NMR Processing: First Point Correction

 

The Fourier-transformation is suited for dealing with periodic functions, but has problems with discontinuities, e.g., the first point of the FID can be considered as a sudden raise in intensity when coming from the end of the FID (with about zero intensity). The effect is to produce a baseline offset of the spectrum that is proportional to the total integral of the spectrum (the first point of the FID represents the total integral of the spectrum) [5].

Note: a constant offset of all points of the FID causes a center spike of one point exactly in the middle of the spectrum (see DC-offset). In contrast, one point in the FID affects all points in the spectrum (here a spectrum offset).

If the first point of the FID is replaced by the average of the first and last point, which usually corresponds to about half the initial data value (FCOR = 0.5), this discontinuity is minimized. Here is an example of a non-digitally filtered dataset that has been processed (with ft only) with first-point correction (FCOR = 0.5) or without (FCOR = 1.0)
 

An analogon can be found in image processing, called Gibbs phenomenon, where the data reduction of the JPEG algorithm produces artefacts in areas of high contrast differences, e.g., in black-and-white line drawings:

For 1D NMR spectra, this baseline offset is not really a problem, because a baseline correction is usually part of the processing, especially if integration is planned. In 2D spectra, however, this offset may vary from row to row in the indirect dimension, thereby creating t1-ridges [1,3]:

Digitally-filtered FIDs (DIGMOD = digital) start with the digital filter function that basically looks like a reversed FID. Therefore, there is no discontinuity. However, if they are part of a 2D or 3D experiment, a correction could be necessary in the indirect dimensions. Occasionally, comments in pulse programs may give a hint. In many cases, the ridges can also be reduced by abs2 followed by abs1.

Literature:

1. Hoffman, Levy, Progr. NMR Spectrosc. 1991, 23, 211; Doi: 10.1016/0079-6565(91)80005-M
2. Otting, Widmer, Wagner, Wüthrich, J. Magn. Reson. 1986, 66, 187; DOI: 10.1016/0022-2364(86)90122-8
3. Pelczer, Szalma, Chem. Rev. 1991, 91, 1507; DOI: 10.1021/cr00007a012
4. Freeman, A Handbook of Nuclear Magnetic Resonance, Longman Scientific & Technical, Burnt Mill, UK, 1988
5. Hoch, Stern, NMR Data Processing, Wiley, New York, 1996, ISBN 0-471-03900-4