Mössbauer study of oxo derivatives of iron in the Fe2O3-Na2O2 system. Pankratov D.A. //Inorganic Materials, 2014. V. 50. № 1. P. 82-89
As a rule, interpretation of poorly resolved Mössbauer spectra is an intricate, multiparametric problem. A measured spectrum is fitted using a particular model based on available information about the composition, structure, and properties of the sample. A criterion for whether the spectrum has been adequately decomposed into components is the agreement between the obtained parameters and expected ones for realistic physical and chemical models. A criterion for sufficiency is provided by statistical parameters of the deviation of results of the proposed model from experimental data. Clearly, a larger number of subspectra in a model and a smaller number of constraints for their components ensure better fitting quality. In some cases, when the iron atoms in a sample have an inhomogeneous environment and parameters of subspectra vary continuously in some range, or spectra contain overlapping components (resonance lines), a model-free description of spectra is used, instead of models. However, a model-free description of spectra also requires an adequate initial hypothesis as to the properties of the substance to be studied. In some cases, even with well-resolved spectra, an incorrect initial hypothesis may entail inadequate interpretation of experimental data.