The clear advantage of variable slit is to have more intensity at
medium/higher diffraction angles, compared to fixed slit systems adapted
to low starting angles. This is nice for quantitative Rietveld analysis
of samples having important intensities at low as well as at medium/high
angles, e.g. clay minerals (00l and 060 reflections). I personally love
variable slit measurements with our old (conventional detector)
instruments for routine quantitative Rietveld analysis, showing good
peak-background ratio and pure alpha1/alpha2 doublet (because of
graphite monochromator, traditional point detectors), combined with
"acceptable high" intensity from the variable slit.
Of course, any "data treatment" like internal conversion from variable
to fixed-slit data (as recommended/necessary in some old-fashioned
software) before Rietveld analysis should be avoided.
As David wrote, it is no big deal to model the instrumental function for
variable slits (inclusive the intensities) in a fundamental parameter
approach, allowing even a reasonable structure refinement, see
http://www.bgmn.de/vardiv.html
However, practical problems may arise at very high angles resp. very
high opening of the variable slit, if the slit divergence reaches the
magnitude of the tube take-off angle or the breadth of the
diffracted-beam monochromator. This cannot be correctly modeled by the
Monte-Carlo simulation, simply because the intensity distribution within
the broad tube beam bundle and the precise position and dimension of the
curved monochromator are unknown/uncertain. So extreme slit opening
should be avoided even with a fundamental parameter modeling.
Reinhard Kleeberg
David Lee schrieb:
The only advantage I see with a variable slit is that you keep a
constant area on the surface. This
gives a constant sampling volume for samples that are thin compared to
the x-ray penetration depth.
In contrast, the sampling volume is fixed with a thick sample and
fixed slit for typical powder samples.
I agree that the variation in resolution with angle makes Reitveld
analysis much harder. A full fundamental
parameter analysis software that includes the variable slit might work.
David Lee, Ph.D.
DTLee Scientific, llc
http://www.dtlee.com
614-562-6230
On Dec 7, 2009, at 10:03 PM, Russ Field wrote:
HI All
I am seeking opinion on the installation of automated slits on a
Panalytical MPD
Pros and cons.
Comments from a previous post are shown below by an esteemed colleague
I see little use for varable divergence slits for all sorts of
reasons (firstly as there is no rock solid conversion from automatic
to fixed intensities, secondly the resolution changes with angle with
variable slits that can't be modelled
unless a more sofisticated model than that of Highcore Plus
is used).
Regards
Russell
Russell Field BSc (Hons),
Scientific Officer,
Dept of Physical Geography,
Macquarie University,
NSW 2019
==============================
Ph: 02 9850 8341
Fax: 02 9850 8420
Mobile No. 0417 681 959
Email: rfi...@els.mq.edu.au
