Knowledge Bank

We have recorded the negative ion photoelectron (photodetachment) spectra of the gas-phase negative ions $N2O-,CO2-,(N2O)2-$ and $(CO2)2-$. Nitrous oxide and carbon dioxide are isoelectronic molecules whose negative ions have bent geometries. Due to the large structural differences between these ions and their linear neutrals, appreciable Franck-Condon overlap is not expected during photodetachment transitions between the lowest lying levels of these negative ions and their neutrals. The photoelectron spectra which we obtain for $N2O-$ and $CO2-$ are consistent with this expecttion. Both spectra contain a single broad peak which we interpret primarily as a progression in the bending mode of the neutral. Some structure is visible on this peak in both spectra, it being more pronounced in the $CO2-$ spectrum. The maxima of these spectra correspond to electron binding energies of $\sim 1.5eV$ for $N2O-$ and $\sim 1.4eV$ for $CO2-$. These values are interpreted as the vertical detachment energies of $N2O-$ and $CO2-$. In contrast to the similarities between the photoelectron spectra of $N2O-$ and $CO2-$, we find substantial differences between the spectra of $(N2O)2-$ and $(CO2)2-$. The photoelectron spectrum of $(N2O)2-$ is similar to that of $N2O-$ in that it consists of a single broad peak. Its maximum, however, is shifted by $\sim 0.2eV$ to lower electron kinetic energies relative to the maximum in the $N2O-$ spectrum, and this represents an increase in the electron binding energy by a corresponding amount. We interpret the $(N2O)2-$ spectrum as arising from the photodetachment of an ionic species which is best described as a bent $N2O-$ ion solvated by a neutral linear $N2O$ molecule, i. e. $N2O-(N2O)1$. The 0.2 eV shift is a rough measure of the ion-solvent dissociation energy of $N2O-(N2O)1$ into $N2O-$ and $N2O$. By contrast, the maximum in the $(CO2)2-$ photoelectron spectrum has been shifted out of our spectral range. This indicates that the electron binding energy of any maximum in the $(CO2)2-$ spectrum is larger than that for the maximum in the $CO2-$ spectrum by at least 1 eV, and that there are substantial differences between $(CO2)2-$ and $(N2O)2-$. While these observations do not definitively distinguish between a simple solvated-ion vs. a more chemical ``oxalate-like"" bonding picture for $(CO2)2-$, they do suggest that if $(CO2)2-$ is a solvated ion, the interaction energy between $CO2-$ and $CO2$ is $>1eV$.