Nmr Data Processing, Automatic Phase Correction And Implementation

Nowadays, Nuclear Magnetic Resonance is used widely in Chemistry, Physics, Biology etc as one of the most powerful tools of structure characterization. Since the result of FFT of FID rarely exhibits the desired pure absorption line, phase correction of FT NMR spectra becomes a necessity in NMR data processing. Though manual phase correction can give satisfying results, the requirement of automatic process for NMR spectra become higher and higher with the gradually popularizing of auto batch test and multi-dimensional NMR.Thus automatic phase correction comes up in this way and becomes a focus. A number of methods for automated phase correction, such as integration, entropy minimization, line shape analysis of DISPA etc, have been proposed with the goal to replace the conventional, manual approach. In this thesis, some typical ones are implemented with C++, then a series of experiments are carried out to test them. Further, several original methods of automatic phase correction are proposed, implemented and tested thoroughly, including peak curve fitting, minimization of peak regularity, minimization of baseline concussion and connecting line of equally spaced points. Each method has its own advantages in one or more aspects concerning precision, speed, correctable phase errors, insensitivity to defects in signals such as peak overlap, incomplete peak, noise, peak asymmetry, baseline distortion etc.A novel method for automatic phase correction of NMR spectra based on DISPA circle, namely Connect Line of Equally Spaced Points (CLESP), is described in detail and proves to be the best for gerneral use. In this method, phase angles of peaks are determined from the slopes of the connect lines of two peak points with equal data point density on each side of the peak. The calculated phase angles of peaks are then used to extract the zero-order and first-order phase correction of the whole spectra by means of linear regression. In addition to the advantages of being non-iterative, fast, accurate, straightforward, easy to implement, and applicable to partial peaks, CLESP is proved to be robust with respect to noise, baseline distortion, peak overlaps, and poor digitization of NMR signals. CLESP algorithm has been successfully implemented in our home-made NMR data-processing software ECNMR.There is still much work to do to further put forward the method of automated phase correction. Several potentials are also discussed in this thesis.