Design of Modified Polymer Membranes Using Machine Learning

Surface modification is an attractive strategy to adjust the properties of polymer membranes. Unfortunately, predictive structure-processing-property relationships between the modification strategies and membrane performance are often unknown. One possibility to tackle this challenge is the application of data-driven methods such as machine learning. In this study, we applied machine learning methods to data sets containing the performance parameters of modified membranes. The resulting machine learning models were used to predict performance parameters, such as the pure water permeability and the zeta potential of membranes modified with new substances. The predictions had low prediction errors, which allowed us to generalize them to similar membrane modifications and processing conditions. Additionally, machine learning methods were able to identify the impact of substance properties and process parameters on the resulting membrane properties. Our results demonstrate that small data sets, as they are common in materials science, can be used as training data for predictive machine learning models. Therefore, machine learning shows great potential as a tool to expedite the development of high-performance membranes while reducing the time and costs associated with the development process at the same time.

Solution Atomic Layer Deposition of Smooth, Continuous, Crystalline Metal-Organic Framework Thin Films

For the first time, a procedure has been established for the growth of surface-anchored metal-organic framework (SURMOF) copper(II) benzene-1,4-dicarboxylate (Cu-BDC) thin films of thickness control with single molecule accuracy. For this, we exploit the novel method solution atomic layer deposition (sALD). The sALD growth rate has been determined at 4.5 Å per cycle. The compact and dense SURMOF films grown at room temperature by sALD possess a vastly superior film thickness uniformity than those deposited by conventional solution-based techniques, such as dipping and spraying while featuring clear crystallinity from 100 nm thickness. The highly controlled layer-by-layer growth mechanism of sALD proves crucial to prevent unwanted side reactions such as Ostwald ripening or detrimental island growth, ensuring continuous Cu-BDC film coverage. This successful demonstration of sALD-grown compact continuous Cu-BDC SURMOF films is a paradigm change and provides a key advancement enabling a multitude of applications that require continuous and ultrathin coatings while maintaining tight film thickness specifications, which were previously unattainable with conventional solution-based growth methods.

Ultrafast Interface Charge Separation in Carbon Nanodot-Nanotube Hybrids

Carbon dots are an emerging family of zero-dimensional nanocarbons behaving as tunable light harvesters and photoactivated charge donors. Coupling them to carbon nanotubes, which are well-known electron acceptors with excellent charge transport capabilities, is very promising for several applications. Here, we first devised a route to achieve the stable electrostatic binding of carbon dots to multi- or single-walled carbon nanotubes, as confirmed by several experimental observations. The photoluminescence of carbon dots is strongly quenched when they contact either semiconductive or conductive nanotubes, indicating a strong electronic coupling to both. Theoretical simulations predict a favorable energy level alignment within these complexes, suggesting a photoinduced electron transfer from dots to nanotubes, which is a process of high functional interest. Femtosecond transient absorption confirms indeed an ultrafast (<100 fs) electron transfer independent of nanotubes being conductive or semiconductive in nature, followed by a much slower back electron transfer (≈60 ps) from the nanotube to the carbon dots. The high degree of charge separation and delocalization achieved in these nanohybrids entails significant photocatalytic properties, as we demonstrate by the reduction of silver ions in solution. The results are very promising in view of using these "all-carbon" nanohybrids as efficient light harvesters for applications in artificial photocatalysis and photosynthesis.