Perspective

http://pubs.acs.org/journal/aelccp

Zero-Gap Electrochemical CO2 Reduction

Cells: Challenges and Operational Strategies
for Prevention of Salt Precipitation
Mark Sassenburg, Maria Kelly, Siddhartha Subramanian, Wilson A. Smith, and Thomas Burdyny*

Cite This: ACS Energy Lett. 2023, 8, 321−331 Read Online

See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

ACCESS Metrics & More Article Recommendations

ABSTRACT: Salt precipitation is a problem in electrochemical

Downloaded via VITO NV on July 13, 2023 at 15:39:50 (UTC).

CO2 reduction electrolyzers that limits their long-term

durability and industrial applicability by reducing the active
area, causing flooding and hindering gas transport. Salt crystals
form when hydroxide generation from electrochemical reac-
tions interacts homogeneously with CO2 to generate substantial
quantities of carbonate. In the presence of sufficient electrolyte
cations, the solubility limits of these species are reached,
resulting in “salting out” conditions in cathode compartments.
Detrimental salt precipitation is regularly observed in zero-gap
membrane electrode assemblies, especially when operated at
high current densities. This Perspective briefly discusses the
mechanisms for salt formation, and recently reported strategies
for preventing or reversing salt formation in zero-gap CO2 reduction membrane electrode assemblies. We link these
approaches to the solubility limit of potassium carbonate within the electrolyzer and describe how each strategy separately
manipulates water, potassium, and carbonate concentrations to prevent (or mitigate) salt formation.

T he electrochemical CO2 reduction reaction (CO2RR)

provides a pathway toward a more CO2 neutral
society. Although still in its infancy, the potential for
this technology to develop further has led to improvements in
the product selectivity, activity, and stability of CO2RR
MEA architectures where a liquid catholyte is removed, salt
precipitation is common and highly disruptive to steady
performance.
An exchange MEA is the most common MEA architecture
used in CO2RR electrolyzers. The cathode side uses a porous
electrolyzers. Much has been adopted from the already GDE and is fed with a gaseous stream of CO2 that can be dry
matured electrochemical hydrogen evolution reaction (HER) or humidified. The anode of the exchange MEA contacts a
field, where performance metrics such as >1 A cm −2 liquid anolyte that provides reactants for the anode reaction
conversion and >10,000 h lifetime are easily surpassed.1 By (typically oxygen evolution) and serves as a water source for
adopting technical features like the gas diffusion electrode the membrane.12 MEAs with a gaseous anode feed (also
(GDE)2−6 and membrane electrode assembly (MEA)7 cell known as full MEAs) have been demonstrated for
architecture, the field of CO2 reduction has achieved CO2RR,13−16 but reports on these systems are limited and
industrially relevant current densities (>200 mA cm−2) while fall outside the primary scope of this Perspective.
retaining selective conversion. These improvements were in The use of GDEs in MEAs is the feature which enables
part realized by utilizing highly alkaline electrolytes, such as elevated current densities by reducing the liquid diffusion
KOH, to limit the competing HER4,8−10 and humidifying the length of CO2 from the gas phase to the catalyst surface.
CO2 gas stream to manage water availability to the cathode
and membrane.11,12 However, since being incorporated into Received: August 22, 2022
more industrial reactors, additional challenges have been found Accepted: November 15, 2022
which impact the long-term stability and economic feasibility Published: December 5, 2022
of CO2RR. In particular, the precipitation of salts within the
reactor leads to operational failures which diminish the
potential impact of this technology. In higher energy efficiency
© 2022 The Authors. Published by
American Chemical Society https://doi.org/10.1021/acsenergylett.2c01885
321 ACS Energy Lett. 2023, 8, 321−331
ACS Energy Letters http://pubs.acs.org/journal/aelccp Perspective

Figure 1. Schematic representation of the cascade of reactions and ion transport in an exchange MEA leading to salt formation on the
cathode composed of a catalyst layer (CL) and gas-diffusion layer (GDL). The inserted graph shows the change in ion concentrations
occurring near the cathode. After both CO32− and K+ concentrations reach critical levels, the precipitation of K2CO3 starts to occur.

However, the production of hydroxide as a byproduct of the underlying principles of local alkalinity, water, and ion
CO2RR during water-splitting, and the use the KOH as an transport can generally be translated to MEA systems.
anolyte, result in a highly alkaline local environment17−19 in Several operational approaches have been deployed in
the cathode compartment of the electrolyzer. The excess CO2 literature to maintain long-term CO2 electrolysis without salt
which is enabled by the gas-diffusion layer then simultaneously formation. In essence, however, each of these strategies work
provides a route toward salt formation through the production toward a similar goal and separately prevent salt formation by
of (bi)carbonates (Figure 1). lowering either [K+], [CO32−], or [K2CO3] in the cathode
In MEAs, these carbonate salts can form in the cathode flow compartment. Some are “active” approaches that require a
field, on the gas side of the cathode, within the GDE, and on periodic change in the operational state of the electrolyzer.
the membrane side of the electrode in systems using both Others are “passive” approaches that are in effect at all times.
alkaline and near-neutral anolytes.7,12,20,21 The deposits block Here, we group the strategies presented in literature into four
general categories. (1) Passive Anolyte Approach: the anolyte
the initially porous GDE and cause the pressure within the
concentration is decreased, or the cation identity is changed, to
cathode chamber to increase as gas flow is progressively
keep the accumulation of cations at the cathode surface below
restricted by the salts.20,22 The presence and formation of salt the critical salting out concentration. (2) Active Dissolution
also restricts access of CO2 to the catalyst, leading to increased Approach: the cathode is periodically pulsed with water or an
hydrogen Faradaic efficiencies. Although salt precipitation has equivalent solvent to dissolve accumulated salts and increase
been observed in other alkaline electrochemical systems,23,24 water availability. Alternatively, while feeding a deionized water
its prevalence in CO2 reduction electrolyzers comes from the anolyte, the cathode is periodically flushed with an “activation”
interplay of 3 components essential to CO2RR: the reactant solution to provide cations near the cathode surface. (3) Active
CO2 gas, the proton-source (H2O or HCO3−),25−27 and a Pulse Approach: the MEA is operated in a pulsed electrolysis
cation that assists in catalysis.28−30 Several citations used in the mode where periodically switching to a low applied potential
work presented here make use of 3 compartment flow- allows accumulated cations and carbonate ions to diffuse away
cells22,31,32,41,56 and even fully aqueous setups,25−30,47 in which from the cathode, thereby keeping their concentration below
mass transport can be quantitatively different. Nevertheless, critical levels. (4) Passive Membrane Approach: the MEA
322 https://doi.org/10.1021/acsenergylett.2c01885
ACS Energy Lett. 2023, 8, 321−331
ACS Energy Letters http://pubs.acs.org/journal/aelccp Perspective

membrane and its components are chosen to reduce ion limits of carbonate salts, each by targeting either the cation,
migration to and accumulation at the cathode. anion, or water concentrations.
First, to explain how salt formation takes place, we look at
In essence each of these strategies the conversion of CO2-to-CO on a Ag catalyst in an alkaline
environment. During electrolysis, some of the CO2 fed into the
work toward a similar goal and sepa- system is converted to CO as described by the cathodic half
rately prevent salt formation by low- reaction:
ering either [K+], [CO32−], or [K2CO3] in
CO2 (aq) + H 2O(l) + 2e CO(aq) + 2OH (aq)
the cathode compartment. (1)
This Perspective reflects on these operational strategies for For each converted CO2 molecule, two hydroxide ions are
avoiding or reversing salt formation in CO2 electrolyzers. We produced when in a neutral or alkaline pH environment. In
discuss each of these approaches in-depth next to the addition to making the environment more alkaline, OH− also
phenomena causing salt formation to highlight that all participates in the unwanted homogeneous conversion of CO2
strategies work toward the same goal of avoiding the solubility to bicarbonate and carbonate (depending on the exact pH):

CO2 (g) + 2OH (aq) JoooooooooK HCO3 (aq) + OH (aq) JoooooooooK CO32 (aq) + H 2O(l)
pK a = 6.35 pK a = 10.3 (2)

Since CO2 gas is abundantly present and hydroxides are Many studies have examined the effect of different salt
continuously produced, the effectively utilized amount of CO2 cations on the performance of CO2RR systems, but here we
gas for CO2RR can drop down to ∼30% due to dissolution, focus on the implications of K+ as it is the most studied salt
while up to ∼70% of CO2 is converted into carbonates that can cation. These conclusions can be generalized to other cations,
fuel salt formation.31,32 Multiphysics models developed by albeit with different solubility limits potentially changing the
Weng et al. and Kas et al. have also determined the maximum primary location of salt formation in the cathode compartment.
CO2 utilization efficiency to be ∼50% for an exchange MEA While the chemical reactions in eqs 1−3 describe how ions
system and a GDE with a flowing catholyte, respectively.33,34 are formed and precipitate into salts, the Nernst−Planck
While this is a significant problem on its own in terms of CO2 equation then describes the transport and accumulation of ions
utilization efficiency, another issue is the accumulation of across the electrochemical system:
carbonate at the cathode due to reaction 2.
The third reaction to consider is the combination of C(x) zF (x )
J (x) = D + DC + C (x )
accumulating carbonate ions near the gas−liquid interface and x RT x
the cations (i.e., K+) that are used to improve ionic
flux = diffusion + migration + convection (4)
conductivity and stabilize CO2 reduction intermediates. Since
the cathode is negatively charged during electrolysis and where J(x) is the flux of an ionic species, D is its diffusivity
hydroxide ions are being produced, migration of cations from constant, dC/dx is the concentration gradient, z is its
the anolyte past the membrane leads to a gradually increasing electronic charge, F is Faraday’s constant, R is the ideal gas
concentration near the cathode to maintain charge neutrality constant, T is the temperature, dϕ/dx is the electrical potential
within the system. Ultimately the high concentrations of gradient, and v(x) is the fluid velocity. Near the electrode
cations and carbonates exceed the solubility limit (1096 g/L or surface where the fluid velocity ν is negligible (Cν(0) = 0), this
7.93 M K2CO3 at 20 °C in pure water)35 and lead to the equation states that in a steady state system where there is no
formation of salts: net flux of ionic species (J(x) = 0), the electromigration of
potassium ions toward the negative cathode has to equalize
CO32 (aq) + 2K+(aq) K 2CO3(s) (3) with the diffusion of high concentrations back to the
(relatively) low concentration bulk.
It is most accurate to use the solubility product constant Within a zero-gap system the concentrations of ionic species
(Ksp) to define the conditions for K2CO3 precipitation. are then determined by the applied reaction rate, the anolyte
However, K2CO3 is highly soluble, and at saturation the concentration, and the diffusion, migration, and convection
solution would deviate from ideal solution behavior. For driven ionic transport through the cathode region, membrane,
simplicity, the remainder of this review will use the solubility of and anode region. While carbonate forms easily as gaseous
K2CO3 in units of molarity to describe the conditions for CO2 reacts with the OH− product (eq 2), a zero-gap system
precipitation with the disclaimer that greater concentrations of typically has limited potassium ions initially at the cathode.
potassium and carbonate could lead to earlier than described Moreover, the majority of reported zero-gap systems utilize
salt formation. For this reason, operational strategies should anion exchange membranes, which should be repellant to
aim to keep both K+ < 15.86 M and CO32− < 7.93 M to avoid cations.37 Driven by the high concentration of negative charges
the solubility product from exceeding the solubility limit. at the cathode, counterion transport of potassium across the
In addition to K2CO3, KHCO3 and K4H2(CO3)3·1.5H2O20 anion exchange membrane is facilitated through electro-
have also been detected by ex situ XRD in MEA cathodes. osmotic drag as depicted in Figure 1. In conjunction with
KHCO3 and K4H2(CO3)3·1.5H2O can form by CO2 sorption water transport, partially neutralized potassium ions are able to
of solid K2CO3, so it is proposed that K2CO3 initially cross the membrane and accumulate at the cathode.
precipitates then reacts with excess CO2 in the gas stream to In order to avoid potassium carbonate precipitation in a
form other carbonate salts.20,36 strongly alkaline system, the concentrations of both CO32− and
323 https://doi.org/10.1021/acsenergylett.2c01885
ACS Energy Lett. 2023, 8, 321−331
ACS Energy Letters http://pubs.acs.org/journal/aelccp Perspective

K+ must be kept below 7.93 and 15.86 M, respectively. effects of ion migration from the anode to the cathode.
Although these concentrations are much higher than the ∼1 M Migration is then balanced by diffusion of cations from the
K+ of typical CO2RR electrolytes, the substantial production of cathode to the anode. Combined, the accumulation of
hydroxide and carbonate at elevated current densities creates potassium at the cathode is maintained below the solubility
such an environment, as was computationally hypothesized by limit of K2CO3, thereby preventing salt precipitation (Figure
several catalyst layer concentration models.17,38,39 2a).
The experiences of rapid salt formation at industrially Liu et al. showed that reducing the anolyte concentration to
relevant current densities (e.g., 50 min for a 2 M KOH anolyte 10 mM KHCO3 instead of the typical 1 M concentration
operating at 100 mA cm−2)40 indicate that the migration term allowed stable operation for 3800 h (200 mA cm−2, 3 Vcell).43
of cations toward the cathode is larger than the diffusion term In this situation, the diffusion and migration terms equalize
in eq 4. Once salting out conditions are met, nucleation occurs and keep the potassium concentration below the critical salting
and rapid growth of crystals is observed into the cathode pores out condition. However, the use of a lower anolyte
and flow field until salts block gas flow altogether. concentration also increased the overall cell resistance, leading
To achieve an operational lifetime in the range of hydrogen to higher cell potentials. Similarly, Endrő di et al. observed that
electrolyzers (>10,000 h), methods for the prevention (or decreasing the electrolyte concentration prolongs electrolyzer
reversal) of salt formation in CO2RR MEA systems need to be operation at the expense of current density. When operating an
developed and improved. However, MEA designs to prevent MEA at 3.1 Vcell, the current density with a 0.1 M KOH
carbonate precipitation faces several challenges with various anolyte was 300 mA cm−2 but dropped to 100 mA cm−2 in a
trade-offs for performance and durability. Any change made to deionized water anolyte.20 The drop in current density when
suppress salt formation often contributes to other negative using a pure water feed can again be attributed to its low
effects such as electrolyte flooding,41 loss of CO2RR selectivity conductivity: electrochemical impedance spectra of both cells
over HER,42 increase in cell voltage,43 or increased down time indicated a 3 to 4 times larger charge transfer resistance in the
of the reactor for cleaning or pulsed electrolysis modes.44 MEA fed with pure water compared to 0.1 M KOH.
Thus, implementation of engineering and design methods for However, the performance of CO2RR MEAs using a pure
precipitation prevention results in a complex optimization water anolyte has been improved using novel membranes and
problem of many MEA operational factors. ionomers. For example, Yin et al. used a quaternary ammonia
In the past decade of CO2RR research, salt precipitation in poly(N-methylpiperidine-co-p-terphenyl) polymer as both
CO2 electrolyzers with GDEs has not been studied extensively anion exchange membrane and cathode ionomer in an MEA
despite being a commonly observed phenomenon. Only a few operating with pure water anolyte. The system achieved 100
papers have mentioned salt formation and its importance in mA cm−2 at 2.25 V for over 100 h with CO FE consistently
operations, while fewer provide empirical engineering solutions greater than 90%.45 The same system reached 500 mA cm−2
to obtain longer stability. By analyzing the research that has and ∼90% FE at 3 V and 60 °C, although long-term durability
at this current density was not reported. By avoiding the use of
By analyzing the research that has an alkaline electrolyte and consequently the introduction of
sought to overcome salt precipitation metal cations, the authors were able to prevent salt
precipitation entirely. Notably, it is generally agreed upon
we were able to identify 4 main that small amounts of alkali metal cations are needed to
categories of engineering solutions. increase the system conductivity and stabilize the CO2RR
These approaches include (i) passively intermediates,29 so the mechanisms for CO2RR in systems
with deionized water anolytes should be further investigated.
modifying the anolyte concentration O’Brien et al. suggests such systems without a mobile cation
and composition, (ii) actively dissolving can still achieve high CO2RR selectivity if the fixed positive
salts at the cathode, (iii) actively charges in the anion exchange membrane are able to stabilize
pulsing the electrolyzer, and (iv) pas- the CO2 reduction intermediates instead.46
These examples demonstrate the trade-off between salt
sively modifying the MEA. precipitation and cell voltage when lowering the anolyte
concentration. Thus, for the issue of salt prevention, the
sought to overcome salt precipitation we were able to identify question is whether it is economically beneficial to prevent salt
4 main categories of engineering solutions. These approaches precipitation by using dilute electrolytes that will increase the
include (i) passively modifying the anolyte concentration and overall cell potential. As more data on long-term testing of
composition, (ii) actively dissolving salts at the cathode, (iii) CO2RR electrolyzers becomes available, technoeconomic
actively pulsing the electrolyzer, and (iv) passively modifying analyses should consider the trade-off between cell potential
the MEA. Collectively, these strategies tackle the same issue of and cell lifetime which is influenced by salt precipitation. As an
preventing potassium and carbonate from simultaneously alternative approach to limit potassium crossover from the
reaching their critical concentrations. anode, the properties of anion exchange membranes
themselves could also play an important role. By varying the
1. PASSIVE ANOLYTE APPROACH: CATION thickness, water permeability, hydration, and ionic resistan-
CONCENTRATION AND IDENTITY ces,47,48 modifications of the membrane may limit potassium
The first option presented to reduce salt formation is to crossover without reduction of anolyte concentration.
decrease the concentration of cations in the electrolyte, or Salt precipitation may also be controlled by altering the
eliminate them entirely from the system (illustrated in Figure cation identity of the anolyte. Cofell et al. observed that
2). From a mass transport perspective, a lower bulk switching the electrolyte from KOH to CsOH in a flow cell
concentration of K+ in the anolyte reduces the transport resulted in smaller, well-dispersed bicarbonate crystal deposits
324 https://doi.org/10.1021/acsenergylett.2c01885
ACS Energy Lett. 2023, 8, 321−331
ACS Energy Letters http://pubs.acs.org/journal/aelccp Perspective

Figure 2. (a) Plot of cathode concentration versus time showing the general trends of K+ and CO32− concentrations at the cathode when the
anolyte concentration is reduced. (b) Schematic depiction of a lower concentration of K+ in the anolyte solution resulting in reduced
electromigration. This enables the balancing between migration and diffusion of K+, keeping the total concentration below the solubility
limit of K2CO3.

and a slowing of the performance degradation caused by the after salts have precipitated in a CO2RR system. Importantly
precipitation of carbonate salts.4 By contrast, the bicarbonate this strategy takes advantage of the fact that the most
deposits formed from the KOH electrolyte covered much detrimental effect of salt formation is blockages of the CO2
larger areas of the cathode and formed fractal-like patterns. diffusion pathways and not necessarily the nucleation of salt
Chiacchiarelli et al. also noted the effect of cation identity on crystals themselves. If the salt crystals at the cathode can then
slowing the formation of deposits on an electrode.49 In their be removed through the timely introduction of a secondary
work, a rotating Sn electrode was submerged in a 0.1 M flow, the operational lifetime of the system can be increased
KHCO3 electrolyte purged with N2. Subsequent electrolysis (Figure 3). Additionally, preventative addition of water to the
resulted in several degradation modes, including alkali deposits cathode region can periodically lower ion concentrations prior
from the electrolyte. The amount of the deposits decreased to salt formation occurring.
based on the cation identity in the order Na+ > K+ > Cs+. This Endrő di et al. performed two experiments to remove the
trend could be explained by the solubility change with cation accumulation of K2CO3 salts in the cathode.42 In the first
identity (Table 1). For carbonates, the solubility (in units of experiment, the cathode gas feed was humidified and heated to
85 °C to increase the water vapor in the cathode flow field and
Table 1. Solubility of (Bi)carbonate Species for Na+, K+, and salt solubility. This approach allowed for stable operation for at
Cs+ Cations least 8 h (at 200 mA cm−2, 3 Vcell) but lowered the selective
salt solubility (M at 20 °C) CO conversion to 65−70% due to the increased water content
which promoted HER. In the second experiment, the cathode
NaHCO3 1.14
KHCO3 2.24
chamber was flushed once per hour with a 50 cm3 deionized
CsHCO3 3.49
water (Tcell = 60 °C). During CO2RR, a continuously decaying
Na2CO3 2.06
current (275−200 mA cm−2) was obtained, which the authors
K2CO3 7.93 attributed to the formation of K2CO3. After each dissolution
Cs2CO3 8.01 step, the reduced current returned to its initial value after
which a new “decay cycle” was initiated. The combination of
lower temperature and salt dissolution resulted in a continuous
molarity) increases in the order Na+ < K+ ≈ Cs+, and for selectivity of 85% CO2-to-CO. The empirically chosen value of
bicarbonates, the trend is Na+ < K+ < Cs+.35 Additionally, 50 cm3 deionized water shows that this method of regeneration
differences in ionic radius, ion hydration, and ion diffusivity is possible but also far from optimized. Later work by the same
have all been suggested to affect the rate of cation and water group cast doubt on cathode rinsing as a viable long-term
transport to the cathode surface and the energies required to technique for removing precipitates since significant pressure is
nucleate and grow a carbonate salt.4,49 These effects of cation necessary to penetrate the hydrophobic cathode and effectively
identity on salt precipitate morphology merit further clean out the precipitated salts.20 Currently, carbon-based
investigation and have yet to be shown in an MEA architecture. GDEs commonly used for CO2RR are only mechanically
robust enough to withstand pressure differences up to 100
2. ACTIVE DISSOLUTION APPROACH: ADDING mbar prior to flooding.41,50 Moreover, droplets that remain in
SOLVENTS TO THE CATHODE the GDL after rinsing can promote HER and limit the free
The second approach to reduce the consequences of salt accessibility of CO2 to the catalyst. The two aforementioned
precipitation works by actively adding solvents to the cathode effects indicate the limited feasibility of dissolution as a viable
region to dissolve and remove precipitates and elevated salt technique to overcome salt formation.
concentrations from near the cathode surface. While Instead of using water to periodically dissolve and flush out
preventing salt formation is ideal, this second strategy already formed salts, increasing the water availability has also
demonstrates how operational performance can be regained been shown as a technique to prevent salt precipitation. In one
325 https://doi.org/10.1021/acsenergylett.2c01885
ACS Energy Lett. 2023, 8, 321−331
ACS Energy Letters http://pubs.acs.org/journal/aelccp Perspective

Figure 3. (a) Plot of cathode concentration versus time showing the general trends of K+ and CO32− concentrations at the cathode during
active flushing of the cathode compartment with water. (b) Schematic depiction of actively mitigating the buildup of ions and nucleation of
crystal seeds on the catalyst by dissolving and removing salt from the cathode with water.

Figure 4. (a) Plots of cathode concentration and cell voltage versus time showing the general trends of K+ and CO32− concentrations at the
cathode during pulsed electrolysis. (b) Schematic depiction of ion-transport during a “pulse” of lower voltage. At the lower regeneration
voltage, the reaction slows down and migration of carbonates and K+ allow the system to partially homogenize before returning to the
operational voltage.

case, De Mot et al. introduced more liquid water to a Sn-based In a separate work, Wheeler et al. humidified the cathode gas
MEA for formate production by injecting a constant stream of feed to reduce the formation of salt precipitates.12 When water
water with the cathode gas feed.40 The water injection rate was is supplied through the gas stream, less water is drawn across
calculated by conducting a water balance on the cathode the anion exchange membrane to facilitate CO2RR. This
compartment, and the authors determined 0.15 mL/min of means that co-ion transport of K+ across the membrane is
additional water was necessary to prevent salt precipitation. reduced, mitigating the accumulation of K+ at the cathode.
This calculation was in good agreement with their However, Mardle et al. noted that humidifying the gas feed
experimental results which found that at 0.2 mL/min of lowers selectivity for CO2RR at higher current densities
water injection, there was no visible salt formation within 1 h
because of flooding of the cathode. Thus, water management is
(although potassium was detected in the electrode pores by
key to not only CO 2 RR performance but also salt
ICP-MS). For comparison, at a 0.1 mL/min water injection
rate, the MEA failed after 50 min because of salt precipitation. precipitation.51 Conversely, others suggest that salt formation
Further increasing the water injection rate decreased the is initially caused by flooding of the electrolyte into the GDE
amount of K+ detected in the cathode GDE but also diluted and then drying of the electrolyte to leave behind salt crystals
the concentration of formate in the product stream. Typically, that subsequently pump more liquid into the GDE.36 So the
concentrated product streams are desired for downstream question remains whether salt formation in the cathode GDE is
processing steps, so this work highlights the potential negative caused by flooding and drying of the electrolyte, by salt crystals
impact of water (and salt) management schemes on product first forming and then pulling liquid in to flood the electrode,
dilution. or a combination of both processes.6,52
326 https://doi.org/10.1021/acsenergylett.2c01885
ACS Energy Lett. 2023, 8, 321−331
ACS Energy Letters http://pubs.acs.org/journal/aelccp Perspective

Figure 5. (a) Plot of cathode concentration versus time showing the general trends of K+ and CO32− concentrations at the cathode when a
BPM is used. (b) Schematic showing effects of a BPM on K+-comigration past the membrane by limiting free ion transport and electro-
osmotic drag. Additionally, CO32− concentrations are reduced by combining with H+ formed at the BPM junction to regenerate CO2. While
changing the MEA recipe delays the accumulation of ions, it does not necessarily prevent critical concentrations from being reached.

The examples discussed above all use a liquid anolyte responsible for the long-term stability of the pulsed electro-
containing KOH or KHCO3 and rely on introducing more lyzer. These works indicate that active manipulation of applied
water to the cathode to flush out salts or limit co-ion current or voltage are viable methods of controlling the pH
migration. Recently, Endrő di et al. have successfully mitigated and ion distribution in an MEA to mitigate salt precipitation.
salt precipitation by taking the opposite approach: feeding the Due to the low number of case studies on altering
cell with a pure water anolyte and periodically “activating” the operational and regenerative voltages as well as cycle durations,
cathode by injecting a small volume of alkali cation containing there is plenty of room for further investigation using this
solutions (10 cm3 of 0.5 M KOH) into the cathode feed.20 approach. To complement the relevance of this direction of
These solutions were 1:3 isopropanol/water mixtures (to help research, future CO2 electrolyzers are likely required to operate
the solution penetrate the hydrophobic GDE) and were intermittently to account for fluctuating power generation from
injected every 12 h of operation. At a constant cell potential of renewable sources.53 However, there may then be too many
3.2 V, initial introduction of the activation solution increased operational constraints from both the electrolyzer and system
jCO from 120 mA cm−2 to 350 mA cm−2. Over the course of perspective to optimize both fully.54
224 h, jCO stabilized to 420 ± 50 mA cm−2 and no salt
precipitation was observed in the cells; however, stable 4. PASSIVE MEMBRANE APPROACH: MEMBRANES
operation over thousands of hours using this technique has AND MATERIALS
not yet been demonstrated. The previous three approaches, while viable to maintain steady
operation, all allowed for the excess formation of carbonate
3. ACTIVE PULSE APPROACH: PULSED species. The operational approaches then provide an engineer-
ELECTROLYSIS ing solution rather than a fundamental solution to the problem
of salt formation. The final approach described here aims to
A third approach to overcome salt precipitation is the use of a reconvert any formed (bi)carbonates back into CO2 by
periodic regeneration voltage to redistribute ions within the providing protons to the cathode chamber through the use
MEA. In this approach the device voltage is ramped up and of a bipolar membrane (BPM) instead of a monopolar
down in a predefined duty cycle, which lowers the operating membrane (Figure 5).55−58 Such an approach then adjusts the
current density and temporarily reduces the formation of physical and chemical components of the MEA itself which
byproduct hydroxide. During the lower voltage cycle the differs from the previous operational approaches.
transport of ions in the system is maintained, however (Figure
4). Migration of K+ from the anolyte is then decreased, while
CO32− has additional time to move to the anode, collectively The operational approaches provide an
decreasing the concentrations of both ions and preventing salt engineering solution rather than a
formation. fundamental solution to the problem
Xu et al. demonstrated the benefits of a recurring of salt formation.
regeneration step where the potential was alternated between
−3.8 Vcell during operation and −2.0 Vcell during regeneration.
Stable operation was maintained for 236 h (out of which 157 h A BPM is composed of both a cation exchange layer (CEL)
were at an operational voltage).44 When the same setup ran and an anion exchange layer (AEL) that are affixed to one
without a regeneration voltage, the system broke down after another. Upon the application of a reversed bias (where the
∼10 h due to salt formation. Subsequent modeling of these cation exchange layer is closest to the cathode and the anion
two systems indicated that electromigration (instead of exchange layer is closest to the anode), water inside the
diffusion) of carbonate ions during the regeneration step is membrane is split into H+ and OH− molecules which migrate
327 https://doi.org/10.1021/acsenergylett.2c01885
ACS Energy Lett. 2023, 8, 321−331
ACS Energy Letters http://pubs.acs.org/journal/aelccp Perspective

to the cathode and anode, respectively. By using a BPM in a from the cathode and protons from an acidic anolyte
MEA for CO2 electrolysis, salt formation is then reduced recombine. Such an approach, while preventing salt precip-
through two different approaches. First, the H+ generated in itation through the use of an acidic anolyte, causes gas
the BPM migrates to the cathode and chemical interacts with evolution in the middle of the membrane.68 In principle, a
(bi)carbonates to regenerate CO2, effectively offsetting the monopolar CEM can also be used to transport H+ ions toward
hydroxide that was generated in eq 1.59−61 And second, as H+ the cathode using an acidic anolyte solution devoid of cations.
becomes the primary charge carrier, the migration of the co- The acidity of the cathode needs to be balanced, however, to
ion K+ from the anolyte is greatly reduced. Both [K+] and avoid excessive proton concentrations which would cause HER
[CO32−] are then reduced using a BPM operating in reversed to dominate CO2 electrolysis.69 Additionally cations are likely
bias as compared to a monopolar membrane. A large factor in necessary for CO2 electrolysis to outcompete hydrogen
the success of using BPM’s to prevent salt formation resides in evolution.
the ability to prevent co-ion crossover of potassium from the A common challenge for CO2RR research is controlling the
anolyte. Such BPM properties have been examined by environment close to the catalyst such that the core
Blommaert et al., who showed that under reversed-bias performance metrics of voltage, current density, selectivity,
conditions water dissociation will dominate K+ co-ion and stability can all be maintained. The issue of salt
crossover at current densities >10 mA cm−2.62 In fact, beyond precipitation in MEA systems is no exception and requires
current densities of 1 mA cm−2, the flux of K+ was shown to be consideration of the electrochemical and chemical reactions
fixed almost independent of the applied current density and occurring in the system, as well as mass transport within each
constituted less than 3% of the charge transported across the component. Herein we identified several mechanisms that lead
membrane. Thus, a BPM is likely to greatly slow salt to salt formation and reviewed four operational techniques for
precipitation by limiting potassium transport to the cathode salt precipitation prevention in neutral and alkaline CO2RR
but by itself will not clearly avoid precipitation. MEAs, all with the goal of lowering cation and/or carbonate
In literature, the reversed-bias BPM approach has been used concentrations near the cathode.
in a number of scenarios with the primary intent to increase The outlook for each of the presented approaches are
CO2 utilization within CO2 electrolysis systems.21,56,63−65 If a promising given the relatively few papers that have tried to
higher fraction of CO2 is used for the electrochemical reaction, directly address salt formation, leaving room for greater
then less CO2 can be permanently converted into carbonate advancements. For example, there remains a large amount of
salts. Interestingly, the BPM configuration does not avoid operating conditions left to be tested, and combining a subset
carbonate formation which could lead to salt formation but of the approaches above is likely to allow for salt formation
instead provides a means of neutralizing the formed carbonate failure to be prevented indefinitely. It is also worth noting that
with protons prior to the anions migrating through the cation the challenges associated with salt and carbonate formation
and anion exchange layers. The BPM approach has then were only noted a few years prior to this article, and there are
allowed for stable operation of >1255 and 2456 hours in two now several proposed solutions, highlighting progress in a
examples, with several others reporting much more stable short period of time. Notably for each of the presented cases,
operation than with anion or cation exchange membranes however, is that system stability was improved at the cost of
alone.21,64,65 The use of BPM’s in reversed bias, however, has decreases in other performance metrics. For example,
an associated energy cost. Specifically, BPMs require increased decreasing the anolyte concentration or using a BPM is
potentials to dissociate water at the anion and cation exchange penalized by higher cell voltages, while periodic operation
membrane interface. Additionally, the presence of two lowers the capacity factor of the electrolyzer. Future work then
membranes causes greater charge ohmic resistance than a needs to evaluate which trade-offs are acceptable at the
singular thinner membrane. Further designs of effective bipolar expense of other metrics.
membranes might help in overcoming the higher cell voltages Looking to the future, we note that operational strategies are
encountered in commercially available BPMs.65 Promising not the only methods available to stop salt precipitation, and
results by Oener et al. for example showed that optimization of we expect materials selection and development to also play a
the BPM through lower thickness, increased AEM/CEM role. Recent reports in flow cells have demonstrated the ability
interface area and an additional water dissociation catalyst of ionomer binders, monolayers, and bilayers to control local
inserted at the cation and anion junction led to overpotentials concentrations of ions in the catalyst layer and influence salt
as low as 10 mV (at 20 mA/cm2).66 As further work continues precipitation.5,70 When developing solutions to overcome salt
on BPMs, their potential to reach elevated current densities at precipitation for CO2RR, researchers can also look to other
lower overpotentials is expected then to increase. fields for inspiration. Research on durable membranes for
A secondary issue with using BPMs in reversed-bias is that water filtration applications has extensively studied material
the cathode conditions become acidic. Without proper control design strategies (i.e., controlling surface charge, roughness,
of the cathode pH and mobile cations, hydrogen evolution can hydrophobicity, etc.) to mitigate membrane fouling by
then outcompete CO2 reduction. Some approaches have used inorganic salts (primarily CaCO3, SiO2, and BaSO4).71 The
a weakly acidic buffer layer to increase the pH to a point where formation of carbonate salts at gas−liquid−solid boundaries is
CO2 electrolysis is favorable again.56,65,67 Such results motivate also of interest to geological carbon storage applications
further reinvestigation into acidic CO2 electrolysis catalysts. whereby CO2 is injected into saline aquifers for sequestra-
While less common, systems for CO2 conversion have also tion.72 Further study into salt nucleation and growth
considered using BPMs in a forward-bias configuration. When mechanisms under CO2RR conditions by operando or in situ
operating a BPM in forward bias (where the AEM is pressed characterization techniques (i.e., atomic force microscopy,
against the cathode compartment instead) some energy can be nano- or microcomputed tomography, X-ray diffraction, etc.)
recovered by the recombination of ions at the cation and anion will also inform the development of both materials-based and
exchange junction. Here CO2 can be regenerated as carbonate operational salt prevention strategies. Lastly, the emerging field
328 https://doi.org/10.1021/acsenergylett.2c01885
ACS Energy Lett. 2023, 8, 321−331
ACS Energy Letters http://pubs.acs.org/journal/aelccp Perspective

of CO2RR under acidic conditions provides another avenue to Author Contributions

avoid the issue of salt precipitation entirely, as the reduction of M.S. and T.B. conceived the project. M.K., S.S., and W.A.S.
hydroxide concentrations leads to significantly less homoge- assisted with the literature review, categorization, and scientific
neous formation of salts.73 Future works should weigh the discussion. M.K. designed all figures. M.S., M.K., and S.S.
advantages and disadvantages of CO2RR in acidic and alkaline wrote the manuscript with contribution and editing from all
environments to determine which system is most desirable for coauthors.
a durable and selective electrolyzer operating at industrially Notes
relevant current densities, that also maintain overall high The authors declare no competing financial interest.
energy and carbon efficiency. Biographies
Salt precipitation is one of the major limitations for the
Mark Sassenburg is a Ph.D. graduate student in the Department of
selective and long-term operation of neutral and alkaline MEA
Chemical Engineering at the Delft University of Technology under
CO2RR electrolyzers. This issue is difficult to avoid because
the supervision of Wilson Smith and Thomas Burdyny. While
the three essential components for CO2RR (CO2 gas, a proton
performing empirical experiments alongside material characterization,
source, and an alkali cation) also directly influence the local he attempts to further understand the effects of operational
concentrations of ions that can precipitate into salt deposits. parameters on CO2 electroreduction catalysts.
Here mechanisms for salt formation are discussed, and four
operational approaches to prevent or reverse salt precipitation Maria Kelly is pursuing her Ph.D. in Chemical Engineering at the
are presented, which can be broken down into either passive University of Colorado, Boulder, and the National Renewable Energy
Laboratory under the supervision of Prof. Wilson Smith. Her primary
system changes or active mediation. Several of these strategies
research focuses on using atomic force microscopy to study CO2
are successful over the course of tens to hundreds of hours;
electroreduction microenvironments.
however, none demonstrate selective system operation on the
order of tens of thousands of hours. We encourage researchers Siddhartha Subramanian is pursuing his Ph.D. in Chemical
to report longer term electrolysis studies using these salt Engineering at the Delft University of Technology under the
precipitation prevention methods and analyze their feasibility supervision of Dr. Thomas Burdyny. Using a combination of
for commercial systems. A combination of operational experiments and transport modelling, his research focuses on
solutions will likely need to be deployed to solve the salt understanding spatial effects in zero gap CO2 electrolyzers at
industrially relevant reaction rates.
precipitation problem.
Wilson A. Smith started his independent career as an Assistant and

■ AUTHOR INFORMATION
Corresponding Author
Associate Professor of Chemical Engineering at the Delft University of
Technology. Wilson is currently a Professor of Chemical Engineering
at the University of Colorado, Boulder, and Senior Scientist at the
Thomas Burdyny − Materials for Energy Conversion and National Renewable Energy Lab, where his work focuses on
Storage (MECS), Department of Chemical Engineering, Delft electrochemical approaches to carbon capture and conversion.
University of Technology, 2629 ZH Delft, The Netherlands; Thomas Burdyny is an Assistant Professor in Chemical Engineering
orcid.org/0000-0001-8057-9558; Email: t.e.burdyny@ at the Delft University of Technology researching new electro-
tudelft.nl chemical energy technologies, with expertise CO2 electrolysis and gas-
Authors diffusion electrodes. His research interests span from nanoscale
phenomena to process scales with motivations to advance a
Mark Sassenburg − Materials for Energy Conversion and
technology’s industrial viability.
Storage (MECS), Department of Chemical Engineering, Delft
University of Technology, 2629 ZH Delft, The Netherlands;
orcid.org/0000-0002-2826-7765
Maria Kelly − Department of Chemical and Biological
■ ACKNOWLEDGMENTS
M.S. acknowledges the Electrons to Chemical Bonds (E2CB,
Engineering and Renewable and Sustainable Energy Institute NWO project number P17-09-01) research programme. T.B.
(RASEI), University of Colorado Boulder, Boulder, Colorado thanks the NWO for funding in the form of a Veni grant
80303, United States; National Renewable Energy (17337). M.K. acknowledges funding from the National
Laboratory, Golden, Colorado 80401, United States Science Foundation Graduate Research Fellowship under
Siddhartha Subramanian − Materials for Energy Conversion Grant DGE2040434. M.K. and W.S. acknowledge funding
and Storage (MECS), Department of Chemical Engineering, from the Liquid Sunlight Alliance, which is supported by the
Delft University of Technology, 2629 ZH Delft, The U.S. Department of Energy, Office of Science, Office of Basic
Netherlands; orcid.org/0000-0002-7992-3849 Energy Sciences, Fuels from Sunlight Hub under Award DE-
Wilson A. Smith − Materials for Energy Conversion and SC0021266. S.S. and T.B. would like to acknowledge the
Storage (MECS), Department of Chemical Engineering, Delft cofinancing provided by Shell and a PPP-allowance from Top
University of Technology, 2629 ZH Delft, The Netherlands; Consortia for Knowledge and Innovation (TKI’s) of the
Department of Chemical and Biological Engineering and Ministry of Economic Affairs and Climate in the context of the
Renewable and Sustainable Energy Institute (RASEI), TU Delft e-Refinery Institute.
University of Colorado Boulder, Boulder, Colorado 80303,
United States; National Renewable Energy Laboratory,
Golden, Colorado 80401, United States; orcid.org/0000-
0001-7757-5281
■ REFERENCES
(1) Buttler, A.; Spliethoff, H. Current Status of Water Electrolysis for
Energy Storage, Grid Balancing and Sector Coupling via Power-to-
Complete contact information is available at: Gas and Power-to-Liquids: A Review. Renew. Sustain. Energy Rev.
https://pubs.acs.org/10.1021/acsenergylett.2c01885 2018, 82, 2440−2454.

329 https://doi.org/10.1021/acsenergylett.2c01885
ACS Energy Lett. 2023, 8, 321−331
ACS Energy Letters http://pubs.acs.org/journal/aelccp Perspective

(2) Lv, J.-J.; Jouny, M.; Luc, W.; Zhu, W.; Zhu, J.-J.; Jiao, F. A Highly (19) Lu, X.; Zhu, C.; Wu, Z.; Xuan, J.; Francisco, J. S.; Wang, H. In
Porous Copper Electrocatalyst for Carbon Dioxide Reduction. Adv. Situ Observation of the pH Gradient near the Gas Diffusion Electrode
Mater. 2018, 30 (49), 1803111. of CO2 Reduction in Alkaline Electrolyte. J. Am. Chem. Soc. 2020, 142
(3) Verma, S.; Hamasaki, Y.; Kim, C.; Huang, W.; Lu, S.; Jhong, H.- (36), 15438−15444.
R. M.; Gewirth, A. A.; Fujigaya, T.; Nakashima, N.; Kenis, P. J. A. (20) Endrő di, B.; Samu, A.; Kecsenovity, E.; Halmágyi, T.; Sebő k,
Insights into the Low Overpotential Electroreduction of CO2 to CO D.; Janáky, C. Operando Cathode Activation with Alkali Metal
on a Supported Gold Catalyst in an Alkaline Flow Electrolyzer. ACS Cations for High Current Density Operation of Water-Fed Zero-Gap
Energy Lett. 2018, 3 (1), 193−198. Carbon Dioxide Electrolysers. Nat. Energy 2021, 6 (4), 439−448.
(4) Cofell, E. R.; Nwabara, U. O.; Bhargava, S. S.; Henckel, D. E.; (21) Yang, K.; Li, M.; Subramanian, S.; Blommaert, M. A.; Smith, W.
Kenis, P. J. A. Investigation of Electrolyte-Dependent Carbonate A.; Burdyny, T. Cation-Driven Increases of CO2 Utilization in a
Formation on Gas Diffusion Electrodes for CO2 Electrolysis. ACS Bipolar Membrane Electrode Assembly for CO2 Electrolysis. ACS
Appl. Mater. Interfaces 2021, 13 (13), 15132−15142. Energy Lett. 2021, 6 (12), 4291−4298.
(5) Nwabara, U. O.; Hernandez, A. D.; Henckel, D. A.; Chen, X.; (22) Vennekötter, J.-B.; Scheuermann, T.; Sengpiel, R.; Wessling, M.
Cofell, E. R.; de-Heer, M. P.; Verma, S.; Gewirth, A. A.; Kenis, P. J. A. The Electrolyte Matters: Stable Systems for High Rate Electro-
Binder-Focused Approaches to Improve the Stability of Cathodes for chemical CO2 Reduction. J. CO2 Util. 2019, 32, 202−213.
CO2 Electroreduction. ACS Appl. Energy Mater. 2021, 4 (5), 5175− (23) Schröder, D.; Sinai Borker, N. N.; König, M.; Krewer, U.
5186. Performance of Zinc Air Batteries with Added K2CO3 in the Alkaline
(6) Pham, T. H. M.; Zhang, J.; Li, M.; Shen, T.-H.; Ko, Y.; Tileli, V.; Electrolyte. J. Appl. Electrochem. 2015, 45 (5), 427−437.
Luo, W.; Züttel, A. Enhanced Electrocatalytic CO2 Reduction to C2+ (24) Naughton, M. S.; Brushett, F. R.; Kenis, P. J. A. Carbonate
Products by Adjusting the Local Reaction Environment with Polymer Resilience of Flowing Electrolyte-Based Alkaline Fuel Cells. J. Power
Binders. Adv. Energy Mater. 2022, 12, 2103663. Sources 2011, 196 (4), 1762−1768.
(7) de Jesus Gálvez-Vázquez, M.; Moreno-García, P.; Xu, H.; Hou, (25) Wuttig, A.; Yoon, Y.; Ryu, J.; Surendranath, Y. Bicarbonate Is
Y.; Hu, H.; Montiel, I. Z.; Rudnev, A. V.; Alinejad, S.; Grozovski, V.; Not a General Acid in Au-Catalyzed CO2 Electroreduction. J. Am.
Wiley, B. J.; Arenz, M.; Broekmann, P. Environment Matters: CO2RR Chem. Soc. 2017, 139 (47), 17109−17113.
Electrocatalyst Performance Testing in a Gas-Fed Zero-Gap Electro- (26) Zeng, J. S.; Corbin, N.; Williams, K.; Manthiram, K. Kinetic
lyzer. ACS Catal. 2020, 10 (21), 13096−13108. Analysis on the Role of Bicarbonate in Carbon Dioxide Electro-
(8) Dinh, C.-T.; Burdyny, T.; Kibria, M. G.; Seifitokaldani, A.; reduction at Immobilized Cobalt Phthalocyanine. ACS Catal. 2020,
Gabardo, C. M.; García de Arquer, F. P.; Kiani, A.; Edwards, J. P.; De 10 (7), 4326−4336.
Luna, P.; Bushuyev, O. S.; Zou, C.; Quintero-Bermudez, R.; Pang, Y.; (27) Chen, Y.; Wrubel, J. A.; Vise, A. E.; Intia, F.; Harshberger, S.;
Sinton, D.; Sargent, E. H. CO2 Electroreduction to Ethylene via Klein, E.; Smith, W. A.; Ma, Z.; Deutsch, T. G.; Neyerlin, K. C. The
Hydroxide-Mediated Copper Catalysis at an Abrupt Interface. Science
Effect of Catholyte and Catalyst Layer Binders on CO2 Electro-
2018, 360 (6390), 783−787.
reduction Selectivity. Chem. Catal. 2022, 2 (2), 400−421.
(9) Garg, S.; Li, M.; Weber, A. Z.; Ge, L.; Li, L.; Rudolph, V.; Wang,
(28) Chen, L. D. Cations Play an Essential Role in CO2 Reduction.
G.; Rufford, T. E. Advances and Challenges in Electrochemical CO2
Nat. Catal. 2021, 4 (8), 641−642.
Reduction Processes: An Engineering and Design Perspective
(29) Monteiro, M. C. O.; Dattila, F.; Hagedoorn, B.; García-Muelas,
Looking beyond New Catalyst Materials. J. Mater. Chem. A 2020, 8
R.; López, N.; Koper, M. T. M. Absence of CO2 Electroreduction on
(4), 1511−1544.
Copper, Gold and Silver Electrodes without Metal Cations in
(10) Burdyny, T.; Smith, W. A. CO2 Reduction on Gas-Diffusion
Electrodes and Why Catalytic Performance Must Be Assessed at Solution. Nat. Catal. 2021, 4, 654−662.
(30) Moura de Salles Pupo, M.; Kortlever, R. Electrolyte Effects on
Commercially-Relevant Conditions. Energy Environ. Sci. 2019, 12 (5),
1442−1453. the Electrochemical Reduction of CO2. ChemPhysChem 2019, 20
(11) Ge, L.; Rabiee, H.; Li, M.; Subramanian, S.; Zheng, Y.; Lee, J. (22), 2926−2935.
H.; Burdyny, T.; Wang, H. Electrochemical CO2 Reduction in (31) Rabinowitz, J. A.; Kanan, M. W. The Future of Low-
Membrane-Electrode Assemblies. Chem. 2022, 8, 663. Temperature Carbon Dioxide Electrolysis Depends on Solving One
(12) Wheeler, D. G.; Mowbray, B. A. W.; Reyes, A.; Habibzadeh, F.; Basic Problem. Nat. Commun. 2020, 11 (1), 5231.
He, J.; Berlinguette, C. P. Quantification of Water Transport in a CO2 (32) Ma, M.; Clark, E. L.; Therkildsen, K. T.; Dalsgaard, S.;
Electrolyzer. Energy Environ. Sci. 2020, 13 (12), 5126−5134. Chorkendorff, I.; Seger, B. Insights into the Carbon Balance for CO2
(13) Wang, G.; Pan, J.; Jiang, S. P.; Yang, H. Gas Phase Electroreduction on Cu Using Gas Diffusion Electrode Reactor
Electrochemical Conversion of Humidified CO2 to CO and H2 on Designs. Energy Environ. Sci. 2020, 13 (3), 977−985.
Proton-Exchange and Alkaline Anion-Exchange Membrane Fuel Cell (33) Weng, L.-C.; Bell, A. T.; Weber, A. Z. Towards Membrane-
Reactors. J. CO2 Util. 2018, 23, 152−158. Electrode Assembly Systems for CO2 Reduction: A Modeling Study.
(14) Shironita, S.; Karasuda, K.; Sato, M.; Umeda, M. Feasibility Energy Environ. Sci. 2019, 12 (6), 1950−1968.
Investigation of Methanol Generation by CO2 Reduction Using Pt/C- (34) Kas, R.; Star, A. G.; Yang, K.; Van Cleve, T.; Neyerlin, K. C.;
Based Membrane Electrode Assembly for a Reversible Fuel Cell. J. Smith, W. A. Along the Channel Gradients Impact on the
Power Sources 2013, 228, 68−74. Spatioactivity of Gas Diffusion Electrodes at High Conversions
(15) Pérez-Rodríguez, S.; Barreras, F.; Pastor, E.; Lázaro, M. J. during CO2 Electroreduction. ACS Sustain. Chem. Eng. 2021, 9 (3),
Electrochemical Reactors for CO2 Reduction: From Acid Media to 1286−1296.
Gas Phase. Int. J. Hydrog. Energy 2016, 41 (43), 19756−19765. (35) CRC Handbook of Chemistry and Physics, 103rd ed.; Rumble, Jr.,
(16) Kriescher, S. M. A.; Kugler, K.; Hosseiny, S. S.; Gendel, Y.; J. R., Ed.; CRC Press/Taylor & Francis: Boca Raton, FL, 2022.
Wessling, M. A Membrane Electrode Assembly for the Electro- (36) Leonard, M. E.; Clarke, L. E.; Forner-Cuenca, A.; Brown, S. M.;
chemical Synthesis of Hydrocarbons from CO2(g) and H2O(g). Brushett, F. R. Investigating Electrode Flooding in a Flowing
Electrochem. Commun. 2015, 50, 64−68. Electrolyte, Gas-Fed Carbon Dioxide Electrolyzer. ChemSusChem
(17) Bohra, D.; Chaudhry, J. H.; Burdyny, T.; Pidko, E. A.; Smith, 2020, 13 (2), 400−411.
W. A. Modeling the Electrical Double Layer to Understand the (37) Garg, S.; Giron Rodriguez, C. A.; Rufford, T. E.; Varcoe, J. R.;
Reaction Environment in a CO2 Electrocatalytic System. Energy Seger, B. How Membrane Characteristics Influence the Performance
Environ. Sci. 2019, 12 (11), 3380−3389. of CO2 and CO Electrolysis. Energy Environ. Sci. 2022, 15, 4440.
(18) Yang, K.; Kas, R.; Smith, W. A. In Situ Infrared Spectroscopy (38) Lees, E. W.; Bui, J. C.; Song, D.; Weber, A. Z.; Berlinguette, C.
Reveals Persistent Alkalinity near Electrode Surfaces during CO2 P. Continuum Model to Define the Chemistry and Mass Transfer in a
Electroreduction. J. Am. Chem. Soc. 2019, 141 (40), 15891−15900. Bicarbonate Electrolyzer. ACS Energy Lett. 2022, 7 (2), 834−842.

330 https://doi.org/10.1021/acsenergylett.2c01885
ACS Energy Lett. 2023, 8, 321−331
ACS Energy Letters http://pubs.acs.org/journal/aelccp Perspective

(39) Hawks, S. A.; Ehlinger, V. M.; Moore, T.; Duoss, E. B.; Beck, V. CO in a Flow Cell with a Bipolar Membrane. ACS Energy Lett. 2018,
A.; Weber, A. Z.; Baker, S. E. Analyzing Production Rate and Carbon 3 (1), 149−154.
Utilization Trade-Offs in CO2RR Electrolyzers. ACS Energy Lett. (57) Chen, Y.; Vise, A.; Klein, W. E.; Cetinbas, F. C.; Myers, D. J.;
2022, 7 (8), 2685−2693. Smith, W. A.; Deutsch, T. G.; Neyerlin, K. C. A Robust, Scalable
(40) De Mot, B.; Ramdin, M.; Hereijgers, J.; Vlugt, T. J. H.; Platform for the Electrochemical Conversion of CO2 to Formate:
Breugelmans, T. Direct Water Injection in Catholyte-Free Zero-Gap Identifying Pathways to Higher Energy Efficiencies. ACS Energy Lett.
Carbon Dioxide Electrolyzers. ChemElectroChem. 2020, 7 (18), 2020, 5 (6), 1825−1833.
3839−3843. (58) Yang, H.; Kaczur, J. J.; Sajjad, S. D.; Masel, R. I. Performance
(41) Baumgartner, L. M.; Koopman, C. I.; Forner-Cuenca, A.; and Long-Term Stability of CO2 Conversion to Formic Acid Using a
Vermaas, D. A. Narrow Pressure Stability Window of Gas Diffusion Three-Compartment Electrolyzer Design. J. CO2 Util. 2020, 42,
Electrodes Limits the Scale-Up of CO2 Electrolyzers. ACS Sustain. 101349.
Chem. Eng. 2022, 10 (14), 4683−4693. (59) Li, Y. C.; Lee, G.; Yuan, T.; Wang, Y.; Nam, D.-H.; Wang, Z.;
(42) Endrő di, B.; Kecsenovity, E.; Samu, A.; Darvas, F.; Jones, R. V.; García de Arquer, F. P.; Lum, Y.; Dinh, C.-T.; Voznyy, O.; Sargent, E.
Török, V.; Danyi, A.; Janáky, C. Multilayer Electrolyzer Stack H. CO2 Electroreduction from Carbonate Electrolyte. ACS Energy
Converts Carbon Dioxide to Gas Products at High Pressure with Lett. 2019, 4 (6), 1427−1431.
High Efficiency. ACS Energy Lett. 2019, 4 (7), 1770−1777. (60) Ma, M.; Kim, S.; Chorkendorff, I.; Seger, B. Role of Ion-
(43) Liu, Z.; Yang, H.; Kutz, R.; Masel, R. I. CO2 Electrolysis to CO Selective Membranes in the Carbon Balance for CO2 Electro-
and O2 at High Selectivity, Stability and Efficiency Using Sustainion reduction via Gas Diffusion Electrode Reactor Designs. Chem. Sci.
Membranes. J. Electrochem. Soc. 2018, 165 (15), J3371. 2020, 11 (33), 8854−8861.
(44) Xu, Y.; Edwards, J. P.; Liu, S.; Miao, R. K.; Huang, J. E.; (61) Kim, J. Y. ‘T.’; Zhu, P.; Chen, F.-Y.; Wu, Z.-Y.; Cullen, D. A.;
Wang, H. Recovering Carbon Losses in CO2 Electrolysis Using a
Gabardo, C. M.; O’Brien, C. P.; Li, J.; Sargent, E. H.; Sinton, D. Self-
Solid Electrolyte Reactor. Nat. Catal. 2022, 5 (4), 288−299.
Cleaning CO2 Reduction Systems: Unsteady Electrochemical Forcing
(62) Blommaert, M. A.; Verdonk, J. A. H.; Blommaert, H. C. B.;
Enables Stability. ACS Energy Lett. 2021, 6 (2), 809−815.
Smith, W. A.; Vermaas, D. A. Reduced Ion Crossover in Bipolar
(45) Yin, Z.; Peng, H.; Wei, X.; Zhou, H.; Gong, J.; Huai, M.; Xiao,
Membrane Electrolysis via Increased Current Density, Molecular Size,
L.; Wang, G.; Lu, J.; Zhuang, L. An Alkaline Polymer Electrolyte CO2 and Valence. ACS Appl. Energy Mater. 2020, 3 (6), 5804−5812.
Electrolyzer Operated with Pure Water. Energy Environ. Sci. 2019, 12 (63) Blommaert, M. A.; Subramanian, S.; Yang, K.; Smith, W. A.;
(8), 2455−2462. Vermaas, D. A. High Indirect Energy Consumption in AEM-Based
(46) O’Brien, C. P.; Miao, R. K.; Liu, S.; Xu, Y.; Lee, G.; Robb, A.; CO2 Electrolyzers Demonstrates the Potential of Bipolar Membranes.
Huang, J. E.; Xie, K.; Bertens, K.; Gabardo, C. M.; Edwards, J. P.; ACS Appl. Mater. Interfaces 2022, 14 (1), 557−563.
Dinh, C.-T.; Sargent, E. H.; Sinton, D. Single Pass CO2 Conversion (64) Eriksson, B.; Asset, T.; Spanu, F.; Lecoeur, F.; Dupont, M.;
Exceeding 85% in the Electrosynthesis of Multicarbon Products via Garcés-Pineda, F. A.; Galán-Mascarós, J. R.; Cavaliere, S.; Rozière, J.;
Local CO2 Regeneration. ACS Energy Lett. 2021, 6 (8), 2952−2959. Jaouen, F. Mitigation of Carbon Crossover in CO2 Electrolysis by Use
(47) Geise, G. M.; Hickner, M. A.; Logan, B. E. Ionic Resistance and of Bipolar Membranes. J. Electrochem. Soc. 2022, 169 (3), 034508.
Permselectivity Tradeoffs in Anion Exchange Membranes. ACS Appl. (65) Xie, K.; Miao, R. K.; Ozden, A.; Liu, S.; Chen, Z.; Dinh, C.-T.;
Mater. Interfaces 2013, 5 (20), 10294−10301. Huang, J. E.; Xu, Q.; Gabardo, C. M.; Lee, G.; Edwards, J. P.; O’Brien,
(48) Reyes, A.; Jansonius, R. P.; Mowbray, B. A. W.; Cao, Y.; C. P.; Boettcher, S. W.; Sinton, D.; Sargent, E. H. Bipolar Membrane
Wheeler, D. G.; Chau, J.; Dvorak, D. J.; Berlinguette, C. P. Managing Electrolyzers Enable High Single-Pass CO2 Electroreduction to
Hydration at the Cathode Enables Efficient CO2 Electrolysis at Multicarbon Products. Nat. Commun. 2022, 13 (1), 3609.
Commercially Relevant Current Densities. ACS Energy Lett. 2020, 5 (66) Oener, S. Z.; Foster, M. J.; Boettcher, S. W. Accelerating Water
(5), 1612−1618. Dissociation in Bipolar Membranes and for Electrocatalysis. Science
(49) Chiacchiarelli, L. M.; Zhai, Y.; Frankel, G. S.; Agarwal, A. S.; 2020, 369 (6507), 1099−1103.
Sridhar, N. Cathodic Degradation Mechanisms of Pure Sn Electro- (67) Yan, Z.; Hitt, J. L.; Zeng, Z.; Hickner, M. A.; Mallouk, T. E.
catalyst in a Nitrogen Atmosphere. J. Appl. Electrochem. 2012, 42 (1), Improving the Efficiency of CO2 Electrolysis by Using a Bipolar
21−29. Membrane with a Weak-Acid Cation Exchange Layer. Nat. Chem.
(50) Legrand, U.; Lee, J. K.; Bazylak, A.; Tavares, J. R. Product 2021, 13 (1), 33−40.
Crossflow through a Porous Gas Diffusion Layer in a CO2 (68) Xu, Y.; Miao, R. K.; Edwards, J. P.; Liu, S.; O’Brien, C. P.;
Electrochemical Cell with Pressure Drop Calculations. Ind. Eng. Gabardo, C. M.; Fan, M.; Huang, J. E.; Robb, A.; Sargent, E. H.;
Chem. Res. 2021, 60 (19), 7187−7196. Sinton, D. A Microchanneled Solid Electrolyte for Carbon-Efficient
(51) Mardle, P.; Cassegrain, S.; Habibzadeh, F.; Shi, Z.; Holdcroft, S. CO2 Electrolysis. Joule 2022, 6 (6), 1333−1343.
Carbonate Ion Crossover in Zero-Gap, KOH Anolyte CO 2 (69) Shafaque, H. W.; Lee, C.; Fahy, K. F.; Lee, J. K.; LaManna, J.
Electrolysis. J. Phys. Chem. C 2021, 125 (46), 25446−25454. M.; Baltic, E.; Hussey, D. S.; Jacobson, D. L.; Bazylak, A. Boosting
(52) Li, M.; Idros, M. N.; Wu, Y.; Burdyny, T.; Garg, S.; Zhao, X. S.; Membrane Hydration for High Current Densities in Membrane
Wang, G.; Rufford, T. E. The Role of Electrode Wettability in Electrode Assembly CO2 Electrolysis. ACS Appl. Mater. Interfaces
Electrochemical Reduction of Carbon Dioxide. J. Mater. Chem. A 2020, 12 (49), 54585−54595.
2021, 9 (35), 19369−19409. (70) Bui, J. C.; Digdaya, I.; Xiang, C.; Bell, A. T.; Weber, A. Z.
(53) Samu, A. A.; Kormányos, A.; Kecsenovity, E.; Szilágyi, N.; Understanding Multi-Ion Transport Mechanisms in Bipolar Mem-
Endrő di, B.; Janáky, C. Intermittent Operation of CO2 Electrolyzers at branes. ACS Appl. Mater. Interfaces 2020, 12 (47), 52509−52526.
Industrially Relevant Current Densities. ACS Energy Lett. 2022, 7 (5), (71) Zhao, S.; Liao, Z.; Fane, A.; Li, J.; Tang, C.; Zheng, C.; Lin, J.;
1859−1861. Kong, L. Engineering Antifouling Reverse Osmosis Membranes: A
(54) Huesman, A. Integration of Operation and Design of Solar Fuel Review. Desalination 2021, 499, 114857.
Plants: A Carbon Dioxide to Methanol Case Study. Comput. Chem. (72) Miri, R.; Hellevang, H. Salt Precipitation during CO2 Storage�
A Review. Int. J. Greenh. Gas Control 2016, 51, 136−147.
Eng. 2020, 140, 106836.
(73) Huang, J. E.; Li, F.; Ozden, A.; Sedighian Rasouli, A.; Garcia de
(55) Li, Y. C.; Zhou, D.; Yan, Z.; Gonçalves, R. H.; Salvatore, D. A.;
Arquer, F. P.; Liu, S.; Zhang, S.; Luo, M.; Wang, X.; Lum, Y.; Xu, Y.;
Berlinguette, C. P.; Mallouk, T. E. Electrolysis of CO2 to Syngas in
Bertens, K.; Miao, R. K.; Dinh, C.-T.; Sinton, D.; Sargent, E. H. CO2
Bipolar Membrane-Based Electrochemical Cells. ACS Energy Lett.
Electrolysis to Multicarbon Products in Strong Acid. Science 2021,
2016, 1 (6), 1149−1153. 372 (6546), 1074−1078.
(56) Salvatore, D. A.; Weekes, D. M.; He, J.; Dettelbach, K. E.; Li, Y.
C.; Mallouk, T. E.; Berlinguette, C. P. Electrolysis of Gaseous CO2 to

331 https://doi.org/10.1021/acsenergylett.2c01885
ACS Energy Lett. 2023, 8, 321−331