Dear Roger (& others),
Assuming you are using Bragg-Brentano geometry - there is no absorption 
correction needed. However, there may be a surface roughness effect if your 
sample is very strongly absorbing. GSAS does have a correction for this if 
needed. Unusually small Uiso's are the main hint that there is a problem. 
Don't invoke surface roughness until all done with everything else. There 
will be very little change in the fit but the Uiso's will be more 
reasonable. It's a pretty touchy refinement so take care with it. On the 
other hand if you are using capillary geometry (synchrotron - which you 
apparently are) then enter muR/lam for the GSAS absorption coefficient. 
Don't refine it though; it 100% correlates with Uiso.
Bob Von Dreele
At 08:16 PM 2/17/00 -0500, you wrote:

>         Should I concern myself about X-ray absorption in Rietveld refinement
>and sturcture solution from powder data?  If so, is there way to input mass
>attenuation coefficients into GSAS?  I have read page 123 of the GSAS 
>TECHNICAL
>MANUAL and seems to say no.  The Debye-Scherrer absorption function is
>suggested only for "a weak absorber" and seems to be an arbitrary function
>derived from neutron diffraction.  The other functions are inappropriate 
>for my
>experimental conditions.  Should I just refine thermal parameters, which are
>presumed highly correlated with absorption, and not worry about it?
>
>I am working with a strongly absrobing material, CuAlI4 in a Debye-Scherrer
>geometry.  I have used mucalc, on Brian Toby's NIST website
>
>http://www.ncnr.nist.gov/programs/crystallography/mucalc.html
>
>as earlier suggested by Lachlan Cranwick on this list.
>
>Using the crystalographic density,  Packed density = 6.673 (g/cc)
>Wavelength = 0.9388 (A)
>
>mu = 497.8 cm-1
>
>Using 0.6 the crystalographic density,  Packed density = 4.003 (g/cc)
>Wavelength = 0.9388 (A)
>
>mu = 298.6 cm-1
>
>Thank you for any experience in this matter.
>
>Roger
>
>
>Roger Sullivan
>Department of Chemistry, Box 8204
>North Carolina State University
>Raleigh, NC 27695
>
>[EMAIL PROTECTED]

Reply via email to