The Nature of the Insulator-to-Metal transition of SrTiO3 crystals

The discovery of the resistive switching phenomenon in  multinary  oxides  (especially in ternary oxides with perovskite structure)  has raised fundamental questions concerning the nature of the reversible insulator-to-meta  (I/M) transition [1,2]. How can a band insulator (e.g. SrTiO3 or BaTiO3) be switched into a  metallic  conductor  under the influence of different stimuli such as the chemical potential gradient or an electrical current  flow. In our opinion, for the description of the nature of this transition a paradigm shift is needed in relation with the semiconducting-like concepts of electrical transport traditionally applied to this class of oxides. Instead of only considering a current flow through the whole bulk we should also accept the possibility of the creation of a network of highly conducting nano-filaments through the matrix of binary and ternary oxides. Reduction or electrodegradation of the crystals transform such a network into metallic nanowires and short-circuit the rest of the non-conducting (insulating part) of the crystal. Without accepting the special role of such filaments (due to the available extended defects in the crystal and their high concentration in the surface layer) in electrical transport or as a sink for oxygen vacancies it seems impossible to explain the I/M transition in, for instance,  SrTiO3. This is due to the fact that the actual concentration measured for defects and corresponding charge carriers, which are made responsible for the I/M transition, is 4 to 5 orders of magnitude lower than predicted by the Mott criterion [3]. For the apparently oxymoronic statement that BaTiO3 is a ferroelectric metal in the surface [4] we need this new composite-like interpretation with an insulating bulk and a network of conducting filaments [3], according to which the reversible I/M transition takes place only along extended defects (here, edge dislocations) while the rest of the material remains a perfect ferroelectric.  In my lecture, I will argue that using a nanosensitive method (e.g. local conductivity AFM (LCAFM), Kelvin probe, AFM- potentiometry, PFM, XPS) we can reveal the “hidden aspects" of the I/M transition on the atomic scale in both models of ternary oxides with perovskite structure.

1)    K.Szot et al, Nature Materials 5, 312, (2006)

2)    Waser et al, Advanced Materials 21, 2632 (2009)

3)     K.Szot et al,  Solid State Physics 65, Chapt.4, (2014)

4)    Y.Watanabe et al,  Physical Review Letters 86,  332 (2001)