| 摘要 |
1,145,954. Semi-conductor devices. RADIO CORPORATION OF AMERICA. 10 Feb., 1967 [17 Feb., 1966], No. 6589/67. Heading H1K. Isolated islands of semi-conductor material supported by insulating material are produced by forming a groove in the front surface of a wafer of semi-conductor material, forming a hole extending from the bottom of the groove through the remaining thickness of the wafer, depositing insulating material in the groove by forcing it through the hole from the back of the wafer and removing the back surface of the wafer to a depth sufficient to isolate the islands. As shown, an integrated circuit assembly is produced by forming devices 4 in the top face of a semi-conductor wafer 2, Fig. 2(a), forming grooves 6 between the devices, Fig. 2(b), forming holes 8 through the wafer by cutting grooves 10 in the bottom face of the wafer, the holes being formed at the intersections of the grooves 6 and 10, Fig. 2(c), placing the top face of the wafer against a polished face of a vitreous carbon plate arranged in a mould, covering with a layer of powdered glass and heating and pressing to force molten glass through the apertures 8 and into the grooves 6, Fig. 2(d), and then lapping away the lower face of the wafer to electrically isolate the device areas, Fig. 2(e). The glass does not adhere to the vitreous carbon plate and this plate prevents the glass from covering the devices. The devices may be produced in the regions of semi-conductor material after they have been isolated. In a second embodiment, Fig. 5 (not shown), devices which are difficult to form in the same body of semi-conductor material, e.g. PNP and NPN devices, are integrated by forming one set of components in a first wafer (20) and the other set of components in a second wafer (22) forming grooves between the components in each wafer, forming holes between the grooves and the backs of each wafer, the devices, grooves and holes being so arranged that the projections of the second wafer can be inserted through the holes in the first wafer so that all the devices are co-planar, and then placing the ends of the projections of both wafers in contact with a vitreous carbon plate and embedding the assembly in glass as in the first embodiment. The back of the assembly is then lapped off to a depth sufficient to electrically isolate the various devices. In modifications of the second embodiment the second wafer may comprise a low resistivity surface or may be of a metal such as nickel to provide conductive regions located between the devices provided by the first wafer, Fig. 7 (not shown). The grooves may be produced by etching, sawing or ultrasonic cutting, and the glass may be melted by RF heating. Alternatively a readily flowing synthetic resin may be used instead of glass.
|
