A query on U, V, W peak fitting vs Fundamental Parameters approach. Is U, V, W
still taken seriously in terms of characterising diffractometers given U, V, W
type approaches seem best described as legacy code and a
throw back from the late 1960's? U, V, W is a mathematical approach that
describes no physical reality - but was good at fitting the peaks and
getting the job done of refining structures(?)
Perhaps a summary from the Size/Strain round robin/conference could
help clarify on this point if this has made it into print yet?
(i.e., if size strain is a mature science - can people get a
roughly consistent set of results and agree on the methematics/
methods used?)
Also, if U, V, W is valid at characterising the machine
contribution to the peak profile, it should automatically include the
low angle peak-assymmetry, as this is caused by the diffractometer
geometry.
----
Traditional approaches that use empirical fitting of peaks - and indirectly
draw all kinds of conclusions on physical reality does seem rather dodgy.
This compares to using an approach that directly measures the diffractometer
constants; and convolutes these into the Rietveld/Peak profiling software
(thus doing away with a requirement for running a posibly dodgy standard)
References and some plots at:
http://www.dl.ac.uk/CCP/CCP14/maths/index.html#funpar
E.g. in the case of the Koalariet Rietveld software for a Philips
diffractometer:
xdd_diffractometer_radius 173
xdd_length_of_tube_filament 12
xdd_sample_length 15
xdd_receiving_slit_length 12
xdd_receiving_slit_width .1
xdd_primary_soller_angle 5
xdd_secondary_soller_angle 5
xdd_divergence_fixed_angle 1
Thus by rigorously and scientifically convoluting the effects
of the diffractometer, you are then free to attempt to model the
effects of the sample (size, strain, etc).
A tutorial on using the fundamental parameters XFIT software
to fit peaks from a laboratory diffractometer is at:
http://www.dl.ac.uk/CCP/CCP14/tutorial/xfit-95/fun1.htm
(One obvious advantage being that low angle peak assymmetry is
automatically calculated (being caused by the diffractometer geometry)
without the requirement of refining extra assymmetry parameters)
Lachlan.
PS: Will Fundamental Parameters win over short-term kludge,
legacy empirical approaches?
Given source code for U,V,W and empirical approaches are
readily available, one should not be surprised if this is the
mainstream of Rietveld for quite some time in existing and new
Rietveld software(?). (It isn't the "best" that wins - but that
which is "available" and does the job - as Microsoft has
demonstrated most amply)
If the original Rietveld code was patented back in the 1960's,
would Hugo now be a millionaire - and us still using the Rietveld
method? Highly doubtful. Odds are Hugo would not be driving
a Rolls Royce and that the method would join the pile of
obscure and ignored scientific methods. (When it comes to the
crunch scientists are peasants and prefer the cheapest option?)
-- Lachlan M. D. CranswickCollaborative Computational Project No 14 (CCP14) for Single Crystal and Powder DiffractionDaresbury Laboratory, Warrington, WA4 4AD U.KTel: +44-1925-603703 Fax: +44-1925-603124 E-mail: l.cranswick@dl.ac.uk Ext: 3703 Room C14CCP14 Webpage (Under heavy reconstruction): http://www.dl.ac.uk/CCP/CCP14/