Comment on ``Raman spectroscopy study of Na x CoO 2 and superconducting Na x CoO 2 ∙yH 2 O ''

The effect of surface degradation of the thermolectric cobaltite on Raman spectra is discussed and compared to experimental results from Co3O4 single crystals. We conclude that on NaCl flux grown NaxCoO2 crystals a surface layer of Co3O4 easily forms that leads to the observation of an intense phonon around 700 cm-1 [Phys. Rev. B 70, 052502 (2004)]. Raman spectra on freshly cleaved crystals from optical floating zone ovens do not show such effects and have a high frequency phonon cut-off at approximately 600 cm -1 [Phys. Rev. Lett 96, 167204 (2006)]. We discuss the relation of structural dimensionality, electronic correlations and the high frequency phonon cut-off of the thermolectric cobaltite.