Noble metal-free hydrazine fuel cell catalysts: EPOC effect in competing chemical and electrochemical reaction pathways

Article abstract of DOI:10.1021/ja111160r

We report the discovery of a highly active Ni-Co alloy electrocatalyst for the oxidation of hydrazine (N₂H₄) and provide evidence for competing electrochemical (faradaic) and chemical (nonfaradaic) reaction pathways. The electrochemical conversion of hydrazine on catalytic surfaces in fuel cells is of great scientific and technological interest, because it offers multiple redox states, complex reaction pathways, and significantly more favorable energy and power densities compared to hydrogen fuel. Structure-reactivity relations of a Ni₆₀Co₄₀ alloy electrocatalyst are presented with a 6-fold increase in catalytic N ₂H₄ oxidation activity over todays benchmark catalysts. We further study the mechanistic pathways of the catalytic N₂H ₄ conversion as function of the applied electrode potential using differentially pumped electrochemical mass spectrometry (DEMS). At positive overpotentials, N₂H₄ is electrooxidized into nitrogen consuming hydroxide ions, which is the fuel cell-relevant faradaic reaction pathway. In parallel, N₂H₄ decomposes chemically into molecular nitrogen and hydrogen over a broad range of electrode potentials. The electroless chemical decomposition rate was controlled by the electrode potential, suggesting a rare example of a liquid-phase electrochemical promotion effect of a chemical catalytic reaction ( EPOC ). The coexisting electrocatalytic (faradaic) and heterogeneous catalytic (electroless, nonfaradaic) reaction pathways have important implications for the efficiency of hydrazine fuel cells.(Figure Presented)

Full text of DOI:10.1021/ja111160r

ARTICLE
pubs.acs.org/JACS
Noble Metal-Free Hydrazine Fuel Cell Catalysts: EPOC Effect in
Competing Chemical and Electrochemical Reaction Pathways
Jean Sanabria-Chinchilla,† Koichiro Asazawa,‡ Tomokazu Sakamoto,‡ Koji Yamada,‡ Hirohisa Tanaka,‡ and
Peter Strasser*,§
†Department of Chemical and Biomolecular Engineering, University of Houston, Houston Texas 77204, United States
‡Frontier Technology R&D Division, Daihatsu Motor Co., Ltd., 3000 Yamanoue, Ryuo, Gamo, Shiga 520-2593, Japan
§The Electrochemical Energy, Catalysis, and Materials Science Laboratory, Department of Chemistry, Chemical Engineering Division,
Technical University Berlin, 10623 Berlin, Germany
S Supporting Information
b
ABSTRACT: We report the discovery of a highly active Ni-Co
alloy electrocatalyst for the oxidation of hydrazine (N2H4) and
provide evidence for competing electrochemical (faradaic) and
chemical (nonfaradaic) reaction pathways. The electrochemical
conversion of hydrazine on catalytic surfaces in fuel cells is of
great scientific and technological interest, because it offers
multiple redox states, complex reaction pathways, and signifi-
cantly more favorable energy and power densities compared to
hydrogen fuel. Structure-reactivity relations of a Ni60Co40
alloy electrocatalyst are presented with a 6-fold increase in
catalytic N2H4 oxidation activity over today’s benchmark cata-
lysts. We further study the mechanistic pathways of the catalytic
N2H4 conversion as function of the applied electrode potential
using differentially pumped electrochemical mass spectrometry
(DEMS). At positive overpotentials, N2H4 is electrooxidized
into nitrogen consuming hydroxide ions, which is the fuel cell-
relevant faradaic reaction pathway. In parallel, N2H4 decom-
poses chemically into molecular nitrogen and hydrogen over a broad range of electrode potentials. The electroless chemical
decomposition rate was controlled by the electrode potential, suggesting a rare example of a liquid-phase electrochemical promotion
effect of a chemical catalytic reaction (“EPOC”). The coexisting electrocatalytic (faradaic) and heterogeneous catalytic (electroless,
nonfaradaic) reaction pathways have important implications for the efficiency of hydrazine fuel cells.
Hydrazine (N2H4) is a high-energy fuel molecule, the catalysis
of which has been studied as early as 1912.1,2 The long interest in
developing high-power density direct hydrazine-air fuel cells3 is
based on their following unique advantages: (i) the theoretical
standard equilibrium hydrazine cell potential of þ1.56 V is
superior to those of other liquid or H2-air fuel cells. The
electrode reactions read
1. INTRODUCTION
Electrochemical catalysis and heterogeneous chemical catalysis
share many elementary principles, such as surface adsorption of
reactants, activated surface reactions of intermediates via transition
states, or surface desorption forming reaction products. They
sharply differ, however, in the nature of their redox electron transfer
across the catalytic interface. Given the absence of charged species,
chemical catalysis uses pressure and temperature as the two key
control parameters to affect the Gibbs free energy of reaction and
the associated Gibbs free energy of activation, that is, the energy of
the activated complex in the transition state. Owing to the presence
of charges species and electrons, electrochemical catalysis leverages
the electrode potential as well in order to affect Gibbs free energies
of reaction and activation. From this, one would hypothesize that, if
electrochemical and chemical reactions share similar intermediates
or transition states, electrode potential could affect and possibly
promote rates of nonfaradaic chemical catalysis, as well. This is
demonstrated here for the electrooxidation of hydrazine on novel
highly active non-noble alloy electrocatalysts.
Anode : N2H4 þ 4OH- / N2 þ 4H2O þ 4e-
Eo ¼ -0:33 V=RHE
ð1Þ
ð2Þ
ð3Þ
Cathode : O2 þ 2H2O þ 4e- / 4OH-
Eo ¼ þ1:23 V=RHE
Cell Reaction : N2H4 þ O2 / N2 þ 2H2O
Eo ¼ þ1:56 V=RHE
Received: December 11, 2010
Published: March 22, 2011
r
2011 American Chemical Society
5425
dx.doi.org/10.1021/ja111160r J. Am. Chem. Soc. 2011, 133, 5425–5431
|

Journal of the American Chemical Society
ARTICLE
(ii) the products of hydrazine oxidation (nitrogen and water) are
neither harmful nor carbon-containing, (iii) the cell operates at
near-ambient temperatures (40-80 °C), (iv) the hydrazine fuel
cell offers superior energy and power densities compared to
hydrogen fuel cell (5400 W h L-1) making it ideally suited for
automotive applications. In a previous report, we have demon-
strated a detoxification technique which safely fixed hydrazine to
the carbonyl groups in the polymer and released hydrazine by
means of a solvent at the time of power generation.4
electrode potential, a rare example of an electric field-based infl-
uence on heterogeneous catalytic reactions (EPOC effect31-38) in
low-temperature liquid electrolyte electrochemistry. We hypothesize
that shared surface transition states and/or reactive intermedi-
ates of the hydrazine molecule is the origin for the EPOC effect.
2. EXPERIMENTAL METHODS
2.1. Catalyst Synthesis. Carbon-supported Ni, Co, and NiM (M =
Co, Cu, Zn, Fe, Ag, Mn, Zr) and CoZn bimetallic catalyst libraries with
stoichiometries from (90:10) to (20:80) were prepared by an impreg-
nation/freeze-drying procedure followed by thermal annealing. Prepara-
tion started with impregnation and sonication of high-surface-area
carbon (HSC) (Ketjenblack EC 300J, Akzo Nobel) with aqueous
metal-salt solutions.39 Solutions were either commercially available,
e.g., ZrO(NO3)2 in 35 wt % nitric acid or made from precursor
compounds [e.g., Ni(NO3)2 6H2O or Fe(NO3)3 9H2O] dissolved in
The hydrazine standard electrode potential is more negative
than that of the hydrogen evolution reaction (HER):
2H2O þ 4e- / 2OH- þ H2 Eo ¼ 0:00 V=RHE
ð4Þ
Therefore, depending on the kinetics of the hydrogen half cell
system on the selected catalyst material, a complex mixed
potential based on the concomitant electrooxidation of aqueous
N2H4 (reaction 1) and the HER may occur.5,6 On noble metals,
such as Pt, an electroless (nonfaradaic) purely chemical catalytic
decomposition pathway was reported:7
3
3
deionized (DI) water (>18.2 MΩ cm) Milli-Q gradient system, Milli-
pore Inc.). Ni and Mn precursors were obtained from Alfa Aesar; all
other metal precursors and solutions were obtained from Sigma-Aldrich,
Inc. A robotic liquid dispenser (Tecan Systems, United States) was
utilized to pipet the desired amount of metal solution onto the carbon
powder substrate for metal impregnation. Once these slurries were
sonicated, the impregnated catalysts were then frozen in liquid N2 and
subsequently freeze-dried under a moderate vacuum (0.055 mbar). All
catalysts were prepared in ∼10 mL quartz vials. Reduction of metal
precursors to the zerovalent state on the carbon support was thermally
driven under a reductive H2 atmosphere (6% H2, balance Ar, Matheson
N2H4 f N2 þ 2H2
ð5Þ
ð6Þ
3N2H4 f 2N2 þ 4NH3
The electrooxidation of hydrazine in alkaline media was
investigated extensively at platinum,8-12 palladium,13-15
gold,12,16 mercury,11,16-18 and silver19 electrodes. Yi et al.19
studied the alkaline hydrazine electrooxidation on Ti-supported
Ag-Ni nanoparticles and reported high catalytic activities for
a Ag86Ni14 alloy. Recently, Ye et al.14 studied Ni-Pd alloy
nanoparticles and found improved hydrazine oxidation activities.
Rosca and Koper20 presented a comparative Tafel-slope study of
hydrazine electrooxidation on Pt in alkaline and acidic solutions.
Using online mass spectrometry, they confirmed earlier reports21
that molecular nitrogen is the major reaction product. Early on, a
nonfaradaic catalytic breaking of the N-H bond on Pt was
hypothesized15,22 yet never confirmed due to the immediate
electrooxidation of surface atomic hydrogen. More recent re-
ports addressed the design of direct hydrazine polymer electro-
lyte membrane fuel cell (PEMFCs),4,7,23-28 comparing different
anode materials7 and anion-exchange membranes.29 Employing
non-noble anode monometallic electrocatalysts, a distinct sensi-
tivity of the cell performance on the alkaline concentration was
detected.27 Despite these recent studies on non-noble electro-
catalysts, a more detailed mechanistic understanding of the
relevant mechanistic reaction pathways is still missing. While
pure Co and Ni catalysts show catalytic activities comparable to
that of Pt30 and are the anode catalysts of choice in alkaline
media, catalytic activity benefits of bimetallic noble metal-free
alloy catalysts have never been investigated for the hydrazine
electrooxidation.
Here, we investigate the electrocatalytic oxidation of hydra-
zine on non-noble bimetallic electrocatalysts, benchmarking
their catalytic activity against state-of-the-art Co and Ni mono-
metallic catalysts. We report a new bimetallic catalyst system with
unprecedented catalytic hydrazine oxidation activity. Using
in situ differentially pumped electrochemical mass spectrometry
(DEMS) we report evidence for a mechanistic competition of
two parallel catalytic hydrazine conversion pathways, one purely
chemical electroless (reaction 5) and the other electrochemical
faradaic (reaction 1). We find that the rate of the electroless
chemical catalysis shows a dependence on the applied external
Tri Gas) at 250 °C for 2.5 h using a Lindberg/blue tube furnace. Final
3
thermal annealing was performed immediately after the reduction step at
800 °C for 8 h. Each bimetallic library was designed to contain eight
catalyst samples. The atomic composition (as at %) of each sample was
varied in 10 at % increments from (90:10) to (20:80). All catalyst
samples contained 30 wt % total metal on HSC support.
2.2. Parallel Electrochemical High Throughput Screening
of Electrocatalysts. A (4 ꢀ 4)16-channel electrochemical electrode
array for parallel testing of alloy electrocatalysts was utilized in this study.
This array was previously described in detail.39 The array consisted of
mirror-polished glassy carbon (GC) electrodes (see Figure S1), onto
which the different catalyst powders were applied as catalyst powder thin
film based on the procedure originally reported by Schmidt et al.40,41
.
Each of these working electrodes was i) in contact with independent
samples of the testing solution, and ii) independently controlled by a
potentiostat.
Due to the use of strong alkaline media, a Zn/ZnO quasi reference
electrode was utilized instead of the conventional Ag/AgCl reference
electrode. The reference electrode comprised a Zn wire immersed in 1.0
M KOH solution, and it was calibrated against a hydrogen reference
electrode (Hydroflex RE, Gaskatel GmbH, Germany). A Pt wire served
as counter electrode. All potentials presented and discussed here are
reported against a reference hydrogen electrode (RHE) at pH 14.
Catalyst inks were prepared in a manner nearly identical to that
previously published.42-44 In short, 10 mg of the desired catalyst was
combined with 7.5 mL of DI water, 1.5 mL of isopropanol, 0.460 mL of
THF, and 40.0 μL of a 5 wt % anionic ionomer solution (Daihatsu
Motor Co.). The ink was then sonicated during 15 min. After sonication,
39.2 μL of the ink was applied onto the GC electrode, resulting in a
loading of 15 μgtotal metal/cm2. The entire array was then left to dry for at
least 1 h.
For each of the eight catalyst samples of each library, two ink
applications were randomly applied in two different GC electrodes.
One catalyst film was immersed in 1.0 M KOH solution [50% w/v aq
sol., Alfa Aesar], whereas the other one was immersed in 0.1 M hydrazine
hydrate/1.0 M KOH solution [60% v/v hydrazine hydrate solution,
5426
dx.doi.org/10.1021/ja111160r |J. Am. Chem. Soc. 2011, 133, 5425–5431

Journal of the American Chemical Society
ARTICLE
Ootsuka Chemistry Inc., Japan]. Each of the 16 chambers in the array
was then bubbled with N2 gas (UHP, Matheson Tri Gas) to remove any
3
unwanted gases from the solutions. After 15 min, the gas was switched to
a “blanketing” flow. The catalyst films were then cycled 1000 times
between -0.25 V and þ0.10 V (vs RHE) at 20 mV/s (durability test)
and then three consecutive linear sweep voltammograms (LSV) were
run from -0.25 V to þ0.10 V (vs RHE) at 5 mV/s. From these LSV
runs, the mean mass activity value for each catalyst sample was calculated
by obtaining the maximum current at þ0.10 V. Mass activity values were
expressed as metal-based mass current (A/gtotal metal).
2.4. Rotating Disk Electrode (RDE) Measurements. RDE
activity measurements were performed in a three-electrode configura-
tion on a 10 mm glassy carbon (GC) disk electrode in 0.1 M HH/1.0 M
KOH electrolyte. Due to the strong alkaline media utilized in this study,
a custom-made temperature-controlled Teflon electrochemical cell was
built for the RDE testing. The ink and electrode thin-film preparation
procedures as well as the measurement protocols of the RDE testing
were identical to those for parallel array testing with the following
exceptions: (i) The temperature was increased from 25 to 60 °C; (ii)
speed rotation was set at 900 rpm. A BiStat potentiostat (BioLogic
Science Instrument) was utilized in the RDE testing. Speed rotation was
set through a rotating electrode spin control (Pine Research Instru-
mentation). A Pt wire was used as counter electrode and a Zn/ZnO
reference electrode (Eo = -0.432 V/RHE, -1.26 V/NHEpH=0). All
potentials are reported with respect to a reversible RHE. Electrochemi-
cal characterization of the catalyst film was performed following the
same experimental protocol used with the combinatorial array.
2.5. Catalyst Characterization by XRD. Laboratory-source
XRD (Siemens D5000 θ/2θ diffractometer) was used to characterize
the electrocatalysts structurally. The diffractrometer is equipped with a
Braun position sensitive detector (PSD) with an angular range of 8°. The
Cu KR source operates at a potential of 35 kV and current of 30 mA.
Diffraction angles (2θ) ranged from 20° to 70°, scanned with step size of
0.02° per step and holding time of 10-30 s per step. Advanced X-ray
solution (X-ray Commander, Bruker AXS) software was used to control
the diffractometer from a desktop computer. The XRD sample holder
was a custom-made 3 cm ꢀ 3 cm plexiglass with a 1 cm width ꢀ 2 cm
length ꢀ 1 mm depth well in the center that holds the powder catalyst
samples. The catalyst powder was poured into the well and carefully
flattened to form a smooth surface, flush with the surface of the
plexiglass. XRD patterns were analyzed by JADE 8.1 software (MDI):
Peak profiles of individual reflections were obtained by a nonlinear least-
squares fit of the Cu KR2 corrected data.
2.6. Differentially Pumped Electrochemical Mass Spectro-
metry (DEMS). DEMS experiments were performed using a differen-
tially pumped ThermoStar GSD 301 Pfeiffer gas analyzer (Pfeiffer
Vacuum, United States) with a liquid-to-gas interface containing a
Silastic tubing (Dow Corning, United States) for diffusion of gaseous
and volatile species into the differentlially pumped area of the spectro-
meter. During measurements, the polymeric probe interface was placed
inside the electrolyte of the electrochemical cell next to the RDE catalyst
interface (about 1-2 cm distance). Instead of N2 gas for blanketing, Ar
Figure 1. Cyclic voltammograms of Ni, Co, and a novel Ni60Co40
electrocatalysts in (a) argon-saturated 1.0 M KOH solution and (b) in
0.1 M hydrazine hydrate N2H4 H2O (HH)/1.0 M KOH solution.
3
Plotted is the metal mass-based catalytic current. Conditions: RDE,
60 °C, 900 rpm, 20 mV/s.
potential scans of Ni (þ0.23 V, weak) and Co (þ0.08 V, weak, and
þ0.25 V, strong) in Figure 1a suggest the formation of higher-
valent Co and Ni surface oxygenates on the basis of the interaction
of OH- with the metal surface according to
M þ OH- / M - OH þ e-
M þ 2OH- / MðOHÞ2 þ 2e-
ð7Þ
ð8Þ
where M represents a surface metal atom. The passivating effect of
the surface metal hydroxide layers for the hydrazine electrooxida-
tion becomes evident when comparing voltammograms a and b in
Figure 1. The electrocatalytic HH oxidation on Ni (Figure 1b) first
rises, then rapidly drops upon Ni passivation above þ0.23 V. On a
Co surface, the HH electrooxidation drops past þ0.43 V, most
likely due to the delayed passivating effect of Co oxygenates based
on their better solubility,45 which is consistent with the larger Co
passivation charge. At potentials above 0.5 V in Figure 1b, HH
electrooxidation is low on both surfaces, with Co showing some
residual activity.
gas (UHP, Matheson Tri Gas) was used when m/z = 28 was monitored.
3
3. RESULTS AND DISCUSSION
After reversing the potential scan at þ0.56 V (Figure 1b), no
hydrazine oxidation was observed during the cathodic scan
(cathodic arrow) evidencing the irreversible nature of the metal
passivation.24 Active Co and Ni interfaces can be recovered by
scanning below 0.0 V/RHE as shown by the reduction peaks in
Figure 1a. Upon repeated passivation-reactivation cycles be-
tween -0.25 V to þ0.56 V, Ni catalysts eventually exhibited
superior HH electrooxidation activities in agreement with pre-
vious reports.30
3.1. Hydrazine Electrooxidation on Ni and Co Electrodes.
Figure 1 compares the initial electrocatalytic behavior of pure Ni
(blue) and pure Co (red) nanoparticles (a) in absence and (b) in
presence of hydrazine hydrate, N2H4 H2O (HH). Also included in
3
Figure.1 is a voltammetry of a Ni-Co alloy, which will be discussed
later. The voltammetric features in Figure 1a represent the
electrode potential-dependent surface chemistries of each catalytic
interfaces. The anodic voltammetric peaks on the positive-going
5427
dx.doi.org/10.1021/ja111160r |J. Am. Chem. Soc. 2011, 133, 5425–5431

Journal of the American Chemical Society
ARTICLE
3.2. Electrocatalytic Hydrazine Oxidation on Non-Noble
Alloys. Eight different bimetallic catalyst libraries (Ni-Co, Ni-
Cu, Ni-Zn, Ni-Fe, Ni-Ag, Ni-Mn, Ni-Zr, and Co-Zn)
were prepared and tested for hydrazine electrooxidation and
benchmarked against pure Co and Ni (Table S1, Supporting
Information [SI]). Each library included eight different binary
compositions at a 10 at % interval (Table S2, SI). A stationary
electrochemical electrode array39,46-49 (Figure S1, SI) was used
to test the catalytic HH oxidation activity and stability of a total of
64 different catalysts.
Table S1 reveals that Ni-Co bimetallic alloys are more active
than Ni and Co and superior to all other bimetallic systems. In
particular, Figure 2a reveals that Co exhibited a low hydrazine
oxidation onset potential, while Ni is more active than Co at
þ0.10 V after potential cycling in agreement with earlier
reports.50 A novel bimetallic Ni60Co40 composition showed
unprecedented HH electrooxidation activity. RDE measure-
ments of scaled-up catalysts (Figure 2b) confirmed the screening
trends for Ni, Co, and the novel Ni60Co40 composition. Our
RDE data suggest a nearly 100% increase in HH electrooxidation
activity on the bimetallic catalyst compared to that of pure Co.
To obtain insight in the atomic structure of the highly active
hydrazine oxidation catalyst, we performed bulk structural X-ray
scattering analysis (Figure 3). All Ni-Co bimetallics were face-
centered cubic (space group Fm3m) substitutional solid solu-
h
tions (disordered alloys). Table S2 (SI) summarizes ICP-OES-
derived compositions of the most active catalysts, showing
that they are in excellent agreement with their nominal library
composition.
The voltammetric features of the Ni60Co40 catalyst (green) in
Figure 1a suggest reduced surface passivation on the anodic scan.
The HH oxidation activity (Figure 1b) is about 6x higher than
that of Ni and about 2ꢀ higher than that of Co. Similar to Ni, the
HH oxidation activity of the alloy peaks at þ0.23 V and drops to
near zero values past 0.5 V. Noteworthy, the HH oxidation onset
potential of -0.15 V of the novel Ni60Co40 catalyst is signifi-
cantly shifted lower electrode potentials compared to pure Co
and Ni offering high cell efficiencies at low current densities.
3.3. Mechanistic Pathways and Electrochemical Promo-
tion of Catalysis (EPOC Effect). The Hydrogen Oxidation Reac-
tion on Ni-Co Alloys. To arrive at a conclusive mechanistic
understanding of major reaction pathways, we first needed to
investigate the hydrogen redox reactivity (eq 4) of our novel
Figure 2. (a) Parallel evaluation of hydrazine electrooxidation activities
of Ni-Co bimetallic catalysts on a stationary electrode array compared
to Co and Ni benchmarks. Shown are anodic voltammetric sweeps after
1000 cycles between -0.25 V and þ0.10 V. The background was
measured in HH-free electrolyte. 0.1 M HH/1.0 M KOH, 25 °C, 5 mV/s.
Inset: Mass activities at þ0.10 V. (b) RDE voltammetry of the novel
Ni60Co40 catalyst confirmed the screening trends from (a); all condi-
tions as in (a). Insets a and b: RDE catalytic mass activities in
A/gtotal metal at þ0.10 V.
Figure 3. Powder X-ray diffraction profiles of the most active set of Ni-Co bimetallic electrocatalysts. The Ni-Co materials exhibit solid solutions.
Ni60Co40 is the most active HH oxidation catalyst. Stick pattern show pure Co (cyan) and pure Ni(green).
5428
dx.doi.org/10.1021/ja111160r |J. Am. Chem. Soc. 2011, 133, 5425–5431

Journal of the American Chemical Society
ARTICLE
Figure 4. Hydrogen oxidation reaction (HOR) and hydrogen evolution
reaction (HER) activity of the Ni60Co40 bimetallic catalyst in H2-
saturated (red) and Ar-saturated (blue) 1 M KOH electrolyte. HER
and HOR on a Pt in H2-saturated KOH is shown for comparison
(black). Ni60Co40 shows negligible activity for HOR and a significant
HER overpotential ΔηHER compared to Pt. conditions: 25 °C, scan rate
50 mV/s.
Ni60Co40 catalyst, that is, the electrode potential-dependent rate
of the hydrogen evolution reaction (HER) and the hydrogen
oxidation reaction (HOR). Figure 4 reports the HOR and HER
behavior of the Ni60Co40 catalyst in H2-saturated electrolyte
(red) compared to a Pt catalyst (black). While Pt shows a
diffusion-limited HOR reactivity beyond þ0.05 V and highly
active HER,51 Ni60Co40 exhibited no detectable activity for the
HOR and a very large overpotential for the HER (blue in
Figure 4). We conclude that the Ni60Co40 alloy has a very low
tendency to adsorb atomic hydrogen,52,53 the key intermediate in
the HOR and HER reaction.
The absence of any HOR activity on the Ni60Co40 was
important, because it enabled us to directly detect and quantify
the nonfaradaic HH decomposition (reaction 5) by measuring
the rate of molecular hydrogen formation. On noble metal
catalysts, such as Pt or Pd, which readily oxidize molecular
hydrogen, this would not be possible.
Differentially Pumped Electrochemical Mass Spectrometry.
Insight into relevant mechanistic reaction pathways of the HH
electrooxidation on Ni60Co40 was achieved by differentially
pumped electrochemical mass spectrometry (DEMS).54,55
DEMS is a powerful analytical technique which probes in situ
volatile chemical products formed at electrified catalytic inter-
faces. Graphs a and b of Figure 5 show the temporal evolution of
the ion mass currents for H2 and N2, respectively, at the electrode
potential of zero faradaic current (open circuit potential, OCP)
(red lines) and at an electrode potential of þ0.10 V/RHE (blue
lines). At OCP, formation of molecular H2 in the presence of
hydrazine evidenced the electroless, nonfaradaic chemical HH
decomposition according to reaction 5. After an initial rapid
increase, the rate of hydrogen formation by the nonfaradaic HH
decomposition gradually decreased before it approached a near-
stationary plateau.
At an electrode potential of þ0.10 V, a faradaic anodic current
associated with the electrocatalytic oxidation of HH to nitrogen
and protons according to reaction 1 is observed. Now, a sustained
increased molecular H2 production rate suggested an enhanced
electroless (nonfaradaic) catalytic HH decomposition.38 The
applied electrochemical potential and the associated faradaic
current appear to promote the purely chemical catalytic HH
decomposition. Based on reaction eqs 1 and 5, the rate of
nitrogen formation should increase at positive electrode
Figure 5. Evolution of (a) hydrogen (m/z = 2) and (b) nitrogen (m/z =
28) ions mass currents at open circuit potentials (red), that is, at zero
faradaic current, and at þ0.10 V/RHE (blue) in the presence (solid) and
absence (dashed) of hydrazine (HH) using DEMS. t = 0 marks the
instant when the catalyst was placed in contact with the solution at the
specified potential. (c) Proposed competing chemical (red, reaction 5)
and electrochemical (blue, reaction 1) HH decomposition pathways.
The presence of the faradaic current was found to enhance the rate of the
chemical nonfaradaic pathway (EPOC effect).
potentials, because the N2 formation according to eqs 1 and 5
will add up. In fact, this behavior is observed in the temporal N2
ion current (m/z = 28) profile at OCP (solid red curve) and at
þ0.1 V/RHE (solid blue curve) in Figure 5b. We want to point
out that the values of the absolute ion currents sensitively depend
on the pressure and ionization conditions inside the instrument
and hence do not allow for a direct quantitative comparison
between species. We also note that the ion current baseline of
m/z = 28 in Figure 5b displayed a high sensitivity to hydrodynamic
perturbations; the insertion of the working electrode into the
5429
dx.doi.org/10.1021/ja111160r |J. Am. Chem. Soc. 2011, 133, 5425–5431

Journal of the American Chemical Society
ARTICLE
electrolyte at t = 0 resulted in reproducible transient decreases in
the current baselines associated with changes in the nitrogen
pressure inside the mass spec ionization chamber.
under operating conditions, despite the expected significant
improvement in overall cell output power.
The EPOC effect. The enhancement of a nonfaradaic catalytic
surface reaction in the presence of an interfacial potential
associated with an electrochemical faradaic current is referred
to as an “electrochemical promotion of chemical catalysis” effect
(EPOC effect).31-38 The EPOC effect was first described by
Vayenas and his group31 in a heterogeneous catalytic reaction
between gaseous adsorbates on metal surfaces supported on solid
electrolytes. The catalytic metals were typically deposited as
porous films onto solid ion conducting electrolytes,31,32,56 such
as O2- conducting yttrium-stabilized zirconium dioxide
(YSZ).57 Electrochemically induced catalytic rate increases up
to a factor of 20034 were reported. The EPOC effect is perceived
as a phenomenon linking electrocatalysis and heterogeneous
catalysis at the molecular level. Earlier mechanistic hypothesis for
the EPOC effect on solid electrolytes repeatedly cited the role of
modifications of the local work function32-34 of the metal surface
by spilled-over oxygen anions and their effect on the chemisorp-
tion strengths of metal surface adsorbates. While the presence of
spilled over charged species on the catalytic metals has been
confirmed by various groups, a very recent review of the EPOC
effect38 has critically revisited the alleged controlling mechanistic
role of the surface work function in the light of modern surface
science. The review concluded that the surface work function
possibly plays a more minor role in our current understanding of
surface catalysis than previously thought. On the other hand,
more recent concepts and descriptors in heterogeneous catalysis,
such as the d-band center position relative to the Fermi level
energy, have gained importance to explain trends and modifica-
tions in catalytic reactivity.58 A rare case of an EPOC effect in a
liquid electrolyte included the chemical and electrochemical
reaction of mixtures of hydrogen and oxygen on Pt catalysts in
alkaline solutions.34,35 Similarly to solid electrolytes, the mod-
ification of the local work function and surface binding strengths
were cited as the origin of the EPOC effect in liquid electro-
lytes.34 However, no clear physical explanation was provided
whether or how charged surface species can influence surface
properties of the metal electrode.
To explain the observed EPOC effect in the present case of
the HH oxidation reaction we hypothesize the existence of a
hydrazine surface adsorbate, which serves as a reactive inter-
mediate for both the faradaic (reaction 1) and the nonfaradaic
(reaction 5) pathway. This hypothesis would imply that the two
reactions pathways originate on structurally similar surface sites
or defects on the Ni-Co electrode and could share a structu-
rally similar transition state. These sites break N-H bonds and
subsequently adsorb hydrazine molecules, such that they can
further react either along the faradaic or the nonfaradaic path-
way. At electrode potentials positive of the OCP, the local work
function of the catalytic surface increases relative to that at
OCP; this lowers the Gibbs free energy of protons, electrons
and of the transition state and foremost drives the electro-
catalytic oxidative pathway.59 A shared or at least structurally
similar transition state of the two pathways would now also
result in a reaction rate increase of the nonfaradaic chemical HH
decomposition and hence could account for the observed
EPOC effect.
4. CONCLUSIONS
We have developed a novel non-noble Ni60Co40 single-phase
disordered alloy catalyst with unparalleled electrocatalytic activ-
ity for the HH electrooxidation (reaction 1) in potential ranges of
relevance to direct alkaline hydrazine fuel cell anodes. Deploy-
ment of this new catalyst in direct hydrazine fuel cells would
result in significant improvements in terms of output power of
these fuel cell devices.
A mechanistic study of the HH oxidation on the novel non-noble
electrocatalyst revealed the presence of parallel reaction pathways.
In particular, combined cyclic voltammetry and DEMS studies
evidenced the existence of two competing HH reaction pathways.
One pathway is an electroless, purely chemical catalytic decom-
position of HH into molecular hydrogen and nitrogen (reaction 5).
The other is the electrocatalytic conversion of HH into nitrogen and
water. The fact that the Ni-Co alloy catalyst does not electro-
oxidize molecular hydrogen enabled a direct observation of the
nonfaradaic decomposition pathways of hydrazine.
Our DEMS study revealed that the rate of the nonfaradaic,
chemical pathway increased in the presence of an anodic faradaic
(electrochemical) HH oxidation current. This suggested an
electrochemical promotion of a chemical catalytic reaction on a
metal electrode interfacing with an electrolyte (EPOC effect).
Typically, EPOC effects were observed at solid electrolyte/metal
interfaces and mostly for chemical reactions involving
oxygen31,56,57,60 at temperatures above 300 °C. The generally
accepted origin of the EPOC effect on metal electrodes sup-
ported on solid oxide electrolytes lies in the spillover of nega-
tively charges oxide species onto the metal catalyst affecting the
local work function and local reactivity. We hypothesize that the
mechanistic origin of the current low-temperature EPOC effect
in a liquid electrolyte lies in shared or structurally similar
activated transition states and/or adsorbed surface intermediates
of hydrazine oxidation and decomposition. As a result of that,
promotion of the electrochemical oxidative pathway by positive
electrode potentials also enhances the rate of the chemical
hydrazine decomposition. This type of EPOC effect would be
generally relevant where reactants of electrochemical half cells
exhibit parallel chemical decomposition pathways, such as duel
path mechanism often found in the electrocatalytic oxidation of
small organic compounds like formic acid. Here, the electro-
chemical oxidation is believed to proceed in parallel with the
nonfaradaic chemical decomposition of the reactant to surface
bound carbon monoxide, CO.61-66
The coexisting nonfaradaic electrooxidation pathways of
hydrazine represent a serious challenge to the fuel-efficiency of
direct hydrazine fuel cells anodes, as fuel is constantly consumed
and decomposes in hydrogen and nitrogen, even under idle OCP
conditions.
’ ASSOCIATED CONTENT
S
Supporting Information. This material is available free of
b
charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Practical implications for hydrazine-air fuel cells employing
the novel Ni-Co anode catalysts include parasitic fuel loss at
open-circuit (rest) conditions as well as reduced fuel efficiency
Corresponding Author
pstrasser@tu-berlin.de
5430
dx.doi.org/10.1021/ja111160r |J. Am. Chem. Soc. 2011, 133, 5425–5431

Journal of the American Chemical Society
ARTICLE
’ ACKNOWLEDGMENT
(33) Ladas, S.; Kennou, S.; Bebelis, S.; Vayenas, C. G. J. Phys. Chem.
1993, 97, 8845–8848.
(34) Neophytides, S. G.; Tsiplakides, D.; Stonehart, P.; Jaksic,
M. M.; Vayenas, C. G. Nature 1994, 370, 45.
This work was supported by the Department of Energy, Office
of Basic Energy Sciences, under a subcontract with Stanford
Accelerator Laboratory (SLAC). P.S. acknowledges financial
support by the Center of excellence (Exzellenzcluster) in cata-
lysis (“UNICAT”) funded by the German National Science
Foundation (DFG) and managed by the Technical University
Berlin.
(35) Neophytides, S. G.; Tsiplakides, D.; Stonehart, P.; Jaksic, M.;
Vayenas, C. G. J. Phys. Chem. 1996, 100, 14803–14814.
(36) Marwood, M.; Vayenas, C. G. J. Catal. 1998, 178, 429–440.
(37) Vayenas, C. G.; Bebelis, S.; Pliangos, C.; Brosda, S.; Tsiplakides, D.
Electrochemical Activation of Catalysis: Electrochemical Promotion and Metal-
Support Interactions; Kluwer Academic/Plenum Publishers: Dordrecht,
Netherlands, 2001.
(38) Imbihl, R. Prog. Surf. Sci. 2010, 85, 241–278.
(39) Neyerlin, K. C.; Bugosh, G.; Forgie, R.; Liu, Z.; Strasser, P.
J. Electrochem. Soc. 2009, 156, B363–B369.
(40) Schmidt, T. J.; Gasteiger, H. A.; Staeb, G. D.; Urban, P. M.;
Kolb, D. M.; Behm, R. J. J. Electrochem. Soc. 1998, 145, 2354–2358.
(41) Schmidt, T. J.; Gasteiger, H. A. In Handbook of Fuel Cells-
Fundamentals, Technology, and Applications; Vielstich, W., Lamm, A.,
Gasteiger, H., Eds.; Wiley: Chichester, 2003; Vol. 2, Chapter 22, p 316.
(42) Koh, S.; Strasser, P. J. Electrochem. Soc. 2010, 157, B585–591.
(43) Liu, Z.; Koh, S.; Yu, C.; Strasser, P. J. Electrochem. Soc. 2007,
154, B1192–B1199.
’ REFERENCES
(1) Schmidt, E. W. Hydrazine and its derivatives: preparation, proper-
ties, applications; John Wiley & Sons: New York, 1984.
(2) Rothgery, E. F. In Kirk-Othmer Encyclopedia of Chemical Tech-
nology; John Wiley & Sons: New York, 2004; Vol. 13, pp 562-607.
(3) Serov, A.; Kwak, C. Appl. Catal., B 2010, 98, 1–9.
(4) Asazawa, K.; Yamada, K.; Tanaka, H.; Oka, A.; Taniguchi, M.;
Kobayashi, T. Angew. Chem., Int. Ed. 2007, 46, 8024.
(5) Susbielles, G.; Bloch, O. C. R. Acad. Sci. 1962, 255, 685–687.
(6) Szpak, S.; Stonehart, P.; Katan, T. Electrochim. Acta 1965,
10, 563–583.
(44) Koh, S.; Strasser, P. J. Am. Chem. Soc. 2007, 129, 12624–12625.
(45) Reddy, A. K. N.; Rao, B. Can. J. Chem. 1969, 47, 2693–2698.
(46) Strasser, P. J. Comb. Chem. 2008, 10, 216–224.
(47) Hagemeyer, A., Strasser, P., Volpe, A. F., Eds. High-Throughput
Screening in Chemical Catalysis - Technologies, Strategies and Applications;
Wiley VCH: Weinheim, 2004.
(48) Strasser, P.; Fan, Q.; Devenney, M.; Weinberg, W. H.; Liu, P.;
Norskov, J. K. J. Phys Chem B 2003, 107, 11013.
(49) Fergie, R.; Neyerlin, K. C.; Bugosh, G.; Liu, Z.; Strasser, P.
Electrochem. Solid-State Lett. 2010, 13, B36–B39.
(50) Asazawa, K.; Yamada, K.; Tanaka, H.; Oka, A.; Taniguchi, M.;
Kobayashi, T. Angew. Chem. 2007, 46, 1–5.
(51) Liu, Z.; Yu, C.; Rusakova, I.; Huang, D.; Strasser, P. Top. Catal.
2008, 49, 241–250.
(52) Greeley, J.; Norskov, J. K.; Kibler, L. A.; El-Aziz, A. M.; Kolb, D. M.
(7) Yamada, K.; Yasuda, K.; Tanaka, H.; Miyazaki, Y.; Kobayashi, T.
J. Power Sources 2003, 122, 132–137.
(8) Petek, M.; Bruckenstein, S. J. Electroanal. Chem. 1973, 47,
329–333.
(9) Kodera, T.; Honda, M.; Kita, H. Electrochim. Acta 1985, 30,
669–675.
(10) Burke, L. D.; O’Dwyer, K. J. Electrochim. Acta 1989, 34,
1659–1664.
(11) Harrison, J. A.; Khan, Z. A. J. Electroanal. Chem. 1970, 26, 1–11.
(12) Conway, B. E.; Marincic, N.; Gilroy, D.; Rudd, E. J. Electrochem.
Soc. 1966, 113, 1144–1158.
(13) Korovin, N. V.; Yanchuk, B. N. Electrochim. Acta 1970,
15, 569–580.
(14) Ye, L. Q.; Li, Z. P.; Qin, H. Y.; Zhu, J. K.; Liu, B. H. J. Power
Sources 2011, 196, 956–961.
(15) Fukumoto, Y.; Matsunaga, T.; Hayashi, T. Electrochim. Acta
ChemPhysChem 2006, 7, 1032–1035.
(53) Nørskov,J.K.;Bligaard,T.;Logadottir,A.;Kitchin,J.R.;Chen,J.G.;
Pandelov, S.; Stimming, U. J. Electrochem. Soc. 2005, 152, J23–J26.
(54) Baltruschat, H. J. Am. Soc. Mass Spectrom. 2004, 15, 1693–1706.
(55) Wang, H.; Loeffler, T.; Baltruschat, H. J. Appl. Electrochem.
2001, 31, 759–765.
1981, 26, 631–636.
(16) Koꢀrínek, K.; Koryta, J.; Musilovꢁa, M. J. Electroanal. Chem. 1969,
21, 319–327.
(17) Karp, S.; Meites, L. J. Am. Chem. Soc. 1969, 84, 906–912.
(18) Eisner, U.; Zemer, Y. J. Electroanal. Chem. 1972, 34, 81–89.
(19) Yi, Q.; Li, L.; Yu, W.; Liu, X.; Zhou, Z.; Nie, H. Rare Met. 2010,
29, 26.
(56) Vayenas, C. G.; Bebelis, S.; Yentekakis, I. V.; Lintz, H. G. Catal.
Today 1992, 11, 303–442.
(57) Baranova, E. A.; Thursfield, A.; Brosda, S.; Foti, G.; Comninellis,
C.; Vayenas, C. G. J. Electrochem. Soc. 2005, 152, E40–E49.
(58) Hammer, B.; Nørskov, J. K. Surf. Sci. 1995, 343, 211–220.
(59) Bard, A. J.; Faulkner, L. R. Electrochemical Methods; Wiley: New
York, 1980.
(60) Jiang, Y.; Yentekakis, I. V.; Vayenas, C. G. Science 1994, 264,
1563.
(61) Lipkowski, J.; Ross, P. N.; Sun, S. G. Studying Electrocatalytic
Oxidation of Small Organic Molecules with In-Situ Infrared spectroscopy. In
Electrocatalysis; Lipkowski, J., Ross, P. N., Eds.; Wiley-VCH: New York,
1998, p 143.
(62) Jarvi, T. D.; Stuve, E. M., Fundamental Aspects of Vacuum and
Electrocatalytic Reactions of Methanol and Formic Acid on Platinum
Surfaces. In Electrocatalysis; Lipkowski, J., Ross, P. N., Eds.; Wiley-VCH:
New York, 1998, p 75.
(63) Herrero, E.; Chrzanowski, W.; Wieckowski, A. J. Phys. Chem.
1995, 99, 10423–10424.
(20) Rosca, V.; Koper, M. T. M. Electrochim. Acta 2008, 53,
5199–5205.
(21) Perek, M.; Bruckenstein, S. J. Electroanal. Chem. 1973, 47, 329.
(22) Kodera, T.; Honda, M.; Kita, H. Electrochim. Acta 1985, 30, 669.
(23) Karami, P. E.; Jafarloo, R.; Dorraji, P. S. Electrochim. Acta 2009,
54, 5721–5726.
(24) Asazawa, K.; Yamada, K.; Tanaka, H.; Taniguchi, M.; Oguro, K.
J. Power Sources 2009, 191, 362–365.
(25) Dantas, L. M. F.; dos Reis, A. P.; Tanaka, S. M. C. N.; Zagal,
J. H.; Chen, Y.-Y.; Tanaka, A. A. J. Braz. Chem. Soc. 2008, 19, 720–726.
(26) Yamada, K.; Asazawa, K.; Yasuda, K.; Ioroi, T.; Tanaka, H.;
Miyazaki, Y.; Kobayashi, T. J. Power Sources 2003, 115, 236–242.
(27) Yin, W. X.; Li, Z. P.; Zhu, J. K.; Qin, H. Y. J. Power Sources 2008,
182, 520–523.
(28) Shao, J. L.; Hai, Y. Q.; Li, Q. Y.; Bin, H. L.; Zhou, P. L. J. Power
Sources 2010, 195, 4135–4138.
(29) Yamada, K.; Yasuda, K.; Fujiwara, N.; Siroma, Z.; Tanaka, H.;
Miyazaki, Y.; Kobayashi, T. Electrochem. Commun. 2003, 5, 892–896.
(30) Asazawa, K.; Yamada, K.; Tanaka, H.; Taniguchi, M.; Oguro, K.
J. Power Sources 2009, 191, 362–365.
(64) Strasser, P.; Eiswirth, M.; Ertl, G. J. Chem. Phys. 1997, 107, 991.
(65) Strasser, P.; Luebke, M.; Raspel, F.; Eiswirth, M.; Ertl, G.
J. Chem. Phys. 1997, 107, 979.
(66) Strasser, P.; Eiswirth, M.; Koper, M. T. M. J. Electroanal. Chem.
1999, 478, 50.
(31) Vayenas, C. G.; Bebelis, S.; Neophytides, S. J. Phys. Chem. 1988,
92, 5083–5085.
(32) Vayenas, C. G.; Bebelis, S.; Ladas, S. Nature 1990, 343, 625.
5431
dx.doi.org/10.1021/ja111160r |J. Am. Chem. Soc. 2011, 133, 5425–5431