BERNHARDT WUENSCH: Materials have been around for a long, long time. Glass making was already a flourishing industry in Egypt in 5,000 BC. Pottery is usually one of the landmarks in the evolution of civilization. In fact we are used to designating the level of a particular culture by the materials they used. Neolithic, bronze age, iron age, and this will probably be known as the silicon age. It is a witty observation I can lay no claim to, but it is valid. But the difference between what we know as materials science and engineering today is that these older crafts, and such they were, were strictly empiricist. They did not understand what was happening. To use ceramics, my field, as an example: the word ceramics is derived from the Greek keramos, which means burnt stuff. You take a plastic blob of clay and throw it in a fire; and suddenly it becomes hard, maintains its shape and becomes impervious to water. Exactly what goes on in that process, called sintering, was not unravelled until the mid-1950s. All sorts of complicated physical and chemical changes take place. The clay minerals that make up most clays, predominantly kaolinite and montmorillonite, are hydrous silicates. The water of crystallization goes off, organic material that is invariably present in clay gets burnt off and then little bridges begin to form between the crystallites. These grow and coarsen, and gradually, rather than having a few grains that are tenuously stuck together, you have grains that are bonded with big vacancies in between the grains. And then the grains grow and the boundaries between them grow and some of these voids become enclosed in the solid. It is a tremendously complicated process driven by diffusion and vaporization processes that we now understand. But the early refractories industries, the clay industries, were based purely on empiricism. And when a particular mine closed down and the supply of kaolin dried up, the processing industry had to start all over again, and work out new procedures from scratch.
In the area of materials, and I would now include metals, ceramics, polymers, and more recently electronic materials, not just semiconductors but also magnetic and optical materials, the four main classes of materials, underwent a transformation from empiricism to something founded on a firm basic understanding. When these materials made the transition from a low-cost, high-volume, primarily structural material (steel in the case of metals; basic structural ceramics, sewer pipes; bath room fixtures, dishware and so on in the case of ceramics) to something that was very high-tech, a case where the development of a particular device was limited by the materials available, and improving them required knowing precisely what went on. And what justified this effort was that this was a transition to a very high-technology, high value added application. So, in the case of metals one went from steel girders as metals of construction to high-temperature alloys that went into jet engines and things of that sort. And to achieve those properties and to push those limits required a fundamental understanding of what was going on. In the case of ceramics that transition came perhaps 10 or 15 years later. And some of the driving forces for that transition...
ARNE HESSENBRUCH: Which decades are you referring to?
BERNHARDT WUENSCH: For metals the 1940s and 1950s, for ceramics the 1950s and 1960s. The real breakthroughs there were first driven by magnetic oxides in the very late 40s early 50s. That began with the radar developed during the war effort. One had to use materials that were magnetic but also electrically insulating because the high-frequency electromagnetic radiation employed would induce currents in an electrically-conducting material causing detrimental heating.
The other great impetus came with the advent of the space program, and in particular the development of materials for rocket nozzles and for reentry vehicles. I was in graduate school in those days, the early 1960s, and worked summers at a company involved in the nosecone project for the Apollo program. The initial approach to the problem, to get rid of all the heat caused by the vehicle entering and reentering the atmosphere, was to make it out of a gigantic casting of copper copper being a good thermal conductor and the heat would thus be conducted away into other regions. But that was not terribly efficient: these things were huge, 10 feet in diameter: gleaming and shining, marvellous pieces of technology and materials processing and development. The next stage, largely developed by the company that I spent the summers with, a collaboration that lasted for 15 years, used ceramic materials because they have much higher melting points. 2500-3000 degrees centigrade is not uncommon for simple ceramic materials such as aluminum oxide and magnesium oxide. Simple low-cost materials, but very refractory, very high melting points. But they have one liability and that is that they are very brittle. So the technique that was worked out by the Space Systems Division of AVCO Corporation, in Wilmington, Mass., was again to make a massive ceramic forging of something like magnesium or aluminum oxide, and then avoided the brittle problem by using a refractory metal such as molybdenum or tungsten, formed in the shape of a honey comb that would provide the strength and resistance to fracture. So this was a composite material. And then, as goes on in many industries, the final and ultimate solution was to make it out of plastic!
To recap: in the early days of the field we saw this transition from an empirically based technology, variable and not controllable, to an era when the fundamental physical properties were understood on the basis of rational chemistry and physics and applied sometimes in extremely complicated contexts to a case where one could design and improve materials and push limits. At that time, the field was aligned very much along specific classes of materials. There was a metals industry, there was a Journal of Metals, there was an American Society for Metals. There were industries: US Steel and so on. They were based on metals, and even particular sorts of metals. The same was true of ceramics. There were very large refractory companies that made abrasives and grinding wheels. Norton in Worcester, Massachusetts, was one example. Their whole product line was crystalline ceramics. Corning Glass, as the name indicates, was working with amorphous oxides.
It happened a little bit later with polymers. The polymer program here at MIT sprung up in very different ways. Curiously, a large part of it was initiated in Mechanical Engineering, simply because the development of the new fibers, nylon and rayon, required a completely different set of machinery to spin and weave these fibers. They had very different physical properties from natural fibers like silk and wool. So again the need for a transition to a new, higher class, engineered sort of material required mechanical engineering people to learn something more about polymers in order to process them. But again, back in that era, apart from synthetic fibers, when you thought of plastics you thought of some inexpensive kids' toy that would break in 36 hours and you did not really care because it cost next to nothing. Or you thought of polystyrene foam coffee cup, but you did not think of high-technology conductive or optical materials. And that has now become the case.
And then finally with the transistor and electronic devices, one of the great revolutions in our culture. The additional class of materials, silicon, electro-optical materials, came into being.
Finally it dawned upon people that even though these were materials with different properties and problems, having to be processed in different ways, everyone was really doing the same thing. They were looking at the connection between structure on all of its scales, from electronic and atomic arrangements to the arrangements of crystallites in a solid to more massive structures such as iron reinforcements in concrete. Everyone was looking at the connection between structure on all its scales and the way a material behaved. Increasingly, engineers had to learn to use these different classes of materials in concert. My favorite example is an electronic chip. It has a ceramic substrate, usually aluminum oxide, on which electrically active semiconductor elements are deposited and then etched and shaped and changed in composition. You need to have that thing talk with the outside world, so you need metal contacts in order to connect the electronic components to some circuitry. And then to protect this very small-scale structure, you encapsulate the whole works in a polymer. So, for someone to design such devices, he or she had to be familiar with all four major classes of materials, and the interfaces between them and how they behaved. The devices would not be possible without that sort of understanding. Or on the more science-oriented activities: someone who is operating a sophisticated instrument like a Secondary-Ion Mass Spectrometer to measure very shallow composition gradients on the scale of Ångstroms, or someone who is using a sophisticated electron microscope of ultrahigh resolution that allows you to actually see the individual atoms (actually more rigorously: projected columns of atoms in thin film form). They could call themselves a metallurgist, a ceramist, or an electronic materials person, merely depending on what sort of material they put in their instrument. Techniques were the same sure: the information they were trying to extract and the uses to which it was to be put differed, but still they would all do the same thing independently of the particular class of materials. So that was really the birth of a materials science. It was when this concern with structure and properties began to transcend different classes of materials and the technology required that materials were used in concert. Now there are companies, particular the semiconductor companies that are involved with all these materials, and people have to understand the basic science of the classes of materials and the engineering.
ARNE HESSENBRUCH: What about the demand for interdisciplinarity? I have heard accounts that materials science as a uniter of all these different strands was primarily a political decision. It is very difficult to unite across departments of a university, and so interdisciplinary laboratories were created from above, in which this interdisciplinary work could take place. In other words, in addition to the development of a logic of the field itself, there may also have been an outside demand pulling the different strands together.
BERNHARDT WUENSCH: Probably one of the great difficulties in any university setting, particularly a busy one like MIT, is that people are involved in a certain set of problems that they are very knowledgeable in and very excited about, and they can find more than enough for themselves to do. This means that you are perfectly happy working in a particular area and perhaps even achieving prominence in it, but you are really too busy to reach out and think about collaboration. This was particularly true when there were walls between academic disciplines and departments that were far more inpenetrable than they are now. There are several observations that might be made. First of all, I think the evolution of materials science and materials engineering as a name in the academic enterprise is something that happened naturally. It was not created by decree. Many of the people who worked in metals had backgrounds in chemistry, and that is also true of many of the people who were transforming ceramics science, developing a ceramics science from empiricism. They primarily had chemistry backgrounds and more rarely a background in physics. But the people doing condensed matter physics were really anxious to be at a frontier and not do practically-oriented engineering investigations. Again the old generalizations don't stand up, but for most people this was the case.
ARNE HESSENBRUCH: When was this?
BERNHARDT WUENSCH: In the case of Kingery this was in the late 1950s, and the developments in steel were about a decade earlier. The Department of Metallurgy gradually became more diverse because the junior people brought in had wider interests than the narrowly metallurgist interest of the people already there. When I entered the Department in 1964, there were two people who had just joined. One was Harry Gatos, one of the founders of the Materials Research Society, one of the first major materials-generic societies. He was brought in from the Lincoln Lab to work on the growth of large single crystals of silicon semiconductor materials which had the purity and the mechanical perfection to have the desired electrical properties. There was a young Assistant Professor whose name was Gus Witt who came to work with an electrochemist. Casting around is probably too strong a term, because he was not looking for something to do. But the person he had come to work with, Phil de Bruyn, had left MIT. He was South African and had a Dutch wife and wanted to return to that cultural setting. So Gus was left without a home. In those days a young faculty member did not usually get a 6 or 7 figure start-up grant of expenses for capital, but rather had to align him- or herself with a senior faculty member who, because of their standing and reputation, were well-endowed with funds. So Gus Witt collaborated with Harry Gatos on the semiconductor program. That was the nucleus of the semiconducting and electronic materials part of the department.
ARNE HESSENBRUCH: And that this was a generational change?
BERNHARDT WUENSCH: Partly a generational change, but partly also driven by this transition from empiricism to a rational scientific basis. One of the things that MIT faculty in that period were very, very good at was this process of breathing new rationality and sound scientific fundamentals into many engineering disciplines that were really not far beyond empiricism. The science, chemistry and physics that they introduced were not really terribly sophisticated by today's standards. They had the knack of seeing the opportunity to bring this rationality into a field which otherwise lacked it. I mentioned the materials people, but there were people doing the same thing in other fields, in electrical and mechanical engineering and so on.
ARNE HESSENBRUCH: And the theoretical tools that they were beginning to share were things like thermodynamics?
BERNHARDT WUENSCH: That is right.
ARNE HESSENBRUCH: Any other theoretical tools that were beginning to be shared in the same period?
The systems that they worked with were relatively simple, chemically and structurally. The geologists however, were concerned with anything that Mother Nature had provided, so they delved into these enormously complicated silicates and other sorts of minerals, impure, all sorts of solid solution and chemical substitutions, incredibly complex structurally. But it was a static record. It was relatively late that they began to actively investigate phase transformations, something dynamic. The metallurgists and ceramists had been working on solid-state diffusion, oxidation, and corrosion processes for 20-25 years before the geologists became interested in measuring diffusion rates in silicates. They did so in order to understand the transformations that took place in rock implacements, particularly under pressure and hydrothermal conditions. And conversely, the aim was also to interpret the thermal history of the rock with knowledge of the kinetic rates. So there were two complementary views: one of complex materials and a static picture along with the materials science people who were working on similar issues but in a more dynamic sense with simple materials.
ARNE HESSENBRUCH: It can almost be mapped by today's standard textbook. There will be chapters on the four major classes of materials: metals, ceramics, and so on; and then there will be chapters on the properties (thermodynamics, electrical, optical, magnetic). And each of these chapters corresponds to the theoretical tools that were beginning to be shared by the communities you mentioned in the 1950s and 1960s. Would that be a reasonable way of describing it?
BERNHARDT WUENSCH: Yes. And another chapter that every textbook on condensed matter has to start out with is structure. This is atomic-scale structure in crystals and amorphous materials, but also microstructure: dislocations, imperfections, and vacancies. These control a great many properties to a larger extent even than the intrinsic properties of the material. Also interfaces. Unfortunately if you are writing a comprehensive textbook you cannot spend too much time on this sort of thing. My field of interest and passion crystallography is one that is often slighted. I took a lot of ribbing from my colleagues when I joined the Department of Metallurgy: You are going to have it easy. There are only three kinds of crystal systems: body centered cubic, face centered cubic, and complex! There was some truth as a basis for that kind of sardonic view. Even today the great majority of materials used in ceramics are cubic. Not all, but the majority are structurally simple.
But then you asked about, since this was becoming valuable in technology, valuable to government enterprises based on those technologies, once the value of this interdisciplinary work was perceived, were there ways to encourage people to work in this interdisciplinary mode? There were indeed. In the late 1950s, there was an effort to do this that arose in the Department of Defense, from an agency then called DARPA (Defense Advanced Research Projects Agency). This was shortly on the heels of the discovery of the transistor and the great potential for this device in electronics and communication, and therefore in defense. The people in DARPA had the wisdom to realize that, in order to capitalize on the electronic properties of semiconductors, one would have to understand the chemistry of producing these materials and the way of growing single crystals of unprecedented perfection. This brought in the chemists, it brought in the crystal growers and the materials scientists. You also needed to understand the physics of these very complicated electronic structures and the way of controlling them through doping and through creation of proper interfaces. The only way to make any progress on this was to bring together in one place people who were interested primarily in solids but who came to this area bringing the perspective, instrumentation, and the skill and expertise of their particular discipline with a view to collaboration in a very close fashion that is, not only formulating but also conducting the research. It was probably the late 1950s, before I came on the scene as a faculty member, that DARPA went around to probably a dozen or more research universities saying: Hey, do we have a deal for you! We would like to come and make a building for you in which you might care to house the majority of your people working on problems in condensed matter. Initially, the emphasis was on semiconductors, because that was the driving force for this investment. I believe twelve sites or so were selected, perhaps initially ten with two created later. But that was the origin of our Building 13, directly behind the great dome. That building was constructed during the early 1960s and we moved into it in 1965. It housed a great part of the Department of Metallurgy and Materials Science, as the name had evolved by then. Each name change was won at great expense by the Department Chair, since obviously people who had been trained as metallurgists resisted. That united them in a way that nothing else in academia could have done. You can imagine the furor arising from our grand alumni when the word metallurgy disappeared altogether.
In any case, the new Building 13, the Center for Materials Science and Engineering housed, on its 4th and 5th floors, a major portion of the Department of Materials Science and Engineering: the entire ceramics and electronic materials groups, the polymers group (which grew and increasingly took residence on the 5th floor). The electrical engineers, those who were interested in the materials aspects of devices, were on the 3rd floor. On the 2nd floor were the administrative headquarters and the condensed matter physics group. The latter has grown considerably since that time. Parenthetically: this new creation was the second major interdisciplinary laboratory at MIT the first was an outgrowth of the Radiation Lab that developed radar in the famous wooden Building 20, since deceased, and Building 24. It is now called the Research Laboratory of Electronics. At the end of the war many of those people who had been brought together for purposes of the defense effort dispersed and went back to their home institutions. But some stayed on and that was the birth of the Research Lab for Electronics. This was another benchmark in the evolution of interdisciplinary work. In this case it was primarily Electrical Engineering and attendant design - some materials work. That came into being shortly after the war, early 1950s.
The charter for the Center for Materials Science and Engineering (and the organization had slightly different names at different institutions) was that first you will conduct materials research in a highly interactive and interdisciplinary mode and faculty from as many as seven different departments at MIT had participated in the programs. These workers were to be grouped together in teams called Areas of Thrust in early semiconductor research, but later broadened considerably. The other thing that was visionary was the realization that in order to conduct first-class research one needed sophisticated, first-class equipment constantly evolving and growing more sophisticated. This was too expensive to obtain for any single faculty member or small group of faculty. And moreover, for it to be effective and working at capability, it required support staff. So, the second mandate besides creating these funded, collaborative, and interdisciplinary research teams was to establish central research facilities. For a time, the Center had as many as 12 different central facilities. The next thing the sponsors did was to seek to convert young faculty to this interdisciplinary mode of operation. In order to do so they provided seed funding. It was an unabashed attempt to seduce young faculty into these interdisciplinary teams!
ARNE HESSENBRUCH: Seed funding refers to a couple of years' funding which would then be taken over by the hosting university?
BERNHARDT WUENSCH: No. The Center operated under a block-funded federal budget. Seed funding was intended to provide an opportunity for a young faculty member, someone who might otherwise not be attracted to participate in the interdisciplinary teams. To develop this connection they wanted to get to them early and make it worth their while to not identify only with their peers in their own department. The hope was to fold such young faculty into one of the Areas of Thrust. There were also some features of this program that were almost without precedent: a great deal of local autonomy was given to the Director of the Center. The Director decided what problems to address on the basis of the resources unique to the particular host university. Rather than having these Areas of Thrust (generally speaking there were three or four) monitored by some program director in Washington, they were monitored locally. And this Director presumably would be gutsy enough to pare off the team members who were not interacting sufficiently or whose expertise was no longer needed for a new direction to be taken. It was a thoroughly enlightened view that somebody on the scene would be better equipped than someone down in Washington. This is not true, even in individual research grants, today. The NSF gives grants on the basis of promise, and once you have received the grant you can change minor details, but you cannot start off doing ceramics and then suddenly decide to do fruit fly genetics at least not without major questioning from the NSF division. Other agencies, the Atomic Energy Commission, and then the Department of Energy, wrote not a grant but a contract: and you had to be very specific about your research. You had to prepare a work statement from which you could not depart without permission of the program director.
This made the Materials Research Laboratories very lean and fast on their feet. If something came up, they could change direction instantly. One of the great success stories here was in high critical-temperature oxide superconductors. When Bednorz and Muller made their earth shaking discovery [of superconductivity in ceramic materials], any number of people dropped what they were doing and began bootlegging efforts to identify these phases and understand this peculiar behavior which is still not well understood. The Director of our Center for Materials Science and Engineering at the time brought together everyone at MIT who was interested in these materials, and obviously that number was rather substantial, and selected individuals to develop an area of thrust. And with one or two months' notice another Area of Thrust was cancelled: one concerned with defects in semiconductors. Now that may seem like a very courageous thing, but in fact many of the people working on these defects were also interested or even more interested in superconducting materials. People were working in everything from growing single crystals of the materials for basic property measurements. People in condensed matter physics ended up using these crystals some very, very critical measurements in magnetic ordering and magnetic correlation lengths. That was the first clue that something in the magnetic moments of copper ions was critical to this behavior. Other materials scientists were interested in learning how to make polycrystalline specimens, so that the superconducting properties could be used in making better magnets. Since these were not metals, unlike all previous superconductors, it was not at all clear how you would wind them into solenoids to make high-strength magnets. That particular part of the effort at our Center led to the first US patent granted for processing of an oxide superconductor, and in fact it eventually led to a very successful spin-off company.
The only other thing that needs to be said about the program is that it evolved over time. The typical tour of duty of the Director was five years because by that time the necessary decisions had offended enough people for a new Director to be brought in. The change came towards the end of the Eisenhower administration, when Eisenhower coined the famous phrase: the military-industrial complex. There was a growing concern that the military had control of and sponsored much work, such as basic research in the space program, and that this belonged more properly in the civilian sector. So, at that time, the leader of the Senate, Mike Mansfield, tagged onto the Defense Appropriation Bill an amendment named after him, that confined the military to weapons systems and maintenance and support of armed forces. But they should not be in a position of supporting basic science. That is also interesting historically because one of the prime sources of funding for basic research prior to that had been the Office of Naval Research. They supported a lot of very fundamental science that had only indirect long-range connections with naval defense systems. So the amendment brought about momentous changes: the National Science Foundation's budget was beefed up; it grew exponentially. I think NASA was set up at approximately this time. It absorbed space programs that had been conducted under the auspices of the Air Force and the Army I think it is fair to say with considerable overlap and duplication. I think that also the Atomic Energy Commission and the Department of Energy came about then too. The military aspects of atomic energy stayed in the hands of the military, but the peaceful uses were spun off to the civilian agency.
What happened to the Materials Research Laboratories? The NSF was very impressed with the success of this approach. The whole justification of it is that the net effect should be far and away above what could be achieved single investigator grants. And these investigators might or might not collaborate. There was arguable evidence that this was the case. The NSF toured these Materials Research Laboratories and announced that they would take over the sponsorship of the program.
Things stayed in this form until about seven years ago (1994), when there was some concern that, successful as the program was, it appeared to have become an old boys' club, in that there was no sunset clause. True, the programs were reviewed every three years with a major site visit lasting several days; proposals were critically reviewed by outside referees, as is the case with any research proposal. Nevertheless, there were some materials activities at universities that were better than others, perhaps because of a critical number of faculty involved and also, one might say, because of the quality of these universities. I am aware that this may sound smug and immodest, but the situation was ripe for sour grapes, and there was a lot of grousing. Politically, a lot of states realized that a way to give the state economy a boost was to start up a research university that attracted industries and created jobs leading to tax revenues and so on. So there were sour grapes. And the MRL program had indeed gone on for a long time. It was competing with a similar program that had been created during the Reagan era. This was created as a response to Reagan's address with words to the effect that we needed to get this country going again, to be competitive and to rev up the economy. The NSF went to Reagan with a proposal to create something called Science and Technology Centers. Very similar in concept to the Materials Research Centers, with two exceptions: 1) they had a sunset clause in them (seven years); and 2) they were to focus on a specific problem, and could not shift and change, grow and evolve. They had a much greater emphasis on the outreach to technology and to the professional community: newsletters, training courses, bridges to industry to implement the technologies were integral to the Science and Technology Centers. So we had MRLs (successful, but oh, what have you done for us lately), and the S&TCs. And about seven years ago, just after a director of the NSF had resigned and the directorship of the Materials Research program had changed hands, Congress decided that the entire program should be looked at anew. As the result of a study that took place over a very, very short period of time, the system was changed into something called the MRSECs (Materials Research Science and Engineering Centers). I love the name because as an acronym it is gender blind: they can be either Mrs. Ecs or Mr. Secs! A masterful political ploy. They supplanted both the Materials Research Laboratories and also the Science and Technology Centers. The funding level remained pretty much at a comparable level. Now there is a notion that, for the older programs, the bar should be raised consecutively at the time of renewals and there should be a chance for creation of some new MRSECs. This is taking place: some of the older MRLs have lost their support and new programs have come into existence, as MRSECs.
So much for institutional support! But that was a conscious effort to advance materials research in general but also to deliberately foster this interdisciplinary, collaborative mode of work on materials.
ARNE HESSENBRUCH: You have talked about the theoretical developments, from the institutional perspective, and the funding. Is there a story to be told about the history of the instrumentation?
BERNHARDT WUENSCH: When one discusses the history of instrumentation one immediately gets very specific. There are tools such as electron optics that have had a wide application from medical science to inorganic materials research: physics, chemistry, and geology and so on. It was a tool with unprecedented resolution. But it gets very specific. It is true of the way I have operated on my own research path. I always kept an ear to the ground to new apparatus that might provide something new in terms of precision, resolution, or more flexibility, or permit doing something that could not previously be done. Often, these new instruments came into being in fields that were distinct from your own area. Apparatus has become increasingly complex, and companies have made new designs and developed them to a high level. The advent of powerful computers have changed research in a major way. The speed and means of collecting data, the way in which it is analysed, the sort of analysis of a problem that can be undertaken have all changed in ways that could not have been anticipated a few years earlier.
This interview has been transcribed by Arne Hessenbruch and proofread by Professor Wuensch.
This page was last updated on 16 June 2001 by Arne Hessenbruch.