Tuning Oxygen Vacancy Diffusion through Strain in SrTiO₃ Thin Films

Abstract

Understanding the diffusion of oxygen vacancies in oxides under different external stimuli is crucial for the design of ion-based electronic devices, improve catalytic performance, etc. In this manuscript, using an external electric field produced by an AFM tip, we obtain the room-temperature diffusion coefficient of oxygen-vacancies in thin-films of SrTiO 3 under compressive/tensile epitaxial strain. Tensile strain produces a substantial increase of the diffusion coefficient, facilitating the mobility of vacancies through the film. Additionally, the effect of tip bias, pulse time, and temperature on the local concentration of vacancies is investigated. These are important parameters of control in the production and stabilization of non-volatile states in ion-based memresistive devices. Our findings show the key role played by strain for the control of oxygen vacancy migration in thin-film oxides.1

Room‐Temperature AFM Electric‐Field‐Induced Topotactic Transformation between Perovskite and Brownmillerite SrFeOx with Sub‐Micrometer Spatial Resolution

“…(Figure 5 a). Although the threshold voltage changes slightly for the films deposited on different substrates, probably reflecting the effect of epitaxial strain in oxygen mobility (NdAlO3, LaAlO3 and SrTiO3 induce a strain on SrFeO3- of -2.5%, -1.8%, and +1.1% respectively), [23] the local expansion induced by the voltage-biased AFMtip is observed irrespective of the substrate used to grow the SrFeO3- PV film. Also, although SrTiO3 is very susceptible to the creation of oxygen vacancies, NdAlO3, LaAlO3 or MgO, are not.…”

“…Therefore, the charged nature of VÖ and their large mobility in the PV phase, should make them amenable to their control by a local electric field applied by a voltage-biased AFM tip, as it was shown in other 3d-oxides. [22][23][24][25] We hypothesize that the small free energy difference between the PV and BM phases in SrFeOx, [13] could lead to the spontaneous rearrangement of the vacancies accumulated by the electric field in a region of the PV film, leading to the local transformation into the BM phase. [26] The evolution of the room temperature Raman spectra for different voltages is shown in Figure 3 a).…”

Ab initio description of oxygen vacancies in epitaxially strained $$\hbox {SrTiO}_{{3}}$$ at finite temperatures

“…However, the nonmonotonic peak-like -dependence of the formation enthalpy that has been measured for biaxially strained STO thin films 34 cannot be satisfactorily explained in terms of such electrostatic arguments (as otherwise the formation energy of oxygen vacancies should increase, rather than decrease, under compressive conditions). Likewise, the partial disagreements between theory and experiments on the diffusion properties of oxygen vacancies 37 – 39 appear to suggest that some key elements might be missing in previous ab initio studies 41 .…”

“…This behaviour is different from zero-temperature first-principles results (also referred to as ab initio ) obtained for other prototype oxide perovskites like and , which indicate that compressive (tensile) biaxial stress typically depletes (promotes) the formation of 35 , 36 . Meanwhile, atomic force microscopy experiments have shown that tensile biaxial strain produces a substantial increase in the mobility of oxygen vacancies whereas small compressive biaxial strain produces an incipient diffusion reduction 37 . These latter experimental observations appear to be in partial disagreement with previous zero-temperature first-principles studies in which it has been concluded that both tensile and compressive biaxial strains tend to promote the migration of oxygen ions 38 , 39 .…”

Ferroelectricity-free lead halide perovskites

“…3a for the CsFAMA perovskite is likely coming from ion migration mediated by strain. 41,42 As a result, we observe: (i) a negative current when a force of 20 mN was initially applied to the sample, triggering ion movement; (ii) an increase in current close to zero as the force was gradually increased from 20 mN to 226 mN in 50 ms, since more ions were activated during this process; (iii) a decrease in current when the force was kept constant at 226 mN, due to a decrease in concentration of available local ions; (iv) a drop in current as the force decreased from 226 mN to 20 mN, leading to less driving force for ion movement and (v) a gradual increase to a plateau as the force was held at 20 mN (gradual movement of ions under a low force).…”