ww.sciencedirect.com

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 7 ( 2 0 1 2 ) 1 4 5 3 6e1 4 5 4 2

Available online at w

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

Comparison of different types of membrane in alkaline directethanol fuel cells

L. An, T.S. Zhao*, Q.X. Wu, L. Zeng

Department of Mechanical Engineering, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China

a r t i c l e i n f o

Article history:

Received 16 May 2012

Received in revised form

17 June 2012

Accepted 27 June 2012

Available online 26 July 2012

Keywords:

Fuel cell

Alkaline direct ethanol fuel cell

Ion exchange membrane

Ionic conductivity

* Corresponding author. Tel.: þ852 2358 8647E-mail address: metzhao@ust.hk (T.S. Zh

0360-3199/$ e see front matter Copyright ªhttp://dx.doi.org/10.1016/j.ijhydene.2012.06.1

a b s t r a c t

It has been understood that the use of cation-exchange membranes (CEM) and alkali-doped

polybenizimidazole membranes (APM) in alkaline direct ethanol fuel cells (DEFC) with an

added base in the fuel exhibits performance similar to the use of anion-exchange

membranes (AEM). The present work is to assess the suitability of the three types of

membrane to alkaline DEFCs by measuring and comparing the membrane properties

including the ionic conductivity, the species permeability, as well as the thermal and

mechanical properties. The comparison shows that: (i) the AEM is still the most promising

membrane for the alkaline DEFC, although the thermal stability needs to be further

enhanced; (ii) before solving the problem of the poor thermal stability of AEMs, the CEM is

another choice for the alkaline DEFC running at high temperatures (<90 �C); and (iii) the

APM can also be applied to the alkaline DEFC operating at high temperatures, but its

mechanical property needs to be substantially enhanced and the species permeability

needs to be dramatically decreased.

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

reserved.

1. Introduction conductivity of state-of-the-art AEMs and ionomers. For this

Alkaline direct ethanol fuel cells (DEFC) have received ever-

increasing attention, mainly due to the increased perfor-

mance as a result of the fast electrochemical kinetics at both

the anode and cathode in alkalinemedia [1]. The conventional

architecture design of alkaline DEFC systems is similar to that

of proton exchange membrane (PEM) fuel cells, in which the

ion transport pathways between the anode and cathode are

formed by the network of dispersed ionomers in the elec-

trodes that are interfaced with the membrane. However, such

a fuel cell system that purely relies on an anion-exchange

membrane (AEM) to conduct ions through the membrane

and ionomers in the electrodes exhibited extremely low cell

performance (the state-of-the-art peak power density is

1.6 mW cm�2 at 60 �C) [2], primarily because of the low

.ao).2012, Hydrogen Energy P05

reason, previous efforts with respect to the development of

AEMeDEFCs have mainly been made to the development of

high-conductivity AEMs and ionomers, as well as highly-

catalytic materials for both the ethanol oxidation reaction

(EOR) and the oxygen reduction reaction (ORR) [3,4]. On the

other hand, however, it has recently been demonstrated

that, even using existing ion conductors and catalysts, an

addition of an alkali (e.g.: NaOH and KOH) to ethanol would

enable AEMeDEFCs to yield extremely high performance

(the state-of-the-art peak power density can be as high as

185 mW cm�2 at 60 �C) [5]. Such a drastic improvement in

performance is basically attributed to the added base, which

not only enables the kinetics of the EOR to be speeded up [6],

but also allows a drastic increase in the ionic transport capa-

bility [7,8]. However, a barrier that limits the cell performance

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

i n t e r n 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 7 ( 2 0 1 2 ) 1 4 5 3 6e1 4 5 4 2 14537

of AEMeDEFCs is that state-of-the-art AEMs do not allow the

fuel cell to operate at high temperatures [9]. For this reason,

the use of alkali-doped polybenizimidazolemembranes (APM)

that can withstand a high operating temperature (typically

90 �C) in alkaline DEFC systems has been attempted [10],

resulting in a so-called APMeDEFC. More recently, for the

reason that cations (Naþ or Kþ) are present in the fuel cell

system, cation-exchange membranes (CEM) have been

applied in the alkaline DEFC system to form another so-called

CEMeDEFC [11]. Previous studies [11e13] of the three types of

alkaline DEFC have demonstrated that there was no

substantial difference in the performance between the

AEMeDEFC, the CEMeDEFC and the APMeDEFC at low oper-

ating temperatures (<60 �C), but the latter two could operate

stably at high operating temperatures (typically 90 �C). Other

than that, no systematic comparisonwith respect to the use of

the three membranes in the fuel cell system with an added

base in ethanol has been made, as shown in Fig. 1. For this

reason, the questions of what type of membrane is more

suitable to the alkaline DEFC with an added base and the pros

and cons of the three types of membrane remain unclear.

Hence, the objective of this work was to make a systematic

comparison of the AEM, APM and CEM in terms of their ionic

conductivity, species permeability, as well as thermal and

mechanical properties.

2. Experimental

2.1. Pre-treatment of the membranes

The AEM (A201), with a thickness of 28 mm, was provided by

Tokuyama; the CEM was a Nafion 211 membrane (thickness:

25 mm), which was treated as a cation conductor; the PBI

membrane (thickness: 30 mm) was provided by Yick-Vic. The

AEM and CEM are immersed in 1.0 M NaOH for 24 h to convert

two membranes to OH� form and Naþ form, respectively [14].

Product

e- O2

Anode Cathode

EtOH + Alkali e-

-2

-3

-23 4eO4HCOOCH5OHOH +

++

+ + +

++→

→

→CHCH:Anode--

22 4OH4eO2HO:Cathode

OH2COOCHOHOOHCHCH:Overall 2-

3-

223

M

OH-

Na+

Fig. 1 e Schematic of the alkaline DEFC system.

After that, both themembranes werewashed in deionized (DI)

water for several times to remove the free alkali remained in

the membranes, and kept in DI water before use [15,16]. It

should be noted that the conductivity of the pure PBI

membranes is extremely low, i.e.: 1 � 10�12 S cm�1 [17]. For

this reason, the PBI membranes were pre-treated by

immersing them in the 6.0 M NaOH solution for 7 days to form

the APMs. After that, the APMs were washed in DI water for

several times to remove the free alkali remained in the

membrane matrix, and then kept in DI water before use.

2.2. Determination of the liquid uptake

The liquid uptake of the membrane, 4, is determined by:

4 ¼ mw �md

md� 100% (1)

withmw andmd denoting themass of the hydratedmembrane

and the mass of the dry membrane, respectively [18]. To

determine md, a membrane sample was dried for 3 h at 80 �C,

and was then weighed with a sealed weighing balance. To

determine mw, the dried membrane sample was immersed in

the 5.0MNaOH solution at room temperature for several days.

Prior to weighing the sample, the surface of the sample was

cleaned by delicate wipers (Kimwipes, Canada).

2.3. Measurement of the ionic conductivity

To measure the ionic conductivities of the AEM, CEM and

APM, the pre-treatedmembraneswere immersed in the NaOH

solutions with different concentrations (from 0 M to 10 M) for

24 h. Prior to measurements, each sample was kept at room

temperature for at least 10 min. Each sample, with an area of

1.0 cm � 1.0 cm, sandwiched between a pair of gold-coated

stainless steel plates, was measured by the electrochemical

impedance spectra (EIS) at the frequency range from 100 kHz

to 1 Hz with 5 mV amplitude [19]. Each sample was measured

at least three times, and an average value was then obtained.

The ionic conductivity was determined by:

s ¼ dR$S

(2)

where s is the ionic conductivity (S cm�1), d is the thickness of

the membrane (cm), R is the membrane resistance obtained

from the real axis intercept of the impedance Nyquist plot (U),

and S is the effective area of the membrane (cm2).

2.4. Determination of the species permeability

Species permeability measurements were made using

a home-made diffusion cell described elsewhere [20], which

consisted of two cylindrical compartments separated by

a membrane with an effective area of 2.0 cm2. One compart-

ment, termed as Compartment A, was filled with 25mL (VA) of

the aqueous solution (ethanol or NaOH), and the other

compartment, termed as Compartment B, was filled with the

same volume (VB ¼ 25 mL) of DI water. The selected

membrane was clamped between two compartments, and the

solutions in the both compartments were kept under the

stirring conditions during the experiment to avoid any

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 7 ( 2 0 1 2 ) 1 4 5 3 6e1 4 5 4 214538

concentration build-ups. To determine the ethanol concen-

tration in Compartment B, each sample (1 mL) was taken every

20min fromCompartment B and the ethanol concentration in

each samplewas determined by the gas chromatography (GC).

Therefore, the ethanol concentration profile (CB) was deter-

mined over a period of time. The ethanol permeability was

calculated by [21]:

P ¼ a� VB � d

A� CA(3)

where P (cm2 s�1) is the species permeability (¼D� KwhereD is

the species diffusivity and K is the partition coefficient), a is the

slope of CB versus time plot, d (cm) is the membrane thickness,

and A (cm2) is the area of the membrane exposed to the solu-

tions. As for determining the NaOH permeability, the NaOH

concentration (CB) in each sample was quantified by the Met-

rohm� 883 basic ion chromatography (IC). Therefore, the NaOH

permeability can also be calculated according to Eq. (3).

3. Results and discussion

3.1. Ionic conductivities

3.1.1. Effect of the NaOH concentrationFig. 2 shows the effect of NaOH concentrations varying from

0 M to 10 M on the ionic conductivities of the three types of

membrane. It can be seen that with increasing the NaOH

concentration, the ionic conductivities of the three

membranes increases first, and then decreases resulting in

a peak ionic conductivity for each membrane. In DI water, the

ionic conductivity of the AEM is about 5.3 mS cm�1, which is

close to that of the CEM (6.2 mS cm�1) but much higher than

that of the APM (0.006 mS cm�1). The reason why both the

AEM and CEM have a higher conductivity in DI water is

because they both have functional groups, i.e.: quaternary

ammonium (QA) and sulfonic acid (SA), which transport

anions and cations, respectively. On the other hand, as the

APM has no functional groups, its ionic conductivity in DI

water is almost zero. This fact indicates that once the free

0 2 4 6 8 10

0

5

10

15

20

25

30

35

40

AEM CEM APM

Ioni

c C

ondu

ctiv

ity,m

S cm

NaOH Concentration, M

Room temperature

Fig. 2 e Effect of the NaOH concentration on the ionic

conductivity.

alkali in the matrix of the APM is washed away by DI water,

the APM will not conduct ions. Therefore, transporting ions

through the APM is achieved by the free alkali in thematrix. In

summary, the ionic conductivity of a membrane in DI water

depends on its functional groups.

Fig. 2 also indicates that the respective peak ionic

conductivity of the AEM, the CEM, and the APM occurs,

respectively, at 5.0 M, 3.0 M, and 7.0 M when the membranes

were immersed in alkaline solutions. The reason why there

exists the peak conductivity is explained as follows. Generally,

ions through the membrane in the alkaline DEFC system are

conducted by both functional groups and alkali-doped free

volumes [22]. Once a membrane is soaked in an alkaline

solution, the membrane state is changed: alkali-doped free

volumes are formed, which can conduct ions under an electric

field in a fashion similar to an aqueous solution [10]. However,

it should be noted that transporting ions by alkali-doped free

volumes has a much lower transport resistance than by

functional groups [23]. For this reason, the ionic conductivity

of the AEM in the 1.0 M NaOH (21.7 mS cm�1) solution is more

than 4 times higher than that in DI water (5.3 mS cm�1), sug-

gesting that the ionic conduction in the AEM is predominated

by alkali-doped free volumes. The existence of the peak ionic

conductivity is attributed to the fact thatw5.0 M is the critical

concentration resulting in the peak ionic conductivity in

NaOH solutions [24].With increasing the NaOH concentration,

the NaOH concentration in the alkali-doped free volumes also

increases, thus speeding up the ionic transport rate through

the AEM. On the other hand, the viscosity of the alkaline

solution increases with increasing the alkaline concentration

[25], increasing the ionic transport resistance and thus

lowering the ionic velocity in the alkaline solution. According

to the theory of electrophoresis, a smaller ionic velocity

results in a smaller ionic mobility, lowering the ionic

conductivity under the same electric field. Hence, when the

NaOH concentration exceeds 5.0 M, the ionic conductivity of

the AEM starts to decrease.

As for the CEM, Fig. 2 shows that the ionic conductivities in

the 1.0 M and 3.0 M NaOH solutions are 8.3 mS cm�1 and

10.7 mS cm�1, respectively, which are higher than that in DI

water (6.2 mS cm�1). The increased ionic conductivity is

attributed to the formation of alkali-doped free volumes in

alkaline solutions. However, when the NaOH concentration

exceeds 3.0 M, the increased viscosity of the alkaline solution

will lower the ionic mobility, thus decreasing the ionic

conductivity. Therefore, 3.0 M is the optimal NaOH concen-

tration for the CEM, which is consistent with the previous

work [26].

With respect to the APM, the ionic conductivity in the 1.0 M

NaOH (3.6 mS cm�1) is much higher than that in DI water

(0.006 mS cm�1), suggesting that the contribution to the ionic

conductivity is completely attributed to alkali-doped free

volumes. However, when the NaOH concentration exceeds

7.0 M, the increased viscosity of the alkaline solution will

reduce the ionic mobility, thus lowering the ionic conduc-

tivity. Therefore, 7.0 M is the optimal NaOH concentration for

the APM yielding the peak ionic conductivity, which is

consistent with the previous work [27]. In summary, the

competition between the favorable effect of increasing the

ionic concentration and the adverse effect of the increased

i n t e r n 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 7 ( 2 0 1 2 ) 1 4 5 3 6e1 4 5 4 2 14539

viscosity in the alkali-doped free volumes of the membrane

results in an optimal NaOH concentration that gives the

highest ionic conductivity for each type of membrane. This

finding is similar to that reported elsewhere [28].

The comparison in the peak ionic conductivity of the three

membranes shows that the AEM has much higher conduc-

tivity (30.1 mS cm�1) than that of both the CEM and APM do

(10.7 mS cm�1 and 17.2 mS cm�1), because both functional

groups and alkali-doped free volumes can conduct anions [22].

Hence, in terms of the ionic conductivity, the AEM is the best

membrane among the three membranes.

3.1.2. Effect of the temperatureFig. 3(a) shows the effect of temperature on the ionic

conductivities of the three types of membrane in the 5.0 M

NaOH solution. It can be seen that the ionic conductivities of

the three membranes increase almost linearly with temper-

ature. In particular, when the temperature increases from

23 �C to 60 �C, the ionic conductivities of the AEM, CEM

and APM increase from 30.1 mS cm�1, 5.6 mS cm�1 and

15.7 mS cm�1 to 54.9 mS cm�1, 11.4 mS cm�1 and

33.7 mS cm�1, respectively. An increase in temperature can

accelerate the molecular motions, thereby enhancing the

ionic transport of the functional groups and increasing the

ionic conductivity [29]. On the other hand, increasing

20 30 40 50 600

10

20

30

40

50

60a

Ioni

c C

ondu

ctiv

ity,m

S cm

Temperature, C

AEM CEM APM

5.0 M NaOH

2.9 3.0 3.1 3.2 3.3 3.4 3.5-5.5

-5.0

-4.5

-4.0

-3.5

-3.0

-2.5

E =15.84 kJ/mol

E =16.97 kJ/mol

AEM CEM APM

ln(Io

nic

cond

uctiv

ity, S

cm

)

1000/T, K

E =13.30 kJ/mol

b

Fig. 3 e Effect of the temperature on the ionic conductivity.

temperature can also increase the ionic mobility in the alkali-

doped free volumes of the membrane, resulting in the

increased ionic conductivity [27]. Therefore, an increase in

temperature not only enhances the ionic transport capability

of the functional groups, but also increases the ionic mobility

in the alkali-doped free volumes. Fig. 3(b) shows the temper-

ature dependence of the ionic conductivities of the three

membranes. It can be seen that the conductivities of the three

membranes show an approximate Arrhenius-type tempera-

ture dependence, and the respective apparent activation

energy is 13.30 kJmol�1 for the AEM, 15.84 kJmol�1 for the CEM

and 16.97 kJ mol�1 for the APM. The apparent activation

energy of the three membranes is comparable, implying that

the three membranes have the similar ionic conduction

mechanism [30], which is consistentwith our recentwork [22].

3.1.3. Effect of the soaking timeIt has been reported that long-time exposure of an AEM in

a concentrated alkaline mediummay cause decomposition of

the functional groups in the membrane, thus decreasing the

ionic conductivity [31]. For this reason, we tested the effect of

the soaking time on the conductivity of the three membranes

in the 5.0 M NaOH solution and the results are shown in Fig. 4.

It is seen that the ionic conductivity of the AEM remains

almost the same after being immersed in the 5.0 M NaOH

solution for 70 days. The reason why the soaking time has an

insignificant effect on the AEM conductivity is because the ion

conduction through themembrane is predominated by alkali-

doped free volumes. For the same reason, the soaking time

also has an insignificant effect on the conductivity of the other

two membranes.

3.2. Liquid uptake

Table 1 shows the liquid uptake of the threemembranes in the

5.0 M NaOH solution. It can be seen that the liquid uptakes of

the AEM and the APM are much higher than that of the CEM.

Generally, a high liquid uptake will lead to a higher ionic

conductivity of the membrane in low-temperature fuel cells.

However, too high liquid uptake may cause dimensional

changes and a serious decrease in the mechanical strength

[32]. Hence, the higher ionic conductivities of the AEM and

APM can be attributed to the higher liquid uptake of the both

membranes. On the other hand, as be discussed in Section 3.4,

the APM has the worst mechanical property, primarily

because of its highest liquid uptake.

3.3. Permeability

3.3.1. EthanolTable 2 shows the respective ethanol permeability through

the three membranes. It can be seen that the ethanol

permeability through the three membranes is similar

(1� 10�7 cm2 s�1) at room temperature. Generally, the ethanol

permeation through the membrane is mainly driven by the

ethanol concentration gradient. In direct methanol fuel cells

(DMFC), the methanol permeability of a Nafion membrane is

about 1 � 10�6 cm2 s�1, which is one order magnitude higher

than the ethanol permeability, mainly because the molecule

size of methanol is smaller than that of ethanol [33].

0 10 20 30 40 50 60 700

5

10

15

20

25

30

35

40

45

50

AEMCEMKPM

mcS

m,ytivitcudnoC

cinoI1-

Soaking Time, Days

Room Temperature5.0 M NaOH

Fig. 4 e Effect of the soaking time on the ionic conductivity.

Table 1 e Liquid uptake of the three membranes.

Electrolyte AEM CEM APM

5.0 M NaOH 54.68% 8.48% 58.12%

30

40

ress

,MPa

AEM CEM APM

5.0M NaOHRoom temperature

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 7 ( 2 0 1 2 ) 1 4 5 3 6e1 4 5 4 214540

Therefore, ethanol crossover in an alkaline DEFC is not as

serious as methanol crossover in DMFCs.

3.3.2. NaOHTable 2 also shows the respective NaOH permeability through

the three membranes. It can be seen that at room tempera-

ture, the NaOH permeability through the APM is as high as

19.83 � 10�8 cm2 s�1, but the NaOH permeability through the

both AEM and CEM is relatively lower, i.e.: 2.541� 10�8 cm2 s�1

and 3.778 � 10�8 cm2 s�1, respectively. The reason why the

NaOH permeability through the both AEM and CEM is rela-

tively lower is because they have functional groups, which can

hinder the ionic transport of cations and anions due to

the charge repulsion, respectively. It should be noted that

NaOH crossover will cover the active surface area in the

cathode of alkaline DEFCs, thereby lowering the cathode

performance [34].

3.4. Mechanical property

The stressestrain curves of the three types of membrane are

shown in Fig. 5, which were obtained at room temperature by

applying a tensile force at a uniform rate to the samples after

Table 2 e NaOH and ethanol permeability of the threemembranes.

PNaOH (10�8 cm2 s�1) PEtOH (10�7 cm2 s�1)

AEM 2.541 1.442

CEM 3.778 1.344

APM 19.83 1.826

soaked in the 5.0 M NaOH solution for 24 h. It can be seen that

the threemembranes show the different tensile strengths, i.e.:

40.2 MPa, 33.0 MPa, and 14.7 MPa for the AEM, CEM, and APM,

respectively, indicating that the membrane structure signifi-

cantly affects its mechanical property. The tensile strength of

the both AEM and CEM is much larger than the APM, sug-

gesting that the both of them have a better mechanical

property. However, previous studies showed that the pure PBI

membrane without any pre-treatment had the high tensile

strength [35]. This fact indicates a compromise between the

mechanical property and the ionic conductivity in the APM

[36]. For the pure PBI membrane, the hydrogen bonding

between eN] and eNHe groups is the dominant force

determining its mechanical strength [37]. After doping with

the alkali, however, the hydrogen bonds will be formed

betweeneN] and alkali, resulting in the decreasedmolecular

cohesion in the APM [37]. On the other hand, the presence of

the free alkali in the matrix of the APM would lead to the

reduced intermolecular forces as a result of the increased

separation distance. Therefore, the tensile strength of the

APM is much smaller than that of the pure PBI membrane.

3.5. Thermal property

Thermogravimetric analysis (TGA) curves of the three

membranes were recorded in nitrogen at a heating rate of

10 �C min�1. Prior to the measurements, the three samples

were placed in a constant-temperature oven (100 �C) to remove

free water for 12 h. It can be seen from Fig. 6 that among the

three types of membrane, the APM shows the best thermal

stability; hence it canbeapplied to relativelyhigh-temperature

fuel cells [38]. On the other hand, the AEM shows the worst

thermal stability, which is consistent with the result reported

elsewhere [3]. Generally, the decomposition process of the

membrane can be divided into three stages [23]: (i) the loss of

free and bondedwater; (ii) the decomposition of the functional

groups; and (iii) the splitting of the main chain of the

membrane. For the three membranes, at temperatures below

150 �C, the weight loss is mainly attributed to the loss of the

bonedwatermolecules. For theAEM, theQA functional groups

are gradually degraded from 150 �C to 200 �C, and then the

0 20 40 60 80 100 120 1400

10

20St

Strain,%

Fig. 5 e Stressestrain curves of the three membranes.

0 100 200 300 400 500 600 700 8000

10

20

30

40

50

60

70

80

90

100 AEM CEM APM

Wei

ght %

Temperature oC

Fig. 6 e TGA curves of the three membranes.

i n t e r n 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 7 ( 2 0 1 2 ) 1 4 5 3 6e1 4 5 4 2 14541

main chain starts and completes the degradation at 350 �C and

500 �C, respectively. For the CEM, there is no obvious degra-

dation of the SA functional groups in its TGA curve, but when

the heating temperature is over 440 �C, the CEM starts the

degradation and eventually completes the degradation at

around 550 �C. For the APM, it is stable until the temperature is

over 570 �C. The total weight loss of the APM at 800 �C is only

30%. Therefore, the APM has the best thermal stability and the

AEM has the worst thermal stability.

4. Concluding remarks

In this work, we compared three types of commercial

membrane, which have been applied in the alkaline DEFC

with an added base in the fuel, in terms of the ionic conduc-

tivity, the species permeability, as well as the mechanical and

thermal properties. It is experimentally found that (i) the AEM

shows the highest ionic conductivity and the best mechanical

property, but the worst thermal stability; (ii) the APM shows

the best thermal stability, but the worst mechanical property

and the highest species permeability; and (iii) the CEM shows

the lowest ionic conductivity, but acceptable thermal stability,

mechanical property, and species permeability. These results

suggest that (i) the AEM is still the most promising membrane

for the alkaline DEFC, although its thermal stability at high

temperatures needs to be further enhanced; (ii) before solving

the problem of the poor thermal stability of AEMs, the CEM

can be another choice for the alkaline DEFC running at high

temperatures (<90 �C); and (iii) the APM can also be applied to

the alkaline DEFC operating at high temperatures, but its

mechanical property needs to be substantially enhanced and

the species permeability needs to be dramatically decreased.

Acknowledgments

The work described in this paper was fully supported by

a grant from the Research Grants Council of the Hong Kong

Special Administrative Region, China (Project No. 623311).

r e f e r e n c e s

[1] Antolini E, Gonzalez ER. Alkaline direct alcohol fuel cells.J Power Sources 2010;195:3431e50.

[2] Li YS, Zhao TS, Liang ZX. Effect of polymer binders in anodecatalyst layer on performance of alkaline direct ethanol fuelcells. J Power Sources 2009;190:223e9.

[3] Varcoe JR, Slade RCT. Prospects for alkaline anion-exchangemembranes in low temperature fuel cells. Fuel Cells 2005;5:187e200.

[4] Bianchini C, Shen PK. Palladium-based electrocatalysts foralcohol oxidation in half cells and in direct alcohol fuel cells.Chem Rev 2009;109:4183e206.

[5] Xu JB, Zhao TS, Li YS, Yang WW. Synthesis andcharacterization of the Au-modified Pd cathode catalyst foralkaline direct ethanol fuel cells. Int J Hydrogen Energy 2010;35:9693e700.

[6] Li YS, Zhao TS, Liang ZX. Performance of alkaline electrolyte-membrane based direct ethanol fuel cells. J Power Sources2009;187:387e92.

[7] An L, Zhao TS, Shen SY, Wu QX, Chen R. Performance ofa direct ethylene glycol fuel cell with an anion-exchangemembrane. Int J Hydrogen Energy 2010;35:4329e35.

[8] An L, Zhao TS, Shen SY, Wu QX, Chen R. An alkaline directoxidation fuel cell with non-platinum catalysts capable ofconverting glucose to electricity at high power output.J Power Sources 2011;196:186e90.

[9] Couture G, Alaaeddine A, Boschet F, Ameduri B. Polymericmaterials as anion-exchange membranes for alkaline fuelcells. Prog Polym Sci 2011;36:1521e57.

[10] Hou H, Sun G, He R, Wu Z, Sun B. Alkali dopedpolybenzimidazole membrane for high performance alkalinedirect ethanol fuel cell. J Power Sources 2008;182:95e9.

[11] An L, Zhao TS. An alkaline direct ethanol fuel cell witha cation exchange membrane. Energy Environ Sci 2011;4:2213e7.

[12] Modestov AD, Tarasevich MR, Leykin AY, Filimonov VY. MEAfor alkaline direct ethanol fuel cell with alkali doped PBImembrane and non-platinum electrodes. J Power Sources2009;188:502e6.

[13] Shen SY, Zhao TS, Xu JB, Li YS. Synthesis of PdNi catalysts forthe oxidation of ethanol in alkaline direct ethanol fuel cells.J Power Sources 2010;195:1001e6.

[14] An L, Zhao TS, Chen R, Wu QX. A novel direct ethanol fuelcell with high power density. J Power Sources 2011;196:6219e22.

[15] An L, Zhao TS. Performance of an alkaline-acid directethanol fuel cell. Int J Hydrogen Energy 2011;36:9994e9.

[16] An L, Zhao TS. A bi-functional cathode structure for alkaline-acid direct ethanol fuel cells. Int J Hydrogen Energy 2011;36:13089e95.

[17] Kongstein OE, Berning T, Børresen B, Seland F, Tunold R.Polymer electrolyte fuel cells based on phosphoric aciddoped polybenzimidazole (PBI) membranes. Energy 2007;32:418e22.

[18] Yang CC. Synthesis and characterization of the cross-linkedPVA/TiO2 composite polymer membrane for alkaline DMFC.J Membr Sci 2007;288:51e60.

[19] Li YS, Zhao TS. Understanding the performance degradationof anion-exchange membrane direct ethanol fuel cells. Int JHydrogen Energy 2012;37:4413e21.

[20] Nasef MM, Zubir NA, Ismail AF, Dahlan KZM, Saidi H,Khayet M. Preparation of radiochemically pore-filledpolymer electrolyte membranes for direct methanol fuelcells. J Power Sources 2006;156:200e10.

[21] Nasef MM, Zubir NA, Ismail AF, Khayet M, Dahlan KZM,Saidi H, et al. PSSA pore-filled PVDF membranes by

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 7 ( 2 0 1 2 ) 1 4 5 3 6e1 4 5 4 214542

simultaneous electron beam irradiation: preparation andtransport characteristics of protons and methanol. J MembrSci 2006;268:96e108.

[22] An L, Zhao TS, Li YS, Wu QX. Charge carriers in alkalinedirect oxidation fuel cells. Energy Environ Sci 2012;5:7536e8.

[23] Wang ED, Zhao TS, Yang WW. Poly (vinyl alcohol)/3-(trimethylammonium) propyl-functionalized silica hybridmembranes for alkaline direct ethanol fuel cells. Int JHydrogen Energy 2010;35:2183e9.

[24] Bousfield WR, Lowry TM. The electrical conductivity andother properties of sodium hydroxide in aqueous solution aselucidating the mechanism of conduction. Philos Trans R SocLond A 1905;204:253e322.

[25] Sipos PM, Hefter G, May PM. Viscosities and densities ofhighly concentrated aqueous MOH solutions (Mþ ¼ Naþ, Kþ,Liþ, Csþ, (CH3)4N

þ) at 25.0 �C. J Chem Eng Data 2000;45:613e7.[26] Cheng H, Scott K. Influence of operation conditions on direct

borohydride fuel cell performance. J Power Sources 2006;160:407e12.

[27] Xing B, Savadogo O. Hydrogen/oxygen polymer electrolytemembrane fuel cells (PEMFCs) based on alkaline-dopedpolybenzimidazole (PBI). Electrochem Commun 2000;2:697e702.

[28] Agel E, Bouet J, Fauvarque JF. Characterization and use ofanionic membranes for alkaline fuel cells. J Power Sources2001;101:267e74.

[29] Zhou J, Unlu M, Vega JA, Kohl PA. Anionic polysulfoneionomers and membranes containing fluorenyl groups foranionic fuel cells. J Power Sources 2009;190:285e92.

[30] Tanaka M, Fukasawa K, Nishino E, Yamaguchi S, Yamada K,Tanaka H, et al. Anion conductive block poly(arylene ether)s:

synthesis, properties, and application in alkaline fuel cells.J Am Chem Soc 2011;133:10646e54.

[31] Vega JA, Chartier C, Mustain WE. Effect of hydroxide andcarbonate alkaline media on anion exchange membranes.J Power Sources 2010;195:7176e80.

[32] Xu S, Zhang G, Zhang Y, Zhao C, Zhang L, Li M, et al. Cross-linked hydroxide conductive membranes with side chainsfor direct methanol fuel cell applications. J Mater Chem 2012;22:13295e302.

[33] Song SQ, Zhou WJ, Liang ZX, Cai R, Sun GQ, Xin Q, et al. Theeffect of methanol and ethanol cross-over on theperformance of PtRu/C-based anode DAFCs. Appl Catal BEnviron 2005;55:65e72.

[34] Yu EH, Scott K, Reeve RW. Application of sodium conductingmembranes in direct methanol alkaline fuel cells. J ApplElectrochem 2006;36:25e32.

[35] Han M, Zhang G, Liu Z, Wang S, Li M, Zhu J, et al. Cross-linkedpolybenzimidazole with enhanced stability for hightemperature proton exchange membrane fuel cells. J MaterChem 2011;21:2187e93.

[36] Hou H, Sun G, He R, Sun B, Jin W, Liu H, et al. Alkali dopedpolybenzimidazole membrane for alkaline direct methanolfuel cell. Int J Hydrogen Energy 2008;33:7172e6.

[37] He R, Li Q, Bach A, Jensen JO, Bjerrum NJ.Physicochemical properties of phosphoric acid dopedpolybenzimidazole membranes for fuel cells. J Membr Sci2006;277:38e45.

[38] Li Q, He R, Jensen JO, Bjerrum NJ. Approaches and recentdevelopment of polymer electrolyte membranes for fuelcells operating above 100 �C. Chem Mater 2003;15:4896e915.