CHAPTER I BACKGROUND AND LITERATURE REVIEW 1.1 Smart Materials Smart materials are different from every day materials due to the flexibility in altering its properties which can be significantly altered in a controlled fashion by external stimuli such as stress, temperature, electricfield, and magnetic field. Representation of the smart materials based on the response to the stimuli is shown in Table I [1]. Table I. Representation of various smart materials based on response/stimuli 1 Electrical Magnetic Optical Mechanical Piezoelectric Electrostrictive Magnetostrictive Magneto chromic Negative Poisson Ratio Mechanical Optical Mechanical Thermo Chromic Shape Memory Thermal Electrical Optical Photoconductor Photo Chromic Optical Optical Mechanical Magneto-Optic Magneto-Rheological Magnetic Optical Mechanical Electro Chromic, Electro Optic, Electro Luminescent Piezoelectric, Electrostrictive Electro-Rheological Electrical Stimuli Property Response
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
CHAPTER I
BACKGROUND AND LITERATURE REVIEW
1.1 Smart Materials
Smart materials are different from every day materials due to the flexibility in altering
its properties which can be significantly altered in a controlled fashion by external stimuli
such as stress, temperature, electricfield, and magnetic field. Representation of the smart
materials based on the response to the stimuli is shown in Table I [1].
Table I. Representation of various smart materials based on response/stimuli
Figure 9. Comparison of the responses of PZT, FNP coated PZT and PZT treated under the magnetic field.
1.3.3 Advantages of Mixing ZnO Nanoparticles with the Ferrofluid
Ferrofluids when subjected to saturation magnetic field, whereby the applied
magnetic field overcomes the surface energy of the ferrofluid and forms ferrous cones.
These ferrous cones are used as low electron source for charge injection devices [23].
The arrays of pointed structure or the cones as shown in the Figure 10 have various
mechanical and electrical properties.
PZT with FNP Treatment
0
50
100
150
200
250
0 5 10 15 20 25
Current (Amps) (Amps)
R E S P O N S E (
mV
)
PZT under magnet
FNP coated PZT
PZT
17
Figure 10. Ferrous cones on the aluminum substrate.
Mechanical and electrical properties could be improved by coating these ferrous
cones with a wide band gap semiconductors like ZnO nanoparticles. Ferrous cones could
be particularly useful for creating emitters to be coated by wide band gap
semiconductors, which can absorb and emit photons in the ultraviolet light bands.
Coating the ferrous cones with the ZnO nanoparticles will be an alternative for the
photocathode systems and source of electrons. This technology has various advantages
as described below [23]:
a) ZnO avoids the oxidizing properties associated with ferrous cones and can be
used as photocathode systems.
b) ZnO in ferrous cones are expected to impact a wide range of electronics and
mechanical equipment such as field emission devices, photocathodes, scanning
18
tunneling microscopes, atomic force microscopes and low-power propulsion
systems.
c) This method is low cost, versatile and consistent.
The challenges are:
How to coat ferrous cones with a semiconductor like ZnO for the above
described advantages?
FNP coating on PZT has increased sensitivity [20]. How will the
sensitivity of the PZT be changed by coating with a FNP and ZnO
mixture?
1.4 Importance of Tribological Analysis
Though PZT has various applications, literature on its tribological properties are
scarce. To enhance the PZT applications in the field of MEMS it is essential to study the
interfacial engineering, plastic deformation, and fracture characteristics to design smart
structures with the ability to resist failure under mechanical and thermal loads. The large
surface to volume ratios and low restoring forces, unwanted adhesion and friction
dominate the performance of the MEMS. The macroscopic contacts of MEMS are
composed of nanoscale asperities. A nanoscale asperity acts as the source of fracture or
cracks due to the high friction at the asperities contact. The characteristics of surface
morphology and properties of single asperity contacts can be studied by atomic force
microscopy (AFM). Based on the above facts, minimizing the effect of adhesion,
friction, and wear is essential for improving MEMS.
Surface morphology of PZT used in the D-RAM applications is correlated to
19
fractal morphology and to evaluate efficiency of the D-RAM component [25]. PZT
domain stability, mobility and structure are a function of friction due to the fluctuating
stresses. Low friction leads to low mobility and decreases the domain walls. Change in
the sensitivity of PZT is a function of concentration of point defects, interaction between
defects and change in the real structure of the PZT. Presence of these defects is going to
affect mobility and stability of domain walls. Predicting the surface defects, interfacial
friction and energy and early elastic deformation with respect to the ramped load at the
PZT surface is essential to define its tribological characteristics [26].
1.5 Justification and Importance of This Work
Improving the piezoelectric property of the PZT is a major concern to expand the
range of applications. FNP coating on PZT is going to provide the additional magnetic
energy and improve the magnetoelectric effect of the PZT and in turn increase the
response of the PZT [21]. The effect of mixing ZnO nanoparticles in a ferrofluid is
studied by coating this mixture on the PZT and subjecting it to power harvesting.
Comparison of the power harvesting capability of plain PZT, FNP coated PZT, and (ZnO
nanoparticle + FNP) mixture coating on the PZT was carried out.
The importance of studying the atomic friction and interfacial atomic defects on
the PZT is understood to an extent, and it is a challenge for nanotribologists. There is a
necessity for development of a high spatial resolution tool to analyse these challenges.
Tribo-diagnostic tool is developed based on the phase fluctuation based mathematics to
analyse the atomic friction and to identify the atomic defects at the sliding interface. This
tool is effectively used on the silver coated PZT by considering its applications to
20
MEMS.
21
CHAPTER II
THEORY
2.1 Tribology
Tribology relates to the study of friction, wear and lubrication at the interface of
two or more surfaces in contact. Tribology is a vast subject and it is important to prevent
failures at the contact interface due to sliding, rolling, or rubbing. In 1966, the lack of
knowledge towards tribology costs 6% of the GDP to the US government, i. e. $200
million in the year 1966. One-third of the world’s energy resources used are subjected to
friction in one form or the other [27]. Research in tribology has gained importance after
these statistics were realized and extensive research on preventing the loss due to friction
and wear at the contact of the rubbing surfaces is ongoing today.
The better performance of the MEMS is possible through the study of tribology at
the microscale or nanoscale leading to a new subject known as nano/micro tribology.
The following Sections deal with the tribology at nanoscale.
2.1.1. Nanoscale Friction
Friction is a part of tribology and is defined as the force opposing the motion of
rubbing surface. There are various laws that define the factors affecting friction. In
1699, Amontons law was stated as, “friction forces are proportional to the load applied
and independent of the surface area”. After some years, Hirn [28] distinguished the
22
friction variation on the unlubricated and lubricated surface. The coefficient of friction
for the lubricated surface was reduced as compared to the unlubricated solid surface.
Hence, Hirn added the effects of velocity, load and surface area on friction. In 1892,
Ewing related the origin of the friction to the surface forces and defined friction as the
opposing molecular force for the applied load, eventually leading to the molecular
displacement.
Tomlinson’s theory correlates the interaction between the two sliding elements at
the molecular level. According to Tomlinson, the energy is dissipated as the molecules
are forced into each other atomic field. This is the theory widely used to analyse the
friction at the nanoscale or atomic scale [28].
Friction at nanoscale does not obey Amonton’s law of friction. Friction at
nanoscale is proportional to the contact area, load applied, contact pressure and surface
forces at the sliding interface [27]. In many cases, simple empirical formulas used for the
describing friction at macroscale is not applied to the nanoscale because of the high
surface area to volume ratio, surface chemistry, adhesion and the roughness effects.
Hence, the friction force varies nonlinearly to the load applied and the models used to
analyze it are explained in the Section 4.1.1 [29]. For example, Figure 11 represents
friction force variation at the nanoscale over an increased load on Carbon-60 (C 60) and
Germanium Sulphide (GeS) surface. The nonlinear response of the friction vs. load is due
to the surface force effect [30].
Corrugations seen on the friction forces which are distinctively different in shape
is called as stick-slip friction. Sliding action of the surface asperities undergoes sudden
slips during sliding motion [31]. The friction due to stick-slip is studied to analyse the
23
friction at nanoscale.
Figure 11. Friction force vs. load [Si tip on the GeS and C60].
2.1.2 Stick-Slip
When the two surfaces slide on each other they can have an abrupt motion or an
irregular motion. This irregular motion is due to the asperities on sliding surface which
leads to stick-slip motion. The sliding process is not a continuous one; the motion
proceeds by jerks. The asperities “stick” together until, as a result of the gradually
increasing pull, there is a sudden break with a consequent very rapid “slip” [28].
Different types of stick-slip signals are described below:
Periodic Stick-Slip: The signal with potential (friction force) maxima and minima has
narrow widths that do not overlap.
Random Stick-Slip: The signal with friction force potential maxima and minima have
wide width and overlap.
Stepped Stick-Slip: There is a step increase in the friction force potential. The saw
toothed maxima, has increased force maximum than the force amplitude before.
24
Figure 12. Different types of stick-slip process.
Stick-slip friction provides information on the interface science in many ways. Some
of the theories on the stick-slip are described as below:
As the atoms in one plane are in equilibrium with the neighboring plane, when
the shear forces are applied at the interface, the atoms deform elastically
called as slip and the prior situation where atomic equilibrium existed is called
as stick. This slip may be related to loss of energy or fracture due to periodic
elastic deformation of the atomic planes [32].
Between the two sliding surfaces, asperities at the interface undergo elastic
deformation known stick” and the release of the energy leads to plastic
deformation known “slip”.
The stick-slip process is a melting-freezing process where the melting is due
to the dislocations of atoms and the freezing is related to the strain hardening
25
of the atoms [33]. In many cases the rapid slip of the upper surface relieves
some of the applied force which eventually falls below another critical value
Fk, the kinetic friction, at that point the film resolidifies and the whole stick-
slip cycle is repeated [27].
Stick-slip behavior varies with applied load, scan speed and scan direction
with respect to crystallographic directions [34].
2.1.3 Advantages and Challenges of Stick-Slip
Stick-slip study at the atomic scale helps in understanding the transitions
between different regimes of atomic scale friction and provides insight into
the origins of friction and may lead to the ability to control it [36].
Determination of variation in the amplitude of the stick-slip is a guide to
predict the intensity of the friction force. Reducing the force amplitude is
going to reduce the friction at the interface.
Understand the transition of stick-slip to continuous sliding at the atomic
friction [37].
Could the extensive study of stick-slip explain the relation between friction
and wear [29]. ?
Slip due to the atomic dislocation could be also due to the atomic defects;
hence, stick-slip could be used for imaging the atomic defects [38].
Some of these above stick-slip challenges are to be addressed to understand the
tribological behavior of the PZT. Further, these tribological characteristics could be used
for designing the PZT based MEMS for sliding applications and the relation between
26
PZT electromechanical efficiency and load acting on it. AFM is used for obtaining the
stick-slip signal at the sliding interface; AFM description is given in the following
Section.
2.1.4 Atomic Force Microscopy (AFM)
AFM is used for measuring the three-dimensional surface topography and
physical properties of a surface using a profiler or sharp tip probe. The probe is placed at
a point where the surface force effect exists. The probe is raster scanned over the surface
by keeping a constant force; this raster scan displays the surface topography. AFM
consists of a cantilever beam with a probe, laser beam, piezotube scanner and a laser
photodetector. The description of the AFM is shown in the Figure 13.
Figure 13. Schematics of AFM.
The probe is very sensitive to surface forces and as the probe scans over the
sample, the probe gets deflected due to the surface forces; the feed back control system is
27
used to keep the probe and sample distance constant. The deflection of the probe will
yield a surface image; AFM has been effectively used in the field of friction analysis at
the nanoscale [38].
Force calibration mode is used in studying the interface science [tip and the
sample surface interaction]. In force calibration mode X and Y voltages applied to the
piezotube is held at zero and a saw tooth voltage is applied to the Z-electrode of the piezo
tube. The force measurement starts with the sample far away and the cantilever tip at its
rest position. As a result of the applied voltage the sample is moved up and down and a
cantilever tip deflection signal is monitored using the photodiode. The force curve, a plot
of the cantilever tip deflection signal as a function of the voltage applied to the piezotube,
is obtained. Figure 14 shows the force separation curve; the arrow head reveals the
direction of the piezo tube travel. At point 1, the tip is off the sample surface. From
point 1 to 2 there is no change in the deflection signal as the piezo tube extends, because
the force is initially zero and the sample has not come into contact with the tip. At point
2, the force between the tip and sample becomes attractive. The attractive force increases
till the point 2’ of the sample-tip, until the sample-tip force becomes repulsive. The
maximum forward deflection of the cantilever multiplied by the spring constant of the
cantilever is the “pull-off” force. At point 3, the deflection signal reaches maximum and
the piezo tube starts to retract. At point 4 the cantilever is not deflected, but due to
adhesion between the tip and the sample. The tip sticks to the sample and the cantilever
is bent down as the piezo retracts. The attractive force between the tip and sample
becomes equal to the spring force at point 5. At point 6 the cantilever has returned to its
28
undeflected state. Cantilever deflection signal remains constant as the piezo continues to
retract to point 7.
Figure 14. Force curve.
AFM is used for measuring friction in the form of friction force microscopy. A 0o
scan angle corresponds to sample scanning in the “y” direction, and a 90o scanning angle
corresponds to the sample scanning perpendicular to the “y” axis in the xy-plane (along x
axis). This scan is called as perpendicular scan, where the scanning tip experiences only
frictional force.
As the tip is perpendicularly scanned over the sample, twisting of the tip takes
place. Twisting of the tip will deflect the laser beam on to the horizontal and vertical
quadrant of the photodetector. The differential signal between the left (L) and right (R)
photo detector is denoted as friction force signal and calculated from {(L-R)/ (L+R)}
[27].
29
High Speed Data Capture is a function in the AFM used for collecting the friction
signal data at an unprecedented time resolution. Friction signal data at an unprecedented
time resolution is going to provide high spatial resolution. This property is important in
analyzing the atomic scale stick-slip friction. Phase fluctuation based processor is used to
analyze this friction signal data collected at an unprecedented time resolution.
2.2 Phase Fluctuation Based Processor
Phase fluctuation based processing produces better gain over the amplitude based
analysis of the signal. Studying the amplitude variation of the acoustic signal to identify
the noise and signal is a traditional method of obtaining high signal-to-noise ratio.
Sometimes the noise also has the same fluctuation as that of the signal and it is difficult
to identify the difference between the signal and noise. Amplitude based signal-to-noise
ratio does not consider the unpredictable dynamic changes in the signal. High signal-to-
noise ratio is obtained by reading the variation of phase angle with respect to time, where
deviation from the aligned phase angle is called as noise. Phase fluctuation based signal
processing provides better spectral and spatial resolution [39].
30
2.2.1 Working Principle
The working principle of phase fluctuation based processor is explained via
Figure 15.
Figure 15. Polar representation of acoustic pressure wave vectors.
Phase fluctuation is defined on three planes for self correction of the phase rate and
direction of phase angle rotation. As phase angle approaches zero, the phase angle
temporal coherence increases with gain (signal-to-noise ratio). The mathematics used in
phase fluctuation is explained based on the Figure 15.
R i, R i-1, R i-2 , where
……………………………………. (2)
Let, ti = Discrete time sample
= Phase angle at a given frequency
= Angular velocity of the vectors
= Angular acceleration of the vectors
31
………………………………………… (3)
…………………………………………… (4)
= Aligned phase angle of the vectors
…………………………………………… (5)
For a coherent signal,
=Constant; = 0
If, 0, measure of incoherence or noise
The phase fluctuation ( ) on the polar plane is defined as the phase angle of
pressure vector is out of uniform phase angle rotation with phase angle and of
the previous two phasors and , as represented by the equation 2. Uniform rate of
phase fluctuation is termed as constant angular velocity; constant angular velocity gives
zero angular acceleration. Any deviation of the angular acceleration from the zero point
towards the positive or negative side gives the information on noise.
Twisting of the cantilever tip corresponds to the friction force experienced by it.
Uniform deflection of the cantilever tip is considered to have uniform phase fluctuation
and hence, called as signal or “stick” and the deviation from this uniformity is considered
to be the surface defect or “slip”.
High spectral resolution and friction signal-to-noise ratio is essential for analyzing
the stick-slip at nanoscale and will solve the challenges mentioned in stick-slip studies.
This phase fluctuation mathematics and HSDC function in the AFM together form the
tribo-diagnostic tool.
32
CHAPTER III
EXPERIMENTAL DETAILS
3.1 Tribo-Diagnostics of PZT
The AFM with HSDC function is used for collecting the friction signal data. This
friction signal data is analysed using the phase fluctuation based mathematics coded in
Matlab. The AFM calibration and HSDC methods followed in this research for
developing the tribo-diagnostic tool are explained below.
3.1.1 AFM Calibration
Atomic Force Microscopy is used for the tribo diagnostics of the PZT smart
material. AFM supplied by the digital instruments with the nanoscope V software is used
for this experiment. Various AFM parameters used for this experiment are shown in the
Table V.
33
Table V. AFM Parameters for studying the friction force signal.
AFM Parameter’s Values
Scanning Mode Contact
Scan Area (nm2) 50 X 6. 2
Scan Rate ( Hz) 1. 99
Ramped Load (nN) 40 to 200
Scan Angle (deg) 90
Lines 256
Points/Line 256
Scanning Velocity (nm/s) 199
3.1.2 Experimental Procedures
High Speed Data Capture (HSDC)
High Speed Data Capture (HSDC) is used for obtaining unprecedented time
resolution of the interfacial friction signal. This is a data collection tool in the
nanoscope V software. HSDC can collect data at 6. 25 M Hz and 512 K Hz.
Considering high resolution requirements, the 6. 25 M Hz option was used for
collecting data [41]. The step-by-step procedure used to collect the friction signal
data is explained below.
PZT surface and silicon nitride [AFM cantilever Tip] is cleaned with
methanol for avoiding the effect of dirt on the interfacial friction.
34
The cleaned PZT surface is glued on the metallic disc and stuck on the
piezotube magnetic disc holder in AFM.
For contact mode the feed back control system with proportional gain = 3 and
integral gain = 2 is set.
The laser point is concentrated on the cantilever tip using upper and lower
knob in the AFM. In the meantime, look at the laser signal on the screen and
try to get the laser point below the center of the cross (to get ~-2 V vertical
deflection and ~0. 0 V horizontal deflection).
Nanoscope V software function “engage” is used to bring contact between the
silicon nitride tip and the silver coated PZT surface. Calculate the deflection
sensitivity using the ramp load function.
Based on the deflection sensitivity of cantilever, the load acting on the sample
is defined; the load ranged from 40 to 200 nN is applied.
At each load applied on the PZT surface, sliding of the silicon nitride
cantilever tip on the PZT surface leads to the interfacial friction.
The twisting of silicon nitride tip during scanning on the PZT surface
generates a differential signal [lateral force deflection] from the left and right
photodiodes, which provide the friction force signal.
HSDC is used to collect this data at a time resolution of 6. 25 MHz.
The friction signal data in the form of ASCII format is saved in the computer
for exporting it to Matlab for further analysis.
Phase Fluctuation Analysis: The importance and theory of the phase
fluctuation based analysis was explained in the Section 2. 2. Phase fluctuation
35
analysis mathematics was coded in Matlab. The interfacial friction signal data
collected at each load ranged from 40 to 200 nN was exported in ASCII
format to Matlab.
The Section below explains the experimental setup for nanoparticle coating to
improve the sensitivity of PZT and for power harvesting.
3.2 Coating Process
This Section explains the experimental set up for nanoparticle coating to improve
the sensitivity of PZT .
3.2.1 Coating Equipment and Procedure:
Coating Materials:
1. Piezofilm (P-5E) with resonance frequency of 1.3 kHz, impedance of 300
ohms, and capacitance of 40 nF at 120 Hz was supplied by Murata Inc. P-
5E thin sheets were used for the power harvesting experiment. The
properties of P-5E piezofilm are given in Table VI [2]. The SEM image of
the P-5E with silver coating is shown in the Figure 16.
Figure 16. SEM image of the silver coated PZT (P-5E).
36
Table VI. Property of the PZT [P-5E].
Relative
dielectric
constant
(ε33 )
Electro-
mechanical
coupling
(K33 )
Piezoelectric
constant(d33)
(10-12 m/V)
Elastic
constant
(10-12
m2/V)
Density
(103Kg/m3)
Bending
strength
(106 N/m2)
1510 62% 271 14. 3 7. 8 11. 3
2) Ferrofluid is a stable colloidal suspension of magnetite
nanoparticles of size 10 nm and coated with the surfactants like
oleic acid to prevent the agglomeration of the magnetite
nanoparticle. The ferrofluid was supplied by the Edmund
Scientific Inc; Table VII contains the properties of ferrofluid
[18].
37
Table VII. Ferrofluid properties as provided by Edmund Scientifics.
Edmund
Scientific
Saturation
magnetization Density Viscosity
Surface
tension Volatility
Ferrofluid 400 Gauss 1.21gm/ml 6cp @
270C
29dynes/cm
9%
(1hr 5000C)
3) Zinc Oxide: Zinc oxide nanoparticles were supplied by the Sigma-Aldrich
[42]. The properties of the ZnO nanoparticles are short listed in Table VIII.
Table VIII. Properties of ZnO nanoparticles.
Material Particle Size Surface Area
ZnO <100nm 15-25 m2/g
4) Ultraviolet (UV) light of wavelength 130 nm was used for curing the
coated PZT for 80 Hrs.
Experimental Procedure:
a) Aluminum Strip: Aluminum alloy 6061 is used for the experiment. It is
usually used in aircraft and automobile manufacturing. An aluminum
substrate of dimension (11X2X. 2) in3 is used as a substrate for the silver
38
coated PZT (P-5E) and nanoparticle coated PZT (P-5E) in power
harvesting.
b) The surface of the aluminum strip is cleaned with alcohol and PZT is
glued on the surface. Non-conductive epoxy resin is used as the glue.
The PZT glued aluminum is shown in the Figure 17.
Figure 17. PZT glued on the aluminum strip.
c) The nanoparticle coating equipment with the UV light, rectangular box
and magnet glued wooden block is shown in the Figure 18.
39
Figure 18. Schematic of the nanoparticle coating equipment.
d) A transparent rectangular box is fabricated using the glass and plastic
material as shown in Figure 19. At the base, a slot of size 12X2 in 2 is
machined to accommodate the PZT glued aluminum strip. The PZT
glued aluminum strip is drenched with the ferrofluid and the ferrofluid
mixed with ZnO.
40
UV Light
Magnet glued wooden board
Rectangular box with an aluminum strip
Figure 19. Rectangular box with the slot to hold the PZT glued aluminum strip.
e) Disc magnets of strength 400 gauss are glued on the small wooden plate
as shown in the Figure 20. Two wooden plates with south and north poles
of the magnets are glued. These magnet glued wooden plates are placed
with the north Pole disc magnets on the top and the south pole disc
magnets below the rectangular box. Hence, the magnetic field passes
through the aluminum strip in the rectangular box.
41
Figure 20. Magnet glued wooden piece to create the magnetic field on the coating surface.
f) ZnO nanopowder is mixed thoroughly in the ferrofluid using a vibrating
mixer. ZnO nanopowder of 6% of the volume of ferrofluid was mixed.
g) The PZT glued aluminum strip is pushed into the rectangle box slot.
h) For FNP coating, the ferrofluid is poured on the PZT glued aluminum
strip in the slot, and the ZnO nanoparticles mixed in ferrofluid solution
are used for ZnO nanoparticle coating.
i) Disc magnet glued wood pieces are placed on the rectangular box. The
magnetic lines of flux pass through the coated PZT glued aluminum, from
the south pole to north pole.
j) Ultraviolet (UV) curing of the nanoparticle coated PZT glued aluminum
is carried out for 80 hrs [20].
k) After 80 hrs of UV curing, the nanoparticle coating is polymerized onto
the PZT material.
42
l) Three each of PZT glued aluminum strips, FNP coated PZT glued
aluminum strips and FNP+ZnO coated PZT glued aluminum strips were
prepared.
The nanoparticle coated PZT glued aluminum plates are shown in Figure 21.
Sometimes the base of PZT glued aluminum strip was coated, to avoid the effect of
coating on the PZT base in the power harvesting, the coating at the base is scraped using
a scrapper.
a)
b)
Figure 21. a) FNP coated PZT glued aluminum strip; b) FNP+ZnO coated PZT glued aluminum strip.
43
3.3 Power Harvesting
This Section presents the equipment and the experimental set up for implementing
the above coated PZT and plain PZT for power harvesting.
3.3.1 Equipment and Experimental Setup
Equipment used for the power harvesting experiment are as listed below:
a) Exciter
b) Function Generator
c) Oscilloscope
a) Exciter: It is an oscillator which oscillates at the input driving frequency.
Exciter has a dc generator fixed to the output shaft. The AC-voltage from the
function generator acts as an input voltage to the exciter, which is then
converted to the oscillation of the shaft through the DC generator. This is
used for oscillating the aluminum plate with either PZT or nanoparticle coated
PZT.
b) Function Generator: It is used for the generating the sinusoidal signal at the
required frequency. It is used for exciting the exciter at the required
frequency.
c) Oscilloscope: It is used to analyze the voltage generated during the power
harvesting experiment.
44
The experimental setup used for the power harvesting experiment is
shown in Figure 22. The step-by-step procedure for the power harvesting setup is
explained below.
Figure 22. Schematic representation of the power harvesting setup.
1) The tip of the PZT glued aluminum plate is fixed on the exciter [basically
a cantilever support].
2) Exciter is connected to the function generator through connecting the
cables.
3) Alligator clips are used as connectors to determine the voltage response of
the PZT glued aluminum plate; they are connected to the oscilloscope.
Exciter
Alligator clip
Oscilloscope
Aluminum strip
Function generator
45
3.3.2 Experimental Procedure
The function generator is used for generating the sinusoidal signal at a frequency
of 90 Hz [kept constant]. As the PZT glued aluminum strip is excited at 90 Hz, due to
piezoelectric property of the PZT, a voltage is generated which is then recorded with the
oscilloscope. Similarly, the FNP coated PZT and FNP+ZnO coated PZT are used to
analyze the importance of nanoparticle coated PZT over plain PZT as related to power
harvesting.
46
.
Chapter IV
RESULTS and DISCUSSION
4.1 Tribo-Diagnostics
Based on the experimental procedure described in Section 3.1, the friction signal
data was collected and analyzed using the tribo-diagnostic tool. The tribo-diagnostic tool
possesses high spatial resolution, spectral representation of the friction signal, and pattern
representation of the stick-slip friction. The results of the tribo-diagnostic tool on the
silver coated PZT surface are described in the following Sections.
4.1.1 Effect of Friction (Atomic Scale) on Increasing Load
Friction force at nanoscale is calculated based on the surface forces during the
sliding of cantilever tip on the silver coated PZT (P-5E) surface. There are two models
used for calculating the nanofriction, the JKR model [Johnson, Kendall and Roberts] and
DMT model [Derjaguin, Muller and Toporov] [45].
The JKR model is appropriate for strongly adhering materials with short range
surface forces. The DMT model is opposite to the JKR model. The DMT model is used
for stiff, weakly adhering materials with long range forces. Since the PZT surface is stiff
and has long range forces, the DMT model is used to predict the nanofriction at the PZT
interface [44].
47
DMT Model [44], [45]:
F= Friction Force, nN
E= Reduced young’s modulus of the tip and PZT interface, Gpa
A=Contact area, nm2
L=Load applied on the PZT surface, nN
= Adhesive force, nN
LC= Pull-off force, nN
= Interfacial shear strength, Gpa
R= Tip radius =0.002 nm
= Shear strength of the PZT (P-5E) material, Gpa
= Shear strength of the silicon nitride tip, Gpa
Friction force at nanoscale is calculated as shown in the equation (6).
………………………………… (6)
Interfacial shear strength is given in the equation 7
……………………………… (7)
Effective shear strength at the interface is given in equation 8
………………………. (8)
Contact area and pull off force is given in the equation 9 and 10
………………………. (9)
48
……………………………. . (10)
Friction force is calculated based on the following data, under the load (L) of 40nN.
=52.749Gpa
=2.110Gpa……………………………… (11)
…………………………… (12)
=1.44*10-17m2……………………… (13)
From (11) and (13), in equation (6)
F=0.3 nN [at a load of 40 nN)
The above mathematics is used for calculating the friction force at each load
(80,120,160 and 200nN) and it is depicted in Table VIII and as shown in Figure 23.
49
Table VIII. Load vs. Friction Force (nN).
Load, L (nN) Friction Force, F (nN)
40 0.3
80 0.48
120 0.63
160 0.77
200 0.91
Figure 23. Load vs. friction force (nN).
50
The interfacial friction force as a function of the load with the topographic image
after each load is represented in the Figure 23. The friction force in the region of 40 to
120 nN is non-linear relates to the surface force effects at low load or high friction at low
load. Beyond 120 nN the friction force is linear till 200 nN.
4.1.2 Specgram and Fast Fourier Transform (FFT) of Friction Signal
The friction signal data [43122 data points] from the AFM was collected in 2.47 s
at a sampling frequency of 17.4 KHz. Each sond of the force signal data is magnified to
17 sec, and totally 42 sec of data is plotted for 2.47 sec. Hence, 58 ms of friction signal
data is analyzed as one second in the specgram. Each 58 ms friction signal data is
divided into 1024 points for recognizing the friction signal temporal coherence.
At 80 nN Figure 24 (a) shows the friction signal in the random stick-slip form
obtained directly by the AFM. This friction signal is represented in the form of a
specgram as shown in Figure 24(b). The magnitude of the cantilever twisting is
represented on the “y- axis” and time (s) on the “x-axis”. The dark red region in the
specgram represents a high friction region and its FFT shows high amplitude. High
amplitude relates to high intensity of twisting of the silicon nitride cantilever tip.
51
Figure 24. a) Friction force signal collected at a time resolution of 6. 25 MHz for a load of 80nN. b) Specgram and FFT of the friction signal at 12-23 Hz for a load of 80nN.
(a) (b)
a) Load: 40nN
52
b) Load: 120nN
c) Load: 160nN
53
Figure 25. a), b) c) and d) Friction signal, specgram and FFT of the friction signal , with an increasing load 40, 120, 160 and 200 nN respectively.
Spectral representation of the friction signal shows the region of high and low
friction. It is hard to differentiate the amplitude of the friction signal variation over the
load ranging from 40 to 200 nN due to lack of spatial resolution. The FFT of this friction
signal as represented above is windowed using the Hilbert transform. This windowed
signal is used for phase fluctuation analysis to obtain better spatial resolution. The above
results are an approach towards better spatial resolution of the friction signal.
4.1.3 Effect of Stick-Slip on Increasing Load
High resolution pattern representation of the silicon nitride tip scanning on the
PZT (P-5E) surface represents the stick-slip behavior of the interfacial friction. Tip
velocity is 199 nm/s; the distance traveled in 58 ms (1024 points) is 11.61 nm. Hence,
100 points represents 11.3 Å. This is represented in the stick-slip chart shown in Figure
d) Load: 200nN
54
26. As the tip is scanned over the PZT surface, slip/atomic displacement or acoustic noise
(due to atomic displacement) is observed. This is represented as a deviation of the
aligned phase fluctuation from zero or the deviation from the uniformity of tip twisting
during the tip sliding on the PZT surface. The region before this fist slip is considered as
the elastic deformation regime. The downward arrows represent the qualitative relation
between the slip and zero magnitude. The pattern as shown in the Figure 26 is a precise
observation of the positional behavior of the tip on the PZT.
Figure 26. The phase and magnitude representation of stick-slip and elastic regime before first slip at a load of 40 nN.
Elastic Regime before first slip
Slip/Atomic Defect
11. 3Å
55
a)
b)
43Å
51Å
56
c)
d)
Figure 27. a), b), c) and d): Represents the stick-slip variation and the elastic regime before first slip due to the increase in load from 80 to 200 nN respectively.
108Å
188Å
57
The pattern described in this Section shows the time and length of the first slip.
There is an increase in the early elastic regime due to increased load ranging from the 40
to 200 nN. The relationship between the elastic deformation regimes with respect to
increasing load is represented in the Section 4.1.5.
4.1.4 Hilbert Transform
The friction signal data was obtained from the AFM at a sampling frequency of
17.4 KHz. Windowing of this friction signal data is carried out using Hilbert transform at
a frequency range of 12-23 Hz. Frequency of the friction signal data is varied as a
function of the phase of the signal and provides high spectral representation. Windowed
friction signal data represents the continuous sliding and random stick-slip as represented
in the Figure 28. For 58 ms (11.61 nm) of friction signal data, the tip has scanned a
length of 11.61 nm on the silver coated PZT. In this 11.61 nm of distance covered; 11 13
Å of the sliding has continuous sliding and 91 Å of sliding has random stick-slip behavior
as shown in Figure 28. Continuous sliding and random stick-slip represents the region of
low and high friction respectively.
58
Figure 28. Continuous sliding and random stick-slip at a load of 40 nN and a band pass frequency of 12-23 Hz.
4.1.5 Early Elastic Deformation Regime at Atomic Scale
The length of early elastic deformation regime as shown in Figure 26 and Figure 27 is
calculated and plotted with the increasing load. Cubic relation between the elastic
deformation regime and load due to curve fitting is shown in the Figure 28.
If y = early elastic regime in Å
X = load in nN
y = 6.4e-005*x3 - 0.017*x2 + 2*x - 47
Stick
Slip
Continuous Sliding Random Stick-Slip
59
Figure 28. Elastic regime vs. load.
4.1.6 Effect of Interfacial Energy on Increasing Load
During scanning, magnitude of the dissipated or interfacial energy (ET) and the
energy for plastic deformation (Ep) is calculated by the equation (14. a.) and (14. b.)
= Tangential force applied on the surface, nN
L= Length of scratch, nm = 50nm
N= Number of scratches= 1024 lines
H= Hardness of the PZT surface, Gpa
Vwear= Volume of the wear, nm3
…………………………… (14. a. )
…………………………… (14. b. )
The linear relationship between the interfacial energy and load is given in the
Figure 29.
60
Table IX. Variation of interfacial energy with respect to increasing load.
Load( ), nN Interfacial Energy( ),
40 2. 048
80 4. 096
120 6. 144
160 8. 192
200 10. 24
Figure 29. Load (nN) vs. interfacial energy (10-12 J).
If, y= interfacial energy (10-12 J) and x= load in nN
From Figure 29,
y = 0. 051*x - 1. 5e-015
61
4.2 Coating
The methods of FNP coating, FNP+ZnO mixing and coating was described
earlier. Results on the geometric structure of the FNP cones and verification of the ZnO
nanoparticles in the FNP cones are described below.
4.2.1 Pattern of Ferrous Cones at Macro Scale
The magnetic field strength is higher than surface energy of the ferrofluid and
hence, the ferrofluid with ferrous nanoparticles form cones along the applied magnetic
field. The ferrous cones are at equidistance and form an array as shown in Figures 30 and
32.
Figure 30. Ferrous cones grown on a glass slide.
62
Figure 31. Array of ferrous cones.
The array of ferrous cones at the macro scale is the magnified image of the
ferrous cones at nanoscale. These ferrous cones act as a low source of electron and
contribute to the flow of charges on the PZT surface. The significance of this equidistant
array in power harvesting has yet to be determined.
4.2.2 Presence of ZnO in the Ferrous Cones.
In Section 3.2.1 the method used in mixing of the ZnO in ferrofluid is explained.
Fluroscene microscope (FM) is used to verify the presence of ZnO nanoparticles in the
ferrous cones. The FNP cones and (FNP+ZnO) cones were exposed to ultraviolet rays
for 20 min. The ferrous nanoparticle cones with the ZnO nanoparticles are shown in the
Figure 32; the tip of the ferrous cones acts as the region of charge concentration and a
low source of electrons. Figure 33 shows the emission of light from the (FNP+ZnO)
cones due to the distribution of the ZnO nanoparticles on the surface. FNP cones after 20
63
min of UV rays exposure, produced a dark picture. Qualitative analysis showed presence
of the ZnO nanoparticles in FNP cones due to the mixing procedure followed in Section
3.2.1.
Figure 32. Array of ferrous nanoparticle and ZnO nanoparticle mixed cones.
Figure 33. Shows the light emitted by the (FNP+ZnO) cones exposed to an UV light of wavelength 130 nm.
64
The intention of the ZnO coating on the FNP was to produce more charges due to
the piezoelectric property and thus control the flow of charges in the FNP cones due to its
semiconducting ability.
4.3 Power Harvesting
The above Section explains the advantages of the FNP coating and ZnO+FNP
coating in the field of charge conduction. In this Section we analyze the advantages of
these coating on the PZT in the field of power harvesting where the magnetic and
piezoelectric property is combined for increasing the electromechanical efficiency of the
PZT.
4.3.1 Effect of PZT
When the PZT glued aluminum was excited at 90 Hz, the voltage generated was
51.1 mV. The magnitude of the voltage at a frequency 90 Hz is shown in Figure 34.
65
Figure 34. The magnitude of the voltage (51. 1 mV) at an excitation frequency of 90 Hz for a plain PZT.
4. 3. 2 Effect of the FNP Coated PZT
FNP coated PZT glued aluminum was excited at 90 Hz and the voltage generated
was 115.4341 mV. This is due to the additional electron source on the surface and may
be due to the magnetoelectric effect. The magnitude of the voltage at a frequency 90 Hz
is shown in Figure 35.
66
Figure 35. The magnitude of the voltage (115.1 mV) at an excitation frequency of 90 Hz for a FNP coated PZT.
4.3.3 Effect of the (FNP+ZnO) Coated PZT
When the (FNP+ZnO) coated PZT glued aluminum was excited at 90 Hz, the
voltage generated was 366.8 mV. This is due to the additional piezoelectric effect on the
PZT surface by the ZnO in the ferrofluid coating. ZnO also acts as an additional source
of electrons, due to its ability to emit charges at room temperature. The magnitude of the
voltage at a frequency 90 Hz is shown in Figure 36.
67
Figure 36. Represents the magnitude of the voltage (366.8 mV) at an excitation frequency of 90 Hz for a FNP +ZnO coated PZT.
All the above results are plotted in the Figure 37. It shows the increase in voltage
response of the PZT due to the nanoparticle coatings. This increase is due to the
magnetoelectric effect or increase in the capacitance of the silver coated PZT (P-5E) due