Anisotropic Defect Diffusion in Layered CsPbBr$_\mathrm{x}$I$_\mathrm{3-x}$ Perovskites

Large-scale molecular dynamics simulations reveal that layered ordering of bromine and iodine anions in CsPbBr$_\mathrm{x}$I$_\mathrm{3-x}$ perovskites induces strongly anisotropic defect diffusion, where migration is facile along the layers but suppressed across them due to directional lattice strain and specific bonding configurations.

The Gist

Imagine a solar cell material as a giant, 3D Lego castle built from tiny blocks. In this specific type of castle, called a "perovskite," the blocks are made of different ingredients: Cesium (Cs), Lead (Pb), and a mix of two types of "glue" atoms—Bromine (Br) and Iodine (I).

The problem is that this castle is a bit shaky. Over time, tiny pieces of the castle (called "defects") start to wander around. When these pieces move, they can break the castle's structure or ruin its ability to turn sunlight into electricity. The researchers wanted to figure out how to stop these wandering pieces from causing trouble.

Here is what they discovered, explained simply:

1. The "Layer Cake" Strategy

Usually, when you mix Bromine and Iodine, they get jumbled up like sprinkles in a cake batter. The researchers tried a different approach: they organized the sprinkles into neat, distinct layers. Imagine a cake where one layer is purely chocolate sprinkles and the next is purely vanilla, stacked perfectly on top of each other.

They found that this "layer cake" structure changes how the wandering pieces move. Instead of wandering in all directions (up, down, left, right, forward, backward), the pieces get stuck moving only sideways along the layers. They are effectively blocked from jumping up or down between the layers.

2. The "Crowded Hallway" (For Cesium)

Think of the Cesium atoms as people trying to walk through a hallway made of octagonal pillars (the Lead-halide blocks).

• In a normal, mixed castle: The pillars are slightly tilted in random directions, creating open doorways in every direction. The Cesium people can walk anywhere easily.
• In the layered castle: Because the layers are different sizes, the pillars in the "Iodine layers" get squeezed and tilted in a very specific, rigid pattern. It's like the pillars have locked their doors shut in the vertical direction. The Cesium people can still shuffle sideways along the floor, but they cannot jump to the next floor. The "gate" to move up or down is jammed shut by the strain of the layers.

3. The "Social Club" (For Halide Glue)

The Bromine and Iodine atoms that wander around (as defects) act a bit like people at a party who only want to hang out with their own kind.

• The Rule: A Bromine defect prefers to form a "double bridge" with another Bromine atom. An Iodine defect wants to pair up with another Iodine.
• The Result: In the layered castle, if a Bromine defect is in a Bromine layer, it can easily hop from neighbor to neighbor because everyone is Bromine. But if it tries to jump into an Iodine layer, it can't find a Bromine partner to hold hands with, so it gets stuck.
• The Twist: Even though the layers are squeezed (strained), the main reason these atoms stay in their own lanes is this "social preference" for their own chemical type. They stick to the layers where their "friends" are.

4. The "Vacancy" (The Empty Seat)

Sometimes, a spot in the castle is empty (a vacancy). Think of this as an empty chair in a crowded theater.

• The Physics: The "Iodine layers" are under a bit of a squeeze (compressive strain), while the "Bromine layers" are stretched out.
• The Effect: The squeeze in the Iodine layers actually makes the empty chairs (vacancies) feel more comfortable and stable there. So, if an empty seat appears, it prefers to stay and move around within the squeezed Iodine layers rather than the stretched Bromine layers.

The Big Takeaway

The researchers showed that by arranging the atoms in neat, alternating layers, they can create a "one-way street" for defects.

• Along the layers: Defects can still move (like cars on a highway).
• Across the layers: Defects are effectively blocked (like a wall).

This is important because if you can stop the defects from moving in the direction that hurts the solar cell (usually moving toward the surface or interfaces), you can make the material more stable and last longer. The paper suggests that by "engineering the strain" (squeezing and stretching the layers just right), you can control exactly where these tiny defects are allowed to go, keeping the solar cell working better for longer.

Technical Summary

Technical Summary: Anisotropic Defect Diffusion in Layered CsPbBrxI3−x Perovskites

Problem Statement
Mixed-halide perovskites (MHPs) offer tunable optoelectronic properties and improved thermal stability compared to organic-inorganic counterparts, yet they face significant challenges regarding structural stability. Specifically, the photoactive black phase of CsPbI3 is metastable at room temperature and prone to transitioning into a non-perovskite yellow phase. While substituting Iodine (I) with Bromine (Br) can stabilize the lattice via the Goldschmidt tolerance factor, it often widens the band gap, reducing suitability for single-junction solar cells. Recent experimental strategies involving layered halide ordering (alternating Br and I layers) have shown promise in reducing band gaps and increasing carrier mobility. However, the impact of such ordering on the kinetics of degradation—specifically the migration of point defects (vacancies and interstitials)—remains unclear. Lattice defects are known to initiate degradation, and understanding how to control their mobility through structural ordering and strain engineering is critical for device longevity.

Methodology
The authors employed large-scale molecular dynamics (MD) simulations using a reactive force field (ReaxFF) to investigate defect migration in CsPbBrxI3−x perovskites.

• Structural Models: The study focused on ordered (od) structures, specifically od-CsPbBr2I and od-CsPbBrI2, where majority and minority halide ions occupy alternating layers. These were compared against unordered (ud) random alloys of the same composition, as well as pristine CsPbBr3 and CsPbI3.
• Simulation Parameters: Simulations were conducted at 700 K using the AMS2024 software package. Systems consisted of 4x4x4 supercells (320 atoms) containing single point defects (Cs, Br, or I vacancies and interstitials).
• Analysis Techniques:
  • Voronoi Decomposition: A dynamic site-based analysis was used to track defect positions. The simulation cell was partitioned into discrete sites based on the instantaneous positions of Pb atoms. Unoccupied sites were identified as vacancies, and doubly occupied sites as interstitials.
  • Diffusion Coefficients: Mean Squared Displacement (MSD) was calculated to determine diffusion coefficients ($D$) in specific Cartesian directions.
  • Strain Isolation: To distinguish between chemical composition effects and mechanical strain, the authors applied controlled biaxial strain to pristine CsPbI3 and CsPbBr3 to replicate the octahedral tilting patterns observed in the layered mixed-halide systems.

Key Contributions and Results

1. Anisotropic Defect Diffusion: The primary finding is that layered halide ordering induces strongly anisotropic defect diffusion. Migration proceeds readily parallel to the halide layers but is strongly suppressed perpendicular to them.
2. Mechanism for Cesium (Cs) Defects:
   • The anisotropy in Cs vacancy and interstitial migration is driven by directional lattice strain and associated octahedral tilting.
   • In ordered structures, the Pb-I layers exhibit significant octahedral tilting (checkerboard pattern) due to the strain of linking differently sized octahedra. This tilting "locks" Pb-I-Pb angles, suppressing the "gate-opening" rearrangement required for Cs ions to jump between sites across the layers.
   • Simulations of strained pristine CsPbI3 confirmed that reproducing this specific tilting pattern is sufficient to induce similar anisotropic diffusion, isolating strain as the governing factor for Cs defects.
3. Mechanism for Halide Defects:
   • Halide Interstitials: Their migration is governed by a combination of strain and preferential local chemical bonding. Interstitials preferentially form "double bridge" configurations with lattice halides of the same species (Br-Br or I-I). Consequently, interstitials are confined to layers where they can maintain these favorable bonding configurations (e.g., Br interstitials in Br-rich layers).
   • Halide Vacancies: Vacancy migration is primarily driven by the energy landscape and strain. Vacancies are energetically favored in Pb-I layers (which are under compressive strain) and migrate predominantly within these layers. Compressive strain stabilizes the vacancy and lowers the diffusion barrier, whereas tensile strain (found in Br layers) hinders it.
4. Ordering vs. Randomness: In ordered structures, the difference in diffusion coefficients between Iodine and Bromine migration is one to two orders of magnitude. In unordered (random) structures, this difference is significantly smaller (factor of 1.2–3.6), as asymmetric local environments allow both halides to participate in diffusion.

Significance and Claims
The paper claims to provide atomistic insights into how strain engineering via halide layering can be used to control defect-mediated transport in perovskites. By demonstrating that layered ordering can effectively confine defect motion to specific crystallographic planes, the authors suggest a design strategy to improve material stability.

• Stability: Suppressing defect migration perpendicular to the layers could mitigate degradation pathways that rely on long-range ion transport.
• Device Design: In operating devices, ionic motion can cause hysteresis or interface degradation. The authors propose that aligning the ordered layers perpendicular to the electric field in a device could suppress unwanted ionic motion.
• Scope: The study emphasizes that while strain is a dominant factor for Cs defects, chemical bonding preferences are equally critical for halide interstitials. The work does not propose new experimental synthesis methods but rather offers a theoretical framework to interpret and guide the optimization of existing layered perovskite structures for high-performance optoelectronic applications.