betaFeedback

WorksPhysics and Astronomy

One-loop fermion-photon vertex in arbitrary gauge and dimensions: a novel approach

Abstract

We compute one-loop electron-photon vertex with fully off-shell external momenta in an arbitrary covariant gauge and space-time dimension. There exist numerous efforts in literature where one-loop off-shell vertex is calculated by employing the standard first order Feynman rules in different covariant gauges and space-time dimensions of interest. The tensor structure which decomposes this three-point vertex into the components transverse and longitudinal to the photon momentum gets intertwined in this first order formalism. The Ward-Takahashi identity is explicitly invoked to untangle the two pieces and the results are expressed in a preferred basis of twelve spin-amplitudes. We propose a novel approach based upon an efficient combination of the first and second order formalisms of quantum electrodynamics to compute this one-loop vertex. Among some conspicuous advantages is the fact that this less known second order formalism separates the spin and scalar degrees of freedom of an electron interacting electromagnetically. More noticeably, the longitudinal and transverse contributions naturally disentangle from the onset in our approach. Moreover, this decomposition leads to identities between one-loop scalar Feynman integrals with higher powers in the propagators and shifted space-time dimensions that can be used to prove the Ward-Takahashi identity at one-loop order without the need to evaluate any Feynman integral. Additionally, this natural decomposition allows us to establish the gauge-independence of the Pauli form factor through explicit cancellations of scalar Feynman integrals that depend on the gauge parameter. These cancellations naturally lead to a compact expression for the Pauli form factor in arbitrary dimensions. Wherever necessary and insightful, we make comparisons with earlier works.

Introduction

Solving any quantum field theory (QFT) is equivalent to computing its Green functions. Three and four-point interaction vertices are the defining Green functions of any QFT within the standard model of particle physics. These vertices appear at the level of the Lagrangian. The simplest three-point vertex in quantum electrodynamics (QED) involves a photon interacting with a fermion, a charged lepton (like electron) or a quark. Therefore, the fermion-photon vertex not only orchestrates the dance of events when purely electromagnetic interactions are involved but it also plays a crucial role in unravelling the internal structure of hadrons. Though this structure and dynamics of hadrons are predominantly determined by quantum chromodynamics (QCD), it is generally probed through electromagnetic interactions of photons with electrically charged quarks which compose all hadrons. One such example is provided by hadron electromagnetic form factors. The quark-photon vertex not only ensures charge conservation at zero photon momentum-transfer through the vector Ward-Takahashi identity (WTI) but its labyrinthine details also explain how the asymptotic limit of these form factors is faithfully approached for large momentum-transfer of the probing photons.

A popular approach to study non-perturbative effects in a QFT in continuum is through the infinite set of Schwinger-Dyson equations (SDEs) which constitute the defining equations of motion of a QFT. Their intricate mathematical structure is such that the SDE of the 2-point Green function is coupled to that of the 3-point Green function through an integral equation, the one of the 3-point function is in turn entangled with that of the 4-point function ad infinitum. Any practical attempt to look for a non-perturbative solution is generally made by proposing a reliable ansatz for the relevant interaction vertices to truncate the infinite tower of these coupled equations to a finite solvable number. It is only natural to assume that any non-perturbative construction, say that of the fermion-photon vertex, must agree with its perturbative expansion in the weak coupling regime. Thus the perturbative knowledge of this vertex can be used as a guiding tool for constructing such an ansatz. In literature, several such studies can be found, spanned over the last four decades. For example,

  • In [1], one-loop QED vertex was computed in the Feynman gauge and four space-time dimensions, i.e., D=4D = 4. A convenient choice of the basis vectors was adopted therein such that the corresponding coefficients are free of kinematic singularities.

  • In Ref. [2], one-loop vertex was reported in the asymptotic limit of momenta where momentum squared in one of the fermion legs is much greater than in the other.

  • The work of Ball and Chiu in Ref. [1] was extended to an arbitrary covariant gauge in Ref. [3]. The appearance of unwanted kinematic singularities for this general choice of an arbitrary covariant gauge was observed. However, this shortcoming was

easily cured by redefining two of the basis vectors as a superposition of the previous set of these vectors.

  • In three space-time dimensions, D=3D = 3, this vertex for massless and massive fermions in an arbitrary covariant gauge was calculated in Refs. [4] and [5], respectively.

  • For scalar QED, the one-loop vertex in arbitrary covariant gauge and dimensions was reported in Ref. [6].

  • A complete calculation of the quark-gluon vertex in an arbitrary covariant gauge and dimensions was carried out in detail in [7]. All spinor QED results can be derived from the general expressions present therein through an appropriate choice of the color factors and selecting DD as desired.

  • In Ref. [8], numerical and analytical expressions for the transverse and longitudinal form factors of the quark-gluon vertex in different kinematical regimes are presented. Again, one can deduce QED results from there.

Following Ref. [1], subsequent works commence with the WTI which relates the full fermion-photon vertex Γμ(p,p′,k)\Gamma^\mu(p,p',k) with the full fermion propagator S(p)S(p) not only at every order of perturbation theory but even more generally, non-perturbatively:

k⋅Γ(p,p′,k)=S−1(p′)−S−1(p) ,(1)k \cdot\Gamma(p,p',k) = S^{-1}(p') - S^{-1}(p) \, , \tag*{(1)}

where kk is the incoming momentum of the photon, pp is that of the fermion and p′p' is the outgoing momentum of the latter. This identity readily allows us to decompose the vertex into two components, the so called longitudinal piece ΓLμ(p,p′,k)\Gamma^\mu_L(p,p',k) and the remaining part ΓTμ(p,p′,k)\Gamma^\mu_T(p,p',k) which is transverse to the photon 4-momentum kμk^\mu3:

Γμ(p,p′,k)=ΓLμ(p,p′,k)+ΓTμ(p,p′,k) ,(2)\Gamma^\mu(p,p',k) = \Gamma^\mu_L(p,p',k) + \Gamma^\mu_T(p,p',k) \, , \tag*{(2)}

where ΓTμ(p,p′,k)\Gamma^\mu_T(p,p',k) satisfies the following restrictions:

k⋅ΓT(p,p′,k)=0 ,ΓT(p,p,0)=0 .(3)k \cdot\Gamma_T(p,p',k) = 0 \, , \qquad\Gamma_T(p,p,0) = 0 \, . \tag*{(3)}

The longitudinal part ΓLμ(p,p′,k)\Gamma^\mu_L(p,p',k) alone satisfies the WTI. Starting from the limiting form of the WTI, namely, the Ward identity

Γμ(p,p,0)=∂∂pμS−1(p) ,(4)\Gamma^\mu(p,p,0) = \frac{\partial}{\partial p_\mu} S^{-1}(p) \, , \tag*{(4)}

Ball and Chiu proposed how to construct the longitudinal component, see [1] for the details and the explicit form of the vertex. We choose to call it the Ball-Chiu vertex and adopt the following notation

ΓLμ(p,p′,k)=ΓBCμ(p,p′,k) .(5)\Gamma^\mu_L(p,p',k) = \Gamma^\mu_{BC}(p,p',k) \, . \tag*{(5)}

We would like to emphasize that neither the method of its construction nor the final form of the vertex is unique. However, it has become a standard practice to decompose the vertex in this manner:

Γμ(p,p′,k)=ΓBCμ(p,p′,k)+ΓTμ(p,p′,k) .(6)\Gamma^\mu(p,p',k) = \Gamma^\mu_{BC}(p,p',k) + \Gamma^\mu_T(p,p',k) \, . \tag*{(6)}

All one-loop results for the vertex are then presented by ensuing the following strategy:

  • The vertex in a particular kinematic configuration, gauge and dimensions is calculated at the one-loop level.

  • Accordingly, one-loop fermion propagator is also computed within the same set of requirements.

  • Using the last result, ΓBCμ(p,p′,k)\Gamma^{\mu}_{BC}(p,p',k) is evaluated at the one-loop level: ΓBCμ 1-loop(p,p′,k)\Gamma^{\mu\,1\text{-loop}}_{BC}(p,p',k).

  • One-loop transverse vertex is obtained by simply subtracting the longitudinal vertex from the full vertex:

    ΓTμ 1-loop=Γμ 1-loop−ΓBCμ 1-loop .\Gamma^{\mu\,1\text{-loop}}_T = \Gamma^{\mu\,1\text{-loop}} - \Gamma^{\mu\,1\text{-loop}}_{BC} \, .
  • This result is finally projected onto the transverse basis proposed in [1] or [3] and the transverse form factors are identified for each basis vector.

The procedure outlined above is cumbersome though straightforward. There is a plethora of available research dedicated to constructing increasingly refined, reliable and physically meaningful ansätze, taking perturbation theory as a guide where key elements of a QFT, such as gauge invariance and renormalization, are satisfied order by order in a systematic manner. A comprehensive list of articles is hard to be provided without intentionally overlooking some of the several efforts in the literature. However, a selected list of references akin to the work we present here is [9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19].

All the one-loop results we have discussed so far are obtained solely through employing the standard first order formalism of QED. In this article, we weave this approach with a relatively less scrutinized second order formalism, Refs. [20, 21, 22, 23]. Explicit calculations presented here convincingly reveal that an adequate merger of both the first and the second order formalisms is particularly advantageous and efficient in the following aspects:

  1. Vertex decomposition: We shall observe in the next section that the electron-photon vertex Vμ(p,p′,k)V^\mu(p,p',k) decomposes naturally into its longitudinal and transverse pieces at one-loop order without requiring a Ball-Chiu like decomposition4:

Vμ(p,p′,k)=VLμ(p,p′,k)+VTμ(p,p′,k) ,(7)V^\mu(p,p',k) = V^\mu_L(p,p',k) + V^\mu_T(p,p',k) \, , \tag*{(7)}

where VTμV^{\mu}_{T} satisfies the transversality condition

k⋅VT(p,p′,k)=0(8)k \cdot V_{T}(p,p',k)=0 \tag*{(8)}

while the longitudinal piece VLμV^{\mu}_{L} satisfies the WTI at the same level of approximation:

k⋅VL(p,p′,k)=e[Σ(p̸)−Σ(̸p′)](9)k \cdot V_{L}(p,p',k)=e\left[\Sigma(\not{p})-\Sigma(\not{p}')\right] \tag*{(9)}

where Σ(p̸)\Sigma(\not{p}) is the fermion self-energy and ee is the usual QED coupling. We believe that this feature of a natural decomposition of the full vertex into its longitudinal and transverse components would continue to persist at higher orders of the electromagnetic coupling. However, only an explicit calculation will be able to confirm this statement.

2. Expeditious computation: By employing an astute concoction of the first and the second order formalisms, we are able to select efficient routes to not only expedite the complete computation of one-loop vertex in arbitrary gauge and dimensions but also write it in a much more compact and concise form. However, in order to make a swift comparison with other results in the literature, we can always project our result onto the Ball-Chiu [1] or Kızılersü-Pennington [3] basis.

An added advantage of using this combined analysis is to keep track of those terms which identically vanish when external momenta go on-shell at all intermediate stages of this calculation. Both in the longitudinal and the transverse components of the vertex the terms which are explicitly irrelevant on-shell are ordered conveniently, e.g., the operator (p̸+m)(\not{p}+m) remains on the far right whereas (p̸−m)(\not{p}-m) is present on the left. Therefore, when on-shell conditions (p̸+m)us(p)=0(\not{p}+m)u_s(p)=0 and uˉs(p)(p̸+m)=0\bar{u}_s(p)(\not{p}+m)=0 are imposed, they vanish. After eliminating these terms and using the on-shell symmetries of the Feynman integrals, we explicitly see the instantaneous cancellations of the integrals that depend explicitly on the covariant gauge parameter ξ\xi, e.g., in the evaluation of the on-shell Pauli form factor. These cancellations naturally lead to a compact expression for it in an arbitrary space-time dimension in terms of scalar integrals with: (i) higher powers of scalar propagators and (ii) shifted dimensions. It hints towards a possible analogous simplification in the evaluation of the anomalous magnetic moment of the charged fermion at higher orders of perturbation theory.

The article is organized as follows: Sec. II contains a summary of the second order formalism. In § 1.2, we apply a convenient combination of first and second order formalisms to the fermion-photon vertex to decompose it into its longitudinal and transverse components. In Sec. IV, we provide explicit confirmation of the fact that the longitudinal part of the vertex is indeed longitudinal, i.e., it satisfies the WTI. We express it in terms of Feynman integrals. Then in Sec. V, transverse contributions of the vertex which were previously obtained in § 1.2, are written in terms of Feynman scalar integrals. Employing these expressions, and the results obtained in Sec. IV for the longitudinal part of the vertex, we also provide the coefficients of the Ball-Chiu basis in terms of scalar integrals. In Sec. VI we calculate the Pauli form factor in DD dimensions, explicitly showing its independence of the covariant gauge parameter ξ\xi by cancellations of the scalar integrals that depend on this parameter in the longitudinal and transverse contributions of the vertex. A compact expression for the Pauli form factor in terms of scalar integrals is obtained. Making use of this general expression, special cases, i.e., D=4D=4 and D=3D=3 are reviewed, recovering the known results in the literature. Concluding remarks with prospects for future endeavors are provided in § 1.6. The manuscript is complemented with three appendices with supplemental information. Appendix A provides explicit difference between the Ball-Chiu and our construction of the longitudinal vertex in perturbation theory. Naturally one is related to the other through the addition of a transverse piece. Appendices B and C detail expressions of the transverse and longitudinal vertices, respectively, at one-loop order, and the scalar integrals involved with relevant identities relating them.

Second order formalism for QED

The second order formalism for QED is based on Feynman rules displayed in Figure 1 which were formally derived in [21]. It provides an efficacious correspondence between spinor and scalar QED. This formalism has been used to compute fermion loops and amplitudes involving external fermions, see [21] and references therein. In this article, we apply this strategy to achieve similar success in the off-shell evaluation of the QED vertex. The only difference is the additional spin factor σμν=12[γμ,γν]\sigma^{\mu\nu}=\frac{1}{2}[\gamma^{\mu},\gamma^{\nu}] that appears in the QED vertex. It disentangles explicit contributions of the electromagnetic interaction of the fermionic particle into scalar and spin degrees of freedom with appropriate usage of Gordon-like identities. In addition to the rules of Figure 1, whenever a fermionic loop appears in a Feynman diagram, an extra factor of −1/2-1/2 must be included in the amplitude. Though the structure of the first order and the second order rules are different, the computed amplitude of a QED process following both formalisms is equivalent.

Image not extracted

Figure 1. Second order rules for spinor QED

As mentioned before, one can derive the second order rules from the first order [21], by formally rewriting the product of the fermion propagator S(p+k)S(p+k) with the first order vertex eγμe\gamma^{\mu} as,

eS(p+k)γμ=e−(p̸+k̸)+m(p+k)2+m2γμ=Ap,kμDp+k(10)eS(p+k)\gamma^{\mu}=e\frac{-(\not{p}+\not{k})+m}{(p+k)^2+m^2}\gamma^{\mu}=\frac{A^{\mu}_{p,k}}{D_{p+k}} \tag*{(10)}

where

Ap,kμ=Bp,kμ+Cpμ,A^{\mu}_{p,k}=B^{\mu}_{p,k}+C^{\mu}_{p},
Bp,kμ=e(2p+k)μ+eσμαkα,B^{\mu}_{p,k}=e(2p+k)^{\mu}+e\sigma^{\mu\alpha}k_{\alpha},
Cpμ=eγμ(p̸+m),C^{\mu}_{p}=e\gamma^{\mu}(\not{p}+m),
Dq=q2+m2.(11)D_q=q^2+m^2. \tag*{(11)}

Here, Dp−1D_{p}^{-1} represents the scalar propagator, and Bp,kB_{p,k} is the three point vertex of the second order rules which naturally separates the scalar and spin electromagnetic interactions. Thus a product of a propagator and a vertex in a first order Feynman diagram can be decomposed into a second order contribution, and a left over first order term given by the operator CpμC_{p}^{\mu}

Now, the four-point scalar vertex that appears in the second order rules arises when there are two consecutive pairs of a fermion propagator and a first order vertex in a Feynman diagram, as shown in Figure 2. An amplitude constructed from a first order Feynman diagram that contains a subdiagram as the one shown in Figure 2 is proportional to e2S(p+k1+k2)γνS(p+k1)γμe^{2}S(p+k_{1}+k_{2})\gamma^{\nu}S(p+k_{1})\gamma^{\mu}. According to Eq. (10), it can be decomposed as,

Image not extracted

Figure 2. Subdiagram that produces the contribution to the four-point scalar vertex of the second order formalism.

e2S(p+k1+k2)γνS(p+k1)γμ=Ap+k1,k2μDp+k1+k2Ap+k1μDp+k1.(12)e^{2}S(p+k_{1}+k_{2})\gamma^{\nu}S(p+k_{1})\gamma^{\mu}=\frac{A^{\mu}_{p+k_{1},k_{2}}}{D_{p+k_{1}+k_{2}}}\frac{A^{\mu}_{p+k_{1}}}{D_{p+k_{1}}}. \tag*{(12)}

Since

Cp+kμAp,kνDp+k=e2γμγν−e2(ημν−σμν),(13)C^{\mu}_{p+k}\frac{A^{\nu}_{p,k}}{D_{p+k}}=e^{2}\gamma^{\mu}\gamma^{\nu}-e^{2}\left(\eta^{\mu\nu}-\sigma^{\mu\nu}\right), \tag*{(13)}

Eq. (12) acquires the following form

References

  1. [1]J. S. Ball and T.-W. Chiu, Phys. Rev. D 22, 2542 (1980).
  2. [2]D. C. Curtis and M. R. Pennington, Phys. Rev. D 42, 4165 (1990).
  3. [3]A. Kizilersu, M. Reenders, and M. R. Pennington, Phys. Rev. D 52, 1242 (1995), arXiv:hep-ph/9503238.DOI
  4. [4]A. Bashir, A. Kizilersu, and M. R. Pennington, (1999), arXiv:hep-ph/9907418.DOI
  5. [5]A. Bashir and A. Raya, Phys. Rev. D 64, 105001 (2001), arXiv:hep-ph/0103036.
  6. [6]A. Bashir, Y. Concha-Sanchez, and R. Delbourgo, Phys. Rev. D 76, 065009 (2007), arXiv:0707.2434 [hep-th].DOI
  7. [7]A. I. Davydychev, P. Osland, and L. Saks, Phys. Rev. D 63, 014022 (2001), arXiv:hep-ph/0008171.DOI
  8. [8]R. Bermudez, L. Albino, L. X. Gutiérrez-Guerrero, M. E. Tejeda-Yeomans, and A. Bashir, Phys. Rev. D 95, 034041 (2017), arXiv:1702.04437 [hep-ph].DOI
  9. [9]D. C. Curtis and M. R. Pennington, Phys. Rev. D 48, 4933 (1993).
  10. [10]C. J. Burden and C. D. Roberts, Phys. Rev. D 47, 5581 (1993), arXiv:hep-th/9303098.DOI
  11. [11]A. Bashir and M. R. Pennington, Phys. Rev. D 50, 7679 (1994), arXiv:hep-ph/9407350.DOI
  12. [12]A. Bashir, A. Kizilersu, and M. R. Pennington, Phys. Rev. D 62, 085002 (2000), arXiv:hep-th/0010210.DOI
  13. [13]A. Bashir and A. Raya, Nucl. Phys. B 709, 307 (2005), arXiv:hep-ph/0405142.DOI
  14. [14]A. Kizilersu and M. R. Pennington, Phys. Rev. D 79, 125020 (2009), arXiv:0904.3483 [hep-th].DOI
  15. [15]A. Bashir, A. Raya, and S. Sanchez-Madrigal, Phys. Rev. D 84, 036013 (2011), arXiv:1108.4748 [hep-ph].DOI
  16. [16]A. Bashir, R. Bermudez, L. Chang, and C. D. Roberts, Phys. Rev. C 85, 045205 (2012), arXiv:1112.4847 [nucl-th].DOI
  17. [17]A. Kizilersu, T. Sizer, M. R. Pennington, A. G. Williams, and R. Williams, Phys. Rev. D 91, 065015 (2015), arXiv:1409.5979 [hep-ph].arxiv.org/abs/1409.5979
  18. [18]L. Albino, A. Bashir, L. X. G. Guerrero, B. E. Bennich, and E. Rojas, Phys. Rev. D 100, 054028 (2019), arXiv:1812.02280 [nucl-th].DOI
  19. [19]L. Albino, A. Bashir, A. J. Mizher, and A. Raya, Phys. Rev. D 106, 096007 (2022), arXiv:2201.01280 [hep-ph].
  20. [20]L. C. Holster, J. Math. Phys. 26, 1348 (1985).
  21. [21]A. G. Morgan, Phys. Lett. B 351, 249 (1995), arXiv:hep-ph/9502230.DOI
  22. [22]N. Ahmadiniaz, V. M. Banda Guzmán, F. Bastianelli, O. Corradini, J. P. Edwards, and C. Schubert, JHEP 08, 049 (2020), arXiv:2004.01391 [hep-th].DOI
  23. [23]N. Ahmadiniaz, V. M. B. Guzman, F. Bastianelli, O. Corradini, J. P. Edwards, and C. Schubert, JHEP 01, 050 (2022), arXiv:2107.00199 [hep-th].DOI
  24. [24]A. I. Davydychev, Phys. Lett. B 263, 107 (1991).
  25. [26]C. Anastasiou, T. Gehrmann, C. Oleari, E. Remiddi, and J. B. Tausk, Nucl. Phys. B 580, 577 (2000), arXiv:hep-ph/0003261.DOI
  26. [27]K. G. Chetyrkin and F. V. Tkachov, Nucl. Phys. B 192, 159 (1981).
  27. [28]F. V. Tkachov, Phys. Lett. B 100, 65 (1981).
  28. [29]O. V. Tarasov, Phys. Rev. D 54, 6479 (1996), arXiv:hep-ph/9606018.
  29. [30]R. N. Lee, Nucl. Phys. B Proc. Suppl. 205-206, 135 (2010), arXiv:1007.2256 [hep-ph].DOI
  30. [31]R. N. Lee, Nucl. Phys. B 830, 474 (2010), arXiv:0911.0252 [hep-ph].DOI
  31. [32]R. N. Lee, (2012), arXiv:1212.2685 [hep-ph].arxiv.org/abs/1212.2685
  32. [33]R. N. Lee, J. Phys. Conf. Ser. 523, 012059 (2014), arXiv:1310.1145 [hep-ph].DOI
  33. [34]M. E. Peskin and D. V. Schroeder, An Introduction to Quantum Field Theory (Westview Press, 1995) reading, USA: Addison-Wesley (1995).
  34. [35]A. K. Das and S. Perez, Physics Letters B 581, 182 (2004).

Paper details

vlnc:W_sm3qpqkf
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