Water Research 89 (2016) 201e209

Contents lists available at ScienceDirect

Water Research
journal homepage: www.elsevier.com/locate/watres

Bipolar membrane electrodialysis for generation of hydrochloric acid

and ammonia from simulated ammonium chloride wastewater
Ya Li a, c, Shaoyuan Shi a, b, *, Hongbin Cao a, b, **, Xinmin Wu c, Zhijuan Zhao a, b,
Liying Wang d
a
National Engineering Laboratory for Hydrometallurgical Cleaner Production Technology, Beijing Engineering Research Center of Process Pollution Control,
Key Laboratory of Green Process and Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, China
b
University of Chinese Academy of Sciences, Beijing 100049, China
c
College of Chemical Engineering, Beijing Institute of Petro-Chemical Technology, Beijing 102617, China
d
Anshan Iron and Steel Group Corporation, Anshan 114021, Liaoning, China

a r t i c l e i n f o a b s t r a c t

Article history: Simulated ammonium chloride wastewater was treated by a lab-scale bipolar membrane electrodialysis
Received 13 July 2015 for the generation of HCl and NH3$H2O and desalination. The influence of initial concentration of NH4Cl,
Received in revised form current density, salt solution volume, initial concentration of acid and base and membrane stack
7 November 2015
structure on the yields of HCl and NH3$H2O was investigated. The current efficiency and energy con-
Accepted 15 November 2015
sumption were also examined under different conditions. The results showed that, at the current density
Available online 22 November 2015
of 48 mA/cm2, the highest concentration of HCl and NH3$H2O with initial concentration of 110 g/L NH4Cl
was 57.67 g/L and 45.85 g/L, respectively. Higher initial concentration of NH4Cl was favor to reduce unit
Keywords:
Ammonium chloride wastewater
energy consumption and increase current efficiency of the BMED system. The membrane stack voltage of
Bipolar membrane electrodialysis BMED increased quickly under constant current when the concentration of NH4Cl contained in the so-
Influence factors lution of salt compartment was depleted below the “inflection point concentration” about 8000 mg/L. It
means that the concentration of NH4Cl below 8000 mg/L was no longer suitable for BMED because of
higher energy consumption. The HCl and NH3$H2O concentration increased more quickly following the
increase of current density. When increasing the volume of NH4Cl, the concentration of HCl and NH3$H2O
also increased. The high initial concentration of acid and base could improve the final concentration of
them, while the growth rate was decreased. Compared with the BMED system with three compartments,
the growth rate of HCl concentration with the two compartments was higher and its unit energy con-
sumption was lower. It meant that the performance of the BMED system could be improved by opti-
mizing operation conditions. The application feasibility of the generation of HCl and NH3$H2O and
desalination of ammonium chloride wastewater by BMED was proved.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction processes containing ammonium chloride have been continuously

Due to the fast development of economy and industrialization,
industrial effluents from chemical fertilizer and rare earth
* Corresponding author. National Engineering Laboratory for Hydrometallurgical
Cleaner Production Technology, Beijing Engineering Research Center of Process
Pollution Control, Key Laboratory of Green Process and Engineering, Institute of
Process Engineering, Chinese Academy of Sciences, Beijing 100190, China.
** Corresponding author. National Engineering Laboratory for Hydrometallurgical
Cleaner Production Technology, Beijing Engineering Research Center of Process
Pollution Control, Key Laboratory of Green Process and Engineering, Institute of
Process Engineering, Chinese Academy of Sciences, Beijing 100190, China.
E-mail address: syshi@ipe.ac.cn (S. Shi).
discharged into the environment and resulted in severe problems
(Wang et al., 2014a; Huang et al., 2012). Ammonium chloride is one
of the crucial sources of ammonia nitrogen and discharging of the
ammonium chloride containing wastewater to the fresh water re-
sources could lead to several environmental problems such as
eutrophication, excessive propagation of algae and accumulation of
chloride ion (Wang et al., 2014a; Li and Qiu, 2013). It was reported
that more than 2.4 million m3 of wastewaters containing ammonia
chloride were discharged in 2013 in China. Therefore, ammonia
nitrogen pollution has become a serious environmental concern in
many newly industrialized countries and the development of
effective technologies for removal of ammonia nitrogen is in high
demand.

http://dx.doi.org/10.1016/j.watres.2015.11.038
0043-1354/© 2015 Elsevier Ltd. All rights reserved.
202 Y. Li et al. / Water Research 89 (2016) 201e209
The technologies of nano-filtration, reverse osmosis and
biodegradation have been investigated and tried to apply in the
treatment of ammonia chloride effluents (De Gisi et al., 2009;
Pe
rez-Gonza
lez et al., 2012; Jin et al., 2013), but those methods
still have some disadvantages. Huang et al., proposed an integrated
method in which evaporation crystallization, ammonia stripping
and break-point chlorination processes were involved to treat the
ammonia chloride containing effluent discharged from rare-earth
separation plants with various concentration (Huang et al., 2012).
Nevertheless, suffering from high energy consumption, secondary
pollution and high capital and process costs, this method was
hindered for the wider applications. Some synergistic effects
combining biodegradation and physical treatments were also
applied to remove the ammonium nitrogen within the wastewater.
The studies revealed that by incorporating anaerobic, anoxic and
oxic processes with nano-filtration and reverse osmosis, 99.8% of
ammonium nitrogen within the coking wastewater was removed
(Jin et al., 2013). However, the concentrated wastewater after nano-
filtration and reverse osmosis processes obtaining higher ammo-
nium and chemical oxygen demand (COD) level requires further
treatments. Ion exchange materials, such as weak and strong acid
cation exchange resins, natural and synthetic zeolite in packed bed
columns were used to concentrate the ammonium contained in
municipal wastewater (Malovanyy et al., 2013). However, the
concentrated liquid of ammonium also needs to be further treated
before discharge. Although a lot of effort was paid to improve the
treatment of industrial wastewater with ammonium chloride, the
environmental pollution resulted from ammonium chloride
effluent and wastewater resources has not been solved completely.
Thus more efforts should have been devoted for the treatment of
ammonia nitrogen wastewaters.
Being a new type of electrodialysis technology, bipolar mem-
brane electrodialysis (BMED) has received intensive research
attention due to its high efficiency to produce acid and base from
the neutral salt under the direct electric field, in which the bipolar
membrane with the interface layer is able to split water into Hþ and
OH ions under the direct electric field (Mafe
et al., 1998). Fruitful
researches have been conducted using this BMED technology for
the production of organic acids (Bailly, 2002; Quoc et al., 2011;
Lameloise and Lewandowski, 2012). It was reported that succinic
acid could be prepared using BMED with high current efficiency
over 90% and low unit energy consumption below 2.3 kWh/kg (Fu
et al., 2014). Moreover, the production of tartaric acid from BMED
method has overcome several disadvantages of the traditional
methods such as high labor intensity, low recovery and notorious
solid pollution (CaSO4) (Xu, 2002; Zhang et al., 2009). In the recent
years, much attention has been paid to the application of BMED
technology in the treatment of industrial effluents (Wei et al., 2011;
Wang et al., 2014b; Graillon et al., 1996; Rehouma et al., 2013). It
was reported that the metal washing step water could be treated by
BMED for the regeneration of inorganic acid and base, and the
concentration of acid and base was up to 1.76 N and 2.41 N,
respectively (Tran et al., 2015). The industrial NaCl stream was
treated by BMED technology with the desalination of 99%, and both
of HCl and NaOH produced from the industrial saline water could
attain the concentration between 1.5 and 2.0 M simultaneously
(Ghyselbrecht et al., 2014, 2013). These studies presented that
BMED technology had a good application prospect in organic acid
Table 1
The main characteristic of mono-polar ion exchange membranes used in the experiments.
Membrane Thickness (mm) Water content (%)
Cation exchange membrane (PE-001) 0.3 35
Anion exchange membrane (JAM-10) 0.24e0.26 20e24
preparation and industrial saline wastewater treatment. However,
reports about the treatment of ammonium chloride effluents by
BMED technology have not been found up to now.
In this study, simulated ammonium chloride wastewater was
treated by a lab-scale BMED for the generations of HCl and
NH3$H2O and desalination at the same time. The influence of initial
concentration of NH4Cl, current density, saline wastewater volume,
initial concentration of acid and base and membrane stack struc-
ture on acid and base generation by BMED was investigated. The
current efficiency, energy consumption, acid and base concentra-
tions under different conditions were examined to test the appli-
cation feasibility of BMED technology in the treatment of simulated
ammonium chloride wastewater and the optimized conditions of
the system were obtained.
2. Materials and methods
2.1. Reagents and ion exchange membranes
The following reagents were used in the experiments: ammo-
nium chloride (solid, 99.5%), sodium sulfate (0.05 mol/L), hydro-
chloric acid (0.05 mol/L) and ammonia solution (0.05 mol/L) were
used to prepare the initial solutions used in salt, acid and base
compartments, respectively. Sodium hydroxide (0.5 mol/L), silver
nitrate (0.1 mol/L), methyl orange (IND), phenolphthalein (IND) and
potassium chromate (10 g/L) were used in the chemical analyses.
All the reagents were produced by Xilong chemical Co. Ltd.
Three ion exchange membranes (anion exchange membranes
JAM-10, bipolar membranes BP-001 and cation exchange mem-
brane PE-001) were used in the experiments, in which JAM-10 and
BP-001 are provided by Guangya Chemical Co. Ltd, China. PE-001 is
provided by Shanghai Xiangfeng Water Treatment Equipment Co.
Ltd, China. The main characteristics of ion exchange membranes
provided by the manufacturers were presented in Tables 1 and 2.
BMED apparatus used in the experiments was provided by
Sanyuanbada Technology Development Co. Ltd, China. The style of
bipolar membrane configuration in this study was BM-A-C-BM
namely three-compartment system including salt, acid and base
compartments shown in Fig. 1(a). It could be seen clearly from
Fig. 1(a) that the simulated ammonium chloride wastewater, HCl
solution, ammonia solution and electrode rinsing solution were
circulated in the three-compartment system driven by four peri-
staltic pumps, respectively. Each ion exchange membrane had an
effective surface area of 50 cm2 and the membrane triples were 10
in a stack. Electrodes included a titanium electrode coated with
ruthenium oxide as the anode and a stainless steel electrode as the
cathode. Anions (Cl) and cations (NHþ 4 ) migrated through anion
exchange membranes into acid compartments and cation exchange
membranes into base compartments under a direct electric field,
respectively. Then the Cl and NHþ þ
4 combined with H and OH

generated from water splitting at the interface of bipolar mem-
brane (BM) to produce HCl and NH3$H2O.
The ammonia, as a weak electrolyte generated in base
compartment, can lead to the high solution resistance of base
compartment to increase energy consumption of the BMED system.
Therefore, the BMED system with two-compartment including salt
and acid compartments was also employed and shown in Fig. 1(b).
The style of bipolar membrane configuration was BM-A-BM, in
Ion exchange capacity (meq/g) Area resistance (U cm2) Transport number (%)
2 ± 0.02 4.5 >98
1.6e1.7 7e10 96
 Y. Li et al. / Water Research 89 (2016) 201e209 203

Table 2
The main characteristics of bipolar membrane used in the experiments.

Membrane Thickness (mm) Water splitting voltage(V) (100 mA/cm2, 0.5 M Na2SO4, 25  C) Burst strength (105 Pa) Area resistance (U cm2)

Bipolar membrane (BP-001) 0.24e0.26 0.9e1.5 2.5e3.0 8e9

using methyl orange as indicator. The concentration of Cl ions in

(a)
salt solution was analyzed using a standard silver nitrate solution as
BM A C BM A C BM
the titrant and using potassium chromate as indicator. The NHþ 4

NH4+ concentration of salt compartment was measured by an ammonia

NH4+
electrode (M124, Mettler Toledo). The conductivity and pH of the
OH- solutions in all the compartments were measured by a
H+ OH- H+ OH- _
+
H+ conductivity-meter (S230 SevenCompact conductivity, Mettler
Cl-
Toledo) and a pH-meter (S220 SevenCompact pH/ion, Mettler
Cl-
Toledo), respectively.
anode cathode
2.2.2. Data analysis
The performance of BMED used in ammonium chloride effluents
treatment by generating of HCl and NH3$H2O from NH4Cl was
Ammo- ERS evaluated in terms of energy consumption (EC) and average current
Acid uium Base 0.05M
HCl NH4OH efficiency (ACE). The average current efficiency was calculated as
chloride Na2SO4
equation (1) (Tran et al., 2015).

zFjC0 V0  Ct Vt j
ACE ¼ (1)
60NIDt

where z is the ions' valence. F is Faraday constant (96485C/mol). C0

(b) and Ct (mol/L) are the concentration of acid or base in the acid or
BM A BM A BM
base compartment at time 0 and t (min), respectively. V0 and Vt are
NH4+ NH4+ the volume (L) of acid or base solution at time 0 and time t (min). N
-
is the number of cell triplets (N ¼ 10). I is the current (A) for this
OH
H+ OH- H+
OH- _ time period. Dt (min) is the time period. t is the time (min).
+
H+ The energy consumption (kWh/kg) was calculated as equation
Cl- Cl- (2) (Tran et al., 2015),
anode cathode Z t
UIdt
0
EC ¼ (2)
60MðCt Vt  C0 V0 Þ
Ammo- ERS
Acid uium 0.05M where U is the voltage (V) of membrane stack at time t. I is the
HCl chloride Na2SO4 current (A). t is the time (min). M is the base or acid's molar mass (g/
mol). C0 and Ct (mol/L) are the concentrations of acid or base in the
acid or base compartment at time 0 (min) and t (min), respectively.
V0 and Vt are the volume (L) of acid or base solution at time 0 and
Fig. 1. Schematic diagrams of the BMED system with three (a) and two compartments
(b).
time t (min).

2.3. Experimental procedures

which NHþ 4 ions were retained in salt compartment and combined
with OH ions generated by BM to produce the base (NH3$H2O).
Driven by three diaphragm pumps, the HCl solution, NH4Cl and
NH3$H2O mixing solution and electrode rinse solution were circu-
lated in two-compartment system, respectively.
To examine the influence of membrane stack structure on the
generation of HCl and NH3$H2O from the solution of NH4Cl by
BMED, the concentration of acid and base, current efficiency and
unit energy consumption and the conductivity of feed solution
were examined with different membrane stack structure.
2.2. Analysis method and calculations
The experiments were performed at constant current densities
controlled by a direct-current (DC) rectifier. Unless otherwise
noted, the simulated ammonium chloride wastewater was circu-
lated through the salt compartment and 0.05 mol/L HCl and
0.05 mol/L NH3$H2O were circulated through the acid and base
compartments, respectively, driven by different pumps at a con-
stant flow rate of 45 L/h in the BMED system with three-
compartment. A 0.05 mol/L Na2SO4, as electrode rinse solution,
was circulated through the electrode compartment at the flow rate
of 20 L/h. The liner velocities of acid, base, and salt solution are
3.1 cm/s and it is calculated as equation (3),

2.2.1. Analysis methods 103  Q

The hydrochloric acid concentration in acid compartment was V¼ (3)
3600  N  W  d
determined by titration with standard sodium hydroxide solution
using phenolphthalein as indicator. The ammonia concentration in where V (cm/s) is the liner velocities, Q (L/h) is the flow rate of acid,
base compartment was titrated with a standard hydrochloric acid base or salt solution (45 L/h). N is the number of cell triplets
204 Y. Li et al. / Water Research 89 (2016) 201e209
(N ¼ 10). W (cm) is the width of flow section area (W ¼ 5 cm).
d (cm) is the spacer thickness between two membranes (0.8 mm).
Unless otherwise noted, the volume of every solution was
700 mL. The voltage of membrane stack, conductivity and pH of salt
solution were measured at a constant interval of 10 min. The
samples of acid and base solutions were withdrawn at a constant
time period of 20 min.
3. Results and discussion
3.1. Influence of initial salt concentration
In order to investigate the effect of ammonium chloride initial
concentration on current efficiency, unit energy consumption, acid
and base yields in a lab-scale BMED system, the experiments were
performed with simulated wastewater containing NH4Cl with
different concentrations such as 60 g/L, 90 g/L and 110 g/L.
Fig. 2 indicated that the changes of membrane stack voltages
and Cl concentration in the salt solution with different concen-
tration of NH4Cl in the BMED system at the current density of
48 mA/cm2. It was found that the membrane stack voltage of BMED
dropped gradually at the beginning of the experiments, however, it
increased sharply at the end of the experiments. The main reason
should be that the initial concentration of acid and base in acid and
base compartments was very small, which resulted in the higher
membrane stack resistance in the BMED system at the beginning of
the experiment, so the membrane stack voltage of the BMED sys-
tem was higher. Along with the production of acid and base, the
decrease of acid and base solutions resistance could lead to the
voltage drop. The lower resistance of solutions in different com-
partments led to the lower voltage of membrane stack at constant
current, and vice versa. It was shown that the changes of acid, base
and salt concentration could affect the changes of membrane stack
voltage because of the changes of solution resistance in different
compartments of BMED, in which the sharp increase of membrane
stack voltage was resulted from the depletion of NH4Cl in salt
compartment at the end of the experiments. So the changes of stack
voltage were related to the increases of HCl and NH3$H2O con-
centration in the acid and base compartments at the earlier stage of
experiments and the depletion of NH4Cl in salt compartment at the
end of experiments.
When the voltages of BMED membrane stack increased quickly,
the samples of the solution in salt compartment were withdrawn
and the concentration of Cl and NHþ 4 was analyzed. It was found
that the concentration of NH4Cl contained in salt solution
Fig. 2. The changes of membrane stack voltages and Cl ions concentrations in salt
solution for the different concentration of NH4Cl in the BMED system.
decreased to less than 8000 mg/L, which was in accordance with
the “inflection point” of stack voltage in the BMED system. The
results revealed that when the NH4Cl concentration within the
effluent was less than of 8000 mg/L, it was no longer suitable to be
treated by BMED, because the lower concentration of NH4Cl con-
tained in salt solution could lead to the higher energy consumption.
The wastewater with lower concentration of NH4Cl could be
desalted further by traditional electrodialysis for the aim of water
reuse, while its concentrated solution might be used again to
generate acid and base by BMED.
Fig. 3 showed that the concentrations of acid (HCl) and base
(NH3$H2O) generated from the simulated wastewater with
different concentration of NH4Cl were increased generally during
the BMED processes. As seen, when the simulated wastewater
containing NH4Cl of 60 g/L, 90 g/L and 110 g/L after continuously
run for 90 or 220 min, the concentrations of acid and base were
29.2 g/L, 46.36 g/L, 57.67 g/L and 26.25 g/L, 34.65 g/L, 45.85 g/L,
respectively. The concentration of HCl obtained could meet the
requirements of resin regeneration (Zhang and Zheng, 2013; Zhan
and Yu, 2010) and chemical cleaning of reverse osmosis mem-
brane (Madaeni and Mansourpanah, 2004; Madaeni and Samieirad,
2010; Sayed et al., 2001), and the NH3$H2O could be used for the
flue gas desulfurization (Ding, 2009). Moreover, the obtained acid
and basic can be concentrated by evaporation or other methods for
the more application. It was also found from Fig. 3 that the in-
creases of acid and base concentrations at the beginning of the
experiments were quicker than that at the later period of experi-
ments, which was resulted from the decrease of NH4Cl concentra-
tion and water electricity penetration in salt compartment (Tanaka,
2004). The decrease of NH4Cl contained in the feed solution
necessarily led to the less of NHþ 
4 and Cl ions to immigrate to the
adjacent acid and base compartments. And along with the trans-
membrane migration of ions, parts of the water existed in the form
of hydrated ions migrated from the salt compartment into the
adjacent acid and base compartments through ion exchange
membranes in DC electric field (i.e. water electricity penetration) to
lead the increase of acid and base solution volumes, which delayed
the increase of HCl and NH3$H2O concentration in the acid and base
compartments. The results meant that higher concentration of
NH4Cl was favor to produce higher concentration of HCl and
Fig. 3. The concentrations of acid (HCl) and base (NH3$H2O) generated from the
different concentrations of NH4Cl by BMED.
 Y. Li et al. / Water Research 89 (2016) 201e209 205

NH3$H2O by BMED. Moreover, the concentration of HCl was obvi-

ously higher than that of NH3$H2O. It was supposed that the
volatility of NH3$H2O led to its loss, thus making the decrease of
NH3$H2O concentration in base compartment, especially in the
prolonged experiment time. The mass balance of NH3 was calcu-
lated to verify the amount of NH3 that is missing. The missing mass
of NH3 were 0.17 g, 0.85 g and 1.36 g with the feed concentration of
60 g/L, 90 g/L and 110 g/L NH4Cl. The results show that the high
concentration of NH3$H2O is favor to promote the volatilization of
NH3. In addition the pH values in base compartment is about 11,
that suggested the concentration of OH is very low and most of the
NH3$H2O exist in molecular state in the base compartment which is
helpful for the volatilization of NH3.
The changes of conductivity and pH of the solution with
different concentration of NH4Cl in salt compartment of the BMED
system were shown in Fig. 4. It was found from Fig. 4(a) that the
conductivity of solution in salt compartment decreased gradually
to zero, suggesting that NH4Cl in salt compartment was almost
depleted at the end of experiments. However, the increases of HCl
and NH3$H2O concentration in acid and base compartments were Fig. 5. The current efficiency and unit energy consumption of acid (HCl) and base
(NH3$H2O) generated from different concentration of NH4Cl by BMED.
limited by the lower concentration of NH4Cl in salt compartment. It
was also proved that the ammonium chloride effluent could be
treated by BMED to realize the goal of NH3$H2O and HCl generation
decrease the unit energy consumption of HCl and NH3$H2O by
and desalination at the same time. The changes of pH value shown
BMED. The energy consumption of 75 g/L Na2SO4 at the current
in Fig. 4(b) indicated that the solutions of salt compartment were
density of 50 mA/cm2 in BMED system was 5.0 kWh/kg (0.49 kWh/
always acidic, in which the acidity of the simulated wastewater in
mol) for the acid (H2SO4) and 4.1 kWh/kg (0.16 kWh/mol) for the
the initial period resulted from the hydrolysis of NH4Cl. While the
base (NaOH) (Tran et al., 2015). While the energy consumption of
decrease of pH in salt compartment might be the leakage of Hþ ions
NH3$H2O with the 60 g/L NH4Cl at the current density of 48 mA/
from acid compartment into adjacent salt compartment during the
cm2 presented in Fig. 5 was 9.1 kWh/kg (0.32 kWh/mol) which is
BMED processes, because the “tunnel effect” of Hþ had led to the
higher than it in literature (Tran et al., 2015). The current efficiency
higher permeability of anion exchange membrane for proton (Fu
of 35 g/L NaCl at the current density of 50 mA/cm2 by BMED was
et al., 2014; Rottiers et al., 2014). The increase of salt solution pH
n
~ ez et al., 2013) which was higher than the current
60e80% (Iba
value at the end of the experiment might be ascribed to the con-
efficiency values in Fig. 5. The higher energy consumption and
centration polarization at constant current, because the concen-
lower current efficiency in the BMED system of NH4Cl might be
tration of NH4Cl in the salt compartment was too low at that period
resulted from that the ammonia is weak electrolyte and it is easy to
(Krol et al., 1999).
volatilize. That resulted to the high solution resistance of the BMED
Fig. 5 presents the current efficiency and unit energy con-
system. In addition, the initial concentration of salt, the current
sumption of acid and base generated from different initial con-
efficiency and unit energy consumption of acid and base in BMED
centration of NH4Cl in the BMED system. It was found that the
were also related to other factors such as the properties of ion
higher initial concentration of NH4Cl contained in the simulated
exchange membrane, membrane stack constructer and operating
wastewater was favor to increase the current efficiency and
conditions (Ghyselbrecht et al., 2014; Wang et al., 2013).

3.2. Effects of current density

The effect of current density on the concentration of acid and

Fig. 4. The conductivity (a) and pH (b) of the solution with the different concentration
of NH4Cl in the salt solution of BMED system.
base, current efficiency and unit energy consumption were exam-
ined in the BMED system with 60 g/L NH4Cl at different current
densities (38 mA/cm2, 48 mA/cm2 and 58 mA/cm2). The results
shown in Fig. 6 indicated that both of the acid and base concen-
tration in the acid and base compartments increased with the in-
crease of current density in the BMED system with 60 g/L NH4Cl. It
was proved that the proper increase of current density was helpful
to improve the performance of BMED system to produce HCl and
NH3$H2O from NH4Cl for shortening the operation time and
reducing the concentration diffusion of ions between the adjacent
compartments of membrane stack. This growth pattern of acid and
base with different current densities is similar to the NaCl system in

n
literature (Iba ~ ez et al., 2013). The concentration of HCl generated
from NH4Cl was close to 31 g/L after continuously running for 120,
90 and 60 min, respectively, at the current densities of 38, 48 and
58 mA/cm2. However the concentration of NH3$H2O was always
slightly lower than the corresponding concentration of HCl for the
different current densities.
Fig. 7 showed that the current efficiency increased and unit
206 Y. Li et al. / Water Research 89 (2016) 201e209
Fig. 6. The concentrations of acid (HCl) and base (NH3$H2O) generated from the so-
lution containing 60 g/L NH4Cl by BMED.
Fig. 7. The current efficiency and unit energy consumption of acid and base generated
from 60 g/L NH4Cl by BMED with different current densities.
energy consumption decreased in the BMED system of acid and
base generated from 6% NH4Cl when the current density increased
from 38 mA/cm2 to 58 mA/cm2. The higher current density could
accelerate the transmembrane migration of ions and the dissocia-
tion of H2O to Hþ and OH ions. Moreover, the higher current
density meant the higher membrane stack voltage, which was
favorable to inhibit reverse diffusion of ions, especially at the end of
the experiments when there was a large concentration gradient
between the adjacent compartments of salt and acid/base. In
addition, the higher current density could shorten the reaction time
to decrease the loss of ammonia due to ammonia volatilization and
improve the production efficiency. Therefore, it is unfavorable to
use too low current density. However, the too high current density
can exceed the limiting current density, which leads to the occur-
rence of side reactions and high energy consumption. In this work
the limiting current density was checked and the selected three
current density values were less than the limiting current density at
the beginning of the experiments. The appropriate high current
density is suitable for the BMED system, especially in the solution
with higher salt concentration.
3.3. Effects of salt solution volume and the initial concentration of
acid and base solution
The increases of acid and base concentration in the BMED sys-
tem could be limited in the lab-scale apparatus when the NH4Cl
was close to zero. Therefore, the effect of salt solution volume on
the concentration of acid and base generated from NH4Cl by BMED
was further examined. The experiments were performed with 60 g/
L NH4Cl, in which both of the volumes of acid and base solutions
were 700 ml and the volume of salt solution was 2100 ml (Vsalt:Vacid
or base ¼ 3:1), and the other conditions were same as the ones in
Section 3.1. The changes of HCl and NH3$H2O concentration in the
BMED systems with different volumes of NH4Cl solution were
presented in Fig. 8. It was shown that when the volume of salt
solution was 2100 mL, the concentrations of acid and base were
48.18 g/L and 43.05 g/L, respectively. Which were higher than the
concentration of acid (29.20 g/L) and base (26.25 g/L) when the
volume of salt solution was 700 mL. The increase of salt solution
volume meant the higher total mole number of NH4Cl, which was
favorable to make more Cl and NHþ 4 to transmembrane migrate
into acid and base compartments in the BMED system. The results
indicated that the concentration of HCl and NH3$H2O generated by
BMED was improved with the increase of salt solution volume.
Therefore, increase the concentration of salt solution (such as the
method in Section 3.1) or increase the volume of salt solution with
lower concentration may be a good way to improve the concen-
tration of HCl and NH3$H2O.
The energy consumption and current efficiency of acid and base
with the different volumes of salt solution were shown in Fig. 9. It
was found that the increase of Vsalt:Vacid or base could cause the
increase of current efficiency and the reduction of unit energy
consumption of acid and base. The current efficiency of HCl had
been improved from 43% to 56% and the energy consumption of
generating HCl had been reduced from 9.0 kWh/kg to 6.1 kWh/kg
by increasing the volume of salt solution from 700 mL to 2100 mL
after run 90 min. The current efficiency and unit energy con-
sumption for the NH3$H2O generation also presented the similar
changes with that of HCl in the BMED system with the different
volume of NH4Cl solution. Increase the volume of salt solution may
Fig. 8. The concentration of acid and base with different volumes of salt solution.
 Y. Li et al. / Water Research 89 (2016) 201e209
Fig. 9. The effect of salt solution volume on the current efficiency and energy
consumption.
be a good method to get higher concentration of HCl and NH3$H2O
from NH4Cl wastewater with BMED.
The experiment with 0.3 mol/L HCl and NH3$H2O was also taken
to investigate the influence of initial HCl and NH3$H2O concentra-
tions on the productions of acid and base in BMED system. The
results of 0.3 mol/L HCl and NH3$H2O concentration compared with
0.05 mol/L HCl and NH3$H2O were showed in Fig. 10. It could be
seen clearly from Fig. 10 that the higher initial concentrations of
acid and base helped to get a higher concentrations of final HCl and
NH3$H2O, while the growth rate of HCl and NH3$H2O concentra-
tions was inhibited by the higher initial concentrations of HCl and
NH3$H2O. The reason may be that the higher initial concentrations
of acid and base led to the increase of transfer resistance from salt
compartment to acid and base compartments which would inhibit
the growth rate of HCl and NH3$H2O. Along with the decrease of
salt solution concentration, the concentrations of acid and base
solution increased and made the higher ion concentrations in the
acid and base compartments. The high concentration gradient
Fig. 10. The influence of initial concentration of acid and base on the production of HCl
and NH3$H2O.
207
between the adjacent compartments could cause the ions back-
diffusion from acid and base compartments to salt compartment.
Therefore, the higher initial concentration of acid and base may be a
method to increase the concentrations of acid and base generated
with BMED, while the growth rate of acid and base would be
slower. That may be not a suitable method to improve the pro-
duction of acid and base from NH4Cl wastewater by BMED.
3.4. Effect of BMED membrane stack
In order to avoid the effect of high resistance of ammonia water
in base compartment on the performance of BMED, the membrane
stack of BMED was modified to two-compartment from three-
compartment shown in Fig. 1(b), in which the cation exchange
membranes were removed and there were only anion exchange
membranes and bipolar membranes that arranged alternately in
the membrane stack, in which the transmembrane migration of
NHþ 4 didn't occurred during the BMED process. In the two-
compartment of BMED system, HCl and the mixing solution of
NH4Cl and NH3$H2O were produced and circulated in the acid and
salt compartments, respectively. The conductivity of the solution in
salt compartment, concentration of HCl, the current efficiency and
energy consumption in the two BMED systems with different
membrane stacks were shown in Figs. 11 and 12.
Fig. 11 showed that the conductivity of mixing solution in salt
compartment in two-compartment system dropped faster than
that in three-compartment system. The main reason was that the
total transmembrane migration number of ions was same in the
two BMED systems at the same current density. In the two-
compartment system, the absence of transmembrane migration
of NHþ 
4 led more Cl ions to immigrate into the acid compartment
through anion exchange membrane. Although there was the same
amount of OH ions produced because of water dissociation by BM,
most OH ions were combined with NHþ 4 into NH3$H2O quickly. So
the decrease of the conductivity in salt compartment in the two-
compartment system was more rapid than that in the three-
compartment system.
Fig. 12(a) showed the HCl concentration of the two BMED sys-
tems. As seen, the concentration of HCl in the two-compartment
system was higher than that in the three-compartment system.
The reason was also that the absence of transmembrane migration
of NHþ 
4 led more Cl ions to immigrate into the acid compartment
through anion exchange membrane which lead to the increase of
HCl in the two-compartment system. While the growth rate of HCl
Fig. 11. The conductivity of the three-compartment and two-compartment system.
208 Y. Li et al. / Water Research 89 (2016) 201e209
Fig. 12. The concentration of HCl (a) and the current efficiency and energy con-
sumption of HCl (b) generated from the BMED system with different membrane stacks.
concentration in the two-compartment system was not much
higher than that in three-compartment system. The reason might
be that a little part of OH ions in the salt compartment instead of
Cl ions transferred through anion exchange membrane to the
adjacent acid compartment and combined with Hþ into H2O. Thus
resulted the growth rate of HCl concentration in the two-
compartment system was not much higher than that in the
three-compartment system.
It was found from Fig. 12(b) that the current efficiency in two-
compartment system was higher than that in three-compartment
system. The lower energy consumption of the two-compartment
system resulted from the lower membrane stack resistance with
the absence of cation exchange membrane and base compartments,
compared with the three-compartment system. Moreover, the
absence of cation exchange membrane could also reduce the
leakage of ions, especially at the end of experiment. That is a reason
for the higher current efficiency of two-compartment system. The
growth rate of HCl concentration in two-compartment system was
higher than that in the three-compartment then could also
improve the current efficiency. Therefore, the BMED system with
two compartments might be more suitable to the application in the
treatment of industrial effluents containing ammonium chloride.
However, less pure NH3$H2O was obtained in the two-
compartment system. The ammonia might be blow-off and adsor-
bed to form NH3$H2O for its application further.
4. Conclusions
The application feasibility of simulated ammonium chloride
wastewater treated by a lab-scale BMED system for the generation
of NH3$H2O and HCl and desalination was proved in the work. The
results showed that the initial concentration of ammonium chlo-
ride, current density, salt solution volume and membrane stack
structure could affect the HCl and NH3$H2O yield, current efficiency
and unit energy consumption in the BMED system. It was shown
that the higher initial concentration of ammonium chloride was in
favor to improve the acid and base concentrations and current ef-
ficiency, while reduce the unit energy consumption. The higher
concentration of HCl and NH3$H2O were 57.67 g/L and 45.85 g/L,
respectively, obtained in the solution of 110 g/L NH4Cl. The mem-
brane stack voltage of BMED rose sharply at constant current
density when the NH4Cl concentration was depleted to below the
“inflection point concentration” about 8000 mg/L, indicating that
the wastewater was no longer suitable to be treated by BMED
because of higher energy consumption. The concentration of HCl
and NH3$H2O increase from 29.20 g/L and 26.25 g/L to 48.18 g/L and
43.05 g/L, respectively, when changing the volume of 60 g/L NH4Cl
from 700 mL to 2100 mL. The high initial concentration of acid and
base could improve the final concentration of them, while the
growth rate was decreased. Compared with the three-
compartment BMED system, it was found that the increase of HCl
concentration in the two-compartment BMED system was obvi-
ously quicker and its unit energy consumption was reduced
because of the lower membrane stack resistance, which was
resulted from the absence of cation exchange membranes and base
compartments.
Acknowledgements
The financial support from National Natural Science Foundation
of China (21076214), Beijing Natural Science Foundation (8132047)
and National Water Pollution Control And Management Technology
Major Projects (2014ZX07201-011-003) are greatly appreciated. We
also thank the anonymous reviewers for their very valuable com-
ments to improve this work.
References
Bailly, M., 2002. Production of organic acids by bipolar electrodialysis: realizations
and perspectives. Desalination 144 (1e3), 157e162.
De Gisi, S., Galasso, M., De Feo, G., 2009. Treatment of tannery wastewater through
the combination of a conventional activated sludge process and reverse osmosis
with a plane membrane. Desalination 249 (1), 337e342.
Ding, Y., 2009. Ammonia Absorption Process for Boiler Flue Gas Desulfurization.
M-sized Nitrogenous Fertilizer Progess 31 (4), 12e14 (in Chinese).
Fu, L., Gao, X., Yang, Y., Aiyong, F., Hao, H., Gao, C., 2014. Preparation of succinic acid
using bipolar membrane electrodialysis. Sep. Purif. Technol. 127, 212e218.
Ghyselbrecht, K., Huygebaert, M., Van der Bruggen, B., Ballet, R., Meesschaert, B.,
Pinoy, L., 2013. Desalination of an industrial saline water with conventional and
bipolar membrane electrodialysis. Desalination 318, 9e18.
Ghyselbrecht, K., Silva, A., Van der Bruggen, B., Boussu, K., Meesschaert, B., Pinoy, L.,
2014. Desalination feasibility study of an industrial NaCl stream by bipolar
membrane electrodialysis. J. Environ. Manag. 140, 69e75.
Graillon, S., Persin, F., Pourcelly, G., Gavach, C., 1996. Development of electrodialysis
with bipolar membrane for the treatment of concentrated nitrate effluents.
Desalination 107 (2), 159e169.
Huang, H., Xiao, X., Yan, B., 2012. Complex treatment of the ammonium nitrogen
wastewater from rare-earth separation plant. Desalin. Water Treat. 8 (1e3),
109e117.
Ib

an~ ez, R., Pe
rez-Gonza
lez, A., Go
mez, P., Urtiaga, A.M., Ortiz, I., 2013. Acid and base
recovery from soften reverse osmosis (RO) brines. Experimental assessment
using model concentrates. Desalination 309, 165e170.
Jin, X., Li, E., Lu, S., Qiu, Z., Sui, Q., 2013. Coking wastewater treatment for industrial
reuse purpose: combining biological processes with ultrafiltration, nano-
filtration and reverse osmosis. J. Environ. Sci. 25 (8), 1565e1574.
Krol, J.J., Wessling, M., Strathmann, H., 1999. Concentration polarization with
monopolar ion exchange membranes: currentevoltage curves and water
dissociation. J. Membr. Sci. 162 (1e2), 145e154.
 Y. Li et al. / Water Research 89 (2016) 201e209
Lameloise, M.-L., Lewandowski, R., 2012. Recovering l-malic acid from a beverage
industry waste water: experimental study of the conversion stage using bipolar
membrane electrodialysis. J. Membr. Sci. 403e404, 196e202.
Li, H., Qiu, Z., 2013. Novel technology for condensation and reclamation of the rare-
earth ammonium chloride wastewater. Mod. Chem. Ind. 33 (6), 101e105 (in
Chinese).
Madaeni, S.S., Mansourpanah, Y., 2004. Chemical cleaning of reverse osmosis
membranes fouled by whey. Desalination 161, 13e24.
Madaeni, S.S., Samieirad, S., 2010. Chemical cleaning of reverse osmosis membrane
fouled by wastewater. Desalination 257, 80e86.
Mafe
, S., Ramırez, P., Alcaraz, A., 1998. Electric field-assisted proton transfer and
water dissociation at the junction of a fixed-charge bipolar membrane. Chem.
Phys. Lett. 294 (4e5), 406e412.
Malovanyy, A., Sakalova, H., Yatchyshyn, Y., Plaza, E., Malovanyy, M., 2013. Con-
centration of ammonium from municipal wastewater using ion exchange pro-
cess. Desalination 329, 93e102.

rez-Gonz

Pe alez, A., Urtiaga, A.M., Ib

an~ ez, R., Ortiz, I., 2012. State of the art and re-
view on the treatment technologies of water reverse osmosis concentrates.
Water Res. 46, 267e283.
Quoc, A.L., Mondor, M., Lamarche, F., Makhlouf, J., 2011. Optimization of electrodi-
alysis with bipolar membranes applied to cloudy apple juice: minimization of
malic acid and sugar losses. Innov. Food Sci. Emerg. Technol. 12 (1), 45e49.
Rehouma, A., Belaissaoui, B., Hannachi, A., Muhr, L., 2013. Bipolar membrane elec-
trodialysis and ion exchange hybridizing for dilute organic acid solutions
treatment. Desalin. Water Treat. 51 (1e3), 511e517.
Rottiers, T., Ghyselbrecht, K., Meesschaert, B., Van der Bruggen, B., Pinoy, L., 2014.
Influence of the type of anion membrane on solvent flux and back diffusion in
electrodialysis of concentrated NaCl solutions. Chem. Eng. Sci. 113, 95e100.
Sayed, S.M., Toraj, M., Mansour, K.M., 2001. Chemical cleaning of reverse osmosis
209
membranes. Desalination 134, 77e82.
Tanaka, Y., 2004. Overall mass transport and solution leakage in an ion-exchange
membrane electrodialyzer. J. Membr. Sci. 235 (1e2), 15e24.
Tran, A.T.K., Mondal, P., Lin, J., Meesschaert, B., Pinoy, L., Van der Bruggen, B., 2015.
Simultaneous regeneration of inorganic acid and base from a metal washing
step wastewater by bipolar membrane electrodialysis after pretreatment by
crystallization in a fluidized pellet reactor. J. Membr. Sci. 473, 118e127.
Wang, X., Wang, Y., Zhang, X., Feng, H., Xu, T., 2013. In-situ combination of
fermentation and electrodialysis with bipolar membranes for the production of
lactic acid: continuous operation. Bioresour. Technol. 147, 442e448.
Wang, X., Li, Q., Wang, N., Wang, L., 2014a. Treatment of industrial wastewater of
ammonium chloride. Coal Chem. Ind. 37 (1), 143e145 (in Chinese).
Wang, M., Wang, K., Jia, Y., Ren, Q., 2014b. The reclamation of brine generated from
desalination process by bipolar membrane electrodialysis. J. Membr. Sci. 452,
54e61.
Wei, Y., Wang, Y., Zhang, X., Xu, T., 2011. Treatment of simulated brominated butyl
rubber wastewater by bipolar membrane electrodialysis. Sep. Purif. Technol. 80
(2), 196e201.
Xu, T., 2002. Electrodialysis processes with bipolar membranes (EDBM) in envi-
ronmental protectionda review. Resour. Conserv. Recycl. 37 (1), 1e22.
Zhan, Y., Yu, J., 2010. Cause analysis and solving countermeasure of the acid and
base consumption in the regeneration of ion exchange resin in the primary
desalting system. Ind. Water Treat. 30 (5), 81e83 (in Chinese).
Zhang, W., Zheng, F., 2013. Influence of regeneration method on regeneration de-
gree of cation exchange resin. Therm. Power Gener. 42 (3), 88e89 (in Chinese).
Zhang, K., Wang, M., Wang, D., Gao, C., 2009. The energy-saving production of
tartaric acid using ion exchange resin-filling bipolar membrane electrodialysis.
J. Membr. Sci. 341 (1e2), 246e251.