We have measured polarized resonance Raman spectra of nickel(II) octaethylporphyrin in CS2 and CH2Cl2 solution at different excitation wavelengths (430−580 nm) and temperatures (190−310 K). The analysis of the spectra revealed that the structure-sensitive Raman lines ν19 and ν10 can be decomposed consistently into two sublines for all excitation wavelengths and temperatures. In the resonance region of the QO and QV bands, the 0−1 and 0−0 resonances in the excitation profiles of the low-frequency (LF) sublines of ν19 and ν10 are red-shifted by (150 (30) cm-1 with respect to the sublines that are at higher frequencies (HF). In accordance with experimental and theoretical results, this indicates that the LF sublines of ν19 and ν10 result from a nonplanar conformer, whereas the HF sublines correspond to an almost planar conformer. The existence of this known conformational equilibrium in solution is further corroborated by the van't Hoff behavior of the intensity ratios I LF/I HF of the sublines of ν19 and ν10. From the straight lines in the van't Hoff plot, we calculate that the nonplanar conformer in solution is energetically favored by about 3.0 kJ/mol.

Spiro and co-workers' have measured and compared the complex Raman spectra of the heme group embedded in yeast iso-1-cytochrome c with some of its isotopomers. Their findings lead them to the conclusion that most of the particular strong DPR dispersions earlier observed in our laboratory for the Raman lines v4, v19, and v21 of horse heart ferrocytochrome c2 result from spectral overlap with adjacent lines rather than from symmetry lowering perturbations. To check the reliability of the reported DPR values, we have reinvestigated the polarization properties of the above Raman lines by measuring polarized Raman spectra of ferrocytochrome c at various excitation wavelengths with comparatively high spectral resolution (i.e., 1.9-2.4 ~m -' ) .~ Thus, we reaffiied the earlier reported DPR dispersion for v21 (at 1312 cm-l) and v4 (at 1360 cm-l), whereas the corresponding DPR dispersion of the very intense ~1 9 band (at 1583 cm-l) was found to result in part from an overlap with the adjacent v2 mode. Moverover, it was shown that other Raman lines (Le., v2, v20) also show significant dispersion of their DPR.Hu and Spiro have now published a reply" in which they claim that the spectral crowding in the Raman spectra of ferrocytochrome is too intense to allow a reliable decomposition into single Raman lines. To support their argument for the case of the Raman line 191, they have measured the polarized spectra of yeast iso-1-ferrocytochrome c and its 2,4-di(a-dl)-isotopomer (Le., deuteration of the CH groups attached at the thioether bonds) between 1200 and 1350 cm-l at 11 K with three different excitation wavelengths. Indeed, the spectra of the natural abundance (n.a.) sample taken at 413 nm (B excitation) show that the v21 band overlaps with at least two lines at 1302 and 1317 cm-l. They are absent in the corresponding spectra of the isotopomer and are therefore assigned to the C,H bending modes of the thioether bridges. The DPR value of the isolated 1 9 1 of the above isotopomer is close to 3 (not 5 , as mentioned in ref 4) at 413, 521, and 531 nm. From this observation the authors conclude that the DPR of v21 does not show any significant dispersion.Spiro and c o -w~r k e r s~,~ have performed their measurements at 11 K to obtain a slightly better separation of overlapping bands. It is normally very difficult to obtain correct DPR values from heme proteins in an aqueous solution under this condition, because the scattered light is partially depolarized by multiple scattering in the ice even if one uses a backscattering geometry. This reduces the apparent DPR dispersion signficantly and makes comparison with our data difficult.In order to check the reliability of our previous analysis of the v21 and other relevant Raman bands, we have subjected our spectra to a new line shape analysis using the program MULTIFIT recently developed in our laboratory, which allows better and faster fitting of experimental spectra than the previously used LAB CALC. The natural Lorentzians of the Raman lines have now been convoluted with Gauss...