Matteo,
The error in the LST file is a statistical error that doesn't include
systematic errors.
Probably the largest of these systematic errors is an underestimation of
the weight fraction of a highly absorbing phase, for instance if you were
using copper radiation and had a first-row transition metal in your sample.
In this situation the absorbing phase can be significantly underestimated
(several percent or higher errors are not uncommon). (Extinction in one or
more of the phases can also cause errors.)
Fortunately there is a nice trick to help with these problems. Add, for
example, 40 weight percent of CaF2 (or another simple material with low
absorption) to your unknown sample. Refine the weight fraction of CaF2.
If you get significantly less that 40 weight percent CaF2, the "missing
mass" can be accounted for by adding mass to the absorbing phase.
In cases I have tried, I also had density measurements from a pycnometer.
I found this to be a very useful independent check of the data.
I hope that helps.
- Kurt Leinenweber
At 06:41 PM 11/9/99 +0100, you wrote:
>
>Hello everybody!
>I am a new reader of this mailing list and an enthusiastic user of GSAS on
>XRPD though not still expert. I have the problem of estimating the wt%
>error of components of a multiphase mixture (that I made). Does anybody
>know how to calculate this error from the data given in the *.LST file? It
>seems that the error I get is too small (!), what could it be responsible
>for it? Any suggestion or information about literature on the issue?
>Thanks
>
>Matteo Zoppi
>[EMAIL PROTECTED]
>
>
>
>
*************************************
Kurt Leinenweber
Dept. of Chemistry
Arizona State University
Tempe, AZ 85287-1604
Phone (480)-965-8853
Fax (480)-965-0474
************************************
