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 e n e r g y 3 3 ( 2 0 0 8 ) 6 9 1 3 – 6 9 2 0

Avai lab le a t www.sc iencedi rec t .com

j ourna l homepage : www.e lsev ier . com/ loca te /he

Concentration of HIx solution by electro-electrodialysis usingNafion 117 for thermochemical water-splitting IS process

M. Yoshidaa,b, N. Tanakaa,*, H. Okudaa, K. Onukia

aJapan Atomic Energy Agency, 4002 Narita-cho, Oarai-machi, Higashiibaraki-gun Ibaraki 311-1393, JapanbCHIYODA CORPORATION, 3-13 Moriya-cho, Kanagawa-ku, Yokohama-shi Kanagawa 221-0022, Japan

a r t i c l e i n f o

Article history:

Received 19 February 2008

Received in revised form

2 September 2008

Accepted 18 September 2008

Available online 5 November 2008

Keywords:

Electro-electrodialysis

HIx solution

Nafion

IS process

Thermochemical hydrogen

production

* Corresponding author. Tel.: þ81 29 266 732E-mail address: tanaka.nobuyuki61@jaea

0360-3199/$ – see front matter Crown Copyright ª 2

doi:10.1016/j.ijhydene.2008.09.035

a b s t r a c t

An experimental study of applying electro-electrodialysis (EED) for improved HI concen-

tration in the HIx solution, a mixture of HI–I2–H2O of approximately quasi-azeotropic

compositions has been carried out in the conditions of around 90 �C and using Nafion 117

and graphite electrodes. A range of 25–80% increase in initial current efficiency of HI

molality in catholyte is measured with the use of EED. In general, the efficiency increases

with increasing iodine molality and weight ratio of anolyte solution to catholyte solution.

The EED performance degrades in time. In some cases, the HI concentration limits are

observed. Electric conductivity of the HIx solution, overvoltage of electrode reaction, and

the membrane voltage drop is measured in a temperature range of 20–120 �C. It is found

that the EED cell voltage, which is an important cell performance parameter, is governed by

the membrane voltage drop.

Crown Copyright ª 2008 Published by Elsevier Ltd on behalf of International Association for

Hydrogen Energy. All rights reserved.

1. Introduction conventional fractional distillation to separating HI from the

Thermochemical hydrogen production processes have been

studied since 1970s for large-scale hydrogen production using

heat with temperature of as high as 1000 �C, which can be

supplied by nuclear energy from High Temperature Gas-

cooled Reactor or solar energy. Of many processes studied, the

iodine–sulfur process proposed by General Atomics [1] has

attracted much interest and intensive studies.

One of the technical challenges of making hydrogen

production efficient with the sulfur–iodine process remains to

be the successful development of an efficient scheme of

separating hydrogen iodide (HI) from HIx solution, a mixture of

HI–I2–H2O. The HIx solution, which is produced in the Bunsen

process reactor, is nearly quasi-azeotropic. Application of the

4; fax: þ81 29 266 7741..go.jp (N. Tanaka).008 Published by Elsevier Ltd o

solution is energy intensive due to the necessity of evaporating

the solvent water. Alternative distillation methods including

the extractive distillation using phosphoric acid [2] and the

reactive distillation in pressurized condition [3–5] have been

investigated. In addition, application of membrane technolo-

gies has also been considered such as the pervaporation [6–8]

and the electro-membrane process called electro-electrodial-

ysis (EED) [9–12].

The concept of the EED, on which this paper is focused,

separates HI from the HIx solution while avoiding evaporation

of water. It increases the HI molality in the solution to over the

level of quasi-azeotropic composition by combining the redox

reaction of iodine–iodide ion at electrodes and the selective

permeation of proton through a separation diaphragm (Fig. 1).

n behalf of International Association for Hydrogen Energy. All rights reserved.

Nomenclature

tþ transport number of proton, dimensionless

b ratio of permeated quantities of water to

proton, dimensionless

F Faraday constant, C/mol

DMcj molar change of j-component in catholyte, mol

I electric current, A

R resistance of membrane, U

t time, s

Dm molality increase, dimensionless

mcHI;0 initial HI molality in catholyte, mol/kg

mcHI HI molality in catholyte, mol/kg

Qe normalized quantity of electricity,

dimensionless

McHI;0 feed HI in catholyte, mol

hQ current efficiency, dimensionless

4 overvoltage of anode or cathode, V

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 e n e r g y 3 3 ( 2 0 0 8 ) 6 9 1 3 – 6 9 2 06914

The HI molality of quasi-azeotropic HIx solution is about

10 mol/kg, which varies with iodine molality and pressure [13–

15]. In spite of the already very high electrolyte concentration,

the feasibility of its further improvement was demonstrated

in beaker-scale experiments using a commercial styrene -

divinyl benzene type cation exchange membrane as the

separation diaphragm and glassy-carbon as the electrode

material [9]. Recently, Hong et al. [10–12], using Nafion 117 as

the separation diaphragm, reported the effects of operation

temperature and iodine molality on the membrane permse-

lectivity in the EED conditions.

Because of its exceptional thermal and chemical stability,

the perfluorinated cation exchange membrane of the

commercially available cation exchange membranes is

expected to be more applicable than other synthetic polymer

membranes to operate in harsh process conditions such as the

present EED condition. In the present study, further experi-

ments are carried out using Nafion to examine its permse-

lectivity under various EED conditions, which might well be

indicated by the current efficiency of the HI molality increase

in catholyte. Another important performance indicator of the

EED is the cell voltage. In the present study, in order to gain

technical insight required to optimize cell design and improve

component performance detailed breakdown of the overall

I-

I2I-

I2

H+

Catholyte(= HIx soln.)(= HIx soln.)

Anolyte

Electrode

Cation exchange membrane

Fig. 1 – Electro-electrodialysis of HIx solution.

cell voltage has been examined by measuring the contributing

factors in an EED cell made of Nafion and graphite electrodes.

2. Experiments

2.1. Test solutions and membrane

The 55–58 wt% hydriodic acid supplied by Kanto Chemical Co.,

INC., the 67 wt% hydriodic acid by Merck Ltd., and the 99.8wt%

iodine by Kanto Chemical Co., INC. were used as received to

prepare HIx solutions to approximately quasi-azeotropic

compositions with I2/HI molar ratios ranging from 1 to 4. The

quasi-azeotropic compositions of HIx solution were estimated

based on Engels et al. [16]. Nafion 117 was provided by E.I.du

Pont de Nemours & Company as the separation diaphragm.

2.2. Concentration by electro-electrodialysis

2.2.1. Experimental setupThe experimental setup for the EED experiments consists of

an EED cell, reservoirs for anolyte and catholyte solutions, two

pumps, and a DC power supply as shown in Fig. 1. A filter-

press type cell shown in Fig. 2 was used as the EED cell which

consists of two electrodes made of isotropic graphite provided

by TOYO TANSO Co., Ltd., two rubber gaskets by Viton� and

a separation diaphragm. The effective area of the membrane

is 80 cm2 and the gasket thickness is 2.0 mm. A Teflon mesh is

inserted in each compartment as a spacer between membrane

and electrode to support the membrane and also to avoid

stagnation. Rubber heaters are used to control the operation

temperature of the EED cell. Glass reservoirs for anolyte

solution and catholyte solution are connected with each

compartment of the cell with Teflon tubes. Peristaltic pumps

are used for circulation of the solutions. The glass reservoirs

are coated with tin oxide type transparent heaters and placed

on hot-plate type magnetic stirrers. Programmable DC Power

source supplied by TAKASAGO LTD. feeds direct current to the

cell in addition to monitoring the cell voltage.

2.2.2. Concentration experimentsThe concentration experiments were carried out under

atmospheric pressure and at constant current density.

Anolyte and catholyte solutions fed to the reservoirs were,

at first, heated while circulating between the cell and the

reservoirs for about 1 h. After the system reached steady state

of prescribed temperature, direct current was supplied with

galvanostatic mode. In the EED operation, the cell voltage was

continuously recorded and the sampling of the solutions was

carried out periodically to monitor the concentration change.

The current supply was terminated when irregular

phenomena occurred such as solid iodine precipitation in the

anolyte solution or boiling of the catholyte.

In all the concentration experiments, identical HIx solu-

tions were used as the feeds for anolyte and catholyte. The

test temperature was 90–100 �C. The applied current density

ranged from 100 to 200 mA/cm2. Composition of the test

solution was analyzed using an automatic potentiometric

titrator (COM-2500, Hiranuma Sangyo Co., Ltd., Japan). NaOH

solution (0.1 mol/d m�3) and Na2S2O3 solution (0.1 mol/d m�3)

(a) (b) (c) (d) (e) (d) (c) (b) (a)

anolyte catholyte

Fig. 2 – Schematic of EED cell (a) end plate; (b) electrode; (c) Teflon mesh; (d) gasket; (e) cation exchange membrane.

Table 1 – EED conditions for concentration experiments.a

Run Currentdensity

Feed for catholyte Feed for anolyte

Weight [HI] [I2] Weight [HI] [I2]

mA/cm2

g mol/kg

mol/kg

g mol/kg

mol/kg

A 100 470 12.1 25.0 470 12.1 25.0

B 100 470 11.7 12.1 468 11.7 12.1

C 100 444 12.5 51.2 443 12.5 51.1

D 200 471 11.9 24.4 469 11.8 24.3

E 100 254 12.4 25.0 999 12.4 25.0

a Separation diaphragm: Nafion 117 (effective surface area

80 cm2.), electrode: isotropic graphite, operation temperature:

95� 5 �C.

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 e n e r g y 3 3 ( 2 0 0 8 ) 6 9 1 3 – 6 9 2 0 6915

were used for measuring the concentration of Hþ and I2,

respectively.

2.2.3. Measurement of transport number of proton (tþ) andratio of permeated quantities of water to proton (b)In some concentration experiments, the weight change of

catholyte solution was measured to evaluate the molar

changes of HI and H2O, so as to evaluate the ‘‘ transport

number of proton (tþ)’’ and the ‘‘ratio of permeated quantities

of water (H2O) to proton (b)’’ defined as follows [9]:

tþ ¼FDMc

HI

It(1)

b ¼FDMc

H2O

Ittþ(2)

Here, the weight of concentrated catholyte was measured by

collecting the solution from the reservoir. The weight of

uncollected solution sticking inside the cell and tubes was

evaluated as the balance between the gross weight of the cell

and tubes in dry condition and that in wet condition after

operation.

2.3. Cell voltage of electro-electrodialysis

Three types of measurements were carried out to discuss the

breakdown of EED cell voltage.

2.3.1. Overvoltage of iodine–iodide redox reaction at graphiteelectrodeCurrent-potential relationship was measured of graphite

electrode in HIx solution in the temperature range of 20–

120 �C. From the difference of the electrode potential at

certain current density and the rest potential, the overvoltage

of iodine–iodide redox reaction at the graphite electrode was

evaluated. Solartron 1280B electrochemical test system was

used for the potential control. Hg/Hg2SO4 system (þ0.615 V vs.

SHE, 25 �C) was used as the reference electrode (RE-2C, BAS

Inc.) and Pt plate was used as the counter electrode.

2.3.2. Electric conductivity of HIx solutionElectric conductivity of HIx solution was measured in the

temperature range of 20 – 120 �C by electrochemical impedance

spectroscopy [17]. Frequency Response Analyzer of Solartron

1280B electrochemical test system was used with frequency of

10–20 kHz, and 0.01 mA amplitude. The calibration was per-

formed using 1 mol/kg KCl solution.

2.3.3. Voltage drop due to membraneUsing the experimental setup described in Section 2.2.1, the

EED cell voltage was measured at certain current density

using identical HIx solution as catholyte and anolyte with and

without inserting the membrane. The difference of the two

voltages corresponds to the sum of IR drop in membrane and

concentration polarization at the boundary of membrane,

which we call ‘‘membrane voltage drop’’ in this paper. The

measurements were carried out in the temperature range of

20 – 110 �C with 19 mm/s nominal flow velocity at the

membrane.

3. Results and discussion

3.1. Concentration by electro-electrodialysis

Table 1 lists the experimental conditions. Figs. 3–7 show the

observed changes in the HI and I2 molality in catholyte and

anolyte.

8

9

10

11

12

13

14

15H

I m

olal

ity

[mol

/kg-

H2O

]

Time [min]

10

15

20

25

30

35

40

45

50

0 10 20 30 40 50 60 70 80

0 10 20 30 40 50 60 70 80

I 2 m

olal

ity

[mol

/kg-

H2O

]

Time [min]

Fig. 3 – Evolution of HI and I2 molality in Run A.

B: catholyte, C: anolyte.

6

7

8

9

10

11

12

13

14

HI

mol

alit

y [m

ol/k

g-H

2O]

5

10

15

20

25

30

0 20 40 60 80 100 120

I 2 m

olal

ity

[mol

/kg-

H2O

]

Time [min]

0 20 40 60 80 100 120

Time [min]

Fig. 4 – Evolution of HI and I2 molality in Run B.

B: catholyte, C: anolyte.

8

9

10

11

12

13

14

15

16

0 5 10 15 20 25 30

HI

mol

alit

y [m

ol/k

g-H

2O]

Time [min]

30

35

40

45

50

55

60

65

70

0 5 10 15 20 25 30

I 2 mol

alit

y [m

ol/k

g-H

2O]

Time [min]

Fig. 5 – Evolution of HI and I2 molality in Run C.

B: catholyte, C: anolyte.

6

8

10

12

14

16

10 15 20 25 30 35 40

10

15

20

25

30

35

40

0 5

0 5 10 15 20 25 30 35 40

HI

mol

alit

y [m

ol/k

g-H

2O]

I 2 m

olal

ity

[mol

/kg-

H2O

]

Time [min]

Time [min]

Fig. 6 – Evolution of HI and I2 molality in Run D.

B: catholyte, C: anolyte.

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 e n e r g y 3 3 ( 2 0 0 8 ) 6 9 1 3 – 6 9 2 06916

10

11

12

13

14

15

16

0 5 10 15 20 25

20

21

22

23

24

25

26

27

28

HI

mol

alit

y [m

ol/k

g-H

2O]

I 2 m

olal

ity

[mol

/kg-

H2O

]

Time [min]

0 5 10 15 20 25

Time [min]

Fig. 7 – Evolution of HI and I2 molality in Run E.

B: catholyte, C: anolyte.

0

0.5

1

1.5

2

2.5

3

3.5

0.7 0.75 0.8 0.85 0.9 0.95 1

[-]

t+ [-]

Fig. 9 – tD and b. ,: I2/HI ratio [ 1 6: I2/HI ratio [ 2, see

Table 2 for the experimental conditions.

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 e n e r g y 3 3 ( 2 0 0 8 ) 6 9 1 3 – 6 9 2 0 6917

In Run A (Fig. 3), identical amounts of quasi-azeotropic HIx

solution with I2/HI ratio of ca. 2 were used as the feeds for

anolyte and catholyte. As seen in the figure, in the course of

the EED operation, the HI molality in catholyte increased to

a maximum value, i.e. the concentration limit, before

decreasing slightly. The EED operation was terminated due to

occurrence of iodine precipitation in anolyte. The HI molality

0

0.05

0.1

0.15

0.2

0.25

0.3

0 0.1 0.2 0.3 0.4 0.5 0.6

Mol

alit

y in

crea

se [

-]

Qe [-]

Fig. 8 – Current efficiency of HI molality increase in

catholyte. ,: Run A, >:Run B, B:Run C, :: Run D, A: Run E.

at the concentration limit was ca. 14.3 mol/kg that is 18%

higher than that in the feed.

Run B (Fig. 4) was carried out under the same conditions as

Run A except the I2 molality. The I2/HI ratio in this case is ca. 1.

The HI molality in catholyte had also showed a maximum

value before the EED terminated due to iodine precipitation.

However, the molality increase was ca. 12% at the maximum,

which is lower than that of Run A.

The same conditions as those of the above Runs were used

with the exception for the I2/HI ratio, which is ca. 4 in Run C

(Fig. 5). Similar to the observations made in Runs A and B, the

maximum value of HI molality in catholyte was reached prior

to the EED termination. The molality increase was ca. 15% in

Run C.

Here, it may be worth noting that, as seen from Figs. 3–5,

the difference of HI molality in catholyte and anolyte showed

a similar value of about 5 mol/kg, irrespective of the iodine

molality at the concentration limit. As for the occurrence of

maximum HI molality, it is supposed that this phenomenon

was caused by the increase of water permeation rate in

Table 2 – Experimental conditions for tD andb measurement.a

Run Catholyte:before

concentration

Catholyte:after

concentration

Quantity ofapplied

electricity

Molality[mol/kg]

Weight[g]

Molality[mol/kg]

Weight [g] [C]

1 HI¼ 11.6 199.06 HI¼ 12.5 204.58 12,320

I2¼ 11.8 I2¼ 8.6

2 HI¼ 12.5 254.93 HI¼ 13.8 256.06 8,300

I2¼ 24.7 I2¼ 21.2

a Current density: 100 mA/cm2, operation temperature: 95� 5 �C.

0

0.1

0.2

0.3

0.4

0.5

0.6

0 20 40 60 80 100 120

Cel

l vol

tage

[V

]

Time [min]

Fig. 10 – Evolution of EED cell voltage. ,: Run A, >:Run B.

0

0.005

0.01

0.015

0.02

0.025

0 20 40 60 80 100

Ove

rvol

tage

[V

]

Temperature [ºC]

: HI=10.8, I2=11.5

: HI=9.9, I2=17.3[mol/kg]

Fig. 12 – Overvoltage of graphite electrode in HIx solution.

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 e n e r g y 3 3 ( 2 0 0 8 ) 6 9 1 3 – 6 9 2 06918

accordance with the increase of the concentration difference

between anolyte and catholyte in the course of the EED

operation. This hypothesis can explain the change pattern of

I2 molality in the catholyte and anolyte, as well. In Run D

(Fig. 6), current density was increased to 200 mA/cm2. In this

case, the HI molality of catholyte increased without showing

maximum until occurrence of iodine precipitation in anolyte,

although the final molality increase was nearly the same with

the maximum molality increase observed in Run A.

In Run E (Fig. 7), the feed amounts of anolyte and catholyte

were changed from those in Run A, so that the feed amount of

anolyte was 4 times larger than that of catholyte. In this case,

HI molality of catholyte increased without showing maximum

until EED operation should have terminated due to boiling of

catholyte when the HI molality in catholyte reached to ca.

15.3 mol/kg.

c

0.284

0.286

0.288

0.29

0.292

0.294

0.296

-200 -150 -100 -50 0 50 100 150 200

Pot

enti

al [

V v

s SH

E]

Current density [mA/cm2]

a

Fig. 11 – Potential vs. current density at graphite electrode

in HIx solution. [HI] : 9.9 mol/kg, [I2]: 17.3 mol/kg,

temperature: 110 8C.

In order to evaluate the current efficiency of HI molality

increase in catholyte, relationship between the HI molality

increase in catholyte and the applied quantity of electricity

was examined for the results of Runs A–E. Fig. 8 shows the

relationships. Here, the molality increase (Dm) and the quan-

tity of electricity (Qe) were normalized as follows.

Dm ¼mc

HI �mcHI;0

mcHI;0

(3)

Qe ¼It=FMc

HI;0

(4)

where, mcHI, mc

HI;0, McHI;0, I, t and F denote HI molality in cath-

olyte [mol/kg], initial HI molality in catholyte [mol/kg], feed HI

in catholyte [mol], electric current [A], operation time [s], and

0

10

20

30

40

50

60

70

80

0 20 40 60 80 100 120

Con

duct

ivit

y [S

/m]

Temperature [ºC]

: HI=11.6, I2=11.7: HI=12.3, I2=24.7: HI=12.9, I2=51.3

[mol/kg]

Fig. 13 – Electric conductivity of HIx solution.

0

0.2

0.4

0.6

0.8

1

1.2

0 20 40 60 100 120

Mem

bran

e vo

ltag

e dr

op [

V]

Temperature [ºC]

: HI=11.7, I2=12.1: HI=12.6, I2=25.3[mol/kg]

80

Fig. 14 – Membrane voltage drop of Nafion 117 in HIx

solution at current density of 100 mA/cm2.

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 e n e r g y 3 3 ( 2 0 0 8 ) 6 9 1 3 – 6 9 2 0 6919

the Faraday constant [C/mol], respectively. Under this

normalization, the ratio Qe/Dm directly represents the current

efficiency of the aimed concentration. In Fig. 8, as for Runs A, B

and C, only the data before reaching concentration limits are

shown, which are of interest in the aimed concentration.

As seen in figure, the current efficiency in the initial stage

of Runs A and D exhibited similar values, the average of which

was ca. 50% in Qe range of 0–0.3, indicating that the current

density affected little on the initial current efficiency in the

tested range.

As for the effect of iodine, comparison of the current effi-

ciency of Run B, ca. 30%, with that of Run A clearly indicated

its positive effect. The presence of iodine favored the perm-

selectivity of Nafion for the present application in terms of

higher selectivity for proton permeation and/or lower

permeation of neutral species (H2O, HI). This observation is

consistent with that reported by Hong et al. [11] based on their

experiments in the range of I2/HI ratio 0.5–2. However, further

improvement of current efficiency could not be observed in

the higher I2/HI ratio of ca.4 (Run C), where the current effi-

ciency was ca. 50% as same as Run B.

In Run E, the current efficiency showed ca. 80%, the

highest value observed in the present study, in the Qe range of

0–0.3. Since the feed composition in Run E is the same with

Run A, the large difference in the current efficiencies of Runs

A and E was due to the smaller difference of the concentra-

tions in anolyte and catholyte in Run E, which suppressed

either increasing of the permeation of neutral species or

lowering of the selectivity of proton permeation in Nafion. In

any case, the results suggest that making up the process

scheme so as to realize a high amount ratio of anolyte to

catholyte may help the efficient concentration. As for the

boiling behavior observed in Run E, it can be avoided by using

pressure-resistant EED cells, by which achievement of higher

concentration is expected than that observed in the present

experiments.

3.1.1. Evaluation of tþ and b

Fig. 9 shows tþ and b. The experimental conditions are listed in

Table 2. As seen in the figure, the effect of the I2/HI ratio for tþwas not clear within the experimental error due to the

chemical analysis and the weight measurement. On the

contrary, b lowered clearly with increasing the I2/HI ratio,

indicating that the permeation of H2O was suppressed by the

presence of iodine.

3.2. Cell voltage of electro-electrodialysis

3.2.1. Evolution of the cell voltage in EED operationFig. 10 shows the evolution of cell voltages in Runs A and B

that were carried out with current density of 100 mA/cm2.

Both cell voltages gradually increased with the progress of

the EED operation. The cell voltage may be classified into

theoretical decomposition voltage, overvoltage of electrode

reaction, ohmic loss in solution, and the membrane voltage

drop. As for the theoretical decomposition voltage in the

present EED system, it is equivalent to the electromotive

force of a concentration cell with liquid junction, which is

governed by the concentration ratio of anolyte and catholyte

and also by the transport number of proton in the membrane

[18]. The theoretical decomposition voltage is zero when EED

starts with the feeds of identical composition for anolyte and

catholyte, and increases with the development of composi-

tion difference between catholyte and anolyte. In contrast,

the initial cell voltage is considered to consist of the over-

voltage, the ohmic loss in solution, and the membrane

voltage drop. Since the initial cell voltage accounted for most

of the EED cell voltage and, in principle, it can be reduced by

improving the performance of cell components, breakdown

of the initial cell voltage is discussed in the following

sections.

3.2.2. OvervoltageFig. 11 shows an example of current–potential relationship of

the graphite electrode in HIx solution from which overvoltage

of redox reaction of iodine–iodide ion was evaluated.

Temperature dependence of the overvoltage at current

density of 100 mA/cm2 is shown in Fig. 12 for two kinds of HIx

solutions with composition of close to quasi-azeotropic ones

with different I2 molality. The overvoltage exhibited lower

value in the solution of higher I2 molality, and decreased

steeply with the increase of temperature. It exhibited rela-

tively small values of lower than 0.005 V at and above 80 �C.

The value accounted for just a few percents of the initial cell

voltage observed in Runs A and B.

3.2.3. Ohmic loss due to solution resistanceElectric conductivities of three kinds of HIx solutions with

composition of close to quasi-azeotropic ones with different I2molality are shown in Fig. 13. In every case, the conductivity

increased almost linearly with temperature, and was in the

range of 50–65 S/m at temperature of 90–100 �C. Using these

data and the distance value between electrodes adopted in the

experiments, the ohmic loss due to solution resistance in

Runs A and B could be estimated to be in the range of 0.06–

0.08 V, which accounted for ca. 20–30% of the initial cell

voltage.

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 e n e r g y 3 3 ( 2 0 0 8 ) 6 9 1 3 – 6 9 2 06920

3.2.4. Membrane voltage dropFig. 14 shows temperature dependence of the membrane

voltage drop at current density of 100 mA/cm2. Larger voltage

drop was observed with higher iodine molality. The

voltage drop decreased with the increase of temperature, as

well. The voltage drop at 90–100 �C was in the range of 0.15–

0.3 V. These values are comparable with the initial cell voltage

observed in Runs A and B, indicating that majority of the

initial cell voltage was due to the membrane voltage drop.

4. Concluding remarks

The EED of HIx solution with the approximately quasi-azeo-

tropic compositions has been examined using Nafion 117 and

graphite electrode. Experiments were conducted under

different operation conditions of current density, amount

ratio of anolyte and catholyte and different iodine molality.

The results showed a wide variety of initial current efficiency

of enhancing HI molality in catholyte. In some cases, the

concentration limits were observed. Useful information has

been obtained for the practical process design: (1) The differ-

ence of HI molality in catholyte and anolyte shows a value of

about 5 mol/kg irrespective of the iodine molality when the HI

molality of catholyte reaches the concentration limit; (2) the

high ratio of anolyte to catholyte is effective and efficient to

increase the HI molality in catholyte.

Minimizing the cell voltage of the EED system is important

to the cell performance efficiency. The results of independent

measurements of the electric conductivity of the HIx solution,

the electrode overvoltage and the membrane voltage drop

revealed that the main contributor to the cell voltage is the

membrane voltage drop.

The EED performance, therefore, depends largely on the

membrane performance, especially the membrane resistivity.

Acknowledgment

The authors express their gratitude to Mr. K. Ikenoya for the

implementation of concentration experiments.

r e f e r e n c e s

[1] Russell Jr. JL, McCorkle KH, Norman JH, Porter JT, Roemer TS,Schuster JR, et al. Water splitting – a progress report. In: Proc.1st World Hydrogen Energy Conf., Miami-Beach, Florida,USA; March 1–3, 1976. 1A105–124.

[2] Norman JH, Besenbruch GE, O’Keefe DR. Thermochemicalwater-splitting for hydrogen production, GRI-80/0105; March1981.

[3] Roth M, Knoche KF. Thermochemical water splitting throughdirect HI-decomposition from H2O/HI/I2 solutions. IntJ Hydrogen Energy 1989;14:545–9.

[4] Berndhaeuser C, Knoche KF. Experimental investigations ofthermal HI decomposition from H2O–HI–I2 solutions. IntJ Hydrogen Energy 1994;19:239–44.

[5] Goldstein S, Borgard JM, Vitart X. Upper bound and bestestimate of the efficiency of the iodine sulphur cycle. IntJ Hydrogen Energy 2005;30:619–26.

[6] Stewart FF, Orme CJ, Jones MG. Membrane processes for thesulfur–iodine thermochemical cycle. Int J Hydrogen Energy2007;32:457–62.

[7] Le Duigou A, Bogard JM, Larousse B, Doizi D, Allen R,Bruce CE, et al. HYTHEC: an EC funded search for a long termmassive hydrogen production route using solar and nucleartechnologies. Int J Hydrogen Energy 2007;32:1516–29.

[8] Caputo G, Felici C, Tarquini P, Giaconia A, Sau S. Membranedistillation of HI/H2O and H2SO4/H2O mixtures for the sulfur–iodine thermochemical process. Int J Hydrogen Energy 2007;32:4736–43.

[9] Onuki K, Hwang GJ, Arifal, Shimizu S. Electro-electrodialysisof hydriodic acid in the presence of iodine at elevatedtemperature. J Membr Sci 2001;192:193–9.

[10] Hong SD, Kim CH, Kim JG, Lee SH, Bae KK, Hwang GJ. HIconcentration from HIx (HI–H2O–I2) solution for thethermochemical water-splitting IS process by electro-electrodialysis. J Ind Eng Chem 2006;12:566–70.

[11] Hong SD, Kim JK, Bae KK, Lee SH, Choi HS, Hwang GJ.Evaluation of the membrane properties with changing iodinemolar ratio in HIx (HI–I2–H2O mixture) solution toconcentrate HI by electro-electrodialysis. J Membr Sci 2007;291:106–10.

[12] Hong SD, Kim JK, Kim BK, Choi SI, Bae KK, Hwang GJ.Evaluation on the electro-electrodialysis to concentrate HIfrom HIx solution by using two types of the electrode. IntJ Hydrogen Energy 2007;32:2005–9.

[13] Neumann D. Phasengleichgewichte von HJ/H2O/J2-Loesungen, Diplomarbeit, RWTH Aachen; Mai 1987.

[14] Hodotsuka M, Yang X, Okuda H, Onuki K. Vapor–liquidequilibria for the Hi + H2O system and the HI + H2O + I2

system. J Chem Eng Data 2008;53:1683–7.[15] Liberatore R, Ceroli A, Lanchi M, Spadoni A, Tarquini P.

Experimental vapour–liquid equilibrium data of HI–H2O–I2mixtures for hydrogen production by sulphur–iodinethermochemical cycle. Int J Hydrogen Energy 2008;33:4283–90.

[16] Engels H, Knoche KF. Vapor pressure of the system HI/H2O/I2

and H2. Int J Hydrogen Energy 1986;11:703–7.[17] Sugimoto K. Fundamentals of theory and analysis of

electrochemical impedance spectroscopy. Zairyo-to-Kankyo1999;48:673–80 [in Japanese].

[18] Tamamushi R. DENKIKAGAKU(Electrochemistry). TokyoKagaku Dojin Pub; 1991 [in Japanese].