from R&D Innovator Volume 2, Number 2
Discovery of Met-Cars
Castleman is Evan Pugh Professor of Chemistry at The Pennsylvania
State University, University Park, PA. He is a fellow of the
American Association for the Advancement of Science and The
American Physical Society, and is senior editor of The
Journal of Physical Chemistry.
just finished some work on the reactions and structure of titanium
species, and we were getting set to look at another molecular
cluster—radon decay products.
But just for curiosity's sake we decided to take a
chance—and wound up finding an entirely new class of compounds.
group focuses on the fascinating phenomenon called clusters, which
are sometimes called the fifth state of matter because their
properties fall between those of gases and those of liquids and
physical and chemical makeup of clusters helps determine their
basic properties. We
have investigated the production of precisely-sized
titanium dioxide clusters for various shades of white pigments.
We've also shown how sulfur dioxide combines with an OH
radical to eventually form sulfuric acid clusters, a key component
of acid rain.
best-know cluster is probably the 60-carbon species called buckminsterfullerene,
whose shape resembles that of geodesic domes.
These so-called buckyballs made headlines a few years ago
and have been touted for many applications.
Buckyballs are clusters comprised of atoms; but clusters
can also be made of molecules.
Our group, for example, has examined clusters of water
vapor molecules in an attempt to find out how they condense into
we look at a new cluster, we search for the most stable
organization of a particular combination of atoms.
We create the cluster by shooting a combination of gases
through a nozzle into a vacuum, then ionizing the material with a
laser beam and gauging the mass of the products with a mass
return to my story: We'd
finished using the laser to probe various titanium reactions,
including the formation of titanium dioxide crystals, and we were
scheduled to begin looking at certain lead ions, to gain an
understanding of how radioactive radon daughter products might be
transported into human lungs, where they could contribute to
See What Happens
spending a few hours to refocus the laser, two of my graduate
students, Baochuan Guo and Kevin Kerns, decided to aim it at a
titanium rod over which ethylene gas was flowing, just
to see what would happen.
We had been investigating the influence of ground state Ti+
in polymerization reactions and wanted to ascertain the effects
due to ions in the excited state.
it was more or less just serendipity that we produced sub-picogram
quantities of a new material.
When we introduced it into a mass spectrometer, it showed a
very strong peak at 528 mass units.
Most clusters produced in this way do not display a single
peak, so we thought we'd done something wrong and immediately
tried the same reaction with a number of other hydrocarbon gases.
All showed the same 528-mass peak.
on earlier work, in which we'd reacted small particles of titanium
with hydrocarbon pollutants, we first thought we might have
produced a polymer of CH2 units. We
immediately discounted this idea since a polymer of single size is
unlikely, but we tested for it anyway.
this might not have been very significant, at least it was simple
enough to test. But
when we substituted a hydrocarbon composed of deuterium (the
one-neutron isotope of hydrogen) as our raw material, we saw no
change in the mass of the product.
Ergo, hydrogen was absent from the mystery material, which
was definitely not a CH2
what was it? I
started thinking it might be a small, 44-carbon buckyball with
mass 528. This seemed
unlikely, however, since the especially prominent peak correlating
to this species had never been reported.
The other alternative—contamination—was less palatable,
and I asked my assistants to clean our equipment.
I couldn't quite get this stuff out of my mind.
As a scientist, I'm always puzzled when I see something I
don't understand—so we tried another reaction, substituting a
different transition metal for titanium.
When Guo and I compared transparencies of the spectrum of
the vanadium-ethylene material with those of the titanium-ethylene
batch, we saw that the peaks had shifted.
This new species was therefore heavier, and that meant
there were titanium atoms in the first reaction product, and
vanadium atoms in the second.
here was something that merited spending some money.
Although I've stayed away from the buckyball scene, I knew
that researchers were seeking to implant metals in them, and now
we had a stable species that apparently contained transition
metals in their very essence.
problem: We still
didn't know what it was. So
we bought some carbon-13 to use in the hydrocarbon raw
material—and made a product that was 12 mass units heavier.
Using only grade-school arithmetic, it was clear that our
material had 12 carbons and 8 metal atoms.
elucidate the structure, we looked at how the new material reacted
with various chemicals. We
tried ammonia, which bonds well to transition metals; the reaction
product had exactly 8 ammonias.
Since we knew that ammonia could only link up to titanium
on the outside of the cluster, we concluded that the
structure must have titanium atoms on the surface.
we learned that the atoms were in equivalent positions (meaning
that every titanium atom was bonded to three carbons, and every
carbon was bonded to two titaniums and one carbon).
That severely constrained the possible structures.
final piece of the puzzle fell into place while I was walking home
from work in January, 1992. I'm
sure any students who were watching me wondered what this
professor was furiously scribbling in the dark—but I was
oblivious to them. My
mind was concentrating on structure, because I had realized the
material was built like a water cluster we'd previously studied. That cluster had 20 water molecules on the outside, and it
was built, like the new species, of pentagonal dodecahedrons.
our mystery material must be composed of balls with 12 pentagonal
faces, each face containing three carbons and two titaniums.
I called Guo (who's now a post-doc with me), and he took
out a modeling set and confirmed the structure was possible.
We began calling these things metallo-carbohedrenes—or
wanted to make sure we were right before going public, so I
offered an escalating reward—which eventually reached a few
hundred dollars —to anyone in the lab who could disprove
our explanation. I
ended up keeping my money.
A few weeks later, we unearthed some more Met-Cars with as many as 22 metal atoms per molecular cluster. Their structures were remarkably different: Instead of a miniature soccer-ball, they had two or more balls that were connected by a relatively strong bond (a shared orbital bond), and importantly, that shared common faces. Their unique growth mechanism as well as their structure establishes these new molecular clusters as unique species and not in any way just variants of the buckyball.
filed a patent on these clusters, and that means we think they
could have useful applications.
For one thing, the abundance of free electrons means they
may be able to serve as dopants in semiconductors.
transition metals are good catalysts, Met-Cars could be used to
promote chemical reactions. And,
if one of the theories is accurate that says superconductivity
occurs at warmer temperatures as the molecular cages get smaller—well, our cages are much
smaller than those of C60 buckyballs.
this point, I think we're close to producing larger quantities of
Met-Cars, which will be essential to checking into practical uses.
basic lesson from all this is pretty simple, and not unique to our
something outside of your formal plans.
Be observant. Be
tenacious—keep returning to puzzle things out.
And always think about the potential significance of your observations.