Controlled dewetting and phase transition hysteresis of VO2 nanostructures

Abstract

As artificial intelligence continues to grow, so does the need for more efficient ways to process data. Besides moving from electronic to photonic circuits, a promising approach is to integrate phase-change materials. Vanadium dioxide (VO$_2$) exhibits an ultrafast, near-room-temperature phase transition, characterized by hysteresis and large optical modulation -- making it a promising candidate for short-term memories and for mimicking neural behavior in brain-like computing systems. While the hysteresis behavior of VO$_2$ has been well studied in thin films and nanostructures, practical control and device integration have been limited only to thin films. Here, we demonstrate control over the phase transitions of VO$_2$ nanocylinders via lithographic patterning, controlled crystallization, and controlled dewetting. Because nanostructures are easier to address and consume less power than films, the ability to fabricate them with tailored geometry and hysteresis properties directly on integrated platforms is a key step toward scalable, energy-efficient memory and neuromorphic photonic devices.