Well, as a student I was interested in, and so tried to study, the history of just about everything, from physiology to philosophy and from meteorology to math to molecules. Speaking of math and molecules, mathematics and chemistry became my major fields, the first because it seemed the most challenging subject and the second because the math needed to be applied to something, and with the oil crisis ongoing I could get interesting work running programs to solve the differential equations modeling coal-liquefaction processes.

Then, for my graduate work (Cornell), I chose Physical Chemistry because it seemed best to combine all those interests, but it meant I had to remedy my deficiencies in physics, which I didn't particularly enjoy in comparison to the magic world of chemistry. And instead of doing mainly theoretical or modeling work, I was attracted to the new laser laboratories, again to work on energy-related projects (photocatalysis), but soon enough I was working mainly on pure molecular spectroscopy (and molecular quantum mechanics) using the new nonlinear optical methods and supercooled gas jets and mass spectrometers. That was a wonderful experience, and I owe a lot to my main thesis advisor (Prof. Ed Grant, now Chemistry Chair at the University of British Columbia) for encouraging me to explore so many areas (My Ph.D. dissertation, "Non-Adiabatic Bound States of Isolated Molecules," won the ACS prize for that year, 1984).

“Over time, many groups accumulated enough experience with diverse nanoparticle materials to realize that the claims about the 'nanocrystal gold molecules' really hold up, which regrettably hasn't been the case for so many other such new materials being promoted out there.”

Looking beyond my immediate thesis research, I was really excited by the prospect that these new molecular-beam and spectroscopic methods could be applied to ever larger, more complex systems, without any discernible upper limit and with incredible sensitivity. In this way, one might ultimately obtain an accurate, molecular-level picture of the inner workings of phenomena that are normally obscured in complex media or are (short-lived) transients in rapidly occurring processes. And this was truly a wide open, new area, at that time, ripe for the picking!

One path was to experiment on weakly bound systems, and this would lead to studies of solvation phenomena and the properties of liquid solutions, and ultimately to molecular biology—proteins and their complexes and such. A second path would focus on strongly bound materials, inorganic clusters as "solids in embryonic forms," and models for surface-chemical processes (such as corrosion and catalysis and sorption). The idea of gaining an understanding of the solid-state and surface phenomena, by exploring well-defined small fragments of solids, was so timely. So, initially pursuing the latter, I went to work first at Exxon's Corporate Research Labs, where they (along with Richard Smalley's lab at Rice) had assembled these special capabilities and staff, to explore metal and carbon clusters under the theme of "Clusters = Molecular-Scale Surfaces." And this was all set to become a tremendously active and exciting area. Just to remind you of one well-known outcome, Kroto et al. [1985] discovered (using the same methods) the special stability of the C60 cluster and so on, and proposed their famous structural models (confirmed as stable fullerene molecules just five years thereafter).

And so when I left Exxon, to join the UCLA faculty, we built up a lab there to pursue this new kind of work, both on molecular clusters as well as the strongly bound variety, i.e., carbon clusters and metal clusters in cluster beams. There we had the good fortune to contribute to many significant developments in this field, including fundamental aspects of cluster structure, thermodynamics (phase transitions), and dynamical properties, as well as outgrowths into related fields—for example, the first isolation of the higher fullerenes and metallofullerenes, and the superconducting fulleride salts. These experiences were crucial to our decision to get involved in the special area of metal nanocrystals.

Well, nanocrystals may be regarded as a special subset of larger clusters, and so this was an extension or outgrowth of our involvement with clusters to larger dimensions and focusing on better-ordered structures. Returning to the carbon analogy: just as the C60 buckyball is related to the familiar graphite (pencil-lead) form of macroscopic carbon, certain special clusters of other elements and compounds were frequently found to be related to the crystalline-ordered phases of the corresponding bulk solids, all from a surprisingly small size of just a few dozen atoms. For example, our UCLA lab worked on clusters of sodium-chloride (aka sea salt or table salt)—which are models for some important marine aerosol processes—and these turn out frequently to assume structures that are essentially just little fragments or crystallites of the much larger familiar crystals, even though half (or more) of the atomic ions are in the surface layer. (Our first paper using the term in this way was "Alkali-Halide Nanocrystals," in Accounts of Chemical Research 26[2]: 49-56, February 1993.)

Metal clusters and nanocrystals was another very active area. And from this activity came some general results, by the ‘90s, in the form of rules for the special stability of certain cluster sizes. These were based either on the close-packing of metal atoms and/or the "electron count" (of the valence or conduction electrons), the latter giving rise to the idea of "superatoms" (see also "quantum dots"). So it was natural to wonder whether, or how, certain of these ultra-stable metal nanocrystallites could be generated (outside the vacuum chambers) in molecular forms, much as the fullerenes are to carbon clusters generally. So, as I said before, the recent experience with the metals-in-fullerenes (metallofullerene compounds) fresh in mind, we were looking for ways to encapsulate and protect the larger metal clusters.

This is where another development comes in: from surface chemistry, the discovery and rapidly accumulating knowledge of self-assembled monolayers (SAMs), with their exquisite control of crystal-surface properties, suddenly made it seem possible to control the surfaces of diverse nanometer-scale crystallites using the same chemistry. Put another way, this seemed to offer a promising new way, well beyond the established colloidal chemistry or synthetic chemistry routes, to manipulate very large clusters or ultra-fine crystallites, and turn their properties into a precision molecular science. So, when we moved the lab from UCLA to GIT, we set out to develop special methods, or instruments, for generating and collecting and isolating nanocrystals protected by self-assembled monolayers, in order that they could be thoroughly investigated, classified, and turned over to chemists and materials scientists generally for all kinds of other purposes.

The term "nanocrystal" is a compound abbreviation for "nanometer-scale crystallite," and thus refers both to a length-scale (one nanometer = a billionth of a meter) of a material structure and to the ordered arrangement of the atoms within the structure, i.e., that it is related to one of the crystalline bulk phases. The existence of nanocrystals of various solids has been known for a long time, e.g. electron-microscope images date to the 1950s, but their poor "handling properties," i.e., their very active surfaces, is one big reason why they hadn't become as well known (until recently) as biomolecules or polymers of about the same dimension. Besides being objects of fundamental interest—as a bridge between the world of molecules and that of crystalline solids—which lie clearly within the interdisciplinary world of "nanoscience," one also has, in a more speculative way, the roles imagined for nanocrystals within a new "nanotechnology." Certainly, various metallic, magnetic and semiconductor nanocrystals figure prominently in schemes for a futuristic nano-electronics (computer circuits, sensors, bio-nano elements, and the like).

Metal clusters and nanocrystals seemed to be those most urgently in need of being rendered into robust molecular forms. (Certainly one hoped to gain a level of control and understanding approaching what was happening in the simpler compound-semiconductor area, which had developed rapidly and mostly independent of the cluster-science field.) Gold is the archetypal metallic element, and had been the most commonly investigated of colloidal metals (since 1856!); it has long played a role in critical microelectronics and relays. It has the important advantage that it is noble, i.e., it does not form stable compounds with most of the substances that it is exposed to, although it does form a range of alloys with other metals and can form a protective plated coating on an even wider range of important (nano-)materials. It has also been thought to have a "simple nature" (bonding and electronic structure), and could thus be used to illuminate many aspects of nano-metals generally. So it seems natural to begin with gold and silver, or at least to use them as reference materials for alloys and bimetallic (coated or plated) metals generally.

This report makes a number of unprecedented claims about the nature of very large gold clusters (100-1000 atoms) when SAMs of sulfur-containing molecules (thiolates) protect their surfaces. Most important (to us) was that certain sizes (or narrow size ranges) were much more abundant and stable than others. In the established colloidal-metal chemistry, there had been no such distinction. Of course, the immediate question was why, i.e., what is the selection principle operating here? If one could determine this, then perhaps the nature of the assemblies generally could be described quite simply. And so, in collaboration with theorists, we made an educated guess about the stable structures, and this generated some nice illustrations of the models, which remain popular today.

Another claim was that these especially stable forms could be accumulated in high yield, and then each one separated from the others, so as to obtain a series of high-purity new substances. In essence, each of these forms might be a new kind of very large molecule, or molecular substance, composed of a specific gold-nanocrystal core (of ordered structure and dimension in the 1.0-3.0 nm diameter range) and a densely organized protective organosulfur outer shell, or SAM. In any case, we showed that they could be handled like pure molecular substances that form ordinary solutions in suitable solvents, or in dry form as molecular crystals—crystals of nanocrystals—and could be analyzed intact by typical molecular-science methods, like mass-spectrometry.

We chose the catchy title, "Nanocrystal Gold Molecules," in an attempt to capture the unprecedented nature of these findings; the adjectives "nanocrystal" and "molecular" being otherwise incompatible descriptions of the same objects.

This was our first real publication on the SAM-protected clusters, after several exciting years of steadily accumulating experience with them, of trying to make sure we had the appropriate methods and interpretations. It finally became urgent to publish a report, even before we had anything approaching a full understanding, mainly because other groups had become active in the area, particularly since mid-1994 when the Liverpool University group (Brust, Schiffrin) reported a simple solution-phase method for generating them. So it would have been irresponsible not to share our particular experiences and perspectives on them.

In fact, we had a great deal of trouble writing it and then getting it published in anything close to the form we wanted to use to describe our findings. It was rejected several times, even after we accommodated the reviewers' or editor's requests to make it look more like a conventional "nanoparticles" paper, e.g. by adding electron-microscope images. It was an profoundly frustrating experience, feeling compelled to publish an unconventional paper in a conventional way, and it gave us a whole new perspective on what it means to discover and report on something so astounding that the experts (editors, reviewers, et al.) do not believe it is possible at all. Ideally, those results and insights—including all their limitations—really should have been written out, very carefully and patiently, in a much longer initial paper, and this would have been less flashy, or provocative.

Well, it has clearly turned into a large and active area, and much of that activity is subsumed into the even larger area of nanoscience and nanotechnology projects so diverse that one could hardly hope to follow it all.

In our lab, and several others, we worked directly to characterize the various properties of the special sizes, working from the larger ones downward to find the smallest of the class. And this meant several years of very busy work, with quite a few publications, each one still preliminary but significant steps towards mapping out the whole class. What would happen is that each line of investigation led to rapid progress up to the stopping point, where the analytical methods were not yet suited to handling such materials. (By 2000 we had seemingly run out of immediate contributions to make, and since then we've worked mainly just to correct errors or omissions in that earlier work.) Nowadays, with better-suited methods, the whole work is done much more efficiently in other labs.

Initially, I believe these early ('96 and immediately following) papers failed to have much impact at all. Other researchers would use one small portion or result of the papers, or cite them for some almost incidental aspect or image contained in the papers, while altogether ignoring the main points. But over time this has gradually changed. One reason for this is that dramatic visual evidence exists, in the form of photographs showing sharply defined, colored bands in the (electrophoresis) gels, to convince anyone immediately that the special sizes do exist, at molecular levels of definition and purity. To chemists, this means they can be exploited with confidence, and also that their selective formation indeed requires explanation.

Another reason is a little harder to explain, but has to do with the "robustness" of the properly prepared SAM-protected metal nanocrystals, compared with other kinds of small-nanoparticle materials that had become available. Over time, many groups accumulated enough experience with diverse nanoparticle materials to realize that the claims about the "nanocrystal gold molecules" really hold up, which regrettably hasn't been the case for so many other such new materials being promoted out there. Of course, we didn't know this, and instead assumed that they would be surpassed very soon, a comprehensive review article would be written, and that no one would need to consult these early or preliminary reports. Instead, with each passing year, it becomes clearer that these are likely to stand out among the best of the new nanomaterials, and that the early papers came surprisingly close to giving a full description of their preparations and remarkable properties.

Also, on the fundamental chemistry and physics side, it has been realized that the SAMs and the gold-cores really aren't so simple as they were imagined (in '95), i.e., they had been over-idealized. So they began to attract ever more attention from those who tried new experiments and theoretical models to explain the existence and properties of the special cluster assemblies. The discovery of new catalysts based on supported gold nanocrystals or clusters has brought to light that the chemistry of zero-valent (metallic) gold is not well understood, and some of this interest has spilled over into gold nanostructures "passivated" by the thiolate SAMs. The latest models include features we never could have imagined, such as gold cores that go (from smallest sizes) from planar to deltrahedral cages, to bilayers, some having chiral or helical character; and, in the adsorbed monolayer, a highly cooperative or polymeric bonding among the metal-thiolate groups that encapsulate the metallic core.

Gold nanocrystals, generally, are already at the heart of commercial catalysts developed in Japan for air-purification. Commercial applications for the thiolate-protected metal nanocrystals have developed more slowly, despite the hype surrounding their potential use in nano-electronic or -sensing devices. (The Nanoprobes Inc. company has long sold bio-functionalized gold clusters, based on phosphorous-group and halide protection.) I would look first for a different line of development, i.e., that it must become possible to order them, in purified and functionalized forms, from a chemical-company catalog, just as one would purchase any specialty chemical substance. And then every physicist or chemist or molecular biologist or materials scientist/engineer can use these as starting materials to explore applications within their own domain of science or technology. At that point one will see whether their special properties make them capable of displacing existing products. Until that day comes, I doubt we'll know much about their commercial potential.

Robert. L. Whetten, Ph.D.
Georgia Institute of Technology
Atlanta, GA, USA