Discovery of Ferroelectric Twin Boundaries in a Photoactive Halide Perovskite

arXiv:2601.12892 · cond-mat.mtrl-sci · Submitted 2026-01-19 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Condensed Matter. The day in condensed matter: materials, phases and the measurements under them.

Kai: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "Discovery of Ferroelectric Twin Boundaries in a Photoactive Halide Perovskite".

Mira: Here's the gist of the paper. Discovery of ferroelectric twin boundaries in a photoactive halide perovskite is significant because it reveals hidden ferroic functionality in halide perovskite semiconductors through domain wall engineering The paper discovers the existence of ferroelastic twin boundaries in vapor-deposited CsPbI3 thin films, parallel to the 110 and 112 planes Remarkably, despite photoactive CsPbI3 being centrosymmetric and non-polar, Pb atoms shift at 110 twin boundaries driving a local ferroelectric-like polarisation These polar twin walls form an intrinsic array of nanoscale functional interfaces, spaced 30-50 nm apart, embedded within the non-polar perovskite lattice In contrast, 112 twin boundaries remain non-polar but strongly suppress octahedral tilt and offcentre Cs atom displacements, revealing a different untapped ferroic degree of freedom These discoveries together uncover previously hidden ferroic functionality in halide perovskite semiconductors, opening opportunities for enhanced conductivity and photovoltaic behaviour through domain wall engineering Non-polar Bulk Characterization The study began by determining the crystal structure, symmetry and polarity of bulk CsPbI3 Using annular dark field scanning transmission electron microscopy (STEM-ADF), they confirmed that CsPbI3 adopts an orthorhombic perovskite structure, commonly referred to as γ-CsPbI3 Quantitative analysis found no detectable Pb off-centre displacements anywhere in the single-crystal regions away from domain walls, confirming there is no spontaneous polarisation in the bulk Consistent with the γ-CsPbI3 phase, Cs shows anti-polar off-centre displacements, that is, Cs atoms on adjacent A-sites shift in opposite directions along the

001: direction These antipolar displacements arise from octahedral tilts rather than spontaneous polarisations Characterization of 110 Twin Boundaries The investigation extended to twin domains and twin boundaries, revealing that the crystallography of the 110 twin boundary in CsPbI3 is illustrated schematically Localised Polarisation at 110 Twin Boundaries showed significant off-centre displacements of the Pb/I atomic columns localised at the twin boundary, with the displacement vector predominantly parallel to

110: , along the boundary plane These displacements are confined to a narrow region approximately four atomic layers wide, indicating that the boundary itself hosts a distinct structural configuration which is polar The peak magnitude of the displacement reaches 15 pm at the twin boundary, clearly distinguishing it from the surrounding bulk domains, where displacements remain negligible Characterization of 112 Twin Boundaries The study also identified a second type of twin boundary in CsPbI3 films, orientated parallel to the 112 plane Interestingly, no measurable displacement of the Pb/I columns relative to the centres of the surrounding octahedra is observed at 112-type twin boundaries, indicating the absence of localised polarisation at 112-type twin boundaries This suppression of Cs atomic displacements at the 112 twin boundary occurs together with the suppression of octahedral tilts at the boundary, as evidenced by measurements of the “ellipticity” of the intensity maxima associated with the Pb/I columns Together, these results suggest that, while the 112 twin boundaries preserve the centrosymmetry of the Pb-I framework, they reduce the antipolar displacements of Cs and the octahedral tilt to accommodate the structural reflection between adjacent twin domains Implications for Perovskite Functionality The distinction in polar behaviour between the 110 and 112 twin boundaries originate from their differing atomic structures The polar ferroelastic domain walls can locally break inversion symmetry and host polar distortions, despite the centrosymmetric, non-polar nature of the surrounding interior matrix This work provides the first direct experimental evidence of such polar domain walls in a photoactive halide perovskite, demonstrating they are not exclusive to oxide perovskites By revealing analogous functionality in the halide perovskite of CsPbI3, a key semiconductor for solar cells and light-emitting devices, our work translates oxide ferroic insight into the realm of halide perovskites Our findings not only reveal the existence of ferroelectric phenomena in halide perovskites but also establish twin boundaries as a promising platform for domainwall engineering to improve photophysical properties and in the design of novel perovskite optoelectronic devices The polarization vectors are consistently oriented parallel to the twin boundary, a characteristic shared with the polar domain walls in the oxide perovskite systems Structural Origin of Polarization

Paper summary: Kai: So, to wrap up what we just heard, this paper introduces the discovery that vapor-deposited CsPbI3 thin films contain ferroelastic twin boundaries parallel to the one thousand and ten and one thousand and twelve planes. The central thesis is that these boundaries host localized ferroelectric-like polarization driven by Pb atoms, even though the parent CsPbI3 material is centrosymmetric and non-polar.

Mira: And what makes it matter is that this reveals a previously hidden ferroic functionality in halide perovskite semiconductors through domain wall engineering. They found these polar twin walls act as nanoscale functional interfaces spaced about thirty to fifty nanometers apart within each grain.

Lev: From my perspective, the key finding is the distinction between the two boundary types: the one parallel to one hundred ten shows localized charge polarization, while the one parallel to one hundred twelve exhibits structural relaxation instead of net charge polarization <ref:2601.12892#pg1>.

Kai: That distinction is crucial because it shows that different twin boundaries have fundamentally different atomic structures at the nanoscale, leading to different physical behaviors in terms of how they interact with light or electricity.

Mira: They also quantified this by showing that at the one thousand and ten boundary, there are local polar distortions generating thin charge-polarised domains just two to four unit cells wide repeating at about thirty to fifty nanometers intervals inside each grain.

Lev: If we take that quantification seriously, it means we are dealing with highly localized polarization effects, which is a different kind of physics than what you might expect from a bulk ferroelectric material.

Kai: And they did confirm the lack of spontaneous polarization in the bulk CsPbI3 regions away from these boundaries. They used scanning transmission electron microscopy to confirm that no detectable Pb off-centre displacements were found anywhere in the single-crystal regions away from those twin boundaries, which rules out simple bulk ferroelectricity.

Mira: So what this means for the broader field is that understanding of ferroic behavior needs to extend beyond oxide perovskites because we are now seeing analogous functionality in CsPbI3.

Lev: And it sets up a clear direction for future research: if we want to leverage this, the focus shifts toward controlling domain wall engineering as a design principle for these materials.

Kai: It’s about using the structural flexibility of the lattice to deliberately create these interfaces that possess specific, localized polar distortions. This is what they mean by realizing nanoscale functional elements in perovskite semiconductors.

Conclusion: Mira: So, looking at the title "Discovery of Ferroelectric Twin Boundaries in a Photoactive Halide Perovskite" and the work by Li et al., it really highlights how we need to broaden our search for ferroic properties in halide perovskites.

Kai: They're proving that these materials aren't inherently non-polar, but that their structural features can be engineered to introduce polar distortions, which is a key insight for future device design.

Lev: The implication for the engineering side is that we now have a specific structural feature—the twin boundaries—that we can target to manipulate photophysical properties by controlling domain wall structure.

Kai: It moves the discussion from simply asking if these materials are ferroelectric to figuring out exactly how and where they can be engineered to achieve desired electronic outcomes.

Mira: In simple terms, this paper shows that even in a centrosymmetric crystal, you can have hidden polar behavior localized at specific structural interfaces.

Lev: That is the core idea: targeting those twin boundaries gives us a tangible platform for manipulating the material's response in ways that aren't possible with bulk properties alone.

School of Physics and Astronomy, Monash University · Department of Physics, University of Oxford · Monash Centre for Electron Microscopy, Monash University · Department of Materials Science and Engineering, Monash University

cond-mat.mtrl-sci

Submitted: 2026-01-19

Updated: 2026-01-19

DOI: 10.1073/pnas.2610011123

License: http://creativecommons.org/licenses/by-nc-nd/4.0/

Importance score: 99/100

The gist: Discovery of ferroelectric twin boundaries in a photoactive halide perovskite is significant because it reveals hidden ferroic functionality in halide perovskite semiconductors through domain wall…

Key concepts

Ferroelastic Twin Boundaries
These are structural interfaces within a material where the crystal lattice forms twin domains. In this study, specific twin boundaries (110 and 112 planes) in CsPbI3 were shown to host local structural distortions that act like ferroelectric polarization, even though the bulk material is non-polar.
Localised Polarization
This refers to a specific type of electrical ordering occurring only at the twin boundaries. At the 110 boundary, Pb atoms shift significantly along a [110] direction, creating a local electric dipole moment. This polarization is confined to a narrow region near the boundary and is absent in the surrounding bulk material.
Domain Wall Engineering
This is the strategy of intentionally controlling and manipulating these twin boundaries. By engineering these interfaces, researchers can create nanoscale functional regions that influence how light interacts with the perovskite, potentially enhancing its conductivity or improving its performance in solar cell applications.
Centrosymmetric Material
CsPbI3 is a material that lacks spontaneous net electrical polarization in its bulk structure. However, the twin boundaries break this symmetry locally. This means that while the overall crystal is centrosymmetric, the interfaces themselves can host polar distortions, which is key to understanding their unique functionality.

Terminology

Summary

Discovery of ferroelectric twin boundaries in a photoactive halide perovskite is significant because it reveals hidden ferroic functionality in halide perovskite semiconductors through domain wall engineering

The paper discovers the existence of ferroelastic twin boundaries in vapor-deposited CsPbI3 thin films, parallel to the 110 and 112 planes Remarkably, despite photoactive CsPbI3 being centrosymmetric and non-polar, Pb atoms shift at 110 twin boundaries driving a local ferroelectric-like polarisation These polar twin walls form an intrinsic array of nanoscale functional interfaces, spaced ~30-50 nm apart, embedded within the non-polar perovskite lattice In contrast, 112 twin boundaries remain non-polar but strongly suppress octahedral tilt and offcentre Cs atom displacements, revealing a different untapped ferroic degree of freedom These discoveries together uncover previously hidden ferroic functionality in halide perovskite semiconductors, opening opportunities for enhanced conductivity and photovoltaic behaviour through domain wall engineering

Non-polar Bulk Characterization

The study began by determining the crystal structure, symmetry and polarity of bulk CsPbI3 Using annular dark field scanning transmission electron microscopy (STEM-ADF), they confirmed that CsPbI3 adopts an orthorhombic perovskite structure, commonly referred to as γ-CsPbI3 Quantitative analysis found no detectable Pb off-centre displacements anywhere in the single-crystal regions away from domain walls, confirming there is no spontaneous polarisation in the bulk Consistent with the γ-CsPbI3 phase, Cs shows anti-polar off-centre displacements, that is, Cs atoms on adjacent A-sites shift in opposite directions along the [001] direction These antipolar displacements arise from octahedral tilts rather than spontaneous polarisations

Characterization of 110 Twin Boundaries

The investigation extended to twin domains and twin boundaries, revealing that the crystallography of the 110 twin boundary in CsPbI3 is illustrated schematically Localised Polarisation at 110 Twin Boundaries showed significant off-centre displacements of the Pb/I atomic columns localised at the twin boundary, with the displacement vector predominantly parallel to [110], along the boundary plane These displacements are confined to a narrow region approximately four atomic layers wide, indicating that the boundary itself hosts a distinct structural configuration which is polar The peak magnitude of the displacement reaches ~ 15 pm at the twin boundary, clearly distinguishing it from the surrounding bulk domains, where displacements remain negligible

Characterization of 112 Twin Boundaries

The study also identified a second type of twin boundary in CsPbI3 films, orientated parallel to the 112 plane Interestingly, no measurable displacement of the Pb/I columns relative to the centres of the surrounding octahedra is observed at 112-type twin boundaries, indicating the absence of localised polarisation at 112-type twin boundaries This suppression of Cs atomic displacements at the 112 twin boundary occurs together with the suppression of octahedral tilts at the boundary, as evidenced by measurements of the “ellipticity” of the intensity maxima associated with the Pb/I columns Together, these results suggest that, while the 112 twin boundaries preserve the centrosymmetry of the Pb-I framework, they reduce the antipolar displacements of Cs and the octahedral tilt to accommodate the structural reflection between adjacent twin domains

Implications for Perovskite Functionality

The distinction in polar behaviour between the 110 and 112 twin boundaries originate from their differing atomic structures The polar ferroelastic domain walls can locally break inversion symmetry and host polar distortions, despite the centrosymmetric, non-polar nature of the surrounding interior matrix This work provides the first direct experimental evidence of such polar domain walls in a photoactive halide perovskite, demonstrating they are not exclusive to oxide perovskites By revealing analogous functionality in the halide perovskite of CsPbI3, a key semiconductor for solar cells and light-emitting devices, our work translates oxide ferroic insight into the realm of halide perovskites Our findings not only reveal the existence of ferroelectric phenomena in halide perovskites but also establish twin boundaries as a promising platform for domainwall engineering to improve photophysical properties and in the design of novel perovskite optoelectronic devices The polarization vectors are consistently oriented parallel to the twin boundary, a characteristic shared with the polar domain walls in the oxide perovskite systems

Structural Origin of Polarization

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Precision structural characterization of halide perovskites can be enhanced by employing low-dose atomic-resolution transmission electron microscopy to directly observe atomic positions and local polar distortions at twin boundaries, as demonstrated in Fig 3C. This capability allows researchers to determine net charge polarisation necessary for ferroelectric phenomena at the nanoscale, moving beyond interpretations based solely on macroscopic symmetry.

  2. Development of domain wall engineering strategies can be advanced by leveraging the discovery that these polar twin walls form an intrinsic array of nanoscale functional interfaces, which can be used to enhance conductivity and photovoltaic behaviour in perovskite semiconductors. Improved AI systems will be able to predict how the specific crystallographic orientation of twin boundaries, such as the interchange between [100]/[010] for the 110-type twins, dictates the resulting local polarization.

  3. Predictive modeling of ferroelectric switching mechanisms can be improved by incorporating structural constraints derived from atomic-resolution measurements, specifically understanding that the distinction in polar behaviour between the 110 and 112 twin boundaries originate from their differing atomic structures. This allows AI to predict which twin type might serve as an initial site for ferroelectric switching by locally reducing the energy barrier for polarisation, based on structural features like octahedral tilts.

Abstract

Halide perovskites have emerged as promising materials for next-generation photovoltaics, laser sources and X-ray detectors. There is intense debate as to whether some photoactive halide perovskites exhibit ferroelectric behaviour and whether it might be possible to utilise the bulk photovoltaic effect to enhance the performance of halide perovskite solar cells. Here, using low-dose scanning transmission electron microscopy, we discover the existence of ferroelastic twin boundaries in vapor-deposited CsPbI3 thin films, parallel to 110 and 112. Remarkably, despite photoactive CsPbI3 being centrosymmetric and non-polar, we observe directly that Pb atoms shift at 110 twin boundaries driving a local ferroelectric-like polarisation. These polar twin walls form an intrinsic array of nanoscale functional interfaces, spaced 30-50 nm apart, embedded within the non-polar perovskite lattice. In contrast, 112 twin boundaries remain non-polar but strongly suppress octahedral tilt and off-centre Cs atom displacements, revealing a different untapped ferroic degree of freedom. These discoveries together uncover previously hidden ferroic functionality in halide perovskite semiconductors, opening opportunities for enhanced conductivity and photovoltaic behaviour through domain wall engineering.

Related papers