Formulation Graphs for Mapping Structure-Composition of Battery Electrolytes to Device Performance.

Advanced computational methods are being actively sought to address the challenges associated with the discovery and development of new combinatorial materials, such as formulations. A widely adopted approach involves domain-informed high-throughput screening of individual components that can be combined together to form a formulation. This manages to accelerate the discovery of new compounds for a target application but still leaves the process of identifying the right "formulation" from the shortlisted chemical space largely a laboratory experiment-driven process. We report a deep learning model, the Formulation Graph Convolution Network (F-GCN), that can map the structure-composition relationship of the formulation constituents to the property of liquid formulation as a whole. Multiple GCNs are assembled in parallel that featurize formulation constituents domain-intuitively on the fly. The resulting molecular descriptors are scaled based on the respective constituent's molar percentage in the formulation, followed by integration into a combined formulation descriptor that represents the complete formulation to an external learning architecture. The use case of the proposed formulation learning model is demonstrated for battery electrolytes by training and testing it on two exemplary data sets representing electrolyte formulations vs battery performance: one data set is sourced from the literature about Li/Cu half-cells, while the other is obtained by lab experiments related to lithium-iodide full-cell chemistry. The model is shown to predict performance metrics such as Coulombic efficiency (CE) and specific capacity of new electrolyte formulations with the lowest reported errors. The best-performing F-GCN model uses molecular descriptors derived from molecular graphs (GCNs) that are informed with HOMO-LUMO and electric moment properties of the molecules using a knowledge transfer technique.

Artificial intelligence driven design of catalysts and materials for ring opening polymerization using a domain-specific language.

Advances in machine learning (ML) and automated experimentation are poised to vastly accelerate research in polymer science. Data representation is a critical aspect for enabling ML integration in research workflows, yet many data models impose significant rigidity making it difficult to accommodate a broad array of experiment and data types found in polymer science. This inflexibility presents a significant barrier for researchers to leverage their historical data in ML development. Here we show that a domain specific language, termed Chemical Markdown Language (CMDL), provides flexible, extensible, and consistent representation of disparate experiment types and polymer structures. CMDL enables seamless use of historical experimental data to fine-tune regression transformer (RT) models for generative molecular design tasks. We demonstrate the utility of this approach through the generation and the experimental validation of catalysts and polymers in the context of ring-opening polymerization-although we provide examples of how CMDL can be more broadly applied to other polymer classes. Critically, we show how the CMDL tuned model preserves key functional groups within the polymer structure, allowing for experimental validation. These results reveal the versatility of CMDL and how it facilitates translation of historical data into meaningful predictive and generative models to produce experimentally actionable output.

Combinatorial Polyacrylamide Hydrogels for Preventing Biofouling on Implantable Biosensors.

Biofouling on the surface of implanted medical devices and biosensors severely hinders device functionality and drastically shortens device lifetime. Poly(ethylene glycol) and zwitterionic polymers are currently considered "gold-standard" device coatings to reduce biofouling. To discover novel anti-biofouling materials, a combinatorial library of polyacrylamide-based copolymer hydrogels is created, and their ability is screened to prevent fouling from serum and platelet-rich plasma in a high-throughput parallel assay. It is found that certain nonintuitive copolymer compositions exhibit superior anti-biofouling properties over current gold-standard materials, and machine learning is used to identify key molecular features underpinning their performance. For validation, the surfaces of electrochemical biosensors are coated with hydrogels and their anti-biofouling performance in vitro and in vivo in rodent models is evaluated. The copolymer hydrogels preserve device function and enable continuous measurements of a small-molecule drug in vivo better than gold-standard coatings. The novel methodology described enables the discovery of anti-biofouling materials that can extend the lifetime of real-time in vivo sensing devices.

Hebbian Learning on Small Data Enables Experimental Discovery of High Tg Polyimides.

We report a study combining computational design and experimental evaluation of polyimides with high glass transition temperatures: Tg between 220 °C and 500 °C. The computational approach is based on the recently introduced competitive learning algorithm, supervised self-organizing maps (SUSI), which we recast as an ensemble method, e-SUSI. We use e-SUSI to solve both unsupervised and supervised/semisupervised learning tasks capturing structure-property relationships of high-Tg polyimides historically studied at Almaden Research Center. Predictors trained on historical data were applied to the combinatorial library of novel polyimides and informed selection of the candidates for synthesis and characterization. In this manner, three new polyimides were prepared with Tg values 281 °C, 282 °C, and 331 °C. The measured values closely agree with the predicted values 273 °C, 311 °C, and 335 °C, respectively. We discuss specific reasons that make the proposed computational design strategy attractive in rapid, deliverable-driven efforts with limited, small-batch data sets.

Radiolysis generates a complex organosynthetic chemical network.

The architectural features of cellular life and its ecologies at larger scales are built upon foundational networks of reactions between molecules that avoid a collapse to equilibrium. The search for life's origins is, in some respects, a search for biotic network attributes in abiotic chemical systems. Radiation chemistry has long been employed to model prebiotic reaction networks, and here we report network-level analyses carried out on a compiled database of radiolysis reactions, acquired by the scientific community over decades of research. The resulting network shows robust connections between abundant geochemical reservoirs and the production of carboxylic acids, amino acids, and ribonucleotide precursors-the chemistry of which is predominantly dependent on radicals. Moreover, the network exhibits the following measurable attributes associated with biological systems: (1) the species connectivity histogram exhibits a heterogeneous (heavy-tailed) distribution, (2) overlapping families of closed-loop cycles, and (3) a hierarchical arrangement of chemical species with a bottom-heavy energy-size spectrum. The latter attribute is implicated with stability and entropy production in complex systems, notably in ecology where it is known as a trophic pyramid. Radiolysis is implicated as a driver of abiotic chemical organization and could provide insights about the complex and perhaps radical-dependent mechanisms associated with life's origins.

Diagnostics of Data-Driven Models: Uncertainty Quantification of PM7 Semi-Empirical Quantum Chemical Method.

We report an evaluation of a semi-empirical quantum chemical method PM7 from the perspective of uncertainty quantification. Specifically, we apply Bound-to-Bound Data Collaboration, an uncertainty quantification framework, to characterize (a) variability of PM7 model parameter values consistent with the uncertainty in the training data and (b) uncertainty propagation from the training data to the model predictions. Experimental heats of formation of a homologous series of linear alkanes are used as the property of interest. The training data are chemically accurate, i.e., they have very low uncertainty by the standards of computational chemistry. The analysis does not find evidence of PM7 consistency with the entire data set considered as no single set of parameter values is found that captures the experimental uncertainties of all training data. A set of parameter values for PM7 was able to capture the training data within ±1 kcal/mol, but not to the smaller level of uncertainty in the reported data. Nevertheless, PM7 was found to be consistent for subsets of the training data. In such cases, uncertainty propagation from the chemically accurate training data to the predicted values preserves error within bounds of chemical accuracy if predictions are made for the molecules of comparable size. Otherwise, the error grows linearly with the relative size of the molecules.

Three-months optical coherence tomography analysis of a biodegradable polymer, sirolimus-eluting stent.

OBJECTIVE: We aimed to assess early neointimal healing by optical coherence tomography (OCT) 3 months after implantation of the ultrathin Orsiro® sirolimus-eluting stent with biodegradable polymer. BACKGROUND: New generations of drug-eluting stents with biodegradable polymer have been developed to avoid the continued vascular irritation of durable polymers. METHODS: In this prospective, open-label study, 34 patients received an Orsiro® sirolimus-eluting stent with biodegradable polymer. In a subgroup of patients (n = 15), the intervention was performed under OCT guidance. All patients underwent OCT-examination at three months. The primary endpoint was 3-month neointimal healing (NIH) score, calculated by weighing the presence of filling defects, malapposed and uncovered struts. Secondary endpoint was maturity of tissue coverage at 3 months. RESULTS: At 3 months, NIH score was 13.7 (5.4-22), covered struts per lesion were 90% (84-97%), malapposed struts were 2.7% (0.8-5.4%) and rate of mature tissue coverage was 47% (42-53%). No target lesion failure occurred up to 12 months. Patients with OCT-guided stent implantation demonstrated a trend toward earlier stent healing as demonstrated by superior NIH scores (angio guided: 17.6% [8.8-26.4]; OCT-guided: 9.8% [4.0-15.5]; mean difference -8, [95%CI: -18.7-2.9], P = 0.123). This group had significantly more covered struts per lesion (angio-guided: 86% [82-90]; 95% [92-99]; mean difference 9% [95%CI: 3-15], P = 0.001). CONCLUSION: The Orsiro® sirolimus-eluting stent with biodegradable polymer shows early vascular healing with a high rate of strut coverage at 3-month follow-up. OCT guided stent implantation had a positive impact on early vascular healing.

Chiral Sugars Drive Enantioenrichment in Prebiotic Amino Acid Synthesis.

Chiral pentose sugars mediate the enantioselective synthesis of amino acid precursors, with the magnitude of the chiral induction dictated by a subtle cooperativity between sugar hydroxyl groups. Ribose and lyxose give opposite chiral preferences, and theoretical calculations reveal the pseudoenantiomeric nature of transition state structures from the two sugars. Prebiotically plausible mixtures of natural d-sugars lead to enantioenrichment of natural l-amino acid precursors. Temporal monitoring and kinetic modeling of the reaction reveal an unusual dynamic kinetic resolution that shifts toward an enantioselective pathway over time, providing an amplification mechanism for the transfer of chiral information. This work adds to growing evidence for synergy in the etiology of the single chirality of the two most important classes of biological molecules, the sugars that make up DNA and RNA and the amino acids that form proteins.

Sustainability of Transient Kinetic Regimes and Origins of Death.

It is generally recognized that a distinguishing feature of life is its peculiar capability to avoid equilibration. The origin of this capability and its evolution along the timeline of abiogenesis is not yet understood. We propose to study an analog of this phenomenon that could emerge in non-biological systems. To this end, we introduce the concept of sustainability of transient kinetic regimes. This concept is illustrated via investigation of cooperative effects in an extended system of compartmentalized chemical oscillators under batch and semi-batch conditions. The computational study of a model system shows robust enhancement of lifetimes of the decaying oscillations which translates into the evolution of the survival function of the transient non-equilibrium regime. This model does not rely on any form of replication. Rather, it explores the role of a structured effective environment as a contributor to the system-bath interactions that define non-equilibrium regimes. We implicate the noise produced by the effective environment of a compartmentalized oscillator as the cause of the lifetime extension.

Uncertainty of prebiotic scenarios: the case of the non-enzymatic reverse tricarboxylic acid cycle.

We consider the hypothesis of the primordial nature of the non-enzymatic reverse tricarboxylic acid (rTCA) cycle and describe a modeling approach to quantify the uncertainty of this hypothesis due to the combinatorial aspect of the constituent chemical transformations. Our results suggest that a) rTCA cycle belongs to a degenerate optimum of auto-catalytic cycles, and b) the set of targets for investigations of the origin of the common metabolic core should be significantly extended.

Interval prediction of molecular properties in parametrized quantum chemistry.

The accurate evaluation of molecular properties lies at the core of predictive physical models. Most reliable quantum-chemical calculations are limited to smaller molecular systems while purely empirical approaches are limited in accuracy and reliability. A promising approach is to employ a quantum-mechanical formalism with simplifications and to compensate for the latter with parametrization. We propose a strategy of directly predicting the uncertainty interval for a property of interest, based on training-data uncertainties, which sidesteps the need for an optimum set of parameters.

Pathways to soot oxidation: reaction of OH with phenanthrene radicals.

Energetics and kinetics of the oxidation of possible soot surface sites by hydroxyl radicals were investigated theoretically. Energetics were calculated by employing density functional theory. Three candidate reactions were selected as suitable prototypes of soot oxidation by OH. The first two, OH + benzene and OH + benzene-phenol complex, did not produce pathways that lead to substantial CO expulsion. The third reaction, OH attack on the phenanthrene radical, had multiple pathways leading to CO elimination. The kinetics of the latter reaction system were determined by solving the master equations with the MultiWell suite of codes. The barrierless reaction rates of this system were computed using the VariFlex program. The computations were carried out over the ranges 1500-2500 K and 0.01-10 atm. At higher temperatures, above 2000 K, the oxidation of phenanthrene radicals by OH followed a chemically activated path. At temperatures lower than 2000 K, chemical activation was not sufficient to drive the reaction to products; reaction progress was impeded by intermediate adducts rapidly de-energizing before reaching products. In such cases, the reaction system was modeled by treating the accumulating adducts as distinct chemical species and computing their kinetics via thermal decomposition. The overall rate coefficient of phenanthrene radical oxidation by OH forming CO was found to be insensitive to pressure and temperature and is approximately 1 × 10(14) cm(3) mol(-1) s(-1). The oxidation of phenanthrene radicals by OH is shown to be controlled by two main processes: H atom migration/elimination and oxyradical decomposition. H atom migration and elimination made possible relatively rapid rearrangement of the aromatic edge to form oxyradicals with favorable decomposition rates. The reaction then continues down the fastest oxyradical pathways, eliminating CO.

Complex Chemical Reaction Networks from Heuristics-Aided Quantum Chemistry.

While structures and reactivities of many small molecules can be computed efficiently and accurately using quantum chemical methods, heuristic approaches remain essential for modeling complex structures and large-scale chemical systems. Here, we present a heuristics-aided quantum chemical methodology applicable to complex chemical reaction networks such as those arising in cell metabolism and prebiotic chemistry. Chemical heuristics offer an expedient way of traversing high-dimensional reactive potential energy surfaces and are combined here with quantum chemical structure optimizations, which yield the structures and energies of the reaction intermediates and products. Application of heuristics-aided quantum chemical methodology to the formose reaction reproduces the experimentally observed reaction products, major reaction pathways, and autocatalytic cycles.

From aromaticity to self-organized criticality in graphene.

The unique properties of graphene are rooted in its peculiar electronic structure where effects of electron delocalization are pivotal. We show that the traditional view of delocalization as formation of a local or global aromatic bonding framework has to be expanded in this case. A modification of the π-electron system of a finite-size graphene substrate results in a scale-invariant response in the relaxation of interatomic distances and reveals self-organized criticality as a mode of delocalized bonding. Graphene is shown to belong to a diverse class of finite-size extended systems with simple local interactions where complexity emerges spontaneously under very general conditions that can be a critical factor controlling observable properties such as chemical activity, electron transport, and spin-polarization.

Quantum Monte Carlo for the x-ray absorption spectrum of pyrrole at the nitrogen K-edge.

Fixed-node diffusion Monte Carlo (FNDMC) is used to simulate the x-ray absorption spectrum of a gas-phase pyrrole molecule at the nitrogen K-edge. Trial wave functions for core-excited states are constructed from ground-state Kohn-Sham determinants substituted with singly occupied natural orbitals from configuration interaction with single excitations calculations of the five lowest valence-excited triplet states. The FNDMC ionization potential (IP) is found to lie within 0.3 eV of the experimental value of 406.1 ± 0.1 eV. The transition energies to anti-bonding virtual orbitals match the experimental spectrum after alignment of IP values and agree with the existing assignments.

Thermal decomposition of pentacene oxyradicals.

The energetics and kinetics of the thermal decomposition of pentacene oxyradicals were studied using a combination of ab initio electronic structure theory and energy-transfer master equation modeling. The rate coefficients of pentacene oxyradical decomposition were computed for the range of 1500-2500 K and 0.01-10 atm and found to be both temperature and pressure dependent. The computational results reveal that oxyradicals with oxygen attached to the inner rings are kinetically more stable than those with oxygen attached to the outer rings. The latter decompose to produce CO at rates comparable to those of phenoxy radical, while CO is unlikely to be produced from oxyradicals with oxygen bonded to the inner rings.

A diffusion Monte Carlo study of the O-H bond dissociation of phenol.

The homolytic O-H bond dissociation energy (BDE) of phenol was determined from diffusion Monte Carlo (DMC) calculations using single determinant trial wave functions. DMC gives an O-H BDE of 87.0 +/- 0.3 kcal/mol when restricted Hartree-Fock orbitals are used and a BDE of 87.5 +/- 0.3 kcal/mol with restricted B3LYP Kohn-Sham orbitals. These results are in good agreement with the extrapolated B3P86 results of Costa Cabral and Canuto (88.3 kcal/mol), the recommended experimental value of Borges dos Santos and Martinho Simões (88.7 +/- 0.5 kcal/mol), and the G3 (88.2 kcal/mol), CBS-APNO (88.2 kcal/mol), CBS-QB3 (87.1 kcal/mol) results of Mulder.