Measurement of Coherent Polarons in the Strongly Coupled Antiferromagnetically Ordered Iron-Chalcogenide ${\mathrm{Fe}}_{1.02}\mathrm{Te}$ using Angle-Resolved Photoemission Spectroscopy

The nature of metallicity and the level of electronic correlations in the antiferromagnetically ordered parent compounds are two important open issues for the iron-based superconductivity. We perform a temperature-dependent angle-resolved photoemission spectroscopy study of ${\mathrm{Fe}}_{1.02}\mathrm{Te}$, the parent compound for iron chalcogenide superconductors. Deep in the antiferromagnetic state, the spectra exhibit a “peak-dip-hump” line shape associated with two clearly separate branches of dispersion, characteristics of polarons seen in manganites and lightly doped cuprates. As temperature increases towards the Néel temperature ($T_N$), we observe a decreasing renormalization of the peak dispersion and a counterintuitive sharpening of the hump linewidth, suggestive of an intimate connection between the weakening electron-phonon ($e$-ph) coupling and antiferromagnetism. Our finding points to the highly correlated nature of the ${\mathrm{Fe}}_{1.02}\mathrm{Te}$ ground state featured by strong interactions among the charge, spin, and lattice and a good metallicity plausibly contributed by the coherent polaron motion.

Figure 1

(a),(e) Fermi surface map of ${\mathrm{Fe}}_{1.02}\mathrm{Te}$ measured with 22 eV excitation energy at (a) $T=10\mathrm{K}$ and (e) $T=90\mathrm{K}$. The photoemission intensity is integrated over a 10 meV window around the $E_F$. (b),(f) Photoemission intensity of the cut along the $\Gamma \mathrm{\text{−}}M$ direction of (a) and (e), respectively. Dashed curves are eye guides to the dispersion bands. (c),(g) Plot of the energy distribution curves (EDCs) around $\Gamma$ along the cuts in (b) and (f). Labeling marks are local maxima of the EDCs after dividing the corresponding Fermi-Dirac function and background subtraction (see below). (d),(h) Plot of the EDCs around $M$ along the cuts in (b) and (f). Labeling marks are local maxima of the EDCs.

Figure 2

(a) Plot of the EDCs around $\Gamma$ along the $\Gamma \mathrm{\text{−}}M$ direction cut measured at 30 K. The data are plotted after normalizing to the intensity at the highest binding energy and dividing the corresponding Fermi-Dirac function. The green EDC is taken as the background to be subtracted from the blue EDCs around $\Gamma$. (b) Plot of the blue EDCs around $\Gamma$ after background subtraction in (a). Red lines are Gaussian and Lorentzian fitting results of the EDCs. Fitted Guassian (Lorentzian) peaks are marked by open circles (triangles). The dashed line denotes the position of the dips in the EDCs, which is 18 meV below $E_F$. (c) Photoemission intensity plot of the EDCs in (b), together with marks labeling the peak positions in the EDCs and MDCs. Peak positions of the MDCs of the $\sigma$ band are determined by fitting to a two-Lorentzian function. The dashed curve is the parabolic fit of the $\sigma$ band MDC peaks.

Figure 3

(a) Plot of the EDCs at the $M$ point at various temperatures. The red EDC is recorded when $T=T_N$. Marks labeling peaks of the $\eta$ and ${\eta }^{\prime }$ bands are local maxima of the EDCs. (b) Plot of the $\eta$ band binding energy and linewidth at the $M$ point together with the (0.5, 0, 0.5) AFM Bragg peak intensity versus temperature. The magnetic peak intensity curve is adapted from a published neutron scattering experiment [37] and roughly proportional to the ordered magnetic moment of Fe.

Figure 4

(a) Photoemission intensity around $\Gamma$ along the $\Gamma \mathrm{\text{−}}M$ direction at various temperatures. Red arrows indicate the positions of the dips in the spectra. (b) The second derivative plot of each intensity plot in (a). White curves are the parabolic fit of the MDC peaks found in (a). (c) Plot of the $\sigma$ band depth (derived from the fitting results in (b)) versus temperature. (d) Magnified plot of photoemission intensity in the $E_F$ vicinity at various temperatures. The data are processed using the similar method as in Fig. 2. Blue marks denote the peaks of the EDCs of the ${\sigma }^{\prime }$ band and the red curve is the parabolic fit of these peaks. (e) Plot of the effective mass of the ${\sigma }^{\prime }$ and $\sigma$ band versus temperature. The effective mass of the ${\sigma }^{\prime }$ and $\sigma$ band is calculated from the EDC and MDC [Fig. 4b] fitting results, respectively. If not shown, the error bars are smaller than the symbol’s size.