RESUMEN
CONTEXT: The study of platinum (Pt) clusters and nanoparticles is essential due to their extensive range of potential technological applications, particularly in catalysis. The electronic properties that yield optimal catalytic performance at the nanoscale are significantly influenced by the size and structure of Pt clusters. This research aimed to identify the lowest-energy conformers for Pt 18 , Pt 19 , and Pt 20 species using Density Functional Theory (DFT). We discovered new low-symmetry conformers for Pt 19 and Pt 20 , which are 3.0 and 1.0 kcal/mol more stable, respectively, than previously reported structures. Our study highlights the importance of using density functional approximations that incorporate moderate levels of exact Hartree-Fock exchange, alongside basis sets of at least quadruple-zeta quality. The resulting structures are asymmetric with varying active sites, as evidenced by sigma hole analysis on the electrostatic potential surface. This suggests a potential correlation between electronic structure and catalytic properties, warranting further investigation. METHODS: An equivariant graph neural network interatomic potential (NequIP) within the Atomic Simulation Environment suite (ASE) was used to provide initial geometries of the aggregates under study. DFT calculations were performed with the ORCA 5 package, using functional approximations that included Generalized Gradient Approximation (PBE), meta-GGA (TPSS, M06-L), hybrid (PBE0, PBEh), meta-GGA hybrid (TPSSh), and range-separated hybrid ( ω B97x) functionals. Def2-TZVP and Def2-QZVP as well as members of the cc-pwCVXZ-PP family to check basis set convergence were used. QTAIM calculations were performed using the AIMAll suite. Structures were visualized with the AVOGADRO code.
RESUMEN
C[C4H4], the simplest compound of the [4]-pyramidane family, has so far eluded experimental characterization, although several of its analogs E[C4(SiMe3)4], in which the E apex atom is a tetrel group element, have been successfully prepared. The non-classical bonding mode of E, similar to that found in propellanes, has prompted a considerable number of theoretical studies to unravel the nature of the apex-base interaction. Here, we contribute to this knowledge by analyzing the electron localization function (ELF) and classical QTAIM descriptors; as well the statistical distribution of electrons in atomic regions by means of the so-called electron distribution functions (EDFs), calculation of multicenter indices (MCI) as aromaticity descriptors and by performing orbital invariant energy decompositions with the interacting quantum atoms (IQA) approach on a series of E[C4(SiMe3)4] compounds. We find that the bonding evolves from covalent to electrostatic as E changes from C to Pb, with an anomaly when E=Si, which is shown to be the most charged moiety, compatible with an aromatic [C4(SiMe3)4]2- scaffold in the pyramidane base.
RESUMEN
Nanoclusters exhibit electronic, optical, and magnetic properties that differ significantly from those of extended and molecular systems with comparable stoichiometries. In this work, we examined the structural, energetic, and electronic characteristics of yttrium-doped boron clusters (YBn, n =2-14) with robust wavefunction analysis tools. Special emphasis is placed on the elucidation of the potential aromatic character exhibited by the resultant molecules and how it can affect their chemical bonding and stability. Our results revealed that the YBn stability is governed by the maximization of the ionic Y-B interactions. This is evidenced from the lowest-energy conformations, which manifest as half-sandwich structures wherein the majority of boron atoms are bonded to yttrium. The stabilization of such chemical contacts comes at the expense of a notorious depletion of the Y local electron density, crystallizing in a considerable ionic character, close to Y2+ + Bn2-. Such a charge transfer is coupled to the enhancement of the electron delocalization within the YBn lattice, resulting in quite remarkable local and global aromatic characters. Altogether, this study shows how the toolkit of real space chemical bonding descriptors can offer valuable insights into the structural and electronic properties, of YBn clusters, contributing to a better understanding of their behavior.
RESUMEN
The x-ray restrained wave function (XRW) method is a quantum crystallographic technique to extract wave functions compatible with experimental x-ray diffraction data. The approach looks for wave functions that minimize the energies of the investigated systems and also reproduce sets of x-ray structure factors. Given the strict relationship between x-ray structure factors and electron distributions, the strategy practically allows determining wave functions that correspond to given (usually experimental) electron densities. In this work, the capabilities of the XRW approach were further tested. The aim was to evaluate whether the XRW technique could serve as a tool for suggesting new exchange-correlation functionals for density functional theory or refining existing ones. Additionally, the ability of the method to address the influences of the crystalline environment was also assessed. The outcomes of XRW computations were thus compared to those of traditional gas-phase, embedding quantum mechanics/molecular mechanics, and fully periodic calculations. The results revealed that, irrespective of the initial conditions, the XRW computations practically yield a consensus electron density, in contrast to the currently employed density functional approximations (DFAs), which tend to give a too large range of electron distributions. This is encouraging in view of exploiting the XRW technique to develop improved functionals. Conversely, the calculations also emphasized that the XRW method is limited in its ability to effectively address the influences of the crystalline environment. This underscores the need for a periodic XRW technique, which would allow further untangling the shortcomings of DFAs from those inherent to the XRW approach.
RESUMEN
Machine-learned computational chemistry has led to a paradoxical situation in which molecular properties can be accurately predicted, but they are difficult to interpret. Explainable AI (XAI) tools can be used to analyze complex models, but they are highly dependent on the AI technique and the origin of the reference data. Alternatively, interpretable real-space tools can be employed directly, but they are often expensive to compute. To address this dilemma between explainability and accuracy, we developed SchNet4AIM, a SchNet-based architecture capable of dealing with local one-body (atomic) and two-body (interatomic) descriptors. The performance of SchNet4AIM is tested by predicting a wide collection of real-space quantities ranging from atomic charges and delocalization indices to pairwise interaction energies. The accuracy and speed of SchNet4AIM breaks the bottleneck that has prevented the use of real-space chemical descriptors in complex systems. We show that the group delocalization indices, arising from our physically rigorous atomistic predictions, provide reliable indicators of supramolecular binding events, thus contributing to the development of Explainable Chemical Artificial Intelligence (XCAI) models.
RESUMEN
Despite the importance of the one-particle picture provided by the orbital paradigm, a rigorous understanding of the spatial distribution of electrons in molecules is still of paramount importance to chemistry. Considerable progress has been made following the introduction of topological approaches, capable of partitioning space into chemically meaningful regions. They usually provide atomic partitions, for example, through the attraction basins of the electron density in the quantum theory of atoms in molecules (QTAIM) or electron-pair decompositions, as in the case of the electron localization function (ELF). In both cases, the so-called electron distribution functions (EDFs) provide a rich statistical description of the electron distribution in these spatial domains. Here, we take the EDF concept to a new fine-grained limit by calculating EDFs in the QTAIM â© ELF intersection domains. As shown in AHn systems based on main group elements, as well as in the CO, NO, and BeO molecules, this approach provides an exquisitely detailed picture of the electron distribution in molecules, allowing for an insightful combination of the distribution of electrons between Lewis entities (such as bonds and lone pairs) and atoms at the same time. Besides mean-field calculations, we also explore the impact of electron correlation through Hartree-Fock (HF), density functional theory (DFT) (B3LYP), and CASSCF calculations.
RESUMEN
Condensing the many physical variables defining a chemical system into a fixed-size array poses a significant challenge in the development of chemical Machine Learning (ML). Atom Centered Symmetry Functions (ACSFs) offer an intuitive featurization approach by means of a tedious and labor-intensive selection of tunable parameters. In this work, we implement an unsupervised ML strategy relying on a Gaussian Mixture Model (GMM) to automatically optimize the ACSF parameters. GMMs effortlessly decompose the vastness of the chemical and conformational spaces into well-defined radial and angular clusters, which are then used to build tailor-made ACSFs. The unsupervised exploration of the space has demonstrated general applicability across a diverse range of systems, spanning from various unimolecular landscapes to heterogeneous databases. The impact of the sampling technique and temperature on space exploration is also addressed, highlighting the particularly advantageous role of high-temperature Molecular Dynamics (MD) simulations. The reliability of the resulting features is assessed through the estimation of the atomic charges of a prototypical capped amino acid and a heterogeneous collection of CHON molecules. The automatically constructed ACSFs serve as high-quality descriptors, consistently yielding typical prediction errors below 0.010 electrons bound for the reported atomic charges. Altering the spatial distribution of the functions with respect to the cluster highlights the critical role of symmetry rupture in achieving significantly improved features. More specifically, using two separate functions to describe the lower and upper tails of the cluster results in the best performing models with errors as low as 0.006 electrons. Finally, the effectiveness of finely tuned features was checked across different architectures, unveiling the superior performance of Gaussian Process (GP) models over Feed Forward Neural Networks (FFNNs), particularly in low-data regimes, with nearly a 2-fold increase in prediction quality. Altogether, this approach paves the way toward an easier construction of local chemical descriptors, while providing valuable insights into how radial and angular spaces should be mapped. Finally, this work opens the possibility of encoding many-body information beyond angular terms into upcoming ML features.
Asunto(s)
Simulación de Dinámica Molecular , Aprendizaje Automático no Supervisado , Distribución Normal , AutomatizaciónRESUMEN
Acetylenes are essential building blocks in modern chemistry due to their remarkable modularity. The introduction of heteroatoms, such as pnictogens (X), is one of the simplest approaches to altering the C≡C bond. However, the chemistry of the resultant dipnictogenoacetylenes (DXAs) is strongly dependent on the nature of X. In this work, rigorous theoretical chemistry tools are employed to shed light on the origin of these differences, providing a detailed evaluation of the impact of X on the geometrical and electronic features of DXAs. Special emphasis is made on the study of the carbene character of the systems through the analysis of the interconversion mechanism between the linear and zigzag isomers. Our results show that second-period atoms behave drastically differently to the remaining X: down the group, a zwitterionic resonance form emerges at the expense of decreasing the carbenoid role, eventually resulting in an electrostatically driven ring closure. Furthermore, our findings pave the way to potentially unveiling novel routes for the promotion of free-radical chemistry.
RESUMEN
Evidence that the electronic structure of atoms persists in molecules to a much greater extent than has been usually admitted is presented. This is achieved by resorting to N-electron real-space descriptors instead of one- or at most two-particle projections like the electron or exchange-correlation densities. Here, the 3N-dimensional maxima of the square of the wavefunction, the so-called Born maxima, are used. Since this technique is relatively unknown to the crystallographic community, a case-based approach is taken, revisiting first the Born maxima of atoms in their ground state and then some of their excited states. It is shown how they survive in molecules and that, beyond any doubt, the distribution of electrons around an atom in a molecule can be recognized as that of its isolated, in many cases excited, counterpart, relating this fact with the concept of energetic promotion. Several other cases that exemplify the applicability of the technique to solve chemical bonding conflicts and to introduce predictability in real-space analyses are also examined.
RESUMEN
The topology of the Ehrenfest force field (EhF) is investigated as a tool for describing local interactions in molecules and intermolecular complexes. The EhF is obtained by integrating the electronic force operator over the coordinates of all but one electron, which requires knowledge of both the electron density and the reduced pair density. For stationary states, the EhF can also be obtained as minus the divergence of the kinetic stress tensor, although this approach leads to well-documented erroneous asymptotic behavior at large distances from the nuclei. It is shown that these pathologies disappear using the electron density functions and that the EhF thus obtained displays the correct behavior in real space, with no spurious critical points or attractors. Therefore, its critical points can be unambiguously obtained and classified. Test cases, including strained molecules, isomerization reactions, and intermolecular interactions, were analyzed. Various chemically relevant facts are highlighted: for example, non-nuclear attractors are generally absent, potential hydrogen-hydrogen interactions are detected in crowded systems, and a bifurcation mechanism is observed in the isomerization of HCN. Moreover, the EhF atomic basins are less charged than those of the electron density. Although integration of the EhF over regions of real space can also be performed to yield the corresponding atomic forces, several numerical drawbacks still need to be solved if electron density functions are to be used for that purpose. Overall, the results obtained support the Ehrenfest force field as a reliable descriptor for the definition of atomic basins and molecular structure.
RESUMEN
Nanoclusters represent a connection between (i) solid state systems and (ii) species in the atomic and molecular domains. Additionally, nanoclusters can also have very interesting electronic, optical and magnetic properties. For example, some aluminium clusters behave as superatoms and the doping of these clusters might strengthen their adsorption capabilities. Thus, we address herein the structural, energetic and electronic characterisation of scandium-doped aluminium clusters (AlnSc (n = 1-24)) by means of density functional theory calculations and quantum chemical topology wave function analyses. We studied the effect of Sc-doping on the structure and charge distribution by considering pure Al clusters as well. The quantum theory of atoms in molecules (QTAIM) reveals that interior Al atoms have large negative atomic charges (≈2a.u.) and hence the atoms surrounding them are considerably electron deficient. The Interacting Quantum Atoms (IQA) energy partition allowed us to establish the nature of the interaction between the Al13 superatom and the Al12Sc cluster with Al to form the complexes Al14 and Al13Sc, respectively. We also used the IQA approach to examine (i) the influence of Sc on the geometry of the AlnSc complexes along with (ii) the cooperative effects in the binding of AlnSc and Aln+1 clusters. We also exploited the QTAIM and IQA approaches to study the interaction of the electrophilic surface of the examined systems with CO2. Overall, we observe that the investigated Sc-doped Al complexes with a marked stability towards disproportionation reactions exhibit strong adsorption energies with CO2. Concomitantly, the carbon dioxide molecule is considerably distorted and destabilised, conditions which might prepare it for further chemical reactions. Altogether, this paper gives valuable insights on the tuning of the properties of metallic clusters for their design and exploitation in custom-made materials.
RESUMEN
The Quantum Theory of Atoms in Molecules (QTAIM) provides an intuitive, yet physically sound, strategy to determine the partial charges of any chemical system relying on the topology induced by the electron density ρ(r) . In a previous work [J. Chem. Phys. 2022, 156, 014112], we introduced a machine learning (ML) model for the computation of QTAIM charges of C, H, O, and N atoms at a fraction of the conventional computational cost. Unfortunately, the independent nature of the atomistic predictions implies that the raw atomic charges may not necessarily reconstruct the exact molecular charge, limiting the applicability of the latter in the chemistry realm. Trying to solve such an inconvenience, we introduce NNAIMGUI, a user-friendly code which combines the inferring abilities of ML with an equilibration strategy to afford adequately behaved partial charges. The performance of this approach is put to the test in a variety of scenarios including interpolation and extrapolation regimes (e.g chemical reactions) as well as large systems. The results of this work prove that the equilibrated charges retain the chemically accurate behavior reproduced by the ML models. Furthermore, NNAIMGUI is a fully flexible architecture allowing users to train and use tailor-made models targeted at any atomic property of choice. In this way, the GUI-interfaced code, equipped with visualization utilities, makes the computation of real-space atomic properties much more appealing and intuitive, paving the way toward the extension of QTAIM related descriptors beyond the theoretical chemistry community.
Asunto(s)
Aprendizaje Automático , Programas Informáticos , Teoría CuánticaRESUMEN
In this perspective, we review some recent advances in the concept of atoms-in-molecules from a real space perspective. We first introduce the general formalism of atomic weight factors that allows unifying the treatment of fuzzy and non-fuzzy decompositions under a common algebraic umbrella. We then show how the use of reduced density matrices and their cumulants allows partitioning any quantum mechanical observable into atomic or group contributions. This circumstance provides access to electron counting as well as energy partitioning, on the same footing. We focus on how the fluctuations of atomic populations, as measured by the statistical cumulants of the electron distribution functions, are related to general multi-center bonding descriptors. Then we turn our attention to the interacting quantum atom energy partitioning, which is briefly reviewed since several general accounts on it have already appeared in the literature. More attention is paid to recent applications to large systems. Finally, we consider how a common formalism to extract electron counts and energies can be used to establish an algebraic justification for the extensively used bond order-bond energy relationships. We also briefly review a path to recover one-electron functions from real space partitions. Although most of the applications considered will be restricted to real space atoms taken from the quantum theory of atoms in molecules, arguably the most successful of all the atomic partitions devised so far, all the take-home messages from this perspective are generalizable to any real space decompositions.
RESUMEN
Chemical bonding theory is of utmost importance to chemistry, and a standard paradigm in which quantum mechanical interference drives the kinetic energy lowering of two approaching fragments has emerged. Here we report that both internal and external reference biases remain in this model, leaving plenty of unexplored territory. We show how the former biases affect the notion of wavefunction interference, which is purportedly recognized as the most basic bonding mechanism. The latter influence how bonding models are chosen. We demonstrate that the use of real space analyses are as reference-less as possible, advocating for their use. Delocalisation emerges as the reference-less equivalent to interference and the ultimate root of bonding. Atoms (or fragments) in molecules should be understood as a statistical mixture of components differing in electron number, spin, etc.
RESUMEN
Real space chemical bonding descriptors, such as the electron localization function or the Laplacian of the electron density, have been widely used in electronic structure theory thanks to their power to provide chemically intuitive spatial images of bonded and non-bonded interactions. This capacity stems from their ability to display the shell structure of atoms and its distortion upon molecular formation. Here, we examine the spatial position of the N electrons of an atom at the maximum of the square of the wavefunction, the so-called Born maximum, as a shell structure descriptor for ground state atoms with Z = 1-36, comparing it to other available indices. The maximization is performed with the help of variational quantum Monte Carlo calculations. We show that many electron effects (mainly Pauli driven) are non-negligible, that Born shells are closer to the nucleus than any other of the examined descriptors, and that these shells are very well preserved in simple molecules.
RESUMEN
Recent discovery of an unusual bond between Na and B in NaBH3- motivated us to look for potentially similar bonds, which remained unnoticed among systems isoelectronic with NaBH3-. Here, we report a novel family of collective interactions and a measure called exchange-correlation interaction collectivity index (ICIXC; [Formula: see text]) to characterize the extent of collective versus pairwise bonding. Unlike conventional bonds in which ICIXC remains close to one, in collective interactions ICIXC may approach zero. We show that collective interactions are commonplace among widely used organometallics, as well as among boron and aluminum complexes with the general formula [Ma+AR3]b- (A: C, B or Al). In these species, the metal atom interacts more efficiently with the substituents (R) on the central atoms than the central atoms (A) upon forming efficient collective interactions. Furthermore, collective interactions were also found among fluorine atoms of XFn systems (X: B or C). Some of organolithium and organomagnesium species have the lowest ICIXC among the more than 100 studied systems revealing the fact that collective interactions are rather a rule than an exception among organometallic species.
RESUMEN
Bimolecular nucleophilic substitution (SN2) and elimination (E2) reactions are prototypical examples of competing reaction mechanisms, with fundamental implications in modern chemical synthesis. Steric hindrance (SH) is often considered to be one of the dominant factors determining the most favorable reaction out of the SN2 and E2 pathways. However, the picture provided by classical chemical intuition is inevitably grounded on poorly defined bases. In this work, we try to shed light on the aforementioned problem through the analysis and comparison of the evolution of the steric energy (EST), settled within the IQA scheme and experienced along both reaction mechanisms. For such a purpose, the substitution and elimination reactions of a collection of alkyl bromides (R-Br) with the hydroxide anion (OH-) were studied in the gas phase at the M06-2X/aug-cc-pVDZ level of theory. The results show that, generally, EST recovers the appealing trends already anticipated by chemical intuition and organic chemistry, supporting the role that SH is classically claimed to play in the competition between SN2 and E2 reactions.
RESUMEN
Within substitution reactions, the Bimolecular Nucleophilic Substitution (SN 2) reaction mechanism is one of the most frequently found and studied ones. Among other factors, the easiness of the SN 2 pathway is classically considered to be determined by steric hindrance. However, the diffuse nature of the latter inevitably darkens these and other arguments holding the pillars of chemical intuition. In this work, we employ the steric energy (EST ) descriptor, formulated within the Interacting Quantum Atoms approach, to offer insights regarding this problem. The steric demands of the substrate, nucleophile and leaving group were studied using the gas-phase SN 2 reaction with different organic skeletons (CH3 , CH3 CH2 , (CH3 )2 CH, (CH3 )3 C, (CH3 )3 CCH2 ) and halogens (F, Cl, and Br) as test-bed systems. Our results show that, according to EST , the SH experienced along these simple reactions fits, in the general case, the trends predicted by a meticulous and rigorous application of chemical intuition. However, steric clash alone should not be considered as the only argument used to explain the easiness of the SN 2 reaction over different electrophiles.
RESUMEN
The interacting quantum atoms (IQA) method decomposes the quantum mechanical (QM) energy of a molecular system in terms of one- and two-center (atomic) contributions within the context of the quantum theory of atoms in molecules. Here, we demonstrate that IQA, enhanced with molecular mechanics (MM) and Poisson-Boltzmann surface-area (PBSA) solvation methods, is naturally extended to the realm of hybrid QM/MM methodologies, yielding intra- and inter-residue energy terms that characterize all kinds of covalent and noncovalent bonding interactions. To test the robustness of this approach, both metal-water interactions and QM/MM boundary artifacts are characterized in terms of the IQA descriptors derived from QM regions of varying size in Zn(II)- and Mg(II)-water clusters. In addition, we analyze a homologous series of inhibitors in complex with a matrix metalloproteinase (MMP-12) by carrying out QM/MM-PBSA calculations on their crystallographic structures followed by IQA energy decomposition. Overall, these applications not only show the advantages of the IQA QM/MM approach but also address some of the challenges lying ahead for expanding the QM/MM methodology.