Engineering and Technology
Revealing the room temperature superplasticity in bulk recrystallized molybdenum
W. Chen, X. Li, et al.
Discover how Wenshuai Chen and colleagues have transformed pure molybdenum into a room temperature superplastic material, overcoming brittleness and enabling exceptional deformability. Their innovative approach could reshape the fabrication of refractory metals for extreme applications.
~3 min • Beginner • English
Introduction
Molybdenum (Mo) is widely used in aerospace, electronics, energy and nuclear applications due to its high-temperature strength, creep resistance, and thermal/electrical conductivity. However, its room-temperature (RT) ductility is poor, limited by intrinsic brittleness and recrystallization-induced intergranular fracture driven largely by oxygen (O) segregation at grain boundaries. Conventional alloying strategies (solution softening, solution/second-phase strengthening) can improve properties but are less effective at very high temperatures (≥1500 °C) and in ultra-high vacuum where reactive alloying elements can form unstable oxides. This raises the question of whether tuning thermomechanical processing and grain boundary chemistry in pure Mo can overcome recrystallization brittleness and enable RT superplasticity in fully recrystallized bulk material. The study aims to engineer GB 'cleanness' (minimizing harmful O segregation), promote favorable texture (<110>//RD) and a high fraction of low-angle grain boundaries (LAGBs), and enhance intragranular dislocation storage/transmission to suppress intergranular fracture and achieve exceptional RT ductility.
Literature Review
Prior work shows pure Mo readily recrystallizes and grain coarsens under thermal/mechanical stimuli, degrading strength and ductility. Alloying approaches include solution softening (e.g., Re additions) to reduce Peierls barrier, solution strengthening via lattice distortion, and second-phase strengthening to pin dislocations and delay recrystallization (e.g., La2O3, ZrC, Al2O3 dispersoids). However, high-temperature deformation tends to be diffusion-controlled, limiting effectiveness, and some alloying strategies are incompatible with extreme high-temperature/ultra-high vacuum environments due to oxide instability. Grain boundary segregation of light elements (O, N, P) in Mo is known to embrittle GBs and promote intergranular fracture; cohesion-enhancing metallic dopants (e.g., Ni, Fe and others) can counteract this effect. Segregation levels typically exceed 1 at% O at GBs in recrystallized Mo in literature. Texture and GB character also influence deformation; <110> orientations are favorable for slip in BCC under tension, and LAGBs can allow dislocation transmission, whereas HAGBs often block dislocations and concentrate stress. The literature thus motivates GB chemistry and structure engineering as a route to ductility without relying on reactive alloying additions.
Methodology
Materials preparation: Pure Mo powders (≥99.9 wt%, Fisher size 3.0 ± 0.2 µm) were cold isostatically pressed at 200 MPa for 10 min into bars, sintered in H2 at 1900 °C for 6 h, turned to 50 mm diameter, and Y-type hot rolled at 1300 °C to hexagonal bars with diameters 18 mm (Sample A, 87% reduction) and 12 mm (Sample B, 94% reduction). Subsequent hydrogen annealing at 800–1900 °C (10 K/min heating, 120 min hold, furnace cool) produced fully recrystallized states above 1000 °C. Impurity contents (ppm) were measured: Sample A—C 20, O 32, N 11, Ni 4.18, Fe 7.31; Sample B—C 20, O 34, N 8, Ni 2.85, Fe 5.45; Mo matrix >99.9 wt%. Densities were near fully dense by Archimedes method.
Mechanical testing: Dog-bone tensile specimens (gauge length 25 mm) were wire-EDM cut and ground. RT quasi-static uniaxial tensile tests were performed at strain rate 3 × 10⁻² s⁻¹ on a WDW3100 universal tester; additional strain-rate tests included 3 × 10⁻³ s⁻¹. A high-resolution extensometer measured yield and ductility. At least two replicates per condition were tested. Metrics recorded: ultimate tensile strength, 0.2% proof strength, total elongation and reduction of area.
Microstructure characterization: Metallography and EBSD (FEI QUAN FEG 450 with EBSD, HKL Channel 5) were used to obtain texture components, GB distributions, grain size distributions and Schmid factor maps. EBSD step size 1 µm, area 1000 µm × 1000 µm, 15 kV. Fractography by FE-SEM (JSM-F100).
TEM/STEM: Thin foils were mechanically thinned to ~70 µm and twin-jet electro-polished (Struers TenuPol-5) at 238 K and 40 V in H2SO4:CH3OH = 130:870 mL. TEM at 200 kV (FEI Tecnai G2 F20) and 300 kV (FEI Titan G2 60–300); HAADF-STEM on probe-corrected FEI Titan G2 80–200 ChemiSTEM at 200 kV. Dislocation densities were determined by surface intersection counts (ρ = 2N/A).
Atom probe tomography (APT): LEAP 4000X Si in laser pulsing mode at 50 K, 200 pJ laser energy, 200 kHz pulse rate, 0.5% detection rate, vacuum ~3 × 10⁻⁷ mbar. Specimens were prepared by FIB annular milling with stepwise reduced voltage/current. Data analyzed with IVAS 3.8.4; GB segregation quantified via Gibbs interfacial excess.
First-principles calculations: DFT (VASP) with GGA-PBE, plane-wave cutoff 500 eV; structures relaxed to energy <1×10⁻⁴ eV and forces <0.01 eV/Å to evaluate effects of GB segregants (e.g., O, N, Ni, Fe) on GB cohesion.
Key Findings
- Processing and microstructure: Y-type hot rolling plus annealing produced fully recrystallized fine-grained pure Mo. Sample B (94% reduction) retained higher <110>//RD texture after recrystallization (74.8% in B-1200) versus A-1200 (51%). LAGB fractions were higher in B-1200 (19% total, including 7% at ~1.5°) than A-1200 (11% total, 2% at ~1.5°). Average grain sizes were comparable (A-1200: 21.7 µm; B-1200: 23.6 µm), with B-1200 having a larger fraction of ≤3 µm grains (6% vs 2%).
- Mechanical properties: As-rolled B achieved 720 MPa UTS and 61% total elongation (uniform ~20%), outperforming as-rolled A (680 MPa, 45%). After full recrystallization (>1000 °C), yield strengths remained ≳400 MPa and fracture strengths ≳510 MPa while ductility diverged: B-series showed RT superplasticity with total elongation >90% between 1000–1700 °C, peaking at 108.7% for B-1200; A-series remained <50% after >1100 °C anneals. B-1000 sustained ~60% uniform elongation at 3 × 10⁻³ s⁻¹. After 1900 °C anneal, both decreased in ductility owing to grain coarsening, yet B-1900 still reached ~50% elongation.
- Fractography: A-1200 fractured predominantly intergranularly with transgranular cleavage and micropores (<1 µm). B-1200 exhibited fibrous tearing in the center and finer cleavage facets at the perimeter, indicating suppressed intergranular cracking and enhanced plasticity.
- Dislocation mechanisms: Upon deformation, A-1200 developed dense short dislocations with tangling; B-1200 formed long straight screw dislocations that weave into ordered dislocation networks, subdividing grains and enhancing storage. At ε ~70%, B-1200 retained the network with dislocation cells; A-1200 showed severe tangling.
- GB-dislocation interactions: In A-1200, both LAGBs and HAGBs blocked dislocations, promoting pile-up and GB cracking. In B-1200, dislocation transmission was GB-angle dependent: easy transmission across LAGBs, increasing impedance with GB angle, and effective blocking at HAGBs; a critical misorientation of ~15° was identified for transmission.
- Schmid factor and texture: Both A-1200 and B-1200 had high average Schmid factors (A: 0.469; B: 0.459). Higher <110>//RD texture in B-1200 correlated with more uniform intragranular slip and better intergranular compatibility, favoring formation/motion of straight screw dislocations and coordinated deformation.
- Grain boundary chemistry (APT): A-1200 LAGB showed strong O segregation (1.22 at%, interfacial excess ~2.79 atoms/nm²). B-1200 LAGB showed ultralow O (~0.19 at%, similar to grain interior) with minor Ni (~0.11 at%). At HAGBs, O in B-1200 was ~0.55 at% vs 1.29 at% in A-1200; N segregated at HAGBs (A: 0.33 at%, B: 0.43 at%); Ni segregation at HAGBs was higher in A-1200 (0.18 at%) than B-1200 (0.06 at%). No significant C or P segregation observed. Overall O+N at HAGBs: A-1200 ~1.62 at% vs B-1200 ~0.98 at%; Ni+Fe: A-1200 ~0.22 at% vs B-1200 ~0.12 at%.
- Cohesion and deformation synergy: Reduced O segregation and trace Ni at GBs enhance GB cohesion and allow dislocation transmission across LAGBs, while high LAGB fraction and <110>//RD texture facilitate ordered dislocation networks and stress relaxation—together suppressing intergranular fracture and enabling RT superplasticity (max 108.7% elongation).
Discussion
The study addresses the central challenge of RT brittleness in recrystallized Mo by simultaneously engineering GB chemistry and crystallographic/GB structure through a scalable thermomechanical route. High <110>//RD texture and a greater fraction of LAGBs in Sample B promote uniform slip and allow dislocations to transmit across LAGBs, reducing pile-up. Ultralow O segregation at GBs, especially at LAGBs (near grain-interior levels), markedly improves GB cohesion. Trace Ni at GBs further assists cohesion and may decrease the Peierls barrier for screw dislocations, supporting the development of long straight dislocations that weave into networks, subdivide grains, and increase storage capacity. The GB-angle-dependent transmission (critical ~15°) provides a mechanism for local stress release during deformation, mitigating strain localization and delaying GB cracking. Collectively, these factors transform the deformation behavior from GB-controlled cracking to intragranular plasticity-controlled flow, producing RT superplasticity in fully recrystallized bulk Mo. The results suggest that GB 'cleanness' and controlled texture/LAGB architecture can be a general strategy for refractory metals where light-element segregation limits ductility.
Conclusion
A low-cost, scalable pathway—powder metallurgy, Y-type hot rolling, and annealing—produces fully recrystallized bulk Mo with stable fine grains and unprecedented RT superplasticity (up to 108.7% elongation). The superplastic behavior arises from the synergy of: (i) ultralow O segregation at GBs (especially LAGBs) with trace Ni co-segregation, enhancing GB cohesion and enabling dislocation transmission; (ii) a high proportion of <110>//RD texture and LAGBs that facilitate formation of ordered, weaving dislocation networks and coordinated intergranular deformation; and (iii) GB-angle-dependent dislocation transmission (critical ~15°) that relieves local stresses. This framework suppresses intergranular fracture that typically plagues recrystallized Mo. The approach provides a broadly applicable strategy for designing ductile, structurally stable refractory metals and alloys for harsh environments. Future work could optimize processing to further tailor GB chemistry/structure, extend to other refractory systems (e.g., W, Nb, Ta), quantify strain-rate sensitivity and cyclic behavior, and elucidate the atomistic origin of long straight screw dislocations.
Limitations
The origin of the observed long straight screw dislocations and their weaving networks remains unresolved. The study focuses on pure Mo with specific impurity levels and processing parameters; generality to other chemistries or processing routes requires validation. While a strain-rate point at 3 × 10⁻³ s⁻¹ is reported, a comprehensive rate/temperature dependence and high-temperature mechanical behavior are not fully mapped here. Residual N segregation persists at HAGBs, which may still affect GB cohesion. Occasional HAGB cracking events remain possible, indicating that complete suppression of intergranular fracture is not achieved in all microstructural regions.
Related Publications
Explore these studies to deepen your understanding of the subject.

