High temperature thermoplastic based on polybenzimidazole and silica nanocomposites

TRABAJO FIN DE MÁSTER Máster en Ingeniería Química HIGH TEMPERATURE THERMOPLASTIC BASED ON POLYBENZIMIDAZOLE AND SILICA NANOCOMPOSITES Memoria y Anexos Autora: Laura Crespo Bartolomé Directora: Dr. Elaine Armelin Diggroc Convocatòria: Junio 2018 Laura Crespo Bartolomé Memoria TFM ii Resum D’entre els polímers termoplàstics, el polibenzimidazol (PBI), abreviatura de poli [2,2’-(m-fenileno)- 5,5’-bisbenzimidazol], és conegut per la seva estabilitat tèrmica i química al estar sotmès a altes temperatures. En general, es fa servir en la fabricació d’eines protectores d'alt rendiment, com ara fibres sintètiques per equips de bombers, roba d’astronautes i soldadors i parets d’avions. No obstant això, la aplicació més recent del polímer PBI és en forma de membrana per a cel·les de combustible de membrana d’electròlit polimèric d’altes temperatures, conegut com HT-PEMFC, per les seves sigles en anglès. Les membranes de PBI han de dopar-se amb àcids inorgànics, com ara l’àcid fosfòric (PA), per tal de millorar les seves propietats conductives de manera substancial. Però, en dopar les membranes, la seva integritat física empitjora, esdevenint un dels seus majors inconvenients per a la aplicació del PBI dopat en dispositius tipus PEMFC. La incorporació de materials inorgànics, com ara diòxid de titani, sílice o argila, en membranes d’intercanvi de protons (PEM) millora substancialment les seves propietats mecàniques i tèrmiques, segons estudis previs; i, per tant, la capacitat de la membrana per aplicacions en piles de combustible d’alta temperatura. En el present estudi, les membranes de PBI han estat preparades per colada de dissolvent després de la incorporació de nano-partícules de polietilenimina (PEI) i sílice químicament modificades (NP Si-PEI). L’ús de PEI (Mw ≈ 800 g/mol) com a agent d’acoplament de la sílice es fa per tal de millorar les interaccions interfacials entre l’entorn orgànic i inorgànic de la matriu del polímer. En aquest projecte es presenta una descripció detallada de la síntesi i els mètodes utilitzats per a la caracterització de les noves membranes. Els resultats demostren que les membranes PBI-Si-PEI modificades químicament presenten millors propietats que les PBI no modificades . A més, les propietats elèctriques també es veuen millorades un cop s’ha dopat la membrana amb l’àcid fosfòric. Paral·lelament, per tal de incrementar la durabilitat i les propietats mecàniques de la membrana de PBI, s’ha realitzat un estudi breu de la incorporació d’un àcid dopant diferent al PA, l’àcid dodecilbenzè sulfònic (DBSA), i es va avaluar les seves propietats conductives. Aquest àcid dopant es, a més a més, un agent plastificant, degut a l’elevat nombre de grups metilens (-CH2) a la seva estructura. High temperature thermoplastic based on polybenzimidazole and silica nanocomposites i Resumen Entre los polímeros termoplásticos, el polibenzimidazol (PBI), abreviatura del poli [2,2’-(m-fenileno)- 5,5’-bisbenzimidazol], es conocido por su excepcional estabilidad térmica y química al ser sometido a altas temperaturas. Por lo general, se emplea para fabricar herramientas protectoras de alto rendimiento como, por ejemplo, fibras sintéticas para equipos de bomberos, trajes de astronautas y soldadores, así como en paredes de aviones. Sin embargo, la aplicación más reciente del polímero PBI es en forma de membrana para celdas de combustible de membrana de electrolito polimérico de alta temperatura, conocido como HT-PEMFC, por sus siglas en inglés. Las membranas de PBI deben doparse con ácidos inorgánicos, como el ácido fosfórico (PA), para mejorar de manera sustancial sus propiedades conductivas. Sin embargo, al dopar las membranas, su integridad mecánica empeora, representando uno de sus mayores inconvenientes para la aplicación del PBI dopado en dispositivos tipo PEMFC. La incorporación de materiales inorgánicos, tales como dióxido de titanio, sílice o arcilla, en membranas de intercambio de protones (PEM) mejora sustancialmente sus propiedades mecánicas y térmicas, según estudios previos; y, por lo tanto, la capacidad de la membrana para aplicaciones en pilas de combustible de alta temperatura. En el presente estudio, las membranas de PBI se prepararon por colada de disolvente después de la incorporación de nanopartículas de polietilenimina (PEI) y sílice químicamente modificadas (NP Si-PEI). El uso de PEI (Mw ≈ 800 g/mol) como agente de acoplamiento de la sílice va dirigido a mejorar las interacciones interfaciales entre el entorno orgánico e inorgánico de la matriz del polímero. En el presente proyecto se presenta una descripción detallada de la síntesis y los métodos utilizados para la caracterización de las nuevas membranas. Los resultados demuestran que las membranas PBI-Si-PEI modificadas químicamente tienen mejores propiedades que las PBI no modificadas. Además, las propiedades eléctricas también se ven mejoradas una vez se ha dopado la membrana con el ácido fosfórico. Paralelamente, con el objetivo de incrementar la durabilidad y las propiedades mecánicas de la membrana de PBI, se ha realizado un estudio breve de la incorporación de un ácido dopante diferente del PA, el ácido dodecilbenceno sulfónico (DBSA), y se evaluaron sus propiedades conductivas. Este ácido dopante es, además, un agente plastificante, debido al elevado número de metilenos (-CH2) en su estructura. Laura Crespo Bartolomé Memoria TFM ii Abstract Among thermoplastics polymers, polybenzimidazole (PBI), short for poly [2, 2’-(m-phenylene)-5,5’- bisbenzimidazole], is known for its exceptional thermal and chemical stability at elevated temperature. It is usually employed to fabricate high-performance protective tools such as synthetic fibers for firefighter’s gear, astronaut and welder’s suits, as well as in aircraft walls, as for examples. However, the most recent application of PBI polymer is as membrane in high-temperature polymer electrolyte membrane fuel cell (HT-PEMFC). PBI membranes must be doped with inorganic acids, like phosphoric acid (PA), to substantially improve their conductivity properties. Nevertheless, when doped, its mechanical integrity gets much worse, representing one of their major handicaps for the application of doped PBI in PEMFC devices. The addition of inorganic materials, such as titanium dioxide, silica or clay, into proton exchange membranes (PEM) is known to substantially improve their mechanical and thermal properties, as reported in previous studies; and, hence the capability of the membrane for high temperature fuel cell applications. In the present study, PBI membranes were prepared by solvent casting after the incorporation of chemically modified polyethyleneimine-silica nanoparticles (Si-PEI NPs). The use of PEI (Mw ≈ 800 g/mol) as silica coupling agent is intended to improve the interfacial interactions between the organic-inorganic surrounding in the polymer matrix. A detailed description of the synthesis and the methods used for the characterization of the new membranes is reported. Results demonstrated that chemically modified PBI-Si-PEI membranes have better properties than the pristine PBI films. Moreover, the conductive properties were also enhanced after the doping process with the PA molecules. Additionally, to the study above mentioned, another dopant molecule composed by dodecylbenzenesulfonic acid (DBSA) was investigated. It consisted in a brief study where the conductive properties were also approached. The DBSA molecules perform as dopant and plasticizer, due to the high number of methylene groups inside its chemical structure. High temperature thermoplastic based on polybenzimidazole and silica nanocomposites iii Acknowledgements First, I would like to thank my tutor, Dr. Elaine Armelin Diggroc, for giving me the opportunity to conduct this research, and for her dedication. I would also like to thank the dedication and daily support in the laboratory of Ms. Brenda Molina. I also would like to acknowledge Mr. Amir Aiman Bin Tahrim (University of Kuala Lumpur, Malaysia) for his help with the PBI synthesis and nanoparticles characterization; as well as Dr. Lourdes Franco and Mr. Francisco José Calvo, for their help with the thermal analyses. This work was supported by MINECO (MAT2015-69367-R) and the Agència de Gestió d'Ajuts Universitaris i de Recerca (2017SGR359). Laura Crespo Bartolomé Memoria TFM iv High temperature thermoplastic based on polybenzimidazole and silica nanocomposites v Glossary Al2O3: aluminion oxide AMS: amine functionalized silane APTES: 3-aminopropyltriethoxysilane APU: auxiliary power unit ASMs: acidic surfactant like molecules CHP: combined heat and power CO2: carbon dioxide CO: carbon oxide DBSA: Dodecylbenzenesulfonic acid DLS: dinamic light scattering EIS: electrochemical impedance spectroscopy EtOH: ethanol FA: formic acid FTIR: fourier-transform infrared spectroscopy HT-PEMFC: high-temperature polymer electrolyte membrane fuel cell ILs: ionic liquids LAMS: long chain amine functionalized modified silica OBA: 4, 4’-oxybis (benzoic acid) OPBI: poly (4,4’-diophenylether-5,5’-bibenzimidazole) ORR: oxygen reduction reaction PA: phosphoric acid Memoria 4 Among the variety of PEMFCs, the high temperature proton exchange membrane fuel cells (HTPEMFCs) are proficient clean energy conversion devices for automotive and stationary applications. HT-PEMFC could mitigate the CO poisoning, humidity and heat management, and sluggish of oxygen reduction reaction (ORR). As mention above, acid doped polybenzimidazoles (PBIs)/functionalized PBIs polymer electrolyte membranes are familiar uses for HT-PEMFC because of high proton conductivity with thermo-mechanical stability. Proton conductivity of PBI membranes is greatly promising by acid doping dimension and cell operating temperature.(9) Figure 2. 2. Proton conductivity mechanism of PA-doped PBI membrane (a) PA-water proton transfer; (b) proton transfer between PA molecules; (c) benzimidazole ring-PA proton transfer (10) PBI membrane needs to be doped with acid in order to facilitate the movement of protons across the membrane. The movement of proton in the acid doped PBI membrane are commonly through Grotthus mechanism (Figure 2.2). Due to the present of benzimidazole rings in the chemical structure, PBI membranes can be doped with different kind of acids. The choice of acid used affects the maximum proton conductivity properties. Savadogo & Xing (10) show that different effect of acid used to dope PBI membrane gives different level of photon conductivity. The proton conductivity changes as follow: H2SO4 > H3PO4 > HClO4 >HNO3 >HCl. Doping PBI with sulphuric acid gives the best proton conductivity however, phosphoric acid are commonly used nowadays are due to sulphuric acid doped membrane still required humidification and phosphoric acid doped membrane are able to operate under anhydrous condition (11). Although increased number of reports in recent years on proton exchange membrane (PEM) developed from nanocomposites of polybenzimidazole (PBI) with inorganic fillers brought hope to end the saga of contradiction between proton conductivity and variety of stabilities, such as mechanical, thermal, chemical, etc.; it still remains a prime challenge to develop a highly conducting PEM with superior aforementioned stabilities. In fact, the very limited understanding of the High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 5 interactions, especially interfacial interaction between PBI and inorganic filler, leads to confusion over the choice of inorganic filler type and their surface functionalities.(12) Taking clue from T. Jana’s early study (12) based on poly (4,4’-diphenylether-5,5’-bibenzimidazole) (OPBI)/silica nanocomposites, where silica nanoparticles modified with short chain amine showed interfacial interaction-dependent properties, in that work they explored the possibility of enhanced interfacial interaction and control over the interface by optimizing the chemistry of the silica surface. 2.1. High temperature thermoplastic: polybenzimidazole (PBI) Polybenzimidazoles are a class of extremely heat-resistant heterocyclic thermoplastics. They are prepared from an aromatic tetraamine and an aromatic dicarboxylic acid or a derivative of it (Figure 2.3). A prominent example is the condensation reaction of diphenyl isophthalate and 3,3’,4,4’- tetraaminodiphenyl which undergo spontaneous cyclization at temperatures around 350 to 400 °C in an inert atmosphere.(13) Figure 2. 3.Chemical repeating unit of PBI (13) PBI is an amorphous thermoplastic polymer with a glass transition temperature of 425-4360C and no melting point, and with heat deflection temperature at 1.8 MPa is about 435°C. It has a good chemical resistance and excellent textile fibre properties (14). Its decomposition temperature is more than 500°C (13). In form of a membrane, PBI has received much attention mainly for use in blood dialysis and reverse osmosis at high temperature and in harsh environment. (15) Chemically, PBI is a basic polymer and can readily react with acids. As a result, various inorganic acids were investigated such as H2SO4, H3PO4, HClO4, HNO3, HBr, HCl, organic acids like CH3SO3H, C2 H5SO3H and aromatic phosphoric acids, as well as polymeric acids. (14) Among the doping acids, phosphoric acid and sulphuric acid were found to give high conductivity. The acids act both as donors and acceptors in proton transfer and therefore allow for proton migration along the anionic chain. H3PO4 is an interesting acid due to its conductivity and thermal stability at temperatures up to 2000C. The PBI cells have been operated at temperatures up to 2000C without humidification of the reactant gases. (9)(14) Memoria 6 Thus, the acid doping level is an important parameter that will affect the performance and durability of the PEMFCs. The acid doping level is defined as the mole number of phosphoric acid per repeat unit of PBI and was found to be of special importance for proton conductivities and mechanical properties (12)(14)(16). The performance of a membrane is dependent on proton conductivity, as mention above, which in turn often depends on its water content. High proton conductivity is supported by high level of water uptake; at the same time, it is also a sign of low-dimensional stability as water influences the polymer microstructure and mechanical properties. Since water is also known to assist the mass transport of methanol and oxygen through the membrane, the water uptake measurements could serve as a quantitative measure of membrane performance for DMFC application as well. Gravimetric technique has been widely used for this purpose. Water uptake measurements are generally done by double weighing. ‘Wet’ weights of the membranes are first measured after equilibrating with water at different temperatures or upon exposure to water vapour at various pressures. The membrane samples are then dried at a temperature above the boiling point of water for a particular period of time and their ‘dry’ weight is measured. (17) Ion exchange capacity is the measure of relative concentration of acid groups within polymer electrolyte membranes. Proton conductivity and water uptake both rely heavily on the concentration of ion conducting units in the polymer membrane. The ion content is characterized by the mass of dry membrane per molar equivalent of ion conductor. Varying the ion content of the membrane can control both its water uptake and conductivity. While it is desirable to maximize the conductivity of membrane by increasing its ion content (decreasing equivalent weight), other physical properties must be considered. Too many ionic groups will cause the membrane to swell excessively with water, which compromises mechanical integrity and durability.(17) Other properties can be taken into account for a more detailed characterization of the PBI membranes, among them: gas and liquid permeability, durability tests, mechanical strength, etc...are some examples. However, in order not to extend too much the introduction, more detailed description of such properties and their complete characterization techniques can be access in some reviews published in the literature.(17)(18) 2.2. Applications of polybenzimidazole (PBI) membranes Polybenzimidazoles are known for their high-strength and high-temperature performance, as explained before. They find applications in many industrial fields including semiconductor, petrochemical and aerospace industries. Major applications include heat resistant apparels, contact High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 7 seals, wafer carriers, membranes for various separation processes, insulator bushings, and thermal isolators.(13) Figure 2. 4. Accumulative published items of the topic “polybenzimidazole” and “fuel cells” indexed within Web of Science (18) One of the main applications of PBI membranes is its use in fuel cells. As seen in the Figure 2.4, there has been an increase on the number of publications that mention both the topic “polybenzimidazole” and “fuel cells”. From a technological application point of view, PBI-based fuel cells seem most suited for stationary power applications, for example, based on the natural gas reforming and combined heat and power (CHP) generation. For automobile applications more challenges exist by considering startup time and thermal/load cycling, however, an auxiliary power unit (APU)-like system is of special interest. Volkswagen has presented at the 2007 Los Angeles Motor Show a hybrid concept car with a PBI stack as a charger for batteries to extend the driving range. As small power units, the PBIbased cells have the potential to integrate with a simplified methanol reformer or metal hydride tank (18). In addition, this proton conducting polymer electrolyte opens the door for many other electrochemical applications that could benefit from or require higher temperatures such as hydrogen gas pumping and purification (19), electrochemical sensors and water electrolysis.(18) Subianto (11) did mention that poly(4,4-diphenyl ether-5,5-bibenzimidazole) or OPBI membrane is one of the PBI derivatives which have excellent properties to be use as membrane in HT-PEMFC. It has gained popularity nowadays (8)(12)(16). OPBI membrane has a readily tuneable molecular weight as well as relatively more flexible backbone due to presence of ether linkage (-O-) in the backbone (Figure 2.5) compared to other PBI structures (Figure 2.3) (8). Introduction of ether linkages may facilitate the entanglement between molecular chains thus forming intertwined polymer networks. The intertwined structured provide a ‘sponge like’ microstructure which provide “free volume” to hold excess acid molecules (20). OPBI membrane also has a better solubility in low boiling point 0 40 80 120 160 200 Published items Year Memoria 8 volatile solvents such as formic acid, employed for its synthesis. In order to improve the membrane durability and the proton conductivity, silica is added into OPBI films to obtain new composite membranes with the aid of silane coupling agents. Figure 2. 5. Molecular structure of poly (4, 4’-diphenylether-5, 5’.bibenzimidazole) (OPBI) 2.3. Polybenzimidazole (PBI) composites Various strategies have been applied to improve the stability and conductivity of PBI membranes including optimization of membrane fabrication techniques, crosslinking of polymer backbone, blending with other polymers, and forming composite structure by incorporation of various inorganic acids. However, the improvements brought by such modifications have been found rather limited and the PBI-based membranes remained demonstrating weakness in mechanical strength when highly loaded with PA and poor endurance when tested for long term.(20)(21) In order to solve this problem, modification of the PBI matrix by addition of inorganic filler is required. Inorganic fillers are typically added to increase the proton conductivity and/or acid uptake of PBI membranes. The combine of hydrophilic inorganic nanomaterial with a PBI creates favorable results, because of their affinity to interact with the water and acid and turn out to be more hydrophilic. Most types of HT-PEMFC membranes have successfully been doped with inorganic filler such as hygroscopic oxides (SiO2, TiO2, ZrO2, Al2O3); montmorillonite, clays, heteropolyacids and zirconium phosphates (ZrP) (22) and zeolites (17). This project focuses on the inorganic fillers and acid/plasticizer component for the obtaining of OPBI composite membranes. Ceramic nanoscale and mesoscale fillers such as titania and zirconia have been used to design composite PA doped PBI-based membranes with appealing properties. Particularly, ZrO2 fillers were recently tested for enhancing the conductivity and stability of PA doped PBI membranes. However, the use of nanoscale ZrO2 filler and the accompanied membrane casting is challenged by agglomeration and precipitation leading to drastic phase inhomogeneity, despite the improvement in the hydromechanical properties by the reinforcement.(21) Among different inorganic additives, SiO2 is well known for its barrier property towards gases and solvents along with its strong H-bonding capability to acids, which is beneficial for the prevention of the acid leaching and improve the proton conductivity of the resulting hybrid membranes compared High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 9 to the virgin membrane (22). It is also reported that the introduction of the silica particles into the PBI matrix improved the thermal stability (22) as well as mechanical properties of the membranes that are important features for HT-PEMFC application. Figure 2. 6. Schematic structure: How silane coupling agent works. The main drawback for SiO2 filler is its insulating capability, for this reason, a low percentage of this compound is suggested to be added to the PBI matrix in order to ensure the conductivity of the membrane. Less than 2,0 wt% was used in this study. As stated earlier, addition of SiO2 did reduce the proton conductivity (23)(24) (reduction about 22 – 28% of conductivity compared to pure-acid doped PBI membrane). In order to overcome this problem, several studies have tried to solve this problem in the past by means of employing different kinds of amine silane coupling agents: (3-aminopropyltriethoxysilane (APTES) and N-(3trimethoxysilylpropyl) diethylenetriamine (TMSPDT) (Figure 2.6) (12)(25). Most of them addressed the issue to improve the proton conductivity without compromising the mechanical stability of the membrane(25). By using amine functionalized (AMS) silane coupling agent, it was found out in the characterization of structure and evaluation of crystallinity of the composite membrane, AMS/OPBI shows several peaks in the wide angle X-ray diffraction (WAXD) patterns while unmodified silica (UMS)/OPBI doesn’t show similar trend. The crystalline peaks indicate an ordered structure due to the fact that the silica particles are highly dispersed in the OPBI matrix with the help of –NH2 moieties of AMS and imidazole moieties on the OPBI. As for thermal stabilities, initial weight loss (20% loss) of composite begin at temperature 191oC for 15 wt% UMS/OPBI, while for 15 wt% AMS/OPBI begin at temperature 2260C and; at a temperature of 5100C, 15 wt% UMS/OPBI show a residual weight loss of 69,46% and 77,99% for 15 wt% AMS/OPBI. It is suggested that a well self-assembled cluster in the nanocomposite membrane is responsible for the improvement in the thermal stabilities(25). On the hand, as for proton conductivity; 20 wt% Coupling agent Inorganic material Organic polymer Memoria 10 AMS/OPBI shows the highest proton conductivity which is 1.23 x 10-1 S·cm-1. As for 15 wt% AMS/OPBI, it has the higher proton conductivity compared to 15 wt% UMS/OPBI which is around 1.23 x 10-1 S·cm-1 where 15 wt% UMS/OPBI has around 0.07 x 10-1 S·cm-1. It was concluded that: morphology of OPBI/silica membrane is dependent upon the functionalized groups of silica surface, amine modified silica (AMS) helps to produce self-assembled clusters formation and unmodified silica (UMS) shows a well-dispersed formation (25). The thermal stabilities and proton conductivity are also improved with the addition of the filler. In the study conducted by Singha et. al. (11), a long amine silica agent was used to functionalize the surface of silica nanoparticle before it was incorporated into OPBI/silica nanocomposite membrane. The long chain amine functionalized modified silica (LAMS) was used because they anticipated improved membrane properties that might arise because of the increase in the chain length and stronger multiple point interaction between three amine groups on the surface of silica and the polymer matrix. In the WAXD patters, crystalline peaks also formed but at the higher concentration (7 wt% of LAMS/OBPI and above). Compared to the previous study described, formation of selfassembled cluster is well defined as they suggested that due to the stronger interaction between three amine functionality and OPBI matrix (25). As for the thermal stabilities, it was reported that the thermal stability increased as the loading of the silica increased (at 15 wt% LAMS/OBPI shows the lesser mass loss (4% weight loss) compare to neat OBPI (7% weight loss) while 15 wt% UMS/OPBI does not show any significant improvement.). In the proton conductivity test, which was conducted at 1600C, the results showed that the conductivity for 15 wt% LAMS/OPBI was 0,181 S·cm-1, while pristine OPBI was 0,118 S·cm-1. However, another study, conducted by Ghosh et al. (26), showed that the conductivity increased as the filler loading increased because the PA doping level was also enhanced. On the other hand, other similar analysis, conducted by Singha et al. (12), showed a different result. Chart 2.1 shows the comparison data between these two studies. They also suggested that the three-amine group on the silica formed hydrogen bonding with OPBI chain to produce crystallites all over the matrix. This crystal formation provides a proton-conducting channel in the matrix that helps an efficient and ease proton hopping across the membrane. They also highlighted that bare silica particles did not create protonconducting pathways, but creates a barrier that cause tortuosity of the diffusive path for the incoming protons. In other words, the proton is forced to wiggle around the particles travelling through a tortuous path, which results in the decreasing of the proton conductivity. High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 11 Chart 2. 1. Comparison of proton conductivity (at 1600C) and PA doping level (mol/RU)a) of AMS/OPBI and LAMS/OPBI nanocomposite membrane. Nanofiller loading (wt%) Proton conductivity at 1600C (S·cm-1) PA doping level (mol/RU) AMS/OPBI composite LAMS/OPBI composite AMS/OPBI composite LAMS/OPBI composite 4 0,062 0,157 21,75 26,7 7 0,073 0,159 22,75 25,3 10 0,093 0,176 24,43 24,3 Note: a) RU: repeat unit By modifying the surface of silica nanoparticle with amino (-NH2) silane agent, there were improvements in terms of better interfacial interaction and better proton conductivity compared to non-modified particle. Both studies concluded that due to the increase in the number of hydrogen bonding between the amino moieties (-NH2) of silane agent and imidazole group of OBPI membrane which create proton conducting channels which are responsible for higher proton conductivity. Hence, it was proposed that the usage of silane coupling agent with different aliphatic amine chain (Figure 2.7), which has the same alkoxy group (methoxy; -OCH3), towards the properties of composite membrane, especially for the HT-PEMFC application. b) a) c) Figure 2. 7. Types of amine silane coupling agent: (a) (3-aminopropyl) trimethoxysilane (APTMS); (b) (3-trimethoxysilylpropyl) diethylenetriamine (TMSPDT); (c) trimethoxysilylpropyl modified polyethyleneimine (TMS-PEI) Memoria 12 2.4. Polybenzimidazole (PBI) composites with plasticizers One of the main problems associated with polybenzimidazoles seems to be their poor solvent solubility, which results from the high degree of molecular rigidity in the backbone and the strong intra-interchain hidrogen bonding interactions, coming in the way of development of new applications (27), as well as the dehydration of phosphoric acid observed at temperatures above 1800C, which results in serious conductivity decrease (28). Attempts have been made to modify the properties of PBI by incorporating groups that make the polymer chain more flexible or other bulky units into the main or side chain to overcome the restricted polymer solubility. Recently, ionic liquids (ILs) have attracted considerable attention in electrochemical applications. ILs are liquid molten salts at room temperature. They are composed of ions and are non-volatile substances. It is also known that ILs can be highly thermal stable at a wide temperature range. Thus, they could be good candidates for high-temperature PEMFCs (28). An example of ionic liquids that has been reported is 1-hexyl-3-methylimidazolium trifluoromethanesulfonate (HMI-Tf), its use lead to the improvement of the thermal stability and its function as a plasticizer and ion carrier enhance the conductivity of the PBI system significantly, at high temperatures. (28) Aiming to improve solubility and proton conductivity, a novel pyridine-containing polybenzimidazole (PDA-PBI) membrane, and incorporating ether groups, was developed. It is reported the polycondensation of 4, 4’-(pyridine-2, 6-diyl-bis (oxy)) dibenzoic acid (PDA) and 3, 3’- diaminobenzidine (DAB) in phosphorous pentoxide/methanesulfonic acid (PPMA). The imparting of ether groups aims to boost the polymer solubility and flexibility. The introduction of pyridine ring is designed to increase the density of basic sites favouring the acid doping and hence improvement in proton conductivity. (27) Furthermore, over the past few years PBI composites have been prepared with acidic surfactant like molecules (ASMs), with the purpose of improving PEM properties especially proton conductivity. Composites are obtained by homogenizing poly (4, 4’-diphenylether-5, 5’-bibenzimidazole) (OPBI) in three different ASMs namely camphorsulfonic acid (CSA), p-toluenesulfonic acid (PTSA) and mono-ndodecyl phosphate (MDP). Mechanical reinforcement was observed, in case of composite membranes, due to an increase on the storage modulus with increasing ASM loading (29). High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 13 3. Objectives The objectives of the present project are: • To enhance the membrane proton conductivity by fabricating a PBI/silica functionalizedcomposite membranes for future HT-PEMFC application. • To improve the surface interface between organic/inorganic composite membrane by applying silane coupling agent. • To improve mechanical properties of the membrane by adding plasticizers. • To study the effect of different coupling agents towards the performance of HT-PEMFC application by measuring their conductivities. Memoria 20 (gas, liquid or solid). The technique is based on the procurement of the radiation dispersed in an inelastic way by a material when a beam of monochromatic light impinges on it. The phenomenon that occurs, with characteristics inherent to the molecular composition and crystal structure of the irradiated sample, is called the Raman effect. (31) RAMAN spectra of the samples studied here were obtained through Renishaw InVia confocal Raman microscope (Figure 4.5), at 785 nm of laser excitation, with an exposure time of 1 s, a laser power of 0.1%, 3 accumulations and in the Raman shift range of 600-1800 cm-1. The samples were evaluated in film shape with a size of 1,5 cm x 7 mm, approximately. Figure 4. 5. Renishaw InVia confocal RAMAN microscope High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 21 Scattering is a general physical process in which some forms of radiation changes its direction and, on occasion, its frequency after interacting with the medium. In light scattering process when a light beam incises in a determinate frame the total number of photons is unaltered, but the number going in forward direction decreases due to redirection of light from scattering interaction. In other words, when light impinges on matter, the electric field of the light induces an oscillating polarization of electrons in the molecules and these molecules provide a secondary source of light and subsequently scattering light. The frequency shifts, the angular distribution, the polarization and the intensity of scattered light are determined by the size, shape and molecular interaction in the scattering material. (32) Dynamic light scattering (DLS) NanoBrook Omni Zeta Potential Analyzer (from Brookheaven Instruments) (Figure 4.6) was employed for particle size measurements using 10 mg/L of Si-PEI dilute aqueous solutions. The samples were in the form of dispersed particles in liquid solution. Figure 4. 6. DLS NanoBrook Omni Zeta Potential Analyzer Memoria 22 4.5.2. Microscopy analysis The scanning electron microscope is a powerful instrument that permits the observation and characterization of heterogeneous organic and inorganic materials and surfaces on a specific scale. The area to be examined is irradiated with a finely focused electron beam, producing secondary and backscattered electrons signals that can be used to examine many characteristics of the sample: composition, surface, topography, crystallography, and others.(33) Inspection of the cryo-fractured surfaces of OPBI films, employing liquid nitrogen, was conducted by scanning electron microscopy (SEM) using a Focus Ion Beam (FIB) Zeiss Neon 40 instrument (Carl Zeiss, Germany) (Figure 4.7) operating at 5kV, equipped with an EDX spectroscopy system. Films were covered with conductive carbon coating, by using Mitec K950 Sputter Coater, before SEM analysis. The samples were used in film shape, with a size of 3x4 mm2, approximately, and varied thicknesses. Figure 4. 7. Focus Ion Beam (FIB) Zeiss Neon 40 instrument High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 23 4.5.3. Wettability When a gas and liquid, separated by their common interface, encounter a solid surface, the contact line between the three phases is called the common line (CL), the contact angle (θ) is the one between the liquid-solid, and liquid-gas interfaces, Figure 4.9. If θ is lower than 900, the surface is called hydrophilic, whereas if θ is greater than 900, the liquid is non-wetting, so the surface is called hydrophobic.(34) Figure 4. 8. Static contact angle measurement (34) Contact angle measurements were performed by employing OCA 20 (DataPhysics Instruments GmbH, Filderstadt) equipment (Figures 4.10 and 4.11) and using the sessile drop method at room temperature. For the static contact angle (sCA) measurements, 0,500 μL droplets of distilled water were dispensed on the respective surfaces. The contact angle values (software SCA 20) were obtained as the average of ten independent measures for each sample, which was flat films with 1,5x1,5 cm2 of area. Memoria 24 Figure 4. 9. OCA 20 equipment for the measurement of WCA (a) Figure 4. 10. OCA 20 equipment for the measurement of WCA (b) 4.5.4. Thermal stability Thermogravimetric analysis (TGA) is a method of thermal analysis in which the mass of a sample is measured over time as the temperature changes. This measurement provides information about physical phenomena, such as phase transitions, absorption and desorption; as well as chemical phenomena including chemisorptions, thermal decomposition, and solid-gas reactions. (35) Thermogravimetric analysis (TGA) was carried out with a DTA1600 thermogravimetric analyser of TA Instruments (Figures 4.12 and 4.13) at a heating rate of 100C/min, under argon atmosphere, from 200C to 11000C. The polymer mass required for this experiment was at about 15 mg in powder shape. High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 25 Figure 4. 11. Detail of the thermogravimetric analyser Figure 4. 12. DTA1600 thermogravimetric analyser High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 27 4.5.5. Water uptake, swelling ratio and swelling volume measurements in water and PA For water uptake, swelling ratio and swelling volume measurements in water and acid, the membranes, with variable dimensions (3-24 mm), were first dried thoroughly in a vacuum oven at 1000C for 3 days. Similar sized pieces of membranes were cut in triplicate and their weights, dimensions and thickness noted. They were then immersed in distilled water for 3 days (for measurements in water) and in PA for 3 days (for measurements in acid). After the specified period, the wet membranes were quickly wiped to remove the surface water (or acid) with filter paper and their weights, dimensions and thickness were noted again. Water uptake, swelling ratio and swelling volume were calculated as: %100 W-W =keWater Upta dw x Wd (Eq, 4.1) %100& x L LL watertioinPASwellingRa d dw   (Eq, 4.2) %100& x V VV waterlumeinPASwellingVo dw   (Eq, 4.3) Where Ww,Lw and Vw are the weight, length and volume of the wet membranes, respectively and Wd, Ld and Vd are the weight, length and volume of dry membranes, respectively. These measurements were carried out in triplicate, independently, with three similar sized pieces of the membranes to check for reproducibility, and the average values were calculated. 4.5.6. Proton conductivity measurement One of the most modern impedance analysis systems is the Electrochemical Impedance Spectroscopy or EIS. It is a relatively new and powerful method of characterizing electrical properties of materials and their interfaces with electronically conducting electrodes. It can be used to investigate the dynamics of bound or mobile charge in the bulk of interfacial regions of any kind of solid or liquid material: ionic, semiconducting, and mixed electronic-ionic or dielectrics. Electrochemical impedance is measured via the current through an electrochemical cell, to which a small excitation AC potential is being applied. The mathematical analysis is done using Fourier series. The EIS experiment involves the application of a sinusoidal electrochemical perturbation (potential or current) to the sample that covers a wide range of frequencies. The multi-frequency excitation allows Memoria 28 the measurement of several electrochemical reactions that take place at different rates and the measurement of the capacitance of the electrode. The most common and standard procedure is to measure the impedance applying a single-frequency voltage or current to the interface of the electrode, and measuring phase shift and amplitude (real and imaginary parts) of the resulting current at said single-frequency, using either analogue circuits or fast Fourier transform (FFT) analysis. (36) Electrochemical Impedance Spectroscopy (EIS) measurements were performed using a capacitor cell and an AUTOLAB-302N potentiostat/galvanostat (Figures 4.14-4.16) operating between the frequency range of 104 Hz and 10-2 Hz and 10 mV of amplitude for the sinusoidal voltage. The volume of the cell is defined by a 15 mm inner diameter and a 2 mm inner depth. The pressure is kept constant and defined by the maximum limits of screws used to close the probe arrangement, Figure 4.17. For the assays, the films were previously cut into disks (area = 1.766 cm2); dry film thickness (DFT) values were measured with a Neurtek Mega-Check pocket FE device. (6) Figure 4. 13. Autolab-302N potentiostat Figure 4. 14. Capacitor cell Figure 4. 15. Cell assembly High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 29 Figure 4. 16. Schematic representation of the cell geometry and membrane arrangement used in the present study: (a) 3D open view of the individual parts of the through-plane impedance cell with their corresponding size (1, 5,40 mm; 2, 2,60 mm; 3, 22,10 mm; 4,32,95 mm; 5, 11,15 mm; 6, 12,00 mm; 7, 5,70 mm; 8, 9,40 mm; 9, 15,00 mm; 10, 11,50 mm; 11, 5,70 mm; 12, 7,70 mm; 13, 2,35 mm;14, 5,45 mm.) (b) Configuration of the closed arrangement probe. (c) Equivalent electrical circuit.(6) The acid loaded membrane was cut into a disc shape and mounted onto the in house built conductivity cell. The membranes were dried at 100°C by heating and holding at 100°C isothermally for 2 hours to remove the water from the membrane. The membrane samples were cooled in a vacuum oven and taken out before conductivity measurement in an effort to keep the samples dry. The conductivities of the samples were obtained from the directcurrent potential difference between the two inner electrodes. The conductivity was calculated with the following equation: RBL D   (Eq, 4.1) ( where, σ is the proton conductivity (S/cm), D is the distance between the electrodes, B and L are the thickness and width of the membranes respectively and R is the resistance obtained from Nyquist plots.(16) For the high temperature assays, the samples were left in a vacuum oven for 48h to stabilize before EIS analysis. Memoria 36 weight mass varying between 12,5 wt.% to 50 wt. % was tried. As can be seen in Figure 5.6a, with the highest content, the film was not even been formed, leaving the material in powder shape adhered to the Petri dish. In Figure 5.6b, the film (OPBI/DBSA (37,5 wt.%) was successfully obtained but it was still very fragile. Finally, the optimal content was found to be 12,5 wt. % of DBSA acid. It was possible to create an integral film (Figure 5.6c), with better mechanical properties in an undoped state than the pristine OPBI or the OPBI/Si-PEI membranes. Figure 5. 6. OPBI tests with DBSA plasticizer: a) OPBI/DBSA (50%); b) OPBI/DBSA (37.5 %); c) OPBI/DBSA (12.5 %). Despite the issues raised with the membrane preparation, it was possible to perform all the necessary tests for the total characterization of the films. 5.3. Characterization of the OPBI films During the following section, the most remarkable results of the characterizations carried out are explained, for instance: FTIR and RAMAN spectroscopic analyses, SEM microscopy evaluation, wettability tests, thermogravimetric study, swelling tests and proton conductivity measurements. 5.3.1. Spectroscopy characterization The FTIR spectra of each pair of films: OPBI, OPBI/Si-PEI and OPBI/DBSA (12,5%) (doped and undoped) generated are included in the Figures 5.7-5.9. In the Figure 5.7, we can appreciate a comparison between the OPBI film undoped and the one doped. One of the main characteristic of the infrared from the undoped films is their broad and less intense absorption bands compared with the doped films. Some of the absorption bands in OPBI doped films are very sharp and well-defined, which main peaks correspond to the characteristic band of phosphoric acid between 3000-2500 and between 1200-500 cm-1. High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 37 As reported in previous works, OPBI shows three characteristic peaks at 3413, 3144 and 3066 cm-1 attributed to non-hydrogen bonded free N-H groups, self-associated hydrogen-bonded N-H groups and stretching modes of aromatic C-H groups, respectively (16). Figure 5. 7. Infrared spectra of OPBI undoped and doped films The infrared spectra of OPBI/Si-PEI films is presented in the Figure 5.8. The peaks at 1273, 1169, 1053 and 794 cm-1 could be assigned to the silica nanoparticles, as seen in the Figure 5.3 from previous sections. On keen observation, we see that the intensity of peaks of =C-N-H functionality (at 2300 cm1) decreases in the silica membranes, suggesting the formation of hydrogen bonds between amine groups of polymer and functional groups of the Si-PEI.(12) Memoria 38 Figure 5. 8. Infrared spectra of OPBI/Si-PEI undoped and doped films The infrared spectra of OPBI films with 12,5 wt.% of DBSA molecules is shown in the Figure 5.9. The peaks at 3393, 2916, 1031 and 1006 cm-1 could be assigned to O-H stretching, C-H stretching of –CH2, S=O stretching and >CH stretching of benzoic rings in DBSA molecule, respectively. Some of these peaks were also observed in the doped films but these bands are collapsed with that from phosphoric acid absorption bands, in the wavenumber between 3000-2500 cm-1 and between 1200500 cm-1. High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 39 Figure 5. 9. Infrared spectra of OPBI+DBSA (12,5%) undoped and doped films Raman spectroscopy was employed to confirm the main absorption bands found in the infrared analysis. The RAMAN spectra of each pair of films: OPBI (doped and undoped), OPBI/Si-PEI (doped and undoped) and OPBI/DBSA (12,5%) undoped, are included in the Figures 5.10 - 5.12. The three spectra of the undoped films show very similar absorption bands due to the low amount of silica-PEI nanoparticles (less than 2 wt.%) and plasticizer present in the OPBI/Si-PEI and OPBI/DBSA films, respectively. However, small differences on the late can be found. For this reason, the first discussion will be related to these three samples. The peaks at 1616, 1577 and 1548 cm-1 could be assigned to C=C stretching from the aromatic compounds. They represent the most intense bands in the Raman spectra. The last one (1548 cm-1), as well as the peak at 1460 cm-1, could also be assigned to the C=N stretching and, to end up, the peak at 964 cm-1 could be assigned to C-H vibration of the benzene ring.(14) On the other hand, by comparing these spectra with doped ones, we can appreciate new and broad absorption bands, in the range of 1300-1500 cm-1. As for the comparison of undoped and doped membranes, it is reported in Figure 5.13 the RAMAN spectra of the pure polyphosphoric acid. In this image it is remarkable the intensity of the peak at Memoria 40 1372 cm-1, which represents the P=O vibrations. Another peak to take into account should be the POH vibration at 911 cm-1. Comparing this spectrum with that of Figures 5.10 and 5.11, we can also see a wide bands between 1337-1397 cm-1 that proves the acid is trapped inside the OPBI films. Figure 5. 10. Raman spectra of OPBI undoped and doped films High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 41 Figure 5. 11. Raman spectra of OPBI/Si-PEI undoped and doped films Figure 5. 12. Raman spectra of OPBI/DBSA (12,5%) undoped films Memoria 42 Figure 5. 13. Raman spectra of polyphosphoric acid 5.3.2. Microscopy analysis In this section, two types of SEM tests are analysed: the morphology of the surface (topography) and the morphology of the fractured films (cross-section). With the first of the techniques what is intended to be observed is the surface of the films so that it could be able to determine if the films are homogeneous or heterogeneous, their composition, if they present precipitated particles or even if they present contaminants. With the cross-section study, the type of fracture of the films could be seen since they are previously cut with the help of liquid nitrogen (i.e. by cryo-fracture). Likewise, it could be determined if the surface of the membranes has the same structure as the interior.  Topography The morphology of the OPBI membranes was studied using scanning electron microscopy technique. SEM images are displayed in Figures 5.14-5.16 and EDX results are shown in the Figure 5.17. The surface of the OPBI membrane is shown in Figure 5.14a and 5.14b. At first glance, it is necessary to emphasize that the surface is quite homogeneous since the convex hemispheres that appear in the image are located only on the surface and not inside the film. It probably arises from a fast High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 43 evaporation process, performed under high vacuum and temperature. This effect was not observed for other samples, therefore, we do not believe it belongs to the film morphology itself. On the other hand, from Figure 5.17b , where the EDX diagram is shown, the appearance of the expected C, O and P atoms must be pointed out due to the chemical composition of the polymer and since the phosphoric acid was used on its manufacture. When the membrane is doped (Figure 5.14c and 5.14d), the hemispheres disappeared given the film a flat and smooth appearance. As mention before, the doping process was performed with phosphoric acid (85%) for 7 days, reason why this time the P peak on Figure 5.17d appears much more intense and sharp than in the pristine OPBI. It should be pointed out that Na atoms in the EDX spectra is not expected to belong to ours samples, instead of it, it probably comes from water contamination. Figure 5. 14. SEM micrographs of OPBI undoped and doped: a) OPBI film (low magnification); b) OPBI film (high magnification); c) OPBI doped film (low magnification); d) OPBI doped film (high magnification). Scale bars inset. In the case of OPBI membranes modified with silica nanoparticles (OPBI/Si-PEI), the surface of the membrane is shown in Figures 5.15a and 5.15b. This membrane is clearly heterogeneous since it is possible to distinguish some particles inside the films. These particles are due both to the silica and the formic acid that after drying in an oven, precipitate. It should be noted that fibre formation (Figure 5.15b, arrow inset) was also an indicative of the well obtaining of the polymer film. However, the strategy of functionalization of silica inorganic particles with polyethyleneimine (PEI) did not Memoria 44 result in a good compatibilization of Si particles inside the polymer matrix. As for the EDX diagram (Figure 5.17f), C and O peaks can be noted both in the white and dark parts of the film but the peak of the silica (Si) can only be detected in the white particles seen in the Figure 5.17e. Thus, it confirms that the silica particles had precipitated during the drying stage. After doping, once again, the film becomes more homogeneous and an intense peak of P appears in the EDX spectrum, Figure 5.17h. Figure 5. 15. SEM micrographs of OPBI/Si-PEI undoped and doped: a and b) OPBI/Si-PEI film (high magnification); c) OPBI/Si-PEI doped film (very low magnification); d) OPBI/Si-PEI doped film (high magnification) Furthermore, the surface of the OPBI membrane modified with a plasticizer (OPBI/DBSA) is shown in Figure 5.16a and 5.16b. As the previous membranes, this one is also heterogeneous before doping and it is believed that the bright parts of the film correspond to the zones where the plasticizer was incorporated. This last statement is confirmed by the EDX chart (Figure 5.17j) since appears atoms appear with more intensity in that part than in the dark one. When doped, heterogeneity disappears (Figure 5.16c and 5.16d) and the P peak on the EDX (Figure 5.17l) appears reasonably, with great intensity. Therefore, SEM analysis confirms the films were well doped, in accordance to previous spectroscopy characterization results. Particularly in the case of OPBI/DBSA doped membranes, those were not possible to study by RAMAN spectroscopy. High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 45 Figure 5. 16. SEM micrographs of OPBI/DBSA (5%) undoped and doped: a) OPBI/DBSA (12,5%) film (low magnification); b) OPBI/DBSA (12,5%) film (high magnification); c) OPBI/DBSA (12,5%) doped film (very low magnification); d) OPBI/DBSA (12,5%) doped film (high magnification) Memoria 52 The weight loss percentage decreased from about 42% for the OPBI/Si-PEI to just about 36% for the neat OPBI, and of 22% for the OPBI with the plasticizer at 11000C, Chart 5.2. This might be caused by two reasons: (i) the presence of silica particles helps the decreasing the moisture absorption tendency of the polymer membranes; and (ii) the presence of both silica particles and the plasticizer make them act as thermal shields, protecting the polymer chains from being subjected to heat. As for the degradation temperatures (Chart 5.2), it is clear that when the degradation is about a 20% of weight, the sample that decomposes at less temperature is the one with the plasticizer (DBSA), while the other two films had more or less the same degradation temperature. Thus, it can indicates that the OPBI/DBSA composite absorbs more water than the other films, on its preparation step; despite all films were previously dried in a vacuum oven at 1100C for 5 days. However, when the decomposition reaches the 50%, the pristine OPBI film has the highest decomposition temperature, being more thermally stable than OPBI/DBSA film and OPBI/Si-PEI membranes. Chart 5. 2. Thermogravimetric data of the films studied in this project Sample Tinitial (°C) T0.2* (°C) T0.5* (°C) Char yield** (%) OPBI 20 235 684 37 OPBI/Si-PEI 19 236 527 42 OPBI/DBSA 12,5% 22 147 568 23 *T0.2 and T0.5 correspond to the temperatures after 20% and 50% of degradation, respectively. **Char yield at 11000C. 5.3.5. Water uptake, swelling ratio and swelling volume It is known that OPBI is hydrophilic and has high affinity for moisture due to the tendency of –NH and acid groups to form hydrogen bonds with water molecules (16). After immersion of the membrane in distilled water for 3 days, it was found that OPBI can absorb about 9% of water with respect of the dry membrane. On the other hand, for the membranes that were doped in PA (85%), immersion must take place in this media, as well, for 5 days to prevent the doping from spoiling. From Chart 5.3, we can see reduction in the swelling ratio capacity of the OPBI and OPBI/Si-PEI membranes when comparing pairs with the same composition, undoped and the ones doped, whereas the OPBI with DBSA plasticizer has the contrary effect. The presence of silica nanoparticles in the polymer matrix would prevent the motion of OPBI chains and makes them more rigid related High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 53 to the pristine OPBI films; thus preventing the separation between the stacked polymer backbones caused by the doping acid molecules. As for the swelling volume, there is a general increase between the films that were not doped and the ones doped with PA. This means that the polymer OPBI doped in its three varieties absorbs greater quantity of liquid than its undoped counterparts do. Additionally, our results is in accordance with previous reported works, where they discussed the difficulty to measure the water uptake of the OPBI membranes as well as the OPBI/Si-PEI films. (12)(14). In fact, there is no record of these samples due to the poor stability of the doped films in water. On the other hand, when the plasticizer is used, the integrity of the film improves and that is the reason why it is more stable and we can obtain a value for the water uptake when the membrane is doped with PA. Comparing the OPBI/DBSA undoped and doped water uptake, the late film has been increased by a factor of 9 times the first. Chart 5. 3. Water Uptake, Swelling Ratio, and Swelling Volume in water of OPBI and nanocomposite membranes.a) Sample Water uptake (%) Swelling ratio (%) Swelling volume (%) OPBI 8,94 (1,29) 19,36 (4,08) 30,32 (7,70) OPBI doped -b) 0,79 (1,11) 60,71 (23,16) OPBI/Si-PEI 45,27 (7,69) 6,72 (2,90) 1,08 (1,21) OPBI/Si-PEI doped - b) 2,03 (1,95) 24,43 (4,01) OPBI/DBSA (12,5%) 3,07 (0,70) 1,63 (0,91) 9,41 (1,76) OPBI/DBSA (12,5%) doped 28,08 (4,08) 2,20 (2,53) 47,93 (17,1) a) The standard deviation of measurements is shown in parentheses. b) The OPBI doped sample was not possible to measure due to its lack of film integrity in water medium. Therefore, to conclude, OPBI/DBSA (12,5 wt.%) seems to have the good solvent swelling properties to be used as PBI-based fuel cell membrane. However, the conductivity data must be approached. Memoria 54 5.3.6. Proton conductivity measurements Prior to the start of the proton conductivity measurements, all the membranes were dried by heating at 1000C to avoid the effect of moisture and to prevent the cell from being oxidized by the action of phosphoric acid. The complete procedure was described on the section 4.5.6. As stated before, OPBI membranes in doped conditions have better mechanical integrity than the undoped ones. It was also reflected in the conductivity experiments, when they had to be fitted inside the capacitor cell described on Figures 4.15-4.16. The proton conductivity of all the composite membranes was measured in the range of 22–1800C by using EIS spectroscopy. Accordingly, two kind of plots can be obtained: the Nyquist and the Bode plots. In this work, only the Bode curves have been compared and the proton conductivities of the OPBI and that of the OPBI composite membranes obtained from the Nyquist plots (not shown here) are reported in the Chart 5.4. In all the three composites, the proton conductivities and the Bode plots of the membranes were compared between doped and undoped states. In previous works, the proton conductivity usually increases with increasing temperature(29). Although, in our case, we observed mostly drops in proton conductivity beyond 100-1800C (Chart 5.4). It is probably due to the instability of the PEM at high temperature for long time (48h for stabilization). By contrary, when comparing the doped and undoped states of one unique system, it can be seen that the proton conductivity is higher for the doped films, as expected for this type of high temperature thermoplastic (Figure 5.23). The values indicate that pristine OPBI doped and OPBI/DBSA doped composites had the better proton conductivities, compared to the other samples. In fact, the proton conductivities of OPBI/DBSA composites are very close to that of pristine OPBI sample, which is also the manifestation of the good stability of this composite. Thus, 12,5 wt.% of DBSA load in PEM, can help the polymer to improve its mechanical properties without decreasing the electrical properties. On the other hand, OPBI/Si-PEI had the worse electrical behaviour due to the inhomogeneity of nanoparticles dispersion, as observed by SEM. The poor mechanical property of this film, prevent it to be measured at high temperatures. Therefore, this composition would not have an effective applicability as PEM. High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 55 Chart 5. 4. Proton conductivity of OPBI, OPBI/Si-PEI and OPBI/DBSA (12,5%) membranes Membrane Membrane state Temperature (0C) Resistance () Thickness (μm) Width (mm) σ (S/cm) OPBI undoped 22 7,41E+05 45,4 15 3,97E-05 100 2,38E+05 1,23E-04 180 1,49E+05 1,97E-04 doped 22 1,99E+04 22,9 14,7 2,98E-03 100 6,88E+03 8,64E-03 180 4,34E+04 1,37E-03 OPBI/DBSA (12,5%) undoped 22 3,94E+05 20,6 14,8 1,66E-04 100 8,73E+06 7,51E-06 180 4,21E+05 1,56E-04 doped 22 4,84E+04 42,9 15 6,42E-04 100 2,79E+05 1,11E-04 180 7,12E+05 4,36E-05 OPBI/Si-PEI undoped 22 4,52E+05 17,1 14,7 1,76E-04 100 - - - - 180 - - - - doped 22 7,00E+05 61,9 15 3,08E-05 100 5,88E+06 61,9 15 3,66E-06 180 - - - - Figure 5. 23. Conductivity measurements: a) OPBI doped and undoped; b) OPBI/DBSA (12,5%) doped and undoped; c) OPBI/Si-PEI doped and undoped Memoria 56 The proton conductivities showed in the Chart 5.4 were measured employing the values of the film resistance (R2) taken from a simple electrical circuit [(R1(R2Q1)]. Therefore, the data provided are not precise. For a better interpretation of these results, we will need devote time to try other electrical circuits that can adjust the experimental data. However, observing the Bode plots (Figures 5.24-5.28), some of our discussion is validated. The phase angle (edge Y2, on right) is an indicative parameter of the insulating or conductive properties of the film, at high frequencies. When this value is close to 90 degrees, the samples are mostly insulating, whereas if it is close to zero degree, it will be conductive. The undoped samples have phase angle starting close to 60-70 degrees and the doped samples dropped to approximately 5 degrees. Figure 5. 24. OPBI undoped Bode curves at different temperatures (22,100 and 1800C) High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 57 Figure 5. 25. OPBI doped Bode curves at different temperatures (22,100 and 1800C) Figure 5. 26. OPBI/DBSA (12,5%) undoped Bode curves at different temperatures (22,100 and 1800C) Memoria 58 Figure 5. 27. OPBI/DBSA (12,5%) doped Bode curves at different temperatures (22,100 and 1800C) Figure 5. 28. OPBI/Si-PEI undoped Bode curves at different temperatures (220C) Figure 5. 29. OPBI/Si-PEI doped Bode curves at different temperatures (22 and 1000C) High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 59 6. Environmental impact analysis This section describes a study of the possible environmental impact caused by the synthesis of the OPBI films, including energy consumption, gas emissions and the elimination of waste generated during the project. In addition, an analysis of the substance’s hazards used during the realization of this project will be carried out following the European Union regulations in force. The regulation is subject to the REACH standards that encompass the regulatory framework for the management of chemical substances. 6.1. Energy consumption and emissions An aspect to take into account in the environmental impact analysis is the generation of carbon dioxide emissions due to the consumption of electricity. This consumption comes from two main components: the drying oven, with a power of 1,4kW and the extractor hood, with a power of 0,25 kW. The electricity observatory, in its January 2016 bulletin (40) states that the annual average for 2012 is 0,146 kg of carbon dioxide per kW·h consumed. For other greenhouse gases such as nitrogen oxides (NOx) and for sulfur dioxide (SO2) values are 0,220 g NOx/kW·h and 0,309 g SO2/kW·h, respectively. If it is considered a treatment of 400 h in the extractor hood and another 650 h in a heating stove, the total consumption results in 1010 kW·h: i) Extractor hood: 0,25 kW·400 h = 100 kW·h ii) Heating stove: 1,4 kW·650 h = 910 kW·h Therefore, the quantities of gases mentioned that are estimated to have been emitted into the atmosphere are: i) 1010 kW·h·0,146 kg CO2/kW·h= 147,46 ≈ 147 kg of CO2 ii) 1010 kW·h·0,309 g SO2/kW·h= 312,09 ≈ 312 g of SO2 iii) 1010 kW·h·0,220 g NOx/kW·h= 222,20 ≈ 222 g of NOx Therefore, this can be considered the global contribution of this project to air pollution in the city of Barcelona and the global greenhouse effect. Memoria 60 6.2. Environmental impact of experimental phase To identify the impacts produce during the investigation, it is necessary to understand that the cause of the environmental impact is the activity and the consequence is the impact. This concept is shown and explained in Figure 6.1. Figure 6. 1. Scheme to identify environmental impacts In concordance with the past concept and considering as main aspect the use of chemical reagents, the principal impacts generated during experimental phase can be analysed in Figure 6.2. In order to prevent these effects was necessary: follow the normal lab safety rules (wear lab coats, gloves and protection glasses), store products and reagents in controlled atmospheres with proper ventilation and taking into account supplier’s recommendations. Figure 6. 2. Causes and effects produced during this investigation Activity Aspect Impact Synthesis and doping of OPBI membranes Use of chemical reagents  Health risk  Pollution  Decrease in air quality  Greenhouse effect Cause Effect Cause Effect High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 61 6.3. Treatment of waste generated The waste generated during the execution of this project was, mainly, non-chlorinated substances, acids and solid waste that were disposed in special containers, depending on their nature. Subsequently, a specialized external company destroys the contents of the containers according to current regulations. 6.4. Substance’s hazards Regulation (EC) Nº 1907/2006 (hereinafter referred to as REACH, acronym of Registration, Evaluation, Authorization and Restriction of chemical substances and mixtures) entered into force on June 1st 2007 and whose main objective is to improve protection for human health and the environment against the risk that the manufacture, commercialization and use of the chemical substances and mixtures can entail.(41) REACH standard applies to all chemical substances present in daily life, either as such, in mixtures or contained in products, and is therefore applicable in economic sectors of diverse nature. To comply with REACH provisions, companies must identify and manage the risks associated with the substances they manufacture and market in the European Union. They must demonstrate how to use these substances safely and communicate all information related to risk management measures to all parties involved. According the current regulation, Chart 6.1 shows the substances classified as dangerous and their pictograms, whereas in the Chart 6.2 their Hazard Statements are indicated, according to the established codes. Chart 6. 1.Classification of the dangerous substances Chemical substance Danger Pictogram Formic Acid GHS02, GHS05, GHS06, GHS07,GHS08 Memoria 68 8.2. Equipment cost The costs per hour of the equipment needed to perform the characterization are shown in Chart 8.2. (42) Chart 8. 2. Cost per hour of the equipment used in this project Equipment Time (h) Cost (€/h) Final Cost (€) Potenciostategalvanostat 12,0 35,0 420 FT-IR spectrophotometer 6,0 22,50 135 Scanning electron microscope 4,5 187,50 843,75 Contact angle meter 5,0 15,0 75 RAMAN Spectroscopy 7,0 56,25 393,75 Thermogravimetric analyser 15,0 12,0 180,0 Particle Size Analizer (Laser Diffraction) 3 60,0 180 TOTAL (€) 2.227,5 8.3. Personnel cost The labour cost was established based on the hours spent on the project and considering a minimum salary of 8 €/h (43), marked by the UPC practices agreement. Chart 8.3 shows the time inverted in the project and the final cost. High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 69 Chart 8. 3. Associated cost of the hours invested in the project Activity Time (h) Cost (€/h) Final cost (€) Bibliographic research 70 8 560 Experimental test 650 8 5.200 Result analysis 190 8 1.520 TOTAL (€) 7.280 8.4. Total cost In general expenses were included electricity, water, basic laboratory and office material. To calculate this value, it is estimated the 10% of the past expenses. As for the consumables, the cost includes laboratory material such as petri dishes, beakers, weight paper… Thus the global cost of the project was 11.235,24 €, the final sum is shown in Chart 8.4. Chart 8. 4. Global cost Expense Final cost (€) Reagents 106,36 Consumables 600 Equipment 2.227,5 Personnel 7.280 General expenses 1.021,38 TOTAL (€) 11.235,24 High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 71 References 1. Lee, K.-S. et al. Synthesis and Characterization of Highly Fluorinated Cross-linked Aromatic Polyethers for Polymer Electrolytes. A: Chemistry of Materials [en línia]. 2010, Vol. 22, núm. 19, p. 5511. Disponible a: https://doi.org/10.1021/cm101405h. 2. Huang, Y. et al. Niobium phosphates as an intermediate temperature proton conducting electrolyte for fuel cells. A: Journal of Materials Chemistry [en línia]. 2012, Vol. 22, núm. 42, p. 22456. Disponible a: http://dx.doi.org/10.1039/C2JM34704K. 3. Nawn, G. et al. Nanocomposite Membranes based on Polybenzimidazole and ZrO2 for HighTemperature Proton Exchange Membrane Fuel Cells. A: ChemSusChem. 2015, Vol. 8, núm. 8, p. 13811393. ISSN 1864564X. DOI 10.1002/cssc.201403049. 4. Samms, S.R., Wasmus, S. i Savinell, R.F. Thermal Stability of Proton Conducting Acid Doped Polybenzimidazole in Simulated Fuel Cell Environments. A: Journal of The Electrochemical Society [en línia]. 1996, Vol. 143, núm. 4, p. 1225. Disponible a: http://jes.ecsdl.org/content/143/4/1225.abstract. 5. Müller, F. et al. New sulfonated polystyrene and styrene-ethylene/butylene-styrene block copolymers for applications in electrodialysis. A: Journal of Physical Chemistry B. 2012, Vol. 116, núm. 38, p. 11767-11779. ISSN 15205207. DOI 10.1021/jp3068415. 6. Müller, F. et al. Measuring the Proton Conductivity of Ion-Exchange Membranes Using Electrochemical Impedance Spectroscopy and Through-Plane Cell. A: The Journal of Physical Chemistry B [en línia]. 2014, Vol. 118, núm. 4, p. 1102-1112. ISSN 1520-6106. DOI 10.1021/jp409675z. Disponible a: http://pubs.acs.org/doi/abs/10.1021/jp409675z. 7. FUEL CELL TODAY. «The Fuel Cell Industry Review 2012». A: Platinum Metals Review [en línia]. 2012, Vol. 56, núm. 4, p. 272-273. ISSN 00321400. DOI 10.1595/147106712X657535. Disponible a: http://openurl.ingenta.com/content/xref?genre=article&issn=00321400&volume=56&issue=4&spage=272. 8. Sannigraphi, A. et al. How the Monomer Concentration of Polymerization Influences Various Properties of Polybenzimidazole: A Case Study with Poly(4,4’-diphenylether-5,5’-bibenzimidazole). A: Journal of Applied Polymer Science. 2008, Vol. 111, p. 2194-2203. ISSN 09673911. DOI 10.1002/app. 9. Haque, M.A. et al. Acid doped polybenzimidazoles based membrane electrode assembly for high temperature proton exchange membrane fuel cell: A review. A: International Journal of Hydrogen Energy [en línia]. Elsevier Ltd, 2017, Vol. 42, núm. 14, p. 9156-9179. ISSN 03603199. DOI 10.1016/j.ijhydene.2016.03.086. Disponible a: http://dx.doi.org/10.1016/j.ijhydene.2016.03.086. 10. BZ, X. and Savadogo, O. The effect of acid doping on the conductivity of polybenzimidazole (PBI). A: Journal of New Materials for Electrochemical Systems. 1999, Vol. 2, p. 95. 11. Subianto, S. Recent advances in polybenzimidazole/phosphoric acid membranes for hightemperature fuel cells. A: Polymer International. 2014, Vol. 63, núm. 7, p. 1134-1144. ISSN 10970126. DOI 10.1002/pi.4708. Memoria 72 12. Singha, S. i Jana, T. Structure and Properties of Polybenzimidazole/Silica Nanocomposite Electrolyte Membrane: In fl uence of Organic/Inorganic Interface. A: . 2014, 13. Ellis, B. i Smith, R. Polymers: A Property Database, Second Edition [en línia]. 2, ilustra. CRC Press, 2008. ISBN 9781420005707. Disponible a: https://books.google.es/books?id=S-TKBQAAQBAJ. 14. Li, Q. et al. Water uptake and acid doping of polybenzimidazoles as electrolyte membranes for fuel cells. A: Solid State Ionics. 2004, Vol. 168, núm. 1-2, p. 177-185. ISSN 01672738. DOI 10.1016/j.ssi.2004.02.013. 15. Wang, K.Y., Chung, T.S. i Qin, J.J. Polybenzimidazole (PBI) nanofiltration hollow fiber membranes applied in forward osmosis process. A: Journal of Membrane Science. 2007, Vol. 300, núm. 1-2, p. 612. ISSN 03767388. DOI 10.1016/j.memsci.2007.05.035. 16. Singha, S. and Jana, T. Structure and properties of polybenzimidazole / silica nanocomposite electrolyte membrane : Influence of organic / inorganic interface. A: . p. 1-4. 17. Peighambardoust, S.J., Rowshanzamir, S. i Amjadi, M. Review of the proton exchange membranes for fuel cell applications [en línia]. Elsevier Ltd, 2010. ISBN 2177491223. DOI 10.1016/j.ijhydene.2010.05.017. Disponible a: http://dx.doi.org/10.1016/j.ijhydene.2010.05.017. 18. Li, Q. et al. High temperature proton exchange membranes based on polybenzimidazoles for fuel cells. A: Progress in Polymer Science (Oxford). 2009, Vol. 34, núm. 5, p. 449-477. ISSN 00796700. DOI 10.1016/j.progpolymsci.2008.12.003. 19. Perry, K.A., Eisman, G.A. i Benicewicz, B.C. Electrochemical hydrogen pumping using a hightemperature polybenzimidazole (PBI) membrane. A: Journal of Power Sources. 2008, Vol. 177, núm. 2, p. 478-484. ISSN 03787753. DOI 10.1016/j.jpowsour.2007.11.059. 20. Li, X., Chen, X. and Benicewicz, B.C. Synthesis and properties of phenylindane-containing polybenzimidazole (PBI) for high-temperature polymer electrolyte membrane fuel cells (PEMFCs). A: Journal of Power Sources [en línia]. 2013, Vol. 243. Disponible a: https://doi.org/10.1016/j.jpowsour.2013.06.033%0AUR. 21. Nasef, M.M. et al. Enhancement of performance of pyridine modified polybenzimidazole fuel cell membranes using zirconium oxide nanoclusters and optimized phosphoric acid doping level. A: International Journal of Hydrogen Energy [en línia]. Elsevier Ltd, 2016, Vol. 41, núm. 16, p. 6842-6854. ISSN 03603199. DOI 10.1016/j.ijhydene.2016.03.022. Disponible a: http://dx.doi.org/10.1016/j.ijhydene.2016.03.022. 22. Devrim, Y., Devrim, H. i Eroglu, I. Polybenzimidazole/SiO2hybrid membranes for high temperature proton exchange membrane fuel cells. A: International Journal of Hydrogen Energy. 2016, Vol. 41, núm. 23, p. 10044-10052. ISSN 03603199. DOI 10.1016/j.ijhydene.2016.02.043. 23. Chuang, S.W., Hsu, S.L.C. i Liu, Y.H. Synthesis and properties of fluorine-containing polybenzimidazole/silica nanocomposite membranes for proton exchange membrane fuel cells. A: Journal of Membrane Science. 2007, Vol. 305, núm. 1-2, p. 353-363. ISSN 03767388. DOI 10.1016/j.memsci.2007.08.033. 24. Pu, H. et al. Organic/inorganic composite membranes based on polybenzimidazole and nano- High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 73 SiO2. A: Electrochimica Acta. 2009, Vol. 54, núm. 28, p. 7536-7541. ISSN 00134686. DOI 10.1016/j.electacta.2009.08.011. 25. Ghosh, S. et al. Role of Clays Structures on the Polybenzimidazole Nanocomposites: Potential Membranes for the Use in Polymer Electrolyte Membrane Fuel Cell. A: The Journal of Physical Chemistry C [en línia]. 2011, Vol. 115, núm. 23, p. 11474. Disponible a: https://doi.org/10.1021/jp202672s. 26. Ghosh, S., Maity, S. i Jana, T. Polybenzimidazole/silica nanocomposites: Organic-inorganic hybrid membranes for PEM fuel cell. A: Journal of Materials Chemistry [en línia]. 2011, Vol. 21, núm. 38, p. 14897. ISSN 0959-9428. DOI 10.1039/c1jm12169c. Disponible a: http://xlink.rsc.org/?DOI=c1jm12169c. 27. Fang, J. et al. Preparation and characterization of novel pyridine-containing polybenzimidazole membrane for high temperature proton exchange membrane fuel cells. A: Journal of Membrane Science. 2016, Vol. 502, p. 29-36. ISSN 18733123. DOI 10.1016/j.memsci.2015.12.006. 28. Wang, J.T.W. i Hsu, S.L.C. Enhanced high-temperature polymer electrolyte membrane for fuel cells based on polybenzimidazole and ionic liquids. A: Electrochimica Acta. Elsevier Ltd, 2011, Vol. 56, núm. 7, p. 2842-2846. ISSN 00134686. DOI 10.1016/j.electacta.2010.12.069. 29. Sana, B. i Jana, T. Polybenzimidazole composite with acidic surfactant like molecules: A unique approach to develop PEM for fuel cell. A: European Polymer Journal [en línia]. Elsevier Ltd, 2016, Vol. 84, p. 421-434. ISSN 00143057. DOI 10.1016/j.eurpolymj.2016.09.051. Disponible a: http://dx.doi.org/10.1016/j.eurpolymj.2016.09.051. 30. Chakraborty, D.S. Instrumentation of ftir and its herbal applications. A: World Journal of Pharmacy and Pharmacutical Sciences. 2016, Vol. 5, núm. 3, p. 498-505. 31. Cubells Pérez, M. Análisis De Pigmentos Con Espectroscopía Raman: Determinación TeóricoExperimental De La Temperatura Inducida Por El Láser. A: . 2013, 32. Piella Bagaria, J. PROTEIN-NANOPARTICLE CONSTRUCTS FOR INTRACELLULAR DELIVERY. A: . 2011, núm. February. 33. Goldstein, J. Scanning Electron Microscopy and X-Ray Microanalysis: A Text for Biologists, Materials Scientists, and Geologists. New York, NY: Springer Science & Business Media, 2012. 34. Armelin Diggroc, E. i Molina, B.G. Graft copolymers with a random distribution of pyrrole monomer with hydrophilic functionalities. A: . 2016, 35. Coats, A.W. i Redfern, J.P. Thermogravimetric analysis. A review. A: Analyst [en línia]. 1963, Vol. 88, núm. 1053, p. 906-924. Disponible a: http://dx.doi.org/10.1039/AN9638800906. 36. Espejo Miralles, Z. Investigation of Lab-on-Spoon Low-Power Realization for Smart Kitchen and AAL Scenarios. A: . 2013, p. 1-125. 37. Stöber, W., Fink, A. i Bohn, E. Controlled growth of monodisperse silica spheres in the micron size range. A: Journal of Colloid and Interface Science [en línia]. 1968, Vol. 26, núm. 1, p. 62. Disponible a: http://www.sciencedirect.com/science/article/pii/0021979768902725%0AER. Memoria 74 38. Beganskiene, A. et al. FTIR, TEM and NMR Iinvestigations of Stöber Silica Nanoparticles. A: Materials Science (Medžiagotyra) [en línia]. 2004, Vol. 10, núm. 9, p. 287-290. ISSN 00223093. DOI 10.1016/j.jnoncrysol.2012.11.006. Disponible a: http://internet.ktu.lt/lt/mokslas/zurnalai/medz/pdf/medz0-79/02 Beganskiene 287-290.pdf. 39. Feifel, S.C. i Lisdat, F. Silica nanoparticles for the layer-by-layer assembly of fully electro-active cytochrome c multilayers. A: Journal of Nanobiotechnology. BioMed Central Ltd, 2011, Vol. 9, núm. 1, p. 59. ISSN 1477-3155. DOI 10.1186/1477-3155-9-59. 40. WWF España. Producción total peninsular : Demanda total peninsular : A: . 2016, p. 1-7. 41. Consejo, P.E. y del. Reglamento 1907/2006 del Parlamento Europeo y del Consejo, de 18 de diciembre de 2006, relativo al registro, la evaluación y la restricción de las sustancias y preparados químicos (REACH), por el que se crea la Agencia Europea de Sustancias y Preparados Q. A: Diario Oficial de la Unión Europea [en línia]. 2005, p. 1-3. Disponible a: http://echa.europa.eu/es/regulations/reach;jsessionid=12411E70EB80BD2290B756D69F94EEB3.live 2. 42. [Online]. CRNE, Centre de Recerca de Nanoenginyeria. A: UPC-CRNE. 2016, p. 1. 43. Facultat d’Informàtica de Barcelona. Normativa Prácticas en Empresa UPC. A: UPB-FIB. 2015, núm. [Online], p. 14. High temperature thermoplastic based on polybenzimidazole and silica nanocomposites 75