John Veysey Sanders 1924-1987
This memoir was originally published in Historical Records of Australian Science, vol.8, no.2, 1990.
Family and education
John Veysey Sanders was the younger of two children of Harold William Sanders, Professor of Mathematics at the University of Adelaide and his wife Isobel (née Armstrong). The Sanders family had long-established roots in Adelaide: John Sanders' great-grandfather had come from Cornwall shortly after Adelaide was colonised and was the Board of Works engineer, and his grandfather was a professional organist (rated as the best in Adelaide) who established a business in the city for selling musical instruments. John Sanders' mother (Isobel) came from an ecclesiastical family, her father having been a Church of England clergyman in Perth, Western Australia. She had, however studied Mathematics at the University of Western Australia, and it was at this time that she and John Sanders' father (H.W. Sanders) were married, the latter then being a Senior Lecturer in Mathematics at the University of Western Australia. Isobel's family was clearly exceptional for the time since, not only was it then unusual for a woman to graduate in mathematics, but in addition, one of Isobel's sisters (Florence) was the first woman graduate in geology in Western Australia.
H.W. Sanders and his wife Isobel moved from Perth to Adelaide in February, 1923, where the former had been appointed to a Lectureship in mathematics at the University of Adelaide, and it was in Adelaide on 28 May, 1924, that John Veysey Sanders was born.
From the security of a close-knit family, John Sanders had a sunny and happy childhood. His school education began at Poultney Grammar School, Adelaide, but when the time came to begin his secondary school education he entered St Peters College, Adelaide, as a scholarship entrant. At school he was a competent but unexceptional student. At least in part this was because he had interests in a wide range of other activities, many of a practical sort such as photography and woodwork; however, the truth of the matter seems to be that he rebelled to some extent against school regimentation, and was happiest when he was able to do things on his own initiative and to make his own decisions. His outstanding academic ability did not become evident until he began his lifelong affection for physics at the University of Adelaide.
John Sanders began his undergraduate career at the University of Adelaide in engineering, but he transferred to physics at an early stage because of a preference for fundamental science rather than applied technology. In retrospect, this transition should not be particularly surprising because a desire to get to the heart of a problem was a central feature of all of his work.
In 1947 he graduated from the University of Adelaide with an Honours B.Sc. degree in Physics, and soon afterwards was awarded a CSIR overseas studentship to work for a Ph.D. degree at the University of Cambridge where he joined the group directed by Dr Philip Bowden the Laboratory for the Physics and Chemistry of Rubbing Solids. He was awarded the Ph.D. degree by the University of Cambridge in 1949, and then returned to Australia to join the CSIRO Division of Tribophysics: he remained with this laboratory and its successors (Divisions of Materials Science, and Materials Science and Technology) for his entire scientific career.
There was at this time a close relationship between the Division of Tribophysics in Melbourne and the Laboratory for the Physics and Chemistry of Rubbing Solids in Cambridge, since the immediate precursor to the Division of Tribophysics had been established in the early years of World War 2 by Philip Bowden (an Australian by birth) at the invitation of the Australian Government which found a need for an expert group to solve pressing wartime problems related to the production of aircraft engines and explosive devices. Immediately after the war Philip Bowden returned to Cambridge, and a relationship between the two laboratories evolved naturally. John Sanders was one of several outstanding Australian scientists for whom this relationship facilitated their working towards a Ph.D. degree at Cambridge at a time when this degree was unavailable at Australian universities.
When John Sanders arrived in Cambridge in 1947 he entered Caius College which had been his father's college before him (his father having read for the Mathematics Tripos while on leave from the University of Western Australia in 1920-22). It appears that on the first day that John walked into the Porters Lodge at Caius he was recognized by the Head Porter as a probable son of Harold Sanders whom he remembered from more than two decades before! They were more spacious days than now.
Early in 1949 in Cambridge John Sanders and his fiancée Gloria (née Cleary) were married. Gloria June Cleary also came from South Australia where her family lived in the town of Angaston and where her father was stud-master at the famous stud at Lindsay Park. John and Gloria Sanders had four children: Jeffrey William (born 1950), Toni (1952), Lynne (1954) and Andrew Veysey (1956). It was a close-knit family in which John took great pride and in which he was able to relax from his scientific work. His great interest in the microstructure of gem opals is dealt with in more detail subsequently, but it did lead to his assembling a considerable personal collection, including specimens he obtained himself from personal trips to inland Australia: most of this collection was subsequently lost when his home in Melbourne was burgled. John liked the outdoor life and he also developed a considerable knowledge of Australian native plants about which he was a mine of information when asked.
John Sanders published over 100 original scientific papers, mostly in the areas of surface science, crystallography, solid state physical chemistry, and electron microscopy/electron diffraction. In addition, he made a notable contribution to the study of the microstructure of opal, and he was undoubtedly the leading authority in this field.
His work in the early part of his research career (1947-1956) was centred on the fields of adsorption, the structure of adsorbed surface layers (particularly of long-chain organic molecules), and the study of these phenomena by diffraction methods. His interest in this area can clearly be traced to his original work at Cambridge, and this in turn had its origin in the pioneering work of the groups at Melbourne and Cambridge on problems associated with boundary lubrication.
By the mid-1950s it had, however, become clear that the chemical properties of solid surfaces could be influenced by the structure of the surface itself. This led to an exploration of the problem of the characterization of solid surfaces (particularly of metals), and this remained an important theme of the work for the best part of a decade. To this end, John Sanders recognized that the use of transmission electron microscopy with thin evaporated metallic film specimens was a powerful technique since it then became possible to characterize the specimen in considerable detail in terms of defects such as grain boundaries, twin boundaries, stacking faults, and dislocations. Although these are basically bulk defects, they can be seen to intersect the surface in a very thin specimen, and each has specific consequences for the surface structure. This approach was used with evaporated films of silver deposited on glass or mica, and the nature of the defected surface was then compared with the chemical reactivity of the surface for the catalytic decomposition of formic acid. It was possible to show that surface defects of this sort were not important in controlling this reaction but the reaction kinetics were controlled by the proportion of silver surface present as 111 planes. This work was certainly one of the earliest examples where the optimum surface structure for a catalytic section had been reasonably firmly established.
The identification of defects on solid surfaces remained a main theme of John Sanders' work for a number of years. It became a rather fashionable topic at the time, mainly because of a widespread interest in trying to understand the fundamentals of solid state surface chemistry, particularly in areas such as catalysis and metallic tarnishing reactions. It was possible to demonstrate the presence and morphology of surface defects by decoration methods. For instance, those on silver were decorated with small particles of gold produced by evaporation or small particles of silver sulphide produced by reaction with hydrogen sulphide.
Much of this work on surface defects used evaporated metal films grown epitaxially on substrates such as mica, although detailed structural studies were also carried out on polycrystalline evaporated metal films of the sort then frequently used as model catalysts. Work with epitaxially grown evaporated metal films led to a study of epitaxy itself, particularly the conditions required for its occurrence, and the way in which crystal growth proceeds from the nucleation stage to the formation of a continuous film.
It was, in fact, the work with small crystals formed in the early stages of evaporated metal film growth that led John Sanders towards the use of the electron microscope for studying the structure of catalysts. Many catalysts also have very small particles or very complex morphologies (or both) which demand the ultimate in electron microscope resolution, and this is a field in which John Sanders became pre-eminent. Although initially devoted to structural studies with dispersed metal catalysts, this work also encompassed the structure of metal sols, and was extended to complex catalysts such as those used for hydrodesulphurization based on the Ni/Mo system. As one of the pioneers in the application of modern electron microscopy to the study of catalysts, he produced several definitive reviews which are widely recognized.
As notable and significant as his early studies on the microstructure of metallic surface and catalysts were, many in the scientific world would consider John Sanders' greatest work to be in the field of high resolution electron microscopy and in particular the direct imaging of crystal lattices and their defects. His pioneering work on this with John Allpress in the early 1970s confirmed the role of crystal defects in supporting chemical non-stoichiometry and began a whole new era in structural solid state chemistry. The methods used have since continued to develop and profoundly influence inorganic chemistry in many diverse areas. John's determination to achieve, with other collaborators, a fundamental understanding of lattice imaging ensured that the technique became established at higher and higher resolution and precision. It has thereby been extended to other applications, many of considerable technological importance in present times.
The early work on n-beam lattice imaging of crystal structures typifies the experimental innovation and critical interpretation of results that underlined much of John Sanders' scientific career. This work commenced after the installation of a new electron microscope and the subsequent commissioning of a special goniometer stage. He designed and supervised the construction of this stage to allow the tilting of crystal specimens about two axes normal to the incident electron beam. The conflicting requirements of small bore objective lens pole-pieces (necessary for high resolution) and double-tilt (which inevitably demands more space than a fixed orientation stage) were cleverly achieved. The resulting resolution limit of about 0.7nm was not generally available for crystallographic work at that time, and this was to prove a crucial factor in the subsequent germination of the n-beam lattice imaging technique. Following the development of the necessary apparatus, he proceeded to investigate the influence on image contrast of various instrumental and specimen parameters, and systematic procedures for using the technique successfully were thereby established.
The crystals used in the early observations were primarily of various phases in the systems Nb2O5-WO3 and Nb2O5-TiO2. Traditionally, it was considered that Nb2O5 could accommodate up to about 50 mol.% of WO3 and about 30 mol.% of TiO2 in solid solution without any substantial change of structure. As a result of the persistent and detailed single crystal x-ray diffraction work of the late A.D. Wadsley (also in Melbourne) during the mid-sixties it had become accepted, however, that these ranges of composition included many distinct but closely related phases together with a number of more or less ordered intergrowth structures of intermediate compositions. Each structure was made up of oxygen octahedra basically joined at their corners but interrupted periodically by regions in which the octahedra share edges rather than corners. These regions, known as crystallographic shear (CS) planes, reduce the overall oxygen/metal ratio to less than that in the corner-shared parent structure, and thereby provide a mechanism for deviations from the stoichiometric composition.
The early one-dimensional n-beam images of such structures were obtained by orienting small crystal flakes such that a set of CS planes was perpendicular to the incident electron beam. The resulting images displayed contrast in the form of a set of fringes, sometimes perfectly periodic but also sometimes showing irregularities, depending on the sample. Careful crystallographic and imaging work suggested that there was a one-to-one correspondence between the spacing of the fringes in the image and the position and number of CS planes (and hence sample composition) in the crystal specimens. This direct correspondence between image contrast and structure was not obvious at that stage, and in fact previous experimental observations had correctly cautioned that such correspondence should not be assumed. Later work using two-dimensional images in which two sets of CS planes were orientated perpendicular to the electron beam suggested that, with care, this interpretation could be extended to these much more informative images also.
These studies entailed working at the very limits of performance of the instrumentation. Setting the small crystals into the necessary orientation required careful attention and considerable persistence. The resolution requirements often meant working in the early hours of the morning after the trams in nearby Swanston Street had ceased running so that vibration and other interference had settled to sufficiently low levels. These early results made an enormous impact on the crystallographic community and opened the way for many new electron microscope applications in chemistry, physics, mineralogy and materials science.
Not being totally satisfied with his crystallographic induction method of image interpretation, John Sanders proceeded, with other collaborators in Melbourne, to set up methods for computer calculation of n-beam lattice image contrast. In a series of publications by this team, starting with a seminal paper in 1972, the use of computer simulated images for understanding the details of image contrast and its variations was explored and exposed. The review articles that followed became the definitive prescription for both experimental and theoretical aspects of high resolution electron microscopy of crystals; it has remained so, with updates occasioned by improved instrumental performance, to this day.
The passion John Sanders felt for electron microscopy is well reflected in the efforts he made to disseminate knowledge and practice in the field throughout Australia and elsewhere. He readily welcomed many beginners in the field, both young and established scientists, into his laboratory and patiently communicated his skills and experience to them. He worked tirelessly as a member of the Organising Committee of the Eighth International Conference on Electron Microscopy held in Canberra in 1974 and was Co-editor of the Proceedings of this highly successful meeting. He was an enthusiastic contributor in many different ways to each biennial Australian Conference on Electron Microscopy and later took great interest in the development of electron microscopy in neighbouring countries through his participation in the activities of the workshops and conferences of the Asia-Pacific Societies for Electron Microscopy.
Throughout his scientific life John had a strong interest in the structure and properties of gemstones, particularly the colours displayed by precious opal. His initial optical experiments borrowed on his crystallographic expertise to provide a clear and definitive elucidation of the diffraction processes that generate such colours. He continued these studies with extensive electron microscope investigations on the microstructure of opal; methods of characterisation of the gems based on this work have followed. These discoveries not only gained international scientific acclaim but also generated many associations with gemmologists, prospectors, jewellers and others outside the scientific community. In fact, he became Federal Patron of the Gemmological Association of Australia and for several years conducted theory and practical classes on the properties of gemstones for the Association in Victoria.
Despite these outstanding successes in the fields of chemical crystallography and electron microscopy, John Sanders never lost his longstanding interest in surface science, catalysis, and the structure of catalysts. In particular, he returned to this field in the last decade or so of his scientific career with a splendid series of papers on the crystallography and structure of zeolites. In this work he drew heavily upon the lattice imaging techniques which had been the cornerstone of his previous studies with defected oxide systems, and he was undoubtedly a world leader in the study of these industrially very significant materials. This work was also extended to modified zeolites and to some non-zeolitic layered materials.
For his outstanding contribution to science, John Veysey Sanders was elected Fellow of the Australian Academy of Science in 1980.
John Sanders was a man of modesty, charm, and total integrity, who strongly believed in the old scientific virtues: that scientific merit could be judged in terms of the intrinsic quality of the work, rather than by how loudly one shouts about it. He was a splendid and loyal colleague and friend to all who were privileged to work with him. For a considerable period of time, until shortly before his death, he was Assistant Chief of the CSIRO Division of Materials Science and Technology. Although this involved a significant administrative load and corresponding inroads into the time he had available for his own research work, he undertook this burden with typical enthusiasm and ability, and he did it not so much because he liked administration but because he was a generous spirit who came to believe that he owed it to the organization which had enabled him to pursue his long and productive scientific career.
John Sanders was aware of his potentially terminal illness for about two years before he died (3 December 1987), but during this time he faced the future with remarkable courage. With characteristic enthusiasm he continued his scientific work at a nearly undiminished level, and was in the laboratory planning future research only a few days before his sudden death.
A.E.C. Spargo, School of Physics, University of Melbourne.