Charge transport and magnetization profile at the interface between the correlated metal ${\text{CaRuO}}_3$ and the antiferromagnetic insulator ${\text{CaMnO}}_3$

A combination of spectroscopic probes was used to develop a detailed experimental description of the transport and magnetic properties of superlattices composed of the paramagnetic metal ${\text{CaRuO}}_3$ and the antiferromagnetic insulator ${\text{CaMnO}}_3$. The charge-carrier density and Ru valence state in the superlattices are not significantly different from those of bulk ${\text{CaRuO}}_3$. The small charge transfer across the interface implied by these observations confirms predictions derived from density-functional calculations. However, a ferromagnetic polarization due to canted Mn spins penetrates 3–4 unit cells into ${\text{CaMnO}}_3$, far exceeding the corresponding predictions. The discrepancy may indicate the formation of magnetic polarons at the interface.

Figure 1

(Color online) (a) Reflection high-energy electron-diffraction pattern after growth of the final layer for the $N=10\text{u}\text{.c}\text{.}$ sample. (b) (H0L) scan for the $N=10\text{u}\text{.c}\text{.}$ sample on $\left(H=0\right)$ and slightly off $\left(H=0.01\right)$ the specular rod. (c) HK map on the film (003) Bragg peak showing that the $N=10\text{u}\text{.c}\text{.}$ sample is fully lattice matched to the LAO substrate.

Figure 2

(Color online) Normalized sheet resistance of CMO-CRO superlattices as a function of temperature.

Figure 3

(Color online) Experimental (solid lines) and best-fit calculated (open circles) (a) near normal-incidence reflectivity and [(b) and (c)] ellipsometry spectra of the $N=10$ superlattice, $T=300\text{K}$. The reflectivity spectra of the bare ${\text{LaAlO}}_3$ substrate is shown for comparison (black solid line). Solid lines in ellipsometry spectra are fits to the model discussed in the text.

Figure 4

(Color online) Best-fit model functions ${\sigma }_1\left(\omega \right)$ and ${\epsilon }_1\left(\omega \right)$ for superlattices with $N=4$ and 10 consecutive ${\text{CaRuO}}_3$ u.c., as obtained from FIR ellipsometry data by inversion of the ellipsometric parameters. The dashed lines represent the results of Drude fits, as described in the text. The peak at $220{\text{cm}}^{-1}$ superposed on the broad Drude response is due to a bond-bending lattice vibration of ${\text{CaMnO}}_3$.

Figure 5

(Color online) (a) Absorption and XMCD measured at the $\text{Mn}{L}_3$ edge at $T=10\text{K}$ for $N=6$. The XMCD of signal is sizable $\left(\sim 12\mathrm{\%}\right)$ and consistent with rather large magnetic moment. (b) Absorption and XMCD from the $\text{Ru}{L}_3$ edge for the 4 unit cell ${\text{CaRuO}}_3$ sample. Comparison with reference samples indicates that for both the Mn and Ru, the valence is close to $4+$. The very small apparent Ru XMCD signal is within the systematic error of the measurement.(c) Field dependence of the Mn XMCD measured at 10 K showing that the magnetic moment saturates near 4 T but shows no sign of remanent magnetization.

Figure 6

(Color online) (a) Magnetization versus temperature measured by total moment magnetometry, which show the clear onset of order close to $T_N\sim 120\text{K}$. (b) Field dependence of the Mn XMCD measured at 10 K showing that the magnetic moment saturates near 4 T but shows no sign of remanent magnetization.

Figure 7

(Color online) (a) Specular reflectivity measured at 50 K for $E=639.2\text{eV}$ on the $\text{Mn}{L}_3$ edge (upper trace) and at room temperature with photon energy $E=620\text{eV}$, below the $\text{Mn}{L}_3$ edge (lower trace). The lines are the results of fits to a structural model discussed in the text. The Bragg-peak positions are indicated. (b) Dichroic differences $I^{+}-I^{-}$ measured at the Bragg-peak positions, normalized to the peak intensity. (c) Calculated normalized difference at the same Bragg-peak positions for different layer thicknesses of the interface magnetized regions in ${\text{CaMnO}}_3$.