Melting of Charge/Orbital Ordered States in Nd$_{1/2}$Sr$_{1/2}$MnO$_3$: Temperature and Magnetic Field Dependent Optical Studies

TL;DR

This study explores how temperature and magnetic field influence the charge and orbital order in Nd$_{1/2}$Sr$_{1/2}$MnO$_3$, revealing phase transitions and local ordering phenomena through optical and dielectric measurements.

Contribution

It provides new insights into the melting process of charge/orbital order and the emergence of ferromagnetic metallic states under high magnetic fields using optical and dielectric studies.

Findings

Spectral weight changes up to 4.0 eV explained by polaron picture.

Evidence of local ordered states persisting above charge ordering temperature.

Charge/orbital melting occurs via percolation of ferromagnetic metal domains.

Abstract

We investigated the temperature ($T=$ 15 $\sim $ 290 K) and the magnetic field ($H=$ 0 $\sim $ 17 T) dependent optical conductivity spectra of a charge/orbital ordered manganite, Nd$_{1/2}$Sr$_{1/2}$MnO$_3$. With variation of $T$ and $H$, large spectral weight changes were observed up to 4.0 eV. These spectral weight changes could be explained using the polaron picture. Interestingly, our results suggested that some local ordered state might remain above the charge ordering temperature, and that the charge/orbital melted state at a high magnetic field (i.e. at $H=$ 17 T and $% T=$ 4.2 K) should be a three dimensional ferromagnetic metal. We also investigated the first order phase transition from the charge/orbital ordered state to ferromagnetic metallic state using the $T$- and $H$% -dependent dielectric constants $\epsilon_1$. In the charge/orbital ordered insulating state, $\epsilon_1$ was positive and $d\epsilon_1/d\omega \approx 0$. With increasing $T$ and $H$, $\epsilon_1$ was increased up to the insulator-metal phase boundaries. And then, $\epsilon_1$ abruptly changed into negative and $d\epsilon_1/d\omega >0$, which was consistent with typical responses of a metal. Through the analysis of $% \epsilon_1$ using an effective medium approximation, we found that the melting of charge/orbital ordered states should occur through the percolation of ferromagnetic metal domains.