The commercialization of low temperature fuel cells is restricted by the high cost and low durability of cathode catalysts. Intense efforts have been devoted to tackle this issue by engineering the structure of Pt-based catalysts. Herein, a novel concept towards enhancing the performance of low temperature fuel cell catalysts is proposed, namely by tuning the local active site microenvironment with an immobilized ionic liquid (IL) phase. As demonstrated by the applicant in preliminary work, a suitable IL layer strongly influences the active catalytic site in a very promising manner, apparently via a highly complex interplay of solvent-, ligand- and electrostatic-stabilization effects. As the structural versatility of ILs allows for rational engineering of this modification at molecular level, the proposed project aims for a full scientific exploration of the remarkable activation and stabilization effects in ORR, to enable the realization of an innovative fuel cell cathode with dramatically enhanced performance. To achieve this ambitious goal, a sound fundamental understanding of the interaction of ILs with electrocatalytic sites will be derived by making use of the excellent research infrastructure and longstanding experience in ionic liquid design and catalytic materials at our institute. To demonstrate the general applicability, the deduced principals will also be applied to CO2 electrochemical reduction. The approach will not stop at the design of novel catalyst systems, but will address solutions to ensure long-term stability of the IL modification. To avoid IL leaching from the catalyst over time, the recent success of the applicant in the synthesis of novel core/shell carbon materials will be employed. The IL will be synthesized in situ within a mesoporous core and the steric demanding ions fixed through a molecular sieving shell surrounding each catalyst particle. A model-assisted strategy will be applied for optimization of the core/shell pore structures.

Deep learning approaches, mostly in the form of convolutional neural networks (CNNs), have taken the field of computer vision by storm. While the progress in recent years has been astounding, it would be dangerous to believe that important problems in computer vision are close to being solved. Many canonical deep networks for vision tasks ranging from image understanding to 3D reconstruction or motion estimation perform incredibly well "on dataset", i.e.~in the very setting in which they have been trained. The generalization to novel, related scenarios is still lacking, however. Moreover, large amounts of labeled data are required for training, which are not available in all potential application areas. In addition, the majority of deep networks in computer vision show deficiencies in terms of explainability. That is, the role of network components is often opaque and most deep networks in vision do not output reliable quantifications of the uncertainty of the prediction, limiting the comprehension by users. In this project, we aim to significantly advance deep networks in computer vision toward improved robustness and explainability. To that end, we will investigate structured network architectures, probabilistic methods, and hybrid generative/discriminative models, all with the goal of increasing robustness and gaining explainability. This is accompanied by research on how to assess robustness and aspects of explainability via appropriate datasets and metrics. While we aim to develop a toolbox that is as independent of specific tasks as possible, the work program is grounded in concrete vision problems to monitor progress. We specifically consider the challenges of 3D scene analysis from images and video, including tasks such as panoptic segmentation, 3D reconstruction, and motion estimation. We expect the project to have significant impact in applications of computer vision where robustness is key, data is limited, and user trust is paramount.

Conventional thermal insulation materials like mineral wools, polyurethane, and polystyrene foams are widely used in Europe, but their production significantly contributes to CO2 emissions, generating up to 6.9 times CO2. Additionally, these materials impede air circulation in buildings, leading to poor indoor air quality and discomfort. This project seeks to address the need for carbon-neutral or carbon-negative building materials that not only provide thermal insulation but also enhance indoor air quality through effective air filtration. The focus of this research is on developing lightweight indoor concrete blocks utilizing biochar, a carbon-rich material produced through pyrolysis. Biochar can sequester up to 2.6 times its own weight of CO2 while offering exceptional properties such as air filtration, acoustics, and thermal insulation. The project aims to maximize biochar content in concrete blocks to 50 to 75 vol% to achieve superior air filtration and thermal comfort while maintaining adequate mechanical performance. The methodology involves developing granulated biochar aggregates and combining them with low-carbon binders and biochar fines to develop lightweight, carbon-negative concrete blocks. Various mix designs, curing conditions, and additives will be explored to optimize the blocks' physical, mechanical, and durability properties. Advanced analytical techniques, including microstructural studies and multiscale modelling, will be employed to understand and validate the blocks' performance. This project has the potential to significantly advance sustainable building practices by developing a novel material that aligns with EU sustainability goals, the Paris Agreement, and the European Green Deal. By integrating biochar into building materials, this research offers an innovative approach to improving indoor air quality and energy efficiency, contributing to healthier and more sustainable living environments.

In recent years, the analysis of large data sets is becoming increasingly important in the fields of material science and engineering. There is a strong demand for real-time automated identification algorithms in electron microscopy (EM) for the analysis of atomic-structure, phases, and defects. Unfortunately, it is non-trivial to obtain or extract meaningful scientific information from raw EM output digital data. It requires a tedious process of filtering/fitting and the expertise of a seasoned microscopist. With the rapid development of information technology and computer science, automated computer-assisted analysis of electron microscopy images/data is becoming a reality. In the past decade, different techniques have been developed and applied to digital data analysis. Meanwhile, the rapid development of novel microscopy techniques and instrumentation, e.g. in situ/operando and pixelated detector-based techniques, require high-speed data execution and analysis. Currently, several groups worldwide are concentrating their efforts into implementing machine learning and deep learning algorithms for image/data analysis. However, this is still a very undeveloped direction in the field of electron microscopy for materials science, especially in Europe. According to the Digital Transformation Monitor, artificial intelligence-based technologies will play a major role in future economy. The ability to analyse levels of data that are beyond human comprehension will allow business to personalize experiences, customize products and services and identify growth opportunities with a speed and accuracy that has never been possible before. The objective of this PoC is to generate an innovative software package that enables the analysis of large sets of EM data (i) at high throughput with (ii) low costs, in (iii) a standardized approach and (iv) under operando conditions, based on advanced machine learning algorithms.

This is a project for higher education student and staff mobility between Programme Countries and Partner Countries. Please consult the website of the organisation to obtain additional details.