Recently, a new method (MXAN) of extracting the local structural information available in the XANES spectra has been developed in the framework of the multiple scattering theory and successfully applied to the analysis of several system, both in solid and liquid state. In this paper we show the progress we have done in the MXAN procedure, in particular for what concerns the phenomenological broadening and the electronic charge fitting. We have also implemented parallelization on MXAN code by MPI library for both energy points and annealing for no-structure parameters. This approach reduce the computer time of a factor approximate to 10 (our cluster contains 14 nodes but the computer times scales with the number of nodes) and at the same time increase the accuracy of the method. The new version of the program will be discussed on the basis of several examples.
Progresses in the MXAN Fitting Procedure / K., Hayakawa; K., Hatada; S., DELLA LONGA; D'Angelo, Paola; M., Benfatto. - In: AIP CONFERENCE PROCEEDINGS. - ISSN 0094-243X. - STAMPA. - 882:(2007), pp. 111-113. (Intervento presentato al convegno X-RAY ABSORPTION FINE STRUCTURE - XAFS13: 13th International Conference tenutosi a Stanford, CA; United States nel 9 July 2006 through 14 July 2006) [10.1063/1.2644444].
Progresses in the MXAN Fitting Procedure
D'ANGELO, Paola;
2007
Abstract
Recently, a new method (MXAN) of extracting the local structural information available in the XANES spectra has been developed in the framework of the multiple scattering theory and successfully applied to the analysis of several system, both in solid and liquid state. In this paper we show the progress we have done in the MXAN procedure, in particular for what concerns the phenomenological broadening and the electronic charge fitting. We have also implemented parallelization on MXAN code by MPI library for both energy points and annealing for no-structure parameters. This approach reduce the computer time of a factor approximate to 10 (our cluster contains 14 nodes but the computer times scales with the number of nodes) and at the same time increase the accuracy of the method. The new version of the program will be discussed on the basis of several examples.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
