Massive two-loop Bhabha scattering: the factorizable subset

The experimental precision that will be reached at the next generation of colliders makes it indispensable to improve theoretical predictions significantly. Bhabha scattering (e^+ e^- \to e^+ e^-) is one of the prime processes calling for a better theoretical precision, in particular for non-zero electron masses. We present a first subset of the full two-loop calculation, namely the factorizable subset. Our calculation is based on DIANA. We reduce tensor integrals to scalar integrals in shifted (increased) dimensions and additional powers of various propagators, so-called dots-on-lines. Recurrence relations remove those dots-on-lines as well as genuine dots-on-lines (originating from mass renormalization) and reduce the dimension of the integrals to the generic d = 4 - 2 \epsilon dimensions. The resulting master integrals have to be expanded to ${\it O}(\epsilon)$ to ensure proper treatment of all finite terms.