Surfaces and interfaces play key roles in chemical and material science. Understanding physical and chemical processes at complex surfaces and interfaces is a challenging task. Machine learning provides a powerful tool to help analyze and accelerate simulations. This comprehensive review affords an overview of the applications of machine learning in the study of surfaces and interfaces of chemical systems and materials. We categorize surfaces and interfaces into the following broad categories: solid–solid interface, solid–liquid interface, liquid–liquid interface, surface of solid, surface of liquid, and three-phase interfaces. High-throughput screening, combined machine learning and first-principles calculations, and machine learning force field accelerated molecular dynamics simulations are used to rational design and study physical and chemical processes of surfaces and interfaces in systems such as all-solid-state batteries, solar cells, and heterogeneous catalysis. This review provides detailed and comprehensive information on the applications of machine learning on surfaces and interfaces for chemical and material science.

In this paper, polyethyleneterephthalate (PET) films with and without plasma pretreatment were modified by atomic layer deposition (ALD) and plasma-assisted atomic layer deposition (PA-ALD). It demonstrates that the Al₂O₃ films are successfully deposited onto the surface of PET films. The cracks formed on the deposited Al₂O₃ films in the ALD, plasma pretreated ALD, and PA-ALD were attributed to the energetic ion bombardment in plasmas. The surface wettability in terms of water contact angle shows that the deposited Al₂O₃ layer can enhance the wetting property of modified PET surface. Further characterizations of the Al₂O₃ films suggest that the elevated density of hydroxyl -OH group improve the initial growth of ALD deposition. Chemical composition of the Al₂O₃-coated PET film was characterized by X-ray photoelectron spectroscopy, which shows that the content of C 1s reduces with the growing of O 1s in the Al₂O₃-coated PET films, and the introduction of plasma in the ALD process helps the normal growth of Al₂O₃ on PET in PA-ALD.

Superhydrophobic antibacterial fabric possesses properties of antibacterial and antibacterial adhesion and shows huge demand in the field of medical textiles. However, current technologies are unable to fully address this. Hence, a simple method is highly desirable. Herein, the pristine polyester (PET) fabric is immersed into the solution containing ZnO nanoparticle and polydimethylsiloxane (PDMS), and the fiber surfaces are uniformly covered by a ZnO-PDMS layer after being treated by low pressure Ar plasma. The weight gain rate of the treated fabric is 3.5%, which is basically unchanged, and the air permeability, moisture permeability, and tensile properties of the fabric are basically not affected. It is found that the water contact angle (WCA) of the fabric is over 162.7°and sliding angle (SA) is less than 10°. The stable binding of PDMS and PET fibers induces a robust superhydrophobicity; even after 300 washing cycles and 600 friction cycles, it still remains superhydrophobic. The antibacterial rates of Escherichia coli and Staphylococcus aureus before washing were 99.89% and 99.85%, respectively, and after 100 cycles of washing, the antibacterial rates decreased to 99.36% and 99.17%, respectively. Therefore, it shows a good development prospect in the application of protecting clothing or textiles that require good antibacterial properties (such as bed sheets, duvet covers, etc.).