R. Stephen Berry, James Franck Distinguished Service Professor Emeritus of the Department of Chemistry and the James Franck Institute, MacArthur fellow, Guggenheim fellow, Sloan fellow, American Academy of the Arts and Sciences fellow and former vice president, member and former home secretary of the National Academy of Science, member of the American Philosophical Society, foreign member of the Royal Danish Academy of Sciences, former special advisor to the director for national security at Argonne National Laboratory, is the author of many articles and books, including Understanding Energy: Energy, Entropy and Thermodynamics for Everyman (1991), Water and Energy as Linked Resources (co-author, 1978), TOSCA: Optimizing the Mix of Fossil and Nuclear Plants from Total Social Cost (1979), and Physical Chemistry (1980). Still actively seeking answers to a range of questions daily, he also plays the recorder and collects art.
What are you working on these days?
I’m not working very hard because I don’t have graduate students or postdocs anymore, but the main thing I’m working on is using an approach we developed to connect the micro and macro approaches to physical problems by examining phenomena that are well-described by the macro picture but break down for very small systems. We got into this because we were studying atomic and molecular clusters for a long time. Some early simulations showed—and we verified and explained and explored the phenomenon—that very small systems, clusters of atoms or molecules, 10-20 atoms, do not obey what had been accepted as a universal rule, the Gibbs’ phase rule, discovered by J. Willard Gibbs at Yale. It is absolute doctrine—the simplest formula in all of natural science. It just says that the number of degrees of freedom that you can vary is given by the number of individual components in the sample minus the number of phases (solid, liquid, gas, or different forms of solids), plus two. That was a great discovery, and it’s a great mystery.
Where does the two come from?
The two gives the right answer! It’s this rule that says that water—liquid and ice—can be in equilibrium at only a single temperature at a given pressure. It’s exactly 32 Fahrenheit or 0 Celsius at 1 atmosphere of pressure. Change the pressure, and you have to change the temperature. But it turns out that very small systems violate this rule, and you can have solid and liquid in observable amounts over a range of temperature and pressure. And so we were able to explain this on the basis that the Gibbs’ phase rule, which is valid for large systems, is specifically a consequence of large numbers. When numbers are large—you’re typically dealing with 10^10 to 10^20 atoms (a grain of sand is 10^20 atoms)—although formally the passage from solid to liquid is smooth, it actually occurs over such a tiny range that you couldn’t possibly observe anything except at the point where the two are in equilibrium. The Gibbs’ phase rule is exact in the sense that the smooth change between phases is in an unobservably narrow region. But when you have very small systems, then the difference in free energy between the solid and liquid when they’re in disequilibrium can be small enough that you can observe both phases—you can have water crystals and water liquid in equilibrium in a grain, in a tiny particle, over a range of temperatures and pressures, which the Gibbs’ phase rule says you can’t do.
We worked on a variety of things and figured out a basic reason for this, and then we got around to asking a question that we could have asked much earlier on: what determines the largest size of the smallest system for which you could actually observe the violation of the Gibbs’ phase rule? What fixes that approximate size depends on what your capabilities are for making observations. But if you say, for example that I could see 10% of the unfavored form in the presence of the favored form, that tells you that if you could see the violation of the Gibbs’ phase rule with up to 75 atoms, but you wouldn’t be able to see it with 100 atoms. It’s too sharp. The range where the solid and liquid coexist is too narrow to be observable. The governing factor that determines that maximum observable size is the inverse of the entropy change between the phases. When you go from solid to liquid, the entropy increases. One over the entropy is the physical quantity that determines how big is the biggest you can observe.
At the last session of [physics professor] Leo Kadanoff’s 75th birthday meeting, someone from Urbana said, “You know, one of the big open questions in physics is matching the macro and micro approaches.” We had just recently published our first paper on this maximum size phenomenon, so I had this sudden realization that we had found a way to address the question of matching the macro and micro by asking where the macro broke down. We found it for one phenomenon, but there are lots of phenomena for which the macro description doesn’t work for small systems. For example, the solid form of almost all pure substances is a crystal. But when you get to very small systems, you don’t get crystals; you get polyhedra or icosahedra. An icosahedron is not a crystal. It’s not infinitely repeatable. You can have 13 atoms, for example, with 1 in the center and 12 around it in the form of an icosahedron, or you can have a second icosahedral cell that gives you a total of 55 atoms, and you can build icosahedral shells out and out and out, but those aren’t crystals. A crystal doesn’t have a center. At what size does any given substance take on its crystalline bulk form instead of its small polyhedral structure for clusters? If you have a crystal and you add 1 or 10 or 100 more atoms or molecules to it, it goes linearly with the amount of stuff. But when you have a polyhedron, you don’t have monotonic behavior because you get closed shells, and you add things onto the outside of the closed shells, and the behavior is non-monotonic. When the structure depends sensitively on the number of things, it isn’t a monotonic dependence anymore. When does the monotonic behavior break down? These are open questions.
Most small systems melt at temperatures well below the bulk temperature. At what size do they start to melt like the bulk? Martin Jarroldand his group in Indiana found that there are a couple of strange materials—tin and gallium—for which small clusters melt at temperatures much higher than the bulk. It’s weird! They actually did experiments looking at the size dependence of that melting, and they found the approximate size at which things start to melt like the bulk, but their experiments didn’t answer the question of why. It’s basically that the nature of the bonding depends on the number of atoms you have and the way the atoms bond together.
Is there a similarity between gallium and tin?
Eh, both weird! The basic big question which I find challenging for any given phenomenon is, how do you approach the macro-micro question by asking where does the macro break down—and why?
How long have you been working on the macro to micro problem?
A dozen years at least. I can look it up.
So hardly your whole scientific career—
Oh, I’ve been working on all sorts of things. For example, one area that I worked in for many years, though I’m not working on it now, though I certainly haven’t lost interest in it, was the one area that I knew that I would never study! Absolutely!
What is it?
Thermodynamics! Because when I finished grad school I knew absolutely for sure that subject was completely finished. There was nothing more to be done! And it was lovely, and it was elegant, but it was finished!
How could you say that?
It was so elegant. My interest in it only arose when I moved to Chicago in 1964. And it was a consequence of my dislike—my actual anger—at the level of pollution, air pollution, especially, in Chicago. It was a filthy city! The air was gray, and it smelled bad. It was toxic. It was really awful. And I got very angry! I thought I was psychologically prepared for it, but when I got here and lived in it, I realized I wasn’t. I actually wrote a letter to Mayor Richard J. Daley which started out, “Dear Mayor Daley, you live like a pig.”
Did he write you back?
The letter got more rational after that, but I said I had heard that there were things being done to try to address the pollution issue, but I didn’t see any signs of it. I had the intuitive sense to send a copy of it to our alderman, who was Leon Despres. The reply I got was an invitation to me and to Alderman Despres to visit the city’s air pollution facilities.
Were there really facilities for air pollution?
There was a laboratory. And it was so clean and neat and elegant, it was clear nothing was happening. At that point, I began to get very involved as a citizen at a time when there was an actively growing interest in addressing the environment. In the 1960s, there was a whole societal tipping point, and environment and pollution changed from being for nerds only to being politically acceptable. It culminated at the end of that decade in the Environmental Protection Act and the Clean Air Act. But I got involved with citizens’ groups, even testifying at the city council. It was pretty interesting.
They said they didn’t have enough staff in the environmental department to monitor air pollution sources, so we made a proposal to train volunteers. They wouldn’t have any enforcement authority, but they could report to people who did. So we proposed this at a meeting at the city council, and they carefully scheduled so we were the very last people to speak that day. There were only three aldermen left, one who was chairing the meeting—a charming Irishman who later went to prison (he was actually a friend of Mayor Daley, but when he was convicted, Daley wouldn’t stand up for him at all)—Mike Bilandic, who later became mayor, and Leon Despres. I made the proposal, and Bilandic said, “I think that’s a really dumb proposal.” But this Irishman from the north side winked and looked at me and said, “I think it’s a pretty good proposal!” Whereupon, Bilandic said, “Hey, that’s a good idea!” But it never happened. Anyway, it didn’t need to, because the federal law changed!
About that time, I realized if you wanted to address the problem, you needed to do more than put precipitators on power plants. You had to somehow use energy more efficiently. So we began by looking at how energy was used. The first thing we did was a study of the energy used to produce the automobile, starting with the ore in the ground and ending with the final disposal. We weren’t looking at the transportation aspect; we were looking at the manufacture and the disposal. We looked at each step and compared the actual energy and free energy with the ideal thermodynamic limit with the idea that where the difference was greatest, you had the greatest leverage for technological change. So that was the first study of what became life cycle analysis. And we went from looking at the automobile to looking at all sorts of things, comparing plastic and paper bags, for example, and people moved to Argonne and began doing it regularly there.
We published quite a number of papers on these analyses—actually the very first one we did, the automobile, I submitted to Science, and they wouldn’t publish it, so it came out in the Bulletin of Atomic Scientists.
How did you get the information to the people who needed it?
Oh, we would talk to people in the industry.
They took your suggestions?
Yes, they were interested. One time somebody said, “Why are you comparing the actual energy and free energy with the ideal thermodynamic limit, which is based on an infinitely slow process? Who would wait for a delivery of a car by a manufacturer who makes his cars reversibly?” And that raised a very interesting question in my mind. In traditional thermodynamics, you use thermodynamic potentials to determine the ultimate best performance that you can get. You use different potentials for different kinds of processes. But there’s a whole class of thermodynamic potentials that are based on the infinitely slow process. And the question that was stimulated by that challenge was, is it possible, or under what conditions was it possible to define and construct and evaluate the analogue of the traditional thermodynamic potentials for a process that’s constrained to operate in a finite time? This led to our asking questions about existence—we were proving existence theorems—and developing methods to construct these finite-time thermodynamic potentials. It has become a whole field of thermodynamics, finite-time thermodynamics.
I have spent many years working on thermodynamics.
Was there something else you thought you were going to work on instead?
I did my thesis on electronic structure calculations. I always was interested in both experiments and theory. One of my favorite experiments was colliding positive and negative ions at low relative velocities and watching them neutralize as the electron jumped from the negative ion to the positive one. You can see it because the electron jumps, for example, from an oxygen negative ion to a sodium positive ion, and it goes to an excited state, so the sodium emits light. You can detect the light with photocells. We had two beams that went into a magnetic field, which bent them 180 degrees so they went in one direction and came out the opposite direction, arranged so they merged, and the relative velocity was very low. I always have loved that experiment!
So it slows them down?
No, they’re going at almost the same speed. It’s how fast they’re moving relative to each other that matters. That [experiment] came at a time when I would go out with my graduate students and postdocs on Fridays for lunch. One of the places we liked to go was a Chinese restaurant on 63rd Street called Tai Sam Yon. So we named the apparatus Twin Accelerating Ion Sources Analyze Magnetically Yielding Optically-Excited Neutrals—TAISAMYON! And we put that in the paper when we published it!
Anyway, I’ve done a lot experimentally and theoretically in things like atomic collisions and atomic photoionization. And for a while I got contaminated by biology—I collaborated with Karl Freed and Tobin Sosnick on some protein work. We looked at things like how certain kinds of things, like exposed hydrogen bonds on the surface of the protein, act as an attractive site for pulling things together.
What comes first for you, theory or experiments?
When I was a graduate student, I was trying to do both, but my thesis was all theoretical because there was a problem [with] the apparatus for the experimental work I was trying to do. I foolishly was trying to make my apparatus out of glass instead of metal, and I was able to have the glassblower at Harvard make the apparatus. He showed it to me when he finished it, and all he had to do was anneal it overnight, but the annealing thermostat didn’t work, and the apparatus melted. I came in the next morning to pick it up, and instead of being round it was sort of flat.
And you gave it up immediately?
Yes, exactly. I got so discouraged I decided I would just do a theoretical thesis. But actually I had a temporary instructorship after I finished my PhD at Harvard, and in that period, I actually did experimental spectroscopy in collaboration with a man named Bill Klemperer, a longtime friend since graduate school—and Stuart Rice! They were looking at the vapor of alkali halides like sodium chloride in infrared, and there was one discrepancy between what they were finding and a much, much earlier paper from the 1920s about UV spectroscopy, so I did the UV experiment and found it really was consistent with their infrared work, so we were collaborating even in the 1950s.
How is it to have a working relationship for such a long time?
We haven’t written a paper together in awhile. But we will!
You didn’t always intend to solve practical problems with your science, did you?
I was never averse to it. But I think it was the environmental challenge when I came to Chicago that really got me interested in that whole aspect of using science. One of the people I worked with was David Currie from the Law School. [Once] somebody in the administration decided there should be a public debate about controlling air pollution. So they asked [law professor] George Anastaplo and me to be on one side, and on the other side, there was Harold Demsetz, Ronald Coase, and Milton Friedman. It was a formal debate—resolved that Commonwealth Edison should be required to put precipitators and air pollution controls on their power plants. We had the affirmative, and they had the negative. We knew how they would argue, but they didn’t know how we would argue. That we knew. But there was another advantage we had that we didn’t realize—they were so overconfident, they didn’t do any homework! They just thought they knew what the answers were. We were able to show that the costs of controls were much less than the cost of the damages, and we won the debate. You can imagine how I felt, winning a debate from Milton Friedman and Ronald Coase! Harold Demsetz was the first speaker, and the first thing he said was, “We do not yet have enough air pollution.” Demsetz was a fantastic teacher in the sense that he was just great at getting students to challenge their unexamined accepted ideas.
After that debate, three of us decided to give a course together on environmental management—David Currie, Dan Janzen,an ecologist who left for Michigan at the end of that year, and myself. We didn’t know where it should go in the curriculum, but we thought maybe Urban Studies. I called up the chair of Urban Studies, and he said, “Hey, that’s a great idea! I’ll give you a course number.” I think we gave the course three times before David Currie went to Springfield to head the state pollution control board, and George Tolley from economics took over from him. George and I gave that course for many years. We got tired of it and we stopped, but he and I have been giving a course for the past six or seven years now on energy and energy policy together. We started with about 35 students, and this past fall we had 135 students. We try to get as many students as we can from the natural sciences because the majority comes from economics. We make them work in interdisciplinary teams, and each team has to write a professional quality term paper. A surprising number of students in the course have decided to go into public policy work. It’s been a lot of fun.
How did you become interested in science?
As a small kid, I was given a microscope, and I really liked that. And one Christmas I remember waking up at about four in the morning and looking at the pile of presents and finding this box. I could smooth the paper and read that it was a chemistry set. I think I must have been about 11. As a teenager, I read a lot of books about magic in a bottle and crucibles,[1] so I knew chemistry was about explosions and the structure of the atom. I loved explosions. I had a lab in the basement. At that time, anybody, including a teenage kid, could go to a chemical supply house and buy anything. So we would buy sodium. You could throw a chunk of sodium in the water in the gutter and watch it go zzzzt . . . BOOM!! I also learned about physics from my mother’s high school physics book, which was just terrible. Physics was about ladders leaning against walls. So physics was boring and chemistry was interesting—I knew that from what I read! And it wasn’t until I got to college that physics was also interesting. I decided then I wanted to work on the border of physics and chemistry.
If you could work on anything right now, what would you be working on?
I would certainly have students working on the problem of size-dependent properties [or] the way that electron-electron interactions determine the structure of an atom. The pictures out in the hall from my office essentially demonstrate that atoms with two valence electrons, like magnesium or barium, are much like molecular states—you can assign vibrational and rotational quantum numbers to the electrons collectively. In many ways viewing the electrons as molecule-like in these two-electron shells is a better approximation than imagining them swimming independently. This led to a question that we never got around to answering: suppose you have four electrons like the carbon atom. Is the structure of the ground state of a carbon atom really a tetrahedral shape? Linus Pauling said that to get a tetrahedron, you have to promote the atom. I’ve come to disbelieve that. I think the electron-electron correlation and the repulsion of the electrons is enough to make the carbon look tetrahedral in its ground state. But that’s only my guess. You’d find out by looking carefully in the right way examining the structure of a well-represented carbon atom. That’s one of the things I’d do.
Are you worried about what will happen to environmental protections under the current administration?
I’m terrified. If this president and his head of the EPA carry through with what they’ve threatened, it could lead to global disaster. It could cause environmental consequences that would lead to many phenomena that would destroy our civilization. If sea levels were to rise a foot-and-a-half, we’d lose coastal cities, and the climate extremes could become so severe that we’d have far more storms than we’re having now—and we just had tornadoes yesterday! Look: no snow in January or February here in Chicago! The blindness of the present administration to this phenomenon and their absolute refusal to recognize the facts we observe—even if you’re skeptical of the models—could be suicidal.
What can we do?
Take a bigger view—the whole human species, the whole of life on earth is a tiny transient in the bigger scheme of things. So why worry? If it’s not an asteroid that makes us extinct, we can do it ourselves!
(Irene C. Hsiao, The Chemists Club, Winter 2017)
[1] Milton M. Silverstein, Magic in a Bottle (New York: Macmillan, 1942); Bernard Jaffe, Crucibles: The Lives and Achievements of the Great Chemists (New York: Simon & Schuster, 1930).