A series of Pd/g-C₃N₄ catalysts were synthesized using different graphitic carbon nitride precursors, and were found to exhibit significant variations in catalytic performance for solvent-free selective catalytic oxidation of benzyl alcohol. Through comprehensive characterization (XRD, N₂-BET, ICP-AES, TEM, and XPS), the experimental results found that the nitrogen chemical configuration and surface Pd²⁺concentration critically determined the catalytic efficiency. Among the various nitrogen species, N-(C)₃-type nitrogen demonstrated the strongest influence on catalytic activity, which was positively correlated with its abundance in g-C₃N₄ matrices. In particular, g-C₃N₄ derived from dicyandiamide contained the highest N-(C)₃ type nitrogen content. When serving as Pd nanoparticle support, this material simultaneously achieved the best Pd²⁺ surface concentration and catalytic performance compared to the other g-C₃N₄-supported catalysts.

The isobaric vapor-liquid equilibrium data for the four systems consisting of 1-octene + 2-methyl-thiophene, cyclohexene + 2-methylthiophene, 2,3-dimethyl-1-butene + 2-methylthiophene, and 2,3-dimethyl-2-butene + 2-methylthiophene were determined by using a modified Rose-Williams equilibrium still at 90.00 kPa. All the vapor-liquid equilibrium data measurements passed the thermodynamic consistent test by employing the Herington method. The azeotropic behavior was observed in the 1-octene + 2-methyl-thiophene system. Additionally, the experimental data were correlated by the Wilson model, and they were also compared with the original UNIFAC and UNIFAC Dortmund predictive models. All the results indicated that the measured data were in good agreement with the calculated values obtained from the Wilson. Meanwhile, the UNIFAC Dortmund model can provide better prediction result than the original UNIFAC model.