[Paper Review] On the massive two-loop corrections to Bhabha scattering

M. Czakon, J. Gluza+1 more|ArXiv.org|Nov 15, 2005

Particle physics theoretical and experimental studies3 citations

TL;DR

This paper presents a systematic approach to computing the massive two-loop virtual corrections in Bhabha scattering, focusing on the most challenging topology—B5l3m—using a synergistic method combining Mellin-Barnes representations and differential equations. The authors derive analytical expressions for master integrals up to $\mathcal{O}(1/\epsilon)$, including the complete five-master system, and validate results via numerical checks, advancing the path toward a full two-loop calculation for precision luminosity measurements at future $e^+e^-$ colliders.

ABSTRACT

We overview the general status of higher order corrections to Bhabha scattering and review recent progress in the determination of the two-loop virtual corrections. Quite recently, they were derived from combining a massless calculation and contributions with electron sub-loops. For a massive calculation, the self-energy and vertex master integrals are known, while most of the two-loop boxes are not. We demonstrate with an example that a study of systems of differential equations, combined with Mellin-Barnes representations for single masters, might open a way for their systematic calculation.

Motivation & Objective

• To address the missing two-loop box diagrams in the massive calculation of Bhabha scattering, a key component for achieving $10^{-4}$ accuracy in luminosity measurements at future $e^+e^-$ colliders.
• To develop a systematic method for evaluating the most complex master integrals, particularly the five-master system B5l3m, which remains a bottleneck in the full two-loop computation.
• To combine Mellin-Barnes representations and differential equations to solve master integrals where standard DE techniques face high coupling complexity.
• To validate analytical results through numerical checks using sector decomposition, ensuring reliability for physical applications.

Proposed method

• A hybrid approach is employed: Mellin-Barnes representations are used to solve the most singular master integral B5l3md2, reducing the system from five to three coupled differential equations.
• Differential equations are then solved for the remaining master integrals, including B5l3md3[-1], using a parametrization in terms of Mandelstam variables $s = -(1-x)^2/x$, $t = -(1-y)^2/y$.
• The solution for B5l3md2 is derived analytically up to $\mathcal{O}(1/\epsilon)$, involving harmonic polylogarithms (HPLs) and generalized HPLs.
• The singularities of the remaining master integral B5l3md1 are determined algebraically by replacing dotted diagrams with numerator-dependent integrals.
• The method is cross-validated by numerical sector decomposition, confirming consistency across analytical and numerical approaches.
• All results are expressed in terms of standard HPLs and generalized HPLs, enabling implementation in Monte Carlo programs.

Experimental results

Research questions

• RQ1 Can the five-master system B5l3m, the most complex topology in two-loop Bhabha scattering, be systematically solved despite high coupling and kinematic complexity?
• RQ2 Can Mellin-Barnes representations be effectively used to decouple and solve the most singular master integral, thereby simplifying the full system of differential equations?
• RQ3 How can analytical results for master integrals be validated for accuracy and consistency in the context of massive two-loop amplitudes?
• RQ4 To what extent do the derived results for B5l3md2, B5l3md3[-1], and B5l3md1 contribute to the completeness of the two-loop virtual correction in Bhabha scattering?
• RQ5 Can the combined use of Mellin-Barnes and differential equations provide a robust framework for solving other high-multiplicity master integrals in multi-scale field theories?

Key findings

• The master integral B5l3md2 is analytically solved up to $\mathcal{O}(1/\epsilon)$ using Mellin-Barnes techniques, yielding an expression involving harmonic polylogarithms and a $1/\epsilon^2$ divergence.
• The differential equation for B5l3md3[-1] is solved explicitly, resulting in a closed-form expression proportional to $H[0,x]H[0,y]$ with rational prefactors in $x$ and $y$.
• The singular part of B5l3md1 is determined algebraically by replacing dotted diagrams with numerator integrals, leading to a complex expression involving $H[0,x]$, $H[0,y]$, and $\xi_2$ terms.
• The analytical results for all three master integrals are cross-checked against numerical sector decomposition, confirming consistency and reliability.
• The solution of B5l3md2 enables a reduction of the original five-master system to a three-master system, significantly simplifying further computation.
• The entire approach provides a viable pathway toward computing the remaining massive two-loop box diagrams, completing the virtual two-loop correction for Bhabha scattering.