Peak analysis position

Figure 10.4. MGZ analysis of an undrawn polyamide 6 fiber, (a) Plot of the peak maximum position as a function of the detector coordinates (x,y). Relations to the parameters of the ellipse and the structure (long periods Lm, Lq) are indicated, (b) Separation of meridional long period, Lm, from the long period L . (courtesy S. MURTHY)...

Peak by peak analysis is widely used in X-ray diffraction and significant developments in digital calculations have made it possible to define very precisely the actual peak profiles. It sometimes happens, however, that the peaks overlap too much for it to be possible to separate their contributions. It then becomes necessary to provide additional information which makes it possible to correlate either the relative positions of the different peaks or their intensities. This is referred to as whole pattern fitting. 

Since the user will never be totally satisfied with automatic analysis all the time, it is very important to allow him to assume control over all parts of the quantification whenever necessary. For example, the peak analysis routine of FSUNMR performs an automatic quantification by default. The user is given a wide range of options when the computations are complete. These include insertion and deletion of peaks, manual declaration of a threshold, peak position calibration, and a complete set of display options. Good peak analysis routines can save the spectroscopist a lot of time without... 

The range of relaxation times allowed in the fitting was usually between 0.5 ps and 1 s with a density of 12 points per decade. Relaxation rates are obtained from the moments of the peaks in the relaxation time distribution or, if the peaks overlap, from the peak maximum position. With a broad distribution of relaxation times, these inversion methods yield multiple peaks in the "unsmoothed" analysis. The "smoothing" parameter (P) was selected as 0.5 in all cases, after it was established that the number of peaks did not increase with further increase in smoothing. As a further check, an analysis was made on a simulated correlation function consisting of a broad continuous distribution of relaxation times with noise added equal to the residuals from the analysis of the experimental correlation curve. REPES recovers the original distribution except when a very low smoothing parameter (P 0) is used. 

---