betaFeedback

WorksPhysics and Astronomy

Time inversion symmetry in the Dirac and Schrödinger-Pauli theories

Abstract

The Schrödinger-Pauli theory is generally believed to give a faithful representation of the nonrelativistic and weakly relativistic limit of the Dirac theory. However, the Schrödinger-Pauli theory is fundamentally incomplete in its account of broken time inversion symmetry, e.g., in magnetically ordered systems. In the Dirac theory of the electron, magnetic order breaks time inversion symmetry even in the nonrelativistic limit, whereas time inversion symmetry is effectively preserved in the Schrödinger-Pauli theory in the absence of spin-orbit coupling. In the Dirac theory, the Berry curvature 1/(2m2c2)1/(2m^2c^2) is thus an intrinsic property of nonrelativistic electrons similar to the well-known spin magnetic moment eℏ/(2m)e\hbar/(2m), while this result is missed by the nonrelativistic or weakly relativistic Schrödinger-Pauli equation. In ferromagnetically ordered systems, the intrinsic Berry curvature yields a contribution to the anomalous Hall conductivity independent of spin-orbit coupling.

While the famous Dirac equation provides a relativistically invariant formulation of quantum mechanics, the Schrödinger-Pauli equation with a Zeeman term and a spin-orbit coupling (SOC) term is generally believed to properly represent the nonrelativistic or weakly relativistic limit of the Dirac equation [1]. Time inversion symmetry (TIS) is a fundamental symmetry of nature that is broken by magnetic order. Broken TIS represents the fundamental cause for all physical phenomena that distinguish magnetic systems from nonmagnetic systems [2]. This includes, e.g., the anomalous Hall effect [3] and the magnetoelectric effect [2].

This Letter demonstrates that the Dirac theory accounts for the breaking of TIS in magnetically ordered structures even in the nonrelativistic limit and without SOC. The distinction between systems preserving TIS and systems breaking TIS is lost in the nonrelativistic Schrödinger theory; and it can only be re-introduced in the weakly relativistic Pauli theory via SOC. Our findings are relevant in the context of recent Schrödinger-Pauli theories of magnetic order distinguishing between nonrelativistic magnetic phenomena arising in the absence of SOC and phenomena that do require SOC [4, 5].

For conceptual simplicity, we use a 2D model to illustrate the qualitative differences between the Dirac theory and Schrödinger-Pauli theory when applied to nonrelativistic magnetic electron systems. Ignoring the zz component of motion, the 4×44 \times4 Dirac Hamiltonian becomes

H4×4=(Δ+00cp−0Δ−cp+00cp−−Δ−0cp+00−Δ+),(1)H_{4\times4}=\begin{pmatrix} \Delta^{+} & 0 & 0 & cp_{-} \\ 0 & \Delta^{-} & cp_{+} & 0 \\ 0 & cp_{-} & -\Delta^{-} & 0 \\ cp_{+} & 0 & 0 & -\Delta^{+} \end{pmatrix}, \tag*{(1)}

where p±=px±ipyp_{\pm}=p_x\pm ip_y and Δ±=mc2±δ\Delta^{\pm}=mc^{2}\pm\delta. The parameter δ≪mc2\delta\ll mc^{2} represents a simple model for an exchange field arising from magnetic order. We ignore any potential VV. By regrouping rows and columns appropriately, the Hamiltonian (1) can be rewritten in unitarily equivalent, block-diagonal form as

H~4×4=(H2×2+00H2×2−)(2)\widetilde{H}_{4\times4}=\begin{pmatrix} H^{+}_{2\times2} & 0 \\ 0 & H^{-}_{2\times2} \end{pmatrix} \tag*{(2)}

with

H2×2±=(Δ±cp∓cp±−Δ±).(3)H^{\pm}_{2\times2}=\begin{pmatrix} \Delta^{\pm} & cp_{\mp} \\ cp_{\pm} & -\Delta^{\pm} \end{pmatrix}. \tag*{(3)}

Here H2×2±H^{\pm}_{2\times2} are decoupled 2×22 \times2 Hamiltonians for spin up and spin down. In the limit δ→0\delta\to0, we have (H2×2±)∗=H2×2∓≠H2×2±(H^{\pm}_{2\times2})^{*}=H^{\mp}_{2\times2}\ne H^{\pm}_{2\times2}, i.e., the decoupled Hamiltonians are inherently complex, indicating that the Hamiltonians H2×2±H^{\pm}_{2\times2} individually always break TIS. Thus, very generally, the (backward) unitary time evolution under (H2×2±)∗(H^{\pm}_{2\times2})^{*} cannot annul the (forward) unitary time evolution under H2×2±H^{\pm}_{2\times2}, as it is the case in systems preserving TIS.

Unlike the Schrödinger-Pauli theory, the Dirac theory does not explicitly contain a magnetic moment of the electron. But the electrons’ orbital motion gives rise to an orbital magnetic moment of the eigenstates of H2×2±H^{\pm}_{2\times2} that can be evaluated with the modern theory of orbital magnetization [6, 7] [8], yielding (here for the positive-energy eigenstates ∣ψ±⟩|\psi^{\pm}\rangle)

μ±(p)=±ieℏ2[∇pψ±]×[H2×2±−E±(p)][∇pψ±]z(4)\mu^{\pm}(p)=\pm\frac{i e\hbar}{2}\left[\nabla_p\psi^{\pm}\right]\times\left[H^{\pm}_{2\times2}-E^{\pm}(p)\right]\left[\nabla_p\psi^{\pm}\right]_z \tag*{(4)}
=±eℏc2Δ±2(Δ±2+c2p2)≈±eℏmc42(m2c4+c2p2).(5)=\pm\frac{e\hbar c^{2}\Delta^{\pm}}{2\left(\Delta^{\pm2}+c^{2}p^{2}\right)}\approx\pm\frac{e\hbar mc^{4}}{2\left(m^{2}c^{4}+c^{2}p^{2}\right)}. \tag*{(5)}

As expected, the p→0p\to0 limit of the orbital magnetic moment in the Dirac theory equals the well-known spin magnetic moment

μs±=±eℏ2m(6)\mu_s^{\pm}=\pm\frac{e\hbar}{2m} \tag*{(6)}

in the nonrelativistic Schrödinger-Pauli theory [9, 10]. Similarly, we can evaluate for the eigenstates ∣ψ±⟩\lvert\psi^{\pm}\rangle of H2×2±\mathcal{H}^{\pm}_{2\times2} the Berry curvature [6, 11] [8]

Ω±(p)=i⟨∇pψ±|×|∇pψ±⟩z(7)\Omega^{\pm}(\boldsymbol{p})=i\left\langle\nabla_{\boldsymbol{p}}\psi^{\pm}\middle|\times\middle|\nabla_{\boldsymbol{p}}\psi^{\pm}\right\rangle_z \tag*{(7)}
=±c2Δ±2(Δ±2+c2p2)3/2≈±mc42(m2c4+c2p2)3/2(8)=\pm\frac{c^2\Delta^{\pm}}{2\left(\Delta_{\pm}^{2}+c^2p^2\right)^{3/2}}\approx\pm\frac{mc^4}{2\left(m^2c^4+c^2p^2\right)^{3/2}} \tag*{(8)}

that determines the anomalous Hall conductivity [3]. In the nonrelativistic limit p→0p\to0, we obtain [12]

Ω±(0)=±c22Δ±2≈±12m2c2.(9)\Omega^{\pm}(0)=\pm\frac{c^2}{2\Delta^{\pm2}}\approx\pm\frac{1}{2m^2c^2}. \tag*{(9)}

The Berry curvature (9) is an intrinsic property of the nonrelativistic electron like the famous spin magnetic moment (6). An imbalance Δn\Delta n between spin-up and -down states in a ferromagnet thus implies, besides a net orbital magnetic moment, a net intrinsic anomalous Hall conductivity ∼(e2/ℏ)(ℏ/mc)2Δn/2\sim(e^2/\hbar)(\hbar/mc)^2\Delta n/2. Using typical numbers, this simple estimate yields conductivities that are too small to explain experimentally observed values, but it illustrates the fundamental difference between models preserving TIS and models breaking TIS. Different from earlier work [3], SOC due to a potential VV is not necessary for this contribution to the anomalous Hall conductivity.

In the nonrelativistic Schrödinger theory, the Hamiltonians H2×2±\mathcal{H}^{\pm}_{2\times2} are replaced by

H1×1±=p22m±δ.(10)\mathcal{H}^{\pm}_{1\times1}=\frac{p^2}{2m}\pm\delta. \tag*{(10)}

These Hamiltonians are real, (H1×1±)∗=H1×1±(\mathcal{H}^{\pm}_{1\times1})^*=\mathcal{H}^{\pm}_{1\times1}, i.e., they preserve TIS individually even in the presence of an exchange-field δ\delta—as noted previously in the context of electronic-structure calculations for magnetic systems in, e.g., Refs. [4, 13, 14]; see also Ref. [15] reviewing the general context. This is closely related to the well-known fact that nondegenerate eigenfunctions of Hamiltonians H1×1±\mathcal{H}^{\pm}_{1\times1} are real upto an overall phase factor. Beyond that, TIS of spin-decoupled Schrödinger-type 1×11\times1 crystal Hamiltonians H1×1±\mathcal{H}^{\pm}_{1\times1} [15] implies that the Bloch eigenfunctions of H1×1±\mathcal{H}^{\pm}_{1\times1} can be classified according to Herring’s cases (a), (b), and (c) [16], which is a unique property of crystal Hamiltonians preserving TIS. This happens even if these decoupled Hamiltonians H1×1±\mathcal{H}^{\pm}_{1\times1} ultimately represent opposite spin channels in (collinear) magnetically ordered structures. Therefore, such Hamiltonians cannot account for magnetic phenomena like the Berry curvature (5) and the associated anomalous Hall conductivity.

A nonzero Berry curvature representing the breaking of TIS is usually re-introduced into the weakly relativistic Pauli theory via SOC (arising from gradients of a potential VV) that represents an off-diagonal coupling between the blocks H1×1+\mathcal{H}^{+}_{1\times1} or H1×1−\mathcal{H}^{-}_{1\times1} [3]. Unlike the intrinsic Berry curvature (5), the curvature in the Pauli theory thus depends on the potential VV.

Similar to the anomalous Hall effect, it has previously been assumed that microscopic theories of the magnetoelectric effect require SOC [17]. The magnetoelectric effect exists in materials that break both TIS and space inversion symmetry [2]. In fact, such microscopic theories can again be developed independent of SOC as long as the broken inversion symmetries are otherwise taken into account. However, in this case, the necessary details are more intricate [18], and a full discussion of these will be presented elsewhere.

In conclusion, a proper account of magnetic phenomena is naturally achieved in electronic-structure calculations based on the Dirac theory where TIS breaking manifests itself via orbital equilibrium currents. Such fully relativistic calculations of magnetic systems, though less common than calculations based on the Schrödinger-Pauli theory, have been reported, e.g., in Refs. [19, 20]. However, such a description does not lend itself to a decoupling of real-space order and magnetic order that is assumed, e.g., in applications of spin-group theories [4, 5, 21].

Acknowledgments

RW and UZ acknowledge discussions with D. Culcer, T. Jungwirth, J. Sinova, and L. Šmejkal. RW also acknowledges discussions with G. Bihlmayer and D. Vanderbilt. Work at Argonne was supported by DOE BES under Contract No. DE-AC02-06CH11357.

References

References

  1. [1]P. Strange, Relativistic Quantum Mechanics (Cambridge University Press, Cambridge, 1998).
  2. [2]L. D. Landau and E. M. Lifshitz, Electrodynamics of Continuous Media, 2nd ed. (Pergamon, Oxford, 1984).
  3. [3]N. Nagaosa, J. Sinova, S. Onoda, A. H. MacDonald, and N. P. Ong, Anomalous Hall effect, Rev. Mod. Phys. 82, 1539 (2010).DOI
  4. [4]L. Šmejkal, J. Sinova, and T. Jungwirth, Beyond conventional ferromagnetism and antiferromagnetism: A phase with nonrelativistic spin and crystal rotation symmetry,
  5. [5]L. Šmejkal, J. Sinova, and T. Jungwirth, Emerging research landscape of altermagnetism, Phys. Rev. X 12, 040501 (2022).arxiv.org/abs/2204.10844
  6. [6]D. Xiao, W. Yao, and Q. Niu, Valley-contrasting physics in graphene: Magnetic moment and topological transport, Phys. Rev. Lett. 99, 236809 (2007).
  7. [7]R. Resta, Electrical polarization and orbital magnetization: the modern theories, J. Phys.: Condens. Matter 22, 123201 (2010). In Eqs. (3a) and (5a), the bra and ket are three-component vectors, and the symbol “×” denotes the vector product of such vectors.DOI
  8. [9]Footnote [12] in J. Shi, G. Vignale, D. Xiao, and Q. Niu, Quantum theory of orbital magnetization and its generalization to interacting systems, Phys. Rev. Lett. 99, 197202 (2007).
  9. [10]Appendix C in Ref. [18].
  10. [11]D. Culcer, A. H. MacDonald, and Q. Niu, Anomalous Hall effect in paramagnetic two-dimensional systems, Phys. Rev. B 68, 045327 (2003).DOI
  11. [12]M.-C. Chang and Q. Niu, Berry curvature, orbital moment, and effective quantum theory of electrons in electromagnetic fields, J. Phys.: Condens. Matter 20, 193202 (2008).DOI
  12. [13]D. Gosálbez-Martínez, I. Souza, and D. Vanderbilt, Chiral degeneracies and Fermi-surface Chern numbers in bcc Fe, Phys. Rev. B 92, 085138 (2015).
  13. [14]L.-D. Yuan, Z. Wang, J.-W. Luo, E. I. Rashba, and A. Zunger, Giant momentum-dependent spin splitting in centrosymmetric low-Z antiferromagnets, Phys. Rev. B 102, 014422 (2020).DOI
  14. [15]G. Bihlmayer, Density functional theory of magnetism, in Handbook of Magnetism and Advanced Magnetic Materials, Vol. I, edited by H. Kronmüller and S. Parkin (Wiley, Chichester, UK, 2007) pp. 1–26.DOI
  15. [16]C. Herring, Effect of time-reversal symmetry on energy bands of crystals, Phys. Rev. 52, 361 (1937).DOI
  16. [17]S. Dong, H. Xiang, and E. Dagotto, Magnetoelectricity in multiferroics: a theoretical perspective, Natl. Sci. Rev. 6, 629 (2019).DOI
  17. [18]R. Winkler and U. Zülicke, Collinear orbital antiferromagnetic order and magnetoelectricity in quasi-two-dimensional itinerant-electron paramagnets, ferromagnets, and antiferromagnets, Phys. Rev. Res. 2, 043060 (2020).
  18. [19]S. Lowitzer, D. Ködderitzsch, and H. Ebert, Coherent description of the intrinsic and extrinsic anomalous Hall effect in disordered alloys on an ab initio level, Phys. Rev. Lett. 105, 266604 (2010).DOI
  19. [20]D. Wangenknecht, L. Šmejkal, Z. Kašpar, J. Sinova, T. Jungwirth, J. Kudrnovský, K. Carva, and I. Turek, Temperature-dependent resistivity and anomalous Hall effect in NiMnSb from first principles, Phys. Rev. B 99, 174433 (2019).
  20. [21]D. B. Litvin and W. Opechowski, Spin groups, Physica 76, 538 (1974).DOI

Paper details

vlnc:W_9k25s578
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents