Electro-catalytic conversion of ethanol in solid electrolyte cells for distributed hydrogen generation

Article abstract of DOI:10.1016/j.electacta.2016.07.062

The global interest in hydrogen/fuel cell systems for distributed power generation and transport applications is rapidly increasing. Many automotive companies are now bringing their pre-commercial fuel cell vehicles in the market, which will need extensive hydrogen generation, distribution and storage infrastructure for fueling of these vehicles. Electrolytic water splitting coupled to renewable sources offers clean on-site hydrogen generation option. However, the process is energy intensive requiring electric energy >4.2?kWh for the electrolysis stack and?>6?kWh for the complete system per m³ of hydrogen produced. This paper investigates using ethanol as a renewable fuel to assist with water electrolysis process to substantially reduce the energy input. A zero-gap cell consisting of polymer electrolyte membrane electrolytic cells with Pt/C and PtSn/C as anode catalysts were employed. Current densities up to 200?mA?cm^?2 at 70?°C were achieved at less than 0.75?V corresponding to an energy consumption of about 1.62?kWh?m^?3 compared with >4.2?kWh?m^?3 required for conventional water electrolysis. Thus, this approach for hydrogen generation has the potential to substantially reduce the electric energy input to less than 40% with the remaining energy provided by ethanol. However, due to performance degradation over time, the energy consumption increased and partial oxidation of ethanol led to lower conversion efficiency. A plausible ethanol electro-oxidation mechanism has been proposed based on the Faradaic conversion of ethanol and mass balance of the by-products identified and quantified using ¹H nuclear magnetic resonance spectroscopy and gas chromatography.

Full text of DOI:10.1016/j.electacta.2016.07.062

Electrochimica Acta 212 (2016) 744–757
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Electro-catalytic conversion of ethanol in solid electrolyte cells for
distributed hydrogen generation
a
a
a,
b
HyungKuk Ju , Sarbjit Giddey , Sukhvinder P.S. Badwal *, Roger J. Mulder
a
CSIRO Energy, Private Bag 10, Clayton South, Victoria 3169, Australia
CSIRO Manufacturing, Private Bag 10, Clayton South, Victoria 3169, Australia
b
A R T I C L E I N F O
A B S T R A C T
Article history:
Received 27 January 2016
Received in revised form 11 July 2016
Accepted 12 July 2016
The global interest in hydrogen/fuel cell systems for distributed power generation and transport
applications is rapidly increasing. Many automotive companies are now bringing their pre-commercial
fuel cell vehicles in the market, which will need extensive hydrogen generation, distribution and storage
infrastructure for fueling of these vehicles. Electrolytic water splitting coupled to renewable sources
offers clean on-site hydrogen generation option. However, the process is energy intensive requiring
electric energy >4.2 kWh for the electrolysis stack and >6 kWh for the complete system per m3 of
hydrogen produced. This paper investigates using ethanol as a renewable fuel to assist with water
electrolysis process to substantially reduce the energy input. A zero-gap cell consisting of polymer
electrolyte membrane electrolytic cells with Pt/C and PtSn/C as anode catalysts were employed. Current
Available online 12 July 2016
Keyword:
Electrocatalyst
PEM Electrolysis
Bioethanol reforming
Ethanol oxidation mechanism
Catalytic fuel processing
ꢀ
2
ꢁ
densities up to 200 mA cm at 70 C were achieved at less than 0.75 V corresponding to an energy
ꢀ
3
ꢀ3
consumption of about 1.62 kWh m compared with >4.2 kWh m required for conventional water
electrolysis. Thus, this approach for hydrogen generation has the potential to substantially reduce the
electric energy input to less than 40% with the remaining energy provided by ethanol. However, due to
performance degradation over time, the energy consumption increased and partial oxidation of ethanol
led to lower conversion efficiency. A plausible ethanol electro-oxidation mechanism has been proposed
based on the Faradaic conversion of ethanol and mass balance of the by-products identified and
quantified using 1H nuclear magnetic resonance spectroscopy and gas chromatography.
ã 2016 Elsevier Ltd. All rights reserved.
1. Introduction
intensive [1–3]. Hydrogen at present is produced by natural gas
2
reforming and coal gasification leading to large CO and pollutant
From the >60 million tons per annum of hydrogen produced
globally, over 90% is consumed for ammonia and fertilizer
production, in oil refineries, pharmaceutical industry and for
production of chemicals such as methanol. In future, the demand
for hydrogen as a sustainable fuel for long-term storage and
subsequent utilization in fuel cells for power generation for
residential and transport applications will grow. The hydrogen/fuel
cell systems offer a number of advantages such as distributed
power generation, high efficiency for combined heat and power
gas (NO , SO , and fine particulates) emissions [3]. The electrolytic
x
x
water splitting at low temperature produces pure hydrogen, but is
a very energy intensive process due to the sluggish kinetics of the
oxygen evolution reaction (OER) requiring >6 kWh electric power
3
input per m of hydrogen produced [4–6]. In terms of thermody-
namic energy considerations, the reversible voltage of the overall
ꢁ
ꢁ
water electrolysis reaction is 1.23 V (E rev = ꢀ
DG/nF) at 25 C and
1 atm. However the minimum voltage so-called thermo-neutral
DH/nF) required for the electrolysis process to
start is 1.48 V, corresponding to 3.24 kWh m electric input at
ꢁ
voltage (E the = ꢀ
ꢀ3
generation, zero CO
2
and pollutant emissions at the point of use,
ꢁ
low mechanical noise and excellent load following capability
required for automotive applications. Current hydrogen produc-
tion processes are highly energy and greenhouse gas emission
25 C and 1 atm [4,7]. In order to operate the electrolysis cell at
practical current densities (a measure of hydrogen production
rate), the actual cell voltage required can be in excess of 1.9 V due to
the electrode polarization, electrolyte resistance and other
resistive cell components, thus increasing the electric power input
ꢀ3
required to around 4.2 kWh m for the electrolysis stack and to
ꢀ3
*
Corresponding author.
over 6 kWh m for the system once auxiliary system components
E-mail address: Sukhvinder.Badwal@csiro.au (S.P.S. Badwal).
http://dx.doi.org/10.1016/j.electacta.2016.07.062
013-4686/ã 2016 Elsevier Ltd. All rights reserved.
0

H.K. Ju et al. / Electrochimica Acta 212 (2016) 744–757
745
are taken into consideration. Recently, to reduce the anode
2. Experimental
polarization losses due to the slow OER, many researchers have
focused on exploring new approaches for example using water
vapor as a feedstock [8], activated multiwall carbon nanotube
catalysts for OER [9], Pt-group-metal catalysts combined with an
anion exchange membrane alkaline electrolysis [10].
2.1. Chemicals and materials
All reagents used in this work were of analytical grade. High
purity 40% Pt/C (E-TEK) and a bimetallic alloy of 20% PtSn/C (Pt:
Sn = 3:1 by atomic ratio, Premetek) were used for preparing
catalyst inks for the electrodes. Nafion membrane 115 from
As another alternative to reduce the energy input, the
utilization of alcohols (hydrocarbon solution) with co-electrolysis
of water the so-called organic solution assisted water electrolysis
has been proposed as a viable option. After the publication of a
report by NASA’s Jet Propulsion Laboratory [11] on methanol
assisted water co-electrolysis using Pt-based catalysts in polymer
electrolyte membrane (PEM) based cells, a few studies have been
made in this area on organic solution assisted co-electrolysis
process using methanol [12–20], ethanol [21–23], bio-ethanol [24],
formic acid [25–27], glycerol [28], and ethylene glycol [29] as the
hydrocarbon source. Among all alcohols, hydrogen production
from ethanol is a very attractive candidate owing to its well-
established supply chain from various biomass sources (renewable
source), relatively high hydrogen to carbon ratio (six H atoms), less
toxicity, and ease of transportation and storage [30,31].
TM
ꢀ1
DuPont (typical thickness: 127
m
m, conductivity: 0.1 S cm at
ꢁ
25 C) was used as an electrolyte (PEM) and Nafion solution (5 wt%,
1100EW, Ion Power, Inc., USA) as an ionomer for preparing the
electrodes. Ethanol (CH
were purchased from Merck, Germany. All solutions were prepared
with ultra-pure water using de-ionized water (18 M cm) filtering
system via a Millipore 0.22 m filter. Anode and cathode gas
diffusion layers (GDL) selected were respectively hydrophilic (SGL
25AA, Germany, no wet proof, thickness: 190 m), and hydropho-
3 2 2 2 2 4
CH OH), 2-propanol, H O and H SO
V
m
m
bic (Toray TGP-H-120, Japan, 30 wt% wet proof, thickness: 370
carbon papers.
mm)
2.2. Membrane electrode assembly preparation
The possible reactions of organic solution assisted water
electrolysis are summarized in Table 1, including reversible cell
A membrane electrode assembly (MEA) consisted of a hydrogen
electrode (cathode), PEM and an oxygen/carbon dioxide electrode
(anode). The catalyst inks were prepared by mixing appropriate
amounts of catalysts, ultra-pure water, 2-propanol, and Nafion
ionomer. The electrodes were prepared by spraying catalyst ink on
ꢁ
ꢁ
voltage (E rev), thermoneutral voltage (E the), and theoretical
energy consumption for the electrolysis process in each case.
The major merit for this approach is that the anode overpotential
ꢁ
can be dramatically reduced to below 0.5 V (E theforethanol = 0.26 V)
ꢁ
ꢀ2
as compared to for water electrolysis (E theforwater = 1.48 V). Thus,
the GDL using conventional spray method. Pt/C (2 mg cm ) or
ꢀ2
this approach for hydrogen generation has the potential to
substantially reduce the electric energy input (>60%).
PtSn/C (2 mg cm ) were used as anode catalysts and Pt/C
ꢀ
2
(0.5 mg cm ) as a cathode catalyst. The hydrophilic carbon paper
for anode was used to facilitate the diffusion of ethanol solution to
the electrode/membrane reaction interface, while the hydrophobic
Inthework reportedinthispaper, investigations have beenmade
onethanol assistedwater electrolysis(EAWE)using PEMbasedzero-
gap cell configuration. In order to make an assessment of the
practical viability of the process, the effect of different cell operating
conditions using Pt/C and PtSn/C catalysts has been studied. The
results are also reported on understanding the degradation
behaviour of the electrolysis cell by performing extended stability
tests and a detailed post mortem analysis of electrodes. A plausible
ethanol electro-oxidation mechanism has been proposed based on
the Faradaic conversion of ethanol and mass balance of the by-
products identified and quantified using 1H nuclear magnetic
resonance spectroscopy and gas chromatography.
GDL was used for cathode to improve H
2
release from the
electrolyte/electrode interface and diffusion through the pores of
the carbon paper to interconnect flow channels [16,17]. In order to
+
clean the polymer membrane, and convert to the acid H form, the
membrane was pre-treated by boiling in 5 wt% H
ultra-pure water and then keeping in 0.5 M H SO
2
O
2
and rinsing in
ꢁ
2
4
for 1 hr at 90 C.
The pre-treated Nafion membranes were stored in ultra-pure
water bath. The MEA was assembled by hot pressing the electrodes
ꢁ
on both sides of the dried Nafion 115 membrane at 130 C under
applied pressure of 5 MPa. For the conventional PEM water
Table 1
Theoretical reactions for organic solution assisted water electrolysis.
ꢁ
ꢁ
ꢁ
Power consumption based on E the
Feed solution
Methanol
Reaction
No. of electron
6
E
rev/V
E
the/V
ꢀ
3
/
kWh m
Anode CH OH þ H O ! CO þ 6Hþ þ 6eꢀ
0.02
0.08
ꢀ0.17
0.22
1.23
0.23
0.26
ꢀ0.02
0.09
1.48
0.50
0.57
<0.1
0.20
3.24
3
2
2
þ
ꢀ
Cathode 6H þ 6e ! 3H
2
Overall CH
3
OH þ H
2
O ! CO
2
þ 3H
2
Anode C H OH þ 3H O ! 2CO þ 12Hþ þ 12eꢀ
Ethanol
12
2
2
5
2
2
þ
ꢀ
Cathode 12H þ 12e ! 6H
2
Overall C
2
H
5
OH þ 3H
2
O ! 2CO
2
þꢀ6H
2
þ
Formic acid
Glycerol
Water
Anode HCOOH ! CO þ 2H þ 2e
2
þ
ꢀ
Cathode 2H þ 2e ! H
2
Overall HCOOH ! CO
2
þ H
2
Anode C H O þ 3H O ! 3CO þ 14Hþ þ 14eꢀ
14
2
3
8
8
þ
2
ꢀ
2
Cathode 14H þ 14e ! 7H
2
Overall C
3
H
8
O
8
þ 3H
2
O ! 3CO
2
þꢀ 7H
2
þ
Anode H O ! 2H þ 1=2O þ 2e
2
2
þ
ꢀ
Cathode 2H þ 2e ! H
Overall H þ 1=2O
O ! H
2
2
2
2
*
Note:
The
power consumption
is
directly
proportional
to
the
cell
voltage
according
to
Faraday’s
law
À
Á
ꢀ3
nF
1
ꢀ3
ꢀ3
Energyconsumption=kWhm ¼ VcellIt ¼ Vcell
ꢂ
ꢂ
¼ 2:19 kWh m ꢂVcell
.
Conventional water electrolyzer requires normally up to 4.2 kWh
m
3
600
24:47
ꢀ
1
3
ꢀ1
ꢀ1
ꢁ
(
46 kWh kg , at applied voltage of 1.8–2 V) of electrical energy consumption. H
2
= 24.47 m mo1
and density = 0.09 g L
(at 25 C and 1 atm).

746
H.K. Ju et al. / Electrochimica Acta 212 (2016) 744–757
electrolysis cell, the oxygen electrode was prepared by brush
coating Ir on the Pt sputtered Ti mesh, and hydrogen electrode was
prepared by following the same procedure as given above for the
ethanol assisted electrolysis. A more detailed description of this
MEA fabrication has been reported in a previous study [32]. The
four packed columns to separate H
2
, CO
2
, ethylene (C
2 4
H ), ethane
(C ), acetylene (C ), O , methane (CH
2
H
6
2
H
2
2
4
), and CO and a dual
thermal conductivity detector (TCD). The liquid byproducts from
anode and cathode chambers were analyzed by H nuclear
magnetic resonance (NMR; Bruker Av400 NMR spectrometer)
spectroscopy operating at 400.13 MHz for H observation and
1
2
1
active area of the cells used in this study was 9 cm .
equipped with an inverse 5 mm broadband probe. Samples were
held at 298.1 ꢃ0.1 K during data acquisition and spectra were
acquired without deuterium field locking. The water signal was
suppressed using shaped pulse pre-saturation. The relaxation
2
.3. Electrolysis setup and evaluation
The electrolysis cells were assembled by employing two
titanium plates, each of 5 mm thick with parallel flow channels
2.0 mm wide, 1.0 mm deep, and spaced 2.0 mm apart, ca. 60% of
delay used was >5 ꢂ T
1
for all signals to ensure complete relaxation
and, therefore, accurate peak integrations.
(
the area devoted to fluid flow) on both sides of the MEA. The
schematic of the alcohol assisted water electrolysis setup is shown
in Fig. 1. Ethanol solution was circulated in the anode chamber of
the cell with the help of a peristaltic pump (Masterflex L/S Easy-
The surface morphology and the elemental qualitative analysis
of the electrode catalyst layer were carried out via field emission
scanning electron microscopy (FESEM; Merlin-Zeiss Ultra-Plus
with a Gemini II column, Germany) equipped with a back-
scattered-electron detector (BSE; Zeiss) and with an energy
ꢀ1
Load II, Cole-Parmer) at a flow rate of 20 mL min , while the H
produced in the cathode chamber was passed through
2
2
a
dispersive X-ray spectroscopy (EDX; X-Max 80 mm , Oxford
dehumidifier to separate moisture from hydrogen. All experiments
were carried out under an ambient pressure and the temperature
of the solutions circulated in the anode chamber was controlled
Instruments, UK). The crystallinity and phase analysis of the
MEA were performed using X-ray diffraction (XRD; D2 Phaser,
Bruker AXS, Germany). The surface chemical composition and
electronic state of the electrode catalyst layer were analyzed via X-
ray photoelectron spectroscopy (XPS) using an AXIS Ultra DLD
spectrometer (Kratos Analytical Inc., UK) with a monochromated
ꢁ
ꢁ
between 40 C and 70 C as measured at the inlet to the anode
chamber. The electrolysis cells were operated in a test station
previously described [33]. A DC power supply (Powerbox PBX,
Australia) was used to supply voltage to the cell. The voltage-
current characteristics of the cell were obtained by setting the
value of the current and monitoring the cell voltage by a digital
multimeter (Keithley 192 DMM). The extended performance
stability tests were carried out using a VersaSTAT 4 (Princeton
Applied research, USA). The electrochemical impedance spectros-
copy (EIS; IM6, Zahner, Germany) was performed under potentio-
static mode from open circuit voltage (OCV) to 0.9 V at excitation
amplitude of 20 mV. The frequency range chosen for the EIS was
from 10 kHz to 50 mHz with 10 steps per decade at both lower limit
and above 66 Hz (total 55 points) to collect more data at the low
frequency end. In addition, all EIS measurements were conducted
at almost identical times to demonstrate quasi-steady-state
condition [34].
Al Ka source at a power of 144 W (12 kV x 12 mA). During XPS
analysis, the total pressure of the main vacuum chamber was
ꢀ9
ꢀ8
typically between 10
and 10 mbar. Survey spectra were
acquired at a pass energy of 160 eV. Binding energies were
referenced to the graphitic carbon 1 s peak at 284.5 eV.
3. Results and discussion
3.1. Effect of cell operating conditions and type of anode catalyst
3.1.1. Temperature
An electrolysis cell with Pt/C catalyst on anode and cathode
with 1 M ethanol solution was operated at different temperatures.
Fig. 2(a) compares the voltage-current characteristics of the cell at
different temperatures. It is clear from the figure that with increase
in temperature, higher current densities were achieved at the same
cell voltage. For example at 1 V, the increase in temperature from
2.4. Product analysis and MEA characterization
ꢁ
ꢁ
ꢀ2
Hydrogen production rate was carefully measured using an in-
40 C to 70 C almost doubled the current density from 78 mA cm
ꢀ
2
house made Bunte burette to minimize back pressures into the
cathode and it was also cross-checked by a digital volumetric flow
meter (Defender 510, BIOS International Corp., USA). The analysis
of anode and cathode gases was carried out by a customized gas
chromatography (GC; Clarus 500, Perkin Elmer, USA) equipped
with the 2 channel sample switching valve coupled with bypassed
to 155 mA cm . This can be attributed to improvements in ethanol
oxidation reaction (EOR) kinetics and mass-transport of ethanol to
the electrode/membrane interface with increase in temperature.
Another observation that can be made from this figure is the start
of electrolysis process that occurs at significantly lower voltages
compared to 1.48 V as is the case for conventional electrolysis.
Fig. 1. A schematic illustration of the alcohol assisted PEM water electrolysis set-up with circulating solution.

H.K. Ju et al. / Electrochimica Acta 212 (2016) 744–757
747
current densities (in the very low current density region) appear to
touch the Y-axis at voltages closer to the thermo-neutral voltage of
0.26 V for ethanol assisted electrolysis. A similar trend in the V–I
curves was observed, showing a negligible effect of the ethanol
concentration increase from 1 M to 8 M. A similar behavior has
been observed by Caravaca et al. [21,22] who reported that the
influence of the ethanol concentration on the electrolysis
performance was negligible below 0.8 V, and at higher applied
voltages and high concentration of ethanol (above 6 M), the current
densities slightly decreased. There are number of factors that need
to be mentioned here to understand this behavior. The increase in
ethanol concentration leads to higher uptake of water by the
ethanol molecules due to the hygroscopic property of alcohols
(
methanol and ethanol), thus resulting in increase in ethanol
crossover. The decrease in the proportion of water in the
concentrated solutions results in decrease in the membrane ionic
conductivity [21,22,35,36]. These factors appear to negate any
effect of the increase in ethanol molar concentration on V–I curves
as seen in Fig. 2(b). However, the increase in ethanol concentration
gradient that results in faster diffusion of ethanol molecules to the
electrode/electrolyte reaction interface would improve the elec-
trolysis cell performance. Other factors that can also play a
significant role in the cell performance are hydrophilicity of GDL,
solution feeding rate, the thickness of catalyst and diffusion layers
on the carbon paper support and type and loading of catalyst.
Furthermore, the ethanol crossover from the anode to the cathode
across the PEM increases with higher ethanol concentrations,
leading to undesirable reaction phenomena such as mixed
potential and catalyst poisoning at the cathode [37,38]. Although
fuel consumption at the cathode due to ethanol oxidation reaction
(
EOR) is also an issue in direct ethanol fuel cells (DEFCs), in EAWE,
EOR at the cathode is not expected due to the absence of any
oxidants (air and O ) in the electrolysis system.
2
3
.3. Anode catalyst
A comparison of the V–I characteristics of electrolysis cells with
Pt/C and PtSn/C as the anode catalyst and 2 M ethanol solution at
ꢁ
ꢁ
6
0 C and 70 C are shown in Fig. 2(c). The figure also shows V–I
characteristics of an electrolysis cell (Ir-Pt as anode catalyst
2
(
0.4 mg/cm ) deposited on a titanium mesh and Pt/C as a cathode
catalyst deposited on a carbon paper) operated without the
presence of ethanol (pure water only). The data showed a
considerably lower voltage for similar current densities for EAWE
ꢁ
due to lower thermo-neutral voltage (E the = 0.26 V) for ethanol
assisted electrolysis, compared to that for the water electrolysis
ꢁ
(
E
the = 1.48 V) as discussed previously in the introduction section.
Interestingly, the performance of the Sn-modified Pt electro-
ꢀ
2
catalyst (Pt loading = 1.5 mg cm
) exhibited lower operating
voltages and significantly enhanced electro-catalytic activity for
electrochemical EOR. Onset potential with this binary PtSn catalyst
was significantly lower (<0.2 V) than with only Pt catalyst (<0.5 V).
This is also reflected in an initial sharp increase in voltage in the
low current density region using Pt/C due to the activation
polarization loss caused by lower catalytic activity of Pt/C than of
PtSn/C. In general, for both Pt/C and PtSn/C anodes for EAWE, the
operating voltage increased more rapidly with increasing current
density compared to that for normal water electrolysis, due to the
poor OER/ethanol oxidation activity and sluggish kinetics, despite
the use of high loadings of noble metal catalysts [4–6].
Fig. 2. Polarization curves for the alcohol assisted water electrolysis comparing the
activity of Pt-based catalysts under different conditions. (a) The effect of
ꢁ
temperatures (40–70 C) with 1 M ethanol solution and (b) the effect of ethanol
ꢁ
concentration (1–8 M) at 70 C, with 40 wt% Pt/C catalyst. (c) Comparison of the
electrochemical performance of 2 M ethanol solution with 40 wt% Pt/C and 20 wt%
ꢁ
ꢁ
PtSn/C catalysts at 60 C and 70 C; water electrolysis performance using Ir-Pt
catalyst on Ti mesh is also given as a reference. Overall average error for each data
point in Figures (a) to (c) is less than 5%.
3.2. Ethanol concentration
In order to study the effect of molar concentration of ethanol
For the EAWE process, there was around two-fold increase in
current densities at all recorded voltages using the Sn-modified Pt
catalyst compared to Pt/C and the current density of up to
275 mA cm at cell voltage of 1 V and 70 C was achieved. Such
current densities achieved using the binary PtSn catalyst can
probably be ascribed to the enhanced the C ꢀꢀ H bond dissociation
solution on the EAWE, another cell with Pt/C catalyst on anode and
ꢁ
cathode was operated at 70 C with ethanol solutions of different
ꢀ
2
ꢁ
concentrations circulated in the anode chamber. Fig. 2(b) com-
pares the voltage-current characteristics obtained for 1–8 M
concentration ethanol solutions. The curves extrapolated to zero

748
H.K. Ju et al. / Electrochimica Acta 212 (2016) 744–757
and/or the C ꢀꢀ C bond cleavage, as well as in facilitating the
oxidation of intermediates formed due to increase in the
adsorption of water molecules, which is commonly known to
occur due to the combination of the electronic effect and the
bifunctional mechanism [24,39–42]. However, for PtSn/C anode,
the curves at the high current density region appears to form a
steeper slope probably due to the accumulation of intermediate
poisoning species or mass transport type limitation at the anode.
Another observation that can be made from the V–I curves for
PtSn/C catalyst is that the onset potential for EAWE is even lower
(
<0.2 V) than the thermo-neutral voltage (0.26 V). This has been
also reported by other investigators from their work on
electrochemical ethanol oxidation studies and DEFCs, where the
ethanol oxidation onset potential for the PtSn based catalysts was
observed to be in the vicinity of 0.15 V [41–47]. It is possible that
during heating of ethanol solution in the presence of PtSn catalyst
some products such as acetaldehyde and acetic acid are formed
which have lower onset potential for their oxidation (see Section
3.4 below).
Table 2 compares the Faradaic efficiencies, onset potentials for
ꢀ
2
hydrogen evolution, cell voltages at 200 mA cm current density
and the power consumption per unit volume and weight of the
hydrogen produced for the two types of anode catalysts and for the
conventional electrolysis process. The Faradic efficiency for
electrolysis was calculated based on the applied current density
and the measured hydrogen flow rate from the hydrogen chamber
of the electrolysis cell. The measured hydrogen production rate is
close to theoretical quantities with reasonably stable Faradic
efficiency excluding values in the very low current density region
ꢀ2
(
<50 mA cm ) due to the generated water by crossover to cathode
that creates back pressure in the exhaust line. An energy
consumption regarding the hydrogen production achieved was
ꢀ3
about 1.62 kWh m at 0.74 V with a current density of 200 mA
ꢀ
2
Fig. 3. Extended performance testing of the alcohol assisted water electrolysis cells
using 2 M ethanol at 70 C at a constant voltage of (a) 0.85 V using Pt/C for 4 hr and
PtSn/C for 12 hr and (b) 0.95 V using PtSn/C for around 125 hr. *: In region IV, after
ꢄ6 hr of operation the cell was unloaded (open circuit) for a period of time.
cm
using ethanol with the Sn-modified Pt/C electrocatalyst,
ꢁ
leading to a saving of over 50% electrical energy input, compared
ꢀ3
with 3.39 kWh m at 1.55 V required for the conventional water
electrolysis at this low current density. Normally commercial PEM
ꢀ
2
ꢀ2
ꢀ2
based water electrolysis systems operate at 1 A cm or higher
gradual but continuous drop to about 35 mA cm and 25 mA cm
ꢀ
3
current densities and the energy consumption is >4.2 kWh m
.4. Extended performance and degradation behaviour
In order to obtain further assessment of the EAWE technology,
.
respectively. The cell with Pt/C catalyst was tested up to 4 hr,
whereas the testing of the cell with PtSn/C catalyst continued for
12 hr. For the cell with PtSn/C anode, after 4 hr (region II), the
operation was interrupted briefly and started again without
changing the ethanol solution. A small recovery in the current
density was observed, however, the degradation behavior seem to
follow the same trend as before this treatment as clear from the
figure. Such a decay in performance may be associated with the
poisoning of Pt-based catalyst surface by strongly adsorbed
3
extended performance tests were conducted in a chronoamper-
ꢁ
ometry mode at 0.85 V and 0.9 V using 2 M ethanol at 70 C.
Fig. 3(a) shows the variation in current density with time at a
constant voltage of 0.85 V for electrolysis cells with Pt/C and PtSn/C
as the anode catalysts. As previously observed in the V–I
characteristics, the PtSn/C catalyst showed much higher current
intermediate species such as CO, ꢀꢀ CH
x
, and ꢀꢀ CH
x
O which cannot
be easily removed from the catalyst surface even at the higher
applied voltage of 0.85 V [21,45,50–55]. The second treatment after
8 hr (region III) of cell operation involved replacing the complete
ethanol solution with a fresh solution. Now this treatment
ꢀ2
density at the beginning (initial current density; j
i
= 98 mA cm
= 54 mA cm ), owing to its prevalent catalyst
combination that favors dissociative ethanol adsorption
41,44,48,49]. However, the current density for both cells using
)
ꢀ
2
compared to Pt/C (j
i
[
produced a different current ꢀ time behavior. The j jumped to
i
ꢀ
2
ꢀ2
PtSn/C and Pt/C catalysts showed a rapid initial drop and then a
5
4 mA cm and then gradually dropped to 32 mA cm in 4 hr,
Table 2
ꢁ
Summary of electrolysis performance using three different catalysts at 70 C.
ꢀ
3
ꢀ1
Catalyst
Feed solution
Faradic current efficiency for gases/% at 1 A
Onset potential/V
Best performance
Power consumption/kWh m (kWh kg
)
ꢀ
2
/
V at 200 mA cm
Ir-Pt/Ti mesh
Pt/C
PtSn/C
Pure water
2 M ethanol
2 M ethanol
98
96
97
1.44
0.49
0.15
1.55
1.15
0.74
3.39 (37.66)
2.52 (28.00)
1.62 (18.00)
*
Note: Faradic current efficiency of hydrogen was calculated based on the measured gas flow rate at the cathode (hydrogen electrode) when 1 A of current was passed through
the electrolysis cell.

H.K. Ju et al. / Electrochimica Acta 212 (2016) 744–757
749
which is somewhat similar to the current density achieved over a
similar duration as per region I in Fig. 3(a). It seems that by
replacing the ethanol solution with fresh solution, a more
sustained performance is a result of removal of the poisoning
intermediates dissolved in the ethanol solution and accumulated
at the catalyst surface. Regarding the overall degradation behavior,
it seems that two different degradation mechanisms are affecting
the ethanol electrolysis: i) the initial rapid drop in current density
is due to possibly the poisoning of the catalyst and its deactivation
due to the irreversibly adsorbed intermediates at the catalyst
surface; ii) the gradual degradation was caused by the decrease in
the ethanol concentration and continuous accumulation of
intermediate reaction by-products, such as acetaldehyde and
acetic acid (see discussion below), and the continuous recycling of
ethanol solution containing these by-products of incomplete
ethanol electrochemical oxidation. These observations are consis-
tent with literature reports. For example a similar hypothesis,
based on the accumulation of acetaldehyde and/or acetic acid in
the solution and suppression of further electrochemical reaction
on the catalyst surface resulting in hindered C ꢀꢀ C bond breakage,
has also been discussed using a sum-frequency generation
spectroscopy [53], a rotating disk electrode and cyclic voltammetry
[56], and potentiodynamic and potentiostatic on-line differential
electrochemical mass spectrometry measurements [57]. Our
experimental observations are also in good agreement with recent
studies in an alkaline electrochemical oxidation of ethanol in term
of the catalyst deactivation behaviour [58,59].
Another EAWE cell was operated continuously for up to a period
of around 125 hr to assess the commercial viability of the process
by employing regeneration process (replacing the ethanol solution
with a fresh solution after every 24 hr or so), and the results are
shown in Fig. 3(b). Thus each region marked as IV, V, VI, VII and VIII
starts with a fresh ethanol solution. The cell applied potential was
changed to 0.9 V to assist in reducing the surface coverage of the
strongly adsorbed intermediate reaction species. At each regener-
ation step, current densities showed similar value of ca. 145 mA
ꢀ
2
cm at the start and then it decreased sharply down to about
ꢀ2
ꢀ2
60 mA cm
and then stabilized at around 50 mA cm , with
ꢁ
Fig. 4. Nyquist plots of alcohol assisted water electrolysis cell with PtSn/C anode at 70 C showing effect of (a) cell applied voltage, (b) time including fitting of the impedance
data to the equivalent circuit shown in (d), and (c) replacement of spent ethanol solution with a fresh one. Before impedance data were recorded, the cell had undergone
accumulative performance evaluation for 50 hr as shown by Region VI in Fig. 3(b).

7
50
H.K. Ju et al. / Electrochimica Acta 212 (2016) 744–757
degradation rate of ca. 1.7% hr 1. As the performance seems to
almost recover on changing the ethanol solution with a fresh
solution, the degradation seems to be due to the poisoning of the
catalyst and due to intermediates and by-products formed in the
ꢀ
arc, the size of the LF arc changed significantly with the change in
cell operating conditions, and more particularly as noticed on
changing the anode chamber ethanol solution. Therefore, the LF arc
can be attributed to the anodic reactions in the EAWE, and the
diameter of this arc can be considered to be the anodic charge
transfer resistance (Rct,anode). The impedance data obtained were
modelled with Z-view impedance modelling software to further
elucidate ethanol oxidation mechanism. The equivalent circuit
shown in Fig. 4(d) was used to fit our impedance data as it best
described the possible physical processes occurring in the
electrochemical cell under various operating conditions. Similar
equivalent circuits have been used before by other investigators
especially to explain the inductive loop in the low frequency region
which is most likely associated with adsorption of CO-like
intermediates [60,61]. The data fitted reasonably well to this
equivalent circuit as shown by a representative graph (impedance
data obtained after 4 hr operation at 0.9 V) in Fig. 4(b).
Comparing the spectra obtained at different applied potentials,
it was observed that the anode charge-transfer resistance,
decreases systematically with increase in the applied potential.
In Fig. 4(a), an increase in cell potential from 0.7 V to 0.9 V almost
halves the anodic charge transfer resistance with more conspicu-
ous decrease when the cell potential was changed from 0.5 V to
0.7 V. This clearly demonstrates the effect of increasing the cell
potential on enhancing the reaction kinetics. Another observation
that is obvious is the effect on the anodic charge transfer resistance
of cell operating time at a constant applied potential. Comparing
the spectra obtained at 0.9 V at the beginning and after 24 hr
solution by the electro-catalytic EOR (C
2
fuels). Thus, it is
concluded that the EAWE cells can be operated for extended
periods of time at a reasonable current density as long as the
solution is replaced from time to time with a fresh ethanol
solution.
Electrochemical impedance spectroscopy was carried out, after
0 hr of operation of the cell (Region VI in Fig. 3(b)), in the
potentiostatic mode to understand the reaction mechanism and
the ethanol assisted electrolysis operation as a whole. Fig. 4(a)–(c)
show the Nyquist plots obtained for the above electrolysis cell
5
ꢁ
employing PtSn/C anode with 2 M ethanol solution at 70 C at
different cell voltages and operating times. All impedance spectra
showed two arcs (a smaller arc on the high frequency (HF) side and
a large arc on the low frequency (LF) side) and an inductive loop at
the low end of the impedance spectra. The ohmic resistance value
as read from the intersection of HF arc with real axis is around
2
0
.2
V
-cm and is mainly due to the electrolyte membrane
resistance with some contribution from electrodes/diffusion
layers. Despite the change in the applied potential and other
operating conditions, the Nyquist plots indicate
polarization resistance produced by the HF arc. This is most likely
a similar
2
associated with cathode (Rct,cathode: ꢄ1.3
V-cm ) as evidenced by
the negligible effect of the regeneration process (anode chamber
solution change), operating time and cell loading. Unlike the HF
ꢁ
Fig. 5. SEM-BSE images with EDX element maps of C, O, S, F, Sn, and Pt for the PtSn/C anode GDL sample following testing at constant voltage of 0.9 V at 70 C for ꢄ125 hr.

H.K. Ju et al. / Electrochimica Acta 212 (2016) 744–757
751
Table 3
operation showed an increase in the anodic charge transfer
resistance by a factor of about three, as shown in Fig. 4(b). This
appears to be due to the detrimental effect of accumulation with
time of ethanol oxidation intermediates in the solution and other
poisonous by-products on the catalyst surface. Further, Fig. 4(c)
shows the effect at 0.9 V of replacing the partially spent ethanol
solution, after 24 hr operation in Region VI, with a fresh solution.
The impedance spectrum was almost restored to its original state
Structural composition using SEM-EDX analysis for as prepared and after tested for
25 hr anode GDL.
1
Element
As prepared
SEM–EDX/wt%
After 125 hr testing
SEM–EDX/wt%
C
O
F
S
Pt
Sn
Total
77.5
4.2
6.6
0.6
8.6
2.5
100
79.7
4.5
3.7
0.5
9.0
2.6
100
(
before the 24 hr operation or at the beginning of Region VI). The
anodic charge transfer resistance started increasing again as
shown in Fig. 4(c). From this observation it is clearly obvious that
the degradation behavior observed over the 24 hr time period is
related to the reaction intermediates and by-products accumulat-
ed on the electrocatalyst surface and in the solution. There are
several possible factors that contribute to the changes in the LF arc
shape like an ‘evening primrose flower’. At low applied potentials,
the corresponding current density is relatively low and the LF arc
shows a large blooming arc due to sluggish adsorption/desorption
reaction kinetics while the closing up of the LF arc is attributed to
the reduction of the Rct,anode with improvement in the EOR kinetics.
In addition, the Nyquist plots exhibited a clockwise pseudo-
inductive loop at the LF end of the spectra. We hypothesize that
this inductive behavior in the fourth quadrant may be caused due
to the production of more active available (OHads species) leading
prepared PtSn/C catalyst on GDL (Fig. 6), following performance
evaluation for 12 hr (Fig. 6(c)) as reported above in Fig. 3(a) and
after performance evaluation for 125 hr (Fig. 6(d)) as per Fig. 3(b).
In Table 3, the surface chemical compositions of as prepared
and around 125 hr tested PtSn/C anode catalyst on GDL as
determined by SEM-EDX analysis are given. The PtSn catalyst
and carbon support show a stable composition before and after
testing. The Pt:Sn weight ratios were also close to the nominal
values of 3:1. The loss of F ions might be due to gradual
decomposition or leaching out of F from the outer ionomer layer
during the long-term stability test.
XRD was performed to determine the changes in crystallinity
and phases present in the EAWE electrode material before and
after operation and the diffractograms are shown in Fig. 7.
Polycrystalline Pt (JCPDS #4-0802) diffraction peaks correspond-
to more reaction intermediates (COads, ꢀꢀ CHxads, and ꢀꢀ CH
catalyst surface and blocking some of active anode sites [62,63].
The further oxidation of these species to CO is very sluggish in the
x
Oads) at
2
ethanol oxidation reaction and most likely is a rate-limiting step.
Thus, the pseudo-inductive loop at the low frequency is due to
poisoning of the catalyst by CO-like intermediates. The size of the
inductive loop may depend on the reaction potential and the
operating period owing to accumulated adsorbed species and EOR
pathway. These results are in good agreement with the above
hypothesis and with other findings in the literature on electro-
chemical oxidation of ethanol [62,63].
ꢁ
ꢁ
ꢁ
ing to Pt (111) at 39.76 , (200) at 46.24 , and (220) at 67.45 were
3.5. Catalyst characterization and product analysis with mass balance
The morphology and physiochemical changes in the anode have
been characterized both before and after operation of the cell as
per Fig. 3(b) in order to further understand the EAWE process and
the degradation behavior. Fig. 5 shows SEM-BES images with EDS
mapping of the bulk composition of PtSn/C catalyst distribution on
GDL of the electrolysis cell tested for around 125 hr. The EDS
images of different elements show that PtSn/C catalyst was
uniformly distributed throughout the GDL indicating that the
catalyst remained stable during extended performance testing. The
stability of the micro-structure of electrode and catalyst layer
during the 125 hr was also confirmed by FESEM (see Fig. 6). The
high magnification images showed very similar morphology for as
Fig. 7. (a) XRD patterns for determining the crystalline properties of PtSn/C anode.
(
i) As prepared PtSn/C on the GDL; after testing for 12 hr from (ii) electrode and (iii)
membrane sides; and after testing for around 125 hr from (iv) electrode and (v)
membrane sides.
Fig. 6. FESEM images of (a) as received GDL SGL 25AA carbon paper without any catalyst; (b) freshly prepared PtSn/C catalyst on GDL, (c) after performance evaluation for
2 hr as per Fig. 3(a), and (d) after performance evaluation for around 125 hr as per Fig. 3(b).
1

7
52
H.K. Ju et al. / Electrochimica Acta 212 (2016) 744–757
observed in both as prepared and ꢄ125 hr tested catalyst. The
sharp peak observed in both samples in Fig. 7(i), (ii), and (iv) at
discreet particles containing both Sn and oxygen (Fig. 5 and
Table 3) and also confirmed by XPS data in Fig. 8(c) (see below).
Note that XRD of catalyst material from the Nafion membrane side
taken after 12 hr and around 125 hr of operation of the EAWE
ꢁ
around 27 is most likely related to graphite and is a remnant of the
ꢁ
carbon paper used for the GDL. The broad amorphous feature at 25
is associated with the Vulcan XC-72 catalyst support material. In
showed no crystal phase that could be associated with Sn or SnO
2
,
Fig. 7(i), there are a number of reflections associated with SnO
including (110) at 26.61 , (101) at 33.89 , (200) at 37.95 , and (220)
at 54.76, which can be matched to JCPDS #41-1445. The presence
2
,
but only very small peaks associated with SnO are observed after
2
ꢁ
ꢁ
ꢁ
testing the EAWE electrode side (Fig. 7(ii) and (iv)). In addition,
there is no shift in the position of the Pt peaks that suggests no
reaction between the Sn and Pt phases. We propose that the Sn
oxide has been reduced partially in-situ during EAWE especially at
the ethanol (fuel)/PtSn catalyst/membrane reaction interface at
2
of SnO is consistent with the SEM-EDX performed which showed
ꢁ
temperatures below 100 C.
XPS studies were carried out to investigate the surface
oxidation state of Sn catalyst before and after performance testing
of the catalyst as per Fig. 3(b) and the XPS wide-scan spectra in the
binding energy range from 0 eV to 800 eV is shown in Fig. 8(a). It is
clearly obvious that all peaks, assigned to identify the overall
surface element composition, in the as-prepared catalysts also
existed in the spectra after testing. The XPS narrow scan spectra for
Pt and Sn showed that there is no shift of binding energy of Pt 4f5/2
(
74.5 eV) and 4f7/2 (71.3 eV) peaks in the before and after tested
samples, as shown in Fig. 8(b). However, as shown in Fig. 8(c), the
binding energy of Sn 3d5/2 peak shifted to a lower value. The Sn 3d5/
2
peak located at 487.5 eV was assigned to oxidize Sn in the as-
prepared anode, while Sn 3d5/2 peak observed at 486.9 eV for the
tested anode appeared in between the peak for metallic Sn and
SnO
gradually reduced to lower oxidation state (SnO
electrolysis. Zhu et al. [42] suggested that only metallic Sn can
form an alloy phase with Pt while SnO species stay in a non-
2
. It appears that the SnO
2
catalyst surface on the GDL is
x
)
during
2
alloyed state. However, there was no evidence from XRD of a PtSn
alloy formation. The relative lower intensity of Sn 3d and negative
shift of the binding energy is most likely due to the PtSn catalyst at
the surface having higher metallic coordination and consequent
less oxygen coordination following testing of MEA for 125 hr
[64,65].
In general, Sn would convert to SnO
2
under the anodic
polarization conditions (both DEFC and EAWE) due to its high
operating overpotential than the standard redox potential for the
2
+
2+
4+
Sn/Sn (E = ꢀ0.13 V vs SHE) and Sn /Sn (E = +0.15 V vs SHE) [66].
However, House et al. [67] reported that at pH lower than about 2,
the SnO
such as Sn(OH)
2
may be unstable and exists in unstable hydroxide forms
and Sn(OH) . Thus, it has been suggested that
2
4
these hydroxides may assist with ethanol oxidation reaction but
also leading to partial reduction of Sn oxidation state. This behavior
may be one of the contributing factor for the degradation of EOR
activity and may have detrimental effect on the extended
performance of CAWE, in good agreement with above catalyst
characterization results and in accord with the suggested role of Sn
oxide in the electrochemical systems [55,66,68].
During extended testing of the EAWE cell test as per Fig. 3(b)
from 0 to 24 hr, EAWE products were analyzed on both anode (EOR
side) and cathode (hydrogen production) sides by using a 2 channel
1
GC-TCD and a H liquid-state NMR to identify and quantify the
reaction gas and liquid products, respectively. The hydrogen purity
collected from cathode was 98.7% with small amounts of CO
(0.03%), and Ar/O (0.28%) as determined by the obtained GC
data. There is no reason for CO or C to be formed on the
2
(1%),
C
2
H
2
2
2
2 2
H
cathode side and thus we speculate that these gases permeated
from the anode to the cathode side possibly due to the high flow
ꢀ
1
rate (20 ml min ) of ethanol solution. On the anode side, very low
CO yield of only 0.3% was observed partially due to permeation of
2
the gas to the cathode side and also due to very low flow rate of the
product gas as compared to the results presented by some other
investigators [24,48,50,52], where higher CO
2
was reported but
Fig. 8. XPS spectra of the before and after testing PtSn/C anode for around 125 hr. (a)
XPS wide-scan spectra, (b) Pt 4f spectrum, (c) Sn 3d spectrum.
under different operating regimes. It is worth mentioning here that

H.K. Ju et al. / Electrochimica Acta 212 (2016) 744–757
753
the anode effluent of the EAWE cell. Based on NMR data, the fuel
ethanol) conversion efficiency was 17.24% using 2 M ethanol
operated for 24 hr. Initial concentration of ethanol was 2 M (or
2.5 ml EtOH in 450 ml total solution) thus 9.05 ml of ethanol was
converted over the 24 hr period. One mole of ethanol (MW:
(
5
ꢀ
1
ꢀ3
4
6 g mol and density: 0.789 g cm ) is equivalent to 58.3 ml. Thus
the number of moles of ethanol (DN) converted were calculated as
follows:
D
N ¼ Q=neF
5
2:5 mlð2MEtOHin450 mltotal solutionÞ
¼
¼
17:24% ꢂ
5
8:3 mlð1 moleEtOHÞ
1:552 ꢂ10ꢀ1moles
ð1Þ
The average number of electrons (n
e
) transferred in all ethanol
oxidation reactions were calculated from the total charge in
Coulombs (Q = 55,106C) passed during the EAWE for 24 hr
(Fig. 3(b), region V) and the total number of ethanol moles
converted as below:
5
5; 106C
n
e
¼
¼
1
ꢀ1
1
:552 ꢂ 10ꢀ mol ꢂ 96; 485Cmol
3:68 electron transfer
ð2Þ
F is Faraday constant of 96,485C molꢀ1. The number of electrons
transferred for EOR can also be determined based on the molar
ratio of each final product determined from the NMR data.
According to the NMR data, from the 17.24% ethanol that was
converted to other products after 24 hr (Fig. 9(b)): 73% was
converted to acetic acid (net electron transfer of 2.92e with the
reaction requiring 4 electrons); 6% was converted to acetaldehyde
(
involving 2e) corresponding to 0.12e; 19% was converted to
ethane-1,1-diol (involving 2e) corresponding to 0.38e. The balance
of 2% ethanol was converted to some unidentified products and
may involve 2 or 4e. For the unidentified products, an average of 3
electron transfer reactions were assumed corresponding to 0.06e.
Given the small percentage of unidentified products, this value
would make little difference to overall value of electrons
transferred. Thus overall electrons transferred correspond to
Fig. 9. 1H liquid-state NMR (400 MHz) data at 298 K. (a) Overlain NMR spectra of
liquid samples extracted at t = 0; t = 4 hr; t = 8 hr; t = 24 hr (as per Fig. 3 (b)). The
vertical scale of each spectrum has been amplified in order to more clearly show the
reaction products formed. (b) Liquid chemical product concentration (primary Y-
axis) and ethanol conversion rate (secondary Y-axis) for liquid effluent at the anode,
collected at different operation time intervals.
3
.48e in reasonable agreement with 3.68e transfer calculated
from Faraday’s law above. According to the above discussion, it can
be anticipated that the whole EOR is a complex multistep process
and the reaction intermediate products may also result in
secondary oxidation reaction.
2
the electro-catalytic ethanol oxidation to CO can be strongly
affected by multiple parameters such as reaction temperature,
applied potential, selectivity of catalyst, and ethanol concentration
as observed by different investigators [45,50,57].
We can also estimate the quantity of hydrogen production using
two different calculation approaches. Firstly, the hydrogen
production was obtained from the total charge in Coulomb during
the EAWE for 24 hr. From the total charge of 55,106C, the moles of
hydrogen generated were calculated to be 0.2856 (Q/nF) which
equated to 6.99 L at room temperature. Secondly the total
hydrogen production was also estimated from the proportion of
ethanol conversion based on NMR results. In this estimation, we
assumed all unknown product is also ethane-1,1-diol. One mole of
ethanol converted to CH CHO would produce one mole of H as per
1
The H NMR spectra revealed three main products in ethanol
solution on the anode side as shown in Fig. 9(a). The spectra in this
Figure are the water-suppressed spectra and are used here so that
the products can be clearly seen. The peaks centered at 3.49 ppm
(
q, CH
2 3
) and 1.03 ppm (t, CH ) could be assigned to ethanol. The
major reaction products were acetaldehyde (peaks at 9.51 ppm
(
(
(
q, CH) and 2.08 ppm (d, CH
s, CH )). In addition, another set of peaks centered at 1.15 ppm
d, CH ) and at 5.08 ppm (q, CH) can be attributed to the ethane-
,1-diol (1,1-ethanediol). It should be noted that ethane-1,1-diol
3
)) and acetic acid (peak at 1.94 ppm
3
3
3
2
1
reaction (3) below. Thus 0.55 ml of ethanol would produce 0.23 L of
H . Similarly 6.6 ml ethanol would produce 5.54 L H (based on two
was not observed initially after the 4 hr test but was detected
following 8 hr of testing. The oxygenated product of ethane-1,1-
diol might be yielded by the reaction of acetaldehyde with
absorbed water molecules on electro-catalyst [46,69].
2
2
moles of H2 from one mole of ethanol converted to CH COOH
3
(reaction (4) below) and 1.9 ml of ethanol would produce 0.8 L H2
(based on one mole of ethanol converted to CH CH(OH) (reaction
3
2
The concentrations of known compounds were determined
using Bruker’s ERETIC2 quantification tool which correlates the
absolute intensities of two different NMR spectra [70]. The NMR
spectra of the reference solution of known concentration (96.4 mM
(5) below). Thus the total volume of H2 estimated from NMR
results is 6.57 L which compares very favorably with 6.99 L
obtained from the Coulomb’s law considering the complex
electrochemical reactions and formation of some gas products
(e.g. CO , C H ) including vaporization.
3
1,4-dioxane in CDCl ) and the samples were collected under
2
2 2
identical conditions. The concentration of the samples were
obtained by comparison with the reference solution. Fig. 9(b)
shows the relative quantities of the liquid chemical products from
Based on the product analysis data for actual quantity and the
measured total charge, a mass balance was performed. The
calculations show that this EAWE system can produce hydrogen

7
54
H.K. Ju et al. / Electrochimica Acta 212 (2016) 744–757
ꢀ
1
gas ca. 0.3 L hr
at 0.9 V and an average current of ꢄ0.64 A
oxidation of ethanol generating 12-electrons via one-step reaction
(
Fig. 3(b)), corresponding to an energy consumption of 1.92 kWh
(6) can be significantly limited under low temperature electrolysis
ꢀ
3
ꢁ
m
. This is slightly higher in comparison to the previous energy
conditions (<100 C) due to poor catalytic activity of most electro-
ꢀ3
consumption of 1.62 kWh m calculated from the V–I profile at
the beginning of the experiment. This discrepancy is due to the
performance degradation of the cell during the extended test.
Furthermore it is to be noted that if all ethanol had been
catalysts to break the C ꢀꢀ C bonds [21,52]. For such reasons,
previous reported results of ethanol assisted water electrolysis
show lower current densities and faster degradation than
methanol assisted water electrolysis [21,22].
completely converted to CO
2
(12-electron transfer path), only
Herein, we propose a possible overall electro-catalytic ethanol
oxidation mechanism in EAWE using binary PtSn catalyst based on
results presented above along with thermodynamic data and
previous studies as shown in Fig. 10 [24,45–48,52–55,71]. The
proposed mechanism includes the more favorable formation of
1/3rd of ethanol would be required to produce hydrogen rather
than actually the amount of ethanol consumed. Thus due to
performance degradation over time and partial oxidation of
ethanol, the ethanol assisted water electrolysis process has a very
low efficiency.
ꢀ
oxygenated species (OH /OHads) at the Sn-modified Pt electro-
catalyst surface which allows the electrochemical EOR to liquid
products of acetaldehyde, acetic acid, and ethane-1,1-diol. This is
due to improved C ꢀꢀ H bond dissociation according to the
bifunctional catalyst reaction mechanism [24,39–42]. The pres-
ence of Sn on the Pt-based catalyst in DEFCs is known to promote
further electro-catalytic oxidation which could convert acetalde-
3
.6. Electrochemical oxidation mechanism of ethanol in EAWE process
As we discussed above, the generally accepted mechanism for
EOR in EAWE is a multistep process with the reaction inter-
mediates and undesirable by-products. These incomplete electro-
oxidation reactions are dependent on the cell potential, tempera-
ture, and the type/properties of electro-catalysts employed such as
Ru, Sn, or Rh with Pt catalyst. As detailed in Table 4, the electro-
oxidation reaction of ethanol can be divided into two major paths
as follows [24,46,50]:
hyde and CO-like intermediate species to acetic acid as the SnO
species can readily provide OH species [42,48,72]. However, the
2
ꢀ
complete electrochemical ethanol oxidation to CO
2
seems to be
ꢁ
very limited within the operating condition (E > 0.9 V at 70 C) of
EAWE and most likely is a rate-limiting step (e.g. 10 and
12-electron transfer paths in Fig. 10). Based on the results
presented here and data reported in the literature [41,47,53], it
is most likely due to slow kinetics of complete electrochemical EOR
path of reaction (6) and different and conflicting reaction potential
Incomplete electrochemical oxidation path
CH
CH
3
CH
CH
2
OH ! CH
3
CHO þ 2Hþ þ 2eꢀ Eo ¼ 0:36V
ð3Þ
the
required for C ꢀꢀ C bond cleavage (E < 0.4 V) and CO oxidation to CO
2
COOH þ 4Hþ þ 4eꢀ Eo ¼ 0:23V
3
2
OH þ H
2
O ! CH
3
the
(E > 0.7 V).
ð4Þ
ð5Þ
4
. Conclusion
Electrolytic water splitting coupled to renewable sources offers
þ
CH
3
CH
2
OH þ H
2
O ! CH
3
CHðOHÞ þ 2H þ 2eꢀ
2
clean on-site hydrogen generation and storage option, however,
the process is energy intensive requiring >4.2 kWh for the
electrolysis stack and >6 kWh of electric energy input for the
complete system per m of hydrogen produced. In this study, the
utilization of renewable ethanol to assist with water electrolysis
process was investigated with a view to reduce the energy input
Complete electrochemical oxidation path
CH
3
CH
2
OH þ 3H
2
O ! 2CO
2
þ 12Hþ þ 12eꢀ
Eo ¼ 0:26V
the
3
ð6Þ
The incomplete reactions (3) to (5) lead to the formation of by-
products such as acetaldehyde, acetic acid, and ethane-1,1-diol.
Further oxidation of acetaldehyde may also occur as per the
following reaction:
3
per m of hydrogen produced. Two different catalysts, Pt/C and
PtSn/C were employed on the anode side and investigations were
carried out as a function of time, temperature and ethanol
concentration. Faradaic efficiency for hydrogen generation was
close to 97% and current densities of up to 200 mA cm at 70 C
have been achieved at less than 0.75 V corresponding to an energy
COOH þ 2Hþ þ 2eꢀ
Eo
CH
3
CHO þ H
2
O ! CH
3
¼ 0:10V
ð7Þ
the
ꢀ2
ꢁ
ꢀ3
ꢀ3
consumption of about 1.62 kWh m compared with >4.2 kWh m
CH
3
CHO þ 3H
2
O ! 2CO
2
þ 10Hþ þ 10eꢀ Eo ¼ 0:65V
ð8Þ
the
required for conventional water electrolysis cells. Binary PtSn/C
catalyst exhibited significantly enhanced electro-catalytic activity
and had a better tolerance to poisoning intermediate reaction
An important question arises: what are the key factors affecting
the selectivity of the reaction path for the EOR. The complete
Table 4
ꢁ
ꢁ
Summary of possible electrochemical reaction of ethanol with the reversible cell voltage (E rev) and thermoneutral voltage of (E the).
ꢁ
ꢁ
the/V
Reaction
No. of electrons
2
Temperature
E
rev/V
E
ꢁ
C
2
C
2
C
2
H
H
H
5
5
5
OH ! CH
3
CHO þ H
2
25 C
0.18
0.12
0.08
0.04
0.08
0.02
ꢀ0.02
ꢀ0.05
0.47
0.36
0.36
0.23
0.12
0.26
0.16
0.10
ꢁ
1
27 C
ꢁ
OH þ H O ! CH
2
3
COOH þ 2H
2
4
25 C
ꢁ
1
27 C
ꢁ
OH þ 3H
2
O ! 2CO
O ! CH COOH þ H
CHO þ 3H O ! 2CO þ 5H
2
þ 6H
2
12
2
25 C
ꢁ
1
27 C
ꢁ
CH
CH
3
3
CHO þ H
2
3
2
25 C
ꢁ
1
27 C
ꢀ0.12
ꢁ
2
2
2
10
25 C
0.65
ꢁ
1
27 C
0.41
0.52

H.K. Ju et al. / Electrochimica Acta 212 (2016) 744–757
755
Fig. 10. Schematic diagram for overall proposed mechanism of electrocatalytic oxidation of ethanol on PtSn binary catalyst in the ethanol assisted water electrolysis process.
species compared with Pt/C catalyst. However, only a very small %
of CO was produced in the electrolytic process under the
Acknowledgements
2
conditions investigated, providing little evidence of C ꢀꢀ C bond
This research project has been supported by CSIRO Office of the
Chief Executive (OCE) Postdoctoral Fellowship. The authors would
like to thank Dr Ani Kulkarni and Dr Christopher Munnings for
reviewing this manuscript and Daniel Fini and Gary Paul for
providing technical assistance with some experiments. The
authors would also like to acknowledge contribution from Mark
Greaves (SEM, EDX analysis) and Dr Thomas Gengenbach (XPS
analysis) at CSIRO Manufacturing.
break or complete oxidation of ethanol. The major byproduct
1
analyzed by H nuclear magnetic resonance and gas chromatogra-
phy, was acetic acid with small quantities of acetaldehyde and
ethane-1,1-diol, leading to the loss of ethanol energy unless some
value can be placed on these byproducts. Due to incomplete
oxidation of ethanol, a substantial amount of the energy in ethanol
was not recovered in the form of hydrogen. Based on mass balance,
it was established that if all ethanol had been completely converted
2
to CO (12-electron transfer path), only 1/3rd of ethanol would be
required to produce hydrogen rather than actually the amount of
ethanol consumed. Also during extended tests of up to 125 hr, the
cell degradation occurred due to poisoning of the catalyst by
intermediates and by-products formed in the process, although a
part of the degradation appeared to reverse on replenishing the
solution with fresh ethanol solution. A plausible ethanol electro-
oxidation mechanism has been proposed based on the Faradaic
conversion of ethanol and mass balance of the identified by-
products. Thus due to performance degradation over time and
partial oxidation of ethanol, the ethanol assisted water electrolysis
process has a very low efficiency. Better catalyst and process
conditions have the potential to make further improvements to
ethanol conversion and overall process efficiency. Preliminary
economic assessment was performed on hydrogen production
with bio-ethanol assisted water electrolysis by taking into
consideration ethanol cost and reduced energy consumption.
References
[
[
1] J.A. Turner, Sustainable Hydrogen Production, Science 305 (2004) 972–974.
2] R.D. Cortright, R.R. Davda, J.A. Dumesic, Hydrogen from catalytic reforming of
biomass-derived hydrocarbons in liquid water, Nature 418 (2002) 964–967.
[3] S.P.S. Badwal, S. Giddey, C. Munnings, A.I. Bhatt, A.F. Hollenkamp, Emerging
electrochemical energy conversion and storage technologies, Front. Chem. 2
(
2014) 79.
[4] S.P.S. Badwal, S. Giddey, C. Munnings, Hydrogen production via solid
electrolytic routes, WIREs Energy Environ. 2 (2013) 473–487.
[5] M. Carmo, D.L. Fritz, J. Mergel, D. Stolten, A comprehensive review on PEM
water electrolysis, Int. J. Hydrogen Energy 38 (2013) 4901–4934.
[6] J. Lee, B. Jeong, J.D. Ocon, Oxygen electrocatalysis in chemical energy
conversion and storage technologies, Curr. Appl. Phys. 13 (2013) 309–321.
[7] S. Giddey, A. Kulkarni, S.P.S. Badwal, Low emission hydrogen generation
through carbon assisted electrolysis, Int. J. Hydrogen Energy 40 (2015) 70–74.
[8] J.M. Spurgeon, N.S. Lewis, Proton exchange membrane electrolysis sustained
by water vapor, Energy Environ. Sci. 4 (2011) 2993–2998.
[9] X. Lu, W.L. Yim, B.H.R. Suryanto, C. Zhao, Electrocatalytic Oxygen Evolution at
Surface-Oxidized Multiwall Carbon Nanotubes, J. Am. Chem. Soc. 137 (2015)
2901–2907.
ꢀ
1
However, to meet US DOE cost target of US$ 3–3.7 kg for central
or distributed hydrogen production, several factors such as lower
ethanol production and renewable energy costs and significantly
higher ethanol conversion efficiency must co-contribute.
[
10] C.C. Pavel, F. Cecconi, C. Emiliani, S. Santiccioli, A. Scaffidi, S. Catanorchi, M.
Comotti, Highly Efficient Platinum Group Metal Free Based Membrane-
Electrode Assembly for Anion Exchange Membrane Water Electrolysis, Angew.
Chem. 126 (2014) 1402–1405.

7
56
H.K. Ju et al. / Electrochimica Acta 212 (2016) 744–757
[
11] B.J. Nakamura, S.R. Narayanan, T.I. Valdez, W. Chun, Making hydrogen by
electrolysis of methanol, NASA Tech Brief 26 (2002).
[39] A.O. Neto, R.R. Dias, M.M. Tusi, M. Linardi, E.V. Spinacé, Electro-oxidation of
methanol and ethanol using PtRu/C, PtSn/C and PtSnRu/C electrocatalysts
prepared by an alcohol-reduction process, J. Power Sources 166 (2007)
87–91.
[
12] Z. Hu, M. Wu, Z. Wei, S. Song, P.K. Shen, Pt-WC/C as a cathode electrocatalyst for
hydrogen production by methanol electrolysis, J. Power Sources 166 (2007)
4
58–461.
[40] A. Kowal, M. Li, M. Shao, K. Sasaki, M.B. Vukmirovic, J. Zhang, N.S. Marinkovic, P.
[
13] T. Take, K. Tsurutani, M. Umeda, Hydrogen production by methanol–water
solution electrolysis, J. Power Sources 164 (2007) 9–16.
Liu, A.I. Frenkel, R.R. Adzic, Ternary Pt/Rh/SnO
2
electrocatalysts for oxidizing
ethanol to CO , Nat. Mater. 8 (2009) 325–330.
2
[
14] G. Sasikumar, A. Muthumeenal, S.S. Pethaiah, N. Nachiappan, R. Balaji, Aqueous
methanol eletrolysis using proton conducting membrane for hydrogen
production, Int. J. Hydrogen Energy 33 (2008) 5905–5910.
[41] C. Lamy, S. Rousseau, E.M. Belgsir, C. Coutanceau, J.-M. Léger, Recent progress in
the direct ethanol fuel cell: development of new platinum–tin electrocatalysts,
Electrochim. Acta 49 (2004) 3901–3908.
[
15] C.R. Cloutier, D.P. Wilkinson, Electrolytic production of hydrogen from aqueous
acidic methanol solutions, Int. J. Hydrogen Energy 35 (2010) 3967–3984.
16] S. Uhm, H. Jeon, T.J. Kim, J. Lee, Clean hydrogen production from methanol–
water solutions via power-saved electrolytic reforming process, J. Power
Sources 198 (2012) 218–222.
17] A.T. Pham, T. Baba, T. Sugiyama, T. Shudo, Efficient hydrogen production from
aqueous methanol in a PEM electrolyzer with porous metal flow field:
Influence of PTFE treatment of the anode gas diffusion layer, Int. J. Hydrogen
Energy 38 (2013) 73–81.
18] A.T. Pham, T. Baba, T. Shudo, Efficient hydrogen production from aqueous
methanol in a PEM electrolyzer with porous metal flow field: Influence of
change in grain diameter and material of porous metal flow field, Int. J.
Hydrogen Energy 38 (2013) 9945–9953.
[42] M. Zhu, G. Sun, Q. Xin, Effect of alloying degree in PtSn catalyst on the catalytic
behavior for ethanol electro-oxidation, Electrochim. Acta 54 (2009)
1511–1518.
[43] C. Lamy, E.M. Belgsir, J.M. Leger, Electrocatalytic oxidation of aliphatic alcohols:
application to the direct alcohol fuel cell (DAFC), J. Appl. Electrochem. 31
(2001) 799–809.
[44] W.J. Zhou, S.Q. Song, W.Z. Li, Z.H. Zhou, G.Q. Sun, Q. Xin, S. Douvartzides, P.
Tsiakaras, Direct ethanol fuel cells based on PtSn anodes: the effect of Sn
content on the fuel cell performance, J. Power Sources 140 (2005) 50–58.
[45] N. Nakagawa, Y. Kaneda, M. Wagatsuma, T. Tsujiguchi, Product distribution and
the reaction kinetics at the anode of direct ethanol fuel cell with Pt/C, PtRu/C
and PtRuRh/C, J. Power Sources 199 (2012) 103–109.
[46] I. Kim, O.H. Han, S.A. Chae, Y. Paik, S.-H. Kwon, K.-S. Lee, Y.-E. Sung, H. Kim,
Catalytic Reactions in Direct Ethanol Fuel Cells, Angew. Chem. 123 (2011)
2318–2322.
[47] F. Vigier, C. Coutanceau, F. Hahn, E.M. Belgsir, C. Lamy, On the mechanism of
ethanol electro-oxidation on Pt and PtSn catalysts: electrochemical and in situ
IR reflectance spectroscopy studies, J. Electroanal. Chem. 563 (2004) 81–89.
[48] S. Rousseau, C. Coutanceau, C. Lamy, J.-M. Léger, Direct ethanol fuel cell (DEFC):
Electrical performances and reaction products distribution under operating
conditions with different platinum-based anodes, J. Power Sources 158 (2006)
18–24.
[
[
[
[
19] S. Tuomi, A. Santasalo-Aarnio, P. Kanninen, T. Kallio, Hydrogen production by
methanol–water solution electrolysis with an alkaline membrane cell, J. Power
Sources 229 (2013) 32–35.
[
20] S.P. Sethu, S. Gangadharan, S.H. Chan, U. Stimming, Development of a novel
cost effective methanol electrolyzer stack with Pt-catalyzed membrane, J.
Power Sources 254 (2014) 161–167.
[
21] A. Caravaca, F.M. Sapountzi, A. de Lucas-Consuegra, C. Molina-Mora, F. Dorado,
J.L. Valverde, Electrochemical reforming of ethanol–water solutions for pure
H
2
production in a PEM electrolysis cell, Int. J. Hydrogen Energy 37 (2012)
9
504–9513.
[49] F. Colmati, E. Antolini, E.R. Gonzalez, Effect of temperature on the mechanism
[
[
[
22] A. Caravaca, A. de Lucas-Consuegra, A.B. Calcerrada, J. Lobato, J.L. Valverde, F.
Dorado, From biomass to pure hydrogen: Electrochemical reforming of bio-
ethanol in a PEM electrolyser, Appl. Catal. B Environ. 134–135 (134) (2013)
3
of ethanol oxidation on carbon supported Pt, PtRu and Pt Sn electrocatalysts, J.
Power Sources 157 (2006) 98–103.
[50] J. Friedl, U. Stimming, Model catalyst studies on hydrogen and ethanol
3
02–309.
oxidation for fuel cells, Electrochim. Acta 101 (2013) 41–58.
23] Y.X. Chen, A. Lavacchi, H.A. Miller, M. Bevilacqua, J. Filippi, M. Innocenti, A.
Marchionni, W. Oberhauser, L. Wang, F. Vizza, Nanotechnology makes biomass
electrolysis more energy efficient than water electrolysis, Nat. Commun. 5
[51] S.C. Zignani, V. Baglio, J.J. Linares, G. Monforte, E.R. Gonzalez, A.S. Aricò,
Endurance study of a solid polymer electrolyte direct ethanol fuel cell based on
a Pt–Sn anode catalyst, Int. J. Hydrogen Energy 38 (2013) 11576–11582.
[52] R.M. Antoniassi, A. Oliveira Neto, M. Linardi, E.V. Spinacé, The effect of
acetaldehyde and acetic acid on the direct ethanol fuel cell performance using
(
2014) 4036.
24] C. Lamy, T. Jaubert, S. Baranton, C. Coutanceau, Clean hydrogen generation
through the electrocatalytic oxidation of ethanol in a Proton Exchange
Membrane Electrolysis Cell (PEMEC): Effect of the nature and structure of the
catalytic anode, J. Power Sources 245 (2014) 927–936.
2
PtSnO /C electrocatalysts, Int. J. Hydrogen Energy 38 (2013) 12069–12077.
[53] R.B. Kutz, B. Braunschweig, P. Mukherjee, R.L. Behrens, D.D. Dlott, A.
Wieckowski, Reaction pathways of ethanol electrooxidation on polycrystalline
platinum catalysts in acidic electrolytes, J.Catal. 278 (2011) 181–188.
[54] S.C.S. Lai, M.T.M. Koper, Electro-oxidation of ethanol and acetaldehyde on
platinum single-crystal electrodes, Faraday Discuss. 140 (2008) 399–416.
[55] F.C. Simões, D.M. dos Anjos, F. Vigier, J.-M. Léger, F. Hahn, C. Coutanceau, E.R.
Gonzalez, G. Tremiliosi-Filho, A.R. de Andrade, P. Olivi, K.B. Kokoh,
Electroactivity of tin modified platinum electrodes for ethanol
electrooxidation, J. Power Sources 167 (2007) 1–10.
[
[
[
25] E.Ö. Kiliç, A.S. Koparal, Ü.B. Ö gꢀ ütveren, Hydrogen production by
electrochemical decomposition of formic acid via solid polymer electrolyte,
Fuel Process. Technol. 90 (2009) 158–163.
26] W.L. Guo, L. Li, L.L. Li, S. Tian, S.L. Liu, Y.P. Wu, Hydrogen production via
electrolysis of aqueous formic acid solutions, Int. J. Hydrogen Energy 36 (2011)
9
415–9419.
27] C. Lamy, A. Devadas, M. Simoes, C. Coutanceau, Clean hydrogen generation
through the electrocatalytic oxidation of formic acid in a Proton Exchange
Membrane Electrolysis Cell (PEMEC), Electrochim. Acta 60 (2012)
[56] G. Wu, R. Swaidan, G. Cui, Electrooxidations of ethanol, acetaldehyde and
acetic acid using PtRuSn/C catalysts prepared by modified alcohol-reduction
process, J. Power Sources 172 (2007) 180–188.
112–120.
[
[
28] A.T. Marshall, R.G. Haverkamp, Production of hydrogen by the electrochemical
reforming of glycerol–water solutions in a PEM electrolysis cell, Int. J.
Hydrogen Energy 33 (2008) 4649–4654.
29] A. de Lucas-Consuegra, A.B. Calcerrada, A.R. de la Osa, J.L. Valverde,
Electrochemical reforming of ethylene glycol. Influence of the operation
parameters, simulation and its optimization, Fuel Process. Technol. 127 (2014)
[57] S. Sun, M. Heinen, Z. Jusys, R.J. Behm, Electrooxidation of acetaldehyde on a
carbon supported Pt catalyst at elevated temperature/pressure: An on-line
differential electrochemical mass spectrometry study, J. Power Sources 204
(2012) 1–13.
[58] A. de Lucas-Consuegra, A.R. de la Osa, A.B. Calcerrada, J.J. Linares, D. Horwat, A
novel sputtered Pd mesh architecture as an advanced electrocatalyst for highly
efficient hydrogen production, J. Power Sources 321 (2016) 248–256.
[59] D. Basu, S. Basu, Performance studies of Pd–Pt and Pt–Pd–Au catalyst for
electro-oxidation of glucose in direct glucose fuel cell, Int. J. Hydrogen Energy
37 (2012) 4678–4684.
13–19.
[
30] S.P.S. Badwal, S. Giddey, A. Kulkarni, J. Goel, S. Basu, Direct ethanol fuel cells for
transport and stationary applications—A comprehensive review, Appl. Energy.
145 (2015) 80–103.
[
31] M. Ni, D.Y.C. Leung, M.K.H. Leung, A review on reforming bio-ethanol for
[60] A. Kulkarni, S. Giddey, S.P.S. Badwal, G. Paul, Electrochemical performance of
direct carbon fuel cells with titanate anodes, Electrochim. Acta (2014) 34–43.
[61] E.P. Freitas, CO and methanol electrooxidation on Pt/Ir/Pt multilayers
electrodes, Electrochim. Acta (2009) 1999–2003.
hydrogen production, Int. J. Hydrogen Energy 32 (2007) 3238–3247.
[
32] S. Giddey, F.T. Ciacchi, S.P.S. Badwal, High purity oxygen production with a
polymer electrolyte membrane electrolyser, J. Membr. Sci. 346 (2010)
2
27–232.
[62] S.S. Gupta, J. Datta, A comparative study on ethanol oxidation behavior at Pt
and PtRh electrodeposits, J. Electroanal. Chem. 594 (2006) 65–72.
[63] A.R. Bonesi, M.S. Moreno, W.E. Triaca, A.M.C. Luna, Modified catalytic materials
for ethanol oxidation, Int. J. Hydrogen Energy 35 (2010) 5999–6004.
[64] A.K. Shukla, A.S. Aricò, K.M. El-Khatib, H. Kim, P.L. Antonucci, V. Antonucci, An
X-ray photoelectron spectroscopic study on the effect of Ru and Sn additions to
platinised carbons, Appl. Surf. Sci. 137 (1999) 20–29.
[65] S. Uhm, H. Jeon, J. Lee, Electrocatalytic Oxidation of HCOOH on an
Electrodeposited AuPt Electrode: its Possible Application in Fuel Cells, J.
Electrochem. Sci. Technol. 1 (2010) 10–18.
[
[
33] S. Giddey, S.P.S. Badwal, Polymer electrolyte membrane fuel cell as a hydrogen
flow rate monitoring device, Ionics 19 (2012) 523–528.
34] N. Wagner, E. Gülzow, Change of electrochemical impedance spectra (EIS) with
time during CO-poisoning of the Pt-anode in a membrane fuel cell, J. Power
Sources 127 (2004) 341–347.
35] G. Sasikumar, A. Muthumeenal, S.S. Pethaiah, N. Nachiappan, R. Balaji, Aqueous
methanol eletrolysis using proton conducting membrane for hydrogen
production, Int. J. Hydrogen Energy 33 (2008) 5905–5910.
[
[
36] S. Uhm, H.J. Lee, J. Lee, Understanding underlying processes in formic acid fuel
cells, Phys. Chem. Chem. Phys. 11 (2009) 9326–9336.
[66] J. Mann, N. Yao, A.B. Bocarsly, Characterization and Analysis of New Catalysts
[
37] S. Kontou, V. Stergiopoulos, S. Song, P. Tsiakaras, Ethanol/water mixture
for a Direct Ethanol Fuel Cell, Langmuir 22 (2006) 10432–10436.
[67] C.I. House, G.H. Kelsall, Potential-pH diagrams for the Sn/H O-Cl system,
2
1
permeation through a Nafion based membrane electrode assembly, J. Power
Sources 171 (2007) 1–7.
Electrochim. Acta 29 (1984) 1459–1464.
[
38] K. Taneda, Y. Yamazaki, Study of direct type ethanol fuel cells: Analysis of
anode products and effect of acetaldehyde, Electrochim. Acta 52 (2006)
[68] S. Lee, J.D. Ocon, Y. Son, J. Lee, Alkaline CO
2
Electrolysis toward Selective and
Continuous HCOO–Production over SnO
2
Nanocatalysts, J. Phys. Chem. C 119
1627–1631.
(2015) 4884–4890.

H.K. Ju et al. / Electrochimica Acta 212 (2016) 744–757
757
[
69] A. Scheithauer, E. von Harbou, H. Hasse, T. Grützner, C. Rijksen, D. Zollinger, W.
[71] G.A. Camara, T. Iwasita, Parallel pathways of ethanol oxidation: The effect of
ethanol concentration, J. Electroanal. Chem. 578 (2005) 315–321.
[72] S.C. Zignani, V. Baglio, J.J. Linares, G. Monforte, E.R. Gonzalez, A.S. Aricò,
1
13
R. Thiel, H- and C NMR spectroscopic study of chemical equilibria in the
system acetaldehyde + water, AIChE J. 61 (2015) 177–187.
[
70] G. Wider, L. Dreier, Measuring protein concentrations by NMR spectroscopy, J.
Am. Chem. Soc. 128 (2006) 2571–2576.
x
Performance and selectivity of Pt Sn/C electro-catalysts for ethanol oxidation
prepared by reduction with different formic acid concentrations, Electrochim.
Acta 70 (2012) 255–265.