Spectroscopy by integration of frequency and time information (SIFT) and fast acquisition of dark spectra (FADS)

PHYS 243

Yoh Matsuki and Judith Herzfeld, herzfeld@brandeis.edu. Department of Chemistry, Brandeis University, Waltham, MA 02454-9110
Amplitudes in the time and frequency domains are linearly related and knowledge of amplitudes in one domain provides constraints on amplitudes in the other. In particular, knowledge of amplitudes at N frequency points provides a set of N linear equations that obviates the acquisition of amplitudes at N time points. Here we demonstrate how the integration of frequency and time domain information can reduce the amount of time information that needs to be acquired in FT-NMR. In particular, we make use of the fact that signals tend to cluster in regions of the spectrum, leaving large expanses predictably bare (or “dark”). This information can be used to compute amplitudes in the time domain either analytically or by numerical iteration, with no assumptions or adjustable parameters. The computational “SIFTing” of non-uniformly sampled data is very rapid and preserves chemical shifts and relative signal intensities.