Why direct and post-refinement determinations of absolute structure may give different results

Sep 192016
 

Acta. Cryst. (2016) B72 661-683 (Feature Article) [ doi:10.1107/S2052520616012890 ]

surfaceDirect determination of the Flack parameter as part of the structure refinement procedure usually gives different, though similar, values to post-refinement methods. The source of this discrepancy has been probed by analysing a range of data sets taken from the recent literature. Most significantly, it was observed that the directly refined Flack (x) parameter and its standard uncertainty are usually not much influenced by changes in the refinement weighting schemes, but if they are then there are probably problems with the data or model. Post-refinement analyses give Flack parameters strongly influenced by the choice of weights. Weights derived from those used in the main least squares lead to post-refinement estimates of the Flack parameters and their standard uncertainties very similar to those obtained by direct refinement. Weights derived from the variances of the observed structure amplitudes are more appropriate and often yield post-refinement Flack parameters similar to those from direct refinement, but always with lower standard uncertainties. Substantial disagreement between direct and post-refinement determinations are strongly indicative of problems with the data, which may be difficult to identify. Examples drawn from 28 structure determinations are provided showing a range of different underlying problems. It seems likely that post-refinement methods taking into account the slope of the normal probability plot are currently the most robust estimators of absolute structure and should be reported along with the directly refined values.

Publisher’s copy

BCA Spring Meeting 2011

Apr 112011
 

The BCA Spring Meeting was held at the University of Keele in Staffordshire.  Contributions from Chem. Cryst. included:

Richard I. Cooper, Amber L. Thompson & David J. Watkin
Standard Uncertainties and Experimental Design (Poster)

Amber L. Thompson & David J. Watkin
“Unpublishable” Data:  Does My R-factor Look Big in This? (Poster)

David Watkin
What’s all this MoOing about? (Presentation)

Handling Cell Errors in Crystallographic Data

Apr 152010
 

Cryst. Rev. (2010), 16(2), 133-144.    [ doi:10.1080/08893110903483238 ]

Recent developments in crystallographic error analysis are described. This report provides an introduction to more-formal work, originally published in Ref. (Haestier, J., J. Appl. Cryst. 2009, 42, 798). Prior to the main discussion, a brief overview of the normal distribution and error-propagation is provided, with some simplified examples to demonstrate the effects of covariance terms on error calculations. A new method is described for absorbing the cell-parameter errors into the variance-covariance matrix of the refined parameters. Problems occur for monoclinic and triclinic cell settings as the crystallographer must ‘choose’ how errors on the cell angles affect the coordinates. The choice has no effect on error-calculations on the internal geometry (bond-lengths, angles and torsion angles) of the structure, but may introduce a bias if the errors are used as weights.

An Unusual Methylene Aziridine Refined in P21/c and the Nonstandard Setting P21/n

Nov 182009
 

Acta Cryst. (2009), C65, o635-o638.    [ doi:10.1107/S0108270109046952 ]

The unusual methylene aziridine 6-tert-butyl-3-oxa-2-thia-1-azabicyclo[5.1.0]oct-6-ene 2,2-dioxide, C9H15NO3S, was found to crystallize with two molecules in the asymmetric unit. The structure was solved in both the approximately orthogonal and the oblique settings of space group No. 14, viz. P21/n and P21/c, respectively. A comparison of these results clearly displayed an increase in the correlation between coordinates in the ac plane for the oblique cell. The increase in the corresponding covariances makes a significant contribution to the standard uncertainties of derived parameters, e.g. bond lengths. Since there is yet no CIF definition for the full variance-covariance matrix, there are clear advantages to reporting the structure in the nonstandard space-group setting.

Publisher’s copy: IUCr

ECM 25

Aug 212009
 

The 25th European Crystallographic Meeting was held in the Harbiye Museum and Cultural Centre in the beautiful city of Istanbul.  It was a very eventful week, and contributions to the conference made by Chem. Cryst. include:

N. David Brown, James Haestier, Mustapha Sadki, Amber L. Thompson & David J. Watkin
A Further Improved Structure Matching Algorithm (Poster)

James Haestier
Handling of Cell Errors and their Effect on Derived Parameters (Poster)

Mustapha Sadki
New Modelling for Disordered Atoms in Free-form Based Hybrid Refinement and Visual Representation (Poster)

Mustapha Sadki
The Application of Novel Modelling and Refinement Strategies to Crystallography (Presentation)

Amber L. Thompson & David J. Watkin
Absolute Configuration Determination – Is there More Information in the Data? (Poster)

David J. Watkin, Richard I. Cooper & Amber L. Thompson
CRYSTALS:  Refinement and Validation Tools (Poster)

David Watkin
The Future of Small Molecule Software (Session Chair)

Handling Cell-Parameter Errors in Crystallographic Data

Aug 012009
 

J. Appl. Cryst. (2009), 42, 798-809.    [ doi:10.1107/S0021889809024376 ]

A new method is presented for handling errors on crystallographic data. In single-crystal diffraction experiments, two variance-covariance matrices are present, one for the cell parameters and the second for the refined parameters (atomic coordinates and anisotropic displacement parameters). These two matrices can be combined so that errors on derived parameters, such as bond distances, bond angles and TLS tensors, may be calculated more simply. The new method works for all space groups but there are limitations on its application to triclinic space groups. The method allows errors to be transformed between space groups.

Publisher copy: IUCr

BCA Spring Meeting 2009

Apr 212009
 

The 2009 British Crystallographic Association Spring Meeting was held at the University of Loughborough.  Contributions from Chem. Cryst. included:

N. David Brown, James Haestier, Mustapha Sadki, David J. Watkin & Amber L. Thompson
A Further Improved Structure Matching Algorithm (Poster)

James Haestier
Effects of Cell Errors on Derived Parameters (YC Presentation)

James Haestier
Computation of Cell Errors Effects on Derived Parameters (Poster)

Mustapha Sadki & David J. Watkin
New Framework for Reliable Refinement Data Types (Presentation)

Error Estimates on Bond-length and Angle Corrections from TLS Analysis

Apr 082008
 

J. Appl. Cryst. (2008), 41, 531-536.    [ doi:10.1107/S0021889808005463 ]

Librational motion within a crystal structure distorts the measured bond distances and angles from their physical values. TLS analysis of a rigid molecule or a rigid part of a molecule allows the calculation of bond-length and angle corrections. Until now, no estimate of the error on these corrections has been available. A method is presented for propagating the errors on the anisotropic displacement parameters (ADPs) to the bond-length and angle corrections which are a function of the libration tensor. The numerical significance of approximations made during the calculation is discussed.

Publisher copy: IUCr

BCA Spring Meeting 2008

Apr 012008
 

The Annual Meeting of the British Crystallography Association was held in York.  Contributions from Chem. Cryst. included:

Anna Collins, Richard I. Cooper, Andrew R. Cowley, David J. Watkin
2-(1-Hydroxy-2-methyl-propyl)-2,5-dihydro-furan-2-carboxylic acid diisopropylamide:  A study of a phase transition to a pseudosymmetrical Z′=2 structure (Poster)

James Haestier, Mustapha. Sadki, Amber L. Thompson and David J. Watkin
Cell Parameter Standard Uncertainties and their Effect on Crystallographic Data (Poster)

Mustapha Sadki, James Haestier, Amber L. Thompson & David J. Watkin
Computational Infrastructure for Bridging the Gap between Previous and Future Generations of Crystallographers (Poster)