A career in NMR: An interview with Gareth Morris

Abstract

The short answer of course is that I did not. Like a lot of the best things in life, it happened by accident. When I was an undergraduate student in my first degree, which was chemistry, I had to choose between various research projects. One of them was with my then Physical Chemistry tutor who I knew was very good, and it involved Fourier transforms, radio frequency electronics and computing. Those were all things that I'd known something about from interests that I pursued in secondary school. I am afraid that I thought “if I do that project I can carry on being a lazy student.” Then I found out that actually research was great fun and I worked harder than I had ever worked in my life on it, but it was a stroke of enormous good luck to be there at the right time, when multiple pulse NMR was just starting to blossom, and working for a wonderful supervisor who was very supportive and very flexible. The first one that worked was DANTE and that came about partly by accident. I had been trying to teach myself about Fourier Transfer NMR, about the theory of NMR, and in particular trying to understand the way that the signals we excite depend on the form of excitation we apply. I had come to the conclusion that if we were to apply radiation in the form of short, weak, equally spaced pulses that we could actually achieve selective excitation. That's now something absolutely routine, we have got much better ways of doing it, but at that stage it was a very strange thing to want to do because the whole point of Fourier transfer NMR was that it excited all of the frequencies at once. I had been musing on this idea for a whilst, trying to do calculations and in particular trying to understand the relationship between the non-linear spin response and the engineer's view of NMR, which would be linear systems theory. I talked a little bit about it to Ray and he'd been a little bit discouraging, saying that “the NMR response is very nonlinear - I don't know how useful these ideas are.” So I did not actually do any experiments until something else that I had been planning to do suddenly went wrong and I found myself with about 2 or 3 months of time left to finish my undergraduate project. With no results and no prospect of any results, the next day I came into the lab and programmed the pulse sequence. It worked beautifully and a few weeks later Ray talked about it at the ENC. It was very, very quick, and it is very nice when something works first time. That is a little bit like asking, which is your favourite child? It is a dangerous question to ask. One that was special was INEPT. This came out of writing my doctoral thesis, as part of which I had done experiments on heteronuclear correlation and everyone had been surprised at how quick these experiments were. These were relatively low field, 80 MHz proton, experiments on carbon-13 at natural abundance, and yet we could get good spectra in 2 or 3 minutes. So as part of my thesis I wrote an analysis of why the sensitivity was good. Seeing that one of the effects was that polarisation transfer from proton to carbon-13 it was then obvious that we could get this sensitivity advantage without actually doing a two-dimensional experiment at all. Again, it was an experiment that came very quickly and worked very quickly (and was something that I did on my own as Ray was taking a “French holiday,” which means that you just disappear when the weather is nice. We were on our own in Oxford for a month over the summer and I just did this experiment and it gave a nice result. Ray came back and I think I showed it to him and to Howard Hill. Howard said that would be nice for nitrogen 15, and we wrote the paper up. The least successful experiment was almost certainly the one that led to DANTE. I should explain that at that time Ray had an unusual approach to undergraduate research projects. When you arrived in his lab he would normally disappear, and so the student would ask other members of the group “Will Doctor Freeman be in soon?” and we would say “No, well, he may not be in for a week or two - why don't you just learn how to use the machine and learn a little bit about the area.” After about 2 weeks you would see them realise that Ray wasn't going to turn up until they had thought of something useful to do. It was a heavy burden in the first few weeks of a project, but it meant that every student that came through did something original, did something new, and of course Ray was always there to help out and to make sure that they got a good result in the end but they felt a real sense of ownership of the project. When I arrived, before this policy really got going, Ray just said “Oh, get to learn how to use the machine, see how you get on” and so I went to the library and I worked out a nice project for myself; I was going to build a new piece of equipment, a high-powered radio frequency amplifier which was designed to do spin locking for measuring T1rho in a high-resolution machine, which has less high power. The idea was that it would have a power supply which would quickly ramp down the voltage so that the RF power would decrease during the spin lock, but that would mean that you started at a high spin lock power and then you can move down to a low spin lock power at which you can maintain the lock without generating oscillations in the signal. It was a perfectly reasonable idea, and on modern equipment it would work fine, and I built the equipment. I think it was February 28, 1976; I came into the lab late one evening, connected up the equipment, hit the switch to change the timing circuitry from the standard Varian one to one that Geoffrey Bodenhausen had made, and the amplifier blew up! The next morning I came in and programmed the DANTE experiment, because if that did not work I really would have no results, and one of those lucky chances, a failed experiment, pushed me to do something new and that turned out to be very useful. I'd say “Thank you.” They have given me a wonderful scientific life over the last 40 years. There is a cliché in English that the one thing that Napoleon asked of his generals is that they should be lucky (I have never met a French person that recognises the quotation!). The most important thing in science is to be lucky. Also important, of course, is to recognise your luck when you see it. There have been so many times when people have failed to realise when they have found something important. Be lucky, know when you have been lucky, and be grateful. I should say lectures from Richard Ernst or something like that, but the truth is actually my first conference. This was a NATO Advanced Study Institute, and it was being held on the North Coast of Sicily in a new hotel in 1976. I had spent the summer travelling around Turkey with a friend, and then came back overland by train. I was working my way down Italy, Rimini, Arezzo and various other places, and everywhere I went, I saw posters on the walls that had the name of my hotel and the date of my conference on, but the posters did not say the NATO Advanced Study Institute on Less Receptive Nuclei, they said La Donna Ideale (The Ideal Woman). I thought “I am going to get this conference only to find that it has been cancelled and this other event has taken over”! There was no email in those days, so I would have no way of knowing. In fact when I got there I found that it was a big hotel complex and the two events were running in parallel, and it was a surreal experience. On the one hand, you had a bunch of scruffy academics talking about NMR, and on the other hand, you had all these beauty queens and their minders, and swim suit contests, table laying contests, it was like something out of the 19th century but it was enormous fun, and I think that most of the people who were at that conference still remember it. In one sense that is a very easy question, because the most rapidly expanding fronts of knowledge nowadays are biological, but when we choose a research field what's more important is where we can make a contribution. If everybody is running to bioinformatics then maybe there is less opportunity to carve out an individual career, so I do not think that is at all an easy question to answer. Of course most people answer it in the same way I did at that stage, which is pretty much at random. Things are governed by chance. Certainly biology, the relationship between structure and function; I am sure that there are great advances to come in our ability to use computers to explain structure and function. Perhaps some unexpected directions, you just have to look at the success of Google Translate: decades of work went into attempting to write logical, structured, intelligent attempts at machine translation, and they were largely unsuccessful. Then someone came along and just said “well, let's just throw the corpus of translations at a machine and let's see what it comes back with.” I am sure that machine learning has a lot to teach us, but whether we can understand its lessons is a more difficult question, I think. I enjoy music, although I do not have time to play much anymore; and I try to stay sane by growing vegetables. Quite often before work I will go and dig potatoes or hoe the beans or whatever. I would make a real attempt to understand the current NMR manufacturer's technology. When I was a young graduate student I got to know probably as much about the inside of a CFT-20 spectrometer as many of the engineers who worked on it, and for the first 10 or 20 years of my career I could be fairly confident that I understood pretty much everything that was going on in a spectrometer. But now, there are no circuit diagrams, you frequently cannot get access to source code, and in any case there are very large quantities of code. It is much more difficult to get a real understanding of what is happening, and yet it is still enormously important, because progress often comes from identifying the limitations of equipment and finding ways to circumvent them. If I had a year I think I would probably approach Bruker and say “look, how much are you willing to share with me? I really would like to understand what is going on in my machines, to see if I can make them work better.” I think if I'm allowed to take someone from the past I would take Johan Sebastian Bach—somebody who in his music, it seems to me, embodies most of the best things about humanity. There's a balance between emotion and intellect, between profundity and playfulness. As experimental scientists, we live with imperfections. Whenever we do an experiment, we know it is imperfect—things go wrong, our understanding, our insight, is imperfect—and in the music of composers like Bach and Mozart we see at least the possibility of perfection. I find that inspiring. If I knew I would not be telling! No one ever got rich predicting the future of NMR. Many years ago, I was invited to visit Varian in Palo Alto, and I drove down from the University of British Colombia. I drove down I-5 all the way to Palo Alto and there was another visitor there at the same time—I think it was Luciano Mueller—and they sat us down and they asked us; “do you think it is worth Varian building a small angle phase shifter for the spectrometers?” We both said “It would be really nice to have that, but I guess not many people would use it”—and now of course this is a standard feature on every spectrometer on every radio frequency channel. One of the beauties of NMR is that it constantly surprises us, and I never know from year to the next what I am going to be working on. I have very little idea on the future of NMR but I am modestly confident that it will remain interesting and relevant. Interview by Roberto R. Gil, Laura Castanar, and Paul Trevorrow. EUROMAR, Nantes, France, July 4, 2018