• Abstract and Figures
• Public Full-text

Triviality of quantum electrodynamics revisited D. Djukanovic,1 J. Gegelia,2, 3 and Ulf-G. Meißner4, 2 1Helmholtz Institute Mainz, University of Mainz, D-55099 Mainz, Germany 2Institute for Advanced Simulation, Institut f¨ur Kernphysik and J¨ulich Center for Hadron Physics, Forschungszentrum J¨ulich, D-52425 J¨ulich, Germany 3Tbilisi State University, 0186 Tbilisi, Georgia 4Helmholtz Institut f¨ur Strahlen- und Kernphysik and Bethe Center for Theoretical Physics, Universit¨at Bonn, D-53115 Bonn, Germany (Dated: 31 May, 2017) Quantum electrodynamics is considered to be a trivial theory. This is based on a number of evidences, both numerical and analytical. One of the strong indications for triviality of QED is the existence of the Landau pole for the running coupling. We show that by treating QED as the leading order approximation of an effective field theory and including the next-to-leading order corrections, the Landau pole is removed. Therefore, we conclude that the conjecture, that for reasons of self- consistency, QED needs to be trivial is a mere artefact of the leading order approximation to the corresponding effective field theory. PACS numbers: 03.70.+k , 11.10.Gh, 14.70.Bh The concept of triviality in quantum field theories orig- the problem of triviality. To that end we analyse the inates from papers by Landau and collaborators studying contributions of the next-to-leading order interaction, i.e. the asymptotic behaviour of the photon propagator in dimension five operator, the well-known Pauli term. We quantum electrodynamics (QED) [1, 2] (for a review see start with the most general U(1) locally gauge invariant e.g. Ref. [3]). Resumming the leading logarithmic con- effective Lagrangian of the electron (and the positron) tributions they found that the photon propagator has a field ψ interacting with the electromagnetic field Aµ pole at large momentum transfer. If this pole persists 1 µν also in non-perturbative calculations then to avoid the = F Fµν + ψ¯ (iD/ m) ψ apparent inconsistency QED has to be a non-interacting, L −4 − iκ µ ν ν µ i.e. trivial, theory. In calculations applying a finite cutoff + ψ¯(γ γ γ γ )ψ Fµν + ho , (1) this problem manifests itself as a singularity in the bare 2 − L coupling for a finite value of the cutoff. It is therefore where m is the (bare) electron mass, e is the (bare) elec- impossible to remove the cutoff unless the renormalized tromagnetic charge, Fµν = ∂µAν ∂ν Aµ, Dµ = ∂µ ieAµ coupling vanishes. and contains an infinite number− of terms with− oper- Lho The standard model, in a modern point of view, is ators of order six and higher. We assume that the con- a leading order approximation to an effective field the- tributions of these terms in the photon self-energy are ory (EFT) [4]. While the effective Lagrangian contains suppressed compared to those of the Pauli term. We re- an infinite number of local interactions consistent with mark here that the standard QED Lagrangian given by the underlying symmetries, at low energies the contribu- the first line of Eq. (1) describes the experimental data tions in physical quantities of the interactions with cou- very well. From the modern point of view this is be- pling constants of negative mass dimensions (i.e. non- cause the contributions of terms in the second line are arXiv:1706.10039v1 [hep-th] 30 Jun 2017 renormalizable interactions in the traditional sense) are beyond the current accuracy of the data. In particular, suppressed by powers of the energy divided by a large the anomalous magnetic moment of the electron in stan- scale characterising those degrees of freedom which are dard QED gets contributions only from loop diagrams not explicitly taken into account in the effective theory. and its calculated value agrees with the experiment very In the framework of EFT the solution of the leading or- well suggesting that the Pauli term is suppressed by a scale larger than 4 107 GeV [4]. der Wilson renormalization group equations might be ob- × structed at very large cutoffs, however, this should not be The scale-dependent renormalized running coupling 2 1 a hard problem because of irrelevant interactions or omit- eR(q ) can be defined by the following relation ted fields being important at short distances [4]. There- 1 qµqν fore inconsistencies in the renormalization group analysis Dµν (q) e2 = gµν e2 (q2), (2) of renormalizable quantum field theories, like QED or φ4 −q2 − q2 R theory of self-interacting scalars, might be absent in the corresponding EFT framework. In this letter we address the consequences of treating 1 We carried out all calculations in Landau gauge, however, the QED as a leading order approximation of an EFT for results are gauge independent. 2 In renormalizable theories only logarithmic divergences contribute to the renormalization of coupling constants and therefore there is a direct correspondence between the Gell-Mann-Low [5] and the Wilsonian renormaliza- tion group approaches [6]. As a result, the presence of the Landau pole in the expression of the running renor- malized coupling authomatically leads to the pole in the bare coupling as a function of the cutoff parameter. How- ever, in an EFT with non-renormalizable interactions the FIG. 1. Two-loop diagrams contributing in the photon self- direct link between the two renormalization group equa- energy. The solid and wiggly lines correspond to the electron tions is lost and therefore the Wilsonian renormalization and photon propagators, respectively. group approach requires additional study of the cutoff dependence. We investigate the cutoff dependence of the bare electromagnetic coupling in our model by apply- ing the higher derivative regularization [7, 8] which pre- serves the local U(1) gauge invariance. Notice that di- mensional regularization is not suitable here as it discards the power-law divergences. In addition to the fields in conventional QED, we introduce scalar (i.e. commuting) ghost fields ξ¯ and ξ which regulate the one-loop counter term diagrams contributing to the photon self-energy at FIG. 2. One loop counter term diagrams contributing in the two-loop order. The effective Lagrangian generating one photon self-energy at two-loop order. The solid and wiggly and two-loop diagrams, contributing to our calculation of lines correspond to the electron and photon propagators, re- the photon self-energy up to two-loop e2κ2 order, which spectively. The crosses denote the counter terms of the one- are all finite for finite Λ, is given by: loop order. Diagrams with ghost propagators are not shown. 2 2 1 µν ∂ = F 1+ Fµν LHDR −4 Λ2 3 µν 1 D2 where D (q) is the dressed propagator of the bare pho- + ψ¯ (iD/ m) 1+ ψ + h.c. ton field. Calculating the dressed propagator at one-loop 2 − Λ2 2 2 order we obtain for q m : 1 D2 3 − ≫ + ξ¯(iD/ m) 1+ ξ + h.c. 2 − Λ2 2 2 2 er iκ eR(q )= 2 2 2 2 . (3) µ ν ν µ er +2κ q −q 2 + ψ¯(γ γ γ γ )ψ Fµν + ho. (5) 1 2 ln 2 + cR q − 12π m 2 − L The bare electromagnetic coupling as a function of the Here, e is the renormalized coupling at q2 = m2 for r cutoff satisfies a renormalization group equation which c = 0, where c is a renormalized coupling− constant R R up to the level of accuracy of our calculation has the of the higher-order Lagrangian . It is suppressed by Lho form: two orders of some large scale. For κ = cR = 0 the running coupling has the well-known pole singularity at dα(Λ) = A α2(Λ)+κ2α2(Λ)Λ2(2A +A +2A ln Λ/m), (the Landau pole) d ln Λ 1 2 3 3 (6) 2 q2 = m2 exp 12π2/e2 . (4) where α(Λ) = e /(4π), the coefficient A1 is given by one- L − r loop diagrams and A2 and A3 are extracted from two- While this pole appears at extremely high energies, re- loop calculations. Notice that there are no power-law ducing its practical importance to nil, it is still a problem divergences at one-loop order and all terms suppressed if present in the full theory [3]. For reasonable values of by powers of m/Λ have been dropped in our calculations κ 1/ q2 the Landau pole is absent remedying the as they are negligible for large values of Λ. ≫ − L inconsistencyp at the level of an EFT. The solution to Eq. (6) is given by α α(Λ) = 0 , (7) 1 α ln Λ (A + A κ2Λ2)+ α ln Λ0 (A + A κ2Λ2) α A κ2(Λ2 Λ2) − 0 m 1 3 0 m 1 3 0 − 0 2 − 0 3 where α0 = α(Λ0) is the bare coupling at some fixed and A2 terms and the negative sign of A3 guarantees that cutoff m< Λ0 < Λ. α(Λ) has no pole. For Λ Λ we have ≫ 0 α To conclude, we have shown that the problem of triv- α(Λ) = 0 . (8) Λ 2 2 2 2 iality in QED can be attributed to QED being a lead- 1 α0 ln (A1 + A3κ Λ ) α0A2κ Λ − m − ing order approximation of an effective field theory. We For κ = 0 and positive A1 the expression in Eq. (8) has have shown that already at next-to-leading order, i.e. a pole at adding the Pauli term to QED, the Landau pole disap- pears thereby obviating the need for QED to be a trivial 1 ΛP = m exp . (9) theory. hA1α0 i This pole, if remaining in the non-perturbative full ex- This work was supported in part by DFG and pressions of the bare coupling, prevents the Λ limit NSFC through funds provided to the Sino-German CRC → ∞ 110 “Symmetries and the Emergence of Structure in unless α0 0, thus leaving us with a non-interacting theory.

Load more

  3

Copy Link