# Fileset

[JSSEC_Nor2020_Article_CSPPSU-SPOSS.pdf](https://mdr.nims.go.jp/filesets/675b082e-421b-4ea0-b705-50211c7e08e2/download)

## Creator

Nor Azureen Mohamad Nor, Juhana Jaafar, [Je-Deok Kim](https://orcid.org/0000-0003-4301-1044)

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[IMPROVED PROPERTIES OF SULFONATED OCTAPHENYL POLYHEDDRAL SILSEQUIOXANE CROSSLINK WITH HIGHLY SULFONATED POLYPHENYLSULFONE AS PROTON EXCHANGE MEMBRANE](https://mdr.nims.go.jp/datasets/aaf54676-194a-4bdd-9693-b899cb155c48)

## Fulltext

Improved properties of sulfonated octaphenyl polyhedral silsequioxane cross-link with highly sulfonated polyphenylsulfone as proton exchange membraneORIGINAL PAPERImproved properties of sulfonated octaphenyl polyhedralsilsequioxane cross-link with highly sulfonated polyphenylsulfoneas proton exchange membraneNor Azureen Mohamad Nor1,2,3 & Juhana Jaafar3 & Je-Deok Kim1,2Received: 3 March 2020 /Revised: 3 April 2020 /Accepted: 5 April 2020# Springer-Verlag GmbH Germany, part of Springer Nature 2020AbstractThis study explored the concept of improving the properties of the cross-linkedmembrane using a 1.5-nm closed cage octaphenylpolyhedral silsesquioxane (POSS) form of nanosilica carrying the sulfonic acid group. POSS functioned with SO3H groups(SPOSS) at 0, 1, 2, and 5 wt% were cross-linked with water-soluble sulfonated polyphenylsulfone (SPPSU) polymer. The cross-linking between SPPSU and SPOSS was accomplished through the interchain condensation of sulfonic acid functionalitiesinitiated by thermal curing treatment. In this study, a covalently cross-linked membrane was obtained under stepwise thermalcuring from 80 to 180 °C. Upon curing at 180 °C, the SPPSU-SPOSS showed considerable improvement on the membraneproton conductivity under low and high RH (%) conditions compared with the pristine SPPSUmembrane. Themembrane protonconductivity shows similar patterns with the membrane water uptake as the presence of water greatly influences the cross-linkedmembrane. The proton conductivity of the SPPSU cross-linked with 1 wt% SPOSS that was conducted under low RH (%) and atelevated temperature exhibited about six times higher proton conductivity as compared with pristine SPPSU membrane.However, increasing the loading of SPOSS beyond 1 wt% significantly dropped the membrane water uptake and protonconductivity due to SPOSS aggregation, blocking the hydrophilic domains in the polymer matrix. The results indicated thatthe incorporation of SPOSS in the SPPSUmembrane by curing at 180 °C exhibit improvement on membrane water managementand proton conductivity as compared with the pristine SPPSU membrane.Keywords Sulfonated PPSU . Proton exchangemembrane . Sulfonated POSS . Cross-linking . Fuel cellIntroductionThe scarcity of fossil fuel as energy sources, together with thedestruction of environmental issues due to the emission oftoxic gasses, has raised the adopted need for clean and greentechnologies for power generation. Fuel cells are nowadaysbecoming a rising technology that has been widely exploreddue to promising clean and efficient energy conversions.Despite modern world technology, fuel cells have known toscience for more than 150 years. Though generally considereda curiosity in the 1800s, fuel cells became the subject of in-tense research and development during the 1900s [1]. Fuelcells offer several advantages over conventional powersources include reducing dependence on fossil fuels, longuseful life, high efficiency, relatively safe, essential zero tox-icity, minimal maintenance costs, and carbon emission–free.This cell system also prepared a platform for a clean, quiet,and highly efficient process in the conversion of the fuel to* Juhana Jaafarjuhana@petroleum.utm.my* Je-Deok KimKIM.Jedeok@nims.go.jp1 Hydrogen Production Materials Group, Center for Green Researchon Energy and Environmental Materials, National Institute forMaterials Science (NIMS), 1-1 Namiki, Tsukuba, Ibaraki 305-0044,Japan2 Polymer Electrolyte Fuel Cell Group, Global Research Center forEnvironmental and Energy based on Nanomaterials Science(GREEN), National Institute for Materials Science (NIMS), 1-1Namiki, Tsukuba, Ibaraki 305-0044, Japan3 Advanced Membrane Technology Research Centre (AMTEC),School of Chemical & Energy Engineering, Faculty of Engineering,Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysiahttps://doi.org/10.1007/s10008-020-04594-2Journal of Solid State Electrochemistry (2020) 24:1185–1195/Published online: 4 2020Mayhttp://crossmark.crossref.org/dialog/?doi=10.1007/s10008-020-04594-2&domain=pdfmailto:juhana@petroleum.utm.mymailto:KIM.Jedeok@nims.go.jpenergy via an electrochemical process. In addition to low orzero emissions, benefits include high efficiency and reliability,multi-fuel capability, flexibility, durability, scalability, andease of maintenance [2].The fuel cell system is classified primarily by the type ofelectrolyte that has employed. The type of electrolyte usedwill determine the type of electrochemical reaction that hap-pens in the cell, the catalyst needed, the operating temperature,and the reactant required in the cell [3, 4]. Polymer electrolytemembrane fuel cells employed proton exchange membrane(PEM) as an electrolyte that is operating at intermediate tem-perature (80 to 120 °C), and low relative humidity (% RH)conditions have been extensively studied to be applied in au-tomobile transportation. Since the radiators for automobilesare designed to operate at an intermediate temperature, de-signed PEM should be complying with the same temperaturerange. Nafion (perfluorosulfonic acid)-based polymer electro-lyte membrane by DuPont is the most and common commer-cially available PEM in the market that has high hydrolyticand oxidative stability also excellent proton conductivity [5].The state of the art of the PEM fuel cel l–basedperfluorosulfonic acid ionomer that has lower proton conduc-tivity along with rising temperature and water managementissues has limited the application of the Nafion to be appliedat intermediate temperature ranges [6]. A series of studieshave been performed for the development of alternativePEMs using sulfonated hydrocarbon polymer for the applica-tion at intermediate temperature applications with excellentproton conductivity and water mobility.Polyphenylsulfone (PPSU) represents a family of hydro-carbon polymer that has been widely explored as an alterna-tive to proton-conducting membrane instead of expensiveperfluorinated membranes. PPSU has excellent thermal stabil-ity and appropriate mechanical strength with high proton con-ductivity, which increases along the degree of sulfonation [7].Unfortunately, highly sulfonated materials suffered from me-chanical and chemical deterioration due to excess membraneswelling in water causing a reduction in proton conductivity[8]. The mechanical weaknesses of the highly sulfonatedPPSU have initiated several attempts to prepare a more stableproton-conducting membrane. Recently, the cross-linkingprocess has been studied to improve the poor dimensionalstability and mechanical properties of the highly sulfonatedPPSU through bridging links to the reactive sulfonic acidfunctions without deterioration of proton conductivity [9]. Itis practicable to improve the mechanical strength and surpassmembrane swelling in water by cross-linking technique.However, the cross-linking between sulfonic acid groups hasreduced the acid function of the proton transport channel [10].The formation of micro-phase-separated morphology withnanochannel for improving proton conductivity has obtainedby blending ionomeric polymers with organic-inorganic hy-brid materials.This study explored the properties of composite sulfonatedpolyphenylsulfone (SPPSU) membrane blending withorganic-inorganic hybrid materials. The composite membranehaving inorganic materials is believed to avoid the trade-offbetween the water swelling and mechanical properties of themembrane. 1.5-nm closed cage polyhedral oligomericsilsesquioxane (POSS) form of nanosilica is one kind oforganic-inorganic hybrid material that becomes attractive ma-terials to be explored as filler. POSS is an organic-inorganichybrid material that combines the inorganic characteristicspresented by the siloxane bond (Si–O–Si) with silica as themain chain [11]. POSS has a general formula of RSiO1.5where R may be hydrogen or organic functional group thatattaches to the Si. POSS particles that are dispersed in thepolymer matrix can produce a nanoscale organic-inorganicinterface that influences the hydrophilic channel of PEM.POSS molecules that carry different types of organic substit-uents on its outer surface make this nanostructure compatiblewith many polymers. The organic substituents can also bemodified to enhance the compatibility with a specific polymermatrix. In this study, the water management and physico-chemical characteristics of the composite SPPSU-SPOSS–cross-linked membrane comprising different loading ofsulfonated POSS are presented and discussed.Experimental procedureMaterialsThe commercial polyphenylsulfone (PPSU, Solvay Radel R-5000 NT, MW ~ 50,000) purchased from Solvay SpecialtyPolymers, Japan, was used as the polymer backbone in themembrane formation. Highly sulfonated PPSU polymer wassulfonated using sulfuric acid (H2SO4, ~ 98%) purchasedfrom Sigma-Aldrich as a sulfonating agent. Meanwhile, thecommercial polyoctahedral silsesquioxane (POSS, HybridPlastics Inc., U. S, FW ~ 1033.53) was used as an inorganicfiller. Chlorosulfonic acid was used as a sulfonating agent todevelop sulfonated POSS.Preparation of highly sulfonated polyphenylsulfonePPSU polymer was first dried in an oven at 80 °C for about48 h to remove moisture contents. Dried PPSU polymer wasadded into 1 L of sulfuric acid (H2SO4, ~ 98%) at 60 °C in anoil bath. The solution was continuously stirred for 48 h andmaintain at 60 °C. The sulfonated polymer was recovered byprecipitating the sulfonic acid solution into the huge excess ofice. The resulted precipitate sulfonated PPSUwas then filteredusing mild vacuum filtration and then was subjected to neu-tralize by excess amounts of deionized water in a dialysis-tubing cellulose membrane until pH 7. The resulting SPPSUJ Solid State Electrochem (2020) 24:1185–11951186polymer was then dried at 80 °C until it was completely driedand ready to use. The ion exchange capacity (IEC) value of theprepared SPPSU polymer was estimated using titration tech-nique to determine the degree sulfonation of SPPSU. Thedegree sulfonation of prepared SPPSU corresponds to 1.86in which there are about two sulfonic acid groups attachedto one repeating unit of the SPPSU polymer matrix.Preparation of sulfonated POSSOctaphenyl-POSS (3 g, 2.9 mmol) was first dried in an oven at80 °C for more than 24 h to remove moisture content.Chlorosulfonic acid (ClHO3S, 60 mL, 900 mmol) was addedto dried POSS and was stirred for 24 h at 50 °C in an oil bathunder Ar gas conditions. Unreacted ClHO3S was removed byvacuum distillation at 110 °C. The crude products werewashed with water several times, and the excess water wasevaporating under normal heating. The brownish liquid prod-ucts designated as SPOSS were formed, and the IEC valuesusing the titration method show a value of 7.84 meq/g.SPPSU-SPOSS–cross-linked membrane preparationThe SPPSU-SPOSS–cross-linked membrane was preparedunder a single casting step using a slow evaporation techniqueto develop a dense membrane structure. The dope solution formembrane preparation was carried out by varying the differentSPOSS wt% loading. Firstly, 10% of SPPSU was dissolved inDMSO, and the dope solution was continuously stirred for24 h until it became homogeneous. Then, different wt% ofSPOSS were mixed in the solution and stirred for another24 h to produce a homogenous solution. In the preparationof the cross-linked membrane, the SPPSU-SPOSS dope solu-tion was first spread in the glass petri dish that acquires themembrane thickness about ~ 70 μm. The cast membrane wasdried in the drying oven at 80 °C for 24 h until the solvent wascompletely removed. The nanocomposite membrane was thenundergoing the thermal cross-linking process to improve themechanical properties of the membrane. The membranes werefurther heated in air at 120 °C (24 h), 160 °C (24 h), and180 °C (24 h). After the thermal cross-linking process, themembrane was then activated in a different solution. The re-sidual acids and water-extractable molecules will be removedfrom the post-activation step. The cross-linking membranewas then immersed in 0.5 M of NaOH (80 °C, 2 h), boilingwater (2 h), 1 M of H2SO4 (80 °C, 2 h), and boiling water(2 h). Then, the activated membrane was dried at room tem-perature for further characterizations.CharacterizationsFourier transform infrared spectroscopy (FTIR) of the pre-pared membrane was obtained by attenuated total reflection(ATR) with an infrared (IR) spectrophotometer (Nicolet-6700,Thermo Scientific) in the frequency range of 4000–400 cm−1.The IEC values of the sulfonated POSS were determinedusing the titration method. SPOSS was mixed in 20 mL of2 M NaCl for about 24 h under continuous stirring to replacethe protons with sodium ions. The solution was then titratedwith 0.01 M NaOH solution until the pH turned to 7. The IECvalue of the test sample was calculated using the following Eq.(1):IEC meq=gð Þ ¼ cv=wdry ð1Þwhere c (molar) is titration solution (0.01 M), v (mL) is thevolume of the neutralized NaOH, and Wdry (g) is the dryweight of the sample tested.Meanwhile, the water uptake of the prepared membranewas determined by the membrane weight variation beforeand after hydration. The membrane was first dried in the ovenat 80 °C for 24 h. The dry membrane was weighted and de-noted asWdry. Then, the dry membrane was immersed in boil-ing water for 1 h. After that, the wet membrane was removedfrom boiling water, and the membrane was mopped with blot-ting paper to remove the excess water on the membrane sur-face. The wet membrane was weighed and denoted as Wwet.The water uptake of the prepared sample was calculated as thefollowing Eq. (2):Water uptake %ð Þ ¼ Wwet−Wdry� �=Wdry� � ð2ÞThe hydration number of the membrane (λ) which wasclassified as the average number of water molecules perconducting functional group was calculated using the nextEq. (3):Hydration number λð Þ ¼ mol H2Omol acid groups¼ 1000 Wwet−Wdry� �18WdryIECð3ÞThe swelling ratio of the prepared sample was determinedbased on the membrane variation before and after hydrationon the membrane thickness and dimensions. The swellingratio was determined based on the following Eqs. (4) and (5):Thickness; S:R %ð Þ ¼ twet−tdry� �=tdry� �� 100 ð4ÞDimensions; S:R %ð Þ ¼ lwet−ldry� �=ldry� �� 100 ð5Þwhere twet and lwet was the membrane thickness and lengthafter immersing in boiling water for 1 h while tdry and ldry werethe membrane thickness and length in dry conditions.J Solid State Electrochem (2020) 24:1185–1195 1187The mechanical properties were measured by determiningthe tensile strength and elongation at the break of cross-linkedmembranes. The measurement was performed by using a ten-sion test machine (Shimadzu, EZ-S) at room temperature witha load cell of 100 N at a displacement rate of 0.5–1.0 mm/min.The samples were cut by using a super dumbbell cutterSDMP-1000 (Dumbbell Co.) having a width around 2.7–4.8 mm. At least three sample measurements were conductedfor each sample.Thermal properties of the SPPSU and SPPSU-SPOSS–cross-linkedmembrane was analyzed using thermogravimetryanalysis (TGA, STA 8000, Perkin Elmer) with alumina pansat a heating rate 10 °C/min from 30 to 800 °C under flowing ofnitrogen gas. The membrane was dried at 80 °C for about 24 hbefore measurement.The proton conductivity of the SPPSU membrane wasmeasured using a 4-point probe impedance spectroscopy byMTS 740 test system (Scribner Associates, Inc.) with a phase-sensitive multimeter (model PSM1735, Newtons4th Ltd.)combined with an impedance analysis interface. The conduc-tivities were measured at a temperature of 80 °C and 120 °C atdifferent % RHs (40, 60, 80, 90). Frequency ranges of 1 Hz to1MHz and a peak-to-peak voltage of 10mVwere used for theimpedance measurements.Single-cell performance testThe thickness of the tested membrane was approximately42 μm. 0.5 mg/cm2 40 wt% Pt/C was used as a single-cellelectrode catalyst, and an active electrode area is 2.25 cm2.The membrane electrode assembly (MEA) was prepared byhot pressing the membrane between the anode and cathodelayer at 75 kg/cm2 at room temperature for 1 min. The I–Vpolarization curves were recorded by a single-cell test at 80 °Cunder 60% RH and 100% RH. The flow rate of hydrogen wasfixed at 100 mL/min and 100 mL/min for the pure oxygen.The backpressure was atmospheric pressure. The I–V perfor-mance was compared with commercial Nafion 117 (Sigma-Aldrich; thickness, 177.8 μm).Results and discussionsThe prepared sulfonated SPPSU polymer in this study washighly soluble in water. It is well understood that a highlyswelled membrane is troublesome to the mechanical stabilityof a membrane. In order to overcome the issue, highlysulfonated PPSU polymer was cross-linked with SPOSS byheat treatment. Figure 1 a and b show the photographic imagesof the SPPSUmembrane heating at 80 °C and 180 °C. SPPSUmembrane heating at 80 °C, resulting in a transparent mem-brane while the color of the SPPSU membrane heating at180 °C turns to light brown. After heating at 80 °C and180 °C, the membrane was smooth and very flexible. Themembrane solubility in water was tested by immersing bothmembranes in the water at room temperature. SPPSU mem-brane heated at 80 °C was immediately dissolved in waterwhile the SPPSU membrane heated after 180 °C showed nodifference in the membrane appearance. The improvement inthe mechanical strength and dimensional stability of the mem-brane heating at 180 °C suggested that heat promoted thecross-linking between the SPPSU polymer matrixes. Thecross-linking SPPSU membrane further illustrated by theFTIR spectra shown in Fig. 1c. IR spectra displayed a broadpeak around 3411 cm−1 assigned to O–H stretching, and1693 cm−1 of O–H bending vibration from sulfonic acidgroups interacts with water molecule for SPPSU-80 °C. TheO–H peak of hydroxyl groups was decreased for SPPSU-180 °C, indicating that dehydration of water as the cross-linking temperature increased. The characteristic peak at1411 cm−1 and 1376 cm−1 was attributed to SO3H vibration.The intensity peak of SO3H for the SPPSU-180 °C was re-duced compared with SPPSU-80 °C. These indicate thatSPPSU heat at 180 °C has promoted the cross-linking withinthe SPPSU polymer matrix that simultaneously improves themechanical strength and dimensional stability of themembrane.The FTIR spectra were measured to explore the structuralproperties of the SPPSU-SPOSS cross-linking membrane.The IR spectra of the cross-linked membrane are illustratedin Fig. 2. The absorption peak of the SPPSU-SPOSS–cross-linked membrane shows almost identical spectra with pristineSPPSU-cross-linked membrane. The characteristic peak at1488 cm−1, 1411 cm−1, and 1373 cm−1 attributed to thestretching vibration of the C=C bond of aromatic PPSU. AsSPOSS was incorporated into the SPPSU polymer matrix, thedensities of the C=C vibration peak were reduced. The strongsignal at 1010 cm−1 assigned to the vibration peak of symmet-ric stretchings of the sulfonic acid group. The densities of thecharacteristic peak of the symmetric sulfonic acid group de-creased as the SPOSS loading was increased. It might be dueto the higher number of sulfonic acid groups taking part incross-linking reaction during heat treatment. Although it is notquantitatively reported, the spectroscopic observation indi-cates that there is some modification in the SPPSU polymermatrix chemical structure with the incorporation of SPOSS.The chemical structure of highly sulfonated PPSU and possi-ble cross-linking mechanism of SPPSU and SPOSS are illus-trated in Fig. 3. The unique structure of organic and inorganichybrid of SPOSS with its nanosized structure and rich surfacefunctional sulfonate group can improve the number of chem-ical bonding between SPOSS and SPPSU as a cross-linkingreaction under heat treatment occurs mainly between the sul-fonic acid groups.The most important characteristic that can reflect the per-formance of a PEM is proton conductivity conducted throughJ Solid State Electrochem (2020) 24:1185–11951188the water content of a membrane. The proton conductivity isespecially dependant on the water content and water uptake ofthe membrane. Proton transportation through the PEM is fun-damentally dependent on the content and distribution of waterwithin the membrane since the dominant proton transportationrequired water as a medium [12]. The water contents in themembrane affect the rate of proton transfer and hydrogen per-meability throughout the electrolyte layer. Lower water uptakeand excess water uptake can cause reduction of the protontransportation rate at lower humidity conditions and deterio-rate mechanical strength, respectively. Thus, the optimumamount of membrane water uptake is required to give a bal-ance between those properties. Figure 4 shows the water up-take and water content of the SPPSU-SPOSS cross-link mem-branes. The water uptake for SPPSU-0% SPOSS, SPPSU-1%SPOSS, SPPSU-2% SPOSS, and SPPSU-5% SPOSS was de-termined to be 43.0%, 127.7%, 79.8%, and 33.1%, respective-ly. Adding 1 wt% of SPOSS can cause water uptake of morethan 100% in the cross-linked membrane. However, increas-ing the SPOSS loading of more than 1 wt% contributes to asignificant reduction in water uptake. It is indicated that theparticles are aggregated and blocking the hydrophilic domainsin the polymer matrix [13]. Hydration number was calculatedas a parameter to define the number of water molecules foreach sulfonic acid groups. As can be seen in Fig. 4, the hy-dration number of SPPSU-incorporated SPOSS content washigher than the pristine SPPSU. It was advocated that theSPOSS was bearing a sulfonic acid group, enhancing thenumber of water content inside the cross-linked membranefor better proton transportation.Meanwhile, ion exchange capacity (IEC) of the cross-linkedmembrane depends on the content of the active sulfonicacid group for proton transportation [14]. Table 1 shows theIEC, swelling ratio, and conductivity values of the cross-linked membrane. When the SPOSS was incorporated intothe SPPSU, there is a significantly decrease in the IEC. Itcould be related to the decrease of the available number ofthe sulfonic acid group per unit mass due to the –SO3H grouptaking part in cross-linking reaction under thermal treatment.It leads to having the lower availability of the exchangeableproton transport. Despite having lower IEC with increasingloads of SPOSS, all the composite membranes show a highervalue of hydration number compared with pristine SPPSU.Water uptake and membrane swelling become essential pa-rameters that need to be determined in which this property isdirectly affecting almost entire membrane properties [15].Although water uptake of the SPPSU-SPOSS membrane ishigher than the pristine SPPSU membrane, the swelling on4000 3500 3000 2500 2000 1500 1000(a)721cm-11376cm-11411cm-11693cm-1)u.a(rebmunevaWWavelength (cm-1)(a) SPPSU-180 oC(b) SPPSU- 80 oC3411cm-1(b)a)b)In waterc)Fig. 1 Photographic image of SPPSU membrane after heating at a 80 °C, b 180 °C, and c FTIR of SPPSU membrane2000 1500 10001372 cm-11411 cm-1(d)(c)(b)(a) SPPSU-5% SPOSS(b) SPPSU-2% SPOSS(c) SPPSU-1% SPOSS(d) SPPSU-0% SPOSS)u.a(ecnattimsna rTWavenumber (cm-1)(a)1488 cm-1 1010 cm-1Fig. 2 IR spectra of SPPSU-SPOSS–cross-linked membrane at 2000–525 cm−1J Solid State Electrochem (2020) 24:1185–1195 1189the membrane dimension is not so much different. It might bedue to accumulation and the cross-linking effect of SPOSS inthe SPPSU polymer matrix that prevents the membrane fromswelling. This characteristic has confirmed with the mechan-ical strength properties that will be further discussed in Fig. 5.The SPPSU-SPOSS–cross-linked membrane was furthercharacterized for its mechanical properties. Proton-conducting membrane should have excellent mechanical sta-bility as high temperature and humidity in an operating cellsystem significantly make a change in membrane mechanicalproperties due to degradation or aging [16]. Figure 5 showsthe stress-strain curve of the SPPSU-SPOSS–cross-linkedmembrane. The mechanical strength and ultimate elongationof the SPPSU membrane are 47.77 MPa and 74.11%, respec-tively. SPPSU-cross-linked membrane shows a characteristicof an elastomeric membrane with significantly lower strengthbut more excellent elongation. Meanwhile, the membranestrength and ultimate elongation at break of the SPPSU-SPOSS–cross-linked membrane lie between 26.62 to52.30 MPa and 36.06 to 72.98%, respectively. The mechani-cal properties of the SPPSU-cross-linked membrane werefound to be affected by the chemical composition of SPOSSand cross-linked structure. From Fig. 5, 2 wt% of SPOSSloading behaved like elastomeric materials with lowerstrength but mostly excellent in membrane strain.Meanwhile, when the loading increased up to 5 wt%SPOSS, the membrane shows tough characteristics and visi-bly more rigid compared with lower SPOSS loading. Themechanical behavior indicates that SPOSS effectively sup-presses excessive water absorption and increases the tensilestrength of the cross-linked membrane [17].Thermal management is one of the critical characteristicsfor longer terms, and stable performance in PEM fuel cell(PEMFC) as membrane materials sometimes loses their per-formance due to thermal degradation. Generally, the thermalanalysis could provide information on the temperature-dependent properties of the membrane. The thermal propertiesof the SPOSS- and SPPSU-SPOSS–cross-linked membraneswere measured using the TGA analysis. Figure 6 illustratesTG curves for SPOSS- and SPPSU-SPOSS–cross-linkedmembranes obtained under the nitrogen gas atmosphere. Thetemperature at 5% mass loss (Td5%) was determined to mea-sure the resistance of the material towards initial thermal deg-radation [18]. The Td5% of SPOSS measured by TG analysiswas 265 °C. The TG curves of SPOSS are illustrated inFig. 6a. SPOSS is bearing silica as the main chain havinghigher thermal retention as the materials decompose onlyabout 38.6% after being heated up to 800 °C. Higher thermalO SSO3HHO3SOOnSO3HSO3HSO3HPOSSSO3H 8SO2SO3HSO2SO2SO2SO3HHO3SSO3HSO3HO2SSO3HSO3HSO3HHO3SSO2SO2HO3SSO2SO2SO2SO2SO2SO3HHO3SSO3HSO3HSO2HO3S SO3HSulfonated PPSU (SPPSU)Sulfonated POSS (SPOSS)Fig. 3 Chemical structure of highly sulfonated PPSU with the possible cross-linking structure of the cross-linked membrane during thermal treatment0% SPOSS 1% SPOSS 2% SPOSS 5% SPOSS50100150%,ekatpUretaW01020304014.727.037.3,tnetnoCret aW11.9Fig. 4 Water uptake and water content of the SPPSU-SPOSS–cross-linked membraneJ Solid State Electrochem (2020) 24:1185–11951190degradation of SPOSS offers advantages over the cross-linkedmembrane as it can improve the thermal properties of theSPPSU polymer [19].SPPSU membrane without incorporating SPOSS fillershows obviously two steps of degradation. The first degrada-tion temperature is around 267.9 °C and the second is at501.6 °C. These degradation steps depicted towards the sub-stitution of SO3H groups and thermal degradation of the poly-mer backbone, respectively. These thermal characteristics arecomparable with the previously reported study [20]. The tem-perature at which a Td5% mass loss of SPPSU-0% SPOSS was300.6 °C. As 1 wt% of SPOSS was incorporated, the Td5%degradation was increased to 330.6 °C. The existence ofSPOSS within the SPPSU polymer matrix had slightly de-layed the oxidative degradation of the polymer matrix in situwith improving the thermal stability of the cross-linked mem-brane [15]. This might be due to the nano reinforcement effectof SPOSS that is responsible for increasing the initial decom-position of the cross-link membrane [21]. As the SPOSS load-ing increased, the Td5% reduced to 312.3 °C and 311.5 °C forSPPSU-2%SPOSS and SPPSU-5%SPOSS, respectively.There were not many significant changes in the thermal deg-radation of the SPPSU-SPOSS–cross-linked membranecompared with the SPPSU membrane. The TGA curves areillustrated in Fig. 6b. The thermal degradation temperature ofthe cross-linked membrane due to the evaporation of watermolecules interacting with sulfonate groups was higher thanthe pristine SPPSU membrane which might be due to thecross-linking effect between SPOSS and SPPSU polymer thatresists the degradation of the water molecule.A polymer electrolyte membrane having high proton con-duction properties is in strong demand to be applied in the fuelcell system [22]. In particular, the high proton conductivemembrane is needed to obtain high voltage per current densityin the unit cell [23]. The proton conductivity was measuredfrom the resistance of the membrane against the current flow.In this study, the proton conductivity of the cross-linked mem-brane was measured using the 4-point probe impedance spec-troscopy under different relative humidity (% RH) conditions.The proton conductivity of the cross-linked membrane carry-ing different loadings of SPOSS is illustrated as in Fig. 7. Theproton conductivity was enhanced about six times at 40% RHat 80 °C over 1 wt% of SPOSS incorporated into the SPPSUpolymer. At a higher temperature of 120 °C, the proton con-ductivity of SPPSU-1% SPOSS enhanced more than ten timesat low relative humidity. Alternatively, POSS grafted withsulfonic acid groups can provide external proton channelsfor nanocomposite membranes which has resulted in improve-ments on the proton conductivity values of the membrane[24].Unfortunately, as the loading of the SPOSS increased, theproton conductivity values gradually decreased. It is sug-gested that higher content of SPOSS resulting in particle ag-gregation corresponds to reduce the active site for protontransportation [25]. These results of proton conductivity areequivalent to the membrane water uptake in which the con-ductivity decreased upon higher SPOSS loading. It suggestedthat the proton conduction requires water-assisted pathwaysfor proton transportation, and water acts as a proton carrier.The proton will travel along with the hydrogen-bonded ionicchannels, and proton conductivity highly depends on the con-nectivity of the hydrated domains [26]. Therefore, it resultedin proton conductivity values of the SPPSU-SPOSS compos-ite membrane following the same trend, based on the waterTable 1 IEC, water uptake,swelling ratio, and conductivity ofthe cross-linked membranesSample IEC (meq/g) SR (t/l) (%) Conductivity, S/cm at80 °CConductivity, S/cm at120 °C40% RH 90% RH 40% RH 90% RHSPPSU-0% SPOSS 2.00 16.4/11.0 0.0011 0.0112 0.0006 0.0131SPPSU-1% SPOSS 1.90 26.2/40 0.0065 0.0379 0.0066 0.0498SPPSU-2% SPOSS 1.64 23.6/30 0.0042 0.0313 0.0037 0.0361SPPSU-5% SPOSS 1.25 11.1/7.7 0.0009 0.0094 0.0006 0.01030 20 40 60 80 100020406080(b)(d)(c))aPM(ssertSStrain (%) (a) SPPSU-1% SPOSS (b) SPPSU-2% SPOSS (c) SPPSU-5% SPOSS (d) SPPSU-0% SPOSS(a)Fig. 5 Stress-strain of SPPSU-SPOSS–cross-linked membrane withdifferent loadings of SPOSSJ Solid State Electrochem (2020) 24:1185–1195 1191uptake results in which higher membrane water uptake result-ed in higher proton conductivity. From the obtained results,incorporating SPOSS in the SPPSU polymer matrix indeedgreatly improved the properties of the highly sulfonatedPPSU as a proton-conducting membrane.Based on the physicochemical properties of the mem-branes, SPPSU-2% SPOSS nanocomposite membrane gener-ated a significant improvement as a proton exchange mem-brane compared with SPPSU-1% SPOSS and SPPSU-5%SPOSS. Therefore, the performance of SPPSU-2% SPOSSnanocomposite membrane was further characterized in termsof its cell voltages and power densities as a function of currentdensity in a single-cell PEMFC. In this study, the single-cellperformance was tested at the standard operating temperatureof the PEMFC system, which is at 80 °C. The effect of thedifferent % RH conditions (60% RH and 100% RH) on thecell performance was also analyzed. Figure 8a shows thesingle-cell voltages and power density of SPPSU-2%SPOSS concerning the current density under 60% RH and100% RH conditions. At 100% RH, SPPSU-2% SPOSSnanocomposite membrane showed better performance ascompared with 60% RH. The collected voltage potential ofthe cell at 200 mA/cm2 is 0.48 Vand 0.56 V for 60% RH and100% RH, respectively. The fact is that the decreasing outputvoltage was due to significant ohmic losses in the MEA, inwhich increasing ohmic losses caused the drop in output volt-age [27]. The ionic resistance dominates the ohmic losses inthe PEM fuel cell, representing resistance against proton trans-fer from anode to cathode due to the membrane interfaces[28]. Ionic resistance is closely related to the membrane thick-ness and proton conductivity in which ionic resistance in-creased when the proton conductivity is reduced [29]. Thus,these behaviors are consistent with the lower proton conduc-tivity values of SPPSU-2% SPOSS nanocomposite membranemeasured at low % RH conditions resulting in lower potentialvoltage. The peak current density of the cross-linked mem-brane at fully hydrated conditions reached up to 244.44 mA/cm2 and 297.75 mA/cm2, respectively. The maximum powerdensity of the cross-linked membrane at 100% RH conditionsexhibited as high as 133.51 mW/cm2, which are higher than at60% RH conditions (101.84 mW/cm2). Low-humidity condi-tions in the cell may increase the resistance in the membranethat can restrict transportation, causing performancedegradation.a) b)200 400 600 8006080100265 oC)%(thgieWTemperature (oC)Td5%200 400 600 800406080100)%(thgieWTemperature (oC) (a) SPPSU-0% SPOSS (b) SPPSU-1% SPOSS (c) SPPSU-2% SPOSS (d) SPPSU-5% SPOSSTd5% (SPPSU-0%SPOSS): 300.6 oCTd5% (SPPSU-1%SPOSS): 330.4 oCTd5% (SPPSU-2%SPOSS): 312.3 oCTd5% (SPPSU-5%SPOSS): 311.5 oCTd5% (b)(c)(d)(a)Fig. 6 TGA curves for a SPOSS- and b SPPSU-SPOSS–cross-linked membranes40 60 80 10010-310-210-1(c)(a)(d))mc/S(ytivitcudnoCRH (%) (a) SPPSU-0% SPOSS(b) SPPSU-1% SPOSS (c) SPPSU-2% SPOSS (d) SPPSU-5% SPOSST = 80 oC(b)40 60 80 10010-310-210-1(d)(a)(c))mc/S(ytivitcudnoCRH (%) (a) SPPSU-0% SPOSS (b) SPPSU-1% SPOSS(c) SPPSU-2% SPOSS(d) SPPSU-5% SPOSS(b)T = 120 oCa) b)Fig. 7 Proton conductivity of the SPPSU-1%, SPPSU-2%, and SPPSU-5% SPOSS membranes at a 80 °C and b 120 °CJ Solid State Electrochem (2020) 24:1185–11951192Under optimized conditions of single-cell PEMFC per-formance test, the I–V polarization curves of SPPSU-2%SPOSS nanocomposite membrane were compared withthe commercial Nafion 117 and illustrated as in Fig. 8b.From Fig. 8b, SPPSU-2% SPOSS nanocomposite mem-brane showed good performance compared with Nafion117 measured at 80 °C under fully hydrated conditions.The peak current density of SPPSU-2% SPOSS nanocom-posite membrane reached up to 297.80 mA/cm2 comparedwith 257.8 mA/cm2 for Nafion 117. The maximum powerdens i ty of SPPSU-2% SPOSS exhib i t ed abou t133.51 mW/cm2, which is higher than the maximum pow-er density for Nafion 117 (111.76 mW/cm2). The perfor-mance of the SPPSU-2% SPOSS nanocomposite mem-brane was further tested under the long-term durabilityof the MEAs to predict the feasibility of the preparedmembrane towards fuel cell applications. The potentialvoltage against times of the SPPSU-2% SPOSS mem-branes is illustrated in Fig. 8c. It shows that SPPSU-2%SPOSS nanocomposite membrane exhibited good electro-chemical stability under constant current (0.1 A) at oper-ating conditions of 80 °C and 100% RH. Only small volt-age drop was observed after 8 h of operations. It isinteresting to state that the MEA using SPPSU-2%SPOSS nanocomposite membrane shows good electro-chemical properties under operating conditions of 80 °Cand 100% RH, which are comparable with commercialNafion 117.ConclusionsThe research results on the effect of various loadings ofSPOSS in highly sulfonated PPSU membrane subjectedunder thermal cross-linking reaction as the proton-conducting membrane was studied. The properties of theSPPSU-SPOSS-cross-linked membrane was studied byvarying the loading of the SPOSS (1%, 2%, and 5%).SPPSU-SPOSS–cross-linked membrane shows higher wa-ter uptake compared with pristine SPPSU-cross-linkedmembrane. Increasing SPOSS loading of more than1 wt%, the water uptake and IEC values were decreased.The mechanical characteristics of the SPPSU-SPOSS–cross-linked membrane were found to be affected by thechemical composition of SPOSS as the mechanicalstrength was improved with the incorporation of SPOSS.a) b)0 100 200 300 400 5000.00.20.40.60.81.0 60% RH 100% RHCurrent Density (mA/cm2))V(egatloV020406080100120140160ytisneDrewoP( mW/cm2)0 100 200 300 400 5000.00.20.40.60.81.0 SPPSU-2% SPOSS Nafion 117Current Density (mA/cm2))V(egatloV020406080100120140160ytisneDrewoP(mW/cm2)0 1 2 3 4 5 6 7 8 9 100.00.20.40.60.81.0)V(egatloveerfRiTime (h) SPPSU-2% SPOSSc)Fig. 8 Polarization curves of a SPPSU-2% SPOSS nanocompositemembrane at 80 °C, 60% RH and 100% RH. b Comparison betweenSPPSU-2% SPOSS with commercial Nafion 117 at 80 °C under 100%RH conditions. c Voltage stability of SPPSU-2% SPOSS nanocompositemembranes under 0.1-A constant current for 8 hJ Solid State Electrochem (2020) 24:1185–1195 1193The proton conductivity of the SPPSU-SPOSS–cross-linked membrane also improves compared with pristineSPPSU membrane, indicating that adding SPOSS contrib-utes to more significant numbers of available sulfonic acidgroups for proton transportation. The PEMFC perfor-mance of SPPSU-2% SPOSS nanocomposite membraneshowed a good performance compared with Nafion 117measured at 80 °C under fully hydrated conditions. Basedon these findings, the SPPSU-SPOSS–cross-linked mem-brane yielded a significant improvement, which improvesin membrane mechanical strength and higher proton con-ductivity as compared with the SPPSU-cross-linkedmembrane.Acknowledgments The authors would like to express their gratitude tothe Ministry of Higher Education (MOHE), Universiti TeknologiMalaysia (UTM) and Research Management Centre (RMC), UTM, forsupporting the research management activities. The authors would alsolike to acknowledge support by the International Cooperative GraduateSchool (ICGS) Fellowship under the “Universiti Teknologi Malaysia-NIMS Cooperative Graduate School Program” to conduct research inthe National Institute of Materials Science (NIMS), Tsukuba, Japan.Funding information This work was supported by Ministry of HigherEducation (MOHE) under project grant MRUN (R.J130000.7851.4L880)and Universiti Teknologi Malaysia (UTMPR:Q.J130000.2851.00L22,UTM-TDR:Q.J130000.3551.06G88). This work also was supported bythe MEXT Program for the Development of Environmental Technologyusing Nanotechnology.References1. Kirubakaran A, Jain S, Nema RK (2009) A review on fuel celltechnologies and power electronic interface. Renew Sust EnergRev 13(9):2430–24402. Miyake J, Taki R,Mochizuki T, Shimizu R, Akiyama R, UchidaM,Miyatake K (2017) Design of flexible polyphenylene protonconducting membrane for next generation fuel cells. Sci Adv 3:1–83. Muhmed SA, Nor NAM, Jaafar J, Ismail AF, Othman MHD,Rahman MA, Aziz F, Yusof N (2019) Emerging chitosan and cel-lulose green materials for ion exchange membrane fuel cell: a re-view. Energy Ecol Environ 2019:1–234. Tada M, Uruga T, Iwasawa Y (2015) Key factors affecting theperformance and durability of cathode electrocatalysts in polymerelectrolyte fuel cells characterized by in situ real time and spatiallyresolved XAFS techniques. Catal Lett 145(1):58–705. Kyu T, Nazir NA (2013) Supramolecules impregnated proton elec-trolyte membranes. Curr Opin Chem Eng 2(1):132–1386. Park JW,Wycisk R, Pintauro PN, Yarlagadda V, Nguyen TV (2016)Electrospun nafion®/polyphenylsulfone composite membranes forregenerative hydrogen bromine fuel cells. Materials 9:1–157. Urena N, Perez-Prior MT, Rio CD, Varez A, Sanchez JY, Iojoiu C,Levenfeld B (2019) Multiblock copolymers of sulfonated PSU/PPSU poly (ether sulfone) s as solid electrolytes for proton ex-change membrane fuel cells. Electrochim Acta 302:428–4408. Kim YS, Pivovar BS (2010) Moving beyond mass-based parame-ters for conductivity analysis of sulfonated polymers. Annu RevChem Biomol Eng 1(1):123–1489. Kim JD, Ghil LJ (2016) Annealing effect of highly sulfonatedpolyphenylsulfone polymer. Int J Hydrog Energy 41(27):11794–1180010. Kamel MSA, Mohamed HFM, Abdel-Hamed MO, Abdel-HadyEE (2019) Characterization and evaluation of nafion HP jpJPs pro-ton exchange membrane: transport properties, nanostructure, mor-phology, and cell performance. J Solid State Electrochem 23(9):2639–265611. Zhang J, Chen F, Ma X, Guan X, Chen D, Hickner MA (2015)Sulfonated polymers containing polyhedral oligomericsilsesquioxane (POSS) core for high performance proton exchangemembranes. Int J Hydrog Energy 40(22):7135–714312. ZhengY, Ash U, Pandey RP, OziokoAG, Ponce-Gonzalez J, HandlM, Weissbach T, Varcoe JR, Holdcraft S, Liberatore MW, HiesgenR, Dekel DR (2018) Water uptake study of anion exchange mem-branes. Macromolecules 51(9):3264–327813. Mazzapioda L, Panero S, Navarra MA (2019) Polymer electrolytemembranes based on nafion and a superacidic inorganic additive forfuel cell application. Polymers 11(5):914 (1-10)14. Yen YC,YeYS, Cheng CC, Lu CH, Tsai LD,Huang JM, Chang FC(2010) The effect of sulfonic acid groups within a polyhedral olig-omeric silsesquioxane containing cross-linked proton exchangemembrane. Polymer 51(1):84–9115. Salarizadeh P, Javanbakht M, Pourmahdian S (2017) Enhancing theperformance of speek polymer electrolyte membranes using func-tionalized TiO2 nanoparticles with proton hopping sites. RSC Adv7(14):8303–831316. Shi S, Chen G, Wang Z, Chen X (2013) Mechanical properties ofnafion 212 proton exchange membrane subjected to hygrothermalaging. J Power Sources 238:318–32317. Kim K, Heo P, HwangW, Baik JH, Sung YE, Lee JC (2018) Cross-linked sulfonated poly (arylene ether sulfone) containing a flexibleand hydrophobic bishydroxy perfluoropolyether cross-linker forhigh-performance proton exchange membrane. Appl MaterInterfaces 10(26):21788–2179318. Mokhtaruddin SR, Mohamad AB, Loh KS, Kadhum AAH (2016)Thermal properties and conductivity of nafion-zirconia compositemembrane. Malays J Anal Sci 20(3):670–67719. Awang N, Jaafar J, Ismail AF (2018) Thermal stability and watercontent study of void free electrospun speek/cloisite membrane fordirect methanol fuel cell application. Polymers 10:1–1620. Matsushita S, Kim JD (2018) Organic solvent-free preparation ofelectrolyte membranes with high proton conductivity using aromat-ic hydrocarbon polymers and small cross-linker molecules. SolidState Ionics 316:102–10921. Wu Y, Li L, Feng S, Liu H (2013) Hybrid Nanocomposite based onnovolac resin and octa (phenyl) polyhedral oligomericsilsesquioxanes (POSS): miscibility, specific interactions, andthermomechanical properties. Polym Bull 70(12):3261–327722. Lee CH, Park HB, Lee YM, Lee RD (2005) Importance of protonconductivity measurement in polymer electrolyte membrane forfuel cell application. Ind Eng Chem Res 44(20):7617–762623. Khabibullin A, Minteer SD, Zharov I (2014) The effect of sulfonicacid group content in pore-filled silica colloidal membranes on theirproton conductivity and direct methanol fuel cell performance. JMater Chem A 2(32):12761–1276924. Kim SW, Choi SY, Rhee HW (2018) A novel sPEEK nanocompos-ite membrane with wee-controlled sPOSS aggregation in tunablenanochannels for past proton conduction. Nanoscales 10(38):18217–1822725. Yameen B, Kaltbeitzel A, Glasser G, Langner A, Muller F, GoseleU, Knoll W, Azzaroni O (2010) Hybrid polymer-silicon protonconducting membranes via a pore-filing surface-initiated polymer-ization approach. ACS Appl Mater Interfaces 2(1):279–287J Solid State Electrochem (2020) 24:1185–1195119426. Sengupta S, Lyulin AV (2018) Molecular dynamics simulations ofsubstrate hydrophilicity and confinement effects in capped nafionfilms. J Phys Chem B 122(22):6107–611927. Francia C, Ijeri VS, Specchia S, Spinelli P (2011) Estimation ofhydrogen crossover through nafion membranes in PEMFCs. JPower Sources 196(4):1833–183928. O’Hayre R, Cha S, ColellaW, Prinz FB (2009) Fuel cell fundamen-tals, Second edn. Wiley, New York29. Niya SMR, Hoorfar M (2014) Process modelling of the ohmic lossin proton exchange membrane fuel cells. Electrochim Acta 120:193–203Publisher’s note Springer Nature remains neutral with regard to jurisdic-tional claims in published maps and institutional affiliations.J Solid State Electrochem (2020) 24:1185–1195 1195 Improved... Abstract Introduction Experimental procedure Materials Preparation of highly sulfonated polyphenylsulfone Preparation of sulfonated POSS SPPSU-SPOSS–cross-linked membrane preparation Characterizations Single-cell performance test Results and discussions Conclusions References