The systems differ in terms of the structure of the solvation cluster formed around the excess electron.

The existence of the dipole-bound anion of uracil was first predicted theoretically in our group¹ and subsequently detected in gas phase experiments by the groups of Schermann² and Bowen.³

Subsequent experiments concerning anion formation of hydrated molecules of nucleic acid bases⁴ showed that hydration changes the nature of the complex with an excess electron anion and converts the very weakly-dipole-bound uracil anion to a more stable conventional covalent anion.

While for the former a sharp narrow peak at 93 ± 7 meV was observed, the latter produced a broad feature characteristic to a covalent anion.⁴

The broad feature did not show any vibrational structure and was noticeably shifted towards higher electron binding energies than the feature corresponding to the dipole-bound anion of the monomer.

The stabilization of an excess electron attached to a closed-shell system by solvating the anion with polar molecules is an effect that can be easily understood based on the analysis of the electrostatic interactions occurring between the excess electron and the solvating systems.

The transformation of a dipole-bound anion to a covalent anion upon solvation, however, may also involve, apart from the stabilization of the excess electron by the direct interaction with solvating systems, an additional effect resulting from the structural conformation of the solvated covalent anion being more stable than the conformation of the solvated dipole-bound anion.

Our previous calculations of the covalent electron attachment to uracil indicated that such an attachment causes a puckering distortion of the uracil ring.

If solvation stabilizes the puckered uracil structure more than the planar structure of the uracil dipole-bound anion, the covalent anion may become more stable.

The additional stabilization due to solvation may also result from a stronger interaction of the solvent with the more localized covalently-attached excess electron than with the much more diffused dipole-bound electron.

For the [uracil]⁻·water complex, we have already studied one such configuration where the dipole-bound excess electron of uracil was localized between the uracil molecule and the water molecule.⁵

We showed that such a configuration is stable, the presence of water stabilizes the dipole-bound excess electron and makes it more localized.

It is also interesting to investigate structures of possible molecular electron traps formed by molecular complexes.

For example, electron trapping is relevant to molecular charge conductivity and charge localization in such systems as DNA and other biopolymers.

All structures reported in this work are available via e-mail from the corresponding author upon request.

These features differentiate the structure found here from the structure of the uracil·H₂O anion described by Dolgounitcheva et al.⁸

In their structure the water molecule had an in-plane hydrogen bond to one of the oxygens in uracil and its contact with the π excess electron was reduced.

There are a number of ways an excess electron attached to a uracil molecule can interact with two H₂O molecules.

Both the dipole-bound σ excess electron and the covalently attached π-electron have to be considered as possible targets of hydration since in both cases the stabilization effect of hydration should render the excess electrons more strongly attached to the uracil moiety.

One can envision the following configurations of the two H₂O molecules interacting with the dipole-bound excess electron of the uracil anion: a configuration where both waters point the positive ends of their dipoles at the electron; a configuration where an H₂O dimer points its dipole at the excess electron; a configuration where the excess electron is dipole-bound to the complex of uracil with a water molecule and is solvated by the other water; and a configuration where an H₂O dimer is H-bonded to uracil and at the same time solvates the dipole-bound electron of uracil.

An electron covalently attached to a uracil molecule in a π-state can be hydrated by two water molecules in the following ways: two waters can approach uracil perpendicularly to its ring, one on each side of the ring; two waters can independently approach the excess electron from one side of the ring; a water dimer can approach the electron perpendicularly to the uracil ring.

There are, perhaps, some other possibilities.

In our previous studies we have determined that this set of additional orbitals should be sufficient to describe the ground states of the dipole-bound excess electron in systems with dipole moments similar to the dipole moments of the complexes considered in the present work.⁹

By including gaussians with very small exponents in the basis we allowed the excess electron to escape from the system, if such a process would lower the system's total energy.

Thus, the possibility that the excess electron stays confined to the system due to an overly confining orbital basis and not due to higher stability of the anion than the neutral cluster is eliminated.

As mentioned before,¹⁰ in the covalent anion of uracil its ring is noticeably puckered.

This system has been considered in a number of works including, for example, recent studies by van Mourik¹¹ and Gadre et al.¹²

Since the optimizations of the uracil·(H₂O)₂ cluster in the present calculations were only initiated with the geometries of the A–E anions, the equilibrium structures found do not describe the complete topology of the uracil·(H₂O)₂ potential energy surface.

Two series of three calculations were undertaken for each of the two systems.

In the above approach we assumed that the basis set effect does not change much in moving from the anion to the neutral system and, thus all the above calculations could be performed for the neutral forms of the molecules forming anions A and C.

The calculations performed in this work describe an interesting topology of the configurational space of the anion consisting of a uracil molecule and one and two H₂O molecules.

We have only considered structures where either one or both waters are in direct contact with the excess electron attached to the uracil, and for the anion involving a single water molecule we only considered a structure with a covalently attached electron in a π-state because the other isomers of this system had been studied before.⁵

The uracil·(H₂O)₂ system is an interesting model of an electron trap formed by clusters of polar molecules.

We hope the present calculations will inspire further experimental attempts to study the configurational isomerism of anions of hydrated uracil in the gas phase.