Publication date: 15th December 2025
Two-dimensional perovskite nanoplatelets (NPLs) promise atomically precise emission control—but achieving that precision in the flask is anything but trivial. Their formation teeters on a fine balance between solvent polarity, precursor chemistry, and the split-second timing of an antisolvent drop. In this talk, I will show how data and diffraction together can turn this delicate art into a predictable science. Using in situ X-ray scattering and photoluminescence, we uncover how emissive nanocluster intermediates evolve into either rods or platelets depending solely on the antisolvent’s dipole moment and hydrogen-bonding strength—essentially, how chemistry “decides” the dimensionality.[1] Building on this mechanistic insight, our machine-learning platform Synthesizer transforms these parameters into predictive maps of color, linewidth, and quantum yield.[2] Within just a few syntheses, it delivers nm-level precision over emission and aspect ratio, all under ambient conditions. Together, these studies define a quantitative recipe for crafting bright, narrow, and stable 2D NPLs—and hint at a future where algorithms, not trial-and-error, steer nanocrystal growth.
