Nanometer-scale materials can provide smaller devices than those currently available.
Ames Research Center, Moffett Field, California
Recently, one-dimensional (1-D) nanostructures, such as nanowires and nanotubes, have become the focal point of research in nanotechnology due to their fascinating properties. These properties are intrinsically associated with low dimensionality and small diameters, which may lead to unique applications in various nanoscale devices. It is generally accepted that 1-D nanostructures provide an excellent test ground for understanding the dependence of physical, electrical, thermal, optical, and mechanical properties on material dimensionality and physical size. In particular, 1-D semiconductor nanostructures, which exhibit different properties as compared with their bulk or thin film counterparts, have shown great potential in future nanoelectronics applications in data storage, computing, and sensing devices.
Phase-change materials (PCMs) are among the most promising media for nonvolatile, re-writable, and highly durable data storage applications. Phase change materials based on the Ge–Sb–Te (GST) multi-element alloy system and other materials, such as In2Se3, have been extensively studied and have been found to be suitable for optical and electrical memories. Resistive switching PRAM devices, using thin-film form of these materials, have been shown to provide faster write/read, improved endurance, and simpler fabrication as compared with the transistor-based nonvolatile memories.
Although substantial improvements in the performance of thin-film-based PRAM have been made over the past decade, a number of issues remain; most notably, large programming current, limited cyclability, and scaling problems when moving to increasingly smaller dimensions. There is particularly a concern about the crystalline-to-amorphous phase transition when high current is required for material melting. The Joule heating effect may cause excessive power dissipation and inter-cell thermal interference, presenting problems for further memory scaling.
The phase transition behavior of nanoscale materials may overcome these limitations. For example, the melting temperature Tmelt of GST nanowire (e.g. 385 °C for 80-nm diameter) is reduced from that of bulk GST (632 °C), a phenomenon consistent with a previous report for semiconductor nanocrystals that reports a decrease of Tmelt with decreasing nanocrystal size. The thermal conductivity of a low-dimensional structure is also reduced relative to its bulk counterpart. The reduced Tmelt thermal conductance and smaller material volume jointly contribute to reduction of the threshold energy required for structural phase transition in nanostructures.
A bottom-up synthesis strategy provides a route to prepare the above phase change nanowires. Some elements or binary compound 1-D nanostructures have been synthesized by thermal evaporation, laser-ablation, chemical vapor deposition (CVD), and metalorganic chemical vapor deposition (MOCVD) methods. One of the challenging issues in the field of PCM nanostructures is the synthesis of high-quality, high-purity binary and ternary alloy nanowires suitable for assembling devices on large wafers for high memory density.
The present invention provides a method for growing 1-D phase change nanowires of GST alloys and similar alloys, such as GeTe and Sb2Te3, or Ge, Te, and Sb element powders on any kind of substrates, using a thin film or nanoparticles of at least one of Au, Ni, Ti, Cr, In, Sb, Ge, and Te as a growth catalyst. The atomic ratio in the alloy nanowires can be controlled by controlling the composition of feedstock in the starting mixture. For example, Ge2Sb2Te5 nanowires can be grown from the mixture of GeTe and Sb2Te3 with 2:1 mole ratio. The length of the nanostructures depends upon the growth time applied.