🔬 materials science

First Principles Thermodynamics of Zr B Segregation at Grain Boundaries in Recycled Nd2Fe14B

This study employs first-principles DFT+U calculations to demonstrate that both Zr and B exhibit a strong thermodynamic preference for grain boundaries in recycled Nd2Fe14B magnets, establishing a pathway for the co-localization of these elements that stabilizes the intergranular microstructure and supports high coercivity.

Original authors: Avik Mahata, Miha Zakotnik

Published 2026-09-09

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Avik Mahata, Miha Zakotnik

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Permanent magnets are the invisible engines behind much of modern life, powering everything from the motors in electric cars to the turbines that generate wind energy. At the heart of the most powerful of these magnets is a material called neodymium-iron-boron. While the main body of the magnet is made of large crystals of this material, the secret to its strength lies in the thin, invisible boundaries where these crystals meet. These boundaries act like the mortar between bricks; if they are weak or disordered, the entire structure loses its ability to resist demagnetization, especially when heated. In recent years, scientists have discovered a way to recycle old magnets by breaking them down and reassembling them, but this process often scrambles the delicate arrangement of these boundaries, leaving the new magnet weaker than the old one. Researchers have found that adding tiny amounts of zirconium and boron during recycling can help fix this, causing tiny, needle-like particles to form exactly where the crystals meet. These particles seem to hold the structure together, but until now, no one knew exactly why nature chose to build them in those specific spots rather than scattering them randomly throughout the material.

To solve this mystery, a team of researchers turned to a powerful type of computer simulation that allows them to watch atoms behave as if they were real, without needing a physical laboratory. They focused on the specific question of why zirconium and boron, which are present in very small amounts, decide to gather at the edges of the crystals instead of staying mixed in the middle. Using a method that accounts for the complex magnetic behavior of the atoms involved, they built digital models of the magnet's interior and its boundaries. They then tested what would happen if they placed individual atoms of boron, zirconium, and other elements in these different locations. The goal was to measure the energy required to keep these atoms in the middle of a crystal versus holding them at the boundary. In the world of atoms, systems naturally seek the lowest energy state, much like a ball rolling to the bottom of a hill. By calculating these energy levels, the researchers could predict where the atoms would prefer to settle.

The simulations revealed a clear and surprising hierarchy of preferences. The researchers found that boron atoms have a very strong, natural tendency to rush to the boundaries, lowering their energy significantly by doing so. Zirconium atoms showed a similar, though slightly weaker, desire to be at the edges. When the researchers placed both elements together in a configuration that mimics the structure of the tiny particles seen in real magnets, the combination still preferred the boundary. This means that even though the total amount of zirconium in the magnet is tiny—only about 0.1 percent—the material naturally funnels it toward the crystal edges where it can team up with boron. The computer models showed that once these elements meet at the boundary, they form a stable, low-energy arrangement that helps lock the structure in place. This explains how a magnet with very little zirconium can still develop distinct, concentrated particles at its boundaries: the boundary itself acts as a magnet for these specific atoms, pulling them out of the bulk material and gathering them together.

The study also looked at how these atomic changes affect the magnet's overall strength. A major concern in magnet science is that adding foreign elements might disrupt the magnetic alignment of the iron atoms, which would weaken the magnet. However, the simulations showed that when zirconium and boron gather at the boundary, they do not disturb the magnetic order of the surrounding crystals. The iron atoms inside the crystals remain strongly aligned, preserving the magnet's power, while the boundary region becomes a stable, reinforced zone. This finding connects the dots between the microscopic behavior of atoms and the macroscopic performance of the magnet. It suggests that the reason recycled magnets can be made strong again is not just because of the new particles, but because the thermodynamics of the material naturally drive the necessary ingredients to the right place. The boundary acts as a preferred reservoir, collecting the scarce elements and organizing them into a structure that stabilizes the grain boundaries, preventing the crystals from growing too large and losing their magnetic grip.

While the computer models provided a clear picture of why these atoms gather, the researchers noted that the actual formation of the particles involves a complex dance of heat and movement that happens over time, which the simulation did not fully capture. The study focused specifically on the energy preferences that set the stage for this process. It confirmed that the boundary is the energetically favored location for these elements, but it did not calculate the speed at which they move or the exact steps of how the final particles crystallize. The results serve as a fundamental explanation for a phenomenon observed in the lab: why a small amount of zirconium can lead to a large, organized improvement in the magnet's structure. By understanding that the boundary is a thermodynamic sink for these elements, scientists can better design recycling processes that rely on this natural tendency to produce high-quality magnets from old ones, ensuring that the valuable materials inside are not just recovered, but restored to their full potential.

Technical Summary: First-Principles Thermodynamics of Zr–B Segregation at Grain Boundaries in Recycled Nd2Fe14B

Problem Statement
The coercivity and thermal stability of Nd2Fe14B permanent magnets are critically dependent on the chemistry and structure of their grain boundaries (GBs) and intergranular (IG) regions. In recycled magnets, recovering the hard-magnetic Nd2Fe14B phase alone often fails to restore the optimal intergranular microstructure required for high performance. Recent experimental studies on recycled Nd2Fe14B magnets containing approximately 0.1 at.% Zr revealed the emergence of nanoscale ZrB2 precipitates at grain boundaries and triple junctions. These precipitates were associated with the suppression of grain coarsening and enhanced coercivity. However, the experimental data could not explain the atomistic origin of this preferential localization: why do Zr and B accumulate at intergranular regions to form ZrB2-like structures when the bulk Zr concentration is low and uniform? Existing first-principles studies of Nd–Fe–B interfaces have been limited, often focusing on single chemistries, idealized interfaces, or utilizing open-core approximations for rare-earth 4f states, leaving a gap in understanding the coupled thermodynamics of multi-species segregation (specifically B, Zr, and Dy) within a consistent electronic framework.

Methodology
To address this gap, the authors employed spin-polarized density-functional theory with a Hubbard correction (DFT+U) to calculate the atomistic thermodynamics of defect segregation.

Models: The study utilized 136-atom supercells representing both bulk Nd2Fe14B and an idealized $\Sigma1$ (001) rotational grain boundary. The GB model was constructed by rotating the upper half of the bulk cell by 90° about the [001] axis.
Defect Configurations: Five composition-matched bulk/GB pairs were analyzed: pristine, interstitial Boron ( $B_{int}$ ), substitutional Zirconium ( $Zr_{Nd}$ ), substitutional Dysprosium ( $Dy_{Nd}$ ), and a local Zr–B complex ( $Zr$ – $2B$ ) representing a ZrB2-like motif.
Electronic Structure: The calculations used the Vienna Ab initio Simulation Package (VASP) with the PBE generalized-gradient approximation and the rotationally invariant DFT+U formulation (Dudarev et al.) with $U_{eff} = 6$ eV applied to Nd and Dy 4f states. This approach maintained a correlated-electron description across all configurations, avoiding open-core approximations.
Validation: A rigorous magnetic-state consistency check was performed. The study verified that all defect-containing structures remained within the same high-moment Fe-sublattice regime as the pristine reference, ensuring that energy differences reflected true site preferences rather than artifacts of differing magnetic ground states.
Thermodynamic Metrics: Segregation energies ( $E_{seg}$ ) were calculated to determine site preference, and grain-boundary excess energies ( $\gamma_{GB}$ ) were computed to assess interfacial stabilization.

Key Results
The calculations established a clear thermodynamic hierarchy for the segregation of B, Zr, Dy, and Zr–B complexes:

Strong Boron Segregation: Interstitial Boron exhibits the strongest preference for the grain boundary, with a segregation energy of $E_{seg}(B) = -1.59$ eV/defect. The formation energy of B becomes negative at the boundary ($-0.58$ eV/B) compared to positive in the bulk ( $+1.01$ eV/B), indicating the boundary acts as a thermodynamic sink for excess B.
Independent Zirconium Preference: Zr also shows a direct thermodynamic preference for the interface, with $E_{seg}(Zr) = -0.77$ eV/defect.
Stability of Zr–B Coordination: When Zr and B are combined in a ZrB2-like local configuration, the boundary preference is retained ( $E_{seg} = -0.32$ eV/defect), though the driving force is weaker than for isolated species. This indicates that the interface remains favorable as Zr–B coordination develops.
Dysprosium Contrast: In contrast to Zr and B, Dy exhibits a positive segregation energy ( $E_{seg} = +1.90$ eV/defect) in this idealized zero-temperature model, suggesting it is energetically stabilized in the matrix rather than the boundary under these specific conditions.
Interfacial Stabilization: The presence of B, Zr, and the ZrB2-like motif reduces the grain-boundary excess energy relative to the pristine interface (by $-0.33$, $-0.16$, and $-0.20 $J m$ ^{-2}$, respectively), whereas Dy increases it.
Magnetic Compatibility: All boundary-localized defects perturb the normalized magnetization less than their bulk counterparts. The Fe-sublattice moments and spin-resolved d-states remain consistent with the high-moment ground state, confirming that Zr–B segregation does not destabilize the magnetic character of the surrounding grains.

Significance and Claims
The paper claims to provide the first-principles thermodynamic explanation for the experimentally observed redistribution of Zr and B in recycled Nd2Fe14B magnets. The primary contribution is the establishment of a thermodynamic pathway where the grain boundary acts as a preferential reservoir for both Boron and Zirconium. This explains how a low global Zr concentration (~0.1 at.%) can result in locally concentrated ZrB2 precipitates: the boundary energetically favors the accumulation of both species, facilitating their co-localization and subsequent ZrB2-like coordination.

The authors emphasize that these results clarify the "preceding atomistic step" in the microstructural evolution of recycled magnets. While the calculations do not model the kinetics of precipitate growth or the direct calculation of coercivity, they demonstrate that the formation of ZrB2-rich boundary structures is thermodynamically driven by the segregation of constituent elements. This segregation stabilizes the intergranular region and is magnetically compatible with the high-moment Nd2Fe14B phase, providing a microscopic basis for the enhanced microstructural stability and coercivity observed in recycled magnets containing Zr. The study distinguishes this thermodynamic driving force from the complex processing environment (finite temperature, diffusion gradients) that ultimately dictates the final distribution of elements like Dy.

More like this