New divided route to III-Vs on silicon
A team of scientists from Philips and the Kavli Institute of Nanoscience Delft claims the world first growth of III-V semiconductor nanowires on germanium and silicon substrates (Nature Materials, November 2004). The scientists believe that this is an important step towards the integration of the superior high frequency and optoelectronic properties of III-V semiconductors with the huge silicon technology base of the semiconductor industry.
The team grew the III-V material in a "bottom up" approach - instead of growing a layer over the entire substrate and removing the parts that are not needed, the nanowires are only grown at substrate locations where needed. Because this results in many small individual structures rather than one large connected layer, the mechanical stress in the substrate is relieved and perfect epitaxial growth can be achieved.
Until now, III-V semiconductors could not be fabricated on silicon or other group IV materials by the conventional fabrication route of thin-film deposition and lithographic structuring. Fundamental issues such as lattice and thermal expansion mismatch prevent the crystallographic structure of the substrate being copied in a layer grown on top ("epitaxial growth"). Such a growth mode is essential to produce the required material properties such as a low interface resistance.
The new nanowire growth method uss a vapour-liquid-solid (VLS) method, where metal (gold) seeds are deposited (using conventional lithography) at substrate locations where the nanowires should grow. Then the semiconductor material is applied to the substrate in vapour form. The vapour dissolves into the metal seeds, and when this mixture becomes oversaturated, growth of the semiconducting material in the form of a nanowire starts. "Although this process is not new, our team of Philips and the Kavli Institute of Nanoscience was the first to apply it successfully to grow III-V materials on silicon and germanium substrates," says Dr Erik Bakkers, senior scientist at Philips Research and first author of the article in Nature Materials.
The team showed that perfect epitaxial growth, with atomically smooth interfaces and low contact resistance could be reached, providing an excellent base to explore these materials in devices such as transistors, integrated circuits and light-emitting diodes.
III-V semiconductors such as gallium-arsenide (GaAs) or indium-phosphide (InP) are attractive candidates for high-frequency (high-bandwidth connectivity) or optoelectronic (integrated optics or LEDs) applications. Until now, devices using these materials are grown onto substrates of the same class, making them expensive. The Nature Materials article describes the growth of InP nanowires on germanium substrates. In the meantime, the team has also succeeded in growing InP on silicon substrates.
