Systematic comparison of approximations and functionals in first-principle calculations of aluminum-based III-V ferroelectric nitrides
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Condensed Matter. The day in condensed matter: materials, phases and the measurements under them.
Kai: We're looking at the paper: "Systematic comparison of approximations and functionals in first-principle calculations of aluminum-based III-V ferroelectric nitrides".
Paper summary: Kai: So wrapping up this paper, "Systematic comparison of approximations and functionals in first-principle calculations of aluminum-based III-V ferroelectric nitrides", what's the bigger picture for us here >
Mira: It's about establishing a consistent and predictive framework for these nitride ferroelectrics by systematically comparing different ways to do those first-principles calculations >
Lev: It means that if we are designing new materials, we have to be very careful about which approximations and functionals we pick because they aren't all giving us the same results >
Kai: The authors point out that alloying with Sc and B drives qualitatively different structural changes, which is driven by whether you increase or decrease coordination tendencies >
Mira: They establish clear structure-property relationships here, offering guidance for designing next-generation nitride ferroelectrics with tailored functionalities based on those specific coordination effects >
Lev: For the experimental side of things, this means we need to be more precise in our material synthesis because the theoretical predictions about phase stability and polarization are quite sensitive to these alloying elements >
Kai: The paper suggests that SQS combined with SCAN is the best choice among all studied combinations based purely on energetic considerations for getting a consistent treatment >
Mira: That's the main recommendation, but it also shows that even with that combination, there are still gaps between theory and experiment, especially regarding critical concentrations of phase transitions >
Lev: So what they didn't cover, or where they flag limitations is that this is all within a unified forty-eight-atom supercell framework > <ref:2605.24334#pg1,within a unified 48-atom>
Kai: Right. It’s a limitation on the structural complexity; it’s not testing infinite size systems, which limits how much detail we can get from the simulation >
Mira: And another thing is that their results are based on specific DFT calculations, so they are still fundamentally limited by the underlying exchange correlation functional chosen >
Lev: So to summarize, this paper is a strong piece of work because it gives us a way to compare these methods and tells us which combination of SQS and SCAN provides the most consistent theoretical treatment for these materials >
Conclusion: Kai: So we've compared different math methods for these aluminum nitride ferroelectrics and found that SQS combined with SCAN seems to be the most reliable way to get a consistent picture of what's happening on the atomic level.
Mira: Right, so this paper is really digging into how different ways of calculating things—the approximations and the functionals—affect our predictions for these materials, Al one x sc one x b one nitride.
Lev: From an error correction standpoint, that consistency is huge because if we can't trust the underlying calculation method itself, then all our simulated hardware designs based on those calculations are just guessing.
Kai: It’s about making sure the theoretical framework actually matches what you see when you build and measure these things in the lab.
Mira: The implication here is that for designing new nitride ferroelectrics, we need to be very careful about picking our tools because they aren't all giving us the same results.
Lev: Specifically, this means we have a better idea of which calculations are actually worth spending time on before trying to run them on real hardware where the overhead gets heavy.
Kai: The authors also point out that even with their best combination, like SQS and SCAN, there's still a gap between what the theory predicts and what experiments actually show about things like phase stability.
Mira: They found that sometimes PBE or PBESol functionals make big quantitative differences even when they look similar on the surface.
Lev: That’s a warning for us; it means we can’t just trust that one functional blindly without checking it against other methods, especially when we're dealing with phase transitions like going from wurtzite to rocksalt.
Kai: So the main point of this paper is to show that we need a rigorous comparison of these calculation techniques before we can say what the structural and electronic properties really are for these alloys.
Mira: And that’s why figuring out which approximation and functional combination gives you the most consistent, predictable treatment is so important for moving forward in this area.
Lev: It sets a baseline for us on what the best starting point is before we try to push those calculations toward experimental reality, which will be our next big hurdle.
Smart Ferroic Materials center, Institute for Nanoscience & Engineering and Department of Physics, University of Arkansas · Department of Physics, Guangdong Technion – Israel Institute of Technology · Department of Metallurgical and Materials Engineering, Colorado School of Mines · Materials Science Center, National Laboratory of the Rockies · Department of Materials Science and Engineering, Tel Aviv University · Université Paris-Saclay, CentraleSupélec, CNRS
cond-mat.mtrl-sci
Submitted: 2026-05-23
Updated: 2026-11-17
License: http://creativecommons.org/licenses/by/4.0/
Importance score: 85/100
The gist: Systematic comparison of approximations and functionals in first-principle calculations of aluminum-based III-V ferroelectric nitrides is a comprehensive study revisiting first-principles predictions…
Key concepts
- SQS
- Special Quasirandom Structures are a method used in first-principles calculations to model chemical disorder. It helps simulate how random alloying elements (like Sc or B) are distributed within the crystal lattice without needing to calculate every possible arrangement, making simulations more realistic for real materials.
- SCAN Functional
- SCAN is an advanced exchange-correlation functional used in density functional theory calculations. The study found that SCAN performs better than other tested functionals (like PBE and PBESol) because it offers a more accurate description of the electronic interactions within these complex nitride materials.
- Ferroelectric Phase Stability
- This refers to how different chemical compositions affect which crystal structure (like wurtzite or rocksalt) is most stable. The study showed that adding Sc or B changes the coordination preferences, leading to distinct structural and polarization responses in AlN alloys.
- Exchange-Correlation Approximations
- These are mathematical approximations within quantum mechanics used to simplify the complex interactions between electrons in a material. The paper compared various approximations (like VCA) to see which ones best capture the true physics of the aluminum-based nitride systems.
Terminology
Summary
Systematic comparison of approximations and functionals in first-principle calculations of aluminum-based III-V ferroelectric nitrides is a comprehensive study revisiting first-principles predictions for structural, ferroelectric, and electronic properties in Al1-xScxN and Al1-xBxN using various approximations and functionals. p. 0 (Dated: May 26, 2026)
We revisit first-principles predictions of structural, ferroelectric, and electronic properties in
aluminum-based III–V nitride alloys, focusing on Al1-xScxN and Al1-xBxN. Using density functional theory within a unified 48-atom supercell framework, we systematically assess the role of
chemical disorder and exchange–correlation approximations by comparing the virtual crystal approximation (VCA) and special quasirandom structures (SQS), as well as PBE, PBESol, SCAN, and
SCAN+rVV10 functionals.Where the paper says it
Methodological Comparison
The study employs a unified 48-atom supercell framework to compare multiple competing polymorphs, including wurtzite, rocksalt, zincblende, and hexagonal phases on an equal footing The researchers systematically assess the role of chemical disorder and exchange–correlation approximations by comparing the virtual crystal approximation (VCA) and special quasirandom structures (SQS), as well as PBE, PBESol, SCAN, and SCAN+rVV10 functionals They demonstrate that even amongst the similar PBE and PBESol functionals, big quantitative and qualitative differences emerge
Structural Phase Stability
The analysis reveals qualitatively different structural and ferroelectric responses driven by increased or decreased coordination tendencies with Sc and B alloying Key findings regarding phase stability include:
-
For Al1-xScxN, the VCA or SQS PBESol strongly underestimate the stability domain of the ferroelectric wurtzite phase compared to SQS PBE or SQS SCAN
-
The 5-fold coordinated hexagonal phase predicted in 2002 is a low-energy metastable state between the four-fold coordinated ferroelectric wurtzite phase and the six-fold coordinated rocksalt phase near the transition point upon increasing the Sc content
-
Al1-xBxN shows a much faster destabilization of the wurtzite ferroelectric phase, with bond breaking which strongly distorts the wurtzite structure (with enhanced polarization) and eventually favor a zincblende phase and a threefold coordinated hexagonal layer phase
Functional Performance
The comparison of exchange-correlation functionals highlights SCAN as superior, with SQS combined with SCAN providing the most consistent theoretical treatment Specific comparisons showed that PBESol significantly underestimates the critical concentration from the WZ to the RS phase compared to experimental reports Furthermore, PBE predicts that the HX phase may be the most stable phase in a very narrow concentration window which has never been observed experimentally
Electronic and Polarization Properties
The study quantifies structural parameters, showing that for Al1-xScxN, the c/a ratio steadily decreases with increasing Sc content in WZ (meanwhile the internal parameter u increases), suggesting qualitatively a decrease in polarization In contrast, Al1-xBxN depicts an opposite trend, that is a slight increase of c/a and an almost constant internal parameter u, which may indicate a stable or increasing spontaneous polarization The calculated spontaneous polarization in Al1-xScxN steadily decreases from 135.1 µC/cm2 in pure AlN to 108. p. 9 In Al1-xScxN, the polarization steadily decreases from
135. 1 µC/cm2
in pure AlN to 108. In Al1-xScxN, the polarization steadily decreases from
135. 1 µC/cm2
in pure AlN to 108.Where the paper says it
Where the paper says it
Bond Distribution and Density of States
Analysis of bond lengths reveals that Sc–N average bond length in Al1-xScxN is about 0.2 Å longer than the Al–N bonds across all four structures The density of states analysis shows that Sc d states start contributing significantly to both the top of the valence bands and the bottom of the conduction bands, indicating that Sc tends to hybridize with nitrogen p states more strongly than Al states In contrast, when doping with B, little modification to the valence bands can be observed The HX and RS phases show a larger bandwidth in their conduction bands compared to the WZ and ZB phases
Conclusion
Overall, the study provides a coherent and quantitatively grounded picture of alloying effects in III–V nitride ferroelectrics, while clarifying the methodological requirements for accurate first-principles predictions These findings offer guidance for the design and optimization of nitride-based ferroelectric materials The SCAN exchange–correlation functionals, in conjunction with Special Quasirandom Structures, seems to be a better choice among all studied combination based on energetic considerations alone
ACKNOWLEDGMENTS
This work is supported by a grant from the U.S. Department of Energy through grant agreement no. p. 6 Surprisingly, and although the c/a ratio is almost identically
well described by VCA and SQS, VCA quantitatively and
qualitatively differs from SQS predictions of the individual lattice constants a and c (see middle and right lower
panels in Fig. In fact, VCA calculations grossly overestimate the a and c lattice constants in (Al,B)N compared to SQS, despite showing good relative agreement
in describing the energy of the WZ phase in Fig. 3Department of Metallurgical and Materials Engineering,
Colorado School of Mines, Golden, Colorado 80401, USA. 4Materials Science Center, National Laboratory of the Rockies, Golden, Colorado 80401, USA. Systematic comparison of approximations and functionals in first-principle
calculations of aluminum-based III-V ferroelectric nitrides
Alejandro Mercado Tejerina,1 Peng Chen,1, 2 Keisuke Yazawa,3, 4
Andriy Zakutayev,4 Laurent Bellaiche,1, 5 and Charles Paillard1, 6
1Smart Ferroic Materials center, Institute for Nanoscience & Engineering and Department of Physics,
University of Arkansas, Fayetteville, AR 72701, USA. 2Department of Physics, Guangdong Technion – Israel Institute of Technology, Shantou, Guangdong Province, China. Systematic comparison of approximations and functionals in first-principle
calculations of aluminum-based III-V ferroelectric nitrides
Alejandro Mercado Tejerina,1 Peng Chen,1, 2 Keisuke Yazawa,3, 4
Andriy Zakutayev,4 Laurent Bellaiche,1, 5 and Charles Paillard1, 6
1Smart Ferroic Materials center, Institute for Nanoscience & Engineering and Department of Physics,
University of Arkansas, Fayetteville, AR 72701, USA. 2Department of Physics, Guangdong Technion – Israel Institute of Technology, Shantou, Guangdong Province, China.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Improved AI systems will be able to predict experimental ferroelectric Wurtzite phase stability windows more accurately by using SCAN exchange-correlation functionals combined with Special Quasirandom Structures (SQS). This is because the paper states,
SQS combined with SCAN provides the most consistent theoretical treatment of emerging nitride-based ferroelectric materials.
-
Improved AI systems will gain better predictive power regarding structural phase transitions in Al1-xScxN by understanding that
the VCA or SQS PBESol (a popular functional) strongly underestimate the stability domain of the ferroelectric wurtzite phase in Al1-xScxN compared to SQS PBE or SQS SCAN.
-
Improved AI systems will be able to distinguish between structural effects driven by chemical disorder and those driven by exchange-correlation approximations when modeling nitride ferroelectrics. This is because the study
highlights the critical importance of both local disorder and exchange–correlation treatment in predicting the functional properties of III–V nitride ferroelectrics.
-
Improved AI systems will be able to predict the qualitative structural response of Al1-xBxN based on boron content by recognizing that "Al1-xBxN shows a much faster destabilization of the wurtzite ferroelectric phase, with bond breaking which strongly distorts the wurtzite structure (with enhanced polarization) and eventually favor a zincblende phase and a threefold coordinated hexagonal layer phase."
-
Improved AI systems will be able to select the most reliable exchange-correlation functional for nitride ferroelectrics by prioritizing SCAN, as it
provides the most consistent description of phase energetics and structural properties.
Abstract
We revisit first-principles predictions of structural, ferroelectric, and electronic properties in aluminum-based III-V nitride alloys, focusing on Al1-xScxN and Al1-xBxN. Using density functional theory within a unified 48-atom supercell framework, we systematically assess the role of chemical disorder and exchange-correlation approximations by comparing the virtual crystal approximation (VCA) and special quasirandom structures (SQS), as well as PBE, PBESol, SCAN, and SCAN+rVV10 functionals. We demonstrate that, even amongst the similar PBE and PBESol functionals, large quantitative and qualitative differences emerge. In particular, VCA or SQS PBESol modeling (a popular functional) underestimates the stability domain of the ferroelectric wurtzite phase in Al1-xScxN compared to SQS PBE or SQS SCAN. We demonstrate that the 5-fold coordinated hexagonal phase predicted in 2002 by Farrer and Bellaiche [Phys. Rev. B 66, 201203] is a metastable state whose energy lies between that of the four-fold coordinated ferroelectric wurtzite phase and the energy of the six-fold coordinated rocksalt phase below the transition point. In contrast, Al1-xBxN shows a much faster destabilization of the wurtzite ferroelectric phase, with bond breaking which strongly distorts the wurtzite structure (with enhanced polarization) and eventually favors a zincblende phase and a threefold coordinated hexagonal layer phase. Our analysis highlights the critical importance of both local disorder and exchange-correlation treatment in the quantitative prediction of the functional properties of III-V nitride ferroelectrics.
Related papers
- Structural contribution to light-induced gap suppression in Ta 2 NiSe 5
- Anatomy of fast current-induced skyrmion motion in synthetic antiferromagnets
- Magnetic control of an exciton-polariton condensate in a van der Waals magnet
- High-pressure elastic properties of GeO2 polymorphs up to 120 GPa
- Decoupling Strain-Rate Sensitivity and Deformation Length Scale Effects in Neutron-Irradiated Tungsten: A Coupled Nano-Indentation, HR-EBSD and Crystal Plasticity Study
- Magnetism modulation in Co3Sn2S2 by current-assisted domain wall motion