copyright by gaurav gupta 2008
TRANSCRIPT
The Dissertation Committee for Gaurav Gupta
certifies that this is the approved version of the following dissertation:
Design of Novel Catalysts by Infusion of Presynthesized
Nanocrystals into Mesoporous Supports
Committee:
_________________________________ Keith P. Johnston, Supervisor _________________________________ Brian A. Korgel _________________________________ Arumugam Manthiram _________________________________ Richard M. Crooks _________________________________ Miguel Jose Yacaman
Design of Novel Catalysts by Infusion of Presynthesized
Nanocrystals into Mesoporous Supports
by
Gaurav Gupta, B.Tech.
Dissertation
Presented to the Faculty of the Graduate School of
the University of Texas at Austin
in Partial Fulfillment
of the Requirements
for the Degree of
Doctor of Philosophy
The University of Texas at Austin
August 2008
v
Acknowledgements
This thesis is the result of interactions and discussions with several people,
and I would like to take this opportunity to thank them individually. First I would like
to thank my advisor, Dr. Keith Johnston. In addition to the scientific guidance he
provided, he also encouraged exploring new and novel research ideas. I would also
like to thank my dissertation committee for their contribution towards this thesis.
This thesis is written in collaboration with many individuals and I would like
to thank them starting with Dr. Keith Stevenson who provided valuable guidance as
well as use of the lab equipments for the fuel cell work. I would like to thank Mehul
Patel for having intellectually stimulating discussions which significantly enriched
this thesis. Thanks to Dan Slanac, Pavan Kumar, Andrew Heitsch, Cynthia Stowell,
Dr. Aaron Saunders, Dr. Domingo Ferrer, Dr. Xiaoxia Gao, Dr. Parag Shah, Dr.
Steven Swinnea and Jaclyn Wiggins for their invaluable help in the several key
experiments. I also had the pleasure of having insightful discussions with Arindam
Sarkar, Dr. Hiroshi Uchida, Dr. Won Ryoo, and Corinne Atkinson.
Special thanks for the support and friendship of the lunch group involving
Manas Shah, Gaurav Goel, Mehul Patel, Amogh Prabhu, John Keagy, Dale Simpson
and Landry Khounlavong.
Exceptional thanks to my parents and my brother for their support, patience
and understanding throughout these years.
vi
Design of Novel Catalysts by Infusion of Presynthesized
Nanocrystals into Mesoporous Supports
Publication No. ___________
Gaurav Gupta, Ph.D.
The University of Texas at Austin, 2008
Supervisor: Keith P. Johnston
Traditionally, supported metal catalysts have been synthesized by reduction of
precursors directly over the support. In these techniques, it is challenging to control
the metal cluster size, composition and crystal structure. Herein, we have developed a
novel approach to design catalysts with controlled morphologies by infusing
presynthesized nanocrystals into the supports. High surface area mesoporous
materials, including graphitic carbons, have been utilized for obtaining a high degree
of metal dispersion to enhance catalyst stabilities and activities. Gold and iridium
nanocrystals have been infused in mesoporous silica with loadings up to 2 wt % using
supercritical CO2 as an antisolvent in toluene to enhance the van der Waals
interactions between nanocrystals and the silica. The iridium catalysts show high
catalytic activity and do not require high temperature annealing for ligand removal, as
ligands bind weakly to the iridium surface. To further enhance metal loadings to >10
% in the catalysts, short-ranged interactions between the metal nanocrystals and the
support are further strengthened with weakly binding ligands to expose more of the
metal surface to the support. For pre-synthesized FePt nanocrystals, coated with oleic
acid and oleylamine ligands, high loadings >10 wt % in mesoporous silica are
achieved, without using CO2. The strong metal-support interactions favor FePt
vii
adsorption on the support and also enhance stability against sintering at high
temperatures. High resistance to sintering favors formation of the FePt intermetallic
crystal structure with <4 nm size upon thermal annealing at 700 °C. The fundamental
understanding of the metal-support interactions gained from these studies is then
utilized in the design of highly stable Pt and Pt-Cu electrocatalysts with controlled
size, composition and alloy structure supported on graphitized mesoporous carbons
for oxygen reduction. The resistance of the graphitic carbons to oxidation coupled
with strong metal-support interactions mitigate nanoparticle isolation from the
support, nanoparticle coalescence, Pt dissolution and subsequent Ostwald ripening
and thus enhance catalyst stability. The control of the Pt nanocrystal morphology with
high concentrations of highly active (111) surface leads to 25% higher activities than
commercial Pt catalysts. Furthermore, the catalyst activities obtained for Pt-Cu
catalysts are 4-fold higher than Pt catalysts due to strained Pt shell generated from
electrochemical dealloying of copper from the nanoparticle surface.
viii
Table of Contents
List of Tables………………………………………………………………………...xii
List of Figures………......….………………………………………………………xiii
Chapter 1: Introduction ...………………………………………………………… 1 1.1 Nanocrystal Synthesis ……………………………….…………………. 3 1.2 Mesoporous Supports ………………………………………………….. 5 1.3 Infusion of Presynthesized Nanocrystals into Mesoporous Supports ...... 9 1.4 Electrocatalysts for Oxygen Reduction Reaction: Activity and Stability ……………………………………………………………….. 13 1.5 Objectives ……………………………………………………………... 14 1.6 Dissertation Outline………….…………………………………………15 1.7 References………………………………………………………………18
Chapter 2: Enhanced Infusion of Gold Nanocrystals into Mesoporous Silica with Supercritical Carbon Dioxide ………………………………….27 2.1 Introduction……………………………………………………………..28 2.2 Experimental……………………………………………………………32 2.2.1 Nanocrystal Synthesis………………………………………...32
2.2.2 Mesoporous Silica Synthesis………………………………….32 2.2.3 Infusion Procedure……………………………………………33 2.3 Results…………………………………………………………………..35 2.3.1 Nanoparticle Characterization………………………………..35 2.3.2 Characterization of Mesoporous Silica…………………….....37 2.3.3 Nanoparticle Dispersibility in Toluene-CO2............................43 2.3.4 Effect of CO2 on Loading of Gold Nanocrystals in Silica……44 2.3.5 Effect of Pore Size and Concentration on the Loading……….46 2.3.6 Kinetic Study of the Infusion Process………………………...46
2.3.7 TEM Micrographs of Infused Silica…………………………..49 2.3.6 Effect of Conventional Liquid Anti-Solvent on Infusion……..49
2.4 Discussion………………………………………………………………51 2.4.1 Equilibrium Nanocrystal Loading……….……………………28 2.4.2 Transport Aspects of Nanocrystal Loading…………………..52 2.5 Conclusions……………………………………………………………..58 2.6 References………………………………………………………………59 Chapter 3: Infusion of Presynthesized Iridium Nanocrystals into Mesoporous Silica for High Catalyst Activity……………………………………..65 3.1 Introduction……………………………………………………………..66
ix
3.2 Experimental……………………………………………………………69 3.2.1 Materials……………………………………………………...69 3.2.2 Iridium Nanocrystal Synthesis………………………………..70 3.2.3 Mesoporous Silica Synthesis ………………………………...70 3.2.4 Infusion of Ir into Mesoporous Silica with CO2..…………….71 3.2.5 Determination of Nanocrystal Precipitation Boundary..……...74 3.2.6 Catalysis ……………………………………………………...76 3.3 Characterization………………………………………………………..76 3.3.1 Nitrogen Porosimetry…………………………………………76 3.3.2 Transmission Electron Microscopy………………………..…76 3.3.3 Gas Chromatography/ Mass Spectroscopy…………………...77 3.4 Results and Discussion…………………………………………………77 3.4.1 Nanoparticle Characterization……………………………..…77 3.4.2 Characterization of Mesoporous Silica……………………….80 3.4.3 Nanocrystal Dispersibility in Toluene-CO2 Mixtures………...80 3.4.4 Nanocrystal Loading in Pores using CO2…….………………82 3.4.5 Visualization of Nanocrystals in Mesoporous Silica…………85 3.4.6 Decene Hydrogenation in Ir/MPS Catalysts………………….88 3.4.7 Turnover Number for uncalcined Ir/MPS…………………….91 3.4.8 Effect of Catalyst Pretreatment on Catalytic Activity………..91 3.4.9 Stability of Supported Nanocrystals during the Reaction…….94 3.4.10 Activation Energy of the Decene Hydrogenation…………...95 3.5 Conclusions……………………………………………………………..99 3.6 References……………………………………………………………..100
Chapter 4: Stable Ordered FePt-Mesoporous Silica Catalysts with High Loadings……………………………………………………………...106 4.1 Introduction……………………………………………………………107 4.2 Experimental…………………………………………………………..109 4.2.1 FePt Nanocrystal Synthesis…………………………………109 4.2.2 Mesoporous Silica Synthesis………………………………..110 4.2.3 Infusion of FePt in Mesoporous Silica………………………110 4.2.4 Calcination…………………………………………………..111 4.2.5 Transmission Electron Microscopy…………………………111 4.2.6 X-ray diffraction…………………………………………….111 4.2.7 Catalysis……………………………………………………..112 4.3 Results…………………………………………………………………112 4.3.1 Structure of Nanocomposites after Infusion of FePt………...112 4.3.2 Stability of the FePt Nanocomposites……………………….118 4.3.3 Crystal Structure of the Bimetallic FePt after Annealing…...120 4.3.4 Activity and Stability of Catalyst for Decene Hydrogenation123 4.4 Discussion……………………………………………………………..125 4.4.1 Mechanism for Metal loading and Stability…………………125
x
4.4.2 Design of Bimetallic Nanocrystals…….……………………127 4.4.3 Catalytic Activity of Supported Nanocrystals………………130 4.5 Conclusions……………………………………………………………131 4.6 References……………………………………………………………..132
Chapter 5: Highly Stable ORR Electrocatalysts from Presynthesized Pt Nanocrystals Supported on Graphitic Mesoporous Carbons…….141 5.1 Introduction……………………………………………………………142 5.2 Experimental Section………………………………………………….145 5.3 Results…………………………………………………………………150 5.4 Discussion……………………………………………………………..169 5.4.1 Carbon Oxidation and its Effect on Electrocatalyst Stability.169 5.4.2 Catalyst Synthesis and Stability by Pt-Carbon Interactions...170 5.4.3 Control of Pt Nanoparticle Morphology for High Activity…174 5.5 Conclusions……………………………………………………………175 5.6 References……………………………………………………………..176
Chapter 6: Highly Stable and Active Electrocatalysts for Oxygen Reduction: Presynthesized Bimetallic Nanocrystals on Graphitic Carbon…..183 6.1 Introduction……………………………………………………….…...184 6.2 Experimental…………………………………………………………..186 6.3 Results…………………………………………………………………190 6.3.1 Structure of the Pt-Cu on Mesoporous Carbon Catalysts…...190 6.3.2 Activity and Stability of Pt-Cu Catalysts for ORR………….202 6.4 Discussion……………………………………………………………..205 6.4.1 Carbon Corrosion and Impact on Stability………………….205 6.4.2 Metal-Support Interactions: Catalyst Synthesis and Stability.208 6.4.3 Dealloying – Structure and Activity Enhancement…………211 6.4.4 Composition of Bimetallic Nanocrystals before Dealloying 212 6.5 Conclusions……………………………………………………………214 6.6 References …………………………………………………………….215
Chapter 7: Conclusions and Recommendations………………………………..223 7.1 Conclusions……………………………………………………………223 7.1.1 Infusion of Metal Nanocrystals in Mesoporous Silica by CO2223 7.1.2 Synthesis of Stable and Ordered FePt Bimetallic Catalysts...224 7.1.3 Stable Pt Electrocatalysts Supported on Graphitic Carbon ...224 7.1.4 Highly Stable and Active Bimetallic Pt-Cu for PEMFC……225 7.2 Recommendations……………………………………………………..226 7.2.1 Design of Core-Shell Bimetallic Catalysts for ORR………..226 7.2.2 Design of Active Catalysts for Water-Gas Shift Reaction…..226 7.3 References……………………………………………………………..227
xi
Appendix A: Infusion of FePt Nanocrystals in Mesoporous Silica……………229 A.1 Characterization………………………………………………………229 A.2 Results and Discussion………………………………………………..230 A.3 References……………………………………………………………..241
Appendix B: Highly Stable and Active Pt-Cu Electrocatalysts for ORR……..243 B.1 Characterization………………………………………………………243 B.2 Results and Discussion………………………………………………..244 B.3 References……………………………………………………………..257
Bibliography……………………………………………………………………….258
Vita…………………………………………………………………………………284
xii
List of Tables
Table 2.1 Methods for loading of metal nanoparticles in mesoporous silica…..29
Table 2.2 Properties of mesoporous silica……………………………………...40
Table 2.3 Overall Hamaker constants A132 between silica surface and gold
core mediated by supercritical CO2-toluene solution. Component
1 is gold, 2 is SiO2, and 3 is a mixture of toluene and CO2………….54
Table 2.4 Overall Hamaker constants A132 between silica surface and gold
core mediated by methanol-toluene solution. Component 1 is gold
, 2 is SiO2, and 3 is a mixture of toluene and methanol……………...56
Table 3.1 Activity and turnover frequency for decene hydrogenation of 1.3%
Ir supported on mesoporous silica relative to 5 % Pd on alumina. ….92
Table 5.1 Properties of mesoporous carbon…………………………………...153
Table 5.2 Loadings of infusion of Pt nanoparticles in carbon supports……….156
Table 5.3 Weight loss of the carbon supports at 195°C air for 16 days due to
carbon oxidation catalyzed by platinum……………………………..162
Table 5.4 Electrochemical data for 20% Pt on mesoporous carbons before and
after ADT potential cycling between 0.5 – 1.2 V at 50 mV/s for 1000
cycles in 0.1M HClO4 saturated with argon………………………...163
Table 6.1 Properties of mesoporous carbon…………………………………...191
Table 6.2 Infusion of Pt-Cu nanoparticles in mesoporous carbon supports.
All infusions were done on 50 mg carbon and nanoparticles were
dispersed in 3 mL of toluene………………………………………..196
Table 6.3 Composition and structural analysis of 14% Pt20Cu80/DMC-2000
before and after dealloying…………………………………………201
Table 6.4 Activity and stability of dealloyed Pt-Cu catalysts supported on
mesoporous carbons………………………………………………...206
Table A.1 Loading of FePt nanocrystals in mesoporous silica as a function of
time and concentration……………………………………………...240
xiii
List of Figures
Figure 1.1 Different techniques for the synthesis of heterogeneous catalysts…….2
Figure 1.2 Arrested precipitation technique of nanocrystal synthesis…………….4
Figure 1.3 Synthesis of bimetallic catalysts via wetness-impregnation and
infusion of presynthesized nanocrystals into porous supports………...6
Figure 1.4 Synthesis approaches for mesoporous materials………………………8
Figure 1.5 Influence of CO2 concentration on nanoparticle loading……………..11
Figure 1.6 TEM images of FePt nanocrystals (3.6 nm) after infusion into
mesoporous silica (7 nm). (a) HAADF STEM planar view; (b)
HRTEM………………………………………………………………12
Figure 2.1 Schematic of reaction infusion apparatus……………………………34
Figure 2.2 Calibration curve of absorbance versus concentration measured at
500 nm for gold nanocrystal solutions……………………………….36
Figure 2.3 Intensity I(q) vs q for gold nanocrystals dispersed in toluene
determined by SAXS………………………………………………....38
Figure 2.4 SAXS for calcined mesoporous silica SBA-15 (a), L-MCM41 (b) and
S-MCM41 (c)………………………………………………………...39
Figure 2.5 N2 isotherms and pore size for SBA-15 (a), L-MCM41 (b), and
SBA-15 (c)…………………………………………………………...42
Figure 2.6 Loading of gold nanocrystals versus vol% CO2 for 24 hour
impregnation period in L-MCM41. () 1 atm and 23°C ;() 241 atm
and 35°C……………………………………………………………..45
Figure 2.7 Variation of loading of gold nanocrystals with the pore
diameter of silica……………………………………………………...47
Figure 2.8 Loading versus time for S-MCM41. (a) Loading versus t1/2 for
up to 24 hours) with a correlation coefficient of 99.5%. (b)
xiv
Loading versus time approaching equilibrium……………………....48
Figure 2.9 TEM of gold impregnated SBA-15 (a), L-MCM41 (b), and
S-MCM41 (c)………………………………………………………...50
Figure 3.1 Schematic of the infusion apparatus…………………………………72
Figure 3.2 Calibration curve of UV- Vis absorbance versus concentration for
iridium nanocrystal dispersions in toluene, measured at 500 nm…....73
Figure 3.3 Schematic of apparatus for determining the dispersibility of the
iridium nanocrystals in CO2-toluene solvent………………………...75
Figure 3.4 GC-MS Calibration curve of a mixture of 1-decene and 1-decane….78
Figure 3.5 (a) TEM micrograph of Iridium nanocrystals stabilized by TOAB
and (b) corresponding histogram for Iridium nanocrystals………….79
Figure 3.6 (a) N2 adsorption-desorption isotherms for mesoporous silica MPS
and (b) corresponding pore size distributions for MPS……………...81
Figure 3.7 Dispersibility of Ir-TOAB nanocrystals in CO2-toluene mixed
solvent measured by UV-Vis spectroscopy………………………….83
Figure 3.8 (a) Z contrast STEM (Planar view) of MPS infused with Ir-TOAB;
(b) Out of focus view of (a); (c) top down view of (a); and (d) TEM
Micrograph of MPS infused with Ir-TOAB nanocrystals without any
pre-treatment…………………………………………………………87
Figure 3.9 Comparison of catalyst activities for decene hydrogenation with
Ir-MPS (without any pretreatment) and an industrial catalyst,
Pd-Al2O3……………………………………………………………..89
Figure 3.10 Calibration curve of UV- Vis absorbance versus concentration for
Iridium nanocrystal dispersions in 1-decene, measured at 500 nm….96
Figure 3.11 TEM micrographs of Ir-MPS (without any pretreatment) after 1-
decene hydrogenation………………………………………………..97
Figure 3.12 Arrhenius plot of decene hydrogenation over untreated Ir-MPS…….98
xv
Figure 4.1 (a) HRTEM images of pre-synthesized FePt nanocrystals prior to
infusion. (b) Histogram of nanocrystal size………………………...113
Figure 4.2 Loading of FePt nanocrystals within mesoporous silica in 10 minutes
as a function of initial concentration of nanocrystals………………114
Figure 4.3 TEM images of 7.3 wt % FePt nanocrystals after infusion into
mesoporous silica. (a) HAADF STEM planar view; (b) HRTEM;
(c) Histogram of nanocrystal size using (a)………………………...116
Figure 4.4 7.3 wt % FePt nanocrystals in mesoporous silica calcined at
700°C for 4 h: (a) HAADF- STEM; (b) Histogram of nanocrystals
after calcination determined from (a). (c) Filtered image of random
portion of (a) to eliminate shot noise; (d) Intensity profile of FePt
marked (A) in image (c); (e) Intensity profile of FePt marked
(B) in image (c)…………………………………………………….119
Figure 4.5 (a) XRD diffraction of FePt nanocrystals in silica before and after
annealing at 700 °C. (b) Convergent beam electron diffraction
of FePt nanocrystals in silica……………………………………….121
Figure 4.6 Turn over frequency for 1-decene hydrogenation by 5 wt % FePt
in mesoporous silica………………………………………………..124
Figure 5.1 (a) UV-Vis spectra of Pt nanoparticles dispersed in toluene before
and after infusion in AMC and (b) calibration curve for Pt
nanoparticles using known concentrations and measuring
absorbance at 800 nm………………………………………………148
Figure 5.2 SAXS of mesoporous carbons……………………………………...151
Figure 5.3 Nitrogen porosimetry of the mesoporous carbons and their pore size
distributions in the inset for GMC (A) and AMC (B)……………...152
Figure 5.4 XRD of the mesoporous carbon samples without added Pt………..155
Figure 5.5 TEM micrographs of 20% Pt on (A) GMC uncalcined and (B)
GMC calcined at 300 °C, (C) AMC calcined at 500 °C. (D), (E),
xvi
(F) Histograms for (A), (B) and (C), respectively. More than 50
nanoparticles were counted in each case…………………………...158
Figure 5.6 (A) TGA for the pure (blank) carbons and 20% Pt infused
mesoporous carbons AMC and GMC calcined at 500 and 300 °C
in N2/H2 atmosphere and (B) Differential thermal gravimetry (DTG)
for the Pt on AMC and GMC. Inset shows the magnified section of
(B) where onset of oxidation occurs for AMC…………………….159
Figure 5.7 XRD spectra of 20% Pt on (a) GMC and (b) AMC, before and
after ADT (for conditions in Table 5.4). Calcinations were done in
7% H2-93%N2 environment………………………………………...161
Figure 5.8 XPS spectra before and after ADT for 20% Pt supported on AMC
and GMC. (A),(B): C1s region for AMC and GMC, respectively.
(C),(D): Pt4f region for AMC and GMC, respectively…………….165
Figure 5.9 Electrochemical studies for 20% Pt on mesoporous carbon before
and after ADT cycling for conditions in Table 5.4 (A)CV of GMC
at 20 mV/s (B) CV of AMC at 20 mV/s (C) Linear Scan Voltammetry
(LSV) of GMC at 5 mV/s rotated at 1600 rpm (D) LSV for AMC
under identical conditions………………………………………….166
Figure 5.10 Tafel Slopes for Pt-GMC calcined at 300°C (A) Mass activity (B)
Specific activity and Pt-AMC calcined at 500°C (C) Mass Activity
and (D) Specific Activity…………………………………………...168
Figure 6.1 XRD spectra of the disordered mesoporous carbons annealed at
different temperatures………………………………………………193
Figure 6.2 Raman spectra of the disordered mesoporous carbons annealed at
different temperatures………………………………………………194
Figure 6.3 Cyclic voltammetry of Pt20Cu80 nanoparticles supported on
DMC-2000 during electrochemical dealloying at 20 mV/s in argon
atmosphere………………………………………………………….198
xvii
Figure 6.4 XRD spectra of the Pt20Cu80 nanocrystals supported on DMC-2000
before and after electrochemical dealloying of copper (a); peak
analysis of the XRD spectra of the as-synthesized Pt20Cu80
nanoparticles (b); and electrochemically dealloyed nanoparticles
with a final composition of Pt85Cu15………………………………..199
Figure 6.5 TEM micrographs and histograms of Pt20Cu80 on DMC-2000
(A,C) as-synthesized and calcined at 350 °C (B,C) Electrochemically
dealloyed……………………………………………………………203
Figure 6.6 Linear scan voltammetry of the dealloyed Pt-Cu nanoparticles
supported on DMC at 5 mV/s with rotation rate of 1600 rpm……...204
Figure 6.7 Potential-Activity plots of dealloyed Pt20Cu80 nanoparticles
supported on DMC-2000 and DMC-850 compared to commercial
Pt (ETEK) catalysts (a) Pt mass-based activities and (b) Pt specific
surface area activities……………………………………………….207
Figure A.1 SAXS on mesoporous silica of type SBA-15. Center-to-center
spacing between cylindrical pores is 8.5 nm………………………..231
Figure A.2 (a) Adsorption isotherms for SBA-15 measured by nitrogen
porosimetry. (b) Pore size distribution for SBA-15 measured by
nitrogen porosimetry………………………………………………..232
Figure A.3 HAADF-STEM micrographs of 7.3 wt % FePt nanocrystals in
mesoporous silica calcined at 700 oC in 7%H2/93%N2…………….233
Figure A.4 Total interaction energy as a function of separation distance and
metal nanocrystal size. Inset shows the van der Waals energy as a
function of separation distance for different nanocrystal sizes……..236
Figure A.5 Low resolution TEM of 7.3 wt % FePt in mesoporous silica………238
Figure A.6 Calibration of UV-Vis absorbance of FePt…………………………239
Figure B.1 Nitrogen adsorption-desorption curves (a) and corresponding pore
xviii
size distributions (b) for the disordered mesoporous carbons…….....245
Figure B.2 TEM micrographs (A) Low resolution (B) High-resolution and (C)
Histogram, of as-synthesized Pt20Cu80 nanoparticles. The average
nanoparticle size is 2.85±0.38 nm…………………………………..246
Figure B.3 (a) Effect of infusion of the Pt20Cu80 nanoparticles in DMC-850 on
absorbance as determined by UV-Visible spectrophotometry (b)
Calibration curve of the UV-Visual absorbance vs. concentration of
Pt20Cu80 nanoparticles in toluene, measured at 800 nm…………….247
Figure B.4 First 20 cyclic voltammetry profiles of Pt20Cu80 nanoparticles
supported on DMC-2000 during electrochemical dealloying at 20
mV/s in argon atmosphere. The black arrows indicate the direction in
which the peak shifts as cycling progresses………………………...249
Figure B.5 Cyclic voltammogram of 20% monometallic Pt on DMC-2000
scanned at 100 mV/s in argon atmosphere…………………………250
Figure B.6 XPS of 20% Pt supported on graphitized mesoporous carbon……..251
Figure B.7 TGA of Pt nanoparticles supported on disordered mesoporous
carbons……………………………………………………………...252
Figure B.8 Cycling voltammograms for Pt20Cu80 on DMC 2000 before and after
ADT potential cycling……………………………………………...253
Figure B.9 Peak analysis of the XRD spectra of the as-synthesized Pt75Cu25
nanoparticles………………………………………………………..254
1
Chapter 1
Introduction
A major goal in catalysis is the rational design of active catalyst sites to
produce high turnover numbers, selectivities and stabilities. 1-4 The arrangement of
metal clusters on supports which governs metal-support interactions has a profound
effect on catalytic performance. Metal nanocrystals with controlled morphology offer
high fractions of (1) coordinatively unsaturated surface atoms and (2) highly active
bifunctional perimeter sites at the metal-support interface.5,3 It is thus desirable to
design heterogeneous catalysts with control over the morphology of both the
nanocrystals and the support. Traditionally, heterogeneous catalysts are synthesized
by precursor reduction1,6,7, ion exchange8,9 and chemical vapor/fluid deposition.10-12
In these techniques, it is challenging to control the metal cluster size, alloy
composition, crystal structure and morphology. Metal precursors upon reduction may
sinter to form large nanocrystals with low active surface area and consequently show
poor activities (Figure 1.1). Synthesizing multimetallic catalysts by in-situ reduction
of multiple precursors on the supports may result in high compositionally
polydispersity, which is undesirable.13-15 By-products may be formed during the
reduction reaction and may be detrimental for the activity or selectivity of catalysts.
Another approach for synthesizing supported catalysts is to synthesize the
support structure in the presence of a dispersion of pre-synthesized nanocrystals.16,17
In this technique it can be difficult to achieve high loadings of dispersed nanocrystals
within the pores. In addition, the presence of the particles can interfere with the
synthesis of the desired pore structure and conversely, the synthesis of the
mesoporous material can change the morphology of the nanocrystals. The
nanocrystals may be completely trapped in the support matrix and may become
inaccessible to the reactants during the reaction. The ligands stabilizing the
2
Figure 1.1
Different techniques for the synthesis of heterogeneous catalysts
+ Metal precursor
Reduction
Metal oxide precursor (like TEOS)
Precursor reduction
Support synthesis in nanoparticle solution
+
3
nanocrystals may desorb and adsorb onto the substrate during the reaction, leading to
destabilization of the nanocrystals.
Yet another novel technique for catalyst design is to infusion of the
presynthesized nanocrystals in the catalyst supports (Figure 1.1).3,18,19 The
decoupling of the nanocrystal synthesis step and the infusion step, leads to exquisite
control of the nanocrystal size, morphology and dispersibility within the pores. An
advantage of this approach is that a large number of well-known synthetic methods
may be utilized for the design of the nanocrystals20 as well as that of support
materials.21 Prior to infusion, the particles are easily isolated from unwanted
byproducts, cleaned, separated by size, and the ligands may be exchanged if needed.
1.1 Nanocrystal Synthesis
The ability to control the uniformity of the size, shape, composition, and
surface properties of metal nanocrystals is desirable for optimizing their performance
in the catalysis. Recently, colloidal nanocrystals with controlled structure and
morphology have been synthesized by arrested precipitation technique using
stabilizing ligands (Figure 1.2).18,22-29 A typical synthesis system for colloidal
nanocrystals consists of three components: metal precursors, surfactants (ligands) and
solvents.20 The formation of the nanocrystals involves two steps: nucleation and
growth. In the nucleation step, precursors decompose or react, typically at a relatively
high temperature, to form a supersaturation of monomers followed by a burst of
nucleation of nanocrystals. The adhesion energy of ligands to the surface needs to be
such regions of the nanocrystal surface are transiently accessible for growth, yet
entire crystals are, on average, monolayer- protected to block aggregation.
In arrested precipitation technique size, morphology and composition can be
controlled by adjusting the reaction conditions such as time, temperature, precursor
concentration, surfactant type, and surfactant concentration.30-33 When the
nanocrystals are pre-synthesized, heterogeneous sites on a support are not present,
4
Figure 1.2
Arrested precipitation technique of nanocrystal synthesis.
Metal precursor + Ligand + Solvent
Reduction
Arrested Growth Metal nanoparticle dispersion
3.6 nm FePt nanoparticles
5
which could otherwise lead to uncontrolled composition, resulting from competition
between nucleation in solution and on the surface (Figure 1.3). When the rates of
precursor decomposition and equilibrium solubilities of the two resulting metals are
similar, formation of binary metal nuclei is favored.31 Even when these properties are
somewhat dissimilar, the nuclei formed from the first metal may capture newly
formed atoms of the second metal before it nucleates, which would otherwise lead to
two separate populations of growing nanocrystals.34 As metal from the more rapidly
decaying precursor is incorporated into the nanoparticle, it will be depleted in the
solution relative to the second metal. Thus, the probability of adding the second
component will increase. In essence, this may be considered to be composition
focusing by analogy to size focusing. The temperature is often a key variable to
produce rapid decomposition of both precursors to achieve focusing of both size and
composition, along with sufficient dynamic solvation by the ligand to arrest growth.
Once the particles are pre-synthesized, they may be loaded physically into the pores
of the support at completely different concentrations than those used in the synthesis
of the particles. Thus, both the pre-synthesis and infusion steps may be optimized in
ways not available when precursors are reduced within pores of the support.
1.2 Mesoporous Supports
An emerging frontier in catalysis is the design of extremely active and stable
catalysts by depositing pre-synthesized nanocrystals with desired morphologies onto
high surface area supports with ordered mesopores.34-38 Mesoporous materials of
silica such as MCM-41 and SBA-15 and various metal oxides39 have been
synthesized with surfactant templates to produce periodic ordered cylindrical pores
with diameters in the range of 2 to 20 nm with internal surface areas up to 1000
m2/g.40-42 The pore diameter is ideal for dispersing metal nanocrystals with diameters
in the same size range and for the transport of reactants and products. The high
dispersion of the nanocrystals within the extremely high surface area pores will
6
Figure 1.3
Synthesis of bimetallic catalysts via wetness-impregnation and infusion of
presynthesized nanocrystals into porous supports.
Metal-1 precursor
Metal-2 precursor
Reduction over Supports
Heterogeneous Nucleation
Support
Growth
Support
Support
Homogeneous Nucleation
Growth
Metal-1 precursor
Metal-2 precursor
Reduction in solvent in presence of ligands
Support
Infusion
7
minimize contact between the nanocrystals to prevent sintering. Particles in adjacent
pores are completely isolated. Particles in a single cylindrical pore are separated over
a long pore length. The convex surfaces on the nanocrystals favored contact with the
cylindrical mesopores with concave curvature, enhancing the geometric interfacial
area which favors strong metal-support interactions, thereby facilitating high stability
of nanocrystals towards desorption or sintering.43 These strong interactions can be
used to manipulate the crystal structure of nanoparticles, while controlling the
nanoparticle size. For example, ordered intermetallic FePt alloy with <4 nm size is
achieved by transformation of the disordered face-centered-cubic FePt nanocrystals
supported in mesoporous silica by annealing at 700 oC. However, the surfaces of both
the metal nanocrystal and support are often both convex for traditional supports
leading to a smaller degree of interfacial contact and poor metal-support interactions.
For traditional supports, the nanocrystals are supported on the external surface of
support particles and have a greater tendency to contact other nanocrystals and grow
by sintering.
Typically, mesoporous materials are synthesized by a soft surfactant
templating approach, where long-chained surfactants are used as a template.21 These
surfactants tend to form complex micelles in solution whose morphology can be
controlled by altering surfactant concentration. The morphology of mesoporous
materials can in turn be tuned by controlling the morphology of the surfactant
micelles. When silica/metal-oxide precursors are introduced in the micelles,
mesostructures are formed by a cooperative self-assembly of surfactant/precursor ion
pairs that arrange themselves in various ordered arrays (Figure 1.4). Upon
condensation of the metal-oxide oligomers and subsequent surfactant removal by
calcination or ion exchange, the mesoporous materials with surface areas up to 1000
m2/g and pore diameter up to 20 nm are formed. For example, cylindrical
mesoporous silica of type SBA-15 with pore size ~ 10 nm is synthesized using
amphiphilic surfactant pluronic P123 (EO20PO70EO20) as a template and
tetraethoxysilane as a silica precursor (Figure 1.4).42
8
(C) (D)
Figure 1.4
(A, B) Synthesis approaches for mesoporous materials. The mesostructure can be
formed previously (A) or by a cooperative process between surfactants and metal-
oxide precursors. Figure taken from Soler-Illia et al.21 TEM images of mesoporous
silica of type SBA-15 with 6 nm pores (top view) (C) and planar view (D). Images
taken from Zhao et al.42
9
Another approach for synthesizing mesoporous materials is to cross-link the
metal-oxide precursors within the preformed mesoporous materials like silica which
are used as hard templates, followed by etching away the template.44,45 In this way the
morphology of the mesoporous template is transferred to the new mesoporous metal-
oxide materials. The first self-supported highly ordered mesoporous carbon of type
CMK-1 with 3-d cubic morphology was synthesized by using 3-D highly ordered
mesoporous silicate material, MCM-48, as a hard template.46 Since then several
different morphologies of mesoporous carbons have been synthesized by hard
templating method.47-49
1.3 Infusion of Presynthesized Nanocrystals into Mesoporous Supports
Relatively few studies have examined infusion of pre-synthesized
nanocrystals into porous materials with carrier solvents.18,19,50 There are two primary
challenges in the infusion of pre-synthesized nanocrystals into a nanoporous matrix.
First, the nanocrystal dispersion must be transported through the pores by favorable
capillary wetting, and the nanocrystals must not block small pore entrances. Once the
nanocrystals penetrate the pores, the second challenge is to achieve sufficient
adsorption of the nanocrystals from the solution phase to the substrate surface. As the
surface of the metal is covered with the stabilizing ligands which may bind very
strongly to the metal, like thiol to gold surface, short-ranged metal-support
interactions which facilitate adsorption may be highly screened. To achieve the metal
loadings >0.5 wt %, the long-ranged van der Waals (vdW) interactions be used;
However, the solvent must allow for enough attraction between the nanocrystals and
substrate to achieve sufficient adsorption, while simultaneously preventing
flocculation of nanocrystals in the solvent phase. The nanocrystals must be dispersed
in “good” solvent for the stabilizing ligands to prevent flocculation. However, good
solvents may be expected to cause the interactions between the nanocrystals and the
surface of the nanoporous material to be repulsive or weakly attractive.30 The
10
resulting weak adsorption of the particles inside the pores may produce low loadings.
Thus, the solvent must be tuned to provide a judicious balance between nanocrystal-
nanocrystal versus nanocrystal-substrate interactions.
The use of supercritical CO2 as an antisolvent helps overcome both transport
and thermodynamic challenges in nanoparticle infusion into porous materials (Figure
1.5).18 Supercritical CO2 is added to an organic solvent like toluene to achieve high
levels of infusion of pre-synthesized nanocrystals into mesoporous silica by
enhancing long-ranged interactions. CO2 has an unusually low polarizability/density
(very weak VdW forces) and upon addition to toluene forms a mixture of low
Hamaker constant. This consequently weakens the vdW forces for the intervening
solvent and results in a stronger interaction between the metal and support thus
facilitating nanocrystal adsorption on the porous surface and enhancing the loadings
upto 5 wt %.18,50 Furthermore, the loading may be tuned over a wide range by
controlling the van der Waals interactions by varying the CO2 composition.
Yet another technique to further enhance the loadings over 5 wt % not
typically achieved by tuning the long-ranged van der Waals interactions using CO2, is
to strengthen short-ranged interactions between the metal and support.19 This is done
by strategic selection of weakly binding ligands for nanocrystal stabilization like
amines and fatty acids that known to bind to metals like Pt or Au weakly relative to
ligands like thiols.51 For example, surface coverage of dodecylamine on Pt is 46%52
compared to 75% for dodecanethiol on Ag26. The lower surface coverage exposes
barer metal surface for the formation of specific short-ranged interactions with the
silica. These short-ranged interactions could facilitate the loadings of ~20 wt %,
relative to the much lower loadings for gold capped with dodecanethiol,18 where the
surface coverage is higher. For example, Figure 1.6 shows the FePt nanoparticles
with 8 wt % loadings supported on mesoporous silica. The convex surfaces on the
nanocrystals favored contact with the cylindrical mesopores with concave curvature,
further enhancing metal-support interactions.43
11
Figure 1.5
Influence of CO2 concentration on nanoparticle loading. (A) Optimized addition of
the CO2 in toluene promotes nanocrystal adsorption on the supports while
simultaneously keeping nanocrystals dispersed in the solvent. (B) Too much addition
of CO2 in toluene leads to nanocrystal flocculation and poor loadings.
10 μm
~5 nm
Nanocrystals dispersed in CO2-toluene
10 μm
~5 nm
(A)
(B)
12
(a) (b)
Figure 1.6
TEM images of FePt nanocrystals (3.6 nm) after infusion into mesoporous silica (7
nm). (a) HAADF STEM planar view; (b) HRTEM; Figures taken from Chapter 4.
13
1.4 Electrocatalysts for Oxygen Reduction Reaction: Activity and
Stability
The high dispersion of catalysts on mesoporous supports is beneficial for
electrocatalysis, for example for the oxygen reduction reaction for Pt on C. Polymer
electrolyte membrane fuel cells (PEMFC) are considered a potential alternative
source of energy due to high efficiencies and clean operation.53-55 Commercialization
of PEMFCs is still a challenge due to the high cost of platinum based catalysts as well
as poor catalyst stability.56-58 The performance of the platinum catalysts needs to be
increased to 4-5 times that of current commercial catalysts in order to meet the
performance targets for fuel cell cathodes.53 To achieve activity enhancement,
platinum based bimetallic alloys, Pt-M (M=Co, Ni, Fe, V, Ti), have been extensively
studied and activity enhancements up to 2.5 times that of pure Pt have been
achieved.59-65 Recently, Strasser and co-workers synthesized novel Pt-Cu catalysts on
carbon with a pronounced activity enhancement of >4 time that of pure Pt
catalysts.66,67 The copper was electrochemically etched from the nanoparticle surface
to form a strained Pt shell. However, the catalysts were synthesized by sequential
wetness impregnation and led to catalysts with multiple alloys phases. Another
significantly bigger challenge hindering the fuel cell technology is the poor catalyst
stability.53,56,68-70 For automotive applications, catalyst support stability at 1.2V is
required for an accumulated life time of a vehicle, with a maximum allowable
performance loss of <30 mV at 1.5A/cm2 (< 1 µV/cycle target).56 Traditionally
amorphous carbons, such as Vulcan XC-72 carbon, easily oxidize at potentials of
1.2V, leading to significant activity losses due carbon oxidation, metal nanoparticle
coalescence and metal dissolution.56
The challenges in stability and activity can be ameliorated by the infusion of
presynthesized nanocrystals with controlled size, composition, crystal phases and
morphology in the graphitic mesoporous carbon supports. Graphitized carbons are
more resistant to oxidation than amorphous carbons due to significantly reduced
14
number of oxygen sites on the surface.69,71-73 In addition, the abundant π electron sites
on the surface interact strongly with platinum, thus strengthening the metal-support
interactions, as demonstrated by XPS and electron-spin resonance (ESR).69,71,73
Strong metal-support interactions are expected to mitigate coalescence , dissolution
and subsequent Ostwald ripening of the Pt. Thus, oxidation resistant graphitic
supports with strong Pt-support interactions are expected to enhance the stability of
ORR catalysts.
1.5 Objectives
The objective of this thesis is to synthesize supported catalysts with control
over the morphology of the nanoparticles as well as of the supports for achieving high
catalyst activities and stabilities. To accomplish this, nanocrystals with controlled
size, composition and alloy structure have been presynthesized using stabilizing
ligands and then infused within high surface area mesoporous supports. One of the
key objectives of this thesis is to synthesize the catalysts with high metal loadings.
Thus, the effect of the ligand-metal binding strength on metal loadings was studied.
Nanocrystals stabilized with strongly binding ligands preclude short-ranged metal-
support interactions and lead to poor loadings. For achieving significant metal
loadings of up to 5 wt % for strongly bound nanocrystals, the use of supercritical CO2
as an antisolvent in toluene to enhance the long-ranged metal-support interactions is
explored. Subsequently, to achieve the nanocrystal loadings ≥5 wt %, weakly binding
hydrocarbon ligands that allow short-ranged metal-support interactions are used for
nanocrystal synthesis. These weakly binding hydrocarbon ligands are demonstrated to
be removed from the metal surface by mild thermal treatment without perturbing the
morphology of the metal or the support. The easy ligand removal frees metal sites and
facilitates high catalyst activities.
Another key focus of the thesis is enhancement of the catalyst stability.
Immobilization of the nearly spherical nanocrystals in concave mesopores with
15
similar radii enhances the interfacial area and favors strong metal-support
interactions, relative to other support geometries. The high stabilities resulting from
these strong interactions have been shown to be beneficial in the transformation of the
bimetallic nanoparticle structure from a disordered to an ordered intermetallic alloy
upon annealing at 700 oC without perturbing the size. The fundamental understanding
of the metal-support interactions is then utilized in the design of highly stable Pt
electrocatalysts with controlled size and morphology for oxygen reduction using
graphitized mesoporous carbons as supports. The high surface area and resistance to
oxidation of the graphitic carbon69 in addition to the strong binding to Pt69 has been
demonstrated to mitigate nanoparticle isolation, sintering and dissolution, and hence
enhance catalyst stability. Finally, highly stable Pt-Cu catalysts on mesoporous
graphitic carbons for the oxygen reduction have been synthesized to achieve activities
4-fold that of pure Pt, by selective electrochemical dealloying of copper. The
relationship between the catalyst activity and strain in the Pt shell produced by copper
dealloying has been investigated by controlling the composition of the single-phased
Pt-Cu nanoparticles.
1.6 Dissertation Outline
Chapter 2 represents the first example of infusion of presynthesized gold
nanocrystals into mesoporous silica using sc-CO2 as an antisolvent in toluene. The
van der Waals interactions between gold nanocrystals, stabilized with dodecanethiol,
and silica are strengthened upon addition of CO2 to toluene. The enhanced long
ranged interactions facilitate adsorption of nanocrystals on the silica thus leading to
metal loadings of ~ 2 wt %. Concentration of CO2 in toluene needs to be tuned to
promote nanocrystal adsorption without flocculating them in the solvent. A simple
van der Waals model for calculating the overall Hamaker constant between the metal
nanocrystal and silica wall, with an intervening solvent, is developed and
16
qualitatively predicts the linear increase in loadings with increase of CO2
concentration in toluene, as seen experimentally.
While Chapter 2 examined just the synthesis of the gold catalysts with
significant metal loadings, in Chapter 3 iridium catalysts are synthesized using
supercritical CO2 and investigated for hydrogenation reactions. The
tetraoctylammonium bromide ligands for iridium nanoparticle stabilization are
strategically chosen such that they bind sufficiently weakly to the nanoparticle
surface, so that it becomes unnecessary to pre-treat the catalyst at high temperature to
remove the ligand, for activation of the catalysts. The iridium catalysts are used for 1-
decene hydrogenation and show good catalyst activities which are higher than
homogeneous iridium catalysts and twice better than commercial Pd-Al2O3 catalysts.
The catalyst does not show any enhancement of the activity when pre-treated under
high temperatures or when annealed in supercritical CO2 indicating migration of
weakly binding ligands on to high surface area supports after nanoparticle adsorption.
Whereas CO2 was used to facilitate nanocrystal adsorption on the support in
the above two chapters, it was not needed in Chapters 4 to 6, since the interactions of
the nanocrystals with the support were stronger as a consequence of uncovered metal
on the particle surface. Chapter 4 demonstrates for the first time, a novel concept to
form supported catalysts by infusion of pre-synthesized bimetallic nanocrystals into
ordered mesoporous supports, facilitated by short-ranged metal-support interactions.
For pre-synthesized FePt nanocrystals (<4 nm), coated with oleic acid and oleylamine
ligands in toluene, high loadings above 10 wt % are achieved in order of minutes. The
strong metal-support interactions are favored by the low coverage of the weakly
bound ligands and enhanced interfacial area between spherical nanocrystals and
concave mesopores. These strong interactions facilitate phase transformation of FePt
from a disordered phase (FCC) to ordered intermetallic phase (FCT) by annealing at
temperatures of 700 °C while preventing any significant sintering. The calcined FePt
catalysts are used for 1-decene hydrogenation and exhibit 6-fold higher catalyst
17
activity (TOF =30 s-1) than that of a commercial Pd-alumina catalyst with stability for
multiple reactions.
Whereas Chapters 3-4 examined hydrogenation reactions, Chapters 5-6
examine electrocatalysis with similar types of pre-synthesized nanocrystal catalysts,
but with mesoporous carbon supports. In Chapter 5, highly stable Pt electrocatalysts
have been prepared by the infusion of pre-synthesized nanocrystals with controlled
size (<4 nm) and morphology onto the ordered graphitic mesoporous carbon. Weakly
binding ligands (dodecylamine) stabilized the nanocrystals, while exposing sufficient
amount of bare metal surface to facilitate strong Pt-support interactions, resulting in
nanocrystal loadings of 20 wt % with high metal dispersion. The mass activity for the
oxygen reduction reaction (ORR) at 0.9V reaches 0.14 A/mgPt for Pt on the graphitic
mesoporous carbon, relative to ~0.11 A/mgPt for Pt on commercial amorphous
Vulcan XC-72. The loss in ORR catalyst activity for Pt supported on graphitic
mesoporous carbons under an accelerated durability test of potential cycling between
0.5 to 1.2 V for 1000 cycles, is extremely small, <2%, compared to a 30% loss for Pt
on amorphous mesoporous carbons and 70% typically for commercial Pt on
amorphous carbon catalysts. The highly oxidation resistant graphitized carbon
coupled with the strong Pt interactions with the π sites of graphite, prevents metal
sintering, Pt dissolution, and Ostwald ripening, resulting in high catalyst stability.
Chapter 6 is a follow up study of Chapter 5 where, highly stable and active
electrocatalysts have been prepared by the infusion of pre-synthesized bimetallic Pt-
Cu nanocrystals with controlled size, composition and alloy structure onto graphitic
mesoporous carbon. Electrochemically dealloying via leaching of Cu from the Pt-Cu
nanoparticle surface generates strained Pt rich shell raising the catalyst activity for
ORR 4-folds over commercial Pt catalysts. The low polydispersity in the size and
composition during pre-synthesis is maintained through the dealloying stage,
resulting in a Pt shell with a well controlled level of strain that favors high ORR
activity. The catalysts supported on graphitic mesoporous carbons are stable towards
potential cycling between 0.5- 1.2V for 1000 cycles and show a loss of <2 % activity,
18
compared to a 40% loss for identical nanoparticles on amorphous mesoporous
carbons. The high stability is due to oxidation resistant carbon and strong Pt-
π interactions.
The dissertation conclusion and recommendations for future work are given in
Chapter 7. Supporting information for Chapter 4 can be found in Appendix A, while
Appendix B contains supplementary information for Chapter 7.
1.7 References
(1) Wakayama, H.; Setoyama, N.; Fukushima, Y. Size-Controlled
Synthesis and Catalytic Performance of Pt Nanoparticles in Micro- and Mesoporous
Silica Prepared Using Supercritical Solvents. Adv. Mater. 2003, 15, 742-745.
(2) Rolison, D. R. Catalytic nanoarchitectures - The importance of nothing
and the unimportance of periodicity. Science (Washington, DC, United States) 2003,
299, 1698-1702.
(3) Rioux, R. M.; Song, H.; Hoefelmeyer, J. D.; Yang, P.; Somorjai, G. A.
High-Surface-Area Catalyst Design: Synthesis, Characterization, and Reaction
Studies of Platinum Nanoparticles in Mesoporous SBA-15 Silica. Journal of Physical
Chemistry B 2005, 109, 2192-2202.
(4) Liang, C.; Dai, S. Synthesis of Mesoporous Carbon Materials via
Enhanced Hydrogen-Bonding Interaction. Journal of the American Chemical Society
2006, 128, 5316-5317.
(5) Cho, A. Catalysis news: Connecting the dots to custom catalysts.
Science (Washington, DC, United States) 2003, 299, 1684-1685.
(6) Lensveld, D. J.; Mesu, J. G.; van Dillen, A. J.; de Jong, K. P. The
application of well-dispersed nickel nanoparticles inside the mesopores of MCM-41
by use of a nickel citrate chelate as precursor. Studies in Surface Science and
Catalysis 2002, 143, 647-657.
19
(7) Zong, Y.; Watkins, J. J. Deposition of copper by the H2-assisted
reduction of Cu(tmod)2 in supercritical carbon dioxide: Kinetics and reaction
mechanism. Chemistry of Materials 2005, 17, 560-565.
(8) Singh, U. K.; Albert Vannice, M. Liquid-Phase Hydrogenation of
Citral over Pt/SiO2 Catalysts. Journal of Catalysis 2000, 191, 165-180.
(9) Goodenough, J. B.; Hamnett, A.; Kennedy, B. J.; Manoharan, R.;
Weeks, S. A. Methanol oxidation on unsupported and carbon supported platinum +
ruthenium anodes. Journal of Electroanalytical Chemistry and Interfacial
Electrochemistry 1988, 240, 133-145.
(10) Hayek, K.; Goller, H.; Penner, S.; Rupprechter, G.; Zimmermann, C.
Regular Alumina-Supported Nanoparticles of Iridium, Rhodium and Platinum Under
Hydrogen Reduction: Structure, Morphology and Activity in the Neopentane
Conversion. Catalysis Letters 2004, 92, 1-9.
(11) Serp, P.; Hierso, J. C.; Feurer, R.; Kihn, Y.; Kalck, P.; Faria, J. L.;
Aksoylu, A. E.; Pacheco, A. M. T.; Figueiredo, J. L. Single-step preparation of
activated carbon supported platinum catalysts by fluidized bed organometallic
chemical vapor deposition. Carbon 1999, 37, 527-530.
(12) Kim, K. T.; Kim, Y. G.; Chung, J. S. Effect of surface roughening on
the catalytic activity of Pt-Cr electrocatalysts for the oxygen reduction in phosphoric
acid fuel cell. Journal of the Electrochemical Society 1995, 142, 1531-1538.
(13) Raghuveer, V.; Manthiram, A.; Bard, A. J. Pd-Co-Mo Electrocatalyst
for the Oxygen Reduction Reaction in Proton Exchange Membrane Fuel Cells.
Journal of Physical Chemistry B 2005, 109, 22909-22912.
(14) Antolini, E. Formation of carbon-supported PtM alloys for low
temperature fuel cells: a review. Materials Chemistry and Physics 2003, 78, 563-573.
(15) Koh, S.; Leisch, J.; Toney, M. F.; Strasser, P. Structure-Activity-
Stability Relationships of Pt-Co Alloy Electrocatalysts in Gas-Diffusion Electrode
Layers. Journal of Physical Chemistry C 2007, 111, 3744-3752.
20
(16) Konya, Z.; Puntes, V. F.; Kiricsi, I.; Zhu, J.; Alivisatos, P.; Somorjai,
G. A. Novel Two-Step Synthesis of Controlled Size and Shape Platinum
Nanoparticles Encapsulated in Mesoporous Silica. Catalysis Letters 2002, 81, 137-
140.
(17) Konya, Z.; Puntes, V. F.; Kiricsi, I.; Zhu, J.; Ager, J. W., III; Ko, M.
K.; Frei, H.; Alivisatos, P.; Somorjai, G. A. Synthetic Insertion of Gold Nanoparticles
into Mesoporous Silica. Chemistry of Materials 2003, 15, 1242-1248.
(18) Gupta, G.; Shah, P. S.; Zhang, X.; Saunders, A. E.; Korgel, B. A.;
Johnston, K. P. Enhanced Infusion of Gold Nanocrystals into Mesoporous Silica with
Supercritical Carbon Dioxide. Chemistry of Materials 2005, 17, 6728-6738.
(19) Gupta, G.; Patel, M. N.; Ferrer, D.; Heitsch, A. T.; Korgel, B. A.;
Johnston, K. P. Stable Ordered FePt Mesoporous Silica Catalysts with High Loadings
Chemistry of Materials 2008.
(20) Yin, Y.; Alivisatos, A. P. Colloidal nanocrystal synthesis and the
organic-inorganic interface. Nature (London, United Kingdom) 2005, 437, 664-670.
(21) de Soler-Illia, G. J.; Sanchez, C.; Lebeau, B.; Patarin, J. Chemical
Strategies To Design Textured Materials: from Microporous and Mesoporous Oxides
to Nanonetworks and Hierarchical Structures. Chemical Reviews (Washington, DC,
United States) 2002, 102, 4093-4138.
(22) Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R.
Synthesis of Thiol-Derivatised Gold Nanoparticles in a Two-Phase Liquid-Liquid
System. J. Chem. Soc., Chem. Commun. 1994, 7, 801-802.
(23) Sun, S. Recent advances in chemical synthesis, self-assembly, and
applications of FePt nanoparticles. Advanced Materials (Weinheim, Germany) 2006,
18, 393-403.
(24) Sun, S.; Fullerton, E. E.; Weller, D.; Murray, C. B. Compositionally
controlled FePt nanoparticle materials. IEEE Transactions on Magnetics 2001, 37,
1239-1243.
21
(25) Korgel, B. A.; Fitzmaurice, D. Small-angle x-ray-scattering study of
silver-nanocrystal disorder-order phase transitions. Physical Review B: Condensed
Matter and Materials Physics 1999, 59, 14191-14201.
(26) Korgel, B. A.; Fullam, S.; Connolly, S.; Fitzmaurice, D. Assembly and
Self-Organization of Silver Nanocrystal Superlattices: Ordered \"Soft Spheres\".
Journal of Physical Chemistry B 1998, 102, 8379-8388.
(27) Knecht, M. R.; Garcia-Martinez, J. C.; Crooks, R. M. Hydrophobic
Dendrimers as Templates for Au Nanoparticles. Langmuir 2005, 21, 11981-11986.
(28) Kim, Y.-G.; Garcia-Martinez, J. C.; Crooks, R. M. Electrochemical
Properties of Monolayer-Protected Au and Pd Nanoparticles Extracted from within
Dendrimer Templates. Langmuir 2005, 21, 5485-5491.
(29) Ferrer, D.; Blom, D. A.; Allard, L. F.; Mejia, S.; Perej-Tijerina, E.;
Yacaman, M. J. Atomic structure of three-layer Au/Pd nanoparticles revealed by
aberration-corrected scanning transmission electron microscopy. Journal of Materials
Chemistry 2008, 18, 2442.
(30) Israelachvili, J. Intermolecular & Surface Forces; 2nd ed.; Academic
Press, Inc.: San Diego, 1992.
(31) Sun, X.; Kang, S.; Harrell, J. W.; Nikles, D. E.; Dai, Z. R.; Li, J.;
Wang, Z. L. Synthesis, chemical ordering, and magnetic properties of FePtCu
nanoparticle films. Journal of Applied Physics 2003, 93, 7337-7339.
(32) Santiago, E. I.; Varanda, L. C.; Villullas, H. M. Carbon-Supported Pt-
Co Catalysts Prepared by a Modified Polyol Process as Cathodes for PEM Fuel Cells.
Journal of Physical Chemistry C 2007, 111, 3146-3151.
(33) Alden, L. R.; Roychowdhury, C.; Matsumoto, F.; Han, D. K.;
Zeldovich, V. B.; Abruna, H. D.; DiSalvo, F. J. Synthesis, Characterization, and
Electrocatalytic Activity of PtPb Nanoparticles Prepared by Two Synthetic
Approaches. Langmuir 2006, 22, 10465-10471.
22
(34) Chan, K.-Y.; Ding, J.; Ren, J.; Cheng, S.; Tsang, K. Y. Supported
mixed metal nanoparticles as electrocatalysts in low temperature fuel cells. Journal of
Materials Chemistry 2004, 14, 505-516.
(35) Lu, Y. Surfactant-templated mesoporous materials: from inorganic to
hybrid to organic. Angewandte Chemie, International Edition 2006, 45, 7664-7667.
(36) Schueth, F. Non-siliceous Mesostructured and Mesoporous Materials.
Chemistry of Materials 2001, 13, 3184-3195.
(37) Tanev, P. T.; Pinnavaia, T. J. Mesoporous Silica Molecular Sieves
Prepared by Ionic and Neutral Surfactant Templating: A Comparison of Physical
Properties. Chemistry of Materials 1996, 8, 2068-2079.
(38) Tanev, P. T.; Pinnavaia, T. J. A neutral templating route to
mesoporous molecular sieves. Science (Washington, D. C.) 1995, 267, 865-867.
(39) Chen, Y.; Yang, F.; Dai, Y.; Wang, W.; Chen, S. Ni@Pt core-shell
nanoparticles: synthesis, structural and electrochemical properties. Journal of
Physical Chemistry C 2008, 112, 1645-1649.
(40) Yang, P.; Zhao, D.; Margolese, D. I.; Chmelka, B. F.; Stucky, G. D.
Block Copolymer Templating Syntheses of Mesoporous Metal Oxides with Large
Ordering Lengths and Semicrystalline Framework. Chemistry of Materials 1999, 11,
2813-2826.
(41) Li, Y.; Fu, Q.; Flytzani-Stephanopoulos, M. Low-temperature water-
gas shift reaction over Cu- and Ni-loaded cerium oxide catalysts. Applied Catalysis,
B: Environmental 2000, 27, 179-191.
(42) Yang, H.; Yan, Y.; Liu, Y.; Zhang, F.; Zhang, R.; Meng, Y.; Li, M.;
Xie, S.; Tu, B.; Zhao, D. A Simple Melt Impregnation Method to Synthesize Ordered
Mesoporous Carbon and Carbon Nanofiber Bundles with Graphitized Structure from
Pitches. Journal of Physical Chemistry B 2004, 108, 17320-17328.
(43) Haller, G. L. New catalytic concepts from new materials:
understanding catalysis from a fundamental perspective, past, present, and future.
Journal of Catalysis 2003, 216, 12-22.
23
(44) Liang, C.; Li, Z.; Dai, S. Mesoporous carbon materials: synthesis and
modification. Angewandte Chemie, International Edition 2008, 47, 3696-3717.
(45) Lee, J.; Kim, J.; Hyeon, T. Recent progress in the synthesis of porous
carbon materials. Advanced Materials (Weinheim, Germany) 2006, 18, 2073-2094.
(46) Ryoo, R.; Joo, S. H.; Jun, S. Synthesis of Highly Ordered Carbon
Molecular Sieves via Template-Mediated Structural Transformation. Journal of
Physical Chemistry B 1999, 103, 7743-7746.
(47) Gierszal, K. P.; Kim, T.-W.; Ryoo, R.; Jaroniec, M. Adsorption and
Structural Properties of Ordered Mesoporous Carbons Synthesized by Using Various
Carbon Precursors and Ordered Siliceous P6mm and Ia.hivin.3d Mesostructures as
Templates. Journal of Physical Chemistry B 2005, 109, 23263-23268.
(48) Ryoo, R.; Joo, S. H. Nanostructured carbon materials synthesized from
mesoporous silica crystals by replication. Studies in Surface Science and Catalysis
2004, 148, 241-260.
(49) Joo, S. H.; Choi, S. J.; Oh, I.; Kwak, J.; Liu, Z.; Terasaki, O.; Ryoo, R.
Ordered nanoporous arrays of carbon supporting high dispersions of platinum
nanoparticles. [Erratum to document cited in CA135:184349]. Nature (London,
United Kingdom) 2001, 414, 470.
(50) Gupta, G.; Stowell, C. A.; Patel, M. N.; Gao, X.; Yacaman, M. J.;
Korgel, B. A.; Johnston, K. P. Infusion of Pre-synthesized Iridium Nanocrystals into
Mesoporous Silica for High Catalyst Activity. Chemistry of Materials 2006, 18,
6239-6249.
(51) Jana, N. R.; Peng, X. Single-Phase and Gram-Scale Routes toward
Nearly Monodisperse Au and Other Noble Metal Nanocrystals. Journal of the
American Chemical Society 2003, 125, 14280-14281.
(52) Fu, X.; Wang, Y.; Wu, N.; Gui, L.; Tang, Y. Surface Modification of
Small Platinum Nanoclusters with Alkylamine and Alkylthiol: An XPS Study on the
Influence of Organic Ligands on the Pt 4f Binding Energies of Small Platinum
Nanoclusters. Journal of Colloid and Interface Science 2001, 243, 326-330.
24
(53) Gasteiger, H. A.; Kocha, S. S.; Sompalli, B.; Wagner, F. T. Activity
benchmarks and requirements for Pt, Pt-alloy, and non-Pt oxygen reduction catalysts
for PEMFCs. Applied Catalysis, B: Environmental 2005, 56, 9-35.
(54) Dresselhaus, M. S.; Thomas, I. L. Alternative energy technologies.
Nature (London, United Kingdom) 2001, 414, 332-337.
(55) Steele, B. C. H.; Heinzel, A. Materials for fuel-cell technologies.
Nature (London, United Kingdom) 2001, 414, 345-352.
(56) Mathias, M. F.; Makharia, R.; Gasteiger, H. A.; Conley, J. J.; Fuller, T.
J.; Gittleman, C. I.; Kocha, S. S.; Miller, D. P.; Mittelsteadt, C. K.; Xie, T.; Yan, S.
G.; Yu, P. T. Two fuel cell cars in every garage? Electrochemical Society Interface
2005, 14, 24-35.
(57) Wagner, F. T.; Gasteiger, H. A.; Makharia, R.; Neyerlin, K. C.;
Thompson, E. L.; Yan, S. G. Catalyst development needs and pathways for
automotive PEM fuel cells. ECS Transactions 2006, 3, 19-29.
(58) Gasteiger, H. A. Proton exchange fuel cell materials and R&D needs
for future market success. Electrochemistry (Tokyo, Japan) 2007, 75, 103.
(59) Stamenkovic, V. R.; Fowler, B.; Mun, B. S.; Wang, G.; Ross, P. N.;
Lucas, C. A.; Markovic, N. M. Improved Oxygen Reduction Activity on Pt3Ni(111)
via Increased Surface Site Availability. Science (Washington, DC, United States)
2007, 315, 493-497.
(60) Stamenkovic, V. R.; Mun, B. S.; Arenz, M.; Mayrhofer, K. J. J.;
Lucas, C. A.; Wang, G.; Ross, P. N.; Markovic, N. M. Trends in electrocatalysis on
extended and nanoscale Pt-bimetallic alloy surfaces. Nature Materials 2007, 6, 241-
247.
(61) Xiong, L.; Manthiram, A. Nanostructured Pt-M/C (M = Fe and Co)
catalysts prepared by a microemulsion method for oxygen reduction in proton
exchange membrane fuel cells. Electrochimica Acta 2005, 50, 2323-2329.
(62) Xiong, L.; Manthiram, A. Effect of atomic ordering on the catalytic
activity of carbon supported PtM (M = Fe, Co, Ni, and Cu) alloys for oxygen
25
reduction in PEMFCs. Journal of the Electrochemical Society 2005, 152, A697-
A703.
(63) Wakabayashi, N.; Takeichi, M.; Uchida, H.; Watanabe, M.
Temperature Dependence of Oxygen Reduction Activity at Pt-Fe, Pt-Co, and Pt-Ni
Alloy Electrodes. Journal of Physical Chemistry B 2005, 109, 5836-5841.
(64) Yano, H.; Kataoka, M.; Yamashita, H.; Uchida, H.; Watanabe, M.
Oxygen Reduction Activity of Carbon-Supported Pt-M (M = V, Ni, Cr, Co, and Fe)
Alloys Prepared by Nanocapsule Method. Langmuir 2007, 23, 6438-6445.
(65) Ye, H.; Crooks, R. M. Effect of Elemental Composition of PtPd
Bimetallic Nanoparticles Containing an Average of 180 Atoms on the Kinetics of the
Electrochemical Oxygen Reduction Reaction. Journal of the American Chemical
Society 2007, 129, 3627-3633.
(66) Koh, S.; Strasser, P. Electrocatalysis on Bimetallic Surfaces:
Modifying Catalytic Reactivity for Oxygen Reduction by Voltammetric Surface
Dealloying. Journal of the American Chemical Society 2007, 129, 12624-12625.
(67) Mani, P.; Srivastava, R.; Strasser, P. Dealloyed Pt-Cu Core-Shell
Nanoparticle Electrocatalysts for Use in PEM Fuel Cell Cathodes. Journal of
Physical Chemistry C 2008, 112, 2770-2778.
(68) Antolini, E.; Salgado, J. R. C.; Gonzalez, E. R. The stability of Pt-M
(M=first row transition metal) alloy catalysts and its effect on the activity in low
temperature fuel cells. Journal of Power Sources 2006, 160, 957-968.
(69) Yu, X.; Ye, S. Recent advances in activity and durability enhancement
of Pt/C catalytic cathode in PEMFC. Journal of Power Sources 2007, 172, 145-154.
(70) Shao, Y.; Yin, G.; Gao, Y.; Shi, P. Durability study of Pt/C and
Pt/CNTs catalysts under simulated PEM fuel cell conditions. Journal of the
Electrochemical Society 2006, 153, A1093-A1097.
(71) Coloma, F.; Sepulveda-Escribano, A.; Rodriquez-Reinoso, F. Heat-
treated carbon blacks as supports for platinum catalysts. Journal of Catalysis 1995,
154, 299-305.
26
(72) Yu, X.; Ye, S. Recent advances in activity and durability enhancement
of Pt/C catalytic cathode in PEMFC. Journal of Power Sources 2007, 172, 133-144.
(73) Shao, Y.; Yin, G.; Gao, Y. Understanding and approaches for the
durability issues of Pt-based catalysts for PEM fuel cell. Journal of Power Sources
2007, 171, 558-566.
27
Chapter 2
Enhanced Infusion of Gold Nanocrystals into Mesoporous Silica with
Supercritical Carbon Dioxide
Gold nanocrystal dispersions in toluene-CO2 mixtures were infused into cylindrical
pores in mesoporous silica to achieve high loadings over 2 wt % in 24 hours. The
nanocrystals were highly dispersed according to transmission electron microscopy
and the loadings approached equilibrium. In contrast, the loadings were small for
infusion with pure toluene or toluene mixed with an antisolvent, methanol. The
differences in loading were correlated with the long-ranged van der Waals forces
between the gold and silica through the intervening solvent. These van der Waals
forces became stronger as CO2 was added to toluene, as a consequence of a reduction
in the Hamaker constant of the mixed intervening solvent, resulting in stronger
nanocrystal adsorption. The decoupling of the nanocrystal synthesis step and the
infusion step, leads to exquisite control of the nanocrystal size, morphology and
dispersibility within the pores. The simplicity of the method allows for the facile
production of nanocrystal/silica composites for applications such as catalysis and
optoelectronics.
Large portions of this chapter have been previously published as Gupta, G.; Shah, P. S.; Zhang, X.; Saunders, A. E.; Korgel, B. A.; Johnston, K. P. Enhanced Infusion of Gold Nanocrystals into Mesoporous Silica with Supercritical Carbon Dioxide. Chemistry of Materials 2005, 17, 6728-6738.
28
2.1 Introduction
Decreasing the size of metals and semiconductors leads to unique size
dependent electronic, optical and catalytic properties, which deviate from the bulk.1-4
One of the key challenges is to control the arrangement of nanocrystals in various
substrates including block copolymer patterned templates,5 nanoscale features
patterned by e-beam lithography6 and nanoporous materials7 as described in Table
2.1. Reactive precursors may be deposited into a porous material from either the gas
phase or via a wetting liquid in order to synthesize nanocrystals with the pores.8-13
Here, the size and the shape of the pores may be used as a template to control the
particle size and shape, to form spherical particles10 or in the case of cylindrical pores,
nanowires.9,11,12,14 The diffusion of precursors and nanocrystals through the solvent
may be undesirably slow for mesopores. It took about 3 weeks for HAuCl4 aqueous
precursor solution to completely infiltrate mesoporous silica,13 perhaps indicating
diffusional resistances of entrapped air in regions of variable curvature.15
Relatively few studies have examined infusion of pre-synthesized
nanocrystals into porous materials with carrier solvents. Such nanocrystals have been
imbibed into the pores by capillary forces, in conjunction with additional driving
forces including vacuum induction and electrophoresis.16 With this technique, it can
be challenging to fill the pores uniformly. Nanocrystals synthesized in the aqueous
core of the reverse micelles have subsequently been infused into nanoporous
matrices.17,18 The ability to isolate the particles from the unreacted precursors and to
keep the particles coated with surfactant can be challenging in these micelles. Yet
another approach is to entrap pre-synthesized nanocrystals during the synthesis of a
nanoporous matrix, although the nanocrystals can in some cases perturb the geometry
of the matrix.7,19
There are two primary challenges in the infusion of pre-synthesized
nanocrystals into a nanoporous matrix. First, the nanocrystal dispersion must be
transported through the pores by favorable capillary wetting, and the nanocrystals
29
Nanoporous Material
Material impreg-nated
Pore diameter
(nm)
Nano- particle
Size (nm)Loading Method of
loading Reference
FSM-16 Pt 2.4 2.3±0.1 0.9 wt % Reduction of precursor
8
FSM-16 Pt 3.5 2.9±0.16 2.9 wt % Reduction of precursor
8
Thiol modified MCM41
CdS 3.8 3.3 -
Impregnation of
nanocrystals from reverse
micelle
18
Planar-like monolithic Au 5.7 5.2 -
Reduction of precursor by sonification
13
HMM-2 Au d100=2.2 3.2±0.5 2 wt % Reduction of Precursor
72
SBA-15 Pt d100= 10.23 2-5
1017
particle/gm
silica
Silica synthesis in presence of nanocrystals
62
Porous Alumina
membrane Au 50 15 -
Impregnation of
nanocrystals by vacuum induction
16
SBA-15 Au 8.8 2.2±0.3 2.3 wt %Impregnation
of nanocrystals
♦
MCM-41 Au 2.9 2.2±0.3 2.0 wt % -do- ♦ MCM-41 Au 2.4 2.2±0.3 1.7 wt % -do- ♦ * Time and temperature are only shown for the infusion of either the precursor or the pre-synthesized silica. ♦ shows results for our work.
Table 2.1
Methods for loading of metal nanoparticles into mesoporous silica.
30
must not block small pore entrances. Once the nanocrystals penetrate the pores, the
second challenge is to achieve sufficient adsorption of the nanocrystals from the
solution phase to the substrate surface. A major dilemma must be addressed. The
solvent must allow for enough attraction between the nanocrystals and substrate to
achieve sufficient adsorption, while simultaneously preventing flocculation of
nanocrystals in the solvent phase. The nanocrystals must be dispersed in “good”
solvent for the stabilizing ligands to prevent flocculation. However, good solvents
may be expected to cause the interactions between the nanocrystals and the surface of
the nanoporous material to be repulsive or weakly attractive.20 The resulting weak
adsorption of the particles inside the pores may produce low loadings. Thus, the
solvent must be tuned to provide a judicious balance between nanocrystal-nanocrystal
versus nanocrystal-substrate interactions.
The penetration of reactive precursors into high aspect ratio features21 and
nanoporous materials such as activated carbon22 is particularly efficient with
supercritical fluids, due to high diffusion coefficients and low viscosities. Examples
include the synthesis of porous silica fibers23 and metal oxide fibers24 within activated
carbon fiber templates. Silicon and germanium nanowires have been synthesized in
mesoporous silica from reactive precursors using sc-hexane and sc-CO2,
respectively.25,26 High diffusivities in Sc-CO2 facilitate transport of the precursors
into the pores. The miscibility between CO2 and air or organic solvents often leads to
favorable contact angles and wetting, as in the case of CO2 enhanced oil recovery in
small sandstone pores27 and in the cleaning of low k nanoporous dielectric insulators. 28
CO2 is used extensively to swell organic solvents to lower the solvent strength
to precipitate polymers from solution29 and to form nanoparticles of many types of
materials30-33 including pharmaceuticals. The influence of CO2 as an antisolvent on
the solubility of solutes34 and on spectral shifts of solvatochromic probes35,36 has been
determined over a wide range in CO2 concentration. CO2-expanded liquids (CXLs)
have been exploited for homogeneous catalytic oxidations37,38 and for hydrogenation
31
and hydroformylation reactions.39,40 The presence of CO2 in the mixed medium
increases the solubility of O2 by 100 times and also improves diffusion rates. Tunable
properties of CXLs has been utilized for the rapid and precise size selection of the
nanocrystals.41
Herein, we show that the use of supercritical CO2 as an antisolvent helps
overcome both transport and thermodynamic challenges in nanoparticle infusion into
porous materials. Our objective is to add CO2 to an organic solvent, toluene, to
achieve high levels of infusion of pre-synthesized nanocrystals into mesoporous silica
within a 24 hour period. The CO2 concentration in toluene is tuned to produce
sufficiently strong nanocrystal-silica interactions for high nanocrystal adsorption in
the pores without flocculating the nanocrystals in the solvent phase. At 35°C and
pressures of this study above 68 bar,42 CO2 forms a homogeneous one-phase solution
with toluene. Dodecanethiol capped metal nanocrystals cannot be dispersed in pure
CO2;4 however they are soluble in our CO2-toluene mixtures. The loading of the gold
particles was determined by measuring spectrophotometrically, the reduction in the
absorbance of gold in the supernatant after separation from the mesoporous silica
phase. As a control, the infusion was also studied (1) without CO2 present and (2)
with methanol as an antisolvent in toluene. The diameters of the porous cylinders in
the mesoporous silica were varied from 2.5 nm to 9 nm. The kinetics of infusion was
studied over a period of 3 to 72 hours, both with and without CO2, to understand the
effect of changes in the transport properties of the nanocrystals. The effects of CO2 on
both the kinetics and the final loading are analyzed relative to non-expanded liquid
solvents, toluene or toluene-methanol mixtures. In the first part of the discussion
section, we analyze the equilibrium loading and then examine the kinetic aspects in
the second section.
A key advantage of impregnating pre-synthesized nanocrystals into pores is
that the nanocrystal synthesis may be controlled with stabilizing ligands to produce
the desired size, crystallinity, shape, and surface properties. Numerous synthetic
schemes have been reported.2 The nanocrystal morphology may be chosen
32
independently of the pore morphology, as long as the crystals are smaller than the
pores. Prior to infusion, the particles may be isolated from unwanted byproducts,
cleaned, separated by size, and the ligands may be exchanged if desired for
alternative ligands. It would be difficult if not impossible to execute these post-
reaction steps if the nanocrystals were synthesized inside the nanoporous material.
Finally, synthesis inside nanoporous materials can be perturbed by surface chemistry
due to the substrate, for example, capping ligands may adsorb to the substrate rather
than the nanocrystals.
2.2 Experimental
2.2.1 Nanocrystal Synthesis
Gold nanocrystals were synthesized by a two-phase (organic-water) arrested
precipitation technique,2 using dodecanethiol (C12H25SH) as a capping agent. Initially,
36 ml of aqueous (0.03 M) hydrogen tetrachloroaurate(III) trihydrate (HAuCl4.3H2O)
was combined with 25 mL of toluene solution containing 2.7 gm of phase transfer
catalyst, tetraoctylammonium bromide. After stirring for 1 h, the organic phase, with
the transferred gold, was collected. The gold salt was then reduced using 15 mL of an
aqueous sodium borohydride (NaBH4) solution (0.44 M) resulting in gold
nanocrystals dispersed in toluene, protected by the phase transfer catalyst. To this
solution, 240 mL of dodecanethiol was added immediately and stirred over night. The
organic phase rich in gold nanocrystals was collected. Methanol was added to the
organic phase and the solution was centrifuged to separate the nanocrystals. The
nanocrystals were washed and redispersed in 20 mL of toluene.
2.2.2 Mesoporous Silica Synthesis
Mesoporous silica SBA-15 was synthesized via a procedure similar to that
reported by Zhao et al.43 A solution of 1.6 g EO20PO70EO20 : 7 mL HCl: 8 mL TEOS:
38 mL H2O was prepared, stirred for 24 h at 50°C and then heated at 100°C for 24 h.
33
The resulting slightly yellow powder was dispersed in ethanol, filtered and then again
heated at 100°C for 24 h. The yellow product was then calcined at 550°C.
Mesoporous silica MCM-41 was synthesized by S+X-I+ assembly via a
procedure similar to that reported by Tanev et al.44 CnH2n+1N(CH3)3Br template
(n=10, 16) was used as a structure directing agent. Depending on the chain length of
the surfactant, MCM-41 silica with different pore diameters was obtained. Small pore
diameter silica (S-MCM41) was obtained with surfactant chain length n=10 while
large pore diameter silica (L-MCM41) was obtained with chain length n=16.
For the synthesis of S-MCM41, a solution of 0.8 g C10H21N(CH3)3Br: 7 mL
HCl: 5 mL TEOS: 38 mL H2O was prepared, stirred overnight at ambient temperature
and then heated at 100°C for 6 h. The ratio of the components in the solution is
different than that reported by Tanev et al.44 The while powder was then dispersed in
ethanol, filtered and dried at 100°C for 24 hours. The collected solid was then
calcined at 600°C. The L-MCM41 was synthesized similarly to S-MCM41 except
that a solution with composition 1.6 g C16H33N(CH3)3Br: 7 mL HCl: 8 mL TEOS: 38
mL H2O was used.
2.2.3 Infusion Procedure
In a typical experiment, 20 mg of mesoporous silica and 2.5 mL of gold nanocrystals
dispersed in toluene, with a concentration of 0.3 mg/ml, were loaded in the reaction
cell with a working volume (volume with piston in the cell) of 27 mL. 0-5 mL of
toluene was added in the cell. The cell was then sealed and then an amount of CO2
equivalent to 2.5 – 7.5 mL at 241 bar (3500 psi) and 35°C was added so as to produce
a total volume of reaction mixture of 10 mL. The cell was then pressurized to 241 bar
with CO2 at the backside of the piston using an ISCO model 100 DX Computer
Controlled syringe pump. The contents of the cell, in a water bath, were stirred using
a magnetic stir bar that was inserted inside the cell (Figure 2.1). The reaction was
maintained at 35°C and 241 bar (3500 psia) and allowed to proceed for 24 h. After
the reaction, the cell was cooled to 0°C and CO2 was then vented as a vapor from the
35
top of the cell, leaving most of the other components in the cell. Lowering the
temperature to 0°C reduces the loss of toluene vapor along with CO2. The supernatant
was collected and filtered, using a polycarbonate 0.05 μm filter, to separate the
mesoporous silica. The volume of the supernatant was found to be slightly less than
the total volume of gold feed solution and toluene added to the cell because of some
toluene loss along with CO2. The loss of toluene was usually about 5-6 vol%. The
extent of incorporation of gold nanocrystals in the silica was determined by
subtracting from the initial mass of the nanocrystals, the final mass of the
nanocrystals recovered in the supernatant. The absorbance of the gold nanocrystals
dispersed in toluene before and after infusion was measured at a wavelength of 500
nm using a Cary 500 UV-Vis-NIR spectrophotometer to determine the nanocrystal
mass. An optical path length of 1 cm was used. A standard calibration curve was
generated at 500 nm using known concentrations of gold nanocrystals in toluene as
shown in Figure 2.2. A typical starting gold concentration was 0.3 mg/ml and after
infusion the value decreased to 0.195 mg/ml for a loading of 2 wt % and to 0.17
mg/ml for a loading of 2.23 wt %. The experiments were repeated several times to
check for the reproducibility, which was on the order of +0.3 in wt % units.
2.3 Results
2.3.1 Nanoparticle Characterization
SAXS was used to characterize mesoporous silica structure as well as the gold
nanocrystal dispersions. The intensity of scattered radiation, I(q), is proportional to a
shape factor and structure factor, P(q) and S(q), respectively: )()()( qSqPqI ∝ . Here,
q is the wave vector, defined as q = 4π/λ sin(θ), where λ is the wavelength (0.154
nm) and 2θ is the scattering angle. The wave vector is also inversely proportional to a
characteristic distance, d, in the scattering system: q = 2π/d. The shape factor is
related to the size and shape of the scatterers, whereas the structure factor contains
information about interactions or ordering within the system.
36
0
0.1
0.2
0.3
0.4
0.5
0.6
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
Abs
orba
nce
at 5
00 n
m
gold concentration (mg/ml)
Figure 2.2
Calibration curve of absorbance versus concentration measured at 500 nm for gold
nanocrystal solutions.
37
For dilute solutions of non-interacting nanocrystals S(q) = 1, and the scattered
intensity relates only the shape factor of the particles dispersed in solution , i.e. I(q) ∝
P(q).48-51 The shape factor for spherical particles is given analytically
by52,53 ( )[ ]23)/()cos()sin(3)( qRqRqRqRqP −= , where R is the nanocrystal radius.
The effects of sample polydispersity on the scattered intensity can be taken into
account using ∫∝ RRqPRNqI d)()()( 6 , where N(R) is a normalized size
distribution.48,54 Here the size distribution is assumed to be Gaussian, where Ravg and
σ are the average radius and standard deviation, respectively. Fitting this model to the
scattering data from the toluene dispersion of gold nanocrystals (Figure 2.3) indicates
that the gold nanocrystals were relatively monodisperse with a mean diameter of
2.2±0.3 nm. The curve in Figure 2.3 does not show any peaks, indicating weak
interactions between the particles due to the good solvation of the ligands, as seen
previously for nanocrystals in pure organic solvents.54 In contrast, strong attractive
interactions of perfluoropolyether coated gold nanoparticles have been observed with
SAXS in CO2 due to the stronger Hamaker interactions between the cores, and the
limited solvation of the ligands, both resulting from the weak solvent strength.55
2.3.2 Characterization of Mesoporous Silica
SAXS was used to characterize the mesoporous silica powders prior to
infusion. Highly oriented pores, arranged in a 2-D hexagonal lattice, give rise to a
series of Bragg peaks in the scattering pattern. Figure 2.4 and Table 2.2 give the small
angle XRD pattern for three mesoporous silicas, SBA-15 (a), L-MCM41 (b) and S-
MCM41 (c). For SBA-15, the four well-resolved peaks are indexable as the (100),
(110), (200), and (210) reflections associated with p6mm hexagonal symmetry. One
additional weak peak is found at q(nm-1) =1.82 and corresponds to a (300) reflection.
Presence of multiple sharp peaks indicates that calcined SBA-15 has a high degree of
hexagonal mesoscopic organization. The first three peaks for SBA-15 are located at
38
0.01
0.1
1
10
0 0.5 1 1.5 2 2.5 3 3.5
I(q)
q (nm-1)
Figure 2.3
Intensity I(q) vs q for gold nanocrystals dispersed in toluene determined by SAXS.
39
10
100
1000
104
105
106
107
0 0.5 1 1.5 2 2.5 3 3.5 4
I(q)
q (nm -1)
a
b
c
100
110200
210
100 110
100
Figure 2.4
SAXS for calcined mesoporous silica SBA-15 (a), L-MCM41 (b) and S-MCM41 (c).
40
* Average length of the pore was considered 10 microns for all three mesoporous silica Table 2.2
Properties of mesoporous silica.
41
q100 = 0.61, q110 = 1.06, q200 = 1.22 nm-1 with the peak position ratio of 1: 3 : 2,
which is consistent with the literature.56-58 The distance d100 between adjacent planes
is given by d100 = 2π/ q100 = 10.29 nm. The unit cell parameter a0 = 3/2 100d = 11.9
nm gives the center-to-center spacings of the pores and agrees fairly with the
literature.43
L-MCM41 shows two indexable peaks corresponding to (100) and (110)
reflections. The peaks are broad and show diffuse scattering centered at q = 1.63 nm-1
corresponding to reflection from the (100) plane, which manifests a lower degree of
order as compared to SBA-15. The distance d100 between adjacent planes is 3.85 nm
and the unit cell parameter a0 = 4.44 nm for L-MCM41. S-MCM41 exhibits single
d100 reflections accompanied by a diffuse scattering centered at ~ 2.95 nm with a unit
cell parameter of 3.4 nm. Pinnavaia et al. have demonstrated that “single-reflection”
MCM-41 still can have short ranged hexagonal symmetry.44,59 They show that MCM-
41 synthesized by acidic S+X-I+ pathway with surfactant chain length <14 result in
poorly ordered products. For S-MCM41, the lower degree of order can be attributed
to the small chain length of only 10 for the structure directing agent.
Figure 2.5 gives the adsorption-desorption isotherms for the three calcined
silica samples. Representative nitrogen adsorption/desorption isotherms and
corresponding pore size distribution (BJH method) are shown in Fig 2.5(a). SBA-15
had a mean pore diameter of 8.8 nm, BET surface area of 700 m2/g and a pore
volume of 1.09 cc/gm (refer Table 2.2), consistent with the literature.43 Three well
distinguished regions of the adsorption-isotherm are evident: (i) monolayer-multilayer
adsorption, (ii) capillary condensation and (iii) multilayer adsorption on the outer
particle surfaces. The approximate pore size calculated using the BJH analysis is
significantly smaller than the center-to-center distance between adjacent pores
determined by SAXS, because the later includes the thickness of the pore wall. From
the difference in these two values, the thickness of the pore wall is estimated to be 3.1
nm for SBA-15 which is in excellent agreement with the values reported by Zhao et
42
0
100
200
300
400
500
600
700
800
0 0.2 0.4 0.6 0.8 1
Vol
(cc/
g)
P/Po
Desorption
Adsorption
0
0.005
0.01
0.015
0.02
0.025
0 50 100 150 200
Dv(
r) [c
c Å
-1 g
-1]
Pore Diameter (Å) (a)
0
50
100
150
200
250
300
350
400
0 0.2 0.4 0.6 0.8 1
Vol
(cc/
g)
P/Po
Adsorption
Desorption
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
15 20 25 30 35 40
Pore
Vol
ume
(cc
Å1
g1 )
Pore Diameter (Å) (b)
140
160
180
200
220
240
260
0 0.2 0.4 0.6 0.8 1
Vol
(cc/
g)
P/Po
Adsorption
Desorption
0
0.005
0.01
0.015
0.02
0.025
0 10 20 30 40 50
Pore
Vol
ume
(cc
Å-1
g-1)
Pore Diameter (Å) (c)
Figure 2.5
N2 isotherms and pore size for SBA-15 (a), L-MCM41 (b) and S-MCM41 (c).
43
al.43 The adsorption-desorption isotherm for L-MCM41 and S-MCM41 and their
corresponding pore size distributions, determined by SF method,46 are shown in
Figures 2.5(b) and Figure 2.5(c) respectively and Table 2.2.
As can be seen in Figure 2.5(a), SBA-15 shows a type-H1 hysteresis loop with
capillary condensation at a relative pressure of P/Po= 0.7.43 The step in the hysteresis
loop is very steep with a jump from 300 cc/gm to 700 cc/gm due to capillary
condensation. The slope and the height of the step are clear indications of very well
defined mesopores with narrow pore size distribtution,44 which is further
substantiated by the multiple peaks arising in SAXS (Figure 2.4).
MCM-41 (Figures 2.5(b) and 2.5(c)) shows a type 4 adsorption isotherm44
with a small degree of hysteresis. This behavior is typical of MCM-41.44 The pore
sizes of 2.5 and 2.9 nm are smaller as compared to that of SBA-15. The adsorption
increases gradually with an increase of P/Po without a steep jump, which suggests that
the pores have a wider pore size distribution than for SBA-15. The surface areas
obtained for L-MCM41 and S-MCM41 were 825 and 738 m2/g, respectively, and are
typical for MCM-41 mesoporous silica. The adsorption volume, for P/Po approaching
unity, becomes the pore volume due to complete condensation of N2. S-MCM41 to
SBA-15 (refer Table 2.2) resulting in increase of the adsorption volumes at STP.
2.3.3 Nanoparticle Dispersibility in Toluene-CO2
To date only fluorinated ligands have been shown to disperse nanocrystals in
pure CO2.3,4,60,61 However, we show that standard alkane capped nanocrystals may be
dispersed in the toluene-CO2 co-solvent system. Nanocrystal dispersions were
analyzed with a Cary 3E UV-Vis spectrophotometer in a vertically mounted variable-
volume high-pressure optical cell with a path length of 2 cm. The temperature was
controlled at 35°C with heating tape and the pressure was 241 bar. At a concentration
of CO2 in the solution of 50 vol%, the absorbance at 500 nm remained 1.92±0.02 for
a period of 8 hrs indicating a stable dispersion.
44
2.3.4 Effect of CO2 Concentration on Loading of Gold Nanocrystals in
Mesoporous Silica
Figure 2.6 shows the loadings of the nanocrystals in mesoporous silica as a
function of CO2 concentration. For nanocrystals dispersed in pure toluene, there is
virtually no infusion into the pores, as loadings were less than 0.1 wt %. The loadings
increase linearly with CO2 concentrations of 25% and 50% and then level off at the
highest value of 75%. It is not possible to go to higher CO2 concentrations, as the
particles become unstable and start to precipitate, indicating insufficient solvation of
the hydrocarbon capping ligands.
To investigate the effect of hydrostatic pressure on trapped air in the pores,
infusion was conducted in pure toluene at 241 bar, in absence of carbon dioxide. Air
may potentially be trapped between two liquid domains at both ends of a cylindrical
pore. A very small loading of 0.3 wt % was observed indicating that trapped air had
little effect. The pores in the silica are interconnected to each other via small diameter
channels. Perhaps the high pressure forced some of the particles into these
interconnects between pores.15
For CO2-toluene mixtures, the relatively large loadings above 2% comparable
to those obtained by the other infusion techniques in Table 2.1. We estimate, based on
the molecular weight of the nanocrystals (50,000 gm/mole of nanocrystals), a loading
of about 1017-1018 nanocrystals per gm of silica, for each of our types of silica. A
rough estimate of the order of number of pores per gm of silica (N) can be made from
the specific surface area. For a surface area (S) of 1000 m2/g and a radius (r) of 1 nm
and pore length (l) of 10 microns (assuming that the pore length is equal to the
particle size), we estimate from N=S/(2πrl) that there are about 1016 -1017 pores per
gram of silica. According to this estimate, our loading was about 10-100 particles per
pore on average. Konya et al62 obtained 1016 particles per gm of mesoporous silica,
roughly equal to 2 particles per pore, which corresponds to 1016 pores per gram,
similar to our value. It is important to note that we have obtained about 10 times
45
0
0.5
1
1.5
2
2.5
3
0 25 50 75 100Wt%
Gol
d Lo
adin
g (m
g A
u/m
g Si
lica)
Vol% CO2
Figure 2.6
Loading of gold nanocrystals versus vol% CO2 for 24 hour impregnation period in L-
MCM41. () 1 atm and 23°C ;() 241 atm and 35°C.
46
higher loadings than Konya et al.62 At higher loadings Konya et al62 observed that the
crystallinity of the silica decreased and worm like channels appeared due to the
presence of the nanocrystals.
2.3.5 Effect of Pore Size and Concentration of Nanocrystals on the Loading
Figure 2.7 shows that the loadings of nanocrystals in the S-MCM41, L-
MCM41 and SBA-15 silicas from pure toluene dispersions were negligible. However,
with the addition of 50% CO2 the loadings were large in all cases and increased
modestly with the pore diameter from 1.7% to 2.4%. The loading of the 2.2 nm
particles into the silica with 2.4 nm pores was well over half the loading for silica
with 8.8 nm pores. Thus the diameters of the pores along the pore length appeared to
be relatively uniform such that the particles did not block many narrow regions. The
surface areas were similar for the three types of silica. The concentration of gold in
the feed solution was increased ten fold for L-MCM41 to 3 mg/ml, as shown in
Figure 2.7. The loading increased a moderate amount from 1.97% to 2.37%.
2.3.6 Kinetic Study of the Infusion Process
For pure toluene, the nanocrystal loadings were negligible even after 7 days.
Detailed kinetic studies were performed for 50% CO2 with the silica with the smallest
pores, S-MCM41. Here the mean diameter of the gold nanocrystals was 1.95 nm with
a standard deviation/mean of 14% in contrast with the somewhat larger nanocrystals
for our other systems. The loadings increased with time but appeared to reach an
asymptote at 72 hours suggesting an approach to equilibrium (Figure 2.8). Over the
first 24 hours, where most of the infusion occurred, the loading was linear in t1/2 as
shown in Figure 2.8(a).
47
0
0.5
1
1.5
2
2.5
3
0 2 4 6 8 10
Wt%
Gol
d L
oadi
ng
Diameter of silica pore (nm)
Figure 2.7
Variation of loading of gold nanocrystals with the pore diameter of silica. The hollow
squares () are for loadings at 50 vol% CO2, solid circles () for 0 vol% CO2 and ♦
for loading at 50 vol% CO2 with the feed at 10 times higher concentration.
48
(a)
0
0.5
1
1.5
2
2.5
3
3.5
0 10 20 30 40 50 60 70 80
Wt%
Gol
d Lo
adin
g
Time (hr)
(b)
Figure 2.8
Loading versus time for S-MCM41. (a) Loading versus t1/2 for up to 24 hours with a
correlation coefficient of 99.5%. (b) Loading versus time approaching equilibrium.
0
0.5
1
1.5
2
2.5
0 1 2 3 4 5
y = 0.45826x R= 0.99541
Wt%
Gol
d L
oadi
ng
sqrt (t) (hr0.5)
49
2.3.7 TEM Micrographs of Infused Silica
Figure 2.9 shows typical images of gold nanocrystals infused in the three
types of silica. The pore sizes and structures appear to be uniform. In each case, large
numbers of highly dispersed nanocrystals are present in each pore, consistent with the
estimate of about 100 particles/pore above. The presence of large numbers of
nanocrystals throughout each pore corroborates the relatively high loadings measured
by spectroscopy in Table 2.1. The presence of large number of particles in the pores
was most likely for SBA-15 where the pore diameter is more than 4 times the
diameter of the nanocrystals. Hence, the adsorption of a nanocrystal in the pore would
be less likely to block the entry of a second nanocrystal. However, in the case of L-
MCM41 and S-MCM41, where the ratio of the diameter of pore to the diameter of the
particle is less than 2, the chances of pore blocking were much larger. Thus, the
presence of large numbers of particles in each pore suggests that the adsorbed
particles are mobile and can move toward the center of the pore by diffusion on the
surface and/or in the bulk.
2.3.8 Effect of a Conventional Liquid Anti-Solvent on Infusion
The effect of a liquid antisolvent, methanol, on the loading was used as a
control experiment to shed further insight into the above experiments for the
supercritical antisolvent, CO2. Polar solvents are utilized as antisolvents for size
selective precipitation of alkanethiol coated nanocrystals.2 The addition of methanol
may influence the nanocrystal-silica and -solvent interactions, but will not lower the
viscosity of the solvent as is the case for CO2. It was observed visually that gold
nanocrystals at a concentration of 0.3 mg/ml dispersed in 50% methanol
concentrations were stable for at least 24 hours. At a level of 55% methanol-45%
toluene, the dispersions became unstable with a visible settling front in less than one
hour. In order to adsorb nanocrystals in the pores without aggregation in the solvent
mixture, we chose a methanol concentration of 50% for the infusion experiments.
51
An amount of 20 mg of S-MCM41 was added to a 10ml stirred solution of 1.95 nm
gold nanocrystals (0.4 mg/ml). After 24 hours, the loading was 0.56±0.50 %. After 7
days it reached 0.80±0.18 %, suggesting that the differences for methanol and CO2
are not due simply to kinetic factors. These two loadings are within the experimental
uncertainty indicating that the system approached equilibrium within 24 hours. These
loadings are less than half of the values obtained with CO2.
2.4 Discussion
We begin by demonstrating that toluene wets the pores rapidly, and thus
capillary wetting does not limit the infusion of the nanocrystals. The capillary forces
become substantial for the cylindrical nanopores as described by the Laplace pressure
RP LV θγ cos2
=Δ (2.1)
where γLV is the surface tension of the liquid, R is the pore radius, and θ is the contact
angle between the surface of the pore and the liquid. Toluene with a surface tension
of 28 dynes/cm, wets silica with a contact angle of about 6º.20 The Laplace pressure
for wetting of a 4 nm pore for a toluene-silica-air interface of 278 bars provides a
strong driving force for capillary wetting. On the basis of the conservative assumption
that the initial air in the pores does not dissolve in toluene (if toluene is already pre-
saturated with air), the air will be compressed from 1 bar to 278 bar and occupy a
negligible volume, even if it is not displaced through one of the pore ends. Thus
almost all of the pore volume will be available to the toluene. The capillary wetting
by each of the other solvent mixtures in this study will also be favorable, as the
contact angles will be well below 90o.
The rate at which the liquid solvent penetrates the pores can be estimated
using the Washburn equation, (combination of Laplace and Poiseuille equations)63
hR
dtdh
ηθγ
4cos
= (2.2)
52
where η is the bulk viscosity. For cylindrical pores that extend the length of the
particle, h is on the order of 20 μm. The time to fill a 4 nm diameter pore, is on the
order of the order of 10-4 seconds. Thus, the diffusion and adsorption of the
nanocrystals in the pores required several orders of magnitude more time than solvent
infusion.
2.4.1 Equilibrium Nanocrystal Loading
In each of the solvents, pure toluene, toluene-CO2, and toluene-methanol, the
nanocrystal loading did not increase significantly after 24 hours, indicating an
approach to equilibrium. For toluene-CO2, the similar loadings for the three types of
mesoporous silica are consistent with the similar surface areas, despite the difference
in pore diameters. The difference in pore diameter did not appear to cause an
appreciable difference in the loading kinetics. The similar loadings for the similar
surface areas suggest that the loadings approached equilibrium. The equilibrium
adsorption of a nanocrystal in a pore onto the silica surface may be expected to be
driven primarily by the chemical potential of the nanocrystals in the solvent and long-
ranged van der Waals forces between the particles and silica surface, since the ligands
are nonpolar. For simplicity, we treat the silica surface as a flat wall. The Hamaker
constant A132, for a gold particle (1) and silica surface (2) across a medium (toluene-
CO2 mixture, 3) may be approximated by20
))(( 33223311132 AAAAA −−= (2.3)
where Aii is the Hamaker constant for pure i, which can be estimated using a
simplification of the Lifshift theory.20
( )( ) 2/322
2222
2163
43
vacuumy
vacuumye
vacuumy
vacuumybyy
nn
nnhTkA
+
−+⎟
⎟⎠
⎞⎜⎜⎝
⎛
+
−=
νεεεε
(2.4)
53
where eν is the maximum electronic ultraviolet adsorption frequency, typically
assumed to be 15103× s-1.4 The A for gold, 19104 −× J,20 the ε and n for silica, 3.91 and
1.45, respectively64,65 and for toluene, 2.38 and 1.496, respectively66,67 were taken
from the literature. The Hamaker constants for silica and toluene calculated from
equation (2.2) are 201097.5 −× J and 201098.6 −× J, respectively.
The A33 was determined for the binary solvent mixture from equation 2.4
along with the Lorenz-Lorenz mixing rule,
∑ +−=+− 2222 )2/()1()2/()1( iiimixmix nnnn φ (2.5)
where φi is the volume fraction of component i. An equivalent expression for the
dielectric constant is obtained by replacing n with ε1/2 in equation 2.5. The n and ε for
CO2 were obtained as a function of density from the following.68
34242
2
10171.310412.107016.021
rrrnn ρρρ −− ×−×+=
+− (2.6)
202602.02386.01 rr ρρε +=− (2.7)
where /r cρ ρ ρ= . At the reaction conditions of 241 Bar and 35°C, the n and ε for
CO2 were determined to be 1.2 and 1.12, respectively. The A for the pure CO2 was
calculated from equation 2.2 to be 201044.1 −× J. The results for A33 are summarized
in the Table 2.3.
The Hamaker constant for the gold-silica interaction across pure toluene, A132, turns
out to be negative, indicating an unfavorable driving force for adsorption of the
54
Vol. % CO2
A33 × 1020 (J) A132 × 1020 (J)
100 1.44 6.38
75 2.81 3.57
50 4.19 1.70
25 5.56 0.34
0 6.93 -0.31
Table 2.3
Overall Hamaker constants A132 between silica surface and gold core mediated by
supercritical CO2-toluene solution. Component 1 is gold, 2 is SiO2, and 3 is a mixture
of toluene and CO2.
55
nanocrystals. The very low loading of nanocrystals from pure toluene is consistent
with this result. With the addition of CO2, A132 becomes increasingly positive, which
would lead to an increase in the adsorption as observed. The loadings are correlated
with the particle-silica van der Waals forces as characterized by A132 as listed in
Table 2.3, which becomes more attractive with an increase in CO2 concentration. In
addition, the solvation of the nanocrystals in the solvent phase decreases with added
CO2 increasing the chemical potential of the nanocrystals. Both of these factors will
produce stronger adsorption resulting in higher loadings, as observed.
This type of analysis was also being applied to methanol as an antisolvent. For
20% methanol, A132 becomes positive, indicating attractive van der Waals forces
(Table 2.4). The A132 value with 50% methanol is much weaker than in the case of
only 25% CO2. Therefore, the much weaker adsorption for methanol as an antisolvent
relative to CO2 is consistent with the A132 values.
Pure CO2, with an exceptionally low polarizability/volume60 has a much lower
A than methanol or just about any other antisolvent. For example, CO2 produces a
greater reduction in A when add to toluene than does methanol. As the A of the
intervening solvent, A33, decreases, A132 always becomes less negative or more
positive in equation 2.1, thus favoring adsorption. Therefore, CO2 provides an
opportunity to achieve a greater range in the tunability of a mixed solvent, which in
our case plays a key role in achieving high nanocrystal loadings.
2.4.2 Transport Aspects of Nanocrystal Loading
Koone and Zerda69 reported that the diffusion coefficient of water in a
nanoporous glass of diameter 2.9 nm is of the order of 10104 −× m2/s, while the
diffusion coefficient of water in the unbound system was 9102.2 −× m2/s. A much
greater reduction in diffusion rate may be expected for the nanocrystals in the pores
due to the much higher aspect ratio or solute size to pore diameter. Vigne-Maeder et
al. performed molecular dynamics simulations of the diffusion of gases in zeolites.70
56
Vol. % methanol A33 × 1020 (J)** A132 × 1020
(J)
50
5.84
0.52
40
6.05
0.34
30
6.27
0.17
20
6.49
0.003
10
6.71
-0.16
0
6.93
-0.23
** Hamaker constant of solvent interacting with itself in vacuum.
Table 2.4
Overall Hamaker constants A132 between silica surface and gold core mediated by
methanol-toluene solution. Component 1 is gold, 2 is SiO2, and 3 is a mixture of
toluene and methanol.
57
The primary mechanism for transport was adsorption followed by surface diffusion to
a pore mouth, rather than direct entrance to the pore. The importance of surface
diffusion increased as the aspect ratio of particle diameter to pore diameter increased.
The nanocrystals diffuse from the entrance of the pore at both ends towards
the center along the concentration gradient. The diffusion will take place both in the
solvent within the pore and on the surface. In Figure 2.8(a), the loading of the
nanocrystals is plotted as a function of t1/2 with a correlation coefficient of 0.995.
According to the Stokes-Einstein relationship the diffusion coefficient of 2 nm gold
nanocrystals in bulk toluene is 10102 −× m2/s. However, the actual diffusion
coefficient in the pores was lower due to confinement in that it took about 24 hours to
approach the equilibrium loading as seen in figure 2.8(b). The effective diffusion
coefficient in the pores of length l, including surface diffusion,71 may be
approximated by the equation Dtl = . For 10 micron long pores in the particles, the
diffusion coefficient was calculated to be 10-15 m2/s.
When the diameter of the pore approaches that of the particle, the mass
transfer resistance for the entry of the particle into the pore increases significantly and
diffusion through the pore becomes considerably difficult. The aspect ratios, or
particle/pore diameters are 0.9, 0.76, and 0.25 for our three silicas. The slight increase
in the loadings with the pore diameter (Figure 2.7) can be attributed to the larger
pores and the increase in the structural ordering. The degree of order was shown to
increase with an increase in pore size. The presence of well-ordered uniform
cylindrical pores would lead to less pore blockage than disordered pores. As seen
from the SAXS data, the degree of long-range ordering is highest for SBA-15
followed by L-MCM41, and finally S-MCM41 with limited order. It is observed that
the Even with relatively little order in S-MCM41, the loadings are significant,
indicating that kinetic limitations were minor.
The similar loadings for the three silica supports with similar surface areas
suggest that blocking of pores by nanocrystals is not significant. In the case of S-
58
MCM41 where the aspect ratio of the particle to the pore is 0.9, surface diffusion
must have been significant to prevent blocking of the pores. For this high aspect ratio,
the polydispersity was rather low, otherwise, many smaller pores would not have
been accessible.
2.5 Conclusions
A novel approach is presented to infuse pre-synthesized gold nanocrystals into pre-
formed mesoporous silica by tuning the solvent quality with CO2 as an antisolvent.
The concept of decoupling the nanocrystal synthesis step and the infusion step,
provides exquisite control of the nanocrystal size, morphology and dispersibility
within the pores, without perturbing the morphology of the mesoporous silica.
High loadings of nanocrystals in mesoporous silica over 2 wt % were obtained
in carbon dioxide-toluene mixtures in 24 hours. It is estimated that roughly 10-100
nanocrystals were infused per pore inside the mesoporous silica. The loading
approached equilibrium and the individual nanocrystals were highly dispersed
according to TEM. In contrast, the loadings were small with pure toluene or toluene
mixed with the antisolvent methanol. The differences in loading were correlated with
the long-ranged van der Waals forces between the gold and silica through the
intervening solvent, as described by the Hamaker constant, A132. As CO2 has an
unusually low polarizability/density (very weak VdW forces), it may be added to a
liquid solvent to form a mixture with an unusually low Hamaker constant. The
weaker vdW forces for the intervening solvent result in a stronger interaction between
the gold and silica and a higher chemical potential for the gold in the mixed solvent.
Both factors raise the nanocrystal adsorption on the surface and thus the equilibrium
loading. Furthermore, the loading may be tuned over a wide range by varying the CO2
composition. The difference in pore size had only a small effect on the loadings, even
when the aspect ratios reached 0.9, indicating pore blockage was minor. Thus it
59
appeared that the adsorbed nanocrystals were mobile, despite the proximity of the
nanocrystals to the surface even at the center of a pore.
The simplicity of this infusion method, in that well-known independent
syntheses may be utilized for both the nanocrystals and silica, allows for the facile
production of nanocrystal/silica composites for applications such as catalysis and
optoelectronics. The approach is general and may be applied to the wide varieties of
nanocrystals and mesoporous materials that are available or are being discovered.
2.6 References
(1) Murray, C. B.; Norris, D. J.; Bawendi, M. G. J. Am. Chem. Soc. 1993,
115, 8706-8715.
(2) Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R. J.
Chem. Soc., Chem. Commun. 1994, 7, 801-802.
(3) Shah, P. S.; Husain, S.; Johnston, K. P.; Korgel, B. A. J. Phys. Chem.
B 2001, 105, 9433-9440.
(4) Shah, P. S.; Husain, S.; Johnston, K. P.; Korgel, B. A. J. Phys. Chem.
B 2002, 106, 12178-12185.
(5) Skaff, H.; Lin, Y.; Dinsmore, A.; Russell, T.; Emrick, T. Abstracts of
Papers, 224th ACS National Meeting, Boston, MA, United States, August 18-22, 2002
2002, COLL-026.
(6) Weimann, T.; Geyer, W.; Hinze, P.; Stadler, V.; Eck, W.; Golzhauser,
A. Microelectronic Engineering 2001, 57-58, 903-907.
(7) Konya, Z. P., Victor F.; Kiricsi, Imre; Zhu, Ji; Alivisatos, Paul;
Somorjai, Gabor A. Catalysis Letters 2002, 81, 137-140.
(8) Wakayama, H.; Setoyama, N.; Fukushima, Y. Adv. Mater. 2003, 15,
742-745.
(9) Han, Y.-J.; Kim, J. M.; Stucky, G. D. Chem. Mater. 2000, 12, 2068-
2069.
60
(10) Fukuoka, A.; Araki, H.; Sakamoto, Y.; Sugimoto, N.; Tsukada, H.;
Kumai, Y.; Akimoto, Y.; Ichikawa, M. Nano Letters 2002, 2, 793-795.
(11) Coleman, N. R. B.; Ryan, K. M.; Spalding, T. R.; Holmes, J. D.;
Morris, M. A. Chem. Phys. Lett. 2001, 343, 1.
(12) Coleman, N. R. B.; O'Sullivan, N.; Ryan, K. M.; Crowley, T. A.;
Morris, M. A.; Spalding, T. R.; Steytler, D. C.; Holmes, J. D. J. Am. Chem. Soc.
2001, 123, 7010.
(13) Chen, W.; Cai, W.; Zhang, L.; Wang, G.; Zhang, L. J. Colloid and
Interface Sci. 2001, 238, 291-295.
(14) Ryan, K. M.; Erts, D.; Olin, H.; Morris, M. S.; Holmes, J. D. J. Am.
Chem. Soc. 2003, 125, 6284.
(15) Lake, L. W. Enhanced Oil Recovery; Prentice Hall, PTR, 1996.
(16) Hornyak, G.; Kroll, M.; Pugin, R.; Sawitowski, T.; Schmid, G.; Bovin,
J.-O.; Karsson, G.; Hofmeister, H.; Hopfe, S. Chemistry--A European Journal 1997,
3, 1951-1956.
(17) Hirai, T.; Okubo, H.; Komasawa, I. J. Phys. Chem. B 1999, 103, 4228-
4230.
(18) Hirai, T.; Okubo, H.; Komasawa, I. J. Colloid and Interface Sci. 2001,
235, 358-364.
(19) Konya, Z. P., Victor F.; Kiricsi, Imre; Zhu, Ji; Ager, Joel W., III; Ko,
Moon Kyu; Frei, Heinz; Alivisatos, Paul; Somorjai, Gabor A. Chem. Mater. 2003, 15,
1242-1248.
(20) Israelachvili, J. Intermolecular & Surface Forces; 2nd ed.; Academic
Press, Inc.: San Diego, 1992.
(21) Watkins, J. J.; Blackburn, J. M.; McCarthy, T. J. Chemistry of
Materials 1999, 11, 213-215.
(22) Wakayama, H.; Fukushima, Y. Ind. Eng. Chem. Res. 2000, 39, 4641-
4645.
61
(23) Fukushima, Y.; Wakayama, H. J. Phys. Chem. B 1999, 103, 3062-
3064.
(24) Wakayama, H.; Itahara, H.; Tatsuda, N.; Inagaki, S.; Fukushima, Y.
Chem. Mater. 2001, 13, 2392-2396.
(25) Coleman, N. R. B.; Morris, M. A.; Spalding, T. R.; Holmes, J. D. J.
Am. Chem. Soc. 2001, 123, 187-188.
(26) Ziegler, K. J.; Polyakov, B.; Kulkarni, J. S.; Crowley, T. A.; Ryan, K.
M.; Morris, M. A.; Erts, D.; Holmes, J. D. Journal of Materials Chemistry 2004, 14,
585-589.
(27) Van Bergen, F.; Gale, J.; Damen, K. J.; Wildenborg, A. F. B. Energy
(Amsterdam, Netherlands) 2004, 29, 1611-1621.
(28) Zhang, X.; Pham, J. Q.; Martinez, H. J.; Wolf, P. J.; Green, P. F.;
Johnston, K. P. Journal of Vacuum Science & Technology, B: Microelectronics and
Nanometer Structures--Processing, Measurement, and Phenomena 2003, 21, 2590-
2598.
(29) McClellan, A. K.; McHugh, M. A. Polymer Engineering and Science
1985, 25, 1088-1092.
(30) Gallagher, P. M., Coffey, M.P., Krukonis, V.J. and Klasuts, N. In
Supercritical Fluid Science and Technology; Johnston, K. P. a. P., J.M.L., Ed.;
American Chemical Society: Washington, D.C., 1989; Vol. 406, pp 334-354.
(31) Dixon, D. J.; Bodmeier, R. A.; Johnston, K. P. AIChE J. 1993, 39, 127.
(32) Randolph, T. W.; Randolph, A. D.; Mebes, M.; Yeung, S.
Biotechnology Progress 1993, 9, 429-435.
(33) Yeo, S.-D.; Debenedetti, P. G.; Radosz, M.; Schmidt, H.-W.
Macromolecules 1993, 26, 6207-6210.
(34) Dixon, D. J.; Johnston, K. P. AIChE Journal 1991, 37, 1441-1449.
(35) Maiwald, M.; Schneider, G. M. Berichte der Bunsen-Gesellschaft
1998, 102, 960-964.
(36) Kelley, S. P.; Lemert, R. M. AIChE Journal 1996, 42, 2047-2056.
62
(37) Wei, M.; Musie, G. T.; Busch, D. H.; Subramanian, B. J. Am. Chem.
Soc. 2002, 124, 2513-2517.
(38) Wei, M.; Musie, G. T.; Busch, D. H.; Subramanian, B. Green Chem.
2004, 6, 387-393.
(39) Xie, X.; Liotta, C. L.; Eckert, C. A. Ind. Eng. Chem. Res. 2004, 43,
7907-7911.
(40) Sellin, M. F.; Webb, P. B.; Cole-Hamilton, T. J. Chem. Commum.
2001, 781-782.
(41) McLeod, M. C.; Anand, M.; Kitchens, C. L.; Roberts, C. B. Nano
Letters 2005, 5, 461-465.
(42) Ng, H.-J.; Robinson, D. B. Journal of Chemical and Engineering Data
1978, 23, 325-327.
(43) Zhao, D.; Feng, J.; Huo, Q.; Melosh, N.; Frederickson, G. H.;
Chmelka, B. F.; Stucky, G. D. Science (Washington, D. C.) 1998, 279, 548-552.
(44) Tanev, P. T.; Pinnavaia, T. J. Chemistry of Materials 1996, 8, 2068-
2079.
(45) Barrett, E. P.; Joyner, L. G.; Halenda, P. P. Journal of the American
Chemical Society 1951, 73, 373-380.
(46) Saito, A.; Foley, H. C. AIChE Journal 1991, 37, 429-436.
(47) Horvath, G.; Kawazoe, K. J. Chem. Engg. Japan 1983, 16, 470-475.
(48) Lifshin, E. X-ray Characterization of Materials; Wiley-VCH: New
York, 1999.
(49) Korgel, B. A.; Fitzmaurice, D. Physical Review B: Condensed Matter
and Materials Physics 1999, 59, 14191-14201.
(50) Mattoussi, H.; Cumming, A. W.; Murray, C. B.; Bawendi, M. G.;
Ober, R. Physical Review B: Condensed Matter and Materials Physics 1998, 58,
7850-7863.
(51) Korgel, B. A.; Fullam, S.; Connolly, S.; Fitzmaurice, D. Journal of
Physical Chemistry B 1998, 102, 8379-8388.
63
(52) Guinier, A.; Fournet, G. Small-Angle Scattering of X-Rays; Wiley:
New-York, 1955.
(53) Glatter, O.; Kratky, O. Small Angle X-ray Scattering; Academic Press
Inc: New-York, 1982.
(54) Saunders, A. E.; Korgel, B. A. Journal of Physical Chemistry B 2004,
108, 16732-16738.
(55) Saunders, A. E.; Shah, P. S.; Park, E. J.; Lim, K. T.; Johnston, K. P.;
Korgel, B. A. J. Phys. Chem. B 2004, 108, 15969-15975.
(56) Lee, J.; Park, Y.; Kim, P.; Kim, H.; Yi, J. Journal of Materials
Chemistry 2004, 14, 1050-1056.
(57) Ehrburger-Dolle, F.; Morfin, I.; Geissler, E.; Bley, F.; Livet, F.; Vix-
Guterl, C.; Saadallah, S.; Parmentier, J.; Reda, M.; Patarin, J.; Iliescu, M.;
Werckmann, J. Langmuir 2003, 19, 4303-4308.
(58) Kresge, C. T.; Leonowickz, M. E.; Roth, W. J.; Beck, J. S. Nature
1992, 359, 710.
(59) Tanev, P. T.; Pinnavaia, T. J. Science (Washington, D. C.) 1995, 267,
865-867.
(60) Shah, P. S.; Holmes, J. D.; Doty, R. C.; Johnston, K. P.; Korgel, B. A.
J. Am. Chem. Soc. 2000, 122, 4245-4246.
(61) Shah, P. S.; Novick, B. J.; Hwang, H. S.; Lim, K. T.; Carbonell, R. G.;
Johnston, K. P.; Korgel, B. A. Nano Letters 2003, 3, 1671-1675.
(62) Konya, Z. P., Victor F.; Kiricsi, Imre; Zhu, Ji; Alivisatos, A. Paul;
Somorjai, Gabor A. Nano Letters 2002, 2, 907-910.
(63) Hiemenz, P. C. Principles of Colloid and Surface Chemistry.; Marcel
Dekker, Inc.: New York, 1986.
(64) Hourri, A.; St-Arnaud, J. M.; Bose, T. K. Journal of Chemical Physics
1997, 106, 1780-1785.
(65) Burns, R. C.; Graham, C.; Weller, A. R. M. Mol. Phys. 1986, 59, 41.
64
(66) Rubio, J. E. F.; Arsuaga, J. M.; Taravillo, M.; Baonza, V. G.; Caceres,
M. Experimental Thermal and Fluid Science 2004, 28, 887-891.
(67) Williams, J. W.; Krchma, I. J. J. Am. Chem. Soc. 1926, 48, 1888-1896.
(68) Lewis, J. E.; Biswas, R.; Robinson, A. G.; Maroncelli, M. Journal of
Physical Chemistry B 2001, 105, 3306-3318.
(69) Koone, N. D.; Zerda, T. W. Journal of Non-Crystalline solids 1995,
183, 243-251.
(70) Vigne-Maeder, F.; Amrani, S. E.; Gelin, P. J. Catal. 1992, 134, 536.
(71) Zhang, Z.; Lagally, M. G. Science 1997, 276, 377-383.
(72) Fukuoka, A.; Araki, H.; Kimura, J.-i.; Sakamoto, Y.; Higuchi, T.;
Sugimoto, N.; Inagaki, S.; Ichikawa, M. Journal of Materials Chemistry 2004, 14,
752-756.
65
Chapter 3
Infusion of Pre-synthesized Iridium Nanocrystals into Mesoporous
Silica for High Catalyst Activity
Traditionally, finely dispersed metal catalysts have been formed by reduction
of precursors within supports. A new concept for designing catalysts with enhanced
activities and selectivities is to infuse pre-synthesized nanocrystals, with well-defined
morphologies, into ordered mesoporous materials. The decoupling of nanocrystal
synthesis and infusion provides exquisite control of the nanocrystal size, morphology,
and dispersibility within the pores. A dispersion of iridium nanocrystals was infused
into mesoporous silica by expanding the solvent toluene with supercritical CO2. To
achieve high nanocrystal loadings, up to 1.3 wt %, the solvent quality was tuned to
strengthen the interactions of the nanocrystals with the pore walls, but without
precipitating the nanocrystals in the bulk solvent. Z-contrast STEM indicates
conclusively that the iridium nanocrystals were located within the pores and not on
the external silica surface. High catalytic activity was observed for 1-decene
hydrogenation, which is consistent with a high degree of dispersion of the 4.5 nm
nanocrystals throughout the pores, as observed by TEM. A maximum turnover
frequency (TOF) of 16 s-1 was measured, which was higher than the initial TOF for
homogeneous catalysis with the same nanocrystals in 1-decene. The iridium catalysts
do not require pre-treatment to remove the tetraoctylammonium bromide ligands to
achieve activation, as the ligands bind weakly to the iridium surface. Consequently,
the activity was not enhanced when calcined at 500oC in nitrogen.
Large portions of this chapter have been previously published as Gupta, G.; Stowell, C. A.; Patel, M. N.; Gao, X.; Yacaman, M. J.; Korgel, B. A.; Johnston, K. P. Chemistry of Materials 2006, 18, 6239-6249.
66
3.1 Introduction
In catalysis, substantial improvements in turnover rates and selectivities are
possible with metal nanocrystals containing high concentrations of active surface
sites.1,2 Nanocrystal growth may be controlled in solution with stabilizing ligands to
achieve specific size, crystallinity, shape, and surface properties.3 The design of the
surface sites on the nanocrystals may be guided by fundamental studies of reaction
mechanisms on well-defined surfaces on single crystals.4,5 Immobilization of the
nanocrystals within the pores of a mesoporous support helps maintain high surface
area by preventing coalescence of the nanocrystals during reaction.6 The size of the
mesopores may be designed to influence reaction selectivity and catalyst activity by
limiting access to the nanocrystals for certain reactants and products, as a function of
their size. Finally, the immobilization on the support facilitates catalyst recovery and
recycles.
Traditionally, supported catalysts are synthesized by precursor reduction,6-8
ion exchange9,10 and chemical vapor/fluid deposition.11-13 In these techniques, it is
challenging to control the metal cluster size and surface morphology. One approach
to metal nanocrystal encapsulation within a mesoporous support is to form the
mesoporous structure in the presence of a dispersion of pre-synthesized
nanocrystals.14,15 In this technique it can be difficult to achieve high loadings of
dispersed nanocrystals within the pores. In addition, the presence of the particles can
interfere with the synthesis of the desired pore structure and conversely, the synthesis
of the mesoporous material can change the morphology of the nanocrystals. For
example, the ligands stabilizing the nanocrystals may desorb and adsorb onto the
substrate during the reaction, leading to destabilization of the nanocrystals.
Furthermore, the surfactant templates used for the mesoporous material may interact
with the nanocrystal surfaces and modify their morphologies.
An alternative approach is to decouple nanocrystal synthesis from the
synthesis of the mesoporous material.16,17 An advantage of this approach is that a
67
large number of well-known synthetic methods may be utilized for the design of the
nanocrystals18 and of the mesoporous materials.19 Prior to infusion, the particles are
easily isolated from unwanted byproducts, cleaned, separated by size, and the ligands
may be exchanged if desired for alternative ligands. Here it is not necessary to
remove excess nanocrystal precursors and reaction by-products from the pores as in
traditional techniques such as wet impregnation.20 After synthesis and purification,
the nanocrystals may be infused into the pores of a pre-synthesized mesoporous
material by a physical technique without modifying the structure of either material.
Rioux et al16 infused platinum nanocrystals coated with polyvinylpyrrolidone
in an aqueous dispersion into high surface area mesoporous silica. Under sonication,
pore loadings reached approximately 1wt %. Without sonication, the nanocrystals
were not able to infuse into the pores and remained coated to the outside of the silica.
The nanocrystals penetrated the pores for a 1:1 mixture of water and ethanol, but not
with pure water. The observed difference in penetration of the pores was attributed to
the difference in surface tensions of water and ethanol.
In our previous study,17 gold nanocrystals coated with dodecanethiol were
infused into mesoporous silica to achieve highly dispersed nanocrystals with loadings
over 2 wt %. The infusion required the addition of supercritical CO2. CO2 has a low
polarizability relative to hydrocarbon solvents, and serves as an antisolvent for the
alkyl-coated nanocrystals.21-23 (However, nanocrystals dispersions have been obtained
in pure CO2 at high temperatures24 and in very dilute dispersions.25) For toluene
expanded with CO2, it is possible to tune the solvent strength to raise the long-ranged
van der Waals interactions of the nanocrystals with the pore walls, and thus the
nanocrystal loading, without flocculating the nanocrystals in the solution phase.17
This goal was not achieved by adding methanol as an antisolvent to toluene.
The exact location of nanocrystals within the mesopores versus on the
external surface is an important issue in metal/mesoporous silica catalysts. Rioux et
al.16 showed that nanoparticles, upon annealing at 400 °C, formed nanorods with a
conformation related to the geometry of the cylindrical mesopores. However, the
68
possibility of nanoparticles on the outside surface was not ruled out. To the best of
our knowledge, the location of pre-synthesized nanocrystals infused into mesoporous
materials has not been determined conclusively in previous studies. Midgley et al.26
showed that the exact location of nanocrystals within mesoporous silica can be
determined by Z-contrast scanning transmission electron microscopy (STEM) with a
high angle annular dark field (HAADF) detector. The nanocrystals were synthesized
inside the mesoporous silica by thermal reduction of a precursor. The use of a
HAADF detector has the advantage that the images do not suffer from diffraction
contrast. Furthermore, the contrast between heavy and light elements is enhanced as
the signal is proportional to the square of the atomic number.27
In a recent study of Pt nanocrystals infused into mesoporous silica,16 the as-
synthesized supported catalyst was inactive because the strongly bound polymeric
shell blocked the active sites on the metal surface.28 Calcination at 450 °C and above
for 12 hours was necessary to remove the polymeric coating and activate the catalyst.
High temperature annealing can potentially result in oxidation of the nanocrystals,
can significantly change the nanocrystal size and surface morphology, and can also
result in aggregation and coalescence, which reduces the available surface area of the
catalyst.29 The nature of ligand binding also has a large effect on homogeneous
catalysis with nanoparticle solutions. Stowell et al.30 showed that the catalytic activity
of Ir nanocrystals in solution for hydrogenation of 1-decene was influenced markedly
by the nature of the binding of the stabilizing ligands. Strong binding ligands
passivated the surface and inhibited catalytic activity, while nanocrystals with weakly
bound ligands showed much greater catalytic activity. However, as weakly bound
ligands desorbed from the nanocrystals surface into the solvent, the nanocrystals
aggregated and the activity declined.
Herein, we show that pre-synthesized iridium nanocrystals stabilized by
weakly bound tetraoctylammonium bromide (TOAB) ligands may be infused into
pre-synthesized mesoporous silica to produce a highly active catalyst for 1-decene
hydrogenation. Z-contrast STEM with an HAADF detector is used to discriminate
69
between nanocrystals within the mesopores and on the external silica surface. We
chose to study decene hydrogenation, given the previous data for homogeneous
catalysis with Ir nanocrystals in solution stabilized by TOAB.30 The activity is
compared for the original infused catalyst without pre-treatment and for catalysts in
which the ligands were removed by annealing at high temperature in air and nitrogen,
or by extraction with compressed carbon dioxide. The objective was to identify a
ligand with sufficiently weak binding to the surface, such that it becomes unnecessary
to pre-treat the catalyst to remove the ligand. In order to prevent aggregation of the
nanocrystals in solution during infusion, excess ligand was added to the solution. The
effect of CO2 concentration on the stability of the nanocrystal dispersion was
investigated to identify the optimum conditions for the infusion. The stability of the
catalyst was investigated by recovering the catalyst and performing a second reaction.
The activation energy of decene hydrogenation reaction was determined over a
temperature range of 40 °C to 75 °C. The catalytic activities of the Ir nanocrystals
infused into the mesoporous silica supports were compared with a commercial Pd
hydrogenation catalyst supported on alumina used under the same reaction
conditions.
3.2 Experimental
3.2.1 Materials
All chemicals were used as received. Tetraoctylammonium bromide or TOAB
(C32H68N Br, 98 %), decane (99 %), 1-decene (94 %) and 1,2 hexadecanediol (90%)
were purchased from Aldrich. (Methylcyclopentadienyl) (1,5-cyclooctadiene) iridium
( (C6H7)(C8H12)Ir, 99 %) was purchased from Strem. Toluene (99.9%) and
concentrated liquid HCl (normality 12.1) were obtained from Fisher Scientific.
Ethanol (Absolute 200 proof) was obtained from Aaper alcohol. Methanol (purity >
99.8%) was obtained from EM Sciences. Tetraethoxysilane (TEOS), and dioctylether
(97%) were obtained from Fluka Chemika and Poly(ethylene oxide)-b-
70
poly(propylene oxide)-b-poly(ethylene oxide) EO~20PO~70EO~20 (Pluronic P-123)
from BASF Corporation. CO2 (purity > 99.99%) (Matheson Gas Products) was used
as received. Polycarbonate filters (0.05 micron) were obtained from Osmonics. 5%
Palladium on alumina (Pd-Al2O3) commercial catalyst was obtained from Alfa Aesar.
Water was doubly distilled and deionized.
3.2.2 Iridium Nanocrystal Synthesis
Iridium nanocrystals with TOAB ligands were synthesized by a method
developed by Stowell et al.30 using a 25 ml- three neck, round bottom flask. 0.2 g of
hexadecanediol, 7 ml of dioctylether, 0.2 g of (methylcyclopentadienyl) (1,5-
cyclooctadiene) iridium, and 0.76 g of TOAB were measured into the flask. The
solution was freeze-pump-thawed for three cycles using liquid nitrogen, and heated
up to 270 °C for 30 minutes. The product was a black liquid and the particles were
isolated using one rinse in ethanol as an antisolvent. Rinsing the solution multiple
times in ethanol resulted in the removal of capping ligands from the iridium surface,30
so exposure to ethanol was kept at a minimum. The nanocrystals were finally
dispersed in toluene before the infusion procedure. Excess TOAB ligand (10 mg/ml)
was dissolved in the nanocrystal dispersion so as to prevent desorption of the ligands
and hence prevent aggregation of the particles.
3.2.3 Mesoporous Silica Synthesis
Mesoporous silica SBA-15 (referred to as MPS) was synthesized with a block
copolymer template similarly to the reported procedure of Zhao et al.31,32 A solution
of 1.6 g EO20PO70EO20 : 7 mL HCl: 8 mL TEOS: 38 mL H2O was prepared, stirred
for 24 h at 50°C and then heated at 100°C for 24 h. The resulting slightly yellow
powder was dispersed in ethanol, filtered and then again heated at 100 °C for 24 h.
Calcination was carried out by slowly increasing temperature from room temperature
to 500 °C in 4 h and heating at 500 °C for 6 h.
71
3.2.4 Infusion of Ir into Mesoporous Silica with CO2
In a typical experiment, 100 mg of mesoporous silica and 6 mL of iridium
nanocrystals dispersed in toluene, with a concentration of 0.4 mg/ml, were loaded
into a variable-volume reaction cell17 with a working volume (volume with piston in
the cell) of 27 mL. An amount of CO2 equivalent to 4 mL at 241 bar (3500 psia) and
35 °C was added so as to produce a total volume of reaction mixture of 10 mL. The
cell was then pressurized to 241 bar with CO2 at the back side of the piston using an
ISCO model 100 DX computer controlled syringe pump. The contents of the cell, in a
water bath, were stirred with a magnetic stir bar (Figure 3.1). The reaction was
maintained at 35 °C and 241 bar (3500 psia) and allowed to proceed for 24 h. After
the reaction, the cell was cooled to 0°C and CO2 was then vented as a vapor from the
top of the cell at the vapor pressure, 35 bar. The concentrations of organics in the CO2
vapor were small at 0 °C such that losses were minimized. The supernatant was
collected and filtered, to separate the Ir-MPS. Ir-MPS was cleaned by washing with 2
ml of pure toluene to remove the nanocrystals adsorbed on the outer surface of the
silica. Hence the resulting nanocomposite contained Ir nanocrystals immobilized
within the pores.
The extent of incorporation of iridium nanocrystals in the silica was
determined by subtracting from the initial mass of the dispersed nanocrystals in
toluene, the final mass of the nanocrystals recovered in the supernatant after
filtration.17 A correction was not made for the nanocrystals removed by washing with
pure toluene. The absorbance of the iridium nanocrystals dispersed in toluene before
and after infusion was measured at a wavelength of 500 nm using a Cary 500 UV-
Vis-NIR spectrophotometer with an optical path length of 1 cm. A standard
calibration curve was generated at 500 nm using known concentrations of iridium
nanocrystals in toluene as shown in Figure 3.2. The experiments were repeated
several times to check for the reproducibility in the nanocrystal loading, which was
on the order of +0.2 in wt % units, or 15 % of the final Ir mass. The different factors
72
Figure 3.1
Schematic of the infusion apparatus.
CO2 Cylinder ISCO Pump
Variable Volume View Cell PT
Stir Bar Piston
73
0
0.05
0.1
0.15
0.2
0.25
0.3
0 0.1 0.2 0.3 0.4 0.5
Abs
orba
nce
at 5
00 n
m
Concentration of Iridium (mg/ml)
Figure 3.2
Calibration curve of UV- Vis absorbance versus concentration for iridium nanocrystal
dispersions in toluene, measured at 500 nm.
74
which cause uncertainty in final iridium loadings are volume measurement (1%),
weight of silica (1%), uncertainty in absorbance (1%) and uncertainty in linearity (2-
3%) of calibration curve. The uncertainty of each of these individual factors results in
uncertainty on nanoparticle loading. Total uncertainty in loadings is reported as the
sum of the uncertainty due to each individual factor, which is approximately close to
10-15%. The uncertainty in this method of loading determination is comparable to
that of elemental analysis (10% error), which makes it equally effective for estimating
the metal loadings. The concentration of nanocrystals washed from the outer surface
of the silica with pure toluene was measured spectrophotometrically. The absorbance
measured was less than 0.01 and within the detection limit indicating a maximum of
0.1 mg of nanocrystals. The low concentration of nanocrystals is consistent with a
ratio of external to internal surface area of mesoporous silica on the order of 10-4
based on the pore diameter (~1 nm) and particle size (~10 μm). Hence if the loading
of nanocrystals is uniform on external and internal areas, then the loading on the
external surface would be negligible (~10-4%).
3.2.5 Determination of Nanocrystal Precipitation Boundary
The effect of CO2 concentration was determined on the stability of the
nanocrystal dispersions in the mixed solvent with toluene. Nanocrystal dispersions
were analyzed at 35 oC and 241 bar with a Cary 3E UV-Vis spectrophotometer in a
vertically mounted variable-volume high-pressure optical cell with a path length of 2
cm as described previously.33 The schematic is shown in the Figure 3.3. 6 mL of 0.4
mg/mL concentration of iridium nanocrystals in toluene with 10 mg/mL of excess
ligands in the solvent was loaded in the front part of the cell. The contents were
stirred using a magnetic stir bar. The computer controlled ISCO syringe pump was
used to introduce a known amount of CO2 in the front part of the cell, and the cell
was stirred for 30 s to mix the solvents. The stirring was stopped to allow the particles
75
Figure 3.3
Schematic of apparatus for determining the dispersibility of the iridium nanocrystals
in CO2-toluene solvent.
Light Source
UV-Vis Spectrophotometer
Detector
CO2 Syringe pump
Variable volume View Cell
76
to settle and the absorbance was measured for 20 min. Another aliquot of CO2 was
added and the procedure was repeated.
3.2.6 Catalysis
1-decene hydrogenation reactions were carried out in a 25 ml round bottom
flask batch reactor. Supported iridium nanocrystals were added to neat 1-decene to
create a 1500:1 decene to iridium mass ratio. Hydrogen gas at 5 psig was bubbled into
the dispersion. The system was stirred using a magnetic stir bar at high speed to raise
mass transfer rates. After the system was well saturated with hydrogen for 20 min. at
23 °C, the temperature was increased to the reaction temperature of 75 °C, and
hydrogen flow was maintained throughout the reaction. Aliquots of the solution (100
μl) were removed every hour for next four hours and stored at room temperature in
sealed vials for a maximum of 4 hours before analysis by gas chromatography.
3.3 Characterization
3.3.1 Nitrogen Porosimetry
N2 porosimetry was used to measure the surface area and the pore size
distribution of the mesoporous silica with a high speed surface area BET analyzer
(NOVA 2000, Quantachrome Instruments, Boynton Beach, FL) at a temperature of
77 K. Silica samples were pretreated at 100 ºC for 24 h in vacuum immediately prior
to data collection. Pore size distributions were analyzed by using the adsorption
branch of the isotherm. The Barrett- Joyner- Halenda (BJH) method34 was used for
pore size distribution.
3.3.2 Transmission Electron Microscopy
Low resolution pictures of mesoporous silica, infused with iridium
nanocrystals, were visualized by TEM on a Phillips EM280 microscope with a 4.5 Å
point-to-point resolution operated with an 80 kV accelerating voltage. High resolution
77
transmission electron microscopy (HRTEM) was performed using a JEOL 2010F
TEM operating at 200 kV. Z- Contrast STEM was performed on FEI TECNAI G2
F20 X-TWIN transmission electron microscope using a HAADF (High-angle annular
dark field) STEM detector with a 2 Å point-to-point resolution. Images were obtained
primarily with a GATAN digital photography system. Silica particles were deposited
from a dilute chloroform solution onto 200 mesh carbon-coated copper TEM grids
(Electron Microscopy Sciences).
3.3.3 Gas Chromatography/ Mass Spectroscopy
The extent of catalytic conversion of 1-decene was determined with a
Finnigan MAT GCQ gas chromatograph with mass spectrometer detector (GC/MS).
The column used was DB5-MS (5% phenyl-methylpolysiloxane, J & W scientific)
with a diameter of 0.32 mm and a length of 30 meter. The GC was used in positive
electron ionization mode with electrons accelerated at 70eV. 0.1μl of the undiluted
sample was injected using a split injection technique with an autosampler. The
retention time of the 1-decene and decane peaks was approximately 30 seconds. The
GCMS was calibrated to determine the relative amounts of 1-decene (MW = 140) and
decane (MW = 142) in each aliquot as shown in Figure 3.4. The curve was highly
linear with a correlation coefficient of 0.9995. 3.4 Results and Discussion
3.4.1 Nanoparticle Characterization
Figure 3.5(a) shows a representative TEM image of TOAB-stabilized Ir
nanocrystals. The particles were spherical in shape. By counting over 200 particles,
the average diameter was found to be 4.5 nm with a standard deviation of 0.6 nm (see
Figure 3.5(b)). Nanoparticles in the image appear slightly larger than the original due
to pixel aberration caused by image magnification.
78
0
0.1
0.2
0.3
0.4
0.5
0 0.2 0.4 0.6 0.8 1
Peak
Are
a (1
-dec
ene)
/ P
eak
Are
a (1
-dec
ene)
+ (d
ecan
e)
Volume % decane
Figure 3.4
GC-MS calibration curve of a mixture of 1-decene and 1-decane.
79
(a)
0
20
40
60
80
100
120
3 3.5 4 4.5 5 5.5
Num
ber
of P
artic
les
Nanoparticle diameter
(b)
Figure 3.5
(a) TEM micrograph of iridium nanocrystals stabilized by TOAB and (b)
corresponding histogram for iridium nanocrystals.
80
3.4.2 Characterization of Mesoporous Silica
Representative nitrogen adsorption/desorption isotherms and corresponding
pore size distributions (BJH method34) are shown in Figure 3.6. MPS has a mean pore
diameter of 6.5 nm, BET surface area of 742 m2/g and a pore volume of 0.98 cm3/g.
As well cited in literature for type SBA-15 mesoporous silica31 three distinct regions
are evident in the adsorption isotherm: (i) monolayer-multilayer adsorption, (ii)
capillary condensation and (iii) multilayer adsorption on the outer particle surfaces.
As can be seen in Figure 3.6, MPS shows a type-H1 hysteresis loop with capillary
condensation at an activity P/Po= 0.7.31 The step in the hysteresis loop is very steep
with a jump from 300 cm3/g to 500 cm3/g indicating capillary condensation. The
slope and the height of the step are clear indications of very well defined mesopores
with narrow pore size distribution.35 In our previous study17 we measured interpore
spacing of mesoporous silica type SBA-15 synthesized exactly as in the current study,
using small angle X-ray scattering (SAXS). SAXS showed multiple peaks indicating
a long range arrangement of pores with a pore diameter of 8.8 nm and a wall
thickness of 3.1 nm, typically associated with SBA-15 silica.31
3.4.3 Nanocrystal Dispersibility in Toluene-CO2 Mixtures
The nanocrystals are stabilized in the dispersion by steric repulsion from the
ligands adsorbed to the surface. The interactions between the ammonium bromide
head group and metal surfaces like iridium30 and gold36 have been found to be
relatively weak. For Ir, weak adsorption of TOAB ligands results in visible
aggregation of the nanocrystals over a period of 12 h. Ligand desorption can be
reduced by adding excess free ligand to the solvent. Mattoussi et al.37 examined this
concept quantitatively with SAXS studies on CdSe nanocrystals stabilized by
trioctylphosphine and trioctylphosphine oxide (TOP/TOPO) dispersed in hexane. In
the case of pure solvent with no excess ligands, the nanocrystals exhibited attractive
81
100
200
300
400
500
600
0 0.2 0.4 0.6 0.8 1
Vol
ume
(cc/
g)
Relative Pressure (P/Po)
(a)
0
0.005
0.01
0.015
0.02
0.025
0.03
0 50 100 150 200
Ads
orpt
ion
Dv(
d) (c
c/A
/g)
Pore diameter (A)
(b)
Figure 3.6
(a) N2 adsorption-desorption isotherms for mesoporous silica MPS and (b)
corresponding pore size distributions for MPS.
82
interactions. However, the addition of free ligands eliminated interparticle attractions
and the nanocrystal interactions eventually became repulsive. In this study, ~10
mg/mL of TOAB was added to the Ir nanocrystal dispersion in toluene with a
concentration of 0.5 mg/ml and nanocrystals were found to be stable for several days
on the basis of visual observation.
Figure 3.7 shows the absorbance at 500 nm of the Ir nanocrystal dispersion as
a function of solvent composition and time. As the total amount of nanocrystals
loaded in the variable-volume cell remains constant, the absorbance decreases as the
concentration of Ir is decreased by the addition of CO2 to the dispersion. It was
determined that the absorbance remained constant with time for less than 50% CO2
(w/w). When the CO2 concentration reached 50%, the absorbance began to decrease
modestly with time. At 55% CO2, absorbance decreased faster, indicating faster
aggregation and settling of particles. CO2 is a poor solvent for the hydrocarbon
ligands, and the insufficiently solvated ligands do not provide enough steric
stabilization to counter the attractive van der Waals forces between the cores. Since
large particles possess greater interparticle van der Waals attractions, they precipitate
first upon lowering the solvent quality.38 On the basis of these results, the loading of
the Ir particles for catalytic studies was performed at 40% CO2, where the absorption
was constant, indicating a stable dispersion.
3.4.4 Nanocrystal Loading in Pores using CO2
The Ir nanocrystals dispersed in toluene-CO2 solvent were infused into the
mesoporous silica over a 24 hour period. In our previous study,17 we showed that 24
hours was sufficient for reaching a maximum loading, with little increase from 24 to
72 hrs. We assumed 24 hours was also sufficient in the current study, since the metal
cores in each study had nearly the same Hamaker constants (gold vs. iridium) and
since both studies used hydrocarbon ligands. The loading of the iridium nanocrystals
within the mesoporous silica measured spectrophotometrically was about 1.3±0.2 wt
% (g Ir/g silica) for a CO2 concentration of 40%. For the nanocrystals dispersed in
83
Figure 3.7
Dispersibility of Ir-TOAB nanocrystals in CO2-toluene mixed solvent measured by
UV-Vis spectroscopy.
84
pure toluene without CO2, the loading was negligible. The high loading suggests that
the nanocrystals are mobile on the silica pore surface with the ability to diffuse deeply
into the pores without causing pore blockage.
The equilibrium adsorption of a nanocrystal onto the silica surface within a
pore is a function of the chemical potential of the nanocrystals in the solvent and the
long-ranged van der Waals forces between the nanocrystals and silica surface,
mediated by the intervening solvent.17 The van der Waals force of attraction between
the particle and the silica wall is directly proportional to the Hamaker constant of the
system, A132, for an iridium particle (1) and silica surface (2) across an intervening
medium (toluene or toluene-CO2 mixture, 3). It may be approximated by39
))(( 33223311132 AAAAA −−= (3.1)
where Aii is the Hamaker constant for pure i, which can be estimated using a
simplification of the Lifshift theory.39
Ayy =34
kbTεy −εvacuum
εy + εvacuum
⎛
⎝ ⎜ ⎜
⎞
⎠ ⎟ ⎟
2
+3hν e
16 2ny
2 − nvacuum2( )2
ny2 + nvacuum
2( )3 / 2 (3.2)
where eν is the maximum electronic ultraviolet adsorption frequency, typically
assumed to be 15103× s-1.23 The A for iridium, 19104 −× J,39 the ε and n for silica, 3.91
and 1.45, respectively40,41 and for toluene, 2.38 and 1.496, respectively42,43 were taken
from the literature. The Hamaker constants for silica and toluene calculated from
equation (3.2) are 201097.5 −× J and 201098.6 −× J, respectively.
For pure toluene without CO2, A132 is -0.31×10-20 J. The negative sign in the
overall Hamaker constant indicates that the interactions between the iridium
nanoparticle and the silica wall were repulsive or unfavorable for particle adsorption.
Hence, even though the nanocrystals can diffuse inside the pores along with the
solvent and fill the pores, the weak adsorption onto the silica surface may be expected
85
to result in low loadings as observed experimentally. If the dispersion of nanocrystals
in the pores contained the bulk concentration of nanocrystals, the calculated loading
would be less than 0.1 wt %, as the internal pore volume was only 1% of the entire
volume in the system.
CO2 has an unusually low polarizability/density (very weak vdW forces), and
may be added to a liquid solvent to form a mixture with an unusually low Hamaker
constant.17 The Hamaker constant for pure CO2 was calculated to be only 201044.1 −×
J.17 The Hamaker constant of the intervening mixed-solvent A33, was determined for
the binary solvent mixture with the Lorenz-Lorenz mixing rule as described
previously.17 The interactions between the nanocrystals and silica become strongly
attractive as the Hamaker force of the mixed solvent is reduced by adding CO2.17 For
40% CO2 in the intervening solvent, A132 was determined to be 4.1×10-20 J, indicating
strong attractive interactions of the Ir nanocrystals with the silica surface. The high Ir
loading of 1.3 wt.% seen experimentally were consistent with this prediction, as was
the low loading for pure toluene, where A132 was weakly repulsive.
The low Hamaker constant of the mixed solvent, and the weaker solvation of
the ligands also strengthens the interactions between the nanocrystals in solution, as
was seen in Figure 3.7. However, iridium nanocrystals form a stable dispersion with
40% CO2 in the solvent mixture. Thus, the solvent strength may be tuned with CO2 to
achieve strong adsorption on the silica surface, and thus high loading of nanocrystals,
without aggregating the nanocrystals in solution. The ability to fine tune the solvent
strength with CO2 to achieve high nanocrystal loadings has also been observed for Au
nanocrystals.17 Given these results for gold, it is to be expected that Ir could also be
infused into silica, given the similar Hamaker constants for Au and Ir, and the similar
hydrocarbon tails in the ligands used in both studies.
3.4.5 Visualization of Nanocrystals in Mesoporous Silica
High resolution Z-contrast STEM was utilized to confirm that most of the
nanocrystals are dispersed within the pores, rather than on the external surface of the
86
particles. The contrast variation in the STEM is proportional to the mass density of
the region multiplied by the atomic number squared (Z).27 Regions of high density are
shown in white. Figure 3.8(a) shows the planar view of the nanocrystals aligned
within the pores. Iridium has higher atomic number and mass density with respect to
the pore walls composed of silica, which in turn are denser than empty pores. Hence,
the iridium nanocrystals are observed as spherical bright dots, silica walls appear as
discrete white lines and empty pores appear as dark spaces between the silica walls.
Most of the nanocrystals are aligned within the empty pores along the pore wall while
a small fraction appears to be in the pore wall. The nanocrystals, which appear to be
on the pore walls, could have been in many cases located inside the secondary pores
(which act as interconnects between the pores) within the pore walls, which can be as
large as 4 nm in diameter for mesoporous silica of type SBA-15.44
Figure 3.8(b) is an out-of-focus image of 3.8(a) showing two different planes
of ordered pores (as indicated by arrows) with iridium nanocrystals inside the pores.
The planes were too close together to come from two different particles, but instead
indicated a fault line in the surfactant template for a single particle. Thus these
nanocrystals were in the interior of a particle in the pores and not on the surface,
further supporting the observations in Figure 3.8(a). Figure 3.8(c) shows the top down
view of the pores where we see the pore opening on the silica surface, with pores
arranged in hexagonal assembly. The arrows indicate the nanocrystals that are located
within the dark empty pores. Almost all the nanocrystals are dispersed within the
pores rather than on the external surface, consistent with the much higher internal
surface area as compared to the external area. The micrographs thus conclusively
prove that almost all the iridium nanocrystals are present within the pores.
Further evidence that the nanocrystals are mostly within the mesopores is
provided by the theoretical estimate of the close-packed monolayer coverage for
nanocrystals on the external silica surface. The particle diameter was measured to be
87
(a)
(b)
(c)
(d)
Figure 3.8
(a) Z contrast STEM (planar view) of MPS infused with Ir-TOAB; (b) Out of focus
view of (a); (c) top down view of (a); and (d) TEM micrograph of MPS infused with
Ir-TOAB nanocrystals without any pre-treatment. Scale bars in (a-c) are 10 nm.
88
greater than 10 micron by laser light scattering (Malvern Mastersizer-S, Malvern
Instruments Inc., Southborough, MA). The average molecular weight of 4.5 nm Ir
nanocrystals was approximated to be about 400,000 g/mol. The external surface area
of 10 μm nonporous silica spheres is 0.3 m2/g. From geometrical calculations, it was
found that in a hexagonal unit lattice 45% of the surface area was occupied by pores.
With this correction for 7 nm pores, the maximum external surface area of the porous
silica particles was determined to be 0.16 m2/g. In addition, the packing density of
0.906 (maximum density considering hexagonal close packing) was assumed for the
monolayer assembly, which arises because of the voids between the spherical
nanocrystals. The 1.3 wt % loading corresponded to 161096.1 × nanocrystals per gm
silica, while only 15103.4 × nanocrystals were required to form a close packed
monolayer on the external silica surface. The 1.3 wt % loading of nanocrystals was
therefore approximately 4.5 times that of monolayer coverage on the external surface.
Thus, even if external surface area was covered by a monolayer of iridium
nanocrystals, 78% of the nanoparticles would still be inside the pores, consistent with
the observations by Z-contrast microscopy. The high loading in the pores suggests
that the adsorbed particles are mobile and can move toward the center of the pore by
diffusion on the surface and/or in the bulk, without blocking the pores.
Figure 3.8(d) shows a typical HRTEM image (without Z contrast) of Ir
nanocrystals infused into mesoporous silica. The nanocrystals are highly dispersed,
although the resolution is considerably lower than in the Z contrast images in Figures
3.8 (a-c).
3.4.6 Decene Hydrogenation in Ir/MPS catalysts
The rates for 1-decene hydrogenation with uncalcined Ir/MPS are shown in
Figure 3.9. The conversion was linear in time. As the reaction was conducted in 1-
decene as a neat solvent, the reaction rate was zero order with respect to 1-decene.
Given that the solvent was the reactant, the mass transfer resistances for 1-decene to
89
0
100
200
300
400
500
600
0 1 2 3 4 5 6
Mol
dec
ene
reac
ted/
g c
atal
yst
Time (hr)
Ir-MPS (No pre-treatment)
Pd-Alumina
Figure 3.9
Comparison of catalyst activities for decene hydrogenation with Ir-MPS (without any
pretreatment) and an industrial catalyst Pd-Al2O3.
90
the surface were negligible. The surface of the catalyst was constantly saturated with
1-decene, eliminating any concentration dependence with time. The linearity also
suggested that the catalyst was not deactivated over the course of the experiment.
Hydrogen was continuously bubbled into the reaction mixture at 760 torr to maintain
a constant H2 concentration in the liquid phase. The rate of 1-decene hydrogenation is
a function of reaction temperature, concentration of hydrogen in the liquid phase and
the 1-decene concentration as follows
constCCfRTE
Ar Hdecenea
decene =⋅⎟⎠⎞
⎜⎝⎛ −
⋅=− ),(exp2
(3.3)
where A is the pre-exponential factor, Ea is the activation energy, Cdecene is the
concentration of neat 1-decene and CH2 is the concentration of dissolved hydrogen in
the liquid phase. In our case, all three factors remain constant; hence the rate of
reaction remains constant.
A turnover frequency (TOF) can be defined as the moles of reactant consumed
per unit time per mole of available active sites of the catalyst30
⎟⎠⎞
⎜⎝⎛
⎟⎟⎠
⎞⎜⎜⎝
⎛=
dtdm
NTOF decene
site
.1 (3.4)
where ( )site p ucN n A /A= , Ap is the surface area of the nanocrystal determined from
TEM, Auc is the area of one unit cell on the surface and n is the number of Ir atoms in
a unit cell face. The number of surface unit cells is simply Ap/Auc. The number of
active sites may only be approximated. Even if the moles of binding sites are
measured for a certain gas via chemisorption, the active sites for the catalytic reaction
may not be exactly the same sites.28 We chose to assume that the TOAB ligands did
not block any active sites as has been assumed in similar studies.28,45 It was assumed
that all of the surface Ir atoms were active, resulting a conservative lower limit for the
TOF via equation 3.4. The number of surface atoms was determined from the total
91
surface area of the nanocrystals, the number of iridium unit cells on the surface, a
lattice length of 0.384 nm, and the number of iridium atoms within the FCC unit cell.
3.4.7 Turnover Number for Uncalcined Ir-MPS
The TOF for uncalcined Ir-MPS in the 1-decene hydrogenation was 16 s-1.
The highly dispersed nanocrystals, with an average diameter of 4.5 nm and minimal
aggregation enabled the high catalyst activity. The use of weakly binding ligands for
nanoparticle stabilization in the bulk solvent, which readily desorb from the surface
of the Ir nanocrystals dispersed on the support, may be expected to produce a large
number of active sites and thus high catalyst activity.
For metal nanocrystals with very high activities, it can be challenging to
prevent poisoning of the surface during reaction resulting in reduced catalytic
activity or in some cases complete deactivation.28 Catalyst deactivation can be
particularly severe for concentrated reactants, in our case pure reactant.46 In order to
determine if Ir-MPS undergoes any surface poisoning during decene hydrogenation,
the Ir-MPS was recovered and used for a second catalytic reaction. Table 3.1
illustrates that the activity only dropped by 7% in the second reaction, within
experimental error, and remained much higher than the industrial catalyst in the first
reaction.
3.4.8 Effect of Catalyst Pretreatment on Catalytic Activity
Although, capping ligands are necessary for dispersing the nanocrystals
during infusion, they potentially may have a negative effect on the catalytic activity if
they block active sites on the Ir surface. Stowell et al.30 showed that iridium
nanocrystals stabilized with strongly bound ligands like oleic acid/oleylamine were
essentially catalytically inactive, whereas those stabilized with weakly bound ligands
like TOAB provided much higher catalytic activity. The TOF in decene
hydrogenation improved significantly with successive reactions, from 5 s-1 to 124 s-1,
92
Catalyst Temp
(oC)
TOF
(mol decene
reacted/ mol
surface atoms
of catalyst /sec)
Activity
(mol of decene
reacted / g of
metal /sec× 104)
Ir-MPS (uncalcined) 75 16.3 1.66
Ir-MPS (2nd run
uncalcined) 75 15.1 1.54
Ir-MPS (calcined in
inert atmosphere,
nitrogen at 500C for 6
hrs)
75 15.4 1.59
Ir-MPS (annealed in
CO2 at 4000psi, 60C
for 12 hrs)
75 17.1 1.75
Ir-MPS (calcined at
250C for 6 hrs in air) 75 11 1.12
Ir-MPS(calcined at
500C for 6 hrs in air) 75 4 0.41
Ir-MPS (uncalcined) 50 5.7 0.58
Ir-MPS (uncalcined) 40 2.4 0.25
5% Pd on Alumina
(Industrial Catalyst) 75 - 0.92
Table 3.1
Activity and turnover frequency for decene hydrogenation of 1.3% Ir supported on
mesoporous silica relative to 5 % Pd on alumina.
93
as the weakly bound ligands desorbed from the surface. However, the nanocrystals
eventually aggregated from too much ligand desorption resulting in a loss in surface-
to-volume ratio and activity.30 This aggregation may be mitigated by supporting the
nanocrystals within mesoporous silica.
Table 3.1 compares the TOFs, with 760 torr hydrogen pressure at 75 °C unless
otherwise specified, for decene hydrogenation for Ir-MPS catalysts pre-treated in
various ways. The TOF obtained with the untreated Ir-MPS was 16±1 s-1. The TOF
was essentially the same when the catalyst was calcined in an inert nitrogen
atmosphere at 500 °C for 6 hours. This temperature was sufficient to desorb and
volatilize the TOAB ligands. The nanocrystals were immobilized on the support, as
they did not show any deactivation from sintering upon calcination at 500 oC. The
catalysts were also calcined in air at two different temperatures, 250 ˚C and 500˚C.
The TOF decreased from 16.5 to 11 s-1 when calcined at 250˚C, and to 4, less than a
quarter of the original activity, at 500 oC. Given that the TOF did not change after
calcination in N2 at 500 oC, the loss in activity upon calcination in air may be
attributed to oxidation of the surface of the Ir nanocrystals. Annealing in supercritical
CO2 at 275 bar pressure and 60 °C for 12 hours produced a slight increase in the
catalytic activity. The solubility of the ionic molecule TOAB in supercritical CO2
may be expected to be low and hence 0.1 gm of 1.3% Ir-MPS was annealed in
approximately 25 gm of supercritical CO2.
In the case of supported nanocrystals on mesoporous silica, we observe that
high catalytic activity may be obtained without thermal pre-treatment to remove the
ligands from the surface. Furthermore, the high activity remains constant even after
pretreatment by a) thermal calcination in N2 and b) CO2 extraction. Thus any ligands
on the surface prior to these pretreatments did not inhibit catalyst activity, as was the
case for polymeric stabilizers on Pt nanocrystals.28 Further evidence will be provided
to explain this highly beneficial result.
The TOAB ligands will partition between the bulk phase, the nanocrystals and
the mesoporous silica during the infusion process and during the catalytic reaction.
94
The mesoporous silica with a surface area of 742 m2/g has a large number of sites for
ligand adsorption. The total surface area of the adsorbed nanocrystals was only 3.5
m2. Suppose that all of the TOAB adsorbed exclusively on the silica. The area for a
linear dodecanethiol ligand has been reported to be approximately 20 A2 per
molecule.47 If it is assumed that the area of TOAB (highly branched) is twice that of
dodecanethiol, we estimate that the total amount of TOAB required for monolayer
coverage on the mesoporous silica is about 1.7 g TOAB/ g silica. Consequently, even
if all the TOAB in the system (60 mg added) adsorbed exclusively on the mesoporous
surface, only 35% of the silica sites would be covered. Thus, it is likely that a large
fraction of the weakly bound ligands on the dilute nanocrystals within the pores will
migrate to the vacant sites on the silica and reduce the coverage of ligands on the
surface. Ligand desorption from the nanocrystals will increase the number of
catalytically active sites.
TOAB’s weak binding and highly branched architecture will leave more
catalytic sites available relative to ligands such as oleic acid and oleylamine. The
volume occupied by the small ammonium bromide head groups will be relatively
small for TOAB relative to the larger volume of the four hydrocarbon tails. In
essence, the bulky tails will leave gaps between the head groups containing active
catalytic sites for the small hydrogen molecules. Consequently, H2 atoms may still
chemisorb on the surface and react very close to the binding site of the TOAB
underneath the larger cone of the tail groups. The transfer of ligands from the
nanocrystals to the silica surface and the gaps between the headgroups of the ligands
is likely to play an important role in the high activity of the catalysts without pre-
treatment.
3.4.9 Stability of Supported Nanocrystals During the Reaction
It was determined that the metal nanocrystals do not desorb from the
mesoporous silica upon exposure to the 1-decene during 4 hours of reaction at 75 °C.
After the reaction, the mixture was filtered and the supernatant was recovered and
95
analyzed spectrophotometrically at 500 nm. The absorbance was below the detectable
limit within the experimental uncertainty (<0.005). From an extrapolation of the
calibration curve (Figure 3.10), the concentration of the desorbed iridium
nanocrystals was determined to be less than 0.01 mg/ml, or less than 3% of the total
amount of catalyst. The stability of the catalyst against nanocrystal desorption is
further supported by the fact that the TOF for the second catalytic run was almost the
same as that of the original run, as shown in Table 3.1. Figure 3.11 shows the TEM
micrograph of Ir-MPS after reaction of 1-decene with hydrogen for 4 hours. The
micrograph shows that the particles were well dispersed and did not aggregate,
further indicating the stability of the nanocrystals within mesoporous silica. The
expected desorption of the ligand from the adsorbed nanoparticle surface onto silica
would increase the attractive van der Waals attraction between the particle and the
silica, further enhancing its binding to the surface.
3.4.10 Activation Energy of the Decene Hydrogenation
The apparent activation energy for the decene hydrogenation by Ir-MPS was
measured over a temperature range of 40-75 ˚C as shown in Table 3.1. As shown in
the Arrhenius plot for the TOF16 (Figure 3.12), the activation energy was
approximately 13.5 kcal/mol. To further put our results in perspective, the catalytic
activity of Ir-MPS was compared with a typical industrial hydrogenation catalyst (5%
palladium on Alumina). Decene hydrogenation reactions were performed under
identical conditions with the same mass of metal as shown in Figure 3.9. The rate of
decene hydrogenation is constant for Pd-alumina as was observed above for Ir-MPS
for the same reasons. For Ir-MPS, the activity was 117 mole decene reacted/g Ir/hr
versus only 65 for Pd. For the Pd catalyzed reaction of 1-decene, the activation
energy was 17 kcal/mol.48 The higher activity for Ir-MPS vs. Pd-alumina, by a factor
of 1.8, is caused in part by the high surface area on the order of 80 nm2 for 5 nm in
diameter Ir nanocrystals. The greater fraction of surface atoms of metal, as
characterized by the higher surface area, increases the number of active sites.
96
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.11
0.12
0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
Abs
orba
nce
at 5
00 n
m
Conc of Ir in decene (mg/ml)
Figure 3.10
Calibration curve of UV- Vis absorbance versus concentration for iridium nanocrystal
dispersions in 1-decene, measured at 500 nm.
98
0.5
1
1.5
2
2.5
3
2.9 2.95 3 3.05 3.1 3.15 3.2
ln (T
OF)
1/T * 1000 (Kelvin-1)
Figure 3.12
Arrhenius plot of decene hydrogenation over untreated Ir-MPS.
99
3.5 Conclusions
Herein, we show a novel approach for designing catalysts with high activities
by infusing pre-synthesized iridium nanocrystals, with well-defined surface
morphologies, into ordered mesoporous silica. Decoupling the synthesis step and the
infusion step enables control of the iridium nanocrystal size, morphology, catalytic
properties and dispersibility within the mesopores. A dispersion of iridium
nanocrystals may be infused into mesoporous silica by expanding the solvent,
toluene, with supercritical CO2. High loadings of the nanocrystals are achieved by
tuning the solvent strength with CO2, in order to strengthen the interactions of the
nanocrystals with the pore walls, without flocculating the nanocrystals in the bulk
solvent. With Z-contrast STEM using a HAADF detector, it was demonstrated clearly
that the iridium nanocrystals are highly dispersed within the pores, with relatively few
particles on the external particle surface. The ability to maintain the pre-designed
nanocrystal size during infusion and to achieve high loadings within the pores, as
characterized by Z-contrast TEM, are important milestones in the development of
highly active and selective catalysts. The high surface area of the nanocrystals
produced high turnover frequencies (TOF) of 16 s-1 for decene hydrogenation, even
without thermal pre-treatment to remove the TOAB ligands. The TOF was higher
than the initial TOF for homogeneous catalysis30 with the same nanocrystals
dissolved in 1-decene. Thermal calcination at 500 °C or CO2 annealing at 275 bars
had no impact on catalyst activity, which is not surprising given the high activity of
the original catalyst.
The nanocrystals remain dispersed in the pores and do not aggregate or desorb
into the solvent during the reaction, as observed by post reaction TEM and UV-Vis
spectrophotometric analysis of the product solution. Weakly bound ligands on the
nanocrystal surface, such as TOAB, may be expected to migrate to the vacant sites on
the silica, significantly reducing the ligand coverage on the surface and hence
increasing the catalyst activity. Additionally, tetraoctylammonium bromide ligands
100
with bulky tails will leave gaps between the head groups allowing small hydrogen
molecules to chemisorb on the surface and react very close to the binding site of the
TOAB. Desorption of ligands from the adsorbed nanoparticle surface, onto the
support and into the decene during reaction, will enhance metal binding to the surface
and aid catalyst stability. This general infusion method for controlling the size,
surface morphology and stabilizing ligands of the nanocrystals, independently of the
synthetic procedure for the mesoporous material, offers broad opportunities in
designing novel highly active catalysts.
3.6 References
(1) Scott, R. W. J.; Wilson, O. M.; Crooks, R. M. Synthesis,
Characterization, and Applications of Dendrimer-Encapsulated Nanoparticles.
Journal of Physical Chemistry B 2005, 109, 692-704.
(2) Rolison, D. R. Catalytic nanoarchitectures - The importance of nothing
and the unimportance of periodicity. Science (Washington, DC, United States) 2003,
299, 1698-1702.
(3) Kanaras, A. G.; Soennichsen, C.; Liu, H.; Alivisatos, A. P. Controlled
Synthesis of Hyperbranched Inorganic Nanocrystals with Rich Three-Dimensional
Structures. Nano Letters 2005, 5, 2164-2167.
(4) Farias, M. H.; Gellman, A. J.; Somorjai, G. A.; Chianelli, R. R.; Liang,
K. S. The coadsorption and reactions of sulfur, hydrogen and oxygen on clean and
sulfided molybdenum(100) and on molybdenum disulfide(0001) crystal faces.
Surface Science 1984, 140, 181-196.
(5) Strongin, D. R.; Carrazza, J.; Bare, S. R.; Somorjai, G. A. The
importance of C7 sites and surface roughness in the ammonia synthesis reaction over
iron. Journal of Catalysis 1987, 103, 213-215.
101
(6) Wakayama, H.; Setoyama, N.; Fukushima, Y. Size-Controlled
Synthesis and Catalytic Performance of Pt Nanoparticles in Micro- and Mesoporous
Silica Prepared Using Supercritical Solvents. Adv. Mater. 2003, 15, 742-745.
(7) Lensveld, D. J.; Mesu, J. G.; van Dillen, A. J.; de Jong, K. P. The
application of well-dispersed nickel nanoparticles inside the mesopores of MCM-41
by use of a nickel citrate chelate as precursor. Studies in Surface Science and
Catalysis 2002, 143, 647-657.
(8) Zong, Y.; Watkins, J. J. Deposition of copper by the H2-assisted
reduction of Cu(tmod)2 in supercritical carbon dioxide: Kinetics and reaction
mechanism. Chemistry of Materials 2005, 17, 560-565.
(9) Singh, U. K.; Albert Vannice, M. Liquid-Phase Hydrogenation of
Citral over Pt/SiO2 Catalysts. Journal of Catalysis 2000, 191, 165-180.
(10) Haran, N. P.; Bhagade, S. S.; Nageshwar, G. D. Catalysis by ion
exchange resins - alkylation. Chemicals & Petro-Chemicals Journal 1978, 9, 21-28.
(11) Hayek, K.; Goller, H.; Penner, S.; Rupprechter, G.; Zimmermann, C.
Regular Alumina-Supported Nanoparticles of Iridium, Rhodium and Platinum Under
Hydrogen Reduction: Structure, Morphology and Activity in the Neopentane
Conversion. Catalysis Letters 2004, 92, 1-9.
(12) Serp, P.; Hierso, J. C.; Feurer, R.; Kihn, Y.; Kalck, P.; Faria, J. L.;
Aksoylu, A. E.; Pacheco, A. M. T.; Figueiredo, J. L. Single-step preparation of
activated carbon supported platinum catalysts by fluidized bed organometallic
chemical vapor deposition. Carbon 1999, 37, 527-530.
(13) Chu, H. P.; Lei, L.; Hu, X.; Yue, P.-L. Metallo-Organic Chemical
Vapor Deposition (MOCVD) for the Development of Heterogeneous Catalysts.
Energy & Fuels 1998, 12, 1108-1113.
(14) Konya, Z.; Puntes, V. F.; Kiricsi, I.; Zhu, J.; Alivisatos, P.; Somorjai,
G. A. Novel Two-Step Synthesis of Controlled Size and Shape Platinum
Nanoparticles Encapsulated in Mesoporous Silica. Catalysis Letters 2002, 81, 137-
140.
102
(15) Konya, Z.; Puntes, V. F.; Kiricsi, I.; Zhu, J.; Ager, J. W., III; Ko, M.
K.; Frei, H.; Alivisatos, P.; Somorjai, G. A. Synthetic Insertion of Gold Nanoparticles
into Mesoporous Silica. Chemistry of Materials 2003, 15, 1242-1248.
(16) Rioux, R. M.; Song, H.; Hoefelmeyer, J. D.; Yang, P.; Somorjai, G. A.
High-Surface-Area Catalyst Design: Synthesis, Characterization, and Reaction
Studies of Platinum Nanoparticles in Mesoporous SBA-15 Silica. Journal of Physical
Chemistry B 2005, 109, 2192-2202.
(17) Gupta, G.; Shah, P. S.; Zhang, X.; Saunders, A. E.; Korgel, B. A.;
Johnston, K. P. Enhanced Infusion of Gold Nanocrystals into Mesoporous Silica with
Supercritical Carbon Dioxide. Chemistry of Materials 2005, 17, 6728-6738.
(18) Yin, Y.; Alivisatos, A. P. Colloidal nanocrystal synthesis and the
organic-inorganic interface. Nature (London, United Kingdom) 2005, 437, 664-670.
(19) de Soler-Illia, G. J.; Sanchez, C.; Lebeau, B.; Patarin, J. Chemical
Strategies To Design Textured Materials: from Microporous and Mesoporous Oxides
to Nanonetworks and Hierarchical Structures. Chemical Reviews (Washington, DC,
United States) 2002, 102, 4093-4138.
(20) Shyue, J.-J.; De Guire, M. R. Single-Step Preparation of Mesoporous,
Anatase-Based Titanium-Vanadium Oxide and Its Application. Journal of the
American Chemical Society 2005, 127, 12736-12742.
(21) Shah, P. S.; Holmes, J. D.; Doty, R. C.; Johnston, K. P.; Korgel, B. A.
Steric Stabilization of Nanocrystals in Supercritical CO2 Using Fluorinated Ligands.
J. Am. Chem. Soc. 2000, 122, 4245-4246.
(22) Shah, P. S.; Husain, S.; Johnston, K. P.; Korgel, B. A. Nanocrystal
Arrested Precipitation in Supercritical Carbon Dioxide. J. Phys. Chem. B 2001, 105,
9433-9440.
(23) Shah, P. S.; Husain, S.; Johnston, K. P.; Korgel, B. A. Role of Steric
Stabilization on the Arrested Growth of Silver Nanocrystals in Supercritical Carbon
Dioxide. J. Phys. Chem. B 2002, 106, 12178-12185.
103
(24) Lu, X.; Ziegler, K. J.; Ghezelbash, A.; Johnston, K. P.; Korgel, B. A.
Synthesis of Germanium Nanocrystals in High Temperature Supercritical Fluid
Solvents. Nano Letters 2004, 4, 969-974.
(25) Bell, P. W.; Anand, M.; Fan, X.; Enick, R. M.; Roberts, C. B. Stable
Dispersions of Silver Nanoparticles in Carbon Dioxide with Fluorine-Free Ligands.
Langmuir 2005, 21, 11608-11613.
(26) Midgley, P. A.; Weyland, M.; Thomas, J. M.; Johnson, B. F. G. Z-
Contrast tomography: a technique in three-dimensional nanostructural analysis based
on Rutherford scattering. Chemical Communications (Cambridge, United Kingdom)
2001, 907-908.
(27) Midgley, P. A.; Weyland, M. 3D electron microscopy in the physical
sciences: the development of Z-contrast and EFTEM tomography. Ultramicroscopy
2003, 96, 413-431.
(28) Yoo, J. W.; Hathcock, D. J.; El-Sayed, M. A. Propene hydrogenation
over truncated octahedral Pt nanoparticles supported on alumina. Journal of Catalysis
2003, 214, 1-7.
(29) Scott, R. W. J.; Wilson, O. M.; Crooks, R. M. Titania-Supported Au
and Pd Composites Synthesized from Dendrimer-Encapsulated Metal Nanoparticle
Precursors. Chemistry of Materials 2004, 16, 5682-5688.
(30) Stowell, C. A.; Korgel, B. A. Iridium Nanocrystal Synthesis and
Surface Coating-Dependent Catalytic Activity. Nano Letters 2005, 5, 1203-1207.
(31) Zhao, D.; Feng, J.; Huo, Q.; Melosh, N.; Frederickson, G. H.;
Chmelka, B. F.; Stucky, G. D. Triblock copolymer syntheses of mesoporous silica
with periodic 50 to 300 angstrom pores. Science (Washington, D. C.) 1998, 279, 548-
552.
(32) Zhao, D.; Huo, Q.; Feng, J.; Chmelka, B. F.; Stucky, G. D. Nonionic
triblock and star diblock copolymer and oligomeric surfactant syntheses of highly
ordered, hydrothermally stable, mesoporous silica structures. Journal of the American
Chemical Society 1998, 120, 6024-6036.
104
(33) Ryoo, W.; Webber, S. E.; Johnston, K. P. Water-in-Carbon Dioxide
Microemulsions with Methylated Branched Hydrocarbon Surfactants. Industrial &
Engineering Chemistry Research 2003, 42, 6348-6358.
(34) Barrett, E. P.; Joyner, L. G.; Halenda, P. P. The determination of pore
volume and area distributions in porous substances. I. Computations from nitrogen
isotherms. Journal of the American Chemical Society 1951, 73, 373-380.
(35) Tanev, P. T.; Pinnavaia, T. J. Mesoporous Silica Molecular Sieves
Prepared by Ionic and Neutral Surfactant Templating: A Comparison of Physical
Properties. Chemistry of Materials 1996, 8, 2068-2079.
(36) Saunders, A. E.; Sigman, M. B., Jr.; Korgel, B. A. Growth Kinetics
and Metastability of Monodisperse Tetraoctylammonium Bromide Capped Gold
Nanocrystals. Journal of Physical Chemistry B 2004, 108, 193-199.
(37) Mattoussi, H.; Cumming, A. W.; Murray, C. B.; Bawendi, M. G.;
Ober, R. Properties of CdSe nanocrystal dispersions in the dilute regime: Structure
and interparticle interactions. Physical Review B: Condensed Matter and Materials
Physics 1998, 58, 7850-7863.
(38) McLeod, M. C.; Anand, M.; Kitchens, C. L.; Roberts, C. B. Precise
and Rapid Size Selection and Targeted Deposition of Nanoparticle Populations Using
CO2 Gas Expanded Liquids. Nano Letters 2005, 5, 461-465.
(39) Israelachvili, J. Intermolecular & Surface Forces; 2nd ed.; Academic
Press, Inc.: San Diego, 1992.
(40) Hourri, A.; St-Arnaud, J. M.; Bose, T. K. Dielectric and pressure virial
coefficients of imperfect gases: CO2-SF6 mixtures. Journal of Chemical Physics
1997, 106, 1780-1785.
(41) Burns, R. C.; Graham, C.; Weller, A. R. M. Mol. Phys. 1986, 59, 41.
(42) Rubio, J. E. F.; Arsuaga, J. M.; Taravillo, M.; Baonza, V. G.; Caceres,
M. Refractive index of benzene and methyl derivatives: temperature and wavelength
dependencies. Experimental Thermal and Fluid Science 2004, 28, 887-891.
105
(43) Williams, J. W.; Krchma, I. J. Dielectric constants of binary mixtures.
J. Am. Chem. Soc. 1926, 48, 1888-1896.
(44) Joo, S. H.; Choi, S. J.; Oh, I.; Kwak, J.; Liu, Z.; Terasaki, O.; Ryoo, R.
Ordered nanoporous arrays of carbon supporting high dispersions of platinum
nanoparticles. [Erratum to document cited in CA135:184349]. Nature (London,
United Kingdom) 2001, 414, 470.
(45) Le Bars, J.; Specht, U.; Bradley, J. S.; Blackmond, D. G. A Catalytic
Probe of the Surface of Colloidal Palladium Particles Using Heck Coupling
Reactions. Langmuir 1999, 15, 7621-7625.
(46) Ha, S.; Larsen, R.; Masel, R. I. Performance characterization of Pd/C
nanocatalyst for direct formic acid fuel cells. Journal of Power Sources 2005, 144,
28-34.
(47) Korgel, B. A.; Fullam, S.; Connolly, S.; Fitzmaurice, D. Assembly and
Self-Organization of Silver Nanocrystal Superlattices: Ordered \"Soft Spheres\".
Journal of Physical Chemistry B 1998, 102, 8379-8388.
(48) Goyal, H. B.; Gupta, A. K. Kinetics of liquid-phase hydrogenation of
C10-C16 olefins on a palladium-alumina catalyst. Indian Journal of Technology
1989, 27, 360-362.
106
Chapter 4
Stable Ordered FePt-Mesoporous Silica Catalysts with High
Loadings
A new concept is presented to form catalysts by infusion of pre-synthesized
bimetallic nanocrystals into ordered mesoporous supports. For pre-synthesized FePt
nanocrystals (<4 nm), coated with oleic acid and oleylamine ligands in toluene, high
loadings above 10 wt % were achieved in 10 minutes. The strong metal-support
interactions were favored by the low coverage of the weakly bound ligands. The
nanocrystals were highly dispersed within the pores as indicated by HAADF-STEM
and X-ray diffraction (XRD) and stable against sintering at 700 °C and desorption
into polar and nonpolar solvents at room temperature. A phase transformation from a
disordered phase (FCC) to ordered phase (FCT) was observed upon thermal
annealing at 700 °C without sintering, as confirmed by convergent beam electron
diffraction and XRD. The calcined FePt catalyst exhibited six fold higher catalyst
activity (TOF =30 s-1) than that of a commercial Pd-alumina catalyst for liquid 1-
decene hydrogenation and was stable for multiple reactions. The decoupling of
nanocrystal synthesis and infusion provides exquisite control of the nanocrystal size,
alloy structure, binding to the support and dispersibility within the pores, offering
broad opportunities for enhanced catalyst activities, selectivities and stabilities.
Large portions of this chapter have been previously published as Gupta, G.; Patel, M. N.; Ferrer, D.; Heitsch, A. T.; Korgel, B. A.; Johnston, K. P., Chem. Mater., 2008, ASAP.
107
4.1 Introduction
Highly active and stable catalysts may be designed by controlling the size and
composition of supported metal nanoparticles.1-8 For supported catalysts synthesized
by precursor reduction,1,9 chemical vapor/fluid deposition10 or coprecipitation,5,11 it is
difficult to control the size and surface morphology of the metal clusters as well as
the binding to the surface. An emerging concept in catalyst design is to pre-synthesize
metal nanocrystals coated with stabilizing ligands to control the morphology and then
to infuse the particles onto high surface area ordered mesoporous supports.3,12,13 The
highly uniform and tunable size and curvature of 1-D and 3-D pores in ordered
mesoporous substrates14 facilitates dispersion and stability of the nanoparticles.15-
17,18,19
The metal loading of catalysts formed with pre-synthesized nanocrystals has
been found to be highly dependent on the stabilizing ligands on the particle surface.
Infusion of Pt nanocrystals coated with polyvinylpyrollidone (PVP) ligands into
mesoporous silica with sonication yielded a metal loading of approximately 1 wt %.3
Gold12 nanocrystals capped with dodecanethiol and iridium20 with
tetraoctylammonium bromide were loaded into mesoporous silica up to 2.5 wt %.
The low molecular weight ligands facilitated ligand removal. The high catalytic
activity for 1-decene hydrogenation without calcination suggests that the weakly
bound ligand tetraoctylammonium bromide did not cover the active sites. Recently,
Jung et al21 showed that FePt nanocrystals could be infused in hydrophobic
mesoporous silica, although the metal loading was not given. An alternative approach
is to synthesize the mesoporous support in the presence of a nanocrystal
dispersion.22,23 However, the presence of the nanocrystals can interfere with the
mesopore structure and conversely, the synthesis of the mesoporous material can alter
the morphology and surface composition of the nanocrystals. To date, high metal
loadings >10 wt % have received little attention for the infusion technique, but would
be of interest in various applications.6,24
108
Multimetallic alloy nanocrystals often enable enhanced catalyst activities as a
consequence of synergistic interactions between atoms25,26 including formation of
ordered intermetallic nanocrystals.6,26,27 The high annealing temperatures required to
form these intermetallic alloys often leads to particle sintering.28-30 Traditionally,
bimetallic nanocrystals have been synthesized by reduction of multiple precursors
over a metal support28 or by decomposition of multimetal carbonyl precursors31.
Thomas and coworkers decomposed specially synthesized multi-metal carbonyl
precursors thermally over mesoporous silica to form nanoclusters25,32 composed of
10-20 atoms with a size of ~1 nm. Whereas these catalysts have great potential, larger
metal nanocrystals ≥3 nm are in some cases more catalytically active than small metal
clusters.33
The reduction of multiple metal precursors within porous supports often
produces, high polydispersity and high variability in particle composition.34 Without
the use of stabilizing ligands, the nanocrystal size may exceed 10 nm and even reach
the size of the pores of the support, as has been observed for FePt,30 NiFe28 and
PtCo35 in mesoporous silica. For example, FePt nanoparticles grew to a size of 6 nm
in mesoporous silica with 9 nm pores. Here the nanoparticles may block some of the
pores. The varying rates of nucleation and growth for each metal precursor over the
heterogeneous sites on the supports may lead to highly variable composition.34 For
example, the composition varied for Pt:Co between 44:56 and 97:3, and the size
reached 6-10 nm.35 Potentially, these limitations may be overcome by infusion of pre-
synthesized bimetallic nanocrystals into the porous support, although this approach
has received little attention. For example, FePt nanocrystals have been synthesized
with precise control over the nanocrystal size, shape and composition by arrested
growth precipitation.2,36,37
The primary objectives of this study were to: (1) design active and stable
catalysts at high loadings above 10 wt % by infusion of pre-synthesized FePt
nanocrystals on mesoporous silica and (2) to transform the crystal structure to a face
centered tetragonal (FCT) intermetallic alloy upon annealing at 700 oC without
109
perturbing the size. These objectives were facilitated by designing stabilizing ligands
to strengthen short-ranged interactions between bare Pt and silica, which have been
characterized by XPS,38 EELS39 and wettability studies40 at high temperatures.
Because various types of ligands such as dodecanethiol12 and polyvinylpyrrolidone3
screen these interactions, we chose to investigate weakly bound oleic acid and
oleylamine low molecular weight ligands that expose a greater fraction of the metal
surface domains. The large interfacial area between the spherical nanocrystals and
concave mesopores with similar diameters favors nanocrystal-support interactions,
relative to other support geometries.
The size and composition of the nanocrystals was determined by X-ray
diffraction and high angle annular dark field-scanning transmission electron
microscopy (HAADF-STEM). The kinetics and equilibrium loading of nanocrystals
in mesoporous silica were studied as a function of the initial nanocrystal
concentration over a period from 10 minutes to 16 hours. The nanocrystals within the
pores were differentiated from those on the external silica surface by HAADF-STEM.
Finally, the catalyst activity and stability for liquid phase hydrogenation of 1-decene
were examined both before and after calcination.
4.2 Experimental Section
4.2.1 FePt Nanocrystal Synthesis
3–4 nm spherical FePt nanocrystals were synthesized with the high
temperature arrested precipitation method developed be Sun et al.36 The reaction was
carried out under a N2 environment with a Schlenk Line. The nanocrystals were
prepared by mixing 0.098 g (0.25 mmol) platinum (II) acetylacetonate, 0.195 g (0.77
mmol) 1,2–hexadecanediol, and 10 mL of dioctyl ether in a 50 mL three-neck flask.
The stirred solution was heated to 100 °C under N2 flow, and 0.065 mL (0.50 mmol)
iron (0) pentacarbonyl, 0.085 mL (0.26 mmol) oleylamine, and 0.08 mL (0.25 mmol)
oleic acid were injected. The temperature was raised at a rate of 20 °C/min to the
110
reflux temperature of dioctyl ether (297 °C) and the reaction solution was incubated
for 30 minutes. The solution was allowed to cool to room temperature and excess
ethanol was used to precipitate the FePt nanocrystals which were recovered after
centrifugation (10 min, 8000 rpm). The nanocrystals were washed twice by
redispersing them in hexane and repeating the precipitation procedure. No additional
size selective precipitation was necessary to obtain the monodispersity.
4.2.2 Mesoporous Silica Synthesis
Mesoporous silica SBA-15 was synthesized with a block copolymer template
with the exact procedure of Zhao et al.18,41 A solution of 1.6 g EO20PO70EO20:7 mL
HCl:8 mL TEOS:38 mL H2O was prepared, stirred for 24 h at 50 °C and then heated
at 100 °C for 24 h. The resulting slightly yellow powder was dispersed in ethanol,
filtered and then again heated at 100 °C for 24 h. Calcination was carried out by
slowly increasing the temperature from room temperature to 550 °C in 4 h and
heating at 550 °C for 6 h.
4.2.3 Infusion of FePt into Mesoporous Silica
To prepare the supported nanocrystal composites, 10 mg of mesoporous silica
was mixed with 3 mL of FePt nanocrystal dispersion in toluene of known
concentrations from 0.25 to 2 mg/mL. The contents were then stirred for ~10 min.
After stirring, the mesoporous silica infused with FePt nanocrystals was separated
from the nanocrystal supernatant by filtration with a 0.2 μm PTFE filter. The extent
of incorporation of FePt nanocrystals in the silica was determined by subtracting from
the initial mass of the dispersed nanocrystals in toluene, the final mass of the
nanocrystals recovered in the supernatant after filtration.12 The absorbance of the FePt
nanocrystals dispersed in toluene before and after infusion was measured at a
wavelength of 500 nm using a Cary 500 UV-Vis-NIR spectrophotometer with an
optical path length of 1 cm. A standard calibration curve was generated with known
111
concentrations of FePt nanocrystals in toluene as shown in Appendix A, Figure A.6.
The experiments were repeated three times to check for the reproducibility in the
nanocrystal loading, which was on the order of ± 0.2 in wt % unit. Total uncertainty
in loadings determined with this technique was approximately 12% as explained in
detail previously.20 The uncertainty for this method is comparable to that for
elemental analysis (10% error). The stability of the nanocrystals immobilized within
the pores was studied by suspending the nanocomposite in pure ethanol. The
suspension was sonicated for 2 h after which the nanocomposite was separated by
vacuum filtration, redispersed in ethanol and analyzed by using UV-Vis spectroscopy.
4.2.4 Calcination
To modify the crystal structure of FePt, nanocomposites were calcined at 700
°C in N2/H2 atmosphere for 4 h. Temperature was ramped from ambient to 700 °C in
20 min. To activate the FePt for catalysis, nanocomposites were calcined in stagnant
air at 500 °C for 4 h. Temperature was ramped from ambient to 500 °C in 15 min.
The nanocomposites were washed for 5 min in ethanol.
4.2.5 Transmission Electron Microscopy
Microscopy was performed on a FEI TECNAI G2 F20 X-TWIN TEM using a
high-angle annular dark field detector (HAADF). HAADF images were obtained with
a 2 Å resolution. High resolution transmission electron microscopy (HRTEM) was
performed using a JEOL 2010F TEM operating at 200 kV. Pictures were obtained at
the optimum defocus condition. Nanocomposites were deposited from a dilute
ethanol solution onto 200 mesh carbon-coated copper TEM grids.
4.2.6 X-ray Diffraction
Wide angle X-ray diffraction was performed with samples on a quartz slide
using a Bruker-Nonius D8 Advance diffractometer. Samples were typically scanned
for 12 h at a scan rate of 12 deg/min with 0.02 deg increments. The average
112
nanocrystal size was estimated from the Scherrer equation by fitting the XRD line
shapes using JADE software (by Molecular Diffraction Inc).
4.2.7 Catalysis
1-decene hydrogenation reactions were carried out in a 25 ml round bottom
flask batch reactor as reported previously.20 Supported FePt nanocrystals were added
to neat 1-decene to create a 2000:1 decene to iridium mass ratio. Reaction conversion
was identified using GC-MS.
4.3. Results
4.3.1 Structure of Nanocomposites after Infusion of FePt Nanocrystals
Figure 4.1 shows a representative high resolution TEM of FePt of the original
nanocrystals prior to infusion, along with a histogram of the particle number
distribution. The particles are spherical in shape and crystalline as can be seen from
the lattice fringes in HRTEM. The lattice spacing is estimated to be 2.24 Ǻ, which
matches to the (111) plane as reported by Sun.36 Upon counting 250 particles, the
average diameter was found to be 3.6 ± 0.5 nm as shown in the histogram. The
mesoporous silica as synthesized is of type SBA-15 with hexagonally ordered 1D
cylindrical mesopores. The mesoporous silica had a surface area of 540 m2/g, a mean
pore diameter of 5.5 nm and a wall thickness of 4.3 nm (Appendix A, Figure A.1 and
A.2 in supplementary information).
The pre-synthesized FePt nanocrystals were infused into the mesoporous
silica for 10 min. Figure 4.2 shows the loadings reached 14 wt % as a function of
initial nanocrystal concentration in toluene. The loading markedly exceeds previous
values which was a maximum of 2.5 wt % using scCO2 as an anti-solvent.12 The
loadings may be tuned linearly as a function of the initial FePt concentration. At low
initial concentrations (~0.25 mg/ml), nearly all of the FePt nanocrystals in the bulk
113
(a)
0
20
40
60
80
100
2 2.5 3 3.5 4 4.5 5 5.5 6
Num
ber
of P
artic
les
FePt Diameter (nm)
(b)
Figure 4.1
(a) HRTEM images of pre-synthesized FePt nanocrystals prior to infusion. (b)
Histogram of nanocrystal size. Average diameter is 3.6 ± 0.5 nm. 250 nanocrystals
were counted for the histogram.
114
6
8
10
12
14
16
0 0.5 1 1.5 2 2.5
FePt
Loa
ding
(wt%
)
Initial Concentration FePt (mg/mL)
Figure 4.2
Loading of FePt nanocrystals within mesoporous silica in 10 minutes as a function of
initial concentration of nanocrystals.
115
were adsorbed within the mesoporous silica, leaving a clear supernatant. At a higher
initial concentration of 2 mg/ml, the final concentration of nanocrystals after infusion
was only ~0.37 mg/ml in the supernatant, indicating over ¾ of the original
nanocrystals were adsorbed. The loadings for FePt nanocrystals were measured as a
function of time to determine the kinetics of nanocrystal infusion (Appendix A,
Table A.1). For an initial concentration of 0.7 mg/ml, loadings increased from 8.9 wt
% to 11.4 wt % when the time of infusion was increased from 10 min to 16 h,
suggesting an approach to equilibrium. For higher initial concentrations of 2 mg/ml,
the loadings increased slightly from 13.8 wt % to 14.3 wt % over this time interval,
which is within the experimental reproducibility. Thus, the level of loading
approached a near equilibrium value on the order of minutes.
High resolution HAADF-STEM was used to visualize the FePt nanocrystals
supported on the mesoporous silica prior to thermal annealing. The contrast variation
is proportional to the mass density of the region multiplied by the atomic number Z
squared.32 Regions of high atomic number appear in white. The image of the system
may be understood as a convolution of the probe, the SiO2 support and the metal
nanocrystals.
I (total) = Iprobe θ IMetal θ ISiO2 ………………………(4.1)
where the symbol θ represents the convolution operation. The interpretation of the
image can be complicated. However the SiO2 porous structure is periodic and
relatively straightforward to understand and allows the interpretation of the HAADF
images. On the other hand, the particles are located in both black and white fringes,
with somewhat similar sizes in each domain. Figure 4.3(a) shows a planar view of the
FePt nanocrystals aligned within the pores at 7.3 wt % metal loading after infusion.
FePt nanocrystals are observed as spherical bright dots, silica walls appear as white
lines and empty pores appear as dark spaces between the silica walls, as expected due
to differences in atomic number. The nanocrystals are highly dispersed without any
significant aggregation. The FePt nanocrystals associated more with the dark line
regions are more likely to be in the pores and are estimated to be 65% of the
116
(a) (b)
0
10
20
30
40
50
2 2.5 3 3.5 4 4.5 5 5.5
Num
ber
of P
artic
les
FePt nanocrystal size (nm)
(c)
Figure 4.3
TEM images of 7.3 wt % FePt nanocrystals after infusion into mesoporous silica.
(a) HAADF STEM planar view; (b) HRTEM; (c) Histogram of nanocrystal size using
(a) for 100 nanocrystals.
20 nm
117
total infused nanocrystals, while those in the white regions can be either on the
surface or in the pore just above or below the pore wall or in the secondary
interconnecting pores.19 The secondary pores of size up to 3.6 nm are present in SBA-
15 mesoporous silica as reported by Joo and Ryoo. Most of the nanocrystals are
aligned with the pores along the pore wall and show a definite orientation parallel to
the pores, suggesting that they are in the pores, and not the external surface. In
addition, many of the particles in the white domains are also in the pores. The FePt
nanocrystals in the tail end of the distribution with 5 nm metal cores (and 1.8 nm
longs ligands on the metal surface) will not be able to enter the pores with a diameter
of 5.5 nm. These larger nanocrystals may only be adsorbed on the external surface
and appear in the white regions of the pore walls. As a result, the average size of
some of the nanocrystals that appear on the white lines or pore walls is larger (>4 nm)
than that of the nanocrystals in the dark spaces or pores (~ 3.5 nm).
The average interparticle distance between the nanocrystals in Figure 4.3(a) is
between 5-10 nm. Assuming a uniform distribution of nanocrystals for a 7.3 wt %
loading in the cylindrical mesopores, the average interparticle distance was estimated
to be 35 nm. The STEM micrograph is necessarily obtained near an edge of a particle
where the particle is not too thick. Thus, this difference in interparticle distance
suggests that nanocrystals are more concentrated closer to the pore openings than in
the center of the pore, as they have to diffuse longer distances to reach the center of
the pore. The bright field high resolution image of the FePt nanocrystals in Figure
4.3(b) indicates that the nanocrystals retain their size and crystallinity as confirmed
by the presence of lattice fringes. The lattice spacing of these supported FePt was
estimated to be 2.28 Ǻ, which is comparable to the value of 2.24 Ǻ (for (111) plane)
estimated for unsupported nanocrystals.
A calculation of the surface coverage of the nanocrystals may be used to
provide further evidence that the nanocrystals are in the pores and not merely
adsorbed on the external surface. The external surface area of mesoporous silica is
118
extremely small, ~0.3 m2/g based on a particle size estimate of 10 μm. For an
area/particle of 101057.1 −× m2 on the surface, a closely packed monolayer of FePt
nanocrystals would only produce a loading of <0.5 wt %. The observed loading of 10
wt % would require at least 20 layers of nanocrystals, if all the particles were located
on the external surface. As can be seen in Figure 4.4(a), the thickness of the
nanocrystal layer on the external surface is only a few layers thick, far below 20.
Thus, the STEM–HAADF images and this calculation suggest that most of the
particles were adsorbed in the mesopores and not on the surface.
4.3.2 Stability of the FePt Nanocomposites
The stability of the supported FePt nanocrystals towards desorption in polar
and nonpolar solvents were studied in ethanol and toluene separately. 20 mg of 7.3 wt
% FePt in mesoporous silica was dispersed in 5 mL of solvent and sonicated for 2 h at
ambient temperature. Using UV-Vis spectroscopy, the absorbance was less than 0.01.
Thus, a maximum of less than 2.5 % of the total infused nanocrystals for both
solvents could have desorbed, a very minor amount, within the error limit of the
instrument.
The stability of supported FePt nanocrystals towards sintering and
coalescence at high temperature was characterized by annealing the nanocomposite at
700°C as shown in the STEM-HAADF images in Figure 4.4(a). Most of the
nanocrystals are well dispersed without any noticeable aggregation. From the TEM
image in Fig 4.4(a), pore collapse or structural damage is not present for the
mesoporous silica thermally treated at 700 °C. The size of the larger nanocrystals in
the distribution is still within the range of the as synthesized nanocrystal dispersion
shown in Figure 4.3(c). The size distribution of annealed supported nanocrystals as
seen in the histogram (Figure 4.4(b)) was very similar to that of as synthesized
nanocrystals prior to infusion. However, there are a few regions in the STEM image
which show sintering of nanocrystals, as examined in greater detail in the
119
(a) (b)
(c) (d)
(d) (e)
Figure 4.4
7.3 wt % FePt nanocrystals in mesoporous silica calcined at 700 °C for 4 h: (a)
HAADF- STEM; (b) Histogram of nanocrystals after calcination determined from (a).
116 nanocrystals were counted. Average size is 3.7 ± 0.7 nm.; (c) Filtered image of
random portion of (a) to eliminate shot noise; (d) Intensity profile of FePt marked (A)
in image (c); (e) Intensity profile of FePt marked (B) in image (c).
Inte
nsity
(arb
itrar
y u
Length (arbitrary units)
Inte
nsity
(arb
itrar
y u
Length (arbitrary units)
Inte
nsity
(arb
itrar
y un
its)
Length (arbitrary units)
Inte
nsity
(arb
itrar
y un
its)
Length (arbitrary units)
0
10
20
30
40
50
2 3 4 5 6
Num
ber o
f Par
ticle
s
FePt nanocrystal size (nm)
120
supplementary information (Appendix A, Figure A.3). These regions are likely on the
external surface of the silica and not within the pores. The STEM image in Figure
4.4(a) was filtered to eliminate shot noise, as shown in Fig 4.4(c). Due to the fact that
most of the particles are of the same composition, the STEM-HAADF profile
represents the profile of the size and shape of the nanocrystals. The intensity profile
of larger particles in region A (Figure 4.4(d)) can be compared with the smaller
particles in region B (Figure 4.4(e)). The particles in A are more intense and are less
spherical than the particles in B. The size of the particles in A approaches the pore
size of 5 nm. Therefore, it is reasonable to assume that the particles in A wetted the
external surface of the silica during thermal annealing. Figure 4.4(d) shows two small
nanocrystals near each other. In contrast with the large type A particles, the size of
the type B nanocrystals is ~3 nm, well below the pore size. If the B particles were on
the surface of the silica, it is likely that the particle would sinter at the high annealing
temperature. Thus, it is likely that the B particle was in a pore. From a statistical
analysis of the number of particles of type A and B in a sample of 106 nanocrystals,
85 correspond to type B, based on a size cutoff of 4.5 nm. With the STEM analysis of
the samples before and after calcination, along with the calculation of the particle
surface coverage, it is evident a large fraction (at least 80%, a conservative estimate)
of the particles was in the pores. Furthermore, nanoparticles on the external surface
will sinter on annealing as shown in the supplementary figure (Appendix A, Fig A.3)
and on the top edge of the Fig 4.4(a). The small degree of sintering shown in STEM
images and in the XRD results indicate that a large fraction of the particles were
protected in the pores.
4.3.3 Crystal Structure of the Bimetallic FePt after Annealing
To facilitate the ordering of atoms in the supported FePt nanocrystals, the
samples were annealed at 700 °C. The change in crystal structure of FePt nanocrystals
on thermal annealing was observed with wide-angle XRD and convergent beam
electron diffraction (CBED) studies. Figure 4.5(a) gives the XRD
121
0
100
200
300
400
500
600
10 20 30 40 50 60 70 80 90
2 theta
Inte
nsity
Uncalcined FePt on MPS
calcined FePt MPS 700C
(a)
(b) Figure 4.5
(a) XRD diffraction of FePt nanocrystals in silica before and after annealing at 700
°C. The (111) peak to the right shift after annealing supports the phase transition from
FCC phase of uncalcined nanocrystals to FCT (L10). (b) convergent beam electron
diffraction of FePt nanocrystals in silica before and after annealing also confirms
phase transformation from FCC to FCT phase.
220
111 200
220
111
200
Amorphous Silica
311
311
122
patterns of infused FePt nanocrystals in mesoporous silica, before and after thermal
annealing. In the as-synthesized uncalcined nanocomposite, peaks are observed at
2θ= 40.2°, 46.7° and 67.6° and 83° are assignable to the (111), (200), (220) and (311)
orientations for FePt. A broad peak is found at 2θ= 22.5°, which can be assigned to
amorphous silica. The FePt peaks can be indexed to a FCC phase, which is reported
to be a disordered phase.36 From the XRD spectra according to the Scherrer equation,
the nanocrystal size is estimated to be 4.0 nm and is in close agreement with the FePt
particle size before infusion measured using TEM. The small nanoparticle sizes lead
to weaker and broader peaks than for larger particles.
The XRD spectra for the annealed FePt nanocomposite indicates a (111) peak
shift by 0.5°.36 The full width half maxima of (111) peak of annealed FePt
nanocomposite does not show any significant narrowing, as compared to the as-
synthesized sample, indicating coalescence of nanocrystals was insignificant,
consistent with the STEM results. The nanocrystal size after sintering from peak
fitting was ~4.1 nm, close to uncalcined nanocrystal size. Hence, the FePt
nanocrystals underwent ordering to the FCT phase upon thermal annealing without
any appreciable sintering. The (110) peak for FCT phase is very broad and weak but
we used another technique of convergent beam electron diffraction to further study
the phase transformation.
The technique of convergent beam electron diffraction (CBED) was employed
to confirm the structural changes of nanocrystals loaded in the silica. This is the an
accurate method to determine three-dimensional information about the reciprocal
lattice point and space group symmetry details, lattice parameters from fine regions
and atomic positions within a unit cell.42,43 Electron diffraction patterns recorded from
a single supported FePt nanocrystal are shown in Figure 4.5b. The patterns are
indexed for the single nanocrystal before and after thermal treatment at 700 °C, to
identify the change in crystal structure. CBED of a single uncalcined FePt nanocrystal
supported on mesoporous silica confirms the FCC phase of the nanocrystals
123
undergoes a phase transformation to FCT phase upon thermal annealing (Figure
4.5(b)).
4.3.4 Activity and Stability of Catalysts for 1-decene Hydrogenation
The turnover frequency (TOF) for the FePt supported on mesoporous silica,
calculated from moles of 1-decene reacted per second per mole of active sites on the
nanocrystal surface, as described earlier20 is shown in Figure 4.6. The blank
mesoporous silica did not have any catalytic effect for decene hydrogenation.
Conversion for 1-decene by blank silica was <1% after 4 h (wt of decene: wt of silica
= 10:1), and is within the error limits of the GC-MS. The number of active sites for
FePt was calculated for a particle size of 3.6 nm, as determined from TEM in Figure
4.3(a). To ensure complete ligand removal, the catalyst was calcined at 500 °C. The
temperature was much higher (~100 °C ) than where the ligands are removed, as
determined by TGA.44 The reaction was carried out in neat 1-decene. It was zero
order with respect to decene and with a linear conversion in time. A large increase in
catalyst activity for fresh FePt catalyst was observed after thermal oxidation. In order
to test the stability of the FePt catalyst, it was recovered and used for two more
successive reactions. The TOF doubled in the second hydrogenation reaction to a
high level of 30 s-1, which is twice that observed for the 4.5 nm iridium catalyst,20 and
did not show any significant increase for the third reaction. The corresponding
specific activity was 8.5×10-4 moles/g-s, which was 6 times higher than the industrial
catalyst 5 wt % Pd-alumina; however, the metals were different in each case. As the
size of the Pd nanoparticles is not known, TOF for Pd-alumina cannot be calculated.
The uncalcined as synthesized FePt nanocomposite exhibited significant catalyst
activity (~33% relative to a 5% Pd-alumina industrial catalyst). Thus, the passivating
ligands did not block the entire metal surface. The reproducibility in the TOFs was
124
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
5% FePtcalcined -1st time(Fresh)
5% FePtcalcined-2nd time
(Used once)
5% FePtcalcined-3rd time(Usedtwice)
5% FePtuncalcined
(Fresh)
5% FePtuncalcined-2nd time
Tur
n ov
er fr
eque
ncy
(Sec
-1)
Figure 4.6
Turn over frequency for 1-decene hydrogenation by 5 wt % FePt in mesoporous
silica.
125
determined by synthesizing new batch of 5% FePt-silica catalyst and then repeating
the reactions (fresh, washed once and washed 2nd time). The reproducibility for all
cases was within 5% of the 1-decene conversion and is within the detection error of
GC-MS. During infusion, the weakly bound ligands may desorb and adsorb on the
silica with a much larger surface area.20 The catalyst activity doubled relative to the
first reaction when the uncalcined FePt nanocomposites were used for a second
reaction. The performance of the FePt catalyst after first cycle was three times better
than the commercial catalyst (5 % Pt on alumina). The significant increase in the
activity of the calcined catalyst for the second run may likely be attributed to the
removal of carbonaceous contaminants that block active sites from the metal surface.
The lack of change during the third run suggested that these contaminants were
completely removed.
4.4. Discussion
4.4.1 Mechanism for Metal Loading and Stability
The inherent stability of the nanocrystals against desorption into polar
solvents at low temperature or sintering at 700 oC may be attributed to two key
factors, electronic and geometrical effects. Zheng and Stucky have reported that gold
nanocrystals stabilized with thiol ligands adsorbed on non-porous substrates, such as
silica and titania, adsorb weakly and desorb readily into a polar solvent such as
ethanol.7
Metal-support interactions play a key role in both the activity and the thermal
stability of the nanocrystal catalysts. Yu et al. studied the thermal stability of Pt
nanocrystals supported on silica thin films using in-situ TEM.40 The particles began
to change shape at 350 oC, and spread on the surface above 500 oC. The wetting of
the surface by the nanocrystals is attributed to the interfacial mixing of Pt and silica,
as a result of strong metal-support interactions. Klie et al. characterized Pt-silica
specific short range interactions using electron energy loss spectroscopy.39 Strong
silica charge transfer interactions with Pt were identified using X-ray photoelectron
126
spectroscopy (XPS).38 Similar strong interactions between Pt atoms on the surface of
FePt nanocrystals and silica may be expected to aid both the infusion process and the
stabilization of the particles upon thermal annealing.
Long-ranged electrostatic and van der Waals interactions also contribute to the
interaction between a nanoparticle and the support. FePt nanocrystals did not did not
migrate under a field of 166 V/cm, indicating minimal charging. Silica in nonpolar
media acquires negative charges45 but the surface charge density is very small, with
each charge separated by ~100 nm. Metal nanoparticles are extremely polarizable
and form a large induced dipole in presence of a charge.46 The attractive charge-
induced dipole interactions between charged silica and a single metal nanocrystal
(Appendix A, Figure A.4) are proportional to the cube of the nanocrystal radius,
whereas the long-ranged van der Waals interactions were found to be weakly
repulsive in toluene. The total long-ranged interaction energy (charge-induced dipole
+ van der Waals) is only weakly attractive for the 3.6 nm FePt nanocrystals
(Appendix A, Figure A.4). However, with the sparse charge densities for silica (20-
200 charges/μm2) the total loading would be only 0.5 wt % if one nanocrystal
interacted with each charge on the silica. Consistent with this prediction, we have
shown that 2 nm gold nanocrystals stabilized by dodecanethiol adsorb weakly on
silica with small loadings of <0.3 wt %.20 Therefore, the high levels of adsorption for
both the FePt and Pt nanocrystals (up to 20%) in this study must be driven primarily
by the short-ranged interactions described above.38-40
The hydrocarbon capping ligands on the metal surface must not screen the
strong short-ranged interactions with the silica. Amines and fatty acids are known to
bind to Au more weakly relative to thiols.47 For example, surface coverage of
dodecylamine on Pt is 46%48 compared to 75% for dodecanethiol on Ag.49 The lower
surface coverage exposes barer metal surface for the formation of specific short-
ranged interactions with the silica. These short-ranged interactions could explain the
loadings of ~20 wt %, relative to the much lower loadings for gold capped with
dodecanethiol, where the surface coverage is higher. Furthermore, more of the weakly
127
bound amine ligands will migrate from the nanocrystals to the silica surface and
expose even more surface area than in the case of the strongly bound thiols. The large
thermodynamic driving force for adsorption in the FePt case led to near equilibrium
loadings in only 10 minutes versus 24 h for the dodecanethiol capped gold.20 The
short-ranged specific interactions were sufficient to prevent sintering from
mobilization of nanocrystals at high temperatures up to 700 °C and from desorption
into both polar (ethanol) and nonpolar (toluene) solvents upon sonication. The
controlled particle size and composition were favored by the metal-support
interactions ease of ligand removal at moderate temperatures (<400 °C) and the
removal of byproducts of nanocrystal synthesis prior to infusion. The convex surfaces
on the nanocrystals favored contact with the cylindrical mesopores with concave
curvature.50 In contrast, the surfaces of both the metal nanocrystal and support are
often both convex for traditional supports leading to a smaller degree of interfacial
contact. The beneficial geometrical effect of enhanced interfacial area on
nanocomposite stability has been reported by Rolison and co-workers for 1-10 wt %
gold-titania composites.51,52
Previously, loadings only reached ~2.5 wt % for gold nanocrystals, coated
with strongly bound alkanethiols, infused into silica with CO2 as an antisolvent in
toluene.12 The addition of CO2 lowers the Hamaker constant of the intervening
solvent and strengthens the van der Waals interactions of the particles with the silica,
thus raising the nanocrystal loading.12 In the present study, it was not necessary to add
CO2 to drive the nanocrystal loadings since the particle-support interactions were
strong as a consequence of the weakly bound ligands.
4.4.2 Design of Bimetallic Nanocrystals
The results in Figs. 4.1, 4.3 and 4.5 indicate that the size, composition and
crystal structure of the bimetallic nanocrystals were relatively uniform. These
properties may be shown to be favored by the arrested precipitation in the presence of
stabilizing ligands.53,54 Here, the dynamic solvation of the nanocrystal surface by the
128
ligands may be tuned to produce regions of the nanocrystal surface accessible for
growth, while providing enough steric repulsion to inhibit nanocrystal aggregation.
The temperature may be varied to adjust the precursor decomposition and
consequently the supersaturation, nucleation and growth rates. At sufficiently high
temperature and monomer concentrations, the balance between interfacial energy and
supersaturation produces a small critical size where the growth rate is zero.54 Here,
the nanocrystal diameter may be focused to narrow polydispersity, as seen in Figure
4.1, as the growth rate is higher for the smaller nanocrystals than the larger ones.54
The focusing is optimal if the monomer concentration and dynamic solvation
maintain the average nanocrystal size slightly larger than the critical size.
A variety of previous studies indicate that the size and composition of
bimetallic nanocrystals may be controlled over a significant range in the arrested
precipitation technique by varying the temperature and precursor concentrations.36,55-
57 When the rates of precursor decomposition and equilibrium solubilities of the two
resulting metals are similar, formation of binary metal nuclei is favored.36 Even when
these properties are somewhat dissimilar, the nuclei formed from the first metal may
capture newly formed atoms of the second metal before it nucleates, which would
otherwise lead to two separate populations of growing nanocrystals.2 In essence, this
may be considered to be composition focusing by analogy to size focusing. As metal
from the more rapidly decaying precursor is incorporated into the nanoparticle, it will
be depleted in the solution relative to the second metal. Thus, the probability of
adding the second component will increase. This process will focus the overall
composition as well as the particle size. The temperature is a key variable to produce
rapid decomposition of both precursors to achieve focusing of both size and
composition, along with sufficient dynamic solvation by the ligand to arrest growth.
Because our FePt nanocrystals were pre-synthesized, heterogeneous sites on a support
were not present, which could otherwise lead to uncontrolled composition, resulting
from competition between nucleation in solution and on the surface. Once the
particles are pre-synthesized, they may be loaded physically into the pores of the
129
support at completely different concentrations than those used in the synthesis of the
particles. Thus, both the pre-synthesis and infusion steps may be optimized in ways
not available when precursors are reduced within pores of the support.
A variety of factors complicate control over the size, composition and
morphology of bimetallic nanocrystals synthesized by the traditional technique of
precursor reduction directly on a support.34 The precursors may be reduced either
simultaneously or sequentially.58-61 The diffusional resistance of the precursors
through the pores may prevent the production of large supersaturation values that
favor control of nucleation, growth and size focusing. For simultaneous impregnation,
the metal ions may separate chromatographically as they diffuse into the pores,
leading to large variation in nanocrystal composition.58 For sequential impregnation,
the second metal may form a shell around the first metal, or even nucleate on separate
heterogeneous sites on the support.59 High temperature annealing may promote
alloying between metals, but can also cause sintering of the nanoparticles.58 Without
the use of stabilizing ligands, dynamic solvation is not present to aid size and
composition focusing. The precursor concentrations of the metals must be chosen to
control the particle size and composition, as well as the final metal loadings on the
support. For example, low precursor concentrations for low loadings may produce
insufficient supersaturation and thus defocusing,54 whereas high loadings may require
long reaction times, which may lead to multiple nucleation events resulting in
polydispersity. Another challenge is that the reactants and/or by-products formed
during the reduction step may prevent adhesion of the metal particles on the support,
resulting in limited catalyst stability. In the pre-synthesis technique, the particles may
be isolated from unwanted byproducts, cleaned, and separated by size, and the ligands
may be exchanged if desired prior to infusion.
Thermal annealing above ~550 °C may be utilized to order the atoms in the
FePt nanocrystals to form an intermetallic alloy.36 The reduction in the melting point
of the nanocrystals can cause spreading of the nanocrystals as they wet the support.40
The resulting surface diffusion may lead to coalscence.62 This undesired coalescence
130
appeared to be mitigated with the pre-synthesized nanocrystals, indicating strong
metal-support interactions (Figure 4.4). Ligand removal at moderate temperatures
below the melting point further strengthens interactions between the bare metal and
the support.
High temperature annealing is not always needed to make ordered
multimetallic nanocrystals by the “polyol process”. Ordered intermetallic
nanocrystals have been synthesized in solvents including ethylene glycol63-65 or
tetraethylene glycol66,67 with a control over the size at 100-300 °C. These ordered
nanoparticles may then be infused into the support to form catalysts without the need
for high temperature thermal annealing.
4.4.3 Catalytic Activity of Supported Nanocrystals
FePt catalysts were highly active and stable for 1-decene hydrogenation for at
least three reaction cycles. The enhanced stability may be attributed to the strong
short-ranged metal-support interactions, which prevent desorption or
coalescence/aggregation of the metals during the reaction. For the hydrogenation of
1-hexene, the TOF of bimetallic Pd-Ru nanoparticles exceeded by 10 times that of
individual Pd and Ru nanoparticles indicating synergism between these two types of
atoms.68 Similar effects have been reported for other bimetallic catalysts.69 The
uniform composition of the pre-synthesized bimetallic nanocrystals in the current
study and the ability to control the intermetallic alloy structure may be expected to be
beneficial for achieving these types of synergistic effects on catalytic activity.
The FePt particle size formed by precursor reduction of 6 nm in mesoporous
FePt SBA-15 silica without ligands was much larger than the 3.6 nm particles formed
in the current study by infusion of ligand stabilized particles. Furthermore, upon
ordering the FePt by thermal annealing at 650 °C, the size increased modestly from 6
to 7.2 nm. As the particles approach the pore size of the silica, blocking of pores may
limit mass transfer during catalysis. Furthermore, the surface area/volume is two
times larger for our 3.6 nm particles versus 7.2 nm particles.
131
Reducing the size of the metal nanoparticles below 5 nm significantly
increases the fraction of surface atoms and hence the mass catalytic activity. For the
small 3.6 nm FePt nanocrystals, the morphologies of the crystal facets may be
designed to produce more active surfaces than for large particles.70 For example,
atoms present on the jagged “step edges” of small gold nanoparticles (<5 nm) were
found to be highly catalytically active for CO oxidation.2 Pre-synthesis of the
bimetallic nanocrystals may enable greater control over these types of active sites
versus traditional precursor reduction techniques, with the potential of higher catalyst
activity and selectivity.
4.5 Conclusions
Herein, we have demonstrated a novel method to design highly stable and
active catalysts composed of bimetallic nanocrystals on mesoporous supports with a
controlled intermetallic alloy structure. The catalysts were formed by infusion of pre-
synthesized nanocrystals with well-controlled size and composition. The arrested
growth precipitation favors focusing of the size (<4 nm) as well as the composition of
the nanocrystals. The stabilizing ligands control dynamic solvation, supersaturation,
nucleation and growth, as reflected in the small nanocrystal size. The low surface
coverage of the weakly bound ligands resulted in strong metal-support interactions40
as indicated by high metal loadings above 10 wt % in only 10 min and stability
against sintering upon thermal annealing at 700 °C. These interactions are further
increased by the enhanced interfacial contact area between the metal nanocrystals and
concave mesopores with similar radii. For strongly binding ligands such as
dodecanethiol20 or polymers such as polyvinylpyrrolidone3 with high surface
coverages, high loadings were not achieved. According to HAADF-STEM, more than
80% of the nanocrystals were within the pores, relative to the external surface,
consistent with the resistance to sintering at 700 °C. Finally, the calcined FePt
catalyst was found to exhibit six fold higher catalyst activity than commercial Pd-
132
alumina catalyst for liquid 1-decene hydrogenation and was stable for multiple
reactions. The decoupling of nanocrystal synthesis and infusion provides exquisite
control of the nanocrystal size, alloy structure, binding to the support and
dispersibility within the pores, relative to conventional precursor reduction
techniques. The ability to design catalyst morphology and crystal structure with pre-
synthesized bimetallic nanocrystals offers novel opportunities for enhanced catalyst
activities, selectivities and stabilities.
4.6 References
(1) Wakayama, H.; Setoyama, N.; Fukushima, Y. Size-Controlled
Synthesis and Catalytic Performance of Pt Nanoparticles in Micro- and Mesoporous
Silica Prepared Using Supercritical Solvents. Adv. Mater. 2003, 15, 742-745.
(2) Chan, K.-Y.; Ding, J.; Ren, J.; Cheng, S.; Tsang, K. Y. Supported
mixed metal nanoparticles as electrocatalysts in low temperature fuel cells. Journal of
Materials Chemistry 2004, 14, 505-516.
(3) Rioux, R. M.; Song, H.; Hoefelmeyer, J. D.; Yang, P.; Somorjai, G. A.
High-Surface-Area Catalyst Design: Synthesis, Characterization, and Reaction
Studies of Platinum Nanoparticles in Mesoporous SBA-15 Silica. Journal of Physical
Chemistry B 2005, 109, 2192-2202.
(4) Akita, T.; Okumura, M.; Tanaka, K.; Kohyama, M.; Haruta, M.
Analytical TEM observation of Au nano-particles on cerium oxide. Catalysis Today
2006, 117, 62-68.
(5) Meunier, F. C.; Tibiletti, D.; Goguet, A.; Shekhtman, S.; Hardacre, C.;
Burch, R. On the complexity of the water-gas shift reaction mechanism over a
Pt/CeO2 catalyst: Effect of the temperature on the reactivity of formate surface
species studied by operando DRIFT during isotopic transient at chemical steady-state.
Catalysis Today 2007, 126, 143-147.
133
(6) Stamenkovic, V. R.; Mun, B. S.; Arenz, M.; Mayrhofer, K. J. J.;
Lucas, C. A.; Wang, G.; Ross, P. N.; Markovic, N. M. Trends in electrocatalysis on
extended and nanoscale Pt-bimetallic alloy surfaces. Nature Materials 2007, 6, 241-
247.
(7) Zheng, N.; Stucky, G. D. A General Synthetic Strategy for Oxide-
Supported Metal Nanoparticle Catalysts. Journal of the American Chemical Society
2006, 128, 14278-14280.
(8) Kang, S.; Jia, Z.; Zoto, I.; Reed, D.; Nikles, D. E.; Harrell, J. W.;
Thompson, G.; Mankey, G.; Krishnamurthy, V. V.; Porcar, L. Sintering behavior of
spin-coated FePt and FePtAu nanoparticles. Journal of Applied Physics 2006, 99,
08N704/701-708N704/703.
(9) Koh, S.; Toney, M. F.; Strasser, P. Activity-stability relationships of
ordered and disordered alloy phases of Pt3Co electrocatalysts for the oxygen
reduction reaction (ORR). Electrochimica Acta 2007, 52, 2765-2774.
(10) Song, Z.; Cai, T.; Hanson, J. C.; Rodriguez, J. A.; Hrbek, J. Structure
and Reactivity of Ru Nanoparticles Supported on Modified Graphite Surfaces: A
Study of the Model Catalysts for Ammonia Synthesis. Journal of the American
Chemical Society 2004, 126, 8576-8584.
(11) Li, Y.; Fu, Q.; Flytzani-Stephanopoulos, M. Low-temperature water-
gas shift reaction over Cu- and Ni-loaded cerium oxide catalysts. Applied Catalysis,
B: Environmental 2000, 27, 179-191.
(12) Gupta, G.; Shah, P. S.; Zhang, X.; Saunders, A. E.; Korgel, B. A.;
Johnston, K. P. Enhanced Infusion of Gold Nanocrystals into Mesoporous Silica with
Supercritical Carbon Dioxide. Chemistry of Materials 2005, 17, 6728-6738.
(13) Tsung, C.-K.; Hong, W.; Shi, Q.; Kou, X.; Yeung, M. H.; Wang, J.;
Stucky, G. D. Shape- and orientation-controlled gold nanoparticles formed within
mesoporous silica nanofibers. Advanced Functional Materials 2006, 16, 2225-2230.
(14) Lee, J.; Kim, J.; Hyeon, T. Recent progress in the synthesis of porous
carbon materials. Advanced Materials (Weinheim, Germany) 2006, 18, 2073-2094.
134
(15) Morey, M. S.; Davidson, A.; Stucky, G. D. Silica-based, cubic
mesostructures: synthesis, characterization and relevance for catalysis. Journal of
Porous Materials 1998, 5, 195-204.
(16) Hanrahan, J. P.; Copley, M. P.; Ziegler, K. J.; Spalding, T. R.; Morris,
M. A.; Steytler, D. C.; Heenan, R. K.; Schweins, R.; Holmes, J. D. Pore Size
Engineering in Mesoporous Silicas Using Supercritical CO2. Langmuir 2005, 21,
4163-4167.
(17) Chen, Y.; Yang, F.; Dai, Y.; Wang, W.; Chen, S. Ni@Pt core-shell
nanoparticles: synthesis, structural and electrochemical properties. Journal of
Physical Chemistry C 2008, 112, 1645-1649.
(18) Zhao, D.; Feng, J.; Huo, Q.; Melosh, N.; Frederickson, G. H.;
Chmelka, B. F.; Stucky, G. D. Triblock copolymer syntheses of mesoporous silica
with periodic 50 to 300 angstrom pores. Science (Washington, D. C.) 1998, 279, 548-
552.
(19) Joo, S. H.; Choi, S. J.; Oh, I.; Kwak, J.; Liu, Z.; Terasaki, O.; Ryoo, R.
Ordered nanoporous arrays of carbon supporting high dispersions of platinum
nanoparticles. Nature (London, United Kingdom) 2001, 412, 169-172.
(20) Gupta, G.; Stowell, C. A.; Patel, M. N.; Gao, X.; Yacaman, M. J.;
Korgel, B. A.; Johnston, K. P. Infusion of Pre-synthesized Iridium Nanocrystals into
Mesoporous Silica for High Catalyst Activity. Chemistry of Materials 2006, 18,
6239-6249.
(21) Jung, J. S.; Lim, J. H.; Malkinski, L.; Vovk, A.; Choi, K. H.; Oh, S. L.;
Kim, Y. R.; Jun, J. H. Magnetic properties of structurally confined FePt nanoparticles
within mesoporous nanotubes. Journal of Magnetism and Magnetic Materials 2007,
310, 2361-2363.
(22) Konya, Z.; Puntes, V. F.; Kiricsi, I.; Zhu, J.; Ager, J. W., III; Ko, M.
K.; Frei, H.; Alivisatos, P.; Somorjai, G. A. Synthetic Insertion of Gold Nanoparticles
into Mesoporous Silica. Chemistry of Materials 2003, 15, 1242-1248.
135
(23) Song, H.; Rioux, R. M.; Hoefelmeyer, J. D.; Komor, R.; Niesz, K.;
Grass, M.; Yang, P.; Somorjai, G. A. Hydrothermal Growth of Mesoporous SBA-15
Silica in the Presence of PVP-Stabilized Pt Nanoparticles: Synthesis,
Characterization, and Catalytic Properties. Journal of the American Chemical Society
2006, 128, 3027-3037.
(24) Raghuveer, V.; Manthiram, A.; Bard, A. J. Pd-Co-Mo Electrocatalyst
for the Oxygen Reduction Reaction in Proton Exchange Membrane Fuel Cells.
Journal of Physical Chemistry B 2005, 109, 22909-22912.
(25) Raja, R.; Khimyak, T.; Thomas, J. M.; Hermans, S.; Johnson, B. F. G.
Single-step, highly active, and highly selective nanoparticle catalysts for the
hydrogenation of key organic compounds. Angewandte Chemie, International Edition
2001, 40, 4638-4642.
(26) Casado-Rivera, E.; Volpe, D. J.; Alden, L.; Lind, C.; Downie, C.;
Vazquez-Alvarez, T.; Angelo, A. C. D.; DiSalvo, F. J.; Abruna, H. D. Electrocatalytic
Activity of Ordered Intermetallic Phases for Fuel Cell Applications. Journal of the
American Chemical Society 2004, 126, 4043-4049.
(27) Xu, Y.; Ruban, A. V.; Mavrikakis, M. Adsorption and dissociation of
O2 on Pt-Co and Pt-Fe alloys. Journal of the American Chemical Society 2004, 126,
4717-4725.
(28) Folch, B.; Larionova, J.; Guari, Y.; Datas, L.; Guerin, C. A
coordination polymer precursor approach to the synthesis of NiFe bimetallic
nanoparticles within hybrid mesoporous silica. Journal of Materials Chemistry 2006,
16, 4435-4442.
(29) Vondrova, M.; Klimczuk, T.; Miller, V. L.; Kirby, B. W.; Yao, N.;
Cava, R. J.; Bocarsly, A. B. Supported Superparamagnetic Pd/Co Alloy Nanoparticles
Prepared from a Silica/Cyanogel Co-gel. Chemistry of Materials 2005, 17, 6216-
6218.
(30) Kockrick, E.; Krawiec, P.; Schnelle, W.; Geiger, D.; Schappacher, F.
M.; Poettgen, R.; Kaskel, S. Space-confined formation of FePt nanoparticles in
136
ordered mesoporous silica SBA-15. Advanced Materials (Weinheim, Germany) 2007,
19, 3021-3026.
(31) Hermans, S.; Raja, R.; Thomas, J. M.; Johnson, B. F. G.; Sankar, G.;
Gleeson, D. Solvent-free, low-temperature, selective hydrogenation of polyenes using
a bimetallic nanoparticle Ru - Sn catalyst. Angewandte Chemie, International Edition
2001, 40, 1211-1215.
(32) Thomas, J. M.; Johnson, B. F. G.; Raja, R.; Sankar, G.; Midgley, P. A.
High-Performance Nanocatalysts for Single-Step Hydrogenations. Accounts of
Chemical Research 2003, 36, 20-30.
(33) Valden, M.; Lai, X.; Goodman, D. W. Onset of catalytic activity of
gold clusters on titania with the appearance of nonmetallic properties. Science
(Washington, D. C.) 1998, 281, 1647-1650.
(34) Zhou, B.; Hermans, S.; Somorjai, G. A. Nanotechnology in Catalysis;
Kluwer Academic/ Plenum 2004; Vol. 1.
(35) King, N. C.; Blackley, R. A.; Wears, M. L.; Newman, D. M.; Zhou,
W.; Bruce, D. W. The synthesis of mesoporous silicates containing bimetallic
nanoparticles and magnetic properties of PtCo nanoparticles in silica. Chemical
Communications (Cambridge, United Kingdom) 2006, 3414-3416.
(36) Sun, X.; Kang, S.; Harrell, J. W.; Nikles, D. E.; Dai, Z. R.; Li, J.;
Wang, Z. L. Synthesis, chemical ordering, and magnetic properties of FePtCu
nanoparticle films. Journal of Applied Physics 2003, 93, 7337-7339.
(37) Liang, C.; Dai, S. Synthesis of Mesoporous Carbon Materials via
Enhanced Hydrogen-Bonding Interaction. Journal of the American Chemical Society
2006, 128, 5316-5317.
(38) Komiyama, M.; Shimaguchi, T. Partial reduction of Si(IV) in SiO2
thin film by deposited metal particles: an XPS study. Surface and Interface Analysis
2001, 32, 189-192.
137
(39) Klie, R. F.; Disko, M. M.; Browning, N. D. Atomic Scale
Observations of the Chemistry at the Metal-Oxide Interface in Heterogeneous
Catalysts. Journal of Catalysis 2002, 205, 1-6.
(40) Yu, R.; Song, H.; Zhang, X.-F.; Yang, P. Thermal Wetting of Platinum
Nanocrystals on Silica Surface. Journal of Physical Chemistry B 2005, 109, 6940-
6943.
(41) Zhao, D.; Huo, Q.; Feng, J.; Chmelka, B. F.; Stucky, G. D. Nonionic
triblock and star diblock copolymer and oligomeric surfactant syntheses of highly
ordered, hydrothermally stable, mesoporous silica structures. Journal of the American
Chemical Society 1998, 120, 6024-6036.
(42) Williams, D. B.; Carter, C. B. Transmission Electon Microscopy;
Springer: New York, 1996.
(43) Ferrer, D.; Torres-Castro, A.; Gao, X.; Sepulveda-Guzman, S.; Ortiz-
Mendez, U.; Jose-Yacaman, M. Three-layer core/shell structure in Au-Pd bimetallic
nanoparticles. Nano Letters 2007, 7, 1701-1705.
(44) Zhang, L.; He, R.; Gu, H.-C. Oleic acid coating on the monodisperse
magnetite nanoparticles. Applied Surface Science 2006, 253, 2611-2617.
(45) Papirer, E. Adsorption on Silica Surfaces CRC, 2000; Vol. 90.
(46) Israelachvili, J. Intermolecular & Surface Forces; 2nd ed.; Academic
Press, Inc.: San Diego, 1992.
(47) Jana, N. R.; Peng, X. Single-Phase and Gram-Scale Routes toward
Nearly Monodisperse Au and Other Noble Metal Nanocrystals. Journal of the
American Chemical Society 2003, 125, 14280-14281.
(48) Fu, X.; Wang, Y.; Wu, N.; Gui, L.; Tang, Y. Surface Modification of
Small Platinum Nanoclusters with Alkylamine and Alkylthiol: An XPS Study on the
Influence of Organic Ligands on the Pt 4f Binding Energies of Small Platinum
Nanoclusters. Journal of Colloid and Interface Science 2001, 243, 326-330.
138
(49) Korgel, B. A.; Fullam, S.; Connolly, S.; Fitzmaurice, D. Assembly and
Self-Organization of Silver Nanocrystal Superlattices: Ordered \"Soft Spheres\".
Journal of Physical Chemistry B 1998, 102, 8379-8388.
(50) Haller, G. L. New catalytic concepts from new materials:
understanding catalysis from a fundamental perspective, past, present, and future.
Journal of Catalysis 2003, 216, 12-22.
(51) Pietron, J. J.; Stroud, R. M.; Rolison, D. R. Using Three Dimensions in
Catalytic Meso-porous Nanoarchitectures. Nano Letters 2002, 2, 545-549.
(52) Rolison, D. R. Catalytic nanoarchitectures - The importance of nothing
and the unimportance of periodicity. Science (Washington, DC, United States) 2003,
299, 1698-1702.
(53) Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R.
Synthesis of Thiol-Derivatised Gold Nanoparticles in a Two-Phase Liquid-Liquid
System. J. Chem. Soc., Chem. Commun. 1994, 7, 801-802.
(54) Yin, Y.; Alivisatos, A. P. Colloidal nanocrystal synthesis and the
organic-inorganic interface. Nature (London, United Kingdom) 2005, 437, 664-670.
(55) Sra, A. K.; Ewers, T. D.; Schaak, R. E. Direct Solution Synthesis of
Intermetallic AuCu and AuCu3 Nanocrystals and Nanowire Networks. Chemistry of
Materials 2005, 17, 758-766.
(56) Santiago, E. I.; Varanda, L. C.; Villullas, H. M. Carbon-Supported Pt-
Co Catalysts Prepared by a Modified Polyol Process as Cathodes for PEM Fuel Cells.
Journal of Physical Chemistry C 2007, 111, 3146-3151.
(57) Alden, L. R.; Roychowdhury, C.; Matsumoto, F.; Han, D. K.;
Zeldovich, V. B.; Abruna, H. D.; DiSalvo, F. J. Synthesis, Characterization, and
Electrocatalytic Activity of PtPb Nanoparticles Prepared by Two Synthetic
Approaches. Langmuir 2006, 22, 10465-10471.
(58) Ponec, V.; Bond, G. C. Cataysis by Metals and Alloys; Elsevier
Netherlands, 1995.
139
(59) Venezia, A. M.; Liotta, L. F.; Deganello, G.; Schay, Z.; Guczi, L.
Characterization of Pumice-Supported Ag-Pd and Cu-Pd Bimetallic Catalysts by X-
Ray Photoelectron Spectroscopy and X-Ray Diffraction. Journal of Catalysis 1999,
182, 449-455.
(60) Konopny, L. W.; Juan, A.; Damiani, D. E. Preparation and
characterization of g-Al2O3-supported Pd-Mo catalysts. Applied Catalysis, B:
Environmental 1998, 15, 115-127.
(61) Batista, J.; Pintar, A.; Gomilsek, J. P.; Kodre, A.; Bornette, F. On the
structural characteristics of g-alumina-supported Pd-Cu bimetallic catalysts. Applied
Catalysis, A: General 2001, 217, 55-68.
(62) Wang, Z. L.; Petroski, J. M.; Green, T. C.; El-Sayed, M. A. Shape
Transformation and Surface Melting of Cubic and Tetrahedral Platinum Nanocrystals.
Journal of Physical Chemistry B 1998, 102, 6145.
(63) Jeyadevan, B.; Hobo, A.; Urakawa, K.; Chinnasamy, C. N.; Shinoda,
K.; Tohji, K. Towards direct synthesis of fct-FePt nanoparticles by chemical route.
Journal of Applied Physics 2003, 93, 7574-7576.
(64) Iwaki, T.; Kakihara, Y.; Toda, T.; Abdullah, M.; Okuyama, K.
Preparation of high coercivity magnetic FePt nanoparticles by liquid process. Journal
of Applied Physics 2003, 94, 6807-6811.
(65) Roychowdhury, C.; Matsumoto, F.; Mutolo, P. F.; Abruna, H. D.;
DiSalvo, F. J. Synthesis, Characterization, and Electrocatalytic Activity of PtBi
Nanoparticles Prepared by the Polyol Process. Chemistry of Materials 2005, 17,
5871-5876.
(66) Minami, R.; Kitamoto, Y.; Chikata, T.; Kato, S. Direct synthesis of
L10 type Fe-Pt nanoparticles using microwave-polyol method. Electrochimica Acta
2005, 51, 864-866.
(67) Leonard, B. M.; Bhuvanesh, N. S. P.; Schaak, R. E. Low-temperature
polyol synthesis of AuCuSn2 and AuNiSn2: Using solution chemistry to access
140
ternary intermetallic compounds as nanocrystals. Journal of the American Chemical
Society 2005, 127, 7326-7327.
(68) Raja, R.; Hermans, S.; Shephard, D. S.; Johnson, B. F. G.; Raja, R.;
Sankar, G.; Bromley, S.; Thomas, J. M. Preparation and characterisation of a highly
active bimetallic (Pd-Ru) nanoparticle heterogeneous catalyst. Chemical
Communications (Cambridge) 1999, 1571-1572.
(69) Scott, R. W. J.; Sivadinarayana, C.; Wilson, O. M.; Yan, Z.; Goodman,
D. W.; Crooks, R. M. Titania-Supported PdAu Bimetallic Catalysts Prepared from
Dendrimer-Encapsulated Nanoparticle Precursors. Journal of the American Chemical
Society 2005, 127, 1380-1381.
(70) Min, M.-K.; Cho, J.; Cho, K.; Kim, H. Particle size and alloying
effects of Pt-based alloy catalysts for fuel cell applications. Electrochimica Acta
2000, 45, 4211-4217.
141
Chapter 5
Highly Stable ORR Electrocatalysts from Pre-synthesized Pt
Nanocrystals Supported on Graphitic Mesoporous Carbons
Highly stable Pt electrocatalysts for oxygen reduction have been prepared by the
infusion of pre-synthesized nanocrystals with controlled size (<4 nm) and surface
morphology into a highly graphitic, ordered 3d cubic mesoporous carbon with 7 nm
pores. The weakly bound dodecylamine ligands expose a sufficient number of Pt
atoms on the nanoparticle surface that bind strongly to the carbon surface to facilitate
high Pt loading up to 20 wt % with high metal dispersion, as indicated by XRD and
TEM. The mass activity for the oxygen reduction reaction (ORR) at 0.9 V reaches
0.14 A/mgPt for Pt on the graphitic mesoporous carbon, 25% larger than for Pt on
commercial amorphous Vulcan XC-72. The loss in ORR catalyst activity for Pt
supported on graphitic mesoporous carbons under an accelerated durability test of
potential cycling between 0.5 to 1.2 V for 1000 cycles, is extremely small, <2%,
compared to a 30% loss for Pt on amorphous mesoporous carbons and a 70% loss that
is typical for commercial Pt on amorphous carbon catalysts. The high resistance of
the graphitized carbon against oxidation, as characterized by thermo gravimetric
analysis, coupled with the strong and electron transfer Pt-C interactions, inhibit
nanoparticle isolation, nanoparticle coalescence, Pt dissolution and Ostwald ripening,
resulting in high catalyst stability. The large interfacial contact area between the
metal nanocrystals and concave mesopores with comparable radii, along with the high
Pt dispersion on the large surface area (>320 m2/g) supports further lower the
tendency of Pt coalescence. The ability to disperse pre-synthesized metal nanocrystals
with controlled size and surface morphologies onto mesoporous graphitic carbon
supports offers the potential for new classes of highly stable and active catalysts.
The contents of this chapter are being prepared for a manuscript.
142
5.1 Introduction
A significant challenge hindering the development of catalysts for the oxygen
reduction reaction in proton exchange membrane fuel cells, especially in
transportation applications, is the catalyst stability towards potential cycling.1-5 The
catalyst must be stable at 1.2 V for 100 h corresponding to the accumulated life time
of the vehicle, with a maximum allowable performance loss of <30 mV at 1.5A/cm2
(<1 µV/cycle target).4,5 However, Pt catalysts supported on traditional amorphous
carbons, containing oxygenated surface groups, undergo significant losses at 1.2 V
due to carbon oxidation to CO/CO2. Pt is known to catalyze the oxidation of the
support. As the carbon oxidatively degrades, the supported nanoparticles become
isolated, resulting in a loss in activity.1,5 Aside from carbon degradation, potential
cycling between 0.5 and 1.2 V causes Pt sintering as well as dissolution.6 For
example, Pt on Vulcan XC-72 carbon showed a 70% loss in the electrochemical area
upon potential cycling between 0.5 and 1.2 V for 1000 cycles.7 The presence of
oxygen species withdraws electrons from the Pt and weakens the binding of the Pt to
the carbon, thus facilitating Pt sintering and dissolution.8
The stability of Pt catalysts may be enhanced with corrosion resistant
graphitized carbons.5 These carbons are considerably more crystalline than typical
more amorphous carbons such as Vulcan XC-72 and contain very few oxidizable
polar groups on the surface. Consequently graphitized carbons are far more resistant
towards oxidation catalyzed by platinum.6,9,10 The abundant π sites on the surface of
graphitized carbon interact strongly with Pt as demonstrated by XPS and electron-
spin resonance (ESR) studies.3,6,9 The Pt-C interactions are partially covalent, due to
electron delocalization between the p-orbital of the π sites and Pt d-orbital, and are
partially ionic due to electron transfer from Pt to the carbon.11 Strong metal-support
interactions may inhibit Pt oxidation and dissolution during potential cycling and
lower the degree of Ostwald ripening and coalescence.6 Platinum supported on carbon
nanotubes (CNT) with graphitic surfaces is more stable than Pt on amorphous
143
carbon.12-14 For example, 30% loss in ECSA occurred for CNT compared to 70% loss
for VulcanXC-72 during potential cycling between 0.5 – 1.2 V for 1000 cycles. This
advancement is an important step in meeting the stability criteria for viable fuel cell
catalysts.5
The traditional technique of wetness impregnation may be utilized to form Pt
nanoparticles by reduction of Pt precursors on amorphous carbon, however, this
approach is limited for graphitic carbon supports.15 For amorphous carbon, polar
oxygenated surface species are required for adsorption of ionic platinum precursors,
such as hexachloroplatinic acid, to produce dispersed metal nanoparticles on the order
of 2-3 nm.3 However, the limited number of oxygen sites on graphitized carbon is
insufficient to achieve sufficient precursor adsorption. Consequently, reduction of
these weakly binding precursors leads to large nanoparticle sizes (>5 nm).6 To
prevent sintering, graphitic supports typically require surface modification, such as
refluxing in nitric acid, to introduce polar oxygen groups on the carbon surface to
facilitate binding of the precursors.16-19 However, after such treatments, the modified
graphitic supports are more susceptible to oxidation and the surface oxygen groups
weaken the final Pt nanocrystal-support interactions, raising the tendency towards Pt
dissolution and coalescence.14
An emerging concept in catalyst design is to pre-synthesize colloidal metal
nanocrystals coated with stabilizers, and then to infuse them onto high surface area
supports.20-23 The decoupling of nanocrystal synthesis and infusion provides exquisite
control over the nanocrystal size, morphology, and dispersibility within the porous
support. This technique has been reported for the preparation of fuel cell catalysts on
amorphous carbon, where the nanoparticles were stabilized with ligands24-27 or
microemulsions.28,29 High metal loadings have been achieved for nanoparticles
stabilized by weakly bound hydrocarbon capping ligands.21 The high loadings
suggested that the low metal surface coverage of these weakly bound ligands did not
screen the strong short-ranged interactions with the support. Furthermore, the ligands
may be removed at temperatures of ≤500 °C to activate the metal surface prior to
144
catalysis.4,21,23 For example, highly active FePt catalysts with high loadings have been
synthesized by infusing pre-synthesized nanocrystals coated with weakly bound oleic
acid and oleylamine ligands in mesoporous silica.21
The primary objective of this study was to design highly stable and active
ORR Pt catalysts with controlled size and morphology supported on mesoporous
graphitic carbon formed from mesophase pitch. The catalyst stability was tested under
relatively severe conditions, 1000 cycles between 0.5 and 1.2 V at 50 mV/s. Pt
nanocrystals were pre-synthesized by arrested growth precipitation with
dodecylamine ligands to control the size and surface morphology, and then infused
into graphitic mesoporous carbons. The large pore size of the mesoporous carbons (7-
10 nm) was achieved by eliminating the microporosity in the walls of mesoporous
silica template, to prevent formation of interconnecting carbon rods.30 Graphitic
supports are highly resistant to carbon oxidation3 and produce strong interactions with
Pt.6 The strong metal-support interactions will be shown to enhance catalyst stability
by inhibiting Pt dissolution, Ostwald ripening and coalescence. The concave pores in
the carbon were designed to achieve large interfacial area with the 3 nm spherical Pt
nanocrystals to strengthen metal-support interactions. Furthermore, the high
dispersion of the Pt nanoparticles on the high surface area (350 m2/g) graphitic carbon
with the 7 nm mesopores lowers the tendency for Pt coalescence. Weakly binding
ligands (dodecylamine) were chosen to stabilize the Pt nanoparticles to expose a
greater fraction of the metal surface to the graphitic π sites, to promote adsorption of
the nanocrystals at loadings up to 20 wt %. The low molecular weight ligands were
removed by mild thermal calcination to activate the catalyst while attempting to avoid
changing the nanocrystal size and morphology. The catalyst stabilities under potential
cycling are compared with those of previous studies on amorphous mesoporous
carbons.24,31-35 To the best of our knowledge, use of highly graphitic mesoporous
carbon as a support for synthesis of highly stable Pt catalysts for ORR has not been
reported.
145
The porous structure of the mesoporous carbons are characterized by nitrogen
porosimetry and small angle X-ray scattering (SAXS). The degree of graphitization of
the mesoporous carbons is characterized by X-ray diffraction (XRD) and the
oxidative stability by thermal gravimetric analysis. The size of the supported Pt
nanoparticles is characterized by XRD, TEM and electrochemical surface area
(ECSA). Pt-C interactions are quantified by XPS. In addition, the size of
nanoparticles and surface characteristics of the nanoparticles are characterized by
TEM, XRD and XPS to provide insight into the catalyst activity and stability.
5.2 Experimental Section
All chemicals were used as received. Platinum (IV) chloride (99.9%), ,
dodecylamine (98%), perchloric acid 70% (99.999% purity), sodium borohydride
(98%), tetradecylammonium bromide (98%), anhydrous ethanol, Pluronic P123
(EO20PO70EO20), n-butanol (99.9%), 5 wt % nafion solution in lower alcohols were
purchased from Aldrich. Toluene (99.9%) was obtained from Fisher Scientific,
ethanol (Absolute 200 proof) from Aaper alcohol. Polycarbonate filters (0.05 micron)
were obtained from Osmonics. High purity water (resistance ~18 MΩ) was used.
Platinum nanoparticles stabilized by dodecylamine were synthesized using the
phase-transfer method developed by Wikander at al.36 15mL of 60mM aqueous
solution of platinum chloride was combined with 105mL of toluene containing 1.5 of
TDAB (phase transfer catalyst). The mixture was stirred for 2 hours leading to a
platinum transfer from the aqueous phase to an organic phase. 3.4 g of dodecylamine
in 15mL of toluene was added to the organic phase while it stirring. Platinum was
then reduced by the drop wise addition NaBH4 (0.57 g NaBH4 dissolved in 25mL
water). Upon reduction, the solution turned dark brown and was stirred for 2h, before
initiating the cleaning phase. For cleaning, 500mL of ethanol was added to the
toluene phase (reduced to approximately 1/10th the original volume) and then kept
overnight in a freezer (-10 °C). The platinum nanocrystals precipitated at the bottom
146
of the breaker and were collected by decanting and dispersed in toluene. The platinum
nanocrystals were further washed by adding excess ethanol (toluene:ethanol=1:35)
and then dispersed in toluene.
The ordered mesoporous carbons were synthesized by the recipe described in
detail by using mesoporous silica of type KIT-637 as a template.30,38 Briefly, the
mesoporous silica was synthesized by mixing Pluronic P123 and n-butanol in
aqueous HCl solution. After complete dissolution, tetraethyl orthosilicate (TEOS,
98% Acros) was added into the mixture at 308 K. The molar composition of the
starting reaction mixture was 0.017 P123: 1.2 TEOS: 1.31 BuOH: 1.83 HCl: 195
H2O. The mixture was magnetically stirred for 1 day at the same temperature.
Subsequently, the mixture was heated for 1 day in an oven for hydrothermal
treatment. After the hydrothermal treatment, the solid product was recovered by
filtration, washed with ethanol and then calcined at 823K to obtain mesoporous silica.
The mesoporous silica was then mixed with carbon precursor (furfural alcohol or
mesophase pitch) and p-toluene sulfonic acid at room temperature (1g silica: 0.458
mL carbon precursor: 1.97 g p-toluene sulfonic acid). The resultant sample was
heated at 308 K for 1 h, at 373 K for 1h and then for 2 h at 623 K. Carbonization was
performed by heating to 1173 K at 3.7 K/min in argon. The silica template was then
removed using HF to recover mesoporous carbon. Mesoporous carbon synthesized
with furfural alcohol is more amorphous39 and is denoted as AMC and that with
mesophase pitch is more graphitic and is denoted as GMC.
The infusion of the Pt nanoparticles in mesoporous carbon was done by a
similar procedure as previously reported in Gupta et al.21 30 mg of mesoporous
carbon was mixed with ~3 mL of Pt nanocrystal dispersion in toluene of known
concentrations ~4-10 mg/mL. The contents were then stirred for 2 days. After
stirring, the mesoporous carbon infused with Pt nanocrystals was separated from the
nanocrystal supernatant by filtration with a 0.2 μm PTFE filter. The extent of
incorporation of Pt nanocrystals in the carbon was determined by subtracting from the
147
initial mass of the dispersed nanocrystals in toluene, the final mass of the nanocrystals
recovered in the supernatant after filtration.22 The absorbance of the Pt nanocrystals
dispersed in toluene before and after infusion was measured using a Cary 500 UV-
Vis-NIR spectrophotometer with an optical path length of 1 cm. Figure 5.1a shows a
representative absorbance curve for the Pt nanoparticles.
A standard calibration curve with a correlation coefficient of 0.998 was
generated with known concentrations of Pt nanocrystals in toluene for a wavelength
of 800 nm (Figure 5.1b). Typically loadings of 20% metal on mesoporous carbons for
both AMC and GMC were obtained after infusion. Total uncertainty in loadings
determined with this technique was approximately 12% as explained in detail
previously.23 To activate the catalysts, ligands were removed by calcining the catalyst
at 300-500 °C for 2h. The temperature was ramped at 16 °C/min. After calcination,
the catalysts were washed with 100 mL of ethanol.
N2 sorption analysis was performed on a Micromeritics ASAP 2010
volumetric analyzer. The specific surface area was calculated using the BET method
from the nitrogen adsorption data in the relative range (P/P0) of 0.05-0.30. The
samples were outgases at 200 °C in the degas port of the adsorption analyzer. The
total pore volume was determined from the amount of N2 uptake at P/P0 = 0.95. The
pore size distribution was derived from the adsorption branch of the isotherm based
on the BJH model.
Wide angle X-ray diffraction was performed with samples on a quartz slide
using a Bruker Nokius instrument using Cu Kα radiation with wavelength = 1.54 Å.
Samples were scanned for 12 h at a scan rate of 12 °/min with 0.02 deg increments.
The average nanocrystal size was estimated from the Scherrer equation using JADE
software (by Molecular Diffraction Inc). Small angle X-ray diffraction was collected
using Pohang Light Source in the reflection mode.
Thermogravimetric analyses (TGA) were made with a Perkin Elmer TGA 7
instrument with ~2 mg catalyst samples. All samples were initially heated to 50 °C,
148
(a)
Figure 5.1
(a) UV-Vis spectra of Pt nanoparticles dispersed in toluene before and after infusion
in AMC and (b) calibration curve for Pt nanoparticles using known concentrations
and measuring absorbance at 800 nm. Correlation coefficient was 0.9998.
00.5
11.5
22.5
33.5
4
400 500 600 700 800Wavelength (nm)
Abs
orba
nce
Before infusion
After infusion
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0 0.2 0.4 0.6
Conc of Pt‐DDA (mg/ml)
Absorbance at 800
nm
149
held at that temperature for 10 min, and heated to 900 °C at 20 °C/min. Isothermal
weight loss studies were conducted by holding the samples of 100 mg weight at an
elevated temperature of 195 °C for 16 days, similarly to as described by Stevens and
Dahn.40
Low resolution TEM was conducted on a Phillips EM280 microscope with a
4.5Å point-to-point resolution operated with a 80 kV operating voltage. Platinum
catalyst was deposited from a dilute ethanolic solution onto a 200 mesh carbon-coated
copper TEM grid (Electron Microscopy Sciences).
A glassy carbon rotating disk electrode (0.196 cm2, from Pine Instruments)
was used as a substrate for the catalyst and was polished with alumina suspension of
0.3µm followed by 0.05 µm to give a mirror finish. 1.0 mg of Pt catalyst was
suspended by sonicating in 300µL of anhydrous ethanol and 300 µL of 0.15 wt %
nafion solution (diluted from 5% nafion stock solution). A total of 10 µL of Pt
catalyst suspension was pipetted onto the substrate and was dried at room temperature
for 15 min.
Electrochemical measurements were carried out using a three-compartment
electrochemical cell with a Pt wire and Hg/Hg2SO4 serving as the counter and the
reference electrodes, respectively. Electrochemical measurements were recorded
using CHI 832 electrochemical analyzer. All potentials cited in this article are
reported with respect to NHE.
After preparation of the catalyst on the substrate, the catalyst was cleaned by
cycling potential between 0.05 and 1.24 V at 200 mV/s for 100 cycles in argon
saturated 0.1M HClO4 electrolyte. The area under the desorption peak of
underpotentially deposited hydrogen in a CV (scanned at 100 mV/sec between 0.04
and 1.24V), was used to evaluate electrochemically active surface area (ECSA) of the
Pt catalyst on the electrode. Hydrodynamic voltammograms at rotating disk electrode
were recorded by anodically scanning its potential from 0.04 to 1.24 V at 5 mV/s in
oxygen saturated 0.1 M HClO4. Accelerated durability tests (ADT) were employed by
150
cycling the electrode potential between 0.5 and 1.2 V at 50 mV/s for 1000 cycles in
argon saturated atmosphere over a period of 7 h.
5.3 Results
The structure of the ordered mesoporous carbons was studied using small
angle X-ray scattering. For both AMC and GMC, multiple reflections associated with
the I4132 space-group symmetry (3d cubic morphology) are observed (Figure 5.2)
and indicate a long range ordering of pores. The first three reflections appear at 2θ =
0.68°, 1.2° and 1.4° and 2θ = 0.68°, 1.22° and 1.43° for GMC and AMC,
respectively. The d-spacings corresponding to the reflections are calculated using
Bragg’s Law ( )sin(.2 θλ dn = ).The d-spacings are estimated to be 130, 74 and 64 Å
for GMC and 130, 72.3 and 62 Å for AMC for the first three reflections. The ratio of
the d-spacings d1:d2:d3 is thus 3:2:32 for both AMC and GMC, and can be
indexed as 110, 211 and 220 planes of I4132 tetragonal geometry which is consistent
with the literature.30
Figures 5.3(a) and 5.3(b) gives the adsorption-desorption isotherms and
corresponding pore size distributions (in inset) for GMC and AMC, respectively. For
both carbons, three well-distinguished regions of the adsorption isotherm are evident:
(i) monolayer- multilayer adsorption, (ii) capillary condensation, and (iii) multilayer
adsorption on the outer particle surfaces. The adsorption isotherms show a type H1
behavior with capillary condensation at P/Po ~ 0.6 for AMC and of type H2 for GMC
with onset of capillary condensation at P/Po ~ 0.5. The slope and almost vertical step
in the hysteresis loops suggest well-defined mesopores with a narrow pore size
distribution as reflected by the pore size distributions.
The surface area, pore volume and mean pore size for GMC and AMC are
shown in Table 5.1. The surface area for AMC and GMC are 978 and 337 m2/g and
are significantly higher than VulcanXC-72 (~250 m2/g). Lower surface areas and pore
volume for GMC relative to AMC can be attributed to the fact that mesophase pitch
152
(a)
(b)
Figure 5.3
Nitrogen porosimetry of the mesoporous carbons and their pore size distributions in
the inset for GMC (A) and AMC (B).
153
Carbon Surface
Area [m2/g]
Total pore
volume [cc/g]
Pore size
(BJH) [nm]
La (XRD)
[Å]
Lc (XRD)
[Å]
d (XRD)
[Å]
AMC 978.5 1.7 9.5 37 7 3.57 GMC 337.4 0.43 7.1 55 16 3.42
Table 5.1
Properties of mesoporous carbons.
154
contains a considerable amount of large polyaromatic compounds, which after
carbonization, yield mesoporous carbon with little microporosity, as reported in the
literature.41 AMC, on the other hand, is synthesized from the furfural alcohol and
contains significant number of micropores, as indicated by the large uptake of
nitrogen at P/Po → 0, and thus has higher surface area. GMC has a predominant pore
size of 7.1 nm but also shows a slight hump at ~5 nm too while AMC has a pore size
of 9.5 nm. The large pores are facilitated by the lack of microporosity in the
mesoporous silica template which prevents formation of the interconnecting carbon
rods. Upon removal of the silica template, the enantiomeric carbon frameworks join
pair-wise leading to pore enlargement.30
XRD spectra of the carbons at each of the processing temperatures are shown
in Figure 5.4. The peak at ~25° corresponds to the graphitic <002> plane on carbon.
The narrower taller <002> peak is shown for the GMC indicating a larger degree of
graphitization, relative to the broader diffuse peak for the AMC. The layer dimension
parallel to the basal place, La, and perpendicular, Lc, were calculated from the <10>
and <002> reflections respectively (Table 5.1), according to the relations (La =
1.84λ/Bcos(θ) and Lc = 0.89λ/Bcos(θ),42,43 where λ is the wavelength of the X-rays
(1.54Å), B is is the full angular width at half max, and θ is the Bragg angle. The
increased layer dimensions (La and Lc) for GMC relative to AMC indicate a greater
degree of graphitization; there is longer range ordering in GMC due to inherent
graphitic character of the mesophase pitch used as precursor for synthesis of GMC .
The interplanar d-spacings calculated from the <002> reflection using Bragg’s law
(nλ = 2dsin(θ)) are shown in Table 5.1. There is a decrease in d-spacing for GMC
relative to AMC, indicating a tighter packing of the carbon planes, again consistent
with an increase in degree of graphitization.43
The loading of the Pt nanocrystals in mesoporous carbons is given in Table
5.2. For AMC, loadings of 23 wt % are attained after Pt dispersion at concentration of
4 mg/mL is infused in the support for 2 days. The final concentration after infusion is
155
Figure 5.4
XRD of the mesoporous carbon samples without added Pt.
0
200
400
600
800
1000
1200
1400
1600
1800
2000
0 20 40 60 80 100
AMC
GMC
2θ
Inte
nsity
002
0
200
400
600
800
1000
1200
1400
1600
1800
2000
0 20 40 60 80 100
AMC
GMC
2θ
Inte
nsity
0
200
400
600
800
1000
1200
1400
1600
1800
2000
0 20 40 60 80 100
AMC
GMC
2θ
Inte
nsity
002
156
Support Time of infusion (days)
Initial Conc
(mg/ml)
Final Conc.
(mg/ml)
Loading† (wt %)
Loading‡ (wt %)
AMC 2 4 <0.1 23±1 25 GMC 2 4 2.5 10±1 - GMC 3 10 6.8 20±3 22
* All infusions were done on 30 mg carbon and nanoparticles were dispersed in 3 mL of solvents † Loading calculated by UV-Vis ‡ Loading by TGA
Table 5.2
Loadings for infusion of Pt nanoparticles in carbon supports
157
<0.1 mg/ml and is within the experimental error, indicating almost all the
nanocrystals are infused in the mesopores. For GMC, 10 wt % loadings are achieved
under identical conditions of infusion and are lower than AMC. The smaller pore size
and surface area of the GMC relative to AMC may lead to lower loadings. However,
20 wt % loadings are achieved by increasing the starting concentration of Pt
nanocrystals to 10 mg/mL. The final concentration of nanoparticles after infusion was
6.8 mg/mL, indicating that 32% of the initial Pt nanoparticles are infused in the pores.
As-synthesized unsupported Pt nanoparticles showed an average size of
2.3±0.3 nm. Figure 5.5 shows TEM images of the Pt nanoparticles supported on
mesoporous carbons. Pt nanoparticles are highly dispersed on GMC and do not show
any sintering or aggregation upon thermal treatment for ligand removal, as indicated
by the similar histograms (Figure 5.5d, e) for the nanoparticle sizes before and after
thermal calcination. For GMC, the average Pt nanoparticle size for as-synthesized
nanocomposites is 3.05±0.42 nm which upon thermal calcination at 300 °C changes
to 3.37±0.65 nm. The slight increase in the particle size for uncalcined supported Pt
can be attributed to the low image contrast between the Pt and the carbon, which
makes the particles appear larger.23 The marginal increase in the nanoparticle size
during thermal calcination indicates strong Pt interactions with graphitic GMC. For
AMC, the ligands are removed at 500 °C and the average nanoparticle size is
3.55±0.65 nm after annealing, indicating little sintering. Thus the average Pt size on
AMC and GMC after calcination was comparable.
Figure 5.6(a) shows the TGA curves for blank AMC and GMC as well these
carbons with 20% Pt infused into the pores. Each curves exhibits four distinctive
regions; the stable region with no weight loss at low temperatures, a transition region
with gradual weight loss, a steep weight loss region, and a plateau after all the carbon
is oxidized. Figure 5.6(b) shows the derivative of Figure 5.6(a), that is differential
thermal gravimetry (DTG). The onset of carbon oxidation for Pt-AMC occurs at
relatively low temperature of ~320 °C (as shown in inset of 6b), and is slightly better
than the value of ~250 °C for 20% Pt on Vulcan XC-72.44 GMC on the other hand
158
Figure 5.5
TEM micrographs of 20% Pt on (A) GMC uncalcined and (B) GMC calcined at 300 °C, (C) AMC calcined at 500 °C. (D), (E), (F) Histograms for (A), (B) and (C), respectively. More than 50 nanoparticles were counted in each case.
0
5
10
15
20
25
2 2.4 2.8 3.2 3.6 4 4.4 4.8
Cou
nt
Pt nanoparticle size (nm)
0
5
10
15
20
25
1.5 2.2 2.9 3.6 4.3 5
Cou
nt
Pt nanoparticle size (nm)
0
5
10
15
20
25
1.5 2.2 2.9 3.6 4.3 5 5.7 6.4
Cou
nt
Pt Np Size (nm)
A B
C D
E F
159
Figure 5.6
(A) TGA for the pure (blank) carbons and 20% Pt infused mesoporous carbons AMC
and GMC calcined at 500 and 300 °C in N2/H2 atmosphere respectively and (B)
Differential thermal gravimetry (DTG) for the Pt on AMC and GMC. Inset shows the
magnified section of (B) where onset of oxidation occurs for AMC.
0
20
40
60
80
100
200 300 400 500 600 700 800Temp (C)
Wt %
Pt-AMCPt-GMCBlank AMC
Blank GMC
A
B
160
does not show any carbon degradation until a much higher temperature of 430 °C.
The AMC losses most of the carbon steeply at 400 °C (Figure 5.6(b)), which is lower
than the onset oxidation temperature for GMC. GMC looses most of the weight in
two phases at temperatures of 550 and 700 °C. The final stable weight for the AMC
and GMC is 23 and 21 wt % respectively and matches well with that determined from
the UV-Vis spectroscopy (Table 5.2). The onset temperatures for the blank AMC and
GMC are 420 °C and 510 °C respectively, significantly higher than corresponding
carbons with Pt. Similarly, blank GMC and AMC has a steep weight loss at 590 and
630 °C respectively, again higher than with carbons with Pt.
The relative stabilities of the mesoporous carbons and Vulcan XC-72 towards
accelerated thermal oxidation in presence of Pt at elevated temperature of 195 °C is
reported over an extended period of 16 days in Table 5.3. Vulcan XC-72 showed a
significant loss of 8.84 wt % due to carbon oxidation to CO/CO2, consistent with the
previously reported results.40 In contrast, the weight losses for Pt-AMC and GMC
were 2.95 and 0.71 wt %, respectively, much lower than Vulcan carbon. The values
were negligible for the three carbons without Pt, indicating Pt catalyzes the oxidation
as shown previously.40
Figure 5.7 gives the XRD spectra of Pt on mesoporous carbons before and
after ADT cycling. Pt crystalline peaks observed at 2θ= 39.8°, 46.5° and 67.8° and
81.6° may be assigned to the (111), (200), (220) and (311) orientations for an FCC
crystal phase. The nanocrystal size was determined by the Scherrer equation using the
(111) peak. For GMC, the size before ADT was determined to be 3.2 nm in good
agreement with the average size of 3.4 nm determined from TEM. Upon ADT
cycling, the Pt diameter increased slightly to 3.6 nm, an increase of ~10%. All of the
peaks for the orientations in the FCC crystal phase remained. For AMC, upon ADT
cycling the peaks become sharper and narrower, indicating significant increase in Pt
nanoparticle size. An additional peak at 86.2° corresponding to (222) plane appears.
The nanoparticle size for AMC increased from 3.6 nm to 5.6 nm upon ADT, an
increase of over 55%, significantly larger than in the case of GMC.
161
400
600
800
1000
1200
1400
1600
1800
30 40 50 60 70 80 90
before ADTafter ADT
Inte
nsity
2 theta
111
200220
311
(a)
(b)
Figure 5.7
XRD spectra of 20% Pt on (a) GMC and (b) AMC, before and after ADT
(for conditions in Table 5.4). Calcinations were done in 7% H2-93% N2 environment.
0
2000
4000
6000
8000
1 104
30 40 50 60 70 80 90
Before ADTPost ADT
Inte
nsity
2 Theta
111
200
220 311
111
111200
220 311
222
162
Type of Carbon % Loss of Carbon
20% Pt - Vulcan XC-72 8.84
20% Pt – A MC 2.95
20% Pt - GMC 0.71
Blank Vulcan XC-72 <0.5
Blank AMC <0.5
Blank GMC <0.5
Table 5.3
Weight loss of the carbon supports at 195 °C air for 16 days due to carbon oxidation
catalyzed by platinum.
163
Support
Calcination
Temp (°C)
Dxrd†
(nm) CSAxrd
†
(m2/g)
I † (A/mg
Pt)
I † (µA/cm2
)
ECSA†
(m2/g)Dxrd
‡
(nm) CSAxrd
‡
(m2/g)
I ‡ (A/mg
Pt)
I ‡ (µA/cm
2)
ECSA‡ (m2/g)
GMC 300 3.2 87 0.14 208 70 3.6 79 0.14 203 69
GMC 350 - - 0.13 170 74 - - 0.135 171 76
GMC 400 - - 0.1 155 62 - - 0.12 149 71
GMC Uncalcined - - 0.042 82 52 - - 0.046 84 54
AMC 500 3.3 84 0.14 218 65 5.6 51 0.095 202 46
VC-XC 721 - 3 93 0.11 200 69 5.382 52 0.035 ~180 21
Table 5.4
Electrochemical Data for 20% Pt on mesoporous carbons before and after ADT
potential cycling between 0.5 and 1.2 V at 50 mV/s for 1000 cycles in 0.1 M HClO4
saturated with argon.
164
XPS in Figure 5.8a for AMC and 8b for GMC show the C 1s spectra before and after
accelerated durability testing, whereas the Pt 4f spectra are shown in Figures 5.8c for
AMC and 5.8d for GMC. The C 1s spectra for GMC does not show any change after
potential cycling. However, for AMC, two new pronounced peaks appear at higher
binding energies relative to the original peak at 284.4 eV. The peak appearing at
288.5 eV can be attributed to the formation of COOH groups on the surface due to
carbon oxidation.45 The O/C ratio increases by 100% upon potential cycling for AMC
compared to only 25% for GMC. The binding energy for the Pt 4f7/2 peak for GMC
(72.1) is larger than for both unsupported Pt (71.0 eV) and Pt on AMC (71.2).46,47
For both AMC and GMC the Pt 4f7/2 peak does not change significantly upon ADT
cycling.
The electrochemical stabilities of the Pt-GMC and AMC catalysts were
studied for ADT cycling between 0.5 and 1.2V. Figures 5.9a (GMC), and 5.9b
(AMC) give CVs of the Pt catalyst before and after ADT potential cycling. The
ECSA of the platinum nanocrystals was estimated from the area of the anodic peak
corresponding to H-desorption, QH, appearing between 0.04 and 0.4 V
by13
cvmQECSAPt
H
..=
where mpt is the mass of platinum on the electrode, ν is scan rate (V/s) and c is a
constant defined by the charge required to oxidize a monolayer of hydrogen on Pt
(0.21 mC/cm2). The ECSA for 20% Pt on GMC calcined at 300 °C was 70 m2/g and
was comparable to that of commercial Pt catalysts (65-68 m2/g).1 A modestly smaller
value of 52 m2/g was observed for the uncalcined Pt-GMC (Table 5.4). Thus, this
modest decrease in the ECSA indicates that the dodecylamine ligands present on the
Pt nanocrystal surface block only a small fraction of the active Pt sites. For
calcination at 400 °C, the ECSA of 62 m2/g is slightly lower, perhaps indicating a
small degree coalescence of nanoparticles. The ECSA of 20% Pt nanoparticles
supported on AMC calcined at 500 °C was 65 m2/g, comparable to that of 20% Pt on
GMC calcined at 300 °C.
165
Figure 5.8 XPS spectra before and after ADT for 20% Pt supported on AMC and GMC. (A),(B): C1s region for AMC and GMC, respectively. (C),(D): Pt4f region for AMC and GMC, respectively.
85 80 75 70 65
Nor
mal
ized
Inte
nsity
Binding Energy (eV)
GMC GMC post ADT
85 80 75 70 65
Nor
mal
ized
Inte
nsity
Binding Energy (eV)
AMC AMC post ADT
294 292 290 288 286 284 282 280 278 276
Nor
mal
ized
Inte
nsity
Binding Energy (eV)
AMC AMC post ADT
294 292 290 288 286 284 282 280 278 276
Nor
mal
ized
Inte
nsity
Binding Energy (eV)
GMC GMC post ADT
A B
C D
166
Figure 5.9
Electrochemical studies for 20% Pt on mesoporous carbon before and after ADT
cycling for conditions in Table 5.4 (A) CV of GMC at 20 mV/s (B) CV of AMC at
20 mV/s (C) Linear Scan Voltammetry (LSV) of GMC at 5 mV/s rotated at 1600
rpm (D) LSV for AMC at same conditions
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
0 0.2 0.4 0.6 0.8 1 1.2 1.4
E (vs NHE)
I (m
A/c
m2
geo)
Before ADT cycling
After ADT cycling
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
0 0.2 0.4 0.6 0.8 1 1.2
E (vs NHE)
I (m
A/c
m2)
-6
-5
-4
-3
-2
-1
0
1
0.2 0.4 0.6 0.8 1 1.2
E (vs NHE)
I (m
A/c
m2
elec
trod
e)
Before ADT cycling
After ADT cycling
-7
-6
-5
-4
-3
-2
-1
0
1
0.2 0.4 0.6 0.8 1 1.2
E (vs NHE)
I (m
A/ c
m2
elec
trod
e)
Before ADT cycling
After ADT cycling
C D
167
After ADT cycling, the CV for Pt-GMC overlaps to the curve for the as-synthesized
catalyst, indicating negligible loss in stability with a constant ECSA, consistent with
the XRD results. The 20% Pt on AMC sample however loses substantial ECSA, with
a 30% drop as shown in Figure 5.9(b). Furthermore, 20% Pt on Vulcan-XC 72 shows
a loss of 70% ECSA under identical ADT cycling conditions.7 Thus, the trend in
stability for the catalysts can be established as Pt-GMC >> Pt-AMC > Pt-VulcanXC-
72.
Figure 5.9c (GMC and 9d (AMC) show the representative hydrodynamic
voltammogram obtained for the ORR before and after ADT cycling. The diffusion-
limited current densities at highly reducing potentials were 5.6 and 6 mA/cm2 for Pt
supported on GMC and AMC, respectively, both before and after ADT. These current
densities are comparable to the 6 mA/cm2 for a smooth polycrystalline Pt-disk
electrode, demonstrating negligible mass-transport effects. For Pt-GMC, the
hydrodynamic voltammograms after ADT overlaps with the before ADT, indicating
no change in catalyst activity, consistent with the results of ECSA as shown in Figure
5.9a, b. However, for Pt-AMC the hydrodynamic voltammograms shifts to lower
potentials in the kinetic region between 0.6-1.0 V after ADT relative to as-
synthesized, indicative of activity losses.
Figure 5.10 shows the Tafel slopes of mass activity and specific area activities
for the Pt supported on GMC and AMC, in the kinetically controlled regime, before
and after 0.5 – 1.2 V potential cycling. The activities are reported in Table 5.4 for a
potential of 0.9V. The mass activities for both the Pt-GMC calcined at 300 °C and
calcined Pt-AMC before cycling are 0.14 A/mg Pt, about 30% higher than that of
commercial 20% Pt-Vulcan carbon catalyst.1 Thus the degree of graphitization of the
support did not appear to influence the initial mass activity. Similarly, the specific
area activities for both Pt-GMC and AMC catalysts before cycling were ~200 µA/cm2
comparable to the value for commercial Pt-Vulcan XC-72 catalyst. The mass and area
activities for Pt-GMC before ADT testing decreased modestly with a calcination
temperature of 400 °C, consistent with a decrease in ECSA. For each of the calcined
168
0.84
0.86
0.88
0.9
0.92
0.94
0.96
0.01 0.1 1
Mass activity (A/ mg Pt)
E (v
s N
HE)
before stability
After stability
0.84
0.86
0.88
0.9
0.92
0.94
0.96
10 100 1000Specific activity (µ
A/ cm2 Pt)
E (v
s N
HE)
before stability
After stability
(A) (B)
0.84
0.86
0.88
0.9
0.92
0.94
0.96
10 100 1000
Specific Current Activity (µA/cm2 Pt)
E (v
s N
HE)
Before cycling
After cyling
(C) (D)
Figure 5.10
Tafel slopes for Pt-GMC calcined at 300 °C (A) mass activity (B) specific activity
and Pt-AMC calcined at 500 °C (C) mass activity and (D) specific activity
0.84
0.86
0.88
0.9
0.92
0.94
0.96
0.01 0.1 1
Mass Current Activity (A/ mg Pt)
E (v
s N
HE)
Before ADT cycling
After ADT cycling
169
Pt-GMC catalysts, the catalyst activity and ECSA did not decrease after ADT cycling,
indicating extremely high stability relative to the AMC and Vulcan catalysts. For Pt-
AMC, a significant ~30% loss in mass activity occurred during ADT, which
correlates well with the 30% loss in ECSA (Figure 5.10). An even more substantial
loss in these properties was observed for the Vulcan carbon catalyst.
5.4 Discussion
5.4.1 Carbon Oxidation and its Effect on Electrocatalyst Stability
The XRD spectra in Figure 5.4 indicates that AMC is significantly more
amorphous (broad <002> peak) relative to GMC (narrow <002> peak). In lieu of this,
AMC is expected to have a significant number of surface oxygen groups (carbonyl,
carboxylate, ester-like and alcohol).6,25 The presence of Pt atoms on amorphous
carbon accelerates carbon oxidation, which is facilitated by the surface oxygen groups
acting as reactive centers. This is confirmed by carbon weight loss under isothermal
treatment at 195 °C for Pt-AMC and Pt-Vulcan XC-72 (Table 5.3).6,48 The weight
loss for Vulcan carbon is significantly higher than that for AMC, suggesting that
Vulcan carbon contains a significantly higher proportion of oxygen groups.49
Similarly, Pt catalyzed oxidation of AMC and Vulcan XC-72 initiated on surface
oxygen sites may occur when subjected to highly oxidative potentials (up to 1.2V) in
corrosive, acidic environments. Carbon oxidation leads to Pt nanoparticle isolation
and is possibly one of the several factors leading to the loss in ECSA during potential
cycling, as seen for AMC and Vulcan carbon. The oxidation of the AMC after
potential cycling between 0.5 and 1.2 V is reflected by the manifestation of peaks
within the 287-293 eV binding energy range for the C 1s XPS spectra (Figure 8).
These peaks are commonly referred to as the carbon “oxide shoulders”.19,50 The
increase in the atomic ratio of O/C for Pt/AMC is 1.6 times that of Pt/GMC,
indicating that the degree of oxidation of AMC is higher than that of GMC, which is
consistent with the previous results.19 As the degree of oxidation/degradation of the
170
carbon increases due to higher number of oxygen groups, the activity of the catalyst
decreases due to a loss of active sites from nanoparticle isolation, as was the case for
Vulcan carbon.
The support degradation and consequent activity loss is alleviated for GMC
due to its higher degree of graphitization, which is expected to reduce the amount of
surface oxygen species (defect sites) as displayed in Table 5.4.5 Graphitization of the
mesoporous carbon enhances the basal layer dimension, as determined from XRD,
and thus leads to the reduction of the number of edge sites. This reduces the number
of dangling bonds (located on the edge sites), which readily react to form surface
oxides, thus mitigating carbon oxidation by Pt.3 This is confirmed by the insignificant
weight loss of carbon, (<0.7 wt %) for 20% Pt on GMC held isothermally at 195 °C
(Figure 5.7). In addition, the onset temperature of oxidation for GMC (~430 °C) is
higher compared to AMC (~310 °C), further demonstrating the high corrosion
resistance of graphitic carbon towards oxidation.. The high thermal stability of
graphitized carbons is expected to lead to high electrochemical stability towards
potential cycling (Table 5.4). After potential cycling, no “oxide shoulder” peaks
between 287-293 eV are observed for GMC, due to inherent resistance to oxidation
afforded to the support from graphitization. Graphitized mesoporous carbons thus
remain intact and do not oxidize upon exposure at 1.2V, preventing the electrical
isolation of the Pt nanoparticles and maintaining stable activities.
5.4.2 Carbon Synthesis and Stability by Pt-Carbon Interactions
The strong binding interaction between the Pt nanoparticles and GMC is
evidenced by the shift in the XPS Pt 4f7/2 peak to a higher binding energy (72.1 eV)
as compared to both unsupported Pt (71.0 eV) and Pt on amorphous AMC (71.2 eV)
(Figure 5.8).46,47 This shift in binding energy can be attributed to a change in the
coordination of the Pt due to the interaction of the metal with the carbon support.
While the exact nature of the Pt interaction with graphite is still a topic of
considerable discussion, it is reported that the delocalized electrons in the p-orbital of
171
the π sites on graphite overlap with the d-orbital of the Pt ([Xe] 4f14 5d9 6s1),
indicating that the nature of the binding is partially covalent.11 This occurs due to π-
backbonding, where the electron in the d-orbital of the metal is shared with a π*
antibonding orbital from the graphitic support. In addition, partial electron transfer
from the Pt to the graphitic surface has also been determined by previous XPS11 and
DFT51 studies. Thus, the strong partially covalent bond due to metal/support orbital
overlap (π backbonding) coupled with partial charge transfer between Pt and
graphitic surfaces, enables strong metal-support interactions and thus facilitates high
Pt loading.
The strategic selection of the weakly binding dodecylamine ligands with low
surface coverage for Pt stabilization (wt. of ligand on Pt was 25% as determined from
TGA) exposes sufficient bare metal to facilitate the strong metal-support
interactions.8,21,52 The low surface coverage of the metal by the ligand is reflected by
an ECSA of 52 m2/g for uncalcined Pt on GMC, which is 75% of the ECSA for Pt
calcined at 300 °C (70 m2/g Pt). All the ligands are expected to be removed at 300 °C
as dodecylamine starts oxidizing at 180-200 °C (as seen in TGA of pure Pt
nanoparticles). These results indicate that the ligand only covers ~25% of the Pt
active sites prior to calcination. The Pt-carbon interactions are further strengthened by
large interfacial contact area between the convex nanoparticles and the concave
mesopores of similar radii.21
The strong Pt interactions with graphitic carbon above, prevent the
coalescence of Pt nanoparticles during potential cycling, as seen by <10% change in
Pt size (ECSA and CSA) for GMC. These interactions restrain migration of the
nanoparticles on the carbon surface that would otherwise produce particle growth via
coalescence. The resistance to oxidative surface degradation of the graphitic carbon
helps keep the nanoparticles anchored to the surface and further prevents particle
migration. Finally, the high dispersion of the Pt on the high surface area mesoporous
supports lowers the tendency of Pt coalescence due to the large interparticle distances
172
and tortuous pathways in the 3D pores. All of these factors contribute to the
exceptional stability during potential cycling for the ORR reaction. In contrast, the
size increases significantly for Pt supported on AMC (>50%). For carbons with
significant amorphous character, the presence of oxygen complexes on the surface
withdraws electrons from the π sites, thus weakening the binding of the Pt to the
carbon surface.9 As the degree of graphitization increases, thereby raising the Pt-C
binding, sintering of metal nanoparticles has been shown to decrease by Coloma and
et al.9
During ADT, the stability in Pt size and ORR activity further indicate limited
dissolution of Pt from the metal surface (Table 5.4). At potentials above 0.7 V, Pt is
oxidized to PtO and forms multilayers of PtO at even higher oxidation potentials
(>1.0 V). PtO reacts with the hydrogen ions of the acidic electrolyte to form Pt2+ (PtO
+ 2H+ = Pt2+ + H2O). These Pt2+ ions can then migrate into the solution and may be
reduced during the reduction cycle of ADT to form Pt clusters which are not
supported on the carbon and hence do not participate in the ORR, thus causing loss in
ECSA and activity. Alternatively, Pt2+ can dissolve in the electrolyte and then
redeposit on other Pt nanoparticles causing 3D Ostwald ripening and consequent loss
in ECSA.
Given the stability in ECSA, CSA and ORR activity for the GMC, as
observed through XRD, Pt dissolution and Ostwald ripening appeared to be
insignificant, as a result of the resistance of the carbon to oxidation and strong metal-
support interactions. Although, the effect of metal-support interactions on the Pt
dissolution is not well studied, it is plausible that the C-Pt binding due to π
backbonding synergistically alters the electronic structure of the Pt metal, causing the
observed shift to higher binding energies for the Pt 4f7/2 XPS spectra for Pt on GMC,
thereby creating a stabilizing effect against Pt oxidation. The delocalization of
electron density in the orbital overlap between the Pt and the π sites on the carbon
lowers the tendency to form a Pt-O bond as the potential becomes more positive,
173
resulting in a decrease in Pt dissolution. A related method to alter the Pt electronic
structure to make it less resistant to dissolution is galvanic deposition of Au clusters
on Pt nanoparticles.53 For the more amorphous AMC carbon, the Pt-C interactions are
weaker (as shown by the binding energy in the Pt 4f7/2 XPS spectra as well as the
catalyst stability in ADT) whereby the weaker Pt- π site overlap makes the Pt more
susceptible to oxidation to PtO.
In Table 5.4 for Pt on Vulcan C, the loss is 70% in ECSA compared to only
55% in CSA. The higher loss for ECSA may be attributed to isolated nanoparticles
upon carbon corrosion that are no longer electrochemically active, as observed
previously.19 For AMC, the smaller and more similar losses in CSA (35%) and ECSA
(30%) indicate a smaller degree of nanoparticle isolation, consistent with the greater
resistance to C oxidation in Table 5.3, as well as the higher degree of graphitization,
which favors strong Pt-C interactions.
The lattice orientation of the graphene layers in GMC and AMC with respect
to the pore walls is difficult to characterize given the complexity of the bicontinuous
cubic mesostructure pore structure as reported previously.54 From the high stability
in the ORR activity and morphology of Pt during ADT, it appears that the graphene
layers were accessible in the pore walls to the Pt particles.
Traditional wetness impregnation techniques pose a variety of challenges in
the synthesis of the fuel cell catalysts with graphitic supports. The lack of polar
oxygenated sites on the graphitic surfaces precludes electrostatic interactions between
ionic metal precursors and the supports.6 Consequently, metal precursors nucleate and
grow to large nanoparticle sizes (>5nm) upon reduction limiting metal dispersion and
thus lowering ORR activities.6 To enhance the electrostatic interactions of the ionic
precursors with the supports, the graphitic carbons may be oxidized, for example,
with nitric acid to generate polar oxygen groups.16-19 Platinum nanoparticles with <3
nm size were synthesized on nitric acid treated graphitized carbon nanotubes (CNT)
by wetness impregnation and showed 15% loss in ECSA upon potential cycling
174
between 0.6 – 1.2 V for 1000 cycles.14 Though graphitization of the CNT partially
ameliorated the catalyst activity relative to Vulcan carbon, (which shows ~70% loss),
oxidation of the graphitized CNT generated corrosion prone sites on the surface and
may have led to the activity loss. The greater stability in ECSA for GMC in the
current study (essentially no loss) may be attributed to the greater dispersion and
separation of Pt particles in the 3D concave internal pores versus the external surface
of convex CNTs, as well as the smaller amount of oxygen species on the surface (no
nitric acid treatment) .
5.4.2 Control of Pt Nanoparticle Morphology for High Activity
Not only does the control of the Pt nanoparticle morphology in the pre-
synthesis technique offer a route to achieve high catalyst stabilities, but also high
activities. Thus, the lattice planes on the Pt particle surface may be designed in the
original arrested growth precipitation and maintained during infusion and calcination,
at relatively low temperatures (300 °C), as was shown by XRD. For ORR in HClO4
electrolyte, the (111) phase is more active than (110), which in turn is more active
than (100).1,55 In Figure 5.4, the intensity of the (111) peak is much stronger than the
(200) peak (which is associated with the (100) peak)55 indicating a higher fraction of
(111) over (100) facets (Figure 5.7). The FCC structure does not include a (110)
surface, further maximizing the (111) surface fraction. Furthermore, for the
polyhedral shape of the Pt particles according to TEM (Figure 5.5) the (111) surface
is dominant, consistent with the XRD results.55 Thus, the ability to design a high
fraction of highly active (111) sites in pre-synthesis, and to maintain these sites
through infusion and calcination, is an important strategy for raising catalyst activity.
This approach was also successful for the AMC catalyst. The activities are
comparable for AMC and GMC, indicating that metal-support interactions do not
appear to influence the catalytic behavior of platinum, consistent with previous
reports.9
175
5.5 Conclusions
Herein, we have designed highly stable Pt electrocatalysts for ORR on ordered
graphitized mesoporous carbon with controlled nanocrystal size and surface
morphology by infusion of pre-synthesized Pt nanoparticles. The loss in ORR catalyst
activity for Pt supported on graphitic mesoporous carbons under an accelerated
durability test of potential cycling between 0.5 to 1.2 V for 1000 cycles, is extremely
small, <2%, consistent with the lack of change in Pt size as observed by XRD, ECSA
and CSA. In comparison, the loss is 30% for Pt on amorphous mesoporous carbons
and 70% on commercial Pt on Vulcan carbon. The highly graphitic supports, with
relatively few surface oxygen sites, as shown by XRD and XPS, are highly resistant
to oxidation. This stability of the support minimizes nanoparticle isolation from the
surface and facilitates anchoring of the nanoparticles to the surface to inhibit
coalescence. The strong metal-support interactions from electron delocalization
between d-p orbitals of Pt and π sites and partial charge transfer from Pt to C as
observed by XPS51 and DFT11 calculations, mitigate nanoparticle migration and
coalescence, as well as Pt dissolution and subsequent Ostwald ripening. The high
dispersion of the Pt on the high surface area mesoporous supports with tortuous
pathways in the 3D pores, as well as the large interfacial contact area between the
metal nanocrystals and 7 nm concave mesopores with relatively similar radii further
lower the tendency of Pt coalescence. For the less graphitic AMC and Vulcan XC-72,
the greater losses activity results from carbon oxidation and weaker Pt-C interactions.
The arrested growth precipitation of Pt nanocrystals to form 3 nm Pt particles
with high concentrations of the highly active (111) surface leads to 25% higher
activities than typically seen for Pt formed on Vulcan carbon. The low surface
coverage of the weakly bound ligands favors strong Pt-C interactions8 as indicated by
high metal loadings of 20 wt % and ECSA of 52 m2/g. The active sites on the pre-
synthesized nanocrystals may be maintained during the facile removal of the weakly
bound ligands at low temperatures. In contrast, it is challenging to achieve high
176
loadings of Pt on highly graphitic carbon with low oxygenation by precursor
reduction, given weak adsorption of precursors on the carbon.3 The ability to control
catalyst size and surface morphology with pre-synthesized nanocrystals on highly
graphitic mesoporous carbons offers novel opportunities for designing highly active
and stable catalysts including electrocatalysts for oxygen reduction in PEMFCs.
5.6 References
(1) Gasteiger, H. A.; Kocha, S. S.; Sompalli, B.; Wagner, F. T. Activity
benchmarks and requirements for Pt, Pt-alloy, and non-Pt oxygen reduction catalysts
for PEMFCs. Applied Catalysis, B: Environmental 2005, 56, 9-35.
(2) Yu, X.; Ye, S. Recent advances in activity and durability enhancement
of Pt/C catalytic cathode in PEMFC. Journal of Power Sources 2007, 172, 133-144.
(3) Shao, Y.; Yin, G.; Gao, Y. Understanding and approaches for the
durability issues of Pt-based catalysts for PEM fuel cell. Journal of Power Sources
2007, 171, 558-566.
(4) Wagner, F. T.; Gasteiger, H. A.; Makharia, R.; Neyerlin, K. C.;
Thompson, E. L.; Yan, S. G. Catalyst development needs and pathways for
automotive PEM fuel cells. ECS Transactions 2006, 3, 19-29.
(5) Mathias, M. F.; Makharia, R.; Gasteiger, H. A.; Conley, J. J.; Fuller, T.
J.; Gittleman, C. I.; Kocha, S. S.; Miller, D. P.; Mittelsteadt, C. K.; Xie, T.; Yan, S.
G.; Yu, P. T. Two fuel cell cars in every garage? Electrochemical Society Interface
2005, 14, 24-35.
(6) Yu, X.; Ye, S. Recent advances in activity and durability enhancement
of Pt/C catalytic cathode in PEMFC. Journal of Power Sources 2007, 172, 145-154.
(7) Waje, M. M.; Li, W.; Chen, Z.; Larsen, P.; Yan, Y. Effect of scan
range on Pt surface area loss in potential cycling experiments. ECS Transactions
2007, 11, 1227-1233.
177
(8) Yu, R.; Song, H.; Zhang, X.-F.; Yang, P. Thermal Wetting of Platinum
Nanocrystals on Silica Surface. Journal of Physical Chemistry B 2005, 109, 6940-
6943.
(9) Coloma, F.; Sepulveda-Escribano, A.; Rodriquez-Reinoso, F. Heat-
treated carbon blacks as supports for platinum catalysts. Journal of Catalysis 1995,
154, 299-305.
(10) Antolini, E. Formation, microstructural characteristics and stability of
carbon supported platinum catalysts for low temperature fuel cells. Journal of
Materials Science 2003, 38, 2995-3005.
(11) Cuong, N. T.; Chi, D. H.; Kim, Y.-T.; Mitani, T. Structural and
Electronic Properties of Ptn (n=3, 7, 13) clusters on metallic single wall carbon
nanotube. Physica Status Solidi 2006, 13, 3472-3475.
(12) Kongkanand, A.; Kuwabata, S.; Girishkumar, G.; Kamat, P. Single-
Wall Carbon Nanotubes Supported Platinum Nanoparticles with Improved
Electrocatalytic Activity for Oxygen Reduction Reaction. Langmuir 2006, 22, 2392-
2396.
(13) Chen, Z.; Deng, W.; Wang, X.; Yan, Y. Durability and activity study
of single-walled, double-walled and multi-walled carbon nanotubes supported Pt
catalyst for PEMFCs. ECS Transactions 2007, 11, 1289-1299.
(14) Wang, J.; Yin, G.; Shao, Y.; Wang, Z.; Gao, Y. Investigation of
Further Improvement of Platinum Catalyst Durability with Highly Graphitized
Carbon Nanotubes Support. Journal of Physical Chemistry C 2008, 112, 5784-5789.
(15) Wakayama, H.; Setoyama, N.; Fukushima, Y. Size-Controlled
Synthesis and Catalytic Performance of Pt Nanoparticles in Micro- and Mesoporous
Silica Prepared Using Supercritical Solvents. Adv. Mater. 2003, 15, 742-745.
(16) Rajalakshmi, N.; Ryu, H.; Shaijumon, M. M.; Ramaprabhu, S.
Performance of polymer electrolyte membrane fuel cells with carbon nanotubes as
oxygen reduction catalyst support material. Journal of Power Sources 2005, 140,
250-257.
178
(17) Matsumoto, T.; Komatsu, T.; Nakano, H.; Arai, K.; Nagashima, Y.;
Yoo, E.; Yamazaki, T.; Kijima, M.; Shimizu, H.; Takasawa, Y.; Nakamura, J.
Efficient usage of highly dispersed Pt on carbon nanotubes for electrode catalysts of
polymer electrolyte fuel cells. Catalysis Today 2004, 90, 277-281.
(18) Li, X.; Hsing, I. M. Electrooxidation of formic acid on carbon
supported PtxPd1-x (x = 0-1) nanocatalysts. Electrochimica Acta 2006, 51, 3477-
3483.
(19) Shao, Y.; Yin, G.; Gao, Y.; Shi, P. Durability study of Pt/C and
Pt/CNTs catalysts under simulated PEM fuel cell conditions. Journal of the
Electrochemical Society 2006, 153, A1093-A1097.
(20) Rioux, R. M.; Song, H.; Hoefelmeyer, J. D.; Yang, P.; Somorjai, G. A.
High-Surface-Area Catalyst Design: Synthesis, Characterization, and Reaction
Studies of Platinum Nanoparticles in Mesoporous SBA-15 Silica. Journal of Physical
Chemistry B 2005, 109, 2192-2202.
(21) Gupta, G.; Patel, M. N.; Ferrer, D.; Heitsch, A. T.; Korgel, B. A.;
Johnston, K. P. Stable Ordered FePt Mesoporous Silica Catalysts with High Loadings
Chemistry of Materials 2008.
(22) Gupta, G.; Shah, P. S.; Zhang, X.; Saunders, A. E.; Korgel, B. A.;
Johnston, K. P. Enhanced Infusion of Gold Nanocrystals into Mesoporous Silica with
Supercritical Carbon Dioxide. Chemistry of Materials 2005, 17, 6728-6738.
(23) Gupta, G.; Stowell, C. A.; Patel, M. N.; Gao, X.; Yacaman, M. J.;
Korgel, B. A.; Johnston, K. P. Infusion of Pre-synthesized Iridium Nanocrystals into
Mesoporous Silica for High Catalyst Activity. Chemistry of Materials 2006, 18,
6239-6249.
(24) Ding, J.; Chan, K.-Y.; Ren, J.; Xiao, F.-s. Platinum and platinum-
ruthenium nanoparticles supported on ordered mesoporous carbon and their
electrocatalytic performance for fuel cell reactions. Electrochimica Acta 2005, 50,
3131-3141.
179
(25) Vijayaraghavan, G.; Stevenson, K. J. Synergistic Assembly of
Dendrimer-Templated Platinum Catalysts on Nitrogen-Doped Carbon Nanotube
Electrodes for Oxygen Reduction. Langmuir 2007, 23, 5279-5282.
(26) Liu, Z.; Koh, S.; Yu, C.; Strasser, P. Synthesis, dealloying, and ORR-
electrocatalysis of PDDA-stabilized Cu-rich Pt alloy nanoparticles. Journal of the
Electrochemical Society 2007, 154, B1192-B1199.
(27) Tian, Z. Q.; Jiang, S. P.; Liu, Z.; Li, L. Polyelectrolyte-stabilized Pt
nanoparticles as new electrocatalysts for low temperature fuel cells. Electrochemistry
Communications 2007, 9, 1613-1618.
(28) Xiong, L.; Manthiram, A. Nanostructured Pt-M/C (M = Fe and Co)
catalysts prepared by a microemulsion method for oxygen reduction in proton
exchange membrane fuel cells. Electrochimica Acta 2005, 50, 2323-2329.
(29) Raghuveer, V.; Ferreira, P. J.; Manthiram, A. Comparison of Pd-Co-
Au electrocatalysts prepared by conventional borohydride and microemulsion
methods for oxygen reduction in fuel cells. Electrochemistry Communications 2006,
8, 807-814.
(30) Kim, T.-W.; Solovyov, L. A. Synthesis and characterization of large-
pore ordered mesoporous carbons using gyroidal silica template. Journal of Materials
Chemistry 2006, 16, 1445-1455.
(31) Joo, S. H.; Choi, S. J.; Oh, I.; Kwak, J.; Liu, Z.; Terasaki, O.; Ryoo, R.
Ordered nanoporous arrays of carbon supporting high dispersions of platinum
nanoparticles. [Erratum to document cited in CA135:184349]. Nature (London,
United Kingdom) 2001, 414, 470.
(32) Choi, W. C.; Woo, S. I.; Jeon, M. K.; Sohn, J. M.; Kim, M. R.; Jeon,
H. J. Platinum nanoclusters studded in the microporous nanowalls of ordered
mesoporous carbon. Advanced Materials (Weinheim, Germany) 2005, 17, 446-451.
(33) Nam, J.-H.; Jang, Y.-Y.; Kwon, Y.-U.; Nam, J.-D. Direct methanol
fuel cell Pt-carbon catalysts by using SBA-15 nanoporous templates.
Electrochemistry Communications 2004, 6, 737-741.
180
(34) Liu, S.-H.; Lu, R.-F.; Huang, S.-J.; Lo, A.-Y.; Chien, S.-H.; Liu, S.-B.
Controlled synthesis of highly dispersed platinum nanoparticles in ordered
mesoporous carbons. Chemical Communications (Cambridge, United Kingdom)
2006, 3435-3437.
(35) Wikander, K.; Ekstroem, H.; Palmqvist, A. E. C.; Lundblad, A.;
Holmberg, K.; Lindbergh, G. Alternative catalysts and carbon support material for
PEMFC. Fuel Cells (Weinheim, Germany) 2006, 6, 21-25.
(36) Wikander, K.; Petit, C.; Holmberg, K.; Pileni, M.-P. Size Control and
Growth Process of Alkylamine-Stabilized Platinum Nanocrystals: A Comparison
between the Phase Transfer and Reverse Micelles Methods. Langmuir 2006, 22,
4863-4868.
(37) Kim, T.-W.; Kleitz, F.; Paul, B.; Ryoo, R. MCM-48-like Large
Mesoporous Silicas with Tailored Pore Structure: Facile Synthesis Domain in a
Ternary Triblock Copolymer-Butanol-Water System. Journal of the American
Chemical Society 2005, 127, 7601-7610.
(38) Ryoo, R.; Joo, S. H.; Jun, S. Synthesis of Highly Ordered Carbon
Molecular Sieves via Template-Mediated Structural Transformation. Journal of
Physical Chemistry B 1999, 103, 7743-7746.
(39) Kim, T.-W.; Park, I.-S.; Ryoo, R. A synthetic route to ordered
mesoporous carbon materials with graphitic pore walls. Angewandte Chemie,
International Edition 2003, 42, 4375-4379.
(40) Stevens, D. A.; Dahn, J. R. Thermal degradation of the support in
carbon-supported platinum electrocatalysts for PEM fuel cells. Carbon 2004, 43, 179-
188.
(41) Gierszal, K. P.; Kim, T.-W.; Ryoo, R.; Jaroniec, M. Adsorption and
Structural Properties of Ordered Mesoporous Carbons Synthesized by Using Various
Carbon Precursors and Ordered Siliceous P6mm and Ia.hivin.3d Mesostructures as
Templates. Journal of Physical Chemistry B 2005, 109, 23263-23268.
181
(42) Tian, J. M.; Wang, F. B.; Shan, Z. H. Q.; Wang, R. J.; Zhang, J. Y.
Effect of preparation conditions of Pt/C catalysts on oxygen electrode performance in
proton exchange membrane fuel cells. Journal of Applied Electrochemistry 2004, 34,
461-467.
(43) Kinoshita, K. Carbon: Electrochemical and Physicochemical
Properties; John Wiey and Sons, 1988.
(44) Baturina, O. A.; Aubuchon, S. R.; Wynne, K. J. Thermal Stability in
Air of Pt/C Catalysts and PEM Fuel Cell Catalyst Layers. Chemistry of Materials
2006, 18, 1498-1504.
(45) Darmstadt, H.; Roy, C.; Kaliaguine, S.; Choi, S. J.; Ryoo, R. Surface
Chemistry of Ordered Mesoporous Carbons. Carbon 2002, 40, 2673-2683.
(46) Fu, X.; Wang, Y.; Wu, N.; Gui, L.; Tang, Y. Surface Modification of
Small Platinum Nanoclusters with Alkylamine and Alkylthiol: An XPS Study on the
Influence of Organic Ligands on the Pt 4f Binding Energies of Small Platinum
Nanoclusters. Journal of Colloid and Interface Science 2001, 243, 326-330.
(47) Antolini, E.; Giorgi, L.; Cardellini, F.; Passalacqua, E. Physical and
morphological characteristics and electrochemical behavior in PEM fuel cells of
PtRu/C catalysts. Journal of Solid State Electrochemistry 2001, 5, 131-140.
(48) Hodgkins, R. P.; Ahniyaz, A.; Parekh, K.; Belova, L. M.; Bergstroem,
L. Maghemite Nanocrystal Impregnation by Hydrophobic Surface Modification of
Mesoporous Silica. Langmuir 2007, 23, 8838-8844.
(49) Stevens, D. A.; Hicks, M. T.; Haugen, G. M.; Dahn, J. R. Ex Situ and
In Situ Stability Studies of PEMFC Catalysts. Journal of the Electrochemical Society
2005, 152, A2309-A2315.
(50) Kuo, T.-C.; McCreery, R. L. Surface Chemistry and Electron-Transfer
Kinetics of Hydrogen-Modified Glassy Carbon Electrodes. Analytical Chemistry
1999, 71, 1553-1560.
(51) Yang, D. Q.; Zhang, G. X.; Sacher, E.; Jose-Yacaman, M.; Elizondo,
N. Evidence of the Interaction of Evaporated Pt Nanoparticles with Variously Treated
182
Surfaces of Highly Oriented Pyrolytic Graphite. Journal of Physical Chemistry B
2006, 110, 8348-8356.
(52) Jana, N. R.; Peng, X. Single-Phase and Gram-Scale Routes toward
Nearly Monodisperse Au and Other Noble Metal Nanocrystals. Journal of the
American Chemical Society 2003, 125, 14280-14281.
(53) Zhang, J.; Sasaki, K.; Sutter, E.; Adzic, R. R. Stabilization of Platinum
Oxygen-Reduction Electrocatalysts Using Gold Clusters. Science (Washington, DC,
United States) 2007, 315, 220-222.
(54) Yang, H.; Yan, Y.; Liu, Y.; Zhang, F.; Zhang, R.; Meng, Y.; Li, M.;
Xie, S.; Tu, B.; Zhao, D. A Simple Melt Impregnation Method to Synthesize Ordered
Mesoporous Carbon and Carbon Nanofiber Bundles with Graphitized Structure from
Pitches. Journal of Physical Chemistry B 2004, 108, 17320-17328.
(55) Wang, C.; Daimon, H.; Onodera, T.; Koda, T.; Sun, S. A general
approach to the size- and shape-controlled synthesis of platinum nanoparticles and
their catalytic reduction of oxygen. Angewandte Chemie, International Edition 2008,
47, 3588-3591.
183
Chapter 6
Highly Stable and Active Electrocatalysts for Oxygen Reduction:
Pre-synthesized Bimetallic Nanocrystals on Graphitic Carbon
Extremely stable and highly active electrocatalysts have been prepared by the
infusion of pre-synthesized bimetallic nanocrystals with controlled size (<4 nm),
composition and alloy structure onto graphitic mesoporous carbon. The low coverage
of weakly bound oleic acid and oleylamine ligands on the nanocrystal surface
facilitates strong metal-support interactions, resulting in nanocrystal loadings up to
~15 wt % with high dispersion as indicated by XRD and TEM. Electrochemically
dealloying via leaching of Cu from the Pt20Cu80 nanoparticle surface generates
Pt85Cu15 with a Pt rich shell. The specific current for the oxygen reduction reaction
(ORR) reaches 0.46 A/mg Pt as a result of the strained Pt shell, relative to 0.11 A/mg
Pt for pure Pt nanocrystals on amorphous carbon. For Pt-Cu supported on graphitic
mesoporous carbons, the loss in ORR catalyst activity over 1000 cycles from 0.5 to
1.2 V was extremely small, <2%, compared to a 40% loss for identical Pt-Cu
nanoparticles on amorphous mesoporous carbons. A typical loss for Pt on a
commercial amorphous carbon is 70%. The high resistance to oxidation of
graphitized carbon and the strong interactions of the π electrons with Pt prevent metal
sintering, Pt dissolution, and Ostwald ripening, resulting in high catalyst stability. The
high activities and stabilities for the ORR approach the levels targeted for practical
proton exchange membrane fuel cells. The ability to design catalysts composed of
pre-synthesized bimetallic nanocrystals with controlled size, composition and surface
properties onto high surface area mesoporous supports including graphitic carbon
offers the potential for new classes of highly stable and active catalysts.
The contents of this chapter are being prepared for a manuscript.
184
6.1 Introduction
The commercialization of polymer electrolyte membrane fuel cells (PEMFC)
depends upon improvement in catalyst activities by about 4-5 times relative to pure Pt
for the oxygen reduction reaction.(ORR).1-5 The activity may be raised up to about
2.5 times that of pure Pt with platinum based bimetallic alloys, Pt-M (M=Co, Ni, Fe,
V, Ti).6-13 Recently, Strasser and co-workers synthesized novel Pt-Cu catalysts on
carbon with a pronounced activity enhancement of >4 time that of pure Pt
catalysts.14,15 The copper was electrochemically etched from the nanoparticle surface
to form a strained Pt shell,15-19 with a reduced Pt-Pt distance, which alters the rate of
oxygen chemisorption and coverage by the intermediate oxygen species, and
enhances the activity for the ORR.
The second major challenge is to increase the catalyst stability upon exposure
to potential cycles encountered, for example, in automotive applications.2,5,20-22 Here,
catalyst stability at 1.2 V is required for an accumulated life time of 100 h, with a
maximum allowable performance loss of <30 mV at 1.5A/cm2 (<1 µV/cycle target).5
Pt supported on traditional primarily amorphous carbons show significant losses at
1.2 V from carbon oxidation as well as Pt dissolution upon potential cycling. For
example, for Pt on Vulcan XC-72 carbon, a 70% loss in electrochemical area was
observed upon potential cycling between 0.5 and 1.2 V for 1000 cycles.23 In highly
acidic environments, polar sites on amorphous carbons are oxidized at potentials ≥1.2
V leading to metal nanoparticle isolation and activity loss.2,5 The oxidation also
weakens the Pt-support interactions resulting in Pt dissolution and metal sintering.24
Graphitized carbons are more resistant to oxidation than amorphous carbons since
they contain much fewer defect sites and polar species on the surface.21,25-27 In
addition, the abundant π electron sites on the surface interact strongly with platinum,
thus strengthening the metal-support interactions, as demonstrated by XPS and
electron-spin resonance (ESR).21,25,27 Strong metal-support interactions reduce the
sintering, and coalescence of Pt. Charge transfer from the graphite to Pt28, inhibits Pt
185
oxidation and subsequent dissolution. Thus, corrosion resistant graphitic supports
with strong Pt-support interactions are expected to enhance the stability for the ORR.
Metals such as Pt are not dispersed efficiently on graphitic carbon upon
reduction of ionic precursor by traditional techniques such as wetness impregnation
.21,26,29 These precursors interact too weakly with graphitized supports, which do not
have a sufficient number of polar sites. For bimetallic catalysts, multiple metal
precursors are reduced on the supports either simultaneously30,31 or sequentially14,32.
This often produces nanoparticles with high polydispersity in size and composition,
with multiple alloy phases.33 The lack of uniformity results from varying rates of
nucleation and growth for each metal precursor in solution and on the support, as well
as non-uniform interactions with the heterogeneous support.15,17 These interactions
will be particularly weak for graphitic carbon. It is critical to control the size as well
as composition of bimetallic nanoparticles to achieve high catalyst activities. For
example, Pt-Co catalysts show enhanced activities over pure Pt at compositions of
Pt3Co but no appreciable enhancement at other compositions.9
An emerging concept in catalyst design is to pre-synthesize metal nanocrystals
with controlled size and composition via arrested growth precipitation with stabilizing
ligands and to infuse these nanocrystals onto high surface area supports.34-40 The
arrested growth precipitation step provides much greater control over the nanocrystal
size, composition, alloy structure than traditional techniques where reduction in the
presence of the support complicates nucleation, growth and the binding of stabilizing
ligands. For example, highly stable FePt catalysts with a controlled size, composition
and intermetallic alloy structure have been synthesized by pre-synthesizing FePt
nanocrystals and then infusing them into mesoporous silica.35 Weakly bound low
molecular weight ligands have been designed to stabilize the nanocrystals while
exposing significant area of the metal to the support for sufficient binding to achieve
loading up to 15 wt %.36,37 The weak binding facilitates ligand removal and catalyst
activation at low temperatures without perturbing the nanocrystal structure
significantly.
186
The primary objective of this study was to design highly stable Pt-Cu
catalysts on mesoporous graphitic carbons for the ORR, even for activities 4-fold the
typical values on Pt. Pt-Cu nanocrystals were pre-synthesized to controlled the
nanocrystal size, composition, and alloy structure, and then infused into the
mesoporous carbon to attempt to achieve high dispersion and nanocrystal stability.
We hypothesize that the high surface area and resistance to oxidation of the graphitic
carbon21 in addition to the strong binding of Pt21 as well as copper,41 will mitigate Pt-
Cu sintering and dissolution during potential cycling. The large interfacial area
between the pre-synthesized relatively spherical nanocrystals and the concave
mesopores further favors strong metal-support interactions. Weakly binding ligands
(oleic acid and oleylamine) were chosen to stabilize the Pt-Cu nanoparticles to expose
a significant fraction of the copper rich surface domains to the graphitic π sites
needed to promote adsorption and high loading of the nanocrystals. The low
molecular weight ligands were removed by calcination at relatively low temperature
to ameliorate changes in nanocrystal structure. The nanoparticles were subsequently
dealloyed electrochemically to selectively remove copper from the surface in order to
strain the platinum shell for the enhancement of the catalyst activity, as reported
previously.14,15 The degree of graphitization of the mesoporous carbons (without
metal) was characterized by XRD and Raman spectroscopy and the oxidative stability
by thermal gravimetric analysis. In addition, the size and composition of Pt-Cu
nanoparticles were characterized before and after electrochemical dealloying by
TEM, XRD and SEM-EDX to provide insight into the activity for the ORR reaction
and the stability under potential cycling between 0.5 and 1.2 V for 1000 cycles at 50
mV/s.
6.2 Experimental
All chemicals were used as received. Platinum (II) acetylacetonate (97%),
Copper bis(2,2,6,6-tetramethyl-3,5-heptanedionate), Oleylamine (70%), Oleic Acid
187
(99%) diphenyl ether (98% purity), perchloric acid 70% (99.999% purity), anhydrous
ethanol, and 5 wt % nafion solution in lower alcohols were purchased from Aldrich.
Toluene (99.9%) was obtained from Fisher Scientific, and ethanol (Absolute 200
proof) from Aaper alcohol. Polycarbonate filters (0.05 micron) were obtained from
Osmonics. High purity water (resistance ~18 MΩ) was used. Oxygen (research grade,
99.999% purity) and argon (research grade, 99.999% purity) were obtained from
Praxair.
Pt-Cu particles were synthesized by employing the modified polyol process
described by Sun et al.42 The synthesis was carried out in a 50 mL three-neck flask
under an argon atmosphere using standard Schlenk line techniques. For the
preparation of Pt20Cu80 nanoparticles, 0.16 mmol Platinum acetylacetonate and 0.50
mmol Copper bis(2,2,6,6-tetramethyl-3,5-heptanedionate) were separately dissolved
in 10mL of diphenyl ether and injected into the flask. The solution was continuously
stirred and heated to 80 °C using a heating mantle. At 80 °C, 0.5 mmol of the
stabilizing agents, oleic acid and oleyl amine, were injected. The temperature was
then raised to the reflux temperature of diphenyl ether (~260 °C). At around 215 °C a
sudden change in the reaction mixture was observed as the solution turned yellow and
then dark brown, indicating the reduction of the precursors. The reaction mixture was
maintained at 260 °C for 30 minutes and then cooled down to room temperature. The
dark dispersion was removed from the flask and the particles were isolated by the
addition of 50mL of acetone and centrifugation. The nanoparticles were further
cleaned by redispersing the Pt20Cu80 particles in 1mL of toluene and excess acetone
and centrifuging again. The other compositions of Pt-Cu were synthesized by
changing the molar ratios of the precursors.
Disordered mesoporous carbons (DMC) were prepared via self-assembly of
block copolymer and phenolic resin under acidic conditions as described in our
previous reports.44,45 Typically, 16.5 g of resorcinol and 16.5 g of F127
(EO106PO70EO106) were dissolved in 67.5 mL of EtOH and 67.5 mL of HCl aqueous
solution (3.0 M). To this solution, 19.5 g of formaldehyde (37 wt % in H2O) was
188
added with stirring at room temperature. As the polymerization reaction of resorcinol
and formaldehyde proceeded, self-assembly of in situ formed phenolic resin and F127
induced a phase separation. The polymer-rich gel phase was collected in the bottom
of centrifuge tubes by centrifugation at 9500 rpm for 4 min after the mixture was
stirred for 30 min. The gel was redissolved in a mixed solvent of 18 g of THF and 12
g of EtOH and the mixture was loaded on a substrate, dried at room temperature
overnight, and cured at 80 °C for 24h. Carbonization was carried out under N2
atmosphere at 400 °C for 2 h with a heating rate of 1 °C/min, which was followed by
further treatment at 850 °C for 3 h with a heating rate of 5 °C/min. To enhance the
degree of graphitization, the heat treatment of mesoporous carbons was carried out in
a high-temperature furnace (Thermal Technology Inc., CA) under helium atmosphere
at desired temperatures for 1 h with a heating ramp of 20 °C/min. The mesoporous
carbons heat treated at temperatures of 850 °C, 2000 °C and 2600 °C will be referred
to as DMC-850, DMC-2000 and DMC-2600 respectively for the rest of the article.
The conductivity of the DMC-2000 and DMC-2600 is ~300 S/cm, about 100 times
higher than that of Vulcan XC-72 carbon46 (3.0 S/cm). The conductivity of the less
graphitic DMC-850 is only 2.0 S/cm. The pores of the mesoporous carbons were
disordered in nature as peaks were not present in small angle X-ray scattering (not
shown).
The infusion of the Pt-Cu nanoparticles into mesoporous carbon was done by
a similar procedure as reported in Gupta et al.38 25 mg of mesoporous carbon was
mixed with 3 mL of a Pt-Cu nanocrystal dispersion of known concentration, ~20
mg/mL, in toluene. The contents were stirred for 4 days. The product was separated
from the nanocrystal supernatant by filtration with a 0.2 μm PTFE filter. The extent
of incorporation of Pt-Cu nanocrystals in the carbon was determined by subtracting
from the initial mass of Pt-Cu in toluene, the final mass recovered in the supernatant
after filtration.36 The masses were determined by absorbance of the Pt-Cu
nanocrystals with a Cary 500 UV-Vis-NIR spectrophotometer with an optical path
189
length of 1 cm (Appendix B, Figure B.2). A standard linear calibration curve was
generated with known concentrations of Pt-Cu nanocrystals in toluene for a
wavelength of 800 nm with correlation coefficient of 0.998 (Appendix B, Figure B.2).
Loadings of ~15% metal on mesoporous carbons were obtained after infusion (Table
6.2). The individual uncertainties in the loadings produced by factors of volume
measurement, weight of carbon, absorbance, and linearity of the calibration curve are
(+2%), (+1%), (+5%) and (+3%) respectively. The total uncertainty is calculated
conservatively by adding the individual uncertainties in loadings and is ~11%, as
explained in detail previously.37 To activate the Pt-Cu catalysts ligands were
removed by calcining the catalyst at 350 °C for 2h. The temperature was ramped at 16
°C/min. After calcination, the catalysts were washed with 100 mL of ethanol.
Wide angle X-ray diffraction was performed with samples on a quartz slide
using a Bruker-Nonius D8 Advance diffractometer. Samples were scanned for 12 h at
a scan rate of 12 deg/min with 0.02 deg increments. The average nanocrystal size was
estimated from the Scherrer equation using JADE software (by Molecular Diffraction
Inc).
Raman spectra were acquired using a Renishaw inVia Microscope using a
514.5 nm Argon laser at 50% power with a 50x aperture. Four acquisitions were
taken for each sample from 500 to 2000 cm-1 for 50 seconds. The spectra were fit
using the PeakFitTM software, following the convention set forth by Cuesta et. al47 of
fitting five Gaussians at 1624, 1583, 1487, 1351, and 1220 cm-1, corresponding to the
D’, G, D”, D, and I bands, respectively. All spectral fits had correlation factors (R2)
greater than 0.98.
A glassy carbon rotating disk electrode (0.196 cm2, from Pine Instruments)
was used as a substrate for the catalyst. It was polished with a 0.3µm followed by a
0.05 µm alumina suspension to give a mirror finish. 1.0 mg of Pt-Cu catalyst was
suspended by sonication in 300µL of anhydrous ethanol and 300 µL of 0.15 wt %
nafion solution (diluted from 5% nafion stock solution). A total of 10 µL of Pt-Cu
190
suspension was pipetted onto the substrate and was dried at room temperature for 15
min.
Electrochemical measurements were carried out using a three-compartment
electrochemical cell with a Pt wire counter electrode and a Hg/Hg2SO4 reference
electrode. Electrochemical measurements were recorded using CHI 832
electrochemical analyzer (CH instruments Inc., Austin). All potentials cited in this
article are normalized with respect to NHE.
After preparation of the catalyst on the substrate, the catalyst was dealloyed to
leach off copper by cycling the potential between 0.04 and 1.24 V at 200 mV/s for
500 cycles in argon saturated 0.1M HClO4 electrolyte.1 The area under the desorption
peak of underpotentially deposited hydrogen in a cyclic voltammogram (scanned at
100 mV/sec between 0.04 and 1.24V), was used to evaluate electrochemically active
surface area (ECSA) of the Pt catalyst on the electrode. Hydrodynamic
voltammograms at rotating disk electrode were recorded by anodically scanning the
potential from 0.04 to 1.24 V at 5 mV/s in oxygen saturated 0.1 M HClO4. The
oxygen pressure was 1 atm. Accelerated durability tests (ADT) were employed by
cycling the electrode potential between 0.5 and 1.2 V at 50 mV/s for 1000 cycles in
an argon saturated atmosphere for a total time of 7h.22
6.3 Results
6.3.1 Structure of the Pt-Cu on Mesoporous Carbon Catalysts
The surface area, pore volume and mean pore size for the three carbons
measured by nitrogen porosimetry (see supplementary information) is given in Table
6.1. As the calcination temperature for the carbons increases from 850 °C to 2600 °C,
both the surface areas and the pore volume decrease as expected. DMC-850 and
DMC-2000 show a surface area of 577 and 344 m2/g respectively, which are
191
Carbon Surface
Area (m2/g)
Pore volume
Pore size (nm)
La (Å)
Lc (Å)
d (Å)
DMC 850 577 0.70 8.1 33.8 16 3.81
DMC 2000 344 0.57 8.0 35.4 22 3.63
DMC 2600 262 0.39 7.9 43.2 30 3.35
Table 6.1
Properties of mesoporous carbon.
192
significantly higher than commercial Vulcan XC-72 (~250 m2/g) supports. The pore
size for all the three mesoporous carbons was ~8 nm and did not show any shrinkage
in the pore diameter or structural collapse with increasing temperature, manifesting
the inherent stability of the supports. The pores of the mesoporous carbons were
disordered in nature as peaks were not present in small angle X-ray scattering.
XRD spectra of the carbons at annealed at varied temperatures are shown in
Figure 6.1. The peak at ~25° corresponds to the graphitic <002> plane on carbon. A
sharp <002> peak is associated with a higher graphitic character as compared to a
broad diffuse peak seen in amorphous carbons.48 With increasing annealing
temperature, the <002> peak becomes sharper and narrower, indicating an increase in
the degree of graphitization. The layer dimension parallel to the basal place, La, and
perpendicular, Lc, were calculated from the <10> and <002> reflections respectively,
according the relations,48,49
La = 1.84λ/Bcos(θ) and
Lc = 0.89λ/Bcos(θ)
where λ is the wavelength of the x-rays (1.54A), B is is the full angular width at half
max, and θ is the Bragg angle (Table 6.1). The monotonic increase in the values of La
and Lc along with increasing temperature indicates progressively enhanced degree of
graphitization of carbon supports. Table 6.1 also lists the interplanar d-spacing
calculated from the <002> reflection using Bragg’s law (nλ = 2dsin(θ)). There is a
monotonic decrease in d-spacing with increasing temperature, indicating a tighter
packing of the carbon planes, again consistent with an increasing graphitization.48
The first order Raman spectra for DMC-850, 2000, and 2600 are shown in
Figure 6.2. There are two peaks apparent in the spectra for all three samples,
corresponding to the D band at 1355 cm-1 and the G band at 1583 cm-1. Increasing the
annealing temperature for the carbons enhances the overlap between the two bands,
making it necessary to deconvolute the FWHM values by fitting five bands at 1624,
1583, 1487, 1355, and 1220 cm-1, corresponding to the D’, G, D”, D, and I bands,
193
0
5000
1 104
1.5 104
2 104
2.5 104
0 20 40 60 80 100
DMC 850 °CDMC 2000 °CDMC 2600 °C
Inte
nsity
2 theta
002
100
004
Figure 6.1
XRD spectra of the disordered mesoporous carbons annealed at different
temperatures.
194
0
2000
4000
6000
8000
1 104
500 1000 1500 2000
DMC 850DMC 2000DMC 2600
Inte
nsity
Raman Shift (cm-1)
Figure 6.2
Raman spectra of the disordered mesoporous carbons annealed at different
temperatures.
195
respectively, following the convention of Cuesta et. al.47 The D band is often referred
to as the “disorder” band and is commonly believed to be a Raman inactive mode that
becomes active as the disorder in the system is increased, or likewise as the symmetry
at or near a crystalline edge is reduced.50,51 The G band is present in all sp2 hybridized
carbon materials and arises from the E2g vibrational mode. By determining the area
under the peak for the D band and the G band, ID/IG (ratio of amorphous to graphitic
character) were estimated and decreased monotonically as 1.30, 1.24 and 1.02 for
DMC-850, 2000 and 2600 respectively. The decreasing ID/IG with increasing carbon
temperature is indicative of improving degree of graphitization. The relative levels of
graphitization can be discerned by estimating the in plane crystalline length, La (=
4.4/(Id/Ig))52 in nanometers. As the ID/IG ratio decreases, La increases (La=3.38, 3.74,
4.32 nm for DMC 850, 2000 and 2600 respectively) as graphitic crystallites grew in
size with increasing carbon treatment temperature.
Two different compositions of Pt-Cu nanoparticles were synthesized by using
the molar precursor ratio of Pt:Cu ~24:76 and 50:50 respectively. After synthesis of
the nanoparticles, the atomic/molar compositions were determined to be Pt20Cu80 and
Pt45Cu55 respectively using SEM-EDX, and matches well with the starting precursor
molar ratio. For DMC-850 where the loading was 17% (Table 6.2), about ~25% of
total initial nanoparticles were infused in the mesopores. To enhance the loadings for
the more graphitic DMC-2000, the initial concentration of Pt20Cu80 was increased to
30 mg/mL and the time of infusion was raised to 5 days. For DMC-2600 with the
highest degree of graphitization, loading were limited for Pt20Cu80, but were higher
for Pt45Cu55 since Pt binds more strongly to graphitic C. In contrast, for amorphous
Vulcan XC-72, almost all the Pt20Cu80 nanoparticles were adsorbed in a relatively
short time span of 2 days and at a low starting concentration of 5 mg/mL, suggesting
that the polar sites on the amorphous carbon facilitated adsorption. However, the
desired loading could be obtained on the DMC-2000 with higher concentrations to
drive the thermodynamics and longer times.
196
Nanoparticles Support
Time of infusion (days)
Initial Conc
(mg/ml)
Final Conc.
(mg/ml)
Loading (wt %)
Pt20Cu80 DMC-
850 3 20 15.5 17
Pt20Cu80 DMC-2000 3 20 18 8
Pt20Cu80 DMC-2000 5 30 26 14
Pt20Cu80 DMC-2600 7 30 29 <5
Pt45Cu55 DMC-2600 7 30 26 14
Pt20Cu80 Vulcan-XC72 2 5 <0.1 19
Table 6.2
Infusion of Pt-Cu nanoparticles in mesoporous carbon supports. All infusions were
done on 50 mg carbon and nanoparticles were dispersed in 3 mL of toluene.
197
The Pt-Cu catalysts were electrochemically dealloyed to leach off copper from
the nanoparticle surface and produce a highly active strained Pt surface.1 Figures 6.3
(and Appendix B, Figure B.4) show voltammograms for Pt20Cu80/SDC-2000 catalyst
during dealloying cycles for slow potential cycling between 0.04 and 1.24 at 20
mV/s,. The first dealloying cycle does not show a characteristic anodic stripping peak
between 0.04 and 0.4V, associated with underpotentially deposited hydrogen on
platinum surfaces. Here, the surface of the Pt-Cu nanoparticles is primarily copper
and contains very little Pt. This result is consistent with the segregation of the copper
on the surface in Cu rich Pt-Cu alloys, given its low surface energy.53 A broad anodic
peak appeared at 0.3 V and extended almost to 0.85 V indicating dissolution of the
copper. After 3 dealloying cycles, two distinct copper dissolution peaks at 0.3 V and
0.7V emerge. With the progression of the cycles, the peak intensity and area of
copper dissolution peaks decreased as the surface became enriched with Pt, indicated
by the appearance and growth of an anodic hydrogen desorption peak between 0.04
and 0.4V. After about 500 cycles, the cyclic voltammograms reached a steady state or
equilibrium behavior with the signature of a pure Pt surface (Appendix B, Figure
B.5), with the familiar H2 ad/desorption range (0.05-0.4 V) as well as the Pt oxide
formation and desorption peaks at 0.83 and 0.72 V on the anodic and cathodic scan,
respectively.
The dealloying mechanism changes significantly upon progressive Cu
removal. After just 20 dealloying cycles, copper dissolution peaks at 0.3 and 0.7 V
essentially disappear, as the surface becomes composed of almost pure Pt. However,
the “strongly” and “weakly” bound hydrogen desorption peaks associated with Pt
<100> and <111> respectively, are not present, indicating an amorphous structure.54
After the completion of 500 dealloying cycles, these peaks become present at 0.16
and 0.21 V (Figure 6.3) indicating crystalline platinum.55 Thus, while the initial
dealloying cycles facilitate Cu removal from the surface, the later dealloying cycles
facilitate the structural reordering of the platinum atoms upon crystallization.
198
-1.20
-1.00
-0.80
-0.60
-0.40
-0.20
0.00
0.20
0.40
0.60
0.80
0.000 0.200 0.400 0.600 0.800 1.000 1.200
Potential vs NHE
Cur
rent
(mA
)/ cm
2 ge
o
cycle 1
cycle 20
cycle 500
cycle 2
Figure 6.3
Cyclic voltammetry of Pt20Cu80 nanoparticles supported on DMC-2000 during
electrochemical dealloying at 20 mV/s in argon atmosphere.
Cycle 500 Cycle 1
Cycle 2
Cycle 20
199
1000
1500
2000
2500
3000
20 30 40 50 60 70 80 90 100
before dealloying after dealloying
Inte
nsity
2 theta
111
311220
311
220200
111
(a) (b)
(C)
Figure 6.4
XRD spectra of the Pt20Cu80 nanocrystals supported on DMC-2000 before and after
electrochemical dealloying of copper (a); peak analysis of the XRD spectra of the as-
synthesized Pt20Cu80 nanoparticles (b); and electrochemically dealloyed nanoparticles
with a final composition of Pt85Cu15.
200
Figure 6.4(a) shows XRD spectra of the Pt20Cu80/DMC-2000 before and after
dealloying. Before dealloying, the <111> peak approaching that of pure copper is
present., The peaks at 43.1°, 69° and 83.1°, are associated with the <111>, <220> and
<311> crystal planes of FCC lattice. The phases of the nanocrystal were investigated
by peak fitting using JADE software, as shown in Figure 6.4(b) and Table 6.3. A
single-phase Cu rich alloy with a peak at 42.7° was present, along with a small Pt rich
phase as indicated by the peak at 40.1°. From the relative areas, the Cu rich phase was
~95% with only about ~5% of a Pt rich phase. The peak position can be used to
determine the phase compositions with Vegard’s Law.56,57 The composition of the
FCC-disordered Pt–Cu alloy can be estimated from observed unit cell parameters
according to ac = a0 − k. xCu where a0= 3.923 Å is the unit cell parameter of pure Pt,
ac is the measured cell parameter of a Pt–Cu alloy, xCu is the molar ratio of Cu in the
binary alloys, and k is a constant (~0.308 Å for the Pt-Cu system).58 The unit cell
parameter ac for the Pt-Cu alloy can be related to the lattice spacing for 111 peak and
2θ according to
)sin(2
3/
111
111
θλ CuPtc
CuPt
d
ad−
−
=
= (6.1)
The composition of the copper rich phase was Pt17Cu83, while that of the platinum
rich phase was ~Pt90Cu10. The composition of the copper rich phase agrees well with
that determined by SEM-EDX. The nanoparticle size was estimated to be 3.0 nm
from the Scherrer equation.
After dealloying, the <111> peak shifts to lower 2θ values relative to the
<111> peak before dealloying, as copper is leached from the nanoparticle surface
(Figure 6.4(a)). After dealloying four peaks appear at 40.54°, 46.74°, 68.4° and
81.26°, corresponding to the <111>, <200>, <220>, and <311> FCC lattice planes,
respectively. As shown in Figure 6.4(c), only one platinum-rich alloy phase was
present in the dealloyed nanoparticles. The composition was estimated to be Pt85Cu15
201
Catalyst DTEM (nm)
DXRD (nm)
CSATEM (m2/g)
SEM-EDX
Compo-sition
Bravais lattice
Lattice Parameter
(A)
Composition
(Vegard’s law)
Ratio of 111 peak areas
Before dealloying 3.25 3.0 86 Pt20Cu80
FCC1 FCC2
3.663 3.881
Pt17Cu83 Pt90Cu10
FCC1:FCC2= 19:1
After dealloying 2.95 2.7 95 Pt85Cu15 FCC 3.876 Pt85Cu15 -
Table 6.3
Composition and structural analysis of 14% Pt20Cu80/DMC-2000 before and after
dealloying.
202
using Vegard’s Law (Table 6.3). It was in good agreement with the bulk composition
of Pt82Cu18 determined from SEM-EDX. The nanocrystal size after dealloying from
the Scherrer equation was 2.8 nm indicating a 10% reduction after dealloying.
Figure 6.5 gives the TEM micrographs and the corresponding histograms of
Pt20Cu80 on DMC-2000 before and after electrochemical dealloying. The
nanoparticles are well dispersed on the support with little aggregation or sintering
upon ligand removal or during the dealloying process. The average nanoparticle size
increases slightly from 2.85 nm (Appendix B, Figure B.2) to 3.25 nm (Figure 6.5)
upon thermal calcination for ligand removal. It decreased from 3.25 nm to 2.95 nm
upon dealloying (Table 6.3) in good agreement with the sizes from XRD. The
chemical surface area (CSA) after dealloying was determined from the particle sizes
(CSA= 3/(ρPt.rnp))and was estimated to be ~95-100 m2/g by XRD and TEM (Table
6.3).
6.3.2 Activity and Stability of Pt-Cu Catalysts for ORR
The electrochemical surface area (ECSA=QH/mpt.ν.c) of the platinum enriched
nanocrystals after dealloying was estimated from the area of the anodic H2 desorption
peak between 0.04 and 0.4V,59 where QH is the charge for desorption, mpt is the mass
of platinum on the electrode, ν is scan rate (V/s) and c is a constant defined by the
charge required to oxidize a monolayer of hydrogen on Pt (0.21 mC/cm2). The
estimated ECSA of 83m2/g Pt was comparable or slightly better than that of
commercial Pt catalysts.23 The ECSA was ~85% of the chemical surface area
estimated by XRD and TEM (Table 6.3), indicating that 85% of the total Pt sites
participate in the ORR.
Figure 6.6 shows a representative linear scan voltammogram obtained for the
oxygen reduction reaction with dealloyed Pt-Cu/DMC-2000. The diffusion-limited
current densities were 6.3, 6.0 and 5.6 mA/cm2 for Pt20Cu80 / DMC-2000, Pt20Cu80 /
DMC-850 and Pt45Cu55 / DMC-2000, respectively. These results are similar to the
203
(A)
(B)
(C)
(D)
Figure 6.5
TEM micrographs and histograms of Pt20Cu80 on DMC-2000 (A,C) as-synthesized
and calcined at 350 °C (B,C) Electrochemically dealloyed. Average nanoparticle size
changed from 3.25 nm to 2.95 nm after dealloying.
0
2
4
6
8
10
12
14
2 2.6 3.2 3.8 4.4 5
Cou
nt
PtCu Nanoparticle size (nm)
0
2
4
6
8
10
12
1.75 2.25 2.75 3.25 3.75 4.25 4.75
Cou
nt
Nanoparticle Size (nm)
204
-7
-6
-5
-4
-3
-2
-1
0
1
0 0.2 0.4 0.6 0.8 1 1.2
14% Pt20Cu80 / DMC-200017% Pt20Cu80 / DMC-85014% Pt45Cu55 / DMC-2600
I ( m
A/c
m2 e
lect
rode
)
Potential (vs NHE)
Figure 6.6
Linear scan voltammetry of the dealloyed Pt-Cu nanoparticles supported on DMC at
5 mV/s with rotation rate of 1600 rpm.
205
value of 6 mA/cm2 for a smooth polycrystalline Pt-disk electrode, demonstrating
negligible mass-transport effects.
Figure 6.7 shows the Tafel slopes for the dealloyed Pt20Cu80 on DMC-2000
and DMC-850 catalysts, in the kinetically controlled regime relative to a commercial
platinum catalyst. Both the mass and surface activities for the dealloyed Pt20Cu80 are
significantly better than those of the commercial Pt catalyst, as seen in previous
studies with similar Pt-Cu composition.1,12 These values are also reported in Table 6.4
at 0.9V. The value of 0.45 A/mg Pt for dealloyed Pt20Cu80 on DMC-850 and DMC-
2000 was ~4 times better than the commercial Pt catalyst (~0.11 A/mg Pt). For these
two supports, mass activities did not change with the degree of graphitization.
Similarly, the area activities were more than three times greater than that of the
commercial Pt catalyst. The mass activity was only half as large for Pt45Cu55
supported on DMC-2600. For Pt75Cu25 nanoparticles supported on DMC-850, the
activity was 0.16 A/mg Pt, marginally higher than for pure Pt.
The activities of the catalysts were measured after ADT cycling (Figure 6.7
and Table 6.4). Both Pt20Cu80 / DMC-2000 and Pt45Cu55 / DMC-2600 did not show
any significant losses in mass activity or ECSA after ADT cycling, indicating high
catalyst stabilities for these highly graphitic supports. The loss was significant, about
40%, for Pt20Cu80 on the less graphitic DMC-850. However, the loss in ECSA was
only 15%. The loss was even greater, >50%, for Pt20Cu80 on Vulcan carbon after ADT
cycling. Pt20Cu80 supported on DMC-2600 did not show any loss in catalyst activity,
consistent with the high corrosion resistant of carbon.
6.4 Discussion
6.4.1 Carbon Corrosion and Impact on Stability
Pt-Cu supported on the DMC-850 support showed significant loss in activity
upon ADT potential cycling. DMC-850 is amorphous in nature as shown in Figure
6.1 and 6.2. It is oxidized when subjected to potentials ≥1.2 V in highly corrosive
206
Sample* I (A/mg Pt)
Before ADT†
I (µA/cm2) Before ADT†
ECSA before ADT
(m2/g Pt)
I (A/mg Pt)
after ADT†
I (µA/cm2) after ADT†
ECSA after ADT
(m2/g Pt)17% Pt20Cu80 /
DMC- 850 0.46 630 73 0.27 435 62
14% Pt20Cu80 / DMC - 2000 0.45±0.03 537±25 83±4 0.45±0.03 540±28 84±4
14% Pt45Cu55- DMC-2600 0.23 382 59 0.22 332 66
5% Pt20Cu80 / DMC- 2600 0.16 232 68 0.15 150 66
20% Pt20Cu80 / vulcan-XC72 0.42 495 68 0.20 198 38
* All samples have been dealloyed by potential cycling
† All activities are calculated at 0.9V
Table 6.4
Activity and stability of dealloyed Pt-Cu catalysts supported on mesoporous carbons.
207
(a)
(b)
Figure 6.7
Potential-Activity plots of dealloyed Pt20Cu80 nanoparticles supported on
DMC-2000 and DMC-850 compared to commercial Pt (ETEK) catalysts (a) Pt mass-
based activities and (b) Pt specific surface area activities.
208
acidic environments as reported previously for other amorphous carbons.21,28 The
surface oxygen species on the amorphous carbon acts as reactive centers and
accelerate carbon oxidation.26,27 Pt atoms on the carbon surface (after dealloying Pt-
Cu) catalyze the carbon oxidation.60 We found that 20% Pt on amorphous Vulcan
XC-72 lost about 9% of its weight under isothermal conditions of 195 °C in 16 days
due to carbon loss. Degradation of carbon leads to metal nanoparticle isolation and
partly contributes to activity losses during ADT potential cycling (Figure 6.7).
The support degradation and consequent activity loss were alleviated by
graphitization of the carbon support as seen for DMC-2000 and 2600 (Table 6.4).21
The high temperature graphitization reduces the number of surface oxygen on the
surface and thus increases resistance to oxidation.28 For example, the onset
temperature for 20% Pt on graphitized DMC-2000 is ~180 °C higher than for 20% Pt
on amorphous DMC-850, as seen by TGA (Appendix B, Figure B.7) indicating
significantly higher oxidation resistance. The weight loss of carbon for 20% platinum
on the more graphitic carbons under isothermal conditions at 195 °C for 16 days was
insignificant, (<0.7 wt %, which is within the error limit) also indicating high
oxidation resistance. Graphitized mesoporous carbons remain intact and do not
oxidize even at 1.2V, thus preventing the electrical isolation of the dealloyed Pt-Cu
nanoparticles as indicated by the insignificant loss in ECSA and ORR activity (Table
6.4).
6.4.2 Metal-Support Interactions: Catalyst Synthesis and Stability
The high metal loadings of Pt-Cu nanoparticles infused into the graphitized DMC-
2000 and DMC-2600 were facilitated by the short-ranged interactions between the
metal and the support, as shown previously for metals on silica supports.38 The
surface of the Pt20Cu80 nanoparticles is Cu-rich as observed from the initial dealloying
curves (Figure 6.3). Relatively strong interactions between Cu and graphitic fullerene
due to charge transfer from Cu to fullerene have been identified by means of a high
frequency Raman spectra.42 The peak width of the pentagon-pinch mode becomes
209
wider and the peak position shifts to the lower frequency as the atomic ratio of Cu/C
increases, indicating strong metal-support interactions. These short-ranged
interactions enable the adsorption of Cu-rich Pt-Cu nanoparticles on the mesoporous
carbon supports to render high loadings (Table 6.2).
A key challenge to adsorb high levels of nanoparticles onto carbon is to
overcome screening of the short-ranged interactions by the hydrocarbon capping
ligands. The strategic selection of the weakly binding oleic acid/oleyl amine ligands
with low surface coverage (wt of ligand on Pt-Cu was 20% as determined from
TGA), exposes enough bare metal to facilitate strong metal (primarily Cu)-support
interactions.24,38,61 These short-ranged interactions are further strengthened by the
favorable geometric contact area between the convex nanoparticles and the concave
mesopores.38 In addition, the weak binding allows facile removal of the ligands at a
relatively low temperature of 350 °C, without causing any significant nanoparticle
sintering.
After dealloying, the Pt-rich surface (Figure 6.3) binds even more strongly to
graphite than the Cu-rich surface in the original particles. Pt binds very strongly with
the graphitized carbons, as indicated by the shifting of the binding energy for the Pt
4f7/2 peak to higher energy value (refer XPS data in Appendix B, Figure B.6)
compared to both unsupported Pt and Pt on amorphous carbon.62,63 The exact nature
of the Pt interaction with graphite is still a topic of considerable discussion. However,
it is reported that the delocalized electrons in the p-orbital of the π sites on graphite
overlap with the d-orbital of the Pt, indicating that the nature of the binding is
partially covalent.29 In addition, partial electron transfer from the Pt to the graphitic
surface has also been determined by previous XPS29 and DFT64 studies. Thus, the
strong partially covalent bond due to orbital overlap coupled with partial charge
transfer between Pt and graphitic surfaces, enables strong metal-support interactions
and thus facilitates high catalyst stability (Table 6.4). The metal-support interactions
prevent the coalescence of dealloyed Pt nanoparticles during stability cycling,
210
consistent with no loss in electrochemical surface area after ADT (Figure 6.7),
typically observed for amorphous carbons.20 Similar beneficial effect of strong Pt-
graphite interactions for preventing coalescence of Pt nanoparticles at high
temperatures has been reported by Coloma et al.25
The stability in ECSA, CSA and ORR activity for the dealloyed Pt-Cu on
DMC-2000 and DMC-2600 suggest insignificant Pt dissolution and/or subsequent
Ostwald ripening, as a result of the resistance of the carbon to oxidation and strong
metal-support interactions. The relatively monodisperse size of the dealloyed particles
further limits Ostwald ripening. Although, the effect of metal-support interactions on
Pt dissolution is not well studied, it is plausible that the C-Pt binding synergistically
alters the electronic structure of the Pt, causing the observed shift to higher binding
energies for the Pt 4f7/2 XPS spectra (Appendix B, Figure B.6), thereby creating a
stabilizing effect against Pt oxidation. The delocalization of electron density in the
orbital overlap between the Pt and the π sites on the carbon lowers the tendency to
form a Pt-O bond as the potential becomes more positive, resulting in a decrease in Pt
dissolution. A related method to alter the Pt electronic structure to make it less
resistant to dissolution is galvanic deposition of Au clusters on Pt nanoparticles.17 The
large interfacial contact area between the nanoparticles and concave mesopores with
relatively similar radii, as well as the high dispersion of the nanoparticles on the high
surface area mesoporous supports with tortuous pathways in the 3D pores coupled
with the strong metal-graphite interactions appeared to prevent Pt dissolution,
Ostwald ripening and nanoparticle coalescence.
Although traditional wetness impregnation techniques are applicable for
producing catalysts with relatively amorphous carbons, a variety of challenges must
be addressed for highly graphitic supports. The lack of polar and charged oxygenated
sites on the graphitic surface precludes strong ion-dipole and electrostatic interactions
with ionic metal precursors.26 Consequently, the weakly interacting metal precursors
often lead to undesirably large nanoparticle sizes (>5 nm) upon reduction.26 To
211
enhance precursor interactions, the graphitic carbons may be oxygenated with acids.18
However, these oxygen groups weaken the interactions of the final metal
nanoparticles with graphitic π electrons and make the carbon more susceptible to
oxidation in fuel cells.
6.4.3 Dealloying – Structure and Activity Enhancement
The straining of the Pt shell from the underlying Cu in the dealloyed
nanoparticles has been shown to raise the catalyst activity markedly.1,12 However, the
relationship between strain and catalyst activity can be highly complex for multiphase
binary catalysts. Suppose a mixture of compositions of nanoparticles were dealloyed.
The Pt shells from starting particles with too little copper may not be strained enough
to produce high activity. For particles with too much copper, the final shell may have
too little Pt or the strain be too high. Thus, for multiphase catalysts, it may be difficult
to understand the relationship between activity and strain and to optimize the activity.
The low polydispersities in the composition of the essentially single-phase
nanoparticles in this study are beneficial for probing these phenomena, and
ultimately, for optimizing catalyst activity. In order to control, characterize and
optimize the strain in the Pt shell, it would be beneficial to form bimetallic
nanoparticles with low polydispersity in composition after dealloying. Clearly,
dealloying of single-phase particles would be more likely to produce such particles.
In this section, we discuss the effect of dealloying on particle morphology, and in the
next section we describe the control of the composition prior to dealloying.
The strain upon dealloying will lower the d-band centers relative to the Fermi
level and consequently enhance the ORR activity according to DFT studies.65-67 This
lowering reduces the adsorption energy of oxygen on the Pt-rich surface. The
resulting reduction in the coverage of Pt surface sites by oxygen intermediate species4
enhances the activity.
212
The dealloying of the Pt20Cu80 nanoparticles to Pt85Cu15 results in an almost
pure Pt surface with a copper enriched core (Figure 6.3 and Appendix B, Figure B.5),
as previously hypothesized by Strasser and others.1 A smaller lattice parameter was
observed for the Pt-rich particles after dealloying relative to pure Pt nanoparticles
(Table 6.3). Thus the strain from the copper core reduced Pt-Pt distances on the
particle surface and enhanced the activity. Furthermore, the catalyst activity increases
with increasing strain as reflected by the increase in activity from 0.16 A/ mg Pt to
0.45 A/ mg Pt for essentially single-phased Pt75Cu25 versus Pt20Cu80 nanoparticles,
respectively, prior to dealloying. The lack of any significant increase in ECSA upon
dealloying indicated that the activity increase was not be caused by surface
roughening (see supplementary information for further details) consistent with
previous reports.1
During dealloying at a given potential, the copper will thus be leached at
similar rates from particle to particle when the compositional polydispersity is low.
Hence, the dealloyed particles may be expected to also have low compositional
polydispersity and thus a similar degree of strain in the Pt shell for all nanoparticles,
as was observed. The dealloying process may be controlled more precisely than for
polydisperse nanoparticles with variable leaching rates. With these well-defined and
uniform particles, the relationship between catalyst activity and strain may be more
clearly elucidated. This fundamental understanding is beneficial for targeting the
optimum strain level to maximize catalyst activity.
6.4.4 Composition of Bimetallic Nanocrystals before Dealloying
The ability to achieve low polydispersity in particle composition in the
arrested growth precipitation, as shown by XRD in Figure 6.4, led to the high degree
of control of the composition and morphology of the particles after dealloying. In
contrast, dealloying of multiphase nanoparticles has the potential to produce a range
of different nanoparticle compositions with widely varying catalyst activities.
213
It is difficult to synthesize single-phased bimetallic catalyst with narrow
composition polydispersity and controlled size by the traditional wetness-
impregnation technique.1,12,31,68,69 For sequential impregnation the second metal
precursor is infused and reduced on pre-formed particles of the first metal on the
support. However, the second precursor may nucleate separately on the
heterogeneous support sites, instead of on the first metal resulting in a multi-phased
catalyst,33,34 as was seen for Pt-Cu nanoparticles.1 Three distinct composition phases
of Pt68Cu32, Pt24Cu76 and pure Cu were observed upon reduction of Cu precursor on
supported Pt particles.1 Upon dealloying, a very high activity enhancement of >4-fold
(0.5 A/mg Pt) relative to pure Pt was observed for this multi-phased catalyst. The
relationship between dealloying of the multiple phases and the varying Pt strain
depending upon the initial phase composition can be highly complex. For
simultaneous impregnation, the metal ions may separate chromatographically as they
diffuse into the pores, and produce large compositional polydispersity.31,32,69,70 For
example, PtNi synthesized by simultaneous impregnation produced two different
phases of NiO and Pt94Ni6 for a precursor ratio of Pt(70):Ni(30).71
The formation of essentially single-phase nanoparticles with narrow
compositional polydispersity in this study is favored by the arrested precipitation
technique of synthesis in the presence of stabilizing ligands.39,72,73 Both pure platinum
and copper exhibit FCC crystal structure which is favorable for attempting to prevent
phase separation. The precursor concentrations can be tuned to control nucleation and
growth rates to focus the nanocrystal size to narrow polydispersity, as shown in Table
6.2, facilitated by the dynamic solvation of the nanocrystal surface by the ligands.
The precursors are simultaneously reduced in absence of the carbon supports, which
may otherwise cause heterogeneous nucleation of multiple phases as is often seen in
the wetness impregnation technique.1,12,31,68,69 The composition may be varied by
varying the precursor concentrations to produce essentially single-phase particles (as
shown in Table 6.2 for Pt-Cu). The dealloying of these essentially single-phase
214
particles leads to better control of strain in the Pt shell which is beneficial both for
fundamental understanding and improvement of catalyst activity.
6.5 Conclusions
Herein, we have demonstrated a novel presynthesis/infusion method to design
highly stable and active bimetallic catalysts on graphitized mesoporous carbon. The
unprecedented stability of these highly graphitic ORR catalysts is indicated by a loss
in ECSA of <1% and a loss in catalytic activity of less than 2% after 1000 cycles
between 0.5 and 1.2 V, versus 70% loss for a typical commercial carbon, Vulcan XC-
72. The high resistance of the highly graphitic supports, with very few surface oxygen
sites, to oxidation to CO2 at 1.2 V for 7 hours, prevents nanoparticle isolation from
the surface. The strong covalent and charge transfer interactions between the π
electrons on the graphitic surface and the enriched Pt surface after dealloying mitigate
nanoparticle sintering, Pt dissolution and Ostwald ripening. The strong metal-carbon
interactions are further enhanced by the large interfacial contact area between the
metal nanocrystals and concave mesopores with relatively similar radii. Removal of
90% of the copper from these nanoparticles via dealloying produced essentially
single-phased Pt85Cu15 with large strain in the Pt shell, resulting in a 4-fold activity
enhancement for ORR over pure Pt nanoparticles. This large enhancement
corroborates previous results for Pt-Cu nanoparticles on Vulcan XC-72 carbon.1 The
arrested growth precipitation of Pt-Cu nanoparticles with stabilizing ligands favors
focusing of the size (<3 nm) as well as the composition, each with narrow
polydispersity, indicating essentially single-phase nanoparticles, as shown by XRD
and TEM. The low surface coverage of the weakly bound ligands resulted in strong
Cu-support interactions24 as indicated by high Pt-Cu loadings up to 15 wt %. The
ability to design and control the size and composition of uniform bimetallic
nanocrystals with the pre-synthesis concept and to achieve strong binding to graphitic
mesoporous carbon offers novel opportunities for understanding fundamental
215
catalytic mechanisms. This understanding will be beneficial for designing highly
active and stable catalysts, including electrocatalysts for oxygen reduction in PEM
fuel cells.
6.6 References
(1) Koh, S.; Strasser, P. Electrocatalysis on Bimetallic Surfaces: Modifying
Catalytic Reactivity for Oxygen Reduction by Voltammetric Surface Dealloying.
Journal of the American Chemical Society 2007, 129, 12624-12625.
(2) Stamenkovic, V. R.; Fowler, B.; Mun, B. S.; Wang, G.; Ross, P. N.; Lucas, C.
A.; Markovic, N. M. Improved Oxygen Reduction Activity on Pt3Ni(111) via
Increased Surface Site Availability. Science (Washington, DC, United States) 2007,
315, 493-497.
(3) Zhang, J.; Vukmirovic, M. B.; Xu, Y.; Mavrikakis, M.; Adzic, R. R.
Controlling the catalytic activity of platinum-monolayer electrocatalysts for oxygen
reduction with different substrates. Angewandte Chemie, International Edition 2005,
44, 2132-2135.
(4) Stamenkovic, V. R.; Mun, B. S.; Arenz, M.; Mayrhofer, K. J. J.; Lucas, C. A.;
Wang, G.; Ross, P. N.; Markovic, N. M. Trends in electrocatalysis on extended and
nanoscale Pt-bimetallic alloy surfaces. Nature Materials 2007, 6, 241-247.
(5) Gasteiger, H. A. Proton exchange fuel cell materials and R&D needs for
future market success. Electrochemistry (Tokyo, Japan) 2007, 75, 103.
(6) Salgado, J. R. C.; Antolini, E.; Gonzalez, E. R. Structure and Activity of
Carbon-Supported Pt-Co Electrocatalysts for Oxygen Reduction. Journal of Physical
Chemistry B 2004, 108, 17767-17774.
(7) Koh, S.; Leisch, J.; Toney, M. F.; Strasser, P. Structure-Activity-Stability
Relationships of Pt-Co Alloy Electrocatalysts in Gas-Diffusion Electrode Layers.
Journal of Physical Chemistry C 2007, 111, 3744-3752.
216
(8) Xiong, L.; Manthiram, A. Nanostructured Pt-M/C (M = Fe and Co) catalysts
prepared by a microemulsion method for oxygen reduction in proton exchange
membrane fuel cells. Electrochimica Acta 2005, 50, 2323-2329.
(9) Xiong, L.; Manthiram, A. Effect of atomic ordering on the catalytic activity of
carbon supported PtM (M = Fe, Co, Ni, and Cu) alloys for oxygen reduction in
PEMFCs. Journal of the Electrochemical Society 2005, 152, A697-A703.
(10) Wakabayashi, N.; Takeichi, M.; Uchida, H.; Watanabe, M. Temperature
Dependence of Oxygen Reduction Activity at Pt-Fe, Pt-Co, and Pt-Ni Alloy
Electrodes. Journal of Physical Chemistry B 2005, 109, 5836-5841.
(11) Yano, H.; Kataoka, M.; Yamashita, H.; Uchida, H.; Watanabe, M. Oxygen
Reduction Activity of Carbon-Supported Pt-M (M = V, Ni, Cr, Co, and Fe) Alloys
Prepared by Nanocapsule Method. Langmuir 2007, 23, 6438-6445.
(12) Mani, P.; Srivastava, R.; Strasser, P. Dealloyed Pt-Cu Core-Shell Nanoparticle
Electrocatalysts for Use in PEM Fuel Cell Cathodes. Journal of Physical Chemistry C
2008, 112, 2770-2778.
(13) Koh, S.; Strasser, P. Electrocatalysis on bimetallic surfaces: modifying
catalytic reactivity for oxygen reduction by voltammetric surface dealloying. J Am
Chem Soc FIELD Full Journal Title:Journal of the American Chemical Society
FIELD Publication Date:2007, 129, 12624-12625. FIELD Reference Number:
FIELD Journal Code:7503056 FIELD Call Number:.
(14) Mani, P.; Srivastava, R.; Yu, C.; Strasser, P. In-situ, in-layer de-alloying of Pt-
M intermetallics for high performance PEMFC electrode layers: MEA activity and
durability studies. ECS Transactions 2007, 11, 933-939.
(15) Strasser, P.; Koh, S.; Yu, C. Voltammetric surface dealloying of Pt bimetallic
nanoparticles: a novel synthetic method towards more efficient ORR electrocatalysts.
ECS Transactions 2007, 11, 167-180.
(16) Koh, S.; Yu, C.; Strasser, P. De-alloyed Pt-M nanoparticle electrocatalysts for
efficient electroreduction of oxygen: structural-activity-stability relationship. ECS
Transactions 2007, 11, 205-215.
217
(17) Zhang, J.; Sasaki, K.; Sutter, E.; Adzic, R. R. Stabilization of Platinum
Oxygen-Reduction Electrocatalysts Using Gold Clusters. Science (Washington, DC,
United States) 2007, 315, 220-222.
(18) Wang, J.; Yin, G.; Shao, Y.; Wang, Z.; Gao, Y. Investigation of Further
Improvement of Platinum Catalyst Durability with Highly Graphitized Carbon
Nanotubes Support. Journal of Physical Chemistry C 2008, 112, 5784-5789.
(19) Shao, Y.; Yin, G.; Gao, Y.; Shi, P. Durability study of Pt/C and Pt/CNTs
catalysts under simulated PEM fuel cell conditions. Journal of the Electrochemical
Society 2006, 153, A1093-A1097.
(20) Ferreira, P. J.; la O, G. J.; Shao-Horn, Y.; Morgan, D.; Makharia, R.; Kocha,
S.; Gasteiger, H. A. Instability of Pt/C electrocatalysts in proton exchange membrane
fuel cells: A mechanistic investigation. Journal of the Electrochemical Society 2005,
152, A2256-A2271.
(21) Mathias, M. F.; Makharia, R.; Gasteiger, H. A.; Conley, J. J.; Fuller, T. J.;
Gittleman, C. I.; Kocha, S. S.; Miller, D. P.; Mittelsteadt, C. K.; Xie, T.; Yan, S. G.;
Yu, P. T. Two fuel cell cars in every garage? Electrochemical Society Interface 2005,
14, 24-35.
(22) Waje, M. M.; Li, W.; Chen, Z.; Larsen, P.; Yan, Y. Effect of scan range on Pt
surface area loss in potential cycling experiments. ECS Transactions 2007, 11, 1227-
1233.
(23) Gasteiger, H. A.; Kocha, S. S.; Sompalli, B.; Wagner, F. T. Activity
benchmarks and requirements for Pt, Pt-alloy, and non-Pt oxygen reduction catalysts
for PEMFCs. Applied Catalysis, B: Environmental 2005, 56, 9-35.
(24) Yu, R.; Song, H.; Zhang, X.-F.; Yang, P. Thermal Wetting of Platinum
Nanocrystals on Silica Surface. Journal of Physical Chemistry B 2005, 109, 6940-
6943.
(25) Coloma, F.; Sepulveda-Escribano, A.; Rodriquez-Reinoso, F. Heat-treated
carbon blacks as supports for platinum catalysts. Journal of Catalysis 1995, 154, 299-
305.
218
(26) Yu, X.; Ye, S. Recent advances in activity and durability enhancement of Pt/C
catalytic cathode in PEMFC. Journal of Power Sources 2007, 172, 145-154.
(27) Yu, X.; Ye, S. Recent advances in activity and durability enhancement of Pt/C
catalytic cathode in PEMFC. Journal of Power Sources 2007, 172, 133-144.
(28) Shao, Y.; Yin, G.; Gao, Y. Understanding and approaches for the durability
issues of Pt-based catalysts for PEM fuel cell. Journal of Power Sources 2007, 171,
558-566.
(29) Cuong, N. T.; Chi, D. H.; Kim, Y.-T.; Mitani, T. Structural and Electronic
Properties of Ptn (n=3, 7, 13) clusters on metallic single wall carbon nanotube.
Physica Status Solidi 2006, 13, 3472-3475.
(30) Wakayama, H.; Setoyama, N.; Fukushima, Y. Size-Controlled Synthesis and
Catalytic Performance of Pt Nanoparticles in Micro- and Mesoporous Silica Prepared
Using Supercritical Solvents. Adv. Mater. 2003, 15, 742-745.
(31) Raghuveer, V.; Ferreira, P. J.; Manthiram, A. Comparison of Pd-Co-Au
electrocatalysts prepared by conventional borohydride and microemulsion methods
for oxygen reduction in fuel cells. Electrochemistry Communications 2006, 8, 807-
814.
(32) Freund, A.; Lang, J.; Lehmann, T.; Starz, K. A. Improved Pt alloy catalysts
for fuel cells. Catalysis Today 1996, 27, 279-283.
(33) Min, M.-K.; Cho, J.; Cho, K.; Kim, H. Particle size and alloying effects of Pt-
based alloy catalysts for fuel cell applications. Electrochimica Acta 2000, 45, 4211-
4217.
(34) Antolini, E. Formation of carbon-supported PtM alloys for low temperature
fuel cells: a review. Materials Chemistry and Physics 2003, 78, 563-573.
(35) Rioux, R. M.; Song, H.; Hoefelmeyer, J. D.; Yang, P.; Somorjai, G. A. High-
Surface-Area Catalyst Design: Synthesis, Characterization, and Reaction Studies of
Platinum Nanoparticles in Mesoporous SBA-15 Silica. Journal of Physical Chemistry
B 2005, 109, 2192-2202.
219
(36) Gupta, G.; Shah, P. S.; Zhang, X.; Saunders, A. E.; Korgel, B. A.; Johnston,
K. P. Enhanced Infusion of Gold Nanocrystals into Mesoporous Silica with
Supercritical Carbon Dioxide. Chemistry of Materials 2005, 17, 6728-6738.
(37) Gupta, G.; Stowell, C. A.; Patel, M. N.; Gao, X.; Yacaman, M. J.; Korgel, B.
A.; Johnston, K. P. Infusion of Pre-synthesized Iridium Nanocrystals into
Mesoporous Silica for High Catalyst Activity. Chemistry of Materials 2006, 18,
6239-6249.
(38) Gupta, G.; Patel, M. N.; Ferrer, D.; Heitsch, A. T.; Korgel, B. A.; Johnston, K.
P. Stable Ordered FePt Mesoporous Silica Catalysts with High Loadings Chemistry of
Materials 2008.
(39) Chen, Y.; Yang, F.; Dai, Y.; Wang, W.; Chen, S. Ni@Pt core-shell
nanoparticles: synthesis, structural and electrochemical properties. Journal of
Physical Chemistry C 2008, 112, 1645-1649.
(40) Vijayaraghavan, G.; Stevenson, K. J. Synergistic Assembly of Dendrimer-
Templated Platinum Catalysts on Nitrogen-Doped Carbon Nanotube Electrodes for
Oxygen Reduction. Langmuir 2007, 23, 5279-5282.
(41) Ye, H.; Crooks, R. M. Effect of Elemental Composition of PtPd Bimetallic
Nanoparticles Containing an Average of 180 Atoms on the Kinetics of the
Electrochemical Oxygen Reduction Reaction. Journal of the American Chemical
Society 2007, 129, 3627-3633.
(42) Hou, J. G.; Li, X.; Wang, H.; Wang, B. Cu-C60 interaction and nano-
composite structures. Journal of Physics and Chemistry of Solids 2000, 61, 995-998.
(43) Sun, X.; Kang, S.; Harrell, J. W.; Nikles, D. E.; Dai, Z. R.; Li, J.; Wang, Z. L.
Synthesis, chemical ordering, and magnetic properties of FePtCu nanoparticle films.
Journal of Applied Physics 2003, 93, 7337-7339.
(44) Liang, C.; Dai, S. Synthesis of Mesoporous Carbon Materials via Enhanced
Hydrogen-Bonding Interaction. Journal of the American Chemical Society 2006, 128,
5316-5317.
(45) Wang, X. Q.; Liang, C.; Dai, S. Langmuir 2008, in press.
220
(46) Pantea, D.; Darmstadt, H.; Kaliaguine, S.; Summchen, L.; Roy, C. Electrical
conductivity of thermal carbon blacks. Influence of surface chemistry. Carbon 2001,
39, 1147-1158.
(47) Cuesta, A.; Dhamelincourt, P.; Laureyns, J.; Martinez-Alonso, A.; Tascon, J.
M. D. Raman microprobe studies on carbon materials. Carbon 1994, 32, 1523-1532.
(48) Kinoshita, K. Carbon: Electrochemical and Physicochemical Properties; John
Wiey and Sons, 1988.
(49) Tian, J. M.; Wang, F. B.; Shan, Z. H. Q.; Wang, R. J.; Zhang, J. Y. Effect of
preparation conditions of Pt/C catalysts on oxygen electrode performance in proton
exchange membrane fuel cells. Journal of Applied Electrochemistry 2004, 34, 461-
467.
(50) Nemanich, R. J.; Solin, S. A. First- and second-order Raman scattering from
finite-size crystals of graphite. Physical Review B: Condensed Matter and Materials
Physics 1979, 20, 392-401.
(51) Maldonado, S.; Morin, S.; Stevenson, K. J. Structure, composition, and
chemical reactivity of carbon nanotubes by selective nitrogen doping. Carbon 2006,
44, 1429-1437.
(52) Tuinstra, F.; Koenig, J. L. Raman spectrum of graphite. Journal of Chemical
Physics 1970, 53, 1126-1130.
(53) Ruban, A. V.; Skriver, H. L.; Norskov, J. K. Surface segregation energies in
transition-metal alloys. Physical Review B: Condensed Matter and Materials Physics
1999, 59, 15990-16000.
(54) Mukerjee, S. Particle size and structural effects in platinum electrocatalysis.
Journal of Applied Electrochemistry 1990, 20, 537-548.
(55) Kinoshita, K. Particle size effects for oxygen reduction on highly dispersed
platinum in acid electrolytes. Journal of the Electrochemical Society 1990, 137, 845-
848.
(56) Vegard, L. Formation of mixed crystals by the contact of solid phases.
Zeitschrift fuer Physik 1921, 5, 393-395.
221
(57) Vegard, L. The constitution of mixed crystals and the space occupied by
atoms. Zeitschrift fuer Physik 1921, 5, 17-26.
(58) Liu, Z.; Koh, S.; Yu, C.; Strasser, P. Synthesis, dealloying, and ORR-
electrocatalysis of PDDA-stabilized Cu-rich Pt alloy nanoparticles. Journal of the
Electrochemical Society 2007, 154, B1192-B1199.
(59) Chen, Z.; Deng, W.; Wang, X.; Yan, Y. Durability and activity study of
single-walled, double-walled and multi-walled carbon nanotubes supported Pt
catalyst for PEMFCs. ECS Transactions 2007, 11, 1289-1299.
(60) Stevens, D. A.; Hicks, M. T.; Haugen, G. M.; Dahn, J. R. Ex Situ and In Situ
Stability Studies of PEMFC Catalysts. Journal of the Electrochemical Society 2005,
152, A2309-A2315.
(61) Jana, N. R.; Peng, X. Single-Phase and Gram-Scale Routes toward Nearly
Monodisperse Au and Other Noble Metal Nanocrystals. Journal of the American
Chemical Society 2003, 125, 14280-14281.
(62) Fu, X.; Wang, Y.; Wu, N.; Gui, L.; Tang, Y. Surface Modification of Small
Platinum Nanoclusters with Alkylamine and Alkylthiol: An XPS Study on the
Influence of Organic Ligands on the Pt 4f Binding Energies of Small Platinum
Nanoclusters. Journal of Colloid and Interface Science 2001, 243, 326-330.
(63) Antolini, E.; Giorgi, L.; Cardellini, F.; Passalacqua, E. Physical and
morphological characteristics and electrochemical behavior in PEM fuel cells of
PtRu/C catalysts. Journal of Solid State Electrochemistry 2001, 5, 131-140.
(64) Yang, D. Q.; Zhang, G. X.; Sacher, E.; Jose-Yacaman, M.; Elizondo, N.
Evidence of the Interaction of Evaporated Pt Nanoparticles with Variously Treated
Surfaces of Highly Oriented Pyrolytic Graphite. Journal of Physical Chemistry B
2006, 110, 8348-8356.
(65) Xu, Y.; Ruban, A. V.; Mavrikakis, M. Adsorption and dissociation of O2 on
Pt-Co and Pt-Fe alloys. Journal of the American Chemical Society 2004, 126, 4717-
4725.
222
(66) Greeley, J.; Krekelberg, W. P.; Mavrikakis, M. Strain-induced formation of
subsurface species in transition metals. Angewandte Chemie, International Edition
2004, 43, 4296-4300.
(67) Schlapka, A.; Lischka, M.; Gross, A.; Kasberger, U.; Jakob, P. Surface Strain
versus Substrate Interaction in Heteroepitaxial Metal Layers: Pt on Ru(0001).
Physical Review Letters 2003, 91, 016101/016101-016101/016104.
(68) Zhou, B.; Hermans, S.; Somorjai, G. A. Nanotechnology in Catalysis; Kluwer
Academic/ Plenum 2004; Vol. 1.
(69) Fernandez, J. L.; Raghuveer, V.; Manthiram, A.; Bard, A. J. Pd-Ti and Pd-Co-
Au Electrocatalysts as a Replacement for Platinum for Oxygen Reduction in Proton
Exchange Membrane Fuel Cells. Journal of the American Chemical Society 2005,
127, 13100-13101.
(70) Vasiliev, M. A. Surface effects of ordering in binary alloys. Journal of
Physics D: Applied Physics 1997, 30, 3037-3070.
(71) Antolini, E.; Salgado, J. R. C.; dos Santos, A. M.; Gonzalez, E. R. Carbon-
supported Pt-Ni alloys prepared by the borohydride method as electrocatalysts for
DMFCs. Electrochemical and Solid-State Letters 2005, 8, A226-A230.
(72) Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R. Synthesis of
Thiol-Derivatised Gold Nanoparticles in a Two-Phase Liquid-Liquid System. J.
Chem. Soc., Chem. Commun. 1994, 7, 801-802.
(73) Yin, Y.; Alivisatos, A. P. Colloidal nanocrystal synthesis and the organic-
inorganic interface. Nature (London, United Kingdom) 2005, 437, 664-670.
223
Chapter7
Conclusions and Recommendations
7.1 Conclusions
7.1.1 Infusion of Metal Nanocrystals in Mesoporous Silica using sc-CO2
A novel approach is presented to infuse presynthesized gold nanocrystals into
mesoporous silica by tuning the solvent quality with CO2 as an antisolvent. The
concept of decoupling the nanocrystal synthesis step and the infusion step provides
exquisite control of the nanocrystal size, morphology and dispersibility within the
pores, without perturbing the morphology of the mesoporous silica. High loadings of
the gold and iridium nanocrystals in mesoporous silica (~2 wt %) are achieved by
tuning the solvent strength with CO2, in order to strengthen the van der Waals
interactions of the nanocrystals with the silica pore walls, without flocculating the
nanocrystals in the bulk solvent. The loadings do not show significant dependence on
the pore size of the silica as well as concentration of nanocrystals in the bulk solvent.
The high surface area of the iridium nanocrystals produced high turnover frequency
(TOF) of 16 s-1 for decene hydrogenation and is higher than the TOF for
homogeneous iridium catalysts. Strategic selection of weakly binding ligand
(tetraoctylammonium bromide) for iridium stabilization facilitated catalyst activation
without any thermal calcination. Weakly bound ligands on the nanocrystal surface
may be expected to migrate on to the vacant sites on the silica, significantly reducing
the ligand coverage on the surface and hence increasing the catalyst activity.
Desorption of ligands from the adsorbed nanoparticle surface, onto the support and
into the decene during reaction, will enhance metal binding to the surface and aid
catalyst stability.
224
7.1.2 Synthesis of Stable and Ordered FePt Bimetallic Catalysts
Highly stable and active catalysts composed of bimetallic nanocrystals on
mesoporous supports with a controlled intermetallic alloy structure have been
designed for the first time. Whereas CO2 was used to facilitate gold nanocrystal
adsorption on the support as shown in Chapter 2, it is not needed for FePt coated with
weakly binding ligands, since the interactions of the nanocrystals with the support
were stronger as a consequence of uncovered metal on the particle surface. The
catalysts are formed by infusion of pre-synthesized FePt nanocrystals, with well-
controlled size and composition. The arrested growth precipitation favors focusing of
the size (<4 nm) as well as the composition of the nanocrystals. The stabilizing
ligands control dynamic solvation, supersaturation, nucleation and growth, as
reflected in the small nanocrystal size. The low surface coverage of the weakly bound
ligands resulted in strong metal-support interactions as indicated by high metal
loadings above 10 wt % in only 10 min and stability against sintering upon thermal
annealing at 700 °C. These interactions are further increased by the enhanced
interfacial contact area between the metal nanocrystals and concave mesopores with
similar radii. Finally, the calcined FePt catalyst exhibits 6-fold higher catalyst activity
than commercial Pd-alumina catalyst for liquid 1-decene hydrogenation and was
stable for multiple reactions.
7.1.3 Stable Pt Electrocatalysts Supported on Graphitic Carbon
A novel presynthesis/infusion method to design highly stable Pt
electrocatalysts on ordered graphitized mesoporous carbon with controlled
nanocrystal size and morphology has been demonstrated for the first time. The
graphitic supports with very few surface oxygen sites, offers high resistance to
oxidation to CO2, even at potentials of 1.2 V for 7 hours. Lack of support oxidation
prevents nanoparticle isolation from the surface and consequent activity loss. The
strong metal- support interactions between the π sites on the graphitic surface and the
225
Pt surface mitigate nanoparticle sintering, Pt dissolution and Ostwald ripening, as
indicated by a loss in ECSA of <1% and a loss in catalytic activity of <2% after 1000
cycles between 0.5 and 1.2 V. These metal-carbon interactions are further enhanced
by the large interfacial contact area between the metal nanocrystals and concave
mesopores with relatively similar radii. For amorphous carbons such as Vulcan XC-
72, carbon oxidation and Pt dissolution lead to a loss in ORR activity of 70% upon
potential cycling. The arrested growth precipitation of Pt nanocrystals with 3 nm size
and high concentrations of the active (111) surface leads to 25% higher activities than
typically seen for Pt formed on Vulcan carbon. The low surface coverage of the
weakly bound ligands favors strong Pt-C interactions as indicated by high metal
loadings of 20 wt %.
7.1.4 Highly Stable and Active Bimetallic Pt-Cu for PEMFC
Highly active bimetallic catalysts on graphitized mesoporous carbon with
controlled nanocrystal size, composition and alloy structure and high stability have
been synthesized. Electrochemically leaching of copper from the nanocrystals
generates a strained Pt shell, enhancing the activity for oxygen reduction over 4-fold
relative to Pt catalysts, as observed previously for Vulcan XC-72 carbon. The arrested
growth precipitation of Pt-Cu nanoparticles with stabilizing ligands favors focusing of
the size (<3 nm) as well as the composition of the nanocrystals resulting in a single
crystal phase. The low surface coverage of the weakly bound ligands resulted in
strong Cu-support interactions as indicated by high Pt-Cu loadings up to 15 wt %.
The strong interactions between the π electrons on the graphitic surface and the
enriched Pt surface after dealloying mitigate nanoparticle sintering, Pt dissolution and
Ostwald ripening during potential cycling between 0.5 and 1.2 V, as previously
shown in Chapter 5. The ability to design catalyst morphology and crystal structure
with pre-synthesized bimetallic nanocrystals on graphitic mesoporous carbon offers
226
novel opportunities for highly active and stable catalysts including electrocatalysts for
oxygen reduction in PEMFCs.
7.2 Recommendations
7.2.1 Design of Core-Shell type bimetallic catalysts for ORR
It is shown in Chapter 6 that Pt-Cu nanoparticles after electrochemical
dealloying of copper show activity enhancement of 4x over pure Pt catalyst possibly
due to formation of a core shell structure, with an copper-rich core and a Pt metal
shell.1 However in using the Pt-Cu catalyst as the cathode in the fuel cell stack, a
practical challenge needs to be overcome. The copper ions that are dealloyed will
leach into the proton conducting membrane (like nafion) and may stick to the
negative sites of the membrane. Consequently, the resistance of the membrane for the
transport of the H+ to the cathode significantly increases which limits the performance
of the PEMFC. To overcome this challenge, it would be desirable to design catalysts
with core-shell nanoparticles composed of pure copper core and a pure Pt shell. These
core shell particles may potentially display high catalyst activities due to the strain
generated on the Pt shell by the pure copper core, analogous to the dealloyed Pt-Cu
systems but without any catalyst activation step for copper removal.
The area of core-shell nanoparticles as ORR catalysts is largely unexplored.
Recently, there have been a few advances in the development of core-shell
nanoparticles.2,3 By synthesizing and testing various Cu-Pt core-shell nanoparticles,
relationships between the strain on the Pt surface and the core copper can be
elucidated. In this way, the structure of nanoparticles in relation to their catalytic
activity can be studied in a facile manner.
7.2.2 Design of Active Catalysts for Water-Gas Shift Reaction
227
As per the recommendation of Department of Energy, there is a need for (1)
design and assembly of nanoscale catalysts for energy related reactions, (2)
understanding and control of catalyst activities and selectivities for energy related
reactions, and (3) development of concepts in catalysis to meet major energy research
needs. The key reactions in hydrogen production are the low temperature water gas
shift (WGS) reaction, CO + H2O = CO2 + H2 and preferential oxidation of CO
(PROX) in order to increase H2 conversions.4-6 The CO level must be lowered to
about 1% in the WGS reaction and to 50 ppm in PROX to prevent poisoning of
catalysts in fuel cells. Existing catalysts are not selective and stable enough to meet
these goals. The novel “bottom up design” concept to prepare catalysts that has been
developed for fuel cells as shown in Chapters 4-6 can be used to design WGS
catalysts with much higher performance and lower amounts of expensive noble
metals including Au and Pt. Pre-synthesized metal nanocrystals will be infused into
ordered mesoporous metal oxide supports to achieve extraordinary control of the
morphology of the active sites at the metal-metal oxide interface. The structure and
bonding at the metal-metal oxide interface will be characterized by X-ray absorption
spectroscopy (XAS), electron energy-loss spectroscopy (EELS) and other techniques
to understand the catalytic mechanisms.4,5,7,8 This understanding of structure-activity
relationships for these catalysts with well-defined active sites will guide the discovery
novel generations of highly active, selective and stable catalysts for hydrogen related
reactions.
7.3 References
(1) Koh, S.; Strasser, P. Electrocatalysis on Bimetallic Surfaces:
Modifying Catalytic Reactivity for Oxygen Reduction by Voltammetric Surface
Dealloying. Journal of the American Chemical Society 2007, 129, 12624-12625.
(2) Muraviev, D. N.; Macanas, J.; Parrondo, J.; Munoz, M.; Alonso, A.;
Alegret, S.; Ortueta, M.; Mijangos, F. Cation-exchange membrane as nanoreactor:
228
Intermatrix synthesis of platinum-copper core-shell nanoparticles. Reactive &
Functional Polymers 2007, 67, 1612-1621.
(3) Chen, Y.; Yang, F.; Dai, Y.; Wang, W.; Chen, S. Ni@Pt core-shell
nanoparticles: synthesis, structural and electrochemical properties. Journal of
Physical Chemistry C 2008, 112, 1645-1649.
(4) Mathias, M. F.; Makharia, R.; Gasteiger, H. A.; Conley, J. J.; Fuller, T.
J.; Gittleman, C. I.; Kocha, S. S.; Miller, D. P.; Mittelsteadt, C. K.; Xie, T.; Yan, S.
G.; Yu, P. T. Two fuel cell cars in every garage? Electrochemical Society Interface
2005, 14, 24-35.
(5) Wakayama, H.; Setoyama, N.; Fukushima, Y. Size-Controlled
Synthesis and Catalytic Performance of Pt Nanoparticles in Micro- and Mesoporous
Silica Prepared Using Supercritical Solvents. Adv. Mater. 2003, 15, 742-745.
(6) Fu, X.; Wang, Y.; Wu, N.; Gui, L.; Tang, Y. Surface Modification of
Small Platinum Nanoclusters with Alkylamine and Alkylthiol: An XPS Study on the
Influence of Organic Ligands on the Pt 4f Binding Energies of Small Platinum
Nanoclusters. Journal of Colloid and Interface Science 2001, 243, 326-330.
(7) Tibiletti, D.; Fonseca, A. A.; Burch, R.; Chen, Y.; Fisher, J. M.;
Goguet, A.; Hardacre, C.; Hu, P.; Thompsett, D. DFT and in situ EXAFS
investigation of gold/ceria-zirconia low-temperature water gas shift catalysts:
identification of the nature of the active form of gold. J. Phys. Chem. B 2005, 109,
22553-22559. FIELD Reference Number: FIELD Journal Code:101157530 FIELD
Call Number:.
(8) Schwartz, V.; Mullins, D. R.; Yan, W.; Chen, B.; Dai, S.; Overbury, S.
H. XAS Study of Au Supported on TiO2: Influence of Oxidation State and Particle
Size on Catalytic Activity. Journal of Physical Chemistry B 2004, 108, 15782-15790.
229
Appendix A
Infusion of FePt Nanocrystals in Mesoporous Silica
These results relate to FePt nanocrystals infused in mesoporous silica, as
described in Chapter 4.
A.1 Characterization
N2 porosimetry was used to measure the surface area and the pore size
distribution of the mesoporous silica with a high speed surface area BET analyzer
(NOVA 2000, Quantachrome Instruments, Boynton Beach, FL) at a temperature of
77 K. Silica samples were pretreated at 100 ºC for 24 h in vacuum immediately prior
to data collection. Pore size distributions were analyzed by using the adsorption
branch of the isotherm. The Barrett- Joyner- Halenda (BJH) method1 was used for
pore size distribution.
Small angle X-ray scattering (SAXS) was used to determine the long ranged
structure and the center-to-center distance between the pores of synthesized
mesoporous silica. SAXS was performed using a rotating copper anode generator
(Bruker Nonius) operated at 3.0 kW. Scattered photons were collected on a multiwire
gas-filled detector (Molecular Metrology, Inc.). The scattering angle was calibrated
using a silver behenate (CH3(CH2)20COOAg) standard.
Reproduced with permission from Gupta, G.; Patel, M. N.; Ferrer, D.; Heitsch, A. T.; Korgel, B. A.; Johnston, K. P.; Chemistry of Materials 2008, 106, 12178-12185, Supplementary Information. Copyright 2008 American Chemical Society.
230
A.2 Results and Discussion
SAXS was used to characterize the mesoporous silica prior to infusion as
shown in Figure A.1. Highly oriented pores, arranged in a 2-D hexagonal lattice, give
rise to a series of Bragg peaks in the scattering pattern. For mesoporous silica, the
three well-resolved peaks are indexable as the (100), (110), (200) reflections
associated with p6mm hexagonal symmetry. Presence of multiple sharp peaks
indicates that calcined mesoporous silica has a high degree of hexagonal mesoscopic
organization. The three peaks for SBA-15 are located at q100 = 0.74, q110 = 1.28, q200 =
1.48 nm-1 with the peak position ratio of 1: 3 :2, which is consistent with the
literature2-4. The distance d100 between adjacent planes is given by d100 = 2π/q100 =
8.5 nm. The unit cell parameter a0 = 3/2 100d = 9.8 nm gives the center-to-center
spacing of the pores.
Representative nitrogen adsorption/desorption isotherms and corresponding
pore size distributions (BJH method1) are shown in Figure A.2. Mesoporous silica has
a mean pore diameter of 5.5 nm, BET surface area of 541 m2/g and a pore volume of
0.62 cm3/g. The pore size distribution is narrow as indicated by the slope and step in
the adsorption isotherms. The wall thickness for the silica is 4.3 nm, which is
determined by subtracting pore diameter from the unit cell parameter as determined
from SAXS.
It is estimated that for 1 gm of mesoporous silica with a loading of 7.3% of
FePt, the surface area of FePt considering there are no ligands on the surface, is just
8.3 m2. Assuming 40% coverage of ligands, the surface area of FePt will be just 5 m2.
A small change in surface area of ~ 5 m2 due to infusion of the nanocrystals will be
difficult to be identified by nitrogen porosimetry.
Figure A.3 shows the HAADF-STEM micrographs of the FePt nanoparticles
infused in mesoporous silica, after calcination at 700 °C in 7%H2/93%N2 atmosphere.
Most of the nanoparticles were well dispersed in the pores and did not show any
significant change in size upon annealing. However, there are small regions in the
231
0.00E+00
5.00E+02
1.00E+03
1.50E+03
2.00E+03
2.50E+03
3.00E+03
3.50E+03
0 0.5 1 1.5 2 2.5 3
2 theta
Inte
nsity
Figure A.1
SAXS on mesoporous silica of type SBA-15. Center-to-center spacing between
cylindrical pores is 8.5 nm.
232
Adsorption Isotherm
0
50
100
150
200
250
300
350
400
450
0.00E+00 2.00E-01 4.00E-01 6.00E-01 8.00E-01 1.00E+00
Relative Pressure (P/P0)
Volu
me
(cc/
g)
Adsorption
Desorption
(a)
Pore size distribution for mesoporous silica
-5.00E-030.00E+005.00E-031.00E-021.50E-022.00E-022.50E-023.00E-023.50E-024.00E-02
-10 10 30 50 70 90 110 130 150
Pore diameter (A)
Ads
orpt
ion
dV(D
) (cc
/A/g
)
(b)
Figure A.2
(a) Adsorption isotherms for SBA-15 measured by nitrogen porosimetry. (b) Pore size
distribution for SBA-15 measured by nitrogen porosimetry.
233
Figure A.3
HAADF-STEM micrographs of 7.3 wt % FePt nanocrystals in mesoporous silica
calcined at 700 oC in 7%H2/93%N2.
Nanoparticles on external surface aggregated at high temperature calcination
Nanoparticles within the pores do not show aggregation
234
nanocomposite, where the nanoparticles grew to 20 nm. These particles were most
likely on the external surface of the silica and not protected by the concave curvature
of the internal pores.
We develop a model where contributions from van der Waals interactions and
charge-induced dipole interactions are used to estimate the total long ranged
interaction potential between the metal nanocrystal and silica. The charge-induced
dipole interaction varies as ~ 1/R4 (where R is the separation distance between the
silica and nanocrystal surface) and decays rapidly with distance.5 Although the charge
on mesoporous silica could not be measured, values from literature for silica particles
dispersed in benzene give a charge density between 20 to 200 charges/μm2,
depending on the size of the particle.6 Assuming the charges are spaced equally apart,
the charge to charge distance is approximately 80 to 250 nm. With a 1/R4 dependence
on distance, it can be assumed that the metal nanocrystal interacts primarily with the
closest point charge at small distances. The interaction energy for charge-induced
dipole is given by5
420
2
dipole induced-charge )4(2q V
Rrεπεα
= (A.1)
where q is the charge, εr is the relative dielectric constant, ε0 is the vacuum
permittivity, and α is the polarizability of the metal and is given by
3
0
.4
arεπεα
= (A.2)
where a is radius of the metal nanocrystal. Hence Vcharge-induced dipole varies as a3.
The van der Waals interaction potential between the ligand bound metal core and
silica surface mediated by toluene is given by7,8
(A.3)
(A.4) 1
ln21
11)( , ++
++=
ij
ij
ijijji x
xxx
xH
( )[ ])()()(121
2,11,1 xHAAxHAV SiltolueneLigSiltolueneMetalSiltolueneLigvdW −−−−−− −+−=
235
(A.5)
where ALig-toluene-Sil and AMetal-toluene-Sil are the mixed Hamaker constants and l is the
ligand length. The Hamaker constant, Axyz, for a medium x and medium z interacting
across solvent y, may be approximated by5,
))(( yyzzyyxxxyz AAAAA −−= (A.6)
where Axx, Ayy and Azz are Hamaker constants for pure components.
The total long interaction potential (VvdW + Vcharge-induced dipole) is plotted as a
function of the separation distance between the metal nanocrystal and silica for
different nanocrystal diameters (Figure A.4). The van der Waals contribution to the
total interaction potential is weakly negative on the order of 10-4 to 10-5 eV for 4.6 to
2.2 nm particle sizes at small separation distances. Hence, the attractive interactions
are caused chiefly by the charge-induced dipole interactions, which are orders of
magnitude stronger than van der Waals. From Figure A.4, 2.2 nm nanocrystals do not
have a high enough attractive potential for silica in the toluene medium. Even at a
separation distance of 2 nm, which is comparable to the ligand length of
dodecanethiol or oleic acid/oleyl amine stabilizing the nanocrystal, the attractive
potential is much smaller than (3/2)kBT (kB is the Boltzmann constant). Hence,
nanocrystal adsorption on silica is very weak and nanocrystals can desorb easily by
thermal energy associated with the 3 degrees of freedom for the nanocrystal. This
result was seen previously with insignificant loadings of less than 0.1 wt.% for 2.2
nm gold nanocrystals in mesoporous silica with a pore diameter of 8.8 nm, as
reported in our earlier publication.9
alR
xa
Rx2
;2 1211
+==
236
-0.25
-0.2
-0.15
-0.1
-0.05
0
0 2 4 6 8 10 12
2.2 nm3.6 nm4.6 nm(3/2)kT
Long
rang
ed In
tera
ctio
n En
ergy
(eV)
Separation Distance (nm)
0 100
5 10-5
1 10-4
1.5 10-4
2 10-4
0 2 4 6 8 10 12
Van
der W
aals
Ene
rgy
(eV)
Figure A.4
Total interaction energy as a function of separation distance and metal nanocrystal
size. Inset shows the van der Waals energy as a function of separation distance for
different nanocrystal sizes.
237
As the size of the nanocrystal increases from 2.2 nm to 3.6 nm for FePt nanocrystals
in this study, the magnitude of the attractive charge-induced dipole interactions
increases significantly, and the net interaction potential becomes larger than (3/2)kBT.
At a separation distance of 2.7 nm, a distance greater than the ligand length, the total
interaction potential is high enough for the FePt nanocrystal to adsorb strongly with
the silica surface and cannot be easily desorbed by Brownian motion.
Figure A.5 shows the low resolution bright field TEM image for FePt in
mesoporous silica. The magnification is low and hence it is difficult to visualize the
FePt nanocomposite morphology. HAADF-STEM on the other hand provides much
better contrast and magnification for visualizing the FePt nanocrystals supported in
mesoporous silica. Figure A.6 shows the calibration curve for the FePt nanocrystals.
Table A.1 shows the loadings as a function of initial FePt concentration as
well as time of infusion. For a given concentration, there is no significant change in
loadings observed when time is increased from 10 min to 960 min. At low
concentration of 0.24 mg/ml, all the nanocrystals are adsorbed in the silica in the first
10 min, resulting in a clear supernatant. For higher concentration of 0.7 mg/ml, the
supernatant is not completely clear. Loadings increase slightly from 9 wt % to 11 wt
% when the time of infusion is increased from 10 min to 960 min. At 2 mg/ml, there
is no change in loadings after 16 h. Thus it can be concluded that the equilibrium
loadings of FePt in mesoporous silica are obtained in the order of minutes.
239
Calibration at 500 nm
y = 7.6529xR2 = 1
0
0.5
1
1.5
2
2.5
3
3.5
0 0.1 0.2 0.3 0.4 0.5
Conc (mg/ml)
Abs
orba
nce
Figure A.6
Calibration of UV-Vis absorbance of FePt.
240
Wt % loading of FePt-Oleic acid/Oleylamine
Initial Nanocrystal Concentration (mg/mL)
Time (min) 0.24 0.70 2.0
10 7.3 ± 0.2 8.9 ± 0.2 13.8 ± 1.3
30 ― 9.0 ± 0.4 ―
960 ― 11.4 ± 0.4 14.3 ± 0.8
Table A.1
Loading of FePt nanocrystals in mesoporous silica as a function of time and
concentration.
241
A.3 References
(1) Barrett, E. P.; Joyner, L. G.; Halenda, P. P. The determination of pore
volume and area distributions in porous substances. I. Computations from nitrogen
isotherms. Journal of the American Chemical Society 1951, 73, 373-380.
(2) Lyu, Y.-Y.; Yi, S. H.; Shon, J. K.; Chang, S.; Pu, L. S.; Lee, S.-Y.;
Yie, J. E.; Char, K.; Stucky, G. D.; Kim, J. M. Highly Stable Mesoporous Metal
Oxides Using Nano-Propping Hybrid Gemini Surfactants. Journal of the American
Chemical Society 2004, 126, 2310-2311.
(3) Ehrburger-Dolle, F.; Morfin, I.; Geissler, E.; Bley, F.; Livet, F.; Vix-
Guterl, C.; Saadallah, S.; Parmentier, J.; Reda, M.; Patarin, J.; Iliescu, M.;
Werckmann, J. Small-Angle X-ray Scattering and Electron Microscopy Investigation
of Silica and Carbon Replicas with Ordered Porosity. Langmuir 2003, 19, 4303-4308.
(4) Kresge, C. T.; Leonowickz, M. E.; Roth, W. J.; Beck, J. S. Nature
1992, 359, 710.
(5) Israelachvili, J. Intermolecular & Surface Forces; 2nd ed.; Academic
Press, Inc.: San Diego, 1992.
(6) Papirer, E. Adsorption on Silica Surfaces CRC, 2000; Vol. 90.
(7) Nir, S. Van der Waals interactions between surfaces of biological
interest. Progress in Surface Science 1977, 8, 1-58.
(8) Meli, L.; Li, Y.; Lim, K. T.; Johnston, K. P.; Green, P. F. Phase
Inversion in Diblock Copolymer Nanocomposite Films. In Preparation 2007.
(9) Gupta, G.; Shah, P. S.; Zhang, X.; Saunders, A. E.; Korgel, B. A.;
Johnston, K. P. Enhanced Infusion of Gold Nanocrystals into Mesoporous Silica with
Supercritical Carbon Dioxide. Chemistry of Materials 2005, 17, 6728-6738.
(10) Sun, S.; Fullerton, E. E.; Weller, D.; Murray, C. B. Compositionally
controlled FePt nanoparticle materials. IEEE Transactions on Magnetics 2001, 37,
1239-1243.
242
(11) Sun, S.; Murray, C. B.; Weller, D.; Folks, L.; Moser, A. Monodisperse
FePt nanoparticles and ferromagnetic FePt nanocrystal superlattices. Science
(Washington, D. C.) 2000, 287, 1989-1992.
243
Appendix B
Highly Stable and Active Pt-Cu Electrocatalysts for ORR
The contents of this Appendix are supplementary to the data shown in Chapter 6 on
the synthesis of bimetallic Pt-Cu electrocatalysts.
B.1 Characterization
N2 sorption analysis was performed on a Micromeritics Germini analyzer at -
196 °C (77 K). The specific surface area was calculated using the BET method from
the nitrogen adsorption data in the relative range (P/P0) of 0.06-0.30. The samples
were heated overnight at 100°C for degassing. The total pore volume was determined
from the amount of N2 uptake at P/P0 = 0.95. The pore size distribution was derived
from the adsorption branch of the isotherm based on the BJH model.
Low-resolution pictures of mesoporous silica, infused with gold nanocrystals,
were visualized by TEM on a Phillips EM280 microscope with a 4.5 Å point-to-point
resolution operated with a 80 kV operating voltage. High resolution transmission
electron microscopy (HRTEM) was performed using a JEOL 2010F TEM operating
at 200 kV. Pictures were obtained at the optimum defocus condition.
Nanocomposites were deposited from a dilute ethanol solution onto 200 mesh carbon-
coated copper TEM grids.
XPS was acquired using a Kratos AXIS Ultra DLD spectrometer equipped
with a monochromatic Al X-ray source (Al Kα, 1.4866 keV) and 20 eV path energy
for high resolution elemental scans and 160 eV for survey scans conducted at angle of
45°. The sample is prepared by depositing the catalysts (1-2 mg) on the double sided
copper tape and then entered in the vacuum chamber.
244
TGA was conducted using Perkin–Elmer TGA7 equipment under oxygen
flow. Samples were heated to 750 °C at the rate of 20 °C/min. Samples were held at
50 °C for 10 min before increasing the temperature.
B.2 Results and Discussion
Figure B.1 gives the adsorption-desorption isotherms for the three DMCs and
corresponding pore size distributions. For all the carbons, three well-distinguished
regions of the adsorption isotherm are evident: (i) monolayer- multilayer adsorption,
(ii) capillary condensation, and (iii) multilayer adsorption on the outer particle
surfaces. The adsorption isotherms show a type H1 behavior with capillary
condensation at P/Po ~ 0.6 for DMC-850 and DMC-2000, while DMC-2600 likely
shows a H2 hysteresis behavior. The slope and almost vertical step in the hysteresis
loops suggest well-defined mesopores with a narrow pore size distribution as shown
in Figure B.1b.
Figure B.2 gives the low resolution TEM micrographs of unsupported
Pt20Cu80 nanoparticles. The nanoparticles are relatively monodisperse with a average
diameter of 2.85±0.38 nm as seen in the histogram. The high resolution images show
lattice fringes indicating that the crystalline nature of nanoparticles. The lattice d-
spacing is 2.25Å and can be indexed for a Pt (111) phase.
Figure B.3(a) gives the absorbance of the Pt20Cu80 nanoparticles before and
after infusion after dilution by a common factor. After infusion, the absorbance of the
nanoparticle effluent solution is 23% lower than that before infusion, as the
nanoparticles leave the solution and adsorb onto the mesoporous carbon (DMC-
850).The initial concentration of nanoparticles was 20 mg/mL. The change in
concentration after infusion was determined by using a calibration curve generated by
measuring the absorbances at 800 nm for several dilutions of Pt20Cu80 nanoparticles at
245
0.0 0.2 0.4 0.6 0.8 1.00
50
100
150
200
250
300
350
400
450
500
550
Am
ount
Ads
orbe
d (c
m3
STP
/g)
P/P0
DMC 850 DMC 2000 DMC 2600
0 2 4 6 8 10 12 14 16 18 200
2
PSD
Pore Diameter (nm)
(a) (b)
Figure B.1
Nitrogen adsorption-desorption curves (a) and corresponding pore size distributions
(b) for the disordered mesoporous carbons.
246
(a) (b)
(c)
Figure B.2
TEM micrographs (A) Low resolution (B) High-resolution and (C) Histogram, of as-
synthesized Pt20Cu80 nanoparticles. The average nanoparticle size is 2.85±0.38 nm.
0
5
10
15
20
1.75 2 2.25 2.5 2.75 3 3.25 3.5 3.75
Cou
nt
PtCu Nanoparticle size
247
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
500 550 600 650 700 750 800
Wavelength [nm]
Abs
orba
nce
Before Infusion
After Infusion
(a)
0
0.5
1
1.5
2
2.5
0 0.2 0.4 0.6 0.8 1 1.2Concentration (mg/ml)
Abs
orba
nce
(b)
Figure B.3
(a) Effect of infusion of the Pt20Cu80 nanoparticles in DMC-850 on absorbance as
determined by UV-Visible spectrophotometry (b) Calibration curve of the UV-Visual
absorbance vs. concentration of Pt20Cu80 nanoparticles in toluene, measured at 800
nm.
248
known concentrations and was found to be 15.5 mg/mL. The absorbance-
concentration calibration curve is linear (Figure B.3 (b)) with a correlation coefficient
of 0.998.
Figure B.4 shows the dealloying curves for Pt20Cu80. Cycle 500 in Figure B.4
approaches pure Pt as shown in Figure B.5.
Figure B.6 shows the XPS spectra of the pure Pt nanoparticles of 3 nm size
supported on graphitized carbon. The binding energy for the Pt 4f7/2 peak is 72.1 eV
as seen in Figure B.6. The binding energy is shifted to higher level by 1.1 eV relative
to bulk unsupported Pt (71.0 eV) and by 0.8 eV relative to Pt supported on
commercial amorphous carbon (~71.4 eV. As the nanoparticle size for the Pt is ~ 3
nm, the shift in the binding energy to higher value can not be attributed to the particle
size effects, which are observed for 1-2 nm nanoparticles but become insignificant for
>2 nm nanoparticles. The binding energy shift to higher levels can hence be attributed
to the enhanced Pt-carbon interactions facilitated by the π sites charge transfer to the
Pt metal. It can hence be deduced that the metal-support interactions for Pt on
graphitic supports are stronger than Pt on amorphous carbons and play a key role in
enhancing the catalyst stability towards ADT potential cycling (Figure 6.7).
Figure B.7 shows the TGA curves for 20% Pt on DMC-850 and DMC-2000
mesoporous carbons. The onset of carbon oxidation for DMC-850 is gradual and
occurs at relatively low temperature of ~ 300 °C for DMC 850, and is comparable to
that of 20% Pt on vulcan–XC72 which also starts oxidizing at ~ 300 °C.1 DMC-2000
on the other hand does not show any carbon degradation for DMC 2000 until 600 °C.
Thus graphitic DMC-2000 is significantly more stable than DMC-850 as well as
Vulcan carbon towards carbon oxidation by Pt. High stability of the carbon enables
complete retention in activity upon ADT potential cycling for the dealloyed
Pt20Cu80 on DMC-2000 catalysts (as seen in Figure 6.7).
Figure B.8 shows the cyclic voltammograms of the dealloyed Pt80Cu20 on
DMC-2000 before and after ADT cycling. The cyclic voltammograms show two
249
Figure B.4
First 20 cyclic voltammetry profiles of Pt20Cu80 nanoparticles supported on DMC-
2000 during electrochemical dealloying at 20 mV/s in argon atmosphere. The black
arrows indicate the direction in which the peak shifts as cycling progresses.
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
0.000 0.250 0.500 0.750 1.000 1.250E (V) vs SHE
I (m
A/c
m2
geo)
Cycle 5Cycle 3Cycle 2Cycle 1 Cycle 6 Cycle 10
Cycle 20
250
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
0 0.2 0.4 0.6 0.8 1 1.2
E (vs NHE)
I (A
)
Figure B.5
Cyclic voltammogram of 20% monometallic Pt on DMC-2000 scanned at 100 mV/s
in argon atmosphere.
252
Figure B.7
TGA of Pt nanoparticles supported on disordered mesoporous carbons
Figure B.7
TGA of Pt nanoparticles supported on disordered mesoporous carbons.
0
20
40
60
80
100
120
0 200 400 600
Temperature (C)
Wei
ght % DMC 2000
DMC 850
Onset for DMC-2000
Onset for DMC-850
253
-1.2
-0.7
-0.2
0.3
0.8
1.3
0 0.25 0.5 0.75 1 1.25
E (vs NHE)
I (A)
Before ADT cycling After ADT cycling
Figure B.8
Cycling voltammograms for Pt20Cu80 on DMC 2000 before and after ADT potential
cycling
254
peaks in the underpotentially deposited hydrogen region between 0.05 and 0.3V
associated with the weak and strong desorption of hydrogen. The multiple crystal
facets present on the crystalline dealloyed Pt80Cu20 nanoparticles bind with hydrogen
with varying strength and hence lead to two desorption peaks. After ADT cycling
between potentials of 0.5V and 1.2V for 1000 cycles, the area under the hydrogen
desorption peak remains unchanged, indicating no loss in platinum area due to
sintering, coalescence, Pt dissolution or Ostwald ripening.
The potential range between 0.75- 0.85V where water activation leads to
surface Pt-OH formation according to Pt + H2O = Pt-OH + H+ + e- is shifted to higher
potentials of ~0.85V for dealloyed Pt80Cu20 as compared to that of pure Pt (0.8V,
Figure B.5). Anodic shift of the peak potentials for the dealloyed catalysts compared
to pure Pt suggested a delayed formation of Pt-OH surface species during anodic
sweeps. This upward shift of the Pt-OH onset potential for Pt alloy electrocatalysts is
either considered the cause for enhanced ORR activity2,3 or perceived more of an
effect of a reduced Pt-O chemisorption energy.4
The change in composition from Pt20Cu80 to ~Pt20Cu4 (≡ Pt85Cu15) during
dealloying may be compared with the shrinkage in nanoparticle size, by examining
the relevant lattice parameters. Assuming close packing of atoms, the ratio of final
volume (Vf) to initial volume (Vi), Vf/Vi= (NPt,f.aPt + NCu,f.aCu) /(NPt,i.aPt + NCu,i.aCu =
0.28) where the atomic volumes aPt = 1.51 cc/mole and aCu = 1.18 cc/mole and the
numbers of atoms NPt,i , NCu,i , NPt,f, NCu,f are 20, 80, 20, and 4, respectively.
Consequently, the particle diameter ration df/di = (Vf/Vi)1/3 = (1.245/0.349)1/3 = 0.66
which corresponds to a shrinkage of 34%. However, from XRD and TEM, the
shrinkage in the diameter is 10-12%. This discrepancy can be addressed by including
the effect of lattice restructuring and expansion during copper removal. The lattice
parameters were 3.610 Å for bulk Cu, 3.668 Å for the original Pt-Cu nanoparticles,
3.876 Å for the dealloyed particles and 3.923 Å for bulk Pt (Table 3). The net
expansion in the unit cell lattice is thus estimated to be about 8%, ((3.876/3.668)-
1)*100. Thus, the net estimated reduction of the nanoparticle size was 34% - 8% =
255
26%, whereas the experimentally observed reduction was 12%. The difference
reflects the experimental uncertainty in size measurement by TEM and lattice
parameters by XRD.
257
B.3 References
(1) Baturina, O. A.; Aubuchon, S. R.; Wynne, K. J. Thermal Stability in
Air of Pt/C Catalysts and PEM Fuel Cell Catalyst Layers. Chemistry of Materials
2006, 18, 1498-1504.
(2) Stamenkovic, V. R.; Fowler, B.; Mun, B. S.; Wang, G.; Ross, P. N.;
Lucas, C. A.; Markovic, N. M. Improved Oxygen Reduction Activity on Pt3Ni(111)
via Increased Surface Site Availability. Science (Washington, DC, United States)
2007, 315, 493-497.
(3) Stamenkovic, V. R.; Mun, B. S.; Arenz, M.; Mayrhofer, K. J. J.;
Lucas, C. A.; Wang, G.; Ross, P. N.; Markovic, N. M. Trends in electrocatalysis on
extended and nanoscale Pt-bimetallic alloy surfaces. Nature Materials 2007, 6, 241-
247.
(4) Norskov, J. K.; Rossmeisl, J.; Logadottir, A.; Lindqvist, L.; Kitchin, J.
R.; Bligaard, T.; Jonsson, H. Origin of the Overpotential for Oxygen Reduction at a
Fuel-Cell Cathode. Journal of Physical Chemistry B 2004, 108, 17886-17892.
258
Bibliography Ajayan, P. M.; Ebbesen, T. W. Reports on Progress in Physics 1997, 60, 1025-1062.
Ajayan, P. M.; Iijima, S. Nature (London, United Kingdom) 1993, 361, 333-4.
Akita, T.; Okumura, M.; Tanaka, K.; Kohyama, M.; Haruta, M. Catalysis Today
2006, 117, 62-68.
Alden, L. R.; Roychowdhury, C.; Matsumoto, F.; Han, D. K.; Zeldovich, V. B.; Abruna, H. D.; DiSalvo, F. J. Langmuir 2006, 22, 10465-10471.
Alivisatos, A. P. Science 1996, 271, 933-937.
Allum, K. G.; Hancock, R. D.; Howell, I. V.; Lester, T. E.; McKenzie, S.; Pitkethly,
R. C.; Robinson, P. J. Journal of Organometallic Chemistry 1976, 107, 393-
405.
Allum, K. G.; Hancock, R. D.; Howell, I. V.; Lester, T. E.; McKenzie, S.; Pitkethly,
R. C.; Robinson, P. J. Journal of Catalysis 1976, 43, 331-8.
Antolini, E. Materials Chemistry and Physics 2003, 78, 563-573.
Antolini, E. Journal of Materials Science 2003, 38, 2995-3005.
Antolini, E.; Giorgi, L.; Cardellini, F.; Passalacqua, E. Journal of Solid State
Electrochemistry 2001, 5, 131-140.
Antolini, E.; Salgado, J. R. C.; dos Santos, A. M.; Gonzalez, E. R. Electrochemical
and Solid-State Letters 2005, 8, A226-A230.
Antolini, E.; Salgado, J. R. C.; Giz, M. J.; Gonzalez, E. R. International Journal of
Hydrogen Energy 2005, 30, 1213-1220.
Antolini, E.; Salgado, J. R. C.; Gonzalez, E. R. Journal of Power Sources 2006, 160,
957-968.
Araki, H.; Fukuoka, A.; Sakamoto, Y.; Inagaki, S.; Sugimoto, N.; Fukushima, Y.;
Ichikawa, M. Journal of Molecular Catalysis A: Chemical 2003, 199, 95-102.
Argo, A. M.; Gates, B. C. Abstracts of Papers, 221st ACS National Meeting, San
Diego, CA, United States, April 1-5, 2001 2001, CATL-053.
259
Argo, A. M.; Odzak, J. F.; Lai, F. S.; Gates, B. C. Nature (London, United Kingdom)
2002, 415, 623-626.
Barrett, E. P.; Joyner, L. G.; Halenda, P. P. Journal of the American Chemical Society
1951, 73, 373-80.
Bartholin, M.; Graillat, C.; Guyot, A. Journal of Molecular Catalysis 1981, 10, 377-
84.
Batista, J.; Pintar, A.; Gomilsek, J. P.; Kodre, A.; Bornette, F. Applied Catalysis, A:
General 2001, 217, 55-68.
Baturina, O. A.; Aubuchon, S. R.; Wynne, K. J. Chemistry of Materials 2006, 18,
1498-1504.
Beard, B. C.; Ross, P. N., Jr. Journal of the Electrochemical Society 1990, 137, 3368-
74.
Beckman, E. J. Journal of Supercritical Fluids 2004, 28, 121-191.
Bell, A. T. Science (Washington, DC, United States) 2003, 299, 1688-1691.
Bell, P. W.; Anand, M.; Fan, X.; Enick, R. M.; Roberts, C. B. Langmuir 2005, 21,
11608-11613.
Bergeron, B. V.; Kelly, C. A.; Meyer, G. J. Langmuir 2003, 19, 8389-8394.
Blackburn, J. M.; Long, D. P.; Cabanas, A.; Watkins, J. J. Science FIELD Publication
Date:2001 Oct 5, 294, 141-5. FIELD Reference Number: FIELD Journal
Code:0404511 FIELD Call Number:.
Bockstaller, M. R.; Thomas, E. L. Journal of Physical Chemistry B 2003, 107, 10017-
10024.
Bond, G. C.; Rank, J. S. Proc. Intern. Congr. Catalysis, 3rd, Amsterdam, 1964 1965,
2, 1225-36, discussion 1236-7.
Bond, G. C.; Wells, P. B. Journal of Catalysis 1966, 5, 419-27.
Borodko, Y.; Ager, J. W., III; Marti, G. E.; Song, H.; Niesz, K.; Somorjai, G. A.
Journal of Physical Chemistry B 2005, 109, 17386-17390.
Brehm, A.; Busch, V.; Wilken, J. Chemie Ingenieur Technik 1993, 65, 573-5.
260
Brunauer, S.; Emmett, P. H.; Teller, E. Journal of the American Chemical Society
1938, 60, 309-19.
Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R. J. Chem. Soc.,
Chem. Commun. 1994, 7, 801-802.
Burns, R. C.; Graham, C.; Weller, A. R. M. Mol. Phys. 1986, 59, 41.
Cabanas, A.; Long, D. P.; Watkins, J. J. Chemistry of Materials 2004, 16, 2028-2033.
Cable, R. E.; Schaak, R. E. Chemistry of Materials 2005, 17, 6835-6841.
Campbell, C. T.; Parker, S. C.; Starr, D. E. Science (Washington, DC, United States)
2002, 298, 811-814.
Carlock, J. T. 1979, 4 pp.
Casado-Rivera, E.; Volpe, D. J.; Alden, L.; Lind, C.; Downie, C.; Vazquez-Alvarez,
T.; Angelo, A. C. D.; DiSalvo, F. J.; Abruna, H. D. Journal of the American
Chemical Society 2004, 126, 4043-4049.
Casula, M. F.; Corrias, A.; Falqui, A.; Serin, V.; Gatteschi, D.; Sangregorio, C.; de
Fernandez, C.; Battaglin, G. Chemistry of Materials 2003, 15, 2201-2207.
Chan, K.-Y.; Ding, J.; Ren, J.; Cheng, S.; Tsang, K. Y. Journal of Materials
Chemistry 2004, 14, 505-516.
Chao, M.-C.; Lin, H.-P.; Mou, C.-Y.; Cheng, B.-W.; Cheng, C.-F. Catalysis Today
2004, 97, 81-87.
Chen, J. Y.; Pan, F. M.; Chang, L.; Cho, A. T.; Chao, K. J. Journal of Vacuum
Science & Technology, B: Microelectronics and Nanometer Structures--
Processing, Measurement, and Phenomena 2005, 23, 2034-2040.
Chen, M.; Kim, J.; Liu, J. P.; Fan, H.; Sun, S. Journal of the American Chemical
Society 2006, 128, 7132-7133.
Chen, M.; Liu, J. P.; Sun, S. Journal of the American Chemical Society 2004, 126,
8394-8395.
Chen, M. S.; Goodman, D. W. Science (Washington, DC, United States) 2004, 306,
252-255.
261
Chen, W.; Cai, W.; Zhang, L.; Wang, G.; Zhang, L. J. Colloid and Interface Sci.
2001, 238, 291-295.
Chen, Y.; Yang, F.; Dai, Y.; Wang, W.; Chen, S. Journal of Physical Chemistry C
2008, 112, 1645-1649.
Chen, Z.; Deng, W.; Wang, X.; Yan, Y. ECS Transactions 2007, 11, 1289-1299.
Cho, A. Science (Washington, DC, United States) 2003, 299, 1684-1685.
Choi, W. C.; Woo, S. I.; Jeon, M. K.; Sohn, J. M.; Kim, M. R.; Jeon, H. J. Advanced
Materials (Weinheim, Germany) 2005, 17, 446-451.
Chou, N. H.; Schaak, R. E. Journal of the American Chemical Society 2007, 129,
7339-7345.
Chu, H. P.; Lei, L.; Hu, X.; Yue, P.-L. Energy & Fuels 1998, 12, 1108-1113.
Chytil, S.; Glomm, W. R.; Vollebekk, E.; Bergem, H.; Walmsley, J.; Sjoeblom, J.;
Blekkan, E. A. Microporous and Mesoporous Materials 2005, 86, 198-206.
Claus, P.; Brueckner, A.; Mohr, C.; Hofmeister, H. Journal of the American Chemical
Society 2000, 122, 11430-11439.
Coleman, N. R. B.; Morris, M. A.; Spalding, T. R.; Holmes, J. D. J. Am. Chem. Soc.
2001, 123, 187-188.
Coleman, N. R. B.; O'Sullivan, N.; Ryan, K. M.; Crowley, T. A.; Morris, M. A.;
Spalding, T. R.; Steytler, D. C.; Holmes, J. D. J. Am. Chem. Soc. 2001, 123,
7010.
Coleman, N. R. B.; Ryan, K. M.; Spalding, T. R.; Holmes, J. D.; Morris, M. A. Chem.
Phys. Lett. 2001, 343, 1.
Coloma, F.; Sepulveda-Escribano, A.; Fierro, J. L. G.; Rodriguez-Reinoso, F. Applied
Catalysis, A: General 1996, 148, 63-80.
Coloma, F.; Sepulveda-Escribano, A.; Rodriquez-Reinoso, F. Journal of Catalysis
1995, 154, 299-305.
Colvin, V. L.; Schlamp, M. C.; Alivisatos, A. P. Nature 1994, 370, 354-357.
Cuesta, A.; Dhamelincourt, P.; Laureyns, J.; Martinez-Alonso, A.; Tascon, J. M. D.
Carbon 1994, 32, 1523-32.
262
Cuong, N. T.; Chi, D. H.; Kim, Y.-T.; Mitani, T. Physica Status Solidi 2006, 13,
3472-3475.
Dabbousi, B. O.; Bawendi, M. G.; Onitsuka, O.; Rubner, M. F. Appl. Phys. Lett.
1995, 68, 1316-1318.
Darmstadt, H.; Roy, C.; Kaliaguine, S.; Choi, S. J.; Ryoo, R. Carbon 2002, 40, 2673-
2683.
de Soler-Illia, G. J.; Sanchez, C.; Lebeau, B.; Patarin, J. Chemical Reviews
(Washington, DC, United States) 2002, 102, 4093-4138.
Debenedetti, P. G.; Tom, J. W.; Kwauk, X.; Yeo, S. D. Fluid Phase Equilibria 1993,
82, 311-21.
DeSimone, J. M.; Maury, E. E.; Menceloglu, Y. Z.; McClain, J. B.; Romack, T. J.;
Combes, J. R. Science (Washington, DC, United States) 1994, 265, 356-9.
Diaz, E.; Ordonez, S.; Vega, A.; Coca, J. Microporous and Mesoporous Materials
2005, 77, 245-255.
Dickson, J. L.; Gupta, G.; Horozov, T. S.; Binks, B. P.; Johnston, K. P. Langmuir
2006, 22, 2161-2170.
Ding, J.; Chan, K.-Y.; Ren, J.; Xiao, F.-s. Electrochimica Acta 2005, 50, 3131-3141.
Dixon, D. J.; Bodmeier, R. A.; Johnston, K. P. AIChE J. 1993, 39, 127.
Dixon, D. J.; Johnston, K. P. AIChE Journal 1991, 37, 1441-1449.
Dresselhaus, M. S.; Thomas, I. L. Nature (London, United Kingdom) 2001, 414, 332-
337.
Dumas, J. M.; Rmili, S.; Barbier, J. Journal de Chimie Physique et de Physico-
Chimie Biologique 1998, 95, 1650-1665.
Eckert, C. A.; Liotta, C. L.; Bush, D.; Brown, J. S.; Hallett, J. P. Journal of Physical
Chemistry B 2004, 108, 18108-18118.
Ehrburger-Dolle, F.; Morfin, I.; Geissler, E.; Bley, F.; Livet, F.; Vix-Guterl, C.;
Saadallah, S.; Parmentier, J.; Reda, M.; Patarin, J.; Iliescu, M.; Werckmann, J.
Langmuir 2003, 19, 4303-4308.
Ertl, G.; Knozinger, H.; Weitkamp, J.; Editors Preparation of Solid Catalysts, 1999.
263
Fan, H.; Yang, K.; Boye Daniel, M.; Sigmon, T.; Malloy Kevin, J.; Xu, H.; Lopez
Gabriel, P.; Brinker, C. J. Science FIELD Full Journal Title:Science (New
York, N.Y.) FIELD Publication Date:2004 2004, 304, 567-71. FIELD
Reference Number: FIELD Journal Code:0404511 FIELD Call Number:.
Fan, J.; Boettcher, S. W.; Stucky, G. D. Chemistry of Materials 2006, 18, 6391-6396.
Farias, M. H.; Gellman, A. J.; Somorjai, G. A.; Chianelli, R. R.; Liang, K. S. Surface
Science 1984, 140, 181-96.
Fernandez, J. L.; Raghuveer, V.; Manthiram, A.; Bard, A. J. Journal of the American
Chemical Society 2005, 127, 13100-13101.
Ferreira, P. J.; la O, G. J.; Shao-Horn, Y.; Morgan, D.; Makharia, R.; Kocha, S.;
Gasteiger, H. A. Journal of the Electrochemical Society 2005, 152, A2256-
A2271.
Ferrer, D.; Blom, D. A.; Allard, L. F.; Mejia, S.; Perej-Tijerina, E.; Yacaman, M. J.
Journal of Materials Chemistry 2008, 18, 2442.
Ferrer, D.; Torres-Castro, A.; Gao, X.; Sepulveda-Guzman, S.; Ortiz-Mendez, U.;
Jose-Yacaman, M. Nano Letters 2007, 7, 1701-1705.
Folch, B.; Larionova, J.; Guari, Y.; Datas, L.; Guerin, C. Journal of Materials
Chemistry 2006, 16, 4435-4442.
Ford, D. M.; Glandt, E. D. J. Mem. Sci. 1995, 107, 47-57.
Freund, A.; Lang, J.; Lehmann, T.; Starz, K. A. Catalysis Today 1996, 27, 279-83.
Fu, Q.; Saltsburg, H.; Flytzani-Stephanopoulos, M. Science (Washington, DC, United
States) 2003, 301, 935-938.
Fu, X.; Wang, Y.; Wu, N.; Gui, L.; Tang, Y. Journal of Colloid and Interface Science
2001, 243, 326-330.
Fujdala, K. L.; Tilley, T. D. Journal of Catalysis 2003, 216, 265-275.
Fukuoka, A.; Araki, H.; Kimura, J.-i.; Sakamoto, Y.; Higuchi, T.; Sugimoto, N.;
Inagaki, S.; Ichikawa, M. Journal of Materials Chemistry 2004, 14, 752-756.
Fukuoka, A.; Araki, H.; Sakamoto, Y.; Sugimoto, N.; Tsukada, H.; Kumai, Y.;
Akimoto, Y.; Ichikawa, M. Nano Letters 2002, 2, 793-795.
264
Fukushima, Y.; Wakayama, H. J. Phys. Chem. B 1999, 103, 3062-3064.
Gallagher, P. M., Coffey, M.P., Krukonis, V.J. and Klasuts, N. 1989, 406, 334-354.
Gasteiger, H. A. Electrochemistry (Tokyo, Japan) 2007, 75, 103.
Gasteiger, H. A.; Kocha, S. S.; Sompalli, B.; Wagner, F. T. Applied Catalysis, B:
Environmental 2005, 56, 9-35.
Ghosh, A.; Ranjan Patra, C.; Mukherjee, P.; Sastry, M.; Kumar, R. Microporous and
Mesoporous Materials 2003, 58, 201-211.
Gierszal, K. P.; Kim, T.-W.; Ryoo, R.; Jaroniec, M. Journal of Physical Chemistry B
2005, 109, 23263-23268.
Glatter, O.; Kratky, O. Small Angle X-ray Scattering, Academic Press Inc: New-
York, 1982.
Gonzalez, O.; Lujano, J.; Pietri, E.; Goldwasser, M. R. Catalysis Today 2005, 107-
108, 436-443.
Goyal, H. B.; Gupta, A. K. Indian Journal of Technology 1989, 27, 360-2.
Greeley, J.; Krekelberg, W. P.; Mavrikakis, M. Angewandte Chemie, International
Edition 2004, 43, 4296-4300.
Grobert, N.; Mayne, M.; Walton, D. R. M.; Kroto, H. W.; Terrones, M.;
Kamalakaran, R.; Seeger, T.; Ruhle, M.; Terrones, H.; Sloan, J.; Dunin-
Borkowski, R. E.; Hutchison, J. L. Chemical Communications (Cambridge,
United Kingdom) 2001, 471-472.
Gross, A. Topics in Catalysis 2006, 37, 29-39.
Guinier, A.; Fournet, G. Small-Angle Scattering of X-Rays, Wiley: New-York, 1955.
Gupta, G.; Patel, M. N.; Ferrer, D.; Heitsch, A. T.; Korgel, B. A.; Johnston, K. P.
Chemistry of Materials 2008.
Gupta, G.; Shah, P. S.; Zhang, X.; Saunders, A. E.; Korgel, B. A.; Johnston, K. P.
Chemistry of Materials 2005, 17, 6728-6738.
Gupta, G.; Stowell, C. A.; Patel, M. N.; Gao, X.; Yacaman, M. J.; Korgel, B. A.;
Johnston, K. P. Chemistry of Materials 2006, 18, 6239-6249.
265
Ha, S.; Larsen, R.; Masel, R. I. Journal of Power Sources 2005, 144, 28-34.
Hall, S. C.; Subramanian, V.; Teeter, G.; Rambabu, B. Solid State Ionics 2004, 175,
809-813.
Haller, G. L. Journal of Catalysis 2003, 216, 12-22.
Han, Y.-J.; Kim, J. M.; Stucky, G. D. Chem. Mater. 2000, 12, 2068-2069.
Hanrahan, J. P.; Copley, M. P.; Ziegler, K. J.; Spalding, T. R.; Morris, M. A.;
Steytler, D. C.; Heenan, R. K.; Schweins, R.; Holmes, J. D. Langmuir 2005,
21, 4163-4167.
Haran, N. P.; Bhagade, S. S.; Nageshwar, G. D. Chemicals & Petro-Chemicals
Journal 1978, 9, 21-8.
Hayashi, T.; Tanaka, K.; Haruta, M. Journal of Catalysis 1998, 178, 566-575.
Hayek, K.; Goller, H.; Penner, S.; Rupprechter, G.; Zimmermann, C. Catalysis
Letters 2004, 92, 1-9.
Hermans, S.; Raja, R.; Thomas, J. M.; Johnson, B. F. G.; Sankar, G.; Gleeson, D.
Angewandte Chemie, International Edition 2001, 40, 1211-1215.
Hiemenz, P. C. Principles of Colloid and Surface Chemistry, Marcel Dekker, Inc.:
New York, 1986.
Hillenbrand, L. J.; Lacksonen, J. W. Journal of the Electrochemical Society 1965,
112, 245-9.
Hirai, T.; Okubo, H.; Komasawa, I. J. Colloid and Interface Sci. 2001, 235, 358-364.
Hirai, T.; Okubo, H.; Komasawa, I. J. Phys. Chem. B 1999, 103, 4228-4230.
Hirsch, P. B.; et al. Electron Microscopy of Thin Crystals, 1977.
Hodgkins, R. P.; Ahniyaz, A.; Parekh, K.; Belova, L. M.; Bergstroem, L. Langmuir
2007, 23, 8838-8844.
Hornyak, G.; Kroll, M.; Pugin, R.; Sawitowski, T.; Schmid, G.; Bovin, J.-O.;
Karsson, G.; Hofmeister, H.; Hopfe, S. Chemistry--A European Journal 1997,
3, 1951-1956.
Horvath, G.; Kawazoe, K. J. Chem. Engg. Japan 1983, 16, 470-475.
266
Hou, J. G.; Li, X.; Wang, H.; Wang, B. Journal of Physics and Chemistry of Solids
2000, 61, 995-998.
Hourri, A.; St-Arnaud, J. M.; Bose, T. K. Journal of Chemical Physics 1997, 106,
1780-1785.
Huang, M. H.; Choudrey, A.; Yang, P. Chemical Communications (Cambridge) 2000,
1063-1064.
Huang, S.-M.; He, B.-L. Reactive Polymers 1994, 23, 11-18.
Hyun, C.; Lee, D. C.; Israel, C.; Korgel, B. A.; de Lozanne, A. IEEE Transactions on
Magnetics 2006, 42, 3799-3802.
Iijima, S. Nature (London, United Kingdom) 1991, 354, 56-8.
Imagawa, Y.; Kojima, K.; Seto, H. 2005, 17 pp.
Israelachvili, J. Intermolecular & Surface Forces, Academic Press, Inc.: San Diego,
1992.
Iwaki, T.; Kakihara, Y.; Toda, T.; Abdullah, M.; Okuyama, K. Journal of Applied
Physics 2003, 94, 6807-6811.
Jana, N. R.; Peng, X. Journal of the American Chemical Society 2003, 125, 14280-
14281.
Janssen, A. H.; Yang, C. M.; Wang, Y.; Schueth, F.; Koster, A. J.; De Jong, K. P.
Journal of Physical Chemistry B 2003, 107, 10552-10556.
Jarmer, D. J.; Lengsfeld, C. S.; Randolph, T. W. Journal of Supercritical Fluids 2003,
27, 317-336.
Jeyadevan, B.; Hobo, A.; Urakawa, K.; Chinnasamy, C. N.; Shinoda, K.; Tohji, K.
Journal of Applied Physics 2003, 93, 7574-7576.
Jiang, Y.-X.; Weng, W.-Z.; Si, D.; Sun, S.-G. Journal of Physical Chemistry B 2005,
109, 7637-7642.
Johnston, K. P.; Shah, P. S. Science (Washington, DC, United States) 2004, 303, 482-
483.
267
Jongste, H. C. d.; Kuijers, F. J.; Ponec, V. Preparation of Catalysts, Elsevier:
Amsterdam, 1976.
Joo, S. H.; Choi, S. J.; Oh, I.; Kwak, J.; Liu, Z.; Terasaki, O.; Ryoo, R. Nature
(London, United Kingdom) 2001, 412, 169-172.
Joo, S. H.; Choi, S. J.; Oh, I.; Kwak, J.; Liu, Z.; Terasaki, O.; Ryoo, R. Nature
(London, United Kingdom) 2001, 414, 470.
Jung, J. S.; Lim, J. H.; Malkinski, L.; Vovk, A.; Choi, K. H.; Oh, S. L.; Kim, Y. R.;
Jun, J. H. Journal of Magnetism and Magnetic Materials 2007, 310, 2361-
2363.
Kanaras, A. G.; Soennichsen, C.; Liu, H.; Alivisatos, A. P. Nano Letters 2005, 5,
2164-2167.
Kaneda, M.; Tsubakiyama, T.; Carlsson, A.; Sakamoto, Y.; Ohsuna, T.; Terasaki, O.;
Joo, S. H.; Ryoo, R. Journal of Physical Chemistry B 2002, 106, 1256-1266.
Kang, S.; Jia, Z.; Zoto, I.; Reed, D.; Nikles, D. E.; Harrell, J. W.; Thompson, G.;
Mankey, G.; Krishnamurthy, V. V.; Porcar, L. Journal of Applied Physics
2006, 99, 08N704/1-08N704/3.
Kawi, S.; Gates, B. C. Journal of Catalysis 1994, 149, 317-25.
Kelley, S. P.; Lemert, R. M. AIChE Journal 1996, 42, 2047-2056.
Khlobystov, A. N.; Britz, D. A.; Wang, J.; O'Neil, S. A.; Poliakoff, M.; Briggs, G. A.
D. Journal of Materials Chemistry 2004, 14, 2852-2857.
Kim, K.-M.; Kyu, K.-S.; Park, N.-G.; Park, Y.-J.; Chung, S.-H. 2002, 6 pp.
Kim, K. T.; Kim, Y. G.; Chung, J. S. Journal of the Electrochemical Society 1995,
142, 1531-8.
Kim, T.-W.; Kleitz, F.; Paul, B.; Ryoo, R. Journal of the American Chemical Society
2005, 127, 7601-7610.
Kim, T.-W.; Solovyov, L. A. Journal of Materials Chemistry 2006, 16, 1445-1455.
Kim, Y.-G.; Garcia-Martinez, J. C.; Crooks, R. M. Langmuir 2005, 21, 5485-5491.
268
Kim, Y.-J.; Kang, J. G.; Lim, S.-H.; Song, J. H. Bulletin of the Korean Chemical
Society 2005, 26, 1129-1131.
King, N. C.; Blackley, R. A.; Wears, M. L.; Newman, D. M.; Zhou, W.; Bruce, D. W.
Chemical Communications (Cambridge, United Kingdom) 2006, 3414-3416.
Kinoshita, K. Carbon: Electrochemical and Physicochemical Properties, John Wiey
and Sons: 1988.
Kinoshita, K. Journal of the Electrochemical Society 1990, 137, 845-8.
Klee, D.; Hilgers, C. 2002, 26 pp.
Klie, R. F.; Disko, M. M.; Browning, N. D. Journal of Catalysis 2002, 205, 1-6.
Knecht, M. R.; Garcia-Martinez, J. C.; Crooks, R. M. Langmuir 2005, 21, 11981-
11986.
Kockrick, E.; Krawiec, P.; Schnelle, W.; Geiger, D.; Schappacher, F. M.; Poettgen,
R.; Kaskel, S. Advanced Materials (Weinheim, Germany) 2007, 19, 3021-
3026.
Koh, S.; Leisch, J.; Toney, M. F.; Strasser, P. Journal of Physical Chemistry C 2007,
111, 3744-3752.
Koh, S.; Strasser, P. Journal of the American Chemical Society 2007, 129, 12624-
12625.
Koh, S.; Toney, M. F.; Strasser, P. Electrochimica Acta 2007, 52, 2765-2774.
Koh, S.; Yu, C.; Strasser, P. ECS Transactions 2007, 11, 205-215.
Komiyama, M.; Shimaguchi, T. Surface and Interface Analysis 2001, 32, 189-192.
Kongkanand, A.; Kuwabata, S.; Girishkumar, G.; Kamat, P. Langmuir 2006, 22,
2392-2396.
Kongkanand, A.; Martinez Dominguez, R.; Kamat, P. V. Nano Letters 2007, 7, 676-
680.
Konopny, L. W.; Juan, A.; Damiani, D. E. Applied Catalysis, B: Environmental 1998,
15, 115-127.
269
Konya, Z.; Puntes, V. F.; Kiricsi, I.; Zhu, J.; Ager, J. W., III; Ko, M. K.; Frei, H.;
Alivisatos, P.; Somorjai, G. A. Chemistry of Materials 2003, 15, 1242-1248.
Konya, Z.; Puntes, V. F.; Kiricsi, I.; Zhu, J.; Alivisatos, P.; Somorjai, G. A. Catalysis
Letters 2002, 81, 137-140.
Konya, Z. P., Victor F.; Kiricsi, Imre; Zhu, Ji; Ager, Joel W., III; Ko, Moon Kyu;
Frei, Heinz; Alivisatos, Paul; Somorjai, Gabor A. Chem. Mater. 2003, 15,
1242-1248.
Konya, Z. P., Victor F.; Kiricsi, Imre; Zhu, Ji; Alivisatos, A. Paul; Somorjai, Gabor
A. Nano Letters 2002, 2, 907-910.
Konya, Z. Z., Ji; Szegedi, Agnes; Kiricsi, Imre; Alivisatos, Paul; Somorjai, Gabor A.
Chem. Commun. 2003, 3, 314-315.
Koone, N. D.; Zerda, T. W. Journal of Non-Crystalline solids 1995, 183, 243-251.
Korgel, B. A.; Fitzmaurice, D. Physical Review B: Condensed Matter and Materials
Physics 1999, 59, 14191-14201.
Korgel, B. A.; Fullam, S.; Connolly, S.; Fitzmaurice, D. Journal of Physical
Chemistry B 1998, 102, 8379-8388.
Kresge, C. T.; Leonowickz, M. E.; Roth, W. J.; Beck, J. S. Nature 1992, 359, 710.
Kruk, M.; Jaroniec, M.; Ko, C. H.; Ryoo, R. Chemistry of Materials 2000, 12, 1961-
1968.
Krukonis, V. 1984.
Kuo, T.-C.; McCreery, R. L. Analytical Chemistry 1999, 71, 1553-1560.
Lagerev, S. P.; Babak, S. F. Zhurnal Obshchei Khimii 1937, 7, 1661-3.
Lai, F. S.; Gates, B. C. Nano Letters 2001, 1, 583-587.
Lake, L. W. Enhanced Oil Recovery, Prentice Hall, PTR: 1996.
Lang, H.; May, R. A.; Iversen, B. L.; Chandler, B. D. Journal of the American
Chemical Society 2003, 125, 14832-14836.
Le Bars, J.; Specht, U.; Bradley, J. S.; Blackmond, D. G. Langmuir 1999, 15, 7621-
7625.
270
Lee, D. C.; Mikulec, F. V.; Pelaez, J. M.; Koo, B.; Korgel, B. A. Journal of Physical
Chemistry B 2006, 110, 11160-11166.
Lee, J.; Kim, J.; Hyeon, T. Advanced Materials (Weinheim, Germany) 2006, 18,
2073-2094.
Lee, J.; Park, Y.; Kim, P.; Kim, H.; Yi, J. Journal of Materials Chemistry 2004, 14,
1050-1056.
Lensveld, D. J.; Mesu, J. G.; van Dillen, A. J.; de Jong, K. P. Studies in Surface
Science and Catalysis 2002, 143, 647-657.
Leonard, B. M.; Bhuvanesh, N. S. P.; Schaak, R. E. Journal of the American
Chemical Society 2005, 127, 7326-7327.
Lewis, J. E.; Biswas, R.; Robinson, A. G.; Maroncelli, M. Journal of Physical
Chemistry B 2001, 105, 3306-3318.
Li, H.; Wang, R.; Hong, Q.; Chen, L.; Zhong, Z.; Koltypin, Y.; Calderon-Moreno, J.;
Gedanken, A. Langmuir 2004, 20, 8352-8356.
Li, X.; Hsing, I. M. Electrochimica Acta 2006, 51, 3477-3483.
Li, Y.; Fu, Q.; Flytzani-Stephanopoulos, M. Applied Catalysis, B: Environmental
2000, 27, 179-191.
Liang, C.; Dai, S. Journal of the American Chemical Society 2006, 128, 5316-5317.
Liang, C.; Li, Z.; Dai, S. Angewandte Chemie, International Edition 2008, 47, 3696-
3717.
Lifshin, E. X-ray Characterization of Materials, Wiley-VCH: New York, 1999.
Lin, S.; Yin, C.-y.; Wang, L.-f.; Jiang, D.-z.; Xiao, F.-s. Shandong Daxue Xuebao,
Lixueban 2005, 40, 101-104.
Liu, S.-H.; Lu, R.-F.; Huang, S.-J.; Lo, A.-Y.; Chien, S.-H.; Liu, S.-B. Chemical
Communications (Cambridge, United Kingdom) 2006, 3435-3437.
Liu, Z.; Koh, S.; Yu, C.; Strasser, P. Journal of the Electrochemical Society 2007,
154, B1192-B1199.
Lu, X.; Ziegler, K. J.; Ghezelbash, A.; Johnston, K. P.; Korgel, B. A. Nano Letters
2004, 4, 969-974.
271
Lu, Y. Angewandte Chemie, International Edition 2006, 45, 7664-7667.
Maiwald, M.; Schneider, G. M. Berichte der Bunsen-Gesellschaft 1998, 102, 960-
964.
Maldonado, S.; Morin, S.; Stevenson, K. J. Carbon 2006, 44, 1429-1437.
Mani, P.; Srivastava, R.; Strasser, P. Journal of Physical Chemistry C 2008, 112,
2770-2778.
Mani, P.; Srivastava, R.; Yu, C.; Strasser, P. ECS Transactions 2007, 11, 933-939.
Mathias, M. F.; Makharia, R.; Gasteiger, H. A.; Conley, J. J.; Fuller, T. J.; Gittleman,
C. I.; Kocha, S. S.; Miller, D. P.; Mittelsteadt, C. K.; Xie, T.; Yan, S. G.; Yu,
P. T. Electrochemical Society Interface 2005, 14, 24-35.
Matsumoto, T.; Komatsu, T.; Nakano, H.; Arai, K.; Nagashima, Y.; Yoo, E.;
Yamazaki, T.; Kijima, M.; Shimizu, H.; Takasawa, Y.; Nakamura, J. Catalysis
Today 2004, 90, 277-281.
Matthey, D.; Wang, J. G.; Wendt, S.; Matthiesen, J.; Schaub, R.; Laegsgaard, E.;
Hammer, B.; Besenbacher, F. Science (Washington, DC, United States) 2007,
315, 1692-1696.
Mattoussi, H.; Cumming, A. W.; Murray, C. B.; Bawendi, M. G.; Ober, R. Journal of
Chemical Physics 1996, 105, 9890-9896.
Mattoussi, H.; Cumming, A. W.; Murray, C. B.; Bawendi, M. G.; Ober, R. Physical
Review B: Condensed Matter and Materials Physics 1998, 58, 7850-7863.
Mattoussi, H.; Radzilowski, L. H.; Dabbousi, B. O.; Thomas, E. L.; Bawendi, M. G.;
Rubner, M. F. J. Appl. Phys. 1998, 83, 7965-7974.
McClellan, A. K.; McHugh, M. A. Polymer Engineering and Science 1985, 25, 1088-
92.
McLeod, M. C.; Anand, M.; Kitchens, C. L.; Roberts, C. B. Nano Letters 2005, 5,
461-465.
Meli, L.; Li, Y.; Lim, K. T.; Johnston, K. P.; Green, P. F. In Preparation 2007.
Meunier, F. C.; Tibiletti, D.; Goguet, A.; Shekhtman, S.; Hardacre, C.; Burch, R.
Catalysis Today 2007, 126, 143-147.
272
Midgley, P. A.; Weyland, M. Ultramicroscopy 2003, 96, 413-431.
Midgley, P. A.; Weyland, M.; Thomas, J. M.; Johnson, B. F. G. Chemical
Communications (Cambridge, United Kingdom) 2001, 907-908.
Min, M.-K.; Cho, J.; Cho, K.; Kim, H. Electrochimica Acta 2000, 45, 4211-4217.
Minami, R.; Kitamoto, Y.; Chikata, T.; Kato, S. Electrochimica Acta 2005, 51, 864-
866.
Mohr, C.; Hofmeister, H.; Radnik, J.; Claus, P. Journal of the American Chemical
Society 2003, 125, 1905-1911.
Molnar, E. v.; Tasi, G.; Konya, Z.; Kiricsi, I. Catalysis Letters 2005, 101, 159-167.
Morey, M. S.; Davidson, A.; Stucky, G. D. Journal of Porous Materials 1998, 5, 195-
204.
Morgenstern, K.; Rosenfeld, G.; Poelsema, B.; Comsa, G. Physical Review Letters
1995, 74, 2058-2061.
Mukerjee, S. Journal of Applied Electrochemistry 1990, 20, 537-48.
Mukerjee, S.; Srinivasan, S. Journal of Electroanalytical Chemistry 1993, 357, 201-
24.
Mukerjee, S.; Srinivasan, S.; Soriaga, M. P. Journal of the Electrochemical Society
1995, 142, 1409-22.
Muraviev, D. N.; Macanas, J.; Parrondo, J.; Munoz, M.; Alonso, A.; Alegret, S.;
Ortueta, M.; Mijangos, F. Reactive & Functional Polymers 2007, 67, 1612-
1621.
Murray, C. B.; Norris, D. J.; Bawendi, M. G. J. Am. Chem. Soc. 1993, 115, 8706-
8715.
Nam, J.-H.; Jang, Y.-Y.; Kwon, Y.-U.; Nam, J.-D. Electrochemistry Communications
2004, 6, 737-741.
Napolsky, K. S.; Eliseev, A. A.; Knotko, A. V.; Lukahsin, A. V.; Vertegel, A. A.;
Tretyakov, Y. D. Materials Science & Engineering, C: Biomimetic and
Supramolecular Systems 2003, C23, 151-154.
273
Nemanich, R. J.; Solin, S. A. Physical Review B: Condensed Matter and Materials
Physics 1979, 20, 392-401.
Ng, H.-J.; Robinson, D. B. Journal of Chemical and Engineering Data 1978, 23, 325-
327.
Niesz, K.; Grass, M.; Somorjai, G. A. Nano Letters 2005, 5, 2238-2240.
Nir, S. Progress in Surface Science 1977, 8, 1-58.
Norskov, J. K.; Rossmeisl, J.; Logadottir, A.; Lindqvist, L.; Kitchin, J. R.; Bligaard,
T.; Jonsson, H. Journal of Physical Chemistry B 2004, 108, 17886-17892.
Oh, S.; Niu, Y.; Crooks, R. M. Langmuir 2005, 21, 10209-10213.
Ozkaya, D.; Zhou, W.; Thomas, J. M.; Midgley, P.; Keast, V. J.; Hermans, S.
Catalysis Letters 1999, 60, 113-120.
Pantea, D.; Darmstadt, H.; Kaliaguine, S.; Summchen, L.; Roy, C. Carbon 2001, 39,
1147-1158.
Papirer, E. Adsorption on Silica Surfaces CRC: 2000.
Paulus, U. A.; Wokaun, A.; Scherer, G. G.; Schmidt, T. J.; Stamenkovic, V.;
Radmilovic, V.; Markovic, N. M.; Ross, P. N. Journal of Physical Chemistry
B 2002, 106, 4181-4191.
Pietron, J. J.; Stroud, R. M.; Rolison, D. R. Nano Letters 2002, 2, 545-549.
Pitkethly, R. C.; McKenzie, S.; Allum, K. G. 1974, 17 pp.
Platonova, O. A.; Bronstein, L. M.; Solodovnikov, S. P.; Yanovskaya, I. M.;
Obolonkova, E. S.; Valetsky, P. M.; Wenz, E.; Antonietti, M. Colloid and
Polymer Science 1997, 275, 426-431.
Ponec, V.; Bond, G. C. Cataysis by Metals and Alloys, Elsevier
Netherlands, 1995.
Qi, Z.-m.; Honma, I.; Zhou, H. Applied Physics Letters 2006, 88, 053503/1-053503/3.
Raghuveer, V.; Ferreira, P. J.; Manthiram, A. Electrochemistry Communications
2006, 8, 807-814.
274
Raghuveer, V.; Manthiram, A.; Bard, A. J. Journal of Physical Chemistry B 2005,
109, 22909-22912.
Raja, R.; Hermans, S.; Shephard, D. S.; Johnson, B. F. G.; Raja, R.; Sankar, G.;
Bromley, S.; Thomas, J. M. Chemical Communications (Cambridge) 1999,
1571-1572.
Raja, R.; Khimyak, T.; Thomas, J. M.; Hermans, S.; Johnson, B. F. G. Angewandte
Chemie, International Edition 2001, 40, 4638-4642.
Rajalakshmi, N.; Ryu, H.; Shaijumon, M. M.; Ramaprabhu, S. Journal of Power
Sources 2005, 140, 250-257.
Rajca, I. 1982, 4 pp.
Randolph, T. W.; Randolph, A. D.; Mebes, M.; Yeung, S. Biotechnology Progress
1993, 9, 429-435.
Rindfleisch, F.; DiNoia, T. P.; McHugh, M. A. Journal of Physical Chemistry 1996,
100, 15581-15587.
Rioux, R. M.; Song, H.; Hoefelmeyer, J. D.; Yang, P.; Somorjai, G. A. Journal of
Physical Chemistry B 2005, 109, 2192-2202.
Rolison, D. R. Science (Washington, DC, United States) 2003, 299, 1698-1702.
Roychowdhury, C.; Matsumoto, F.; Mutolo, P. F.; Abruna, H. D.; DiSalvo, F. J.
Chemistry of Materials 2005, 17, 5871-5876.
Ruban, A. V.; Skriver, H. L.; Norskov, J. K. Physical Review B: Condensed Matter
and Materials Physics 1999, 59, 15990-16000.
Rubio, J. E. F.; Arsuaga, J. M.; Taravillo, M.; Baonza, V. G.; Caceres, M.
Experimental Thermal and Fluid Science 2004, 28, 887-891.
Ryan, K. M.; Erts, D.; Olin, H.; Morris, M. S.; Holmes, J. D. J. Am. Chem. Soc. 2003,
125, 6284.
Ryoo, R.; Choi, M.; Choi, S. H. 2005, 13 pp.
Ryoo, R.; Joo, S. H. Studies in Surface Science and Catalysis 2004, 148, 241-260.
275
Ryoo, R.; Joo, S. H.; Jun, S. Journal of Physical Chemistry B 1999, 103, 7743-7746.
Ryoo, W.; Webber, S. E.; Johnston, K. P. Industrial & Engineering Chemistry
Research 2003, 42, 6348-6358.
Saito, A.; Foley, H. C. AIChE Journal 1991, 37, 429-36.
Salgado, J. R. C.; Antolini, E.; Gonzalez, E. R. Journal of Physical Chemistry B
2004, 108, 17767-17774.
Santiago, E. I.; Varanda, L. C.; Villullas, H. M. Journal of Physical Chemistry C
2007, 111, 3146-3151.
Saunders, A. E.; Korgel, B. A. Journal of Physical Chemistry B 2004, 108, 16732-
16738.
Saunders, A. E.; Shah, P. S.; Park, E. J.; Lim, K. T.; Johnston, K. P.; Korgel, B. A. J.
Phys. Chem. B 2004, 108, 15969-15975.
Saunders, A. E.; Sigman, M. B., Jr.; Korgel, B. A. Journal of Physical Chemistry B
2004, 108, 193-199.
Schlamp, M. C.; Peng, X.; Alivisatos, A. P. J. Appl. Phys. 1997, 82, 5837-5842.
Schlapka, A.; Lischka, M.; Gross, A.; Kasberger, U.; Jakob, P. Physical Review
Letters 2003, 91, 016101/1-016101/4.
Schlogl, R.; Abd Hamid Sharifah, B. Angew Chem Int Ed Engl 2004, 43, 1628-37.
Schueth, F. Chemistry of Materials 2001, 13, 3184-3195.
Schwartz, V.; Mullins, D. R.; Yan, W.; Chen, B.; Dai, S.; Overbury, S. H. Journal of
Physical Chemistry B 2004, 108, 15782-15790.
Scott, R. W. J.; Sivadinarayana, C.; Wilson, O. M.; Yan, Z.; Goodman, D. W.;
Crooks, R. M. Journal of the American Chemical Society 2005, 127, 1380-
1381.
Scott, R. W. J.; Wilson, O. M.; Crooks, R. M. Journal of Physical Chemistry B 2005,
109, 692-704.
Scott, R. W. J.; Wilson, O. M.; Crooks, R. M. Chemistry of Materials 2004, 16, 5682-
5688.
Segalman, R. A. Materials Science & Engineering, R: Reports 2005, R48, 191-226.
276
Sellin, M. F.; Webb, P. B.; Cole-Hamilton, T. J. Chem. Commum. 2001, 781-782.
Serp, P.; Hierso, J. C.; Feurer, R.; Kihn, Y.; Kalck, P.; Faria, J. L.; Aksoylu, A. E.;
Pacheco, A. M. T.; Figueiredo, J. L. Carbon 1999, 37, 527-530.
Shah, P. S.; Holmes, J. D.; Doty, R. C.; Johnston, K. P.; Korgel, B. A. J. Am. Chem.
Soc. 2000, 122, 4245-4246.
Shah, P. S.; Husain, S.; Johnston, K. P.; Korgel, B. A. J. Phys. Chem. B 2001, 105,
9433-9440.
Shah, P. S.; Husain, S.; Johnston, K. P.; Korgel, B. A. J. Phys. Chem. B 2002, 106,
12178-12185.
Shah, P. S.; Novick, B. J.; Hwang, H. S.; Lim, K. T.; Carbonell, R. G.; Johnston, K.
P.; Korgel, B. A. Nano Letters 2003, 3, 1671-1675.
Shao, Y.; Yin, G.; Gao, Y. Journal of Power Sources 2007, 171, 558-566.
Shao, Y.; Yin, G.; Gao, Y.; Shi, P. Journal of the Electrochemical Society 2006, 153,
A1093-A1097.
Shevchenko, E. V.; Talapin, D. V.; Kotov, N. A.; O'Brien, S.; Murray, C. B. Nature
(London, United Kingdom) 2006, 439, 55-59.
Shibata, G.; Deguchi, M.; Taomoto, A.; Ozaki, T. 2004, 8 pp.
Shukla, A. K.; Ravikumar, M. K.; Roy, A.; Barman, S. R.; Sarma, D. D.; Arico, A.
S.; Antonucci, V.; Pino, L.; Giordano, N. Journal of the Electrochemical
Society 1994, 141, 517-22.
Shyue, J.-J.; De Guire, M. R. Journal of the American Chemical Society 2005, 127,
12736-12742.
Sinfelt, J. H.; Taylor, W. F.; Yates, D. J. C. Journal of Physical Chemistry 1965, 69,
95-101.
Singh, U. K.; Albert Vannice, M. Journal of Catalysis 2000, 191, 165-180.
Sinha, A. K.; Seelan, S.; Tsubota, S.; Haruta, M. Topics in Catalysis 2004, 29, 95-
102.
277
Skaff, H.; Lin, Y.; Dinsmore, A.; Russell, T.; Emrick, T. Abstracts of Papers, 224th
ACS National Meeting, Boston, MA, United States, August 18-22, 2002 2002,
COLL-026.
Somorjai, G. A. Catalysis Letters 2005, 101, 1-3.
Somorjai, G. A.; Contreras, A. M.; Montano, M.; Rioux, R. M. Proceedings of the
National Academy of Sciences of the United States of America 2006, 103,
10577-10583.
Somorjai, G. A.; Rioux, R. M.; Grunes, J. Clusters and Nano-Assemblies: Physical
and Biological Systems, [International Symposium], Richmond, VA, United
States, Nov. 10-13, 2004 2005, 97-125.
Song, H.; Rioux, R. M.; Hoefelmeyer, J. D.; Komor, R.; Niesz, K.; Grass, M.; Yang,
P.; Somorjai, G. A. Journal of the American Chemical Society 2006, 128,
3027-3037.
Song, H.; Rioux, R. M.; Kim, F.; Somorjai, G. A.; Yang, P. Abstracts of Papers,
227th ACS National Meeting, Anaheim, CA, United States, March 28-April 1,
2004 2004, INOR-877.
Song, Z.; Cai, T.; Hanson, J. C.; Rodriguez, J. A.; Hrbek, J. Journal of the American
Chemical Society 2004, 126, 8576-8584.
Spange, S.; Simon, F.; Heublein, G.; Jacobasch, H. J.; Boerner, M. Colloid and
Polymer Science 1991, 269, 173-8.
Sra, A. K.; Ewers, T. D.; Schaak, R. E. Chemistry of Materials 2005, 17, 758-766.
Srivastana, R.; Mani, P.; Hahn, N.; Strasser, P. Angewandte Chemie, International
Edition 2007, 46, 8988-8991.
Stamenkovic, V. R.; Fowler, B.; Mun, B. S.; Wang, G.; Ross, P. N.; Lucas, C. A.;
Markovic, N. M. Science (Washington, DC, United States) 2007, 315, 493-
497.
Stamenkovic, V. R.; Mun, B. S.; Arenz, M.; Mayrhofer, K. J. J.; Lucas, C. A.; Wang,
G.; Ross, P. N.; Markovic, N. M. Nature Materials 2007, 6, 241-247.
Steele, B. C. H.; Heinzel, A. Nature (London, United Kingdom) 2001, 414, 345-352.
278
Steigerwalt, E. S.; Deluga, G. A.; Lukehart, C. M. Journal of Physical Chemistry B
2002, 106, 760-766.
Stevens, D. A.; Dahn, J. R. Carbon 2004, 43, 179-188.
Stevens, D. A.; Hicks, M. T.; Haugen, G. M.; Dahn, J. R. Journal of the
Electrochemical Society 2005, 152, A2309-A2315.
Stowell, C. A.; Korgel, B. A. Nano Letters 2005, 5, 1203-1207.
Strasser, P. Journal of Combinatorial Chemistry 2008, 10, 216-224.
Strasser, P.; Koh, S.; Yu, C. ECS Transactions 2007, 11, 167-180.
Strongin, D. R.; Carrazza, J.; Bare, S. R.; Somorjai, G. A. Journal of Catalysis 1987,
103, 213-15.
Subramanian, B.; Rajewski, R. A.; Snavely, K. J. Pharm. Sci. 1997, 86, 885-890.
Subramoney, S. Advanced Materials (Weinheim, Germany) 1998, 10, 1157-1171.
Sun, Q.; Sun, R.-A. Chemical Research in Chinese Universities 2002, 18, 307-310.
Sun, S. Advanced Materials (Weinheim, Germany) 2006, 18, 393-403.
Sun, S.; Fullerton, E. E.; Weller, D.; Murray, C. B. IEEE Transactions on Magnetics
2001, 37, 1239-1243.
Sun, S.; Murray, C. B.; Weller, D.; Folks, L.; Moser, A. Science (Washington, D. C.)
2000, 287, 1989-1992.
Sun, X.; Kang, S.; Harrell, J. W.; Nikles, D. E.; Dai, Z. R.; Li, J.; Wang, Z. L. Journal
of Applied Physics 2003, 93, 7337-7339.
Tanev, P. T.; Pinnavaia, T. J. Chemistry of Materials 1996, 8, 2068-2079.
Tanev, P. T.; Pinnavaia, T. J. Science (Washington, D. C.) 1995, 267, 865-7.
Tang, Z.; Geng, D.; Lu, G. J Colloid Interface Sci 2005, 287, 159-66. .
Taylor, W. F.; Sinfelt, J. H.; Yates, D. J. C. Journal of Physical Chemistry 1965, 69,
3857-63.
Thomas, J. M.; Johnson, B. F. G.; Raja, R.; Sankar, G.; Midgley, P. A. Accounts of
Chemical Research 2003, 36, 20-30.
Thomas, L.; Lionti, F.; Ballou, R.; Gatteschi, D.; Sessoli, R.; Barbara, B. Nature
(London) 1996, 383, 145-147.
279
Tian, J. M.; Wang, F. B.; Shan, Z. H. Q.; Wang, R. J.; Zhang, J. Y. Journal of Applied
Electrochemistry 2004, 34, 461-467.
Tian, Z. Q.; Jiang, S. P.; Liu, Z.; Li, L. Electrochemistry Communications 2007, 9,
1613-1618.
Tibiletti, D.; Fonseca, A. A.; Burch, R.; Chen, Y.; Fisher, J. M.; Goguet, A.;
Hardacre, C.; Hu, P.; Thompsett, D. J. Phys. Chem. B 2005, 109, 22553-22559.
Tsang, S. C.; Chen, Y. K.; Harris, P. J. F.; Green, M. L. H. Nature (London) 1994,
372, 159-62.
Tsubota, S.; Cunningham, D. A. H.; Bando, Y.; Haruta, M. Studies in Surface Science
and Catalysis 1995, 91, 227-35.
Tsubota, S.; Haruta, M.; Kobayashi, T.; Ueda, A.; Nakahara, Y. Studies in Surface
Science and Catalysis 1991, 63, 695-704.
Tsung, C.-K.; Hong, W.; Shi, Q.; Kou, X.; Yeung, M. H.; Wang, J.; Stucky, G. D.
Advanced Functional Materials 2006, 16, 2225-2230.
Tuinstra, F.; Koenig, J. L. Journal of Chemical Physics 1970, 53, 1126-30.
Valden, M.; Lai, X.; Goodman, D. W. Science (Washington, D. C.) 1998, 281, 1647-
1650.
Van Bergen, F.; Gale, J.; Damen, K. J.; Wildenborg, A. F. B. Energy (Amsterdam,
Netherlands) 2004, 29, 1611-1621.
Vasiliev, M. A. Journal of Physics D: Applied Physics 1997, 30, 3037-3070.
Vegard, L. Zeitschrift fuer Physik 1921, 5, 17-26.
Vegard, L. Zeitschrift fuer Physik 1921, 5, 393-5.
Venezia, A. M.; Liotta, L. F.; Deganello, G.; Schay, Z.; Guczi, L. Journal of
Catalysis 1999, 182, 449-455.
Vigne-Maeder, F.; Amrani, S. E.; Gelin, P. J. Catal. 1992, 134, 536.
Vijayaraghavan, G.; Stevenson, K. J. Langmuir 2007, 23, 5279-5282.
Vondrova, M.; Klimczuk, T.; Miller, V. L.; Kirby, B. W.; Yao, N.; Cava, R. J.;
Bocarsly, A. B. Chemistry of Materials 2005, 17, 6216-6218.
280
Wagner, F. T.; Gasteiger, H. A.; Makharia, R.; Neyerlin, K. C.; Thompson, E. L.;
Yan, S. G. ECS Transactions 2006, 3, 19-29.
Waje, M. M.; Li, W.; Chen, Z.; Larsen, P.; Yan, Y. ECS Transactions 2007, 11,
1227-1233.
Wakabayashi, N.; Takeichi, M.; Uchida, H.; Watanabe, M. Journal of Physical
Chemistry B 2005, 109, 5836-5841.
Wakayama, H.; Fukushima, Y. Ind. Eng. Chem. Res. 2000, 39, 4641-4645.
Wakayama, H.; Fukushima, Y. Industrial & Engineering Chemistry Research 1344,
ACS ASAP.
Wang, A.-Q.; Chang, C.-M.; Mou, C.-Y. Journal of Physical Chemistry B 2005, 109,
18860-18867.
Wang, C.; Daimon, H.; Onodera, T.; Koda, T.; Sun, S. Angewandte Chemie,
International Edition 2008, 47, 3588-3591.
Wang, C.; Zhu, G.; Li, J.; Cai, X.; Wei, Y.; Zhang, D.; Qiu, S. Chemistry--A
European Journal 2005, 11, 4975-4982.
Wang, J.; Yin, G.; Shao, Y.; Wang, Z.; Gao, Y. Journal of Physical Chemistry C
2008, 112, 5784-5789.
Wang, K.; Morris, M. A.; Holmes, J. D. Chemistry of Materials 2005, 17, 1269-1271.
Wang, X. Q.; Liang, C.; Dai, S. Langmuir 2008, in press.
Wang, Z. L.; Petroski, J. M.; Green, T. C.; El-Sayed, M. A. Journal of Physical
Chemistry B 1998, 102, 6145.
Wark, M.; Wellmann, H.; Rathousky, J. Thin Solid Films 2004, 458, 20-25.
Watkins, J. J.; Blackburn, J. M.; McCarthy, T. J. Chemistry of Materials 1999, 11,
213-215.
Wei, M.; Musie, G. T.; Busch, D. H.; Subramanian, B. Green Chem. 2004, 6, 387-
393.
Wei, M.; Musie, G. T.; Busch, D. H.; Subramanian, B. J. Am. Chem. Soc. 2002, 124,
2513-2517.
281
Weimann, T.; Geyer, W.; Hinze, P.; Stadler, V.; Eck, W.; Golzhauser, A.
Microelectronic Engineering 2001, 57-58, 903-907.
Wikander, K.; Ekstroem, H.; Palmqvist, A. E. C.; Lundblad, A.; Holmberg, K.;
Lindbergh, G. Fuel Cells (Weinheim, Germany) 2006, 6, 21-25.
Wikander, K.; Petit, C.; Holmberg, K.; Pileni, M.-P. Langmuir 2006, 22, 4863-4868.
Williams, D. B.; Carter, C. B. Transmission Electon Microscopy, Springer: New
York, 1996.
Williams, J. W.; Krchma, I. J. J. Am. Chem. Soc. 1926, 48, 1888-96.
Wynne, K. J.; Shenoy, S.; Fujiwara, T.; Irie, S.; Woerdeman, D.; Sebra, R.; Garach,
A.; McHugh, M. Polymer Preprints (American Chemical Society, Division of
Polymer Chemistry) 2002, 43, 888.
Xie, X.; Liotta, C. L.; Eckert, C. A. Ind. Eng. Chem. Res. 2004, 43, 7907-7911.
Xiong, L.; Manthiram, A. Journal of the Electrochemical Society 2005, 152, A697-
A703.
Xiong, L.; Manthiram, A. Electrochimica Acta 2005, 50, 2323-2329.
Xu, Y.; Ruban, A. V.; Mavrikakis, M. Journal of the American Chemical Society
2004, 126, 4717-4725.
Xu, Z.; Xiao, F. S.; Purnell, S. K.; Alexeev, O.; Kawi, S.; Deutsch, S. E.; Gates, B. C.
Nature (London) 1994, 372, 346-8.
Yan, X. M.; Kwon, S.; Contreras, A. M.; Koebel, M. M.; Bokor, J.; Somorjai, G. A.
Catalysis Letters 2005, 105, 127-132.
Yang, C.-m.; Kalwei, M.; Schuth, F.; Chao, K.-j. Applied Catalysis, A: General 2003,
254, 289-296.
Yang, D. Q.; Sacher, E. Surface Science 2002, 516, 43-55.
Yang, D. Q.; Zhang, G. X.; Sacher, E.; Jose-Yacaman, M.; Elizondo, N. Journal of
Physical Chemistry B 2006, 110, 8348-8356.
Yano, H.; Kataoka, M.; Yamashita, H.; Uchida, H.; Watanabe, M. Langmuir 2007,
23, 6438-6445.
282
Ye, H.; Crooks, R. M. Journal of the American Chemical Society 2007, 129, 3627-
3633.
Ye, H.; Scott Robert, W. J.; Crooks Richard, M. Langmuir 2004, 20, 2915-20. .
Yeo, S.-D.; Debenedetti, P. G.; Radosz, M.; Schmidt, H.-W. Macromolecules 1993,
26, 6207-6210.
Yin, Y.; Alivisatos, A. P. Nature (London, United Kingdom) 2005, 437, 664-670.
Ying, J. Y.; Mehnert, C. P.; Wong, M. S. Angewandte Chemie, International Edition
1999, 38, 56-77.
Yoo, J. W.; Hathcock, D. J.; El-Sayed, M. A. Journal of Catalysis 2003, 214, 1-7.
Yu, R.; Song, H.; Zhang, X.-F.; Yang, P. Journal of Physical Chemistry B 2005, 109,
6940-6943.
Yu, X.; Ye, S. Journal of Power Sources 2007, 172, 145-154.
Zanella, R.; Giorgio, S.; Henry, C. R.; Louis, C. Journal of Physical Chemistry B
2002, 106, 7634-7642.
Zhang, G.; Yang, D.; Sacher, E. Journal of Physical Chemistry C 2007, 111, 565-570.
Zhang, J.-M.; Ma, F.; Xu, K.-W. Applied Surface Science 2004, 229, 34-42.
Zhang, J.; Han, B.; Hou, Z.; Liu, Z.; He, J.; Jiang, T. Langmuir 2003, 19, 7616-7620.
Zhang, J.; Lima, F. H. B.; Shao, M. H.; Sasaki, K.; Wang, J. X.; Hanson, J.; Adzic, R.
R. Journal of Physical Chemistry B 2005, 109, 22701-22704.
Zhang, J.; Sasaki, K.; Sutter, E.; Adzic, R. R. Science (Washington, DC, United
States) 2007, 315, 220-222.
Zhang, L.; He, R.; Gu, H.-C. Applied Surface Science 2006, 253, 2611-2617.
Zhang, W.-H.; Daly, B.; O'Callaghan, J.; Zhang, L.; Shi, J.-L.; Li, C.; Morris, M. A.;
Holmes, J. D. Chemistry of Materials 2005, 17, 6407-6415.
Zhang, Z.; Lagally, M. G. Science 1997, 276, 377-383.
Zhao, D.; Feng, J.; Huo, Q.; Melosh, N.; Frederickson, G. H.; Chmelka, B. F.;
Stucky, G. D. Science (Washington, D. C.) 1998, 279, 548-552.
Zhao, D.; Huo, Q.; Feng, J.; Chmelka, B. F.; Stucky, G. D. Journal of the American
Chemical Society 1998, 120, 6024-6036.
283
Zhao, D.; Yang, P.; Margolese, D. I.; Stucky, G. D. Chemical Communications
(Cambridge) 1998, 2499-2500.
Zhao, L.; Zhu, G.; Zhang, D.; Chen, Y.; Qiu, S. Journal of Solid State Chemistry
2005, 178, 2980-2986.
Zheng, N.; Stucky, G. D. Journal of the American Chemical Society 2006, 128,
14278-14280.
284
Vita
Gaurav Gupta was born in Gwalior, India to Ramesh Chandra and Archana
Gupta. Upon graduating from Jawahar Lal Nehru High School, Bhopal, India in 1999,
he enrolled at Indian Institute of Technlogy (IIT) in Mumbai, India. While at IIT
Mumbai, he conducted undergraduate research and a senior research project under the
direction of Dr. Jayesh Bellare. He received his B.Tech. in Chemical Engineering in
May of 2003. He enrolled at The University of Texas at Austin in September of 2003
for graduate studies. As a graduate student he conducted research under the
supervision of Dr. Keith P. Johnston. He will receive his Ph.D. in Chemical
Engineering in August 2008. He will join ExxonMobil Upstream research center as a
senior research scientist in Nov. 2008.
Permanent Address: 50, Amaltas Phase-2,
Chunna Bhatti, Kolar Road
Bhopal, Madhya Pradesh, Pin Code- 462016
India
This dissertation was typed by the author.