First-principles band alignment engineering in polar and nonpolar orientations for wurtzite AlN, GaN, and B$_x$Al$_{1-x}$N alloys

This study employs advanced computational methods to determine and analyze the polar and nonpolar band alignments of wurtzite B$_x$Al$_{1-x}$N alloys, revealing composition-dependent type I or II alignments and surface polarity effects that provide critical design guidelines for high-electron-mobility transistors and ultraviolet optoelectronic devices.

The Gist

Imagine you are an architect trying to build a super-fast, super-efficient electronic city. To do this, you need to stack different layers of materials on top of each other, like a skyscraper made of different types of glass and steel. For these layers to work together, the "energy floors" inside them need to line up perfectly. If the floors don't match, the electricity (the people walking through the building) gets stuck, falls down a hole, or bounces back the wrong way.

This paper is about designing the blueprints for a specific, ultra-modern building material called Boron Aluminum Nitride (BxAl1−xN). This material is like a "super-glass" that can handle extreme heat and block electricity very well, making it perfect for next-generation electronics and deep-ultraviolet light devices.

Here is what the researchers did, explained simply:

1. The Problem: The "Floor" Mismatch

The researchers wanted to know exactly how the energy floors of this new "super-glass" line up when stacked against two other common materials: Aluminum Nitride (AlN) and Gallium Nitride (GaN).

Think of Band Alignment as the height of the floor in a building.

• Valence Band: The floor where people (electrons) usually hang out.
• Conduction Band: The ceiling or the next floor up where people can run freely.

If you stack two materials and their floors don't match, the electrons get confused. The researchers needed to calculate these heights precisely to tell engineers how to build devices that work.

2. The Challenge: The "Spinning Top" Effect

Calculating these heights is tricky because these materials are polar. Imagine a spinning top that has a built-in electric charge at its top and bottom. When you try to measure the "floor height" of a spinning top, the charge messes up your ruler.

• The Old Way: Previous methods tried to measure these materials by ignoring the spin, which led to wrong answers.
• The New Trick: The authors used a clever "passivation" technique. Imagine putting a special, invisible "cap" (called pseudohydrogen) on the top and bottom of the material slice. This cap neutralizes the spinning charge, allowing them to measure the floor heights accurately without the ruler getting confused.

3. The Two Angles: Looking from the Top vs. the Side

The researchers looked at the material from two different angles, like looking at a brick from the top (the c-plane) or from the side (the a-plane).

• The Top View (Polar c-plane):

  • When they mixed a little bit of Boron into the Aluminum Nitride (low amounts), the floors lined up almost perfectly (near-zero difference). This is great for letting electrons flow smoothly.
  • When they added more Boron, the floors started to shift. Sometimes the new material's floor was higher, sometimes lower. This creates a "staggered" effect (Type II alignment), which is useful for trapping electrons in specific spots.
  • Surprise: They found that the "floor height" depends heavily on how the atoms are arranged. If the atoms are slightly squished or twisted (tetrahedral distortion), the floor height changes.

• The Side View (Non-polar a-plane):

  • Here, the rules changed. As they added more Boron, the "floor" (Valence Band) dropped lower and lower, while the "ceiling" stayed roughly the same.
  • This creates a situation where the material acts like a natural slide for electrons. The researchers noted that for high Boron content, the material even has "negative electron affinity," which is like having a floor that is so low it naturally pushes electrons out into the air. This could be used to make spontaneous electron emitters.

4. The "Magic" of Boron

The paper highlights that Boron is the secret ingredient.

• Low Boron: The material acts very much like Aluminum Nitride.
• High Boron: The material behaves more like Boron Nitride, which has a very different energy structure.
• The Twist: The relationship isn't a straight line. At certain middle amounts of Boron, the atoms get "squished" (distorted), causing the energy floors to jump up or down unexpectedly.

5. Checking the Work

The researchers compared their computer calculations with real-world experiments done by other scientists.

• The Good News: Their numbers matched the real-world experiments very well, especially for the "Top View" (c-plane) materials.
• The Warning: They also tried an older, simpler method (called the SSE approach) that ignores the surface angles. They found that this old method was often wrong because it missed the "spinning top" effects and the specific way the atoms are arranged on the surface.

The Bottom Line

This paper provides the first accurate "blueprints" for how to stack this new Boron-Aluminum-Nitride material with existing ones.

• For Engineers: It tells them that by tweaking the amount of Boron and choosing the right angle (top or side view), they can design devices that either trap electrons tightly (for LEDs) or let them fly freely (for high-speed transistors).
• The Takeaway: You can't just guess how these materials stack; you have to account for the "spin" of the material and the exact angle you are looking at it, or your electronic city will have mismatched floors and won't work.

Technical Summary

Technical Summary: First-Principles Band Alignment Engineering in Polar and Nonpolar Orientations for Wurtzite AlN, GaN, and BxAl1−xN Alloys

Problem Statement
Boron aluminum nitride (BxAl1−xN) is a promising ultra-wide bandgap material (6.2–7.4 eV) for next-generation deep-ultraviolet optoelectronics and high-power electronics, offering tunable properties and compatibility with III-nitride semiconductors. However, the design of heterojunction devices using these alloys is hindered by a lack of systematic knowledge regarding their band alignments. Experimental determination is challenging due to difficulties in synthesizing high-purity, phase-pure, and high-crystallinity BxAl1−xN samples. Furthermore, existing theoretical studies are limited; previous work often relies on empirical solid-state energy (SSE) approaches that neglect surface effects, which are critical for polar materials where spontaneous polarization induces significant electric fields. Consequently, the band alignments for BxAl1−xN across the full composition range (x = 0 to 1) and for both polar (c-plane) and nonpolar (a-plane) orientations remain largely unexplored.

Methodology
The authors employ a first-principles computational framework combining Density Functional Theory (DFT) and many-body perturbation theory (specifically the GW0 method) to compute valence and conduction band alignments.

• Structural Models: The study investigates 17 ground-state structures of BxAl1−xN, along with GaN, AlN, and wurtzite BN (w-BN). To probe the low-boron regime, new models for x = 0.083 and x = 0.125 were generated based on lowest formation energy configurations.
• Surface Simulation: A two-step process is utilized to determine band edges relative to the vacuum level. First, bulk valence band maximum (VBM) and conduction band minimum (CBM) energies are calculated. Second, finite-slab models are simulated to align the macroscopic average potential with the vacuum energy.
• Polarization Handling: A critical methodological advancement involves the use of a novel passivation scheme (Yoo et al.) for polar c-plane slabs. This scheme employs pseudohydrogen (psH) atoms with modified valency dependent on spontaneous polarization ($P_s$) to neutralize the electric fields generated by polarization-induced charge displacement, a long-standing challenge in computing band alignments for polar III-nitrides.
• Computational Details: Calculations were performed using the VASP package with the PBE functional and PAW formalism. GW0 corrections were applied to PBE bandgaps to ensure accuracy. A total of 104 distinct surface configurations (c-plane N-surfaces, c-plane M-surfaces, and a-plane surfaces) were analyzed.

Key Results

• Polar (c-plane) Alignments:
  • N-surfaces: For low boron compositions (x < 0.333), BxAl1−xN/AlN heterostructures exhibit near-zero valence band offsets (VBO), consistent with the common-anion rule. As x increases to intermediate ranges (0.333 ≤ x ≤ 0.667), the VBM rises above that of AlN, correlating with reduced tetrahedrality (bond distortion) in the bulk structures. For high x (x > 0.667), the VBM drops below that of AlN due to high boron content and restored tetrahedrality.
  • M-surfaces: The electronic properties of M-terminated (metal-terminated) surfaces differ drastically from N-surfaces. The GaN M-surface VBM and CBM are approximately 1.62 eV higher than the N-surface, leading to a predicted AlN/GaN VBO of 1.85 eV, significantly higher than the 0.44 eV found for N-surfaces.
  • Composition Dependence: The band alignments show strong nonlinearity with respect to boron fraction, particularly in the intermediate x range, driven by structural tetrahedrality.
• Nonpolar (a-plane) Alignments:
  • The average VBM of a-plane BxAl1−xN decreases roughly linearly with increasing x, while the CBM remains relatively constant due to the widening bandgap.
  • These interfaces generally exhibit Type-I band alignments with AlN. However, the results show large standard deviations (up to 1.4 eV) for the same composition, attributed to surface stoichiometry and bond distortions (dipoles) near the surface.
  • High-x a-plane surfaces can exhibit negative electron affinity, suggesting potential for spontaneous electron emission.
• Comparison with Experiment and SSE: The computed offsets show good agreement with available experimental data (e.g., AlN/GaN VBO of 0.44 eV matches XPS measurements). When compared to the empirical SSE approach, the first-principles surface-dependent calculations reveal significant discrepancies, particularly for B-containing compounds, highlighting the limitations of bulk-only empirical models in predicting surface-dependent band offsets.

Significance and Claims
The authors claim this work provides the first surface-dependent predictions of BxAl1−xN band alignments across the complete composition range using first-principles simulations. By addressing the challenges of spontaneous polarization in polar slabs and accounting for surface stoichiometry in nonpolar slabs, the study offers essential design guidelines for integrating BxAl1−xN into advanced semiconductor technologies.

Specifically, the results suggest that:

1. BxAl1−xN/AlN heterostructures with low boron content are well-suited for high-electron-mobility transistors (HEMTs) and ultraviolet light-emitting diodes (LEDs) due to their low valence band offsets and high conduction band offsets.
2. Band alignment tuning is achievable by modulating the boron fraction and controlling interfacial stoichiometry, particularly in Type-I and Type-II heterostructures.
3. Surface polarity and structural distortion are critical factors that cannot be ignored; the study demonstrates that neglecting surface effects (as in SSE models) leads to inaccurate predictions for these materials.

The paper concludes that these findings represent a foundational step toward establishing the performance of BxAl1−xN-based heterojunction devices, moving beyond empirical estimates to rigorous, surface-aware theoretical predictions.