Molten salt shielded synthesis of the nanolaminated transition metal boride Fe2AlB2

Fe2AlB2 is one of many ternary transition metal borides called MAB phases, which attracted interest owing to its magnetocaloric effect and magnetic properties. Herein, the molten salt-shielded synthesis (MS3) of Fe2AlB2 powder is studied using potassium bromide “KBr” in an open-air atmosphere. The synthesis process of the MAB phase, from Fe, Al, and B elemental powders, was studied with and without Sn additive in the temperature range of 900–1200 °C. The obtained powders were analyzed using XRD, TDA, GTA, and SEM analysis. The corresponding results revealed a successful synthesis of nearly pure Fe2AlB2 phase at 1000 °C for only one hour of holding time, beyond which FeB impurities form. These results show the efficiency of KBr (with Sn as a synthesis aid “additive”) in increasing the reactivity of this MAB phase in terms of the lowest synthesis temperature and time reported to date.

So far, single-phase Fe 2 AlB 2 is hardly achievable compared to other MAB phases. Its synthesis needs a longer sintering time to get highly pure samples. Lewis et al. [21] synthesized Fe 2 AlB 2 by arc melting at 1000 °C for 14 days of holding time. Similarly, using the same synthesis method, Ten et al. [23] successfully synthesized a nearly pure Fe 2 AlB 2 phase in only 7 days. The excess of Al in their starting powder mixture (Al/B/Fe = 3/2/2, in a molar ratio) lowered both synthesis time and temperature from 1000 to 900 °C. Yet, some Al 13 Fe 4 impurities were still present in the final powder.
For rapid synthesis of Fe 2 AlB 2 , some reports showed that hot pressing (HP) is more effective than arc melting. Li et al. [24] prepared bulk Fe 2 AlB 2 using HP for a 30-min sintering time at 1200 °C. Moreover, Liu et al. [25] used pressureless sintering in the temperature range of 1000-1200 °C for only 1 h. Both previous sintering processes showed less sintering time or temperatures. However, an extra grinding step is required to get powders from bulk samples. Additionally, employing an inert gas is indispensable to protect oxidation-prone materials during synthesis, which leads to a high production cost.
To overcome these manufacturing issues, the molten salt methods are the most adequate for ceramic powder synthesis at low temperatures. Two different methods can be distinguished: molten salt-assisted method and molten salt-shielded synthesis (MS 3 ).
The first method includes mixing salt with the initial starting powders to enhance reactivity. Several MAX and MAB phases have been produced by this method. B. Wang et al. [26] synthesized V 2 AlC ternary carbide by the molten NaCl-assisted method. They showed that the ionic liquid environment when salt melts helps the formation of V 2 AlC. The synthesis temperature was lowered from 1400 °C without salt assistance to 1050 °C when using NaCl. Tin et al. [27] reported the successful synthesis of the Cr 2 AlC phase in a binary NaCl/KCl system at 1000 °C. Other ternary compounds such as Ti 3 SiC 2 [28] using NaCl molten salt and Cr 2 AlB 2 [29] using eutectic NaCl/KCl have been successfully synthesized using this method. In all reported cases, the presence of salts reduced either or both synthesis temperature and time. Therefore, highly pure powder products were formed. Even though the molten salt-assisted method increased the sinterability of several MAX and MAB phases, the use of inert protective gas is still mandatory.
The second method was recently discovered by Dash et al. [30]. They showed a better alternative to the protecting inert gas atmosphere by a molten salt-shielded synthesis (MS 3 ) process. Unlike the previously mentioned molten saltassisted method, where inert atmosphere is required. This process uses salt not only as a reaction medium but also to protect the materials from oxidation during processing in an open-air atmosphere. Cold-pressed salts can form a gas-tight encapsulation around the initial powders, inhibiting air entry. They were able to synthesize different ternary transition metal compounds (Ti 3 SiC 2 , Ti 2 AlN, TiC, and MoAlB) as well as pure Ti. Recently, M. Dey et al. [31] synthesized Fe 2 AlB 2 , Mn 2 AlB 2 , and MoAlB powders by molten salt-shielded synthesis with different salts (NaCl, KBr, and eutectic NaCl-KCl). Their experimental method differs from Dash et al.'s method [30]. The salts were used only as an airtight shield and were not mixed with the initial powders. Their study showed that MAB phases could be synthesized in a salt medium without using any shielding gas. Despite all these advances in reducing synthesis temperature and time, the production process is still energy intensive. All attempts to further reduce them, even with small variations, are welcome.
In this work, we present a possibility to reduce the synthesis time of Fe 2 AlB 2 using the MS 3 method in KBr molten salt with and without Sn additive at a 900-1200 °C temperature range. To study phase purity, synthesized powders were characterized using X-ray diffraction (XRD), scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS), thermo-differential analysis (TDA), and thermogravimetric analysis (TGA).

Experimental procedures
Fe 2 AlB 2 phase was synthesized using Fe (purity 99.5%, Bio-Chem), Al (purity 98%, BioChem), B (purity 95%, Alpha Aesar), and Sn (purity 98.9%, BioChem) with a molar ratio of Fe = 2, Al = 2, B = 2, and Sn = 0.2. Potassium bromide "KBr" (Honeywell, purity 99.9%) in a 1:1 weight ratio was added to the resulting mixture. Excess of Al is often used in the synthesis of MAB phases. It minimizes the content of byproducts and compensates its loss due to high-temperature evaporation [23,32]. The amount of Al was chosen based on Fig. 1 a Schematic view of the Fe 2 AlB 2 synthesis by the molten salt shielded synthesis "MS 3 " process using potassium bromide "KBr." b Schematic view of the molten salt-assisted method the previous work on molten salt-assisted synthesis of the Cr 2 AlB 2 MAB phase [29].
The starting powders were initially mixed using an agate mortar for 1 h, then mixed in a polyethylene jar containing zirconia balls for 12 h with a multidirectional mixer. The powders were uni-axially cold pressed in a stainless-steel die under 200 MPa pressure to form compacts of 16 mm diameter and 5 mm height. The compacts were further encapsulated with KBr salt in a larger steel die of 32 mm diameter. Shortly after, the sample emerged in a KBr salt bed inside a cylindrical alumina crucible.
The samples were heated in a resistance furnace (NABERTHERM B180) in an open-air atmosphere at a 900-1200 °C temperature range. The heating rate was 5 °C min −1 until reaching the synthesis temperature and held for 1 h. The samples were allowed to cool inside the furnace until room temperature. The encapsulation is essential to prevent the sample from oxidation during the salt premelting stage [30,33,34]. Once the salt bed melts, it protects the sample from oxidation by creating a liquid barrier between the ambient air and the sample [30].
To collect the Fe 2 AlB 2 powder, the alumina crucible was submerged in boiling distilled water inside a beaker. After salt dissolution, the remaining solution was stirred for 1 h. The powder was then filtered, washed five times with deionized water, and dried at 80 °C for 18 h. Figure 1 shows the main difference between the MS 3 process (Fig. 1a) and the molten salt-assisted method (Fig. 1b).
To study the phase purity, synthesized powders were characterized by XRD analysis using a Bruker Advance 8 diffractometer with α-Cu radiations. 2θ° ranged from 5 to 80° with a 0.02° step size and a step time of 1 s. SEM (Quanta 650), equipped with an EDS (Brucker X-Flash 6/10), was used for morphological and phase analysis. EDS analysis was performed using 10 kV voltage and 90 s acquisition time. The weight change and the phase transformations were determined by thermogravimetric analysis (TGA) and differential thermal analysis (DTA). The apparatus used was SETARAM Labsys EVO in the temperature range of 50-1200 °C under N 2 gas and using 10 °C/min heating rate. Figure 2 shows the XDA of the synthesized Fe 2 AlB 2 powder starting from Fe/Al/B (2/2/2, molar ratio). The temperature varied between 900 and 1200 °C with (Fig. 2a) and without (Fig. 2b) Sn addition. For the synthesized sample at 900 °C using 0.2 Sn, the pattern shows two major phases: FeB and FeAl 6 (Fig. 2a). Some peaks of Fe 2 AlB 2 and Sn are also present. At 1000 °C, the XRD revealed successful synthesis of predominantly single-phase Fe 2 AlB 2 powder in the Fe-Al-B system. Fe 2 AlB 2 is the predominant phase, accompanied by Sn. When rising temperatures up to 1100 °C, Fe 2 AlB 2 is still the major phase. However, some small peaks of FeB and FeAl 6 are present. Up to 1200 °C, even though Fe 2 AlB 2 was predominant, it is discernible that FeB and FeAl 6 peaks intensity increased. According to the above results, it can be concluded that highly pure Fe 2 AlB 2 powder is synthesized using Fe/Al/B/Sn mixture at 1000 °C.

Results and discussion
To confirm if Sn powder influenced Fe 2 AlB 2 formation, a mixture of 2Fe/2Al/2B was heated by keeping the same previous conditions (without using Sn additive). The formation of Fe 2 AlB 2 starts at 900 °C. However, the major phases are FeB and FeAl x , as shown in Fig. 2b. For the samples produced at 1000 and 1100 °C, a large quantity of unreacted FeB, Fe 2 B, and FeAl x remained in the samples with relatively higher Fe 2 AlB 2 content. Increasing temperature up to 1200 °C showed a continued consumption of FeB, FeAl x , Fig. 2 XRD patterns of a 2Fe/2Al/2B/0.2Sn mixture and b 2Fe/2Al/2B mixture synthesized by the MS 3 process using KBr salt at 900, 1000, 1100, and 1200 °C, for 1 h synthesis time and Fe 2 B, as shown by the decrease of the intensities of respective phases' peaks. The major phase, in this case, was Fe 2 AlB 2 , as shown by its relatively higher peak intensities.
Based on the aforementioned results, it is reasonable to conclude that even at 1200 °C, the reaction is still in progress. Knowing that the synthesis of single-phase Fe 2 AlB 2 requires a very long time to be achieved [31], according to the present XRD results, as shown in Fig. 2a, the sample synthesized by adding Sn powder was nearly single-phase Fe 2 AlB 2 . The powder was obtained at the lowest temperature and time reported to date. Recently, M. Dey et al. [31] reported that Fe 2 AlB 2 could be produced in a KBr salt bath at 1000 °C. Yet, it took 12 h synthesis time to be achieved in a single phase. It must be noted that in their report, the synthesis was done without salt assisting. The salt was used only to protect initial powders from oxidation and was not mixed with reactive powders.
It can reasonably be concluded that the Sn additive played a major role during Fe 2 AlB 2 formation, as seen by the increased reactivity of the starting powders. Our results are in contestation with the study done by Jie Lui et al. [25]. They studied the effect of Sn on the final purity of the Fe 2 AlB 2 phase using a mixture of 2Fe/xAl/2B/0.2 Sn, where (x = 1, 1.3, and 1.5). They concluded that the Sn additive could not effectively improve the purity of Fe 2 AlB 2 . Their conclusion was based only on one synthesis temperature (at 1150 °C for 1 h). Besides, they used a smaller amount of Al in the starting powders without salt assisting. Other conditions compared to Jie. Liu et al. [25] work might have influenced the Fe 2 AlB 2 formation: (1) the difference in initial powder particle size [35]; (2) the reaction medium (the KBr salt bath and its presence in the initial reactive powders); (3) the heating rate. We have shown in our XRD results  Fig. 2b) that up to 1100 °C, some peaks of FeB and FeAl x appear and increase with further increase in temperature. It probably means that Fe 2 AlB 2 started decomposing [36], which is likely the case in Jie Lui et al. study [25]. Hence, it is reasonable to conclude that Sn powder is an effective synthesis aid that accelerates the formation of Fe 2 AlB 2 using the MS 3 process. The Fe 2 AlB 2 powder produced using this method is less expensive than pressureless sintering using an argon atmosphere. It provides a higher-purity product in a shorter time.
To have a better understanding of the Fe 2 AlB 2 formation mechanism, TDA/TGA curves in the 200-1200 °C temperature range are shown in Fig. 3. For the Fe/Al/B/Sn mixture, there are two endothermic and three exothermic peaks, as shown in Fig. 3a. The first endothermic peak is observed at 230 °C, which corresponds to the melting point of Sn. While the second peak, around 738 °C, corresponds to the melting point of potassium bromide. Besides, the two exothermic peaks at 511 °C and 641 °C could be attributed to the formation of Fe-Al and Fe-B phases, respectively [37]. The last exothermic peak at 1170 °C could probably indicate the decomposition of the Fe 2 AlB 2 phase [6]. On the other hand, the synthesized sample without using Sn additive has one endothermic peak at 740 °C, which corresponds to the melting point of KBr. The two exothermic peaks at 613 and 727 °C might be assigned to Fe-Al and Fe-B temperature formation, respectively.
The synthesis reaction of Fe 2 AlB 2 starts by Al reacting with Fe to form the FeAl 6 phase, as shown in Eq. (1). This reaction took place below the melting temperature of Al. It might explain the absence of the endothermic peak of the Al melting point (667 °C) in the TDA/TGA curves (Fig. 3). Subsequently, Fe reacts with B to form the FeB phase (Eq. 2). Both phases were seen to react at lower temperatures than reported earlier [37]. This was probably triggered by the presence of both KBr salt and Sn additive in the reaction medium. Our results agree with a previously reported study on the Fe 2 AlB 2 formation mechanism [38]. The Fe 2 AlB 2 reaction path consists of the consumption of transient phases (FeB, Fe x Al y , and B) to form the Fe 2 AlB 2 phase. Even though the authors used FeB and Al powders as starting reactants, Al did not intercalate into the FeB structure.
(1) Fe + 6Al = FeAl6 Previously, it was shown that the KBr molten salt accelerates the reaction of several MAX and MAB phases [30]. According to the present TDA results, the formation of Fe-Al intermetallic and Fe-B in the sample with the Sn presence (Fig. 3a) did form at lower temperatures (511 and 641 °C, respectively) compared to the sample synthesized without Sn (613 and 727 °C, respectively). Hence, it is reasonable to conclude that the Sn additive is a good choice to lower the formation temperature of the Fe 2 AlB 2 compound since it decreases the formation temperature of its reactive intermediate phases (FeAl 6 and FeB).
TGA curves (Fig. 3c and d) show a decrease in mass of both samples (2Fe/2Al/2B/Sn and 2Fe/2Al/2B) by about 50% during heating above 740 °C, which is mainly caused by the evaporation of liquid KBr. After the loss of all KBr, the weight of both samples is stable [30].
SEM micrographs of the synthesized sample at 1200 °C without Sn addition are shown in Fig. 4. Fe 2 AlB 2 particles are randomly shaped, with traces of spindles identified by EDS as Fe x Al y O z (spinel oxides). Moreover, elemental mapping conducted on the same area (Fig. 5b-d) confirms the successful synthesis of the Fe 2 AlB 2 phase. Figure 5 shows SEM micrographs of the sample synthesized at 1000 °C with Sn addition. The synthesized powders have a sub-spherical shape with an average particle size of  Fig. 5a. Higher magnification of X area (Fig. 5b) shows the sub-spherical powders to be agglomerates of smaller particles bonded together by what resembles a resolidified phase. EDS analysis (spot B), conducted on the latter, revealed it to contain mostly Sn (Sn 95 at. %, Al 3 at. %, Fe 1.5 at. %, and B 0.5 at. %). Note that EDS is not effective in quantifying boron accurately [39,40]. Moreover, spot C (small particles) contains Fe, Al, and B. For a better presentation, EDS-elemental mapping was performed on the sample. It shows elements distribution of Fe, B, Al, and Sn in Fig. 5c-f, respectively. It gives a clearer view of Fe 2 AlB 2 particles surrounded by resolidified Sn. This agglomeration is due to the low solidification temperature of liquid Sn during the cooling process.
Elemental mapping of non-agglomerated free Fe 2 AlB 2 particles synthesized with Sn additive at 1000 °C is shown in Fig. 6a-e. The presence of another phase that contains only Al and B was detected, probably Al x B y intermetallic (spot E: Al 62 at. %, B 37 at. %, and Fe 0.6 at. %). This latter did not appear in the XRD diffractogram owing to its small content in the synthesized powder. This phase was not considered in this work and was not reported previously. Thus, more work is needed to show whether this phase plays a role in the reaction mechanism of Fe 2 AlB 2 formation.
It has been shown elsewhere that Sn is a good synthesis additive during the synthesis of Ti 3 AlC 2 MAX phase [41][42][43]. It reduced the synthesis temperature and increased the purity of the final products. The principal role of the Sn Fig. 6 a SEM micrographs of synthesized Fe 2 AlB 2 with Sn additive at 1000 °C. SEM-EDS elemental mapping of: b aluminum, c iron, d tin, and e boron additive is to inhibit the thermal explosion during exothermic reactions in the Ti-Al-C system. Thus, it decreases the amount of TiC impurities in the final Ti 3 AlC 2 powders [43,44]. Same as Ti-Al-C ternary system, milled Fe/Al/B powder's reactions are all exothermic [25,45]. Jie Liu et al. [25] expected that adding Sn additive into the starting powders could reduce local heat generation during these exothermic reactions. Based on our SEM micrographs and XRD results, as shown in Figs. 5 and 2a, respectively. It can be said that the Sn additive did not diffuse and formed a solid solution, as in the case of MAX phases in the Ti-Al-C system. Indeed, the Sn can dissolve into the crystal structure of Ti 3 AlC 2 and Ti 2 AlC, forming Ti 3 Al (1_x) Sn x C 2 and Ti 2 Al (1_y) Sn y C (x + y = 0.1) solid solutions [43,46]. In the Fe-Al-B system, Sn powder seems to fasten the reaction by lowering the formation temperature of reactive intermediate phases and does not interact with Fe 2 AlB 2 to form unwanted solid solutions. Besides, Sn could be dissolved using a proper chemical etching, which could be useful to purify the final Fe 2 AlB 2 powder.

Conclusion
In this work, the effect of Sn addition and synthesis temperature on the formation of Fe 2 AlB 2 compound was studied by the MS 3 method using KBr salt. The main conclusions are as follows: • High purity of ternary layered Fe 2 AlB 2 powder was successfully synthesized with a mixture of 2Fe/2Al/2B with Sn additive at 1000 °C. Above this temperature, Fe 2 AlB 2 decomposes to Fe x Al y intermetallics and iron borides. • The sample synthesized at 1200 °C without Sn addition also reveals the production of the Fe 2 AlB 2 phase. However, a considerable amount of impurities are still present. • Sn reduces the formation temperature of the intermediate FeB and Fe x Al y phases. Thus, Fe 2 AlB 2 forms in higher purity. • The formation mechanism of the Fe 2 AlB 2 compound starts with the reaction between Al and Fe to form FeAl 6 . At a higher temperature, the remaining iron will react with B to form FeB. The two phases will react and form Fe 2 AlB 2 .