the rational design of selective electrocatalysts for renewable …349502/fulltext.pdf · the...
TRANSCRIPT
The Rational Design of Selective Electrocatalysts for Renewable Energy Devices
by Daniel F. Abbott
B.S. in Chemistry, Framingham State University
A dissertation submitted to
The Faculty of
the College of Science of
Northeastern University
in partial fulfillment of the requirements
for the degree of Doctor of Philosophy
March 24, 2015
Dissertation directed by
Sanjeev Mukerjee
Professor of Chemistry and Chemical Biology
ii
Acknowledgements
Firstly, I would like to thank Sanjeev Mukerjee for taking me into his lab as a young and
inexperienced researcher. Over the past five years I have had many incredible opportunities and
experiences due to your resourcefulness. You developed my skills as a researcher by challenging
my train of thought and you simultaneously provided me with opportunities and tools to better
myself. I am particularly grateful for the experiences provided at the Army Research Laboratory,
the many trips to Brookhaven National Labs, and for opportunity to study abroad in the Czech
Republic.
My deepest thanks to Petr Krtil for providing me with the opportunity to complete part of my
Ph.D. studies at the J. Heyrovský Institute in Czech Republic. Not only did you continuously
challenge my fundamental understanding of electrocatalysis, but you also helped me make the
most of my experience in the Czech Republic by understanding and supporting my desire to
travel and experience the culture while the opportunity was there. I would also like to extend my
thanks to Hana Hoffmannová and Antonín Trojánek among others for making my experience at
the Heyrovský Institute so memorable.
My most sincere gratitude to my parents, Leon and Cynthia Abbott. You have always been
supportive of my decisions throughout life and have always displayed such pride in my
achievements. I regret that my father is not here to see me today; I know how proud he would
have been. I would also like to thank my brother and sister, James and Emily Abbott, for being
the best siblings I could ask for.
iii
I would like to thank my dissertation committee members, David Budil and Max Diem, for their
time and advice over the years. Many thanks are also extended to the Department of Chemistry
and Chemical Biology, including Cara Shockley, Andrew Bean, and Alex Henriksen for their
continued help.
I would like to give special thanks and recognition to Kara Strickland, Michael Bates, and
Urszula Tylus. We've shared each other's highs and lows over the past 5 years and if it weren't
for the understanding and support of each other then I'm sure we would have had several (more)
mental breakdowns by now. I would also like to thank the rest of my colleagues, past and
present, for the their knowledge, help, and support, including Qingying Jia, Iromie Gunasekara,
Mehmet Nurullaah Ates, Gizem Aysal, Ryan Pavlicek, Shraboni Ghoshal, Jingkun Li, Nagappan
Ramaswamy, Christopher Allen, Matthew Trahan, Jaehee Hwang, William Fowle, and Robert
Allen.
I would like to thank my incredible roommates Chrystine Reilly, Louis DeVita, Noah Grabeel,
Karen Turner, and Rose Fieschko. You've made my living experience in Boston the best that it
could have been. Never did I imagine that I would be living with such a group of fun and caring
people. I'll always be grateful for our family dinners, nights on the porch, Sunday fundays, and
miscellaneous shenanigans on Forbes Street and around JP.
Without the support of my friends I'd be lost. I would like to express my deepest gratitude to my
closest friends Jordan Guertin, Steve Boulay, Kevin Leis, Andrew Fuhrman, David Manchester,
Sean Milan, Robert Duca, Patrick Flaherty, and Dawn Mahoney, for their continued friendship
iv
and unwavering support over the past decade and more. You've all been there to help me
celebrate the best of times and to pick me up during the worst of times. I cannot express how
much I appreciate you all.
I express my deepest gratitude to my friends abroad. Grant Philips and his lovely wife, Kateřina
Phillips, for their loving friendship, generosity, and hospitality during my stay in Czech
Republic. I met you both towards the beginning of my stay and you helped create the lasting
experience I returned with. I'm ever grateful for our many trips to Znojmo, whether for the
Vinobraní or for Christmas; you and your family always made me feel welcome. Anna Havlová
for her love and support throughout my last months in Prague and for joining me in Boston for as
long as you could. My travel companions, Esther Serrano, Liina Ränkel, and Pavel Havlíček for
the many adventures we shared throughout Czech Republic and beyond - and also for those still
to come! I also can't forget my good friends Anna Fodor, Jack Hudson, Peterz Gerhard, Phil
D'Esposito, Veronika Hanakova, and Cesar Figueroa for our numerous dinner dates, game
nights, and late nights.
Finally, I would like to recognize the influence of my former mentors. My highschool chemistry
teacher, Susan Seery, who inspired me pursue my B.S. in Chemistry with her extreme
enthusiasm for chemistry and the physical sciences. My undergraduate research advisor,
Catherine Dignam, for taking the time to work with me and exposing me to the world of
research. Carol Russell for convincing me to pursue my Ph.D. Her genuine concern for her
students and her love of physical chemistry undoubtedly helped lead me to where I am today.
v
Abstract of Dissertation
The rational design of electrocatalysts is paramount to the development of
electrochemical devices. In particular, modifications to the structure and electronic properties of
a particular catalyst can have a strong influence on the activity and selectivity towards various
electrochemical reactions or pathways. In many cases this can lead to a particular reaction
pathway being opened or closed, the formation of intermediates being stabilized or inhibited, the
adsorption of poisonous species being mitigated, or the removal of poisonous species being
promoted. In the this dissertation the design and characterization of catalysts for electrochemical
devices (fuel cells, electrolyzers, and hydrogen pumps) will be discussed with regards to
tailoring the selectivity in order to promote or inhibit certain electrochemical reactions. The
electrochemical reactions of primary interest will include the methanol oxidation reaction
(MOR), the oxygen reduction reaction (ORR), oxygen evolution reaction (OER), and hydrogen
oxidation reaction (HOR).
Chapter 2 introduces and discusses specific issues hindering the advancement of anion-
exchange membrane (AEM)-based direct alcohol fuel cells (DAFCs). Specifically, alkaline
DAFCs experience severe potential losses at low current densities in comparison to the proton
exchange membrane (PEM)-based analogues. Quaternary ammonium cations are known to
specifically adsorb to the catalyst surface and induce electrostatic effects that result in a loss of
active surface area, which limits OH- adsorption and leads to a significant drop in methanol
oxidation current. Characterization of the anode electrode-polymer electrolyte interface is
performed with specific attention paid to the adsorption of quaternary ammonium cations present
in the polymer electrolyte. This research also aims to develop and investigate catalysts that
exhibit improved methanol oxidation activity and high resistance to specific adsorption of
vi
quaternary ammonium in alkaline media. The developed Pt/NiPb/C catalyst shows a significantly
higher electrochemical activity of than that of the commercial Pt/C and PtRu/C electrocatalysts.
In Chapter 3 nanocrystalline ruthenium dioxide and doped ruthenia of the composition
Ru1-xMxO2 (M = Co, Ni, Zn) are prepared and the corresponding oxygen reduction activity and
selectivity is evaluated in alkaline media. In general, the ruthenium based oxides show a strong
preference towards the 2-electron oxygen reduction pathway to hydrogen peroxide at low
overpotentials. However, the selectivity shifts at higher overpotentials towards the complete 4-
electron reduction pathway to H2O. It is shown that Ni- and Co-doped ruthenia continue to
produce significant amounts of peroxide at high overpotentials whereas the Zn- and nondoped
materials prefer the 4-electron reduction pathway. DFT-based analyses on ruthenium based
oxides show that the suppression of the 4-electron reduction pathway on Ni and Co-doped
catalysts can be accounted for by the presence of the Ni and Co cations in the cus binding sites.
Chapter 4 again examines the selectivity of ruthenium based oxides. In this chapter
nanocrystalline Mg-doped ruthenium dioxide catalysts with the formula Ru1-xMgxO2 are
synthesized. The chlorine and oxygen evolution (CER/OER) selectivity of the synthesized
materials in chloride containing media is then related to the catalyst local structure. Through X-
ray absorption spectroscopy (XAS) it is shown that the magnesium ions are not distributed
homogeneously in the material but exist in Mg-rich clusters. The refinement of the Mg EXAFS
functions shows that the Mg-rich clusters contain Mg in a highly strained environment similar to
that of the rutile-type structure at low Mg concentrations (< 10%). As the Mg content is
increased (> 10%) the Mg environment shifts to an ilmenite-type inclusion. Although the Mg
containing catalysts show lower overall CER/OER activities compared with the non-doped
ruthenia, they exhibit a preference for the chlorine evolution process. The observed shift in
vii
selectivity is primarily attributed to the opening of a reaction pathway for chlorine evolution
associated with presence of Mg modified active sites.
In Chapter 5 the reformate tolerance of carbon-supported Pt-based alloys is investigated
under electrochemical hydrogen pump conditions. The CO and CO2 tolerance of the
electrocatalysts under HOR conditions is evaluated and correlated to the composition of the
catalyst. Specifically, the incorporation of sites more oxophilic than Pt at the catalyst surface
promote the activation of water at lower overpotentials. This leads to the removal of adsorbed
CO at lower overpotentials, therefore allowing for higher rates of hydrogen oxidation in the
presence of CO/CO2 and improving the overall cell performance.
Chapter 6 will summarize the conclusions of each research chapter while providing
further insight. Additional methods of experimentation capable of providing valuable
information and extending the current findings will be discussed.
viii
Table of Contents
Acknowledgements ii
Abstract of Dissertation v
Table of Contents viii
List of Figures xiii
List of Tables xvii
List of Abbreviations and Symbols xviii
Chapter 1 Introduction 1
1.1 Importance of Renewable Energy Technology 1
1.2 Electrochemical Energy Conversion and Purification Devices 2
1.2.1 Fuel Cells 2
1.2.2 Electrolyzers 5
1.2.3 Electrochemical Hydrogen Pump 7
1.3 Electrocatalysis 8
1.3.1 Electrochemistry Fundamentals 8
1.3.2 Electrical Double Layer 10
1.3.3 Cyclic Voltammetry 14
1.3.4 Rotating Ring Disk Electrode (RRDE) Method 16
1.4 Differential Electrochemical Mass Spectrometry (DEMS) 18
1.5 X-Ray Absorption Spectroscopy (XAS) 20
1.6 Density Functional Theory (DFT) 26
1.7 Scope of Dissertation 31
ix
1.8 References 33
Chapter 2 Analysis of Double Layer and Adsorption Effects at the Alkaline 38
Polymer Electrolyte-Electrode Interface and the Development of a
Quaternary Ammonium Poisoning Resistant Electrocatalyst for
Methanol Oxidation
2.1 Introduction 38
2.2 Experimental 41
2.2.1 Electrochemical Characterization of the Anode-Polymer 41
Electrolyte Interface
2.2.2 Preparation of Pt/NiPb/C Electrocatalyst 42
2.2.3 XRD and TEM Characterization 43
2.2.4 Electrochemical Characterization of Pt/NiPb/C 43
2.3 Results and Discussion 44
2.3.1 Electrochemical Characterization of the Anode-Polymer 44
Electrolyte Interface
2.3.2 XRD and TEM Characterization 52
2.3.3 Electrochemical Characterization of Pt/NiPb/C 54
2.4 Conclusions 61
2.5 Acknowledgements 62
2.6 References 63
Chapter 3 Oxygen Reduction on Nanocrystalline Ruthenia - Local 68
Structure Effects
3.1 Introduction 68
x
3.2 Experimental 70
3.2.1 Materials Preparation 70
3.2.2 XRD, XPS, and SEM Characterization 71
3.2.3 Electrochemical Measurements 72
3.2.4 DFT Analysis of Oxygen Reduction 73
3.3 Results and Discussion 73
3.3.1 XRD and SEM Characterization 73
3.3.2 Electrochemical Measurements 76
3.3.3 DFT Analysis of Oxygen Reduction 83
3.4 Conclusions 89
3.5 Acknowledgements 90
3.6 References 91
Chapter 4 Selective Chlorine Evolution Catalysts Based on Mg-Doped 95
Nanoparticulate Ruthenium Dioxide
4.1 Introduction 95
4.2 Experimental 97
4.3.1 Materials Preparation 97
4.3.2 XRD, XPS, and SEM Characterization 98
4.3.3 Local Structure Characterization 99
4.3.4 Electrochemical and DEMS Measurements 100
4.3 Results and Discussion 101
4.3.1 XRD, XPS, and SEM Characterization 101
4.3.2 Local Structure Characterization 105
xi
4.3.3 Electrochemical and DEMS Measurements 111
4.3.4 Discussion 116
4.4 Conclusions 118
4.5 Acknowledgements 119
4.6 References 120
Chapter 5 Reformate Tolerant Pt-based Catalysts for the 123
Electrochemical Hydrogen Pump
5.1 Introduction 123
5.2 Experimental 125
5.2.1 Catalyst Preparation 125
5.2.2 Physical Characterization 126
5.3.2 Electrochemical Cell Polarization Measurements 127
5.3 Results and Discussion 128
5.3.1 Physical Characterization 128
5.3.2 Cell Polarization Measurements 131
5.4 Conclusions 139
5.5 Acknowledgements 140
5.6 References 141
Chapter 6 Dissertation Summary, Conclusions, and Future Directions 146
6.1 Summary 146
6.2 Chapter Synopses and Future Directions 146
6.2.1 Chapter 2 - Analysis of Double Layer and Adsorption 146
Effects at the Alkaline Polymer Electrolyte-Electrode
xii
Interface and the Development of Quaternary
Ammonium Poisoning Resistant Electrocatalyst for
Methanol Oxidation
6.2.2 Chapter 3 - Oxygen Reduction on Nanocrystalline 148
Ruthenia - Local Structure Effects
6.2.3 Chapter 4 - Selective Chlorine Evolution Catalysts 149
Based on Mg-Doped Nanoparticulate Ruthenium
Dioxide
6.2.4 Chapter 5 - Reformate Tolerant Pt-based Catalysts 150
for the Electrochemical Hydrogen Pump
6.3 Concluding Remarks 151
6.4 References 153
xiii
List of Figures
Figure 1.2.1. (a) Schematic of a PEM-based H2/O2 fuel cell. (b) Schematic of an AAEM-based
H2/O2 fuel cell. .................................................................................................................................4
Figure 1.2.2. Schematic of an AAEM-based MeOH/O2 fuel cell. ..................................................5
Figure 1.2.3. Schematic of an acid-based water electrolysis cell. ..................................................6
Figure 1.2.4. Schematic of a PEM-based electrochemical hydrogen pump with a contaminated
H2 feed to the anode and producing purified H2 at the cathode. ......................................................8
Figure 1.3.1. Representation of the double-layer interface based upon the Gouy-Chapman-Stern
model showing the electrode surface (M), the Inner Helmholtz Plane (IHP), Outer Helmholtz
Plane (OHP), and Diffuse Layer. ..................................................................................................11
Figure 1.3.2. Cyclic voltammogram of a polished polycrystalline Pt disk electrode in Ar
saturated 0.1 M KOH. Scan rate of 20 mV s-1
. The anodic scan is shown in blue; the cathodic
scan is shown in red. ......................................................................................................................15
Figure 1.3.3. Cyclic voltammogram of a polished polycrystalline Pt disk electrode in Ar
saturated 0.1 M KOH + 10 mM K3Fe(CN)6. Scan rate of 20 mV s-1
. ...........................................16
Figure 1.3.4. Depiction of RRDE tip. ...........................................................................................17
Figure 1.4.1. (a) Schematic of an electrochemical cell designed for use with a DEMS apparatus,
(b) Photo of DEMS apparatus used, and (c) Flow chart of DEMS................................................20
Figure 1.5.1. Schematic of the typical XAS experimental setup. .................................................21
Figure 1.5.2. (a) Example of a normalized XAS spectrum for RuO2 showing both the XANES
and EXAFS regions; (b) EXAFS spectrum of RuO2 shown in k-space; (c) Fourier transform of k-
space EXAFS spectrum to R-space; (d) XANES spectrum of RuO2 showing the pre-edge and
absorption edge regions; (e) Local structure of RuO2 ((110) plane shown in blue) used to model
EXAFS data. ..................................................................................................................................23
Figure 1.5.3. Example of a single photoelectron backscattering event and a multiple
backscattering event that contribute to the oscillating EXAFS signal. ..........................................24
Figure 1.6.1. Example of free energy diagrams for the reduction of O2 on RuO2. The dotted line
represents the equilibrium potential of the reduction of O2 to H2O2. ............................................29
Figure 1.6.2. Volcano plot for the 2-electron (blue) and 4-electron (red) reduction of O2 to H2O2
and H2O, respectively, using the binding energy of OOH as a descriptor. The dotted lines
represent the equilibrium potentials for the reduction products. ...................................................31
Figure 2.1. Percent loss in current density as a function of contaminant concentration in 0.1 M
KOH + 0.5 M MeOH + x mM contaminant. Glassy carbon disk electrode (0.247 cm2) with 5 μL
xiv
of 30% Pt/C ink deposited (Loading = 7.5 ugPt/cm2). Steady state currents obtained from
chronoamperometry at 0.6 V vs. RHE at 900 seconds. .................................................................45
Figure 2.2. Expected potential profile of anode double layer interface in NaOH, TMAOH, and
polymer electrolyte solutions when the electrode potential is more negative than PZC[16]. .......47
Figure 2.3. CV of Fe(CN)63-/4-
redox couple in Ar saturated 0.1M NaOH + 10mM K3Fe(CN)6
solution. Scans performed on glassy carbon disk electrode with 5 μL of 5 wt.% AS4 ionomer
deposited on surface. Geometric surface area = 0.247 cm2. Scan rate = 20 mV/s. ......................49
Figure 2.4. CV of Co(NH3)62+/3+
redox couple in Ar saturated 0.1M NaOH + 10mM
Co(NH3)6Cl3 solution. Scans performed on glassy carbon disk electrode with 5 μL of 5 wt.%
AS4 ionomer deposited. Geometric surface area = 0.247 cm2. Scan rate = 20 mV/s. .................50
Figure 2.5. Expected potential profile of anode double layer interface in the presence and
absence of anion exchange ionomer when the electrode potential is more positive than PZC. ....51
Figure 2.6. Powder XRD patterns for nanocrystalline Pt/C and Pt/NiPb/C samples....................53
Figure 2.7. TEM images of the Pt/NiPb/C catalyst. Insert is higher magnification. ....................54
Figure 2.8. Electrochemical measurement of Pt/C, PtRu/C, and Pt/NiPb/C catalysts in 0.1 M
KOH and 1.0 M methanol at 298 K. Current densities are normalized to the geometric surface
area. (a) CV measurements at a scan rate of 20 mV/s. (b) Chronoamperometric tests with 50 mV
potential steps and displaying the steady-state current value after 180 s. ....................................56
Figure 2.9. Cyclic voltammograms of 40% Pt/C and Pt/NiPb/C catalysts with Nafion or AS4
ionomer used as a catalyst binder. Scans were taken in Ar saturated 0.1 M KOH + 0.5 M
methanol at 298 K at a scan rate of 10 mV/s. Cyclic voltammograms collected at a scan rate of
10 mV/s in Ar saturated 0.1 M KOH are shown in the insert. ......................................................60
Figure 2.10. Electrochemical impedance spectroscopy of 40% Pt/C and Pt/NiPb/C with either
Nafion or AS4 ionomer used as a catalyst binder in 0.1 M KOH + 0.5 M MeOH. EIS data was
collected at E = 0.70 V vs. RHE from 32 kHz to 1 Hz with a amplitude of 10 mV. The insert
shows the high frequency resistance. .............................................................................................60
Figure 3.1. Powder X-ray diffraction patterns for nanocrystalline RuO2 and Ru1-xMxO2 (M = Ni,
Co, Zn) after annealing in air for 1 hour at 400 °C ........................................................................74
Figure 3.2. SEM images of nanocrystalline (a) RuO2, (b) Ru0.90Zn0.10O2, (c) Ru0.80Zn0.20O2, (d)
Ru0.95Ni0.05O2, (e) Ru0.90Ni0.10O2, and (f) Ru0.90Co0.10O2 after annealing at 400 °C in air for 1
hour. ..............................................................................................................................................75
Figure 3.2. ORR polarization curves and ring currents at 1600 rpm for RuO2 and Ru1-xMxO2 (M
= Ni, Co, Zn) electrodes at 20 mV s-1
in O2 saturated 0.1 M NaOH. Ering = 1.1 V vs. RHE. ........77
Figure 3.4. Potential dependence of the average number of electrons transferred during oxygen
reduction on RuO2 and Ru1-xMxO2 (M = Ni, Co, Zn) electrodes. Presented data were calculated
using Koutecky-Levich equation. ..................................................................................................79
xv
Figure 3.5. Phenomenological mechanism of oxygen reduction according to reference[41]. ......80
Figure 3.6. |ID/IR| vs. ω-1/2
plots for (a) RuO2, (b) Ru0.80Zn0.20O2, and (c) Ru0.90Ni0.10O2.
Presented data were extracted from RRDE experiments carried out in O2 saturated 0.1 M NaOH.
........................................................................................................................................................80
Figure 3.3. Potential dependence of the rate constants for the reduction of O2 to H2O (k1), of O2
to H2O2 (k2), and H2O2 to O2 (k3) on nanocrystalline ruthenia based catalysts. The presented data
correspond to experiments carried out in O2 saturated 0.1 M NaOH at 1600 rpm. .......................81
Figure 3.4. The potential of equal rate in 2- and 4-electron reduction for different ruthenia based
catalysts. .........................................................................................................................................82
Figure 3.5. Surface Pourbaix diagram for RuO2. Detailed description of the diagram
construction is given in the supplementary information. ...............................................................84
Figure 3.6. Surface Pourbaix diagram for Ni-doped RuO2. Detailed description of the diagram/s
construction is given in the supplementary information. ..............................................................85
Figure 3.11. Free energy diagrams for the reduction of O2 on three catalytic active sites, the Ru
cus site on conventional ruthenia (green), the cus Ru site on Ni doped RuO2 (magenta) and the
cus Ni site on Ni doped RuO2 (blue). The dotted line represents the equilibrium potential of the
reduction O2 to H2O2. The key difference is the binding of O on the Ni cus site compared to the
Ru cus sites. ...................................................................................................................................86
Figure 3.7. Volcano plot for the 2-electron (blue) and 4-electron (red) reduction of O2 to H2O2
and H2O, respectively, using the binding energy of OOH as a descriptor. The dotted lines
represent the equilibrium potentials for the reduction products. In the case of the Ni-doped
ruthenia the limiting over-potential for both possible reaction sites (Rucus and Nicus) are shown
along with that of conventional ruthenia. ......................................................................................87
Figure 4.1. Powder X-ray diffraction patterns for nanocrystalline RuO2 and Ru1-xMxO2 after
annealing in air for 1 hour at 400 °C. ..........................................................................................102
Figure 4.2. Survey scans (a) of Ru0.8Mg0.2O2 before (1) and after (2) electrochemical
experiments. Panes (b) and (c) contain fitted high resolution spectra of Ru 3d + C 1s (b) and Ru
4s + Mg 2s photoelectrons, respectively. The high resolution spectra of the Ru0.8Mg0.2O2 after
electrochemical experiments do not show significant differences from those plotted in panes (b)
and (c). .........................................................................................................................................103
Figure 4.3. SEM images of nanocrystalline (a) RuO2, (b) Ru0.95Mg0.05O2, (c) Ru0.90Mg0.10O2, and
(d) Ru0.80Mg0.20O2 after annealing in air for 1 hour at 400 °C. ....................................................105
Figure 4.4. EXAFS functions extracted from the X-ray absorption spectra of the Ru1-xMgxO2
(0<x<0.2) measured on the Ru K edge (a) and Mg K edge (b). Actual Mg content is shown in the
Figure legend. ..............................................................................................................................106
Figure 4.5. (a) A typical example of the non-linear least square fit of the Ru EXAFS function of
Ru0.95Mg0.05O2; (b) A typical example of the non-linear least square fit of the Ru EXAFS
xvi
function of Ru0.90Mg0.10O2. The square symbols represent the experimental data, the red line
denotes the best fit. .....................................................................................................................107
Figure 4.6. Linear scan voltammograms of the oxygen evolution on MgxRu1-xO2electrodes
(0<x<0.2) recorded in 0.1 M HClO4 at a polarization rate of 5 mV s-1
. The curve assignment is
given in the Figure legend............................................................................................................111
Figure 4.7. Chloride concentration dependence of the oxygen evolution (a) and chlorine
evolution (b) contributions to the overall current response of MgxRu1-xO2 electrodes to anodic
polarization in chloride containing acid media. The presented values correspond to potentiostatic
experiments at 1.25 V vs. Ag/AgCl. ...........................................................................................112
Figure 4.8. Composition dependence of the Mg doped ruthenia selectivity towards chlorine
evolution. The data correspond to potentiostatic experiments at 1.25 V (Ag/AgCl). The actual
chloride concentrations are given in the Figure legend. ..............................................................113
Figure 4.9. Potential dependence of the Mg doped ruthenia selectivity towards chlorine
evolution. The data were extracted from potentiostatic experiments in 0.1M HClO4 containing 10
mM NaCl (a) and 50 mM NaCl (b). ............................................................................................114
Figure 4.10. Time course of DEMS-based signals of potentiostatically generated oxygen (blue)
and chlorine (red) for RuO2 (a), Ru0.95Mg0.05O2 (b), Ru0.90Mg0.10O2 (c), and Ru0.80Mg0.20O2 (d).
Signals were recorded in 0.1 M HClO4 containing10 mM NaCl; the potential perturbation
corresponded to a step from 0.70 V to 1.25 V vs. Ag/AgCl. .......................................................115
Figure 5.1. Powder XRD patterns for electrocatalyst samples. ..................................................129
Figure 5.2. HRTEM images of the synthesized PtNi/C (a) and PtNiRu/C materials. ................131
Figure 5.3. Anode half-cell polarization curves of an electrochemical hydrogen pump with the
anode and cathode being fed pure H2. Polarization measurements were collected at a cell
temperature of 85°C, 100% relative humidity, and 25/30psi backpressure on the anode/cathode,
respectively. .................................................................................................................................132
Figure 5.4. Anode half-cell polarization curves in an electrochemical hydrogen pump with the
cathode being fed hydrogen and the anode being fed by hydrogen containing 300ppm CO (filled
circles) or a reformate gas mixture containing 100 ppm CO, 15% CO2, 1% CH4, 45% H2, and N2
balance (empty circles). Polarization measurements were collected at a cell temperature of 85°C,
100% relative humidity, and 25/30psi backpressure on the anode/cathode, respectively. ..........135
Figure 5.5. Anode half-cell polarization curves in an electrochemical hydrogen pump with the
cathode being fed pure hydrogen and the anode being fed by hydrogen containing 50% molar
CO2. Polarization measurements were collected at a cell temperature of 85°C, 100% relative
humidity, and 25/30psi backpressure on the anode/cathode, respectively. .................................138
xvii
List of Tables
Table 1.2.1. Typical operating conditions and properties for various types of fuel cells. .............2
Table 2.1. XRD and electrochemical results for various catalysts. ...............................................53
Table 3.1. Results of the analysis of the powder diffraction data for RuO2 and doped RuO2
samples. ..........................................................................................................................................75
Table 4.1. Results of the analysis of the powder diffraction data of the MgxRu1-xO2 catalysts .102
Table 4.2. XPS based surface metal content in Ru0.8Mg0.2O2. ....................................................103
Table 4.3. (a) Results of the NLLS fit of the EXAFS functions obtained from the Ru Kedge X-
ray absorption spectra of the MgxRu1-xO2 (0 < x < 0.2) catalysts. CN denotes the coordination
number and d stands for the bonding length; (b) Theoretically conceived local structures of the
Mg in rutile and ilmenite type MgRuO oxides and results of the NLLS fit of the EXAFS
functions obtained from the Mg K edge X-ray absorption spectra of the MgxRu1-xO2 (0<x<0.2)
catalysts. The theoretically conceived data are given in italics. Symbol assignment are the same
as in the case of Table 3a. ............................................................................................................108
Table 5.1. Results of XRD and HRTEM physical characterization. ...........................................130
Table 5.2. Performance and operating conditions of ECHPs reported in the literature ..............133
xviii
List of Abbreviations and Symbols
α Charge Transfer Coefficient, Position of Transition
State Along Reaction Coordinate
βi Width of Diffraction Peak at Half Max Intensity
η Electrochemical Overpotential
ω Rotation Rate
υ Kinematic Viscosity of the Electrolyte
ϕM Inner-potential of the Metal Electrode
ϕS Potential at the Diffuse Layer
ϕ1 Potential at the Inner-Helmholtz Plane
ϕ2 Potential at the Outer-Helmholtz Plane
θ Angle of Incidence/Coverage
λ Wavelength of incident radiation
μ Absorption Coefficient
μ(E) Absorption Coefficient at energy E
μ0(E) Absorption Coefficient at energy E0
χ(E) EXAFS function at energy E
χ(k) EXAFS function at wavenumber k
μm Micrometer (10-6
meters)
σ2 Debye-Waller Factor
Ω Ohms
δ Phase Shift
∆E Reaction Energy
ΔEp Peak Separation Potential
∆G°, ∆G0 Standard Gibbs Free Energy
∆G Gibbs Free Energy
∆GpH Gibbs Free Energy at pH, pH
∆GE Gibbs Free Energy at an electrode potential E
∆H° Standard Enthalpy Change
∆H Enthalpy of Reaction
∆S Entropy Change
∆ZPE Zero Point Energy Difference
Å Angstroms (10-10
meters)
A Geometric Electrode Area, Pre-exponential Factor
atm Atmospheric Pressure
aO Chemical Activity of Oxidant Species
aR Chemical Activity of Reductant Species
C Coulombs
°C Degrees Celsius
CO Concentration of Molecular Oxygen
Concentration of Hydroxide at OHP
Concentration of Hydroxide in bulk solution
cm centimeter
xix
D Diffusion Coefficient
Di Crystallite Domain Size
e Charge of an Electron
e- Electron
eV Electron Volts
E° Standard Electrode Potential
E Electrode Potential, or Energy
E0 Binding Energy, Activation Energy of Reference
Reaction
Ea Anodic Peak Potential, Activation Energy
Ec Cathodic Peak Potential
Eonset Onset Potential
Epeak Peak Potential
Erxn Standard Electrode Potential of Reaction
EPZC Electrode Potential at Potential of Zero Charge
ERing Ring Electrode Potential
EDisk Disk Electrode Potential
f Scattering Amplitude
F Faraday’s Constant
ħ Planck’s Constant
Hz Hertz
i Current Density
ik Kinetic Current Density
ilim Limiting Current Density
ipeak Peak Current Density
i0 Exchange Current Density
it0 True Exchange Current Density
iR Ring Current
iD Disk Current
I0 Intensity of Incident X-ray Beam
It Intensity of Transmitted X-ray Beam
Ir Intensity of Reference X-ray Beam
J Current Density
k photoelectron wavenumber, Boltzmann constant,
or rate constant
k0 Apparent Rate Constant
kt0 True Rate Constant
M Molarity
mA Milliamp
mV Millivolt
me Mass of an Electron
n Number of Electrons Transferred
nm Nanometer (10-9
meters)
N Coordination Number
ppm Part per Million
xx
psi Pounds per Square Inch
qM
Charge on the Metal Electrode
R Universal Gas Constant, Bond Length
T Temperature
V Volts
Welec Electrical Work
x Thickness of Sample
z Charge on an Ion
ads adsorbate
AAEM Alkaline Anion Exchange Membrane
AAEMFC Alkaline Anion Exchange Membrane Fuel Cell
AEI Anion Exchange Ionomer
AFC Alkaline Fuel Cell
ASE Atomic Simulation Environment
BET Brunauer–Emmett–Teller
CA Cyclic Amperometry
CE Counter Electrode
CER Chlorine Evolution Reaction
cus Coordinated Unsaturated Site
CV Cyclic Voltammetry
DEMS Differential Electrochemical Mass Spectrometry
DFT Density Functional Theory
DMFC Direct Methanol Fuel Cell
ECHP Electrochemical Hydrogen Pump
EDAX/EDS Energy Dispersive Analysis of X-rays
EIS Electrochemical Impedance Spectroscopy
EOR Ethanol Oxidation Reaction
EXAFS Extended X-ray Absorption Fine Structure
FC Fuel Cell
fcc Face Centered Cubic
FRA Frequency Response Analyzer
GC Glassy Carbon
GDE Gas Diffusion Electrode
GDL Gas Diffusion Layer
GCS Guoy-Chapman-Stern
hcp Hexagonal Close Packed
HOR Hydrogen Oxidation Reaction
HRTEM High Resolution Transmission Electron Microscope
HT Heat Treated
IHP Inner-Helmholtz Plane
MCFC Molten Carbonate Fuel Cell
MEA Membrane Electrode Assembly
MeOH Methanol
MOR Methanol Oxidation Reaction
MSCV Mass Spectrometric Cyclic Voltammogram
xxi
MWCNT Multi-wall Carbon Nanotubes
NASA National Aeronautics and Space Administration
NHE Normal Hydrogen Electrode
NSLS National Synchrotron Light Source
OCP Open Circuit Potential
OER Oxygen Evolution Reaction
OHP Outer-Helmholtz Plane
ORR Oxygen Reduction Reaction
PAFC Phosphoric Acid Fuel Cells
PBI Polybenzimidazole
PSA Pressure Swing Adsorption
PTFE Polytetrafluoroethylene
PZC Potential of Zero Charge
PZFC Potential of Zero Free Charge
PZTC Potential of Zero Total Charge
PEM Proton Exchange Membrane
PEMFC Proton Exchange Membrane Fuel Cell
QA+ Quaternary Ammonium Ion
RH Relative Humidity
RHE Reversible Hydrogen Electrode
RPM Revolutions Per Minute
RE Reference Electrode
RRDE Rotating Ring-Disk Electrode
SEM Scanning Electron Microscopy
SHE Standard Hydrogen Electrode
SOFC Solid Oxide Fuel Cell
TEM Transmission Electron Microscopy
upd Underpotential Deposition
WE Working Electrode
XANES X-ray Absorption Near Edge Spectroscopy
XAS X-ray Absorption Spectroscopy
XRD X-ray Diffraction
XPS X-ray Photoelectron Spectroscopy
1
Chapter 1
Introduction
1. Introduction
1.1 Importance of Renewable Energy Technology
The production of energy from clean and renewable sources is crucial to the future
development of human civilization. Given the fact that the burning of traditional fossil fuels is
contributing to the ever-growing issue of climate change and is generally unsustainable, the
adoption of renewable and sustainable energy sources is of paramount importance[1]. Hydrogen
is an alternative source of energy that can be obtained from several resources, such as the
electrolytic splitting of water or the reforming of hydrocarbons (e.g. natural gas, propane,
methanol). When hydrogen is produced and reclaimed using the energy obtained from renewable
sources such as wind, solar, hydro, etc. it offers a means of escaping the paradigm of burning
carbon-based fossil fuels. The clean energy expended to produce hydrogen can then later be
reclaimed using fuel cell technology for anything from portable electronics, to automotive
vehicles, to grid-scale power. This can ultimately lead to a reduction of CO2 emissions and oil
dependency. To develop and influence a hydrogen economy free of fossil fuels, efficient and cost
effective methods for hydrogen production, purification, storage, transport, and conversion are
imperative.
The following introductory sections provide a brief overview of the electrochemical
technologies targeting the production, purification, and consumption of hydrogen (electrolyzers,
electrochemical hydrogen pumps, and fuel cells). Moreover, they introduce the fundamental
concepts and techniques required for the development and characterization of electrocatalysts for
2
these devices. The primary focus of subsequent chapters will be on the rational design of
electrocatalysts tailored to address specific issues hindering the development of hydrogen-based
technologies.
1.2 Electrochemical Energy Conversion and Purification Devices
1.2.1 Fuel Cells
Fuel cells are electrochemical devices that convert chemical energy into electrical energy.
There are several major types of fuel cells, including Proton Exchange Membrane Fuel Cells
(PEMFCs), Phosphoric Acid Fuel Cells (PAFCs), Solid Oxide Fuel Cells (SOFCs), and Alkaline
Anion Exchange Membrane Fuel Cells (AAEMFCs). The electrolyte, charge carrier, and
operating temperature of a fuel cell are dependent upon the type of fuel cell being considered.
The major types of fuel cells are summarized in Table 1.2.1[2].
Table 1.2.1. Typical operating conditions and properties for various types of fuel cells.
PEMFC AAEMFC AFC PAFC MCFC SOFC
Electrolyte Polymer
membrane
Polymer
membrane
Liquid
KOH
Liquid
H3PO4
Molten
carbonate Ceramic
Operating Temp (°C) 80 60 60-220 200 650 600-1000
Charge Carrier H+ OH
- OH
- H
+ CO3
2- O2
-
One focus of the work presented in the following chapters is the electrocatalysis
encountered in direct methanol-based AAEMFCs. AAEMFCs offer several advantages over the
PEMFC analogues, including the use of non-noble metals, reduced electro-osmotic drag,
improved kinetics in the case of ORR and MOR, and the lack of carbonate precipitation issues[3-
6]. The progressive development of alkaline anion exchange polymers has led to a growing
interest in the development of AFCs that use an AEM in the past 10 years. This technology
3
clearly surpasses the liquid KOH electrolyte used in previous systems[7, 8], including those used
by NASA in the Gemini space programs[9, 10].
Both PEMFCs and AAEMFCs can use H2 (or small alcohols) as a fuel and O2 or air as an
oxidant. The difference lies within the charge carrier and the electrochemistry at both the anode
and the cathode. In a PEMFC, hydrogen is oxidized to protons at the anode, which are then
transported across the PEM to the cathode where they combine with reduced oxygen to form
H2O (See Figure 1.2.1a). In an AAEMFC, hydroxide ions are produced at the cathode instead of
producing protons at the anode. In this case molecular oxygen is reduced to hydroxide ions,
which are then transported across the AAEM to the anode where they combine with hydrogen to
form water (See Figure 1.2.1b). The half cell reactions for anode and cathode for the PEMFC
and AAEMFC can be seen below[11]:
PEMFC:
Anode: H2 → 2H+ + 2e
- E° = 0.000 V vs. SHE (1.2.1)
Cathode: 1/2O2 + 2H+ + 2e
- → H2O E° = 1.229 V vs. SHE (1.2.2)
AAEMFC:
Anode: H2 + 2OH- → 2H2O + 2e
- E° = -0.828 V vs. SHE (1.2.3)
Cathode: 1/2O2 + H2O + 2e- → 2OH
- E° = 0.401 V vs. SHE (1.2.4)
4
Figure 1.2.1. (a) Schematic of a PEM-based H2/O2 fuel cell. (b) Schematic of an AAEM-based
H2/O2 fuel cell.
A schematic of an AEM-DMFC can be seen in Figure 1.2.2. In this case, the methanol
oxidation reaction (MOR) is the primary reaction that occurs at the anode (Eq. 1.2.8)[12], while
oxygen reduction (Eq. 1.2.4) remains the primary reaction at the cathode. It should be noted that
partial oxidation of methanol can occur, which produces unwanted intermediates such as formate
and carbon monoxide as shown in Eq. 1.2.5 and Eq. 1.2.6, respectively. Alkaline Exchange
Membrane Direct Methanol Fuel Cells (AEM-DMFCs) will be discussed in more detail in
Chapter 2.
CH3OH + OH- → CH3Oads + H2O + e
- (1.2.5)
CH3Oads + 3OH- → COads + 3H2O + 3e
- (1.2.6)
COads + 2OH- → CO2 + H2O + 2e
- (1.2.7)
CH3OH + 6OH- → CO2 + 5H2O + 6e
- E° = -0.810 V vs. SHE (1.2.8)
5
Figure 1.2.2. Schematic of an AAEM-based MeOH/O2 fuel cell.
1.2.2 Electrolyzers
Water electrolysis essentially reverses the direction of the electrochemical process
employed in H2/O2 fuel cells. Electrolyzers use electrical energy to drive chemical reactions, i.e.
convert electrical energy to chemical energy. Water electrolysis, the splitting of water into its
constituent elements, H2 and O2, has been known for over 200 years and has been used as a
method to produce hydrogen industrially since 1888. Different electrolytic processes are also
widely used, such as HCl electrolysis to produce hydrogen and chlorine from concentrated HCl
solutions, or the chlor-alkali process used to produce NaOH and Cl2 from concentrated brine
(NaCl) solutions. In water electrolysis, molecular oxygen is produced at the anode via the
oxidation of water. This in turn generates protons, which then cross the membrane to the cathode
where they recombine to form hydrogen gas (See Figure 1.2.3). The half cell reactions are
presented below:
6
Anode: H2O → 1/2O2 + 2H+ + 2e
- E° = 1.229 V vs. SHE (1.2.9)
Cathode: 2H+ + 2e
- → H2 E° = 0.00 V vs. SHE (1.2.10)
It should be noted that water electrolysis is sensitive to the presence of contaminants in
solution, particularly that of chloride. Given the relatively close standard electrode potentials of
both oxygen evolution (Eq. 1.2.9) and chlorine evolution (Eq. 1.2.11) and the fact that OER
requires a significant overpotential, one must anticipate the simultaneous production of both
oxygen and chlorine at reasonable operating potentials when chloride is present. Since chloride
contamination is one of the more serious issues in conventional water electrolysis, the
development of selective catalysts for OER/CER will be discussed in Chapter 4.
Anode: 2Cl- → Cl2 + 2e
- E° = 1.358 V vs. SHE (1.2.11)
Figure 1.2.3. Schematic of an acid-based water electrolysis cell.
7
1.2.3 Electrochemical Hydrogen Pump
Hydrogen purification is crucial to the advancement and integration of hydrogen-based
fuel cells. When hydrogen is produced via gassification of coal or natural gas, it typically
requires several stages of purification before it can be consumed commercially. Common
contaminants from steam reforming processes include CO2, CO, and methane. CO is particularly
problematic when considering the Pt-based catalysts typically employed in fuel cell applications
since it binds strongly to the Pt surface, effectively blocking the surface sites available for
hydrogen oxidation. CO2 is also a problematic contaminant in fuel cell applications as it can be
converted to CO via the reverse water gas shift reaction (See Eqs. 1.2.12 and 1.2.13). Current
state of the art technologies for hydrogen purification include pressure swing adsorption (PSA),
palladium membrane filters, cryogenic separation, and chemical adsorption methods based on
alkaline or ammine scrubbing[13-15]. The focus here will be on the electrochemical purification
of hydrogen, which was first pioneered by Sedlak in 1981[16].
CO2 + H2 ⇌ CO +H2O (1.2.12)
CO2 + 2M-Hads ⇌ 2M-COads + H2O + M (1.2.13)
An electrochemical hydrogen pump (ECHP) is very similar to traditional proton
exchange membrane fuel cells (PEMFCs) in terms of electrochemistry and in terms of hardware.
The electrochemical reaction at the anode is hydrogen oxidation (Eq. 1.2.14), while the reaction
at the cathode is hydrogen evolution (Eq. 1.2.15) instead of ORR. An electrochemical hydrogen
pump operates on the principle that the hydrogen oxidation and reduction reactions are extremely
facile on Pt, requiring a very small overpotential to drive the overall reaction. This may offer a
cost effective means of purifying hydrogen from various sources given that PSA and cryogenic
8
separation methods are typically energy intensive and require several steps. A schematic of a
hydrogen pump can be seen in Figure 1.2.4 in which a contaminated stream of hydrogen is fed to
the anode while purified hydrogen is produced at the cathode. The focus in subsequent chapters
will be on the development of CO and CO2 tolerant Pt-based catalysts to be used at the anode in
an ECHP.
Anode H2 → 2H+ + 2e
- E° = 0.000 V vs. SHE (1.2.14)
Cathode 2H+ + 2e
- → H2 E° = 0.000 V vs. SHE (1.2.15)
Figure 1.2.4. Schematic of a PEM-based electrochemical hydrogen pump with a contaminated
H2 feed at the anode and producing purified H2 at the cathode.
1.3 Electrocatalysis
1.3.1 Electrochemistry Fundamentals
Electrochemistry is a branch of chemistry that is largely concerned with the
thermodynamics and kinetics of chemical redox reactions (i.e. the Gibb's free energy associated
9
with the oxidation or reduction of a chemical species) in relation to the electrical work associated
with the addition or removal of electrons. The Gibb's Free Energy of a system can be described
by the following equation[2]:
(1.3.1)
where ∆G° is the standard Gibb's Free Energy, R is the ideal gas constant, T is the absolute
temperature, and aO and aR are the chemical activity of oxidant and reductant species in a
reaction such as that shown below:
O + n → R (1.3.2)
The change in Gibb's Free Energy of a system is also equal to the electrical work performed by
the system, or the negative of the work done on the system:
(1.3.3)
where n is the number of electrons transferred, F is Faradays constant (96,485 C/mol), and Erxn is
the standard electrochemical potential associated with a given reaction. By substituting Eq. 1.3.3
into Eq. 1.3.1, it is possible to derive the Nernst equation (See Eq. 1.3.4), which relates the half-
cell reduction potential to the redox composition of the system at a given temperature and
pressure.
(1.3.4)
where E is the half-cell reduction potential and E° is the standard electrode potential.
The kinetics of a standard one electron transfer reaction, such as that shown in Eq. 1.3.2,
can be described using the Butler-Volmer equation:
(1.3.5)
10
where i0 is the exchange current density, α is the transfer coefficient, F is Faraday's constant, η is
the electrochemical overpotential, R is the ideal gas constant, and T is the absolute temperature.
The Butler-Volmer is a fundamental equation in electrochemistry that describes the relationship
between the current and overpotential in the absence of mass-transport limitations. This equation
can be simplified in high overpotential regions to yield the Tafel equation:
(1.3.6)
A plot of η vs. ln(i) then allows for valuable information regarding the charge transfer coefficient
and exchange current density to be extracted.
1.3.2 Electrical Double Layer
The structure of the electrical double layer and how it is altered at different electrolyte/Pt
interfaces will be considered in Chapter 2. The development of a suitable model to accurately
describe the interface between the electrode surface and the electrolyte has taken place over the
past 150 years. The structure of the double layer is generally believed to consist of three distinct
regions: the Inner Helmholtz Plane (IHP), the Outer Helmholtz Plane (OHP), and the diffuse
layer (See Figure 1.3.1). Species located in the IHP are said to be specifically adsorbed and can
only approach the electrode surface to a finite distance, x1. This distance is defined by the locus
of the centers of the specifically adsorbed ions. Specifically adsorbed ions in the IHP experience
strong short-range interactions with the charged metal surface. The Outer Helmholtz plane
(OHP) is defined at a distance, x2, which is determined by how close ions or molecules can
approach the electrode before desolvating and adsorbing to the surface[11]. This is considered
the plane of closest approach and is limited by the thickness of the compact layer. Ions found in
the OHP are said to be nonspecifically adsorbed since their interaction with the charged metal
11
surface occurs only through long-range electrostatic forces. Beyond the OHP, nonspecifically
adsorbed ions are distributed throughout a three dimensional region known as the diffuse layer,
which extends from the OHP to the bulk solution. Figure 1.3.1 shows the three regions of the
double layer for a negatively charged electrode surface.
Figure 1.3.1. Representation of the double-layer interface based upon the Gouy-Chapman-Stern
model showing the electrode surface (M), the Inner Helmholtz Plane (IHP), Outer Helmholtz
Plane (OHP), and Diffuse Layer.
The point at which an electrode surface is free of adsorbates and free of excess charge
(i.e. completely neutral) is defined as the potential of zero free charge (PZFC). The potential at
which any excess charge on the electrode surface is balanced by ions in solutions (specifically
12
adsorbed or not) is defined as the potential of zero total charge (PZTC). It is important to note
that most electrode surfaces are not well defined (i.e. polycrystalline electrodes having a certain
roughness). In this case it is generally accepted that the potential of zero charge (PZC) is the
potential at which the differential capacitance experiences a minimum in dilute electrolyte
solutions[11].
In the absence of specific adsorption, the voltage drop between the metal electrode and
the solution occurs across the compact and diffuse layers, as described by the Gouy-Chapman-
Stern model, with the OHP defined by the locus of centers of hydrated ions[11]. It is important to
note that the thickness of the double layer is dependent upon the concentration of the electrolyte
solution. At high electrolyte concentrations, the double layer becomes more compact because
there is a higher concentration of ions within close proximity of the surface to form the compact
layer. The effect is that the potential drop across the double layer occurs primarily in the compact
layer with little contribution from the diffuse layer. At more dilute electrolyte concentrations, or
ones where the ions have limited mobility, a greater thickness is needed in order to accumulate
enough charge to counterbalance the charge of the metal surface. The effect is that the potential
drop across the diffuse layer becomes important. In this case the electrode reaction is no longer
driven by the potential difference ϕM - ϕS, where ϕM is the potential of the metal electrode and
ϕS is the potential of the solution. Instead, the electroactive species at the OHP experiences a
potential difference (ϕM - ϕ2), where ϕ2 is the potential at the OHP. In addition, charged
electroactive species migrate under the influence of the electric field within the diffuse layer. The
concentration of electroactive species located at the OHP will then differ from that immediately
outside the in the diffuse layer by a factor of exp(-zϕ2F/RT), where z is the charge on the ion, F is
Faraday’s constant, R is the ideal gas constant, and T is the absolute temperature.
13
In general, cations will typically be drawn to the electrode surface when the surface has a
negative charge (E < EPZC) whereas anions will be attracted to a positively charged electrode
surface (E > EPZC). Similarly, cations will be repelled from a positively charged surface (E >
PZC) and anions will be repelled from a negatively charged surface (E < PZC). Consequently, in
the case of E < EPZC the positive charge contributed by cations in the IHP will create a potential
shift in ϕ2 such that species in the OHP plane will experience a potential more positive than ϕ1
(See Figure 1.3.1). Similarly, at positive potentials anions will populate the IHP creating a
negative shift in ϕ2. These effects can have a large influence on the kinetics of a given reaction
since ϕ2 varies with the electrode potential and electrolyte concentration.
The described effect of the double layer on kinetics is often referred to as either the
Frumkin effect or the ϕ2 effect. In general, the overall result is that i0 and k
0 are dependent upon
the variation of ϕ2 with E - EPZC and concentration[17-19]. The Frumkin correction can be used
to relate the apparent rate constant with the true rate constant:
(1.3.6)
where k0 is the apparent rate constant,
is the true rate constant, α is the transfer coefficient, ϕ2
is the potential at the OHP, F is Faraday's constant, R is the ideal gas constant, T in the absolute
temperature, and z is the charge on the ion. The potential driving the electrode reaction at the
OHP is corrected for by the first exponential term. This is also valid for outer sphere electron
transfer reactions. The concentration profile for any charged species is corrected for by the
second exponential term. As follows from Eq. 1.3.6, the true exchange current density can also
be determined:
(1.3.7)
where i0 is the apparent exchange current density and
is the true exchange current density.
14
While ϕ2 can be predicted using the Gouy-Chapman-Stern model in the absence of
specific adsorption, it becomes more complex in the presence of specific adsorption of charged
species e.g. quaternary ammoniums. The adsorption of electro-inactive species create significant
changes in the structure of the double layer that differ from the non-adsorbing case[11, 17, 19].
In this case the actual potential at the OHP cannot be clearly defined. Instead a qualitative
assessment must be made. Furthermore, the adsorption of electro-inactive species results in
blockage of the electrode surface. This inhibits the reaction rate independent of ϕ2. In general,
the blockage effect only becomes dominant at high concentrations of the electro-inactive species.
At low concentrations of the adsorbing species the reaction rate is affected by the ϕ2 effect.
1.3.3 Cyclic Voltammetry (CV)
Cyclic voltammetry is perhaps one of the most widely used techniques used for
electrochemical characterization. Cyclic voltammetry is commonly used in the characterization
of electrocatalytic materials and the study of oxidation/reduction processes that the occur at the
electrode-electrolyte interface. Consider a polished polycrystalline Pt electrode in Ar saturated
0.1 M KOH solution, such as that shown in Figure 1.3.2. Saturating the electrolyte with argon, or
any inert gas, allows for the Pt surface to be studied in detail without the interference of
dissolved oxygen. The cathodic scan in the cyclic voltammogram reveals information regarding
the reduction processes occurring at the Pt surface while the anodic scan reflects the oxidation
processes. Generally speaking, a typical CV of a polycrystalline surface can be divided into three
regions. The features observed in Region I are characteristic of the adsorption and desorption of
hydrogen. Region II is referred to as the "double-layer" region. This is due to the fact that there
are no faradaic reactions occurring in this potential regime; only the double-layer capacitance is
15
observed. Finally, Region III is defined by the formation and reduction of surface oxides. As one
scans anodically, features corresponding to the formation of PtOH and PtO2 are observed and
eventually the evolution of O2 can be seen at potentials positive of 1.6 V vs. RHE. On the
cathodic scan, features corresponding to the reduction of surface oxides can be seen in Region
III. Again Region II corresponds to the absence of faradaic processes on Pt. Continuing to
cathodically scan into Region I, the formation of Pt-H is observed before the eventual formation
and desorption of H2 at potentials negative of 0.05 V vs. RHE.
Figure 1.3.2. Cyclic voltammogram of a polished polycrystalline Pt disk electrode in Ar
saturated 0.1 M KOH. Scan rate of 20 mV s-1
. The anodic scan is shown in blue; the cathodic
scan is shown in red.
Cyclic voltammetry is also used in the study of electroactive species dissolved in solution
at the electrode surface. Consider again a polycrystalline Pt electrode submerged in Ar saturated
0.1 M KOH, however, this time with the addition of 10 mM K3Fe(CN)6. The outer-sphere
oxidation/reduction behavior of this species can be examined by scanning the appropriate
16
potential regime as shown in Figure 1.3.3. The peak corresponding to the oxidation of
[Fe(CN)6]4-
to [Fe(CN)6]3-
can be seen at 1.2210 V. Similarly, the peak corresponding to the
reduction of [Fe(CN)6]3-
to [Fe(CN)6]4-
can be seen at 1.1398 V. The separation of peaks is
approximately 81 mV, which corresponds a one-electron transfer reaction. This redox reaction is
said to exhibit a Nernstian behavior. It should be noted that a redox couple displaying ideal
Nernstian behavior would have a peak separation of 56.5 mV/n, where n is the number of
electrons transferred.
Figure 1.3.3. Cyclic voltammogram of a polished polycrystalline Pt disk electrode in Ar
saturated 0.1 M KOH + 10 mM K3Fe(CN)6. Scan rate of 20 mV s-1
.
1.3.4 Rotating Ring Disk Electrode (RRDE) Method
The Rotating Ring Disk Electrode (RRDE) is a common method used in electrochemistry
to study the products and intermediate species generated during an electrochemical reaction
under convective diffusion as the potential of the working electrode is varied. This technique has
been used extensively to study the oxygen reduction reaction (ORR) due to its ability to detect
17
any hydrogen peroxide intermediate species produced[20-23]. A schematic of the rotating ring
disk apparatus can be seen in Figure 1.3.2. Essentially, the electrode tip consists of a disk
electrode (carbon, platinum, gold, or other material) embedded in the center of a Teflon housing.
Surrounding the disk electrode is a secondary ring electrode, the potential of which is controlled
separately. The ring electrode material is typically either gold or platinum. The tip is fitted into
the shaft of the rotator, which controls the speed at which the electrode can be rotated and thus
controls the transport of electroactive species to the electrode surface. The electrodes are
connected to a potentiostat/galvanostat that is used to control the potential of each electrode
while monitoring the current or vice versa. The tip of the electrode is submerged in an electrolyte
solution containing the supporting electrolyte and electroactive species. Ink suspensions
containing the catalyst of interest can be made and deposited on the disk of the electrode to
provide a thin film of the catalyst under investigation.
Figure 1.3.4. Depiction of RRDE tip.
18
The advantage of the RRDE setup is that it allows an electroactive species to be
oxidized/reduced at the disk electrode and then the intermediate species generated are
reduced/oxidized at the ring electrode. The transport of species to the disk electrode is controlled
by varying the rotation rate, which also controls the rate at which species are swept away radially
from the disk to the secondary ring electrode. By controlling the rotation rate of the electrode and
monitoring the disk current, one can apply the Koutecký-Levich equation to obtain information
about the number of electrons transferred in a reaction or the diffusion coefficient of the
electroactive species in solution[11]:
(1.3.8)
where ik represents the kinetic current, n is the number of electrons transferred, F is Faraday's
constant, A is the electrochemically active surface area, DO is the diffusion coefficient of the
reactant species, ω is the angular rotation rate of the electrode, ν is kinematic viscosity of the
electrolyte, and CO is the concentration of the reactive species.
1.4 Differential Electrochemical Mass Spectrometry (DEMS)
Differential Electrochemical Mass Spectrometry (DEMS) is a unique technique that
allows for the on-line detection and identification of volatile products and intermediate species of
faradaic reactions. When used in parallel with cyclic voltammetry, DEMS is able to provide on-
line detection of species generated at the electrode surface as a function of potential, resulting in
a mass spectrometric cyclic voltammogram (MSCV)[24]. Other electrochemical techniques,
such as chronoamperometry or chronopotentiometry, can also be combined with DEMS.
Reaction products can be detected somewhat quantitatively, allowing for the distinction between
the main and side reactions products.
19
The DEMS setup is fairly simple, consisting of a 3-electrode electrochemical cell (See
Figure 1.4.1a) separated via a porous PTFE membrane from the apparatus (See Figure 1.4.1b),
which consists of turbomolecular dragging pump station, a rotary vane vacuum pump, and a
quadrupole mass spectrometer. A flow chart of the typical DEMS setup can be seen in Figure
1.4.1c.
Species generated at the electrode surface diffuse through a porous PTFE membrane via
needle valve to a vacuum environment where they are then transported to the mass spectrometer.
The chemical species are ionized in the ion source before being separated by mass and
transported to the detector. Simultaneously, the potential of the working electrode can be held
constant or varied to provide the necessary electrochemical perturbation. DEMS has been used
among others to study the electrochemical decomposition of alcohols[25, 26], quantifying
simultaneous oxygen and chlorine evolution[27-30], and the exchange of oxygen between the
electrode catalyst material and solvent[31].
20
Figure 1.4.1. (a) Schematic of an electrochemical cell designed for use with a DEMS apparatus,
(b) Photo of DEMS apparatus used, and (c) Flow chart of DEMS.
1.5 X-ray absorption Spectroscopy (XAS)
X-ray absorption spectroscopy is a technique that relies on the use of high intensity X-
rays, typically generated by a synchrotron light source, to excite the core level electrons (K, L,
and/or M shells) of a particular element in a sample from the ground state into an excited
electronic state or into the continuum[32]. Promoting an electron to an excited state leaves an
empty core shell hole behind, which is quickly filled (~10-15
s) by an electron from a higher
energy orbital[33]. The relaxation of an electron from a higher energy orbital to a lower one
results in the release of energy in the form of fluorescence. The amount of energy released is
equal to the energy difference between the two orbitals and is element specific, therefore the
21
fluorescence radiation emitted as a result of the transition is characteristic of the absorbing
element.
Not all of the incident X-ray radiation is absorbed by the sample and the remainder is
transmitted through the sample. Absorption is a process of finite probability and the amount of
X-ray radiation transmitted through the sample is exponentially dependent upon the thickness of
the sample and the X-ray absorption coefficient:
(1.5.1)
where It is the intensity of the transmitted radiation, I0 is the intensity of the incident radiation, x
is the thickness of the sample, and μ is the X-ray absorption coefficient. The X-ray absorption
coefficient describes the probability that an X-ray will be absorbed by a given element or sample.
This is an element-specific quantity that increases with the binding energy of core-level electrons
and thus increases with atomic number.
The typical setup for an XAS experiment is shown in the schematic below:
Figure 1.5.1. Schematic of the typical XAS experimental setup.
The incident X-ray radiation passes through an ionization chamber that measures the radiation
flux, or initial intensity, I0. The radiation then hits the sample, producing photoelectrons (X-ray
fluorescence) that can be detected by a fluorescence detector set up at a 45° angle from the
sample. The remaining radiation that is transmitted through the sample then enters a second
22
ionization chamber to measure the transmitted radiation, It. Typically a reference foil is placed
after the secondary ionization chamber and a third ionization chamber is used to measure the
intensity of the X-rays transmitted through the reference material, Ir. The reference signal is
important in ensuring that the position of the monochromator does not shift during the
measurement. By rearranging Eq. 1.5.1, the X-ray absorption coefficient can be determined
based upon the measured intensity of the incident radiation and the intensity of the of radiation
transmitted:
(1.5.2)
Similarly, μ can also be determined from the intensity of the X-ray fluorescence since X-ray
absorption is proportional to the generation of a core hole and the subsequent relaxation process:
(1.5.3)
As can be inferred from the above equation, a sharp increase in μ is observed when the incident
radiation is equal to the energy required to excite promote a core level electron. This can be
observed in a typical XAS spectrum (See Figure 1.5.2b) and the feature is referred to as the
absorption edge. The absorption edge energy is specific to the binding energy of the electrons for
each element.
23
Figure 1.5.2. (a) Example of a normalized XAS spectrum for RuO2 showing both the XANES
and EXAFS regions; (b) EXAFS spectrum of RuO2 shown in k-space; (c) Fourier transform of k-
space EXAFS spectrum to R-space; (d) XANES spectrum of RuO2 showing the pre-edge and
absorption edge regions; (e) Local structure of RuO2 ((110) plane shown in blue) used to model
EXAFS data.
XAS spectra are generally divided into two regions: the X-ray Absorption Near-Edge
Structure (XANES) and the Extended X-Ray Absorption Fine Structure (EXAFS) (See Figure
1.5.2). The XANES portion of the spectrum can be used to obtain information regarding the
oxidation state and coordinate chemistry of the absorbing atoms. EXAFS is used to obtain
information about the local molecular structure of the element of interest within a sample. The
XANES region ranges from approximately -50 to +100 eV of the absorption edge.
The EXAFS region begins at approximately 50 eV+ of the absorption edge. The EXAFS
region is sensitive to the surrounding environment because at excitation energies higher than the
24
absorption edge the photoelectron is ejected from the absorbing atom into the continuum. The
ejected photoelectrons are of dual nature and can be described as spherical waves propagating
from the absorbing atom that scatter off of the surrounding atoms. The phase of the scattered
photoelectron wave is determined by its wavelength and the distance between the propagating
atom and the scattering atoms. The minima and maxima observed in the oscillating EXAFS
signal are dependent upon the phase of the outgoing photoelectron wave and the scattered wave
at the absorbing atom. A minimum occurs when the waves are out of phase; a maximum occurs
when the waves are in phase. There are generally many scattering paths a photoelectron can take
depending upon the nature of neighboring atoms and the distances between them, although some
pathways are degenerate (See Figure 1.5.3). Neighboring atoms at the same interatomic distance
from the absorbing atom all contribute to the same component of the EXAFs signal. Any
neighboring atoms at the same distance from the absorbing atom are all part of the same shell
and the total number of atoms in a shell determines the coordination number.
Figure 1.5.3. Example of a single photoelectron backscattering event and a multiple
backscattering event that contribute to the oscillating EXAFS signal.
25
The EXAFS function, χ(E), describes the oscillatory part of the of the absorption
spectrum (see Figure 1.5.2) and is defined as:
(1.5.4)
where µ(E) is the observed absorption coefficient and µ0(E) is the absorption coefficient at an
energy equivalent to the absorption edge, E0. EXAFS is more commonly interpreted in terms of
the emitted photoelectron energy rather than the energy of the incident X-ray radiation. The
transformation to k-space (See Figure 1.5.2b), where k is the wave number of the photoelectrons,
takes place through the following relationship:
(1.5.5)
where me is the mass of an electron and is Planck's constant. The EXAFS can then be
expressed as the sum of the contribution from all scattering paths of the photoelectron:
(1.5.6)
where Nj is the coordination number of the neighboring atom, Rj is the distance to the
neighboring atom, σj2 is the mean-square disorder of the neighboring distance, k is the photo-
electron wavenumber, λ(k) is the wavelength of the radiation, fj(k) is the scattering amplitude,
and δj(k) is the phase shift.
A Fourier transform of the data expressed in k-space (momentum space) leads to an
expression of the data in R-space (direct space), where R represents the radial distance of the
neighboring atom from the absorbing atom. An example of the EXAFS spectrum in R-space can
be seen in Figure 1.5.2c. Theoretical EXAFS spectra can be generated from a chosen structural
model (see Figure 1.5.4e) using software packages such as IFFEFIT[34]. The theoretical EXAFS
26
spectra can then be fitted to the measured spectra as the parameters of the model are refined. The
model shown in Figure 1.5.4e is a representation of the RuO2 local structure that was used to fit
the EXAFS spectrum shown in Figure 1.5.2a.
1.6 Density Functional Theory (DFT)
Density Functional Theory (DFT) is a quantum mechanical computational approach used
in many branches of chemistry. In electrocatalysis, DFT serves a great purpose in terms of
modeling electrode surfaces and the electrocatalytic reactions that take place at the electrode-
electrolyte interface. In this respect, DFT is a valuable tool for understanding, predicting, and
designing catalyst materials. It has already been shown that DFT can be used to successfully
model electrocatalytic reactions such as ORR, OER, and CER on a variety of metal and metal
oxide surfaces[35-40].
The activity of different catalysts towards various chemical reactions can be predicted
computationally by considering the relationship between the adsorption energies of adsorbates in
addition to the Brønsted-Evans-Polanyi relation between reaction energies and the reaction
barriers. Classically, the Brønsted-Evans-Polanyi (BEP) relation describes the linear trend
between the activation energy and the reaction energy for reactions belonging to the same
class[41, 42]. The relationship can be expressed as:
(1.6.1)
where Ea is the activation energy, E0 is the activation energy of a reference reaction of the same
class, ΔH is the enthalpy of reaction, and α characterizes the position of the transition state along
the reaction coordinate (0 ≤ α ≤ 1). This linear relationship serves as a convenient way to
calculate the activation energy of reactions within a distinct family. The activation energy can
27
then be used to extract information regarding the kinetics of a particular reaction by applying the
Arrhenius equation:
(1.6.2)
where k is rate constant, A is pre-exponential factor, R is the Ideal Gas Constant, and T is the
absolute temperature.
In the context of heterogeneous catalysis, it was recognized that BEP relations exist
between the activation energy for the dissociative chemisorption of a number of molecules and
the reaction energy[43-45]. Essentially, this links the height of the activation barrier for a given
reaction to the driving force of the reaction. DFT calculations have been able to take advantage
of BEP relations such that a single parameter can be used to describe the catalytic activity of
many different materials towards a particular reaction.
In the case of the ORR, the binding energies of O*, OH*, and OOH* intermediate species
to the metal/metal oxide surface were found to serve as reliable descriptors for the
electrochemical activity. However, it is important to note that the binding energies of these
intermediates (*OH and *OOH) cannot be varied independently due to the scaling relationships
between the binding of related intermediates[40]. The thermodynamic model employed for the
ORR allows free energy of reaction intermediates to be shifted with respect to several
parameters, including pH, potential, concentration, and electric field strength. Essentially, the
Gibb's Free Energy of the reaction intermediates can be calculated by[35]:
(1.6.3)
where ∆E is the reaction energy calculated via DFT, ∆ZPE is the difference in zero point energy,
T is the absolute temperature, and ∆S is the change in entropy. Both ∆ZPE and ∆S are calculated
by DFT using the vibrational frequencies and standard tables for gas molecules. The typical
28
method employed for such calculations sets the reference potential to be that of the standard
hydrogen electrode (SHE). This allows for the chemical potential of H+ + e
- to be related to that
of 1/2H2 in the gas phase. Considering the pH of the electrolyte to be pH 0, the reaction free
energy of 1/2H2 → H+ + e
- is also zero for 1 bar of H2 at 298 K and an electrode potential of E =
0. Therefore, the free energy of the reaction *AH → A + 1/2H2 can be calculated since ∆G0 =
∆G at standard conditions. At a pH other than pH 0, the free energy of H+ ions can be corrected
by considering the concentration dependence of entropy:
(1.6.4)
From Eq. 1.6.4, the reaction free energy can then be calculated by:
(1.6.5)
where ∆GE = -eE is used to account for the effect of a bias on all states involving an electron at
the electrode by shifting the energy of this state.
Modeling the ORR at the electrode surface follows the associative mechanism of oxygen
reduction (Eqs. 1.6.6a-1.6.6d), which divides the 4-electron transfer reaction into a series of four
consecutive, concerted single electron and proton transfer steps[46, 47]. This follows from the
Marcus theory of electron transfer given the fact that it is generally more energetically favorable
to proceed through a series of single electrons transfers in which intermediates are formed than it
is to simultaneously transfer two electrons[40].
O2(g) + 4H+ + 4e
- + * → OOH* + 3H
+ + 3e
- (1.6.6a)
OOH* + 3H+ + 3e
- → O* + H2O(l) + 2H
+ + 2e
- (1.6.6b)
O* + H2O(l) + 2H+ + 2e
- → OH* + H2O(l) + H
+ + e
- (1.6.6c)
OH* + H2O(l) + H+ + e
- → * + 2H2O(l) (1.6.6d)
29
The calculated free energy associated with each individual electron/proton transfer step can be
used to construct a free energy diagram, which shows the free energy of each step vs. the
reaction coordinate as shown in Figure 1.6.1[48].
Figure 1.6.1. Example of free energy diagrams for the reduction of O2 on RuO2. The dotted line
represents the equilibrium potential of the reduction of O2 to H2O2.[48]
In general, the diagrams can be interpreted by considering the sequence of free energies
associated with each electron/proton transfer at the given driving force (E). According to the free
energy diagrams shown in Figure 1.6.1[48], at a potential of E = 0.8 V the reduction of O2 is
unfavorable. This is evidenced by the fact that not all steps are downhill in free energy
(following the reaction coordinate from right to left), specifically the final reduction of *OH to a
free surface site * (Step 1.6.6d). This indicates that the oxygen species are bound too strongly to
the RuO2 surface. It can also be observed that the first reduction step (Step 1.6.6a) forms a
strongly bound *OOH intermediate that continues to be more stable than the formation of H2O2
at most reasonable potentials. It allows to predict that the formation of H2O2 is not favorable
until the free energy associated with the formation of *OOH is higher than that of equilibrium
potential for the reduction of O2 to H2O2, EO2/H2O2 (i.e. the step associated with *OOH must be
above the dotted line in Figure 1.6.1). It is not until lower potentials (higher overpotentials) that
all steps are downhill in free energy and the complete 4-electron reduction of O2 can proceed, for
30
instance, at E = 0.5 V. It is not until E = 0.14 V that the formation of H2O2 becomes
thermodynamically possible, as can be seen in the free energy diagram for E = 0 V.
An ideal ORR catalyst should facilitate the complete 4e- reduction to H2O just below the
equilibrium potential of 1.23 V. Similarly, an ideal catalyst for O2 reduction to H2O2 would
facilitate the reaction just below the equilibrium potential of 0.68 V. Since the binding of *OH
and *OOH are related by a constant value (scaling relationship) and therefore cannot be varied
independently[40], it follows that there cannot be a single ideal catalytic site that binds all
oxygen intermediates optimally. A consequence of this scaling relationship is that ORR is
limited by either one of two steps: the activation of O2 to *OOH (Step 1.6.6a) or the removal of
adsorbed *OH (Step 1.6.6d). In the event that a catalyst binds oxygen intermediates too strongly
Step 1.6.6d will be the limiting step, whereas binding oxygen intermediates too weakly results in
Step 1.6.6a being the limiting step. This is essentially a formulation of the Sabatier principle[49],
which concludes that an ideal interaction of the adsorbed intermediate with the catalyst (neither
too weak or too strong) is desired.
An activity plot based on the DFT thermodynamic analysis is shown in Figure 1.6.2. It is
well known for heterogeneous catalysis that a volcano-type relationship is established when the
activity of a catalyst is plotted as a function of some property of that catalyst[50-52]. This
quantitative representation of the Sabatier principle has been reported in the literature for the
HER[53] and ORR[38, 54, 55]. In this case, the catalyst activity for the 2e- and 4e
- ORR
pathways is calculated in terms of (over)potential and the ability of the catalyst to bind *OOH
was used as the activity descriptor (note that the scaling relationship between oxygen
intermediates allows for the activity to be described using a single parameter, which again in this
case is the *OOH binding energy[56]) . The right (weak binding) leg of the volcano represents
31
the activation of O2 to *OOH (Step 1.6.6a). This step is the same in both the 4e- reduction of O2
to H2O and the 2e- reduction to form H2O2. The left (strong binding) leg of the volcano
represents the removal of *OH (Step 1.6.6d) in the 4e- reduction to H2O and the removal of
*OOH for the 2e- reduction to H2O2[38]. As follows from the Sabatier principle, an ideal catalyst
displaying an ideal binding (neither too strong nor too weak) for the oxygen intermediates would
be found at the apex of the volcano. Considering the ORR on RuO2 in Figure 1.6.2, it is shown
that RuO2 suffers from binding *OOH too strongly. This indicated that Step 1.6.6a is the limiting
step for ORR on RuO2. The ORR on doped and non-doped RuO2 will be considered in greater
detail in Chapter 3.
Figure 1.6.2. Volcano plot for the 2-electron (blue) and 4-electron (red) reduction of O2 to H2O2
and H2O, respectively, using the binding energy of OOH as a descriptor. The dotted lines
represent the equilibrium potentials for the reduction products.
1.7 Scope of Dissertation
This dissertation focuses on various aspects of electrocatalytic devices for hydrogen
energy storage and purification. In Chapter 2, alkaline direct methanol fuel cells will be
discussed in regards to the potential losses encountered in an AEM environment. This chapter
32
identifies the structure of the double layer, electrostatic effects, and the adsorption of quaternary
ammonium species as causes of the observed potential losses which need to be removed. It also
aims at developing a supported Pt-based electrocatalyst for these applications. In Chapters 3 and
4 the effect of the electrocatalyst local structure on the selectivity in electrocatalytic reactions
relevant to electrolysis and fuel cells (i.e. oxygen evolution, chlorine evolution, and oxygen
reduction) will be discussed. Particularly, correlations between the variations of the RuO2 local
structure induced by transition metal doping and the electrochemical selectivity are examined on
the basis of XAS, DFT, and electrochemical data. Finally, Chapter 5 addresses the issue of
hydrogen purification in an electrochemical hydrogen pump. Pt-based catalysts tailored to
withstand CO/CO2 poisoning associated with the use of reformate gas mixtures are treated in
detail. Chapter 6 expands upon the research presented and offers insight to future directions of
hydrogen based technologies.
33
1.8 References
1. Thomas, S. and M. Zalbowitz Fuel Cells - Green Power.
2. O'Hayre, R., et al., Fuel Cell Fundementals. 2006, Hoboken, NJ: John Wiley & Sons.
3. Scott, K., et al., Performance of a direct methanol alkaline membrane fuel cell. J. Power
Sources, 2008. 175: p. 452-457.
4. Tripkovic, A.V., et al., Methanol electrooxidation on supported Pt and PtRu catalysts in
acid and alkaline solutions. Electrochim. Acta, 2002. 47: p. 3707-3714.
5. Antolini, E. and E.R. Gonzalez, Alkaline direct alcohol fuel cells. J. Power Sources,
2010. 195: p. 3431-3450.
6. Spendelow, J.S. and A. Wieckowski, Electrocatalysis of oxygen reduction and small
alcohol oxidation in alkaline media. Phys. Chem. Chem. Phys., 2007. 9: p. 2654-2675.
7. Varcoe, J.R. and R.C.T. Slade, Prospects for alkaline anion-exchange membranes in low
temperature fuel cells. Fuel Cells (Weinheim, Ger.), 2005. 5(2): p. 187-200.
8. Hickner, M.A., A.M. Herring, and E.B. Coughlin, Anion Exchange Membranes: Current
Status and Moving Forward. J. Polym. Sci., Part B: Polym. Phys., 2013. 51: p. 1727-
1735.
9. Kinoshita, K., Electrochemical Oxygen Technology. The ECS Series of Texts and
Monographs. 1992, New York, NY: John Wiley & Sons, Inc.
10. McLean, G.F., et al., An assessment of alkaline fuel cell technology. Int. J. Hydrogen
Energy, 2002. 27: p. 507–526.
11. Bard, A.J. and L.R. Faulkner, Electrochemical Methods: Fundamentals and Applications.
2nd Edition ed. 2001, New York: John Wiley & Sons, Inc.
12. Yu, E.H. and K. Scott, Development of direct methanol alkaline fuel cells using anion
exchange membranes. J. Power Sources, 2004. 137: p. 248-256.
34
13. Liu, K., C. Song, and V. Subramani, Hydrogen and Syngas Production and Purification
Technologies. 2009, Hoboken, NJ: A Joint Publication of the Center for Chemical
Process Safety of the American Institute of Chemical Engineers and John Wiley & Sons,
Inc. 564.
14. Grashoff, G.J., C.E. Pilkington, and C.W. Corti, The Purification of Hydrogen. Platinum
Metals Rev., 1983. 27: p. 157-169.
15. Uehara, I., Separation and Purification of Hydrogen, in Energy Carriers and Conversion
Systems with Emphasis on Hydrogen, T. Ohtaaytexasphltw, Editor. 2008, Eolss
Publishers: Paris, France.
16. Sedlak, J.M., J.F. Austin, and A.B. LaConti, Hydrogen Recovery and Purification Using
the Solid Polymer Electrolyte Electrolysis Cell. Int. J. Hydrogen Energy, 1981. 6: p. 45-
51.
17. Devanathan, M.A.V. and M.J. Fernando, Specific adsorption of tetra-alkyl-ammonium
iodides at the mercury-water interface and the structure of the electrical double layer.
Trans. Faraday Soc., 1962. 58: p. 368-381.
18. Frumkin, A.N., Influence of cation adsorption on the kinetics of electrode processes.
Trans. Faraday Soc., 1959. 55: p. 156-167.
19. Breiter, M., M. Kleinerman, and P. Delahay, Structure of the Double Layer and Electrode
Processes. J. Am. Chem. Soc., 1958. 80: p. 5111-5117.
20. Appleby, A.J. and M. Savy, Kinetics of oxygen reduction reactions involving catalytic
decomposition of hydrogen peroxide. Applications to porous and rotating ring-disk
electrodes. J. Electroanal. Chem., 1978. 92: p. 15-30.
21. Hsueh, K.L., D.T. Chin, and S. Srinivasan, Electrode kinetics of oxygen reduction. A
theoretical and experimental analysis of the rotating ring-disk electrode method. J.
Electroanal. Chem., 1983. 153: p. 79-95.
22. Markovic, N.M., H.A. Gasteiger, and P.N. Ross, Oxygen Reduction on Platinum Low-
Index Single-Crystal Surfaces in Alkaline Solution - Rotating Ring Disk Studies. J. Phys.
Chem., 1996. 100: p. 6715-6721.
35
23. Paulus, U.A., et al., Oxygen reduction on a high-surface are Pt/Vulcan carbon catalyst: a
thin-film rotating ring-disk electrode study. J. Electroanal. Chem., 2001. 495: p. 134-145.
24. Baltruschat, H., Differential electrochemical mass spectrometry. J. Am. Soc. Mass.
Spectrom., 2004. 15: p. 1693-1706.
25. Wang, H., Z. Jusys, and R.J. Behm, Ethanol oxidation on carbon-supported Pt, PtRu,
and Pt3Sn catalysts: A quantitative DEMS study. J. Power Sources, 2006. 154: p. 351-
359.
26. He, Q., et al., Highly Stable Pt−Au@Ru/C Catalyst Nanoparticles for Methanol Electro-
oxidation. J. Phys. Chem. C, 2013. 117: p. 1457-1467.
27. Arikawa, T., Y. Murakami, and Y. Takasu, Simultaneous determination of chlorine and
oxygen evolving at RuO2/Ti and RuO2-TiO2/Ti anodes by differential electrochemical
mass spectroscopy. J. Appl. Electrochem., 1998. 28: p. 511-516.
28. Macounova, K., et al., Parallel oxygen and chlorine evolution on Ru1−xNixO2−y
nanostructured electrodes. Electrochim. Acta, 2008. 53: p. 6126-6134.
29. Petrykin, V., et al., Local structure of Co doped RuO2 nanocrystalline electrocatalytic
materials for chlorine and oxygen evolution. Catal. Today, 2012. 202(63-69).
30. Petrykin, V., et al., Zn-doped RuO2 electrocatalysts for selective oxygen evolution:
relationship between local structure and electrocatalytic behavior in chloride containing
media. Chem. Mater. , 2011. 23: p. 200-207.
31. Macounova, K., M. Makarova, and P. Krtil, Oxygen evolution on nanocrystalline RuO2
and Ru0.9Ni0.1O2 electrodes – DEMS approach to reaction mechanism determination.
Electrochem. Commun., 2009. 11: p. 1865-1868.
32. Yano, J. and V.K. Yachandra, X-ray Absorption Spectroscopy. Photosynth. Res., 2009.
102: p. 241-254.
33. Kelly, S.D., D. Hesterberg, and B. Ravel, Analysis of Soils and Minerals Using X-Ray
Absorption Spectroscopy. Methods of Soil Analysis: Part 5. Mineralogical Methods.
2008, Madison, WI USA: Soil Science Society of America, Inc.
36
34. Newville, M., IFEFFIT : interactive XAFS analysis and FEFF fitting. J. Synchrotron
Radiat., 2001. 8: p. 322-324.
35. Rossmeisl, J., et al., Electrolysis of water on oxide surfaces. J. Electroanal. Chem., 2007.
607: p. 83-89.
36. Hansen, H.A., et al., Electrochemical chlorine evolution at rutile oxide (110) surfaces.
Phys. Chem. Chem. Phys., 2010. 12: p. 283-290.
37. Man, I.C., et al., Universality in Oxygen Evolution Electrocatalysis on Oxide Surfaces.
Chem. Cat. Chem., 2011. 3: p. 1159-1165.
38. Viswanathan, V., et al., Unifying the 2e– and 4e– Reduction of Oxygen on Metal
Surfaces. The Journal of Physical Chemistry Letters, 2012. 3(20): p. 2948-2951.
39. Halck, N.B., et al., Beyond the volcano limitations in electrocatalysis - oxygen evolution
reaction. Phys. Chem. Chem. Phys., 2014. 16: p. 13682.
40. Koper, M.T.M., Thermodynamic theory of multi-electron transfer reactions: Implications
for electrocatalysis. J. Electroanal. Chem., 2011(660): p. 254-260.
41. Evans, M.G. and M. Polanyi, Further considerations of the thermodynamics of chemical
equilibria and reaction rates. Trans. Faraday Soc., 1936. 32: p. 1333-1360.
42. Brønsted, J.N. and K.J. Pedersen, in Zeitschrift für Phys. Chemie, Stöchiometrie und
Verwandtschaftslehre. 1924. p. 185–235.
43. Nørskov, J.K., et al., Universality in Heterogeneous Catalysis. J. Catal., 2002. 209: p.
275-278.
44. Liu, Z.-P. and P. Hu, General trends in CO dissociation on transition metal surfaces. J.
Chem. Phys., 2001. 114: p. 8244-8247.
45. Logadottir, A., et al., The Bronsted-Evans-Polanyi Relation and the Volcano Plot for
Ammonia Synthesis over Transition Metal Catalysts. J. Catal., 2001. 197: p. 229-231.
37
46. Nørskov, J.K., et al., Origin of the Overpotential for Oxygen Reduction at a Fuel-Cell
Cathode. J. Phys. Chem. B, 2004. 108: p. 17886−17892.
47. Karlberg, G.S., J. Rossmeisl, and J.K. Nørskov, Estimations of electric field effects on the
oxygen reduction reaction based on the density functional theory. Phys. Chem. Chem.
Phys., 2007. 9: p. 5158−5161.
48. Abbott, D.F., et al., Oxygen reduction on nanocrystalline ruthenia – local structure
effects. RSC Adv., 2015. 5(1235-1243).
49. Sabatier, F., La catalyse en chimie organique. 1920, Berauge, Paris.
50. Boudart, M., Handbook of Heterogeneous Catalysis. 1997, Weinheim: Wiley-VCH.
51. Boudart, M. and G. Djéga-Mariadassou, Kinetics of Heterogeneous Catalytic Reactions.
1984, Princeton, NJ: Princeton Univ. Press.
52. Dumesic, J.A., et al., The Microkinetics of Heterogeneous Catalysis. 1993, Washington,
D.C.: Am. Chem. Soc.
53. Greeley, J., et al., Computational High-throughput Screening of Electrocatalytic
Materials for Hydrogen Evolution. Nat. Mater., 2006. 5: p. 909-913.
54. Stamenkovic, V., et al., Changing the Activity of Electrocatalysts for Oxygen Reduction
by Tuning the Surface Electronic Structure. Angew. Chem., Int. Ed., 2006. 45: p. 2897-
2901.
55. Greeley, J., et al., Alloys of Platinum and Early Transition Metals as Oxygen Reduction
Electrocatalysts. Nat. Chem., 2009. 1: p. 552-556.
56. Viswanathan, V., et al., Universality in Oxygen Reduction Electrocatalysis on Metal
Surfaces. ACS Catal., 2012. 2: p. 1654-1660.
38
Chapter 2
Analysis of Double Layer and Adsorption Effects at the Alkaline Polymer Electrolyte-
Electrode Interface and the Development of a Quaternary Ammonium Poisoning Resistant
Electrocatalyst for Methanol Oxidation
2.1 Introduction
Over the years methanol electro-oxidation in alkaline media has been the subject of
considerable research interest for applications such as anion exchange membrane direct methanol
fuel cells (AEM-DMFCs)[1-6]. The use of alkaline media in direct methanol fuel cell
applications provides several advantages over the proton exchange membrane (PEM)-based
analogues. It is well known that the process of methanol dehydrogenation and the removal of
COads intermediate species in alkaline media is far more facile in comparison with acidic
electrolytes due to the abundance of hydroxide anions[3, 7-9]. Furthermore, the use of alkaline
media opens up the possibility of using transition metal based electrocatalysts that would
otherwise be unstable and subject to dissolution in acidic media[10, 11].
A major obstacle for the advancement of AEM-DMFCs involves the development of
stable anion-exchange membranes and ionomers with sufficient ionic conductivity[12, 13].
Recent efforts typically rely on the use of quaternary ammonium species, e.g.
tetramethylammonium, as charge transfer groups for conducting hydroxide ions between the
cathode and anode[13-15]. Although progress has been made over the past 5-10 years, the ionic
conductivity and stability of anion exchange polymers still needs considerable improvement. The
addition of alkali metal hydroxide salts to the liquid fuel feed has been reported to significantly
improve ionic conductivity and thus the cell performance, although this is highly undesirable as
39
it ultimately leads to carbonate precipitation issues[12]. In addition to low ionic conductivity and
low thermal stability, the quaternary ammonium groups used to conduct hydroxide ions in AEMs
can also induce potential losses through electrostatic interactions and adsorption, i.e. poisoning,
of the catalyst surface, especially in the case of Pt[16].
The quaternary ammonium poisoning issue that arises with the use of Pt stems from both
the solvation energy of the quaternary ammonium species and the potential of zero charge (PZC)
of the Pt surface. Quaternary ammonium species are characterized as weak adsorbates that are
poorly solvated in water[17]. The low solvation energy of quaternary ammonium species leads to
favorable electrostatic interactions with a surface of the opposite charge, in contrast to alkali
metal cations in which the solvation energy is higher than that of electrostatic adsorption. The
potential of zero charge of Pt also plays an important role in the adsorption of such species. The
PZC of Pt has been reported as -0.4 V vs. NHE (0.37 V vs. RHE) at pH = 12[18] and -0.11 V vs.
NHE (0.55 V vs. RHE) at pH = 11.2[19, 20]. Below the PZC of Pt, the negatively charged
electrode surface favors the adsorption of weakly solvated and positively charged quaternary
ammonium species. The electrostatic adsorption of the charge transfer groups ultimately blocks
the catalyst surface, both decreasing the active surface area and preventing the adsorption of OH-
. It is not until higher potentials (E > PZC) that the positively charged moieties are repelled from
the electrode surface and sufficient adsorption of OH- begins to occur. This has a considerable
effect on the electrocatalysis of small alcohol molecules, i.e. MeOH[16], therefore anode
catalysts for use in an AEM environment must exhibit a high tolerance to quaternary ammonium
poisoning in order to provide high performance.
Studies have shown that alloying Pt with other metals can significantly improve the
catalytic activity and CO tolerance in comparison with pure Pt catalysts[21-23]. The increase in
40
activity is generally associated with increasing the oxophilicity of the catalyst surface, allowing
for the adsorption of hydroxide species at lower overpotentials to aid in the removal of adsorbed
CO from adjacent Pt sites, i.e. the bifunctional mechanism. Several metal oxide supports have
been proposed in the past, including those based on Ni[24, 25], Sn[26-28], Ce[24], Mn[24, 29],
Ti[30-32], Ru[33, 34], and Nb[33-35]. Among the materials investigated, Ni has been studied
extensively. In the case of Pt-Ni alloys the enhanced activity was explained in terms of the bi-
functional mechanism and was also supported by DFT calculations[36]. It has been shown that
Ni serves to optimize the electronic structure of Pt sites via the ligand effect, thus lowering the
binding energy of COads and promoting its removal. The former enhancement is likely due to the
fact that Ni forms a Ni(OH)2 passivation layer in the alkaline environment[25]. Pb is also a
promising candidate as it offers a significant enhancement in the electro-oxidation of alcohols
(MeOH, EtOH, etc.) and other organic fuels[23, 29, 37-41]. It was suggested that ordered
intermetallic phases, such as Pt-Pb, exhibited improved catalytic activity for MOR as evidenced
by a lower onset potential[42].
This chapter will begin with a review of the research conducted to characterize the
adsorption issues arising at the anode-polymer electrolyte interface[16]. In the following sections
the extent of ammonium poisoning in alkaline media will be addressed both quantitatively and
qualitatively. Furthermore, the electrostatic effects of adsorbed quaternary ammonium species
will be examined through the use of transition metal redox probes. In response to the poisoning
issues encountered in an alkaline AEM environment, we then report herein the characterization
and electrochemical methanol oxidation performance of Pt deposited on the NiPb/C support
(denoted as Pt/NiPb/C) in alkaline media. This catalyst exhibits high electrocatalytic methanol
oxidation activity and a high resistance to quaternary ammonium poisoning.
41
2.2 Experimental
2.2.1 Electrochemical Characterization of the Anode-Polymer Electrolyte Interface
Ammonium contamination studies were conducted on Pt/C modified glassy carbon
electrodes. A catalyst ink was prepared from 25 mg of 30% Pt/C (BASF) in 10 mL Millipore
H2O and 10 mL isopropyl alcohol. The ink was sonicated for 30 minutes. Exactly 5 μL of ink
was deposited on a glassy carbon rotating disk electrode (Ageo = 0.247 cm2) to yield a total
catalyst loading of 7.5 μgPt/cm2. Prior to ink deposition, the electrode was polished with 0.05 μm
alumina slurry and sonicated twice in a sonic bath for 30 seconds each in a 50:50 solution of
Millipore H2O and isopropyl alcohol. The reference electrode used was a reversible hydrogen
electrode (RHE) prepared from a solution of 0.1 M KOH. At the beginning of each experiment,
the electrode was cycled 20 times from 0.05 V to 1.2 V at 50 mV/s in 0.1 M KOH followed by
five cycles at 20 mV/s. Absolute methanol was then added to the KOH solution to obtain a total
MeOH concentration of 0.5 M. Methanol oxidation was then performed by holding the potential
at 0.6 V vs. RHE for 900 seconds. Subsequently, an aliquot of a given contaminant was added to
bring the total contaminant concentration to 1 mM and the potential was again held at 0.6 V for
900 seconds. This procedure was repeated for the remaining contaminant concentrations of 5, 10,
20, 40, 60, and 120 mM. A new ink coating was used in the study of each contaminant.
Contaminants investigated included tetramethylammonium hydroxide (TMAOH, 25% w/w aq) ,
tetraethylammonium hydroxide (TEAOH, 35% w/w aq), tetra-n-propylammonium hydroxide
(TPAOH, 40% w/w aq.), and benzyltrimethylammonium hydroxide (BTMAOH, 25% w/w aq.)
all from Alfa Aesar. Electrochemical measurements were made using an Autolab
Potentiostat/Galvanostat Model PGSTAT30 (Metrohm USA).
42
The electrochemical behavior of the transition metal complexes was studied at an alkaline
ionomer modified glassy carbon electrode. A glassy carbon rotating disk electrode (Ageo = 0.247
cm2), after polishing and pretreatments described above, was allowed to dry after deposition of 5
μL of 5 wt.% anion exchange ionomer (AS4, Tokuyama Corp., Japan) on the surface. The
electrode was placed in an argon saturated 0.1 M NaOH solution and subjected to 20 cycles
from 0.05 V to 1.2 V at 50 mV/s and five cycles at 20 mV/s. 10 mM of K3Fe(CN)6 was then
added to the solution and the solution was purged with argon for an addition 10 minutes before
collecting data. This experiment was repeated using 10 mM Co(NH3)6Cl3 in place of K3Fe(CN)6.
2.2.2 Preparation of Pt/NiPb/C Electrocatalyst
The layered Pt/NiPb/C electrocatalyst reported herein was synthesized via the sequential
aqueous impregnation method using NaBH4 as a reducing agent. The catalyst consists of 20
wt.% Pt deposited on a composite support consisting of 60 wt.% Ni-Pb (2:1 atomic ratio Ni:Pb)
on carbon (Black Pearls Carbon-2000, Cabot Corp.). The carbon black was dispersed in ~200
mL solution containing stoichiometric amounts of Ni(NO3)2 and Pb(NO3)2 (Alfa Aesar, 99.99%
metal basis) metal salt precursors in ultra-pure water (18.2 MΩ, Milli-Pore filtration system).
The suspension was homogenized for 30 min before drop-wise addition of an aqueous NaBH4
solution (3 mL, 2.5 molar excess) and subsequent stirring for 1 h. The product was filtered and
rinsed 3 times with Milli-Pore water. The resulting NiPb/C powder was then dried in a vacuum-
oven at 80 oC overnight. Platinum was deposited on the NiPb/C support by repeating the above
procedure: The NiPb/C powder was dispersed in ~100 mL of aqueous H2PtCl6 (Alfa Aesar,
99.99% metal basis) solution and stirred for 30 min, followed by drop-wise addition of NaBH4
solution (3 mL, 2.5 molar excess). The final product was stirred for 1 h, filtered and rinsed with
43
ultra-pure water, and then dried in the vacuum-oven at 80 oC overnight. Commercial 46.2 wt.%
Pt/C (Tanaka Corp, Japan), 73.5 wt.% PtRu/C (Tanaka Corp, Japan), and 40 wt.% Pt/C (ETEK,
USA) electrocatalysts were used as received.
2.2.3 XRD and TEM Characterization:
The dispersion of Pt on the NiPb/C support was observed using transmission electron
microscopy (TEM, JEOL 2100). Powder X-ray diffraction (XRD) analysis of catalyst samples
was carried out with a Rigaku Ultima IV diffractometer using CuKα radiation (λ = 1.542 Å)
operating at 40 kV and 44 mA with a 0.05o step and 5 sec hold per step in order to obtain high
resolution XRD signals. XRD analysis was performed using Rigaku PDXL software.
2.2.4 Electrochemical Characterization of Pt/NiPb/C
Electrochemical measurements were performed in a conventional single-compartment,
three-electrode cell at ambient temperature (298 K). The glassy carbon working electrode (0.247
cm2 surface area) was polished with 0.05 µm alumina paste and rinsed with Milli-Pore water
prior to catalyst ink deposition. Catalyst ink suspensions, which consisted of an appropriate
amount of catalyst, isopropyl alcohol, ultra-pure water, and 5 wt.% AS4 anion exchange ionomer
solution (Tokuyama Corp., Japan), were ultrasonically mixed for at least 30 min. Exactly 5 μL of
ink was drop-cast on the mirror-polished glassy carbon working electrode to achieve a mass
loading of 8 μgPt/cm2. The reference electrode was a reversible hydrogen electrode (RHE) in 0.1
M KOH solution and Pt mesh was used as the counter electrode. All potentials are reported
versus the RHE scale. Before measurements, the 0.1 M KOH electrolyte was purged with Ar for
30 min.
44
Cyclic voltammetry (CV), chronoamperometry (CA), and electrochemical impedance
spectroscopy (EIS) measurements were performed using an AutoLab PGSTAT30 (Metrohm,
USA) and the Nova 1.10 software package. Cyclic voltammograms were collected within a
potential range of 0.05 V to 1.00 V (PtRu/C was measured from 0.05 V to 0.80 V) at 20 mV s-1
.
The CA results were generated within a potential range of 0.40 V to 0.80 V using 50 mV step
size and 180 s hold time. Additional CV measurements were made to investigate the effect of the
anion exchange ionomer on the MOR activity. In this case, 50 μL of either 5 wt.% Nafion
solution or 5 wt.% AS4 solution was used in the catalyst ink suspensions described above. Cyclic
voltammograms were collected in Ar purged 0.1 M KOH and then in 0.1 M KOH + 0.5 M
MeOH at a scan rate of 10 mV/s. The EIS measurements were recorded within a 32 kHz to 1 Hz
frequency range at E = 0.70 V with an amplitude of 10 mV. EIS data was analyzed using non-
linear least-squares fit of Nyquist plot to simple Randles cell circuit.
2.3 Result and discussion
2.3.1 Electrochemical Characterization of the Anode-Polymer Electrolyte Interface
In order to address and quantify the issue of quaternary ammonium poisoning on
platinum, the oxidation of methanol on platinum was studied as a function of the ammonium
group adsorbed on the surface of the electrode. The first three cations studied were quaternary
ammoniums with different length alkyl chains (methyl, ethyl and propyl). The fourth one was
benzyltrimethylammonium hydroxide. The hydroxide salt of these quaternary ammonium cations
was added to a 0.5 M MeOH solution. The percent loss in current density after 900 sec potential
control at 600 mV (vs. RHE) in 0.1 M KOH was then recorded as a function of quaternary
ammonium as shown in Figure 2.1. In general, all ammonium cations showed a significant
45
adsorption effect on the platinum surface resulting in a severe loss of methanol oxidation
activity. The decrease in current became more pronounced at higher concentrations of the added
cations. TMA+ showed the smallest drop in current, whereas the other three cations with bulkier
substitute groups resulted in substantially greater loss in current. This observation is consistent
with previous literature that showed a higher surface inhibition with the longer alkyl length on
the ammonium group[43].
Conc. of Contaminant [mM]
0 20 40 60 80 100 120
% L
oss in
Cu
rre
nt
De
nsity
-100
-80
-60
-40
-20
0
TMAOH
TEAOH
TPAOH
BTMAOH
Figure 2.1. Percent loss in current density as a function of contaminant concentration in 0.1 M
KOH + 0.5 M MeOH + x mM contaminant. Glassy carbon disk electrode (0.247 cm2) with 5 μL
of 30% Pt/C ink deposited (Loading = 7.5 ugPt/cm2). Steady state currents obtained from
chronoamperometry at 0.6 V vs. RHE at 900 seconds.
These results with different quaternary ammonium cations in aqueous electrolytes show
that quaternary ammonium ion adsorption on Pt surfaces lowers the rate of methanol oxidation
by blockage of the active catalyst surface area. However, the solid polymer electrolyte, ion-
conducting medium in a polymer-based fuel cell is different from that of free ions in an aqueous
46
solution. A fuel cell electrode consists of quaternary ammonium cations with little chain segment
mobility tethered to a polymer backbone. A more realistic representation of the reaction medium
in a fuel cell on a planar electrode surface was investigated by Unlu et al.[16], in which an anion
exchange polymer with quaternary ammonium ion sites, poly-tetramethylammonium hydroxide
(PTMAOH), was used as the polymer electrolyte without additional alkali metal hydroxide. This
polymer electrolyte provides high anion conductivity with high viscosity in the electrolyte phase,
as would occur in a fuel cell electrode.
Methanol oxidation is an inner-sphere electron transfer reaction and methanol is a neutral
molecule. Therefore, the Frumkin effect described in Chapter 1 (See Section 1.3.2) does not
directly affect the methanol species. Instead, the oxidation of alcohols on platinum requires the
presence of adsorbed hydroxide species to proceed through a Langmuir-Hinshelwood
mechanism. Transport of hydroxide anions across the diffuse layer and subsequent specific
adsorption on the electrode surface are necessary for alcohol oxidation. The concentration of
hydroxide at the OHP, OHP
OHC , available for adsorption is controlled by the bulk concentration,
charge of the ion, and ϕ2, as shown in below in Eq. 2.1.
2expOHP bulk
OH OH
zFC C
RT
(2.1)
The double layer structure of an electrode-electrolyte interface significantly differs in
polymer electrolytes and aqueous electrolytes. The double layer structures in NaOH, TMAOH,
and polymer electrolyte surfaces are depicted for a negatively charged electrode surface in
Figure 2.2. When the electrode potential is negative of PZC in a NaOH solution, there will be a
potential gradient from the electrode to the bulk solution potential. In TMAOH solution, TMA+
species are adsorbed on the Pt surface, particularly because ϕM < ϕS, forming a compact inner
47
layer of positive charges. In contrast, TMA+ moieties tethered to a polymer chain are not as
mobile as free TMA+ causing the excess charge in polymer electrolytes to be distributed across
the diffuse layer. The mobile TMA+ cations can form a more compact double layer due to their
mobility. These variations in the structure of the double layer result in large differences in ϕ2.
The order of ϕ2 would be 2 2 2
T N P in TMAOH, NaOH, and Polymer electrolytes,
respectively. However, even though the relatively more positive value of 2
T would favor the
transport of hydroxide ions to the OHP, this factor is outweighted by the blockage of the
electrode surface, which decreases the available catalyst surface area and therefore the oxidation
rate.
Figure 2.2. Expected potential profile of anode double layer interface in NaOH, TMAOH, and
polymer electrolyte solutions when the electrode potential is more negative than PZC[16].
48
When the TMAOH and polymer electrolyte solutions are considered, the TMA+ cation is
more mobile than the tethered polymer electrolyte cation which could lead to a greater degree of
adsorption and loss of catalyst surface area, however, in the case of Unlu et al.[16] the polymer
electrolyte cations show smaller currents. The lower reaction rate in polymer electrolyte solutions
is probably due to the more negative value of 2
P than that of 2
T . The negative value of 2
P
inhibits the transport of hydroxide to the electrode surface, limiting the reaction rate. This effect
on hydroxide transport and adsorption explains the enhancement in the performance when alkali
metal hydroxide is added to the fuel, as shown in the literature[16]. The addition of NaOH to
polymer electrolytes introduces two effects; (i) the bulk concentration of OH- ions increases in
Eq. 2.1 and (ii) the mobile Na ions lead to a decrease in the thickness of diffuse layer, resulting
in lower ϕ2. These changes lead to higher apparent reaction rates.
The proposed ϕ2 effect on hydroxide transport and subsequent adsorption are difficult to
quantify. Instead, charged redox couples were added to the electrolyte to help identify changes
and probe the double layer structure and consequent ϕ2 effect. In this study, the Fe(CN)63-/4-
and
Co(NH3)62+/3+
redox couples were employed in order to probe the electrode surface for evidence
of the ϕ2 effect in anion exchange ionomers and membranes. Cyclic voltammetry was used to
observe the redox behavior of each couple on a glassy carbon electrode in the presence and
absence of an anion exchange ionomer film (Tokuyama, AS4 which contains quaternary
ammonium cation ion-exchange groups). Figure 2.3 displays the cyclic voltammograms for the
Fe(CN)63-/4-
redox couple on a GC electrode with and without a deposited film of AS4 ionomer.
Typically, the current magnitude cannot be used to make a direct comparison between solution
and AEM environment because both concentration and diffusion coefficients change. However,
the difference between the oxidation and reduction peaks can be evaluated to define a change in
49
reaction rates. From Figure 2.3, it can be seen that the peak separation (ΔEp) decreases from
279.5 mV in the absence of AS4 to 131.0 mV in presence of the AS4 ionomer film. In contrast,
Figure 2.4 shows the CV profiles for the Co(NH3)62+/3+
redox couple on a GC electrode with and
without a deposited film of AS4 ionomer. The current densities are lower with AS4 ionomer than
without ionomer because AS4 ionomer mostly excludes positively charged species, decreasing
the concentration on the electrode surface, i.e. lower current. However, it is critical to note that
there is a slight increase in ΔEp from 65.3 mV to 83.9 mV in the presence of ionomer.
Figure 2.3. CV of Fe(CN)63-/4-
redox couple in Ar saturated 0.1M NaOH + 10mM K3Fe(CN)6
solution. Scans performed on glassy carbon disk electrode with 5 μL of 5 wt.% AS4 ionomer
deposited on surface. Geometric surface area = 0.247 cm2. Scan rate = 20 mV/s.
50
Figure 2.4. CV of Co(NH3)62+/3+
redox couple in Ar saturated 0.1M NaOH + 10mM
Co(NH3)6Cl3 solution. Scans performed on glassy carbon disk electrode with 5 μL of 5 wt.%
AS4 ionomer deposited. Geometric surface area = 0.247 cm2. Scan rate = 20 mV/s.
For these redox couples, the electrode potentials are more positive than PZC (ca. 0.55 V vs.
RHE[19, 20]) at the electron transfer potentials and the reaction is believed to proceed through
an outer-sphere electron transfer. The double layer structure differs from the methanol oxidation
region where the electrode potential is more negative than PZC. Figure 2.5 shows the double
layer structure when the electrode potential is greater than PZC in the presence and absence of
AEM ionomer. As a distinct difference, the diffuse layer in an AEM becomes thicker relative to
that in the absence of an AEM. Consequently, ϕ2 is more positive in the presence of AEM. For a
negatively charged redox couple, both exponential terms favor the electron transfer, i.e. higher
kinetics. This is consistent with the lower peak splitting for Fe(CN)63-/4-
in Figure 2.3. However,
for positively charged species, the driving force still favors the reaction but the transport of these
species is inhibited. This inhibition factor is greater than the enhancement in driving force,
51
resulting in lower reaction kinetics. This is consistent with the greater peak splitting in Figure
2.4.
Figure 2.5. Expected potential profile of anode double layer interface in the presence and
absence of anion exchange ionomer when the electrode potential is more positive than PZC[16].
Now that the issues surrounding the specific adsorption of quaternary ammonium species
have been addressed both qualitatively and quantitatively we will consider the rational design of
a Pt-based catalyst capable of delivering high methanol electro-oxidation activity while showing
a high resistance to quaternary ammonium poisoning. The following sections will detail the
synthesis and characterization of a highly oxophilic NiPb/C support for Pt nanoparticles.
52
2.3.2 XRD and TEM Characterization
The powder X-ray diffraction patterns of the layered Pt/NiPb/C and commercial Pt/C
samples are shown in Figure 2.6. Both samples display the anticipated peaks at approximately
39.5 and 46.0 degrees, which are characteristic of the Pt fcc (111) and (220) crystallite
reflections. The absence of any significant change in the lattice parameter (See Table 2.1) or shift
in the primary fcc diffracting regions of the layered Pt/NiPb/C material relative to that of Pt/C
suggests that the deposited Pt exists as a separate phase and is not alloyed with the NiPb/C
support to any significant extent. In addition, the layered Pt/NiPb/C sample exhibits features
characteristic of a hydro-cerussite Pb3(CO3)2(OH)2 phase. Although crystalline Ni/Ni-Ox phases
were not detected in the XRD analysis, the slight shift of the characteristic hydro-cerussite
reflections to higher angles suggests that some Ni is likely incorporated into the Pb crystallite
phase. Additionally, some Ni may also exist as an amorphous surface oxide layer. It should also
be noted that a PtNiPb/C material was synthesized and characterized by Chen et al.[44] that did
not contain hydro-cerrusite reflections in the diffractograms but did show a shift in the Pt fcc
reflections (and thus incorporation of Ni and Pb into the platinum crystal lattice).
The corresponding coherent domain sizes as calculated by the Scherrer equation are
shown in Table 2.1. From Table 2.1 it can be seen that the calculated crystallite domain sizes are
nearly the same for both the commercial Pt/C and the PtRu/C material (1.7 and 1.5 nm,
respectively), while the size of the Pt/NiPb/C crystallites are more than twice the size of the
commercial samples. This indicates that the surface area of the Pt/NiPb/C sample is likely much
less than that of the commercial samples. As a result, any enhancement in MOR activity for
Pt/NiPb/C with respect to that of Pt/C or PtRu/C is unlikely to be a result of increased surfaces
area.
53
Figure 2.6. Powder X-ray diffraction patterns for nanocrystalline Pt/C and Pt/NiPb/C samples.
Table 2.1. XRD and electrochemical results for various catalysts.
XRD results Electrochemical results
Catalyst Lattice
parameter (nm)
Coherent domain
size (nm)
Eonset
(V)
Epeak
(V)
iPeak
(mA/cm2)
iat 0.7V
(mA/cm2)
Pt/C 0.3918 1.7 0.356 0.959 12.98 2.20
PtRu/C 0.3868 1.5 0.299 0.761 2.72 2.40
Pt/NiPb/C 0.3912 4.4 0.252 0.918 15.73 8.61
Representative TEM images of the layered Pt/NiPb/C electrocatalyst are shown in Figure
2.7. From Figure 2.7 it can be seen that the nanoparticles are well dispersed on the carbon
support. The average size of the supported nanoparticles obtained from TEM analysis is less than
5 nm, which is in agreement with the coherent domain size calculated from the Scherrer
equation. The high dispersion of nanoparticles is likely a major factor contributing to the high
MOR activity, though it is not the only source responsible for the enhancement in activity.
54
Figure 2.7. TEM images of the Pt/NiPb/C catalyst. Insert is higher magnification.
2.3.3 Electrochemical Characterization of Pt/NiPb/C
The electrochemical methanol oxidation response of the corresponding catalysts in 0.1 M
KOH and 1.0 M methanol solution are shown in Figure 2.8a. The MOR activity of the layered
Pt/NiPb/C is compared with the commercial Pt/C and PtRu/C catalysts. The layered Pt/NiPb/C
material displays the highest overall MOR activity among these catalysts. It shows the highest
peak current (ipeak) and the lowest onset potential (Eonset) (See Table 2.1). Additionally, the
chronoamperometric response was measured, which provides further data on catalyst activity and
stability. The steady state current measured at 180 s for each potential step can be seen in Figure
2.8b. The same general trend in methanol oxidation activity that is shown by the cyclic
voltammograms is observed again here, with the Pt/NiPb/C material showing higher oxidation
currents across the entire potential range examined. From Figure 2.8a and Table 2.1 it can be
seen that the layered Pt/NiPb/C and PtRu/C have a similar onset potential, with methanol
oxidation beginning at roughly 250 and 300 mV vs. RHE, respectively, which is approximately
55
50 to 100 mV lower than Pt/C. Although the onset of methanol oxidation occurs at roughly the
same potential for Pt/NiPb/C and PtRu/C, the methanol oxidation current is considerably higher
for Pt/NiPb/C across the entire MOR potential range. In general, the low activity of PtRu/C can
be attributed to the weak adsorption of methanol on Ru sites and the dilution of Pt surface sites
required for methanol dehydrogenation[8]. Furthermore, it is interesting to note that the peak
current potential (Epeak) for Pt/NiPb/C is positioned nearly 80 mV beyond the Epeak for PtRu/C
even though Eonset for the two materials is about the same. On the basis of the bifunctional
mechanism, this suggests that OH adsorption (and subsequent COads removal) begins at roughly
the same potential for both Pt/NiPb/C and PtRu/C. However, at more positive potentials it is
likely that the higher surface coverage and formation of more strongly bound oxygenated species
begins to inhibit further removal of the COads intermediate on PtRu/C, thus decreasing the
availability of active sites for MOR[8]. In the case of Pt/NiPb/C, however, it appears that
sufficient surface coverage and binding of OHads exists over a wider potential range and that the
formation of strongly bound oxo species does not begin to inhibit MOR catalysis until higher
potentials.
56
Figure 2.8. Electrochemical measurement of Pt/C, PtRu/C, and Pt/NiPb/C catalysts in 0.1 M
KOH and 1.0 M methanol at 298 K. Current densities are normalized to the geometric surface
area. (a) CV measurements at a scan rate of 20 mV/s. (b) Steady state currents obtained after 180
s from chronoamperometric tests ranging from 400 to 800 mV with 50 mV steps.
Enhancement of MOR activity has been previously reported for PtNi-based catalysts[36,
45, 46]. Sun et al.[36] reported higher specific activities for PtmNin/C and decreased CO
adsorption when compared to Pt/C, which was attributed to modification of the Pt electronic
structure and weakened interactions between PtmNin clusters and adsorbed CO. Wu et al.[45]
reported lower onset potentials (by approx. 50 mV) and higher peak current densities (approx.
2x) for PtNi (4:1) supported on MWCNTs over Pt/MWCNTs. Furthermore, Jiang et al. have
shown that Ni-Ox electrodes can oxidize MeOH at much higher anodic potentials, with the
Ni2+/3+
redox transition occurring at nearly the same potential as the MOR[36]. Lyons et al. have
also explained that in the alkaline environment, Ni and other transition metals assume a hydrous
oxy-hydroxide layer during potential cycling[46]. This hydrous oxy-hydroxide layer should
exhibit sufficient chemical reversibility (similar to the Ti, Nb & W bronzes described by Jaksic
et al.[47]) to facilitate spillover of OHad – thus promoting complete MOR.
57
Even greater enhancement of alcohol oxidation activity has been reported for Pb-
containing catalysts[37, 38, 40, 41, 44, 48, 49]. Sun et al.[38] reported that the Pt mass
normalized peak current density for PtPb/C was nearly 2.1 times that of Pt/C. Furthermore,
Matsumoto et al.[37] reported BET surface area normalized currently densities for PtPb/C that
were nearly 8 times higher than those of PtRu/C at the same potential. In both the cases the MOR
onset potential for PtPb was reported to be significantly lower (by 100-150 mV) than those in the
case of Pt/C or PtRu/C. Chen et al.[44] reported a Pt-Ni-Pb/C catalyst that reached peak current
densities in acidic media that were twice as high as Pt/C, though the dramatic shift to a more
negative MOR onset potential relative to Pt/C was not observed as in alkaline media[37, 38].
Table 2.1 shows that the layered Pt/NiPb/C material displays a peak current density
approximately 1.2 times that of Pt/C, although the current density at intermediate potentials, i.e.
0.7 V vs. RHE (see Table 2.1) is up to 4 times higher than that of Pt/C. The enhancements
reported for Pt-Pb catalysts have been described in terms of the bifunctional effect and the
electronic effect, although Sun et al.[38] claimed that the bifunctional effect is unlikely to play a
major role and attributed the observed enhancements to a partial charge transfer from Pb to Pt,
resulting in a modification of the Pt d-orbitals and a subsequent decrease in CO adsorption. It is
important to note that these reports have dealt mostly with single-phase bitmetallic PtPb or PtNi
alloys, which is not the case for the layered Pt/NiPb/C material described herein that includes a
hydro-cerussite (Pb3(CO3)2(OH)2 phase in addition to the Pt fcc crystal phase. The effect of Pb in
the Pt/NiPb/C catalyst has not been fully understood yet and further evaluations of the
synergistic effects of Ni and Pb will likely require the employment of DFT calculations.
The interfacial effects of the anion exchange ionomer (AEI) and the extent of poisoning
resulting from the presence of quaternary ammonium species were investigated by monitoring
58
the electrochemical response of each catalyst in the presence and absence of AEI (AS4, which
contains quaternary ammonium cation ion-exchange groups) via cyclic voltammetry and
electrochemical impedance spectroscopy (EIS). The methanol electro-oxidation curves for 40%
Pt/C and Pt/NiPb/C with and without AEI in 0.1 M KOH + 0.5 M methanol solution are shown
in Figure 2.9. A 5 wt.% Nafion solution was used as a catalyst binder in the absence of AEI.
From Figure 2.9 it can be seen that the Pt/NiPb/C catalyst with AEI exhibits a significant
increase in methanol oxidation current when compared to the MOR activity in the absence of
AEI, indicating that the presence of quaternary ammonium facilitates the diffusion of OH- to the
Pt surface. This also indicates that the AEI is not blocking the active catalyst surface area. This is
in contrast to what is observed for Pt/C. In Figure 2.9 it is shown that the presence of AEI
decreases the overall magnitude of the methanol oxidation current for Pt/C. The insert in Figure
2.9 shows the cyclic voltammograms of both Pt/C and Pt/NiPb/C in 0.1 M KOH without the
addition of MeOH. From Figure 2.9 it can be seen that the addition of AEI to the catalyst ink
decreases the active catalyst surface area on Pt/C, as evidenced by the loss of charge in the Hupd
region. Although there are no strong Hupd features for Pt/NiPb/C, there is an increase in the
hydrogen oxidation current in the presence of AEI following evolution of hydrogen from the
cathodic scan.
It is interesting to note that the formation of surface oxides is also enhanced in the
presence of the ionomer for Pt/C. This is likely due to the fact that as one goes positive of the
PZC on Pt (E > ~0.55 V vs. RHE) the adsorbed quaternary ammonium are repelled by the
increasing positive charge of the catalyst surface. As quaternary ammonium are repelled from the
Inner Helmholtz Plane (IHP), it allows OH- (which is now attracted by the positively charged
surface) to access the IHP. The potential regime for MOR on Pt/C only ranges from
59
approximately 0.4 to 0.9 V vs. RHE. The onset of MOR therefore coincides with the adsorption
region of OH- while the formation of the surface oxide layer at more positive potentials (E > 0.9
V) inhibits the oxidation MeOH, resulting in a sharp decrease in the MOR activity[8, 50].
Therefore, when the potential is sufficiently positive to repel adsorbed quaternary ammonium (E
> 0.55 V) the electrode potential is already nearing the regime where it begins to bind oxygen too
strongly to allow for optimal MeOH coverage. In the case of Pt/NiPb/C, the support has a very
high oxo/hydrophilic character that likely inhibits quaternary ammonium adsorption in this
potential regime while allowing for enhanced OH- adsorption and transport to adjacent Pt MOR
active sites. The effects can also be seen in the EIS (See Figure 2.10), where the presence of AEI
decreases both the high frequency resistance and the charge transfer resistance on Pt/NiPb/C.
The accompanying decrease in the charge transfer resistance suggests that not only does the
Pt/NiPb/C catalyst have a high resistance to quaternary ammonium poisoning, but the AEI in fact
exhibits a significant promoting effect on MOR activity. The opposite case is observed for Pt/C,
where the high frequency resistance and charge transfer resistance are increased. In the case of
Pt/C this suggests that the AEI is blocking the active surface area and decreasing the number of
sites available for OH- adsorption. This effect is in agreement with Unlu et al.[16], where it was
determined that the specific adsorption of quaternary ammonium and possible ϕ2 effects all work
to inhibit MOR until the anode potential becomes positive of PZC.
60
Figure 2.9. Cyclic voltammograms of 40% Pt/C and Pt/NiPb/C catalysts with Nafion or AS4
ionomer used as a catalyst binder. Scans were taken in Ar saturated 0.1 M KOH + 0.5 M
methanol at 298 K at a scan rate of 10 mV/s. Cyclic voltammograms collected at a scan rate of
10 mV/s in Ar saturated 0.1 M KOH are shown in the insert.
Figure 2.10. Electrochemical impedance spectroscopy of 40% Pt/C and Pt/NiPb/C with either
Nafion or AS4 ionomer used as a catalyst binder in 0.1 M KOH + 0.5 M MeOH. EIS data was
collected at E = 0.70 V vs. RHE from 32 kHz to 1 Hz with a amplitude of 10 mV. The insert
shows the high frequency resistance.
61
The dramatic improvement in MOR activity of the Pt/NiPb/C over the commercial Pt/C
and PtRu/C is likely due to the morphology of the catalyst itself. The NiPb/C support appears to
improve the quaternary ammonium poisoning resistance due to the high oxo/hydrophilicity and
preference of both Ni and Pb to form surface hydroxide and oxyhydroxide species in alkaline
media. Pt is known to be poisoned by quaternary ammonium species in addition to the COads
intermediate in MOR. Therefore, the presence of an oxophilic metal in close proximity to Pt sites
is necessary to aid not only in the removal of COads but also to inhibit the adsorption of
quaternary ammonium species. Although PtRu/C is well known to promote the removal of CO,
the alloyed material may not provide enough Ru sites in close proximity to Pt sites to effectively
inhibit quaternary ammonium adsorption and provide sufficient OH- spillover. We suggest that
by depositing highly dispersed Pt nanoparticles onto the highly oxophilic NiPb/C support, as
opposed to alloying or depositing oxophilic metals onto Pt particles, that we effectively inhibit
the specific adsorption of quaternary ammonium on Pt active sites while providing a high rate of
OH- transport from the surrounding support, thus significantly improving the MOR activity in
alkaline media.
2.4 Conclusion
Quaternary ammonium cations in solution were shown to inhibit the oxidation of
methanol through specific adsorption. Specific adsorption, migration in the diffuse layer due to
hydroxide repulsion away from the electrode, and possible ϕ2 effects, all work against an
efficient electrode structure until the anode potential becomes positive of PZC. In this study, the
effect of PZC on anode operation was explored. Transition metal redox complexes in a polymer
electrolyte showed behavior consistent with the ϕ2 effect.
62
The deposition NiPb onto a carbon support prior to Pt deposition can significantly
improve the Pt electrochemical performance for methanol oxidation versus Ru-alloying. TEM
images of Pt/NiPb/C catalyst indicate that the Pt nanoparticles are well dispersed on the NiPb/C
support. In the case of Pt/NiPb/C, the addition of anion exchange ionomer to the catalyst ink
suspension is an important factor for enhancing the electrochemical performance. The CV, CA,
and impedance results with/without AEI indicate that NiPb/C support imparts a strong resistance
to quaternary ammonium poisoning of the Pt catalyst while dramatically improving MOR
activity. It is proposed that the AEI facilitates both diffusion of OH- from the bulk electrolyte
through the ionomer and surface diffusion of OHads from the NiPb interlayer. Further analysis to
elucidate the origin of the synergistic effects of NiPb/C on Pt is currently in progress.
2.5 Acknowledgments
The authors wish to thank the Army Research Office for the financial support under a
single investigator grant. The authors gratefully acknowledge the financial support from the U.S.
Army and DuPont Corporation especially Dr. Deryn Chu (US Army) and Shoibal Banerjee
(DuPont). Materials support from Tokuyama Corp. is also gratefully acknowledged. Murat Ünlü
and Paul A. Kohl of Georgia Institute of Technology and Nagappan Ramaswamy of
Northeastern University are gratefully acknowledged for their significant contributions to the
development of interfacial model of the anode-polymer electrolyte interface. Additionally,
Myoungseok Lee and Michael Bates of Northeastern University are gratefully acknowledged for
their assistance in the development and electrochemical characterization of the Pt/NiPb/C
catalyst.
63
2.6 References
1. Liu, J.P., et al., Kinetics of ethanol electrooxidation at Pd electrodeposited on Ti.
Electrochem. Commun., 2007. 9: p. 2334-2339.
2. Hou, H.Y., et al., Alkali doped polybenzimidazole membrane for alkaline direct methanol
fuel cell. Int. J. Hydrogen Energy, 2008. 33: p. 7172-7176.
3. Jayashree, R.S., et al., Air-Breathing Laminar Flow-Based Direct Methanol Fuel Cell
with Alkaline Electrolyte. Elelctrochem. Solid-State Lett., 2006. 9: p. A252-A256.
4. Zhu, C., S. Guo, and S. Dong, Rapid, General Synthesis of PdPt Bimetallic Alloy
Nanosponges and Their Enhanced Catalytic Performance for Ethanol/Methanol
Electrooxidation in an Alkaline Medium. Chem. Eur. J., 2013. 19: p. 1104-1111.
5. Figueiredo, M.C., et al., Pt catalysts modified with Bi: Enhancement of the catalytic
activity for alcohol oxidation in alkaline media. J. Catal., 2014. 312: p. 78-86.
6. Awasthi, R. and R.N. Singh, Graphene-supported Pd–Ru nanoparticles with superior
methanol electrooxidation activity. Carbon, 2013. 51: p. 282-289.
7. Samant, P.V. and J.B. Fernandes, Enhanced activity of Pt(HY) and Pt-Ru(HY) zeolite
catalysts for electrooxidation of methanol in fuel cells. J. Power Sources, 2004. 125: p.
172-177.
8. Tripkovic, A.V., et al., Methanol electrooxidation on supported Pt and PtRu catalysts in
acid and alkaline solutions. Electrochim. Acta, 2002. 47: p. 3707-3714.
9. Prabhuram, J. and R. Manoharan, Investigation of methanol oxidation on unsupported
platinum electrodes in strong alkali and strong acid. J. Power Sources, 1998. 74: p. 54-
61.
10. Singh, R.N., A. Singh, and Anindita, Electrocatalytic activity of binary and ternary
composite films of Pd, MWCNT and Ni, Part II: Methanol electrooxidation in 1M KOH.
Int. J. Hydrogen Energy, 2009. 34: p. 2052-2057.
64
11. Spendelow, J.S. and A. Wieckowski, Electrocatalysis of oxygen reduction and small
alcohol oxidation in alkaline media. Phys. Chem. Chem. Phys., 2007. 9: p. 2654-2675.
12. Varcoe, J.R. and R.C.T. Slade, Prospects for alkaline anion-exchange membranes in low
temperature fuel cells. Fuel Cells (Weinheim, Ger.), 2005. 5(2): p. 187-200.
13. Hickner, M.A., A.M. Herring, and E.B. Coughlin, Anion Exchange Membranes: Current
Status and Moving Forward. J. Polym. Sci., Part B: Polym. Phys., 2013. 51: p. 1727-
1735.
14. Lee, H.-C., et al., Anion exchange membranes based on novel quaternized block
copolymers for alkaline direct methanol fuel cells. RCS Adv., 2014. 4: p. 10944-10954.
15. Janarthanan, R., et al., A Direct Methanol Alkaline Fuel Cell Based on Poly(phenylene)
Anion Exchange Membranes. J. Electrochem. Soc., 2014. 161: p. F944-F950.
16. Unlu, M., et al., Analysis of Double Layer and Adsorption Effects at the Alkaline Polymer
Electrolyte-Electrode Interface. J. Eletrochem. Soc., 2011. 158: p. B1423-B1431.
17. Anson, F.C., Patterns of Ionic and Molecular Adsorption at Electrodes. Accounts Chem.
Res., 1975. 8: p. 400-407.
18. Lunarska, E., Hydrogen Degradation of Ferrous Alloys, ed. R.A. Oriani, J.P. Hirth, and
M. Smialowski. 1985, Park Ridge, New Jersey: Noyes Publications. 89.
19. Gileadi, E., S.D. Argade, and J.O. Bockris, Potential of Zero Charge of Platinum and Its
pH Dependence. J. Phys. Chem., 1966. 70: p. 2044.
20. Bockris, J.O., S.D. Argade, and E. Gileadi, Determination of Potential of Zero-Charge on
Solid Metals. Electrochim. Acta, 1969. 14: p. 1259.
21. Caillard, A., et al., Structure of Pt/C and PtRu/C catalytic layers prepared by plasma
sputtering and electric performance in direct methanol fuel cells (DMFC). J. Power
Sources, 2006. 162: p. 66-73.
65
22. Wang, D.L., L. Zhuang, and J.T. Lu, An Alloying-Degree-Controlling Step in the
Impregnation Synthesis of PtRu/C Catalysts. J. Phys. Chem. C., 2007. 111: p. 16416-
16422.
23. Halder, A., et al., In situ X-ray absorption spectroscopy on probing the enhanced
electrochemical activity of ternary PtRu@Pb catalysts. Electrochim. Acta, 2013. 108: p.
288-295.
24. Ye, K.-H., et al., Stability analysis of oxide (CeO2, NiO, Co3O4 and Mn3O4) effect on
Pd/C for methanol oxidation in alkaline medium. Electrochim . Acta, 2013. 90: p. 108-
111.
25. Shen, P.K., et al., Electro-oxidation of Methanol on NiO-Promoted Pt/C and Pd/C
Catalysts. Electrochem. Solid State Lett., 2006. 9: p. A39-A42.
26. Gadgil, M.M., R. Sasikala, and S.K. Kulshreshtha, CO oxidation over Pd/SnO2 catalyst.
J. Mol. Cat., 1994. 87: p. 297-309.
27. Schryer, D.R., et al., Effects of pretreatment conditions on a platinum/tin dioxide catalyst
for the oxidation of carbon monoxide in carbon dioxide lasers. J. Catal., 1990. 122: p.
193-197.
28. Sekizawa, K., et al., Low temperature oxidation of methane over Pd catalyst supported
on metal oxides. Catal. Today, 2000. 59: p. 69-74.
29. Huang, Y., J. Cai, and Y. Guo, Roles of Pb and MnOx in PtPb/MnOx-CNTs catalyst for
methanol electro-oxidation. Int. J. Hydrogen Energy, 2012. 37: p. 1263-1271.
30. Zhang, H., et al., Enhanced electrocatalytic performance for methanol oxidation on
PteTiO2/ITO electrode under UV illumination. Int. J. Hydrogen Energy, 2010. 35: p.
13290-13297.
31. Antolini, E. and E.R. Gonzalez, Ceramic materials as supports for low-temperature fuel
cell catalysts. Solid State Ionics, 2009. 180: p. 746-763.
66
32. Yang, L., W. Yang, and Q. Cai, Well-Dispersed PtAu Nanoparticles Loaded into Anodic
Titania Nanotubes: A High Antipoison and Stable Catalyst System for Methanol
Oxidation in Alkaline Media. J. Phys. Chem. C, 2007. 111: p. 16613-16617.
33. Sasaki, K. and R.R. Adzic, Monolayer-Level Ru- and NbO2-Supported Platinum
Electrocatalysts for Methanol Oxidation. J. Eletrochem. Soc., 2008. 155: p. B180-B186.
34. Konopka, D.A., et al., Platinum Supported on NbRuyOz as Electrocatalyst for Ethanol
Oxidation in Acid and Alkaline Fuel Cells. J. Phys. Chem. C, 2011. 115: p. 3043-3056.
35. Chhina, H., S. Campbell, and O. Kesler, Ex Situ and In Situ Stability of Platinum
Supported on Niobium-Doped Titania for PEMFCs. J. Electrochem. Soc., 2009. 156: p.
B1232-B1237.
36. Jiang, Q., et al., Promoting Effect of Ni in PtNi Bimetallic Electrocatalysts for the
Methanol Oxidation Reaction in Alkaline Media: Experimental and Density Functional
Theory Studies. J. Phys. Chem. C., 2010. 114: p. 19714-19722.
37. Matsumoto, F., Ethanol and Methanol Oxidation Activity of PtPb, PtBi, and PtBi2
Intermetallic Compounds in Alkaline Media. Electrochemistry, 2012. 80: p. 132-138.
38. Jiang, Q., et al., Experimental and density functional theory studies on PtPb/C bimetallic
electrocatalysts for methanol electrooxidation reaction in alkaline media. Electrochim .
Acta, 2011. 56: p. 6431-6440.
39. Li, G. and P.G. Pickup, The promoting effect of Pb on carbon supported Pt and Pt/Ru
catalysts for electro-oxidation of ethanol. Electrochim. Acta, 2006. 52: p. 1033-1037.
40. Zhang, L.J., Z.Y. Wang, and D.G. Xia, Bimetallic PtPb for formic acid electro-oxidation.
J. Alloys Compd., 2006. 426: p. 268-271.
41. Lei, H.-W., H. Hattori, and H. Kita, Electrocatalysis by Pb adatoms of HCOOH
oxidation at Pt(111) in acidic solution. Electrochim. Acta, 1996. 41(10): p. 1619-1628.
67
42. Rivera, E.C., et al., Electrocatalytic Activity of Ordered Intermetallic Phases for Fuel
Cell Applications. J. Am. Chem. Soc., 2004. 126: p. 4043-4049.
43. Gierst, L., J. Tondeur, and E. Nicolas, DOUBLE COUCHE ELECTROCHIMIQUE ET
CINETIQUE DES REACTIONS DELECTRODE - ELUCIDATION DU MODE
DACTION DES CATIONS TETRAALKYLAMMONIUM. Journal of Electroanalytical
Chemistry, 1965. 10(5-6): p. 397-&.
44. Chen, M., et al., Investigation of the Pt–Ni–Pb/C ternary alloy catalysts for methanol
electrooxidation. Electrochem. Commun., 2008. 10: p. 443-446.
45. Xiong, L., et al., Pt–Ni alloy nanoparticles supported on multiwalled carbon nanotubes
for methanol oxidation in alkaline media. J. Solid State Electrochem., 2013. 17: p. 805-
810.
46. Lyons, M.E.G. and M.P. Brandon, A comparative study of the oxygen evolution reaction
on oxidized nickel, cobalt and iron electrodes in base. J. Electroanal. Chem., 2010. 641:
p. 119-130.
47. Jaksic, J.M., et al., Novel spillover interrelating reversible electrocatalysts for oxygen
and hydrogen electrode reactions. J. Phys. Chem. C., 2010. 114: p. 18298-18312.
48. Wang, J., et al., Synthesis and electrochemical study of Pt-based nanoporous materials.
Electrochim . Acta, 2008. 53: p. 6944-6952.
49. Shiller, P. and A.B. Anderson, Effects of chemisorbed and substitutional 0, I, and II Ge,
Sn, and Pb on CO adsorption on Pt( 111) : molecular orbital theory. Surf. Sci., 1990.
236: p. 225-232.
50. Tripkovic, A.V., et al., Kinetic and mechanistic study of methanol oxidation on a Pt(110)
surface in alkaline media. Electrochim . Acta, 1998. 44: p. 1135-1145.
68
Chapter 3
Oxygen Reduction of Nanocrystalline Ruthenia - Local Structure Effects
3.1 Introduction
The fuel cell related electrocatalytic processes based on controlled hydrogen oxidation
and oxygen reduction have recently gained importance mainly in connection with the increasing
utilization of renewable energy sources. Despite efforts devoted to the optimization of existing
systems, the performance of real fuel cells still lags behind the expectations and the cathodic
oxygen reduction is seen as the performance limiting process. Simultaneously, the
electrochemical fuel cell reactions can also be generally employed in the energy storage mode
using the excess electricity or solar energy to generate energetically useful hydrogen (produced
along with the oxygen). This leads to the introduction of the regenerative fuel cell concept[1, 2].
It also needs to be stressed that the regenerative fuel cell applications have sparked extensive
catalyst development primarily for the oxygen evolution/reduction processes.
Oxygen electrochemistry, including oxygen evolution as well as reduction, represents the
simplest example of multiple electron charge transfer processes which have been extensively
studied both experimentally as well as theoretically[3, 4]. In contrast to the development of
suitable catalysts for independent oxygen evolution (OER) or oxygen reduction (ORR), the
catalysts' application for regenerative fuel cells faces significant restrictions in terms of
minimizing the energetic barriers of both kinetically irreversible processes. The fact that the
oxygen evolution reaction proceeds solely on oxide covered surfaces disfavors the use of metal
catalysts which are reported to be of superior activity in oxygen reduction. The oxide activity in
69
the oxygen evolution was traditionally investigated in both acidic as well as alkaline media on
various systems based on oxide of ruthenium[5-8], iridium[5, 7-9], cobalt[5, 10] or
manganese[5]. Oxygen reduction studies on oxides are less frequent and are generally restricted
to alkaline media. Oxygen reduction has been studied on rutile[11, 12], spinel[13, 14],
perovskite[15] and pyrochlore[16, 17] structural types based on ruthenium, manganese, nickel,
cobalt and iridium oxides. The investigated oxide catalysts were the subject of electrochemical
characterization which was phenomenologically analyzed in order to explain the possible
reaction pathways leading to both 4-electron and 2-electron oxygen reduction processes. In
contrast to the studies of oxygen evolution, no detailed investigations aiming at the role of the
catalyst structure, including the local structure of the oxygen reduction active site, have been
reported so far.
The theoretical approach allowing for the generalization of oxygen electrochemistry on
oxides based on DFT calculations was recently reported [4, 18, 19]. The DFT calculations
identify the active sites for oxygen activation and the charge transfer to so-called coordination
unsaturated sites (cus), the surface population of which is a function of the surface orientation.
The cus surface sites feature (n - 1) oxygen bonding partners, where n is equal to number of
metal-oxygen bonds present in the oxide bulk. It is believed that only cus sites can form the atop
reaction intermediate(s), which are essential to the oxygen electrochemistry. In this respect one
can easily predict that the catalytic activity and selectivity of oxide catalysts may be altered if
one controls the population and stacking of the cus sites at the oxide surface. This trend has been
shown for oxygen evolution on heterostatically doped ruthenia when the incorporation of lower
valency cations, such as Ni[20-23], Co[24-27], Fe[28] or Zn[29, 30] into ruthenia framework
70
resulted in changes to both the activity and selectivity of anodic processes including oxygen and
chlorine evolution. Similar systematic studies focused on other oxide systems are, so far, lacking.
This paper focuses on the role of the local structure of the oxide catalysts in the oxygen
reduction reaction. We report on the ORR activity of model nanocrystalline ruthenia catalysts
with local structure controlled by doping with Ni, Co, and Zn. The observed electrocatalytic
activity and selectivity are related to the actual local structures and rationalized using DFT-based
thermodynamic analysis of the oxygen reduction process.
3.2 Experimental
3.2.1 Materials Preparation
Ruthenium dioxide and doped samples of the composition Ru1-xMxO2 (M = Co, Ni, Zn)
were synthesized using the spray-freezing freeze-drying method as described in references[30,
31]. Generally, an 8 mM solution was prepared by dissolving the appropriate amount of
Ru(NO)(NO3)3 (31.3% Ru, Alfa Aesar) in 100 mL of Millipore H2O. In the case of doped
materials, a stoichiometric amount of the appropriate transition metal salt was added to the
solution. Zinc-doped samples were prepared from the acetate precursor, Zn(C2H3O2)2 • 2H2O
(99.5% ACS reagent grade, Fluka). Cobalt- and nickel-doped samples were prepared from the
nitrate salts, Co(NO3)2 • 6H2O and Ni(NO3)2 • 6H2O (99.999% trace metal basis, Sigma Aldrich),
respectively. The starting solution was then sprayed into liquid N2. The resulting ice slurry was
collected in an aluminum tray pre-cooled with liquid N2 and quickly transferred to a freeze-dryer
(FreeZone Triad Freeze Dry System 7400030, Labconco) pre-cooled to -30°C. The frozen
solvent was sublimated at reduced pressure (≈1.0 Pa) while the temperature was ramped
71
according to the following program: -30°C (1h), -25°C (5h), -20°C (4h), -15°C (6h), 30°C (4h).
After drying, the resulting powder was annealed in air at 400°C for 1 hour.
3.2.2 XRD, XPS, and SEM Characterization
The crystallinity of sample powders was characterized using a Rigaku Miniflex 600
powder X-ray diffractometer with CuKα radiation operating at 40 kV and 15 mA. The average
sample compositions were evaluated with X-ray energy dispersive spectroscopy using a Hitachi
S4800 scanning electron microscope (SEM) equipped with a Nanotrace EDX detector (Thermo
Electron). Sample compositions did not deviate significantly from the projected ones. Particle
size was evaluated by analyzing SEM images and averaging the size of 300 randomly chosen
particles. The X-ray photoelectron spectra (XPS) of the prepared materials were measured using
a modified ESCA 3 MkII multitechnique spectrometer equipped with a hemispherical electron
analyzer operating in the fixed transmission mode. Al Kα radiation was used for electron
excitation. The binding energy scale was calibrated using the Au 4f7/2 (84.0 eV) and Cu 2p3/2
(932.6 eV) photoemission lines. The spectra were collected at a detection angle of 45° with
respect to the macroscopic surface normal. The studied materials were characterized using
survey scan spectra and high resolution spectra of overlapping Ru 3d + C 1s photoelectrons, Ru
4s, Zn 2s and O 1s photoelectrons. The spectra were curve fitted after subtraction of Shirley
background using the Gaussian–Lorentzian line shape and nonlinear least-squares algorithms.
Quantification of the elemental concentrations was accomplished by correcting the photoelectron
peak intensities for their cross sections and for the analyzer transmission function. The typical
error of quantitative analysis by XPS is ~10%.
72
3.2.3 Electrochemical Measurements
The electrochemical oxygen reduction activity of the prepared materials was assessed in a
three-electrode single-compartment cell with a rotating ring-disk electrode (RRDE) setup (Pine
Instruments, USA). The potential was controlled using an Autolab PGSTAT30 (EcoChemie, The
Netherlands). Catalyst ink suspensions were prepared by sonicating 9.8 mg RuO2 or Ru1-xMxO2
(M = Co, Ni, Zn) with 5.00 mL Millipore water, 4.95 mL isopropyl alcohol, and 50 μL of 5 wt.%
Nafion® ionomer solution as a binder until the suspension was well dispersed. A 10.0 μL aliquot
of the ink was drop cast on a 0.196 cm2 glassy carbon disk electrode equipped with a platinum
ring to yield a total catalyst loading of approximately 50 μg cm-2
. All experiments were
conducted at room temperature in 0.1 M NaOH prepared from sodium hydroxide pellets
(semiconductor grade, 99.99%, Sigma-Aldrich). A platinum wire served as the counter electrode
and a saturated calomel electrode (SCE) served as the reference electrode. All potentials reported
are quoted against RHE. Electrolyte solutions were saturated with O2 for 30 minutes prior to
oxygen reduction measurements. The measured oxygen reduction currents were corrected for the
contribution of the capacitive current by subtracting the cyclic voltammograms obtained under
identical conditions in Ar saturated solution. Cyclic voltammograms were recorded at a scan rate
of 20 mV s-1
and the potential of the platinum ring electrode was held at 1.1 V vs. RHE during all
measurements. The ring collection efficiency was determined to be 0.275 according to the
procedure described in reference[32].
73
3.2.4 DFT Analysis of Oxygen Reduction
The thermodynamic analysis of the ORR on ruthenia based [110] surfaces was addressed
using GPAW (grid-based projector-augmented wave) DFT based code[33] together with the
ASE (atomic simulation environment)[34]. For all surfaces the exchange correlation functional,
revised Perdew Burke Ernzerhof[35], was used. The grid spacing selected was 0.18 and the
Brillouin zone was sampled using a 4 x 4 x 1 Monkhorst–Pack grid. The two model systems, the
non-doped and Ni-doped ruthenia [110], were approximated using a 2 x 1 and a 3 x 1 supercell,
respectively, with four atomic trilayers and with the bottom two trilayers fixed. The remaining
layers and adsorbates were relaxed until the residual forces in all directions were less than 0.05
eV Å-1
. The positions of the Ni atoms were modeled using the approach described in
reference[33]. The calculations containing Ni were spin-polarized.
3.3 Results and Discussion
3.3.1 XRD and SEM Characterization
X-ray diffraction patterns of all studied materials are shown in Figure 3.1. In all cases the
recorded patterns conform to a single phase tetragonal structure of the rutile type identical with
that of RuO2 (PDF file #431027). The average size of coherent crystallite domains was evaluated
using the Scherrer formula:
(3.1)
74
where Di is the size of the crystallite domain, λ is the wavelength of the incident radiation (CuKα
= 1.540598 Å), βi is the width of the diffraction peak at half maximum intensity measured in
radians, and θi is the angle of the hkl reflection.
Figure 3.8. Powder X-ray diffraction patterns for nanocrystalline RuO2 and Ru1-xMxO2 (M = Ni,
Co, Zn) after annealing in air for 1 hour at 400 °C
The average coherent domain size ranged between 4.3 and 5.7 nm (see Table 3.1).
Representative SEM images of the doped ruthenia are summarized in Figure 3.2. The particle
sizes evaluated from SEM micrographs agree with the coherent domain size values (See Table
3.1). Average sample compositions did not deviate from the projected ones and are listed in
Table 3.1.
The surface composition of all doped samples reflects the metastable character of the
materials and previous thermal treatment which result in a dopant enrichment of the surface
layer.[21, 36] This effect is most pronounced in the case of the Zn doped materials when the
75
actual surface compositions of both studied materials correspond to Ru0.73Zn0.27O2.43 and to
Ru0.63Zn0.37O2.23 for the materials with nominal Zn contents of 0.1 and 0.2, respectively. It needs
to be noted that the apparent excess of the oxygen in the surface composition can attributed to
surface OH groups as well as to adsorbed water.
Table 3.1. Results of the analysis of the powder diffraction data for RuO2 and doped RuO2
samples.
Actual Composition Coherent Domain
Size [nm]
Strain
[%] a [Å] c [Å]
Particle
Size [nm]
RuO2 5.7 0.46 4.470 3.120 7.6 ± 2.2
Ru0.9Zn0.1O2-z 4.9 0.61 4.526 3.108 8.9 ± 2.2
Ru0.82Zn0.18O2-z 5.5 0.00 4.519 3.099 5.8 ± 1.5
Ru0.95Ni0.05O2-z 5.3 0.00 4.515 3.096 7.2 ± 1.4
Ru0.91Ni0.09O2-z 5.0 0.00 4.501 3.079 7.9 ± 2.3
Ru0.90Co0.10O2-z 4.3 0.21 4.505 3.081 7.4 ± 1.8
Figure 3.2. SEM images of nanocrystalline (a) RuO2, (b) Ru0.90Zn0.10O2, (c) Ru0.80Zn0.20O2, (d)
Ru0.95Ni0.05O2, (e) Ru0.90Ni0.10O2, and (f) Ru0.90Co0.10O2 after annealing at 400 °C in air for 1
hour.
76
3.3.2 Electrochemical Measurements
All prepared ruthenia materials are active ORR catalysts in alkaline media. The ORR
polarization curves for RuO2 and Ru1-xMxO2 (M = Ni, Co, Zn) samples are shown in Figure 3.3.
The disk current (iD) which reflects the oxygen reduction shows a pronounced peak at
approximately 0.40 V to 0.55 V before approaching a mass transport controlled region. The disk
current feature can be tentatively associated with a change in the Ru oxidation state from Ru(IV)
to Ru(III)[11, 37]. This process is usually connected with cation insertion into the oxide structure
to balance the charge in cationic and anionic sub-lattices[36]. The behavior giving rise to the
peak in the disk current is also manifested in the ring current, indicating a pronounced formation
of hydrogen peroxide in this potential region. The hydrogen peroxide formation in the 0.40 V to
0.55 V interval seems to be unaffected by the chemical composition of the catalyst. In addition,
the formation of hydrogen provide seems to be suppressed with increasing rotation rate. The
precise mechanism of this reduction process is, however, not evident.
77
Figure 3.9. ORR polarization curves and ring currents at 1600 rpm for RuO2 and Ru1-xMxO2 (M
= Ni, Co, Zn) electrodes at 20 mV s-1
in O2 saturated 0.1 M NaOH. Ering = 1.1 V vs. RHE.
The overall ORR activity of the doped ruthenia catalysts is lower than that of the non-
doped ruthenia. The ORR activity as reflected in the disk currents (iD) generally decreases for the
Co- and Ni-doped samples. The corresponding ring currents (iR) are, however, higher than that of
the non-doped ruthenia, particularly at high overpotentials (i.e. at potentials negative to 0.4 V vs.
78
RHE). This shows a pronounced tendency of Co and Ni-doped materials to produce H2O2
namely at high overpotentials (n ranging between 3.0 and 3.4). In contrast, the Zn-doped
materials show a preference for the 4-electron reduction pathway with n values ranging between
approximately 3.6 to 3.8 while the activity remains comparable to that of the non-doped ruthenia.
Also the selectivity of the doped ruthenia in oxygen reduction is controlled by the doping process
itself rather than by the actual dopant content.
The observed behavior reflects the surface sensitivity of oxygen reduction on oxide
surfaces, which can be related to the surface local structure. Quantitative visualization of this
behavior is shown in Figure 3.4, which plots the potential dependence of the average number of
electrons transferred to an oxygen molecule on different doped ruthenium dioxide materials as
calculated from the Koutecky-Levich equation[38]:
(3.2)
where F is Faraday's constant, A is the geometric area of the electrode, D is the diffusion
coefficient (1.90 × 10-5
cm2 s
-1), ν is the kinematic viscosity (8.70 × 10
-3 cm
2 s
-1), and C is the
bulk concentration of O2 (1.22 × 10-6
mol cm-3
)[39].
79
Figure 3.4. Potential dependence of the average number of electrons transferred during oxygen
reduction on RuO2 and Ru1-xMxO2 (M = Ni, Co, Zn) electrodes. Presented data were calculated
using Koutecky-Levich equation.
It has to be stressed that in contrast to the behavior known for metal electrocatalysts in
acid media, the oxygen reduction on ruthenia based catalysts apparently forms primarily
hydrogen peroxide, namely at low overpotentials. The observed selectivity of ruthenia-based
catalysts in ORR shows a more complex potential dependence which can be treated either by a
phenomenological or a local structure sensitive approach.
Assuming a general phenomenological model of the oxygen reduction mechanism as
proposed previously (See Fig. 3.5)[40, 41], oxygen can be reduced to water (4-electron
pathway) either directly or sequentially with H2O2 as the main adsorbed intermediate.
In principle, H2O2 either desorbs and can be detected on the ring or can be further
reduced to water in the second 2-electron reduction process. The measured disk current
80
summarizes the current contributions from the complete 4-electron reduction to H2O and the 2-
electron reduction to H2O2 while the recorded ring current is proportional only to the amount of
oxygen reduced to H2O2. In this respect a ratio of iD/iR can be used as an indicator of the actual
mechanism which should yield a straight line proportional to k1/k2 when plotted against ω-1/2
(See
Fig. 3.6)[41].
Figure 3.5. Phenomenological mechanism of oxygen reduction according to reference[41].
Figure 3.6. |ID/IR| vs. ω-1/2
plots for (a) RuO2, (b) Ru0.80Zn0.20O2, and (c) Ru0.90Ni0.10O2.
Presented data were extracted from RRDE experiments carried out in O2 saturated 0.1 M NaOH.
The actual iD/iR data deviate from linearity (see Figure 3.6) as can be expected since the
formalism incorporated in the scheme depicted in Fig. 3.5 disregards the nature of the individual
reaction steps composing both 2- and 4-electron reduction pathways and their different
dependence on the electrode potential.
81
The individual rate constants k1, k2, and k3 were evaluated from ORR data assuming that
all three processes proceed simultaneously and that the values of k-1, k-2, and k-3, corresponding
to reversed reactions, are negligible. The adsorption of oxygen on the electrode surface is also
assumed to proceed sufficiently fast. The potential dependence of the rate constants for all
considered catalysts in the overall oxygen reduction mechanism is shown in Figure 3.7.
Figure 3.10. Potential dependence of the rate constants for the reduction of O2 to H2O (k1), of O2
to H2O2 (k2), and H2O2 to O2 (k3) on nanocrystalline ruthenia based catalysts. The presented data
correspond to experiments carried out in O2 saturated 0.1 M NaOH at 1600 rpm.
It seems that the conversion of H2O2 to H2O through the series pathway (k3) is negligible
on all electrode surfaces at high overpotentials. It has to be noted, that although the k3 values are
negligible with respect to k1 and k2 there is a significant difference between k3 of the Zn-doped
and non-doped samples and those obtained for the Ni- and Co-doped samples. The values of k3
observed for Ni- and Co-doped samples are approximately one order of magnitude lower and
82
seem to correspond to decreased tendency of these materials to reduce oxygen through the 4-
electron pathway.
As follows from Figure 3.7, the conversion of O2 to H2O2 appears to be the dominant
process on ruthenium based oxides at low overpotentials. In this respect the reduction behavior
of the ruthenia differs significantly from that of metals which prefer the 4-electron reduction at
low overpotentials. The role of the chemical composition in selectivity of doped catalysts
towards 2- and 4-electron reduction pathways can be visualized by the potential at which the
catalytic system shows the same preference for the 4-electron and 2-electron reaction pathways,
i.e. potential at which k1/k2 = 1 (see Figure 3.8).
A fundamental description of the oxygen reduction on oxide surfaces can be based on the
thermodynamic analysis of the observed trend, which highlights the enhanced tendency of the
Ni- and Co-doped materials to form hydrogen peroxide compared with non-doped and Zn-doped
ruthenia and apparently reflects the local structure of the doped ruthenium oxides.
Figure 3.11. The potential of equal rate in 2- and 4-electron reduction for different ruthenia
based catalysts.
83
3.3.3 DFT Analysis of Oxygen Reduction
A fundamental description of the oxygen reduction on oxide surfaces can be based on the
thermodynamic analysis of the process utilizing the DFT modeling. Reverting to the formalism
used for the oxygen evolution reaction we can describe the overall reduction process as a
sequence of four consecutive concerted electron/proton transfers - if one aims for the complete 4-
electron reduction - or of two consecutive electron/proton transfers if hydrogen peroxide is
considered as the reaction product. The results of the DFT investigations of ORR reduction on
ruthenium dioxide based catalysts are summarized in Figures 3.9-3.12.
A systematic description of the stable surface structures at different potentials represents
a prerequisite step in the theoretical investigation of oxygen reduction on an oxide surface, which
in this case is the [110] rutile surface of ruthenia. This procedure results in computational
Pourbaix diagrams where the stable surface at any given potential features the highest
stabilization (i.e. the most negative surface energy) of the system.
Bearing in mind that the [110] oriented surface of a rutile type oxide features the
transition metal cations in two local environments, cus and bridge, one can visualize the surface
of non-doped ruthenium dioxide as changing from the surface structure characterized by
protonated oxygen on cus sites and deprotonated oxygens in bridge sites (region C in Figure 3.9)
to the surface featuring vacant cus and protonated oxygens in bridge sites (region A in Figure
3.9). Since the ORR was not observed at potentials positive to 0.7 V (vs. RHE) one can restrict
the DFT investigations of the oxygen reduction on conventional RuO2 to the surface stable in the
region (A). In the case of doped ruthenia (as shown in the case of the Ni doped material
presented in Figure 3.10) one needs to consider the complexity arising from the chemical
composition when both types of transition metal cations enter the cus and bridge positions. This
84
variability in the chemical composition also increases the number of distinctive oxygen atoms
available at the surface, the binding energy of which depends on their nearest neighbors.
Although the electrode potential dependent variability of the surface structure is more
pronounced in this case, the stable structures predicted from the potential range in which the
ORR proceeds is qualitatively the same and corresponds to vacant cus sites complemented by
protonated oxygen atoms connecting the bridge sites (see structure D in Figure 3.10).
Figure 3.12. Surface Pourbaix diagram for RuO2. Detailed description of the diagram
construction is given in the supplementary information.
85
Figure 3.13. Surface Pourbaix diagram for Ni-doped RuO2. Detailed description of the
diagram/s construction is given in the supplementary information.
The DFT models predict that the entire process begins with oxygen adsorption at
coordination unsaturated (cus) cationic sites. The behavior of both the non-doped and the doped
ruthenia catalysts is controlled by the local structure and depends on the nature of the cation
residing in the cus site as well as on the electrode potential. In the case that the cus site is
occupied by a ruthenium cation (which are present on all investigated catalysts) the first electron
reduction forms a rather strongly bound *OOH intermediate, which is more stable than the
hydrogen peroxide at most reasonable electrode potentials (see Figure 3.11). Consequently, the
further reduction of the *OOH intermediate located on Ru cus site cannot form hydrogen
peroxide unless one uses a rather strong external electric field to weaken the *OOH binding to
the surface. The actual potential(s) at which hydrogen peroxide formation becomes
thermodynamically allowed are given in the legend of the Figure 3.11.
86
Figure 3.11. Free energy diagrams for the reduction of O2 on three catalytic active sites, the Ru
cus site on conventional ruthenia (green), the cus Ru site on Ni doped RuO2 (magenta) and the
cus Ni site on Ni doped RuO2 (blue). The dotted line represents the equilibrium potential of the
reduction O2 to H2O2. The key difference is the binding of O on the Ni cus site compared to the
Ru cus sites.
In the case that the cus site is occupied by an heteroatom, e.g. Ni or Co, (see Figure 3.11)
one observes a significantly weaker binding of the *OOH and *O compared with the Ru
occupied cus sites. This fact decreases the potential at which the reduction on Nicus starts to
contribute to the overall reduction process. The weak interaction of the *OOH with the
heteroatom-containing cus site restricts the presence of such an adsorbate in the potential region
with low total surface coverage, i.e. to relatively high over-potentials. It needs to be noted
though, that the formation of hydrogen peroxide from *OOH confined on a heteroatom occurs at
much more positive potentials than in the case of *OOH confined to Ru-containing cus sites and
further reduction of the *OOH intermediate can proceed via the 4-electron or 2-electron
reduction pathway with approximately the same probability.
Figure 3.12 shows the dependence of the electrode potential needed to drive the oxygen
reduction on oxide based surfaces either via the 4-electron (red) or 2-electron (blue) reaction
pathway as a function of the reaction descriptor - i.e. adsorption energy of the *OOH
intermediate. It needs to be noted that in a similar manner one may describe the reaction with the
87
adsorption of *OH due to the interdependence of the adsorption energies of the intermediate
formed in the first and third charge transfer step[3, 33].
Figure 3.14. Volcano plot for the 2-electron (blue) and 4-electron (red) reduction of O2 to H2O2
and H2O, respectively, using the binding energy of OOH as a descriptor. The dotted lines
represent the equilibrium potentials for the reduction products. In the case of the Ni-doped
ruthenia the limiting over-potential for both possible reaction sites (Rucus and Nicus) are shown
along with that of conventional ruthenia.
Such a dual volcano plot has been used with great success in literature[42, 43]. The
volcano curves presented in Figure 3.12 clearly show a quantitative prediction of the
thermodynamic preference of the 4-electron reduction pathway over the 2-electron reduction on
strongly adsorbing cus sites. As follows from Figure 3.11, oxygen reduction on Ni-doped
ruthenia should proceed at slightly more positive potentials compared with the non-doped
ruthenia as long as the cus sites are occupied with Ru cations. The reduction process on Ru
occupied cus sites should show a pronounced preference for 4-electron reduction and the
formation of hydrogen peroxide should be excluded for potentials positive of 0.14 V or 0.43 V
(vs. RHE) for non-doped ruthenia and the Ni-doped material, respectively (see Figure 3.11). The
88
easier formation of hydrogen peroxide on the Ni-doped material should be compensated by an
earlier onset of the oxygen reduction process as predicted for Ni-doped material. In the case of
weakly adsorbing sites, e.g. in the case of Ni cus sites - there is no apparent thermodynamic
preference for either the 4- or 2-electron reduction pathway. The DFT model predicts the onset
of the oxygen reduction process to occur at potentials comparable with the ORR on non-doped
ruthenia. The formation of hydrogen peroxide is possible at significantly more positive potentials
(see Figure 3.11)
Analyzing the experimental behavior of the ruthenium dioxide based catalysts for the
oxygen reduction process in the light of the DFT results one can qualify the existence of two
classes of catalysts – one favoring the 4-electron reduction (non-doped RuO2 and Zn-doped
RuO2) and another showing significant activity in hydrogen peroxide production (Ni- and Co-
doped ruthenia). Realizing that the Zn present in the Zn-doped ruthenia is itself redox inactive
one can assume that the catalysts in the first group have all active cus sites occupied by Ru
regardless of the actual chemical composition. The confinement of the catalytic activity on Ru
itself justifies the selectivity towards 4-electron reduction pathway as it is shown in Figure 3.4
and Figure 3.8. In the case of the Co- and Ni-doped ruthenia the significant amount of hydrogen
peroxide formed in the process can be attributed primarily to the Ni/Co cus sites although the Ru
cus sites also contribute to the hydrogen peroxide formation at lower potentials. In contrast to the
complementary oxygen evolution process, the Ni (or Co) ions located in the bridge sites, which
play a crucial role in the complementary anodic process[33], apparently have no effect on the
oxygen reduction activity of these materials. A different role of the catalysts local structure in
oxygen reduction is not entirely surprising given the irreversibility of oxygen
evolution/reduction.
89
The DFT calculations, however, fail to explain pronounced formation of the hydrogen
peroxide on all ruthenium based catalysts at low overpotentials (0.55-0.40 V) when the hydrogen
peroxide on Ru cus sites should be thermodynamically excluded. Given the relatively short
timescale of the RRDE experiments one may therefore suggest that the system fails to reach the
thermodynamically stable surface structure on the experimental timescale and the hydrogen
peroxide is released from meta-stable intermediates not reflected in the DFT calculations.
3.4 Conclusions
Nanocrystalline ruthenia based electrocatalysts offer a convenient model for investigating
the role of the local structure in the oxygen reduction on oxide electrodes. The oxygen reduction
related activity of RuO2 is comparable with that of the doped ruthenia . The selectivity of doped
ruthenia catalysts differs from that of the RuO2 in which the non-doped as well as Zn-doped
catalysts prefer 4-electron oxygen reduction while the Ni- and Co-doped ruthenia produce
significant amount of hydrogen peroxide. The observed selectivity trends can be rationalized
using a thermodynamic analysis of the oxygen reduction process based on DFT calculations.
The DFT based analysis confines the oxygen reduction activity to cus sites, the
occupancy of which controls the selectivity of the oxygen reduction process. Oxygen reduction
on non-doped ruthenium dioxide is controlled by the fourth electron transfer. Doping the
ruthenium dioxide shifts the potential control to the first electron transfer. This trend can be
attributed to decreasing occupancy of the cus sites with ruthenium. The strong adsorption of the
*OOH intermediate on the Ru cus site steers the reaction mechanism towards 4-electron
reduction pathway. Incorporation of reactive transition metal cations into bridge sites has a
90
negligible effect on the ORR activity. A confinement of the reactive transition metal into cus
sites weakens the adsorption of the reaction intermediates and opens the 2-electron reaction
pathway at relatively low overpotentials.
3.5 Acknowledgments
This work was supported by the European Commission within the Initial Training
Network ELCAT (Project No. 214936). The support of the Danish Ministry of Science,
Technology and Innovation though the CASE is also gratefully acknowledged. Niels Bendtsen
Halck and Jan Rossmeisl of Technical University of Denmark are gratefully acknowledged for
performing all DFT calculations and for their assistance in the subsequent analysis. The authors
would like to thank Valery Petrykin of the J. Heyrovský Institute of Physical Chemistry for his
significant contributions to the EXAFS-based analysis of the doped ruthenia and Zdeněk Bastl,
also from the J. Heyrovský Institute of Physical Chemistry, for his assistance with the XPS
measurements.
91
3.6 References
1. Chen, G., S.R. Bare, and T.E. Mallouka, Development of Supported Bifunctional
Electrocatalysts for Unitized Regenerative Fuel Cells. J. Electrochem. Soc., 2002(149):
p. A1092-A1099.
2. Zhang, Y., et al., Deposited RuO2–IrO2/Pt electrocatalyst for the regenerative fuel cell.
Int. J. Hydrogen Energy, 2007. 32(400-404).
3. Koper, M.T.M., Thermodynamic theory of multi-electron transfer reactions: Implications
for electrocatalysis. J. Electroanal. Chem., 2011(660): p. 254-260.
4. Man, I.C., et al., Universality in oxygen evolution electrocatalysis on oxide surfaces.
Chem. Cat. Chem., 2011. 3: p. 1159-1165.
5. Trasatti, S., Electrocatalysis in the anodic evolution of oxygen and chlorine. Electrochim
. Acta, 1984. 29: p. 1503-1512.
6. Wohlfahrt-Mehrens, M. and J. Heitbaum, Oxygen evolution on Ru and RuO2 electrodes
studied using isotope labeling and on-line mass spectrometry. J. Electroanal. Chem.,
1987. 237: p. 251-260.
7. Lyons, M.E.G. and S. Floquet, Mechanism of oxygen reactions at porous oxide
electrodes. Part 2—Oxygen evolution at RuO2, IrO2 and IrxRu1-xO2 electrodes in aqueous
acid and alkaline solution. Phys. Chem. Chem. Phys., 2011. 13: p. 5314-5335.
8. Song, S., et al., Electrochemical investigation of electrocatalysts for the oxygen evolution
reaction in PEM water electrolyzers. Int. J. Hydrogen Energy, 2008. 33: p. 4955-4961.
9. Fierro, S., et al., Investigation of the oxygen evolution reaction on Ti/IrO2 electrodes
using isotope labelling and on-line mass spectrometry. Electrochem. Commun., 2007. 9:
p. 1969-1974.
10. Esswein, A.J., et al., Size-Dependent Activity of Co3O4 Nanoparticle Anodes for Alkaline
Water Electrolysis. J. Phys. Chem. C, 2009. 113: p. 15068-15072.
11. Chang, C.C. and T.C. Wen, Kinetics of oxygen reduction at RuO2-coated titanium
electrode in alkaline solution. J. Appl. Electrochem., 1997. 27: p. 355-363.
92
12. Chang, C.-C. and T.-C. Wen, Kinetics of Oxygen Reduction at IrO2-Coated Titanium
Electrode in Alkaline Solution. J. Electrochem. Soc., 1996. 143: p. 1485-1491.
13. Cheng, F., et al., Rapid room-temperature synthesis of nanocrystalline spinels as oxygen
reduction and evolution electrocatalysts. Nature Chemistry, 2011. 3: p. 79-84.
14. Hamdani, M., R.N. Singh, and P. Chartier, Co3O4 and Co-based Spinel Oxides for
Bifunctional Oxygen Electrodes. Int. J. Electrochem. Sci., 2010. 5: p. 556-577.
15. Suntivich, J., et al., Design principles for oxygen-reduction activity on perovskite oxide
catalysts for fuel cells and metal–air batteries. Nat Chem, 2011. 3(7): p. 546-550.
16. Horowitz, H.S., J.M. Longo, and H.H. Horowitz, Oxygen electrocatalysis on some oxide
pyrochlores. J. Electrochem. Soc., 1983. 130: p. 1851-1859.
17. Egdell, R.G., et al., Electrochemistry of ruthenates: Part 1 - Oxygen reduction on
pyrochlore ruthenates. J. Chem. Soc., Faraday Trans. 1, 1983. 79: p. 893-912.
18. Garcia-Mota, M., et al., Tailoring the Activity for Oxygen Evolution Electrocatalysis on
Rutile TiO2(110) by Transition-Metal Substitution. Chem. Cat. Chem., 2011. 3: p. 1607-
1611.
19. Rossmeisl, J., et al., Electrolysis of water on oxide surfaces. J. Electroanal. Chem., 2007.
607: p. 83-89.
20. Macounova, K., et al., Parallel oxygen and chlorine evolution on Ru1−xNixO2−y
nanostructured electrodes. Electrochim. Acta, 2008. 53: p. 6126-6134.
21. Petrykin, V., et al., Local structure of nanocrystalline Ru1-xNixO2-δ dioxide and its
implications for electrocatalytic behavior - an XPS and XAS study. J. Phys. Chem. C,
2009. 113: p. 21657-21666.
22. Krstajic, N. and S. Trasatti, Cathodic Behavior of RuO2-Doped Ni/Co3O4 Electrodes in
Alkaline Solutions: Surface Characterization. J. Electrochem. Soc., 1995. 142: p. 2675-
2681.
93
23. Krstajic, N. and S. Trasatti, Cathodic behaviour of RuO2-doped Ni/Co3O4 electrodes in
alkaline solutions: hydrogen evolution. J. Appl. Electrochem., 1998. 28: p. 1291-1297.
24. Silva, L.M.D., J.F.C. Boodts, and L.A. DeFaria, ‘In situ’ and ‘ex situ’ characterization of
the surface properties of the RuO2 (x) + Co3O4 (1-x) system. Electrochim. Acta, 2000.
45: p. 2719-2727.
25. Silva, L.M.D., J.F.C. Boodts, and L.A.D. Faria, Oxygen evolution at RuO2(x)+Co3O4(1-
x) electrodes from acid solution. Electrochim. Acta, 2001. 46: p. 1369-1375.
26. Silva, L.M.D., L.A.D. Faria, and J.F.C. Boodts, Electrochemical impedance
spectroscopic (EIS) investigation of the deactivation mechanism, surface and
electrocatalytic properties of Ti/RuO2(x)+Co3O4(1-x) electrodes. J. Electroanal. Chem.,
2002. 532: p. 141-150.
27. Petrykin, V., et al., Local structure of Co doped RuO2 nanocrystalline electrocatalytic
materials for chlorine and oxygen evolution. Catal. Today, 2012. 202(63-69).
28. Macounová, K., et al., Influence of Oxygen on Reactivity of Ru1−xFexO2−y-Doped
Materials. J. Electrochem. Soc., 2008. 11: p. F27-F29.
29. Petrykin, V., et al., Zn-doped RuO2 electrocatalysts for selective oxygen evolution:
relationship between local structure and electrocatalytic behavior in chloride containing
media. Chem. Mater. , 2011. 23: p. 200-207.
30. Petrykin, V., et al., Tailoring the selectivity for electrocatalytic oxygen evolution on
ruthenium oxides by zinc substitution. Angew. Chem., 2010. 49: p. 4813-4815.
31. Tretyakov, Y.D. and O.A. Shlyakhtin, Recent progress in cryochemical synthesis of oxide
materials. J. Mater. Chem., 1999. 9: p. 19-24.
32. Paulus, U.A., et al., Oxygen reduction on a high-surface are Pt/Vulcan carbon catalyst: a
thin-film rotating ring-disk electrode study. J. Electroanal. Chem., 2001. 495: p. 134-145.
33. Halck, N.B., et al., Beyond the volcano limitations in electrocatalysis - oxygen evolution
reaction. Phys. Chem. Chem. Phys., 2014. 16: p. 13682.
94
34. Bahn, S.R. and K.W. Jacobsen, An object-oriented scripting interface to a legacy
electronic structure code. Comput. Sci. Eng., 2002. 4: p. 56-66.
35. Hammer, B., L.B. Hansen, and J.K. Norskov, Improved adsorption energetics within
density-functional theory using revised Perdew-Burke-Ernzerhof functionals. Phys. Rev.
B: Condens. Matter, 1999. 59: p. 7413-7421.
36. Macounová, K., et al., Electrochemical Behavior of Nanocrystalline Ru0.8Me0.2O2 − x
( Me = Fe , Co , Ni ) Oxide Electrodes in Double-Layer Region. J. Electrochem. Soc.,
2007. 154(A1077-A1082).
37. Pourbaix, M., Atlas of Electrochemical Equilibria in Aqueous Solutions. 1966, Oxford,
U.K.: Pergamon Press.
38. Bard, A.J. and L.R. Faulkner, Electrochemical Methods: Fundamentals and Applications.
2nd Edition ed. 2001, New York: John Wiley & Sons, Inc.
39. Markovic, N.M., H.A. Gasteiger, and P.N. Ross, Oxygen Reduction on Platinum Low-
Index Single-Crystal Surfaces in Alkaline Solution - Rotating Ring Disk Studies. J. Phys.
Chem., 1996. 100: p. 6715-6721.
40. Damjanovic, A., M.A. Genshaw, and J.O.M. Bockris, Distinction between intermediates
produced in main and side electrodic reactions. J. Phys. Chem., 1966. 45: p. 4057-4059.
41. Hsueh, K.L., D.T. Chin, and S. Srinivasan, Electrode kinetics of oxygen reduction. A
theoretical and experimental analysis of the rotating ring-disk electrode method. J.
Electroanal. Chem., 1983. 153: p. 79-95.
42. Siahrostami, S., et al., Enabling direct H2O2 production through rational electrocatalyst
design. Nat. Mater., 2013. 12: p. 1137-1143.
43. Viswanathan, V., et al., Unifying the 2e– and 4e– Reduction of Oxygen on Metal
Surfaces. The Journal of Physical Chemistry Letters, 2012. 3(20): p. 2948-2951.
95
Chapter 4
Selective Chlorine Evolution Catalysts Based on Mg-Doped Nanoparticulate Ruthenium
Dioxide
4.1 Introduction
Understanding the electrocatalytic behavior of transition metal oxides for the oxygen
evolution (OER) and chlorine evolution reactions (CER) is fundamental to the chlor-alkali
industry as well as water electrolysis. Iridium and ruthenium based oxides have long been the
center of attention in this respect since they represent the industrial benchmark catalysts due to
their exceptional activity and stability for these reactions[1-3]. Although the traditional approach
in the development of oxide electrocatalysts aims primarily at the optimization of the oxide’s
activity, the issue of systematic selectivity control receives more attention in connection with the
rational design of oxygen evolution catalysts and with the rational control of the electrocatalytic
processes. Both anodic gas evolution processes are among the least thermodynamically favored
redox reactions as shown by the rather high standard electrode potential of each reaction:[4]
O2 + 4H+ + 4e
− ⇌ 2H2O E° = 1.229V (4.1.1)
Cl2(g) + 2e− ⇌ 2Cl
− E° = 1.358V (4.1.2)
Accepting the fact that the industrially relevant production of each gas requires a non-zero
overpotential, one has to anticipate a simultaneous production of oxygen and chlorine at most
potentials, making chloride contamination one of the more serious issues in water electrolysis.
The activity of ruthenium dioxide in both gas evolving reactions was systematically
investigated by means of the density functional theory (DFT) based thermodynamic analysis.
96
DFT analysis has been able to offer some insight regarding the competitive nature of both
processes, confirming that the OER and CER are strongly correlated on rutile-type oxides[5].
According to the DFT analyses, the oxygen evolution reaction proceeds primarily through the
formation of surface confined hydroxo, oxo, or peroxo intermediate species on pentacoordinated
Ru sites, also denoted as coordination unsaturated sites (cus)[6, 7]. Similarly, chlorine evolution
on ruthenium dioxide is most likely to proceed via the formation of peroxo species bridging two
adjacent cus sites (O2,cus) which are subsequently subject to oxidative attack by Cl−[5]. This
theoretical description was recently challenged by Over et. al which assumes that cus confined
oxo species are the most stable surface structure at the given conditions[8].
The theoretical analysis also shows that the activity in chlorine evolution can be
described with the same reaction descriptor as the oxygen evolution process, which in this case is
the oxygen adsorption energy[5]. Despite the less favorable standard potentials, the theory
confirms that the CER process is more facile than oxygen evolution and that the theoretical
limiting overpotential required to drive CER is lower than that of OER in the whole range of
oxygen binding energies[6]. This fact should lead to a complete suppression of the oxygen
evolution activity on oxide surfaces even at moderate chlorine concentrations.
Keeping in mind the interdependence of the oxygen and chlorine evolution processes one
may anticipate that the catalyst selectivity may be affected by controlling the oxygen binding
energy to the oxide surface in a manner similar to that employed for controlling the oxygen
evolution activity. Cationic substitution turns out to be a versatile approach in controlling the
oxygen binding energy and has been demonstrated in the case of ruthenium dioxide modified
with Ni[9-12], Co[11-17], Zn[18], Sn[19-21], and Fe[22]. Experiments combining voltammetric
methods with differential electrochemical mass spectrometry (DEMS) have shown that the
97
incorporation of the heterovalent cations generally affects the selectivity of the ruthenia-based
catalyst in oxygen and chlorine evolution[9, 23, 24]. A general description of the relationships
between the nature of the modified ruthenia-based catalyst and its selectivity is so far missing.
The issue of the oxide catalyst selectivity is addressed in this paper which also extends
the previous studies on the selectivity of rutile-type oxides. In the current work we describe the
synthesis of Mg-modified ruthenium dioxide containing various amounts of Mg and match their
activity and selectivity in parallel oxygen and chlorine evolution with detailed structural
correlations using extended X-ray absorption fine structure (EXAFS).
4.2 Experimental
4.2.1 Materials Preparation
Ruthenium dioxide and Ru1-xMgxO2 materials were synthesized using the spray-freezing
freeze-drying method as outlined in reference [25]. Aqueous solutions (8 mM) were prepared
from ruthenium (III) nitrosylnitrate (31.3% Ru, Alfa Aesar) and magnesium acetate tetrahydrate
(Puratronic®, 99.997% metals basis, Alfa Aesar) in 100 mL of Millipore H2O. The solutions
were then sprayed into liquid N2 to create fine ice particles. The resulting ice slurry was
transferred to an aluminum tray precooled with liquid N2 and placed in the freeze dryer
(FreeZone Triad Freeze Dry System 7400030, Labconco) precooled to −30 °C. The pressure was
decreased to approximately 1.0 Pa and the temperature was ramped according to the following
program: −30 °C (2h), −25 °C (5h) −20 °C (6h), −15 °C (5h), 30 °C (4h). Afterwards, the
resulting powder was annealed in the furnace at 400 °C for one hour.
98
4.2.2 XRD, XPS, and SEM Characterization
Crystallinity and phase purity of the prepared materials was checked using a Rigaku
Miniflex 600 powder X-ray diffractometer with CuKα radiation. Morphology of the synthesized
catalysts was characterized using a Hitachi S4800 scanning electron microscope (SEM) and a
Nanotrace EDX detector (Thermo Electron) was used to evaluate the average sample
compositions by X-ray energy dispersive spectroscopy. Particle size was evaluated by analyzing
SEM images and averaging the size of 300 randomly chosen particles.
The X-ray photoelectron spectra (XPS) of the prepared materials and of the electrodes
which were used in electrochemical experiments were measured using a modified ESCA 3 MkII
multitechnique spectrometer equipped with a hemispherical electron analyzer operating in the
fixed transmission mode. Al Kα radiation was used for electron excitation. The binding energy
scale was calibrated using the Au 4f7/2 (84.0 eV) and Cu 2p3/2 (932.6 eV) photoemission lines.
The spectra were collected at a detection angle of 45° with respect to the macroscopic surface
normal. The studied materials were characterized using survey scan spectra and high resolution
spectra of overlapping Ru 3d + C1s photoelectrons, Ru 4s, Mg 2s, O 1s photoelectrons and Mg
KL23L23 Auger electrons. The spectra were curve fitted after subtraction of Shirley background
using the Gaussian−Lorentzian line shape and nonlinear least-squares algorithms. Quantification
of the elemental concentrations was accomplished by correcting the photoelectron peak
intensities for their cross sections and for the analyzer transmission function. The typical error of
quantitative analysis by XPS is ∼10%.[26]
99
4.2.3 Local Structure Characterization
EXAFS spectra were measured at the Photon Factory synchrotron of the High Energy
Accelerator Organization (KEK) in Tsukuba, Japan. The spectra were measured in transmission
mode at the Ru K edge at beam-line AR-NW10A (Si(311) monochromator); the Mg K
absorption edge was measured in total electron yield mode at the BL-11A beam-line (grazing
incidence monochromator). Ru K scans extended to 20 Å−1
and Mg K data were limited to 15
Å−1
. Each spectrum was recorded at four different scanning step sizes: the preedge region from
500 to 50 eV was scanned in 6.5 eV steps to enable background subtraction; in the 50 eV pre-
edge and 100 eV post-edge range a step size of 0.4–0.5 eV was used to acquire the XANES part
of the spectra, while a 2.5–3.0 eV and 7.0 eV scanning step was maintained in the post-edge
regions of 100–500 eV and above 500 eV, respectively.
All data handling pre-requisite to the local structure refinement of the extended X-ray
absorption fine structure (EXAFS) functions (i.e., normalization, smoothing and background
subtraction, the Fourier transforms of the spectra and windowing) was done in the IFEFFIT
software package.[27] The photoelectron wave vector k for the Fourier transform of spectra was
kept within the range of k = 3–14Å−1
for Ru-EXAFS and k = 3–12 Å−1
for Mg-EXAFS. The k-
weighting factor of 2 was applied. For the analysis of the local structure of Ru1-xMgxO2 materials
a full-profile refinement of the EXAFS spectra by non-linear least squares (NLLS) minimization
in the R-space with a k-weighting factor equal to 2 was carried out using the Artemis NLLS
module of the IFEFFIT package. The theoretical model was generated using FEFF6.2 library.
100
4.2.4 Electrochemical and DEMS Measurements
The electrocatalytic activity and selectivity in the oxygen evolution and chlorine
evolution reactions were evaluated using a combination of potentiostatic experiments with
differential electrochemical mass spectrometry (DEMS).Working electrodes were prepared from
water and isopropanol based catalyst suspensions by sedimentation on Ti mesh (Goodfellow,
20% open area). Catalyst suspensions were pre pared by mixing 10.0 mg RuO2 or Ru1-xMgxO2
with 2.0 mL millipore H2O and 2.0 mL isopropyl alcohol with subsequent homogenization in an
ultrasonic bath and drop cast onto the Ti mesh current collector. The electrodes were dried in air
at 60 °C repeatedly until the desired catalyst loading (1–2 mg cm−2
) was obtained. Electrodes
were annealed at 400 °C in air for 2 hours to stabilize the catalyst layer. The measured current
values were normalized with respect to the actual surface area based on the known mass of the
catalyst and particle size based specific surface area.
Electrochemical measurements were performed in a three electrode arrangement in a
home-made Kel-F single compartment cell using a Pt auxiliary electrode and sat. Ag/AgCl
reference electrode in 0.1 M HClO4. The reference electrode itself was placed outside the cell to
avoid chloride leakage; the conductive connection of the reference electrode was achieved by a
Luggin capillary. The potential of the working electrode in all experiments was controlled using
a PAR 263A potentiostat. The DEMS apparatus consisted of a PrismaTM QMS200 quadrupole
mass spectrometer (Balzers) connected to a TSU071E turbomolecular drag pumping station.
101
4.3 Results and Discussion
4.3.1 XRD, XPS, and SEM Characterization
X-ray diffraction patterns for the synthesized materials are shown in Figure 4.1. The
recorded patterns conform to a single phase tetragonal structure of the rutile-type as in RuO2
(PDF file #431027). The incorporation of Mg into the rutile structure results in a change of the
unit cell parameters which can be evaluated from the diffraction patterns. Assuming the
oxidation state of Mg in the prepared material to be II, its incorporation into the RuO2 framework
should results in a slight expansion of the unit cell volume in the Mg-doped ruthenia
(anticipating Mg to maintain octahedral coordination) due to the slightly larger ionic radius of
the divalent magnesium (0.72 Å) with respect to that of the tetravalent ruthenium (0.62 Å)[26].
This trend, however, is not experimentally encountered (see Table 4.1). Incorporation of
magnesium leads to an expansion of the lattice parameter a (which increases monotonously with
increasing Mg content) and is accompanied by a significant shortening of the lattice parameter c,
which further decreases with increasing Mg content. Although the unit cell volume of Mg-
modified materials increases with increasing Mg content, its concentration dependent increase
does not compensate for the initial decrease in unit cell volume associated with the incorporation
of Mg. Similar unit cell shortening in the [001] direction has been previously reported for Ni-
doped ruthenia and can be taken as an indicator that Mg is not evenly distributed in the resulting
material[10]. The observed trend in the unit cell volume may also be interpreted in terms of the
Mg local environment, specifically that of the oxygen coordination number with increasing Mg
content. Such an interpretation should be, however, supported by local structure data based, e.g.
on X-ray absorption spectroscopy. It needs to be noted that this change should be most
pronounced in materials with a low magnesium content (x < 0.1). The single phase nature of the
102
recorded XRD patterns suggests that the actual chemical composition agrees with the projected
composition. The magnesium seems to be homogeneously distributed between bulk and surface
of the nanoparticles as evidenced by the XPS-based surface composition and projected average
composition (see Figure 4.2 and Table 4.2). This agreement is, however, weakened by the fact
that the Mg XPS data are not available for samples with a Mg content lower than 0.2.
Figure 4.1. Powder X-ray diffraction patterns for nanocrystalline RuO2 and Ru1-xMxO2 after
annealing in air for 1 hour at 400 °C.
Table 4.1. Results of the analysis of the powder diffraction data of the MgxRu1-xO2 catalysts
Sample a [Å] c [Å] V [Å3]
Coherent domain
size [nm]
Particle size
[nm] Strain (%)
RuO2 4.454(15) 3.21(8) 63.68 4.3 9.6 ± 3.0 1.0(3)
Ru0.95Mg0.05O2 4.478(13) 3.12(2) 62.56
5.9 5.3 ± 1.4 0.7(4)
Ru0.90Mg0.10O2 4.505(3) 3.105(4) 63.02
4.6 5.0 ± 1.3 1.0(5)
Ru0.80Mg0.20O2 4.509(3) 3.108(3) 63.19
12.0 5.5 ± 1.8 0.3(2)
103
Figure 4.2. Survey scans (a) of Ru0.8Mg0.2O2 before (1) and after (2) electrochemical
experiments. Panes (b) and (c) contain fitted high resolution spectra of Ru 3d + C 1s (b) and Ru
4s + Mg 2s photoelectrons, respectively. The high resolution spectra of the Ru0.8Mg0.2O2 after
electrochemical experiments do not show significant differences from those plotted in panes (b)
and (c).
Table 4.2. XPS based surface metal content in Ru0.8Mg0.2O2.
Sample XRu XMg
As synthesized 0.80 0.20
After Electrochemistry 0.79 0.21
104
The relatively broad diffraction peaks are compatible with the anticipated nanocrystalline
character of the materials. The average coherent domain size of the prepared materials can be
approximated using the Scherrer formula:
(4.3.1)
where Di is the size of the crystallite domain, λ is the wavelength of the incident radiation (CuKα
= 1.540598 Å), βi is the width of the diffraction peak at half maximum intensity measured in
radians, and θi is the diffraction angle. The diffraction peak analysis yields an average coherent
domain size ranging between 4 and 12 nm (see Table 4.1). The nanocrystalline nature of the
prepared materials is also confirmed in the SEM micrographs (see Fig. 4.3). Regardless of the
actual composition, the Mg-doped ruthenia forms isometric crystals with an average size of
approximately 5 nm (see Table 4.1) complemented with larger particles of approximately 40 nm.
The coarse 40 nm particles are, however, significantly less represented. Despite the actual
underrepresentation of the coarse nanoparticles, they form the principal diffracting regions
controlling the shape of the diffraction peaks, particularly in the material containing 20% Mg.
This fact results in a significant discrepancy between the coherent domain size and the actual
particle size such that the coherent domain size exceeds the particle size based on analysis of the
micrographs.
105
Figure 4.3. SEM images of nanocrystalline (a) RuO2, (b) Ru0.95Mg0.05O2, (c) Ru0.90Mg0.10O2, and
(d) Ru0.80Mg0.20O2 after annealing in air for 1 hour at 400 °C.
4.3.2 Local Structure Characterization
Information complementary to the diffraction is extractable from the X-ray absorption
spectra measured on the Ru and Mg K edges (See Figure 4.4). The experimentally observed Ru
K edge position of the Mg doped ruthenia is not sensitive to the actual Mg content and is equal to
22133 eV. This value agrees well with that reported for ruthenium dioxide in the literature[9,
13]. The corresponding EXAFS functions as they are presented in Figure 4.4a are also
insensitive to the magnesium content and agree qualitatively with the Ru EXAFS functions
106
reported for doped ruthenia previously[10]. The observed behavior is generally compatible with
the non-homogeneous distribution of magnesium in the RuO2 matrix, resulting in the formation
of Mg enriched clusters, the nature of which may depend on the overall Mg content. The
magnesium K edge position in the X-ray absorption spectra of the Mg doped ruthenia shifts with
increasing Mg content from ca. 1310 eV (material containing 5% Mg) to 1308 eV (material
containing 20% Mg). The Mg content-dependent shift of the Mg K edge position may be
interpreted either in terms of a change in oxidation state or in terms of a change in local
structure. The observed Mg absorption edge values, primarily those for the materials with low
total Mg content, exceed that observed for the MgO standard (1309 eV) and could be formally
assigned to an increase in the oxidation state. Keeping in mind the alkaline earth nature of
magnesium, however, the assignment of the observed trend in the position of the Mg absorption
edge energy to the overall effects of bonding arrangements in the vicinity of Mg seems to be
more likely. The Mg EXAFS functions obtained from the Mg K edge X-ray absorption spectra,
in contrast to the Ru EXAFS functions, are affected by the actual Mg content (see Figure 4.4b).
Figure 4.4. EXAFS functions extracted from the X-ray absorption spectra of the Ru1-xMgxO2
(0<x<0.2) measured on the Ru K edge (a) and Mg K edge (b). Actual Mg content is shown in the
Figure legend.
107
The Ru EXAFS functions can be refined assuming a structural model based on ruthenium
dioxide with a rutile structure. A typical example of the refinement for the Ru EXAFS functions
is plotted in Figure 4.5a; the results of the EXAFS function refinement are summarized in Table
4.3a. It may be concluded that the bonding arrangements in the ruthenium environment are
essentially unperturbed by the incorporation of Mg. The Ru-O as well as the Ru-Me (Me = Ru or
Mg) bonding distances and the Ru coordination arrangements do not differ significantly from
those of pure ruthenium dioxide. The presence of Mg in the direct neighbor metal positions with
respect to the absorbing ruthenium is lower than that corresponding to the average chemical
composition.
Figure 4.5. (a) A typical example of the non-linear least square fit of the Ru EXAFS function of
Ru0.95Mg0.05O2; (b) A typical example of the non-linear least square fit of the Ru EXAFS
function of Ru0.90Mg0.10O2. The square symbols represent the experimental data, the red line
denotes the best fit.
108
Table 4.3. (a) Results of the NLLS fit of the EXAFS functions obtained from the Ru Kedge X-
ray absorption spectra of the MgxRu1-xO2 (0 < x < 0.2) catalysts. CN denotes the coordination
number and d stands for the bonding length; (b) Theoretically conceived local structures of the
Mg in rutile and ilmenite type MgRuO oxides and results of the NLLS fit of the EXAFS
functions obtained from the Mg K edge X-ray absorption spectra of the MgxRu1-xO2 (0<x<0.2)
catalysts. The theoretically conceived data are given in italics. Symbol assignment are the same
as in the case of Table 3a.
(a)
(b)
Sample Bond CN Error d [Å] Error Sample Bond CN Error d [Å] Error
Ru0.95Mg0.05O2 Ru-O 2.0 0.2 1.937 0.006 Rutile Mg-O 2 1.94
4.0 0.4 1.982 0.006 4 1.98
Ru-Ru 1.8 0.4 3.100 0.010 Mg-Ru 2 3.14
Ru-O 4.0 0.4 3.410 0.007 Mg-O 4 3.41
Ru-Ru 7.1 1.5 3.552 0.006 Mg-Ru 8 3.55
Ru-Mg 0.8 0.9 3.552 0.006 Ilmenite Mg-O 3 1.95
Ru0.90Mg0.10O2 Ru-O 2.0 0.2 1.938 0.007 3 2.17
4.0 0.3 1.982 0.007 Mg-Ru 1 3.01
Ru-Ru 2.1 0.6 3.101 0.009 3 3.05
Ru-O 4.0 0.3 3.408 0.007 Mg-O
3,40
3 3.40
Ru-Ru 6.3 2.0 3.556 0.008
Mg-Ru 6 3.36
Ru-Mg 1.7 1.5 3.556 0.008 Ru0.90Mg0.10O2 Mg-O 6.1 0.5 2.07 0.02
Ru0.80Mg0.20O2 Ru-O 2.0 0.3 1.937 0.008 Mg-Ru 2.9 0.9 3.08 0.03
4.0 0.6 1.981 0.008 Mg-Mg 4.0 1.4 3.38 0.03
Ru-Ru 2.2 0.8 3.101 0.010 Mg-O 5.1 1.2 3.315 0.08
Ru-O 4.0 0.6 3.407 0.008 Mg-Ru 4.0 0.5 3.55 0.03
Ru-Ru 6.5 2.1 3.557 0.009 Ru0.80Mg0.20O2 Mg-O 6.5 1.2 2.00 0.05
Ru-Mg 1.5 1.5 3.667 0.009 Mg-Ru 2.2 0.7 2.88 0.06
Mg-Mg 1.8 0.5 3.07 0.04
Mg-O 4.0 1.0 3.52 0.07
Mg-Ru 1.5 0.8 3.37 0.04
-
M
g
M
g
-
M
g
M
g
-
M
g
M
m
m
m
m
2.0 0.8 3.56 0,04
The refinement of the Mg EXAFS functions is only possible for the materials containing
at least 10% Mg in the cationic positions. The quality of the XAS spectra for materials with
lower Mg content was not sufficient for the full refinement. The refinement of the Mg local
structure using a rutile-type structural model does not give convergent results as can be seen by
109
the absence of the characteristic cation-cation interactions corresponding to bonding distances of
ca. 3.1 and 3.5 Å. It ought to be stressed that resolved scattering due to the cation-cation
interactions with a bonding distance in the range of 3 to 4 Å is characteristic for most of the
binary and ternary oxide structural types and its absence suggests a rather complex and
disordered nature of the Mg local environment. The actual Mg-O bonding distances in the first
coordination shell increase with increasing Mg content from ca. 2.00 (x = 0.1) to 2.07 Å (x =
0.2). This trend reflects a gradual relaxation of the Mg environment from the bonding
arrangements native to rutile framework of RuO2. The trend in the Mg-O bonding distance
reflects the general incompatibility of Mg with the RuO2 environmental confinement which
brings substantial strain to the material and has to be stabilized by the small size of the Mg
cluster as well as the overall particle size. The Mg-O bond length observed in materials with low
overall Mg content is comparable with that of the Ru-O bond in the ruthenium dioxide. It is,
however, also approximately 0.1 Å shorter that the Mg-O bonding distance in the
thermodynamically stable cubic magnesium oxide (2.12 Å). The refined Mg-O bonding distance
trend seems to be in accordance with the observed shift in the Mg absorption edge toward higher
energies. It is, however, unrealistic to assign the observed behavior to the removal of a third
electron from Mg and we attribute the experimental behavior rather to the strain imposed by the
adjacent Ru-rich rutile-type matrix.
A full refinement of the Mg EXAFS functions requires the formulation of a convenient
structural model reflecting the overall chemical composition of the prepared materials. Since Ru
and Mg do not form stable double oxides one has to base an applicable local structure model on
the double oxides existing in the Mg-Ti-O ternary system[28]. Although magnesium and
titanium form two stable ternary oxides conforming to the spinel and ilmenite structural models,
110
the absence of the pronounced scattering between 3 and 4 Å seems to disagree with both possible
structural models for the Mg EXAFS function refinement (see Figure 4.5b).
A satisfactory fit of the experimental data can be achieved assuming that the local
environment of Mg bears features of both the ruthenia host (rutile structural model) and of the
stable ternary Mg-Ru-O oxide conforming to an ilmenite structural model. Typical result of the
full refinement for the Mg EXAFS functions is shown in Figure 4.5b; the parameters of the
refinement are included in Table 4.3. As shown by the parameters of the refinement summarized
in Table 4.3, the observed increase in the Mg-O bonding distance with increasing overall Mg
content from ca. 2.00 to ca. 2.07 Å can therefore be viewed as a relaxation from when the Mg
originally confined in the rutile-like oxygen coordination gradually relaxes to a local structure
similar to that in stable Mg oxide(s)[29]. The refined EXAFS functions do not reflect any change
in the Mg-O coordination suggested by the composition dependence of the unit cell volume. The
possibility of a change in the Mg-O coordination, specifically at low Mg contents, however,
cannot be ruled out since the EXAFS functions of Ru1-xMgxO2 with x < 0.10 are not available.
It needs to be stressed that the Mg-O bonding distance in all prepared materials remains
shorter than that in Mg-O or Mg-Ti-O and the Mg local environment remains rather strained. The
cationic arrangement also shows a gradual development between bonding distances and
coordination numbers characteristic for a rutile-like structure (Ru0.9Mg0.1O2) and those
approaching the values expected for ilmenite (Ru0.8Mg0.2O2). It needs to be stressed that although
the nearest cationic coordination in the latter material starts to resemble the ilmenite, there still
remain observable cation–cation interactions characteristic of a rutile-like structure. This
behavior is not surprising since both structural types are related as it was shown, e.g. in reference
[18].
111
4.3.3 Electrochemical and DEMS Measurements
The Mg doped materials are active catalysts for both oxygen and chlorine evolution. The results
of the electrochemical characterization are summarized in Figures 4.6-4.10. The incorporation of
Mg into the rutile structure apparently decreases the activity of the ruthenia-based catalysts for
the oxygen evolution process (see Figure 4.6). Such behavior contradicts that of the Ni or Co
containing analogs[10, 13]. The observed behavior seems to reflect the fact that the cation
perturbing the local structure of the catalyst is itself catalytically inactive (i.e. cannot enter redox
reactions and become a binding site of the reaction intermediates). The decrease in the oxygen
evolution activity shows, however, a non-monotonous dependence on the actual Mg content with
a maximum activity observed for the material with an overall Mg content of x = 0.10. This
behavior cannot be directly linked with the refined local structure and its explanation will most
likely need to employ advanced theoretical approaches.
Figure 4.6. Linear scan voltammograms of the oxygen evolution on MgxRu1-xO2electrodes
(0<x<0.2) recorded in 0.1 M HClO4 at a polarization rate of 5 mV s-1
. The curve assignment is
given in the Figure legend.
112
The overall activity of the Mg modified ruthenia and pure ruthenia toward parallel
oxygen and chlorine evolution can be seen in Figure 4.7. The measured faradaic current shows a
linear dependence on the chloride concentration in the electrolyte solution. The increase in the
overall activity is also accompanied with a significant change in the material’s selectivity. While
approximately 30% of the charge being passed at a concentration of 300 mM Cl− can be
attributed to oxygen evolution on the non-doped ruthenia (see Figure 4.8), in the case of the Mg
doped materials the yield of the oxygen drops below 20% for a chloride concentration of 50 mM
and the chlorine evolution becomes practically quantitative at higher chloride concentrations.
Figure 4.7. Chloride concentration dependence of the oxygen evolution (a) and chlorine
evolution (b) contributions to the overall current response of MgxRu1-xO2 electrodes to anodic
polarization in chloride containing acid media. The presented values correspond to potentiostatic
experiments at 1.25 V vs. Ag/AgCl.
113
Figure 4.8. Composition dependence of the Mg doped ruthenia selectivity towards chlorine
evolution. The data correspond to potentiostatic experiments at 1.25 V (Ag/AgCl). The actual
chloride concentrations are given in the Figure legend.
The selectivity of the Mg modified ruthenia toward chlorine evolution seems to be little
affected by the catalyst’s composition; at higher chloride concentrations the selectivity is also
independent of the electrode potentials. A potential dependence of the catalyst’s selectivity
remains, however, observable at low to medium chloride concentrations (10 to 50 mM) when all
catalysts seem to favor the chlorine evolution at lower overpotentials. The selectivity of non-
doped catalysts, however, changes toward oxygen evolution at more positive potentials (see
Figure 4.9). Detailed information on the nature of the selectivity of both Mg doped as well as
non-doped ruthenia can be obtained by analyzing the time courses of the oxygen and chlorine
evolution at constant potential in solution containing 10 mM of sodium chloride (see Figure
4.10). The DEMS signals corresponding to both anticipated reaction products start to rise
simultaneously after the application of the potential step and the chlorine evolution signal attains
114
a steady state value after 10 and 40 s. These steady state values reflect equalization of the rate of
chlorine production and its removal into the DEMS apparatus. The fact that the chlorine
evolution signal attains a steady state value before the oxygen evolution signal indicates that the
oxygen evolution process does not affect the rate of chlorine evolution. This type of behavior
contradicts the assumed strict competition of both anodic processes for the same active sites at
the catalyst’s surface. The chlorine and oxygen evolution processes apparently use different
surface structures. The obtained data also suggest that the chlorine evolution pathway is
kinetically less hindered than that used for oxygen evolution. It needs to be noted that the active
sites used in chlorine evolution can also catalyze the oxygen evolution process if chlorine
evolution becomes transport limited.
Figure 4.9. Potential dependence of the Mg doped ruthenia selectivity towards chlorine
evolution. The data were extracted from potentiostatic experiments in 0.1 M HClO4 containing
10 mM NaCl (a) and 50 mM NaCl (b).
115
Figure 4.10. Time course of DEMS-based signals of potentiostatically generated oxygen (blue)
and chlorine (red) for RuO2 (a), Ru0.95Mg0.05O2 (b), Ru0.90Mg0.10O2 (c), and Ru0.80Mg0.20O2 (d).
Signals were recorded in 0.1 M HClO4 containing 10 mM NaCl; the potential perturbation
corresponded to a step from 0.70 V to 1.25 V vs. Ag/AgCl.
The conditions of the electrochemical experiments are generally incompatible with the
sample composition in general. To establish a general validity of the electrochemical data all
experiments were run with fresh electrodes and the duration of the experiments was kept short
and never exceeded 1000 s. The surface composition after the electrochemistry as reflected in the
XPS spectra agrees well with the initial one (see Tables 4.1 and 4.2). Although this fact validates
the relevance of the presented electrochemical data it does not allow to make any predictions
related to the practical suitability of the prepared materials in industry-like applications where
the catalyst should exhibit stability on significantly longer time-scales (~108 s).
116
4.3.4 Discussion
Traditionally, the theoretical description of the oxygen evolution process on rutile-type
ruthenium dioxide designates the active sites as transition metal atoms in so-called cus positions.
The significant enhancement of the oxygen evolution activity achieved by the modification of
the rutile structure with Ni or Co is then attributed to the activation of the bridge sites for a
proton transfer by incorporation of the additional transition metal[30]. The incorporation of
alkaline earth metal cations, such as Mg, however, is unlikely to improve the catalytic activity of
the rutile structure even if present in the cus positions. The catalytic cycle for oxygen evolution
formally requires the active metal cations to be oxidizable if they should contribute to the overall
activity. This can be hardly achievable given the oxidation state of Mg in the prepared materials.
This fact is reflected in the drop in oxygen evolution activity which therefore results from a
decrease in the available active Ru cus sites via dilution effect.
The link between the selectivity of parallel oxygen and chlorine evolution with the
catalyst’s local structure has not been established. Although the selectivity of similar systems,
e.g. Co[13] or Ni[9], doped ruthenia has been experimentally assessed in parallel oxygen and
chlorine evolution, the presented results were interpreted mainly in terms of morphology rather
than local structure[24]. Although the theory suggests that superiority of the chlorine evolution
should be structure insensitive, the experimental studies on Ni, Co, and Zn[18] doped ruthenia
are rather equivocal. In this regard one may consider two types of structures attributable to
dopant incorporation. The first type represents Ni and Co doped ruthenia which generally form
surface structures characterized by isolated or stacked Ni and Co cations which reside in a
strained rutile-like environment[9, 13]. The stacking of the dopant atoms along the c axis shifts
the selectivity toward chlorine evolution[9]. Breaking the dopant arrangements along the c axis,
117
e.g. due to shear plane formation[10], shifts the selectivity toward oxygen evolution[9]. The Zn
doped materials, on the other hand, form local structures characterized by partially or fully
broken stacking of cus and bridge sites[18] and their unusual selectivity toward oxygen
evolution was previously attributed to the structural hindrance of forming peroxo bridges
between adjacent cus sites.
The Mg doped materials, from a structural point of view, represent a transition in the
local structure from where the dopant stacking in the bridge and cus sites is essentially
unperturbed (characteristic for pure RuO2 and for Ni and Co doped ruthenia) to a structure
containing dopant enriched ilmenite-like inclusions characteristic for Zn doped ruthenia. A
purely structural comparison, therefore, suggests the electrocatalytic behavior of the Mg doped
ruthenia to fall between that of the non-doped and Zn doped ruthenia. This trend seems to be
fully reflected in oxygen evolution data when both Zn2+
and Mg2+
suppress the activity in the
oxygen evolution. The adherence of the behavior of Mg doped ruthenia observed in chloride
containing solutions to the theory predictions is less pronounced. The Mg doped materials show
a preference for chlorine evolution which peaks for the material featuring about 5% Mg in
cationic positions (see Figure 4.8). Although the detailed local structure information for this
material is missing it may be envisaged that this material shows the closest similarity to the non-
doped ruthenia. The moderate selectivity shift toward oxygen evolution observed for the
materials with higher Mg content reflects the development of the catalysts’ local structure toward
ilmenite-type inclusions. Breaking the stacking of cationic positions decreases the possibility to
form peroxo bridges between adjacent cus sites which were proposed to be primary chlorine
evolution active sites[18]. The shift of the selectivity of Mg doped ruthenia toward oxygen
evolution, however, lags behind that of the Zn doped ruthenia for chlorine evolution. The precise
118
nature and functionality of the chlorine evolution active sites in Mg doped ruthenia cannot be,
however, described experimentally with sufficient precision and needs to be backed with
complementary DFT-based studies modeling both surface stability and activity of doped ruthenia
based catalysts.
4.4 Conclusions
Nanocrystalline Mg doped ruthenium dioxide catalysts were prepared by the spray-
freezing freeze-drying technique. Regardless of the average chemical composition, the prepared
materials are of nanocrystalline character with an average particle size of approximately 5 nm.
Despite the larger radius of the Mg cation compared with the tetravalent ruthenium cation, the
observed decrease in the unit cell volume suggests a nonhomogeneous distribution of Mg in the
structure with a local environment corresponding to that of a disordered ilmenite structure. Such
an arrangement is also supported by the analysis of the Mg EXAFS functions. The incorporated
Mg apparently exists in a significantly strained environment as shown by the shift in the Mg K
absorption edge energies as well as the unusually short Mg-O bonding distances which do not
exceed 2.07 Å. The prepared materials are active in both oxygen as well as in chlorine evolution
reaction. While the oxygen evolution activity is suppressed on Mg doped ruthenia, the chlorine
evolution is enhanced compared to the conventional ruthenia.
4.5 Acknowledgements
This work was supported by the European Commission within the Initial Training
Network ELCAT (Project No. 214936). The synchrotron measurement time was provided by the
KEK of Japan within the projects 2013R-35 and 2009R-29. Valery Petrykin of the J. Heyrovský
119
Institute of Physical Chemistry and Maki Okube of the Tokyo Institute of Technology are
gratefully acknowledged for their assistance with the synchrotron measurements and additional
thanks to Valery Petrykin for performing the EXAFS data analysis. The authors would also like
to thank Zdeněk Bastl of the J. Heyrovský Institute of Physical Chemistry for his assistance with
the XPS measurements and analysis.
120
4.6 References
1. Trasatti, S., Electrocatalysis in the anodic evolution of oxygen and chlorine. Electrochim
. Acta, 1984. 29: p. 1503-1512.
2. Trasatti, S., Electrocatalysis: understanding the success of DSA®. Electrochim . Acta,
2000. 45: p. 2377-2385.
3. Burke, L.D., et al., The oxygen electrode. Part 8.—Oxygen evolution at ruthenium
dioxide anodes. J. Chem. Soc., Faraday Trans. 1, 1977. 73: p. 1659-1671.
4. Bard, A.J. and L.R. Faulkner, Electrochemical Methods: Fundamentals and Applications.
2nd Edition ed. 2001, New York: John Wiley & Sons, Inc.
5. Hansen, H.A., et al., Electrochemical chlorine evolution at rutile oxide (110) surfaces.
Phys. Chem. Chem. Phys., 2010. 12: p. 283-290.
6. Man, I.C., et al., Universality in Oxygen Evolution Electrocatalysis on Oxide Surfaces.
Chem. Cat. Chem., 2011. 3: p. 1159-1165.
7. Rossmeisl, J., et al., Electrolysis of water on oxide surfaces. J. Electroanal. Chem., 2007.
607: p. 83-89.
8. Exner, K.S., et al., Chlorine Evolution Reaction on RuO2(110): Ab initio Atomistic
Thermodynamics Study – Pourbaix Diagrams. Electrochim . Acta, 2014. 120: p. 460-466.
9. Macounova, K., et al., Parallel oxygen and chlorine evolution on Ru1−xNixO2−y
nanostructured electrodes. Electrochim. Acta, 2008. 53: p. 6126-6134.
10. Petrykin, V., et al., Local structure of nanocrystalline Ru1-xNixO2-δ dioxide and its
implications for electrocatalytic behavior - an XPS and XAS study. J. Phys. Chem. C,
2009. 113: p. 21657-21666.
11. Krstajic, N. and S. Trasatti, Cathodic Behavior of RuO2-Doped Ni/Co3O4 Electrodes in
Alkaline Solutions: Surface Characterization. J. Electrochem. Soc., 1995. 142: p. 2675-
2681.
121
12. Krstajic, N. and S. Trasatti, Cathodic behaviour of RuO2-doped Ni/Co3O4 electrodes in
alkaline solutions: hydrogen evolution. J. Appl. Electrochem., 1998. 28: p. 1291-1297.
13. Petrykin, V., et al., Local structure of Co doped RuO2 nanocrystalline electrocatalytic
materials for chlorine and oxygen evolution. Catal. Today, 2012. 202(63-69).
14. Silva, L.M.D., J.F.C. Boodts, and L.A. DeFaria, ‘In situ’ and ‘ex situ’ characterization of
the surface properties of the RuO2 (x) + Co3O4 (1-x) system. Electrochim. Acta, 2000.
45: p. 2719-2727.
15. Silva, L.M.D., J.F.C. Boodts, and L.A.D. Faria, Oxygen evolution at RuO2(x)+Co3O4(1-
x) electrodes from acid solution. Electrochim. Acta, 2001. 46: p. 1369-1375.
16. Silva, L.M.D., L.A.D. Faria, and J.F.C. Boodts, Electrochemical impedance
spectroscopic (EIS) investigation of the deactivation mechanism, surface and
electrocatalytic properties of Ti/RuO2(x)+Co3O4(1-x) electrodes. J. Electroanal. Chem.,
2002. 532: p. 141-150.
17. Hummelgård, C., et al., Physical and electrochemical properties of cobalt doped
(Ti,Ru)O2 electrode coatings. Mater. Sci. Eng., B, 2013. 178: p. 1515-1522.
18. Petrykin, V., et al., Zn-doped RuO2 electrocatalysts for selective oxygen evolution:
relationship between local structure and electrocatalytic behavior in chloride containing
media. Chem. Mater. , 2011. 23: p. 200-207.
19. Gaudet, J., et al., Physicochemical Characterization of Mixed RuO2-SnO2 Solid
Solutions. Chem. Mater., 2005. 17: p. 1570-1579.
20. Wu, X., et al., Nano-crystalline RuxSn1-xO2 powder catalysts for oxygen evolution
reaction in proton exchange membrane water electrolysers. Int. J. Hydrogen Energy,
2011. 36: p. 14796-14804.
21. Xiong, K., et al., Sn and Sb co-doped RuTi oxides supported on TiO2 nanotubes anode
for selectivity toward electrocatalytic chlorine evolution. J. Appl. Electrochem., 2013. 43:
p. 847-854.
22. Macounová, K., et al., Influence of Oxygen on Reactivity of Ru1−xFexO2−y-Doped
Materials. J. Electrochem. Soc., 2008. 11: p. F27-F29.
122
23. Makarova, M., et al., The electrocatalytic behavior of Ru0.8Co0.2O2-x - the effect of
particle shape and surface composition. Electrochim . Acta, 2008. 53: p. 2656-2664.
24. Petrykin, V., et al., Tailoring the selectivity for electrocatalytic oxygen evolution on
ruthenium oxides by zinc substitution. Angew. Chem., 2010. 49: p. 4813-4815.
25. Tretyakov, Y.D. and O.A. Shlyakhtin, Recent progress in cryochemical synthesis of oxide
materials. J. Mater. Chem., 1999. 9: p. 19-24.
26. Briggs, D. and J.T. Grant, Surface Analysis by Auger and X-ray Photoelectron
Spectroscopy. 2003, Cromwell Press, Trowbridge: IM Publications and SurfaceSpectra
Ltd.
27. Newville, M., IFEFFIT : interactive XAFS analysis and FEFF fitting. J. Synchrotron
Radiat., 2001. 8: p. 322-324.
28. Shannon, R.D., Revised Effective Ionic Radii and Systematic Studies of Interatomic
Distances in Halides and Chalcogenides. Acta. Cryst., 1976. A32: p. 751-767.
29. Wyckoff, R.W.G., The Structure of Crystals. 2nd ed. 1935, New York, USA: Reinhold
Publishing Corporation.
30. Halck, N.B., et al., Beyond the volcano limitations in electrocatalysis - oxygen evolution
reaction. Phys. Chem. Chem. Phys., 2014. 16: p. 13682.
123
Chapter 5
Reformate-Tolerant Pt-based Catalysts for the Electrochemical Hydrogen Pump
5.1 Introduction
Efficient and cost-effective methods for hydrogen purification are important for several
applications and the development of such methods is critical to the overall advancement of
hydrogen fuel-based technologies. Many industrial processes, such as steam reforming of
gasified coal and natural gas in conjunction with the water gas shift reaction, produce gas
mixtures of CO2, H2, water vapor, CO, and other trace impurities. Current state of the art
technologies for hydrogen purification and separation include chemical adsorption methods
based on alkaline or ammine scrubbing, pressure swing adsorption (PSA), membrane separation
purifiers (polymer, ceramic, or palladium based), and cryogenic separation. Pressure swing
adsorption is the most widely used method in the large scale separation and purification of
hydrogen[1-3]. Although this method is able to produce hydrogen at high purity (>99.9%) with
moderate to high recovery rates (70-95%), it can also require several wash columns to remove
CO and CO2 [3-6]. Similarly, cryogenic separation also has the advantage of providing hydrogen
at a high purity (90-99%) with an even higher recovery rate (95%), but this method requires a
high energy input to condense out impurities in the contaminated gas stream[5, 7]. Membrane
separation technologies are generally used for the final purification step and require a relatively
clean hydrogen stream (>98%) to produce highly purified hydrogen (99.9999%), The downside
to membrane separation techniques is that they generally require the use of elevated temperatures
(≥ 300 °C) to operate. Furthermore, the membrane separators often require high pressure inlet
gas feeds, requiring additional energy input, and can experience hydrogen embrittlement[3, 5].
124
An attractive alternative to the conventional hydrogen separation and purification
technologies is the electrochemical hydrogen pump (ECHP). The concept of hydrogen
purification using a PEM-based ECHP was first demonstrated by Sedlak et al.[8] in the early
1980's. Since then it has been shown by Gardner et al. that it was possible to recover 80% of
hydrogen with an energy efficiency of 80% from H2/CO2 mixtures in a single cell while pulsing
the anode potential[9]. More recently, Benziger et al. reported that a multistage electrochemical
hydrogen pump operating with programmed voltage pulsing was able to recover >98% of
hydrogen from H2/CO2 gas mixtures with energy efficiencies >92%[10]. In addition to being an
efficient means of obtaining highly purified H2, it has also been demonstrated that
electrochemical hydrogen pumps offer an efficient means of hydrogen compression [11-14] and
hydrogen recirculation in a fuel cell stack[15]. The revitalized interest in this technology within
the past few years has led to large advancements of the ECHP and it is likely that higher
recovery rates and energy efficiencies could be achieved with further development.
There have been several studies recently focusing on high temperature hydrogen pumps
using polybenzimidazole (PBI) membranes[16, 17]. Some investigations have also looked at
phosphoric acid [5], molten carbonate[18, 19], and solid-acid membrane[11] analogues.
However, there is a lack of literature regarding low temperature (< 100 °C) hydrogen pump
systems utilizing CO-tolerant electrocatalysts at the anode[9, 20]. Over the last several decades
there has been a surge of literature exploring the use of reformate-tolerant catalysts in proton-
exchange membrane fuel cells (PEMFCs), including both gas-fed and liquid-fed fuel cells such
as direct-methanol fuel cells[21-28]. The current work aims to extend the previous findings and
demonstrate the use of Pt-based alloys as efficient reformate-tolerant catalysts suitable for the
electrochemical hydrogen pump.
125
5.2 Experimental
5.2.1 Catalyst Preparation
The PtRu/C (41.4 wt.% Pt, 32.1 wt.% Ru) and Pt/C (46.2 wt.% Pt) catalysts used in this
work were obtained from Tanaka Corporation, Japan. The PtSn/C (20 wt.%) with an atomic ratio
of 3:1 (Pt:Sn) was obtained from E-TEK Corporation.
PtNi/C (40 wt. %) with an atomic ratio of 1:1 (Pt:Ni) was synthesized by aqueous
precipitation of Ni2+
onto Pt/C. The Pt/C precursor was synthesized by dissolving an appropriate
amount of H2Pt(OH)6 (DF Goldsmith Chemical and Metal, USA) in a 6% H2SO3 solution. The
solution was stirred at room temperature overnight and then diluted with Millipore H2O to bring
the total volume to 50 mL. Approximately 700 mg of Vulcan XC-72R (Cabot Corp., USA) was
added to the flask and ultrasonically dispersed for 30 minutes before adjusting the pH to 3.0 with
1.0 M NaOH solution. A stoichiometric amount of 30% H2O2 relative to H2Pt(OH)6 was then
added dropwise to the slurry and the pH was periodically adjusted to 3.0 with 1.0 M NaOH three
more times. Afterwards, the slurry was boiled for one hour and cooled to room temperature
before filtering and washing with 500 mL of Millipore H2O. The resulting powder was dried
overnight at 110 °C and then ultrasonically dispersed in 50 mL of Millipore H2O. An appropriate
amount of Ni(NO3)2•6H2O (Alfa Aesar, USA) was added to the slurry and stirred for 30 minutes.
The Ni was then precipitated onto the Pt/C support in the form of an oxy/oxy-hydroxide layer by
dropwise addition of 1.0 M NaOH until the pH was adjusted to 8.5. The slurry was boiled for 30
minutes then cooled to room temperature, filtered, and washed with 500 mL Millipore H2O. The
resulting powder was dried at 110 °C overnight and then annealed at 800 °C under 5% H2/Ar for
two hours. An acid wash treatment (stirring in 1.0 M HNO3 at 70 °C overnight) was used to
leach out any excess Ni.
126
Pt3Mo (40 wt. %) was synthesized by aqueous impregnation of Pt/C with MoO3. The
Pt/C precursor was synthesized according to the above procedure. A solution of H2MoO4 was
prepared and an appropriate amount was added to a slurry of Pt/C in Millipore H2O. The slurry
was stirred for 24 hours and then allowed to settle before filtering, washing 200 mL Millipore
H2O, and drying at 110 °C overnight. The resulting powder was annealed in 5% H2/Ar at 500 °C
for 6 hours.
Ternary alloy catalyst PtRuNi/C was synthesized in-house starting with the acid washed
PtNi/C precursor. Ruthenium was deposited by dispersing the PtNi/C in Millipore H2O and
dissolving an appropriate amount of RuCl3•xH2O. The ruthenium was then reduced by passing a
stream of hydrogen through the stirred solution for approximately 2 hours. Afterwards, the
catalyst powder was filtered, washed, and dried at 80 °C overnight. The dried powder was
annealed in 5% H2/Ar for 3 hours at 300 °C. The final metal loading was calculated to be 49
wt.% with an atomic ratio of 1:1:1 (Pt:Ru:Ni).
5.2.2 Physical Characterization
The crystallinity and particle size of the carbon-supported nanoparticles were
characterized using a Rigaku Ultima IV X-ray diffractometer with Cu Kα radiation (λ = 1.5418
Å) operating at 44 kV and 40 mA. Samples were scanned from 20° < 2θ < 90° and the recorded
patterns were matched against the PDXL database. The average sample composition of
synthesized binary and ternary alloy nanoparticles was confirmed with X-ray dispersive
spectroscopy using a Hitachi S-4800 scanning electron microscope (SEM) equipped with an
EDAX Sapphire Si(Li) detector. In addition, HRTEM images of the synthesized PtNi/C and
PtNiRu/C materials were obtained with JEOL JEM-2010F Field Emission Electron Microscope.
127
5.2.3 Electrochemical Cell Polarization Measurements
Membrane electrode assemblies (MEAs) were prepared using the corresponding anode
catalyst and a Pt/C cathode. A total metal loading of approximately 0.5 mg cm-2
was used for all
anode and cathode electrodes. To prepare the anode and cathode electrodes, catalyst ink
suspensions were prepared in a water-alcohol mixture with the required amount of ionomer
solution (Nafion D-521, 5 wt.%) and sprayed onto a Sigracet gas diffusion layer (GDL).
Electrodes were coated with an interfacial layer of 0.5 mgNafion cm-2
before hot pressing. Hot
pressing the electrodes with a Nafion 212 membrane was carried out at 500 psi and 135 °C for
four minutes. MEA testing was performed in a single cell with graphite serpentine flow fields.
All MEAs were humidified with N2 (100% RH) at a cell temperature of 85 °C for at least 90
minutes before activating with H2/H2 (25/30 psi backpressure on anode/cathode) for a minimum
of 90 minutes. All subsequent measurements were taken at a cell operating temperature of 85 °C
and 100% RH with a backpressure of 25 psi on the anode and 30 psi on the cathode. Hydrogen
pump polarization curves were recorded using an Autolab Potentiostat/Galvanostat Model
PGSTAT30 with Booster 20A module (Metrohm, USA) and the Nova 1.10 software package.
Cell activation was carried out by holding the cell potential at 25 mV for one hour and
then holding at 50 mV for an additional 30 minutes. Galvanostatic polarization was performed
by ramping the cell current to 7.0 A at 70 mA steps. After recording the cell polarization curve
under H2, the anode gas feed was switched to a reformate gas mixture (100 ppm CO, 15% CO2,
1% CH4, 45% H2, balance N2) for 30 minutes before repeating the cell polarization curve
measurement. The anode and cathode gas feeds were switched to nitrogen and hydrogen,
respectively, between polarization measurements to allow for half-cell polarization and to
remove any adsorbed CO. Finally, a hydrogen gas mixture containing 300 ppm CO was passed
128
over the anode for 30 minutes before collecting the final polarization measurements. A similar
set of cell measurements was made using a new set of MEAs to investigate the CO2 tolerance of
each catalyst. The cells were first polarized while flowing H2 over both the anode and the
cathode. Afterwards, the anode gas feed was switched to 50:50 (molar ratio) H2:CO2 and the cell
polarization was recorded again.
Impedance measurements were recorded between polarization measurements in order
determine the internal cell resistance. Impedance measurements were made potentiostatically at
the open circuit potential (OCP) from 20 kHz to 0.1 Hz with a 10 mV amplitude. All reported
potentials are quoted in the reversible hydrogen electrode (RHE) scale and have been corrected
for iR losses.
5.3 Results and Discussion
5.3.1 Physical Characterization
The powder XRD patterns for each electrocatalyst sample tested are shown in Figure 5.1.
The Pt/C (Tanaka, Japan) sample shows the anticipated diffraction pattern regions characteristic
of the fcc (111) and (220) reflections appearing at 39.51 and 46.05 degrees, respectively. The
expected shift in the primary diffracting regions due to the incorporation of dopant atoms is
observed in the commercially available Pt3Sn/C and Pt2Ru3/C samples. The peaks in the Pt3Sn/C
diffraction pattern shift to lower 2θ values as expected due to the larger atomic radius of Sn
relative to Pt. Similarly, the peaks in the Pt2Ru3/C sample were shown to shift to higher 2θ
values due to the smaller atomic radius of Ru. In addition to shifting to higher angles, some Ru
may exist in the hexagonal phase as evidenced by the broad overlap between the peaks at 40.48
and 44.87 degrees. This is in agreement with Tripkovic et al.[29]. Due to the similar size of the
129
Pt (1.39 Å) and Mo (1.39 Å) atomic radii, no significant shift was observed in the diffraction
peaks for this sample. A significant difference was observed, however, in the diffraction patterns
of the synthesized PtNi/C and PtNiRu/C materials. The broad shoulder towards higher angles
associated with each peak indicates the presence non-alloyed platinum in addition to the alloyed
PtNiRu or PtNi supported particles. In the case of PtNiRu/C, a shoulder at 38.48 degrees and a
small peak at 43.91 degrees indicates the presence of some Ru in the hexagonal phase, similar to
the commercial Pt2Ru3/C material.
Figure 5.1. Powder XRD patterns for electrocatalyst samples.
130
Table 5.1. Results of XRD and HRTEM physical characterization.
Electrocatalyst
Sample
Lattice
Type
Atomic
Ratio
XRD Analysis
TEM
Particle Size
/ Å
Lattice
Parameter
/ Å
d(111)
/ Å Particle
Size / Å
d-spacing
/ Å
Pt/C fcc - 17.3 3.9184 2.279
Pt2Ru3/C fcc 2:3 15.1 3.8679 2.226
Pt3Mo/C fcc 3:1 40.3 3.9209 2.265
PtNi/C fcc 1:1 44.7 3.8250 2.208
56.2 ± 16.9 2.25 ± 0.03
Pt3Sn/C fcc 3:1 15.6 3.9916 2.308
PtRuNi/C fcc 1:1:1 40.7 3.8269 2.204
59.2 ± 15.8 2.20 ± 0.11
The coherent domain size of each sample as calculated by the Scherrer equation can be
seen in Table 5.1. Representative HRTEM images of the PtNi/C and PtNiRu/C electrocatalysts
synthesized in-house can be seen in Figure 5.2. From Figure 5.2 it can be seen that the
nanoparticles are well dispersed on the carbon support. The average particle sizes of these
samples was also calculated by measuring over 300 randomly selected particles and are shown in
Table 5.1. The average size of the supported nanoparticles calculated from TEM analysis is
between 5.6 and 6.0 nm, which is only slightly larger than those calculated from the Scherrer
equation.
131
Figure 5.2. HRTEM images of the synthesized PtNi/C (a) and PtNiRu/C materials.
5.3.2 Cell Polarization Measurements:
Shown in Figure 5.3 are the half-cell anode polarization curves for hydrogen oxidation.
All polarization curves are normalized to the Pt metal-loading, since in most cases Pt provides
the only catalytically active site for hydrogen oxidation. In addition, all polarization curves were
corrected for iR-losses using the high frequency resistance measurement determined via
electrochemical impedance spectroscopy (EIS). This configuration allows for a direct
comparison of the different anode catalyst activities toward hydrogen oxidation for an
electrochemical cell operating in the hydrogen pump mode. As expected, pure Pt/C shows one of
the highest activities for the hydrogen oxidation reaction (HOR). It should be noted that the
PtRu/C and PtNiRu/C show a higher activity due to the fact that both samples contained some
Ru in the hcp phase (as evidenced by XRD in Figure 5.1), which is known to show sufficient
HOR activity between 0 to 300 mV vs. RHE (below the point of OHads nucleation on Ru)[30].
The other materials investigated showed a lower HOR activity in the presence of pure hydrogen.
132
This is due to the fact that the other metals present in the Pt-based alloy are not HOR active and
thus dilute the sites available for hydrogen oxidation. Although the polarization curves are
normalized to the Pt-loading, one cannot assume a completely homogeneous distribution of each
metal within the nanoparticles or at the particle surface, therefore the number of surface sites
available for HOR may be less than anticipated. In all cases it can be seen that high specific
currents are obtained at very low overpotentials due to the extremely facile kinetics of the HOR
on Pt and Pt-based catalysts[31].
Figure 5.3. Anode half-cell polarization curves of an electrochemical hydrogen pump with the
anode and cathode being fed pure H2. Polarization measurements were collected at a cell
temperature of 85°C, 100% relative humidity, and 25/30psi backpressure on the anode/cathode,
respectively.
Other PEM-based ECHPs have been reported previously. Shown in Table 5.2 are the
various operating conditions and current densities for ECHPs operating with pure H2 supplied to
the anode. Comparatively, the current density reported for the Pt anode in Figure 5.3 is
133
considerably higher than those reported in the literature. However, it is important to note that in
all cases the operating temperature used in this study was higher than the others reported.
Additionally, the use of a thinner membrane (50.8 μm for Nafion 212 vs. 127 μm for Nafion
115) also contributes significantly considering that the reported current densities were not iR-
corrected.
Table 5.2. Performance and operating conditions of various ECHPs reported in the literature
Reference Operating
Temp / °C
Anode Loading /
mgPt cmgeo-2
Membrane
J @ 150 mV /
A cmgeo-2
Figure 5.3 (Pt/C) 85 0.5 Nafion 212 1.5
Gardner et al.[9] 20 1.0
(1.5 mgPtRu cm2)
Nafion 115 0.31
Benziger et al.[10] 50 0.4 Nafion 115 0.65
Fateev et al.[13] 75 0.8 Nafion 117 0.51
Barbir et al.[15] 30 0.8 3M PFSA 1000 EW (28 μm) 0.60
Shown in Figure 5.4 are the anode half-cell polarization curves for hydrogen oxidation
with an anode gas feed of either H2 containing 300 ppm CO or a reformate gas mixture
containing 100 ppm CO, 15% CO2, 1% CH4, 45% H2, and N2 to balance. PtRu/C and PtNiRu/C
again show the highest activity, and thus the highest tolerance to CO/CO2 poisoning. The high
activity of these two materials can be attributed to the presence of Ru. It has been reported by
Tripković et al. that Pt2Ru3/C is generally a poor catalyst for methanol oxidation, but shows an
extremely high CO-tolerance[32]. They attributed the excellent CO tolerance not to the twinned
structure of Pt2Ru3 but rather to the presence of a Ru-rich (hcp) phase in the catalyst, such as that
observed in the XRD profiles in Figure 5.1. It is widely accepted that the mechanism for CO-
removal is dependent upon water activation at Ru sites to form OHads, otherwise known as the
bifunctional mechanism[33, 34]. PtMo/C also shows a relatively high tolerance to CO and the
134
reformate mixture. It has been shown in previous studies that PtMo/C has a comparable CO-
tolerance to PtRu/C and in some cases even shows a higher tolerance[21]. The mechanism for
PtMo/C, although similar to that of PtRu/C, is more dependent on the turnover of the Mo(IV/VI)
redox couple. The lower activity of PtNiRu/C compared to that of PtRu/C is likely due to
presence of Ni, which appears to dilute the surface Pt sites required for hydrogen adsorption in
addition to diluting the Ru surface sites needed for OH adsorption to remove COads. This dilution
effect is also evidenced by the lower activity of PtNi/C vs. Pt/C, in which case the surface Ni
atoms are inactive for hydrogen oxidation and offer no means of promoting COads removal via
the bifunctional mechanism. It has been reported by Mohsen et al.[35] that the presence of Ni
lowers the d-band center of Pt and aids in CO2 tolerance, though no correlation was found
between the CO tolerance and Pt d-band center and PtNi/C. As was also reported by Mohsen et
al.[35], Pt-based anode electrocatalysts that rely on the bifunctional mechanism for CO removal,
such as PtRu/C, PtMo/C and PtSn/C, typically show a high CO tolerance but show a low
tolerance to CO2. In contrast, Pt-based catalysts that drastically improve CO2 tolerance by
altering the electronic properties of Pt, such as PtNi/C or PtCo/C, generally show a very poor
tolerance to CO poisoning. The behavior observed in Figure 5.4 seems to indicate that the anode
electrocatalyst activity is heavily dependent upon the presence of CO and that the presence of
CO2 or CH4 in the reformate mixture is not the dominant poisoning species.
135
Figure 5.4. Anode half-cell polarization curves in an electrochemical hydrogen pump with the
cathode being fed hydrogen and the anode being fed by hydrogen containing 300ppm CO (filled
circles) or a reformate gas mixture containing 100 ppm CO, 15% CO2, 1% CH4, 45% H2, and N2
balance (empty circles). Polarization measurements were collected at a cell temperature of 85°C,
100% relative humidity, and 25/30psi backpressure on the anode/cathode, respectively.
It is well known that CO2 can be converted to CO via the reverse water gas shift reaction
by either of the following two pathways[36-40]:
CO2 + H2 ⇌ CO +H2O (5.1)
CO2 + 2M-Hads ⇌ 2M-COads + H2O + M (5.2)
Such a conversion occurs at the electrode interface when CO2 adsorption occurs adjacent to H2
adsorption. It has been shown that CO2 can poison Pt-based catalysts that otherwise show a high
tolerance to CO poisoning beyond the levels accounted for by simple dilution[37, 38].
136
Furthermore, the trends observed for CO tolerance in Pt-based alloys do not necessarily apply
when considering CO2 tolerance[41-43].
The hydrogen oxidation activity of four different Pt-based catalysts in the presence of an
H2-CO2 (50:50 molar ratio) gas mixture at the anode can be seen in Figure 5.5. It can be seen
from Figure 5.5 that the extent of catalyst deactivation induced by CO2 poisoning is less than that
observed for CO (See Figure 5.4). Although the extent of poisoning is less in the case of CO2,
the same trend observed for CO tolerance is observed again here, with Pt2Ru3/C showing the
highest HOR activity and thus the highest CO2 tolerance followed by PtRuNi/C, Pt/C, and
PtNi/C. This is in contrast to the results reported by Mohsen et al.[35], where PtNi/C was shown
to have the highest tolerance to CO2 and PtRu/C was shown to have one of the lowest tolerances.
It is important to note that Strasser et al.[44] has shown that unsupported PtxNiy nanoparticles
undergo selective dissolution when exposed to corrosive anodic conditions. Although it was
concluded that the resulting Pt-enriched core-shell structures show higher catalytic activity for
the oxygen reduction reaction (ORR), this may not be the case for HOR. It is well known that the
surface structure of PtxNiy is heavily dependent on the Ni content, the material preparation
method, and the electrode potential[44-46]. It has been shown by Hoffmannová et al.[46] that
hydrogen adsorption on PtNi triggers Ni segregation to the surface while on the other hand the
Ni is confined primarily to the subsurface when the electrode is polarized to the double-layer
region. The discrepancies observed in the CO2 tolerance between the PtNi/C sample reported
herein and that reported by Mohsen et al.[35] are likely due to differences in the surface
composition of the catalysts, which may be due to synthetic route chosen or the electrode history
prior to polarization measurements. As mentioned previously, Ni segregation occurs when the
electrode is polarized to the hydrogen adsorption region, which would decrease HOR Pt-based
137
specific activity (hence the lower activity than Pt) as Ni is significantly less catalytically active
for the HOR than Pt. It is likely that the nonhomogeneous distribution of Ni under these
circumstances would also mitigate any promotional effects on the Pt d-band center that are
believed to increase the CO2 tolerance[35], as such effects (in the case of ORR) are known to
occur when the Ni is located subsurface and a Pt skin is formed on the surface[46]. It should be
noted that two other PtxNiy/C samples were also prepared in-house using the same method
reported above and were subsequently tested in the ECHP (not shown). The additional samples
displayed nearly identical HOR activities and CO/CO2 tolerances, indicating that the poor CO2
tolerance is not batch-specific.
The improved CO2 tolerance of Pt2Ru3/C, similar to the improved CO and reformate
tolerance observed in Figure 5.4, is attributed to the activation of water via the bifunctional
mechanism. The presence of Ru lowers the potential at which water is activated, decreasing the
onset of surface OH formation, which subsequently removes adsorbed CO formed by the reverse
water gas shift reaction on adjacent Pt sites. This is also the case for the PtNiRu/C material,
although again the Ni dilutes the key catalytic sites on the particle surface reducing the overall
activity and CO2 tolerance. Ideally, the presence of Ni would lower the Pt d-band center aiding in
the removal of CO2 while the presence of surface Ru sites adjacent to Pt would aid in the
removal of CO via the bifunctional mechanism. It possible that with more development an
improved ternary alloy (PtNiRu/C) could be formed that offers superior tolerance to both CO and
CO2.
138
Figure 5.5. Anode half-cell polarization curves in an electrochemical hydrogen pump with the
cathode being fed pure hydrogen and the anode being fed by hydrogen containing 50% molar
CO2. Polarization measurements were collected at a cell temperature of 85°C, 100% relative
humidity, and 25/30psi backpressure on the anode/cathode, respectively.
Although a direct comparison cannot be made between the PEM-based ECHPs reported
in the literature for CO/CO2 tolerance due to the large difference in gas mixture composition,
there have been reports of PBI membrane-based ECHPs tested in the presence of reformate gas
mixtures similar to those employed here. For instance, Thomassen et al.[17] reported an ECHP
operating with a PBI membrane at 160 °C while the anode was exposed to an unhumidified gas
mixture consisting of 44% H2, 35% N2, 21% CO2, 100 ppm CO. A current density of
approximately 1.25 A cm-2
was obtained at an anode overpotential of 180 mV, compared to an
overpotential of 240 mV (360 mV uncorrected) at the same current density for the cell operating
with a Pt2Ru3/C anode reported in Figure 5.4. Additionally, Perry et al.[16] reported a PBI
membrane-based ECHP with a 1.0 mgPt cm-2
loading at the anode operating at 160 °C that
139
attained a current density of approximately 2.0 A cm-2
at 150 mV under a humidified pure
hydrogen feed or 0.95 A cm-2
at 150 mV under unhumidified conditions. This is comparable to
the performance of the Pt2Ru3/C catalyst presented in Figure 5.3, which reached a current density
of 1.5 A cm-2
at 150 mV (uncorrected). In either case, the higher current densities attained by the
PBI-based ECHPs indicate that the elevated operating temperature (160 °C in PBI vs. 80 °C in
PEM) may offer reformate tolerances that are beyond what is attainable with reformate-tolerant
catalysts in low temperature ECHPs. This is largely due to the fact that CO adsorption on Pt is
thermodynamically unfavorable at elevated temperatures as evidenced by the rather negative
standard entropy value[47, 48].
5.4 Conclusions
Of the Pt-based electrocatalysts reported herein, Pt2Ru3/C has proven to be a promising
CO/CO2 tolerant anode catalyst for a PEM-based cell operating in the ECHP mode. This agrees
well with previous findings in the literature that investigated the use of PtRu/C as a reformate
tolerant catalyst in cells operating in the PEMFC mode. The high tolerance of Pt2Ru3/C is
attributed to the activation of water on Ru sites at much lower overpotentials, which then permits
the removal of CO adsorbed on adjacent Pt sites, i.e. the bifunctional effect. The presence of Ni
in the PtNi/C and PtNiRu/C samples was not shown to enhance the CO2 tolerance as was
reported previously. However, it is possible that with more development and fine tuning of the
Ni distribution that ternary alloys such as PtNiRu/C could offer superior reformate tolerances
that exceed Pt2Ru3/C.
140
5.5 Acknowledgements
The authors gratefully acknowledge the financial support from the Department of Energy,
under the auspices of a Small Business Innovative Research grant lead by Proton Onsite. The
authors would also like to thank Shraboni Ghoshal of Northeastern University for her assistance
with the hydrogen pump cell measurements.
141
5.6 References
1. Ruthven, D.M., S. Farooq, and K.S. Knaebel, Pressure swing adsorption. 1993, New
York, N.Y.: VCH Publishers.
2. Sircar, S., Pressure Swing Adsorption. Ind. Eng. Chem. Res., 2002. 41: p. 1389-1392.
3. Liu, K., C. Song, and V. Subramani, Hydrogen and Syngas Production and Purification
Technologies. 2009, Hoboken, NJ: A Joint Publication of the Center for Chemical
Process Safety of the American Institute of Chemical Engineers and John Wiley & Sons,
Inc. 564.
4. Uehara, I., Separation and Purification of Hydrogen, in Energy Carriers and Conversion
Systems with Emphasis on Hydrogen, T. Ohtaaytexasphltw, Editor. 2008, Eolss
Publishers: Paris, France.
5. Farooque, M., A. Kush, and S. Abens, Novel Electrochemical Hydrogen Separation
Device Using Phosphoric Acid Membrane Cell Sep. Sci. Technol., 1990. 25: p. 1361-
1373.
6. Grande, C.A., Advances in Pressure Swing Adsorption for Gas Separation. ISRN
Chemical Engineering, 2012. 2012: p. 1-13.
7. Grashoff, G.J., C.E. Pilkington, and C.W. Corti, The Purification of Hydrogen. Platinum
Metals Rev., 1983. 27: p. 157-169.
8. Sedlak, J.M., J.F. Austin, and A.B. LaConti, Hydrogen Recovery and Purification Using
the Solid Polymer Electrolyte Electrolysis Cell. Int. J. Hydrogen Energy, 1981. 6: p. 45-
51.
9. Gardner, C.L. and M. Ternan, Electrochemical separation of hydrogen from reformate
using PEM fuel cell technology. J. Power Sources, 2007. 171: p. 835-841.
10. Abdulla, A., et al., Efficiency of Hydrogen Recovery from Reformate with a Polymer
Electrolyte Hydrogen Pump. AIChE J., 2011. 57: p. 1767-1779.
11. Papandrew, A.B., et al., Electrochemical Hydrogen Separation via Solid Acid
Membranes. J. Electrochem. Soc., 2014. 161: p. F679-F685.
142
12. Strobel, R., et al., The compression of hydrogen in an electrochemical cell based on a PE
fuel cell design. J. Power Sources, 2002. 105: p. 208-215.
13. Grigoriev, S.A., et al., Description and characterization of an electrochemical hydrogen
compressor/concentrator based on solid polymer electrolyte technology. Int. J. Hydrogen
Energy, 2011. 36: p. 4148-4155.
14. Rohland, B., et al., Electrochemical Hydrogen Compressor. Electrochim. Acta, 1998. 43:
p. 3841-3846.
15. Barbir, F. and H. Gorgun, Electrochemical hydrogen pump for recirculation of hydrogen
in a fuel cell stack. J. Appl. Electrochem., 2007. 37: p. 359-365.
16. Perry, K.A., G.A. Eisman, and B.C. Benicewicz, Electrochemical hydrogen pumping
using a high-temperature polybenzimidazole (PBI) membrane. J. Power Sources, 2008.
177: p. 478-484.
17. Thomassen, M., E. Sheridan, and J. Kvello, Electrochemical hydrogen separation and
compression using polybenzimidazole (PBI) fuel cell technology. J. Nat. Gas Sci. Eng.,
2010. 2: p. 229-234.
18. Weaver, D. and J. Winnick, Performance of an Electrochemical Membrane H2S
Separator. J. Electrochem. Soc., 1992. 139: p. 492-498.
19. Alexander, S.R. and J. Winnick, Removal of Hydrogen Sulfide from Natural Gas through
an Electrochemical Membrane Separator. AIChE J., 1994. 40: p. 613-620.
20. Ibeh, B., C. Gardner, and M. Ternan, Separation of hydrogen from a hydrogen/methane
mixture using a PEM fuel cell. Int. J. Hydrogen Energy, 2007. 2007(32): p. 908-914.
21. Mukerjee, S., et al., Electrocatalysis of CO Tolerance by Carbon-Supported PtMo
Electrocatalysts in PEMFCs. J. Electrochem. Soc., 2004. 151: p. A1094-A1103.
22. Antolini, E., Platinum-based ternary catalysts for low temperature fuel cells Part II.
Electrochemical properties. Appl. Catal., B, 2007. 74: p. 337-350.
143
23. Mu, R., et al., Synergetic Effect of Surface and Subsurface Ni Species at Pt-Ni Bimetallic
Catalysts for CO Oxidation. J. Am. Chem. Soc., 2011. 133: p. 1978-1986.
24. Ioroi, T., et al., Enhanced CO-Tolerance of Carbon-Supported Platinum and
Molybdenum Oxide Anode Catalyst. J. Electrochem. Soc., 2003. 150: p. A1225-A1230.
25. Gasteiger, H.A., N.M. Markovic, and J. Philip N. Ross, Electrooxidation of CO and
H2/CO Mixtures on a Well-Characterized Pt3Sn Electrode Surface. J. Phys. Chem., 1995.
99: p. 8945-8949.
26. Schubert, M.M., et al., Bimetallic PtSn catalyst for selective CO oxidation in H2-rich
gases at low temperatures. Phys. Chem. Chem. Phys., 2001. 3: p. 1123-1131.
27. Giorgi, L., et al., H2 and H2/CO oxidation mechanism on Pt/C, Ru/C and Pt-Ru/C
electrocatalysts. J. Appl. Electrochem., 2001. 31: p. 325-334.
28. Ralph, T.R. and M.P. Hogarth, Catalysis for Low Temperature Fuel Cells. Part II: The
Anode Challenges. Platinum Metals Rev., 2002. 46: p. 117-135.
29. Tripkovic, A.V., et al., Study of the kinetics and the influence of Biirr on formic acid
oxidation at Pt2Ru3/C. Electrochim. Acta, 2007. 53: p. 887-893.
30. Gasteiger, H.A., N.M. Markovic, and J. Philip N. Ross, H2 and CO Electrooxidation on
Well-Characterized Pt, Ru, and Pt-Ru. 1. Rotating Disk Electrode Studies of the Pure
Gases Including Temperature Effects. J. Phys. Chem. C, 1995. 99: p. 8290-8301.
31. Sheng, W., H.A. Gasteiger, and Y. Shao-Horn, Hydrogen Oxidation and Evolution
Reaction Kinetics on Platinum: Acid vs Alkaline Electrolytes. J. Electrochem. Soc., 2010.
157: p. B1529-B1536.
32. Tripkovic, A.V., et al., Methanol electrooxidation on supported Pt and PtRu catalysts in
acid and alkaline solutions. Electrochim. Acta, 2002. 47: p. 3707-3714.
33. Watanabe, M. and S. Motoo, Electrocatalysis by ad-atoms. II. Enhancement of the
oxidation of methanol on platinum by ruthenium ad-atoms. J. Electroanal. Chem.
Interfacial Electrochem., 1975. 60: p. 267-273.
144
34. Watanabe, M. and S. Motoo, Electrocatalysis by ad-atoms. III. Enhancement of the
oxidation of carbon monoxide on platinum by ruthenium ad-atoms. J. Electroanal. Chem.
Interfacial Electrochem., 1975. 60: p. 275-283.
35. Ehteshami, S.M.M., et al., The role of electronic properties of Pt and Pt alloys for
enhanced reformate electro-oxidation in polymer electrolyte membrane fuel cells.
Electrochim. Acta, 2013. 107: p. 155-163.
36. Tingelof, T., et al., The influence of CO2, CO and air bleed on the current distribution of
a polymer electrolyte fuel cell. Int. J. Hydrogen Energy, 2008. 33: p. 2064-2072.
37. Bruijn, F.A.d., et al., The influence of carbon dioxide on PEM fuel cell anodes. J. Power
Sources, 2002. 110: p. 117-124.
38. Bellows, R.J., E.P. Marucchi-Soos, and D.T. Buckley, Analysis of Reaction Kinetics for
Carbon Monoxide and Carbon Dioxide on Polycrystalline Platinum Relative to Fuel Cell
Operation. Ind. Eng. Chem. Res., 1996. 35: p. 1235-1242.
39. Giner, J., Electrochemical reduction of CO2 on platinum electrodes in acid solutions.
Electrochim. Acta, 1963. 8: p. 857-865.
40. Smolinka, T., et al., CO2 reduction on Pt electrocatalysts and its impact on H2 oxidation
in CO2 containing fuel cell feed gas – A combined in situ infrared spectroscopy, mass
spectrometry and fuel cell performance study. Electrochim. Acta, 2005. 50: p. 5189-
5199.
41. Ball, S., et al., The Proton Exchange Membrane Fuel Cell Performance of a Carbon
Supported PtMo Catalyst Operating on Reformate. Electrochem. Solid-State Lett., 2002.
5: p. A31-A34.
42. Janssen, G.J.M., Modelling study of CO2 poisoning on PEMFC anodes. J. Power
Sources, 2004. 136: p. 45-54.
43. Urian, R.C., A.F. Gulla, and S. Mukerjee, Electrocatalysis of reformate tolerance in
proton exchange membranes fuel cells: Part I. J. Electroanal. Chem., 2003. 554: p. 307-
324.
145
44. Tuaev, X., et al., In Situ Study of Atomic Structure Transformations of Pt-Ni Nanoparticle
Catalysts during Electrochemical Potential Cycling. ACS Nano, 2013. 7: p. 5666-5674.
45. Stamenkovic, V., et al., Surface segregation effects in electrocatalysis: kinetics of oxygen
reduction reaction on polycrystalline Pt3Ni alloy surfaces. J. Electroanal. Chem., 2003.
554: p. 191-199.
46. Hoffmannov , H., et al., Surface Stability of Pt3Ni Nanoparticulate Alloy
Electrocatalysts in Hydrogen Adsorption. Langmuir, 2013. 29: p. 9046-9050.
47. Li, Q., et al., The CO Poisoning Effect in PEMFCs Operational at Temperatures up to
200°C. J. Electrochem. Soc., 2003. 150: p. A1599-A1605.
48. Dhar, H.P., L.G. Christner, and A.K. Kush, Nature of CO Adsorption during H2
Oxidation in Relation to Modeling for CO Poisoning of a Fuel Cell Anode. J.
Electrochem. Soc., 1984. 134: p. 3021-3026.
146
Chapter 6
Dissertation Summary, Conclusions, and Future Directions
6.1 Summary
Several topics have been introduced and discussed throughout the course of this
dissertation. Although each topic is somewhat unique, the underlying theme of electrocatalytic
selectivity is present. Chapter 2 considered the rational design of catalysts showing high MOR
activity while inhibiting quaternary ammonium poisoning, whereas Chapters 3 and 4 examined
the local structure effects of RuO2 on the selectivity of oxygen and chlorine electrocatalysis, and
finally Chapter 5 considered optimizing the reformate tolerance of Pt catalysts through alloying
with transition metals. All of the topics introduced address various issues encountered in
electrocatalysis, though the primary objective was to understand and develop ways of altering the
selectivity of the electrocatalyst to promote (or inhibit) specific electrochemical processes.
6.2 Chapter Synopses and Future Directions
6.2.1 Chapter 2 - Analysis of Double Layer and Adsorption Effects at the Alkaline Polymer
Electrolyte-Electrode Interface and the Development of a Quaternary Ammonium
Poisoning Resistant Electrocatalyst for Methanol Oxidation
The aim of Chapter 2 was to address the issues that arise when transitioning from an acidic,
PEMFC environment to an alkaline AEMFC. The development of stable and highly conductive
membranes is still a challenge, although the development of electrocatalysts that operate well
with the current state-of-the-art anion-exchange polymers is imperative[1, 2]. Here the issue of
quaternary ammonium poisoning that arises with the use of AEMs was addressed while
147
considering the nature of the catalyst itself. The specific adsorption of quaternary ammonium
species on Pt-based anode catalysts is a serious issue that results in significant performance
losses[3]. This is largely due to a combination of the favorable electrostatic interactions of poorly
solvated quaternary ammonium species with the anode surface and the potential of zero charge
of Pt. Here we introduced the idea of tailoring a highly oxophilic transition metal support to help
mitigate this issue. It was shown that the deposition of oxophilic metals, i.e. Ni and Pb, to the
underlying carbon support can drastically improve the MOR activity while inhibiting the specific
adsorption of quaternary ammonium species on the supported Pt nanoparticles.
The high MOR activity and resistance to quaternary ammonium poisoning of the
Pt/NiPb/C was demonstrated in Chapter 2. A more in-depth analysis of the structure, however, is
so far lacking. In order to truly elucidate the nature of the electrocatalytic enhancements
observed it may be necessary to employ a combined EXAFS and DFT approach. The detailed
structural information obtained from EXAFS would offer valuable insight regarding the
distribution of Ni and Pb in the highly oxophilic support. Currently it is known that the Pb exists
in a hydro-cerrusite phase that is separate from the Pt fcc phase, as evidenced by XRD analysis,
but the location of Ni is uncertain and can only be speculated. Furthermore, the interaction and
distribution of Pt on the NiPb support is still not completely understood. DFT analysis would
offer valuable information regarding the synergystic effects between Ni and Pb in addition to
information regarding the binding energies of quaternary ammonium and MeOH species with the
supported Pt nanoparticles and NiPb support.
148
6.2.2 Chapter 3 - Oxygen Reduction on Nanocrystalline Ruthenia - Local Structure Effects
Chapters 3 and 4 both examine the impact of doping transition metals into the rutile
ruthenia on the selectivity towards electrocatalytic reactions. In the case of Chapter 3, the effect
of doping varying amounts of Ni, Co, or Zn into RuO2 was shown to induce changes on the
selectivity of the ORR. Through the interpretation of electrochemical data and DFT calculations,
it was determined that the incorporation of Ni or Co into the cus active site positions results in a
shift of the ORR selectivity towards the 2-electron reduction pathway to H2O2. By lowering the
binding energy of the *OOH intermediate species, the generation of H2O2 becomes more
thermodynamically favorable at lower overpotentials. In the case of Zn doped ruthenia, the
disruption of the surface to an ilmenite-type facet disrupts the ordered stacking of bridge and cus
sites such that the 4-electron reduction pathway to H2O is favored. Considering the fact that Zn is
itself redox inactive, the Zn-doped material most closely resembles the non-doped RuO2 since
the catalytic activity is confined to Ru active sites in the cus position. The effect of the dopant
metals in the bridge positions, however, was shown to have a negligible effect. This is in contrast
to what is observed for the anodic process of oxygen evolution, where the presence of Ni or Co
in the bridge position was found to have a profound effect[4].
Although the doped ruthenia examined in Chapter 4 are by no means ideal ORR
catalysts, the results of this study can still be extended to other catalytic systems. The ORR is a
widely studied reaction of paramount importance for many electrochemical systems. The
opening and closing of catalytic pathways through the optimization of the local structure is
something to be considered in the rational design of any catalyst. Here it was shown that
introducing disruptions in the ordered stacking of active sites is enough to guide the direction of
the overall reduction process.
149
6.2.3 Chapter 4 - Selective Chlorine Evolution Catalysts Based on Mg-Doped Nanoparticulate
Ruthenium Dioxide
Chapter 4 again dealt with the doped ruthenia. In this case the effects of Mg-doping on
the selectivity of ruthenia towards the oxygen and chlorine evolution reactions was examined. It
was shown through X-ray diffraction that the synthesized materials are single phase and conform
to a tetragonal oxide of the rutile structural type. The use of DEMS and EXAFS was able to
provide valuable information regarding the selectivity with respect to the structural changes
induced with varying Mg-content. The refinement of the EXAFS data was able to confirm that
the Mg is not homogeneously distributed throughout the materials and that Mg-rich clusters are
formed. Furthermore, the Mg resides in a rather strained rutile-type structure at low Mg
concentrations while shifting to an ilmenite-type inclusion at higher Mg concentrations (Mg >
10%). All Mg modified materials are active in the oxygen evolution and chlorine evolution
reactions. Although the Mg containing catalysts show lower overall activities compared with the
non-doped ruthenia, they feature enhanced selectivity toward the chlorine evolution process,
which is attributed primarily to the opening of a reaction pathway for chlorine evolution
associated with presence of Mg modified active sites.
The results of this study offer insight regarding the fundamental nature of the Mg-doped
ruthenia in relation to the local structure. Similar to the doping effects discussed in Chapter 3, the
results here indicate that creating changes in the local structure can modify the selectivity of the
catalyst. Chapter 3 focused more on steering the selectivity of a single reaction towards the
production of the intermediate species. Here it is shown that the competitive nature of two
separate reactions can be controlled by altering the structure of the active sites involved.
150
6.2.4 Chapter 5 - Reformate Tolerant Pt-Based Catalysts for the Electrochemical Hydrogen
Pump
The reformate tolerance of Pt and various Pt-based alloys was examined in Chapter 5. In
this chapter special attention was paid to the dopant metal introduced. Prior studies have
indicated that the introduction of transition metals such as Ni, Co, and Fe improve the CO2
tolerance of Pt. Similarly, prior studies have concluded that the CO tolerance of Pt can be greatly
enhanced through the introduction of oxophilic Ru sites (via bifunctional mechanism) or Mo
sites (via turnover of the Mo4+/6+
redox couple). On this basis, the original goal was to develop a
Pt-based ternary metal catalyst that would take advantage of the bifunctional effect as well as the
electronic effects in order to offer an unprecedented tolerance to reformate gas mixtures
containing both CO and CO2. This concept has been demonstrated previously, though in a H2/O2
PEMFC environment as opposed to a H2/H2 ECHP setting[5]. Throughout this project various
catalysts were synthesized and tested, including PtMoNi/C, PtMoCo/C, PtRuNi/C, PtMoRu/C,
and PtRuCo/C. Of all these attempts, the PtRuNi/C showed the highest hydrogen oxidation
activity in the presence of CO/CO2 and the smallest potential losses. Therefore this was the only
synthesized ternary material discussed in detail in Chapter 5. However, even the reformate
tolerance of PtRuNi/C fell short of that measured for the commercial Tanaka PtRu/C sample.
The structure of the Pt-based alloy is believed to have a significant impact on the
reformate tolerance. Although previous studies have indicated that PtNi/C shows a high
tolerance to CO2, the results of the hydrogen pump cell testing conducted in Chapter 5 show that
PtNi/C has of the lowest CO2 tolerances. A review of the literature indicates that the PtNi
structure is highly dependent upon the preparation of the material, Ni content, the electrode
151
history, and the operating potential[6-8]. Of the catalysts examined, the commercial PtRu/C was
shown to have the highest CO tolerance and the highest tolerance to CO2. Although detailed
structural information is not available for the prepared materials, it can be speculated that the
small particle size and high dispersion of Ru on Pt is responsible for the high reformate tolerance
observed.
It is possible that with more development a Pt-based ternary catalyst could show superior
CO/CO2 tolerance with respect to the PtRu/C sample tested. Although many materials were
synthesized and developed, an in-depth analysis of the catalyst structure was never conducted.
With the aid of XAS it is possible to obtain detailed structural information that could be
correlated with the reformate tolerance. Optimization of the Ni and Ru content and distribution
are necessary to gain the highest catalytic enhancement.
6.3 Concluding Remarks
The development of electrochemical technologies targeting the production, purification,
and consumption of hydrogen (electrolyzers, electrochemical hydrogen pumps, and fuel cells)
are crucial to future of sustainable energy. Electrochemical hydrogen purification is still in its
infancy and has not yet received considerable research interest. With the development of low-
loading, high surface area, reformate tolerant catalysts it is possible that this technology may be
able to compete with the current state-of-the-art hydrogen purification methods. Similarly, AEM-
based direct methanol fuel cells also require a significant amount of research interest and
improvement before commercialization can be considered. We have demonstrated that
quaternary ammonium resistant catalysts can be developed, however, the in situ cell performance
has yet to be demonstrated. In addition, improvements to the ionic conductivity and stability of
152
AEMs still remains a major challenge. Although great strides have been made over the years,
there is still much work to be done before the widespread adoption of electrochemical hydrogen
pumps and AEM-based fuel cells can occur. From an electrocatalysis standpoint, it will be
interesting to see how these technologies develop and progress over the years to come.
153
6.4 References
1. Varcoe, J.R. and R.C.T. Slade, Prospects for alkaline anion-exchange membranes in low
temperature fuel cells. Fuel Cells (Weinheim, Ger.), 2005. 5(2): p. 187-200.
2. Hickner, M.A., A.M. Herring, and E.B. Coughlin, Anion Exchange Membranes: Current
Status and Moving Forward. J. Polym. Sci., Part B: Polym. Phys., 2013. 51: p. 1727-
1735.
3. Unlu, M., et al., Analysis of Double Layer and Adsorption Effects at the Alkaline Polymer
Electrolyte-Electrode Interface. J. Eletrochem. Soc., 2011. 158: p. B1423-B1431.
4. Halck, N.B., et al., Beyond the volcano limitations in electrocatalysis - oxygen evolution
reaction. Phys. Chem. Chem. Phys., 2014. 16: p. 13682.
5. Ehteshami, S.M.M., et al., The role of electronic properties of Pt and Pt alloys for
enhanced reformate electro-oxidation in polymer electrolyte membrane fuel cells.
Electrochim. Acta, 2013. 107: p. 155-163.
6. Tuaev, X., et al., In Situ Study of Atomic Structure Transformations of Pt-Ni Nanoparticle
Catalysts during Electrochemical Potential Cycling. ACS Nano, 2013. 7: p. 5666-5674.
7. Stamenkovic, V., et al., Surface segregation effects in electrocatalysis: kinetics of oxygen
reduction reaction on polycrystalline Pt3Ni alloy surfaces. J. Electroanal. Chem., 2003.
554: p. 191-199.
8. Hoffmannov , H., et al., Surface Stability of Pt3Ni Nanoparticulate Alloy
Electrocatalysts in Hydrogen Adsorption. Langmuir, 2013. 29: p. 9046-9050.