Specific surface area and carbon content of specimens
Materials with nano-scale components as reinforcement phase are adding new dimensions to composite materials, thereby fascilating major improvement in functional and structural properties. Ceramic nanocomposites are the blends of different ceramic matrices with nanometer sized functional particles. The advantages of these nanocomposites include: improved mechanical properties, surface properties, decreased permeability to gases, water and hydrocarbons, higher thermal stability and heat distortion temperature, higher chemical resistance, smoother surface appearance and higher thermal conductivity. The incorporation of only a few percent of nano-sized particles makes great property changes and formerly unachievable property combinations possible.
The ceramic nanocomposites can be devided into into three types; intragranular composite, intergranular composite and nano/nano composite (Niihara, 1991) according to their microstructures. In intra and intergranular nano-composite, the nano-sized particles are dispersed mainly within the matrix grains or at the grain boundaries of the matrix, respectively. The aim of these composites is to improve the mechanical properties such as hardness, facture strength, toughness and also high temperature mechanical properties such as hardness, strength, creep and fatigue facture resistances. On the other hand, nano/nano composites are composed of the dispersoids and matrix grains within the nanometer size. The primary purpose of this type of nano-composite is to add new functions such as machinability and super plasticity like metals to ceramics.
1.1. Synthesis of ceramic nanocomposite
Chemical Vapor Deposition (CVD) is a very preferable method to disperse the nano-sized second phases into the matrix grains or at the grain boundaries (Niihara & Hirai, 1986). Si3N4/TiN was the first system prepared by CVD. However, the CVD process is not applicable to fabricate the large and complex shaped component for the mass production and also it is very expensive. Processing route is another technique to prepare ceramic nanocomposites. Following the initial work of (Niihara & Nakahira, 1988), several research groups have tried to synthesize the nanocomposites using processing route. There are several methods involve in processing route such as conventional powder processing (Borsa et al., 1999; Carroll et al., 1996; Niihara et al., 1989; Shapiro et al., 2009), sol-gel processing (Liu et al., 2006; Xu et al., 1994) and polymer processing (Borsa & Brook, 1995; Galusek et al., 2007). The ceramic nanocomposites can be synthesized using microwave plasma (Vollath et al., 1997; Vollath & Szabó, 2006). In this process, the matrix material is coated with a layer of second phase material. The main advantage of this technique is that the reaction product does not form hard agglomerates because of the specific conditions during synthesis. Oxide and nitride based ceramic nanocomposites have been synthesized using microwave plasma technique (Vollath & Sickafus, 1993; Vollath & Sickafus, 1992). Recently developed sparked plasma synthesis (SPS) is an advanced technique for the synthesis of ceramic nanocomposites. The technique has a great advantage over conventional sintering technique, since the whole process can be complete in a few minutes, thus allowing nanosized grains of the sintering powder, and its crystal structure as well, to be retained in the sintered body. Various ceramic nanocomposites have been synthesized using SPS in literature (Dusza et al., 2009; Kumari et al., 2009; Trombini et al., 2009).
Furthermore, using all the above synthesis techniques for preparing ceramic nanocomposite, the nanoscaled reinforced particles are difficult to disperse uniformly on the micro-scale matrix particles. This problem is ascribed to the fact that nano-scale particles agglomerate easily due to the interaction between the particles. The agglomeration promote the generation of voids during the densification and microstructural in homogeneity. Kunii & Levenspiel, 1977 have recognized that a fluidized bed reactor can supply an environment with a uniform temperature and concentration of the coating precursor, which can provide the possibility of a good dispersion of reinforcing nano-particles in a matrix. A precursor vaporized at a low temperature is the major characteristic of the metal–organic chemical vapor deposition (MOCVD) process. The combination of conventional fluidized bed technology with standard chemical vapor deposition has been proven to be an effective method to deposit particles (Chen & Wei, 2002; Tsugeki et al., 1993).
1.2. Characteristics of ceramic nanocomposite
Microstructure and mechanical properties such as micro hardness, facture strength, and facture toughness are the main characteristics of ceramic nanocomposite materials. Apart from this, the nanocomposites also exhibit electro conductive, wear resistance, creep resistance and high temperature performance. One of the major characteristics and the greatest disadvantages of ceramics is their brittleness. Researchers have tried to compensate these disadvantages and to improve the strength of the nanocomposite. It has been observed that the facture strength of the brittle materials can only be improved by the increase in facture toughness or by reduction in critical flaw size (Davidge, 1979). Therefore, much effort has been invested in advanced processing technology to reduce the size and density of processing flaws (Lange, 1989). However, the design of tougher, flaw-tolerant ceramics is a more interesting approach for wider industrial applications. The facture toughness can be increased by incorporating various energy-dissipating components such as whiskers, platelets or particles into the ceramic microstructure (Becher, 1991; Lange, 1973). The reinforcements serve to deflect the crack or to provide bridging elements hindering further opening of the crack. Another concept is to incorporate metallic ligaments into the ceramic matrix (Sigl et al., 1988) to form crack bridging elements that absorb energy by plastic deformation.
1.3. Mechanisms of ceramic nanocomposite
In 1997, Sternitzke reviewed the modeling (strengthening and toughening mechanisms) of nanocomposites and divided the mechanisms into three groups; c-mechanism, K-mechanism and grain boundary strengthening mechanism. The c-mechanism is based on the fact that the matrix becomes refined following the adding of nano-sized silicon carbide. A refinement of the grain size leads to smaller critical flaw size and higher strength. K-mechanism relates to R-curve behavior, crack deflection, and crack bowing during a crack extension. Ohji et al., 1998 proposed a particle-bridge mechanism, where there exists crackface shielding when nano-size particles bridge the crack surfaces. Crack deflection and crack bowing are related to the interactions of a crack front with second-phase inclusions which depend on the differences in the thermoelastic properties of the matrix and inclusions. Levin et al., 1995 and Sekino et al., 1997 explained that the strengthening mechanism of nanocomposites results from residual stresses around the second-phase particles, and matrix weakening and grain boundary strengthening produce the change of the fracture mode. Only nano-sized dispersed particles can improve the strength and the fracture toughness.
2. Alumina/chromium carbide nanocomposite
Alumina is one of the important materials in structural ceramic application because of its excellent mechanical properties, good chemical stability and high temperature characteristics (Ghate et al., 1975; Wang & Hsu, 1996). Its intrinsic brittleness and relatively poor reliability however, made the toughening of alumina ceramics an important and challenging area of research now days. The incorporation of secondary ceramic reinforcement phases (e.g. particulates, fibers, or platelets) has been proven to be an easy, safe and economically toughening technique for alumina ceramics (Chou & Green, 1992; Lio et al., 1989). There are several approaches by the researchers to toughen the alumina matrix with addition of different secondary phases. Among all the approaches, SiC reinforcement particles on Al2O3 composites mostly improve the mechanical properties (Anya, 1999; Sternitzke, 1997). But the toughness enhancement is not significant for these composites. On the other hand, incorporating chromium carbide particles into Al2O3 matrix has been successfully achieved for toughening purposes (Fu et al., 1994; Huang et al., 2000; Lii et al., 1999). These aluminum oxide/chromium carbide composites demonstrate superior mechanical properties to those of matrix phase due to the high Young’s modulus of chromium carbide and its outstanding ability to resist high temperature erosion at temperatures up to 1000 C. In addition, the good electrical conductivity of chromium carbide makes electrical discharge machining possible.
Besides second phase particles, solid solution strengthening is another mechanism for the Al2O3 matrix. Chromia (Cr2O3) has been used to improve the physical properties of Al2O3 (Bradt, 1967). As Cr2O3 has the same corundum crystal structure similar to Al2O3, Al2O3–Cr2O3 can form substitutional solid solution in all ranges at high temperature. The addition of Cr2O3 was found to increase the hardness, tensile strength, and thermal shock resistance of Al2O3.
This chapter will emphasize on the processing of Alumina/Chromium carbide ceramic matrix composite by metal oxide chemical vapor deposition (MOCVD) in fluidized bed and analysis of these composites using various characterization techniques. The synthesis parameters, microstructure and mechanical properties of the alumina/chromium carbide prepared by three different processes such as (i) Hot press sintering, (ii) Carbothermal reduction, and (iii) Spark plasma sintering, have been investigated and summarized in details in the following sections.
2.1. Experimental details
This section describes the raw powders, different types of synthesis procedures such as MOCVD, Hot press sintering (HP), Carbothermal reduction, Spark plasma sintering (SPS), and various characterization techniques such as X-ray diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), and energy dispersive spectroscopy (EDS), X-ray photoelectron spectroscopy (XPS) and electron energy loss spectroscopy (EELS) used in the current work. The flow chart showing the detailed experimental procedure is given in Fig.1.
2.1.1. Powder and sample preparation
Chromium hexacarbonyl (Cr(CO)6, 99%, Strem Chemicals Co., USA) was used as the precursor of chromium oxide in the MOCVD process. Aluminum oxide powder with the average particle size of about 0.2–0.4 μm (A16-SG, Alcoa, USA) was used as the matrix powder. A vaporized precursor carried by He gas (99.9% pure) was introduced into the fluidized bed reactor for the MOCVD process. Based on the results of Lander & Germer, 1947, the precursor container was kept at 75˚C in a vacuum (10 torr) in the present experiment. To investigate the influence of the temperature on the product deposited on the fluidizing Al2O3 particles, the Cr(CO)6 vapor was decomposed in the fluidized chamber at 300 and 400 °C, respectively. The apparatus consisted of six main components: gas supply, MOCVD precursor, fluidized bed reactor, power supply, temperature controller, cold trap and vacuum system, as shown schematically in Fig. 2. (Lin et al., 2006a). The pressure in the chamber was measured with a pressure meter. A rotary vacuum pump was set up with a cold trapping system. The temperature controllers (TC) were connected to a heating system that controlled the temperature of the fluidized reactor and precursor container.
2.1.2. Characterization of the fluidized powder
The fluidized powders were analyzed by x-ray diffraction (XRD; Rigaku D/MaxII, Japan), and transmission electron microscopy (TEM; Hitachi FE-2000, Japan) equipped with energy dispersive x-ray spectroscopy (EDS; UK). X-ray photoelectron spectroscopy (XPS; VG Scientific 210, UK) was used to determine the coating phases by binding energy. A Brunauer–Emmett–Teller (BET) instrument Micromeritics Gemini 2360, USA) was used to measure the specific surface area by nitrogen adsorption, and C/O analyzer (LECO CS-244, USA) was used to measure the carbon content of coated powder by a combustion method. Fig. 3 shows the XRD patterns of decomposed Cr (CO)6 prepared in fluidized bed at 300˚C and 400˚C. It is observed that at 300˚C, the powder is completely amorphous in nature, where as at 400˚C, the powder consists of Cr2O3 along with CrC1-x phase (Lin et al., 2006b).
The pyrolysis of precursor Cr(CO)6 at 300˚ C results the formation of Cr2O3, C-Cr and C-C bonds observed from XPS spectra (Fig. 4). This figure displays the XPS spectra of the Cr 2p and C 1s regions of the as-deposited powder prepared at 300 °C. Fig. 4a shows two peaks corresponding to the spin-orbit splitting 2p1/2 (right) and 2p3/2 (left) of Cr with bonding energies of 586.3 and 576.6 eV, respectively. The band shift of these two peaks is 9.7 eV, which is in good agreement with previously reported data (Wagner et al., 1979) and confirms the existence of Cr2O3 particles in the as-deposited powder. The XPS spectra of the C 1s regions in Fig. 4b provide the evidence for which at least two forms of carbon in the as-deposited powder exist. One is free carbon (C–C) at 284.6 eV, and the other carbon bonded to chromium atoms (C–Cr) at 283.5 eV.
The decomposed precursor deposits uniformly over Al2O3 and the deposited nano particles (~ 30 nm) are shown in the TEM micrograph (Fig. 5a). The insert shows the TEM pattern of the deposited nano particles, which is amorphous in nature. From the EDS spectrum shown in Fig. 5b, it is confirmed that the nanoparticles consist of mostly Cr and O along with small amount of C. The presence of Cu is the contribution from Cu grid.
Table 1 shows the carbon content of the samples prepared at different fluidized temperature. The carbon contain is 0.75 and 0.23% for the samples fluidized at 300 and 400˚C, respectively. It indicates that the as-deposited amorphous powder prepared at 300 ˚C has more carbon content than the crystalline powders prepared at 400˚C. This was probably because of the amorphous powder has specific surface area (25.5m2/g) larger than the crystalline powder (11.2 m2/g).
|BET surface area|
|300 C, 2h|
400 C, 2h
3. Carbothermal treatment of Cr2O3 particles dispersed on Al2O3 particles
The carbothermal treatment of the prepared Cr2O3/Al2O3 powders was first performed in a graphite furnace at a variety of temperatures from 700 to 1150 ˚C in a vacuum condition (10−3 Torr) for 2h. Fig. 6. shows the XRD patterns of the fluidizing powders treated from 700 to 1150 ◦C in the graphite furnace for 2 h. The Cr2O3 peaks seen in the XRD patterns clearly indicate that the amorphous Cr2O3 has crystallized, but does not carbonize at temperatures under 1000 ˚C. As reported by Bouzy et al., 1993 and Bewilogua et al., 1988, an annealing treatment causes a transformation of the metastable carbide CrC1−x into the stable carbide phase Cr3C2. Here, Cr3C2 peaks are not observed in the XRD patterns shown in Fig. 6a–6d, consequently, suggesting that the CrC1−x content of the decomposed Cr (CO)6 is too little to be found in XRD patterns. According to the XRD pattern shown in Fig. 6e, Cr2O3 reacts with carbon and transforms into Cr3C2 when the treatment temperature is 1150 ˚C. This carbothermal reaction process ( Storms, 1967) can be shown as the following equation:
3.1. The formations of Cr-carbide/Al2O3 nanocomposite and Cr2O3/Al2O3 solid solution
After carbothermal treatment, the fluidized powders undergone hot pressed sintering at 1400˚ C. For this process, the 200 mesh fluidized powder was first die-pressed, and then put into a BN-coated graphite die and hot-pressed at a pressure of 25MPa in a HP furnace (High-multi 5000, Fujidempa Kogyo Co., Ltd., Japan) at 1400˚C under vacuum (5×10−4 torr). The detailed sintering conditions of the different samples are listed in Table 2.
|Samples||Sintering powders||Sintering conditions|
|Pure alumina powder|
Composite powders fluidized at 300 C for 1 h
|1400C/1h, pressure 25 MPa|
1000C/1h, 1400C/1h, 25 MPa
1150C/1h, 1400C/1h, 25 MPa
1400C/1h, pressure 25 MPa
Fig. 7 shows the XRD patterns of the samples thermally treated under different conditions. The Si peaks shown in the XRD pattern were used for calibration. First, Fig. 7a and 7b indicate that the Al2O3 peaks for sample 1 shifted to lower angles than those for ALO. For pure ALO, the pure alumina powder was hot pressed at 1400˚C, while for sample 1, the fluidized powder was pre-sintered at 1000 ˚C before the hot pressing at 1400 ˚C. During the pre-sintering at 1000 ˚C, most of the Cr2O3 reacted with Al2O3 to form an Al2O3–Cr2O3 Solid solution and thus the Al2O3 peaks shifted to lower angles in the XRD pattern. Furthermore, the color of the pieces is dark red (Nassau, 1983) as observed in sample 1, due to the (Al, Cr)2O3 solid solution reaction product. Secondly, a comparison with the peaks of ALO reveals that the Al2O3 diffraction peaks of sample 2, as shown in Fig. 7c, shift to lower angles, occurring simultaneously with the peaks of Cr3C2. Consequently, besides forming a solid solution, some of the Cr2O3 reacts with carbon as equation (Eq. 1) to form chromium carbide Cr3C2 when the powders have been pre-sintered at 1150 ˚C (Lin et al., 2007). Finally, Fig. 7d shows the XRD patterns of sample 3, which is similar to the results of sample 2. In addition to forming a solid solution, some of the Cr2O3 was also carbonized. However, it transformed into mixed phases of Cr3C2 and Cr7C3 when the fluidized powders were hot pressed at 1400 ˚C. Berger et al., 2001 reports that Cr7C3 is formed at an elevated temperature (>1150 ˚C). Moreover, the TG/DTA results of our paper (Lin et al., 2005) also indicate that the generation temperature of Cr7C3 is about 1170 ˚C. This carbothermal reaction process (Bradt, 1967) can be shown as reaction (Eq. 2). Consequently, not only Cr3C2, but also Cr7C3 is formed in sample 3, hot pressed at 1400 ˚C.
Fig. 8a shows the SEM micrograph of pure Al2O3 and 8b is the High Angle Annular Dark Field scanning transmission electron microscopy (HAADF STEM) image of sample 1, showing almost all the Cr3+ replaced the Al3+ and formed an Al2O3–Cr2O3 solid solution. According to Bondioli et al., 2000, at temperatures of over 1000 ˚C the complete ranges of substitutional solid solutions are obtained. Fig. 8c and 8d are the HAADF STEM images of samples 2 and 3, respectively, showing that in addition to forming a solid solution, nanosized chromium carbide particles also disperse uniformly in the alumina matrix. Comparing the HAADF STEM images of samples, sample 2 has smaller alumina grain size (0.7m) than others (sample 1: 1.5m, sample 3: 0.9m.) This is because it has much nanosized particles on the Al2O3 grain boundaries, which effectively inhibit the grain growth. The volume percentages of the reinforced particles showing in Fig. 8c and 8dare 4.5 and 1.5 vol%, respectively. Conversely, the Al2O3 grains of sample 1 have a larger growth size as a result of the diffusion of Cr atoms inducing grain boundary migration (DIGM) (Han et al., 1995). The drive force for the migration is believed to rise from the coherency strain in the Cr2O3 diffusion zone in front of the migrating grain boundaries.
There are two HRTEM micrographs of interfaces shown in Fig. 9. One is the interface between the Cr3C2 and Al2O3 and the other is the interface between Cr7C3 and Al2O3. The interface between Al2O3 and Cr3C2 is non-coherent, as shown in Fig. 9a. The smaller illustrations in Fig. 9a show Al2O3 with hexagonal and Cr3C2 with orthorhombic structure. The lattice spacing of the Al2O3 (0 1 4) is 0.255 nm and that of the Cr3C2 (0 0 2) is 0.115 nm. The spacing difference between these two planes is large. However, the interface between Al2O3 and Cr7C3 is semi-coherent, as shown in Fig. 9b. Structures of both Al2O3 and Cr7C3 are hexagonal. The lattice spacing of the Al2O3 plane (1 0 2) and Cr7C3 plane (20 1) are 0.349 and 0.360 nm, respectively. There is little difference between the lattice spacing of these two hexagonal planes.
3.2. Carbothermal reduction of Cr2O3 using methane as carbonization source
The conventional route for Cr3C2 composite synthesis mainly involves the reduction of chromium oxide by methane gas (Anacleto & Ostrovski, 2004). However, besides the three stable chromium carbides, Cr3C2, Cr7C3, and Cr23C6, several metastable carbides such as Cr3C, CrC1−y, Cr5C2 and Cr3C2−x also have been reported, which emanate during different preparation methods and for different Cr/C contents (Bewilogua et al., 1988; Bouzy et al., 1993; Inoue & Masumoto, 1979). Lerch & Rousset, 1994 indicated that the CrO1.9 with high surface area (200–350 m2/g) reduced at 700 °C in an atmosphere of methane and hydrogen mixture leading to the formation of stable Cr3C2 and metastable Cr3C2−x. The metastable Cr3C2−x powder was also prepared by Loubière et al., 1995 & 1996 using CH4–H2 atmosphere from the carburization of metastable chromium oxide. The metastable Cr3C2−x is a Re3B-type structure, readily formed by carburization of high specific surface area chromium oxide and usually co-existent with Cr2O3, Cr3C2 and free carbon.
In the previous sections, we have discussed the synthesis of Cr2O3/Al2O3 composite powders by the decomposition of chromium hexacarbonyl in a fluidized bed. This was followed by the hot-pressing of the composite powders in a graphite mode eventually leading to its transformation into a Cr3C2/Al2O3 nanocomposite. One of the major drawbacks in this synthesis process was the formation of undesirable Al2O3- Cr2O3 solid solution at a high temperature, which leads to abnormal grain growth of the Al2O3 matrix and eventually decreased the desired Cr3C2 contents in the composite. In order to carburize the chromium oxide completely, a new approach is reported by using the carburization process of nanosized chromium oxide formed from a metal-organic chromium hexacarbonyl precursor in a carburizing mixture of CH4/H2 gas. For this process, the as received powder after fluidization was carburized in an alumina tube furnace in the presence of a methane–hydrogen mixture (1:9) at 700–850 °C for 5 h. The H2 restrain the CH4 to decompose fast and using CH4 as carbon source, the carbothermal reduction reaction becomes
Fig. 10 shows the XRD patterns of the decomposed powder carburized at 700–850 °C for 5 h in CH4–H2 gas. At the carburization temperature of 700 °C, the pattern displays sharp peaks that index with the Cr2O3 phase. At the increased carburization temperature of 750 °C, the intensity of Cr2O3 peaks decreases significantly accompanied by some incipient low intensity peaks which could be indexed with metastable Cr3C2−x. When the temperature reaches to 800 °C, the Cr2O3 phase disappears, concomitant with a complete transformation to the Cr3C2−x phase. Further increase of the carburization temperature to 850 °C, leads to the formation of the stable Cr3C2 phase along with a small amount of Cr3C2−x metastable phase (Wang et al., 2010).
For further validation of carburization process, the powders for different carburization temperature were characterized by TEM. Fig. 11a shows the TEM micrograph of the powder treated at 700 °C, where the particle size is found to be around 40 nm. The diffraction pattern verifies its phase for the chromium oxide Cr2O3, (rhombohedral structure). The high resolution image shows a clear atomic image inside the powder. A thin amorphous layer is visible in a different contrast outside the boundary surface. This can be attributed to the adsorption of carbon deposit on the powder surface. Fig. 11b shows the TEM and HRTEM of powdered samples carburized at 800 °C for 5 h. The spherical shape has transformed to faceted irregular grain. The diffraction pattern confirms that the phase is attributed to metastable Cr3C2−x, which is in accordance with the XRD result. The HRTEM image shows that there are several carbon nanofilms formed around the surface of the powder surface. The Fast Fourier transform (FFT) pattern of these layers, shows two obvious halo spots that are attributed to the graphene structure of carbon nanofilms. The TEM micrographs of the powders, carburizated at 850 °C, show that the powder is encapsulated by layers of graphene, as shown in Fig. 11c. The diffraction pattern verifies that the powder comprises the stable Cr3C2 phase.
The phase transformation from Cr2O3 to Cr3C2 via an intermediate state Cr3C2-x has been identified using electron-energy loss spectroscopy (EELS). The EELS of prepared samples were characterized by field emission gun scanning transmission electron microscopy (FEG-STEM ) equipped with energy dispersive X-ray detector (EDS) and energy filter (Gatan). Fig.12 shows the EELS spectra of Cr L edge for the sample of the decomposed precursor and the samples carbonized at 800 °C and 850 °C. The chromium L edges have features comprising two sharp L2 and L3, known as “white lines” (Lozzi et al., 1991) and a continuum background following the edge. The sharp double peaks are due to transitions from the 2p1/2 and 2p3/2 core levels towards 3d states above E f. The sharp peaks of L2 and L3 at threshold are due to the transitions from the 2pl/2 core levels to 3d3/2 states and the 2p3/2 core levels to 3d3/2 3d5/2, respectively. The L3/L2 white line ratio is correlated to the electron occupancy and spin pairing in 3d band. The variations of the ratio are due to the interaction between chromium and surrounding atoms. A change of the L2, L3 edge intensities in the samples indicates a variation of the d holes because the edge intensity is proportional to the number of the empty final states available. The ratio of L3/L2 of sample shown in Fig. 12a is higher than that of carbonized samples. According to the results of Arévalo-López & Alario-Franco, 2009, the ratio of L3/L2 for the Cr2O3 is about 1.6 and from the report of Fan et al., 1999, the ratio of L3/L2 for the Chromium carbide is about 1. A comparison between Fig.12b and 12c shows L3 edge of the sample carbonized at 850 °C is 577.72eV, while that of sample carbonized at 800 °C is 575.42eV. The shift in energy (2.3 eV) is called the chemical shift. This result is due to the transformation from the metastable carbide to stable carbide, because the EELS represent the difference in energy between a core-level initial state and the lowest energy final state of an excited electron. But the energy loss close to 577.72 eV for the decomposed precursor (Daulton et al., 2002), which consists of Cr2O3 is same as the sample carbonized at 850˚ C. It is observed that there is no chemical shift between this two species because Cr2O3 and Cr3C2 are both chemically stable compounds.
The π electron formed sp2 bonding observed from the EELS spectra of C K edges as shown in Fig. 12. The peak of sp2 bonding is sharper when the carbonized temperature increases from 800 °C to 850 °C. The sharpening of sp2 peak is attributed to the formation of graphite like carbon according to Fan et al., 1999. In Cr3C2, C is graphite like rather than diamond like, which comprising of sp3 bonded carbon atoms. For the as decomposed precursor, the content of carbon was little and formed amorphous phase. More carbon produced from the pyrolysis of methane coated on the Cr2O3 surface and then formed graphite phase and metastable carbide of Cr3C2-x, when the sample was treated at 800 °C for 5 hours. The increase of treated temperature to 850 °C, enhanced the process of carbonizing reaction leading to more graphite formation followed by transformation of metastable Cr3C2-x to stable carbide of Cr3C2 (Lin et al., 2011).
Based on these above results, one could hypothesize the following phase transformation steps: In the first step, the methane decomposes to elemental carbon and hydrogen (Muradov, 1998 & 2001). The carbon reacts with the Cr2O3 at temperatures higher than 800 °C, and produce partial Cr3C2 phase and CO(g). The CO(g) can also carburize the Cr2O3 further. There are three kinds of carbon sources in the system viz. free carbon from the decomposition of methane, CO from above reaction and C-C bonding from the precursor. Finally, the unreacted redundant carbon sources would deposit on the surface of the powder as the lamellar graphene. The graphene deposited during the carburization has an important role for the formation of meta-stable Cr3C2-x and stable Cr3C2. Initially, carbon gets absorb on the surface of chromium oxide at low temperature. As the temperature rises, substantial amount of carbon is available, leading to metastable chromium carbide. Finally, as the temperature increases further, the redundant carbon forms the graphene layers. Further thickening of the graphene films, generates a mechanical stress (Loubière et al., 1996) which assists the transformation of meta-stable Cr3C2-x state to stable Cr3C2 phase.
3.3. Microstructure and mechanical properties of Cr3C2/Al2O3 nanocomposites
As discussed in the previous sections, the carbonization treatment of initial powders without using methane as the carbon source leads to the formation of (Cr3C2, Cr7C3)/Al2O3 nanocomposites, where as carbonization using methane as carbon source leads to Cr3C2/Al2O3 nanocomposites, which is the desired material. In order to improve the mechanical properties of these Cr3C2/Al2O3 nanocomposites, the carbonized powders were undergone hot-press sintering at 1400 ˚C for 1 hr in vacuum.
Fig. 13 shows the SEM micrographs of monolithic Al2O3 and Cr3C2/Al2O3 nanocomposites carbonized at different temperatures for 5 hours and hot-pressed at 1400˚C for 1 hr in vacuum. One can see that there is abnormal grain growth of Al2O3 shown in Fig. 13a. In case of Cr3C2/Al2O3 nanocomposite, the dispersion secondary phase, Cr3C2 is uniform (Fig. 13b) and the abnormal grain growth has been inhibited. The grain size distribution of monolithic Al2O3 and Cr3C2/Al2O3 nanocomposites has been shown in Fig. 14. The nanocomposite exhibits low grain size than that of pure alumina. It implies that addition of Cr3C2 decreases the grain size and inhibits the growth of alumina matrix.
The TEM micrograph of Cr3C2/Al2O3 nanocomposites shown in Fig.15 illustrates two types of inclusion of Cr3C2 onto the Al2O3 matrix i.e. inter-type inclusion (Fig. 15a) and intra-type inclusion (Fig. 15b). Fig. 16 shows the SEM micrographs of the fracture surface of monolithic Al2O3 and Cr3C2/Al2O3 nanocomposites under the same carbonization condition. The facture surface of pure alumina is intergranular in nature, where as transgranular facture surface is observed in Cr3C2/Al2O3 nanocomposites.
The basic mechanical properties such as hardness, fracture strength and toughness of pure alumina as well as of this nanocomposite have been analyzed and mentioned in table 3. The fracture strength was determined by 4-point bending (Shimadzu AG-IS 100KN, Japan). The toughness was measured by Single-Edge-Precracked Beam (SEPB). A precrack with depth of 0.254mm was conducted in the center of sample by a diamond cutter with a thickness of 0.15mm. It is observed that the nanocomposite exhibits improved hardness, fracture strength and facture toughness in comparison to monolithic alumina. The relationship between the microstructure and improved mechanical property has been investigated and described below.
(Indentation Fracture method)
The nanocomposites consisting of Cr3C2 particles dispersed within Al2O3 matrix grains, result in the generation of thermally induced residual stresses after hot-pressed sintering at 1400C. The mismatch in thermal expansion co-efficient between the matrix (α-Al2O3= 8.4*10-6/ °C), and the dispersed particles (Cr3C2=11.2*10-6/ °C) yields highly localized residual stresses around the particles. These stresses reduce quickly as distance from the boundary increases because of the nano-sized particles, which can generate only small defects such as dislocations shown in Fig. 17a in close vicinity to the particles (Awaji et al., 2002; Choi & Awaji, 2005). In the monolithic alumina, crack propagates along with the grain boundaries due to the existence of defects and the compress residual stresses resulting from anisotropic thermal expansion, Young’s modulus along the crystal axes, and crystallographic misorientation across the grain boundaries. The fracture toughness of grain boundaries is usually lower than that within the grains. However, in nanocomposites, the dislocations around the Cr3C2 particles release residual stresses in the Al2O3 matrix. Consequently, the defect size along the grain boundaries is reduced. Also, the dislocations serve as origins of small stress concentrations, and create nano-cracks around the propagating crack tip. These nano-cracks slightly reduce the strength of the alumina matrix, while reduction of both the residual stresses along the grain boundaries and the strength in the matrix lead to a change in the fracture mode from that of the intergranular fracture in monolithic alumina to that of transgranular fracture in nanocomposites, as shown in Fig. 16b. In addition, due to the effect of nano-cracking, the step-wise fracture surface rather than a planar cleavage plane is observed in the nanocomposites (Choi et al., 2004), as shown in Fig. 17b. The reinforcement of the Cr3C2 particles, small grain size of Al2O3 matrix and reduction of both the defect size along the grain boundaries and the tensile residual stresses in the matrix grains by dislocations result in improvement of the strength of nanocomposites.
In the case of Cr3C2-Al2O3 solid solution described in the previous section, the Cr3+ ions increase the growth rate of Al2O3 grain because of the coherency strain energy at grain boundary (Harabi & Davies, 1995; Paek et al., 1996; Riu et al., 2000). However, this phenomenon is not found in the nanocomposites. It is attributed that the secondary particles of Cr3C2 inhibits the growth of alumina matrix. The fracture strength is increased by the grain-boundary modification caused the compressive stress generated by the substitution of bigger Cr3+ ion (0.076 nm) in place of smaller Al3+ ion (0.068nm). The localized compressive stress helps in strengthening the grain boundary, and then the fracture strength increases by this effect (Lii et al., 1999). The higher toughness of the nanocomposites relative to monolithic alumina may be connected with the stepped transgranular cleavage. The direction of crack propagation within the matrix grains is affected by these dislocations.
4. Densification behavior of Al2O3-Cr2O3/Cr3C2 nanocomposites prepared by Spark Plasma Sintering
Compared with conventional hot pressed sintering, spark plasma sintering (SPS) has recently attracted increasing interest due to its ability of rapidly heating the powder compact to high temperatures and consolidating it to high densities within very short time. SPS utilizes applied pressure along with temperature and a pulsed direct current to densify materials. SPS is heralded for its promise in retaining nanostructure features, producing grain boundaries devoid of impurities as well as many other benefits (Kim et al., 2007; Munir et al., 2006; Shen et al., 2002).
As described earlier, the initial powders used for this sintering process were Al2O3-Cr2O3, which were prepared by metal organic chemical vapor deposition (MOCVD) in a fluidized bed. The nanosized composite powders fabricated in the fluidized bed for 30 and 60 minutes were named as S-30 and S-60, respectively. The pressure of reaction chamber was controlled at 10 torr, and the reaction temperature was kept at 300 °C. The fluidized powders were densified by a SPS process (SPS-515S, Shumitomo, Japan). The 1.5gram of powder (pure Al2O3, S-30, S-60) was put into a graphite mold of 15.5mm diameter. The uniaxial pressure of 50Mpa was imposed on the powder and the vacuum level was less than 6 Pa. The heating rate from room temperature to 600 °C and from 600 °C to sintering temperature was 200 °C/min and 100 °C/min, respectively with a holding time of 10 minutes. There were four different sintering temperature conditions of 1200 °C, 1250 °C, 1300 °C and 1350 °C conducted in this study.
4.1. Effect of SPS sintering temperature on the color changes of compacts
The color of compacts prepared under different SPS conditions is shown in Fig.18. It is observed that the color of compacts sintered at 1200 °C is green; the color of 1250 °C compact is purple; while that of 1300 °C and 1350 °C are black. Fig. 19 indicates the XRD patterns of the compacts processed by various SPS temperatures. Besides Al2O3 peaks, the Cr2O3 peaks are found in the compact sintered at 1200 °C. Compared with the peak of pure-Al2O3, the Al2O3 peaks of compact sintered at higher temperature have slight shifted to lower angles. It means only a little solid-solution of Al2O3-Cr2O3 has formed in this compact. This result is special to the compacts fabricated by conventional densification methods in which significant pink solid-solution of Al2O3-Cr2O3 was formed. According to Bondioli et al.,2000, the complete ranges of Al2O3-Cr2O3 solid solution can be obtained for the reaction temperature higher than 1000 °C.
It is suggested that due to the higher heating rate and short holding time, the most of Cr3+ doesn’t have enough time to diffuse into Al2O3 matrix and remains as Cr2O3 phase after the SPS process. The color of pure Cr2O3 is green, and so the composite is with green appearance. For the compact sintered at 1250 °C, the Cr2O3 peaks are disappeared in the XRD pattern because more Cr2O3 react with Al2O3 matrix to form pinker solid solution in this higher sintering temperature. According to Equation 4 given in (Kubaschewski et al. 1979), chromium carbide would form probably when the temperature is higher than 1150 °C. It is believed that there is little black carbide formed when sintered at 1250 °C, so the compact exhibits purple color. It is reasonable that the more Cr3+ diffuse into Al2O3 matrix with the increase in sintering temperature. In comparison to other compacts, more Al2O3 peaks of the compact sintered 1350 °C have shifted to lower angles observed from XRD patterns. It indicates that more solid solution is formed at a higher temperature. The amount of the solid solution of Al2O3-Cr2O3 and Cr3C2 increases with the increase of sintering temperature.
The black color of the compact prepared at 1350 °C is relevant to the formation of chromium carbide. More Cr3C2 is formed as the temperature is over than 1150 °C. Cr3C2 nanoparticle is observed for the compacts prepared at 1300 °C and 1350 °C. The TEM diffraction pattern shows the particle with Cr3C2 phase as shown in Fig. 20. Although the peak of Cr3C2 peaks are not obvious in the XRD patterns in Fig. 19., but based on the Equation 1, sintering at a higher temperature is beneficial to form Cr3C2. In our previous study (Lin et al., 2007), Cr3C2 peaks are easy to be found in the XRD patterns as compacts prepared by HP densification in which the vacuum maintained at about 10-5 torr, however in this SPS process that is about 6 pa (~10-2 torr). According to the report of Chu & Rahmel, 1981, carbon would be consumed by oxygen in a higher oxygen pressure in SPS process and inhibits the formation of Cr3C2.
4.2. Effect of SPS sintering temperature on shrinkage behavior
Fig. 21 shows the sintering shrinkage curves of the pure Al2O3, nanocomposite powders of S-30 and S-60 during the SPS densification in the temperature range 1200 °C ~1350 °C with holding time of 10minutes. The y-axis ΔL/Lo (%) represents the shrinkage profile directly during densification of the powders in real-time. The trend of Al2O3 curve becomes flat at 1200 °C, but the curves of S-30 and S-60 still trended up. This means the shrinkage Al2O3 has already stopped, but S-30 and S-60 continue shrinking during the period of holding time.
This result is consistent with the apparent density shown in Fig. 22. It shows that pure Al2O3 exhibits a higher density than S-30 and S-60, for the sintering temperature at 1200 °C and 1250 °C. The amount of shrinkage of S-30 is more significant than S-60. The temperature at which the shrinkage vanishes for S-30 and S-60 is 1300 °C and 1350 °C, respectively as shown in Fig. 21c and 21d. The curves of pure Al2O3, S-30 and S-60 are trended down for the holding temperature at 1350 °C. This signifies that all the powder compacts undergone thermal expansion after completion of the sintering shrinkage. The densities of S-30 and S-60 are increased significantly when the temperature rises to 1300 °C and 1350 °C.
Fig. 21d illustrates that the pure Al2O3 starts to densify at about 950 °C and finishes at about 1300 °C. For S-60 and S-30, the densification temperature starts at about 1050 °C and 1100 °C, separately, and finishes at 1350 °C for both. Actually S-60 finishes its densification after holding about 50 seconds. The second phase particles decrease the diffusivity of grain boundary and lattice, and result in the retardation of densification (Chae et al., 2006). According to XRD spectrum shown in Fig. 19, the delay on the densification process for the nanocomposites is due to the Cr2O3 particles coated on Al2O3 as sintered at 1200 °C. Cr2O3 plays as a second phase to inhibit the grain growth during densification. While some of Cr2O3 changes to Cr3C2 to be the second phase as temperature increases to 1300 °C and 1350 °C (Lin et al., 2011).
4.3. Effect of SPS sintering temperature on microstructures and mechanical properties
The fine grains of Al2O3-Cr2O3/Cr-carbide nanocomposites were prepared by employing spark plasma sintering (SPS) technique. The basic mechanical properties like hardness, fracture strength and toughness were analyzed. The fracture surfaces of the nanocomosites are shown in Fig. 23. It indicates that the pure alumina has larger grain size than that of S-30 and S-60 densification samples. The nanosized reinforced particles (Cr3C2) are found in the alumina matrix. S-60 densification sample has smallest grain size due the more amounts of secondary phase particles to inhibit its grain growth. The hardness, fracture strength and toughness of the S-30 and S-60 nanocomposites are present in Fig. 24. The S-60 nanocomposite exhibits highest fracture strength and toughness. The increase in hardness, fracture strength and toughness of nanocomposites over monolithic alumina is due to the strengthening of secondary phase of Cr3C2 particles and solid solution of Al2O3-Cr2O3. The co-relation between the microstructure and mechanical properties is similar to that of nacocomposites prepared using hot press sintering described in section 3.3.
Nanoscaled Cr2O3 powder with an average particle size of 20–40 nm, coated on alumina particles, has been produced by means of chemical vapor deposition in a fluidized chamber, using the pyrolysis of Cr(CO)6 precursor. The pyrolysis of Cr (CO)6 at 300 and 400 °C produces amorphous and crystalline Cr2O3 particles, respectively. The decomposed precursor consists of Cr2O3, CrC1-x and C. To prepare nanoscaled chromium carbide powder from the nanometer-sized Cr2O3, carbonizing behavior of the Cr2O3 particles was investigated using a graphite furnace at 1150 C. It was found that, when amorphous Cr2O3 powders were carbonized at 1150 °C, the powder was transformed into Cr3C2, while the crystalline Cr2O3 was transformed into a mixture of Cr7C3 and Cr3C2. These transformations were confirmed by X-ray diffraction (XRD), transmission electron microscopy (TEM), and energy dispersive spectroscopy (EDS). A solid solution of Al2O3–Cr2O3 and an Al2O3–Cr2O3/Cr3C2 nanocomposite were formed when these fluidized powders were pre-sintered at 1000 and 1150 C, respectively before hot-pressing at 1400 C. In addition, an Al2O3–Cr2O3/Cr-carbide (Cr3C2 and Cr7C3) nanocomposite was formed when the particles were directly hot pressed at 1400 C.
The undesirable solid solution leads to abnormal grain growth of the Al2O3 matrix and decreases the desired Cr3C2 contents in the composite. In order to avoid this drawback, carbothermal reduction process was adopted using methane as the carbonization source. The carburization process involved the sequential deposition of carbon on the outer surface of the Cr2O3 powder followed by carbon diffusion into the powder, leading to the formation of metastable Cr3C2−x phase and stable Cr3C2 in the temperature range 700 -850 C. The phase transformation from Cr2O3 to Cr3C2 has been characterized by XRD, X-ray photoelectron spectroscopy (XPS) and electron energy loss spectroscopy (EELS). It is observed that the formation of carbon nanofilms surrounding the carbide crystallites provides the stress and assists the phase transformation from metastable Cr3C2−x to stable Cr3C2.
The densification behavior of Al2O3-Cr2O3/Cr3C2 nanocomposites has been studied using the recently developed SPS technique. Different colors of the compacts such as green, purple and black were observed during densification process at different SPS temperatures from 1200 C to 1350 C. These changes in color were relevant to the existence of secondary phase of green Cr2O3, pink solid solution of Al2O3-Cr2O3 and black Cr3C2 at different SPS temperature. This SPS process reduces the abnormal grain growth of Al2O3, which appears during hot press sintering and Al2O3-Cr2O3/Cr3C2 nanocomposites exhibit small grain size in this process. The microstructure and basic mechanical properties such as hardness, fracture strength and toughness of pure alumina as well as this nanocomposite have been analyzed. The microstructure of dislocation, transgranular and step-wise fracture surface were observed in the nanocomposites. The nanocomposites show facture strength (780MPa) and fracture toughness (4.8 Mpa.m1/2), which is higher than monolithic alumina. In comparison with monolithic alumina, the Al2O3-Cr2O3/Cr3C2 nanocomposites exhibit higher hardness, fracture strength and toughness due to the strengthening of secondary phase of Cr3C2 particles and solid solution of Al2O3-Cr2O3.
Authors are thankful to National Science Council of Taiwan for its financial support under the contract No: 99-2923-E-006- 002-MY3 to carryout the present work.