A complete set of vibrational spectra obtained from several spectroscopic techniques, i.e., neutron inelastic scattering (NIS), Raman scattering, and infrared absorption (IR), has been used in order to assign the vibrational modes of uracil on the basis of an ab initio scaled quantum mechanical (SQM) force field. NIS, Raman, and IR spectra of polycrystalline uracil recorded at T ) 15 K from native and N-deuterated species provide complementary data for analysing different groups of molecular vibrational modes. Solid-state spectra have been completed with various Raman (λ exc ) 257 nm, 514.5 nm, and 1.06 µm, and IR spectra in aqueous solutions. Both phases allowed effects of the environment on the vibrational modes, related to either strong (crystal) or weak (solution) hydrogen bondings, to be shown. In addition, the various laser excitations allowed the wavelength dependence of the Raman cross sections of in-plane characteristic modes to be observed. Due to the large NIS incoherent cross section of protons, the intense NIS bands are those arising from the vibrational modes containing hydrogen motions. The molecular fundamental wavenumbers calculated at the SCF+MP2 level, by using different types of molecular orbitals, have first been compared with the experimental wavenumbers taken from gas phase or Ar-matrix isolated uracil. Then the force field has been scaled in order to improve the agreement with experimental data from solid and aqueous phases. In the scaling procedure, the standard Pulay method was reliable for the in-plane vibrational modes, whereas it failed to scale successfully the out-of-plane vibrational modes. Consequently another scaling method, which consists of refining the nondiagonal elements of the internal force field matrix, has been used. On the basis of this procedure for out-of-plane modes, the simulation of the NIS intensity related to the C-H wagging motions could be performed without any particular difficulty. However some difficulties still exist for the N-H wagging motions which are largely perturbed by hydrogen bonding and packing effects in solid phase.

This work validated a SERS probe able to compare adenyl reactivity in DNA and RNA. A Creighton silver colloid including adenine (A) [or 2 -deoxyadenosine 5 -phosphate (pdA)] from 2 × 10 −3 to 2 × 10 −8 M is stabilized in the absence or presence of chloride. Concentration-dependent SER spectral profiles reveal how A may interact with (Ag) n + sites. At concentration ≥2 × 10 −5 M adsorption of (A) n clusters prevents the colloid from undergoing salt effects. Adsorption via N1/N3 is allowed whereas C6NH 2 is involved in self-association. At [A] <2 × 10 −5 M with chloride, hydrogen bonding between chloride and the C6NH 2 group enhances C6N electronegativity, which assists C6N/N7 cooperative adsorption. Complex A(Cl − ) entities compete with individual chloride ions for adsorption on silver. Very similar C6N/N7 adenyl adsorption occurs for pdA but only above 2 × 10 −5 M. Chloride-adenyl bonding is reduced and pdA self-association is weaker than adenine self-association. Steric factors, repulsive electrostatic forces and phosphate competitive reactivity with respect to chloride may explain the much higher pdA concentration needed to saturate the silver surface compared with A. Mg 2+ -phosphate complexation entails concentration-dependent opposite effects on adenyl reactivity with (Ag) n + sites. Cytosine, thymine and guanine base or corresponding nucleotides deliver weaker SER spectra and much higher SERS responses for chloride adsorption compared with A or pdA. This reveals a weaker adsorption of the oxo bases, assumed to result from alternative oxo and nitrogen interactions with the (Ag) n + sites.