Pushing forward 2D/3D contacts simulations

By Tom Sharp on September 23, 2016

In a recent paper in Journal of Physics: Condensed Matter, a new approach enabling fully ab initio simulations of the contacts between 3D and 2D materials was proposed. We find out more from the author Dr Artem Fediai:

Most emerging 2D materials possess high electron mobility and effective gate control. They are considered to be the main contenders to substitute silicon in channel field-effect transistors. However, even the best 2D material is worth nothing if we fail to provide its low-ohmic contact to an external voltage source made of a bulk (or 3D) metal.

In silicon technology, where both the channel and the electrodes are bulk materials, minimization of the contact resistance is resolved trivially: the Si substrate is highly doped at the contact with a metal. Thus, once good adhesion is achieved, the source/drain contacts in Si MOSFET are low ohmic.

This does not work for 2D/3D contacts as 2D material have no bulk, which makes doping them extremely challenging and harmful as it “kills” the main its advantage – high mobility. This calls for new strategies in designing 2D/3D contacts. It should be emphasized that these are a new type of a contact in electronics, and cannot be fully understood and described with the concepts borrowed from the theory of conventional 3D/3D contacts (such as the ‘Schottky barrier’, the ‘transmission line method’, etc.). Likewise, conventional language cannot be successfully used to understand and thus to improve 2D/3D contacts. New theories and methods of description are thus required.

An example of (a) two local contacts (a standard molecular electronics set-up: an organic molecule between the two metallic electrodes), (b) an extended contact between 2D and 3D materials (2D material shown is a black phosphorus). Current flow is shown with the yellow arrows. Image taken from Fediai et al. J. Phys. Cond. Mat. 28 395303 © IOP Publishing, All Rights Reserved.

To understand 2D/3D contact properties at a fundamental level using models with no additional assumptions, we use the non-equilibrium Green function method combined with the density functional theory (DFT) – hereinafter, DFT+NEGF. The standard combination of DFT+NEGF used in a variety of codes, but it has been primarily aimed to simulate the geometry of a local contact (figure 1a), whereas 2D/3D planar contacts belong to the extended type. The corresponding geometrical difference is obvious from figure 1. It is experimentally shown that the current inflow in the extended contact is gradual (yellow arrows in figure 1b) that requires the simulation of the long (up to 100 nm) contacts, which is prohibitively numerically expensive when using standard NEGF+DFT methods. Our method pushes the limits of the NEGF+DFT method, which allows simulating relevant contact lengths from 1 to 100 nm with reasonable numerical resources. As a result, it is now possible to ab-initio simulate how scaling of the contact length influences corresponding electrical characteristics. This enables realistic predictions of the scaling properties of the emerging 2D materials, which may stimulate further miniaturization in electronics.

Artem Fediai received his PhD degree in solid-state electronics from the Kyiv Polytechnic Institute (Ukraine) in 2012. As a postdoc, he joined the Chair of Materials Science and Nanotechnology of the Dresden University of Technology in 2013. He supports the Center for Advancing Electronics Dresden (cfaed) in developing atomistic simulation approaches, which enable an ab-initio design of the emerging electronic components.

Dmitry A Ryndyk received his PhD in 2000 at the Institute for Physics of Microstructures of the Russian Academy of Sciences. As a postdoc Dmitry joined the group of Joachim Keller at the Institute for Theoretical Physics, University of Regensburg. From 2004 he continued to work in Regensburg in the groups of Gianaurelio Cuniberti and Klaus Richter and finished the Habilitation in the field of quantum transport at the atomic and molecular scale. From 2012 till 2016 Dmitry Ryndyk was a senior scientist and group leader at the Institute for Materials Science, TU Dresden. From 2016 he is an assistant professor (Privatdozent) at the Bremen Center for Computational Materials Science (BCCMS) and the Institute for Theoretical Physics, University of Bremen. The main research interests include the theory of quantum transport and many-body nonequilibrium effects at nanoscale, topological states of matter, planar and surface molecular electronics, 2D materials, nanoscale device modeling, computational platform for nanoscale and materials modeling, nonequilibrium superconductivity.

Gianaurelio Cuniberti has held the Chair of Materials Science and Nanotechnology at the Dresden University of Technology since 2007. He leads the Nanobiomaterials Department of the Max Bergmann Center of Biomaterials Dresden and is the founding director of the Dresden Center for Computational Materials Science (DCCMS). He studied Physics at the University of Genoa and at the University of Hamburg and was visiting scientist at MIT and the Max Planck Institute for the Physics of Complex Systems Dresden. From 2003 to 2007 he was at the head of a Volkswagen Foundation Junior Research Group at the University of Regensburg. He is distinguished visiting Professor at the Division of IT Convergence Engineering of POSTECH, the Pohang University of Science and Technology and Adjunct Professor for the Department of Chemistry at the University of Alabama.