The significance of the subnanometre regime is difficult to overstate. In 2003, a landmark theoretical study by Junquera and Ghosez in Nature examined ultrathin ferroelectric films sandwiched between electrodes and identified a critical thickness of roughly three unit cells, below which the depolarizing field generated by incomplete charge screening would suppress the spontaneous polarization entirely. That prediction set a de facto benchmark for the community: to build ever-denser ferroelectric memories, researchers would need either better electrode screening, new material chemistries, or structural tricks that stabilize polar phases where conventional ones fail.

Gallium oxide is an unusual candidate for this challenge. In its most common form, the beta phase, Ga2O3 is a monoclinic, centrosymmetric crystal with a wide bandgap of about 4.8 electronvolts, properties that have made it a darling of power electronics but seemingly disqualifying for ferroelectricity, which requires a non-centrosymmetric structure with two energetically degenerate polar states. The material does, however, host a rich family of polymorphs, including metastable phases with different coordination geometries, and this structural flexibility is precisely what strain engineering exploits. By imposing mechanical strain, either through epitaxial growth on a mismatched substrate or through the confinement inherent in ultrathin layers, researchers can destabilize the equilibrium phase and drive the crystal into alternative structures that would not exist in bulk form.

According to the Nature Electronics analysis by Zhao Guan and Ni Zhong of East China Normal University, writing in the journal’s News and Views section, the central result they discuss involves a gallium oxide layer just half a unit cell thick. At this extreme dimension, strain engineering converts the material into a zincblende-type phase, a cubic crystal structure in which the gallium and oxygen sublattices are displaced relative to one another in a way that permits a switchable polarization. Crucially, the polarization can be reversed by an applied voltage of less than one volt, a coercive voltage low enough to be directly relevant to low-power electronic devices and compatible with the operating budgets of modern integrated circuits.

The evidence for ferroelectricity at this scale rests on what the analysis describes as both microscopic and macroscopic measurements. Microscopic probes can resolve the local atomic displacements that define the polar zincblende phase, confirming that the crystal structure itself has transformed and that the gallium and oxygen atoms occupy positions consistent with a non-centrosymmetric lattice. Macroscopic measurements, in turn, demonstrate the functional signature: a switchable polarization response that reverses under an applied electric field, the defining behavior of a ferroelectric. Together, these two lines of evidence address the most common objection to claims of ultrathin ferroelectricity, namely that apparent switching signals might arise from extraneous effects such as charge injection, electromigration or interfacial chemistry rather than genuine lattice polarization.

The low coercive voltage deserves particular attention from a device engineering standpoint. In conventional ferroelectric films, the switching voltage scales with thickness, so that aggressive miniaturization paradoxically demands enormous electric fields to reverse the polarization, fields that can exceed the breakdown strength of surrounding dielectrics and electrodes. A subnanometre film that switches below one volt inverts that logic: the same dimensional scaling that makes the film attractive for density also makes it attractive for energy efficiency. If the result can be reproduced in manufacturable device geometries, it points toward ferroelectric memory elements that operate at voltages comparable to those of mainstream CMOS logic, removing one of the long-standing barriers to integrating ferroelectrics directly into advanced chip stacks.

The work also connects to a broader wave of interest in two-dimensional and quasi-two-dimensional ferroelectrics. The analysis situates the gallium oxide result alongside recent studies of ultrathin oxide systems, including work by Zhao and colleagues published in Physical Review B in 2021 on the theoretical behavior of confined ferroelectric phases, and a 2026 Science Advances study by Shen and coauthors. Additional recent contributions from Wang, Sun and Mei in Advanced Functional Materials and from Jiang and colleagues in Nature Electronics have explored related routes to functional polarization in dimensionally confined oxides. Taken together, this literature suggests a field in transition, moving from the pessimistic critical-thickness picture of the early 2000s toward a toolkit of strain, epitaxy and dimensional confinement that can stabilize polar order in structures once thought impossible.

For gallium oxide specifically, the implications extend beyond memory. The material’s ultra-wide bandgap already underpins a growing industry in high-power transistors, solar-blind ultraviolet photodetectors and radiation-hardened electronics, and a ferroelectric polymorph accessible through strain engineering would add non-volatile functionality to a material system that is already technologically mature in other respects. Device designers have long sought ways to combine power handling with memory and sensing on a single platform; a strain-stabilized ferroelectric phase of Ga2O3 offers a conceptual route to exactly that integration, allowing polarization-based state variables to coexist with the high-voltage capability of the beta phase on related growth platforms.

Significant challenges remain before the laboratory demonstration translates into products. Strain engineering at the half-unit-cell level demands exquisite control of growth conditions, substrate choice and interface quality, and the metastable zincblende phase must remain stable through the thermal budgets of real fabrication processes. Endurance, retention and fatigue, the classic reliability metrics of ferroelectric devices, have yet to be established for this system, and the depolarizing effects that once doomed ultrathin ferroelectrics will still need careful management through electrode and interface design. The analysis by Guan and Zhong makes clear, however, that the conceptual barrier has fallen: ferroelectricity is not intrinsically incompatible with the subnanometre scale. As the authors frame it, a half-unit-cell layer of gallium oxide, reshaped by strain into a switchable polar phase and flipped by less than a volt, redefines what engineers can expect from the smallest ferroelectric structures, and it will likely energize a new round of experimental and theoretical work aimed at turning subnanometre polarization from a scientific curiosity into a working component of future electronics.

Subject of Research: Subnanometre-scale ferroelectricity induced by strain engineering in ultrathin gallium oxide layers

Article Title: Subnanometre ferroelectricity in strain-engineered gallium oxide

Article References: Guan, Z., & Zhong, N. (2026). Subnanometre ferroelectricity in strain-engineered gallium oxide. Nature Electronics. https://doi.org/10.1038/s41928-026-01702-4

Image Credits: AI Generated

DOI: 10.1038/s41928-026-01702-4

Keywords: ferroelectricity, gallium oxide, strain engineering, zincblende phase, two-dimensional materials, ultrathin films, coercive voltage, Nature Electronics, memory devices, polarization switching, wide bandgap semiconductors, critical thickness