A new single crystal from isotopically enriched silicon was used to determine the Avogadro constant N A by the x-ray-crystal density method. The new crystal, named Si28-23Pr11, has a higher enrichment than the former 'AVO28' crystal allowing a smaller uncertainty of the molar mass determination. Again, two 1 kg spheres were manufactured from this crystal. The crystal and the spheres were measured with improved and new methods. One sphere, Si28kg01a, was measured at NMIJ and PTB with very consistent results. The other sphere, Si28kg01b, was measured only at PTB and yielded nearly the same Avogadro constant value. The mean result for both 1 kg spheres is N A = 6.022 140 526(70) × 10 23 mol −1 with a relative standard uncertainty of 1.2 × 10 −8 . This value deviates from the Avogadro value published in 2015 for the AVO28 crystal by about 3.9(2.1) × 10 −8 . Possible reasons for this difference are discussed and additional measurements are proposed.

The latest determination of the Avogadro\ud constant, carried out by counting the atoms in a pure silicon\ud crystal highly enriched in 28Si, reached the target 2 × 10−8\ud relative uncertainty required for the redefinition of the kilogram\ud based on the Planck constant. The knowledge of the isotopic\ud composition of the enriched silicon material is central; it is\ud measured by isotope dilution mass spectrometry. In this work,\ud an independent estimate of the 30Si mole fraction was obtained\ud by applying a relative measurement protocol based on\ud Instrumental Neutron Activation Analysis. The amount of 30Si\ud isotope was determined by counting the 1266.1 keV γ-photons\ud emitted during the radioactive decay of the radioisotope 31Si\ud produced via the neutron capture reaction 30Si(n,γ)31Si. The\ud x(30Si) = 1.043(19) × 10−6 mol mol−1 is consistent with the value currently adopted by the International Avogadro\ud Coordination

This paper focuses on the use of the spreadsheet technique to set up the uncertainty budget for the k0-standardisation method in Neutron Activation Analysis. The adopted measurement model included most of presently recognized error sources and was written to limit the covariances between input quantities. The calculations were implemented in a worksheet file and tested in a multi-elemental analysis of a biological material. Besides, it was demonstrated that the k0-standardisation turns to the relative-standardisation when the monitor element corresponds to the analyte element. The developed worksheet is available and suitable for the analysis of other materials in different experimental conditions.