Vanadium nitride (VN_x) powder was synthesized by heating a compact of V₂O₅ and BN powders with Na metal at 600–1100°C under an Ar atmosphere. The process proceeded via a metathesis reaction using BN as the nitrogen source, accompanied by the formation of Na₃BO₃ as a by-product (3V₂O₅ + 5BN + 15Na → 6VN_0.83 + 5Na₃BO₃). For the samples heated at or above 900°C, nearly single-phase VN_x powder was obtained after washing with water to remove the by-product. Rietveld analysis and elemental analysis revealed that the product was nitrogen-deficient VN_x with x = 0.79–0.88, and the particle size increased from 0.5–1 µm (900°C) to 2–5 µm (1100°C) with increasing heating temperature. The molten Na₃BO₃ by-product acted as a flux, promoting both the formation and grain growth of VN_x. Furthermore, reacting a V₂O₅ compact with a BN crucible at the contact interface produced coarse, octahedral euhedral crystals of VN_x (10–50 µm). This method allows for the synthesis of VN_x powders and crystals at lower temperatures than conventional methods and is effective for controlling both nitrogen composition and particle size.

Binary phase diagrams of Group 4 (Ti, Zr, Hf), Group 5 (V, Nb, Ta), and Group 6 (Cr, Mo, W) transition metal carbides were calculated using the CALPHAD method with FactSage and Thermo-Calc software. The calculated results were compared with classical experimental phase diagrams compiled by Suzuki, referencing data from Storms and Rudy. The calculated diagrams showed good agreement with experimental data regarding liquidus shapes and invariant reaction temperatures, such as eutectic and peritectic points, across all systems. However, discrepancies were observed in the homogeneity ranges of MC-type carbides (Group 4 and 5); the calculated single-phase regions tended to be narrower and shifted toward stoichiometric compositions compared to experimental reports. Additionally, differences were noted in the stability of specific phases, such as ζ-phases in V-C and Nb-C systems, which are often stabilized by impurities in experiments, and the treatment of certain carbides as line compounds in databases. These results suggest that while thermodynamic databases are powerful tools for predicting phase equilibria, careful consideration of solid solution ranges and phase modeling assumptions is necessary for material design. Furthermore, future challenges, including improving homogeneity range accuracy, evaluating impurity effects, and refining phase modeling for order-disorder transformations, are identified to enhance database reliability.