The growth of in situ sol-gel silica in natural rubber. Synthesis, morphological and mechanical characterization of the composites

Great interest was recently devoted to the use of inorganic particles as a reinforcing filler in tires for the automotive industry. In fact,the reinforcing of elastomers by the addition of these fillers affects the stiffness, strength, elongation at break, fracture toughness, energy dissipation (rolling resistance), friction to ice or wet grip, wear (abrasion resistance)and on the material processability. Therefore, the technology improvement of tires currently has to satisfy the requirements of sustainable development in order to reduce fuel consumption, environmental pollution and acoustic noise. Carbon black has been the only additive used for this purpose for a long time but silica is now becoming the alternative reinforcing filler; as it offers a lot of advantages, such as better rolling resistance and reduction of the heat buildup; moreover, it can be suitably employed in all cases where the black color is not required. Natural rubber-silica (NR-SiO2) composites are usually prepared by mechanical mixing. Unfortunately, silica particles have a strong tendency to interact with each other within the elastomeric matrix, favoring the inhomogeneous dispersion due to particle tendency to agglomerate and in principle to reduce the filler-rubber interaction. Significant contributions to overcome the disadvantages derived from filler-filler interaction of the silica particles; they are obtainable by enhancing the filler-rubber interaction, which causes additional cross-linking in the rubber structure. The enhancement of filler-rubber interaction is obtained by the use of coupling agents which interact with both the polymer (hydrophobic) and the silica(hydrophilic) surface groups,due to the presence of different functionalities at the ends of the molecules. An alternative approach and object of the present Ph.D thesis,is to prepare composites by in situ formation of the silica filler particles by sol-gel hydrolysis and condensation of tetraethoxysilane(TEOS. Therefore,the aim of this work is to prepare silica-natural rubber (NR-SiO2) composites, with the intention of improving the properties which normally the compound, prepared by mechanical blending possesses. Several factors like nature of the solvent, nature of the catalyst, medium pH, molar ratio between alkoxysilane and water or solvent, gelling and drying temperatures are fundamental sol-gel parameters which can modify the process of the nanocomposite preparation in order to optimize the better homogeneity of the filler distribution inside the rubber matrix. Therefore, considering the filler particle growth in situ in the polymeric matrix, it is possible to distinguish two possible preparation techniques which differ in some of the factors affecting the sol-gel process: aqueous and non-aqueous in situ sol-gel methods. In addition to these factors that influence the preparation of the composite, the amount of the filler present in the silica-rubber composite affects the final performance. In fact, the enhancement in mechanical properties can be achieved when the composite contains the optimal amount of the filler required to form a continuous filler network within the rubber matrix; filler amounts less of this percolation threshold show mostly constant and poor mechanical properties. The presence of a continuous filler network and its homogeneity depends on the filler characteristics, such as size and shape of the particles and on the in situ composite preparation method used. The formation of a convenient filler network is in turn relatable to the physical and chemical interactions among the particles and among their aggregates (filler-filler interaction) and to the chemical and physical interactions between the particles and the matrix (filler-rubber interaction). The presence of functional groups on the particle surface can significantly influence the interface between the filler particles and the rubber matrix; consequently, modifying the filler-filler and filler-rubber interaction leads to the variation of the reinforcement level of the composites. With the intention to investigate and to rationalize the effects induced by surface functionalization of the filler on the properties of rubber composites, during the aqueous sol-gel synthesis of the silica the surface was functionalized by using trialkoxysilane having different functional groups. These functionalities were selected among those which are suitable for promoting the formation of differently shaped silica particles or for modulating the filler-filler and the filler-rubber interactions. Silica particles were modified by molecules containing alkylthiol, thiocarboxylate, alkyldisulfide, alkyltetrasulfide (silica functionalized with containing S groups); vinyl, propyl, octyl chains, alkylamine, alkylcyanate and alkylisocyanate groups (silica functionalized with containing N groups). For this purpose a mixture of TEOS and TMSPM ((3-mercaptopropy)trimethoxysilane) or TESPD (bis(3-triethoxysilylpropyl)disulfide) or TESPT (bis(3-triethoxysilylpropyl) tetrasulfide) or NXT (3-octanoylthio-1propyltriethoxy) or VTEOS (vinyltriethoxysilane) or PTEOS (triethoxy(propyl)silane) or OCTEOS (triethoxy(octyl)silane) or APTEOS ((3-aminopropyl)triethoxysilane) or CPTEOS (3-cyanopropyltriethoxysilane) or ICPTEOS (3-(triethoxysilyl)propylisocyanate) were introduced in NR latex (containing 60 % dry rubber, 39.3 % of water and 0.7% of NH3) during the aqueous sol-gel process. The functionalized molecules were selected with the aim of promoting the formation of different shapes on silica particles (anisotropic or spherical), moreover of modifying the filler-filler and filler-rubber interaction through the chemical functionalities of substituents. In the aqueous in situ sol-gel method, the presence of large amounts of water helps the silica precursors to react quickly in the early stage of the synthesis in rubber matrix, allowing thus to increase the hydroxyl group numbers on particle surfaces, making them less compatible with the rubber and in this way favoring particle aggregation. Through the non aqueous in situ sol gel method, the oxygen present in silica nanoparticles is provided by a suitable reaction and not, as in the aqueous in situ sol gel method, by the water solvent. During this method, the addition of two simple solvents on the metal oxide precursor can generate water in situ that can start sol-gel hydrolysis and condensation reactions. In order to work in an environment without the presence of the water as initial reactant, the silica-rubber composite by the non-aqueous in situ sol-gel method was prepared starting from a solution of dry NR with toluene (inert solvent) and TEOS, which was added to formic acid and alcohol (ethanol of benzylalcohol). Therefore, well defined amounts of water were formed through the esterification reaction produced by formic acid and alcohol, which control the formation of metal oxide growth within the rubber matrix. Moreover, in order to understand better the morphology of the silica particle growth in NR through the in situ sol-gel method, bare silica and functionalized silica powders were prepared by using the same method without the rubber. These powders were morphologically characterized with the intention of more easily evaluating the shape, the size, the surface area, the effect of the metal oxide precursor and the modification of the silica surface. Regarding the composites, the amount of silica was determined by thermogravimetric analysis(TGA) in air. The stability and the reactivity of the functional groups and the hydrolysis rate of the alkoxy groups of the trialkoxysilanes in the early stage of the sol-gel reaction were evaluated by ATR-FTIR. The homogeneity of the particle dispersion, the dimension and shape of silica aggregates were investigated by SEM and TEM, to draw appropriate relations between the filler morphology and the reinforcement of the elastomeric network. The efficacy of the filler network in reinforcing the rubber matrix was assessed by swelling measurements. The Electron Spin Resonance(ESR) behavior of nitroxide radical, introduced as spin probe in order to check the rigidity of the rubber chains, was also investigated. The static and dynamic-mechanical properties of the nanocomposites, both uncured and vulcanized, were investigated and discussed referring to the network morphology, allowing to suggest a connection between the silica precursors used and the functional properties and the amount of the filler. The vulcanization kinetics was also studied, as well as the homogeneity of the dispersion of the filler and the rubber networks. Hardness, abrasion resistance, tensile analysis and compression set results of vulcanized composites were discussed taking into account the structural, morphological and mechanical characteristics determined before. Silica-natural rubber composites for tire applications with a controlled composition and morphology was obtained by in situ sol-gel method and compared with conventional mechanical blending prepared by using silica Rhodia and dried NR. Regarding the in situ sol-gel method of nanocomposite preparation, two different synthetic approaches were carried out by starting from the same silica precursor: aqueous and non aqueous methods. Deep investigation on the relationship between composition, size and morphology of the silica particles, dispersion network degree and dynamic mechanical behavior of uncured composites allowed to rationalize the final performance of the cured composites. In fact, the efficacy of silica filler to improve the mechanical properties of the tires for the automotive industry is related both to the interaction among the silica particles and aggregates (filler-filler interaction) and to their capability to interact with the rubber matrix (filler-rubber interaction). Therefore, the characteristics of the silica such as shape, size, surface area, nature of the surface (OH group, functional molecules and amount), degree of aggregation lead to modify the nature of the interface with the rubber and also the homogenous distribution of the filler network. The obtained results for aqueous in situ sol-gel silica-natural rubber preparation led to the following conclusions: - The aqueous sol-gel method is a promising procedure to prepare nanocomposites from silica precursor TEOS mixed to trialkoxisilanes having different functional groups when the filler precursor doesn’t react quickly through hydrolysis and condensation reactions in water environment and allow to control the filler-filler and filler-rubber interactions. - The functional groups from different substituted silicon alkoxide precursors promote during the sol-gel process the formation of different silica particle shapes and modify the filler-filler and filler-rubber interaction.In particular,the precursor functionalities induced the formation of anisotropic shaped silica particles, unlike the spherical ones derived from TEOS. - Different precursors give rise to particle networks with different degrees of continuity, depending on physical and chemical properties of particles they originate. Spherical or slightly anisotropic particles with homogeneous size show the best self assembling behavior and form continuous networks. When particles contain surface groups able to interact with each other (e.g. hydroxyl, amino and thiol groups), the formation of chemical bonds makes the filler-filler interaction stronger along the network. - The network continuity within the composite is the main prerequisite to obtain strong rubber reinforcement. Not homogeneous distribution of the filler and irregular segregation of particles induce large voids in the network, lowering the dynamic-mechanical properties. However, it appears that a strong chemical interaction among particles can balance the absence of a fully continuous network, preserving high storage modulus. This is the case of the network in NR-TMSPM, where the thiol functionality assists the bonding interaction. On the contrary, the large voids observed in the network of NR-OCTEOS are not compensated by bonding interactions among particles, which are absent or very weak due to the presence of the surface alkyl groups. - The different filler-rubber interaction due to substituents able to chemically interact with the polymer, promotes the homogeneous distribution of the filler particles even if its contribution to the reinforcement properties is less effective than that to the filler-filler interaction. Thus filler-filler interaction governs the dynamic-mechanical properties of silica rubber composites either through the shape induced physical interactions responsible for the network formation, or by the chemical interaction among particle surface groups. - To sum up when a choice among different functionalized silica is required, the suggestion is to look for well assembled and continuous filler networks, eventually assisted by chemical interaction among particles. The second condition seems very important, in fact in the case of NXT network, the homogeneous particle distribution is not sufficient to guarantee strength; instead the dynamic-mechanic behavior of NR-NXT is enhanced after thermal treatment which allows the sulphur-rubber interaction. - Regarding the properties of the vulcanized composites, it is evident that the preparation of the silica in situ improve the rigidity, the hardness, the tensile strength and reduce the dissipation of the energy, which means an improvement in rolling resistance. The obtained results for non-aqueous in situ sol-gel silica-natural rubber preparation led the following conclusions: -The non-aqueous in situ sol-gel synthesis is a procedure to prepare nanocomposites allowing the in situ growth of small particles in rubber matrix in the absence of large amount of water. In addition, due to the slow hydrolysis and condensation reaction rate of the precursors it possible to control at molecular level the growth of the metal oxide particles and therefore the filler-filler interaction and the dispersion in rubber matrix. - Non-aqueous in situ sol-gel silica-natural rubber nanocomposites were prepared by in situ esterification reaction between formic acid and ethylalcohol or benzylalcohol which produce controlled amount of water for the hydrolysis and condensation reaction. Therefore inside the dried NR it is possible to growth high amount of silica particles well distributed in a network were the particles are less aggregated and show lower filler-filler interaction in comparison with the composite prepared by conventional blending route. - Silica-natural rubber composites were obtained containing large amount of well dispersed silica (75 phr or 43% wt) without using coupling agents. -The nature of the alcohol involved in the esterification reaction influences the shape and the size of the silica prepared in swollen natural rubber. Silica particles produced in ethanol environment are bigger than those produced in presence of benzylalcohol, and create an effective reinforcement of the rubber when their amount is higher than 60 phr (37% wt). In spite of the different particle dimensions, silica produced from the two different routes are equally highly dispersed in the rubber matrix, due to the lower hydrophilicity of the surface, and a large amount of filler is required to form an efficient silica network able to reinforce the composite. -This behavior, peculiar of silica particles prepared by non aqueous in situ method, confirm that even if a homogeneous distribution of the particles in the matrix is required to obtain an efficient and strong reinforcement of the rubber, highly separated particles with low filler-filler interaction hinder an efficient filler network. On the other side the non aqueous synthesis allows to load a large amount of silica in the rubber, which could not be loaded with the traditional blending method without using a compatibilizer or disperdent agent.