Synthesis, microstructure and properties of Ti–Al porous intermetallic compounds prepared by a thermal explosion reaction

Abstract

Ti–Al porous intermetallic compounds were prepared by a simple and energy-saving process of thermal explosion (TE) reactions. The effects of the Ti/Al molar ratios (including Ti : Al = 3 : 1, 2 : 1, 1 : 1, 1 : 2 and 1 : 3) on the temperature profiles, phase compositions, pore characteristics, density, expansion behaviors and oxidation resistance were investigated. The ignition temperatures were between 630–653 °C and the combustion temperatures were between 737–1161 °C, both of which are higher than the target furnace temperature. Porous Ti–Al alloys displayed a high open porosity of 35–56%. The pores were from non-fully dense green compacts and expansion behaviors of TE. When the target furnace temperature was set at 650 °C, the specimen with Ti and Al in a 1 : 2 ratio possessed the highest open porosity of 53.89% and the lowest density of 1.58 g cm⁻³. When the target furnace temperature was set at 700 °C, the specimen with Ti and Al in a 1 : 3 ratio possessed the highest open porosity of 56.27% and the lowest density of 1.44 g cm⁻³. The porous specimens with Ti and Al in a 1 : 3 ratio exhibited the best resistance to oxidation at 650 °C in air.

---

1. Introduction

According to the IUPAC classification, porous materials are divided into three classes: microporous, mesoporous and macroporous.¹ Microporous materials are composed of a solid (body) base and interconnected pores with molecular dimensions (less than 2 nm), which are widely used in the applications of heterogeneous catalysis, adsorption and gas storage.² Mesoporous materials have a distribution of pores in the range between 2 and 50 nm.³ Yanagisawa et al. first reported the preparation of mesoporous silica in 1990.⁴ In recent years, many works have been devoted to the synthesis and preparation of mesoporous materials. Duan et al.⁵ fabricated silicon oxycarbide ceramics by pyrolysis of polysiloxane, and they changed the pyrolysis temperature to adjust the surface areas and pores size distributions. Franceschini and co-workers successfully synthesized the mesoporous Pt and Pt/Ru alloy using a F127® template.⁶ Owing to the development of synthesis technology and the properties of mesoporous materials, the materials have been widely applied in the fields of chromatography, catalysis and electrochemistry.^1,3

Similar to microporous and mesoporous materials, macroporous materials also have attracted increasing attention and applications. Macroporous materials include porous organic and inorganic materials. However, porous organic materials are only applied in the fields of water treatment and biology due to their poor mechanical properties and thermostabilities at high temperatures and high pressures, their poor resistance to environmental corrosion and their negligible resistance to organic solvents.⁷ Compared with organic materials, porous ceramics exhibit special properties, including low densities, large specific surface areas, and excellent strengths and stabilities in high temperatures,^8,9 allowing for their use in molten metal filtration, thermal insulation, bio-scaffolds for tissue engineering, etc.^10–12 Meanwhile, their poor mechanical properties at room temperature, such as the brittleness and poor weldability of porous ceramics, limit their applications. In contrast, intermetallic compounds contain a mixture of metallic and covalent bonds¹³ and have the common advantages of both metals and ceramics.^14–17 Furthermore, porous alloy materials (including Ni–Ti,¹⁴ Fe–Al,¹⁵ Ni–Al¹⁶ and Ti–Al alloys^18,19) offer many distinct advantages over other materials that allows for their use in a broad range of applications.

Ti–Al intermetallic compounds have superior advantages, e.g., low densities and outstanding corrosion resistances.²⁰ Moreover, Ti–Al porous intermetallic compounds can be applied in the fields of filtration,¹⁹ catalysis,²¹ and thermal insulation.²² According to previous research,^23,24 three different intermetallic compounds, Ti₃Al, TiAl and TiAl₃, were obtained in the Ti–Al system; of these compounds, TiAl₃ possesses a lower density, a better resistance to oxidation and a higher strength compared with the others due to the high concentration of aluminum. It represents significant progress to develop Ti–Al porous materials based on TiAl₃ for filtration, especially where resistance to oxidation and strength are the prime requirements.^24,25 He et al. developed a new method based upon the Kirkendall effect to fabricate Ti–Al porous alloys, and the total sintering time would last for 12.3–19.8 h because the heating rate was set at 2 °C min⁻¹ throughout the whole process.²⁶ Yang¹⁹ and Liang²⁷ prepared porous Ti–Al intermetallic compounds by reactive diffusivities and the Kirkendall effect, and the total sintering times were 8.7 h and 16.25 h, which were processes that consumed a large amount of energy. Therefore, it is necessary to determine a new manufacturing method with a short sintering time and low energy consumption to prepare Ti–Al porous intermetallic compounds.

Until now, many technologies have been employed to prepare porous materials, among which the self-propagating high-temperature synthesis (SHS) is an attractive method. SHS is an autogenous process and derives its energy from the exothermic reactions from the reactants.²⁸ SHS offers several advantages over traditional synthetic methods.²⁹ But, SHS is a self-propagating method that is ignited locally at the contact point of the specimen with an ignition coil and quickly spreads to the entire specimen section in the form of a propagating combustion wave.³⁰ It is easy to form the structural imperfections of the materials during the SHS reaction, such as deformation, lamination and cracking.³¹ However, the thermal explosion (TE) reaction is a simultaneous reaction that involves a whole model of the combustion synthesis, i.e., a rapid heating of the entire specimen to the target temperature simultaneously with the synthesis that occurs throughout the body. Furthermore, TE can avoid the deformation and cracks of the specimen during the sintering process because TE is generally conducted in a furnace where all of the reactants are heated simultaneously.

In this paper, we report a simple and energy-saving approach to prepare porous Ti–Al intermetallic compounds by TE reaction. We produced the TE curves of Ti and Al with different molar ratios. Meanwhile, the effects of Ti/Al molar ratios on the phase compositions, density, expansion behaviors, pore characteristics and resistance to oxidation were investigated.

2. Experimental procedure

Titanium powders (∼300 mesh, 99.9% purity) and aluminum powders (∼200 mesh, 99.0% purity) were mixed with five different molar ratios, including Ti : Al = 3 : 1, 2 : 1, 1 : 1, 1 : 2 and 1 : 3, which corresponds with aluminum contents of 15.82 wt%, 21.99 wt%, 36.05 wt%, 52.99 wt% and 62.48 wt%, respectively. The mixture powders were wet blended in a planetary ball mill at 450 revolutions per minute for 240 minutes and were dried at 50 °C for 12 h, and cold-pressed into cylindrical compacts of ∼16 mm-diameter and ∼2.3 mm-height at 200 MPa. The TE reaction was conducted in a tubular furnace under argon atmosphere (see ESI†). When the target furnace temperature was pre-set as 650 °C, i.e., the tubular furnace was heat from the room temperature to 650 °C with a constant heating rate of 17 °C min⁻¹. As seen in Fig. S1,† a pair of WRe3–WRe25 thermocouple wires with a diameter of 0.1 mm was placed between two similar specimens to record the actual TE reaction temperatures at a frequency of 300 Hz. The heating was stopped when the TE reaction was observed (an abrupt temperature rise), and then the specimens were cooled in the furnace to the room temperature.

The phase composition of the TE products was investigated by X-ray diffraction (XRD) on a Bruker D8 ADVANCE X-ray diffraction machine with Cu target (λ = 0.15406 nm). The optical metallography of the polished specimens was performed on an Olympus optical microscope. Quanta 250 scanning electron microscope (SEM) equipped with UANTAX400 energy dispersive spectrometer (EDS) was employed to examine the microstructure of the porous materials. The dimensions of the specimens before and after the sintering were measured to determine the expansion behavior, which was observed by the changes in volume as well as in the axial direction and radial direction expansion ratios. The density of the porous alloys was measured with the Archimedes method. The open porosity (θ) was calculated by the expression: θ = (M₂ − M₁)/(Vρ).¹⁸ Where M₁ is the mass of a dried porous bodies and M₂ is the mass of the porous bodies filled with the wax, V is the external volume of the porous bodies, and ρ is the density of the max.

The oxidation experiments were conducted at 650 °C for 4 cycles for duration 96 h under air atmosphere. Each cycle represented 24 h of exposure at 650 °C. The mass of Ti–Al porous monoliths was measured before and after each cycle. The oxidation kinetics of the Ti–Al porous materials were determined by measuring the change in mass every 24 hours as a function of the exposure time.

3. Results and discussion

3.1 Characteristics of the TE reaction

Fig. 1 represents the TE profiles and the partial magnified exothermic peaks of the specimens with Ti/Al molar ratios of 3 : 1, 2 : 1, 1 : 1, 1 : 2 and 1 : 3. Thermal explosion includes two important parameters, i.e., the ignition temperature (T_ig) and the combustion temperature (T_c), which are marked in Fig. 1. In the early period of the TE curves, the temperature increased slowly, and the reaction was ignited when the temperature reached T_ig. Then, an abrupt and dramatic increase in the temperature appeared before the temperature reached T_c. Finally, the compact was cooled to the furnace temperature, and the TE reaction was accomplished. In Fig. 1a–e, the temperature–time profiles were obtained when the target furnace temperature was 650 °C. The T_ig values of the TE reactions were 634 °C (Fig. 1a), 630 °C (Fig. 1b), 630 °C (Fig. 1c), 649 °C (Fig. 1d), 653 °C (Fig. 1e). As seen in Fig. 1e, the ignition temperature (653 °C) is a slightly higher than the target furnace temperature (650 °C), and this is due to the furnace temperature stability of ±3 °C. The ignition temperatures fluctuated slightly approximately 630 °C when the Ti/Al ratios were 3 : 1, 2 : 1 and 1 : 1 (Fig. 1a–c). The T_ig increased with an increase in the Al content, and it was more difficult for the specimens to be ignited when the Ti/Al ratios were 1 : 2 and 1 : 3. Moreover, the T_ig was around the furnace temperature (∼650 °C) and approached the melting point of Al (660 °C), which implied that the TE reaction mechanism may involve solid Ti–liquid Al even if the T_ig was below the melting point of aluminum.³² To further investigate the exothermic action of Ti : Al = 1 : 3, the target furnace temperature was set at 700 °C and the green compacts were once again synthesized. The temperature profile is shown in Fig. 1f, and the T_ig values of the TE reactions were 679 °C.

Fig. 1 The actual temperature–time profiles (left) and the magnified exothermic peaks (right).

As shown in Fig. 1a–e, the T_c values were 869 °C (Fig. 1a), 1068 °C (Fig. 1b), 1161 °C (Fig. 1c), 915 °C (Fig. 1d) and 737 °C (Fig. 1e), which were higher than the furnace temperature (650 °C), indicating that Ti–Al system showed an obvious heat release phenomenon during the TE process. The specimen with a Ti : Al = 1 : 3 showed a remarkably low T_c compared with the other specimens, i.e., 737 °C. Additionally, the combustion temperatures T_c first increased and then decreased with an increase in the Al content (Fig. 1a–e). When the Ti/Al ratios changed from 3 : 1 to 1 : 1, the combustion temperature increased gradually, and this may be because that more compounds were formed and less unreacted Ti was retained in the final products with an increase in the Al content (XRD results in Fig. 2). The residual Ti in the final products acts as a diluent during the TE, reduces the combustion synthesis temperatures³³ and extends the ball milling time of the mechanically alloying.³⁴ Jo et al. reported that with the addition of 10 wt% diluent (MoSi₂) to a mixture of Mo and Si, the calculated adiabatic temperature decreased from 1943 K to 1748 K and the powder compact did not undergo a combustion reaction.³³ However, the T_c showed a tendency to decrease when the Ti/Al ratios changed from 1 : 1 to 1 : 3, which may be explained by the fact that the more aluminum that was absorbed required more heat to melt, which ultimately reduced the maximum temperature in the Ti–Al system. Bertolino³² reported that the maximum measured temperatures were 1200, 1400, and 1150 °C for Ti : Al = 3 : 1, 1 : 1 and 1 : 3, respectively, which is consistent with the present observations. Moreover, the chemical synthesis time from T_ig to T_c can be obtained from Fig. 1. When the target temperature was 650 °C and the Ti/Al ratios changed from 3 : 1 to 1 : 1, the synthesis times were 6 s, 4 s and 3 s, which corresponded with T_c values of 869 °C, 1068 °C and 1161 °C, respectively. However, with an increase in the Al content from Ti : Al = 1 : 1 to 1 : 3, the synthesis time was 3 s, 17 s and 226 s, which corresponded to T_c values of 1161 °C, 915 °C and 737 °C, respectively. The results indicated that the higher combustion temperature and the more severe exothermic reaction, the shorter the chemical synthesis time would be. It should also be noted that, for the specimens with Ti/Al molar ratios of 1 : 3, under the two furnace temperature conditions (Fig. 1e and f, furnace temperatures of 650 °C and 700 °C), both of the two specimens showed a TE phenomenon and they were made into Ti–Al porous monoliths. However, the TE profiles showed obvious differences. When the target furnace temperature was set at 650 °C, the TE reaction was ignited at 653 °C and took a long time (226 s) to finish the chemical synthesis with a low T_c, 737 °C. However, when the target furnace temperature was set at 700 °C, the TE reaction was ignited at 679 °C, and the synthesis occurred in a short time (10 s) with a high T_c, 1004 °C. Thus, the latter released more heat and possessed a higher combustion temperature than the former. Additionally, from the partial magnified images of the exothermic peaks in Fig. 1 before the combustion temperatures T_c were reached, obvious fluctuations and multiple temperature peaks were observed in the curves, and this phenomenon may be related to a multistep chemical reaction in the Ti–Al system.^17,27,35 To investigate the characteristics and mechanisms of the thermal explosion reaction, the products from different Ti/Al molar ratios were identified by X-ray diffraction (Fig. 2).

Fig. 2 XRD patterns of the TE reaction products.

3.2 Phase transformation

Fig. 2 shows XRD patterns of the TE products. In the specimens with the Ti : Al = 3 : 1, 2 : 1 and 1 : 1, the TE products were composed of three different intermetallic compounds, Ti₃Al, TiAl and TiAl₃ (Fig. 2a–c). The peaks of elemental Ti were found to be very strong and Ti was the predominant phase with an abundant TiAl phase and minor Ti₃Al and TiAl₃ phases, shown in Fig. 2a. In the specimen with Ti : Al = 2 : 1, the products included various phases like the specimen with Ti : Al = 3 : 1, with TiAl being the predominant phase, however, the peaks of Ti were still identifiable and strong in Fig. 2b. In Fig. 2c, though TiAl was the predominant compound, the peaks of Ti₃Al were obvious and those of Ti were weak. This meant that the residual Ti in the TE products decreased with the increase in the Al content (Fig. 2a–c). In the specimens with Ti : Al = 1 : 2 and 1 : 3, the TE reaction product was found to be solely the TiAl₃ compound, except for a small amount of residual Ti (Fig. 2d–f). Additionally, a comparison of Ti : Al = 1 : 3 specimens at two reaction conditions was conducted and Fig. 2e shows the XRD patterns of the specimen fabricated at 650 °C with low T_c (Fig. 1e) while Fig. 2f shows the XRD patterns of the specimen fabricated at 700 °C with high T_c (Fig. 1f). It can be found that the product phase was solely TiAl₃ regardless of the synthesis methods or the reaction parameters, indicating that the chemical synthesis reaction was completed under different target furnace temperatures and the same products were achieved. Comparing the intensity of the peaks of the residual Ti in Fig. 2e and f, it can be observed that the peak of Ti in Fig. 2f was weaker than those in Fig. 2e, indicating that the reaction was more complete when the furnace temperature was set at 700 °C.

Dahms et al.³⁵ reported a Ti–Al reaction mechanism of Ti + Al → TiAl₃ → Ti₃Al + TiAl + TiAl₂ → TiAl, but there was no detectable TiAl₂ phase in previous reports^17,18,27 or in the present research (Fig. 2). Gachon et al.³⁶ reported the mechanism of reaction and phase formation of intermetallic Al–Ti films, showing that there were two major reaction stages and two major stages of heat release during the formation of the final product phases of TiAl and TiAl₃ in the reactive multilayer films. Xu et al.³⁷ found that TiAl was the second phase formed after the formation of TiAl₃, whereas Peng et al.³⁸ believed that Ti₃Al should be the second phase due to it possessing a larger negative free energy than that of TiAl. Thus, the reactions are complex in Ti–Al system. Furthermore, to investigate the reactions of the Ti–Al–Nb system, Liang et al.²⁷ used the differential scanning calorimeter (DSC) to test the specimens with a heating rate of 20 °C min⁻¹, which was close to the heating rate of 17 °C min⁻¹ in present work. They summarized the three stages of the phase transformations during the sintering process, including Ti + 3Al → TiAl₃ between 600–720 °C, TiAl₃ + 2Ti → 3TiAl between 720–1000 °C and TiAl + 2Ti → Ti₃Al between 1000–1300 °C.²⁷ It can be observed from the magnifications of the exothermic peaks in temperature–time profiles (Fig. 1) that there were obvious fluctuations and multiple temperature peaks at approximately 800–900 °C. Bertolino reported similar behavior and observed a noticeable change in slope at approximately 870 ± 60 °C during the combustion synthesis.³² This may be due to the heat release actions in the multistep reactions of Ti–Al system. Multistep reactions involving phase transformations and heat release may result directly in the fluctuations of the temperature curves.³⁹

3.3 Expansion behavior and density

After the TE reaction, specimens with different Ti/Al molar ratios had a variety of macroscopic expansions, but all of the specimens maintained the original cylindrical shape and did not crack, as shown in Fig. 3a. Additionally, Fig. 3b shows the expansion characteristics as a function of Al content. When the target furnace temperature was set at 650 °C, the volume expansion first increased and then decreased with an increase in the Al content. Particularly, the volume expansion ratio reached a maximum value of 71.35% when the Al content was 52.99 wt% (Ti/Al molar ratio 1 : 2), and the axial and radial direction expansion behaviors were similar to the volume expansion behavior, reaching peak values of 19.76% and 19.62%, respectively. The expansions weakened at the Al content of 62.48 wt% (Ti/Al molar ratio 1 : 3), reducing to 31.87% (volume expansion), 9.84% (axial expansion) and 9.57% (radial expansion). However, when the target furnace temperature was set at 700 °C, the expansion ratios of the Ti : Al = 1 : 3 specimen appeared to change drastically and these ratios were significantly higher than the ratios at the target furnace temperature of 650 °C, reaching values of 60.48% (volume expansion), 15.77% (axial expansion) and 17.74% (radial expansion) (Fig. 3b). The different expansion ratios of both specimens of Ti : Al = 1 : 3 may be due to the TE reaction as well as the exothermic action. Combining the XRD results (Fig. 2e and f) and the TE curves (Fig. 1e and f), both specimens of Ti : Al = 1 : 3 showed the same products, so it can be speculated that the differences in the expansion behaviors were determined by the reaction process and the heat release rather than the final phase components. The latter specimen (700 °C) showed a more noticeable expansion phenomenon due to the high T_ig, high T_c and short reaction time (Fig. 1f).

Fig. 3 Expansion behaviors of the TE reaction specimens: (a) macroscopic expansion behavior; (b) volume, axial and radial expansion ratios (hollow symbols represent the results from the 700 °C furnace temperature).

The various expansion behaviors result in the different densities of the Ti–Al porous materials, as shown in Fig. 4. The density first decreased and then increased with an increase in the Al content, showing an opposite trend compared with the expansion ratios (Fig. 3b). For the specimen with the molar ratio of 1 : 2, the density of porous monoliths reached the lowest value of 1.58 g cm⁻³ while the expansion ratio reached the highest value of 71.35%. At the 650 °C target temperature, when the Ti/Al ratio was reduced to 1 : 3, the density of the monoliths increased to 2.16 g cm⁻³. However, the density was remarkably reduced when the target temperature was set at 700 °C, achieving the minimum value of 1.44 g cm⁻³. Just like with the expansion ratios, the differences between the two specimens of Ti : Al = 1 : 3 are due to the heat release reaction. The latter showed a higher ignition temperature, combustion temperature and shorter reaction time, which lead to more serious expansion behaviors and a lower density, whereas the former underwent a mild expansion phenomenon.

Fig. 4 Density of the TE reaction specimens (hollow symbols represent the results from the 700 °C furnace temperature).

3.4 Pore structure and elemental analysis

As indicated above, the expansion behaviors and the densities were dramatically affected by the Ti/Al molar ratios and the furnace temperature, therefore the pore structures of these porous materials may be similarly influenced by the Ti/Al molar ratios and furnace temperatures as well.

The skeletons and pore morphologies are shown in Fig. 5, including the SEM fracture images (Fig. 5a) and optical microscope images (Fig. 5b) of the porous Ti–Al alloys fabricated by thermal explosion. Fig. 5a shows that all of the fractures were brittle without melting–casting characteristics. It can be observed in Fig. 1 that the maximum combustion temperature was 1161 °C, which was lower than the melting points of final products, e.g., TiAl₃ (1350 °C), TiAl (1460 °C), Ti₃Al (1600 °C), and residual Ti (1678 °C). Therefore, the porosity should be influenced by the Ti/Al molar ratios and the reaction parameters. At the 650 °C target temperature, the monoliths had an interconnected and uniform porous structure (Fig. 5b). The open porosity of the Ti–Al porous alloys first increased and then decreased with an increase in the Al content (Fig. 6), and the quantity variation of the total porosity were similar to the open porosity. Moreover, with the increase in the Al content, the open porosities were measured to be approximately 37.94%, 40.81%, 48.75%, 53.89% and 35.25%, which corresponded with the specimens of Ti : Al = 3 : 1, 2 : 1, 1 : 1, 1 : 2 and 1 : 3, respectively. This meant that the open porosity of the specimen with Ti : Al = 1 : 3 reduced sharply to 35.25%, and the decline was consistent with the results of the Kirkendall effect.¹⁸ However, at the 700 °C target temperature, the open porosity of Ti : Al = 1 : 3 showed an abrupt increase (Fig. 5b), and the open porosity reached 56.27% (Fig. 6), the maximum value in all of the specimens. The reason behind the different pore characteristics of both of the Ti : Al = 1 : 3 specimens can be attributed to the expansion behaviors due to the heat release phenomenon.

Fig. 5 Pore structures of the TE reaction specimens: (a) SEM fracture micrographs; (b) optical microscope images of the polished specimens.

Fig. 6 Open porosity of the Ti–Al porous materials prepared via TE reaction (hollow symbols represent the results from the 700 °C furnace temperature).

EDS was used to confirm the presence of the elements in the TE reaction products. Fig. 7 showed the element composition in the specimen with Ti/Al molar ratios 1 : 3 (furnace temperature 700 °C) detected by EDS. Fig. 7a illustrated the magnification morphology of fracture surface of porous Ti–Al monolith, which implied that the TE products were particle structures. Fig. 7b showed that Al mainly distributed in the region of the particle products, while Fig. 7c showed that Ti distributed evenly on the fracture surface. Fig. 7d showed that particle products mainly included Ti and Al elements, confirming that the particles were Ti–Al compounds. The EDS point analysis results (as shown in Fig. S2†) confirmed that the atomic ratios of Ti : Al were in relatively close agreement with 1 : 3. Combined with XRD pattern (Fig. 2f), it was suggested that Ti–Al particles were TiAl₃ compounds. Meanwhile, there were clear indications of some golden Ti elements around the product particles (Fig. 7d), indicating that the golden Ti elements were residual Ti. These were in agreement with the results of the XRD patterns, and the TE products were composed of TiAl₃ compounds with a small amount of residual Ti. The surface EDS spectrum and composition confirmed that the atomic ratio of Ti : Al was 28.19 : 71.81 (Fig. 7e), which was also close to stoichiometric proportions of nominal compositions. To the ideal result, the atomic ratio of Ti : Al should be equal to 1 : 3 in sintered specimen, but the EDS result was not consistent completely with the ideal result, which was due to the semi-quantitative analysis of EDS. As known, the EDS result was hard to match with the actual composition.^23,40 For other TE specimens, the EDS results were similar to that of Ti : Al = 1 : 3 (700 °C).

Fig. 7 Fracture surface EDS of porous specimen with Ti : Al = 1 : 3 (furnace temperature 700 °C).

3.5 Oxidation properties

The mass gain observed in Fig. 8 illustrates the oxidation kinetics of the Ti–Al porous materials, and all of the oxidation kinetic curves exhibited parabolic-like behavior, indicating that the Ti–Al porous materials have superior high-temperature oxidation resistance. The mass gain after 96 h reached the maximum values of 2.1416 kg m⁻², 1.0298 kg m⁻², 0.3576 kg m⁻², 0.4032 kg m⁻², 0.1851 kg m⁻² and 0.2996 kg m⁻² (700 °C), which corresponded with the specimens with Ti/Al molar ratios of 3 : 1, 2 : 1, 1 : 1, 1 : 2, 1 : 3 and 1 : 3 (furnace temperature of 700 °C), respectively. These curves demonstrate that the porous alloys prepared from Ti : Al = 3 : 1 and 2 : 1 exhibited a poorer oxidation resistance than the other four specimens, which had larger mass gain during the oxidation process, indicating that they were not suitable for a high temperature environment. In contrary, the curves of 1 : 1, 1 : 2 and 1 : 3 show lower mass gain that represent superior high-temperature oxidation resistances. In particular, Fig. 8 shows that the specimens of Ti : Al = 1 : 3 have the least mass gain regardless of the furnace temperature conditions, revealing their excellent resistance to oxidation. Even though the specimen of 1 : 3 possesses a maximum porosity at 700 °C, it still has outstanding oxidation resistance. The larger mass gain in of 3 : 1 and 2 : 1 were ascribed to the redundant titanium after the TE reaction because the titanium was easily oxidized to form TiO₂ during the oxidation process,⁴¹ which was not beneficial to the high temperature oxidation. The improvement of the oxidation resistance with the increase in the Al content should be due to the decrease of the residual Ti and the increase of the TiAl₃ compound (Fig. 2c–f).^23,24

Fig. 8 Mass gain of the Ti–Al porous monoliths oxidized at 650 °C in air.

3.6 Discussion

According to the phase identifications (Fig. 2d–f), it can be determined that TiAl₃ is the main product phase in the Ti–Al systems of the Al-rich specimens by the TE reaction. Additionally, the Al-rich specimens with Ti/Al molar ratios 1 : 2 and 1 : 3 (furnace temperature 700 °C) possessed a larger open porosity, were solely comprised of the TiAl₃ phase and possessed superior high-temperature oxidation resistance, indicating that the Al-rich specimens are suitable for use in high temperature environments. In the present work, we discussed the reaction mechanisms and pore formation models of the Ti–Al porous alloys made from the Al-rich specimens (i.e., Ti/Al molar ratios 1 : 2 and 1 : 3).

Traditionally, in the range from the room temperature to the ignition temperature T_ig, the reactive diffusion and the Kirkendall effect of the solid Ti–solid Al phase occurred due to the large differences of the diffusion coefficients of Ti and Al.^18,26 He et al. fabricated Ti–Al porous alloys between 550–650 °C for 60–240 min, and the specimens were held for enough time in order for pores to be generated through the Kirkendall effect, and then the specimens were finally sintered at elevated temperatures ranging from 800 to 1300 °C for 30–300 min. The heating rate was set at 2 °C min⁻¹ throughout the whole process.²⁶ However, in the present study, the short time (i.e., approximately 38 min from the room temperature to ignition temperatures due to the heating rate of 17 °C min⁻¹) and the weak interdiffusion between Ti and Al resulted in the limitation of Kirkendall effect, and the Kirkendall effect should be ignored in the Ti–Al system during the TE reaction.⁴²

When the specimen was heated to the ignition temperatures, a sudden slope appeared and the temperatures of specimen increased sharply, implying that the Al powders started to melt. The reaction between the solid Ti and liquid Al was violent due to the liquid aluminum possessing a much higher reactivity than solid aluminum. The solid Ti powders dispersed throughout the liquid Al, and TiAl₃ was nucleated and quickly grew in the molten aluminum.^32,42 The large amounts of heat released from the TiAl₃ formation in the Ti–Al system lead to an instant rise in the temperature–time profiles from the ignition temperature to the combustion temperature. The consumption of the aluminum resulted in the formation of the in situ pores in the original aluminum particle sites. After T_c, the specimen was cooled to the furnace temperature. During the TE procedure, Ti and Al reacted violently, and the specimens released massive amounts of heat and gas, which caused the formation of a large number of pores in the specimen.

Combined with the reaction mechanism in the Ti–Al system, in the present work we proposed a pore-formation model of the Ti–Al porous materials prepared by a thermal explosion reaction, as observed in Fig. 9:

Fig. 9 The schematic diagram of pore-formation model of the Ti–Al porous alloys.

(a) The green compacts formed by the cold-pressed method could not be fully dense. Many pores exist among the Ti and Al powder particles inside of the green compacts. In the present experiments, the total porosities of the green compacts with Ti/Al molar ratios of 3 : 1, 2 : 1, 1 : 1, 1 : 2 and 1 : 3 were measured to be 35.48%, 31.25%, 29.48%, 24.90% and 20.67%, respectively, which corresponds to 61.56%, 68.75%, 70.52%, 75.10% and 79.33%, respectively, of the relative density. Thus, there are many pores in the green compacts, which are the important sources for pores in the Ti–Al porous materials (Fig. 9a).

(b) Before T_ig, with the increase of the furnace temperature, the gases by absorbed by the Ti and Al powder particles diffused from the compacts, which formed vacancies among the particles, and the vacancies connected the original pores in the green compact to form interconnected pores (Fig. 9b).

(c) When the temperature was increased to the ignition temperature, the green compact was quickly ignited and the TE reaction occurred rapidly. The reaction of the solid Ti and the liquid Al takes place quickly to generate the Ti–Al compounds. The obvious expansion behaviors were observed in all of the specimens due to the serious exothermic nature of the reaction, which increased the pores in the monoliths. Moreover, the large phase grain was broken into many fine grains from the pressure of the liquid Al, which formed small interconnected pores.^27,43 The pores that were derived from the TE reaction between the solid Ti and liquid Al were due to the consumption of the liquid Al, where the consumption of the Al results in the formation of in situ pores in the original Al particle sites (Fig. 9c).

(d) After TE, Ti–Al intermetallics formed the skeleton of the network structures of the porous monoliths, as shown in Fig. 5b. In the final monoliths, there are mainly three structures, including Ti–Al intermetallics skeleton, large pores among skeletons and small pores in skeletons. At the same time, the pores combined with each other to form interconnected pores (Fig. 9d).

Compared with the traditional vacuum sintering^{17–19,26,27} to fabricate Ti–Al porous alloys, TE reaction is found to be more promising. The method of vacuum sintering takes longer sintering time (8.7–19.8 h) to prepare porous Ti–Al, which consumes a large amount of energy.^19,26,27 TE reaction only takes less than 40 min (from room temperature to 650 °C or 700 °C with a constant heating rate of 17 °C min⁻¹) to fabricate porous Ti–Al. The open porosity of TE specimens is slightly higher than the previous literature.^17,18 The open porosity of Ti : Al = 1 : 1 reaches 48.75%, which is higher than the porosity of 45.9%.¹⁷ And the open porosity of Ti : Al = 1 : 3 reaches 56.27%, which is the same to the specimen sintered by vacuum sintering, only approximating to 56%.¹⁸ Thus, the present work opens the way to synthesize high porosity Ti–Al intermetallics using an energy and time saving route.

4. Conclusions

(1) Ti–Al porous alloys were prepared by the simple and energy-saving thermal explosion reaction. The obvious heat release phenomenon was observed, and all of the porous monoliths kept their original cylindrical shape and did not crack after the thermal explosion.

(2) At the 650 °C target furnace temperature, the chemical synthesis time from T_ig to T_c of Ti : Al = 1 : 1 reached a minimum value of 3 s, and the time increased to a maximum value of 226 s for Ti : Al = 1 : 3. The actual combustion temperature first increased and then decreased with an increase in the Al content. The specimen of Ti : Al = 1 : 1 possessed the maximum combustion temperature of 1161 °C. The porous monoliths fabricated from Ti : Al = 1 : 2 possessed the largest porosity, the lowest density when compared with the other four specimens and exhibited good resistance to oxidation.

(3) At the 700 °C target furnace temperature, the specimen of Ti : Al = 1 : 3 possessed a higher ignition temperature, 679 °C, a higher combustion temperature, 1004 °C, a shorter synthesis time, 10 s, a lower density, 1.44 g cm⁻³, and a larger open porosity, 56.27%, than the other specimen of Ti : Al = 1 : 3 at the 650 °C target temperature. Both the specimens of Ti : Al = 1 : 3 at the different furnace temperatures showed an excellent resistance to oxidation.

(4) The original porosities of the green compacts were between 20.67–35.48%, and these were the important sources of the pores for the Ti–Al materials. The obvious expansion behaviors that occurred during TE further improved the porosities of the Ti–Al porous monoliths.

Acknowledgements

This work was supported by the Fundamental Research Funds for the Central Universities (2012QNA02), National Natural Science Foundation of China (U1361213 and 51476184), Jiangsu Province Science Fund for Distinguished Young Scholars (BK20140005) and the Opening Research Fund of State Key Laboratory of Porous Metal Materials (PMM-SKL-3-2012).

Notes and references

1. D. Rath, S. Rana and K. M. Parid, RSC Adv., 2014, 4, 57111–57124.
2. A. Del Regno and F. R. Siperstein, Microporous Mesoporous Mater., 2013, 176, 55–63.
3. A. Walcarius, Chem. Soc. Rev., 2013, 42, 4098–4140.
4. T. Yanagisawa, T. Schhimizu, K. Kuroda and C. Kato, Bull. Chem. Soc. Jpn., 1990, 63, 988–992.
5. L. Q. Duan, Q. S. Ma and Z. H. Chen, J. Eur. Ceram. Soc., 2013, 33, 841–846.
6. E. A. Franceschini, G. A. Planes, F. J. Williams, G. J. Soler-Illia and H. R. Corti, J. Power Sources, 2011, 196, 1723–1729.
7. P. S. Liu and K. M. Liang, J. Mater. Sci., 2001, 36, 5059–5072.
8. P. Colombo, Philos. Trans. R. Soc., A, 2006, 364, 109–124.
9. P. Sooksaen and S. Karawatthanaworrakul, Appl. Clay Sci., 2015, 104, 295–302.
10. A. R. Studart, U. T. Gonzenbach, E. Tervoort and L. J. Gauckler, J. Am. Ceram. Soc., 2006, 89, 1771–1789.
11. R. F. Chen, C.-A. Wang, Y. Huang, L. G. Ma and W. Y. Lin, J. Am. Ceram. Soc., 2007, 90, 3478–3484.
12. E. C. Hammel, O. L.-R. Ighodaro and O. I. Okoli, Ceram. Int., 2014, 40, 15351–15370.
13. D. Nguyen-Manh, M. J. Cawkwell, R. Gröger, M. Mrovec, R. Porizek, D. G. Pettifor and V. Vitek, Mater. Sci. Eng., A, 2005, 400–401, 68–71.
14. S. K. Sadrnezhaad and S. A. Hosseini, Mater. Des., 2009, 30, 4483–4487.
15. P. Z. Shen, Y. H. He, H. Y. Gao, J. Zou, N. P. Xu, Y. Jiang, B. Y. Huang and C. T. Liu, Desalination, 2009, 249, 29–33.
16. D. Q. Li, H. B. Guo, D. Wang, T. Zhang, S. K. Gong and H. B. Xu, Corros. Sci., 2013, 66, 125–135.
17. F. Yang, M. Tane, J. P. Lin, Y. H. Song and H. Nakajima, Mater. Des., 2013, 49, 755–760.
18. Y. Jiang, Y. H. He, N. P. Xu, J. Zou, B. Y. Huang and C. T. Liu, Intermetallics, 2008, 16, 327–332.
19. F. Yang, L. Q. Zhang, J. P. Lin, Y. F. Liang, Y. H. He, S. L. Shang and Z.-K. Liu, Intermetallics, 2013, 33, 2–7.
20. Z. Zheng, Y. Jiang, H. X. Dong, L. M. Tang, Y. H. He and B. Y. Huang, Trans. Nonferrous Met. Soc. China, 2009, 19, 581–585.
21. H. R. Gong, Y. H. He and B. Y. Huang, Appl. Phys. Lett., 2008, 93, 101907–101909.
22. Y. H. Wang, J. P. Lin, Y. H. He, C. K. Zu and G. L. Chen, J. Alloys Compd., 2010, 492, 213–218.
23. C. Angeles, G. Rosas and R. Perez, Mater. Chem. Phys., 1998, 56, 262–265.
24. X. W. Huo, S. Q. Wang and K. M. Chen, J. Aeronaut. Mater., 2007, 27, 71–76.
25. S. C. Ferreira, L. A. Rocha, E. Ariza, P. D. Sequeira, Y. Watanabe and J. C. S. Fernandes, Corros. Sci., 2011, 53, 2058–2065.
26. Y. H. He, Y. Jiang, N. P. Xu, J. Zou, B. Y. Huang, C. T. Liu and P. K. Liaw, Adv. Mater., 2007, 19, 2102–2106.
27. Y. F. Liang, F. Yang, L. Q. Zhang, J. P. Lin, S. L. Shang and Z.-K. Liu, Intermetallics, 2014, 44, 1–7.
28. S. K. Mishra, V. Gokuul and S. Paswan, Int. J. Refract. Met. Hard Mater., 2014, 43, 19–24.
29. J. Y. Xu, B. L. Zou, S. M. Zhao, Y. Hui, W. Z. Huang, X. Zhou, Y. Wang, X. L. Cai and X. Q. Cao, Ceram. Int., 2014, 40, 15537–15544.
30. P. Z. Feng, W. S. Liu, A. Farid, J. Wu, J. N. Niu, X. H. Wang and Y. H. Qiang, Adv. Powder Technol., 2012, 23, 133–138.
31. J. J. Moore and H. J. Feng, Prog. Mater. Sci., 1995, 39, 243–273.
32. N. Bertolino, M. Monagheddu, A. Tacca, P. Giuliani, C. Zanotti and U. Anselmi Tamburini, Intermetallics, 2003, 11, 41–49.
33. S. W. Jo, G. W. Lee, J. T. Moon and Y. S. Kim, Acta Mater., 1996, 44, 4317–4326.
34. P. Z. Feng, A. Farid, X. H. Wang, W. S. Liu, J. Wu, S. Zhang and Y. H. Qiang, J. Alloys Compd., 2010, 494, 301–304.
35. M. Dahms, J. T. Jewett and C. Michaelsen, Z. Metallkd., 1997, 88, 125–130.
36. J.-C. Gachon, A. S. Rogachev, H. E. Grigoryan, E. V. Illarionova, J.-J. Kuntz, D. Yu. Kovalev, A. N. Nosyrev, N. V. Sachkova and P. A. Tsygankov, Acta Mater., 2005, 53, 1225–1231.
37. L. Xu, Y. Y. Cui, Y. L. Hao and R. Yang, Mater. Sci. Eng., A, 2006, 435–436, 638–647.
38. L. M. Peng, J. H. Wang, H. Li, J. H. Zhao and L. H. He, Scr. Mater., 2005, 52, 243–248.
39. A. Biswas, Acta Mater., 2005, 53, 1415–1425.
40. C. D. Brunetta, J. A. Brant, K. A. Rosmus, K. M. Henline, E. Karey, J. H. MacNeil and J. A. Aitken, J. Alloys Compd., 2013, 574, 495–503.
41. T. Shimizu, T. Iikubo and S. Isobe, Mater. Sci. Eng., A, 1992, 153, 602–607.
42. S. Y. Jiang, S. C. Li and L. Zhang, Trans. Nonferrous Met. Soc. China, 2013, 23, 3545–3552.
43. H. J. Huang, H. Y. He, Y. Jiang, B. Y. Huang and N. P. Xu, Chin. J. Mater. Res., 2007, 21, 337–342.

---

Footnote

† Electronic supplementary information (ESI) available: Fig. S1 and S2. See DOI: 10.1039/c5ra04047g

---