Determination of Burgers vectors of dislocations in monoclinic $\beta$-Ga$_2$O$_3$ crystals by large-angle convergent-beam electron diffraction

This study demonstrates that large-angle convergent-beam electron diffraction (LACBED), utilizing a dual lattice basis approach to bypass the need for a metric tensor, can effectively and unambiguously determine the Burgers vectors of dislocations in monoclinic $\beta$-Ga$_2$O$_3$ crystals, with results validated by weak-beam dark-field imaging.

The Gist

Imagine a crystal of $\beta$-Ga$_2$O$_3$ (a special material used for making powerful, efficient electronics) as a giant, perfectly stacked library of books. In a perfect library, every book sits in a neat, straight row. But in real life, things get messy. Sometimes, a book is shoved in the wrong spot, or a whole row is shifted. In the world of crystals, these "messy spots" are called dislocations.

To fix the library or understand why it's not working right, you need to know exactly how the books are messed up. You need to know the direction and size of the shift. In physics, this "shift" is called the Burgers vector.

The Problem: A Twisted Library

Most materials have a simple, box-like structure (like a standard grid). But $\beta$-Ga$_2$O$_3$ is different; it has a monoclinic structure. Think of this not as a neat grid of boxes, but as a stack of books that are slightly tilted and leaning against each other.

Because the "books" are leaning, the usual math tools scientists use to measure the shifts (called "metric tensors") become complicated and hard to use. It's like trying to measure the distance between two leaning shelves using a ruler meant for straight walls; the angles make the math messy.

The Solution: A New Way to Count

The researchers in this paper wanted to prove they could still measure these shifts accurately, even in this "leaning" crystal. They used a technique called LACBED (Large-Angle Convergent-Beam Electron Diffraction).

Here is the simple analogy for how LACBED works:
Imagine shining a flashlight through a stained-glass window. If there is a crack in the glass (a dislocation), the light pattern changes. Specifically, the crack creates a series of "kinks" or "nodes" (little breaks) in the light lines.

The magic rule the scientists used is: The number of kinks tells you the size of the shift.

• If you see 2 kinks, the shift is a certain size.
• If you see -3 kinks (a specific direction of shift), it's a different size.

The big breakthrough in this paper is showing that you don't need the complicated "leaning shelf" math to count these kinks. Because of a special relationship between the crystal's physical shape and the way light bounces off it, the scientists could count the kinks and solve the puzzle using simple, straight-line math, just like they would for a normal, box-shaped crystal.

The Experiment: Making a Mess on Purpose

To test this, the scientists didn't just look at random messes. They created their own:

1. The Indent: They took a tiny, super-hard diamond tip (like a very sharp needle) and pressed it into the crystal surface. This is called "nanoindentation."
2. The Damage: This pressure created a cluster of dislocations (messy shifts) right under the tip, spreading out like cracks in a windshield.
3. The Scan: They sliced the crystal open and used an electron microscope to take "photos" of the light patterns (LACBED) around these cracks.

The Results: Counting the Kinks

They picked 8 specific cracks (labeled D-1 to D-8) and counted the kinks in the light patterns for three different angles.

• The Math: They set up three simple equations based on the number of kinks they saw.
• The Answer: When they solved the equations, every single crack had the exact same "shift" vector: [0 1 0].

To double-check their work, they used a different method called WBDF (Weak-Beam Dark-Field imaging). This is like looking at the cracks in shadow.

• When they looked at the cracks from one angle, the shadows disappeared (meaning the shift was parallel to the light).
• When they looked from another angle, the shadows were clear.
• This shadow test confirmed exactly what the "kink counting" method found: All the cracks were shifting in the same direction.

The Bottom Line

This paper proves that even though $\beta$-Ga$_2$O$_3$ has a weird, tilted crystal structure, scientists can use the "kink counting" method (LACBED) to accurately measure how the crystal is broken. They showed that they don't need complex, messy math to do it; the standard, simple counting method works perfectly.

This is important because knowing exactly how these crystals are broken helps engineers understand how to make better, more reliable power electronics in the future. But for now, the main achievement is simply proving that the "kink counting" tool works on this specific, tricky material.

Technical Summary

1. Problem Statement

• Material Context: Monoclinic $\beta$-Ga$_2$O$_3$ is a promising wide-bandgap semiconductor (4.5 eV) for ultra-high-voltage power devices. However, high-quality bulk crystals still contain significant dislocation densities ($\sim 10^4$ cm$^{-2}$), which degrade device performance.
• The Challenge: To understand and mitigate defect impacts, it is crucial to determine not just the direction, but also the magnitude and sense of the Burgers vector ($\mathbf{b}$) of dislocations.
• Limitations of Existing Methods:
  • Weak-Beam Dark-Field (WBDF) Imaging: While standard for dislocation analysis, WBDF relies on the invisibility criterion ($\mathbf{g} \cdot \mathbf{b} = 0$). This allows determination of the direction of $\mathbf{b}$ but makes it difficult to determine its magnitude or sign unambiguously.
  • Crystal Structure Complexity: $\beta$-Ga$_2$O$_3$ has a monoclinic (non-orthogonal) crystal structure. Traditional methods for calculating the inner product $\mathbf{g} \cdot \mathbf{b}$ often require a metric tensor, which complicates analysis compared to cubic or hexagonal systems.
• Gap: The applicability of Large-Angle Convergent-Beam Electron Diffraction (LACBED) for determining Burgers vectors in non-orthogonal $\beta$-Ga$_2$O$_3$ had not been sufficiently investigated.

2. Methodology

The study employed a combined experimental and theoretical approach:

• Sample Preparation:
  • Used a Sn-doped (2$\bar{1}$01)-oriented $\beta$-Ga$_2$O$_3$ single-crystal wafer (EFG-grown, 680 $\mu$m thick).
  • Introduced controlled dislocations via nanoindentation using a Berkovich indenter (50 mN load) to create a localized strain field.
  • Prepared cross-sectional Transmission Electron Microscopy (TEM) specimens using Focused Ion Beam (FIB) milling perpendicular to the [102] direction.
• Theoretical Framework (Non-Orthogonal Systems):
  • The authors established a method to calculate the inner product $\mathbf{g} \cdot \mathbf{b}$ in non-orthogonal systems without explicitly using a metric tensor.
  • They utilized the dual relationship between real-space lattice bases ($\mathbf{a}_i$) and reciprocal-space bases ($\mathbf{g}_i$), defined by $\mathbf{g}_i \cdot \mathbf{a}_j = 2\pi\delta_{ij}$.
  • This allows the calculation $\mathbf{g} \cdot \mathbf{b} = 2\pi N$ (where $N$ is an integer) to be performed directly using integer coefficients of the basis vectors, simplifying the analysis for monoclinic crystals.
• Experimental Techniques:
  • LACBED: Performed on a JEOL JEM-2100Plus (200 kV). The number of nodes ($n$) where the dislocation line intersects a Laue reflection line was counted. The relation $\mathbf{g} \cdot \mathbf{b} = n$ was used to solve for the Burgers vector components.
  • WBDF: Used for validation under two-beam conditions ($\mathbf{g} = \bar{2}01$ and $\mathbf{g} = 020$) with a $g/3g$ weak-beam condition.

3. Key Contributions

1. Theoretical Adaptation: Successfully demonstrated that the metric tensor is unnecessary for calculating $\mathbf{g} \cdot \mathbf{b}$ in monoclinic systems by leveraging the dual basis relationship. This makes LACBED analysis mathematically straightforward for $\beta$-Ga$_2$O$_3$.
2. Methodological Validation: Proved that LACBED can be effectively applied to monoclinic $\beta$-Ga$_2$O$_3$ to unambiguously determine the full Burgers vector (direction, magnitude, and sense).
3. Experimental Demonstration: Applied the method to nanoindentation-induced dislocations, solving the Burgers vectors for eight distinct dislocations (D-1 to D-8) and confirming the results against WBDF imaging.

4. Results

• LACBED Analysis:
  • For dislocation D-1, LACBED patterns were acquired using three different reciprocal lattice vectors ($\mathbf{g}_1, \mathbf{g}_2, \mathbf{g}_3$).
  • The observed node counts were $n_1 = 2$, $n_2 = -3$, and $n_3 = -2$.
  • Solving the resulting system of linear equations yielded a Burgers vector of $\mathbf{b} = [0\bar{1}0]$.
  • Analysis of dislocations D-2 through D-8 consistently revealed Burgers vectors of $\mathbf{b} = \langle 010 \rangle$.
• WBDF Verification:
  • Under $\mathbf{g} = \bar{2}01$, the dislocations showed very weak contrast (consistent with $\mathbf{g} \cdot \mathbf{b} \approx 0$).
  • Under $\mathbf{g} = 020$, the dislocations showed strong contrast (consistent with $\mathbf{g} \cdot \mathbf{b} \neq 0$).
  • This contrast behavior perfectly matched the $\mathbf{b} = \langle 010 \rangle$ result derived from LACBED, validating the LACBED method.
• Observation: All analyzed dislocations around the indentation possessed the same Burgers vector direction, suggesting a specific slip system activation under the indentation load, though the specific mechanism was noted as outside the scope of this study.

5. Significance

• Defect Characterization: This work provides a robust, reliable tool for characterizing dislocations in $\beta$-Ga$_2$O$_3$, a critical step toward improving crystal growth processes and device reliability.
• Beyond Direction: Unlike WBDF, this method allows researchers to determine the magnitude and sense of the Burgers vector. This is vital for understanding phenomena analogous to those in SiC and GaN (e.g., hollow micropipes or basal plane dislocation expansion), where the specific nature of the dislocation core dictates device failure modes.
• General Applicability: The theoretical framework regarding dual bases offers a simplified approach for analyzing vector inner products in any non-orthogonal crystal system, potentially benefiting the study of other complex semiconductor materials.
• Future Impact: By enabling precise defect mapping, this technique supports the development of process control guidelines to reduce dislocation density, directly contributing to the realization of high-efficiency, ultra-high-voltage power conversion devices.