Different ways of calculating Coulomb and dipole sums over crystal lattices are analyzed comparatively. It is shown that the currently alleged disagreement between various approaches originates in ignoring the requirement for the self-consistency of surface conditions, which are of fundamental importance due to the long-range nature of the bulk interactions that these sums describe. This is especially true of surfaces arising when direct sums for infinite translation-invariant structures are truncated. The charge conditions for actual surfaces being self-consistently adjusted to the bulk state are formally the same as those on the truncation surface, consistent with the concept of the thermodynamic limit for the bulk-state absolute equilibrium and with the fact that the surface energy contribution in this case is, naturally, statistically small compared to the bulk contribution. Two-point multipole expansions are briefly discussed, and the problems associated with the boundary of their convergence circle are pointed out.

PACS: 41.20.Cv, 61.50.Lt, 75.10.−b, 81.40.Rs (all) DOI: 10.1070/PU2004v047n10ABEH001720 URL: https://ufn.ru/en/articles/2004/10/a/ 000226730200001 2004PhyU...47..965K Citation: Kholopov E V "Convergence problems of Coulomb and multipole sums in crystals" Phys. Usp. 47 965–990 (2004) BibTexBibNote ® (generic)BibNote ® (RIS)MedlineRefWorks Оригинал: Холопов Е В «Проблемы сходимости кулоновских и мультипольных сумм в кристаллах» УФН 174 1033–1060 (2004); DOI: 10.3367/UFNr.0174.200410a.1033