Absence of self-heated bistable resistivity in ${\text{La}}_{0.7}{\text{Sr}}_{0.3}{\text{MnO}}_3$ films up to high current densities

The $E\left(J\right)$ characteristics of well-characterized fully strained epitaxial ${\text{La}}_{0.7}{\text{Sr}}_{0.3}{\text{MnO}}_3$ films have been investigated up to high current densities $\left(J\sim {10}^4–{10}^5\text{A}/{\text{cm}}^2\right)$. Different electric contact resistances and different current path configurations have been used to sort out the role of Joule self-heating from contact heating in the transport properties. It is demonstrated through macroscopic transport and conductive scanning force microscopy measurements that contact heating may lead to nonlinear and irreversible $E\left(J\right)$ characteristics when high contact resistances are used. Low dissipative contact power measurements are crucial to define the upper $J$ limits to keep linear and reversible $E\left(J\right)$ characteristics. We demonstrate that Joule self-heating in ${\text{La}}_{0.7}{\text{Sr}}_{0.3}{\text{MnO}}_3$ thin films only induces moderate warming at high temperatures while achieving bistable resistivity at low temperatures would require the use of very high current densities $\left(J\sim 5\times {10}^5\text{A}/{\text{cm}}^2\right)$.

Figure 1

(Color online) Main panel: temperature dependence of the resistivity under 0 and 5 T magnetic fields (right axis) and the magnetization in a magnetic field of 5 mT (left axis). Inset: temperature dependence of the magnetoresistance at 5 T and the magnetic moment derivative respect temperature dm/dT.

Figure 2

(Color online) $E\left(J\right)$ characteristic curves measured in a four-points configuration for LSMO films with (a) high resistivity contacts (nonlithographed film, $\text{DCP}\sim 17\text{W}{\text{cm}}^2$); (b) low resistivity contacts at 5 K $\left(\text{DCP}\sim {10}^{-11}\text{W}{\text{cm}}^2\right)$ and 300K $\left(\text{DCP}\sim {10}^{-12}\text{W}{\text{cm}}^2\right)$ in a lithographed film. Note that data for different field values overlap at this scale; (c) for different dissipative contact powers in lithographed ($\text{DCP}\sim 0.5\text{W}{\text{cm}}^2$, ● and $\text{DCP}\sim {10}^{-12}\text{W}{\text{cm}}^2$, ○) and nonlithographed ($\text{DCP}\sim {10}^{-7}\text{W}{\text{cm}}^2$, ◼) films. The maximum Joule heating density achieved in these measurements is in the range of $Q\sim 2\times {10}^{12}\text{J}{\text{m}}^{-3}$.

Figure 3

(Color online) Top: (a) optical micrograph of a lithographed sample and (b) magnification of the lithographed tracks. Bottom: SFM images of (c) topography and (d) local conductivity map at $V_{\text{tip}}=-1V$ measured at the lithographed LSMO bridge. The nonconducting character of the bare STO substrate contrasts with the conducting character observed in the LSMO thin film track. The maximum current response corresponds to saturation of the current amplifier, fixed to 100 nA in the experimental set up used.

Figure 4

(Color online) $E\left(J\right)$ curves measured at 300 K and zero magnetic field in nonlithographed LSMO thin films up to $J_{\text{max}}=5\times {10}^4\text{A}{\text{cm}}^{-2}$ (a) and $J_{\text{max}}={10}^5\text{A}{\text{cm}}^{-2}$ (b). The calculated temperature increases based on a steady heat transfer through the substrate is indicated by a color code as indicated. (c) Main panel: Representation of the two members of Eq. (3), i.e., the resistivity under zero magnetic field data of Fig. 1 (●) and ${\kappa }_{\text{s}}/J^2$ $t_{\text{s}}{t}_{\text{f}}$ $\left(T-T_{\text{B}}\right)$ for lithographed and nonlithographed films. For $T_{\text{B}}=300\text{K}$, the three straight lines with decreasing slope, each for the experimental values $J=2\times {10}^4\text{A}{\text{cm}}^{-2}$, $J=5\times {10}^4\text{A}{\text{cm}}^{-2}$, and $J=1\times {10}^5\text{A}{\text{cm}}^{-2}$, respectively, were used to obtain the corresponding $T_{\text{f}}$. The steeper line corresponds to the lithographed film. The inset is a magnification of the 300 K region to show the small rise in temperature. For comparison, two extreme cases are shown, in the main panel, for $T_{\text{B}}=5\text{K}$: the maximum $J$ value employed in the presented experiments ($J=2\times {10}^5\text{A}{\text{cm}}^{-2}$, large slope line) and a typical $J$ value ($J=5\times {10}^5\text{A}{\text{cm}}^{-2}$, low slope line) required to have bistability (see text).

Figure 5

(Color online) (a) Current SFM image $\left(5\mu \text{m}\times 5\mu \text{m}\right)$ of a LSMO film surface where Au dots have been deposited (see text for details). The Au dots are easily identified at a small sensing bias. (b) Typical $I\text{−}V$ curve measured under direct contact between tip and LSMO surface. (c) Typical $I\text{−}V$ curve measured through a contact of the tip on top of the Au nanodots. Flat regions response in the current are due to the saturation of the current amplifier, fixed to 100 nA in the experimental set up used.