Active solid-state devices (e.g. – transistors – solid-state diode – Field effect device – Junction field effect transistor
Reexamination Certificate
2001-11-09
2003-10-07
Wilson, Allan R. (Department: 2815)
Active solid-state devices (e.g., transistors, solid-state diode
Field effect device
Junction field effect transistor
C257S287000
Reexamination Certificate
active
06630698
ABSTRACT:
The present invention concerns a semiconductor device with a semiconductor body having a blocking pn-junction, a first zone of a first conductivity type, which is connected to a first electrode and abuts one of the zones of a second conductivity type opposite the first conductivity type forming the blocking pn-junction, and with a second zone of the first conductivity type, which is connected to a second electrode, whereby the side of the zone of the second conductivity type facing the second zone forms a first surface and in the region between the first surface and a second surface, which lies between the first surface and the second zone, areas of the first and of the second conductivity type are nested.
Such semiconductor devices are also known as compensation devices. Such compensation devices are, for example, n- or p-channel MOS field effect transistors, diodes, thyristors, GTOs, or other components. In the following, however, a field effect transistor (also referred to briefly as “transistor”) is assumed as an example.
There have been various theoretical investigations spread over a long period of time concerning compensation devices (cf. U.S. Pat. No. 4,754,310 and U.S. Pat. No. 5,216,275) in which, however, specifically, improvements of the on-resistance RDS(on) but not of stability under current load, such as, in particular, robustness with regard to avalanche and short circuit in the high-current operation with high source-drain voltage, are sought.
Compensation devices are based on mutual compensation of the charge of n- and p-doped areas in the drift region of the transistor. The areas are spatially arranged such that the line integral above the doping along a line running vertical to the pn-junction in each case remains below the material-specific breakdown voltage (silicon: approximately 2×10
12
cm
−2
). For example, in a vertical transistor, as is customary in power electronics, p- and n-columns or plates, etc. may be arranged in pairs. In a lateral structure, p- and n-conductive layers may be stacked on each other laterally alternating between a groove with a p-conductive layer and a groove with an n-conductive layer (cf. U.S. Pat. No. 4,754,310).
By means of the extensive compensation of the p- and n-doping, the doping of the current-carrying region (for n-channel transistors, the n-region; for p-channel transistors, the p-region) can be significantly increased, whereby, despite the loss in current-carrying area, a clear gain in on-resistance R
DS
(on) results. The blocking capability of the transistor depends substantially on the difference between the two dopings. Since, because of the reduction of the on-resistance, a doping higher by at least one order of magnitude of the current-carrying area is desirable, control of the blocking voltage requires controlled adjustment of the compensation level, which can be defined for values in the range ≦±10%. With a greater gain in on-resistance, the range mentioned becomes even smaller. The compensation level is then definable by
(p-doping−n-doping)
-doping
or by
charge difference/charge of one doping area.
Other definitions are, however, possible.
It is an object of the present invention to provide a robust semiconductor component of the kind initially mentioned, to be firstly distinguished by a high “avalanche” ruggedness and high current load capacity before and/or during breakdown and secondly simple to produce with reproducible properties in view of technological latitudes of fluctuation of manufacturing processes.
This object is accomplished according to the invention, in a semiconductor component of the kind initially mentioned, in that the regions of the first and second types of conductivity are so doped that charge carriers of the second conductivity type predominate in regions near the first surface and charge carriers of the first conductivity type in regions near the second surface.
Preferably, the regions of the second conductivity type do not extend as far as up to the second zone, so that between said second surface and the second zone, a weakly doped region of the first conductivity type remains. It is possible, however, to allow the width of this region to go to “zero.” The weakly doped region, however, provides certain advantages, such as enhancement of the barrier voltage, “smooth” profile of the electrical field strength, or improvement of commutation properties of the inverse diode.
In another refinement of the invention, it is provided that between the first and second surfaces, a degree of compensation effected by the doping is so varied that atomic residues of the second conductivity type dominate near the first surface and atomic residues of the first conductivity type near the second surface. In other words, there are sequences of p, p
−
, n
−
, n or n, n
−
, p
−
, p layers between the two surfaces.
Advantageous improvements of the semiconductor device according to the invention (hereinafter also referred to as “compensation device”) are disclosed by the other dependent claims.
The effect of the areas nested in each other, alternating different conductivity types, on the electrical field, is, in contrast to a conventional DMOS transistor, for example, as follows (“lateral” and “vertical” refer in the following to a vertical transistor):
(a) There is a cross-field, “lateral” to the direction of the connection between the electrodes, the strength of which depends on the proportion of the lateral charge (line integral perpendicular to the lateral pn-junction) relative to the breakdown charge. This field leads to the separation of electrons and holes and to a reduction in the current-carrying cross-section along the current paths. This fact is of primary significance for the understanding of the processes in avalanche, of the breakdown characteristic curve, and of the saturation region of the output characteristics diagram.
(b) The “vertical” electrical field parallel to the direction of the connection between the electrodes is determined locally by the difference between the adjacent dopings. This means that with an excess of donors (n-loaded distribution: the charge in the n-conductive areas exceeds the charge of the p-areas) on the one hand, a DMOS-like field distribution (maximum of the field on the blocking pn junction, decreasing field in the direction of the opposing back of the device) appears, whereby the gradient of the field is, however, clearly less than would correspond to the doping of the n-area alone. On the other hand, however, by overcompensation of the n-conductive area with acceptors, a field distribution rising in the direction of the back is possible (p-loaded distribution: excess of acceptors compared to the donors). In such a design, the field maximum lies at the bottom of the p-area. If the two dopings are exactly compensated, there is a horizontal field distribution.
With an exact horizontal field distribution, the maximum of the breakdown voltage is obtained. If the acceptors or the donors predominate, the breakdown voltage drops in each case. If the breakdown voltage is then plotted as a function of the degree of compensation, a parabolic characteristic is obtained.
Constant doping in the p- and n-conductive areas or even a locally varying doping with periodic maxima of equal height results in a comparatively sharply pronounced maximum of the “compensation parabola”. For the benefit of a “production window” (including the fluctuations of all relevant individual processes), a comparatively high breakdown voltage must be steered for in order to obtain reliable yields and production reliability. Consequently, the objective must be to make the compensation parabola as flat and as broad as possible.
When the blocking voltage is applied to the device, the drift region, i.e., the region of the areas of opposite doping arranged in pairs, is cleared of mobile charge carriers. The positively charged donor cores and the negatively charged acceptor cores remain in the spreading space charge region. They then determine the co
Ahlers Dirk
Deboy Gerald
Rueb Michael
Strack Helmut
Weber Hans
Infineon AG
Wilson Allan R.
LandOfFree
High-voltage semiconductor component does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with High-voltage semiconductor component, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and High-voltage semiconductor component will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3138627