BN/Graphene/BN sandwiches for high mobility nanodevices

点击翻译按钮获取中文摘要

The mobility of carriers in graphene is commonly used as a figure of merit for the quality of a device. Typical mobilities of up to 10000 cm.V-1s-1 can be achieved in macroscopic structures on SiO2 substrates and up to 200.000 cm.V-1s-1 has been reported in large suspended structures, however this method significantly reduces lithographic options. The search for a more suitable substrate dielectric has recently resulted in a major breakthrough when hexagonal boron nitride (h-BN) was found to significantly increase the electronic quality of graphene devices yielding mobilities of about 140.000 cm.V-1s-1 close to the charge neutrality point. While BN has therefore been seen as a success in terms of material quality, recent experiments on nanostructures of graphene BN substrates revealed no significant improvement in mobilities and disorder density compared to similar structures on SiO2. A general speculation in the field is that the quality is limited by edge disorder, which plays an increasing role as devices are made smaller. While this is a possibility, it has never been confirmed experimentally, and it is also well know that the mobility of even large devices can be seriously deteriorated by anything less than ultra-clean lithography. We therefore postulate that the limiting factor in these BN-graphene nanostructures is indeed lithographically induced, and propose a new processing scheme, based on the basic idea of sandwiching the graphene layer between two protective BN sheets, in order to allow an essentially perfectly clean process, and hopefully produce nanodevices of unprecedented mobility.