Dissociation Of Chemical Bonds Research Articles

Effect of chemical action on shear thickening polishing of YAG

The shear-thickening polishing (STP) method, a flexible polishing technique developed in recent years, was utilized to obtain high surface quality of yttrium aluminum garnet (YAG) crystal effectively. The effect of polishing slurry pH value, which characterizes the chemical action between slurry and YAG crystal, on the material removal rate and surface roughness of YAG crystal was investigated, and the mechanism of the chemical action between YAG and slurry in STP process was revealed through X-ray photoelectron spectroscopy (XPS) analysis. The influence of polishing speed and abrasive concentration, which were relating with mechanical action, on the polishing effects was also studied. The experimental results show that when the pH value of the polishing slurry is lower than 11, the dissociation of chemical bonds and the formation of new bonds on the surface of the YAG crystal are predominantly found, resulting in a lower material removal rate. While the pH value of polishing slurry reaches 13, a reaction layer consisting of Al2SiO5 and Y2SiO5, which has lower hardness than abrasive, is formed on the surface, and this reaction layer can be removed by mechanical action efficiently. The surface roughness Sa decreases from 60 ± 10 nm to 0.38 ± 0.06 nm in 15 min with MRR reaches 117.30 nm/min, which is 4.3 times higher than that of chemical mechanical polishing. Based on the results of XPS analysis, the material removal mechanism of YAG polished by STP is proposed. As revealed by the results, STP with chemical action provide a high efficiency polishing method for the high surface quality YAG crystal.

Enhanced production of hydrogen via catalytic methane decomposition on a Pt7-Ni (110) substrate: a reactive molecular dynamics investigation

Abstract Numerous researchers in the energy field are engaged in a competitive race to advance hydrogen as a clean and environmentally friendly fuel. Studies have been conducted on the different aspects of hydrogen, including its production, storage, transportation and utilization. The catalytic methane decomposition technique for hydrogen production is an environmentally friendly process that avoids generating carbon dioxide gas, which contributes to the greenhouse effect. Catalysts play a crucial role in facilitating rapid, cost-effective and efficient production of hydrogen using this technique. In this study, reactive molecular dynamics simulations were employed to examine the impact of Pt7 cluster decoration on the surface of a Ni (110) catalyst, referred to as Pt7-Ni (110), on the rates of methane dissociation and molecular hydrogen production. The reactive force field was employed to model the atomic interactions that enabled the formation and dissociation of chemical bonds. Our reactive molecular dynamics simulations using the Pt7-Ni (110) catalyst revealed a notable decrease in the number of methane molecules, specifically ~11.89 molecules per picosecond. The rate was approximately four times higher than that of the simulation system utilizing a Ni (110) catalyst and approximately six times higher than that of the pure methane, no-catalyst system. The number of hydrogen molecules generated during a simulation period of 150 000 fs was greater on the Pt7-Ni (110) surface than in both the Ni (110) and pure methane systems. This was due to the presence of numerous dissociated hydrogen atoms on the Pt7-Ni (110) surface.

Comparison of H2O2 and H2O oxidations on TDMAT absorbed on silicon(100) surface during reaction step of ALD–TiO2 process: A DFT study

Atomic layer deposition (ALD) is a thin film technology with high potential in nano fabrication. There is industrial demand to produce an ALD–TiO2 thin film using a low temperature process for applications of bending electronic devices and biomaterials. TDMAT has gained the most attention as a suitable titanium source to grow an ALD–TiO2 film at extremely low temperatures. In this process, almost all published reports have focused on metal-oxide precursors. The current work investigates the effects of oxidizing agents. For the first time, a theoretical investigation was carried out on the oxidation of H2O2 and H2O on TDMAT absorbed on a silicon (100) surface during the reaction step of the ALD–TiO2 process. We simulated the oxidation process, which continued from the adsorption step of the ALD–TiO2 process, where the TDMAT was used as a titanium precursor. The cluster of the Si9H12 =O2 =Ti[N(CH3)]2 surface model was used to represent the remaining TDMAT molecule absorbed on the Si(100) substrate. The reaction mechanisms of H2O2 oxidizing on this surface were evaluated based on their reaction pathways. The reaction pathway of H2O oxidizing on this surface was calculated at the same level to compare the effect of oxidizers. The DFT at the B3LYP 6–311 G+ (2d,p) and 6–31 G(d,p) theoretical levels were used to calculate the potential energy surface and to optimize the electronic structures in this process. It was found that the reaction mechanism depended on the nature of chemical bond dissociation within the oxidizer molecules during the oxidation process. In the oxidation of H2O2, the H2O2 molecule contained O–O and O–H bonds, which were dissociated during the oxidation process, resulting in different product structures. Only the O–O bond dissociation pathway produced the TiO2 structure. The two species of TDMAT ligands at the surface were completely oxidized using a single H2O2 molecule. The H2O2 oxidation produced more –OH groups (double) compared to H2O oxidation. In H2O oxidation, the H2O molecule contained only the O–H bond within its molecule and required activation energies significantly less than for all pathways of H2O2. The HOMO–LUMO properties and IR frequency characteristics (cm−1) of the plausible product structures were discussed. The results should be useful in guiding the selection of suitable precursors for use in the co–reactant system of the ALD–TiO2 process.

Toward DMC Accuracy Across Chemical Space with Scalable Δ-QML.

In the past decade, quantum diffusion Monte Carlo (DMC) has been demonstrated to successfully predict the energetics and properties of a wide range of molecules and solids by numerically solving the electronic many-body Schrödinger equation. With O(N3) scaling with the number of electrons N, DMC has the potential to be a reference method for larger systems that are not accessible to more traditional methods such as CCSD(T). Assessing the accuracy of DMC for smaller molecules becomes the stepping stone in making the method a reference for larger systems. We show that when coupled with quantum machine learning (QML)-based surrogate methods, the computational burden can be alleviated such that quantum Monte Carlo (QMC) shows clear potential to undergird the formation of high-quality descriptions across chemical space. We discuss three crucial approximations necessary to accomplish this: the fixed-node approximation, universal and accurate references for chemical bond dissociation energies, and scalable minimal amons-set-based QML (AQML) models. Numerical evidence presented includes converged DMC results for over 1000 small organic molecules with up to five heavy atoms used as amons and 50 medium-sized organic molecules with nine heavy atoms to validate the AQML predictions. Numerical evidence collected for Δ-AQML models suggests that already modestly sized QMC training data sets of amons suffice to predict total energies with near chemical accuracy throughout chemical space.

The unusual effect of paraffin wax on the thermal decomposition of cyclotetramethylenetetranitramine.

The practical application of paraffin wax (PW) in cyclotetramethylenetetranitramine (HMX)-based polymer-bonded explosive (PBX) requires an understanding of its effect on the thermal decomposition of HMX. In this work, by comparing the thermal decomposition of HMX and a HMX/PW mixture, combined with crystal morphology analysis, molecular dynamics simulation, kinetic analysis, and gas product analysis, the unusual phenomenon and mechanism of the PW effect on the thermal decomposition of HMX were evaluated. During the initial decomposition, PW infiltrates the crystal surface of HMX, reduces the energy barrier for chemical bond dissociation, and induces the decomposition of molecules on HMX crystals, resulting in a lower initial decomposition temperature. PW consumes the active gas produced by HMX with further thermal decomposition and inhibits the dramatic increase of the HMX thermal decomposition rate. In decomposition kinetics, this effect is manifested as PW inhibits transition from an n-order reaction to an autocatalytic reaction.

A Comparative Study on the Combustion Chemistry of Two Bio-hybrid Fuels: 1,3-Dioxane and 1,3-Dioxolane.

Bio-hybrid fuels are a promising solution to accomplish a carbon-neutral and low-emission future for the transportation sector. Two potential candidates are the heterocyclic acetals 1,3-dioxane (C4H8O2) and 1,3-dioxolane (C3H6O2), which can be produced from the combination of biobased feedstocks, carbon dioxide, and renewable electricity. In this work, comprehensive experimental and numerical investigations of 1,3-dioxane and 1,3-dioxolane were performed to support their application in internal combustion engines. Ignition delay times and laminar flame speeds were measured to reveal the combustion chemistry on the macroscale, while speciation measurements in a jet-stirred reactor and ethylene-based counterflow diffusion flames provided insights into combustion chemistry and pollutant formation on the microscale. Comparing the experimental and numerical data using either available or proposed kinetic models revealed that the combustion chemistry and pollutant formation differ substantially between 1,3-dioxane and 1,3-dioxolane, although their molecular structures are similar. For example, 1,3-dioxane showed higher reactivity in the low-temperature regime (500-800 K), while 1,3-dioxolane addition to ethylene increased polycyclic aromatic hydrocarbons and soot formation in high-temperature (>800 K) counterflow diffusion flames. Reaction pathway analyses were performed to examine and explain the differences between these two bio-hybrid fuels, which originate from the chemical bond dissociation energies in their molecular structures.

Design and directed evolution of noncanonical β-stereoselective metalloglycosidases

Metallohydrolases are ubiquitous in nearly all subclasses of hydrolases, utilizing metal elements to activate a water molecule and facilitate its subsequent dissociation of diverse chemical bonds. However, such a catalytic role of metal ions is rarely found with glycosidases that hydrolyze the glycosidic bonds in sugars. Herein, we design metalloglycosidases by constructing a hydrolytically active Zn-binding site within a barrel-shaped outer membrane protein OmpF. Structure- and mechanism-based redesign and directed evolution have led to the emergence of Zn-dependent glycosidases with catalytic proficiency of 2.8 × 109 and high β-stereoselectivity. Biochemical characterizations suggest that the Zn-binding site constitutes a key catalytic motif along with at least one adjacent acidic residue. This work demonstrates that unprecedented metalloenzymes can be tailor-made, expanding the scope of inorganic reactivities in proteinaceous environments, resetting the structural and functional diversity of metalloenzymes, and providing the potential molecular basis of unidentified metallohydrolases and novel whole-cell biocatalysts.

39.2: Charge‐Transfer Exciton Manipulation Based on Hydrogen Bond of Iridium (III) Complexes for Highly Efficient Blue OLED

Phosphorescent iridium emitters have played the significant role in the development of organic light-emitting diodes (OLEDs). In contrast to the green- and red-emitting phosphors, the design and synthesis of high-performance blue emitters remains as a major challenge, owing to the fact that the annihilation of long-lived triplet excited states in the emissive layer, such as triplet-triplet annihilation (TTA) and triplet-polaron annihilation (TPA), could lead to the occurrence of chemical bond dissociation, which eventually caused the degradation of the OLEDs. In this work, blue-emitting vacuum-deposited OLEDs based on these iridium complexes have been realized, in which the devices retain a high EQEmax and CE of 22.01 % and 56.54 cd A−1, respectively, and their efficiency roll-off value around 10 % at the high luminescence level of 1000 cd m−2. To our knowledge, the max brightness value of the device is over 50000 cd m−2, which is the brightness record in the blue phosphorescent OLED range.

On the specific heat and mass loss of thermochemical transition

Poly(vinyl alcohol) PVA is a synthetic polymer used in a variety of applications and exhibits unusual thermal behavior. In this work, differential scanning calorimetry (DSC), isothermal and non-isothermal thermogravimetric analysis (TGA), Fourier transform infrared spectroscopy (FTIR), optical microscopy, viscometry and kinetic modeling were used to study the thermal behavior of PVA for providing further insights about the thermochemical transition. It is shown that the value of specific heat of thermochemical transition is comparable to theoretical values derived from chemical bond dissociation energy values. Activation energy's contribution to the specific heat of thermochemical transition seems to be low. Contradictions among the conclusions from various analytical techniques, regarding the glass transition of PVA, are discussed and interpreted. Settlement of these contradictions leads to the conclusion that like PVA's melting point, its glass transition is also a thermochemical transition but with no detectable mass loss.

Insights into the effects of dynamic high-pressure microfluidization on the structural and rheological properties of rapeseed protein isolate

The application of dynamic high-pressure microfluidization (DHPM) provides interesting modifications in food structures. However, the effects of DHPM on the structural and rheological properties of rapeseed protein isolate (RPI) were scarcely investigated. In this study, the average hydrodynamic size of RPI treated by DHPM significantly declined from 239.2 nm to 170 nm with the pressure and time rising to 60 MPa and 2 min. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis found that DHPM did not affect the molecular mass of the RPI. The changes in the intrinsic fluorescence and circular dichroism spectra showed that tertiary and secondary structures of RPI were altered, as proved by the higher fluorescence intensity and greater conversion of random coils and β-sheets into α-helices compared with the control (unprocessed RPI). The spatial unfolding of the RPI structure and the dissociation of chemical bonds influenced its rheological behavior. All RPI suspensions (15%, w/v) processed by DHPM had lower apparent viscosity, moduli, higher gelling temperature (up to 69.7 °C) than control, and the differences in rheological performance were closely dependent on the treatment pressure and time. In conclusion, through this study of the effects of DHPM on RPI, DHPM was shown to be a potential method for modifying proteins and reducing the consistency of high-protein fluid foods.

N- and B-doped fullerene as peroxidase- and catalase-like metal-free nanozymes with pH-switchable catalytic activity: A first-principles approach

Metal-free nanozymes are designed to function as peroxidase or catalase, controllable by pH conditions. Using first-principles density functional theory (DFT) calculations, we propose that heteroatomic doping of N and B to fullerene is promising to form active sites for catalyzing oxygenate intermediates for H2O2 decomposition. In neutral pH with bare surfaces, both N- and B-doped fullerenes show peroxidase-like catalytic activity rather than a catalase-like one. The presence of water, at neutral pH, drives peroxidase-like catalysis on the B-doped fullerene with marginal energy, in which the B atom functions as a strong electrophile that firmly adsorbs OH* and promotes O-O chemical bond dissociation. Interestingly, acidic, and basic environments tune their selectivity toward peroxidase- and catalase-like catalytic activity, respectively, mediated by pre-adsorbed H* and OH*. The DFT calculations demonstrate the applicability of N- and B-doped fullerenes by showing significantly decreased dissociation energy barriers of H2O2 into 2OH during the oxidation of 3,3′,5,5′-tetramethylbenzidine (TMB). We believe that the results can provide a useful guide for designing highly active and selective carbon-based nanozymes toward H2O2 decomposition.

Optical van-der-Waals forces in molecules: from electronic Bethe-Salpeter calculations to the many-body dispersion model

Molecular forces induced by optical excitations are connected to a wide range of phenomena, from chemical bond dissociation to intricate biological processes that underpin vision. Commonly, the description of optical excitations requires the solution of computationally demanding electronic Bethe-Salpeter equation (BSE). However, when studying non-covalent interactions in large-scale systems, more efficient methods are desirable. Here we introduce an effective approach based on coupled quantum Drude oscillators (cQDO) as represented by the many-body dispersion model. We find that the cQDO Hamiltonian yields semi-quantitative agreement with BSE calculations and that both attractive and repulsive optical van der Waals (vdW) forces can be induced by light. These optical-vdW interactions dominate over vdW dispersion in the long-distance regime, showing a complexity that grows with system size. Evidence of highly non-local forces in the human formaldehyde dehydrogenase 1MC5 protein suggests the ability to selectively activate collective molecular vibrations by photoabsorption, in agreement with recent experiments.

Probing of sequential atmospheric degradation of chlorine radical initiated 1,8-cineole in the presence of O2 and NO radical with the emission of secondary pollution

1,8-cineole is an essential monoterpene cyclic ether which is released into the troposphere by many types of plants. It interacts with several atmospheric oxidants because of which is removed from the troposphere via oxidation. The oxidation of 1,8-cineole with Cl radical and the subsequent addition of atmospheric O2 and NO radical with the intermediates are studied using the quantum chemical method. Further, the thermodynamic parameters of 1,8-cineole, such as enthalpy and Gibbs free energy are calculated for all initial and subsequent reactions to facilitate perspicacity. The dissociation and formation of chemical bonds during H abstraction from 1,8-cineole at C2, C6, and C8 sites are described using Mayer bond order analysis. The reaction force analysis demonstrates that the structural rearrangement is dominant with the yield percentages of 85%, 50.80%, and 96.9% over electron reordering with the yield percentages of 15%, 49.19%, and 3.03% respectively in the H abstraction reaction of 1,8-cineole. In the temperature range of 278–350 K, the total CVT/SCT rate constant is calculated to be 2.94 × 10−12 cm3/molecule/sec, which is consistent with the experimentally available value of 2.2 × 10−10 cm3/molecule/sec. At 298 K, branching ratios of rate constant of alkyl radical intermediates I1A, I1B, and I1C are calculated with the percentage of 42.19%, 21.52%, and 36.29% respectively, which suggest that the Cl addition to the C2 site contributes more to the total rate constant rather than the other two sites (C6 and C8). The lifetime of 1,8-cineole is calculated to be 5.2 weeks, implies that the 1,8-cineole may be readily destroyed in the atmosphere after it is released. Secondary pollutants formed from this degradation mechanism may be harmful to the environment and the living things.

Influence of reactive oxygen species concentration and ambient temperature on the evolution of chemical bonds during plasma cleaning: a molecular dynamics simulation.

The research on plasma chemistry involved in the formation and dissociation of abundant chemical bonds is fundamental to developing plasma cleaning. To understand the influence of reactive oxygen species' concentration and ambient temperature on the evolution behavior of the chemical bond during plasma cleaning, microscopic reaction models between organic contaminants and reactive oxygen species were established and performed by reactive molecular dynamics. Dibutyl phthalate, as a representative organic contaminant, was selected as the research object. The simulation results suggested that hydrogen bonds between hydroxyl radicals reduced the mobility of reactive species, resulting in the cleaning ability of hydroxyl radicals being much lower than atomic oxygen and ozone radicals. The concentration of reactive species dominated the efficiency of plasma cleaning, and the increase in ambient temperature further improved the cleaning ability. C-H, C-C and C-O bonds were gradually oxidized to C[double bond, length as m-dash]C, C-O, C[double bond, length as m-dash]O and O-H bonds by hydrogen abstraction reaction during the reaction of reactive species with organic contaminants. An increase in ambient temperature induced the possibility of benzene ring destruction under the action of reactive species, which was considered a method of complete dissociation of aromatic hydrocarbons.

Study on Chemical Bond Dissociation and the Removal of Oxygen-Containing Functional Groups of Low-Rank Coal during Hydrothermal Carbonization: DFT Calculations.

The molecular structure model of lignite was constructed, and the dissociation and removal mechanism of different C–O bonds and oxygen-containing functional groups was investigated using density functional theory (DFT) calculations. First, the bond order and bond dissociation enthalpy (BDE) were analyzed to predict the strength of different chemical bonds, and differences in the BDE and bond order were related to the difference in the fragment structure and electronic effects. The first group to break during hydrothermal carbonization (HTC) is the methyl of Ph(CO)O–CH3, followed by the C–O of CH3–OC(O)OH; the hydroxyl in Ph–OH is the most thermally stable group, followed by the hydroxyl in CH3OC(O)–OH. In addition, the orbital localization analysis has also been carried out. All three chemical bonds of Ph(CO)OCH3 show the characteristics of σ bond, while Ph(C=O)OCH3 and Ph(CO)–OCH3 with the Mayer bond order (MBO) greater than 1 also contains certain π bond characteristics. The lignite van der Waals (vdW) surface electrostatic potential (ESP) was constructed and visualized, and the results showed that the oxygen-containing functional groups mainly contributed to the area with a large absolute ESP. Finally, weak interactions between water molecules and lignite at different sites were described by independent gradient model (IGM) analysis. Models A, B, and E formed weak interactions with the hydrogen bond as the main force; model E showed the weakest hydrogen bond, while model C showed van der Waals interaction as the dominant force. In addition, some steric effect was also observed in model D.

First Principles Dynamics Study of Excited Hole Relaxation in DNA.

A recent theoretical work showed that ion irradiation generates excited holes deep within the valence band of DNA. In this work, we investigate the excited hole relaxation toward HOMO using a first-principles computational method following such ionization events. The excited hole relaxation is found to depend significantly on the energetic position of the excited hole generated. The relaxation process is found to be an order of magnitude slower for holes that are generated deeper than 20 eV than those generated within 10 eV, where the probability for the initial ionization events is the highest. However, the excited holes that are generated in different spatial moieties such as DNA nucleotide bases and phosphate backbones do not show noticeable differences in terms of the relaxation time. Our work also shows that decoherence due to nuclei dynamics slows down the relaxation by a factor of two or more. At the same time, the relaxation time is found to be less than a couple of picoseconds, much shorter than typical timescales associated with chemical bond dissociation.

Study of carbon dioxide adsorption on H-saturated porous graphene sheet and its separation from nitrogen using functional density theory and reactive molecular dynamics

The adsorption of carbon dioxide molecules on four H-saturated porous graphene sheets with different pore sizes and a poreless graphene sheet was investigated and compared with the adsorption of nitrogen molecules on them. Reactive molecular dynamics was used in this study, which took into account the possibility of chemical bond formation and dissociation as well as the effects of polarity. This research demonstrates that all porous graphene sheets and non-cavity graphene sheets absorb carbon dioxide molecules more than nitrogen molecules and can be used to separate these two gases. However, the size and shape of the cavities have no significant impact on gas molecule adsorption on these plates.

(Invited) A Multi-Scale Infrastructure for Automating Materials Science Computations

In this talk, I will present our group’s effort to build a high-throughput infrastructure for the automated execution of materials science computations and analysis. This work is an extension of open-source Python packages that have been developed in Materials Project [1]: (1) pymatgen [2] for structure representation and input/output files generation and handling, (2) FireWorks [3] for managing workflows over computing resources, and (3) Custodian [4] for monitoring inevitable errors during the simulations and applying on-the-fly fixes. In our work, we benefit from these three libraries to build well-tested workflows for speeding up the prediction and screening of materials properties relevant to various applications with a focus on battery systems. At the backend, the infrastructure interfaces with the Gaussian [5] software which enables electronic structure calculations of chemical systems, and LAMMPS [6] open-source code for molecular dynamics (MD) simulations. In addition to its ability to handle the generation of input/data files and parsing of output files from the mentioned computational software, the infrastructure allows management of the collected data and storing it in MongoDB [7], a NoSQL database program using JSON-like documents with flexible schema. Examples of implemented Gaussian-based workflows include the calculation of electrostatic partial charges, NMR chemical shift, binding energy, ionization potential, electron affinity, and bond dissociation energy. Each derived property is saved in its own collection with auxiliary information like molecular metadata (smiles representation, chemical formula, ...), which makes it easy to query and data-mine structure-property relationships. The user can tune the calculations by overriding default workflow parameters, for example, the functional and basis set, by-passing selected steps, or packing many jobs over multiple nodes for supercomputing resources. LAMMPS workflows allow the execution of MD simulations in different ensembles and analysis of the dumped trajectories for various dynamical and structural properties. The infrastructure enables coupling first-principles calculations with classical MD simulations to address chemical and physical phenomena occurring in a given system at different length and time scales. This approach aids in the development of databases required for training machine-learning models to accelerate the design and screening of materials with optimized properties for various materials science applications.References Jain, A., et al., Commentary: The Materials Project: A materials genome approach to accelerating materials innovation. Apl Materials, 2013. 1(1): p. 011002.Ong, S.P., et al., Python Materials Genomics (pymatgen): A robust, open-source python library for materials analysis. Computational Materials Science, 2013. 68: p. 314-319.Jain, A., et al., FireWorks: a dynamic workflow system designed for high‐throughput applications. Concurrency and Computation: Practice and Experience, 2015. 27(17): p. 5037-5059.Custodian. <https://github.com/materialsproject/custodian>.Frisch, M.J., et al., Gaussian 16 Rev. C.01. 2016: Wallingford, CT.Plimpton, S., Fast parallel algorithms for short-range molecular dynamics. 1993, Sandia National Labs., Albuquerque, NM (United States).MongoDB Inc., M., 2014.

Amino acids adsorption onto the (111) surface of cubic zirconia: a density functional theory study

The adsorption of six amino acids onto (111) surface of cubic ZrO2 was theoretically investigated using density functional theory (DFT). DFT calculations with D3 dispersion correction were carried out with the PBE exchange–correlation functional. Mechanism, molecular adsorption structures, and energies were calculated. The (111) surface of cubic ZrO2 was modeled using a slab with 9 atomic layers (81 atoms). The slab thickness was 7.3 A, and a vacuum separation of 12 A was used with periodic boundary conditions. The results showed that the order of the molecular adsorption energies is ASP > HYP > GLY > ALA > LYS > PRO. Furthermore, adsorption of amino acids on the zirconia surface showed complete hydrogen transfer from GLY, ALA, HYP, and ASP to the surface. For LYS and PRO, hydrogen transfer does not occur. Topological analysis of the electron localization function (ELF) was applied to study the nature of the molecule-surface interactions. It was found that the dissociation of molecules and strong chemical bonds of adsorbed atoms to surface oxygen atoms play an important role in the adsorption mechanism.

Automation of Active Space Selection for Multireference Methods via Machine Learning on Chemical Bond Dissociation.

Predicting and understanding the chemical bond is one of the major challenges of computational quantum chemistry. Kohn-Sham density functional theory (KS-DFT) is the most common method, but approximate density functionals may not be able to describe systems where multiple electronic configurations are equally important. Multiconfigurational wave functions, on the other hand, can provide a detailed understanding of the electronic structures and chemical bonds of such systems. In the complete active space self-consistent field (CASSCF) method, one performs a full configuration interaction calculation in an active space consisting of active electrons and active orbitals. However, CASSCF and its variants require the selection of these active spaces. This choice is not black box; it requires significant experience and testing by the user, and thus active space methods are not considered particularly user-friendly and are employed only by a minority of quantum chemists. Our goal is to popularize these methods by making it easier to make good active space choices. We present a machine learning protocol that performs an automated selection of active spaces for chemical bond dissociation calculations of main group diatomic molecules. The protocol shows high prediction performance for a given target system as long as a properly correlated system is chosen for training. Good active spaces are correctly predicted with a considerably better success rate than random guess (larger than 80% precision for most systems studied). Our automated machine learning protocol shows that a "black-box" mode is possible for facilitating and accelerating the large-scale calculations on multireference systems where single-reference methods such as KS-DFT cannot be applied.