Model Independent Bounds on Kinetic Mixing
6/4/2010
271 citations (255 excluding self-citations). Established the baseline experimental constraints on dark photons that the subsequent generation of dark sector searches was designed to improve upon.
The Problem
A dark photon — a new U(1) gauge boson that kinetically mixes with the Standard Model hypercharge — is one of the simplest and most motivated extensions of the Standard Model. It appears naturally in string compactifications, in models of dark matter with a dark sector, and as a generic consequence of any hidden gauge symmetry. By 2010, dark photon searches were becoming a major experimental program, but there was no systematic compilation of the existing constraints from decades of e+e- experiments across the full mass range.
The Key Idea
The paper is deliberately model-independent: it assumes only kinetic mixing and nothing else about the dark photon’s interactions. Using data from e+e- experiments spanning 1 GeV to 1 TeV (BaBar, Belle, CLEO, LEP, and others), it extracts bounds on the mixing parameter ε as a function of the dark photon mass. The result is ε < 0.03 across most of the mass range, regardless of any additional interactions the dark photon might have. This clean, model-independent bound gave experimentalists a clear target: any search that couldn’t beat ε ~ 0.03 wasn’t adding new information.
Impact
The paper established the baseline that the next generation of dark photon experiments was designed to beat. The “New Light Weakly Coupled Particles” community report (921 citations) from the 2013 Snowmass process, the US Cosmic Visions dark matter report (914 citations), and the MATHUSLA physics case (536 citations) all reference these bounds as the starting point for dark sector sensitivity projections. Curtin, Essig, and others’ study of dark photons at the LHC and 100 TeV colliders (481 citations) directly compares projected collider sensitivity against the e+e- bounds from this paper.
The bounds shaped the experimental program in a specific way: they showed that the parameter space above ε ~ 0.03 was already excluded, directing new experiments toward the 10-3 to 10-8 regime where discovery was still possible. Beam-dump reanalyses (341 citations) used the paper’s methodology to extract hidden photon limits from historical experiments. New proposals — LDMX (288 citations), SeaQuest dark sector searches (234 citations), and electron beam-dump experiments for MeV-GeV dark matter (204 citations) — were designed to probe the open parameter space below the bounds this paper established. The SIMP spectroscopy paper (181 citations), which connects directly to the SIMP Miracle, uses a kinetically mixed dark photon as the mediator between the dark and visible sectors, with the mixing parameter constrained by these bounds.
Recollections
Anson Hook and Eder Izaguirre were my graduate students at Stanford. The paper grew out of a simple frustration: the dark photon parameter space was being discussed at every dark sector workshop, but nobody had systematically compiled what was already known from existing e+e- data. The calculations were not difficult — the hard part was tracking down the experimental results from decades of different experiments, each with different luminosities, energy ranges, and systematic treatments, and combining them into a consistent set of bounds. The model-independent framing, assuming only kinetic mixing and nothing else, was a deliberate choice to make the bounds as broadly applicable as possible. It’s the same philosophy that drove the Jets+MET and Simplified Models work: present results in a form that the widest possible community can use.