Recent developments and applications of the electron nuclear dynamics (END) theory [1] will be presented [2-4]. END is a time-dependent, variational, non-adiabatic method for chemical dynamics that evaluates potential energy and interatomic forces “on the fly”, without employing predetermined potential energy surfaces. The simplest-level END (SLEND) [1] adopts a nuclear classical-mechanics description (as the zero-width limit of frozen Gaussian wave packets) and an electronic single-determinantal wavefunction. In the SLEND framework, three new developments will be presented:
1. The use of various types of coherent-states (CS) sets [4] to describe all types of particles (nuclei and electrons) and of degrees of freedom (translational, rotational [5], vibrational [6], and electronic). The CS sets conveniently represent the SLEND trial wavefunction and can mediate between classical and quantum descriptions. For instance, the rotational [5] and canonical [6] CS sets permit reconstructing rovibrational quantum properties from the SLEND nuclear classical dynamics. Conversely, a new CS set [2] participates in a valence-bond approach to a classical-electrostatics/charge-equilibration model based on the Sanderson principle of electronegativity equalization.
2. A new time-dependent Kohn-Sham density-functional-theory (KSDFT) method in the SLEND framework: END/KSDFT [3], which incorporates electron correlation effects absent in SLEND.
3. A new implementation of effective core potentials into SLEND and END/KSDT to treat large systems.
The new developments are implemented in our code: PACE (Python Accelerated Coherent-states Electron-nuclear dynamics) that utilizes several computer-science technologies [code parallelization, compute unified devise architecture (CUDA) devices, etc.]. The new developments are applied to the following chemical systems:
1. High-energy collisions of protons with water clusters (water radiolysis) and with DNA components (direct DNA damage): these processes are highly relevant to proton cancer therapy.
2. Various proton-molecule reactions [7,8] with an emphasis on accurately predicting rovibrational, energy-transfer, and electron-transfer properties.
3. Various chemical reactions involving large reactants, such as Diels-Alder and SN2 reactions inter alia.
Results of the above simulations compare well with available experimental results.
References:
[1] E. Deumens, A. Diz, R. Longo, Y. Öhrn, Rev. Mod. Phys. 66 (1994) 917.
[2] J. A. Morales, J. Phys. Chem. A 113 (2009) 6004.
[3] S. A. Perera, P. M. McLaurin, T. V. Grimes, J. A. Morales, Chem. Phys. Lett. 496 (2010) 188.
[4] J. A. Morales, Mol. Phys. 108, 3199 (2010)
[5] J. A. Morales, E. Deumens, Y. Öhrn, J. Math. Phys. 40 (1999) 766.
[6] J. A. Morales, A. Diz, E. Deumens, Y. Öhrn, J. Chem. Phys. 133 (1995) 9968.
[7] B. Maiti, R. Sadeghi, A. Austin, J. A. Morales, Chem. Phys. 340 (2007) 105.
[8] B. Maiti, P. M. McLaurin, R. Sadeghi, S. A. Perera, J. A. Morales, Int. J. Quant. Chem. 109 (2009) 3026. |