---

Excited State Solvation Dynamics

Nonequilibrium, experimental techniques may be combined with ab-initio molecular dynamics. One example for this is local dielectric spectroscopy, where the nonequilibrium situation is generated by photoexcitation of probe molecules. Experimentally the evolution of the emission wavelength of a fluorescent dye can be used to monitor its excited state solvation dynamics. Indirectly this gives access to the local THz spectrum around the fluorescent probe molecule. Ab initio molecular dynamics (AIMD) can simultaneously provide solute/solvent dynamics in the electronically excited state and provide the fluorescence wavelength via TD-DFT calculations.

As the experimental timescale is in the picosecond range, the experiment may be directly simulated by AIMD. In one example, we have simulated the relaxation of water around the probe N-methyl-6-oxyquinolinium betaine (MQ). The connection between MQ's experimental TDSS data and the local THz spectrum of water has been established by simple dipolar continuum theory. Our simulations confirm, that the dipole moment of MQ is drastically reduced upon electronic excitation. The experimental Stokes shift could be reproduced, then decomposed into molecular components (see Fig. 1), with the water molecules in the dipole field having a significant contribution to the relaxation. We found the probe molecule to be influenced by more than the first solvation shell, thus giving experimentalists information about the range and region of their measurements.

Figure 1: The first solvation shell of MQ and its components: Water molecules in torus around N-O (yellow beads), H2O molecules hydrogen bonded to the MQ-oxygen terminus (beads), and entire first solvation shell (sticks)

In various systems, acidic properties emerge when the system is electronically excited. Although the time scale attributed to the electronic excitations is usually on the order of femtoseconds, these processes can, in general, trigger much slower processes. Having several time scales at play makes it extremely hard to theoretically study the triggered processes. Electronic excitations can be studied using various methods, such TD-DFT. However, even in the case of TD-DFT, reaching long simulation times for proper sampling of nuclear motions is practically tedious, especially when studying systems in the condensed phase. On the other hand, computational methods which facilitate long simulation times, such as DFT-based AIMD, usually lack a proper description of the electronic excited states. Recently, we have presented a method[5] with a final goal of increasing the computational efficiency of DFT-based calculations of the excited states. We focus on obtaining proton transfer energy at the S1 excited state through actual density functional theory calculations at the T1 state with additional optimized effective potentials (Fig. 2). The potentials are optimized as such to reproduce the excited-state energy surface obtained using TD-DFT and equation of motion coupled cluster methods, but can be generalized to other more accurate quantum chemical methods. The presented method is not only suitable for studies on excited-state proton transfer and ion mobility in general systems but can also be extended to investigate more involved processes, such as photo-induced isomerization.

Figure 2: Ground-state gas-phase structures of phenol+water (a) and 7HQ+water (b) complexes optimized at the DFT level with ωB97X-D3 XC functional. Also shown are the OH dissociation curves for (c) the phenol+water [Figs. 2(a)] and (d) 7HQ+water [Fig. 2(b)]. Solid triangles (green crosses) show the energy values calculated at the DFT (TDDFT) level using ωB97X-D3 functional at the S0 (S1) state. Hollow triangles represent the values obtained using the BLYP functional at the corresponding T1 states without using extra effective potentials (T1/orig.). Solid circles in blue and red show the energies obtained at the T1 state using the optimized effective potential parameters (T1/opt.) for C, N, O, and acidic H. Gray rectangles in (a) show the energies obtained after optimizing the effective potential parameters for C and O only.

References:

[1] C. Allolio, M. Sajadi,  N.P. Ernsting and D. Sebastiani, “An Ab Initio Microscope: Molecular  Contributions to the Femtosecond Time-Dependent Fluorescence Shift of a  Reichardt-Type Dye”, Angew. Chem. Int. Ed. 52, 1813-1816 (2013)   

[2] C. Allolio and D.  Sebastiani, “Approaches to the solvation of the molecular probe  N-methyl-6-quinolone in its excited state”, Phys. Chem. Chem. Phys. 13, 16395-16403 (2011).   

[3] G. Bekçioglu, F.  Hoffmann, and D. Sebastiani, “Solvation-dependent latency of photoacid  dissociation and transient IR signatures of protonation dynamics,” J. Phys. Chem. A 119, 9244 (2015).
   
[4]  M. Ekimova, F. Hoffmann, G. Bekcioglu-Neff, A. Rafferty, O. Kornilov,  E.T. Nibbering, and D. Sebastiani, “Ultrafast proton transport between a  hydroxy acid and a nitrogen base along solvent bridges governed by  hydroxide/methoxide transfer mechanism,” J. Am. Chem. Soc. 141, 14581 (2019).    

[5]  P. Partovi-Azar and D. Sebastiani, Optimized effective potentials to  increase the accuracy of approximate proton transfer energy calculations  in the excited state, J. Chem. Phys. 152, 064101 (2020).   

Zum Seitenanfang