• Abstract and Figures
• Public Full-text

Corrections to Laser Electron Thomson Scattering Thomas Heinzl Anton Ilderton School of Computing and Mathematics Department of Applied Physics Plymouth University Chalmers University of Technology Drake Circus, Plymouth PL4 8AA, UK SE-412 96 Gothenberg, Sweden [email protected] [email protected] We discuss classical and quantum corrections to Thomson scattering between an electron and a laser. For radiation reaction, nonlinear, and quantum effects we identify characteristic dimensionless parameters in terms of which we determine the leading order correction terms. 1 Introduction result, the Thomson cross section takes on the numeri- −24 2 cal value σTh ' 6:6 × 10 cm ' 0:66 barn which is Thomson scattering is the classical process of light be- not small by particle physics standards. ing deflected by a charged obstacle of typical size smaller The differential cross section may be derived in a than the wavelength of the incident radiation. For defi- similar vein. For this one needs the angular distribution niteness we will henceforth assume the charge to be an of the radiated power (1) in direction l which is [1] electron and the incoming wave stemming from a laser beam. The scattered radiation may be viewed as re- dP e2 rad = (l · v_ )2 ; (5) sulting from a two-step process: (i) the acceleration of dΩ (4π)2c3 the charge due to the electromagnetic field it encoun- ters and (ii) the bremsstrahlung emitted in consequence. or, upon employing the equation of motion, with E = (In fact, this is what happens when solving the classical E, equations in powers of the coupling: at zeroth order, dP e4 rad = E2 1 − (l · )2 : (6) there is only acceleration, while at first order, there is dΩ (4π)2m2c3 emission.) The instantaneously radiated power is then given by Larmor's formula which, in its nonrelativistic Averaging over polarisations and introducing the scat- incarnation and using Heaviside-Lorentz units, reads [1] tering angle θ one has 2 e2 1 + cos2 θ P = v_ 2 : (1) h1 − (l · )2i = : (7) rad 3 4πc3 pol 2 Expressing the acceleration by means of the nonrel- This angular dependence was originally found by Thom- ativistic Lorentz force (neglecting the v × B term), son when calculating the mean \rate at which energy is mv_ = eE, E being the electric field of the incoming streaming through unit area" [2]. Dividing by the in- wave, the radiated power becomes coming flux, S = cE2, finally yields the differential cross section, 4 2 e 2 dσTh 1 2 2 Prad = 2 3 E : (2) = r (1 + cos θ) : (8) 3 4πm c dΩ 2 e This may be turned into a cross section (units of area) Integrating over angles we reobtain (4). Clearly, the arXiv:1307.0406v1 [hep-ph] 1 Jul 2013 upon dividing by the energy flux, i.e. the modulus of the differential cross section is more useful in general as it Poynting vector, which for a plane wave (PW) is conveys more, namely spectral, information. It is instructive to rederive (4) in a fully covariant PW S = c jE × Bj = cE2 : (3) way. The radiated power may be inferred from the proper time derivative of the wave 4-momentum which Hence, we obtain the Thomson cross section, can be expressed in terms of the electron 4-velocity uµ [3], 2 2 Prad 8π e 8π 2 2 σ := = =: r ; (4) µ 2 e 2 µ Th 2 e P_ = − u_ u : (9) S 3 4πmc 3 3 4πc5 2 2 with the classical electron radius, re ' 3 fm. At this Dotting in u, using u = c and the covariant equation µ µν distance, the Coulomb energy between electrons equals of motion, mu_ = (e=c)F uν , one finds the radiated mc2, and (with the benefit of hindsight) one ends up power, with a distance scale that is actually typical for strong _ µα ν interactions (1 fm being roughly the nucleon size). As a Prad = u · P = σThuµF Fα uν =c =: σThw0 : (10) For a plane wave (or, more general, null field), the square Here we have introduced the laser frequency in the in- of the field strength tensor coincides with the energy- stantaneous rest frame, !0. For a plane wave depending momentum tensor, only on the invariant phase k · x with k2 = 0, we can write !0 = k · u. In this case the solution of (15) implies µα ν PW µν F Fα = c T ; (11) !0 = k · u = k · u0 + O(τ0) where u0 is the initial four- velocity [11]. Thus, !0 is only conserved in the absence so that w0 in (10) may be interpreted as the energy of radiation reaction [12]. flux density of the electromagnetic wave as `seen' by the Assuming that our equations of motion both receive 2 electron in its instantaneous rest frame. This together corrections of order τ0 we should rewrite the cross sec- with (10) yields a nice expression for the Thomson cross tion as 2 2 sections in terms of Lorentz invariants, σRR = σTh 1 − !0τ0 : (17) u · P_ The cross section (16) may alternatively be obtained by σ = ; (12) considering an electron scattering off a bound charge, Th u T µν u µ ν with the interaction described by an additional har- the numerator and denominator being the field energy monic damping force, FRR = −δmv. The cross section rate of change and flux density in the co-moving frame is (see [1], Ch. 45), of the electron, respectively. σTh σRR = 2 2 ; (18) 1 + δ =!0 2 Radiation reaction which we can match to (16) by identifying 2 The two-step procedure mentioned in the introduction δ = !0τ0 : (19) breaks down (or rather, becomes insufficient) when the energy loss due to radiation becomes substantial such For a later comparison with the quantum effects it is that the produced radiation field back-reacts on the par- useful to rewrite the small parameter !0τ0 in terms of 2 ticle motion. In this case, one has to solve a modified the fine structure constant, α = e =4π~c, and the di- equation of motion taking back-reaction into account mensionless parameter and named after Lorentz, Abraham and Dirac (LAD) k · u ! eζ ! [4, 5, 6]. Introducing the time parameter ν := ~ 0 = ~ 0 = ~ ; (20) 0 mc2 mc2 mc2 2 −23 τ0 := re=c ' 10 s ; (13) which measures the energy of the laser photons (as seen 3 by the initial-state electron) in units of the electron rest it is most compactly written as [3] mass. In the last expression we have introduced rapid- ity ζ via the electron gamma factor, exp ζ = γ(1 + β), e to write ν in terms of the lab frequency, !. Obviously, mu_ µ = F µ + τ mu¨µ = F µν u + τ mu¨µ : (14) 0 0 c ν 0 one has 2 This equation, being third order in time derivatives !0τ0 = αν0 ; (21) ... 3 (¨u = x ), has pathological features such as runaway so- where the factors of ~ cancel on the right-hand side as lutions and preacceleration which can be traded for each they must for a purely classical parameter. Let us es- other but not entirely removed while insisting on (14) timate the magnitude of this quantity for a relativistic { see the lucent discussion in [7]. The loophole is to electron (β ' 1) colliding with an optical laser [12, 13]. modify the equation via iteration, i.e. by expressing mu¨ 2 −6 ζ In this case one has ν := ~!=mc ' 10 and e ' 2γ. on the right-hand side in terms of the Einstein-Lorentz Thus, ! τ ' αγν ' 10−8γ, which implies that classical _ µ 0 0 force, mu¨ = F +O(τ0) which yields the Landau-Lifshitz radiation reaction will be substantial for electron ener- equation [8] gies of 102 TeV which is in the far quantum regime. So, unless we find a way of classically boosting the radia- mu_ µ = F µ + τ mF_ µ + O(τ 2) : (15) 0 0 tion reaction its observation will not become feasible. A possible means towards this end is nonlinearity [14, 15]. This equation has been rederived (but only to order τ0) using adiabatic perturbation theory [9] or a sophisti- cated classical regularisation procedure [10]. For plane 3 Nonlinear Effects wave backgrounds there is an exact analytic solution [11] Clearly, either equation must lead to a modification Laser intensity is measured in terms of the r.m.s. energy of the Thomson cross section. This has already been gained by an electron traversing a laser wave length, in worked out in Dirac's seminal paper [6] who found the units of the electron rest energy. This leads to the di- following simple result, mensionless laser amplitude σTh eErmsλ eErms σRR = 2 : (16) a0 = = ; (22) 1 + (!0τ0) mc2 mc! which may be expressed covariantly using the energy section (27), and hence the two will be difficult to dis- density of (10) such that [16] tinguish experimentally. This statement holds for circu- lar polarisation, but for linear polarisation the situation e2hu T µν u i a2 = µ ν rms : (23) is different and higher harmonics have been unambigu- 0 m2c2(k · u)2 ously identified from their angular radiation patterns [19]. When the laser beam is focussed down to the diffraction 2.0 limit this may be turned into a rule-of-thumb estimate of a in terms of laser power P measured in petawatts 0 L 1.5 (PW), namely 2 3 1.0 a0 ' 5 × 10 PL=PW : (24) For powers of 10 PW to be expected in the near future 0.5 this implies a0 values of a few hundred.

Load more

  5

Copy Link