The Avogadro problem: summary of tests on crystal imperfections

International measurements of the molar volume of silicon on the basis of more than 15 different crystals have confirmed the Committee on Data for Science and Technology (CO-DATA) value of the Avogadro constant published in 1986. Only one discrepancy still exists and cannot yet be completely explained. In this paper, we summarize results of tests that relate to the perfection of the crystals used. With high probability, it turns out that imperfections of the crystal material itself and not the uncertainties of the measurement procedures are responsible fur this inconsistency. At present, the lack of exact knowledge of the degree of crystallographic perfection and, in particular, of the isotopic composition, hampers the progress of the crystal route for a new definition of the kilogram.