|
By Rishi Gupta and Richard Stallcup, Zyvex
Instruments
Download a printable copy here.
Introduction
Over the last few years, researchers have developed many
novel nanostructures, such as tubes, coils, and wires. Much
of advanced materials research is being focused on the creation
of stronger and lighter materials using these structures1.
Nanotubes and wires are also of interest for the engineering
of molecular electronics2, biomedical devices3,
and nanoelectromechanical systems4. Theoretical
models have predicted that these structures exhibit astounding
mechanical and electrical properties5,6, but characterizing
the physical properties of isolated nanostructures in
situ has presented a challenge to the research community.
Many researchers currently employ arduous lithography and
microscopy steps to characterize electrical properties of
carbon nanotubes. Dispersed nanotubes are first deposited
onto a substrate and isolated nanotubes are then found by
atomic force or scanning electron microscopy. Once a single
nanotube has been located, metal pads are deposited in just
the right area so that electrical measurements can be carried
out using microprobes. Devices are made using similar methods7.
In contrast, nanomaterial measurements can be conducted in
free space using the Zyvex S100 Nanomanipulator, which eliminates
the need for complex processing that can affect the physical
properties of the material. The structures can be held, stressed,
twisted, rotated, cut, and otherwise manipulated as if they
were macroscale objects. The Zyvex S100 Nanomanipulator System
effectively solves the problems afflicting researchers working
to characterize nanoscale materials.
The Zyvex S100 Nanomanipulator System is designed specifically
for manipulating and characterizing nanostructures and can
serve as a platform to engineer nanoelectromechanical systems.
The S100 System is a piezo-motor based manipulation system,
which offers less than 5 nanometers of precision and 12 millimeters
of travel range. The center stage rotates 360 degrees clockwise
and counterclockwise with a resolution of 3 microradians.
Each positioner has 5 electrically isolated input/output lines
that can interface with test bench hardware through BNC connectors
on the control cabinet. Coupled with low noise amplifiers
and the NanoEffector™ suite of end-effectors, the S100
is capable of performing mechanical and electrical characterization
on nanostructures in scanning electron microscopes (SEM).
The S100 is capable of performing many types of characterization.
For nanowires, the current-voltage (IV) response is typically
measured to determine the conductance (or inverse of resistance).
This is done by contacting the nanowire with the S100 probes,
which are connected to a current source and meter. For semiconducting
wires, the IV response is non-linear and is used to determine
the saturation limit of the nanowire. The S100 can also perform
measurements to determine the field-induced current in a nanowire.
By applying a field to a nanotube suspended between two probes,
the induced current can be measured using current-meters.
The S100 is also capable of performing mechanical measurements
on nanostructures. Typical mechanical measurements that can
be performed with the S100 are Young’s modulus, breaking
strain, and compression strain. Each can be taken by using
cantilevered probes of known moduli, or by utilizing Zyvex
Corporation’s extensive line of microelectromechanical
systems (MEMS) -based devices.
One of the purposes of this application note is to describe
methods for mounting a multiwalled carbon nanotube (MWNT)
to an end-effector, which is the first step towards characterizing
any isolated structure or engineering a nano-scale device.
This note presents a discussion of the necessary materials,
the set-up of the experiment, and the detailed procedure for
mounting the nanotube. The reader will also learn about electron
beam induced deposition (EBID), which is used as an adhesive
binder, much as a weld is used to hold two objects together.
Though this note discusses MWNTs, one can use this procedure
to mount very small structures of different shapes, sizes,
and composition. Finally, procedures are detailed which are
useful for characterizing physical properties of nanostructures,
such as conductance and Young’s modulus.
Multiwalled Carbon Nanotube
The nanostructure discussed in this note is an individual
MWNT. There are many suppliers and manufacturers of nanotubes.
Unless the nanotubes are in aligned arrays, they will usually
need to undergo some processing to loosen and untangle them
prior to manipulation and testing. Many processes exist to
unbundle tubes, such as sonication and treatment with surfactants,
but these processes will not be discussed. See references
eight and nine for additional information.
The nanotubes should be on a wire or flat surface. It is
important that the long axes of the nanotubes protrude from
the surface into free space and be orthogonal to the electron
beam of the microscope. This allows the probe to gain access
to the nanotubes, and it will also help the user to image
the tubes under the SEM (Figure 1). A good
candidate nanotube tends to be long, relatively thin, and
disentangled from bundles or neighboring tubes.
Figure 1: Multiwalled carbon nanotubes
(Alfa Aesar).
End-Effectors
End-effectors are defined as the “tools” at the
business end of the manipulator. The user selects the type
of end-effector according to his/her specific application.
The S100 functions with various types of end-effectors.
• NanoEffector™ probes are used for manipulation
tasks that require probes with very small tip radii; they
can be created with a radius of curvature from hundreds of
nanometers to less than 15 nm (Figure 2).
The material of the probes is selected depending on the specific
application. Tungsten probes are used for picking and placing
of nanotubes; palladium (Pd) wires are well suited for electrical
measurements and probing nanodevices (Pt and Pd coated tungsten
wires are an alternative).
Figure 2: Zyvex NanoEffector™
probe
• MEMS-based grippers are used for grasping nanotube
bundles and for mechanical characterization work.
• Cantilevered probes with quantifiable moduli are
often used for mechanical characterization (Figure
3).10
Figure 3: Atomic force microscope tip
(TappiMode, MikroMasch).
The NanoEffector™ product suite is available from
Zyvex Corporation. Call 877-ZYVEX99 (877-998-3999) ext. 271
or email sales@zyvex.com for more information.
Set-up
It is very important to set up the S100 properly before starting.
All of the probes must make electrical contact to the cabinet.
NanoEffector™ probes must be handled carefully to avoid
damaging the tips.
The nanotubes need to be on a surface that is accessible
to one or more of the probes. The rotational stage is a good
place to mount the tubes, as opposed to mounting them in one
of the end-effector plugs. In either case, the tubes should
protrude from the surface so that they are normal to the electron
beam and clear of obstructions. This will allow the probe
to approach them while imaging at high resolution.
Once the S100 is in the SEM chamber, verify that the feed-through
connectors are mounted properly. Always follow the procedure
for handling and operating the S100 as prescribed in the S100
User’s Manual.
Electron Beam Induced Deposition
EBID is a by-product of electron microscopy, but serves as
a useful “welding” agent for nanotubes. EBID is
caused by the dissociation of surface adsorbed molecules (hydrocarbons,
for example) by high-energy electrons. EBID layers will act
as a binder as the deposited amorphous material coats the
nanotube.
Optimizing the Deposition
Koops, et al., have investigated this phenomenon
for a number of years11. From the literature, we
see that the density N of adsorbed molecules on the
substrate surface varies with time:
dN/dt = g * F * (1 – N/N0)
– N/ t – q * N
* f, (AppEq 1)
using g as the sticking coefficient, F
as the molecular flux density arriving on the substrate, N0
as the molecule density in a monolayer, t as the
mean lifetime of the adsorbed molecule, q as the
cross section for dissociation of the adsorbed molecules under
electron bombardment, and f as the electron flux
density. The layer growth rate, R, is
R = v * N * q * f,
(AppEq 2)
where v is the volume occupied by a dissociated
molecule or its fractions.
From AppEq 2, we can see that the growth rate depends on
the cross section for dissociation of molecules in the path
of the electron beam. From quantum mechanics, we know that
the differential cross section for scattering is inversely
dependent upon the energy of the incident particle. One should
adjust the accelerating (EHT) voltage of the electron beam
to optimize the deposition efficiency. As a rule, the lower
the EHT value, the higher the electron cross section, allowing
for a higher probability of electron-molecule interaction.
However, there is a point of diminishing returns. The cross
section will drop off at very low energies. Also, a lower
mean free path will increase the EBID efficiency. Higher pressures
result in an increased probability for electron-molecule collisions.
EBID is not a “fire and forget” technique. If
the SEM image drifts away from the target region, material
will be deposited onto undesirable areas. This could ruin
the experiment, so care must be exercised to avoid image drift.
Applying EBID to weld a nanotube to a probe is described later
in this document.
An Alternative to EBID
Scientists have reported that an effective method for attaching
nanotubes to probes is to dip the probes in glue before making
contact with nanotubes. For example, Dai, et al.
demonstrated attachment of nanotubes to scanning probe microscopy
probes by bringing the probe tip in contact with carbon tape
in situ12. This method is easily performed
with the S100.
For this process, carbon tape (acrylic adhesive) is necessary
in addition to the aforementioned materials. Apply the carbon
tape to an unsharpened wire and place it into one of the inputs
on the end-effector plug. It can also be placed on the center
stage. Bring the probe into contact with the adhesive. Figures
4 shows what a bead of adhesive looks like under
electron microscopy;
Figure 4: Electron micrograph of a
hook-shaped sharpened tungsten wire on approach to a bead
of adhesive. The bright region is charge build-up.
Figure 5 shows the hook-shaped probe immersed
in the adhesive. Retract the probe and approach the target
nanotube. The drawback of using the adhesive is less control
over where, when, and in what state the nanotube sticks to
the probe. It may also be difficult to ensure ohmic contact
to the probe even though the carbon tape tends to be electrically
conductive.
Figure 5: The tungsten wire submerged
in the acrylic adhesive.
If the SEM has lithography capabilities or a focused ion
beam, metals may be used as the binding material. The focused
ion beam works in a similar manner as EBID, but one should
follow the manufacturer’s instructions. An integrated
lithography system can be very useful for depositing and patterning
materials other than carbon.
Performing the Experiment
This section will describe the procedure for making contact
to the nanotube and detail the processes needed to optimize
the visual feedback.
SEM Modes of Operation
There are different modes of operation for each microscope
(i.e., reduced scan or spot mode). One can observe the motion
of the manipulator and the nanotubes with high resolution
in near real-time by utilizing the “reduced scan size”
operating mode, where the electron beam raster scans across
a smaller area. The ideal mode of operation for applying EBID
is “spot mode,” where the electron beam is fixed
upon a single point.
Approaching the Carbon Nanotube
The best way to start approaching a carbon nanotube (CNT)
is to select a target nanotube, as defined above. Begin by
using the coarse positioners to bring the end-effector within
10 microns of the nanotube (the range of the fine positioners
is 100 microns in the X- and Z-direction and 10 microns in
the Y-direction using the “Local Coordinate System”).
The coordinate system is best described by taking the positive
Y-direction to be towards the center stage for each positioner.
The positive Z-direction is out of the plane of the center
stage. The X-direction is then orthogonal to the Y- and Z-directions.
Figure 6 shows a probe that can reach a bundle
of nanotubes using the fine positioners.
Figure 6: On approach to the nanotube
(circled) using the coarse positioners (Mitsui CNTs, Leo 1550
FE-SEM, Zyvex S100).
The nanotube is best approached from underneath. This way,
you can see it deflect when contact is made, and the probe
will be in the best position to apply EBID (Figure
7).
Figure 7: Fine positioner adjustments
to maneuver close to the nanotube.
Using the Z-axis of the fine positioner, slowly bring the
end-effector into contact with the nanotube (Figure
8). Verify contact by watching the nanotube deflect.
Figure 8: Contact to the nanotube.
If the nanotube is loose enough, the user can remove it from
the surface via van der Waals forces alone (van
der Waals forces arise from induced dipoles in neighboring
molecules that result in a relatively weak attraction). At
this point, retract the probe by moving the fine positioner
in the negative Y direction. The nanotube should slip out,
being stuck to the probe. If the nanotube needs to be bound
more strongly to the probe, try performing electron beam induced
deposition to affix it to the probe.
Applying EBID
To apply EBID, use either a reduced scan size or the spot
mode of operation. If possible, decrease the electron beam
voltage (lower electron energy increases the electron cross
section, increasing the probability of dissociation from electron-molecule
collisions). We have found that 1.5 – 5.0 kV works well
for us. While in focus on the nanotube and the probe, scan
the reduced area for several minutes. The electron beam will
deposit amorphous carbon material, which tends to be ubiquitous
even in high vacuum environments in the form of hydrocarbons.
The rate of EBID deposition will depend on the hydro-carbon
partial pressure in your vacuum system. In particularly clean
systems, the rate of material deposition may be low. This
material will bind the nanotube to the probe. Retract the
probe as described above once sufficient material has been
deposited (Figures 9-12).
Figure 9: Apply EBID.
Figure 10: Begin pulling the nanotube
from the surface.
Figure 11: The nanotube loosened and
beginning to come out.
Figure 12: The nanotube free from the
surface and attached to the probe.
Towards Characterization
The S100 can perform electrical and mechanical measurements
on nanostructures. The suite of Zyvex NanoEffectors™
includes probes, MEMS grippers, and AFM cantilevers. Electrical
and mechanical measurements can be carried out by utilizing
the unique functional characteristics of each.
The electrical properties of multiwalled carbon nanotubes
have been studied. Current experimental procedures for measuring
the electrical properties utilize lithography processes that
not only require time and a fabrication facility but also
run the risk of altering the very physical properties that
need to be measured13. The S100 can be used to
make electrical measurements without the complexity of standard
experimental techniques.
Electrical measurements can be accomplished using two or
four probes, depending on user preference. Figure
13 shows a nanotube in a four-probe configuration.
Four-probe electrical measurements are made to compensate
for contact resistance between the forcing probes14.
Figure 13: Four-probe configuration
for electrical characterization of a multiwalled carbon nanotube.
A current is supplied between the outermost probes. The voltage
is then measured between the innermost probe. The IV response
depends on the type of nanowire. If the wire is metallic,
then the conductance is ascertained by Ohm’s law, and
a plot of the current versus resistance will be linear.
Multiwalled nanotubes however, are complex structures due
to the layering of metal and semiconducting tubes. The contribution
from inner layers of the multiwalled nanotube can be gauged
by successively removing the outer layers13,15.
CNTs have quantized conductance based on the electrode position
on the nanotube16. The S100 can position probes
at different points along the nanotube to observe the quantized
conductance of the nanotube. Another interesting characterization
metric is the field-induced electrical response of nanostructures.
The S100 can apply an external field to a nanowire that is
suspended between two probes. The induced current can be measured
with a low-noise ammeter.
Mechanical measurements are made by using cantilevered probes.
The nanotube is suspended between the two cantilevered probes,
which have disparate moduli. Applying tension to the nanotube
by displacing the probes with the S100 positioners causes
the weaker cantilever to bend. This bend magnitude can be
measured in the SEM and a force can be calculated, which translates
to the modulus of the nanotube. Also, the breaking strain
and compression strain can be calculated this way9,17.
A compression strain can be ascertained by bringing the cantilevers
together with the fine positioners. A rigid nanowire will
cause the cantilevers to bend until the force on the nanowire
causes it to buckle. This force is the compressive force and
can be determined by measuring the displacement of the cantilevers
as the load is applied. This process will not work for nanowires
and nanotubes that are very flexible. In fact, long carbon
nanotubes will bend before buckling. Rigid nanostructures
such as silicon or boron nitride nanowires are better candidates
for AFM cantilever based compression testing.
MEMS-based grippers can be used to characterize the force-deflection
characteristics of nanotubes18. Mechanical characterization
of nanotubes with MEMS devices has been performed using complex,
AFM-based manipulation systems that require long experiment
times and costly equipment19. The S100 can position
nanotubes between silicon-based, thermally-actuated beams.
The beams, or flexures, can be calibrated for moduli in the
SEM, and are actuated with a DC voltage source. When the nanotube
is positioned between the beams and the beams are displaced,
the modulus of the nanotube can be calculated. This is advantageous
over other methods since the calculations for the modulus
are not solely dependent upon observations of the cantilever
displacement ascertained by visual feedback from the SEM.
The silicon flexures can be calibrated for displacement per
unit voltage, and the supplied voltage measured during the
experiment translates into the force applied to the nanotube.
The mechanical properties of the nanotube can also be ascertained
from its parametric resonance frequency. A nanotube will oscillate
at a particular frequency and amplitude depending on its length
and diameter. This resonance frequency translates to a Young’s
modulus value20. Once a nanotube is affixed to
a probe, it can be stimulated for resonance using a second
S100 probe (Figure 14). The two probes are
connected by an alternating current source, and the nanotube
feels the effects from the alternating field.
Figure 14: A boron nanowire in resonance.
The nanowire is affixed to the top probe and is being excited
by field effect from the probe on the right.
Conclusions
The Zyvex platform of nanomanipulator tools was designed
to facilitate the physical property characterization of nanoscale
structures. The procedure for making contact with a nanotube
is straightforward and makes use of the most fundamental capabilities
of the S100. Nanotube and nanowire research has long been
stifled by the inability of scientists to isolate individual
nanostructures. The Zyvex S100 Nanomanipulator simplifies
the task by enabling picking, placing, stretching, bending,
and stimulating nanotubes with relative ease. The procedures
developed are directly applicable for manipulating and characterizing
nanostructures of different materials such as germanium or
boron. Interfacing to test bench hardware is seamlessly achieved
through an easily accessible BNC patch panel. The multiple
probe configurations facilitate two, three, and four probe
electrical measurements on nanostructures, nanomaterials,
and nanodevices. Coupled with Zyvex Corporation’s NanoEffector™
suite of tools, the S100 is the most versatile instrument
for novel materials characterization and nanoscale device
engineering on the market.
References
1. A. Nakamura, et al., Nature Materials 2 453 (2003)
2. M.S. Fuhrer, et al., Science 288 494 (2000)
3. A. Star, et al., Nano Letters 3 459 (2003)
4. A.M. Fennimore, et al., Nature 424 408 (2003)
5. V.M. Harik, Comput. Mater. Sci. 24 328 (2002)
6. R. Saito, et al., Appl. Phys. Lett. 60 2204 (1992)
7. A.M. Fennimore, et al., Nature 424 408 (2003)
8. Y. Sato, et al., J. Phys. Chem. B 105 3387 (2001)
9. R. Murphy, et al., J. Phys. Chem. B 106 3087 (2002)
10. M-F. Yu, et al., Science 287 637 (2000)
11. H.W.P. Koops, et al., J. Vac. Sci. Technol. B
6 477 (1988)
12. H. Dai, et al., Nature 384 147 (1996)
13. P.G. Collins, et al., App. Phys. A, 74 329 (2002)
14. J. Hochberg, et al., http://www.zyvex.com/Products/
KZ00_001a.html
15. J. Cummings, et al., Nature 406 586 (2000)
16. S. Frank, et al., Science 280 1744 (1998)
17. X. Chen, et al., Nano Letters 3 1299 (2003)
18. A. Hartman, http://www.zyvex.com/Prodcuts/MMIC_001a.htm.
19. P.A. Williams, et al., Appl. Phys. Lett. 80 2574
(2002)
20. M-F. Yu, et al., Phys. Rev. B 66 073406 (2002)
back to top
| ©
Copyright 2008, Zyvex Instruments. All Rights Reserved. |
|
