operando fuel cell spectroscopy - northeastern university › files › neu:747 › fulltext.pdf ·...
TRANSCRIPT
1
OPERANDO FUEL CELL SPECTROSCOPY
A dissertation presented
by
Ian Michael Kendrick
to
The Department of Chemistry and Chemical Biology
In partial fulfillment of the requirements for the degree of
Doctor of Philosophy
in the field of
Chemistry
Northeastern University
Boston, Massachusetts
April 24, 2013
2
OPERANDO FUEL CELL SPECTROSCOPY
by
Ian Michael Kendrick
ABSTRACT OF DISSERTATION
Submitted in partial fulfillment of the requirements
for the degree of Doctor of Philosophy in Chemistry
in the College of Science of
Northeastern University
April 24, 2013
3
Abstract
The active state of a catalyst only exists during catalysis(1)
provided the motivation for
developing operando spectroscopic techniques. A polymer electrolyte membrane fuel cell
(PEMFC) was designed to interface with commercially available instruments for acquisition of
infrared spectra of the catalytic surface of the membrane electrode assembly (MEA) during
normal operation. This technique has provided insight of the complex processes occurring at the
electrode surface. Nafion, the solid electrolyte used in most modern-day polymer electrolyte
membrane fuel cells (PEMFC), serves many purposes in fuel cell operation. However, there is
little known of the interface between Nafion and the electrode surface. Previous studies of
complex Stark tuning curves of carbon monoxide on the surface of a platinum electrode were
attributed the co-adsorption of bisulfite ions originating from the 0.5M H2SO4 electrolyte used in
the study(2)
. Similar tuning curves obtained on a fuel cell MEA despite the absence of
supplemental electrolytes suggest the adsorption of Nafion onto platinum(3)
. The correlation of
spectra obtained using attenuated total reflectance spectroscopy (ATR) and polarization
modulated IR reflection-absorption spectroscopy (PM-IRRAS) to a theoretical spectrum
generated using density functional theory (DFT) lead to development of a model of Nafion and
platinum interaction which identified participation of the SO3- and CF3 groups in Nafion
adsorption.
The use of ethanol as a fuel stream in proton exchange membrane fuel cells provides a
promising alternative to methanol. Relative to methanol, ethanol has a greater energy density,
lower toxicity and can be made from the fermentation of biomass(4)
. Operando IR spectroscopy
was used to study the oxidation pathway of ethanol and Stark tuning behavior of carbon
4
monoxide on Pt, Ru, and PtRu electrodes. Potential dependent products such as acetaldehyde,
acetic acid and carbon monoxide are identified as well as previously unobserved peaks
corresponding to adsorbed ethanol.
A modification to the operando fuel cell design allowed for acquisition of Raman spectra.
A confocal Raman microscope enabled characterization of the MEA through depth profiling.
The potential dependent peaks of an Fe-Nx/C catalyst were identified and compared to the
theoretical spectra of the proposed active sites. It was determined that oxygen adsorbed onto
iron/iron oxide carbon nanostructures were responsible for the experimentally obtained peaks.
This finding was supported by additional Raman studies carried out on a catalyst with these
active sites removed through peroxide treatments.
1. Topsoe, H., Developments in operando studies and in situ characterization of heterogeneous catalysts. Journal of Catalysis, 2003. 216(1-2): p. 155-164.
2. Stamenkovic, V., et al., Vibrational properties of CO at the Pt(111)-solution interface: the anomalous stark-tuning slope. Journal of Physical Chemistry B, 2005. 109(2): p. 678-680.
3. Kendrick, I., et al., Elucidating the Ionomer-Electrified Metal Interface. J. Am. Chem. Soc., 2010. 132(49): p. 17611-17616.
4. Lamy, C. and Leger, J.M., FUEL-CELLS - APPLICATION TO ELECTRIC VEHICLES. Journal De Physique Iv, 1994. 4(C1): p. 253-281.
5
Acknowledgements
I am extremely grateful for the opportunity to have been invited to pursue my graduate degree
at Northeastern University. I could not have gone this far alone and wish to acknowledge the
help and support I have received over the years.
My advisor Dr. Eugene Smotkin for his constant support, advice and for pushing me further
than I ever thought possible.
Graham Jones for all of the opportunities have gave me during the first half of my graduate
career.
My thesis committee: Drs. Graham Jones, Sanjeev Mukerjee and Max Deim.
Lab members past and present especially: Sara Evarts, Adam Yakaboski, Jonathan Doan,
Mike Bates and Mike Finch. Thank you for all of the support and friendship.
I am fortunate to have made many wonderful friends that I have made here at Northeastern,
especially Meghan Johnston. Thank you for reminding me that work and research isn’t
everything.
Maiann Good for her constant support and understanding.
6
Table of contents
Abstract ........................................................................................................................................... 2
Acknowledgements ......................................................................................................................... 5
Table of contents ............................................................................................................................. 6
List of Figures ................................................................................................................................. 7
List of Tables ................................................................................................................................ 10
List of Schemes ............................................................................................................................. 11
List of Abbreviations .................................................................................................................... 12
Chapter 1: Introduction ................................................................................................................. 14
Chapter 2: Complex Stark Tuning of CO ..................................................................................... 26
Chapter 3: Elucidating the ionomer-metal interface ..................................................................... 35
Chapter 4: Minimal Vibrational Mode Analysis of Nafion Infrared Spectroscopy ..................... 51
Chapter 5: Operando Infrared Spectroscopy of Ethanol Oxidation in Polymer Electrolyte Fuel
Cells .............................................................................................................................................. 68
Chapter 6: Operando Raman Spectroscopy of a non-Pt Cathode Nafion Membrane Electrode .. 87
7
List of Figures Figure 1.1 PEM fuel cell schematic .............................................................................................. 14
Figure 1.2. Schematic of operando IR cell developed by Fan et al .............................................. 17
Figure 1.3. Operando XAS spectroscopy fuel cell and schematic (from ref. 28) ......................... 17
Figure 1.4. Operando IR-XAS fuel cell schematic. ...................................................................... 18
Figure 1.5. Cell installed in a diffuse reflectance accessory ......................................................... 19
Figure 1.6. IR-XAS cyclic voltammogram. ................................................................................. 20
Figure 1.7. IR-XAS fuel cell polarization curve. ......................................................................... 20
Figure 2.1. Potential-dependent spectra of CO on Pt. .................................................................. 29
Figure 2.2. Stark tuning plots of linearly adsorbed CO adsorbed on Pt at 50 °C. ....................... 29
Figure 2.3. CO electro-oxidation onset potentials versus adsorption potentials .......................... 30
Figure 2.4. Operando CO/Pt Stark tuning ..................................................................................... 31
Figure 2.5. The potential at which CO begins to oxidize as a function of temperature. ............. 32
Figure 3.1. DFT calculated normal modes and Nafion ATR spectrum ....................................... 40
Figure 3.2. Theoretical and experimental spectra of Nafion ....................................................... 41
Figure 3.3. Normal mode coordinate animation snapshots of the Nafion side-chain anion and
backbone fragment ........................................................................................................................ 44
Figure 3.4. Gaussian 03 Viewer Nafion-Pt interface model ........................................................ 48
Figure 4.1. Flow chart for generation of Nafion repeat group MVM spectra. ............................. 55
Figure 4.2. Definition of the internal coordinates for a) C3v, b) C2v, and c) the Nafion Backbone.
....................................................................................................................................................... 58
Figure 4.3. Top: DFT calculated normal modes decomposed into MVM spectra. Composite
DFT lines show contributions of minimal vibrational modes to each normal mode. ................... 60
Figure 4.4: The ATR-IR spectrum of Teflon superimposed over Nafion backbone MVMs. ...... 61
8
Figure 4.5: ATR spectrum of PFMHSA (left) and PFEBSA (right) superimposed over SO3- νas,
SO3- νs, CF3 νas, CF3 νs MVMs. ..................................................................................................... 61
Figure 5.1. Stark tuning curve of CO adsorbed onto Pt and potential dependent spectra of a Pt
black electrode in the presence of ethanol vapor. ......................................................................... 72
Figure 5.2. Stark tuning curve of CO adsorbed onto PtRu potential dependent spectra of an
unsupported PtRu electrode in the presence of ethanol vapor. ..................................................... 73
Figure 5.3. A general mechanism for the oxidation of ethanol on Pt. ......................................... 75
Figure 5.4. Potential dependent spectra of a Pt black electrode in the presence of ethanol vapor.
....................................................................................................................................................... 76
Figure 5.5. Potential dependent spectra of a Ru black electrode in the presence of ethanol vapor.
....................................................................................................................................................... 79
Figure 5.6. Potential dependent spectra of a PtRu black electrode in the presence of ethanol
vapor ............................................................................................................................................. 81
Figure 6.1. Exploded view of operando Raman cell components. ............................................... 90
Figure 6.2. Operando Raman cell beneath a confocal Raman microscope. ................................ 90
Figure 6.3: Total free energy, in Hartrees, for the hypothetical Fe-Nx/C active site at various spin
states. ............................................................................................................................................. 92
Figure 6.4. Schematic outlining the concept of depth profiling a membrane electrode assembly
using a confocal Raman microscope. Depth dependent spectra of a membrane electrode
assembly consisting of an Fe/N/C catalyst and Nafion.. .............................................................. 94
Figure 6.5. The potential dependent Raman spectra of an iron based non-PGM cathode catalyst
obtained under O2. ........................................................................................................................ 96
Figure 6.6. Raman fuel cell polarization curve obtained under oxygen with a MEA consisting of
a Pt anode and an Fe-Nx/C cathode............................................................................................... 97
Figure 6.7. Potential dependent control experiments of a non-PGM cathode catalyst. (Left):
Spectra obtained under O2 with a catalyst prepared without iron. (Right): Spectra obtained under
N2 catalyst prepared with iron....................................................................................................... 98
Figure 6.8. DFT generated theoretical spectra of an iron based non-PGM catalyst with a N4
pyridinic active site. . ................................................................................................................. 100
Figure 6.9. DFT generated theoretical spectra of an iron based non-PGM catalyst with a N4
pyrrolic active site. ...................................................................................................................... 100
9
Figure 6.10. DFT generated theoretical spectra of an iron nanoparticle with and without adsorbed
oxygen ......................................................................................................................................... 103
Figure 6.11. Potential dependent Raman spectra of non-PGM catalyst treated with H2O2 to
remove nanoparticles. Spectra were obtained under presence of oxygen. ................................ 104
Figure 6.12. Fuel cell polarization curve obtained under oxygen with a MEA consisting of a Pt
anode and an Fe-Nx/C cathode subjected to the structural distortion treatment. ........................ 105
10
List of Tables
Table 3.1. PM-IRRAS and DFT IR adsorption peaks and assignments. ………………………45
Table 3.2. Average partial charges of selected Nafion segments. ……………………………...47
Table 4.1. Internal coordinate system MVMs of the most prominent IR modes for Nafion. …...57
Table 4.2 Pure mode assignments from visualization of normal mode animations in comparison
to MVM assignments of this work. ……………………………………………………………..63
11
List of Schemes Scheme 3.1. Segment and atom labeling for Nafion. …………………………………………42
Scheme 4.1: Nafion chemical repeat unit. ……………………………….……………………53
12
List of Abbreviations
ATR attenuated total reflection
CCM catalyst coated membrane
CO carbon monoxide
CV cyclic voltammogram
DEMS differential electrochemical mass spectroscopy
DFT density functional theory
DMFC direct methanol fuel cell
FTIR Fourier transform infrared spectroscopy
IR infrared
GDL gas diffusion layer
MEA membrane electrode assembly
MVM minimal vibrational mode
NBO natural bond order
NHE normal hydrogen electrode
Non-PGM non-platinum group metal
ORR oxygen reduction reaction
PEMFC proton exchange membrane fuel cell
PFEBSA perfluoro(2-ethoxybutane) sulfonic acid
PFMHSA perfluoro(3-methyl-2,4-dioxahexane) sulfonic acid
PM-IRRAS polarization modulated infrared reflection-absorption spectroscopy
Psig pounds per square inch (gauge)
RHE reversible hydrogen electrode
13
Sccm standard cubic centimeter per minute
TR transmission spectrum
XAS X-ray absorption spectroscopy
14
Chapter 1: Introduction
That the active state of a catalyst exists only during catalysis(1) is succinct rationale for
operando methods of catalyst characterization. The primary challenge to operando
spectroscopy is conversion of a practical device into a spectroscopic cell with minimal
perturbation of device functionality. Figure 1.1 schematizes the terminal end-cell of a fuel cell
stack.(2)
Figure 1.1 PEM fuel cell schematic (From ref. 2)
Graphite flow-field plates distribute fuel and oxidant to the 5-layer membrane electrode
assembly (MEA). An ionomer membrane (e.g., Nafion) supports electrocatalytic layers that
contact gas diffusion layers (i.e., porous carbon paper or cloth) that are optimized for reactant
transport and electronic conductivity. MEA fabrication methods have been reviewed.(3)
Catalyst particles (carbon supported or metal blacks) (4) are dispersed in alcoholic solutions of
solubilized ionomer. These “inks” are deposited onto the gas diffusion layers and hot pressed to
the membrane. Alternatively, a catalyst-coated-membrane (CCM) can be prepared by
15
immobilizing the membrane on a heated vacuum table (NuVant Systems Inc., Crown Point, IN)
for direct ink deposition onto the membrane. Catalytic layers are a complex blend of ionomer,
catalyst particles and, sometimes at the cathode, Teflon dispersion. The relative amounts depend
on the catalyst composition and device application. The final step of MEA preparation occurs in
the operating fuel cell. The hot pressing of the gas diffusion electrode and/or ink drying on the
heated vacuum table causes delamination of the ionomer from the catalyst particles. The
conditioning of the MEA in the operating fuel cell re-wets the catalyst with ionomer and
removes the deep oxides that are typical of as-prepared catalysts.(5) There are no intended
“triple phases”. The catalyst active area must be coated with a sub-micron gas permeable, ion
conducting layer.(6) A Nafion layer on Pt has been shown to enhance electrocatalysis.(6, 7)
Ionomer electrolytes have no mobile ions other than protons or hydroxide ions. Supplemental
electrolytes such as aqueous H2SO4 or HClO4 contribute mobile ions that competitively adsorb
onto the surface.(8-10) Supplemental aqueous electrolytes preclude fuel cell operation at the high
end of relevant temperatures (e.g. 70-120 oC).
Operando spectroscopy requires control of the temperature, flow rate and the humidity of
the anode and cathode reactant streams while potential dependent spectra are acquired. Cell
component materials must not fluoresce at energies similar to the X-ray edge energies of the
catalysts. The uniformity of the polymer electrolyte resistance, governed by the ionomer
membrane thickness (e.g., 7 mil for Nafion 117)(11) depends on careful water management and
proper flow field design. Whether studying anode or cathode catalysts, the counter electrode can
serve as the auxiliary and the reference electrode (counter-reference electrode).(12) Gurau
delivered hydrogen to the counter-reference electrode while acquiring liquid feed direct
methanol fuel cell anode polarization curves.(13) Pure water can be delivered to the counter-
16
reference electrode: Hydrogen evolution at the counter electrode and the hydrogen ion activity,
set by the ionomer equivalent weight and state of hydration, poises(14) the counter-reference
electrode. Although the counter-reference is polarized at high currents, the alternative of
developing a 3rd
electrode as a reference is more complex than correcting for reference electrode
polarization. Fortunately, the exchange current density for the hydrogen electrode is many orders
of magnitude larger than that for the oxygen reduction reaction.(15) On a practical level, the
use of the fuel cell counter-reference affords greater reproducibility between laboratories.
Operando infrared spectroscopy of the catalyst-electrolyte interface is ideal for adsorbate
characterization. Adsorbates effect dynamic changes on the catalyst surface such as surface
restructuring where surface atoms can be entirely displaced in order to make stronger surface-
adsorbate bonds.(16) These changes can be observed, in operando and in situ spectroscopy, as
changes in the vibrational modes of either the surface or the adsorbate atoms. Fan introduced
operando fuel cell spectroscopy in 1996.(17, 18) He incorporated a CaF2 window on a fuel cell
flow field, cut a slot into the GDL to expose the catalytic layer, and installed the modified fuel
cell into a 1990’s model Harrick Praying Mantis (Pleasantville, NY) diffuse reflection IR
accessory. Figure 1.2 schematizes the cell used to identify methanol oxidation products,
including formic acid, in spectra acquired over a range of temperatures from 25 - 90oC with
steady state reactant flow to both electrodes. Bo used the same cell to measure Stark tuning
rates(19) of CO/Pt at 50oC .(20)
17
Figure 1.2. Schematic of operando IR cell developed by Fan et al
Viswananthan(21) and Stoupin(22) introduced hydrogen and liquid feed direct methanol fuel
cell (DMFC) operando X-ray absorption spectroscopy respectively using the cell of Fig. 1.5.
They end-milled a rectangular core out of both flow field blocks and endplates. Palladium was
used at the cathode to mitigate interference with the absorption edge of the Pt based anode
catalysts. The “cored” cell design has since been used by a number of workers.(23-27)
Figure 1.3. Operando XAS spectroscopy fuel cell and schematic (from ref. 28)
A number of operando XAS(28-35) studies followed that of Viswananthan with many
focused on the oxidation state of the metal components. All found that at relevant fuel cell anode
operating potentials, platinum is metallic. The oxidation of CO was studied both on Pt(32), and
CO, D2O
inlet
D2, D2O
heating pad
Nafion
membrane
Exhaust
CaF2
Incident IR Integrating sphere
Exhaust
PtRu
Pt
CO, D2O
inlet
D2, D2O
heating pad
Nafion
membrane
Exhaust
CaF2
Incident IR Integrating sphere
Exhaust
PtRu
Pt
CO, D2O
inlet
D2, D2O
heating pad
Nafion
membrane
Exhaust
CaF2
Incident IR Integrating sphere
Exhaust
PtRu
Pt
CO, D2O
inlet
CO, D2O
inlet
D2, D2O
heating pad
Nafion
membrane
Exhaust
CaF2
Incident IR Integrating sphere
Exhaust
PtRu
Pt
CO, D2O
inlet
D2, D2O
heating pad
Nafion
membrane
Exhaust
CaF2
Incident IR Integrating sphere
Exhaust
PtRu
Pt
18
PtRu,(28, 31, 36, 37) confirming the electronic benefits of Ru as a co-catalyst for CO oxidation.
Studies of adsorbed oxygen reduction reaction intermediates on fuel cell catalysts show that
adsorption is both potential-dependent and site-specific.(28, 38-41)
The Viswananthan cell was designed for transmission X-ray absorption measurements.
Fluorescence measurements provide better signal-to-noise because interference due to cell
components is minimized. The most recent cell design (Fig. 1.8) combines features of the
Viswananthan cell(21) with the specular reflectance infrared cell reported by Fan et al.(18) The
IR-XAS cell enables operando XAS (in transmission and fluorescence) and specular reflectance
FTIR spectroscopy. The top flow field accommodates a CaF2 window for IR reflectance
studies.(20, 42) A pin-style upper flow field optimizes flow distribution around the CaF2
window inset. The CaF2 window can be removed when the working electrode of interest is an
air-breathing electrode.
Figure 1.4. Left: IR-XAS cell with DE9 connector for electrodes, heater cartridge and thermister.
Right: 1) X-ray transmission, 2) Reflectance IR or fluorescence X-ray, 3) CaF2 window housing,
4) Teflon gasket, 5) gas outlet insert, 6) slider assembly, 7) thermocouple/heater cartridge port,
8) lower flow field, 9) MEA, 10) upper flow field, 11) top plate.
19
The IR-XAS cell slides into a Pike Technologies (Watertown, WI) diffuse reflectance
accessory (with a modified front plate) that is accommodated by most commercial FTIR
instruments (Fig. 1.9).
Figure 1.5. Cell installed in Pike Technologies (Watertown, WI) diffuse reflectance accessory in
a Bruker Optics (Billerica, MA) Vertex 70 FTIR Spectrometer.
A slot under the slider assembly ensures precise positioning of the cell under the Diffuse-IR
accessory integrating mirrors. Figure 1.10 shows background CV (solid) and the CO stripping
wave (dotted) obtained at 50 oC. The CO stripping wave (10 mV/sec) extends from 600 - 900
mV. Figure 1.11 is a performance curve of a Pt/Pt electrode
20
Figure 1.6. Cyclic voltammetry (10mV/s) of fuel cell working electrode (5-cm2 geometric)
under humidified nitrogen at 50°C (solid). Humidified H2 (50 sccm) at the counter-reference.
CO stripping wave (dashed).
Figure 1.7. IR-XAS fuel cell polarization curve.
0.05 0.00 -0.05 -0.10 -0.15 -0.20 -0.25 -0.30 -0.35
0.2
0.4
0.6
0.8
1.0
Po
ten
tial (m
V)
Current (A)
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
-0.02
-0.01
0.00
0.01
0.02
0.03
0.04
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
-0.02
-0.01
0.00
0.01
0.02
0.03
0.04
Curr
ent
(mA
)
Potential (mV)
21
Here the first fully operando FTIR and X-ray absorption studies of fuel cells is described.
The features of these legacy cells are then combined into a single cell that can accommodate both
FTIR and X-ray absorption spectroscopy of operating fuel cells. Operando FTIR spectroscopy,
complemented with density functional theory (DFT) and polarization modulated infrared
reflection absorption spectroscopy (PM-IRRAS) of ionomer-platinum interfaces, suggests a
model for self-assembly of Nafion onto a Pt surface. The assignment of bands responsible
Nafion’s self-assembly onto Pt motivated the development of an automated method to identify
functional group contributions to theoretical vibrational modes. Separating these contributions
to into their own spectra enables the prediction of changes in experimentally obtained spectra of
Nafion caused by derivitization, state-of-hydration and ion exchange.
The utility of operando FTIR spectroscopy is exemplified with the study of electro-
oxidation of ethanol. The oxidation pathways of ethanol on Pt, Ru and PtRu electrodes are
elucidated. The bipolar nature of peaks associated with adsorbed species enables differentiation
from desorbed species without the use of a polarized light source allows for a comprehensive
assessment of oxidation processes with a single technique. Previous in-situ FTIR studies of
ethanol oxidation use HClO4 as an electrolyte yielding an intense peak at 1033 cm-1
. Because
operando spectra are acquired without mobile anions, a peak associated with O-adsorbed ethanol
is observed for the first time. The oxidation mechanisms elucidated through operando
spectroscopy is consistent with the established mechanisms and has been used to characterize
ethanol oxidation on experimental catalysts.
A simple modification to the design of the slider assembly enabled the collection of
Raman spectra. The use of a confocal Raman microscope allowed for a complete
characterization of the MEA. This new technique was used to identify potential dependent
22
features of the Raman spectra of the surface of iron/nitrogen based supported oxygen reduction
catalysts. The exact nature of the active site and mechanism of these catalysts is still disputed.
The latest theory is that in acidic media, the ORR proceeds via a two-step mechanism: a Fe-N4/C
moiety reduces oxygen to hydrogen peroxide and peroxide is reduced to water on a second active
site. Recent studies suggest that the active site consists of a metal/metal oxide cluster.(44) DFT
was used to generate theoretical spectra of two proposed Fe-N4/C active sites and one Fe
nanocluster. Theoretical spectra were generated for these models both in the presence and
absence of adsorbed O2. The position of the potential dependent features of the Raman spectra
matches the position of Fe-O stretching on the Fe nanocluster theoretical spectrum. The
potential dependent peaks are missing in the Raman spectra Fe-Nx/C catalysts treated to remove
the Fe/FeOx moieties.
References:
1. Topsoe, H., Developments in operando studies and in situ characterization of heterogeneous catalysts. Journal of Catalysis, 2003. 216(1-2): p. 155-164.
2. Smotkin, E.S., FTIR and X-Ray Absorption Spectroscopy of Operating Fuel Cells, in In-Situ Spectroscopic Studies of Adsorption at the Electrode and Electrocatalysis, S.-G. Sun, P.A. Christensen, and A. Wieckowski, Editors. 2007, Elsevier: Oxford, UK. p. 247-272.
3. Gottesfeld, S. and Zawodzinski, T.A., in Polymer Electrolyte Fuel Cells, R.C. Alkire and D.M.K.a.C.W.T. H. Gerischer, Editors. 1997, Wiley-VCH. p. 229-244.
4. Liu, L., et al., Carbon supported and unsupported Pt-Ru anodes for liquid feed direct methanol fuel cells. Electrochimica Acta, 1998. 43(24): p. 3657-3663.
5. Viswanathan, R., et al., In-Situ XANES of Carbon-Supported Pt-Ru Anode Electrocatalyst for Reformate-Air Polymer Electrolyte Fuel Cells. Journal of Physical Chemistry B, 2002. 106(13): p. 3458-3465.
6. Liu, L., et al., Methanol oxidation on nafion spin-coated polycrystalline platinum and platinum alloys. Electrochemical and Solid State Letters, 1998. 1(3): p. 123-125.
7. Ploense, L., et al., Spectroscopic study of NEMCA promoted alkene isomerizations at PEM fuel cell Pd-Nafion cathodes. Solid State Ionics, 2000. 136-137: p. 713-720.
23
8. Stamenkovic, V., et al., Vibrational properties of CO at the Pt(111)-solution interface: the anomalous stark-tuning slope. Journal of Physical Chemistry B, 2005. 109(2): p. 678-680.
9. Kendrick, I., et al., Elucidating the Ionomer-Electrified Metal Interface. J. Am. Chem. Soc., 2010. 132(49): p. 17611-17616.
10. Lewis, E.A., et al., Operando X-ray absorption and infrared fuel cell spectroscopy. Electrochimica Acta, (0).
11. The first two digits are the equivalent weight of Nafion divided by 100; the last digit is the thickness of Nafion in thousandths of an inch.
12. Rivera, H., et al., Effect of sorbed methanol, current, and temperature on multicomponent transport in nation-based direct methanol fuel cells. Journal of Physical Chemistry B, 2008. 112(29): p. 8542-8548.
13. Gurau, B. and Smotkin, E., Methanol crossover in direct methanol fuel cells: a link between power and energy density. J. Power Sources, 2002. 112: p. 339-352.
14. Bard, A. and Faulkner, L., Electrochemical Methods: Fundamentals and Applications. 2001: John WIley & Sons.
15. Bard, A.J., Electrochemical methods : fundamentals and applications / Allen J. Bard, Larry R. Faulkner, ed. L.R. Faulkner. 1980, New York :: Wiley.
16. Somorjai, G.A., Introduction to Surface Chemistry and Catalysis. 1994, New York, NY: John Wiley & Sons, Inc.
17. Fan, Q., et al., In situ FTIR-diffuse reflection spectroscopy of the anode surface in a direct methanol/oxygen fuel cell. Journal of the Electrochemical Society, 1996. 143(2): p. L21-L23.
18. Fan, Q.B., Pu, C., and Smotkin, E.S., Ins situ Fourier transform infrared-diffuse reflection spectroscopy of direct methanol fuel cell anodes and cathodes. Journal Of The Electrochemical Society, 1996. 143(10): p. 3053-3057.
19. Chang, S.C. and Weaver, M.J., In-Situ Infrared-Spectroscopy at Single-Crystal Metal-Electrodes - An Emerging linke between Electrochemical and Ultrahigh-Vacuum Surface Science. Journal of Physical Chemistry, 1991. 95(14): p. 5391-5400.
20. Bo, A.L., et al., In situ Stark effects with inverted bipolar peaks for adsorbed CO on Pt electrodes in 50 degrees C direct methanol fuel cells. Journal Of Physical Chemistry B, 2000. 104(31): p. 7377-7381.
21. Viswanathan, R., Liu, R., and Smotkin, E.S., In situ x-ray absorption fuel cell. Review of Scientific Instruments, 2002. 73(5): p. 2124-2127.
22. Stoupin, S., et al., Pt and Ru X-ray Absorption Spectroscopy of PtRu Anode Catalysts in Operating Direct Methanol Fuel Cells. Journal of Physical Chemistry B, 2006. 110(20): p. 9932-9938.
23. Russell, A.E. and Rose, A., X-ray absorption Spectroscopy of low temperature fuel cell catalysts. Chemical Reviews, 2004. 104(10): p. 4613-4635.
24. Witkowska, A., et al., Pt–Co cathode electrocatalyst behaviour viewed by in situ XAFS fuel cell measurements. Journal of Power Sources, 2008. 178(2): p. 603-609.
25. Principi, E., et al., An XAS experimental approach to study low Pt content electrocatalysts operating in PEM fuel cells. Phys Chem Chem Phys, 2009. 11(43): p. 9987-95.
24
26. Roth, C., et al., Pt-Ru fuel cell catalysts subjected to H2, CO, N2 and air atmosphere: An X-ray absorption study. Applied Catalysis, A: General, 2007. 319: p. 81-90.
27. Tada, M., et al., In situ time-resolved dynamic surface events on the Pt/C cathode in a fuel cell under operando conditions. Angewandte Chemie-International Edition, 2007. 46(23): p. 4310-4315.
28. Roth, C., et al., Determination of O[H] and CO Coverage and Adsorption Sites on PtRu Electrodes in an Operating PEM Fuel Cell. J. Am. Chem. Soc., 2005. 127(42): p. 14607-14615.
29. Wiltshire, R.J.K., et al., A PEM fuel cell for in situ XAS studies. Electrochimica Acta, 2005. 50(25-26): p. 5208-5217.
30. Roth, C., et al., In-situ XAFS fuel cell measurements of a carbon-supported Pt-Ru anode electrocatalyst in hydrogen and direct methanol operation. Physical Chemistry Chemical Physics, 2002. 4(15): p. 3555-3557.
31. Scott, F.J., Roth, C., and Ramaker, D.E., Kinetics of CO Poisoning in Simulated Reformate and Effect of Ru Island Morphology on PtRu Fuel Cell Catalysts As Determined by Operando X-ray Absorption Near Edge Spectroscopy. J. Phys. Chem. C, 2007. 111(30): p. 11403-11413.
32. Maniguet, S., Mathew, R.J., and Russell, A.E., EXAFS of carbon monoxide oxidation on supported Pt fuel cell electrocatalysts. Journal of Physical Chemistry B, 2000. 104(9): p. 1998-2004.
33. Singh, J., et al., In situ XAS with high-energy resolution: The changing structure of platinum during the oxidation of carbon monoxide. Catalysis Today, 2009. 145(3-4): p. 300-306.
34. Ramaker, D.E. and Koningsberger, D.C., The atomic AXAFS and [capital Delta][small mu ] XANES techniques as applied to heterogeneous catalysis and electrocatalysis. Physical Chemistry Chemical Physics, 2010. 12(21): p. 5514-5534.
35. Croze, V., et al., The use of in situ X-ray absorption spectroscopy in applied fuel cell research. Journal of Applied Electrochemistry, 2010. 40(5): p. 877-883.
36. Russell, A.E., et al., In situ X-ray absorption spectroscopy and X-ray diffraction of fuel cell electrocatalysts. Journal of Power Sources, 2001. 96(1): p. 226-232.
37. Scott, F.J., Mukerjee, S., and Ramaker, D.E., CO Coverage/Oxidation Correlated with PtRu Electrocatalyst Particle Morphology in 0.3 M Methanol by In Situ XAS. Journal of The Electrochemical Society, 2007. 154(5): p. A396-A406.
38. Teliska, M., O'Grady, W.E., and Ramaker, D.E., Determination of H Adsorption Sites on Pt/C Electrodes in HClO4 from Pt L23 X-ray Absorption Spectroscopy. J. Phys. Chem. B, 2004. 108(7): p. 2333-2344.
39. Teliska, M., et al., Site-Specific vs Specific Adsorption of Anions on Pt and Pt-Based Alloys. J. Phys. Chem. C, 2007. 111(26): p. 9267-9274.
40. Teliska, M., et al., In situ determination of O(H) adsorption sites on Pt based alloy electrodes using X-ray absorption spectroscopy. Proceedings - Electrochemical Society, 2005. 2003-30(Fundamental Understanding of Electrode Processes): p. 212-216.
25
41. Roth, C. and Ramaker, D.E., 3 XAS Investigations of PEM Fuel Cells, in Interfacial Phenomena in Electrocatalysis, C.G. Vayenas, Editor. 2011, Springer New York. p. 159-201.
42. Gasteiger, H.A., et al., Activity benchmarks and requirements for Pt, Pt-alloy, and non-Pt oxygen reduction catalysts for PEMFCs. Applied Catalysis B: Environmental, 2005. 56(1-2): p. 9-35.
43. FTIR studies require a slot cut into the GDL to enable beam access to the catalytic surface.
44. Olson, T.S., et al., Bifunctional Oxygen Reduction Reaction Mechanism on Non-Platinum Catalysts Derived from Pyrolyzed Porphyrins. Journal of the Electrochemical Society, 2010. 157(1): p. B54-B63.
26
Chapter 2: Complex Stark Tuning of CO
2.1 Introduction The chemisorption of CO on a metal surface results from electron transfer from the 5σ
orbital of CO to the metal. Additionally, electrons from dz metal orbital back donate in the CO
2π* antibonding orbital of CO.(1) The degree of back donation can be modulated when the
metal surface is used as an electrode. The adsorption of CO onto a metal surface can be directly
studied through in-situ infrared spectroscopy.(2-4) At more positive potentials, the bond order is
increased due to 2π* antibonding orbital being less occupied. This results in a shift in the peak
position known as Stark tuning. The peak position of adsorbed CO provides information on
adlayer structures.(5-7) In this study, the Start tuning of CO peaks adsorbed onto a Pt electrode
will be examined in a working fuel cell. The complex behavior elucidates co-adsorption of the
polymer electrolyte and enthalpies of adsorption will be correlated to adsorption potentials.
2.2 Experimental
Membrane electrode assembly preparation: Nafion-117 (E. I. DuPont) was immersed in
boiling ~8 M nitric acid for 20 min, rinsed with Nanopure™ water, and finally immersed in
boiling water for one hr. Catalyst inks were prepared as previously described.(8, 9) Briefly, Pt
black (Johnson Matthey) was dispersed in 5 wt% Nafion ionomer solution (Sigma Aldrich,
Milwaukee, WI) diluted with Nanopure™ water and isopropanol. Inks were applied directly to a
5 cm2 area of Nafion immobilized on a temperature controlled vacuum table (NuVant Systems
Inc., Crown Point, IN) at 70° C. The catalyst loadings were 4mg/cm2 of Pt black. The carbon
paper gas diffusion layers (Toray Industries, Tokyo, Japan) were blocked with Vulcan XC-72
(Cabot Corporation, Billerica, MA).
27
Operando IR spectroscopy: Spectra from 800-4000 cm-1
, at 4 cm-1
resolution, were acquired
with a Vertex 70 Spectrometer (Bruker Optics, Billerica, MA) and analyzed with Opus 6.5™
software. The potential dependent IR specular reflectance spectra of adsorbed CO (COads) was
obtained for CO dosing potentials (Eads) of 100 to 400 mV vs. RHE in 100 mV increments. Prior
to acquisition of the background IR spectrum, the cell was brought to 50 °C with H2 (50 sccm) at
the counter-reference electrode and N2 at the working electrode (200 sccm) for 15 min. The
working electrode feed was switched to CO (40 sccm) at the selected adsorption potential for 15
min prior to purging the working electrode with N2 (200 sccm). The potential was set to 100
mV prior to acquisition of four signal-averaged (250 scans) spectra, at 50mV increments, until
the CO vibrational bands were no longer observable.
2.3 Results and discussion
Potential dependent spectra of CO adsorbed at 100 mV vs. NHE are shown in Figure 2.1. The
corresponding Stark tuning plots for CO adsorbed at 100, 200, 300 and 400 mV (Fig. 2.2) show
the effects of co-adsorption on the CO stretching frequencies. These results are similar to those
of Stamenkovic et al. studying CO on Pt(111) in 0.5 M H2SO4.(5) Stamenkovic correlated the
complex Stark tuning of COads stretching frequencies (νCO) to the compression/dissipation of
COads islands: A linear Stark tuning region with a subtle νCO blue shift from the extrapolated
linear region was followed by a precipitous drop in νCO, followed by an upturn. The potential
where νCO precipitously drops (Eonset) correlates with COads oxidation. Stamenkovic identified the
co-adsorbate as HSO3- originating from the 0.5 M H2SO4 electrolyte used in their study.
Stamenkovic attributed the blue shift from the linear region to the oxidation of a small amount of
adsorbed CO caused by the presence of activated water adsorbed onto defect sites. Kendrick,
using no supplemental electrolyte (i.e., operando conditions), identified co-adsorbates as the
28
Nafion sulfonate exchange group and the side chain CF3 group. COads oxidation, induced by OH-
adsorption, diminishes dipole-dipole coupling and thus precipitously decreases νCO. The upturn
is ascribed to increased adsorption of sulfonate species relative to OH-, which reestablishes
repulsive dipole interactions that compress COads islands and increases νCO.
The Stark tuning rates of this work, 6.8 0.6 cm-1
/mV, are within the range of 2.5 cm-1
/V to
18.3 cm-1
/V of previous operando Stark tuning studies of direct methanol fuel cells at potentials
negative of 0.5 V vs. NHE.42
The wavelength precision of an FTIR is determined by the stability
of the reference HeNe laser. Virtually all FTIR spectrometers manufactured today are capable of
0.1 cm-1
or better precision on the wavelength axis. The Bruker Vertex 80V is a research
spectrometer that is routinely precise to better than 0.07 cm-1. This does not mean that spectral
features are resolved to this extent, but it does mean that band positions can reliably be reported
to this level of precision.
2200 2175 2150 2125 2100 2075 2050 2025 2000
Wavenumber (cm-1)
100 mV
150 mV
200 mV
250 mV
300 mV
350 mV
400 mV
450 mV
500 mV
29
Figure 2.1. Potential-dependent spectra of CO on Pt; Adsorption potential of 100 mV operating
at 50 °C.
Figure 2.2. Stark tuning plots of linearly adsorbed CO adsorbed on Pt at 50 °C.
Figure 2.3 shows how Eonset dependends on Eads. At 400 mV, Eonset and Eads coincide because at
higher potentials COads favors adsorption sites with higher enthalpies of adsorption (e.g., steps
and kinks) (10, 11) and there is no thermodynamic force driving migration to sites of lower
ΔHads. At an Eads of 300 mV, the Eonset exceeds Eads by 60 mV because the distribution of
adsorption sites are more heterogeneous and there is a driving force for COads on terraces to
migrate to steps, kinks and ad atoms. Thus the difference between Eonset on Eads would be
100 200 300 400 500 6002074
2076
2078
2080
2082
2084
2086
2088
2090
2092
Wavenum
ber
(cm
-1)
Potential (mV)
100 200 300 400 500 6002074
2076
2078
2080
2082
2084
2086
2088
2090
2092
Wavenum
ber
(cm
-1)
Potential (mV)
100 200 300 400 500 6002074
2076
2078
2080
2082
2084
2086
2088
2090
2092
Wavenum
ber
(cm
-1)
Potential (mV)
100 200 300 400 500 6002074
2076
2078
2080
2082
2084
2086
2088
2090
2092W
avenum
ber
(cm
-1)
Potential (mV)
400 mV 300 mV
100 mV 200 mV
30
expected to increase as Eads is decreased. At Eads of 200 mV and lower, the Eonset is leveled to 320
mV
Figure 2.3. CO electro-oxidation onset potentials versus adsorption potentials
31
Stark tuning plots of CO/Pt obtained at 30° C, 50° C and 70° C (Fig. 2.4) are similar to those of
figure 2.2.
Figure 2.4. Operando CO/Pt Stark tuning. (12)
100 200 300 400 500 600
2072
2074
2076
2078
2080
2082
2084
2086
2088
Waven
um
ber
(cm
-1)
Potential (mV)
0 100 200 300 400 500 600
2072
2074
2076
2078
2080
2082
2084
2086
2088
Waven
um
ber
(cm
-1)
Potential (mV)
0 100 200 300 400 500 600
2072
2074
2076
2078
2080
2082
2084
2086
2088
Waven
um
ber
(cm
-1)
Potential (mV)
30° C
50° C
70° C
32
Figure 15 shows that Eonset decreases linearly (slope = -2.88 cm-1
/K) with temperature. The
relationship of this linear variation to the kinetics of the inner sphere processes(13) will be
addressed in future work.
Figure 2.5. The potential at which CO begins to oxidize as a function of temperature.
Figures 2.2 and 2.4 show an adsorption process despite the lack of mobile ions typical of
aqueous sulfuric acid. The operando spectroscopy suggests a need for elucidation of Nafion
functional groups responsible for the complex Stark tuning plots.
2.4 Conclusion
Stark tuning of CO adsorbed at 100, 200, 300 and 400 mV vs. the fuel cell counter-
reference electrode (i.e., NHE) is interpreted in terms of the difference between the oxidation
30 40 50 60 70
280
300
320
340
360
380
400
420
Po
ten
tia
l (m
V)
Temperature (°C)
Slope: -2.88
R2: 0.9911
33
onset potential and the potential at which the CO was adsorbed. At 400 mV, the oxidation onset
potential (Eonset) is coincident with the adsorption potential (Eads) because CO preferentially
adsorbs only on sites with high Hads and there is no driving force for migration. As Eads, Eonset
decreases relative to Eads as CO migrates to sites of higher enthalpies of adsorption. At Eads at or
below 200 mV the Eonset is leveled to 325 mV.(14)
References:
1. Blyholder, G., Molecular orbital view of chemisorbed carbon monoxide. Journal of Physical Chemistry, 1964. 68(10): p. 2772-8.
2. Beden, B., et al., Infrared Study of Adsorbed Species on Electrodes - Adsorption of Carbon-Monoxide on Pt, Rh, and Au. Journal of Electroanalytical Chemistry, 1982. 142(1-2): p. 345-356.
3. Russell, J.W., et al., Infrared-Spectrum of CO on a Platinum-Electrode in Acidic Solution. Journal of Physical Chemistry, 1982. 86(16): p. 3066-3068.
4. Golden, W.G., Dunn, D.S., and Overend, J., A Method for Measuring Infrared Reflection-Absorption Spectra of Molecules Adsorbed on Low-Area Surfaces at Monolayer and Submonolayer Concentrations. Journal of Catalysis, 1981. 71(2): p. 395-404.
5. Stamenkovic, V., et al., Vibrational properties of CO at the Pt(111)-solution interface: the anomalous stark-tuning slope. Journal of Physical Chemistry B, 2005. 109(2): p. 678-680.
6. Villegas, I. and Weaver, M.J., Carbon-Monoxide Adlayer Structures on Platinum(111) Electrodes - A Synergy Between in-situ Scanning-Tunneling-Microsopy and Infrared-Spectroscopy. Journal of Chemical Physics, 1994. 101(2): p. 1648-1660.
7. Yoshimi, K., Song, M.B., and Ito, M., Carbon monoxide oxidation on a Pt(111) electrode studied by in-situ IRAS and STM: Coadsorption of CO with water on Pt(111). Surface Science, 1996. 368: p. 389-395.
8. Gurau, B. and Smotkin, E., Methanol crossover in direct methanol fuel cells: a link between power and energy density. J. Power Sources, 2002. 112: p. 339-352.
9. Stoupin, S., et al., Pt and Ru X-ray Absorption Spectroscopy of PtRu Anode Catalysts in Operating Direct Methanol Fuel Cells. Journal of Physical Chemistry B, 2006. 110(20): p. 9932-9938.
10. Kim, C.S. and Korzeniewski, C., Vibrational Coupling as a Probe of Adsorption at Different Structural Sites on a Stepped Single-Crystal Electrode. Analytical Chemistry, 1997. 69(13): p. 2349-2353.
11. Yoshinobu, J., et al., Lateral displacement by transient mobility in chemisorption of CO on Pt(997). Phys Rev Lett, 2003. 90(24): p. 248301.
12. Kendrick, I., et al., Elucidating the Ionomer-Electrified Metal Interface. J. Am. Chem. Soc., 2010. 132(49): p. 17611-17616.
34
13. Bard, A.J., Inner-Sphere Heterogeneous Electrode Reactions. Electrocatalysis and Photocatalysis: The Challenge. Journal of the American Chemical Society, 2010. 132(22): p. 7559-7567.
14. Iwasita, T. and Pastor, E., A DEMS and FTIR Spectroscopic Investigation of Adosbed Ethanol on Polycrystaline Platinum. Electrochimica Acta, 1994. 39(4): p. 531-537.
35
Chapter 3: Elucidating the ionomer-metal interface
3.1 Introduction
A solubilized version of Nafion(1) (ionomer solution) is often used to prepare catalyst “inks”
that are directly painted or decal transferred to the membrane.(2) The ionomer-metal interface
formed after evaporation of the ink solvent is central to PEMFC electrocatalysis. For example, a
spin coated Nafion layer on polycrystalline Pt enhances electrocatalysis.(3, 4) Little is known
about ionomer-metal interfaces. Markovic and co-workers probed Pt-electrolyte interfaces by
measurements of CO oxidation currents, in sulfuric, perchloric and KOH solutions, synchronized
with IR absorption-reflection spectroscopy of linear (νCOl) and bridge bound (νCO
b) COads on
Pt(111).(5) More recently their electrochemical studies on Pt(hkl)-Nafion interfaces suggest that
the Nafion sulfonate group adsorbs onto the Pt surface.(6) A fuel cell membrane electrode
assembly uniquely enables study of the ionomer metal interface without interferences due to
mobile anions characteristic of aqueous acidic electrolytes. Operando fuel cell infrared (IR)
spectroscopy was introduced by Fan et al.(7) In this report the aggregate of operando
spectroscopy of fuel cell membrane electrode surfaces, attenuated total reflectance spectroscopy
of Nafion 117, polarization modulated IR reflection absorption spectroscopy of Nafion spin-
coated onto Pt, and density functional theory calculated Nafion spectra suggest a model for the
Pt-Nafion interface the includes the Nafion CF3 group as an important co-adsorbate at the
ionomer-Pt interface.
3.2 Experimental:
Attenuated total reflectance (ATR) spectroscopy: A surface pressure of 815 psi was
maintained over the 1.8 mm diameter ATR crystal. Spectra were obtained using a Bruker™
36
Vertex 70 and Vertex 80V vacuum FTIR spectrometer (Bruker, Billerica, MA). A MIRacle™
ATR accessory (Pike Technologies Spectroscopic Creativity, Madison, WI) with a ZnSe ATR
crystal was used. The spectra were signal averaged from 100 scans at 4 cm-1
resolution with a
dry-air purge at ambient temperature. Atmospheric compensation (to eliminate H2O and CO2
interference in the beam path) was used in all measurements. Data processing for all infrared
data was done with the Bruker™ OPUS 6.5™ software.
Preparation of arc-melt Pt: The preparative method for arc melted electrodes has been
described.(8) Briefly, the arc-melter (Materials Research Furnaces, Sun Cook, NH) was charged
with 3 mm Pt shot (99.9+%, Sigma-Aldrich, St. Louis, MO) The chamber was evacuated to -29
psig and purged with argon three times. The Pt was arc-melted at 75 amps under an Ar bleed.
The chamber was vented to flip and arc-melt the sample three times. The Pt slug was epoxied
(Devcon HP250, Danvers, MA) to a modified glass syringe barrel, cut flat using a diamond cut-
off saw (Buehler IsoMet 1000, Lake Bluff, IL), and finally polished to a mirror finish using 0.05
µm aluminum oxide (Magner Scientific, Dexter, MI). The electrode was sonicated in
Nanopure™ water (Milli-Q, Billerica, MA) for 10 min. Nafion ionomer solution (20µL) was
pipetted onto the Pt electrode assembly mounted on an inverted electrode rotator (Pine
Instrument Company, Grove City, Pa) and then rotated (1000 rpm, one min).
Polarization modulated infrared reflection absorption spectra (PM-IRRAS): The Vertex
80V spectrometer was equipped with a Hinds II/ZS50 photoelastic modulator (Hinds
Instruments, Hillsboro, OR), SR830 lock-in amplifier (Stanford Research Systems, Sunnyvale,
CA) and a D3131\6 MCT detector (Infrared Associates, Stuart, FL). The angle of incidence was
60° and the photoelastic modulator frequency was 50.14 kHz. The PM-IRRAS cell design have
37
been reported.(9, 10) Spectra were averaged (710 scans; 4 cm-1
resolution). Li-exchanged
Nafion was prepared by soaking Nafion samples in 0.1M salt solutions.
Computational method: Unrestricted DFT(11, 12) with the X3LYP(13) functional was used
for geometry optimization and calculations of the normal mode frequencies and corresponding
IR spectra of the deprotonated and protonated Nafion side-chain and backbone segment. The
backbone-segment terminal ends were substituted with CH3 groups to eliminate computational
interference with the Nafion CF3 group. Jaguar 6.5 (Schrodinger Inc., Portland, OR) was used
with the all-electron 6-311G**++ Pople triple- basis set (“**” and “++” denote polarization
(14) and diffuse (15) basis set functions, respectively). Output files were converted to vibrational
mode animations using Maestro (Schrodinger Inc). Calculations were carried out on a 55 node
(dual core Xeon processors with 4GB RAM) High Performance Computing Cluster at the
University of Texas, Pan American.
Operando spectroscopy: Temperature dependant COads Stark tuning data were acquired by
operando specular reflectance IR spectroscopy using a cell (Fig. 1.8) based on the design of Fan
et al.(7) The cell, controlled by an EZstat potentiostat (NuVant Systems Inc), interfaces to a
diffuse reflectance accessory (Pike Technologies, Madison, WI) installed on the Vertex 70
Spectrometer. The IR beam accesses the working electrode surface through a CaF2 window
inserted into the upper flow field and a small slot in the carbon gas diffusion layer. The lower
flow field-electrode serves as both a hydrogen reference and counter electrode when charged
with hydrogen. The small CO oxidation currents do not measurably polarize the hydrogen
counter electrode. The working electrode was cycled (50 times) from 0 to1.2 V vs. the hydrogen
counter electrode. Spectra were obtained by averaging 250 scans at 4 cm-1
resolution. The cell
was brought to the desired temperature and potential (300 mV) for acquisition of reference
38
spectra. Carbon monoxide was passed over the working electrode for 15 min. The cell was
purged with N2 (15 min) prior to setting the potential to 100 mV. Replicates of four spectra were
acquired at 50 mV increments until the CO vibrational bands were no longer observable.
3.3 Results and Discussion:
Operando Spectroscopy: Stark tuning plots of COads on Pt in a fuel cell operated at 30° C,
50° C and 70° C (Fig. 2.4) show a remarkable similarity to plots obtained by Stamenkovic et al.
in sulfuric acid.(5) They correlated complex potential dependences (Stark tuning), of COads
vibrational frequencies (i.e., dνCOl/dE and dνCO
b/dE) in 0.5 M H2SO4, to the
compression/dissipation of COads islands: After a linear region from 0.1 to 0.3 V (dνCOl/dE = 31
cm-1
/V), a subtle νCOl blue shift from the extrapolated linear region was followed by a precipitous
drop, initiating at 0.5 volts, that finally upturns at 0.65V. They attribute this behavior to COads
island compression due to repulsive dipole interactions with co-adsorbed bisulfate ion initially
observable at 0.35V. COads oxidation initiating at 0.5 volts, induced by OH- absorption,
diminishes dipole-dipole coupling(16) and thus decreases dνCOl/dE. The upturn is attributed to
increased absorption of HSO3- relative to OH
-, which reestablishes repulsive dipole interactions
that compress COads islands and increase νCOl. The plots in Figure 2.4 demonstrate an adsorption
phenomenon onto Pt despite the lack of mobile ions typical of dilute sulfuric acid solutions. The
fuel cell operando spectroscopy suggests a need for elucidation of Nafion functional groups
responsible for modulation of CO/Pt interactions (i.e., Stark tuning).
Analysis of IR spectra: The DFT calculated spectrum of a 55-atom Nafion side-chain and
backbone segment provides 159 normal mode frequencies and intensities. Figure 3.1 shows the
theoretically derived peak positions and intensities (black lines) superimposed upon the ATR
spectrum (red line) of hydrated Nafion. ATR is a spectroscopic technique that requires that the
39
sample come into with what’s known as an internal reflection element (IRE). When light with
an angle of incidence θA passing through a transparent material passes through another material
with a different index of refraction, the resulting light will be refracted at angle θB. If θA exceeds
a critical angle θc, the light will be reflected internally. The value of θc is defined as:
equation 1
Where nA is the index of refraction of the first material and nB is the index of refraction of the
second. In the case of ATR, material A is the IRE and material B is the sample. As shown in
equation 1, the IRE must have an index of refraction greater than the sample for light to be
reflected internally at the interface. Even though the light is reflected internally within the IRE,
energy is lost via an evanescent wave emanating from the IRE perpendicular to the surface,
penetrating the sample. The evanescent wave’s depth of penetration is calculated by:
√
equation 2
Where λ is the wavelength. Given that the equation 2 is linear with only variable during an
acquiring a spectrum is λ, the loss of peak intensity at higher wavenumbers can easily be
compensated for.
40
Figure 3.1. DFT calculated normal modes (black lines) and Nafion ATR spectrum (red).
PM-IRRAS enhances (relative to the ATR) vibrational modes of functional groups ordered by
the Pt surface. PM-IRRAS is technique that enhances peaks associated with molecules ordered
by a surface. A polarizer generates light perpendicular the plane of incidence (p-polarized). An
angle of incidence higher than ~50° is required. At angles below 50°, the electric vector of p-
polarized light undergoes a phase change of 180° resulting in destructive interference between
the incident and reflected rays. The optimum angle of incidence of IRRAS techniques is usually
around 70°. A dipolar molecule at the metal surface results in an image dipole inside the
metal.(17) When the molecule is oriented parallel to the metal, the negative charge induces a
positive charge in the image dipole. Thus, upon absorption, there is no net change in dipole
moment. When the molecule’s dipole moment is perpendicular to the surface, is in the case of
41
adsorbed molecules, the image dipole is such that the net change in dipole moments creates an
additive effect (18)
Figure 3.2 shows the ATR spectrum (red), the PM-IRRAS spectra of Nafion-H/Pt interface
(grey line), Nafion-Li/Pt interface of Li+ exchanged Nafion (blue line) and 6 selected (from the
159 calculated) DFT calculated frequencies and intensities.
Figure 3.2. Theoretical and experimental spectra. ATR of hydrated Nafion (red); PM-IRRAS of
Nafion-H on Pt (grey); PM-IRRAS of Nafion-Li on Pt (blue); Selected DFT peaks (black lines 1-
6).
Scheme 3.1 is the Nafion structure with functional groups labeled for ease of discussion. The
low-frequency ATR band (Fig. 3.2) at 971 cm-1
(corresponding to theoretical 984 cm-1
; line-1)
and the 1056 cm-1
band (corresponding to theoretical 1059 cm-1
; line-3) have recently been
thoroughly assigned by Webber et al.(19) They obtained high resolution transmission spectra of
hydrated and thoroughly dehydrated Nafion and analyzed them in the context of the
spectroscopy of the short chain ionomer (formerly DOW membrane), the Nafion sulfonyl
fluoride(20) and the Nafion sulfonyl imide.(21) Animations of the DFT calculated internal
coordinates reveal that the observed 1056 cm-1
and 971 cm-1
peaks both have internal coordinates
42
resulting from the mechanical coupling of the adjacent the sulfonate and COC (A) ether link.(19)
Thus these peaks shift concertedly with changes in the sulfonate environment. Consider the
ATR and PM-IRRAS spectra of the protonic form of Nafion (Fig. 3.2, grey line). The 1056 cm-1
and 971 cm-1
peaks concertedly shift to higher frequencies in the PM-IRRAS because of the
interaction of the sulfonate functional group with the Pt surface. A similar effect is observed
with Li+ exchange of the adsorbed Nafion (blue line).
Scheme 3.1. Segment and atom labeling for Nafion
A convention for correlating PM-IRRAS enhanced peaks to the calculated DFT peaks would
enable identification of functional groups ordered by the Pt surface: The association of observed
PM-IRRAS peaks with DFT peaks, assigned by visualization of mechanically coupled internal
coordinates,(19, 22) provides the basis for such a convention. Normal mode coordinate
animations (generated by Maestro from DFT output files) explicitly show how neighbor
functional groups (called out in scheme 1) are mechanically coupled. The calculated internal
coordinates are viewed in the context of calculated normal modes of relevant small molecules
43
(e.g., triflic acid, CF3OCF3, 10-carbon CF2 backbone, etc.) hereafter are referred to as “pure
modes,” which serve as the basis-elements for assigning DFT calculated normal modes
associated with observed peaks. Figure 3.3 shows the assignments of the 6 selected DFT peaks
and snapshots of the corresponding Maestro animations. The atoms contributing to the
dominating motion (black circles) and the next most significant atom motions (dotted circles)
comprise pure modes that form the basis for the assignments. An alternate strategy for
determining the dominant mode is to consider the contribution to the potential energy surface on
an atom by atom basis.(23) While this may change the selection of the dominant mode, it does
not alter what pure modes contribute to the assignments. The correlation of the DFT to PM-
IRRAS peaks (Fig. 3.2) and the resulting assignments in terms of the mechanically coupled
modes are tabulated in Table 3.1.
44
Figure 3.3. Normal mode coordinate animation snapshots of the Nafion side-chain anion and
backbone fragment (see scheme 1). Left and right views are extrema positions of the vibrational
mode. Functional groups associated with the dominant internal coordinates and next most
significant motions are designated by solid and dotted boundary lines respectively.
2. 992 cm-1 CF2
(BBdef) + COC(B) δs
6. 1322 cm-1 CF2 δs (BBdef)
1. 984 cm-1 SO3
- s + COC (A) as + COC (B) ρr
3. 1059 cm-1 COC (A) as + SO3
- s
4. 1168 cm
-1 CF2 δs (BBdef) + CF2 ρr (SCdef) + COC(A)
5. 1254 cm-1 CF3
as + COC (A) δs + COC (B) as
45
Symmetric stretch, s; Asymmetric stretch, as
Wagging, ; Scissoring, δs; Twisting, ; Rocking, ρr
Backbone deformation, BBdef; Side-chain deformation, SCdef; Backbone Stretching, BBstre
Table 3.1. PM-IRRAS and DFT IR adsorption peaks and assignments.
The pure mode peak assignments (Table 1) elucidate functional groups ordered by the Pt
surface. The rational for the key functional group assignments (Table 1) is supported by the
overlap of the DFT calculated peak positions with the PM-IRRAS peaks. Consider the DFT and
PM-IRRAS peaks in the contexts of the bulk-Nafion ATR and the report by Cable et al.(20) that
the 1056 cm-1
and 971 cm-1
peaks shift with alterations of the sulfonate group environment. The
bulk ATR peak at 1056 cm-1
(red), the PM-IRRAS peak of protonated Nafion adsorbed on Pt
(grey line) at 1061 cm-1
and the PM-IRRAS peak of lithiated Nafion adsorbed on Pt (blue line) at
1077 cm-1
(Fig. 3.2) confirm that Pt surface atoms induce frequency shifts, as do extent-of-
hydration(19) and ion exchange of Nafion.(20) Thus the PM-IRRAS enhances bulk-Nafion-
modes that are shifted due to functional group interactions with Pt. Less explicit than the 1056
cm-1
peak, are PM-IRRAS peaks derived from bulk-Nafion-modes that are convoluted within the
Nafion ATR broad envelop region (1100 - 1300 cm-1
), in particular the 1164 and 1260 cm-1
PM-
IRRAS peaks. Di Noto et al.(24) extensively deconvoluted the broad envelop region. Their
Wavenumber (cm-1) Pure Mode Components
PM-IRRAS DFT
1 971 984 SO3- s + COC(A) as + COC(B) ρr
2 984 992 CF2 (BBdef) + COC(B) δs
3 1061 1059 COC(A) as + SO3- s
4 1164 1168 CF2 δs (BBstre) + CF2 (BBdef) ρr + COC(A)
5 1260 1254 CF3 as + COC(A) δs + COC(B) δs
6 1322 1322 CF2 δs (BBdef)
46
resulting peak library includes 1148, 1245 cm-1
, which could be reconciled with an association of
the DFT peaks (Fig. 3.2, line-4 and line-5 ) with the shifted PM-IRRAS peaks at 1164 and 1260
cm-1
.
Nafion/Pt adsorption model: The animation of the theoretical peak at 1254 cm-1
(Fig. 3.2,
line-5), associated with PM-IRRAS peaks at 1260 cm-1
(blue and grey lines), suggests that the
CF3 internal coordinates dominate the normal mode. The insensitivity of the 1201 cm-1
peak, to
ion exchange, suggests that the internal coordinates are not substantially coupled to the sulfonate
group. The 1260 cm-1
band-intensity is over an order-of-magnitude greater than that the cluster
of peaks (i.e., associated with theoretical lines 1 and 2) that are mechanically coupled to the
sulfonate pure mode: The CF3 functional group is a co-adsorbate of comparable importance to
the sulfonate exchange group in formation of the Nafion/Pt interface. Further support for this
model is provided by Mulliken population(25) analysis. Atomic charges of the 55 Nafion
fragment atoms were calculated. Table 3.2 shows the average charges of the backbone, side
chain and CF3 group fluorine atoms and the average charges on the sulfonate oxygen atoms. The
charges for chemically equivalent atoms (e.g., CF3 fluorine and sulfonate oxygen atoms) differ
because the calculations are done for the lowest energy Newman projections where the atomic
environments are different for chemically equivalent atoms because of the absence of symmetry
in the full molecule. The chemically equivalent atoms have smaller charge standard deviations
as would be expected. The average charge of the CF3 fluorine atoms are the highest amongst the
three classes of fluorine atoms (Table 2) and are about 18% that of the sulfonate oxygens.
47
Segment Backbone (F) Side-chain (F) CF3 (F) Sulfonate (O)
13 atoms 8 atoms 3 atoms 3 atoms
Avg. Partial Charge -0.0665 -0.0816 -0.0876 -0.4879
Standard Deviation 0.027 0.055 0.013 0.012
Table 3.2. Average partial charges of selected Nafion segments
A Gaussian 03 Viewer (Gaussian, Wallingford, CT) used to construct a 2-equivalent (1100
g/equiv) model of Nafion 117, enables rotation of dihedral angles while maintaining the native
bond angles associated with each and every functional group. The CF3 and SO3- groups, oriented
with the two planes defined by the CF3 fluorine and sulfonate oxygen atoms parallel to a Pt
surface, effect ordering of the CF2 backbone segments with respect to the Pt surface. Ordering of
the CF2 groups would be expected to yield PM-IRRAS peaks. The PM-IRRAS peak at 1164 cm-
1 is associated with the theoretical peak (line-4) at 1168 cm
-1. The line-4 animation shows that
CF2 backbone internal coordinates dominate the 1168 cm-1
mode, supporting the suggestion of
ordered CF2 groups. Figure 3.4 is the Gaussian View model resulting from orienting the CF3 and
SO3- groups for adsorption to the Pt surface. The numbers (yellow) associate DFT calculated IR
peaks (line-1 – 6, Fig. 3.2) and associated PM-IRRAS peaks with regions of order induced by the
CF3 and SO3- functional group adsorbates.
48
Figure 3.4. Gaussian 03 Viewer Nafion-Pt interface model. Oxygen (red), Sulfur (yellow),
Fluorine (light blue), Carbon (grey), Pt (dark blue).
The ordering of the backbone CF2 groups in the Gaussain model is a natural consequence
of adjusting the dihedral angles of the anchoring groups for adsorption, while maintaining
functional group native bond angles. Thus the aggregate of the Stark tuning data of Figure 2.3,
the PM-IRRAS and DFT calculations support Figure 3.4 as a model for Nafion functional group
adsorption to Pt. The details of exactly how adsorbed CF3 functional groups influence the
operando Stark tuning curves is not yet established. The low density of functional group
adsorption sites, relative to the number of backbone CF2 groups suggests an explanation as to
why Nafion is observed to enhance electrode processes.(3, 26) The methodology of assigning IR
bands in the context of mechanically coupled internal coordinates of neighboring functional
groups, and correlating those assignments to functional groups interactions with metal surfaces,
has broad applications towards characterization of ionomeric interfaces.
2.4 Conclusion:
Operando IR spectroscopy, PM-IRRAS of Nafion-Pt interfaces, and ATR spectroscopy of
Nafion, correlated with DFT calculated normal mode frequencies confirm that Nafion side-chain
sulfonate and CF3 co-adsorbates are structural components of the Nafion-Pt interface. These
5 1
2,4,6
3
49
“anchoring” functional groups reduce degrees of freedom available for backbone and side-chain
CF2 dynamics. The partial ordering of Nafion CF2 groups is supported by observed PM-IRRAS
and DFT calculated peaks possessing vibrational internal coordinates dominated by, and
mechanically coupled to side-chain CF2 group motions.
References
1. Martin, C.R., Rhoades, T.A., and Ferguson, J.A., Dissolution of perfluorinated ion-containing polymers. Analytical Chemistry, 1982. 54(9): p. 1639-1641.
2. Wilson, M.S. and Gottesfeld, S., Thin-Film Catalyst Layers For Polymer Electrolyte Fuel-Cell Electrodes. Journal of Applied Electrochemistry, 1992. 22(1): p. 1-7.
3. Liu, L., et al., Methanol oxidation on nafion spin-coated polycrystalline platinum and platinum alloys. Electrochemical and Solid State Letters, 1998. 1(3): p. 123-125.
4. Ploense, L., et al., Spectroscopic study of NEMCA promoted alkene isomerizations at PEM fuel cell Pd-Nafion cathodes. Solid State Ionics, 2000. 136-137: p. 713-720.
5. Stamenkovic, V., et al., Vibrational properties of CO at the Pt(111)-solution interface: the anomalous stark-tuning slope. Journal of Physical Chemistry B, 2005. 109(2): p. 678-680.
6. Subbaraman, R., et al., Three Phase Interfaces at Electrified Metal-Solid Electrolyte Systems 1. Study of the Pt(hkl)-Nafion Interface. Journal of Physical Chemistry C, 2010. 114(18): p. 8414-8422.
7. Fan, Q., et al., In situ FTIR-diffuse reflection spectroscopy of the anode surface in a direct methanol/oxygen fuel cell. Journal of the Electrochemical Society, 1996. 143(2): p. L21-L23.
8. Ley, K.L., et al., Methanol oxidation on single-phase Pt-Ru-Os ternary alloys. Journal of the Electrochemical Society, 1997. 144(5): p. 1543-1548.
9. Kunimatsu, K., et al., Carbon-Monoxide Adsorption on a Plationum-Electrode Studied by Polarization Modulated FT-IRRAS.1. CO Adsorbed in the Double-Layer Potential Region and Its Oxidation in Acids. Langmuir, 1985. 1(2): p. 245-250.
10. Kunimatsu, K., Infrared Spectroscopic Study of Menthanol and Formic-Acid Adsorbates on a Platinum-Electrode .1. Comparison of the Infrared-Adsorption Intensities of Linear CO(A) Derived From CO, CH3OH AND HCOOH. Journal of Electroanalytical Chemistry, 1986. 213(1): p. 149-157.
11. Hohenberg, P. and Kohn, W., Inhomogeneous Electron Gas. Physical Review B, 1964. 136(3B): p. B864-&.
12. Kohn, W. and Sham, L.J., Self-Concsistent Equations Including Exchange And Correlation Effects. Physical Review, 1965. 140(4A): p. 1133-&.
13. Xu, X., et al., An extended hybrid density functional (X3LYP) with improved descriptions of nonbond interactions and thermodynamic properties of molecular systems. Journal of Chemical Physics, 2005. 122(1): p. 14.
50
14. Frisch, M.J., Pople, J.A., and Binkley, J.S., Self-Consistent Molecular-Orbital Methods 25. Supplementary Functions For Gaussian-Basis Sets. Journal of Chemical Physics, 1984. 80(7): p. 3265-3269.
15. Clark, T., et al., Efficient Diffuse Function-Augmented Basis-Sets For Anion Calculations 3. The 3-21+G Basis Set For 1ST-Row Elements, LI-F. Journal of Computational Chemistry, 1983. 4(3): p. 294-301.
16. Persson, B.N.J. and Ryberg, R., Vibrational Interaction Between Molecules Adsorbed On A Metal-Surface - The Dipole-Dipole Interaction. Physical Review B, 1981. 24(12): p. 6954-6970.
17. King, F.W., Duyne, R.P.V., and Schatz, G.C., Theory of Raman scattering by molecules adsorbed on electrode surfaces. The Journal of Chemical Physics, 1978. 69(10): p. 4472-4481.
18. Griffiths, P.R. and de Haseth, J.A., Specular Reflection, in Fourier Transform Infrared Spectrometry. 2006, John Wiley & Sons, Inc. p. 277-301.
19. Webber, M., et al., Mechanically Coupled Internal Coordinates of Ionomer Vibrational Modes. Macromolecules, 2010. 43(13): p. 5500-5502.
20. Cable, K.M., Mauritz, K.A., and Moore, R.B., Effects of hydrophilic and hydrophobic counterions on the Coulombic interactions in perfluorosulfonate ionomers. Journal of Polymer Science, Part B: Polymer Physics, 1995. 33(7): p. 1065-72.
21. Byun, C.K., et al., Infrared Spectroscopy of Bis (perfluoroalkyl)sulfonyl Imide Ionomer Membrane Materials. Journal of Physical Chemistry B, 2009. 113(18): p. 6299-6304.
22. Warren, D.S. and McQuillan, A.J., Infrared spectroscopic and DFT vibrational mode study of perfluoro(2-ethoxyethane) sulfonic acid (PES), a model Nafion side-chain molecule. Journal of Physical Chemistry B, 2008. 112(34): p. 10535-10543.
23. Johansson, P. 2010: Gothenburg, Sweden. 24. Di Noto, V., et al., Structure, properties and proton conductivity of Nafion/ (TiO2)center
dot(WO3)(0.148) (psi TiO2) nanocomposite membranes. Electrochimica Acta, 2010. 55(4): p. 1431-1444.
25. Mulliken, R.S., Electronic population analysis on LCAO-MO [linear combination of atomic orbital-molecular orbital] molecular wave functions. I. Journal of Chemical Physics, 1955. 23: p. 1833-40.
26. Ploense, L., et al., Proton spillover promoted isomerization of n-butylenes on Pd-black cathodes/Nafion 117. Journal of the American Chemical Society, 1997. 119(47): p. 11550-11551.
51
Chapter 4: Minimal Vibrational Mode Analysis of Nafion Infrared
Spectroscopy
4.1 Introduction:
Ionomer membranes have a rapidly growing range of applications, including separators,
reaction media, and electrolytes that motivate the investigation of structural variations to
accommodate new functionalities.(1-10) Ionomers typically feature a backbone structure with
side-chains terminated by an ion exchange group. Infrared spectroscopy (IR) is an important tool
for correlating the response of the ionomer exchange group and backbone to ion exchange and
absorption of solutions of molecular species. The infrared spectra of polymers, in general are
deceptively simple. Bower and Maddams explain this by considering a polymer as a series of
chemical repeat units,(11) noting that the wavelength of the IR radiation absorbed by the
polymer is orders of magnitude larger than the repeat unit dimensions. Thus, each wavelength
will reflect the interaction between the IR radiation and an ensemble of repeat units with the
number of normal modes reduced to about 3n where n is the number of atoms in the repeat unit
(vs. 3N-6 where N is the number of atoms in the molecule): The chemical repeat unit of Nafion
(Scheme 1) would give rise to well over 100 normal modes (with intensity) within the fingerprint
region (900-1400 cm-1
). Yet there are only six bands and associated shoulders within the finger
print region. It is difficult to assign individual bonds in this region than at higher frequencies due
to the different bending vibrations within the molecule. In the past, researchers have used the
distinct –OH stretching bands at higher frequencies to study hydration.(12, 13) The discussion
of the fingerprint region is generally limited to the well defined peaks at 1060, 969 and 980 cm-1
and their relationship with the SO3- and ether groups. Analysis of the broad envelope region
52
from 1070-1350 cm-1
is incomplete or simply assigned as backbone vibrations. Despite over
7000 publications on Nafion since 1975,(14) no consensus has been reached regarding the
assignment of the fingerprint region.
Using density functional theory (DFT), the theoretical spectrum of a 55-atom Nafion proton
dissociated repeat unit (scheme 1, right) was generated.(15) The 159 normal mode spectrum was
accompanied by an eigenvector animation for each normal mode. These animations support the
notion introduced by Cable et al.(16) and Warren and McQuillan(17) that normal modes should
be assigned as group modes rather than the prevalent single-functional-group assignments. .
The contributions of different functionalities and their vibrations (e.g. stretching, scissoring, etc)
were assigned through visualization of each normal animation. (18) This method is lacking
given that the process is time consuming, precludes the assignment of the relative contribution of
each functionality, and is ultimately subjective. The thorough assignment of polymer vibrational
spectra demands a quantitative method of assigning each theoretical normal mode. This work
describes a method which assigns the fingerprint region of the Nafion theoretical spectrum.
Visualization of normal mode animations are used only to identify functional groups
participating in a group mode. For each normal mode, the selected functional groups are then
represented as subsets of the generalized coordinates. These subsets provide a set of eigenvectors
now introduced as minimal vibrational modes (MVMs). The MVMs are coded as color bars,
where bar-lengths scale with the MVM contribution to the normal mode. The addition of a
categorical MVM axis clarifies the complex effects of the mechanical coupling in another
dimension as well. Along an MVM axis, the distribution of selected functional group MVM
amongst all normal modes is easily visualized (Fig. 3 top) vida infra.
53
Scheme 4.1: Nafion chemical repeat unit.
4.2 Experimental:
IR Spectroscopy: Attenuated total reflectance (ATR) spectra of Nafion 117 were obtained
using a Bruker™ Vertex 70 and Vertex 80V vacuum FTIR spectrometer (Bruker, Billerica, MA)
equipped with a MIRacle™ ATR accessory (Pike Technologies Spectroscopic Creativity,
Madison, WI) using a ZnSe ATR crystal with 45° beveled edges. A surface pressure of 815 psi
was maintained over the 1.8 mm diameter ATR crystal. Atmospheric compensation (to
eliminate H2O and CO2 interference in the beam path) was used in all measurements and the
spectra were corrected to take into account for the depth of penetration of the IR beam. Spectra
were signal averaged (100 scans at 4 cm-1
resolution) using a DLaTGS detector. Data processing
was done with the Bruker™ OPUS 6.5™ software.
Density Functional Theory: Unrestricted DFT(19, 20) with the X3LYP(21) functional was
used for geometry optimization and normal mode calculations of the Nafion side-chain and
backbone-segment. The Nafion backbone-segment is artificially terminated to yield the
computational chemical repeat unit used in this study. CH3 groups were used to terminate the
backbone segment rather than CF3 groups in order to avoid computational interference with the
Nafion side-chain CF3 group (i.e. the only type of perfluorinated methyl group in the molecule).
54
Jaguar 6.5 (Schrodinger Inc., Portland, OR) was used with the all-electron 6-311G**++ Pople
triple- basis set (“**” and “++” denote polarization (22) and diffuse (23) basis set functions,
respectively). Output files were converted to vibrational mode animations using Maestro
(Schrodinger Inc.). Calculations were carried out on a 55 node (dual core Xeon processors with
4GB RAM) High Performance Computing Cluster at the University of Texas, Pan American.
Automated Normal Mode Assignments: The normal mode displacements were calculated
using our normal mode output file.(15) A Mathematica (Wolfram Res., Champaign, IL) program
calculates normal mode displacements as:
and Eq. (1)
where xi are the atomic coordinates of the geometry optimized repeat-unit and N is the number
atoms of the repeat unit. The generalized coordinates (q) are defined in terms of and xi. The
MVMs are described in terms of generalized coordinate subsets (đ) obtained from q. The MVMs
(e.g., bending, symmetric, asymmetric, wagging, twisting, rocking, and scissoring), are
represented by a đ, are assigned a color. The đ2 are calculated and normalized to the largest đ
2 on
a per-color basis. The program color-codes DFT line spectra by re-normalizing the đ2 on a per
mode basis. The methodology is summarized (Fig. 1). To demonstrate the reproducibility of this
technique, the theoretical vibrational spectrum of benzyltrimethylammonium hydroxide was
assigned. Benzyltrimethylammonium hydroxide was selected because it has been suggested that
it is the ion exchange head group in commercially available anionic exchange membranes
(AEMs).
55
Figure 4.1. Flow chart for generation of Nafion repeat group MVM spectra.
Consider the normal modes dominated by the Nafion side-chain groups (e.g., SO3- , CF2,
CF, CF3 and COC(A,B)). In the case of C3v type functional groups (SO3– and CF3) the
symmetric stretch is the average of the S-O or C-F bonds as given by Eq. 2 and shown on Fig.
2 (a).
Table 1 shows the MVM internal coordinate system of the most prominent IR modes for
Nafion. Figure 2 shows the definition of the functional group internal coordinates used in this
work. An internal coordinate that involves a distance between two atoms is defined as,
| | | |, Eq. (2)
56
where is the vector defined by the atomic coordinates of the geometry optimized repeat-
unit, and is the vector defined by
of Eq. 1. Similarly, an internal coordinate that
involves an in-plane angle between three atoms is defined as,
| |-| |, Eq.
(3)
where is the angle formed by and , and is the angle formed by and
. Rocking γ is represented by the summation of the angles formed by
and and by
and . An out-of-plane angle is formed between
and , whereas the angles
δi’ measuring twisting and scissoring modes are formed by and for i=k, l.
57
Table 4.1. Internal coordinate system MVMs of the most prominent IR modes for Nafion.*
*In plane angles as follows: Out-of-plane angles as follows: is the angle formed by
and . For BB is the average between angles formed by
and and
by and . See Fig.2 for definitions of distances and bond angles.
Symmetry
type
Functional
Group
Minimal Vibrational Modes
C3V
SO3– and CF3
Symmetric stretch = average among d1, d2, and d3.
Asymmetric stretch = all sums of d1, d2, and d3, where one of the
factors is multiplied by -1.
Symmetric bend = average among α1, α2, and α3.
C2V
CF2 and COC Symmetric stretch = average between d1 and d2.
Asymmetric stretch = -d1+d2.
Wagging= average between δ1’ and δ2’.
Twisting= δ1’- δ2’.
Scissoring= δ.
Rocking= γ.
C1 Backbone BB Stretches = sum of e4,16, e16,28, e4,13, e13,19, e16, 22, e22,7, e13,25,
e19,10, and e7, 10, sum of e4,16, -e16,28, e4,13, -e13,19, e16, 22, -e22,7, -e13,25, -e19,10, and e7,
10,
sum of e4,16, -e16,28, -e4,13, e13,19, e16, 22, -e22,7, -e13,25, e19,10, and e7, 10,
sum of e4,16, e16,28, e4,13, e13,19, -e16, 22, -e22,7, -e13,25, -e19,10, and -e7,
10,
sum of e4,16, -e16,28, -e4,13, e13,19, -e16, 22, e22,7, e13,25, -e19,10, and e7, 10.
BB out-of-plane deform. = sum of 7,4,10 , 22,16,7
, 19,13,10 ,
52,28,22 ,and 48,25,19
,
sum of 7,4,10 , , 22,16,7
, 19,13,10 , 52,28,22
, and 48,25,19 ,
sum of 7,4,10 , 22,16,7
, 19,13,10 , 52,28,22
,and 48,25,19 ,
sum of 7,4,10 , 22,16,7
, 19,13,10 , 52,28,22
,and 48,25,19 ,
sum of 7,4,10 , , 22,16,7
, 19,13,10 , 52,28,22
,and 48,25,19 .
58
Figure 4.2. Definition of the internal coordinates for a) C3v, b) C2v, and c) the Nafion Backbone.
4.3 Results and discussion:
Figure 3 (top) shows the đ2 contributions of each normal mode distributed along a MVM
categorical axis. For example, the 982*, 984* and 1059
* cm
-11 normal modes all have SO3
-
MVMs. The utility of the MVM categorical axis can be seen with the MVMs corresponding
with the Nafion backbone. The perfluorinated backbone, when considered separately from the
Nafion side-chain, shares similar morphology with Teflon®. Figure 4 (left) shows MVMs
corresponding to backbone stretching and out-of-plane bending overlaid upon the ATR-IR
spectrum of Teflon®. Clusters of backbone MVMs are found at wavenumbers coincident with
experimentally obtained IR bands and have similar relative intensities. The ATR-IR spectra of
perfluoro(3-methyl-2,4-dioxahexane) sulfonic acid (PFMHSA) and perfluoro(2-ethoxybutane)
1 ATR wavenumbers are denoted with xATR, high-resolution transmission: xTR, DFT calculated: x*. Unsuperscripted values are generic values for
discussion.
59
sulfonic acid (PFEBSA) obtained by Danilczuk et al.(24) allows for a similar analysis of the Nafion
side-chain. Figure 5 shows the ATR-IR spectra of PFMHSA and PFEBSA, small molecule analog
of the Naifon side-chain, superimposed over the CF3 and SO3 peaks MVM peaks. Upon initially
viewing the PFMHSA spectrum, it would be tempting to correlate the 1228ATR
cm-1
to the SO3-
νas MVM2 found at 1223* cm-1. However, the diminished intensity of the corresponding PFEBSA
band at 1225ATR
cm-1
suggests that this band should be associated with a CF3 MVM rather than a
SO3-. This is supported by the proximity of these bands to 1201* cm
-1 which is the DFT peak
with the largest CF3 νas MVM contribution. Similarly, the 1196/1197ATR cm-1 bands should be
associated with the SO3- νas MVM of 1224* cm-1 because their intensity is not diminished through
derivitization. The transposition of these two MVM’s relative to their occurrence in the
experimentally obtained spectra is not surprising: Scaling factors inherent in DFT calculations can
shift calculated peaks (vida infra). It should be noted that comparison of these specific MVM’s to
experimentally obtained spectra is enabled by their abundance in a narrow region of the
theoretically obtained vibrational spectrum. The broad occurrence of MVM’s corresponding to
other functional groups (e.g. CF2(1),(2) and COC(A),(B), preclude similar analysis.
2 νs: symmetric stretching; νas asymmetric stretching; ω: wagging; δs: scissoring; τ: twisting; ρr: rocking.
60
Figure 4.3. Top: DFT calculated normal modes decomposed into MVM spectra (i.e., s(CF3)
light orange; CF3 as, dark orange; SO3- s, light brown; etc.) along the categorical axis. Middle:
Nafion ATR spectra (black). Composite DFT lines show contributions of minimal vibrational
modes to each normal mode. Bottom: Color code legend for minimum vibrational modes.
Str: unspecified stretching; Str, OOP: Stretching out-of-plane.
61
Figure 4.4: The ATR-IR spectrum of Teflon superimposed over Nafion backbone MVMs.
Figure 4.5: ATR spectrum of PFMHSA (left) and PFEBSA (right) superimposed over SO3- νas,
SO3- νs, CF3 νas, CF3 νs MVMs. ATR spectra are reproduced from reference (24).
The use of MVM spectra as a tool for the interpretation of Nafion spectra is exemplified
by analysis of the multiplet between 940-1000 cm-1
and the 1060 band. Aqueous solvation of
the COC(A) group has been invoked to explain the concerted shift of the 1060 cm-1
and 970 cm-
1 bands with changes in the sulfonate environment.(16) This is reconciled by the fact that all
models of hydrated Nafion have the sulfonate group immersed within the confines of an aqueous
phase.(16, 25-29) However, Hamrock suggests that solvation of COC(A) is unlikely because of
the electron withdrawing effects of the fluorine atoms.(30) This motivated our Natural Bond
Orbital (NBO) analysis(31) of dimethyl ether versus the perfluorinated analog. The charge
62
density of oxygen in the perfluorinated analog is 48 % that of oxygen in dimethyl ether.3 Thus
solvation of the COC(A) group in water, or complexation to cations is unlikely. Consideration
of the Nafion spectrum as a series of mechanically coupled group modes allows for proper
analysis of the shift in the 970 cm-1
and 1060 cm-1
bands. Specifically, any normal mode with
SO3- MVM’s are expected to concurrently vary with changes in the exchange group environment
or changes due to synthetic modification of the ionomer. The composite spectrum (Fig. 3 middle)
provides an overview of where these regions of the spectrum are. This is why the 970 cm-1
and
the 1060 cm-1
bands shift concertedly with changes in the sulfonate environment (e.g. they both
vanish with rigorous dehydration(15)). The 983ATR
cm-1
peak is another example where
composite spectra can be used to predict which peaks are sensitive to derivitization. Figure 3
(middle) shows that COC(A) νas and CF3 νs MVM’s contribute to both 970* cm-1
and 973* cm-1
.
It is therefore reasonable to assume that if these functionalities were not present, there would be
significant change in the experimentally obtained spectrum in this region. This is observed in the
spectrum of PFEBSA where the 985 cm-1
peak found in the PFMHSA spectrum missing. The
presence of the 962ATR
cm-1
peak in the PFEBSA also supports the correlation of the Nafion
970ATR
cm-1
to a SO3 νs MVM.
Previously, we assigned the 984* cm-1
mode to primarily SO3- s with a substantially
smaller contribution from COC(A) as through visualization of normal mode animations.(15)
The MVM spectra show that CF2 motions are greater contributor to 984* cm-1
than the COC(A)
as. Additionally, the s(CF3) contribution to the multiplet of bands from 940 – 1010 cm-1
was
elucidated by the MVM spectra. The animations showed the dominant 1059* cm-1
pure mode as
as(COC(A)) with a lesser contribution from SO3-. The quantitative MVMs show the SO3
- νs
3 NBO 5.0 is incorporated within the Jaguar code that is commercially available from Schrodinger.
63
đ2contributing 2% to the normal mode. The importance of this 2% is scaled by the fact that the
S-O bonds have the largest force constants. Table 2 shows the advantage of MVM assignments
over previous assignments based on visualization of eigenvector animations.
DFT tends to overestimate vibrational frequencies by a few percent: Scaling factors are
applied to calculated peak frequencies.(33) However the applications of scaling factors to
calculated spectra of chemical repeat units has caveats. Bower and Maddams(11) explain that
polymer chemical repeat units are not physical repeat units: Hydrated ion exchange membranes
have geometrical irregularities that include interfacial regions separating complex aqueous
confines from hydrophobic regions.(28, 34, 35) A chemical repeat unit can be approximated as a
separate molecule in its own local environment. This is applicable to the motion of side-groups,
with little motion of the backbone. Broad infrared absorption peaks are expected because groups
of the same atoms in separate repeat units have frequencies that vary as a consequence of the
different physical environments. Atoms in close proximity to the ion exchange site would exhibit
more substantial frequency shifts with changes in the environment.(36, 37) This can scramble the
order that normal modes appear in DFT line spectra in comparison to the order experimentally
observed. This severely limits the reliability of using DFT intensities and frequencies for
deconvolution and assignment of resulting bands.
Table 4.2 Pure mode assignments from visualization of normal mode animations in comparison to
MVM assignments of this work.
ν (cm-1
) (15) Legacy assignments Pure mode components MVM assignments 984* COC(A) νas SO3
- νs, COC(A) νas, COC(B) ρr (15, 18) SO3- νs, CF2(2) ω, CF2 (1) νs, BBStr,OOP, COC(A) νas,
1059* SO3- νs COC(A) νas, SO3
- νs (15, 18) COC(A) νas, CF2(2) νs, COC(B) νs, SO3- νs, CF3 νas,
1168* - BBStr, CF2 ρr, COC(A) ω(32) CF2(2) νas, COC(A) ω, COC(B) τ, BBStr,OOP, SO3- νas
1254* - CF3 νas, COC(A) δS, COC(B) δS25
CF3 νas, COC(B) ω, BBStr,OOP,COC(A) ω, C1-C2Str
1299* - C1-C2Str, BBStr(32) C1-C2Str, BBStr, COC(A) δS, C3-C4Str, COC(B) δS
1302* - BBStr, C1-C2Str(32) BBStr, C1-C2Str, COC(B) δS, COC(A) δS, C3-C4Str
1322* - BBStr,OOP(32) BBStr, COC(B) ω, CF3 νas, COC(A) νas, CF2(2) νas
1357* - BBStr,OOP, COC(B) ρr(32) BBStr, COC(B) δS, CF3 νs bend, COC(A) τ, C1-C2Str
64
Experimental data are essential for correlation of DFT calculated peaks to experimental
bands. Consider the 940 – 1010 cm-1
multiplet, and the1060 cm-1
band. If the criterion of
nearness of a theoretical peak to an experimental band is used, the 984* cm-1
mode would be
correlated with the green band centered at 983TR
cm-1
. However, the band at 983TR
cm-1
is
entirely insensitive to changes to the SO3- environment. The MVM spectra show that the 984*
cm-1
mode is predominantly an SO3- s mode. Thus we associate the 984* cm
-1 with the
experimental 969TR
cm-1
peak because of its high sensitivity to ion exchange and changes in the
state-of-hydration.(15) The correlation of the sulfonate MVM segments at 973*, 982*, and 984*
cm-1
(light brown) with the 969TR
cm-1
and/or the 971ATR
cm-1
bands of our previous studies(15,
18) are further supported by the extensive SO3
- MVMs within the multiplet of bands between 940
cm-1
and 1010 cm-1
.
Fortunately the ion exchange group, sensitive to the state of hydration and the nature of
the counter-ions, facilitates the correlation of theoretical lines to experimental bands.(15, 18) A
direct correlation of theoretical lines to experimental bands without additional data (e.g., ion
exchange, dehydration or synthetic derivatization) would require knowledge of the physical
repeat unit under the conditions in which the spectra are obtained. Molecular dynamics may be a
source of more realistic physical repeat units.(38, 39)
4.4 Conclusion:
Single-functional-group normal mode assignments have been a persistent source of confusion
to the ionomer electrolyte community. The ionomer side-chain functional groups are distributed
between the restricted confines of aqueous and hydrophobic regions. It has previously been
impossible to explain how variations in the exchange group environment cause concerted shifts
of peaks in disparate parts of the IR spectra within the framework of individual functional group
65
assignments. The mechanical coupling of the internal coordinates of side-chain functional
groups that can reside in different phases (e.g., aqueous or hydrophobic) explains the complex
behavior of IR peaks that accompanies ion exchange, variations in the state-of-hydration and
chemical derivatization. The representation of functional group contributions to normal modes,
as subsets of the generalized coordinates, provides a set of eigenvectors now introduced as
minimal vibrational modes. The addition of a categorical axis representing minimal vibrational
modes coded as colors adds a dimension (i.e., wavenumbers, intensities, colors) to IR
spectroscopy that clarifies the complex effects of the mechanical coupling of internal coordinates
of functional groups in ionomer electrolytes.
References:
1. Hickner, M.A., Ion-containing polymers: new energy & clean water. Materials Today,
2010. 13(5): p. 34-41. 2. Hamrock, S.J. and Yandrasits, M.A., Proton exchange membranes for fuel cell
applications. Polymer Reviews, 2006. 46(3): p. 219-244. 3. Li, S.K., Zhu, H., and Higuchi, W.I., Enhanced transscleral iontophoretic transport with
ion-exchange membrane. Pharmaceutical Research, 2006. 23(8): p. 1857-1867. 4. Allegrezza, A.E., Jr., et al., Chlorine resistant polysulfone reverse osmosis modules.
Desalination, 1987. 64: p. 285-304. 5. Mauritz, K.A. and Moore, R.B., State of understanding of Nafion. Chemical Reviews,
2004. 104(10): p. 4535-4585. 6. Elabd, Y.A., et al., Transport properties of sulfonated poly (styrene-b-isobutylene-b-
styrene) triblock copolymers at high ion-exchange capacities. Macromolecules, 2006. 39(1): p. 399-407.
7. Elabd, Y.A. and Hickner, M.A., Block Copolymers for Fuel Cells. Macromolecules, 2011. 44(1): p. 1-11.
8. Ploense, L., et al., Proton spillover promoted isomerization of n-butylenes on Pd-black cathodes/Nafion 117. Journal of the American Chemical Society, 1997. 119(47): p. 11550-11551.
9. Liu, Z.J., et al., Reductive dehalogenation of gas-phase chlorinated solvents using a modified fuel cell. Environmental Science & Technology, 2001. 35(21): p. 4320-4326.
10. Hernandez-Pagan, E.A., et al., Resistance and polarization losses in aqueous buffer-membrane electrolytes for water-splitting photoelectrochemical cells. Energy & Environmental Science, 2012. 5(6): p. 7582-7589.
66
11. Bower, D.I. and Maddams, W.F., The Vibrational Spectroscopy of Polymers. 1989, Cambridge: Cambridge University Press.
12. Korzeniewski, C., Snow, D.E., and Basnayake, R., Transmission infrared spectroscopy as a probe of Nafion film structure: Analysis of spectral regions fundamental to understanding hydration effects. Applied Spectroscopy, 2006. 60(6): p. 599-604.
13. Basnayake, R., Wever, W., and Korzeniewski, C., Hydration of freestanding nation membrane in proton and sodium ion exchanged forms probed by infrared spectroscopy. Electrochimica Acta, 2007. 53(3): p. 1259-1264.
14. ISI Web of Knowledge Home Page,, 2009. 15. Webber, M., et al., Mechanically Coupled Internal Coordinates of Ionomer Vibrational
Modes. Macromolecules, 2010. 43(13): p. 5500-5502. 16. Cable, K.M., Mauritz, K.A., and Moore, R.B., Effects of hydrophilic and hydrophobic
counterions on the Coulombic interactions in perfluorosulfonate ionomers. Journal of Polymer Science, Part B: Polymer Physics, 1995. 33(7): p. 1065-72.
17. Warren, D.S. and McQuillan, A.J., Infrared spectroscopic and DFT vibrational mode study of perfluoro(2-ethoxyethane) sulfonic acid (PES), a model Nafion side-chain molecule. Journal of Physical Chemistry B, 2008. 112(34): p. 10535-10543.
18. Kendrick, I., et al., Elucidating the Ionomer-Electrified Metal Interface. J. Am. Chem. Soc., 2010. 132(49): p. 17611-17616.
19. Hohenberg, P. and Kohn, W., Inhomogeneous Electron Gas. Physical Review B, 1964. 136(3B): p. B864-&.
20. Kohn, W. and Sham, L.J., Self-Concsistent Equations Including Exchange And Correlation Effects. Physical Review, 1965. 140(4A): p. 1133-&.
21. Xu, X., et al., An extended hybrid density functional (X3LYP) with improved descriptions of nonbond interactions and thermodynamic properties of molecular systems. Journal of Chemical Physics, 2005. 122(1): p. 14.
22. Frisch, M.J., Pople, J.A., and Binkley, J.S., Self-Consistent Molecular-Orbital Methods 25. Supplementary Functions For Gaussian-Basis Sets. Journal of Chemical Physics, 1984. 80(7): p. 3265-3269.
23. Clark, T., et al., Efficient Diffuse Function-Augmented Basis-Sets For Anion Calculations 3. The 3-21+G Basis Set For 1ST-Row Elements, LI-F. Journal of Computational Chemistry, 1983. 4(3): p. 294-301.
24. Danilczuk, M., et al., Understanding the fingerprint region in the infra-red spectra of perfluorinated ionomer membranes and corresponding model compounds: Experiments and theoretical calculations. Journal of Power Sources, 2011. 196(20): p. 8216-8224.
25. Lehmani, A., Durand-Vidal, S., and Turq, P., Surface morphology of Nafion 117 membrane by tapping mode atomic force microscope. Journal of Applied Polymer Science, 1998. 68(3): p. 503-508.
26. Moore, R.B., III and Martin, C.R., Morphology and chemical properties of the Dow perfluorosulfonate ionomers. Macromolecules, 1989. 22(9): p. 3594-9.
27. Gierke, T.D. and G. E. Munn, F.C.W., The morphology in Nafion perfluorinated membrane products, as determined by wide- and small-angle x-ray studies. Journal of Polymer Science: Polymer Physics Edition, 1981. 19(11): p. 1687-1704.
67
28. Hsu, W.Y. and Gierke, T.D., Ion transport and clustering in nafion perfluorinated membranes. Journal of Membrane Science, 1983. 13(3): p. 307-326.
29. Yeager, H.L. and Steck, A., Cation and Water Diffusion in Nafion Ion Exchange Membranes: Influence of Polymer Structure. Journal Of The Electrochemical Society, 1981. 128(9): p. 1880-1884.
30. Hamrock, S. 2010: St. Paul, MN. 31. Weinhold, F. and Landis, C., Natural bond order and extensions of localized bonding
concepts. Chem. Educ. Res. Pract. Eur, 2001. 2(2): p. 91-104. 32. Byun, C.K., et al., Thermal Processing as a Means to Prepare Durable, Submicron
Thickness Ionomer Films for Study by Transmission Infrared Spectroscopy. Analytical Chemistry, 2012. 84(19): p. 8127-8132.
33. Dimakis, N., et al., A band dispersion mechanism for Pt alloy compositional tuning of linear bound CO stretching frequencies. Journal of Physical Chemistry B, 2005. 109(5): p. 1839-1848.
34. Elliott, J.A. and Paddison, S.J., Modelling of morphology and proton transport in PFSA membranes. Physical Chemistry Chemical Physics, 2007. 9(21): p. 2602-2618.
35. Gebel, G. and Lambard, J., Small-angle scattering study of water-swollen perfluorinated ionomer membranes. Macromolecules, 1997. 30(25): p. 7914-7920.
36. Moukheiber, E., et al., Investigation of ionomer structure through its dependence on ion exchange capacity (IEC). J. Membr. Sci., 2012. 389: p. 294-304.
37. Kumari, D., Vibrational spectroscopy of ion exchange membranes. 2011, Northeastern University.
38. Brandell, D., et al., Molecular dynamics studies of the Nafion (R), Dow (R) and Aciplex (R) fuel-cell polymer membrane systems. Journal of Molecular Modeling, 2007. 13(10): p. 1039-1046.
39. Brandell, D., Karo, J., and Thomas, J.O., Modelling the Nafion (R) diffraction profile by molecular dynamics simulation. Journal of Power Sources, 2010. 195(18): p. 5962-5965.
68
Chapter 5: Operando Infrared Spectroscopy of Ethanol Oxidation
in Polymer Electrolyte Fuel Cells
5.1 Introduction:
Ethanol and methanol are an attractive alternative to hydrogen gas in proton exchange
membrane (PEM) fuel cells. When compared to hydrogen gas, the production, storage and
delivery of these fuels can be achieved with less modification to existing infrastructure. The
theoretical mass energy density for methanol and ethanol is 6.1 and 8.0 kWh/kg, respectively,
close to that of gasoline. (1) Ethanol is significantly less toxic to methanol and can be obtained
from biomass. The rate of ethanol crossover across a proton exchange membrane is lower than
that of methanol.(2) The crossover of ethanol or methanol reduces the number of sites for
oxygen reduction and creates a mixed potential at the cathode reducing cell voltage.(3) While
platinum is generally considered the best catalyst for the oxidation of ethanol and methanol,
cleaving the C-C bond presents a significant challenge with oxidizing ethanol. The cleavage of
the C-C bond leads to the complete oxidation of ethanol yielding CO2 and generating 12
electrons. Acetic acid is the result of an incomplete oxidation pathway yielding four electrons.
Several studies have investigated the oxidation pathway of ethanol on a Pt electrode using both
in-situ Fourier transform infrared spectroscopy (FTIR) (4-13) and differential electrochemical
mass spectroscopy (DEMS)(4, 9, 12, 14, 15). Each of these techniques has their drawbacks. It is
difficult to study the evolution of acetic acid in ethanol oxidation using DEMS due to its low
volatility. In-situ IR techniques using liquid electrolytes often have uneven current distribution
in the thin layer between the working electrode and the IR window resulting in observing
oxidation products associated with lower potentials despite the cell being set to a higher
69
potential.(16) Peaks in the IR spectrum resulting from the electrolytes used in these in-situ
methods (e.g. sulfuric and perchloric acids) often overlap portions of the IR spectrum useful in
identifying products of ethanol oxidation. It is better to study the products of ethanol oxidation
under conditions associated with an operating fuel cell. Operando IR spectroscopy allows for the
acquisition of IR spectra under actual operating conditions without interference from
supplemental electrolytes. In spectra acquired using operando spectroscopy adsorbed species
are observed as bipolar peaks,(17) often eliminating using polarization techniques in
differentiating between adsorbed and desorbed compounds. The oxidation of ethanol on Pt, Ru
and PtRu catalysts are studied using operando spectroscopy. All of the principle products of
ethanol oxidation are detected using a single technique under real world operation conditions.
This demonstrates operando spectroscopy’s viability in the study of new ethanol oxidation
catalysts.
5.2 Experimental
Membrane electrode assembly preparation: Nafion 117 (Dupont, was immersed in
boiling 8 M nitric acid for 20 minutes, rinsed with Nanopure™ water and immersed in boiling
water for one hour. Catalyst inks are comprised of the catalyst nanoparticles (4 mg/cm2) and 5
wt% Nafion ionomer solution (Sigma Aldrich, Milwaukee, WI) dispersed in Nanopure™ water.
Inks were applied to a 5 cm2 portion of Nafion immobilized on a heated vacuum table at 70 °C.
Carbon paper gas diffusion layers (Toray Industries, Tokyo, Japan) are placed between each flow
field and electrode during fuel cell assembly. When charged with hydrogen, a Pt electrode
served at both the counter and reference electrode. The working electrodes used for this study
were: Pt (Johnson Matthey), Ru (Sigma-Aldrich) and PtRu (1:1, Johnson-Matthey). Membrane
electrode assemblies (MEA’s) were initially conditioned in a fuel cell operating at 50 °C by
70
cycling the potential from 800 mV to 600 mV at a rate of 40 mV/min five times. Anode and
cathode reactant streams were humidified H2 (50 sccm) and air (200 sccm) respectively.
Operando Spectroscopy: Operando specular reflectance IR spectra were obtained using
a fuel cell described by Lewis et al.(18) Briefly, a CaF2 window in the upper flow field and an
aperture in the gas diffusion layer, exposes the working electrode to the IR beam. The cell
interfaces with a commercially available diffuse reflection stage (Pike Technologies, Madison,
WI) connected to a Vertex 70 spectrometer (Bruker). Spectra were obtained by averaging 100
scans at 4 cm-1
resolution using a liquid nitrogen cooled MCT detector. The cell, operating at 50
°C, was fed humidified H2 (50 sccm) and N2 (200 sccm) to the counter/reference and working
electrodes, respectively. Immediately prior to acquiring spectra, the working electrode was
conditioned by cycling the potential from 0 mV to 1200 mV 50 times. Ethanol vapor was fed to
the cell using a modified GOW- MAC 350 gas chromatograph (Bethlehem, PA) being fed 10 M
ethanol at a rate of 2.5 µL/min with a N2 carrier gas (60 sccm). Once the ethanol feed began, the
cell potential was held at 0 mV and equilibrated for 30 minutes and a reference spectrum was
obtained. Potential dependent spectra were obtained at 100 mV increments between 100 mV and
900 mV. Before increasing the potential, reference spectra were taken 0 mV to maintain a
consistent baseline and to reduce contamination from atmospheric CO2 and water.
5.3 Results and discussion:
At low potentials adsorbed carbon monoxide is produced via the breaking of the C-C
bond of the adsorbed acetaldehyde species.(19, 20) Adsorbed species are observed in operando
FTIR spectra as bipolar peaks.(17) Carbon monoxide adsorption to the catalytic surface inhibits
catalysis by blocking sites for the adsorption of ethanol. At higher potentials, adsorption of
hydroxide allows for the oxidation of CO to CO2 freeing up sites for additional
71
ethanol/acetaldehyde adsorption.(11) The potential dependent shift in the CO peak, referred to
as Stark tuning, is the result of back donation of dπ electrons from the electrode into the 2π*
orbital of CO.(21) Figure 1 (left) shows the Stark tuning curve of CO adsorbed onto a Pt
electrode. The complex Stark tuning behavior of CO on Pt in an acidic medium has been
previously discussed.(22, 23) Briefly, the peak position of adsorbed CO increases linearly until
the potential is high enough to allow for adsorption of –OH and subsequent oxidation of CO
causing a drop in peak position caused by decreased dipole-dipole interactions. At higher
potentials, the adsorption of the Nafion sulfonate and CF3 groups onto Pt re-establishes dipole-
dipole interaction between the Nafion adsorbed species and the remaining clusters of adsorbed
CO thereby increasing the CO peak position.(23) The Stark tuning curve and potential dependent
spectra of figure 1 agree with previous studies reporting the onset of CO oxidation at 500 mV
with a sharp increase in oxidation at 650 mV.(24, 25) The Operando Stark tuning curves
demonstrate the sensitivity of adsorbed peak position to dipole-dipole interactions. The amount
of CO oxidized is low enough to relax the dipole-dipole interaction (i.e. lower peak position), yet
the amount of CO2 observed in the FTIR spectrum were barely noticeable at potentials lower
than 700 mV.
Alloying Ru to Pt yields a catalyst showing both a high activity for ethanol oxidation
while allowing for CO oxidation at much lower potentials. Ruthenium on its own, shows poor
kinetics for breaking the C-C bond of ethanol but adsorbs –OH at much lower potentials (300
mV) than Pt (26, 27). In the literature, there are two generally accepted theories to explain this
phenomenon. The alloying of Ru to Pt contributes to the electron density of Pt thereby
weakening Pt-CO bond.(28-32) This theory is supported by DFT studies showing either no Pt-C
bond contacting or bond elongation expected with electron back donation.(33) Alternatively, the
72
bifunctional theory states that Ru activates water at lower potentials yielding a tightly bound Ru-
OH species. The adsorbed CO on Pt sites to interact with Ru-OH sites allowing for its oxidation
to CO2.(27, 34-39) The spectra of figure 2 show the potential dependent spectra of ethanol on a
PtRu electrode (1:1). The spectra agree with Lima and Gonzales(40), whose electrochemical
studies of ethanol studies on PtRu show an onset of CO oxidation at 400 mV.
Figure 5.1. Stark tuning curve of CO adsorbed onto Pt (left). Potential dependent spectra of a
Pt black electrode in the presence of ethanol vapor (right). This region of the spectrum focuses
on peaks associated with adsorbed CO (~2065 cm-1
) and CO2 (2350 cm-1
).
73
Figure 5.2. Stark tuning curve of CO adsorbed onto PtRu (right). Potential dependent spectra of
an unsupported PtRu electrode in the presence of ethanol vapor (left). This region of the
spectrum focuses on peaks associated with adsorbed CO (~2060 cm-1
) and CO2 (2350 cm-1
).
Oxidation Mechanism
The oxidation mechanism of ethanol on Pt electrodes has been extensively studied.(2, 11,
12, 41-48) The potential dependent products of ethanol oxidation are acetaldehyde, acetic acid
and CO2. In a generally accepted mechanism, the initial step of ethanol oxidation is its
adsorption at the carbon alpha to the oxygen. The transfer of two electrons from the adsorbed
species yield acetaldehyde which dissociates from the surface. The acetaldehyde then re-adsorbs
at low potentials at which point one of two things can happen: 1) The C-C bond is broken
yielding CO or 2) The acetaldehyde is oxidized to acetic acid. In the former scenario, the
ethanol is completely oxidized yielding twelve electrons whereas in the latter, only 4 electrons
are produced. This is summarized in figure 3 below. Desorbed acetaldehyde is generally
74
observed around 400 mV with acetic acid in lower amounts. Bands corresponding to both
products intensify at 600 mV coinciding with and increased availability in adsorption sites due to
the oxidation of adsorbed CO.(13) Equally important are adsorbed species. As previously
discussed, ethanol is adsorbed at low potentials at the alpha carbon and two Pt atoms.(11, 12)
The band corresponding to adsorbed C-adsorbed ethanol is observed 1256 cm-1
.(12)
Acetaldehyde can be adsorbed at the carbon, oxygen or both.(49-53) All three modes of
adsorption yield bands from 1645-1665 cm-1
. The two oxygen atoms of acetic acid adsorb to a
Pt surface via two proposed mechanisms: Each oxygen is bound to its own unique Pt atom or
both oxygens are bound to a single Pt atom.(54) Surface bound acetic acid is referred to as
adsorbed acetate. In the case of Pt single-crystal electrodes, the bands corresponding to adsorbed
acetate form generally form between 300-350 mV but in the case of Pt(110), adsorbed acetate
can be found at 100 mV(54). Adsorbed acetate bands are found at 1395-1420 cm-1
. (7, 54-56)
75
Figure 5.3. A general mechanism for the oxidation of ethanol on Pt. For brevity, adsorbed
species are depicted only once despite the existence of several bonding configurations.
Reproduced from (11, 57)
In order fully elucidate the oxidation mechanism of ethanol on Pt, or any other catalyst, it
is not enough to detect the presence of desorbed species but the potential dependent evolution of
adsorbed species must also be identified. An advantage of studying ethanol oxidation using
operando techniques is that both of these types of products can be observed. Figure 4 is a series
of spectra taken of a Pt electrode in the presence of ethanol vapor under normal operating
conditions.
76
Figure 5.4. Potential dependent spectra of a Pt black electrode in the presence of ethanol vapor.
Vapor phase acetaldehyde and acetic acid were obtained ex-situ.
1900 1800 1700 1600 1500 1400 1300 1200 1100 1000
1394
1627
1258
1066
1369
1763
Wavenumber (cm-1)
Acetic Acid
Acetaldehyde
900 mV
800 mV
700 mV
600 mV
500 mV
400 mV
300 mV
200 mV
0 mV
100 mV
77
The spectra in figure 5.4 illustrate and follow the accepted mechanism for ethanol oxidation on
Pt. In order to distinguish between adsorbed and desorbed species, recall that bands
corresponding to adsorbed species are observed as bipolar bands.(17) At 100 mV bands
corresponding to adsorbed CO are observed. Also at 100 mV, bipolar peaks at 1258 cm-1
and
1066 cm-1
appear. These to peaks correspond to C and O adsorbed ethanol respectively.(12)
While the observation of the 1258 cm-1
has been reported, this is the first time that a peak
corresponding to O adsorbed ethanol has been observed.(12) Iwasita and Pastor reported the
possibility of O adsorbed ethanol theorizing a corresponding band from 1000 cm-1
and 1050 cm-
1, noting that a dipole moment of a C-O bond perpendicular to the surface would be small and
difficult to observe. Previous in-situ ethanol FTIR studies on Pt used HClO4 as an electrolyte.
ClO4- peaks are observed at 1033 cm
-1 (58) and dominate this region of the spectrum making the
observation of O adsorbed ethanol under non-operando conditions unlikely. Bipolar peaks
appearing at 1627 cm-1
are associated with adsorbed acetaldehyde(50, 51) evolve at the same
potentials as peaks associated with adsorbed ethanol. Adsorbed acetic acid is observed starting
at 200 mV around 1394 cm-1
. This peak position is slightly lower than that reported by Shin et
al, however, the potential at which these peaks evolve is consistent with previous studies and
subsequent experiments in this report (vida infra).(7, 54) At 400 mV, small amounts of adsorbed
CO begin to oxidize to CO2 decreasing CO coverage and allowing for other adsorbates. This is
correlated in the spectra of figure 4 by a marked increase in the intensities of peaks
corresponding to all the other adsrobed species. Acetaldehyde is the intermediate produced
between adsorption of ethanol and the cleavage of the C-C bond. Therefore it is not unexpected
that desorbed acetaldehyde is detected in potentials above 400 mV with and marked increase in
production above 600 mV coinciding with CO oxidation. Since Pt is shows activity for cleaving
78
the C-C bond of ethanol, CO2 is observing in signficant ammounts. It is difficult to ascertain the
amount of desorbed acetic acid produced using this technique. Any carbonyl peaks arising from
acetic acid (1730 cm-1
) would be overwhelmed by carbonyl peaks related to acetaldehyde.
To demonstrate its efficacy on catalysts outside of Pt, the operando methodolgy was used
to study ethanol oxidation on Ru and PtRu. Figure 5 is a series of a Ru catalyst in the presence
of ethanol at increasing potentials. Recall that compared to Pt, Ru displays lower activity
towards cleaving the C-C bond of ethanol. Therefore it reasonable to expect a majority of the
peaks in the spectra of figure 5 to correspond to acetic acid and its adsorbed intermediates.
Peaks corresponding to ethanol adsorption on Ru (1258 cm-1
and 1065 cm-1
) do not arise until
possibly 300 mV with O adsorption favored compared to Pt. The two bands used to identify
acetic acid in figure occur at 1730 cm-1
and 1172 cm-1
with adsorbed acetate appearing at 1417
cm-1
. The presense of adsorbed water is needed for the oxidation of acetaldehyde to acetic acid.
Because Ru adsorbs water at lower potentials than Pt, the oxidation of acetaldehyde will also
occur at lower potentials. This explains why peaks corresponding to desorbed acetaldehyde are
observed at significantly lower intensities compared to Pt.
79
Figure 5.5. Potential dependent spectra of a Ru black electrode in the presence of ethanol vapor.
Vapor phase acetaldehyde and acetic acid were obtained ex-situ.
The potential dependent spectra of a PtRu (1:1) catalyst in the presense of ethanol vapor
is shown in figure 6. The bifunctional mechanism ethanol oxidation will take place on the Pt
sites while Ru sites adsorb activated water species(59). The result is a lower potential at which
1900 1800 1700 1600 1500 1400 1300 1200 1100 1000
1065
1172
1276
1417
Acetic Acid
Acetaldehyde
900 mV
800 mV
700 mV
600 mV
500 mV
400 mV
300 mV
200 mV
100 mV
Wavenumber (cm-1)
0 mV
80
CO oxidizes, lowering coverage and freeing up sites for further ethanol oxidation. As shown
earlier in figure 2, CO2 evolves at 200 mV on PtRu coinciding with the plateau of the Stark
tuning curve. The shift in the CO band position signifies a relaxed dipole-dipole interaction from
lower CO coverage.(22) However, despite a lower CO coverage, there are less Pt sites making
complete oxidation of ethanol not as favorable than on pure Pt. This is exemplified by the
substantially lower signal for bipolar peaks indicative of ethanol adsorption. Consequently, there
is stronger signal of acetaldehyde relative to that of Pt (Fig 4) as there are fewer sites for
acetaldehyde to readsorb for complete oxidation.
81
Figure 5.6. Potential dependent spectra of a PtRu black electrode in the presence of ethanol
vapor. Vapor phase acetaldehyde and acetic acid were obtained ex-situ.
1900 1800 1700 1600 1500 1400 1300 1200 1100 1000
Wavenumber (cm-1)
Acetic Acid
13731056
1763
Acetaldehyde
900 mV
800 mV
700 mV
600 mV
500 mV
400 mV
300 mV
200 mV
100 mV
0 mV
1256
82
5.4 Conclusion
Operando fuel cell spectroscopy allows for elucidation of the ethanol oxidation pathway
on a variety of cataysts. Both adsorbed and desorbed species can be observed and differentiated
without employing a polarized light source. By studying this process with an electrolyte system
without any mobile anions, previously unreported peaks associated with O-adsorbed ethanol
have been observed. Under these conditions it is clear that the complete oxidation of ethanol is
favored over partial pathway. The appearance of strong absorptions consitant with the presence
CO, CO2 and acetaldehyde support this claim. This methodology operates under conditions that
reflect real world conditions providing data with more relevancy than in-situ techniques and will
be a useful tool in characterizing new catalysts.
References:
1. Lamy, C. and Leger, J.M., FUEL-CELLS - APPLICATION TO ELECTRIC VEHICLES. Journal De Physique Iv, 1994. 4(C1): p. 253-281.
2. Wang, J., Wasmus, S., and Savinell, R.F., Evaluation of Ethanol, 1-Propanol, and 2-Propanol in a Direct Oxidation Polymer-Electrolyte Fuel Cell. J. Electrochem. Soc., 1995. 142(12): p. 4218.
3. Rivera, H., et al., Effect of Sorbed Methanol, Current, and Temperature on Multicomponent Transport in Nafion-Based Direct Methanol Fuel Cells. The Journal of Physical Chemistry B, 2008. 112(29): p. 8542-8548.
4. Bittins Cattaneo, B., et al., Intermediates and Products of Ethanol Oxidationon Platinum in Acid-Solution. Berichte Der Bunsen-Gesellschaft-Physical Chemistry Chemical Physics, 1988. 92(11): p. 1210-1218.
5. Leung, L.W.H., Chang, S.C., and Weaver, M.J., Real-Time FTIR Spectroscopy as an Electrochemical Mechanistic Probe - Electrooxidation of Ethanol and Related Species on Well-Defined Pt(111) Surfaces. Journal of Electroanalytical Chemistry, 1989. 266(2): p. 317-336.
6. Holze, R., On the Adsorption and Oxidation of Ethanol on Platinum as Studied with in-situ IR Spectroscopy. Journal of Electroanalytical Chemistry, 1988. 246(2): p. 449-455.
7. Shin, J., et al., Elementary steps in the oxidation and dissociative chemisorption of ethanol on smooth and stepped surface planes of platinum electrodes. Surface Science, 1996. 364(2): p. 122-130.
83
8. Vigier, F., et al., Development of anode catalysts for a direct ethanol fuel cell. Journal of Applied Electrochemistry, 2004. 34(4): p. 439-446.
9. Iwasita, T., Fuel cells: Spectroscopic studies in the electrocatalysis of alcohol oxidation. Journal of the Brazilian Chemical Society, 2002. 13(4): p. 401-409.
10. Iwasita, T., et al., A SNIFTIRS Study of Ethanol Oxidation on Platinum. Electrochimica Acta, 1989. 34(8): p. 1073-1079.
11. Hitmi, H., et al., A kinetic analysis of the electro-oxidation of ethanol at a platinum electrode in acid medium. Electrochimica Acta, 1994. 39(3): p. 407-415.
12. Iwasita, T. and Pastor, E., A DEMS and FTIR Spectroscopic Investigation of Adosbed Ethanol on Polycrystaline Platinum. Electrochimica Acta, 1994. 39(4): p. 531-537.
13. Xia, X.H., Liess, H.D., and Iwasita, T., Early stages in the oxidation of ethanol at low index single crystal platinum electrodes. Journal of Electroanalytical Chemistry, 1997. 437(1-2): p. 233-240.
14. de Souza, J.P.I., et al., Electro-oxidation of ethanol on Pt, Rh, and PtRh electrodes. A study using DEMS and in-situ FTIR techniques. Journal of Physical Chemistry B, 2002. 106(38): p. 9825-9830.
15. Willsau, J. and Heitbaum, J., Elementary Steps of Ethanol Oxidation on Pt in Sulfuric-Acid as Evidenced by Isotope Labeling. Journal of Electroanalytical Chemistry, 1985. 194(1): p. 27-35.
16. Iwasita, T. and Vielstich, W., The electrochemical oxidation of ethanol on platinum: a SNIFTIRS study. Journal of electroanalytical chemistry and interfacial electrochemistry, 1988. 257(1-2): p. 319-324.
17. Burgi, T., ATR-IR spectroscopy at the metal-liquid interface: influence of film properties on anomalous band-shape. Physical Chemistry Chemical Physics, 2001. 3(11): p. 2124-2130.
18. Lewis, E.A., et al., Operando X-ray absorption and infrared fuel cell spectroscopy. Electrochimica Acta, 2011. 56(24): p. 8827-8832.
19. Shao, M.H., et al., In situ ATR-SEIRAS study of electro oxidation of dimethyl ether on a Pt electrode in acid solutions. Electrochemistry Communications, 2005. 7(5): p. 459-465.
20. Souza-Garcia, J., Herrero, E., and Feliu, J.M., Breaking the Oxidation Reaction on Platinum Electrodes: Effect of Steps and Ruthenium Adatoms. ChemPhysChem, 2010. 11(7): p. 1391-1394.
21. Blyholder, G., Molecular orbital view of chemisorbed carbon monoxide. Journal of Physical Chemistry, 1964. 68(10): p. 2772-8.
22. Stamenkovic, V., et al., Vibrational properties of CO at the Pt(111)-solution interface: the anomalous stark-tuning slope. Journal of Physical Chemistry B, 2005. 109(2): p. 678-680.
23. Kendrick, I., et al., Elucidating the Ionomer-Electrified Metal Interface. J. Am. Chem. Soc., 2010. 132(49): p. 17611-17616.
24. Salgado, J.R.C., et al., Carbon monoxide and methanol oxidation at platinum catalysts supported on ordered mesoporous carbon: the influence of functionalization of the support. Physical Chemistry Chemical Physics, 2008. 10(45): p. 6796-6806.
84
25. Bellows, R.J., Marucchi-Soos, E.P., and Buckley, D.T., Analysis of Reaction Kinetics for Carbon Monoxide and Carbon Dioxide on Polycrystalline Platinum Relative to Fuel Cell Operation. Industrial & Engineering Chemistry Research, 1996. 35(4): p. 1235-1242.
26. Tanaka, S., et al., Preparation and evaluation of a multi-component catalyst by using a co-sputtering system for anodic oxidation of ethanol. Journal of Power Sources, 2005. 152(1): p. 34-39.
27. Lu, C., et al., UHV, Electrochemical NMR, and Electrochemical Studies of Platinum/Ruthenium Fuel Cell Catalysts. Journal of Physical Chemistry B, 2002. 106(37): p. 9581-9589.
28. Goodenough, J.B., et al., Methanol Oxidation on Unspported and Carbon Supported Pt + Ru Anodes. Journal of Electroanalytical Chemistry, 1988. 240(1-2): p. 133-145.
29. Iwasita, T., Nart, F.C., and Vielstich, W., An FTIR Study of the Catalytic Activity of a 85-15 Pt-Ru Alloy for Methanol Oxidation. Berichte Der Bunsen-Gesellschaft-Physical Chemistry Chemical Physics, 1990. 94(9): p. 1030-1034.
30. Krausa, M. and Vielstich, W., Study of the Electrocatalytic Influence of Pt/Ru and Ru on the Oxidation of Residues of Small Organic-Molecules. Journal of Electroanalytical Chemistry, 1994. 379(1-2): p. 307-314.
31. Frelink, T., Visscher, W., and vanVeen, J.A.R., Measurement of the Ru surface content of electrocodeposited PtRu electrodes with the electrochemical quartz crystal microbalance: Implications for methanol and CO electrooxidation. Langmuir, 1996. 12(15): p. 3702-3708.
32. Frelink, T., Visscher, W., and Vanveen, J.A.R., On the Role of Ru and Sn as Promoters of Methanol Electrooxidation Over Pt. Surface Science, 1995. 335(1-3): p. 353-360.
33. Dimakis, N., et al., A band dispersion mechanism for Pt alloy compositional tuning of linear bound CO stretching frequencies. Journal of Physical Chemistry B, 2005. 109(5): p. 1839-1848.
34. Wang, J.X., et al., In situ X-ray reflectivity and voltammetry study of Ru(0001) surface oxidation in electrolyte solutions. Journal of Physical Chemistry B, 2001. 105(14): p. 2809-2814.
35. Watanabe, M. and Motoo, S., ELECTROCATALYSIS BY AD-ATOMS .1. ENHANCEMENT OF OXIDATION OF METHANOL ON PLATINUM AND PALLADIUM BY GOLD AD-ATOMS. Journal of Electroanalytical Chemistry, 1975. 60(3): p. 259-266.
36. Watanabe, M. and Motoo, S., ELECTROCATALYSIS BY AD-ATOMS .2. ENHANCEMENT OF OXIDATION OF METHANOL ON PLATINUM BY RUTHENIUM AD-ATOMS. Journal of Electroanalytical Chemistry, 1975. 60(3): p. 267-273.
37. Watanabe, M. and Motoo, S., ELECTROCATALYSIS BY AD-ATOMS .3. ENHANCEMENT OF OXIDATION OF CARBON-MONOXIDE ON PLATINUM BY RUTHENIUM AD-ATOMS. Journal of Electroanalytical Chemistry, 1975. 60(3): p. 275-283.
38. Tremiliosi, G., et al., Reactivity and activation parameters in methanol oxidation on platinum single crystal electrodes 'decorated' by ruthenium adlayers. Journal of Electroanalytical Chemistry, 1999. 467(1-2): p. 143-156.
39. Davies, J.C., Hayden, B.E., and Pegg, D.J., The modification of Pt(110) by ruthenium: CO adsorption and electro-oxidation. Surface Science, 2000. 467(1-3): p. 118-130.
85
40. Lima, F.H.B. and Gonzalez, E.R., Ethanol electro-oxidation on carbon-supported Pt–Ru, Pt–Rh and Pt–Ru–Rh nanoparticles. Electrochimica Acta, 2008. 53(6): p. 2963-2971.
41. Rightmire, R.A., et al., Ethyl Alcohol Oxidation at Platinum Electrodes. Journal of the Electrochemical Society, 1964. 111(2): p. 242-247.
42. Pastor, E. and Iwasita, T., D/H Exchange of Ethanol at Platinum-Electrodes. Electrochimica Acta, 1994. 39(4): p. 547-551.
43. Lamy, C., et al., Recent advances in the development of direct alcohol fuel cells (DAFC). J. Power Sources, 2002. 105: p. 283-296.
44. Souza, J.P.I., et al., Performance of a co-electrodeposited Pt-Ru electrode for the electro-oxidation of ethanol studied by in situ FTIR spectroscopy. Journal of Electroanalytical Chemistry, 1997. 420(1-2): p. 17-20.
45. Delime, F., Leger, J.M., and Lamy, C., Enhancement of the electrooxidation of ethanol on a Pt-PEM electrode modified by tin. Part I: Half cell study. Journal of Applied Electrochemistry, 1999. 29(11): p. 1249-1254.
46. Lamy, C., Belgsir, E.M., and Leger, J.M., Electrocatalytic oxidation of aliphatic alcohols: Application to the direct alcohol fuel cell (DAFC). Journal of Applied Electrochemistry, 2001. 31(7): p. 799-809.
47. Bonarowska, M., Malinowski, A., and Karpinski, Z., Hydrogenolysis of C-C and C-Cl bonds by Pd-Re/Al2O3 catalysts. Applied Catalysis a-General, 1999. 188(1-2): p. 145-154.
48. Aboulgheit, A.K., Menoufy, M.F., and Elmorsi, A.K., Hydroconversion of N-Heptane on Catalysis Containing Platinum, Rhenium and Platinum Rhenium on Sodium Mordenite. Applied Catalysis, 1990. 61(2): p. 283-292.
49. McCabe, R.W., DiMaggio, C.L., and Madix, R.J., Adsorption and reactions of acetaldehyde on platinum(S)-[6(111) .times. (100)]. The Journal of Physical Chemistry, 1985. 89(5): p. 854-861.
50. Zhao, H.B., Kim, J., and Koel, B.E., Adsorption and reaction of acetaldehyde on Pt(111) and Sn/Pt(111) surface alloys. Surface Science, 2003. 538(3): p. 147-159.
51. Rodriguez, J.L., et al., Reaction intermediates of acetaldehyde oxidation on Pt(111) and Pt(100). An in situ FTIR study. Langmuir, 2000. 16(12): p. 5479-5486.
52. Perez, J.M., et al., Adsorption of Acetaldehyde on Pt(100) and (111) Faces - A Semiempirical Quantum Mechanical Study. Surface Science, 1990. 235(2-3): p. 307-316.
53. Shekhar, R., et al., Adsorption and reaction of aldehydes on Pd surfaces. Journal of Physical Chemistry B, 1997. 101(40): p. 7939-7951.
54. Rodes, A., Pastor, E., and Iwasita, T., An FTIR study on the adsorption of acetate at the basal planes of platinum single-crystal electrodes. Journal of Electroanalytical Chemistry, 1994. 376(1–2): p. 109-118.
55. Shao, M.H. and Adzic, R.R., Electrooxidation of ethanol on a Pt electrode in acid solutions: in situ ATR-SEIRAS study. Electrochimica Acta, 2005. 50(12): p. 2415-2422.
56. Corrigan, D.S., et al., Adsorption of acetic acid at platinum and gold electrodes: a combined infrared spectroscopic and radiotracer study. The Journal of Physical Chemistry, 1988. 92(6): p. 1596-1601.
57. Vigier, F., et al., Electrocatalysis for the direct alcohol fuel cell. Topics in Catalysis, 2006. 40(1-4): p. 111-121.
86
58. Karelin, A.I., Grigorovich, Z.I., and Rosolovskii, V.Y., Vibrational spectra of perchloric acid-I. Gaseous and liquid HClO4 and DClO4. Spectrochimica Acta Part A: Molecular Spectroscopy, 1975. 31(5–6): p. 765-775.
59. Scott, F.J., Mukerjee, S., and Ramaker, D.E., CO coverage/oxidation correlated with PtRu electrocatalyst particle morphology in 0.3 M methanol by in situ XAS. Journal of the Electrochemical Society, 2007. 154(5): p. A396-A406.
87
Chapter 6: Operando Raman Spectroscopy of a non-Pt Cathode
Nafion Membrane Electrode
6.1 Introduction:
The increasing cost of Pt has motivated the development of non-platinum group metal
(non-PGM) compounds to catalyze the reduction of molecular oxygen to water in proton
exchange membrane fuel cells (PEMFC). The ability of Fe and Co N4 macrocycles to catalyze
the oxygen reduction reaction (ORR) has been known since the 1960s.(1, 2) The activity of
these catalysts was subsequently improved through pyrolysis in an inert atmosphere. (3-10)
Despite their increased activity, the cost of the macrocycle precursors made their use impractical.
More recently, it has been demonstrated that compounds capable of catalyzing the ORR can be
made through pyrolysis of cost effective transition metal, nitrogen and carbon precursors.(11-14)
Pyrolysis of these materials results in a catalyst M-Nx active site embedded in a graphite plane
(M-Nx/C).(15-20)
There is no generally consensus regarding the nature of the active complex of M-Nx/C
catalysts. The proposed structure of this active site could exist in two possible configurations M-
N2/C and M-N4/C.(21, 22) In the case of the former, the iron would be bound to two pyridinic
nitrogens at the edge of a graphene crystallite(21, 23) while he latter describes the active site as
an iron atom bound to four nitrogens at the center of porphyrin moiety.(22) Given that previous
studies show iron coordinated by four nitrogens(24-30) in a phenanthrolinic configuration,22,24
Dodelet further proposed a Fe-N2+2/C active site where the iron is bound to four nitrogens from
two separate graphene crystallites.(31) Additionally, a second active site has been proposed
consisting of metal nanoparticles and/or oxides (M/MOx) encapsulated in a carbon
88
nanostructure.(32, 33) These M/MOx structures have been shown stable even subjected to
prolonged periods of time in an acidic environment.(32)
The proposed mechanism in acidic media for the ORR on M-Nx/C catalysts is a two-step
process where oxygen is first reduced to hydrogen peroxide followed a reduction to water.
Ramaswamy et al demonstrated using a Fe-N4 porphryin catalyst that in, acidic media, the Fe-N4
active site is incapable of reducing hydrogen peroxide to water.(34) This suggests the existence
of second active site which catalyzes the second step of the ORR. While Jaouen et al(19) noted
the catalytic activity from the presence of M/MOx complexes, it is not until recently that their
role as a second active site for the reduction of peroxides has been considered.(35, 36)
In an effort to gain insight into the reduction mechanism active sites of these new
catalysts, a fuel cell was developed for the purpose of obtaining Raman spectra of a Fe-Nx/C
catalyst under normal operating conditions. The goal was to identify and characterize the
potential-dependent peaks arising during catalysis. The confocal Raman microscope allowed for
depth profiling of the membrane electrode assembly (MEA) allowing for comprehensive
characterization of the MEA. Density functional theory was used to calculate the theoretical
vibrational spectrum of the hypothetical Fe-N2+2 and Fe-N4 active site along with a cluster of Fe
atoms to represent a nanopartricle. For each of these models, spectra were generated both in the
presence and absence of adsorbed O2. A comparison between these two theoretical spectra
elucidates which peaks corresponds to O2 adsorption.
6.2 Experimental:
Membrane electrode assembly preparation: Nafion 117 (Wilmington, DE) was
pretreated by immersing sheets in boiling 8 M nitric acid followed by two hours in boiling
Nanopure™ water. Non-PGM Catalysts were prepared using the method described by
89
Barton.(37) Briefly, Ketjen black 600JD, iron acetate (0.75 wt % Fe) and melamine (6.3 wt %
N) was heat treated at 800 °C. Catalyst inks were prepared by diluting the catalyst in 5 wt %
Nafion ionomer solution (Sigma Aldrich, Milwaukee, WI), water and isopropyl alcohol. 4
mg/cm2 Pt black (Johnson Matthey) and 1.6 mg non-PGM catalyst was used as the anode and
cathode catalysts, respectively. Catalyst inks were applied to a 5 cm2 area of a sheet of Nafion
117 immobilized on a heated vacuum table (NuVant Systems, Inc., Crown Point, IN). MEA’s
were initially conditioned by cycling the potential from 200-800 mV in the operando Raman cell
operating at 50 °C with humidified H2 (50 sccm) and O2 (200 sccm) flowing over the anode and
cathode, respectively.
Cell design: The operando Raman fuel cell design is based on the operando IR-XAS cell
described by Lewis et al.(38) Briefly, the upper flow field connects to the working electrode and
contains an aperture that accommodates a G.E. 124 fused quartz window. The lower flow field
connects to the counter/reference electrode. An aluminum housing slider (Fig 1.) delivers the
reactant feeds to the graphite flow fields and contains the heating and temperature control
elements. A DB9 connector seated in the housing slider connects the cell to the potentiostat.
The slider is dimensioned to position the catalytic surface at the working distance of the
microscope objective (Fig 2).
90
Figure 6.1. Exploded view of operando Raman cell components. (1) Top plate, (2) upper flow
field, (3) membrane electrode assembly, (4) lower flow field, (5) Assembly gasket, and (6)
assembly stage.
Figure 6.2. Operando Raman cell beneath a confocal Raman microscope.
91
Operando Raman spectroscopy: All operando Raman spectra were acquired using a
WITec Inc. (Ulm, Germany) Confocal Raman Microscope (CRM 200). A 488 nm (23 mW)
solid state laser (WITec Inc.) was used as the excitation source, which was coupled into a Zeiss
(Thornwood, NY) microscope via a 50 m wavelength-specific single-mode optical fiber. The
incident laser beam was focused onto the sample using a Nikon (Tokyo, Japan) Fluor (10x/0.25,
WD: 7.00 mm) objective. The Raman backscattered radiation was focused through a
holographic notch filter, onto a 50 m multimode optical fiber, and into a 300 mm focal length
monochromator (600/mm grating, blazed at 500 nm). The Raman spectrum was detected via a
back-illuminated, deep-depletion CCD camera (1024 x 128 pixels) operating at -82C. Single
Raman spectra were aquired for 15 s.
Prior to obtaining spectra, the cell potential was cycled from 0 to 1200 mV at 50 °C with
humidified H2 (50 sccm) and N2 (200 sccm) fed to the counter/reference and working electrodes,
respectively. The working electrode reactant feed was then switched to humidified O2 (200
sccm). Raman spectra were obtained between 1100 mV and 0 mV and collected at decreasing
100 mV increments.
Density Functional Theory: The hypothetical active sites of a Fe-Nx/C catalyst were
used as the input structure to be modeled in the presence and absence of O2. Calculations were
performed using Jaguar 7.2 (Schrodinger Inc., Portland, OR) at the X3LYP/LACV3P++** (“**”
and “++” denote polarization (39) and diffuse (40) basis set functions, respectively) level of
theory. Output files were converted to vibrational mode animations using Maestro (Schrodinger
Inc.). Calculations were carried out on a 55 node (dual core Xeon processors with 4GB RAM)
High Performance Computing Cluster at the University of Texas, Pan American. By varying the
spin state of catalyst and calculating the total energy values of each, DFT shows that the ground
92
state of [catalyst] as having a spin multiplicity of 3. The calculated values are depicted in Figure
3. All three cases have the lowest energy spin state as a triplet. For the lowest energy state,
normal mode calculations were performed
Figure 6.3: Total free energy, in Hartrees, for the hypothetical Fe-Nx/C active site at various spin
states.
93
Structural distortion treatment procedure of Fe-Nx/C: The Melamine-Fe catalyst was
incorporated on MEA specifically designed for in-situ operando cell. The ready MEA electrode
was assembled into a flow-through half-cell which was equipped with carbon cloth as counter
electrode and RHE reference. While the treatment, acid-based electrolyte saturated with oxygen,
0.5 M H2SO4 was circulating through the cell. The half-cell was connected to an Autolab
(Ecochemie Inc. model-PGSTAT 30) and subjected to CV-cycling at potentials range of 0.05-1.1
V (vs. RHE). The structural alteration of Melamine-Fe-C catalyst was achieved by addition of a
controlled amount of 35% hydrogen peroxide to the electrolyte (pH=1) reservoir. After the
treatment, the half-cell was reassembled and the electrode was rinsed with Milipore DI water and
dried prior further studies.
6.3 Results
There are several advantages to studying the ionomer-metal interface of a fuel cell using
Raman spectroscopy. A hydrated membrane is essential for the normal operation of a proton
exchange membrane fuel cell. In previous FTIR operando studies the reactant stream
humidification, flow rate and cell temperature needed to be rigorously controlled or potentially
useful regions of the IR spectrum would be obliterated from peaks resulting from an excess or
deficit of water. Peaks in Raman spectra related to water do not appear in significant intensity in
the region of interest for this study making the acquisition and analysis of spectra comparatively
simple. The confocal microscope allows for depth profiling of the membrane electrode assembly
providing a comprehensive representation of the interface. Figure 4 outlines the utility of this
technique. In order to obtain a reference point, the microscope was positioned along the z-axis
such that the spectrum of bulk Nafion was the strongest and will be referred to as +0 µm (all
subsequent positions on the z-axis will refer to points above +0 µm). The signal for bulk Nafion
94
remains prevalent at positions from up to +100 µm. At this point peaks related to the carbon
support begin to appear. The two prominent features of the graphite are the E2g vibration mode
1360 cm-1
disorder peak (D peak) at 1600 cm-1
.(16) The D peak arises from a A1g breathing
mode is observed at the edges of graphene planes on clusters smaller than 200 Å.(16) Residual
peaks from Nafion are due the Nafion dispersion used to prepare catalyst inks. Peaks from
molecular oxygen also evolve beginning at +100 µm and continue to intensify as the distance
from the reference point increases.
Figure 6.4. (Left) Schematic outlining the concept of depth profiling a membrane electrode
assembly using a confocal Raman microscope. (Right) Depth dependent spectra of a membrane
electrode assembly consisting of an Fe/N/C catalyst and Nafion. The focal position where the
spectrum of bulk Nafion is most intense serves as the reference point.
Potential dependent changes to the surface of a Fe-Nx/C cathode during ORR were
assessed using operando Raman spectroscopy. Figure 5 is a series of spectra acquired at
95
decreasing potentials in the presence of oxygen. Spectra were obtained at 350 µm above the
reference point. Potential dependent peaks at 631 cm-1
, 600 cm-1
, and 564 cm-1
appear at 700
mV and are present in subsequent spectra taken at lower potentials. The polarization curve
shown in figure 6 indicates that current is produced at potentials of 700 mV and lower. It is
therefore reasonable to correlate these peaks to catalytic activity. This experiment was
duplicated on the same catalyst in the presence of nitrogen and using a catalyst prepared without
the iron precursor. In the resulting spectra shown, in figure 7, the 631 cm-1
, 600 cm-1
, and 564
cm-1
peaks to do not appear in either of these control experiments.
96
Figure 6.5. The potential dependent Raman spectra of an iron based non-PGM cathode catalyst
obtained under O2.
1300 1200 1100 1000 900 800 700 600 500 400
1300120011001000900800700600500400
564 cm-1
600 cm-1
0mV
100mV
200mV
300mV
400mV
500mV
600mV
700mV
800mV
900mV
1000mV
1100mV
629 cm-1
97
Figure 6.6. Raman fuel cell polarization curve obtained under oxygen with a MEA consisting of
a Pt anode and an Fe-Nx/C cathode.
98
Figure 6.7. Potential dependent control experiments of a non-PGM cathode catalyst. (Left):
Spectra obtained under O2 with a catalyst prepared without iron. (Right): Spectra obtained under
N2 catalyst prepared with iron.
Theoretical spectra generated using DFT are useful for analyzing experimentally acquired
vibrational spectra. These spectra provide the approximate location of each normal mode along
with an animation which describes the vibration. Scaling factors inherent in DFT calculations
generally preclude an exact agreement between theoretically and experimentally obtained data:
peak assignments must therefore be supported with additional experimental data.(41)
Generating a theoretical spectrum of a hypothetical active site a Fe-Nx/C catalyst will provide a
basis for locating regions of the Raman spectra where peaks corresponding to oxygen adsorption
may be found. The selection of the hypothetical active site provides a challenge in this analysis.
1200 1000 800 600 400
Wavenumber (cm-1)
0mV
100mV
200mV
300mV
400mV
500mV
600mV
700mV
800mV
900mV
1100mV
1000mV
629 cm-1
600 cm-1 564 cm
-1
1200 1000 800 600 400
600 cm-1
564 cm-1
629 cm-1
1100mV
Wavenumber (cm-1)
0mV
100mV
200mV
300mV
400mV
500mV
600mV
700mV
800mV
900mV
1000mV
1100mV
99
In order to ensure that both forms of nitrogen are considered (pyridinic and pyrrolic), Fe-Nx/C
proposed active sites were modeled. Theoretical spectra both of these models were generated,
each with O2 adsorbed onto the iron and without. Figure 8 shows the series of theoretical spectra
representing the Fe-N2+2/C active site with pyridinic nitrogens and figure 9 shows the N4 pyrrolic
active site.
100
Figure 6.8. (Top) DFT generated theoretical spectra of an iron based non-PGM catalyst with a
N4 pyridinic active site. (Bottom) Molecular representations of the theoretical catalyst with and
without adsorbed oxygen.
Figure 6.9. (Top) DFT generated theoretical spectra of an iron based non-PGM catalyst with a
N4 pyrrolic active site. (Bottom) Molecular representations of the theoretical catalyst with and
without adsorbed oxygen.
101
The theoretical spectra of the hypothetical Fe-N/C catalyst active sites in the presence
and absence of adsorbed oxygen were used to elucidate which normal modes corresponded to
oxygen adsorption. Specifically, differences between the two spectra could be used to identify
normal modes corresponding with Fe-O and O-O stretching. Observing peaks associated with
O-O stretching is unlikely given that they are found in regions of the spectra dominated by E2g
and D graphene peaks. The spectra of figure 8, corresponding to a Fe-N2+2/C pyridinic active
site, show only a handful of normal modes present in the oxygenated spectra that are missing the
deoxygenated spectra. Normal mode animations show that Fe-N vibrations are responsible for
the 843 cm-1
and 930 cm-1
modes while the 1216 cm-1
peak is from O-O stretching. A complete
inspection of all of the DFT normal modes of the oxygenated sample show Fe-O stretching at
372 cm-1
. Similarly the differences between the Fe-N4/C pyrrolic active site spectra (mainly the
regions near 608 cm-1
, 725 cm-1
, 788 cm-1
, 866 cm-1
, 1009 cm-1
, and 1105 cm-1
) also correspond
with Fe-N vibrations. The peaks associated with O-O are found at 1592 cm-1
and Fe-O stretching
are found at 148 cm-1
, 192 cm-1
, and 223 cm-1
. In both models Fe-O peaks are at frequencies
which are too low to be associated with the experimentally obtained potential dependent peaks.
An explanation for position of the experimentally obtained potential dependent peaks is
that they result oxygen adsorbed onto Fe/FeOx nanostructures. This can be observed in figure
10, as the Fe-O stretch from the theoretical spectrum of an iron nanoparticle cluster with
adsorbed O2 appears at 537 cm-1
. As previously discussed, additional experimental data is
required for associating real peaks with theoretical normal modes. To that end, a Fe-Nx/C
electrode was subjected to the structural distortion treatment described in the experimental
section. This treatment, as described by Tylus et al., contributes to the dissolution of Fe/FeOx
nanostructures.(42) The potential dependent Raman spectra of the Fe-Nx/C catalyst after the
102
peroxide/CV treatment is shown in figure 11. The fact that none of the potential dependent
features of figure 5 are present in figure 11, suggests that they result from oxygen species
adsorbed onto Fe nanoparticles and not the Fe-Nx moiety. A polarization curve of the
peroxide/CV treated catalyst (Fig. 12) shows that despite a higher overpotential, there is only a
small decrease in current density at lower potentials when compared to the intact catalyst. A
study by Olson et al proposes a bifunctional mechanism where a M-Nx moiety reduces O2 to
H2O2 and the reduction of H2O2 to water is catalyzed by M/MOx clusters.(35) The idea that
H2O2 reduced by M/MOx clusters is supported by the observations of Guillet et al who reported a
92% selectivity in the production of H2O2 on a Co-based porphyrin catalysts prepared under
conditions that limited formation Co/CoOx nanostructures.(36)
103
Figure 6.10. (Top) DFT generated theoretical spectra of an iron nanoparticle with and without
adsorbed oxygen. (Bottom) Molecular representations of iron nanoparticle with and without
adsorbed oxygen.
104
Figure 6.11. Potential dependent Raman spectra of non-PGM catalyst treated with H2O2 to
remove nanoparticles. Spectra were obtained under presence of oxygen.
1200 1000 800 600 400
0mV
100mV
200mV
300mV
400mV
500mV
600mV
700mV
800mV
900mV
1000mV
1100mV
Wavenumber (cm-1)
105
Figure 6.12. Fuel cell polarization curve obtained under oxygen with a MEA consisting of a Pt
anode and an Fe-Nx/C cathode subjected to the structural distortion treatment.
6.4 Conclusions:
Operando Raman spectroscopy to observe potential-dependent peaks on a Fe-Nx/C
catalyst in the presence of oxygen. It is reasonable to associate these peaks with catalysis as their
appearance coincides with current production. A proposed mechanism for the ORR on these
catalysts suggests two active sites: Fe-Nx moieties where O2 is reduced to H2O2 and Fe/FeOx
nanostructures that catalyze the reduction of H2O2 to water. Positions of Fe-O stretching modes
observed in theoretical spectra of oxygen adsorbed onto hypothetical models of both active sites
suggests the experimentally obtained peaks are associated with the latter. Additional Raman
studies on an Fe-Nx/C catalyst with these Fe/FeOx nanostructures removed, do not show the
potential dependent peaks observed on the intact catalyst.
References:
5 0 -5 -10 -15
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Po
ten
tia
l (m
V)
Current Density (mA/cm2)
106
1. Jasinski, R.J., New fuel cell cathode catalyst. Nature (London, U. K.), 1964. 201(Copyright
(C) 2012 American Chemical Society (ACS). All Rights Reserved.): p. 1212-13. 2. Jahnke, H., Schonborn, M., and Zimmermann, G., Organic dyestuffs as catalysts for fuel
cells. Top Curr Chem, 1976. 61(Copyright (C) 2012 U.S. National Library of Medicine.): p. 133-81.
3. Wiesener, K., et al., N4 macrocycles as electrocatalysts for the cathodic reduction of oxygen. Materials Chemistry and Physics, 1989. 22(3–4): p. 457-475.
4. Bagotzky, V.S., et al., Electrocatalysis of the oxygen reduction process on metal chelates in acid electrolyte. Journal of Power Sources, 1978. 2(3): p. 233-240.
5. Tarasevich, M.R. and Radyuskina, K.A., Pyropolymers of N4-complexes: Structure and elegtrocatalytic properties. Materials Chemistry and Physics, 1989. 22(3–4): p. 477-502.
6. Wiesener, K., N4-chelates as electrocatalyst for cathodic oxygen reduction. Electrochimica Acta, 1986. 31(8): p. 1073-1078.
7. Van Der Putten, A., et al., Oxygen reduction on pyrolysed carbon-supported transition metal chelates. Journal of Electroanalytical Chemistry and Interfacial Electrochemistry, 1986. 205(1–2): p. 233-244.
8. Scherson, D., et al., Transition metal macrocycles supported on high area carbon: Pyrolysis—mass spectrometry studies. Electrochimica Acta, 1986. 31(10): p. 1247-1258.
9. Scherson, D.A., et al., Cobalt tetramethoxyphenyl porphyrin—emission Mossbauer spectroscopy and O2 reduction electrochemical studies. Electrochimica Acta, 1983. 28(9): p. 1205-1209.
10. Franke, R., Ohms, D., and Wiesener, K., Investigationof the Influence of Thermal-Treatment on the Properties of Carbon Materials Modified by N-4-Chelates for the Reduction of Oxygen in Acidic Media. Journal of Electroanalytical Chemistry, 1989. 260(1): p. 63-73.
11. Gupta, S., et al., Heat-treated polyacrylonitrile-based catalysts for oxygen electroreduction. Journal of Applied Electrochemistry, 1989. 19(1): p. 19-27.
12. Ohms, D., et al., Influence of metal-ions on the electrocatalytic oxygen reduction of carbon materials prepared from pyrolyzed polyacrylonitrile. Journal of Power Sources, 1992. 38(3): p. 327-334.
13. Alves, M.C.M., et al., Origin of the electrocatalytic properties for oxygen reduction of some heat-treated polyacrylonitrile and phthalocyanine cobalt compounds adsorbed on carbon-black as probed by electrochemistry and X-ray absorption-spectroscopy. Journal of Physical Chemistry, 1992. 96(26): p. 10898-10905.
14. Dodelet, J.-P., Oxygen Reduction in PEM Fuel Cell Conditions: Heat-Treated Non-Precious Metal-N4 Macrocycles and Beyond, in N4-Macrocyclic Metal Complexes, J. Zagal, F. Bedioui, and J.-P. Dodelet, Editors. 2006, Springer New York. p. 83-147.
15. Lefevre, M., et al., Iron-Based Catalysts with Improved Oxygen Reduction Activity in Polymer Electrolyte Fuel Cells. Science, 2009. 324(5923): p. 71-74.
16. Jaouen, F., Charreteur, F., and Dodelet, J.P., Fe-based catalysts for oxygen reduction in PEMFCs - Importance of the disordered phase of the carbon support. Journal of the Electrochemical Society, 2006. 153(4): p. A689-A698.
107
17. Jaouen, F., et al., Heat-treated Fe/N/C catalysts for O-2 electroreduction: Are active sites hosted in micropores? Journal of Physical Chemistry B, 2006. 110(11): p. 5553-5558.
18. Lefevre, M., Dodelet, J.P., and Bertrand, P., Molecular oxygen reduction in PEM fuel cells: Evidence for the simultaneous presence of two active sites in Fe-based catalysts. Journal of Physical Chemistry B, 2002. 106(34): p. 8705-8713.
19. Jaouen, F., et al., Oxygen reduction catalysts for polymer electrolyte fuel cells from the pyrolysis of iron acetate adsorbed on various carbon supports. Journal of Physical Chemistry B, 2003. 107(6): p. 1376-1386.
20. Artyushkova, K., et al., XPS structural studies of nano-composite non-platinum electrocatalysts for polymer electrolyte fuel cells. Topics in Catalysis, 2007. 46(3-4): p. 263-275.
21. Lefevre, M., Dodelet, J.P., and Bertrand, P., O-2 reduction in PEM fuel cells: Activity and active site structural information for catalysts obtained by the pyrolysis at high temperature of Fe precursors. Journal of Physical Chemistry B, 2000. 104(47): p. 11238-11247.
22. Bouwkamp-Wijnoltz, A.L., et al., On active-site heterogeneity in pyrolyzed carbon-supported iron porphyrin catalysts for the electrochemical reduction of oxygen: An in situ Mossbauer study. Journal of Physical Chemistry B, 2002. 106(50): p. 12993-13001.
23. Faubert, G., et al., Oxygen reduction catalysts for polymer electrolyte fuel cells from the pyrolysis of FeII acetate adsorbed on 3,4,9,10-perylenetetracarboxylic dianhydride. Electrochimica Acta, 1999. 44(15): p. 2589-2603.
24. Yuasa, M., et al., Modifying carbon particles with polypyrrole for adsorption of cobalt ions as electrocatatytic site for oxygen reduction. Chemistry of Materials, 2005. 17(17): p. 4278-4281.
25. Maruyama, J. and Abe, I., Fuel cell cathode catalyst with heme-like structure formed from nitrogen of glycine and iron. Journal of the Electrochemical Society, 2007. 154(3): p. B297-B304.
26. Bouwkamp-Wijnoltz, A.L., et al., Electrochemical reduction of oxygen: an alternative method to prepare active CoN4 catalysts. Electrochimica Acta, 1999. 45(3): p. 379-386.
27. Bron, M., et al., EXAFS, XPS and electrochemical studies on oxygen reduction catalysts obtained by heat treatment of iron phenanthroline complexes supported on high surface area carbon black. Journal of Electroanalytical Chemistry, 2002. 535(1-2): p. 113-119.
28. Bron, M., et al., Thermogravimetry/Mass Spectrometry Investigations on the Formation of Oxygen Reduction Catalysts for PEM Fuel Cells on the Basis of Heat-Treated Iron Phenanthroline Complexes. Fuel Cells, 2002. 2(3-4): p. 137-142.
29. Schulenburg, H., et al., Catalysts for the oxygen reduction from heat-treated iron(III) tetramethoxyphenylporphyrin chloride: Structure and stability of active sites. Journal of Physical Chemistry B, 2003. 107(34): p. 9034-9041.
30. Koslowski, U.I., et al., Nature of the catalytic centers of porphyrin-based electrocatalysts for the ORR: A correlation of kinetic current density with the site density of Fe-N-4 Centers. Journal of Physical Chemistry C, 2008. 112(39): p. 15356-15366.
31. Charreteur, F., et al., Fe/N/C non-precious catalysts for PEM fuel cells: Influence of the structural parameters of pristine commercial carbon blacks on their activity for oxygen reduction. Electrochimica Acta, 2008. 53(6): p. 2925-2938.
108
32. Wu, G., et al., High-Performance Electrocatalysts for Oxygen Reduction Derived from Polyaniline, Iron, and Cobalt. Science, 2011. 332(6028): p. 443-447.
33. Faubert, G., et al., Heat-treated iron and cobalt tetraphenylporphyrins adsorbed on carbon black: Physical characterization and catalytic properties of these materials for the reduction of oxygen in polymer electrolyte fuel cells. Electrochimica Acta, 1996. 41(10): p. 1689-1701.
34. Ramaswamy, N. and Mukerjee, S., Fundamental Mechanistic Understanding of Electrocatalysis of Oxygen Reduction on Pt and Non-Pt Surfaces: Acid versus Alkaline Media. Advances in Physical Chemistry, 2012. 2012: p. 17.
35. Olson, T.S., et al., Bifunctional Oxygen Reduction Reaction Mechanism on Non-Platinum Catalysts Derived from Pyrolyzed Porphyrins. Journal of the Electrochemical Society, 2010. 157(1): p. B54-B63.
36. Guillet, N., et al., Electrogeneration of hydrogen peroxide in acid medium using pyrolyzed cobalt-based catalysts: Influence of the cobalt content on the electrode performance. Journal of Applied Electrochemistry, 2006. 36(8): p. 863-870.
37. Nallathambi, V., et al., Nitrogen Precursor Effects in Iron-Nitrogen-Carbon Oxygen Reduction Catalysts. Electrochemical and Solid State Letters, 2011. 14(6): p. B55-B58.
38. Lewis, E.A., et al., Operando X-ray absorption and infrared fuel cell spectroscopy. Electrochimica Acta, 2011. 56(24): p. 8827-8832.
39. Frisch, M.J., Pople, J.A., and Binkley, J.S., Self-Consistent Molecular-Orbital Methods 25. Supplementary Functions For Gaussian-Basis Sets. Journal of Chemical Physics, 1984. 80(7): p. 3265-3269.
40. Clark, T., et al., Efficient Diffuse Function-Augmented Basis-Sets For Anion Calculations 3. The 3-21+G Basis Set For 1ST-Row Elements, LI-F. Journal of Computational Chemistry, 1983. 4(3): p. 294-301.
41. Dimakis, N., et al., A band dispersion mechanism for Pt alloy compositional tuning of linear bound CO stretching frequencies. Journal of Physical Chemistry B, 2005. 109(5): p. 1839-1848.
42. Tylus, U., et al., Elucidating Active Site in Oxygen Reduction at M-Nx and M/MxOy-based Non-Noble Metal Electrocatalysts., in Manuscript in Preparation.