Atomic Structure of Grain Boundaries, Dislocations and Associated Strain in Templated Co-evaporated Photoactive Halide Perovskites

This study employs tailored low-dose electron microscopy to reveal the atomic structures of grain boundaries, edge dislocations, and associated strain fields in templated FA0.9Cs0.1PbI3-xClx perovskite films, providing critical insights into how specific defects influence charge transport and device performance.

The Gist

The Big Picture: Building a Better Solar City

Imagine you are trying to build a super-efficient solar city using a special type of "magic brick" called Perovskite. These bricks are amazing at capturing sunlight and turning it into electricity. However, for a long time, these cities had a major problem: the bricks didn't fit together perfectly.

When you lay down millions of these bricks to make a solar panel, they don't form one giant, smooth block. Instead, they form a patchwork of many small, separate "neighborhoods" (called grains). Where these neighborhoods meet, there are messy borders called Grain Boundaries.

Think of these borders like the seams between two patches of a quilt. If the seams are messy, the quilt is weak, and the "electricity traffic" gets stuck or lost at the seams. This makes the solar panel less efficient and causes it to break down faster.

The Experiment: The "Templated" Construction Site

In this study, scientists tried a new construction trick. They used a template—a very thin, invisible guide layer underneath the bricks. Think of this like laying down a perfectly flat, pre-drawn grid on the ground before building.

The Result: The new bricks lined up much better! They all stood up straight (pointing in the same direction), which is great for electricity flow. But, the scientists wanted to know: Did the seams between the neighborhoods become perfect, or were they still messy?

To find out, they used a special "super-microscope" that uses very gentle beams of electrons (so it doesn't melt the delicate bricks) to look at the atomic level.

What They Found: The Three Types of "Seams"

The scientists discovered that even with the template, the neighborhoods still met in three different, imperfect ways. Here is what they found, using simple analogies:

1. The "Random Collision" Seams (High-Angle Boundaries)

Imagine two crowds of people walking toward each other. One group is facing North, and the other is facing East. When they meet, they crash into each other at a weird angle.

• What the paper says: Most of the time, the grains just hit each other at random angles.
• The Problem: The atoms at the seam don't line up at all. It's like trying to zip two zippers together where the teeth are completely different shapes. This creates "dangling bonds" (loose, unconnected atoms) that act like potholes, trapping electricity and causing the solar cell to lose power.

2. The "Mirror Image" Seams (Special Symmetrical Boundaries)

Sometimes, two neighborhoods meet in a very specific way, like a perfect mirror reflection.

• What the paper says: They found a special type of seam (called a $\Sigma5$ boundary) where the atoms line up perfectly every fifth spot.
• The Good News: These are much cleaner than the random ones. It's like two puzzle pieces that fit together almost perfectly.
• The Bad News: Even these "good" seams aren't perfect. There are still tiny gaps and stress points that can cause trouble.

3. The "Twisted Roads" (Dislocations and Strain)

This is the most interesting part. Sometimes, the "roads" inside the neighborhoods get twisted.

• The Analogy: Imagine a stack of pancakes. If you push the top half of the stack slightly to the right, the layers don't line up anymore. You create a "kink" or a dislocation.
• The Strain: The paper found that around these kinks, the material is either being squished (compressed) or stretched (tensioned).
  • Analogy: Think of a rubber band. If you stretch it, it's under tension. If you push it, it's compressed. The atoms in the solar cell are being squeezed and pulled apart in these twisted areas.
• The Consequence: This stretching and squeezing changes the energy of the material, creating "traps" where electricity gets stuck and disappears as heat instead of power.

The "Intruder" Brick (PbI2)

They also found a few stray bricks that were the wrong type entirely (called PbI2).

• The Analogy: Imagine you are building a wall with red bricks, but someone accidentally slipped in a few blue bricks.
• The Problem: Even though there were very few of these blue bricks, they didn't fit well with the red ones. Where they met, the wall was bumpy and stressed, creating more traps for electricity.

The Bottom Line: Why This Matters

The Good News: The new "template" method works! It forces the solar bricks to stand up straight, which is a huge improvement.

The Bad News: Even with the template, the "seams" between the bricks are still messy. They are full of twists, turns, and stress points that steal electricity.

The Takeaway:
The scientists realized that to make the perfect solar cell, we can't just use templates to line up the bricks; we need to figure out how to make the seams disappear entirely. Ideally, we want to build a solar city made of one giant, seamless crystal block, rather than a patchwork of many small neighborhoods.

This paper is like a detailed map showing exactly where the potholes are in our solar city, so engineers can finally figure out how to pave over them and build a perfect road for electricity.

Technical Summary

1. Problem Statement

Metal halide perovskites (OIMHPs) have emerged as leading candidates for next-generation photovoltaics, with power conversion efficiencies (PCE) exceeding 27% in single-junction cells. However, their performance and stability are limited by structural defects, particularly grain boundaries (GBs) and intra-grain defects (dislocations, stacking faults).

• The Gap: While templated co-evaporated films of FA$_{0.9}$Cs$_{0.1}$PbI$_{3-x}$Cl$_x$ have shown improved grain orientation and efficiency, the underlying atomic mechanisms governing defect formation and strain at these boundaries remain unclear.
• The Challenge: Conventional electron microscopy techniques (like EBSD or standard STEM) often deliver high electron doses that damage these sensitive materials, causing structural degradation, ion migration, and radiolysis. This makes it difficult to observe the true atomic structure of GBs and defects without altering them.

2. Methodology

The authors employed a suite of ultra-low-dose electron microscopy techniques to preserve the structural integrity of the perovskite film while achieving atomic resolution.

• 4D-STEM in SEM: A four-dimensional scanning transmission electron microscopy (4D-STEM) setup was built within a Scanning Electron Microscope (SEM).
  • Advantage: Lower electron energy (reducing knock-on damage) and a large field of view compared to TEM-based 4D-STEM.
  • Process: A 2 nm collimated probe scanned a 5 µm × 5 µm area, recording 40,000 diffraction patterns to map grain orientation and phase without significant beam damage.
• Low-Dose High-Resolution STEM (TEM): Ultra-low-dose Low-Angle Annular Dark Field (LAADF) imaging was used in a Transmission Electron Microscope (TEM) to resolve atomic structures at GBs and dislocations.
• Data Analysis:
  • Butterworth Filtering: Applied to LAADF images to enhance contrast while preserving atomic details.
  • Geometrical Phase Analysis (GPA): Used on raw (unfiltered) data to map local strain fields ($\epsilon_{xx}$, $\epsilon_{yy}$, $\epsilon_{xy}$) around defects.
  • Simulations: Density Functional Theory (DFT) and simulated electron diffraction patterns were used to validate experimental observations of specific boundary structures (e.g., Coincident Site Lattice boundaries).

3. Key Contributions

• Development of a Low-Dose Workflow: Demonstrated a robust methodology combining SEM-based 4D-STEM and TEM-based low-dose LAADF to characterize radiation-sensitive perovskites without inducing beam damage.
• Atomic-Scale Defect Mapping: Provided the first detailed atomic structures of various GB types (high-angle, low-angle, CSL, pseudo-twin) and their associated strain fields in templated FA$_{0.9}$Cs$_{0.1}$PbI$_{3-x}$Cl$_x$ films.
• Strain-Defect Correlation: Established a direct link between specific atomic configurations (e.g., edge dislocations, stacking faults) and localized compressive/tensile strain fields.

4. Key Results

A. Grain Orientation and Growth Mechanism

• Preferred Orientation: The templated films exhibit a strong preferred orientation along the 〈001〉 zone axis perpendicular to the substrate.
• Arbitrary Rotation: While aligned vertically, grains are randomly rotated about the 〈001〉 axis.
• Growth Model: This structure supports a Volmer–Weber growth mechanism, where grains nucleate independently and grow perpendicular to the substrate, colliding at arbitrary angles.
• Phase Purity: The film is predominantly cubic perovskite ($Pm\bar{3}m$), with only ~0.22% of the area containing hexagonal/trigonal PbI$_2$ phases.

B. Grain Boundary (GB) Structures

1. High-Angle GBs (Random):
   • Most adjacent grains have arbitrary misorientation angles (e.g., 20.8°, 26°, 36.9°, 41.7°).
   • These boundaries are incoherent, lacking amorphous layers but containing point defects and dangling bonds that likely act as recombination centers.
2. Coincident Site Lattice (CSL) GBs:
   • Observed a $\Sigma5$ (310)/[001] boundary with a 36.9° misorientation.
   • This boundary is atomically flat and coherent, featuring mirror symmetry and a specific arrangement of "kite-shaped" structural units. It represents a lower-energy interface compared to random GBs.
3. Low-Angle GBs:
   • Boundaries with misorientation <15° (e.g., 5.3°, 7°) are composed of arrays of edge dislocations.
   • Two types of edge dislocation cores were identified: one involving removed half-planes and another involving shifted cores, both creating dangling bonds.

C. Strain Fields and Intra-Grain Defects

• Edge Dislocation Strain: GPA strain mapping revealed that edge dislocations create a distinct strain field: compressive strain on the side with the extra half-plane and tensile strain on the opposite side.
• Pseudo-Twin Boundaries: Observed 90° planar defects (antiphase boundaries) parallel to {110} planes. These involve edge-sharing octahedra (unlike the standard corner-sharing) and create "zig-zag" structures.
  • These defects induce significant local strain and are associated with stacking faults parallel to {100} planes.
  • Strain fields at these defects show alternating positive/negative transitions across dislocation cores.
• Perovskite/PbI$_2$ Interface: While some interfaces appear locally coherent, atomic resolution revealed that most trigonal PbI$_2$ inclusions create edge dislocations and significant lattice mismatch strain at the interface, suggesting they are not benign.

5. Significance and Implications

• Defect Impact on Performance: The study confirms that while the templating layer improves overall grain orientation, the resulting boundaries (both high and low angle) and intra-grain defects introduce dangling bonds, trap states, and localized strain. These factors are likely to act as recombination centers, limiting the theoretical efficiency of the solar cells.
• Strain Engineering: The detailed mapping of strain fields provides a framework for understanding how local lattice distortions modify bandgaps and charge transport.
• Future Directions: The findings suggest that to achieve single-junction efficiencies approaching the Shockley-Queisser limit, future growth strategies must aim to minimize these defects, potentially by moving toward single-crystal films or optimizing the templating process to reduce arbitrary grain rotations and dislocation densities.
• Methodological Impact: The successful application of low-dose 4D-STEM and LAADF sets a new standard for characterizing beam-sensitive materials, ensuring that observed structures reflect the material's native state rather than beam-induced artifacts.