AEM faculty spotlight:
When engineers design new bridges, they consult reference tables. AEM Professor Richard James seeks to create such a reference on a slightly smaller scale, that of atoms and molecules. James, a Distinguished McKnight and Russell J. Penrose Professor, has been interested in developing unusual new materials. Recently, his work has been motivated by his studies of the virus bacteriophage T4. Viruses like T4 are built of components that he calls "objective structures." To picture an objective structure, imagine a group of identical people sitting in chairs that are arranged in a perfect circle. Each person sees exactly the same environment. Professor James discusses these and more in the following Q&A.
What kinds of research interest you these days?
My work is about the discovery of new materials. In my lab we make alloys from elemental materials, guided by theoretical developments. My students and I develop theory to understand the behavior of materials. We take theoretical predictions to the lab and make the materials. As you can imagine, it can be incredibly exciting, as when we actually discover a new material with some unexpected properties, or deeply humbling, when we realize that we have left an important factor out of the theory.
I hesitate to expose the deeply humbling moments, but much of our work is on shape memory materials, where a large change of shape is induced by changing the temperature. This behavior is caused by a phase transformation. Recently, guided by theory, we realized that if a material would have such a phase transformation, and also have the property that one phase is magnetic and the other is not, we could make an interesting energy conversion device. It is a fantastic material: put it on the table and it is non magnetic. Heat it up just a few degrees and it becomes a strong magnet. Now imagine wrapping a coil around this material. As it is being heated through the phase transformation, the magnetization changes rapidly, dM/dt is large. From the relation B = H + M, the quantity dM/dt is partitioned between dB/dt and dH/dt, where H is the magnetic field and B is the magnetic induction. This partitioning is governed by a theory called micromagnetics. It turns out that with a well-chosen shape of the specimen the big dM/dt produces a big dB/dt. Then, by Faraday's law, dB/dt = - curl E, an electric field E is produced. This field drives a current in a surrounding coil. So, we've produced a new device that turns heat directly into electricity.
How would you look for such a material?
Partly, we rely on collaborations with physicists who do "first principles calculations" to help identify favorable structures for properties like ferromagnetism or ferroelectricity. Another crucial aspect of these materials is reversibility. Lots of materials have phase transformations, but few are highly reversible. Of course, it is much more interesting when we have this high degree of reversibility, as one can then go easily back and forth between the phases. One of our recent breakthroughs is a new understanding of reversibility, completely contrary to what is written in textbooks. It is a quantitative understanding, so we could go to the lab and systematically change composition to achieve these special conditions that theoretically give high reversibility. We've been doing this a lot, and the results are amazing: in some cases the size of the hysteresis (a measure of reversibility) can be decreased by a factor of more than 10. There are fascinating new families of shape memory alloys that are emerging from this work. My postdoc Vijay Srivastava discovered the magnetic alloy I spoke about earlier by using this strategy.
Tell me a bit about objective structures.
This is a new direction in my research, and, as usual, I am beginning with the development of theory. Imagine a structure which is built of identical molecules, each with, say, 100 atoms, numbered 1 to 100. Imagine that you sit on the 27th atom of one of the molecules. Take a look around. Nearby, you see other atoms of the same molecule. Further away, you see other atoms of the other molecules. Now sit on the 27th atom of a different molecule and, again, look around. If you orient yourself in just the right way, you see exactly the same environment, just as in the circle of chairs. If this happens for corresponding atoms in every molecule of the structure, then I call that an objective structure. Lots of the parts of viruses have these structures. But also, most of the structures adopted by elements in the Periodic Table are objective structures, with one atom per molecule, including exotic things like buckyballs and carbon nanotubes.
What can we learn from such structures?
For one thing, these are the natural structures to search for special physical properties like ferromagnetism and ferroelectricity. For ferromagnetism, if one atom wants to have unpaired spins, that is, to be magnetic, all of the atoms will also want to be magnetic, because of the way objective structures are built. People in biology take natural structures and study them: it is considered uninteresting in biology if one violates conditions that occur in vivo. But I like to think in more of an engineering way about this, and the first step is to understand how they're put together. Objective structures offer a special window into the building of molecular structures.
How would someone go about building something like a virus?
When engineers build a bridge, they go to tables of I-beams and trusses and, from stress analysis, they understand how to use these in a reliable, efficient structure. What I would like is to have that kind of quantitative information for molecules, so you could build structures out of molecules like you build bridge structures. We think that at molecular level, the useful analogs of I-beams and trusses are objective structures. We also think we know how to build them, in principle. The method is called self-assembly. We arrange the positions and orientations of bonding sites on the molecule to be given by the formulas we have developed for objective structures. We find that such "designed molecules" have an amazing tendency to self-assemble. We have a long way to go, but we think one might be able to build a structure which exactly matches some part of a virus. By doing that you may be able to make it so the virus naturally binds the structure, thereby disabling the virus. You might be able to build a tubular structure whose molecules exactly match the inside of a carbon nanotube. Suitably functionalized, it could be a template for the large-scale growth of carbon nanotubes.
How do you find objective structures? Do they all look like carbon nanotubes or viruses?
No, there are many others! Recently, together with Prof. Ryan Elliott and our postdoc Kaushik Dayal (now Assistant Professor at CMU), we set ourselves the task of finding every objective structure. I was on sabbatical in Germany, and I think I invested more than 1000 man-hours in this one calculation! But, we succeeded in finding an explicit formula for every objective structure. This is a basis for a systematic study of their properties.
Last Modified: Wednesday, 05-Oct-2011 13:22:02 CDT -- this is in International Standard Date and Time Notation