Stepwise Neutral−Ionic Phase Transitions in a Covalently Bonded Donor/Acceptor Chain Compound

Abstract

Neutral (N)-ionic (I) transitions in organic donor (D)/acceptor (A) charge-transfer complexes are intriguing because a 'reservoir of functions' is available. For systematically controlling N-I transitions, tuning the ionization potential of D and the electron affinity of A is extremely important. However, the effect of Coulomb interactions, which likely causes a number of charge-gap states at once in a system bringing about stepwise transitions, is a long-standing mystery. Here, we show definite evidence for stepwise N-I transitions caused by contributions from anisotropic interchain Coulomb interactions in a metal-complex-based covalently bonded DA chain compound, [Ru(2)(2,3,5,6-F(4)PhCO(2))(4)(DMDCNQI)]·2(p-xylene) (1; 2,3,5,6-F(4)PhCO(2)(-) = 2,3,5,6-tetrafluorobenzoate; DMDCNQI = 2,5-dimethyl-N,N'-dicyanoquinonediimine), where the [Ru(2)(II,II)(2,3,5,6-F(4)PhCO(2))(4)] moiety has a paddlewheel diruthenium(II,II) motif with a Ru-Ru bond. An intermediate-temperature phase involving self-organized N and I chains was observed in the temperature range between 210 K (= T(2)) and 270 K (= T(1)) with N phase at T > T(1) and I phase at T < T(2). Accompanying the charge transitions, the spin-ground states as well as the ferrimagnetic ordering in the I phase vary. The stepwise feature of the N-I transition with a highly sensitive magnetic response should bring about new dynamical functionalities associated with charge, spin, and lattice.

Magnetic Sponge with Neutral–Ionic Phase Transitions

“…The χ and χT data were simulated at temperatures ranging from 110 to 190 K using an alternating chain model with S i = 3/2 and S i +1 = 1/2 in the Hamiltonian H = −2 J Σ N i =1 S i ·S i +1 ,46 and an adequate parameter set was obtained as g [Ru2] = 2.181(2); g Rad = 2.0 (fix); and J / k B = −105.2(6) K (Figure 4a). The spins of [Ru 2 II,III ] + and TCNQ(EtO) 2 •− were strongly antiferromagnetically coupled, and the magnitude of J was in good agreement with that reported previously for chain compounds 11, 47. The strong antiferromagnetic coupling can be caused by a considerable overlap between the frontier π* singly occupied molecular orbital (SOMO) of [Ru 2 II,III ] + and the π* SOMO of TCNQ(EtO) 2 •− .…”

“…Table S2 (Supporting Information) summarizes the selected bond distances and angles. Although TCNQ(EtO) 2 can serve as a tetradentate ligand with four cyano groups, the moiety in 1‐DCE acted as a linear‐type bidentate ligand with only two cyano groups in the trans position, affording a DA alternating chain motif with an [—(Ru 2 )—{TCNQ(EtO) 2 }—] repeat similar to 0 ,11 and an ionic DA chain compound with bis(1,2,5‐thiadiazolo)tetracyanoquinodimethane (BTDA‐TCNQ) 37. The chains run along the <110> direction, affording a chain‐aggregated layer on the (001) plane (hereafter known as a chain layer; see Figure 1b–d).…”

“…As a result of cooling, the average bond length decreased to around 2.02 Å in the temperature range of 237–213 K. At temperatures less than 213 K, the Ru—O eq bond length was maintained at ≈2.02 Å for [Ru 2 II,III ] + . In addition, the axial bond length (Ru—N ax ), which was also sensitive to the oxidation state of the [Ru 2 ] unit,11 suddenly became shorter on cooling at temperatures less than 237 K (Table S3 (Supporting Information) and inset of Figure 2a). In addition, the Ru—Ru bond length exhibited a small anomaly around 230 K (Table S3 (Supporting Information) and inset of Figure 2a) 11…”

“…This example represents a new research field in organic electron‐donor (D)–acceptor (A) systems,1, 2, 3, 4, 5, 6, 7, 8 with the discovery of the first compound tetrathiafulvalene–chloranil in the class of alternating π‐stacking DA systems by Torrance et al9, 10 However, these organic DA systems were of low interest from the viewpoint of spin systems because of the sole production of S = 1/2 antiferromagnetic pairs in the I‐state involving the structural distortion (i.e., dimerization). Conversely, a covalently bonded DA system exhibiting the N–I phase transition has been observed for a metal complex chain with the D and A of [Ru 2 (2,3,5,6‐F 4 PhCO 2 ) 4 ] and DMDCNQI, respectively, [Ru 2 (2,3,5,6‐F 4 PhCO 2 ) 4 DMDCNQI]·2( p ‐xylene) ( 0 , 2,3,5,6‐F 4 PhCO 2 − = 2,3,5,6‐tetrafluorobenzoate, DMDCNQI = 2,5‐dimethyl‐ N , N ′‐dicyanoquinonediimine) 11. Ferrimagnetic correlation between the spins of S = 3/2 for [Ru 2 II,III ] + and S = 1/2 for DMDCNQI •− in the I‐phase is produced via an electron transfer from the paramagnetic N‐phase with S = 1 of [Ru 2 II,II ] without dimerization.…”

“…This result implied that the electron transfer from D to A (or from A − to D + ) is also possibly sensitive to the environment in which D and A are located, because this could be associated with the Madelung stabilization in the I‐phase, and not just sensitive to the balance of intrinsic potentials for D/A, i.e., the ionization potential ( I D ) of D and the electron affinity ( E A ) of A. Indeed, the transition temperature ( T c ) is sensitive to the application of hydrostatic pressure and partial chemical modification in the same system 11, 12, 13. These results have motivated the research of chemically switchable N–I phase transitions; as the previous compound represented the only case wherein chains with multiple spin states exhibit the N–I phase transition, new systems have been considered for developing guest‐induced magnetic change (i.e., magnetic sponge) phenomena associated with the N–I phase transition.…”