Two kilometers beneath the Canadian Shield, inside a nickel mine turned physics laboratory, a set of silicon and germanium crystals has been chilled to roughly 15 millikelvins, just a sliver above the coldest temperature nature allows. The SuperCDMS experiment at SNOLAB reached that milestone on March 17, 2026, according to SLAC National Accelerator Laboratory, and in doing so activated one of the most sensitive instruments ever built to search for dark matter.
Dark matter makes up roughly 27 percent of the universe’s total mass and energy, according to measurements of the cosmic microwave background, yet no detector has ever registered a confirmed signal from one of its particles. Most large-scale searches, such as the LZ and XENONnT experiments, use liquid xenon to look for relatively heavy candidates known as WIMPs. SuperCDMS takes a different approach: its superconducting sensors are tuned to catch far lighter particles, some with masses below a single proton, that would slip past heavier-target detectors entirely.
From factory floor to mine shaft
The detector towers at the heart of SuperCDMS were fabricated and tested at SLAC in Menlo Park, California, then shipped to SNOLAB’s underground campus in Sudbury, Ontario, beginning in May 2023. What followed was a nearly three-year integration effort: wiring thousands of readout channels, sealing the cryostat, and executing a multi-stage cooldown that separately chilled cables and crystal arrays before bringing the full system to its operating range of 15 to 30 millikelvins.
At those temperatures, the germanium and silicon crystals become superconducting. A dark matter particle nudging even a single atom in the lattice would deposit enough energy to break that superconducting state in a tiny region, producing a measurable electrical signal. The principle is elegant, but making it work two kilometers underground, where logistics are constrained and every material must be screened for trace radioactivity, required years of preparation.
Rehearsals underground and lessons learned
Before committing the full detector payload, the collaboration ran dress rehearsals at two underground test facilities. The Cryogenic Underground Test Facility (CUTE), housed at SNOLAB itself, allowed researchers to validate cryogenic procedures and measure background radiation rates under the same rock overburden that shields the main experiment. A 2023 preprint authored by collaboration members describes how CUTE reduced commissioning risk by catching problems in a controlled setting before they could stall the larger project.
A parallel effort at Fermilab’s NEXUS underground test stand uncovered a subtler threat. Researchers operating prototype SuperCDMS HVeV detectors noticed time-correlated signals appearing across multiple sensors simultaneously, a pattern inconsistent with dark matter but consistent with stray light. They traced the source to luminescence in the printed circuit board holders surrounding the detectors. Ordinary PCB materials, bombarded by low levels of ambient radiation, were glowing faintly at wavelengths the sensors could pick up, mimicking the very signals the experiment was designed to find. Identifying that noise source before the full SNOLAB run informed decisions about which materials and shielding geometries to use in the final configuration.
A prototype’s proof of concept
While the main array was being assembled, a single SuperCDMS HVeV prototype detector was already producing physics results underground. A blind-analysis search using just 6.1 gram-days of exposure set exclusion limits on dark-matter-electron scattering at roughly MeV-scale masses and placed new constraints on hypothetical dark photons and axionlike particles. The paper, posted to the arXiv preprint server and reported as accepted by Physical Review D, has not yet been independently confirmed as appearing in the journal’s published volume. The results are modest in absolute terms, reflecting the tiny mass and short run time of a single prototype, but they demonstrate that the sensor technology works and that the collaboration can execute the kind of rigorous, blinded statistical analysis required for a credible discovery claim.
“Getting to 15 millikelvins underground is something the team has been working toward for years,” said Richard Partridge, SuperCDMS project director at SLAC, in the laboratory’s announcement. “Now the real work of understanding our detectors in their final home begins.”
Scaling from a gram-scale prototype to the full multi-kilogram SNOLAB array is not a simple multiplication exercise. Long-term detector stability, correlations between towers, and new background sources unique to the larger setup will all affect real-world performance. Still, the prototype results offer a concrete benchmark and a reason for cautious optimism.
What commissioning means, and what comes next
Reaching base temperature is a critical milestone, but it is not the finish line. SuperCDMS is now in commissioning, a phase that involves calibrating energy scales at millikelvin temperatures, characterizing each detector’s response to known particle sources, and validating the data-quality cuts that will separate potential dark matter signals from mundane backgrounds. That process can take months.
No primary source has yet disclosed live exposure times, initial noise performance, or preliminary data yields from the full array since the March 17 cooldown. The collaboration’s projected sensitivity, described in a 2017 design study in Physical Review D, forecasts world-leading reach for low-mass dark matter, but those projections rest on assumptions about background levels and energy thresholds that only commissioning data can confirm or revise. Real-world surprises, like the PCB luminescence discovered at NEXUS, could shift the actual sensitivity in either direction.
SuperCDMS enters a competitive landscape. Experiments like CRESST-III and EDELWEISS also target sub-GeV dark matter using cryogenic techniques, while liquid-xenon giants LZ and XENONnT continue to push sensitivity for heavier candidates. What sets SuperCDMS apart is the combination of its sensor design, optimized for the lightest mass ranges, and its location beneath two kilometers of Canadian rock, which filters out the cosmic rays that plague surface laboratories.
If commissioning proceeds smoothly and backgrounds match or beat design expectations, the first physics results from the full SuperCDMS SNOLAB array could arrive within the next year or two. A confirmed detection would rank among the most consequential discoveries in modern physics. Even a null result at the projected sensitivity would eliminate large swaths of theoretical parameter space, sharpening the target for the next generation of searches. Either way, the crystals are cold, the sensors are listening, and the hunt is on.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
