Silane coupling agents are typically used in the surface treatment of glass fibers with the desired purpose of tailoring fiber-matrix behavior. Organofunctional silanes can undergo hydrolysis and subsequent condensation with hydroxyl groups on a glass fiber surface to yield a desired chemical reactivity. For a composite requiring high strength, strong interfacial adhesion can be designed by making a fiber highly reactive with the matrix resin. Subsequently, a high degree of coupling with the matrix phase is decidedly effective in transferring stress between the matrix and the fibers. Conversely, composites requiring high toughness and energy absorption capacity can be achieved through poor interfacial bonding by creating chemically non-reactive fiber surfaces. In these systems, fibers can easily debond and pull away from the matrix, allowing energy to be dissipated through frictional losses. It has been previously demonstrated that a fiber sizing package formulation containing a mixture of silane coupling agents that are both reactive and non-reactive towards the matrix phase, in conjunction with the nanoscale colloidal silica, yields a simultaneously high strength and tough composite. However the complex relationships between silane coupling agent mix ratios, overall silane coupling agent concentrations, colloidal silica reactivity, and colloidal silica concentration are not fully understood.
The overall goal of this research is to control the co-deposition of silane coupling agent and colloidal silica on a glass surface to control the nanotexturing roughness enhancement. Ultimately, this should translate to further refinement in the balance between impact energy absorption and mechanical strength properties in glass fiber reinforced composites. The first step of this work was to determine which factors of the silane and silica treatments were most influential in controlling the silica deposition on a glass surface. To accomplish this task, we used silicon wafers as model surfaces to imitate the chemistry that occurs on glass fibers. The basis of the silane chemistry involves controlling the number of amine-functional groups on the fiber surface. Propyltrimethoxylsilane (PTMO) and aminopropyltrimethoxysilane (APS) are silanes that contain one propyl group and one primary amine, respectively. By the varying the ratio of these silanes in solution, we can control the concentration of amine groups on the fiber surface. The amine groups on the fiber surface act as potential sites for epoxy-amine reactions. Two routes were chosen for affixing silica to the surface using the amine groups. In the first route, a reverse silane reaction is carried out; where the epoxy groups of a non-hydrolyzed glycidylpropyltrimethoxysilane (GPS) react with the amine groups on the fiber surface. Upon hydrolysis, the three methoxy groups of the silane are converted to hydroxyl groups that can undergo a condensation reaction with the silica. The second route for silica deposition involves the functionalization of the silica. In a separate step, hydrolyzed GPS is reacted with the silica to yield epoxy-functional silica. The epoxy groups on the silica can react with the amine groups on the glass surface.
In addition to the silica reactivity, two other aspects of the aforementioned chemistry that can affect silica deposition are silica concentration in solution and the number of amine receptor groups available for reaction (controlled through varying the APS:PTMO ratio). Preliminary data has been obtained for systems with varying silica concentration. Silicon wafers treated with 1 wt. % APS were subjected to treatments of 0.01, 0.1, 0.5, and 1.0 wt. % of epoxy-functional silica. Scanning electron micrographs were transferred to imaging software to quantify the surface coverage of silica. Surface coverage was measured at 14.2 ± 0.8, 34.5 ± 2.1, 41.6 ± 2.0, and 27.2 ± 2.6 area % for the 0.01, 0.1, 0.5, and 1.0 wt. % silica solutions, respectively. Particle agglomeration was observed at 1.0 wt. % silica, which explains the lower value for surface coverage.
The surface coverage of silica on the glass surface influences the overall surface roughness. Atomic force microscopy was used to correlate surface coverage to surface roughness. The lowest RMS roughness value was 6.18 ± 0.34 nm, given by the 1.0 wt. % silica solution. Minimal increases in roughness to 6.68 ± 0.42 nm and 6.92 ± 0.18 nm were exhibited by reducing silica concentration to 0.5 and 0.1 wt. %, respectively. The lowest silica concentration (0.01 wt. %) yielded the roughest surface at 9.52 ± 0.40 nm. As the particles become more sparsely dispersed, the height variation of the surface increased. For surfaces with higher surface coverage, height variation was reduced by the adjacent particles, thus reducing overall roughness.
Fiber pull-out during an impact event induces friction that manifests itself as a shear force. Therefore, future work will entail measuring the mode II properties of these systems. A napkin ring test is one test that measures pure mode II properties. In this test, silicon wafers will be subjected to various silane chemistries that yield surfaces of different roughness. A cylindrical ring will be bonded to the silicon wafer surface with an epoxy adhesive. As torsional shear is applied to the ring, the torque will be measured to calculate the shear stress. Another measurement of mode II behavior will be the end-notch flexure (ENF) test. In this test, an epoxy adhesive will be sandwiched between two glass adherends. The adherends will undergo various silane chemistries prior to testing. A starter crack will be used to initiate delamination in the adhesive layer. When loaded in a three point bend, an almost pure shear prevails at the tip of the mid-plane delamination.