Higher-order epitaxy: A pathway to suppressing structural instability and emergent superconductivity

Abstract

Molecular beam epitaxy enables the growth of thin film materials with novel properties and functionalities. Typically, the lattice constants of films and substrates are designed to match to minimise disorders and strains. However, significant lattice mismatches can result in higher-order epitaxy, where commensurate growth occurs with a period defined by integer multiples of the lattice constants. Despite its potential, higher-order epitaxy is rarely used to enhance material properties or induce emergent phenomena. Here, we report single-crystalline FeTe films grown via 6:5 commensurate higher-order epitaxy on CdTe(001) substrates. Scanning transmission electron microscopy reveals self-organised periodic interstitials near the interface, arising from higher-order lattice matching. Synchrotron x-ray diffraction shows that the tetragonal-to-monoclinic structural transition in bulk FeTe is strongly suppressed. Remarkably, these films exhibit substrate-selective two-dimensional superconductivity, likely due to suppressed monoclinic distortion. These findings demonstrate the potential of higher-order epitaxy as a tool to control materials and inducing emergent phenomena.

Figures (12)

• Figure 1: Crystal structures, characterisation, and substrate-selective superconductivity of FeTe films.a, High-temperature tetragonal and b, low-temperature monoclinic crystal structures of FeTe VESTA. The red boxes indicate the unit cell. The in-plane anisotropy $a/b$ and the tilt angle $\pi/2 - \beta$ are slightly exaggerated to emphasize the monoclinic distortion. c, Crystal structure of CdTe. d, Crystal structure of SrTiO$_3$ (STO). e, X-ray diffraction (XRD) $\theta$-2$\theta$ profile of for a FeTe/CdTe film. The blue and magenta arrows indicate FeTe (0 0 $n$) and CdTe (0 0 2$m$) reflections, respectively, where $n$ and $m$ are integers. The asterisk demotes an impurity phase of NaCl-type CdTe. f, XRD rocking curves for the FeTe (0 0 1) reflection for FeTe/CdTe (red) and FeTe/STO (grey) films. g, XRD azimuthal profiles for the asymmetric Bragg points of FeTe (1 0 4) (blue) and CdTe (1 1 5) (magenta). h, Atomic force microscopy image for an FeTe/CdTe film after vacuum annealing. The scale bar represents 0.5 $\mu$m. i, Temperature dependence of resistivity for FeTe films of 1000 nm thickness grown on CdTe and STO.
• Figure 1: Characterisation of an FeTe/STO film.a, XRD $\theta$-2$\theta$ profile. Blue and orange arrows indicate Bragg peaks of FeTe and STO, respectively. Cyan arrows indicate peaks of FeTe$_2$ capping layer. b, XRD azimuthal profiles of asymmetric Bragg reflections FeTe (1 0 4) (blue) and STO (1 0 3) (orange). c, Atomic force microscopy image of an as-grown FeTe/STO film. The scale bar represents 0.5 $\mu$m.
• Figure 2:
• Figure 2: Real-space observation of periodic interstitial modulation near FeSe$_{0.1}$Te$_{0.9}$/CdTe interface.a, Cross-sectional HAADF-STEM image near an FeSe$_{0.1}$Te$_{0.9}$/CdTe interface. Yellow, brown, and magenta spheres indicate Te/Se, Fe, and Cd atoms, respectively. Red circles represent interfacial interstitials. b, Integrated-intensity profiles along the cut A and B, as indicated in a. The dots above the profiles indicate the equilibrium position for each atom. Grey areas indicate the sites where the interfacial ions approach most closely. Red arrows represent interfacial interstitials. The dashed line denotes the amplitude modulation extracted by a sinusoidal fitting with a wave length of $6a_{\rm FST[100]}$.
• Figure 3: Suppression of monoclinic lattice distortion in FeTe/CdTe films.a-c, XRD $\theta$-2$\theta$ profiles around the FeTe (0 0 7) Bragg reflection for a FeTe1000/STO, b FeTe1000/CdTe, and c FeTe40/CdTe. Each profile is vertically shifted for clarity. Grey and red curves are Gaussian fittings for the tetragonal and monoclinic phases, respectively. Vertical dashed lines are guides for eyes. d, Temperature dependence of the lattice parameter $c$ deduced from the FeTe (0 0 7) reflections. Orange, cyan, and purple markers denote FeTe1000/STO, FeTe1000/CdTe, and FeTe40/CdTe, respectively. White and black squares are $c$ for bulk single crystals reported by neutron neutron2009bao and XRD XRD2011Xiao studies. Cyan and purple lines are simulations for the Poisson ratios $\nu$ = 0.4 and 0.5, respectively, assuming perfect clamping to the CdTe substrate. The data highlighted in red and grey areas are those for tetragonal and monoclinic phases, respectively. The error bars represent the standard deviation derived from fittings. e, Fe moments $\mu$ and energy difference between AFM and paramagnetic phases, $E_{\rm AFM} - E_{\rm PM}$, as a function of out-of-plane compressive strain $-\Delta c/c$, obtained by first-principles calculation. f, Temperature dependence of the volume fraction of the monoclinic phase, $V_{\rm M}/(V_{\rm T} + V_{\rm M})$. The inset schematically depicts a film with a phase separation into monoclinic and tetragonal. The lines are guides for the eyes. The error bars represent the standard deviation derived from fittings.