On the Physics of Sub-Micron Semiconductor Devices

Abstract

Considerable evidence now exists that we are on the threshold of a second industrial revolution ushered in by the recent development and massive proliferation of modern semiconductor micro-electronics. The impetus from economic, defense and technological interests is pushing towards ever increasing micro-miniaturization (Ballantyne, 1979, Einspruch, 1979). So far the achievement of large-scale-integration (LSI) has been sustained by a relatively sound base of reliable scientific knowledge in the fields of materials science, fabrication processes, microelectronics and semiconductor physics. The advent of high resolution electron, X-ray, molecular and ion-beam lithographic techniques is currently opening prospects for very-large-scale-integration (VLSI) in which individual feature sizes down to the molecular scale of 10–20 nanometers are not inconceivable. Indeed, simple experimental structures of 0.1 to 0.01 micrometers have already been achieved (Broers et al., 1978). However, these new developments are considerably hampered by a large gap in our understanding of non-equilibrium semiconductor heterostructures on scales intermediate to the true atomic scale (≲ 10 Angstroms) and the bulk solid-state macro-scale (≳ lμm = 10,000 Angstroms). It is already apparent that simple down-scaling of processing, device-function and performance, bulk physics, etc., is not adequate; nor indeed, is a straight-forward up-scaling of known atomic-scale phenomena.