- The paper demonstrates that aluminum-assisted thermal annealing efficiently induces a controllable PV-to-BM phase transition in LSMO films with reduced thermal budgets.
- Detailed XRD, XAS, and EELS analyses reveal significant lattice expansion, pronounced Mn loss, and cation redistribution impacting magnetic and transport properties.
- The findings indicate that tailoring oxygen vacancy profiles via getter-assisted processing can enhance integration in oxide electronics and related device applications.
Topotactic Phase Transitions in Epitaxial La₀.₇Sr₀.₃MnO₃₋δ Films via Oxygen Getter-Assisted Thermal Annealing
Introduction and Context
This study addresses the topotactic phase transition from the perovskite (PV, ABO3) to the brownmillerite (BM, ABO2.5) structure in La0.7Sr0.3MnO3−δ (LSMO), an archetype complex oxide with strong coupling between magnetic, electronic, and lattice degrees of freedom. Oxygen vacancy engineering in transition metal oxide thin films is central to tuning functional properties, as oxygen stoichiometry directly controls the electronic bandwidth, manganese valence, and consequently, magnetic and transport properties. Traditionally, triggering PV-to-BM transitions in LSMO demands high temperatures and long annealing times, limiting process flexibility and integrability, especially for device applications.
The reported work proposes and experimentally validates an efficient thermal annealing protocol utilizing aluminum as an in situ oxygen getter, achieving rapid and controllable PV-to-BM transitions at reduced thermal budgets. The approach's efficacy is systematically compared against the canonical annealing processes, with particular focus on lattice and cationic rearrangements, as well as the emergent changes to magnetic and transport properties, using a comprehensive suite of characterization methodologies.
Experimental Procedure and Structural Evolution
Epitaxial LSMO films were deposited on SrTiO3 (STO) substrates using both High Oxygen Pressure Sputtering and Pulsed Laser Deposition for optimized surface quality. Al foil, with its native oxide mechanically removed, serves as a robust oxygen getter when exposed to LSMO under high vacuum and elevated temperatures. This configuration efficiently lowers the oxygen chemical potential, significantly promoting the formation and migration of oxygen vacancies into the LSMO.
X-ray diffraction data unambiguously identify the characteristic signatures of the phase transition:
- As-grown PV-LSMO exhibits a (002) Bragg reflection corresponding to c=3.853 Å.
- Annealing at 350℃ for 12 hours in the presence of Al results in a lattice expansion to c=3.934 Å (E-PV350), indicative of oxygen vacancy formation but no long-range BM order.
- Annealing at 400℃ for 12 hours leads to the emergence of BM superlattice reflections (BM400), denoting the formation of the BM structure with alternating MnO6 octahedral and MnO4 tetrahedral layers and a full topotactic transition.
Surface and subsurface probing via X-ray reflectivity and AFM reveals nontrivial thickness reduction and increased roughness upon transition, as well as strong hints of cation migration and segregation, particularly of Mn. The inefficacy of simple multilayer models in describing post-annealed reflectivity strongly suggests complexity beyond mere oxygen loss, implicating cation rearrangement and surface amorphization.
Cation Redistribution and Electronic/Magnetic Phase Evolution
Comprehensive analysis of cation profiles by Rutherford Backscattering Spectroscopy establishes sizable Mn loss: up to 43.3% reduction in the BM400 structure compared with pristine PV-LSMO, with STEM-EDS mapping confirming segregation and surface depletion of manganese. This results in the formation of distinct layers: an amorphous, La/Mn/O-deficient topmost region; a segregated Mn surface region; and a crystalline BM-LSMO layer above the substrate.
Crucially, depth-resolved EELS oxidation state mapping shows a significant gradient: surface and amorphous regions are dominated by Mn2+, while the crystalline BM phase is primarily Mn3+. This multi-valency reflects substantial oxygen vacancy densities and cation redistribution under the imposed thermodynamic gradients by the O2 getter.
SQUID magnetometry and resistivity measurements reveal strong property tuning in tandem with the topotactic restructuring:
- PV-LSMO displays canonical double-exchange-driven ferromagnetism (TC = 327 K) and a metallic-insulator transition.
- The oxygen-deficient E-PV350 phase exhibits antiferromagnetic characteristics with suppressed overall magnetization, consistent with charge localization and broken double-exchange pathways due to Mn2+/Mn3+ enrichment.
- After full transition to BM400, long-range magnetic order signatures vanish, and the system becomes highly insulating, as expected for a BM lattice with highly reduced, disordered cations.
High-resolution XAS corroborates these findings through detailed line shape and branching ratios: increased L3/L2 ratios at the Mn L-edge indicate high-spin, low valence Mn dominates the near-surface region post-annealing. Loss of fine structure in the O K-edge implies amorphization and/or loss of long-range crystallinity at the interface.
Comparative Efficiency and Broader Implications
The Al-assisted vacuum annealing not only reduces the required transition temperature but also enables PV-to-BM transitions in micron-sized bulk powder (550℃, 20 hours), a considerable improvement over conventional protocols typically restricted to much thinner films. This points to substantially enhanced oxygen chemical activity in the closed system, which may generalize to other perovskite oxide systems (e.g., La1−xSrxCoO3, SrFeO3−δ, etc.).
The observed cation redistribution and surface amorphization have major implications for devices where interfaces and surface reactivity are critical. For example, in catalysts or solid oxide fuel cell electrodes, such reconstructed surfaces may modulate ionic conductivity, catalytic activity, or passivation behaviors. From a practical perspective, the demonstrated protocol is compatible with wafer-scale or batch processing, enabling integration into complex oxide electronics or neuromorphic devices that exploit resistive switching or spintronic phenomena.
Theoretical Speculations and Future Developments
The strong coupling between oxygen migration, cation mobility, and emergent collective order highlights the necessity for mesoscale modeling that includes both anion and cation sublattice kinetics. The spatially inhomogeneous redistribution of cations observed here is likely to play a significant role in the long-term stability, fatigue resistance, and reversibility of functionality in oxygen-vacancy-engineered systems. Further, the multi-step nature of the magnetic transitions and the formation of amorphous/cation-deficient surface layers raise open questions about possible emergent phenomena (e.g., exchange bias, spin glassy states) when such engineered heterostructures are coupled to other functional layers.
The application range for this getter-driven technique encompasses topotactic transitions in other challenging oxide chemistries, including nickelates and cobaltites with more complex redox landscapes. Additionally, combination with in situ probes (e.g., environmental TEM, time-resolved synchrotron spectroscopy) could elucidate the mechanistic pathways of cation and anion migration, with further feedback to materials design.
Conclusion
This study establishes that aluminum-assisted thermal vacuum annealing substantially enhances the tunability and efficiency of topotactic PV-to-BM transitions in epitaxial LSMO thin films and bulk powders. The phase change is accompanied by a transition from ferromagnetic metallic to antiferromagnetic (or magnetically undefined) insulating states, coupled with pronounced cation segregation and surface amorphization. These findings not only broaden the methodology for phase control in oxide electronics but also underscore the need for advanced characterization and theoretical description of coupled anion-cation redox and morphological evolution in out-of-equilibrium oxide systems.