40Analysis of the Experimental Dataand one polynomial (5th degree). The obtained values with the two different baseline corrections differ considerably strong for the integrated intensities for all bands. The band width is less influenced (exception: the small band around 280 cm-1 which is hardly recognizable) but the peak position is nearly unaffected. To be able to perform a comparative analysis of the spectra it was necessary to use the same baseline correction scheme for all spectra.(a) Lineal Baseline correctionBand position (cm-1) 74 104 293 483 502 Intensity (a. u.) 5493 11302 620 6368 16595 Width at half Integrated maximum (cm-1) intensity (a. u.) 20 50402 46 241056 29 8631 40 125341 14 120369(b) 5th degree Polynomial Baseline correctionBand position (cm-1) 73 103 280 486 502 Intensity (a. u.) 2798 4851 2344 6066 16124 Width at half maximum (cm-1) 16 42 67 35 13 Integrated intensity (a. u.) 21018 95270 74191 110349 108626Figure 4.6: a) Lineal and b) Polynomial baseline subtraction and fitting of the resulting bands of the liquid NdI3(0.5)-NdI2(0.5) spectrum, in the quartz cell. The spectrum was recorded with the Ar laser (514.5 nm), 4.05 notch filter angle, without polarizer and an integration time of 120 s. The presented data are just an approximation; for this reason, the error for each date does not appear in this table.After this correction, we were able to fit the peaks using different band shapes, like Gaussian, Lorentzian, or a combination of both. For this objective, we used the Origin庐 software package. An example is given in Fig 4.7.
