A series of novel PBI/SrTiO₃ nanocomposite membranes composed of polybenzimidazole (PBI) and strontium titanate (SrTiO₃) with a perovskite structure were fabricated with various concentrations of SrTiO₃ through a solution casting method. Various characterization techniques such as proton nuclear magnetic resonance, thermogravimetric analysis, atomic force microscopy (AFM) and AC impedance spectroscopy were used to investigate the chemical structure, thermal, phosphate absorption and morphological properties, and proton conductivity of the fabricated nanocomposite membranes. The optimized PBI/SrTiO₃-8 polymer nanocomposite membrane containing 8wt% of SrTiO₃ showed a higher proton conductivity of 7.95 × 10^-2 S/cm at 160^oC compared to other nanocomposite membranes. The PBI/SrTiO₃-8 composite membrane also showed higher thermal stability compared to pristine PBI. In addition, the roughness change of the polymer composite membrane was also investigated by AFM. Based on these results, nanocomposite membranes based on perovskite structures are expected to be considered as potential candidates for high-temperature PEM fuel cell applications.

Keywords: Polybenzimidazole, Perovskite structure, Nanocomposite membrane, Thermal stability

Due to the recent rapid increase in industrial and domestic energy consumption and environmental pollution, researchers around the world are being required to develop renewable and sustainable energy storage and conversion technologies. Recently, sustainable energy conversion devices, including proton exchange membrane fuel cell (PEMFC), have attracted considerable attention in an electrical vehicle (EV), industrial plant, unmanned aerial vehicles (UAV_S), emergency backup power generation, and portable device applications due to their high efficiency, environmental friendliness, simplicity, quick startup, and low cost. Various investigations have been conducted in recent decades to optimize and develop PEMFC applications, along with the study of proton exchange membranes (PEMS) being particularly important [1-4]. At present, Nafion is used as a benchmark proton exchange membrane in fuel cell applications due to its chemical, thermal and mechanical stabilities along with high proton conductivity.
However, the Nafion and fluorinated polymers have high-cost, low conductivity, and low durability at high temperatures [5]. In recent decades, intensive efforts have been made to develop low-cost polymer-based membranes to replace Nafion in the fuel cell field. At present, organic-inorganic polymer nanocomposites have attracted considerable attention due to their improved physical, mechanical, thermal, and proton conductivity properties. These non-fluorinated membranes of poly(ether sulfone), poly(benzimidazole), poly(ether ether ketone), and poly(imide) are effectively used in PEM fuel cell application [6-9].
Since inorganic nanofillers provide high thermal and mechanical stability and pristine polymers provide flexibility and multi-functional reactivity, the combination of inorganic nanoparticles and polymer membranes provide improved aforementioned properties and it could be used for high temperature (HT) PEMFCs [10-15].
Poly(benzimidazole) (PBI) is an aromatic heterocyclic polymer, which contains both proton acceptor (-N=) and proton donor (-NH-) hydrogen bonding sites in the polymer backbone that is an excellent candidate for use in HT-PEMFCs due to their excellent thermal and mechanical stability. Generally, PBI is a better electronic and ionic insulator, while doping with acid shows excellent proton conductivity properties. PBI doping with phosphoric acid (PA) is an interesting technique for HT-PEMFC applications due to its high boiling point temperature, non-volatile property, high thermal stability, and high proton conductivity [16,17]. On the other hand, PA molecules with amphoteric nature and high dielectric constant at low humidity are suitable properties for substituting H₂O molecules in HT-PEMFC, which can create hydrophilic channels for cation exchange in the PA doped PBI membranes.
Perovskite is a stable type of oxide with the general formula ABO₃ can be a suitable choice for this purpose. A (cation A is 12-coordinated by oxygen, has a large ionic radius, and has a size similar to that of an oxygen ion) and B (cation B is 6-coordinated by oxygen and has a small ionic radius located in the octahedral holes between the closed pack AO layers) are suitable cations for the perovskite structure in the ABO₃ formula of perovskite [18].
Perovskite structures obtained from transition elements such as Ba, Ra, and Srcan be used as hygroscopic materials to improve the humidity sensing performance [19]. The proton conduction mechanism of the materials when interacting with water is well known as cation conductors [20]. Studies on perovskite-type ABO₃ materials have improved in recent years owing to their high proton conductivity in a humid atmosphere at high temperatures (1000^oC). The perovskite structures have high thermal, chemical, and mechanical stabilities. Perovskite-based SrTiO₃ displays a cubic structure with a lattice parameter of 3.91 Å and space group Pm3m [21]. In the regular octahedral coordination of the SrTiO₃ lattice, A-site occupies strontium cation, B-site occupies titanium, and oxide anion occupies periphery. It provides excellent proton transfer under certain atmospheric and doping conditions. Therefore, studies to improve the properties of cross-linked, composite or/and hybrid polymers are being conducted [22,23]. Among several types of PA-doped perovskites, such as PBI/Fe₂TiO₅ [24], PBI/SrCeO₃ [25], and PBI/BaZrO₃ [26], have been considered as potential materials to enhance the proton conductivity of composite membranes.
Based on the studies mentioned above, we designed PBI/SrTiO₃ nanocomposite membranes with different weight percentages of SrTiO₃ (2wt%, 4wt%, 6wt%, 8wt%, and 10wt%) and fabricated them by the solution casting method. In addition, through acid uptake, thermal stability, and proton conductivity tests of the fabricated polymer nanocomposite membranes, particularly the optimized PBI/SrTiO₃(8wt%) nanocomposite membrane, was evaluated to have improved thermal stability, proton conductivity, and surface roughness compared to other nanocomposites. We consider that the PBI/SrTiO₃ perovskite-based polymer nanocomposite membranes could be potential candidates for HT-PEMFC applications.

2.1 Materials
3,3-Diaminobenzidine (DAB, 99%), phosphoric acid (PA, 85%, HPLC grade), titanium isopropoxide (99%), isophthalic acid (IPA, 99%), sodium bicarbonate (NaHCO₃, 99%), methanol (99%), strontium nitrate (SrNO₃, 99%), potassium hydroxide (KOH, 98%) were purchased from Sigma-Aldrich, India. Polyphosphoric acid (PPA, 85%), N,N-dimethylformamide (DMF, 85%), N,N-dimethylacetamide (DMAc, 85%), N-methyl-2-pyrrolidone (NMP, 85%), tetrahydrofuran (THF, 85%) were procured from Alfa-Aesar, India. Deionized (DI) water was used for all experimental procedure.

2.2 Preparation of PBI
The PBI was synthesized from DAB (4.32 g, 0.012 mmol), IPA (1.99 g, 0.012 mmol) in 150 mL of PPA as a condensing agent by the synthesis method of Scheme 1. That is, both DAB and IPA were added into a 300 mL two-necked flask equipped with a mechanical stirrer under a nitrogen (N₂) atmosphere. The mixture was then stirred at 220^oC after 24 h, then an appropriate amount of DI water was added and neutralized with the aid of NaHCO₃. The obtained product was then washed with methanol and dried overnight in a vacuum oven at 120^oC [27,28]. The molecular weight (M_W) of synthesized PBI was 75 kDa, which was calculated by Ubbelohde viscometric studies.

2.3 Synthesis of SrTiO₃
The synthesis of SrTiO₃ was carried out using a 1:1 stoichiometric ratio of metal cations of SrNO₃ and titanium (IV) isopropoxide. The metal cation precursors in the above ratio were separately dissolved in DI water to form a transparent homogeneous solution under constant stirring for 4 h. A similar mole % of KOH solution was then added to the metal cation solution to control the pH level. The resulting product was then heated in a tubular furnace at 1000^oC in a N₂ atmosphere for 6 h. The obtained samples were collected and various characterization techniques were applied.


Scheme 1. Preparation of the PBI

2.4 Preparation of PBI/SrTiO₃–X composite membranes
As shown in Scheme 2, PBI/SrTiO₃-X nanocomposite membranes were prepared by the solution casting method. First, a proper amount of prepared PBI was dissolved in DMAc with stirring at 100~120^oC. Thereafter, different weight percentages of SrTiO₃ (2wt%, 4wt%, 6wt%, 8wt%, and 10wt%) nanoparticles were added into the above PBI polymer solution, and the mixture was stirred at room temperature for 1 day. The obtained brown solution was poured into a Petri dish and kept at 160^oC in an oven for 1 day. The resulting membranes were removed from the Petri dish and immersed in the PA solution. The obtained PBI/SrTiO₃-X (X is 2, 4, 6, 8, 10wt%) polymer composite membranes were stored in vacuum desiccators to avoid moisture absorptions. The thickness of the prepared polymer composites was measured at various positions using a digital screw gauge and thickness was about 50 µm.

2.5 Characterizations
Proton nuclear magnetic resonance (¹H NMR, JNM-ECA600-600 MHz) was used to ensure the chemical structure of the polymers using deuterated DMSO-d₆ as a solvent. The morphology and surface roughness of the pristine PBI and PBI/SrTiO₃ composite membranes were analyzed using an atomic force microscope (AFM, multimode 8, Bruker Corporation) in tapping mode. Thermogravimetric analysis (TGA) was performed using SDT Q600 systems. The TGA measurement was measured by heating from 25 to 1000^oC under an N₂ atmosphere at a heating rate of 10^oC/min.

2.6 Measurements
A piece of the composite membrane having a size of 1×1 cm was immersed in an 85% PA concentration solution for 1 day. The PA doping level of the membranes was defined as follows. The number of the PBI repeating units contained in 1 mole of PA is as shown in equation (1).



Where w_o and w₁ are the weights of the dry membrane and doped membrane, respectively. Values 98 and 308 are the molecular weights of the PA and PBI repeat units, respectively. Leaching test is a method of measuring the residual amount of PA after washing the PA-doped PBI/SrTiO₃ nanocomposite membrane with hot water (the membrane is immersed in hot DI water at 90^oC for 2 h). It is considered as one of the major decomposition factors of the PBI/SrTiO₃ composite membranes in fuel cells. AC impedance measurements were performed on polymer electrolytes using a computer-controlled μ-Auto lab Potentiostat/Galvanostat in the frequency range of 0.1 Hz to 10 MHz with a voltage amplitude of 50 mV [29]. Measurements were carried out in the potentiostat mode.



Where σ is the ionic conductivity value (S/cm), L is the thickness of the membrane film, R is the bulk resistance of the composite membrane film, and A (1 cm × 1 cm) is the area of the sample.


Scheme 2. Preparation of PBI/SrTiO3–Xcomposite membranes

3.1 NMR analysis of pure PBI
¹H NMR spectroscopy is used as a scrutinizing technique to confirm the presence of the polymer structure. The ¹H NMR spectrum of PBI with DMSO-d₆ solvent is shown in Fig. 1. The main peak 13.25 ppm is the amine protons of the benzimidazole moiety and peaks 7.59 to 9.15 ppm represent the aromatic groups of the polymer [28]. The results suggest that the carboxyl and amine groups are involved in the condensation process and copolymerization of IPA and DAB to PBI.

3.2 PA doping, leaching, and solubility
PBI is a heterocyclic aromatic hydrocarbon polymer with a proton conductivity value of 1 × 10^-4 S/cm. Therefore, the pristine PBI is not suitable for the PEM manufacturing process. After all, PA doping of PBI is necessary to improve proton conductivity. Moreover, PA act as a proton-conducting carrier, a hygroscopic compound, exhibiting proton conductivity in both anhydrous and hydrated states and does not require water for proton conduction in the composites [30]. Fig. 2 shows the PA uptake results of PBI and PBI/SrTiO₃ nanocomposite membranes. The incorporation of the perovskite SrTiO₃ nanoparticles with different wt% in the polymer electrolytes improved the ability to trap PA and thus enhanced the proton conductivity of the composite membranes. Therefore, all PBI/SrTiO₃ composite membranes showed high PA uptake and proton conductivity. Fig. 2 compares the acid doping percentage of pristine PBI and PBI/SrTiO₃-X (PBI < PBI/SrTiO₃-2 < PBI/SrTiO₃-4 < PBI/SrTiO₃-6 < PBI/SrTiO₃-8). However, the further addition of PBI along with SrTiO₃-10 markedly decreased the PA uptake value due to particle agglomeration and higher loading [31]. These results indicated that the amount of PA doping level is an important characteristic of proton conductivity.
The homogeneous and fine dispersion of nanoparticles in the polymer matrix increases the interface between the polymer and inorganic components even at high temperatures, effectively retaining the acid, thereby increasing the proton conductivity. The PA-doped polymer nanocomposite membranes were immersed in hot DI water for 3 h. The remained acids are related similarly to those described above. As shown in Fig. 3, the uniform homogeneous dispersion of SrTiO₃ in the polymer composites increases the interface between the proton-conducting polymer and perovskite nanoparticles even at high temperatures, which helps to effectively retain the acid and increase the proton conductivity [32]. The acid leaching test results in Fig. 3 show that the 8wt% SrTiO₃ membrane provides a high percentage of retention. Solubility is an initial checking process in the post-polymerization process. The pristine PBI (3 mg) was dissolved in 1 mL of various solvents such as THF, DMF, NMP, and DMAc. Yellow color transparent and flexible PBI and PBI/SrTiO₃ composite membranes were obtained from DMAc solvent.

3.3 Morphological study
Fig. 4 shows the 3 dimensional AFM images for pristine PBI and PBI/SrTiO₃-8 composite membrane. Fig. 4(A) shows the pristine PBI membrane has a uniform surface and no pores and defects observed in the image. It is due to the corresponding hydrophobic PBI matrix. Fig. 4(B) shows more roughness compared to pristine PBI. The average roughness (R_a) of the PBI/SrTiO₃-8 composite was calculated by AFM analysis, which indicates a quantitative idea of the difference in surface morphology. The R_a value of the PBI/SrTiO₃-8 composite membrane was found to be 25.13 nm. This indicates that the higher roughness of the PBI/SrTiO₃-8 composite membrane can enhance the uptake of PA molecules and the proton conductivity of the composite membrane because of its high surface area.

3.4 Thermal stability
Thermal stability of the composite membranes was performed using TGA analysis at a heating rate of 10^oC/min in an N₂ atmosphere. The TGA curve of prepared composite membranes is shown in Fig. 5. It can be seen that the PA doped PBI/SrTiO₃-8 nanocomposite membrane exhibits high thermal stability compared to other nanocomposite membranes. The initial weight loss between 100~150^oC occurs due to the absorbed OH group molecules in the polymer composites [33]. The second weight-loss occurring at 160~200^oC is owing to thermal changes in the PA to form pyrophosphoric and triphosporic acids as shown in below following equations [34].



The third major weight loss that occurs at above 600^oC is due to the main polymer degradation [35]. Since the incorporation of SrTiO₃ nanoparticles into the PBI polymer matrix increases the thermal stability of composite membrane, the obtained results explain that TGA shows high thermal properties of the PBI/SrTiO₃-8 nanocomposite membrane for HT-PEMFC.

3.5 AC impedance
The Nyquist plots of PA doped PBI with various concentrations of SrTiO₃ at 160^oC are shown in Fig. 6. From Fig. 6, x-axis Z’ is denoted as real part and the imaginary part is –Z’’ the bulk resistance R can be taken from intercept on the x-axis. Javan bakht et al. reported the proton conductivity of pristine PA-PBI was measured AC impedance method and the proton conductivity value of 5.12 × 10^-3 S/cm at 160^oC [25]. Proton conductivities were calculated from the bulk resistance of the composite membranes of Fig. 6 using equation (2). The maximum proton conductivity of the nanocomposite membrane was obtained at 7.95 × 10^-2 S/cm at 160^oC for the PBI/SrTiO₃-8 sample. It involves protons exchange between phosphoric acid, H₂PO^4- or HPO₄^2- and PBI. Additionally, strong acids can form polymeric complexes due to hydrogen bonding interactions or acid-base interactions between the imidazole group of PBI and the acid molecules from the oxygen to the nearest neighbor ion [36,37]. Moreover, the sites corresponding to the ABO₃ perovskite structure correspond to A-cation at the vertices of the unit cell, B-site cation in the center, and oxygen ion at the face of the cubic structure [38,39]. In particular, the cubic SrTiO₃ perovskite structure exhibits proton conductivity at high temperatures [40]. The proton conductivity of the nanocomposite increases with an increasing amount of SrTiO₃ nanofiller at 160^oC. This enhancement of the proton conductivity of the PBI/SrTiO₃-8 sample is attributed to the hygroscopic properties of SrTiO₃ particles in the membrane. However, the conductivity value of the PBI/SrTiO₃-10 polymer composite membrane is decreased because some of the amino groups in PBI chains are blocked by SrTiO₃, which reduces the binding ability of the PBI chains to the PA groups.

The present study demonstrated the synthesis of PA doped PBI/SrTiO₃–X nanocomposite membranes and their characterization via NMR, physicochemical properties, TGA, AFM, and proton conductivity. The synergetic combination of the organic and inorganic phases of PBI/SrTiO₃ resulted in significant improvements. The highest PA uptake of 120% was observed for PBI/SrTiO₃-8 polymer electrolytes. TGA analysis revealed that these PBI nanocomposite membranes can stand temperatures of up to 200^oC. The R_a value of the nanocomposite polymer electrolyte over the scanned area is around 25.13 nm. The highest proton conductivity was obtained at 7.95 × 10^-2 S/cm at 160^oC for PBI/SrTiO₃-8 polymer composite membrane. These results show that the acid absorption and proton conductivity of polymer composite membranes were higher than that of the pristine PBI membrane. These PBI/SrTiO₃ based composite membranes are promising membranes for high-temperature PEMFC applications.

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and Future Planning (NRF-2020R1A2B5B01001458). This work was supported by grants from the Medical Research Center Program (NRF‐2017R1A5A2015061) through the National Research Foundation (NRF), which is funded by the Korean government (MSIP).