Qualidade do refinamento do método de Rietveld em amostras de solo

• Ano de defesa: 2012
• Autor(a) principal: Leite, Wellington Claiton
• Orientador(a): Saab, Sérgio da Costa
• Banca de defesa: Brinatti, André Maurício , Antonio, Selma Gutierrez
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: UNIVERSIDADE ESTADUAL DE PONTA GROSSA
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciências
• Departamento: Fisica
• País: BR
• Palavras-chave em Português: difração de Raios X; fluorescência de Raios X; análise multivariada; fração silte do solo
• Palavras-chave em Inglês: X-Ray Difraction; X-Ray Fluorescence and Multivariate Analysis; silt soil.
• Área do conhecimento CNPq: CNPQ::CIENCIAS EXATAS E DA TERRA::FISICA
• Link de acesso: http://tede2.uepg.br/jspui/handle/prefix/898

The Rietveld Method (RM) is a tool commonly used in the study of polycrystalline material.This method coupled with data from X-ray disraction (XRD) has been shown to be esective in the qualitative and quantitative study of various samples. In the literature there are several computer programs that perform the algorithm of the method, however, each program has implemented diferent functions its the data, including the profile function. One issue to be considered in the refinements MR structures are proposed for the XRD pattern, appropriate structural models are desirable for the correct description of the total difraction pattern, more realistically describing the study sample. With regard to the quality of settings, using indices of disagreements (index R), these are used to evaluate the quality of the final refinement as well as their convergence. However,its strong dependence on the quality of equipment used and the characteristics of sample preparation, these become the major indexes (index S) doubtful about the quality of refinement. The objective of this study was to compare the results of the refinement program GSAS and DBWS, from these results to compare the crystal structure of the adjustments made by the two programs and structural adequacy of the samples studied. And also propose a new way of assessing the quality of refinement, independent of data collection for soil samples. Samples were characterized from silt (particle size between 2-20 cm) of a Dystrophic Red Latosol submitted to management systems: conventional tillage (CT), no-tillage (NT) and minimum tillage (MT). The XRD revealed the presence of the minerals gibbsite, quartz, kaolinite, halloysite, goethite, anatase, hematite, muscovite and vermiculite. For quantification by RM phase, the programs agree to the majority phase. In DBWS was not possible to refine the structures completely, the rates of disagreement values converged to unsatisfactory. The profile function used in this program, failed to adjust the enlargements of peaks of phases at low concentration and presents a good fit of the majority phases. In GSAS, it was possible to refine all variables. For the GSAS refinement of the levels of disagreement were compared to R indices generated by DBWS in GSAS these indices were more satisfactory. From this information the diference curve was analyzed, it was found that it follows a Gaussian distribution. The mean and fwhm of the Gaussian distribution, were used as parameters for evaluating the quality of refinement. The technique of X-ray fluorescence indicated the presence of oxides of Al, Si, Fe, Ti, K, Mg, Mn, P, Ca and Na in order of decreasing oncentration. Calculations performed concentration of oxides showed that among the results of the GSAS and DBWS compared with XRF show that the GSAS presents the results more similar than FRX. A tool used in the statistical analysis of multivariate data is the Principal Components Analysis, it is able to check the degree of correlation between samples, using this technique was verified that the samples differ due to their chemical composition.