Engineering Small-Molecule Proton-Transfer Ferroelectrics by Crystal Structure Prediction: Design Limits at the Salt-Cocrystal Boundary.

Organic molecular ferroelectrics hold significant potential in organic electronics due to their chemical tunability and straightforward fabrication methods. Among these, acid-base proton-transfer (PT) salts are notable for their low coercive fields and fast switching capabilities but are limited by relatively low spontaneous polarization. Using smaller molecular species can in principle increase the polarization, but requires both stabilization of the monovalent salt state and crystal packing that supports ferroelectric PT pathways. Using a crystal-structure prediction (CSP)-based design combined with density functional theory (DFT), we investigate 30 combinations of molecular acids and bases aimed at enhancing the dipole density. We identified several crystal structures with PT-capable hydrogen-bonding networks, and in our initial DFT ranking, three candidate ferroelectric packings and one antiferroelectric. Subsequent experimental work on two representative systems, while confirming the ability of CSP to predict PT-capable packing motifs, found neutral cocrystals rather than the desired monovalent salts supporting ferroelectricity. More detailed computational analysis traced the disparity to the relative stability of protonation states, which is strongly sensitive to the exchange-correlation functional and to vibrational zero-point energy contributions. Thus, while our CSP study correctly identified proton-transfer crystal packing motifs, it failed at the level of protonation-state stability, which we found to be strongly influenced by exchange-correlation choice and vibrational free energy.