SE438577B - SEMICONDUCTOR HIGH VOLTAGE DEVICE - Google Patents

Description

H0 aoosvus-6 g 2 handlebars and cathode. This limitation in turn means that the device can only be used at relatively low operating voltages when it comes to connecting two such devices in antiparallel connection, i.e. with each cathode connected to the other anode. Such a circuit would be useful if it could be designed as a bi-directional semiconductor switch for high voltage. Additional problems are that in the ideal case the base region should be heavily doped to avoid penetration from the anode to the handlebars; however, this leads to a low breakthrough voltage between anode and cathode. Widening of the base region limits the penetration effect, but at the same time increases the resistance of the device in the "ON" state.

It is desirable to provide a semiconductor switch that can be easily integrated so that two or more switches can be fabricated simultaneously on a common substrate and so that each switch can block relatively high voltages in a bi-directional manner.

Summary of the Invention The present invention relates to a structure having a semiconductor body whose bulk is of a first conductivity type and which has a main surface, the semiconductor body being made on a semiconductor support (substrate) of the opposite semiconductor type. In the semiconductor body there are a first and a second region which are separated from each other. The first region is of the first conductivity type and the second region is of the opposite conductivity type.

Each region has a part of the same which extends to the main- Both regions have relatively low resistivity in relation to said first and second surface. to the main mass (bulk) of the semiconductor body. regions serve as anode and cathode in the structure. Separate lower-resistive electrical contacts are connected to the anode region and the cathode region. The structure is arranged so that during operation a bidirectional charge carrier injection is obtained. the substrate and the substrate are arranged to enable an electrode to be connected thereto to serve as a guide in the structure. When this structure is suitably constructed, it can operate as a switch which is characterized by a low impedance current path between anode and cathode in the ON state (the conductive state) and a high impedance current path between anode and cathode in the OFF state (the blocked state). The potential applied to the control determines the state of the switch device. In the ON state, bi-directional charge carrier injection takes place which lowers the resistance between the anode and l The body is in contact with 10 15 20 25 _m 35- H0.-._..- ._.._...____ 8005746-6 cathodef In another embodiment of the invention, a semiconductor body as described above, having said first and second regions, as if on its part present on the main surface, is surrounded by a semiconductor region of the opposite conductivity type. A plurality of semiconductor bodies, each with a surrounding special semiconductor region, are made in a common wafer of the first conductivity type, with parts of the semiconductor brie separating all semiconductor bodies from each other.

These structures, which will hereinafter be referred to as Gated Diode Switches (GDS), can, when properly designed, block relatively large potential differences between anode and cathode in the OFF state, regardless of polarity, and they can in ON- - the condition conduct relatively large currents with a relatively small voltage drop between the anode and cathode.

The bidirectional blocking characteristics of these GDS structures make them particularly useful in many applications. Two of the GDS described above can be connected to the guides connected to each other and to each other's anode to the other's cathode. This combination forms a bidirectional high voltage switch. can be manufactured on a common semiconductor wafer to form junction points, or two GDS can be manufactured with a common guide to form a bidirectional switch.

The invention will be described in more detail in the following in connection with exemplary embodiments shown in the accompanying drawing with Figs.

Fig. 1 shows a structure in accordance with an embodiment of the invention. Fig. 2 shows a proposed electrical symbol for the structure shown in Fig. 1. §Ig¿_§ is a top view of a structure in accordance with another embodiment of the invention. Fig. U shows a structure in accordance with a further embodiment of the invention.

Fig. 5 shows a structure in accordance with yet another embodiment of the invention. Fig. 6 shows a structure in accordance with a further embodiment of the invention, and Fig. 7 finally shows a structure in accordance with a further embodiment of the invention.

Detailed Description Fig. 1 shows a semiconductor structure 10 comprising two substantially identical controlled diode switches GDS1 and GDS2 which are shown within dashed rectangles and which are both made in a semiconductor wafer or a semiconductor substrate 12. The semiconductor structure 10 has a main surface 11. is of a first conductivity type 'Aggregate with a plurality of GDS 10 15 20 25 30 35 H0 80057116-6' H and acts as a common control and as support for GDS1 and GDS2.

An epitaxial layer of the opposite conductivity type relative to the substrate 12 is isolated by semiconductor regions 20 to form semiconductor bodies 16 and 16a. Many bodies 16, 16a may be made on the substrate 12 in addition to the two shown. The regions '20 are of the same conductivity type as the substrate 12 but have a higher disturbance concentration and extend from the main surface 11 down to the substrate 12. of the same conductivity type as the body 16 but with higher disturbance con- A semiconductor region 22 is of the same conductivity type as a semiconductor- Within the body 16 there is also a semiconductor anode region 18 centration. body 16 but has lower resistivity than body 16. -cathode region 24 is part of the region 22 and has a portion extending to the major surface 11. The region 24 is of the same conductivity type as the regions 20 and has substantially the same disturbance concentration as these. The electrodes 28, 32 and 30 make low-resistance contact with the regions 18, 23 and 23, respectively. 20. The region 20 makes low-resistive contact with the substrate 12. The electrode 30 is thus in low-resistive contact with the substrate 12 and serves as a common control electrode for GDS1 and GDS2. An electrode 38, which may be provided but which does not necessarily have to be present, is shown between the anode electrode 28 and the cathode electrode 32. The electrode 38 is electrically connected to the substrate by an electrical connection to the electron 30.

.In the body 16a there are semiconductor regions 18a, 22a and 2Ha. The electrodes 28a, 32a and 30 are connected to the regions 18a, 24 and 24, respectively. 20.

These regions substantially correspond to corresponding regions in the body 16. An insulating layer 26 electrically insulates all of the electrodes described above from those parts of the structure 10 to which they are not intended to be electrically connected.

In an illustrative embodiment, the substrate 12 is of the conductivity type n, the regions 20 and 2Ä (2Ua) of the conductivity type n +, the body 16 (16a) of the conductivity type p-, the region 18 (18a) of the conductivity type p +, the region 22 (22a) of the conductivity type p and with lower resistivity than the body 16 (16a), and the electrodes 28 (28a), 32 (32a) and 30 are of aluminum. In this embodiment, the anode electrode 28 is electrically coupled to the cathode electrode 32a, and the cathode electrode 32 is connected to the anode electrode 28a.

Suggested electrical symbols for GDS1 and GDS2 are shown in Fig. 2.

The connection terminals to the anode, cathode and guide in GDS1 are 28, 32 resp. The corresponding connection terminals for GDS2 are 28a, 32a and 30. This combination of GDS1 and GDS2 acts as a bidirectional 81o 15 20 25 30 35 H0 s 8005746-6 switch that can double-sided block potentials regardless of whether the anode or cathode in any of the the controlled diode switches have the more positive potential. 2 In GDS1 and GDS2, both are largely identical and function in essentially the same way. Consequently, the following description of GDS1 also applies to GDS2. When GDS1 is in the ON state (conductive), it is characterized by a relatively low resistive current path between the anode region 18 and the cathode region ZH and when in the OFF state (blocking), it is characterized by a significantly higher I ON state. potential in typis- Holes are injected into the body impedance. cases at or below the potential of the anode 28. 16 from the anode region 18 and electrons are injected into the body 16 from the cathode region 2D. These holes and electrons may be present in sufficient numbers to form a plasma which modulates the conductivity of the body 16. This results in a decrease in the resistance of the body 16, so that the resistance between the anode region 18 and the cathode region 2U is relatively low when GDS1 operates in ON the condition. This method, in which both holes and electrons act as charge carriers, is called bidirectional charge carrier injection. The type of structure described here is called "controlled diode switch1" (GDS).

The region 22 contributes to limiting the penetration through a depletion layer formed during operation between the region 20 and the substrate 12 and the cathode region 24. The region 22 also contributes to the formation of a surface inversion layer between the regions_2M and 20. In addition, it allows the anode region 18 to be placed and the cathode region 2H are relatively close to each other. This results in a relatively low resistance between the anode region 18 and the cathode region ZÅ in the ON state.

Lead between the anode region 18 and the cathode region 24 is blocked or broken if the potential on the gate electrode 30 is sufficiently much more positive than the potential on the anode electrode 28 and the cathode electrode 32.

The amount of positive excess potential required to terminate the conductive state depends on the geometry and disturbance concentration levels in structure 10. This positive gate potential causes the portion of the body 16 located between the gate region 12 and a portion of the oxide layer 26 to be depleted. so that the potential in this part of the body is more positive than the potential on the anode region 18 and on the cathode region 2H. This positive potential barrier prevents perforation from the anode region 18 to the cathode region 2ü. It also causes the electrons emitted in the cathode region (2ü) to be collected before they can reach the anode region 18. This entails essentially Ioosvas-sp as a constriction of the body 16 against the dielectric layer '2É in the bulk part of the body located between the anode region (18) and the cathode region (2ü) and extending from the region 12 to the dielectric skin 26. The use of the electrode 38 reduces the size of the potential required for the termination of the management license.

In the OFF state, GDS1 can double-sidedly block relatively high potentials between the anode region and the cathode region, regardless of which of these regions has the most positive potential.

When GDS1 is in the ON state, the pn layer diode formed by the body 16 and the region 20 is biased in the forward direction.

Current limiting means (not shown) can be used to limit the conductivity through the forward biased dicde. It is not necessary for GDS1 and GDS2 to have their anodes and cathodes interconnected. GDS1 or GDS2 can be used individually, but the gate electrodes are common.

Fig. 3 shows a top view of a preferred embodiment of a bidirectional GDS semiconductor structure 100 having the front structure. The structure 100 is similar to the structure 10 except that the anode direction tends to be set. and the cathode region are curved. limit local voltage field concentrations that cause voltage breakthroughs and it magnifies the perimeter common to the anode region and the cathode region to ensure low resistance in the ON state, so that the device can be used for high current operation. The structure 100 has been fabricated on an n-type substrate having a thickness of H57 to 599 micrometers and a conductivity of 1015 to 1016 interfaces / cm 3. The bodies 160 and 160a are of the conductivity type and have a thickness of 30 to RO micrometers, a width of 720 micrometers, a length of 910 micrometers and a substance concentration in the range 5 - 9 X 10 13 / cm 3. The curved anode regions 180 and 180a are of the conductivity type p + and have a thickness of 2 to mik micrometers and a disturbance concentration of 1019 / cm 3. The curved cathode regions 2H0 and 2U0a are of the conductivity type n + with a thickness of 2 to 4 micrometers and an impurity concentration of 1019 / cm3. The total length and width of the manufactured circuit is 1910 micrometers times 1300 micrometers. The gap between the anode and the cathode is in a typical case 120 micrometers.

Some of the fabricated structures contained conductor regions 380, 380a with a width of 60 micrometers, while others did not contain any such regions. The structures made without the regions 380, 380a required 22 volts higher potential on the guide than on the anode to put the distance between the anode and cathode in a non-conductive state.

The structures fabricated with conductor regions 380, 380a required a potential excess of only 7.5 volts above the anode potential to provide non-conductive states. The fabricated structure was found to be able to block 300 volts and conduct 500 milliamps with a voltage drop between anode and cathode of 2.2 volts. This structure could operate at surges of 10 amps with a duration of one millisecond.

Fig. U shows a structure very similar to the structure 10 and all parts of which are substantially identical or similar to those of the structure 10 have been given the same reference numerals with two zeros added at the end. The basic difference between the structures 1000 and 10 is that in the semiconductor regions 22, 22a present in the structure 10 of the structure 10 of Fig. 1 have been eliminated. Purpose distance between the regions 2H00, 2ü00a and the region 2000 provides sufficient protection against depletion layer penetration to the regions 2400, 2U00a and makes it possible to use the structure 1000 as a high voltage switch.

Fig. 5 shows a structure 10,000 which is very similar to the semiconductor structure 10 and all parts of which are substantially identical to the corresponding ones in the structure 10 are provided with the same reference numerals with the addition of three zeros at the end. The fundamental difference between structures 10,000 and 10 is the use of semiconductor protection regions H0, U0a which surround regions 2N, 000, 2U, 000a and are separated from them by parts of the bodies 16,000, 16,000a. The protection ring regions H0, ü0a provide the same type of protection against surface layer inversion as the regions 22, 22a in the structure 10. The protection is assumed to be most complete in some cases for providing a semiconductor switch which can be used for high voltages. The protective rings H0, ß0a are of the same conductivity type as the bodies 16,000, 16,000a but have lower resistivity. The protective rings H0, B0a can be extended (according to what is shown in dashed lines) so that they are in contact with the cathode regions zmooo, zlnoooa. Fig. 6 shows a structure 100,000 similar to structure 10.

All parts of the structure 100,000 which are similar or substantially identical to the corresponding parts of the structure 10 have been given the same reference numerals with the addition of four zeros at the end.

One difference between structure 100,000 and structure 10 is the use of semiconductor protection ring regions H00, H00a which enclose the cathode regions 2N0,000, 2H0,000a. The protective rings H00, U00a are similar to half- I are extended so that they make contact with the cathodes 2ü0,000, 2U0,000a. -10 15 20 025 30 35 40 80057lf6 ~ 6 dare ~ protection ring regions H0, H0a in structure 10,000. The dashed lines show the part of the protective rings 400, Ä00a shows that they can In the combination of regions 220,000, 220,000a and the protective rings 400, H00a provides protection against inversion in the bodies 160,000, 160,000a, especially between the control region 200,000 and cathode region 2U0,000, 2U0,000a, and provides protection against penetration from the depletion layer to the cathode region 2ü0,000, 2N0,000a. This type of double protection around the cathode region 2H0,000, 2U0,000a is the preferred protection structure. The regions 220,000, 220,000a and Ä00, H00a are all of the same conductivity type as the bodies 160,000, 160,000a but with low resistivity. The regions N00, 400a have lower resistivity than the regions 220,000, 220,000a. Another difference between structure 100,000 and structure 10 is the semiconductor regions 70, 70a, which are of the same conductivity type as the cathode regions 2ü0,000, 2H0,000a. The regions 70, 70a are in electrical contact with the electrodes 380,000, 380,000a and act as top control electrodes. The use of the control regions 70, 70a results in a reduction of the potential required to prevent conduction between the anode regions 180,000, 180,000a and the cathode regions 2H0,000, 240,000a.

Fig. 7 illustrates a semiconductor structure 42 containing a plurality of substantially identical controlled diode switches ("GDS") of which only two, GDS3 and GDS4 (shown in dashed rectangles) are shown. which is of a first conductivity type and which has a main surface H6. Within a part of the substrate H4 there are separate regions H8 and H8a which are of opposite conductivity type in relation to the substrate UH and which are separated from each other by parts of the substrate UH and by regions 50 of the same conductivity type as the substrate UR but with a higher concentration of interfering substance.Regions 50 can be arranged or omitted as desired, Substantially identical semiconductor bodies 52 and 52a are located in regions H8 and 48a, respectively.

The bodies 52 and 52a are of the same conductivity type as the substrate HH. Within the body 52 there is an anode region SU which is of has the same conductivity type as the body 52 but has a higher concentration of interfering substance.

Within the body 52 there is also a region which is of the same conductivity type as the body 52 but with a higher concentration of disturbance and which is separated from the region 54 by parts of the body 52. A cathode region 58 is present within a part of the region 56 and is different from the body 52 through parts of the region 56. The cathode region 58 is of the same conductivity type as the region H8. The electrodes 60, 62, 64 and 66 are in low resistive contact with the regions 48, 54, 58 and 50. If the regions 50 are eliminated, the electrode 66 contacts the region UB directly or via a low-resistance semiconductor region (not shown) as well as the region SU, but is enclosed in a part of the substrate 4%. An insulating layer 68, typically silicon dioxide, electrically insulates all electrodes in the structure N2 from the main surface H6 except in the regions in which low resistive contact is desired.

The body 52a, the regions Säa, 56a and 58a and the electrodes 60a, 62a and ößa in GDSU are substantially identical to the corresponding regions in GDS3. The substrate UR is typically kept at the most negative potential available. This involves biasing the pn junction formed by the regions 48, 48a and the substrate HU so that all the GDS present in the substrate HN are junction-isolated from each other. In GDS3 and GDSH operate in substantially the same manner as that of GDS1 and GDS2 in Fig. 1. The region RB serves as the board, while the regions 54 and 58 serve as the anode and cathode. It should be noted that the gate regions H8 and H8a are geometrically and electrically separated and that GDS3 and GDSH can thus operate substantially completely independently of each other, since the respective gates, anodes and cathodes are electrically separated. The structure H2 thus enables the manufacture of an assembly of GDS, whereby each GDS can operate independently of all other GDS in the assembly.

The embodiments described herein are intended as illustrative examples of the general principles of the invention. For example, the disturbance concentration levels, the spaces between the regions and other dimensions of the regions can be changed to enable significantly higher operating voltages and currents than what has been described.

A dielectric layer may be provided between the regions M8 and ä8a and the region UU or the dielectric layer may be provided as a replacement for the regions HN and 50. In addition, other types of dielectric materials, for example silicon nitride, may be used instead of silica. Conductive regions, such as 38 in Fig. 1, may be included in the structures of Figs. 3, N, 5, 6 and 7. Regions 56 and 56a may be eliminated. This reduces the ability of the resulting GDS structure to operate at high voltages; however, the distances between anode and cathode and between adjacent GDS structures can be increased to increase the useful voltage ranges.

Furthermore, the regions 56 and 56a can be replaced with protective rings, for example aoosvas-6 10 of the type shown around the cathode 2H, O00 in Tig. In addition, a region similar to region 220,000 and a guard ring similar to guard ring H00 of Fig. 6 may be provided in place of regions 56; 56a in Fig. 7; The electrodes can consist of doped polysilicon, gold, titanium or other types of conductors. The conductivity type for all semiconductor substrates and regions can be reversed provided that the voltage polarities are duly altered in a manner well known to those skilled in the art. In such a case, the regions 18, 18a, 180, 180a, 1800, 1800a, 18,000, 18,000a, SU, Süa, 180,000 and 180,000a become cathodes, while the regions ZU, 2üa, 2ü0, 2N0a, ZUOO, 2Ä00a, 2H, 00O, 2ü, 000a, 58, 58a, 2U0,000 and 2N0,000a become anodes. It should be noted that the structures according to the invention enable operation with both alternating current and direct current.

Claims (8)

n 8ÛÛ57k6-6 Patentkrav A semiconductor assembly comprising one or more semiconductor switching devices comprising a semiconductor body (16) of which a bulk portion is of a first conductivity type, a first region (18) of the first conductivity type, a second region (ZH) of a second conductivity type, opposite to the first conductivity type, and a control region (20) of the second conductivity type, the first region, the second region and the control region having a lower resistivity than the bulk portion and being separated apart from portions of the semiconductor body bulk portion (16), and the parameters of the device are such that when a first voltage is applied to the control region, a depletion region is formed in the semiconductor body, which region substantially prevents current from flowing between the first and the second region and that, when a second voltage is applied to the control region and when matched voltages are applied to the first and second regions, a relatively low resistive current path a is formed between the first and second regions by double-sided charge carrier injection, the device being characterized in that each of the first and second regions and the control region have a surface located on a first major surface of the semiconductor body ( 16) and that the control region is a semiconductor part (12) which is in contact with the semiconductor body along a second surface which is opposite to the first surface. An assembly according to claim 1, characterized in that the semiconductor body (16) comprises a defined third region (22) of the first conductivity type and having a resistivity lying between the bulk portion of the semiconductor body (16) and the first region (18), wherein the third region (22) is arranged to surround the second region (2U). An assembly according to claim 1, characterized in that the control region (H8 in Fig. 7) is located between the bulk portion (52) of the semiconductor body and a brioc substrate portion (UH) of the first conductivity type. ' U. An assembly according to claim 2, characterized in that the conductivities of the semiconductor body (16), the first region (18), the second region (25) and the third region (22) are p-, p +, n + and p, respectively. An assembly according to claim 1, characterized in that the control region is common to at least two switching devices with the first region (18) in one switching device connected to the second region (2Ua) in the second switching device and the second region (ZH) of the first switching device connected to the first region (18a) of the second switching device. A unit according to claim 3, characterized in that a plurality of switching devices (GDS3, GDSH) are arranged on tray substrate (HH) and that the tray substrate comprises a plurality of regions (50) of the first conductivity type. but with lower resistivity than the brick substrate. Aggregate according to claim 1, characterized in that the semiconductor body (1b, 000) contains a defined fourth region (HO) which surrounds but is not in contact with the second region, the fourth region being of the first conductivity type and having lower resistivity than the bulk portion (16,000). An assembly according to claim 2, characterized in that the semiconductor body comprises a fourth region (400) which is part of the third region (220,000) and which surrounds but is not in contact with the second region (240,000), the fourth region being of the first type of conductivity and has lower resistivity than the third region. , __._____ í ..-._-... ___... a.