Skip to main content
Log in

Equivalent Forms of Dirac Equations in Curved Space-times and Generalized de Broglie Relations

  • Particles and Fields
  • Published:
Brazilian Journal of Physics Aims and scope Submit manuscript

Abstract

One may ask whether the relations between energy and frequency and between momentum and wave vector, introduced for matter waves by de Broglie, are rigorously valid in the presence of gravity. In this paper, we show this to be true for Dirac equations in a background of gravitational and electromagnetic fields. We first transform any Dirac equation into an equivalent canonical form, sometimes used in particular cases to solve Dirac equations in a curved space-time. This canonical form is needed to apply Whitham’s Lagrangian method. The latter method, unlike the Wentzel–Kramers–Brillouin method, places no restriction on the magnitude of Planck’s constant to obtain wave packets and furthermore preserves the symmetries of the Dirac Lagrangian. We show by using canonical Dirac fields in a curved space-time that the probability current has a Gordon decomposition into a convection current and a spin current and that the spin current vanishes in the Whitham approximation, which explains the negligible effect of spin on wave packet solutions, independent of the size of Planck’s constant. We further discuss the classical-quantum correspondence in a curved space-time based on both Lagrangian and Hamiltonian formulations of the Whitham equations. We show that the generalized de Broglie relations in a curved space-time are a direct consequence of Whitham’s Lagrangian method and not just a physical hypothesis as introduced by Einstein and de Broglie and by many quantum mechanics textbooks.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others

Notes

  1. More generally, a common conception is that classical physics emerges from quantum mechanics in the limit as Planck’s constant \(\hbar \) approaches zero. However, the limit \(\hbar \rightarrow 0\) “is not well defined mathematically unless one specifies what quantities are to be held constant during the limiting process” [4]. It is interesting to note that the most classical behaving Gaussian wave functions, the coherent states of the ordinary harmonic oscillator, whose expected position and momentum obey classical equations by Ehrenfest’s theorem, do not resemble wave packets in the limit \(\hbar \rightarrow 0\) [5].

  2. Note, however, that the electron’s magnetic moment, predicted by the Dirac equation, is not contained in the wave packet approximation. Indeed, quoting from Ref. [11]: “The uncertainty principle, together with the Lorentz force, prevents spin-up and spin-down electrons from being separated by a macroscopic field of the Stern-Gerlach type.” In practice, wave packet splitting in Stern–Gerlach experiments is only observed using neutral atoms or molecules, which are undisturbed by the Lorentz force [4]. In Section 3, the wave packet approximation is expressed by neglecting in the Lagrangian the variation in the amplitude of the wave function as compared to the variation of its phase. This leads to wave packet equations which do not involve the electron’s magnetic moment.

  3. In the case of a Majorana representation of the Dirac field, the coefficient matrices \(\left( \gamma ^{\mu }, A \right) \) are pure imaginary, and the Dirac equation is real. In this case, the spin connection matrices Γ μ are real.

  4. Because S depends on the space-time point, Schlütert, Wietschorke, and Greiner call S a “local transformation” [6, 7].

  5. The notion of equivalence, here, is somewhat different than the notion used in Ref. [25], where equivalence was applied only to classify the coefficient fields \( \left( \gamma ^{\mu } , A \right) \) without requiring the existence of a map Ψ→S  − 1 Ψ between the solutions of two Dirac equations.

  6. A globally defined generalized Dirac Lagrangian has the form

    $$\label{Lagrangian-intrinsic} L=\ \frac{i}{2} \left [(\psi,{\not}\mathcal{D}\psi)-({\not}\mathcal{D}\psi,\psi)+2im(\psi ,\psi) \right]. $$
    (10)

    where \({\not}\mathcal{D}\) is a Dirac operator acting on the cross sections ψ of a spinor bundle \({\sf E}\) over the space-time, and ( , ) denotes a hermitizing metric on the fibers of \({\sf E}\). See Ref. [20], Sect. 2.1 and the references therein for the precise definitions. Once any coordinate chart of the space-time and any frame field on the spinor bundle have been chosen, one gets the local expression of the global generalized Lagrangian as (9). In particular, in (9) and in the rest of this paper, Ψ is the column vector made with the components of ψ in the chosen frame field on \({\sf E}\). See Ref. [20], Sect. 2.2.

  7. As stated in Section 2, two Dirac equations that are classically equivalent need not be equivalent with respect to their quantum mechanical energy–momentum operators [2527]. Clearly, the Whitham approximation also distinguishes them. A striking example is a Dirac equation with a Mashhoon term [10, 36, 37]. Applying the Whitham approximation directly to a Dirac equation with a Mashhoon term, without first transforming the Dirac field Ψ, does not produce wave packet motion along classical paths.

  8. Note that the probability current J μ = Ψ +  μΨ is invariant under local similarity transformations S defined in (5), so that when transforming a Dirac equation into canonical form, it is only \(J^\mu _s\) and \(J^\mu _c\) which change their form.

  9. The Legendre transformation is its own inverse (see [40], pages 563–565). Thus, an inverse Legendre transformation is also a Legendre transformation.

References

  1. D. Wick, The Infamous Boundary: Seven Decades of Controversy in Quantum Physics, (Birkhäuser, Boston, 1995), pp. 24 and 50−53

    Book  Google Scholar 

  2. M.P. Silverman, More than One Mystery: Explorations in Quantum Interference, (Springer, New York, 1995), p. 1–8

    Book  Google Scholar 

  3. J. Audretsch, Trajectories and spin motion of massive spin-1/2 particles in gravitational fields, J. Phys. A: Math. Gen. 14, 411–422 (1981)

    Article  ADS  Google Scholar 

  4. L.E. Ballentine, Quantum Mechanics: A Modern Development (World Scientific, Singapore 2001), pp. 230 and 388–390

    Google Scholar 

  5. D. Park, Classical Dynamics and its Quantum Analogues (Springer, Berlin, 1990), pp. 49–51

    Book  MATH  Google Scholar 

  6. P. Schlüter, K.H. Wietschorke, W. Greiner, The Dirac equation in orthogonal coordinate systems: I . The local representation, J. Phys. A: Math. Gen. 16, 1999–2016 (1983)

    Article  ADS  MATH  Google Scholar 

  7. P. Schlüter, K.H. Wietschorke, W. Greiner, The Dirac equation in orthogonal coordinate systems: II . The two center Dirac equation, J. Phys. A: Math. Gen. 16, 2017—2034 (1983)

    Article  ADS  MATH  Google Scholar 

  8. G.C. McVittie, Dirac Equation in General Relativity, R. Astron. Soc, 92, 868–877 (1932)

    ADS  Google Scholar 

  9. G.B. Whitham, Linear and Nonlinear Waves, (Wiley-Interscience, New York, 1995), pp. 363–402

    Google Scholar 

  10. F.W. Hehl, W.T. Ni, Inertial effects of a Dirac particle, Phys. Rev. D 42, 2045–2048 (1990)

    Article  ADS  Google Scholar 

  11. J. Kessler, Polarized Electrons, (Springer, Berlin, 1976), pp. 2–6

    Google Scholar 

  12. M. Arminjon, On the relation Hamiltonian − wave equation, and on non-spreading solutions of Schrödinger’s equation, Il Nuovo Cimento 114B, 71–86 (1999)

    ADS  Google Scholar 

  13. W. Pauli, Contributions mathématiques à la théorie des matrices de Dirac, Ann. Inst. Henri Poincaré 6, 109–136 (1936)

    MathSciNet  Google Scholar 

  14. W. Kofink, Zur Mathematik der Diracmatrizen: Die Bargmannsche Hermitisierungsmatrix A and die Paulische Transpositionsmatrix B, Math. Z. 51, 702–711 (1949)

    Article  MathSciNet  MATH  Google Scholar 

  15. H. Weyl, Elektron und Gravitation, Z. Phys. 56, 330–352 (1929)

    Article  ADS  MATH  Google Scholar 

  16. V.A. Fock, Geometrisierung der Diracschen Theorie des Elektrons, Z . Phys. 57, 261–277 (1929)

    Article  ADS  MATH  Google Scholar 

  17. D.R. Brill, J.A. Wheeler, Interaction of Neutrinos and Gravitational Fields, Rev. Mod. Phys. 29, 465–479 (1957)

    Article  MathSciNet  ADS  MATH  Google Scholar 

  18. T.C. Chapman, D.J. Leiter, On the generally covariant Dirac equation, Am. J. Phys. 44, 858–862 (1976)

    Article  MathSciNet  ADS  Google Scholar 

  19. M. Arminjon, Dirac-type equations in a gravitational field, with vector wave function, Found. Phys. 38, 1020–1045 (2008). arXiv:gr-qc/0702048v4

    Article  MathSciNet  ADS  MATH  Google Scholar 

  20. M. Arminjon, F. Reifler, Four-vector vs. four-scalar representation of the Dirac wave function, Int. J. Geom. Methods Mod. Phys. 9(4), 1250026 (2012) [23 pages]. arXiv:1012.2327v2 [gr-qc]

    Article  MathSciNet  Google Scholar 

  21. M. Arminjon, F. Reifler, Basic quantum mechanics for three Dirac equations in a curved spacetime, Braz. J. Phys. 40, 242—255 (2010). arXiv:0807.0570v3 [gr-qc]

    Article  ADS  Google Scholar 

  22. M. Arminjon, F. Reifler, Dirac Equation: Representation Independence and Tensor Transformation, Braz. J. Phys. 38, 248–258 (2008). arXiv:0707.1829v3 [quant-ph]

    Article  ADS  Google Scholar 

  23. K. Gödel, An Example of a New Type of Cosmological Solutions of Einstein’s Field Equations of Gravitation, Rev. Mod. Phys. 21, 447–450 (1949)

    Article  ADS  MATH  Google Scholar 

  24. M. Arminjon, F. Reifler, General reference frames and their associated space manifolds, Int. J. Geom. Methods Mod. Phys. 8, 155–165 (2011). arXiv:1003.3521v2 [gr-qc]

    Article  MathSciNet  MATH  Google Scholar 

  25. M. Arminjon, F. Reifler, A non-uniqueness problem of the Dirac theory in a curved spacetime, Annalen der Physik 523, 531–551 (2011). arXiv:0905.3686 [gr-qc]

    Article  MathSciNet  ADS  Google Scholar 

  26. M.V. Gorbatenko, V.P. Neznamov, Solution of the problem of uniqueness and Hermiticity of Hamiltonians for Dirac particles in gravitational fields, Phys. Rev. D 82, 104056 (2010). arXiv:1007.4631 [gr-qc]

    Article  ADS  Google Scholar 

  27. M.V. Gorbatenko, V.P. Neznamov, Uniqueness and self-conjugacy of Dirac Hamiltonians in arbitrary gravitational fields, Phys. Rev. D 83, 105002 (2011). arXiv:1102.4067v1 [gr-qc]

    Article  ADS  Google Scholar 

  28. A.O. Barut, I.H. Duru, Exact solutions of the Dirac equation in spatially flat Robertson-Walker space-times, Phys. Rev. D 36, 3705–3711 (1987)

    Article  MathSciNet  ADS  Google Scholar 

  29. G.V. Shishkin, V.M. Villalba, Neutrino in the presence of gravitational fields: Exact solutions, J. Math. Phys. 33, 4037–4045 (1992). arXiv:hep-th/9307061

    Article  MathSciNet  ADS  MATH  Google Scholar 

  30. V.M. Villalba, Dirac spinor in a nonstationary Gödel type cosmological universe, Mod. Phys. Lett. A 8, 3011–3015 (1993). arXiv:gr-qc/9309019

    Article  MathSciNet  ADS  MATH  Google Scholar 

  31. E. Montaldi, A. Zecca, Neutrino Wave Equation in the Robertson-Walker Geometry, Int. J. Theor. Phys. 33, 1053–1062 (1994)

    Article  MathSciNet  Google Scholar 

  32. R. Portugal, Exact solutions of the Dirac equations in an anisotropic cosmological background, J. Math. Phys. 36, 4296–4300 (1995)

    Article  MathSciNet  ADS  MATH  Google Scholar 

  33. P.D. Lax, Hyperbolic Partial Differential Equations, (Am. Math. Soc., 2006), pp. 61

  34. R. Geroch, Partial Differential Equations of Physics, (1996). arXiv:gr-qc/9602055v1

  35. L. Ryder, Introduction to General Relativity, (Cambridge University Press, Cambridge, 2009), pp. 397, 416–418

    Book  MATH  Google Scholar 

  36. B. Mashhoon, Neutron Interferometry in a Rotating Frame of Reference, Phys. Rev. Lett. 61, 2639–2642 (1988)

    Article  ADS  Google Scholar 

  37. L. Ryder, Spin-rotation coupling and Fermi-Walker transport, Gen. Rel. Gravit. 40, 1111–1115 (2008)

    Article  MathSciNet  ADS  MATH  Google Scholar 

  38. M. Arminjon, Dirac Equation from the Hamiltonian and the Case with a Gravitational Field, Found. Phys. Lett. 19, 225–247 (2006). arXiv:gr-qc/0512046v2

    Article  MathSciNet  MATH  Google Scholar 

  39. B.L. Moiseiwitsch, Variational Principles, (Dover, Mineola, 2004), pp. 45-50

    MATH  Google Scholar 

  40. O.D. Johns, Analytic Mechanics for Relativity and Quantum Mechanics, (Oxford University Press, Oxford, 2005), pp. 267—271 and 563–565

    Book  Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Frank Reifler.

Appendix: Proof of Theorem 1

Appendix: Proof of Theorem 1

To prove Theorem 1 of Section 2, we will use the following corollary of a deep theorem of linear hyperbolic partial differential equations:

Theorem 0

Let M 1 , M 2 , ⋯ , M n be complex d×d matrix functions that depend smoothly on n + 1 independent real variables t , x 1 , x 2 , ⋯ , x n in a slab − T ≤ t ≤ T , x ∈ R n , denoted as I×R n . Suppose that M 0 , M 1 , M 2 , ⋯ , M n are Hermitian matrices and furthermore assume that M 0 is positive definite. Let \( F = F\left( S \right) \) be a homogeneous linear function of complex d×d matrices S , as well as having explicit dependence on t , x 1 , x 2 , ⋯ , x n . Then the complex linear hyperbolic system

$$ M_{0} \frac{\partial S}{\partial t} + \sum _{j=1}^{n} M_{j} \frac{\partial S}{ \partial x_{j} } = F\left( S \right) \label{GrindEQ__73_} $$
(81)

has a smooth d×d complex matrix-valued solution \( S {\bf :} {\bf I}\times {\bf R}^{{\rm n}} \to {\it M}\left( {\it C} {\it ,} {\it d} \right) \) which equals the identity matrix at t = 0 , as its prescribed smooth initial data.

Proof

Theorem 0 is Corollary 3 in Ref. [20], Sect. 6, which is based on a theorem of Lax [33] (see also Ref. [34]).□

Proof of Theorem 1 of Section 2

First note from (4) that the matrices B μ ≡  μ are Hermitian matrices. Then, note that the conditions stated in Theorem 1 for the metric components g μν imply that B 0 ≡  0 is a positive definite matrix by Theorem 6 of Ref. [22], Appendix B. By Theorem 3 of Ref. [21], Sect. 3.4, Eq. (54), a local similarity transformation T of the second kind, takes a generalized Dirac equation of the form (11) into a Dirac equation of the normal form (14), if and only if T satisfies the following partial differential equation:

$$ B^{\mu } D_{\mu } T= -\frac{1}{2} \left( D_{\mu } B^{\mu } \right) T . \label{GrindEQ__74_}$$
(82)

If we can solve (82) for such a local similarity transformation T , then the transformed coefficient fields will be as in (5) and (8):

$$\begin{array}{lll} \widetilde{\gamma }^{\mu } &=& T^{-1} \gamma ^{\mu } T, \\ \widetilde{A}\quad &=&\quad T^{+} A T, \\ \widetilde{\Gamma }_{\mu } &=& \Gamma _{\mu } . \end{array} $$
(83)

Now from (18), a local similarity transformation S of the first kind takes a Dirac equation of the normal form (14) into a normal QRD–0 equation (16), if and only if S satisfies the following partial differential equation:

$$\widetilde{B}^{\mu } \partial _{\mu } S= -\widetilde{B}^{\mu } \widetilde{\Gamma }_{\mu } S, \label{GrindEQ__76_}$$
(84)

where \( \widetilde{B}^{\mu } \equiv \widetilde{A}\widetilde{\gamma }^{\mu } \). The matrices B μ and \( \widetilde{B}^{\mu } \) are Hermitian, and moreover, the matrices B 0 and \( \widetilde{B}^{0} \) are positive definite. Thus, the two systems (82) and (84) have same form as the complex linear hyperbolic system in (81).

Let χ : UR ×R 3 , mapping \( X \to \left( t , {\bf x} \right) \), be the coordinate chart that we assume to be defined on U . Then, consider the projection map π : R ×R 3R taking \( \left( t , {\bf x} \right) \to t \). Let X 0 ∈ U and let \( t_{0} = \pi \circ \chi \left( X_{0} \right) \in {\bf R} \). Let \( {\bf M} = \left( \pi \circ \chi \right)^{-1} \left( t_{0} \right) \). Note that X 0 ∈ M ⊂ U . It will suffice to prove that there exist nonsingular solutions T and S of the systems (82) and (84) that are both defined in a common open neighborhood \( \widetilde{{\bf W}} \) of X 0 ∈ U.

By Theorem 0, the Cauchy problem for (82) with the smooth initial data T| M  = 1 4 has a smooth solution T in an open neighborhood W # of X 0 . Denote by W the open subset of W # in which T is a nonsingular matrix, so that T  − 1 exists. Note that X 0 ∈ W since X 0 ∈ M and T| M  = 1 4 . Thus, W is an open neighborhood of X 0 such that the local similarity transformation T (and its inverse T  − 1 ) is well defined on W , and hence, from Eq. (83), the complex linear hyperbolic system (84) is well defined on W .

By Theorem 0, the Cauchy problem for (84) with the smooth initial data S| M ∩ W  = 1 4 has a smooth solution S in an open neighborhood \( \widetilde{{\bf W}}^{{\bf \# }} \subset {\bf W} \) of X 0 . Denote by \( \widetilde{{\bf W}} \) the open subset of \( \widetilde{{\bf W}}^{{\bf \# }} \) in which S is a nonsingular matrix so that S  − 1 exists. Note that \( X_{0} \in \widetilde{{\bf W}} \) since X 0 ∈ M ∩ W and S| M ∩ W  = 1 4 . Thus, \( \widetilde{{\bf W}} \subset \widetilde{{\bf W}}^{{\bf \# }} \subset {\bf W} \) is an open neighborhood of X 0 such that both the local similarity transformation S and the local similarity transformation T (and their inverses S  − 1 and T  − 1 ) are well defined on \( \widetilde{{\bf W}} \), and therefore the local similarity transformation T ∘ S (and its inverse S  − 1 ∘ T  − 1 ) is also well defined on \( \widetilde{{\bf W}} \).□

Rights and permissions

Reprints and permissions

About this article

Cite this article

Arminjon, M., Reifler, F. Equivalent Forms of Dirac Equations in Curved Space-times and Generalized de Broglie Relations. Braz J Phys 43, 64–77 (2013). https://doi.org/10.1007/s13538-012-0111-0

Download citation

  • Received:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s13538-012-0111-0

Keywords

Navigation