Steering Solid–Solution Equilibria to Tame Phase Complexity and Instability in Dion-Jacobson Quasi-2D Halide Perovskites

Dion-Jacobson (DJ) quasi-two-dimensional (2D) halide perovskites (HPs) have emerged as a robust platform for optoelectronic applications, yet their crystallization pathways and subsequent phase evolution remain insufficiently understood. Herein, we integrate robotic high-throughput synthesis with automated spectral diagnosis and multimodal characterization to investigate 1,4-butanediammonium-based HPs containing methylammonium (MA) or formamidinium (FA) across composition and temperature. Tracking phase evolution and cation partitioning reveals competition between lattice stabilization, coarsening, and reconstruction within a coupled heterogeneous reaction system. Continued MA incorporation changes both the solids and surrounding solution, whereas greater initial FA capture does not ensure higher- n phase retention. Comparative extraction and spacer-treatment experiments support an additional pathway linking δ-containing FAPbI 3 structures to FA release and lower- n reconstruction. These findings provide experimental insight into the intricate solid–solution chemical equilibria underlying crystal maturation beyond nucleation. Exploiting this understanding, machine-learning-assisted multi-objective screening identifies synthesis conditions that sustain homogeneous, spectrally ‘frozen’ emission, guiding the formation of chemically stabilized, structurally robust HP frameworks. Applying the selected formulation to HP interfaces yields efficient solar cells with markedly improved thermal stability. Overall, this integrated experiment-analysis workflow translates solid–solution equilibrium chemistry into data-guided control of phase complexity, offering a strategy for durable HP optoelectronics.