Dependence of the structural, transport and magnetic properties of Tl(1-y)Fe(2-z)(Se(1-x)S(x))2 with isovalent substitution of Se by S.

The effect of selenium substitution by sulfur in the Tl(1-y)Fe(2-z)Se(2) antiferromagnet was studied by x-ray and electron diffraction, magnetization and transport measurements. Tl(0.8)Fe(1.5)(Se(1-x)S(x))(2) (nominal composition) solid solution was synthesized in the full x range (0 ≤ x(S) ≤ 1) using the sealed tube technique. No superconductivity was found down to 4.2 K in the series despite the fact that the optimal crystallographic parameters, determined by Rietveld refinements, are reached in the series (i.e. the Fe-(Se, S) interplane height and (Se, S)-Fe-(Se, S) angle for which the critical superconducting transition T(c) is usually maximal in pnictides). A quasi-full Tl site (y ~ 0.05) compared to significant alkaline deficiency (y = 0.2-0.3) in analogous A(1-y)Fe(2-z)Se(2) (A = K, Rb, Cs), and the resulting differences in iron valency, density of states and doping, are suggested as an explanation for this absence of superconductivity. Transmission electron microscopy confirmed the existence of an ordered iron vacancies network in the samples of the Tl(0.8)Fe(1.5)(Se(1-x)S(x))(2) series in the form of the tetragonal √5a × √5a × c superstructure (I4/m). The Néel temperature (T(N)) indicating the onset of antiferromagnetism order in this √5a × √5a × c supercell is found to decrease from 450 K in the selenide (x = 0) to 330 K in the sulfide (x = 1). Finally, we demonstrate a direct linear relationship between T(N) and the Fe-(Se, S) bond length (or Fe-(Se, S) height).