Dear Dip Das, usually a linear background is sufficient for pre-edge region, but you may also use a polynom of a higher degree. However, in the latter case, the extrapolation to the post-edge region may cause problems, especially if the energy range is large. I would strongly recomment to use free available software such as the Athena/Artemis/Demeter packages for EXAFS and XANES data analysis. Kind regards, Dirk
Taking the Victoreen coefficients from the International Tables of x-ray Crystallography I had good approximation of the energy dependence of the attenuation coefficient u(E).
Intern. Tables of X-Ray Crystallography Vol. III. Tables 3.2.2C.
Or just fitting by u(E)= a/E3 +b/E4. So you get easily rid of ugly scaling factors. If the pre-edge energy range is comparable to the EXAFS / XANES range the extrapolation into the range above the edge is no problem (to my experience).
But for the XANES energy range only, which is small compared to the EXAFS range, I think Dirk is right. Just take a linear approximation.