1. Water Hexamer Anion Electron attachment and detachment ?

2. Water Hexamer Water hexamer smallest cluster whose local minima have 3D structure Many experimental and theoretical studies, eg. D.C. Clary and J. K. Gregory J Phys. Chem. A 101 6813 1997 Find cage type isomer global minimum For formation of the anion (H 2 O) - n n=6 is a ‘magic number’. It gives a strong peak in the mass spectrum. Others are: n=2,7,11. Also investigated by photoelectron spectroscopy, N. I. Hammer et al. J. Phys. Chem. A 109 7896 2005 Found evidence for surface bound state with AA motif Calculations suggest in smaller water clusters, eg. tetramer AA motif is robust, eg. M. A. Johnson et al. Science, 306, 675, 2004; J.M. Herbert PCCP 8 68, 2006 Water hexamer smallest cluster whose local minima have 3D structure Many experimental and theoretical studies, eg. D.C. Clary and J. K. Gregory J Phys. Chem. A 101 6813 1997 Find cage type isomer global minimum For formation of the anion (H 2 O) - n n=6 is a ‘magic number’. It gives a strong peak in the mass spectrum. Others are: n=2,7,11. Also investigated by photoelectron spectroscopy, N. I. Hammer et al. J. Phys. Chem. A 109 7896 2005 Found evidence for surface bound state with AA motif Calculations suggest in smaller water clusters, eg. tetramer AA motif is robust, eg. M. A. Johnson et al. Science, 306, 675, 2004; J.M. Herbert PCCP 8 68, 2006

3. Neutral Isomers Book Cage Boat Prism Chair

4. Vertical Binding Energies IsomerE/kcalmol -1 VEBE/eV Chair8.010.12 Boat1.87-0.035 Prism0.95-0.031 Book0.65-0.075 Cage0.0-0.075 Neutral Energies on AMEOBA force-field parameterized to reproduce results from MP2/CBS calculations, VEBE = E(neut) -E(an) Anion energies calculated using MP2/6-311(1+,3+)G** Negative VEBE would be unbound in complete BS limit

5. AA Water Hexamer Anion E(neut)/kcal mol -1 12.7 VEBE/eV0.389 Weakly bound water cluster anion Calculations by Hammer suggest that the AA type isomer observed experimentally has the following geometry Weakly bound water cluster anion Calculations by Hammer suggest that the AA type isomer observed experimentally has the following geometry

6. Other Anion Form E(neut)/kcal mol -1 2.33 VEBE/eV0.049 Anion SOMO Does not show any isomerization in 20 AIMD trajectories on the anion PES - each ~1500fs at 50K e - not associated with 1 water molecule - bound by collective action

7. Electron Detachment Assume e - detached with no change in configuration of nuclei Find Minimum Energy Pathway to the nearest local minimum on the NEUTRAL surface Gradient following method - TINKER 12.6 kcal mol -1 0.95 kcal mol -1 0.0 kcal mol -1 AA Open Prism Cage -e -

8. MEP to Open Prism MD simulation at 1K to show change in configuration along MEP of neutral surface Changes within 700fs - fast even at low temperature Little change in configuration

9. Open Prism E(neut)/kcal mol -1 0.95 VBE/eV-0.016 Anion SOMO

10. t/fsVBE/eV  /D 3400.03515.430 3420.03585.443 11400.05005.510 11460.05625.602 11500.05275.550 11510.05115.460 12860.04005.417 12870.04125.425 12880.04105.403 Fluctutions in Dipole Moment Classical MD simulation on neutral surface at 100K using TINKER and AMOEBA force-field shows large fluctuations in dipole moment Standard deviation of dipole moment,  Dipole = 0.71D

11. Snapshots from trajectory 340fs Anion SOMO 1146fs Anion SOMO Ionize with no change in configuration

12. Sudden Approximation ? eV ??? Anion surface Neutral surface +e - Run AIMD simulation on anion surface to find unknown minimum Prism Open Prism

13. Conformational Change From ionized snapshot at 340fs - run AIMD on the anion surface MP2/6-311(+1,+3)G** Would expect AA isomer formation since calculations on tetramer have shown AA isomer robust and evidence from photoelectron spectroscopy

14. ‘Product’ on Anion Surface VBE = 0.29eV SOMO of ??? Not a local minimum Other trajectories show this behavior None show formation of AA type isomer Is this typical? Have we sampled sufficiently? How can we reconcile this with the experimental evidence for the AA configuration? at 230fs

15. Possible Explanations 1)Formation of the AA isomer a comparatively rare event, we need more sampling. Since a 200fs trajectory takes > 1 week at this level of theory this would be difficult. Perhaps a MODEL HAMILTONIAN approach such as the Drude model of Jordan et al. could reduce the cost. 2) Formation of the AA state goes through the cavity bound state as an intermediate 3) AA formation takes place through an isomer which has fewer initial hydrogen bonds, eg. linear or chair

16. Summary Low energy structures in neutral due to much hydrogen bonding Electron capture can occur in these due to dipole fluctuations So far AIMD calculations show the formation of cavity bound states No AA type formation observed. Is this as a result of insufficient sampling? Model Hamiltonian type calculations needed to increase efficiency of AIMD calculations Low energy structures in neutral due to much hydrogen bonding Electron capture can occur in these due to dipole fluctuations So far AIMD calculations show the formation of cavity bound states No AA type formation observed. Is this as a result of insufficient sampling? Model Hamiltonian type calculations needed to increase efficiency of AIMD calculations

17. Acknowlegements I, for one, welcome our new Overlords John Herbert Leif Jacobson Adrian Lange OSC Ohio State University