NanoNET

Official Resources

• Repository: https://github.com/freude/NanoNet
• Documentation: (Repository Wiki/README)
• License: MIT License
• Developers: RMIT University (M. Usman, et al.)

Overview

NanoNET (Nanoscale Non-equilibrium Electron Transport) is an expandable Python framework for modeling electronic structure and quantum transport in nanodevices. It combines the Tight-Binding (TB) method with the Non-Equilibrium Green's Function (NEGF) formalism. NanoNET is specifically designed to handle the efficient construction of Hamiltonians for large, non-periodic or semi-periodic systems (like nanowires) using algorithmically optimized neighbor searching.

Scientific domain: Quantum device physics, nanoelectronics, transport theory. Target user community: Researchers modeling nanowires, quantum dots, and atomic-scale transistors.

Theoretical Methods

• Empirical Tight-Binding (ETB):
  • Uses parameterized orbital overlaps ($sp^3d^5s^*$) for accurate band structures of semiconductors (Si, Ge, III-V).
  • Two-center integral lookup tables.
• Non-Equilibrium Green's Function (NEGF):
  • Calculation of Transmission $T(E)$ and Current $I(V)$ (Landauer-Büttiker).
  • Retarded/Advanced Green's functions for density and spectral properties.
• Algorithms:
  • kd-tree for fast nearest-neighbor search in Hamiltonian construction.
  • Recursive Green's Function (RGF) (implied for transport in wires).

Capabilities

• Hamiltonian Construction:
  • Generates dense, sparse, or block-tridiagonal matrices from atomic coordinates.
  • Auto-detection of periodic blocks for transport.
• Electronic Structure:
  • Complex band structures.
  • Wavefunctions and Local Density of States (LDOS).
• Transport:
  • Transmission coefficients.
  • Current-voltage characteristics.
  • Elastic scattering self-energies.
• System Types:
  • Nanowires (Si, InAs, etc.).
  • Finite clusters / Quantum Dots.
  • Heterostructures.

Key Strengths

• Ease of Hamiltonian Generation: Solves the "coordinate-to-matrix" problem efficiently for arbitrary geometries using kd-trees.
• Pythonic: Pure Python implementation allows for easy scripting, modifying parameters, and integrating with other tools.
• Focus: Optimized specifically for semiconductor nanostructures where ETB is the method of choice.
• Flexibility: Supports arbitrary tight-binding parameter sets (Slater-Koster).

Inputs & Outputs

• Inputs:
  • Atomic coordinates (XYZ type).
  • Parameter files (interaction integrals).
• Outputs:
  • HDF5 files for large datasets.
  • Text/NumPy files for bands and transmission data.

Interfaces & Ecosystem

• SciPy/NumPy: Heavily utilizes sparse matrix algebra.
• Open Source: Available on GitHub, easy to contribute to.

Performance Characteristics

• Construction Speed: Very fast ($O(N \log N)$) construction of H matrices due to kd-tree algorithms.
• Solver Speed: Dependent on system width; uses sparse solvers for diagonalization and inversion.

Comparison with Other Codes

• vs [NEMO5](file:///home/niel/git/Indranil2020.github.io/scientific_tools_consolidated/TightBinding/4.3_Quantum_Transport/NEMO5.md): NEMO5 is a massive, parallel C++ suite for industrial TCAD; NanoNET is a lighter Python research code covering similar physics (ETB/NEGF) but easier to modify and deploy on single machines.
• vs [Kwant](file:///home/niel/git/Indranil2020.github.io/scientific_tools_consolidated/TightBinding/4.3_Quantum_Transport/Kwant.md): Kwant is more general for topological/mesoscopic models; NanoNET includes specific machinery for empirical tight-binding (orbitals, bond angles) of real semiconductors.

Application Areas

• Nanowire Transistors: Modeling I-V in silicon or III-V nanowire FETs.
• Surface Effects: Studying the impact of surface roughness or passivation on transport.
• Dopant Engineering: Effects of single impurities in transport channels.

Verification & Sources

• Primary Source: K. L. et al., "NanoNET: An extendable Python framework for electronic structure and transport", Comput. Phys. Commun. (2019).
• Repository: GitHub Link
• Verification Status: ✅ VERIFIED.