University of Southern California Dissertations and Theses

Synthesis, characterization, and mechanical properties of nanoporous foams

(USC Thesis Other)

Synthesis, characterization, and mechanical properties of nanoporous foams

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
SYNTHESIS, CHARACTERIZATION, AND MECHANICAL PROPERTIES OF
NANOPOROUS FOAMS

by

I-Chung Cheng

_____________________________________________________________

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In partial fulfillment of the
Requirements for the degree
DOCTOR OF PHILOSOPHY
(MATERIALS SCIENCE)

May 2013

Copyright2013 I-Chung Cheng
ii

Dedication

To my parents, my wife, and my son, who always support and love me
iii

Acknowledgments

First of all, I would like to thank my advisor Prof. Andrea Hodge, who has always
believed in me and given me courage during tough times. This thesis won’t be
possible without Prof. Hodge’s great patience, unconditional support and guidance,
especially when I felt lost and like I was going nowhere throughout this work. It was
her dedication that helped me to move forward and think like a scientist.
I would also like to thank Prof. Edward Goo who provided me the opportunity to
enter USC and guided me through the first semester. It was with his help that I found
Prof. Hodge as my advisor. I would like to thank Prof. Steven Nutt, Prof. Edward
Goo, and Prof. Michael Kassner for accepting the request to serve on my dissertation
defense committee.
Special thanks to my colleagues Tim Furnish, who always listened and talked to me
when I needed him, and Mikhail Polyakov who was always willing to read through
my writings. Also, thanks to the other members in Prof. Hodge’s group: Dr. Anahita
Navid, Dr. Vinod Tim Nayar, Motunrayo Onafalujo, Derrick C. Ike, Thien Phan,
Yifu Zhao, Leonardo Velasco, Shima Haghighat, Nathan Heckman, and Theresa
Juarez for always being kind to me and treating me like a family member.
Additionally, thanks must be given to John Curulli at CEMMA, Dr. Kenneth
Vecchio at UCSD, Dr. Javier Garay at UCR, Dr. Steve Cronin at USC, Dr. Oliver
iv

Franke at USC and the funding from Air Force, AFOSR grant number FA9550-10-1-
0478.
Finally, special thanks to my parents, Shyh-Jehng and Shu-Ming, for their love and
support for all these years, and, most important of all, thanks to m y wife, Ling-Chen,
who always encourages me and supports me at all times.

v

Table of Contents
Dedication.....................................................................................................................ii
Acknowledgments.......................................................................................................iii
List of Tables.............................................................................................................viii
List of Figures..............................................................................................................ix
Abstract......................................................................................................................xv
Chapter 1. Introduction.......................................................................................1
Chapter 2. Background........................................................................................5
2.1 Cellular solids.....................................................................................................5
2.2 Foam processing.................................................................................................7
2.2.1 Dealloying...................................................................................................7
2.2.2 Free corrosion and electrochemically driven dealloying ........................... 11
2.2.3 Heat treatments .......................................................................................... 13
2.3 Foam mechanical properties.............................................................................15
2.3.1 Deformation of foams ................................................................................ 18
2.4 Foam applications.............................................................................................26
Chapter 3. Experimental procedures and materials.....................................31
3.1 Synthesis of NP Au foams................................................................................31
3.2 Synthesis of other metal foams systems...........................................................35
3.2.1 System selection ........................................................................................ 35
3.2.2 Precursor alloys preparation ...................................................................... 38
3.2.3 Electrochemical setup ................................................................................ 40
3.2.4 Preliminary results ..................................................................................... 41
3.2.5 Post dealloying heat treatments ................................................................. 43
3.3 Characterization methods..................................................................................44
vi

3.3.1 SEM/EDX (Scanning Electron Microscope/Energy Dispersive X-ray) .... 44
3.3.2 XRD (X-Ray Diffraction) .......................................................................... 44
3.3.3 Raman Spectroscopy ................................................................................. 44
3.4 Mechanical testing............................................................................................45
3.4.1 Nanoindentation ......................................................................................... 45
3.4.2 Micro Vickers hardness testing ................................................................. 47
Chapter 4. Morphology, Oxidation and Mechanical Behavior of Nanoporous
Cu Foams...................................................................................................................48
4.1 Introduction.......................................................................................................48
4.2 Experimental details..........................................................................................49
4.2.1 Foam processing ........................................................................................ 49
4.2.2 Foams characterization and testing ............................................................ 50
4.3 Results and Discussion......................................................................................52
4.3.1 Ligament size and morphology ................................................................. 52
4.3.2 Oxide Formation ........................................................................................ 56
4.3.3 Nanoindentation ......................................................................................... 61
4.3.4 Heat treatments in vacuum ........................................................................ 64
4.4 Summary...........................................................................................................68
Chapter 5. High temperature morphology and stability of nanoporous Ag
foams..........................................................................................................................69
5.1 Introduction.......................................................................................................69
5.2 Experimental Details.........................................................................................70
5.3 Results and discussion......................................................................................73
5.3.1 Ligament size and morphology ................................................................. 73
5.3.2 Heat treatments in argon/Oxide formation and decomposition ................. 78
5.3.3 Heat treatments in vacuum/Thermal evaporation of silver ........................ 80
5.4 Summary...........................................................................................................81
Chapter 6. Strength scale behavior of nanoporous Ag, Pd, Cu foams...........82
6.1 Introduction.......................................................................................................82
vii

6.2 Experimental details..........................................................................................83
6.3 Result and discussions......................................................................................83
6.4 Summary...........................................................................................................92
Chapter 7. Conclusions.......................................................................................93
Bibliography...............................................................................................................96
Appendices...............................................................................................................104
Appendix A: Summary of Ti samples...................................................................104
Appendix B: Summary of Cu samples..................................................................106
Appendix C: Summary of Ag samples.................................................................108
Appendix D: Summary of Pd samples..................................................................109
Appendix E: Summary of Ni samples...................................................................110
Appendix F: Summary of Zn samples..................................................................111

viii

List of Tables

Table 1: Electrochemical series of metals [58] ............................................................ 8
Table 2: Summary of Au samples .............................................................................. 34
Table 3: List of initial alloy systems used in this study ............................................. 38
Table 4: Nanoindentation results ............................................................................... 63
Table 5: Ligament sizes of NP Cu foams after heat treatments in vacuum ............... 67
Table 6: Ligament size results for NP Ag foams ....................................................... 78
Table 7: Foam hardness and ligament strength table ................................................. 92
ix

List of Figures

Figure 1: Cross-sectional SEM images of (a) macroporous Ni foam [43], and (b) NP
Au foam. ....................................................................................................................... 6

Figure 2: Schematic illustration shows the current/potential behavior of an alloy
undergoing selective dissolution. The dashed vertical lines indicate typical
ambiguity in defining a critical potential [93].............................................................. 9

Figure 3: SEM plan views of (a) amorphous Pt
0.1
Si
0.9
(0 V bias) prior to dealloying,
(b) the resulting NP Pt isotropic open-cell foam, (c) amorphous Pt
0.34
Si
0.66
(0 V bias)
prior to dealloying, (d) the resulting preferential dealloying along columnar
boundaries, and (e) amorphous Pt
0.33
Si
0.67
alloy (0 V bias) prior to dealloying with (f)
preferential dealloying around pre-existing features [3]. ........................................... 11

Figure 4: Representative SEM images of a 25% relative density NP Au foams
synthesized using (a) free corrosion and (b) electrochemically driven dealloying [44].
.................................................................................................................................... 12

Figure 5: Cross-sectional SEM images of a dealloyed 30% Au foam by free
corrosion (a) before heat treatment, and after annealing at (b) 400 °C and (c) 600 °C
[10]. ............................................................................................................................ 13

Figure 6: Raman spectra for chemically prepared (a) Ag
2
O and (b) AgO powders
[26]. ............................................................................................................................ 14

Figure 7: Experimental values for foam yield stress for all samples, normalized by
the yield stress for fully dense Au. The solid line presents the Gibson and Ashby
prediction for a gold foam [41]. ................................................................................. 17

Figure 8: Schematic compressive stress-strain curves of (a) a elastomeric foam and
(b) a elastic-plastic foam [29]. ................................................................................... 18

Figure 9: Schematics of (a) open-cell foam model and three deformation modes (b)
linear elasticity, (c) elastic buckling, and (d) plastic collapse [29]. ........................... 19

x

Figure 10: (a) Single beam with two fixed ends and (b) unit-cell beam model account
for surface effects. Also the prediction of beam deflections of two beam models was
presented using (c) t=10 nm (d) 1 µm [112]. ............................................................. 21

Figure 11: The critical average stresses of ligament buckling predicted by the Euler–
Bernoulli model and Timoshenko model with respect to the ligament size. [112].... 22

Figure 12: Deformation of a ligament is modeled as the plastic deformation of a
cantilever. Plastic deformation is confined to a plastic hinge and the deformation
causes a lattice rotation through an angle θ [21] ........................................................ 24

Figure 13: Relationship of ligament size to ligament yield stress for nanoporous gold
foams obtained by nanoindentation and by column microcompression testing [41] . 25

Figure 14: Illustration of the electrochemical setup for nanoporous Au and the
actuation result [108].................................................................................................. 27

Figure 15: Illustration of surface-chemistry-driven actuation in nanoporous gold.
Nanoporous Au sample mounted in a viscous-flow reactor. Adsorbate-induced
dimensional changes of the nanoporous Au sample are measured by dilatometry
[10]. ............................................................................................................................ 28

Figure 16: The time dependence of the CO
2
concentration at the reactor outlet
reveals the activation phase and long-term activity of a nanoporous gold catalyst (at
50 °C). Inset: schematic illustration of the reactor [120]. .......................................... 29

Figure 17: SERS spectra (632.8 nm excitation) of a 10
-6
M crystal violet/methanol
solution as a function of the annealing temperature of np-Au. The methanol-related
peak is labeled ‘‘MeOH’’ [9]. .................................................................................... 30

Figure 18: Characteristic polarization curves of Au and Au
30
Ag
70
in 70% nitric acid,
performed at USC. ..................................................................................................... 32

Figure 19: Cross-sectional SEM images of a 30% relative density Au foams using (a)
free corrosion and (b) electrochemically driven dealloying ...................................... 33

xi

Figure 20: NP Au samples before and after dealloying. ............................................ 33

Figure 21: EDX data from the center point of NP Au cross section .......................... 34

Figure 22: Phase diagrams of (a) Cu/Al and (b) Ag/Al systems [66]. ....................... 36

Figure 23: Potential-pH equilibrium diagram for (a) Cu-water and (b) Al-water
systems at 25 °C [75]. ................................................................................................ 37

Figure 24: Arc melter used in present study, located at UCSD Nano Engineering
department. ................................................................................................................. 39

Figure 25: Schematic of the set up for electrochemically driven dealloying ............. 40

Figure 26: SEM images (a) Top view (b) cross-sectional view for Ti
35
Al
65
dealloyed
in 1M NaOH. .............................................................................................................. 42

Figure 27: (a) Top view (b) cross-sectional SEM images for Ni
30
Mn
70
dealloyed in
1M (NH
4
)
2
SO
4
............................................................................................................ 42

Figure 28: High vacuum tube furnace with flowing argon setup at USC. ................. 43

Figure 29: (a) Series of load vs. displacement curves collected on the surface of
nanoporous Au using a Berkovich tip, and the (b) SPM and (c) SEM images of a
residual Berkovich indent [7]. .................................................................................... 46

Figure 30: Top view SEM image of Cu
30
Al
70
base alloy, which contains a Cu
30
Al
70

solid phase and a CuAl
2
phase (as indicated). ............................................................ 49

Figure 31: Cross sectional SEM images of a 20% relative density nanoporous copper
foam dealloyed from Cu
20
Zn
80
(a) as prepared and after heat treatments at (b) 200°C
(c) 400°C (d) 600°C. .................................................................................................. 53

xii

Figure 32: Cross sectional SEM images of a 30% relative density nanoporous copper
foam dealloyed from Cu
30
Al
70
(a) as prepared and after heat treatments at (b) 200°C
(c) 400°C (d) 600°C. .................................................................................................. 54

Figure 33: Cross sectional SEM images of a 35% relative density nanoporous copper
foam dealloyed from Cu
35
Zn
65
(a) as prepared and after heat treatments at (b) 200°C
(c) 400°C (d) 600°C ................................................................................................... 54

Figure 34: XRD results for (a) 20% (b) 30% (c) 35% relative density nanoporous
copper foams as prepared (AP) and after 200, 400, and 600°C heat treatments. ...... 59

Figure 35: Raman spectra for (a) 20% (b) 30% (c) 35% relative density of
nanoporous copper foams as prepared (AP) and after 200, 400, and 600°C heat
treatments. .................................................................................................................. 60

Figure 36: Nanoindentation tests of a 30% relative density copper foam after a 200°C
heat treatment (a) representative loading/unloading curves (b) SEM images of typical
nanoindentation array (c) increased magnification of a single indent ....................... 62

Figure 37: XRD results for (a) 20% (b) 30% relative density nanoporous copper
foams as prepared (AP) and after 200, 400, and 600°C heat treatments in vacuum.
The results for (c) 35% relative density as prepared nanoporous copper foams and
after 400°C heat treatments in vacuum were also presented...................................... 65

Figure 38: Cross sectional SEM images of a 20% relative density nanoporous copper
foam dealloyed from Cu
20
Zn
80
(a) as prepared and after heat treatments in vacuum at
(b) 200°C (c) 400°C (d) 600°C. ................................................................................. 66

Figure 39: Cross sectional SEM images of a 30% relative density nanoporous copper
foam dealloyed from Cu
30
Al
70
(a) as prepared and after heat treatments in vacuum at
(b) 200°C (c) 400°C (d) 600°C. ................................................................................. 66

Figure 40: Cross sectional SEM images of a 35% relative density nanoporous copper
foam dealloyed from Cu
35
Zn
65
(a) as prepared and after heat treatments at (b) 400°C.
.................................................................................................................................... 67

xiii

Figure 41: Top view SEM image of Ag
25
Al
75
base alloy, which contains an Ag
25
Al
75

solid phase (flat area) and an Ag
2
Al phase (as indicated). ......................................... 71

Figure 42: A photo of NP Ag samples before and after dealloying. .......................... 72

Figure 43: Cross-sectional SEM images and representative distribution of ligament
size for 25% relative density nanoporous silver foams dealloyed from Ag
25
Al
75
,
using different electrolytes: (a) (c) 5 wt% HCl (sample 1) and (b) (d) 1 M NaOH
(sample 2). .................................................................................................................. 75

Figure 44: Cross-sectional SEM images of 25% relative density nanoporous silver
foams dealloyed using 5 wt% HCl (sample 1) and heat treated under vacuum at (a)
200 °C (b) 400 °C (c) 600 °C and under flowing argon at (d) 200 °C (e) 400 °C (f)
600 °C. ....................................................................................................................... 76

Figure 45: Cross-sectional SEM images of 25% relative density nanoporous silver
foams dealloyed using 1 M NaOH (sample 2) and heat treated under vacuum at (a)
200 °C (b) 400 °C (c) 600 °C and under flowing argon at (d) 200 °C (e) 400 °C (f)
600 °C. ....................................................................................................................... 77

Figure 46: Raman spectra and XRD result for the foams dealloyed using 5wt% HCl ,
sample 1 (a and b, respectively) and dealloyed using 1M NaOH, sample 2 (c and d,
respectively) as prepared (AP) and after 200, 400, and 600 °C heat treatments under
flowing argon. ............................................................................................................ 79

Figure 47: Cross-sectional SEM images of nanoporous foam synthesized in this
study. Cu foams dealloyed from (a) a Cu
20
Zn
80
alloy, (b) a Cu
30
Al
70
alloy, (c) a
Cu
35
Zn
65
alloy. Ag foams dealloyed from Ag
25
Al
75
alloy using two different
electrolytes (d) 5wt% HCl and (e) 1M NaOH and (f) a Pd foam dealloyed from a
Pd
30
Al
70
alloy. ............................................................................................................ 85

Figure 48: Representative distribution of ligament size of the (a-c) Cu, (d-e) Ag and
(f) Pd foams. ............................................................................................................... 85

Figure 49: XRD results of nanoporous Cu, Pd and Ag foams ................................... 86

xiv

Figure 50: Representative SEM images of residual impressions on nanoporous (a)
Cu, (b) Ag performed by nanoindenation and on a (c) nanoporus Pd foam performed
by Vickers indentation. .............................................................................................. 87

Figure 51: Foam hardness values vs. ligament size for Au, Ag, Cu, and Pd
nanoporous foams. Predictions of modified elastic buckling models and plastic
collapse models are also shown. ................................................................................ 88

xv

Abstract

Nanoporous (NP) metal foams, including Au, Cu, Ag, and Pd, have been
successfully synthesized using different processes, such as free corrosion,
electrochemically driven dealloying, and subsequent heat treatments. To investigate
the effect of the synthesis processes on the yield strength and microstructure, the
hardness of NP Au, Cu, Ag, and Pd foams with ligament sizes ranging from 20-220
nm was measured using nanoindentation and micro Vickers hardness testing, while
the morphology and ligament sizes were determined by scanning electron
microscopy (SEM). The results demonstrate that for all these materials, the yield
strength of the NP foams was higher than that predicted by macroporous foam scale
equations. In addition, combining the results with previous data for NP Au, it was
observed that the yield strength of foams increases as ligament size decreases, which
emphasizes the significant effect of the ligament size. The results were also
compared to several predictions for foam deformation models, including elastic
buckling and plastic collapse models. Overall, the deformation mode, regardless of
the type of metal, seems to more closely follow the plastic collapse models.
Deviations from the models are attributed to irregular cell geometry and relative
density estimations.
Additionally, the effect of oxide formation in relation to the NP foam structure
for Cu and Ag was investigated as a function of heat treatments. The formation of
oxide was characterized using X-ray diffraction (XRD) and Raman spectroscopy.
xvi

The variations in ligament size, morphology and yield strength were investigated for
all heat treated foams. For NP Cu foams, the results demonstrate that the formation
of oxides was observed after any heat treatments above 200 °C under argon. These
Cu oxides initiated at the surface layer and can grow in thickness with increasing
heat treatment temperature, leading to a complete oxide foam. It was determined
that the yield strength of Cu oxide foams increased ~ 40 times upon heat treatments,
as compared to Cu foams. For the NP Ag foams, oxidation was observed after heat
treatments in argon, but was limited to only a surface layer. This thin oxide layer
decomposed into silver and oxygen at 600 °C. Furthermore, NP Ag foams appear to
show relatively high thermal stability compared to other metal foams. The heat
treatments on NP Ag foams were not as effective at changing the ligament size as the
use of different electrolytes. However, the results demonstrated that both Cu and Ag
foams retain their structural integrity even after oxidation.

1

Chapter 1. Introduction
Many novel properties have been observed for materials with the microstructure
or morphology of the materials having characteristic sizes at the nano scale, such as
nanocrystalline, nanotwinned, and nanowire materials [14, 27, 45, 65, 69, 82, 84, 90,
111, 115]. Among the materials with nanoscale features, NP metal foams have
attracted considerable attention due to their high surface to volume ratio and the
potential to be high strength, lightweight functional materials [4, 7, 10, 11, 17, 19,
36, 38, 39, 50, 56, 67, 79, 105, 108, 109, 120]. Thus, NP foams present a new field
of research and applications in comparison to macroporous foams [2, 29, 43, 100].
However, the study of NP foams is very challenging since the synthesis of uniform
inter-connected nano-sized ligaments is highly dependent on the materials and the
processing techniques; and therefore, a good understanding of the processing
methods is very important. Among the NP foam synthesis techniques, dealloying is
one of most promising methods to synthesize NP metal foams with open porosity
[12, 13, 18, 20, 22, 33, 34, 47, 57, 60, 78, 79, 88, 95, 98, 116-118]. The main
advantage of this method is due to its simplicity and consistency of synthesizing
uniform NP structures.
Prior studies based on NP Au foams show that the ligament and pore sizes of
the foams can be controlled, depending on the dealloying techniques, time,
electrolytes and post heat treatments [13, 23, 31, 33, 35, 46, 61, 62, 94, 95, 98, 104].
Therefore, one has to consider the role of surface diffusion, percolation, dealloying
threshold, part limit, and standard reduction potential of each metal element involved
2

during the dealloying process [58, 72, 73, 75, 76, 91, 104, 116]. However, all these
parameters are often coupled with the pH values of electrolytes, anode and cathode
reactions, applied potential and oxidized state for each metal element, especially for
the metals which have native oxides. Consequently, due to the existence of
numerous parameters that affect the properties of the NP foams, the processing of
other (less noble) materials using dealloying techniques is still mostly based on trial
and error.
A comprehensive study which correlates dealloying conditions and foam
morphology to yield strength and ligament size is presented here for nanoporous Cu,
Ag, and Pd foams with 65-80% porosity. These metals were chosen due to their
relative high standard reduction potentials [58], whereby the basic requirement for
the dealloying technique of having large potential differences between alloy
elements can be met. Cu/Zn, Cu/Al, Ag/Al, and Pd/Al alloys were used as the
precursor alloys in this project. In addition, the effect of oxide on the NP foams can
be verified by heat treating NP Cu and Ag foams, since both Cu and Ag have stable
oxides [1, 28, 30, 48, 52, 53, 63, 74, 81, 85, 86, 89, 107, 114, 119]. To date, the
formation of oxides on NP foams is still not well understood and further research is
essential since at the nanoscale, an oxide layer can become a large portion of overall
ligament. In order to take into account the effect of oxides, the morphology and
oxide content of the foams were characterized by SEM, XRD, and Raman
spectroscopy. By studying the effects of oxides on the ligaments and its relation to
3

the ligament strength, the controlling parameters for the overall NP foam properties
in non-noble metal foams can be identified.
With the goal of investigating the effects of characteristic size, such as ligament
and pore size, on the foam properties, the foam yield strength as a function of
ligament size was studied. To date, significant progress has been made toward
understanding the mechanical properties of NP foams using nanoindentation, with
emphasis in NP Au foams [41, 87, 106]. In contrast, for non-noble metals, which
have different and more complex oxidized states, only a few studies have been
performed focusing on the foam strength [40, 78]. Nanoindentation is a promising
technique to investigate the NP foam hardness, which is related to the foam yield
strength; therefore, it is a valuable method for nanostructured materials. The aim is
to build a bridge between microstructure, ligament size, and mechanical properties,
with the purpose of predicting material behavior from NP foams.
As the next step, the yield strength results that were obtained were incorporated
into current deformation models. In general, two failure modes have been reported
on open-cell foams: elastic buckling and plastic collapse [24, 29]. However, the
dominant failure mechanism of NP foams is not yet understood. Therefore, it is
essential to investigate the deformation behavior of NP foams, specifically in the
context of cell geometry, node and surface effects. To this extent, several
predictions from the modified deformation models, including elastic buckling and
plastic collapse models [21, 41, 112], were compared to ligament size data.
Therefore, a comprehensive study of the mechanical behavior of NP foams,
4

including their respective deformation models, is presented for a wide range of
systems.

5

Chapter 2. Background
2.1 Cellular solids
Foams have special physical, mechanical and thermal properties, which allow
many new applications as compared to fully dense solids. For example, foams can
be used for the design of light, stiff components such as sandwich due to their low
densities, while their low stiffness is good for cushioning and energy absorption
applications [29]. Furthermore, different foam morphologies can be used for
different applications. Open-cell foams are permeable and can have very high
specific areas, in contrast to closed cell foams, which provide good mechanical
properties and are mostly used in structural, load-bearing applications [7, 25, 41, 43,
100, 103, 106].
This thesis focuses on the open cell porous metal foams, especially with pore
and ligament sizes at the nanoscale, typically less than 100 nm. In Figure 1, a
comparison between macro and nanoporous foams is shown in the cross-sectional
SEM images of a macroporous Ni foam with ligament sizes around 500 µm (Fig. 1a)
and a nanoporous Au foam with ligament sizes around 50 nm (Fig. 1b). Porous
metal foams have unique properties as compared to polymer and ceramic foams [54].
For example, the mechanical strength, thermal stability, and electrical conductivities
of metal foams are higher than those of polymer foams. Meanwhile, they can deform
plastically and absorb energy as opposed to ceramic foams. Metal NP foams which
6

have a high specific surface area allow for applications in electrodes, catalyst,
sensors, actuators and filtration [9, 10, 50, 109].

Figure 1: Cross-sectional SEM images of (a) macroporous Ni foam [43], and (b)
NP Au foam.

Among the NP foam studies, nanoporous Au has been used as model system
due to several reasons. First, the ligament size is very uniform and can be
manipulated from 10 to 1000 nm without changing the relative density or relative
geometry of the material. Second, monolithic NP Au foams can be simply
synthesized by dealloying Au/Ag alloys. Third, Au does not form an oxide layer so
that the post annealed process can be applied to coarsen the ligament sizes.
Furthermore, NP Au has recently attracted considerable interest fueled by its
potential use in actuator, sensor, and catalyst applications [8]. The following
sections present a summary of the foam processing with emphasis on dealloying
techniques.

7

2.2 Foam processing
Macroporous metal foams, which have pore sizes larger than 1 µm, can be
synthesized using either liquid or solid state processing [29]. For instance,
aluminum foams can be made by the mechanical agitation of a mixture of liquid
aluminum and silicon carbide particles. Using electroless deposition,
electrochemical deposition or chemical vapor deposition, a metallic layer onto an
open-cell polymer foam substrate can be coated to give metal foams. Solid state
processes usually use the powder sintering method, which mixes the powdered metal
with a spacing agent during sintering. However, these techniques are limited to the
synthesis of macroporous foams. Recently, many researchers have been focused on
the synthesis of NP foams [33, 42, 51, 88, 99]. Among them, dealloying techniques
were proven to be a processing tool for producing bicontinuous nanoporosity with
open pores in three dimensions [23].
2.2.1 Dealloying
Dealloying is defined as the selective dissolution of one or more components
from a metallic alloy [22]. Typically, the less noble components are removed, and
the more noble components remain behind. In such cases, there must be a significant
difference in the reversible metal/metal ion potential of the metal in the alloy (Table
1), so that the less noble component is soluble in its oxidized state, and the more
noble, nonoxidized component is free to diffuse along the alloy/electrolyte interface.
The most well-known system which meets the criteria is Ag/Au [44]. Au is more
electrochemically noble and inert in electrolytes than Ag, and therefore the
8

dissolution current is solely due to Ag dissolution. Additionally, Au does not readily
form complexes which would prevent surface diffusion and inhibit the formation of a
porous structure [44, 87, 96, 101].
Table 1: Electrochemical series of metals [58]
Materials Half reaction Potential E° (V)
Gold Au
3+
+ 3 e
−
⇄ Au(s) +1.53
Platinum Pt
2+
+ 2 e
−
⇄ Pt(s) +1.188
Palladium Pd
2+
+ 2 e
−
⇄ Pd(s) +0.915
Silver Ag
+
+ e
−
⇄ Ag(s) +0.7996
Copper Cu
2+
+ 2 e
−
⇄ Cu(s) +0.340
Hydrogen 2 H
+
+ 2 e
−
⇄ H
2
(g) 0
Nickel Ni
2+
+ 2 e
−
⇄ Ni(s) −0.25
Iron Fe
2+
+ 2 e
−
⇄ Fe(s) -0.44
Chromium Cr
3+
+ 3 e
−
⇄ Cr(s) -0.74
Zinc Zn
2+
+ 2 e
−
⇄ Zn(s) -0.7618
Manganese Mn
2+
+ 2 e
−
⇄ Mn(s) -1.185
Titanium Ti
2+
+ 2 e
−
⇄ Ti(s) -1.63
Aluminum Al
3+
+ 3 e
−
⇄ Al(s) -1.66

The morphological evolution of dealloying has been extensively studied, and
many models have been proposed to explain the dealloyed structure as well as the
dealloying mechanism, critical potential, and part limit [23, 72, 73]. Pickering
performed a comprehensive study of polarization behavior of various alloys [72, 73].
9

The critical potential is defined as the potential above which there is selective
dissolution of one or more elements in the alloy, leading to the formation of a porous
structure. A schematic illustration of a polarization curve is presented in Figure 2,
which indicates the range of typical critical potentials. The critical potential
corresponds to that associated with the knee in the curve and is not sharply defined.
The shape of the knee is affected by sweep rate, alloy, and electrolyte composition
[93].

Figure 2: Schematic illustration shows the current/potential behavior of an alloy
undergoing selective dissolution. The dashed vertical lines indicate typical
ambiguity in defining a critical potential [93].

10

The parting limit, or dealloying threshold, is the critical content of the reactive
alloy component above which dealloying can occur. Dealloying threshold can be
explained with the use of percolation theory. The basis of percolation theory is that,
in order to remove all of the A atoms from an A
x
B
1-x
alloy, there must exist a
continuous connected path of A atoms throughout the structure. This provides a
continuous active path for the dissolution of A atoms to occur, as well as for the
penetration of the electrolyte into the structure. The model developed by Sieradzki
et al in 1989 incorporated both percolation and surface diffusion and arrived at
dealloying thresholds and morphologies similar to those observed experimentally
[92]. Other models were later developed to explain the porous morphology which
involve the injection of regions of negative curvature into the surface as the pores
form [91], and spinodal decomposition of the alloy at the solid-electrolyte interface
during dealloying [22].
In addition to the starting alloy composition, studies show that the
microstructure of the precursor alloys also have an effect on the morphology of the
nanoporous foams [3, 13, 103]. Figure 3 presents nanoporous Pt foams synthesized
using various starting alloys. The observed irregular geometry of the Pt foams,
synthesized using the different pretreatments of the starting alloys, can lead to
different foam properties and applications.

11

Figure 3: SEM plan views of (a) amorphous Pt
0.1
Si
0.9
(0 V bias) prior to
dealloying, (b) the resulting NP Pt isotropic open-cell foam, (c) amorphous
Pt
0.34
Si
0.66
(0 V bias) prior to dealloying, (d) the resulting preferential dealloying
along columnar boundaries, and (e) amorphous Pt
0.33
Si
0.67
alloy (0 V bias) prior
to dealloying with (f) preferential dealloying around pre-existing features [3].

2.2.2 Free corrosion and electrochemically driven dealloying
There are two methods of dealloying by which a number of nanoporous metal
foams have been successfully synthesized the precursor alloys: free corrosion and
electrochemically driven dealloying [31, 44, 57, 61, 77, 116]. Free corrosion is
simply placing the alloys into suitable electrolytes, based on their st
andard reduction potentials, and waiting for the chemical reactions to be completed.
In most cases, the less noble elements of the alloys are oxidized into metal ions and
dissolved in the solution, accompanied with hydrogen ions being reduced into
hydrogen bubbles. The reaction is completed once there are no bubbles inside the
12

solution. For electrochemically driven dealloying, an external potential is applied to
the precursor alloys with the assistance of a potentiostat. This applied potential can
enhance or inhibit the corrosion rate, depending on whether the applied potential is
above or below the critical potential. For electrochemically driven dealloying, the
applied potential should always be higher than the critical potential, therefore leading
to higher corrosion rates and shorter dealloying times. Foams synthesized by the
electrochemically driven method typically have smaller pore sizes compared to those
processed by free corrosion by a factor of about three, due to the higher over-
potential [41]. Figure 4 shows the representative cross-sectional SEM images of a
25% relative density NP Au foam using free corrosion in 67-70% HNO
3
(Fig. 4a),
and using electrochemically driven dealloying with an applied potential of 600 mV
in 1 mole HNO
3
and 0.01 mole AgNO
3
(Fig. 4b)

[44]. The average ligament size of
the free corrosion foam and electrochemically driven foam is 5 and 30 nm,
respectively.

Figure 4: Representative SEM images of a 25% relative density NP Au foams
synthesized using (a) free corrosion and (b) electrochemically driven dealloying
[44].
13

2.2.3 Heat treatments
In addition to the different dealloying techniques, heat treatments can also be
used to change the ligament size. Ligament sizes ranging from 50 to 1000 nm can be
achieved by various heat treatment temperatures, in which higher post-dealloying
temperature results in larger ligaments. Figure 5 shows the different ligament sizes
of a 30% relative density NP Au foam dealloyed by free corrosion before and after
annealing at different temperatures. Au does not oxidize in air, allowing the pore
size to be varied by a simple furnace anneal. At elevated temperatures, the
pore/ligament dimensions increase due to Ostwald ripening [102]. However, the
heat treatments used to change the ligament sizes of non-noble foams should be
conducted under flowing argon or vacuum, especially the oxide formation can make
it very challenging to predict the ligament coarsening and structure stability.
Furthermore, the formation of an oxide layer would have a significant effect on the
overall foam properties since the ligament is at nano scale.

Figure 5: Cross-sectional SEM images of a dealloyed 30% Au foam by free
corrosion (a) before heat treatment, and after annealing at (b) 400 °C and (c)
600 °C [10].
14

For those metal foams which have a native oxide layer, the oxide content of the
foams should be carefully characterized. For example, the silver–oxygen system
(Ag–O) contains several defined compounds, including Ag
2
O, AgO, Ag
3
O
4
and
Ag
2
O
3
[1, 64, 70, 81, 85, 107]. Among these oxides, the most thermodynamically
stable is Ag
2
O, with AgO existing at low temperatures and high oxygen partial
pressures. Using XRD, Raman, and FT-IR spectroscopy, the spectra-structure
relationship of AgO and Ag
2
O have been established [107]. Figure 6 shows Raman
spectra for chemically prepared (a) Ag
2
O and (b) AgO powders. On the basis of this
information, the oxide content of the as-prepared and post-annealed NP foams can be
monitored.

Figure 6: Raman spectra for chemically prepared (a) Ag
2
O and (b) AgO
powders [26].

15

Although oxide formation may add to the complexity of NP foams, their
existence is not necessary detrimental. The Ag–O system has drawn considerable
attention due to the use of silver as a catalyst and sensor for ethylene oxidation and
carbon monoxide detection [1]. Silver oxide is also utilized as cathodes in silver
oxide/zinc alkaline batteries. Moreover, Silver oxide thin films have been
investigated for their potential use in photovoltaic applications, the synthesis of
superconductors, surface enhanced Raman spectroscopy (SERS), and super
resolution near field structures (super-RENS) in optical memories. They have also
been applied in plasmon photonic devices [1]. A number of applications also can be
found for Cu oxides [53, 114, 119], and the NP metal oxide foams have the potential
to be a lightweight functional material.

2.3 Foam mechanical properties
It is essential to fully understand the mechanical behavior of foams before they
can be used efficiently. The mechanics of macro foams have been extensively
studied by Gibson and Ashby [29]. One of the most important factors in determining
the foam behavior is by using the foam’s relative density, ρ*/ρ
s
(the density of the
cellular material, ρ*, divided by that of the solid from which the cell walls are made,
ρ
s
). In addition, the degree to which the cells are open or closed, and their shape
anisotropy ratio R
12
and R
13
are also important. The crucial cell-wall properties are
the solid density, ρ
s
, the Young’s modulus, E
s
, and the yield strength, σ
s
. Foam
properties are analyzed in terms of these parameters and are compared with
16

experimental data to give equations suitable for design. The Gibson-Ashby foam
scaling equations for the two most often reported mechanical properties are:
𝐸 ∗
= 𝐶 �
𝐸 �
�
�
∗
�
�
�
�
(1)
𝜎 ∗
= 𝐶 �
𝜎 �
�
�
∗
�
�
�
� � ⁄
(2)
where * denotes foam properties and s denotes solid properties. C
1
and C
2
include
all of the geometric constants of proportionality.
It is expected that nanoporous foams would behave differently from
macroporous foams. A comprehensive study on the relationship between yield
strength, relative density and ligament sizes was presented for nanoporous Au foams
[41]. The ligament sizes ranging from 10-900 nm were controlled by dealloying
techniques and subsequent heat treatments. A summary of NP Au foam yield
strengths, which were converted from the hardness values measured by
nanoindentation, and the Gibson and Ashby prediction for yield strength (Eq. 2) are
shown in Figure 7. The main issue regarding foam indentation is an assumption on
the relation between indentation hardness and the yield stress. In the case of plastic
indentation, it has been demonstrated that for foams with relative density less than
30%, the hardness (H) is equal to the yield strength (σ
y
), rather than H ≅ 3σ
y
which
characterizes a fully dense material [29]. In the case of porous metals, deformation
under the indenter is not constrained by the surrounding material due to
densification. Thus the indentation test acts like a uniaxial compression test, and the
17

yield strength σ is simply equal to the measured hardness [7]. It was observed that at
the nanoscale, the strength was not only controlled by the relative density, but also
by the ligament length scale. Meanwhile, the high yield strength of the Au nano
ligament is comparable to the ideal shear strength of Au (5 GPa) in the absence of
dislocations [6]. Several predictions from the deformation models accounting for
surface and size effects were compared with the experimental data to explain the
unexpected high ligament strengths.

Figure 7: Experimental values for foam yield stress for all samples, normalized
by the yield stress for fully dense Au. The solid line presents the Gibson and
Ashby prediction for a gold foam [41].

18

2.3.1 Deformation of foams
In general, the compressive deformation of macroporous open-cell foams
contains three regimes: linear elasticity, collapse plateau, and densification [29].
Figure 8 shows typical schematic stress-strain curves for the foams. Two
deformation mechanisms can be observed when the stress exerted on the foams
exceeds the linear elastic regime: plastic yielding and elastic buckling. As
macroporous open-cell foams can be modeled as inter-connected beams, Figure 9 (a)
shows a schematic of the model with a beam of length l and square cross-section of
side t [29]. Following this model, the linear elasticity can be attributed to the
bending of the beams, as shown in Figure 9 (b). The relation between the Young’s
modulus for the foam and for the bulk material was presented in section 2.3,
Equation (1).

Figure 8: Schematic compressive stress-strain curves of (a) a elastomeric foam
and (b) a elastic-plastic foam [29].

19

Figure 9: Schematics of (a) open-cell foam model and three deformation modes
(b) linear elasticity, (c) elastic buckling, and (d) plastic collapse [29].

Figure 9 (c) and (d) shows the schematics of elastic buckling and plastic
collapse, respectively. If the load exerted on the beams makes them move in a
nonparallel direction relative to the force, they will buckle and initiate the elastic
20

collapse of the foam. On the other hand, plastic collapse occurs when the load
exerted on the cell edges exceeds its fully plastic moment [29]. However, given that
many novel properties of NP foams have been attributed to the characteristic sizes of
the ligaments and pores [21, 41, 112], whether these models can still apply to NP
foams is still a topic of debate. Many factors such as surface tension, surface
elasticity, strain gradient, and ligament size, need to be taken into account [8, 21, 25,
37, 41, 67, 106, 112].
For example, the models by Xia et al address the elastic modulus and critical
buckling behavior of nanoporous materials using the theory of surface elasticity,
combined with two beam models: the Euler-Bernoulli beam model, which covers the
case of small deflections of a beam which is subjected to lateral loads only, and the
Timoshenko beam model, which accounts for shear deformation [112]. Figure 10 (a)
presents a Timoshenko model with a surface layer thickness t
1
by considering the
effects of surface stresses and surface elasticity. In the case of small deflection with
two fixed ends, the nanobeam can be affected by the residual surface tension as a
distributed transverse load. Figure 10 (c) and (d) illustrate the predictions of two
beam models accounting for surface effects subjected to a transverse concentrated
force at the midpoint, where the length to width ratio is 3 and t = 10 nm and 1 µm.
Results show that smaller characteristic sizes and aspect ratios would lead to greater
influences of surface and shear deformation effects. Using the same models, the
critical average stress for the elastic collapse of the NP unit cell is given by the
modified equations 3 and 4.
21

Figure 10: (a) Single beam with two fixed ends and (b) unit-cell beam model
account for surface effects. Also the prediction of beam deflections of two beam
models was presented using (c) t=10 nm (d) 1 µm [112].

𝜎 ��
∗
=
�
�
�
�
� � �
�
( �� )
∗
( � � � )
�
�
�
(Euler model) (3)
𝜎 ��
∗
=
�
( � � � )
�
�
� �
�
�( �� )
∗
� �
�
� � �
�
( �� )
∗
+ 𝐻 �
� (Timoshenko model) (4)
where (EI)* and V are functions of t, and l and t are the length and width of the
beam, respectively. For these models, (t/l)
2
is determined by the relative density,
while the width of the beam, t , is taken to be the ligament size [112]. The material
parameters of single-crystalline bulk Au are taken as follows: bulk Young’s Modulus
22

E
0
= 70 GPa, shear modulus G = 27.2 GPa, surface Young’s modulus E
s
= 3.63
Nm
-1
, residual surface stress τ
0
= 1.4 Nm
-1
, and the (t/l)
2
is determined by the relative
density ρ/ρ
0
= 0.3 taken as an example. Figure 11 presents the critical average stress
of ligament buckling predicted by both beam models as a function of ligament size,
which shows a stronger ligament size dependence for the Euler’s model as the
ligament size falls into the range of 10 nanometers.

Figure 11: The critical average stresses of ligament buckling predicted by the
Euler–Bernoulli model and Timoshenko model with respect to the ligament size.
[112]

23

Another mechanism that has been proposed is that during plastic deformation of
the foams, the struts deformation can be model as the bending of cantilevers. Dou et
al [21] stated that the lattice near the plastic hinge region undergoes rotation at a
constant curvature through an angle θ, as shown in Figure 12. It is assumed that
such rotations result in a uniform hardening in both the compressive and tensile sites
of the curved region [21]. It is also assumed that, for a nano ligament with a
diameter below 100 nm, dislocations are most likely encountering a free surface
instead of interacting with other mobile dislocations during plastic deformation.
Based on these assumptions, a modified plastic collapse model accounting for the
effect of plastic strain gradients in the deformation hinges was established [21]:
𝜎 �
= 𝜎 �
+ 𝑘 �
𝜇 �
�
�
(5)
where σ
1
is the yield stress of the ligament in the absence of any strain gradient, k
3

incorporates Taylor’s constant and all geometrical or shape terms consistent with the
ligaments in the nanoporous structure, µ is shear modulus, b is Burgers vector, and t
is the strut width [21]. The ligament strength can be predicted by the equation using
the following data: shear modulus µ = 26.0 GPa, Burgers vector b = 0.2885 nm, and
the optimized values of k
3
= 0.473 and σ
1
= 0. This model should be equally
applicable to all nanoporous materials that deform through a similar mechanism,
although it was firstly developed for the nanoporous Au.
24

Figure 12: Deformation of a ligament is modeled as the plastic deformation of a
cantilever. Plastic deformation is confined to a plastic hinge and the
deformation causes a lattice rotation through an angle θ [21].

Hodge et al [41] also presented an empirical plastic collapse model by fitting
the data in Figure 13. All ligament strength values were calculated from the
hardness data using Eq. 2. Therefore, a modified Gibson/Ashby equation which
incorporates a ligament size effect similar to a Hall-Petch-type effect is expressed as:
𝜎 ��
= � 𝜎 �
+ 𝑘 ��
∙ 𝐿 � � � ⁄
� (6)
where σ
0
is related to the bulk material yield strength and k
Au
is a material constant,
L is the ligament size. Results show that NP Au has a k value of 9821 MPa, which is
similar to that of nanocrystalline Au films with 26-60 nm grains [82].
25

Figure 13: Relationship of ligament size to ligament yield stress for nanoporous
gold foams obtained by nanoindentation and by column microcompression
testing [41]

To date, the majority of the work concerning mechanical behavior has been on
nanoporous Au foams; however, a comprehensive study of the ligament size effect
on a wide range of metal systems is still lacking. Given the geometry, aspect ratio,
and node effect of nanoporous foams, the failure mechanism of the foams is not well
understood. Whether the current deformation models developed for the open-cell
foams can still apply for the foams at the nano scale is still unclear. The surface
effects, including surface stress and surface elasticity, and the ligament size effect,
which might affect the dislocation motion, also need to be discussed.
26

2.4 Foam applications
For macroporous foams, there are four major areas of application: thermal
insulation, packaging, structural, and buoyancy [29]. As the ligament and pore sizes
decreased to the nano length scale, there are other areas of application which are
important and growing. Furthermore, nanoporous Au foam has been found to be a
promising material for actuation, sensing, green chemistry and energy harvesting [9,
10, 108, 110, 120]. The following presents a summary of the main current
applications for nanoporous foams.

Electrochemical actuator
Materials that can reversibly change their dimensions upon the application of an
external stimulus, such as an applied voltage, are used as actuators in many
applications. The best known examples are piezoelectric and electrostrictive
ceramics, but conducting polymers and carbon nanotubes have recently also been
proposed. However, voltage induced dimension changes of comparable magnitude
have not been reported in metals. Weissmuller et al [108] showed that reversible
strain amplitudes comparable to those of commercial piezoceramics can be induced
in metals by introducing a continuous network of nanometer-sized pores with a high
surface area and by controlling the surface electronic charge density through an
applied potential relative to an electrolyte impregnating the pores. The setup and the
results were displayed in Figure 14, showing that nanoporous gold can be an
effective actuator.
27

Figure 14: Illustration of the electrochemical setup for nanoporous Au and the
actuation result [108].

Chemical sensor/actuator
Studies by Biener et al [10] demonstrated that high-surface-area materials such
as nanoporous gold can be used as a surface chemistry-driven actuator. For example,
they showed reversible strain amplitudes by alternating exposure of nanoporous Au
to ozone and carbon monoxide. The effect can be explained by adsorbate-induced
changes of the surface stress, and can be used to convert chemical energy directly
into a mechanical response, thus making NP Au a candidate for surface-chemistry-
driven actuator and sensor technologies, as shown in Figure 15 [10].

potentiostate
electrolyte
counter
electrode
ref.
electrode
np-Au
displacement
sensor

28

Figure 15: Illustration of surface-chemistry-driven actuation in nanoporous
gold. Nanoporous Au sample mounted in a viscous-flow reactor. Adsorbate-
induced dimensional changes of the nanoporous Au sample are measured by
dilatometry [10].

Pollution control
NP Au with a sponge like morphology has a high catalytic activity for CO
oxidation at ambient pressures and temperatures down to -20 °C [120]. In addition,
NP Au has a better thermal stability than that of Au particles. Zielasek et al studied
the catalytic oxidation of CO with synthetic air in a continuous-flow reactor at
normal pressure in the temperature range -20 to 50°C [120]. The reactor is
illustrated schematically in the inset of Figure 16. An infrared gas analyzer (IRGA)
was used to quantify the amount of CO
2
produced during the reaction. In addition,
the reaction was followed by detecting the heat evolved during the exothermic
oxidation of CO, as the temperature difference between the catalyst and the reactor.
The amount of CO
2
detected by IRGA for a freshly prepared sample of nanoporous
gold is plotted versus time in Figure 16.

29

Figure 16: The time dependence of the CO
2
concentration at the reactor outlet
reveals the activation phase and long-term activity of a nanoporous gold
catalyst (at 50 °C). Inset: schematic illustration of the reactor [120].

Green chemistry
Nanoporous Au is a new catalyst with a stable structure that is active without any
support. It catalyzes the selective oxidative coupling of methanol to methyl formate
with high selectivity [110]. The proposed mechanism of selective oxidation of
methanol on Au surfaces is that Methanol is activated by surface oxygen and bonded
at the surface as methoxy. Subsequent deprotonation leads to the aldehyde. Fast
reaction of the highly reactive aldehyde with further methoxy leads to the coupling
product methyl formate (HCO
2
CH
3
). In the case of excess oxygen, the aldehyde can
be further oxidized, resulting in CO
2
formation. In terms of thermodynamics, the
total oxidation product (CO
2
) is strongly favored. Gold exhibits a remarkable
selectivity toward partial oxidation products, distinguishing it from other transition
metals [110].
30

Surface enhanced Raman scattering (SERS) sensor
The size effect can be used to fine-tune the optical properties of NP Au. For
example, SERS enhancement factors in the range of 10
9
–10
11
were achieved by
tuning the average pore width of NP Au from 10 to 250 nm. The correlation
between pore width and SERS response is shown in Figure 17. Nanoporous Ag and
Cu foams were also proven to be an active substrate for the SERS enhancement [11,
19, 38, 55, 79, 94].

Figure 17: SERS spectra (632.8 nm excitation) of a 10
-6
M crystal
violet/methanol solution as a function of the annealing temperature of np-Au.
The methanol-related peak is labeled ‘‘MeOH’’ [9].

31

Chapter 3. Experimental procedures and materials
This chapter presents the NP foam synthesis methods and the techniques for
analysis of the chemical reactions during dealloying. Characterization tools to study
the foam morphology, oxide content, and hardness are also presented.
3.1 Synthesis of NP Au foams
As discussed in the background, both free corrosion and electrochemically
driven dealloying can be used for the synthesis of NP foams. For electrochemically
driven dealloying, the first step in processing NP foams is setting up the potentiostat.
A three-electrode electrochemical cell controlled by a potentiostat (Gamry Reference
3000) was used for these experiments. The polarization scan of samples was exerted
using Pt wire as a counter electrode and Ag/AgCl as reference electrode, with the
assistance of VFP600 software. We used the polarization curves of Au and
Au
30
Ag
70
in 70% HNO
3
with the scan rate of 5 mV/sec to make sure the whole
electrochemical system was working correctly. Figure 18 shows the characteristic
polarization curves of both Au and Au
30
Ag
70
, which indicate the critical potential of
around 0.9 V. Above this critical potential, the current density of pure Au starts to
decrease and is ready to passive, while that of the Ag rich alloy still goes up and
leads to higher corrosion rate. Furthermore, using this setup with an applied
potential of 1.15 V, NP Au foams with average ligament size of 15 nm was
successfully synthesized by electrochemically driven dealloying Au
30
Ag
70
. The
result was also compared to the free corrosion Au foams, which were synthesized
placing Au
30
Ag
70
samples into 70% HNO
3
for 48 hours. Figure 19 shows the
32

Figure 18: Characteristic polarization curves of Au and Au
30
Ag
70
in 70% nitric
acid, performed at USC.

representative cross-sectional SEM images of both electrochemically driven and free
corrosion foams, which indicates the average ligament size of electrochemically
driven foam is about three times smaller than that of free corrosion foam. The
corrosion rate of electrochemically driven dealloying is much faster that it only took
3 hours to complete the reaction, while the dealloying time for free corrosion took 48
hours. A longer dealloying time allows the more noble atoms to diffuse, thus
coarsening the ligament. The relative density was calculated from the weight loss
during dealloying process, while the volume changes are negligible. Figure 20
shows a photo of NP Au samples before and after dealloying, while the EDX result
from the center point of the NP Au cross-sections (Figure 21) confirms that only Au
33

was remaining in the overall foam structure. Table 2 shows a summary of Au
samples being synthesized in this project.

Figure 19: Cross-sectional SEM images of a 30% relative density Au foams
using (a) free corrosion and (b) electrochemically driven dealloying

Figure 20: NP Au samples before and after dealloying.
34

Figure 21: EDX data from the center point of NP Au cross section

Table 2: Summary of Au samples
Sample
name
Electrolyte Applied
Potential
(mV)
Dealloying
Time (hr)
Foam wt%
after
dealloying
Ligament
size (nm)
AJ-7 70% HNO
3
N/A 49 52% 57
AJ-8 70% HNO
3
N/A 49 43% NA
AJ-9 70% HNO
3
N/A 49 43% NA
AJ-10 70% HNO
3
N/A 49 43% NA
AJ-11 70% HNO
3
N/A 48 44% NA
AJ-12 70% HNO
3
N/A 48 49% NA
IC-1 70% HNO
3
1050 24 51% 37
IC-2 70% HNO
3
1000 18 43% NA
IC-3 70% HNO
3
1150 3 43% 15
Precursor alloy is Au
30
Ag
70
(atomic percent)
35

3.2 Synthesis of other metal foams systems
3.2.1 System selection
As mentioned in section 2.2.1, the system selection for dealloying process
requires a large potential difference between the alloy elements, a proper
composition to meet parting limit, and suitable electrolytes for chemical reactions.
Here, a more detail discussion is presented including phase diagrams of alloys, and
chemical reaction in the electrolytes with different pH values and applied potentials,
known as Pourbaix diagrams, since the reaction in the non-noble metal foam systems
is more complicated during dealloying process. Figure 22 presents the phase
diagrams of Cu/Al and Ag/Al systems, which indicates no complex phase being
present throughout whole compositions, while the potential-pH equilibrium diagrams
of Cu and Al, shown in Figure 23, give us the idea of theoretical conditions of
corrosion, immunity, and passivation of the metals. For example, Al is ready to
dissolve in strong acid or alkaline solutions, at a pH value lower than 4 or higher
than 8, while Cu stays immune at a given condition because of higher standard
reduction potential. However, these diagrams only predict the reactions for pure
metals, the practical reactions strongly depend on the aqueous species, electrolyte
concentrations, temperatures, ion activities, and alloy compositions [5, 75, 76, 97].
Nevertheless, on the basis of the information, we can still select the possible system
for the dealloying process, as shown in Table 3. In this table, ΔE
0
represents the
standard potential difference between the alloy elements. The section 3.2.2 will
address the detail of precursor alloy preparations.
36

Figure 22: Phase diagrams of (a) Cu/Al and (b) Ag/Al systems [66].

37

Figure 23: Potential-pH equilibrium diagram for (a) Cu-water and (b) Al-water
systems at 25 °C [75].

(a)
(b)
38

Table 3: List of initial alloy systems used in this study
Alloy ΔE
0
(V) Dealloying process Applied potential
Cu
20
Zn
80
1.056 Free corrosion in H
3
PO
4
N/A
Cu
35
Zn
65
1.056 Electrochemically driven in HCl -0.2 V
Cu
30
Al
70
2.000 Free corrosion in HCl N/A
Ti
35
Al
65
0.030 Electrochemically driven in HCl 2.8V
Ni
12
Mn
88
0.935 Free corrosion in (NH
4
)SO
4
N/A
Ni
36
Fe
64
0.190 Free corrosion in H
3
PO
4
N/A
Pd
30
Ag
70
0.116 Free corrosion in HNO
3
N/A
Ag
25
Al
75
2.439 Free corrosion in HCl/NaOH N/A
Ag
10
Cu
90
0.459 Free corrosion in H
2
O
2
N/A
Ti
90
Al
6
V
4
0.530 Electrochemically driven in HCl 2.8V
Zn
50
Al
50
0.944 Free corrosion in NaOH N/A
Pd
30
Al
70
2.575 Electrochemically driven in HCl 0.2V
Ni
30
Mn
70
0.935 Free corrosion in (NH
4
)SO
4
N/A

3.2.2 Precursor alloys preparation
For this project, the synthesis of nanoporous foams has been performed using
several systems, including Ti/Al, Ag/Al, Cu/Al, Cu/Zn, Pd/Ag, Ni/Mn, Ni/Fe,
Ag/Cu, Zn/Al, Pd/Al and Ti/Al/V (Table 3). These starting alloys were either
commercially available or prepared using an arc melter. The Cu/Zn, Pd/Ag, Ni/Mn,
Ni/Fe, Ag/Cu, and Ti/Al/V alloys were purchased from Goodfellow Cambridge Ltd
39

(Huntingdon, England). The Cu/Al, Ag/Al, Ti/Al, Zn/Al, and Pd/Al alloy ingots
were arc melted from Cu (99.999%), Ag (99.99%), Ti (99.995%), Zn (99.99%) and
Al (99.999%) and homogenized under argon. Figure 24 presents the photo of arc
melter used in this study. Samples approximately 300 µm thick were cut from the
alloy ingots, polished on one side, and then heat treated for 30 minutes at 200 °C to
relieve stress.

Figure 24: Arc melter used in present study, located at UCSD Nano Engineering
department.

40

3.2.3 Electrochemical setup
The Cu
35
Zn
65
, Pd
30
Al
70
, and Cu
30
Al
70
alloys were dealloyed by a selective
electrochemically driven process in suitable electrolytes, as shown in Table 3. As
mentioned in section 3.1, a calibrated three-electrode electrochemical cell controlled
by a potentiostat (Gamry Reference 3000) was used for these experiments.
Dealloying was performed at room temperature, using a platinum wire counter
electrode and a standard Ag/AgCl reference electrode. A schematic of the
electrochemical set up is presented in Figure 25. The alloy samples were held at a
given applied electrochemical potential, with the assistance of VFP600 software, for
a period of 1-2 days until the measured dissolution current was negligible. Other

samples were dealloyed by free corrosion using suitable electrolytes during a period
of 2-3 days until no visible hydrogen bubbles and no further weight loss was
detected.

Figure 25: Schematic of the set up for electrochemically driven dealloying
41

3.2.4 Preliminary results
During dealloying process, even if the precursor alloys and the electrolytes meet
all the requirements, not all of them produced a foam structure. For example, Ti/Al
only dealloyed on the surface and the reaction was hampered due to the formation of
a passive layer. Other factors such as having an alloy with laminar microstructure
having higher corrosion resistant also affected the dealloying processing. Figure 26
(a) presents a typical porous structure from top view of Ti
35
Al
65
sample after an
electrochemical driven dealloying process in 1 M NaOH (See Appendix A).
However, the cross-sectional view shows only 50 nm depth from the surface which
was porous (Figure 26 (b)). Ag/Pd resulted in simultaneous dissolution of Ag and
Pd, due to the relative similar reduction potentials of the two elements. A similar
result was observed in Ni/Fe and Ag/Cu system. As mentioned in the Chapter 2, the
parting limit varies for different alloys, depending on the relative reduction potentials
of the alloy components, oxide formation, breakdown of the electrolyte, and other
factors. In this project, the alloy elements with a difference in standard reduction
potential below 0.53V did not dealloy into a porous structure. Another issue was
observed for the Ni
12
Mn
88
alloy, which was dealloyed into Ni powder because it does
not have enough materials to form a foam structure. Therefore, Ni
30
Mn
70
alloy was
purchased from Goodfellow Inc. and dealloyed in 1M (NH
4
)
2
SO
4
. Results show that
although the Mn atoms in the alloy were completely dissolved in the solution and the
remaining Ni atoms were not decomposed into powder, the cross-sectional SEM
image shows no foam structure, as in Figure 27. Ni atoms did not diffuse into an
42

inter-connected ligament structure during the corrosion process. However, we did
have some success in the synthesis of NP Cu, Ag, and Pd foams, which will be
addressed in Chapter 4-6, respectively.

Figure 26: SEM images (a) Top view (b) cross-sectional view for Ti
35
Al
65

dealloyed in 1M NaOH.

Figure 27: (a) Top view (b) cross-sectional SEM images for Ni
30
Mn
70
dealloyed
in 1M (NH
4
)
2
SO
4

43

3.2.5 Post dealloying heat treatments
For the foams which were successfully synthesized, each sample was divided
into several slices, and each slice was subjected to heat treatments in argon or
vacuum for 30 minutes at 200, 400 and 600 °C to produce the NP foams with a wide
range of pore/ligament sizes and study the effects of an oxide. The heat treatments
were performed using the tube furnace shown in Figure 28 with the argon flowing
setup or high vacuum pump setup. This furnace can achieved a vacuum of 10
-5
Torr
within 2 hours to prevent foam contamination and oxidation.

Figure 28: High vacuum tube furnace with flowing argon setup at USC.
44

3.3 Characterization methods
3.3.1 SEM/EDX (Scanning Electron Microscope/Energy Dispersive X-ray)
The surface morphologies of the samples were characterized using a high
resolution SEM JSM-7001 (JEOL, USA). To investigate the foam microstructure,
ligament and pore size, cross-sections of the foam were prepared by fracturing the
sample into two pieces. At least three areas of the cross-section, close to the top
surface, center point, and close to the bottom surface, were imaged to examine the
overall foam structures and its uniformity. EDX was also utilized at the same areas
to confirm the depletion of the less noble element.
3.3.2 XRD (X-Ray Diffraction)
XRD was used to confirm presence of a single element after dealloying and the
oxide formation after heat treatments. In the current study, the NP foams were
ground into powders, and placed in a Rigaku Ultima-IV X-Ray diffraction machine.
XRD was performed using Cu Kα radiation with the wavelength of 1.5 Å. A series
of 2θ scans from 20° to 80° with the rate of 4°/min were performed to provide the
XRD patterns.
3.3.3 Raman Spectroscopy
Several spectroscopic techniques are available for the analysis of materials and
chemicals. Among them, Raman spectroscopy uses Raman scattering of light by a
material, where the light is scattered inelastically compared to the more prominent
elastic Rayleigh scattering [26]. This inelastic scattering causes shifts in wavelength,
45

which can then be used to deduce information about the material. Properties of the
material can be determined by analysis of the spectrum, and/or it may be compared
with a library of known spectra to identify a substance.
In order to investigate the possible formation of oxide on the NP foams, Raman
spectra were obtained for all foams. For Raman spectroscopy measurements, a 532
nm Spectra-Physics solid state laser is collimated and focused through a Leica
DMLM microscope with a 100x objective. The laser beam spot size is around 500
nm which covers several ligaments, and the skin depth for Ag and Cu is around 5 nm.
Raman spectra of all foams were collected in a Renshaw InVia Raman
microspectrometer with a PRIOR ProScan II high precision microscope stage to
control the position of the incident laser with respect to the foams.
3.4 Mechanical testing
3.4.1 Nanoindentation
Nanoindentation is a small-scale deformation technique which pushes a probe
of known 3-dimensional geometry using a high sensitivity actuator controlling
displacement into a sample while recording the resulting force and displacement
signals. The nanoindenter used in this study is a Hysitron Triboindenter (Hysitron
Inc., Minneapolis, Minnesota) and uses a high-precision motor to bring the probe tip
close to the sample surface. This method uses the force-displacement curve to
measure the sample stiffness and hardness. As the probe is a 3-dimensional shape,
the equation also must take into account the projected cross- sectional area as it
46

varies with depth, A(h). The contact depth measures how deep into the sample the
probe indents beyond the surface contact point. The probe will press a maximal load,
Fmax, into the sample and the hardness value, H, is measured using H = Fmax/A(h).
The hardness of nanoporous foams was tested by depth sensing nanoindentation
with a Berkovich tip (radius of ~ 200 nm). Indentations were performed on the
planar, “polished surfaces” (polishing was performed prior to dealloying) of the
samples. All foam nanoindentation experiments were performed using a constant
loading rate of 500 µN/s, with loads ranging from 2000 to 6000 µN. A minimum of
25 indents were performed on each sample. Figure 29 shows a representative
load/unloading curves and the Scanning Probe Microscopy (SPM)/SEM images of a
residual indent [7].
.
Figure 29: (a) Series of load vs. displacement curves collected on the surface of
nanoporous Au using a Berkovich tip, and the (b) SPM and (c) SEM images of a
residual Berkovich indent [7].
47

3.4.2 Micro Vickers hardness testing
For the foams with the pore sizes above 200 nm, the hardness values acquired
from the nanoindentation tests were no longer valid, due to the pore size being
comparable to tip radius of a Berkovich tip, which is around 200 nm. In that case,
the hardness values were measured using micro Vickers hardness testing (Leco
LM100) instead. The micro Vickers testing allows a relative large pore size and a
higher surface roughness. A minimum of 10 indents were performed for all foams,
using a peak load of 10 gf with a 10 second loading time.

48

Chapter 4. Morphology, Oxidation and Mechanical Behavior of
Nanoporous Cu Foams
4.1 Introduction
As discussed in chapter 2, most studies of NP metal foams have been focused
on NP Au foams as a model system. However, the performance of non-noble metal
foams is still unclear. Among the NP metal foam systems, NP Cu has several
attributes that make it attractive for a study, such as, low density, high modulus of
elasticity, and relatively lower cost of precursor materials [40]. Moreover, in
comparison to previous work on NP Au, Cu has a native oxide and can be oxidized
during the dealloying or heat treatments process, which makes it an ideal candidate
to study the effect of oxide on overall foam properties.
In this chapter, NP Cu foams with relative densities ranging from 20% to 35%
were successfully synthesized by dealloying processes. Ligament sizes, ranging
from 60-220 nm, were found to be tunable by heat treatments. A variety of precursor
alloys were used to establish the relationship between morphology and stability. We
describe a novel method for the synthesis of nanoporous Cu oxide foams by heat
treating nanoporous copper foams. Several techniques including nanoindentation are
used to compare changes in morphology and behavior between NPC foams and NPC
oxide foams.
49

4.2 Experimental details
4.2.1 Foam processing
The NP Cu samples were processed by electrochemically driven dealloying and
free corrosion of Cu
20
Zn
80
, Cu
30
Al
70
and Cu
35
Zn
65
. The CuZn alloys were purchased
from Goodfellow Cambridge Ltd (Huntingdon, England). The Cu
30
Al
70
alloy ingots
were arc melted from Cu (99.999%) and Al (99.999%) and homogenized for 24
hours at 530 °C under argon. Samples approximately 300 µm thick were cut from
the alloy ingots, polished on one side, and then heat treated for 30 minutes at 200 °C
to relieve stress. Figure 30 shows the microstructure of the Cu
30
Al
70
base alloy,
which contains two phases, a Cu
30
Al
70
solid phase and a CuAl
2
phase. All
characterization was focused on the Cu
30
Al
70
Phase. For the Cu/Zn system, only one
phase was observed; γ phase for Cu
35
Zn
65
, and ε phase for Cu
20
Zn
80
.

Figure 30: Top view SEM image of Cu
30
Al
70
base alloy, which contains a
Cu
30
Al
70
solid phase and a CuAl
2
phase (as indicated).
50

The Cu
35
Zn
65
and Cu
30
Al
70
alloys were dealloyed by a selective
electrochemically driven process in a solution of 5wt% HCl, while a solution of
85wt% H
3
PO
4
was used for Cu
20
Zn
80
. A three-electrode electrochemical cell
controlled by a potentiostat (Gamry Reference 3000) was used for these experiments.
Dealloying was performed at room temperature, using a platinum wire counter
electrode and a standard Ag/AgCl reference electrode. The alloy samples were held
at an applied electrochemical potential of -200 mV for a period of 1-2 days until the
measured dissolution current was negligible. Cu
20
Zn
80
and Cu
30
Al
70
samples were
also dealloyed by free corrosion using 85wt% H
3
PO
4
and 5wt% HCl solutions,
respectively, during a period of 2-3 days until no visible hydrogen bubbles and no
further weight loss was detected. Free corrosion procedures did not yield any
significant weight loss for Cu
35
Zn
65
.
In order to produce samples with a wide range of pore/ligament sizes, each
sample (20%, 30% and 35% relative density) was divided into four slices, and each
slice was subsequently subjected to heat treatments in argon for 30 minutes at 200,
400 and 600 °C.
4.2.2 Foams characterization and testing
Scanning electron microscopy (SEM) was employed for microstructural
characterization. Energy-dispersive X-ray (EDX) spectra were collected for all
nanoporous Cu samples, confirming that the remaining Al or Zn concentrations was
less than 1.0 at% after dealloying. Additionally, the measured mass before and after
51

dealloying was utilized to further verify sample density. The thickness of each
sample was measured before and after dealloying; dimensional changes were
determined to be negligible.
As discussed in section 3.3, the oxide content of nanoporous copper was
determined by Raman spectroscopy and X-ray powder diffraction (XRD). For
Raman spectroscopy measurements, a 532 nm Spectra-Physics solid state laser is
collimated and focused through a Leica DMLM microscope with a 100x objective.
The laser beam spot size is around 500 nm which covers several ligaments, and the
skin depth for copper is around 5 nm. Raman spectra of NP Cu were collected in a
Renshaw InVia Raman microspectrometer with a PRIOR ProScan II high precision
microscope stage to control the position of the incident laser with respect to the NP
Cu. For the XRD studies, the samples were ground to a fine powder, and placed on a
microscope glass slide in a Rigaku diffractometer (Ultima IV, Japan).
The hardness of nanoporous foams was tested by depth sensing nanoindentation
using a Hysitron Triboindenter with a Berkovich tip (radius of ~ 200 nm).
Indentations were performed on the planar, “polished surfaces” (polishing was
performed prior to dealloying) of the samples. All foam nanoindentation
experiments were performed using a constant loading rate of 500 µN/s, with loads
ranging from 2000 to 6000 µN. A minimum of 25 indents were performed on each
sample. Foams heat treated above 600 °C, were not tested by nanoindentation due to
surface roughness and large pore sizes as compared to the Berkovich tip radius.
52

4.3 Results and Discussion
4.3.1 Ligament size and morphology
To obtain uniform nanoporosity by dealloying, it is essential to select suitable
copper alloys as precursors, which satisfy the basic requirement: a large
electrochemical potential difference between copper and the other alloying
components [11]. In this study, we selected Cu
20
Zn
80
, Cu
30
Al
70
, and Cu
35
Zn
65
as the
precursor materials. The standard reversible potentials of Zn and Al are -0.762 V
and -1.66 V (vs. standard hydrogen electrode (SHE)), respectively, whereas for Cu it
is 0.342 V [58]. Numerous acid solutions were tested as potential electrolytes to
selectively etch the precursors. It was found that H
3
PO
4
, and HCl can effectively
produce uniform nanoporosity for Cu
20
Zn
80
and Cu
30
Al
70
. However, the Cu
35
Zn
65

alloy did not yield a significant weight loss under free corrosion. Therefore, only
electrochemical dealloying was used to produce 35% relative density NP Cu.
The pore size and ligament size of a nanoporous foam can be controlled by the
dealloying technique as well as by heat treatments [8]. In Figures 31, 32, and 33,
SEM cross-sectional images show the pore size and ligament size of the copper
foams with different relative densities after heat treatments at 200, 400 and 600 °C.
Figure 31(a) corresponds to as prepared 20% relative density foam, dealloyed from
Cu
20
Zn
80
which exhibits a smooth ligament morphology, with an average ligament
size of 60 nm. This sample was formed after 3 days of free corrosion dealloying
using an 85wt% H
3
PO
4
solution. Figures 31 (b) - 31 (d) show the microstructure
53

after 200, 400, and 600 °C heat treatments in high purity argon atmosphere. The
ligament sizes of the heat treated copper foams coarsen from 65 nm to 150 nm, and
the morphology resembles interconnected particles. Figure 32 (a-d) shows the cross-
sectional images of the 30% relative density copper foam dealloyed from Cu
30
Al
70
under free corrosion, before and after heat treatments. These foams have a spherical
ligament structure with sizes ranging from 90 nm to 110 nm, which suggests a
relatively stable ligament size and morphology. Figure 33 (a) shows the cross-
sectional image of the 35% relative density copper foam processed from Cu
35
Zn
65
by
electrochemical dealloying and after heat treatments (Figure 33b-d). The
morphology differs from that of the previous two foams, however, it is still
composed of nano-sized ligaments and pores. After heat treatment, the ligament
sizes range from 100 nm to 220 nm.

Figure 31: Cross sectional SEM images of a 20% relative density nanoporous
copper foam dealloyed from Cu
20
Zn
80
(a) as prepared and after heat treatments
at (b) 200°C (c) 400°C (d) 600°C.
54

Figure 32: Cross sectional SEM images of a 30% relative density nanoporous
copper foam dealloyed from Cu
30
Al
70
(a) as prepared and after heat treatments
at (b) 200°C (c) 400°C (d) 600°C.

Figure 33: Cross sectional SEM images of a 35% relative density nanoporous
copper foam dealloyed from Cu
35
Zn
65
(a) as prepared and after heat treatments
at (b) 200°C (c) 400°C (d) 600°C
55

Previous NP Cu foams studies did not discuss either heat treatments or oxide
formation [11, 40, 78, 117]. However, our as-prepared structures morphologies,
agree with these previous research which showed that dealloying in different
electrolytes leads to significantly different ligament morphologies and sizes. For
example, Cu
30
Mn
70
dealloyed in a 1.3 pH HCl solution exhibits a smooth ligament
morphology, with an average ligament diameter of 125 ± 30 nm, while dealloying in
citric acid results in ligaments with diameters of 80 ± 20nm, which are not as smooth
or uniform as those formed by the HCl electrolyte [40]. Chen et al demonstrated
that the average pore size can be tuned from 15 to 120 nm by controlling the
dealloying time using a Cu
30
Mn
70
alloy [11]. Melt-spun Al-Cu alloy ribbons were
also dealloyed in 5wt% HCl solution and exhibit an open, bicontinuous ligament
structure with length scales of 100-500 nm for various relative densities [78].
Monolithic NPC ribbons and bulk NPC were fabricated by free corrosion of Mg-Cu
alloys in 5wt% HCl solution and presented a typical bicontinuous ligament-channel
structure with length scales of 148±35 nm for Cu
33
Mg
67
, 175 ± 27 nm for Cu
40
Mg
60

and 211 ± 37 nm for Cu
50
Mg
50
[117]. Additionally, it has been reported that
electrochemically driven dealloying significantly increased the dealloying rate, thus
affecting the final ligament structure (See Figure 33a) [20]. The ligament sizes are
typically three times smaller than that of free corrosion [40, 44],

and by increasing
the dealloying potential the ligament size can decrease to 4-5 nm [71]. However,
electrochemically driven dealloying can lead to more defects such as surface cracks
[71, 88].
56

4.3.2 Oxide Formation
As previously mentioned, to the author’s knowledge, there have been no prior
studies on the oxide formation in NPC foams. Given that Cu has a native oxide
(approximately 4 nm) and further oxide growth is expected as a function of the heat
treatments, we calculated the oxidation growth of copper between 100 °C and 600 °C
as a function of temperature and time using the following equation [80]:

( )
0
2 1
exp d t
T R
Q
A t d
oxide
+ ×






×
−
= (7)
where d
oxide
(t) is the thickness of the formed copper oxide as a function of time, R is
the gas constant 8.314 × 10
-3
kJ/K mol, T is the temperature and t is time in minutes,
d
0
is the initial copper oxide thickness, and the activation energy Q was found
empirically to be 33.1 kJ/mol, 42.21 kJ/mol, and 78.9 1 kJ/mol for 200 °C, 400 °C,
and 600 °C, respectively [80, 83]. The initial coefficient A is in units of Å/min
½
with
values ranging from 5.518 × 10
5
to 6.658 × 10
7
[80, 83].
The calculated oxide thickness layer after 30 min heat treatments at 200 °C, 400
°C and 600 °C is around 70 nm, 145 and 690, respectively. These thickness values
are an upper boundary approximation since the heat treatments were performed in a
high purity argon atmosphere and the parabolic rate equations assumes oxidation in
air. Therefore, it could be assumed that after a 600 °C heat treatment, the copper
foams should completely convert into copper oxide foams. A more direct
approximation can be found by taking into account the density changes from the
57

transformation of Cu into CuO and Cu
2
O and assuming that all measured mass gain
is due to oxide formation; the oxide layer thickness can then be calculated by the
following formula [43]:
O Cu CuO
Cu
m
m
t t
2
,
ρ
ρ ∆
= ∆
(8)
where Δt is the oxide layer thickness, t is the strut-wall thickness, Δm/m is the foam
relative mass gain, and the densities are ρ
Cu
= 8.95 g/cm
3
, ρ
Cu
2
O
= 6.0 g/cm
3
, and
ρ
CuO
= 6.48 g/cm
3
. Depending on the oxide identified by the XRD results, one can
use either the density value for CuO or Cu
2
O. Using ligament sizes as the strut-wall
thicknesses, the oxide thickness given by Eq. (2) ranges from 17 to 36 nm for 20%
relative density foams, 21 to 50 nm for 30% relative density foams, and 19 to 130
nm for 35% relative density foams. However, since changes in dimensions are
expected even without the oxide [44], these results only yield a lower bound
approximation.
In order to verify our previous oxide calculations, the oxide content was
examined by Raman spectra and XRD. Previous studies in bulk copper showed that
Raman Spectroscopy can be used to follow the formation of the different copper
oxides in situ [86]. The changes in the Raman spectra with increasing temperature
reflect the different stages of copper oxidation, with the formation of Cu
2
O between
70

°C and 130 °C, and the formation of CuO at temperatures higher than 260 °C.
From XRD studies, it was observed that copper films formed Cu
2
O at temperatures
58

below 275 °C, while oxidation above 325 °C produced a CuO phase [89]. However,
the oxidation of open-cell nanoporous copper foams has not been previously
measured. The XRD patterns obtained for 20%, 30%, and 35% relative density
copper foams are presented in Figure 34. For the 20% relative density foam, the
XRD data shows that the as-prepared sample is mostly copper with small Cu
2
O
peaks; after a 200 °C heat treatment, mixed peaks of Cu and Cu
2
O are observed;
while after 400 and 600 °C heat treatments, only CuO peaks are visible. For the 30%
relative density copper foam, the XRD data is similar to the data for 20% relative
density foam. However, for the 35% relative density foam, Cu peaks were still
present after 200 and 400 °C heat treatments and no longer appeared after a 600 °C
heat treatment. Further oxide analysis of the foam surface was performed by Raman
spectroscopy (Figure 35). For the 20% relative density foam (Figure 35a), the
Raman spectra show that the as-prepared sample is mostly copper without Cu
2
O or
CuO peaks; while after a 200 °C heat treatment, mixed peaks of CuO and Cu
2
O are
observed. After 400 and 600 °C heat treatments, the Raman spectra peaks were
slightly shifted but still showed both CuO and Cu
2
O peaks. For the 30% relative
density copper foam (Figure 35b), the Raman spectra were similar to the data for the
20% sample. For the 35% relative density copper foam (Figure 35c), the Raman
spectra show no CuO and Cu
2
O peaks for the as prepared sample, however, the
signal is more noisy than for the other copper foams. Furthermore, after a 200 °C
heat treatment, both CuO and Cu
2
O peaks were present. Therefore, the XRD and
Raman results appear to be consistent; the 35% samples seem to have oxide
59

formation mostly on the surface with some residual copper left even after 400 °C
heat treatments.

Figure 34: XRD results for (a) 20% (b) 30% (c) 35% relative density
nanoporous copper foams as prepared (AP) and after 200, 400, and 600°C heat
treatments.
60

Figure 35: Raman spectra for (a) 20% (b) 30% (c) 35% relative density of
nanoporous copper foams as prepared (AP) and after 200, 400, and 600°C heat
treatments.
61

4.3.3 Nanoindentation
In order to verify the extent of the oxide, nanoindentation tests provide hardness
values which allow for a comparison to the results from Raman spectroscopy and
XRD. Previous research on nanoporous gold foams [41], demonstrated that, at the
nanoscale, strength is controlled not only by the relative foam density, but also by
the ligament length scale. For NPC foams, there was only one previous study on a
30% relative density sample with reported hardness values of 128 ± 37 MPa and
ligament sizes of 135 ± 31nm [40]. In figure 36, representative loading-unloading
curves for a 30% relative density sample (heat treated at 200 °C) including SEM
micrographs of the residual impressions of an array of indents are presented. It
should be noted that the deformation behavior is similar to the study by Hayes et al
[40]. However, it remains unclear how the mechanical strength of a heat treated
NPC foam (containing oxides: CuO and Cu
2
O) would change as the oxide content is
increased and the foams is no longer purely metallic.
According to the oxide handbook [83], the hardness values of CuO and Cu
2
O
are between 2050-2490 and 2010-2030 MPa, respectively, while the hardness value
of Cu is ~ 755-980MPa [16]. Therefore, changes in the foam hardness values can
allow us to estimate the oxide contribution as a function of the ligament size. In
Table 4, the hardness values for various relative densities copper foams with and
without oxide are presented. The hardness values increased significantly after 200
°C and 400 °C heat treatments for the 20% and 30% relative density foam, in the
range of 14.7 to 575 MPa and 19.7 to 529 MPa respectively. The substantial
62

Figure 36: Nanoindentation tests of a 30% relative density copper foam after a
200°C heat treatment (a) representative loading/unloading curves (b) SEM
images of typical nanoindentation array (c) increased magnification of a single
indent
63

Table 4: Nanoindentation results
Relative density
[%]
Ligament size
[nm]
Heat treatment
[
O
C]
Hardness
[MPa]
20 60 As prepared 14.7 ± 2.6
20 65 200 220 ± 58
20 75 400 575 ± 188.9
20 150 600 NA
30 90 As prepared 19.7± 0.2
30 105 200 278.2± 43.1
30 105 400 529± 108.3
30 110 600 NA
35 100 As prepared 30.7 ± 1.8
35 140 200 31.3 ± 5.9
35 180 400 33.7± 5.2
35 220 600 NA

increase in hardness values can be explained by the formation of the oxide and the
nanoscale ligament size. However, the hardness values of the 35% relative density
copper foams did not increase for samples heat treated at 200 °C and 400 °C. This
result seems to agree somewhat with the XRD data (figure 34c) where the spectra
shows that copper peaks are still present even after a 400 °C heat treatment. The
estimate of the oxide thickness for the 35% foam as compared to the ligament size,
indicates that a complete transformation into a copper oxide foam is not expected;
however, an oxide is present and the reason for a lack of increase in hardness needs
further study. Perhaps the initial morphology and ligament stability can be exploited
64

so that both Cu and Cu oxide can be synthesized having a wide range of ligament
sizes.
4.3.4 Heat treatments in vacuum
In order to produce NP Cu foams with a wide range of ligament sizes without
the formation of oxides, heat treatments of nanoporous Cu foams were performed in
a high vacuum furnace, which can achieve a vacuum of 1x10
-5
Torr. Figure 37
shows XRD data for 20% and 30% relative density as-prepared NP copper foams
and after 200, 400, and 600°C heat treatments in vacuum, while for 35% relative
density foams being limited to one heat treatment due to the sample size, only a heat
treatment temperature at 400 °C XRD data is presented. The overall results show
that the peaks of Cu foams after heat treatments in vacuum are similar to those of as
prepared foams. No further oxidation on the foams was observed. Furthermore,
Figures 38-40 show the representative cross-sectional SEM images for these foams
before and after heat treatments. It was observed that the changes in average
ligament size and morphology of the foams are insignificant. An average ligament
size for 20% relative density Cu foams is around 60 nm for both as-prepared and
heat treated foams, even though the ligament size of the foam after a 600 °C heat
treatment is not as uniform as that of the as-prepared foam. The diffusion inside Cu
ligament may occur during the heat treatments. For 30% relative density foams, the
average ligament size increases from 90 to 100 nm after a 600°C heat treatment,
while that of a 35% relative density foam increases from 100 to 115 nm after heat
treated at 400°C, as listed in Table 5. Since the ligament size of NP Cu foams does
65

not change significantly as expected, it appears to show better thermal stability in
high vacuum than under argon.

Figure 37: XRD results for (a) 20% (b) 30% relative density nanoporous copper
foams as prepared (AP) and after 200, 400, and 600°C heat treatments in
vacuum. The results for (c) 35% relative density as prepared nanoporous
copper foams and after 400°C heat treatments in vacuum were also presented.
66

Figure 38: Cross sectional SEM images of a 20% relative density nanoporous
copper foam dealloyed from Cu
20
Zn
80
(a) as prepared and after heat treatments
in vacuum at (b) 200°C (c) 400°C (d) 600°C.

Figure 39: Cross sectional SEM images of a 30% relative density nanoporous
copper foam dealloyed from Cu
30
Al
70
(a) as prepared and after heat treatments
in vacuum at (b) 200°C (c) 400°C (d) 600°C.
67

Figure 40: Cross sectional SEM images of a 35% relative density nanoporous
copper foam dealloyed from Cu
35
Zn
65
(a) as prepared and after heat treatments
at (b) 400°C.

Table 5: Ligament sizes of NP Cu foams after heat treatments in vacuum
Relative density
[%]
Ligament size
[nm]
Heat treatment
[
O
C]
20 60 As prepared
20 60 200
20 60 400
20 62 600
30 90 As prepared
30 95 200
30 95 400
30 100 600
35 100 As prepared
35 NA 200
35 115 400
35 NA 600

68

4.4 Summary
We successfully synthesized 20-35% relative density NPC foams by a
dealloying process. The ligament sizes, ranging from 60-220 nm, can be controlled
by the dealloying techniques or subsequent heat treatments. However, the formation
of oxides during heat treatment occurs even in an atmosphere of flowing high purity
argon. XRD and Raman results show that copper oxide peaks are present after heat
treatments above 200 °C. The foams were shown to retain their structural integrity
even after oxidation. Nanoporous copper oxide foams can thus be formed by heat
treating nanoporous copper.
Furthermore, testing results show that the 35% relative density NPC foam is
different from the 20% and 30% foams. XRD data show that copper peaks are
present for the 35% foam, even after a 400 °C heat treatment, while for the 20% and
30% foams, only CuO peaks are visible after a 400 °C heat treatment. By
comparison of the estimated oxide thickness to the actual ligament sizes, it is
assumed that the 35% foam is not completely converted to copper oxide after heat
treatments below 400 °C. The hardness values of the 20% and 30% foams increases
~ 40 times upon heat treatment, indicative of oxide formation, while that of the 35%
foam remains relatively unaffected. In addition, the oxide formation can be avoided
during the heat treatments using a high vacuum chamber. The ligament sizes show
no significant change after heat treatments under vacuum.

69

Chapter 5. High temperature morphology and stability of
nanoporous Ag foams
5.1 Introduction
In this chapter we emphasized on the processing and the thermal stability of
NP Ag foams. For nanoporous silver (NPS) foams, Ag-Al, Ag-Zn, Ag-Mg and Ag-
Cu alloys were selected to investigate dealloying routes [46, 55, 57, 79]. During the
dealloying of Ag-based alloys, it has been found that the parameters, such as the
choice of electrolyte, time, and driving condition (free corrosion or potentiostatic),
are important factors which allow one to control the ligament size and morphology.
Subsequent heat treatment has also been successfully used for Cu and Au foams to
change the ligament size [13, 41]. However, there has been no previous studies on
heat treatments of NPS, and since silver has a native oxide, heat treatments can result
in further oxidation which must be monitored.
It should be mentioned that oxide formation is not necessarily detrimental,
since porous metal oxides have the potential for applications as lightweight
functional materials, particularly when they are arranged into structures with high
surface areas [74]. For example, AgO was applied as a cathode material in zinc
silver oxide button cell batteries for better performance with regard to voltage
regulation and storage life [1], while Ag
2
O is extensively utilized in the fields of
oxidation catalysis for carbon monoxide and numerous other volatile organic
compounds [107]. However, silver oxide decomposes easily into metallic silver and
70

oxygen when heated, even at low temperatures [52], and given the high surface area
of the NPS, the effect of decomposition on the structure could be important.
In this work, NPS foams with 25% relative density (75% porosity) were
successfully synthesized by a dealloying process. Ligament sizes, ranging from 20-
175 nm, were found to be tuneable by the use of different electrolytes or subsequent
heat treatments. In order to investigate the effect of temperature on the morphology
and to minimize the formation of silver oxides, heat treatments in both vacuum and
flowing argon were employed. Several techniques, including Raman spectroscopy,
were used to compare changes in morphology and thermal stability of NPS foams
including the formation of silver oxides.
5.2 Experimental Details
As discussed in Chapter 3, Ag
25
Al
75
alloy ingots were arc melted from Ag
(99.99%) and Al (99.999%) and homogenized for 24 h at 530 °C under argon
(99.999%). Samples approximately 300 µm thick were cut from the alloy ingots,
polished on one side, and then heat treated for 30 min at 200 °C to relieve stresses.
Figure 41 shows the microstructure of the base alloy, which contains two phases, an
Ag
25
Al
75
solid phase and an intermediate δ phase (Ag
2
Al). All characterization was
focused on the solid Ag
25
Al
75
phase.
The Ag
25
Al
75
samples were dealloyed by free corrosion using either 5 wt%
HCl (sample 1) or a 1M NaOH solution (sample 2) during a period of 4 to 48 hours,
until no visible hydrogen bubbles nor further weight loss could be detected. In order
to produce specimens with a wide range of pore/ligament sizes, both sample 1 and
71

sample 2 were divided into seven slices. One slice was left as-prepared. Three of the
slices were subjected to heat treatments for 30 min at 200, 400, and 600 °C in argon
(99.999%). The other three slices were heat treated for 30 min at 200, 400, and 600
°C under vacuum (10
-5
torr).

Figure 41: Top view SEM image of Ag
25
Al
75
base alloy, which contains an
Ag
25
Al
75
solid phase (flat area) and an Ag
2
Al phase (as indicated).

Scanning electron microscopy (SEM) was employed for microstructural
characterization. Energy-dispersive X-ray (EDX) spectra were collected for all
nanoporous Ag foams, confirming that the remaining Al concentrations were less
than 1 at% after dealloying. Additionally, the measured mass before and after
dealloying was utilized to verify sample density. The thickness of each sample was
measured before and after dealloying; dimensional changes were determined to be
negligible. Figure 42 shows a photo of samples before and after dealloying.
72

Figure 42: A photo of NP Ag samples before and after dealloying.

The oxide content of the NPS was determined by Raman spectroscopy and
X-ray diffraction (XRD). For Raman spectroscopy measurements, a 532 nm
Spectra-Physics solid state laser was collimated and focused through a Leica DMLM
microscope with a 100x objective lens. The laser beam spot size is around 500 nm,
which covers several ligaments, and the skin depth for silver is approximately 2 nm.
Raman spectra of NPS samples were collected in a Renshaw InVia Raman
microspectrometer with a PRIOR ProScan II high precision microscope stage to
control the position of the incident laser with respect to the NPS. For the XRD
studies, the samples were ground to a fine powder and placed on a glass microscope
slide in a Rigaku diffractometer (Ultima IV, Japan).
73

5.3 Results and discussion
5.3.1 Ligament size and morphology
To obtain uniform nanoporosity, it is essential to select suitable silver alloys as
precursors which satisfy the basic requirement: a large potential difference between
silver and the other alloying components [23]. In this study, Ag
25
Al
75
was selected
as a precursor material. The standard reversible potentials of silver and aluminium
are 0.8 V and -1.66 V (vs. standard hydrogen electrode), respectively [58].
Numerous acid and alkaline solutions were tested as potential electrolytes to
selectively etch the precursor alloy. It was found that HCl and NaOH can effectively
produce uniform nanoporosity for Ag
25
Al
75
.
In Figure 43, cross-sectional SEM images show the pore sizes and ligament
sizes of the silver foams with 25% relative density, using different electrolytes.
Figure 43a corresponds to as prepared silver foams (denoted as sample 1), dealloyed
by 5 wt% HCl, which exhibit a smooth ligament morphology, with an average
ligament size of 115 nm. An average ligament size of 20 nm can be observed in
figure 43 b which corresponds the as prepared foams dealloyed by 1 M NaOH
(denoted as sample 2). The ligament sizes were determined using the SEM images.
At least 50 ligaments were measured in each sample. Figure 43 c and d represents
the distribution of the ligament size for the as-prepared samples 1 and 2,
respectively. The as-prepared sample morphologies, agree with previous studies,
which show that dealloying in different electrolytes leads to significantly different
ligament morphologies and sizes. For example, Ag
30
Al
70
has been dealloyed by 1.0
74

and 2.5 wt% HCl solutions with the ligament size about 150-300 and 80-190 nm,
respectively [79]. It has been found that chlorine ion may greatly promote the
surface diffusion of Ag atoms, leading to the coarsening of the porous structure.
However, previous studies discussed neither heat treatments nor oxide
formation. Therefore, in order to determine the high temperature morphology and
thermal stability of the silver foams, both of the as prepared samples were heat
treated at 200, 400, and 600 °C in vacuum and in high purity argon. Representative
cross-sectional SEM images are shown in Figures 44-45. The ligament sizes,
ranging from 20 to 165 nm, increase with heat treatment temperature in argon (Table
1), which agrees with similar observations on nanoporous copper foams [13]. Under
vacuum, the ligament sizes only increased at 400 °C. While at 600 °C, the ligament
sizes decreased compared to the sizes at 400°C. The effect of silver oxide formation
and decomposition at elevated temperature and the thermal evaporation of silver in
high vacuum will be discussed further throughout this manuscript. However,
according to the changes in the ligament sizes, sample 2, which was dealloyed by
NaOH, appears to be more thermally stable than sample 1, which was dealloyed by
HCl.

75

Figure 43: Cross-sectional SEM images and representative distribution of
ligament size for 25% relative density nanoporous silver foams dealloyed from
Ag
25
Al
75
, using different electrolytes: (a) (c) 5 wt% HCl (sample 1) and (b) (d) 1
M NaOH (sample 2).

76

Figure 44: Cross-sectional SEM images of 25% relative density nanoporous
silver foams dealloyed using 5 wt% HCl (sample 1) and heat treated under
vacuum at (a) 200 °C (b) 400 °C (c) 600 °C and under flowing argon at (d) 200
°C (e) 400 °C (f) 600 °C.
77

Figure 45: Cross-sectional SEM images of 25% relative density nanoporous
silver foams dealloyed using 1 M NaOH (sample 2) and heat treated under
vacuum at (a) 200 °C (b) 400 °C (c) 600 °C and under flowing argon at (d) 200
°C (e) 400 °C (f) 600 °C.

78

Table 6: Ligament size results for NP Ag foams
Electrolyte
used
Heat treatment
[
O
C]
Ligament size after each heat
treatment [nm]
Vacuum Argon
HCl As prepared 115±30 115±30
HCl 200 150±31 155±35
HCl 400 175±45 160±36
HCl 600 150±36 165±36
NaOH As prepared 20±4 20±4
NaOH 200 25±5 25±5
NaOH 400 27±6 27±7
NaOH 600 25±5 30±7

5.3.2 Heat treatments in argon/Oxide formation and decomposition
Oxide formation on the NPS surface is expected given that silver oxide,
including Ag
2
O and AgO, can be synthesized by thermal oxidation of silver [85].
Therefore, in order to verify the presence of an oxide, samples were examined by
XRD and Raman spectra. Previous studies in nano-crystalline silver oxide films and
polycrystalline powders of silver oxides showed that XRD and Raman spectra can be
used to follow the phase changes of different silver oxides [81, 107]. The Raman
spectra and XRD patterns obtained for 25% relative density as-prepared foams with
different ligament sizes are presented in figure 46. For both of the as-prepared
foams, the Raman spectra shows silver without any silver oxide peaks; while after a
79

200 °C heat treatment, only mixed AgO and Ag
2
O peaks are observed for both
samples. At 400 °C, the silver oxide peaks become less clear, but can still be
indentified; while at 600 °C, no oxide peaks were observed. It has been reported that
silver oxide (AgO and Ag
2
O) decomposes easily to metallic silver and oxygen when
heated, even at low temperatures ranging from 174 to 400 °C [52, 74]. Waterhouse
et al. demonstrated that the thermal reduction of AgO to Ag
2
O was effected above
120 °C and completed by 200 °C, while after heating to 400 °C, complete thermal
decomposition of the Ag
2
O to Ag and O
2
was observed [107].

Figure 46: Raman spectra and XRD result for the foams dealloyed using 5wt%
HCl , sample 1 (a and b, respectively) and dealloyed using 1M NaOH, sample 2
(c and d, respectively) as prepared (AP) and after 200, 400, and 600 °C heat
treatments under flowing argon.
80

For the XRD result, only silver peaks are present for both of the as-prepared
foams even after a heat treatment up to 600 °C. These results combined with the
Raman spectra, seemed to indicate that the heat treated foams are mainly composed
of metallic silver with only a slight oxide component on the foam surface, since the
skin depth of silver for Raman spectra was approximately 2 nm.
5.3.3 Heat treatments in vacuum/Thermal evaporation of silver
In order to tune the ligament size over a wide range and control oxide
formation at the same time, heat treatments in vacuum were also employed. The
XRD and Raman spectra show no oxide peaks at any heat treatment up to 600 °C
(not shown here). However, given silver’s low vapour pressure, it is important to
consider the possible thermal evaporation of the material under vacuum. The
thermal evaporation equation of silver in vacuum is presented as [64]:
J = A exp (-Q/RT) (9)
Where J is evaporation rate in g/cm
2
min, R is the gas constant 1.986 x 10
-3

kcal/Kmol, T is the temperature, and the activation energy Q was found empirically
to be 67.6 ± 1.4 kcal/mol. The coefficient A has units of g/cm
2
min with the value
ranging from 2.40 x 10
8
to 2.19 x 10
10
.
The calculated evaporation rate of silver after heat treatments at 200, 400,
and 600 °C in vacuum is 5.4 x 10
-14
, 6.7 x 10
-5
, and 5.6 g/cm
2
min, respectively,
assuming the maximum value for the A coefficient. At 200 and 400 °C heat
treatments, the evaporation of silver due to a 30 min heat treatment should be
81

negligible. However, a heat treatment at 600 °C for 30 min could have an effect,
since the evaporation thickness of the silver layer is calculated to be ∼ 16 nm, which
is comparable to the ligament sizes for sample 2. Therefore, compared to the
ligament size after heat treated at 400°C, it seems that the observed decreased in
ligament size and weight loss at 600 °C under vacuum is due to the evaporation rate,
although all SEM images showed that the integrity of the foam structure remains.
5.4 Summary
We have successfully synthesized nanoporous Ag foams by a dealloying
process. The ligament size was controlled by using different electrolytes and
subsequent heat treatments. However, the heat treatments were not as effective at
changing the ligament size as the use of different electrolytes. NPS foams appear to
show relatively high thermal stability compared to other metal foams.
Oxidation of NPS was observed after heat treatments in argon, but was limited
to only a surface layer. This thin oxide layer was decomposed into silver and oxygen
at 600 °C. At the same temperature, the decreased ligament size was observed under
vacuum due to the thermal evaporation rate of silver. Overall, the choice of argon or
vacuum for the heat treatment did not greatly affect the foams morphology.

82

Chapter 6. Strength scale behavior of nanoporous Ag, Pd, Cu
foams
6.1 Introduction
As discussed in chapters 4 and 5, we were able to synthesize NP Cu and Ag
foams with different relative densities and ligament sizes. Here, we will address an
additional synthesis process for NP foam systems - NP Pd. The hardness values of
the NP Ag, Cu, and Pd foams are compiled here for the first time, in order to
evaluate their behavior in comparison to NP Au foams.
In general, for open cell macroporous foams, two main deformation
mechanisms have been reported under compression: elastic buckling and plastic
collapse (see Chapter 2). It is unclear if those mechanisms still apply for
nanoporous foams given the cell geometry, aspect ratio and nanoscale size.
Recently, several studies have focused on the failure mechanisms of nanoporous
foams with the goal of accounting for size effects and surface effects [21, 41, 112].
The elastic buckling Euler-Bernoulli and Timoshenko beam models were modified
to incorporate surface effects for nanoporous Au [112]. The plastic collapse model
from Gibson and Ashby, which considers the plastic collapse mechanism to be
controlled by a plastic hinge at the rigid nodes [29], has also been modified to
include ligament and node nanoscale size effects [21, 41].
In this study, the mechanical behavior of multiple nanoporous metal foam
systems, including Ag, Cu, and Pd, with ligament sizes ranging from 20-115 nm are
83

evaluated using nanoindentation and micro Vickers hardness tests. The results are
compared to nanoporous Au data and to several available compression deformation
models. Thus a comprehensive study of the effect of ligaments size on a wide range
of nanoporous metals systems is presented.
6.2 Experimental details
The Ag and Cu foams were dealloyed from Ag
25
Al
75
, Cu
20
Zn
80
, Cu
30
Al
70
and
Cu
35
Zn
65
alloys following the procedures described in previous chapters [12, 13]. A
nanoporous Pd foam was processed by electrochemically driven dealloying using a
Pd
30
Al
70
alloy. The Pd/Al ingot was arc melted from Pd (99.95%) and Al (99.999%)
and homogenized at 600°C for 24 hr. A three-electrode electrochemical cell
controlled by a potentiostat (Gamry Reference 3000) was used for these experiments.
Dealloying was performed at room temperature, using a platinum wire counter
electrode and a standard Ag/AgCl reference electrode, with 5wt% HCl as an
electrolyte. The alloy samples were held at an applied electrochemical potential 200
mV for a period of 1-2 hours until the measured dissolution current was negligible.
6.3 Result and discussions
All nanoporous metal foams were characterized using Scanning Electron
Microscopy (SEM) and Energy-Dispersive X-ray (EDX), while the oxide content of
the foams was determined by Raman spectroscopy and X-Ray diffraction (XRD).
Figures 47 and 48 present the cross-sectional SEM images of the nanoporous Cu, Ag
and Pd foams and their corresponding ligament size distribution. Figure 47 (a-c)
show three types of Cu foams using different alloys as precursors which yield
84

different morphologies and porosities [13]. Figure 47(a) corresponds to a 20%
relative density foam, dealloyed from Cu
20
Zn
80
which exhibits a smooth ligament
morphology, with an average ligament size of 60 nm. Figure 47 (b) shows the image
of the 30% relative density foam dealloyed from Cu
30
Al
70
. This foam has a spherical
ligament structure with ligament sizes of about 90 nm. Figure 48 (c) shows the
image of the 35% relative density copper foam processed from Cu
35
Zn
65
by
electrochemical dealloying, which has a different morphology from that of the
previous two foams, and an average ligament size of 100 nm. In Figure 47 (d) and
(e), the SEM images show two different pore sizes and ligament sizes for the Ag
foams both with 25% relative density. Fig 47 (d) corresponds to the silver foams,
dealloyed using the solution of 5 wt% HCl, which exhibit a typical ligament
morphology, with an average ligament size of 115 nm. A representative Ag foam
dealloyed using a 1 M NaOH electrolyte is shown in Figure 47 (e) with a
corresponding ligament size of ~ 20 nm. These results are in agreement with
reported studies which showed that the pore size and ligament size of nanoporous
foam can be controlled by the dealloying conditions, such as electrolytes types,
solution concentration, dealloying time and applied potentials [11, 23, 40, 105, 113].
In Fig 45 (f), the Pd foam show an average ligament size of 115nm with a rough
ligament surface. The oxide content of all the foams was determined by XRD and
Raman spectroscopy; the Cu, Ag, and Pd foams used in this study showed no oxide
peaks by both techniques. Figure 49 shows the XRD results of NP Ag, Cu, and Pd
foams. Further details, including the Raman spectra, can be found in Chapter 4-5
85

[12, 13]. For the Au foam results used here, no oxide content is expected and has not
been reported [7, 41]. Volume changes in all the foams were measured to be
negligible.

Figure 47: Cross-sectional SEM images of nanoporous foam synthesized in this
study. Cu foams dealloyed from (a) a Cu
20
Zn
80
alloy, (b) a Cu
30
Al
70
alloy, (c) a
Cu
35
Zn
65
alloy. Ag foams dealloyed from Ag
25
Al
75
alloy using two different
electrolytes (d) 5wt% HCl and (e) 1M NaOH and (f) a Pd foam dealloyed from a
Pd
30
Al
70
alloy.

Figure 48: Representative distribution of ligament size of the (a-c) Cu, (d-e) Ag
and (f) Pd foams.
86

Figure 49: XRD results of nanoporous Cu, Pd and Ag foams

The hardness of the nanoporous foams was measured by depth sensing
nanoindentation using a Hysitron Triboindenter with a Berkovich tip (radius of ~ 200
nm) and micro Vickers hardness testing (Leco LM100). Indentations were
performed on the planar, polished surfaces (polishing was performed prior to
dealloying) of the samples. A minimum of 25 indents were performed for the
nanoindentation experiments using a constant loading rate of 500 µN/s, with loads
ranging from 2000 to 6000 µN. Micro Vickers hardness tests were employed for
nanoporous Pd foams and Ag foams synthesized using HCl as an electrolyte since
the surface of those samples was too rough for the nanoindentation. For the Vickers
87

hardness tests, a minimum of 10 indents were performed on each sample, using a
peak load of 10 gf with a 10 second loading time. Although the hardness vs. yield
strength conversion is still a topic of debate [49], in this study, for all values
presented we will use the assumption of the hardness being equal to yield strength,
(H = σ
y
,) as discussed in several other publications [7, 32, 41]. This assumption can
be made, because the ligament and pore structure remain undisturbed outside the
indent contact area as can be observed in Figures 50 (a-c), which show representative
SEM micrographs of the residual indents on Ag, Cu, and Pd foams.

Figure 50: Representative SEM images of residual impressions on nanoporous
(a) Cu, (b) Ag performed by nanoindenation and on a (c) nanoporus Pd foam
performed by Vickers indentation.

In order to correlate the properties of nanoporous foams to the ligament and
pore sizes [21, 41, 112], it is important to understand their deformation behavior at
the nanoscale. As mentioned in the earlier, macroporous foam deformation can take
place either by elastic buckling or plastic collapse, but the extent to which those
models apply to nanoporous foams is still a topic of debate. In Figure 51, current
models and experimental data on nanoporous metal foams are presented as a
88

Figure 51: Foam hardness values vs. ligament size for Au, Ag, Cu, and Pd
nanoporous foams. Predictions of modified elastic buckling models and plastic
collapse models are also shown.

function of ligament size. It should be noted that the data is plotted as the foam
hardness at a given ligament size in order to present the raw data. The four models
presented were calculated using a relative density of 30% in order to calculate foam
strength rather than ligament strength. The two solid lines at the top of Figure 51
present the models by Xia et al, which are based on the elastic modulus and critical
buckling behavior of nanoporous materials using the theory of surface elasticity,
combined with two beam models; the Euler-Bernoulli model, which covers the case
89

of small deflections of a beam which is subjected to lateral loads only, and the
Timoshenko model, which accounts for shear deformation [112]. For both modified
models, the critical average stress during the microstructural collapse of the
nanoporous unit cell is given by the following equations:
𝜎 ��
∗
=
�
�
�
�
� � �
�
( �� )
∗
( � � � )
�
�
�
(Euler model) (10)
𝜎 ��
∗
=
�
( � � � )
�
�
� �
�
�( �� )
∗
� �
�
� � �
�
( �� )
∗
+ 𝐻 �
� (Timoshenko model) (11)
where (EI)* and V are functions of t, and l and t are the length and width of the
beam, respectively. For these models, (t/l)
2
is determined by the relative density,
while the width of the beam, t , is taken to be the ligament size. Further details can
be found elsewhere in Chapter 2 section 2.3.1. There appears to be a stronger
ligament size dependence for the Euler’s model as the ligament size falls into the
range of 10 nm.
Beyond the elastic regime, Gibson and Ashby modeled the plastic deformation
of the foams as being driven by bending of cantilevers. Plastic collapse occurs at the
rigid nodes, where the generated moment is greatest, and a fully developed plastic
hinge is created when the moment exerted exceeds the fully plastic moment of the
cell edges [29]. Dou et al developed a modified plastic collapse model for
nanoporous foams in order to account for the effect of plastic strain gradients that
exist in the deformation hinges [21]:
90

𝜎 �
= 𝜎 �
+ 𝑘 �
𝜇 �
�
�
(12)
where σ
1
is the yield stress of the ligament in the absence of any strain gradient, k
3

incorporates Taylor’s constant and all geometrical or shape terms consistent with the
ligaments in the nanoporous structure, µ is shear modulus, b is Burgers vector, and t
is the strut width. The results for this model are shown in Figure 51, using the
parameters described in section 2.3.1 [21]. Although this model was developed to
explain the mechanical behavior of nanoporous gold, it should be equally applicable
to other nanoporous materials that deform through a similar plastic collapse
mechanism.
The study by Hodge et al [41] presents a modified Gibson/Ashby empirical
model which incorporates a ligament size effect similar to a Hall-Petch-type effect
and is expressed as:
𝜎 ��
= � 𝜎 �
+ 𝑘 ��
∙ 𝐿 � � � ⁄
� (13)
Where σ
0
is related to the bulk material yield strength and k
Au
is a material constant,
L is the ligament size. Overall the nanoporous data values for all samples fall near
the predicted trends of the two models which assumed plastic collapse. Some of the
scattered from the models is probably due to the 30% relative density calculation and
also due to errors in the relative density measurement of each foam. It has been
reported that alloy composition and sample shrinkage would lead to an
underestimated relative density value [59]. For example, the different morphologies
91

of the copper foams, which were synthesized using different precursor alloys, could
lead to the different foam strength values. Although, the above factors would alter
the absolute value of a given foam strength, the trends are very clear and show a
strong relationship to the plastic collapse models only.
Following the trend observed from Figure 51 for plastic collapse, one can then
calculate the ligament strengths of various metal systems using the appropriate
equation

�
∗
�
�
= 𝐶 �
�
∗
�
�
�
� � ⁄
(14)
Where σ
*
is foam yield strength, σ
s
is bulk metal yield strength (in this case, the
ligament strength), ρ
*
/ρ
s
is relative density, and C is geometry constant [29]. For
open cell foams, the constant C is 0.3. The results for the calculated ligament
strength are presented in Table 7. For the Ag foams with 25% relative density, the
ligament strengths values are higher than the bulk Ag yield strength (55MPa)[15] but
lower than the yield strength measured on a Ag nanowire (7.3GPa) [111]. The Pd
ligament strength is similar to the 1 to 1.8GPa yield strength measured from
nanostructured Pd contacts [68]. For nanoporous copper foams, there was one
previous study on a 30% relative density sample with a reported foam hardness value
of 128 ± 37 MPa for ligament sizes of 135 ± 31nm [40]. In this study, the calculated
copper ligament strength is lower than the ideal shear strength of Cu (2.16 - 3.45
GPa)[115] but higher than the yield strength of fully annealed polycrystalline Cu (55
92

MPa). To the authors’ knowledge, no yield strength data for nanoporous Ag and Pd
foams has been previously reported.

Table 7: Foam hardness and ligament strength table
Metal
foams
Relative
density
[%]
Ligament
size
[nm]
Foam
Hardness
[MPa]
Ligament
strength
[MPa]
Test Method
Cu 20 60 ± 12 14.7 ± 2.6 548±97 Nanoindentation
30 90 ± 18 19.7± 0.2 400±4 Nanoindentation
35 100 ± 21 30.7 ± 1.8 494±29 Nanoindentation
Ag 25 20 ± 4 70.0 ± 5.2 1867±139 Nanoindentation
25 115 ± 30 27.3± 2.2 728±59 Micro Vickers
Pd 30 115 ± 27 45.7± 1.8 927±37 Micro Vickers

6.4 Summary
In summary, the Ag, Cu, Pd and Au nanoporous foams hardness values seem to fit
the plastic collapse models rather than the elastic buckling models. Using the
observed trend, the ligament strength was calculated and compared to the
corresponding bulk yield strength. Although cell geometry, surface effects and node
effects are important for nanoporous cell foams, plastic collapse appears to be the
dominant mechanism regardless of the metal. Further models are still necessary to
account for the irregular cell geometry observed in nanoporous materials.

93

Chapter 7. Conclusions
Processing of NP foams using dealloying techniques has been widely
investigated in many studies with an emphasis on the correlation between synthesis
conditions and resulting microstructure. However, the effect of oxide on the foam
structure and mechanical properties needs further investigation.
In this thesis, the correlation between the dealloying conditions, such as free
corrosion and electrochemically driven dealloying, and the foam microstructure and
yield strength was investigated for NP Au, Ag, Cu, and Pd foams. It was shown that
the foam yield strengths are directly affected by the characteristic size of the
nanoscale features the foams, such as ligament and pore size. In the case of NP Cu,
the study of oxide formation and the correlation between oxide content to the foams
microstructure and resulting yield strength demonstrated the contribution of the
oxide to the overall strength.
Taking into account the importance of characteristic size effects at the nano
scale, the measured hardness values of the foams were compared with modified
deformation models that were altered for nano materials. Successful prediction of
the foam yield strength is very challenging due to the irregular cell geometry, surface
effects, and node effects. However, by examining the foam hardness values of
various metal systems with a wide range of ligament sizes, the deformation mode,
regardless of the type of metal, seems to more closely follow the plastic collapse
models.
The main contributions of this study are summarized as following:
94

 Several precursor alloy systems which meet the basic dealloying
requirements were utilized for the synthesis of NP foams, but not every
precursor alloy produced a NP foam. As observed previously the relationship
between dealloying conditions and resulting foam microstructures for non-
noble metal foams demonstrates that the part limit, dealloying threshold, and
potential differences of the precursor alloy strongly influence the synthesis of
NP foams. In addition, the microstructure was shown to depend on the
formation of a passive layer, the microstructure of the precursor alloys, and
the oxidized state of the metal elements involved in the dealloying process.
 On the basis of the above information, we have successfully synthesized NP
Au, Ag, Cu, and Pd foams using different dealloying techniques. Ligament
sizes ranging from 15-220 nm were measured using SEM. The cross-
sectional SEM images demonstrate that the final foam structure depends on
the precursor alloys.
 Raman Spectra and XRD patterns of NP Ag and Cu foams show that oxides
can be detected in the foam structures after a heat treatment. For Cu foams,
both CuO and Cu
2
O were observed for all heat treatments above 200 °C, with
increasing temperatures resulting in thicker oxide layers. The effect of
oxides was further evaluated using nanoindentation testing for the Cu
samples, with hardness values increasing ~ 40 times upon heat treatment.
However, the oxide layer of NP Ag foams was limited to the surface, and
95

decomposed after a 600 °C heat treatment. The thermal stability of NP Ag
was found to be higher than that of other metal foams.
 The yield strengths of NP pure metal foams, such as Cu, Ag, and Pd foams,
were also measured using indentation tests. Since the majority of the work
concerning mechanical behavior was studied using NP Au, this is the first
reported comprehensive study of yield strength with multiple metal systems.
 In order to verify the failure mode of the nanoporous foams, the hardness
values were compared to Gibson-and-Ashby equations and two deformation
models proposed for the foam structure: elastic buckling and plastic collapse.
These models were further modified to account for size and surface effects.
Both Euler and Timoshenko beam models were utilized to develop the
models. Overall, the deformation mode, regardless of the type of metal,
seems to more closely follow the plastic collapse models. Deviations from
the models are attributed to the observed irregular cell geometry and relative
density estimations.

96

Bibliography
[1] M. F. Al-Kuhaili, J. P h y s. D Appl. Phys. 40 (2007) p.2847.

[2] E. Amsterdam, H. van Hoorn, J. T. M. De Hosson and P. R. Onck, Adv. Eng.
Mater. 10 (2008) p.877.

[3] A. Antoniou, D. Bhattacharrya, J. K. Baldwin, P. Goodwin, M. Nastasi, S. T.
Picraux and A. Misra, Appl. Phys. Lett. 95 (2009) p.073116.

[4] V. Bansal, H. Jani, J. Du Plessis, P. J. Coloe and S. K. Bhargava, Adv. Mater.
20 (2008) p.717.

[5] B. Beverskog and I. Puigdomenech, J. Electrochem. Soc. 144 (1997) p.3476.

[6] J. Biener, A. M. Hodge and A. V. Hamza, Appl. Phys. Lett. 87 (2005)
p.121908.

[7] J. Biener, A. M. Hodge, A. V. Hamza, L. M. Hsiung and J. H. Satcher, J.
Appl. Phys. 97 (2005) p.024301.

[8] J. Biener, A. M. Hodge, J. R. Hayes, C. A. Volkert, L. A. Zepeda-Ruiz, A. V.
Hamza and F. F. Abraham, Nano Lett. 6 (2006) p.2379.

[9] J. Biener, G. W. Nyce, A. M. Hodge, M. M. Biener, A. V. Hamza and S. A.
Maier, Adv. Mate. 20 (2008) p.1211.

[10] J. Biener, A. Wittstock, L. A. Zepeda-Ruiz, M. M. Biener, V. Zielasek, D.
Kramer, R. N. Viswanath, J. Weissmuller, M. Baumer and A. V. Hamza, Nat.
Mater. 8 (2009) p.47.

[11] L. Y. Chen, J. S. Yu, T. Fujita and M. W. Chen, Adv. Funct. Mater. 19
(2009) p.1221.

[12] I. C. Cheng and A. M. Hodge, Phil. Mag. Lett. (2012) submitted

[13] I. C. Cheng and A. M. Hodge, Adv. Eng. Mater. 14 (2012) p.219.

[14] S. G. Corcoran, S. R. Brankovic, N. Dimitrov and K. Sieradzki, in Thin-
Films - Stresses and Mechanical Properties Vii, Vol. 505 (R. C. Cammarata,
M. Nastasi, E. P. Busso, and W. C. Oliver, eds.), Materials Research Society,
Warrendale, 1998, p. 77.

97

[15] C. A. Daniels, Metals structure and properties, Abyss books, Washington,
D.C., 2006.

[16] J. R. Davis, Copper and copper alloys, ASM International, Material Park,
OH, 2001.

[17] E. Detsi, E. De Jong, A. Zinchenko, Z. Vukovic, I. Vukovic, S. Punzhin, K.
Loos, G. ten Brinke, H. A. De Raedt, P. R. Onck and J. T. M. De Hosson,
Acta Mater. 59 (2011) p.7488.

[18] E. Detsi, Z. Vukovic, S. Punzhin, P. M. Bronsveld, P. R. Onck and J. T. M.
De Hosson, Crystengcomm 14 (2012) p.5402.

[19] Y. Ding and M. W. Chen, MRS Bull. 34 (2009) p.569.

[20] Y. Ding, Y. J. Kim and J. Erlebacher, Adv. Mater. 16 (2004) p.1897.

[21] R. Dou and B. Derby, J. Mater. Res. 25 (2010) p.746.

[22] J. Erlebacher, M. J. Aziz, A. Karma, N. Dimitrov and K. Sieradzki, Nature
410 (2001) p.450.

[23] J. Erlebacher and R. Seshadri, MRS Bull. 34 (2009) p.561.

[24] H. L. Fan and D. N. Fang, Mater. Design 30 (2009) p.1441.

[25] X. Q. Feng, R. Xia, X. Li and B. Li, Appl. Phys. Lett. 94 (2009) p.011916.

[26] J. R. Ferraro, K. Nakamoto and C. W. Brown, Introductory Raman
Spectroscopy, Academic Press, San Diego, CA, 2003.

[27] Y. Gan and J. K. Chen, Appl. Phys. A-Mater. 95 (2009) p.357.

[28] Z. H. Gan, G. Q. Yu, B. K. Tay, C. M. Tan, Z. W. Zhao and Y. Q. Fu, J.
Phys. D Appl. Phys. 37 (2004) p.81.

[29] L. J. Gibson and M. F. Ashby, Cellular solids: structure and properties,
Cambridge University Press, Cambridge, New York, 2001.

[30] S. Grugeon, S. Laruelle, R. Herrera-Urbina, L. Dupont, P. Poizot and J. M.
Tarascon, J. Electrochem. Soc. 148 (2001) p.A285.

[31] M. Hakamada and M. Mabuchi, J. Alloys Compd. 479 (2009) p.326.

98

[32] M. Hakamada and M. Mabuchi, Scr. Mater. 56 (2007) p.1003.

[33] M. Hakamada and M. Mabuchi, Nano Lett. 6 (2006) p.882.

[34] M. Hakamada and M. Mabuchi, J. Alloy Compd. 485 (2009) p.583.

[35] M. Hakamada and M. Mabuchi, Mater. Trans. 50 (2009) p.431.

[36] M. Hakamada, H. Nakano, T. Furukawa, M. Takahashi and M. Mabuchi, J.
Phys. Chem. C 114 (2010) p.868.

[37] M. Hakamada, M. Takahashi, T. Furukawa and M. Mabuchi, Phil. Mag. 90
(2010) p.1915.

[38] R. B. Han, H. Wu, C. L. Wan and W. Pan, Scr. Mater. 59 (2008) p.1047.

[39] M. Haruta, Chemphyschem 8 (2007) p.1911.

[40] J. R. Hayes, A. M. Hodge, J. Biener, A. V. Hamza and K. Sieradzki, J. Mater.
Res. 21 (2006) p.2611.

[41] A. M. Hodge, J. Biener, J. R. Hayes, P. M. Bythrow, C. A. Volkert and A. V.
Hamza, Acta Mater. 55 (2007) p.1343.

[42] A. M. Hodge, J. Biener, L. L. Hsiung, Y. M. Wang, A. V. Hamza and J. H.
Satcher, J. Mater. Res. 20 (2005) p.554.

[43] A. M. Hodge and D. C. Dunand, Metall. Mater. Trans. A-Phys. Metall.
Mater. Sci. 34A (2003) p.2353.

[44] A. M. Hodge, J. R. Hayes, J. A. Caro, J. Biener and A. V. Hamza, Adv. Eng.
Mater. 8 (2006) p.853.

[45] A. M. Hodge, Y. M. Wang and T. W. Barbee, Scr. Mater. 59 (2008) p.163.

[46] H. Ji, X. G. Wang, C. C. Zhao, C. Zhang, J. L. Xu and Z. H. Zhang,
Crystengcomm 13 (2011) p.2617.

[47] F. L. Jia, C. F. Yu, K. J. Deng and L. Z. Zhang, J. Phys. Chem. C 111 (2007)
p.8424.

[48] X. C. Jiang, T. Herricks and Y. N. Xia, Nano Lett. 2 (2002) p.1333.

99

[49] H. J. Jin, L. Kurmanaeva, J. Schmauch, H. Rosner, Y. Ivanisenko and J.
Weissmuller, Acta Mater. 57 (2009) p.2665.

[50] H. J. Jin and J. Weissmuller, Adv. Eng. Mater. 12 (2010) p.714.

[51] H. J. Jin and J. Weissmuller, Science 332 (2011) p.1179.

[52] B. V. L'Vov, Thermochim. Acta 333 (1999) p.13.

[53] M. H. Lee, K. T. Kim, T. Gemming, D. J. Sordelet and J. Eckert, J. Appl.
Phys. 108 (2010) p.064303.

[54] L. P. Lefebvre, J. Banhart and D. C. Dunand, Adv. Eng. Mater. 10 (2008)
p.775.

[55] G. J. Li, X. P. Song, Z. B. Sun, S. C. Yang, B. J. Ding, S. Yang, Z. M. Yang
and F. Wang, Solid State Sci. 13 (2011) p.1379.

[56] W. J. Li, H. Y. Ma, L. H. Huang and Y. Ding, Phys. Chem. Chem. Phys. 13
(2011) p.5565.

[57] Z. Q. Li, D. Y. Wang, B. Q. Li and X. Lu, J. Electrochem. Soc. 157 (2010)
p.K223.

[58] D. R. Lide (ed.), CRC Handbook of Chemistry and Physics, CRC press, Ann
Arbor, 1995.

[59] R. Liu and A. Antoniou, Scr. Mater. 67 (2012) p.923.

[60] W. B. Liu, S. C. Zhang, N. Li, J. W. Zheng and Y. L. Xing, J. Electrochem.
Soc. 158 (2011) p.D91.

[61] W. B. Liu, S. C. Zhang, N. Li, J. W. Zheng and Y. L. Xing, Micropor.
Mesopor. Mat. 138 (2011) p.1.

[62] W. B. Liu, S. C. Zhang, N. Li, J. W. Zheng and Y. L. Xing, Corrosion Sci. 53
(2011) p.809.

[63] Y. L. Liu, L. Liao, J. C. Li and C. X. Pan, J. Phys. Chem. C 111 (2007)
p.5050.

[64] R. M. Lowe, Acta Metall. Mater. 12 (1964) p.1111.

100

[65] M. Lucas, A. M. Leach, M. T. McDowell, S. E. Hunyadi, K. Gall, C. J.
Murphy and E. Riedo, Phys. Rev. B 77 (2008) p.245420.

[66] T. B. Massalski, H. Okamoto, P. R. Subramanian and L. Kacprzak (ed.),
Binary alloy phase diagrams, ASM international, Materials Park, Ohio, 2001.

[67] A. Mathur and J. Erlebacher, Appl. Phys. Lett. 90 (2007) p.061910.

[68] T. Matsuda and T. Kizuka, Jpn. J. Appl. Phys. Part 1 - Regul. Pap. Brief
Commun. Rev. Pap. 46 (2007) p.4370.

[69] M. T. McDowell, A. M. Leach and K. Gaill, Nano Lett. 8 (2008) p.3613.

[70] A. J. Nagy, G. Mestl, D. Herein, G. Weinberg, E. Kitzelmann and F. Schlogl,
J. Catal. 182 (1999) p.417.

[71] S. Parida, D. Kramer, C. A. Volkert, H. Rosner, J. Erlebacher and J.
Weissmuller, Phys. Rev. Lett. 97 (2006) p.035504.

[72] H. W. Pickering, Corrosion Sci. 23 (1983) p.1107.

[73] H. W. Pickering, Mater. Sci. Eng. A-Struct. Mater. Prop. Microstruct.
Process. 198 (1995) p.213.

[74] J. F. Pierson and C. Rousselot, Surf. Coat. Tech. 200 (2005) p.276.

[75] M. Pourbaix, Lectures on electrochemical corrosion, Plenum Press, New
York-London, 1973.

[76] M. Pourbaix, J. Electrochem. Soc. 123 (1976) p.C25.

[77] D. V. Pugh, A. Dursun and S. G. Corcoran, J. Mater. Res. 18 (2003) p.216.

[78] Z. Qi, C. C. Zhao, X. G. Wang, J. K. Lin, W. Shao, Z. H. Zhang and X. F.
Bian, J. Phys. Chem. C 113 (2009) p.6694.

[79] H. J. Qiu, Z. H. Zhang, X. R. Huang and Y. B. Qu, Chemphyschem 12
(2011) p.2118.

[80] M. Ramirez, L. Henneken and S. Virtanen, Appl. Surf. Sci. 257 (2011)
p.6481.

[81] P. N. Reddy, A. Sreedhar, M. H. P. Reddy, S. Uthanna and J. F. Pierson,
Cryst. Res. Technol. 46 (2011) p.961.
101

[82] S. Sakai, H. Tanimoto and H. Mizubayashi, Acta Mater. 47 (1998) p.211.

[83] G. V. Samsonov, The Oxide Handbook, IFI/Plenum, New York, 1973.

[84] P. G. Sanders, J. A. Eastman and J. R. Weertman, Acta Mater. 45 (1997)
p.4019.

[85] J. M. J. Santillan, L. B. Scaffardi, D. C. Schinca and F. A. Videla, J. Opt. 12
(2010) p.045002.

[86] R. Schennach and A. Gupper, in Materials, Technology and Reliability for
Advanced Interconnects and Low-K Dielectrics-2003, Vol. 766 (A. J.
McKerrow, J. Leu, O. Kraft, and T. Kikkawa, eds.), Materials Research
Society, Warrendale, 2003, p. 427.

[87] E. Seker, J. T. Gaskins, H. Bart-Smith, J. Zhu, M. L. Reed, G. Zangari, R.
Kelly and M. R. Begley, Acta Mater. 55 (2007) p.4593.

[88] N. A. Senior and R. C. Newman, Nanotechnology 17 (2006) p.2311.

[89] N. A. M. Shanid and M. A. Khadar, Thin Solid Films 516 (2008) p.6245.

[90] C. J. Shute, B. D. Myers, S. Xie, T. W. Barbee, A. M. Hodge and J. R.
Weertman, Scr. Mater. 60 (2009) p.1073.

[91] K. Sieradzki, J. Electrochem. Soc. 140 (1993) p.2868.

[92] K. Sieradzki, R. R. Corderman, K. Shukla and R. C. Newman, Philos. Mag.
A-Phys. Condens. Matter Struct. Defect Mech. Prop. 59 (1989) p.713.

[93] K. Sieradzki, N. Dimitrov, D. Movrin, C. McCall, N. Vasiljevic and J.
Erlebacher, J. Electrochem. Soc. 149 (2002) p.B370.

[94] T. T. Song, Y. L. Gao, Z. H. Zhang and Q. J. Zhai, Crystengcomm 13 (2011)
p.7058.

[95] Y. Sun and T. J. Balk, Scr. Mater. 58 (2008) p.727.

[96] Y. Sun, K. P. Kucera, S. A. Burger and T. J. Balk, Scr. Mater. 58 (2008)
p.1018.

[97] S. Tamilmani, W. Huang, S. Raghavan and R. Small, J. Electrochem. Soc.
149 (2002) p.G638.
102

[98] X. L. Tan, K. Li, Y. Liu, W. D. Wu, J. S. Luo and Y. J. Tang, Rare Metal
Mat. Eng. 39 (2010) p.2011.

[99] B. C. Tappan, S. A. Steiner and E. P. Luther, Angew. Chem.-Int. Edit. 49
(2010) p.4544.

[100] C. Tekoglu and P. R. Onck, J. Mech. Phys. Solids 56 (2008) p.3541.

[101] S. Van Petegem, S. Brandstetter, R. Maass, A. M. Hodge, B. S. El-Dasher, J.
Biener, B. Schmitt, C. Borca and H. Van Swygenhoven, Nano Lett. 9 (2009)
p.1158.

[102] P. W. Voorhees, J. Stat. Phys. 38 (1985) p.231.

[103] T. Wada, A. D. Setyawan, K. Yubuta and H. Kato, Scr. Mater. 65 (2011)
p.532.

[104] X. G. Wang, Z. Qi, C. C. Zhao, W. M. Wang and Z. H. Zhang, J. Phys.
Chem. C 113 (2009) p.13139.

[105] X. G. Wang, W. M. Wang, Z. Qi, C. C. Zhao, H. Ji and Z. H. Zhang, J. Alloy
Compd. 508 (2010) p.463.

[106] X. S. Wang and R. Xia, Europhys. Lett. 92 (2010) p.16004.

[107] G. I. N. Waterhouse, G. A. Bowmaker and J. B. Metson, Phys. Chem. Chem.
Phys. 3 (2001) p.3838.

[108] J. Weissmuller, R. N. Viswanath, D. Kramer, P. Zimmer, R. Wurschum and
H. Gleiter, Science 300 (2003) p.312.

[109] A. Wittstock, J. Biener and M. Baumer, Phys. Chem. Chem. Phys. 12 (2010)
p.12919.

[110] A. Wittstock, V. Zielasek, J. Biener, C. M. Friend and M. Baumer, Science
327 (2010) p.319.

[111] B. Wu, A. Heidelberg, J. J. Boland, J. E. Sader, X. M. Sun and Y. D. Li,
Nano Lett. 6 (2006) p.468.

[112] R. Xia, X. D. Li, Q. H. Qin, J. L. Liu and X. Q. Feng, Nanotechnology 22
(2011)

103

[113] C. X. Xu, Y. Y. Li, F. Tian and Y. Ding, Chemphyschem 11 (2010) p.3320.

[114] M. Yin, C. K. Wu, Y. B. Lou, C. Burda, J. T. Koberstein, Y. M. Zhu and S.
O'Brien, J. Am. Chem. Soc. 127 (2005) p.9506.

[115] Y. H. Yue, P. Liu, Z. Zhang, X. D. Han and E. Ma, Nano Lett. 11 (2011)
p.3151.

[116] Z. H. Zhang, Y. Wang, Z. Qi, W. H. Zhang, J. Y. Qin and J. Frenzel, J. Phys.
Chem. C 113 (2009) p.12629.

[117] C. C. Zhao, Z. Qi, X. G. Wang and Z. H. Zhang, Corrosion Sci. 51 (2009)
p.2120.

[118] C. C. Zhao, X. G. Wang, Z. Qi, H. Ji and Z. H. Zhang, Corrosion Sci. 52
(2010) p.3962.

[119] M. Zhou, Y. A. Gao, B. X. Wang, Z. Rozynek and J. O. Fossum, Eur. J.
Inorg. Chem. (2010) p.729.

[120] V. Zielasek, B. Jurgens, C. Schulz, J. Biener, M. M. Biener, A. V. Hamza and
M. Baumer, Angew. Chem.-Int. Edit. 45 (2006) p.8241.

104

Appendices
Appendix A: Summary of Ti samples

Sample
name
Composition
Percentage
(at%)
Electrolyte Applied
Potential
(mV)
Dealloying
Time (hr)
Sample wt%
after
dealloying
AJ-1 Ti
50
Al
50
70% HNO
3
N/A 96 97%
AJ-2 Ti
50
Al
50
38% HCl N/A 24 Totally
dissolved
AJ-3 Ti
50
Al
50
38% HCl N/A 48 Totally
dissolved
AJ-4 Ti
50
Al
50
85% H
3
PO
4
N/A 24 98%
AJ-5 Ti
50
Al
50
2M H
2
SO
4
N/A 24 94%
AJ-6 Ti
50
Al
50
38% HCl N/A 3 66%
IC-4 Ti
35
Al
65
1M NaOH 2800 13.5 100%*
IC-5 Ti
50
Al
50
1M NaOH 2800 18 100%*
IC-6 Ti
90
Al
6
V
4
1M NaOH 2800 4 102%
IC-7 Ti
90
Al
6
V
4
1M NaOH 2900 7 102%
IC-8 Ti
90
Al
6
V
4
1M NaOH 2800 18 100%*
IC-9 Ti
90
Al
6
V
4
10M NaOH 2800 13.5 101%
IC-10 Ti
90
Al
6
V
4
10M NaOH 2800 22.5 102%
IC-11 Ti
35
Al
65
1M NaOH 2800 36 100%*
IC-12 Ti
35
Al
65
10M NaOH 2800 17.5 100%*
IC-13 Ti
50
Al
50
10M NaOH 2800 12 101%
IC-14 Ti
35
Al
65
5M HNO
3
2400 42 67%
IC-15 Ti
50
Al
50
5M HNO
3
2600 21 87%
IC-16 Ti
90
Al
6
V
4
38% HCl 2800 40 97%
IC-17 Ti
90
Al
6
V
4
1M NaOH 2800 44 101%
IC-18 Ti
90
Al
6
V
4
1M NaOH
80°C
2800 16 101%
IC-19 Ti
90
Al
6
V
4
38% HCl 2800 4.5 60%
IC-20 Ti
35
Al
65
1M NaOH 2800 22 Too brittle
IC-21 Ti
90
Al
6
V
4
1M NaOH 2800 17.5 106%
105

Appendix A: Summary of Ti samples – Continue

*100% means no weight loss

Sample
name
Composition
Percentage
(at%)
Electrolyte Applied
Potential
(mV)
Dealloying
Time (hr)
Sample wt%
after
dealloying
IC-22 Ti
90
Al
6
V
4
1M NaOH 2800 0.5 158%
IC-23 Ti
90
Al
6
V
4
1M NaOH
65°C
1500 1.25 98%
IC-24 Ti
90
Al
6
V
4
1M NaOH
65°C
1500 4 97%
IC-25 Ti
50
Al
50
1M NaOH
65°C
1500 2.5 100%*
106

Appendix B: Summary of Cu samples

Sample
name
Composition
Percentage
Electrolyte Applied
Potential
(mV)
Dealloying
Time (hr)
Sample wt%
after
dealloying
IC-26 Cu
20
Zn
80
1M NaOH NA 43 98%
IC-27 Cu
35
Zn
65
1M NaOH NA 43 100%
IC-28 Cu
30
Al
70
38% HCl NA 16 dissolved
IC-31 Cu
30
Al
70
1M NaOH NA 84 59%
IC-32 Cu
20
Zn
80
70% HNO
3
NA 0.5 dissolved
IC-33 Cu
35
Zn
65
70% HNO
3
NA 0.5 dissolved
IC-36 Cu
30
Mn
70
1M
(NH
4
)
2
SO
4

NA 75 42%
IC-43 Cu
20
Zn
80
7.5% HCl NA 132 29%
IC-44 Cu
35
Zn
65
7.5% HCl NA 132 96%
IC-45 Cu
30
Al
70
1M NaOH NA 132 56%
IC-46 Cu
30
Mn
70
1M
(NH
4
)
2
SO
4

NA 132 41%
IC-54 Cu
30
Al
70
5% HCl NA 16 53%
IC-58 Cu
30
Mn
70
1M
(NH
4
)
2
SO
4

-650 19.5 48%
IC-59 Cu
30
Al
70
5% HCl -500 3.5 60%
IC-60 Cu
30
Al
70
5% HCl NA 20 48%
IC-61 Cu
30
Al
70
5% HCl NA 22 64%
IC-64 Cu
30
Al
70
5% HCl NA 24 52%
IC-69 Cu
30
Al
70
5% HCl -500 13.5 52%
IC-70 Cu
30
Al
70
5% HCl NA 24 55%
IC-80 Cu
30
Al
70
5% HCl NA 45 55%
IC-81 Cu
30
Al
70
5%HCl 200 0.5 Become
powders
IC-82 Cu
30
Al
70
5%HCl 50 0.5 49%
IC-83 Cu
20
Zn
80
85%H
3
PO
4
NA 46 24%
IC-84 Cu
30
Al
70
5%HCl NA 48 50%
IC-85 Cu
30
Mn
70
1M
(NH
4
)
2
SO
4

NA 24 46%
107

Appendix B: Summary of Cu samples - Continue

Sample
name
Composition
Percentage
Electrolyte Applied
Potential
(mV)
Dealloying
Time (hr)
Sample wt%
after
dealloying
IC-86 Cu
20
Zn
80
85%H
3
PO
4
-200 18 50%
IC-89 Cu
20
Zn
80
5%HCl NA 51 74%
IC-90 Cu
20
Zn
80
85%H
3
PO
4
NA 51 39%
IC-91 Cu
30
Al
70
5%HCl -200 2.5 48%
IC-92 Cu
35
Zn
65
5%HCl -200 1.5 35%
IC-93 Cu
20
Zn
80
5%HCl -200 3 17%
IC-94 Cu
20
Zn
80
5%HCl -200 3 21%
IC-95 Cu
20
Zn
80
5%HCl -200 2 27%
IC-96 Cu
35
Zn
65
5%HCl -200 1.5 35%
IC-97 Cu
35
Zn
65
5%HCl -200 2 34%
IC-100 Cu
20
Zn
80
85%H
3
PO
4
-200 19.5 49%
IC-101 Cu
20
Zn
80
85%H
3
PO
4
NA 64 20%
IC-102 Cu
30
Al
70
5%HCl NA 29 53%

108

Appendix C: Summary of Ag samples

Sample
name
Composition
Percentage
Electrolyte Applied
Potential
(mV)
Dealloying
Time (hr)
Sample wt%
after
dealloying
IC-29 Ag
25
Al
75
38% HCl
90°C
NA 16 dissolved
IC-30 Ag
25
Al
75
1M NaOH NA 84 67%
IC-35 Ag
90
Cu
10
70% HNO
3
NA 0.5 dissolved
IC-38 Ag
10
Cu
90
1M NaOH NA 69 100%
IC-50 Ag
25
Al
75
1M NaOH NA 132 66%
IC-51 Ag
10
Cu
90
5% HCl NA 132 98%
IC-53 Ag
25
Al
75
5% HCl NA 16 99%
IC-55 Ag
25
Al
75
5% HCl NA 16 61%
IC-57 Ag
25
Al
75
5% HCl 100 0.6 70%
IC-66 Ag
25
Al
75
5% HCl NA 24 65%
IC-67 Ag
25
Al
75
5% HCl NA 24 65%
CuAg1 Ag
10
Cu
90
85% H
3
PO
4
NA 72 99%
CuAg2 Ag
10
Cu
90
H
2
SO
4
NA 72 100%
CuAg3 Ag
10
Cu
90
10% H
2
O
2
NA 72 100%
IC-103 Ag
25
Al
75
5% HCl NA 66 64%
IC-114 Ag
25
Al
75
5% HCl NA 48 63%
IC-115 Ag
25
Al
75
5% HCl -100 1.5 65%
IC-116 Ag
25
Al
75
5% HCl NA 1.5 68%
IC-117 Ag
25
Al
75
5% HCl NA 67 61%
IC-118 Ag
25
Al
75
5% HCl NA 4 58%
IC-119 Ag
25
Al
75
1M NaOH NA 4 58%
IC-120 Ag
25
Al
75
5% HCl -100 0.8 66%

109

Appendix D: Summary of Pd samples

Sample
name
Composition
Percentage
Electrolyte Applied
Potential
(mV)
Dealloying
Time (hr)
Sample wt%
after
dealloying
IC-34 Pd
30
Ag
70
70% HNO
3
NA 7 dissolved
IC-39 Pd
30
Ag
70
1M NaOH NA 48 100%
IC-49 Pd
30
Ag
70
5% HCl NA 132 97%
IC-108 Pd
30
Ag
70
10% HNO
3
NA 21 100%
IC-39 Pd
30
Ag
70
20% HNO
3
1000 20 95%
IC-109 Pd
30
Ag
70
70% HNO
3
1100 2 dissolved
IC-129 Pd
30
Al
70
5% HCl NA 21 73%
IC-134 Pd
30
Al
70
1M NaOH NA 43 87%
IC-134 Pd
30
Al
70
5% HCl NA 24 92%
IC-137 Pd
30
Ag
70
5% HCl 200 1.5 68%
IC-138 Pd
30
Al
70
5% HCl 100 1 65%
IC-139 Pd
30
Al
70
5% HCl 100 1 66%
IC-140 Pd
30
Al
70
5% HCl 100 1.25 73%
IC-141 Pd
30
Al
70
5% HCl 100 1.25 72%
IC-142 Pd
30
Al
70
5% HCl 100 1 62%

110

Appendix E: Summary of Ni samples

Sample
name
Composition
Percentage
Electrolyte Applied
Potential
(mV)
Dealloying
Time (hr)
Sample wt%
after
dealloying
IC-37 Ni
12
Mn
88
1M
(NH
4
)
2
SO
4

NA 25 become
powders
IC-40 Ni
36
Fe
64
1M
(NH
4
)
2
SO
4

NA 51 96%
IC-41 Ni
12
Mn
88
1M
(NH
4
)
2
SO
4

NA 12 become
powders
IC-42 Ni
36
Fe
64
1M NaOH NA 72 99%
IC-47 Ni
12
Mn
88
1M
(NH
4
)
2
SO
4

NA 1 43%
IC-48 Ni
36
Fe
64
70% HNO
3
NA 132 81%
IC-107 Ni
36
Fe
64
85% H3PO4 NA 21 100%
IC-107 Ni
36
Fe
64
85% H3PO4 -100 20 21%
IC-110 Ni
12
Mn
88
1M
(NH
4
)
2
SO
4

1100 1 become
powders
IC-121 Ni
30
Mn
70
1M
(NH
4
)
2
SO
4

NA 67 48%
IC-122 Ni
30
Mn
70
1M
(NH
4
)
2
SO
4

-650 22 41%
IC-123 Ni
30
Mn
70
1M
(NH
4
)
2
SO
4

-650 24 41%
IC-124 Ni
30
Mn
70
1M
(NH
4
)
2
SO
4

-650 15 42%
IC-125 Ni
30
Mn
70
1M
(NH
4
)
2
SO
4

-50 2 13%
IC-126 Ni
30
Mn
70
1M
(NH
4
)
2
SO
4

-500 6 24%
IC-127 Ni
30
Mn
70
1M
(NH
4
)
2
SO
4

-600 9 39%
IC-128 Ni
30
Mn
70
1M
(NH
4
)
2
SO
4

-550 6 19%
IC-130 Ni
30
Mn
70
1M
(NH
4
)
2
SO
4

-600 5 45%
IC-135 Ni
30
Mn
70
1M NaOH 600 5 99%
IC-136 Ni
30
Mn
70
85% H3PO4 NA 1 dissolved
111

Appendix F: Summary of Zn samples

Sample
name
Composition
Percentage
Electrolyte Applied
Potential
(mV)
Dealloying
Time (hr)
Sample wt%
after
dealloying
IC-73 Zn
35
Al
65
85%
H
3
PO
4

NA 48 100%
IC-111 Zn
50
Al
50
1M
(NH
4
)
2
SO
4

NA 52 100%
IC-112 Zn
50
Al
50
1M NaOH NA 52 62%

Abstract (if available)

Abstract Nanoporous (NP) metal foams, including Au, Cu, Ag, and Pd, have been successfully synthesized using different processes, such as free corrosion, electrochemically driven dealloying, and subsequent heat treatments. To investigate the effect of the synthesis processes on the yield strength and microstructure, the hardness of NP Au, Cu, Ag, and Pd foams with ligament sizes ranging from 20-220 nm was measured using nanoindentation and micro Vickers hardness testing, while the morphology and ligament sizes were determined by scanning electron microscopy (SEM). The results demonstrate that for all these materials, the yield strength of the NP foams was higher than that predicted by macroporous foam scale equations. In addition, combining the results with previous data for NP Au, it was observed that the yield strength of foams increases as ligament size decreases, which emphasizes the significant effect of the ligament size. The results were also compared to several predictions for foam deformation models, including elastic buckling and plastic collapse models. Overall, the deformation mode, regardless of the type of metal, seems to more closely follow the plastic collapse models. Deviations from the models are attributed to irregular cell geometry and relative density estimations. ❧ Additionally, the effect of oxide formation in relation to the NP foam structure for Cu and Ag was investigated as a function of heat treatments. The formation of oxide was characterized using X-ray diffraction (XRD) and Raman spectroscopy. The variations in ligament size, morphology and yield strength were investigated for all heat treated foams. For NP Cu foams, the results demonstrate that the formation of oxides was observed after any heat treatments above 200 °C under argon. These Cu oxides initiated at the surface layer and can grow in thickness with increasing heat treatment temperature, leading to a complete oxide foam. It was determined that the yield strength of Cu oxide foams increased ~ 40 times upon heat treatments, as compared to Cu foams. For the NP Ag foams, oxidation was observed after heat treatments in argon, but was limited to only a surface layer. This thin oxide layer decomposed into silver and oxygen at 600 °C. Furthermore, NP Ag foams appear to show relatively high thermal stability compared to other metal foams. The heat treatments on NP Ag foams were not as effective at changing the ligament size as the use of different electrolytes. However, the results demonstrated that both Cu and Ag foams retain their structural integrity even after oxidation.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Metallic syntactic foams synthesis, characterization and mechnical properties

PDF

Development and characterization of hierarchical cellular structures

PDF

Synthesis and mechanical behavior of highly nanotwinned metals

PDF

Structure and behavior of nano metallic multilayers under thermal and mechanical loading

PDF

Modeling of bicontinuous nanoporous materials and investigation of their mechanical and hydraulic properties

PDF

Mechanical behavior and deformation of nanotwinned metallic alloys

PDF

Processing and properties of phenylethynyl-terminated PMDA-type asymmetric polyimide and composites

PDF

Mechanical behavior and microstructure optimization of ultrafine-grained aluminum alloys and nanocomposites

PDF

Microstructural evolution and mechanical properties of a copper-zirconium alloy processed by severe plastic deformation

PDF

Processing, mechanical behavior and biocompatibility of ultrafine grained zirconium fabricated by accumulative roll bonding

PDF

The role of nanotwins and grain boundary plane in the thermal, corrosion, and sensitization behavior of nanometals

$Fatigue and fracture of pultruded composite rods developed for overhead conductor cables$

PDF

Fatigue and fracture of pultruded composite rods developed for overhead conductor cables

PDF

Thermal properties of silicon carbide and combustion mechanisms of aluminum nanoparticle

PDF

High temperature creep behaviors of additively manufactured IN625

PDF

Shock wave response of in situ iron-based metallic glass matrix composites

PDF

Investigation of the influence of severe plastic deformation on the microstructure and mechanical properties of Al-7136 alloy

PDF

Synthesis and mechanical evaluation of micro-scale truss structures formed from self-propagating polymer waveguides

PDF

A comparative study of plasma conditions, microstructure and residual stress in sputtered thin films

PDF

Mechanical behavior of materials in extreme conditions: a focus on creep plasticity

PDF

The effect of lattice structure and porosity on thermal conductivity of additively-manufactured porous materials

Asset Metadata

Creator Cheng, I-Chung (author)

Core Title Synthesis, characterization, and mechanical properties of nanoporous foams

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Materials Science

Publication Date 03/28/2013

Defense Date 03/11/2013

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag dealloying,foams,mechanical properties,nanoporous,OAI-PMH Harvest,thermal stability

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Hodge, Andrea M. (committee chair), Goo, Edward K. (committee member), Kassner, Michael E. (committee member), Nutt, Steven R. (committee member)

Creator Email ichungch@usc.edu,ichungcheng@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-228012

Unique identifier UC11292850

Identifier usctheses-c3-228012 (legacy record id)

Legacy Identifier etd-ChengIChun-1489.pdf

Dmrecord 228012

Document Type Dissertation

Rights Cheng, I-Chung

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA