Combinatorial Substrate Epitaxy and Computationally Guided Epitaxial Synthesis Investigations of the AEMnO3 (AE = Ca, Sr, and Ba) Polymorphs

The alkaline earth manganese oxides, AEMnO₃ (AE = Ca, Sr, and Ba), have attracted the attention of many scientific communities due to the rich coupling between their structural, electronic, and magnetic properties. Specifically of interest in this work is the study of their different structures and the manipulation of synthesis conditions to stabilize their metastable structures as thin films using pulsed laser deposition. Please note that for the rest of this document, AEMnO₃ will referred to as AMnO₃, where A = Ca, Sr, or Ba. As the A-cation size increases from Ca to Sr to Ba, it becomes more difficult to stabilize the cubic (3C) perovskite structure, and so a hexagonal structure forms instead. While CaMnO₃ and SrMnO₃ can both form in the 3C structure (it is the metastable phase for SrMnO₃), BaMnO₃ has only ever been stabilized in a variety of closely related hexagonal structures. The majority of this work, then, investigates the epitaxial polymorph stability of Ba_xSr_1−xMnO₃ (BSMO) (0 ≤ x ≤ 1) on various substrates and orientations, with the ultimate goal of pushing the boundary of phase stability closer to 3C BaMnO₃ for this system. Additionally, density functional theory (DFT) was used to model the epitaxial polymorph competition between the cubic and hexagonal structures of the (Sr/Ba)MnO₃ system, and unraveled the different thermodynamic contributions involved during nucleation. While 3C BaMnO₃ was never successfully stabilized, the conclusions of this work never imply that it is impossible. Careful control of the thermodynamics and kinetics during growth, along with appropriate substrate selection, may yet result in a 3C BaMnO₃ film, even if it is only a few nm thick. 

In the first study, combinatorial substrate epitaxy (CSE) was used to investigate the orientation relationships (ORs) between 60 nm (Ca/Sr)MnO₃ films and polycrystalline hexagonal (4H) SrMnO₃ and 3C SrTiO₃ substrates. Electron backscatter diffraction (EBSD) data was used to determine that there was only one general OR for all regardless of the film, substrate, orientation, or structure. This OR is called the eutactic OR because it aligns nearly close-packed (eutactic) planes and directions between the substrate and film. In the SrMnO₃ film on SrTiO₃, only a few grains were found to be stabilized in the metastable 3C structure while the rest of the film grains were stabilized in the stable 4H polymorph. The 3C grains were oriented near (100) on near (100) SrTiO₃ grains (a cube-on-cube alignment). This suggests that in these growth conditions (850 ^◦C, 2 mTorr pure O₂), penalties of higher interfacial energy and/or strain energies between polymorphic perovskites adopting the eutactic OR are not significant enough to overcome the volumetric formation energy of the stable 4H phase, except for very special orientations. 

The second study expands on the results for SrMnO₃ on polycrystalline SrTiO₃ from the first study. CSE was used to study the epitaxial phase competition between SrMnO₃ polymorphs as a function of substrate orientation, temperature, and oxygen pressure. EBSD data from the 60 nm films was analyzed using a dictionary-based indexing technique to determine that the eutactic OR was still the preferred OR for both polymorphs across all conditions and is a strong driver for epitaxy. An increase in substrate temperature (to 900 ^◦C) and then a decrease in oxygen pressure (using 2 mTorr 1% O₂/99% N₂) extended the 3C stability region on substrate orientations from (100), to (110), and finally, to (111), when the 3C was stable on all SrTiO₃ substrate orientations. In these conditions where the 3C is the stable phase, metastable 4H was then epitaxially stabilized on non-basal orientations of a 4H SrMnO₃ substrate. These results demonstrate how CSE can be used to quickly understand phase competition and prepare novel metastable films on complex oxide substrates. Additionally, CSE can be used to gauge the stability of the BSMO system with increasing Ba-content, since the substitution of Ba into SrMnO₃ increases the metastability of the 3C phase. The extent of 3C orientations on which metastable 3C films are stabilized indicates that a BSMO film with higher Ba-content is possible. To demonstrate this, a BSMO (x = 0.5) film was grown on polycrystalline SrTiO₃ using the same conditions and only (100) 3C BSMO was stabilized. Since the (100) is always the first orientation to stabilize the 3C phase, x = 0.5 is likely near the limit of Ba substitution under these conditions. 

After understanding how deposition conditions and substrate orientation affect 3C epitaxial stability in the BSMO system, we push the phase boundary in the third study to x = 0.6 for the first time. Using regular pulsed laser deposition (rPLD), 3C Ba_0.6Sr_0.4MnO₃ could be epitaxially vi stabilized on DyScO₃ (101)_o substrates at 900 ^◦C in 2 mTorr of 0.1% O₂/99.9% N₂. However, the 3C phase was mixed with the 4H polymorph, for films 24 nm thick and above, and the films were relatively rough. To improve flatness and phase purity, changes in growth kinetics were investigated and interval PLD (iPLD) was especially effective. In iPLD, deposition is interrupted after completion of ≈ a monolayer, and the deposit is annealed for a specific period of time before repeating. Both film flatness, in agreement with prior iPLD results, and, more importantly, the volume of the 3C polymorph improved with iPLD, resulting in flat, 40 nm, single-phase films. iPLD also leads to epitaxial stabilization of 3C Ba_0.6Sr_0.4MnO₃ on many orientations of polycrystalline GdScO₃, an indication of robust stability. The results imply iPLD improves persistent nucleation of metastable phases and that even more highly metastable films may be realized (e.g., higher Ba contents in BSMO). 

In the final study, DFT was used in the computation of thermodynamic terms relevant to the competition between epitaxial polymorphs during nucleation of (Sr/Ba)MnO₃ on (100), (110), and (111) cubic (Sr/Ba)TiO₃ substrates. Values for volumetric formation energies, volumetric strain energies, and area-specific interface energies were computed for different polytypes (3C, 4H, and 2H) and were incorporated in a standard (capillarity) model for epitaxial nucleation. Experimental ORs for SrMnO₃ were used in the construction of strained and interface cells for (Sr/Ba)MnO₃. For 3C polytypes, the OR is simply cube-on-cube, or (111)[110]_{3C, film}||(111)[110]_{3C, sub}, and is isostructural with cubic (Sr/Ba)TiO₃. For 4H/2H polytypes, the orientation relationship is (001)[100]_{4H/2H, film}||(111)[110]_{3C, sub}, which can only be modeled with a coherent interface on the (111) substrate. Results indicate that 3C SrMnO₃ has increased energies (becomes less stable) on moving from SrTiO₃ substrate orientations (100) to (110) to (111), consistent with experimental observations. For BaMnO₃, similar trends are predicted, although no experimental data is available for comparison. We use the DFT results to discuss the different thermodynamic contributions to polytype stability, and assess the feasibility of stabilizing a 3C BaMnO₃ film on (Sr/Ba)TiO₃ substrates.