existing powder diffraction software to help battle ever more complex crystallographic problems. L. M. D. Cranswick, CCP14 (Collaborative Computation Project No 14 for Single Crystal and Powder Diffraction) Department of Crystallography; Birkbeck College, University of London, Malet Street, Bloomsbury, London, WC1E 7HX, UK. E-mail: [email protected]
88
Embed
SIG_9.1: Crystallographic Computing: Advanced Computational Tools for Small Molecules and Materials (Monday, 27th August 2001) New features in existing.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
SIG_9.1: Crystallographic Computing: Advanced Computational Tools for Small Molecules and Materials (Monday, 27th August 2001)
New features in existing powder diffraction software to help battle ever more complex crystallographic problems.
L. M. D. Cranswick, CCP14 (Collaborative Computation Project No 14 for
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 2
Talk Aims• Show that within a year there have been
developments in many freely available powder diffraction programs to enable the tacking of more complex / difficult / tedious powder diffraction and crystallographic problems.
• Demonstrate how some of this functionality can make the difference and possibly help put your intractable problems from the “intractable bin” into the “doable folder”.
• Show that what might be considered superficial software modifications can make a major impact to assist in solving intractable problems.
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 3
Notes Free Zone - they are on the web
http://www.ccp14.ac.uk/poster-talks/ecm2001/
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 4
A snapshot of programs(can’t do justice to all the updates in freely available software)
• Chekcell Powder Indexing Tool
• Rietica Rietveld for mass Le Bail fitting to obtaining cell volumes and follow reactions
• GSAS Rietveld for restrained refinement of large inorganic polymeric framework structrures
• Platon’s Addsym for finding higher symmetry in polymeric inorganics - from a) initial hand built models and b) Rietveld refined structures
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 5
New features in Chekcell and Rietica • Chekcell
– http://www.ccp14.ac.uk/tutorial/lmgp/
– New: Incorporates Ton Spek and A. Meetsma’s Le Page *** (the new addition that can really count) sub-cell / super-cell searching
– New: GUI Cell transformation
– New: Misc bug fixes improving usability and reliability
• Rietica Rietveld – http://www.rietica.org/
– New: Easy to use mass Le Bail fitting of angular and energy dispersive data
– All files are ASCII Files
– Handles alpha 1 /alpha 2 (if relevant)
– Flexible : can manually edit Le Bail HKL files to add or remove HKLs and change intensities of individual HKLs
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 6
New features in Chekcell and Rietica
• Why Bother?
• Project headed by Professor Dave Walker of the Lamont-Doherty Earth Observatory of Columbia University, New York, USA
• Use energy dispersive X-ray diffraction; and high pressure / high temperature phase transitions to help determine the volume of Oxygen at high pressure and temperature (~550°C and 2 to 9 GPa). Then see if the volume this has interesting implications for Oxygen being involved in the Earth’s Outer Core
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 7
New Features in Chekcell and Rietica:
What do we need from the diffraction data?Accurate Unit Cell volumes to
obtain the “equations of state” of the phases of interest (how the volume of the phases change with pressure and temperature)
T - RbClO4 B2-RbCl + 2O2
4 R-KClO3 = 3 O-KClO4 + B2-KCl
O-KClO4 = B2-KCl + 2 O2
O2 from the differences in the unit cell volumes
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 8
Problems that need to be overcome by (Crysfire); Chekcell and Rietica
• Energy Dispersive Diffraction setup and calibration can be very ad-hoc and problematic
• Possible Detector instability over short time spans
• Possible long term non-linearity of the detector
• (more neurotic of these problems only obvious when doing whole pattern Le Bail fitting)
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 9
Problems that need to be overcome by (Crysfire); Chekcell and Rietica
• Poor resolution data with multiple sources of spurious peaks
• Phase transitions give unknown cells
• Got to track how diffraction patterns are changing through the experiments
Experimental configuration of the “Walker Cell”
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 10
Problems that need to be overcome by (XFIT, Crysfire,
Refcel, UNITCELL); Chekcell and Rietica(No Struggle - No Joy!)
• Overlapping multi-phase powder patterns
• Large amounts of raw diffraction data! (collecting 3 patterns each 5 to 300 seconds)
• Intensities are near meaningless– no incident intensity spectrum
– particle statistics problems
– preferred orientation problems
– X-ray absorption
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 11
• Graphical Interpretation of powder indexing results• Full Windows Graphical User Interface (GUI)
• Automatic Cell searching and spacegroup assignment
• Sorting lists of trial cells based on several criteria
• Reads the following raw powder diffraction file formats:
– Bruker RAW, Philips RD, RIET7 and CPI
• Reads the following peak files:
– Bruker DIF, Philip DI, XFIT TXT, Winfit DAT, Column format, Crysfire CDT
• Reads Crysfire Summary SUM files. As well as Crysfire summary files produced for the individual indexing programs:
– dicvol, ito, treor, taup, lzon, fjzn, kohl
Chekcell - Jean Laugier and Bernard Bochuhttp://www.ccp14.ac.uk/tutorial/lmgp/
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 12
Open a data file
(optional if you only have a peak listing)
Routine operation of Chekcell (1 of 5)
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 13
Routine operation of Chekcell (2 of 5)
Open the peak listing
(In this case generated by the XFIT program)
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 14
Routine operation of Chekcell (3 of 5)
Import the Crysfire summary file listing of found trial cells.
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 15
Routine operation of Chekcell (4 of 5)
Play with the list of trial cells and spacegroups and hopefully obtain a good cell that is also the “true” cell.
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 16
Routine operation of Chekcell (4 of 5)
•Automatic Cell and Spacegroup searching
– can trudge through a single selected unitcell; or over 1000s of trial cells looking for the best cell and spacegroup combination based on parsimony of extra reflections criteria.
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 17
Chekcell: Major new feature of Chekcell Porting and “integration” of Ton Spek and A.
Meetsma’s Le Page
•Obtaining the Reduced Cell – which many powder indexing
•Efficient Sub-cell and super-cell searching, then easy reviewing of newly derived cells within the Chekcell interface
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 18
Chekcell: GUI Cell transformation•Easily transform cells and test them withing Chekcell •Knows about common transformations•Can manually look at sub-cells and super-cells
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 19
Chekcell: result of using Le Page
An example:
•Orthorhombic cell with good FOM (Figure of Merit)
•Le Page combined with automatic “Best Solution” easily finds a better hexagonal cell based on parsimony of extra reflections criteria
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 20
1. Chekcell: indexing unknown cells from unexpected phase transitions in high pressure experiments
• While quality of Energy Dispersive diffraction data is low: indexing is doable thanks to Crysfire and Chekcell.
• Due to LePage, can have more confidence in finding a good cell; and checking for other sub-cells and super-cells.
• In this screen image, a new monoclinic cell has been found
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 21
2. Chekcell: re-indexing lost cells (maybe transformed) from big pressure jumps in high pressure experiments
due to racing a synchrotron beam-dump.
• Can also use “expected volumes” as a guide in refinding cells.
• Re-found a monoclinic cell despite I/I20 = 19 (19 out of the 20 first peaks were indexed)
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 22
3. Chekcell: re-index again (big jumps due to trying to beat a synchrotron beam dump)
• Again, re-found a monoclinic cell despite I/I20 = 16! (only 16 out of the first 20 peaks were indexed)
• (With 23 starting peaks)
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 23
Chekcell / LePage Summary:
•These programs give you the maximum chance of indexing unknown unit-cells from powder diffraction data, even of low data quality.
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 24
Rietica Rietveld for Mass Le Bail fitting to get cell vol.Example of 3 phase setup : KClO3; KClO4, B2-KCl
• Easy to use and setup via GUI
• Le Bail is Structureless whole profile fitting - just need cell and spacegroup
• Easy to add and delete structures
• All files are ASCII files (Data, HKL and input file) which can edited manually if required or convenient
• Auto-marquardt damping for initial unstable refinement if required
• Impurities can be very obvious when Le Bail fitting
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 25
Rietica Rietveld for Mass Le Bail fitting : Rietica Database of Structure
• Can store structures and Le Bail derived unit cells at various pressures for later retrieval
• Add to the database at the click of a button
• Select from Database option with the Phase dialog box.
• When having 100s to 1000s of datasets, ease and speed are very important.
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 26
Rietica Rietveld for Mass Le Bail fittingAt simplest: 3 step process after initial setup has been done
• Some beamlines give the option of converting into 2-theta space or refining native in KeV
• (2-theta can be convenient for using search match and related software)
• Le Bailing Sequence:1. Copy over INP and HKL file (using windows explorer)
2. Perform whole profile LB fit
3. Access / plot results - (check need to add or delete phases)
4. Repeat above
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 27
Rietica Rietveld for Mass Le Bail fitting 1 of 10
• Before the phase transition
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 28
Rietica Rietveld for Mass Le Bail fitting2 of 10
• In the phase transition
• No completely freestanding peak for KCl
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 29
Rietica Rietveld for Mass Le Bail fitting3 of 10
• Repeat as required
• No completely freestanding peak for KCl
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 30
Rietica Rietveld for Mass Le Bail fitting4 of 10
• Repeat as required
• No completely freestanding peak for KCl
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 31
Rietica Rietveld for Mass Le Bail fitting5 of 10
• Repeat as required
• Still no completely freestanding peak for KCl
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 32
Rietica Rietveld for Mass Le Bail fitting6 of 10
• Repeat as required
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 33
Rietica Rietveld for Mass Le Bail fitting7 of 10
• Repeat as required
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 34
Rietica Rietveld for Mass Le Bail fitting8 of 10
• Repeat as required
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 35
Rietica Rietveld for Mass Le Bail fitting9 of 10
• Repeat as required
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 36
Rietica Rietveld for Mass Le Bail fitting10 of 10
•Done!!!
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 37
Rietica Rietveld for Mass Le Bail :Graphing up the resultsUsing Le Bail fitting Using Traditional Methods
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 38
New Rietica Macro Language - RIET BASIC
• In theory, mass Le Bailing of the easy parts of a temperature / pressure run can be done in a fully automatic mode
• On present set of EDX data, have found it is best to not refine too quickly and instead continually check individual results and Le Bail fits.
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 39
1. Rietica and Le Bailing : Short-term non-linearities of an Energy Dispersive Detector(?)
• NaCl / Halite at Room Temperature and Pressure (thus is not crystal strain)
• Very nasty miss-fits on peak positions
• Possible short term non-linearity in the detector.
• Possible instability in the detector
• Combination of both - and more?
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 40
2. Rietica and Le Bailing : Short-term non-linearities of an Energy Dispersive Detector(?)
Le Bail fittingNon-unit weighting of reflections
Isotherm data of different temperatures colliding
Inappropriate to use Le Bail fitting on this data
Though Le Bail can detect these problems on the beam-line!!
Traditional Unitcell refinement Unit weighting of reflections
(over a wide KeV range - the data is “on average” linear)
Isotherm data no longer overlapping
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 41
Chekcell and Rietica: results of the volume of Oxygen and the earth’s other core• Energy Dispersive XRD: vol O2 ~10
cc/mol
• Established vol O2 from shockwave experiments: ~15 cc/mol (50% difference)– 15 cc/mol means there cannot be
Oxygen
– 10 cc/mol means there can be Oxygen and has implications for possible transport mechanisms between the lower mantle and outer core.
• Shockwave / Molecular Dynamics / Impulsively Stimulated Scattering Measurements (~15 cc/mol) vs Energy Dispersive Data (~10 cc/mol) beyond the scope of this talk
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 42
GSAS for large “restraint” Rietveld refinements of polymeric inorganics
• by Bob von Dreele and Alan Larsen - ftp://ftp.lanl.gov/public/gsas/
• EXPGUI GSAS Graphical user interface by Brian Toby
• Structure Solution: Ian Grey and Gus Mumme (CSIRO, Melbourne)
– Atom Ordering: Gus Mumme and Ian Grey (CSIRO, Melbourne)
– Tilt Model: Ian Grey and Gus Mumme (CSIRO, Melbourne)
– Basic structure from chemical crystallographic arguments
– Starting atom ordering model from Shelxl and CSRIET structure refinement
– (Complex tilt model obtained from refining the perovskite subcell and looking at the Fourier difference map for peaks giving an indication of tilt directions and magnitudes)
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 57
Setting up a Restrained Rietveld refinement of > 300 atom polymeric inorganic structures in GSAS
• Initial problem is setting everything up and the control of the refinement.
• But need the flexibility to change as you may only find new information while refining the structure. (change of spacegroup being the most obvious) – Eric Dowty’s Cryscon can be very quick and effective for transforming
structures from one cell and spacegroup to another. http://www.shapesoftware.com/#anchor_cryscon
• One solution is to use a spreadsheet program that allows for flexibility and extendability (MS-EXCEL for Windows)
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 58
Restrained refinement in GSAS: setting up an EXCEL spreadsheet
• CCP14 based tutorial and example files at : http://www.ccp14.ac.uk/solution/gsas/restrained_inorganic.html
• Create GSAS script commands in EXCEL and save into txt MACRO files for importing into GSAS.
– (if applicable - model in arbitrary co-ordinates)
– Structure co-ordinates (if applicable - generated from previous model)
– Crystals (Watkin, Cooper, et al) structure co-ordinates for generating bond restraint lists for conversion by Scott Belmonte’s “coue” software into GSAS macro format.
– Total cell contents and/or charge balance constraints
– If applicable, other scripts for controlling GSAS
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 59
GSAS: setting up an EXCEL spreadsheet: Example Atom Co-ordinates
• Calculate or import an atom list.
• Inputting the atom list via EXPEDT
– k l a
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 60
GSAS: setting up an EXCEL spreadsheet: Input for Crystals to generate GSAS bond restraints lists
• Just use EXCEL “=“ commands to copy over values from GSAS atom list
• Got to keep the atom numbers the same as GSAS atoms controls are based on the atom numbers
• Inputting bond length restraints in EXPEDT:
– k l s d
• Again more details given at:http://www.ccp14.ac.uk/solution/gsas/
restrained_inorganic.html
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 61
GSAS: setting up an EXCEL spreadsheet: Dual Occupancy Atom Constraints (in expedt: k l a k)
• Just copy and paste to extend the atom constraints to the desired number of atom sites. In this example for 108 atoms which would be painful to set up manually in GSAS.
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 62
GSAS: setting up an EXCEL spreadsheet: Total Cell Content Restraints (in expedt: k l s c)
• Extend as required depending on the number of atoms you wish to link into the constraint.
• The total number of atoms on the defined sites in this example is restrainted to 81
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 63
GSAS: setting up an EXCEL spreadsheet: Charge Balance Restraints (in expedt: k l s c)
• Just use the Total Cell Contents Restraint which does not care what the value physically represent.
• To enforce charge balance just put the ionization value of each atom site - which should sum to zero.
• Simple example in the screen dump.
• The Total Chemistry Restraint will performs the sum calculation for you and tell you what the present sum is in the restraints menu.
• Even if you don’t use the restraint, inputting this can be a useful, quick check that you have charge balance.
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 64
GSAS: importing all these macro files into GSAS
• Copy the Excel cells to a TXT file.
• Use an Editor such as PFE to replace all TABS with a space:
• Use the @r - import macro file command while in the relevant menu
• Done!
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 65
GSAS: turning off /on histograms, restraints, atoms
• Easily done in GSAS’s EXPEDT
– k l s
– u - The use command
• e.g., turn off X-ray data (histogram 1), turn on bond restraints (histogram 2) and neutron data (histogram 3)
k l s ! Go into EXPEDT restraints menu
u 1 ! Use histogram 1 - X-ray Data
n ! No
u 2 ! Use histogram 2 - Bond Restraints
y ! Yes
u 3 ! Use histogram 3 - Neutron Data
y ! Yes
x x x ! Exit EXPEDIT menus and quit
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 66
GSAS: using restraints as diagnostics• Restraints don’t have to be enabled to exist inside GSAS
• Restraints can also be used diagnostically even if they are not enabled
• E.g., Charge Balance :
• Following screen dumps shows that the occupancies of the various metals are such that the total cell is out of charge balance.
– In EXPEDT: k l s c l ! last l is for list
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 67
GSAS: spiraling in on the minimum and final structure
• May have to refine in quite a spiral manner
• Initially - have found circular refinement to be more controllable than combined refinement to be more controllable
– heavy atoms using X-ray data
– Oxygens using neutron data
• Restraints - problem with cemented structure vs having the structure fall apart– If restraints cement the structure, turning off restraints; doing small numbers of
cycles and manually resetting errant atoms can get things rolling quite nicely.
– Relaxing restraints sometimes can improve the regular co-ordination. !?
• If structure goes haywire, it can be easier to start again from the initial model
• Wall paper bubble effect - tighten one errant atom causes 1 to 4 more to go out
• Interested in other people’s experiences here for large polymeric structures
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 68
GSAS: lessening the tedium: batch file control of the refinement :
• can be better to look after set periods of time and reset wandering atoms - which can wander even if constrained!?
– Following examples for DOS / Windows (Linux / Unix is more flexible)
– The @m command can be used to log a command sequence to a file to server as the starting point for editing a custom macro file.
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 69
GSAS: lessen the tedium: batch file control of the refinement : (EXPEDT Example: script to increase the refined angular range)
• In this case setting up histogram 1 (powder XRD)
• File titled 70_2t_x.txt
• The BAT control file would then enabled (“use”) all relevant histograms and run powpref after this file before going through a genles run.
p ! powder setup
h ! histogram setup
e 1 ! edit histogram 1 (X-ray)
t ! set end of two theta
70 ! set 70 degrees
/ ! confirm this then exit
x
x
x
x
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 70
GSAS: lessen the tedium: batch file control of the refinement : (EXPEDT Example: changing the bond restraint weighting)
• File titled bondr10.txt
k ! don't make a backup of the file
l ! least squares
s ! soft constraints
d ! bond length restraints
f 10 ! set bond restr. weighting to 10
x ! exit expedt
x
x
x
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 71
GSAS: lessen the tedium: batch file control of the refinement :(EXPEDT Example: only refining on neutron data)
• File titled neuonly.txt
k
l ! least squares
s ! soft restraints menu
u 1 ! use histogram 1 - X-rays
n ! set to no
u 3 ! use histogram 3 - neutron
y ! set to yes
x
x
x
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 72
GSAS: lessen the tedium: batch file control of the refinement :(EXPEDT Example: refine heavy metal framework atoms)
• File titled
ref_fmh.txt
k ! don't make a backup of the file
l ! least squares
a ! atom options
v 1:324 -x ! get rid of any X parameters refining
q ! quit atom list
v 1:324 -u ! get rid of any UISO parameters refining
q ! quit atom list
v 1:324 -f ! get rid of any FRAC parameters refining
q ! quit atom list
v 29:108 x ! set framework M sites refine on coords
q ! quit atom list
x
x
x
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 73
GSAS: lessen the tedium: batch file control of the refinement :(EXPEDT Example: refine light metal framework atoms)
• File titled
ref_fml.txt
k ! don't make a backup of the file
l ! least squares
a ! atom options
v 1:324 -x ! get rid of any X parameters refining
q ! quit atom list
v 1:324 -u ! get rid of any UISO parameters refining
q ! quit atom list
v 1:324 -f ! get rid of any FRAC parameters refining
q ! quit atom list
v 1:28 x ! set framework M sites refine on coords
q ! quit atom list
x
x
x
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 74
GSAS: lessen the tedium: batch file control of the refinement :(EXPEDT Example: refine interstitial metal atoms - less vacancies)
• File titled
ref_int.txt
k ! don't make a backup of the file
l ! least squares
a ! atom options
v 1:324 -x ! get rid of any X parameters refining
q ! quit atom list
v 1:324 -u ! get rid of any UISO parameters refining
q ! quit atom list
v 1:324 -f ! get rid of any FRAC parameters refining
q ! quit atom list
v 133:144 x ! set interstitial Met to refine
v 147:154 x ! set interstitial Met to refine
v 157:162 x ! set interstitial Met to refine
x
x
x
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 75
GSAS: lessen the tedium: batch file control of the refinement :(EXPEDT Example: refine oxygens)
• File titled
refoxy.txt
k ! don't make a backup of the file
l ! least squares
a ! atom options
v 1:324 -x ! get rid of any X parameters refining
q ! quit atom list
v 1:324 -u ! get rid of any UISO parameters refining
q ! quit atom list
v 1:324 -f ! get rid of any FRAC parameters refining
q ! quit atom list
v 163:324 x ! set oxygens to refine
q ! quit atom list
x
x
x
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 76
GSAS: lessen the tedium: batch file control of the refinement :(EXPEDT Example: refine UISO)
• File titled
refoxy.txt
k ! don't make a backup of the file
l ! least squares
a ! atom options
v 1:324 -x ! get rid of any X parameters refining
q ! quit atom list
v 1:324 -u ! get rid of any UISO parameters refining
q ! quit atom list
v 1:324 -f ! get rid of any FRAC parameters refining
q ! quit atom list
v 1:324 u ! set UISO of all atoms to refine
q ! quit atom list
x
x
x
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 77
GSAS: lessen the tedium: batch file control of the refinement :(Create a Batch File to follow your desired command sequence)
• File titled: refine.bat• In early stage of
refinement, a “cycle” of refinement involves:– Set bond restraint weight
– Enable only X-ray data
– refine framework metal atoms
– refine interstitial atoms
– Enable only Neutron data
– refine light oxygen atoms
– refine constrained UISO
• Batch files can be made more complicated with respect to logging what is happening
expedt model < batchjob\bondr1.txt
echo step 1 bond weight > autolog.txt
expedt model < batchjob\xrayonly.txt
echo step 2 XRAYS ONLY >> autolog.txt
expedt model < batchjob\ref_fhm.txt
genles model
echo step 3 completed >> autolog.txt
expedt model < batchjob\ref_flm.txt
genles model
echo step 4 completed >> autolog.txt
expedt model < batchjob\neuonly.txt
echo step 5 NEUTRON ONLY >> autolog.txt
expedt model < batchjob\refoxy.txt
genles model
echo step 6 completed >> autolog.txt
expedt model < batchjob\refuiso.txt
genles model
echo step 7 completed >> autolog.txt
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 78
GSAS : If the refinement is not happy
• The tools are there for difficult refinements, but:– May not be easy!– May only find problems when in the middle of the refinement
• Impurities in many and varied forms
• XRD and Neutron data are different enough
– different impurities showing up
– different systematic effects
– different volumes of sample having their effect
• Wrong spacegroup
• Non-optimum atom ordering
• Non-optimum tilt model
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 79
GSAS : “wall paper bubble effect trying to keep all bonds physically reasonable
• One problem trying to keep everything physically reasonable: trying to fix gives a “wall paper bubble effect” - fixed one bond causes 1 to 5 other bonds to become physically unreasonable.
• Suggestions on this appreciated (though this could be implying data purity problems).
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 80
GSAS : Combined refinement on both XRD and Neutron(bond length problems - “bubbles” - to stomp on)
May be problems with overlapping impurities (possible impurity problems only found during the refinement).
X-ray 7.6% R(F**2)
(~3.8% R Bragg)
Neutron 4.2% R(F**2)
(~2.1% R Bragg)
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 81
Ton Spek’s Platon Addsym - finding extra symmetry and exploring symmetry in simple to complex polymeric
inorganics
• WWW: http://www.cryst.chem.uu.nl/platon/
• One of the best programs for finding missing symmetry from crystal structure information.
• Now has enhanced subcell finding
• Now has enhanced defaults for inorganics
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 82
Ton Spek’s Platon Addsym - spectacular example
• Triclinic Cell - P1– (96 independent
Carbon atoms)
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 83
Ton Spek’s Platon Addsym - spectacular exampleRun Platon’s Addsym in default mode
• Addsym finds the true cell - which is a 1/8th subcell– (3 independent
Carbon atoms)
– Monoclinic P 21/c
– (compared to 96 atom starting structure)
ECM 2001: "New features in existing powder diffraction software"Lachlan M. D. Cranswick ([email protected]) http://www.ccp14.ac.uk
Slide 84
Platon’s Addsym - (1 of 2) finding extra symmetry in inorganics and minerals - P1
triclinic starting structure.
• Unpublished Mineral Example– Default Addsym gives C2/C