Quantitative use of electron energy-loss spectroscopy Mo-M2,3 edges for the study of molybdenum oxides.

Because of the large energy separation between O-K and Mo-L2,3 edges, extracting precise and reliable chemical information from core-loss EELS analyze of molybdenum oxides has always been a challenge. In this regard Mo-M2,3 edges represents an interesting alternative as they are situated close to the O-K edges. They should allow thus the extraction of a wealth of chemical information from the same spectra. However the main difficulty to overcome in order to work properly with these edges is the delayed maxima of the Mo-M4,5 edges which hinders the automated background subtraction with the usual inverse power low function. In this study we propose another background subtraction method specifically designed to overcome this obstacle and we apply it to the study of MoO3 and MoO2. We are able to show that quantitative chemical information can be precisely and accurately determined from the joined analyze of O-K and Mo-M2,3 edges. In particular k-factors are derived as a function of the integration window width and standard errors close to 2% are reported. The possibility to discriminate the two oxides thanks to chemical shifts and energy-loss near-edge structures is also investigated and discussed. Furthermore the M3/M2 ratios are derived and are found to be strongly dependent on the local chemical environment. This result is confirmed by multiplet calculations for which the crystal field parameters have been determined by ab initio calculations. The whole methodology as well as the conclusions presented in this paper should be easily transposable to any transitions metal oxides of the 4d family. This work should open a new and easier way regarding the quantitative EELS analyses of these compounds.