Speciation of Ru Molecular Complexes in a Homogeneous Catalytic System: Fingerprint XANES Analysis Guided by Machine Learning

X-ray absorption spectroscopy is a powerful tool for the characterization of local atomic structure. Commonly, bond lengths and coordination numbers are extracted from extended energy region spectrum (extended fine structure, EXAFS). However, many diluted systems, such as homogeneous catalysts, with low concentration active component under in situ or operando conditions, one cannot collect sufficient EXAFS data quantitative analysis. Considering case ruthenium-based catalyst, where ligand surrounding ruthenium atoms can change Br to CO depending on reaction we establish here an effective machine learning approach based descriptor analysis spectral features. After training procedure, algorithm predicts both distances ligands. The prediction quality was verified by means cross-validation procedure applied mixture compounds validated experimental spectra reference RuBr3 [RuBr2(CO)3]2 complexes. This work describes practical route improve classical fingerprint linear combination fit more sophisticated science algorithms.