Hi Max
I have put together some macros and an INP file that approximates ADPs from a site. It uses 7 atoms with beqs for approximating/describing a single site with ADPs. The approximation is accurate and the INP file demonstrates this. Because beqs are used, the approximation can be used in PDF refinement.
Some main points when using the approximation for PDFs:
- Number of ADP parameters becomes 7 instead of 6.
- Broadening due to ADPs in G( r) is implied.
- Asymmetry at low r is also implied.
- This approach works in Version 7
The asymmetry seen at low r is typically difficult to do; this approach however implicitly contains the asymmetry.
The attached INP file has three modes of operation as follows:
- CREATE_WITH_ADPS_ creates simulated single crystal data with normal APDs for one atom.
- FIT_ALTERNATIVE_ADPS_ fits to the simulated data using the six atoms approach. This refinement then saves the calculated pattern created from the six atoms to create sim-2.hkl.
- DETERMINE_ADPs_FFROM_ALTERNATIVE_ADPS_ fits normal ADPs to sim-2.hkl. The fitted parameters from this fit is then seen to be very close to the original ADPs. Preliminary analysis indicates that this similarity is seen for a large range of ADPs.
Cheers
Alan