# Fileset

[20240324 REVISED MANUSCRIPT CLEAN2 (J. Membr. Sci.).pdf](https://mdr.nims.go.jp/filesets/93525791-6647-4c44-ba42-a035276fd9b8/download)

## Creator

Fatin Bazilah Fauzi, [Je-Deok Kim](https://orcid.org/0000-0003-2442-1013)

## Rights

[Creative Commons BY-NC-ND Attribution-NonCommercial-NoDerivs 4.0 International](https://creativecommons.org/licenses/by-nc-nd/4.0/)

## Other metadata

[Comprehensive studies on sulfonated octaphenyl polyhedral silsesquioxane (SPOSS) using sulfuric acid: Structural analysis and composite crosslinked SPPSU/SPOSS membranes](https://mdr.nims.go.jp/datasets/60b088db-2777-4ceb-a87c-8d5dc52043b8)

## Fulltext

Microsoft Word - 20240324 REVISED MANUSCRIPT CLEAN2 (J. Membr. Sci.).docx1   1 Title: Comprehensive studies on sulfonated octaphenyl polyhedral silsesquioxane (SPOSS) using 2 sulfuric acid: Structural analysis and composite crosslinked SPPSU/SPOSS membranes 3  4 Authors: 5 Fatin Bazilah Fauzia 6 Email: fatinbazilah89@gmail.com 7  8 Je-Deok Kima,b 9 Email: kim.jedeok@nims.go.jp 10  11 Affiliations: 12 aFunctional Clay Materials Group, Functional Materials Research Center, National Institute for 13 Materials Science, 1−1 Namiki, Tsukuba 305−0044, Japan 14  15 bEnvironmental Circulation Composite Materials Group, Functional Materials Field, Research Center 16 for Electronics and Optical Materials, National Institute for Materials Science, 1−1 Namiki, Tsukuba 17 305−0044, Japan 18  19 Corresponding author: 20 Je-Deok Kim – Environmental Circulation Composite Materials Group, Functional Materials Field, 21 Research Center for Electronics and Optical Materials, National Institute for Materials Science, 1−1 22 Namiki, Tsukuba 305−0044, Japan; orcid.org/ 0000-0003-4301-1044; Email: 23 kim.jedeok@nims.go.jp 24   25 2  Comprehensive studies on sulfonated octaphenyl 1 polyhedral silsesquioxane (SPOSS) using sulfuric 2 acid: Structural analysis and composite crosslinked 3 SPPSU/SPOSS membranes 4 Fatin Bazilah Fauzi,a  Je-Deok Kim,a,b* 5  6 a Functional Clay Materials Group, Functional Materials Research Center, National Institute for 7 Materials Science, 1−1 Namiki, Tsukuba 305−0044, Japan 8 b Environmental Circulation Composite Materials Group, Functional Materials Field, Research 9 Center for Electronics and Optical Materials, 1−1 Namiki, Tsukuba 305−0044, Japan 10  11 Abstract 12 We prepared sulfonated octaphenyl polyhedral silsesquioxane (SPOSS) with its structure intact by 13 using sulfuric acid (H2SO4) at 40 °C. We discussed the effects of sulfonation using either 14 chlorosulfonic acid (ClSO3H) and H2SO4 on SPOSS structure and its incorporation into crosslinked 15 sulfonated polyphenylsulfone (CSPPSU) composite polymer electrolyte membranes (PEM). An in-16 depth structural characterization using 29Si NMR and FTIR spectroscopies, MALDI-TOF mass 17 spectrometry, and TGA show that phenyl loss occurred in SPOSS prepared using ClSO3H, resulting in 18 a low ion exchange capacity (IEC) of 1.7 meq/g and a sulfonation degree (DS) of 1.9. In contrast, the 19 phenyl remained intact in the SPOSS prepared using H2SO4. Well-dispersed 348-nm-SPOSS-H2SO4 20 was found to have an IEC of 2.4 meq/g and DS of 3.1 The resultant hybrid PEM of CSPPSU/SPOSS 21 using SPOSS-H2SO4 had improved swelling, mechanical and chemical strength, and conductivity. 1 22 wt% of SPOSS-H2SO4 loading had the highest conductivity (13.7 mS/cm) at 120 ºC with 40% RH. 23 This shows the importance of the SPOSS nanocomposite structure in dispersions in the SPPSU 24 polymer matrix and its performance as PEM. We report on an improved direct sulfonation process at 25 a milder condition but sulfonation yield was improved. 26  27 Keywords: Polymer electrolyte membrane; sulfonated octaphenyl polyhedral oligomeric 28 silesquioxane; nanocomposite; crosslinked SPPSU/SPOSS membrane. 29   30 3  1. Introduction 1 Greenhouse gas emission is a significant environmental problem, and 16% of worldwide CO2 was 2 produced from the transport sector in 2016 [1]. Green vehicles like plug-in hybrid and electric cars are 3 already marketed as highly efficient and zero-emission alternatives. However, due to low battery 4 performance, most are still limited to light-duty vehicles. A battery must be repeatedly recharged, 5 resulting in decreased battery life and frequent replacement [2]. Compared to batteries, polymer 6 electrolyte membrane fuel cells (PEMFC) have a higher energy density and faster charging rate. Recent 7 PEMFCs have shown improvements and have started to exceed the batteries currently used [3]. 8 PEMFCs are simple, light, and small, making them suitable for heavy-duty, long-range, high-9 utilization transportation means, such as buses, trucks, and trains [4,5].  10 Currently, commercially available polymer electrolyte membranes (PEMs) comprise 11 poly(perfluorosulfonic acid) (PFSA), like Nafion, which has a well-connected ion path resulting in 12 high conductivity in humid conditions [6,7]. In addition, it has high chemical resistance and 13 mechanical strength due to the strong C–F bonds. However, the biggest challenges for PFSA are its 14 poor performance at high temperatures (>80 °C) and low relative humidity (20% RH) [8-11]. Nafion 15 conductivity was 6.6 mS/cm under those conditions [12]. Moreover, PFSA has a high production cost, 16 and the waste is potentially hazardous [11,13,14]. To overcome the drawbacks of PFSA, the use of a 17 sulfonated aromatic hydrocarbon-based polymer as a greener and more economical PEM material has 18 attracted tremendous interest. 19 Results using sulfonated aromatic hydrocarbons such as polyimide [15-17], polyetheretherketone 20 [18-21], poly(arylene ether sulfone) [21,22], polyethersulfone [23,24] and polysulfone [25,26], have 21 shown that these materials have good potential as alternatives to Nafion. The molecular chain 22 containing an aromatic structure offers thermal and mechanical stability up to 200 °C, which is 23 promising for high-temperature applications [27,28]. The polyphenylsulfone (PPSU) backbone 24 includes electron-rich sites favorable for sulfonation, resulting in more sulfo groups per polymer [29]. 25 Thus, sulfonated PPSU (SPPSU) based membranes have high ion exchange capacities (IECs). For 26 instance, SPPSU prepared from the monomers bis(4-fluorophenyl)sulfone and 4,4’-biphenol have high 27 conductivities in the range of 100–200 mS/cm, which are higher than that for Nafion. However, the 28 sulfonation process is complicated, and the membrane’s durability remains an issue due to extreme 29 swelling and poor strength [30].  30 Kim et al. have comprehensively reported on developments in the large-scale fabrication of 31 crosslinked SPPSU (CSPPSU) membranes for many applications [30-34]. The effects of the co-solvent 32 and annealing and activation processes on the SPPSU membranes are discussed, and they have been 33 shown to affect the conductivity. A large-area (280 cm2) CSPPSU membrane activated to remove the 34 by-products has been reported to have a swelling of 43% and a high tensile modulus of 48 MPa. The 35 fuel cell using this CSPPSU has been shown to be stable up to 4000 h. However, the IEC was less than 36 2 meq/g, and the conductivity was 16 mS/cm (80 °C, 85% RH) [12].  37 Hybrid composite membranes can overcome low proton conductivities while enhancing the 38 4  membrane strength. Commonly used materials are inorganic composites with a high surface area and 1 thermal and mechanical strengths [35-37]. These composite additives also improve membrane water 2 retention and maintain ionic channels at high temperatures [38]. Hygroscopic oxides such as SiO2, 3 CeO2, TiO2, and ZrO2 were reported to enhance the proton conductivity at low relative humidity 4 conditions and the water retention of membranes [39]. However, the weak interaction between the 5 polymer matrix and such composites makes homogeneous mixing difficult. In an effort to improve the 6 weak interaction, recent reports on sulfonated carbon nanostructures as additives were shown to 7 improve the strength and performance of aromatic hydrocarbon-based PEM [40-42]. Currently, 8 controlling the size and aggregation of nanomaterials adequate to the ionic channel remains 9 challenging [43]. 10 Hydrophilic hybrid inorganic-organic particles, such as sulfonated octaphenyl polyhedral 11 oligomeric silsesquioxane (SPOSS), show promising IECs due to the eight phenyl rings available for 12 sulfonation. SPOSS has been commonly prepared using the superacid ClSO3H. This method produces 13 smaller particle sizes with a high degree of sulfonation, reaching 8 [44-46]. However, controlling the 14 sulfonation process at room temperature or higher is challenging, resulting in low yields [38]. Proper 15 control of the sulfonation conditions at a lower temperature is necessary. Membranes prepared via 16 sulfonation with a mixture of chloroform and diluted ClSO3H have been reported to have IECs less 17 than 0.5 meq/g [38, 47-49]. There has not been a thorough discussion on the sulfonation effects on the 18 OPOSS structure. Although several studies have highlighted the agglomeration issue involving SPOSS, 19 further studies are needed. 20 Previously, we have reported the preparation of SPOSS with an IEC of 7.84 meq/g. CSPPSU 21 composite membranes fabricated with 2 wt% of SPOSS have fuel cell performances comparable to 22 Nafion 117 at 80 °C (complete hydration). The SPOSS was sulfonated using ClSO3H at 50 °C for 1 23 day, followed by vacuum distillation at 110 °C before washing and drying. Although highly sulfonated 24 SPOSS was prepared, the SPOSS nanoparticles in the SPPSU polymer matrix were agglomerated and 25 unevenly distributed [50]. 26 In this report, we directly sulfonated commercially available OPOSS and thoroughly discuss the 27 structures of SPOSS prepared using H2SO4 (pKa = –2.8) and ClSO3H (pKa = –6.5). By using H2SO4, 28 the sulfonation was more controllable, and the SPOSS suffered less damage, as confirmed by using 29 NMR and FTIR spectroscopies, TGA, and MALDI-TOF mass spectrometry. Nearly double the number 30 of sulfonic moieties retained in SPOSS prepared using H2SO4 compared to ClSO3H. In comparison 31 with the CSPPSU membrane reported by Kim et al. [12], the conductivity of the CSPPSU/SPOSS-32 H2SO4 composite membrane increased by 2.7 times to 43 mS/cm at 80 °C and 85% RH and by 5.5 33 times at 120 °C and 40% RH to 13.7 mS/cm, which is comparable to Nafion212. When ClSO3H was 34 used to prepare the SPOSS, the tensile modulus of the CSPPSU/SPOSS decreased below 700 MPa up 35 to 10 wt%. However, when H2SO4 was used, it exceeded 1000 MPa. In-depth comparisons of both 36 sulfonation methods for preparing SPOSS and its loading in the CSPPSU/SPOSS composite 37 membranes show the importance of the SPOSS structural integrity in composite CSPPSU membranes. 38 5   1 2. Experimental 2 2.1 Materials 3 Polyphenylsulfone (PPSU; Radel R-5000 NT; Solvay Specialty Polymers Japan K.K.; Mw = 50,000), 4 octaphenyl polyhedral oligomeric silsesquioxane (OPOSS; C48H40O12Si8; Hybrid Plastic Inc.), 5 Nafion® PFSA membrane of NR-212 (~50 μm) and NR-211 (~25 μm) from DuPont, USA, 6 concentrated sulfuric acid (H2SO4; 98.8%; Nacalai Tesque, Inc; CAS: 7664-93-9), chlorosulfonic acid 7 (ClSO3H; 97.0%; Nacalai Tesque, Inc; CAS: 7790-94-5), dimethyl sulfoxide (DMSO; Sigma-Aldrich 8 Co., Ltd.; CAS: 67-68-5), sodium hydroxide (NaOH; 97.0%; Nacalai Tesque, Inc, Japan; CAS: 1310-9 73-2), and sodium chloride (NaCl; 99.5%; Nacalai Tesque, Inc, Japan; CAS: 7647-14-5) were used 10 without any further purification. 11  12 2.2 Synthesis of SPPSU and SPOSS 13 Commercial PPSU was sulfonated through an electrophilic aromatic substitution reaction using a 14 concentrated H2SO4 at 40–60 °C for 24–48 h, as shown in Fig. 1. The details of the sulfonation are 15 described in a previous study [32]. The molecular weights of the SPPSUs were determined to be 16 395,692 (40 °C; 24 h) and 215,604 (60 °C; 48 h) by using gel permeation chromatography (TOSOH 17 HLC-8220 GPC; Shimadzu GPC 850D column) at 60 °C with N,N-dimethyl formamide (DMF) as the 18 eluent.  19 The sulfonation of OPOSS was performed using either concentrated H2SO4 or ClSO3H. First, 8 g 20 of OPOSS was dried at 80 °C for 48 h and then mixed with 400 ml of H2SO4 or ClSO3H while stirring 21 in a digital oil bath (SOR-243D; SANSYO Co, Ltd.) under the conditions listed in Table 1. The solution 22 was cooled in an ice bath, producing a white precipitate, which later was collected using vacuum 23 suction filtration. The crude white product was transferred to a dialysis tubing cellulose membrane 24 (Spectra/Por 6; molecular weight cut-off, MWCO = 1 kD) and repeatedly washed with deionized water 25 (PURELAB Option-R7, 15 MΩ cm at 25 °C) until the pH was 7. The solution was again filtered using 26 filter paper (Advantec 5A) to remove the large impurities. The filtrate was heated on a hot plate to 27 obtain a concentrated SPOSS/H2O solution.  28  29 2.3 Ion exchange capacity (IEC) and degree of sulfonation (DS) of SPPSU and SPOSS 30 The IECs of the SPPSU and SPOSS polymers were determined using a titration method. Solid 31 SPOSS was prepared by drying the SPOSS/H2O solution on a hot plate. SPOSS and SPPSU (5 mg) 32 were immersed in 20 ml of 2 M NaCl for 24 h. The resulting acidic solution was titrated with 0.01 M 33 NaOH. This method maximized the dissociation of the sulfonic acid group and gave the most accurate 34 measurement of the IEC value [51]. IEC (milliequivalent per gram, meq/g) was calculated using eq. 1 35 (C is the concentration; V is the volume; W is the weight): 36 IEC meq/g 𝐶  M  𝑉  ml /𝑊  g          (1) 37 The degree of sulfonation (DS) was calculated using eq. 2, where Mw is the molecular weight of the 38 6  repeating unit of the polymer (R.U. of SPPSU = 400.45, SPOSS = 1033.53) and the sulfonic acid unit 1 (SO3 = 80.06). The IEC and DS of the SPPSU are the average of two SPPSU samples prepared at 40 °C 2 for 24 h, and at 60 °C for 48 h. 3 DS   . . .  .              (2) 4 1H nuclear magnetic resonance (NMR) of SPPSU was performed using a similar approach to our 5 previous report [31]. In eq.3, DS was calculated based on the peak area analysis of 1H NMR spectrum 6 plotted in Fig. S1, where Aabc and Ade are the sums of the area of the proton peaks of a,b,c,d and e 7 found at the chemical shift range of 7.0 to 8.5 ppm [52,53]. The DS obtained from 1H NMR and 8 titration methods were 2.1 and 1.9, respectively. The sulfonation details, IECs, and DSs of the SPPSU 9 and SPOSS are listed in Table 1.   10 DS     /  /               (3) 11  12 2.4 Structural characterizations of SPOSS 13 Solid sampling matrix-assisted laser desorption/ionization mass spectroscopy (MALDI-MS; JMS-14 S3000 SpiralTOF™; JEOL, Ltd.), 29Si NMR spectroscopy (Avance 400 MHz; Bruker), and Fourier-15 transform infrared spectroscopy (FTIR, Nicolet 6700 FT-IR Spectrometer; Thermo Fisher Scientific) 16 were performed to identify the type of molecular species, the bonds, and the structures of SPOSS 17 nanoparticles. The thermal stability was determined using thermogravimetric analysis (TGA) (DTG 18 60AH; Shimadzu) in flowing nitrogen and oxygen gas with a flow rate of 600 sccm. The temperature 19 increased at a rate of 4 °C/min. The thicknesses of the dried and swollen membranes were measured 20 using a Digimatic thickness gauge (547-401 ABSOLUTE; Mitutoyo, Japan). SPOSS prepared using 21 H2SO4 was dispersed in water, and the particle size was measured using dynamic light scattering (DLS, 22 Particle size Analyzer ELSZneo; Laser wavelength: 660 nm). 23  24 2.5 CSPPSU/SPOSS composite membrane preparation 25 First, solvent exchange was performed to prepare various concentrations of SPOSS/DMSO from 26 SPOSS/H2O. DMSO (5 g) was added into SPOSS/H2O while stirring at 80–100 °C until the water was 27 removed entirely. Separately, 0.3 g of SPPSU was dissolved in 3 g of DMSO. To this solution (10 wt% 28 SPPSU/DMSO), SPOSS/DMSO was slowly added while rapidly stirring, and the mixture was stirred 29 for a few days until a clear, well-dispersed cast solution (5 g) was obtained. The SPPSU/SPOSS/DMSO 30 solution was cast in a 70 mm petri dish. The 4-step drying and crosslinking process was performed at 31 Table 1 Conditions for the sulfonation process of the SPPSU and SPOSS by using H2SO4 and ClSO3H along with their IEC and DS values. Polymer Acid Temp. (ºC) Time (h) IEC (meq/g) DS SPPSU H2SO4 40 – 60 24 – 48 3.3 1.9 SPOSS ClSO3H –10 24 1.7 1.9 SPOSS H2SO4 40 48 2.4 3.1 7  60, 120, 160, and 180 °C for 24 h each in an oven (DKN 302; Yamato Scientific Co., Ltd.). ~50 µm-1 thickness of CSPPSU/SPOSS composite membranes with SPOSS loading in the range of 1–50 wt% 2 were prepared and then activated by immersion into 0.5 M NaOH, followed by 1 M H2SO4 for 24 h 3 each, alternating with thoroughly washing in DI water. Activation is a crucial procedure to remove the 4 by-products obstructing the ion-conducting path [12]. Lastly, the activated CSPPSU/SPOSS 5 membranes were dried at RT for 24 h. 6  7 2.6 CSPPSU/SPOSS composite membrane characterizations 8 TGA, X-ray fluorescence spectroscopy (XRF; ZSX PRIMUS II; Rigaku), and a Tension Test 9 Machine (EZ-S, Shimadzu) were used to evaluate the thermal/mass properties, elemental analysis, and 10 the stress-strain of the CSPPSU/SPOSS composite membranes. Microstructure studies were done 11 using small angle X-ray scattering (SAXS) in the beamline of BL-6A (Wavelength: 0.15 nm; Camera 12 length: 1000 mm; KEK Tsukuba, Japan). The average spacing calculated from Bragg’s equation is 13 ascribed to the average dimension of the ionic cluster using eq. 4. L is the average spacing (nm), and 14 q is the scattering vector, which is expressed as eq. 5, with 2θ as the scattering angle and λ as the X-15 ray wavelength [12,54,55]. 16  Fig. 1 Schematic representation of the structure of SPPSU, SPOSS and crosslinked SPPSU/SPOSS composite membrane. 8  𝐿 = 2𝜋 𝑞⁄                  (4) 1 𝑞 4𝜋 𝜆 𝑠𝑖𝑛𝜃⁄                 (5) 2  3 2.7 Ion exchange capacity (IEC), degree of crosslinking (DCL), water uptake (WU), hydration number 4 (λ), and oxidation stability of CSPPSU/SPOSS composite membrane 5 The IEC of the composite membrane before the crosslinking was estimated using eq. 6 (m is the 6 mass fraction) [45]. The IEC after the crosslinking was determined using a titration method (eq. 1). 7 IEC  meq/g IEC  𝑚  IEC   𝑚       (6) 8 Eq. 7 was used to calculate the composite membrane’s DCL from the changes in the membrane’s 9 IEC before and after crosslinking. 10 DCL % IEC  IEC /IEC 100          (7) 11 Water uptake (WU) was calculated from the weight difference between the dry and swollen state 12 using eq. 8. The membrane was dried at 80 °C for 24 h to measure dry weight, and Wdry and Wwet were 13 measured with the membranes in a swollen state after being immersed in water for 1 h at RT and 100 14 ºC. Hydration number (λ) was then calculated using eq. 9: 15 WU % 𝑊   𝑊 /𝑊   100          (8) 16 𝜆 H O / SO H 1000 𝑊  𝑊 / 18 𝑊 IEC        (9) 17 A simple Fenton’s test was done to evaluate the oxidative stability of the CSPPSU/SPOSS 18 composite membranes following the previous report [12]. The interaction of aqueous H2O2 with the 19 Fe2+ ion formed peroxyl/hydroxyl radicals, which react with the polymer and initiate its chemical 20 decomposition [13].  21  22 2.8 Proton conductivity of CSPPSU/SPOSS composite membrane 23 Proton conductivities were measured using 4-probe impedance spectroscopy (1 Hz–1 MHz; peak-24 to-peak voltage, VPP = 10 mV; MTS, Scribner Associates, Inc.) in the temperature range of 40–120 °C 25 with a series of relative humidities from 0% to 85%. The chamber was pressurized to 130 kPa for T > 26 100 °C measurements. The proton conductivities were calculated on the basis of the resistances 27 obtained from the Nyquist plot (d is the thickness of the membrane (cm); A is the cross-sectional area 28 of the membrane (cm2); R is the resistance (Ω)). 29 𝜎 S/cm 𝑑/ 𝐴 𝑅                (10) 30 The activation energy, (Ea, kJ/mol) was calculated from Arrhenius plots (eq. 11) for the proton 31 conductivities at 40, 60 and 80 °C in the relative humidity range of 20%–80%. Here, σ is the proton 32 conductivity, R is the gas constant (8.314 J/mol K), T is the temperature (K), and σo is the pre-33 exponential factor (S/cm). 34 𝑙𝑛 𝜎 𝑙𝑛 𝜎               (11) 35  36 2.9 Single cell performance of CSPPSU/SPOSS composite membrane 37 9  Large area membranes of 120 cm2 was prepared by using a 10-cm-length knife applicator (Tester 1 Sangyo CO., Ltd., Japan) using an automatic film-coating apparatus (KIPAE, KP-3000VH). Unlike 2 small area membranes, casting solutions with higher concentration of 31 wt% in DMSO was used to 3 fabricate membranes of CSPPSU and 1 wt% CSPPSU/SPOSS-H2SO4, casting at 60 ºC (0.2-mm-gap; 4 20 mm/min speed). The membrane proceeded to a 5-step annealing treatment of 60 ºC, 120 ºC, 160 5 ºC, and 180 ºC for 90 minutes each step and at 200 ºC for 12 hours. These crosslinked membranes 6 were activated for 1 h at 80 ºC in both 0.3 M of NaOH followed by 0.5 M of H2SO4. Membranes were 7 washed until it became neutral in water after each step. After activation process, it was dried at RT for 8 24 h. Good, uniform, and transparent membranes with thicknesses of 30–40 μm were successfully 9 prepared at a shorter annealing time and resisted high-temperature activation.  10 A single cell performance was performed on fuel cell system from AutoPEM of Toyo Corporation, 11 Japan. The catalyst electrodes were 0.3 mg/cm2 Pt/C/ionomer (ionomer/carbon=1) on GDL electrode. 12 The membrane electrode assemblies (MEA) have an effective electrode area of 4 cm2. The cell 13 operated at atmospheric pressure, 80 °C temperature, and 100% RH. Feed gas supplied were 50 sccm 14 pure H2 and 100 sccm pure O2 to the anode and cathode, respectively. Detailed information on the 15 procedure can be found in our previous report [12]. 16  17 3. Results and discussion 18 3.1 Structural characterization of SPOSS 19 Thermal decomposition of OPOSS and SPOSS in N2 was observed using TGA, as shown in Fig. 2a. 20 The first derivative weight loss curves indicate the temperatures of the maximum rate of change 21 corresponding to the weight loss curves. In addition, TGA was performed under oxidative conditions 22 using air, as shown in Fig. S2. The details of the weight loss percentage, the maximum degradation 23 (Tmax), and the onset (Tonset) temperatures are listed in Table 2. The OPOSS had good thermal stability 24 up to 350 °C, and its Tonset was 428 °C. Then, a two-step decomposition occurred with a total weight 25 reduction of 29.3%. Decomposition of OPOSS under N2 only allows sublimation and condensation to 26 occur [49,50]. The Tmax was determined from the first derivative weight loss plot. The first Tmax was 27 428 °C, and the second was 595 °C. These steps represent the condensation of the phenyl groups and 28 sublimation of the T8-caged POSS structure. Large char residues of about 70% yield at 600 °C are 29 typical for POSS and depend strongly on decomposing the substituted grafted organic groups [56-58]. 30 Longer aliphatic and aromatic substituents cause the polymerization of POSS, resulting in carbon 31 entrapment and a larger amount of residue [59]. 32 10  When SPOSS was prepared using H2SO4, a three-step decomposition was observed. The first 1 weight loss of 5.6% was observed at 84 °C, and this was due to the evaporation of the bulk water 2 associated with the sulfonic acid groups. Then, a weight loss of 3.2% occurred in the region of 150–3 350 °C due to the decomposition of sulfonic acid groups [13]. Weight losses of 19% were observed 4 from the decomposition of the OPOSS backbone at Tmax values of 428 and 564 °C. The final residue 5 was 72.2% of the original mass. The TGA curve for SPOSS-H2SO4 has a profile similar to that of the 6 unsulfonated OPOSS. This indicates that the original OPOSS structure is maintained even after 7 sulfonation. As shown later using MALDI-TOF, the OPOSS structure remained intact even after 8 sulfonation using H2SO4 (Fig. 2b–2c).  9 In contrast, the TGA for SPOSS sulfonated by ClSO3H only has a two-step weight loss. It starts 10 with a weight loss of 5% at 84 ºC from water attached to the sulfonic acid groups. The sulfonic acid 11 gradually decomposed from 150–350 °C, resulting in a weight loss of 1.7%, almost half that of SPOSS-12 H2SO4. This reduced weight loss confirmed the lower degree of sulfonation and IEC values (Table 1). 13 Only a one-step decomposition of the OPOSS backbone was observed at a Tmax of 503 °C with a weight 14 loss of 16.2%, leaving a residue of 77%. The Tonset for SPOSS-H2SO4 was 125 °C, whereas it was 15 145 °C for the SPOSS-ClSO3H. Attachment of the -SO3H accelerates the decomposition of the polymer, 16 leading to a lower Tonset [60]. This one-step decomposition differed significantly from those for OPOSS 17 and SPOSS-H2SO4, strongly suggesting damage to the original silsesquioxane structure. A detailed 18 analysis of its structure will be provided later. 19  Fig. 2 a) TGA curves of SPOSS-H2SO4 and SPOSS-ClSO3H with comparison to OPOSS in N2 at 4 ºC/min heating rate, MALDI-TOF mass spectrum of b) positive ions using DCTB (trans-2-[3-(4-tert-butylphenyl)-2-methyl-2-propenylidene]malononitrile) matrix and, c) negative ions using IAA (3-indoleacrylic acid) matrix of SPOSS-H2SO4 (Only peaks with intensity values above 300).  Table 2 TGA data of the OPOSS and SPOSS prepared by H2SO4 and ClSO3H. Sample Weight loss (%) Tmax (°C) Tonset (°C) Residue (%) RT–150 °C 150–350 °C 350–700 °C OPOSS - - 29.3 428, 595 428 70.7 SPOSS-H2SO4 5.6 3.2 19.0 84, 428, 564 125 72.2 SPOSS-ClSO3H 5.0 1.7 16.2 84, 503 145 77.1 11  The MALDI-TOF mass spectra of the SPOSS sulfonated using H2SO4 are shown in Fig. 2b and 2c. 1 Please see the supporting information for the full range of the mass spectra (Fig. S3 and S4). In the 2 positive ion mode in Fig. 2b, strong peaks appeared corresponding to the intact structure of OPOSS 3 (m/z = 1055 for OPOSS-Na+) [61] and OPOSS-SO3H (m/z = 1157 for OPOSS-SO3Na-Na+). 4 Sulfonation of the phenyl group increased the mass by 80. Simultaneously, side reactions resulting in 5 broken cage structures of OPOSS are believed to have occurred in the presence of strong acids. Cage 6 opening begins with bond cleavage of the Si-O bond through hydrolysis, as shown in eq. 12 [62]. This 7 cleavage causes the molecular weight to increase by 17. In addition, further hydrolysis could result in 8 the complete elimination of a Si-Ph vertex (eq. 13). In this case, the weight will be reduced by 108. 9 SiOSi H O SiOH  HOSi             (12) 10 Ph Si O 3H O  Ph Si O OH  PhSi OH          (13) 11 Thus, the phenyl loss and fragmentation of the Si-O-Si cage structure are expected to yield 12 fragments with m/z values of 1174 (Ph8Si8O11OH2-Na+), 996 (Ph7Si8O12OH-Na+), and 954 13 (Ph7Si7O9OH3-Na+) [63]. However, there are no such peaks apart from a weak peak at 955 m/z 14 corresponding to the fragment shown in eq. 12 (Fig. S3). The broken structure was detected at m/z = 15 955, as shown in Fig. S5a. 16 The negative ion MALDI-TOF mass spectrum using an IAA (3-indoleacrylic acid) matrix was also 17 acquired. The noise was much higher, and the intensity was lower in the negative ion mode due to 18 measurement parameters and settings favoring higher signal intensities of the cations, disregarding the 19 different mechanisms [64]. We re-plot the MALDI-TOF mass spectra to increase the intensity higher 20 than 300 (Fig. 2c). Peaks for complete cages of OPOSS-SO3– (m/z = 1111) and OPOSS-(SO3H)2-H– 21 (m/z = 1191.95) were observed. Fragments believed to originate from the open cage structures were 22 also detected. The fragments associated with the degree of sulfonation from one to three were observed 23 at m/z values of 1146.31, 1213.93, and 1294.89, respectively (Fig. S5b–d)). The MALDI-TOF in the 24 negative ion mode was used to confirm that sulfonation degrees of up to 3 (m/z = 1294.89 for 25 Ph8Si8O11OH2-SO3H3-H–) were obtained. Both positive and negative mode MALDI-TOF mass spectra 26 show that most of the phenyl groups remain intact in the structure of SPOSS-H2SO4. 27 Fig. 3a shows 29Si NMR spectra of the SPOSS prepared using H2SO4 and ClSO3H. The quaternary 28 (Qn) and trifunctional (Tn) structures were confirmed to be present in the SPOSS. Qn is the notation 29 for Si(Si-O)n(OR)4–n (n = 1–4), and Tn is the notation for PhSi(Si-O)n(OR)3–n (n = 1–3) [65]. For the 30 original OPOSS, only a single T3 peak for the PhSi(Si-O)3 structure appeared at –78 ppm, 31 corresponding to an eight vertex Ph-Si moiety. After sulfonation, hydrolysis of the Si-O-Si bond 32 produced Q2, Q3, and Q4 peaks at –91, –101, and –111 ppm, respectively. Concurrently, T1 and T2 33 peaks due to the dissociation of the phenyl groups, affording incomplete Ph-Si-O-Si bonds, were 34 observed at –67 ppm and –77 ppm. The significant difference in the T3 peak intensity indicates phenyl 35 loss during sulfonation using ClSO3H [38, 48]. A sharp T3 peak for SPOSS-H2SO4 remained, 36 12  suggesting a higher proportion of phenyl-Si bonds.  1 The control of the sulfonation process is essential as the Si-C bonds of the aromatic structure are 2 very sensitive to the strong electrophile substitution process [48]. Aromatic ring cleavage or Si 3 replacement with electrophiles has been reported to occur when a mixture of ClSO3H and chloroform 4 is used [38,48]. As a result, the acid capacities of these sulfonation agents’ resulting acids are lower 5 than that for complete sulfonation with no phenyl loss. In this case, complete sulfonation was one 6 SO3H per phenyl. However, only 30%–40% of the phenyl groups remain attached to the POSS siloxane 7 skeleton [48]. A schematic of the aromatic electrophilic substitution of the OPOSS is illustrated in Fig. 8 3a, along with the 29Si NMR spectra. Meta-substitution generally occurs when the sulfonic acid 9 electrophile replaces the hydrogen atom [38,49]. However, high acid concentration and sulfonation 10 temperatures cause the phenyl to dissociate through ipso-substitution of the Si atoms, producing an 11 unfunctionalized phenyl group [48]. Thus, the loss of the phenyl group led to drops in the IEC and 12 degree of sulfonation. Sulfonation of OPOSS using H2SO4 mainly occurred on the meta-position of 13 the phenyl ring with the electrophile of sulfonic acid. The phenyl remained bonded with the Si–O–Si 14 moiety of POSS, as confirmed by the intensity of the T3 peak in the spectrum. 15 FTIR spectra for OPOSS and SPOSS prepared using H2SO4 and ClSO3H are compared in Fig. 3b. 16 In the IR spectrum of OPOSS, a strong Si−O−Si peak at 1091 cm–1 (stretching vibrations) and an Si-17 phenyl peak in the range of 760–690 cm–1 (bending vibrations) were observed. Weak peaks for the Si-18 phenyl moiety appeared in the range of 1425–1590 cm–1 (bending vibrations), and aromatic C=C/C−H 19 bonds appeared in the range of 3000–3100 cm–1 (stretching vibrations) [49,66]. After the sulfonation 20 process, the peaks for the Si−phenyl and Si−O−Si moieties were broader due to the introduction of the 21 −SO3H group. The appearance of out-of-plane bending at 936 cm–1 indicates meta-substitution of the 22 aromatic ring [38,67]. The sharp peak at 3750 cm–1 for SPOSS was assigned to an isolated Si-OH 23 group on silica [66]. Some of the cage structures of the POSS have been damaged during the 24 sulfonation, producing incomplete Si−O−Si bonds, which is supported by the 29Si NMR spectra. A 25 broad OH peak appeared around 3400 cm–1, showing hydrogen bonding with the sulfonic group [38]. 26 These agree with the higher IEC of the SPOSS-H2SO4.  27 Fig. 3c shows normalized FTIR spectra in the range of 900–1300 cm–1, which were assigned to the 28 Si−O−Si moiety. A peak for an unsulfonated OPOSS was observed at 1091 cm–1. The peak for the 29 SPOSS-ClSO3H shifted from 1091 cm–1 to 1100 cm–1, and that for SPOSS-H2SO4 shifted to 1033 cm–30 1. These were assigned to the symmetric and asymmetric vibrations of the −SO3H groups [46,49]. 31 From the low T3 peak intensity for SPOSS-ClSO3H in the 29Si NMR spectra, phenyl rings were lost, 32 causing a low degree of sulfonation (1.9). The TGA plot showed slight weight loss from 150–350 °C 33 (i.e., SO3H decomposition). A structure for SPOSS-ClSO3H is shown in Fig. 3d, where several phenyl 34 groups are dissociated and replaced by the hydroxyl functional groups and only one or two sulfonated 35 phenyl groups. This structure is consistent with SPOSS-ClSO3H being hydrophilic. In contrast, 36 SPOSS-H2SO4 retains most of its phenyl; about three phenyls have been sulfonated. The SPOSS-37 H2SO4 is relatively hydrophobic. We will show later that the swelling of the CSPPSU composited 38 13  membrane using SPOSS-ClSO3H is much larger than that of SPOSS-H2SO4.  1 We used MALDI-TOF mass spectrometry and NMR spectroscopy to show that the SPOSS 2 structures upon sulfonation by either H2SO4 or ClSO3H were different. Sulfonation using H2SO4 is a 3 more controllable process at higher temperatures with minimal damage to the OPOSS structure than 4 that using ClSO3H. Although some of the cage structure became damaged, we minimized the phenyl 5 loss from the siloxane bond and increased the sulfonation yield of SPOSS-H2SO4 (Fig. 3d). 6  7 3.2 IEC, DS, water uptake and stress-strain of crosslinked SPPSU/SPOSS membrane 8 SPOSS prepared from either H2SO4 or ClSO3H were used to fabricate crosslinked SPPSU 9 membranes and their IEC, DS, WU, and stress-strain characteristics were evaluated. Fig. 4a and 4b 10  Fig. 3 a) 29Si NMR spectra with structure assignment on the distinguishing peaks and the schematic diagram of the electrophilic substitution process using H2SO4 and ClSO3H, (b) FTIR spectra of OPOSS (black), SPOSS-H2SO4 (blue) and -ClSO3H (red), c) Zoom in of normalized FTIR spectra at Si-O-Si peaks region, d) Proposed structure of SPOSS-H2SO4 and SPOSS-ClSO3H after the sulfonation. 14  compared their IECs before crosslinking (black dashed line) as estimated using eq. 6 to their IECs after 1 crosslinking (black solid line) as measured by titration. The degree of crosslinking was then calculated 2 from these values using eq. 7. After crosslinking, a maximum IEC of CSPPSU composite membrane 3 was obtained at 10 wt% and 5 wt% for SPOSS-H2SO4 and SPOSS-ClSO3H, respectively. Degree of 4 crosslinking decreased to less than 5% as SPOSS concentration increased up to 50 wt%. The smaller 5 degree of crosslinking as SPOSS content increased suggests the formation of aggregates within the 6 composite polymers causes sulfonic acid groups unavailable as crosslinking sites [45]. Composite 7 CSPPSU membranes from SPOSS-ClSO3H have a higher tendency to agglomerate more than from 8 SPOSS-H2SO4. 9  Fig. 4c shows the water uptake evaluation performed at room temperature and 100 °C for these 10 membranes. Composite membranes prepared using SPOSS-H2SO4 have low water uptake that does 11 not exceed 100% at room temperature and 300% at 100 °C for all concentration ranges. In contrast, 12 the addition of SPOSS-ClSO3H into the membrane resulted in a large water uptake. At 50 wt% SPOSS-13 ClSO3H, water uptake was 2500% even at room temperature. Composite membranes prepared with 10 14 wt% or higher loading of SPOSS-ClSO3H were found to disintegrate at 100 °C. The structure of 15 SPOSS prepared by ClSO3H is much more hydrophilic and weaker than those prepared by H2SO4. 16 Fig. 4d and 4e show the tensile stress and strain of the CSPPSU/SPOSS composite membranes. The 17 average stress and strain strength of the CSPPSU membranes were determined to be 29 MPa and 118%, 18 respectively. The stress strength of CSPPSU/SPOSS-H2SO4 increased with an increase in the loading 19 and decreased when it was larger than 5 wt%. At 5 wt%, the stress strength increased by 21% to 35 20 MPa with a strain strength of 107%. In contrast, the mechanical strength of CSPPSU/SPOSS-ClSO3H 21 decreased with an increase in the SPOSS-ClSO3H loading. Adding 1 wt% SPOSS-ClSO3H already 22 caused decreases in the tensile stress and strain by 31% and 15%, respectively. CSPPSU/SPOSS-23 ClSO3H is 10 MPa weaker than CSPPSU/SPOSS-H2SO4. The tensile modulus was calculated from 24 the stress-strain curve shown in Fig. 4f. The tensile modulus was improved by 40% when 1 wt% of 25 SPOSS-H2SO4 was incorporated into the CSPPSU polymer. In other words, the mechanical strength 26 of the CSPPSU composite membrane increases when SPOSS-H2SO4 is used, and it is stronger than 27 SPOSS-ClSO3H.  28 CSPPSU/SPOSS-H2SO4 composite membrane has improved water swelling and high strength. The 29 improvements in mechanical strength also strongly suggest crosslinking between the SPPSU polymer 30 matrix and SPOSS additive. At 0% SPOSS, uncrosslinked SPPSU has a stress strength of 3 MPa, 31 which increases to 24 MPa after crosslinking. With the addition of 1% SPOSS, the crosslinked 32 membrane has a higher stress strength of 32 MPa. Identifying SO2 crosslinks between SPPSU and 33 SPOSS remains challenging since the concentration of these crosslinks is too low to be directly 34 observed from FTIR or NMR [68]. The structure of SPOSS-H2SO4 was shown to be structurally intact 35 with the phenyl groups still available, and a much higher degree of sulfonation was achieved. We 36 believe that these factors are responsible for the significant improvements in the mechanical strength 37 of the composite membranes prepared with SPOSS-H2SO4 because the higher DS increased SPOSS 38 15  dispersibility within SPPSU polymer matrix, and more sulfonic acid group sites are available to 1 crosslink to SPPSU. The membrane characteristics of the CSPPSU/SPOSS-ClSO3H composite 2 membranes are opposite, i.e., they are weaker due to extreme swelling and low strength. These results 3 show the importance of the SPOSS structure as the nanocomposite in CSPPSU composite membranes. 4 Phenyl loss and attachment of hydroxyl functional groups in the SPOSS-ClSO3H disrupt the 5 sulfonation process and weaken the CSPPSU/SPOSS composite membrane. 6  7 3.3 Proton conductivity of crosslinked SPPSU/SPOSS membrane 8 Fig. 5a shows the proton conductivity of CSPPSU/SPOSS-H2SO4 at 80 °C against the SPOSS 9 loading from 0 wt% to 50 wt%. CSPPSU membrane had a proton conductivity of 52.8 mS/cm at 85% 10 RH. The conductivity of 1 wt% SPOSS-H2SO4 was found to be 46.7 mS/cm, and it gradually increased 11 with the SPOSS loading. Only small changes of conductivity for all membranes were observed in the 12 high humidity since the hydration numbers were all in the range of 50 – 60. Similar conductivities 13 have been reported for measurements under full humidity conditions [46]. On the other hand, at higher 14 temperature of 120 °C (Fig. 5b), the SPOSS loading significantly affected the conductivity. The 15 CSPPSU membrane had a conductivity of 7.8 mS/cm at 40% RH. The 1 wt% of SPOSS-H2SO4 16  Fig. 4 IEC of the CSPPSU composite membrane using (a) SPOSS-H2SO4, (b) SPOSS-ClSO3H (dashed line represents IEC before crosslinking (eq. 5)), (c) comparison of the WU of CSPPSU/SPOSS-H2SO4 and CSPPSU/SPOSS-ClSO3H at RT and 100 °C (inset: enlargement of the WU plot for CSPPSU/SPOSS-H2SO4), and comparisons of (d) the average stress, (e) average strain and (f) tensile modulus of CSPPSU/SPOSS-H2SO4 and CSPPSU/SPOSS-ClSO3H composite 16  composite membrane had the highest proton conductivity of 13.7 mS/cm, a 75% improvement over 1 that of CSPPSU. Incorporating 1 wt% SPOSS-H2SO4 improves the conductivity, making it comparable 2 to Nafion212 (13.2 mS/cm). When the SPOSS loading was higher than 1 wt%, the conductivities 3 decreased. 4 Fig. 5d and 5e show the conductivities of CSPPSU/SPOSS-ClSO3H at 80 and 120 °C against the 5 SPOSS loading from 0 wt% to 50 wt%. The conductivities increased with an increase in the SPOSS 6 loading. At 50 wt% SPOSS-ClSO3H loading, the highest conductivities of 58.49 mS/cm (85% RH) 7 and 10.8 mS/cm (40% RH) at 80 and 120 °C, respectively, were obtained.  8 The activation energy (Ea), which is the minimum energy needed for the proton transfer from one 9 free site to another. For the CSPPSU/SPOSS composite and Nafion 212 membranes were determined 10 from temperature-dependent Arrhenius plots at 80% RH in the temperature range of 40‒80 °C. The 11 lower Ea, the better. The linear fitting (R2 > 0.91) of the Arrhenius plot in Fig. 5c and 5f is equal to 12 −Ea/R (eq. 11). The CSPPSU membrane has a larger Ea (28.8 kJ/mol) than Nafion212 (19.3 kJ/mol), 13 which is caused by the smaller ionic clusters. Incorporating 1 wt% SPOSS-H2SO4 into SPPSU reduced 14 the Ea to 26.6 kJ/mol. Since similar Ea values were obtained at higher loading, a loading of 1 wt% is 15 sufficient to improve proton transfer. The composite membrane with 1 wt% SPOSS-ClSO3H had an 16 Ea of 22.4 kJ/mol, and it increased with an increase in the SPOSS loading to 28.8 kJ/mol at 50 wt%. 17 Although the CSPPSU/SPOSS-ClSO3H composite membranes have lower Ea values, they are too 18 weak to be used practically due to excessive water swelling and low mechanical strength. 19  Fig. 5 (a–c) Proton conductivities at 80 and 120 °C and activation energy of CSPPSU/SPOSS-H2SO4, (d–f) Proton conductivities at 80 and 120 °C and activation energy of the CSPPSU/SPOSS-ClSO3H in the T range of 40–80 °C at 80% RH. 17  At 120 °C, 1 wt% CSPPSU/SPOSS-H2SO4 was found to have the highest conductivity at 40% RH. 1 A high SPOSS loading causes the aggregation of the SPOSS nanoparticles. A large aggregation of 2 SPOSS initiates winding of and dead ends in the nano channels [69]. This causes the blocking of the 3 hydrophilic domains and increases the tortuosity of the ionic path, resulting in a decrease in the 4 conductivity [38,45-47]. Kugarajah et al. [47] have discussed that the aggregation pattern of the cluster 5 network of POSS is similar to sulfonated POSS. Strong hydrogen bonds between Si-O-Si and SO3H 6 groups cause the aggregation of sulfonic moieties in SPOSS. However, in this study, there was no 7 visible agglomeration observed on the membrane’s surface, even at 50 wt% of SPOSS loading, as 8 shown in Fig. S6a. SPOSS prepared using H2SO4 promotes the dispersion of the nanoparticles in the 9 polymer matrix as the 30 wt% SPOSS in SPPSU/SPOSS cast solution was clear and transparent (Fig. 10 S6b). Well-dispersed inorganic composites in the membrane are essential to enhance the synergistic 11 effects of microscopic composite-type membranes on the thermal, mechanical, and chemical stabilities, 12 which are not easy to achieve [35]. 13 The visible agglomeration on the surface of the composite membranes at specific loadings of 14 SPOSS has been reported when ClSO3H was used to prepare the SPOSS [38,45,47,70]. The SPOSSs 15 were not well dispersed and agglomerated with each other after the heat treatment process. In our 16 previous study on the CSPPSU/SPOSS composite membrane, agglomeration occurred even at 2 wt% 17 loading, as shown in Fig. S6c [50]. SPOSS was prepared using ClSO3H at 50 °C. Although a high IEC 18 of 7.86 was obtained, the dispersion of the SPOSS nanocomposite in the polymer matrix was poor.  19  20 3.4 Structural analysis and mechanical strength of crosslinked SPPSU/SPOSS membrane  21 The density of the sulfonic acids and the nanochannel for the ion mobility affect the conductivity 22 of polymer membranes. Although higher IEC values are obtained with increased SPOSS loadings, 23 poor mobility of the ion transfer inside the nanochannel can decrease the proton conductivity. The 24 microstructure and reorganization of CSPPSU/SPOSS-H2SO4 composite membranes were analyzed 25  Fig. 6 (a) SAXS pattern of the membranes with SPOSS loadings of 1 wt% and 5 wt%, (b) FTIR spectra, and (c) TGA curves of 1 wt% to 50 wt% CSPPSU/SPOSS-H2SO4 composite membranes. 18  by using small-angle-X-ray-scattering (SAXS) with loadings of 1 wt% and 5 wt%, as shown in Fig. 1 6a. Two peaks appeared in the SAXS spectra of a swollen CSPPSU/SPOSS-H2SO4 membrane. The 2 appearance of a shoulder peak between 0.3 nm–1 and 0.8 nm–1, known as the matrix knee, revealed the 3 inter-crystalline spacing between the crystalline domains of the polymer matrix (L = 20.9–7.9 nm). 4 The characteristic ionomer peak at 1 nm–1 < q < 2 nm–1 was attributed to a self-organized ionic cluster 5 network under hydrated conditions. The scattering intensity of the ionomer peak is directly 6 proportional to the difference in electron density of the polymer backbone and ionic clusters in SPOSS 7 and SPPSU [13,54,55,71,72]. The ionic cluster size was calculated using the q value obtained from the 8 plot made using eq. 4. Kim et al. [12] have reported that the CSPPSU membrane ionic cluster is 3.9 9 nm in size. Incorporating 1 wt% SPOSS-H2SO4 increased the size of the ionic cluster to 4.6 nm, and a 10 loading of 5 wt% increased the cluster size to 5.5 nm. The addition of the SPOSS-H2SO4 increased the 11 size of the ionic cluster, the IEC, and the hydration number of the CSPPSU/SPOSS-H2SO4 composite 12 membranes. Although higher loadings increase the ionic cluster size, the proton conductivity at 120 °C 13 decreased when SPOSS-H2SO4 loading > 1 wt%.  14 Kim et al. [45] have discussed the importance of controlling the aggregation in the nanochannel for 15 good proton conductivity. From SAXS analysis, an SPOSS amount higher than 1.5 wt% in SPEEK 16 composite membranes decreases the proton conductivity due to the aggregation of SPOSS in the 17 nanochannel. An SPOSS loading of 5 wt% crowds and expands the nanochannel, which lowers the 18 proton conductivity of the SPEEK/SPOSS composite membranes. Proton transfer in CSPPSU/SPOSS 19 membranes with high water uptake occurs via a vehicular mechanism [8,73]. In our case, 1 wt% of 20 SPOSS-H2SO4 is the optimal concentration in the CSPPSU/SPOSS membrane to improve the 21 membrane strength and proton conductivity. 22 FTIR spectra of the CSPPSU and CSPPSU/SPOSS-H2SO4 membranes with different SPOSS 23 loadings are shown in Fig. 6b. The presence of sulfonic acid groups was evidenced by the appearance 24 of the peaks at 3400 (O−H vibrations), 1017, and 1238 cm–1 (SO3− vibrations). The intensities of the 25 O-H peaks were significantly reduced after adding the SPOSS-H2SO4 nanoparticles into the CSPPSU 26 membrane. The addition of SPOSS-H2SO4 improved the membrane’s water swelling and lowered 27 SO3H peaks due to the crosslinking among the moieties [50]. The C−H absorption peaks at 874, 829, 28 and 686 cm–1 were attributed to the C−H bond from the SPPSU polymer. Absorption peaks due to the 29 C=C bonds in the benzene ring appeared at 1467 and 1581 cm–1. The peaks at 1147 and 1088 cm–1 30 were attributed to S=O bonds in the PPSU polymer backbone.  31 Fig. 6c shows the thermal decomposition of the CSPPSU/SPOSS-H2SO4 membranes by using TGA. 32 Three weight loss stages were observed with thermolysis of distinct groups. The first step was the 33 decomposition of water of H3O+ associated with SO3– to SO3H moieties at 56 °C. The second weight 34 loss was the partial decomposition of SO3H to SO2 and ∙OH at a maximum temperature of 267 °C, and 35 the third weight loss was due to the decomposition of Ar-SO2 in the PPSU backbone to Ar∙ and SO2∙ 36 at 516 °C [13,32,50,74]. The addition of SPOSS-H2SO4 increases the thermal stability of the CSPPSU 37 19  membranes from silicon in the POSS structures [47]. An SPOSS-H2SO4 loading of 1 wt% in CSPPSU 1 increased the T90% from 75 to 82 °C. 2 The Si content, which comes from the siloxane structure in the SPOSS, in the CSPPSU/SPOSS-3 H2SO4 composite membranes before and after activation was evaluated using X-ray fluorescence 4 (XRF), as shown in Table 3. Before the activation, the Si content in the membranes was 10,000 ppm 5 at an SPOSS-H2SO4 loading of 5 wt%. However, only 50 ppm of the Si remained in the membrane 6 after activation. The same Si content was observed when the SPOSS loading was 10 wt%. When the 7 SPOSS loading was 50 wt%, the Si content was only 100 ppm. This shows that the SiO2 leaches during 8 the NaOH activation process with SO2 as a by-product. A large amount of the SPOSS nanocomposite 9 is unnecessary since the amount of Si remains stagnant when SPOSS loading is higher than 5 wt%. In 10 other words, material waste can be prevented since the SPOSS nanocomposite content does not affect 11 the composite polymer much. This shows that only a small amount of the nanocomposite is 12 incorporated into the polymer structure of the composite membranes. The amount of the 13 nanocomposite should be in the order of parts per million (ppm). However, more studies on the range 14 that might substantially improve the polymer membrane are needed. 15 Numerous studies have been reported on sulfonated POSS, which is used as a nanocomposite in 16 hydrocarbon polymers. Table 4 compares the conductivity, swelling, and mechanical strength of the 17 SPOSS composite polymer membranes in this study with previously reported membranes. In our 18 research, we found that 1 wt% of the SPOSS composite was sufficient for the composite CSPPSU 19 membranes. In comparison with our previous study, using sulfuric acid in the sulfonation process of 20 the OPOSS is more controllable and causes less damage to the OPOSS structure. This significantly 21 affects the distribution of the nanoparticles in the polymer matrix, and no visible agglomeration was 22 observed on the membrane, even at a loading of 50 wt%. Agglomeration of SPOSS is a big problem 23 since the nanoparticle distribution in the polymer plays a vital role in the membrane performance and 24 strength. While conductivity results at 80 °C and >80% RH were shown to be much lower than other 25 membranes in literature, not many studies measured conductivities at high-temperature and low-26 humidity conditions. At 120 °C and 40% RH, our CSPPSU/1 wt% SPOSS-H2SO4 has a very 27 comparable conductivity of 13.7 mS/cm to Nafion212 (13.2 mS/cm) and other SPOSS composites 28 aromatic hydrocarbon membranes. The oxidative stability of the CSPPSU membrane against Fenton’s 29 reagent improved from 2 to 4 h with the addition of SPOSS, suggesting future research on the CSPPSU 30 composite membrane. 31 Table 3 XRF measurement of the amount of Si in the CSPPSU/SPOSS composite membranes before and after activation process. SPOSS (wt%) Si content (ppm) 5 (No activation) 10,000 5 50 10 50 50 100 20   1 3.5 Single cell performance of crosslinked SPPSU/SPOSS membrane 2 The evaluation results for single cell performance of CSPPSU membranes with and without 1 wt% 3 SPOSS were compared to the Nafion211 membrane. Each membranes have a thickness of 36, 31, and 4 25 µm. The electrochemical performances of the Pt/C/ionomer catalyst electrode were assessed by 5 plotting the polarization curve and measuring the electrochemically active surface area (ECSA). Fig. 6 7a shows the I-V characteristic of a single cell at 80 °C and 100% RH, where ViR-free and iR-loss were 7 plotted as the y-axes. Open circuit voltage (OCV) for all MEAs was in the range of 0.9 to 1.0 V, 8 representing good assembly of PEM and the electrodes [75]. The current densities at 0.8 ViR-free were 9 66.6 mA/cm2, 118.3 mA/cm2, and 247.7 mA/cm2 for CSPPSU/1wt%SPOSS-H2SO4, CSPPSU, and 10 Table 4 IEC, conductivity, water uptake, stress-strain and oxidative stability data from the previous studies on the SPOSS composite membranes.  Sample IEC (meg/g) σ1 (mS/cm) σ2 (mS/cm) WU (%) RT, 100 °C SS (MPa, %) Roxidation (%) Ref.  CSPPSU 2.9 52.8 7.8 135, 293 29, 118 92 (2h) a  CSPPSU/ 1 wt%SPOSS 3.0 46.7 13.7 120, 275 32, 98 100 (4h) a  CSPPSU/ 5 wt%SPOSS 3.0 38.9 7.3 98, 245 35, 107 94 (4h) a  Nafion212 0.9 33.0a 13.2a 50, - 16, >150 95 (48h)76 12 2023 SPFEK/ 0.03 mg/cm2 SPOSS - 110 - 35.3, - 47 95 (0.7h) 77 2021 Nafion/  12 wt%SPOSS - 123 - 14, - 25, 100 96 (36h) 78 2020 CSPPSU/ 1 wt%SPOSS 1.9 37.9 6.6 128, - 45, 66 - 50 2020 CSPPSU/ 5 wt%SPOSS 1.3 9.4 0.6 33, - 52, 36 - 50 2020 SPAES/ 2 wt%SPOSS 2.1 163 - 48, 82 28, 8 <100 (2h) 75 2020 SPEEK/ 5 wt% SPOSS 1.8 13.1b - 36, - - - 47 2018 SPEEK/ 1.5 wt%SPOSS 1.5 97 - 31, - 37, 70 - 45 2017 PAES/ 14 wt%SPOSS 1.4 142 - 21, 59 14, 45 95 (1h) 79 2014 Nafion/ 2 wt% SPOSS 1.45 24 - 55, - - - 67 2013 SPEEK/ 2 wt%SPOSS - - 0.036c 20.5, 80.3 - - 70 2010 CSPPSU/ 20 wt%SPOSS 1.8 71b - - - - 46 2008 CSPPSU/ 20 wt%SPOSS - 66 - 76 25, 4 - 44 2008 CSPPSU/ 10 wt%SPOSS - 68 - 59 39, 9 - 44 𝜎  at 70‒80 °C (85–100% RH); 𝜎  at 120 °C (40% RH); WU is water uptake; SS is stress-strain; Roxidation is oxidation stability; aOur data; bat room temperature; cat 100 °C 21  Nafion211, respectively. The voltage losses at 0.5 A/cm2 were 143.6 mΩ, 102.2 mΩ, and 67.6 mΩ for 1 CSPPSU/1wt% SPOSS-H2SO4, CSPPSU, and Nafion211, respectively. ECSA determined from CV 2 measurement (Fig. 7b) were 65 m2/g, 68 m2/gPt, and 92 m2/gPt for CSPPSU/1wt%SPOSS-H2SO4, 3 CSPPSU, and Nafion211, respectively. Although the conductivity of the CSPPSU and 4 CSPPSU/1wt%SPOSS-H2SO4 membranes were higher than that of Nafion211 (Table 4), the fuel cell 5 performance of the MEAs using CSPPSU and CSPPSU/1wt% SPOSS membranes was lower than that 6 of Nafion211 due to the high iR loss (Fig. 7a). Furthermore, CV results show that the catalyst reaction 7 at the anode and cathode electrodes of the MEA using CSPPSU/1wt% SPOSS membrane is lower than 8 those of CSPPSU and Nafion211 membranes (Fig. 7b). The hydrogen crossover observed from the 9 linear sweep voltammetry (LSV) are shown in Fig. 7c. Membranes of CSPPSU and 10 CSPPSU/1wt%SPOSS-H2SO4 have a low current density of 0.3 mA/cm2 at 0.4 V. However, Nafion211 11 has a higher current density at 1.8 mA/cm2 at 0.4 V. CSPPSU, and CSPPSU/1wt%SPOSS-H2SO4 have 12 a higher resistance to the hydrogen crossover in the fuel cell system. The results show a need for a 13 high-performance fuel cell system using CSPPSU/1wt% SPOSS membrane to improve the interface 14 resistance and catalyst reaction in the MEA. 15  16 4. Conclusions 17 In summary, direct sulfonation of OPOSS using H2SO4 was more controllable and could be 18 performed under milder conditions to prevent damage to the SPOSS composite structure. Well-19 dispersed 348 nm-hydrophobic-SPOSS-H2SO4 with an IEC of 2.4 meq/g and DS of 3.1 was prepared. 20 Structural analysis using 29Si NMR and FTIR spectroscopies, MALDI-TOF mass spectrometry, and 21 TGA confirmed that the phenyl remained intact in the T8-cage structure of OPOSS. In addition, 22 incorporating 1 wt% SPOSS-H2SO4 into SPPSU improved the stability and strength of the crosslinked 23 SPPSU/SPOSS composite membranes. The water uptake was reduced by 11%, and the tensile modulus 24 was enhanced by 40% over the CSPPSU membrane. The proton conductivity at 120 °C and 40% RH 25 was 13.7 mS/cm, which is similar to Nafion211 (13.2 mS/cm). From SAXS analysis, the ionic cluster 26  Fig. 7 (a) Polarization (b) cyclic voltammetry (CV), and (c) linear sweep voltammetry (LSV) of hydrogen crossover current curves of CSPPSU, CSPPSU/1 wt% SPOSS and Nafion211 at 80 ºC, 100% RH. 22  expanded proportionally with the SPOSS loading. However, SPOSS loadings higher than 1 wt% crowd 1 the expanded ionic cluster, resulting in decreased conductivities. The improvements in the 2 CSPPSU/SPOSS-H2SO4 composite membrane show the importance of the SPOSS structure as the 3 nanocomposite. The sulfonation process using different acids is promising for producing good 4 dispersions and intact SPOSS structure, which is a priority for PEM nanofillers.   5  6 CRediT authorship contribution statement 7 F.B. Fauzi: Methodology, Investigation, Formal analysis, Writing - Original Draft. J.-D. Kim: 8 Conceptualization, Supervision, Funding acquisition, Methodology, Investigation, Formal analysis, 9 Writing - Review & Editing. 10  11 Acknowledgement 12 This research was partly supported by the New Energy and Industrial Technology Development 13 Organization (NEDO) through the “Collaborative Industry-Academia-Government R&D Project for 14 Solving Common Challenges Toward Dramatically Expanded Use of Fuel Cells and Related 15 Equipment”. Special thanks to Sachiko Takao for her assistance in the stress-strain measurements. 16 Small-angle X-ray scattering (SAXS) measurements were performed at the Photon Factory (beamline 17 BL6A) of KEK (proposal no. 2021G517). J.-D. Kim thanks Akihiro Ohira of the National Institute of 18 Advanced Industrial Science and Technology (AIST) for their help with the SAXS measurements. 19  20 References 21  22 [1] H. Ritchie, M. Roser, P. Rosado, CO₂ and Greenhouse Gas Emissions. 23 https://ourworldindata.org/co2-and-greenhouse-gas-emissions, 2023 (accessed 1 July 2023). 24 [2] R. Vinodh, R. Atchudan, H.-J. Kim, M. Yi, Recent Advancements in Polysulfone Based 25 Membranes for Fuel Cell (PEMFCs, DMFCs and AMFCs) Applications: A Critical Review, 26 Polymers. 14 (2022) 300. 27 [3] Z.P. Cano, D. Banham, S. Ye, A. Hintennach, J. Lu, M. Fowler, Z. Chen, Batteries and fuel 28 cells for emerging electric vehicle markets, Nat. Energy. 3 (2018) 279–289. 29 [4] Y. Wang, K.S. Chen, J. Mishler, S.C. Cho, X.C. Adroher, A review of polymer electrolyte 30 membrane fuel cells: Technology, applications, and needs on fundamental research, Appl. 31 Energy. 88 (2011) 981–1007. 32 [5] B.G. Pollet, S.S. Kocha, I. Staffell, Current status of automotive fuel cells for sustainable 33 transport, Curr. Opin. Electrochem. 16 (2019) 90–95. 34 [6] M.A. Hickner, H. Ghassemi, Y.S. Kim, B.R. Einsla, J.E. McGrath, Alternative Polymer 35 Systems for Proton Exchange Membranes (PEMs), Chem. Rev. 104 (2004) 4587–4612.  36 [7] Y. Wang, D.F. Ruiz Diaz, K.S. Chen, Z. Wang, X.C. Adroher, Materials, technological status, 37 and fundamentals of PEM fuel cells – A review, Mater. Today. 32 (2020) 178–203. 38 23  [8] K. Firouz Tadavani, A. Abdolmaleki, M.R. Molavian, M. Zhiani, A Promising Proton-1 Exchange Membrane: High Efficiency in Low Humidity, ACS Appl. Energy Mater. 1 (2018) 2 2464–2473. 3 [9] W. Jia, B. Tang, P. Wu, Novel Composite Proton Exchange Membrane with Connected Long-4 Range Ionic Nanochannels Constructed via Exfoliated Nafion–Boron Nitride Nanocomposite, 5 ACS Appl. Mater. Interfaces. 9 (2017) 14791–14800. 6 [10] H. Nguyen, F. Lombeck, C. Schwarz, P.A. Heizmann, M. Adamski, H.-F. Lee, B. Britton, S. 7 Holdcroft, S. Vierrath, M. Breitwieser, Hydrocarbon-based PemionTM proton exchange 8 membrane fuel cells with state-of-the-art performance, Sustain. Energy Fuels. 5 (2021) 3687–9 3699. 10 [11] X. Ge, F. Zhang, L. Wu, Z. Yang, T. Xu, Current Challenges and Perspectives of Polymer 11 Electrolyte Membranes, Macromolecules. 55 (2022) 3773–3787. 12 [12] J.-D. Kim, A. Ohira, H. Nakao, Chemically Crosslinked Sulfonated Polyphenylsulfone 13 (CSPPSU) Membranes for PEM Fuel Cells, Membranes. 10 (2020) 31. 14 [13] A. Kusoglu, A.Z. Weber, New Insights into Perfluorinated Sulfonic-Acid Ionomers, Chem. Rev. 15 117 (2017) 987–1104. 16 [14] M. Feng, R. Qu, Z. Wei, L. Wang, P. Sun, Z. Wang, Characterization of the thermolysis products 17 of Nafion membrane: A potential source of perfluorinated compounds in the environment, Sci. 18 Rep. 5 (2015) 9859. 19 [15] N. Asano, M. Aoki, S. Suzuki, K. Miyatake, H. Uchida, M. Watanabe, Aliphatic/Aromatic 20 Polyimide Ionomers as a Proton Conductive Membrane for Fuel Cell Applications, J. Am. 21 Chem. Soc. 128 (2006) 1762–1769. 22 [16] Q. Yuan, P. Liu, G.L. Baker, Sulfonated polyimide and PVDF based blend proton exchange 23 membranes for fuel cell applications, J. Mater. Chem. A. 3 (2015) 3847–3853. 24 [17] C.H. Lee, C.H. Park, Y.M. Lee, Sulfonated polyimide membranes grafted with sulfoalkylated 25 side chains for proton exchange membrane fuel cell (PEMFC) applications, J. Memb. Sci. 313 26 (2008) 199–206. 27 [18] R.P. Mariappan, C. Liu, G. Cao, R.P. Manimuthu, Tailoring SPEEK/SPVdF-co-HFP/La2Zr2O7 28 Ternary Composite Membrane for Cation Exchange Membrane Fuel Cells, Ind. Eng. Chem. 29 Res. 59 (2020) 4881–4894. 30 [19] A.R. Kim, M.B. Poudel, J.Y. Chu, M. Vinothkannan, R. Santhosh Kumar, N. Logeshwaran, B.-31 H. Park, M.-K. Han, D.J. Yoo, Advanced performance and ultra-high, long-term durability of 32 acid-base blended membranes over 900 hours containing sulfonated PEEK and quaternized 33 poly(arylene ether sulfone) in H2/O2 fuel cells, Compos. Part B Eng. 254 (2023) 110558. 34 [20] P. Xing, G.P. Robertson, M.D. Guiver, S.D. Mikhailenko, K. Wang, S. Kaliaguine, Synthesis 35 and characterization of sulfonated poly(ether ether ketone) for proton exchange membranes, J. 36 Memb. Sci. 229 (2004) 95–106. 37 [21] R. Hara, N. Endo, M. Higa, K. Okamoto, X. Zhang, H. Bi, S. Chen, Z. Hu, S. Chen, Polymer 38 24  electrolyte fuel cell performance of poly(arylene ether ketone)-graft-crosslinked-1 poly(sulfonated arylene ether sulfone), J. Power Sources. 247 (2014) 932–938. 2 [22] J. Kim, S. Hwang, Y.G. Jeong, Y.S. Choi, K. Kim, Cross-Linked Sulfonated Poly(arylene ether 3 sulfone) Membrane Using Polymeric Cross-Linkers for Polymer Electrolyte Membrane Fuel 4 Cell Applications, Membranes. 13 (2023) 7. 5 [23] M.L. Di Vona, G. Alberti, E. Sgreccia, M. Casciola, P. Knauth, High performance sulfonated 6 aromatic ionomers by solvothermal macromolecular synthesis, Int. J. Hydrogen Energy. 37 7 (2012) 8672–8680. 8 [24] W. Mabrouk, L. Ogier, S. Vidal, C. Sollogoub, F. Matoussi, M. Dachraoui, J.F. Fauvarque, 9 Synthesis and Characterization of Polymer Blends of Sulfonated Polyethersulfone and 10 Sulfonated Polyethersulfone Octylsulfonamide for PEMFC Applications, Fuel Cells. 12 (2012) 11 179–187. 12 [25] C. Simari, M. Prejanò, E. Lufrano, E. Sicilia, I. Nicotera, Exploring the Structure–Performance 13 Relationship of Sulfonated Polysulfone Proton Exchange Membrane by a Combined 14 Computational and Experimental Approach, Polymers. 13 (2021) 959. 15 [26] C. Simari, E. Lufrano, A. Brunetti, G. Barbieri, I. Nicotera, Polysulfone and organo-modified 16 graphene oxide for new hybrid proton exchange membranes: A green alternative for high-17 efficiency PEMFCs, Electrochim. Acta. 380 (2021) 138214. 18 [27] D.W. Shin, M.D. Guiver, Y.M. Lee, Hydrocarbon-Based Polymer Electrolyte Membranes: 19 Importance of Morphology on Ion Transport and Membrane Stability, Chem. Rev. 117 (2017) 20 4759–4805. 21 [28] K. Jiao, J. Xuan, Q. Du, Z. Bao, B. Xie, B. Wang, Y. Zhao, L. Fan, H. Wang, Z. Hou, S. Huo, 22 N.P. Brandon, Y. Yin, M.D. Guiver, Designing the next generation of proton-exchange 23 membrane fuel cells, Nature. 595 (2021) 361–369. 24 [29] S. Swaby, N. Ureña, M. Teresa Pérez-Prior, C. del Río, A. Várez, J.-Y. Sanchez, C. Iojoiu, B. 25 Levenfeld, Proton conducting sulfonated polysulfone and polyphenylsulfone multiblock 26 copolymers with improved performances for fuel cell applications, J. Ind. Eng. Chem. 122 27 (2023) 366–377. 28 [30] Y. Zhang, J. Kim, K. Miyatake, Effect of thermal crosslinking on the properties of sulfonated 29 poly(phenylene sulfone)s as proton conductive membranes, J. Appl. Polym. Sci. 133 (2016). 30 [31] J.-D. Kim, L.-J. Ghil, Annealing effect of highly sulfonated polyphenylsulfone polymer, Int. J. 31 Hydrogen Energy. 41 (2016) 11794–11800. 32 [32] S. Matsushita, J.-D. Kim, Organic solvent-free preparation of electrolyte membranes with high 33 proton conductivity using aromatic hydrocarbon polymers and small cross-linker molecules, 34 Solid State Ionics. 316 (2018) 102–109. 35 [33] J. Kim, A. Donnadio, M. Jun, M.L. Di Vona, Crosslinked SPES-SPPSU membranes for high 36 temperature PEMFCs, Int. J. Hydrogen Energy. 38 (2013) 1517–1523. 37 [34] J. Kim, A. Ohira, Crosslinked Sulfonated Polyphenylsulfone (CSPPSU) Membranes for 38 25  Elevated-Temperature PEM Water Electrolysis, Membranes. 11 (2021) 861. 1 [35] C.H. Park, C.H. Lee, M.D. Guiver, Y.M. Lee, Sulfonated hydrocarbon membranes for medium-2 temperature and low-humidity proton exchange membrane fuel cells (PEMFCs), Prog. Polym. 3 Sci. 36 (2011) 1443–1498. 4 [36] C.Y. Wong, W.Y. Wong, K. Ramya, M. Khalid, K.S. Loh, W.R.W. Daud, K.L. Lim, R. 5 Walvekar, A.A.H. Kadhum, Additives in proton exchange membranes for low- and high-6 temperature fuel cell applications: A review, Int. J. Hydrogen Energy. 44 (2019) 6116–6135. 7 [37] T. Xu, Ion exchange membranes: State of their development and perspective, J. Memb. Sci. 263 8 (2005) 1–29. 9 [38] S. Subianto, M.K. Mistry, N.R. Choudhury, N.K. Dutta, R. Knott, Composite Polymer 10 Electrolyte Containing Ionic Liquid and Functionalized Polyhedral Oligomeric Silsesquioxanes 11 for Anhydrous PEM Applications, ACS Appl. Mater. Interfaces. 1 (2009) 1173–1182. 12 [39] H. Ranganathan, M. Vinothkannan, A.R. Kim, V. Subramanian, M. Oh, D.J. Yoo, Simultaneous 13 improvement of power density and durability of sulfonated poly(ether ether ketone) membrane 14 by embedding CeO2-ATiO2 : A comprehensive study in low humidity proton exchange 15 membrane fuel cells, Int. J. Energy Res. 46 (2022) 9041–9057. 16 [40] M. Vinothkannan, A.R. Kim, S.K. Ryu, D.J. Yoo, Structurally modulated and functionalized 17 carbon nanotubes as potential filler for Nafion matrix toward improved power output and 18 durability in proton exchange membrane fuel cells operating at reduced relative humidity, J. 19 Memb. Sci. 649 (2022) 120393.  20 [41] X. He, G. He, A. Zhao, F. Wang, X. Mao, Y. Yin, L. Cao, B. Zhang, H. Wu, Z. Jiang, Facilitating 21 Proton Transport in Nafion-Based Membranes at Low Humidity by Incorporating 22 Multifunctional Graphene Oxide Nanosheets, ACS Appl. Mater. Interfaces. 9 (2017) 27676–23 27687. 24 [42] A.R. Kim, M. Vinothkannan, S. Ramakrishnan, B.-H. Park, M.-K. Han, D.J. Yoo, Enhanced 25 electrochemical performance and long-term durability of composite membranes through a 26 binary interface with sulfonated unzipped graphite nanofibers for polymer electrolyte fuel cells 27 operating under low relative humidity, Appl. Surf. Sci. 593 (2022) 153407. 28 [43] L. Liu, Y. Lu, Y. Pu, N. Li, Z. Hu, S. Chen, Highly sulfonated carbon nano-onions as an 29 excellent nanofiller for the fabrication of composite proton exchange membranes with enhanced 30 water retention and durability, J. Memb. Sci. 640 (2021) 119823. 31 [44] C. Hartmann‐Thompson, A. Merrington, P.I. Carver, D.L. Keeley, J.L. Rousseau, D. Hucul, K.J. 32 Bruza, L.S. Thomas, S.E. Keinath, R.M. Nowak, D.M. Katona, P.R. Santurri, Proton‐33 conducting polyhedral oligosilsesquioxane nanoadditives for sulfonated polyphenylsulfone 34 hydrogen fuel cell proton exchange membranes, J. Appl. Polym. Sci. 110 (2008) 958–974. 35 [45] S.-W. Kim, S.-Y. Choi, H.-W. Rhee, A novel sPEEK nanocomposite membrane with well-36 controlled sPOSS aggregation in tunable nanochannels for fast proton conduction, Nanoscale. 37 10 (2018) 18217–18227. 38 26  [46] B. Decker, C. Hartmann-Thompson, P.I. Carver, S.E. Keinath, P.R. Santurri, Multilayer 1 Sulfonated Polyhedral Oligosilsesquioxane (S-POSS)-Sulfonated Polyphenylsulfone (S-PPSU) 2 Composite Proton Exchange Membranes, Chem. Mater. 22 (2010) 942–948. 3 [47] V. Kugarajah, S. Dharmalingam, Sulphonated polyhedral oligomeric 4 silsesquioxane/sulphonated poly ether ether ketone nanocomposite membranes for microbial 5 fuel cell: Insights to the miniatures involved, Chemosphere. 260 (2020) 127593. 6 [48] E. Carlier, A. Revillon, A. Guyot, P. Baumgartner, Functional silica supported polymers IV. 7 Synthesis and catalytic activity of silica-grafted sulfonated polyphenylsilsesquioxane, React. 8 Polym. 21 (1993) 15–25. 9 [49] V. Söderholm, J. Esteban, D. Vogt, Synthesis of a H-Sulfo-POSS catalyst and application in the 10 acetalization of glycerol with 2-butanone to yield a biofuel additive, Catal. Sci. Technol. 11 11 (2021) 4529–4538. 12 [50] N.A.M. Nor, J. Jaafar, J.-D. Kim, Improved properties of sulfonated octaphenyl polyhedral 13 silsequioxane cross-link with highly sulfonated polyphenylsulfone as proton exchange 14 membrane, J. Solid State Electrochem. 24 (2020) 1185–1195. 15 [51] N. Gospodinova, E. Tomšík, O. Omelchenko, How strong are strong poly(sulfonic acids)? An 16 example of the poly(2-acrylamido-2-methyl-1-propanesulfonic acid), Eur. Polym. J. 74 (2016) 17 130–135. 18 [52] A. Noshay, L.M. Robeson, Sulfonated polysulfone, J. Appl. Polym. Sci. 20 (1976) 1885–1903. 19 [53] A. Dyck, D. Fritsch, S.P. Nunes, Proton‐conductive membranes of sulfonated 20 polyphenylsulfone, J. Appl. Polym. Sci. 86 (2002) 2820–2827. 21 [54] Z. Jiang, X. Zhao, Y. Fu, A. Manthiram, Composite membranes based on sulfonated poly(ether 22 ether ketone) and SDBS-adsorbed graphene oxide for direct methanol fuel cells, J. Mater. Chem. 23 22 (2012) 24862. 24 [55] B.G. Choi, J. Hong, Y.C. Park, D.H. Jung, W.H. Hong, P.T. Hammond, H. Park, Innovative 25 Polymer Nanocomposite Electrolytes: Nanoscale Manipulation of Ion Channels by 26 Functionalized Graphenes, ACS Nano. 5 (2011) 5167–5174. 27 [56] R.A. Mantz, P.F. Jones, K.P. Chaffee, J.D. Lichtenhan, J.W. Gilman, Thermolysis of Polyhedral 28 Oligomeric Silsesquioxane ( POSS ) Macromers and POSS - Siloxane Copolymers, 4756 (1996) 29 1250–1259. 30 [57] A. Fina, D. Tabuani, F. Carniato, A. Frache, E. Boccaleri, G. Camino, Polyhedral oligomeric 31 silsesquioxanes (POSS) thermal degradation, Thermochim. Acta. 440 (2006) 36–42. 32 [58] H. Fan, R. Yang, Thermal decomposition of polyhedral oligomeric octaphenyl, 33 octa(nitrophenyl), and octa(aminophenyl) silsesquioxanes, J. Therm. Anal. Calorim. 116 (2014) 34 349–357. 35 [59] H. Chabane, S. Livi, H. Benes, C. Ladavière, P. Ecorchard, J. Duchet-Rumeau, J. Gérard, 36 Polyhedral oligomeric silsesquioxane-supported ionic liquid for designing nanostructured 37 hybrid organic-inorganic networks, Eur. Polym. J. 114 (2019) 332–337. 38 27  [60] M.J. Martínez-Morlanes, A.M. Martos, A. Várez, B. Levenfeld, Synthesis and characterization 1 of novel hybrid polysulfone/silica membranes doped with phosphomolybdic acid for fuel cell 2 applications, J. Memb. Sci. 492 (2015) 371–379. 3 [61] L. Li, X. Li, R. Yang, Mechanical, thermal properties, and flame retardancy of PC/ultrafine 4 octaphenyl‐POSS composites, J. Appl. Polym. Sci. 124 (2012) 3807–3814. 5 [62] M. Cypryk, Y. Apeloig, Mechanism of the Acid-Catalyzed Si−O Bond Cleavage in Siloxanes 6 and Siloxanols. A Theoretical Study, Organometallics. 21 (2002) 2165–2175. 7 [63] E.S. Park, H.W. Ro, C. V. Nguyen, R.L. Jaffe, D.Y. Yoon, Infrared Spectroscopy Study of 8 Microstructures of Poly(silsesquioxane)s, Chem. Mater. 20 (2008) 1548–1554. 9 [64] S. Mädler, K. Barylyuk, E. Boeri Erba, R.J. Nieckarz, R. Zenobi, Compelling Advantages of 10 Negative Ion Mode Detection in High-Mass MALDI-MS for Homomeric Protein Complexes, 11 J. Am. Soc. Mass Spectrom. 23 (2012) 213–224. 12 [65] T. Nakai, Solid-state 29Si NMR for Evaluating Microscopic Sites of Glass Materials: 29Si 13 Measurements in Multinuclear NMR, Most Efficient Measurements, Quantitative 14 Measurements, Advanced Measurements, New Glas. 28 (2013) 17–28. 15 [66] P.J. Launer, B. Arkles, Infrared analysis of organosilicon compounds: spectra-structure 16 correlations, in: Silicon Compd. Silanes Silicones, Gelecst Inc, Morrisville, PA, 2008: pp. 223–17 226. 18 [67] C. del Río, E. Morales, P.G. Escribano, Nafion/sPOSS hybrid membranes for PEMFC. Single 19 cell performance and electrochemical characterization at different humidity conditions, Int. J. 20 Hydrogen Energy. 39 (2014) 5326–5337. 21 [68] D. Joseph, N.N. Krishnan, D. Henkensmeier, J.H. Jang, S.H. Choi, H.-J. Kim, J. Han, S.W. 22 Nam, Thermal crosslinking of PBI/sulfonated polysulfone based blend membranes, J. Mater. 23 Chem. A. 5 (2017) 409–417. 24 [69] W. Wu, Y. Li, J. Liu, J. Wang, Y. He, K. Davey, S. Qiao, Molecular‐Level Hybridization of 25 Nafion with Quantum Dots for Highly Enhanced Proton Conduction, Adv. Mater. 30 (2018) 26 1707516. 27 [70] D. Gupta, A. Madhukar, V. Choudhary, Effect of functionality of polyhedral oligomeric 28 silsesquioxane [POSS] on the properties of sulfonated poly(ether ether ketone) [SPEEK] based 29 hybrid nanocomposite proton exchange membranes for fuel cell applications, Int. J. Hydrogen 30 Energy. 38 (2013) 12817–12829. 31 [71] K. Schmidt-Rohr, Q. Chen, Parallel cylindrical water nanochannels in Nafion fuel-cell 32 membranes, Nat. Mater. 7 (2008) 75–83. 33 [72] M.-M. Schiavone, D.H. Lamparelli, Y. Zhao, F. Zhu, Z. Revay, A. Radulescu, The Effects of 34 Temperature and Humidity on the Microstructure of Sulfonated Syndiotactic–polystyrene Ionic 35 Membranes, Membranes. 10 (2020) 187. 36 [73] F. Beyraghi, S.H. Mirfarsi, S. Rowshanzamir, A. Karimi, M.J. Parnian, Optimal thermal 37 treatment conditions for durability improvement of highly sulfonated poly(ether ether ketone) 38 28  membrane for polymer electrolyte fuel cell applications, Int. J. Hydrogen Energy. 45 (2020) 1 13441–13458. 2 [74] N. Lebert, N. Gay, E. Richaud, Mechanical and macromolecular changes during thermal 3 degradation of polyphenylsulfone, Polym. Degrad. Stab. 195 (2022) 109812. 4 [75] P. Zhang, W. Li, L. Wang, C. Gong, J. Ding, C. Huang, X. Zhang, S. Zhang, L. Wang, W. Bu, 5 Polydopamine-modified sulfonated polyhedral oligomeric silsesquioxane: An appealing 6 nanofiller to address the trade-off between conductivity and stabilities for proton exchange 7 membrane, J. Memb. Sci. 596 (2020) 117734. 8 [76] E. Lufrano, I. Nicotera, A. Enotiadis, M.H. Ur Rehman, C. Simari, Elucidating the Water and 9 Methanol Dynamics in Sulfonated Polyether Ether Ketone Nanocomposite Membranes Bearing 10 Layered Double Hydroxides, Membranes. 12 (2022) 419. 11 [77] Z. Li, X. Hao, M. Xiao, S. Huang, D. Han, S. Wang, Y. Meng, Surface-Densified Non-12 Fluorinated Proton Exchange Membrane Used for Direct Methanol Fuel Cell, J. Electrochem. 13 Soc. 170 (2023) 064502. 14 [78] Z. Li, X. Hao, G. Cheng, S. Huang, D. Han, M. Xiao, S. Wang, Y. Meng, In situ implantation 15 of cross-linked functional POSS blocks in Nafion® for high performance direct methanol fuel 16 cells, J. Memb. Sci. 640 (2021). 17 [79] Z. Wu, Y. Tang, D. Sun, S. Zhang, Y. Xu, H. Wei, C. Gong, Multi-sulfonated polyhedral 18 oligosilsesquioxane (POSS) grafted poly(arylene ether sulfone)s for proton conductive 19 membranes, Polymer. 123 (2017) 21–29. 20  21