Synthesis of Triarylsulfonium Salts with Sterically Demanding Substituents and Their Alkaline Stability

Abstract

As cationic functional groups with excellent alkaline resistance that are potentially applicable to building blocks of robust anion exchange membrane (AEM) materials for water splitting and fuel cell modules, we describe the synthesis of triarylsulfonium (TAS) salts bearing sterically demanding substituents by the reaction of arynes with diaryl sulfides/sulfoxides and by the Friedel–Crafts reaction of diaryl sulfoxides. The TAS cations possessing three substituted benzene rings, such as tris(2,5-dimethylphenyl)sulfonium and bis(2,5-dimethylphenyl)mesitylsulfonium, were effectively produced through the appropriate choice of reactions and reagents. The alkaline stability of the TAS cations thus obtained was evaluated from their time-course ¹H NMR spectra in 1 M KOH/CD₃OD, from which the alkaline resistance of the TAS cations increased dramatically as the steric bulkiness of the aromatic substituents attached to the TAS cations increased. Among them, bis(2,5-dimethylphenyl)mesitylsulfonium was found to exhibit 25 times higher alkaline resistance performance compared to benzyltrimethylammonium, a conventional quaternary ammonium cation. The decomposition mechanism of the TAS cations in the basic methanol media was studied in detail, and it was concluded that the decomposition occurred by the nucleophilic ipso-substitution by the methoxide anions.

Introduction

Fuel cells and water electrolyzers are important for the realization of clean energy conversion systems that do not emit carbon dioxide. In particular, those composed of polymer electrolytes have attracted special attention for numerous applications, including vehicles, due to their suitability for manufacturing compact devices. In recent decades, polymer electrolyte fuel cells/water electrolyzers using proton exchange membranes (PEMs) have been widely studied. However, it is practically important to overcome economic issues of high operating costs since noble metal catalysts such as Pt are required as electrode materials. In contrast, those using anion exchange membranes (AEMs), which consist of polymers with cationic functional groups, are expected to be suited to fabricate low-cost devices since they can employ base metals such as Ni and Fe as electrode materials. To realize the AEM-based devices, however, it is very important to develop AEMs with high alkaline and oxidative stability.¹⁻⁴

Concerning the degradation of AEMs (i.e., polymers containing cationic functional groups), both the decomposition of the polymer backbones and the cationic functional groups should be taken into consideration. Many efforts have been made to suppress the degradation of polymer backbones, in which the durability toward both hydrolysis and oxidation has been improved by the structural design of the polymers.^5,6 Besides, studies on alkaline-tolerant ion-exchange groups have also focused on realizing robust AEMs for energy devices.

Tetraalkylammoniums, such as benzyltrimethylammonium (BTMA) are commonly used as cationic functional groups; however, they are susceptible to degradation by the S_N2 reaction and, hence, have poor long-term stability.^7,8 For example, it has been reported that approximately 90% of BTMA underwent degradation after 30 d at 80 °C in 1 M KOH/CD₃OH, while complete degradation took place after 20 d at 80 °C in 2 M KOH/CD₃OH.⁷ In addition, membranes composed of polymers containing the corresponding BTMA cations have also been reported to degrade under alkaline conditions, where their ion-exchange capacity (IEC) and ionic conductivity decreased to less than 50% by immersion in an aqueous 1 M KOH solution at 80 °C for 30 d.⁹ These examples strongly indicate the requirement for developing alkaline-tolerant cationic functional groups for the long-term operation of the AEMs.

Recently, several groups have been working on new organic cations that have improved alkaline stability.⁹ Within the series of nitrogen-based cations, alkyl-substituted ammoniums with acyclic,^8,10−14 cyclic,^15−22,38 and spirocyclic structures,^10,20,23,24 aryl-substituted ammoniums,²⁵⁻²⁷ and N-conjugated cations such as imidazoliums,³⁹⁻⁴⁴ triazoliums,⁷ and guanidiniums^36,37 have been synthesized, and their alkaline stability has been studied systematically. Among them, N-n-hexyl-N-methylpiperidinium, 6-azaspiro[5.5]undecan-6-ium, and 1,3-di-n-butyl-2-(2,6-dimethylphenyl)-4,5-diphenylimidazolium cations proved to exhibit remarkably high alkaline stability, in which almost no decomposition of the cations takes place over 30 d at 80 °C in 2 M KOH/CD₃OH (94%, 96%, and >99% remaining, respectively).⁷ These results indicate that the introduction of sterically demanding substituents and the delocalization of the positive charge are effective in facilitating alkaline stability.

Within the group 15 elements, cationic functional groups with phosphorus centers are also attractive for designing alkaline-stable cations for AEM applications. Previously, phosphorus-based cations such as benzyltrimethylphosphoniums, triarylmonoalkylphosphoniums,^28,29 and tetra(dialkylamino)phosphoniums³⁰ were developed, in which tetrakis[cyclohexyl(methyl)amino]phosphonium and tetrakis(pyrrolidine-1-yl)phosphonium showed no decomposition (>99% remaining) over 30 d at 80 °C in 2 M KOH/CD₃OH.⁷

In the case of the group 16 elements, a few examples of sulfur-based cations have been studied as alkaline stable functional groups for AEM applications. Noonan et al. synthesized dialkylamino-substituted sulfonium and sulfoxonium cations with high alkaline-resistant properties. For example, tris(piperidino)sulfoxonium ([O = S(pip)₃]⁺) showed 95% remaining over 30 d at 80 °C in 2 M KOH/CD₃OH. Its high alkaline stability is supposed to be due to the delocalization of the positive charge on both the S and N atoms.³¹

Some simple sulfonium salts with aliphatic and aromatic substituents have also been studied as cations for AEM materials. For example, Kim et al. reported the stable ionic conductivity of a membrane prepared from poly(fluorenyl ether sulfone) containing methyl-substituted sulfonium units throughout the immersion in an aqueous 1 M KOH solution at 60 °C over 5 d.³² Yan et al. reported the synthesis and alkaline-resistant properties of (4-methoxy-3-methylphenyl)diphenylsulfonium and (4-methylphenyl)diphenylsulfonium, in which the former, with the electron-donating methoxy group, did not degrade in 1 M KOH/D₂O at 60 °C for 10 d, whereas the latter decomposed completely under the same conditions.³³ Although these results may indicate that alkaline-stable cations are obtainable by introducing appropriate substituents, at least within the authors’ knowledge, further studies have scarcely been performed yet.

Recently, we described the synthesis of fully aromatic substituted phosphoniums, tetraarylphosphonium (TAP) cations, with the expectation that they may resist decomposition through nucleophilic substitution, elimination, and ylide formation reactions while minimizing the risk of oxidation. Their alkaline stability was proven to be strengthened with the increase of steric bulkiness around the cationic center of the TAP cations. Since TAS cations would also be an attractive platform to avoid decomposition and oxidation, we herein describe the synthesis of a series of TAS salts carrying sterically demanding substituents (Figure 1). Their alkaline-resistant performance was also studied in detail for AEM applications.

Figure 1.

Previously reported alkaline stable cationic functional groups^{7,10,16,29−31,33−35} and targets of the present study.

Results and Discussion

Synthesis of TAS Salts by Reactions of Diaryl Sulfides/Sulfoxides with Arynes

According to Peng and coworkers, the reaction of sulfides or sulfoxides with arynes generated in situ from 2-(trimethylsilyl)phenyl trifluoromethanesulfonate and its derivatives produces triarylsulfonium trifluoromethanesulfonate (TAS•OTf).^45,46 Following their procedure, the reaction of diaryl sulfides (2A–2D) or sulfoxides (3A–3D) possessing various substituents with aryne precursors (1a and 1b) was carried out in the presence of cesium fluoride, as shown in Scheme 1.

Scheme 1. Synthesis of Triarylsulfonium (TAS) Salts.

The sterically less demanding aryne precursor (1a) was first employed in the reaction with 2A–2D. The reaction proceeded smoothly in the cases of 2A, 2B, and 2C to produce the corresponding sulfonium salts (TAS-aA•OTf, TAS-aB•OTf, and TAS-aC•OTf) in excellent yields (94%, 86%, and 79%, respectively; Table 1). On the contrary, a sulfonium salt (TAS-aD•OTf) was not obtained by the reaction of a sulfide with sterically demanding dimesityl substituents (2D), which is in good accordance with the report by Peng and coworkers.⁴⁵

Table 1. Synthesis of TAS Salts by Reaction of Arynes with Diaryl Sulfides^a.

^aIsolated yield after purification by SiO₂ column chromatography.

The use of a more sterically demanding 3,6-dimethyl-substituted aryne precursor (1b) also proved to give the corresponding sulfonium salts (TAS-bA•OTf and TAS-bB•OTf) from 2A and 2B in 44% and 36% isolated yields, respectively. The yields of the products (TAS-bA•OTf and TAS-bB•OTf) were relatively lower in comparison with those of the corresponding salts produced from 1a. A sulfide with more sterically demanding diaryl substituents (2C) gave an insufficient result, where the sulfonium salt (TAS-bC•OTf) was obtained in below 10% yield. Similar to the case of the reaction with 1a, no product was obtained in the case of 2D.

The reaction of sulfoxides with arynes has also been reported by Peng and coworkers, where triarylsulfonium salts possessing two aromatic substituents derived from the aryne components are produced by the insertion of arynes.⁴⁶ With an expectation of producing more sterically demanding sulfonium salts, the reaction of diaryl sulfoxides (3B–3D) was carried out by using the aryne precursor with dimethyl substituents (1b) (Scheme 2). As expected, the corresponding sulfonium salts (TASO-bB•OTf–TASO-bD•OTf) were obtained in moderate to excellent yields (53%–96%, Table 2), which is higher in comparison with those for the salts (TAS-bB•OTf–TAS-bD•OTf) produced from the sulfides (2B–2D, vide supra). It is of note that the sterically demanding dimesityl sulfoxide (3D) also produced a sulfonium salt carrying two aromatic groups derived from the aryne species (TASO-bD•OTf) in an excellent yield, which is in accordance with Peng’s reported results, while the corresponding sulfide (2D) could not yield the sulfonium salt as mentioned above.⁴⁵

Scheme 2. Synthesis of Aryloxy-Substituted Triarylsulfonium (TASO) Salts.

Table 2. Synthesis of TASO Salts by the Reaction of Arynes with Diaryl Sulfoxides^a.

^aIsolated yield after purification by SiO₂ column chromatography.

Synthesis of TAS Salts by Friedel–Crafts Reactions

As an alternative synthetic method for more sterically demanding triarylsulfonium salts, the possibility of the molecular design based on the Friedel–Crafts reaction was examined. The Friedel–Crafts reaction of aromatic compounds with diaryl sulfoxides has been widely studied. However, to the best of the authors’ knowledge, few reports have attempted to prepare triarylsulfonium salts with sterically demanding substituents. Fortunately, the reaction proceeded smoothly even in the presence of such substituents on both aromatic compounds and diaryl sulfoxides. For example, the reaction of bis(2,5-dimethylphenyl) sulfoxide (3C) with p-xylene affords TAS-bC•OTf in 60% yield in the presence of trifluoromethanesulfonic anhydride (Tf₂O) (Scheme 3 and Table 3). The yield of TAS-bC•OTf is higher than that prepared by the aryne route using 2C and 1b (below 10%, vide supra). The result encouraged us to synthesize a more sterically demanding triarylsulfonium salt using mesitylene and 2C, from which TAS-cC•OTf was obtained in a high yield (86%). All of the TAS salts prepared in this study are solid materials, although they did not give good crystals for X-ray crystallographic analyses. Accordingly, full characterization of the TAS salts was performed by ¹H NMR, ¹³C NMR, and ¹⁹F NMR spectra and high-resolution mass analyses.

Scheme 3. Synthesis of TAS Salts from Diaryl Sulfoxides.

Table 3. Synthesis of TAS by the Reaction of Aromatic Compounds with Diaryl Sulfoxide^a.

^aIsolated yield after purification by SiO₂ column chromatography.

Evaluation of Alkaline Stability of TAS Cations

As mentioned in the introduction, Yan et al. have described the alkaline stability of two kinds of triarylsulfonium salts, (4-methoxy-3-methylphenyl)diphenylsulfonium and (4-methylphenyl)diphenylsulfonium, based on degradation experiments in aqueous alkaline media. Since the degradation behavior of many kinds of cationic functional groups has been studied in nonaqueous methanol alkaline solutions, where the degradation is dramatically accelerated under nonaqueous conditions (e.g., the degradation of triphenylsulfonium occurs 10⁶ times faster in ethanol compared to water, as reported by Oae and Khim),⁴⁷ the alkaline stability of the present sulfonium salts was assessed in nonaqueous methanol alkaline solutions. Specifically, the alkaline stability of the TAS cations was evaluated in 1 M KOH/CD₃OD solutions at 80 °C, and the degradation behavior was monitored by time-course ¹H NMR spectra. In the case of triphenylsulfonium (TAS-aA), peaks of aromatic protons of TAS-aA at 7.82–7.94 ppm disappeared completely within 2 d, indicating that TAS-aA degraded completely under the examined conditions. In this case, new peaks were observed at 7.25–7.35 and 6.90–6.94 ppm, respectively, which indicates that the degradation mechanisms may involve the ipso-attack of the methoxide anion (vide infra).

When TAS-aA was heated independently at 80 °C for 24 h in 1 M KOH/CH₃OH, and the Et₂O-soluble part collected was subjected to GC analysis, peaks assignable to diphenyl sulfide and anisole were detected (Figure S40). The purification of the Et₂O-soluble part by column chromatography provided diphenyl sulfide in 92% yield. Accordingly, the peaks observed at 7.25–7.35 and 6.89–6.92 ppm in the ¹H NMR spectrum are most probably attributable to diphenyl sulfide and anisole-d₃ (C₆H₅–OCD₃), respectively.

According to Khim et al., the decomposition of TAS-aA in basic conditions takes place by the nucleophilic attack of the anions on the S⁺ and the ipso-position of aromatic substituents (routes A and C, respectively).⁴⁷ Regardless of the degradation routes, diphenyl sulfide is produced, while the production of anisole may occur through routes B and C. In the case of tris(2,5-dimethylphenyl)sulfonium (TAS-bC), the degradation proved to be decelerated dramatically as monitored by the time-course ¹H NMR spectra, most probably owing to the sterically demanding substituents (Figure 2). After 1 day, almost no decomposition occurred, judging from the peaks of the aromatic protons. However, the H/D exchange proved to take place almost quantitatively at the o-methyl (the peak highlighted gray in Figure 2) but not at the m-methyl moieties. Over the prolonged period, new peaks assignable to the degraded products are observable in the ¹H NMR spectrum. For example, the ¹H NMR spectrum of the sample obtained after the alkaline treatment over 10 d exhibited new peaks highlighted in green and orange attributable to degradation products besides the original peaks highlighted in blue for the starting TAS-bC (Figure 2). In this case, the residual amount of TAS-bC was determined to be 54% from the integral ratio of the peaks for the aromatic protons at the p-positions (the doublet peak at 7.65 ppm) relative to that of 1,4-dioxane added as an internal standard (the singlet peak at 3.70 ppm). The peaks highlighted in green are assignable to bis(5-methyl-2-(methyl-d₃)phenyl) sulfide (5 in Figure 2) by comparing the chemical shifts of the peaks for bis(2,5-dimethylphenyl) sulfide measured in 1 M KOH/CD₃OD (Figure S42). The peaks at 6.96 (d, J_H–H = 7.6 Hz), 6.70 (s), 6.63 (d, J_H–H = 7.6 Hz), and 2.29 ppm highlighted in orange can be ascribable to partially deuterated 1-methoxy-2,5-dimethylbenzenes, since the ¹H NMR spectrum of an authentic nondeuterated 1-methoxy-2,5-dimethylbenzene taken in 1 M KOH/CD₃OD exhibited peaks in the same chemical shift regions, i.e., at 6.93 (d, J_H–H = 7.6 Hz), 6.67 (s), 6.61 (d, J_H–H = 7.6 Hz), 3.78 (s), 2.29 (s), and 2.12 (s) ppm (Figure S42).

Figure 2.

¹H NMR (400 MHz) spectra of samples obtained after the alkaline treatment of TAS-bC in a 1 M KOH/CD₃OD solution at 80 °C.

In the GC/EI-MS analysis of the sample obtained after the alkaline degradation experiment, 2-(methoxy-d₃)-1-(methyl-d₃)-4-methylbenzene (6 in Figure 3) and/or 2-(methoxy-d₃)-1-methyl-4-(methyl-d₃)benzene (7 in Figure 2) with a chemical formula of C₉H₆D₆O (M⁺ = 142) could be detected besides bis(5-methyl-2-(methyl-d₃)phenyl) sulfide (5) (Figure S43). On the basis of the plausible degradation mechanisms for TAS-aA, methoxy-d₃-substituted products would be produced from routes B and/or C. That is, the deprotonation of the aromatic C–H at the o-position brings about the production of 5 and an aryne, which may undergo the subsequent attack of CD₃O^– followed by the deuteronation to produce 2-(methoxy-d₃)-1-methyl-4-(methyl-d₃)-3-deuteriobenzene (8) and 2-(methoxy-d₃)-4-methyl-1-(methyl-d₃)-3-deuteriobenzene (9) with a chemical formula of C₉H₅D₇O (route B) and/or the ipso-attack of CD₃O^– toward the 5-methyl-2-(methyl-d₃)phenyl substituent that gives rise to 6 with a chemical formula of C₉H₆D₆O (route C). The result obtained from the GC/EI-MS measurement indicates that the reaction path shown in route C most probably operates in the present degradation process.

Figure 3.

Plausible degradation paths of TAS-bC in the presence of CD₃O^–: the nucleophilic attack of CD₃O^– at the cationic center (route A), deprotonation of the aromatic C–H at the o-position (route B), and ipso-attack of CD₃O^– (route C).

Figure 4a,c shows the time-course degradation behavior of the TAS cations and the TASO cations, respectively, in 1 M KOH/CD₃OD at 80 °C. The degradation rate of the TAS cations proved to decrease with the increase in the number of methyl groups on their aromatic substituents (i.e., the increase in the steric bulkiness of the TAS cations). For example, the unsubstituted TAS-aA (i.e., triphenylsulfonium) completely degraded within 2 d, while approximately 30% of TAS-aB, having two methyl groups at the o-positions of the aromatic substituents, remained under the same conditions. It was found that TAS cations with a much larger number of methyl substituents, such as TAS-bB, TAS-bC, and TAS-cC, degraded more slowly, and their degradation occurred much slower than BTMA, one of the commonly employed ammonium building blocks. It is of note that TAS-cC exhibits no significant degradation over 7 d. If TAS-cC was exposed for a longer period, slow degradation could be observed (91% and 82% remaining after 30 and 90 d, respectively, in 1 M KOH/CD₃OH at 80 °C). Under much more forced alkaline conditions, i.e., in 2 M KOH/CD₃OH at 80 °C, a substantial amount of the cation could still survive (69% and 35% remaining after 30 and 90 d, respectively, Figure S44 and Table S1), while BTMA exhibited complete degradation within 20 d under the same conditions.⁷ The enhanced alkaline tolerance could likewise be observed in the series of aryloxy-substituted triarylsulfonium (TASO) cations, where TASO-bB, TASO-bC, and TASO-bC exhibited 62%, 77%, and 82% remaining, respectively, after 7 d in 1 M KOH/CD₃OD at 80 °C (Figure 4c).

Figure 4.

Degradation curves of TAS cations (a) and TASO cations (c) in 1 or 2 M KOH/CD₃OD or CD₃OH at 80 °C, and the corresponding kinetic plots (b and d, respectively).

Based on the degradation mechanism described for TAS-bC (vide supra), the kinetic equation for the degradation of the TAS cations will be expressed using a kinetic constant (k) as follows.

1

Since [MeO^–] is large enough with respect to [TAS] that it can be treated as a constant value, the kinetic equation of the decomposition can simply be expressed as follows.

2

As plotted in Figure 4b,d, the k’ values clearly show the different degradability of TAS cations, from which k^–1 was calculated by the following equation.

3

As summarized in Table 4, TAS-cC was estimated to have alkaline tolerances approximately 25 times and 430 times higher as compared to BTMA and TAS-aA having no methyl group on the aromatic substituents, judging from their k^–1 values.

Table 4. k’ and k^–1 of Various TAS Cations in the Alkaline Degradation Experiments in 1 M KOH/CD₃OD.

	k’	k–1
TAS-aA	2.8 × 10–5	3.6 × 104
TAS-aB	6.5 × 10–6	1.5 × 105
TAS-bB	1.4 × 10–6	7.1 × 105
TAS-bC	7.0 × 10–7	1.4 × 106
TAS-cCa	1.3 × 10–7 (2 M)	1.5 × 107
TASO-bB	8.5 × 10–7	1.2 × 106
TASO-bC	3.6 × 10–7	2.2 × 106
TASO-bD	4.0 × 10–7	2.8 × 107
BTMA	1.7 × 10–6	5.9 × 105

^aEvaluated in 2 M KOH/CD₃OH.

Conclusions

In this study, various triarylsulfonium (TAS) salts with sterically demanding substituents were synthesized based on the addition reactions of diaryl sulfide/sulfoxide to arynes and the Friedel–Crafts reaction of diaryl sulfoxide. In the addition reactions to arynes, diaryl sulfoxides proved to give TAS salts in higher yields compared with the cases of diaryl sulfides with corresponding substituents. The Friedel–Crafts reaction was found to provide more sterically bulky TAS salts in good yields. The alkaline stability of the TAS cations was studied by monitoring their degradation behavior in 1 M KOH/CD₃OD, from which the degradation could be suppressed dramatically by increasing the steric bulkiness of the substituents. The detailed study on the degradation products revealed the plausible degradation mechanism, where the nucleophilic ipso-substitution of the methoxide anions toward the aromatic substituents takes place to give diaryl sulfide and anisole derivatives. Within the cations studied here, the bis(2,5-dimethylphenyl)mesitylsulfonium (TAS-cC) cation exhibited the best alkaline resistance, which corresponds to approximately 25 times higher stability than the commonly used benzyltrimethylammonium (BTMA) cation.

Since the polymers carrying the alkaline-tolerant TAS units have great potential for robust anion exchange membrane applications, the design and synthesis of TAS-containing polymers are currently being investigated.

Data Availability Statement

The data underlying this study are available in the published article and its Supporting Information.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.joc.4c03147.

• Detailed description of the experimental section (materials, preparation of diaryl sulfides/sulfoxides and TAS salts, and their characterization). Characterization of degradation products from TAS-aA and TAS-bC (¹H NMR, GC, GC/EI-MS). Detailed description of evaluation of alkaline stability of all cations (PDF)

Author Contributions

R.H.: validation, analysis. S.I.: validation. I.T.: methodology, conceptualization, validation, resources, writing—review and editing, visualization, supervision, project administration, funding acquisition.

The authors declare no competing financial interest.