Active solid-state devices (e.g. – transistors – solid-state diode – Integrated circuit structure with electrically isolated... – With pn junction isolation
Reexamination Certificate
2002-03-07
2003-01-07
Chaudhuri, Olik (Department: 2813)
Active solid-state devices (e.g., transistors, solid-state diode
Integrated circuit structure with electrically isolated...
With pn junction isolation
C257S549000, C257S550000
Reexamination Certificate
active
06504230
ABSTRACT:
BACKGROUND OF THE INVENTION
Field of the Invention
The present invention relates to a compensation component and to a method for fabricating it.
As is known, the on-state losses in metal-oxide semiconductor (MOS) transistors are composed of losses in the channel between a source zone and a drain zone and of resistive losses in the drift region that serves for taking up a space charge zone in an off-state situation of the MOS transistor. In high-voltage MOS transistors, the proportion of the resistive losses that is caused by the drift region is actually the proportion that is particularly high and dominant.
In order to reduce the resistive losses in the drift region in high-voltage MOS transistors, compensation components were developed. In the case of the latter, highly n-conducting regions and highly p-conducting regions are situated next to one another in the vertical direction in the drift region. For this purpose, by way of example, pillar-type highly p-doped regions are introduced into a highly n-doped semiconductor body. In this case, the net doping is virtually compensated averaged over the drift region horizontally. In other words, in the above example, the doping of the p-doped pillar-type regions practically compensates for the doping of the n-conducting semiconductor body.
If a reverse voltage is applied to the compensation component, then a significant part of the opposite charge of the ionized dopant atoms is situated in the same horizontal plane, so that the electric field strength is reduced little in the vertical direction between the two main surfaces of the semiconductor body. In other words, only a small resultant gradient of the electric field strength is present here in the vertical direction. Therefore, in the vertical direction, the reverse voltage can be dissipated over a smaller thickness of the drift region of the compensation component.
However, since a higher effective n-type doping is available in the drift region in the on-state situation, the compensation components are distinguished by drastically lower losses in the on state in comparison with conventional MOS transistors equal in area. Compensation components thus have a considerably reduced on resistance.
Compensation components are complicated to fabricate, due to the alternating structure of the p-conducting and n-conducting regions in the drift region, that is to say due to an alternating p
/p
. . . structure in the lateral direction.
In the prior art there are two different methods for fabricating such alternating p
/p
. . . structures of the compensation components.
In the preferred method, multistage epitaxy processes with interposed implantations are employed. In specific terms, in this case n-conducting epitaxial layers are applied to an n
+
-conducting silicon substrate, and each epitaxy process is followed by an implantation of boron atoms at locations lying one above the other, with the result that, in the event of a subsequent thermal treatment, the boron implantations lying one above the other form a pillar-type p-conducting region in an n-conducting region.
In the other customary method, deep trenches are introduced into a silicon body of one conductivity type and are subsequently filled with silicon of the other conductivity type.
What is common to both known methods is that they require, for every chip size in a desired voltage class, an exactly adapted substructure in the silicon of the drift region and their processing is extremely complicated and hence expensive.
Despite this considerable disadvantage of complicated processing and a high outlay, no thought has been given heretofore to configuring a compensation component and also a method for fabricating it in a different way such that these disadvantages can be overcome.
SUMMARY OF THE INVENTION
It is accordingly an object of the invention to provide a compensation component and a method for fabricating it which overcome the above-mentioned disadvantages of the prior art devices and methods of this general type, in which complicated and expensive processing can be dispensed with.
With the foregoing and other objects in view there is provided, in accordance with the invention, a compensation component. The compensation component contains a semiconductor body made of silicon, an n-conducting drift zone formed in the semiconductor body, and a p-conducting dopant introduced into the n-conducting drift zone. The p-conducting dopant forms cluster-shaped regions in the n-conducting drift zone and is selected such that a first gap between an acceptor level and a valence band edge is greater than a second gap between a donor level and a conduction band edge.
What is essential to the present invention is the basic idea either of introducing into the usually n-conducting basic doping of the drift zone homogeneously distributed p-conducting atoms with about the same doping concentration as the n-conducting basic doping or else of departing from the concept of an n-conducting basic body and introducing into a p-conducting silicon body a relatively rapidly diffusing dopant with donor properties, such as, in particular, sulfur and/or selenium, in order thus to produce the desired n-conducting pillar-type regions with the rapidly diffusing dopant.
The text below will first discuss the first alternative, that is to say the homogeneous distribution of atoms of one conductivity type in a basic doping of the other conductivity type of the drift region. Although the text below explains in detail how p-doping or p-conducting atoms, namely in particular indium atoms, thallium atoms and palladium atoms, are introduced into a customary n-conducting doping of a drift zone made of phosphorus, for example, the conductivity types can also be reversed, if appropriate, given corresponding selection of the atoms. The same also applies, of course, to the second method explained in more detail further below.
In the selection of the dopant for the first method, that is to say for the introduction of p-doping atoms into a customary n-type doping of the drift zone, care must be taken to ensure that the gap between the acceptor energy level and the valence band edge of silicon is larger than about 150 meV, so that only a very small proportion of the acceptor atoms are ionized at room temperature in thermal equilibrium. In the forward direction of the compensation component, only a small proportion of the n-conducting doping of the drift region is thus compensated, so that the compensation component, in particular a transistor, has the desired low on-state losses.
Moreover, the gap between the acceptor level and the valence band edge should be larger than the gap between the donor level and the conduction band edge of silicon. This is because in the off-state situation a space charge zone is built up which has the effect that the holes liberated during the ionization of the acceptors immediately blow away and cannot interact with other acceptor rumps. Thus, all the acceptors are then ionized in a short time, so that the donors are compensated in the volume of the drift region. In other words, the conditions prevailing are similar to those in customary compensation components.
An essential advantage of the first method resides in a particularly simple and hence inexpensive process control in comparison with the prior art with epitaxy processes and implantations or deep trenches. The p-conducting dopant, that is to say in particular indium, thallium and palladium, can readily be produced at the same time as the n-type doping during the deposition of the epitaxial layer, so that the implantations can be obviated. In addition, an epitaxial wafer treated in this way can be used as a basic material for all chip sizes of a voltage class, which considerably simplifies the logistics and allows the turnaround times to be shortened. This is because it is possible, during the deposition process for the epitaxial layers, to alter the dopant composition by way of the thickness of the epitaxial layers and thus to set the component properties accordingly.
As an al
Deboy Gerald
Mauder Anton
Schulze Hans-Joachim
Strack Helmut
Chaudhuri Olik
Greenberg Laurence A.
Infineon - Technologies AG
Locher Ralph E.
Stemer Werner H.
LandOfFree
Compensation component and method for fabricating the... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Compensation component and method for fabricating the..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Compensation component and method for fabricating the... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3016981