Because of its central role in basically all aspects of science, water is certainly one of the most extensively investigated substances, from a theoretical point of view. Many properties have been, in fact, theoretically addressed both in the isolated and condensed phases. Nevertheless, many aspects are still not completely understood and represent the focus of active theoretical interest. Among them, one of the most appealing is certainly the understanding of the electronic properties, in particular the photoabsorption features, in condensed phase. Liquid water experimentally shows, under ambient conditions, the 0-1 absorption maximum at 147 nm, that is, 88 kJ mol À1 shifted toward the blue with respect to the corresponding absorption in vacuum. [1][2][3] This blue-shift is known to be more pronounced in ice than in liquid water, [4] and it is also present in small water clusters. [5] From these observations, it is well-established that such a blue-shift is to be mainly ascribed to the short contacts of the excited molecule with its solvation shell (the water dipole moment undergoes an inversion upon 0-1 excitation [6] ). However, only a few theoretical studies have been so far devoted to modelling water photoabsorption in the condensed phase. [7][8][9] The computational methods available nowadays are, in fact, able to provide extremely accurate information about the photoexcitation of isolated molecules. However, there are still many difficulties in modelling the same phenomenon in the condensed phase. The inclusion of electronic degrees of freedom (necessary for studying an electronic excitation) into a simulation of a large number of molecules (necessary for a reliable modelling of a condensed phase) is, in fact, still challenging from a computational point of view. In this context, we recently proposed a theoretical computational approach, the perturbed matrix method (PMM), [10,11] whose main computational feature is the possibility of including, into a classical simulation algorithm, electronic degrees of freedom. In a number
[a] M. Aschi,