Assessment of the data repeatability of x-ray diffraction study for silicon nitride powders of different dispersion

Abstract

X-ray diffraction methods are indispensable for studying crystalline materials and provide determination of the phase composition, internal stresses and the parameters of unit cells of crystalline phases, as well as the crystalline size, which is related to the average grain size and affects the width of the diffraction peaks. We present the results of evaluating the repeatability of the X-ray diffraction data obtained for silicon nitride powders Si 3 N 4 of different particle size and different initial phase ratios α-, β-Si 3 N 4 . The relative error of measuring the diffraction peaks was determined using the intensity detector and the absolute error of calculating the unit cell parameters was determined by the Rietveld method. The average particle size of the initial powders was analyzed using scanning electron microscopy. X-ray diffraction studies were performed according to the Bragg – Brentano scheme using CuKα radiation (λ = 1.5406 Å). A series of three experiments was performed for each powder. It is shown that the relative error of intensity measurements with a detector does not exceed 2% for peaks corresponding to 3σ criterion, and the absolute error of the determination of the unit cell phase parameters by the Rietveld method is 0.001 Å. The results obtained can be used to assess the stability of the diffractometer both for the samples based on silicon nitride and materials of different composition and structure, especially for submicron samples. In the latter case, the error of the parameters obtained by X-ray phase analysis can be taken into account without resorting to statistical estimates, as in the method of least squares.