Chapter 1: Carbon Nanotube-Reinforced Polymers 4.3.2 Nanotube- Matrix Interaction Another aim of this research was to achieve a good bonding between the nanotubes and the matrix. A load transfer between them can only be achieved with a sufficiently good bonding, which should result in an improvement of the fracture mechanical properties. TEM provides a possibility of obtaining useful qualitative information concerning the nanotube-epoxy matrix interaction Oxidized MWCNT were dispersed in an epoxy matrix by the above- mentioned techniques and visualized in Figure 12 as a series ofTEM micrographs. 9 The TEM foil was loaded in the microscope by electron beam heating and it seen that a nanotube embedded in the epoxy matrix bridges a void. Electron beam- induced void growth leads to an increasing load and to a rupture of the epoxy ridge" Due to the low nanotube/matrix bonding no reinforcement of the bridging polymer occurs, but one can observe a pull-out of the nanotube from the matri Figure 12. Transmission electron micrographs of oxidized CNT in an epoxy matrix. void and crack growth results from thermal loading of the specimen with the electric beam. a tube bridging a crack is clearly seen; the weak nanotube-matrix bonding leads to pull-out. 9 The conclusion can be drawn that the pull-out of oxidized CNT, regardless of the presence of carboxylic groups on their surfaces, points to an insufficient matrix bonding. Since the void formation and the loading of the test TEM foils took place by electron beam heating, only qualitative assessments can be made 4.3.3 Functionalization Only by a substantial improvement of the CNT-matrix bonding, an adequate reinforcement and, especially, an improvement of the composite fracture strength can be achieved. Therefore, the functionalization of the CNT surfaces is a key issue in order to realize a CNT reinforcement. The importance of functionalization
Chapter 1: Carbon Nanotube-Reinforced Polymers 13 4.3.2 Nanotube-Matrix Interaction Another aim of this research was to achieve a good bonding between the nanotubes and the matrix. A load transfer between them can only be achieved with a sufficiently good bonding, which should result in an improvement of the fracture mechanical properties. TEM provides a possibility of obtaining useful qualitative information concerning the nanotube-epoxy matrix interaction. Oxidized MWCNT were dispersed in an epoxy matrix by the abovementioned techniques and visualized in Figure 12 as a series of TEM micrographs? The TEM foil was loaded in the microscope by electron beam heating and it is seen that a nanotube embedded in the epoxy matrix bridges a void. Electron beaminduced void growth leads to an increasing load and to a rupture of the epoxy "bridge". Due to the low nanotubelmatrix bonding no reinforcement of the bridging polymer occurs, but one can observe a pull-out of the nanotube from the matrix. Figure 12. Transmission electron micrographs of oxidized CNT in an epoxy matrix. Void and crack growth results from thermal loading of the specimen with the electric beam. A tube bridging a crack is clearly seen; the weak nanotube-matrix bonding leads to pull-out? The conclusion can be drawn that the pull-out of oxidized CNT, regardless of the presence of carboxylic groups on their surfaces, points to an insufficient matrix bonding. Since the void formation and the loading of the test TEM foils took place by electron beam heating, only qualitative assessments can be made. 4.3.3 Functionalization Only by a substantial improvement of the CNT-matrix bonding, an adequate reinforcement and, especially, an improvement of the composite fracture strength can be achieved. Therefore, the functionalization of the CNT surfaces is a key issue in order to realize a CNT reinforcement. The importance of functionalization
Polymer Composites: from Nano- to Macroscale in the achievement of a strong CNT-matrix bonding was shown by Frankland et al. via a molecular simulation. These authors have shown that an improvement in mechanical properties can already be achieved if less than 1% of the carbon atoms of the CNT form reactive bridges with the matrix a breakthrough in the development of CNT-reinforced epoxies could be made by chemical functionalization with multifunctional amines. In addition to he improvement of the bonding to the matrix, a better dispersion of the CNT could also be observed. A schematic representation of the functionalization process is shown in Figure 13.9 After the oxidative treatment of the nanotubes, resulting in the development of carboxylic groups(Scheme(1)they can react in a second MWCNT H2SO,/HNO3 (1) step with multifunctional amines and form ionic bonds(Scheme(2)) MWCNT oH球~NHN (2) MW NH NH H3N By the addition of the epoxy resin, the reactive free amino functions on the CNt surfaces form covalent bonds, improving the nanotube-matrix bonding(Scheme( 3) Figure 13. Schematic representation of the functionalization process, showing the entire cycle from the cleaning to the composite manufacturing
14 Polymer Composites: from Nano- to Macroscale in the achievement of a strong CNT-matrix bonding was shown by Frankland et al. via a molecular simulation." These authors have shown that an improvement in mechanical properties can already be achieved if less than 1% of the carbon atoms of the CNT form reactive bridges with the matrix. A breakthrough in the development of CNT-reinforced epoxies could be made by chemical functionalization with multifunctional amines. In addition to the improvement of the bonding to the matrix, a better dispersion of the CNT could also be observed. A schematic representation of the functionalization process is shown in Figure 13.9 After the oxidative treatment of the nanotubes, resulting in the development of carboxylic groups (Scheme (1)) they can react in a second step with multifunctional amines and form ionic bonds (Scheme (2)). By the addition of the epoxy resin, the reactive free amino functions on the CNT surfaces form covalent bonds, improving the nanotube-matrix bonding (Scheme (3)). Functionalization Figure 13. Schematic representation of the functionalization process, showing the entire cycle from the cleaning to the composite manufacturing
Chapter 1: Carbon Nanotube-Reinforced Polymers WCNT- NH H3N MWC H Amino functions were chosen because they are present in the hardener as well The composites produced were observed in a TEM; for the first time matrix-rich areas could be seen alongside the surfaces of the nanotubes, together with crack bridging and telescopic pull-outs, as described in detail in the next section Mic 4.4.1 Matrix Bonding to the Nanotubes After the functionalization, CNT could be found which were entirely covered with the epoxy matrix, as shown in Figure 14a, where a nanotube sticks in the epoxy, a cone of the matrix is formed, and an epoxy layer spreads along the nanotube up to its tip, where the cap is also covered by epoxy. Due to the oxidation (leading to the formation of carboxylic groups)and the functionalization, the matrix can now directly react with the tube, forming covalent bonds 00 Figure 14. Transmission electron micrographs of functionalized carbon nanotubes in an epoxy matrix showing(a) the cone and the formation of a cap of matrix, and(b)the total coverage of the tube by the epoxy polymer
Chapter 1: Carbon Nanotube-Reinforced Polymers 15 Amino functions were chosen because they are present in the hardener as well. The composites produced were observed in a TEM; for the first time matrix-rich areas could be seen alongside the surfaces of the nanotubes, together with crack bridging and telescopic pull-outs? as described in detail in the next section. 4.4 Microscopy 4.4.1 Matrix Bonding to the Nanotubes After the functionalization, CNT could be found which were entirely covered with the epoxy matrix, as shown in Figure 14a, where a nanotube sticks in the epoxy, a cone of the matrix is formed, and an epoxy layer spreads along the nanotube up to its tip, where the cap is also covered by epoxy. Due to the oxidation (leading to the formation of carboxylic groups) and the functionalization, the matrix can now directly react with the tube, forming covalent bonds. Figure 14. Transmission electron micrographs of functionalized carbon nanotubes in an epoxy matrix showing (a) the cone and the formation of a cap of matrix, and (b) the total coverage of the tube by the epoxy polymer?
Polymer Composites: from Nanoto macroscale 4.4.2 Crack Bridging and Telescopic Pull-Outs A further indication of the improved CNT-epoxy bonding can be found in Figure 14b, where a nanotube bridges a pore. It is clearly seen that the tube is completely covered with matrix resin. This result alone should suggest an improvement of the composite fracture toughness, because the nanotube is now able to bridge microcracks The improved bonding due to functionalization is visualized by the telescopic pull-out of nanotubes in Figure 15. Here again, a pore bridged by a matrix-covered nanotube and a multiple-wall carbon nanotube nicely bonded to the matrix are seen. due to the void growth and the good CNT-matrix bonding the outer layer of the tube remains bonded to the bridge, but the inner tubes are telescopically pulled out of the outer ones. This is possible because only weak van der Waals forces are present between the various concentric tubes and only the outer tube is covalently bonded to the matrix. In such a pull-out process, at least parts of the outer tube fail, indicating that a load transfer from the matrix to the outer tube occurs and that, due to the strong covalent bonding to the epoxy matrix, a consequent pull-out of the inner tubes takes place. This result can be further corroborated by the fact shown in Figure 16; a nanotube has been pulled out of the epoxy matrix. The outer shell of the tube remains in the epoxy matrix and continues into the void, whereas the inner tubes had been pulled out It can be assumed that for the adequate mechanical reinforcement of a I matrIX MWCNT might not be the proper choice. The use of single-or double-wall ca nanotubes might be more appropriate to achieve a good reinforcement. Figure 15. Transmission electron micrograph of functionalized carbon nanotubes. A nanotub bonded to a matrix crack bridge is pulled out of its outer shell ( telescopic pull-out). I
16 Polymer Composites: from Nano- to Macroscale 4.4.2 Crack Bridging and Telescopic Pull-Outs A further indication of the improved CNT-epoxy bonding can be found in Figure 14b, where a nanotube bridges a pore. It is clearly seen that the tube is completely covered with matrix resin? This result alone should suggest an improvement of the composite fracture toughness, because the nanotube is now able to bridge microcracks. The improved bonding due to functionalization is visualized by the telescopic pull-out of nanotubes in Figure 15. Here again, a pore bridged by a matrix-covered nanotube and a multiple-wall carbon nanotube nicely bonded to the matrix are seen. Due to the void growth and the good CNT-matrix bonding, the outer layer of the tube remains bonded to the bridge, but the inner tubes are telescopically pulled out of the outer ones. This is possible because only weak van der Waals forces are present between the various concentric tubes and only the outer tube is covalently bonded to the matrix. In such a pull-out process, at least parts of the outer tube fail, indicating that a load transfer from the matrix to the outer tube occurs and that, due to the strong covalent bonding to the epoxy matrix, a consequent pull-out of the inner tubes takes place. This result can be further corroborated by the fact shown in Figure 16; a nanotube has been pulled out of the epoxy matrix. The outer shell of the tube remains in the epoxy matrix and continues into the void, whereas the inner tubes had been pulled out. It can be assumed that for the adequate mechanical reinforcement of a matrix, MWCNT might not be the proper choice. The use of single- or double-wall carbon nanotubes might be more appropriate to achieve a good reinforcement. Figure 15. Transmission electron micrograph of functionalized carbon nanotubes. A nanotube bonded to a matrix crack bridge is pulled out of its outer shell (telescopic pull-out).I8
Chapter 1: Carbon Nanotube-Reinforced polymers Figure 16. Transmission electron micrograph of a functionalized carbon nanotube in an epoxy matrix. The outer shell of this tube remains in the matrix and continues into the pore, while the inner tubes have been pulled out. 9 4.5 Thermal and Mechanical Properties Figurel7a, b shows the results of the dynamic measurements of the complex modulus and the loss factor as a function of temperature. The thermo-mechanical properties were evaluated in order to verify the expected influence of the functionalization of CNT. 18 They did not show a general trend of the effect of the nanotubes on the complex modulus at temperatures below room temperature. This observation can be explained by the small amounts of nanotubes acting as a filler Assuming a homogeneous dispersion, it can be expected that the nanotubes do not show an orientation and that only a small increase in the complex modulus should be observed Fluence of the carbon nanotubes on the complex modulus composite is observed above room temperature. The increasing amounts of carbon nanotubes tend to result in a shift of the glass transition temperature towards higher values and an increase of the loss modulus E"( Figure 18a, b) This increase in thermal stability can be interpreted as a reduction of the mobility of the polymeric matrix through the nanotubes. This would explain the tendency in the shift of Te. An even stronger effect was measured in composites containing functionalized nanotubes. Un e the c non-functionalized CnT, these samples showed an almost linear dependence tween the nanotube content and the shift of T, figure that covalent bonds between the amino-functions on the surface of the nanotubes and the epoxy matrix lead to an even stronger reduction of the matrix mobility, he latter expressing itself in a stronger shift of T. The different behavior of the two sample series is a further evidence of the influence of the chemical unctionalization of the surface on the interfacial adhesion between the nanotubes and the epoxy resin. a detailed investigation should be carried out of the effect of
Chapter 1: Carbon Nanotube-Reinforced Polymers 17 Figure 16. Transmission electron micrograph of a functionalized carbon nanotube in an epoxy matrix. The outer shell of this tube remains in the matrix and continues into the pore, while the inner tubes have been pulled out.g 4.5 Thermal and Mechanical Properties Figurel7a,b shows the results of the dynamic measurements of the complex modulus and the loss factor as a function of temperature. The thermo-mechanical properties were evaluated in order to verify the expected influence of the functionalization of CNT.18 They did not show a general trend of the effect of the nanotubes on the complex modulus at temperatures below room temperature. This observation can be explained by the small amounts of nanotubes acting as a filler. Assuming a homogeneous dispersion, it can be expected that the nanotubes do not show an orientation and that only a small increase in the complex modulus should be observed. A strong influence of the carbon nanotubes on the complex modulus of the composite is observed above room temperature. The increasing amounts of carbon nanotubes tend to result in a shift of the glass transition temperature towards higher values and an increase of the loss modulus E" (Figure 18a,b). This increase in thermal stability can be interpreted as a reduction of the mobility of the polymeric matrix through the nanotubes. This would explain the tendency in the shift of T,. An even stronger effect was measured in composites containing functionalized nanotubes. Unlike the case of the composites containing non-functionalized CNT, these samples showed an almost linear dependence between the nanotube content and the shift of T, (Figure 19). It can be assumed that covalent bonds between the amino-functions on the surface of the nanotubes and the epoxy matrix lead to an even stronger reduction of the matrix mobility, the latter expressing itself in a stronger shift of T,. The different behavior of the two sample series is a further evidence of the influence of the chemical functionalization of the surface on the interfacial adhesion between the nanotubes and the epoxy resin. A detailed investigation should be carried out of the effect of