QM/MM Nonadiabatic Dynamics: the SHARC/COBRAMM Approach

Abstract

We present the SHARC/COBRAMM approach to enable easy and efficient excited-state dynamics simulations at different levels of electronic structure theory in the presence of complex environments using a quantum mechanics/molecular mechanics (QM/MM) setup. SHARC is a trajectory surface-hoping method that can incorporate the simultaneous effects of nonadiabatic and spin–orbit couplings in the excited-state dynamics of molecular systems. COBRAMM allows ground- and excited-state QM/MM calculations using a subtractive scheme, with electrostatic embedding and a hydrogen link-atom approach. The combination of both free and open-source program packages provides a modular and extensive framework to model nonadiabatic processes after light irradiation from the atomistic scale to the nano-scale. As an example, the relaxation of acrolein from S1 to T1 in solution is provided.

Sampling effects in quantum mechanical/molecular mechanics trajectory surface hopping non-adiabatic dynamics

“…Highlights in the QM/MM implementation include (i) a subtractive scheme for the calculation of the energy and the gradient of the electronic states [ 8 ], with the implementation of an electrostatic embedding scheme [ 27 ] to account for the polarization of the QM part due to the presence of the surrounding MM environment, included in the adopted Hamiltonian as point charges; (ii) the inclusion in the MM gradient of a state-specific term due to the force induced by the QM region on the point charges; (iii) the QM molecule surrounded by a droplet of homogeneous radius of water molecules and, in order to address the lack of periodical boundary conditions, the external shell of water molecules is kept frozen to furnish a constant potential and keep the droplet stable. Additional information and the full implementation can be found in the original publication [ 21 ].…”

“…The non-adiabatic coupling was approximated by evaluating the time-derivative coupling through wave function overlaps [ 34 ]. The energy-based decoherence correction scheme [ 35 ], with a decoherence parameter of 0.1, and an atom masking to exclude the solvent molecules from the velocities rescaling procedures was employed [ 21 ].…”

“…We limit the solvent role to electrostatic effects, yet analogous for each set of generated initial conditions. For these calculations, we shall use our recently developed scheme to run QM/MM non-adiabatic dynamics simulations [ 21 ] based on SHARC [ 22 , 23 ] and COBRAMM [ 24 , 25 ] software, which provide TSH and the QM/MM implementations, respectively. We will show how changing the initial momenta, either due to manual selection or through energy exchange and equilibration between the QM and the MM region, will lead to very different and interesting results that should be kept in mind when deciding how to sample initial conditions for QM/MM TSH simulations.…”

Scalable Ehrenfest Molecular Dynamics Exploiting the Locality of Density-Functional Tight-Binding Hamiltonian

“…A well-known example of this kind of approach is the quantum mechanics/molecular mechanics (QM/MM) method. , In the QM/MM-based nonadiabatic MD simulations, the color center is subjected to the nonadiabatic MD treatment based on quantum chemical calculations, while the rest of the system is simulated with the classical MD. These are generally used to study the condensed-phase excited-state dynamics. − In addition, we have developed the excited-state MD methods in the framework of the divide and conquer (DC) method; it is a linear scaling quantum chemical calculation technique that divides the system into a set of small subsystems, where the density matrix of the entire system is obtained by merging the local matrices with each other. , To adapt the DC method to the excited-state MD simulations, the excited-state calculation is performed on a single subsystem that includes the color center. , The DC-based excited-state MD method has further been extended to surface-hopping simulations to treat the nonadiabatic phenomena. , The abovementioned approaches reduce the complexity of the problem by assuming that only a local part of the system is electronically excited.…”

INAQS, a generic interface for non-adiabatic QM/MM dynamics: Design, implementation, and validation for GROMACS/Q-CHEM simulations

“…energies, gradients, nonadiabatic couplings) are typically computed via the standalone approach. 45,[47][48][49][50][51][52][53] To the authors knowledge, and with the exception of some in-house codes, 54,55 integrating the nonadiabatic effects into an existing MD code (2) is rarely employed. 56 But, this approach has several distinct advantages: reduced data transfer and efficient schemes for storing and analyzing long trajectories; the widespread availability of established FFs for solvents of all kinds; the use of advanced sampling and metadynamic schemes.…”