JPS592382B2 - reverse conducting thyristor - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention provides a new structure for a high performance reverse conduction thyristor capable of controlling switching of large amounts of power at high speed.

Reverse conduction thyristors are mainly used in various inverter and chopper fields, including vehicle choppers. As practical research in these fields progresses, there is a strong demand for an increase in the control power of reverse conduction thyristors, and for this purpose, it has become inevitable to increase the diameter of semiconductor wafers. However, when designing and manufacturing reverse conduction thyristors, as the diameter of the semiconductor wafer increases, fundamental problems that affect device performance arise, and it is difficult to achieve a given increase in control power by simply increasing the wafer diameter. is difficult. In other words, the crystal defects and crystal distortion that occur in semiconductor wafers during heat treatment processes such as diffusion tend to be unevenly distributed as the diameter of the wafer becomes larger. Accurate control of carrier life, which is strongly influenced, becomes very difficult.

[Conventional example]

FIG. 1 shows a cross section of a typical example of a conventional reverse conduction thyristor.

The plane is rotationally symmetrical. The conventional reverse conduction thyristor shown in FIG. 1 includes an n-type first base layer NB, a p-type first emitter layer PE Four layers of the second base layer PB and the n-type second emitter layer NE provided in an annular manner between the center and the periphery of its surface, that is, the second main surface, are successively formed adjacent to each other, and between them, the first Pn Junction J1, second Pn junction J2, third
This constitutes a Pn junction J3. The lower part of the annular n-type second emitter layer constitutes a thyristor part 11, and the surrounding part consisting of the first base layer NB and second base layer PB constitutes a diode part 12, both of which are arranged in antiparallel. It is built into the reverse conduction thyristor wafer 13 due to the following relationship. Second
Inside the emitter layer NE, there is a fourth annular layer spaced apart from this.
An emitter layer NE' is provided and has a junction J4 with the second base layer PB. The first main surface 16 of this reverse conduction thyristor J-HA 15 is generally provided with a first main electrode 1 which becomes an anode of an MO plate or the like through a thin layer of A1 or AlSi.
7 is alloy-brazed, and this also serves as a support plate. On the other hand, a ring-shaped second main electrode 19 is provided in common to the thyristor part 11 and the diode part 12 of the second main surface 18, and is also provided in the central part of the second main surface 18 where the second base layer PB is exposed. A gate electrode 21 is provided, and an auxiliary thyristor electrode 20 is provided at a position closer to the thyristor portion 11 on the outer annular fourth emitter layer NE'. The second main electrode 19 for sorting and the gate electrode 21 are generally formed of a vapor deposited layer of Al or the like. As described above, in the reverse conduction thyristor wafer 13, an arrangement structure in which the gate electrode 21 is provided in the center, the auxiliary thyristor portion 13 and the main thyristor portion 11 outside the gate electrode, and the diode portion 12 on the outermost side has been most commonly used.

The reason for this is that in order to speed up the spread of the turn-on region, it is most advantageous to place the gate electrode in the center of the semiconductor wafer (center gate structure), and the thyristor portion was naturally placed around it. In addition, since the center gate structure allows the element structure to be rotationally symmetrical, it is easy to work in the manufacturing process, especially in the electrode type and assembly process, resulting in increased manufacturing yield and high reliability. That's what it means. However, semiconductor wafers have large diameters, especially 50m7! When the diameter becomes larger than L, the following problems occur with this structure.

Generally, when a semiconductor wafer is heat-treated in an impurity diffusion process, an oxide film formation process, etc., many crystal defects are likely to occur, especially near the periphery of the semiconductor wafer. It has been known that this phenomenon becomes more pronounced as the diameter increases, making it difficult to uniformly introduce crystal defects throughout the wafer. Therefore, in the conventional structure example shown in FIG. 1, more crystal defects occur in the diode portion 12 than in the thyristor portion 11, and as a result, the carrier life due to crystal defects is longer in the thyristor portion 11 and shorter in the diode portion 12. . In addition, reverse conduction thyristors are often used for high-speed switching, and a short turn-off time is required.
Processes have been introduced to control carrier life using heavy metal diffusion methods such as gold, radiation irradiation methods, and heat treatment methods. However, the effects of these methods are closely dependent on crystal defects, and under the above-mentioned non-uniform distribution of crystal defects, the non-uniformity of the carrier life is further exacerbated. Therefore, in the characteristics of the obtained reverse conduction thyristor, the main thyristor portion 11 provided near the center
carrier life will not be too short. For this reason, although the turn-off time of the reverse conduction thyristor is not so shortened, the carrier life of the diode portion 12 provided in the other peripheral portion becomes shorter than necessary. As a result, leakage current in the diode portion increases, resulting in a fundamental problem of lowering the withstand voltage.
The present invention eliminates the drawbacks of conventional thyristors and provides a new element structure for obtaining a reverse conduction thyristor that uses a large-diameter semiconductor wafer and has both high-speed switching performance and high withstand voltage performance while being rated for large currents. This is what I am trying to do.

A feature of the present invention is that the gate region, thyristor portion, and diode portion that constitute a reverse conduction thyristor are arranged in a new manner on the wafer, with the main thyristor portion being placed in the outer peripheral area of the semiconductor wafer, and the diode portion being placed inside the main thyristor portion. By arranging the auxiliary thyristor section in the center and the gate region in the central region, the crystal defects that often occur in the outer peripheral region of the wafer can be effectively utilized to improve the characteristics of the thyristor section, and by placing the gate in the center. The feature of the center gate structure that the turn-on region spreads quickly as described above is fully utilized, and leakage of reactive current from the auxiliary thyristor electrode to the second base layer of the diode portion is prevented, and turn-on by the auxiliary thyristor is prevented. It is designed to be carried out effectively.

Next, the present invention will be described with reference to embodiments shown in FIG. 2a and subsequent figures.

A feature of the present invention is that from an electric path provided across the soil of the second base layer, such as an electric path connecting the auxiliary thyristor electrode and the gate auxiliary electrode provided inside the thyristor part, to the second base layer. The purpose is to prevent invalid current leakage of
In order to increase the impedance of the leakage current path, a separation layer having a conductivity type opposite to that of the second base layer is provided on the second base layer below the crossing current path.

The present invention will be explained below with reference to Examples.

In the first embodiment of the present invention, the diameter is 60 mm and the specific resistance is 130 mm.
~150Ω? A reverse conduction type thyristor was fabricated on a silicon wafer using known manufacturing methods such as the diffusion method, epitaxial method, and metallization method.

The structure is shown in plan view in FIG. 2a and in side cross-section in FIGS. 2b and 2c. FIG. 2b is a sectional view taken along the line bb in FIG. 2a, and FIG. 2c is a sectional view taken along the line aa in FIG. 2a. As shown in the above figures, in this embodiment, an n+ type layer is provided on the first principal surface side of the n type first base layer NB on the wafer 105, and an annular p type first emitter layer PE is provided on the outer periphery of the n+ type layer. A p-type third emitter layer PE/ is provided at the center, a p-type second base layer PB is provided in contact with the first base layer NB, and an annular n-type second emitter layer NE is provided at the outer periphery thereof. An annular fourth emitter layer NE' is provided at the center. However, the second emitter layer NE has two arcuate cutouts 115 and a radial cutout 116 extending from a part of each arc toward the center of the wafer.
is provided. In the reverse conduction thyristor, the first emitter layer PE, the n+ layer, the first base layer NB, the second Venice layer PB, and the second emitter layer NE in the peripheral part constitute the main thyristor part 101, and the n+ layer and 1st
The base layer NB and the second base layer PB are the diode portion 1
02, and a third emitter layer PE' inside thereof.
and the n+ layer, the first base layer NB, the second base layer PB, and the fourth
The emitter layer NE constitutes an auxiliary thyristor portion 104. Further, the second base layer PB, the first base layer NB<15n+ layer, and the third emitter layer PE' directly below the gate electrode 109 provided in the central portion constitute a gate region 103. Above n
The + layer is a layer having a higher impurity concentration than the first base layer NB, and by providing this layer, the reverse conduction thyristor is equivalently made into a known Pin structure, and the second base layer is a high proportional resistance layer.
The purpose is to enable the thickness of the base layer PB to be reduced and to reduce the on-voltage without impairing the withstand voltage. This n+ layer is provided by a conventional pitaxial method. On the first main surface 106 of the composite thyristor wafer 105, an anode electrode 107 of an MO plate is alloy soldered via an A1 vapor deposition layer, and on the second main surface 108, a central gate electrode 109 is formed by Al vapor deposition. A cathode electrode 110 provided in an annular shape on the outside thereof and having arcuate and radial openings, and a fourth emitter layer N
An auxiliary thyristor electrode 111 provided in an annular shape on E′, an auxiliary electrode 112 provided in the arc-shaped opening and entangled in the thyristor part 1, and an auxiliary thyristor electrode 112 provided in the arc-shaped opening and entangled in the thyristor part 1, Electrode 111 and auxiliary electrode 1
12 are respectively formed.

The arc-shaped auxiliary electrode 112 is located on the arc-shaped cutout 115 in the second emitter layer, and the radial electric path 1
13 is located on the radial cutout 116. Here, prior to the step of forming each electrode on the second main surface, a second base layer PB is applied to the surface portion of the second base layer PB in at least a part of the area under the location where the electric circuit 113 is arranged. An n-type isolation layer ND opposite to the above is provided, and a Pn junction J6 is formed between it and the second base layer. A is an anode terminal connected to the anode electrode 17, K is a cathode terminal connected to the cathode electrode 19, and G is a gate terminal connected to the gate electrode 21. In this device, if a main voltage with a polarity that makes the former positive is applied between the anode terminal A and cathode terminal K, and a gate voltage with a polarity that makes the former positive is applied between the gate terminal G and the cathode terminal K, as is well known, First, the auxiliary thyristor portion 104 is turned on and an on-current of 1 AT flows.

A part of the on-state current 1AT of this auxiliary thyristor portion 104 is leaked from the auxiliary thyristor electrode 111 to the electric path 11.
3→second base layer PB→cathode electrode 110. However, in this case, the electric line 113
A separation layer ND is provided on most of the contacting portion of the second base layer PB and the second base layer PB.
is provided, and its Pn junction J6 is in the opposite direction to the above leakage current and exhibits a very high impedance. Therefore, the electric path 113 → second base layer PB → cathode electrode 1
The impedance for the reactive current of the 10 paths is the electric path 1.
13→auxiliary electrode 112→second base layer PB→second emitter layer NE→cathode electrode 110, the impedance becomes significantly higher than the effective trigger current impedance for turn-on. As a result, the above-mentioned reactive current can be made very small, and most of the on-state current 1AT of the auxiliary thyristor section 104 can be used as the trigger current of the thyristor section 101. In order not to impair other characteristics of the reverse conduction thyristor, the depth Xjl of the separation layer ND should be shallower than the depth Xj2 of the second emitter layer NE for the following reasons.

That is, when the separation layer ND is arranged under the electric circuit 113 as shown in the figure, the part A where the separation layer ND and the first emitter layer PE face each other, and if an opening is formed, this part is Pnpn 4. It consists of layers and is equivalent to a thyristor structure.
Therefore, when a small number of carriers serving as diode current carriers enter this portion from the diode portion 102, this overlapping portion exhibits a thyristor function and is turned on, causing an undesirable phenomenon in which the commutation of the main thyristor portion 101 fails. is likely to occur. This phenomenon is caused by the separation layer ND
This is likely to occur if the impurity concentration of the emitter layer is high or if the layer thickness Xjl is thicker than the original emitter layer. Therefore, in order to prevent the above-mentioned undesirable turn-on phenomenon, it is necessary to
It is effective to make the depth Xjl shallow to make it difficult for the thyristor function to occur in this part. In addition, separation layer ND
It is also effective to set the impurity concentration as low as possible. In this example, the surface impurity concentration of the second emitter layer NE is 1 x 1020 cm-3, and the depth Xj2 of the same layer is 20 μ.
In contrast, the surface impurity concentration of the separation layer ND is 1×10190
By setting frL-3 and the same depth Xjl to 10μ, the effect of reducing the reactive gate current, which is an effect of the present invention, could be fully exhibited without lowering the commutation ability of the reverse conduction thyristor.

Furthermore, it is of course also effective to shape and arrange each part so that the separation layer ND and the second emitter layer PE do not face each other. In order to further improve the Di/Dt tolerance of the reverse conduction thyristor, it is effective to make the auxiliary electrode 112 intricately intertwine within the thyristor portion 101 and to lengthen the length of the cathode facing the auxiliary electrode 112. As is well known. However, this configuration has the drawback that the area occupied by the auxiliary electrode 112 increases, resulting in a decrease in the effective area of the thyristor, which inevitably necessitates a reduction in current capacity. Therefore, as a method for solving this effective area problem while exhibiting the separation layer effect of the present invention, an example in which a separation region between the thyristor portion and the diode portion is utilized will be shown in the second embodiment below. 3rd a and 3rd b
The figures are a plan view of a reverse conduction thyristor according to a second embodiment of the present invention, viewed from the cathode side, and a-a of Fig. 3a.
FIG. In this reverse conduction thyristor, the arrangement order of the thyristor portion 201, diode portion 202, auxiliary thyristor portion 204, and gate region 203 from the outer periphery to the center of the wafer 210 is the same as in the first embodiment. However, an isolation region 205 is additionally provided between the thyristor portion 201 and the diode portion 202. An electric path 206 connecting the auxiliary thyristor electrode 213 and the auxiliary electrode 207 is provided using this separation region. In this embodiment, as in the previous embodiment, in order to increase the impedance of the path from the electric path 206 to the second base layer PB to the second emitter layer NE of the diode portion 202 to the cathode electrode 208, dotted lines are shown in FIG. 3a. Arc-shaped electric circuit 2 as shown in the layout
06 and the second main surface 211 of the composite wafer 210, an n-type separation layer ND opposite to the second base layer is provided on the surface portion of the second base layer PB under the electric circuit. In general, the isolation region does not have the function of either a thyristor portion or a diode portion, and the existence of this isolation region has conventionally been wasteful from the standpoint of effective use of the wafer area for improving the power capacity of the device. However, in this embodiment, a large number (six in the figure) of auxiliary electrodes 207 are inserted into the thyristor portion 201 by utilizing the separation region, and each auxiliary electrode 207 is connected to the second emitter layer. This allows the thyristor portion 201 to be removed without significantly reducing the effective cathode area.
The length of the portion where the auxiliary electrode faces the cathode electrode 208 is made as long as possible. As a result, we succeeded in improving the turn-on switching characteristics of the reverse conduction thyristor, such as improving the rated critical on-current increase rate (Di/Dt withstand capability) and high frequency characteristics. On the other hand, in this embodiment, the isolation region 205 is a PnP layer consisting of a first emitter layer PEln+ layer, a first base layer NB, a second base layer PBl isolation layer ND.
It forms a thyristor-like structure of n4 layer structure. Therefore, when a minority carrier serving as a diode current carrier enters this isolation region 205 from the diode portion 202, this portion exhibits a thyristor function and turns on.
There is a concern that the commutation ability of the main thyristor portion 201 will decrease. Therefore, as described in the first embodiment, there is a method of reducing the depth of the NB layer, a method of lowering the impurity concentration of the ND layer, or a method of reducing the depth of the NB layer or the impurity concentration of the ND layer, or Methods such as removing a portion of the first emitter layer PE in the facing portion or reducing the depth or concentration of that portion so as not to face each other are also effective in solving the above problem. In addition,
In this embodiment, the isolation region 205 is provided rotationally symmetrically, but the cathode electrode 208 is connected continuously from the thyristor portion 201 to the diode portion 202 without providing an electric path 206 on a portion thereof. It is set up.

This is a structure for facilitating the workability of electrode terminal connection in the assembly process, for example, electrode connection by pressure contact, and the shape is not restricted in any way by the essence of the present invention. Therefore, the present invention can be implemented with any desired electrode shape in addition to the illustrated embodiment shape, and similar effects can be achieved.

It goes without saying that the present invention can sufficiently exhibit the same effect even in the case where the p-type and n-type of the reverse conduction thyristor shown in the embodiments are completely reversed.

Furthermore, although the above embodiments show the case where a silicon wafer is used, the effects of the present invention will not be lost even if other semiconductor wafers are used.

In the above embodiment, it was shown that the Au diffusion method was used to shorten the turn-off time, but other metals such as heavy metal impurity diffusion methods such as Cu, Pt, and Zn,
It was confirmed that the effects of the present invention can be exhibited in the same manner as in the above embodiments even when radiation irradiation, rapid heating and cooling, etc. are used, or even when these methods are not used to the utmost.

In this invention, as described in the above embodiment, the separation layer ND is provided on at least a part of the surface of the base layer PB under the electric circuits 113, 206, etc. provided on the second base layer PB. In a configuration in which a thyristor portion is provided on the outer periphery of the wafer and a diode portion is placed inside the thyristor portion as in the embodiment, in order to improve the turn-on characteristics and voltage resistance, an electric circuit is connected across the diode portion or further across a part of the cathode portion. 113, 206, etc., there is no risk of turn-on failure due to reactive current due to the provision of this electric path, and the desired performance can be achieved.

[Brief explanation of drawings]

FIG. 1 is a side cross-sectional view of a conventional reverse conduction thyristor, FIG. 2a is a plan view of the reverse conduction thyristor according to the first embodiment of the present invention, viewed from the cathode side, and FIG. 2b is a bb section of FIG. 2a. Figure 2c is a cross-sectional view taken along the a-a section in Figure 2a, Figure 3a is a plan view of the reverse conduction thyristor of the second embodiment as seen from the cathode side, and Figure 3b is a cross-sectional view taken along line a in Figure 3a. -a
FIG. A is the anode terminal, K is the cathode terminal, G is the gate terminal, PE is the first emitter layer, NB is the first base layer, PB is the second base layer, NE is the second emitter layer, and PE' is the third emitter layer. , NE' is the fourth emitter layer, J1 to J6 are junctions, 17, 107, 211 are anode electrodes, 19, 110
, 208 are force sword electrodes, 21, 109, 212 are gate electrodes, 20, 111, 213 are auxiliary thyristor electrodes,
ND is a separation layer provided on the second base layer PB under the electric circuits 113, 206, 11, 101, 201 are thyristor parts,
114, 205 are isolation regions, 12, 102, 202 are diode portions, 13, 104, 204 are auxiliary thyristor portions, and 14, 103, 203 are gate region portions.

Claims (1)

[Claims] 1. A first emitter layer of a first conductivity type, a first base layer of a second conductivity type, a second base layer of the first conductivity type, and a second emitter layer of the second conductivity type.
a thyristor part consisting of four layers of emitter layers adjacent to each other in sequence; a diode part consisting of two layers, the first base layer and the second base layer; the third emitter layer of the first conductivity type; and the first base layer. and the second base layer and a fourth layer of the second conductivity type.
A composite thyristor wafer includes an auxiliary thyristor portion including four emitter layers successively adjacent to each other; a first electrode adjacent to a first main surface where the first emitter layer and the second base layer are exposed; a second main electrode connected to the thyristor portion and the diode portion on a second main surface opposite to the first main surface; a gate electrode and an auxiliary electrode connected to the second base layer on the second main surface; , an auxiliary thyristor electrode connected to the fourth emitter layer on the second main surface;
At least one connection path is formed that is spaced apart from the second main electrode and is installed across the diode portion, and the connection path electrically connects the auxiliary thyristor electrode and the auxiliary electrode. Furthermore, a second conductivity type separation layer is provided in the second base layer to form a pn junction by being buried in at least a part of the area under the connection electric path, and the electric path is provided on the second conductivity type separation layer. A high impedance section is provided in the path of the connection circuit - the diode part - the second electrode, and from the center to the periphery of the composite thyristor wafer, the gate electrode, the auxiliary thyristor part, the diode part, and the thyristor are connected. A reverse conduction thyristor characterized in that parts are arranged in sequence. 2. The reverse conduction thyristor according to claim 1, wherein the depth of the separation layer is shallower than the depth of the second emitter layer. 3. The reverse conducting thyristor according to claim 1, wherein the surface impurity concentration of the separation layer is lower than that of the second emitter layer. 4. The reverse conduction thyristor according to claim 1, wherein the connection circuit is installed in at least a part of a separation region between the thyristor part and the diode part. 5. The reverse conduction thyristor according to claim 1, wherein the minority carrier lifetime is shortened by treating the composite thyristor wafer by a heavy metal diffusion method, a heat treatment method, or a radiation irradiation method. 6. The reverse conduction thyristor according to claim 1, wherein the first emitter layer is disposed in a shape such that it does not face the separation layer. 7. The reverse conduction thyristor according to claim 1, wherein the separation layer is shaped so as not to face the first emitter layer.