Translating reproducible phase-separated texture in manganites into reproducible two-state low-field magnetoresistance: An imaging and transport study

In thin-film ${\text{La}}_{0.6}{\text{Ca}}_{0.4}{\text{MnO}}_3$, conducting-tip and magnetic force microscopy reveal a pattern of nanoscale phase separation that is reproducible across cooling runs. This pattern represents the intersection of buried three-dimensional filamentary ferromagnetic metallic pathways with the sample surface. As an interlayer in current-perpendicular-to-the-plane trilayer devices, this phase-separated material magnetically decouples ferromagnetic metallic ${\text{La}}_{0.7}{\text{Ca}}_{0.3}{\text{MnO}}_3$ electrodes which switch sharply. This yields sharp two-state low-field magnetoresistance that is also reproducible across cooling runs. The reproducibility and the magnitude of the resistance jump are linked to highly resistive $\left(\sim {10}^{-12}\Omega {\text{m}}^2\right)$ constrained domain walls in the pathways of the phase-separated interlayer. Phase separation is normally associated with high-field colossal magnetoresistance and, therefore, its exploitation here to produce low-field effects is unusual.

Figure 1

Magnetic and electrical properties of an unpatterned 55 nm ${\text{La}}_{0.6}{\text{Ca}}_{0.4}{\text{MnO}}_3$ film. (a) Magnetization $M$ at 48 K, and (b) resistivity $\rho$ at 78 K versus $H$ applied along [100].

Figure 2

(Color online) Conducting-tip atomic force microscopy (CAFM) images of the 55 nm ${\text{La}}_{0.6}{\text{Ca}}_{0.4}{\text{MnO}}_3$ film. Two-point resistance $R$ measurements between the conductive nanoscale tip and a biased electrical contact located remotely on the film surface were normalized by the median value $\overline{R}$ for each image. The main image was taken below $T_p$ at 77 K $\left(\overline{R}=6\text{M}\Omega \right)$ and the inset was taken above $T_p$ at 293 K $\left(\overline{R}=25\text{k}\Omega \right)$. When the sample is macroscopically metallic below $T_p$, the electronic structure exhibits phase separation on a range of length scales.

Figure 3

(Color online) Magnetic force microscopy (MFM) images of a single $8\times 8\mu {\text{m}}^2$ area of the same 55 nm ${\text{La}}_{0.6}{\text{Ca}}_{0.4}{\text{MnO}}_3$ film studied in Fig. 2. All images are displayed on an identical contrast scale. Images (a)–(d) show zero-field data at 77 K after different magnetic and thermal histories. After cooling from RT in $H=0$, the magnetic contrast in (a) is significantly reduced by the application and removal of a saturating field of 36 mT along [100] prior to the collection of image (b). Images (c) and (d) correspond to a repeat of this run. The insets, all from a single $1.2\times 1.2\mu {\text{m}}^2$ area, show more clearly that the magnetic structure is subject to strong run-to-run variations after cooling in $H=0$, (a), (c) but is recovered with remarkable reproducibility by the application of the field (b), (d).

Figure 4

Magnetic measurements of an unpatterned ${\text{La}}_{0.7}{\text{Ca}}_{0.3}{\text{MnO}}_3\left(29\text{nm}\right)/{\text{La}}_{0.6}{\text{Ca}}_{0.4}{\text{MnO}}_3\left(29\text{nm}\right)/{\text{La}}_{0.7}{\text{Ca}}_{0.3}{\text{MnO}}_3\left(72\text{nm}\right)$ trilayer. $M\left(H\right)$ along [100] at 79 K shows two abrupt jumps associated with top $\left(2.9\pm 0.2\text{mT}\right)$ and bottom $\left(0.44\pm 0.10\text{mT}\right)$ layer switching. The more diffuse switching in 35–60 mT is due to the FMM regions of the ${\text{La}}_{0.6}{\text{Ca}}_{0.4}{\text{MnO}}_3$ interlayer. Upper inset: $\omega \text{−}2\theta$ x-ray scan about ${\text{NdGaO}}_3$ (004) taken with a Philips X’Pert high-resolution diffractometer ($\lambda =1.54\text{Å}$, Ge monochromator). The observation of peaks for each component of the trilayer confirms the epitaxy of the heterostructure. Lower inset: Trilayer schematic (${\text{La}}_{1-x}{\text{Ca}}_x{\text{MnO}}_3$ is denoted LCMO $x$ for clarity).

Figure 5

(Color online) CPP electrical measurements after zero-field cooling to 25 K of a $5\times 5\mu {\text{m}}^2$ mesa device patterned from a ${\text{La}}_{0.7}{\text{Ca}}_{0.3}{\text{MnO}}_3\left(20\text{nm}\right)/{\text{La}}_{0.6}{\text{Ca}}_{0.4}{\text{MnO}}_3\left(20\text{nm}\right)/{\text{La}}_{0.7}{\text{Ca}}_{0.3}{\text{MnO}}_3\left(50\text{nm}\right)$ trilayer in which all layers are macroscopically metallic at the measurement temperature. The schematic shows contact pads (thick lines) and voltage $V$ and current $I$ terminals. Each of the three $R\left(H\right)$ sweeps was measured after an excursion to RT and we plot $\Delta R=R\left(H\right)\text{−}R\left(0\right)$ to eliminate small extrinsic run-to-run variations in top contact resistance attributed to microstructural changes and other aging effects. For the three sweeps, $R\left(0\right)=97.9\Omega$ (black), $96.2\Omega$ (red), and $95.8\Omega$ (blue). The switching (jump magnitude and switching fields) is highly reproducible from run to run and two distinct states are seen at $H=0$ (minor loop, inset). The upper switching field corresponds to the top electrode (see main text).