Transformation from D0₂₂ to L1₂ in Al₃Ti by Fe Addition for Enhanced Wear Resistance

Abstract

The addition of third elements may help transform brittle D0₂₂-structured lightweight Al₃Ti to a relatively ductile L1₂-structured (Al, M)₃Ti (where M represents the third elements), thus increasing the ductility at the expense of hardness. Such a transformation could benefit the wear resistance of the alloy due to improved toughness if a proper balance between the hardness and ductility is achieved. In this work, a D0₂₂-predominant Al₃Ti alloy (S-Al₃Ti) and an L1₂-predominant (Al, Fe)₃Ti alloy (S-Al₆₇Ti₂₅Fe₈) were fabricated by arc melting. Change in wear resistance, corresponding to a D0₂₂-to-L1₂ transformation, caused by the addition of Fe as a representative third element, was investigated and compared with the wear resistance of a commercial Al-matrix composite reinforced by 30 wt.% SiC particles (S-Al/SiC_p) as a reference material. It was observed that wear of the S-Al₃Ti resulted from abrasion involving synergistic oxidation, leading to a larger volume loss. In contrast, the softer S-Al₆₇Ti₂₅Fe₈ showed enhanced wear resistance, benefiting from improved toughness with reasonable hardness. During the wear testing, both the alloys exhibited better performance than S-Al/SiC_p, a well-known lightweight composite. This study highlights that D0₂₂-to-L1₂ transformation enhances wear resistance due to increased toughness which can be adjusted using the addition of a third element.

1. Introduction

Al-Ti intermetallic compounds are very attractive candidates for the structural components used at elevated temperatures due to their low density, high melting point and high specific strength [1,2,3]. The long-range ordering in the intermetallic compounds effectively impedes atomic diffusion and pin dislocations, leading to a high resistance to softening at elevated temperatures, compared to solid-solution alloys [4,5]. According to the Al-Ti binary phase diagram, four Al-Ti intermetallic compounds, i.e., AlTi, Al₂Ti, Al₃Ti and Ti₃Al, are promising for high-temperature applications [6], and Al₃Ti is the most appealing due to its advantageous density [7].

Binary Al₃Ti may have two commonly encountered crystal structures; one is a stable tetragonal D0₂₂ structure with a certain degree of anisotropy due to its low symmetry, and the other is a metastable cubic L1₂ structure with relatively isotropic properties due to its higher symmetry [8]. Figure 1 illustrates the crystal structures of D0₂₂ Al₃Ti and L1₂ Al₃Ti. Yamaguchi et al. [9] reported that the primary deformation mode of D0₂₂ Al₃Ti during compression below 620 °C is <111> [112] twinning, with a dislocation slip activated at higher temperatures, resulting in improved ductility. The brittleness of the D0₂₂ structure at room temperature, because of its insufficient slip systems [10], remains a significant challenge for practical applications of Al₃Ti [11].

The D0₂₂ structure is more ionic with stronger bonding than that of L1₂ structure. In the D0₂₂ structure, Ti atoms gain charges and Al atoms lose charges [12]. Alloying Al₃Ti with a small amount of one or multiple transition metal elements, such as Cr, Mn, Fe, Ni, Cu, and Ag [13,14,15,16,17,18,19,20,21], can transform the tetragonal D0₂₂ structure into the cubic L1₂ structure. These less electropositive third elements may replace Al, taking electrons from Al and suppressing the charge transfer from Al to Ti, which stabilizes the L1₂ structure [22,23]. The cubic L1₂ structure provides more slip systems for plastic deformation, thus enhancing the ductility. The D0₂₂-to-L₁₂ transformation can help achieve improved plasticity, which is beneficial to practical applications of Al₃Ti. In general, the improvement in plasticity is achieved at the expense of strength and hardness, which may lead to negative influence on some properties such as the resistance to wear [24]. A theoretical study on tetragonal-D0₂₂ and cubic-L1₂ structures via calculations with density functional theory by Boulechfar et al. [25] demonstrated that the L1₂ phase showed lowered brittleness and hardness as well, compared to the D0₂₂ structure. Milman et al. [26] compared the hardness of tetragonal Al₃Ti with that of cubic L1₂ Cr-doped Al₆₁Cr₁₂Ti₂₇ and Mn-doped Al₆₆Mn₁₁Ti₂₃, and observed that the hardness decreased by about 40% with Cr doping and 60% with Mn doping, respectively.

Wear, as one of the material failure processes, occurs when two surfaces rub each other, causing surface damage or failure. Estimates show that up to 80% of machine parts failure involves wear [27]. The economic loss caused by friction and wear reaches 2~7% of the gross domestic product for industrial countries [27]. Wear is a complex failure process, influenced by many factors, e.g., environmental aggressiveness, the wear loading condition, and microstructure modification, etc. [28,29,30,31,32], which could be mitigated using different approaches, including alloying, microstructure engineering, heat treatment, and texture engineering, etc. [33,34,35,36,37]. Mechanical properties, such as properly balanced hardness/strength and ductility/toughness, are indispensable to desirable wear resistance. Given the improved ductility and reduced strength/hardness by the transformation from the hard but brittle D0₂₂ Al₃Ti to the ductile L1₂ (Al, M)₃Ti (where M represents ternary elements), the corresponding wear resistance is accordingly affected, which should be adjustable by tuning the balance between hardness and ductility or toughness. However, there are few studies on the wear behaviors of D0₂₂ Al₃Ti and L1₂ (Al, M)₃Ti reported in the literature. It is, thus, of importance to investigate how the wear resistance is affected by the phase transformation.

The objectives of this study are (1) to investigate the wear behavior of tetragonal D0₂₂ Al₃Ti and that of cubic L1₂ (Al, M)₃Ti with improved ductility but reduced hardness, and (2) to reveal the underlying mechanism for corresponding variations in microstructure and relevant properties in order to further improve or optimize the alloy system. For details, the following tasks were undertaken in this study:

(a)

Fe was selected as a representative third element to be alloyed into Al₃Ti, turning the D0₂₂-structured Al₃Ti to a ductile L1₂ structure (Al, M)₃Ti. Fe was selected for this study based on the consideration of its abundant resource, low cost, and the industrial relevance or interest in ferrous alloys. It should be indicated that adding Fe would not affect the alloy’s resistance to corrosion, since the alloy contains passive elements Al and Ti of high concentrations, which generate a protective passive film that can prevent continuous corrosion when used in corrosive environments;

(b)

The wear resistances of the two fabricated alloys were evaluated, in comparison to that of a well-known 30 wt.% SiC particle-reinforced Al-matrix composite as a reference material;

(c)

Corresponding changes in the microstructure, mechanical properties, and wear behavior were analyzed in detail in order to elucidate the underlying wear mechanisms for further improvement or optimization.

2. Materials and Methods

Two alloys, with nominal compositions of Al₃Ti and Al₆₇Ti₂₅Fe₈ (in molar ratio or at.%, denoted as S-Al₃Ti and S-Al₆₇Ti₂₅Fe₈, respectively), were fabricated using an arc melting process. High-purity Al, Ti and Fe powders (≥99.9%) were mixed and pressed into disk-like bulks, followed by arc melting in a water-cooled copper hearth using an arc melting furnace (MRF Inc., Allenstown, NH, USA) in a protective argon atmosphere. The sample preparation procedures are pictorially illustrated in Figure 2. Both of the samples were re-melted four times to minimize the chemical inhomogeneity.

The crystal structure and phases of the alloys were analyzed using an X-ray diffractometer (Rigaku XRD Ultima IV, Tokyo, Japan) at a scanning speed of 4°/min. The microstructure was characterized using a scanning electron microscope (Zeiss EVO M10 SEM, Carl Zeiss AG, Oberkochen, Germany) equipped with energy-dispersive X-ray spectroscopy (EDS) after mechanically grinding and polishing.

The macro-hardness of the alloys was measured using a ZHR Rockwell hardness tester (ZwickRoell, Kennesaw, GA, USA) with a conical diamond indenter under a load of 60 kgf at room temperature. Each polished sample was tested by indentation in 5 randomly selected positions to obtain the average hardness, and error bars were then determined based on the standard deviation. The indentation marks made by a diamond indenter with a square pyramid shape under a load of 15 kgf were observed to visually compare the toughness of the samples.

Sliding wear tests were performed using a ball-on-disk tribometer (CSEM Instruments, Peseux, Switzerland). A silicon nitride (Si₃N₄) ball with a diameter of 3 mm was used as the counter-face to wear the alloy samples at room temperature. The samples were ground and polished before the wear tests. The duration of each test was 30 min. The applied normal load was 3 N. The rotary speed was 191 rpm with a circular path of about 2 mm in diameter. The wear test for each sample was repeated three times to obtain the average value of wear volume loss. The volume losses from the wear tracks and the worn surface morphologies of the samples were automatically measured and analyzed, respectively, using a three-dimensional (3D) optical profilometer (ZeGage™ Pro, Zygo, Middlefield, CT, USA) with installed software (version 1.14.38, ZeMaps). The worn surfaces were further characterized using SEM with EDS.

3. Results

3.1. Microstructure Characterization

The XRD pattern of S-Al₃Ti is displayed in Figure 3a. The main peaks of the D0₂₂ Al₃Ti phase are accompanied by several small peaks of a second phase which is identified as pure Al, indicating that the D0₂₂ Al₃Ti phase is predominant in the S-Al₃Ti alloy. The microstructure of S-Al₃Ti was examined by SEM in the back-scattered electron (BSE) mode, and is illustrated in Figure 3b,c. One may see that the second phase is distributed along grain boundaries, which accounts for approximately 3 vol.%. The EDS results shown in Figure 3d–f confirm the presence of Al as second phase, along with the predominant D0₂₂ Al₃Ti phase.

The XRD pattern of S-Al₆₇Ti₂₅Fe₈ is shown in Figure 4a. The main peaks come from the L1₂ phase, accompanied by several small peaks of second phase(s), indicating that S-Al₆₇Ti₂₅Fe₈ was dominated by the L1₂ phase. The microstructure of S-Al₆₇Ti₂₅Fe₈ observed with the SEM in BSE mode is shown in Figure 4b,c. One may see that there are two second phases: one is a light gray phase and the other is a bright one inside the light gray phase, which account for approximately 7 vol.% in total. The EDS results presented in Figure 4d–g confirm that the L1₂ predominant phase has a composition of (Al, Fe)₃Ti, and show that the second phases are Fe-rich with depleted Ti.

In binary Al₃Ti, the D0₂₂ structure is more stable, in which the charge transfers from Al to Ti, contributing to strong bonding strength. When the third element, which is less electropositive than Al, is added, it substitutes for Al and takes electrons from Al and consequently suppresses the charge transfer from Al to Ti. As a result, the L1₂ structure (Al,M)₃Ti becomes more stable than D0₂₂ (Al,M)₃Ti, i.e., the L1₂ structure is stabilized by adding third elements. More discussions with more details and references have been given in Section 4. The phase transformation from D0₂₂ to L1₂ certainly affects the mechanical properties, i.e., the ductility increases at the expense of hardness. The corresponding change in the balance between the hardness and ductility consequently influences the combination of hardness and toughness and, thus, the wear resistance. The resultant variations in wear resistance caused by the phase transformation are reported in the next section.

Figure 5 illustrates the microstructure of the 30 wt.% SiC particle-reinforced Al-matrix composite (denoted as S-Al/SiC_p), which was used as a reference material when evaluating the wear resistance of the D0₂₂ Al₃Ti-predominant S-Al₃Ti alloy and the L1₂ (Al, Fe)₃Ti-predominant S-Al₆₇Ti₂₅Fe₈ alloy. As the SEM-BSE image in Figure 5a shows, dense particles in dark were distributed in the light gray matrix. According to the corresponding EDS maps shown in Figure 5b–d, the dark particles were SiC and the light gray matrix was Al. The SiC particles significantly hardened the Al matrix, playing an important role in contributing to the wear resistance of the S-Al/SiC_p composite.

3.2. Wear Behavior

The Rockwell hardness values of three samples, S-Al₃Ti, S-Al₆₇Ti₂₅Fe₈, and S-Al/SiC_p as a reference material, were measured at room temperature using a conical diamond indenter under a load of 60 kgf. Results of the tests are displayed in Figure 6. As shown, S-Al₃Ti exhibited the highest hardness amongst the three samples. S-Al₆₇Ti₂₅Fe₈ showed a slightly lower hardness, compared to that of S-Al₃Ti, suggesting that the transformation of the D0₂₂ phase predominant in S-Al₃Ti to L1₂ (Al, Fe)₃Ti in S-Al₆₇Ti₂₅Fe₈ did not decrease the hardness much. Although having densely distributed reinforcing SiC particles, the hardness of the S-Al/SiC_p composite was markedly lower than those of the two alloys.

In this study, S-Al/SiC_p was used as a reference material for the comparison purpose. S-Al₃Ti and S-Al₆₇Ti₂₅Fe₈ had relatively low densities (~3.17 and 3.48 g/cm³, respectively). The present study was conducted with the motivation to develop Al₃Ti-based lightweight alloys for tribological applications. Since the commercial S-Al/SiC_p possessed a good wear resistance with its density around 2.84 g/cm³, close to those of the above-mentioned alloys, selecting this well-known material as a reference material should facilitate this comparison with relatively quantitative information.

Indents caused by a diamond indenter with a square pyramid shape under a load of 15 kgf were observed under SEM in the second electron (SE) mode. The obtained images are displayed in Figure 7, showing that S-Al₃Ti (Figure 7a) is the hardest, S-Al₆₇Ti₂₅Fe₈ (Figure 7b) is less hard, and Al/SiC_p (Figure 7c) is the softest among the three materials. Comparing the morphologies of the indents on S-Al₃Ti and S-Al₆₇Ti₂₅Fe_8, the zone around the indent on the former showed some wrinkles and micro-cracks, while the latter had much less these features, indicating that S-Al₃Ti is brittle or less ductile, compared to S-Al₆₇Ti₂₅Fe₈.

The volume losses of the samples after sliding wear testing under 3 N in the ambient environment for 30 min are shown in Figure 8a, and the wear track profiles are displayed in Figure 8b₁–d₂. In spite of the lower hardness, the alloy S-Al₆₇Ti₂₅Fe₈ demonstrated a higher wear resistance than the harder alloy S-Al₃Ti, which should benefit from its L1₂ (Al, Fe)₃Ti phase with higher plasticity than that of the harder but less ductile D0₂₂ Al₃Ti phase in S-Al₃Ti. A combination of strength and ductility/plasticity can make a reasonably hard material tougher against wear attacks. The composite S-Al/SiC_p showed the lowest wear resistance due to its lowest hardness among the three materials.

In order to elucidate the wear mechanisms for different performances of the three materials during wear tests, their worn surface morphologies were observed and analyzed, as illustrated in Figure 9 and Figure 10. S-Al₃Ti showed its worn surface involving wear and oxidation caused by frictional heating [38,39], as evidenced by the detected oxygen distribution shown in the inset of Figure 9a₁, and the oxide scale spallation shown in Figure 9a₂,a₃ (yellow arrow). In addition, surface fragmentation was also observed on S-Al₃Ti near the area where a piece of oxide scale was removed, as shown in Figure 9a₂ (magenta arrow), due to its lower plasticity or brittleness, compared to that of S-Al₆₇Ti₂₅Fe₈. In contrast, the worn surface of S-Al₆₇Ti₂₅Fe₈ showed more abrasive wear characteristics with scratching grooves as illustrated in Figure 9b₁. Oxidation occurred on the surface as well, but it appeared to be much less, as one may see in the oxygen distribution shown in the inset of Figure 9b₁. Very tiny craters were observed on or around the Fe-rich second phase (marked by yellow circles) connecting to the dominant L1₂ (Al, Fe)₃Ti phase, as shown in Figure 9b₂, which could be an indication of the brittleness of the Fe-rich Ti-depleted second phases in comparison with that of the L1₂ phase. Due to the mechanical difference between the second phase and the L1₂ phase, the local stress concentrations induced by incompatible deformations of the two phases under the wearing stress could induce local beak-off, which could trigger crack initiation. During the continuous wear process, micro-cracks could propagate and lead to eventual wear volume loss.

With higher hardness, S-Al₃Ti was supposed to be subject to less abrasive wear damage when compared to S-Al₆₇Ti₂₅Fe₈. However, the wear damage to S-Al₃Ti is severer than that to S-Al₆₇Ti₂₅Fe₈, as demonstrated by Figure 9a₁,b₁. This happened, as discussed earlier. The less protective oxide scale on S-Al₃Ti with obvious spallation and surface fragmentation could make the alloy less resistant to wear attack. The oxide spall-off along with abraded material pieces of material may act as debris to result in additional wear damage [40,41]. Thus, the brittleness of the substrate, oxidation, and the spallation of the oxide scales could be responsible for the larger wear volume loss of S-Al₃Ti, compared to that of S-Al₆₇Ti₂₅Fe₈.

Regarding the wear of S-Al₆₇Ti₂₅Fe₈ with decreased hardness and increased toughness, less oxidation was involved, evidenced by less oxygen contents inside and outside the wear track (see the inserted figure in Figure 9b₁), and no oxide scale spallation was observed (see Figure 9b₁,b₂). The reduced brittleness with maintained reasonable hardness led to reduced abrasive wear. Consequently, S-Al₆₇Ti₂₅Fe₈ showed a lower volume loss or higher wear resistance, compared to S-Al₃Ti. Regarding oxidation, as shown in Figure 9b₃, oxygen was still detected, but oxidation was markedly less, which can be ascribed to two possible reasons: (1) the oxide film was compacted and protective, so the oxide growth was slow, leading to the formation of a thin oxide film; and (2) the thin oxide film was removed during the wear testing, and re-growth was slow due to higher protectiveness of the oxide film. Bear in mind that a protective oxide film can form rapidly in the initial stage and the following growth becomes slow since the protective film more effectively hinders the diffusion of metallic atoms and oxygen, the continuous oxidation can thus be stopped. A protective oxide film helps reduce wear and minimize synergistic attacks by wear and oxidation. However, it is not very clear why the Fe addition improved the oxidation behavior or the protectiveness of the oxide scale on S-Al₆₇Ti₂₅Fe₈. Analyzing the influences of Fe on the stability or cohesive energy, the adherence and mechanical properties of the oxide scale on the S-Al₆₇Ti₂₅Fe₈ substrate would help understand the role that the Fe addition plays. These will be included in our follow-up studies.

It should also be mentioned that no sign of adhesive wear was noticed. In the obtained EDS maps, Si was not observed due to its insufficient content, implying that the adhesive transfer from the Si₃N₄ ball to the sample under study was negligible. There were also no noticeable patches of material peel-off on the alloy samples, i.e., alloy transfer. Thus, adhesive wear should have little contribution to wear in the present case, and abrasive wear was the dominant wear mechanism with influences of synergistical oxidation for both S-Al₃Ti and S-Al₆₇Ti₂₅Fe₈ at different extents.

The worn surface of S-Al/SiC_p, as a reference material, which exhibited the lowest wear resistance among the three samples due to its lower hardness, is illustrated in Figure 10. The characteristics of abrasive wear with obvious plastic deformation were observed as shown in Figure 10a. The enlarged image in Figure 10b reveals dense wear debris, abrasive grooves, and some adhesive patches. During the wear test, the soft Al matrix experienced severe plastic deformation under the wearing force, forming flakes that adhered to the surface during repeated sliding when oxidation was also involved, as evidenced by the oxygen distribution shown in the inset of Figure 10a. In S-Al/SiC_p, the Al matrix may not be strong enough to effectively maintain the reinforcing SiC particles during the wear process. Thus, SiC particles may be either detached from the Al matrix or extruded along with the plastically deformed Al matrix, as illustrated in Figure 10c,d, thereby failing to effectively protect the surface from wear attack. As a result, S-Al/SiC_p showed considerably larger volume loss and lower wear resistance than both S-Al₃Ti and S-Al₆₇Ti₂₅Fe₈.

4. Discussion

The D0₂₂ structure is proven to be the most stable structure of stoichiometric Al₃Ti alloy, while the L1₂ structure is a metastable structure which may be obtained under non-equilibrium processing condition, such as rapid solidification [42,43] and mechanical alloying [44,45]. The metastable L1₂ structure prepared by mechanical alloying may transform (through D0₂₃ structure) to stable D0₂₂ structure by heating processes [44,46]. First-principals calculations suggest that, in the D0₂₂ structure, Ti atoms gain charges and Al atoms lose charges, while in L1₂ structure, the situation is opposite [12]. As indicated earlier, adding an appropriate amount of less electropositive third elements such as Fe and Cr can replace Al, take electrons from Al, and consequently suppress the charge transfer from Al to Ti, thus stabilizing the cubic L1₂ structure and leading to the transformation of the D0₂₂ structure to the L1₂ one [22,23]. The resultant L1₂ structure keeps stable in an as-cast and homogenized state [16], indicating that adding proper third elements is an effective approach to obtain a stable L1₂ structure near stoichiometric Al₃Ti composition, leading to improved plasticity. The above points have also been confirmed by the present study on the modification of Al₃Ti with Fe addition for enhanced wear resistance.

For dry sliding wear of materials in the ambient environment, mechanical properties, such as strength and ductility that, along with hardness, determine the toughness of a material, are considered to be the most important factors which govern the wear resistance. The D0₂₂-to-L1₂ transformation from S-Al₃Ti to S-Al₆₇Ti₂₅Fe₈ improved the ductility (revealed by indentation marks in Figure 7) without losing much hardness (see Figure 6), leading to enhanced wear resistance, as shown in Figure 8. In addition to Fe, other elements such as Cr, Mn, Ni, Cu, and Ag can also stabilize L1₂ (Al, M)₃Ti due to the difference in electronegativity between the constituent atoms, and the energy difference between the D0₂₂ and L1₂ structures [22,47]. Although little research on the wear behavior of D0₂₂ Al₃Ti and L1₂ (Al, M)₃Ti is reported, the beneficial effects of these elements on the wear resistance of Al₃Ti can be expected based on the present study on the role of Fe addition in improving the wear resistance of the Al₃Ti alloy.

In this work, both S-Al₃Ti and S-Al₆₇Ti₂₅Fe₈ were subject to surface abrasion and oxidation to different extents. Although S-Al₃Ti is harder than S-Al₆₇Ti₂₅Fe₈, it showed lower wear resistance than the latter, ascribed to the fact that its higher brittleness and formation of a less protective oxide scale resulted in more wear damage, as described earlier [48,49]. It should also be mentioned that a brittle substrate may deteriorate the adherence of its oxide scale due to lowered flexibility that may elevate the interfacial stress if the chemical interaction and lattice mismatch between the oxide scale and substrate remain the same or similar, thus accelerating the oxide scale spallation and promoting the wear-oxidation synergy, leading to more wear for the brittle or less tough S-Al₃Ti, compared to that of the more plastic S-Al₆₇Ti₂₅Fe₈. In addition, the oxide scale spall-off mixed with wear debris could result in extra wear damage [41].

5. Conclusions

In this work, two Al₃Ti-based alloys, S-Al₃Ti containing ~97 vol.% of D0₂₂ Al₃Ti phase, and S-Al₆₇Ti₂₅Fe₈ containing ~93 vol.% of L1₂ (Al, Fe)₃Ti phase, were fabricated using an arc melting process, with the main objective being to investigate how the wear resistance of Al₃Ti was influenced by the transformation from the D0₂₂ to the L1₂ structure caused by alloying with a third element, Fe. Their wear resistances were evaluated and compared with a commercial Al-matrix composite reinforced by 30 wt.% SiC particles (S-Al/SiC_p) as a reference material. The following conclusions were drawn:

(a)

Alloying Al₃Ti with Fe, which may partially substitute Al, successfully transformed the D0₂₂ Al₃Ti to L1₂ (Al, Fe)₃Ti, leading to improved ductility or toughness at the expense of hardness;

(b)

S-Al₃Ti experienced oxide scale spallation and surface fragmentation due to its brittleness, resulting in higher wear volume loss;

(c)

S-Al₆₇Ti₂₅Fe₈ exhibited lowered abrasive wear due to a reasonable combination of hardness and ductility/toughness, thus increasing its wear resistance;

(d)

Both the S-Al₃Ti and S-Al₆₇Ti₂₅Fe₈ alloys exhibited higher wear resistances than the commercial S-Al/SiC_p composite. The (Al, M)₃Ti alloy series can be further optimized and is promising as lightweight materials for tribological applications.

The reported in this article is a preliminary study showing the potential anti-wear or tribological applications of lightweight Al₃Ti-based L1₂-structure material fabricated by arc melting. In the follow-up work, efforts will be made to further improve or optimize the base alloy by adding several third alloying elements with appropriate compositions to improve the toughness. In addition, potential combinations of L1₂-structure Al₃Ti-based lightweight material with other materials such as carbides as the reinforcing phases should be explored to develop composites with a superior strength–toughness combination for tribological applications under specific or extreme conditions.