DE10245089A1 - Doping process used in the production of a transistor, IGBT, thyristor or diode comprises preparing a semiconductor body, producing crystal defects in the body, introducing hydrogen ions into the body, and heat treating - Google Patents

Abstract

Doping process comprises preparing a semiconductor body (2) with a base dopant of first conductivity, producing crystal defects in the semiconductor body, introducing hydrogen ions from a first surface (3, 4) into the semiconductor body, and heat treating in which temperature and duration are selected so that hydrogen atoms are introduced over the whole crystal defect region of the semiconductor body. An Independent claim is also included for a semiconductor component produced by the above process.

Description

The invention relates to a doping process for the generation of gradually increasing, deep doping profiles in a semiconductor body and a semiconductor device according to the preamble of the claim 9. The invention further relates to a semiconductor component according to claim 17 and a method for its production according to claim 26.

Such a semiconductor device can be of any design, i.e. it can be one Transistor, an IGBT, a thyristor, a diode or the like act. The following is an example of a semiconductor device assumed a bipolar high-voltage component - for example an IGBT without, however, the invention on this semiconductor device to restrict.

With such high-voltage components the lightly doped inner zone is designed for reverse voltage take. In the event of passage, the inner zone is “flooded” with charge carriers and thus has a very low rail resistance, which is much lower than with a comparable unipolar semiconductor component. Indeed must be used for every switching operation, i.e. in the event of a transition from open mode in lock mode, this excess cargo again from the flooded Inner zone are removed, so that a space charge zone at all can arise and thus the reverse voltage on the semiconductor device can build up. The time until the space charge zone builds up has and the associated increase in reverse voltage, will by the total amount of the flood charge that over the reverse current flows, limited. On the one hand, the semiconductor device should switch application-related not too steep slope of the reverse voltage edge and have the current edge. On the other hand, it is necessary that the reverse flow is not suddenly stops to flow since this is too undesirable high overvoltages to lead would. These surges and the too high voltage or current steepness can in Extreme case for destruction lead of the semiconductor device. Such destruction to avoid the semiconductor component must be ensured that even at maximum reverse voltage, the space charge zone Cargo in the flooded The inner zone is not complete eliminates.

This is typically achieved by a suitably chosen topology of the semiconductor component, two of which are briefly explained below:
On the one hand, semiconductor components are known which have a so-called “non-punch-through” technology, in which the inner zone is so large that the space charge zone no longer reaches a highly doped zone adjacent to the inner zone at full reverse voltage. On the other hand, this is achieved by so-called "punch-through" semiconductor components, in which a highly doped layer of the same line type, the so-called buffer layer, is provided in the inner zone. The purpose of this buffer layer is to prevent the electric field from penetrating at full reverse voltage. Both approaches have in common that the low-doped inner zone must be made thicker than is required for the pure blocking properties due to the requirements for the switching behavior, as a result of which the transmission and switching losses are above the theoretically achievable, optimal value.

As a non-punch-through IGBT or as a punch-through IGBT Trained semiconductor components are, for example, in Jens Peer Stengl, Jenö Tihanyi, Power MOSFET practice, Pflaum-Verlag, Munich, 1992, described in particular on pages 105-106.

In the case of semiconductor components based on epitaxially manufactured base material was built Doping in the inner zone has also been in the past increasing depth in the semiconductor device gradually increased. This leads to a slowed down expansion of the space charge zone into the depth of the Semiconductor device in and reduced by the resulting Thickness of the component to an accompanying simultaneous Improvement in switching and transmission losses. In the near-surface Areas of the inner zone, i.e. directly under the blocking pn junction, the doping remains low, whereby the blocking capacity remains high.

For the manufacture of such an epitaxial layer will be successive Epitaxial layers applied to the semiconductor body with an embossed doping, where the doping concentration in the successively applied Epitaxial layers are gradually reduced. Such an epitaxial manufactured base material in which the doping concentration in the direction of the blocking pn junction is lower in stages, is relatively complex to manufacture. Especially for high-voltage components that have a reverse voltage of 300 - 600 volts or above must be designed the corresponding epitaxial layer is made very thick, whereby the semiconductor component is disproportionately expensive.

Often, however, it is not a step-wise increase, but rather a continuous increase or decrease in the doping. This could only be achieved approximately using epitaxy and would also be extremely complex in terms of production technology and thus expensive. For these reasons, a compromise is made in the case of such semiconductor components, in which the step height of the doping is chosen to be so great that the electrical properties are still acceptable and the costs for the production of the semiconductor components produced in this way are still within an acceptable range. Nevertheless, it would be desirable to have one Finding a method with which a semiconductor component with a continuously increasing doping concentration, in particular in the deep inner zone, could be provided. Furthermore, it would be desirable for this doping concentration to be largely locally limited to a desired semiconductor structure. The epitaxy for the production of such a doping should largely be avoided for the reasons mentioned.

For high-voltage semiconductor components Depending on the application, coordination is also required on the one hand with regard to smooth switching behavior and on the other is optimized with regard to minimal switching losses. A gentle one Shutdown behavior is typically determined by the dopant concentration in the at the blocking pn junction adjacent inner zone set. It is particularly advantageous if the dopant concentration at the blocking pn junction preferably is low and increases in the depth of the semiconductor device. In contrast, to reduce switching losses, it is necessary in the inner zone an area with a doping that is as possible the intrinsic carrier density at minimum thickness, as well as another area that ensures a hard field stop, provide. As a compromise between these two opposing ones So far, epitaxial layers have been used, their doping stepwise to the semiconductor surface decreases. The switching behavior is determined by the thickness and the level of the doping of the individual epitaxial layers in the inner zone. On this allowed the thickness of the inner zone and thus the switching losses of the semiconductor device with still acceptable switching properties can be significantly reduced.

Here, too, there is a need, the doping concentration preferably steadily to the surface of the semiconductor body to increase without using epitaxy.

The present invention lies therefore the task of a method and a semiconductor device Provide that on epitaxy to produce areas with increasing or decreasing doping concentration can be dispensed with can. Furthermore, the concentration of these doping areas should be as possible steady, i.e. increase continuously.

According to the invention, this object is achieved by a Method with the features of claim 1 and by a Semiconductor component with the features of claim 9 solved.

• (a) Bereitstellen eines Halbleiterkörper mit einer Grunddotierung (11) eines ersten Leitungstyps,
• (b) Erzeugen von Kristallschäden in dem Halbleiterkörper;
• (c) Einbringen von Wasserstoffionen ausgehend von einer ersten Oberfläche in den Halbleiterkörper;
• (d) anschließendes Durchführen eines Temperaturprozesses, wobei die Temperatur und die Dauer dieses Prozesses so gewählt sind, dass sich die eingebrachten Wasserstoffatome wenigstens annäherungsweise über den gesamten Kristallschäden aufweisenden Bereich des Halbleiterkörpers umverteilen.

The following is accordingly provided: A doping method for producing deep doping profiles that continuously increase in the vertical direction at least in sections in a semiconductor body, with the method steps:

• (a) providing a semiconductor body with a basic doping ( 11 ) a first line type,
• (b) generating crystal damage in the semiconductor body;
• (c) introducing hydrogen ions into the semiconductor body from a first surface;
• (d) subsequently performing a temperature process, the temperature and the duration of this process being selected such that the hydrogen atoms introduced are redistributed at least approximately over the entire region of the semiconductor body which has crystal damage.

A semiconductor device, in particular produced by a method according to any one of the preceding claims, which arranged in a semiconductor body having a first and a second surface is, with an inner zone, which is a basic doping of the first conductivity type has, with one on the first surface and on the inner zone adjacent first zone of the first conductivity type, its doping concentration is bigger than that of the inner zone, with a second adjacent to the inner zone Zone of the second conduction type, at least in the inner zone a third zone of the first conduction type is provided, the one higher Has doping concentration as the inner zone, the doping concentration the third zone in the direction of the first or the second surface continuously increasing The particular advantage of the method according to the invention consists in the production of components in the vertical direction, i.e. into the depth of the semiconductor body, gradually increasing or decreasing doping profiles at relatively low temperatures.

Doping is carried out using the inventive method only where there are defects due to the implantation or radiation have formed that subsequently be decorated with hydrogen. Here is the maximum doping in the area of the penetration depth of the hydrogen ions, because in this area the majority of the implanted ions are stopped. Towards the irradiated surface this doping gradually, i.e. continuously. The gradient this decrease and thus also the maximum of the doping can over the Temperature and the duration of the tempering can be controlled as a result the redistribution of the implanted hydrogen via diffusion processes and Complex formation is controlled. In addition, the exactly known range of the implanted ions the penetration depth control the doping effect as a relatively sharply delimited step.

The particular advantage here lies in particular in the fact that the interaction of the implanted hydrogen atoms with the defects in the transmission, which is necessary for the production of the donors generation of donors in the lower region of the semiconductor body, which is therefore not irradiated, is largely avoided in the region of the semiconductor body. This means that the hydrogen atoms themselves, of course, do not have a doping effect. They only get their doping effect in connection with the defects created by the radiation, for example vacancies. In the case of a masked implantation of hydrogen atoms, a spread of these donors produced in this way in the lateral direction of the semiconductor body is therefore largely avoided in the required temperature step, since the radiation defects can only scatter around a relatively small area around the bullet channel of the implanted or irradiated atoms. In this way, doping regions that are sharply delimited in depth and laterally can be generated.

By the method according to the invention can be irradiated depending on which side of the pane will produce continuously increasing or decreasing doping profiles. Such doping profiles have so far only been generated by a doping epitaxy. However, this had the disadvantage that it would at most result in a stepwise increasing doping profile was generated. The process according to the invention allows continuous generate increasing doping profiles that do not have such steps. The very complex epitaxy can advantageously be dispensed with here become.

The increase or decrease in the doping can additionally or alternatively also implemented in stages become. However, an increasing or decreasing doping profile can be obtained, for example, by multiple implantation different energies and / or doping doses a desired one Doping profile are generated. This in turn can be done without epitaxy produce.

The total amount of additional generated Doping is particularly about controlled the dose of the implanted hydrogen atoms. The depth of penetration, in which the doping maximum is located is determined by the Implantation or radiation energy set. By means of the described Technology can be used in the case of implanted hydrogen ions e.g. up to 500 μm generate deep n-doped layers.

For the doping is preferably introduced into the semiconductor body protons. Instead of using hydrogen ions to create the damaged zone, from which the n-doping is formed by tempering can additionally or alternatively Helium radiation or implantation can also be used namely for the case that the radiation-related defects with a subsequent to the radiation Hydrogen plasma treatment at temperatures between 250 ° C and 500 ° C for some Be decorated with hydrogen for hours. However, doing so will have the same reach a higher one Implantation or radiation energy required. Because helium ions clearly larger than Hydrogen ions are, however, to generate the implantation damage a significantly lower dose is required.

Since the hydrogen redistribution in generally made at relatively low temperatures in the range of 450-550 ° C becomes the carrier life in the semiconductor body, defects caused by the formation of recombination centers (e.g. double spaces) was lowered by this tempering raised again significantly. For example, at 500 ° C again almost the life of the parent carrier, that existed before the irradiation. At such low temperatures the structures of a semiconductor component, such as whose Gate oxide or its contact metallizations, not or almost not damaged. For these reasons can the inventive method relatively at the end of the entire manufacturing process, in which almost all semiconductor structures have already been manufactured.

If necessary, before doping e.g. mask made by lithography or a metal screen on the semiconductor body applied. The invention in the Semiconductor body thus introduced ions only reach the areas of the semiconductor body that located directly under the uncovered areas of this mask are. Since this means that only these areas of the semiconductor body pass through the implantation or radiation are damaged and a lateral Scattering negligible is low, only these areas damaged by the implantation have a manufactured according to the invention Endowment on. This allows it to be moved laterally and vertically generate sharply defined, previously defined doping regions.

The semiconductor body is before the doping Requires a thickness of less than 140 μm, especially less than 70 μm, thinly ground or thinly etched. The Thickness and the basic doping depend on the desired barrierability of the component and the application-related switching behavior selected.

The doping is advantageously carried out by means of proton implantation and especially in the case of an implantation with protons by means of a plasma treatment following the implantation.

The semiconductor component according to the invention can be designed, for example, as a pin power diode in which the manufactured according to the invention additional Doping is provided in the inner zone receiving the reverse voltage is. Here the doping concentration in the inner zone decreases blocking pn transition forth in the depth of the semiconductor body continuously towards one as possible to enable the diode to switch off gently. However, too Applications conceivable in which the doping of the blocking pn junction decreases with increasing depth of the drift zone.

Another advantageous The semiconductor component is designed as a power transistor trained in which the additional manufactured according to the invention Doping is arranged in the drift zone. The funding takes here from the blocking pn junction forth into the depth of the semiconductor body continuously.

The maximum locking ability of power semiconductors is known for the specific Resistance of the silicon base material, often also ohms of the silicon base material designated, limited. About the blocking ability of power semiconductor components can increase, for example Ohmkeit, that is the specific resistance of the semiconductor base material can be increased. For example, high-blocking power semiconductor components based on silicon base material with an ohmic value of up to 1000 Ωcm manufactured. Another increase however, the ohmic nature of the silicon base material causes in addition to The corresponding silicon wafers also become more expensive the physical and electrical properties of the corresponding silicon semiconductor body. Though can be silicon semiconductor body provide with an ohmicity of the base material of over 1000 Ωcm, but such are Silicon wafers, on the one hand, are very expensive to manufacture, and on the other hand the accuracy as well as the homogeneity of the target deteriorates resistivity over the entire width of the silicon wafer with increasing high resistance of the semiconductor body. In practice, the use of silicon base material with higher ohmic values especially not very useful because of the intrinsic temperature of the base material is already in the operating temperature range.

Another negative aspect of Use of high-resistance base material is a possible deterioration the dynamic properties of the corresponding semiconductor component. For example can by emitter injection, avalanche or by radiation caused events generates an increased density of free charge carriers in the semiconductor body become. As relative to this with the low basic funding small concentrations of free charge carriers can suffice electrical field strengths occurring during operation are undesirably modified considerably become. In extreme cases, this can lead to the permissible operating voltage must be significantly reduced. But this means, as it were, that the corresponding semiconductor component is oversized accordingly must be taken to counteract the effects mentioned in the operating case.

For the reasons mentioned to achieve a higher blocking capability of a power semiconductor component, the ohmic nature of the silicon base material not increased arbitrarily.

For these reasons, increasing the maximum blocking capability of a power semiconductor component today typically the thickness which increases the reverse voltage receiving drift path. In order to however, there is an increase in the on-resistance and therefore the switching-related power loss.

Another alternative to increasing the maximum blocking capability a power semiconductor component is the transition from basic material Silicon carbide (SiC). Today, high-blocking systems can already be Manufacture diodes from silicon carbide, but in particular exist in the manufacture of bipolar or controllable by field effect Silicon carbide base material transistors considerable technological Problems in addition to the high cost of the basic material of a marketable Realization of these semiconductor devices are currently opposed.

The present is based on this The invention is also based on the object of a semiconductor component and a method for producing such a semiconductor component to provide which is a higher blocking capability and has improved robustness in operation.

This object is achieved by a Semiconductor component with the features of claim 20 and a method for producing such a semiconductor device solved with the features of claim 29.

Accordingly, it is a semiconductor device provided which has a first and a second surface Semiconductor body is arranged with an inner receiving a reverse voltage Area which has a basic doping of the first conductivity type, with at least a first area of the second conduction type, the in the vertical direction in the inner region of the semiconductor component added , the first region being arranged in the inner region in this way and the inner area and the first area are each one have such a doping concentration that in each case Area located above and below the first area of the inner area with maximum reverse voltage present the maximum electric field strength associated with the basic doping of the areas is reached (Claim 17).

The idea on which this invention is based is to arrange at least one doped region of the opposite conductivity type in the vertical direction approximately in the middle of the voltage-receiving layer within the voltage-receiving layer. This additional layer of the opposite conduction type divides the voltage-absorbing layer to a certain extent into at least two subareas, in the case of a single additional layer into an upper and a lower subarea. The basic funding in these sub-areas is increased so that in the two additional ones inserted above and below Layers of the opposite line type with maximum reverse voltage present just the maximum electrical field strength associated with the corresponding basic doping is reached. The thickness of these two partial areas should be chosen as small as possible in order to keep the losses of the semiconductor components as low as possible. The advantage of such an arrangement according to the invention is that this method, for example when using a 1000 Ωcm base material instead of the usual reverse voltage of approximately 13 KV, can achieve a reverse voltage of 20 KV. Conversely, 13 KV components could also be realized with a much lower-resistance silicon base material, which advantageously significantly increases the robustness. If several layers are used, the basic doping can be increased further or the blocking capacity can be further increased. The additional layers can also be interrupted, ie present in island form.

Such a power semiconductor device can, for example, as a high-performance thyristor, IGBT, power MOSFET or power diode. However, other uses would be conceivable. In the case of a thyristor, the one according to the invention is additional Layer in the n-doped base zone, which is the voltage-absorbing layer forms, contain. In the case of an IGBT and power MOSFET the additional Layer in the drift zone designed as a stress-absorbing layer contain.

Further advantageous configurations and developments of the invention are the dependent claims as well the description with reference to the drawing.

The invention is described below of the embodiments shown in the figures of the drawing explained in more detail. It shows:

1 in a schematic partial section (a), a first embodiment of a semiconductor component according to the invention and its dopant concentration (b);

2 a schematic partial section of a second embodiment of a semiconductor device according to the invention;

3 in a schematic partial section (a), a third exemplary embodiment of a semiconductor component according to the invention and its dopant concentration (b);

4 that of the dopant concentration as a function of the dopant dose introduced into a semiconductor body in the case of a proton implantation according to Wondrak and own results;

5 in semi-logarithmic plots, some curves of the dopant concentration depending on the distance to the rear contact;

6 in a schematic partial section (a) a further embodiment of a semiconductor component according to the invention and the course of the electric field in the operating case (b);

7 based on partial sections (a) - (d) an inventive method for producing a semiconductor device according to the invention accordingly 6 .

8th the basic relationship between the dopant concentration of the hydrogen-induced donors and the temperature of the temperature treatment following the introduction of the hydrogen ions.

In all figures of the drawing are same or functionally identical elements - unless otherwise stated is with have been provided with the same reference numerals.

1 shows a schematic partial section (a) of a first exemplary embodiment of a semiconductor component according to the invention and the dopant concentration (b) of this semiconductor component.

This in 1 (a) with reference numerals 1 designated semiconductor device is designed as a pin power diode. The diode 1 has a semiconductor body 2 , for example made of silicon, with a first surface 3 and a second surface 4 on. On the first surface 3 is a cathode electrode connected to the cathode connection K. 5 applied. On the second surface 4 is an anode electrode connected to the anode terminal A. 6 applied. The semiconductor body 2 has an n-doped cathode zone 7 that to the first surface 3 adjacent, as well as a p-doped anode zone 8th that to the second surface 4 adjacent to. Katodenzone 7 and anode zone 8th are through an n-doped inner zone 9 spaced from each other. An interface between the inner zone 9 and anode zone 8th defines a pn junction 10.

The inner zone 9 has been produced by the method according to the invention and has a gradually, ie continuously increasing dopant concentration. This is based on 1 (b) , which shows the dopant concentration N _D , N _A in the semiconductor body as a function of the depth x.

• – Ein Halbleiterkörper 2 wird bereitgestellt, der bereits eine n+-dotierte Katodenzone 7 und auch eine p-dotierte Anodenzone 8, die beispielsweise mittels Diffusion, Ionenimplantation oder Epitaxie erzeugt wurden, aufweist. Es wäre auch denkbar, die Anodenzone 8 relativ am Ende des Herstellungsprozesses zu erzeugen.
• – Von der zweiten Oberfläche 4 her werden Wasserstoffionen in die Tiefe des Halbleiterkörpers 2 implantiert oder mittels Plasmabehandlung eindiffundiert, wobei vor der Diffusion ausgehend von der zweiten Oberfläche eine defekterzeugende Implantation, beispielsweise von Heliumatomen, erfolgt. Für diese Tiefenimplantation von Wasserstoffionen wird typischerweise eine Hochenergieimplantation durchgeführt. Dadurch lassen sich beispielsweise bei einer Energie von 2 MeV Eindringtiefen von etwa 50 μm erreichen. Durch die Implantation bzw. Bestrahlung wird der Halbleiterkörper 2 geschädigt, d.h. es werden Defekte im Halbleiterkörper 2 erzeugt. Für den Fall, dass man ein Dotierungsprofil gemäß Kurve 12 erzeugen will, muss die Bestrahlungsenergie so hoch gewählt werden, dass die Eindringtiefe der Protonen mindestens der Summe der Dicken der beiden Zonen 8, 9 entspricht. Im Falle einer Heliumbestrahlung von der zweiten Oberfläche 4 her werden Wasserstoffionen von der ersten Oberfläche 3 her und/oder der zweiten Oberfläche 4 her mittels Plasmabehandlung eindiffundiert.
• – Im Anschluss an die Implantation bzw. Plasmabehandlung wird der Halbleiterkörper einer Temperaturbehandlung bei z.B. 500°C (Temperung), typischerweise einige Stunden lang, unterworfen. Dabei diffundieren die implantierten Wasserstoffionen und lagern sich an die Defekte an und es werden Komplexe gebildet. Diese Komplexe wirken dotierend. Es handelt sich dabei um Donatoren, also um n-Dotierungen. Zur durchstrahlten zweiten Oberfläche 4 hin nimmt diese zusätzliche n-Dotierung 12 graduell ab. Der Mechanismus der Erzeugung einer n-Dotierung durch Wasserstoffbestrahlung ist beispielsweise beschrieben in: Wolfgang Wondrak: "Erzeugung von Strahlenschäden in Silizium durch hochenergetische Elektronen und Protonen", Inaugural-Dissertation, Johann-Wolfgang-Goethe-Universität, Frankfurt, oder in Kozlov, Kozlovski: "Doping of Semiconductors Using Radiation Defects Produced by Irradiation with Protons and Alpha Particles", Semiconductors, Vol. 35, No. 7, Seiten 735-761 und Seiten 769-795. 4 zeigt die Dotierstoffkonzentration in Abhängigkeit von der eingebrachten Dotierstoffdosis im Falle einer Protonenimplantation nach Wondrak. Nach dieser Kurve nimmt die maximal erreichbare Dotierungskonzentration mit zunehmender Dotierstoffdosis ebenfalls zu.
• – Im Anschluss daran oder alternativ auch bereits in einem früheren Verfahrensschritt lassen sich die Kathodenelektrode 5 und die Anodenelektrode 6 beispielsweise durch Abscheidung auf die jeweiligen Oberflächen 3, 4 erzeugen.

The semiconductor body 2 originally has a relatively low basic n-count 11 (dashed line). This basic doping is in the cathode zone 7 and the anode zone 8th due to the much larger n- or p-doping in these zones 7 . 8th superimposed. In the inner zone 9 this n-basic doping is gradually changed to a pn junction by an inventive method 10 declining n-doping 12 (dotted line) overlaid. The following is a method according to the invention for producing such a semiconductor component with one in the inner zone 9 continuously increasing doping explained:

• - A semiconductor body 2 is provided, the be an n + -doped cathode zone 7 and also a p-doped anode zone 8th , which were generated for example by means of diffusion, ion implantation or epitaxy. It would also be conceivable the anode zone 8th relatively at the end of the manufacturing process.
• - From the second surface 4 forth are hydrogen ions in the depth of the semiconductor body 2 implanted or diffused in by means of plasma treatment, a defect-producing implantation, for example of helium atoms, starting from the second surface before the diffusion. A high-energy implantation is typically carried out for this deep implantation of hydrogen ions. This enables penetration depths of around 50 μm to be achieved with an energy of 2 MeV, for example. The semiconductor body becomes by the implantation or irradiation 2 damaged, ie there are defects in the semiconductor body 2 generated. In the event that you have a doping profile according to the curve 12 to generate, the radiation energy must be chosen so high that the penetration depth of the protons is at least the sum of the thicknesses of the two zones 8th . 9 equivalent. In the case of helium radiation from the second surface 4 forth are hydrogen ions from the first surface 3 forth and / or the second surface 4 diffused here by means of plasma treatment.
• - Following the implantation or plasma treatment, the semiconductor body is subjected to a temperature treatment at, for example, 500 ° C. (annealing), typically for a few hours. The implanted hydrogen ions diffuse and attach to the defects and complexes are formed. These complexes have a doping effect. These are donors, i.e. n-dopants. To the irradiated second surface 4 this additional n-doping 12 gradually decreases. The mechanism of generating n-doping by means of hydrogen radiation is described, for example, in: Wolfgang Wondrak: "Generation of radiation damage in silicon by high-energy electrons and protons", inaugural dissertation, Johann Wolfgang Goethe University, Frankfurt, or in Kozlov, Kozlovski : "Doping of Semiconductors Using Radiation Defects Produced by Irradiation with Protons and Alpha Particles", Semiconductors, Vol. 35, No. 7, pages 735-761 and pages 769-795. 4 shows the dopant concentration as a function of the introduced dopant dose in the case of a proton implantation according to Wondrak. According to this curve, the maximum achievable doping concentration also increases with an increasing dopant dose.
• - Following this or alternatively in an earlier process step, the cathode electrode can be removed 5 and the anode electrode 6 for example by deposition on the respective surfaces 3 . 4 produce.

Instead of a continuously increasing or decreasing doping 12 corresponding 1 (b) can also be stepped in the direction of the zone by the method according to the invention 7 increasing doping 13 corresponding 2 produce. The profile in 2 shows several levels 13 each of one level 14 are separated. Within an area 13 takes the doping concentration towards the zone 7 as in 1 (b) continuously. The radiation or implantation dose and the corresponding energies are selected so that the desired step profile is generated. It can be used to generate step-like rising, step-falling and any other step-like profiles.

3 shows a schematic partial section (a) of a third exemplary embodiment of a semiconductor component according to the invention and its dopant concentration (b). That with reference numbers 20 designated semiconductor device is designed as a DMOS-FET.

The semiconductor body 2 corresponding 3 (a) points one to the second surface 4 adjacent, heavily n-doped drain zone 21 on. The drain zone 21 is over a large area on the second surface 4 applied drain metallization 22 connected to the drain terminal D. On the opposite surface 3 is one (or more) p-doped body zones 23 in the semiconductor body 2 embedded. In one body zone each 23 is one (or more) heavily n-doped source zone 24 embedded. The body zones 23 and source zones 24 can in a known manner by ion implantation and / or diffusion into the semiconductor body 2 be introduced. Furthermore, there is a gate electrode connected to the gate terminal G. 25 intended. The gate electrode 25 is about an in 2 gate oxide, not shown 26 from the semiconductor body 2 spaced and arranged so that in an area of the body zone 23 when a positive potential is applied to the gate connection G, a current-carrying channel caused by charge inversion can form. There is also a source metallization 27 , which is connected to the source connection S, provided the body zone 23 and source zone 24 electrically contacted via a shunt.

The body zone 23 is from the drain zone 21 due to a weakly n-doped inner zone 28 , which forms the drift zone and which serves to receive a reverse voltage, spaced apart. The interface between the body zone 23 and inner zone 28 defines a blocking pn transition 29 , The inner zone 28 first shows a basic funding 11 on (dashed line in 3 (b) ) by additional doping 12 (dotted line in 3 (b) ) is superimposed. This additional endowment 12 the inner zone 28 is again produced by a doping method according to the invention, which essentially corresponds to the above 1 described method corresponds.

In contrast to the one in 1 illustrated embodiment, however, the doping concentration increases continuously depending on the depth. For this purpose, the hydrogen atoms or helium atoms are on the first surface 3 in the semiconductor body 2 implanted, in the latter case the hydrogen atoms by means of plasma treatment from the surface 4 and / or the surface 3 be brought into this. The doping course of the additional doping can be determined by a suitable choice of the implantation parameters, ie the dose of the introduced atoms as well as the implantation energy and the annealing temperature 12 be set appropriately. The desired gradient of the doping concentration in the inner zone is therefore thereby 28 adjustable. The inner zone 28 thus points in the area of the pn junction 29 their lowest doping concentration, which rises continuously in the vertical component direction, so that the doping concentration in the region of the inner zone 28 to the drain zone 21 adjacent, is highest. As a result, the two requirements mentioned at the outset, namely the lowest possible doping in the region of the pn junction 29 to ensure a high reverse voltage and the high doping concentration required to reduce switching losses due to a low on-resistance Ron.

Of course, it would also be conceivable that the additional doping 12 in the other direction, i.e. to the second surface 4 towards, decreases (not shown in the figures) when, for example, from the surface 4 is irradiated here.

Beyond that the present invention is also very advantageous for manufacturing a so-called field stop area in a semiconductor device use. A field stop is an area in which a deep diffusion area is provided, in which a low dopant concentration with a higher dopant concentration nearer the surface is combined. The course of the dopant concentration in one deeply diffused area can be according to the available Technology can be designed in any way.

For the design of semiconductor components, in particular of higher blocking bipolar power semiconductors, such as ICGBTs, thyristors and diodes, the power loss that occurs during operation must be reduced on the one hand and a smooth switching behavior must be ensured on the other hand. The dimensions of the less doped layer can be used to reduce the losses 2 between the zones 7 and 8th , which is flooded by forward carriers, can be reduced. For a smooth switching behavior, it is necessary for a (further) part of this flood charge to be cleared out for a (further) rise in the reverse voltage on the component until the maximum reverse voltage is present on the component. As a result, the steepness of the voltage rise is limited and an undesired cut-off of the backflow is excluded.

A smooth switching behavior with a low component thickness can be achieved by the n-doping in the drift zone 2 gradually increases towards the back of the component, which means that increasing the width of the space charge zone until the breakdown voltage is reached is increasingly braked but not completely stopped.

5 shows some basic courses of the dopant concentration ND (logarithmically plotted) of such a field stop area depending on the distance x to the rear contact. The rear contact is here at x = 0 μm. The doping regions have been produced here, for example, by deep hydrogen implantation and / or by hydrogen plasma diffusion after a defect-producing implantation, for example of helium atoms. The doping dose to be selected for the deeply diffused region is particularly important. The integral dopant concentration in this area should exceed the breakdown charge, which is approximately QBR = 1,310 ¹² cm ^-2 to QBR = 1.8 10 ¹² cm ^-2 . This area on the back lies accordingly 5 approximately atx = –5 μm, i.e. approx. 5 μm in front of the rear of the pane.

A suitable choice of the dopant dose introduced into the semiconductor body ensures that the electric field extends relatively far into the field stop area. This results in a gentle field stop. In the rear area of this field stop zone, which typically extends about 5 μm from the rear of the wafer into the semiconductor body, an additional dopant dose is kept available. With the steeper increase in the dopant concentration it is achieved that the electric field is stopped reliably even in the event of full reverse voltage and also in the presence of smaller defects in the area of the gentle field stop immediately in front of the rear side emitter. The defects in the area of the gentle field stop are expressed in a locally lower than in 5 dopant concentration shown in the area of the gentle field stop, which leads to the fact that the integral of the charge in this area does not exceed the breakthrough charge and thus leads to reduced barrier properties. The hard field stop serves to improve the blocking yield of the semiconductor body, since smaller defects can never be completely avoided, particularly during implantation. Especially with larger semiconductor components, the prey can be significantly affected in this way. However, the dopant concentration and the dopant dose should be chosen to be sufficiently low to ensure the efficiency of the back not to reduce the emitter of the semiconductor component.

In contrast, in the German patent application DE 100 53 445 A1 proposed a two-stage dopant concentration in which the higher doping is to be carried out very flat, but relatively highly doped. The goal in the DE 100 53 445 A1 consists in reducing the efficiency of the rear emitter. The resulting peak in the dopant concentration, however, does not fulfill the function of improving the blocking yield, or does so less satisfactorily, since even small defects in the semiconductor body can lead to penetration of the electrical field and thus to an increase in the reverse current. However, these defects are insignificant for forward operation, since their proportion is negligibly small in relation to the entire area of the semiconductor component.

In the above embodiments described vertically formed semiconductor components. However, the invention is not limited exclusively to vertical semiconductor components, but rather could with the corresponding adjustment of the structures also to the lateral Semiconductor components and so-called up-drain semiconductor components apply.

The invention is not limited to the exemplary embodiments according to the 1 - 3 limited. Instead, a large number of new component variants can be specified, for example, by exchanging the conductivity types n for p and vice versa and by varying the doping concentrations. To illustrate the concept of the invention, the invention was only exemplarily illustrated using a pin power diode or a D-MOSFET. Rather, the invention could of course also be extended to any other semiconductor components, for example to any MOSFETs, J-FETs, thyristors, IGBTs, any diode structures, compensation semiconductor components and the like.

In summary it can be stated be that by introducing hydrogen into the depth of a Semiconductor body in complete Moving away from known semiconductor devices doping complexes can be produced at very low temperatures, which are exclusively in the areas of the semiconductor body are arranged in which defects were generated. The hydrogen can for example by proton implantation or after generation defects, for example by means of helium radiation, by a Hydrogen plasma treatment in the depth of the semiconductor body be introduced.

6 shows in a schematic sub-step (a) an embodiment of another semiconductor device according to the invention and the corresponding course of the electric field E (b) in the operating case of this semiconductor device as a function of the depth x and with positive and negative voltage U _AK , -U _AK between anode and cathode.

This in 6 (a) with reference numerals 100 designated semiconductor device is designed as a high-voltage thyristor. The thyristor 100 has a semiconductor body 101 , for example made of silicon, with a first surface 102 and a second surface 103 on. To the first surface 102 borders a p-doped cathode-side base zone 104 on. In the base zone 104 is also on the first surface 102 adjacent, strongly n-doped cathode-side emitter zone 105 embedded. The emitter zone 105 is with the cathode connection K, the base zone 104 connected to a gate terminal G. A strongly p-doped anode-side emitter zone borders on the anode side 106 to the second surface 103 , The emitter zone 106 is connected to the anode connection A. Between the base zone 104 and the anode-side emitter zone 106 is the stress-absorbing layer 107 arranged. The stress-absorbing layer 107 forms the anode-side base zone and has a weakly n-doped basic doping.

The thyristor 100 in 6 (a) has an additional layer according to the invention 108 on. The extra layer 108 is in the vertical direction of the thyristor 100 within the stress-absorbing layer 107 arranged and has a p-doping, ie a doping of the opposite conductivity type to the basic doping. The extra layer 108 divides the voltage-absorbing layer 107 in two areas 109 . 110 ,

The basic funding in the sub-areas 109 . 110 is chosen so that above and below the additional layer 108 with the maximum reverse voltage applied, the maximum electric field strength associated with the corresponding basic doping is just reached ( 6 (b) ), the thickness of these two sections 109 . 110 is chosen as low as possible in order to keep the losses of the semiconductor component as low as possible.

The semiconductor component according to the invention has so to some extent an approximately symmetrical shape, with the one in the voltage-absorbing Layer contains additional layer so to speak forms the axis of symmetry or surface of symmetry. Such in leaves about symmetrical structure for example, by using wafer bonding. Such a wafer bonding process is described, for example, in Q.-Y. Tong, U. Gösele, Semiconductor Wafer Bonding, A Wiley-Interscience Publication, John Wiley & Sons, 1999.

7 shows, using partial sections (a) - (d), a method according to the invention for producing a semiconductor component according to the invention accordingly 6 ,

When using this wafer bonding process, two silicon wafers are initially used 113 . 114 provided ( 7 (a) ).

On the first silicon wafer 113 becomes the part on the cathode side 104 . 105 of the thyristor 100 made in the usual way. In the same way, the anode-side part is in the usual way on the second silicon wafer 114 manufactured ( 7 (b) ).

Before bonding by means of wafer bonding, one or both of the surfaces on the anode or cathode side are placed 102 . 103 opposite surfaces 111 . 112 the silicon wafers implants ions of the opposite conductivity type as the basic doping ( 7 (c) ). The typical dose used to create the additional layer is in the range of a few 10 ¹² cm ^-2 .

The implantation layer thus formed forms the corresponding additional zone after the subsequent connection process ( 7 (d) ). For this purpose, the two silicon wafers are bonded using wafer bonding. To produce a high-temperature stable connection between the two silicon wafers, the two are the anode or cathode-side surfaces 102 . 103 opposite surfaces 111 . 112 placed on top of one another and the entire arrangement is subjected to a high-temperature step in order to produce a stable connection. This high temperature step also ensures that the ions just implanted for the additional layer 108 be electrically activated and the semiconductor body 101 is healed. Finally, the in 7 Electrodes for anode and cathode, not shown, can be produced in a known manner.

The basic doping of the semiconductor body is typically 101 from weakly n-doped base material. To create the additional layer 108 one or both of the silicon wafers is typically implanted with boron ions. In a very advantageous embodiment, a boron implantation can be used to produce the additional layer 108 can also be dispensed with if the high-temperature step to produce a stable connection is carried out at a temperature in the range from approximately 800-900 ° C. At these temperatures, a sufficiently high p-doped layer automatically results 108 between the wafers during wafer bonding. The reason for this is that there is generally enough boron in the laboratory air to produce such an additional p-type doping 108.

The invention is not intended to apply solely to a single additional layer 108 within the stress-absorbing layer 107 be limited, rather it would also be conceivable to have two or more additional layers 108 of the opposite conduction type within the voltage receiving layer 107 to provide in the 6 and 7 not shown, which are arranged alternately in the vertical direction next to each other and by sub-areas 109 . 110 . 111 . 112 ... are objected to. Such a structure is for example in the European patent EP 344 514 B1 described. The subject of EP 344 514 B1 is herewith regarding the structure of the semiconductor components described there, in particular with regard to the structure of a voltage-absorbing layer 107 in which a variety of additional layers 108 are arranged, fully included in the present patent application.

However, it is important to ensure here that the dimensioning of the basic doping of the voltage-absorbing layer 107 and the distance between the individual additional layers 107 is selected from one another such that the maximum possible critical electrical field strength is achieved in each case. This could, for example, improve the dynamic robustness of power diodes and the short-circuit behavior of IGBTs.

However, the semiconductor components just mentioned, ie a large number of additional layers, can be used 108 have, by means of the wafer bonding method described above, or only with considerable effort. However, these structures could be produced, for example, by epitaxy or a combination of wafer bonding and epitaxy. However, the maximum barrier capability is severely limited by the maximum thickness of such a base layer that can be achieved by means of epitaxy.

An advantageous alternative is obtained if instead of using an n-doped base material for the voltage-absorbing layer 107 a p-doped base material is used and thus the doping conditions in 7 be reversed. The p-doped base material thus forms the layers 109 . 110 within the stress-absorbing layer 107 , The n-doping of the partial area 108 is generated according to the invention by means of hydrogen implantation. To do this, high-energy hydrogen ions are introduced into the semiconductor body 101 implanted that the semiconductor body 101 severely damage in the area of the penetration depth of the implanted ions, so that defects are generated. In a subsequent temperature process, complexes are formed that have an n-doping effect. The temperature step should be selected in such a way that the implanted hydrogen cannot shift significantly and thus a relatively narrow n-doped area 108 results.

The advantage of using the Hydrogen implantation is mainly that relatively easy n-doped layers down to depths of more than 500 μm with a suitable one Choice of implantation energy can be generated. The depth of the so generated n-doped layers would over the Choice of the implantation energy given and the concentration of the respective n-doping the implantation dose can be set.

Instead of realizing one or more additional layers 108 By masking these structures, several stored Do lattice or island structures can be realized.

Finally shows 8th the basic relationship between the doping concentration of the "hydrogen-induced donors" generated by the protein radiation and subsequent temperature treatment in a semiconductor body. In the example, the radiation energy of the protons is chosen such that the so-called "end-of-range" range is approximately 30 μm. Irradiation defects occur in the area between the irradiated surface, which is denoted by the zero point of the scale, and the end-of-range area. The hydrogen ions introduced into the end-of-range area diffuse during the subsequent temperature process in the direction of the irradiated surface, the extent of this diffusion and thus the doping concentration in this irradiated area being dependent on the temperature during the temperature process and its duration. The temperature should be in a range between about 300 ° C and 550 ° C to allow the formation of hydrogen-induced donors.

8th shows the doping curve for three different temperatures T1, T2, T3, which are between 400 ° C and 500 ° C, where T1 <T2 <T3 applies. It can be seen from this that the doping concentration, in particular in the end of range, decreases with increasing temperature within the range from 400 ° C. to 500 ° C., the duration of the temperature treatment being the same in all cases.

Depending on the duration of the subsequent heat treatment and the chosen temperature stock up in the end-of-ange The present invention based on the description above so the principle the invention and its practical application as best as possible to explain, however leaves the present invention is of course within the scope of the expert Modify action and knowledge in a suitable manner.

1

pin power diode

2

Semiconductor body

3

first surface

4

second surface

5

cathode

6

anode electrode

7

Katodenzone

8th

anode zone

9

inner zone

10

pn junction

11

basic doping

12

additional doping

13

graduated additional endowments

14

step

20

D-MOSFET

21

drain region

22

drain metallization, drain

23

Body zone

24

source zone

25

gate electrode

26

gate oxide

27

source zone

28

Inner zone, drift region

29

pn junction

100

thyristor

101

Semiconductor body

102

first surface

103

second surface

104

cathode side base zone

105

cathode side emitter region

106

anode side emitter region

107

stress-bearing Layer, anode-side Ba

siszone

108

additional layer

109 110

subregions the stress-absorbing layer

111, 112

surfaces

113 114

Silicon wafers, wafer

S

source terminal

D

drain

G

gate terminal

K

cathode

A

anode

Claims (29)

Doping method for generating deep doping profiles that continuously increase in the vertical direction at least in sections ( 12 . 13 ) in a semiconductor body ( 2 ) with the method steps: (a) providing a semiconductor body ( 2 ) with a basic funding ( 11 ) a first line type ;; (b) generating crystal damage in the semiconductor body ( 2 ); (c) introduction of hydrogen ions from a first surface ( 3 . 4 ) in the semiconductor body ( 2 ); (d) subsequently performing a temperature process, the temperature and the duration of this process being selected such that the hydrogen atoms introduced are at least approximately over the entire area of the semiconductor body which has crystal damage ( 2 ) redistribute. A method according to claim 1, characterized in that the temperature process at a temperature less than 550 ° C, in particular of less than 450 ° C, he follows. The method of claim 1 or 2, wherein the generation the crystal damage and the introduction of the hydrogen atoms during a process step takes place in which hydrogen ions are implanted in the semiconductor body become. The method of claim 1 or 2, wherein the crystal damage by an ion implantation can be generated and the hydrogen ions then diffused in become. The method of claim 4, wherein the hydrogen ions by a plasma treatment in the semiconductor body ( 2 ) are introduced. A method according to claim 4 or 5, in which for generating the crystal damage Helium ions are implanted. Method according to one of Claims 3 to 6, in which, in order to generate the crystal damage, the ions into the semiconductor body at energies of more than 1 MeV, in particular more than 2 MeV ( 2 ) are introduced. Method according to one of the preceding claims, characterized characterized that for the doping is a multiple implantation at different energies and / or different doping doses is carried out. Method according to one of the preceding claims, characterized in that a mask for structuring the doping regions onto the first surface ( 3 . 4 ) is applied, ions over the first surface ( 3 . 4 ) in the semiconductor body ( 1 ) are implanted and then the mask again from the first surface ( 3 . 4 ) is replaced. Method according to one of the preceding claims, characterized in that the semiconductor body ( 2 ) is ground and / or etched before the additional doping to a thickness of less than 140 μm, in particular less than 70 μm. Semiconductor device ( 1 . 20 ), in particular produced with a method according to one of the preceding claims, which comprises a first and a second surface ( 3 . 4 ) having semiconductor body ( 2 ) is arranged with an inner zone ( 9 . 28 ), which is a basic funding ( 11 ) of the first conduction type, with a to the first surface ( 3 . 4 ) and the inner zone ( 9 . 28 ) adjacent first zone ( 7 . 21 ) of the first conductivity type whose doping concentration is greater than that of the inner zone ( 9 . 28 ), with one to the inner zone ( 9 . 28 ) adjacent second zone ( 8th . 23 ) of the second line type, characterized in that in the inner zone ( 9 . 28 ) at least a third zone ( 12 . 13 ) is provided which has a higher doping concentration than the inner zone ( 9 . 28 ), the doping concentration of the third zone ( 12 . 13 ) in the direction of the first or the second surface ( 3 . 4 ) continuously increases at least in sections. Semiconductor component according to claim 10, characterized in that the semiconductor component as a diode ( 1 ) is formed and the second zone ( 8th ) to the second surface ( 4 ) adjoins. Semiconductor component according to one of claims 10 or 11, characterized in that the doping of the third zone ( 12 . 13 ) is an n-doping, which results from the basic doping ( 11 ) and an additional funding ( 12 ), which is formed by complexes of radiation defects and hydrogen atoms. Semiconductor component according to one of claims 11 to 13, characterized in that the radiation defects by a Hydrogen implantation creates and the hydrogen atoms through this hydrogen implantation are introduced. Semiconductor component according to one of claims 11 to 13, characterized in that the radiation defects by a Ion implantation, preferably a helium implantation and the hydrogen atoms are introduced by a plasma treatment are. Semiconductor component according to one of claims 11 to 15, characterized in that in the inner zone ( 9 . 28 ) a variety of third zones ( 13 ) are provided, with a transition ( 14 ) from neighboring third zones ( 13 ) one level each ( 14 ) and where the doping concentration of these third zones ( 13 ) alternating continuously and stepwise in the direction of a surface ( 3 . 4 ) decreases or increases. Semiconductor component according to one of claims 11 to 16, characterized in that the semiconductor device as a power transistor ( 20 ) is formed, the doping concentration of the third zone ( 12 ) in the direction of the second surface ( 4 ) increases. Semiconductor component according to one of claims 11 to 16, characterized in that the semiconductor component as a line transistor ( 20 ) is formed, the doping concentration of the third zone ( 12 ) in the direction of the first surface ( 3 ) increases. Semiconductor component according to one of Claims 11 to 18, characterized in that the semiconductor body ( 1 ) has a thickness of less than 140 μm, in particular less than 70 μm. Semiconductor device ( 100 ) which in a first and a second surface ( 102 . 103 ) having semiconductor body ( 101 ) is arranged, with an inner region receiving a reverse voltage ( 107 ), which has a basic doping of the first conductivity type, with at least one first area ( 108 ) of the second conduction type, which goes vertically into the inner area ( 107 ) of the semiconductor component is inserted, characterized in that the first region ( 108 ) in the inner area ( 107 ) is arranged in this way and the inner area ( 107 ) as well as the first area ( 108 ) each have such a doping concentration that in each case above and below the first region ( 108 ) arranged area ( 109 . 110 ) of the inner area ( 107 ) with maximum applied reverse voltage (UAK) just that for the basic doping of the areas ( 109 . 110 ) appropriate maximum electric field strength is reached. Semiconductor component according to Claim 20, characterized in that a single first region ( 108 ) is provided which covers the inner area ( 107 ) in two areas ( 109 . 110 ) divides by the first area ( 108 ) are spaced from each other. Semiconductor component according to one of Claims 20 to 21, characterized in that the first region ( 108 ) approximately in the middle of the inner area ( 107 ) is arranged. Semiconductor component according to one of claims 20 to 22, characterized in that the doping of the first region (108 ) is a p-doping and that the doping of the areas ( 109 . 110 ) is an n-doping. Semiconductor component according to one of Claims 20 to 23, characterized in that the doping concentration in the inner region ( 107 ) is less than the doping concentration in the first region ( 108 ). Semiconductor component according to one of claims 20 to 24, characterized in that a to the first surface ( 102 ) as well as the inner area ( 107 ) adjacent second area ( 104 ) of the second line type, in which a third area ( 105 ) of the first line type is embedded, the second area ( 104 ) to a gate connection (G) and the third area ( 105 ) are connected to a cathode connection (K) and that one to the second surface ( 103 ) as well as the inner area ( 107 ) adjacent third area ( 106 ) of the second conductivity type, which is connected to an anode connection (A), is provided. Semiconductor component according to one of claims 20 to 25, characterized in that the semiconductor component as a thyristor ( 100 ) is trained. Semiconductor component according to one of claims 20 to 25, characterized in that the semiconductor component is designed as a power MOSFET is. Semiconductor component according to one of claims 20 to 25, characterized in that the semiconductor device as a power diode is trained. Method for manufacturing a semiconductor device ( 100 ) according to one of claims 20 to 28 by means of wafer bonding, characterized by the following method steps: - Two semiconductor bodies ( 113 . 114 ) with a basic funding of the first line type are provided; - Before bonding using wafer bonding, at least one surface ( 111 . 112 ) of a semiconductor body ( 113 . 114 ) Implants, diffused or applied dopants of the second conductivity type; - The two semiconductor bodies ( 113 . 114 ) with their surface ( 111 . 112 ), on or into which the dopants were introduced, placed one on top of the other; - To create a high-temperature stable connection between the two semiconductor bodies ( 113 . 114 ) they are subjected to a high temperature step.