CET-vol 105


 
 
 
 
 
 
 
 
 
 
                                                                                                                                                                DOI: 10.3303/CET23105090 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Paper Received: 7 March 2023; Revised: 29 April 2023; Accepted: 25 April 2023 
Please cite this article as: Filingeri A., Culcasi A., Mueller S., Lopez J., Tamburini A., Linnartz C., Cortina J.L., Cipollina A., Micale G., 2023, An 
Experimental Investigation of Electrodialysis with Bipolar Membranes Long-run Performances with Real Solutions from Saltworks, Chemical 
Engineering Transactions, 105, 535-540  DOI:10.3303/CET23105090 
  

 CHEMICAL ENGINEERING TRANSACTIONS  
 

VOL.  105, 2023 

A publication of 

 

The Italian Association 

of Chemical Engineering 

Online at www.cetjournal.it 

Guest Editors: David Bogle, Flavio Manenti, Piero Salatino 

Copyright © 2023, AIDIC Servizi S.r.l. 

ISBN 979-12-81206-04-5; ISSN 2283-9216 

An Experimental Investigation of Electrodialysis with Bipolar 

Membranes Long-Run Performances with Real Solutions 

from Saltworks 

Antonia Filingeria, Andrea Culcasia, Sarah Muellerb, Julio Lopezc, Alessandro 

Tamburinia, Christian Linnartzb, José Luis Cortinac, Andrea Cipollinaa, Giorgio 

Micalea 
aDipartimento di Ingegneria, Università degli Studi di Palermo, Viale delle Scienze Ed. 6, 90128 Palermo, Italy 

bDepartment of Chemical Engineering, RWTH Aachen University  
cChemical Engineering Department, UPC-BarcelonaTECH, Barcelona, Spain 

andrea.cipollina@unipa.it  

Electrodialysis with Bipolar Membranes (EDBM) is a promising process to produce acidic and basic solutions 

that are important in circular treatment chains. However, there is a lack of knowledge regarding the long-run 

performance of EDBM units fed multi-ionic solutions. This study aims to fill this gap by performing a long-run 

test, via an innovative feed and bleed configuration, with real brines coming from Trapani (Italy) saltworks and 

pre-treated to remove Mg and Ca. A laboratory EDBM unit with 5 triplets with an active membrane area of 100 

cm2 was used. The results show that a steady-state concentration of 0.44 mol/L on average of OH- and H+ (in 

the form of NaOH and HCl) was reached in both acid and base outlet solutions, respectively, with a maximum 

deviation of 2 %. Stable performances were observed with a current efficiency of 69 % and 56 % and specific 

energy consumption of 1.78 kWh/kg and 2.19 kWh/kg on average for base and acid, respectively. These results 

demonstrate the potential of EDBM processes to maintain competitive and efficient performance. 

1. Introduction 

Electrodialysis with Bipolar Membranes (EDBM) has emerged as a promising technology for circular treatment 

schemes due to its ability to convert salty streams into acidic and basic streams (Tongwen, 2022). The EDBM 

stack consists of repeating units of anion exchange, cation exchange, and bipolar membranes (AEM, CEM, and 

BPM) arranged in triplets and placed between an anode and a cathode (Culcasi et al., 2022). When an electric 

field is applied, anions move toward the positive anode, while cations move toward the negative cathode. The 

selectively charged membranes control the transport of ions through them (Parnamae et al., 2021). The BPM 

generates H+ and OH- ions by allowing the water-dissociation reaction to occur at the junction of its cationic 

and anionic layers (Mareev et al., 2020).EDBM has been proposed for integration into seawater reverse osmosis 

(SWRO) desalination plants to convert brine into valuable chemicals, including HCl, NaOH, and a diluted brine. 

Chen et al. (Chen et al., 2018) investigated the use of mono-selective membranes (ACS and CIMS) by Astom 

to valorize NaCl solutions (105 g/L) in a lab-scale unit (189 cm2). Authors achieved concentrations of 1.9 mol/L 

HCl and 2.2 mol/L NaOH and evaluated the impact of competing ions, including calcium, magnesium, and 

sulphate, on the products purity. The base compartment had concentrations below 2 mg/L of calcium and 

magnesium, while the acidic compartment had 14 mg/L of sulphate, resulting in purities higher than 99.9% for 

both solutions.Herrero-González et al. (Herrero-Gonzalez et al., 2020) utilized Ralex (CM-PP and AM-PP) and 

Fumatech (Fumasep FBM) membranes in a lab-scale unit (100 cm2) to generate NaOH and HCl from NaCl 

solutions. The authors tested different constant and variable current densities ranging from 500 A/m2 to 1,000 

A/m2, with a fixed volume ratio of salt to acid/base of 20. They were able to achieve solutions containing 10.2 

wt. % HCl (~3 mol/L) and 9.7 wt. % NaOH (~2.5 mol/L), with a Specific Energy Consumption (SEC) of 43.5 

kWh/kg HCl.Song, Chae and Bang (Song, Chae and Bang, 2021) Song, Chae and Bang (Song, Chae and Bang, 

535

mailto:andrea.cipollina@unipa.it


2021) carried out tests via EDBM with a post-carbonated brine (0.9–1.2 mol/L NaCl, 74 mg/L Mg) using 

Neosepta membranes (CMB, AHA, and BP1). The tests were performed at volume ratios of 5:10:2 

(salt/acid/base), producing 1.8 mol/L NaOH. The authors reported a Current Efficiency (CE) of 48% and a SEC 

of 4.85 kWh/kg NaOH. Despite the possibility of valorizing brines, there are still significant knowledge gaps that 

hinder the large-scale application of EDBM, such as (i) the influence of multi-ionic feed solution on membrane 

performance, (ii) the lack of data on long-term operation (Blommaert et al., 2021), (iii) the study of scaling forming 

agents in terms of operational issues, and (iv) the need for developing new membranes with enhanced 

separation properties, lower costs, and reduced electrical resistance. Long-term tests have been conducted 

using Reverse Electrodialysis (RED) to produce energy from the salinity gradient (Luque Di Salvo et al, 2018, 

Cosenza et al., 2022). However, a proper strategy is needed to avoid channel clogging and membrane fouling. 

The authors recommended daily flushing in short consecutive pulses at very high flow rates to prevent the 

accumulation of matter inside the channels, alkaline backwashing to remove membrane fouling, and water 

acidification to avoid scaling. No side effects of the chemical treatments were reported. 

This study aimed to investigate the performance of EDBM in a long-run test using a real saltwork brine previously 

treated to remove magnesium and calcium. An innovative feed and bleed mode configuration with brine fed to 

both salt and base compartments was tested. Both dynamic and steady-state phases were investigated, 

focusing on the concentration profiles, SEC and CE. 

1.1 SEArcularMINE project 

This study was performed within the framework of the SEArcularMINE project, which is funded by the European 

Union (https://searcularmine.eu). The aim of this project is to develop innovative technologies that can be used 

to recover valuable minerals from the bitterns produced by traditional saltworks in the Mediterranean basin. The 

project specifically aims to recover critical raw materials (CRMs), such as Mg(II), Li(I), B(III), and trace elements 

(TEs). To achieve this goal, a circular scheme will be implemented to maximize resource efficiency, recovery 

efficiency, and economic viability. Additionally, one of the main objectives of the project is to produce the 

necessary chemicals (HCl and NaOH) on-site from exhausted brines using EDBM. 

2. Material and methodology 

A long-run experimental campaign was carried out with an EDBM unit fed by real saltwork brine to produce 

acidic and alkaline solutions at a desired target concentration. 

2.1 Material and experimental setup 

An ED-100-3-10 EDBM unit (Fumatech, Germany) was assembled using five repeating units, including 

Fumatech membranes: each repeating unit comprises three membranes (FKB-PK-130 CEMs, FAB-PK-130 

AEMs, and FBM-PK BPMs) with an active area of 0.01 m2 and three spacer-filled channels. The same type of 

AEM (i.e. FAB-PK-130 AEMs) was used as end membrane to separate the last compartments from the 

electrodes. PVC/ECTFE spacers (500 µm thick) were used to separate the membranes and create the solution 

compartments. Titanium-DSA electrodes were used as both the cathode and anode. To measure the voltage, 

platinum wires (99.9% metal basis, Alfa Aesar, 127 µm diameter) were assembled in the EDBM stack, excluding 

the fraction of the total voltage consumed at the electrodic compartments. 

The experimental campaign was conducted using real brines from Trapani saltwork, which had been pre-treated 

to remove magnesium and calcium. The pre-treatment step involved mixing the saltwork bittern with a spiked 

alkaline solution comprising NaOH micro-pearls (technical grade, Inovyn), NaCl (>99.5% purity, Saline di 

Volterra s.r.l., Italy), Na2SO4 (technical grade, CR GRUPO CRIMIDESA), and KCl (technical grade, Tor). The 

pre-treatment step resulted in the precipitation of Mg and Ca as Mg(OH)2 and Ca(OH)2, separated by 

sedimentation. The clarified solution was a 0.1 mol/L OH- solution, used as the brine feed for the EDBM unit. 

Table 1 shows the compositions of the saltwork bittern, the spiked alkaline solution, and the clarified brine. For 

electrode rinse solution (ERS), a 0.5 mol/L FeCl2/FeCl3 (ACS grade, Chemsolute, Germany) and 0.6 mol/L HCl 

(ACS Reagent 37%, Honeywell, Fluka) was used. 

Table 1. Composition of the main elements constituting the solution involved in the pre-treatment step. 

Solution 
Na+ 

[mol/L] 

K+ 

[mol/L] 

Mg2+ 

[mol/L] 

Ca2+ 

[mol/L] 

Cl- 

[mol/L] 

SO4
2- 

[mol/L] 

OH- 

[mol/L] 

Saltwork bittern 2.99 0.27 1.60 4.65 10-3 5.60 0.45 - 

Alkaline spiked solution 3.38 0.53 <lq* <lq* 3.09 0.32 0.5 

Clarified brine 3.74 0.51 <lq* <lq* 3.06 0.35 0.10 

*below quantification limit (lq) 

536



Figure 1 illustrates the experimental setup featuring a feed and bleed configuration for the acid, base, and salt 

compartments and a closed-loop configuration for the ERS. SEKO Kronos 50 FF10 peristaltic pumps were used 

to introduce the inlet solution in the feed and bleed configuration. In contrast, YT15 and BT601S (Lead Fluid, 

CO LTD, China) recirculation pumps were utilized to ensure an inlet flow rate of 75 mL/min for the salt, base, 

and acid compartments to feed the ERS solution in a closed loop. A higher ERS flow rate of 600 mL/min was 

established to promote the iron redox reactions at the electrodes. 

The clarified brine was employed as the inlet for both the base and salt compartments. Indeed, in the 

SEArcularMINE treatment chain, no pure NaOH solution is required. Softened tap water was used as the inlet 

solution for the acid compartment. Cartridge filters with a 5 µm rating were installed in all lines, downstream the 

recirculation pumps. Cewal pressure gauges were incorporated upstream the stack inlet to monitor the system 

pressure drop. 

 
Figure 1. Schematic representation of the experimental setup, including the laboratory EDBM unit and primary 

auxiliaries. 

The long-term experiment was conducted at a current density of 300 A/m2 using a DC power supply (1902B, 

B&K precision). Two multimeters (Fluke 175) were connected in series and parallel to the electric circuit to 

measure the electric current and the corresponding voltages at both platinum wires and electrodes. The 

sampling points at the outlet of the base and acid solutions were equipped with a T-connection and manual 

valves. A WTW 314i portable pH and conductivity meter was employed to monitor the evolution of the solutions 

pH and conductivity. Samples of 40 mL were collected from the outlet sampling point.  

2.2 Experimental procedure and analytical techniques 

External and internal leakage tests were performed using demineralized water in all the four compartments 

operated in closed-loop mode at the operating feed flow rates for at least one hour. The mass variation was 

monitored using precision scales (KERN KB, Max 10,000 g, d=0.1 g), and leakage was considered acceptable 

due to its value below 5% of the relative mass (1,000 g) over one hour. 

To evaluate the system long-term performance, a 75-hour feed and bleed experiment was conducted at a 

current density of 300 A/m2, aiming to achieve a steady-state OH- concentration of 0.5 mol/L. Initially, all 

compartments were filled with the appropriate inlet solution (the brine for the salt and base compartments and 

softened tap water for the acid compartment). After purging the system from bubbles, the electric current and 

inlet flow rates were set (as shown in Table 2). The latter flow rates were estimated using simplified mass 

balances, assuming a current efficiency of 70%. Specifically, a ~30% lower value of acid flow rate (13 mL/min) 

compared to the base flow rate (18 mL/min) was set to account for the different inlet concentrations of the acid 

and base streams. Indeed, the acid and base concentrations were supposed to increase from 0 mol/L to 0.5 

mol/L and from 0.1 mol/L to 0.5 mol/L, respectively. Therefore, a lower flow rate of acid was required to ensure 

a higher concentration difference between the inlet and the outlet streams.   

 

 

Acid 
in

Brine 
in

+-

Constant current
circuit

Acid 
out

Base 
out

Sample 
Base out

Sample 
Acid out

Collector salt Collector ERS

PI

PI

PI

Recirculation
pumps

PI

Feed 
pumps

Feed 
storage tank

Product
storage tank

EDBM 
stack

537



Table 2. Flow rates for acid, base, salt and ERS solutions. 

Solution Acid Base Salt ERS 

Feed and bleed [mL/min] 13 18 6 - 

Inlet to the stack[mL/min] 75 75 75 500 

 

Samples were collected every 30 minutes in the first two hours and then every hour for the first dynamic part of 

the test in which the OH- and H+ concentrations increased. The total dynamic part was completed in 6 hours as 

it can be observed in Figure 2. Then, the stability of the steady-state concentration over time was monitored by 

taking samples twice per day. The stability of inlet/outlet feed and bleed flow rates were checked twice per day 

by manual measurements of the outlet volume for 90 minutes. Voltage values at both stack and platinum wires 

were measured during the sampling procedure. Manual titration with standard solutions of 0.05 mol/L Na2CO3 

and 0.1 mol/L HCl were used to measure the acid and base concentration. Methyl Orange solution (ACS dye 

content at 85 % wt., SIGMA ALDRICH) at 10 % wt. was used as a pH indicator. Metrohm 882 compact IC was 

used for ion chromatograph analysis. Metrosep C 4. 250/4.0 and Metrosep A Supp 5 – 250/4.0 were assembled 

in the IC as cation and anion columns. The mobile eluents of the IC were 5.5 mmol/L H3PO4 for the cation 

column and a mixture of 1.0 mmol/L NaHCO3 and 3.2 mmol/L Na2CO3 for the anion column. 

2.3 Performance parameters 

The Current Efficiency (CE), expressed as a percentage, represents the proportion of applied current converted 

into the production of H+ and OH- ions in the acid and base compartments, respectively. CE can be calculated 

using Equation (1): 

CE = 
𝑧 𝐹 𝑄𝑠𝑜𝑙  (𝐶𝑠𝑜𝑙,𝑜𝑢𝑡 − 𝐶𝑠𝑜𝑙,𝑖𝑛)

3,600 𝑁 𝐼   
 100 (1) 

where 𝑧 (-) is the ion valence, 𝐹 is the Faraday constant (equal to 96,485 C/mol), 𝑄𝑠𝑜𝑙 (L/h) is the volumetric 

flow rate averaged between the inlet and outlet of the feed and bleed system, 𝐶𝑠𝑜𝑙,𝑖𝑛 and 𝐶𝑠𝑜𝑙,𝑜𝑢𝑡 (mol/L) indicate 

the solution concentrations of acid or base at the inlet and outlet of the feed and bleed system, respectively, 𝑁 

is the number of triplets, and 𝐼 (A) is the applied current. 

The Specific Energy Consumption (SEC), expressed in kWh/kg, represents the energy required by the EDBM 

unit to produce 1 kg of product (HCl or NaOH). SEC can be calculated using Equation (2): 

SEC = 
I U

𝑀𝑤 𝑄𝑠𝑜𝑙  (𝐶𝑠𝑜𝑙,𝑜𝑢𝑡 − 𝐶𝑠𝑜𝑙,𝑖𝑛)  
 (2) 

where U (V) is the voltage at the platinum wires, and 𝑀𝑤 (g/mol) is the molecular weight (for HCl or NaOH). The 

use of the voltage at the platinum wires allows for the exclusion of the contribution of the electric compartments, 

which can have a high impact on the total voltage of a laboratory-scale unit, as in the present case, while they 

can be negligible at the pilot or industrial scale. 

3. Result and discussion 

Figure 2a illustrates the concentration profiles of OH- and H+ ions as a function of time. At the beginning of the 

test, the concentration profile increased from 0.1 mol/L and 0 mol/L for OH- and H+, respectively, up to a steady-

state value. Initially, a large portion of the solution exiting the stack was recirculated and mixed with the low 

concentrated inlet solution of the feed and bleed system. This caused the concentration of the solution entering 

the stack to increase, raising the outlet concentration to a steady-state condition. The dynamic part of the test 

lasted for about three hours. After that, the acid and base streams reached an average concentration of 0.44 

mol/L, slightly lower than the expected 0.5 mol/L target. The concentration was kept constant with a maximum 

deviation of 2% for both solutions. Importantly, these steady-state product concentrations were obtained using 

two different values of the inlet flow rates for the acid and base solutions as mentioned in Section 2.  

Figure 2b shows the CE profile during time for both the base and acid compartments. Similarly to the 

concentration profile, CE increased during the dynamic part of the test due to the increase in the outlet 

concentration. In the steady-state regime, CE was 61–75 % and 49–61 % for the base and acid compartments, 

respectively. The relatively low efficiency could be attributed to non-ideal phenomena such as diffusion and 

partial neutralization of the produced solution. Additionally, the feed and bleed configuration forces the system 

to constantly operate at the target concentration, favouring non-ideal phenomena that are typically the higher, 

the higher the concentration. The higher CE for the base compared to the acid can be explained by the larger 

mobility of H+ compared to OH-.  

538



This observation was supported by the transport of H+ towards the salt compartment, which neutralized and 

acidified the slightly alkaline brine, used as the inlet solution of the salt compartment, from pH 13 (0.1 mol/L OH- 

concentration as reported in Table 1) to pH 1.1 (i.e., H+ concentration of 0.08 mol/L). 

Figure 2c reports the SEC trend during the time referred to NaOH or HCl products. In the dynamic part of the 

test, a significant reduction of SEC was observed, which could be attributed to a reduction of the voltage 

measured at the platinum wires, referred to the 5 triplets, from 13.2 V up to 8.5 V, due to the initial increase in 

acid conductivity, and an increase in the efficiency with which NaOH or HCl were produced. The SEC trend 

mirrored the CE behaviour, and higher values of SEC were always observed for the acid, which showed lower 

CE than the base. In the second part of the test, SEC was 1.56–1.84 kWh/kg and 1.94–2.35 kWh/kg for NaOH 

and HCl products, respectively.It is worth noting that in the acid solution, Cl- was the major anion present as it 

represented ~99 % of the total anions in the solution, while SO4
2- reached a maximum concentration of 0.01 

mol/L. Conversely, apart from the H+ produced via water dissociation in the BPM, Na+ and K+ were present at 

0.09 mol/L and 0.01 mol/L, respectively, as they diffused from both the saline and alkaline channels. 

Overall, Figure 2 shows stable performances for all investigated parameters. No scaling or fouling effects were 

observed, when the stack was inspected, and the pressure drop remained stable. 

 

Figure 2. a) Concentration, b) Current Efficiency and c) Specific Energy Consumption as functions of time for 

both acid and base solutions. A different abscissa was used to magnify the long-run test first hours, where the 

system dynamic behaviour was observed. 

4. Conclusions 

In this study, a laboratory-scale EDBM unit was tested in feed and bleed operation mode for 75 hours using pre-

treated real brine from a saltwork. A new configuration for the EDBM was presented that utilized the 

concentrated pre-treated brine as feed for both salt and base compartments. 

During the transient start-up, which lasted approximately three hours, the acid and base concentrations 

increased to a steady-state value of 0.44 mol/L on average, with a maximum deviation of 2 %. CE reflected the 

ion concentration trend, with an initial increase up to 61–75 % and 49–61 % for base and acid, respectively. The 

lower CE of the acid compartment can be attributed to the high mobility of H+ ions, which partially diffused 

towards the salt compartment, resulting in the acidification of the salty stream from pH 13 to pH 1.1 (i.e., 0.08 

mol/L H+). 

a) b)

c)

Base

Acid

C
u

rr
e

n
t

Ef
fi

ci
e

n
cy

[%
]

C
o

n
ce

n
tr

at
io

n
s

[m
o

l/
L]

Sp
e

ci
fi

c
En

e
rg

y 
C

o
n

su
m

p
ti

o
n

[k
W

h
/k

g]

Time [h] Time [h]

Time [h]

539



SEC reduced during the transient start-up of the test due to increased production and lower voltage, resulting 

into steady-state values of 1.56–1.84 kWh/kg and 1.94–2.35 kWh/kg for NaOH and HCl products, respectively. 

Importantly, no performance loss (e.g. pressure drop increases) were recognized during the long-run. 

Overall, this study demonstrates the potential of EDBM units to be integrated into circular treatment chains, 

producing continuous acidic and alkaline solutions while maintaining stable performance. 

Nomenclature

AEM – anion exchange membrane 

BPM – bipolar membrane 

CEM – cation exchange membrane 

EDBM – Electrodialysis with bipolar membrane 

RED – reverse Electrodialysis 

SWRO – seawater reverse osmosis 

sol - solution 

in – inlet of the feed and bleed system 

out – outlet of the feed and bleed system 

C – molar concentration, mol/L 

CE – current efficiency, % 

F – Faraday constant, C/mol 

I – current intensity, A 

N, number of triplet, - 

Mw – molecural weight, g/mol 

Q – volumetric flow rate, L/h 

SEC – specific energy consumption, kWh/kg 

U – voltage, V 

z – ion valence, - 

 

Acknowledgments 

This project has received funding from the European Union’s Horizon 2020 research and innovation program 

under grant agreement no. 869467 (SEArcularMINE). This output reflects only the author’s view. The European 

Health and Digital Executive Agency (HaDEA) and the European Commission cannot be held responsible for 

any use that may be made of the information contained therein.  

The work of J. López was supported within the scope of Margarita Salas fellowship, financed by the Ministerio 

de Universidades (Spain) and European Union – NextGenerationEU. 

References 

Blommaert, M. A. et al. (2021) ‘Insights and Challenges for Applying Bipolar Membranes in Advanced 

Electrochemical Energy Systems’, ACS Energy Letters, 6, pp. 2539–2548. doi: 

10.1021/acsenergylett.1c00618. 

Chen, B. et al. (2018) ‘Selectrodialysis with bipolar membrane for the reclamation of concentrated brine from 

RO plant’, Desalination. Elsevier B.V., 442, pp. 8–15. doi: 10.1016/j.desal.2018.04.031. 

Cosenza, A. et al. (2022) ‘Power Production from Produced Waters via Reverse Electrodialysis: A Preliminary 

Assessment’, Energies, 15(11). doi: 10.3390/en15114177. 

Culcasi, A. et al. (2022) ‘A comprehensive multi-scale model for bipolar membrane electrodialysis (BMED)’, 

Chemical Engineering Journal. Elsevier B.V., 437(P1), p. 135317. doi: 10.1016/j.cej.2022.135317. 

Herrero-Gonzalez, M. et al. (2020) ‘Highly concentrated HCl and NaOH from brines using electrodialysis with 

bipolar membranes’, Separation and Purification Technology. Elsevier B.V., 242, p. 116785. doi: 

10.1016/j.seppur.2020.116785.https://searcularmine.eu, accessed in March 2022.  

Luque Di Salvo, J. et al. (2018) ‘Long-run operation of a reverse electrodialysis system fed with wastewaters’, 

Journal of Environmental Management. Elsevier Ltd, 217, pp. 871–887. doi: 

10.1016/j.jenvman.2018.03.110. 

Mareev, S. A. et al. (2020) ‘A comprehensive mathematical model of water splitting in bipolar membranes: 

Impact of the spatial distribution of fixed charges and catalyst at bipolar junction’, Journal of Membrane 

Science. Elsevier B.V., 603(December 2019), p. 118010. doi: 10.1016/j.memsci.2020.118010. 

Pärnamäe, R. et al. (2021) ‘Bipolar membranes: A review on principles, latest developments, and applications’, 

Journal of Membrane Science, 617. doi: 10.1016/j.memsci.2020.118538. 

Song, K., Chae, S. C. and Bang, J. H. (2021) ‘Separation of sodium hydroxide from post-carbonation brines by 

bipolar membrane electrodialysis (BMED)’, Chemical Engineering Journal. Elsevier B.V., 423(May), p. 

130179. doi: 10.1016/j.cej.2021.130179. 

Tongwen, X. (2002) ‘Electrodialysis processes with bipolar membranes (EDBM) in environmental protection - A 

review’, Resources, Conservation and Recycling, 37(1), pp. 1–22. doi: 10.1016/S0921-3449(02)00032-0 

  

540


	111filingeri.pdf
	An Experimental Investigation of Electrodialysis with Bipolar Membranes Long-Run Performances with Real Solutions from Saltworks