Intermetallic Compounds: TiAl, NiAl, and MoSi₂ for High-Temperature Structures

Intermetallic compounds occupy a unique position in the materials landscape: they offer property combinations — high melting point, good oxidation resistance, low density, and ordered-structure strengthening — that neither conventional metals nor ceramics alone can match. Their defining characteristic is long-range chemical ordering, in which constituent atoms occupy specific sublattice positions rather than distributing randomly as in a solid solution. This ordering is the source of both their exceptional high-temperature strength and their notorious room-temperature brittleness — the central engineering challenge that has defined four decades of intermetallic research. This article provides a rigorous technical treatment of the three most practically important structural intermetallic systems — gamma-TiAl, NiAl, and MoSi₂ — covering crystal structure, intrinsic deformation mechanisms, alloying strategies, processing routes, and the aerospace turbine engine applications that drive their development.

1. What Defines an Intermetallic Compound

An intermetallic compound is a stoichiometric or near-stoichiometric phase formed between two or more metallic elements in which the constituent atoms occupy specific, crystallographically distinct sublattice positions — in contrast to a random solid solution where solute atoms are distributed statistically across the host lattice sites. This long-range chemical order (LRO) is the defining feature and the origin of all the properties that make intermetallics both attractive and challenging as structural materials.

1.1 Long-Range Order and the Ordering Energy

The degree of long-range order is quantified by the Bragg-Williams order parameter S:

Bragg-Williams Long-Range Order Parameter:

  S = (p − x) / (1 − x)

  where:
    p = fraction of A-atoms on A-sublattice sites (correct occupancy)
    x = overall fraction of A-atoms in the alloy

  S = 1.0  → Perfect LRO (all atoms on correct sublattice sites)
  S = 0.0  → Complete disorder (random solid solution)
  0 < S < 1 → Partial order (decreases continuously through the
               order-disorder transition temperature Tᵇ)

Ordering energy per mole (mean-field approximation):
  ΔGᵓᵃᵈ = −z · Nᵃ · εᵓᵃ · S² / 2

  where:
    z     = number of nearest neighbours
    Nᵃ    = Avogadro's number
    εᵓᵃ  = ordering energy = −(Vᵃᵄ − (Vᵃᵃ + Vᵄᵄ)/2)
             positive ε → A-B pairs energetically favoured → ordering
             negative ε → A-A or B-B pairs preferred → phase separation

For TiAl, NiAl, and MoSi₂, the ordering energy ε is large and positive — A-B nearest-neighbour pairs are strongly preferred — giving high order-disorder transition temperatures (above the melting point in all three cases, meaning these phases are fully ordered at all solid-state temperatures).

1.2 Antiphase Boundaries and Dislocation Physics

The consequence of long-range order for dislocation glide is profound. In a disordered FCC or BCC metal, a dislocation with Burgers vector b = ½⟨110⟩ (for FCC) can glide freely on {111} planes — the lattice looks identical on both sides of the glide plane. In an ordered structure like L1₀ TiAl, the same dislocation with b = ½⟨110⟩ leaves an antiphase boundary (APB) in its wake — a planar fault where the ordering is locally reversed and like atoms become nearest neighbours. The energy of this APB (γ_APB, typically 0.2–0.8 J/m² for intermetallics) is a penalty that must be paid for dislocation glide.

The result is that ordinary dislocations in intermetallics are replaced by superdislocations — pairs of ordinary dislocations coupled by the APB ribbon between them. The pair glides together as a unit, with the leading dislocation creating the APB and the trailing dislocation restoring order. Superdislocations are wider, heavier, and much less mobile than ordinary dislocations — particularly at low temperatures where cross-slip and thermally activated processes are frozen out. This is the atomistic origin of intermetallic brittleness: the stress required to move a superdislocation exceeds the cohesive strength of grain boundaries before significant plastic deformation can occur.

2. Gamma-TiAl (γ-TiAl) — The Commercial Intermetallic

The titanium-aluminium system contains several intermetallic phases, but two dominate structural applications: α₂-Ti₃Al (D0₁₉ hexagonal) and γ-TiAl (L1₀ tetragonal). Commercial alloys based on γ-TiAl contain approximately 44–52 at% Al and invariably include ternary and quaternary additions to improve properties. The γ-TiAl phase occupies the composition range ~49–65 at% Al at room temperature, with the two-phase (γ + α₂) field used in engineering alloys to obtain lamellar microstructures.

2.1 Crystal Structure and Deformation Mechanisms

The L1₀ structure of γ-TiAl is a slightly tetragonally distorted FCC lattice (c/a = 1.02, nearly cubic) in which alternating {001} planes are occupied entirely by Ti atoms or entirely by Al atoms. The small tetragonal distortion from the cubic ideal is significant — it removes the cubic symmetry and reduces the number of crystallographically equivalent slip systems.

Deformation in γ-TiAl occurs by four mechanisms, each contributing differently to overall plasticity:

1. Ordinary dislocations with b = ½⟨110⟩ on {111} planes — leave APB; glide as pairs. Provides basal-plane plasticity; requires cross-slip to contribute independent slip systems.
2. Superdislocations with b = ⟨101⟩ on {111} and {010} planes — restore order, no APB in wake, but very high Peierls stress on {010} planes at low temperature. Contributes fewer independent systems than needed for polycrystalline ductility.
3. Mechanical twinning on {111}⟨11¯2⟩ — critical supplementary deformation mode; twinning does not require the 5 independent slip systems of the von Mises criterion and can accommodate significant strain. Twinning tendency is promoted by lower Al content and by Cr additions (which lower stacking-fault energy).
4. Ordinary dislocations with b = ½⟨11¯2⟩ on {111} planes — relatively rare; important in specific grain orientations.

Von Mises Criterion for Polycrystalline Ductility:

  Nᵄᵇᵈ ≥ 5 independent slip systems required for general plastic deformation
  (5 components of strain tensor, minus hydrostatic component = 5 independent)

  γ-TiAl (L1₀) slip system count:
    ½‹110›{111} ordinary dislocations:  4 independent systems
    ‹101›{111} superdislocations:         4 independent systems (overlapping directions)
    Combined unique independent systems: ~ 4 (below von Mises criterion of 5)
    → Must be supplemented by mechanical twinning to approach ductility

  Consequence: Room-temperature elongation typically 0.5–3% in best alloys
               vs. 20–30% for FCC metals with 12 slip systems

2.2 Intrinsic Brittleness — Three Contributing Mechanisms

The room-temperature brittleness of γ-TiAl (fracture toughness K_IC = 10–20 MPa√m in fully lamellar alloys; 12–17 MPa√m in duplex; elongation 0.5–3%) has three independent contributing mechanisms that must be addressed separately:

Mechanism 1 — Limited independent slip systems: As established above, the von Mises criterion is barely met by combining ordinary dislocation glide and twinning. Any local grain orientation where twinning is geometrically disfavoured will deform only by slip, with fewer than 5 independent systems — leading to strain incompatibility at grain boundaries and intergranular fracture.

Mechanism 2 — High APB energy on {010} planes: The APB energy on {010} planes in γ-TiAl is approximately 0.5–0.7 J/m² — significantly higher than for {111} planes (~0.3 J/m²). This makes superdislocation glide on {010} planes (which is needed for certain grain orientations) energetically expensive and thermally activated at temperatures below ~600°C. Below 600°C, thermally activated processes are too slow and the lattice resists these superdislocations — contributing to the characteristic sharp drop in ductility at room temperature.

Mechanism 3 — Environmental (moisture) embrittlement: γ-TiAl alloys tested in air at room temperature show significantly lower ductility than in vacuum or dry argon. The embrittlement originates from the reaction of exposed Al-rich surfaces (at grain boundaries, fresh crack surfaces) with atmospheric moisture: 2Al + 3H₂O → Al₂O₃ + 6H_ads. Atomic hydrogen absorbed into the metal lattice diffuses to grain boundary triple points and high-stress regions ahead of crack tips, reducing the grain boundary cohesive energy and promoting intergranular fracture. This is directly analogous to the hydrogen-assisted cracking mechanism in high-strength steels described in our hydrogen-induced cracking article, and explains why TiAl components must be handled carefully to avoid surface contamination.

2.3 Engineering Alloy Compositions and Microstructure

2.4 The Role of Alloying Additions

Gamma-TiAl alloy design has progressed through four “generations” since the 1970s, each targeting a specific limitation of the previous generation through systematic alloying:

3. Processing Routes for Gamma-TiAl Components

The inherent brittleness of γ-TiAl at room temperature severely constrains conventional processing. Forging, rolling, and extrusion must occur at temperatures within specific phase field windows where adequate plasticity exists (typically 1100–1300°C), and machining of the finished component is extremely difficult due to low ductility and high work hardening. The principal production routes each make different trade-offs between microstructural quality, geometric capability, cost, and yield.

3.1 Investment Casting

Investment casting is the commercially dominant route, used by GE Aviation for GEnx and LEAP engine low-pressure turbine (LPT) blades. TiAl alloy is vacuum induction melted (VIM) to prevent oxidation contamination, then cast into preheated ceramic investment shells (typically yttria-coated to prevent Al-reduction of silica-based shells). The casting is allowed to solidify, then hot isostatic pressed (HIP) at approximately 1200°C and 175 MPa for 2–4 hours to close solidification shrinkage porosity. Final heat treatment at 1000–1250°C controls the γ/α₂ lamellar colony size and volume fraction.

3.2 Powder Metallurgy and Spark Plasma Sintering

Pre-alloyed γ-TiAl powder is produced by electrode induction gas atomisation (EIGA — a contactless, skull-melting technique that avoids ceramic contamination) or plasma rotating electrode process (PREP). The resulting spherical powder (15–150 μm diameter) is consolidated by:

• Spark plasma sintering (SPS): Powder loaded into graphite die, heated by pulsed DC current at 50–100°C/min to 1200–1300°C under 50–100 MPa uniaxial pressure; full density in 5–15 minutes. SPS produces fine, equiaxed microstructures (grain size 5–20 μm) with excellent property uniformity. Primary limitation is die geometry restricts near-net shape capability.
• Hot isostatic pressing (HIP) of powder: Powder in sealed metal canister pressed at 1200°C/175 MPa for 4 hours; produces fully dense billets for subsequent machining or isothermal forging.

3.3 Additive Manufacturing — Electron Beam Powder Bed Fusion

Laser-based powder bed fusion (L-PBF) is problematic for γ-TiAl because the high thermal gradients produce residual tensile stresses that exceed the alloy’s fracture toughness, causing cracking during or immediately after deposition. Electron beam powder bed fusion (EB-PBF, commercially: Arcam / GE Additive) resolves this by operating with the entire powder bed preheated to 900–1000°C — within the ductile regime of TiAl — reducing thermal gradients and residual stress below the cracking threshold.

EB-PBF of TiAl produces a lamellar microstructure (because the slow solidification from a 900°C preheated bed mimics furnace cooling through the γ+α₂ field) with properties approaching those of investment casting in most respects, but with significantly greater geometric freedom — overhanging features and complex internal cooling channels achievable by casting only with ceramic cores. The technology is being evaluated for second-generation LEAP/GEnx successor engine components.

4. NiAl — High Melting Point, High Conductivity, Unresolved Brittleness

NiAl is a B2-structured (CsCl-type) intermetallic with nominal composition 50 at% Ni, 50 at% Al. Its phase field in the Ni-Al binary is surprisingly wide — from ~45 to ~59 at% Al at room temperature — because the B2 structure tolerates significant antisite disorder (atoms on wrong sublattice) without losing its cubic symmetry or its metallic character. This wide homogeneity range is exploited in alloy design: compositions on the Ni-rich side have better ductility but lower oxidation resistance; Al-rich compositions have better Al₂O₃ scale formation.

4.1 Exceptional Properties and Fundamental Limitations

NiAl Key Properties Compared with Ni Superalloy (IN738LC):

  Property              NiAl (B2)           IN738LC (Ni superalloy)
  ─────────────────────────────────────────────────────────────────
  Melting point (°C)   1638                1315 (solidus)
  Density (g/cm³)      5.9                 8.11
  E (GPa)               190–270            200–215
  κ (W/m·K)           75–80             10–14
  CTE (10⁻⁶/°C)         13.5               12.6
  YS RT (MPa)           150–350          900–1000
  Elongation RT (%)     0 (polycrystal)     3–5
  Kᵢᵇ RT (MPa√m)     4–6 (polycrystal)  15–30
  Oxidation limit (°C)  ~1400               ~1050–1100
  DBTT (°C)            ~300–400          No DBTT (FCC)
  ─────────────────────────────────────────────────────────────────
  The thermal conductivity advantage (5–8×) means NiAl blades could use
  thinner thermal barrier coatings or fewer cooling holes for the same
  metal temperature — potentially enabling higher turbine inlet temperatures
  with lower cooling air consumption.

The fracture toughness of polycrystalline NiAl at room temperature is only 4–6 MPa√m — below the practical minimum for structural aerospace components (~15 MPa√m). Single-crystal NiAl shows orientation-dependent toughness: the ⟨100⟩ orientation has K_IC ~15 MPa√m (acceptable), while ⟨110⟩ is ~5 MPa√m (brittle). This extreme anisotropy and the lack of polycrystalline toughness are why NiAl remains a research material for primary turbine blade service despite its extraordinary property profile.

4.2 Toughening Strategies for NiAl

Four toughening strategies have been investigated, with limited but instructive success:

• Ductile phase reinforcement: Incorporating a second ductile metallic phase (Mo, W, Re) as fibres or particles in an NiAl matrix. Mo fibres in NiAl increase fracture toughness to 10–14 MPa√m at room temperature by crack bridging. The primary limitation is density increase from the refractory reinforcement and the difficulty of processing to achieve uniform distribution.
• Macroalloying with Cr and Fe: Adding 5–10 at% Cr creates a dual-phase microstructure (B2-NiAl + BCC-Cr precipitates). Cr precipitation provides ductile phase toughening and somewhat improves room-temperature elongation (to ~1–2%). The addition also improves oxidation resistance by providing Cr₂O₃ in the oxide scale.
• Single-crystal growth: Directional solidification in the ⟨100⟩ orientation eliminates grain boundaries (the weak link in polycrystalline NiAl) and exploits the highest-toughness crystal orientation. Toughness reaches 15–20 MPa√m ⟨100⟩ — adequate — but single-crystal NiAl blade casting is technically much harder than for Ni superalloys, and thermal fatigue performance is inadequate without significant further alloying.
• Precipitation strengthening: Heusler-ordered (L2₁) Ni₂AlTi precipitates in NiAl-Ti alloys provide high-temperature strengthening analogous to γ’ in nickel superalloys. The Ni₂AlTi phase has a coherent interface with B2-NiAl and produces a positive strength anomaly (strength increasing with temperature up to ~700°C) that is highly desirable for turbine blade service.

5. MoSi₂ — Refractory Intermetallic for Ultra-High Temperature

Molybdenum disilicide (MoSi₂) occupies the extreme end of the intermetallic temperature spectrum. Its melting point of 2030°C and the formation of a continuously healing, crystalline SiO₂-rich scale in air provide oxidation protection to approximately 1700°C — far beyond the capability of any nickel superalloy or TiAl alloy. These properties make MoSi₂ and its composites the primary candidate for the transition from metallic to ceramic-based high-temperature systems at temperatures above 1200°C.

5.1 The Pest Phenomenon

5.2 MoSi₂ Heating Elements and Structural Research

The commercially dominant application of MoSi₂ is as a resistance heating element for high-temperature furnaces (marketed as Kanthal Super, Globar-type, or equivalent by multiple manufacturers). Operating in air at temperatures up to 1850°C, MoSi₂ elements provide resistive heating with high power density and good long-term stability. The commercial product is typically sintered MoSi₂ + 1–2 wt% SiO₂ binder in the form of U-shaped or straight rods.

For structural applications at 1200–1600°C, MoSi₂-based composites are studied: SiC-reinforced MoSi₂ matrix composites improve fracture toughness (from the intrinsic ~3–5 MPa√m) to ~6–10 MPa√m through crack deflection and SiC whisker bridging, while maintaining high oxidation resistance. Mo₅Si₃-MoSi₂ two-phase alloys provide better creep resistance than monolithic MoSi₂ above 1200°C. These composite systems are under evaluation for next-generation hypersonic vehicle leading edges, combustion chamber liners, and gas turbine tip shrouds operating in the 1200–1500°C range.

6. Comparative Properties and Application Map

7. Aerospace Turbine Engine Applications

The primary driver for γ-TiAl development has always been the turbine engine low-pressure turbine (LPT) — the last turbine stage before the exhaust nozzle, operating at temperatures of 650–850°C, where nickel superalloys are over-engineered for temperature capability but their density (8.2 g/cm³) imposes significant centrifugal loading on the turbine disc. The half-density of γ-TiAl (3.8 g/cm³) allows either lighter blades for the same disc, or a smaller disc (and thus a shorter, lighter engine) for the same blade mass. The cascading weight savings from lighter rotating components propagate through the entire structural loading chain of the engine.

7.1 Commercial Milestones

GE Aviation introduced γ-TiAl LPT blades (6th-stage, GEnx-2B engine for Boeing 747-8) into commercial passenger aircraft service in 2011 — the first structural use of an intermetallic compound in a commercial turbine engine. The alloy used was Ti-48Al-2Cr-2Nb (at%) with 0.2 at% B for grain refinement, investment cast, HIPped, and machined. Each blade saved approximately 45% mass versus the equivalent IN718 blade, translating to ~200 kg total engine weight reduction on the GEnx-2B.

The LEAP engine (CFM International, powering Boeing 737 MAX and Airbus A320neo) followed with γ-TiAl LPT blades in the 4th and 5th stages, further extending the application temperature range and validating the technology across a second engine architecture. Airbus A320neo aircraft powered by LEAP-1A engines have accumulated hundreds of millions of operating hours with γ-TiAl blades.

7.2 Future Engine Integration — Higher Temperatures

Next-generation open fan (CFM RISE program) and ultra-high bypass ratio turbofan concepts push LPT operating temperatures to 900–950°C, above the current capability of standard Ti-48-2-2 alloys. The high-Nb TNM alloy and 4th-generation compositions with optimised Nb, Mo, W, and C additions are the primary candidates for these higher-temperature stages, alongside TBC-coated γ-TiAl (applying a 100–200 μm YSZ thermal barrier coating to the TiAl substrate). The connection to the broader superalloy ecosystem — understanding where γ-TiAl fits relative to conventional nickel superalloys — is covered in our related articles on HAZ microstructure in nickel alloy welds and martensite formation context for understanding ordered-phase stability.

Recommended References