- Conference date: 15–18 September 2009
- Location: Karlsruhe (Germany)
Recently a considerable interest has been triggered in the investigation of the composition of individual particles by X‐ray fluorescence microanalysis. The sources of these micro‐samples are mostly diversified. These samples come from space dust, air and ash, soil as well as environment and take the shape of a sphere or an oval. In analysis this kind of samples the geometrical effects caused by different sizes and shapes influence on accuracy of results. This fact arises from the matrix effect. For these samples it is not possible to find analytically a solution of equation taking into account an absorption of X‐rays. Hence, a way out is to approximate the real sample shape with the other one or to use Monte Carlo (MC) simulation method. In current work authors utilized the iterative MC simulation to assess an elemental percentage of individual particles. The set of glass micro‐spheres, made of NIST K3089 material of known chemical composition, with diameters in the range between 25 and 45 μm was investigated. The microspheres were scanned with X‐ray tube primary radiation. Results of MC simulation were compared with these of some analytical approaches based on particle shape approximation. An investigation showed that the low‐Z elements (Si, Ca, Ti) were the most sensitive on changes of particle shape and sizes. For high‐Z elements (Fe—Pb) concentrations were nearly equal regardless of method used. However, for the all elements considered, results of MC simulation were more accurate then these of analytical relationships taken into comparison.
Data & Media loading...
Article metrics loading...