A critical component for understanding the properties of materials, and enabling the development of new materials is the ability to characterize them in detail as well as understand why they form. While this may appear to be a combination of two disparate concepts, in many respects they are not and should be considered as synergistic. Approached from the characterization side, better tools allow one to answer more fundamental scientific questions about why a particular structure is formed. Approached from the other side, the underlying scientific questions can drive what types of characterization is needed. Frequently the underlying science can be best revealed by theoretical calculations, particularly density functional calculations which despite some limitations can probe many important questions.

On particular area of interest is surface structures. It has now become almost conventional when proposing a model for a surface reconstruction to perform a density functional theory (DFT) calculation. What one wants to know is whether the proposed positions are plausible, i.e. the difference between them and refined DFT positions is not too large, as well as whether energetically the structure is plausible. For this one needs to have answered three fundamental questions:

• What is the most appropriate DFT functional to use; there are many in the literature.

• What are the errors in the energies? These numbers are rarely analyzed or published and from an experimental viewpoint a measurement without errors is marginal. Obviously only with knowledge of the errors in the energies can one determine if a structure is plausible.

• What are the basic simple structures against which one wants to compare a reconstruction? Since often reconstructions are variants/superstructures based upon simple 1x1 units, this information is also needed a-priori to aid in solving reconstructions, particularly ones with large unit cells.

• Do the theoretical calculations agreement between experimental and theoretical results for surface structures, particularly oxides, and in many cases they do not. Are the surface structures obtained experimentally kinetically metastable which, if annealed for long enough, would transform? Are there fundamental problems with the theoretical calculations?

1. The Small Unit Cell Reconstructions of SrTiO3 (111).
   L. D. Marks, A. N. Chiaramonti, F. Tran and P. Blaha, Surface Science, 603, 2179 (2009)
   
2. Water-driven structural evolution of the polar MgO (111) surface: An integrated experimental and theoretical approach
   J. Ciston, A. Suramanian, L. D. Marks, Physical Review B, 79, 085421 (2009)
   
3. Diffraction refinement of localized antibonding at the Si (111) 7x7 surface
   J. Ciston, A. Suramanian, I. K. Robinson, and L. D. Marks, Physical Review B, 79, 193302 (2009)
   
4. Force calculation for orbital-dependent potentials with FP-(L)APW+lo basis sets
   F. Tran, J. Kunes, P. Novák, P. Blaha, L.D. Marks, and K. Schwarz, Computer Physics Communications  179: p.784-790, 2008