Hi Hana,

Hi list, I'm looking for a way to make LCF with standards that were measured at different occasions and beam-lines. I did read previous comments regarding this issue, but still need to clarify:

There is no a priori reason why you cannot do so. Two beamlines that use the same mono crystal should measure the same data on the same sample, within measurement uncertainty, of course. The problem is that differences betwee the beamlines (differences in resolution, differences in detector linearity, different gremlin species mucking up the works, etc) result in systematic uncertainty in your results. As long as you are mindful of this systematic uncertainty and honest about reporting it, there is no reason that LCF cannot be done effectively.

1. Does limiting the range to the smallest range help in this case (if spectra were measured with the same steps size, and a calibration standard was used)?

Well, putting the data on a common energy grid is certainly essential. If you cannot do so (using a calibration standard or otherwise), then you cannot do LCF. As for limiting the data range, I am not clear how that helps. If the problem is different resolutions, then any data range of any size that includes the edge will be effected. I suppose that if one of the beamline has a horrible problem with runout error, then limiting the data range might help. But does anyone's beamline really have that problem?

2. For harder situation; when different steps were used - I found that sixpack (according to its manual) interpolating all spectra to the same energy grid - does this solve the problem?

I can't comment well on Sixpack, but I will say that Athena always interpolates regardless of the provenance of the data. It is unclear to me how the fit could be constructed in a numerical sense if the data are not interpolated onto the same grid. B -- Bruce Ravel ------------ bruceravel1@gmail.com Homepage --------------- http://xafs.org/BruceRavel EXAFS software --------- http://github.com/bruceravel/demeter/