Coexistence of superconductivity and density wave in ${\mathrm{Ba}}_2{\mathrm{Ti}}_2{\mathrm{Fe}}_2{\mathrm{As}}_4\mathrm{O}$: An optical spectroscopy study

We performed an optical spectroscopy measurement on single crystals of ${\mathrm{Ba}}_2{\mathrm{Ti}}_2{\mathrm{Fe}}_2{\mathrm{As}}_4\mathrm{O}$, which is a newly discovered superconductor showing a coexistence of superconductivity and density wave orders. The study reveals spectral changes associated with both density wave and superconductivity phase transitions. The density wave phase transition at $T_{\mathrm{DW}}\approx 125$ K leads to the reconstruction of Fermi surfaces which removes about half of Drude spectral weight. The ratio of $2{\Delta }_{\mathrm{DW}}/k_B{T}_{\mathrm{DW}}\approx 11.9$ is considerably larger than the mean-field value based on the weak-coupling BCS theory. At the lowest temperature in the superconducting state, further spectral change associated with the superconducting condensate is observed. The low frequency optical conductivity could be well modeled within the Mattis-Bardeen approach with two isotropic gaps of ${\Delta }_1\left(0\right)=3.4$ meV and ${\Delta }_2\left(0\right)=7.9$ meV. The superconducting properties of the ${\mathrm{Ba}}_2{\mathrm{Ti}}_2{\mathrm{Fe}}_2{\mathrm{As}}_4\mathrm{O}$ compound are similar to those of ${\mathrm{BaFe}}_{1.85}{\mathrm{Co}}_{0.15}{\mathrm{As}}_2$.

Figure 1

Optical reflectivity $R\left(\omega \right)$ in the frequency range between 30 and 3000 ${\mathrm{cm}}^{-1}$ at five representative temperature. The inset: $R\left(\omega \right)$ at 300 K up to 35000 ${\mathrm{cm}}^{-1}$ in a linear frequency range.

Figure 2

The real part of optical conductivity ${\sigma }_1\left(\omega \right)$ for ${\mathrm{Ba}}_2{\mathrm{Ti}}_2{\mathrm{Fe}}_2{\mathrm{As}}_4\mathrm{O}$ below 3000 ${\mathrm{cm}}^{-1}$. The inset: ${\sigma }_1\left(\omega \right)$ up to 35 000 ${\mathrm{cm}}^{-1}$ for 300K.

Figure 3

Cutoff frequency dependent spectral weight at four different temperature. Inset: the normalized spectral weight $W_s\left(30\text{K}\right)/W_s\left(300\text{K}\right)$ up to 10 000 ${\mathrm{cm}}^{-1}$.

Figure 4

The experimental data of ${\sigma }_1\left(\omega \right)$ at 300 and 30 K together with the Drude-Lorentz fits.

Figure 5

The real part of in-plane optical conductivity of ${\mathrm{Ba}}_2{\mathrm{Ti}}_2{\mathrm{Fe}}_2{\mathrm{As}}_4\mathrm{O}$ at 8 K together with two gap fits and mid-infrared component.