[Application of wavelet transform to the isolation of EXAFS oscilations from experiment data]

It is of significant importance to isolate lambda(k) form the total absorption cofficient data micro(E) for the analysis of EXAFS spectrum. Generally, least-squares procedure of polynomial spline or B spline was used for removing the smooth sbsorption background, which has been proved to be effective. But it suffers in that the trial of knot points is time-consuming and it is poor in reproducibility. In this paper, a new chemometrics method--wavelet transform analysis was used to separate the lambda(k) and background absorption from micro(E) and it is proved that the technique is a very good method for the analysis of EXAFS spectrum.