Revealing Polar Walls in CsPbBr3: Herringbone Structure and Formation Pathway

arXiv:2609.30486 · cond-mat.mtrl-sci · Submitted 2026-09-24 · 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: "Revealing Polar Walls in CsPbBr3".

Mira: Here's the gist of the paper. Domain walls in ferroic materials can host nanoscale functionalities absent from the periodic crystal, offering new opportunities to control electronic transport and polarization in semiconductor devices<ref:2609.30486#pg2> Metal halide perovskites demonstrate important photophysical properties; however, their ferroic properties remain unclear<ref:2609.30486#pg2> Here we discover a hierarchical herringbone network of polar walls within the non-polar halide perovskite CsPbBr3 comprising ferroelastic twin walls and antiphase walls<ref:2609.30486#pg2> Atomicscale imaging reveals that both wall types comprise nanoscale regions whose inversion-symmetry is broken compared to the surrounding bulk lattice<ref:2609.30486#pg2> In-situ heating experiments show that these interfaces form sequentially during symmetry-lowering phase transitions, revealing a different formation pathway compared with ‘conventional’ oxide perovskites<ref:2609.30486#pg2> These polar nano-walls, together with their enclosed topology, may impact carrier transport<ref:2609.30486#pg2> This reveals a previously unrecognized structural motif in CsPbBr3, providing new insights into their photophysical and optoelectronic behavior<ref:2609.30486#pg2> Discovery of the Hierarchical Structure The study reveals that polycrystalline CsPbBr3 thin films host a hierarchical herringbone network of ferroelastic twin walls (TWs) and antiphase walls (APWs)<ref:2609.30486#pg2> This structure is composed of three distinct domain walls: “the three distinct domain walls: 112 TWs and 002 APW(I) and 110 APW(II) where the Cs displacement field is locally suppressed and switched across the wall”<ref:2609.30486#pg2> The primary ferroelastic twin walls correspond to “the primary domain walls correspond to 112 ferroelastic twin planes, representing a 90° (measured as 88.2°) twin relationship”<ref:2609.30486#pg2> In contrast, the secondary walls lie along “the secondary walls lie approximately along 110 or 002 planes and do not change the crystal orientation and strain state across the interface”<ref:2609.30486#pg2> Atomic Structure Determination Atomic-resolution imaging confirms that both the TWs and APWs host pronounced local structural changes<ref:2609.30486#pg2> In orthorhombic CsPbBr3, “antipolar Cs displacements (R5+, following the notation of Miller and Love) coupled with octahedral tilts (R4+ and M3+) stabilize the centrosymmetric lattice and suppress bulk ferroelectricity”<ref:2609.30486#pg2> At the walls, “the Cs displacement progressively decreases toward the interface, accompanied by relaxation of the octahedral tilts”<ref:2609.30486#pg2> This structural reconstruction is dominated by Cs rather than Pb displacements relative to the centrosymmetric positions<ref:2609.30486#pg2> The antipolar atomic displacement gradient breaks the inversion symmetry constraint imposed by the bulk lattice, generating a locally polar structure that is otherwise forbidden in the centrosymmetric crystal<ref:2609.30486#pg2> Formation Pathway Analysis In-situ heating experiments show that these walls form sequentially during symmetry-lowering phase transitions<ref:2609.30486#pg2> Specifically, “TWs vanish and reappear during the cubic-tetragonal transition, whereas APWs emerge only during the tetragonal-orthorhombic transition”<ref:2609.30486#pg2> This sequential evolution suggests that “the TWs form first as the lattice lowers from cubic to tetragonal symmetry, while APWs subsequently develop within the twin domains when the orthorhombic structure appears”<ref:2609.30486#pg2> The density, orientation and connectivity of the wall network should be tunable through composition, growth conditions, strain and thermal annealing<ref:2609.30486#pg2> Implications for Optoelectronics All three types of domain wall have the potential to host a localized electric field due to the local symmetry breaking and incomplete cancellation of antipolar displacements at domain walls<ref:2609.30486#pg2> This mechanism of incomplete cancellation of antipolar A-site displacements has been used to induce ferroelectricity in engineered oxide superlattices<ref:2609.30486#pg2> Because the wall network is governed by phase-transition symmetry and kinetics, its density, orientation and connectivity should be tunable providing opportunities to engineer domain walls as a new structural element in halide perovskite optoelectronic and photonic devices<ref:2609.30486#pg2> CsPbBr3 samples were fabricated directly on TEM grids through a “previously reported vapor co-deposition method”<ref:2609.30486#pg2> Transmission electron microscopy (TEM) was used to examine the microstructure<ref:2609.30486#pg2> Quantitative analyses of the location of intensity maxima in STEM-ADF images were performed using an open-source Python package Atomap<ref:2609.30486#pg2> In-situ heating experiments were conducted using a Tecnai F20 FEG-TEM equipped with a Gatan furnace holder<ref:2609.30486#pg2> The electron dose was minimized by using dose-fractionation<ref:2609.30486#pg2> In summary, we reveal the presence of a hierarchical herringbone domain structure with polar domain walls in polycrystalline CsPbBr3 thin films and determine the atomic structure and formation pathway of these domain walls<ref:2609.30486#pg2> Low-dose TEM reveals a network of “112 ferroelastic twin walls interwoven with 110 and 002 antiphase walls”<ref:2609.30486#pg2> In situ heating shows that these walls form sequentially during symmetry-lowering phase transitions<ref:2609.30486#pg2> Furthermore, all three types of domain wall have the potential to host a localized electric field due to the local symmetry breaking and incomplete cancellation of antipolar displacements at domain walls<ref:2609.30486#pg2> How the training signal is built The structure and stability of domain walls are expected to evolve with changes in crystal symmetry<ref:2609.30486#pg2> This evolution has rarely been observed directly in halide perovskites, limiting our understanding of how domain-wall microstructure forms and how their formation can be engineered<ref:2609.30486#pg2> The coupling between A-site displacement and octahedral tilting governs the atomic reconstruction of the walls<ref:2609.30486#pg2> The displacement gradient produces a residual Cs displacement of up to 6.5 pm between each pair of antipolar sites<ref:2609.30486#pg2> This displacement can be converted into a projected polarization Ps with the equation Ps = 1/V Σ δiZi, where V is the volume of unit cell, δi is the displacement, and Zi is the Born effective charge of atom i<ref:2609.30486#pg2> Where the gains come from The topological connectivity of this array may constrain transport pathways in optoelectronic devices such as perovskite light-emitting diodes<ref:2609.30486#pg2> The wall network is governed by phase-transition symmetry and kinetics, its density, orientation and connectivity should be tunable through composition, growth conditions, strain and thermal annealing<ref:2609.30486#pg2> This provides opportunities to engineer domain walls as a new structural element in halide perovskite optoelectronic and photonic devices<ref:2609.30486#pg2> The authors acknowledge use of facilities within the Monash Centre for Electron Microscopy<ref:2609.30486#pg2> Pruning redundant reasoning tokens The domain walls in CsPbBr3 host a hierarchical herringbone network of ferroelastic twin walls (TWs) and antiphase walls (APWs)<ref:2609.30486#pg2> This structure has a fundamentally different formation pathway compared to conventional oxide perovskites<ref:2609.30486#pg2> Unlike the herringbone pattern in ferroelectric oxide perovskites, such as BaTiO3 and PbTiO3, this network forms in a centrosymmetric halide perovskite where bulk ferroelectric polarization cannot exist<ref:2609.30486#pg2> The microstructure therefore arises not from conventional 90° and 180° ferroelectric domains, but from coupled ferroelastic and antiphase domain formation<ref:2609.30486#pg2> Pruning redundant reasoning tokens The lattice parameter map in the x-direction measured by geometric phase analysis shows a color scale ranging from-5% to 5% (left to right)<ref:2609.30486#pg2> The swap in lattice parameter across the ferroelastic twin walls is illustrated by the measurement of lattice spacing by geometric phase analysis<ref:2609.30486#pg2> This swap in lattice parameter across the ferroelastic twin walls is illustrated by the measurement of lattice spacing by geometric phase analysis<ref:2609.30486#pg2> Pruning redundant reasoning tokens The crystal symmetry of MHPs is strongly temperature dependent<ref:2609.30486#pg2> Bulk CsPbBr3 adopts an orthorhombic Pbnm phase at room temperature, transforms to a tetragonal P4/mbm phase at 88°C, and becomes cubic Pm3̅m phase above 130°C<ref:2609.30486#pg2> This observation indicates that the ferroelastic and antiphase domains originate from symmetry breaking associated with the lower-symmetry phases<ref:2609.30486#pg2> Pruning redundant reasoning tokens The authors acknowledge use of facilities within the Monash Centre for Electron Microscopy, a node of Microscopy Australia<ref:2609.30486#pg2> The Thermo Fisher Scientific Spectra φ TEM was funded by ARC LE170100118 and the FEI Titan3 80-300 FEG-TEM was funded by ARC LE0454166<ref:2609.30486#pg2> Additional TEM characterization of ferroelastic-domain scaling; simulated diffraction patterns; geometric-phase analysis; atomic-resolution characterization of twin and antiphase walls; STEM-ADF mistilt simulations; structural models; correlated Cs-displacement and octahedral-tilt analysis; and additional in situ heating data (PDF)<ref:2609.30486#pg2> The material is available free of charge via the Internet at http://pubs.acs.org<ref:2609.30486#pg2> Additional TEM characterization of ferroelastic-domain scaling; simulated diffraction patterns; geometric-phase analysis; atomic-resolution characterization of twin and antiphase walls; STEM-ADF mistilt simulations; structural models; correlated Cs-displacement and octahedral-tilt analysis; and additional in situ heating data (PDF)<ref:2609.

Paper summary: Kai: So, we’re looking at this paper, "Revealing Polar Walls in CsPbBr3: Herringbone Structure and Formation Pathway." The main idea here is that domain walls in materials with ferroic properties can host things that the perfect crystal doesn't have—specifically controlling electronic transport and polarization in semiconductor devices <ref:2609.30486#pg2>.

Mira: Right, so they’re looking at metal halide perovskites, which are interesting because they show good photophysical properties but we don't really know much about their ferroic behavior yet <ref:2609.30486#pg2>. The core claim is that they found a hierarchical herringbone network of polar walls in the non-polar CsPbBr3, made up of ferroelastic twin walls and antiphase walls <ref:2609.30486#pg2>.

Lev: That structure, having these specific domain types, could really impact carrier transport because these walls have enclosed topology <ref:2609.30486#pg2>. I’m wondering how this physical arrangement translates into something useful for actual device performance on real hardware.

Kai: Exactly, Lev. The paper claims that atomic-scale imaging shows both the twin walls and the antiphase walls have local structural changes compared to the bulk lattice <ref:2609.30486#pg2>. They found that in orthorhombic CsPbBr3, specific Cs displacements coupled with octahedral tilts stabilize a centrosymmetric lattice and suppress bulk ferroelectricity <ref:2609.30486#pg2>.

Mira: That’s the crucial part for me. The authors point out that at the walls, the Cs displacement decreases toward the interface, and you see a relaxation of those octahedral tilts <ref:2609.30486#pg2>. What’s really important is that this antipolar atomic displacement gradient breaks the inversion symmetry constraint imposed by the bulk lattice, creating a locally polar structure that shouldn't exist in a centrosymmetric crystal <ref:2609.30486#pg2>.

Lev: Breaking that symmetry at an interface could lead to localized electric fields, which is how you induce ferroelectricity in engineered oxide superlattices <ref:2609.30486#pg2>. But they also found a specific formation pathway during heating experiments that’s different from what we see in conventional oxide perovskites <ref:2609.30486#pg2>.

Kai: That formation pathway is key because it suggests how these walls actually grow, and the paper says the twin walls vanish and reappear during the cubic to tetragonal transition, while the antiphase walls only emerge during that tetragonal to orthorhombic transition <ref:2609.30486#pg2>.

Mira: So it implies that the twin walls form first when you lower from cubic symmetry, and then the antiphase walls develop within those twin domains when you get to the orthorhombic structure <ref:2609.30486#pg2>. That sequence is a big piece of information for understanding their stability across different phases.

Lev: From an engineering standpoint, if we want to control these structures, the paper notes that the density and orientation of this wall network can be tuned through composition, growth conditions, strain, and thermal annealing <ref:2609.30486#pg2>. That tunability is what makes it potentially useful for fabrication.

Kai: And when you think about those tunable properties, we’re not just talking about a structure; we’re looking at how the topological connectivity of this array might constrain transport pathways in things like perovskite light-emitting diodes <ref:2609.30486#pg2>.

Paper summary: Mira: That connectivity is what makes it relevant to optoelectronics, suggesting that engineering these domain walls could be a new structural element in those devices <ref:2609.30486#pg2>. It moves beyond just having bulk ferroelectricity and into controlling the interfaces themselves.

Lev: I think the main thing we need to consider is how you actually measure this on real hardware, since atomic-scale imaging is really difficult to achieve in a scalable way <ref:2609.30486#pg2>. We need to see if those localized electric fields are strong enough, or if the energy barrier for forming these walls is too high for practical applications.

Kai: That brings us into the second part of this paper, which focuses on the atomic structure determination using transmission electron microscopy <ref:2609.30486#pg2>. They used low-dose TEM to look at thin films, around two hundred nanometers thick, and found that the herringbone domain pattern is consistent regardless of thickness <ref:2609.30486#pg3>.

Mira: They confirmed the presence of both types of walls—the ferroelastic twin walls and the antiphase walls—in this atomic resolution image, which is a big step since direct atomic-scale evidence was previously scarce <ref:2609.30486#pg2>. The Fourier transform analysis of that image showed mirror symmetry at the wall interfaces <ref:2609.30486#pg3>.

Lev: That consistency across different film thicknesses is important because it suggests this isn't some thickness-dependent artifact, which would be a nice win for manufacturing control <ref:2609.30486#pg2>. But what about the quantitative data on the displacement gradient itself?

Kai: The paper does provide some detail on that displacement; they found that there’s a residual Cs displacement of up to six point five picometers between each pair of antipolar sites <ref:2609.30486#pg2>. They then show how you can convert this into a projected polarization, which is the equation Ps equals one over V summed over i times delta i Zi <ref:2609.30486#pg2>.

Mira: That conversion formula tells us that the polarization is directly related to how much that atomic displacement changes between those specific sites, scaled by the volume of the unit cell and some atomic charge factor <ref:2609.30486#pg2>. It’s essentially mapping structural deformation to an electrical property.

Lev: So if we take that displacement information, how does that map onto the actual operational requirements for error correction? We need to know the energy cost of these localized polar regions when you try to run quantum operations on them <ref:2609.30486#pg2>.

Kai: That’s a question for Lev. The paper hints that because this wall network is governed by symmetry and kinetics during phase transitions, we can engineer its density and connectivity to tune the resulting electric fields <ref:2609.30486#pg2>.

Mira: It suggests that manipulating the growth conditions—like strain or thermal annealing—could be a way to precisely control where these walls form and how connected they are <ref:2609.30486#pg2>. That’s the pathway to engineering them as structural elements.

Paper summary: Lev: If we could reliably control that connectivity, it would mean designing a substrate or interface that forces these walls into a specific configuration, which is exactly what we need for predictable error correction overhead <ref:2609.30486#pg2>. The limitation the authors mention is that they are studying vacuum-deposited films, and we don't know how this behavior scales up to thicker, more complex devices.

Kai: Exactly, so the paper provides a model based on thin films where we can observe these structural motifs directly, but scaling that understanding to a two hundred nanometer laser cavity or something larger is where the next challenge lies <ref:2609.30486#pg2>.

Mira: So, in summary, this work on "Revealing Polar Walls in CsPbBr3: Herringbone Structure and Formation Pathway" reveals a hierarchical network of ferroelastic twin walls and antiphase walls in CsPbBr3 <ref:2609.30486#pg2>. They discovered the atomic structure and formation pathway of these polar domain walls <ref:2609.30486#pg2>.

Lev: The main point is that these walls host a localized electric field because of the local symmetry breaking and incomplete cancellation of antipolar displacements at those boundaries <ref:2609.30486#pg2>. This mechanism, based on phase transition symmetry, gives us a way to engineer domain walls as structural elements in future halide perovskite optoelectronic and photonic devices <ref:2609.30486#pg2>.

Kai: So the title points directly to the discovery of this herringbone structure and how it forms <ref:2609.30486#pg1>. It connects the material's crystal structure evolution to nanoscale features <ref:2609.30486#pg2>.

Mira: And it matters because metal halide perovskites are important for light-emitting applications, and understanding these polar walls means we can potentially control their electronic transport in a way we couldn't before <ref:2609.30486#pg1>. It opens up new ways to design these materials for better optoelectronic performance.

Lev: The implication is that the density, orientation, and connectivity of this wall network should be tunable through composition, growth conditions, strain and thermal annealing <ref:2609.30486#pg2>. That tunability is what gives us a handle on controlling the structure for practical applications <ref:2609.30486#pg2>.

Kai: So, we’re seeing a link between how the crystal symmetry changes during phase transitions and the formation of these nanoscale polar features <ref:2609.30486#pg2>. It's a structural motif that was previously unrecognized in CsPbBr3 <ref:2609.30486#pg2>.

Mira: And it tells us that even in a non-polar bulk material, you can get these nanoscale functionalities by exploiting the interfaces and the way the structure lowers its symmetry <ref:2609.30486#pg1>. It’s about finding new physics at the boundaries of what we thought was just a simple crystal structure.

Lev: I’d say it's interesting because it suggests that instead of focusing solely on bulk properties, we should be looking at these domain wall structures as active components in our device design <ref:2609.30486#pg2>. It changes how we think about materials for quantum hardware too, where interface control is everything.

Kai: Right, so the paper shows us that there's a pathway to engineer these domain walls to control polarization and transport in semiconductor devices <ref:2609.30486#pg1>. That’s the big picture here for us on this research.

Conclusion: Kai: So, to wrap up, this paper is about finding these specific domain walls in cesium lead bromide perovskite that act like little polar regions inside a material that isn't usually thought of as ferroelectric.

Mira: Yeah, they found this hierarchical herringbone network with twin walls and antiphase walls that have these localized electric fields because of how the atoms are displaced at the boundaries.

Lev: So it’s not just about having a big bulk polarization, but how those interfaces themselves can host something new for electronics.

Kai: Right, so the title is 'Revealing Polar Walls in CsPbBr3: Herringbone Structure and Formation Pathway,' and it points to this specific structure being key.

Mira: It means we need to look at these interfaces instead of just the bulk crystal when we think about controlling electronic transport or polarization in devices.

Lev: The authors show that they can tune the density and shape of this wall network using things like how much material you use, the strain you put on it, or even how much heat you apply during growth.

Kai: That tunability is what makes it interesting for hardware, because if we can control the structure of these walls, we might be able to engineer them as a new structural element in optoelectronic devices.

Mira: It connects the way the crystal symmetry lowers during phase transitions directly to the formation of these nanoscale features, which is a pathway that's different from what happens in traditional oxide perovskites.

Lev: That phase transition dependence is important because it means we have kinetic control over how these things form, which is something you need when building real hardware for error correction.

Kai: So the paper shows this isn't just some random defect; it’s a structured network that has a specific formation sequence during heating that we can use to design better materials.

Mira: It opens up new ways to think about how we can exploit symmetry breaking at interfaces to get localized functions in materials where bulk ferroelectricity is hard to achieve.

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

cond-mat.mtrl-sci

Submitted: 2026-09-24

Updated: 2026-09-24

Comments: 11 pages, 5 figures

DOI: 10.1073/pnas.2610011123

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

Importance score: 99/100

The gist: Domain walls in ferroic materials can host nanoscale functionalities absent from the periodic crystal…

Key concepts

Ferroelastic Twin Walls (TWs)
These are domain walls where the crystal lattice undergoes a specific strain change, like a ~90° twin relationship. In CsPbBr3, these walls are part of the hierarchical network and form during symmetry-lowering phase transitions, acting as distinct structural elements in the material.
Antiphase Walls (APWs)
These walls lie along specific crystallographic planes (like 110 or 002) and do not change the crystal orientation or strain across the interface. They are also part of the hierarchical network and emerge during a different phase transition than twin walls.
Incomplete Cancellation of Antipolar Displacements
At domain walls, atomic displacements (specifically Cs displacements) do not perfectly cancel out as they would in the bulk material. This incomplete cancellation breaks inversion symmetry locally, creating a region with a net polarization that is normally forbidden in the centrosymmetric crystal.
Hierarchical Herringbone Network
This describes the complex arrangement of three types of domain walls—112 TWs, 002 APW(I), and 110 APW(II)—forming a structured pattern resembling a herringbone design. This network is governed by phase-transition symmetry and kinetics, allowing for tunable properties.

Terminology

Summary

Domain walls in ferroic materials can host nanoscale functionalities absent from the periodic crystal, offering new opportunities to control electronic transport and polarization in semiconductor devices<ref:2609. p. 0

Where the paper says it

Revealing Polar Walls in CsPbBr3: Herringbone Structure and Formation Pathway Weilun Li1 , Qimu Yuan2 , Michael B. Johnston2 , Joanne Etheridge1,3,4* 1 School of Physics and Astronomy, Monash University, Victoria, 3800, Australia 2 Department of Physics, University of Oxford, Clarendon Laboratory, Oxford OX1 3PU, United Kingdom 3 Monash Centre for Electron Microscopy, Monash University, Victoria, 3800, Australia 4 Department of Materials Science and Engineering, Monash University, Victoria, 3800, Australia Keywords: metal halide perovskites; polar domain walls; ferroelasticity; antiphase boundaries; symmetry-breaking; transmission electron microscopy Abstract Domain walls in ferroic materials can host nanoscale functionalities absent from the periodic crystal, offering new opportunities to control electronic transport and polarization in semiconductor devices.

30486#pg2> Metal halide perovskites demonstrate important photophysical properties; however, their ferroic properties remain unclear<ref:2609. p. 0
Where the paper says it

Johnston2 , Joanne Etheridge1,3,4* 1 School of Physics and Astronomy, Monash University, Victoria, 3800, Australia 2 Department of Physics, University of Oxford, Clarendon Laboratory, Oxford OX1 3PU, United Kingdom 3 Monash Centre for Electron Microscopy, Monash University, Victoria, 3800, Australia 4 Department of Materials Science and Engineering, Monash University, Victoria, 3800, Australia Keywords: metal halide perovskites; polar domain walls; ferroelasticity; antiphase boundaries; symmetry-breaking; transmission electron microscopy Abstract Domain walls in ferroic materials can host nanoscale functionalities absent from the periodic crystal, offering new opportunities to control electronic transport and polarization in semiconductor devices. Metal halide perovskites demonstrate important photophysical properties; however, their ferroic properties remain unclear.

30486#pg2> Here we discover a hierarchical herringbone network of polar walls within the non-polar halide perovskite CsPbBr3 comprising ferroelastic twin walls and antiphase walls<ref:2609. p. 0
Where the paper says it

Metal halide perovskites demonstrate important photophysical properties; however, their ferroic properties remain unclear. Here we discover a hierarchical herringbone network of polar walls within the non-polar halide perovskite CsPbBr3 comprising ferroelastic twin walls and antiphase walls.

30486#pg2> Atomicscale imaging reveals that both wall types comprise nanoscale regions whose inversion-symmetry is broken compared to the surrounding bulk lattice<ref:2609. p. 0
Where the paper says it

Here we discover a hierarchical herringbone network of polar walls within the non-polar halide perovskite CsPbBr3 comprising ferroelastic twin walls and antiphase walls. Atomicscale imaging reveals that both wall types comprise nanoscale regions whose inversion-symmetry is broken compared to the surrounding bulk lattice.

30486#pg2> In-situ heating experiments show that these interfaces form sequentially during symmetry-lowering phase transitions, revealing a different formation pathway compared with ‘conventional’ oxide perovskites<ref:2609. p. 0
Where the paper says it

Atomicscale imaging reveals that both wall types comprise nanoscale regions whose inversion-symmetry is broken compared to the surrounding bulk lattice. In-situ heating experiments show that these interfaces form sequentially during symmetry-lowering phase transitions, revealing a different formation pathway compared with ‘conventional’ oxide perovskites.

30486#pg2> These polar nano-walls, together with their enclosed topology, may impact carrier transport<ref:2609. p. 0
Where the paper says it

In-situ heating experiments show that these interfaces form sequentially during symmetry-lowering phase transitions, revealing a different formation pathway compared with ‘conventional’ oxide perovskites. These polar nano-walls, together with their enclosed topology, may impact carrier transport.

30486#pg2> This reveals a previously unrecognized structural motif in CsPbBr3, providing new insights into their photophysical and optoelectronic behavior<ref:2609. p. 0
Where the paper says it

These polar nano-walls, together with their enclosed topology, may impact carrier transport. This reveals a previously unrecognized structural motif in CsPbBr3, providing new insights into their photophysical and optoelectronic behavior.

30486#pg2>

Discovery of the Hierarchical Structure

The study reveals that polycrystalline CsPbBr3 thin films host a hierarchical herringbone network of ferroelastic twin walls (TWs) and antiphase walls (APWs)<ref:2609.30486#pg2> This structure is composed of three distinct domain walls: “the three distinct domain walls: 112 TWs and 002 APW(I) and 110 APW(II) where the Cs displacement field is locally suppressed and switched across the wall”<ref:2609. p. 2

Where the paper says it

We note that minor measurement artefacts may arise from slight crystal misorientation (Fig. These observations demonstrate that the hierarchical wall architecture in CsPbBr3 consists of three distinct domain walls: {112} TWs and {002} APW(I) and {110} APW(II) where the Cs displacement field is locally suppressed and switched across the wall.

30486#pg2> The primary ferroelastic twin walls correspond to “the primary domain walls correspond to 112 ferroelastic twin planes, representing a 90° (measured as 88. p. 2
Where the paper says it

Importantly, it also reveals a strain gradient at the ferroelastic twin wall (Fig. The primary domain walls correspond to {112} ferroelastic twin planes, representing a ~90° (measured as 88.

2°) twin relationship”<ref:2609. p. 2
Where the paper says it

Fourier transform analysis of the atomic-resolution image (Fig. 1F) shows mirrored spots across the wall, confirming a twin relationship (Fig.

30486#pg2> In contrast, the secondary walls lie along “the secondary walls lie approximately along 110 or 002 planes and do not change the crystal orientation and strain state across the interface”<ref:2609. p. 2
Where the paper says it

2°) twin relationship. In contrast, the secondary walls lie approximately along {110} or {002} planes and do not change the crystal orientation and strain state across the interface, indicating that they are structurally distinct from the ferroelastic twin walls (Fig.

30486#pg2>

Atomic Structure Determination

Atomic-resolution imaging confirms that both the TWs and APWs host pronounced local structural changes<ref:2609.30486#pg2> In orthorhombic CsPbBr3, “antipolar Cs displacements (R5+, following the notation of Miller and Love) coupled with octahedral tilts (R4+ and M3+) stabilize the centrosymmetric lattice and suppress bulk ferroelectricity”<ref:2609. p. 6

Where the paper says it

Atomic-resolution imaging further shows that both the TWs and APWs host pronounced local structural changes. In the orthorhombic Pbnm phase, antipolar Cs displacements (𝑅5 + , following the notation of Miller and Love34) coupled with octahedral tilts (𝑅4 + and 𝑀3 +) stabilize the centrosymmetric lattice and suppress bulk ferroelectricity35,36 .

30486#pg2> At the walls, “the Cs displacement progressively decreases toward the interface, accompanied by relaxation of the octahedral tilts”<ref:2609. p. 6
Where the paper says it

In the orthorhombic Pbnm phase, antipolar Cs displacements (𝑅5 + , following the notation of Miller and Love34) coupled with octahedral tilts (𝑅4 + and 𝑀3 +) stabilize the centrosymmetric lattice and suppress bulk ferroelectricity35,36 . At the walls, the Cs displacement progressively decreases toward the interface, accompanied by relaxation of the octahedral tilts.

30486#pg2> This structural reconstruction is dominated by Cs rather than Pb displacements relative to the centrosymmetric positions<ref:2609. p. 2
Where the paper says it

The gradual reduction and reversal of this displacement order parameter indicates that the walls locally modify the Pbnm symmetry, allowing Cs cations to relax toward central positions. This reconstruction is dominated by Cs rather than Pb displacements relative to the centrosymmetric positions (Fig.

30486#pg2> The antipolar atomic displacement gradient breaks the inversion symmetry constraint imposed by the bulk lattice, generating a locally polar structure that is otherwise forbidden in the centrosymmetric crystal<ref:2609. p. 6
Where the paper says it

The wall region therefore acts as a nanoscale structural transition zone in which the orthorhombic order parameters are reduced, consistent with a Landau–Ginzburg description of continuous order parameter variation across domain walls. The antipolar atomic displacement gradient breaks the inversion symmetry constraint imposed by the bulk lattice, generating a locally polar structure that is otherwise forbidden in the centrosymmetric crystal.

30486#pg2>

Formation Pathway Analysis

In-situ heating experiments show that these walls form sequentially during symmetry-lowering phase transitions<ref:2609.30486#pg2> Specifically, “TWs vanish and reappear during the cubic-tetragonal transition, whereas APWs emerge only during the tetragonal-orthorhombic transition”<ref:2609. p. 4

Where the paper says it

Close examination reveals that the two types of domain walls evolve differently during these transitions. TWs vanish and reappear during the cubic-tetragonal transition, whereas APWs emerge only during the tetragonal-orthorhombic transition (Fig.

30486#pg2> This sequential evolution suggests that “the TWs form first as the lattice lowers from cubic to tetragonal symmetry, while APWs subsequently develop within the twin domains when the orthorhombic structure appears”<ref:2609. p. 4
Where the paper says it

TWs vanish and reappear during the cubic-tetragonal transition, whereas APWs emerge only during the tetragonal-orthorhombic transition (Fig. This sequential evolution suggests that the TWs form first as the lattice lowers from cubic to tetragonal symmetry, while APWs subsequently develop within the twin domains when the orthorhombic structure appears.

30486#pg2> The density, orientation and connectivity of the wall network should be tunable through composition, growth conditions, strain and thermal annealing<ref:2609. p. 7
Where the paper says it

Furthermore, all three types of domain wall have the potential to host a localized electric field due to the local symmetry breaking and incomplete cancellation of antipolar displacements at domain walls. Because the wall network is governed by phase-transition symmetry and kinetics, its density, orientation and connectivity should be tunable through composition, growth conditions, strain and thermal annealing, providing opportunities to engineer domain walls as a new structural element in halide perovskite optoelectronic and photonic devices.

30486#pg2>

Implications for Optoelectronics

All three types of domain wall have the potential to host a localized electric field due to the local symmetry breaking and incomplete cancellation of antipolar displacements at domain walls<ref:2609.30486#pg2> This mechanism of incomplete cancellation of antipolar A-site displacements has been used to induce ferroelectricity in engineered oxide superlattices<ref:2609. p. 6

Where the paper says it

At both TWs and APWs, neighboring Cs columns exhibit unequal displacement amplitudes across the interface, preventing complete cancellation of the antipolar A-site distortions present in the bulk lattice. This mechanism of incomplete cancellation of antipolar A-site displacements has been used to induce ferroelectricity in engineered oxide superlattices, where alternating 1ABO/1A’BO3 layers break the symmetry that enforces antipolar compensation in the parent lattice37 .

30486#pg2> Because the wall network is governed by phase-transition symmetry and kinetics, its density, orientation and connectivity should be tunable providing opportunities to engineer domain walls as a new structural element in halide perovskite optoelectronic and photonic devices<ref:2609. p. 7
Where the paper says it

Furthermore, all three types of domain wall have the potential to host a localized electric field due to the local symmetry breaking and incomplete cancellation of antipolar displacements at domain walls. Because the wall network is governed by phase-transition symmetry and kinetics, its density, orientation and connectivity should be tunable through composition, growth conditions, strain and thermal annealing, providing opportunities to engineer domain walls as a new structural element in halide perovskite optoelectronic and photonic devices.

30486#pg2>

Methods Summary

CsPbBr3 samples were fabricated directly on TEM grids through a “previously reported vapor co-deposition method”<ref:2609.30486#pg2> Transmission electron microscopy (TEM) was used to examine the microstructure<ref:2609. p. 1

Where the paper says it

Here we address these questions in vacuum-deposited CsPbBr3 thin films, an important platform for scalable and stable perovskite optoelectronics and photonics3 , using low-dose transmission electron microscopy (TEM). Results Herringbone ferroelastic domain We first examine the microstructure of polycrystalline CsPbBr3 thin films with a thickness of ~200 nm, comparable to typical optical gain media in laser cavities.

30486#pg2> Quantitative analyses of the location of intensity maxima in STEM-ADF images were performed using an open-source Python package Atomap<ref:2609. p. 7
Where the paper says it

Images were acquired at 300 kV, using an 18 mrad probeforming aperture, and 39-200 mrad detector collection angle. Quantitative analyses of the location of intensity maxima in STEM-ADF images were performed using an open-source Python package Atomap38, where both the precise position and shape of atomic columns are measured using 2D elliptical Gaussian distributions.

30486#pg2> In-situ heating experiments were conducted using a Tecnai F20 FEG-TEM equipped with a Gatan furnace holder<ref:2609. p. 7
Where the paper says it

Frames were carefully inspected to confirm that no structural changes occurred and then summed up to increase the signal-to-noise ratio. In-situ heating experiments were conducted using a Tecnai F20 FEG-TEM equipped with a Gatan furnace holder.

30486#pg2> The electron dose was minimized by using dose-fractionation<ref:2609. p. 7
Where the paper says it

Kinematic diffraction pattern simulations were performed using SingleCrystal software. To obtain STEM-ADF images suitable for quantitative measurements while preserving the structural integrity of the material, the electron dose was minimized by using dose-fractionation: 10 frames were acquired, each with fluence calibrated using a direct-counting detector and set to 1500 e/Å2 .

30486#pg2>

Conclusion Summary

In summary, we reveal the presence of a hierarchical herringbone domain structure with polar domain walls in polycrystalline CsPbBr3 thin films and determine the atomic structure and formation pathway of these domain walls<ref:2609.30486#pg2> Low-dose TEM reveals a network of “112 ferroelastic twin walls interwoven with 110 and 002 antiphase walls”<ref:2609. p. 6

Where the paper says it

Conclusions In summary, we reveal the presence of a hierarchical herringbone domain structure with polar domain walls in polycrystalline CsPbBr3 thin films and determine the atomic structure and formation pathway of these domain walls. Low-dose TEM reveals a network of {112} ferroelastic twin walls interwoven with {110} and {002} antiphase walls.

30486#pg2> In situ heating shows that these walls form sequentially during symmetry-lowering phase transitions<ref:2609. p. 6
Where the paper says it

Low-dose TEM reveals a network of {112} ferroelastic twin walls interwoven with {110} and {002} antiphase walls. In situ heating shows that these walls form sequentially during symmetry-lowering phase transitions, while atomic-resolution imaging reveals suppressed antipolar Cs displacements and relaxed octahedral tilts at the wall regions.

30486#pg2> Furthermore, all three types of domain wall have the potential to host a localized electric field due to the local symmetry breaking and incomplete cancellation of antipolar displacements at domain walls<ref:2609. p. 7
Where the paper says it

Furthermore, all three types of domain wall have the potential to host a localized electric field due to the local symmetry breaking and incomplete cancellation of antipolar displacements at domain walls. Because the wall network is governed by phase-transition symmetry and kinetics, its density, orientation and connectivity should be tunable through composition, growth conditions, strain and thermal annealing, providing opportunities to engineer domain walls as a new structural element in halide perovskite optoelectronic and photonic devices.

30486#pg2>

How the training signal is built

The structure and stability of domain walls are expected to evolve with changes in crystal symmetry<ref:2609.30486#pg2> This evolution has rarely been observed directly in halide perovskites, limiting our understanding of how domain-wall microstructure forms and how their formation can be engineered<ref:2609. p. 1

Where the paper says it

Because ferroelastic domains originate from symmetry-lowering transformations, the structure and stability of domain walls are expected to evolve with changes in crystal symmetry16,17,19,23 . However, this evolution has rarely been observed directly in halide perovskites, limiting our understanding of how domain-wall microstructure forms and how their formation can be engineered.

30486#pg2> The coupling between A-site displacement and octahedral tilting governs the atomic reconstruction of the walls<ref:2609. p. 3
Where the paper says it

Similar behavior is observed at the TW and APW(II) (Fig. This coupling between A-site displacement and octahedral tilting governs the atomic reconstruction of the walls.

30486#pg2> The displacement gradient produces a residual Cs displacement of up to 6. p. 3
Where the paper says it

This coupling between A-site displacement and octahedral tilting governs the atomic reconstruction of the walls. The displacement gradient produces a residual Cs displacement of up to 6.

5 pm between each pair of antipolar sites<ref:2609. p. 3
Where the paper says it

The displacement gradient produces a residual Cs displacement of up to 6. 5 pm between each pair of antipolar sites.

30486#pg2> This displacement can be converted into a projected polarization Ps with the equation Ps = 1/V Σ δiZi, where V is the volume of unit cell, δi is the displacement, and Zi is the Born effective charge of atom i<ref:2609. p. 3
Where the paper says it

5 pm between each pair of antipolar sites. This displacement can be converted into a projected polarization 𝑃𝑠 with the equation 𝑃𝑠 = 1 𝑉 ∑ 𝛿𝑖𝑍𝑖 , where 𝑉 is the volume of unit cell, 𝛿𝑖 is the displacement, and 𝑍𝑖 is the Born effective charge of atom 𝑖.

30486#pg2>

Where the gains come from

The topological connectivity of this array may constrain transport pathways in optoelectronic devices such as perovskite light-emitting diodes<ref:2609.30486#pg2> The wall network is governed by phase-transition symmetry and kinetics, its density, orientation and connectivity should be tunable through composition, growth conditions, strain and thermal annealing<ref:2609. p. 7

Where the paper says it

Furthermore, all three types of domain wall have the potential to host a localized electric field due to the local symmetry breaking and incomplete cancellation of antipolar displacements at domain walls. Because the wall network is governed by phase-transition symmetry and kinetics, its density, orientation and connectivity should be tunable through composition, growth conditions, strain and thermal annealing, providing opportunities to engineer domain walls as a new structural element in halide perovskite optoelectronic and photonic devices.

30486#pg2> This provides opportunities to engineer domain walls as a new structural element in halide perovskite optoelectronic and photonic devices<ref:2609. p. 7
Where the paper says it

Furthermore, all three types of domain wall have the potential to host a localized electric field due to the local symmetry breaking and incomplete cancellation of antipolar displacements at domain walls. Because the wall network is governed by phase-transition symmetry and kinetics, its density, orientation and connectivity should be tunable through composition, growth conditions, strain and thermal annealing, providing opportunities to engineer domain walls as a new structural element in halide perovskite optoelectronic and photonic devices.

30486#pg2> The authors acknowledge use of facilities within the Monash Centre for Electron Microscopy<ref:2609. p. 8
Where the paper says it

Notes The authors declare no competing interests. Acknowledgements The authors acknowledge use of facilities within the Monash Centre for Electron Microscopy, a node of Microscopy Australia.

30486#pg2>

Pruning redundant reasoning tokens

The domain walls in CsPbBr3 host a hierarchical herringbone network of ferroelastic twin walls (TWs) and antiphase walls (APWs)<ref:2609.30486#pg2> This structure has a fundamentally different formation pathway compared to conventional oxide perovskites<ref:2609. p. 5

Where the paper says it

Discussion Our results reveal that CsPbBr3 films contain a hierarchical herringbone network of ferroelastic twin walls (TWs) and antiphase walls (APWs). This herringbone structure has a fundamentally different formation pathway compared to conventional oxide perovskites.

30486#pg2> Unlike the herringbone pattern in ferroelectric oxide perovskites, such as BaTiO3 and PbTiO3, this network forms in a centrosymmetric halide perovskite where bulk ferroelectric polarization cannot exist<ref:2609. p. 5
Where the paper says it

This herringbone structure has a fundamentally different formation pathway compared to conventional oxide perovskites. Unlike the herringbone pattern in ferroelectric oxide perovskites, such as BaTiO3 29,30 and PbTiO3 31–33 , this network forms in a centrosymmetric halide perovskite where bulk ferroelectric polarization cannot exist.

30486#pg2> The microstructure therefore arises not from conventional 90° and 180° ferroelectric domains, but from coupled ferroelastic and antiphase domain formation<ref:2609. p. 5
Where the paper says it

Unlike the herringbone pattern in ferroelectric oxide perovskites, such as BaTiO3 29,30 and PbTiO3 31–33 , this network forms in a centrosymmetric halide perovskite where bulk ferroelectric polarization cannot exist. The microstructure therefore arises not from conventional 90° and 180° ferroelectric domains, but from coupled ferroelastic and antiphase domain formation.

30486#pg2>

The lattice parameter map in the x-direction measured by geometric phase analysis shows a color scale ranging from-5% to 5% (left to right)<ref:2609.30486#pg2> The swap in lattice parameter across the ferroelastic twin walls is illustrated by the measurement of lattice spacing by geometric phase analysis<ref:2609. p. 2

Where the paper says it

1F) shows mirrored spots across the wall, confirming a twin relationship (Fig. The swap in lattice parameter across the ferroelastic twin walls is illustrated by the measurement of lattice spacing by geometric phase analysis (Fig.

30486#pg2> This swap in lattice parameter across the ferroelastic twin walls is illustrated by the measurement of lattice spacing by geometric phase analysis<ref:2609. p. 2
Where the paper says it

1F) shows mirrored spots across the wall, confirming a twin relationship (Fig. The swap in lattice parameter across the ferroelastic twin walls is illustrated by the measurement of lattice spacing by geometric phase analysis (Fig.

30486#pg2>

The crystal symmetry of MHPs is strongly temperature dependent<ref:2609.30486#pg2> Bulk CsPbBr3 adopts an orthorhombic Pbnm phase at room temperature, transforms to a tetragonal P4/mbm phase at 88°C, and becomes cubic Pm3̅m phase above 130°C<ref:2609. p. 4

Where the paper says it

Evolution of domain walls across phase transitions The crystal symmetry of MHPs is strongly temperature dependent. Bulk CsPbBr3 adopts an orthorhombic Pbnm phase at room temperature, transforms to a tetragonal P4/mbm phase at ~88°C, and becomes cubic 𝑃𝑚3̅𝑚 phase above ~130°C17,19,28 .

30486#pg2> This observation indicates that the ferroelastic and antiphase domains originate from symmetry breaking associated with the lower-symmetry phases<ref:2609. p. 4
Where the paper says it

Upon heating to 150 °C, corresponding to the cubic phase, these domains disappear (Fig. This observation indicates that the ferroelastic and antiphase domains originate from symmetry breaking associated with the lower-symmetry phases.

30486#pg2>

The authors acknowledge use of facilities within the Monash Centre for Electron Microscopy, a node of Microscopy Australia<ref:2609.30486#pg2> The Thermo Fisher Scientific Spectra φ TEM was funded by ARC LE170100118 and the FEI Titan3 80-300 FEG-TEM was funded by ARC LE0454166<ref:2609. p. 8

Where the paper says it

Acknowledgements The authors acknowledge use of facilities within the Monash Centre for Electron Microscopy, a node of Microscopy Australia. The Thermo Fisher Scientific Spectra φ TEM was funded by ARC LE170100118 and the FEI Titan3 80-300 FEG-TEM was funded by ARC LE0454166.

30486#pg2> Additional TEM characterization of ferroelastic-domain scaling; simulated diffraction patterns; geometric-phase analysis; atomic-resolution characterization of twin and antiphase walls; STEM-ADF mistilt simulations; structural models; correlated Cs-displacement and octahedral-tilt analysis; and additional in situ heating data (PDF)<ref:2609. p. 7
Where the paper says it

ASSOCIATED CONTENT Supporting Information This material is available free of charge via the Internet at http://pubs. Additional TEM characterization of ferroelastic-domain scaling; simulated diffraction patterns; geometric-phase analysis; atomic-resolution characterization of twin and antiphase walls; STEM-ADF mistilt simulations; structural models; correlated Cs-displacement and octahedral-tilt analysis; and additional in situ heating data (PDF).

p. 7
Where the paper says it

ASSOCIATED CONTENT Supporting Information This material is available free of charge via the Internet at http://pubs. Additional TEM characterization of ferroelastic-domain scaling; simulated diffraction patterns; geometric-phase analysis; atomic-resolution characterization of twin and antiphase walls; STEM-ADF mistilt simulations; structural models; correlated Cs-displacement and octahedral-tilt analysis; and additional in situ heating data (PDF).

30486#pg2> The material is available free of charge via the Internet at http://pubs. p. 7
Where the paper says it

For in-situ movies, images were acquired continuously at a rate of 1 frame/s. ASSOCIATED CONTENT Supporting Information This material is available free of charge via the Internet at http://pubs.

acs.org<ref:2609.30486#pg2> Additional TEM characterization of ferroelastic-domain scaling; simulated diffraction patterns; geometric-phase analysis; atomic-resolution characterization of twin and antiphase walls; STEM-ADF mistilt simulations; structural models; correlated Cs-displacement and octahedral-tilt analysis; and additional in situ heating data (PDF)<ref:2609.

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. Automated structural defect prediction for metal halide perovskites can be enhanced by incorporating knowledge of polar domain wall formation pathways such as herringbone network of polar walls and ferroelastic twin walls. This allows AI systems to predict the resulting electronic transport and polarization properties based on the observed hierarchical architecture.

  2. Improved AI systems can perform in-situ phase transition monitoring during growth processes by analyzing structural changes related to symmetry lowering, such as how TWs vanish and reappear during the cubic-tetragonal transition or how APWs emerge only during the tetragonal-orthorhombic transition. This enables real-time feedback for optimizing film growth conditions.

  3. AI models can be trained to predict localized electric fields within domain walls by learning from atomic displacement gradients, specifically identifying how the antipolar atomic displacement gradient breaks the inversion symmetry constraint imposed by the bulk lattice and causes a residual dipole moment. This capability extends to predicting functionalities in devices like perovskite light-emitting diodes.

  4. Systems can be developed to tune domain wall properties through compositional and thermal annealing parameters by correlating these inputs with observed structural outcomes, as suggested by the paper's conclusion that the density, orientation and connectivity should be tunable through composition, growth conditions, strain and thermal annealing. This allows for targeted engineering of optoelectronic device performance.

Abstract

Domain walls in ferroic materials can host nanoscale functionalities absent from the periodic crystal, offering new opportunities to control electronic transport and polarization in semiconductor devices. Metal halide perovskites demonstrate important photophysical properties; however, their ferroic properties remain unclear. Here we discover a hierarchical herringbone network of polar walls within the non-polar halide perovskite CsPbBr3 comprising ferroelastic twin walls and antiphase walls. Atomic-scale imaging reveals that both wall types comprise nanoscale regions whose inversion-symmetry is broken compared to the surrounding bulk lattice. In-situ heating experiments show that these interfaces form sequentially during symmetry-lowering phase transitions, revealing a different formation pathway compared with 'conventional' oxide perovskites. These polar nano-walls, together with their enclosed topology, may impact carrier transport. This reveals a previously unrecognized structural motif in CsPbBr3, providing new insights into their photophysical and optoelectronic behavior.

Sources

Related papers