Phase-sensitive experiments

One problem with the gradient approach described above for COSY is that it necessarily precludes the acquisition of high-resolution phase-sensitive data sets by selecting only one pathway in ti. Since this leads to phase-modulated data it provides only phase-twisted lineshapes. Recall that phase-sensitive ac- 

---

In Section IV we quantify the relation of the information-rich phase of the scattering wavefunction to the observable 8(E) of Eq. (5). Here we proceed by connecting the two-pathway method with several other phase-sensitive experiments. Consider first excitation from g) into an electronically excited bound state with a sufficiently broad pulse to span two levels, Ea and ),... 

For 2D phase sensitive experiments a similar editing technique is straightforwardly implemented in the same way. However, because of the wide range of V( Sn, H) coupling constant values even more drastic phase problems appear. The resulting lineshape distortions usually obscure the spectral information desired for structural purposes. Accordingly, 2D spectra transformed in the magnitude mode usually fulfil spectroscopic requirements. 

Since the pulse sequence is the same for EXSY and NOESY, NOESY (or ROESY) cross peaks might be mistaken for EXSY cross peaks. They can be distinguished in the phase-sensitive experiment, since EXSY and ROESY peaks have opposite phases, as do EXSY and NOESY peaks in the fast motion regime. For example, two resolved OH or NH resonances may exhibit EXSY cross peaks from slow proton exchange. These peaks could be mistakenly taken to be NOESY peaks and interpreted incorrectly in terms of stereochemistry. 

It is also important that LP not be abused. A sufficient number of increments must be taken from which the FID s can confidently be extended. A total of 64 increments has, for example, been found to be insufficient, while LP s have successfully been carried out with 96 increments. A good practice is to acquire at least 128 increments for accurate prediction. A second concern is that LP not be extended too far (e.g., 128 points predicted to 4,096). W. F. Reynolds (2002) has found that, as a general rule, data presented in the phase-sensitive mode can be predicted fourfold (e.g., 256 data points can be predicted to 1,024), while absolute-value data can be extended twofold, 256 points to 512. A significant exception to the fourfold rule for phase-sensitive experiments concerns the H-detected, heteronuclear chemical-shift correlation experiments. In marked contrast to COSY and HMBC spectra, for which the interferograms are frequently composed of many signals, those of HMQC and HSQC spectra constitute only one (due to the directly attached C). LP s up to sixteen-fold can be performed in these experiments (Sections 7-8a and 7-8b). 

Quadrature detection As described above, some care is required when employing gradients in phase-sensitive experiments. This either requires that gradients are not placed within time domain evolution periods or that the echo-antiecho approach be used. 

The comparison of 2D spectra is often simplified by the decomposition or splitting of a 2D data matrix into a series of ID spectra. Time domain data can be used to optimize weighting functions prior to processing the 2D data matrix whilst frequency domain data can be used in the evaluation and development of modified pulse sequence. ID spectra can also be used to optimize the phase correction in a phase sensitive experiment. 

The HMQC sequence does not necessarily have to be based on the coherence transfer step shown in the sequence scheme above. The coherence transfer can also be generated by a DEPT element which has the advantage in a phase sensitive experiment of labelling the signal phase according to the multiplicity of the IH spin groups. 

A wide range of proton and carbon NMR techniques, almost exclusively using Fourier Transform (FT) spectroscopy, are presented in applications to support proposed chemical structures, with multinuclear NMR being used where appropriate. Polarisation transfer experiments (e.g. DEPT) are used to indicate carbon multiplicity and the use of two-dimensional techniques, which facilitate signal assignment, is seen more frequently in applications. Typical techniques include H- H correlation (COSY and related phase-sensitive experiments) proton-carbon heteronuclear correlation (including inverse... 

---