ww.sciencedirect.com

i n t e r n a t i o n a l j o u rn a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 7 3e8 1

Available online at w

journal homepage: www.elsevier .com/locate/he

Efficient hydrogen production from aqueous methanol ina PEM electrolyzer with porous metal flow field: Influenceof PTFE treatment of the anode gas diffusion layer

Anh Tuan Pham, Tomohiro Baba, Tatsuki Sugiyama, Toshio Shudo*

Graduate School of Science and Engineering, Tokyo Metropolitan University, 1-1 Minamiosawa, Hachioji, Tokyo 192-0397, Japan

a r t i c l e i n f o

Article history:

Received 24 April 2012

Received in revised form

8 October 2012

Accepted 10 October 2012

Available online 8 November 2012

Keywords:

Hydrogen production

Methanol electrolysis

Porous metal flow field

Proton exchange membrane (PEM)

Gas diffusion layer

Polytetrafluoroethylene (PTFE)

* Corresponding author. Tel./fax: þ81 426772E-mail address: shudot@tmu.ac.jp (T. Shu

0360-3199/$ e see front matter Copyright ªhttp://dx.doi.org/10.1016/j.ijhydene.2012.10.0

a b s t r a c t

In a proton exchange membrane (PEM) methanol electrolyzer, the even supply of reactant

to and the smooth removal of carbon dioxide from the anode are very important in order to

achieve a high hydrogen production performance. An appropriate design of flow field and

gas diffusion layer (GDL) is a key factor in satisfying the above requirements. Previous

research has shown that hydrogen production performance of the PEM methanol elec-

trolyzer cell was largely improved with a porous flow field made of sintered spherical metal

powder compared with a conventional groove type flow field. Based on this improvement,

the current study investigated the influence of polytetrafluoroethylene (PTFE) treatment of

the anode GDL on hydrogen production performance of the PEM methanol electrolyzer

with porous metal flow fields. Influences of operating conditions such as methanol

concentration and cell temperature with the flow field were also investigated.

Copyright ª 2012, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights

reserved.

1. Introduction describes an aqueous methanol electrolysis process with the

Hydrogen is one of the most promising energy carriers for the

future owing to its clean and sustainable energy. Among

many methods to produce hydrogen, water electrolysis with

a proton exchange membrane (PEM) [1e9] is one of the most

convenient ways since it offers a number of advantages for

the hydrogen production such as high gas purity and

compatibility with renewable energy sources (e.g. wind, solar).

Besides the water, aqueous methanol can also be electrolyzed

to generate hydrogen [10e14]. With lower Gibbs free energy

for methanol, hydrogen production performance in the

aqueous methanol electrolysis proceeds at much lower

voltage than that in the water electrolysis [10e14]. Fig. 1

715.do).2012, Hydrogen Energy P36

following reactions:

Anode :CH3OHþH2O/CO2þ6Hþ þ6e E0a ¼0:016V vs: SHE

(1)

Cathode : 6Hþ þ 6e/3H2 E0c ¼ 0 V vs: SHE (2)

Overall : CH3OHþH2O/3H2 þ CO2 E0 ¼ 0:016 V vs: SHE

(3)

The standard potential for the methanol oxidation reaction

(MOR) shown in Reaction (1) is only 0.016 V vs. the standard

hydrogen electrode (vs. SHE) at 298 K, while that for hydrogen

reduction reaction (HRR) shown in Reaction (2) is 0 V (vs. SHE).

ublications, LLC. Published by Elsevier Ltd. All rights reserved.

Table 1 e Specification of the MEAs.

MEA1 MEA2 MEA3

Polymer membrane Nafion 117 (175 mm)

Reaction area 25 cm2 (5 5 cm)

Anode catalyst PteRu 1.0 mg/cm2

Cathode catalyst Pt 1.0 mg/cm2

PTFE content in

the anode GDL

0 wt% 5 wt% 10 wt%

PTFE content in

the cathode GDL

10 wt%

6e- 6e-

3H26H+6H+CO2

CH3OH + H2O

Anode GDL

Cathode GDL

Cathode CL

Anode CL Membrane

Power supply

Fig. 1 e Aqueous methanol electrolysis.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 7 3e8 174

As a result of the substantially reduced operating voltage,

the energy requirement from methanol electrolysis can be

reduced significantly. In addition to this, the corrosion in the

PEM methanol electrolysis is less severe than that in the PEM

water electrolysis which needs more than 1.23 V to start

producing hydrogen. Carbon-based and stainless steel mate-

rials, therefore, can be used in the PEM methanol electrolysis,

while these materials which are vulnerable to corrosion at

high operating voltage cannot be used in the anode section of

the PEM water electrolysis [7].

At the anode of a PEM methanol electrolyzer cell, aqueous

methanol is transferred to the catalyst layer and decomposed

to carbon dioxide molecules. Because of having the same

function of the anode, the setup for the PEM methanol elec-

trolyzer cell can be quite similar to that for a direct methanol

fuel cell (DMFC) as shown in Fig. 3. It is well known that the

sluggish kinetics of the MOR [15e18], the permeation of

methanol from the anode to the cathode (i.e. methanol

crossover) [15e18], and cathode flooding [15] are key barriers

that has hampered the development of DMFC technology.

Especially, the crossover of methanol causes a severe prob-

lems including a mixed potential on the cathode (i.e.

decreasing the cathode potential), low fuel utilization effi-

ciency, etc. Because of this, most DMFCs uses relatively thick

membrane (e.g. Nafion 117) [19,20] and low methanol

concentration (i.e. <10 wt%) [21] to lower the methanol

crossover. Unlike the DMFC, water is not produced in the

cathode of PEM methanol electrolyzer cell and the flooding is

not a problem. Therefore, the major challenges for better

performance in the PEM methanol electrolyzer cell exist only

on the anode side. Besides improving the sluggish kinetics of

the MOR, the mass transport related to supply of aqueous

methanol and removal of CO2 is also highly concerned as in

the DMFC. Previous research has shown that hydrogen

production performance of the PEMmethanol electrolyzer cell

was largely improved with a porous flow field (PFF) made of

sintered spherical metal powder (SMP) compared with

a conventional groove type flow field [14]. This is attributed to

an increase in effective electrode area by using the porous

material which enables the flow field to supply reactant

evenly to the electrode and remove carbon dioxide smoothly.

A lower interfacial contact resistance with the SMP is also

a reason for the improvement.

Besides the flow field, an appropriate design of the anode

gas diffusion layer (GDL) shown in Fig. 1 is also an interesting

topic for better performance in the PEMmethanol electrolyzer

cell. Since having the same setup as above mentioned,

amembraneelectrodeassembly (MEA) forDMFChasbeenused

in the previous research by the authors [14].While the GDLs on

both sides of MEA for DMFC are usually coated with a poly-

tetrafluoroethylene (PTFE), there have been attempts to opti-

mize the amount of PTFE in the anode GDL in order to improve

the cell performance of DMFC [22e26]. Scott et al. [22] and

Oedegaard et al. [23] found that adding PTFE to the anode GDL

might lead to better gas transfer in the liquid phase and have

a positive effect on the cell performance of a DMFC. Krishna-

murthy and Deepalochani [24] concluded that anode micro-

porous layer and anode GDL should be coated with an appro-

priate amount of PTFE to achieve an ideal cell performance.

Meanwhile, Gogel et al. [25] and Xu et al. [26] found that the

anode GDL of a DMFC need not be wet-proofedwith PTFE from

the view point of enhancing the mass transport of aqueous

methanol solution. So far, there has been no clear consensus

showingwhether thePTFEcontent in theanodeGDLofDMFCis

needed or not. To our knowledge, the influence of PTFE treat-

ment of the anode GDL on the hydrogen production perfor-

mance of the PEM methanol electrolyzer cell has not been

addressedyet. In this study, theMEAswithdifferent amountof

PTFE in the anode GDL were used in experiments to analyze

their influence on hydrogen production performance of the

PEM methanol electrolyzer cell with porous metal flow field.

Moreover, the influences of operating conditions such as the

methanol concentration and the cell temperature on the

performance with the flow field were also investigated.

2. Experimental

2.1. Membrane electrode assembly

Specifications of theMEAs tested in this research are shown in

Table 1. The MEAs were made with 5 layers including anode

GDL, anode catalyst layer (CL), electrolytemembrane, cathode

CL, and cathode GDL as depicted in Fig. 1. Carbon paper

(Toray TGP-H-90) with the thickness of 0.28 mm was used as

GDL on both sides of the three MEAs. The PTFE contents in the

Table 2 e Specification of the tested sintered sphericalmetal powder for flow field.

Composition Fee19Cre12Nie2Mo (JIS SUS316L)

Grain diameter 350e500 mm

Pore ratio 48%

i n t e r n a t i o n a l j o u rn a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 7 3e8 1 75

anode GDL of the three MEAs were varied from 0 to 5 and to

10 wt%, while that in the cathode GDL of three MEAs were the

same amount of 10 wt% as listed in Table 1. A Nafion 117

polymer membrane was used as the electrolyte for all the

MEAs. The catalyst layers were coated on both sides of the

membrane with the reaction area of 25 cm2. All the three

MEAs had the same catalyst loadings. The catalyst loading on

the anode was 1.0 mg/cm2 (PteRu), and that on cathode was

1.0 mg/cm2 (Pt). The catalysts were with carbon support and

the percentage of PteRu or Pt to carbon powderwas 50wt%. At

present, the use of binary PteRu catalyst on the anode

[17,18,27e32] is found to be the most active manner to

improve the sluggish kinetics of the MOR. During the meth-

anol oxidation reaction, methanol adsorbs on Pt surfaces and

undergoes through a sequence of dehydrogenation steps to

form adsorbed CO (i.e. COads) as shown in Reaction (4). COads

strongly adsorbs onto the Pt catalyst and can poison the anode

catalyst if only a pure Pt catalyst is used. Consequently, the

surface area of catalyst is reduced and the cell performance is

degraded. Hence, it is important to remove COads for recov-

ering the catalytic activity of Pt/C. The addition of a second

metal (i.e. Ru) to the Pt catalyst tends to improve the problem

of poisoning by carbon monoxide since Ru act as promoting

sites to oxidize water as in Reaction (7). The hydroxyl radical

(i.e. OHads) from the oxidized watermolecule oxidizes COads to

produce carbon dioxide in Reaction (8).

The bi-functional mechanism of the PteRu catalyst

mentioned above is regulated by the operating potential and

temperature. Iwasita [29] has reported that, at the potential

below 0.4 V, Pt is a good catalyst for methanol adsorption, but

not for water dissociation; while Ru is able to dissociate water

but it cannot adsorb methanol. However, the bi-functional

mechanism of the PteRu catalyst is of limited use with

further increase in potential and temperature. It is well known

that Pt can dissociate water well at high potential [29], and as

reported by Gasteiger et al. [27] and Kauranen et al. [28], Ru can

adsorb methanol at high temperature of above 60 C. For

different operating conditions (i.e. potential and temperature),

theMOR can undergoes through different reactions as follows:

Ptþ CH3OH/PtðCOÞads þ 4Hþ þ 4e (4)

Ruþ CH3OH/RuðCOÞads þ 4Hþ þ 4e (5)

PtþH2O/PtðOHÞads þHþ þ e (6)

RuþH2O/RuðOHÞads þHþ þ e (7)

ðCOÞads þ ðOHÞads/CO2 þHþ þ e (8)

Besides the COads, formaldehyde and formic acid can also be

formed as intermediate species during the MOR [18,29e32]. As

seen in the Reaction (1), the MOR requires the molecular ratio

of water to methanol to be 1:1. A complete 6 electrons anodic

oxidation to CO2 can be completed as the molecular ratio of

water to methanol has to be much higher than the stoichio-

metric 1:1 ratio [18,32]. In contrast, if the water in the anode

catalyst layer is insufficient, the MOR will be incomplete to

form either formic acid or formaldehyde as shown in Reac-

tions (9) and (10):

CH3OHþH2O/HCOOHþ 4Hþ þ 4e (9)

CH3OH/HCOHþ 2Hþ þ 2e (10)

If the low methanol concentration is applied to the PEM

methanolelectrolysis, themolecularratioofwatertomethanol is

muchhigher than thestoichiometric1:1 ratio.Consequently, the

formationofeitherformicacidorformaldehydecanbeneglected.

2.2. Porous flow field

Table2showsspecificationsofthesinteredmetalpowderporous

materialusedforflowfieldof thePEMmethanolelectrolyzercell

in this study. The metal powder was made of a corrosion-

resisting type stainless steel (JIS SUS316L; Fee19Cre12Nie2Mo)

by an inert gas atomization method. Spherical metal powders

with 350e500 mm grain diameter were placed in the separator

block and sintered each other and also sintered to the separator

block. The sintered metal powder is in a porous structure with

a porosity of around 48%. The surface image of the porous

structurewhichwas obtainedwith a laser scanningmicroscope

(Keyence VK-9700) is shown in Fig. 2. The separator block was

madeofthesamematerialasthemetalpowderandhadtheflow

field depth of 2mm. The flowfield volumewas 2.9 cc.

2.3. Experimental setup

This study used a single cell for the experiments. Setup of the

tested hydrogen production cell including end plates, gaskets,

collectors, separators, and MEA is schematically shown in

Fig. 3. The silicon rubber gaskets were used for sealing the cell.

The tightening torque of cell assembling bolts was set evenly

at 120 cNm for all the experiments. The electrolysis was

conducted by applying electric current from a DC power

supply to the electrolyzer at a constant voltage mode to

generate hydrogen at the cathode. Temperatures of the cell

and aqueous methanol were controlled during experiments

by putting them into a constant temperature vessel (As one

DOV-450P). Methanolewatermixtureswere fed into the anode

by using a peristaltic pump, the cell temperature was

measuredwith a thermocouple. AC impedancemeter (Kikusui

KFM2030) was used to measure the cell impedance with

a frequency range from 3000 to 0.1 Hz.

3. Results and discussion

3.1. Influence of PTFE treatment of the anode GDL

3.1.1. Hydrogen production performanceFig. 4 shows the influence of PTFE treatment of the anode GDL

on hydrogen production performance of the PEM methanol

0

0.05

0.10

0.15

0.20

0.3 0.4 0.5 0.6 0.7

10wt% PTFE in the anode GDL

0wt%

5wt%

Aqueous methanol concentration: 2.6MAnode feed rate: 10cc/minCell temperature: 303K

Cell voltage [V]

Cu

rren

t d

en

sity [A

/cm

2

]

Fig. 4 e Influence of PTFE treatment of the anode GDL on

the hydrogen production performance with 2.6 M aqueous

methanol concentration.Fig. 2 e Surface of the tested sintered spherical metal

powder for flow field.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 7 3e8 176

electrolyzer cell with the three MEAs shown in Table 1. The

porous metal flow field which exhibited higher hydrogen

production performance than the conventional groove type

flow field [14] was employed in this study. Aqueous methanol

solution with 2.6 M concentration was fed to the anode at the

flow rate of 10 cc/min. Temperatures of the cell and aqueous

methanol solution were set at 303 K. The obtained results

indicate that current density of the PEMmethanol electrolyzer

cell is increased by reducing the amount of PTFE in the anode

GDL. The MEA without PTFE in the anode GDL shows the

highest current density. Because the current density is theo-

retically proportional to the rate of hydrogen production, this

signifies that hydrogen production performance of the PEM

methanol electrolyzer cell is improved by reducing the

amount of PTFE in the anode GDL, and the MEA without PTFE

in the anode GDL exhibits the best hydrogen production

performance.

3.1.2. Impedance analysisElectrochemical impedance spectroscopy is an effective tool

for investigating the phenomena in fuel cell [33e35] or PEM

electrolysis cell [2,5,6,8]. Since the influence of PTFE amount

Gasket

Anode inlet

End plate

Collector

Separator

Carbon paper

Anode outlet

Cathode outlet

End plate

Collector

Separator

Carbon paper

Cathode outletMEA

Fig. 3 e Setup of the tested electrolyzer cell with proton

exchange membrane.

on cell can be recognized by the changes in cell resistance and

reaction resistance obtained from impedance analysis, the

electrochemical impedance spectroscopy was applied to the

electrolyzer cell in this study. Fig. 5 shows the AC impedance

of the cell measured at the current density of 0.2 A/cm2 for the

three cases with different PTFE contents of 0, 5, and 10 wt% in

the anode GDL. The impedance data were obtained at the

frequencies from 3000 to 0.1 Hz. The figure is a ColeeCole plot

showing the imaginary part Im Z and real part Re Z of the

measured complex impedance Z:

Z ¼ ReZþ jImZ (11)

The first intersection of the arc and horizontal axis corre-

sponds to the cell resistance including membrane, electrodes,

gas diffusion layers, separators, and collectors. While the

difference between the low and high frequency intercept

corresponds to the reaction resistance which includes acti-

vation resistance and diffusion resistance. As seen in Fig. 5,

the cell resistances decrease with reducing the amount of

PTFE in the anode GDL. The reduced cell resistances can be

mainly caused by a lower interfacial contact resistance

between the MEA and anode flow field because the difference

among three setups is just the amount of PTFE in the anode

Aqueous methanol concentration: 2.6MAnode feed rate: 10cc/minCurrent density: 0.2A/cm2

Cell temperature: 303K5wt% PTFE in the anode GDL

10wt%

0wt%

5

10

15

010 15 20 25 30 35

ReZ

-Im

Z

f=0.8Hz

f=2000Hz

Fig. 5 e AC impedance data with different amount of PTFE

in the anode GDL.

Methanol concentration [M]

0.55

0.60

0.65

0.70

0.75

0 1 2 3 4 5 6

with PTFE (MEA3)

w/o PTFE (MEA1)

Anode feed rate: 10cc/minCurrent density: 0.15A/cm2

Cell temperature: 303K

Cell vo

ltag

e [V

]

Fig. 6 e Influence of PTFE treatment of the anode GDL on

the hydrogen production performance with different

aqueous methanol concentrations.

333K40

45

50

55

60

65

70

75

0 1 2 3 4 5 6

303K

Surface tension (mN/m)303K 333K

Water 71.18 66.36Methanol 22.02 19.27

Methanol concentration [M]

Su

rface ten

sio

n [m

N/m

]

Fig. 7 e Surface tension of aqueous methanol solutions.

i n t e r n a t i o n a l j o u rn a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 7 3e8 1 77

GDL. Due to the non-conductive property of PTFE, the increase

in amount of PTFE on the anode GDL surface results in higher

interfacial contact resistance between the MEA and anode

flow field. In addition, water content inside the membrane

may also contribute to the difference in cell resistances. By

reducing the amount of PTFE in the anode GDL, the water

content inside the membrane can be increased. This may

result in higher ionic conductivity of the membrane [37] and

lead to the reduction in cell resistances.

With reducing the amount of PTFE in the anode GDL, the

reaction resistances are also lowered as shown in Fig. 5. The

low frequency range of the ColeeCole plot represents the

diffusion resistance, and the high frequency range represents

the activation resistance. The lower reaction resistance with

reduced PTFE conditions is mostly caused by the reduction in

diffusion resistancewhich appears on the right side of the arc.

An increased wettability of the anode GDL with reducing PTFE

amount can be a main reason for the decreased diffusion

resistance. With the increased wettability of GDL, more

aqueous methanol solution can be diffused through the GDL

to the anode catalyst layer. Here, the removal of CO2 may

adversely affect the diffusion resistance. Some reports indi-

cate that the appearance of PTFE in the anode GDLmay lead to

a better gas transfer in DMFC [22,23]. However, the effect on

diffusion of aqueous methanol solution by the increased

wettability was more dominant as shown in Fig. 5.

The results of impedance analyses suggest that the lower

cell resistance and reaction resistance are the reasons for the

improved hydrogen production by reducing the amount of

PTFE in the anode GDL. Therefore, from a view point of

improving the hydrogen production performance by reducing

the cell resistance and enhancing the supply of reactants to

the anode catalyst layer, the carbon paper without PTFE

treatment can be preferred to be the anode GDL of PEM

methanol electrolyzer cell employing the PFF made of SMP.

3.2. Influence of aqueous methanol concentration

Fig. 6 shows the influence of PTFE treatment of the anode GDL

on hydrogen production of the PEMmethanol electrolyzer cell

with different aqueous methanol concentrations including 1,

2, 2.6, 3, 4, and 6 M at the current density of 0.15 A/cm2. Two

typical MEAs with different PTFE content of 0 and 10 wt% in

the anode GDL were used in the experiments of this section.

The anode feed rate and cell temperature were set at the same

conditions as the experiments shown in Fig. 4. From the

results, it could be firstly seen that the MEA without PTFE in

the anode GDL (i.e. MEA1) exhibited lower applied voltage

compared to the MEA with 10 wt% PTFE in the anode GDL

(i.e. MEA3) at any tested aqueous methanol concentrations.

Because the current density is theoretically proportional to

the rate of hydrogen production, the lower applied voltage

results in higher hydrogen production efficiency.

For both MEAs, the applied voltage decreases with

increasing methanol concentration in the low concentration

range. This may be related to the change in surface tension of

aqueous methanol solution with concentration shown in

Fig. 7. The surface tension in the figure was estimated based

on the experimental data and correlation reported by Vazquez

et al. [36]. Due to the lower surface tension of methanol

compared to water, higher concentration solution has lower

surface tension and higher wettability. The higher wettability

may allowmore aqueousmethanol diffuse through the GDL to

the anode catalyst layer and result in lower applied voltage at

the same current density.

In the high concentration range, the applied voltage tends

to increase with increasingmethanol concentration in spite of

the decrease in surface tension. The change in membrane

conductivity with methanol concentration can be a reason for

this trend. Mukundan et al. [38] reported that proton

conductivity of Nafion 117 immersed in 50 wt% aqueous

methanol was about 35 mS/cm at the temperature of 303 K

which is about 50 mS/cm lower than that in liquid water. As

known, Nafion membrane is described as a series of clusters

interconnected by narrow pores. The hydrophilic regions

around the clusters of sulphonated side chains can lead to the

absorption of large quantities of water. Within these hydrated

regions, the Hþ ions are weakly attracted to the SO3 group and

are able to move. Although the hydrated regions are some-

what separate, it is still possible for the Hþ ions to move

through the supporting long molecule structure of Nafion

membrane if the effective hydrated regions are large enough

to facilitate their movement [15]. Due to the lower polarity of

methanol compared to water, the increase in methanol

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 7 3e8 178

concentration may increase the cluster size of Nafion

membrane [12], and thus may reduce the effective hydrated

regions. This may decrease ionic conductivity of the electro-

lyte membrane and contribute to the increased resistance of

electrolyzer cell as plotted in Fig. 8. The increased cell resis-

tance by increasing methanol concentration leads to higher

applied voltages in the high concentration range as shown in

Fig. 6.

The applied voltage with the MEA1 in Fig. 6 is lowest at

2.6 M aqueous methanol concentration, while that with the

MEA3 is lowest at 3M. The difference in this “minimum-point”

between theMEA1 andMEA3 can be explained by the different

amount of PTFE in the anode GDL. Due to its higher wetta-

bility, the MEA1 which has no PTFE treatment of the anode

GDL may allow more aqueous methanol solution diffuse

through the GDL to the anode catalyst layer compared to the

MEA3 at the same aqueousmethanol concentration. However,

more aqueous methanol solution diffused to the anode cata-

lyst layer may also result in reducing the ionic conductivity of

the electrolyte membrane as above mentioned, thus lowering

hydrogen production performance of the PEM methanol

electrolyzer cell at higher aqueous methanol concentration.

For this reason, the optimum methanol concentration with

the MEA3 is higher than that with the MEA1.

3.3. Influence of cell temperature

Fig. 9 shows the influence of PTFE treatment of the anode GDL

on hydrogen production performance of the PEM methanol

electrolyzer cell with different cell temperatures. Aqueous

methanol solution with 2.6 M concentration was fed to the

anode with the feed rate of 10 cc/min. As seen in Fig. 9, even

with or without PTFE treatment of the anode GDL, the current

densities at the cell temperature of 333 K are found to be

significantly higher than those at the cell temperature of 303 K

when the same voltages are applied. Regardless the cell

temperature, the casewith theMEA1 exhibits higher hydrogen

production performance than the case with the MEA3.

The measured AC impedance shown in Fig. 10 can explain

the improved hydrogen production performance by

increasing cell temperature in Fig. 9. Firstly, with increasing

temperature, the kinetic energy of reaction and the catalytic

11

12

13

14

1 2 2.6 3 4 6

Anode feed rate: 10cc/minCurrent density: 0.15A/cm2

Cell temperature: 303Kw/o PTFE (MEA1)

Methanol con

Ce

ll re

sis

ta

nc

e [m

Ω]

centration [M]

Fig. 8 e Cell resistance as a function of aqueous methanol

concentrations.

activity are enhanced, while the energy difference resulting

from the Gibbs free energy of products and reactants in the

Reaction (3) is decreased towards the negative side (i.e. reac-

tion tends to occur spontaneously). Because of this, the acti-

vation resistance is reduced at higher temperature. In

addition, as can be seen in Fig. 7, the surface tension of

aqueous methanol solution decreases at higher temperature.

The lower surface tension would allow more aqueous meth-

anol solution diffuse through the anode GDL to the anode

catalyst layer, thus reducing the diffusion resistance of reac-

tant. Due to these facts, the reaction resistance which

includes activation resistance and diffusion resistance is

reduced significantly at higher temperature. Secondly, it is

known that the water uptake (i.e. the number of water mole-

cules per sulfonic acid site) increases with temperature,

resulting in higher water content inside themembrane [37]. In

addition, even with the same value of water uptake, the ion

conductivity of Nafionmembrane increases with temperature

[37,38]. The increase in water content and ion conductivity

results in lower ionic resistance of membrane, thus the cell

resistance is reduced at higher temperature. The above anal-

ysis suggests that the reduction in ionic resistance of

membrane and reaction resistance can be the reasons for the

improved hydrogen production performance of each MEAs at

higher cell temperature.

Fig. 11 explains the improved hydrogen production

performance with reducing the amount of PTFE in the anode

GDL at the cell temperature of 333 K. From the results, it can be

seen that the cell resistance decreases with reducing the

amount of PTFE in the anode GDL. The decrease in interfacial

contact resistance and the increase in water content by

reducing the PTFE mentioned in Section 3.1.2 can be the

reasons for the reduced cell resistance. Meanwhile, unlike the

trend for the cell temperature of 303 K shown in Fig. 5, the

reaction resistance is almost similar for the two caseswith the

MEA1 and the MEA3 at higher cell temperature. The lower

surface tension of aqueousmethanol solutionwith increase in

temperature shown in Fig. 7 may be a reason for the insig-

nificant difference in reaction resistance between the two

cases. With the lower surface tension, aqueous methanol is

easily to diffuse through the anode GDL to the anode catalyst

layer, even with PTFE. This may mitigate the advantage of the

0

0.05

0.10

0.15

0.20

0.3 0.4 0.5 0.6 0.7

with PTFE(MEA3)

w/o PTFE(MEA1)

Cell voltage [V]

303K

333K

Cu

rren

t d

en

sity [A

/cm

2

]

Anode feed rate: 10cc/minAqueous methanol concentration: 2.6M

Fig. 9 e Influence of PTFE treatment of the anode GDL on

the hydrogen production performance with different cell

temperatures.

co

nversio

n efficien

cy [%

]

Anode feed rate: 10cc/minAqueous methanol concentration: 2.6M

Co

nsu

med

electric en

erg

y [kW

h/N

m

0.50

0.75

1.00

1.25

1.50with PTFE

(MEA3)

w/o PTFE(MEA1)

80

85

90333K

303K

0

5

10

15

20

0 5 10 15 20 25 30 35ReZ [mΩ]

-Im

Z [m

Ω]

Aqueous methanol concentration: 2.6MAnode feed rate: 10cc/minCurrent density: 0.2A/cm2

w/o PTFE (MEA1)

333K 303K

f=1.58Hzf=2000Hz

Fig. 10 e AC impedance data with different cell

temperatures.

i n t e r n a t i o n a l j o u rn a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 7 3e8 1 79

non-PTFE anode GDL on aqueous methanol diffusion through

the GDL at higher cell temperature, and the reaction resis-

tance is almost similar for the two cases.

E

nerg

y

Current density [A/cm ]

0.05 0.10

75

0.15 0.2070

Fig. 12 e Consumed electric energy and energy conversion

efficiency of the PEM methanol electrolyzer cell with and

without PTFE in the anode GDL at different cell

temperatures.

3.4. Energy conversion efficiency

Fig. 12 shows the consumed electric energy and energy

conversion efficiency for the results in Fig. 9. The energy

conversion efficiency in the figure was calculated with the

higher heating value (HHV) of produced hydrogen and

consumed energy as follows:

h ¼ HHV of produced H2

HHV of consumed MeOHþ Consumed electric energy

(12)

The rates of hydrogen production and methanol

consumption for the Eq. (12) were estimated by the Reaction

(3) with the measured current densities in Fig. 9, while the

HHV values of methanol and hydrogen were at standard

condition (i.e. 298 K, 1 atm).

The results in Fig. 12 indicate that energy conversion effi-

ciency of the PEM methanol electrolyzer cell with the MEA1 is

higher than that with the MEA3 at the both cell temperatures

0

5

10

15

20

0 5 10 15 20 25 30 35

-Im

Z [m

Ω]

ReZ [mΩ]

Aqueous methanol concentration: 2.6MAnode feed rate: 10cc/minCurrent density: 0.2A/cm2

Cell temperature: 333K

w/o PTFE(MEA1) with PTFE

(MEA3)

f=1.58Hzf=2000Hz

Fig. 11 e AC impedance data with the MEA1 and MEA3 at

cell temperature of 333 K.

of 303 and 333 K. Regardless the PTFE treatment of the anode

GDL, the energy conversion efficiency is improved by

increasing cell temperature. The combination of the non-PTFE

anode GDL and the increased cell temperature exhibits the

highest energy conversion efficiency among the four cases

tested here. The smaller applied voltage shown in Fig. 9 which

lowers the consumed electric energy in Fig. 12 is the reason for

the improved energy conversion efficiency. The PEM meth-

anol electrolyzer cell achieves quite high energy conversion

75

80

85

90

w/o PTFE with PTFE w/o PTFEwith PTFE

Aqueous methanol concentration: 2.6MAnode feed rate: 10cc/minCurrent density: 0.2A/cm2

En

erg

y co

nversio

n efficien

cy [%

]

Fig. 13 e Comparison of energy conversion efficiency at the

cell temperatures of ( ) 303 K and ( ) 333 K.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 7 3e8 180

efficiency at low current densities. When increasing the

current density, the energy conversion efficiency tends to

decrease. However, the use of the non-PTFE anode GDL or the

increased cell temperature can ease this deterioration.

Fig. 13 shows the comparison of the energy conversion effi-

ciency at the current density of 0.2 A/cm2 for the four cases in

Fig. 12. By employing the porous flow field made of sintered

spherical stainless steel powder, the cell without the PTFE treat-

mentoftheanodeGDLcanachieveaveryhighenergyconversion

efficiency of approximately 84.2% at a cell temperature of 333 K.

The utilization of awaste heat at this relatively low temperature

can be amethod to save the energy for producinghydrogen.

4. Conclusion

This study newly investigated the influence of PTFE treatment

of the anode GDL on the hydrogen production performance of

PEM methanol electrolyzer cell with porous metal flow field.

Results derived from the experiments can be summarized as

follows:

(1) Hydrogen production performance of the PEM methanol

electrolyzer cell was improved by reducing the amount of

PTFE in the anode GDL. From a view point of improving the

hydrogen production performance, the carbon paper

without PTFE treatment can be preferred to be the anode

GDL of PEM methanol electrolyzer cell employing the PFF

made of SMP.

(2) Themeasured AC impedance data indicated that the lower

cell resistance and reaction resistancewere the reasons for

the higher hydrogen production performance by reducing

the amount of PTFE in the anode GDL. The lower cell

resistance can be mainly explained by the decrease in

interfacial contact resistance by reducing the PTFE which

is non-conductive material and increases the interfacial

contact resistance between the GDL surface and flow field,

while the lower reaction resistance can be attributed to the

increase in wettability of the anode GDL by reducing the

PTFE.

(3) Hydrogen production performance of the cell was

improved by increasingmethanol concentration in the low

concentration range. The lower surface tension of the

solution with increasing methanol concentration can be

a reason for this improvement. In the high concentration

range, the hydrogen production performance tended to

decrease with increasing methanol concentration. This

may be explained by the decrease in the ionic conductivity

of the electrolyte membrane by containing methanol.

(4) Increase in cell temperature improved the hydrogen

production performance in the PEM methanol electrolyzer

cell due to the decreases in ionic resistance of membrane

and reaction resistance. By reducing the PTFE, the cell

performance was also improved at the elevated tempera-

ture. However, the decrease in reaction resistancewas less

significant and the reduced contact resistance was the

main reason for this improvement.

(5) Energy conversion efficiency of the PEM methanol elec-

trolyzer cell was improved by reducing the amount of PTFE

in the anode GDL and increasing cell temperature. By

employing the porous flow fieldmade of sintered spherical

stainless steel powder, the PEM methanol electrolyzer cell

achieved very high energy conversion efficiency of

approximately 84.2% at a cell temperature of 333 K and

a current density of 0.2 A/cm2.

Acknowledgement

The authors express their gratitude to The Sanyo Special Steel

Co., Ltd for help with the sintered spherical metal powder

stainless steel material used for the porous flow field in this

research.

r e f e r e n c e s

[1] Grigoriev SA, Millet P, Volobuev SA, Fateev VN. Optimizationof porous current collectors for PEM water electrolysers.International Journal of Hydrogen Energy 2009;34:4968e73.

[2] Siracusano S, Baglio V, Blasi AD, Briguglio N, Stassi A,Ornelas R, et al. Electrochemical characterization of singlecell and short stack PEM electrolyzers based on a nanosizedIrO2 anode electrocatalyst. International Journal of HydrogenEnergy 2010;35:5558e68.

[3] Ito H, Maeda T, Nakano A, Hasegawa Y, Yokoi N, Hwang CM,et al. Effect of flow regime of circulating water on a protonexchange membrane electrolyzer. International Journal ofHydrogen Energy 2010;35:9550e60.

[4] Grigoriev SA, Porembskiy VI, Korobtsev SV, Fateev VN,Aupretre F, Millet P. High-pressure PEM water electrolysisand corresponding safety issues. International Journal ofHydrogen Energy 2011;36:2721e8.

[5] Siracusano S, Blasi AD, Baglio V, Brunaccini G, Briguglio N,Stassi A, et al. Optimization of components and assemblingin a PEM electrolyzer stack. International Journal ofHydrogen Energy 2011;36:3333e9.

[6] Millet P, Mbemba N, Grigoriev SA, Fateev VN, Aukauloo A,Etievant C. Electrochemical performances of PEM waterelectrolysis cells and perspectives. International Journal ofHydrogen Energy 2011;36:4135e42.

[7] Selamet OF, Becerikli F, Mata MD, Kaplana Y. Developmentand testing of a highly efficient proton exchange membrane(PEM) electrolyzer stack. International Journal of HydrogenEnergy 2011;36:11480e7.

[8] Su H, Bladergroen BJ, Linkov V, Pasupathi S, Ji S. Study ofcatalyst sprayed membrane under irradiation method toprepare high performance membrane electrode assembliesfor solid polymer electrolyte water electrolysis. InternationalJournal of Hydrogen Energy 2011;36:15081e8.

[9] Ito H, Maeda T, Nakanoa A, Hwang CM, Ishida M, Kato A,et al. Experimental study on porous current collectors of PEMelectrolyzers. International Journal of Hydrogen Energy 2012;37:7418e28.

[10] Narayanan SR, Chun W, Nakamura BJ, Marino S, Valdez TI.US Patent #6, 299, 744; 2001.

[11] Take T, Tsurutani K, Umeda M. Hydrogen production bymethanolewater solution electrolysis. Journal of PowerSources 2007;164:9e16.

[12] Sasikumar G, Muthumeenal A, Pethaiah SS, Nachiappan N,Balaji R. Aqueous methanol electrolysis using protonconducting membrane for hydrogen production.International Journal of Hydrogen Energy 2008;33:5905e10.

[13] Cloutier CR, Wilkinson DP. Electrolytic production ofhydrogen from aqueous acidic methanol solutions.International Journal of Hydrogen Energy 2010;35:3967e84.

i n t e r n a t i o n a l j o u rn a l o f h y d r o g e n en e r g y 3 8 ( 2 0 1 3 ) 7 3e8 1 81

[14] Pham Anh T, Shudo T. Efficient hydrogen production fromaqueous methanol in a proton exchange membraneelectrolyzer with porous metal flow fields. InternationalJournal of Automotive Engineering, in press.

[15] Larminie J, Dicks A. Fuel cell systems explained. 2nd ed. JohnWiley & Sons, Ltd, ISBN 0-470-84857-X; 2003.

[16] Wasmus S, Kuver A. Methanol oxidation and directmethanol fuel cells: a selective review. Journal ofElectroanalytical Chemistry 1999;461:14e31.

[17] Liu H, Song C, Zhang L, Zhang J, Wang H, Wilkinson DP. Areview of anode catalysis in the direct methanol fuel cell.Journal of Power Sources 2006;155:95e110.

[18] Zhao TS, Yang WW, Chen R, Wu QX. Towards operatingdirect methanol fuel cells with highly concentrated fuel.Journal of Power Sources 2010;195:3451e62.

[19] Shudo T, Suzuki K. Performance improvement in directmethanol fuel cell using a highly porous corrosion-resistingstainless steel flow field. International Journal of HydrogenEnergy 2008;33(11):2850e6.

[20] Shudo T. Performance improvement in direct methanol fuelcell using a sintered spherical metal powder porous flowfield. Fuel Cell 2009;8(4):61e5.

[21] Faghri, Li X, Bahrami H. Recent advances in passive andsemi-passive direct methanol fuel cells. InternationalJournal of Thermal Sciences 2012;62:12e8.

[22] Scott K, Taama WM, Argyropoulos P. Material aspects of theliquid feed direct methanol fuel cell. Journal of AppliedElectrochemistry 1998;28:1389e97.

[23] Oedegaard A, Hebling C, Schmitz A, Moller-Holst S, Tunold R.Influence of diffusion layer properties in a low temperatureDMFC. Journal of Power Sources 2004;127:187e96.

[24] Krishnamurthy B, Deepalochani S. Effect of PTFE content onthe performance of a direct methanol fuel cell. InternationalJournal of Hydrogen Energy 2009;34:446e52.

[25] Gogel V, Frey T, Yongseng Z, Friedrich KA, Jorissen L, Garche J.Performanceandmethanolpermeation indirectmethanol fuelcells: dependence on operating conditions and on electrodestructure. Journal of Power Sources 2004;127:172e80.

[26] Xu C, Zhao TS, Ye Q. Effect of anode backing layer on the cellperformance of a direct methanol fuel cell. Journal ofElectrochimica Acta 2006;51:5524e31.

[27] Gasteiger HA, Markovic N, Ross PN, Cairns EJ. Temperaturedependent methanol electro-oxidation on well-characterized PteRu alloys. Journal of The ElectrochemicalSociety 1994;141:1795e803.

[28] Kauranen PS, Skou E, Munk J. Kinetics of methanol oxidationon carbon-supported Pt and Pt þ Ru catalysts. Journal ofElectroanalytical Chemistry 1996;404:1e13.

[29] Iwasita T. Electrocatalysis of methanol oxidation.Electrochimica Acta 2002;47:3663e74.

[30] Dillon R, Srinivasan S, Arico AS, Antonucci V. Internationalactivities in DMFC R&D: status of technologies and potentialapplications. Journal of Power Sources 2004;127:112e26.

[31] Petrii OA. PteRu electrocatalysts for fuel cells:a representative review. Journal of Solid StateElectrochemistry 2008;12(5):609e42.

[32] Ren X, Becerra JJ, Hirsch RS, Gottesfeld S. Directoxidation fuel cell operating with direct feed ofconcentrated fuel under passive water management. USPatent 7407721 B2.

[33] Muller JT, Urban PM, Holderich WF. Impedance studies ondirect methanol fuel cell anodes. Journal of Power Sources1999;84:157e60.

[34] Chakraborty D, Chorkendorff I, Johannessen T.Electrochemical impedance spectroscopy study of methanoloxidation on nanoparticulate PtRu direct methanol fuel cellanodes: kinetics and performance evaluation. Journal ofPower Sources 2006;162:1010e22.

[35] Lai CM, Lin JC, Hsueh KL, Hwang CP, Tsay KC, Tsai LD, et al.On the electrochemical impedance spectroscopy of directmethanol fuel cell. International Journal of Hydrogen Energy2007;32:4381e8.

[36] Vazquez G, Alvarez E, Navaza JM. Surface tension ofalcohol þ water from 20 to 50C. Journal of ChemicalEngineering Data 1995;40:611e4.

[37] Zawodzinski Jr TA, Springer TE, Davey J, Jestel R, Lopez C,Valerio J, et al. A comparative study of water uptake by andtransport through lonomeric fuel cell membranes. Journal ofThe Electrochemical Society 1993;140:1981e5.

[38] Mukundan R, Kim YS, Garzon F, Pivovar B. Freeze/thaweffects in PEM fuel cells. ECS Transactions 2006;1(8):403e13.