Machine Learning-Enabled Prediction and High-Throughput Screening of Polymer Membranes for Pervaporation Separation.

Pervaporation (PV) is considered as a robust membrane-based separation technology for liquid mixtures. However, the development of PV membranes is impeded largely by the lack of adequate models capable of reliably predicting the performance of PV membranes. In this study, we collect an experimental data set with a total of 681 data samples including 16 polymers and 6 organic solvents for a wide variety of water/organic mixtures under various operating conditions. Then, two types of machine learning (ML) models are developed for prediction and high-throughput screening of polymer membranes for PV separation. Based on the intrinsic properties of polymer and solvent (water contact angle of polymer and solubility parameter of solvent) as gross descriptors, the first type accurately predicts PV separation performance (total flux and separation factor). The second type is based on the molecular representation of polymer and solvent, giving accuracy comparable to the first type, and applied to screen ∼1 million hypothetical polymers for PV separation of water/ethanol mixtures. With a threshold of 700 for the PV separation index, 20 polymers are shortlisted, with many surpassing experimental samples. Among these, 10 are further identified to be synthesizable in terms of a synthetic complexity score. The ML models developed in this study would facilitate the optimization of operating conditions and accelerate the development of new polymer membranes for high-performance PV separation.

Free-standing homochiral 2D monolayers by exfoliation of molecular crystals.

Two-dimensional materials with monolayer thickness and extreme aspect ratios are sought for their high surface areas and unusual physicochemical properties1. Liquid exfoliation is a straightforward and scalable means of accessing such materials2, but has been restricted to sheets maintained by strong covalent, coordination or ionic interactions3-10. The exfoliation of molecular crystals, in which repeat units are held together by weak non-covalent bonding, could generate a greatly expanded range of two-dimensional crystalline materials with diverse surfaces and structural features. However, at first sight, these weak forces would seem incapable of supporting such intrinsically fragile morphologies. Against this expectation, we show here that crystals composed of discrete supramolecular coordination complexes can be exfoliated by sonication to give free-standing monolayers approximately 2.3 nanometres thick with aspect ratios up to approximately 2,500:1, sustained purely by apolar intermolecular interactions. These nanosheets are characterized by atomic force microscopy and high-resolution transmission electron microscopy, confirming their crystallinity. The monolayers possess complex chiral surfaces derived partly from individual supramolecular coordination complex components but also from interactions with neighbours. In this respect, they represent a distinct type of material in which molecular components are all equally exposed to their environment, as if in solution, yet with properties arising from cooperation between molecules, because of crystallinity. This unusual nature is reflected in the molecular recognition properties of the materials, which bind carbohydrates with strongly enhanced enantiodiscrimination relative to individual molecules or bulk three-dimensional crystals.

Rapid Screening of Metal-Organic Frameworks for Propane/Propylene Separation by Synergizing Molecular Simulation and Machine Learning.

At present, 100 000+ metal-organic frameworks (MOFs) have been synthesized, and it is challenging to identity the best candidate for a specific application. In this study, MOFs are rapidly screened via a hierarchical approach for propane/propylene (C3H8/C3H6) separation. First, the adsorption capacity and selectivity of C3H8/C3H6 mixture in "Computation-Ready, Experimental" (CoRE) MOFs are predicted via a molecular simulation (MS) method. The relationships between separation metrics and structural factors are established, and top-performing CoRE MOFs are identified. Then, machine learning (ML) models are trained and developed upon the CoRE MOFs using pore size, pore geometry, and framework chemistry as feature descriptors. By introducing binned pore size distributions and geometric descriptors, the accuracy of ML models is substantially improved. The feature importance of the descriptors is physically interpreted by the Gini impurities and Shapley Additive Explanations. Subsequently, the ML models are used to rapidly screen experimental "Cambridge Structural Database" (CSD) MOFs and hypothetical MOFs for C3H8/C3H6 separation. In the CSD MOFs, the out-of-sample predictions are found to agree well with simulation results, demonstrating the excellent transferability of the ML models from the CoRE to CSD MOFs. Moreover, nine CSD MOFs are identified to possess separation performance superior to top-performing CoRE MOFs. Finally, the similarity and diversity among experimental and hypothetical MOFs are visualized and compared by the t-Distributed Stochastic Neighbor Embedding (t-SNE) feature projections. Remarkably, the CoRE and CSD MOFs are revealed to share a close similarity in both chemical and geometric feature spaces. By synergizing MS and ML, the hierarchical approach developed in this study would advance the rapid screening of MOFs across different databases toward industrially important separation processes.

Chiral BINOL-Based Covalent Organic Frameworks for Enantioselective Sensing.

Covalent organic frameworks (COFs) have emerged as a novel platform for material design and functional explorations, but it remains a challenge to synthetically functionalize targeted structures for task-specific applications. Optically pure 1,1'-bi-2-naphthol (BINOL) is one of the most important sources of chirality for organic synthesis and materials science, but it has not yet been used in construction of COFs for enantioselective processes. Here, by elaborately designing and choosing an enantiopure BINOL-based linear dialdehyde and a tris(4-aminophenyl)benzene derivative or tetrakis(4-aminophenyl)ethene as building blocks, two imine-linked chiral fluorescent COFs with a 2D layered hexagonal or tetragonal structure are prepared. The COF containing flexible tetraphenylethylene units can be readily exfoliated into ultrathin 2D nanosheets and electrospun to make free-standing nanofiber membrane. In both the solution and membrane systems, the fluorescence of COF nanosheets can be effectively quenched by chiral odor vapors via supramolecular interactions with the immobilized BINOL moieties, leading to remarkable chiral vapor sensors. Compared to the BINOL-based homogeneous and membrane systems, the COF nanosheets exhibited greatly enhanced sensitivity and enantioselectivity owing to the confinement effect and the conformational rigidity of the sensing BINOL groups in the framework. The ability to place such a useful BINOL chiral auxiliary inside open channels of COFs capable of amplifying chiral discrimination of the analytes represents a major step toward the rational synthesis of porous molecular materials for more chirality applications.

Molecular Simulation and Analysis of Sorption Process toward Theoretical Prediction for Liquid Permeation through Membranes.

The need to understand and describe permeation through membranes has driven the development of many well-established transport models. The modeling parameters such as solubility, diffusivity, and permeability represent the intrinsic nature of molecular interactions between membrane and permeants. In this study, we report a simulation and analysis methodology for liquid permeation. On the basis of a single simulation of liquid sorption process into a membrane, the solubility and diffusivity are estimated simultaneously; then, the permeability is predicted by the solution-diffusion model. The methodology is applied to water permeation through two representative membranes: a polymer of intrinsic microporosity (PIM-1) and a zeolitic imidazolate framework (ZIF-96). For amorphous PIM-1 membrane, the predicted water permeability agrees perfectly with simulation. For crystalline ZIF-96 membrane, water permeability is fairly well predicted. Furthermore, water dynamics in the membranes is analyzed by simulation trajectories and water structure is characterized by hydrogen bonds. Together with these microscopic insights, this study provides a simple theoretical approach to quantitatively describe water sorption, diffusion, and permeation, and it can be further applied to other liquid permeation (e.g., organic solvent nanofiltration).

Sedimentological constraints on the initial uplift of the West Bogda Mountains in Mid-Permian.

The Late Paleozoic is considered to be an important stage in the evolution of the Central Asian Orogenic Belt (CAOB). The Bogda Mountains, a northeastern branch of the Tianshan Mountains, record the complete Paleozoic history of the Tianshan orogenic belt. The tectonic and sedimentary evolution of the west Bogda area and the timing of initial uplift of the West Bogda Mountains were investigated based on detailed sedimentological study of outcrops, including lithology, sedimentary structures, rock and isotopic compositions and paleocurrent directions. At the end of the Early Permian, the West Bogda Trough was closed and an island arc was formed. The sedimentary and subsidence center of the Middle Permian inherited that of the Early Permian. The west Bogda area became an inherited catchment area, and developed a widespread shallow, deep and then shallow lacustrine succession during the Mid-Permian. At the end of the Mid-Permian, strong intracontinental collision caused the initial uplift of the West Bogda Mountains. Sedimentological evidence further confirmed that the West Bogda Mountains was a rift basin in the Carboniferous-Early Permian, and subsequently entered the Late Paleozoic large-scale intracontinental orogeny in the region.