How does the loop momentum scale differ from direct production thresholds for TeV-scale mediators?

Your question asks about a technical distinction in collider phenomenology that sits at the boundary of what these preprints directly address.

The short answer: Loop momentum scaling and tree-level thresholds represent fundamentally different regimes. At tree level—direct s-channel production—a mediator becomes kinematically accessible once the collision energy exceeds its mass; this is the "threshold" effect [AG-2024.07-1229]. But when the mediator appears in loop diagrams (virtual loops inside Feynman graphs), the momentum flowing through the loop can be much larger than the mediator's mass without violating any kinematic constraint. This means loop-induced processes can probe mediator couplings even far below the direct production threshold, and the cross section's dependence on collision energy differs from the tree-level case.

The archive papers illustrate this principle across several contexts:

Loop-induced effects at colliders: In top-quark production, tree-level mediator exchange and one-loop contributions both matter [AG-2024.07-1229]. The loop contributions "demonstrate a significant contribution from diagrams involving the dark matter mediator" even when direct production would be suppressed—showing that loop momentum can effectively overcome threshold barriers.

Higher-order electromagnetic moments: Dark photon production through loop-induced interactions (dipole moments, anapole moments) from portal matter creates momentum-dependent Lorentz structures absent at tree level [AG-2025.02-233]. This changes the kinematic scaling of the cross section itself.

Form factor calculations: Two-loop QCD amplitudes for dark matter annihilation are decomposed into scalar form factors [AG-2025.11-1908], which encode how the amplitude depends on the loop momentum scale—a different question from whether the process is kinematically allowed.

The key physics insight: loop processes "turn on" gradually with collision energy (scaling determined by loop integrals), while tree-level thresholds are sharp. This allows experiments to constrain mediators both above and below their mass threshold, with different sensitivity patterns.

However, none of these preprints directly compares the two scaling regimes quantitatively for TeV-scale mediators, so a more rigorous treatment of this specific scaling relationship isn't available in the archive.