Epitaxial Sr(Sn, Ge)$_{x}$Ti$_{1-x}$O$_{3}$ buffer layers for continuous strain engineering on SrTiO$_{3}$ substrates

Epitaxial strain plays a key role in determining the structure and functionality of thin films, with the choice of substrate being traditionally used to control the magnitude of the applied strain. However, even in the large family of perovskite materials, this allows for only a limited, discrete set of strain states to be achieved. Here we report on an approach to controlling epitaxial strain for the growth of perovskite materials by involving a single SrTiO${3}$ substrate (the most available perovskite in single crystal form) and a buffer layer that consists of the solid solution Sr(Sn, Ge)${x}$Ti${1-x}$O${3}$, of which the lattice parameter can be tuned in a continuous fashion, from 3.880 Å up to 4.007 Å, while maintaining coherent epitaxial growth on SrTiO${3}$ with high quality interfaces. Using a BaTiO${3}$ overlayer as a model system, we show that changes to the buffer layer composition, i.e. increase of in-plane lattice parameter, change the strain state of BaTiO$_{3}$ from fully relaxed, through highly compressively strained, to an exotic state showing ‘inverted’ epitaxy in which the buffer layer is relaxed from the substrate but lattice matched to the overlayer.

💡 Research Summary

The authors present a novel strategy for continuous epitaxial strain engineering that overcomes the limitations of conventional substrate‑selection approaches. By depositing a solid‑solution buffer layer of Sr(Sn,Ge)ₓTi₁₋ₓO₃ (SSGTO) on a single SrTiO₃ (STO) substrate, they demonstrate that the in‑plane lattice parameter can be tuned continuously from 3.880 Å to 4.007 Å simply by varying the Sn or Ge content (x). Bulk powder X‑ray diffraction confirms a linear Vegard‑type dependence: Sn substitution expands the lattice (Sn⁴⁺ larger than Ti⁴⁺) while Ge substitution contracts it (Ge⁴⁺ smaller than Ti⁴⁺).

Pulsed laser deposition (PLD) was used to grow a series of SSGTO films on STO(001) under identical growth conditions, followed by a BaTiO₃ (BTO) overlayer of comparable thickness. Reflection high‑energy electron diffraction (RHEED) oscillations persisted throughout deposition, indicating layer‑by‑layer growth, and atomic‑force microscopy shows atomically flat surfaces with clear step‑terrace structures. For Sn concentrations above 75 % the growth mode transitions to step‑flow, suggesting that the large lattice mismatch between the buffer and the substrate allows the buffer to relax while still providing a good lattice match to the BTO film.

Temperature‑dependent 2θ–ω X‑ray scans reveal that the out‑of‑plane lattice parameter of BTO varies systematically with the buffer composition. When x ≥ 0.45, BTO remains fully strained to the SSGTO layer, displaying a reduced c‑axis; for x < 0.45 the BTO layer begins to relax, and its c‑axis expands. The Curie temperature (T_C) of BTO, extracted from the intersection of linear fits to the lattice‑parameter‑versus‑temperature curves, increases linearly with compressive strain, reaching values well above the bulk transition (130 °C) and exceeding 500 °C for the most highly strained samples. This confirms that the buffer layer can be used to fine‑tune functional properties that are highly strain‑sensitive.

Scanning transmission electron microscopy (STEM) combined with quantitative STEM‑fit analysis provides atomic‑scale maps of both in‑plane and out‑of‑plane lattice parameters across the heterostructure. In the Ge‑rich sample (x = 0.4) the SSGTO layer is fully coherent with the STO substrate, but the large mismatch to BTO generates interfacial defects and a relaxed BTO film. In contrast, the Sn‑rich sample (x = 0.85) shows a nearly perfect lattice match between BTO and the buffer, with only minor strain relaxation in the upper part of the BTO layer. The crossover composition around x ≈ 0.45 yields a “inverted epitaxy” situation: the buffer layer is slightly relaxed from the substrate yet its lattice constant is precisely intermediate between STO and BTO, allowing both interfaces to be coherently matched.

The work demonstrates that a single, commercially available STO substrate can serve as a universal platform for strain engineering across a wide range of perovskite oxides. By simply adjusting the Sn/Ge ratio in the SSGTO buffer, researchers can select any desired in‑plane lattice constant within a ~3 % window, eliminating the need for multiple substrates with discrete lattice parameters. Moreover, because STO can be epitaxially grown on Si(111), this approach opens a pathway toward integrating high‑quality oxide heterostructures with silicon‑based technology. The SSGTO buffer itself may also be functional (e.g., as a transparent conducting oxide), adding further versatility. In summary, the paper provides a practical, scalable method for continuous strain tuning, validates it with a well‑studied ferroelectric system (BaTiO₃), and outlines broad implications for oxide electronics, ferroelectrics, and other strain‑sensitive functional materials.