Accelerated design of functional 2D materials by leveraging foundation models in a closed-loop approach (2DFound)

Background 

The reshaping of the biosphere’s functions presents profound challenges to Earth’s environmental conditions. Addressing present and future needs requires innovative technologies, where novel materials, particularly two-dimensional (2D) materials, will play a key role.  

With their exceptional properties emanating from a high surface-to-volume ratio, 2D materials are transformative factors in energy storage, catalysis, carbon capture, pollutant degradation, and more.  

Research Question  

The 2DFound project will establish a world-leading research environment with a mission to realize next generations of 2D materials from selective etching, and to develop machine learning (ML) methods for accelerating their discovery. For groundbreaking materials development, it is timely to initiate a large-scale project targeting sustainable and applicable 2D materials. 

2DFound will advance ML-based predictive models and in situ characterization tools to identify 3D precursors and verify the 2D derivatives through tailored synthesis protocols. Further, the project will analyze their structure and properties down to the atomic level. The approach is to merge ML-driven materials prediction and verification, enabling inverse materials design to create materials tailored to specific properties. 

Aim  

Building on earlier breakthroughs in synthesis of other advanced 2D materials from similar methods, like boridene and goldene, 2DFound aims to establish a cutting-edge research environment to discover new families of 2D materials.  

An ultimate goal is inverse design, to provide materials on demand. This is facilitated by a foundation model, the 2DFound Model (2DFM), which integrates novel atomistic models tailored to 2D materials. It builds on predictive and generative capabilities, including dynamical models of reactions on surfaces, guided by multidimensional data from electron microscopy. 

Synergy and Team 

The project integrates six excellent research divisions (AIICS, CVL, STIMA, Materials Design, Thin Film Physics, and Theoretical Physics) at three departments (IDA, ISY, and IFM) at Linköping University into a new interdisciplinary research environment. In addition, Swedish industry partners are engaged in the project.  

This interdisciplinary team will push the boundaries of 2D materials research, recruiting new PhD students and postdoctoral researchers to establish a dynamic WASP-WISE research platform at Linköping University. 

Sustainability 

While all future needs and challenges cannot be predicted, solutions will benefit from new technologies, components, and devices. Novel materials will be a vital part of this development. In particular, 2D materials hold extraordinary properties for energy storage, carbon capture, catalysis, pollutant degradation, and beyond. 

By pioneering next generation 2D materials and machine learning methods, the project addresses pressing challenges in sustainability, energy, and environmental applications.