Embed Size (px)
Citation preview
Purdue UniversityPurdue e-Pubs
Open Access Theses Theses and Dissertations
January 2016
INVESTIGATION OF THE MECHANICALPROPERTIES OF NANOSTRUCTUREDMATERIAL DEPOSITED BY LASER-INDUCED CHEMICAL SOLUTIONDEPOSITION AND POLYMER NANO-COMPOSITESCheng PengPurdue University
Follow this and additional works at: https://docs.lib.purdue.edu/open_access_theses
This document has been made available through Purdue e-Pubs, a service of the Purdue University Libraries. Please contact [email protected] foradditional information.
Recommended CitationPeng, Cheng, "INVESTIGATION OF THE MECHANICAL PROPERTIES OF NANOSTRUCTURED MATERIAL DEPOSITEDBY LASER-INDUCED CHEMICAL SOLUTION DEPOSITION AND POLYMER NANO-COMPOSITES" (2016). Open AccessTheses. 1119.https://docs.lib.purdue.edu/open_access_theses/1119
Graduate School Form30 Updated
PURDUE UNIVERSITYGRADUATE SCHOOL
Thesis/Dissertation Acceptance
This is to certify that the thesis/dissertation prepared
By
Entitled
For the degree of
Is approved by the final examining committee:
To the best of my knowledge and as understood by the student in the Thesis/Dissertation Agreement, Publication Delay, and Certification Disclaimer (Graduate School Form 32), this thesis/dissertation adheres to the provisions of Purdue University’s “Policy of Integrity in Research” and the use of copyright material.
Approved by Major Professor(s):
Approved by:Head of the Departmental Graduate Program Date
Cheng Peng
INVESTIGATION OF THE MECHANICAL PROPERTIES OF NANOSTRUCTURED MATERIAL DEPOSITED BYLASER-INDUCED CHEMICAL SOLUTION DEPOSITION AND POLYMER NANO-COMPOSITES
Master of Science in Industrial Engineering
C.Richard LiuChair
Joseph F. Pekny
Ramses V. Martinez
Wenzhuo Wu
C.Richard Liu
Abhijit Deshmukh 12/2/2016
i
i
INVESTIGATION OF THE MECHANICAL PROPERTIES OF NANOSTRUCTURED
MATERIAL DEPOSITED BY LASER-INDUCED CHEMICAL SOLUTION
DEPOSITION AND POLYMER NANO-COMPOSITES
A Thesis
Submitted to the Faculty
of
Purdue University
by
Cheng Peng
In Partial Fulfillment of the
Requirements for the Degree
of
Master of Science in Industrial Engineering
December 2016
Purdue University
West Lafayette, Indiana
ii
ii
ACKNOWLEDGEMENTS
I would like to take this opportunity to express my sincere gratitude to C. Richard Liu,
professor of Industrial Engineering, for providing me with all the support and valuable
guidance though this venture. I would also like to thank Prof. Ramses Martinez, Prof.
Joseph Pekny and Prof. Wenzhuo Wu for serving on my advisory committee and providing
their inputs. I am also extremely thankful to Rohit Voothaluru and Zhikun Liu, PhD
students from IE, for sharing their expertise and altruistic help during this research. Last I
gratefully acknowledge the support from National Science Foundation (NSF) Award
CMMI 1562960.
iii
iii
TABLE OF CONTENTS
Page
ABSTRACT ........................................................................................................................ v
CHAPTER 1. INTRODUCTION ..................................................................................... 1
1.1 Nanostructured Material Deposited by Laser-induced Chemical Deposition ............1
1.2 Polymer Nano-composites Synergized with Nanofillers ...........................................3
CHAPTER 2. MATERIALS, LASER-INDUCED CHEMICAL DEPOSITION
PROCESS AND EXPERIMENTS FOR THE DEPOSITED MATERIAL ....................... 5
2.1 Materials Selection .....................................................................................................5
2.2 Sample Preparation and Processing Method ..............................................................5
2.3 Material Characterization ...........................................................................................7
2.4 Mechanical Tests and Results ..................................................................................10
2.4.1 Micro-hardness Test ...................................................................................... 10
2.4.2 Nano-indentation Test ................................................................................... 11
2.4.3 Bending Test .................................................................................................. 23
2.4.4 Qualitative Adhesion Test ............................................................................. 29
2.4.5 Porosity Estimation ........................................................................................ 30
CHAPTER 3. MATERIALS, DISPERSION PROCESS AND EXPERIMENTS FOR
POLYMER NANO-COMPOSITES ................................................................................. 31
3.1 Materials Selection ...................................................................................................31
3.2 Sample Preparation and Processing Method ............................................................31
3.3 Mechanical Tests and Results ..................................................................................33
3.3.1 Static Tensile Test .......................................................................................... 33
3.3.2 Theoretical Analysis of Young’s modulus .................................................... 35
3.3.3 Dynamic tensile fatigue test ........................................................................... 38
iv
iv
Page
CHAPTER 4. DISCUSSION ......................................................................................... 40
4.1 Copper Nanoparticles Coating Deposited by Laser-induced Chemical Deposition 40
4.2 Epoxy Nano-composites Synergized with Carbon Nanofillers ...............................42
CHAPTER 5. CONCLUSIONS ..................................................................................... 45
REFERENCES ................................................................................................................. 47
v
v
ABSTRACT
Peng, Cheng. M.S.I.E., Purdue University, December 2016. Investigation of the
Mechanical Properties of Nanostructured Material Deposited by Laser-induced Chemical
Solution Deposition and Polymer Nano-composites. Major Professor: C. Richard Liu.
In this study, the mechanical properties of the deposited coating consisted of copper
nanoparticles and then the polymer carbon-based nano-composites are explored
respectively through various mechanical tests. In the first part, laser-induced chemical
solution deposition is introduced as a recently developed nano-manufacturing technique to
deposit thin film of copper nanoparticles on the copper substrate. In order to assess the
performance and properties of such porous nanostructured materials deposited by this
method, the micro-structure of deposited material is characterized by SEM and its
mechanical properties are investigated by a variety of experiments such as micro-hardness
test, nano-indentation test, bending test and adhesion test. The mechanical properties of
metals with surface deposition have been shown to be inherently strong to allow effective
usage in industrial and other applications. In the second part, different types of nano-
composites are studied: polymer matrix incorporated with two comparable nanoscale
additives. The popular carbon nano-tube and graphene nano-platelets are introduced into
epoxy matrix. Uniaxial tensile test and dynamic fatigue tensile test as well are conducted
to evaluate the tension properties and performance of different polymer nano-composites.
vi
vi
Both nanofillers show a decent improvement in ultimate tensile strength and Young’s
modulus, especially for graphene nano-platelets which are particularly helpful in adding
longevity of the fatigued composites.
Keywords: mechanical properties, laser-induced chemical solution deposition, copper
nanoparticle coating, polymer nano-composites, graphene nano-platelets
1
1
CHAPTER 1. INTRODUCTION
1.1 Nanostructured Material Deposited by Laser-induced Chemical Deposition
Physical Vapor Deposition (PVD) and Chemical Vapor Deposition (CVD) are two
top-used deposition techniques to deposit thin films or coating architectures onto various
substrate surface, especially for semiconductor industry. PVD is practically a vaporization
coating technique based upon the vaporized form of the desired coating material and
deposition of it onto substrate. It involves only physical processes such as thermal
evaporation and sputtering. PVD coating are commonly used to improve the products
hardness, wear resistance or oxidation resistance in industrial field like automotive,
aerospace and medical application. On the other hand, CVD is basically the formation of
solid materials via reaction of precursors or chemicals that contained the required
constituent on the substrate surface to produce desired coating. CVD is widely used in
semiconductor industry, especially for depositing super-thin (atomic level) coating.
Both PVD and CVD have been studied and well developed for many years. L.A.
Dobrzański and his group compared the properties of different coatings on cemented
carbide and cermet substrate when the coating deposition was carried out by PVD method
[1]. P.K Mehrotra and D.T. Quinto used a specialized method to measure the adhesion,
micro hardness and fracture toughness of CVD coating [2]. They also studied the
differences in the microstructure and stress state of CVD and PVD coating respectively [3].
2
2
However, both PVD and CVD have their disadvantages. For PVD, the rate of coating is
usually very slow and the homogeneity of deposition thickness is difficult to achieve.
Moreover, this process is typically operated at very high temperature and vacuum condition
which requires appropriate cooling system and special attention. For CVD, the
manufacturing process is relatively complex and there will be lots of toxic or corrosive
gasses emitted from the reaction of precursors and chemicals.
Recently, Chemical Solution Deposition (CSD) or Sol-Gel process which uses
liquid phase as a transfer media has become another dominating deposition technique other
than PCD and CVD. CSD process and chemical solution deposition of electronic oxide
films had been studied by R.W Schwartz and his group [4]. T. Schneller and his partner
also introduced the application of CSD [5].
In this work, we employ a new technique of deposition: laser-induced chemical
solution deposition which is similar to CSD, but uses laser-induced thermal shock to
accelerate the production rate. During this process, a laser beam is focused on the top
surface of substrate and initiate the chemical reactions within a tiny area. This processing
method costs less time and money, along with producing less heat and poisonous gaseous
by-products than the conventional methods during the material deposition. Such
advantages make the laser deposition easy to conduct and environmentally friendly. Thus
it is important to understand whether the nano-structured materials created have the basic
strength for supporting different application intended. In the following sections, we will
explore the mechanical properties of the thin film (coating) deposited by laser-induced
chemical deposition, such as hardness, modulus and fatigue behavior.
3
3
1.2 Polymer Nano-composites Synergized with Nanofillers
Polymer nano-composites, especially the ones with carbon-based nanofillers have
received prominent attention and been researched by considerable groups [6-8] over the
last decades. It was observed in several paper [9-10] that the polymer composites with only
few percent weight fraction of nanofillers had improved their mechanical properties as well
as thermal conductivity dramatically. These nanofillers-reinforced composites therefore
have many potentials in variety of structural applications, such as aircraft and electronics.
Traditionally, carbon nano-tube is the most studied and utilized nanostructure as additive
to enhance the mechanical properties of nano-composites. However, there are always some
key factors limiting its applications: the cost of production is very high and its dispersion
in polymer matrix is poor. Lately, graphene has been developed as another promising
carbon material, and deemed as an alternative or supplementary nanofillers to carbon nano-
tube.
Graphene is an allotrope of carbon in the form of an in-plane dimensional,
atomically thin, hexagonal lattice where every carbon atom forms each vertex. Graphene
has excited the scientist and emerged as an outstanding material due to its strong
nanofillers-matrix adhesion, high specific area and aspect ratio. It also possesses
extraordinary mechanical properties and conducts heat and electricity more efficiently.
Previously, graphene building blocks was also proposed to construct other carbon-based
nanostructure like carbon nanotube by rolling it up and vice versa. Now graphene is
produced by many methods, but the most popular and potential one is thermal reduction of
4
4
graphite oxide when bulk quantities of graphene nanoplatelets are needed [11]. Graphene
nanoplatelets (GNPs) are multiple stacks of individual layers of graphene sheet.
In this study, I use both multi-walled carbon nanotubes (MWNTs) and graphene
nanoplateltes (GNPs) respectively as additives to reinforce the polymeric matrix. My goal
is to determine the effects of different nanofillers on the composites’ mechanical behavior.
Epoxy is selected as the polymer matrix since it is a necessary technological material for
composite applications ranging from paints and coatings to adhesives used in industrial
engineering. In order to alleviate the poor condition of dispersion, acetone is also
incorporated as surfactant during mixing process. At a fixed weight fraction of MWNTs
and GNPs, we can compare the difference of these two nano-composites under tensile load
and find out which nanofiller provide the polymer nano-composite a better performance.
For each type of nanofillers, the filling fraction of the nanofiller has a great impact on the
performance of composite as well.
5
5
CHAPTER 2. MATERIALS, LASER-INDUCED CHEMICAL DEPOSITION
PROCESS AND EXPERIMENTS FOR THE DEPOSITED MATERIAL
2.1 Materials Selection
C110 copper rectangular bar (thickness=1/16 inches, width=1/2 inches) was
obtained as substrate from MSC Industrial Supply Co., NY, USA. Other chemicals
including copper chloride dihydrate, EDTA, sodium hydroxide, formaldehyde solution,
nitric acid and deionized water were purchased from local chemical store.
2.2 Sample Preparation and Processing Method
The copper substrates were milled to dog-bone shape (table.1, Fig. 1) for uniaxial
tensile test and laser deposition. After milling, all dog-bone shape samples were annealed
in a vacuum chamber at 600°C for 1 hour in order to remove strain hardening and residual
stress effects, followed by polishing them with sand paper.
Fig. 1. Dog-bone shape sample designed based on ASTM E8 [12]
6
6
Table 1. Values of each individual dimensions labeled in figure.1
Dimension Designation Value
W (Width of gauge section) 8 mm
R (Radius of Fillet) 6mm
G (Gauge Length) 25 mm
L (Overall Length) 100 mm
A (Length of the Reduced Section) 32mm
B (Length of Grip Section) 21mm
C (Width of Grip Section) 12.7mm
T (thickness) 1.5875mm
Copper nanoparticles were deposited on the substrate within an aqueous electrolyte
solution. The electrolyte solution was prepared by adding 0.2 g copper chloride dihydrate,
0.7 g EDTA, 0.3 g sodium hydroxide and 4.5 ml 36.5% formaldehyde liquid into 50 ml
deionized water. Copper chloride dihydrate served as the copper (Cu2+ ion) source.
Formaldehyde in alkaline medium acted as the reduction agent. EDTA was the complexing
agent of Cu2+ ion to prevent the precipitation of copper hydroxide and its chemical reaction
can be written as: Cu (EDTA) 2-+2HCHO+4OH-→Cu+EDTA4-+H2+2HCOO-+2H2O [13].
Before laser deposition, copper substrates were rinsed by nitric acid (10mM) and deionized
water to remove oxide layers. After the preparation of electrolyte solution, one rinsed
copper substrate was put inside the solution. The laser assisted electroless deposition of
copper nanoparticles was conducted with an ytterbium pulsed fiber laser and its conditions
was listed as Table. 2:
Table 2. Laser conditions for the deposition process
Laser power 5 W
Wavelength 1064 nm
Frequency 50 kHz
Pulse duration 100 ns
Beam size 1 mm
Scan speed 0.5 mm/s
7
7
2.3 Material Characterization
Top surface of the deposited sample was imaged by scanning electron microscope
(SEM). After 4 hours of laser assisted electroless deposition, the SEM images of the
deposited copper nanoparticles were shown in Fig. 2. The diameters of copper
nanoparticles were about 50nm estimated from SEM images by software ImageJ and the
total thickness of coating was about 3μm. A thickness of 8μm coating was achieved after
8 hours of deposition. A schematic diagram of the micro-structure was also depicted in Fig.
3.
A uniaxial tensile test of dog-bone copper substrate was performed in the first place
in order to get its elastic modulus as a benchmark. From the elastic regime of stress-strain
curve (Fig. 4.), the copper substrate’s Young’s Modulus was evaluated to be around 100
GPa.
Fig. 2. SEM images of the deposited layer: (a) top view (5000X) (b) top view
(80000X)
8
8
Fig. 3. Scheme of microstructure of deposited layer
Fig. 4. Uniaxial tensile test of copper substrate: stress vs. strain
-10
0
10
20
30
40
50
60
70
80
0 0.002 0.004 0.006 0.008 0.01
stre
ss (
MP
a)
strian
stress-strain
9
9
All samples’ number and their morphology of deposition were marked as follows:
(1) Sample 1 (Fig. 5.): rectangular sample with 3μm coating (40mm12.7mm).
Fig. 5. Overview and SEM images of 3μm coating: (a) overview of sample 1, (b) top
view (30000X) of 3μm coating, (c) side view (4000X), (d) side view (7000X)
(2) Sample 2 and sample 3: both are dog-bone samples with 3μm coating.
(3) Sample 4 (Fig. 6.) and sample 5: both are dog-bone samples with 8μm coating.
Fig. 6. (a) Overview of sample 4, (b) SEM image of 8μm coating: side view (3000X)
10
10
2.4 Mechanical Tests and Results
To investigate feasibility of laser-induced chemical solution deposition or
mechanical properties of the deposited layers, we performed micro-hardness test, nano-
indentation test and adhesion test.
2.4.1 Micro-hardness Test
Hardness is a measure of how resistant solid material is to different kinds of
permanent shape change during compressing force is applied. Our micro-hardness test was
designed based on ASTM standard B933 [14]. In order to minimize the penetration depth
and effect of substrate, we used a calibrated Leco M-400-H micro-hardness tester with
Knoop indenter, which was particularly designed for thin sheets and small indentation. The
test specimen was put on a plasticine hold to ensure that its top surface was stable and
perpendicular to the axis of the indenter. The pyramidal diamond indenter was forced into
the top surface of test specimen under an applied load=25gf, magnification=40, and
holding time=10s. The length of long diagonal of the indentation was measured down to
0.1μm through microscope and its hardness values were calculated from the applied
indentation force divided by the resulting projected area. Additional tests were made by
spacing the indentations with appropriate distance so that adjacent tests did not interfere
with each other. In our case, I measured independently 10 times and took the average for
each specimen.
The Knoop hardness (HK) relates to the applied load, indent area and geometry of
Knoop indenter, thus can be calculated by Equation 1. To convert HK to GPa, just multiply
11
11
HK value by 0.009807. The results of micro-hardness tests were shown as follows as Table.
3.
Equation 1: Calculation of Knoop hardness, where P is the applied load in g, D is the
measured length of long diagonal in micrometer.
HK = 14229 ×P
D2 [kgf/mm2]
Table 3. Knoop Hardness values and its standard deviation for different samples HK Hardness (GPa) STD (GPa)
Copper substrate (as-received)
142.1 1.39 0.047
Copper substrate (annealed)
100.8 0.99 0.08
Sample 1 (3μm) 183.8 1.8 0.161 Sample 2 (3μm) 126 1.24 0.06 Sample 4 (8μm) 104.5 1.02 0.084
2.4.2 Nano-indentation Test
Nano-indentation test is developed to measure the hardness of small volumes of
material. Unlike traditional indentation technique (macro or micro indentation) which is
limited by large and varied indenter tip shapes, nano-indentation possesses a smaller
precise tip shape and higher spatial resolution. Above all, nano-indentation test provides
real time load-displacement data when indentation is in progress and homogenized
properties could therefore be studied particularly in composite materials. During a nano-
indentation test, a prescribed load is applied to an indenter in contact with the specimen
surface while the applied load and penetration are continuously recorded. If the properties
and geometry (e.g. area to depth ratio) of the indenter are known, the indentation hardness
and modulus can be derived by the Oliver and Pharr method [15]. For Berkovich indenter,
the hardness and modulus is given by Equation 2. Since the modulus of diamond indenter
12
12
is very high and Poisson’s ratio is always small, the indentation (reduced) modulus is
usually considered as the elastic modulus of test specimen approximately.
Equation 2: Hardness and Modulus calculated from a Berkovich indenter, where P is the
applied load and hc is the contact depth of penetration. dP/dh is the slope of the initial
portion of the unloading curve.
H =𝑃
24.5ℎ𝑐2
and E =1
2
𝑑𝑃
𝑑ℎ
√𝜋
√24.5ℎ𝑐2
The nano-indentation experiments were performed under ambient conditions using
a TI-950 TriboIndenter (Hyistron Inc, MN, USA) equipped with a three-sided pyramidal
Berkovich probe (AA11051214). A fused quartz sample was used for the standard
calibration of its tip area function and instrument frame compliance prior to testing. The
nano-indentation tests were load-controlled through a partial load function (Fig. 7) and the
samples under such a loading-unloading condition would behave like Fig. 8.
Fig. 7. Load profile: applied load vs. time
0
1000
2000
3000
4000
5000
6000
7000
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
load
(uN
)
time(s)
load-time
13
13
Fig. 8. Load vs. indentation depth recorded for copper substrate
For each test sample, we designed 49 (7×7 array) different indent locations (Fig.9)
with a spacing of 15μm. We could get 5 hardness and modulus values respectively for each
location from one loading-unloading curve, and therefore 245 hardness and 245 modulus
values totally for each sample. By averaging the hardness and modulus, the test results
were summarized in Table. 4. Center region was the middle area of coating which was
fully deposited with nanoparticles, while transition region is the edge of coating which was
partially deposited.
Fig. 9. Nano-indentation grid of each specimen
-1000
0
1000
2000
3000
4000
5000
6000
7000
-50 0 50 100 150 200 250 300 350
load
(um
)
contact depth(nm)
load-depth
0
20
40
60
80
100
0 20 40 60 80 100
indent grid (μm)
14
14
Table 4. Modulus, hardness and their standard deviation of different specimens
Modulus
(GPa) STD of E
Hardness (GPa)
STD of H
Copper substrate as-received 96.077 7.33 1.82 0.38 Sample 1 (3μm) 87.53 35.38 0.93 0.62
Sample 4 (8μm) (center) 74.24 18.75 1.22 0.43 Sample 4 (8μm) (transition) 67.09 14.81 0.87 0.31
All the data points of each specimen were demonstrated in Fig. 10-17.
(1) Data plots of copper substrate:
Fig. 10. Hardness vs. depth of copper substrate (49 positions, 245 data points)
0
1
2
3
4
5
6
0 100 200 300 400
Har
dn
ess
depth(nm)
Hardness-depth
2000μN
3000μN
4000μN
5000μN
6000μN
15
15
0
20
40
60
80
100
120
140
0 25 50 75 100 125 150 175 200
Mo
du
lus(
GP
a)
depth(nm)
Modulus-depth(2000μN)(a)
0
20
40
60
80
100
120
140
0 25 50 75 100 125 150 175 200 225 250
Mo
du
lus(
GP
a)
depth(nm)
Modulus-depth(3000μN)(b)
0
20
40
60
80
100
120
0 25 50 75 100 125 150 175 200 225 250 275
Mo
du
lus(
GP
a)
depth(nm)
Modulus-depth(4000μN)(c)
16
16
Fig. 11. Modulus vs. depth of copper substrate under 5 different loading-unloading
steps: (a) max applied load=2000μN, (b) max applied load=3000μN, (c) max applied
load= 4000μN, (d) max applied load=5000μN, (e) max applied load= 6000μN
0
20
40
60
80
100
120
0 25 50 75 100 125 150 175 200 225 250 275 300 325
Mo
du
lus(
GP
a)
depth(nm)
Modulus-depth(5000μN)(d)
0
20
40
60
80
100
120
0 25 50 75 100 125 150 175 200 225 250 275 300 325 350 375
Mo
du
lus(
GP
a)
depth(nm)
Modulus-depth(6000μN)(e)
17
17
(2) Data plots of sample 1 (3μm coating):
Fig. 12. Hardness vs. depth of sample 1 (49 positions, 245 data points)
0
0.5
1
1.5
2
2.5
3
3.5
0 200 400 600 800 1000 1200 1400
Har
dn
ess
depth(nm)
Hardness-depth
2000μN
3000μN
4000μN
5000μN
6000μN
0
50
100
150
200
0 100 200 300 400 500 600 700 800 900 1000
Mo
du
lus(
GP
a)
depth(nm)
Modulus-depth(2000μN)
0
50
100
150
200
250
0 100 200 300 400 500 600 700 800 900 1000 1100
Mo
du
lus(
GP
a)
depth(nm)
Modulus-depth(3000μN)
18
18
Fig. 13. Modulus vs. depth of sample 1 under 5 different loading-unloading steps: (a)
max applied load=2000μN, (b) max applied load=3000μN, (c) max applied load=
4000μN, (d) max applied load=5000μN, (e) max applied load= 6000μN
0
50
100
150
200
250
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
Mo
du
lus(
GP
a)
depth(nm)
Modulus-depth(4000μN)
0
50
100
150
200
250
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300
Mo
du
lus(
GP
a)
depth(nm)
Modulus-depth(5000μN)
0
50
100
150
200
250
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300
Mo
du
lus(
GP
a)
depth(nm)
Modulus-depth(6000μN)
19
19
(3) Data plots of sample 4 (8μm coating, center region):
Fig. 14. Hardness vs. depth of sample 4 (49 positions, 245 data points)
0
0.5
1
1.5
2
2.5
3
0 200 400 600 800
Har
dn
ess
depth(nm)
Hardness-depth
2000μN
3000μN
4000μN
5000μN
6000μN
0
20
40
60
80
100
120
140
160
0 50 100 150 200 250 300 350 400 450
Mo
du
lus(
GP
a)
depth(nm)
Modulus-depth(2000μN)
0
20
40
60
80
100
120
140
0 50 100 150 200 250 300 350 400 450 500
Mo
du
lus
(GP
a)
depth (nm)
Modulus-depth(3000μN)
20
20
Fig.15. Modulus vs. depth of sample 4 (center region) under 5 different loading-
unloading steps: (a) max applied load=2000μN, (b) max applied load=3000μN, (c) max
applied load= 4000μN, (d) max applied load=5000μN, (e) max applied load= 6000μN
0
20
40
60
80
100
120
140
0 50 100 150 200 250 300 350 400 450 500 550
Mo
du
lus
(GP
a)
depth (nm)
Modulus-depth(4000μN)
0
20
40
60
80
100
120
0 50 100 150 200 250 300 350 400 450 500 550 600 650 700
Mo
du
lus
(GP
a)
depth (nm)
Modulus-depth(5000μN)
0
20
40
60
80
100
120
0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800
Mo
du
lus
(GP
a)
depth (nm)
Modulus-depth(6000μN)
21
21
(4) Data plots of sample 4 (8μm coating, transition region):
Fig.16. Hardness vs. depth of sample 4 (49 positions, 245 data points)
0
0.5
1
1.5
2
0 100 200 300 400 500 600 700
Har
dn
ess
depth(nm)
Hardness-depth
2000μN
3000μN
4000μN
5000μN
6000μN
0
20
40
60
80
100
120
0 50 100 150 200 250 300 350 400 450
Mo
du
lus（
GP
a）
depth（nm）
Modulus-depth(2000μN)
0
20
40
60
80
100
120
0 50 100 150 200 250 300 350 400 450 500 550
Mo
du
lus
(GP
a)
depth (nm)
Modulus-depth(3000μN)
22
22
Fig.17. Modulus vs. depth of sample 4 (transition region) under 5 different loading-
unloading steps: (a) max applied load=2000μN, (b) max applied load=3000μN, (c) max
applied load= 4000μN, (d) max applied load=5000μN, (e) max applied load= 6000μN
0
20
40
60
80
100
120
0 50 100 150 200 250 300 350 400 450 500 550 600
Mo
du
lus
(GP
a)
depth (nm)
Modulus-depth(4000μN)
0
20
40
60
80
100
120
0 50 100 150 200 250 300 350 400 450 500 550 600 650
Mo
du
lus
(GP
a)
depth (nm)
Modulus-depth(5000μN)
0
20
40
60
80
100
0 50 100 150 200 250 300 350 400 450 500 550 600 650 700
Mo
du
lus
(GP
a)
depth (nm)
Modulus-depth(6000μN)
23
23
2.4.3 Bending Test
Three-point bending test could give us estimated values for modulus of elasticity
in bending and flexural stress-strain response of tested specimen. We chose bending test
here instead of tensile test to exam the mechanical properties because our coating was very
thin and more sensitive to flexural stress change. But the test results could also be
influenced by the ambient environment and loading rate.
2.4.3.1 Analysis of Two-layer Composite Beam
For a traditional three-point bending test, the stress and modulus can be easily
obtained from simple beam theory as Equation 3.
Equation 3. Flexural stress, strain and modulus. Where L=span length, b=specimen width,
h=specimen thickness, P=load increment, =deflection increment
Flexural stress：σ =3PL
2h2b
Flexural strain：ϵ =6hδ
L2
Flexural modulus：E =PL3
4h3bδ
However, for our case -a composite beam-these expressions should be modified a
little due to the inhomogeneity. The scheme of the bending test is shown in Fig. 18 and 19
for a double-layer composite beam. The neutral axis from bottom is expressed as Equation
4 and then the second moment of inertia can be calculated from neutral axis by using
parallel axis theorem as Equation 5. Finally, the relation between the apparent modulus of
composite beam and the modulus of coating can be expressed as Equation 7 [16].
24
24
Fig. 18. Scheme of three point bending test
Fig. 19. Scheme of flexural stress and strain by eCourses Kurt Gramoll
(www.ecourses.ou.edu/cgi-bin/ebook.cgi?topic=me&chap_sec=06.1&page=theory)
25
25
Equation 4: Neutral axis from bottom. Where h1 and h2 is the thickness of copper substrate
and coating respectively, b is the width of the cross-section.
𝑦𝑐 =(h1
2 + h2) ∙ (n × b × h1) +h2
2 ∙ (b × h2)
(n × b × h1) + (b × h2)=
1
2
nh12 + 2nh1h2 + h2
2
nh1 + h2
n =𝐸𝑐𝑜𝑝𝑝𝑒𝑟
𝐸𝑐𝑜𝑎𝑡𝑖𝑛𝑔
Equation 5: Second moment of inertia
I = [bh2
3
12+ bh2 ∙ (yc −
h2
2)2] + [
nbh13
12+ nbh1 ∙ (h2 +
h1
2− yc)2]
Equation 6: Tension and compression stress at bottom and top surface (negative sign
means the stress direction).
σcoating = −My
I and σcopper = −n
My
I
M =P
4L at midspan (bending moment)
y = distance from neutral axis
Equation 7: relation between the apparent modulus of composite beam with 2 layers and
the modulus of coating.
EappIapp = EcoatingI
Iapp =bh3
12
26
26
2.4.3.2 Static Bending Test
Our bending test was performed by using MTS 810 hydraulic testing system. First of
all, the copper substrate was tested at a specified load rate of 0.1mm/min until permanent
(plastic) deformation occurred. The bending test results (Fig. 20.) presented some variation
or noise of the output signal due to the small thickness of sample and applied load.
Fig. 20. Static bending test for copper substrate: stress vs. strain
The apparent flexural modulus of copper substrate calculated from the slope of stress
vs. strain curve was around 40MPa. Based on the yield strength of copper substrate, we
designed a similar static bending test for sample 4 whose coating thickness was 8μm. The
load was applied below the yield strength so that the sample only deformed within the elastic
regime. The test result was shown in Fig. 21 and its apparent modulus obtained from the
slope of this curve was about 34 MPa which was slightly lower than that of copper substrate.
This might be attributed to the small modulus of coating compared to the dense substrate.
These moduli computed from curving fitting need more experiments to verify their
-20
0
20
40
60
80
100
120
140
-0.005 0 0.005 0.01 0.015 0.02 0.025 0.03
stre
ss(M
Pa)
strain
stress-strain
27
27
repeatability in the future because the applied load and thickness of coating were relatively
small, which made the difference between samples inconspicuous.
Fig. 21. Static bending test for sample 4: stress vs. strain
2.4.3.3 Fatigue Bending Test
To inspect the fatigue behavior of samples, our bending fatigue test was performed
on sample 4 by using cyclic displacement control. The test specimen was placed
symmetrically on two supports and then loaded by a loading nose midway between the
supports. The span length=30mm, displacement range=0-0.1mm (within elastic regime),
frequency=10Hz. After 700000 cyclic loading, the coating surface was examined through
SEM (Fig. 22). Later after 1 million cycles, another image was taken again on the same spot
(Fig. 23). Comparing those images, we could see the peeling-off (circles in Fig. 22) of some
nanoparticles.
0
10
20
30
40
50
60
70
0 0.0005 0.001 0.0015 0.002 0.0025
stre
ss(M
Pa)
strain
stress-strain
28
28
Fig. 22. Overview and SEM images of nanoparticle coating after 700000 cycles: (a)
overview of sample 4, (b) top view (600X), (c) top view (1100X), (d) top view (8000X)
Fig. 23. SEM images of nanoparticle coating after 1 million cycles: (a) top view
(1500X), (b) top view (6000X)
29
29
2.4.4 Qualitative Adhesion Test
To determine if the coating is properly adhered to the top smooth surface, adhesion
test is used to assess the resistance of deposition or coating from substrate. Our qualitative
adhesion test was designed based on ASTM B571 [17]: a hardened steel tool which was
ground to a sharp point (e.g. chisel) was used to draw some lines and rectangular grid pattern
on the coating surface with a distance of 3mm between the lines. When drawing the lines,
sufficient pressure force was applied to cut through the coating till the substrate in a single
stroke. If any parts of the deposition between the lines broke away from copper substrate, the
adhesion was not adequate.
After a satisfactory adhesion was exhibited, I placed a CHT tape (M.E. Taylor
Engineering Inc., MD, USA), which possessed an adhesion bond strength of 60g/mm, onto
a clean grid area with firm finger pressure. Shortly afterward, the tape was removed by
grabbing one free end and peeling it off quickly. The adhesion was not enough if the tape
had deposited particles that came from that grid area adhering to it. From the test result (Fig.
24), the adhesion strength was adequate as no coating broke away between the scribed lines
and no copper particles adhering to the tape.
Fig. 24. Scratched sample and adhesive tape after testing
30
30
2.4.5 Porosity Estimation
Porosity is a measure of empty space in a matter and usually defined as a fraction of
the volumes of voids over total volume. Since the deposited coating consists of numerous
copper nanoparticles, its porosity or relative density might be derived by approximate
calculation. First of all, the weight of pure copper substrate was measured by XP26 Mettler
Toledo micro-balance (Mettler-Toledo LLC, OH, USA), and the value was 14.817055 g.
After laser deposition, the same sample (sample 5) was measured again and its weight
became 14.821675 g. By subtracting the weight of copper substrate, we could know that the
net weight of deposited layer was about 4.62 mg. For a fully dense part, the weight of
deposited layer which could be estimated by multiplying its volume (1.68mm3) and copper
density (8.96g/cm3) was 15 mg. And the relative density was also calculated as:
4.62/1.68=2.75g/cm3. This estimation of relative density based on the volume of deposited
coating was very rough, because it was assumed to be a regular rectangle and its volume
simply came from lengthwidththickness. Sample 5 was also sent out to measure the
porosity of coating commercially but the company proved that not being able to measure it.
So a more accurate model is needed to estimate the porosity of thin coating and this is our
future work.
31
31
CHAPTER 3. MATERIALS, DISPERSION PROCESS AND EXPERIMENTS FOR
POLYMER NANO-COMPOSITES
3.1 Materials Selection
Multi-walled carbon nanotubes (MWNTs) and graphene nanoplatelets (GNPs)
were purchased from Cheap Tubes Inc. (USA). The properties of MWNTs were provided
from the company: outer diameter=10-20nm, inner diameter=5-10nm, purity>95wt%,
length=10-30um, Ash<1.5wt%, specific surface area>233m2/g, density=2.1g/cm3. The
properties of GNPs were also provided by the company: lateral dimensions=1-2µm,
purity>99wt%, number of layers<4, average thickness<4nm, SSA> 750 m2/g. The epoxy
(system 2000 epoxy resin), curing agent (2120 epoxy hardener) and acetone were all
supplied from Fibre Glast, USA. Mold Max®25 was obtained from Smooth-On, Inc. (PA,
USA) to make silicone rubber mold.
3.2 Sample Preparation and Processing Method
The dispersion process was modified based on Rafiee’s methodology [18]. A
preferred amount of nanofillers (MWNTs or GNPs) was measured and dispersed in acetone
(ratio: 100ml acetone to 0.1g nanofillers) by an ultrasonic probe sonicator at high amplitude
in an ice bath. After 1.5 hours of dispersion, a corresponding amount (30g) of epoxy resin
was added to the previous mixture followed by the sonication again for another 1.5 hours.
32
32
In order to remove the acetone in the mixture subsequently, the container was placed on a
hot (80℃) magnetic plate with a Teflon-coated bar stirring inside for 3 hours under fume
hood. When heating was done, the mixture was put in vacuum chamber for 12 hours to
ensure that all of the residual acetone had been evaporated. The mixture was then taken out
and added with low-viscosity epoxy hardener (mix ratio is 100:27 by weight). A high-
speed shear mix machine was used to mix slurry thoroughly for 5 minutes at 2000rpm. The
mixture was again placed in the vacuum chamber to get rid of air bubbles for 1.5 hours.
Finally, the epoxy-nanofillers composite was poured into the mold (Fig. 25) and cured at
room temperature for 24 hours, followed by 4 hours of post cure at 90℃ in a heating oven.
Pure epoxy sample was also prepared by the similar procedure but without adding any
nanofillers.
Fig. 25. Silicon rubber mold and epoxy composite with nano-additives
33
33
3.3 Mechanical Tests and Results
3.3.1 Static Tensile Test
To research the nanofillers’ influence on tensile properties for epoxy composites, we
compared the pure epoxy with epoxy nano-composites containing various weight fraction
(0.1%, 0.3% and 0.5%) of MWNTs and GNPs. A low weight fraction of nanofillers was
selected for the sake of uniform dispersion. Quasi-static tension tests were performed on a
MTS QTest/50LP test machine. Extension rate was set to be 0.05 inch/min before the
uniaxial tensile test, and the sample’s extension was measured by a clip-on type extensometer
over a gauge length of 1 inch. The uniaxial tensile test results were shown in table. 5 and Fig.
26-27.
Table. 5. ultimate tensile strength (UTS) and Young’s modulus of pure epoxy and epoxy
nano-composites from tensile test
Materials UTS (MPa) Young’s Modulus
(GPa) Elongation (in)
pure epoxy 52.5 3.232 0.013
0.1wt%MWNTs 79.7 3.575 0.03
0.3wt%MWNTs 58.5 4.375 0.013
0.5wt%MWNTs 53 4.603 0.01
0.1wt%GNPs 59.8 3.294 0.025
0.3wt%GNPs 58.7 3.572 0.026
0.5wt%GNPs 58.4 4.44 0.014
34
34
Fig. 26. ultimate tensile strength of pure epoxy and epoxy nano-composites with different
weight fraction of additives
In terms of ultimate tensile strength, epoxy nano-composites with different
nanofillers additives out-performed the pure epoxy overall. The tensile strength of 0.1%
MWNTs/epoxy composites were about 50% higher than that of pure epoxy. In contrast,
0.1% GNPs/epoxy composites acquired 14% increase in the tensile strength. Higher
nanofillers loadings also displayed increased tensile strength when compared to the
baseline pure epoxy, but the improvement was not as impressive as that from 0.1% weight
fraction of nanofillers. In summary, the epoxy nano-composites offered the maximum
reinforcement in ultimate tensile strength at relatively lower nanofiller loading. Although
the promotion of ultimate tensile strength tended to decay as the nanofillers weight fraction
increased, GNPs still held a satisfactory performance at larger additives concentration.
35
35
Figure. 27. Young’s modulus of the pure epoxy and epoxy nanocomposties with
different weight fraction of additives.
Young’s modulus was roughly estimated from the slope of stress-strain curve
through curve fitting. The modulus of nano-composite seemed to increase as the additives
concentration increased. Incorporation of 0.5% weight fraction of MWNTs caused its
Young’s modulus to increase by about 40%. And by introducing the same weight fraction
of GNPs, there was around 36% increment in Young’s modulus. The empirical results
exhibited that the significant enhancement of Young’s modulus occurs with a higher value
of weight fraction.
3.3.2 Theoretical Analysis of Young’s modulus
The Halpin-Tsai model which accounts for the aspect ratio, volume fraction of each
constituent as well as the tensile properties of the matrix and inclusion has been used to
predict the Young’s modulus of reinforced composite material [19]. In this well-established
36
36
model, MWNTs can be treated as randomly oriented fiber lamina, and the Young’s
modulus was estimated from the Equation 8.
Equation 8: Where EC=Young’s modulus of the composite, EM=Young’s modulus of the
epoxy matrix, l=length of MWNTs (20µm), d=average outer diameter of MWNTs
(15nm), Eeq= (2t/r)EMWNT is equivalent modulus of the MWNTs considering the hollow
tube as a solid cylinder, t=nanotube wall thickness (3.75nm), r=nanotube radius (7.5nm),
EMWNT was estimated as 810GPa,VMWNT is the volume fraction of the MWNTs.
𝐸𝐶 =3
8
1 + 2(𝑙𝑀𝑊𝑁𝑇
𝑑𝑀𝑊𝑁𝑇)(
𝐸𝑒𝑞
𝐸𝑀− 1
𝐸𝑒𝑞
𝐸𝑀+ 2
𝑙𝑀𝑊𝑁𝑇
𝑑𝑀𝑊𝑁𝑇
)𝑉𝑀𝑊𝑁𝑇
1 − (
𝐸𝑒𝑞
𝐸𝑀− 1
𝐸𝑒𝑞
𝐸𝑀+ 2
𝑙𝑀𝑊𝑁𝑇
𝑑𝑀𝑊𝑁𝑇
)𝑉𝑀𝑊𝑁𝑇
× 𝐸𝑀 +5
8
1 + 2(
𝐸𝑒𝑞
𝐸𝑀− 1
𝐸𝑒𝑞
𝐸𝑀+ 2
)𝑉𝑀𝑊𝑁𝑇
1 − (
𝐸𝑒𝑞
𝐸𝑀− 1
𝐸𝑒𝑞
𝐸𝑀+ 2
)𝑉𝑀𝑊𝑁𝑇
× 𝐸𝑀
Unlike MWNT is regarded as cylinder fiber, GNP is assumed as rectangular sheet
with width (w), length (l) and thickness (t). To predict the GNP nano-composites modulus,
the Halpin-Tsai model is modified as Equation 9.
Equation 9: Where l=w=1.5µm, t=4nm, EMWNT =1TPa, VGNP= the volume fraction of
GNPs
𝐸𝐶 =3
8
1 + (𝑤 + 𝑙
𝑡 )(
𝐸𝐺𝑁𝑃
𝐸𝑀− 1
𝐸𝐺𝑁𝑃
𝐸𝑀+
𝑤 + 𝑙𝑡
)𝑉𝐺𝑁𝑃
1 − (
𝐸𝐺𝑁𝑃
𝐸𝑀− 1
𝐸𝐺𝑁𝑃
𝐸𝑀+
𝑤 + 𝑙𝑡
)𝑉𝐺𝑁𝑃
× 𝐸𝑀 +5
8
1 + 2(
𝐸𝐺𝑁𝑃
𝐸𝑀− 1
𝐸𝐺𝑁𝑃
𝐸𝑀+ 2
)𝑉𝐺𝑁𝑃
1 − (
𝐸𝐺𝑁𝑃
𝐸𝑀− 1
𝐸𝐺𝑁𝑃
𝐸𝑀+ 2
)𝑉𝐺𝑁𝑃
× 𝐸𝑀
37
37
In order to get the volume fraction VMWNT or VGNP, weigh fraction need to be
converted to volume fraction with the density of constituent for fiber reinforced composites:
𝑉𝐹 =𝜌𝐶
𝜌𝐹𝑀𝐹;
𝜌𝐶 = 𝜌𝐹𝑉𝐹 + 𝜌𝑀𝑉𝑀 = 𝜌𝐹𝑉𝐹 + 𝜌𝑀(1 − 𝑉𝐹);
Here 𝜌𝐶 , 𝜌𝐹 and 𝜌𝑀 mean the density of composite, nanofiller and matrix
respectively. VF and VM represent the volume fraction of naofiller and matrix respectively.
MF is the weight fraction of the nanofiller. By combining the two equations above, volume
fraction can be written as Equation 10.
Equation 10: volume fraction of fiber calculated from weight fraction
𝑉𝐹 =𝑀𝐹
𝑀𝐹 +𝜌𝐹
𝜌𝑀(1 − 𝑀𝐹)
The density of 2000 epoxy resin was 1.135g/cm3 measured by the material supplier.
The density of MWNTs and GNPs were ρMWNT= 2.1g/cm3 and ρGNP=1.9g/cm3 obtained
from the company. Thus the corresponding volume fraction for a 0.5% weight fraction of
nanofillers were calculated as VMWNTs=0.271%, VGNPs=0.299% respectively. The
theoretical prediction of Young’s modulus for GNP composite at that nanofiller loading
level was around 4.04GPa, which underestimated the experimental results by about 10%.
This could probably be a result of the wrinkled surface of GNP which was different from
a regular rectangle presumed in the modified Halpin-Tsai model.
38
38
3.3.3 Dynamic tensile fatigue test
To evaluate the composites’ performance under dynamic fatigue conditions, the
tensile fatigue tests were conducted using a MTS servo-hydraulic fatigue testing machine.
For both MWNTs and GNPs nano-composites, we tested 3 samples to check for
reproducibility at each weight fraction of additives and took average as their fatigue lives.
Cyclic loading condition was set as: Amplitude=0.18 kN, Mean=0.22 kN (stress ratio
R=0.1); frequency=1 Hz. Each sample and the average number of cycles to failure were
recorded in table 6.
Table. 6. Mean fatigue life of pure epoxy and its composites with various weight
fractions of additives
Materials Fatigue life (cycles)
pure epoxy 2327
0.1wt%MWNTs 4104
0.3wt%MWNTs 5675
0.5wt%MWNTs 8000
0.1wt%GNPs 4645
0.3wt%GNPs 15049
0.5wt%GNPs 8549
Comparative enhancement in mean fatigue life was achieved over a wide range of
filling fraction for both MWNTs and GNPs. The fatigue life of composites with 0.5%
weight fraction of MWNTs was improved by 4-fold approximately from 2327 cycles to
8000 cycles as compared to the pure epoxy. For the composites with 0.3% weight fraction
of GNPs, the fatigue life was even enhanced by up to 7-fold. While according to Daniel R.
39
39
Bortz’s paper, the fatigue life of reinforced composites might be sometimes on par with or
even inferior to the pristine material [20]. This phenomenon might probably be ascribed to
the competition between the beneficial reinforcement from isolated nanofillers and
detrimental stress concentration effect caused by agglomeration of nanofillers. So there
would be an optimal loading fraction for every nanofillers.
.
40
40
CHAPTER 4. DISCUSSION
4.1 Copper Nanoparticles Coating Deposited by Laser-induced Chemical Deposition
From the test results of nano-hardness test, samples with nanostructured porous
material have different mechanical properties than the copper substrate. Both 3μm and 8μm
deposited samples show larger variation for hardness and modulus when coming to the
nanoscale. On top of that, the values of hardness and modulus seem to be affected by the
indent positions. This probably results from surface roughness of the tested areas or non-
uniform coating during deposition process, especially when the detecting indenter is only
dozens of nanometers and comparable with nanoparticle size. While for the solid copper
substrate, a much more consistent values are seen with lower variation.
The porosity is spontaneously involved in the results when modulus and hardness
are measured. The average modulus and hardness of deposited coating are much lower than
those of copper solid substrate overall, which is an evidence of the porosity in the deposited
coating [21]. If we look into each indent location on every sample and compare those 5
values of hardness or modulus obtained from 5 loading steps within one single indentation,
it is easy to find out that both modulus and hardness increase slowly as the penetration
depth grows for most indent locations (Fig. 28-Fig. 29). This agrees well with Xi Chen’s
work [22] about porous material and suggests that local densification of porous material
may have occurred in deeper regions underneath the indenter tip during nano-indentation.
41
41
A better convergence of modulus at higher applied load also indicates that homogeneity or
the density of the local neighborhood in the vicinity of the indenter has increased as the
indentation depth has increased. However, there were few indent locations showing a
contrary phenomenon. These might be caused by some intrinsic defects such as vacancy or
surface-connected pores.
Figure. 28. Average hardness vs. contact depth under five different loads
Figure. 29. Average modulus vs. contact depth under five different loads
42
42
When comparing the results of sample 1 and sample 4, we also learn that the nano-
hardness value for sample 4 is slightly higher. The possible reason is that as the coating
thickness increase, localized densification (lower porosity) around the indenter becomes
more significant after a longer time of deposition.
The Knoop hardness (HK) of sample 1 is even higher than that of copper substrate.
This could be ascribed to the substrate interfering. For the optimum accuracy of
measurement, the thickness of coating is usually at least ten times the depth of the
indentation in order to minimize the substrate effect. For our Knoop hardness test on
sample 1, the indentation depth is about 1.5μm which is almost half of the 3μm coating.
While for Knoop hardness test on sample 4, the indentation depth is about 1.9μm which is
only 20% of the 8μm coating. To some extent, the micro-hardness values of sample 1 is
more likely to be influenced by the presence of copper substrate. So the micro-hardness
values of sample 4 are more accurate than that of sample 1, and thus closer to the results
from nano-indentation. In addition, the micro-hardness values are very sensitive to the
applied load and shape of indenter. This can also lead to a large error when considering the
substrate.
4.2 Epoxy Nano-composites Synergized with Carbon Nanofillers
Desirable improvement of tensile properties is seen due to the influence of
nanofillers inside the epoxy matrix for both MWNTs and GNPs composites overall. The
wrinkled and rough surfaces of those nanofillers (Fig. 30), along with their large contact
surface area provided a strong interlock bonding with the polymer matrix and hence a
43
43
superior adhesion at the inclusions/matrix interface. However, a degradation of
performance is also observed after a certain amount of filling fraction is reached. This
deterioration of improvement can be mainly explained by the agglomeration or bad-
dispersion of the nanofillers. The high surface area between nanofillers always results in
larger Van der Waals force and consequently makes it easier to aggregate and stack,
especially for the 2-dimensional GNP as clusters of GNP in the powder are even visible
before dispersion. Those agglomerates would not only debilitate the interfacial bonding
through decreasing the specific surface area, but also form disadvantageous voids and holes
by preventing polymer from flowing into the agglomerates [23]. These defects, contained
in the agglomerates, have a great impact on the performance of nano-composites if at high
concentration.
Fig. 30. SEM image of graphene nano-plates from material supplier Cheat Tube
Under the cyclic-loading condition, both MWNTs and GNPs are capable of
prolonging the fatigue life. But GNPs seem to endow its polymer nano-composite with a
better fatigue property because the responsible mechanisms of crack resisting are not
identical for these two additives. In the case of MWNTs, the prevailing strengthening
44
44
mechanism is crack bridging which means the crack is “bridged” by the carbon nano-tube
near the crack front. Evidence is the pull-out or rupture of the nano-tubes observed at the
fracture surface [24]. When initial crack is advancing, the crack propagation rate will be
slowed since more energy needs to be absorbed to overcome the friction caused by the pull-
out of fibers. While for the GNPs, the predominant mechanism of fatigue suppression
becomes crack deflecting. Crack front will veer out of its original propagation route when
it encounters a series of rigid graphene nano-platelets. The coarse fracture surface observed
from SEM pictures indicates that crack deflecting happened and new fracture surface is
generated during the process of crack continuation [20]. As the crack tip is forced to tilt or
detour around the sheets, it also dissipates more energy for the growth of crack because
crack under mix-mode loading condition requires a larger driving force to propagate than
just under fracture mode I (opening mode). Moreover, the ultrathin planar nanostructure of
GNP, combined with its high aspect ratio and specific surface area, makes it more effective
at crack arresting or deflecting than carbon nano-tube.
45
45
CHAPTER 5. CONCLUSIONS
In the first part, we studied the mechanical properties of a uniquely deposited
coating of copper nanoparticles on copper substrate. The results have shown that there is a
propensity for high scatter and variation in the mechanical response as observed from the
nano-indentation test. The porosity of the deposited coating in the localized testing regions
and the coating thickness have an effect on the mechanical properties of such engineered
deposition and composites. From the results of nano-indentation test, it is observed that the
elastic modulus and hardness of the sample show a strong correlation to the indentation
depth. As the indentation depth increased, the values of the elastic modulus and hardness
increased. This trend could be attributed to either the densification (porosity decrease) of
the porous material underneath the indenter or the presence of copper substrate at higher
indentation depth. At small indentation depth (less than 10%) compared to the coating
thickness, the effect of the substrate material is negligible as the coating is predominantly
responsible for the mechanical response to the micro-indentation. The fatigue test results
present some nano-particles loss or coating peel-off after cyclic loading which imply that
we may have a different mode of fracture or damage than that of a dense part. Meanwhile
the adhesion test results have shown that the nanostructured materials deposited by laser-
induced chemical deposition have a satisfactory bonding strength with the substrate.
46
46
In the second part, we have demonstrated that both tensile properties and fatigue
life have been improved by introducing nanofillers into epoxy matrix. Enhancement in
tensile strength is modest but better at lower nanofillers weight fraction while Young’s
modulus appears to gain notable boost at higher loading fraction. So an equilibrium point
has to be figured out to balance the behaviors of such polymer nano-composites based on
different applications. GNP is more remarkable at fatigue crack pinning over MWNT due
to its unique geometry and toughening mechanism. However, in order to fully realize the
potential of carbon-based composite materials, a crucial issue is to disperse the nanofillers
uniformly in the matrix or namely minimize the effect of agglomeration. In some literatures
[11, 23, 25], this problem can be partly resolved by means of adding functional group or
mixing with other inclusions to facilitate the dispersion in solvent.
47
REFERENCES
51
47
47
REFERENCES
[1] L. A. Dobrzaski, A. liwa, L. W. ukowska, J. Mikula, K. Golombek; Structure and
mechanical properties of PVD coatings for tool materials; Journal of Achievements in
Materials and Manufacturing Engineering 42 (1-2) (2010) 33–41
[2] P. K. Mehrotra, D. T. Quinto; Techniques for evaluating mechanical properties of hard
coatings; Journal of Vacuum Science & Technology A 3, 2401 (1985)
[3] D. T. Quinto; Mechanical property and structure relationships in hard coatings for
cutting tools; Journal of Vacuum Science & Technology A 6 (3) (1988) 2149–2157
[4] R. W. Schwartz, T. Schneller, R. Waser; Chemical solution deposition of electronic
oxide films; Comptes Rendus Chimie 7 (5) (2004) 433–461
[5] T. Schneller, S. Tiedke, R. Waser; Chemical Solution Deposition (CSD) and Quality
Assurance of Electroceramic Thin Films for Information Technology
[6] Nanda Gopal Sahoo, Sravendra Rana, Jae Whan Cho, Lin Li, Siew Hwa Chan; Polymer
nanocomposites based on functionalized carbon nanotubes; Progress in Polymer
Science 35 (2010) 837-867
48
48
[7] Takashi Kashiwagi, Eric Grulke, Jenny Hilding, Katrina Groth, Richard Harris,
Kathryn Butler, John Shields, Semen Kharchenko, Jack Douglas; Thermal and
flammability properties of polypropylene/carbon nanotube nanocomposites; Polymer
45 (2004) 4227–4239
[8] N. Yu and Y.W. Chang; Effects of CNT Diameter on the Uniaxial Stress-Strain
Behavior of CNT/Epoxy Composites; Journal of Nanomaterials 2008 (834248)
[9] Jianfeng Shen, Bo Yan, Tie Li, Yu Long, Na Li and Mingxin Ye; Mechanical,
thermal and swelling properties of poly (acrylic acid)–graphene oxide composite
hydrogels; Soft matter, 2012, 8, 1831
[10] Yuanqing Li, Tianyi Yang, Ting Yu, Lianxi Zheng and Kin Liao; Synergistic effect
of hybrid carbon nantube–graphene oxide as a nanofiller in enhancing the mechanical
properties of PVA composites; Journal of Materials Chemistry, 2011, 21, 10844
[11] F. Yavari and N. Koratkar; Dramatic Increase in Fatigue Life in Hierarchical Graphene
Composites; ACS Appl. Mater. Interfaces, 2010, 2 (10), pp 2738–27
[12] ASTM E8/E8M-13a Standard Test Methods for Tension Testing of Metallic Materials
(2013)
49
49
[13] V. A. Kochemirovsky, L. Menchikov, S. V. Safonov, M. D. Bal’Makov, I. I. Tumkin,
Yu S Tveryanovich; Laser-induced chemical liquid phase deposition of metals:
chemical reactions in solution and activation of dielectric surfaces; RUSS CHEM REV,
2011, 80 (9), 869–882
[14] ASTM B933-14 Standard Test Method for Microindentation Hardness of Powder
Metallurgy (P/M) Materials (2009)
[15] A. C. Fischer-Cripps; Nanoindentation Third Edition; Mechanical Engineering
Series,Springer New York, New York, NY, 2011.35
[16] Shiladitya Paul, James Andrew Curran; Apparent modulus of a composite beam;
University of Cambridge
[17] ASTM B571-97 Standard Practice for Qualitative Adhesion Testing of Metallic
Coatings (2008).
[18] Mohammad A. Rafiee, Javad Rafiee, Iti Srivastava, Zhou Wang, Huaihe Song, Zhong-
Zhen Yu, and Nikhil Koratkar; Fracture and Fatigue in Graphene Nanocomposites;
small 2010, 6, No. 2, 179–183
[19] Mohammad A. Rafiee, Javad Rafiee, Zhou Wang, Huaihe Song, Zhong-Zhen Yu, and
Nikhil Koratkar; Enhanced Mechanical Properties of Nanocomposites at Low
graphene content; ACSNANO VOL. 3 ∙ NO. 12
50
50
[20] Daniel R. Bortz, Erika Garcia Heras, and Ignacio Martin-Gullon; Impressive Fatigue
Life and Fracture Toughness Improvement in Graphene Oxide/Epoxy Composites;
Macromolecules 2012, 45, 238-245
[21] V. Kralik, J. Nemecek; Micromechanical properties of porous material based on metal
foam; Chemicke Listy 105 (2011) S672–S675
[22] X. Chen, Y. Xiang, J. J. Vlassak; Novel technique for measuring the mechanical
properties of porous materials by nanoindentation; Journal of Materials Research 21
(03) (2006) 715–724
[23] Shin-Yi Yang, Wei-Ning Lin, Yuan-Li Huang, Hsi-Wen Tien, Jeng-Yu Wang, Chen-
Chi M. Ma, Shin-Ming Li, Yu-Sheng Wang; Synergetic effects of graphene platelets
and carbon nanotubes on the mechanical and thermal properties of composites; Carbon
49(2011)793-803
[24] A. Zandiatashbar, R. C. Picu, N. Koratkar; Mechanical behavior of epoxy-graphene
Platelets Nanocomposites; Journal of Engineering Materials and Technology JULY
2012, Vol. 134 / 031011-3
[25] Jeffrey R. Potts, Daniel R. Dreyer, Christopher W. Bielawski, Rodney S. Ruoff;
Graphene-based polymer nanocomposites; Polymer 52 (2011) 5-25
