DEMO-QIS · Data-driven Design of Molecular Spin Qubits for Quantum Information Science

DEMO-QIS aims to revolutionise the discovery of molecular spin qubits (MSQs) for quantum information science (QIS). QIS promises technologies ranging from advanced computing to detectors that sense atomic-scale signals in medicine, navigation, and beyond. Realising such breakthroughs requires qubits that are reliable, scalable, and long-lasting. MSQs offer unique advantages: bottom-up design, tunability through chemical modification, compatibility with device integration, and spin-state readout via optically detected magnetic resonance (ODMR). Progress requires two pathways: optimising key figures of merit (FoMs) such as ODMR contrast, photoluminescence (PL) intensity, and coherence time, and exploring new candidates within the vast chemical space. FoMs are interdependent; improving one may compromise another, making trial-and-error ineffective. Intuition-based exploration is inadequate given the vast space. DEMO-QIS will address these challenges by combining density functional theory (DFT) with machine learning (ML). Starting with pentacene derivatives, we first quantify the contrast and PL intensity, generate substitution datasets, and derive design rules. We expand to hydrocarbon databases with supervised ML, enhanced with fine-tuning and few-shot learning, while generative models propose de novo design candidates. The design space will be broadened through radical functionalisation to stabilise high-spin states and by optimising MSQ–host combinations to improve coherence. Finally, we will unify all steps into a transferable pipeline that integrates quantum simulations, predictive modelling, candidate retrieval, and generative design. DEMO-QIS will provide both a workflow for MSQ discovery and candidates for experimental validation. By redefining MSQ research, it will accelerate Europe’s progress in quantum technologies, establish transferable MSQ design methods, and promote interdisciplinary training at the interface of chemistry, physics, and data science.