Fermilab’s Mu2e experiment has reached a critical assembly stage with its 60-ton cosmic-ray veto system in the detector hall, alongside a study accepted for publication reporting test-stand measurements that are extrapolated to a 99.99% cosmic-ray muon veto efficiency for the full four-layer design. The veto system is the last line of defense against cosmic rays that would otherwise swamp the faint signal Mu2e is hunting: a muon converting directly into an electron without producing neutrinos, a process not expected to occur at detectable rates in the Standard Model of particle physics. If that conversion exists at detectable rates, it would be direct evidence of new physics, and the veto’s performance is a key factor in separating a potential discovery claim from background noise.
What the published data actually show
The core evidence comes from a two-year cosmic-ray test-stand campaign whose results have been accepted for publication in Nuclear Instruments and Methods in Physics Research Section A, a standard venue for detector-performance papers. According to the detailed preprint, the collaboration measured muon detection efficiency in individual scintillator layers and then extrapolated those results to the full four-layer veto geometry surrounding the detector. The design target, per that study, is 99.99% veto efficiency, meaning only one in every ten thousand cosmic-ray muons would slip through undetected.
That number is not arbitrary. A separate Mu2e overview, available through a status report, explains why cosmic-ray background must be suppressed by four orders of magnitude for the experiment to reach its sensitivity goal. In practical terms, the detector will sit in a high-intensity muon beam, and any stray cosmic ray that mimics the energy signature of a muon-to-electron conversion would create a false positive indistinguishable from the real thing. The veto system exists to tag and discard those events before they contaminate the data.
The hardware itself consists of long plastic scintillator bars arranged in four overlapping layers, read out by silicon photomultipliers. A 2019 fabrication study described the 60-ton system’s design constraints, including low dead time to keep up with the beam’s intensity. Dead time matters because every microsecond the veto is unable to record creates a window where a cosmic ray could enter unnoticed. The study also documents quality-control procedures for thousands of individual modules, a scale that makes small per-channel inefficiencies add up quickly if not tightly managed.
What is verified so far
Three facts are confirmed by primary institutional sources. First, the tracker, a separate detector component responsible for measuring particle trajectories, was moved into the Mu2e Hall in November 2025. Second, the cosmic-ray veto and the calorimeter are both now in the hall. Third, integration of all three systems has begun, according to a Fermilab announcement. Mu2e leadership stated in that release that the team is “starting to see cosmic-ray tracks” through all components, a sign that the detectors are functioning together for the first time in their final configuration.
The peer-reviewed acceptance of the two-year performance study is also confirmed. That acceptance gives the efficiency claims a level of scrutiny beyond a preprint alone, because the journal’s review process requires independent evaluation of the statistical methods and extrapolation techniques the collaboration used. In addition, the publicly posted manuscript allows outside experts to inspect the full data set, reconstruction algorithms, and systematic-uncertainty estimates that underlie the 99.99% claim.
More broadly, Mu2e is embedded in the ecosystem of laboratories and universities that support open dissemination of results. Fermilab is among the institutions listed as contributors on the arXiv membership roster, which helps explain why the collaboration’s technical documentation appears in preprint form before journal publication. That pattern, preprint first and peer-reviewed article later, is standard in high-energy physics and shapes how early performance numbers circulate in the community.
What remains uncertain
The 99.99% figure itself carries an important caveat. The published study measured single-layer efficiency and then extrapolated to the full system. Extrapolation assumes that each layer operates independently and that edge effects, gaps between modules, and aging of scintillator material do not degrade performance in ways the test stand could not capture. The collaboration has not yet published results from the fully assembled veto running in the Mu2e Hall, where conditions differ from a controlled test stand.
There is also a subtle tension in how different papers frame the target. The performance study and the fabrication paper both describe the design goal as 99.99% efficiency. The Mu2e status paper, however, states that the CRV requires vetoing efficiency “better than 99.99%.” Whether the system needs to hit exactly that threshold or exceed it may seem like a minor distinction, but in a rare-event search where a single misidentified cosmic ray could fake a signal, even a small shortfall in veto power changes the experiment’s discovery reach. No primary source resolves whether the measured extrapolated efficiency meets the “better than” standard or merely the “equal to” standard.
Missing from all available sources are post-integration testing timelines, projected dates for first beam on target, and any cost breakdown for the veto system’s construction. Without those details, it is difficult to assess how close Mu2e is to producing physics results or whether budget pressures could delay operations. The existing documentation focuses on technical readiness and design validation rather than schedule risk or financial constraints.
How to read the evidence
Readers should distinguish between three tiers of evidence in the current Mu2e reporting. The strongest tier is the peer-reviewed performance study, which provides quantitative measurements from a real detector operating over two years. The full technical paper includes module counts, trigger configurations, and data-acquisition details that allow other physicists to reproduce or challenge the analysis. This is the kind of evidence that carries weight in the field, because it ties numerical claims to documented procedures and uncertainties.
The second tier is the Fermilab news release confirming physical milestones: hardware moved, components in place, integration underway. These are verifiable institutional facts, but they describe logistics rather than physics performance. The statement about seeing cosmic-ray tracks is encouraging, because it indicates that readout electronics, timing systems, and reconstruction software are all functioning in concert, yet it does not substitute for a calibrated efficiency measurement in the final detector environment.
The third and weakest tier is the extrapolation from single-layer data to full-system performance. Extrapolation is standard practice in detector physics, and the collaboration’s methods will have been scrutinized during peer review. Still, the gap between test-stand conditions and real operating conditions is where surprises tend to appear. Scintillator aging, temperature fluctuations in the hall, and electromagnetic interference from nearby equipment can all degrade performance compared with idealized assumptions. Until Mu2e publishes an in situ efficiency measurement for the complete cosmic-ray veto, the 99.99% figure should be read as a well-motivated design expectation rather than a fully demonstrated fact.
In that light, the current evidence supports a cautiously optimistic view. The hardware is built, installed, and beginning to see tracks; the design has passed both internal reviews and external peer review; and the extrapolated performance appears to clear the threshold that the collaboration itself has argued is necessary. At the same time, crucial questions about real-world efficiency margins, operational stability over years of running, and the exact headroom above the “better than 99.99%” requirement remain open. For an experiment where a handful of spurious events could blur the line between discovery and disappointment, those details will matter as much as the impressive engineering already on display.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
