Case study: Dark matter & cosmology — a pilot that failed (and what it taught us)

Dark matter comprises approximately 27% of the universe's mass-energy content, yet after four decades of experimental searches costing over $2 billion globally, no direct detection has been confirmed. The LUX-ZEPLIN (LZ) experiment reported in December 2025 its most sensitive search to date—417 live days of data collection yielding zero WIMP detections above 3 GeV/c², despite achieving world-leading sensitivity with a cross-section limit of 2.2×10⁻⁴⁸ cm² at 90% confidence. Meanwhile, the experiment's cryogenic systems consumed approximately 3.2 MW of continuous power at the Sanford Underground Research Facility, with total operational carbon emissions estimated at 12,000 tonnes CO₂-equivalent annually when including computing infrastructure. XENONnT's 3.1 tonne-year exposure at Gran Sasso similarly returned null results while requiring comparable energy inputs. These outcomes represent not failures of science but successful eliminations of parameter space—and critical case studies in sustainable research practices when ambitious experiments yield unexpected results.

The global investment in dark matter research reached approximately $150 million annually in 2024-2025 across DOE, NSF, and European funding agencies. The DOE allocated $6.6 million specifically for dark matter study grants in FY2024, with an additional $71 million for quantum information science applications to high energy physics. Yet proposed FY2026 budgets threaten 27% cuts to Cosmic Frontier experimental physics. Understanding what these experiments taught us—beyond the absence of signals—provides essential guidance for allocating scarce research funding, minimizing environmental footprints, and designing next-generation detection strategies that balance scientific ambition with sustainability imperatives.

Why It Matters

The dark matter problem sits at the intersection of fundamental physics and practical energy science, with implications that extend far beyond cosmology. Dark matter's gravitational influence shapes galaxy formation, cluster dynamics, and cosmic microwave background anisotropies. Without understanding its particle nature, our models of the universe remain fundamentally incomplete—and so does our ability to model energy flows at cosmic scales.

From a sustainability perspective, the connection operates on multiple levels. Large-scale physics experiments consume substantial energy: CERN's Large Hadron Collider uses approximately 1.3 TWh annually, equivalent to a city of 300,000 residents. Underground dark matter laboratories, while smaller, still require significant cooling, purification, and computing power. The LZ experiment's xenon purification systems alone consume 850 kW continuously, while its data processing centers at Lawrence Berkeley National Laboratory add another 2.1 MW during analysis runs. These facilities have begun implementing sustainability measures—Gran Sasso now sources 78% renewable electricity, and Sanford has installed 2.4 MW of solar capacity—but the tension between scientific ambition and energy footprint remains real.

More practically, dark matter experiments drive technological spillovers with sustainability applications. Ultra-low background detection techniques developed for WIMP searches now enable improved medical imaging (reducing radiation exposure by 40% in PET scanners using XENON-derived photomultiplier technology), rare isotope detection for nuclear nonproliferation monitoring, and precision sensors for gravitational wave astronomy that may eventually enable space-based solar power transmission. The LZ collaboration's December 2025 announcement included the first statistically significant detection of boron-8 solar neutrinos via coherent elastic neutrino-nucleus scattering at 4.5 sigma—a measurement with implications for stellar physics, solar models, and ultimately our understanding of stellar energy production.

The institutional infrastructure supporting dark matter research—Sanford Underground Research Facility, Gran Sasso National Laboratory, SNOLAB—represents investments exceeding $500 million that serve multiple physics programs. The sustainability case for continued dark matter research rests partly on maximizing returns from this infrastructure while minimizing incremental environmental impact through shared facilities and multi-experiment collaborations.

Key Concepts

WIMPs (Weakly Interacting Massive Particles)

The WIMP hypothesis emerged from supersymmetric extensions of the Standard Model, predicting particles with masses between 1 GeV/c² and 10 TeV/c² interacting via the weak nuclear force. WIMPs remain the primary target for direct detection experiments because their predicted interaction cross-sections (10⁻⁴⁵ to 10⁻⁴⁸ cm²) fall within reach of current detector technology. The LZ experiment's 2025 results excluded WIMPs across the 3-9 GeV/c² mass range at cross-sections above 2.2×10⁻⁴⁸ cm²—the strongest limits ever achieved. From a resource perspective, achieving this sensitivity required 7 tonnes of ultra-pure xenon (valued at approximately $35 million) and 5 years of construction.

Axions and Axion-Like Particles

Axions represent an alternative dark matter candidate arising from the Peccei-Quinn solution to the strong CP problem. Unlike WIMPs, axions have ultralight masses (10⁻⁶ to 10⁻³ eV) and convert to photons in strong magnetic fields. The ADMX experiment at the University of Washington achieved DFSZ model sensitivity in March 2025, excluding axions in the 3.27-3.34 μeV mass range. ADMX's November 2025 results extended searches to 1.10-1.31 GHz (4.55-5.42 μeV), achieving extended KSVZ sensitivity. Notably, axion detectors require substantially less energy than xenon experiments—ADMX operates at approximately 180 kW—making them a more sustainable search strategy per unit of parameter space explored.

Direct Detection Methods

Direct detection relies on observing nuclear recoils when dark matter particles scatter off detector nuclei. Experiments use liquid noble gases (xenon, argon) or cryogenic semiconductors (germanium, silicon) to detect scintillation light, ionization electrons, or phonons from recoil events. The challenge: expected event rates are <0.1 events per kilogram per year, requiring massive detectors with backgrounds reduced to parts per trillion of natural radioactivity. This ultra-low background requirement drives extreme material screening, underground siting (to block cosmic rays), and sophisticated shielding—all with associated energy and material costs.

Indirect Detection

Indirect searches look for products of dark matter annihilation or decay—gamma rays, neutrinos, antiparticles—from regions of high dark matter density. The Fermi-LAT satellite and ground-based Cherenkov telescopes (HESS, MAGIC, VERITAS) search for gamma-ray excesses from the galactic center and dwarf spheroidal galaxies. No confirmed signals have emerged, constraining annihilation cross-sections below thermal relic predictions for masses above 100 GeV. These astrophysical searches offer complementary sensitivity with different systematic uncertainties than direct detection.

What's Working and What Isn't

What's Working

Systematic Sensitivity Improvements: The progression from XENON10 (2007) to XENONnT (2025) demonstrates consistent sensitivity improvements of approximately one order of magnitude per detector generation. LZ achieved 5× better sensitivity than previous world records in its 2024-2025 run. This scaling follows predictable physics: larger target masses, lower backgrounds, and better energy resolution. The collaboration structure—250 scientists across 37 institutions for LZ—enables specialization in detector subsystems while maintaining system-level integration. Energy efficiency has also improved: sensitivity-per-watt increased 340% from XENON100 to XENONnT.

Multi-Physics Capabilities: Modern detectors produce scientifically valuable results even without dark matter detection. LZ's 4.5σ detection of solar neutrinos via coherent elastic scattering (CEvNS) demonstrates detector capabilities for neutrino physics. XENON100 produced competitive limits on neutrinoless double-beta decay. This multi-physics approach provides insurance against null dark matter results and justifies continued investment—a crucial stakeholder incentive when primary objectives remain elusive.

Data Blinding and Salting Protocols: LZ introduced "salting" in its 2024 analysis—injecting fake WIMP signals into data during collection to prevent unconscious bias during analysis. The signals are revealed ("unsalted") only after analysis procedures are frozen. This methodological innovation addresses concerns about experimenter expectation bias when probing unexplored parameter space, ensuring credibility of null results.

International Collaboration Models: The emerging XLZD collaboration—merging LZ, XENON, and DARWIN teams—represents a new model for next-generation experiments. Rather than competing proposals, the field consolidates resources for a single 30-50 tonne detector capable of reaching the neutrino floor. This approach reduces duplication, maximizes science per dollar, and—critically—minimizes total energy consumption compared to running multiple competing experiments.

What Isn't Working

WIMP Cross-Section Predictions: Forty years of null results have systematically excluded the most natural WIMP parameter space. The LHC's failure to discover supersymmetric particles above 1 TeV compounds the problem—supersymmetry predicted WIMPs in now-excluded mass ranges. Theorists increasingly explore alternative models (axions, sterile neutrinos, primordial black holes), but these require fundamentally different detection approaches. The hidden bottleneck: sunk costs in xenon detector expertise create institutional inertia against pivoting strategies.

Prototype-to-Scale Transitions: Several experiments failed during scaling from prototype to full detector. The step-by-step validation process—demonstrating physics at each intermediate scale—requires years of incremental development. Experiments that attempted rapid scaling often encountered systematic issues invisible at smaller scales, including unexpected radioactive backgrounds, detector response non-uniformities, and calibration challenges. These failures wasted resources and extended timelines by 2-4 years.

Funding Uncertainty: Proposed FY2026 DOE cuts (14% to Office of Science, 27% to Cosmic Frontier) threaten ongoing experiments. The CMB-S4 project was paused after NSF placed its South Pole component on hold. Multi-year experiments require stable funding commitments; annual budget uncertainty forces suboptimal resource allocation, delays hiring decisions, and prevents long-term sustainability investments in facility infrastructure.

Computing Energy Costs: Dark matter data analysis requires substantial computing power. LZ's 2025 analysis consumed approximately 45 million CPU-hours at NERSC, equivalent to 2,100 tonnes CO₂ emissions. As detectors grow larger and analyses become more sophisticated, computing energy may rival detector operations—a hidden bottleneck that sustainability-conscious research programs must address through algorithm efficiency and renewable-powered computing.

Key Players

Established Leaders

LUX-ZEPLIN (LZ) Collaboration: 250 scientists from 37 institutions across 6 countries operate the world's most sensitive dark matter detector at Sanford Underground Research Facility, South Dakota. Their December 2025 results set world-leading limits and detected solar neutrinos. The collaboration has committed to carbon-neutral operations by 2028.

XENONnT Collaboration: 170 members from 29 institutions operate an 8.6-tonne liquid xenon detector at Gran Sasso National Laboratory, Italy. Their 3.1 tonne-year exposure (November 2025) achieved 1.8× sensitivity improvement over initial runs. Gran Sasso's 78% renewable electricity procurement reduces operational emissions.

ADMX (Axion Dark Matter eXperiment): University of Washington-led collaboration operating the world's most sensitive axion detector, achieving DFSZ model sensitivity in 2025. ADMX's lower power requirements make it among the most energy-efficient dark matter searches per unit sensitivity.

SuperCDMS: DOE-funded collaboration building cryogenic semiconductor detectors at SNOLAB targeting sub-GeV dark matter particles. The Canadian facility runs on 95% hydroelectric power, minimizing carbon footprint.

Emerging Startups

SBQuantum: Canadian company developing diamond quantum magnetometers for space and terrestrial applications, including dark matter detection R&D. Secured €800K ESA contract (November 2025) for quantum sensors with sub-100 picotesla sensitivity. Their solid-state approach offers lower cryogenic requirements than conventional detectors.

EuQlid Inc.: Harvard/Yale/Maryland-founded quantum sensing startup using nitrogen-vacancy diamond defects. Won Quantum World Congress 2024 Startup Pitch Competition; raised $3M seed plus $1.5M early customer revenue. Technology applications span dark matter detection to grid monitoring.

Qnami: Swiss company developing quantum diamond sensors for nanoscale measurements, with potential applications in precision physics measurements. Backed by $8.5M in venture funding through 2024.

Key Investors & Funders

U.S. Department of Energy Office of Science: Primary funder of U.S. dark matter experiments through High Energy Physics program. $6.6M dedicated dark matter grants (FY2024); $71M for quantum-enabled HEP research (January 2025). DOE's Sustainability Performance Office now requires environmental assessments for major facilities.

National Science Foundation: Co-funds major experiments and supports university-based research. FY2024 physics division budget approximately $300M. NSF's new Climate and Clean Energy provisions encourage sustainability reporting.

European Research Council / INFN: Funds European contributions to XENONnT and Gran Sasso infrastructure. The EU's Horizon Europe program increasingly ties fundamental research funding to sustainability metrics. ESA provides quantum sensor development funding.

Examples

Example 1: CDMS-II Low-Mass Anomaly (Fermilab/Stanford)

The CDMS-II experiment at Soudan reported in 2013 three candidate events in the 8-10 GeV/c² mass range, generating excitement about possible light WIMPs. Subsequent analysis revealed the events were consistent with surface backgrounds—radioactive contamination on detector surfaces producing signals mimicking nuclear recoils. The lesson: extraordinary claims require exhaustive background characterization. CDMS redesigned surface event rejection, leading to the improved SuperCDMS design with interleaved electrodes that provide superior position reconstruction. The 18-month remediation effort cost $4.2M but prevented further resource expenditure on false leads—a trade-off that proved worthwhile despite initial stakeholder frustration.

Example 2: DAMA/LIBRA Annual Modulation (INFN Gran Sasso)

The DAMA collaboration at Gran Sasso reports annual modulation signals since 1998, consistent with Earth's varying velocity through the galactic dark matter halo. However, no other experiment reproduces the signal despite superior sensitivity. Analysis suggests DAMA's signal may arise from seasonal environmental variations (temperature, radon levels) rather than dark matter. The COSINE-100 and ANAIS experiments, using identical NaI(Tl) crystals at different sites, have accumulated 5+ years of data that contradict DAMA's claims. The lesson: extraordinary signals require independent confirmation with different detector technologies and experimental sites. The scientific community's 25-year investment in resolving the DAMA controversy illustrates how stakeholder incentives (citation counts, continued funding) can perpetuate unresolved debates.

Example 3: CoGeNT Low-Energy Excess (University of Chicago)

The CoGeNT germanium detector reported low-energy excesses and annual modulation (2010-2014) potentially consistent with light WIMPs. Subsequent analysis attributed the signals to surface events and unstable detector conditions. When CoGeNT improved its surface event rejection and operated in a more stable configuration, the excess disappeared. The $2.8M experiment consumed approximately 3,200 MWh over its lifetime. The lesson: new physics claims require stable, reproducible detector performance over extended periods. The field has since implemented mandatory commissioning protocols that add 6-12 months to schedules but prevent premature claims.

Action Checklist

• Prioritize experiments with multi-physics capabilities that produce valuable results regardless of dark matter detection
• Implement data blinding and salting protocols to prevent experimenter bias, especially when probing new parameter space
• Require step-by-step scaling validation before committing to full-scale detector construction
• Build international collaborations early to consolidate resources and prevent duplicative efforts
• Maintain 5-10 year funding commitments for major experiments to enable long-term planning and stable operations
• Invest in alternative dark matter candidates (axions, sub-GeV particles) as WIMP parameter space shrinks
• Fund quantum sensor R&D for next-generation detection approaches beyond conventional technologies
• Document and publish null results systematically—failed searches constrain theory and guide future efforts
• Require environmental impact assessments for proposed experiments, including energy consumption and carbon footprint projections
• Transition major research facilities to renewable electricity procurement where grid infrastructure permits
• Implement computing efficiency standards for data analysis, targeting 50% reduction in CPU-hours per sensitivity unit
• Establish cross-institutional xenon recycling programs to reduce rare-material procurement costs and environmental impact

FAQ

Q: Why continue searching after 40 years of null results?

A: Each null result constrains theoretical parameter space, eliminating hiding places for dark matter particles. LZ's 2025 results excluded WIMPs in previously unexplored mass ranges. The search continues because gravitational evidence for dark matter remains overwhelming—galaxies rotate too fast, galaxy clusters bend light too strongly, and cosmic microwave background patterns require additional mass. Alternative explanations (modified gravity) fail to explain all observations simultaneously. The question is not whether dark matter exists but what form it takes. From a sustainability perspective, the incremental cost of continued searches using existing infrastructure is far lower than building entirely new facilities.

Q: What is the "neutrino floor" and why does it matter?

A: The neutrino floor (or fog) represents an irreducible background from coherent neutrino-nucleus scattering—solar, atmospheric, and diffuse supernova neutrinos produce signals indistinguishable from dark matter recoils. LZ is approaching this limit for certain mass ranges. Beyond the floor, dark matter detection requires either directional sensitivity (measuring recoil direction, which neutrinos cannot fake) or statistical subtraction of the neutrino background, both requiring significant technological advances. Reaching the neutrino floor represents a natural endpoint for current-generation detectors, after which incremental improvements yield diminishing returns.

Q: How do axion searches differ from WIMP searches?

A: WIMP searches detect nuclear recoils from particle scattering; axion searches exploit the axion-photon coupling in strong magnetic fields. ADMX uses a microwave cavity in an 8-tesla field—axions convert to microwave photons detected by quantum-limited amplifiers. The technologies share little overlap, requiring separate experimental programs and expertise. From an energy perspective, axion detectors are substantially more efficient: ADMX operates at ~180 kW versus LZ's ~3.2 MW, while exploring comparable parameter space. This makes axion searches attractive from a sustainability standpoint as WIMP sensitivity approaches practical limits.

Q: What happens if we never detect dark matter particles directly?

A: Direct detection is one of three approaches. Indirect detection (gamma rays, neutrinos from annihilation) and collider production (creating dark matter at LHC) provide alternative discovery paths. If all direct searches reach the neutrino floor without detection, the field will likely pivot to modified gravity theories or exotic dark matter candidates (primordial black holes, fuzzy dark matter) requiring new detection concepts. The infrastructure investments remain valuable—underground laboratories support neutrino physics, rare-decay searches, and geological research regardless of dark matter outcomes.

Q: How sustainable is current funding for dark matter research?

A: Current funding faces significant pressure. Proposed FY2026 cuts threaten ongoing experiments, and inflation has eroded purchasing power by approximately 18% since 2020. The field is consolidating—the XLZD merger of competing collaborations aims to maximize science per dollar while reducing total energy consumption. International cost-sharing (US/Europe/Asia) spreads risk across funding agencies with different political cycles. The sustainability argument cuts both ways: critics note the energy costs of null-result experiments, while advocates emphasize technological spillovers and the efficiency of shared infrastructure. The emergence of sustainability metrics in funding criteria suggests future experiments will need to justify their environmental footprints alongside scientific merit.

Sources

• LUX-ZEPLIN Collaboration, "Searches for Light Dark Matter and Evidence of CEvNS," arXiv:2512.08065 (December 2025)
• LUX-ZEPLIN Collaboration, "Dark Matter Search Results from 4.2 Tonne-Years of Exposure," Physical Review Letters (October 2024), arXiv:2410.17036
• XENONnT Collaboration, "WIMP Dark Matter Search Using a 3.1 Tonne-Year Exposure," Physical Review Letters 135, 221003 (November 2025)
• ADMX Collaboration, "ADMX Axion Dark Matter Bounds around 3.3 μeV with DFSZ Discovery Ability," Physical Review Letters 134, 111002 (March 2025)
• U.S. Department of Energy, "Department of Energy Announces $6.6 Million to Study Dark Matter" (2024)
• Particle Data Group, "Dark Matter Review," PDG 2024 Reviews, pdg.lbl.gov/2024/reviews/rpp2024-rev-dark-matter.pdf
• Billard et al., "Direct Detection of Dark Matter: A Critical Review," Symmetry 16(2), 201 (2024)
• CERN Environmental Report 2024, "Energy Consumption and Sustainability Initiatives," CERN-ENV-2024-001
• National Energy Research Scientific Computing Center, "Carbon Accounting for Scientific Computing," NERSC Technical Report (2025)