Carbon nanofibers replacing graphene oxide in ceramic composites as a reinforcing-phase: is it feasible?

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ Esta es la versión aceptada del artículo publicado en: This is an accepted manuscript of a paper published in: Journal of the European Ceramic Society (2017): 30/01/2025 DOI: https://doi.org/10.1016/j.jeurceramsoc.2017.03.027 Copyright: © 2017 Elsevier Ltd. All rights reserved. El acceso a la versión publicada del artículo puede requerir la suscripción de la revista. Access to the published version may require subscription. “This is an Accepted Manuscript of an article published by Elsevier in Journal of the European Ceramic Society on 2017, available at https://doi.org/10.1016/j.jeurceramsoc.2017.03.027” Carbon nanofibers replacing graphene oxide in ceramic composites as a reinforcing-phase. Is it feasible? Rafael Cano-Crespo1, Bibi Malmal Moshtaghioun1, Diego Gómez-García1,2, Arturo Domínguez-Rodríguez1, Rodrigo Moreno3 1Departamento de Física de la Materia Condensada, Universidad de Sevilla, apartado 1065, 41080 Sevilla, Spain 2Instituto de Ciencia de Materiales de Sevilla, CSIC-USE, Avenida Américo Vespucio 49, 41092 Sevilla, Spain 3Instituto de Cerámica y Vidrio, CSIC, E-28049 Madrid Abstract In recent years, the interest of graphene has increased extraordinarily due to the outstanding properties concurring in this material. In ceramic science, the possibility of combining excellent electrical conductivities together with enhanced mechanical properties has motivated the research in fabrication of graphene oxide-reinforced ceramic composites despite the intrinsic difficulties for sintering. In this work a comparison is made between graphene oxide-reinforced alumina composites and carbon nanofiber-reinforced alumina ones. It will be concluded that the improvement of mechanical properties is scarce, if any. Since carbon nanofibers have also a good electrical conductivity and other advantages such as better processing and dispersibility, it will emphasized their importance for future applications as a replacement of more sophisticated but expensive graphene-based ceramic composites. Keywords: Alumina, graphene oxide, carbon nanofibers, composites, mechanical properties 1. Introduction Advanced ceramics, such as alumina (Al2O3), have been used in materials industry and have potential applications covering high speed cutting tools, dental implants, chemical and electrical insulators, wear resistance parts and some coatings 1 . These applications are due to its excellent properties: high hardness, chemical inertness and high electrical and thermal insulation properties 2 . A usual strategy to improve the properties and minimize their main drawback, i.e. their brittleness, is the addition of a second phase, i.e. by fabrication of composites. In recent years graphene, the two dimensional array of a one-atom thick layer of carbon atoms arranged in a honeycomb lattice, is a promising candidate. Graphene has a large specific surface area (2630 m2 g– 1), high intrinsic electron mobility (2x105 cm2 v− 1 s− 1 ) 3 , 4 , high Young modulus (≈ 1.0 TPa) 5 , a fracture strength of 130 GPa 6 , thermal conductivity (≈ 5000 Wm− 1 K− 1 ) 7 , high optical transmittance (≈ 97.7%) and therefore it is fine for applications such as transparent conductive electrodes 8 , 9 . Chemically modified graphene (CMG) has been studied in the context of many applications, such as polymer composites, energy-related materials, sensors, ‘paper-like´ materials, fieldeffect transistors (FET), and biomedical applications 10 , 11 . Graphene platelets (GPLs) are formed by several layers of graphene with thickness of up to 100 nm 12 which are called graphene nanosheets (GNS), multilayer graphene nanosheets (MGN) or graphene nanoplatelets (GNPs). Carbon-based materials, like carbon fibers or carbon nanotubes (CNTs) have been used in the last decade to improve the mechanical properties of a variety of materials. CNTs, with high tensile strength and stiffness, good flexibility, and low density have been used to this purpose. A large number of studies have been reported. Since the discovery of carbon nanotubes by Iijima in 1991 13 , this material has also promising properties such as high tensile strength (2.7 GPa), a Young modulus of 230 GPa and an electrical resistivity of 1·10-3Ω∙m 14 , 15 . Due to these properties carbon nanotubes have found many applications as electrode material in lithium ion secondary battery, electrode material for supercapacitors, and supporting material for metal nanoparticles 16 . Many studies have been devoted to improvement of the fracture toughness of Al2O3 by CNTs additions. The results show a large scatter in the values of the hardness and fracture toughness, probably due to the very different degrees of CNTs dispersion and the extreme difficulty to achieve a fully homogeneous microstructure. This fact opens several doubts on the reproducibility of the mechanical response of CNT-based composites. A review on the mechanical properties of CNTs-reinforced ceramics is reported by E. Zapata-Solvas et al. 17 Another way to improve the mechanical properties of advanced ceramics materials is by addition of fibers (NF) in a so-called fiber-reinforced ceramic composite, which show significant improvements in toughness over monolithic ceramics. Fibers usually have diameters in-between a nanometer to tens of micrometers and lengths around several micrometers to hundreds of micrometers, which are embedded inside a fine-grained ceramic matrix 18 . Ostertag et al. 19 fabricated SiC-fiber reinforced alumina matrix composites using slip casting and achieved a 30% improvement of fracture toughness over pure material. Hansson et al. 20 sintered SiC- whisker reinforced alumina composites by hot-pressing and this material exhibited a 100% increase in fracture toughness. In the case of carbon nanofibers, their applications are numerous because of their outstanding electrical, mechanical and catalytic properties. A review on the matter is provided elsewhere21. In the field of ceramic science, carbon nanofibers have been considered as a reinforcement phase in potential systems such as AlN because of their superior advantages compared with single-walled or multi-walled carbon nanotubes; i.e. their low price and their good dispersibility22. Carbon nanofiber-reinforced alumina composites have been the object of study and a few results are reported in literature. Maensiri et al23. fabricated these composites by hot-pressing at 1450ºC achieving a mean grain size 1 m. The fracture toughness improvement of 13% is found for a 2.5 vol% CNF-reinforced alumina composite, although the hardness decreased slightly to 16.2 GPa (with respect to a hardness in pure alumina equal to 19.1 GPa). In the case of graphene there are several studies in this field: Liu et al.1 fabricated graphene platelets (GPLs)/Al2O3 system and documented a 27.20% increase in fracture toughness. Yadzani et al.24 reported a 63% increase in fracture toughness by using a hybrid mixture of multiwalled nanotubes and graphene platelets as reinforcement phase in Al2O3 matrix. The role of GPLs and CNTs has been investigated separately. Wang et al.25 prepared graphene nanosheets/Al2O3 composites by SPS and reported an increase of 53% in their fracture toughness. Porwal et al.26 fabricated graphene-reinforced alumina nanocomposites by SPS also and found that toughness increased in all cases compared to that pure alumina (2.80 ± 0.17 MPa∙m1/2). The highest value of toughness was reported in composites of Al2O3 + 0.8 vol% graphene (3.90 ± 0.13 MPa∙m1/2) although in detrimental of the hardness. Kim et al.27 studied graphene/ alumina nanocomposites and reported an enhancement of fracture toughness of 75% (5.60 MPa∙m1/2) compared with that in pure Al2O3. Yazdani et al.24 found that the average fracture toughness of the nanocomposites reached up to 5.7 MPa∙m1/2, against 3.5 MPa∙m1/2 of the pure alumina. Liu et al.1 achieved an increase about 27.20% in fracture toughness for the Al2O3 ceramic composites by adding GPLs. Centeno et al.28 reported that the addition of 2 wt% of graphene in the alumina composite gave rise to a fracture toughness increase up to values as high as 7 MPa∙m1/2. Rincon et al29. have reported the mechanical properties and electrical conductivities of alumina-zirconia composites reinforced with either graphene-oxide or carbon nanofibers. Their mechanical properties changed modestly, with slight decrease of hardness and tiny increment of the fracture toughness. On the contrary, the electrical conductivity increases remarkably, reaching the maximum value for 2 vol% graphene-oxide concentration. The main goal of this study is the analysis of the mechanical enhancement of hardness and fracture toughness in graphene-oxide reinforced alumina ceramics and compared the results with those found in carbon-fiber-reinforced alumina composites. All the composites are fabricated by spark plasma sintering. Since fabrication of the second composite is by far much easier than fabrication of the first one, a critical analysis of the comparison can be useful to decide whether it is worth fabricating graphene-reinforced ceramic composites commercially despite their much higher economical cost. 2. Experimental procedure 2.1 Starting materials -alumina powder (Celarox, Conclea HPA05), with an average particle size of 0.35 µm and a surface area of 9.5 m2/g, was used in this study. Graphene oxide (Nanoinnova Tecnologies), with a thickness of 1-4 µm, a length of 0.7-1.2 nm and a surface area of 1030 m2/g was used as a secondary phase. On the other hand, carbon nanofibers (Grupo Antolin) with an outer diameter of 20-80 nm, a length of more than 30 μm and a surface area of 150-200 m2/g were also used as a secondary phase in another alumina-reinforced composite. A simple calculation shows that the critical volume concentration to cover all the grain boundaries with graphene is given by 1 6 1 −      +  d with ,  and d standing for the graphene density, its surface area and the average grain size of alumina respectively. Assuming that the density of graphene is 2.2 gcm-3, the surface area given above and d500 nm, the optimal ratio for covering all the grain boundaries with graphene would be approximately 0.5 vol% graphene versus 99.5 vol% alumina. Assuming a similar density for the carbon fibers, a similar ratio was considered. However, in the case of fibers, the aspect ratio of the fibers and the possible agglomeration of them cannot assess a full overlapping of the boundaries in this second case. Similarly, agglomeration and losses during sample sintering can take place. To the purpose of fabricating a composite with the optimal mechanical properties, a ratio close to the theoretical one is chosen: 2 vol% graphene. This is consistent with the achievements reported by Maensiri et al. commented previously23. 2.2 Powder mixture preparation Firstly, graphene oxide was dispersed in water and was further sonicated for 2 min and alumina was also sonicated using different times to this end. Solution had a pH of 10 and it was constant during this process. The best conditions to prepare the solution was 45% vol and for 5 minutes of ultrasonication. The final solution was lyophilized. The resultant mixture of powders was 98 % vol -alumina (A) and 2 % vol graphene oxide (GO). This mixture is denoted in this study AGO. In the case of powder nanofibers the powder mixture consisted of 98 vol% - alumina (A) and 2 vol % carbon nanofibers (CNF) and they were prepared similarly. This mixture is denoted in this study ACNF. Pure alumina was prepared from the as-received commercial powders for the sake of comparison. This powder is denoted as A. 2.3 Spark Plasma Sintering and characterization Bulk composite samples were sintered using a SPS process (Dr. Sinter Lab Inc., Model 515 S, Kanagawa, Japan (pulsed high DC current 20V, 0-1500 A). The powder mixtures were poured into a graphite die of 15 mm in diameter. A sheet of graphitic paper was placed between the punch and the powder and between the die and the powder for easy removal of the sintered sample. The sintering process was conducted under a vacuum of 4-6 Pa. A uniaxial pressure of 75 MPa was applied throughout the sintering cycle. The sintering temperature was increased to 1300 ºC at a rate of 100 ºC/min. The holding time was 5 min. The temperature was measured and controlled using an optical pyrometer. Shrinkage, displacement, heating current, and voltage were also recorded during the sintering process. The sintered samples were grounded and polished to 0.5 µm using SiC paper and diamond suspension. The bulk density of the samples was measured by the Archimedes method with distilled water as the immersion medium using 3.98, 2.1 and 2.1 g/cm3 as densities of alumina, graphene oxide and carbon nanofibers respectively. The relative density was calculated by dividing the bulk density with the theoretical density of the powder mixture. Spectroscopy Raman (LabRAM Horiba Jobin Yvon) was carried out in order to study the composition of sintered samples, using a wavelength of 532 nm and a range of Raman shift between 200 and 3500 cm-1. Scanning Electron Microscope (SEM) specimens from the asprepared composites were prepared using conventional methods involving successive steps of grinding and polishing. The specimens were studied using a conventional 30 kV SEM (Hitachi, S5200) in the mode of secondary electrons. Vickers hardness and toughness tests were carried out using a 5 kg-force (49 N) and a time of 10 s. An optical confocal microcope (Leica DCM-3D) was used in order to study form and length of indentions and cracks. It was necessary to etch specimens at 1200 ºC in argon atmosphere for grain boundaries SEM observation. 3. Results and discussion After Spark Plasma Sintering, the relative density of all the sintered composites of alumina was around 99%, in the good range for reproducibility of mechanical properties. Figure 1 shows SEM images of monolithic alumina (A), graphene oxidereinforced alumina (AGO) and carbon nanofiber-reinforced alumina (ACNF) respectively after sintering by SPS. The grain size and the form factor are shown in the Table 1. The AGO specimen displays an average grain size smaller than the one in pure alumina specimen. This is very likely to be a consequence of the inhibition of grain size induced by the graphene platelets. On the contrary, the grain size in the ACNF specimen is similar to that found in the monolithic alumina sample within the experimental uncertainty. This fact can be related to the much less specific surface of the nanofibers, which are less efficient to cover all the grain surfaces. Some evidence of particle pull-out during sample polishing is also observed. The form factor does not seem to depend on the nature of the second phase added to the alumina powders and it corresponds to equiaxial alumina grains as displayed in figure 1. Raman scattering is a very adequate technique for characterization of carbonbased materials. Raman spectra for all the specimens under study and the powders prior to sintering are displayed in figure 2. The presence of a peak at around 1582 cm-1 is associated to the first-order band (E2g2) of single hexagonal crystal graphite (G mode). A band around 1357 cm-1 is called the D mode by disorder-induced scattering, which is due to imperfection or lack of hexagonal symmetry in the carbon structure, resulting in breaking the k-momentum conservation. Two weak bands may be detected at 42 (E2g) and 2724 cm-1. Our Raman spectra are compatible with these features. The increase in the peak at around 2724 cm-1 observed in figures 2b and 2d are associated to oxidation of the carbon phase. This fact shows that graphene has oxidized and probably so does the carbon nanofibers, at least partially23, 27. The hardness and toughness of the specimens were measured by means of the indentation technique (Buehler, model Micromet 5103) with a conventional diamond pyramid indenter. The diagonals of each indentation were imaged by optical microscopy (figures 3a-c). The applied load was 49 N for 10 s according to the standard ASTM E92-72. Length (c) and length (a) were measured and their average values are shown in the table 2. For median / radial cracks the value of c/a ≥ 2.5. In this study c/a  2.7 and therefore the indentions are median/radial cracks. Anstis et al.30 proposed a model to evaluate the fracture toughness in the case of radial/median cracks indentations valid with c > a, for sufficiently well-developed cracks:             =2/3 2/1 016.0 c P H E KIC (1) Where KIC is the fracture toughness, E is the Young modulus, H is the hardness (determined from the load and the size of the indentation imprint), P is the applied load and c the crack length. Table 2 displays the different indentation lengths as well as the values of hardness and toughness for all the sample sets considered in this work. All the composites of alumina exhibit systematically a slightly lower hardness compared to that of pure alumina. In contrast, the final fracture toughness is somewhat higher than the monolithic alumina one. These values are in reasonable agreement with those reported by Kim et al. in their graphene-oxide-reinforced alumina composites27. These authors report values in-between 4.5-4.0 MPam1/2 determined by the SEVNB method. The dependence with the amount of graphene is weak in the range from 0.5 up to 2 vol%. Our values are also consistent with those reported by Centeno et al. determined by the indentation method28(4 MPam1/2). However, the reported values of hardness by these authors are slightly smaller (16 GPa). Regarding the ACNF composite, the hardness measured in our SPS specimens 20.4 GPa is 26% higher than that reported by Maensiri et al, whereas the fracture toughness found in this study is 64% higher than the one found in the specimens prepared by hot-pressing23. One possible explanation for this may be related to the sintering route: the SPS technique reduces drastically the dwell time of the ceramic powders. Therefore, the possible degradation of nanofibers is probably reduced accordingly. In fact, both SPS specimens and hot-pressing ones were consolidated in graphite dies but the SPS ones were heated for 1 min dwell-time, while one hour dwell time was necessary for the second ones. When comparing critically the AGO composites with the ACNF ones, it is possible to conclude that there is not a significant advantage in using graphene oxide for mechanical properties enhancement. Within the limits of the experimental scatter, carbon nanofibers are more efficient to this end. Future research on this matter must include the determination of the electrical conductivity and compared with the values of this quantity observed in graphene-oxide-reinforced alumina composites. 4. Conclusions Fully-dense graphene-reinforced and carbon-fiber-reinforced alumina composites have been fabricated by SPS. The analysis of the room-temperature hardness and toughness allows concluding that both phases give rise to a similar mechanical response with a slight advantage for the nanofiber-reinforced composite. SPS technique is rather efficient to preserve the integrity of a sufficient number of graphene oxide platelets or carbon nanofibers to enhance the mechanical response compared to other conventional sintering procedures. Acknowledgements The authors acknowledge the financial support awarded by the Regional project of Excellence of ''Junta de Andalucía'' P12-FQM-1079 and the Spanish MINECO and FEDER funds through the project MAT2015-71411-R. The help of Saint Job is especially acknowledged. FIGURE CAPTIONS Figure 1: SEM micrographs for pure alumina (a), the AGO specimen (b) and the ANF sample (c). Figure 2: Raman spectra: (a) starting powder of alumina (98 %) + graphene oxide (2 %), (b) sample sintered at 1300 ºC from this powder, (c) starting powder of alumina (98%) + carbon nanofibers (2%) and (d) sample sintered at 1300 ºC from this powder. Figure 3: Indentation imprints corresponding to the A (figure 3a), AGO (figure 3b) and ACNF-sintered (figure 3c) composites. FIGURE 1 (a) (b) (c) FIGURE 2 (a) (b) (c) (d) FIGURE 3 (a) (b) (c)