Journal of Alloys and Compounds 521 (2012) 121–125 Contents lists available at SciVerse ScienceDirect Journal of Alloys and Compounds jou rn al h om epage: www.elsevier.com/locate/jallcom CdI 2 structure type as potential thermoelectric materials: Synthesis and high temperature thermoelectric properties of the solid solution TiS x Se 2-x Franck Gascoin ∗ , Nunna Raghavendra, Emmanuel Guilmeau, Yohann Bréard Laboratoire CRISMAT UMR 6508 CNRS ENSICAEN, 6 boulevard du Maréchal Juin, 14050 Caen Cedex 04, France a r t i c l e i n f o Article history: Received 6 December 2011 Received in revised form 6 January 2012 Accepted 12 January 2012 Available online 28 January 2012 Keywords: Thermoelectric CdI2 type structure Titanium selenide Titanium sulfide Solid solution a b s t r a c t Polycrystalline samples of the solid solution TiS x Se 2-x with x varying from 0 to 2 were prepared using direct high temperature reaction of stoichiometric amounts of the elements. Rietveld refinements of powder X-ray diffraction data are consistent with the existence of a full solid solution. High tempera- ture Seebeck coefficient, electrical resistivity and thermal diffusivity measurements were performed on pellets densified by spark plasma sintering. These measurements reveal that along the solid solution the transport properties vary from the rather metallic and p-type character of TiSe 2 to the semiconducting and n-type of TiS 2 . This change of conduction regime is responsible for the peculiar evolutions of trans- port properties of TiS 0.5 Se 1.5 with increasing temperature that vary somewhat differently than that of the other members of the solid solution. As expected, the disorder generated by the mixed occupancy of the S and Se on the anionic site is responsible for the diminution of the lattice thermal conductivity. A maximum zT above 0.4 at 400 ◦ C is reached for TiS 1.5 Se 0.5 . © 2012 Elsevier B.V. All rights reserved. 1. Introduction Thermoelectricity is nowadays considered as a plausible way to produce “clean” electrical energy from virtually any kind of waste heat [1]. However, the need for always higher device efficiency combined with the mandatory lowering of the cost of the watt thermoelectrically produced, help maintaining upstream material research. Material development thus passes by the discovery of novel phases. In fact, the past few years have witnessed the emer- gence of new families of compounds, some of which are now regarded as promising thermoelectric materials. It is the case of certain Zintl phases such as Yb 14 MnSb 11 and its derivative for high temperature spatial applications [2–6], or some members of the CaAl 2 Si 2 structure type [7–12]. Other such new families are the molybdenum selenides based on Mo 9 Se 11 clusters [13], and lay- ered sulfides [14,15], oxyselenides [16] or selenides [17]. The other route to reach efficient materials, also the object of many efforts worldwide, is by optimizing the properties of known good thermo- electric materials such as Bi 2 Te 3 , PbTe, or SiGe for example. Today, the most “à la mode” utilized method is probably by way of nano- engineering aiming at decreasing the lattice contribution to the thermal conductivity  to further increase the thermoelectric fig- ure of merit zT defined as ˛ 2 T/ with ˛ the Seebeck coefficient,  the electrical resistivity, and T the absolute temperature. To reach ∗ Corresponding author. Tel.: +33 231 452 605; fax: +33 231 951 600. E-mail address: franck.gascoin@ensicaen.fr (F. Gascoin). such an objective there are two general routes, the synthesis of nano-powders followed by their fast densification usually by spark plasma sintering to form nano-structured materials, or the intro- duction of nano-domains within a classical, e.g.; not nanometric, bulk thermoelectric matrix. The production of these nano-objects or nano-domains can be achieved by different methods includ- ing melt-spinning [18], ball-milling [19], microwave processing [20], solid state precipitation [21], spinodal decomposition [22], or solution chemistry [23]. More recently, another way has been demonstrated to improve the material performances, using band structure engineering like by introducing thallium in PbTe thus distorting the electronic density of state [24], or in Na-doped PbTe 1-x Se x taking advantage of the convergence of electronic bands [25]. No matter what method is used in the quest for efficient thermo- electric materials, one common prerequisite for the material is its propensity to incorporate impurities. In other words, if the thermo- electric properties can be optimized it is often through substitution or doping, meaning that the crystal structure of the considered phase must be flexible enough to be capable of accommodating all sorts of impurity such as dopants, nano-particles, nano-domains or simply to easily form solid solutions. For these reasons, it appears that layered structures are ideal candidates that could fulfill all these conditions. For example, several compounds crystallizing in the layered structures type CaAl 2 Si 2 and CdI 2 have respectable thermoelectric properties and more importantly demonstrate that there is a great deal of compounds susceptible of possessing inter- esting thermoelectric properties precisely because of the flexibility 0925-8388/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2012.01.067