Report on the Bragg Lecture of 1997 held in Leeds
GORDON COX, LEEDS, AND THE
INCREASING POWER OF X-RAY STRUCTURE ANALYSIS
D W J Cruickshank, UMIST, Manchester
The prestigious Bragg Lectures, set up 20 years ago in memory of the Braggs, are given at specific intervals in the four centres where the Braggs did their work: Leeds, Cambridge, Manchester, and London. It was extremely fitting therefore, that, when the annual BCA Spring Meeting took place in Leeds, there was a Bragg Lecture as part of the program, and that it was delivered by a crystallographer whose career had started in Leeds.
Durward Cruickshank's lecture was essentially a selective resumé of some highlights of the development of X-ray crystal structure analysis since its beginning, and especially since the advent of the use of Fourier synthesis. This was development with which he had been personally involved over the past five decades. The lecture was largely focussed on the career of Gordon Cox, partly because it was being delivered at Leeds, and partly because of Cruickshank's personal association with Cox, by whom he had been recruited into crystallography.
Gordon Cox, in the late l920s, had joined Sir William Bragg's research team
at the Royal Institution in London. There, Sir William gave Cox the task
of finding the crystal structure of benzene and, with the aid of special
apparatus for cooling to -2l°C, Cox was able to obtain just enough
structural information to be confident that the molecule was indeed planar,
not puckered. In those very early days the power of X-ray structure analysis
was not very great. This interest in the chemical bonding of atoms in
molecules persisted when Cox was recruited, in 1929, by Haworth, in
Birmingham. There, during the next ten years, he investigated a large
variety of organic and inorganic substances, transition metal coordination
compounds in particular. Two especially significant organic compounds whose
structures were derived there then were ascorbic acid and pentaerythritol.
The former was, of course, the biologically important vitamin C; the
latter provided the cast iron proof of the tetrahedral bonding of carbon.
While these, and many other structures were being determined (all rather approximately), what really began to increase the power of X-ray analysis were the advances in method - particularly in the use of 3-dimensional Fourier summations - and the theory underlying the analysis of the data and results. The mid-l930s saw the advent of new techniques for calculating Fourier series, Beevers-Lipson strips most notably; at the same time, the Patterson function was invented; and Harker sections, using 3-dimensional data, added to its power for solving crystal structures. Cox was a pioneer in the employment of 3-dimensional data. His pentaerythritol analysis was his first important use of three dimensions, but his successful analysis of glucosamine, with Jeffrey, a couple of years later, was a much bigger task, and a major success. (By l945, Jeffrey had completed several more 3-D structures.) A serious interruption to research everywhere was caused by the 1939-45 war but, by 1946, normality was being re-established. Around the world many crystallographic schools began to flourish. Cox moved at this time to Leeds, to set up an impressive research team there. The theory, and particularly the mathematics of structure analysis and refinement was greatly advanced by the application of Least Squares (Hughes in the US) and by the introduction of Differential Syntheses (Booth) and statistical theory (Cruickshank). Most importantly, by the mid-l950s, the enormous problem of the length of crystallographic calculations had begun to be lightened by the advent of powerful electronic computing machines. Structures of all kinds, and of mounting complexity, were being solved in laboratories all over the world. Structural data were accumulating. (In the mid-l950s, for example, Pauling and Corey's alpha-helix was the outcome of the spatial detail available by that time, on the shapes and dimensions of specific small molecules.)
One example of the increasing power of X-ray analysis at that time, the mid-l950s - now 40 years ago - concerned benzene, the molecule that Cox had wrestled with at the Royal Institution, 20 years previously. This time, in Leeds, a really good electron density was obtained, the planar ring was confirmed, and the C-C bond length determined with good precision. However, this result revealed a problem; and the solution of this problem (by Cruickshank) created one further advance in crystallographic structure theory. From the crystal structure analysis, the C-C bonds were shorter than what had been measured spectroscopically; the discrepancy was greater than the estimated limits of error. It seemed a nasty problem; but the embarrassment vanished when it was realised that molecular libration was the cause of this apparent shortening, and thus one more new and valuable mathematical tool in the crystallographer's kit became available. Another example of libration, studied at three different temperatures at this time (~l960) was hexamethylene tetramine. Thermal parameters could be linked to other physical properties, such as entropy.
Durward Cruickshank ended his lecture by moving to the present day. The power of X-ray structure analysis has, by this time, increased to an almost incredible degree, thanks to computer controlled data collection, new X-ray sources (especially synchrotron radiation), vastly more rapid, more sophisticated computers and, in addition, still more novel innovations in theory and method - such as the technique of Restrained Refinement.
Molecules up to thousands of times larger than benzene or hexamethylene tetramine are now being structurally mapped in atomic detail, and with precision, by X-ray analysis. One spectacular contemporary example is the detection, by X-ray analysis, of the atomic movements occurring within the myoglobin protein molecule as coordinated carbon monoxide is photolyzed off (by laser pulse) and then returns again... all inside 350 micro-seconds!
John Robertson
Leeds
The Dorothy Hodgkin Lecture Leeds 1997
The 1997 Dorothy Hodgkin Prize winner, Professor Michael Woolfson FRS, entertained a large audience with a lecture entitled "Rock Salt to Viruses". His stated aim was to trace out the history of X-ray crystallography from its beginning 85 years ago in Leeds with the work of the Braggs to the present time. This he accomplished with the help of many photographs of the leading proponents of the discipline, starting with Max von Laue and on through Lawrence Bragg, Bernal, Dorothy as Dorothy Crowfoot, Lipson and so to the post '45 years. At this point the complexity of the structures which could be solved had risen from the Rock Salt of the lecture's title to ones having asymmetric units with many different atoms and tens of positional parameters. Fourier maps and Patterson functions were calculated with the aid of Beevers-Lipson strips before the advent of modern computers.
Mike graduated from Oxford. After National Service he joined Henry Lipson in Manchester in 1949. From then on the lecturer allowed a personal element to enter his tale as he took his audience through the development of structure determination by direct methods, starting with a Harker-Kasper inequality and his own thesis work in which he derived an equation which later led to the Karle and Hauptmann S1 relationship. Mike's external examiner was Dorothy herself who was "unusual, because not only did she ask searching questions, but she carefully wrote the answers down in a note book". He found out later that this was not criticism of his thesis merely that Dorothy intended to try the new ideas to help her elucidate the structure of vitamin B12.
In 1952, Woolfson became assistant to Bill Cochran in Cambridge and their involvement in the development of probability relationships between the signs of three structure factors for a centrosymmetric structure, A watershed event occurred in 1964 with the solution of the non-centrosymmetric structure of L-arginine dihydrate by Isabella and Jerry Karle by the hand application of phase relationships. In the period 1965-70, Mike worked at York with Gabriel Germain and later with Peter Main to harness the computer to direct methods and produce the first fully-autornatic program MULTAN.
Here Mike paused to pay tribute to the great breakthroughs being made in the 1950s in macromolecular crystallography by the solution of the structure of DNA by Crick and Watson using a model building approach and the development of the Multiple Isomorphous Replacement method and the structure of myoglobin by John Kendrew and of haemoglobin by Max Perutz. More recently, protein structure solution has been simplified by the availability of synchrotron sources and the application of anomalous scattering techniques. Structures such as that of the foot-and-mouth virus have now been solved.
The lecturer concluded by examining the question of whether it is worthwhile to develop direct methods to give ab initio solutions of protein structures -" after all, the community is doing pretty well with the tools they have". He continued with typical enthusiasm "worthwhile or not, it is an interesting challenge". Woolfson's conclusion was that high resolution would be necessary and that some help could be found by developing his and Laila Rafat's TRITAN method published in 1988. Even direct methods maps with bad phases have to contain many correct vectors, since the Patterson function is phase independent. These maps contain correctly oriented fragments which may be combined using the values of three-phase invariants to give better estimates. Work continues at York.
Michael Woolfson's last meeting with Dorothy Hodgkin was in Beijing whilst both were attending the 1993 Crystallography Congress where they discussed their common interest in China and Chinese education. "She was a great scientist and, more than that, a great person".
J.B.Forsyth
The Lonsdale Lecture - Leeds 1997
X-RAYS AND DIAMONDS
A.R.Lang, Bristol
The 1997 Lonsdale lecture was given at the BCA spring meeting by
Andrew Lang. The title of the talk "X-rays and diamonds" gave
no real clue to the originality, richness and diversity of the subject
material presented. At the beginning of his lecture Andrew paid tribute to
Kathleen Lonsdale's own contribution to the study of diamonds, however by
the end of a fascinating and at times high speed tour through diamonds,
topography and dynamical diffraction the audience was left in no doubt about
the huge contribution to this field made by the speaker himself.
One of the main aims of the lecturer was to try to explain the nature of diamonds and their defects by combining imaging and topography methods. The 'defect tour' covered trigons (triangular defects first noticed with optical microscopy); stacking faults; spike diffuse reflections; mixed habit growth and lattice parameter measurement by double crystal topography and divergent beam methods.
Andrew described his own development of projection topography as though it were just something else to do before obtaining some striking images of trigon features. Of course this was a significant achievement. A highlight of this talk was the steady stream of beautiful images. The nature of these trigons was first discovered in 1959 and was later confirmed by cathodoluminescence. Stacking faults with a dihedral and trihedral nature were some of the first features to be investigated with synchrotron X-rays. The polarisation and collimation of the photon beam gave superb images, these were compared to images of the same material obtained using birefringence measurements.
Andrew then discussed the measurements of 'spike diffuse' topographs in certain types of diamond. If a crystal plate has a band of low absorption and the remainder is high, the band will appear on average UV transparent. If the same type of crystal plate contains a band rich in platelet defects in the appropriate orientation it will on average appear to give spikes. The central question is how do these platelets form? Andrew went on to discuss how nitrogen impurities can give quite different properties in diamond and could alter growth patterns. Electron Microscope images of 111 faceted bodies and Moire fringe patterns were in some cases better able to show the presence of the high nitrogen regions believed to be responsible for the platelet growth. Another area where synchrotron radiation has made a big impact is in the study of mixed habit growth. Images of faceted and non faceted growth were shown which appeared in cross shapes, the sharpness of the features obtained with synchrotron radiation allowing more detailed interpretation of the films.
Andrew continued his wide ranging talk into the investigation of impurity concentrations and how they can affect the lattice parameter. He reported results from double crystal and Schleiren topographs. The latter give a strong refractive index gradient with changes in nitrogen concentration. High and low lattice parameters show as colour contrast. He also reported different values of nitrogen free diamond lattice parameter obtained by various methods; the values change in the 4th of 5th decimal place (nm) depending upon whether a symmetric or asymmetric refractive index correction was applied.
Finally Andrew discussed divergent beam topography an the use of Kossel
cones.
He cited a 3 beam case where the reciprocal lattice points are
fixed very accurately. These measurements are very sensitive to changes in
lattice parameter and can detect shifts to one part in 100,000. This topic
had first been studied by Kathleen Lonsdale in 1947, the progress made since
then has been very significant. Equally remarkable perhaps was the
achievement if the lecturer who managed to squeeze 50 years of imaging,
topography and dynamical diffraction into as many minutes. One parting
speculation was that the nature of some feature called 'voidites' was still
not well understood, given the nature of diamond perhaps they could be HCN
inclusions -- diamond smugglers beware!
Bob Cernik
Daresbury laboratory
Editor's Note: Congratulations to our BCA Lonsdale lecturer on the award of The Royal Society Hughes Medal for 1997. The citation reads " Professor Andrew Richard Lang, FRS, in recognition of his fundamental work on X-ray diffraction physics and for his developments of the techniques of X-ray topography, in particular in studying defects in crystal structures"
Click here to return to BCA homepage