Sent on behalf of Antonio Cervellino (SLS), who's not in the rietveld list.
Best
Norberto
I have read quickly the paper (not attached but I found it)
It uses a single SRM and a single instrument, and a single geometry,
without reference to the accuracy of its angular calibration.
Therefore I consider it not fully credible.
One would need to compare 2-3 different certified SRMs (e.g. 660a, 640d
- not many peaks maybe - or the NAC). Considerations about
sample-dependent effects would be then assuaged. In the absence of
enough certified SRM,
powders measured also with other instruments (and at synchrotron) would do.
Also some instruments allow using both flat-plate and capillary geometry -
that would be interesting to compare too.
For reference,
our angular calibration (that we repeat with a certain frequency) is
credible to within 0.001 deg,
if you consider thermal expansion of the detector box for Delta T <~ 1 deg.
As for the patent. They use a criterion SUM(ABS(Delta(2theta))=0. This
is the 1-norm of the angle difference.
The 2-norm SQRT(SUM((Delta(2theta)**2))=0 would be as effective or more
and they are not patenting it.
Generally one can use (SUM(ABS(Delta(2theta))**p))**(1/p)=0, where p is
any real number >= 1 (this defines the p-norm). The ABS is necessary if
p is not an even integer.
So if you think that the 1-norm is better, but is patented, then use the
1.001-norm. ;-)
Antonio
____________________________________________________________
Dr. Antonio Cervellino
Beamline scientist – X04SA-MS beamline
Swiss Light Source
Paul Scherrer Institut
WLGA/229
CH-5232 Villigen PSI
Phone : +41 56 310 46 11
E-Mail: antonio.cervell...@psi.ch
<#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Please do NOT attach files to the whole list <alan.he...@neutronoptics.com>
Send commands to <lists...@ill.fr> eg: HELP as the subject with no body text
The Rietveld_L list archive is on http://www.mail-archive.com/rietveld_l@ill.fr/
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
