| 摘要 |
988,367. Semi-conductor devices. TEXAS INSTRUMENTS Inc. Sept. 8, 1961 [Sept. 9, 1960], No. 32336/61. Heading H1K. A semi-conductor device comprises a semiconductor body with a PN junction terminating at one face and there defining an enclosed area within which are disposed a plurality of interconnected elongated diffused zones forming PN junctions with the underlying material. One such device is made from a P type silicon wafer, lapped to a thickness of 12 mils and optically finished, by first diffusing donor impurities into its entire surface to a depth about 3 mils, and then lapping one surface to leave a body 4 mils thick with a PN (collector) junction exposed on its upper face. After heating in steam to form an oxide layer over the entire surface the upper face is coated with a photoresist, exposed to light through a mask with a central E-shaped opaque region, the unexposed region of photoresist dissolved away and the surface etched in hydrofluoric acid to remove the oxide coating in the E-shaped area. Phosphorus is diffused from a phosphorus pentoxide source into the silicon surface thus exposed to form a multi-armed emitter region 0À1 mil thick. To facilitate contact to the P type (base) zone gallium is diffused into the base and emitter zones after removal of the remaining oxide layer. During this process a P + zone is formed on the base zone but the N type zone is extended in depth and its surface remains N type. Layers of resist material are next formed over the areas at which the emitter and collector junctions intersect the wafer surface by a photoresist and masking technique as described above, a nickel layer deposited by electroless plating over the wafer surface remaining exposed and the resist removed. Alternatively only the emitter junction is masked, a nickel layer formed over the rest of the wafer being later removed from the collector junction by etching in nitric acid through a wax mask. In either case, to facilitate adhesion, the nickel may be applied in two stages with an intermediate sintering in nitrogen at 700‹ C. for an hour. The resulting device 24 is mounted as shown in Fig. 16 with layers 15, 17, 19, 21 and 22 of flux, layers 16, 20 and 26 of indium solder, and nickel plated molybdenum plate 18 on a gold plated header 25 and the assembly bonded together by heating in an inert atmosphere. Since the solder does not wet silicon the emitter junction is not shortcircuited during this process. Subsequently wires are cold pressure bonded to the emitter and base contacts thus formed. |
