Click here to return to BCA homepage
That was the hope of BCA President Mike Glazer when, instead of each of the four subject groups inviting a plenary speaker to head their own session, he invited them to nominate a speaker for a "Plenary Session" which he would chair. The theme chosen for this session was "Disorder" and the hope was that diversity of subject matter but unity of theme would combine to provide a session of interest and relevance to all crystallographers.
The session started with Keith Prout (Oxford) on "Slow motion and disorder in molecular crystals". He discussed the use of nuclear magnetic resonance (NMR) and X-ray crystallographic (XRD) methods to study molecular disorder in crystals resulting from activated motions. The different timescales probed by the two techniques leads to a useful complementarity in the study of dynamic processes in these structures. Several examples were discussed including two penicillin derivatives. In the first of these, potassium benzylpenicillin, variable temperature 13C CP/MAS NMR studies were used to try to elucidate possible ring flips, where the XRD only indicated in-plane librations. A similar ring flip was indicated in the related compound potassium phenoxypenicillin, where the NMR studies also revealed a phase transition at 373 K between two structures with four and two molecules respectively in the asymmetric unit. Armed with this information, the XRD measurements confirmed the detail of the structure. The different phenyl group orientations found in the high temperature crystal structure can be interpreted as random "static" disorder, but the spinning side-bands in the solid state NMR shows this to be due to slow motions of the phenyl groups, emphasising the importance of complementary NMR
Several other examples involving the clarification of disorder as "static" or "dynamic" were also given, including examples in organometallic chemistry and of molecules in cavities, with studies of various deoxycholic acid derivatives. In these molecules such as ferrocene and camphor sit in the layers defined by the deoxycholic acid molecules. These complexes show various levels of disorder and exhibit phase transitions. The correspondence between the information from the NMR and the librational motions indicated by the atomic displacement parameters found from the crystallography leads to models interpreting the disorder mechanisms as slow motions of the guest molecules. The final example was of the structure of pyridinium nitrate thiourea clathrate, where several groups exhibit disorder and various types of NMR can be used to elucidate the different order/disorder characteristics of these and hence help to explain the complex series of phase changes undergone by crystals of this type.
Richard Welberry (Australian National University) then talked about "Computer simulation as a tool for the interpretation of diffuse scattering", introducing us initially to the analytical expressions governing the appearance of diffuse scattering in diffraction patterns. After a clear and informative description of the various terms and their origin, the difficulty of describing all of these analytically was explained and the modelling of these in terms of rather fewer parameters was described. This intuitive approach was illustrated in a slide show using optical transforms, which illustrated the effect on the diffraction pattern of various simple types of disorder. This showed how different basic forms of disorder lead to very characteristic shapes of disorder scattering. Knowing these basic shapes can greatly assist in the interpretation of diffuse scattering patterns.
Following this introduction, the curved position sensitive detector (PSD) based X-ray diffraction rig used to measure the diffuse scattering patterns was described. Both the measurement and interpretation of the diffraction patterns were illustrated in a series of examples including KLiSO4, urea inclusion compounds and stabilised zirconias. In the latter, the presence of local oxygen vacancies allows the metal centres to relax from their lattice sites and leads to strong diffuse scattering in the diffraction pattern. A series of models were tried, with the best description of the diffuse scattering being obtained from mainly isolated [111] vacancy pairs, and with consequent movement of neighbouring metal atoms along [110] directions away from the vacancy defect. Introducing correlations between these displacements leads to improvements in the fit between calculated and observed patterns.
The talk finished with the description of a new, more computationally intensive, analysis method, and this was illustrated in the Fe3(CO)12 system. In this method the problem is initially parameterised, and the parameters are optimised to give the structure in best agreement with the data using a least squares method, in which Monte Carlo calculations are performed to determine the structures corresponding to the parameters. Though computationally intensive, there is clearly great potential in this method.
The third lecture in the session was given by Patrick Fairclough (Sheffield), moving us from traditional crystallography into the regime of larger scale structures and small angle scattering with his talk on "Scattering studies of polymer crystallisation". The work is aimed at elucidation of how the formation of crystallinity in different types of polymer affects the material properties. The material strength is ultimately of great relevance to the industrial properties of the polymer.
Various clever experimental approaches to the study of crystallisation in situ were described and the use of both the Bragg and diffuse scattering in small angle X-ray scattering (SAXS) and wide angle X-ray scattering (WAXS) patterns was shown to be important in understanding how the polymer both crystallises and organises itself. As well as crystallisation in diblock copolymers, where the crystallisation is found to be confined to the lamellar regions, the lecture discussed nucleation, crystallisation and growth effects in homopolymers. Much of this work was carried out at the Daresbury Laboratory on the SRS stations 16.1, 8.2 and 2.1, in some cases with DSC performed in parallel, at a typical time resolution of 6 seconds. The combination of WAXS and SAXS can be used to differentiate between the various possible contributory factors: classical nucleation is a first order effect and leads to scattering observable in the WAXS pattern, while spinodal decomposition is a second order effect and perturbs the SAXS pattern. In the initial experiments on homopolymers the sample was cooled at a rate of 200C per minute and the SAXS signal was found to develop first, indicating spinodal decomposition to be the dominant process at higher temperatures with growth switching in at 140C. The experimental set-up has subsequently been refined with the installation of a polymer extruder on the beamline, the adjustable position of this being used to change the temperature of the sample in the beam. In these experiments, the SAXS is still always found to develop before the WAXS, indicating that from the melt the spinodal scattering develops at high temperature followed at lower temperatures by the development of crystalline peaks. The conclusions of this work are that long range order develops prior to nucleation and therefore that spinodal decomposition plays an integral role in the crystallisation process.
The talk concluded with an account of the planned study of the silk fibres produced by another, more natural, extruder, that of the web producing duct of a spider. To date these measurements have been carried out on polymers extracted from various stages in the "production" process, and these are found to be more crystalline as one extracts material from further through the gland. The extraordinary strength of these fibres makes their study and understanding an exciting prospect, if potentially less than welcome for the in situ spider.
The session was ended by Keith Wilson (York) who spoke about the various types of disorder in protein structures and how it is to be both appreciated and treated with caution depending on the type of information one is attempting to extract from the structural analysis. In "Proteins: selected disorder" we were treated initially to a clear explanation of the susceptibility of these structures to disorder, with the typical three dimensional protein structure being held together by very weak interactions, typically totalling around 20 kcal mol-1, compared with the denatured protein. This energy is equivalent to only around three hydrogen bonds, meaning that such structures are barely stable at room temperature, and the crystallisation process can therefore seriously perturb the interactions holding the structure together.
"Disorder" in small molecule structures is far more tractable than in proteins, where the presence typically of around 50% water, which itself is largely disordered, can severely limit the available data. For large unit cell problems of very ordered materials high flux can lead to an increase in the data resolution, for example Laue data to 0.67 Å resolution was collected from L-asparagine, and could be used for charge density work. For structures with intrinsic disorder the scattering is intrinsically weak and the only solution is either better crystals or collecting data at low temperature. A typical diffraction pattern illustrated this, showing that often only 3 Å data or poorer is available even on the most intense sources due to intrinsic disorder in the protein crystal structure. However, information available in the diffraction pattern measured to different resolutions is vital to unravelling the various different aspect of structural order and one should always aim to get the best possible resolution data to reduce "blur" in the refinement, even if the ultimate aim of the experiment is not an atomic resolution structure.
Various examples were shown where the interpretation of disorder played an important role in understanding the structure. The disorder was treated in three broad categories: disorder of parts of the protein molecule itself, solvent disorder near the protein and the presence of alternate solvent networks (illustrated for example in work on vitamin B12) - the talk stressed that interpretation of solvent structure is an immense field in protein crystallography. Two different aspects of side chain disorder were discussed. First, the freezing out at low temperature of two conformations of an apparently vibrating group - if this involves the active site then one conformation can be favourable for hydrogen bonding and one not, giving important clues to the binding mechanism. Secondly, alternative conformations, which are not the dominant conformation at in vivo temperatures, can sometimes be frozen out at low temperature and hence important information can be masked in such low T studies.
Something for everyone? Certainly. It would be an unappreciative crystallographer who did not gain much from this session and it is to be hoped that the experiment will be repeated in the future.
Chick Wilson
[email protected]
As well as the joint BCA session on 'Disorder' reported above another experimental format was investigated during the Spring Meeting in St Andrews. Instead of having a BCA Plenary lecture following the AGM we had a "Question and Answer" session chaired by Bob Gould. The questions were solicited from participants earlier in the Meeting.
Q:Where do I find a "Teach yourself crystallography" text at an
affordable price foran absolute beginner?
This apparently straightforward question led to a number of texts being
suggested, including Lesley Dent Glasser's "Crystallography and its
Applications", but many seemed to be out of print. The discussion then
proceeded to analyse the question: what was affordable for students? Could
an absolute beginner obtain useful knowledge of the area simply from a text
book? What was meant by crystallography? One contributor pointed out that
trying to learn crystallography in this way was analogous to trying to learn
brain surgery: in response came the suggestion that it was perhaps more
analogous to learning about the function of the brain.
(Editor's
Note: Further suggestions are welcome at any time, they will be added to the
book list on the BCA Web site at
http://gordon.cryst.bbk.ac.uk/BCA/Cnews/books/books.html#List)
Q: What should/could BCA members do during next year's National week
of Science and Technology, SET99. (or any other time!) to raise interest in
crystallography amongst the general public?
A number of suggestions were made in response, including the organisation of crystal-growing competitions which have already proved successful in some regions. The BCA has been investigating the possibility of an issue of Royal Mail stamps to coincide with the IUCr Congress but the issues for 1999 now seem to have been allocated. Under the general theme of "Time", there will be stamps devoted to "scientists", but only as one of twelve occasionally curious categories one of which is "patients". The observation that the topic for SET98 was 'the paranormal' led to the suggestion that talking about "crystals" could be misunderstood.
Q: What is the smallest size of single crystal from which diffraction
data can be collected nowadays?
Marjorie Harding suggested that the answer might be around 0.5 microns,
but this would apply to a simple inorganic material such as CaF2. With
increasing complexity larger crystals would be needed as the mean reflection
intensity decreases as the number of atoms increases.
Q:Mike Glazer makes a good case for more use of the polarising
microscope. Can anyone offer advice for the building of apparatus suitable
for colour-blind crystallographers?
This question required some clarification: the apparatus referred to is a special polarising microscope which can separate the normal birefringence colours of a crossed polars image into three components, representing the transmission of the light intensity, the magnitude of the birefringence, and its orientation: these can then be displayed as false colour representations. (see front cover of 'Crystallography News' no 58 Sept 96.) The solution for colour-blind users is to change the colour representation, for example by using greyscales.
Q:Acta C makes its acceptance criteria more and more
stringent. Is this because they only want boring, easy structures to be
published there?
With at least three Acta C Co-Editors in the audience this question provided good targets. Does the journal discriminate against chemically interesting results which have, for example, been extracted with skill and patience from unpromising crystals? These charges were answered by pointing out that the criteria were those which should be met by an average study but that extenuating circumstances were taken into account, for example where a crystal diffracted weakly the resolution limit might be relaxed. The Editorial in the January 1998 issue of Acta C contains examples of "valid extenuating circumstances" and "unacceptable excuses". Speaking as an Editor of Z. Krist., Chris Gilmore suggested he would consider structures at which Acta C might turn up its nose.
Other questions concerned achiral molecules in enantiomorphous space groups; the importance of polarity in space groups like Pnc21; the interpretation of Flack parameter values; the importance of correct weighting of reflections whether you are refining on F or F2; when a structure can be considered as refined to convergence; and the reasons for the prevalence of the space group P21/c - is this entirely due to its inherently efficient translational symmetry, or because many of us like it so much?
There was as much subsequent discussion of the session as of any questions discussed during it. Should it be repeated at a future Meeting? If so, how could it be improved? Criticisms of this first attempt included the timing (it would be better before the AGM, or as an evening session held in the bar); the content (too many items on chemical crystallography); some questions were too elementary, others too advanced for discussion in such a forum; the mixture was not suitable) and the structure (would it be better to have a panel answering questions before they are thrown open for contributions from the audience?).
A.J.Blake,
University of Nottingham
Last update: 7 Jul 98
Click here to return to BCA homepage