Anion Exchange Ionomers: Design Considerations and Recent Advances ‐ An Electrochemical Perspective

Alkaline‐based electrochemical devices, such as anion exchange membrane (AEM) fuel cells and electrolyzers, are receiving increasing attention. However, while the catalysts and membrane are methodically studied, the ionomer is largely overlooked. In fact, most of the studies in alkaline electrolytes are conducted using the commercial proton exchange ionomer Nafion. The ionomer provides ionic conductivity; it is also essential for gas transport and water management, as well as for controlling the mechanical stability and the morphology of the catalyst layer. Moreover, the ionomer has distinct requirements that differ from those of anion‐exchange membranes, such as a high gas permeability, and that depend on the specific electrode, such as water management. As a result, it is necessary to tailor the ionomer structure to the specific application in isolation and as part of the catalyst layer. In this review, an overview of the current state of the art for anion exchange ionomers is provided, summarizing their specific requirements and limitations in the context of AEM electrolyzers and fuel cells.


Introduction
Hydrogen is emerging as an important part of the energy transition.In particular, electrochemical technologies, such as electrolyzers and fuel cells are receiving attention as promising candidates for the renewable production of hydrogen and the clean production of electricity, respectively.For the case of fuel cells, the most studied and technologically developed is the proton exchange membrane fuel cell (PEMFC).However, the high requirement of platinum to catalyze the sluggish oxygen reduction reaction has attracted more interest in its anion exchange membrane counterpart (AEMFC) for which transition metals-based catalysts tion reaction) catalyst, ORR catalyst, and AEM have been thoroughly studied, the anion exchange ionomer (AEI) has been largely overlooked.A similar case can be made for anion exchange membrane (AEM) electrolyzers, for which hydrogen is evolved at the cathode, with the co-production of hydroxide ions, which are transported through the ion exchanged membrane to the anode where oxygen is produced (Figure 1a).AEM electrolyzers hold significant advantages compared to their PEM and KOH counterparts, including the opportunity of reaching practical operating current densities with PGM catalysts, with a non-corrosive liquid phase and of using inexpensive flow field materials. [2]However, similar to AEM fuel cells, AEM electrolyzers are also limited by the low conductivity and stability of the membrane and ionomer.IN addition, best AEM electrolyzer performance relies on the addition of free electrolyte (KOH) to the water feed. [2]Although the reasons behind this are not completely understood, it is likely a result of catalyst-ionomer interactions, which highlights the need of expanding the very limited research on anion exchange ionomer.
Significant efforts have been made in recent years to improve the stability and ionic conductivity of anion exchange membranes, leading to the commercialization of a wide variety of AEMs.On the other hand, research of AEI has been more limited and most of the structural optimization has been focused on the membrane side.While the same polymer can be potentially used as membrane and ionomer, AEMs and AEIs play different roles in the electrochemical devices often leading to different limiting factors and their optimal values.For example, while membranes should limit gas crossover, high oxygen permeability is a desired property of AEI in the cathode of fuel cells.Moreover, while membranes are usually tens of micrometers thick, ionomers usually surround catalyst particles with thin layers of a few nanometers, [3] leading to significant changes in their ionic conductivity and water uptake.Finally, ionomers play an important role as binders and can be determinant in the stability of the catalyst layer, which is not the case for AEMs.
Therefore, while a lot of insights can be drawn from the work on AEMs, anion exchange ionomers should be independently developed and optimized.Even in the preliminary screening of electrocatalysts, an appropriate ionomer should be chosen and its effect should be taken into consideration; while, the vast majority of rotating disc electrode and a considerable portion of fuel cell and electrolyzer studies in alkaline conditions still utilize the commercially available proton conducting Nafion as ionomer.
In this review, we will discuss the structures, properties, and effects of anion exchange ionomers on electrochemical devices, with a focus on AEM fuel cells and with an eye on other electrochemical devices utilizing anion exchange ionomers, such as AEM electrolyzers and CO 2 reduction electrochemical cells.For a more focused discussion of AEM electrolyzers, we direct the reader to a recent review by Holdcroft and co-workers. [4]As this review will show, some of their conclusions in the context of electrolyzers are also applicable to fuel cells, mainly the importance of developing ex situ techniques, to be able to study in isolation: i) the binding ability of the ionomer with the catalyst, ii) the role of AEI morphology and distribution on the device performance, and iii) specific ionomer-catalyst interactions.
In this review, we will start by providing an overview of the state-of-the-art and commercially available anion exchange ionomers, summarizing the chemical structure of the most reported polymer backbones and functional groups.We will then discuss design considerations for AEI, including ionic conductivity, gas solubility, water management, the effect on electrode morphology, and ionomer-specific stability.Although it won't be explicitly discussed due to a lack of research on the topic, the analysis offered here is also relevant for CO 2 reduc-tion electrochemical cells, which operate under mild alkaline conditions.

Common Structures of Anion Exchange Ionomers
Figure 2 summarizes the structure of the most commonly reported polymer backbones and functional groups for anion exchange membranes and ionomers.
The polymer backbone has a strong influence on the mechanical and chemical properties of AEM and AEI.More specifically for ionomers, the choice of backbone controls the hydrophilicity of the polymer and the ability to bind the catalyst and create a uniform catalyst layer.[7][8] Reported backbones for anion exchange ionomers include non-fluorinated, partially fluorinated, and fully fluorinated polymers, such as poly(ethylene) (PE), polystyrene (PS), polysulfones (PS), poly(phenyleneoxide)s (PPO), poly(phenylene)s (PP), poly(fluorene) (PF), polybenzylimidazole (PBI), and poly(ethylene-co-tetrafluoroethylene) (EFTE).Table 1 summarizes the structure of the most recently reported anion exchange ionomers and their performance in alkaline fuel cells.
On the other side, functional groups control the ion exchange capacity (IEC), ionic conductivity (), and most of the degradation of anion exchange ionomers.The most commonly reported functional groups are quaternary ammonium (TMA) and N-containing cations, for example based on pyrrolidonium (PY), piperidinium (PRD), and imidazolium (IM) moieties.[11][12] A direct comparison of the performance of different functional groups is challenging and will be discussed in detail in the following subsection.The upcoming sections will discuss two categories of ionomers that have recently received attention: radiation-grafted ionomers and poly(ionic liquid)s and give an overview of the anion exchange ionomers commercially available.

Performance Comparison of OH − Conductive Functional Groups
As mentioned above, a direct comparison of different OH − conductive groups is challenging because the change in functional groups often affects the molecular weight and backbone of the polymer.Recently, Biancolli et al. used radiation grafting to synthesize an anion exchange ionomer with different head group chemistries, starting from the same poly(ethylene-co-tetrafluoroethylene) backbone.The functional groups they tested were benzyltrimethylammonium (TMA), benzyl-N-methylpyrrolidinium (MPY), and benzyl-Nmethylpiperidinium (MPRD), as shown in Figure 3a. [23]TMA and MPY were found to offer very similar ion exchange capacity (2.05 and 2.09 mmolg −1 ) and ionic conductivity (46 and 43 mS cm −1 ), but MPY suffered from higher water uptake, almost double that of TMA.MPRD, on the side, offered significantly lower ion exchange capacity (1.73 mmol g −1 ) and ionic conductivity (31 mS cm −1 ).They tested different combinations of the so-obtained anion exchange membrane and ionomers.Interestingly, they found that the best ionomer depended on the choice of membrane and that using the same functional groups for the AEM and AEI did not always provide the highest performance.Perhaps more surprisingly, they obtained the highest performance when using the MPRD membrane despite the lower ionic conductivity (Figure 3a). [23]part from the variety of backbones, functional groups, morphology, and molecular weight reported, the differences in testing protocols make it exceptionally challenging to compare the performance of different anion exchange ionomers.Table 1 summarizes some of the highest reported activities of AEMFCs when using anion exchange ionomers.As it can be observed, most  OH - conductivity [mS cm -1 ] Peak power density [mW cm -2 ] -80 °C

Membrane
Fuel cell operating conditions [23]   Poly(ethylene-co-tetrafluoroethylene) -1200 MPRD [23] Temperature= 60 °C Catalyst= Pt/C (40% JM) cathode, PtRu/C (50% Pt, 25%Ru) anode Cat.loading= 0.4 mg cm −2 , anode and cathode I/C ratio=0.25 [24]   Poly(cis-stilbene-co- reports utilized different and in-house built membranes, making it impossible to isolate the effect of AEIs from that of the membrane.Moreover, testing parameters such as temperature, relative humidity, gas flowrate, ionomer to catalyst ratio, catalyst composition, and ink formulation can have a significant impact on the performance of the fuel cell, and standardized values do not exist for most of these parameters.The same is valid for AEM electrolyzers, for which standardized testing parameters and catalyst and ionomer loadings have not been established.Nevertheless, targeted studies focusing on the effect of single parameters on the performance can provide valuable insight into the ideal composition of anion exchange ionomers, as will be discussed in the next section of this review.

Radiation-Grafted Anion Exchange Ionomers
Radiation-grafted anion exchange ionomers have reported some of the highest performances in fuel cells. [17,23,29]Radiationgrafting is a facile technique that allows incorporation of the anion-conductive side chains on pre-fabricated polymeric films.On top of ease of fabrication and scale-up, this approach enables the addition of different functional groups to the same backbones, allowing direct comparison of their performance and stability.This technique can also be adjusted to functionalize the surface of polymers with selected functional groups (─COOH, ─OR, ─OH, ─NH 2 , ─R, and ─SO 3 H), without affecting the properties of the bulk material. [30]Several anion exchange membranes and ionomers have been reported using this technique, based on a poly(ethylene-co-tetrafluoroethylene) backbone, with a selection of functional groups, including trimethyl ammonium, [9,17,19,31,32] pyrrolidonium, [9,23,32] piperidinium, [9,23,32,33] morpholinium, [9] pylrrolidinium, [9] piperazinium, [9] and imidazolium. [18]The most common structures and synthesis pathway are shown in Figure 3c.Wang et al. recently reported a non-fluorinated radiation grafted anion exchange polymer, based on a high density poly ethylene backbone, which was tested as membranes, in conjunction with an EFTE-based anion exchange ionomer, in an AEM fuel cell, achieving a remarkable peak power density of 2550 mW cm −2 . [34]For more information about radiation-grafted anion exchange polymers, we direct the reader to a recent review by Yang and co-workers. [30]

Poly(Ionic Liquid)s
Ionic liquids have received increasing attention in electrocatalysis, thanks to their low vapor pressure, benign environmental impact, excellent thermal and electrochemical stability, and large design freedom.][38][39] Poly(ionic liquid)s (PILs) differ from common ion conductive polymers because they normally feature hydrophobic anions (such as PF 6 and NTF 2 ) and organic cations, mainly based on imidazolium or pyrrolidonium.Compared to the already versatile ionic liquids, PILs have even more design parameters, such as the polymer chain length, the monomer ratio, and the monomer order (block co-polymers vs random copolymers), making them an ideal platform to independently tune the many properties relevant for ionomers, such as gas, water, and ion transport.Despite the high design freedom of poly(ionic liquid)s, their application as anion exchange ionomers has been very limited.To the best of our knowledge, there is only a report of poly(ionic liquid)s been tested as anion exchange ionomers.
In the work of Antonietti and co-workers, they used an ionomer obtained from the polymerization of 1-ethyl-3-vinylimidazolium bromide, followed by an anion exchange to bis(trifluoromethane sulfonyl)imide (TFSI), as a binder for the oxygen evolution catalyst NiCoO 2 .They reported that the poly(ionic liquid) offers higher oxidation stability and lifetime compared to the commercial Nafion. [39]42][43][44] For example, Li et al. added poly(ionic liquid)s to a quaternized poly (2,6-dimethyl-1,4-phenylene oxide), reporting a 92% increase in hydroxide conductivity, attributed to the creation of structure hydroxide transfer channels. [41]A similar phenomenon was observed when using a mixture of the PolyIL poly(1-vinyl-3imidazole-co-styrene), and poly(vinyl alcohol) (PVA) was used to prepare an anion exchange membrane for water electrolysis.
Increasing loading of the poly(ionic liquid) was found to lead to phase separation between the hydrophobic and hydrophilic polymer domains, leading to well-connected hydrophilic channels and an ionic conductivity of 90 mS cm −1 . [43]Lower swelling and higher chemical stability have also been reported when adding polyILs to PVA in anion exchange membranes. [44]Other examples exist of poly(ionic liquid)s being employed as proton exchange ionomers.Among these, Snyder and co-workers exploited the design freedom offered by polyILs to design a triblock-co-polymer (Figure 3d), where each monomer was selected to provide a different property: styrene for hydrophobicity, sulfonated units for proton conduction, and fluorinated anions for oxygen solubility. [28]Although this work is focused on proton exchange ionomers, the same approach can be easily applied to anion exchange ionomer.Another design parameter that can be used to tune the properties of PILs is the monomer order.Ye et al. synthesized an anion exchange copolymer based on methyl methacrylate (MMA) and 1-[(2-methacryloyloxy)ethyl]−3-butylimidazolium hydroxide (MEBImOH).They showed that the block-copolymer exhibited hydroxide conductivity over an order of magnitude higher than the random copolymer analogue (Figure 3b). [27]More surprisingly, the block-copolymer also exhibited higher OH − conductivity and lower water uptake than the MEBImOH homopolymer, despite having lower ion exchange capacity.This suggests that the nanoscale morphology can significantly affect the ion transport mechanism.

Commercially Available Anion Exchange Ionomers
A few anion exchange membranes and ionomers are now commercially available and have been investigated for use in AEMbased fuel cells or CO 2 reduction cells in alkaline media.Hereby, we provide an overview of the commercially available anion exchange ionomers.Table 2 summarizes their structure and  ) [45] 19 [45] Sustanion XC-2, XA-9 1.1 [46] 64 (OH − ) [47] 80 [46] Aemion ) [26] 30-60 [26] a) Price from fuel cell store (https://www.fuelcellstore.com/fumion-dispersion-faa-3-solut-10);accessed in June 2023; b) Price from dioxide materials for the XA-9 alkaline ionomer, 5% in ethanol, 25 mL (https://dioxidematerials. com/product/sustanion-xa-9-ionomer-5-in-ethanol/); accessed in June 2023; c) The exact structure is not disclosed; the structure shown here is from the manuscript by Thomas et al; [ 158]   .Adapted with permission. [23]Copyright 2018, Royal Society of Chemistry.b) Effect of monomer order on the conductivity of a block co-polymer.Adapted with permission. [27]Copyright 2013, American Chemical Society.c) Synthesis and structure of radiation-grafted polymer typically reported as anion exchange ionomers.d) Synthesis and structure of the sulfonated, poly(ionic liquid) proton exchange ionomer reported by Snyder and coworkers. [28]e) Synthesis and structure of the anion exchange poly(ionic liquid) ionomer, reported by Ye et al. [27] characteristics; while, their performance in alkaline electrochemical cells is presented in Table 3.

Fumion, Fumatech
Fumatech FAA-3 is an anion exchange ionomer based on a polyaromatic polymer and uses quaternary ammonium as a functional group.It is reported to be soluble in alcohol and provides an ion exchange capacity of 1.7 to 2.2 meq g −1 Park et al. optimized an alkaline water electrolyzer, utilizing Fumatech membrane and ionomers and reaching a current density of 1.5mAcm −2 at 1.9 V. [48] They showed that the morphology of the catalyst layer changed with ionomer content, with porosity increasing for higher loadings.They found the optimal loading to be 20 wt%, where a lower loading caused high charge-transfer resistance; while, a high loading was associated with increased mass transfer and ohmic losses. [48]Fumion FAA-3, in combination with the membrane Fumasep FAA-3, from the same company, had also been tested for fuel cells, with a reported peak power density of 90 mW cm −2 (details in Table 3). [49]inally, Arunkumar et al. reported the use of graphene oxide nanofillers to improve the performance and stability of Fumion anions. [45]They reported that the addition of 1 wt% graphene oxide dramatically increased the ionic conductivity at 90 °C from 29 to 107 mS cm −1 , with improvements also observed for water uptake.

Sustanion, Dioxide Materials
Dioxide materials offer two anion exchange ionomers: Sustanion XC-2 and Sustanion XA-9, optimized for CO 2 reduction and to inhibit water electrolysis.They are both based on a co-polymer of vynilbenzyl methylimidazolium and stryrene.Sustanion XC-2 is slightly less hydrophilic and both are sold in the chlorinated form.A recent publication by Liu et al. gives information about the synthesis (Figure 4) and characterization of this anion exchange ionomer and its associated membrane (Sustanion X24), obtained by cross-linking with divynilbenzene. [47]OH − conductivity was only calculated in the membrane form and was found to be 64 mScm −1 at room temperature. [47]The AEI and AEM were further tested in a CO 2 reduction cell.It was found that the use of Nafion at the anode lead to excessive hydrogen evolution; while, the absence of a binder caused poor catalyst utilization.Finally, the use of the Sustanion ionomer improved the performance, reaching a current density of 300 mA cm −2 at 3 V (details in Table 3). [47]he Sustanion membrane and ionomer have also been tested in water electrolyzers (details in Table 3); [50] while, to the best of our knowledge, the performance of these polymers in fuel cells has not yet been reported.A modified version of the membrane, called Sustanion 37-50 has been proposed, based on the use of tetramethyl imidazole instead of 1-methyl imidazole, to improve its alkaline stability.

Aemion+, Ionomer
Aemion and Aemion+ are two generations of the polyimidazolium-based anion exchange membranes developed by Ionomr.A version of the Aemion+ membrane, reinforced with a fluorine-free polyolefin substrate, has been tested in an AEM water electrolyzer, showing negligible degradation over a period of 8900 h and overall cell degradation of 0.013 mVh −1 at a current density of 600mAcm −2 (details in Table 3).The measurements demonstrated the alkaline stability of the structure but were performed using Nafion as binder on both anode and cathode due to lower dimensional swelling and catalyst delamination. [1]In another study, Aemion+ was used both as anion exchange membrane and ionomer in an alkaline fuel cell.Reinforced Aemion+ was also used as membrane, showing improved stability at 80 °C, to the point that degradation in the catalyst layer surpassed that of the membrane, highlighting the importance of understanding the ionomer degradation in the catalyst layer. [51]gure 4. Schematic of the synthetic route for the production of Sustanion ionomers.Reproduced with permission. [50]Copyright 2018, Frontiers.

PiperION, Versogen
PiperION is the anion exchange ionomer developed by Versogen.It is available in three forms differing for the ion exchange capacity: A5 (2.3 IEC), B5 (2.8 IEC), and C5 (3.5 IEC).They are all based on a poly-aryl piperidinium polymer, which was reported to provide exceptional hydroxide conductivity (up to 200 mS cm −1 ); while, limiting the water uptake below 60%. [26]Thanks to the ether-bond-free aryl backbone, this ionomer also exhibited excellent alkaline stability, maintaining its conductivity and flexibility after 2000 h in 1 m KOH at 100 °C, with only a 3% loss in IEC. [26]The stability at high temperature also expanded the potential operating temperature of fuel cells.The poly-aryl piperidinium polymer was used as both anion exchange membrane and ionomer in an AEM fuel cell at 95 °C, obtaining a peak power density of 920mWcm −2 (details in Table 3). [26]The peak power density in H 2 /O 2 AEM fuel was further improved to a value of 1890 mWcm −2 , by independently optimizing relative humidity and back pressure at the cathode and anode. [52]The use of poly-aryl piperidinium membranes and ionomers was also reported for direct ammonia fuel cells, reaching a peak power of 135 mW cm −2 (details in Table 3). [53]Finally, Endrodi et al. tested the PiperION membrane and ionomer in a CO 2 reduction electrochemical cell, reaching 600 mA cm −2 current density at 3.2 V, with a CO selectivity above 90% (details in Table 3). [54]

Performance Comparison of Commercially Available Ionomers
Endrosi et al. compared the performance of a CO 2 reduction electrochemical cell when using the commercial anion exchange membranes and ionomers Sustanion, PiperION, and Fumasep.While Sustanion offered similar performance to PiperION, the current density obtained with Fumasep was ten times lower.The difference was assigned to the membrane resistance and determined by impedance spectroscopy to be much higher for Fumasep (4 Ω cm 2 ) than for the case of Sustanion (0.85 Ω cm 2 ) or PiperION (0.36 Ω cm 2 ). [54]A similar comparison was performed by Linquist et al. in the context of AEM electrolyzers.They compared the use of Sustanion, Aemion, and PiperION membranes, obtaining very similar results with Sustainion and Aemion; while, the performance was visibly improved by the use of PiperION. [55]omparison of AEI performance, both commercially available and not, is still limited to single independent work but is already to highlight how impactful the anion exchange ionomer can be on the overall activity of electrochemical devices and how the optimal structure is application dependent.

A Note on Ion Exchange for Commercial AEIs
As mentioned above, the commercial anion exchange ionomers are generally delivered in the Cl − and Br − form and the hydroxide conductive form can be obtained by ion exchange.Most commonly, this is achieved by soaking the polymer in concentrated hydroxide solution (1 m KOH) for 24 h at temperatures ranging from 25 °C to 80 °C, followed by washing with water.The ionomer is sometimes exchanged after the deposition of the catalyst layer.For the case of PiperION, a similar process is suggested, which involves soaking the catalyst layer in 0.5 m NaOH or KOH for an hour, followed by a second 1 h wash in a fresh solution.
Despite the simplicity of this ion-exchange step, significant performance differences have been reported with small variations of this procedure.For example, Gatto et al. tested the performance of an AEMFC with Fumatech FAA-3 ionomer and membrane, by varying the ion-exchange time in alkaline solution (1 m KOH) from 1 to 24 h. [56]They found that in the first case, the fuel cell performed better, with a cell potential of 0.656 V@ 200 mA cm −2 , compared to 0.460 V for the second case.This could indicate that the membrane is subject to alkaline degradation, and prolonged exchange times weaken the membrane structure.Sebastian et al. performed ion exchange on a Fumion FAA-3 ionomer after depositing the catalyst layer by first treating it in a KCl solution (0.5 m in water/isopropyl alcohol 75/25 vol%), and subsequently, in a KOH solution (0.1 m in water/isopropyl alcohol 75/25 vol%). [49]They investigated the effect of KCl exchange time and found that most of the bromide is exchanged after 30 min.In terms of performance, in the absence of ion exchange pre-treatment, the activity is much lower; while, the best operation is achieved with two steps KCl exchange. [49]

OH − Conductivity
Depending on the application and operating parameters, ionic conductivity can be one of the most important properties of anion exchange ionomers (AEI).It is even more critical than for proton exchange ionomers (PEI) as the diffusion coefficient of OH − is four times lower than that of H + , causing anion exchange ionomers and membranes to typically exhibit lower ionic conductivities compared to their proton exchange analogues. [1]Despite this, limited data is available on OH − conductivity in ionomers as research has primarily focused on anion exchange membranes.Although AEM can also be used as anion exchange ionomers, their structure and thickness is different, with ionomers generally surrounding catalyst particles with thin layers of a few nanometers, [3] compared to tens of micrometers for the case of membranes.It has been shown that both proton transport and water uptake differ dramatically from films to bulk.For example, Siroma et al. showed that Nafion conductivity decreases dramatically with thickness, to the point that the conductivity of a 100 nm film is about an order of magnitude lower than the bulk (Figure 5a). [59]Thinner films also showed higher apparent activation energy for conduction and the difference was ascribed to the lower water sorption in thin films and to the fact that the 2D layout of thin films disrupts the microphase separated morphology present in bulk membranes. [59]Very limited reports are available on the thickness effect on OH − conductivity for alkaline membranes, partially due to the challenges associated with hydroxide conductivity measurements, which require shielding from CO 2 .Nevertheless, Wang et al. showed that, similar to what was observed for Nafion, ionic conductivity and water uptake of a PFB + anion exchange ionomer decrease with decreasing thickness. [60]igure 5. a) Activation energy and conductivity of Nafion as a function of thickness.Adapted with permission. [59]Copyright 2009, Elsevier.b) Power curves at optimized dew points comparing the deployment of the ionomers shown in the figure, differing by the ratio of hydrophilic and OH − conductive repeating units and hydrophobic repeating units.The red line shows the results obtained with the hydrophobic repeating units on both cathode and anode; the blue line shows the results for the hydrophilic anion exchange ionomer; and the grey line shows the results for the asymmetric deployment of the hydrophilic anion exchange ionomer in the anode and hydrophobic in the cathode.Results obtained with type 1 GDEs, no backpressure; the cell with temperature was 80 °C, H 2 /O 2 flow rates were 1 L min −1 , and AEM was GT64-15.Adapted with permission from Hassan et al. [71] 2020, Wiley.c) Strategies for increasing hydroxide conductivity while maintaining low water uptake.Poly(aril piperidinium) is that backbone that offers the highest conductivity to swelling ratio, as reported by Wang et al. [26] Hu et al. showed that fluorination of the backbone can further improve hydrophobicity, [72] and finally, cross-linking has been shown as an effective technique to control water uptake by Ge et al. [73] Correlation between hydroxide conductivity and water uptake.d) Conductivity as a function of water uptake for a selection of reported anion exchange polymers.The graph shows a general correlation between conductivity and water uptake for side chain (yellow), random (grey), and block (purple) copolymers.Strategies to break this correlation include the use of poly(aryl)piperidinium backbones (in red), fluorinated backbone (blue), and cross-linking (black).Data for commercial anion exchange ionomers are also shown (green).Adapted with permission from Wang et al. [26] 2019, Nature.Fluorinated polymer data are from Hu et al. [72] and cross-linked polymer data are from Ge et al. [73] e)Relationship between state water uptake and its kinetics.Adapted with permission. [74]2018, American Chemical Society.
Apart from the drastic effect that thickness can have on the ionic conductivity of AEM and AEI, the need of optimizing membranes and ionomers separately also arises from the relative importance of these parameters.For example, in the case of fuel cells, the ionic conductivity of the membrane is often of secondary importance compared to its ability to transport water from the anode to the cathode.As a result, performance of AEM may not translate to AEI, for which ionic conductivity might be more important.
Very different is the case of electrolyzers.When dilute KOH is used as an electrolyte, its inherent conductivity may overcome the need of anion exchange ionomers altogether.For example, Wang et al. demonstrated high current density of 2 A cm −2 at 2.1 V, using an ionomer-free plasma-sprayed electrode. [61]On the other side, when DI water is used as a supporting electrolyte, the anion transport provided by the AEI becomes essential. [62]nother important factor to consider is that high ionic conductivity of the membrane alone is not sufficient if the electrodemembrane interface is not optimized.In fact, mass transport phenomena in the electrodes and the lack of a continuous network for the transport of hydroxide ions from the membrane to the catalyst surface have been identified as limiting factors for the performance of alkaline fuel cells. [31]Zeng et al. reported an increase in ECSA at the cathode of an alkaline MEA, when utilizing a cross-linked quaternatized polysulfone, manifesting improved OH − transport network and catalyst utilization. [15]Tee al al. also reported the use of bisphenol-A-diglycidyl ether as an adhesive to chemically bond ionomer, catalyst, and porous transport layer, obtaining a continuous network, which improved both the performance and durability of the catalyst layer. [63]n conclusion, the design of anion exchange ionomers should aim at maximizing OH − conductivity in thin films and consider compatibility with the membrane to create an uninterrupted pathway for the transport of hydroxide ion across the fuel cell.

Water Management
Water management is critical in both AEM fuel cells and electrolyzers.In AEMFCs, water serves as reactant at the cathode and product at the anode.As a result, flooding can limit the performance at the anode by decreasing gas access to the catalyst; while, drying at the cathode can lead to low accessibility to water and hydroxide ions.For the case of electrolyzers, water is consumed at the cathode, but stability is higher if water is supplied at the anode side. [64]Therefore, the water required to support hydrogen evolution at the cathode is supplied by water transport through the membrane, causing mass transfer limitations. [2]IN addition, when OH − ions are transported through anion exchange membranes, up to eight water molecules are carried along by electroosmotic drag, [65] compared to the one or two carried by protons in PEMs, making water management in AEMs more critical, both for the case of electrolyzers and fuel cells. [66]The anion exchange ionomer can have a significant impact on water management and, in general, high water permeability is expected to be beneficial, but the optimum water uptake varies based on the cell and electrode in question. [67]ow water uptake can be particularly critical for oxygen evolution catalysts at the anode of AEM electrolyzers.In this case, drying is not an issue as liquid water is fed at the electrode and produced at the catalyst surface.On the contrary, hydrophobicity is required to allow efficient desorption of gaseous bubble.In fact, it has been reported that hydrophobicity of the ionomer has a much higher effect on oxygen evolution catalysts than ion exchange capacity, which is not as critical when diluted KOH is fed instead of water as the electrolyte can contribute to the conductivity of the catalyst layer. [68,69]Different is the case for hydrogen evolution taking place at the cathode of AEM electrolyzers, for which a higher optimal water activity has been reported.This is because HER electrodes are not externally sup-plied with water and consume water to produce hydrogen.In this case, hydroxide conductivity is much more important than for OER electrodes, and a delicate balance between providing sufficient water supply and avoiding excessive swelling means that an intermediate optimal water uptake of 200% has been reported. [68,70]EM fuel cells have similar requirements as it has been suggested that the oxygen reduction catalysts might favor hydrophobic environment to maximize the reactant concentration; while, the hydrogen oxidation is expected to require a slightly more hydrophilic environment. [71]In a recent study on anion exchange ionomers for AEM fuel cells, Hassan et al. investigated the use block co-polymer, with different percentage of OH − conductive monomers as cathode and anode anion exchange ionomers (Figure 5b).They obtained the highest performance (achieving a peak power density of 3.2Wcm −2 ) when using a combination of the most conductive and hydrophilic ionomer at the anode and hydrophobic at the cathode (Figure 5b). [71]They speculate that the higher water uptake (163%) and IEC (3.74 meq g −1 ) at the anode might help discharge water more quickly to the membrane.
Finally, water uptake can also have an impact on stability as reduction in ionomer swelling can reduce adhesion to the catalyst layer, as summarized in a review by Holdcroft and co-workers. [4]hese observations highlight the necessity of being able to control the water uptake of the anion exchange ionomers, for which several strategies have been proposed.For example, Zeng et al. synthesized a poly(vinylbenzyltrimethylammonium chloride) and showed that the degree of swelling can be tuned by controlling the heat treatment process. [67]However, most approaches to reduce water uptake negatively affect ionic conductivity, as recently shown by Wang et al. [26] (Figure 5d).To overcome this trade-off, Hu et al. showed that fluorination of the polymer backbone (by introducing perfluoroalkylene units, Figure 5c) can also improve the hydrophobicity of anion exchange ionomers.When applied to the anode of fuel cells, this fluorinated ionomer can prevent water flooding and lead to an improve in cell voltage of 140 mV, compared to its hydrophilic counterpart. [72]As the ionomers swell due to the pressure caused by water uptake in the hydrophilic domains, another approach to reduce swelling relies on increasing the elastic modulus of the polymer by nonionic reinforcement [1] or covalent cross-linking. [69,70]In particular, cross-linking has been proven effective in reducing water uptake; while, maintaining high ion exchange capacity (IEC), [69,70] providing a ratio of hydroxide conductivity to water swelling, up to 33. [73] Apart from the steady-state swelling, the ability to transport water quickly is also important and possibly more critical at the device level.Zheng et al. characterized water transport, in a selection of anion-exchange membrane with different backbones and functional groups, observing that the kinetics of water uptake does not directly correlate to the equilibrium value (Figure 5e). [74]Among the AEM membranes tested, ETFE-based ones were found to offer the lowest water uptake, with average IEC and the fastest water transport kinetics, isolating them as an ideal candidate to achieve consistent water management in fuel cells and electrolyzers. [74]onomer loading is also an important parameter in water management as excessive ionomer content can accelerate electrode flooding. [75]In separate works, Yang et al., [76] Kaspar et al., [77] and Park et al. [48] proposed a similar optimal ionomer loading of around 20 wt% of the Tokuyama AS-4, Tokuyama AS-4, and Fumatech FAA3 commercial ionomers, respectively.However, other reports have proposed significantly different optimal values.For example, Cossar et al. found that the optimum Aemion loading for the anode of AEM electrolyzers is 7 wt%. [58]On the contrary, Sebastian et al. reported that the optimum Fumion FAA-3 loading for alkaline fuel cells was 42-50%. [49]t has also been reported that, while membrane swelling has a small effect on durability, a reduction in water uptake of the anion exchange ionomer can increase the lifetime of AEM electrolyzers fourfold in pure water. [78]Similarly, using vapor sorption measurements and adhesion force mapping, Yin et al. showed that cracking of fuel cell catalyst layers is related to ionomer migration and is more pronounced at high humidity. [79]Thus, it is clear that the ability of the anion exchange ionomer to provide consistent water retention does not only affect the performance of electrochemical devices but also its stability, and that the hydrophobicity of the ionomer should be carefully optimized for each electrode.

Oxygen Transport
Unlike anion exchange membranes which should limit gas crossover, anion exchange ionomers in AEM fuel cells should present high H 2 and O 2 permeability to encourage reactant transport to the active sites.In particular, O 2 transport to the cathodic catalyst can be one of the limiting factors when operating fuel cells at high current density.In this case, the same requirement is not valid for electrolyzers, where rapid expulsion of the evolved gas is desired.Therefore, low oxygen and hydrogen solubility are accepted to be beneficial; although, to be best of our knowledge, this aspect has not been studied in the context of electrolyzers.
[82][83][84][85][86] These modifications are normally aimed at disrupting the crystallinity of the polymer, increasing their fractional free volume.As these modifications normally involve the backbone of the polymer alone, the same strategies applied to PEIs should be also applicable to AEIs.While only a handful of examples exist of highly oxygen permeable AEIs, the data available seem to confirm that high void volume increases gas permeability in AEIs.For example, Choi et al. modified a poly(arylene ether sulfone) polymer (BP-PES), with the addition of spirobiindane units (SBI-PES), as shown in Figure 6d. [87]he addition was found to create microporous channels, suitable for gas diffusion and water uptake, obtaining a 5 and 23times increase in oxygen solubility and diffusivity, respectively (Figure 6d). [87]Gas permeability of AEIs was also reported to increase with the flexibility of the polymer chain. [88]Yang et al. also developed a anion exchange ionomer with exceptional hydroxide conductivity of 164 mS cm −1 , using an intrinsically microporous polymer containing the V-shape rigid Troger's base units.The rigidity of the polymer chain and consequent high free vol-ume were reported as the cause of the high conductivity and gas permeability. [76]or example, poly(phenylene oxide) anion exchange ionomers with piperidium functional groups (PPO-DMP) have been reported to provide higher gas permeability and cell performance, compared to their azoni-spiro undecane counterpart (PPO-ASU) (Figure 6b). [89]Yu et al. recently proposed an innovative approach to improve oxygen transport in anion exchange ionomers by controlling their conformation. [24]They synthesized two ionomers based on poly(stilbene-co-diphenyl piperidinium), differing on the isomeric configuration to the stilbene (Figure 6c).They showed that the cis-configuration forms 0.5 nm free volume cavities, which increases oxygen permeability by 93% compared to their trans counterpart and leads to a 54% improvement in peak power density. [24]][39] Snyder and co-workers proposed a triblock-co-polymer, where each section could be optimized to control different properties of the ionomer, such as conductivity and hydrophobicity.In particular, an ionic liquid moiety (vinylbenzylmethylimidazolium bis(trifluoromethylsulfonyl)imide) was used to provide oxygen transport (Figure 6e). [28]Although this work was focused on proton exchange ionomers, the same approach can be used for anion exchange ionomers thanks to the broad design freedom offered by poly(ionic liquid)s.

Ionomer Adsorption
Another factor to consider is that some anion exchange ionomers suffer from ionomer adsorption on the active site, which can limit the performance of the catalysts, particularly for the Pt-catalyzed HOR.Kohl and co-workers were the first to report that adsorption of quaternary ammonium cations can decrease the accessible surface area of the catalyst, inhibiting alcohol oxidation in direct methanol fuel cells. [91]A recent review on the topic highlights the detrimental effect of cation-hydroxide-water co-adsorption and phenyl adsorption on HOR catalysts and on the overall performance of alkaline fuel cells. [92]Cation-hydroxide-water coadsorption has a more significant impact on HOR as it can limit hydrogen diffusion through the adsorbed layer. [91,93][96][97] Mitigation strategies have been identified, such as the use of less phenyl group-adsorbing catalysts.For example, it has been shown that Pt-bimetallic catalysts, such as Pt-Ru, PtNi, and PtMo, limit phenyl adsorption compared to pure platinum. [22]nother possibility is the use of anion exchange ionomers with fewer phenyl moieties, such as polyolefins.Side chain phenyl adsorption can also be limited by cationic substitution [22] or by the use of bulky substituents, such as fluorene. [25]hese phenomena highlight the necessity of optimizing the catalyst layer as a whole, taking into account the interactions between the ionomer and the catalyst, rather than optimizing the individual components separately.c) Improving the oxygen permeability of the anion exchange ionomer SP-DP by controlling its isomeric confuguration. [24]d) Improving the oxygen permeability of the anion exchange ionomer BP-PES by introducing the bulkier monomer SPI. [87]e) Improving the oxygen permeability of the proton exchange ionomer: sulfonated poly(ionic liquid) block co-polymer by the introduction of a monomer, with an oxygenophilic anion. [90]

Stability
The stability of anion exchange membranes and ionomers can be a limiting factor for the performance of electrolyzers and fuel cells and is therefore an essential parameter to consider when designing anion exchange ionomers.Due to the harsh alkaline environment, AEMs and AEIs are susceptible to nucleophilic attack by the hydroxide ion, which takes the form of nucleophilic substitution in cationic tetra-ammonium functional groups and ring opening in imidazolium-based polymers.However, alkaline degradation is not necessarily a good predictor of the stability of the electrochemical devices.In the field of batteries, it is well known that degradation at the electrode interface is unavoidable and leads to the formation of a solidelectrolyte interface (SEI). [98,99]However, rather than causing instability in the battery, this interface can act to stabilize the electrode and prevent further degradation.The same phenomenon has been reported by our group for the electrochemical reduction of nitrogen [100] and could also be happening in the catalyst layer of fuel cells and electrolyzers, where chemical degradation of the anion exchange ionomer might lead to performance stabilization rather than loss.On the contrary, a decrease in performance could result from ionomer degradation caused by electrode-specific factors, such as electrochemical oxidation or physical ageing.Therefore, while the classical stability test (based on immersing the polymer in concentrated KOH solutions) is a good predictor for the durability of membranes, the same cannot be simply translated to ionomers.Rather than focusing on the chemical integrity of the ionomer, research efforts should be directed at investigating the overall stability of the catalyst layers, deconvoluting the effects of catalyst and ionomer degradation.

Stability Against Oxidation
][103] They compared the resistance of AEM electrolyzers assembled with Aemion, Sustainion, and PiperION membranes and ionomers.All showed a similar initial degradation rate of around 10mVh −1 .However, while PiperION stabilized after 60 h, maintaining a voltage below 2.4 V after 180 h and reaching a long term loss of 0.67 mV h −1 , Sustainion did not show any sign of stabilization and reached a cell voltage of 2.9 V in only 40 h of operations.Aemion offered intermediate stability and reached the cut/off limit of 2.9 V after 90 h.Studying the XPS spectra of the gas diffusion electrode before and after a stability test, they reported the appearance of additional C peaks associated with oxidized species and a loss of F and N signals, demonstrating significant oxidation of the anion exchange ionomer, which they identify as the leading cause of the performance loss. [55]Boettcher and co-workers also studied the effect of supporting electrolytes on the stability of commercial anion exchange ionomers under oxidizing potential.PiperION showed good stability in KOH and borate buffer but reported material loss in the carbonate and bicarbonate buffer.Sustainion showed similar stability in KOH and carbonate/bicarbonate; while, little degradation was observed in KOH.Finally, Aemion offered the lowest stability, degrading and dissolving in all the electrolyte tested. [101]lthough ionomer oxidation is now recognized as an important driver of the activity loss of electrolyzer, this oxidation mechanism and its effect on the performance of the catalyst layers are still poorly understood. [104,105]AEI oxidation could happen chemically through reaction with oxygenated species forming at the catalyst surface or electrochemically at the catalyst-ionomer surface.This mechanism could additionally be independent of the catalyst, it could correlate to the ionomer-catalyst interaction strength, [26] or it could be determined by the electrical conductivity of the catalyst. [102]arsh oxidative conditions are also applied at the cathode of fuel cells, for which similar oxidation-driven degradations of the ionomer have been reported.For example, Maurya et al. reported that the most prominent permanent performance losses of AEM fuel cells were caused by phenol formation as a result of the electrochemical oxidation of phenyl groups in the cathodic anion exchange ionomer.The attribution of the phenyl oxidation origin of degradation was confirmed by RDE oxidation of benzyltrimethylammonium hydroxide and by the observations that losses are not recoverable after changing the electrolyte (on the opposite, voltage decay rate goes from 2.3 to 6.6 mV h −1 after 500 h of operation) and are more severe at low current density (i.e., higher potential). [96]Phenyl groups are exceptionally critical as they are most prone to electrochemical oxidation and Maurya et al. have shown that the use of a phenyl-free poly(fluorene) can significantly improve stability, leading to >500 h cathode longevity (Figure 7e).Phenyl oxidation is also detrimental because it releases acidic phenol, which neutralizes the basic electrolyte. [106]s we mentioned in the previous section, phenyl groups are also undesired due to their strong adsorption energy on platinum, which gives an additional research drive to reducing the presence of phenyl groups in the polymeric backbones of ionomer.

Physical Ageing
Physical ageing of the AEI can also be an important factor in the degradation of both AEM electrolyzers and fuel cells.During operation, the anion exchange ionomer re-organizes itself toward a pseudo-equilibrium state, driven not only by thermodynamics but also by degradation and re-arranging of the catalyst itself.This is particularly critical for the case of fuel cells cathodes, where platinum is known to dissolve and agglomerate.Limited work has been done to study this phenomenon as phase structure changes of the ionomer are difficult to detect and decouple from migration of platinum nanoparticles.Nevertheless, ionomer distribution changes [79,107] and crack propagation [108] have been observed as a result of this process (Figure 7f).The ionomer relaxation and re-arrangement have been shown to lead to an initial performance improvement, attributed to an increase in anion transport, followed by a degradation, partially ascribed to physical ageing of the ionomer. [109]or the case of electrolyzers, adhesion of the catalyst and ionomer to the current collector are especially critical as the evolution of hydrogen and oxygen bubbles can cause delamination of the catalyst layer and consequent loss of performance. [110]hysical stability is also a function of ionomer loading, with  [5] Copyright 2016, American Chemical Society.b) Alkaline degradation pathway of cation groups.Adaptedwith permission. [119]Copyright 2018, Wiley-VCH.c) Effect of humidity on degradation of TMBA-remaining reports than <10% ionomer causing rapid catalyst detachment. [70]To overcome this issue, Park et al. developed "self-adhesive" ionomers by adding a bis-phenol-A-diglycidyl ether adhesive to the catalyst ink.They observed that the performance of the electrolyzer improved during a 40 h break-in period, which was attributed to ionomer rearrangement and reported a subsequent minimal degradation rate of 2.5 μV h −1 in the first 200 h. [68]They additionally optimized the adhesive loading and observed that, while for the HER cathode only modest adhesive loading of 0.1mgcm −2 can be used before a significant increase in the ohmic drop, [63] OER anodes benefit from higher loadings of up to 0.6 mg cm −2 which lead a degradation rate of only 0.003 mV h −1 [68] and even cause higher initial activity and shorter break-in periods.
Osmieri et al. reported that, while the use of the commercial anion exchange ionomer PiperION in the catalyst layer of AEM electrolyzers leads to catalyst detachment, the combined used to PiperION and the PEI Nafion gives structural stability to the catalyst, thanks to the superior binding abilities of Nafion. [111]inally, Mayerhöfer et al. showed that heat treatment at 220 °C can also help to increase the elastic modulus of the AEI Aemion, resulting in greater mechanical stability of the catalyst layer. [112]uantification of ionomer adhesion to the catalyst layer is inherently challenging.Tee et al. proposed a qualitative technique to observe ionomer adhesion ex situ, based on the use of adhesive tape, which was pressed on the electrode and then removed by pulling, observing how much of the catalyst layer is removed. [63]As suggested by Mardle et al., standardization of this procedure could provide a quick ex situ way to screen ionomers for adhesiveness. [4]

Alkaline Stability
The alkaline stability of the anion exchange ionomers is greatly influenced by the local environment encountered at the electrodes.In terms of fuel cells, at the cathode, low hydration is encountered due to water consumption by the oxygen reduction reaction, and it has been reported that ionomer degradation is drastically accelerated in low relative humidity conditions (Figure 7c), [42,113,114] that the stability under these conditions cannot be inferred from that at high relative humidity, [42] and that the mechanism might also change. [115]Hydroxide transport would additionally be influenced by the water content, suggesting that cathode ionomers should be specifically designed to operate at low humidity levels.
On the other side, opposite conditions of full hydration and low potential are met at the anode of fuel cells.Although these conditions are more similar to those probed by ex situ measurements, anode-specific mechanisms could still take place, such as those triggered by cation-hydroxide-water co-adsorption on the surface of the HOR catalyst (Figure 7d). [116]or the case of AEM electrolyzers, much higher current densities can be achieved when electrolyte (generally 1 m KOH) is flowed instead of pure water.However, the use of a supporting electrolyte has been reported to increase the performance loss of the electrolyzer more than three times.On the contrary, the degradation rate was reported to decrease from 96 to 40 mV h −1 when KOH was only flowed at the anode. [55]inally, alkaline degradation of AEI can also differ from that of the corresponding AEM, by the simple fact that the anion exchange ionomer is deposited on a carbon support. [117]an et al. reported that, while an mPBI film decomposes into small pieces by hydrolysis and ring cleavage after 24 h in 0.01 m KOH, the same polymer deposited on carbon nanotubes does not show any sign of degradation under the same conditions. [117]his discussion underlined the importance of focusing on the performance loss of the catalyst layer under operating conditions, rather than on the alkaline stability of the anion exchange ionomer.Nevertheless, the extensive work that has been reported on the study on the ex situ alkaline degradation of anion exchange membranes can offer a guide to the design of stable AEIs and to the mitigation of alkaline degradation, in case it is observed to cause performance losses.[120] Here, we will only offer a brief overview of the stability of the polymeric backbones and functional groups most often employed as anion exchange ionomers.
Both the cationic group and polymer backbone have an important role in the degradation pathway (Figure 7a,b).The alkaline stability of polymer backbones is less studied, but a systematic study shows that polymers without aryl ether bonds offer good stability (Figure 7i). [5]On the other side, extensive work has been dedicated to cationic moieties, comparing the stability of different functional groups and the effect of substitutions.For example, Varcoe and co-workers tested a selection of ETFE-polyvenylbenzyl membranes, with different functional groups, showing that both MPY (benzyl-N-methylpiperidinium) and MPRD (benzyl-N-methylpyrrolidinium) offer better alkaline stability, compared to the most commonly employed TMA ion under fully hydrated conditions (Figure 7h). [9]However, the same group showed that, under low hydration conditions, low-density polyethylene-based radiation-grafted anion-exchange membranes with TMA functional groups are more stable than their MPY counterpart, showing that the backbone and hydration conditions can also influence stability of the cationic group. [121]BA fractions as a function of time with different number of water molecules per OH − ; computational results from Dekel et al.Reproduced with permission. [42]Copyright 2017, American Chemical Society.d) CV of Pt in 0.1 m HclO4 and TMAOH electrolytes, showing the effect of cation coadsorption.The CVs were obtained at 25 °C; rotating speed, 900 rpm; and scan rate, 50 mV s −1 .Adapted with permission. [116]Copyright 2016, American Chemical Society.e) Electrochemical oxidation pathway of polymers containing phenyl groups.Adapted with permission. [96]Copyright 2019, Elsevier.f) SEM images showing the effect of anion exchange ionomer physical ageing on the cathode and anode of a fuel cell.Adapted with permission. [108]opyright 2014, Elsevier.g) Structure of incrementally stable cationic functional groups by substitution with phenyl and alkyl substituents. [122]h) IEC of ionomers, before and after alkaline treatment, for a selection of functional groups radiation-grafted of EFTE. [9]i) Stability of polymer backbones to alkaline environment, average molecular weight before and after alkaline treatment.Adapted with permission. [5]Copyright 2016, American Chemical Society.
Yan and co-workers also completed a systematic study of the alkaline stability of several cations by direct polymerization of the relevant monomers, observing lower durability of ethylpyridinium, compared to TMA and a significant dependence of stability on the degree of substitution, independently of the nature of the cationic group. [119]Si et al. investigated the effect of butyl groups at various substitution position of imidazolium cations (N1, C2 and N3), showing that C2 substitutions have the strongest impact on stability; [122] while, other studies show the importance of bulky substituents at the C2 position. [123,124]he effect of C and N substituent on the aromaticity of imidazoles has also been shown computationally. [125]FT simulations have additionally predicted that substitution at the C4 and C5 positions of imidazole should improve its alkaline stability, [46,126,127] and this has been further proven experimentally. [122,128,129]Hugar et al. synthesized an imidazolium cation, stable up to 5 m KOH, at 80 °C, by a combination of C2, C4, C5, and N substitutions (Figure 7g). [129]They reported that 2,6-dimethylphenyl substituents are the most effective at C2 position; while, alkyl substituents perform best on nitrogen. [129]

Electrode Morphology
132][133][134][135][136][137][138] Molecular dynamics simulations and TEM observations have shown that ionomers tend to deposit in the form of thin films on the surface of platinum-based catalysts. [139]The morphology of Pt/C-based electrodes has also been reported to depend on the ionomer content, when using the commercial anion exchange ionomer Tokuyama AS-4. [140]Park et al. also compared the morphology of a catalyst layer obtained with the commercial Fumion ionomer and two in-house poly(phenylene oxide) polymers, differing only for the functional group (Figure 8c-e). [89]Only PPO-DMP was found to produce a uniform layer, while the use of PPO-ASU and Fumion resulted in the formation of cracks.The difference was attributed to the variation in ionomer flexibility. [89]he solvent can also play a role in controlling the electrode morphology, and the optimal ink composition is likely to be dependent on the ionomer used.[143][144][145] Jervis et al. also reported that OH − conductive binders consistently yield poorer electrodes, with large catalyst agglomerates, compared to Nafion(Figure 8a,b), leading to a poorer accessibility to active sites and lower limiting HOR current. [146]he ink composition can be optimized ex situ using techniques such as dynamic light scattering (DLS), rheology studies, and small X-ray scattering (SAXS).DLS and SAXS are used to study the size of catalyst agglomerates in the ink.SAXS is less commonly used because the upper size limit is ≈50 nm; while, larger particles require USAXS at a synchrotron radiation source. [147]DLS, on the other side, is a facile and cheap technique to measure particle size but it requires dilute solution.Due to the high catalyst concentration used in the ink, this technique is most often performed on the solvent and ionomer alone.However, Shukla et al. have shown that dilution of up to 100 times only affects the measured particle size in catalyst inks by less than 20%, showing that this technique applied to diluted inks can provide reliable qualitative information. [138][132][133][134] Using this technique, in combination with ultra-small angle X-ray scattering, Mauger and co-workers showed that Nafion stabilizes carbon dispersions via repulsive electrostatic interactions. [133]More surprisingly, they noticed that the ionomer has an opposite effect on platinum deposited on high surface area carbons, causing the particles aggregates to flocculate. [133]nce again, this shows the importance of comprehensively optimizing the catalyst layer as the optimal solvent and ionomer content depend on both the nature of the catalyst and the ionomer.However, research has so far focused on optimizing ink composition for commercial Pt/C catalysts and Nafion ionomers; [148][149][150][151][152] while, fewer studies exist on PGMfree catalysts [130,136,153] and anion exchange ionomers. [140]In one of the few studies reported on transition metal catalysts, Artyushkova et al. showed that the ionomer interaction with the catalyst depends on the surface chemistry of the catalyst itself. [153]n particular, catalysts rich in active sites, oxygenated functional groups, and cationic nitrogen groups can cause strong interaction with the ionomer side chains, potentially limiting hydroxide transport.In another study on Fe─N─C catalysts, Mauger and co-workers showed that the ideal solvent composition depends on the catalyst concentration, with concentrated inks favoring water-poor dispersion medium due to the hydrophobic nature of the catalyst. [130]They also observed that the particle morphology transitioned among agglomeration, stable dispersion, and flocculation with increasing Nafion concentration (Figure 8g). [130]ptimizing the ink formulation and ionomer content is essential because they affect the catalyst layer morphology, which in turn impacts the transport of water, oxygen, and ions.In terms of ionomer content, a low loading can provide insufficient triplephase boundary and limit the accessibility to the active sites; while, an excessive ionomer content can limit oxygen and water transport.In addition, the ink solvent can modify ionomercatalyst interactions.For example, water-poor ink formulations have been shown to limit ionomer adsorption on Fe─N─C catalysts, resulting in the ionomer being located in the pores of the electrode, which provides good proton transport but high gas transport resistance.Conversely, water-rich inks were found to improve gas transport, providing enhanced fuel cell performance at high relative humidity (Figure 8f). [136]inally, all the studies reported above focus on catalyst layers deposited on a carbon support (usually referred to as CCS).Another common technique to prepare the catalyst layer is by depositing it directly on the membrane, obtaining a catalystcoated membrane, or CCM.CCM has received increasing attention as it normally features higher catalyst utilization, leading to higher electrochemical performance. [154,155]Therefore, the optimization of the ink composition and ionomer should also take into account the choice of deposition method, particularly as CCM might overcome the necessity of an AEI in KOH-fed AEM water electrolyzers. [4]n conclusion, the choice of solvent and ionomer loading can severely impact the morphology and performance of the catalyst  [146] Copyright 2017, The Electrochemical Society.b) The commercial anion-conductive AS-4 from Tokuyama.Adapted with permission. [147]opyright 2017, The Electrochemical Society.c) The commercial anion-conductive Fumion, from Fumatech.Adapted from Park et al. [89] 2020, Elsevier.d) The custom-made anion-conductive PPO-DMP.Adapted with permission. [89]Copyright 2020, Elsevier.e) The custom-made anion-conductive PPO-ASU.Adapted with permission. [89]Copyright 2020, Elsevier.f) Effect of water content in the ink on the resistance of the catalyst layer.For this work, Nafion was used as ionomer and the catalyst was the commercially available Fe─N─C catalyst PMF-011904, from Pajarito Powder LLC.Other than water, the solvent used was IPA.Adapted with permission. [136]Copyright 2020, Elsevier.g) Effect of water content in the ink and ionomer/catalyst ration on the viscosity of the ink (for this work, Nafion was used as ionomer and the catalyst was the commercially available Fe─N─C catalyst PMF-011904, from Pajarito Powder LLC; other than water, the solvent used was IPA).Adapted with permission. [130]Copyright 2020, American Chemical Society.
layer.Although several studies on ink composition have been reported, the optimal formulation is likely to be catalyst and ionomer dependent.Therefore, a careful ink optimization should be implemented when characterizing the performance of novel anion exchange ionomers.

Conclusion
While several studies have been dedicated to anion exchange membranes and the knowledge gained can be transferred to the development of ionomers, AEIs have additional and specific re-quirements and should be independently optimized.In this review, we have provided a summary of the desired properties of ionomers and highlighted the drastic effect that their optimization can have on the performance of fuel cells and electrolyzers.

Performance Comparison in Membrane Electrode Assembly
Fuel cells and electrolyzers are tested with varied and often inhouse made membranes, different catalysts, and a wide range of testing conditions, making it extremely challenging to compare the reported performance of fuel cells and isolate the effect of anion exchange ionomers.A standard testing protocol would allow an easier comparison, but the ideal AEI structure has been shown to be catalyst and membrane dependent.On the other side, halfcell measurements fail to capture the effect of mass transport in real devices.We take the view that the intermediate structure of gas diffusion electrodes could be a valuable technology to study the effect of anion exchange ionomers in isolation and simplify the cross-lab comparison of results. [156,157]

Catalyst Layer Morphology
51][152][131][132][133][134][135][136][137][138] However, very limited work has been dedicated to the effect of anion exchange ionomers on the deposition of transition metal catalysts, which tend to have higher surface area and larger particle size than platinum.Therefore, further investigation should be dedicated to optimizing the ionomer composition and ink formulation for PGM-free catalysts.

OH − Conductivity
Even though OH − conductivity has been thoroughly investigated for anion exchange membranes; very limited data exist on the effect of thickness.The few studies reported on the topic show that ionic conductivity in thin films is dramatically lower, compared to the bulk properties. [59,60]This highlights the importance of characterizing this property at ionomer-relevant thickness.

Water Uptake
The drastically different water management requirements for the cathode and anode of fuel cells urge a separate optimization of the anion exchange ionomer for the two electrodes.In particular, hydrophobic AEIs are desired for the oxygen reduction reaction and the oxygen evolution reaction, [68][69][70][71] which opens the challenge of reducing water uptake while maintaining high ion exchange capacity. [26]Finally, it is necessary to study more in depth, the kinetics of water transport, which has been largely overlooked but could be decisive for water management in alkaline fuel cells.

Gas Permeability
Gas, and in particular, oxygen permeability is another overlooked aspect of anion exchange ionomers particularly in the context of fuel cells.Multiple studies have shown that increasing the fractional free volume can be an effective approach to improve gas transport [24,76,[87][88][89] and we foresee that this unexplored area offers great potential to improve the performance of fuel cells.In this context, knowledge gained from "highly oxygen permeable proton exchange ionomers" could be transferred to AEI.

Stability Against Electrochemical Oxidation
While alkaline stability has been comprehensively investigated and several approaches have been proposed to limit OH − in-duced degradation, degradation pathways specific to the anion exchange ionomers have not received the same attention.In particular, electrochemical oxidation is likely to limit the stability of the cathodic ionomers of fuel cells and the anodic ionomer of electrolyzers [55,96,[101][102][103]106] and the mitigation strategies presently available have not shown sufficient durability.
Maria-Magdalena Titirici has a Ph.D. from the University of Dortmund in Germany followed by a postdoc at the Max-Planck of Colloids and Interfaces where she also become an independent group leader and did her Habilitation.In 2013, Magda moved to Queen Mary University of London as a "reader," and shortly after, was promoted to professor in sustainable materials.In 2019, she moved to the Imperial College London where she currently holds a "Chair in Sustainable Energy Materials."She is also an RAEng chair in Emerging technologies.Magda's research interests are sustainable materials and energy storage and conversion technologies.

Figure 1 .
Figure 1.a) Schematic of an anion exchange electrolyzer.b) Schematic of an anion exchange fuel cell.

Figure 2 .
Figure 2. Summary of most commonly reported polymer backbones and functional groups for anion exchange ionomers.

Figure 3 .
Figure3.a) Fuel cell performance results obtained with the mentioned combination of anion exchange membrane (from left to right MPRD, TMA, and MPY) and anion exchange ionomer (MPRD in blue, TMA in grey, MPY in red).Adapted with permission.[23]Copyright 2018, Royal Society of Chemistry.b) Effect of monomer order on the conductivity of a block co-polymer.Adapted with permission.[27]Copyright 2013, American Chemical Society.c) Synthesis and structure of radiation-grafted polymer typically reported as anion exchange ionomers.d) Synthesis and structure of the sulfonated, poly(ionic liquid) proton exchange ionomer reported by Snyder and coworkers.[28]e) Synthesis and structure of the anion exchange poly(ionic liquid) ionomer, reported by Ye et al.[27]

Figure 6 .
Figure 6.Strategies for increasing oxygen permeability in anion exchange ionomers.a) Comparison of the oxygen permeability of the ionomers, for which structures are shown below.b) Improving the oxygen permeability of the anion exchange ionomer PPO-ASU by changing the functional group to the more flexible piperidinium (PPO-DMP).[89]c) Improving the oxygen permeability of the anion exchange ionomer SP-DP by controlling its isomeric confuguration.[24]d) Improving the oxygen permeability of the anion exchange ionomer BP-PES by introducing the bulkier monomer SPI.[87] e) Improving the oxygen permeability of the proton exchange ionomer: sulfonated poly(ionic liquid) block co-polymer by the introduction of a monomer, with an oxygenophilic anion.[90]

Figure 8 .
Figure 8. SEM image of the catalyst layer obtained with Pt/C and the ionomers: a) the commercial proton-conducting Nafion.Adapted with permission.[146]Copyright 2017, The Electrochemical Society.b) The commercial anion-conductive AS-4 from Tokuyama.Adapted with permission.[147]Copyright 2017, The Electrochemical Society.c) The commercial anion-conductive Fumion, from Fumatech.Adapted from Park et al.[89] 2020, Elsevier.d) The custom-made anion-conductive PPO-DMP.Adapted with permission.[89]Copyright 2020, Elsevier.e) The custom-made anion-conductive PPO-ASU.Adapted with permission.[89]Copyright 2020, Elsevier.f) Effect of water content in the ink on the resistance of the catalyst layer.For this work, Nafion was used as ionomer and the catalyst was the commercially available Fe─N─C catalyst PMF-011904, from Pajarito Powder LLC.Other than water, the solvent used was IPA.Adapted with permission.[136]Copyright 2020, Elsevier.g) Effect of water content in the ink and ionomer/catalyst ration on the viscosity of the ink (for this work, Nafion was used as ionomer and the catalyst was the commercially available Fe─N─C catalyst PMF-011904, from Pajarito Powder LLC; other than water, the solvent used was IPA).Adapted with permission.[130]Copyright 2020, American Chemical Society.

Table 1 .
Summary of anion exchange ionomers reported in literature and their performance in fuel cells.

Table 2 .
List of commercially available anion exchange ionomers, their structure, and reported performance.