Combining Quantum Mechanics and Machine-Learning Calculations for Anharmonic Corrections to Vibrational Frequencies

Prediction of the Infrared Absorbance Intensities and Frequencies of Hydrocarbons: A Message Passing Neural Network Approach

Accurately and efficiently predicting the infrared (IR) spectra of a molecule can provide insights into the structure-properties relationships of molecular species, which has led to a proliferation of machine learning tools designed for this purpose. However, earlier studies have focused primarily on obtaining normalized IR spectra, which limits their potential for a comprehensive analysis of molecular behavior in the IR range. For instance, to fully understand and predict the optical properties, such as the transparency characteristics, it is necessary to predict the molar absorptivity IR spectra instead. Here, we propose a graph-based communicative message passing neural network algorithm that can predict both the peak positions and absolute intensities corresponding to density functional theory calculated molar absorptivities in the IR domain. By modifying existing spectral loss functions, we show that our method is able to predict with DFT-accuracy level the IR molar absorptivities of a series of hydrocarbons containing up to ten carbon atoms and apply the model to a set of larger molecules. We also compare the predicted spectra with those generated by the direct message passing neural network. The results suggest that both algorithms demonstrate similar predictive capabilities for hydrocarbons, indicating that either model could be effectively used in future research on spectral prediction for such systems.

Machine-learning to predict anharmonic frequencies: a study of models and transferability

With more and more accurate electronic structure methods at hand, the inclusion of anharmonic effects in the post-processing of such data towards thermochemical properties is the next step. In this context, the description of anharmonicity has been an important topic of physical chemistry and chemical physics for a long time. In this study, anharmonic frequencies of various hydrogen-halides and halogenated hydrocarbon molecular clusters are calculated using harmonic as well as explicitly anharmonic methods, i.e., normal mode analysis and vibrational self-consistent field. Simple harmonic model based descriptors were used to predict anharmonic frequencies via multilinear regression and gradient boosting regression. Gradient boosting regression is capable of predicting reliable anharmonic data and even the simple multilinear regression model yields reasonable predictions that can account for mode-to-mode couplings. Moreover, the transferability to unseen chemical systems is assessed and it is confirmed that the machine-learned models can be applied to larger, unseen molecules.

Graphormer-IR: Graph Transformers Predict Experimental IR Spectra Using Highly Specialized Attention

Infrared (IR) spectroscopy is an important analytical tool in various chemical and forensic domains and a great deal of effort has gone into developing in silico methods for predicting experimental spectra. A key challenge in this regard is generating highly accurate spectra quickly to enable real-time feedback between computation and experiment. Here, we employ Graphormer, a graph neural network (GNN) transformer, to predict IR spectra using only simplified molecular-input line-entry system (SMILES) strings. Our data set includes 53,528 high-quality spectra, measured in five different experimental media (i.e., phases), for molecules containing the elements H, C, N, O, F, Si, S, P, Cl, Br, and I. When using only atomic numbers for node encodings, Graphormer-IR achieved a mean test spectral information similarity (SISμ) value of 0.8449 ± 0.0012 (n = 5), which surpasses that the current state-of-the-art model Chemprop-IR (SISμ = 0.8409 ± 0.0014, n = 5) with only 36% of the encoded information. Augmenting node embeddings with additional node-level descriptors in learned embeddings generated through a multilayer perceptron improves scores to SISμ = 0.8523 ± 0.0006, a total improvement of 19.7σ (t = 19). These improved scores show how Graphormer-IR excels in capturing long-range interactions like hydrogen bonding, anharmonic peak positions in experimental spectra, and stretching frequencies of uncommon functional groups. Scaling our architecture to 210 attention heads demonstrates specialist-like behavior for distinct IR frequencies that improves model performance. Our model utilizes novel architectures, including a global node for phase encoding, learned node feature embeddings, and a one-dimensional (1D) smoothing convolutional neural network (CNN). Graphormer-IR's innovations underscore its value over traditional message-passing neural networks (MPNNs) due to its expressive embeddings and ability to capture long-range intramolecular relationships.

The effect of machine learning predicted anharmonic frequencies on thermodynamic properties of fluid hydrogen fluoride

Anharmonic effects play a crucial role in determining thermochemical properties of liquids and gases. For such extended phases, the inclusion of anharmonicity in reliable electronic structure methods is computationally extremely demanding, and hence, anharmonic effects are often lacking in thermochemical calculations. In this study, we apply the quantum cluster equilibrium method to transfer density functional theory calculations at the cluster level to the macroscopic, liquid, and gaseous phase of hydrogen fluoride. This allows us to include anharmonicity, either via vibrational self-consistent field calculations for smaller clusters or using a regression model for larger clusters. We obtain the structural composition of the fluid phases in terms of the population of different clusters as well as isobaric heat capacities as an example for thermodynamic properties. We study the role of anharmonicities for these analyses and observe that, in particular, the dominating structural motifs are rather sensitive to the anharmonicity in vibrational frequencies. The regression model proves to be a promising way to get access to anharmonic features, and the extension to more sophisticated machine-learning models is promising.

Exploring the formation of gold/silver nanoalloys with gas-phase synthesis and machine-learning assisted simulations

While nanoalloys are of paramount scientific and practical interest, the main processes leading to their formation are still poorly understood. Key structural features in the alloy systems, including the crystal phase, chemical ordering, and morphology, are challenging to control at the nanoscale, making it difficult to extend their use to industrial applications. In this contribution, we focus on the gold/silver system that has two of the most prevalent noble metals and combine experiments with simulations to uncover the formation mechanisms at the atomic level. Nanoparticles were produced using a state-of-the-art inert-gas aggregation source and analyzed using transmission electron microscopy and energy-dispersive X-ray spectroscopy. Machine-learning-assisted molecular dynamics simulations were employed to model the crystallization process from liquid droplets to nanocrystals. Our study finds a preponderance of nanoparticles with five-fold symmetric morphology, including icosahedra and decahedra which is consistent with previous results on mono-metallic nanoparticles. However, we observed that gold atoms, rather than silver atoms, segregate at the surface of the obtained nanoparticles for all the considered alloy compositions. These segregation tendencies are in contrast to previous studies and have consequences on the crystallization dynamics and the subsequent crystal ordering. We finally showed that the underpinning of this surprising segregation dynamics is due to charge transfer and electrostatic interactions rather than surface energy considerations.

A machine learning potential for simulating infrared spectra of nanosilicate clusters

The use of machine learning (ML) in chemical physics has enabled the construction of interatomic potentials having the accuracy of ab initio methods and a computational cost comparable to that of classical force fields. Training an ML model requires an efficient method for the generation of training data. Here, we apply an accurate and efficient protocol to collect training data for constructing a neural network-based ML interatomic potential for nanosilicate clusters. Initial training data are taken from normal modes and farthest point sampling. Later on, the set of training data is extended via an active learning strategy in which new data are identified by the disagreement between an ensemble of ML models. The whole process is further accelerated by parallel sampling over structures. We use the ML model to run molecular dynamics simulations of nanosilicate clusters with various sizes, from which infrared spectra with anharmonicity included can be extracted. Such spectroscopic data are needed for understanding the properties of silicate dust grains in the interstellar medium and in circumstellar environments.

Machine learning, artificial intelligence, and chemistry: how smart algorithms are reshaping simulation and the laboratory

Machine learning and artificial intelligence are increasingly gaining in prominence through image analysis, language processing, and automation, to name a few applications. Machine learning is also making profound changes in chemistry. From revisiting decades-old analytical techniques for the purpose of creating better calibration curves, to assisting and accelerating traditional in silico simulations, to automating entire scientific workflows, to being used as an approach to deduce underlying physics of unexplained chemical phenomena, machine learning and artificial intelligence are reshaping chemistry, accelerating scientific discovery, and yielding new insights. This review provides an overview of machine learning and artificial intelligence from a chemist’s perspective and focuses on a number of examples of the use of these approaches in computational chemistry and in the laboratory.

Infrared Spectroscopy of Chemically Diverse Carbon Clusters: A Data-Driven Approach

Carbon clusters exhibit a broad diversity of topologies and shapes, encompassing fullerene-like cages, graphene-like flakes, and more disordered pretzel-like and branched structures. Here, we examine computationally their infrared spectra in relation with these structures from a statistical perspective. Individual spectra for broad samples of isomers were determined by means of the self-consistent charge density functional-based tight-binding method, and an interpolation scheme is designed to reproduce the spectral features by regression on a much smaller subset of the sample. This interpolation proceeds by encoding the structures using appropriate descriptors and selecting them through principal component analysis, Gaussian regression or inverse distance weighting providing the nonlinear weighting functions. Metric learning is employed to reduce the global error on a preselected testing set. The interpolated spectra satisfactorily reproduce the specific spectral features and their dependence on the size and shape, enabling quantitative prediction away from the testing set. Finally, the classification of structures within the four proposed families is critically discussed through a statistical analysis of the sample based on iterative label spreading.

Predicting Infrared Spectra with Message Passing Neural Networks

Infrared (IR) spectroscopy remains an important tool for chemical characterization and identification. Chemprop-IR has been developed as a software package for the prediction of IR spectra through the use of machine learning. This work serves the dual purpose of providing a trained general-purpose model for the prediction of IR spectra with ease and providing the Chemprop-IR software framework for the training of new models. In Chemprop-IR, molecules are encoded using a directed message passing neural network, allowing for molecule latent representations to be learned and optimized for the task of spectral predictions. Model training incorporates spectra metrics and normalization techniques that offer better performance with spectral predictions than standard practice in regression models. The model makes use of pretraining using quantum chemistry calculations and ensembling of multiple submodels to improve generalizability and performance. The spectral predictions that result are of high quality, showing capability to capture the extreme diversity of spectral forms over chemical space and represent complex peak structures.

Non long-range corrected density functionals incorrectly describe the intensity of the CH stretching band in polycyclic aromatic hydrocarbons

We present a comprehensive study of the most relevant numerical aspects influencing frequencies and intensities in the infrared spectrum of isolated polycyclic aromatic hydrocarbons (PAHs) regarding the overestimate of the IR CH-stretching bands. We use naphthalene as benchmark and show the validity of our results to different members of the PAH family. Our analysis relies on widely employed density functional theory methods and second-order vibrational perturbational theory for the computation of vibrational eigenstates. We have focused on the elucidation of the origin of the systematic overestimate of the intensities in the CH-stretching region. To rule out nonfundamental numerical errors, we have initially considered the influence of the electronic basis set and various other parameters on the different stages of the vibrational analysis. In a second stage, we have benchmarked the results of different density functional theory functionals with respect to the aforementioned overestimate taken as the ratio between the most prominent features of the spectrum, the CH-bending and the CH-stretching bands. Our results unambiguously indicate that the long-range correction plays a major role in this spurious numerical issue. More specifically, this phenomenon is due to an incorrect description of the charge distribution (and hence dipole) within the symmetrically relevant CH bonds. Long-range correction specifically remedies this issue. It improves the description of the intensities in the stretching region while at the same time it does not perturb significantly the rest of the spectrum. With respect to the frequencies, we have observed an overall improvement when compared to noncorrected functionals.

Comprehensive Assessment of GFN Tight-Binding and Composite Density Functional Theory Methods for Calculating Gas-Phase Infrared Spectra

Vibrational spectroscopy is a valuable and widely used analytical tool for the characterization of chemical substances. We investigate the performance of semiempirical quantum mechanical GFN tight-binding and force-field methods for the calculation of gas-phase infrared spectra in comparison to experiment and low-cost (B3LYP-3c) density functional theory. A data set of 7247 experimental references was used to evaluate method performance based on automatic spectra comparison. Various quantitative spectral similarity measures were employed for the comparison between theory and experiment and for determining empirical scaling factors. It is shown that the scaling of atomic masses provides an accurate yet simple alternative to standard global frequency scaling in density functional theory (DFT) and semiempirical calculations. Furthermore, the method performance for 58 exemplary transition metal complexes was investigated. The efficient DFT composite method B3LYP-3c, that was introduced in the course of this work, was found to be excellently suited for general IR spectra calculations. The GFN1- and GFN2-xTB tight-binding methods clearly outperformed the PMx competitors. Conformational changes were investigated for a subset of the data and are found to have a mediocre strong influence on the simulated spectra suggesting that the corresponding elaborate sampling steps may be neglected in automated compound identification workflows.