Lubricating nanoscale machines: Unusual behavior of highly confined fluids challenges conventional expectations

Contact: John Toon

john.toon@edi.gatech.edu

404-894-6986

Georgia Institute of Technology

The unique and often unexpected properties of fluids confined to very small
spaces will force designers of future nanometer scale devices to reexamine
conventional expectations regarding lubrication and fluid flow.

At these small size scales, considerations pertaining to molecular architecture,
structural conformations and packing — along with the increased importance
of surface roughness, surface tension, frictional losses and fluctuations
— produce dramatic changes in the behavior of lubricants and other fluids.
These considerations come into play as devices approach the size of lubricant
molecules that interact with them.

“We are accumulating more and more evidence that such confined fluids behave
in ways that are very different from bulk ones, and there is no way to extrapolate
the behavior from the large scale to the very small,” said Uzi Landman, director
of the Center for Computational Materials Science at the Georgia Institute
of Technology. “We must find clever ways to harness and control these new
behaviors in order to realize the opportunities in nanotechnology.”

Using supercomputer-based molecular dynamics simulations to model the behavior
of these fluids at the atomic and molecular level, Landman’s research center
has developed a series of predictions that will help guide future device designers.
Some of the theoretical predictions have already been borne out by experimental
results.

Among the predictions:

ï When confined to tight spaces, long-chain lubricant molecules act more
like “soft solids,” forming, for energetic and entropic reasons, ordered layers
that significantly influence the movement of sliding surfaces. This poses
significant challenges in systems such as ultra-high-density computer disk
drives.

ï Confined fluids composed of molecular mixtures segregate themselves by
size, with the longer chain molecules adsorbing near the surfaces and the
smaller ones remaining in the middle region of the confining gap.

ï Liquid jets just a few nanometers in diameter propagate over shorter distances
than predicted by conventional fluid flow equations. Generation of such nanojets
requires special treatment to prevent clogging of the nozzle.

ï Molecular fluids confined between slightly roughened surfaces exhibit a
more liquid-like behavior than when confined by smooth surfaces, resulting
in significant modification of the resistance to sliding.

In addition to pointing out potential issues involved in the nanotribology
of future nanodevices, the simulations — involving as many as hundreds of
thousands molecules — also allow researchers to explore and test potential
solutions.

Writing in journals such as Science, Nature, Physical Review
Letters, Langmuir, and the Journal of Physical Chemistry,
Landman’s research group has reported on the tendency of lubricant molecules
such as hexadecane and other molecular fluids to form highly ordered layers
in planes parallel to the motion of the confining surfaces. On size scales
that approximate multiples of the molecular width, these layered lubricants
appear to increase their viscosity, “becoming, at equilibrium and at various
stages of the sliding motion, liquid-like in the plane parallel to the sliding
surfaces and solid-like in the direction perpendicular to the surfaces,” Landman
said.

This phenomenon manifests itself in several ways, including an increasing
amount of pressure required to squeeze the lubricant out of the confining
spaces. The pressure required shows distinct steps that correspond to the
molecular diameter, suggesting the lubricant is squeezed out layer by layer.

“The confinement of these liquids brings about sluggishness to their response,”
Landman explained. “Viscosity and other concepts that we commonly use are
taken from bulk behavior, and one of the questions we must answer is whether
it is appropriate to adopt the same concepts on the molecular levels.”

Increased friction caused by nanoconfinement-induced layering poses a significant
concern for future devices, but Landman and his colleagues propose several
techniques for countering it:

ï Chemically altering the long-chain molecules to include branched structures
that inhibit the formation of layers. The researchers have shown that a nanoconfined
liquid made of branched alkane molecules has a lower viscocity then a confined
liquid of the same molecular weight but made of straight chain molecules.
This behavior is opposite to that found in much larger environments.

ï Roughening the surfaces of the confining plates to disrupt the molecular
ordering. Instead of forming ordered layers, the molecules closest to the
rough surfaces adhere to them, leaving free-flowing molecules in between.
Consequently, patterning of the surface morphology could be used to control
friction and lubrication processes.

ï Varying the distance between the two confining surfaces in an oscillatory
manner, just enough to keep the lubricant molecules in a “frustated” state
of disorder. Varying the distance by one Angstrom in a 20-Angstrom gap should
be enough to prevent the layering. The frequency of the applied oscillations
depends on the characteristic molecular relaxation times and the viscosity
of the lubricant, which in turn are governed by the nature and structure of
the fluid molecules.

Recent work published by Landman and post-doctoral fellow Michael Moseler
in Science predicts the feasibility of generating nanojets just a few
nanometers in diameter. Such jets could one day be used for printing circuitry
patterns, injecting genes into cells, producing droplets of uniform size and
serving as fuel injectors for nanoengines.

The nanojets, however, would differ significantly from their larger cousins.
For example, nanojets would have to overcome the effects of surface tension
and wetting that are of much less importance at larger scales. In a nanojet,
liquid molecules wet the outer surface of the nozzle, eventually creating
an adsorbed film that clogs the nozzle. To prevent that, the researchers suggest
heating the outer surface of the nozzle to evaporate the condensing films,
or coating the outer surfaces with an anti-stick non-wetting compound.

In a second phase of the nanojet simulations, the researchers reformulated
the traditional hydrodynamics equations to include fluctuations whose influence
becomes dominant at small sizes. The newly derived equations extend hydrodynamics
to the nanoscale, and they were shown by Moseler and Landman to yield results
that agree with their atomistic simulations.

Molecular dynamics simulations allow researchers to study the behavior of
each atom and molecule in a system with very fine resolution in space and
time by integrating the equations of motion with interatomic interactions
derived from quantum mechanical calculations and/or experimental data from
larger systems.

The classical and quantum mechanical simulation methodologies developed by
Landman and his coworkers were the basis for his 2000 Feynman Prize in theoretical
nanotechnology. These “computational microscopies and spectroscopies” allow
scientists to make predictions and draw molecular-based designs that could
guide the fabrication of devices this small.

“In the nanorealm there is a whole new world that is full of surprises and
opportunities,” added Landman.

Research News & Publications Office

Georgia Institute of Technology

430 Tenth Street, N.W., Suite N-116

Atlanta, Georgia 30318 USA

Media Relations Contacts: John Toon (404-894-6986); E-mail: (john.toon@edi.gatech.edu); Fax:
(404-894-4545) or Jane Sanders (404- 894-2214); E-mail: (jane.sanders@edi.gatech.edu);
Fax: (404-894-6983).

Technical Contact: Uzi Landman (404-894-3368); E-mail: (uzi.landman@physics.gatech.edu);
Fax: (404-894-7747). Note: Landman will be unavailable until
February 12.

Visuals Available: Landman and Moseler with nanojet simulations, nanojet
movie, portrait of Landman with early molecular dynamics simulations, molecular
dynamics simulations showing nanojets, ordered layers of lubricants.