Introduction

Titanium silicon carbide (Ti3SiC2) is composed of Si atoms separated from each other by three Ti layers that accumulate C atoms around them (Ref 1). The interesting properties of Ti3SiC2 arise precisely from its crystal structure, in which metallic, covalent and ionic bonds are present. Like metals, it possesses high thermal conductivity (34 W m−1 C−1) and electrical conductivity (4.5 × 106 S/m), good machinability with conventional tools, high resistance to crack propagation, and outstanding resistance to varying ambient operating temperatures (Ref 2). Like ceramics, it is characterized by high mechanical strength at elevated temperatures, a low coefficient of thermal expansion—resulting in increased dimensional stability of machine parts made from it—a high melting point, good corrosion and oxidation resistance, a high modulus of longitudinal elasticity (up to 325 GPa) and high creep resistance. Like other MAX phases, Ti3SiC2 also exhibits high impact strength, the ability to dampen vibrations (Ref 3) and an increased capability to absorb energy, especially in cyclic loading processes.

It is predicted that due to its low density and high mechanical strength parameters, Ti3SiC2 and its composites may replace nickel-based superalloys for the construction of engine components. Being a good thermal conductor with a low coefficient of thermal expansion, it is also a promising alternative to materials commonly used for working at high temperatures, such as bushings, brake disks and heat exchangers (Ref 4).

Ti3SiC2-based materials are commonly manufactured using one of the following methods (Ref 3, 5,6,7): Hot Pressing (HP, RHP, HIP), Self-propagating High-temperature Synthesis (SHS), in-situ, Spark Plasma Sintering (SPS), Vacuum, Discharge or Pressureless Sintering (VS, PDS, PS), Mechanical Alloying (MA), Chemical Vapor Deposition (CVD) and 3D printing. Some of these techniques can also be applied for the fabrication of MAX phase-based composite materials.

However, in the case of SHS products, which are usually highly porous, additional pressure or pressureless infiltration is necessary for the creation of a dense composite material. Infiltration methods have until now been generally reserved for Ti2AlC- or Ti3AlC2-based composite materials (Ref 8, 9). In studies using these, it is reported that residual porosity was one of the most crucial problems. To overcome this issue, squeeze casting infiltration has been proposed as an alternative to conventionally used techniques, as this method ensures the most precise filling of open pores of the preform and high-quality bonding of the reinforcement alloy. This method was previously used by the authors to produce Al-Si matrix composite materials reinforced with MAX phases from a Ti-Al-C system (Ref 10).

The present study aims to develop a new beneficial method, including the synthesis of Ti3SiC2 MAX preforms and subsequent saturation with molten metal. SHS synthesis with correctly attuned parameters can provide a homogeneous porous structure quickly and effectively. With subsequent high-pressure infiltration, these fine structures can be completely saturated, thus reinforcing the Al alloy casting. By performing examinations of microstructure and basic physical properties, it is possible to select the optimized process parameters and confirm the benefits of this method.

Materials and Methods

Commercial powders of Ti (99.5% Ti, −325 mesh, Alfa Aesar), SiC (99.9% SiC, −325 mesh, Alfa Aesar) and graphite (99.5% C, −325 mesh, SGL Carbon Ltd.) were used as starting materials in the molar ratios 3:1.2:1 to prepare a stoichiometric reactant mixture for the fabrication of Ti3SiC2. First, the Ti, SiC and C powders were mixed with balls of ZrO2 for 10-30 min, or for 3-14 h in the case of mechanical activation, with a ball-to-powder ratio (BPR) of 10:1. Subsequently, the powders were uniaxially cold-pressed under a pressure of 450-930 MPa for 60 s, to form cylindrical samples in the shape of pellets with 22 mm diameter, 4 mm of height and mass of 4 g. The prepared samples were subjected to a MASHS process. The glass reaction chamber was filled with alumina fibers, transparent to microwaves, to provide conditions that are nearly adiabatic and to protect the reaction chamber from the effects of extreme temperature. As the Ti/SiC/C combination has a high ignition temperature, additional padding is required. In this technique, called coupled MASHS (Ref 11, 12), the 3Ti-1.2SiC-C compact was placed onto a 2Ti-Al-C compact, the role of which is to boost the reaction above the ignition temperature of the substrates. Microwave magnetron power was applied in the range 300-400 W. The whole reaction took place under the flow of an Ar atmosphere. The temperature was controlled by a Raytek Marathon MM pyrometer with a measuring spot diameter of 0.6 mm. Further processing details are as presented elsewhere for Ti-Al-C systems (Ref 13).

The prepared MAX skeletons were subsequently used for pressure infiltration: squeeze casting with an Al-Si eutectic alloy, AC 44200 (10-13.5% Si, 0.4% Fe, 0.05% Cu, 0.4% Mn, 0.1% Zn, 0.15% Ti, remainder Al). The molten metal (740 °C) was applied under pressure (90 MPa) to the ceramic skeleton, preheated to 750 °C, for 2 min. The resulting composite was analyzed using a Hitachi S-3400 N scanning microscope, a SwiftED3000 chemical analyzer and an Ultima IV x-ray diffractometer (40 kV/40 mA, step 0.04). TEM observation of the samples, ground to a thickness of 5 μm with a Gatan Dimple Grinder and Gatan DuoMil ion polisher, was performed with a Hitachi H-800 transmission microscope (150 kV). The samples were also analyzed by transmission analytical electron microscopy in scanning mode (STEM) with a convergent beam, using a JEOL 2100 electron microscope operating at 200 kV and equipped with a JEOL JED-2300T EDS unit for analytical measurements.

Thermal diffusivity and specific heat were measured using an LFA457/Netzsch laser flash analyzer and an STA 449 F1 Jupiter Netzsch device, respectively. The thermal expansion coefficients were determined using a DIL 402 Netzsch dilatometer. All measurements were carried out under a protective atmosphere of argon.

Mechanical properties, such as instrumental hardness and Young’s modulus, were established with an NHT (Anton Paar) nanoindenter using the Oliver–Pharr method (Ref 14). Indentations were performed with a diamond Berkovich-type indenter for linear loading with a maximum load of 50 mN. Loading and unloading rates were both 100 mN/min.

Results and Discussion

Microwave-Assisted Self-Propagating High-Temperature Synthesis

The fabrication of an open-porous structure of a MAX preform, with sufficient conversion of the substrates and the strength required for pressure infiltration, involves the determination of parameters for the MASHS synthesis. Powders, with a suitable stoichiometric ratio, were mixed for 10-30 min to obtain a homogeneous mixture or were ball-milled for 3-14 h to achieve better contact between the particles and to activate them. The green mixture was uniaxially cold-pressed under a pressure of 450-930 MPa, creating the initial porosity and conditions needed for the self-propagation reaction. Examples of the curves obtained from the samples produced from the green compact mixed and compressed under 930 MPa, and from further samples subjected to mechanical activation in a ball mill (3 hours, BPR 10:1, ZrO2 balls) instead of regular mixing, are presented in Fig. 1. The graph shows that mechanical activation decreases the combustion temperature significantly. However, mechanical activation of the powders caused the resulting preform to be less porous, contrary to the aim of the experiment. Therefore, the conditions selected for further experiments consisted of basic mixing and a relatively high compaction pressure. During pressing, only the Ti particles were plastically deformed, and their oxide films were disrupted. Moreover, the particle interlocking necessary to produce a tough compact was diminished by the carbon, which acts as a kind of solid lubricant. SiC was chosen as the source of Si in the Ti-Si-C combination, because in previous studies it was found advantageous for the production of Ti3SiC2 with high purity (Ref 2). Nevertheless, as SiC particles are far less reactive than the Al or TiC phases, for example, that are present in the Ti-Al-C combination, it is necessary to apply coupled MASHS synthesis. Merzhanov et al. have reported multiple studies of this particular mode of combustion synthesis (Ref 11, 12). To compensate for the evaporation of Si during MASHS synthesis, excess Si was added by altering the molar ratio of the initial mixture to Ti/SiC/C = 3:1.2:1. This composition has also been used for SHS synthesis by Saed et al., who found that it led to the highest Ti3SiC2 content in the synthesized material (Ref 4). The synthesis begins after around 10-15 s of microwave heating. The first peak visible on the curve for the Ti-Si-C combination represents the energy impact from the Ti-Al-C sample (the coupled SHS technique) in which the SHS reaction occurs earliest.

Fig. 1
figure 1

Temperature–time dependences and derivative thermal analysis (dT/dt) of SHS synthesis for powders compacted under 930 MPa: mixed only (930 MPa) and mechanically activated (930 MPa + MA)

Full size image

With the addition of SiC particles, the synthesis proceeds along a slightly different course than for the elementary powders and free silicon. The first observed temperature fluctuation, especially with the mixture mechanically activated (Fig. 1, point A), corresponds to the allotropic transformation Ti(α) → Ti(β) at 882 °C. The more reactive β phase (Ref 15) triggers a synthesis by forming titanium carbide, Ti(β) + C → TiCx, thereby providing the initial energy to make the reaction self-sustaining. Until all of the free C is consumed, the temperature must increase to a sufficient level. At that point, Ti reacts with SiC, releasing the Si. In Ref 15, 16, it is suggested that the reaction proceeds by diffusion and only small quantities of liquid are formed, possibly at the interface. The TEM observations presented below confirm the presence of TiSi2 crystals, which can be formed directly from the melt at 1478 °C, at much lower temperatures than detected in the synthesis. Moreover, the Si released interferes with the TiC produced by the intermediate phase, which can be defined as a solid solution of Si in TiC. The final product of the Ti3SiC2 MAX phase does not grow directly from SHS synthesis; rather, it precipitates from the converted intermediate phase. As an exothermic process, it confirms, based on the deviation observed on the cooling curve (point B), the minimum temperature (1450 °C) for successful sintering (Ref 15).

Microstructure

The process parameters selected should lead to preforms with a structure that is appropriate for metal infiltration. Figure 2(a) presents the overall open porosity of the resulting material, with good homogeneity and regular pore dimensions. The Ti3SiC2 phases typically appear as elongated plate grains, connected together in packets 10-20 μm in length and 1-3 μm in thickness (Fig. 2b). The surfaces of pore walls formed from MAX plates are often covered with characteristic round TiC inclusions. The presence of hard carbides leads to high resistance to brittle fracture, because such particles can block the propagation of cracks within the material (Ref 17).

Fig. 2
figure 2

Overall open porosity of manufactured material after SHS synthesis (a); Ti3SiC2 plates with round inclusions of TiC (b)

Full size image

The microstructure of Al-Si eutectic alloy AC 44200 is shown in Fig. 3. After infiltration, the composites underwent repeated SEM, EDS, XRD and TEM/STEM analyses. The identification of particular phases by EDS analysis is presented in Fig. 4. TiSi2 phase was at first not detected during EDS analysis. Its formation was further confirmed with TEM and STEM analyses. Figure 5(a) shows the microstructure of the Al-Si composite reinforced by the MAX skeleton. Pores with an average size of 50 µm are filled with the Al alloy, which during the final stages of infiltration penetrates even the walls created by the MAX plates. The scattered black spots represent remnants of partly reacted SiC particles. Residual porosity, which amounts to several percent, is located in places where C flakes are present. It is likely that external layers of the large particles reacted during synthesis with the remaining TiC and limited amounts of Ti converted into closed gas pores. As a very high-pressure process, squeeze casting can force the liquid to break the thin walls of closed pores and fill the preform almost completely. MAX plates are often bonded with a thin layer of TiSi2 phase (Fig. 5b). Such stacks formed with narrow gaps between the plates are also effectively saturated with the alloy. Despite extension of the time for which the applied pressure was maintained, to improve the overall filling, no further reaction took place between the Ti3SiC2 preform and the Al-Si matrix. The volumetric ratio of alloy to reinforcement was maintained at about 40:60. Moreover, it can be stated that the Al-Si alloy can be actually improved by squeeze casting, because of its microstructure refinement (Ref 18, 19).

Fig. 3
figure 3

Microstructure of Al-Si eutectic alloy AC 44200

Full size image
Fig. 4
figure 4

Identification of particular phases by EDS analysis (1—Ti3SiC2 MAX phase platelets, 2—TiC inclusion, 3—SiC inclusion, 4—Al-Si matrix after squeeze casting infiltration)

Full size image
Fig. 5
figure 5

Microstructure of the Ti-Si-C MAX/Al-Si composite material: overall view (a); SiC inclusions and Ti3SiC2 plates bonded with TiSi2 layers (b)

Full size image

XRD Analysis

The XRD spectra acquired from the material after MASHS synthesis showed the presence of the main MAX phase of Ti3SiC2 and minor amounts of TiCx, TiSi2 and SiC (Fig. 6). A quantitative Rietveld analysis of the manufactured preforms revealed the following constituents: Ti3SiC2 75.90 wt.%, TiC 14.71 wt.%, TiSi2 8.27 wt.%, SiC 1.12 wt.%. To eliminate the remaining SiC phase, it may be advantageous to limit the amount of SiC in the substrate composition. Unlike in other studies (Ref 5), no Ti5Si3 inclusions were found in the synthesized material. Based on the ratio of the heights of individual reflections and the absence of reflection of the crystal lattice at an angle of 2θ = 46°, the material synthesised from the Ti-Si-C combination is found to be α-Ti3SiC2 (Ref 20).

Fig. 6
figure 6

XRD patterns for samples before and after the squeeze casting pressure infiltration process

Full size image

On examination of the composite comprising the preform infiltrated with Al alloy, it was found that the additional reflections that appeared in the spectra, compared with the previously tested preforms, come only from the matrix material: Al or Si. This indicates the absence of chemical reactions after the infiltration process. In addition, the tests do not show differences between the alloy material within the composite and the reference sample. In the alloy filling the interior pores of the preform, no element diffusion of the reinforcement material was found.

TEM Characterization

A plate-type Ti3SiC2 grain was detected using a selected area diffraction pattern and the sampling planes (001) (d001 = 1.767 nm), (010) (d010 = 0.266 nm) and (011) (d011 = 0.263 nm); see Fig. 7. The main axis of the grain is parallel to the plane (010). Using the dark-field method, a 1-µm-wide precipitation inside the MAX phase was observed, and this was identified as a TiC phase (Fig. 7c, d). The TiC phase is slip-free, but the density of dislocations inside it is higher than in a MAX grain. The lower thermal expansion of TiC compared with its surrounding MAX phases results in the generation of tensile stresses in these precipitations. TEM observations of Ti3SiC2-based ceramics reported by Morgiel et al. (Ref 21) revealed a high density of dislocations and a large number of cracks at the Ti3SiC2/TiC interface. In the present study, most TiC particles were found to be located on the walls of pores, and no cracks were detected.

Fig. 7
figure 7

TEM microstructure of Ti3SiC2 phase after infiltration: bright-field image of Ti3SiC2 plate (a); selected area diffraction pattern obtained in Ti3SiC2 (b); dark-field image of Ti3SiC2, taken in the (010) reflection, showing precipitation inside the MAX phase, marked with a square (c); bright-field image showing the TiC/MAX interface with (010) slip planes in the MAX phase (d)

Full size image

The presence of a TiSi2 intermetallic phase was detected using the sampling planes (220) (d220 = 0.207 nm) and (130) (d130 = 0.156 nm); see Fig. 8. The phase has a plate-like shape (Fig. 8a) and exhibits the effect of internal stress fields and dislocations in the Al matrix. As previously mentioned, a thermal mismatch produces stresses between the MAX preform components. Additionally, during solidification, the Al matrix—with a coefficient of thermal expansion more than twice as large—shrinks significantly and places the preform under compression.

Fig. 8
figure 8

Bright-field image of TiSi2 precipitation close to the Al matrix (a); selected area diffraction pattern obtained from the same place (b)

Full size image

Figure 9 presents an STEM image of the composite with the distribution of C, Al, Si and Ti elements. MAX phase platelets containing Ti, Si and C were identified as included in a matrix of Al and Si crystallites with sizes up to 1 µm. Squeeze casting infiltration is a very fast process and lasts for about 1-2 s. At the end of this stage, the pressure increases abruptly to 90 MPa, pushing the liquid metal into small, nanoscale gaps. Contact with the reinforcement during solidification can be prolonged for tens of seconds, depending on the volume of the casting and heat transfer to the die. In the case of the samples produced here, solidification was calculated to last about 20 s, and for this period, pressure was exerted on the casting. Under these favorable conditions, chemical interaction was restricted and no element transfer or formation of products at the interface was detected. It was shown in a previous study of the interfacial reaction between Ti3SiC2 particles and an Al matrix (Ref 22) that when kept at a temperature of 600 °C for as long as 10 hours the Si atoms from the MAX phase can enter the Al matrix and form an Al(Si) solution. Degradation of the MAX phase needs a higher temperature (700 °C) to be maintained for a long time. It is not ruled out that in the present investigations some Si may have entered the Al-Si matrix at the interface, but this effect is negligible for the silumin group of alloys.

Fig. 9
figure 9

STEM image and corresponding EDS maps showing the distribution of C, Al, Si and Ti in the analyzed area. The RGB image was obtained by superimposing the EDS maps of Al, Si and Ti

Full size image

Mechanical and Thermal Properties

Mechanical testing involved measurements of hardness (HV) and instrumental modulus of longitudinal elasticity (EIT). The results obtained for the preforms and for the composite materials, in comparison with the matrix parameters, are summarized in Table 1. MAX-type phases are less hard than most ceramic materials, but harder than the majority of metals. Due to the nanolaminated structure of MAX phases, the result of a measurement depends to a great extent on its location. The hardness inside one of the MAX phase platelets can be significantly higher than the hardness measured in a larger area (Ref 23). The hardness values determined for preforms are close to the upper limit of the range reported previously by other researchers. In Ref 24, dense polycrystalline Ti3SiC2 of high purity, fabricated by reactive hot isostatic pressing, was found to have a hardness of only 4.0 ± 0.2 GPa. Reinforcing TiSi2–Ti3SiC2 with nanoparticles of SiC improves the hardness to 11-12 GPa (Ref 25), which is only slightly higher than that of the porous preform presented here. Compared with the alloy, the fabricated composite material had a hardness more than five times greater and a Young’s modulus almost six times greater. During solidification, the solid preform induces heterogeneous nucleation and refining of the matrix structure. Moreover, a relatively high volume of the preform results in a considerable reduction in released energy, due to the smaller fraction of liquid and latent heat. Thus, the composite solidified faster, and it was concluded that solid solution strengthening and precipitation strengthening (Si crystallites) both lead to an improvement in mechanical properties.

Table 1 Mechanical properties of the manufactured materials
Full size table

MAX phases have a higher coefficient of thermal conductivity (λ) than binary MX or MA compounds. At room temperature, it lies in the range 12-60 W/m °C, while the thermal expansion coefficient is approximately 5-11 × 10−6 °C−1 (Ref 26). The high thermal conductivity of MAX phases, according to the Wiedmann–Franz law, is directly related to their good electrical conductivity. The overall coefficient of thermal conductivity of a material is the sum of phonon and electron thermal conductivity (Ref 27). MAX phase materials containing Si mainly owe their overall thermal conductivity to electron thermal conductivity (Ref 26). At room temperature, the electron thermal conductivity of Ti3SiC2 is 33 W/m °C, while its phonon thermal conductivity is 1 W/m °C; these combine to give a theoretical overall thermal conductivity of 34 W/m °C (Ref 23). For the majority of MAX phases, apart from Ti4AlN3, Nb2AlC and Ta4AlC3, the thermal conductivity decreases slightly with an increase in temperature (Ref 23); this was also observed in the present study. Table 2 summarizes the thermal properties obtained for materials fabricated in the temperature range 50-500 °C. As temperature increased, with the effects of lattice distortion, molecule vibration and the obstructed flow of free electrons, thermal diffusivity decreased by about 25% in both the Al alloy and the composite. Specific heat increased with increasing temperature, as observed in previous studies (Ref 27).

Table 2 Thermal properties of the alloy and composite at 50-500 °C
Full size table

Based on thermal diffusivity, density (ρ) and specific heat (cp), the thermal conductivity (λ) was determined from the equation λ = cp a ρ. The change in mass, measured by thermogravimetric analysis (TGA), for the Al-Si alloy and Al-Si + Ti-Si-C composite was 0.35% and 0.13%, respectively. Reinforcing the metal matrix involves the formation of new grain boundaries, pores or particles, which naturally reduce thermal conductivity. Conductivity is almost 60% lower in the composite than in the Al-Si alloy, but compared with the base Ti3SiC2 (34-37 W/m °C) (Ref 28) a twofold improvement was observed.

The measured coefficients of thermal expansion confirm a more than twofold improvement in thermal dimensional stability for the fabricated Al-Si/Ti3SiC2 composite compared with the alloy material (Fig. 10). The thermal expansion of the Al-Si/Ti3SiC2 composite material is comparable to that of the reinforcing MAX phase. In the case of the Al-Si alloy, the expansion coefficient shows a slow upward tendency until around 250 °C and then decreases with further increase in temperature. Similar peaks on the curve between 250 °C and 400 °C, observed by Choi et al., were explained by precipitation of the Si phase. Squeeze casting with the use of a metal die proceeds with fast cooling and solidification; therefore, overcooling and saturation of Al with Si may occur. In this case, small Si crystallites were observed in the composite structures. However, the thermal expansion (8.5-10 × 10−6 °C−1) of the composite is almost constant as a function of temperature. That corresponds with the properties of pure MAX phases, which usually lie in the range 5-11 × 10−6 °C−1 (Ref 27), as presented for the temperature range 25-1000 °C by Barsoum et al. (Ref 25) (9.1 × 10−6 °C−1) and Son et al. (Ref 29) (9.1-11 × 10−6 °C−1) and for commercial powders of Ti3SiC2 (Ref 30) (7-9 × 10−6 °C−1). It may be assumed that the Ti3SiC2 skeleton restrains thermal expansion of the matrix, especially at higher temperatures when the Al-Si alloy is more plastic. The yield strength of the alloy quickly decreases above 200 °C (Ref 31) and the compression stress generated by the skeleton results in less overall expansion.

Fig. 10
figure 10

The relative elongation dL/Lo and the coefficient of thermal expansion (CTE) of the matrix (AlSi) and Al-Si + Ti-Si-C (composite)

Full size image

Summary and Conclusions

A novel method for manufacturing Al metal matrix composites based on a Ti-Si-C MAX phase has been successfully developed by combining two techniques: Microwave-Assisted Self-propagating High-temperature Synthesis (MASHS) and squeeze casting infiltration. Both of these methods were modified in order to achieve an open-porous structure and subsequently enable the full infiltration of preforms with the liquid metal. Based on microscopic observations and the evaluation of physical properties, the following conclusions are drawn:

  1. 1.

    MASHS of MAX phases in the Ti-Si-C combination (Ti3SiC2) is accompanied by the formation of inclusions of TiC (14.71 wt.%), TiSi2 (8.27 wt.%) and SiC (1.12 wt.%).

  2. 2.

    Squeeze casting pressure infiltration permits the complete filling of micro- and nanopores of the preform structure that is formed in plates.

  3. 3.

    As regards mechanical properties, the Vickers hardness and Young’s modulus were five and six times higher, respectively, for the composite materials than for the reference Al-Si alloy.

  4. 4.

    The produced composite materials exhibit double the thermal conductivity of Ti3SiC2 and significantly increased thermal stability compared with the alloy. The coefficient of thermal expansion for the composite is half that of the Al-Si alloy.