Odile Eisenstein tells Nicola Nugent about the power of computers in chemistry

Odile Eisenstein

Odile Eisenstein is the head of theoretical chemistry at the University of Montpellier 2, France. She specialises in the use of quantum theoretical methods for the study of catalytic mechanisms. She is an editorial board member for NJC and is on the advisory editorial board of Chem. Soc. Rev.

What inspired you to become a scientist?
At high school, I had an outstanding teacher in physics and chemistry. Plus, I was not bad at science on the whole. It was also clear that there were jobs there! 

What motivated you to specialise in theoretical chemistry?
I studied physics and chemistry, but it turned out that I wasn't very good at physics. So I registered in chemical physics where I had two outstanding professors. One taught a very modern topic at the time, called the Woodward-Hoffmann rules. I thought that this was so good that I had to work with Hoffmann. I became a postdoc with him a couple of years after my PhD with Nguyen Trong Anh and L. Salem.

I like the logic of theoretical chemistry, and the possibility of creativity with images. I felt I had the ability to be accurate with drawings - it wasn't just solving equations. It involves a special blend of mathematical and physical accuracies and qualitative descriptions, often in graphical forms.  This blend is very fulfilling and incredibly powerful for solving chemical problems and communicating the results to a large audience.

I really feel that I am a chemist; I just use computers in place of using experiments. I feel I know the experimental community even better than my own community.

What kind of things do you work on and what are the applications of your research?
I work on catalysis. In general, transition metal chemistry, structure and reactivity, which helps to understand mechanisms. I am not working on a specific type of catalyst. I am interested in anything where I can try to solve the mechanisms with a computational approach. This covers any transition metal and even lanthanides and actinides. I have no prejudice or preference for any element!

What can experimental and theoretical chemists learn from each other?
A lot! I recently got an award from the Dalton Division, so the fact that it was actually an experimental community that gave me an award might be proof that I was useful! There is great need for these two communities to work together and in general this helps to advance the overall field of chemistry. 

What impact have the huge advances in computing had on the work that you do?
I started using very simple qualitative approaches in which I could find conceptual ideas but could not prove them with hard numbers. We were then able to put better numbers into the model, but the models were sometimes very far from the experimental system. For example, phosphine was always PH₃, which is never the phosphine used experimentally. Now we try to zoom in on the real system - why does it work with one molecule and not work with another? 

Of course, very often the solvent is not really represented in an accurate manner. We know there are years of improvement before we really are looking at macroscopic systems.

What will be the next big break through in your field?
The goal is to link the microscopic events to the macroscopic measurement. Typically we look at one molecule, but the reality is that you have Avogadro's number of molecules. It is a big challenge to actually relate one to the other. This will need considerable advances in methodology. The big difficulty right now is to understand entropy change, which will enable us to understand free enthalpy change. If that is possible, we will be able to work with the solvent, and we will be able to look at lots of molecules. A major challenge is the fact that molecules are not isolated in the gas phase at zero degrees Kelvin!

With unlimited computational power, what would be the limits of theoretical chemistry?
Chemistry is much more complex than what you can put into a computer. The computer and the theoretical chemist will never replace the experimental chemist. I am not interested in that world - it should not exist! There will always be something that won't have been put into the model.

What do you think would be the most rewarding aspect of your career over the years?
Involvement with young researchers - students and postdocs. It's nice to see them become independent and successful. And to know that you have inspired them. I would be very unhappy if I left only papers.

What advice would you give to a young researcher at the beginning of their career?
One piece of advice was given to me by Roald Hoffmann - he said you should work well on a topic you like, then learn from that and expand. Don't think that you are immediately going to work on an important topic.  Work on a specific point, which has to have some interest to people, and do it well. Quality is also important. You have to be known for quality work and to be trusted. Produce a good paper that has depth is my biggest advice.

If you weren't a scientist what would you do?
I like movies and spend quite a lot of time in the cinema, but I'm not sure that I would like to be involved in the movie industry. I have no special skill but I am interested in visual art.  I do a little painting and I collect art.