|
Topic Name: The ability of a third-generation nanofountain probe (NFP)
Category: Nanofabrication
Research persons: Horacio Dante Espinosa & his groups
Location: Northwestern University ,
Department of Mechanical Engineering,Micro and Nanomechanics Lab 2145 Sheridan Road Evanston, IL 60208, United States
Details
The delivery, manipulation and assembly of functional materials such as metal
nanoparticles into predefined architectures and patterns is of great interest in
nanotechnology. Nanoscale patterns of nanoparticles have the potential to be
used in miniature electronic circuits or in plasmon waveguides to guide the
transport of electromagnetic energy below the diffraction limit. Nanoparticles
functionalized with biological materials can also be placed between electrodes
for use in biosensing applications.
Researchers from Northwestern University have now demonstrated the ability of a
third-generation nanofountain probe (NFP) to directly deposit gold nanoparticles,
15 nanometers in diameter, onto silicon substrates. The research is published
online by the scientific journal Langmuir.
“Such a direct-write method of deposition provides better control over resultant
patterns and simplifies the process of fabricating functional structures, as
compared to conventional photolithographic or microstamping techniques,” said
Horacio D. Espinosa, professor of mechanical engineering in the Robert R.
McCormick School of Engineering and Applied Science and co-author of the paper.
Espinosa’s group pioneered the development of the nanofountain probe.
The NFP is a cantilevered probe chip that can be mounted on commercial atomic
force microscopy (AFM) equipment. On-chip reservoirs hold liquid inks such as
nanoparticle solutions, which are delivered through enclosed channels to
ring-shaped apertured tips. High throughput microfluidic transport of molecular
inks to AFM tips is of great interest since fluid is a very effective medium for
the direct delivery of molecules, which self-assemble on substrates with very
specific nanoscale architectures.
“The ultimate goal of this project is to develop a robust microsystem platform
for the mass production of nanoscale devices, sensors and structures using
chemicals, biomolecules, nanoparticles, nanotubes and nanowires,” said Espinosa.
Previous versions of the nanofountain probes were shown to be capable of
depositing solutions of fluorescent dyes, alkanethiols and DNA. Among several
design changes, the latest nanofountain probes have deeper microchannels to
allow the facile delivery of larger particles such as gold nanoparticles 15
nanometers in diameter.
Probe-based deposition techniques are amenable to high-resolution, nanoscale and
flexible patterns in which the desired structure can be easily altered at any
time. Dip-pen nanolithography (DPN), in which a commercial AFM probe is coated
with molecules to be deposited, is capable of making high-resolution patterns of
many chemicals and biological materials. However, standard DPN techniques have
not been able to deposit suspensions of solid nanoparticles.
“The nanofountain probe is not only capable of delivering such solutions but can
do so continuously because the inks are contained in reservoirs on the chip,”
said Andrea Ho, a co-author and graduate student in Espinosa’s group.
Because NFPs are batch-fabricated using standard micromachining processes, they
can easily be mass-produced. The current NFPs produce nanoscale patterns with a
linear array of 12 writing tips, but their design allows for straightforward
scaling up to 2D arrays of tips. This would allow for the high-throughput,
parallel deposition of nanoparticles with high resolution.
About Researchers:
Horacio Dante Espinosa
Professor
Phone:
(847) 467-5989
Fax :
(847) 491-3915
Email:
espinosa@northwestern.edu
Office:
Northwestern University
2145 Sheridan Road
Department of Mechanical Engineering
Evanston, IL 60208
Phone: (847) 467-5989
Fax: (847) 491-3915
Email: espinosa@northwestern.edu
URL: http://clifton.mech.northwestern.edu/~espinosa/
Home:
559 Arbor Vitae
Winnetka, IL 60093
Phone: (847) 441-9389
Lab address:
Dynamic Inelasticity Laboratory and
Mechanics of MEMS and Nanosystems
Department of Mechanical Engineering
Northwestern University
2145 Sheridan Rd.
Evanston, IL 60208-3111
Phone: (847) 467-5989
Fax: (847) 491-3540
Andrea Ho,
Graduate Students
Micro and Nanomechanics Lab and
Dynamic Inelasticity Lab
Department of Mechanical Engineering
Northwestern University
2145 Sheridan Rd.
Evanston, IL 60208-3111
Phone: (847) 467-5989
Fax: (847) 491-3915
In The Images:
1.This slide highlights the ultra-nano-crystalline diamond (UNCD) atomic force
microscope (AFM) cantilever. (a) This is a scanning-electron micrograph of a
UNCD cantilever with a tip. (b) This is a scanning electron micrograph of a UNCD
tip after one hour of scanning on a diamond substrate. Inset shows the tip
before the scanning. (c) This is a frictional AFM image of an alkanethiol
monolayer patterned onto a gold substrate with a UNCD tip. Researchers used the
same tip for patterning and imaging. (d) This scanning electron micrograph shows
a commercially available silicon nitride tip after one hour of scanning with the
same parameters used for the UNCD tip in (b). The image shows damage at the tip
apex while the inset shows the tip prior to the test.
2.a longitudinal section of the undeformed and deflected membrane is shown..
3.Stress field at the end of the deflection for point load case.
4.Deformed shape. Comparison for flat and wavy cases. Nothe the different scales
used for the z-axis
5.This slide highlights the components and features of the nanofountain probe
6.This scanning electron micrograph shows the new nanofountain-probe dispensing
tip.
7.The surface of the membrane presents an initial shape that governs the first
part of the load-displacement curve
8.a MTS/ Nano Instruments Nano-indenter XP, with CSM module. Results for
hardness and Young's modulus are given
9.The scanning electron micrograph shows a nanofountain-probe chip, including
the on-chip reservoir.
10.The new nanofountain probe produced these patterns; features are as thin as
40 nanometers.
11.Initial shape of the membrane for line load case. Refinement of the mesh is
shown in the detail
12.Effect of geometry for flat, x-wave and xy-wave for the case of Point load
13.a MTS/ Nano Instruments Nano-indenter XP, with CSM module
14.Modeling of differnet tips. The Berkovich tip was supposed conic instead of
pyramidal
15.Mesh and deformed shape of the a model with the hole pattern for analyzing
the case of point load on these membranes. A quarter of the membrane is
discretized.
Notes-
Nano-indentation-
This method is a derivation of the "hardness test" for bulk materials, performed
over small samples. Nevertheless, instead of indenting and comparing indentation
marks sizes, the force and penetration are measured during the loading-unloading
process, getting a force-displacement curve. A diamond tip penetrates the
surface, driven by a coil component, to measure the elastic and plastic
properties of materials in very small volumes. Several companies are now
building depth-sensing indentation instruments and many research groups have
constructed custom-made devices for studying indentation at the sub-micrometer
level.
MTS/ Nano Instruments Nano-indenter XP, with CSM module. Results for hardness
and Young's modulus are given in the figure-8 & 13.
MDE Tests
Membrane Deflection Experiments are possible using the nanoindenter head, able
to measure loads and displacements very accurately. The experiment consists in
applying a line load over the mid section of the fixed-fixed membrane. In order
to do that, the nanoindenter is used to push the membrane down with a specially
designed diamond tip, 250 microns long. In this manner, simple tension of the
membrane is achieved, an so no assumptions or complicated mathematical models
are required for the data reduction. Assumptions can lead to errors in these
types of problems, since many variables are involved and unknown phenomena could
occur.
The surface of the membrane presents an initial shape that governs the first
part of the load-displacement curve. Then, the membrane behavior is obtained as
it is pulled down over the gap, until the bottom is reached. A third part of the
curve then corresponds to the behavior of the bottom material under indentation.
The membrane is not damaged during the test, and no mark was found after it.
The first test was carried out at MTS/Nano Instruments, Oak Ridge, TN, thanks to
the collaboration of Dr. Warren Oliver and his student, Erik Herbert. Lots of
them are still being carried out there, while they design and build our
AFM-Nanoindenter station. By integrating the AFM to the nanoindenter, a very
complete 3D characterization of the specimen topography can be done before
running the test. The Nanoindenter XP can also be used as an extremely precise
profilometer.
The tests are very accurate and simple. They can also be done dynamically and no
temperature effects can nullify the experiments. The device is designed in the
way that unless a desired thermal drift criterion is reached, the experiment is
not engaged. The thermal drift is corrected on the equipment using a built-in
function that sits the tip on the surface and waits for the drift to be less
than a certain value (in the order of angstroms) to start the test. The gap
under the sample is used to deflect it while continuously measuring the applied
load and displacement.
Other tests over circular and square membranes existing in the same wafer were
performed, since they are made of the same material, and the independence of
stiffness with the shape is going to be investigated
FEM simulations
As previously indicated, one of the main objectives in the mechanical
characterization of MEMS structures is the determination of residual stresses in
the various components. In the case of the MEMS switch, knowledge of the
residual stress in the membrane is crucial.
In this section we discuss the finite element idealization employed in the
analysis of the deflection tests. The bow tie membrane is discretized using
shell elements with variable size in the x-y plane and constant thickness in the
z-direction, as shown. The edges of the membrane along the posts are fixed to
simulate the bonding between the membrane and the aluminum posts. The diamond
wedge-tip is modeled as a rigid body with a tip radius of 40 nm and 80 degree
incline angles. Initial stress in the x-direction are applied with variable
magnitude to examine its effect on the load-deflection curve.
The analysis is performed with Abaqus, version 5.7, in two steps. Nonlinear
geometry is included in the analysis to examine the changes in membrane
stiffness with deflection. The first step consists in equilibrating the initial
stress and achieving contact between the wedge-tip and the membrane. The second
step consists in applying a prescribed displacement to the wedge-tip of 3.5
microns. The membrane constitutive behavior is assumed linear elastic with a
Young's modulus of 70 GPa and a Poisson's ratio of 0.33. These moduli are based
on the measurements previously reported. Moreover, the elastic assumption is
appropriate for the maximum imposed deflection and the estimated aluminum yield
stress.
In this figure-2 a longitudinal section of the undeformed and deflected membrane
is shown. Note that a magnification factor of 5 is used to enhance the
observation of the deformed shape. It is clear that with the exception of small
regions at the fixed end and the point of application of the line load, membrane
stresses dominate.Here the load-displacement curves for various initial stresses
ranging from 0 to 40 MPa are represented. When the initial stress is above 20
MPa, the membrane stiffness does not increase with deflection as observed
experimentally. By contrast, when no residual stress is present or its value
remains low, a few MPa, the membrane stiffness increases progressively with
deformation. When the load-deflection curve, in the presence of low residual
stresses, is compared with the experimentally obtained load-deflection curve, it
is clear that the initial part of the curve is controlled by the initial shape
of the membrane. In fact, the calculations start with a flat membrane while the
tested membrane presented a wavy initial shape with is center displaced up by
approximately 1 micron. If the total force at a displacement of 3 microns is
examined, good agreement with the experimental measurement is observed at low
values of initial stress.
It needs to be pointed, from the discussion in the previous paragraph, that
accounting for the initial membrane shape is very important in view that small
initial offsets in load and displacement result in maximum load changes, at
maximum deflection, of the order of changes predicted by small variations in
membrane residual stress.
Some calculations have been carried out to account for the initial membrane
shape effect in the numerical prediction of the load-deflection curve. One of
the major fining is that initial membrane shape significantly influences the
load-deflection curve. Because of difference in materials thermal expansion
coefficients and device geometry, the membrane has a well-defined initial
geometry. Our study demonstrates that when the membrane has initial deflections,
it may in troduce structural instabilities, which cannot be neglected.
Next figures and animations show the effect of initial membrane shape on the
load-deflection curve and on the final shape of the membranes.
|
