1: \begin{abstract}

2: Using a real-space high order finite-difference approach, we investigate the

3: electronic structure of large spherical silicon nanoclusters. Within Kohn\textendash Sham density functional theory and using pseudopotentials, we report the self-consistent field convergence of a system with over 100,000 atoms: a Si$_{107,641}$H$_{9,084}$ nanocluster with a diameter of

4: 16 nm. Our approach uses Chebyshev filtered subspace iteration to speed-up the convergence of the eigenspace, and blockwise Hilbert space filling curves to speed-up sparse matrix\textendash vector multiplications, all of which is implemented in the PARSEC code. For the largest system, we utilized 2048 nodes

5: (114,688 processors) on the Frontera machine in the Texas Advanced

6: Computing Center. Our quantitative analysis of the electronic structure

7: shows how it gradually approaches its bulk counterpart as a function

8: of the nanocluster size. The band gap is enlarged due to quantum confinement

9: in nanoclusters, but decreases as the system size increases, as expected.

10: Our work serves as a proof-of-concept for the capacity of the real-space

11: approach in efficiently parallelizing very large calculations using high performance computer platforms, which can straightforwardly be replicated in other systems with more than $10^{5}$ atoms.

12: \end{abstract}

13: