Accelerating the Design of Multishell Upconverting Nanoparticles through Bayesian Optimization.

The photon upconverting properties of lanthanide-doped nanoparticles drive their applications in imaging, optoelectronics, and additive manufacturing. To maximize their brightness, these upconverting nanoparticles (UCNPs) are often synthesized as core/shell heterostructures. However, the large numbers of compositional and structural parameters in multishell heterostructures make optimizing optical properties challenging. Here, we demonstrate the use of Bayesian optimization (BO) to learn the structure and design rules for multishell UCNPs with bright ultraviolet and violet emission. We leverage an automated workflow that iteratively recommends candidate UCNP structures and then simulates their emission spectra using kinetic Monte Carlo. Yb3+/Er3+- and Yb3+/Er3+/Tm3+-codoped UCNP nanostructures optimized with this BO workflow achieve 10- and 110-fold brighter emission within 22 and 40 iterations, respectively. This workflow can be expanded to structures with higher compositional and structural complexity, accelerating the discovery of novel UCNPs while domain-specific knowledge is being developed.

Chemical Reaction Networks Explain Gas Evolution Mechanisms in Mg-Ion Batteries.

Out-of-equilibrium electrochemical reaction mechanisms are notoriously difficult to characterize. However, such reactions are critical for a range of technological applications. For instance, in metal-ion batteries, spontaneous electrolyte degradation controls electrode passivation and battery cycle life. Here, to improve our ability to elucidate electrochemical reactivity, we for the first time combine computational chemical reaction network (CRN) analysis based on density functional theory (DFT) and differential electrochemical mass spectroscopy (DEMS) to study gas evolution from a model Mg-ion battery electrolyte─magnesium bistriflimide (Mg(TFSI)2) dissolved in diglyme (G2). Automated CRN analysis allows for the facile interpretation of DEMS data, revealing H2O, C2H4, and CH3OH as major products of G2 decomposition. These findings are further explained by identifying elementary mechanisms using DFT. While TFSI- is reactive at Mg electrodes, we find that it does not meaningfully contribute to gas evolution. The combined theoretical-experimental approach developed here provides a means to effectively predict electrolyte decomposition products and pathways when initially unknown.

Applying Large Graph Neural Networks to Predict Transition Metal Complex Energies Using the tmQM_wB97MV Data Set.

Machine learning (ML) methods have shown promise for discovering novel catalysts but are often restricted to specific chemical domains. Generalizable ML models require large and diverse training data sets, which exist for heterogeneous catalysis but not for homogeneous catalysis. The tmQM data set, which contains properties of 86,665 transition metal complexes calculated at the TPSSh/def2-SVP level of density functional theory (DFT), provided a promising training data set for homogeneous catalyst systems. However, we find that ML models trained on tmQM consistently underpredict the energies of a chemically distinct subset of the data. To address this, we present the tmQM_wB97MV data set, which filters out several structures in tmQM found to be missing hydrogens and recomputes the energies of all other structures at the ωB97M-V/def2-SVPD level of DFT. ML models trained on tmQM_wB97MV show no pattern of consistently incorrect predictions and much lower errors than those trained on tmQM. The ML models tested on tmQM_wB97MV were, from best to worst, GemNet-T > PaiNN ≈ SpinConv > SchNet. Performance consistently improves when using only neutral structures instead of the entire data set. However, while models saturate with only neutral structures, more data continue to improve the models when including charged species, indicating the importance of accurately capturing a range of oxidation states in future data generation and model development. Furthermore, a fine-tuning approach in which weights were initialized from models trained on OC20 led to drastic improvements in model performance, indicating transferability between ML strategies of heterogeneous and homogeneous systems.

Data-Driven Prediction of Formation Mechanisms of Lithium Ethylene Monocarbonate with an Automated Reaction Network.

Interfacial reactions are notoriously difficult to characterize, and robust prediction of the chemical evolution and associated functionality of the resulting surface film is one of the grand challenges of materials chemistry. The solid-electrolyte interphase (SEI), critical to Li-ion batteries (LIBs), exemplifies such a surface film, and despite decades of work, considerable controversy remains regarding the major components of the SEI as well as their formation mechanisms. Here we use a reaction network to investigate whether lithium ethylene monocarbonate (LEMC) or lithium ethylene dicarbonate (LEDC) is the major organic component of the LIB SEI. Our data-driven, automated methodology is based on a systematic generation of relevant species using a general fragmentation/recombination procedure which provides the basis for a vast thermodynamic reaction landscape, calculated with density functional theory. The shortest pathfinding algorithms are employed to explore the reaction landscape and obtain previously proposed formation mechanisms of LEMC as well as several new reaction pathways and intermediates. For example, we identify two novel LEMC formation mechanisms: one which involves LiH generation and another that involves breaking the (CH2)O-C(═O)OLi bond in LEDC. Most importantly, we find that all identified paths, which are also kinetically favorable under the explored conditions, require water as a reactant. This condition severely limits the amount of LEMC that can form, as compared with LEDC, a conclusion that has direct impact on the SEI formation in Li-ion energy storage systems. Finally, the data-driven framework presented here is generally applicable to any electrochemical system and expected to improve our understanding of surface passivation.

Local protein solvation drives direct down-conversion in phycobiliprotein PC645 via incoherent vibronic transport.

The mechanisms controlling excitation energy transport (EET) in light-harvesting complexes remain controversial. Following the observation of long-lived beats in 2D electronic spectroscopy of PC645, vibronic coherence, the delocalization of excited states between pigments supported by a resonant vibration, has been proposed to enable direct excitation transport from the highest-energy to the lowest-energy pigments, bypassing a collection of intermediate states. Here, we instead show that for phycobiliprotein PC645 an incoherent vibronic transport mechanism is at play. We quantify the solvation dynamics of individual pigments using ab initio quantum mechanics/molecular mechanics (QM/MM) nuclear dynamics. Our atomistic spectral densities reproduce experimental observations ranging from absorption and fluorescence spectra to the timescales and selectivity of down-conversion observed in transient absorption measurements. We construct a general model for vibronic dimers and establish the parameter regimes of coherent and incoherent vibronic transport. We demonstrate that direct down-conversion in PC645 proceeds incoherently, enhanced by large reorganization energies and a broad collection of high-frequency vibrations. We suggest that a similar incoherent mechanism is appropriate across phycobiliproteins and represents a potential design principle for nanoscale control of EET.

Aqueous Diels-Alder reactions for thermochemical storage and heat transfer fluids identified using density functional theory.

Thermal storage and transfer fluids have important applications in industrial, transportation, and domestic settings. Current thermal fluids have relatively low specific heats, often significantly below that of water. However, by introducing a thermochemical reaction to a base fluid, it is possible to enhance the fluid's thermal properties. In this work, density functional theory (DFT) is used to screen Diels-Alder reactions for use in aqueous thermal fluids. From an initial set of 52 reactions, four are identified with moderate aqueous solubility and predicted turning temperature near the liquid region of water. These reactions are selectively modified through 60 total functional group substitutions to produce novel reactions with improved solubility and thermal properties. Among the reactions generated by functional group substitution, seven have promising predicted thermal properties, significantly improving specific heat (by as much as 30.5%) and energy storage density (by as much as 4.9%) compared to pure water.

Chemical reaction networks and opportunities for machine learning.

Chemical reaction networks (CRNs), defined by sets of species and possible reactions between them, are widely used to interrogate chemical systems. To capture increasingly complex phenomena, CRNs can be leveraged alongside data-driven methods and machine learning (ML). In this Perspective, we assess the diverse strategies available for CRN construction and analysis in pursuit of a wide range of scientific goals, discuss ML techniques currently being applied to CRNs and outline future CRN-ML approaches, presenting scientific and technical challenges to overcome.

Improving machine learning performance on small chemical reaction data with unsupervised contrastive pretraining.

Machine learning (ML) methods have great potential to transform chemical discovery by accelerating the exploration of chemical space and drawing scientific insights from data. However, modern chemical reaction ML models, such as those based on graph neural networks (GNNs), must be trained on a large amount of labelled data in order to avoid overfitting the data and thus possessing low accuracy and transferability. In this work, we propose a strategy to leverage unlabelled data to learn accurate ML models for small labelled chemical reaction data. We focus on an old and prominent problem-classifying reactions into distinct families-and build a GNN model for this task. We first pretrain the model on unlabelled reaction data using unsupervised contrastive learning and then fine-tune it on a small number of labelled reactions. The contrastive pretraining learns by making the representations of two augmented versions of a reaction similar to each other but distinct from other reactions. We propose chemically consistent reaction augmentation methods that protect the reaction center and find they are the key for the model to extract relevant information from unlabelled data to aid the reaction classification task. The transfer learned model outperforms a supervised model trained from scratch by a large margin. Further, it consistently performs better than models based on traditional rule-driven reaction fingerprints, which have long been the default choice for small datasets, as well as those based on reaction fingerprints derived from masked language modelling. In addition to reaction classification, the effectiveness of the strategy is tested on regression datasets; the learned GNN-based reaction fingerprints can also be used to navigate the chemical reaction space, which we demonstrate by querying for similar reactions. The strategy can be readily applied to other predictive reaction problems to uncover the power of unlabelled data for learning better models with a limited supply of labels.

A chemically consistent graph architecture for massive reaction networks applied to solid-electrolyte interphase formation.

Modeling reactivity with chemical reaction networks could yield fundamental mechanistic understanding that would expedite the development of processes and technologies for energy storage, medicine, catalysis, and more. Thus far, reaction networks have been limited in size by chemically inconsistent graph representations of multi-reactant reactions (e.g. A + B → C) that cannot enforce stoichiometric constraints, precluding the use of optimized shortest-path algorithms. Here, we report a chemically consistent graph architecture that overcomes these limitations using a novel multi-reactant representation and iterative cost-solving procedure. Our approach enables the identification of all low-cost pathways to desired products in massive reaction networks containing reactions of any stoichiometry, allowing for the investigation of vastly more complex systems than previously possible. Leveraging our architecture, we construct the first ever electrochemical reaction network from first-principles thermodynamic calculations to describe the formation of the Li-ion solid electrolyte interphase (SEI), which is critical for passivation of the negative electrode. Using this network comprised of nearly 6000 species and 4.5 million reactions, we interrogate the formation of a key SEI component, lithium ethylene dicarbonate. We automatically identify previously proposed mechanisms as well as multiple novel pathways containing counter-intuitive reactions that have not, to our knowledge, been reported in the literature. We envision that our framework and data-driven methodology will facilitate efforts to engineer the composition-related properties of the SEI - or of any complex chemical process - through selective control of reactivity.

Quantum chemical calculations of lithium-ion battery electrolyte and interphase species.

Lithium-ion batteries (LIBs) represent the state of the art in high-density energy storage. To further advance LIB technology, a fundamental understanding of the underlying chemical processes is required. In particular, the decomposition of electrolyte species and associated formation of the solid electrolyte interphase (SEI) is critical for LIB performance. However, SEI formation is poorly understood, in part due to insufficient exploration of the vast reactive space. The Lithium-Ion Battery Electrolyte (LIBE) dataset reported here aims to provide accurate first-principles data to improve the understanding of SEI species and associated reactions. The dataset was generated by fragmenting a set of principal molecules, including solvents, salts, and SEI products, and then selectively recombining a subset of the fragments. All candidate molecules were analyzed at the ωB97X-V/def2-TZVPPD/SMD level of theory at various charges and spin multiplicities. In total, LIBE contains structural, thermodynamic, and vibrational information on over 17,000 unique species. In addition to studies of reactivity in LIBs, this dataset may prove useful for machine learning of molecular and reaction properties.

BonDNet: a graph neural network for the prediction of bond dissociation energies for charged molecules.

A broad collection of technologies, including e.g. drug metabolism, biofuel combustion, photochemical decontamination of water, and interfacial passivation in energy production/storage systems rely on chemical processes that involve bond-breaking molecular reactions. In this context, a fundamental thermodynamic property of interest is the bond dissociation energy (BDE) which measures the strength of a chemical bond. Fast and accurate prediction of BDEs for arbitrary molecules would lay the groundwork for data-driven projections of complex reaction cascades and hence a deeper understanding of these critical chemical processes and, ultimately, how to reverse design them. In this paper, we propose a chemically inspired graph neural network machine learning model, BonDNet, for the rapid and accurate prediction of BDEs. BonDNet maps the difference between the molecular representations of the reactants and products to the reaction BDE. Because of the use of this difference representation and the introduction of global features, including molecular charge, it is the first machine learning model capable of predicting both homolytic and heterolytic BDEs for molecules of any charge. To test the model, we have constructed a dataset of both homolytic and heterolytic BDEs for neutral and charged (-1 and +1) molecules. BonDNet achieves a mean absolute error (MAE) of 0.022 eV for unseen test data, significantly below chemical accuracy (0.043 eV). Besides the ability to handle complex bond dissociation reactions that no previous model could consider, BonDNet distinguishes itself even in only predicting homolytic BDEs for neutral molecules; it achieves an MAE of 0.020 eV on the PubChem BDE dataset, a 20% improvement over the previous best performing model. We gain additional insight into the model's predictions by analyzing the patterns in the features representing the molecules and the bond dissociation reactions, which are qualitatively consistent with chemical rules and intuition. BonDNet is just one application of our general approach to representing and learning chemical reactivity, and it could be easily extended to the prediction of other reaction properties in the future.

Noncentrosymmetry in new templated gallium fluorophosphates.

Two new noncentrosymmetric polar gallium fluorophosphates have been synthesized under mild hydrothermal conditions through the use of enantiomorphically pure sources of either R-2-methylpiperazine or S-2-methylpiperazine. A centrosymmetric analogue was also prepared using a racemic source of the amine. Novel [Ga(3)F(PO(4))(4)](n)(4n-) layers, constructed from [Ga(3)O(3)F(PO(4))(4)] building units, are observed in all three compounds. The use of racemic 2-methylpiperazine results in crystallographic disorder of the amines and creation of inversion centers, while using a single enantiomer destroys the inversion symmetry and orders the amines. Second harmonic generation measurements were performed on [(R)-C(5)H(14)N(2)](2)[Ga(3)F(PO(4))(4)] x 5.5 H(2)O and [(S)-C(5)H(14)N(2)](2)[Ga(3)F(PO(4))(4)] x 4.75 H(2)O, both of which display type 1 phase-matching capabilities and exhibit activities of approximately 50 x alpha-SiO(2). The structures of these compounds were determined using single crystal X-ray diffraction, infrared spectroscopy, and thermal analyses. [C(5)H(14)N(2)](2)[Ga(3)F(PO(4))(4)] x 5.25 H(2)O, a = 13.0863(5) A, c = 9.9023(4) A, trigonal, P-3 (No. 147), Z = 2; [(R)-C(5)H(14)N(2)](2)[Ga(3)F(PO(4))(4)] x 5.5 H(2)O, a = 13.0887(2) A, c = 29.9439(4) A, trigonal, P3(1) (No. 144), Z = 6; [(S)-C(5)H(14)N(2)](2)[Ga(3)F(PO(4))(4)] x 4.75 H(2)O, a = 13.0871(2) A, c = 29.8350(6) A, trigonal, P3(2) (No. 145), Z = 6.

Disentanglement of excited-state dynamics with implications for FRET measurements: two-dimensional electronic spectroscopy of a BODIPY-functionalized cavitand.

Förster Resonance Energy Transfer (FRET) is the incoherent transfer of an electronic excitation from a donor fluorophore to a nearby acceptor. FRET has been applied as a probe of local chromophore environments and distances on the nanoscale by extrapolating transfer efficiencies from standard experimental parameters, such as fluorescence intensities or lifetimes. Competition from nonradiative relaxation processes is often assumed to be constant in these extrapolations, but in actuality, this competition depends on the donor and acceptor environments and can, therefore, be affected by conformational changes. To study the effects of nonradiative relaxation on FRET dynamics, we perform two-dimensional electronic spectroscopy (2DES) on a pair of azaboraindacene (BODIPY) dyes, attached to opposite arms of a resorcin[4]arene cavitand. Temperature-induced switching between two equilibrium conformations, vase at 294 K to kite at 193 K, increases the donor-acceptor distance from 0.5 nm to 3 nm, affecting both FRET efficiency and nonradiative relaxation. By disentangling different dynamics based on lifetimes extracted from a series of 2D spectra, we independently observe nonradiative relaxation, FRET, and residual fluorescence from the donor in both vase to kite conformations. We observe changes in both FRET rate and nonradiative relaxation when the molecule switches from vase to kite, and measure a significantly greater difference in transfer efficiency between conformations than would be determined by standard lifetime-based measurements. These observations show that changes in competing nonradiative processes must be taken into account when highly accurate measurements of FRET efficiency are desired.