Strong, controllable, robust, and scalable graphene n-doping for optoelectronic and micro/nano-electronic applications

J. Folkson, A. Ashraf, M.D. Eisaman
Stony Brook University(SBU)/Brookhaven National Lab(BNL),
United States

Keywords: graphene, doping, MOSFET, GFET, photovoltaics, sensors


Graphene holds promise for many applications such as microelectronics, optoelectronics, and energy storage due to its outstanding properties. The realization of these applications, however, requires tailoring graphene’s electronic properties via controlled doping (addition or removal of electrons). While achieving strong p-doping of graphene (removing electrons) is straightforward, approaches to n-dope graphene (adding electrons), like chemical doping, have yielded low electron density. Furthermore, chemical doping typically degrades over time and can adversely affect intrinsic graphene’s properties. We recently developed a new approach that yields strong (1.33x1013 electrons/cm2), robust, and spontaneous graphene n-doping on low-cost ($5/m2) soda-lime-glass (SLG) via surface-transfer doping from sodium (Na), without any external chemical, high-temperature, or vacuum processes. The n-doping increases further (2.11x1013 e/cm2) when graphene is transferred onto a p-type copper indium gallium diselenide (CIGS) semiconductor that is itself deposited onto SLG, due to Na diffusion through the CIGS. Using this effect, we have demonstrated the first n-graphene/p-CIGS photovoltaic Schottky junction. For applications to micro/nano-electronics, applying this alkali-metal surface-transfer doping to a wider range of substrates and semiconductors has the potential to enable short-channel, high-speed, graphene-based n-channel MOSFETs, in addition to a wide range of novel sensors and optoelectronic devices. Successfully applying these discoveries will require further optimization to answer some critical questions: (1) What are the details of the doping mechanism? We must gain a more complete understanding of the n-doping mechanism by probing the coordination of Na at the surface of SLG. (2) How can we control the graphene-semiconductor junction properties? In order to control the graphene-semiconductor junction properties, we must have fine control over the doping strength and interfacial recombination rate. Initial experiments suggest that the thickness of a few-nm thick spacer layer between the graphene and semiconductor can drastically reduce recombination and be used to tune the graphene Fermi-level shift. The electronic structure of graphene can be controlled not only through doping, but also by control over the number of graphene layers. (3) How can we extend this technique to a wider array of substrates and semiconductors? In order to ensure the maximum impact on a wide range of technologies, we must extend these results to other substrates. We will test other alkali-containing substrates, such as sodium phosphosilicate glass and flexible glass intended for consumer electronics, and also test devices where the doping originates from the gate dielectric itself, thereby enabling graphene-semiconductor junctions with any desired semiconductor.