“…Actually, in a recent paper, 22 we studied the influence of T as well as the fine size effects on similar samples by deconvoluting the O-H stretching region into four symmetrical profiles in accordance with an interesting model proposed by Walrafen et al 14,15 for bulk H 2 O some years ago. Essentially, the interpretation formulated in our previous work has been partially used here, but with a further and deeper refinement of the aforementioned subbands assignment, having added a new contribution at ¾3020 cm 1 , properly discussed in the framework proposed by Li et al 23 Hence, we best-fitted all the spectra, starting from bulk water passing to the confined sample, both in the case of native and silanised GelSil glass, into five Gaussians, each of them assigned to a particular O-H vibration related to the specific interaction with the absorbing silica substrate. In Fig.…”