mechanical device created from a handful of molecules. Such a device
could be a sensor that can detect infinitesimal traces of chemicals
or biological agents, or it could be an onoff switch, a miniature
building block for creating molecular computers. These ideas are
moving closer to reality because of recent work on mechanically
interlocking molecules at Livermore.
molecules that are joined by covalent bonds (by sharing pairs of
electrons), mechanically interlocking molecules are physically joined,
in the same way as the links in a chain or the rings in the Olympic
Games symbol. Mechanically interlocking molecules are of growing
interest to synthetic chemists, such as Livermores Andrew
Vance, who view them as potential building blocks for future molecular-scale
devicesmotors, sensors, and machines on the nanometer scale.
heads up an effort originally funded through Laboratory Directed
Research and Development to find new mechanically interlocking molecules
that will consistently attach in a single layer on a gold surface.
The team, which includes Vance, physicists Anthony Van Buuren and
Art Nelson, and University of California at Davis physics graduate
student Trevor Willey, is focusing on molecules known as catenanes
and rotaxanes. A catenane has two or more interlocking rings. A
rotaxane consists of a long, straight moleculean axleringed
by a doughnut-shaped molecule. Molecular caps at both
ends of the rotaxanes axle keep the ring from sliding off.
No chemical bond holds ring to ring or ring to axle.
notes, Interlocking molecules in solution are well understood,
but not much is known about them on surfaces. Questions such as
which molecules attach best, whats the best way to determine
how well theyre attached, and how to make the attachments
stable are virtually unexplored.
|(a) Catenanes are formed from
two or more molecular rings interlocked like links in a chain.
(b) Rotaxanes are formed from a ring molecule threaded on an
axle molecule capped by two molecular stoppers.
see what was happening with the synthesized
molecules and the gold surface, the physicists in Andrew
Vances team used two techniquesx-ray absorption
spectroscopy (XAS) and x-ray photoemission spectroscopy
(XPS). Both techniques use x rays from the soft end of
the spectrum (with wavelengths between those of ultraviolet
light and harder, medical x rays). For the photoemission
measurements, the team used the soft x rays created by
a beamline at the Stanford Synchrotron Radiation Laboratory
as well as a newly acquired Physical Electronics Quantum
2000 scanning XPS system here at Livermore.
The two techniques
enable scientists to obtain detailed and specific information
about the monolayers under scrutiny. According to physicist
Trevor Willey, XPS measurements reveal the chemical composition
of whats on the surface as well as the nature of
chemical bonds between the surface and the material. We
used XPS to determine whether the sulfur atom of the thiol
was bound to gold or whether the thiol was just lying
on the surface, essentially unattached, he explains.
Measurements with XAS revealed the orientation of the
molecule. With XAS, we could tell whether the thiol
molecules were standing up, lying down, or leaning in
some direction relative to the gold surface, Willey
says. This also gave us information on how well
ordered the layer wasthat is, whether the attached
molecules were packed together in ordered domains or leaning
randomly every which way.
Looping the Loop
The team started by exploring what kind of molecules
worked best for forming a loop on a gold surface. Forming loops
is the first step toward creating a monolayer of catenanes, in which
each attached loop would thread a ring. Vance explains, We
started at the most fundamental level, looking at how different
molecules attached to the surface and how well they attached.
The challenge was to come up with a molecule that would consistently
attach at not just one but both ends.
The team first tried a linear
dithiol monomer. (A thiol is a molecule that has an atom of sulfur
bound to an atom of hydrogen. This particular monomer had a thiol
at each end; hence, it is a dithiol.) The monomer was, according
to Vance, a floppy molecule. The researchers reasoned
that when the sulfur atom at one end attached to the gold surface,
the molecule would flop over and the sulfur at the other end would
also attach, forming a loop. But measurements taken with x-ray absorption
(XAS) and x-ray photoemission spectroscopy (XPS) at the Laboratory
and at Stanfords Synchrotron Radiation Laboratory revealed
that only about 50 percent of the sulfur atoms had bonded to gold.
(See the box below.) In other words, most of the monomers
had one unattached sulfur, says Van Buuren.
Other measurements indicated
that these monolayers were disordered and had molecules that, on
average, were tilted slightly more than 55 degrees from the surface.
All these data indicated that most monomers were essentially
standing on end on the surface, notes Van Buuren. Because
the concentration of monomers was quite high, the suspicion is that
the monomers packed the surface, leaving little room for them to
flop over and make a loop.
Another set of experiments
used a polymer containing disulfide components. (A disulfide is
two atoms of sulfur bound to each other.) In this case, the scientists
expected the disulfide bonds to cleave, the sulfurs to bind to gold
atoms, and the polymer to form a loop. XPS measurements showed that
the resulting monolayer contained over 90 percent of bound sulfurs,
evidence that nearly all the molecules had successfully formed surface-attached
The presence of a disulfide
made it more likely that both sulfurs would attach to the surface,
says Vance. All this pointed out the importance of designing
molecules that will bind to surfaces in a predictable manner to
form monolayers. In the case of surface-attached loops, simply preparing
compounds with end components that bind well to the surface doesnt
guarantee loop formation. Other factors come into play, including
the solution concentration and the shape of the molecule. Following
these initial results, we also looked at molecules with built-in
turns that encourage loop formation over single attachment.
|The team experimented with
creating loops from different molecules. (a) When the team tried
to attach monomers terminating with sulfurhydrogen bonds,
the vast majority attached at just one end instead of forming
the hoped-for loop on gold. (b) A polymer with disulfide linkages
was far more successful in attaching its sulfurs to the gold
Axles and Rings
Next, team members turned
their attention to attaching rotaxanes to gold. As in the previous
experiments, the team took two different approaches. One involved
an electron-deficient, positively charged T-shaped thiol (which
had the characteristic sulfurhydrogen bond at one end and
an anthracene stopper at the other) and an electron-rich crown ether
ring. In solution, the two molecules are drawn to each other. The
thiol threads the crown ether to form a pseudorotaxane with only
one stopper. Our question was whether the pseudorotaxane would
then attach to the gold, or would the crown ether ring slip off,
leaving only a thiol to attach to the surface, explains Vance.
Vance also synthesized a rotaxane composed of three molecules:
two crown ether rings threaded by an antracene-capped thiol with
a disulfide bond in the middle. From our work with loops,
we felt confident that when the disulfide bonds cleaved, wed
get surface-attached rotaxanes, says Vance.
The teams physicists
took spectra of powder samples of the crown ether rings, anthracene
caps, and rotaxane for reference as well as a spectrum of thiol
attached to gold. When the experimental results were compared to
these control spectra, the spectra from the rotaxane experiments
did indeed show surface-attached rotaxanes. However, the spectra
from experiments with the pseudorotaxane precursor were identical
to that of simple thiol on gold.
The results confirmed
that even though wed set up a process for the thiol to thread
the rings in solution, the rings came off before the sulfur could
attach to the gold, says Vance. For the rotaxane,
the rings were locked into place by the endcaps right up until the
two sulfur atoms cleaved and adsorbed to the gold.
|Results of surface-attached
rotaxane research. Researchers started with thiols in solution
(T-shaped molecules), attached them to gold, and used the resulting
spectrum from x-ray absorption spectroscopy as a reference.
Beginning with crown ether rings and anthracene-capped thiols
in solution, researchers attempted to create pseudorotaxanes
(single rings on axles capped at one end) and attach them to
gold. The resulting spectrum was nearly identical to the reference,
indicating that almost all the rings slipped off before the
thiols attached. But using disulfide rotaxanetwo rings
threaded on a thiol and restrained by anthracene endcapswas
effective in creating a surface-attached rotaxane. The resulting
spectrum shows peaks similar to the reference spectrum for rotaxane
to Future Possibilities
date, the Livermore team is one of only a few groups that has been
successful in repeatedly forming rotaxane monolayers on surfaces.
This consistency in results is important if such surface-attached
molecules are to become molecular machines of the future.
The success of Vance and
his colleagues has led to a collaboration with a group from the
University of California at Los Angeles (UCLA) led by chemistry
professor and researcher Fraser Stoddart, one of the worlds
foremost experts in the synthesis of rotaxanes and catenanes. Vances
team has been taking interlocking molecules created by the UCLA
group, attaching them to surfaces, and using XAS and XPS to examine
In addition, the team continues
its fundamental studies of surface attachment and is beginning to
explore attaching molecules to other surfaces, including silicon.
Furthermore, they are working on ways to create surface-attached
catenanesloops with rings. Beyond this near-term research
are the long-term goals of creating sensors with properties that
can be controlled at the molecular level. Such a sensor might have
arrays of surface-attached catenanes with rotating rings that have
|Members of the interlocking
molecules research team. From left, Trevor Willey, Anthony Van
Buuren, Cheryl Evans, and
By controlling the
rotation of the ring, for instance, we can create an onoff
switch, Vance explains. Suppose you could create a ring
that has a small hydrophobic component. In the presence of a water
molecule, it would spin one way; without water present, it would
spin the other way. You could also create rings that are electrochemically
or optically reactive and turn them on or off by changing the charge
at the surface or by the presence or absence of light. Developing
these switchable features is on our list of plans for the future.
believes the combination of synthetic chemistry and spectroscopy,
chemists and physicists is a critical element in making this research
possible. What were doing now, he continues, is
fundamental science that has intriguing possibilities. In the future,
these surface-attached interlocking molecules could be used in molecular
machines, sensors, and electronics in ways weve yet to even
imagine. It all comes down to being able to understand and
control these small structures on the molecular and atomic level.
And its a combination of chemistry and physics that, in the
end, will make this possible.
Key Words: catananes,
molecular machines, rotaxanes, Stanford Synchrotron Radiation Laboratory,
surface-attached mechanically interlocking molecules, x-ray absorption
spectroscopy (XAS), x-ray photoemission spectroscopy (XPS).
For further information contact Andrew Vance (925) 423-9166 (firstname.lastname@example.org).