Exploring molecular excited states holds immense significance across organic chemistry, chemical biology, and materials science. Understanding the photophysical properties of molecular chromophores is crucial for designing nature-inspired functional molecules, with applications ranging from photosynthesis to pharmaceuticals. Non-adiabatic molecular dynamics simulations are powerful tools to investigate the photochemistry of molecules and materials, but demand extensive computing resources, especially for complex molecules and environments. To address these challenges, the integration of machine learning has emerged. Machine learning algorithms can be used to analyse vast datasets and accelerate discoveries by identifying relationships between geometrical features and ground as well as excited-state properties. However, challenges persist, including the acquisition of accurate excited-state data and managing the complexity of the data. This article provides an overview of recent and best practices in machine learning for non-adiabatic molecular dynamics, focusing on pre-processing, surface fitting, and post-processing of data.

• Publication type
• Journal Article MeSH
• Review MeSH

Simulating the coupled electronic and nuclear response of a molecule to light excitation requires the application of nonadiabatic molecular dynamics. However, when faced with a specific photophysical or photochemical problem, selecting the most suitable theoretical approach from the wide array of available techniques is not a trivial task. The challenge is further complicated by the lack of systematic method comparisons and rigorous testing on realistic molecular systems. This absence of comprehensive molecular benchmarks remains a major obstacle to advances within the field of nonadiabatic molecular dynamics. A CECAM workshop, Standardizing Nonadiabatic Dynamics: Towards Common Benchmarks, was held in May 2024 to address this issue. This Perspective highlights the key challenges identified during the workshop in defining molecular benchmarks for nonadiabatic dynamics. Specifically, this work outlines some preliminary observations on essential components needed for simulations and proposes a roadmap aiming to establish, as an ultimate goal, a community-driven, standardized molecular benchmark set.

• Publication type
• Journal Article MeSH
• Review MeSH

The COLUMBUS program system provides the tools for performing high-level multireference (MR) computations, including the multireference configuration interaction (MRCI) method and its multireference averaged quadratic coupled cluster (MR-AQCC) extension, allowing computations on a wide range of fascinating atomic and molecular systems, including the treatment of open-shells and complicated excited state phenomena. The inclusion of spin-orbit coupling (SOC) directly within the MRCI step enables the description of systems containing heavy elements, such as lanthanides and actinides, whose properties are strongly influenced by SOC. Analytic energy gradients and nonadiabatic couplings at the correlated MRCI level provide the foundation for a variety of dynamics studies, giving insight into ultrafast photochemistry. New and ongoing method developments in COLUMBUS include the computation of spin densities, improved descriptions of ionic states, enhancements to the AQCC method, and the porting of COLUMBUS to graphical processing units (GPUs). New external interfaces enable an enhanced description of electronic resonances and molecules in strong laser fields. This work highlights these new developments while providing a detailed account of the diverse applications of COLUMBUS in recent years.

• Publication type
• Journal Article MeSH

A computational study of I-BODIPY (2-ethyl-4,4-difluoro-6,7-diiodo-1,3-dimethyl-4-bora-3a,4a-diaza-s-indacene) has been carried out to investigate its key photophysical properties as a potential triplet photosensitizer capable of generating singlet oxygen. Multireference CASPT2 and CASSCF methods have been used to calculate vertical excitation energies and spin-orbit couplings (SOCs), respectively, in a model (mono-iodinated BODIPY) molecule to assess the applicability of the single-reference second-order algebraic diagrammatic construction, ADC(2), method to this and similar molecules. Subsequently, time-dependent density functional theory (TD-DFT), possibly within the Tamm-Dancoff approximation (TDA), using several exchange-correlation functionals has been tested on I-BODIPY against ADC(2), both employing a basis set with a two-component pseudopotential on the iodine atoms. Finally, the magnitudes of SOC between excited electronic states of all types found have thoroughly been discussed using the Slater-Condon rules applied to an arbitrary one-electron one-center effective spin-orbit Hamiltonian. The geometry dependence of SOCs between the lowest-lying states has also been addressed. Based on these investigations, the TD-DFT/B3LYP and TD-DFT(TDA)/BHLYP approaches have been selected as the methods of choice for the subsequent nuclear ensemble approach absorption spectra simulations and mixed quantum-classical trajectory surface hopping (TSH) molecular dynamics (MD) simulations, respectively. Two bright states in the visible spectrum of I-BODIPY have been found, exhibiting a redshift of the main peak with respect to unsubstituted BODIPY caused by the iodine substituents. Excited-state MD simulations including both non-adiabatic effects and SOCs have been performed to investigate the relaxation processes in I-BODIPY after its photoexcitation to the S 1 $$ {\mathrm{S}}_1 $$ state. The TSH MD simulations revealed that intersystem crossings occur on a time scale comparable to internal conversions and that after an initial phase of triplet population growth a "saturation" is reached where the ratio of the net triplet to singlet populations is about 4:1. The calculated triplet quantum yield of 0.85 is in qualitative agreement with the previously reported experimental singlet oxygen generation yield of 0.99 ± $$ \pm $$ 0.06.

• Keywords
• excited‐state MD simulations, iodinated BODIPY, photosensitizer, spin–orbit coupling, trajectory surface hopping,
• Publication type
• Journal Article MeSH

The field of nonadiabatic dynamics has matured over the last decade with a range of algorithms and electronic structure methods available at the moment. While the community currently focuses more on developing and benchmarking new nonadiabatic dynamics algorithms, the underlying electronic structure controls the outcome of nonadiabatic simulations. Yet, the electronic-structure sensitivity analysis is typically neglected. In this work, we present a sensitivity analysis of the nonadiabatic dynamics of cyclopropanone to electronic structure methods and nonadiabatic dynamics algorithms. In particular, we compare wave function-based CASSCF, FOMO-CASCI, MS- and XMS-CASPT2, density-functional REKS, and semiempirical MRCI-OM3 electronic structure methods with the Landau-Zener surface hopping, fewest switches surface hopping, and ab initio multiple spawning with informed stochastic selection algorithms. The results clearly demonstrate that the electronic structure choice significantly influences the accuracy of nonadiabatic dynamics for cyclopropanone even when the potential energy surfaces exhibit qualitative and quantitative similarities. Thus, selecting the electronic structure solely on the basis of the mapping of potential energy surfaces can be misleading. Conversely, we observe no discernible differences in the performance of the nonadiabatic dynamics algorithms across the various methods. Based on the above results, we discuss the present-day practice in computational photodynamics.

• Publication type
• Journal Article MeSH

A swarm of semi-classical quantum mechanics/molecular mechanics molecular-dynamics simulations where OM2/MNDO is combined with the Gromacs program for consideration of explicit water is performed, solving the time-dependent Schrödinger equation in each step of the trajectories together with the Tully's fewest switches algorithm. Within this stochastic treatment, time dependent probabilities of the three lowest electronic states are determined. The fact that nucleobases are quickly deactivated is confirmed in the cytosine case where our best lifetime estimation is τ1=0.82 ps for the model with 100 water molecules with the SPCE force field and a time step of 0.1 fs. Lifetimes of the remaining molecules are visibly longer: 5-azacytosine, 2,4-diamino-1,3,5-triazine (DT), and 2,4,6-triamino-1,3,5-triazine (TT) molecules have an S1 → S0 de-excitation time of slightly above 10 ps. The lifetimes of the triazine family increases with the increasing number of exocyclic amino groups, that is, s-triazine < 2-amino-1,3,5-triazine < DT < TT. This can be explained by a higher mobility of the carbon-bonded hydrogen atoms in comparison with heavier amino groups since their movement is slowed down due to a substantially higher mass than hydrogen atoms, which can easier reach the out-of-plane positions required in the conical intersection structures. Moreover, bulkier NH2 ligands suffer due to greater friction caused by the surrounding water environment. These mechanical aspects caused a change in the explored lifetime dependences in comparison with our previous gas-phase study.

• Publication type
• Journal Article MeSH