Detection of orbital fluctuations above the structural transition temperature in the iron pnictides and chalcogenides

We use point-contact spectroscopy (PCS) to probe $AE{\text{Fe}}_{2}A{s}_2$ ($AE=\text{Ca,}\text{Sr,}\text{Ba}$) and ${\mathrm{Fe}}_{1+\mathrm{y}}\mathrm{Te}$. For $AE=\text{Sr},\text{Ba}$ we detect orbital fluctuations above $T_S$ while for $AE=\text{Ca}$ these fluctuations start below $T_S$. Co doping preserves the orbital fluctuations while K doping suppresses it. The fluctuations are only seen at those dopings and temperatures where an in-plane resistive anisotropy is known to exist. We predict an in-plane resistive anisotropy of ${\mathrm{Fe}}_{1+\mathrm{y}}\mathrm{Te}$ above $T_S$. Our data are examined in light of the recent work by Lee and Phillips (arXiv:1110.5917v2). We also study how joule heating in the PCS junctions impacts the spectra. Spectroscopic information is only obtained from those PCS junctions that are free of heating effects while those PCS junctions that are in the thermal regime display bulk resistivity phenomena.

Figure 1

(a) $G\left(V\right)$ for ${\mathrm{BaFe}}_2{\mathrm{As}}_2$. Conductance enhancement with peaks at $\sim$65 mV superimposed on a parabolic background is observed at low temperatures. The peaks move in as the temperature is increased and the enhancement survives well above $T_S$ (red curve). (b),(c) $\mathrm{Ba}{\left({\mathrm{Fe}}_{0.95}{\mathrm{Co}}_{0.05}\right)}_2{\mathrm{As}}_2$ and $\mathrm{Ba}{\left({\mathrm{Fe}}_{0.945}{\mathrm{Co}}_{0.055}\right)}_2{\mathrm{As}}_2$ display a coexistence of magnetism and superconductivity. At low temperatures, clear Andreev peaks are observed [right inset (b); the arrows are pointing out the Andreev peaks]. A conductance enhancement with peaks at $\sim 65$ mV coexists with the Andreev spectra and evolves with temperature as it does for ${\mathrm{BaFe}}_2{\mathrm{As}}_2$. This enhancement increases logarithmically near zero bias [left inset (b)]. (d) The overdoped compound $\mathrm{Ba}{\left({\mathrm{Fe}}_{0.875}{\mathrm{Co}}_{0.125}\right)}_2{\mathrm{As}}_2$ shows Andreev spectra below $T_c$. It does not have conductance peaks at higher bias values like the Co underdoped compounds. (e) The hole doped ${\mathrm{Ba}}_{0.8}{\mathrm{K}}_{0.2}{\mathrm{Fe}}_2{\mathrm{As}}_2$ has a coexistence of superconductivity and magnetism. It shows Andreev spectra below $T_c$ and no higher bias conductance enhancement. (f) The zero-bias conductance (ZBC) vs temperature curves for $\mathrm{Ba}{\left({\mathrm{Fe}}_{0.945}{\mathrm{Co}}_{0.055}\right)}_2{\mathrm{As}}_2$ (black dashed curve) and ${\mathrm{Ba}}_{0.8}{\mathrm{K}}_{0.2}{\mathrm{Fe}}_2{\mathrm{As}}_2$ (blue solid curve) corresponding to (c) and (e). Both compounds have a magnetostructural transition and a superconducting transition. Co doping shows a conductance enhancement that survives beyond $T_S$ while K doping does not. The insets correlate the spectra obtained at different temperatures to the ZBC curves.

Figure 2

(a) Conductance spectra for ${\mathrm{SrFe}}_2{\mathrm{As}}_2$. The conductance enhancement lasts above $T_S$ and the spectra is similar to that of ${\mathrm{BaFe}}_2{\mathrm{As}}_2$ [Fig. 1a]. (b) Conductance spectra for ${\mathrm{CaFe}}_2{\mathrm{As}}_2$. In this case the conductance enhancement disappears before $T_S$. (c) Conductance spectra for ${\mathrm{Fe}}_{1.13}\mathrm{Te}$ shows an enhancement that lasts above $T_S$. (d)–(f) The ZBC (blue dashed curves) and the resistance vs temperature (black solid curves) for ${\mathrm{BaFe}}_2{\mathrm{As}}_2$, ${\mathrm{CaFe}}_2{\mathrm{As}}_2$, and ${\mathrm{Fe}}_{1.13}\mathrm{Te}$. For ${\mathrm{CaFe}}_2{\mathrm{As}}_2$ the enhancement disappears before $T_S$ while for ${\mathrm{BaFe}}_2{\mathrm{As}}_2$ and ${\mathrm{Fe}}_{1.13}\mathrm{Te}$ it lasts into the normal state. The insets correlate the spectra obtained at different temperatures to the ZBC curve. The red curve in (d) is a fit to $\rho ={\rho }_0+A{T}^2$.

Figure 3

(a) Phase diagram for $\mathrm{Ba}{\left({\mathrm{Fe}}_{1-\mathrm{x}}{\mathrm{Co}}_{\mathrm{x}}\right)}_2{\mathrm{As}}_2$. For the underdoped side, a new region is marked (diagonal stripes) indicating the conductance enhancement that sets in above $T_S$. (b) $G\left(V\right)$ above $T_S$ for ${\mathrm{BaFe}}_2{\mathrm{As}}_2$ follows a log dependence from $\sim$40 mV to $\sim$90 mV.

Figure 4

(a) Resistance vs temperature for ${\mathrm{BaFe}}_2{\mathrm{As}}_2$. The bulk resistance always rises with an increase in temperature. A gradient change occurs as the magnetostructural transition is crossed. (b) $dV/dI$ for two junctions on ${\mathrm{BaFe}}_2{\mathrm{As}}_2$. Junction 2 is in the thermal limit and follows the functional form of the bulk resistivity (black solid curve, taken at 7.6 K). The junction resistance rises with increasing voltage and there is a kink at $\sim$52 mV corresponding to being heated across the magnetostructural transition. Junction 1 behaves very differently from bulk resistivity (blue dashed curve, taken at 2.0 K). The junction resistance decreases with an increasing voltage from 0 to $\sim$70 mV, and again for voltages larger than $\sim$198 mV. A lack of agreement between bulk resistivity and $dV/dI$ indicates that junction 1 is free of heating effects. (c) The ZBC curve for junction 2 (black solid curve) follows the trend of the bulk resistivity, while that of junction 1 does not (blue dashed curve). The vertical red dashed line marks the magnetostructural transition temperature. (d) Needle-anvil PCS with a Au tip (red dashed curve, taken at 4.0 K) and soft PCS (black solid curve, taken at 6.84 K) on ${\mathrm{Fe}}_{1.13}\mathrm{Te}$ show very similar spectra.

Figure 5

(a) As the temperature is increased, the peak in $dV/dI$ moves to lower bias voltages. However, it does not disappear at $T_N$ $\sim$ 132 K, and survives until 177 K. (b) The bias voltage at which the peak in $dV/dI$ occurs as a function of temperature. Had the peak been due to heating the junction across $T_N$, it would have disappeared at $T_N$, as shown for thermal limit PCS on magnetic materials by Chien et al. (Ref. 55).

Figure 6

(a) Comparing the ZBC and $dI/dV$ curves for ${\mathrm{BaFe}}_2{\mathrm{As}}_2$. The orange dashed and dark yellow dotted curves are the low-temperature $dI/dV$ curve plotted on a temperature scale using $T_{\max }^2=T_{\mathrm{bath}}^2+V^2/4L$. The orange dashed curve uses $L=2.44\times {10}^{-8}\text{W}\Omega {\text{K}}^{-2}$ while the dark yellow dotted curve uses $L=2.1\times {10}^{-7}\text{W}\Omega {\text{K}}^{-2}$. A lack of agreement between the ZBC (solid green curve) and these extrapolated $dI/dV$ curves shows that the junction is not in the thermal limit. (b) Comparing the ZBC and $dI/dV$ curves for a joule heated junction on ${\mathrm{BaFe}}_2{\mathrm{As}}_2$. The kink in the $dI/dV$ spectrum occurs at the same conductance value at which the ZBC crosses the magnetostructural transition, and the two curves may be made to overlap using $L=3.5\times {10}^{-8}\text{W}\Omega {\text{K}}^{-2}$. The lack of a fit at lower bias voltages/lower temperatures implies that either $L$ is varying with temperature or the junction is not completely in the thermal regime.