Taking a Lead with LEED – a Tale of two Research Groups

In this blog, Prof. Lionel Clarke OBE, a Cavendish alumnus and Former co-Chair, UK Engineering Biology Leadership Council, shares his Cavendish experience. He was at the Cavendish between 1974 and 1981.

“I first visited the new Cavendish Laboratory in the summer of 1974, responding to a Ph.D. studentship advertisement in surface science. I was interviewed by Volker Heine, head of the Theory of Condensed Matter (TCM) group, in the Mott building. His questioning was intense, but I managed to acquit myself well enough to be offered the opportunity.

I was then introduced to John Pendry, who had recently completed his theory of low energy electron diffraction (LEED), devised to help measure how atoms shift their positions and associated electronic structures at the interface between a crystal surface and an adjacent vacuum—invisible to the established processes of x-ray crystallography. He sought experimental data to test his theory. He agreed to supervise my Ph.D. My first encounter with John was daunting: he wouldn’t explain his theory but asked me to study his book and approach him only once fully familiar. It took another three months before I had enough knowledge—and courage—to speak to him again.

Meanwhile, the experimentalists downstairs in the Physics and Chemistry of Solids (PCS) group had just set up a low energy electron diffractometer, but it wasn’t yet functioning well enough to produce theory-relevant data. I didn’t initially appreciate the enormity of the challenge I was taking on, but was assigned a second Ph.D. supervisor, the congenial experimentalist Jim Wilson, and an office in PCS to help develop the system. Prof. David Tabor, then head of PCS, became a mentor and eventually, by default, my experimental supervisor after Jim returned to the US.

The working environments in TCM and PCS felt vastly different, but developing both theory and experiment together was crucial to the success of my project. The lab setup included four towering panels of electronic controls and a chaotic array of wires. My safety briefing — “Beware of the 20,000 V power supply; always wear rubber-soled shoes and keep your left hand in your pocket” — turned out to be good advice, as I can testify from the fact that I am still alive today to recount it.

Early on, I realised we needed the diffractometer to operate with very low energy electron beams—so that it could be most sensitive to small deviations in the surface atom positions relative to their bulk crystal structure. The Earth’s magnetic field was increasingly deflecting the beams as the energies were reduced, and I needed to design a cage of Helmholtz coils to generate an opposing magnetic field to cancel it out. I approached the technical workshop team with my basic design. They were a fantastic group—Alan Peck, one of the team members, eagerly began building it. Unfortunately, nearing completion, the wires began snapping under stress at the corners. We went back to the drawing board. A new design was made and successfully installed—we had learned a lot together in the process, and the shared experience cemented a great working relationship that lasted throughout the seven years I remained at the Cavendish.

The ultra-high vacuum pump of the diffractometer wasn’t quite as “ultra” as desired. The clean crystal surface would become contaminated within hours, so datasets were pieced together from multiple sessions, each requiring baking the system overnight. Meanwhile, I prepared new single crystals—mostly molybdenum or tungsten—polishing them to atomic flatness. After about six months, I had generated enough data to start to test results against the theory.

The initial match between theory and experiment wasn’t good enough to draw conclusions. This time, theory needed adjustment. John Pendry’s theory had been encoded into a massive Fortran 4 program called ‘CAVLEED’ by TCM’s brilliant programmer, Dave Titterington. In those days this existed in the physical form of a suitcase full of punched cards.

At the time, the University had just one mainframe computer, and running my program would have monopolized it for over an hour due to its complex loops and subroutines, triggering the system’s built-in limits to fairly distribute resources amongst its many users—which in turn drastically slowed my code. I was subsequently permitted to use the mainframe computer only in the small hours of the night, reprogramming the code and test parameters by changing one punch card at a time.

This initiated a day/night pattern of working that I was to follow throughout the remainder of my research, steadily  and incrementally improving the experiments and theory, to achieve the necessary convergence between the measured and modelled results which would eventually, after a couple of years, remove the ambiguities and establish with full confidence the shift in atomic positions within the surface layers to the nearest 0.1 Angstroms (10^-11 m) and the basis for my first publication.

Brian Pippard, the Cavendish Professor, had helped design the Cavendish 2 with the view that personal interactions and discussions should be a critical feature. Coffee and lunch were taken at large octagonal tables, encouraging interaction across groups. Regular attendees in addition to TCM’s own Brian Josephson included the low temperature physics group—Neville Mott, David Schoenberg and Sam Edwards, to whom I was introduced by Brian Pippard. As a young researcher, it was deeply motivating to be among such thinkers, occasionally joined by guests including Phil Anderson and even John Bardeen.

These conversations often turned to the value of mathematical elegance in solving complex problems. Listening in as an eavesdropping student nevertheless imbued an invaluable combination of self-criticism and self-belief in my mind. I convinced myself that the lack of elegance in my process (requiring the use of a computer to solve a messy problem was still considered rather poor form by some at the time) might later be resolved by finding a more beautiful mathematical explanation for the “why” in future, following the “what” my digital methods revealed.

The quiet revolution was starting.  Whilst I was still dependent upon my suitcase of punched-cards to demonstrate what could be achieved with access to a powerful computer (powerful at least for the time), a good friend and former PhD student neighbour in PCS, Herman Hauser, was already developing the basis for what soon would be launched as the BBC microcomputer.   The working environment in the Cavendish at that time was both stretching the boundaries of what could be done with ‘old-school’ techniques and laying the foundations for radical new futures.

In retrospect, my modest achievements and massive personal learnings had only been made possible by my being located at the heart of the Cavendish 2, in the mid 70’s, drawing upon all the previous scientific and technical developments and the enthusiastic support of the workshop staff, together with the encouragement of those coffee table discussions with inspirational colleagues and thought leaders.”

Prof. Lionel Clarke OBE, during his time at the Cavendish