Integration of Conventional Electrodialysis and Electrodialysis With Bipolar

8/4/2019 Integration of Conventional Electrodialysis and Electrodialysis With Bipolar

http://slidepdf.com/reader/full/integration-of-conventional-electrodialysis-and-electrodialysis-with-bipolar 1/8

Journal of Membrane Science 365 (2010) 294–301

Contents lists available at ScienceDirect

Journal of Membrane Science

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / m e m s c i

Integration of conventional electrodialysis and electrodialysis with bipolar

membranes for production of organic acids

Yaoming Wang, Xu Zhang, Tongwen Xu ∗

Laboratory of Functional Membranes, School of Chemistry and Materials Science, University of Science and Technology of China,

Hefei, Anhui 230026, People’s Republic of China

a r t i c l e i n f o

Article history:

Received 22 July 2010Received in revised form 6 September 2010

Accepted 11 September 2010

Available online 17 September 2010

Keywords:

Conventional electrodialysis

Electrodialysis with bipolar membranes

Organic acids

Gluconic acid

Ion exchange membrane

a b s t r a c t

To reduce the cost of electrodialysis with bipolar membranes (EDBM) for production of organic acids,

conventional electrodialysis (CED) was integrated with EDBM, i.e., CED supplied concentrated organic

salts as the feed to EDBM. To assess the adaptability of this integration, sodium gluconate was chosen as

a model agent for production of gluconic acid. Results indicated that this integration could achieve an

apparent current efficiencyhigher than 100% andlow energyconsumptiondue to theinward concentra-

tion of gluconate and electrode reactions in CED. Besides, the process cost of CED–EDBM was estimated

to be $0.31kg−1, which was less than that of EDBM ($0.39kg−1). This integration not only made the

production cost-effective but also kept the operation of EDBM stable.

© 2010 Elsevier B.V. All rights reserved.

1. Introduction

To date, electrodialysis with bipolar membranes (EDBM) hasbeen accepted as an environmentally benign acidifying technique

for production of organic acids because it can split water into H+

and OH− inside bipolar membranes and supply H+ and OH− in situ

without introduction of other salts [1–4]. Though EDBM is of the

environmental benignity, it has not been industrialized widely due

to the high process cost. On one hand, the bipolar membrane is

costly ($1350 m−2 in China). On the other, a decrease in current

efficiency and a sharp increase in energy consumption often occur

at a high conversion rate of organic salts in EDBM due to the deple-

tion of salts in the feed compartment and the diffusion of organic

acids. When it comes to this shortcoming, it can be overcome to a

great extent by keeping the salt above a certain concentration in

the feed compartment. For this purpose, conventional electrodial-

ysis (CED) can be employed to concentrate organic salts and feedthe concentrate to EDBM. Such concept hasbeen applied to process

dairy products [5,6] and organic salts [2,7], and acquired positive

effects. Note that these integrations had CED and EDBM operate

separately, i.e., the dilute solution of salt was first concentrated to

a high level in CED and then fed into EDBM.

Different from the reports above, this research integrates CED

and EDBM for production of organic acids, and the synergy of CED

∗ Corresponding author. Tel.: +86 551 360 1587; fax: +86 551 360 1592.

E-mail address: [email protected] (T. Xu).

and EDBM will be the focus of this research. For experiments,

sodiumgluconate (NaGlu) is chosenas a model organic salt, andthe

effects of some parameters, such as thecurrent density of CED, elec-trolyte concentration, and stack configurations, on the production

of gluconic acid (HGlu) will be investigated.

2. Experimental

2.1. Materials

A cation-exchange membrane (Neosepta CMX, Tokuyama

Soda Inc., Japan), an anion-exchange membrane (Neosepta AMX,

Tokuyama Soda Inc., Japan), anda bipolar membrane (NeoseptaBP-

1, Tokuyama Soda Inc., Japan) were used forexperiments,and their

properties are listed in Table 1. The chemicals were of analytical

grade. Distilled water was used throughout.

2.2. Experimental set-up

As shown in Fig. 1, a laboratory-scale setup was used for exper-

iments. The CED and EDBM stacks were composed of one anode

and one cathode as well as membranes (BP, bipolar exchange

membrane; A, anion exchange membrane; C, cation exchange

membrane). For integration, CED is combined with an EDBM stack

of BP–C configuration (Fig. 1a CED–EDBM (BP–C)) and an EDBM

stack of BP–A configuration (Fig. 1b CED–EDBM (BP–A)), respec-

tively. For comparison with the CED–EDBM setup, an EDBM stack

of BP–C configuration(Fig. 1c EDBM (BP–C)) waschosensince BP–C

0376-7388/$ – see front matter © 2010 Elsevier B.V. All rights reserved.

doi:10.1016/j.memsci.2010.09.018

http://dx.doi.org/10.1016/j.memsci.2010.09.018


http://www.sciencedirect.com/science/journal/03767388

http://www.elsevier.com/locate/memsci

mailto:[email protected]



mailto:[email protected]

http://www.elsevier.com/locate/memsci

http://www.sciencedirect.com/science/journal/03767388




Y. Wang et al. / Journal of Membrane Science 365 (2010) 294–301 295

Table 1

Properties of the membranes used in CED and EDBM stack.

Membrane type Thickness (m) IEC (meq g−1) Area resistance (cm2 ) Voltage drop (V) Efficiency (%)

Neosepta BP-1 200–350 – – 1.2-2.2 >98

Neosepta AMX 120–180 1.4–1.7 2.0–3.5 – –

Neosepta CMX 220–260 1.5–1.8 2.0–3.5 – –

The data are collected from the product brochure provided by the company.

configuration is the mostfavorable stackconfiguration for econom-ical consideration [8]. Take the CED–EDBM (BP–C) setup (Fig. 1a)

for an example, there are 3 loops: the concentrate compartment

of CED and the acid compartment of EDBM were connected as

a loop; the other two loops were established between the dilute

compartments of CED and the electrode compartments of EDBM,

respectively. Each loop was equipped with an external 1000 cm3

beaker, allowing for continuous circulation by a submersible pump

(AP1000, Zhongshan Zhenghua Electronics Co. Ltd., China). The

effective membrane area was 7.07 cm2. The electrodes were made

of titanium coated with ruthenium. The electrodes were con-

nectedwith a directcurrent power supply(WYL1703×2, Hangzhou

Yuhang Siling Electronics Co. Ltd., China). Before the current was

applied, the solution of each compartment was circulated for half

an hour, and all the visible gas bubbles were eliminated. Na 2SO4

was added as electrode rinse in the EDBM stack.

Note that the feed in this study is not with pure gluconate ion,

but with electrolyte (Na2SO4) as well. The reason for this is the

low conductivity of gluconate ion and small effective membrane

area of the stack. As the membrane area is directly correlated to

the feed concentration. It is uneconomical to perform this small

scale (membrane area of 7.07cm2) CED–EDBM process with high

feed concentration (higher than 0.5 mol dm−3). If the CED–EDBM

stack in this study is conducted under pure condition (only with

gluconate ion in the feed), the voltage drop of the stack is too much

high (higher than 100 V) and the energy consumption is too much

high correspondingly. Moreover, a large amount of heat is gen-

erated under high voltage condition, which is detrimental to the

Fig. 1. Configurations of CED–EDBM stacks and a single EDBM stack: (a) CED–EDBM (BP–C); (b) CED–EDBM (BP–A); (C) EDBM (BP–C). BP, bipolar membrane; C, cation-

exchange membrane; A, anion-exchange membrane.



296 Y. Wang et al. / Journal of Membrane Science 365 (2010) 294–301

0 20 40 60 80 100 120

0.00

0.02

0.04

0.06

0.08

0 20 40 60 80 100 120

8

10

12

14

16

18

20

22

20mA·cm-2

30mA·cm-2

40mA·cm-2

50mA·cm-2

60mA·cm-2

V o

l t a g e

d r o p o

f C E D

( V )

Time (min)

a

0 20 40 60 80 100 120

12

13

14

15

16

17

18

V o

l t a g e d r o p of E DBM ( V )

b

A c i d c o n c e n t r a

t i o n

(

m o

l · d m

- 3)

Time ( min)

20mA·cm-2

30mA·cm-2

40mA·cm-2

50mA·cm-2

60mA·cm-2

20 30 40 50 60

1.5

2.0

2.5

3.0

3.5

4.0

apparent energy consumption

total energy consumption

c

Current density( mA·cm-2)

E n e r g y c o n s u m p t i o n

(

k W h · k g

- 1)

60

80

100

120

140

apparent current efficiency

absolute current efficiency

C ur r e n

t e f f i c i e n c y ( % )

Fig. 2. Effect of the current density of CED on the production of HGlu: (a) the voltage drops of CED and EDBM; (b) acid yield; (c) energy consumption and current efficiency.

Experimental conditions: CED–EDBM (BP–C); current density of EDBM, 50mA cm−2; electrolyte, 0.1mol dm−3; sodium gluconate, 0.05 mol dm−3; flow rate, 27 dm−3 h−1 .

membrane. Fortunately, this mixture feed does not affect any of

our above analysis, because sulfuric acid is a strong acid and glu-

conic acid is a weak one. Both thedissociation of sodiumsulfate and

sodium gluconate will finally contribute the production of gluconic

acid. Furthermore, the separation of gluconic acid from electrolyte

does not exist in large-scale production, as the feed concentration

can be higher than 2mol dm−3. Control experiments with sodium

sulfate as the only feed solution were also carried out to evaluate

the feasibility of CED–EDBM process.

2.3. Determination of gluconic acid concentration

TheHGlu concentration wasdetermined by titrationwith NaOH

using phenolphthalein (pH 8.0–9.8) as indicator.

2.4. Calculation of current efficiency and energy consumption

Considering the comparison between integrated and separate

stacks, an apparent current efficiency Áapp and absolute current




0 20 40 60 80 100 120

0.00

0.02

0.04

0.06

0.08

0 20 40 60 80 100 1205

10

15

20

25

30

0.05mol·dm-3

0.1mol·dm-3

0.2mol·dm-3

0.3mol·dm-3

0.4mol·dm-3

V o

l t a g e

d r o p o

f C E D s t a c k ( V )

Time (min)

a

0 20 40 60 80 100 1205

10

15

20

25

30

V ol

t a g e d r o p of E DBM s t a c k ( V )

Time (min)

b c

A c i d c o n c e n

t r a t i o n

(

m o

l · d m

- 3)

Time (min)

0.05mol·dm-3

0.10mol·dm-3

0.20mol·dm-3

0.30mol·dm-3

0.40mol·dm-3

0.0 0.1 0.2 0.3 0.4

1

2

3

4

5

6



Electrolyte concentration ( mol·dm-3

)


(

k W h · k g

- 1)

60

80

100

120

140



C ur r e n t e f f i c i e n c y ( % )

Fig. 3. Effect of electrolyte concentration on the integration stack performance: (a) the voltage drop across the CED and EDBM stack; (b) acid yield; (c) energy consumption

andcurrent efficiency. Experiment conditions: stack configuration,CED–EDBM (BP–C);currentdensity of CEDand EDBMstack,50 mAcm−2; sodium gluconate concentration

0.05moldm−3; flow rate, 27 dm−3 h−1.

efficiency Áabs was defined as Eq. (1), which is similar to that of

a common EDBM process [9]:

Áapp =(C t −C 0) zBF NI EDBM t

Áabs =(C t −C 0) zBF

N (I CED

+I EDBM

)t

(1)

where C t and C 0 (mol dm−3) arethe HGlu concentrations at time

t and 0, respectively; z isthe absolutevalence ( z = 1 forgluconicion);

B (dm3) isthe volumeof solutionin each loop; I CED and I EDBM (A)are

the currents of CED and EDBM, respectively; F is the Faraday con-

stant (96,485C mol−1); N is the number of repeating units (N =1).

In this work, t (s) is equal to 7200 s, and the change of fluid volume

in each loop was negligible, that is, B =0.5dm3.




20 30 40 50 602

3

4

5

6a





Current density ( mA·cm-2)

E n e r g y c o n s u

m p

t i o n

(

k W h ·

k g

- 1)

40

60

80

100


0.1 0.2 0.3 0.4 0.52

4

6

8

10

12





Sodium sulfate concentration ( mol·dm-3)

E n e r g y c o n s u m p

t i o n

(

k W h ·

k g

- 1)

b

20

40

60

80

100

C ur r e n

t e f f i c i e n c y ( % )

Fig. 4. Performance of pure feed salt (sodium sulfate) in CED–EDBM (BP–C) stack:

(a) Effect of current density on energy consumption and current efficiency; (b)

effect of feed concentration on energy consumption and current efficiency. Experi-

ment conditions: stack configuration, CED–EDBM (BP–C); current density of EDBM

stack, 50mAcm−2; flow rate, 27dm−3 h−1; sodium sulfate concentration in (a),

0.5moldm−3; current density of CED stack in (b), 50mA cm−2 .

The apparent energy consumption Eapp (kWhkg−1) and total

energy consumptionEtot wascalculatedbyextrapolatingtheresults

for the production of 1 kg of acid based on Eq. (2):

Eapp =

U EDBM I EDBMdt

C t BM

Etot =

(U CEDI CED+U EDBM I EDBM )dt

C t BM

(2)

where U CED and U EDBM (V) are the voltage drops across the CED and

EDBMstacks, respectively; I CED and I EDBM (A)arethecurrentsofCED

and EDBM, respectively; C t (mol dm−3) is the acid concentration at

time t (s); B (dm3) is the volume of acid loop; M is the molecular

weight of HGlu (196.14 g mol−1).

3. Results and discussions

3.1. Effect of the current density of CED on the production of

gluconic acid

Fig. 2 shows the effect of current density of CED on the produc-

tion of gluconic acid by using the CED–EDBM (BP–C) setup. As a

tradeoff between the efficiency of water splitting [10] and the life-

time of bipolar membranes [11], a current density of 50 mA cm−2

was chosen for the operation of EDBM. The concentrations of

Na2SO4 and NaGlu were 0.10 and 0.05 mol dm−3

, respectively.

Fig. 2a shows the effect of current density of CED on the voltage

drop of CED and EDBM stack. The results indicate that the volt-

age drops of both CED and EDBM stack decrease with the elapse of

time. Meantime, the higher the current density is, the higher the

voltage drop of CED stack will be. Normally, the voltage drop of

CED stack should increase with time, which is due to the depleting

of salt in the dilute compartment. The decrease of voltage drop

in the CED stack is attributed to the particular configuration of

CED–EDBM (BP–C) stack. For one thing, H+ generated from the

water dissociation of bipolar membrane is introducing to the CED

stack. The conductivity of gluconic acid is much higher than that of

sodium gluconate (the equivalent conductance of HGlu and NaGlu

is 380×10−4 and 80.5×10−4 s cm2 equiv−1, respectively [12]), the

electrical resistance of CED will decrease with an increase in the

concentration of HGlu. For another, H+ and OH− which is ascribe

to the electrode reactions are also introducing to the CED stack. In

theory, H+ and OH− can be neutralized with each other because the

anode and cathode compartments are circulated in a loop in the

CED stack. However, the mobility of H+ through cation exchange

membrane is much higher than OH− through the anion exchange

membrane, so the electrode reaction generated H+ and OH− can

be distributed in the acid and salt compartment separately. This

assumption has been confirmed from the high apparent current

efficiency of the EDBM stack in the following discussion. To sumup, H+ and OH− comes from water dissociation of bipolar mem-

brane and electrode reactions account for the decreasing of voltage

drop in the CED stack. In the case of EDBM stack, the concentrat-

ing of sodium gluconate in CED and water dissociation of bipolar

membrane leads to a sharply decrease of electrical resistance. The

voltage drops of EDBM among different CED current densities are

not pronounced, which indicates that the contribution of water

dissociation of bipolar membrane to the voltage drop is more sig-

nificant than that of concentrating of gluconate salt.

As shown in Fig. 2b, the acid concentration increases nearly

proportion to time elapses. In addition, the acid concentration

increases with the increase of current density of CED stack. Nor-

mally,theyieldof acid is stem from thewater dissociationof bipolar

membrane orthe water splittingof cationexchangemembrane dueto concentration polarization, both of which are directly related to

thecurrent density ofthe EDBM stack.But thecurrent density ofthe

EDBM stack have been fixed at 50mA cm−2, so the increase of acid

concentration must come from theCED stack.In theCED stack,both

electrode reactions and water splitting of mono-polar membranes

can possibly lead to the yield of acid. However, water splitting of

mono-polar membranes has low efficiency and the generated H+

and OH− recombines [13], so this effect is insignificant. Therefore,

theelectrodes reactionsof CEDstack arecontributeto theenhanced

increase of acid concentration as discussed in the above section.

Fig. 2c shows the effect of current density of CED on the current

efficiency and energy consumption of EDBM stack. The apparent

current efficiencies of EDBM stack increase from 98%to 131% when

current densitiesof CEDincreasefrom 20to 50mA cm−2

, indicatingextra H+ besides bipolar membrane dissociation is also introduced

to the EDBM stack.While the absolute current efficiencies decrease

from 70% to 59% in this time. The compete migration of H +/Na+

through cation exchange membrane and the molecular diffusion of

gluconic acid through the bipolar and cation exchange membrane

determinethat thecurrent efficiencyis lower than 100% at lowCED

current density. As current density increase, electrode generated

H+ increase correspondingly, and then the induction of H+ over-

whelm than that of migration out of the acid compartment. When

the current density of CED stack is in the range of 20–50mA cm−2,

the apparent energy consumption of EDBM stack is between 1.45

and2.00kWhkg−1. Even theenergy consumption ofCED andEDBM

stack are both considered, the total energy consumption is about

2.55–3.70 kWh kg−1

, which is similar to our experimental result of




a single EDBM process (1.55–4.72 kWh kg−1) [14]. But we should

take attention that not only gluconic acid but also concentrated

gluconate salt can be obtained with the similar energy consump-

tion in one step. Furthermore, the total energy consumption of the

CED–EDBM stack increases with the increase of current density of

CED stack, indicating a great part of electrical energy is consumed

to overcome theelectrical resistance of theCED stack as thecurrent

density increases. From the discussion above, an accompanying

CED stack is conducive to the performance of EDBM process, espe-

cially when CED and EDBM are both need to be carried out in some

circumstances.

3.2. Effect of electrolyte concentration on the production of

gluconic acid

Fig. 3a shows the effect of electrolyte concentration on the volt-

age drop of CED and EDBM stack when the current density of CED

and EDBM is 50mA cm−2 and sodium gluconate concentration is

0.05 moldm−3. Results indicate that the voltage drop of both CED

and EDBM stack decreases with an increase of electrolyte concen-

tration. The reason can be ascribed to common sense: the higher

the concentration of electrolyte, the lower the electrical resistance

of the membrane and solution. At the same time, the voltage dropsof CED stack increase in the first 10 min and then decrease with

time elapses in lowelectrolyte concentration. There aremainly two

counteracting effects that account forthis result. On onehand,after

the depletion of ions in the intermediate layer of bipolar mem-

brane, water dissociation takes place, and H+ and OH− begin to fill

the bipolar membrane and decrease the voltage drop across bipo-

lar membrane gradually. On the other hand, the concentration of

gluconate in the acid compartment is higher than that in the dilute

compartment of CED, which is due to the concentrating effect of

CED stack. In CED stack, Na+ and Glu− can migrate through cation

andanion exchange membrane,respectively,and accumulate in the

acid compartment. Meantime, Na+ in the acid compartment trans-

ports through the cation exchange membrane andinto the cathode

compartment of EDBM stack. The amount of Na+ migration out of the acid compartment is more than that of into the compartment,

which is the dominant effect at the beginning of experiment. In

contrast, the generated H+ of bipolar membrane accumulates with

time,the overwhelmingof thefirsteffect leadsto a sharply decrease

of voltage drop. While in the case of high electrolyte concentration,

the change of conductivity due to diluting effect of CED is not pro-

nounced, so an extreme point is not appeared in the voltage drop

of CED stack.

Fig. 3b shows the effect of electrolyte concentration on acid

concentration when the current density of CED and EDBM is

50mAcm−2 andsodium gluconateconcentration is 0.05 mol dm−3.

Similarly to effect of current density of CED stack in Fig. 2b, the

yield of acid increases as time elapses. There is a slightly increase

of acidconcentrationwith the increase of electrolyte concentration.Although the mobility of H+ is higher than Na+, Na+ will have more

predominance in quantity as Na2SO4 concentration increases, so

there is less loss ofH+ for gluconic production and thus an increase

in acid concentration. There have been confirmed from the signifi-

cant increase of current efficiencywith theincreasing of electrolyte

concentration.

Fig. 3c shows the effect of electrolyte concentration on the

current efficiency and energy consumption of EDBM stack when

the current density of CED and EDBM is 50 mA cm−2 and sodium

gluconate concentration is 0.05 moldm−3. The apparent and

total current efficiency of EDBM stack is respective between

110.3–140.3% and 55.2–71.2% when electrolyte concentration is in

the range of 0.05–0.4mol dm−3. The higher concentration of Na+,

the less lose ofH+

through cationexchange membrane of theEDBM

stack, which could be used to explain for the high apparent cur-

rent efficiency. The apparent and total energy consumption for the

production of gluconic acid decrease with the increase of elec-

trolyte concentration, which is in the range of 0.90-2.48 Wh kg−1

and 1.82–5.56 Wh kg−1, respectively. The low energy consumption

of EDBM stack, such as 0.90 Wh kg−1 when electrolyte concentra-

tion is 0.4moldm−3, suggesting a synergistic effect of the inward

integration of CED and EDBM is achieved.

3.3. Performances of pure feed salt (sodium sulfate) in

CED–EDBM (BP–C) stack

To test the feasibility of the combined CED–EDBM under pure

salt conditions, control experiments with sodium sulfate as the

only feed solution are also conducted. Trends similar to the mixed

feed salt process have also been observed in these pure feed salts

(Fig. 4). Energy consumptions and current efficiencies decrease

and increase with the increasing of feed concentration, respec-

tively, which are well consistent with the results of mixed feed

salts (Fig. 3b). Moreover, the performances of CED–EDBM (BP–C)

under pure salt conditions are lower than that of the mixed feed

salts, indicating the significant contribution of the dissociation of

sodium gluconate in mixed feed salts.

3.4. Comparison with CED–EDBM (BP–A) configuration and

process economics

For comparison, another integration configuration CED–EDBM

(BP–A) (Fig. 1b), was also tested for gluconic acid production com-

parison. Fig.5 showsthe energyconsumptionand current efficiency

of the different stack configurations. Among them, CED–EDBM

(BP–C) and CED–EDBM (BP–A) were the integration stacks, while

the EDBM (BP–C) was a single EDBM stack (cf. Figs. 1a–c). The

current density of the all the stacks is 50 mAcm−2; electrolyte

concentration is 0.30 moldm−3; and the sodium gluconate con-

centration is 0.05 moldm−3. The apparent energy consumption

of EDBM stack in both combined cases (CED–EDBM (BP–C) or

CED–EDBM (BP–A)) is less than that the single EDBM (BP–C) pro-

cess. However, when the energy consumption of CED stack is

also considered in the integration process, the CED–EDBM (BP–A)

integration process does not indicate much advantages than a

E D B M

( B P -

C )

C E D - E D

B M ( B P -

C )

C E D - E D

B M ( B P -

A )0

1

2

3

4





(

k W h ·

k g

- 1)

Different cell configuration

E D B M

( B P -

C )

C E D - E D

B M ( B P -

C )

C E D - E D

B M ( B P -

A )40

60

80

100

120

140



Fig. 5. Energy consumption and current efficiency for different cell configurations:

CED–EDBM (BP–C); CED–EDBM (BP–A); EDBM (BP–C). Experiment conditions:

current density of CED and EDBM stack, 50 mAcm−2; electrolyte concentra-

tion, 0.3moldm−3; sodium gluconate concentration 0.05 moldm−3; flow rate,

27dm−3

h−1

.




Table 2

Estimation of process cost.

EDBM (BP–C) CED–EDBM (BP–A) CED–EDBM (BP–C)

Repeating units 1 1 1

Current density (mA cm−2) 50 50 50

Experiment time (min) 120 120 120

Effective membrane area (cm2) 7.07 7.07 7.07

Na2SO4 concentration (mol dm−3) 0.3 0.3 0.3

Fluid flow speed (dm3 h−1) 27 27 27

NaGlu concentration (mol dm−3

) 0.05 0.05 0.05HGlu concentration (mol dm−3 ) 0.040 0.048 0.068

Current efficiency (%) 94.85 91.80 131.04

Energy consumption (kWh kg−1) 2.58 2.88 2.15

Process capacity (kgyear−1) 16.79 20.34 28.98

Electricity charge ($ kg−1) 0.1 0.1 0.1

Energy cost for HGlu ($ kg−1) 0.26 0.29 0.22

Energy cost for the peripheral equipment ($kg−1) 0.01 0.01 0.01

Total energy cost ($ kg−1) 0.27 0.30 0.23

Membrane life and amortization of the peripheral equipment (year) 3 3 3

Monopolar membrane price ($ m−2) 135 135 135

Bipolar membrane price ($ m−2) 1350 1350 1350

Membrane cost ($) 1.05 1.24 1.24

Stack cost ($) 1.57 1.86 1.86

Peripheral equipment cost ($) 2.36 2.79 2.79

Total investment cost ($) 3.94 4.65 4.65

Amortization ($ year−1) 1.31 1.55 1.55

Interest ($ year−1) 0.31 0.37 0.37

Maintenance ($ year−1) 0.39 0.47 0.47

Total fixed cost ($year−1) 2.02 2.39 2.39

Total fixed cost ($ kg−1) 0.12 0.12 0.08

Total process cost ($ kg−1 ) 0.39 0 .42 0.31

single EDBM (BP–C) process, as the total energy consumption is

much higher. Besides, the current efficiency of CED–EDBM (BP–A)

integration process is even lower than the single EDBM (BP–C)

process. The low efficiency of CED–EDBM (BP–A) process can be

explained by its particular stack configuration. In the CED–EDBM

(BP–A) integration process, the migration of Glu− through anion

exchange membrane is much more difficult than Na+ through

cation exchange membrane because the hydrate radii of gluconateion is much larger than that of sodium ion. In addition, the cath-

ode reactions in the EDBM stack generated OH-, which not only

competes with Glu− to go through anion exchange membrane but

also neutralize the H+ generated from BP membrane. These neg-

ative effects account for the high energy consumption and low

current efficiency of the CED–EDBM (BP–A) integration process.

While in the case of CED–EDBM (BP–C) integration process, the

apparent current efficiency is higher than the BP–C configuration

with less energy consumption as well as with concentrated glu-

conate besides.

The process cost is calculated by following the procedure as

reported in the literature [15], and the results are listed in Table 2.

For CED–EDBM (BP–C) integration configuration, the total process

costisestimatedtobe0.31$kg−1

, less than theother twoconfigura-tions (0.39$ kg−1 and 0.42$ kg−1, respectively). The process cost of

CED–EDBM (BP–A) integration configuration is higher than that of

a single EDBM (BP–C) configuration, indicating its synergistic effect

is limited.

4. Conclusions

The integration of conventional electrodialysis (CED) and elec-

trodialysis with bipolar membranes (EDBM) provides an effective

way to produce gluconic acid (HGlu) from sodium gluconate

(NaGlu). During the operations, both integration cell configuration

such as CED–EDBM (BP–C)and CED–EDBM (BP–A)as well as opera-

tionparameters suchas current density of CED stackand electrolyte

concentration play an important role on the integration character-

istics. As proven by the experiments results, the CED–EDBM (BP–C)

configuration is a cost-effective means to produce gluconic acid

from the viewpoint of energy consumption and current efficiency.

The electrode reactions and concentrating of salt in the CED stack,

which leading to an apparent current efficiency higher than 100%

and low energy consumption, are conducive to an EDBM process.

The process cost of CED–EDBM (BP–C) integration configuration is

estimatedto be 0.31$kg−1, less than another integration configura-tionCED–EDBM(BP–A) or a singleEDBM (BP–C) process. Obviously,

the process coupling of CED and EDBM can achieve a synergistic

effect, not only made the production cost-effective but also kept

the operation of EDBM stable.

Note that this work is a preliminary study. There is much more

work to do before bringing the process coupling to industrializa-

tion, such as scale-up the experiment from the now one repeating

unit, purification of gluconic acid from gluconate salt, etc.

Acknowledgments

This research is supported by the National Natural Science

Foundation of China (No. 20636050), the National Natural Sci-

ence Funds for Distinguished Young Scholar, the KnowledgeInnovation Program of the Chinese Academy of Sciences (No.

KSCX2-YW-G-075-25) and Foundations of Educational Committee

of Anhui Province (Nos. ZD200901, KJ2010A330 and KJ2008A69).

The authors thank Dr. C.H. Huang for proofreading the manuscript.

References

[1] L. Bazinet, F. Lamarche, D. Ippersiel, Bipolar-membrane electrodialysis: appli-cationsof electrodialysisin thefood industry,TrendsFood Sci.Technol. 9(1998)107–113.

[2] S. Novalic, T. Kongbangkerd, K.D. Kulbe, Recovery of organic acids with highmolecular weight using a combined electrodialytic process, J. Membr. Sci. 166(2000) 99–104.

[3] C.H. Huang, T.W. Xu, Electrodialysis with bipolar membranes for sustainable

development, Environ. Sci. Technol. 40 (2006) 5233–5243.




[4] T.W. Xu, C.H. Huang, Electrodialysis-based separation technologies: a criticalreview, AIChE J. 54 (2008) 3147–3159.

[5] F.L.T. Shee, P. Angers, L. Bazinet, Precipitationof cheddar cheese wheylipids byelectrochemical acidification, J. Agric. Food Chem. 53 (2005) 5635–5639.

[6] L. Bazinet, D. Ippersiel, F. Lamarche, Recovery of magnesium and protein fromsoy tofu whey by electrodialytic configurations, J. Chem. Technol. Biotechnol.74 (1999) 663–668.

[7] J.S.J. Ferrer, S. Laborie, G. Durand, M. Rakib, Formic acid regeneration by elec-tromembrane processes, J. Membr. Sci. 280 (2006) 509–516.

[8] T.W. Xu, W.H. Yang, Effect of cell configurations on the performance of cit-ric acid production by a bipolar membrane electrodialysis, J. Membr. Sci. 203

(2002) 145–153.[9] T.W. Xu,W.H. Yang, Citric acidproductionby electrodialysiswith bipolarmem-

branes, Chem. Eng. Process. 41 (2002) 519–524.[10] H. Strathmann, J.J. Krol, H.J. Rapp, G. Eigenberger, Limiting current density and

water dissociation in bipolar membranes, J. Membr. Sci. 125 (1997) 123–142.

[11] J.J. Krol, M. Jansink, M. Wessling, H. Strathmann, Behaviour of bipolar mem-branes at high current density water diffusion limitation, Sep. Purif. Technol.14 (1998) 41–52.

[12] D. Lide, Handbook of Chemistry and Physics, CRC Press Inc., Boca Raton, 1993.[13] V.I. Zabolotskii, N.V. Sheldeshov, N.P. Gnusin, Dissociation of water molecules

in systems with ion-exchange membranes, Russ. Chem. Rev. 57 (1988) 801–808.

[14] Y.M. Wang, C.H. Huang, T.W. Xu, Optimization of electrodialysis with bipo-lar membranes by using response surface methodology, J. Membr. Sci. (2010),doi:10.1016/j.memsci.2010.06.049.

[15] H.Strathmann,G.H. Koops,Processeconomicsof electrodialyticwaterdissocia-

tion for the production of acid and base, in: A.J.B. Kemperman (Ed.), Handbookon Bipolar Membrane Technology, Twente University Press, Enschede, 2000,pp. 191–220.

Integration of Conventional Electrodialysis and Electrodialysis With Bipolar

Documents