Minisymposium

Heterogeneous surfaces, such as amorphous silica, are characterized by a variety of local atomic environments. The mutual disposition of individual local environments (i.e., different adsorption sites) impacts the adsorption behavior of gas molecules. The resulting adsorption landscape displays irregularly shaped patterns that must be segmented to resolve the complexity of the adsorption mechanisms. We propose an optimized segmentation protocol that allows the control of the extension and morphology of the segmented patterns. We optimize the segmentation protocol to predict the mean residence time of carbon dioxide within segmented regions, which allows the identification of regions where the residence time does not follow an exponential distribution. Thedeviation from the exponential distribution is the signature of the presence of surface defects on the amorphous surfaces. The residence time distributions serve to implement up-scaled adsorption models such as microkientic models. Next, we investigate the effect of the adsorption sites on the adsorption landscape to find a correlation between highly adsorptive regions and the underlyinglocal environment.

Phosphorus is an essential element for life and a critical component of agricultural fertilizers, yet its sustainable management remains a pressing global challenge. Excess phosphorus runoff contributes to environmental pollution, while limited high-quality phosphate rock reserves raise concerns about future availability. This study explores the intersection of AI and molecular modeling to advance phosphorus sustainability. By leveraging molecular dynamics simulations and AI-driven predictive modeling, we investigate the interactions between phosphorus-binding proteins and various phosphorus species to enhance selective capture and recycling strategies. Machine learning algorithms are applied to predict modular peptide sequences with high binding affinity for phosphorus, facilitating the design of novel biomimetic materials for phosphorus recovery. Additionally, we utilize AI-powered data integration to analyze large-scale phosphorus-related datasets, enabling more efficient resource utilization and policy development. This interdisciplinary approach has the potential to revolutionize phosphorus management by improving recovery efficiency, reducing environmental impact, and ensuring long-term sustainability.

Per- and poly-fluoroalkyl substances (PFAS) have emerged as persistent environmental pollutants, posing significant risks to human health and ecosystems due to their extreme chemical stability (also known as forever chemicals) and bioaccumulation potential. Conventional remediation approaches, such as adsorption and degradation, have shown limited efficacy in fully eliminating PFAS from contaminated water sources. In this perspective talk, we want to explore and explain how applying nanoparticle-based strategies could be a transformative solution for PFAS mitigation. We will discuss the application of iron oxide magnetic clusters for PFAS remediation and address the remaining gaps in the literature. A comprehensive understanding of PFAS-specific adsorption mechanisms, particularly at the molecular level, where computational approaches could provide deeper insights, is needed. Most functionalization strategies focus on simple surface modifications, while dual-functionalized or stimuli-responsive coatings remain underexplored. PFAS molecules exhibit amphiphilic properties. Current functionalization strategies often rely solely on electrostatic interactions, which may be effective for long-chain PFAS but fail to efficiently capture short-chain PFAS, which are more hydrophilic and challenging to remove.Strategies such as coatings incorporating pH-responsive polymers can facilitate pH-controlled PFAS desorption. Also, light-sensitive materials, such as azobenzene-functionalized surfaces, could enable PFAS capture.

Rethinking Sustainability Through the Lens of AI and Nanotechnology