Home Physical Sciences Lorentz Atom Revisited by Solving the Abraham–Lorentz Equation of Motion
Article
Licensed
Unlicensed Requires Authentication

Lorentz Atom Revisited by Solving the Abraham–Lorentz Equation of Motion

  • Jürgen Bosse ORCID logo EMAIL logo
Published/Copyright: July 15, 2017

Abstract

By solving the non-relativistic Abraham–Lorentz (AL) equation, I demonstrate that the AL equation of motion is not suited for treating the Lorentz atom, because a steady-state solution does not exist. The AL equation serves as a tool, however, for deducing the appropriate parameters Ω and Γ to be used with the equation of forced oscillations in modelling the Lorentz atom. The electric polarisability, which many authors “derived” from the AL equation in recent years, is shown to violate Kramers–Kronig relations rendering obsolete the extracted photon-absorption rate, for example. Fortunately, errors turn out to be small quantitatively, as long as the light frequency ω is neither too close to nor too far from the resonance frequency Ω. The polarisability and absorption cross section are derived for the Lorentz atom by purely classical reasoning and are shown to agree with the quantum mechanical calculations of the same quantities. In particular, oscillator parameters Ω and Γ deduced by treating the atom as a quantum oscillator are found to be equivalent to those derived from the classical AL equation. The instructive comparison provides a deep insight into understanding the great success of Lorentz’s model that was suggested long before the advent of quantum theory.

Acknowledgements

This work was supported in part by the German–Brazilian DAAD-CAPES program under the project name “Dynamics of Bose–Einstein Condensates Induced by Modulation of System Parameters”. It was my pleasure to discuss with A. Pelster and J. Akram many aspects of this work. Special thanks go to V. Bagnato and E. dos Santos for their warm hospitality and fruitful discussions during a visit to USP Sao Carlos, Brazil, where part of this work was developed.

Appendix

Unique Solution of the AL Equation of Motion

The unique solution of (26), which is the general solution of the homogeneous equation plus a particular solution of the inhomogeneous equation, may be cast into the following form (tt0),

(a1)xAL(t,t0)=xALh(t,t0)+xALp(t,t0),
(a2)xALh(t,t0)=ϕ(tt0)x0+iχ(tt0)mv0(b0+Γv0+Ω2x0)τ21+4Ω˜2τ2[e(tt0)(Γ+1/τ)+ϕ(tt0)+(Γ+1/τ)iχ(tt0)m]
(a3)xALp(t,t0)=0tt0dt[et(Γ+1/τ)+ϕ(t)+(Γ+1/τ)iχ(t)m]τf(tt)(1+4Ω˜2τ2)m,

where the compact form of the denominator results from applying the first of its identities (33). Here oscillator relaxation and response functions, ϕ(t) and χ(t), and frequency Ω˜ are defined in terms of (Ω, Γ) and are given, respectively, in (7) and (10). The oscillator parameters Ω=Ω(τ, ω0) and Γ=Γ(τ, ω0) are given in terms of the AL parameters (τ, ω0) in (44).

As discussed in Section 3.3 above, (a2) implies that the unique solution of (26) for initial values (x0, v0, b0) will diverge, if (tt0)→∞, because the characteristic polynomial of (26) has a positive root, z2=Γ+1/τ>0. From limt0xAL(t,t0)=, I conclude that a steady-state solution of the AL equation does not exist. A steady-state solution would require that xALh(t,)=0 for generic (x0, v0, b0).

Fourier–Laplace Transform (FLT)

In (13), the Fourier–Laplace transform (FLT) of a bounded function f(t) (|f(t)|≤M<∞) has been introduced

(a4)f˜(z)=dt eitzisΘ(st)f(t),   s=sign[z]0,

which is an analytical function for all complex z outside the real axis. The FLT of f(t) has as a Cauchy integral representation

(a5)f˜(z)=dωπf(ω)ωzz=ω±iof(ω)±if(ω)

with f(ω)=12i[f˜(ω+io)f˜(ωio)] denoting the spectral function, or dissipative part of f˜(ω+io), and f(ω)=12[f˜(ω+io)+f˜(ωio)] denoting the reactive part of f˜(ω+io). Dissipative and reactive parts obey dispersion relations

(a6)f(ω)=dω¯πf(ω¯)ω¯ω,   f(ω)=dω¯πf(ω¯)ω¯ω,

known as Kramers–Kronig relations in physics’ literature.

In general, f(ω)=dω¯πf(ω¯)ω¯ω,   and f(ω)=dω¯πf(ω¯)ω¯ω, will be complex functions of the real variable ω. Functions f(t), which vanish for large |t| (as is the case for response and relaxation functions discussed above), are related to their spectral function by conventional Fourier transform

(a7)f(ω)=dt2eitωf(t),   f(t)=dωπeitωf(ω),

and one easily verifies for the response function χ(t) (7), which is purely imaginary, odd in t, and vanishing for |t|→∞

(a8)χ(ω)=±[χ˜(ω±io)]=χ(ω)=χ(ω),

a spectral function which is real, odd in ω, and 1/2 of the conventional Fourier transform of χ(t). Similarly, the relaxation function ϕ(t) (7), which is real, even in t, and vanishing for |t|→∞, will have a spectral function

(a9)ϕ(ω)=±[ϕ˜(ω±io)]=ϕ(ω)=ϕ(ω)

which is real, even in ω, and just 1/2 of the conventional Fourier transform of ϕ(t). For response and relaxation spectrum, Kubo’s identity takes the simple form: χ″(ω)=ωϕ″(ω)/(mΩ2).

References

[1] R. Grimm, M. Weidemüller, and Y. Ovchinnikov, Adv. At. Mol. Opt. Phys. 42, 95 (2000).10.1016/S1049-250X(08)60186-XSearch in Google Scholar

[2] G. Gilbert, A. Aspect, and C. Fabre, Introduction to Quantum Optics, Campridge University Press, New York 2010.Search in Google Scholar

[3] L. Bergmann and Cl. Schaefer. Lehrbuch der Experimentalphysik, volume II (Elektrizitätslehre). Walter de Gruyter & Co., Berlin 1961.10.1515/9783111441894Search in Google Scholar

[4] P. C. Martin. In C. de Witt and R. Balian, (eds, Problème à N Corps, page 37 ff), Gordon and Breach, New York 1968.Search in Google Scholar

[5] D. J. Griffiths, Introduction to Electrodynamics, 3rd ed., Prentice-Hall Inc. 1999.Search in Google Scholar

[6] J. D. Jackson, Classical Electrodynamics, 3rd ed., John Wiley & Sons Inc., New York 1999.Search in Google Scholar

[7] J. D. Jackson, Classical Electrodynamics, 1st ed., John Wiley & Sons, New York 1962.Search in Google Scholar

[8] P. A. M. Dirac, Math Phys Sci. 167, 148 (1938).Search in Google Scholar

[9] H. Spohn, Dynamics of Charged Particles and their Radiation Field, Cambridge University Press, Cambridge, UK 2004.10.1017/CBO9780511535178Search in Google Scholar

[10] H. Spohn, Europhys. Lett. 50, 287 (2000).10.1209/epl/i2000-00268-xSearch in Google Scholar

[11] F. Rohrlich, Phys. Lett A. 283, 276 (2001).10.1016/S0375-9601(01)00264-XSearch in Google Scholar

[12] L. Landau and E. M. Lifshitz, The Classical Theory of Fields, Pergamon Press, New York 1975.Search in Google Scholar

[13] G. W. Ford and R. F. O’Connell, Phys. Lett A. 157, 217 (1991).10.1016/0375-9601(91)90054-CSearch in Google Scholar

[14] O. Keller, Quantum Theory of Near-Field Electrodynamics, Springer-Verlag, Heidelberg 2011.10.1007/978-3-642-17410-0Search in Google Scholar

[15] J. M. Aguirregabiria, J. Phys. A: Math. Gen. 30, 2391 (1997).10.1088/0305-4470/30/7/019Search in Google Scholar

[16] W. Heitler, The Quantum Theory of Radiation, 3rd ed., Oxford University Press, London 1954.Search in Google Scholar

Received: 2017-3-8
Accepted: 2017-6-3
Published Online: 2017-7-15
Published in Print: 2017-8-28

©2017 Walter de Gruyter GmbH, Berlin/Boston

Downloaded on 9.3.2026 from https://www.degruyterbrill.com/document/doi/10.1515/zna-2017-0182/html
Scroll to top button