🔬 materials science

Mapping the influence of symmetry breaking in structure-property relationships of ABO $_3$ perovskites

This paper addresses the computational challenges in modeling distorted perovskite structures by developing an efficient framework using symmetry-constrained and unconstrained supercell templates to systematically explore composition-dependent distortions and their impact on the stability and physical properties of ABO $_3$ perovskites.

Original authors: Panupol Untarabut, Sylvian Cadars, Fabien Pascale, Sébastien Lebègue, Olivier Masson, Samuel Bernard, Assil Bouzid, Santanu Saha

Published 2026-07-31

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Panupol Untarabut, Sylvian Cadars, Fabien Pascale, Sébastien Lebègue, Olivier Masson, Samuel Bernard, Assil Bouzid, Santanu Saha

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world built from tiny, invisible LEGO bricks. In the realm of materials science, these bricks are atoms, and when they snap together in a specific 3D grid, they form materials called perovskites. These aren't just any bricks; they are the superstars of the energy world, capable of turning sunlight into electricity or helping batteries store power for our phones and cars. For a long time, scientists thought these materials were like perfect, rigid cubes—neat, symmetrical, and unchanging. But in reality, these atomic structures are more like wobbly Jell-O. They love to twist, tilt, and squish themselves into different shapes depending on what ingredients (chemical elements) are mixed in and how hot or cold they are.

The big question has always been: How do these tiny wobbles change the material's superpowers? If you twist the atomic grid just a little bit, does the material become a better conductor? If you squish it differently, does it get stronger? Understanding this "structure-property relationship" is like knowing exactly how to fold a paper airplane to make it fly the farthest. Without this knowledge, we are just guessing which materials will work best for our future energy needs. This is where the story of this research begins: trying to map out the endless ways these atomic LEGO sets can bend and break without falling apart.

The Great Atomic Tilt-Off

Think of perovskite materials as a massive, 3D dance floor made of octahedrons (think of them as 8-sided dice) holding hands. In a perfect, high-symmetry "cubic" phase, everyone is standing perfectly still in a neat grid. But nature hates perfection. Depending on the size of the dancers (the atoms), the floor starts to tilt, rotate, and shift. Some dancers lean left, some lean right, and some even wiggle out of their spots.

In this study, a team of scientists acted like choreographers for this atomic dance. They wanted to figure out which dance moves (distortions) make the material the most stable and which ones give it the best energy-saving properties. To do this, they didn't just guess; they built a massive digital playground. They simulated 15 different "recipes" of perovskites, mixing and matching different atoms for the A and B spots in the formula $ABO_3$ .

The Rules of the Dance Floor

The researchers discovered that the "dance style" depends heavily on a simple number called the Goldschmidt tolerance factor ( $t_G$ ). You can think of $t_G$ as a measure of how well the dancers fit into their assigned spots.

The Tight Squeeze ( $t_G < 1.00$ ): When the dancers are a bit too small for the space, they have to huddle together. This forces the whole structure to tilt and rotate wildly to fill the gaps. The study found that for these "tight squeeze" materials, the most stable and energetic dance move is a specific kind of tilt called the $a^+b^-b^-$ pattern (which translates to an orthorhombic shape). It's like the dancers all agreeing to lean in a coordinated, zig-zag pattern to stay balanced.
The Loose Fit ( $t_G > 1.00$ ): When the dancers are too big for the space, they don't need to tilt as much. Instead, they prefer to wiggle individually. In these cases, the material stays mostly cubic (square), but the central atom (the B-site) shifts slightly off-center, like a dancer stepping out of line to grab a prop.

The Simulation Game

To find these answers, the team played a high-stakes game of "structural exploration." They used two main strategies:

The Rulebook Approach: They tried out 15 specific, pre-defined dance moves (called Glazer templates) that scientists had theorized years ago.
The Chaos Approach: They built giant supercells (huge blocks of the material) and randomly shook the atoms, letting them settle wherever they wanted. This was like throwing a bunch of marbles into a box and seeing how they naturally stack up.

The results were fascinating. They found that for the "tight squeeze" materials, the chaotic shaking almost always led the atoms to settle into that same coordinated zig-zag tilt pattern. However, the size and shape of the box (the supercell) mattered a lot. If the box had an "odd" number of layers in one direction, the atoms got confused and couldn't form the perfect tilt, resulting in a messier, higher-energy state. But if the box had an "even" number of layers, the atoms could easily find their perfect, low-energy dance.

What This Means for the Future

The most exciting part of the paper is that the team didn't just map the problem; they solved it with a new recipe. They realized that you don't need to try millions of random combinations to find the best structure. Instead, they proposed a simple, smart workflow:

Look at the ingredients (calculate the $t_G$ ).
Predict how much the atoms should tilt based on that number.
Build a specific-sized box (a $\sqrt{2} \times \sqrt{2} \times 2$ supercell) that is perfectly sized to hold that tilt.
Let the computer relax the structure.

This "smart recipe" consistently found the lowest-energy, most stable structures for all the materials they tested, beating out the random guessing methods.

The Bigger Picture

The study also looked at how these wobbles affect the material's actual powers, like how much energy it takes to squeeze it (bulk modulus) and how it handles electricity (band gap). They found that the more the structure tilts and distorts, the more these properties change. For example, the "tight squeeze" materials with big tilts had different electrical gaps compared to the "loose fit" ones.

Crucially, the paper also checked what happens when things get hot. In the simulations, heating up the materials made the "wobbly" structures want to straighten out and become more symmetrical again. For some materials, this means they might switch from a wobbly, low-symmetry shape to a perfect cube as the temperature rises.

The Takeaway

This research suggests that we don't need to be overwhelmed by the complexity of these materials. By understanding the simple rules of how atomic size dictates atomic dance moves, we can predict exactly how a perovskite will behave. The team has handed us a reliable map and a compass for navigating the complex world of low-symmetry perovskites. This means scientists can now design better materials for solar panels and batteries much faster, skipping the endless trial-and-error and going straight to the winning dance moves. While these findings are currently based on computer simulations, they provide a robust framework that could soon guide real-world experiments, helping us build a more efficient energy future.

Technical Summary: Mapping the Influence of Symmetry Breaking in Structure-Property Relationships of ABO₃ Perovskites

Problem Statement
Perovskite oxides (ABO₃) are critical for energy applications due to their compositional flexibility, which allows for the stabilization of both high-symmetry (cubic) and low-symmetry phases. While low-symmetry phases, characterized by octahedral tilting and B-site displacements, dominate under ambient conditions, computational studies have historically focused on the ideal cubic phase. Modeling distortions in low-symmetry phases remains challenging due to the absence of a quantitative framework capable of capturing composition-dependent distortions arising from multiple non-equivalent atomic displacement modes. Existing approaches, such as the soft phonon method and large unconstrained supercell relaxations, are computationally expensive. Furthermore, while symmetry-broken structural templates (e.g., Glazer templates) exist, their wide-scale adaptation is hindered by a lack of systematic validation across diverse compositions and an unclear understanding of how distortion patterns correlate with thermodynamic stability and physical properties.

Methodology
The authors developed a computational framework to systematically explore the structural landscape of 15 representative ABO₃ perovskite compositions (A = Ca, Sr, Ba; B = Ti, Zr, Hf, Sn, Nb). The study employed Density Functional Theory (DFT) using the PBE functional and ONCV pseudopotentials. The methodology involved three primary categories of structural templates:

Symmetry-Constrained Templates:
- Glazer Templates: 15 unique tilt systems representing combinations of in-phase (IR) and out-of-phase (OR) octahedral rotations along crystallographic axes.
- Standard Templates: Five reference phases (cubic, tetragonal-I/II, orthorhombic, hexagonal) representing well-known phases and specific distortions like B-site displacement.
Unconstrained Templates:
- Supercell Templates: Large supercells (varying in size, shape, and parity—e.g., $2\times2\times2$ , $3\times2\times2$ , $\sqrt{2}\times\sqrt{2}\times1$ based) with random atomic perturbations to sample the potential energy surface without symmetry constraints.
- Distorted Templates: Specific models isolating single distortion types (IR, OR, or B-site displacement) to quantify their individual contributions.
Analysis and Validation:
- Thermodynamic ranking was performed at 0 K via relative energy ( $\Delta E$ ) calculations.
- Temperature-dependent free energy ( $\Delta G(T)$ ) was computed using the quasi-harmonic approximation (phonons) for selected compositions (CaZrO₃, SrTiO₃, BaTiO₃).
- Structural descriptors were defined to quantify inter-octahedral tilting ( $\langle \sin(\theta_{tilt}) \rangle$ ) and intra-octahedral distortions (B-site off-centering $\langle \Delta r_B \rangle$ and order parameter $\langle q \rangle$ ).
- Physical properties (HOMO-LUMO gap $E_g$ and bulk modulus $B_0$ ) were calculated for various templates.

Key Results

Thermodynamic Stability and $t_G$ : A strong quantitative correlation exists between the Goldschmidt tolerance factor ( $t_G$ $t_{G}$ ) and the stabilization energy of distorted phases.
- Low $t_G$ ( $<1.00$ ): Systems like CaZrO₃ exhibit large stabilization energies (up to $\sim$ 200 meV/atom) relative to the cubic phase. The ground state is consistently the orthorhombic $Pnma$ phase (Glazer $a^+b^-b^-$ ), driven primarily by collective octahedral rotations.
- High $t_G$ ( $\ge 1.00$ ): Systems like SrTiO₃ and BaTiO₃ show minimal energy differences ( $<15$ meV/atom) between phases. For BaTiO₃, B-site off-centering becomes the dominant stabilization mechanism, while SrTiO₃ remains close to the cubic limit with weak rotational instabilities.
Role of Supercell Parity: The choice of supercell size and parity significantly impacts the ability to capture low-energy states.
- Supercells with even parity (e.g., $2\times2\times2$ , $\sqrt{2}\times\sqrt{2}\times2$ ) effectively sample coherent anti-phase tilt patterns (like $Pnma$) and access the global minimum.
- Supercells with odd parity (e.g., $3\times2\times2$ ) frustrate ideal out-of-phase rotations, leading to mixed tilt patterns ( $\pm$ ) and higher energies, failing to reach the true ground state in low- $t_G$ systems.
Temperature Effects:
- CaZrO₃: The orthorhombic $Pnma$ phase remains the ground state up to 2000 K; no phase transition is predicted within the harmonic approximation.
- SrTiO₃: The free energy landscape is extremely flat. The cubic phase is stable at low temperatures in this calculation (contrary to some experimental reports of a transition at 105 K), with low-symmetry phases remaining nearly degenerate.
- BaTiO₃: A tetragonal-to-cubic transition is predicted around 880 K (harmonic approximation), driven by vibrational entropy, though this temperature is higher than the experimental value (398 K), indicating the need for anharmonic corrections.
Structure-Property Relationships:
- Band Gap ( $E_g$ ): Generally increases with symmetry lowering. The difference between cubic and orthorhombic gaps narrows as $t_G$ increases. B-site displacement (e.g., in $P4mm$) significantly alters $E_g$ in Ti/Sn-based perovskites.
- Bulk Modulus ( $B_0$ ): The orthorhombic phase generally exhibits the highest $B_0$ , while the cubic phase has the lowest. Distortions involving B-site displacement tend to reduce $B_0$ .

Proposed Optimization Protocol
Based on the findings, the authors propose a robust, computationally inexpensive protocol for rapid structural exploration:

Calculate $t_G$ to estimate the initial lattice constant.
Construct a $2\times2\times2$ supercell of the ideal cubic phase.
Apply a target rotation pattern (specifically $+--$ or $a^+b^-b^-$ ) with an initial tilt angle derived from a linear model based on $t_G$ .
Introduce small random displacements to A and B sites (0.2 Å) to seed B-site off-centering.
Perform full structural relaxation.
This workflow consistently identifies the lowest-energy phases across the 15 compositions, outperforming random sampling and standard Glazer templates in efficiency and reliability, particularly for low- $t_G$ systems.

Significance and Claims
The paper claims to provide a unified computational framework that bridges the gap between idealized high-symmetry models and the complex reality of distorted low-symmetry perovskites. The authors assert that:

Systematic Understanding: The study clarifies the specific roles of inter-octahedral tilting versus intra-octahedral displacements, showing a shift from rotation-driven stabilization (low $t_G$ ) to displacement-driven stabilization (high $t_G$ ).
Methodological Advancement: The proposed protocol offers a practical, high-throughput solution for generating physically relevant low-symmetry structures, overcoming the computational bottlenecks of random supercell sampling and the limitations of existing symmetry-constrained templates.
Database Correction: By enabling the efficient generation of low-symmetry models, this work aims to alleviate the current bias in open-source materials databases toward ideal cubic phases, thereby enriching the pool of energetically relevant perovskite models for future screening and design.
Property Prediction: The work demonstrates that distortions governing thermodynamic stability also directly control electronic and mechanical properties, suggesting that accurate modeling of symmetry breaking is essential for predicting material performance in real-world applications.

Technical Summary: Mapping the Influence of Symmetry Breaking in Structure-Property Relationships of ABO₃ Perovskites

More like this