-side semiconductor region is set higher than that in the first surface-side semiconductor region, and a resurf layer of the second conductivity type with a lower impurity concentration than that in the first surface-side semiconductor region may be locally provided on the surface of the drift region outside the outermost second surface-side semiconductor region. The resurf layer may be formed to be shallower than the second surface-side semiconductor region and deeper than the first surface-side semiconductor region. In a plan view, the second semiconductor region on the back side does not necessarily have to be formed outside the outermost second semiconductor region on the front side. In a plan view, the minimum width of the second surface side semiconductor region on the front surface side is in the range of 5 μm to 60 μm, and the ratio of the area of the back surface side first semiconductor region to the total area of the back surface side first semiconductor region and the back surface side second semiconductor region on the back surface side may be in the range of 30% to 70%. The combination of the second surface side semiconductor region and the corresponding back surface side first semiconductor region may be provided periodically in a plan view.

Advantages of the Invention

[0013] Since the present disclosure is configured as described above, an RFC diode in which secondary breakdown hardly occurs can be obtained.

Brief Description of the Drawings

[0014] [Figure 1] It is a diagram schematically showing the situation of primary breakdown and secondary breakdown of a diode. [Figure 2] It is a cross-sectional view (part 1) showing the configuration of a semiconductor device according to an embodiment of the present disclosure. [Figure 3] It is a result of calculating the current density distribution flowing at the time of primary breakdown of a semiconductor device according to an embodiment of the present disclosure. [Figure 4] It is a result of calculating the dependence of the primary breakdown voltage BV1 (a) and the secondary breakdown voltage BV2 (b) on the protrusion amount D of the second surface side semiconductor region in a semiconductor device according to an embodiment of the present disclosure. [Figure 5] It is a result of calculating the dependence of the primary breakdown voltage BV1 (a) and the secondary breakdown voltage BV2 (b) on the width W of the second surface side semiconductor region in a semiconductor device according to an embodiment of the present disclosure. [Figure 6] It is a result of calculating the dependence of the primary breakdown voltage BV1 (a) and the secondary breakdown voltage BV2 (b) on the pitch P of the second surface side semiconductor region in a semiconductor device according to an embodiment of the present disclosure. [Figure 7] It is a result of calculating the dependence of the primary breakdown voltage BV1 (a) and the secondary breakdown voltage BV2 (b) on the width WN of the back surface side first semiconductor region in a semiconductor device according to an embodiment of the present disclosure. [Figure 8] This is the result of investigating the relationship between the forward voltage VF and the reverse recovery charge Qrr by changing the back surface n-type ratio in an embodiment of the semiconductor device according to the present disclosure. [Figure 9] This figure shows an example of a planar structure of a semiconductor device according to an embodiment of the present disclosure. [Figure 10] This figure shows another example of a planar structure of a semiconductor device according to an embodiment of the present disclosure. [Figure 11] This is a cross-sectional view (part 2) showing the configuration of a semiconductor device according to an embodiment of the present disclosure. [Figure 12] This is a cross-sectional view (part 3) showing the configuration of a semiconductor device according to an embodiment of the present disclosure. [Figure 13] This figure shows a modified example of another example of the planar structure of a semiconductor device according to an embodiment of the present disclosure. [Figure 14] This is a cross-sectional view showing the configuration of an example of a conventional RFC diode. [Modes for carrying out the invention]

[0015] The following describes a semiconductor device that is an embodiment of this disclosure. In the following drawings, identical or similar parts are denoted by the same or similar reference numerals. However, it should be noted that the drawings are schematic, and the relationship between thickness and planar dimensions, the ratio of the lengths of each part, etc., may differ from reality. Therefore, specific dimensions should be determined by referring to the following explanation. It should also be noted that there are parts where the relationships and ratios of dimensions differ between drawings. Furthermore, the embodiments shown below are illustrative examples of devices for realizing the technical idea of ​​this disclosure, and the technical idea of ​​this disclosure does not limit the shape, structure, arrangement, etc. of the components to those described below. Various modifications can be made to the embodiments of this invention within the scope of the claims. In this disclosure, terms such as "top" and "bottom" are used for convenience of description, and even if they are provided on the side surface, if they are substantially identical to the constituent elements of this disclosure, they fall within the scope of the rights of this disclosure. Also, "top" includes not only cases where it is formed in contact with the object, but also cases where it is formed through another layer. Furthermore, in this disclosure, "connection" is not limited to direct connection; even if a connection is made by interposing something such as a resistor, if it is substantially the same as the constituent elements of this disclosure, it falls within the scope of the rights of this disclosure.

[0016] Figure 1 schematically shows the breakdown characteristics (current-voltage characteristics under reverse bias) of the semiconductor device (diode) discussed here. In Figure 1, the current under reverse bias is shown on a logarithmic scale. Here, primary breakdown is the phenomenon in which the current increases sharply from a state where the reverse current is small to the primary breakdown voltage BV1. Furthermore, when the voltage becomes higher than BV1, negative resistance occurs at the secondary breakdown voltage BV2 due to conductivity modulation caused by a large number of holes injected in the drift region, which causes a particularly large current to flow, and this causes the device to break down. BV1 is, for example, 500V or more, and BV2 is higher than this, and secondary breakdown occurs due to the current that flows due to primary breakdown. Since a particularly large current flows in secondary breakdown, secondary breakdown is a cause of breakdown of the semiconductor device itself and the circuits connected to it.

[0017] In this semiconductor device, the current density of the current flowing during primary breakdown is set to be small, and the subsequent secondary breakdown is made less likely to occur (BV2 is increased). As a result, this semiconductor device is less likely to experience breakdown during secondary breakdown.

[0018] Generally, in semiconductor substrates that constitute such semiconductor devices (RFC diodes or power semiconductor elements), an active region that functions as an active element (diode) and a termination region formed on the outer periphery of the active region to ensure breakdown voltage are provided. Although Figure 14 shows the structure of only the active region, the semiconductor device according to the embodiment of this disclosure will be described in a form that includes both of these regions.

[0019] Figure 2 is a cross-sectional view showing the structure of the semiconductor device 1. Here, the active region R1 is shown on the left side of the figure, and the termination region R2 is shown on the right side. The center of the semiconductor substrate is located on the left side in Figure 2. In reality, the termination region R2 is formed to surround the active region R1 in a plan view; therefore, the structure of the termination region R2 in Figure 1 is symmetrical to that of the active region R1 and is also formed to the left of the active region R1. These specific planar structures will be described later.

[0020] In the active region R1, as with the structure in Figure 2, the back side (lower side in the figure) of the n-type semiconductor layer (n-layer 11) that becomes the drift region is the cathode, and the front side (upper side in the figure) is the anode, similar to the semiconductor device 9 described above. On the cathode side, a high-concentration n-type layer, an n-type cathode layer (back side first semiconductor region) 12, and a p-type cathode layer (back side second semiconductor region) 13 are alternately provided horizontally. These are formed with a thin n-type field stop layer 14 sandwiched between them and the n-layer 11. The impurity concentration of the field stop layer 14 is set to be higher than the impurity concentration of the n-layer 11 and lower than the impurity concentration of the n-type cathode layer 12.

[0021] Furthermore, a p-type anode region (surface-side semiconductor region) is formed across the entire surface of the active region R1. The anode region is provided with a shallowly formed first p-type anode layer (first surface-side semiconductor region) 15 and a deeply formed second p-type anode layer (second surface-side semiconductor region) 16. A cathode electrode (second main electrode) 21 is provided on the back side, and an anode electrode (first main electrode) 22 is provided on the front side. Therefore, in the operation of the semiconductor device 1, a current with rectifying characteristics flows between the anode electrode 22 and the cathode electrode 21, similar to a conventional diode, and the reverse recovery characteristics are improved, particularly due to the presence of the p-type cathode layer 13, as is the case with the technology described in Patent Document 1.

[0022] In Figure 2, only three second p-type anode layers 16 are shown on the surface side, along with the terminal p-type anode layer 16A, which will be described later. However, in the active region R1, a larger number of second p-type anode layers 16 are periodically arranged in the horizontal direction of the paper, and are similarly formed outside the illustrated area on the left. Each second p-type anode layer 16 is actually formed by extending in the direction perpendicular to the paper, as will be described later, and is arranged in the horizontal direction of the paper (the direction of extension and the direction perpendicular to it). Below, we will first describe only the structure of the region in the active region R1 where the second p-type anode layers 16 are arranged, and the structure of the terminal p-type anode layer 16A and the terminal region R2 will be described later. It is even more desirable that the spacing between the second p-type anode layers 16 be uniform.

[0023] On the back side, the n-type cathode layer 12 and the p-type cathode layer 13 are also repeatedly formed in pairs, corresponding to the structure on the front side. In this case, the n-type cathode layer 12 is provided directly beneath the second p-type anode layer 16, and its width is set to be wider than that of the second p-type anode layer 16 so that the second p-type anode layer 16 is included in the n-type cathode layer 12 in a plan view. Therefore, the p-type cathode layer 13 is not provided directly beneath the second p-type anode layer 16. Furthermore, the impurity concentration of the n-type cathode region 12 is higher than that of the p-type cathode region 13, and the impurity concentration of the p-type cathode layer 13 is higher than that of the second p-type anode layer 16.

[0024] This configuration allows for a reduction in the current density during primary breakdown, even if the primary breakdown voltage BV1 remains the same, thereby making secondary breakdown less likely to occur. This point will be explained below.

[0025] In the structure of Figure 14, as described above, the current path during primary breakdown is at the bottom edge of the second p-type anode layer 93B, making it easier for electrons to reach the vicinity of the p-type cathode layer 96 on the back side, thereby making it easier for holes to be injected from the p-type cathode layer 96 into the drift region (n-layer 91). In contrast, in the structure of Figure 2, this current path can be made to extend from the second p-type anode layer 16 to the entire area of ​​the corresponding n-type cathode layer 12 directly below it. This makes it more difficult for electrons to reach the p-type cathode layer 13, making secondary breakdown less likely to occur than in the structure of Figure 14. Furthermore, by providing multiple combinations of the second p-type anode layer 16 and the n-type cathode layer 12 directly below it, the breakdown points can be dispersed, making secondary breakdown even less likely to occur.

[0026] Figure 3 shows the simulation results of the current density distribution in the drift region (n layer 11) during primary breakdown in this structure, with the magnitude of the current density in the semiconductor substrate indicated by the shades of gray. As shown here, during primary breakdown, the current flows not in region X in Figure 14, but in a wide region from directly below the locally deepened second p-type anode layer 16 to the n-type cathode layer 12 below it. Furthermore, multiple second p-type anode layers 16 and an n-type cathode layer 12 directly below each second p-type anode layer 16 are provided, and their width is set wider than the second p-type anode layer 16 so that the second p-type anode layer 16 is included in the n-type cathode layer 12 in a plan view. This disperses these local current paths and further reduces the current density in each current path. As a result, breakdown due to local current concentration can be suppressed. Here, it is desirable that the pitch of the second p-type anode layer 16 in the horizontal direction (the direction of arrangement of the second p-type anode layer 16) is equal to the pitch of the n-type cathode layer 12. Furthermore, the central axis of the second p-type anode layer 16 and the central axis of the n-type cathode layer 12 directly beneath the second p-type anode layer 16 are aligned in a straight line. This makes it more difficult for electrons to reach the p-type cathode layer 13, thus reducing the likelihood of secondary breakdown.

[0027] The simulation results regarding this point will be explained. Figure 4 shows the results of a simulation calculating the relationship between the protrusion D from the first p-type anode layer 15 to the bottom of the second p-type anode layer 16, and the primary breakdown voltage BV1(a) and secondary breakdown voltage BV2(b), when the drift region (n-layer 11) is made of silicon with a resistivity of 23 Ω·cm and its thickness (distance from the second p-type anode layer 16 to the n-type cathode layer 12, etc.) is 50 μm. In this case, the width W of the second p-type anode layer 16 in Figure 2 is 5 μm, and its impurity concentration is 9 × 10⁻⁶. 16 cm -3 The depth of the first p-type anode layer 15 is 4.22 μm, and its impurity concentration is 1.08 × 10⁻⁶. 16 cm -3, the pitch P of the periods of the above-mentioned second p-type anode layer 16 and n-type cathode layer 12 is 100 μm, the width WN of the n-type cathode layer 12 is 50 μm, its depth is 0.55 μm, and its impurity concentration is 2.5×10 19 cm -3 , the width of the p-type cathode layer 13 is the difference between the pitch P of the above-mentioned period and the above-mentioned WN, its depth is 0.45 μm, and its impurity concentration is 1×10 18 cm -3 , the thickness of the field stop layer 14 is 1.25 μm, and its impurity concentration is 9×10 15 cm -3 is set as such. In the subsequent calculation results, unless otherwise specified or a value is clearly stated, the parameters are as described above.

[0028] In FIG. 4, when D = 0, it corresponds to the case where the second p-type anode layer 16 is not provided. From the results of FIG. 4(a), by increasing the protrusion amount D of the second p-type anode layer 16, the primary breakdown voltage BV1 decreases. This is because by increasing the protrusion amount D of the second p-type anode layer 16, the distance between the second p-type anode layer 16 and the n-type cathode layer 12 becomes shorter, and the current path in the n-layer 11 becomes shorter. On the other hand, the secondary breakdown voltage BV2 increases as the protrusion amount D increases, that is, secondary breakdown is less likely to occur. During primary breakdown, if the second p-type anode layer 16 is not provided, electric field concentration occurs directly below the first p-type anode layer 15. If the p-type cathode layer 13 is provided directly below the first p-type anode layer 15, electrons are likely to reach the p-type cathode layer 13, and secondary breakdown is likely to occur. On the other hand, as the protrusion amount D increases, current concentrates directly below the second p-type anode layer 16, so that more electrons easily flow into the n-type cathode layer 12 directly below the second p-type anode layer 16. Therefore, it becomes difficult for electrons to reach the p-type cathode layer 13, and secondary breakdown can be made less likely to occur. To increase the primary breakdown voltage BV1, it is effective to increase the resistivity of the silicon substrate (reduce the impurity concentration of the n-layer 11), and thereby, both the primary breakdown voltage BV1 and the secondary breakdown voltage BV2 can be increased.

[0029] From the results in Figures 4(a) and 4(b), it is preferable that the protrusion amount D is in the range of 1 μm to 5 μm in order to minimize the decrease in the primary breakdown voltage BV1 and increase the secondary breakdown voltage BV2. On the other hand, the depth of the first p-type anode layer 15 is set appropriately to the extent that such a protrusion amount D of the second p-type anode layer 16 can be secured and the characteristics such as the forward voltage of the diode can be secured. In order to secure the above range of protrusion amount D while maintaining these characteristics, it is preferable that the depth of the first p-type anode layer 15 is in the range of 0.2 to 1.0 as the ratio of (protrusion amount D) / (depth of the first p-type anode layer 15).

[0030] Furthermore, Figure 5 shows the results of calculating the primary breakdown voltage BV1(a) and secondary breakdown voltage BV2(b) by changing the width W of the second p-type anode layer 16 in Figure 2. Here, the protrusion amount D of the second p-type anode layer 16 in Figure 2 was set to 2 μm, and the width WN of the n-type cathode layer 12 was set to 50 μm.

[0031] As shown in Figure 5(a), no clear dependence of width W is observed in the primary breakdown voltage BV1. This indicates that primary breakdown occurs at both ends of the second p-type anode layer 16, regardless of the width W of the second p-type anode layer 16. On the other hand, as shown in Figure 5(b), the secondary breakdown voltage BV2 decreases when the width W becomes 60 μm. This is because, as the width W increases, primary breakdown occurs at both ends of the second p-type anode layer 16, making it easier for electrons to reach the p-type cathode layer 13, thus making secondary breakdown more likely.

[0032] From the results in Figure 5(b), it is preferable that the width W of the second p-type anode layer 16 be small. However, if the width W is too small, the current density will be high even if the current path occupies the entire width W of the second p-type anode layer 16. Also, a certain width D is necessary to ensure the protrusion amount D of the second p-type anode layer 16. Therefore, a lower limit of width D is preferably around 5 μm. Combining this with the results in Figure 5(b), a width D in the range of 5 μm to 60 μm, and particularly in the range of 5 μm to 40 μm, is preferable.

[0033] By providing a large number of combinations of the above-described second p-type anode layer 16 and n-type cathode layer 12, the current density in each current path can be reduced, thereby improving the secondary breakdown voltage BV2. To achieve this, reducing the pitch P is effective. Figure 6 shows the results of calculating the dependence of the primary breakdown voltage BV1(a) and secondary breakdown voltage BV2(b) on the above-described pitch P when the horizontal width of the active region R1 in Figure 2 is kept constant, the width W of the second p-type anode layer 16 is 5 μm, the protrusion D is 7 μm, and the width WN of the n-type cathode layer 12 is 50 μm.

[0034] As shown in Figure 6(a), the primary breakdown voltage BV1 is constant in the range of pitch P from 100 μm to 300 μm. This is because, as described above, the primary breakdown mechanism occurs at both ends of the second p-type anode layer 16. On the other hand, the secondary breakdown voltage BV2 decreases as the pitch P increases. This is because a narrower pitch P allows for greater dispersion of local current paths, reducing the current flowing directly beneath the second p-type anode layer 16, thus making secondary breakdown less likely. For this reason, it is preferable to reduce the pitch P to increase the secondary breakdown voltage BV2, but reducing the pitch P makes it difficult to make the n-type cathode layer 12 larger than the second p-type anode layer 16. Furthermore, this affects the ratio of the n-type cathode layer 12 to the p-type cathode layer 13, reducing the efficiency of hole injection from the p-type cathode layer 13 to the n-type layer 11 during reverse recovery, thus degrading the reverse recovery characteristics.

[0035] Next, the width W and protrusion D of the second p-type anode layer 16 were kept constant as described above, and the dependence of the primary breakdown voltage BV1 and secondary breakdown voltage BV2 on the width WN of the n-type cathode layer 12 was investigated. Figure 7 shows the calculation results for BV1(a) and BV2(b) in this case.

[0036] As shown in Fig. 7(a), in this case as well, there is no width WN dependence on the primary breakdown voltage BV1. On the other hand, as shown in Fig. 7(b), the secondary breakdown voltage BV2 increases as the width WN of the n-type cathode layer 12 increases. This is because when W < WN, the current during primary breakdown flows across the entire n-type cathode layer 12, and increasing the width WN of the n-type cathode layer 12 reduces this current density. As a result, it becomes difficult for electrons to reach the p-type cathode layer 13, resulting in less likelihood of secondary breakdown. Therefore, it is preferable for this width WN to be large, but when the width WN is 70 μm or more, the secondary breakdown voltage BV2 does not increase further. On the other hand, when increasing the width WN, it is necessary to increase the pitch P further or narrow the p-type cathode layer 13. Therefore, it is preferable to make the width WN smaller than 70 μm, and correspondingly, after providing a p-type cathode layer 13 with a width of a certain level or more, reduce the pitch P.

[0037] Also, in the structure of Fig. 2 during the forward direction, the n-type cathode layer 12 becomes the main current path on the back side. Therefore, in order to lower the forward resistance (lower the forward voltage VF), in the structure of the combination (for one cycle) of the n-type cathode layer 12 and the p-type cathode layer 13 adjacent to it on the back side, the existence ratio of the n-type cathode layer 12 on the back side (back-side n-type ratio) is defined as the width WN of the n-type cathode layer 12 / the pitch P of the cycle (WN < P), and the larger this back-side n-type ratio, the smaller the forward voltage VF becomes. On the other hand, as described above, the p-type cathode layer 13 is provided to improve the reverse recovery characteristics, so when the back-side n-type ratio is large, the reverse recovery characteristics deteriorate. Therefore, the forward voltage VF and the reverse recovery characteristics are in a trade-off relationship.

[0038] Figure 8 shows the relationship between Qrr (reverse recovery charge: the amount of charge required for the reverse recovery current to disappear during reverse recovery operation) and the forward voltage VF, measured for a sample of reverse recovery characteristics, while varying the back surface n-type ratio from 10% to 70%. From these results, there is a trade-off relationship between VF and Qrr as described above; when the back surface n-type ratio is small, Qrr is small and VF is large. However, compared to the case where the back surface n-type ratio is 30% or more, when the back surface n-type ratio is 20% or less, the rate of decrease in Qrr is small, while the increase in VF is significant. Furthermore, beyond 70%, little improvement in reverse recovery characteristics is obtained. For this reason, there is no advantage to making the back surface n-type ratio smaller than 20%, and it is preferable to keep the back surface n-type ratio in the range of 30-70%.

[0039] The structure in the active region R1 in Figure 2 has been described above. Next, the termination region R2, the structure on the termination region R2 side of the active region R1, and the planar structure of the semiconductor device 1 in Figure 2 will be described. As mentioned above, in this semiconductor device 1, it is preferable for primary breakdown to occur between the second p-type anode layer 16 and the n-type cathode layer 12 directly beneath it, and for each combination thereof, in order to improve the secondary breakdown voltage BV2. For this reason, it is preferable not to cause primary breakdown (secondary breakdown) in the termination region R2.

[0040] At the outermost edge (rightmost end) of Figure 2, a high-concentration n-type terminal n-type layer 17 is formed, set to the terminal potential. The first p-type anode layer 15 extends toward the terminal side (right side in the figure), and a resurf layer (resurf region) 18, which is p-type with a lower impurity concentration than the first p-type anode layer 15, is formed between the first p-type anode layer 15 and the terminal n-type layer 17.

[0041] Furthermore, a terminal p-type anode layer 16A is formed on the outermost side (right side in the figure) of the active region R1, with the same depth and impurity concentration as the second p-type anode layer 16. Since the n-type cathode layer 12 is directly below the terminal p-type anode layer 16A, the terminal p-type anode layer 16A functions similarly to the second p-type anode layer 16 in the active region R1, or the terminal p-type anode layer 16A becomes one of multiple second p-type anode layers 16.

[0042] However, in this structure, primary breakdown is likely to occur directly below the terminal p-type anode layer 16A, which is located at the outermost edge of the active region R1. Therefore, in order to reduce the current density in this area during breakdown, the width of the terminal p-type anode layer 16A in the horizontal direction in Figure 2 is made wider than that of the second p-type anode layer 16. Furthermore, the width of the n-type cathode layer 12 (terminal n-type cathode layer 12A) directly below the terminal p-type anode layer 16A is made wider than the width of the n-type cathode layer 12 located further inward (to the left in the figure). For example, as shown in Figure 2, the p-type cathode layer 13 may be located only inward from directly below the terminal p-type anode layer 16A, and the area directly below and outside the terminal p-type anode layer 16A may be the terminal n-type cathode layer 12A. While primary breakdown is likely to occur at the terminal end of the active region R1, by making the width of the terminal-side p-type anode layer 16A wider than the width of the second p-type anode layer 16, primary breakdown can be induced over a large area of ​​the terminal-side p-type anode layer 16A.

[0043] In Figure 2, an interlayer insulating layer 23 is formed on the resurf layer 18, and a terminal electrode 24, insulated from the anode electrode 22 by interposing this layer, is connected to the terminal n-type layer 17. As mentioned above, Figure 2 shows a cross-sectional structure, but as will be described later, in reality, the terminal n-type cathode layer 12A, terminal p-type anode layer 16A, resurf layer 18, and terminal n-type layer 17 are formed in a ring shape surrounding the active region R1 in a plan view. The terminal electrode 24 and the interlayer insulating layer 23 may also be ring-shaped in a similar manner, but their planar shapes are set as appropriate, as long as the potential of each electrode can be properly controlled.

[0044] The potential of the termination electrode 24 is set to be equal to, for example, the cathode electrode 21, and the above structure controls the potential distribution on the termination side of the semiconductor substrate so that the above operation in the active region R1 is performed appropriately. In this case, the breakdown voltage in the termination region R2 is mainly determined by the resurf layer 18. For this reason, although the description in Figure 2 is simplified, the horizontal length of the resurf layer 18 in the figure is actually set to be sufficiently long so that the breakdown voltage between the anode electrode 22 and the termination electrode 24 is higher than the breakdown voltage between the anode electrode 22 and the cathode electrode 21. Furthermore, it is more preferable to provide a field plate structure on the surface of the resurf layer 18 (in the interlayer insulating layer 23) to improve this breakdown voltage, as described in Japanese Patent No. 3275964, but this description is omitted here.

[0045] If a resurf layer 18 is provided, primary breakdown can occur between the resurf layer 18 or its left edge (active region R1 side) in Figure 2 and the back side, which can cause secondary breakdown in this area. In order to prevent breakdown in the terminal region R2, it is preferable that the second p-type anode layer 16 and the terminal p-type anode layer 16A be formed deeper than the resurf layer 18. That is, primary breakdown occurs at a lower voltage between each second p-type anode layer 16 and the n-type cathode layer 12 directly beneath it in the active region R1 than between the first p-type anode layer 15 and the terminal n-type layer 17 in the terminal region R2 in Figure 2, and between the resurf layer 18 and the n-type cathode layer 12 directly beneath it.

[0046] Figure 9 shows an example of the planar structure of the semiconductor device 1 in Figure 2. In Figure 9, as a plan view of Figure 2 from above, the planar shapes of the second p-type anode layer 16 and the terminal p-type anode layer 16A on the front side (both solid lines) and the planar shapes of the n-type cathode layer 12 and the p-type cathode layer 13 on the back side (both dashed lines) are shown. Here, the planar structure of the active region R1 is mainly shown, and the area outside the active region R1 shown here corresponds to the terminal region R2. Although only 13 dot-shaped second p-type anode layers 16 are provided here, in reality, a larger number of second p-type anode layers 16 are arranged and provided. In addition, the terminal p-type anode layer 16A is provided in a ring shape surrounding the arrangement of second p-type anode layers 16. The first p-type anode layer 15 is not shown in Figure 9, but as shown in Figure 2, the first p-type anode layer 15 is formed over the entire surface of the active region R1. In Figure 9, each second p-type anode layer 16 is connected to the first p-type anode layer 15 (and anode electrode 22) on its surface side, and their potentials become a common anode potential.

[0047] In Figure 9, on the back side, dot-shaped n-type cathode layers 12 correspond to each second p-type anode layer 16 and are formed wider than the second p-type anode layer 16 so as to include them in a plan view. This allows multiple combinations of the second p-type anode layer 16 and the n-type cathode layer 12 directly beneath it to be arranged in a row, thereby increasing the secondary breakdown voltage BV2. The width of the terminal p-type anode layer 16A is formed to be larger than the diameter of the second p-type anode layer 16. Also in Figure 9, multiple n-type cathode layers 12 are formed in the active region R1, and the space between the multiple n-type cathode layers 12 is a p-type cathode layer 13.

[0048] On the other hand, in Figure 9, outside the arrangement of the second p-type anode layer 16, the n-type cathode layer 12 is uniformly formed, including the region directly below the terminal p-type anode layer 16A. Therefore, the p-type cathode layer 13 on the back side is provided only in the active region R1, between the spaced-apart n-type cathode layers 12. Furthermore, the resurf layer 18 and the terminal n-type layer 17 in Figure 2 are formed in an annular shape surrounding the terminal p-type anode layer 16A, further outside of the terminal p-type anode layer 16A.

[0049] In this structure, the reverse recovery characteristics are improved (Qrr is reduced) by providing the p-type cathode layer 13 as described above, and the forward voltage VF can be reduced by increasing the area of ​​the n-type cathode layer 12. In this case, the secondary breakdown voltage BV2 can also be increased as described above. The structure in Figure 2 differs from the number of second p-type anode layers 16 as described above, but corresponds to the right half of the cross-section in the AA direction in Figure 9. The terminal n-type layer 17, resurf layer 18, etc. in Figure 2 are formed in an appropriate ring shape on the outside of the structure in Figure 9.

[0050] In the structure shown in Figure 9, each second p-type anode layer 16 is dot-shaped. However, to widen the current path during primary breakdown and lower the current density, there is a structure in which each second p-type anode layer 16 is striped (long strips in one direction). Figure 10 shows a planar structure of an example of this structure. Here, the pattern of the n-type cathode layer 12 corresponding to the rectangular second p-type anode layer 16 is a larger rectangle that includes it.

[0051] Similar to the structure in Figure 9, the cross-sectional view in Figure 2 corresponds to the right half of the cross-section in the BB direction in Figure 10. Figure 11 shows a cross-sectional view corresponding to the lower half of the cross-section in the CC direction (the longitudinal direction of the second p-type anode layer 16 where the second p-type anode layer 16 is located) in Figure 10, and Figure 12 shows a cross-sectional view corresponding to the right half of the cross-section in the DD direction (the arrangement direction of the second p-type anode layer 16 where the second p-type anode layer 16 is not formed on the surface side and the p-type cathode layer 13 is formed on the back side).

[0052] Note that the planar shape of the terminal p-type anode layer 16A in Figures 9 and 10, and the shapes of the p-type cathode layer 13 and n-type cathode layer 12 in Figure 10, are simplified to a rectangular shape. In this case, primary yielding is particularly likely to occur at the corners of the rectangular shape, so in practice, it is especially preferable to appropriately round these corners to an R shape.

[0053] Figure 13 shows a planar structure of a modified version of the structure in Figure 10. In Figure 10, multiple stripe-shaped second p-type anode layers 16 are independently formed parallel to each other in the active region R1, with the vertical direction in the figure as the longitudinal direction. However, in the structure of Figure 13, these are connected to the terminal p-type anode layer 16A on the outside (top and bottom in the figure). In Figure 13, the p-type cathode layer 13 is provided only on the back side of the region enclosed by the second p-type anode layer 16 and the terminal p-type anode layer 16A, while the rest of the region is an n-type cathode layer 12 (12A). In this structure as well, the n-type cathode layer 12 corresponding to the second p-type anode layer 16 is formed wider than the second p-type anode layer 16. In the structure of Figure 13, since there are no longitudinal ends of the n-type cathode layer 12, the occurrence of breakdown caused by these ends is suppressed, and breakdown can be more easily caused on the terminal p-type anode layer 16A side. Thus, a structure in which each second p-type anode layer 16 is connected at its end can also be used, in which case they can be connected by a terminal p-type anode layer 16A. Accordingly, the shape of the n-type cathode layer 12 can be set as appropriate.

[0054] Furthermore, as described above, it is preferable to provide a field plate structure on the upper side of the resurf layer 18, thereby ensuring withstand voltage without widening the termination region R2. The structure for this purpose can be appropriately set as described in Japanese Patent No. 3275964.

[0055] Furthermore, in the above example, the terminal p-type anode layer 16A, the resurf layer 18, the terminal n-type layer 17, etc., are formed in a ring shape on the outer periphery of the semiconductor substrate. However, depending on the structure in the active region of the semiconductor device, the form of these layers on the edge side of the semiconductor substrate can be appropriately set, for example, depending on where the breakdown voltage is limited. For this reason, although they are provided on the edge side of the active region, their form does not need to be ring-shaped. Alternatively, a groove can be provided from the upper surface of the terminal p-type anode layer 16A and the second p-type anode layer 16, but not penetrating through the terminal p-type anode layer 16A and the second p-type anode layer 16, and the anode electrode 22 can be embedded in the groove. In addition, although the above semiconductor device was described using an RFC diode as an example, the present invention may also be applied to the diode portion of an RC-IGBT.

[0056] Furthermore, other layers can be added or removed as appropriate in the semiconductor substrate. It is also clear that the same configuration can be applied even when all the p-type and n-type elements in the semiconductor are reversed in the above example. [Explanation of Symbols]

[0057] 1.9 Semiconductor Devices 11, 91 n layers (drift region) 12, 92 n-type cathode layer (first semiconductor region on the back side) 12A Terminal side n-type cathode layer (n-type cathode layer) 13. 96 p-type cathode layer (second semiconductor region on the back side) 14 Field Stop Layer 15, 93A First p-type anode layer (first surface-side semiconductor region: surface-side semiconductor region) 16, 93B Second p-type anode layer (Second surface-side semiconductor region: Surface-side semiconductor region) 16A Termination-side p-type anode layer 17 Terminated n-type layer 18. Resurf layer (resurf region) 21, 94 Cathode electrode (second main electrode) 22, 95 Anode electrode (first main electrode) 23 Interlayer insulating layer 24 Termination electrode 93 p-type anode layer R1 active area R2 termination area

Claims

1. A semiconductor device in which current flows between a first main electrode provided on the surface side of a semiconductor substrate and a second main electrode provided on the back side, The aforementioned semiconductor substrate is On the surface side, The drift region of the first conductivity type, A surface-side semiconductor region of a second conductivity type opposite to the first conductivity type is formed on the surface side of the drift region and connected to the first main electrode, It is equipped with, On the reverse side, The first semiconductor region on the back side of the first conductivity type, having a higher impurity concentration than the drift region, The second semiconductor region on the back side of the second conductivity type, However, they are formed alternately in a plan view, and the first semiconductor region on the back side and the second semiconductor region on the back side are connected to the second main electrode. The aforementioned surface semiconductor region is The first surface-side semiconductor region and A second surface-side semiconductor region is formed in contact with the first surface-side semiconductor region and is formed deeper than the first surface-side semiconductor region, It is equipped with, In a plan view, the first semiconductor region on the back side is formed directly below each of the second semiconductor regions on the front side, wider than the second semiconductor region on the front side, and corresponding to each of the second semiconductor regions on the front side.

2. The plurality of the aforementioned second surface-side semiconductor regions are formed in a unidirectional arrangement perpendicular to the longitudinal direction in a plan view. The semiconductor device according to claim 1, characterized in that the width of the outermost second surface-side semiconductor region along the aforementioned one direction is formed to be wider than the width of other second surface-side semiconductor regions.

3. The semiconductor device according to claim 1 or 2, characterized in that, in a plan view, each of the second surface-side semiconductor regions is formed in a dot shape.

4. The semiconductor device according to claim 1 or 2, characterized in that, in a plan view, each of the second surface-side semiconductor regions includes a portion formed in a stripe shape with a common longitudinal direction.

5. The impurity concentration in the second surface-side semiconductor region is set higher than that in the first surface-side semiconductor region. The semiconductor device according to claim 1 or 2, characterized in that a resurf layer having the second conductivity type and a lower impurity concentration than the first surface semiconductor region is locally provided on the surface of the drift region outside the outermost second surface semiconductor region.

6. The semiconductor device according to claim 5, characterized in that the resurf layer is formed to be shallower than the second surface-side semiconductor region and deeper than the first surface-side semiconductor region.

7. The semiconductor device according to claim 1 or 2, characterized in that, in a plan view, the second semiconductor region on the back side is not formed outside the outermost second semiconductor region on the front side.

8. In a plan view, On the surface side, the minimum width of the second surface semiconductor region is in the range of 5 μm to 60 μm. The semiconductor device according to claim 1 or 2, characterized in that the ratio of the area of ​​the first semiconductor region on the back side to the sum of the areas of the first semiconductor region on the back side and the second semiconductor region on the back side is in the range of 30% to 70%.

9. The semiconductor device according to claim 1 or 2, characterized in that the combination of the second surface-side semiconductor region and the corresponding first back-side semiconductor region is arranged periodically in a plan view.

Images