Role of zero-point effects in stabilizing the ground state structure of bulk Fe2P

Soumya S. Bhat, Kapil Gupta, Satadeep Bhattacharjee, Seung‐Cheol LeeJournal of Physics Condensed Matter 2018 · DOI · arXiv

Abstract

The structural stability of $\text{Fe}_2\text{P}$ is investigated in detail using first-principles calculations based on density functional theory. While the orthorhombic C23 phase is found to be energetically more stable in standard calculations, experiments suggest it to be the hexagonal C22 phase. In the present study, we show that in order to obtain the correct ground state structure of $\text{Fe}_2\text{P}$ from first-principles based methods, it is necessary to consider zero-point effects such as zero-point vibrations and spin fluctuations. This study demonstrates an exceptional case where a bulk material is stabilized by quantum effects, which are usually important only in low-dimensional materials. Our results also indicate the possibility of a magnetic field-induced structural quantum phase transition in $\text{Fe}_2\text{P}$, which should form the basis for further theoretical and experimental efforts.

Introduction

In materials science, Density Functional Theory (DFT) is the "gold standard" for predicting crystal structures. However, for $\text{Fe}_2\text{P}$—a material prized for its magnetic and catalytic properties—there is a persistent "theory-experiment gap." Experiments show that at room temperature and pressure, $\text{Fe}_2\text{P}$ adopts a hexagonal structure (the C22 phase). Yet, standard DFT calculations consistently predict that an orthorhombic structure (the C23 phase) should be more stable. While zero-point effects (quantum fluctuations that exist even at absolute zero) are known to stabilize low-dimensional materials like graphene, they are usually ignored in bulk 3D materials. This study investigates whether these quantum nuances are the missing link in explaining the stability of $\text{Fe}_2\text{P}$.

Methodology

The researchers employed a multi-layered theoretical approach using the VASP code with the PBE-GGA functional to calculate the electronic structure and basic total energies of both the C22 and C23 phases. To account for quantum corrections, they performed:

  • Phonon Calculations: Using the finite-displacement method (via the Phonopy code) to determine the vibrational spectrum and calculate the Zero-Point Vibrational Energy ($E_{ZPV}$).
  • Spin Fluctuation Analysis: Accounting for Zero-Point Spin Fluctuations ($E_{ZPSF}$) using a Ginzburg-Landau expansion based on magnetic susceptibility calculated via fixed-moment DFT.

Results and Discussion

The Ground State Discrepancy

Standard DFT calculations find the C23 phase to be more stable than the C22 phase by approximately 10.4 meV/atom. This discrepancy is consistent across various exchange-correlation functionals, highlighting a systemic failure of static lattice approximations for this material.

Energy-volume curves for the C22 and C23 phases. The inset shows the enthalpy difference as a function of pressure, indicating that pressure alone does not favor the C22 phase.

As shown in Figure 1, the red dashed curve representing the C23 phase remains lower in energy than the black C22 curve. Furthermore, the inset shows that the enthalpy difference ($\Delta H$) remains negative, suggesting that hydrostatic pressure does not induce a phase transition to the experimentally observed hexagonal structure at 0 K.

Stabilization via Quantum Effects

The breakthrough in this study comes from the inclusion of zero-point corrections. While vibrational effects ($E_{ZPV}$) favor the C22 phase by 4.6 meV/atom, this is insufficient to overcome the initial 10.4 meV/atom gap. The decisive factor is the zero-point spin fluctuations, which provide a massive 23.6 meV/atom stabilization to the hexagonal C22 phase.

Temperature dependence of (a) vibrational free energy and (b) total Helmholtz free energy (including vibrational but not spin-fluctuation corrections).

Figure 3(b) illustrates that even with vibrational free energy included, the C23 phase (red dashed line) remains slightly lower in energy than the C22 phase (black line) at low temperatures. It is only with the addition of the 23.6 meV/atom spin-fluctuation correction that the hexagonal C22 phase becomes the true ground state, matching experimental observations with a total energy advantage of ~17.8 meV/atom.

Conclusion

This study demonstrates that $\text{Fe}_2\text{P}$ is a rare case where a 3D bulk material is stabilized by quantum zero-point effects, particularly spin fluctuations. This challenges the assumption that such effects are only relevant in 2D or light materials. Furthermore, the distinct magnetic signatures of the two phases (3.01 $\mu_B$/f.u. for C22 vs. 1.75 $\mu_B$/f.u. for C23) suggest the possibility of a magnetic field-induced structural quantum phase transition, offering a new avenue for experimental research.

Summary updated

Abstract

Structural stability of Fe<sub>2</sub>P is investigated in detail using first-principles calculations based on density functional theory. While the orthorhombic C23 phase is found to be energetically more stable, the experiments suggest it to be hexagonal C22 phase. In the present study, we show that in order to obtain the correct ground state structure of Fe<sub>2</sub>P from the first-principles based methods it is utmost necessary to consider the zero-point effects such as zero-point vibrations and spin fluctuations. This study demonstrates an exceptional case where a bulk material is stabilized by quantum effects, which are usually important in low-dimensional materials. Our results also indicate the possibility of magnetic field induced structural quantum phase transition in Fe<sub>2</sub>P, which should form the basis for further theoretical and experimental efforts.
BibTeX
@article{bhat2018role,
  author = {Soumya S. Bhat and Kapil Gupta and Satadeep Bhattacharjee and Seung‐Cheol Lee},
  title = {{Role of zero-point effects in stabilizing the ground state structure of bulk Fe<sub>2</sub>P}},
  journal = {Journal of Physics Condensed Matter},
  year = {2018},
  doi = {10.1088/1361-648x/aabe52},
}
Figure