Knowledge Engineering in Chemistry: From Expert Systems to Agents of Creation.

Passing knowledge from human to human is a natural process that has continued since the beginning of humankind. Over the past few decades, we have witnessed that knowledge is no longer passed only between humans but also from humans to machines. The latter form of knowledge transfer represents a cornerstone in artificial intelligence (AI) and lays the foundation for knowledge engineering (KE). In order to pass knowledge to machines, humans need to structure, formalize, and make knowledge machine-readable. Subsequently, humans also need to develop software that emulates their decision-making process. In order to engineer chemical knowledge, chemists are often required to challenge their understanding of chemistry and thinking processes, which may help improve the structure of chemical knowledge.Knowledge engineering in chemistry dates from the development of expert systems that emulated the thinking process of analytical and organic chemists. Since then, many different expert systems employing rather limited knowledge bases have been developed, solving problems in retrosynthesis, analytical chemistry, chemical risk assessment, etc. However, toward the end of the 20th century, the AI winters slowed down the development of expert systems for chemistry. At the same time, the increasing complexity of chemical research, alongside the limitations of the available computing tools, made it difficult for many chemistry expert systems to keep pace.In the past two decades, the semantic web, the popularization of object-oriented programming, and the increase in computational power have revitalized knowledge engineering. Knowledge formalization through ontologies has become commonplace, triggering the subsequent development of knowledge graphs and cognitive software agents. These tools enable the possibility of interoperability, enabling the representation of more complex systems, inference capabilities, and the synthesis of new knowledge.This Account introduces the history, the core principles of KE, and its applications within the broad realm of chemical research and engineering. In this regard, we first discuss how chemical knowledge is formalized and how a chemist's cognition can be emulated with the help of reasoning algorithms. Following this, we discuss various applications of knowledge graph and agent technology used to solve problems in chemistry related to molecular engineering, chemical mechanisms, multiscale modeling, automation of calculations and experiments, and chemist-machine interactions. These developments are discussed in the context of a universal and dynamic knowledge ecosystem, referred to as The World Avatar (TWA).

Vapor/Vapor-Solid Interfacial Growth of Covalent Organic Framework Membranes on Alumina Hollow Fiber for Advanced Molecular Separation.

Covalent organic frameworks (COFs), known for their chemical stability and porous crystalline structure, hold promises as advanced separation membranes. However, fabricating high-quality COF membranes, particularly on industrial-preferred hollow fiber substrates, remains challenging. This study introduces a novel vapor/vapor-solid (V/V-S) method for growing ultrathin crystalline TpPa-1 COF membranes on the inner lumen surface of alumina hollow fibers (TpPa-1/Alumina). Through vapor-phase monomer introduction onto polydopamine-modified alumina at 170 °C and 1 atm, efficient polymerization and crystallization occur at the confined V-S interface. This enables one-step growth within 8 h, producing 100 nm thick COF membranes with strong substrate adhesion. TpPa-1/Alumina exhibits exceptional stability and performance over 80 h in continuous cross-flow organic solvent nanofiltration (OSN), with methanol permeance of about 200 L m-2 h-1 bar-1 and dye rejection with molecular weight cutoff (MWCO) of approximately 700 Da. Moreover, the versatile V/V-S method synthesizes two additional COF membranes (TpPa2Cl/Alumina and TpHz/Alumina) with different pore sizes and chemical environments. Adjusting the COF membrane thickness between 100-500 nm is achievable easily by varying the growth cycle numbers. Notably, TpPa2Cl/Alumina demonstrates excellent OSN performance in separating the model active pharmaceutical ingredient glycyrrhizic acid (GA) from dimethyl sulfoxide (DMSO), highlighting the method's potential for large-scale industrial applications.

Selective and Stable CO2 Electroreduction to CH4 via Electronic Metal-Support Interaction upon Decomposition/Redeposition of MOF.

The CO2 electroreduction to fuels is a feasible approach to provide renewable energy sources. Therefore, it is necessary to conduct experimental and theoretical investigations on various catalyst design strategies, such as electronic metal-support interaction, to improve the catalytic selectivity. Here a solvent-free synthesis method is reported to prepare a copper (Cu)-based metal-organic framework (MOF) as the precursor. Upon electrochemical CO2 reduction in aqueous electrolyte, it undergoes in situ decomposition/redeposition processes to form abundant interfaces between Cu nanoparticles and amorphous carbon supports. This Cu/C catalyst favors the selective and stable production of CH4 with a Faradaic efficiency of ≈55% at -1.4 V versus reversible hydrogen electrode (RHE) for 12.5 h. The density functional theory calculation reveals the crucial role of interfacial sites between Cu and amorphous carbon support in stabilizing the key intermediates for CO2 reduction to CH4 . The adsorption of COOH* and CHO* at the Cu/C interface is up to 0.86 eV stronger than that on Cu(111), thus promoting the formation of CH4 . Therefore, it is envisioned that the strategy of regulating electronic metal-support interaction can improve the selectivity and stability of catalyst toward a specific product upon electrochemical CO2 reduction.

Chemical Species Ontology for Data Integration and Knowledge Discovery.

Web ontologies are important tools in modern scientific research because they provide a standardized way to represent and manage web-scale amounts of complex data. In chemistry, a semantic database for chemical species is indispensable for its ability to interrelate and infer relationships, enabling a more precise analysis and prediction of chemical behavior. This paper presents OntoSpecies, a web ontology designed to represent chemical species and their properties. The ontology serves as a core component of The World Avatar knowledge graph chemistry domain and includes a wide range of identifiers, chemical and physical properties, chemical classifications and applications, and spectral information associated with each species. The ontology includes provenance and attribution metadata, ensuring the reliability and traceability of data. Most of the information about chemical species are sourced from PubChem and ChEBI data on the respective compound Web pages using a software agent, making OntoSpecies a comprehensive semantic database of chemical species able to solve novel types of problems in the field. Access to this reliable source of chemical data is provided through a SPARQL end point. The paper presents example use cases to demonstrate the contribution of OntoSpecies in solving complex tasks that require integrated semantically searchable chemical data. The approach presented in this paper represents a significant advancement in the field of chemical data management, offering a powerful tool for representing, navigating, and analyzing chemical information to support scientific research.

Automated Rational Design of Metal-Organic Polyhedra.

Metal-organic polyhedra (MOPs) are hybrid organic-inorganic nanomolecules, whose rational design depends on harmonious consideration of chemical complementarity and spatial compatibility between two or more types of chemical building units (CBUs). In this work, we apply knowledge engineering technology to automate the derivation of MOP formulations based on existing knowledge. For this purpose we have (i) curated relevant MOP and CBU data; (ii) developed an assembly model concept that embeds rules in the MOP construction; (iii) developed an OntoMOPs ontology that defines MOPs and their key properties; (iv) input agents that populate The World Avatar (TWA) knowledge graph; and (v) input agents that, using information from TWA, derive a list of new constructible MOPs. Our result provides rapid and automated instantiation of MOPs in TWA and unveils the immediate chemical space of known MOPs, thus shedding light on new MOP targets for future investigations.

Wetting-regulated gas-involving (photo)electrocatalysis: biomimetics in energy conversion.

(Photo)electrolysis of water or gases with water to species serving as industrial feedstocks and energy carriers, such as hydrogen, ammonia, ethylene, propanol, etc., has drawn tremendous attention. Moreover, these processes can often be driven by renewable energy under ambient conditions as a sustainable alternative to traditional high-temperature and high-pressure synthesis methods. In addition to the extensive studies on catalyst development, increasing attention has been paid to the regulation of gas transport/diffusion behaviors during gas-involving (photo)electrocatalytic reactions towards the goal of creating industrially viable catalytic systems with high reaction rates, excellent long-term stabilities and near-unity selectivities. Biomimetic surfaces and systems with special wetting capabilities and structural advantages can shed light on the future design of (photo)electrodes and address long-standing challenges. This article is dedicated to bridging the fields of wetting and catalysis by reviewing the cutting-edge design methodologies of both gas-evolving and gas-consuming (photo)electrocatalytic systems. We first introduce the fundamentals of various in-air/underwater wetting states and their corresponding bioinspired structural properties. The relationship amongst the bubble transport behavior, wettability, and porosity/tortuosity is also discussed. Next, the latest implementations of wetting-related design principles for gas-evolving reactions (i.e. the hydrogen evolution reaction and oxygen evolution reaction) and gas-consuming reactions (i.e. the oxygen reduction reaction and CO2 reduction reaction) are summarized. For photoelectrode designs, additional factors are taken into account, such as light absorption and the separation, transport and recombination of photoinduced electrons and holes. The influences of wettability and 3D structuring of (photo)electrodes on the catalytic activity, stability and selectivity are analyzed to reveal the underlying mechanisms. Finally, remaining questions and related future perspectives are outlined.

π-Diradical Aromatic Soot Precursors in Flames.

Soot emitted from incomplete combustion of hydrocarbon fuels contributes to global warming and causes human disease. The mechanism by which soot nanoparticles form within hydrocarbon flames is still an unsolved problem in combustion science. Mechanisms proposed to date involving purely chemical growth are limited by slow reaction rates, whereas mechanisms relying on solely physical interactions between molecules are limited by weak intermolecular interactions that are unstable at flame temperatures. Here, we show evidence for a reactive π-diradical aromatic soot precursor imaged using non-contact atomic force microscopy. Localization of π-electrons on non-hexagonal rings was found to allow for Kekulé aromatic soot precursors to possess a triplet diradical ground state. Barrierless chain reactions are shown between these reactive sites, which provide thermally stable aromatic rim-linked hydrocarbons under flame conditions. Quantum molecular dynamics simulations demonstrate physical condensation of aromatics that survive for tens of picoseconds. Bound internal rotors then enable the reactive sites to find each other and become chemically cross-linked before dissociation. These species provide a rapid, thermally stable chain reaction toward soot nanoparticle formation and could provide molecular targets for limiting the emission of these toxic combustion products.

Question Answering System for Chemistry.

This paper describes the implementation and evaluation of a proof-of-concept Question Answering (QA) system for accessing chemical data from knowledge graphs (KGs) which offer data from chemical kinetics to the chemical and physical properties of species. We trained the question classification and named the entity recognition models that specialize in interpreting chemistry questions. The system has a novel design which applies a topic model to identify the question-to-ontology affiliation to handle ontologies with different structures. The topic model also helps the system to provide answers with a higher quality. Moreover, a new method that automatically generates training questions from ontologies is also implemented. The question set generated for training contains 432,989 questions under 11 types. Such a training set has been proven to be effective for training both the question classification model and the named entity recognition model. We evaluated the system using other KGQA systems as baselines. The system outperforms the chosen KGQA system answering chemistry-related questions. The QA system is also compared to the Google search engine and the WolframAlpha engine. It shows that the QA system can answer certain types of questions better than the search engines.

Automated Calibration of a Poly(oxymethylene) Dimethyl Ether Oxidation Mechanism Using the Knowledge Graph Technology.

In this paper, we develop a knowledge graph-based framework for the automated calibration of combustion reaction mechanisms and demonstrate its effectiveness on a case study of poly(oxymethylene)dimethyl ether (PODEn, where n = 3) oxidation. We develop an ontological representation for combustion experiments, OntoChemExp, that allows for the semantic enrichment of experiments within the J-Park simulator (JPS, theworldavatar.com), an existing cross-domain knowledge graph. OntoChemExp is fully capable of supporting experimental results in the Process Informatics Model (PrIMe) database. Following this, a set of software agents are developed to perform experimental result retrieval, sensitivity analysis, and calibration tasks. The sensitivity analysis agent is used for both generic sensitivity analyses and reaction selection for subsequent calibration. The calibration process is performed as a sampling task, followed by an optimization task. The agents are designed for use with generic models but are demonstrated with ignition delay time and laminar flame speed simulations. We find that calibration times are reduced, while accuracy is increased compared to manual calibration, achieving a 79% decrease in the objective function value, as defined in this study. Further, we demonstrate how this workflow is implemented as an extension of the JPS.

Multiscale Cross-Domain Thermochemical Knowledge-Graph.

In this paper, we develop a set of software agents which improve a knowledge-graph containing thermodynamic data of chemical species by means of quantum chemical calculations and error-canceling balanced reactions. The knowledge-graph represents species-associated information by making use of the principles of linked data, as employed in the Semantic Web, where concepts correspond to vertices and relationships between the concepts correspond to edges of the graph. We implement this representation by means of ontologies, which formalize the definition of concepts and their relationships, as a critical step to achieve interoperability between heterogeneous data formats and software. The agents, which conduct quantum chemical calculations and derive the estimates of standard enthalpies of formation, update the knowledge-graph with newly obtained results, improving data values, and adding nodes and connections between them. A key distinguishing feature of our approach is that it extends an existing, general-purpose knowledge-graph, called J-Park Simulator (http://theworldavatar.com), and its ecosystem of autonomous agents, thus enabling seamless cross-domain applications in wider contexts. To this end, we demonstrate how quantum calculations can directly affect the atmospheric dispersion of pollutants in an industrial emission use-case.

OntoKin: An Ontology for Chemical Kinetic Reaction Mechanisms.

An ontology for capturing both data and the semantics of chemical kinetic reaction mechanisms has been developed. Such mechanisms can be applied to simulate and understand the behavior of chemical processes, for example, the emission of pollutants from internal combustion engines. An ontology development methodology was used to produce the semantic model of the mechanisms, and a tool was developed to automate the assertion process. As part of the development methodology, the ontology is formally represented using a web ontology language (OWL), assessed by domain experts, and validated by applying a reasoning tool. The resulting ontology, termed OntoKin, has been used to represent example mechanisms from the literature. OntoKin and its instantiations are integrated to create a knowledge base (KB), which is deployed using the RDF4J triple store. The use of the OntoKin ontology and the KB is demonstrated for three use cases-querying across mechanisms, modeling atmospheric pollution dispersion, and as a mechanism browser tool. As part of the query use case, the OntoKin tools have been applied by a chemist to identify variations in the rate of a prompt NOx formation reaction in the combustion of ammonia as represented by four mechanisms in the literature.

Reactivity of Polycyclic Aromatic Hydrocarbon Soot Precursors: Kinetics and Equilibria.

The thermodynamics and kinetics of cross-linking reactions between PAHs of various reactive edge types that are observed in soot precursors are explored using density functional theory. The forward rate constants confirm that reactions involving aryl σ-radicals are faster than others, but rate constants for reactions between aryl σ-radicals and localized π-radicals can be as large or even larger than for two aryl σ-radicals. However, rates for all cross-linking reactions between small PAHs are likely too slow to explain soot formation. The equilibrium constants show that reactions involving σ and π-radical PAHs are the most favorable at flame temperatures. Equilibrium constants for larger PAHs show that the ability to form bonded-and-stacked structures results in enhanced equilibrium constants for the reaction of two large localized π-radicals compared to those for other edge types. This suggests that combined physical and chemical interactions between larger π-radical PAHs could be important in flame environments.

Rational Synthesis of Amorphous Iron-Nickel Phosphonates for Highly Efficient Photocatalytic Water Oxidation with Almost 100 % Yield.

A simple solvent ligation effect was successfully used to disrupt the growth of a model compound, Fe[(OH)(O3 P(CH2 )2 CO2 H)]⋅H2 O (MIL-37), into an extended 2D structure by replacing water with dimethylformamide (DMF) as the solvent during the synthesis. Owing to the lack of -OH group, which provides the corner-sharing (binding) oxygen atoms for the octahedra, an amorphous and porous structure is formed. When Fe3+ is partially replaced by Ni2+ , the amorphous structure remains and the resultant binary metal catalyst displays excellent photocatalytic oxygen evolution activity with almost 100 % yield achieved under visible light irradiation using [Ru(bpy)3 ]2+ as the photosensitizer. This study opens up new possibilities of using the simple solvent effect to synthesize high surface area metal phosphonates for catalytic and other applications.

Topology of Disordered 3D Graphene Networks.

Disordered carbons comprise graphene fragments assembled into three-dimensional networks. It has long been debated whether these networks contain positive curvature, as seen in fullerenes, negative curvature, as proposed for the schwarzite structures, or zero curvature, as in ribbons. We present a mesh-based approach to analyze the topology of a set of nanoporous and glassy carbon models that accurately reproduce experimental properties. Although all three topological elements are present, negatively curved structures dominate. At the atomic level, analysis of local environments shows that sp- and sp^{3}-bonded atoms are associated with line defects and screw dislocations that resolve topological complexities such as termination of free edges and stacking of low curvature regions into ribbons. These results provide insight into the synthesis of porous carbon materials, glassy carbon and the graphitizability of carbon materials.

An Ontology and Semantic Web Service for Quantum Chemistry Calculations.

The purpose of this article is to present an ontology, termed OntoCompChem, for quantum chemistry calculations as performed by the Gaussian quantum chemistry software, as well as a semantic web service named MolHub. The OntoCompChem ontology has been developed based on the semantics of concepts specified in the CompChem convention of Chemical Markup Language (CML) and by extending the Gainesville Core (GNVC) ontology. MolHub is developed in order to establish semantic interoperability between different tools used in quantum chemistry and thermochemistry calculations, and as such is integrated into the J-Park Simulator (JPS)-a multidomain interactive simulation platform and expert system. It uses the OntoCompChem ontology and implements a formal language based on propositional logic as a part of its query engine, which verifies satisfiability through reasoning. This paper also presents a NASA polynomial use-case scenario to demonstrate semantic interoperability between Gaussian and a tool for thermodynamic data calculations within MolHub.

Nanostructure of Gasification Charcoal (Biochar).

In this work, we investigate the molecular composition and nanostructure of gasification charcoal (biochar) by comparing it with heat-treated fullerene arc-soot. Using ultrahigh resolution Fourier transform ion-cyclotron resonance and laser desorption ionization time-of-flight mass spectrometry, Raman spectroscopy, and high resolution transmission electron microscopy we analyzed charcoal of low tar content obtained from gasification. Mass spectrometry revealed no magic number fullerenes such as C60 or C70 in the charcoal. The positive molecular ion m/ z 701, previously considered a graphitic part of the nanostructure, was found to be a breakdown product of pyrolysis and not part of the nanostructure. A higher mass distribution of ions similar to that found in thermally treated fullerene soot indicates that they share a nanostructure. Recent insights into the formation of all carbon fullerenes reveal that conditions in charcoal formation are not optimal for the formation of fullerenes, but instead, curved carbon structures coalesce into fulleroid-like structures. Microscopy and spectroscopy support such a stacked, fulleroid-like nanostructure, which was explored using reactive molecular dynamics simulations.

Optical band gap of cross-linked, curved, and radical polyaromatic hydrocarbons.

In this work, the optical band gaps of polycyclic aromatic hydrocarbons (PAHs) crosslinked via an aliphatic bond, curved via pentagon integration and with radical character were computed using density functional theory. A variety of different functionals were benchmarked against optical band gaps (OBGs) measured by ultraviolet-visible spectroscopy with HSE06 being most accurate with a percentage error of 6% for a moderate basis set. Pericondensed aromatics with different symmetries were calculated with this improved functional providing new scaling relationships for the OBG versus size. Further calculations showed crosslinks cause a small decrease in the OBG of the monomers which saturates after 3-4 crosslinks. Curvature in PAHs was shown to increase the optical band gap due to the resulting change in hybridisation of the system, but this increase saturated at larger sizes. The increase in OBG between a flat PAH and a strained curved one was shown to be equivalent to a difference of several rings in size for pericondensed aromatic systems. The effect of σ-radicals on the optical band gap was also shown to be negligible, however, π-radicals were found to decrease the band gap by ∼0.5 eV. These findings have applications in understanding the molecular species involved in soot formation.

Sphere Encapsulated Monte Carlo: Obtaining Minimum Energy Configurations of Large Aromatic Systems.

We introduce a simple global optimization approach that is able to find minimum energy configurations of clusters containing aromatic molecules. The translational and rotational perturbations required in Monte Carlo-based methods often lead to unrealistic configurations within which two or more molecular rings intersect, causing many of the computational steps to be rejected and the optimization process to be inefficient. Here we develop a modification of the basin-hopping global optimization procedure tailored to tackle problems with intersecting molecular rings. Termed the Sphere Encapsulated Monte Carlo (SEMC) method, this method introduces sphere-based rearrangement and minimization steps at each iteration, and its performance is shown through the exploration of potential energy landscapes of polycyclic aromatic hydrocarbon (PAH) clusters, systems of interest in combustion and astrophysics research. The SEMC method provides clusters that are accurate to 5% mean difference of the minimum energy at a 10-fold speed up compared to previous work using advanced molecular dynamics simulations. Importantly, the SEMC method captures key structural characteristics and molecular size partitioning trends as measured by the molecular radial distances and coordination numbers. The advantages of the SEMC method are further highlighted in its application to previously unstudied heterogeneous PAH clusters.

Improved methodology for performing the inverse Abel transform of flame images for color ratio pyrometry.

A new method is presented for performing the Abel inversion by fitting the line-of-sight projection of a predefined intensity distribution (FLiPPID) to the recorded 2D projections. The aim is to develop a methodology that is less prone to experimental noise when analyzing the projection of axisymmetric objects-in this case, co-flow diffusion flame images for color ratio pyrometry. A regression model is chosen for the light emission intensity distribution of the flame cross section as a function of radial distance from the flame center line. The forward Abel transform of this model function is fitted to the projected light intensity recorded by a color camera. For each of the three color channels, the model function requires three fitting parameters to match the radial intensity profile at each height above the burner. This results in a very smooth Abel inversion with no artifacts such as oscillations or negative values of the light source intensity, as is commonly observed for alternative Abel inversion techniques, such as the basis-set expansion or onion peeling. The advantages of the new FLiPPID method are illustrated by calculating the soot temperature and volume fraction profiles inside a co-flow diffusion flame, both being significantly smoother than those produced by the alternative inversion methods. The developed FLiPPID methodology can be applied to numerous other optical techniques for which smooth inverse Abel transforms are required.

Metal-free carbonaceous electrocatalysts and photocatalysts for water splitting.

Water splitting driven by sunlight or renewable resource-derived electricity has attracted great attention for sustainable production of hydrogen from water. Current research interest in this field is focused on the development of earth-abundant photo- or electrocatalytic materials with high activity and long-term stability for hydrogen and/or oxygen evolution reactions. Due to their unique properties and characteristics, carbon and related carbon-based materials show great potential to replace some of the existing precious metal catalysts in water splitting technology. This tutorial review summarizes the recent significant progress in the fabrication and application of metal-free carbonaceous materials as photo- or electrocatalysts for water splitting. Synthetic strategies and applications of various carbonaceous materials, including graphitic carbon nitride (g-C3N4), graphene, carbon nanotubes (CNTs) as well as other forms of carbon-containing materials, for electrochemical or photochemical water splitting are presented, accompanied by a discussion of the key scientific issues and prospects for the future development of metal-free photo- and electrocatalysts.