Doctor of Philosophy (PhD)


Electrical Engineering

Document Type



This work studies various emerging reduced dimensional materials for very large-scale integration (VLSI) interconnects. The prime motivation of this work is to find an alternative to the existing Cu-based interconnect for post-CMOS technology nodes with an emphasis on thermal stability. Starting from the material modeling, this work includes material characterization, exploration of electronic properties, vibrational properties and to analyze performance as a VLSI interconnect. Using state of the art density functional theories (DFT) one-dimensional and two-dimensional materials were designed for exploring their electronic structures, transport properties and their circuit behaviors. Primarily carbon nanotube (CNT), graphene and graphene/copper based interconnects were studied in this work.

Being reduced dimensional materials the charge carriers in CNT(1-D) and in graphene (2-D) are quantum mechanically confined as a result of this free electron approximation fails to explain their electronic properties. For same reason Drude theory of metals fails to explain electronic transport phenomena. In this work Landauer transport theories using non-equilibrium Green function (NEGF) formalism was used for carrier transport calculation. For phonon transport studies, phenomenological Fourier’s heat diffusion equation was used for longer interconnects. Semi-classical BTE and Landauer transport for phonons were used in cases of ballistic phonon transport regime. After obtaining self-consistent electronic and thermal transport coefficients, an equivalent circuit model is proposed to analyze interconnects’ electrical performances.

For material studies, CNTs of different variants were analyzed and compared with existing copper based interconnects and were found to be auspicious contenders with integrational challenges. Although, Cu based interconnect is still outperforming other emerging materials in terms of the energy-delay product (1.72 fJ-ps), considering the electromigration resistance graphene Cu hybrid interconnect proposed in this dissertation performs better. Ten times more electromigration resistance is achievable with the cost of only 30% increase in energy-delay product. This unique property of this proposed interconnect also outperforms other studied alternative materials such as multiwalled CNT, single walled CNT and their bundles.



Committee Chair

Srivastava, Ashok