Researchers from the University of Nevada Las Vegas, Lawrence Berkeley National Laboratory, International Kazakh-Turkish University, University of California and Argonne National Laboratory have introduced a predictive synthesis framework to boost the spin-relevant performance of chiral 2D metal halide perovskites (MHPs) for next-generation spintronics. Chiral 2D MHPs are promising materials for spin-optoelectronic devices that exploit the electron’s spin degree of freedom, yet their chiroptical response, quantified by the absorption dissymmetry factor (gabs), has shown large variability and poor reproducibility. This has hindered the rational design of reliable spintronic components such as circularly polarized LEDs, photodetectors, and spin filters.
To tackle this challenge, the team built a data-driven framework that directly links synthesis “knobs” to chiroptical properties. Using Pearson’s correlation, ANOVA, and Gaussian process regression, they systematically evaluated how solvent choice, annealing temperature, film thickness, and other structural and morphological factors influence gabs. The analysis reveals solvent choice as the primary driver of variability: acetonitrile (ACN)-processed films consistently exhibit higher and more reproducible gabs values than films fabricated from dimethylformamide (DMF) or ACN:dimethyl sulfoxide (ACN:DMSO) mixtures. For ACN-based films, the model identifies specific annealing temperature and thickness ranges that maximize gabs, providing a clear processing playbook instead of ad hoc optimization.
Films produced from higher-boiling-point solvent systems, on the other hand, show more intricate behavior, where chiroptical response depends on coupled parameters including annealing temperature, excitonic integral intensity, and film texture. By untangling these dependencies, the framework explains why apparently similar processing routes can yield drastically different spin-relevant properties. Importantly for the spintronics community, the established synthesis - property correlations offer a practical roadmap to engineer chiral perovskite layers with tailored spin selectivity and robust circularly polarized light response.
The authors position this methodology as a foundation for rational design and accelerated discovery of chiral semiconductors for spintronics. It opens the door to active-learning and machine learning-driven closed-loop experiments, enabling targeted exploration of new chiral cations, perovskite compositions, and device architectures. Such capabilities could significantly speed up the development of chiral perovskite-based spin filters, spin-LEDs, and other spin-optoelectronic components operating under technologically relevant conditions.