Excited-state energy decomposition analysis (EDA) provides a useful framework to dissect the physical interactions that stabilize molecular complexes in electronically excited states. While ground-state EDA has been widely applied to understand noncovalent interactions and chemical bonding, excited states introduce qualitatively new contributions, including photoexcitation, exciton resonance, and charge-transfer excitations. Recent developments in multistate density functional theory (MSDFT) extends the interpretability of EDA into the excited-state domain, offering mechanistic insight into photophysical and photochemical phenomena. This report summarizes the theoretical foundation of multistate EDA (MS-EDA), defines its key energetic terms, and illustrates its application to several groups of excited-state complexes. It is hoped that MS-EDA can provide interpretable understanding of excited state energies in terms of exciton resonance, superexchange stabilization and orbital and configuraiton delocalization.
Traditional computational methods for molecule design are based on first principles calculation, which places a high demand on computing power. The increasingly powerful machine learning (ML) models have fundamentally transformed this landscape. Statistically, by learning the joint probability distribution between molecular or material structure and targeted properties, generative models can autonomously design numerous novel structures with satisfactory properties. This inverse design strategy clearly outperforms the traditional physics-based methods which requires human expertise and intuition, along with serendipity. To validate the generated molecules or materials for specific properties, classical discriminative models allow for fast large-scale screening of the quantitative structure-activity relationships. Generally, the completely ML-based workflow from generation to validation for the exploration of chemical space is accessible and provides outstanding benefits which traditional computational approaches struggle to achieve. In this review, we summarize recent advances in ML-assisted discovery for transition metal complexes and conclude with several existing challenges which impede the widespread practical applications of this technology to the class of problems.
Over the past 25 years there has been remarkable progress towards accurate description of nonbonded interactions within the context of density functional theory (DFT). Various methods have been devised to capture London dispersion, which is the most exacting contribution to noncovalent interactions; these strategies include both new functionals as well as ad hoc dispersion corrections to existing functionals. At present, it is possible to compute interaction energies for small van der Waals complexes (containing ~20 atoms) to an accuracy of ~0.5 kcal/mol, using a range of dispersion-inclusive DFT methods that are reviewed here. Systematic tests reveal remarkable consistency across different methods, at least for small noncovalent dimers, although the magnitude of the ad hoc dispersion corrections is systematically smaller than benchmark dispersion energies because some dispersion resides within the semilocal exchange-correlation functional, in a manner that is difficult to disentangle. Despite impressive results for small systems, the best contemporary DFT methods afford larger errors in systems with 50-130 atoms, approaching 3-5 kcal/mol as compared to ab initio benchmarks for total interaction energies, although the benchmarks have larger uncertainties in systems of this size. Errors for larger systems vary widely from one DFT method to the next ,with no discernible systematic trend. Nanoscale van der Waals complexes thus represent the next frontier in development of DFT for noncovalent interactions.
Theoretical spectroscopy is a powerful tool in the chemist’s toolbox, providing insight into experimental investigations here on Earth and into the observations of telescopes like the recently-operational JWST. This work presents recent developments in our group focused on extending the application of highly-accurate, theoretical spectroscopic techniques to large molecules. By leveraging composite and hybrid methods based on coupled cluster theory, molecules of study can increase in size from three to five atoms to upwards of ten atoms. In turn, using such results as a training set lays the foundation for methods capable of treating dozens, or potentially hundreds, of atoms in the future.
Biofuels constitute a remarkable, sustainable energy source for a future of clean energy. Efficient separation of biofuel components is critical for its cost-effective utilization. In this chapter, we provide an overview of the recent advancements in atomistic-level modeling and deep learning in the rational design of novel, efficient biofuel separation processes. We will briefly review the fundamental principles of quantum and statistical mechanics frequently employed to highlight their underlying differences. We expand our review to the methodologies for molecular representations and deep learning algorithms applicable to biofuel separations. The applications, successes, and risks of employing density functional theory, ab initio molecular dynamics, classical molecular dynamics, and deep learning are provided to showcase their recent accomplishments in biofuel separation, as well as potential improvements in both methodology and application. Lastly, a vision for the future growth of these methods is illustrated.
Despite extensive studies on the catalytic oxidative coupling of methane (OCM) reaction in the past forty years using a variety of experimental and computational approaches, there remain great challenges in finding optimal catalysts for the commercialization of this highly promising reaction. In this perspective, we first highlight the important insights gained from our recent efforts in employing in situ and Operando spectroscopic techniques and first principles based computational methods to predict spectroscopic and catalytic properties. We then describe future research directions for gaining a deeper understanding of the mechanism and kinetics of the OCM reaction using an integrated spectroscopic and computational approach for the rational design of more effective OCM catalysts with a much higher ethylene yield.
Robust and universally applicable double-hybrid functionals are developed in the last decade for all kinds of electronic structure calculations. The expressions presented here are based on the adiabatic connection model introducing a proper interpolation (quadratic) function and are free of any empirical parameterization. We have also developed extensions to adequately incorporate long-range exchange and (non-covalent) correlation effects, as well as efficient techniques to allow the application of these expressions to considerably large systems. Additionally, the corresponding linear-response time-dependent formalism can also be applied to deal with excited-states calculations. The systematic benchmarking of these models along the years have proved their accuracy on a wide range of chemical systems and properties of the most interest. Thus, they represent a set of useful tools for electronic structure calculations which, furthermore, are widely available in many computational chemistry codes.
The transition from fossil fuels to cleaner energies employing different renewable sources constitutes one of the primary worldwide challenges. The search for appropriate solutions is becoming more urgent in view of the severe consequences of climate change. As for a perspective, stationary energy storage, alkali-ion batteries and hybrid supercapacitors are, among others, considered as efficient and affordable solutions. Alkali-ion batteries have proved to be the most investigated products in the past decade including optimizations for cost, energy density and safety. In this Perspective, a computational approach and its applicability in the inverse material design are presented. This approach includes density functional theory calculations, force field-based determinations and both static and molecular dynamics simulations. As for an illustration, the main properties of a selected series of battery materials, including oxides and sulfides Li2SiO3, Li2SnO3, SrSnO3, and A2B6X13 (A = Li+, Na+, K+; B = Ti4+, Sn4+; X = O2-, S2-), and mixed halide antiperovskite A3OX (A = Li+, Na+; X = Cl-, Br-) are explored in depth using these theoretical approaches. Doping strategies, new dopant incorporation mechanism, treatment with alkali insertion/de-insertion cycle in electrodes, transport properties, as well as thermodynamic stability, are discussed. Theoretical approaches reveal that the oxygen-sulfur exchange in alkali hexatitanates and hexastannates induces remarkable improvement of the required properties for electrode and electrolyte materials. In addition, doping of Li2SiO3 with low Na-concentration enhances the room temperature Li-diffusivity by a reduction of the activation energy. The effects of transition-metal and divalent dopants on the defect chemistry and transport properties of Li2SnO3 are also disclosed. The interstitial trivalent doping mechanism is a friendly synthesis strategy to improve the large-scale diffusion in Li2SnO3. The potential of SrSnO3 as an anode in alkali-ion batteries, and the influence of a particular grain boundary in nanocrystalline antiperovskite A3OX are also revealed by using advanced atomistic simulations. The computational approaches described here provide us with a convenient tool for the determination of the properties of battery materials with high accuracy and for the prediction of characteristics of a new generation of alkali battery materials that could be used in improved technologies.
When chemical reactions are accelerated by a catalyst, entropy differences between reactants and their transient intermediates can be the driving force behind the promotion or inhibition of desired and parasitic chemical pathways. Understanding and controlling catalytic processes therefore requires both a fundamental and practicable understanding of entropy in addition to enthalpy. In unstructured media such as the vapor phase equilibrated with sparsely covered surfaces, entropy can be adequately accounted for by well-established approaches based on translational, rotational, and harmonic vibrational partition functions. However, these approximations become inadequate in more complex condensed phase environments, e.g., solid-liquid interfaces of confined reaction spaces. In this chapter, we provide an overview of the state-of-art in the computational quantification of entropy and its known ramifications on catalysis. The fundamental roles of thermodynamics and kinetics in catalysis are covered in enough detail to appreciate and contextualize the computational methods employed to compute chemically accurate estimates of entropy. These methods are discussed in appropriate detail and range from the ubiquitous harmonic oscillator approximation where entropy unrelated to high frequency oscillations is typically underestimated, to enhanced free energy sampling with molecular dynamics where the desired accuracy must be weighed against the associated computational cost of obtaining it. The rising importance of machine learning and artificial intelligence in accelerating methodological progress in this field is touched upon, as well. Finally, applications, successes, and pitfalls of using these methods are provided to showcase past and present accomplishments while clarifying where improvements in both understanding and methodology are still needed.
Twenty-five years ago, the two main pillars of quantum chemistry-density functional and composite ab initio theories-were recognized with a Nobel Prize in Chemistry awarded to Walter Kohn and John Pople. This recognition sparked intense theoretical developments in both fields. Whereas in 1998, the year the Nobel Prize was awarded, there were only a handful of composite ab initio methods; most notably the Gaussian-n methods (n = 1-3), CBS methods (e.g., CBS-QCI and CBS-APNO), and the focal-point analysis approach, today there are many more families of such methods, including the Weizmann-n, MCCM, HEAT, ccCA, FPD, ATOMIC, INT-MP2-F12, and ChS family of methods, where some of these families include dozens of variants. Overall, there are over 100 contemporary variants of composite ab initio methods to choose from, with many variants implemented as a keyword in popular quantum chemical packages. This situation makes it difficult to choose a proper method for a given chemical system, property, and desired accuracy. This chapter provides an overview of contemporary composite ab initio methods applicable to first- and second-row elements, their main energetic components, and their expected accuracy and applicability. To guide the selection of a suitable method for a given chemical system and desired accuracy, the various methods are classified according to a 'Jacob's Ladder' of composite ab initio methods, from computationally economical methods that are capable of approaching chemical accuracy to computationally demanding methods capable of confident sub-benchmark accuracy.
In this work, we review recent studies in the field of computational solvation. Methods at different scales from ab initio to force field calculations with both implicit solvent and explicit solvent modeling are covered. Developments in data science and computer hardware are also highlighted, as are solvation databases, recommended parameters and benchmarking results. Recent applications employing different solvation modeling techniques are showcased, and remaining challenges are discussed.
The G3(MP2) or G3(MP2)B3 composite correlated molecular orbital theory methods have been used to predict the heats of formation of 28 high energy compounds and their borane adducts. The BdN bond energies are also predicted. The heats of formation of BH3NH3, cyclo-(BH2NH2)(3), and cyclo-(BH2NH2)(3) are updated based on improved heats formation of the boron atom. The G3(MP2) gas phase heats of formation agree reasonably well with the limited amount of reported experimental values for the high energy compounds. The BdN dative bond dissociation energies are in general less than that for BH3-NH3 and in some cases are barely bound with respect to bond dissociation on the free energy scale at 298 K. The heats of formation of the liquids are predicted using the calculated boiling points from COSMO-RS and a modification of the Pictet-Trouton rule for the entropy of vaporization which then allows the heat of vaporization to be estimated as the free energy of a phase change is zero. The heat of formation of the solid was estimated by using 5 kcal/mol for the enthalpy of fusion. Even with this simple estimate, the predicted heats of formation of the solids are in reasonable agreement with the known values. The heats of combustion of the solids were calculated and again reasonable agreement with experiment is found for the limited experimental values.
We review the analytical evaluation of the pressure within the extreme pressure polarizable continuum model for the study of compressed atoms and molecules. Furthermore, we present an effective interpretation of the confinement of the electron density in some noble gas atoms in terms of a simple first-order perturbation analysis of their occupied atomic orbitals.
Coupled cluster Green's function (CCGF) approach has drawn much attention in recent years for targeting the molecular and material electronic structure problems from a many-body perspective in a systematically improvable way. Here, we will present a brief review of the history of how the Green's function method evolved with the wavefunction, early and recent development of CCGF theory, and more recently scalable CCGF software development. We will highlight some of the recent applications of CCGF approach and propose some potential applications that would emerge in the near future.
