Theoretical maps of work-function anisotropies

Work functions of stepped metallic surfaces are modeled to generate work-function anisotropy maps. We show how the work function of any stepped surface can be accurately predicted by interpolating between the work functions of a small number of low-index facets using a set of physically motivated symmetry-adapted basis functions. This technique is applied to the work-function anisotropy of tungsten, where we study the W(110), W(100), W(211), W(310), W(111), and W(321) surfaces from first principles. The subsequently modeled work-function anisotropy map is found to be in excellent agreement with recent experimental maps over the full range of surface orientations.