High-Performance Metal/Carbide Composites with Far-From-Equilibrium Compositions and Controlled Microstructures

The prospect of extending existing metal-ceramic composites to those with the compositions that are far from thermodynamic equilibrium is examined. A current and pressure-assisted, rapid infiltration is proposed to fabricate composites, consisting of reactive metallic and ceramic phases with controlled microstructure and tunable properties. An aluminum (Al) alloy/Ti2AlC composite is selected as an example of the far-from-equilibrium systems to fabricate, because Ti2AlC exists only in a narrow region of the Ti-Al-C phase diagram and readily reacts with Al. This kind of reactive systems challenges conventional methods for successfully processing corresponding metal-ceramic composites. Al alloy/Ti2AlC composites with controlled microstructures, various volume ratios of constituents (40/60 and 27/73) and metallic phase sizes (42–83 μm, 77–276 μm, and 167–545 μm), are obtained using the Ti2AlC foams with different pore structures as preforms for molten metal (Al alloy) infiltration. The resulting composites are lightweight and display exceptional mechanical properties at both ambient and elevated temperatures. These structures achieve a compressive strength that is 10 times higher than the yield strength of the corresponding peak-aged Al alloy at ambient temperature and 14 times higher at 400 °C. Possible strengthening mechanisms are described, and further strategies for improving properties of those composites are proposed.

sintering temperature of Ti 3 AlC 2 to avoid substantial and harmful reactions, and thus Ti 3 AlC 2 particles were barely sintered in those composites 7 . The temperature choice in the powder co-sintering method is dictated by the melting point of the metals, which is normally well below typical sintering temperatures of MAX phases, but high enough to trigger intensive reactions between the constituents. One way to process metal/MAX phase composites with controlled interfacial reactions is to infiltrate molten metals into MAX phase foams. In fact, a pressureless infiltration technique has been used to process Ti 2 AlC/Mg composites that exhibit higher strength and mechanical energy dissipation than other Mg composites 11,12 . However, the pressureless infiltration is not an easy task because of the poor wettability of MAX phase foams with molten metals. The poor wettability slows the infiltration and sometimes makes it impossible. It can also yield a weak bonding between metals and ceramics, resulting in inferior mechanical properties of the composites 13 . This problem can be overcome using pressure infiltration to force molten metals into ceramic foams. However, in many cases the ceramic-metal reaction is so fast that even during pressure infiltration new phases would form, especially at the interface between the MAX phase and the metallic alloy. The new phases not only block pores and thus prevent further infiltration, but also degrade the properties of the constituents in the fabricated composites. Therefore, reducing the chemical reaction is the most challenging issue for fabricating metal/MAX phase composites with far-from-equilibrium compositions. In the present work, an Al alloy/Ti 2 AlC composite is selected as an example of far-from-equilibrium systems to fabricate, because Ti 2 AlC exists only in a narrow region of the Ti-Al-C phase diagram and readily reacts with molten Al at temperatures above 660 °C.
Al is a natural choice to be used in ceramic-metal composites for aerospace and transportation applications, where weight saving is critical. Particularly, the development of new generation aircrafts calls for high-performance Al alloys, especially at elevated temperatures. An approach to obtain improved high-temperature properties is to combine Al alloys with ceramics, such as Al 2 O 3 [14][15][16] , B 4 C [17][18][19] , and SiC 20-24 . However, MAX phases have not been explored for Al-based composites until two recent studies on Ti 3 AlC 2 /Al 7 and Al alloy/Ti 2 AlC composites 25 .
The introduction of MAX phases in Al composites brings several additional advantages, which could not otherwise be obtained using traditional ceramics, e.g. Al 2 O 3 , B 4 C, or SiC. First, a typical MAX phase, like Ti 3 SiC 2 , has a higher fracture toughness (~7 MPa·m 1/2 ) than Al 2 O 3 (~4 MPa·m 1/2 ), B 4 C (~3.7 MPa·m 1/2 ), and SiC (~4.6 MPa·m 1/2 ) that potentially can reduce the sensitivity of composites to brittle fracture. Secondly, unlike traditional ceramics, MAX phases have good transport properties that originate from the atomic bonding with mixed covalent, ionic, and metallic characters. Good transport properties of MAX phases would retain the functional properties of Al, namely good thermal and electrical conductivities. Third, MAX phases can be compressed to stresses as high as 1 GPa and fully recover their original, undeformed shapes upon the removal of the stress while dissipating 25% of the mechanical energy. The addition of MAX phases in Al can introduce mechanical energy dissipation, which is unique when compared to traditional ceramic-Al composites. Lastly, unlike traditional ceramics, MAX phases are machinable even at room temperature, which can significantly reduce the manufacturing cost of the composite.
It was shown in our previous work that a current and pressure-assisted, rapid infiltration is a viable method to produce Al alloy/MAX phase composites 25 . In the present work, the combination of rapid infiltration and MAX phase foams with tailored structures is used to control the microstructure in Al alloy/MAX phase composites. In addition, we characterize the microstructures and perform micro-tomography on the resulting composites to find out the reaction mechanisms during processing, and finally tailor the compressive strength of the composites by controlling their microstructures. We also discuss the possibility of adapting this method in building complex hierarchical structures to mimic natural composites.

Results and Discussion
Microstructural, Compositional, and Phase Analyses. Figure 1a-c show SEM images of the Ti 2 AlC foams with different pore sizes, i.e. 42-83 μ m, 77-276 μ m, and 167-545 μ m, respectively, fabricated using NaCl particles as pore formers 26,27 . All three foams were fabricated using the same volume percent (40 vol.%) of NaCl particles and have comparable overall porosities of 40.8, 41.6 and 39.9 vol.% after pressureless sintering. The infiltration of these foams with Al 6061 alloy using procedures described elsewhere 25 resulted in Al alloy/Ti 2 AlC composites with various sizes of the Al alloy phase, as shown in Fig. 1d-f. The size of the Al alloy phase is identical to the pore size in the Ti 2 AlC foams before infiltration. To illustrate and compare the macroscopic structures, insets in Fig. 1 show photographs of both the foams and the composites. Despite the short infiltration time of approximately 30 seconds, the porosity measurement results (see section "Mechanical Properties under Compression") indicate that more than 94% of the open pores in all foams were infiltrated with the molten Al alloy. In addition, a small amount of reaction could be observed at interfaces between the two major phases on SEM images in Fig. 1. Further analysis of the minor interfacial reaction is presented below. Figure 2a shows X-ray diffraction (XRD) results of the composites in comparison with the starting materials (i.e. Al alloy and Ti 2 AlC). In addition to the major phase (Ti 2 AlC), the commercial MAXthal211 (Sandvik Materials Technology, Hallstahammar, Sweden) powder used in the present study contains Ti 3 AlC 2 and a small amount of Al 2 O 3 (Fig. 2a). Ti 2 AlC, Ti 3 AlC 2 and Al were three major phases in the processed composites, together with a small amount of Al 2 O 3 . All phases identified in the composites were also found in the starting materials, and no new phases were detected in the XRD. However, more detailed compositional and phase analysis of the composites by energy dispersive spectrometry (EDS) (Fig. 2b,c) and electron backscatter diffraction (EBSD) (Fig. 3) clearly showed the presence of an intermetallic (TiAl 3 ) phase, whose amount was most likely too small to be detected by using XRD. Figure 2b shows typical backscattered SEM images of the Al alloy/Ti 2 AlC composite at different magnifications, while the EDS results in Table 1 include the concentration of individual elements, Ti/Al ratio, and Al/O ratio in each spot indicated in Fig. 2b, as well as a summary of the phases detected in the EBSD analysis in Fig. 3. Note that carbon is not shown in the EDS results, because EDS cannot provide carbon content with a high accuracy. The Ti/Al ratios in spots 2 and 3 are 0.35 and 1.96 (EDS results), respectively, which are in good agreement with the Ti/Al ratios in TiAl 3 and Ti 2 AlC, respectively. The presence of those two phases was also confirmed by EBSD, Fig. 3. Similarly, the Al/O ratio of 0.69 in spot 4 suggests Al 2 O 3 phase, whose presence was also confirmed by the EBSD results (Fig. 3). While the intermetallic phase was most likely formed by the reaction between molten Al alloy and MAXthal211, as it is discussed in more detail below, Al 2 O 3 has three possible sources. These sources include raw powder (Fig. 2c), possible surface oxidation of Al alloy during processing, and possible reaction between Al and an oxide layer or oxide adsorbates on the surface of Ti 2 AlC preform.   (Table 1) and EBSD analysis (Fig. 3). (c) A backscattered SEM image of the Ti 2 AlC foam with arrows pointing out to TiAl 2 impurities (stoichiometry determined by EDS).
As a further confirmation of the phase composition of the composites, Fig. 3 shows a phase map and four element maps (Ti, Al, C, and O) of the composites with the sizes of the Al alloy phase ranging from 167 to 545 μ m. Both identified and unindexed phases are color coded in the phase map. As listed in Fig. 3, 82.0% of all phases were identified, and 18.0% were not identified by EBSD. Note that out of the 18.0% unindexed phases, approximately 6.0% is residual porosity in the composites (see section "Mechanical Properties under Compression"). The identified phases include Ti 2 AlC, Ti 3 AlC 2 , Al, TiAl 3 , and Al 2 O 3 . The area percent of each phase is listed in Fig. 3. Note that Al 2 O 3 is present in both ceramic and metallic phase regions, suggesting that the Al 2 O 3 in the ceramic phase region most likely comes from the raw ceramic powder (Fig. 2), whereas the Al 2 O 3 in the metallic phase region should result from surface oxidation of the Al alloy or Al powders used in this study. A significant amount of TiAl 3 , i.e. 7.3% (Fig. 3), was identified either at the interfacial region or in the ceramic phase regions. The TiAl 3 in the interfacial region forms a thin "ring" surrounding the Al alloy, separating it from the ceramic phases. A comparison between the phase map and SEM images of the Ti 2 AlC foam ( Fig. 1) suggests that the locations of the TiAl 3 at the ceramic phase region could formerly be the small pores in the ceramic walls, where molten Al alloy could reach and react with the TiAl 2 found in the Ti 2 AlC foam (Fig. 2c).
The unindexed, black regions in Fig. 3 have three possible sources. First, the black dots with sizes of approximately 10 μ m in the Ti 2 AlC region are most likely pores not infiltrated with the Al alloy. Second, the black area, especially around interfaces, were unidentified, because the area was in electron shadow during EBSD data collection due to a slight height difference between the two phases as a result of their uneven polishing of harder ceramic and softer Al phases. Third, the black, straight lines with lengths of 50-150 μ m are scratches, which were introduced during polishing.
A common reaction phase, titanium aluminide, was found in both the composites presented here and the previously reported Ti 3 AlC 2 /Al composites 7 , as it is expected from the Ti-Al-C phase diagram 28 . However, the composites presented here have significantly less titanium aluminide than previously reported 7 . Thus, the short     Figure 4a shows the typical compressive stress-strain curves at room temperature for Al alloy/Ti 2 AlC (volume ratios 40/60 and 27/73) and Ti 2 AlC, while Fig. 4b summarises their room temperature compressive strengths.   Figure 5b shows that the specific strength, i.e. compressive/yield strength divided by density, of the composite is 335 MPa·cm 3 /g, which is about 5 times that of the Ti 2 AlC foam (62 MPa·cm 3 /g), 3 times that of the peak aged (PA) Al 6061 alloy (115 MPa·cm 3 /g 29 ), 6 times that of previously reported value of the Ti 3 AlC 2 /Al composite (55 MPa·cm 3 /g 7 ), and 5 times the yield strength of the SHT Al 6061 alloy (41 MPa·cm 3 /g). The importance of the later cannot be overemphasized, especially for applications in which lightweight materials that can carry large loads, such as aerospace and transportation, are essential. Figure 5c shows typical compressive stress-strain curves of the composites and PA Al 6061 alloy at 25 °C and at 400 °C. Note that Al alloy creeps at 400 °C, which is approximately 0.7 of its melting point. The ultimate compressive strength of the composites at 400 °C (800 MPa) is more than one order of magnitude the yield strength of Al 6061 alloy at 400 °C (58 MPa). Although the strength of the composites is higher than that of PA Al 6061 alloy at both room and elevated temperatures, the failure strain of the composites is smaller than that of PA Al 6061 alloy. In other words, the ceramic phase strengthens Al alloy at the expense of its ductility.

Effects of the Volume Fraction and Size of the Al Alloy phase on Mechanical Properties.
Thermal Stability. Since the Al alloy/Ti 2 AlC is a far-from-equilibrium system, thermal stability is equally important as mechanical strength at elevated temperatures. Figure 6 shows, side by side, the ultimate compressive strength and backscattered SEM images of the as-processed and heat treated Al alloy/Ti 2 AlC (volume ratio 27/73) composites. After heat treatments at 400 °C (0.7 of the melting point of Al) for one day and six days, the composites retained 91% and 92% of its strength, respectively, suggesting a sustainable mechanical performance against elevated temperatures. The backscattered SEM images of the composites before and after the heat treatment suggest little growth of the reaction phase and no evidence of interface de-bonding or cracks at the micrometer scale, which is in good agreement with the fact that the composites were able to retain more than 90% of its strength after the heat treatment.
The difference of the compressive strength before and after the heat treatment could come from several changes. First, the heat treatment could relax the residual stresses introduced during the rapid cooling, leading to different initial stress states between as-processed and heat treated composites for both metallic and ceramic phases upon loading. Second, the heat treatment could change the small misorientation within Al alloy grains. Third, the heat treatment could dissolve precipitates (if any) in the Al alloy. The last two possible changes would lower the strength of Al alloy in comparison with the Al alloy in the as-processed composites.

Conclusions
The current and pressure-assisted, rapid infiltration was used for producing interpenetrating Al alloy/Ti 2 AlC composites with controlled volume percent and size of the Al alloy phase, which were utilized to tailor compressive mechanical properties of the composites. The major findings are summarized as follows.
1. More than 97% of the open porosity in the ceramic preform was infiltrated with molten metal even after a short processing time of approximately 30 seconds. The results suggest that the rapid infiltration offers an efficient route for producing interpenetrating ceramic-metal composites with customizable structures by controlling the structure of the ceramic preform. 2. Little reaction was observed in the Al alloy/Ti 2 AlC system, suggesting that the rapid infiltration could be used to process far-from-equilibrium composite systems, which could not otherwise be obtained using conventional methods. 3. This method can result in composites with superior properties in comparison with those of its constituents.
The specific compressive strength of the Al alloy/Ti 2 AlC composites is about 10 times and 14 times higher than the specific yield strength of peak-aged Al alloy at room temperature and at 400 °C, respectively. 4. Despite the far-from-equilibrium composition, the Al alloy/Ti 2 AlC composites are thermally stable up to 400 °C (0.7 of the melting point of Al). The Al alloy/Ti 2 AlC (volume ratio 27/73) composites retained more than 90% of its strength after a heat treatment at 400 °C for six days. and pore size (ranging from 42-545 μ m) 26,27 . Only two ratios (20/80 and 40/60) were selected in the present study for proof-of-concept work of rapid infiltration with these Ti 2 AlC foams. Further investigations could be carried out to elaborate the effects of porosity and pore size of the ceramic foams on the infiltration process and properties of the resulting composites. Pore size of the foams was determined by measuring the size of approximately 50 pores in SEM images using the line intercept method, as specified in ASTM E112-13 30 . Four SEM images from randomly selected locations on each sample were used to measure the pore size. Al 6061 alloy discs (McMaster-Carr, GA) with a diameter of 20 mm and a thickness of 4 mm were used for the infiltration process.

Composite Sample Preparation.
A disc (20 mm in diameter and 4 mm in thickness) of Ti 2 AlC foam was "sandwiched" in between two Al alloy discs and placed in a graphite die, followed by infiltration at 750 °C under 5 MPa uniaxial pressure for 1 minute. The "sandwich" set-up enables more uniform infiltration of molten metal. The infiltration was carried out in a spark plasma sintering system (SPS 25-10, GT Advanced Technologies, CA). A direct current was applied from 0 to 1250 A in 4 min and stabilized at 860 A for 1 min at 750 °C; the pulse cycle was 10 ms on and 10 ms off. The chamber was evacuated and held at 10 −6 torr for 10 minutes before heating. The heating rate was 200 °C/minute. It takes less than 10 minutes for a complete infiltration process including heating/ melting, soaking, and cooling/solidification. Graphite foils were applied between samples and graphite die before infiltration. The temperature was calibrated and measured using procedures described elsewhere 8 . Characterization. The density and porosity (both open and closed) were determined by alcohol immersion method based on Archimedes' principle, as specified in ASTM C20-00 31 . The theoretical density values of 4.11 (g/cm 3 ) 32 and 2.70 (g/cm 3 ) 33 for Ti 2 AlC and Al, respectively, were used to calculate the theoretical density of composites using the rule of mixture. It was assumed that the effect of new phases formed by chemical reactions on the theoretical density is negligible. The relative density equals the measured value divided by the ROM value, i.e. 3.55 g/cm 3 and 3.73 g/cm 3 for the Al alloy/Ti 2 AlC composites with volume ratios of 40/60 and 27/73, respectively. The volume of the Al alloy is the measured porosity values of Ti 2 AlC foams, assuming all pores were filled with Al alloy. The residual porosity of the composites was measured but was not taken into account for the volume ratio approximation.
The phase composition of the starting powders and the as-infiltrated composites was determined using an X-ray diffractometer, XRD (D8 Discover, Bruker, WI) with Cu Kα radiation (wavelength = 1.542 Å) at 40 kV and 30 mA. The two theta range was from 8° to 80° with a step size of 0.04° and a step time of 1.5 s. The XRD results were analyzed utilizing the Inorganic Crystal Structure Database (ICSD).
The microstructure, phase composition and distribution were characterized using a Field Emission Scanning Electron Microscopy, FE-SEM (JSM-7500F, JEOL, Tokyo, Japan), equipped with Energy Dispersive Spectroscopy (EDS). Also used was another FE-SEM (Zeiss Ultra Plus, Carl Zeiss, Oberkochen, Germany) equipped with an Oxford Instrument AZtec EDS and a Nordlys-S Electron Backscatter Diffraction (EBSD) system. The accelerating voltage and emission current were 15 kV and 20 mA, respectively. The duration of spot scan of EDS was 60 seconds per spectrum. The EBSD scans were run with an accelerating voltage of 12 kV, an aperture size of 60 μm only one EBSD scan is shown here and a step size of 0.3 μ m. The 3D microstructure showing the phase distribution was obtained using an Xradia MicroXCT-400 Micro-Computed Tomography (Zeiss, Germany). Further composition measurements were carried out using a Cameca SX-100 electron microprobe with a wavelength dispersive spectrometry (WDS) system.
The compressive strength of the samples was measured by using a universal testing machine (MTS810, MTS, MN) at a strain rate of 1.4 × 10 −4 s −1 . The strain was measured using a miniature extensometer (MTS810, MTS, MN) attached to the edges of the compressive grips. All samples were cut by electrical discharge machining to dimensions of 3.5 mm × 3.5 mm × 7 mm to achieve parallelism with less than 1% error of thickness.