No abstract

No abstract

No abstract

We describe an ab initio calculation of the properties of energy loss by electrons in crystalline water using its dielectric response function, e(q), ?), where q and ? are, respectively, the wave vector and frequency. The calculation was performed on a model system (cubic ice) in order to take advantage of its ordered structure (i.e. Bloch's theorem), but also because of evidence that liquid water in biological systems ('structured' water) contains residues with tetrahedral structure (i.e. ice) over time scales of at least 10-11 s. The main features of the calculation are: (a) e(q,w) is evaluated in the random phase approximation (we used the expression given by Ehrenreich and Cohen), (b) the crystal potential is expressed as a sum of water-molecule self-consistent potentials, and (c) wave functions are expanded using tight binding functions (ultimately employing a Gaussian base set). A total of seven states (bands), five occupied and two conduction, are considered. We report the band structure and the density of states of the crystal, as well as values of e(q,w) at selected values of q and w. Results are compared with energy loss measurements and with absorption spectra (XPS, UPS, and VUV data). The possibility of using an empirical combination of molecular potentials as a phenomenological Hamiltonian is also examined.

Understanding the effects of low doses of ionizing radi ation on living systems requires detailed information on electron transport in biomaterials. This, in turn, can be obtained from the wave-vector- and frequency-depen dent dielectric response function of the system, ∈(q,ω), via the energy-loss function, Im[-1/∈( q,ω)]. We describe two different possible approaches to obtaining these functions, one based on the semiempirical tight-binding approximation, the other using Hedin's many-body treatment of quasiparticle states in solids. These methods are exemplified with calculations for cubic ice (as a model for cellular "structured" water) and trans- polyacetylene. The availability of supercomputers makes the application of these techniques feasible.