The synthesis and fabrication of nanoscale materials for new types of electronic and magnetic devices is a central theme in materials science research in this second decade of the 21st century. Given that conventional storage materials are estimated to approach their miniaturization limit by 2016, [1] heightened efforts are being directed at the design and synthesis of new types of bistable nanoscale materials, including those capable of undergoing a change from low to high resistance under the application of an electric field. Such nonvolatile memory devices are capable of operating at increased speeds and require less energy than conventional memory devices. Among the materials being investigated for resistance-based memory are materials that contain organic components and whose properties are influenced by magnetic or electric fields. [2] Materials that respond to the application of an electric field or changes in light, pressure, or temperature are being sought for incorporation into electronic devices with ultrafast operating speeds. [3,4] Examples of molecule-based materials that exhibit fascinating properties are the spin-crossover complex [Fe(picolylamine) 3 Cl 2 (C 2 H 5 OH)], [5,6] the neutralionic transition system TTF-chloranil (TTF = tetrathiafulvalene), [7][8][9] the metallo-organic conductor Cu(DM-DCNQI) 2 , [10][11][12][13][14][15][16] (DM-DCNQI = dimethyl-N,N'-dicyanoquinonediimine) and the salt (EDO-TTF) 2 PF 6 , [17,18] (EDO-TTF = ethylenedioxytetrathiafulvalene). These materials provide compelling evidence for the contention that molecular solids may eventually be useful in device applications.
In terms of electric-field-induced behavior, the most extensively studied examples are the organocyanide-based materials Cu(TCNQ) (TCNQ = 7,7,8,8-tetracyanoquinodi-methane), which exhibits reversible switching from a highresistance state to a conducting state promoted by the application of an electric field or upon irradiation, [19][20][21] and the current-driven conductor K(TCNQ) salt. [22] The latter material is a key member of the binary series of alkali-metal salts of TCNQ that behave as so-called "Mott insulators" at high temperatures, in which the fully reduced radical anions are arranged in columns with evenly spaced TCNQ units. At lower temperatures, these "soft" materials undergo a phase transition in which the TCNQ units are brought into close proximity as a result of p dimerization. The electrons are then trapped in the dimers, the conductivity drops, and the materials pass into the spin-Peierls insulating state.
An approach that we have adopted for discovering conducting TCNQ phases is to capitalize on the rich chemistry of alkali metals while circumventing some issues that hinder their conductivity. In this vein, thallium is an interesting element, since it can behave as a pseudo-alkali metal. In contrast to other Group 13 elements, Tl prefers the 1 + oxidation state (although Tl 3+ is known), and many similarities between the chemistry of alkali-metal ions and Tl + have been noted. [23] The electronegativity of Tl (2.04) is much higher than that of any alkali metal, which should lead to less ionic compounds with smaller band gaps and thus higher carrier mobility. Moreover, unlike alkali metals, Tl + possesses a stereoactive lone pair, which is expected to lead to a greater diversity of structures. [24] Indeed, the viability of this idea was demonstrated by Hünig et al., who reported Tl(DM-DCNQI) 2 , which adopts a 3D metal-organic framework structure and behaves as a one-dimensional metal-like semiconductor (s 300 K = 50 S cm À1 ). [25] With the exception of the aforementioned material, there are no other reported main-group binary phases based on weak interactions with organocyanide molecules. In fact, main-group supramolecular chemistry is largely underdeveloped as compared to that of transition-metal ions. [26][27][28] Herein we describe the first chemistry of the Tl I cation with TCNQ radical anions, the result of which is the discovery of two polymorphs with very different conducting properties.
Slow diffusion of a methanol solution of Li(TCNQ) and an aqueous solution of TlPF 6 leads to the isolation of single crystals of the product Tl(TCNQ), phase I (1). A typical bulk stoichiometric reaction leads to crystals of a second product Tl(TCNQ), phase II (2). An X-ray structural determination revealed that 1 crystallizes in the P2 1 /c space group as a 3D network structure consisting of metal ions arranged in linear strings, each surrounded by four stacks of TCNQ acceptor molecules (Figure 1 a) and with adjacent TCNQ stacks [*] Dr.