Engineers Converting Atomic Force Microscopes into Molecular ‘Milling Machines’

Pratt School of Engineering researchers are at the vanguard of efforts to
remake the "atomic force microscope" (AFM), an instrument typically used to
obtain molecular scale images, into a tool to build precisely aligned
structures at those tiny dimensions.

"I think this will be a very good tool for research in the laboratory because
we should have very good control and get results relatively easily," said
Stefan Zauscher, an assistant professor of mechanical engineering and materials
science who brought his expertise in designing and using atomic force
microscopes to Duke.

Conventional AFMs allow scientists to image surfaces at resolutions approaching
the atomic scale by recording the motions of a sharp pivoting tip while the tip
moves over a surface in a way resembling a phonograph needle's interactions
with a record groove.

AFMs can also measure forces between molecules. And in recent years,
researchers have discovered that, besides imaging surfaces and sensing forces,
AFM's can be used like medicine droppers to deposit "nanoscale" amounts of
materials where desired.

The term nanoscale means billionths-of-a-meter dimensions, which can translate
into structures just 10 hydrogen atoms wide, noted Robert Clark, who is the
Thomas Lord professor of engineering and senior associate dean at Pratt.

Trying to use a conventional AFM for nanoscale construction projects introduces
problems of repeatability, reproducibility, accuracy and fidelity, said
Ashutosh Chilkoti, an associate professor of biomedical engineering and
materials science and engineering.

These issues, "really come from taking existing atomic force microscopes that
were actually designed to do something quite different," said Chilkoti, who has
worked with Zauscher to manipulate proteins with an AFM.

In response to these technical shortcomings, Clark and his laboratory group are
collaborating with Zauscher's team to modify AFMs so they can lay down
materials with more precision.

"We bring to bear control strategies that are beyond the typical instruments,"
said Clark. "I think some of the advances we'll make, both software and
hardware, have potentials for commercialization."

Duke researchers have already used conventional AFMs as molecular construction
tools through a procedure known as "dip pen nanolithography"(DPN) -- so-called
because the process works something like a fountain pen laying down ink.

Originally developed by Northwestern University professor Chad Mirkin, DPN
allows an AFM tip to pick up molecules and deposit them on another surface. The
molecular transfer is greatly aided by the presence of a "nanomeniscus," a
miniscule layer of water that condenses on the AFM tip in a moist atmosphere.

Jie Liu, a Duke assistant professor of chemistry, has used a DPN nanomeniscus
as a "chemical reactor" in which to synthesize electrically conducting organic
polymers out of molecular building blocks dissolved in the water. Liu then uses
the AFM tip to deposit the finished products onto the surfaces of other

The Pratt School's Chilkoti has also experimented with DPN as an alternative
way to do other kinds of microscopic chemical engineering within mere dots of
materials. In one experiment, he deposits tiny samples of protein mixtures on a
surface and then adds water and applies heat to separate one protein molecule
from another.

While Chilkoti has previously used flexible "microstamps" to lay out tiny and
precise patterns of chemicals, DPN allows experimentation with much
smaller-scale features, Zauscher said.

Chilkoti said that an atomic force microscope "allows you to build a structure
very quickly without having to go through expensive fabrication facilities."
For years, those facilities have used masks, chemicals and beams of energy to
etch and layer elaborate microstructures such as those engraved on computer

Using the latest energy sources, such as X-ray lithography, a centralized
fabrication facility can now turn out circuitry of nanoscale size for those
willing to pay and to wait. In contrast, the AFM approach "empowers people to
do nanoscience research in their own labs," Chilkoti said. "It's a relatively
low tech-way of building structures quickly. In principle, you can go in and
build something in an afternoon."

The problem with using a conventional AFM for such fabrication is that its
scanning system typically lacks the precise control and repeatability of
standard fabrication techniques, Zauscher said. For example, he said, in
attempting to place dots of material on a surface, the dots might not be
evenly spaced.

Such relatively poor spacing control is why Zauscher is collaborating with
Clark's Adaptive Systems and Structures Laboratory to build more-precise AFMs
for nanolithography.

Clark's group is designing and building an AFM with a "multi-axis" guidance
system to better control the manipulation of AFM tips. Ultimately, he envisions
a more advanced version that could create nanostructures on surfaces much like
a sophisticated milling machine creates patterns.

"We want to create a simple interface so that you don't have to be an expert in
programming the hardware to actually articulate the instrument," said Clark,
who is a mechanical engineer and materials scientist. Such a simple interface
would enable use of standard computer aided design (CAD) software packages to
draw patterns that the AFM would then automatically and precisely execute.

Another feature might be "adaptive controls," a specialty of Clark's lab, which
provides a way to adjust a controller according to varying conditions. Examples
might be responding to probe tips with different stiffnesses or surface
materials with changing properties.

Clark said his laboratory has already used special control technologies
experimentally for everything from maneuvering aircraft to controlling sound
and vibration.