The Interface Structure of FeSe Thin Film on CaF₂ Substrate and its Influence on the Superconducting Performance

Abstract

The investigations into the interfaces in iron selenide (FeSe) thin films on various substrates have manifested the great potential of showing high-temperature-superconductivity in this unique system. In present work, we obtain FeSe thin films with a series of thicknesses on calcium fluoride (CaF) (100) substrates and glean the detailed information from the FeSe/CaF interface by using scanning transmission electron microscopy (STEM). Intriguingly, we have found the universal existence of a calcium selenide (CaSe) interlayer with a thickness of approximate 3 nm between FeSe and CaF in all the samples, which is irrelevant to the thickness of FeSe layers. A slight Se deficiency occurs in the FeSe layer due to the formation of CaSe interlayer. This Se deficiency is generally negligible except for the case of the ultrathin FeSe film (8 nm in thickness), in which the stoichiometric deviation from FeSe is big enough to suppress the superconductivity. Meanwhile, in the overly thick FeSe layer (160 nm in thickness), vast precipitates are found and recognized as Fe-rich phases, which brings about degradation in superconductivity. Consequently, the thickness dependence of superconducting transition temperature (T) of FeSe thin films is investigated and one of our atmosphere-stable FeSe thin film (127 nm) possesses the highest T/T as 15.1 K/13.4 K on record to date in the class of FeSe thin film with practical thickness. Our results provide a new perspective for exploring the mechanism of superconductivity in FeSe thin film via high-resolution STEM. Moreover, approaches that might improve the quality of FeSe/CaF interfaces are also proposed for further enhancing the superconducting performance in this system.

Enhanced superconductivity induced by several-unit-cells diffusion in an FeTe/FeSe bilayer heterostructure

“…Thus, the discovery of a superconducting transition in a surprisingly thin FeTe/FeSe bilayer structure is unexpected and definitely worth further exploration to clarify the mechanism. Furthermore, the T onset c of #FT-300 is 2.5 K higher than the T onset c = 15.1 K from a 127 nm FeSe film in our recent work [55] [shown by magenta diamonds in Fig. 1(a)] which is one of the highest T c 's that can be obtained in a pristine PLD-FeSe thin film grown on a CaF 2 substrate.…”

“…Figure 1 shows the transport superconducting performance obtained from the two samples in this work together with the , which is defined as the temperature at which the resistivity has dropped to one-half of the value at 25 K. It should be noted that T c and B c2 are both higher in the #FT-300 sample than those from the optimized results obtained in a 127 nm FeSe film in our previous work [55]. Considering that the entire thickness of #FT-300 is no more than 15 nm (including both FeSe and FeTe layers), normally, no superconducting transition would be expected in this sample, which is because traditional FeSe or FeSeTe films with ultrathin thickness always show insulating behavior [55,61], just as in the case of 8 nm FeSe (black triangles) plotted in Fig. 1(a) (excluding the cases of HTS FeSe involving heavy electron doping [28][29][30][31][32][33][34][35][36][37]45,46]), and superconductivity is also absent in pure FeTe thin films down to 2 K [62,63].…”

“…Even though the T c is currently not comparable with the T c record achieved in ultrathin FeSe thin films, a deep investigation into the Fe-Ch thick film is still of great importance to reveal the universal mechanism behind the superconductivity in Fe-Ch and similar systems. Using the pulsed laser deposition (PLD) method, Fe-Ch films with T onset c over 15 K [55] in FeSe, or 21 K [56] in FeSe 0.5 Te 0.5 , can be easily prepared without external electron doping, although there is a minimum thickness for this system to exhibit superconductivity, empirically over 20 nm for FeSe [53,57,58] and over 30 nm for Fe-Se-Te [43,51,[56][57][58][59][60]. Films with less thickness are generally nonsuperconducting or even insulating.…”

“…Nabeshima et al [53] demonstrated that low T c in FeSe films thinner than 100 nm was due to the excess-Fe disorder near the surface of the films. The superconductinginsulating transition (SIT) was ascribed to a disorder-driven effect by Schneider et al [55,56] in their FeSe films prepared by a sputtering technique. More recently, we reported similar SIT behavior and ascribed the insulating resistivity behavior of an 8 nm FeSe film to the severely unbalanced FeSe stoichiometry, which is caused by the calcium selenide (CaSe) interlayer that universally exists at FeSe/CaF 2 interfaces [55].…”

“…The superconductinginsulating transition (SIT) was ascribed to a disorder-driven effect by Schneider et al [55,56] in their FeSe films prepared by a sputtering technique. More recently, we reported similar SIT behavior and ascribed the insulating resistivity behavior of an 8 nm FeSe film to the severely unbalanced FeSe stoichiometry, which is caused by the calcium selenide (CaSe) interlayer that universally exists at FeSe/CaF 2 interfaces [55]. Despite the various viewpoints, the intrinsic reason for the suppression of superconductivity in undoped ultrathin Fe-Ch films still remains elusive, which is preventing the discovery of HTS in this system.…”