The Ti-Al intermetallics are light, high-temperature oxidation resistant materials and can find application in automotive, aerospace and power industry [1] . They could replace Ni superalloys in these applications [1] , as they have almost two times lower density. They are also more sufficient than pure titanium due to their higher elastic modulus and strength and better oxidation resistance at high temperatures [1] . These properties result from the dominant presence of the γ-TiAl phase. In such a way Ti-Al intermetallic is a promising group of light and high-temperature oxidation and creep resistant materials [1] . Moreover in order to improve these interesting properties, some alloying elements could be added. For instance, Nb and Ta addition improves the oxidation and creep resistance at 1000 K [2] [3] . In particular, Nb hinders dislocation slip resulting to the increase of creep resistance at high temperatures and, additionally, contributes to oxidation resistance by supporting the formation of an Al₂O₃ film, a kind of passivation [4] .

In spite of their interesting properties, these materials show insufficient hardness values, so that makes them to be inappropriate for applications towards wear resistance requirements [2] . In order to encounter this weakness, a lot of surface treatments have been developed, included diffusion heat treatments such as aluminizing [5] , nitriding [6] , siliconizing [7] and other techniques as well. Boriding (or boronizing) method is not applied extensively on Ti-Al intermetallics, in spite of this method is widely used in similar applications, even on titanium alloys [8] [9] . By this technique, a layer of metal borides is formed on the surface of the metal with diffusion of boron atoms through the metal lattice.

The most commonly used technique is conventional powder-pack boriding, which is characterized by its simplicity and cost-effectiveness [10] . It is often performed using a powder that contains the boron agent, an activator and a filler material at a high temperature, which causes the boron to be released from the agent by the activator. After the boron is released, it diffuses interstitially into the titanium aluminide lattice.

It is important to emphasize that there is not much data related to the boriding process especially on Ti-Al intermetallics. On this topic a rather new and sufficiently comprehensive work it is found at reference [11] . In our work, we investigated the role of the substrate modification, caused by the heat treatment, and the effect of the residual atmospheric gases in the structure and composition of the coating. Such an investigation on a borided Ti-Al intermetallic alloy, focused on the influence of the substrate, is totally absent from literature.

Figure 1 shows XRD pattern of the substrate material GE48-2-2 Titanium Aluminide before any treatment.

In order to separate overlapping peaks, we proceeded to an additional measurement with five times larger exposure time (5.0 sec) within a 2θ range of 36 - 48 degrees, which is shown in Figure 2. It is clear that, in the pre-treatment substrate, the dominant phase is the γ-Titanium Aluminide (γ-TiAl), but there is a presence of another equilibrium Ti-Al phase consisted of α₂-Ti₃Al. These preliminary measurements were carried out for two reasons: first, to detect the initial phases of the raw material and, second, to have a reference to evaluate the possible modification of the substrate after the boronizing treatment.

In Figure 3 the XRD spectrum of the substrate metal after 10 hours non-boriding annealing treatment at 1073 K, 1173 K and 1273 K is presented. The measurements, which were carried out at room temperature, showed that the substrate was modified and transformed into a pure γ-TiAl material, ensuring that during boriding the substrate consists of just one single phase.

In Figure 4 the XRD spectrum of borided layer is presented. In Figure 5 the pattern of limited 2θ range from 33 to 47 degrees for better analysis is presented. This layer was created with boronizing treatment at 1273 K for 10 hours.

Moreover, a variety of secondary phases was revealed such as TiAl₂ and Ti₅Al₁₁. During the boriding process, a diffusion mechanism seems to hold near the surface, leading to an alteration of the local equilibrium composition, which in turn results in formation of the mentioned above Al-rich phases, appeared to be formed between the TiB₂ needles. It is more probably that the Al-rich phases were formed because of the fact that a part of the Ti atoms was consumed for the formation mainly of TiB₂.

There is no evidence of TiB formation, in spite that elsewhere was referred to be observed [14] [15] [16] . Contrariwise, only TiB₂ was formed as it is clearly observable in Figure 4 and Figure 5. This is a very important detail that contributes to the evaluation of the quality of the coating, because the mechanical properties of TiB phase are inferior in comparison to TiB₂ [17] . The absence of TiB is probably due to the fact that the diffusion of boron was not facilitated through the coating layer. The Boron also seems to form a small amount of a

boride with Chromium, Cr₂B₃ (Figure 5). The Cr₂B₃ is a hard phase that enhances the quality of the coating [18] . Moreover, NbN (Figure 5) was probably formed, which contributes to the superficial strengthening of the material [19] .

Finally, there is a strong evidence of the formation of a hexagonal (graphite-like) Boron Nitride, h-BN (Figure 4). Normally, the BN is formed in the case that an NH₄Cl activator is used [10] . However, in our case is obvious that the presence of BN in the coating is caused by the atmospheric Nitrogen traces, as we did not use a Nitrogen containing activator. The substrate is not sufficiently approachable from the X-ray beam, so the γ-TiAl phase peaks on XRD pattern (Figure 4 & Figure 5) are not enough intense [20] [21] .

For additional aspect of the coating, a metallographic image is presented in Figure 6.

In Figure 7 a SEM image of the coating layer is presented.

The image (Figure 7) revealed a 10 - 15 μm thick TiB₂ layer. Moreover, the linear chemical profile across the coating layer for the elements Ti (light blue), Al (purple), Nb (red), Cr (green) is presented. The coating layer, as shown clearly in Figure 6, seems to consist of two phases. The atomic ratio Al:Ti in the coating, as extracted from elemental analysis via SEM (Figure 7), is significantly diminished. The majority of Ti atoms have being consumed in the formation of the TiB₂ phase. Thus, a large amount of Al atoms were deliberated from the initial γ-TiAl phase. Part of these Al atoms seemed to form the detected via XRD Al-rich phases TiAl₂ and Ti₅Al₁₁. The rest of the Al atoms have been moved towards the surface, following a reverse diffusion path. After that, an oxidation reaction of Al atoms took place and the oxide that was formed peeled off due to its weak adhesion with the coating layer.

From the above described results we deduce the following:

・ The boronizing treatment leads to the formation of a coating layer, that includes the desirable TiB₂ phase with sufficient hardness (up to HV = 1400).

・ The TiB phase is totally absent from the spectrum.

・ The substrate undergoes a modification during the boronizing process, being transformed into a pure γ-TiAl material.

・ The same modification is present during non-boriding annealing of the pure substrate at the same time and temperature conditions. Consequently, this modification is independent of the presence of Boron.

The authors acknowledge Mrs. Dabou Xanthi for carrying out the XRD measurements and Assoc. Prof. Pavlidou Eleni for conducting the SEM measurements. This research has been financially supported by General Secretariat for Research and Technology (GSRT) and the Hellenic Foundation for Research and Innovation (HFRI) (Scholarship Code: 530).

The authors declare no conflicts of interest regarding the publication of this paper.

Zagkliveris, D.I. and Triantafyllidis, G.K. (2018) Investigation of the Boriding Process of the Ti-Al Intermetallic Alloy GE48-2-2 Taking into Account the Probability of the Influence of the Substrate Modification. Materials Sciences and Applications, 9, 873-882. https://doi.org/10.4236/msa.2018.911063