New Features Observed in Self-Absorption-Corrected X-ray Fluorescence Spectra for Ni Complexes with Uncertainties.

We present a new technology for analyzing the molecular structure and in particular subtle conformational differences in Ni complexes using X-ray absorption spectroscopy (XAS), enabling tighter and more robust constraints of structure and dynamic bond lengths. Self-absorption and attenuating effects have a large impact in fluorescence X-ray absorption spectroscopy (XAS), compromising accuracy and insight in structural and advanced analyses. We correct for these dominant systematic effects. We investigate nickel(II) complexes, that is, bis(N-n-propyl-salicylaldiminato) nickel(II), "n-pr", and bis(N-i-propyl-salicylaldiminato) nickel(II), "i-pr", in 15 mM solutions with 0.1% w/w Ni. One is "square-planar" and one is "tetrahedral", with identical coordination numbers. We identify two key sources of uncertainty and provide robust estimates for them, reflecting the quality of the data, and provide meaningful estimates of χr2 suitable for hypothesis testing. We apply significance and model testing for fluorescence data, with direct uncertainty estimates. Two new peaks are revealed in the X-ray absorption fine structure (XAFS) at k ≈ 4.4 and 5.4 Å-1. The high intrinsic accuracy of our processed data allows these features to be well modeled and yields deeper potential insight. Three important notions in the field are addressed: resolvability of shell radii, estimation of the number of independent data points in least-squares or Bayesian analysis, and the effect of uncertainties on the determined structure and the determinability of key structural parameters. Conventional XAFS fitting requires a kmin and a kmax. The origin of these limits is explained from the data, in a quantitative manner. Being able to distinguish the isomers spectroscopically and structurally places strong demands on the data, the uncertainties, and the model interpretation, and this article reports success in this subtle structural identification. Two nearby shells-the innermost two shells-are identified quantitatively, well below the conventional aliasing limit. This illustrates the application of new technology to gain new insight.