Transactions of Tianjin University (2021) 27:202–216

https://doi.org/10.1007/s12209-021-00285-9

REVIEW

Revisiting Chlor‑Alkali Electrolyzers: from Materials to Devices

Kai Li1,2 · Qun Fan1,2 · Hongyuan Chuai1,2 · Hai Liu1,2 · Sheng Zhang1,2 · Xinbin Ma1,2

Received: 6 January 2021 / Revised: 7 February 2021 / Accepted: 17 February 2021 / Published online: 13 April 2021
© The Author(s) 2021

Abstract
As an energy-intensive industry, the chlor-alkali process has caused numerous environmental issues due to heavy electric-
ity consumption and pollution. Chlor-alkali industry has been upgraded from mercury, diaphragm electrolytic cell, to ion
exchange membrane (IEM) electrolytic cells. However, several challenges, such as the selectivity of the anodic reaction, slug-
gish kinetics of alkaline hydrogen evolution, degradation of membranes, the reasonable design of electrolytic cell structure,
remain to be addressed. For these reasons, this paper mainly reviews the research progress of the chlor-alkali industry from
materials to devices, including hydrogen evolution anode, chlorine evolution cathode, IEM, and electrolytic cell system.
Finally, the research directions and prospects in the chlor-alkali industry are proposed for its further improvement.

Keywords  Chlor-alkali · Process · Hydrogen energy · Ion exchange · Membrane · Hydrogen evolution · Chlorine evolution

Introduction emission reduction in the chlor-alkali production process are

important development directions.
The chlor-alkali process is one of the most basic chemi- The chlor-alkali industry has experienced a long devel-
cal industries, mainly producing chlorine (­ Cl2) and sodium opment process from mercury, diaphragm electrolytic cells
hydroxide (NaOH). The global annual production of chlo- to ion exchange membrane (IEM) electrolytic cells. At
rine exceeds 75 million tons. Each ton of chlorine consumes present, membrane electrolytic cell technology accounts
about 2200–2600 kW∙h of electricity, and the global chlor- for approximately 81% of the global chlor-alkali capacity
alkali industry needs to consume over 150 TW∙h of electric- [4]. In mercury electrolysis cells, mercury as the liquid
ity every year, accounting for about 10% of global electric- cathode reacts with sodium ions to form sodium amalgam
ity [1]. Chlorine has been used in a variety of applications, ­(Na+  + Hg + ­e− → ∙NaHg), whereas C ­ l2 is formed on the
including the production of building materials such as poly- anode. Then, NaHg is transferred to another electrolytic cell
vinyl chloride, organic synthesis, metallurgy, water treat- to release N­ a+ (NaHg → ­Na+  + Hg + ­e−), forming NaOH in
ment [2], and the manufacture of titanium dioxide [3]. NaOH the electrolyte [5, 6]. A diaphragm is introduced to sepa-
is also a common chemical raw material that is widely used rate the anode and cathode reactions to prevent the cathode
in the production of detergents, herbicides, pesticides, medi- product (NaOH) and the anode product ­(Cl2) from cross-
cines, plastics, and soaps. However, the current chlor-alkali ing over to generate sodium hypochlorite. The negatively
process is one of the industries with high energy consump- charged membrane can inhibit the back diffusion of ­OH− [7].
tion, releasing large amounts of pollutants and causing seri- Therefore, in diaphragm electrolyzers, asbestos felt is often
ous environmental problems. Therefore, energy saving and used to separate the products ­(Cl2 and ­H2) [8, 9], although
the chemical instability of asbestos itself causes the severe
swelling of pure asbestos diaphragm under a high current
* Hongyuan Chuai load [7]. Therefore, these two technologies above rely on
[email protected]
highly toxic mercury and asbestos respectively, resulting in
* Sheng Zhang serious environmental pollution [1, 10].
[email protected]
Membrane electrolysis cells use highly conductive IEMs
1
Collaborative Innovation Centre of Chemical Science instead of traditional asbestos felt to separate C ­ l2 and H
­ 2,
and Engineering, Tianjin University, Tianjin 300072, China which greatly reduces the operating voltage and pollutant
2
School of Chemical Engineering and Technology, Tianjin emissions. Compared with diaphragm electrolytic cells, the
University, Tianjin 300072, China

13
Vol:.(1234567890)
Revisiting Chlor‑Alkali Electrolyzers: from Materials to Devices 203

membrane counterpart offers high-purity sodium hydrox- former technology is notably higher than that of the oxygen
ide. Thus, the replacement of the diaphragm method with gas produced in the latter. Further coupled with cheap and
ion membrane technology is an inevitable trend. The per- clean electricity generated from renewable energy, chlor-
fluorinated membrane method is currently recognized as the alkali electrolyzers will become a promising green hydrogen
most energy-efficient chlor-alkali process in the world. A production technique.
cation-selective permeation membrane is placed between the In general, all chlor-alkali reactions include three reac-
anode and cathode compartments, and a saturated sodium tions: chlorine evolution reaction (CER) at the anode, hydro-
chloride solution is injected into the anode compartment, gen evolution reaction (HER) at the cathode, and NaOH
which usually produces 32%–35% caustic soda (Fig. 1). This generation in the electrolyte. Figure 1 shows a chlor-alkali
membrane rejects the passage of chloride ions (negatively electrolysis cell. The chlor-alkali electrolyzer is mainly com-
charged) but allows sodium ions (positively charged) to pass posed of three parts: anode, cathode, and IEM. Although the
through. The first-generation membrane can only work at energy consumption of state-of-the-art chlor-alkali electro-
low corrosive concentrations, whereas the state-of-the-art lyzers has been greatly reduced with technological progress,
membrane is made of a perfluorosulfonic acid (PFSA) pol- all of its structural parts still need to be further optimized.
ymer layer, where a polytetrafluoroethylene (PTFE)-rein- An in-depth understanding of the selectivity of the anode
forced fabric and a perfluorocarboxylic acid (PFCA) poly- will help to further develop inexpensive, stable, and efficient
mer are bonded together. This membrane has low electrical chlorine evolution catalysts. The slow hydrogen evolution
resistance and high ion-selective permeability and can work kinetics caused by the alkaline condition of the cathode
in a highly corrosive electrolyte environment. is one of the main reasons for its high energy consump-
Hydrogen, which is usually treated as exhaust gas in tion. Therefore, the research on the alkaline HER cathode
the chlor-alkali industry, may become the major fuel used of chlor-alkali electrolysis cells is important to improve the
on a large scale in the future as the most promising clean efficiency of electric energy utilization. In addition, the use
energy. The chlor-alkali electrolyzer can be used as one of of oxygen depolarized cathodes (ODCs) to reduce the energy
the most promising low-cost hydrogen sources in the near consumption of chlor-alkali electrolysis cells has achieved
term, better than water electrolyzer, considering that the certain results, but their potential remains to be evaluated.
value of chlorine gas and chemical NaOH produced in the An in-depth understanding of IEMs can not only guide spe-
cific industrial operating conditions and optimize the overall
structural design of electrolytic cells to extend the service
life of the membrane but also provide a basis for the design
of high-performance membrane materials. In addition to the
various structural units of the chlor-alkali electrolysis cell,
the entire electrolysis system can still be optimized.

Cathode

In conventional chlor-alkali electrolyzers, HER occurs at the

cathode, which can be employed as one of the most promis-
ing low-cost, green hydrogen-making technology. HER is
complicated in the strong alkaline solution in chlor-alkali
electrolyzers compared with the reaction in acidic solutions.
First, the cathode materials need to withstand the highly
corrosive alkaline conditions. Moreover, the slow kinetics
of HER under alkaline conditions is the main reason for
the high energy consumption. Thus, the design of highly
efficient electrocatalysts under strongly basic conditions is
the key to achieving low-energy chlor-alkali electrolyzers
for green hydrogen production.

Hydrogen Energy

Fig. 1  Schematic of the working principle for a renewable electricity- Hydrogen is a clean fuel and is one of the most promising
powered chlor-alkali electrolyzer energy carriers to replace traditional fossil fuels, possessing

13

204 K. Li et al.

high heat value (120 MJ/kg) and no exhaust emissions (only steam production; (6) hydrogen fuel cell technology for the
produces water as by-products) at the point of use. Currently, construction of hydrogen power plants [24]. Several com-
hydrogen gas is mainly produced through petrifaction, such panies have bottled and marketed hydrogen gas produced in
as the gasification of coal [11], natural gas steam reforming chlor-alkali plants, which is then supplied to fuel cells [25].
[12], and methanol cracking [13], with about 4% of global Hydrogen fuel cells can be well coupled with chlor-alkali
hydrogen produced from water electrolysis [14]. plants, which can recover 20% of the electrical energy and
In a typical gasification process, coal is first gasified and 10% of the thermal energy consumed in chlor-alkali elec-
then undergoes carbon monoxide vapor conversion and pres- trolysis [26].
sure swing adsorption purification to produce hydrogen.
This process is less preferred due to the high temperature HER Fundamentals
required, the pollution created, and complex operation. The
most common method used in hydrogen production is the Historically, HER is the most studied, particularly in acidic
methane steam reforming process, which takes up 50% of solutions. By contrast, limited research has been conducted
the hydrogen produced worldwide [15]. This process is simi- on HER in alkaline solutions, that is, the operation condition
lar to the gasification of coal, and thus, the drawbacks are the of chlor-alkali electrolyzers, mainly due to the complicated
same. Despite the low cost of hydrogen production from fos- reaction mechanism. The possible pathway of HER in highly
sil fuels, the heavy carbon footprint and large investment in alkaline media includes the following steps: electro-reduc-
hardware equipment make them unattractive for future use. tion of water molecules and hydrogen adsorption (Volmer
Three ways are used to produce hydrogen from methanol: reaction, Step 1); electrochemical desorption of hydrogen
methanol cracking, methanol-steam reforming, and partial (Heyrovsky reaction, Step 2); chemical desorption (Tafel
oxidation of methanol [13]. Among these methods, methanol reaction, Step 3).
cracking is the most competitive due to its wider applica-
tion range and single feedstock. Its advantage includes lower M + H2 O + e− ↔ MHads + OH− . (1)
investment cost and less energy consumption than coal and
natural gas. However, the cost of methanol feedstock is high, MHads + H2 O + e− ↔ M + OH− + H2 . (2)
which results in a high unit cost of hydrogen production.
As for water electrolysis, the hydrogen production cost 2MHads ↔ 2M + H2 . (3)
is the highest among different methods mainly due to the
sluggish oxygen evolution reaction (OER) [16], which is HER kinetics in alkaline solution is extremely sluggish
estimated to be between 2.5 and 3.5 ¥/Nm3 hydrogen at the and is usually 2–3 orders of magnitude lower than that of
current price of electricity. Water electrolyzers can become acidic media. The pH-dependent hydrogen binding energy
a common method to produce hydrogen if the electricity (HBE) derived by cyclic voltammetry is considered a rea-
price is decreased considerably by maturing wind and hydro- sonable descriptor. HBE is generally considered to be an
electric power supplies [17]. The chlor-alkali electrolyzer is inherent property of every metal and is independent of pH.
potentially one of the most promising low-cost hydrogen- However, the electrode surface is covered with water. Thus,
making technology, better than water electrolyzer, consider- the adsorption/desorption of hydrogen is possibly accompa-
ing that the value of chlorine gas and chemical NaOH pro- nied by the desorption/resorption of water. Cheng et al. [27]
duced in the former technology is substantially higher than attributed pH dependence to the changes in water adsorp-
the oxygen gas produced by the latter [18, 19]. tion energy, given that the hydrophobicity of Pt electrodes
In modern chlor-alkali plants, unlike the products, increases as the potential becomes negative. Therefore,
namely, chlorine and sodium hydroxide, hydrogen is a by- water adsorption or surface hydrophilicity must be tuned by
product. In the USA, approximately 40%–50% of hydrogen alloying or introducing hydrophilic surface groups (such as
from the chlor-alkali process is combusted for process heat, OH). In addition, several uncertain properties of the inter-
with 30%–40% sold to the merchant hydrogen market and face, such as the Schottky barrier [28] or mismatched lat-
the remainder (10%–30%) vented [20]. In China, nearly tice space [29], can hinder electron transfer and the transfer
1.8 billion m­ 3 hydrogen is wasted in industrial chlor-alkali of reaction intermediates between the two phases, causing
process [21]. Currently, the main applications for hydrogen additional kinetic obstacles.
produced in chlor-alkali electrolyzers include: (1) feedstock
gas for the production of ammonia [22]; (2) hydrocracking HER Catalysts
and other petroleum processing industries [23]; (3) chemi-
cal and organic fine chemical synthesis; (4) production of So far, platinum offers the most superior HER performance,
high-purity compressed hydrogen for direct sale (for elec- but its limited abundance hinders its industrial applications.
tronics, metallurgy, glass, and other industries); (5) fuel for In practical chlor-alkali electrolyzers, low-carbon steel is

13
Revisiting Chlor‑Alkali Electrolyzers: from Materials to Devices 205

employed as cathode material instead of platinum. Two main of N atoms can effectively promote electron transfer from Re
issues need to be addressed with the use of low-carbon steel. to S atoms, leading to the redistribution of electrons, render-
First, during the electrolysis process, the cathode surface ing Re atoms in an electron-deficient state, and increasing
becomes gradually covered by chromium oxide ­(Cr2O3), the hydrophilicity of O. The preferential sites of adsorption
resulting in “blind current” loss. This condition will result in activation optimize the H adsorption/desorption process,
a large increase in overpotential and ultimately the precipita- increase the electronic state near the Fermi level of the mate-
tion of Ca and Mg products. Further, the steel cathode has a rial, and improve the charge transfer.
high corrosion rate due to the thermodynamic instability of
iron in the alkaline solution. Nickel is also the most common ODC
catalyst used as cathode material in alkaline media, such as
Ni foam. The complete replacement of platinum with low- One of the important developments in the chlor-alkali indus-
cost active catalytic materials is one of the most important try is the introduction of ODC, which introduces oxygen
goals of modern electrocatalysis. into the electrolysis chamber through a porous cathode.
Meanwhile, numerous new materials have also been Figure 3 shows the ODC electrolysis cell. If the oxygen
investigated for HER: (i) noble metals, such as Pd, Ir, Ru, reduction reaction (ORR) [35] at the ODC cathode is used
Ag, etc.; (ii) inexpensive transition metal catalysts, includ- to replace the HER in the traditional process, the electrical
ing Fe, Co, Ni, Mn, Cu, Mo, and W, and their heterogamous energy demand can theoretically be reduced by up to 30%
nanostructures; (iii) non-metal catalysts, including B, C, [36], and the operating voltage can be reduced by approxi-
N, P, and S [30]. These results have been summarized in mately 1.0 V, as shown in Eqs. (4) and (5). The ODC process
the previous references [31, 32]. Nitrogen-doped transition has also been widely used in the HCl electrolysis scenario,
metal materials exhibit a tunable local electronic structure which consumes downstream HCl by-products to achieve
and fast electron transfer capability. Lei et al. [33] devel- chlorine recovery [37].
oped an atomically dispersed Ni-anchored nitrogen-doped
HER ∶ 2H2 O + 2e− → H2 + 2OH− . (4)
porous carbon (Ni–NC) matrix, which gave a minimum
overpotential of ~ 150 mV at a current density of 10 mA/
cm2. Experimental observations confirmed that the strong E0 = 0.00 V vs. RHE
chemical coupling between Ni nanoparticles and Ni–NC
regulates the electronic structure, promotes electron transfer ORR ∶ 2H2 O + O2 + 4e− → 4OH− . (5)
on the constructed interface, and optimizes the process of
hydrolysis and O ­ H− adsorption. Nitrogen doping can effec-
tively improve the alkaline HER activity. This atomic-level E0 = 1.23 V vs. RHE
modulation strategy opens up a new way for the develop-
ment of practical electrocatalysts using three-dimensional
metal oxides in alkaline environments. Sun et al. [34] used
N atoms to regulate R ­ eS2. As shown in Fig. 2, the injection

Fig. 2  Schematic of the HER mechanism of a ­ReS2 and b N-ReS2 under alkaline conditions (reprinted from [34] with permission from RSC)

13

206 K. Li et al.

Fig. 3  Comparison of chlor-alkali electrolysis with different active cathodes. a Hydrogen evolving cathode. b ODC (reprinted from [38] with
permission from ACS)

ODC Electrode Structure and Working Principle and thus, decreases the abundance of free water molecules
[47]. The decrease in water activity and oxygen solubility is
ODC is a gas diffusion electrode (GDE) that can significantly also accompanied by the decrease in oxygen mass transport
increase the reaction current densities [39]. The industrial rate [48].
ODC consists of a porous structure of silver particles and Research has focused on the correlation between the
a PTFE supporting layer. The typical operating conditions structural parameters of the porous gas diffusion layers
include 10 M NaOH, temperature range from 80 °C to 90 °C (GDLs) and ORR performances. GDLs with large pores may
[40], and current density of 400–600 mA/cm2 [41]. promote gas supply and further facilitate the mass transport
The ODC performance is related to the diffusion rate of rate of oxygen to achieve an excellent electrode performance
oxygen and water activity [42]. Under industrial operating [49, 50]. Pore volume is another factor that determines the
conditions, ORR only occurs near the gas–liquid interface electrode performance [51]. Nara et al. [52] employed elec-
[35, 43]. Given the water consumption and hydroxide ion trochemical impedance spectroscopy to further determine
accumulation, the phase balance changes significantly with the correlation between the ODC performance and the pore
the current density applied and oxygen solubility in the liq- structures, including the primary and secondary pores within
uid electrolyte. This effect is enhanced at high reaction rates, the catalyst layer. Park and Popov [53] studied the influ-
which leads to complete depletion of oxygen and achieve- ence of PTFE content of ODC on electrode activity and
ment of a limiting current. Röhe et al. [44] established a observed the volcanic dependence of PTFE content in the
dynamic three-phase model of advanced chlor-alkali elec- ODC performance.
trolysis porous ODC, which proved that water transport,
especially that of hydroxide ions in the liquid electrolyte, ODC Electrode Materials
is the overall limiting factor. The oxygen concentration is
rapidly consumed near the gas–liquid interface [45]. Con- 1. Carbon-supported transition metal catalysts
sidering the non-ideal behavior of highly concentrated   Pt/C (usually 20 wt%) is a commonly used catalyst
water, the water activity coefficient largely depends on ion in fuel cells and has good activity and stability in acidic
concentration [46]. As sodium hydroxide concentration media [54–56]. However, metallic Pt is easily corroded
increases, water activity will decrease due to the formation by Cl ions; thus, a large amount of noble Pt will dissolve
of a solvated shell that combines water molecules with ions. in the highly corrosive saturated ­Cl2 electrolyte [57].
For example, the highly concentrated electrolyte binds con- The carbon-based rhodium-based chalcogenide R ­ hxSy/C
siderable water to the solvation shell of solute molecules developed by E-TEK is the only ODC catalyst that main-

13
Revisiting Chlor‑Alkali Electrolyzers: from Materials to Devices 207

tains activity under concentrated C ­ l2 electrolyte, but electrolyte infiltration and distribution during electrolysis.
this catalyst has low intrinsic activity [58]. Most of the The results showed a potential-dependent electrolyte trans-
existing carbon materials are dual-functional catalysts port in the ODC and droplet formation on the gas diffusion
for ODC and chlorine evolution in hydrochloric acid side. Given the importance of determining the electrolyte
electrolysis, such as ­CNx [59], Co-NSCx [60], Fe–N– distribution inside the porous structure, further research on
C [58], etc. The main goal is to improve the tolerance in situ imaging technology is needed to accurately link the
to Cl ions. However, overall activity still needs to be structural and electrochemical performances.
improved.
2. Silver catalysts
  Under the harsh operating conditions of chlor-alkali Anode
electrolysis, carbon materials are easily corroded and
suffer from low stability. Alternatively, silver exhibits Anode Reaction
an oxygen reduction activity similar to that of platinum
under strongly alkaline conditions [61]; thus, silver can In chlor-alkali electrolyzers, CER is the desirable reaction
be used as a suitable substitute for noble platinum cata- at the anode. The current efficiency of chlorine is the main
lysts. measure of anode selectivity, whereas the OER is a side
reaction that competes with CER. OER occurs directly at the
Optimizing silver-based GDEs will reduce energy con- anode, where water is oxidized to generate oxygen [16, 63].
sumption and enhance the long-term stability of chlor-alkali The selectivity of chlorine and oxygen evolution is related
electrolyzers. GDEs are composed of silver particles, a com- to the inherent selectivity of anode materials, process condi-
plex pore system around them, and PTFE at the boundaries tions, and electrolyte compositions.
of the silver particles [62]. Figure 4 shows the three-phase The anode reaction is mainly studied by conventional
interface and GDE electrode. When the hydrophilic silver electrochemical techniques (such as voltammetry and polari-
particles support the penetration of the electrolyte, the pore zation techniques), but theoretical methods are also used
channels containing hydrophobic PTFE keep the gas phase to study the intrinsic selectivity [13–15]. The composition
unobstructed. The existing GDE electrode is improved of the evolved gas is usually determined by gas chroma-
mainly through the increase in the three-phase interface (cat- tography or O­ 2 absorption method, such as that which uses
alyst material, alkaline electrolyte, and ­O2-rich gas phase), pyrogallol. The concentrations of hypochlorite, chloride, and
thereby increasing the utilization rate of the catalyst. How- chlorate can be determined by titration of the electrolyte.
ever, adjusting the three-phase boundary can be a complex In addition, differential electrochemical mass spectrometry
task because the distribution of the electrolyte in the porous can be used to measure the reaction products formed on the
structure is still unknown. To address this issue, Paulisch anode with high sensitivity [64–66].
et al. [48] used X-ray computed tomography to study the
CER Fundamentals

Equations (6) and (7), respectively, list two competing reac-

tions on the anode: CER and OER. The four-electron trans-
fer in OER is thermodynamically more favorable and can
compete with CER, although the two-electron transfer CER
exhibits faster kinetics [67]. The excess oxygen and hydro-
gen products can possibly be combined in the cell to form an
explosive gas mixture, which renders oxygen a safety hazard
[68]. Therefore, chlorine selectivity must be improved.

2H2 O → O2 + 4H+ + 4e− . (6)

EO0 /H O = 1.229 V vs. RHE

2 2

2Cl− ↔ Cl2 + 2e− . (7)

0
Fig. 4  Schematic of three-phase interface and GDE structure ECl /Cl−
= (1.358 + 0.059pH) V vs. RHE
2
(reprinted from [62] with permission from RSC)

13

208 K. Li et al.

Numerous factors, including electrolytes, anode mate- innovation of DSAs is called “one of the greatest technologi-
rials, and operating conditions, such as current densities cal breakthroughs in the field of electrochemistry in the past
and temperature, affect the selectivity of chlorine [68, 70]. 50 years” [72]. Commercial DSA electrodes usually contain
Acidic conditions are favorable for CER, whereas alkaline one or more other doping materials, such as ­TiO2, ­SnO2,
conditions are favorable for OER [68]. The pH correlation ­CoOx, and S ­ bOx, to reduce the amount of precious Ir or Ru
between OER and CER can be represented by the phase [73–75]. Modern DSAs can work at current industrial densi-
diagram of the catalyst surface intermediates in equilibrium ties for more than 10 years [72]. The surface of commercial
with ­Cl− (Fig. 5). Unlike the OER, which is a pH-dependent DSAs is coated with an electrocatalytic layer containing a
reaction, CER is pH independent. Furthermore, increasing mixture of titanium dioxide and ruthenium dioxide with an
the current density can increase chlorine selectivity, which atomic ratio of about 70: 30 and a CER selectivity of about
is related to the suppression of OER in low local pH. 95% [76].
To investigate the CER process at the DSA electrode,
Anode Electrode Material Hansen et al. [69] constructed the Pourbaix diagram of ­IrO2
and ­RuO2 based on density functional theory (DFT) calcu-
Electrode materials with high activity for chlorine are usu- lations. The electrochemical and thermodynamic methods
ally also highly active for oxygen. Although these competing were used to construct the linear relationship of various
reactions are governed by the same factors, chlorine selec- adsorption intermediates, and a possible reaction mechanism
tivity can still be optimized. The two reactions occur at a was deduced. All material dependencies were included in
similar active site or form a common surface intermediate one descriptor, that is, the oxygen-binding energy (Fig. 6).
species, whereas CER and OER produce two competing The unit surface has two bridge sites and two cus sites. The
intermediate products, namely, OCl and OOH, respectively binding of adsorbates at a bridge site is stronger than that at
[71]. a cus site. Thus, most bridge sites are occupied by oxygen.
The phase diagram of ­IrO2 shows that the surface is covered
Commercial Dimensionally Stable Anode (DSA) Electrodes by OH groups at a low potential, which increases the oxida-
­ H− in the solution. The OH adsorbed on
tion potential of O
DSAs, which are mainly composed of noble metal Ir or Ru the bridge site is first converted to O and finally to the ther-
oxides, are currently the most advanced CER catalysts. The modynamically favorable OOH. The direct formation of Cl
adsorbate on the cus site requires a pH less than 3, whereas
the bridge site requires an even lower pH. For a high-perfor-
mance chlorine evolution catalyst, the free energy of forma-
tion of Cl adsorbed species is approximately 0 at a potential
close to 1.36 V. The direct binding of Cl adsorbate at Ir fails
to meet these conditions, whereas the intermediate ­ClOc is
thermodynamically favorable at a potential higher than 1.5 V
in the pH range from 0 to 3. Thus, the CER proceeds in two
steps. ­Oc combines with ­Cl− to form a ­ClOc intermediate,

Fig. 5  Surface phase diagram for ­IrO2(110) in equilibrium with ­Cl−,

­H+, and ­H2O at 298.15  K and aCl− = 1. The adsorbate phases are
shown in the insets. Ir atoms are cyan, O atoms are red, H atoms are Fig. 6  Most stable surface at pH = 0 and aCl− = 1 as a function of
white, and Cl atoms are green (reprinted from [69] with permission potential, U, and the surface reactivity descriptor, ΔE(Oc) (reprinted
from RSC) from [69] with permission from RSC)

13
Revisiting Chlor‑Alkali Electrolyzers: from Materials to Devices 209

which combines with another ­Cl− to form ­Cl2. This mecha- through the Volmer–Tafel mechanism, where the activity
nism is also known as the Volmer–Heyvrosky mechanism. was strongly affected by the concentration of chloride ions,
The surface phase diagram of ­RuO2 is complicated. A pos- showing a stepwise negative reaction order.
sible mechanism of the CER is that O ­ 2cc combines with Highly efficient non-noble metal catalysts have also been
­Cl to form a Cl(O )2 intermediate, and Cl(Oc)2 combines
− c
developed to replace expensive Ru and Ir catalysts. Among
with another C ­ l− to form O
­ c and C ­ l2, that is, the Khrishtalik these catalysts, crystalline transition metal antimonates
mechanism. (TMAs) are active and stable electrocatalysts [83]. Moreno-
Exner et  al. [77] used the ab  initio thermodynamic Hernandez et al. [84] reported that crystalline TMAs, such
method to construct the surface Pourbaix diagram of the as ­NiSb2Ox, ­CoSb2Ox, and ­MnSb2Ox, are intermediate active
CER on R ­ uO2(110) based on the DFT calculations and con- catalysts for the electrochemical oxidation of chloride to
sidered the solvation effect. Under CER conditions, that is, chlorine under industrial chlor-alkali process conditions. The
U > 1.36 V, no ORR occurs on the surface of R ­ uO2(110). Faraday efficiencies of chlorine evolution were as follows:
All uncoordinated Ru surface atoms ­(Rucus) are covered by ­NiSb2Ox (96.0% ± 3.7%), ­CoSb2Ox (97.4% ± 3.0%), and
top oxygen (­ Oot), and the adjacent R ­ u2f atoms are bridged ­MnSb2Ox (89.9% ± 0.8%). X-ray photoelectron spectros-
by uncoordinated surface oxygen ­(Obr). In the ball-and-stick copy showed that the valence state of the transition metal
model, the active surface structure in CER and OER can be remained unchanged, whereas the valence of Sb changed
expressed as ­RuO2(110)–Oot. from 5 to a mixed state of 5 and 3. C ­ oSb2Ox exhibited the
In the first-order reaction kinetics, the reaction mecha- best stability and selectivity, and ­NiSb2Ox showed the high-
nism of R ­ uO2(110)–Oot active surface is considered to be est intrinsic activity. Ha et al. [85] proposed that ­Co3O4
the Volmer–Heyrovsky pathway. The mechanism is mainly nanoparticles have better chlorine evolution selectivity and
divided into the Volmer step of Cl adsorption and desorption stability than DSAs. The reaction proceeds by the Krishtalik
on the ­Rucus–Oot active site and the Heyrovsky step of the mechanism, in which the rate-determining step is the second
adsorption of Cl on the R ­ ucus–OClot precursor to recombine step, that is, the Cl atoms adsorbed on the surface of cobalt
with another Cl ion to form C ­ l2 [70]. This phenomenon is atoms are discharged to form adsorbed ­Cl+ species.
consistent with the recent experimental study of CER on
­RuO2(110) single crystal [78]. The rate-limiting step is the
Heyrovsky step under a low overpotential or Volmer step Membranes
under a high overpotential [79]. OER is a relatively compli-
cated four-electron reaction. This process mainly involves The IEM that is widely used in the chlor-alkali industry
the formation of ­OHot, ­Oot, and O ­ OHot adsorbates on R ­ ucus is a double-layer composite material composed of a thick
sites. sulfonic acid-based layer (PFSA) and a thin carboxylic
DFT calculations showed that the formation of ­OOHot acid-based layer (PFCA), which can be incorporated into
adsorbates is a rate-determining step. Exner et  al. [80] symmetrical, dense, or non-porous membranes. Dense films
proved this finding through experiments. Under industrial are relatively void-free structures, and their permeability is
conditions, the O ­ Clot and O­ OHot adsorbates control the inversely proportional to crystallinity. The early-aged chlor-
activity of CER and OER, respectively [70]. Thus, in addi- alkali electrolysis cell employed PFSA-based IEMs, which
tion to thermodynamic properties, the kinetic-based free suffered from low ion selectivity and current efficiency. By
energy of formation of ­OClot and ­OOHot adsorbents should contrast, the PFCA-based IEM is usually used as proton
also be considered in the improved design of catalysts for exchange membranes in fuel cells [86–88] and has excellent
CER. Loading a single layer of ­TiO2(110) on ­RuO2(110) can current efficiency, but is only suitable for producing alkali
increase the selectivity of CER, and the CER activity is also metal hydroxides with a maximum yield of 35% by weight
improved. Consistent with the theory above, Ti replaces the due to its low ion flux and high operating voltage. PFSA
top Ru atom, which weakens the free adsorption energy of and PFCA are usually combined into a double-layer IEM
­Oot by about 1.5 eV. to improve performance. Operation conditions influencing
membrane performances, such as current densities and water
contents, need to be optimized to advance the performance
Other Metal Oxide Electrodes of chlor-alkali cells.

Ir, whether as a pure oxide or dopant, can maintain a long- Key Parameters
term stability [81]. A series of iridium-based double per-
ovskite ­(Ba2BIrO6, where B = Pr, Nd, La, Sn, Y, Tb, Ce) The more uniform the distributions of temperature and brine
has been synthesized and studied for OER selectivity and concentration in the area near the IEM, the longer the ser-
stability [82]. The results showed that CER was carried out vice life of the IEM. A high current density can increase

13

210 K. Li et al.

production capacity and reduce production costs. However, between ion groups in the cluster due to the expansion
this condition speeds up the gas generation rate and causes phenomenon, which weakens the membrane mechanical
the accumulation of the generated gas at the top of the cell, strength.
resulting in an increase in temperature. Given the increase Mg2+ and ­Ca2+ impurities in the brine easily contami-
in local temperature of electrolyzers, the internal channels nate the ionic membrane, which reduces efficiency, increases
of the membrane will be damaged, a phenomenon called energy consumption, and shortens the service life of the
the “membrane pinhole effect”. In addition, a pH gradient membrane [90]. Ca ions can combine with carbonate to form
is built on the anode side surface of the membrane, whereas triangular ­CaCO3 [9, 93]. This slow process will cause the
the salt concentration in the tank decreases. This condi- efficiency of the electrolytic cell to gradually decrease. The
tion will cause IEMs to blister, delaminate, and be perma- state-of-the-art technology requires the use of an electrolyte
nently destroyed. A high operation current density will also containing calcium and magnesium below 20 ppb to enable
increase the temperature of the ion membrane surface and the membrane function for over 4 years. Moreover, to pre-
the interior of the electrolytic cell. When the cell tempera- vent the formation of bubbles, industry operation often uses
ture is higher than 90 °C, the water in the electrolytic cell various techniques to cover a layer of mineral on the top of
will be severely vaporized, and anolyte vaporization or boil- the membrane. However, the performance of the membrane
ing will deteriorate the performance of the ion membrane. in chlor-alkali electrolytic cells will degrade when the min-
Therefore, information about the distribution of cell temper- eral layer detaches from the surface.
ature, membrane anode surface concentration gradient, and
brine concentration is important to the operation, design, Advanced Membrane Design
and development of highly efficient electrolyzers.
The water content in the membrane is determined by the Figure 7 shows the ion transport in a bilayer membrane.
nature and concentration of ion exchange groups, counte- Under the long-term operation, particularly under abnor-
rions, and the degree of cross-linking of the polymer. Two mal conditions (such as failure of timely water supply in
types of water absorption exist [89]: one is absorption by the the cathode chamber, power interruption, reduction or inter-
film particles; the other is absorption by the combination of ruption of brine feed, etc.), the composite membrane is eas-
physical bonds and ions. The first type of water requires a ily delaminated, and blisters are formed between the two
very low heat; thus, it occurs at low temperatures, around layers, resulting in increase in the operating voltage. The
55–65 °C. The second type needs a slightly higher tempera- delamination between the sulfonic acid and carboxyl layers
ture. Structural properties, such as the equivalent weight of the composite membrane is mainly due to the difference
of the polymer, free volume, the degree of aggregation of in their water permeability. The use of a blend of PFSA and
ionic clusters, or cluster size, will affect water absorption. PFCA as an intermediate layer between the sulfonic acid and
The increase in water absorption will increase the distance

Fig. 7  Multicomponent ion
transport in a bilayer cation-
exchange membrane (reprinted
from [91] with permission from
Springer Nature)

13
Revisiting Chlor‑Alkali Electrolyzers: from Materials to Devices 211

carboxyl layers was proposed to minimize the difference in chlor-alkali industry was a typical membrane-less process,
water transport between the two layers. but the high toxicity of mercury prevented its further devel-
Wang et al. [92] studied the effect of blend layers with opment. However, the use of the previous mercury elec-
different ratios of PFSA to PFCA on the performance of trolysis cell implies that the reversible sodium-ion storage
composite membranes. The distribution morphology of the reaction can decouple the chlor-alkali process. Hou et al.
blend membrane has a great influence on its ion transport [93] proposed the ­Na0.44MnO2 electrode for ­Na+ insertion
performance and water absorption capacity. The maximum and extraction as a redox mediator (Fig. 8). The chlor-alkali
amount of water absorption of PFCA is smaller than that process was decoupled into two separate steps: ­H2/NaOH
of PFSA. However, the diffusion path of water in PFCA is formation and ­Cl2 generation. The first step is the cathodic
straighter than that in PFSA. Thus, a large diffusion coeffi- HER ­(H2O → ­H2) and the anode N ­ a+ deintercalation reac-
cient can reach equilibrium faster in the former. In addition, tion ­(Na0.44MnO2 → ­Na0.44−xMnO2), which produces NaOH
water absorption of the membrane in 32% sodium hydroxide in the electrolyte, as shown in Eqs. (6) and (8), respectively.
solution is considerably less than that in deionized water, The second step is the cathode N ­ a+ intercalation reaction
indicating that the highly concentrated sodium hydroxide ­(Na0.44−xMnO2 → ­Na0.44MnO2) and the production of C ­ l2
solution significantly reduces the water absorption of the (Cl → ­Cl2) (Eqs. (2) and (9), respectively). Technically, the
IEM in the electrolytic cell. As the content of PFCA in the two-step reaction can be carried out in the same electro-
blend film increases, the proton conductivity drops rap- lytic cell. However, as the NaOH concentration increases,
idly because the proton conductivity of sulfonate is higher the CER process at the electrode will shift to the OER side
than that of carbonate. The proton conductivity of PFSA is reaction. Compared with membrane electrolysis, the Faraday
directly related to water content, and the increase in PFCA efficiency of hydrogen production is higher (nearly 100%),
content will reduce the water content. The introduction of whereas the efficiency of chlorine production is about 90.2%,
the blend layer can not only enhance the intermolecular which is lower than that of membrane electrolysis (97.4%).
binding force between PFSA and PFCA layers but also bal- Step 1:
ance the water transfer. Given the different water permeabili-
ties of the two materials, an interlayer stress will be gener- Anode ∶ Na0.44 MnO2 → Na0.44 - x MnO2 +xNa+ +xe− . (8)
ated due to the electro-osmotic water during the electrolysis
process. The blend film can effectively reduce this interlayer Cathode: 2H2 O + 2e− → H2 + 2OH− . (4)
stress, which increases its resistance to peeling damage and
Step 2:
achieves long-term stability.
Anode: 2Cl− ↔ Cl2 + 2e− . (7)

Electrolyzers Cathode ∶ Na0.44-x MnO2 +xNa+ +xe− → Na0.44 MnO2 . (9)

Membrane‑Free Method

Given the above disadvantages of IEMs, membrane-less

electrolysis was proposed as an alternative technology.
The mercury electrolytic cell technology used in the early

Fig. 8  Schematic of the opera-

tion mechanism of a membrane-
free electrolysis cell (reprinted
from [93] with permission from
Nat Commun)

13

212 K. Li et al.

Coupling with Other Systems ­ O2 in saturated solutions

is limited by the slow diffusion of C
[62], and its current density is far below that used in practi-
The ODC electrode can be employed as the cathode to cal applications.
optimize the electrolysis system and improve energy effi-
CO2 + H2 O + 2e− → CO + 2OH− . (10)
ciency. A strategy is to use bifunctional electrodes to switch
between HER and ODC electrodes at any time [94]. How-
ever, a certain downtime will occur when electrode cleaning CO2 + OH− → HCO−3 . (11)
is necessary. Alternatively, chlor-alkali electrolysis cells can
possibly be combined with water electrolyzers or hydrogen CO2 + H2 O + 2e− → HCOO− + OH− . (12)
fuel cells [94]. The O ­ 2 produced in water electrolyzers can
be used to directly supply ODC, thereby saving part of the The GDE structure and flow cell have been developed to
cost. Coupled with the H ­ 2 fuel cell, the generated electrical overcome the slow ­CO2 transport issue. In the GDE structure
energy can be used to compensate for part of the electricity [62], ­CO2 is supplied in a gaseous state, and the path of C
­ O2
consumption in chlor-alkali electrolysis cells. diffusion to the catalyst is extremely short. Thus, its concen-
The combination of the chlor-alkali process with ­CO2 tration is maintained in the catalytic layer of the cathode,
reduction reaction (­CO2RR) into value-added products thereby achieving a high current density. Liu et al. [102]
while achieving negative carbon emissions and mitigating carried out a ­CO2RR experiment in 3.5 wt% (approximately
the greenhouse effect has also been proposed. In a single 0.6 M) NaCl electrolyte with a GDE to simulate the seawater
­CO2 reduction process, ­CO2RR occurs at the cathode [95], medium used in chlor-alkali electrolyzers (Fig. 9). The high
and the anode undergoes the OER process. ­CO2RR accepts formate faradaic efficiency of over 80% and a high formate
electrons to break the C=O bonds and generates a series of production rate of ~0.5 A/cm2 have been achieved within
basic anions ­(OH−, ­HCO3−, and ­HCOO−) [96], as shown in more than 0.5 V potential range. Moreover, the highest for-
Eqs. (10–12), whereas the anode generates acid to neutral- mate selectivity of 95% was achieved in saltwater because
ize the cathode products [95, 97–101]. Utilizing the C ­ O2RR the HER side reaction was deeply suppressed along with the
process to replace the cathode HER in the chlor-alkali elec- increased local pH.
trolytic cells, ­CO2 reduction products, ­Cl2, and NaOH can be
obtained. At present, most electrolytic cells adopt a zero-gap
configuration to minimize the ohmic loss of the electrolyte.
The integration of C ­ O2RR and chlor-alkali can be achieved
in an electrolytic cell similar to the H-cell, but this process

Fig. 9  Schematic model of a the chlor-alkali electrolyzer coupled with ­CO2RR and b the performance with tin oxide catalysts converting ­CO2
into formate (reprinted from [102] with permission from J Mater Chem A)

13
Revisiting Chlor‑Alkali Electrolyzers: from Materials to Devices 213

Conclusion and Outlook Open Access  This article is licensed under a Creative Commons Attri-
bution 4.0 International License, which permits use, sharing, adapta-
tion, distribution and reproduction in any medium or format, as long
The above discussion summarizes the research progress, as you give appropriate credit to the original author(s) and the source,
including those of the anode, cathode, IEM, and electroly- provide a link to the Creative Commons licence, and indicate if changes
sis system, of chlor-alkali electrolyzers. Although the chlor- were made. The images or other third party material in this article are
alkali industry has a history of over 100 years, it remains a included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in
high-energy-consuming industry today, consuming nearly the article’s Creative Commons licence and your intended use is not
10% of global electricity every year. This industry can be permitted by statutory regulation or exceeds the permitted use, you will
improved from the following aspects. need to obtain permission directly from the copyright holder. To view a
For the anode reaction, the mechanism of CER/OER must copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/.
be further understood. For example, combining modern exper-
imental and theoretical methods to study anode activity and
selectivity can provide further insights into the structure-per- References
formance relationship. The combination of theoretical mode-
ling and modern characterization methods (such as X-ray spec- 1. Crook J, Mousavi A (2016) The chlor-alkali process: a review of
history and pollution. Environ Forensics 17(3):211–217
troscopy) can help explore the details of the electrode reaction
2. Fauvarque J (1996) The chlorine industry. Pure Appl Chem
process and improve the selectivity and activity of the anode. 68(9):1713–1720
For the cathode, currently, a huge amount of hydrogen as by- 3. Lakshmanan S, Murugesan T (2014) The chlor-alkali process:
products is being vented to the atmosphere directly due to the work in progress. Clean Technol Environ Policy 16(2):225–234
4. Chen YQ, Manzhos S (2016) Voltage and capacity control of
insufficient market size to consume the hydrogen produced by
polyaniline based organic cathodes: an ab initio study. J Power
the chlor-alkali industry. Moreover hydrogen generated above Sources 336:126–131
is usually mixed with impurities (water vapor, nitrogen, oxy- 5. Cao YL, Yang HX, Ai XP et al (2003) The mechanism of oxygen
gen, and chlorine), which limits its usage. Further purification reduction on ­MnO2-catalyzed air cathode in alkaline solution. J
Electroanal Chem 557:127–134
processes, such as pressure swing adsorption, can produce
6. Cao YL, Xiao LF, Wang W et al (2011) Reversible sodium ion
high-purity hydrogen, which has a large market in electronic insertion in single crystalline manganese oxide nanowires with
industry, petrochemical industry, metal smelting, scientific long cycle life. Adv Mater 23(28):3155–3160
research, and other fields. Particularly, large amount of pure 7. Tan C (2005) Development and application of modified dia-
phragm. Chlor-Alkali Ind 11:21–24 (in Chinese)
hydrogen is needed to feed fuel cells to generate clean electric-
8. Bazinet L, Araya-Farias M (2005) Effect of calcium and carbon-
ity to drive future hydrogen society. ate concentrations on cationic membrane fouling during electro-
For IEMs, one can design a self-cleaning membrane to dialysis. J Colloid Interface Sci 281(1):188–196
avoid the accumulation of precipitation on the surface. The 9. Casademont C, Pourcelly G, Bazinet L (2007) Effect of mag-
nesium/calcium ratio in solutions subjected to electrodialysis:
service life of the membrane can be extended by optimizing
characterization of cation-exchange membrane fouling. J Colloid
the operation conditions, such as the quality of brine. Moreo- Interface Sci 315(2):544–554
ver, the addition of circulating electrolytes can result in the uni- 10. Dötzel O, Schneider L (2002) Non-asbestos diaphragms in chlor-
form distribution of the temperature and concentration of the alkali electrolysis. Chem Eng Technol 25(2):167
11. Stiegel GJ, Ramezan M (2006) Hydrogen from coal gasification:
brine in the chlor-alkali pool. This condition can be realized
an economical pathway to a sustainable energy future. Int J Coal
by a reasonable electrolytic cell design, such as the introduc- Geol 65(3–4):173–190
tion of an external forced circulation or installation of an inner 12. Xu JG, Froment GF (1989) Methane steam reforming, methana-
weir, which cannot eliminate gas accumulation at the top of the tion and water-gas shift: I. Intrinsic kinetics. Aiche J 35(1):88–96
13. de Wild PJ, Verhaak MJFM (2000) Catalytic production of
anode chamber. The passage of gas and liquid into the separa- hydrogen from methanol. Catal Today 60(1–2):3–10
tion chamber must be improved to enhance liquid exhaustion. 14. Rossmeisl J, Logadottir A, Nørskov JK (2005) Electroly-
Chlor-alkali electrolyzers can be coupled with other systems sis of water on (oxidized) metal surfaces. Chem Phys
(such as fuel cells and C ­ O2 electrolyzers). These combined 319(1–3):178–184
15. Kayfeci M, Keçebaş A, Bayat M (2019) Chapter 3-hydrogen
technologies need to be further evaluated based on the relation- production. Solar hydrogen production: processes, systems and
ship between the values of different products from C ­ O2 reduc- technologies (1st Edition):45–83
tion and energy consumption. Moreover, hydrogen, which is 16. Kong FD, Zhang S, Yin GP et al (2013) ­IrO2-graphene hybrid
usually treated as exhaust gas in the chlor-alkali industry, may as an active oxygen evolution catalyst for water electrolysis. Int
J Hydrog Energy 38(22):9217–9222
become the major fuel used on a large scale in the future as a
17. Levene JI, Mann MK, Margolis RM et al (2007) An analysis
promising clean energy. of hydrogen production from renewable electricity sources. Sol
Energy 81(6):773–780
Acknowledgements  The work was supported by the National 18. Lee DY, Elgowainy AA, Dai Q (2017) Life cycle greenhouse gas
Nature Science Foundation of China (Nos. 2193800 and 22078232) emissions of by-product hydrogen from chlor-alkali plants. Office
and the Science and Technology Major Project of Tianjin (Nos. of Scientific and Technical Information (OSTI), USA
18ZXJMTG00180 and 19ZXNCGX00030).

13

214 K. Li et al.

19. Elgowainy A (2017) Resourcing byproduct hydrogen from indus- 39. Erikson H, Sarapuu A, Tammeveski K (2019) Oxygen reduc-
trial operations. H2@Scale Workshop, Houston tion reaction on silver catalysts in alkaline media: a minire-
20. Lee DY, Elgowainy A, Dai Q (2018) Life cycle greenhouse gas view. ChemElectroChem 6(1):73–86
emissions of hydrogen fuel production from chlor-alkali pro- 40. Lipp L, Gottesfeld S, Chlistunoff J (2005) Peroxide forma-
cesses in the United States. Appl Energy 217:467–479 tion in a zero-gap chlor-alkali cell with an oxygen-depolarized
21. Yu H (2018) Hydrogen energy production from chlor-alkali elec- cathode. J Appl Electrochem 35(10):1015–1024
trolyzers. Chlor-Alkali Ind (in Chinese) 41. Kuwertz R, Gonzalez Martinez I, Vidaković-Koch T et  al
22. Ozturk M, Dincer I (2021) An integrated system for ammonia (2013) Energy-efficient chlorine production by gas-phase HCl
production from renewable hydrogen: a case study. Int J Hydrog electrolysis with oxygen depolarized cathode. Electrochem
Energy 46(8):5918–5925 Commun 34:320–322
23. Félix G, Quitian A, Rodríguez E et al (2017) Methods to calcu- 42. Zhang CZ, Fan FRF, Bard AJ (2009) Electrochemistry of
late hydrogen consumption during hydrocracking experiments in oxygen in concentrated NaOH solutions: solubility, diffu-
batch reactors. Energy Fuels 31(11):11690–11697 sion coefficients, and superoxide formation. J Am Chem Soc
24. Sharaf OZ, Orhan MF (2014) An overview of fuel cell technol- 131(1):177–181
ogy: fundamentals and applications. Renew Sustain Energy Rev 43. Zhao DJ, Zhang S, Yin GP et al (2013) Tungsten doped Co-Se
32:810–853 nanocomposites as an efficient non precious metal catalyst for
25. Brinkmann T, Giner-Santonja G, Schorcht F et al (2014) Best oxygen reduction. Electrochim Acta 91:179–184
available techniques (BAT) reference document for the produc- 44. Röhe M, Kubannek F, Krewer U (2019) Processes and their
tion of chlor-alkali. Publications Office of the European Union, limitations in oxygen depolarized cathodes: a dynamic model-
Denmark based analysis. Chemsuschem 12(11):2373–2384
26. Guandalini G, Foresti S, Campanari S et al (2017) Simulation 45. Röhe M, Botz A, Franzen D et al (2019) The key role of water
of a 2 MW PEM fuel cell plant for hydrogen recovery from activity for the operating behavior and dynamics of oxygen depo-
chlor-alkali industry. Energy Procedia 105:1839–1846 larized cathodes. ChemElectroChem 6(22):5671–5681
27. Cheng T, Wang L, Merinov BV et al (2018) Explanation of 46. Chavan N, Pinnow S, Polcyn GD et al (2015) Non-isothermal
dramatic pH-dependence of hydrogen binding on noble metal model for an industrial chlor-alkali cell with oxygen-depolarized
electrode: greatly weakened water adsorption at high pH. J Am cathode. J Appl Electrochem 45(8):899–912
Chem Soc 140(25):7787–7790 47. Clausmeyer J, Botz A, Öhl D et al (2016) The oxygen reduction
28. Wang YH, Chen L, Yu XM et al (2017) Superb alkaline hydro- reaction at the three-phase boundary: nanoelectrodes modified
gen evolution and simultaneous electricity generation by Pt- with Ag nanoclusters. Faraday Discuss 193:241–250
decorated ­Ni3N nanosheets. Adv Energy Mater 7(2):1601390 48. Franzen D, Ellendorff B, Paulisch MC et al (2019) Influence of
29. Wang PT, Jiang KZ, Wang GM et al (2016) Phase and interface binder content in silver-based gas diffusion electrodes on pore
engineering of platinum-nickel nanowires for efficient elec- system and electrochemical performance. J Appl Electrochem
trochemical hydrogen evolution. Angewandte Chemie Int Ed 49(7):705–713
55(41):12859–12863 49. Morimoto T, Suzuki K, Matsubara T et  al (2000) Oxygen
30. Mahmood N, Yao Y, Zhang JW et al (2018) Electrocatalysts reduction electrode in brine electrolysis. Electrochim Acta
for hydrogen evolution in alkaline electrolytes: mechanisms, 45(25–26):4257–4262
challenges, and prospective solutions. Adv Sci 5(2):1700464 50. Tseng CJ, Lo SK (2010) Effects of microstructure characteristics
31. Zheng Y, Jiao Y, Vasileff A et al (2018) The hydrogen evolu- of gas diffusion layer and microporous layer on the performance
tion reaction in alkaline solution: from theory, single crystal of PEMFC. Energy Convers Manag 51(4):677–684
models, to practical electrocatalysts. Angew Chem Int Ed Engl 51. Passalacqua E, Squadrito G, Lufrano F et  al (2001) Effects
57(26):7568–7579 of the diffusion layer characteristics on the performance of
32. Safizadeh F, Ghali E, Houlachi G (2015) Electrocatalysis polymer electrolyte fuel cell electrodes. J Appl Electrochem
developments for hydrogen evolution reaction in alkaline 31(4):449–454
solutions-a review. Int J Hydrog Energy 40(1):256–274 52. Nara H, Momma T, Osaka T (2013) Impedance analysis of the
33. Lei CJ, Wang Y, Hou Y et al (2019) Efficient alkaline hydro- effect of flooding in the cathode catalyst layer of the polymer
gen evolution on atomically dispersed Ni–Nx species anchored electrolyte fuel cell. Electrochim Acta 113:720–729
porous carbon with embedded Ni nanoparticles by accelerating 53. Park S, Popov BN (2009) Effect of cathode GDL characteristics
water dissociation kinetics. Energy Environ Sci 12(1):149–156 on mass transport in PEM fuel cells. Fuel 88(11):2068–2073
34. Sun QZ, Zhang B, Diao LC et al (2020) Engineering the elec- 54. Kiros Y, Pirjamali M, Bursell M (2006) Oxygen reduction elec-
tronic structure of 1T’-ReS 2 through nitrogen implantation trodes for electrolysis in chlor-alkali cells. Electrochim Acta
for enhanced alkaline hydrogen evolution. J Mater Chem A 51(16):3346–3350
8(23):11607–11616 55. Marković NM, Schmidt TJ, Stamenković V et al (2001) Oxygen
35. Zhao DJ, Zhang S, Yin GP et al (2012) Effect of Se in Co-based reduction reaction on Pt and Pt bimetallic surfaces: a selective
selenides towards oxygen reduction electrocatalytic activity. J review. Fuel Cells 1(2):105–116
Power Sources 206:103–107 56. Shao YY, Zhang S, Engelhard MH et al (2010) Nitrogen-doped
36. Kintrup J, Millaruelo M, Trieu V et al (2017) Gas diffusion graphene and its electrochemical applications. J Mater Chem
electrodes for efficient manufacturing of chlorine and other 20(35):7491
chemicals. Electrochem Soc Interface 26(2):73–76 57. Masa J, Xia W, Muhler M et al (2015) On the role of metals
37. Ding JS, Hua WQ, Hu BB et al (2011) Closed loop recycling of in nitrogen-doped carbon electrocatalysts for oxygen reduction.
chlorine for sustainable development of polyurethane industry. Angewandte Chemie Int Ed 54(35):10102–10120
Trans Tianjin Univ 17(4):298–304 58. Li JK, Jia QY, Ghoshal S et al (2017) Highly active and stable
38. Liu JJ, Yang C, Liu CG et al (2014) Design of pore structure in Fe–N–C catalyst for oxygen depolarized cathode applications.
gas diffusion layers for oxygen depolarized cathode and their Langmuir 33(37):9246–9253
effect on activity for oxygen reduction reaction. Ind Eng Chem 59. Mamtani K, Jain D, Co AC et al (2017) Investigation of chloride
Res 53(14):5866–5872 poisoning resistance for nitrogen-doped carbon nanostructures

13
Revisiting Chlor‑Alkali Electrolyzers: from Materials to Devices 215

as oxygen depolarized cathode catalysts in acidic media. Catal 80. Exner KS, Sohrabnejad-Eskan I, Anton J et al (2017) Full free
Lett 147(12):2903–2909 energy diagram of an electrocatalytic reaction over a single-crys-
60. Singh V, Nagaiah TC (2019) In situ incorporation of cobalt nano- talline model electrode. ChemElectroChem 4(11):2902–2908
clusters and nitrogen into the carbon matrix: a bifunctional cata- 81. Spöri C, Kwan JTH, Bonakdarpour A et al (2017) The stability
lyst for the oxygen depolarized cathode and chlorine evolution challenges of oxygen evolving catalysts: towards a common fun-
in HCl electrolysis. J Mater Chem A 7(16):10019–10029 damental understanding and mitigation of catalyst degradation.
61. Moussallem I, Pinnow S, Wagner N et al (2012) Development of Angewandte Chemie Int Ed 56(22):5994–6021
high-performance silver-based gas-diffusion electrodes for chlor- 82. Vos J, Liu ZC, Speck FD et al (2019) Selectivity trends between
alkali electrolysis with oxygen depolarized cathodes. Chem Eng oxygen evolution and chlorine evolution on iridium-based double
Process Process Intensif 52:125–131 perovskites in acidic media. ACS Catal 9(9):8561–8574
62. Burdyny T, Smith WA (2019) ­CO2 reduction on gas-diffusion 83. Moreno-Hernandez IA, MacFarland CA, Read CG et al (2017)
electrodes and why catalytic performance must be assessed Crystalline nickel manganese antimonate as a stable water-oxi-
at commercially-relevant conditions. Energy Environ Sci dation catalyst in aqueous 1.0 M H ­ 2SO4. Energy Environ Sci
12(5):1442–1453 10(10):2103–2108
63. Zhang LH, Fan Q, Li K et al (2020) First-row transition metal 84. Moreno-Hernandez IA, Brunschwig BS, Lewis NS (2019) Crys-
oxide oxygen evolution electrocatalysts: regulation strate- talline nickel, cobalt, and manganese antimonates as electrocat-
gies and mechanistic understandings. Sustain Energ Fuels alysts for the chlorine evolution reaction. Energy Environ Sci
4(11):5417–5432 12(4):1241–1248
64. Petrykin V, Macounová K, Okube M et al (2013) Local structure 85. Ha H, Jin K, Park S et al (2019) Highly selective active chlorine
of Co doped ­RuO2 nanocrystalline electrocatalytic materials for generation electrocatalyzed by ­Co3O4 nanoparticles: mechanis-
chlorine and oxygen evolution. Catal Today 202:63–69 tic investigation through in situ electrokinetic and spectroscopic
65. Halck NB, Petrykin V, Krtil P et al (2014) Beyond the volcano analyses. J Phys Chem Lett 10(6):1226–1233
limitations in electrocatalysis—oxygen evolution reaction. Phys 86. Mogg L, Zhang S, Hao GP et al (2019) Perfect proton selectivity
Chem Chem Phys 16(27):13682–13688 in ion transport through two-dimensional crystals. Nat Commun
66. Macounová KM, Simic N, Ahlberg E et al (2015) Electrochemi- 10:4243
cal water-splitting based on hypochlorite oxidation. J Am Chem 87. Mogg L, Hao GP, Zhang S et al (2019) Atomically thin micas as
Soc 137(23):7262–7265 proton-conducting membranes. Nat Nanotechnol 14(10):962–966
67. Koper MTM (2011) Thermodynamic theory of multi-electron 88. Liang X, Wu L, Xu TW (2018) Role of ionomer in membrane
transfer reactions: implications for electrocatalysis. J Electroanal electrode assembly for proton exchange membrane fuel cells. Sci
Chem 660(2):254–260 Sin-Chim 48(9):1040–1057
68. Karlsson RK, Cornell A (2016) Selectivity between oxygen and 89. Zeynali ME, Mohammadi F, Rabiee A (2015) Structural analy-
chlorine evolution in the chlor-alkali and chlorate processes. sis and defect evaluation of ion exchange composite membranes
Chem Rev 116(5):2982–3028 used in electrolysis of sodium chloride in chlor-alkali process.
69. Hansen HA, Man IC, Studt F et al (2010) Electrochemical chlo- Iran Polym J 24(2):85–93
rine evolution at rutile oxide (110) surfaces. Phys Chem Chem 90. Le Faucheur S, Vasiliu D, Catianis I et al (2016) Environmental
Phys 12(1):283–290 quality assessment of reservoirs impacted by Hg from chlor-
70. Exner KS (2019) Controlling stability and selectivity in the com- alkali technologies: case study of a recovery. Environ Sci Pollut
peting chlorine and oxygen evolution reaction over transition Res 23(22):22542–22553
metal oxide electrodes. ChemElectroChem 6(13):3401–3409 91. Moshtarikhah S, Oppers NAW, Groot MT et al (2017) Mul-
71. Over H (2012) Surface chemistry of ruthenium dioxide in het- ticomponent ion transport in a mono- and bilayer cation-
erogeneous catalysis and electrocatalysis: from fundamental to exchange membrane at high current density. J Appl Electrochem
applied research. Chem Rev 112(6):3356–3426 47(2):213–221
72. Trasatti S (2000) Electrocatalysis: understanding the success of 92. Wang J, Wang XJ, Dou P et al (2015) Morphology and proper-
DSA (R). Electrochim Acta 45(15–16):2377–2385 ties of perfluorosulfonic acid polymer/perfluorocarboxylic acid
73. Nanni LC, Polizzi S, Benedetti A et al (1999) Morphology, polymer blend membranes. Polym Eng Sci 55(1):180–189
microstructure, and electrocatalytic properties of ­RuO2-SnO2 93. Hou M, Chen L, Guo Z et al (2018) A clean and membrane-free
thin films. J Electrochem Soc 146(1):220–225 chlor-alkali process with decoupled C ­ l2 and H
­ 2/NaOH produc-
74. Xiong K, Deng ZH, Li L et  al (2013) Sn and Sb co-doped tion. Nat Commun 9(1):438
RuTi oxides supported on ­TiO2 nanotubes anode for selectivity 94. Brée LC, Bulan A, Herding R et al (2020) Techno-economic
toward electrocatalytic chlorine evolution. J Appl Electrochem comparison of flexibility options in chlorine production. Ind Eng
43(8):847–854 Chem Res 59(26):12186–12196
75. Trasatti S (1984) Electrocatalysis in the anodic evolution of oxy- 95. Kuang SY, Li ML, Xia R et al (2020) Stable surface-anchored
gen and chlorine. Electrochim Acta 29(11):1503–1512 Cu nanocubes for ­CO2 electroreduction to ethylene. ACS Appl
76. Exner KS (2020) Beyond dimensionally stable anodes: single- Nano Mater 3(8):8328–8334
atom catalysts with superior chlorine selectivity. ChemElectro- 96. Guo JH, Sun WY (2020) Integrating nickel-nitrogen doped car-
Chem 7(7):1528–1530 bon catalyzed C ­ O2 electroreduction with chlor-alkali process for
77. Exner KS, Anton J, Jacob T et al (2014) Chlorine evolution reac- CO, ­Cl2 and K­ HCO3 production with enhanced techno-econom-
tion on R­ uO2(110): ab initio atomistic thermodynamics study— ics. Appl Catal B Environ 275:119154
Pourbaix diagrams. Electrochim Acta 120:460–466 97. Zhang S, Fan Q, Xia R et al (2020) C ­ O2 reduction: from homo-
78. Kuo DY, Paik H, Nelson JN et al (2019) Chlorine evolution reac- geneous to heterogeneous electrocatalysis. Acc Chem Res
tion electrocatalysis on ­RuO2(110) and ­IrO2(110) grown using 53(1):255–264
molecular-beam epitaxy. J Chem Phys 150(4):041726 98. Gao DF, Zhang Y, Zhou ZW et al (2017) Enhancing ­CO2 elec-
79. Exner KS, Anton J, Jacob T et al (2016) Full kinetics from first troreduction with the metal–oxide interface. J Am Chem Soc
principles of the chlorine evolution reaction over a ­RuO2(110) 139(16):5652–5655
model electrode. Angew Chem Int Ed 55(26):7501–7504

13

216 K. Li et al.

99. Zhang S, Kang P, Meyer TJ (2014) Nanostructured tin catalysts at Chapel Hill in early 2013 and conducted his postdoctoral research
for selective electrochemical reduction of carbon dioxide to for- concentrated on C ­ O2 electrochemistry in Prof. Thomas J. Meyer’s
mate. J Am Chem Soc 136(5):1734–1737 group. In 2015, he obtained Marie Curie international fellowship for
100. Xia R, Zhang S, Ma XB et al (2020) Surface-functionalized experienced researchers and did his research on proton transfer in 2D
palladium catalysts for electrochemical ­CO2 reduction. J Mater materials in Prof. Andre K. Geim’s group at the University of Man-
Chem A 8(31):15884–15890 chester, United Kingdom. Prof. Zhang’s research efforts are aimed at
101. Liu S, Yang HB, Hung SF et al (2020) Elucidating the electro- developing advanced materials and electrochemical technologies for
catalytic ­CO2 reduction reaction over a model single-atom nickel clean energies, such as carbon dioxide electrolyzers, water electrolyz-
catalyst. Angew Chem Int Ed Engl 59(2):798–803 ers, fuel cells, etc. So far he has published over 50 peer reviewed papers
102. Liu H, Su YQ, Kuang SY et al (2021) High efficient ­CO2 elec- with ~7500 citations in top journals including Nat. Nanotechnol., Proc.
trolysis within a wide operation window using octahedral tin Natl. Acad. Sci. U.S.A., J. Am. Chem. Soc., Nat. Commun., Acc. Chem.
oxide single crystals. J Mater Chem A 9:7848–7856 Res., Angew. Chem. Int. Ed., and so on.

Sheng Zhang  is a full professor

in the School of Chemical Engi-
neering and Technology at Tian-
jin University. His broad
research experiences include
electrochemical engineering,
electrocatalysis, and materials
chemistry and physics. He
earned his B.E., M.E., and Ph.D.
at Harbin Institute of Technol-
ogy. After a short stay at Case
Western Reserve University as a
research associate focusing on
carbon materials, he moved to
the University of North Carolina

13