Characterization and Chemical Stability of Anion Exchange Membranes Cross-Linked with Polar Electron-Donating Linkers

The effect of the cross-linker chemical structure on the properties and chemical stability of anion exchange membrane is the focus of this study. Two different cross-linkers were investigated, one with linear hexyl chain between crosslinking sites, and the other, ether in the center of the alkyl linker. These two cross-linkers have a fundamental difference in their polarity and hydrophilicity. The ether-containing cross-linker is more polar and therefore will improve membrane’s water uptake and conductivity. Swelling and conductivity measurements were performed at various temperatures for both types of samples. While water uptake and conductivity were found to be higher for the ether-based cross-linker, degradation measurements indicated that the membrane that is cross-linked with electron-rich linker degraded in hydroxide faster than the alkyl linker at temperature of 60 ◦ C.

Fuel cell technology has been recognized as a promising clean energy conversion technology in next-generation energy systems. Fuel cells have been undergoing revolutionary developments in the last few decades. [1][2][3] Among the different types of fuel cell systems, proton exchange membrane fuel cells (PEMFC) are the most investigated and developed for low-temperature operation in portable and automotive environments. 4,5 Although there has been great progress in improving the performance and lifetime of PEMFCs and large-scale commercial applications are on the horizon, the largest barriers to wide-spread fuel cell technology are materials availability and cost. Among the limitations of PEMFCs, the dependence on noble metal catalysts is a critical concern. For example, the use of Nafion, the state of the art acidic membrane, generally allows only platinum group metals to be used as stable catalysts for long-term use, thus considerably increasing the cost of PEMFC technology. To overcome the requirement of precious metal catalysts, anion exchange membrane fuel cell (AEMFC) technology has begun to attract recent attention. This new effort focused on AEMFCs is due to several potential advantages that anion exchange membranes hold over their acidic membrane counterparts. In AEMFCs, oxygen reduction kinetics is improved in the high pH environment of the membrane and the use of non-precious metals as electrocatalysts greatly reduces the cost of fuel cell devices. [6][7][8] A highlyconductive, chemically robust anion exchange membrane (AEM) is a crucial component of the fuel cell as it acts as a barrier between the fuel and oxidant streams while simultaneously transporting anions from the cathode to the anode. The membrane component therefore requires considerable and thorough development as AEMs for electrochemical use are not widely commercially available. High hydroxide conductivity, good chemical stability and mechanical strength of the anion exchange membrane are essential for ensuring sustainable and durable operation of the cell. 9 Recently, a variety of AEMs have been studied for potential applications in AEMFCs, 10 including examples where the polymer backbone was polysulfone, 11,12 poly(arylene ether) containing triphenyl methane groups and 13 poly(arylene ether ketone), 14 poly(phenylene), 15,16 poly(phenylene oxide), [17][18][19] poly(styrene), 20,21 poly(ethylene), [22][23][24] or poly(ether imide). 25,26 Most of these AEMs are based on aromatic backbone polymers since these types of polymer architectures are considered to be promising as membrane scaffolds * Electrochemical Society Active Member. z E-mail: eineli@tx.technion.ac.il due to their high thermal stability, excellent mechanical strength, and relatively good chemical resistance under acidic and alkaline conditions. 27,28 In addition to the different chemical backbone structures of AEMs, the covalently attached cations on these polymers were derivatives of ammonium, 29,30 guanidinium, 31,32 phosphonium, [33][34][35] and imidazoilum [36][37][38] groups. It is known that the quaternary ammonium moiety is susceptible to hydroxide attack, resulting in degradation of alkaline membrane at elevated temperature under alkaline conditions, but clear pathways for degradation of polymer backbones and other functional groups of AEMs including non-ammonium cations has not been extensively pursued. In general, the degradation pathways for the tetra-alkyl quaternary ammonium cations include Hofmann degradation, E 2 -hydrogen elimination, SN 2 direct nucleophilic substitution, and nitrogen ylide formation. 39,40 These mechanisms have been verified in ex-situ testing of model compounds and materials, but in-situ degradation assessment in operating cells is still ongoing. Clearly, one of the goals for furthering AEMFC technology is to develop new AEMs that not only have high ionic conductivity but also exhibit improved chemical and mechanical stability in alkaline fuel cell conditions. High anion conductivity will improve the performance of AEM-FCs through increased catalyst layer activity and lower ohmic losses. A common strategy to increase ion conductivity in functionalized polymers is to raise the membrane's ion exchange capacity (IEC) by increasing the concentration of positively charged functional groups that are tethered to the polymer backbone. As IEC increases, however, the mechanical properties of the membrane often suffer, with excessive water swelling, poor mechanical strength, and brittleness when dry. 41 Cross-linking is a common technique to overcome these issues. Indeed, to date, chemical cross-linking was usually used to suppress the water swelling and thus improve the dimensional stability of membranes. Several groups have incorporated cross-linkable moieties in AEMs for example: N,N,N ,N -tetramethyl-1,6-hexanediamine (TMHDA), 42,43,44 dithiol 45 dialdehyde, 41 tetraalkoxysilanes, 46 and epoxy based cross-linkers. 47 Recently, progress in other cross-linking strategies has been reported. For example, Ni et al. and Pan et al. 48,49 showed that self-cross-linking represents a conceptual and methodological improvement in designing AEMs. Cross-linking has also been shown to be an effective strategy to enhance the membrane chemical stability in addition to decreasing the swelling ratio. Self crosslinking based on thermal Friedel-Crafts has also been introduced to yield materials with good thermal and dimensional stability and to Scheme 1. Water (represent in the scheme as circulars) flux inside an alkaline fuel cell. reduce swelling. 50 Cross-linking based on diamines or a mixture of monoamines and diamines were presented by Park et al. 51,42 Their work focused on the optimization of the amination time, reagent and casting solvent on the fuel cell performance. However, the alkaline stability, which is a major challenge in the development of cross-linked AEM, was not reported. Low temperature cross-linking by the addition of double bonds to the polymer structure and cross-linking by Grubbscatalyzed metathesis yielded AEMs with interesting phase separation properties. 52,53 These AEM examples also showed low swelling, good conductivity and reasonable chemical stability with the metathesis cross-linking scheme.
Another challenge in alkaline fuel cell development is improved methodologies for water management in the device. The water movement in an alkaline fuel cell is a key factor responsible for long-term stability and high performance. 54 Therefore, the idea behind the chemical structure of the cross-linker was to improve the performance of the alkaline fuel cell by choosing cross-linker that contains hydrophilic group, having a high tendency to interact with water molecules and retain hydration in the structure. This hydrophilic cross-linker has major influence on a number of factors. Firstly, retention of water will improve the membrane's conductivity. Further, greater hydration will decrease the degradation rate of the membrane in an alkaline environment. If OH − is completely hydrated the nucleophilicity of the ion is stabilized and therefore less reactive, thus the degradation processes involving hydroxide will be less aggressive. Finally, and maybe the most important reason for increased attention on water management in AMFCs is the alkaline fuel cell reactions, shown in Scheme 1. Clearly, hydrophilic molecules inside the cross-linker might "assist" the water molecules to transport from the anode, where they are generated, to the cathode, where they are consumed. It was found that the cell performance can be increased if the water movement from the anode to the cathode through the membrane will be improved, due to the promotion of the cathode relative humidity (RH) and the inhibition of anode flooding. 55 In this work, two types of cross-linked AEMs prepared by soaking chloromethylated aromatic polymer-based membranes in two different cross-linkers; a standard alkyl diamine cross-linker and a diaminoether cross-linker to increase the electron density of the linking group. The effect of the two different cross-linkers on AEM properties and their stability in alkaline and thermal conditions were the foci of this study to learn about how the chemical structure of the cross-linker impacts membrane properties.
Attenuated total reflectance Fourier transformed infra-red spectroscopy (ATR-FTIR).-The spectra of the membranes were obtained with the use of a Nicolet iS5 spectrometer equipped with a DTGS detector. A reflection ATR accessory equipped with a diamond crystal at an incident angle of 45 • was used. The membrane sample was pressed to the crystal by the clamp-kit to ensure reproducible contact between the sample and the ATR crystal. The spectrum was collected after 128 scans with a resolution of 4 cm −1 .
This mode of FTIR was preferred for two main reasons; the simplicity of the measurement in this configuration and the assumption that the initial degradation process will occur predominantly on the surface of the membrane.
Water uptake.-The water uptake of HCO 3 − form membranes, was defined as the weight ratio of the absorbed water to that of the dry membrane as given by: where, m D and m W are the mass of the membrane, before and after water absorption, respectively. The procedure of weighing wet membranes includes surface water elimination by rapid surface drying with a Kimwipe followed by drying in a vacuum oven for 24 h at a temperature of 60 • C.
High resolution scanning electron microscopy (HRSEM).-The analysis was carried out using a Zeiss Ultra-Plus FE-SEM. Samples of membrane were cut into squares of approximately 5 mm × 5 mm and fixed to an aluminum stub with double-sided conductive tape. Then the samples were sputter coated with carbon to improve conductivity. The observation was conducted with an acceleration voltage of 3 kV.
Conductivity.-The ionic conductivity (σ, S cm −1 ) of HCO 3 − form membranes (size: 10 mm × 20 mm) was measured by two probe in-plane impedance spectroscopy using σ = d/L s W s R (d is the distance between reference electrodes, L s and W s are the thickness and width of the membrane, respectively). The membrane impedance was measured over the frequency range from 10 kHz to 10 mHz by impedance spectroscopy (EIS), with an AC amplitude of 10 mV and a 0 mV DC bias, using an PARSTAT 2273 (Princton Applied Research). The resistance of the membranes was determined from the real part of the impedance at the minimum imaginary value. The measurements were conducted under fully hydrated conditions, with the sample membranes being immersed in water at elevated temperatures.
Elemental analysis.-A Flash 200 Thermo Scientific elemental analyzer was used to evaluate the nitrogen and carbon elemental weight percent of the materials. In order to track changes in the cationic group, ratio of nitrogen to carbon weight was calculated via: Thermal gravimetry analysis (TGA).-The thermal stability of the HCO 3 − form membranes was analyzed using a Q50 TGA, TA Instruments Corporation. TGA was used in order to evaluate the shortterm thermal properties and stabilities of the AEMs. The temperature was increased from room temperature to 500 • C at a heating rate of 10 • C min −1 .
Atomic force miroscopy (AFM).-Atomic force miroscopy in tapping mode was performed with a Agilent Technologies 5100 SPM, using micro-fabricated cantilevers with a force constant of approximately 40 N m −1 .
Small-angle X-ray scattering (SAXS).-Small-angle X-ray scattering curves of HCO 3 − form membranes were obtained using a Rigaku (formerly Molecular Metrology) instrument equipped with a pinhole camera with an Osmic microfocus Cu Kα source and a parallel beam optic. Typical counting times for integration over a multiwire area detector were 1 h. Measurements were taken under vacuum conditions. Scattering intensities were normalized for background scattering and beam transmission.
Where, m d is the mass of the dry membrane (dried at 60 • C in a vacuum oven for 24h), V AgNO 3 is the consumed volume of AgNO 3 solution and C AgNO 3 is the concentration of AgNO 3 solution.

Results and Discussion
ATR-FTIR.-The chemical structures of the membranes were investigated by ATR-FTIR and the absorption spectra of the membranes are shown in Figure 1a.
The absorption bands assigned to the polystyrene/poly(vinyl benzyl chloride) backbone structure were observed at 1616 cm −1 , attributed to a ring stretching of carbon-carbon bond, at 1512 cm −1 and 1488 cm −1 due to a ring deformation and stretching of the carbon-carbon bond and two additional small peaks at 1244 cm −1 and 1222 cm −1 assighned to C-H ring, in plane deformation. 56,57 The ether group in the cross-linker has usually two typical bands; one is strong band due to asymmetrical stretching (∼1125 cm −1 ), while the other is weak symmetric band (∼850 cm −1 ). Our FTIR investigation revealed only the strong asymmetric band that is located at 1125 cm −1 (Figure 1b).
Additionally, we will focus on 925 cm −1 peak, which is attributed to the quaternary ammonium (C 4 N + ) group, as this group is responsible for the ionic properties of the membranes. The peak assignment at 1125 cm −1 represents the fundamental difference between the membranes, being reflected in the appearance of the ether group in the AMINOETHER cross-linker, and absent at the TMHDA cross-linker, as can be seen in Figure 1b. The 925 cm −1 C 4 N + peak was chosen to track changes in the cationic group. This group dictates the IEC of the membrane which has a significant influence on the ionic conductivity therefore on the fuel cell performance. Consequently we will follow the changes in this group while being exposed to alkaline and thermal conditions. The relative abundance of the chemical group will assist us to learn about the stability of this group and to perform a quantitative analytical analysis, by measuring and calculating the ratio areas of the absorption peaks.
Water uptake and conductivity.-Water uptake and conductivity of the membrane has a major effect on the performance of the fuel cell. To improve both performance and stability of the alkaline polymer, the water related properties of the membranes should be considered.
It is well-known that AEMs in hydroxide form quickly convert to less conductive CO 3 −2 and HCO 3 − forms when exposed to CO 2 . 58,59 Thus, in order to avoid misleading and confusion in the interpretation of the results (avoid a mixture of anions as [OH − /HCO3 − /CO3 −2 ]), the conductivity and the water uptake were conducted in HCO 3 − form. Figure 2a shows that water uptake of both of the membranes increases with temperature. Additionally, the water uptake of the membrane that is cross-linked with AMINOETHER at room temperature (22 • C) is slightly higher than that of the sample cross-linked with TMHDA. This difference in water uptake can be attributed to the existence of the ether linkage which increases the electron density of the tether and imparts some polar character to the cross-linker. The electronegativity difference between carbon and oxygen (2.55 vs. 3.44 on the Pauling scale) is higher than between carbon and nitrogen (2.55 vs 3.04). Consequently, the membrane that is cross-linked with AMINOETHER has the potential to adsorb more water. As shown in Figure 2a, water uptake of TMHDA cross-linked membrane increased from 10 wt% at room temperature to almost 22 wt% at 60 • C, while the water uptake of AMINOETHER cross-linked membrane increased from 14 wt% at room temperature to 26 wt% at 60 • C.
The ionic conductivity of the membranes as a function of temperature is shown in Figure 2b. We assume that the ether group provided continuous sites for water molecules, which likely promoted hydroxide transport in the membrane system, and thus increased ion conductivity, as can be seen in Scheme 3.
It can be seen that the conductivity of AMINOETHER cross-linked membrane is higher, through the entire temperature range, than that of TMHDA cross-linked membrane. This difference is in agreement with the difference in water uptake measurements. The conductivity of the TMHDA cross-linked sample was found to be 1.7 mS cm −1 at room temperature and increased to 5 mS cm −1 at 60 • C. The conductivity of the sample with the AMINOETHER cross-linker was found to be 2.5 mS cm −1 at room temperature, increasing to 8 mS cm −1 at 60 • C. The conductivity of AMINOETHER cross-linked membrane was larger than the TMHDA cross-linked sample, due to its larger water uptake. Since the ion-conductivity is influenced by polar groups, we assume that the ether group in the AMINOETHER cross-linked membrane caused the membrane to be more polar and therefore, facilitate more water uptake to promote greater conductivity. losses. 60,61 The first decomposition step till 100 • C is due to water evaporation, the second at 135 • C is ascribed to the decomposition of HCO 3 − group. Decomposition of cationic groups is at 220 • C and the decomposition that starts at 420 • C is attributed to the polymer backbone.
Consequently, the presence of different cross-linkers had no significant influence on the thermal stability of the membrane. Therefore, the ether linkage had no impact on the thermal stability of the membrane in the HCO 3 − format. In order to keep the membrane hydrated to exhibit high ionic conductivity, the fuel cell should be operated at temperatures lower than 100 • C. Obviously, all components of the membrane, in their  non-hydrated form, are thermally stable till 100 • C (indicated by a vertical line in Figure 3).

Surface images of the membranes.-
The surface images of the membranes are presented in Figure 4. The surfaces were found to be homogeneous and showed only minor surface defects. The membrane that was cross-linked with TMHDA, displayed more organized structure that seemed to have a nano-porous structure due to the casting conditions. Similar porous structure was seen already in quaternary ammonium functionalized polydimethyl phenylene oxide. 62 Another group that reported on seeing similar porous structure is Nagarale et., 63 which reported on synthesis membrane that were prepared via an aqueous dispersion polymerization route and anionexchange groups were introduced in the membrane matrix by the chemical grafting of 4-vinylpyridine with the desired content. They reported that an increase in vinylpyridine content in the membrane matrix, led to increase in the membrane porosity. The reason for the porous structure in these membranes is probably surface defects that was created due to manufacturing. Thus, these images show that processing and casting conditions of the membranes must be considered in the fabrication of samples.
AFM.-The phase images of the AEMs recorded using tapping mode AFM are presented in Figure 5. The dark regions correspond to the ''soft'' hydrophilic ionic clusters and the bright domains correspond to the "hard" structures of the hydrophobic polymer backbones. 64,65 The cationic groups aggregate to form the hydrophilic domain that is hydrated upon absorption of water. The membranes exhibited small phase separated morphology with slightly different hydrophilic and hydrophobic domain sizes. Connectivity among the hydrophilic domains was observed mainly at the AMINOETHER cross-linked membrane. The hydrophilic domain size of AMINOETHER cross-linked membrane was larger than that of TMHDA cross-linked membrane. The hydrophilic domain sizes seem to be influenced by the cross-linker type. Meaning that the AMINOETHER cross-linked membrane had wider hydrophilic domains than the TMHDA membrane. These wide ionic domains were well connected to one another, and may provide a good ionic transporting pathway. These data on the hydrophilic domain sizes by AFM are in agreement with the water uptake and conductivity measurements.
SAXS data, presented in Figure 6, indicated on no characteristic phase separation for these membranes at the nano-scopic scale. However, it is known that SAXS measurements are not always able to detect phase separation, especially when relatively short side chains are present (weak separation between the two components) 66 the AEMs were evaluated using Elemental Analysis, FTIR, IEC and visual color changes. Elemental analysis.-Elemental chemical analysis is known to be a precise analytical tool for evaluating the elements content in the membranes. As can be seen in Figure 7, both of the membranes showed similar N/C ratio decreases of 20% after being held in 1 M KOH solution. In a more aggressive environment, 10 M KOH solution, TMHDA cross-linked membrane maintained 77% of the starting N/C ratio, while AMINOETHER cross-linked membrane preserved only 72% of the N/C ratio of the original material. The decrease in the N/C ratio can imply on decomposition of the cationic groups by the main degradation mechanisms. 68 The data for the AMINOETHER cross-linked membrane indicated that high pH environment led to more severe degradation in the stability of cationic groups with an electron-donating crosslinker. These results are the first indication for some negative influence of the ether linkage in the cross-linker on the stability of the quaternary ammonium groups under alkaline and thermal conditions. HRSEM.-The surface morphologies of the membranes were investigated after being exposed to alkaline environment. Figure 8 displays SEM images. Surprisingly, the surface of both of the membranes did not undergo any detectable change, or suffer from any surface degradation as cracks or holes subsequent to immersion in 10 M KOH solution at 60 • C for 500 h as was seen in the literature in other membranes after thermal and alkaline treatment. 69 The predominant conclusion from these images shows that the surface morphologies of some membranes can be unaffected after being subjected to high alkaline and thermal conditions. ATR-FTIR.-Chemical changes in the membranes during alkaline stability tests were investigated by ATR-FTIR spectroscopy. The backbone stability was not examined in this paper, due to overlap of the CH 2 vibrations from the backbone and from the cross-linker. The relative abundance of chemical groups was quantified from the areas of the FTIR absorption peaks. The ratios between areas can reflect changes in the chemical structure of the (degraded) membranes. In order to track changes in the quaternary ammonium (QA) group, the peak at 920 cm −1 that was assigned to C 4 N + vibration of the quaternary ammonium group was analyzed. The peak at 1475 cm −1 , assigned to the aromatic stretching of carbon-carbon linkage, showed high stability during the entire experiment and therefore, was used to normalize the spectra. The ratio between these two peaks was measured at different degradation time points in order to evaluate the stability of QA group vs. the aromatic group. Figure 9 shows a decrease of 30% in the QA group at 920 cm −1 corresponding to the C 4 N + stretch of the AMINOETHER cross-linked membrane in 1 M and serious decrease of 40% in 10 M KOH at 60 • C solution after 500 h. On the other hand, TMHDA membrane showed a slight degradation in the 1 M solution, while in the 10 M solution it showed a decrease of 30% in the C 4 N + stretch. These results are in slight disagreement with the elemental analysis results, since elemental analysis measurements provide results related to the N/C decrease, which is attributed to the main degradation cationic pathways (SN 2 and Hoffman elimination mechanisms), whereas FTIR measurements display a decrease in the C 4 N + group, attributed to all possible degradation pathways (Sommelet-Hauser rearrangement, Stevens rearrangement, direct nucleophilic substitution and Hoffmann elimination). Therefore, FTIR presents a higher loss of the cationic groups than the one observed in elemental analysis results. Ion exchange capacity.-The stability of the membranes in terms of their IEC was studied as well. The membrane that was cross-linked with AMINOETHER preserved 70% of its initial IEC in 1 M KOH solution within 500 h, as can be seen in Figure 10a, however in 10 M KOH the membrane showed poor long-term stability and preserved only 60% of its initial IEC values. On the other hand, TMHDA crosslinked membrane preserved 85% of its initial IEC in 1 M KOH solution within 500 h, as can be seen in Figure 10b, and 73% of its IEC in 10 M KOH. These results support the previous assumption that the ether group has a negative influence on the stability of the head groups in a high pH environment.
Visual color variations.-The tolerance and stability of the membranes toward alkaline media was observed visually from variations in color after immersion in 10 M KOH at 60 • C for 500 h. The membrane that is cross-linked with AMINOETHER soaked in 10 M KOH after 500 h presented a color change. With time, the color of the membrane changed from colorless to dark brown, as can be seen in Figures 11a-11b. Merle 70 and Wang 71 have already observed the same effect for polyether sulfone based anion exchange membranes and assigned this color change to a deterioration of the chemical structure, however they could not give support to the color change by proposing a chemical degradation reaction. Li 72 reported on a purple-colored phenyl-trimethyl ammonium (PTMA) functionalized polysulfone anion exchange membranes in the hydroxide form which were obtained by treating membranes in 1 M NaOH at room temperature for 24 h. The color change in the PTMA-polysulfone membranes is presumably due to the decomposition of PTMA under alkaline conditions during the hydroxide conversion step.
Nonetheless, TMHDA membrane did not change its color during the entire degradation experiment, as can be seen in Figures 11c-11d. Therefore, the degradation process of AMINOETHER membrane is different and probably faster from that of the TMHDA membrane. The difference is exclusively related to the difference in the cross-linker structure.
A possible explanation for the severe degradation of the AMINOETHER cross-linked membranes in terms of QA group, might be related to the presence of the ether group, known as an electrondonating group. Meaning, in the process of Hofmann elimination mechanism of the cationic groups, H β become much more acidic, leading to be preferable for OH − nucleophilic attack (Scheme 4). On the other hand, TMHDA cross-linked membrane, is absent of electron-donating group and therefore will degrade in a much slower process.
Unfortunately, we could not verify the creation of C=C bond through the Hoffman elimination mechanism by FTIR investigation.
..   The stretching vibration of the C=C bond (in the cross-linker) is usually gives rise to strong band in the region of 1630-1690 cm −1 , however in this case, this peak is considerably overlapped with 1450-1800 cm −1 region that has vibrations due to carbon-carbon stretching in the aromatic region, and the asymmetric vibration of the CO 2 of the bicarbonate anion (as can be seen from the highlighted region in Figure 12).

Conclusions
The effect of the two different cross-linkers on AEM properties and their stability in alkaline and thermal conditions was investigated and studied. One of the cross-linkers had an amino-ether and the other, had a trymethylhexadecyl amine group. At first, we confirmed the fundamental difference between the two cross-linkers by ATR-FTIR investigation. FTIR peak at 1125 cm −1 assigned to the ether group, appeared in the amino-ether cross-linker, and was absent at the TMHDA cross-linker. Water uptake and conductivity measurements showed that AMINOETHER cross-linked membrane has higher values through the 25-60 • C temperature range than TMHDA cross-linked membranes. AFM data supported these results by demonstrating wider hydrophilic domains for the AMINOETHER crosslinked membrane that contributed to efficient ionic pathways. Hence, the membrane that is cross-linked with AMINOETHER has the potential to adsorb more water and achieve higher conductivity values due to the presence of polar and hydrophilic group. Chemical stability results indicated that in high pH environment, AMINOETHER crosslinked membrane led to more severe degradation in the stability of cationic groups, expressed by a decrease in the N/C ratio in elemental analysis measurements. These results were supported by ATR-FTIR and IEC experiments. IEC results showed that AMINOETHER crosslinked membrane preserved 60% of its initial IEC values within 500 h, while TMHDA preserved 73% of its IEC in the same thermal and alkaline conditions. Our assumption for the fast degradation of the AMINOETHER cross-linked group in alkaline environment is related to presence of ether, known as an electron-donating group. During Hofmann elimination process, ether group led H β to be much more acidic and be preferable for OH − nucleophilic attack.