JPS5932904B2 - bidirectional photothyristor - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a bidirectional photothyristor that performs bidirectional switching operation using an optical trigger signal.

A bidirectional thyristor that switches in both directions based on a trigger signal usually has five successive layers of alternating conductivity, and these five layers have opposite polarities, with both outer layers serving as one end layer. The semiconductor device includes a semiconductor substrate formed with a pair of four-layer regions, a pair of main electrodes in ohmic contact with each outer layer and an adjacent intermediate layer of the semiconductor substrate, and means for applying a trigger signal. In such a thyristor, when a trigger signal is applied between the main electrodes while a voltage is applied such that one main electrode has a higher potential than the other main electrode, the direction from one main electrode to the other main electrode becomes normal. When the four-layer region in the direction becomes conductive, and conversely, when a voltage is applied between the main electrodes so that the other main electrode has a higher potential than one main electrode, and the trigger signal is applied, the voltage from the other main electrode to one of the main electrodes becomes conductive. The four-layer region whose forward direction is the direction toward the main electrode becomes conductive. As a means for applying a trigger signal, a so-called electric gate method is generally used, in which a gate electrode is provided on any layer of a semiconductor substrate, and a gate signal current is caused to flow from this gate electrode to cause a switching operation of the device.

However, the electric gate method has the following drawbacks. (1) In order to perform a symmetrical bidirectional switching operation, it is necessary to provide two gate electrodes, which requires two gate circuits that are electrically insulated from each other. (
2) When using one gate electrode, use two gate electrodes.
The gate electrode is provided in a location close to the four-layer region of the individual, and the gate electrode is in contact with the part of the outer layer exposed on the main surface opposite to the main surface on which the gate electrode is provided, when projected in the stacking direction. Since they are formed so as to overlap with each other, there is a risk of so-called erroneous ignition in which conduction occurs before a gate signal current is applied during commutation. (3) Furthermore, when there is only one gate electrode, it is difficult to obtain symmetrical switching characteristics. (4) Two 4 layers. In order to prevent interference between the regions, when projected in the stacking direction between the four layer regions, both outer layers are separated by a predetermined distance and an isolation region is formed by the middle three layer regions. However, when forming an isolation region, two gate electrodes are required (
It has the drawbacks mentioned in 1). (5) There is a risk of erroneous ignition due to the induction effect. Due to the drawbacks of the electric gate method mentioned above, bidirectional thyristors that perform bidirectional switching operation have been limited to applications that do not require high reliability.
Also, most of them had small current-voltage capacity.

One method that has recently attracted attention as a means for applying a trigger signal is to perform a switching operation by converting the trigger signal into light and irradiating it onto a semiconductor substrate.

There is a so-called optical trigger system. Compared to the electric trigger system, the optical trigger system has the following advantages: (1) two 4-layer regions arranged electrically isolated can be ignited by a single light source electrically insulated from them; One piece is enough. (
2) Since the bidirectional thyristor circuit and the gate circuit are electrically isolated, it is easy to apply the thyristor to high voltage circuits; (3) there is no risk of false firing due to inductive action, etc. There are advantages. However, the optical tricar system also has the following drawbacks. The first drawback is the problem of imbalance in ignition sensitivity between the two 4-layer regions.When igniting a 4-layer region, the carrier generation region effective for ignition is near the reverse biased central junction. Therefore, in order to improve the light ignition sensitivity, it is necessary to irradiate light near the central junction. When attempting to ignite two 4-layer regions by irradiating light from one main surface of the semiconductor substrate, a reverse biased region that is a generation region of a carrier effective for igniting the two 4-layer regions is used. The distance between the central junction and the light source differs by the thickness of the central layer among the five layers. In a 4-layer region with a junction far from the light source, the irradiated light is absorbed in the central layer and decays exponentially, so the light firing sensitivity is significantly lower than in a 4-layer region with a central junction close to the light source. do. In order to eliminate this imbalance in ignition sensitivity, it is conceivable to irradiate light onto the side end face that connects the pair of main surfaces of the semiconductor substrate, or to irradiate both main surfaces with light. In the former case, the disadvantage is that the light source (irradiation area) must be increased or used in two units for ignition to occur; in the latter case, two light sources are required, and it cannot be adopted due to the structure of the device. It has the disadvantage of being difficult. This is the second drawback. To explain the shortcomings of the former case, since bidirectional thyristors adopt a structure in which the outer layer and the adjacent intermediate layer are short-circuited by the main electrode, the two four-layer regions are adjacent to each other. Even if electron-hole pairs are generated at a certain point by the irradiation, they do not flow into the outer layer, but are mostly extracted directly from the intermediate layer to the main electrode, and do not contribute to ignition at all, so the optical trigger signal In order to fire the bidirectional thyristor, it is necessary to apply an optical trigger signal to each of the four-layer regions at a location away from the adjacent portions of the two four-layer regions. For this purpose, two light sources must be provided or a light source with a wide irradiation area must be used. A third drawback is that in order to prevent false ignition during commutation, ignition sensitivity is reduced or two light sources are required. In bidirectional thyristors,
During so-called commutation, when one thyristor changes from a conducting state to a non-conducting state and the other thyristor changes from a non-conducting state to a conducting state, a trigger signal is applied to the other thyristor due to the influence of the residual accumulated carrier of one thyristor. It has the undesirable characteristic of igniting before it is activated.
In order to improve this problem, it is known to form an isolation region between both thyristors that does not operate as a thyristor, and to further distribute heavy metal atoms in that region. This is because when heavy metal atoms are not distributed, the width of the isolation region increases and the utilization rate of the semiconductor substrate decreases. Such bidirectional thyristors require a light source for the trigger signal at a location remote from the isolation region, ie, for each thyristor, due to the short lifetime of the carrier in the isolation region and for the reasons mentioned above. With current technology, it is difficult to distribute heavy metal atoms only in isolated regions when the semiconductor substrate is relatively thick, and the heavy metal atoms are usually distributed throughout the semiconductor substrate. In such a case, no matter where the optical trigger signal is applied, the lifetime of the carrier is short, so ignition cannot be achieved unless the amount of light is increased. Due to the above-mentioned drawbacks, the optical trigger system has not been adopted as a trigger system for high-voltage, large-current bidirectional thyristors, even though it has advantages over the electric trigger system. The object of the present invention is to provide a novel bidirectional photothyristor which has symmetrical bidirectional current-voltage characteristics that are triggered by an optical trigger signal from the same light source, and which has substantially equal optical firing sensitivity in both directions. Our goal is to provide the following.

Another object of the present invention is to provide a novel bidirectional photothyristor that is free from false firing during commutation.

Still another object of the present invention is to provide a novel bidirectional photothyristor that can be applied to high-voltage, large-current circuits.

The bidirectional photothyristor of the present invention that achieves the above object is characterized by arranging two thyristor portions close to each other with opposite polarities and substantially the same characteristics, and beveling the portions of each thyristor portion that are close to each other. The main point is that the optical trigger signal is irradiated on the side surface of the sensor.

This makes it possible to realize a bidirectional photothyristor that has substantially equal light firing sensitivity in both directions using a light trigger signal from the same light source. Another feature of the bidirectional photothyristor of the present invention is that two thyristor parts having opposite polarities are separated from each other. This makes it possible to realize a bidirectional photothyristor with high commutation capability, high breakdown voltage, and large current. Further features of the bidirectional photothyristor of the present invention will become clear from the following description of the embodiments. 1 and 2 show a first embodiment of the bidirectional photothyristor of the present invention, in which 1 has a pair of main surfaces 11 and 12 located opposite each other and a side end surface 13 that communicates with the main surfaces. Then, NE, PBl, NB, .
A first semiconductor body part 2 comprising four layers of PEL has a pair of main surfaces 21, 22 located opposite to each other and a side end surface 23 communicating the main surfaces with each other, with no gap between adjacent layers between the main surfaces. NEs of different conductivity are alternately formed to form a PN junction.
, PB, , NB2, PE2, each layer of these parts is as shown in the drawing: NEl and NE2, PEl and PB2, NBl.
and NB2, and PBl and PE2 are formed to have approximately the same thickness.

This is achieved by forming both parts using the same semiconductor substrate as a starting material, as will be described later. Of course, the concentrations of each layer become approximately equal as a result. 3 and 4 are the PEl layer, the NEl layer and the PB on the main surface of the first semiconductor body portion 1.
An anode electrode and a cathode electrode 5 and 6 are in ohmic contact with the PE2 layer and the NE2 layer and the PB2 layer, respectively, on the main surface of the second semiconductor substrate portion 2;

Although not shown, the cathode electrode 4 and the anode electrode 5 are electrically connected. 7 has the first and second semiconductor substrate portions placed thereon so that the polarities are opposite to each other, that is, the first semiconductor substrate portion 1 has an anode electrode 3, and the second semiconductor substrate portion 2 has a cathode electrode 6, respectively. These are auxiliary support plates placed on the lower side and close to each other.

The first semiconductor substrate portion has side end surfaces 13 inclined with respect to the main surface so that the main surface 11 provided with the anode electrode 3 has a larger area than the main surface 12 provided with the cathode electrode 4, and the second semiconductor substrate portion 2 has side end surfaces 23 inclined with respect to the main surface so that the main surface 22 on which the cathode electrode 6 is provided has a larger area than the main surface 21 on which the anode electrode 5 is provided. A V-shaped groove 8 is formed between the parts. All the PN junctions are exposed at both side end face portions forming this letter-shaped groove 8. Reference numeral 9 denotes a light source for irradiating a part of the surface of the V-shaped groove 8 with an optical trigger signal.

As the light source 9, a solid state light emitting device such as a light emitting diode may be arranged close to the semiconductor substrate, or a window may be provided in the package housing the device and an external light source may be irradiated through the window. Light may be irradiated using a light transmitting member such as a fiberscope. In the figure, a passivation film for stably holding the surface of a semiconductor substrate portion and a package containing a device are omitted. Such a bidirectional photothyristor can be manufactured by forming two four-layer regions with opposite polarities on the same semiconductor substrate and then separating the two regions by forming a groove 8 between them.

That is, as is well known, an N-type silicon substrate is prepared, and P-type impurities are diffused from both main surfaces of the substrate to form PB,
, PEl, PE2, and PB2 are formed simultaneously. after that,
N-type impurities are selectively diffused from both main surfaces to form NEl and NE2.
are formed at the same time. Thereafter, grooves 8 are formed. This manufacturing method, in which two semiconductor body portions having the same characteristics can be obtained, is preferable from the viewpoint of making the characteristics of the two semiconductor body portions (thyristor portions) the same and from the viewpoint of workability. Next, the operation of the bidirectional photothyristor having such a configuration will be explained.

First, a voltage is applied such that the electrodes 4 and 5 have a more negative potential than the electrodes 3 and 6, and in the first semiconductor part, the NEl layer and the PB
The PN junctions Jll and NB between the l layer and the PE layer are forward biased, and the PN junction J,3 between the layers PBL and NB is forward biased.
The PN junction Jl2 between the l layer is in a reverse bias state,
In the second semiconductor body portion, the PN junction J22 between the PB2 layer and the NB2 layer is forward biased, and the PN junction J2 between the NE2 layer and the PB2 layer and the PN junction J22 between the NB2 layer and the PE2 layer are forward biased. Let us explain the case where is in the Yudo Asu state. In this state, when the light source 9 irradiates the surface of the groove 8 with light as a trigger signal, the first semiconductor substrate portion which is in the forward blocking state is ignited. The ignition mechanism will be explained in detail. By irradiation with light, electron-hole pairs are generated in the vicinity of the exposed portion of the groove 8 of the PN junction J,2, and the electrons are collected in the NB layer and the holes are collected in the PB layer, respectively. The holes gathered in the PBL layer are PB,
The potential of the layer is gradually lowered, and the voltage applied to the junction at the right end of the PN junction Jll becomes the barrier voltage (Built-1nu01
When the voltage becomes higher than tage), injection of electrons from the NEl layer to the PBL layer starts at that point. The electrons injected into the PBl layer pass through the PN junction Jl2 and diffuse into the NB layer.Due to the diffusion of these electrons, the potential of the NBl layer increases and P
When the bias voltage of N junction Jl3 becomes higher than the barrier voltage, P
Holes are injected from the El layer to the NBL layer. The injected holes diffuse from the NBL layer to the PBL layer and forward bias the PN junction J, . This facilitates injection of electrons from the NEl layer. When the above operation is repeated and the sum α1+αNPN of the current amplification factors for each transistor becomes greater than 1, the four-layer region consisting of NEl, PB, NBl, and PEl becomes conductive. Next, when a voltage is applied such that the electrodes 4 and 5 have a more positive potential than the electrodes 3 and 6, the same operation as described above is repeated in the second semiconductor substrate portion which is in the forward blocking state. NE2, PB2, NB2, PE
The 4-layer region consisting of 2 becomes conductive. As is clear from the above description, in the bidirectional thyristor of the present invention, grooves are formed between the two semiconductor substrate parts to expose the central junctions Jl2 and J22, respectively, and the grooves are irradiated with light as a trigger signal. The bidirectional photothyristor of the present invention allows the switching operation to be performed with substantially equal sensitivity in both directions using the same light source. Since a groove is formed between the semiconductor base portions to isolate the two portions, false ignition due to residual carrier during commutation is eliminated, and commutation withstand capability can be increased.

Furthermore, according to the bidirectional photothyristor of the present invention, the exposed side end surface of the PN junction (which has a withstand voltage) of each semiconductor base portion is inclined with respect to the PN junction, so that no residue remains during commutation. Erroneous firing due to carrier is eliminated,
It becomes possible to increase the commutation withstand capacity. Furthermore, according to the bidirectional photothyristor of the present invention, P of each semiconductor substrate portion
Since the exposed side end face of the N junction (which has a breakdown voltage) is inclined with respect to the PN junction, it is possible to achieve a high breakdown voltage. Furthermore, according to the bidirectional photothyristor of the present invention, the portion to which the optical trigger signal is applied is the side end face, and the electrode can be attached to the entire main surface.Therefore, the utilization rate of the semiconductor substrate is high, and the current capacity is small and the current capacity is high. It becomes possible to obtain a large device. FIG. 3 shows a second embodiment of the bidirectional photothyristor of the present invention, which provides a device with higher voltage resistance than the first embodiment. The symbols of each part are the same as in the first embodiment. As one means for increasing the breakdown voltage of a semiconductor element, so-called beveling is known, in which the side end surface of a semiconductor substrate where a PN junction is exposed is inclined with respect to the PN junction. There are two types of beveling: one is to slope the side end face so that the cross-sectional area parallel to the PN junction on the side with higher impurity concentration is larger than the other side, and the other is to slope it in the opposite direction. This is called negative beveling. And when obtaining the same breakdown voltage, P
It is also known that the angle between the N-junction and the side end surface (bevel angle) is significantly smaller in the case of negative beveling than in the case of positive beveling. If the side end faces are inclined at the same bevel angle as in the first embodiment, the withstand voltage of the device is determined to be low by the PN junctions Jl2 and J23, which have negative beveling.
The second embodiment is a PN junction Jl2,J with negative beveling.
PN junction Jl3, where the bevel angle of 23 is a positive bevelling,
The bevel angle is smaller than that of J22, and the P is negatively beveled.
The structure is such that the breakdown voltage of the N junction is made close to that of the positively beveled PN junction, thereby increasing the breakdown voltage of the device. Although grooves having such a shape can also be formed by sandblasting, it is preferable to use etching. When using the etching method, first a shallow groove for negative beveling is formed, and then a deep groove for positive beveling is formed. Reference numeral 10 is an insulating material such as silicone rubber or glass that covers the side end surface.

Next, the effects of the present invention will be explained using numerical examples.

The device according to the present invention has a structure shown in FIG.
We prototyped a 100A bidirectional photothyristor. The diameter of the semiconductor substrate is 1.6 cm, the thickness is 350 μ, and the width of the upper end of the groove is 2.2 cm.
It is m1. The prototype photothyristor has a power output of 0.1W (light output of 1
0mW) Bidirectional switching operation is performed by the optical trigger signal from the semiconductor light emitting device, and the current reduction rate is 100A/μ.
It was confirmed that the commutation Dv/Dt withstand capacity of 100V/μs or more was provided when s. On the other hand, in a 3A, 200V bidirectional thyristor that uses an electric gate method, the gate power required for bidirectional switching is 3.
5v x 300mA, and the Dv/Dt tolerance during commutation is at most 20V/μ even when the current reduction rate is 1A/μs.
It is s. In this case, the width between the two thyristors is approximately 1m1.
L isolation regions are formed.

[Brief explanation of the drawing]

FIG. 1 is a plan view showing a first embodiment of the bidirectional photothyristor of the present invention, FIG. 2 is a sectional view taken along the line - in FIG. 1, and FIG. FIG. 4 is a cross-sectional view showing a third embodiment of the present invention. 1...First semiconductor base portion, 2...
Second semiconductor base portion, 3, 4, 5, 6... Electrode, 7... Auxiliary support plate, 8... Groove, 9
······light source.

Claims (1)

[Claims] 1. (a) A first semiconductor body portion having a pair of main surfaces located on opposite sides and having four consecutive layers of PNPN having different conductivity types alternately between the main surfaces; (b) a first semiconductor body portion; (c ) a support member that supports the first semiconductor base portion and the second semiconductor base portion on one main surface side of each; and (d) an outer P exposed on one main surface of the first semiconductor base portion. (e) a first electrode means for electrically connecting the layer and an outer N layer exposed on one major surface of the second semiconductor body portion; and (e) a first electrode means exposed on the other major surface of the first semiconductor body portion. (f) second electrode means for electrically connecting the outer N layer exposed on the other main surface of the second semiconductor body portion and the outer P layer exposed on the other main surface of the second semiconductor body portion; means for irradiating an optical trigger signal to each of the semiconductor substrate portions from the other main surface side, wherein the first and second semiconductor substrate portions are arranged on both main surfaces of the semiconductor substrate of one conductivity type. After sequentially diffusing impurities of the other conductivity type from almost the entire surface and then diffusing impurities of one conductivity type from selected surfaces of both main surfaces, the cross-sectional area of each part parallel to the main surface is diffused from the other main surface. The optical trigger signal is formed by dividing the first semiconductor substrate portion into a shape that becomes larger toward one main surface, and the optical trigger signal connects the two main surfaces of the first semiconductor substrate portion and the second semiconductor substrate portion at locations close to each other. A bidirectional photothyristor characterized in that the side surface of the photothyristor is illuminated.