Theoretical and experimental investigation of the W-Al-B and Mo-Al-B systems to approach bulk WAlB synthesis

Highlights

• Thermodynamic stability of MoAlB and WAlB compared to their most stable competing phases was evaluated by DFT.

• Solid solution (Mo,W)AlB MAB phase compounds were synthesised by solid-state powder metallurgy.

• Descriptions of the mechanisms involved in pressureless and hot-press reacted (Mo,W)AlB compounds are provided.

• (Mo,W)AlB compounds offer a pathway towards development of high W-content bulk MAB phase components.

Abstract

MAB phases are a family of ternary transition metal borides with a layered crystal structure, that provides them with properties likely to be suitable for applications within extreme environments. Here, we report a computational and experimental examination of the W-Al-B system compared to the isostructural Mo-Al-B system. Utilising DFT calculations, WAlB and MoAlB were found to be thermodynamically favourable compared to their most stable competing phases, with respective total energy differences of -0.15 eV and -0.28 eV at 0 K. Partial substitution of W on the Mo-site of MoAlB was observed for certain solid solution compositions. The experimental results indicate that synthesis of (Mo,W)AlB compounds is driven by in-situ formation of metal boride solid solutions, (Mo,W)B, which further react with Al, Mo-Al, W-Al or (Mo,W)-Al compounds to obtain the MAB phase structure. Finally, reactive hot-pressing was shown to be a promising avenue for the production of dense engineering (Mo,W)AlB-containing components.

Introduction

The nuclear, aerospace, mining and chemical processing sectors are examples of industries in which extreme environments are found, requiring high-performance materials, capable of operating under increasingly demanding conditions. Conditions encountered in such environments often generate combinations of damaging effects such as high temperature, oxidation, creep, high-energy radiation, chemical attack, mechanical damage and wear. Therefore, a combination of material properties associated with both metals and ceramics are essential to resist these environments.

Recently, a set of ternary metal borides known as MAB phases have attracted attention due to their combination of ceramic and metallic properties [1,2]. Their structure and properties have similarities to the well-established M_n+1AX_n (MAX) phases (M = transition metal, A = typically group IIIA-VIA element, X = C or N, n = 1, 2, 3) [3,4], a set of nano-laminated ternary carbides and nitrides. The MAX phases have already been shown to have useful properties for extreme applications and it is anticipated that the MAB phases could also find application in these areas. As with the MAX phases, extensive material property testing is required to identify the particular strengths and weaknesses of each MAB phase compound. Therefore, development of robust synthesis methods is necessary.

MAB phases have a nano-laminated crystal structure, composed of trigonal prismatic transition metal boride (M₆B) slabs intercalated by Al layers (A), producing orthorhombic space groups [1]. The set of metals (M) for which MAB phases have been observed is currently limited to Fe, Mn, Cr, Mo and W. The Ru-based MAB phases will not be considered in this work due to their structural differences compared with the aforementioned systems. As a result of their compositional and structural differences, the bond strength in MAB phases has been determined by first principle calculations to be typically stronger than the corresponding bonds in the hexagonal MAX phases [[5], [6], [7], [8]]. Physically, this manifests as differences in material properties, where some MAB phases are found to have approximately twice the compressive strength or hardness of MAX phases and exhibit a lower load dependence effect [[9], [10], [11], [12]].

Some of the properties of MAB phases, which are known to be important for extreme applications, have already been confirmed experimentally. These include: damage tolerance as seen by lack of dominant cracking from the corners of Vickers hardness indents on MoAlB and Fe₂AlB₂ [9,10]; formation of a passivating Al₂O₃ layer on the surface of MoAlB when heated in air [10,[13], [14], [15]]; crack-healing behaviour [14]; and the presence of damage tolerant mechanisms such as grain delamination, grain pull-out and crack deflection [16,17]. However, grain buckling, which has been observed for the MAX phases, has not been identified for MAB phases at this time. Of the currently tested bulk MAB phases, MoAlB has the highest potential operating temperature, with decomposition determined by DTA and TGA to occur at 1435 °C [18] promoting its consideration for high temperature applications. The radiation damage properties of MAX phases have been well-studied and indicate that they may be useful materials for critical nuclear reactor components [[19], [20], [21], [22], [23]]. The similarities between MAB and MAX phases therefore suggest that MAB phases may also possess radiation tolerance. Recently, the first radiation damage study of MAB phases (MoAlB and Fe₂AlB₂) [24], compared their performance to that of certain MAX phases as well as SiC, a benchmark material for nuclear applications. It was found that while MoAlB suffered severe radiation-induced amorphization, Fe₂AlB₂ had similar amorphization resistance to SiC and neither MoAlB or Fe₂AlB₂ succumbed to radiation-induced cracking which was observed in the MAX phases. This study indicated the potential use of MAB phases for certain nuclear fusion and 4th generation fission reactor applications. It is also of great interest for the nuclear fusion industry to develop and characterise the bulk synthesis of the WAlB MAB phase. This compound may provide useful benefits for applications within nuclear fusion reactors as a plasma facing material (PFM) due to the layered crystal structure acting as a defect sink much like the MAX phases, and the useful properties of tungsten allowing it to resist sputtering during plasma-wall interactions which reduce reactor efficiency [25,26]. However, the lack of a synthesis method for the bulk form of WAlB has meant that no such research has been possible.

While the synthesis of bulk MAB phase systems (M = Fe, Mo, Cr, Mn) has been reasonably studied [2], general analysis of the W-Al-B system is lacking. Rieger et al. assessed the W-Al-B system at 1000 °C and found no evidence of any ternary compound formation [27]. Since then, synthesis of WAlB has only been reported in the form of small single crystals within an aluminium flux, rather than in the bulk polycrystalline form [1,[28], [29], [30]]. Ade et al. stated that while WAlB is isostructural with the readily produced MoAlB phase, their formation mechanisms appear to be different [1]. Additionally, Ade et al. concluded that while Zhang et al. and Okada et al. had previously reported the synthesis of WAlB single crystals [[28], [29], [30]], attempts to repeat that work under the described conditions were unsuccessful and they were only able to produce a low yield of single crystals using different synthesis conditions [1]. One notable difference between the methods used by Zhang et al., Okada et al. and Ade et al. was the nature of the starting materials; Zhang et al. reacting pre-formed α-WB single crystals with Al powder [29] and both Okada et al. [28,30] and Ade et al. [1] using elemental W + Al + B powders, all with a large excess of Al. A second difference was the cooling rate used; Ade et al. claimed that extremely slow cooling rates of 2 °C/h from 1550 °C to 660 °C were essential for WAlB formation [1], however Zhang et al. claimed that spontaneous cooling within the furnace from 1500 °C resulted in pure WAlB formation while slower cooling rates of 48 °C/h from 1500 °C to 1000 °C followed by furnace cooling to room temperature lead to the formation of α-WB as a secondary phase [29] and Okada et al. found that a cooling rate of 50 °C/h from 1500 °C to 1000 °C before quenching to room temperature [28] or simply cooling from 1500 °C at 50 °C/h [30] resulted in optimal WAlB phase purity. Therefore, the exact methodologies for synthesis of WAlB are not well-defined and the fact that the only case of pure WAlB synthesis required spontaneous cooling [29] provides some evidence for the metastability of WAlB.

The reaction mechanisms of MoAlB have not been well studied, partially due to difficulties using standard analysis techniques. High-temperature in situ X-ray diffraction (XRD) is difficult to accurately perform because the pressureless solid-state reactions Mo + Al + B→MoAlB and α-MoB + Al→MoAlB result in expansion of the sample during synthesis. Further, in situ neutron diffraction would require the use of starting materials containing the B¹¹ isotope (the only alternative stable boron isotope) due to the large neutron capture cross section of B¹⁰ within standard boron sources. However, consideration of the Mo-B and WB phase diagrams [31,32] and the crystal structure of MoAlB and WAlB provide some evidence for the nature of such reactions. The M₆B slabs within the crystal structure of MoAlB and WAlB are both composed not of the corresponding stable room-temperature tetragonal metal borides (α-MoB & α-WB, space group I₄₁/amd) but instead are composed of the metastable high-temperature orthorhombic metal borides (β-MoB and β-WB, space group cmcm). According to the phase diagrams, the corresponding α↔β transition temperatures for these phases (depending on exact composition) are 1800–2180 °C for MoB and 2100–2170 °C for WB [31,32], while the reaction α-MoB + Al→MoAlB is commonly performed at 1000–1100 °C. This discrepancy could be explained by either 1) a multi-stage synthesis route involving the reactions of, and between, several intermediate compounds or 2) the stabilisation of β-MoB at low temperatures by the presence of Al. This latter phenomenon was first reported by Rieger et al. who showed that the β-MoB phase is stabilised by only a small amount of Al, forming the solid solution compound Mo_0.45B_0.5Al_0.05 (cmcm) [27]. The same Al-stabilising behaviour has not been reported for β-WB. However, Boller et al. showed that a hot-pressed WB sample, annealed and quenched from 1400 °C, retained a transitional crystal structure between α-WB and β-WB [33]. It is hypothesised that with the initial formation of β-MoB, the structure may be easily intercalated with Al to form MoAlB and the fact that the same stabilisation does not occur for β-WB may contribute to the synthesis difficulties of WAlB. While the mechanistic differences between the Mo-Al-B and W-Al-B systems are not currently understood, assessment of the thermodynamic stability of these phases is expected to provide some clarity.

Density functional theory (DFT) is a well-established computational method for modelling material systems based on their electronic and crystallographic structure. This allows for first principle calculations of many aspects of materials, including their material properties, bonding characteristics and thermodynamic properties including total energy of a reaction. The thermodynamic aspect of DFT is used to predict the stability of compounds, which is a powerful tool for assessing the value in experimentally researching new materials, or optimised compositions of solid solution phases, without the need for expensive trial and error-based laboratory work. This feature of DFT has been exploited for theoretical MAX phases, with some of these predictions leading to successful experimental synthesis of new compounds [[34], [35], [36], [37], [38], [39], [40], [41]]. However, while there have been an impressive number of DFT studies on the mechanical and thermal properties as well as bonding behaviour within the MAB phases in the past few years [[5], [6], [7], [8],18,[42], [43], [44], [45], [46], [47], [48], [49], [50], [51]], a much smaller amount of work has been conducted on assessing the stability of MAB phases by analysing the total energy or Gibbs free energy of the MAB phases or their competing compounds [16,18,47,48,51]. To the best of our knowledge, studies by Rajpoot et al. are currently the only reports, experimental or theoretical, of thermodynamic analyses of WAlB used for the purposes of calculating the unit cell volume [47,48]. However, those studies did not address the thermodynamic stability of WAlB compared with competing phases which must be investigated to help answer questions surrounding the bulk synthesis of this compound.

Due to the isostructural nature of WAlB with the readily synthesised MoAlB phase (space group cmcm), the experimental work reported in this paper explores the possibility of forming solid solutions between the two materials within the Mo_1-xW_xAlB (0 ≤ x ≤ 1) system for the temperature range of 1000–1600 °C. It is noted that a similar study was performed by Okada et al., who investigated the Mo_1-xW_xAlB (0 ≤ x ≤ 1) system using single crystal synthesis methods and found complete solubility between MoAlB and WAlB [28]. However, the use of an aluminium flux in the process of forming single crystals results in different reaction mechanisms to standard powder metallurgical processes.

In this paper, we report our DFT analysis which, for the first time, was used to assess the thermodynamic stability of WAlB and MoAlB with respect to their most stable competing phases. Additionally, this work outlines the conditions under which W forms a solid solution with MoAlB via pressureless solid-state synthesis and reactive hot-pressing as pathways to obtain bulk MAB phase materials with near-WAlB stoichiometry.