Synthesis of Carbon Nanotubes on Carbon Fibre for Hierarchical Composites

Carbon fibre (CF) reinforced polymer (CFRP) composites are surpassing steel greatly in terms of strength and stiffness per weight. This makes them particularly attractive when the weight of a part has a major influence on the performance and environmental impact, such as in automotive or airspace applications, wind turbines or sporting goods. The superior mechanical properties of the CFRPs are mostly pronounced along the fibre orientation. However, in load modes such as compression or shear, the CFRP performance is much lower, dominated by the properties of the polymer matrix and the fibre-matrix interface. Therefore, it is critical to enhance the performance of the fibre-matrix interface and of the matrix in order to push the limits of the lightweight CFRP materials. This thesis explores the concept of hierarchical composites based on CFs grafted with carbon nanotubes (CNTs) by chemical vapour deposition (CVD) for an additional reinforcement of the CF-matrix interphase region at the nanoscale, harnessing the extraordinary mechanical properties of CNTs. CNT growth on CFs poses a challenge, particularly due to the CF degradation in the CVD process, which occurs due to migration of iron (the common catalyst of CNT growth by CVD) into the CF. This issue leads to trade-offs between the quality of the grown CNTs and the strength of CF. In this work, we utilised an alumina barrier layer, synthesised on the CF by atomic layer deposition (ALD) for protection of the fibre in the harsh CVD conditions as well as for an effective support of CNT growth. In Chapter 2, we demonstrated that a 12 nm alumina film indeed allows to block the migration of iron into the CF, which results in retained CF tensile properties. Moreover, the mechanisms of the detrimental iron migration into CFs are identified and quantified by means of ptychographic X-ray computed tomography and scanning transmission electron microscopy (STEM). In Chapter 3, we addressed the issue of delamination of the alumina film from the CF upon exposure to high temperatures used for CNT growth. A modified ALD process was proposed, including an ozone treatment for enhancement of covalent bonding between the fibre and the alumina film. The preservation of the CF-alumina interface shear strength was evidenced by single fibre pull-out testing. The findings were supported by fracture analysis of the pull-out surfaces by scanning electron microscopy and energy-dispersive X-ray spectroscopic elemental analysis. In Chapter 4, we reported a method of coating of alumina-buffered surfaces with iron catalyst nanoparticles, which allows for a homogeneous CNT growth over substrates of complex geometries, such as CF fabrics. A dip-coating method was applied to deposit the iron catalyst precursor on the substrates. We examined the influence of the iron nitrate solution concentration and ageing time on the resultant catalyst iron nanoparticle coating quality in terms of nanoparticle sizes, coating density and dispersion. Moreover, we proposed an aminosilane treatment of the alumina surface, which we showed to enhance the coating quality. The catalyst nanoparticle coating quality was quantified by SEM image analysis. We showed a correlation between the improvements in the catalyst coating quality with the enhanced CNT growth uniformity, density and alignment. The excellent CNT growth morphology was demonstrated on flat substrates, as well as on complex surfaces of CF fabrics. We have identified, that the theory of gas transport in fibrous structures was lacking in the literature. In Chapter 5 we derived an analytical model of gas transport in fibrous media from basic physical principles. The model is applicable both at low pressures, when the gas transport occurs in the Knudsen regime, as well as at higher pressures, while an analytical continuous transition towards the viscous gas transport regime was delivered. The applicability of the model presented reaches far beyond the processing of the fibres for composites. The theoretical framework provided in Chapter 5 was validated at nanoscale in Chapter 6, where it was applied in a kinetic model of ALD on CNT mats. The modelling allowed to predict a coating thickness profile of CNTs in a multicycle diffusion-limited ALD experiment. The predicted profile was compared with an experimentally obtained one, giving an excellent agreement. Moreover, the chapter provides numerous new physical insights into the kinetics of ALD coating of porous structures in general. Accounting for all the above, we believe that the work presented in this thesis constitutes a major progress in understanding and development of the CF-CNT hierarchical composites.