A Deep Learning Approach to Design and Discover Sustainable Cementitious Binders: Strategies to Learn From Small Databases and Develop Closed-form Analytical Models

To reduce the energy-intensity and carbon footprint of Portland cement (PC), the prevailing practice embraced by concrete technologists is to partially replace the PC in concrete with supplementary cementitious materials [SCMs: geological materials (e.g., limestone); industrial by-products (e.g., fly ash); and processed materials (e.g., calcined clay)]. Chemistry and content of the SCM profoundly affect PC hydration kinetics; which, in turn, dictates the evolutions of microstructure and properties of the [PC + SCM] binder. Owing to the substantial diversity in SCMs’ compositions–plus the massive combinatorial spaces, and the highly nonlinear and mutually-interacting processes that arise from SCM-PC interactions–state-of-the-art computational models are unable to produce a priori predictions of hydration kinetics or properties of [PC + SCM] binders. In the past 2 decades, the combination of Big data and machine learning (ML)—commonly referred to as the fourth paradigm of science–has emerged as a promising approach to learn composition-property correlations in materials (e.g., concrete), and capitalize on such learnings to produce a priori predictions of properties of materials with new compositions. Notwithstanding these merits, widespread use of ML models is hindered because they: 1) Require Big data to learn composition-property correlations, and, in general, large databases for concrete are not publicly available; and 2) Function as black-boxes, thus providing little-to-no insights into the materials laws like theory-based analytical models do. This study presents a deep learning (DL) model capable of producing a priori, high-fidelity predictions of composition- and time-dependent hydration kinetics and phase assemblage development in [PC + SCM] pastes. The DL is coupled with: 1) A fast Fourier transformation algorithm that reduces the dimensionality of training datasets (e.g., kinetic datasets), thus allowing the model to learn intrinsic composition-property correlations from a small database; and 2) A thermodynamic model that constrains the model, thus ensuring that predictions do not violate fundamental materials laws. The training and outcomes of the DL are ultimately leveraged to develop a simple, easy-to-use, closed-form analytical model capable of predicting hydration kinetics and phase assemblage development in [PC + SCM] pastes, using their initial composition and mixture design as inputs.

Modern materialistic understanding of the “energy” notion and the method of its systemic evaluation in chemicals

It was shown that the traditional approaches to understanding the notion of “energy” as a work or a physical quantity are outdated. For example, R. Feynman noted that “today’s physics does not know what energy is.” Therefore, even now, some researchers believe that “by and large, the concept of energy… is artificial, because unlike matter, of which we can say that it exists, energy is the fruit of human thought.” In contrast to these ideas, the authors showed that energy, like matter, objectively exists in various forms (energy continuum), which differ in structure, and is able to perform different types of work, to determine the forms of interaction and movement of matter in various material systems (substances, material bodies and megamaterial systems). A new scientific foundation for systematization and quantitative evaluation of energy characteristics of chemical compounds was proposed. It is based on a comprehensive assessment of contribution of chemical compounds’ composition and chemical bond type in line with a chemical bond’s unified model and the “System of chemical bonds and compounds” (SCBC). As a result of using this basic scientific innovation, the symbiosis of Mendeleev’s periodic table of atoms (composition – property) and SCBC (composition - chemical bond and structure – property) was realized for the first time. The foundation was laid for creating a database on systemic digitalization and evaluation of energy stored in various chemicals (natural gas, coal, oil, peat, wood, etc.) and the most effective ways of extracting it from these.

Chemical Composition and Low-Temperature Fluidity Properties of Jet Fuels

The physicochemical properties of petroleum-derived jet fuels mainly depend on their chemical composition, which can vary from sample to sample as a result of the diversity of the crude diet processed by the refinery. Jet fuels are exposed to very low temperatures both at altitude and on the ground in places subject to extreme climates and must be able to maintain their fluidity at these low temperatures otherwise the flow of fuel to turbine engines will be reduced or even stopped. In this work, an experimental evaluation of the effect of chemical composition on low-temperature fluidity properties of jet fuels (freezing point, crystallization onset temperature and viscosity at −20 °C) was carried out. Initially, a methodology based on gas chromatography coupled to mass spectrometry (GC–MS) was adapted to determine the composition of 70 samples of Jet A1 and Jet A fuels. This methodology allowed quantifying the content, in weight percentage, of five main families of hydrocarbons: paraffinic, naphthenic, aromatic, naphthalene derivatives, and tetralin- and indane-derived compounds. Fuel components were also grouped into 11 classes depending on structural characteristics and the number of carbon atoms in the compound. The latter compositional approach allowed obtaining more precise model regressions for predicting the composition–property dependence and identifying individual components or hydrocarbon classes contributing to increased or decreased property values.

On effective computations in subsemigroups of affine Cremona semigroup and implentations of new postquantum multivariate cryptosystems

Multivariate cryptography (MC) together with Latice Based, Hash based, Code based and Superelliptic curves based Cryptographies form list of the main directions of Post Quantum Cryptography.Investigations in the framework of tender of National Institute of Standardisation Technology (the USA) indicates that the potential of classical MC working with nonlinear maps of bounded degree and without the usage of compositions of nonlinear transformation is very restricted. Only special case of Rainbow like Unbalanced Oil and Vinegar digital signatures is remaining for further consideration. The remaining public keys for encryption procedure are not of multivariate. nature. The paper presents large semigroups and groups of transformations of finite affine space of dimension n with the multiple composition property. In these semigroups the composition of n transformations is computable in polynomial time. Constructions of such families are given together with effectively computed homomorphisms between members of the family. These algebraic platforms allow us to define protocols for several generators of subsemigroup of affine Cremona semigroups with several outputs. Security of these protocols rests on the complexity of the word decomposition problem, Finally presented algebraic protocols expanded to cryptosystems of El Gamal type which is not a public key system.

Atomically precise metal chalcogenide supertetrahedral clusters: frameworks to molecules, and structure to function

Metal chalcogenide supertetrahedral clusters (MCSCs) are of significance for developing crystalline porous framework materials and atomically precise cluster chemistry. Early research interest focused on the synthetic and structural chemistry of MCSC-based porous semiconductor materials with different cluster sizes/compositions and their applications in adsorption-based separation and optoelectronics. More recently, focus has shifted to the cluster chemistry of MCSCs to establish atomically precise structure–composition–property relationships, which are critical for regulating the properties and expanding the applications of MCSCs. Importantly, MCSCs are similar to Ⅱ-Ⅵ or Ⅰ-ⅡI-Ⅵ semiconductor nanocrystals (also called quantum dots, QDs) but avoid their inherent size polydispersity and structural ambiguity. Thus, discrete MCSCs, especially those that are solution processable, could provide models for understanding various issues that cannot be easily clarified using QDs. This review covers three decades of efforts on MCSCs, including advancements in MCSC-based open frameworks (reticular chemistry), the precise structure–property relationships of MCSCs (cluster chemistry), and the functionalization and applications of MCSC-based microcrystals. An outlook on remaining problems to be solved and future trends is also presented.

Scalable Gaussian processes for predicting the optical, physical, thermal, and mechanical properties of inorganic glasses with large datasets

Scalable Gaussian process for predicting composition–property of glasses with large datasets.