IEDM

Session 35: Modeling and Simulation Progress in Modeling Methodology and Approaches

Wednesday, December 6
Continental Ballroom 1-3
Co-Chairs: Dragica Vasileska, Arizona State University
Stephen Cea, Intel Corp.

1:35 PM
35.1 NEGF based transport modelling with a full-band, pseudopotential Hamiltonian: Theory, Implementation and Full Device Simulations, M. G. Pala, O. Badami*, and D. Esseni*, Université Paris-Saclay, *University of Udine

This paper presents the theory, implementation and application of a new quantum transport, NEGF based modelling approach employing a full-band Empirical Pseudopotential (EP) Hamiltonian. The use of a hybrid real-space$/$plane-waves basis results in a remarkable reduction of the computational burden compared to a full plane waves basis, which allowed us to obtain complete, self-consistent simulations for both FETs and Tunnel FETs in Si or in Ge, and with geometrical features in line with forthcoming CMOS technologies.

2:00 PM
35.2 First-principles based quantum transport simulations of nanoscale field effect transistors (Invited), M. Shin, H.-E. Jung, and S. Jung, Korea Advanced Institute of Science and Technology

We present first-principles density functional theory (DFT) based quantum transport simulations of nanoscale field effect transistors made of Ge, Si, strained-Si, and few-layer black phosphorus channels. The effects of atomistically modeled, crystalline/amorphous SiO2 gate dielectrics on device performance are investigated. A spectral adjustment technique is developed to overcome the band gap underestimation problem of DFT and applied to simulations of tunnel field effect transistors.

2:25 PM
35.3 Dopant diffusion in Si, SiGe and Ge : TCAD model parameters determined with density functional theory, Y. Park, C. Zechner*, Y. Oh**, H. Kim,; I. Martin-Bragado**, E. M. Bazizi, and F. Benistant, GLOBALFOUNDRIES, *Synopsys GmbH, **Synopsys Inc.

We used density functional theory (DFT) to calculate TCAD parameters to describe dopant diffusion in Si, SiGe and Ge. The dopant profile simulated in TCAD with calculated parameters is in good agreement with experiment. It is demonstrated that DFT could help to get unknown TCAD parameters and provide valuable insights on the parameter relations supporting experimental data.

2:50 PM
35.4 First Topography Simulation of SiC-Chemical-Vapor-Deposition Trench Filling, Demonstrating the Essential Impact of the Gibbs-Thomson Effect, K. Mochizuki, S. Ji, R. Kosugi, Y. Yonezawa, and H. Okumura, National Institute of Advanced Industrial Science and Technology

A technology-computer-aided-design-based topography-simulation model is proposed to simulate chemical-vapor-deposition trench filling for SiC superjunction devices. Experimental observations, concerning void formation and mesa overetching, are reproduced for the first time by including the Gibbs-Thomson effect (i.e., the effect of curvature of a growing surface on equilibrium vapor-phase concentration of growing species).

3:15 PM
35.5 A Physics-Based Investigation of Pt-Salt Doped Carbon Nanotubes for Local Interconnects, J. Liang, R. Ramos*, J. Dijon*, H. Okuno**, D. Kalita**, D. Renaud*, J. Lee***, V. P. Georgiev***, S. Berrada***, T. Sadi***, A. Asenov***, B. Uhlig^, K. Lilienthal^, A. Dhavamani^, F. Könemann^^, B. Gotsmann^^, G. Goncalves^^^, B. Chen^^^, K. Teo^^^, R. R. Pandey, and A. Todri-Sanial, CNRS/LIRMM-University of Montpellier, *University Grenoble Alpes/CEA-LITEN, **University Grenoble Alpes/CEA-INAC, ***University of Glasgow, ^Fraunhofer IPMS, ^^IBM Research Zurich, ^^^Aixtron Ltd.

We investigate, by combining physical and electrical measurements together with an atomistic-to-circuit modeling approach, the conductance of doped carbon nanotubes (CNTs) and their eligibility as possible candidate for next generation back-end-of-line (BEOL) interconnects. Ab-initio simulations predict a doping-related shift of the Fermi level, which reduces shell chirality variability and improves electrical conductance up to 90% by converting semiconducting shells to metallic. Circuit-level simulations predict up to 88% signal delay improvement with doped vs. pristine CNT. Electrical measurements of Pt-salt doped CNTs provide up to 50% of resistance reduction which is a milestone result for future CNT interconnect technology.