The overall purpose of this study was to investigate the effects of the magnesium and lithium precursors and the catalyst surface properties on the catalytic performance in the oxidative methylation of acetonitrile. The performance of the catalysts for the oxidative methylation of acetonitrile to acrylonitrile was significantly affected by the Ii precursor, where catalysts prepared with LiCl and LiOH on MgO had the best performance for this reaction. The catalysts＇ activity was virtually unaffected by the source and surface area of the MgO. In contrast, relatively high BET surface areas appeared to hare a negative affect on the catalysts＇ performance by producing less acrylonitrile and more COx. XPS and XRD analyses of Li/MgO-based catalysts indicated that the lithium salts used in the catalyst synthesis predominantly formed mixtures with MgO. It was found that the Li : Cl atomic ratio was 1 : 1 in catalysts prepared with LiCl, even after calcination at 650 degrees C for 14 h in air. XRD analysis of LiCl/MgO catalysts calcined at 650 degrees C suggests that some amount of LiCl precursor on MgO decomposes into Li+O- (Li2O2). The latter species are believed to be responsible for the effective transformation of methane and acetonitrile to acrylonitrile. The binding energy of Ii Is ranges between 57.0 and 52.3 eV. Very interestingly, the binding energy of the Li Is peak observed at 57.0 eV in LiCl/MgO (C1-C6) and LiOH/MgO (C9) catalysts corresponds to Li+O- (Li2O2). This peak was negligible for Li2CO3/MgO (C7 and Cs) and LiNO3/MgO (C10) catalysts. Our proposal for this XPS peak of Li+O- species is supported by the EPR peak at g(perpendicular to) = 2.0544. To the best of our knowledge there is no previous communication of the Ii Is peak which corresponds to Li+O- determined by XPS. Catalytic experiments for the oxidative methylation of acetonitrile to acrylonitrile over the aforesaid catalysts indicate that indeed the catalysts synthesized with LiCl and LIOH promote the desired reaction, or equivalently they are associated with Li+O- species. A direct relation between the EPR peak at g(perpendicular to) = 2.0544 and the lack of CO was established. Although the LiCl/MgO catalysts＇ performance was better than that of other catalysts prepared in this study, it was also the least stable. Observations made during these studies suggest that lithium sublimation occurs, thus resulting in catalyst deactivation. It is remarkable to note that catalysts synthesized with LiNO3 demonstrated superior thermal stability, even under severe calcination conditions. (C) 2000 Academic Press.