Commission on Electron Crystallography

Interview with Sven Hovmöller, joint winner of the 2020 Gjønnes Medal in Electron Crystallography

Sven, we would like to congratulate you on winning this prestigious award.

Thank you! Except for a poster prize once, this is my first scientific award, so I am very pleased. 

Could you tell us a little about your childhood and early education?

I was born in Sweden but as a Danish citizen because my parents had just moved here from Denmark. My father was a meteorologist and as such he worked abroad for year-long positions several times. When I was 12 years old, he worked at the WMO (think WHO but for meteorology) and the whole family spent a full year in Geneva, Switzerland. This was in 1959, before charter tourism had become common, and I had the great pleasure to stay in this very beautiful and exotic country! Perhaps the best was that I went to Ecolint (The International School) in Geneva, in a class with pupils from all over the world: Italy, Egypt, Australia, USA, Sweden, Thailand, Indonesia … . That made me decide I should work with something where I could work in different countries.

What or who inspired you to pursue a career in science?

When I was 14, I started having chemistry as a subject at school and I immediately decided to become a chemist. And I stuck to that dream, something I never regretted, although I later understood that there are numerous other fields of science that are also extremely fascinating. To mention just a few: astronomy with exoplanets, comparative linguistics combined with DNA sequencing as a way to understand how all humans are related and how we migrated all the time.

Originally, I wanted to become a biochemist. For a number of reasons, I had to change subject, department and even university three times and supervisor six (!) times, before I finally took my PhD exam in crystallography, in 1980. Two of my supervisors got new jobs, two got the Nobel Prize (Aaron Klug in Chemistry and Bengt Samuelsson in Medicine, both 1982) and one (Samuelsson) sacked me.

After repeatedly changing supervisor, I decided to finally study in a field where I could combine my interests in chemistry, computing and statistics (number crunching). I asked around and one day someone recommended crystallography. I had never heard of such a topic, but I tried it and I loved it.

When I was a child, my father would always demand the whole family to shut up during the weather report on the radio, which more often than not coincided with dinner time! He had decided, at the age of 8, to become a meteorologist. I always thought his main interest in life was the weather. But when he was over 90, he confessed to me that he wasn’t really interested in the weather at all! He was just interested in all the numbers! As a young boy, he had first seen the exchange rates in their local newspaper; so many numbers, and new numbers every day! At that point, he could have become a banker and the whole family would have become rich. But luckily, he soon found out that the local weather reports contained even more numbers, and so he decided to become a meteorologist. Now that I am 72 and have retired, I can honestly say that I never was very interested in chemistry, BUT THE NUMBERS! I loved them and I still love to look through tables with thousands of lines and dozens of columns, such as you can get in crystallography. And knowing that there are unknown scientific rules and principles hidden inside these numbers – yours to find out if you can just decipher the numbers!

And what attracted you to electron crystallography?

My very first PhD project was in biochemistry, trying to understand the mechanism of the ATPase of the mitochondria. At that time, 1969, almost nothing was known in this field. It wasn’t yet established how many polypeptides formed the ATPase complex, and even less the molecular weight or even the sequence of these polypeptides. When I started my PhD studies in structural chemistry (crystallography) in 1973, I wanted to determine the crystal structure of the ATPase. My supervisor told me that was too hard for a PhD project, so he had me learn the basics in crystallography by determining single-crystal structures of a number of small organic molecules, related to DDT. In 1973, there were a few hundred crystal structures of water-soluble proteins, but not a single membrane protein was solved. It wasn’t even clear what a membrane was. And then came the electron microscopy study by Nigel Unwin and Richard Henderson in Cambridge on the structure of bacteriorhodopsin! Like many others, I thought it might never be possible to make 3D crystals of membrane proteins. Many had tried, but nobody had yet succeeded. So I wrote to their boss, Aaron Klug, and asked to be accepted as a post- or predoc to work on bacteriorhodopsin. The positive reply said they didn’t need more crystallographers on that project at the moment, but I was welcome to work on the structure of TMV (tobacco mosaic virus) and its specific binding to RNA. I was already familiar with the work of Klug on EM studies of viruses, tRNA etc. so of course I was more than delighted to come to work with Aaron Klug. I shared an office with Richard Henderson and eight others, so I got to know a lot also about membrane protein research.

[OUP book]

After finishing my predoc and my PhD in 1980, I set up electron crystallography research in my own lab at Stockholm University, following the principles worked out in Cambridge. Later I developed CRISP and other programs for electron crystallography, based on the Cambridge system but much more user-friendly and faster. I thought it would be necessary to make it easy to learn and use electron crystallography, if this method would ever become spread within new fields of chemistry and physics.

What are the advantages and limitations of electron crystallography when compared with X-ray crystallography?

The main advantage is that the much stronger interaction of electrons with matter makes it possible to study crystals a million times smaller than those needed for X-ray crystallography. Recently, we have shown that with modern computer-controlled electron microscopes and very fast and low-noise electron detectors, it is also possible to collect complete 3D electron diffraction data in just seconds, i.e. much faster than on the most powerful synchrotrons. Another advantage is that unlike X-rays, electrons can be focused into an image. In crystallographic terminology: we can observe experimentally not only the amplitudes (as in diffraction) but also the crystallographic structure factor phases.

Does electron crystallography have a special role to play in coronavirus research and other global problems?

I think most people on the planet see the same background illustration on the television news every day: a coronavirus particle. This is in fact an electron crystallography image, or rather the result of a 3D reconstruction from several EM images of the virus particle. It is today much faster to obtain a 3D structure of huge biological objects, such as viruses, by electrons than by X-rays. The last decade has seen the resolution in electron crystallography reaching beyond 2–3 Å, making it possible to get density maps so detailed that the polypeptide chain can be traced. Obviously, to have a 3D picture of the object you study is a great advantage when you want to plan a strategy for making a vaccine, or you want to understand the mechanism of the virus, in atomic detail.

You and Ute have been members of the IUCr Commission on Electron Crystallography (formerly IUCr Commission on Electron Diffraction). What would you say is the Commission's most important function?

  • Educate next-generation electron crystallographers by arranging schools in electron crystallography
  • Promote sessions in electron crystallography at crystallography conferences
  • Encourage international collaboration within the fast-growing community of electron crystallographers
  • Arrange visits by PhD students and scientists to each other's labs
  • Make sure the methods developed in the various labs become user-friendly and available to the entire community, with X-ray crystallography as a role model.

You have been commended for your role in raising a generation of electron crystallographers through your international schools, many of which have been supported by the IUCr. With the necessary transition to online meetings in 2020, how do you see the future for face-to-face schools?

Nothing beats personal face-to-face interaction. Sometimes even a short conversation with an established person in your field can change the course of life for a young student.

This question also touches on a dark chapter of electron crystallography. I would like to take this opportunity to tell especially young scientists about it. It all started half a century ago.

During the 1960s, Aaron Klug developed optical diffraction of EM images of crystallographic samples. If you shine a laser on a periodic object, you will see a pattern of diffraction spots. You can try this at home: use a laser pointer and shine the laser light through any piece of textile. What you see is actually the Fourier transform of the image. At that time, there were no computers powerful enough to calculate Fourier transforms. Even 10 years later, in 1979, the calculation of a 256 x 256 Fourier transform was an overnight job at the mainframe computer at Cambridge University. Already in 1971 Klug wrote “The methods are closely analogous to those used by X-ray crystallographers to measure periodicities or degrees of order and finally to see the arrangement of atoms in simpler substances. The difference, however, is that there is no phase problem here because the phase information is contained in the electron micrograph” [Phil. Trans. Roy. Soc. Lond. B 261, 173–179 (1971). III. Applications of image analysis techniques in electron microscopy. Optical diffraction and filtering and three-dimensional reconstructions from electron micrographs]. Clearly, Aaron Klug understood this fundamental difference between X-ray diffraction and EM images, already half a century ago, and he received the Nobel Prize in Chemistry in 1982 partly for this. Yet, a large part of the community in physical sciences working on EM was not aware of this. I tried for decades to spread the insights of Klug, as well as the spectacular EM work on membrane protein crystals by Unwin and Henderson, to everybody but with little success.

The scientific community working on EM images of inorganic compounds, such as minerals, semiconductors etc. always claimed that “the phase information is lost in images”. I, quoting Aaron Klug, claimed that “the phase information is present in images”. Any crystallographer will understand the magnitude of a disagreement on this point. When the IUCr in 1993 arranged a Summer School on Electron Crystallography at Tsinghua University, I wasn’t invited. I had taken the effort to fly to Arizona to introduce image processing by Fourier techniques to John Cowley, the honorary chair of this School. Cowley is considered as the father of electron diffraction physics. He stopped me when I said “and then we calculate the Fourier transform of the EM image, and at each diffraction spot we can now directly read out the phases”. “Wait a minute, let me think.” And then after a short moment, “Yes, you are right, please continue”. Obviously, Cowley was impressed by the demonstration and said “You know, next summer there will be an IUCr summer school on electron crystallography. I think you should go there to lecture. Would you be interested?” I was of course exhilarated and said yes. To be honest, my whole trip to Arizona was done only for the chance of hearing those words. Cowley wrote to the other dozen or so invited speakers to ask if I could be added to the list but got the majority vote NO. But I could come to give a computer lab. When I came to the school, there were no computers, no room for computers and no time allocated in the program for my computer lab. Luckily, my then PhD student Xiaodong Zou was also at the school. She immediately rented bikes for us and we cycled to the nearest shop and rented a couple of computers and huge displays. We could then stand in the corridor and run our lab in the coffee breaks.

[Erice 1997]

Electron Crystallography School 1997 in Erice.

The school was then repeated in Bristol in 1994. Two members of the IUCr Commission on Electron Diffraction (as it was then called), Douglas Dorset and Boris Zvyagin, both pioneers of electron crystallography, suggested I should be invited as a lecturer to the Bristol school. The matter was discussed but after one member said “Hovmöller should not be invited because he confuses the students”, I wasn’t invited to that school either. This time I didn’t attempt to sneak in through the back door as I had done in Beijing. Instead I agreed with Xiaodong that we should just arrange our own schools. We then arranged a school at Stockholm University, with the students living in student dormitories and the conference dinner in our home with Xiaodong and her daughter Anna (4 years old) singing traditional Swedish Lucia songs. It was a success. Soon after, people who hadn’t been able to come asked for the next school. We had only planned to have one single event but it became one school after the other, first twice per year and then annually. After about 10 schools, we realized we were becoming something like Ludovico Riva de Sanseverino, who spent half his life arranging the IUCr schools in Erice. We then announced that we didn’t have any patent on making electron crystallography schools but we in fact would welcome any initiative to arrange schools in other countries as well. We would of course be happy to provide lectures, lecture notes, labs and suggest speakers, but the local organisers should arrange the hotels, meals and lecture rooms. This went very well, and by now these schools have been arranged in a dozen cities, most of them in Europe, but some occasional ones in USA, Korea and China. Gradually, local organisers, including Ute Kolb, took over all the duties. A very special highlight was when we managed to convince Ludovico to arrange one of our schools in Erice in 1997! From then it has become a tradition to have this Erice school every seven years (held in 2004, 2011, 2018, the next one in 2025).

How important is the recent initiative by the European Crystallographic Association and other organisations to provide childcare facilities during their meetings?

That’s a very good initiative. My first international conference was in 1973, and then the social program was still called “Ladies program”. At the International Zeolite Conference in Montpellier in 2000, my wife, Xiaodong Zou, had already overtaken me in terms of scientific career so I accompanied her to the conference, looking after our new-born, Linus, so Xiaodong could keep breast-feeding Linus and I could take part in the “Ladies program” (which by that time actually was called Social program in every conference).

What other interests do you pursue?

I am mainly involved in political and social issues since I retired six years ago. I have started a “Chemistry Club” mainly for exceptionally gifted children, aged 5–12. We give lectures at the university level for these kids, and get hundreds of questions during a two-hour lecture. Most lectures are on chemistry, but some are about physics, mathematics or biology. One evening I presented in three Asian languages: Chinese, Japanese and Korean. However, they only had to learn to read Korean in those two hours.

At the other end of the spectrum, I chair a small organization helping Romanian beggars in Sweden to find shelter, learn to write their first name and later to speak and read Swedish, to find a job etc. The European Roma are among the poorest and most discriminated people on Earth. Even today, millions have not attended school for a single day, and hardly any more than eight years.

Of which of your professional achievements do you feel most proud?

I am proud and content that, with others, we have managed to develop electron crystallography from a small field into a mainstream method. In the last decade, we have gone from a very time-consuming (years) and demanding field for 2D projections only to ultra-fast (minutes) 3D structure determination of very complex structures, including proteins. In fields such as zeolites, metal–organic frameworks and pharmaceuticals, important structures that have remained unknown for up to 100 years have now been solved by electron crystallography. Typically, they form crystals too small for single-crystal X-ray crystallography and too complex for X-ray powder diffraction. Some are multiphasic and/or full of defects. For electrons, even the smallest powder particles are single crystals.

[Sven and Linus]

Sven Hovmöller and his son, Linus Hovmöller Zou, solving the crystal structure of PD quasicrystal approximants.

What has made me most happy and proud was when our then 9-year old son, Linus, in 2010 solved a set of very complex quasicrystal approximants, called pseudodecagonal (PD1, PD2 etc.). They are made synthetically in alloys of composition ⁓Al70NixCo30-x. Together with three collaborators, I had spent half a year full-time trying to solve these structures but with no success. Every summer for about eight years, I had more and more desperately tried to solve them, but no. We had hundreds of HRTEM images and electron diffraction patterns, but the structures resisted being solved. Then in 2010, I gave it a try again and got stuck. It was during the summer vacation and Linus was around at home. I asked him “Linus can you come here for a while and help your dad?” He sat down at our huge kitchen table, full of diffraction patterns, EM images and computers. Linus is very clever at pattern recognition and his head was not as completely muddled as mine from looking at and remembering hundreds of these images. So he could object to my suggestions and after a full 8-hour day, we had solved the first two PD structures! We continued the following day and finally managed to solve in total four of these very complex structures. We decided to have Linus included as one of the co-authors, given his critical contribution to finally solving these structures. I told the editor that one of the authors, Linus Hovmöller Zou, was only 9 years old. It took a year before it was published and then the editor informed me and also wrote “We mentioned in the press release that one author is only 9 years old.” And then RING! The phone was on fire, with calls from the BBC, The Times and many Swedish and international newspapers. They got interviews with Linus, in English. We were on television. You can still see the BBC reportage by Googling BBC Linus Hovmöller Zou. As it turns out, Linus is the youngest boy ever to co-author a scientific paper. Only an American girl was a few months younger.

Thank you very much for your time and congratulations once again.

Thank you!


These pages are maintained by the Commission Last updated: 15 Oct 2021