1: \begin{abstract}

2: We present a comparison of methods for treating the electrostatic

3: interactions of finite, isolated systems within periodic boundary

4: conditions (PBCs), within Density Functional Theory (DFT), with particular

5: emphasis on linear-scaling (LS) DFT. Often, PBCs are not physically

6: realistic but are an unavoidable consequence of the choice of basis

7: set and the efficacy of using Fourier transforms to compute the Hartree

8: potential. In such cases the effects of PBCs on the calculations need

9: to be avoided, so that the results obtained represent the open rather

10: than the periodic boundary. The very large systems encountered in

11: LS-DFT make the demands of the supercell approximation for isolated

12: systems more difficult to manage, and we show cases where the open

13: boundary (infinite cell) result cannot be obtained from extrapolation

14: of calculations from periodic cells of increasing size. We discuss,

15: implement and test three very different approaches for overcoming

16: or circumventing the effects of PBCs: truncation of the Coulomb interaction

17: combined with padding of the simulation cell, approaches based on

18: the minimum image convention, and the explicit use of Open Boundary

19: Conditions (OBCs). We have implemented these approaches in the ONETEP

20: LS-DFT program and applied them to a range of systems, including a

21: polar nanorod and a protein. We compare their accuracy, complexity,

22: and rate of convergence with simulation cell size. We demonstrate

23: that corrective approaches within PBCs can acheive the OBC result

24: more efficiently and accurately than pure OBC approaches.

25: \end{abstract}

26: