Interview: D. Phillip Woodruff
2000 Medard W Welch Award Recipient
Interviewed by Ted Madey, 2000
MADEY: Hello. I'm Ted Madey from Rutgers University. Today I'm interviewing Professor D. Phillip Woodruff from Warwick University. Phil Woodruff is the year 2000 winner of the Medard W. Welch award. The citation reads that he's receiving the award "for contributions to the understanding of the geometric properties of clean and adsorbate-decorated surfaces, and for innovative development of surface science techniques." Phil, perhaps you could tell us a little bit about yourself, your educational background, and how you became interested in physics in the first place.
WOODRUFF: Okay. Well, I was born in the Manchester area of England. My key education, my secondary education, was at Manchester Grammar School, which is a slightly famous school because it has a reputation for academic excellence. You had to pass an exam to get into this place, and people came from many miles around to do that. And actually, I started at the age of ten there, so I was kind of a year ahead of myself, which maybe wasn't always good because I wasn't always at the top of the class by any means.
But I developed an interest in science. In a way, the way the school was organized, you tended to be a linguist or a scientist. It was very clear to me that the latter was my natural affinity. I just moved on in that general direction. I think when I was at school, I had it in mind that I'd probably end up in chemistry. Then when I got to the more meaty kind of science, I suddenly decided that physics was much more fascinating. So I went on to Bristol University to take a first degree in physics, and then on to Warwick, at the brand-new University of Warwick, in 1965 to take a PhD.
MADEY: What attracted you to Warwick?
WOODRUFF: Well, I was basically looking for a PhD project and a lot of my colleagues were going into things like high-energy physics and so on. I was really not attracted at that time to the idea of doing a project where I'd be part of a huge team. I really wanted to try and do the kind of science where it was you and the science, as it were. And I talked to some potential people in Bristol. One of them was John Forty, who was about to move to Warwick as the founding professor of physics, and he persuaded me to go with him. So I had actually a kind of strange vision of the Midlands as a deeply industrial area, which is strange because coming from Manchester, it's actually a much more industrial area [laughs]! When I got there, I realized that we were right on the edge of that industrial area, and it's a beautiful area to be. It's where, indeed, lots of American tourists come, to Stratford-on-Avon, which is very close to where I live.
MADEY: But it was at Warwick that you really became interested in surfaces and the structure of surfaces, too.
WOODRUFF: That's right. Yes, my PhD was actually on solidification melting, so I was really interested in the solid-melt interface. I just developed out into the new area of kind of solid-vacuum interfaces. There was a guy in the department, Helmut Mykura, who was just setting up a new LEED [Low Energy Electron Difraction] project, and this looked interesting. So I spent a year post-doc-ing with him on that and then actually got a permanent position in the department and developed that theme.
MADEY: What did you do in your first LEED studies?
WOODRUFF: Well, it was really very much trying to get a handle on a basic understanding of LEED from single-crystal surfaces. We were looking at, I think, copper (111), if I remember. It was about the time of the early development of LEED theory. John Pendry, at that time, was a student of Volker Heine's at Cambridge, and we interacted a little bit with him. So we did kind of simple things, like trying to understand temperature effects in LEED and how the intensities varied, which gave us a handle on Debye-Waller factors, and then also realized that the LEED I-V peaks moved, and that made us realize that we could get some information out of surface thermal expansion coefficients. Then ultimately, we started to try and measure proper quantitative intensity energy data and see if we could match those with theory.
MADEY: So when you joined the faculty at Warwick, then, did you continue initially in the direction mainly of LEED?
WOODRUFF: That's right, yes. It was very much LEED. I went out to get some funding to build a separate LEED system, which would allow us to do quantitative measurements by putting a Faraday cup into the system. That basically developed from there. New things came along - Auger electron spectroscopy came along and we realized maybe one could think about microscopy. So we started out a project in that area and so on, just kind of spread out as the field spread.
MADEY: What was your first experiment in Auger spectroscopy that related to surface structure?
WOODRUFF: That was kind of a coincidence as well, because we'd just built this LEED system with the Faraday cup. The Faraday cup had a pair of grids in the front of it, so it was essentially like a miniature LEED optics. And we'd just seen reports in the literature of people who'd shown that you could measure Auger electron spectroscopy with a LEED optic retarding field analyzers. So we thought, well, maybe we could do the same thing with this little detector and look at the angle dependence. I think, to be fair, at the time it was just curiosity that drove that. We measured some of the Auger peaks from copper and saw these huge angular variations. Then, of course, having found something, we had to try and figure out what it meant.
MADEY: How big were the angular variations?
WOODRUFF: Oh, off the top of my head, a factor of two or three variations. They were big, they had big effects. Basically, I set up a very simple calculation to just look at the diffraction aspect about the final state effect, which was a plane wave single scattering model, which reproduced the kind of phenomenology but not the detailed structure. It basically showed us that the diffraction was clearly an important component, but we were conscious that these were core valence-valence transitions. So getting a proper handle on the initial state was very complicated. That's what made me realize that the experiment we should really be doing was in photoemission from a core level, where one has a much better handle on the initial state.
MADEY: So what was the first core level photoemission experiment that you tried?
WOODRUFF: Ultimately the first photoelectron diffraction experiment was an experiment on sodium on nickel (100), which you may remember was one of the very first systems to be analyzed by LEED, by John Pendry again, and Stig Anderson. That was done with Neville Smith and Mort Traum and Helen Farrell from Bell Labs. I'd actually met Neville at a conference in Warwick in 1975, which turned out to be a rather exciting period. It was a kind of UK national surface science meeting that tended to have quite a lot of Europeans. It was just the year that angle resolved photoemission really came of age. And quite a number of people came over from the States, quite a few from Bell Labs. They were presenting angle resolved valence emission. Or actually Neville had some data on shallow core levels in layer compounds.
MADEY: They were selenides, I believe.
WOODRUFF: That's right. And I remember they had features in the azimuth patterns that Mort Traum used to refer to as ears, noses, and chins [laughs]. We basically just hit it off. Neville was actually from England himself, but I'd never met him before. But we just seemed to gel. We said, well, let's try and figure out how to do this experiment together. So we ended up doing it out on the Tantalus storage ring at Madison.
MADEY: What energy range were you working in there?
WOODRUFF: Oh, that was - I don't remember. The binding energies were 30-50 eV or something like that. So we were working with photon energies around 100-150 eV. That was actually already a tricky energy range for synchrotron radiation. It was something that was only just becoming available. We actually ended up shipping out a monochromator from the U.K. because there was a monochromator that had just been installed on the old electron synchrotron on NINA just in the dying days of NINA, as it were. Then we had a three or four-year dark period before they built the new SRS. So we shipped the monochromator out to Wisconsin and became a kind of general use facility for a couple of years.
MADEY: I see. Then you brought it back later?
WOODRUFF: It came back and actually was doing good service on the SRS until just a year or so ago. But we even made some measurements from that in the last couple of years.
MADEY: But in these first experiments with the sodium on nickel, were you able to do an analysis that confirmed the binding sites?
WOODRUFF: Actually, yes we were. There was a theoretician at Warwick, Brian Holland, who had got interested in the problem. He'd actually been doing the LEED theory calculations. Yes, he did some calculations that basically showed that we could match those data with the known bonding site. It was kind of nice.
MADEY: Then your interest in the use of synchrotron radiation for photoelectron diffraction really bloomed after that.
WOODRUFF: Mm-hmm, yes. There was actually a bit of a gap. We did some more experiments with Wisconsin for a few years, and I'd spent a bit of time on sabbatical leave at Bell Labs doing that. But then we reached a period where what we were really keen to do was to look at things like the carbon 1s and oxygen 1s levels, to look at molecular species of some chemical relevance. But there really was just a lack of decent beam lines around the world in that energy range. The project vaguely went into limbo for a few years.
Then I got together with Alex Bradshaw at the Fritz-Haber Institute in Berlin, and they build up a Fritz-Haber beamline on BESSY (the Berlin electron synchrotron radiation source), which covered precisely this energy range from 200 eV to 1000 eV or so. The first experiment we did was really a very simple experiment to try to just see if we could get some data in this way. But then, alas, I think that initial experiment was not very successful. We got enough data to encourage us to start to do something more serious in that area. It slowly took off when we built up a lot of methodology.
MADEY: What year was that that you actually started collaborating with Bradshaw?
WOODRUFF: I spent a brief period of time at the Fritz-Haber Institute in the early '80s. But then that was really before BESSY was fully operational, and I worked on infrared spectroscopy. Then we started using BESSY together around about 1986. So we've had almost 15 years of this collaboration since then.
MADEY: So you spend quite a lot of time in Berlin, then?
WOODRUFF: Right now I do. Now I spend a week every month there, actually, because Alex Bradshaw has moved on to a position at the Plasma Physics Institute in Munich. I effectively run both ends of that exercise at the moment [laughs].
MADEY: Keeps you rather busy.
WOODRUFF: Yes, that's right.
MADEY: How did you get involved in the X-ray standing waves work?
WOODRUFF: That was another kind of accident, in a way. We built, ...as I was saying, they decided to build the so-called Synchrotron Radiation Source, the SRS, at Daresbury Lab in the UK. I suppose the decision was taken in the mid- to late-'70s. The machine finally came on-stream in '81-'82 or something like that. And during that period, the idea of SEXAFS (surface extended X-ray absorption fine structure) really emerged from Paul Citrine at Bell Labs, who'd been working over here, in the USA.. We decided this looked like an interesting way to get surface structure and that we should try to build a beamline to do that. So I got together a proposal to build a beamline, actually collaborating with David King, who was at Liverpool at the time, and Martin Prutton at York. And we got the funding to build the beamline and set up a program of surface EXAFS. But you know, surface EXAFS is always a struggle. The size of the signals you're working with are rather small. We happened to be making some measurements on chlorine on copper (111) with essentially normal incidence to the surface. About, I don't know, 100-150 eV above the chlorine absorption edge, we saw this rather strange feature in the EXAFS, which was actually quite a bit bigger than the EXAFS. I realized that when I thought about it, when I was working at Bell Labs I'd heard a talk by Gene Golochenko about X-ray standing waves. This just rang bells in my head, and I looked and realized that this feature was at the energy of the Bragg peak from the copper (111). But I was really puzzled by it because this talk by Golochenko had made a big point about how you had to have incredibly perfect crystals and you had to have this incredibly perfect collimated beam because the rocking curve is extremely narrow. I thought, well, metal simple crystals just aren't like that. This beamline is not beautifully parallel.
Then it slowly dawned on us that because we were working at normal incidence, the Bragg condition goes through a turning point, so you have essentially zero gradient in the Bragg condition at normal incidence. So suddenly the experiment at normal incidence is not sensitive to these parameters. Then we basically realized this was X-ray standing waves, and the question was how do we do an experiment to do something useful with this. That probably took another couple of years, actually trying to think how to do that and how we detect the signal and modifying, maybe, the instrumentation a little bit to do that.
So it was a tantalizing period, actually, because during that same period there was a group at the Photon Factory in Japan that obviously realized the same thing and published a spectrum that showed the feature and said this obviously is standing waves and we could use it, but didn't actually use it. So we finally did get the first result out.
MADEY: What year was that when the first result came out?
WOODRUFF: That was around '87 or something like that. It was actually a very similar period to when this sort of thing started to get going at BESSY; that's right.
MADEY: What have some of the systems been that you've used standing waves in? And I would say where do you think it is most successful and most useful?
WOODRUFF: We've tended to - I was mentioning early that when we do the photoelectron diffraction, one of the things we're really interested in is to look at the low atomic number elements, which are chemically very relevant. If you go to the x-ray standing wave, then your X-ray energies are around 3 kV or higher. The cross-section for these very shallow levels is then very low because you're far above the threshold for photo-ionization. So it's much more natural to go to, say, second row elements in the periodic table, because then you can get to the 1s stage with those, things like sulfur and chlorine and phosphorus. And you can ultimately go to heavier atoms as well by looking at shallower levels.
So early on most of the work was on sulfur and chlorine and molecules containing those species. We tried very much to look at molecular systems as well, which was something that we felt was really lacking in a lot of the data. A lot of the great bulk of surface crystallography data is still from LEED. But LEED is a technique which relies on good, long-range order, so it's particularly appropriate for atomic adsorbates. Many molecular systems don't form nice, ordered overlays, so that was a real strength for these two techniques because neither of these techniques rely on long-range order.
MADEY: So you are looking at the local order.
WOODRUFF: Exactly.
MADEY: And they have to be ordered with respect to the crystal lattice.
WOODRUFF: Well, they need to be in well-defined sites. That's right. One of the interesting things about the standing wave is that it does give you some information about the disorder. You just have two fitting parameters, and one of those parameters is information about order or disorder. Although if that parameter says there's a lot of disorder, it becomes quite hard to get a reliable structure.
MADEY: But you also, I think, have taken advantage of the chemical specificity of X-ray standing waves in certain instances. Can you comment on that?
WOODRUFF: Yes, that's right. Again, that's something that really started the photoelectron diffraction, because if you're looking at the photoemission signal and you obviously start to think, "Well, if I have some chemical shifts in those states, then I can look at them separately." And that's something we demonstrated with photoelectron diffraction some years ago, but only with relatively simple model systems. We needed large chemical shifts. Because the nice thing about being the one that allowed us to do these experiments at BESSY was it was very high flux but not high resolution. You couldn't have both at once. But then you realized that if you can monitor the X-ray standing wave absorption by the photoemission signal, then you have the possibly, again, to use those chemical shifts. In fact, we gained access... You know, this combination of high flux and high resolution is essentially a strength of undulator sources on third generation synchrotron radiation sources. In fact, we gained access to a facility in the X-ray energy range, that is the European synchrotron radiation facility in Grenoble, before we actually had access to a similar facility in the soft X-ray end. So in a way, these third-generation experiments are really started with the X-ray standing wave. We don't have such fantastic spectral resolution at 3kV as we can get at a few hundred eV, but even so, it's allowed us to make some progress in that direction.
MADEY: How about beam damage effects working at the extremely intense sources?
WOODRUFF: It's a serious issue, that's right. In the experiments we've done with the standing waves, certainly - well the species that I know you're very familiar with, PF3 - we certainly do see fragmentation of that species in the beam. It turns out that for that particular system, the rate of fragmentation is just about slow enough that you can both exploit it to produce the fragments but still get measurements in the meantime.
MADEY: You have also used ion scattering to learn about the surface structure?
WOODRUFF: Yes, that's right. I'd be hard-pressed to put a date on it, but it's pretty early on in the early '80s that I did get interested in low-energy ion scattering - helium 1kV kind of scattering - I think triggered by the kind of work that Hidde Brongersma had been doing at Philips Labs and Ed Tagelauer and Werner Heiland had been doing in Munich. Indeed, I actually took a trip over, took a week, to go and see these two guys to try and find out what the techniques were really like. Because these new techniques always look wonderful until you get into detail, and you always find the things that go wrong as well as the things that go right. And we did set up some work on low-energy ion scattering, on helium scattering.
Actually, that kind of turned into an interesting bag of worms in its own right, because the idea is just to exploit the elastic shadowing. It was well known that you have a major charge exchange issue there, that low-energy helium ions are very easily neutralized by a metal surface. This was part of the basis of Homer Hagstrom's work on the electron spectroscopy. But it had always been assumed that that was a trajectory-independent process, that it basically just depended on the distance of the ion from the surface, and not on the detailed trajectory. What we came to realize was that it depended on the detailed trajectory. So effectively, atoms on a surface not only produce an elastic shadow, but they produce a kind of inelastic charge neutralization shadow as well. You had to put that into the calculation in some primitive way, because actually if you didn't, you got completely wrong answers.
We had some interesting arguments with some of the practitioners in the field at the time, but subsequent events seemed to indicate we got basically the right idea. But in a way, although that's a very interesting piece of physics, it made the extraction of quantitative structural information much more difficult. It's much more recently that I got interested in significantly higher energy ion scattering, so-called MEIS - medium energy ion scattering, pronounced the Dutch way because they invented the technique - which is using something like 100 kV hydrogen or helium ions. We eventually managed to get funding for a UK national facility about four or five years ago. That was basically, I tried to get some money to do that, and then a guy called Dave Armour working at the University of Salford had vaguely worked in this area for some years, and he basically tried to get some money as well. Eventually, we were kind of put together into a single bid for a national facility that exploited rather fortuitously the fact that they were closing the thing called the Nuclear Structure Facility, a Daresbury lab, which was a large tandem Van Der Graaf for nuclear structure work. The injection system for that, the relatively low-end injection system, turned out to be a rather good accelerator.
MADEY: I see. So that's what you're using now?
WOODRUFF: That's what we're using. They had to invert the polarity so that it produced positive ions. Yes, that's turned out to be a kind of interesting development, which I think has been fairly fruitful in the last two or three years. We've really only just got going at trying to develop our methodology in the last few years. We've been looking very much at rather different systems, at systems associated with surface-alloy formation with things like antimony or copper and silver and lead on nickel and so on. So on these systems, where you actually appear to form two-dimensional alloy faces at the surface... when people first started finding that, it was a bit of a surprise. The antimony-copper-silver case is particularly curious, because it not only forms a surface alloy, but it appears to create a stacking fault between the alloy phase and the substrate, which is really certainly not something we were looking for. That's been an interesting development, as well. I'm trying to get more of a program working on surface alloy phases, which I think is an interesting area, again, from the surface chemistry point of view.
MADEY: So in the last few years, you've actually been spending a good bit of your time working at facilities away from Warwick.
WOODRUFF: That's right, very much so. Well, especially my students and postdocs, but that's certainly involved me in quite a bit of movement as well. That's right, yes. Actually, I would say right now that probably two-thirds of our work is based elsewhere. We have some projects also running at Warwick. Then we have relatively standard surface science instruments at Warwick on which you can kind of rehearse some of these experiments before you go and use this valuable beam time.
MADEY: Phil, I wanted to turn the discussion for just a second, as Phil Woodruff, together with Trevor Delchar, had written the landmark - how would I best call it … a textbook for surface scientists?
WOODRUFF: A graduate monograph.
MADEY: The graduate monograph is widely used in graduate education here in the US and I suppose in the UK as well. Perhaps you could tell us a little bit about how that text came to be.
WOODRUFF: Like most writing projects you ultimately get frustrated that somebody else hasn't done it. Now you even have to think you can do it better than anybody else, right? I had, curiously enough-- It's strange, but after my PhD, I'd worked for three years on this liquid-solid melt interface. I got fairly fired up about that and ended up moving into a different field and actually ended up writing a monograph on that, which was published by Cambridge University Press [CUP]. Then, several years later, when we started to get frustrated that you've got a new graduate student and there was really nothing to hand them that would get them off the ground. You could find some reviews on specific techniques or something, but there really wasn't a kind of reference point for that. I basically decided I'd like to have a go at doing something about that - a mission - and approached people I knew at Cambridge University Press and they said would they be interested in publishing that.
And then I felt it would be good to have a co-author with a rather different background. Trevor came from a much more chemistry background. So I thought that would be a useful complement. He had experience with techniques that I didn't have much experience of. So between us we managed to cover most things. There are a few things we just had to learn quite a lot about.
MADEY: How long did the whole project take?
WOODRUFF: I'm hard-pressed to remember. It was a few years. It actually started, I think, somebody approached me at one point about a handbook of surface science. This was going to have a lot of bibliographic data as well. We played around with that a bit but realized the amount of work involved was just huge, and somehow the reward - We found it's much more entertaining to write about techniques and to try and teach techniques, than just to kind of build bibliographies and handbooks and so on. So we kind of evolved the project, and that's when I went to CUP. Yes, it was a few years.
MADEY: I guess the book was updated in 1994?
WOODRUFF: Yes, it was something like that.
MADEY: Do you have any plans for further updating or expanding of the book?
WOODRUFF: Well, I feel the need, but I just need some time. I'd like to do it, and I can certainly see places where there are parts of the book that I think would change very little and parts that I think would change quite a bit.
MADEY: Would you actually teach a surface science course at Warwick from that book?
WOODRUFF: Yes, I do. Yes, basically from the book, although only from parts of the book. It's just a 15-lecture course, so I basically really talk essentially about LEED and Augeer and XPS and UPS spectroscopies, and not a whole lot else. That really takes up the 15 lectures, so that's really only a couple - I do a little bit of stuff on ion scattering. I think that's relatively simple, conceptually, anyway. It gives a bit of light relief from some of the quantum mechanics.
MADEY: One area in which Phil and I have had strong overlap in the last few years - and as a matter of fact, we also have commonality with our photographer today - is our involvement with the IUVSTA. Jim Lafferty and I are Past-Presidents, and Phil is the current President of the International Union for Vacuum Science, Technique, and Applications. Phil, perhaps you could tell us some of the factors that got you involved in the international vacuum science community.
WOODRUFF: I came in as a National Counselor for the UK through the British Vacuum Council. Maybe I should explain. The British Vacuum Council is a very different organization from the AVS. In particular, it has no members. It's an organization which was established, essentially at the time that IUVSTA was established, as an interface between IUVSTA and those existing societies in the UK that represented the kind of scientific and technological interests of IUVSTA. So it's essentially a committee of representatives from the Institute of Physics, from the Institute of Electrical Engineers, and we used to have the Institute of Metals, which sadly we've lost at the moment. And also, we now have the chemists in there.
I just came along to one or two IUVSTA meetings. What really fired me up - there were certainly aspects of IUVSTA that I found a little bit arcane - but it was about the time that Tony Van Oostrom, who was the scientific director, was trying to get funding out of the Union to set up this workshop program. We actually, with Alex Bradshaw as a matter of fact, put in a bid to organize the very first IUVSTA workshop, which I think you yourself attended, in Portugal. I saw that as a way that IUVSTA could maybe start to get much more involved in some real scientific initiatives. That happened right at the end of the triennium, just before the Cologne Congress. Tony Van Oostrom approached me and said would I be prepared to be the Scientific Secretary? I said, well, okay. Maybe that's somewhere where we could really get this thing going. My involvement -
MADEY: Tony was the Scientific Director.
WOODRUFF: He was, that's right. In fact, as events unfolded, I became much more heavily involved than I anticipated because, if you remember, Tony fell ill very soon after that and actually died only 18 months or two years later. In effect, for that triennium, I really became Acting Scientific Director as well as Scientific Secretary. But in that period, I saw that there was a role to be played and tried to help to develop the union. That was around about the time that you came President as well, which I think was also a significant change in recent history of a pattern of presidents where we actually had presidents that were very seriously active in research still. I saw that as a very positive move that I wanted to contribute to. Somehow that evolved in me being President.
MADEY: I believe that you had served one more term as the Scientific Director before being elected as President-Elect.
WOODRUFF: That's right, yes.
MADEY: What is the current situation with respect to the IUVSTA workshops? How many have been held?
WOODRUFF: I lose track, but it must be round about 30. It's turned into a very successful program, I think. We have six or nine each triennium, of that order. In fact, we basically agreed to slightly reduce the level of sponsorships so we can get more of them. Everybody seems to be very happy with doing it that way around. I think that it provides the seed corn money that allows these things to happen.
MADEY: These workshops, I guess, are held in all areas of vacuum science, techniques, and applications, all over the world.
WOODRUFF: That's right. Absolutely. Probably there were more early on in Europe, but in a sense, that's slightly where, in terms of the number of societies, the centre of gravity of the union is. But yes, there's certainly been several in Japan and I think some over here in the US.
MADEY: Yes. I think there's at least one upcoming in Japan in the near future.
WOODRUFF: Yes, that's right.
MADEY: Are there any major new directions or major new challenges that you see for IUVSTA in the coming triennia?
WOODRUFF: I think a problem - which I have to say I see as a long term problem, but never really figured out how to get to grips with it - is the one of developing nations and how best to help them. The trouble is that all the kind of help they need is basically financial. And one thing that IUVSTA does not have is a lot of money. It's very frustrating that you would like to find a way of being more constructive in that area, but it seems to be very hard to see how to achieve it. I do feel frustrated about that.
I think the other problem is one that probably the AVS has to worry about as well, which is, we have a divisional structure which is in many ways modeled on the way the AVS divisional structure developed. It's easy to create new divisions. But I think sooner or later, one has to look at have we got the right divisions and should we close some down or amalgamate them or rename them? That's a lot harder, to close things down than it is to open them up. We've been struggling with that for some time, and I still don't really see where we're heading. I think we have to keep talking about that.
MADEY: That's a challenge. And that's a challenge within the AVS also.
WOODRUFF: Yes, I'm sure that's right.
MADEY: Well, Phil, I think this has been a very useful and helpful discussion. Are there any other things that we haven't covered that you would like to comment on?
WOODRUFF: No. I think we've covered quite a bit of ground. I thank you for sympathetic interviewing.
MADEY: Thank you.