Experimental verification of SCS calculated rates

The simplicity of SCS is a great advantage when the problem becomes 

The full E-dependence of t (E) has to be found to calculate each of them from Eq. (5.45) and the huge matrix of such constants must be computed 

Since good resolution of the Q-branch is hardly achievable by means of the usual Raman spectroscopy the first verification of this formula was carried out on side branches of anisotropic spectra which are easier to resolve (see Fig. 0.2 and Fig. 3.1). Generally speaking the right formula for component widths of these branches must be separately derived [212] but approximate estimation for the S-branch may be done as proposed in [213]  

It is reasonable when both vibrational and rotational dephasings are negligible. Using this approximation SCS estimations of the rate coefficients of line broadening were compared in [191] with the experimental y -dependence of kj in the S-branch of ty-Ar mixture obtained in [214], 

When calculating the rate constants, two potentials were used the anisotropic 6-12 Lennard-Jones from [209] and the anisotropic Morse [216] for comparison. The results appeared to be very similar, thus indicating low sensitivity of the line widths to the potential surface details. The agreement with experimental data shown in Fig. 5.6(h) is fairly good. Moreover, the SCS approximation gives a qualitatively better approach to the problem than the purely non-adiabatic IOS approximation. As is seen from Fig. 5.6 the significant decrease of the experimental line widths with j is reproduced as soon as adiabatic corrections are made [215]. 

---