Abstract: Along the outline of a first doped region, a first mask is formed. The mask is made up of a dielectric opposed to the oxygen diffusion. Another mask is created within this first mask, using a process of selective thermal oxidation. The second mask is used to implant dopant in a second region which will only be defined along the outlines of the first region.

Abstract: A description is given of two versions of an integrated structure in the emitter switching configuration comprising a high-voltage bipolar power transistor and a low-voltage MOS power transistor. In the vertical MOS version, the emitter region of the bipolar transistor is completely buried, partly in a first N- epitaxial layer and partly in a second N- epitaxial layer; the MOS is located above the emitter region. The bipolar is thus a completely buried active structure. In the horizontal MOS version, in a N- epitaxial layer there are two P+ regions, the first, which constitutes the base of the bipolar transistor, receives the N+ emitter region of the same transistor; the second receives two N+ regions which constitute the MOS source and drain regions, respectively; the front of the chip is provided with metal plating to ensure the connection between the MOS drain and the bipolar emitter contacts.

Abstract: The emitter region of a speed-up transistor is created in a base of a final transistor of a Darlington device and has a relatively low dopant concentration and small thickness.

Abstract: A description is given of two versions of an integrated structure in the emitter switching configuration comprising a high-voltage bipolar power transistor an a low-voltage MOS power transistor. In the vertical MOS version, the emitter region of the bipolar transistor is completely buried, partly in a first N-epitaxial layer and partly in a second N epitaxial layer; the MOS is located above the emitter region. The bipolar is thus completely buried active sturcture. In the horizontal MOS version, in a N-epitaxial layer there are two P+regions, the first, which constitutes the base of the bipolar transistor, receives the N+emitter region of the same transistor; the second receives two N+regions which constitute the MOS source and drain regions, respectively; the front of the chip is provided with metal plating to ensure the connection between the MOS drain and the bipolar emitter contacts.

Abstract: A multiplicity of semiconductor chips constituting the active elements of a semiconductor power device are attached, in a predetermined configuration, to a metal plate which acts as both a support and a first electrical terminal. Two other electrical terminals are formed by two metal electrodes having an interdigitated structure with the fingers of one electrode being inserted between those of the other electrode and which are positioned above the semiconductor chips and have tabs for establishing a connection between the electrodes and the two contact areas of each chip.

Abstract: The invention relates to a method for the manufacture of high voltage semiconductor devices with at least one planar junction with a variable charge concentration.The method consists in doping with impurities of a same type, in a region of monocrystalline semiconductor material, a first zone, and then a second zone which comprises the first, and so on, and in carrying out a subsequent heat treatment so as to provide a planar junction with a stepped profile and a concentration of impurities which decreases from the center to the periphery in a predetemined range. In this way the intensity of the surface electric field, when the junction is reverse biased, is reduced as a result of which it is possible to provide planar junctions having very high breakdown voltages of some thousands of volts.

Abstract: A monolithic semiconductor device including an integrated control circuit and a pair of power transistors in a Darlington configuration integrated in the same chip solves the problem of ON-OFF switching which is prevented by the presence of parasitic transistors within the structure, these transistors preventing the correct operation of the device at saturation. The solution involves a suitable arrangement of the components in the chip, with the output transistor of the Darlington pair disposed in an intermediate position between the drive transistor of the pair and the integrated control circuit. The addition of semiconductor shields, disposed between the output transistor of the Darlington pair and the integrated control circuit further reduces the damaging effects of the parasitic transistors.

Abstract: Filiform elements of predetermined resistivity, e.g. selectively destructible leads of an electrically programmable read-only memory, are formed on a semiconductor substrate such as a silicon body by first depositing thereon a layer of dielectric material such as SiO.sub.2 and topping that layer with a conductive or nonconductive coating which is resistant to a chemical such as hydrofluoric acid capable of attacking the dielectric layer. Next, the top coating is partly destroyed by photolithographic treatment to leave at least one substantially rectangular patch. Thereafter, the dielectric layer is isotropically attacked by the aforementioned chemical with resulting reduction to about half its original thickness and concurrent lateral erosion of a patch-supporting pedestal of that layer whereby channels of generally semicylindrical concavity are formed around the periphery of this pedestal.

Abstract: Filiform elements of predetermined resistivity, e.g. selectively destructible leads of an electrically programmable read-only memory, are formed on a semiconductor substrate such as a silicon body by first depositing thereon a layer of dielectric material such as SiO.sub.2 and topping that layer with a conductive or nonconductive coating which is resistant to a chemical such as hydrofluoric acid capable of attacking the dielectric layer. Next, the top coating is partly destroyed by photolithographic treatment to leave at least one substantially rectangular patch. Thereafter, the dielectric layer is isotropically attacked by the aforementioned chemical with resulting reduction to about half its original thickness and concurrent lateral erosion of a patch-supporting pedestal of that layer whereby channels of generally semicylindrical concavity are formed around the periphery of this pedestal.

Abstract: Two patches of silicon nitride are formed above spaced-apart regions of an n-type substrate (2) on an overlying oxide layer (8) of small thickness. Arsenic ions are then implanted through the oxide layer in substrate areas not covered by the patches whereupon one patch (10a) and an adjoining portion of the oxide layer are covered by a photoresist mask (14), leaving unprotected the second patch (10b) and an oxide portion adjacent thereto. The wafer is then bombarded with boron ions, first at a relatively low energy level to penetrate the last-mentioned oxide portion and then at a higher energy level with additional penetration of the second patch (10b) to form a p-well (18) bounded by a p+ guard zone (20); the previously implanted arsenic ions in the unbombarded area form an n+ guard zone (22).