HAB79: A new molecular dataset for benchmarking DFT and DFTB electronic couplings against high-level ab initio calculations

Using atomic clustering based on structural and electronic descriptors that consider surrounding environment to evaluate local properties of DFT functionals

We developed a method for evaluating the accuracies of the local properties of DFT functionals in detail using a clustering method based on machine learning and structural/electronic descriptors. We generated 36 clusters consistent with human intuition using 30,436 carbon atoms from the QM9 dataset. The results were used to evaluate 13C NMR chemical shifts calculated using 84 DFT functionals. Carbon atoms were grouped based on their similar environments, reducing errors within these groups. This enables more accurate assessment of the accuracy using a specific DFT functional. Therefore, the present atomic clustering provides more detailed insight into accuracy verification.

Efficient calculation of electronic coupling integrals with the dimer projection method via a density matrix tight-binding potential

Designing organic semiconductors for practical applications in organic solar cells, organic field-effect transistors, and organic light-emitting diodes requires understanding charge transfer mechanisms across different length and time scales. The underlying electron transfer mechanisms can be efficiently explored using semiempirical quantum mechanical (SQM) methods. The dimer projection (DIPRO) method combined with the recently introduced non-self-consistent density matrix tight-binding potential (PTB) [Grimme et al., J. Chem. Phys. 158, 124111 (2023)] is used in this study to evaluate charge transfer integrals important for understanding charge transport mechanisms. PTB, parameterized for the entire Periodic Table up to Z = 86, incorporates approximate non-local exchange, allowing for efficient and accurate calculations for large hetero-organic compounds. Benchmarking against established databases, such as Blumberger's HAB sets, or our newly introduced JAB69 set and comparing with high-level reference data from ωB97X-D4 calculations confirm that DIPRO@PTB consistently performs well among the tested SQM approaches for calculating coupling integrals. DIPRO@PTB yields reasonably accurate results at low computational cost, making it suitable for screening purposes and applications to large systems, such as metal-organic frameworks and cyanine-based molecular aggregates further discussed in this work.

Infrared Spectroscopy of Liquid Solutions as a Benchmarking Tool of Semiempirical QM Methods: The Case of GFN2-xTB

The accurate description of large molecular systems has triggered the development of new computational methods. Due to the computational cost of modeling large systems, the methods usually require a trade-off between accuracy and speed. Therefore, benchmarking to test the accuracy and precision of the method is an important step in their development. The typical gold standard for evaluating these methods is isolated molecules, because of the low computational cost. However, the advent of high-performance computing has made it possible to benchmark computational methods using observables from more complex systems such as liquid solutions. To this end, infrared spectroscopy provides a suitable set of observables (i.e., vibrational transitions) for liquid systems. Here, IR spectroscopy observables are used to benchmark the predictions of the newly developed GFN2-xTB semiempirical method. Three different IR probes (i.e., N-methylacetamide, benzonitrile, and semiheavy water) in solution are selected for this purpose. The work presented here shows that GFN2-xTB predicts central frequencies with errors of less than 10% in all probes. In addition, the method captures detailed properties of the molecular environment such as weak interactions. Finally, the GFN2-xTB correctly assesses the vibrational solvatochromism for N-methylacetamide and semiheavy water but does not have the accuracy needed to properly describe benzonitrile. Overall, the results indicate not only that GFN2-xTB can be used to predict the central frequencies and their dependence on the molecular environment with reasonable accuracy but also that IR spectroscopy data of liquid solutions provide a suitable set of observables for the benchmarking of computational methods.

Tunneling-to-Hopping Transition in Multiheme Cytochrome Bioelectronic Junctions

Multiheme cytochromes (MHCs) have attracted much interest for use in nanobioelectronic junctions due to their high electronic conductances. Recent measurements on dry MHC junctions suggested that a coherent tunneling mechanism is operative over surprisingly long long distances (>3 nm), which challenges our understanding of coherent transport phenomena. Here we show that this is due to (i) a low exponential distance decay constant for coherent conduction in MHCs (β = 0.2 Å^-1) and (ii) a large density of protein electronic states which prolongs the coherent tunneling regime to distances that exceed those in molecular wires made of small molecules. Incoherent hopping conduction is uncompetitive due to the large energy level offset at the protein-electrode interface. Removing this offset, e.g., by gating, we predict that the transport mechanism crosses over from coherent tunneling to incoherent hopping at a protein size of ∼7 nm, thus enabling transport on the micrometer scale with a shallow polynomial (∼1/r) distance decay.

Design of new hole transport materials based on triphenylamine derivatives using different π-linkers for the application in perovskite solar cells. A theoretical study

New organic molecules containing five different compounds, commonly called p-linkers, located between the triphenylamine units, were theoretically designed and analyzed in order to be proposed as new hole transport materials (HTMs) in perovskite solar cells, in total ten new molecules were analyzed. The electronic, optical and hole transport properties were determined, similarly, the relationship of these properties with their molecular structure was also investigated by Density Functional Theory (DFT) and Density Functional Tight Binding (DFTB) calculations. Eight of the ten analyzed compounds exhibited the main absorption band out of the visible region; therefore these compounds did not present an overlap with the absorption spectra of the typical methylammonium lead iodide (MAPI) hybrid-perovskite. The results showed that the Highest occupied molecular orbital (HOMO) levels of the compounds are higher than the perovskite HOMO level, and in some cases these are even higher than the Spiro-OMeTAD HOMO. The calculated electronic couplings and the reorganization energy values provided useful information in order to determine if the systems were hole or electron transport materials.

A quantum chemical molecular dynamics repository of solvated ions

The importance of ion-solvent interactions in predicting specific ion effects in contexts ranging from viral activity through to electrolyte viscosity cannot be underestimated. Moreover, investigations of specific ion effects in nonaqueous systems, highly relevant to battery technologies, biochemical systems and colloid science, are severely limited by data deficiency. Here, we report IonSolvR – a collection of more than 3,000 distinct nanosecond-scale ab initio molecular dynamics simulations of ions in aqueous and non-aqueous solvent environments at varying effective concentrations. Density functional tight binding (DFTB) is used to detail the solvation structure of up to 55 solutes in 28 different protic and aprotic solvents. DFTB is a fast quantum chemical method, and as such enables us to bridge the gap between efficient computational scaling and maintaining accuracy, while using an internally-consistent simulation technique. We validate the database against experimental data and provide guidance for accessing individual IonSolvR records.

Measurement(s)	solvation structure
Technology Type(s)	quantum chemistry computational method • Molecular Dynamics

Charge Transport in Organic Semiconductors: The Perspective from Nonadiabatic Molecular Dynamics

Organic semiconductors (OSs) are an exciting class of materials that have enabled disruptive technologies in this century including large-area electronics, flexible displays, and inexpensive solar cells. All of these technologies rely on the motion of electrical charges within the material and the diffusivity of these charges critically determines their performance. In this respect, it is remarkable that the nature of the charge transport in these materials has puzzled the community for so many years, even for apparently simple systems such as molecular single crystals: some experiments would better fit an interpretation in terms of a localized particle picture, akin to molecular or biological electron transfer, while others are in better agreement with a wave-like interpretation, more akin to band transport in metals.

Exciting recent progress in the theory and simulation of charge carrier transport in OSs has now led to a unified understanding of these disparate findings, and this Account will review one of these tools developed in our laboratory in some detail: direct charge carrier propagation by quantum-classical nonadiabatic molecular dynamics. One finds that even in defect-free crystals the charge carrier can either localize on a single molecule or substantially delocalize over a large number of molecules depending on the relative strength of electronic couplings between the molecules, reorganization, or charge trapping energy of the molecule and thermal fluctuations of electronic couplings and site energies, also known as electron–phonon couplings.

Our simulations predict that in molecular OSs exhibiting some of the highest measured charge mobilities to date, the charge carrier forms “flickering” polarons, objects that are delocalized over 10–20 molecules on average and that constantly change their shape and extension under the influence of thermal disorder. The flickering polarons propagate through the OS by short (≈10 fs long) bursts of the wave function that lead to an expansion of the polaron to about twice its size, resulting in spatial displacement, carrier diffusion, charge mobility, and electrical conductivity. Arguably best termed “transient delocalization”, this mechanistic scenario is very similar to the one assumed in transient localization theory and supports its assertions. We also review recent applications of our methodology to charge transport in disordered and nanocrystalline samples, which allows us to understand the influence of defects and grain boundaries on the charge propagation.

Unfortunately, the energetically favorable packing structures of typical OSs, whether molecular or polymeric, places fundamental constraints on charge mobilities/electronic conductivity compared to inorganic semiconductors, which limits their range of applications. In this Account, we review the design rules that could pave the way for new very high-mobility OS materials and we argue that 2D covalent organic frameworks are one of the most promising candidates to satisfy them.

We conclude that our nonadiabatic dynamics method is a powerful approach for predicting charge carrier transport in crystalline and disordered materials. We close with a brief outlook on extensions of the method to exciton transport, dissociation, and recombination. This will bring us a step closer to an understanding of the birth, survival, and annihiliation of charges at interfaces of optoelectronic devices.

Efficient Surface Hopping Approach for Modeling Charge Transport in Organic Semiconductors

The trajectory surface hopping (TSH) method is nowadays widely applied to study the charge/exciton transport process in organic semiconductors (OSCs). In the present study, we systematically examine the performance of two approximations in the fewest switched surface hopping (FSSH) simulations for charge transport (CT) in several representative OSCs. These approximations include (i) the substitution of the nuclear velocity scaling along the nonadiabatic coupling vector (NCV) by rescaling the hopping probability with the Boltzmann factor (Boltzmann correction (BC)) and (ii) a phenomenological approach to treat the quantum feedback from the electronic system to the nuclear system (implicit charge relaxation (IR)) in the OSCs. We find that charge mobilities computed by FSSH-BC-IR are in very good agreement with the mobilities obtained by standard FSSH simulations with explicit charge relaxation (FSSH-ER), however, at reduced computational cost. A key parameter determining the charge carrier mobility is the reorganization energy, which is sensitively dependent on DFT functionals applied. By employing the IR approximation, the FSSH method allows systematic investigation of the effect of the reorganization energies obtained by different DFT functionals like B3LYP or ωB97XD on CT in OSCs. In comparison to the experiments, FSSH-BC-IR using ωB97XD reorganization energy underestimates mobilities in the low-coupling regime, which may indicate the lack of nuclear quantum effects (e.g., zero point energy (ZPE)) in the simulations. The mobilities obtained by FSSH-BC-IR using the B3LYP reorganization energy agree well with experimental values in 3 orders of magnitude. The accidental agreement may be the consequence of the underestimation of the reorganization energy by the B3LYP functional, which compensates for the neglect of nuclear ZPE in the simulations.

Ultrafast estimation of electronic couplings for electron transfer between pi-conjugated organic molecules. II

The development of highly efficient methods for the calculation of electronic coupling matrix elements between the electron donor and acceptor is an important goal in theoretical organic semiconductor research. In Paper I [F. Gajdos, S. Valner, F. Hoffmann, J. Spencer, M. Breuer, A. Kubas, M. Dupuis, and J. Blumberger, J. Chem. Theory Comput. 10, 4653 (2014)], we introduced the analytic overlap method (AOM) for this purpose, which is an ultrafast electronic coupling estimator parameterized to and orders of magnitude faster than density functional theory (DFT) calculations at a reasonably small loss in accuracy. In this work, we reparameterize and extend the AOM to molecules containing nitrogen, oxygen, fluorine, and sulfur heteroatoms using 921 dimer configurations from the recently introduced HAB79 dataset. We find again a very good linear correlation between the frontier orbital overlap, calculated ultrafast in an optimized minimum Slater basis, and DFT reference electronic couplings. The new parameterization scheme is shown to be transferable to sulfur-containing polyaromatic hydrocarbons in experimentally resolved dimeric configurations. Our extension of the AOM enables high-throughput screening of very large databases of chemically diverse organic crystal structures and the application of computationally intense non-adiabatic molecular dynamics methods to charge transport in state-of-the-art organic semiconductors, e.g., non-fullerene acceptors.

Kinetics of trifurcated electron flow in the decaheme bacterial proteins MtrC and MtrF

The bacterium Shewanella oneidensis has evolved a sophisticated electron transfer (ET) machinery to export electrons from the cytosol to extracellular space during extracellular respiration. At the heart of this process are decaheme proteins of the Mtr pathway, MtrC and MtrF, located at the external face of the outer bacterial membrane. Crystal structures have revealed that these proteins bind 10 c-type hemes arranged in the peculiar shape of a staggered cross that trifurcates the electron flow, presumably to reduce extracellular substrates while directing electrons to neighboring multiheme cytochromes at either side along the membrane. Especially intriguing is the design of the heme junctions trifurcating the electron flow: they are made of coplanar and T-shaped heme pair motifs with relatively large and seemingly unfavorable tunneling distances. Here, we use electronic structure calculations and molecular simulations to show that the side chains of the heme rings, in particular the cysteine linkages inserting in the space between coplanar and T-shaped heme pairs, strongly enhance electronic coupling in these two motifs. This results in an [Formula: see text]-fold speedup of ET steps at heme junctions that would otherwise be rate limiting. The predicted maximum electron flux through the solvated proteins is remarkably similar for all possible flow directions, suggesting that MtrC and MtrF shuttle electrons with similar efficiency and reversibly in directions parallel and orthogonal to the outer membrane. No major differences in the ET properties of MtrC and MtrF are found, implying that the different expression levels of the two proteins during extracellular respiration are not related to redox function.