Three-dimensional visualization of lattice defects in $\beta$-Ga$_2$O$_3$ via synchrotron-radiation Borrmann-effect X-ray topo-tomography

This study presents the first demonstration of three-dimensional visualization of dislocations in $\beta$-Ga$_2$O$_3$ using synchrotron-radiation X-ray topo-tomography under Borrmann-effect conditions, enabling depth-resolved analysis of defect propagation in device structures.

The Gist

Imagine you are trying to build a perfect, high-speed highway for electricity. The material you are using is called β-Ga2O3 (beta-gallium oxide). It's a superstar material for next-generation electronics because it can handle huge amounts of power without melting or breaking.

But here's the problem: even the best materials have "potholes." In the world of crystals, these potholes are called lattice defects or dislocations. They are tiny misalignments in the atomic structure. If you have too many of them, your electronic devices (like the power converters in electric cars or solar inverters) will fail or perform poorly.

For a long time, scientists could only see these potholes in 2D, like looking at a flat map of a city. They could see where the potholes were on the surface, but they couldn't tell if a pothole was just a small bump on the road or a deep, dangerous crater that went all the way down to the bedrock.

This paper is about building a 3D "Google Earth" for these tiny atomic potholes.

The Magic Flashlight: Synchrotron Radiation

To see these invisible defects, the researchers used a super-powered flashlight called a synchrotron. This isn't a normal flashlight; it's a giant machine that shoots incredibly bright, focused beams of X-rays.

Normally, X-rays just pass right through solid objects, or they get absorbed (which is how medical X-rays work). But the researchers used a special trick called the Borrmann effect.

Think of the Borrmann effect like a ghostly dance. When X-rays hit a perfect crystal at a very specific angle, they don't just pass through or bounce off; they "dance" through the crystal in a way that makes the perfect parts of the crystal almost invisible to the X-rays. However, if there is a defect (a pothole), the dance gets disrupted. The defect shows up as a dark shadow against a bright background.

The 3D Trick: Spinning the Crystal

Here is the clever part. In the past, scientists could only take a snapshot of the crystal from one angle. If a defect was hidden behind another, they couldn't tell where it was in depth.

In this study, the researchers put the crystal on a high-precision turntable and slowly spun it while taking pictures with their super X-ray camera.

Imagine you are looking at a spinning top with a few stickers on it.

• If you look at it from the front, the stickers might look like they are in a straight line.
• If you look at it from the side, the stickers might look like they are scattered.
• If you spin it and watch how the stickers move relative to each other, you can figure out exactly how deep each sticker is stuck into the top.

The researchers did this with the crystal. By rotating the sample and watching how the "shadows" of the defects moved and changed shape, they could mathematically reconstruct a 3D model of the defects.

What Did They Find?

Using this 3D "X-ray CT scan," they looked at a device called a Schottky Barrier Diode (SBD). This device has two main parts:

1. The Substrate: The thick, bottom layer (the bedrock).
2. The Epilayer: A thin, high-quality layer grown on top (the new highway).

The Big Discovery:
They found that most of the "potholes" (dislocations) weren't vertical tunnels going straight down. Instead, they were mostly flat, horizontal cracks lying on the surface of the layers.

• The Good News: The deep "bedrock" defects rarely reached up to ruin the new "highway" layer.
• The Bad News: The defects right at the border between the bedrock and the highway were the troublemakers. They acted like a messy foundation, causing the new layer to grow with its own tangles of defects.

Why Does This Matter?

Before this, engineers were worried about deep, vertical defects. They were trying to fix the "bedrock" all the way down.

This paper tells us: "Stop worrying about the deep bedrock! Focus on the surface."

If you want to build better, faster, and more reliable electronics, you need to make sure the surface where the new layer is grown is perfectly smooth. If you fix the "interface" (the meeting point between the two layers), the rest of the device will work much better.

In a Nutshell

This paper is the first time anyone has successfully created a 3D movie of atomic defects inside this specific material. It's like giving engineers a pair of 3D glasses that let them see exactly where the weak spots are, so they can fix the right place and build better power electronics for the future.

Technical Summary

1. Problem Statement

• Material Context: $\beta$-Ga$_2$O$_3$ is a critical next-generation material for power electronics due to its wide bandgap (~4.8 eV) and high breakdown field strength.
• The Challenge: Device performance and reliability are severely degraded by lattice defects, particularly dislocations, stacking faults, and voids.
• Limitations of Current Methods:
  • Conventional characterization techniques (selective chemical etching, standard X-ray topography, TEM, phase-contrast microscopy) provide valuable 2D data or limited depth sensitivity but fail to deliver full three-dimensional (3D) structural information.
  • Standard X-ray Computed Tomography (CT) relies on absorption contrast, which is inherently insensitive to lattice defects like dislocations in highly crystalline semiconductors.
  • Existing 3D visualization methods used for materials like GaN and SiC are not readily applicable to $\beta$-Ga$_2$O$_3$.

2. Methodology

The authors developed a novel approach using Synchrotron-Radiation X-ray Topo-Tomography under a specific diffraction condition to achieve 3D visualization.

• Experimental Setup:
  • Source: Synchrotron radiation monochromatic beam ($\lambda = 1.24$ Å) at KEK Photon Factory (BL-14B).
  • Samples:
    1. Bare Sn-doped (001)-oriented $\beta$-Ga$_2$O$_3$ substrate (EFG method).
    2. Schottky Barrier Diode (SBD) wafer with a 12 $\mu$m HVPE-grown epilayer on a similar substrate.
  • Geometry: Transmission X-ray topography (XRT) utilizing the Borrmann effect (anomalous transmission).
    • Reflection planes: (020) with diffraction vector $\mathbf{g} = 020$.
    • Symmetric Laue case geometry.
    • Two-beam condition (excitation of both transmitted o-wave and diffracted h-wave).
• Data Acquisition (Topo-Tomography):
  • The sample was rotated stepwise about the diffraction vector ($\mathbf{g}$-vector), corresponding to the $\chi$-axis of the goniometer.
  • Rotation Range: Images were acquired from $\chi = -20^\circ$ to $+20^\circ$ (for SBDs) and $0^\circ$ to $60^\circ$ (for substrates) with fine increments (0.5° for SBDs).
  • Mechanism: As the sample rotates, the projected position of defects changes relative to the detector based on their depth ($z$-coordinate). By capturing the evolution of dislocation contrast across these angles, depth information is encoded.
• Reconstruction Logic:
  • Mathematical modeling using rotation matrices ($R(\omega)$ and $R(\chi)$) relates the 3D crystal coordinates $(x, y, z)$ to the 2D camera coordinates $(x_1, y_1)$.
  • The variation in the $y_1$ coordinate as a function of $\chi$ allows for the extraction of depth ($z$) information.
  • Full 3D reconstruction is achieved using algorithms based on the inverse Radon transform and Fourier slice theorem (adapted for cone-beam topo-tomography).

3. Key Contributions

• First Demonstration: This is the first study to successfully demonstrate 3D dislocation reconstruction in $\beta$-Ga$_2$O$_3$.
• Depth-Resolved Visualization: The method enables intuitive, depth-resolved observation of dislocations, allowing researchers to distinguish between defects in the epilayer and the substrate within SBD structures without destructive sectioning.
• Validation of Geometry: The study validated that dislocations in $\beta$-Ga$_2$O$_3$ predominantly lie on the (001) plane, as experimental contrast evolution matched geometric simulations perfectly.
• Quantitative Analysis: The technique provides a way to map the spatial distribution and geometrical characteristics of defects (dislocations, voids, inclusions) within the crystal bulk.

4. Key Results

• Dislocation Behavior:
  • Orientation: Most dislocations lie on the (001) plane and extend predominantly along the $\langle 010 \rangle$ direction (b-axis). This aligns with the active slip system $\langle 010 \rangle\{001\}$.
  • Threading Dislocations: Very few threading dislocations (perpendicular to the (001) plane) were observed.
• Epitaxial Propagation:
  • There was no clear evidence of dislocations propagating from deep within the substrate into the epilayer.
  • Only dislocations located very close to the substrate/epilayer interface appeared to influence subsequent epitaxial growth.
  • Implication: In-plane dislocations near the interface are more critical to device performance than threading dislocations for (001)-oriented SBDs.
• Tangled Dislocations:
  • Tangled dislocation complexes in the substrate were observed to induce complex, tangled dislocation structures in the epilayer.
  • While a direct 1-to-1 correlation is difficult due to the complexity, the presence of substrate tangles correlates with epilayer defects.
• Visualization Capability: The method successfully separated specific dislocations (e.g., $e1, e2$ in the epilayer vs. $s1$ in the substrate) by analyzing their relative positional shifts during rotation.

5. Significance

• Device Optimization: By identifying that near-interface in-plane dislocations are the primary drivers of epitaxial defects, this study provides a clear target for improving crystal growth processes and device performance.
• Non-Destructive 3D Characterization: The technique offers a non-destructive alternative to TEM or sectioning, preserving the sample while providing full volumetric defect data.
• Broader Applicability: While demonstrated on $\beta$-Ga$_2$O$_3$, this topo-tomography approach under the Borrmann effect holds strong potential for 3D defect visualization in other highly crystalline semiconductors where absorption-based CT fails.
• Quality Control: The ability to visualize defect propagation (or lack thereof) offers a powerful tool for quality assurance in the manufacturing of high-power electronic devices.