Nonaqueous Synthesis of Low-Vacancy Chromium Hexacyanochromate

Abstract

Prussian blue analogues (PBAs) are a highly tunable family of materials with properties suitable for a wide variety of applications. Although their straightforward aqueous synthesis allows for the facile preparation of a diverse set of compositions, the use of water as the solvent has hindered the preparation of specific compositions with highly sought-after properties. A typical example is Cr[Cr(CN)₆]: its predicted strong magnetic interactions have motivated many attempts at its synthesis but with limited success. The lack of control over vacancies, crystallinity, and the oxidation state has prevented the experimental validation of its theoretical magnetic properties. Here, we report the nonaqueous synthesis of vacancy-suppressed, nanocrystalline chromium hexacyanochromate. The control over vacancies and the oxidation state leads to stronger magnetic interactions with a markedly increased absolute Weiss temperature (Θ = -836(6) K) and magnetic ordering temperature of (240 ± 10) K. Our results challenge the notion of the solvent as merely reaction medium and introduce a pathway for exploring moisture- and air-sensitive PBA compositions.

Figure 1

Physicochemical materials characterization. Model structure (A), PXRD (B) and IR transmission spectrum (C) of (1). Model structure (D), PXRD (E) and IR transmission spectrum (F) of (2). Downward facing arrow illustrates chemical oxidation under extraction of Rb⁺. The diffractograms of both materials can be indexed in the Fm3̅M space group. The oxidation state determines the amount of A⁺ ions inside the structure. Cyanide stretch frequencies are sensitive to the materials’ oxidation states. Higher oxidation states cause shifts of the ν(CN) band toward higher wavenumbers, from 2060 and 1997 cm^–1 for (1) to 2186 cm^–1 for (2).

Figure 2

Transmission electron micrographs of (2). Low-magnification imaging shows particle sizes of 10 to 30 nm (A). Continuous lattice fringes over ≈30 nm indicate the particles’ single-crystalline nature (B). The material’s cubic crystal structure is directly evident via imaging in [100] zone axis (C).

Figure 3

Thermal analysis of (2). The material loses 1.1% of its mass likely due to solvent evaporation before a temperature of 180 °C. At this temperature, the DSC signal displays the onset of an exothermic process. The coupled gas analysis registers the evolution of CO₂ at this temperature, which indicates material decomposition. Beyond 300 °C, the decomposition becomes more severe, with rapid sample mass decay and the typical evolution of HCN from the material. Throughout the measurement, the water signal remained roughly constant at background levels.

Figure 4

Neutron powder diffraction of (2). Temperature-dependent neutron diffraction patterns measured at temperatures from 10 to 300 K (A). The dominant magnetic contribution to the diffraction pattern is observed at the (111) reflection position (6.0 Å, red box). Temperature-dependence of the (111) peak intensity extracted using Le Bail fitting tracks the evolution of long-range magnetic order(B). The magnetic contribution vanishes above (240 ± 10) K, marking the magnetic ordering temperature. Neutron diffraction patterns can be refined using magnetic Rietveld fits in the Fm3̅m space group throughout the measured temperature range (C–E).

Figure 5

Temperature- and field-dependent magnetic measurements of (2). Temperature-dependent susceptibility data represented as χ^–1 vs T (A), χ vs T including results from zero field cooled (ZFC) and field cooled (FC) experiments (B) and χT vs T (C). Field-dependent magnetization curve at 5 K deconstructed into linear and saturating parts (D). The increase of χ at 240 K toward lower temperatures is indicative of the onset of magnetic ordering with noncompensation. In the χT vs T curve, a minimum is found at 265 K, revealing that the ordering is of antiferromagnetic type. In the high-temperature region above 270 K, the data is well described by the Curie–Weiss law, as shown in the χ^–1 vs T curve. From saturation magnetization at 5 K, a noncompensation of 0.038 μ_B per formula unit is determined.