Magnetic order tuned by Cu substitution in Fe${}_{1.1-z}$Cu${}_z$Te

We study the effects of Cu substitution in Fe${}_{1.1}$Te, the nonsuperconducting parent compound of the iron-based superconductor, Fe${}_{1+y}$Te${}_{1-x}$Se${}_x$, utilizing neutron scattering techniques. It is found that the structural and magnetic transitions, which occur at $\sim$60 K without Cu, are monotonically depressed with increasing Cu content. By 10$\%$ Cu for Fe, the structural transition is hardly detectable, and the system becomes a spin glass below 22 K, with a slightly incommensurate ordering wave vector of ($0.5-\delta$, 0, 0.5) with $\delta$ being the incommensurability of 0.02, and correlation length of 12 Å along the $a$ axis and 9 Å along the $c$ axis. With 4$\%$ Cu, both transition temperatures are at 41 K, though short-range incommensurate order at (0.42, 0, 0.5) is present at 60 K. With further cooling, the incommensurability decreases linearly with temperature down to 37 K, below which there is a first-order transition to a long-range almost-commensurate antiferromagnetic structure. A spin anisotropy gap of 4.5 meV is also observed in this compound. Our results show that the weakly magnetic Cu has a large effect on the magnetic correlations; it is suggested that this is caused by the frustration of the exchange interactions between the coupled Fe spins.

Figure 1

Magnetic and structural phase transition in Fe${}_{1.06}$Cu${}_{0.04}$Te. (a) Magnetic susceptibility measured with a 0.1-T filed under zero-field cooling (ZFC) condition. (b) Lattice distortion $s$ defined as $2(a-b)/(a+b)$ obtained from fitting the scans in Fig. 2d. (c) Integrated intensity (left axis) and HWHM (right axis) of the magnetic peak obtained by fitting the $H$ scans through (0.5, 0, 0.5) using Lorentzian functions. (d) A contour map showing a series of $H$ scans through (0.5, 0, 0.5) at different temperatures. Solid lines indicate the magnetic peak position. Vertical dashed lines indicate the transition temperatures as explained in the text.

Figure 2

(a) and (b) Scans along $H$ and $L$ direction through the magnetic peak (0.5, 0, 0.5) at 10, 36, and 37 K in Fe${}_{1.06}$Cu${}_{0.04}$Te. (c) Scans along $H$ direction through (0.5, 0, 0.5) at 3 higher temperatures than in (a). (d) Scans along $H$ direction through the nuclear Bragg peak (201) at 10 and 60 K. Lines through data in (a)–(c) are fits using Lorentzian function. In (d), the 10-K data is fitted with a two-peak Gaussian function and the 60-K data is fitted with a single-peak Gaussian.

Figure 3

Energy scans from 2 to 6 meV at ${\mathbf{Q}}_{\mathbf{0}}=(0.5,0,0.5)$ at 15 and 30 K, and at (0.42, 0, 0.5) at 60 K for Fe${}_{1.06}$Cu${}_{0.04}$Te. Intensity [$I$(${\mathbf{Q}}_{\mathbf{0}},\omega$)] after subtracting the background obtained from the constant-energy scans has been converted to the dynamical susceptibility ${\chi }^{\prime \prime }({\mathbf{Q}}_{\mathbf{0}},\omega )$ using ${\chi }^{\prime \prime }({\mathbf{Q}}_{\mathbf{0}},\omega )$ $=$ $I$(${\mathbf{Q}}_{\mathbf{0}},\omega$)$[1-\exp (-\hslash \omega /k_{\text{B}}T)]$. Arrow indicates the position of the spin anisotropy gap $E_g$.

Figure 4

Scans through (0.5, 0, 0.5) along $H$ direction at 2, 4, and 6 meV at 15 K in Fe${}_{1.06}$Cu${}_{0.04}$Te. A sloping background has been subtracted from all data. Lines through data are fits with Gaussian function.

Figure 5

Short-range incommensurate magnetic order in FeCu${}_{0.1}$Te. (a) Susceptibility measured with 0.1-T filed under ZFC and FC conditions; (b) Intensity vs temperature of the (201) Bragg peak; (c) Integrated intensity (left axis) and HWHM (right axis) obtained by fitting the $H$ scans through the magnetic peak (0.48, 0, 0.5) with Lorentzians. Dashed lines indicate the spin-glass transition temperature.

Figure 6

Scans along $H$ and $L$ direction through the magnetic peak (0.5, 0, 0.5) in FeCu${}_{0.1}$Te.

Figure 7

Magnetic ordering temperature vs Cu doping. Lines are guides to the eye.