Optimization of Pt-Ir on carbon fiber paper for the electro-oxidation of ammonia in alkaline media

Article abstract of DOI:10.1016/j.electacta.2010.04.040

Plating bath concentrations of Pt(IV) and Ir(III) have been optimized as well as the total catalytic loading of bimetallic Pt-Ir alloy for the electro-oxidation of ammonia in alkaline media at standard conditions. This was accomplished using cyclic voltam

Full text of DOI:10.1016/j.electacta.2010.04.040

Electrochimica Acta 55 (2010) 5287–5293
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Optimization of Pt–Ir on carbon fiber paper for the electro-oxidation
of ammonia in alkaline media
Bryan K. Boggs, Gerardine G. Botte∗
Department of Chemical and Biomolecular Engineering, 165 Stocker Center, Ohio University, Athens, OH 45701, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Plating bath concentrations of Pt(IV) and Ir(III) have been optimized as well as the total catalytic loading
of bimetallic Pt–Ir alloy for the electro-oxidation of ammonia in alkaline media at standard conditions.
This was accomplished using cyclic voltammetry, scanning electron microscopy (SEM), energy dispersive
X-ray (EDX), and statistical optimization tools. Concentrations of Pt(IV) and Ir(III) of the plating bath
strongly influence electrode surface atomic compositions of the Pt–Ir alloy directly affecting the electro-
oxidation behavior of ammonia. Several anode materials were studied using cyclic voltammetry, which
demonstrated that Pt–Ir was the most active catalyst for the electro-oxidation of ammonia. Criteria for
optimization were minimizing the climatic oxidation overpotential for ammonia and maximizing the
exchange current density. Optimized bath composition was found to be 8.844 0.001 g L−1 Pt(IV) and
4.112 0.001 g L−1 Ir(III) based on electrochemical techniques. Physical characterization of the electrodes
by SEM indicates that the plating bath concentrations of Pt and Ir influence the growth and deposition
behavior of the alloy.
Received 9 December 2009
Received in revised form 5 February 2010
Accepted 10 April 2010
Available online 18 April 2010
Keywords:
Pt–Ir alloy
Ammonia oxidation
Optimization
Anode material
Hydrogen production
© 2010 Elsevier Ltd. All rights reserved.
1. Introduction
(Eq. (3)), 0.06 V are required [3,7].
2NH3(aq) + 6OH− → N2(g) + 6H2O + 6e−
6H2O + 6e− → 3H2(g) + 6OH−
2NH3(aq) → N2(g) + 3H2(g)
(1)
(2)
(3)
1.1. Ammonia electrolysis
Aqueous ammonia’s high capacity for hydrogen storage has led
to increased interest in using ammonia as an alternative energy
carrier [1,2]. Extracting this hydrogen from ammonia can be accom-
plished through the use of a novel electrochemical approach.
According to Vitse et al. [3,4], ammonia electrolysis consumes
95% less energy than water electrolysis theoretically at stan-
In addition to creating pure hydrogen and nitrogen with >99.99%
Faradaic efficiency at room temperature and pressure [3], elec-
trolyzing ammonia remediates nitrate contamination in ground
and drinking water caused by human and animal excreta. These
contaminations are believed to be an epidemic [8]. Current tech-
nologies used to desalinate nitrates from wastewater are expensive
and have long retention times [9–12]. Ammonia-rich wastewater
from breweries, tanneries, domestic wastewaters, landfills, fertil-
izer plants, brine wastewater, etc. are alkaline in nature [9,12–15].
From an energetic and ecological point of view, ammonia electrol-
ysis and optimization thereof plays an important role in improving
everyday life.
dard conditions and produces hydrogen at a cost of $0.89 kg−1
,
which is significantly less than the $2–3.00 kg−1 goal set forth
by the U.S. Department of Energy (DOE) [5]. Researchers at the
University of Florida, studying the utilization of domestic fuels
for hydrogen production, have found that ammonia-based solar-
powered electrolysis produces the cheapest hydrogen ($ GJ−1
)
compared to all other common hydrogen production technolo-
gies by the year 2024 [6]. Ammonia in alkaline media is oxidized
at the anode (Eq. (1)) at a potential of −0.77 V versus standard
hydrogen electrode (SHE). Alkaline reduction of water occurs on
the cathode (Eq. (2)) and requires −0.83 V versus SHE. Overall
1.2. Electro-oxidation of ammonia: catalyst selection
Assuming negligible kinetic limitations during the electrolysis
of ammonia at 25 ◦C, 1.55 Wh g−1 of hydrogen is required. This
same gram of hydrogen theoretically generates 33 Wh from a pro-
ton exchange membrane fuel cell (PEMFC). Potentially, 31.45 Wh
of net energy are available from an ammonia electrolytic cell
(AEC) and PEMFC coupling. Thermodynamics of ammonia elec-
∗
Corresponding author. Tel.: +1 740 593 9670; fax: +1 740 593 0873.
E-mail address: botte@ohio.edu (G.G. Botte).
0013-4686/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2010.04.040

5288
B.K. Boggs, G.G. Botte / Electrochimica Acta 55 (2010) 5287–5293
trolysis is favorable; however, large anodic overpotentials have
been observed on monometallic Pt electrodes [3,7]. Given a PEM-
FCs 50–60% electrical efficiency [16,17] and the increased energy
consumption for ammonia electrolysis due to the electro-oxidation
overpotential, the possible net energy from the coupling is threat-
ened. There has been a significant amount of research regarding
which catalyst(s) are best suited for ammonia oxidation. Noble
transition metals such as Pt, Rh, Pd, and Ir have demonstrated to
be active for ammonia oxidation whereas coinage metals like Cu,
Ag, and Au are not [18,19]. Electro-catalytic alloys and bimetallic
depositions have proven to improve electro-kinetics of ammonia
oxidation [3]. According to Moran et al. [19], a binary alloy of Pt–Ir
has higher activity in alkaline media than other metals because Ir
is the most selective metal for oxidizing ammonia. Very limited
research regarding this alloyed electro-catalyst in particular has
been performed [3,19–21], which is the aim of the present paper.
1.3. Objectives of the study
Fig. 1. Schematic representation of statistical approach used for optimization. Cen-
tral composite circumscribed (CCC) was the type of CCD used. Each corner of the
square represents full factorial points. Stars represent axial points determined as a
function of alpha. The central circle represents the six central points which are all
at the same conditions making the system more robust.
In the present paper, the most active electro-catalyst anode
material for oxidizing ammonia in alkaline media at standard con-
ditions was determined. Electroplating was the technique used
for preparation of the electrodes. Within this context, the specific
objectives of this paper are:
done to ensure that only the 2 cm × 2 cm CFP was being deposited
on. The electrodes were rinsed with acetone and HPLC-grade ultra-
pure water (Fisher Scientific). They were dried in an oven, and the
electrode weights were recorded. This allowed for catalyst load-
ing determination (electrode weight after plating minus electrode
weight before plating).
1. Determine the most active (minimal overpotential and max-
imum exchange current density) electro-catalyst for the
electro-oxidation of ammonia in alkaline media using carbon
fiber paper (CFP) electrodes.
2. Optimize the plating bath of the most active electro-catalyst as
well as the total catalytic loading based on geometrical surface
area based on minimizing the climatic ammonia oxidation over-
potential and exchange current density.
2.3. Anode catalyst – Objective 1
Table 1 shows the eight mono, bi, and ternary catalyst plat-
ing conditions. The concentration of each metal in the bath
was 0.160 0.001 g L−1. All of the salts were 99.99% pure from
Alfa Aesar. Deposition potentials were experimentally determined
using cyclic voltammetry. The experiments were performed using
an electrochemical setup shown in Fig. S2 of Boggs et al. [22]. All
electrodes in this study were plated potentiostatically with this
same setup. A 2.5-cm stir bar at 60 rpm kept the bath solutions
mixed during experimentation minimizing concentration gradi-
ents. Koslow Scientific supplied the Ag/AgCl reference electrode
(+0.1999 V versus SHE) supported by a home-made Luggin capil-
lary and filled with its respective electrolyte. The tip of the Luggin
capillary was placed 1 mm from the center of the working electrode.
Platinum foil (0.01 cm thick, 99.999% pure from ESPI Metals) acted
as the anode for plating except Ni. For plating Ni, Ni foil (0.127 mm
thick, 99.9%) from Alfa Aesar was utilized. The Ni electrode was
plated using the common Watts bath [23]. All of the plating solu-
tions prepared for this paper were solvated with ultrapure high
performance liquid chromatography (HPLC) water.
3. Quantify and qualify the electrode surface composition and mor-
phology using scanning electron microscopy (SEM) and energy
dispersive X-ray (EDX), respectively.
2. Experimental/materials and methods
2.1. Experimental setup and procedure
All chemicals and supplies were high purity (>99.9%) and
supplied from Alfa Aesar or Fisher Scientific. A Solartron 1281
Multiplexer potentiostat was used for the electrochemical stud-
ies throughout this paper. Statistical electrode optimization was
accomplished using Stat-Ease Design Expert® 7.0. After studying
the electrochemical performance of the electrodes, their surface
morphology and atomic compositions were analyzed using a JEOL
JSM-5300 SEM with a combined EDX from EDAX. Experimen-
tal errors were calculated using equipment uncertainties through
propagation of error.
2.2. Electrode preparation
2.4. Pt–Ir optimization matrix – Objective 2
Electrodes were prepared similar to those presented in Fig. 1
of Boggs and Botte [2]. Ti foil (0.127 mm thick, 99.9% pure from
Alfa Aesar) acted as the current collector. It was cut with a pair
of scissors so that a 2 cm × 2 cm square was open. The remaining
Ti on the sides (0.8 cm) of the square acted as arms, which were
bent in half vertically. A sandwich-style packet 2.8 cm wide and
2 cm high containing two sheets of untreated Toray TGP-H-030
carbon fiber paper (CFP) with Ti gauze (18 mesh 99.9% pure Alfa
Aesar) in between was placed in the square opening. Then, the
half-vertically bent arms were closed and pressed holding the car-
bon fiber paper/Ti gauze packet in. Cellophane tape was used to
mask the exposed Ti foil that is present in the plating bath. This was
A standard response surface methodology (RSM) with central
composite design (CCD) was the statistical design of experiments
used for this process optimization. CCD (shown in Fig. 1) is a full
factorial matrix at multiple levels that builds quadratic models
for the response variables without the need of a complete three-
level experiment and is the most common tool utilized for process
optimization. The two responses of this design are the climatic
ammonia oxidation overpotential (ꢀ) and exchange current den-
sity (i0) both of which can be obtained from cyclic voltammograms.
Minimizing the overpotential will reduce the energy required for
electrolysis and maximizing the exchange current density will

B.K. Boggs, G.G. Botte / Electrochimica Acta 55 (2010) 5287–5293
5289
Table 1
Plating conditions for various metals.
Metal
Anode (foil)
Electrolyte
Salts
Temperature (◦C)
Plating potential
(V versus Ag/AgCl)
Rh
Ru
Pt
Pt–Ir
Ni
RhPt
RuPt
RhPtIr
Pt
Pt
Pt
Pt
Ni
Pt
Pt
Pt
1 M HCl/HPLC
1 M HCl/HPLC
1 M HCl/HPLC
1 M HCl/HPLC
0.5 M B(OH)3/HPLC
1 M HCl/HPLC
1 M HCl/HPLC
1 M HCl/HPLC
RhCl3·3H2O
78
78
78
78
45
78
78
78
−0.12
−0.12
−0.12
−0.12
−0.80
−0.12
−0.10
−0.11
RuCl3·3H2O
H2PtCl6·6H2O
H2PtCl6·6H2O + IrCl3·3H2O
NiSO4·7H2O + NiCl2·6H2O
RhCl3·3H2O + H2PtCl6·6H2O
RuCl3·3H2O + H2PtCl6·6H2O
RhCl3·3H2O + H2PtCl6·6H2O + IrCl3·3H2O
Fig. 2. Methodology for analyzing ammonia oxidation overpotentials and exchange current densities using a cyclic voltammogram.
increase kinetics [24,25]. Fig. 2 shows how these two electro-
chemical characteristics are obtained from cyclic voltammetry. The
overpotentials were obtained versus Hg/HgO accounting for 5 M
KOH. The exchange current densities were obtained from the inter-
cept of the Tafel plots (log current versus overpotential) taken from
the forward scan of the oxidation of ammonia peak.
has a high ratio of surface area to volume and has exhibited minimal
reactivity to a large range of operating conditions. These profiles
were captured using an InfiniteFocus Light Microscope by Alicona.
SEM imaging (c) shows that Pt–Ir deposits on both sides of the CFP
Table 2
Pt–Ir CCD experimental matrix.
Pt(IV) concentration (g L−1), Ir(III) concentration (g L−1), and
electro-catalyst loading (mg cm−2) were the three factors opti-
mized. A schematic representation of CCD is shown in Fig. 1. In
addition to the typical “high” and “low” levels for test matrices, CCD
uses a middle and two axial levels outside the “high” and “low” con-
ditions. The experimental matrix is shown in Table 2. These levels
were chosen based on prior knowledge of plating bath conditions
and loadings [2,7,8]. The experiments were conducted over a 2-day
period in two blocks. Randomization was used to ensure any sys-
tematic effects that may have been present were transformed into
experimental noise. Twelve runs (eight factorial points and four
central points) occurred on Day 1. The remaining eight runs (six
axial points plus 2 more central points) were held on Day 2. Axial
points were chosen using a preset “alpha” of 1.41421 allowing the
system to remain rotatable. A rotatable system refers to the ability
to rotate the design points about the center point and the moments
of distribution of the design remain unchanged [26,27].
Factor
Units
Low level (−1)
High level (+1)
A – Pt(IV) concentration
B – Ir(III) concentration
C – Pt–Ir loading
g L−1
g L−1
(
(
0.001)
0.001)
3.200
3.200
5.5
12.000
12.000
20.0
mg cm−2
(
0.1)
Electrode
Block
Type
Bath ID
Factor
A
B
C
1
2
3
4
5
6
7
8
Day 1
Day 1
Day 1
Day 1
Day 1
Day 2
Day 2
Day 2
Day 2
Day 1
Day 1
Day 1
Day 1
Day 1
Day 1
Day 1
Day 2
Day 2
Day 2
Day 2
Factorial
Factorial
Factorial
Factorial
Center
Axial
Axial
Axial
Axial
Factorial
Factorial
Factorial
Factorial
Center
Center
Center
Axial
1
2
3
4
5
6
7
8
9
1
2
3
4
5
5
5
5
5
5
5
3.200
12.000
3.200
12.000
7.600
0.200
15.000
7.600
7.600
3.200
12.000
3.200
12.000
7.600
7.600
7.600
7.600
7.600
7.600
7.600
3.200
3.200
5.5
5.5
12.000
12.000
7.600
7.600
7.600
0.200
15.000
3.200
3.200
12.000
12.000
7.600
7.600
7.600
7.600
7.600
7.600
7.600
5.5
5.5
12.8
12.8
12.8
12.8
12.8
20.0
20.0
20.0
20.0
12.8
12.8
12.8
0.6
9
10
11
12
13
14
15
16
17
18
19
20
3. Results and discussion
3.1. Active electrode geometric surface area
Axial
Center
Center
24.9
12.8
12.8
Fig. 3 shows (a) a 3D image and (b) surface–depth analysis of
the untreated Toray TGP-H-030 CFP chosen as catalytic support. In
addition to its low cost and great physical properties, this substrate

5290
B.K. Boggs, G.G. Botte / Electrochimica Acta 55 (2010) 5287–5293
Fig. 3. Catalytic substrate analysis. (a) 3D surface image showing different surface heights echoed in the surface profile plot (b). (c) Shows that metallic deposits occur
completely throughout the CFP as well as the exposed surfaces.
support suggesting that total catalytic active area is 8 cm2 from the
two exposed 4 cm2 sheets of CFP. This area was used for current
density and catalytic loading calculations.
and exchange current densities. Ni-only, Rh-only, and Ru-only
did not show signs of oxidizing ammonia which is echoed by
Ge and Johnson [18] and Moran et al. [19]. In terms of mini-
mizing the ammonia oxidation overpotential, catalyst selection
3.2. Possible electro-catalysts for ammonia oxidation
Table 3
Anodic metal comparison for the electro-oxidation of ammonia in alkaline media.
Fig. 4 shows cyclic voltammograms of possible electro-catalyst
candidates for ammonia oxidation in alkaline media. Eight met-
als/combinations were studied. The carbon fiber paper electrodes
were deposited with each respective metal/combination using
the process parameters shown in Table 1. Electrodes were plated
with 20 0.1 mg. At first glance, Pt–Ir offers the largest exchange
current density and second smallest ammonia oxidation over-
potential based on Fig. 4. Table 3 lists the metal/combinations
along with their ammonia oxidation overpotentials versus Hg/HgO
Catalyst
ꢀ (mV) versus SHE ( 0.1)
i0 (mA cm−2) ( 0.1)
Rh
Ru
Pt
Ni
Pt–Ir
Pt–Rh
Pt–Ru
Pt–Ir–Rh
738.4
749.4
498.6
698.1
394.9
455.1
440.6
366.0
0.4
0.3
6.2
0.3
9.3
7.4
1.0
2.5

B.K. Boggs, G.G. Botte / Electrochimica Acta 55 (2010) 5287–5293
5291
Fig. 6. As the Pt atomic composition of the electrode increases, so does the plating
efficiency. This is based on Pt only suggesting that the deposition of Ir decreases the
plating efficiency.
Fig. 4. Anode metal comparison with cyclic voltammetry at 5 mV s−1 and 25 ◦C. A
16 cm2 Ni-foil counter electrode was used. Pt–Ir exhibited the best electrochemi-
cal behavior for oxidizing ammonia based on the criteria of minimizing ammonia
oxidation overpotential and maximizing the Tafel slope.
Fig. 6 which shows that the plating efficiency is higher with pro-
portional increase in the atomic composition of Pt. Also, there is
a proportional trend between the ammonia oxidation overpoten-
tial and the loading. Similarly, an increase in surface composition
of Pt leads to an increase in exchange current densities while an
is ranked as follows Pt–Ir–Rh > Pt–Ru > Pt–Rh > Pt–Ir > Pt. With
regards to maximizing the exchange current density, the ranking is
Pt–Ir > Pt–Rh > Pt > Pt–Ir–Rh > Pt–Ru. Due to the large exchange cur-
rent density and average oxidation overpotential, Pt–Ir was chosen
as the most active and suitable electro-catalyst to further optimize.
3.3. Pt–Ir plating bath optimization
The experimental matrix in Table 2 yielded nine different plat-
ing bath compositions in terms of Pt(IV) and Ir(III) concentrations.
Fig. 5 shows plating bath characterization by cyclic voltammetry for
each of the nine different baths in order to obtain a constant plat-
ing potential for the study. Based on the results of Fig. 5, electrodes
for Pt–Ir optimization were plated at a constant plating potential of
−0.12 V versus Ag/AgCl. New baths were prepared for each of the
twenty electrodes in Table 2 to ensure Pt(IV) and Ir(III) concentra-
tion changes did not affect deposition behavior.
Table 4 shows the overpotential and exchange current densities
for the design matrix. Also included are the atomic compositions
of Pt and Ir. Plating bath efficiencies were calculated based on the
atomic composition of Pt. The atomic composition of Pt and the
plating efficiency are a function of one another. This is verified in
Fig. 5. Plating potential characterization for Pt–Ir optimization experimental
matrix. Cyclic voltammetry was used with a voltage scan rate of 5 mV s−1. The solu-
tions were stirred at 60 rpm and temperature controlled at 78 ◦C. A 16 cm2 Pt foil
was used for the anode.
Fig. 7. Normal probability plots for experimental matrix factors: (a) climatic ammo-
nia oxidation overpotential; (b) ammonia oxidation Tafel slope. The data points are
approximately linear indicating desired normality in the error term.

5292
B.K. Boggs, G.G. Botte / Electrochimica Acta 55 (2010) 5287–5293
Table 4
Experimental matrix results including atomic surface compositions and plating efficiencies.
Electrode
Factors
Climatic ꢀ (mV)
versus SHE ( 0.1)
i0 (mA cm−2) ( 0.1)
Atomic composition
(%) ( 0.2)
Plating efficiency ꢀ
(%) ( 0.4)
A
B
C
Pt
Ir
1
2
3
4
5
6
7
8
3.200
12.000
3.200
12.000
7.600
0.200
15.000
7.600
7.600
3.200
12.000
3.200
12.000
7.600
7.600
7.60
3.200
3.200
5.5
5.5
5.5
5.5
727.4
387.4
457.7
587.0
831.3
625.8
872.8
819.1
685.0
723.7
809.6
677.7
754.9
697.9
741.3
730.4
649.7
712.4
679.5
680.4
0.8
6.0
1.2
1.6
7.6
1.0
9.9
2.4
2.6
5.1
9.4
2.5
4.4
2.0
3.5
1.5
0.1
3.9
1.3
1.7
67.4
74.4
53.7
55.3
78.0
40.1
64.7
72.9
69.6
75.9
70.1
57.4
56.4
51.8
61.1
77.4
65.4
60.7
68.8
61.1
33.5
26.4
47.9
45.8
22.4
60.6
36.7
28.1
31.2
25.6
30.4
43.8
44.7
49.6
39.1
23.0
35.0
40.8
32.4
39.2
50.1
60.5
34.6
33.3
85.3
26.1
44.3
60.4
51.0
64.4
52.1
46.3
42.9
43.4
51.8
61.7
48.0
43.6
62.8
45.7
12.000
12.000
7.600
7.600
7.600
0.200
15.000
3.200
3.200
12.000
12.000
7.600
7.600
7.600
7.600
7.600
7.600
7.600
12.8
12.8
12.8
12.8
12.8
20.8
20.8
20.0
20.0
12.8
12.8
12.8
0.6
9
10
11
12
13
14
15
16
17
18
19
20
7.60
7.60
7.60
7.60
24.9
12.8
12.8
Table 5
ANOVA results for the system responses. Both models suggested are significant
according to a 95% confidence interval.
Response
Source
p-Value
Comment
(Prob > F)
Climatic ꢀ (mV) versus SHE
Model
0.0002
0.0002
0.0024
0.3032
0.9399
Significant
A – Pt conc.
B – Ir conc.
C – Loading
Lack of fit
Not significant
Significant
i0 (mA cm−2
)
Model
0.0124
0.0646
0.0204
0.0442
0.0417
0.0143
0.2857
0.0706
A – Pt conc.
B – Ir conc.
C – Loading
AB
AC
BC
Lack of fit
Not significant
increase in Ir surface composition decreases the ammonia oxida-
tion overpotential. Based on these tradeoffs, optimization of the
bath is necessary.
Fig. 8. 3D response surface plot at
a .
catalytic loading of 5.5 0.1 mg cm−2
Optimization of the plating process parameters indicate that plating bath of
8.844 0.001 g L−1 Pt(IV) and 4.112 0.001 g L−1 Ir(III) should be used to obtain
a
minimal ammonia oxidation overpotential and maximum exchange current
Energy dispersive X-ray spectroscopy was used for each elec-
trode as well. EDX confirmed that Pt–Ir bimetallic alloy had been
plated on each electrode. Quantitativeless atomic compositions
were obtained using a working distance of 10 mm, an acceleration
voltage of 15 kV, and magnification of 750×.
density.
ꢀNH = 0.65 + 0.056A − 0.042B + 0.013C
(5)
(6)
3
Using the response data in Table 4 in Stat-Ease Design Expert®
7.0, statistical models for each response were generated. Eq. (5) is
the linear response model for the oxidation of ammonia overpoten-
tial. Eq. (6) is the quadratic model response for the exchange current
density of Pt–Ir in alkaline ammonia electrolysis. A is the Pt(IV) con-
centration (g L−1), B is the Ir(III) concentration (g L−1), and C is the
catalytic loading (mg cm−2) of Pt–Ir. Table 5 shows ANOVA analysis
for both responses.
i0 = 3.42 − 0.31A − 0.89B + 0.53C − 0.73AB − 2.45AC + 0.97BC
Fig. 7(a) shows the evenly distributed normalized residual plot
for ammonia oxidation overpotential, and (b) shows residual plot
for the exchange current densities. Using the optimization software
Table 6
Numerically optimized process conditions for plating CFP anodes based on desirability.
Pt (g L−1) ( 0.001)
8.844
Ir (g L−1) ( 0.001)
4.112
Loading (mg cm−2) ( 0.1)
5.5
ꢀ (mV) versus SHE ( 0.1)
i0 (mA cm−2) ( 0.1)
Desirability
0.690
682.4
5.1

B.K. Boggs, G.G. Botte / Electrochimica Acta 55 (2010) 5287–5293
5293
described by Myers [27], each response yielded the optimum plat-
References
ing process parameters. An overall process desirability of 0.690
was achieved for minimizing ammonia oxidation overpotential
and maximizing exchange current density of ammonia electroly-
sis. Stat-Ease® combines individual response desirabilities into one
process desirability. A value of 1.000 indicates an ideal case. Fig. 8
is a 3D plot of the optimized process parameters as a function of
desirability.
[1] G. Thomas, G. Parks. Potential roles of ammonia in a hydrogen economy: a
study of issues related to the use of ammonia for on-board vehicular hydrogen
storage, U.S. Department of Energy, 2006.
[2] B.K. Boggs, G.G. Botte, J. Power Sources 192 (2009) 573.
[3] F. Vitse, M. Cooper, G.G. Botte, J. Power Sources 142 (2005) 1.
[4] G.G. Botte, F. Vitse, M. Cooper, Electrocatalysts for the oxidation of ammonia and
their application to hydrogen production, fuel cells, sensors, and purification
processes, US Patent 7,485,211 (2004).
[5] S. Satyapal, J. Petrovic, C. Read, G. Thomas, G. Ordaz, Catal. Today 120 (2007).
[6] S.T. Mirabal, H.A. Ingley, N. Goel, Y. Goswami, Int. J. Power Energy Syst. 24 (2004)
3.
4. Conclusions
[7] M. Cooper, G.G. Botte, J. Mater. Sci. 41 (2006) 17.
[8] E. Bonnin, E. Biddinger, G.G. Botte, J. Power Sources 182 (2008) 284.
[9] C. Alfafara, T. Kwawmori, N. Nomura, M. Matsumura, J. Chem. Technol. Biotech-
nol. 79 (2004) 3.
[10] D.C. Bouchard, M.K. Williams, R.Y. Surampalli, J. Am. Water Works Assoc. 84
(1992) 9.
Pt–Ir as an anode catalyst is the most suitable material for oxi-
dizing ammonia in alkaline media. The concentrations of the Pt–Ir
plating bath play a major role in the electrochemical behavior of
oxidizing ammonia as well as the overall catalytic loading. These
factors have been optimized for producing electrodes that will
electrolyze ammonia with the lowest ammonia electro-oxidation
overpotential and largest exchange current density. Optimized pro-
cess parameters are summarized in Table 6. As a result, lower
energy is consumed during electrolysis and faster kinetics is accom-
[11] A.F. Bouwman, D.S. Lee, W.A.H. Asman, F.J. Dentener, Global Biogeochem. Cycles
11 (1997) 4.
[12] K. Vijayaraghavan, D. Ahmad, R. Lesa, Ind. Eng. Chem. Res. 45 (2006) 20.
[13] L. Shao, P. He, J. Xue, G. Li, Water Sci. Technol. 53 (2006) 11.
[14] L. Szpyrkowicz, S.N. Kaul, R.N. Neti, S. Satyanarayan, Water Res. 39 (2005) 8.
[15] A.G. Vlyssides, P.K. Karlis, N. Rori, A.A. Zorpas, J. Hazard. Mater. 95 (2002) 1–2.
[16] J.J. Hwang, D.Y. Wang, N.C. Shih, J. Power Sources 141 (2005) 1.
[17] M.W. Melaina, Int. J. Hydrogen Energy 28 (2003) 7.
[18] J.S. Ge, D.C. Johnson, J. Electrochem. Soc. 142 (1995) 10.
[19] E. Moran, C. Cattaneo, H. Mishima, B.A. López de, J. Solid State Electrochem. 12
(2008) 5.
[20] K. Endo, Y. Katayama, T. Miura, Electrochim. Acta 50 (2005) 11.
[21] K. Endo, K. Nakamura, Y. Katayama, T. Miura, Electrochim. Acta 49 (2004) 15.
[22] B.K. Boggs, R.L. King, G.G. Botte, Chem. Commun. 32 (2009) 4859.
[23] J.P. Hoare, J. Electrochem. Soc. 133 (1986) 12.
[24] P.T. Kissinger, W.R. Heineman, Laboratory Techniques in Electroanalytical
Chemistry, 2nd ed., Marcel Dekker Inc., New York, NY, 1984.
[25] C.M.A. Brett, A.M.O. Brett, Electrochemistry: Principles, Methods, and Applica-
tions, Oxford Science Publications, 1993.
[26] M.J. Anderson, P.J. Whitcomb, RSM Simplified: Optimizing Processes Using
Response Surface Methodology for Design of Experiments, Productivity Press,
New York, NY, 2005.
plished. According to the analysis, the lowest loading (5.5 mg cm−2
)
was found to be the optimum value when using the plating bath
conditions reported in Table 6. One important consideration is
that the catalyst loadings used (low, medium, and high) were
broad (between 5.5 and 20 mg cm−2), and the results revealed the
lowest loading as the optimum value. However, no lower load-
ings were evaluated. Therefore, there is the potential to obtain
lower loadings while achieving a similar performance, making
the ammonia electrochemical cell less expensive. Thereby, it is
recommended that a narrower Pt–Ir loading window between
0.1 and 5 mg cm−2 be optimized using the optimized plating
bath conditions found in Table 6 (8.844 0.001 g L−1 Pt(IV) and
4.112 0.001 g L−1 Ir(III)).
[27] R.H. Myers, Response Surface Methodology, Allyn and Bacon Inc., Boston, MA,
1971.