EP1008182A1 - High voltage component and method for making same - Google Patents

Abstract

The invention concerns a high voltage component for switching high voltage currents and a method for making such a component. Said component comprises partial components (10) series-mounted and laterally arranged on a self-supporting semiconductor wafer (14), said partial elements, for example, operating a switch from a certain voltage applied between a first bridge cathode (22) and an anode metal layer (7). At least one partial component has a region extending from the semiconductor front face to its rear face.

Description

High-voltage component and method for its production

State of the art

The invention is based on a high-voltage component or a method for producing a high-voltage component according to the type of the independent claims. A high-voltage component is already known from DE 4108611, in which partial components arranged laterally on a carrier plate are connected in series. The partial components are integrated in a thin silicon layer applied to the carrier plate, which preferably has a thickness of approximately 100 micrometers.

Advantages of the invention

In contrast, the high-voltage component according to the invention with the characterizing features of the independent device claim has the advantage of a simple construction, since no carrier plate is necessary due to the use of a wafer that carries itself. This results in a compact, space-saving construction. Another advantage is that the use of an appropriately thick wafer results in a better current distribution, in particular a higher current carrying capacity of the high-voltage component, since a larger volume of semiconductor material can contribute to a briefly high current flow, as is particularly the case when using ignition systems is required for motor vehicles. The method according to the invention with the characteristic features of the The independent method claim has the advantage that it represents a simple method for producing a high-voltage component, the use of a carrier plate in particular being omitted and the component directly, for example without the laborious handling of an insulated carrier plate, after the etching and passivation. B. by Ummolden or by casting, can be packed.

The measures listed in the dependent claims enable advantageous developments and improvements of the high-voltage component or method for its production specified in the independent claims. It is particularly advantageous that the high-voltage component has, in addition to p-doped regions designed as separation diffusion regions, flat p-doped wells. This makes it possible to implement a base width of an NPN sub-transistor which is variable but nevertheless spatially homogeneous in wide areas, for example if such an NPN sub-transistor is part of a thyristor which is the sub-component. In particular, spatially homogeneous small base widths can be set, which lead to large current amplifications in a thyristor as a component, as are required for use to control an ignition current of an ignition coil in the motor vehicle. In addition, despite the use of a thick, self-supporting wafer, a large bottom resistance can be achieved, so that the thyristor is ignited in a reliable manner and when the anode currents are not too high. The process of introducing separation diffusions and also introducing flat p-doped wells allows a simple one

Setting exact base widths when using thick wafers. This can be explained by the fact that diffusion zones, which only protrude flat into the wafer, can be more precisely defined in their geometrical extent than, for example, separation diffusion areas. If the p-well is connected to the p-area designed as a separation diffusion area, this results in a more compact design in the lateral extent, and the manufacturing process is also simplified.

The provision of a resistance range which is in electrical connection with the p-well ensures a controlled temperature response of the electrical resistance, which is important for the ignition of the high-voltage component.

drawing

Embodiments of the invention are shown in the drawing and explained in more detail in the following description. 1 shows a first exemplary embodiment, FIG. 2 shows a second exemplary embodiment, FIG. 3 shows a third exemplary embodiment and FIG. 4 shows a fourth exemplary embodiment.

Description of the embodiments

FIG. 1 shows a high-voltage component according to the invention, which is arranged in a semiconductor wafer 14. The high-voltage component has a plurality of sub-components 10 connected in series. In the exemplary embodiment shown, the component 10 is a thyristor. The wafer 14 is weakly n-doped and has p-regions 2 designed as separation diffusion regions, which completely penetrate the semiconductor wafer 14, as shown in the cross-sectional view in FIG. 1. As a result, the semiconductor wafer 14 is divided into n-doped regions 1. The p-regions 2, like the n-doped regions 1, run essentially parallel to one another, that is to say they form perpendicular to the cross-sectional view shown in FIG extensive parallel stripes. A p-doped well 20 is introduced into each of the n-doped regions 1, which, in contrast to the p-regions 2, does not completely penetrate the semiconductor wafer. The p-wells 20 extend from a left boundary 16 to a right boundary 17, the boundaries extending perpendicular to the cross-sectional view extending essentially parallel to the n-doped regions 1 and the p-regions 2. A strongly n-doped strip 4 is embedded in each of the p-wells 20 and also runs perpendicular to the cross-sectional view parallel to the other areas. A first bridge cathode 22 or second bridge cathode 23, lined with insulation 21, connects the p-well 20 to the p-region 2 of the adjacent sub-component. At the same time, the strongly n-doped strip 4 introduced into the p-well is contacted by the respective bridge cathode. An edge p-region 15 can be contacted via an anode metallization 7. The remaining surface of the front of the high-voltage component, on which no contacts are arranged, is coated with an oxide layer 8 for insulation. The back of the semiconductor wafer 14 is also coated with a further oxide layer 9. The thickness 11 of the wafer from the front to the rear is in a range from, for example, 200 to 500 micrometers, preferably in a range from 200 to 250 micrometers. The high-voltage component provided with connecting wires on the first bridge cathode 22 and on the anode metallization 7 is finally surrounded by a protective material or plastic (connecting wires and packaging are not shown in FIG. 1).

If a voltage is applied in the forward direction between the first bridge cathode 22 and the anode metalization 7, only a reverse current initially flows as long as there is a potential difference between the heavily n-doped strip 4 and the p-well 20, which is less than 0.6 volts. This potential difference is influenced by the so-called bottom penetration of the p-well 20. This is the resistance of the p-well 20 below the heavily n-doped strip 4, which the reverse current must flow through in order to get from the right boundary 17 to the left boundary 16. If the applied voltage rises, the reverse current increases until a potential difference between the p-well 20 and the heavily n-doped strip 4 forms in the vicinity of the right limit 17 (for example 0.6 volts) that the high-voltage component breaks down comes ("overhead firing" of the thyristor). The bottom resistance is in turn influenced by the layer thickness 24 of the p-well 20 underneath the heavily n-doped strip 4. Small layer thickness 24 means large undermight resistance of the p-well 20. If the layer thickness 24 is selected So sufficiently small, a sufficiently large under-resistance can be set, so that not only at too high reverse currents

Overhead ignition occurs. If, on the other hand, no p-well 20 is provided, rather, as is known from the prior art, the heavily n-doped strip 4 is embedded in a region corresponding to the p-region 2, which extends from the front to the rear of the semiconductor wafer With a wafer thickness greater than 100 micrometers it can happen that the bottom resistance is so small that the overhead firing of the thyristor is not carried out correctly or only with a very high anode current. In the concept known from the prior art, it is therefore not possible to use thick, self-supporting semiconductor wafers. In the exemplary embodiment shown, which eliminates this disadvantage, the flat p-wells 20, calculated from the surface of the semiconductor wafer, have a thickness of approximately 20-40 micrometers. The current gain of the individual NPN Partial transistor, which is formed from the heavily n-doped strip 4, the p-well 20 and the n-doped region 1, and thus also the tilting behavior of the thyristor, in particular the speed of the high-voltage component, is strongly affected by the distance between the heavily n-doped strips 4 and the adjacent n-doped region 1 influenced. This distance is just the layer thickness 24 of the p-well 20 and represents the base width of the NPN partial transistor mentioned. For a high switching speed of the

High-voltage component, especially for a large ignition current immediately after the overhead firing of a thyristor, a large current gain of the NPN sub-transistor is necessary, which can be achieved over a small layer thickness 24 and thus a small base width of the NPN sub-transistor. Furthermore, the ratio of the dopant concentrations of the dopant concentration of the heavily n-doped strip 4 to the dopant concentration of the p-well 20 must be suitably high. The structure shown in FIG. 1 also makes it possible to set a homogeneous base width, that is to say the base width is essentially the same at different locations of the NPN junction of the NPN partial transistor. This means that when the thyristor is fired at the right boundary 17, the entire width of the p-well 20 is immediately from the left

Limit 16 to the right limit 17 is used for current amplification, that is, the NPN partial transistor works along its entire spatial extent. The location-independent base width of the NPN partial transistor thus leads to the fact that at a given required

Current carrying capacity, the component 10 designed as a thyristor can be made smaller in its lateral extent, in particular the width of the heavily n-doped strip 4 and the extent of the p-well 20 from the left boundary 16 to the right limit 17 reducible. In addition to the design of the component 10 as a thyristor, the structure according to the invention can also be used for a series connection of light-tipping diodes (ie thyristors that can be fired with a light pulse) or for triac components.

FIG. 2 shows a second exemplary embodiment of a high-voltage component, in which the same reference numerals as in FIG. 1 designate the same parts and are not described again. As an alternative to the embodiment according to FIG. 1, the high-voltage component has a p-well 3 which is connected to the p-region 2 of the adjacent sub-component via a p-doped region 12. Instead of bridge cathodes as in FIG. 1, a first is used as the cathode metallizations in the exemplary embodiment according to FIG

Cathode metallization 5, a second cathode metallization 6 and other metallizations, not shown in FIG. 2, are used, which are not relined by an insulation 21. This is not necessary here either because the cathode metallizations cannot come into contact with the n-doped regions 1 due to the p-doped regions 12. The mode of operation is essentially analogous to the prescribed exemplary embodiment. However, there is still a more compact design since there is no longer a weakly n-doped region which separates the p-wells from the p-regions 2.

FIG. 3 shows a high-voltage component which has partial components with a p-well 3, which is provided with a gate metallization 22a or 23a. This

Gate metallizations are electrically connected to the adjacent bridge cathode 22 and 23 via resistance areas 34 and 35, respectively. In the exemplary embodiment shown, the gate metallization 22a or 23a is designed in the form of a strip perpendicular to the cross-sectional view shown, in particular in parallel to the bridge cathodes, which are also strip-shaped. The gate metallizations contact the flat p-wells near their right boundary 17. The bridge cathodes differ from the bridge cathodes of FIG. 1 in that the PN junction between the heavily n-doped strip 4 and the p-well 3 is not short-circuited by the bridge cathodes in the vicinity of the p-doped region 12, that is to say that instead of the insulation 21, an alternative insulation 32 is provided, which is also partially arranged above the heavily n-doped strip 4. The insulation 32 serves to set a distance between the bottom resistances formed by the flat p-wells and the respectively associated bridge cathodes 22, 23 etc. The resistance areas 34, 35 etc. can be used to set the tilting conditions of the

High-voltage component can also be used. This is very advantageous if the tilting behavior of the high-voltage component should have a very small temperature dependency. The resistance areas 34 and 35 etc. are each connected in parallel with the bottom resistance of the respectively assigned p-well. Since the temperature dependence of the resistance range when using, for example, polysilicon is small in comparison to the temperature dependence of the under-resistance, a small resistance value for the

Resistance ranges the temperature dependency of the resistance of the parallel connection of the resistance range and the bottom resistance, which influences the tilting condition, is determined by the temperature coefficient of the resistance range and is therefore small. This makes the one for overhead lighting one of the

Subcomponents required voltage applied between the first bridge cathode 22 and anode metallization 7 essentially independent of the temperature of the high-voltage component. In a further, simpler embodiment variant, instead of the bridge cathodes as in 3, simple cathode metallizations, as explained in the description of FIG. 2, are used, that is, the alternative insulations 32 are omitted. In order to determine the temperature response of the breakover voltage from the temperature response of the resistance range in this case, the bottom penetration resistance should be chosen larger than in the embodiment shown in Figure 3.

FIG. 4 shows a high-voltage component with separate p-wells 20, similar to the exemplary embodiment shown in FIG. 1, but with resistance regions 34, 35, etc., as have already been explained in the description of FIG. 3. In contrast to the exemplary embodiment in FIG. 3, in the exemplary embodiment in FIG. 4, the bottom resistance and the assigned resistance range are connected in series. Therefore, the value of the resistance range must be chosen so that it is greater than the bottom resistance, so that the resistance range with its small temperature response determines the temperature response of the breakdown voltage. Optionally, the

Gate metalizations 22a and 23a etc. can also be arranged over the p-doped region 12 (this alternative arrangement of the gate metalization is not shown in FIG. 4, only the region 12 is marked here with its reference symbol). In this case they are

Provide bridge cathodes in their strip-shaped extension perpendicular to the cross-sectional view with one or more holes into which the gate metallization is arranged in isolation from the bridge cathode, either as a simply connected area or consisting of several sub-areas and separated by areas which are covered with the metallization of the bridge cathode. The resistance areas occurring in the exemplary embodiments mentioned are preferably made of polysilicon and either between assigned gate Metallization and bridge cathode of the neighboring component arranged, or they are placed in a suitable manner between the bridge cathode and the semiconductor wafer, each isolated from the respectively associated bridge cathode and the semiconductor wafer. The latter arrangement is a preferred embodiment for ensuring the high-voltage strength of the high-voltage component.

Due to the use of thick semiconductor wafers, a separation diffusion method is advantageously used in the method according to the invention for producing the p regions 2, for example. This separation diffusion method consists of applying p-type dopant atoms on both sides of the semiconductor wafer in areas which are provided for the p-type regions 2 and in a further step allowing these dopant atoms to diffuse in until the semiconductor wafer is similar to a zebra pattern consisting of alternating p -doped and weakly n-doped areas. In further steps, p-wells and heavily n-doped strips are introduced, with the p-wells 20 in particular if the p-wells 3 are to be connected to the p-regions 2 via p-doped regions 12, with p -Areas 2 can be produced in the simplest way.

Claims

Expectations 1. High-voltage component on an n-doped semiconductor wafer (14), with at least two partial components connected in series, characterized in that the wafer (14) having a front and a rear side is a self-supporting component and at least one partial component ( 10) has a p-region (2) which extends from the front to the back of the wafer. 2. Component according to claim 1, characterized in that at least one component has a p-doped trough (20 or 3). 3. Component according to claim 2, characterized in that a heavily n-doped strip (4) is embedded in the p-well. 4. The component according to claim 2 or 3, characterized in that the p-well (3) with the p-region (2) of the adjacent sub-component is connected via a p-doped region (12). 5. The component according to claim 2, 3 or 4, characterized in that a cathode metallization (5) or a bridge cathode (22) contacts the p-well (3 or 20) and the p-region (2) of the adjacent sub-component together . 6. The component according to claim 5, characterized in that the p-well has a gate metallization (22a) which is in electrical contact with the cathode metallization (5), for example, with the bridge cathode (22) via a resistance region (34). 7. The component according to claim 6, characterized in that the resistance region consists of polysilicon. 8. Component according to one of the preceding claims, characterized in that the component has an edge p-region which is provided with an anode metallization (7). 9. Component according to one of the preceding claims, characterized in that the back of the wafer is covered with a silicon dioxide layer (9). 10. The component according to one of the preceding claims, characterized in that the sub-component is a thyristor or a thyristor that can be ignited with a light pulse. 11. The component according to one of the preceding claims, characterized in that the component of one Protective material, especially a plastic covering, is surrounded. 12. Component according to one of the preceding claims, characterized by interchanged doping. 13. A method for producing a high-voltage component with sub-components connected in series, characterized in that parallel, strip-shaped p-regions (2) are introduced into the wafer in a weakly n-doped semiconductor wafer (14), which are introduced from the front Reach the back of the wafer and divide the wafer into parallel, weakly n-doped areas (1). 14. The method according to claim 13, characterized in that both sides of the semiconductor wafer are covered with p-doping atoms over the areas provided for the strip-shaped p-regions (2) and that in a further step the doping atoms diffuse in from both sides are formed until p regions (2) extending from the front to the rear of the semiconductor wafer have been formed. 15. The method according to claim 13 or 14, characterized in that in the n-doped regions (1) p-wells (20 or 3) are introduced.