Wrinkle Mediated Phase Transitions in In$_2$Se$_3$

TL;DR

The study demonstrates a non-contact, room-temperature pathway to reversibly switch between the $\alpha$ and $\beta'$ phases of In2Se3 by laser-induced wrinkle formation, bypassing cryogenic steps. Accumulated strain from wrinkling drives the $\beta'$→$\alpha$ transition, which can be repeated via annealing to recover the original phase and domain structure, enabling multiphase heterostructures within a single flake. Domain imaging (PD-PEEM) reveals wrinkle-linked reorganization and sharp lateral $\alpha$/$\beta'$ boundaries, suggesting functional junctions for ferroic devices. The work provides a practical route to engineer ferroic states and phase-change memory architectures in 2D In2Se3 and points to tunable domain patterns via intraflake strain.

Abstract

Crystalline phase transitions in two-dimensional materials enable precise control over electronic and ferroic properties, making them attractive materials for memory and energy storage applications. In$_2$Se$_3$ is particularly promising because its $α$ and $β'$ phases are both stable at room temperature but exhibit distinct ferroic behaviors. However, achieving reliable reversible switching between these states remains challenging. Here, we show that controlled $β'\rightarrowα$ phase transitions in 2D In$_2$Se$_3$ become accessible through laser-induced wrinkling, establishing a room-temperature approach for manipulating ferroic states in In$_2$Se$_3$ thin films. Combined with thermal annealing for phase recovery, this approach eliminates cryogenic steps and mechanical perturbation while harnessing accumulated internal strain to generate multiphase heterostructures and direct domain reorganization. This pathway for phase transitions in In$_2$Se$_3$ opens the door for further development in ferroic device architectures and phase-change memory technologies.

Figures (5)

• Figure 1: (A) Schematic of the laser induced wrinkle phase transition experiment. The laser is focused onto the sample while being monitored by a camera affixed with a microscope objective. (B,C) Cross-polarized optical images of $\beta'$- and $\alpha$-In2Se3 before and after laser exposure. Scalebar is $20\ \mu \text{m}$. (D) Raman spectra before and after wrinkling phase transition. Blue line is the pristine flake before wrinkling threshold experiment and red is after wrinkling. (E) AFM image of a wrinkled flake surface. Inset shows height profile along the line cut. Scalebar is $2\ \mu \text{m}$. (F) Box plots of input energy required to induce wrinkles in flakes on Si and SiO2 substrates. Red lines are medians; boxes and whiskers indicate interquartile ranges. Points are individual measurements.
• Figure 2: (A) Cross-polarized optical images of an In2Se3 flake on Si through two phase cycles consisting of laser exposure and annealing. (B,C) Optical and AFM images of a wrinkled $\alpha$ phase flake after laser exposure and (D,E) the same flake after converting back to $\beta'$ via annealing. (F) Heights of various wrinkles before and after annealing measured with AFM . Scalebars are $20\mu$m for optical images and $1\mu$m for AFM images.
• Figure 3: (A) Raman spectra from the right (red trace) and left (blue trace) sides of the flake in Figure \ref{['fig:phase_cycle']}(A) after phase cycling. (B) Map of integrated Raman intensity for the range (109-119 cm$^{-1}$) highlighted in gray in (A) overlayed over a cross-polarized optical image of the flake. Pink corresponds to greater intensity of the $\beta'$-In2Se3 $A_1$ mode. Scalebar is $10\ \mu$m.
• Figure 4: (a) Static PEEM images of In2Se3 through two phase cycles consisting of laser ($h\nu=3.1$ eV) exposure and annealing. Scalebar is $10\mu$m. (B) Maps of $\theta_\text{TDM}$ at from the marked regions in (A) for each step in the phase cycling process. Scalebar is $2\mu$m. (C) Histogram of domain boundary orientations throughout the phase cycling process.
• Figure 5: Table of Contents Figure