Diffactometer requirements

Lachlan Cranswick ( lachlan@melbpc.org.au )
Sun, 28 Jan 1996 14:13:29 +1000

>Hi alls
>
> We are new users of Rietveld programs; we use it with mixed oxides of
>iron and rare earth. We have a Philips PW 1710 diffractometer with
>automatic divergence slit, a 0.2 mm receiving slit and a 1=BA scatter slit
>(we convert the ADS data to fixed divergence slit using a formula supplied
>in the Philips documentation). We use RIETAN (supplied by Dr. Izumi) and
>LS1 (supplied by Dr. Lutterotti) programmes to fit the data, but we obtain
>always very high values of the termal parameters (between 2 and 6,=20
>sometimes higher).=20
>
> I would like first the best conditions for the diffractometer work, and
>second the best conditions to the data collection. We use a step width of
>0.02=BA and a step time of 10 s between 20 and 100=BA. Sometimes we smooth
>the data to a step width of 0.04 =BA2theta. So, we work with 2500-4000 data
>points. Intensities are higher than 20 000 counts. We don't know if these=
=20
>experimental sets are alrigth for good statistics. Are they O.K.?=20

Apologies if I have missed some follow-ons and this has been duplicated
elsewhere as AXAA 96 in Sydney finished this week. We didn't like the=20
mechanical ADS, as our Rietveld software and other quantitative software=20
could not use the "raw" data, so went back to fixed slits. Thus much of=
this=20
is from memory. So if some of the following is not correct, feel free to=20
put red pen through it with another follow-on.

Shouldn't the scatter slit be 4=BA to match the maximum possible divergence
of the ADS divergence slit?

------------

One suggestion is to set the Philips mechanical ADS to emulate a fixed slit
and collect "fixed slit data" and see if the problem is still there. =20
(We also found the incident beam Sollers slits on the ADS did not collimate=
the=20
X-ray beam as well as the fixed slit incident beam Sollers).

From memory (be sure to check this out with you local Philips Technican)
this
is done by moving the goniometer to 40=BA (SAN 40). This (from memory) is
the
equivalent divergence of a 1 degree "fixed" divergence slit. Lock the ADS
by
tightening the appopriate grub screw, then disengage the ADS clutch from the
chain drive by loosening the grub screws on the clutch. Please think
carefully about what you are doing during this, else non-moving parts
will become moving parts to the detriment of the equipment - or the ADS
can go out of alignment.

>
> If we work at angles lower than 20 =BA2theta the collected intensity
>increases a lot and it is impossible trying to model the background for
>the angular region 4-20 =BA2theta. Is there any correction for this?

As we prefer simple (inherantly reliable?) fixed slit systems, others
may like to comment whether this could be due to alignment or
data conversion to fixed slit equivalent data?

---------

Answers to past queries also mentioned concerns about the mechanical=
movement
of the ADS? Perhaps time for another FAQ?
------------

Cheers,

Lachlan.
Lachlan M.D. Cranswick _--_|\ =20
Melbourne PC User Group, Australia / \
A/H Phone/Fax : (613) 9455-1345 \_.--._/
E-mail : lachlan@melbpc.org.au v
WWW : http://www.unige.ch/crystal/stxnews/stx/volnteer.htm=20