Scalar leptoquark pair production at the LHC: precision predictions in the era of flavour anomalies

Abstract

We comprehensively examine precision predictions for scalar leptoquark pair production at the LHC. In particular, we investigate the impact of lepton t-channel exchange diagrams that are potentially relevant in the context of leptoquark scenarios providing an explanation for the flavour anomalies. We also evaluate the corresponding total rates at the next-to-leading order in QCD. Moreover, we complement this calculation with the resummation of soft-gluon radiation at the next-to-next-to-leading logarithmic accuracy, hence providing the most precise predictions for leptoquark pair production at the LHC to date. Relying on a variety of benchmark scenarios favoured by the anomalies, our results exhibit an interesting interplay between the t-channel diagram contributions, the flavour texture satisfied by the leptoquark Yukawa couplings, the leptoquark masses and their representations under the Standard Model gauge group, as well as the chosen set of parton densities used for the numerical evaluations. The net effect on a cross section turns out to be very non-generic and ranges up to about 60% with respect to the usual nextto-leading-order predictions in QCD (i.e. without any t-channel contribution) for some scenarios considered. Dedicated calculations are thus required for any individual leptoquark model that could be considered in a collider analysis in order to assess the size of the studied corrections. In order to facilitate such calculations we provide dedicated public numerical packages.

Leptoquark manoeuvres in the dark: a simultaneous solution of the dark matter problem and the $$ {R}_{D^{\left(\ast \right)}} $$ anomalies

“…Appendix A is dedicated to technical details about the implementation of the model in numerical tools. Moreover, as a byproduct of our analysis, we discuss in appendix B the impact of t-channel SM lepton exchanges contributions to leptoquark pair production, an effect that is currently ignored in the vast majority of the leptoquark phenomenological and experimental studies and that could be important [36,37].…”

“…For all collider results included in this paper, we convolve leading-order (LO) matrix elements with the NNPDF30_lo_as_0118 [50] set of parton distribution functions (PDF), that we handle through LHAPDF [51]. Since predictions at next-to-leading-order (NLO) in QCD are in the same ballpark as the corresponding LO rates [36,37,[52][53][54], higher-order corrections are not expected to largely modify the LHC constraints that we derive on the model in this work. We therefore employ LO cross sections throughout this study.…”

Leptoquark manoeuvres in the dark: a simultaneous solution of the dark matter problem and the $R_{D^{(*)}}$ anomalies

“…Appendix A is dedicated to technical details about the implementation of the model in numerical tools. Moreover, as a byproduct of our analysis, we discuss in appendix B the impact of t-channel SM lepton exchanges contributions to leptoquark pair production, an effect that is currently ignored in the vast majority of the leptoquark phenomenological and experimental studies and that could be important [26,27].…”

“…For all collider results included in this paper, we convolve leading-order (LO) matrix elements with the NNPDF30_lo_as_0118 [39] set of parton distribution functions (PDF), that we handle through LHAPDF [40]. Since predictions at next-to-leading-order (NLO) in QCD are in the same ballpark as the corresponding LO rates [26,27,[41][42][43], higher-order corrections are not expected to largely 800 900 1000 1100 1200…”

“…The naive scaling of this diagram indeed goes as λ 2 R , which is (naively) assumed to be subdominant relative to the QCD-induced contributions (that scale as the strong coupling α s ). However, the considered solutions to the R D ( * ) anomalies sometimes involve O(1) values for λ R , so that this diagram could have a substantial impact on a cross section [26,27]. Deciding whether this contribution could be neglected is very non-generic, and the issue must be addressed on a case-by-case basis, for any individual benchmark.…”