Re: [sdpd] Re: how to purchase a high resolution diffractometer

[Jon Wright <jpw22...@cam.ac.uk>]

On Wed, 11 Oct 2000, Armel Le Bail wrote:

> 
> >I should also point out that what I meant by higher angles
> >might more strictly be-calleded intermediate angles, i.e. those where
> >the alpha1,2 peak splitting is present but not resolved. Some of the
> >peaks at these sort of angles may be required for indexing and the human
> >eye can be very good (with practice) at distinguishing single from multiple
> >peaks, but less so when two wavelengths are present and all the peaks are
> >multiple by definition.
> >
> >So I beg to differ with you Armel on this point ....
> 
> Well, even if not completely resolved, the alpha-2 component is
> removable by one of the 50 or so alpha-2 correction methods 
> (almost all published in the Journal of Applied Crystallography,
> and all pretending to be the best). This is not a job for the eyes.
> Moreover, when the FWHM is as low as 0.04°(2-theta) in the
> 5-50°(2-theta) range, the alpha-2 component is detectable at angle
> as low as 20°, even by old eyes like mines. The old good Rachinger
> method removes Kalpha-2 almost perfectly, at any angle, so that you
> may obtain the peak positions exactly like if you had a primary beam
> monochromator. A Round Robin about the precision of the extracted
> peak positions either from Kalpha or Kalpha-1 should not conclude
> clearly in favour of the Kalpha-1 option, in my opinion, but you
> may organize one.
> 
> I don't see how your arguments could convince me. Our disagreement
> will continue for ever.
> 
> Best,
> 
> Armel Le Bail
> http://www.cristal.org/course/

It'd also be interesting to see if our what our eyes can see could agree
with physics and mathematics. I prefer when they do :) How do the numbers
compare if you do Pawley fits to the generated datasets and calculate the
number of independent observations (N_ind) in the pattern using Bill
David's method from J.appl.cryst? To my mind this seems like a better
measure of powder data quality. I would guess that when the alpha_1
component of a weak peak is under the alpha 2 component of a stronger peak
it's much harder to see the weaker peak. Hence N_ind will go up if you use
a monochromator and the data is "better", if it was overlapped when the 
alpha_2 was present.

Seems like it's all a question of degree of overlap. If the number of
independent intensities is not too much less than the number of
reflections in the pattern, then increasing the resolution doesn't make
much difference, as shown by Armel's example. It's only when there is a
high degree of peak overlap that you'd expect to see much of a difference
between the experimental methods. 

Jon Wright

============================================================================
         Dept. of Chemistry, Lensfield Road, Cambridge, CB2 1EW







-------------------------- eGroups Sponsor -------------------------~-~>
Restaurants, Movies, Weather, Traffic & More!
Call 1-800-555-TELL.  For more info visit:
http://click.egroups.com/1/9533/14/_/40740/_/971277880/
---------------------------------------------------------------------_->