A significant challenge for the continued deployment of nanotechnology is manufacturing scalability. While it is often possible to demonstrate exciting technological breakthroughs using expensive laboratory demonstrations, the transition to a manufacturing environment can be the principal bottleneck. One such example is the semiconductor quantum dot array in which an ordered assembly of uniformly-sized nanoscale 3D islands is formed by directed deposition on a semiconductor substrate (e.g., germanium (Ge) islands formed on a crystalline silicon (Si) surface). Such arrays have highly promising potential for high density storage and optoelectronic device applications, while in principal being compatible with existing silicon-based microelectronic device fabrication technology. Lithographic substrate patterning, one approach to directed island growth, is limited in terms of manufacturing throughput, constraining the practical application of such technology. Other approaches have relied on natural pattern formation, such as utilizing the stress pattern formed by misfit dislocation arrays in a hetero-substrate, but to date these have led to relatively poor results in terms of island size and spatial distribution uniformity.
In this project, we are investigating a new approach for heteroepitaxial semiconductor island patterning that may be compatible with the high-throughput requirements of nanomanufacturing. The essential concept of the approach is to produce a patterned surface stress on multiple semiconductor substrates using a single pre-fabricated indenter mold that can be reused in a high volume setting. In our investigations, an SiGe alloy wafer is used as a substrate for Ge deposition and subsequent island formation. An Si-based mold consisting of a prepatterned array of nanoindenters is used to apply a patterned stress to the SiGe substrate at elevated temperature. The stress imposed by the indenter array induces a compositional redistribution of the Ge atoms in the substrate, with the larger Ge atoms moving away from the indented regions, as shown in the figure below. Here, the gold-colored atoms represent a rigid indenter material, the red atoms Si, and blue atoms Ge. Once the indenter mold is removed, the substrate retains an “imprint” of the indenter array because the inhomogeneous distribution of the Ge atoms – this imprint will then be used to direct Ge island assembly.
This project is being carried out in collaboration with the group of Professor Sang Han at the University of New Mexico and is supported by the National Science Foundation (CMMI-1068841).