How Do We Search for Dark Matter? Detection Experiments and Methods

Dark matter is the dominant form of matter in the universe — about 5 times more abundant by mass than ordinary atomic matter. It holds galaxies together, sculpts the large-scale structure of the cosmic web, and its density fluctuations seeded the formation of every galaxy, star, and planet in existence. Yet in more than 90 years since Fritz Zwicky first inferred its presence in the Coma Cluster (1933), no one has directly detected a single dark matter particle. We know it exists from its gravitational effects. We know roughly how much of it there is. But we have no confirmed idea what it actually is.

This is not for lack of trying. The search for dark matter is one of the most technically ambitious and scientifically urgent programs in modern physics. Underground laboratories shield detectors from cosmic radiation to detect the faint nuclear recoils of dark matter particles. Space telescopes scan the sky for gamma rays from dark matter annihilation in galactic halos. The Large Hadron Collider smashes protons together at record energies hoping to create dark matter in the lab. Microwave cavities and torsion balances search for ultralight particles in completely different mass ranges. Yet after decades of increasingly sensitive experiments, we remain in the dark.

What Dark Matter Must Be

Before discussing detection, it is useful to clarify what dark matter's properties are known to be from astrophysical observations:

Gravitationally interacting: Dark matter's gravitational effects are well-established and consistent across scales from galactic rotation curves (Rubin & Ford 1970s) to the Bullet Cluster collision (Chandra X-ray + weak lensing, 2006) to the CMB acoustic peaks.

Not baryonic: Big Bang nucleosynthesis constrains the baryon density (ordinary matter) to Ω_b ≈ 0.049 — only ~16% of the total matter density Ω_m ≈ 0.31. The rest (Ω_DM ≈ 0.26) cannot be ordinary atoms.

Non-relativistic at matter-radiation equality ("cold"): The large-scale structure of the universe, particularly the acoustic peak positions in the CMB and the galaxy power spectrum, is consistent with dark matter that was moving slowly (non-relativistically) when structure began forming. "Hot" dark matter (light neutrinos) predicts a spectrum inconsistent with observations at small scales.

Not electromagnetic: Dark matter does not emit, absorb, or scatter light. It is "dark" precisely because it does not interact electromagnetically.

Long-lived: Dark matter must be stable (or extremely long-lived) to persist for the 13.8 billion year age of the universe.

Collisionless (or nearly so): The Bullet Cluster shows that dark matter did not slow down when two galaxy clusters passed through each other — consistent with at most a small self-interaction cross-section.

The Leading Candidate: WIMPs

The most theoretically motivated dark matter candidate has long been the Weakly Interacting Massive Particle (WIMP). WIMPs arise naturally from supersymmetry and other extensions of the Standard Model of particle physics, and they have a compelling "miracle": if dark matter consists of particles with masses of 10–1,000 GeV (similar to weak force mediators) and interaction strengths of electroweak scale, the thermal relic density from the Big Bang would be approximately the right amount to explain the observed dark matter density. This "WIMP miracle" coincidence drove decades of experimental programs.

WIMPs are predicted to:

• Have masses of ~10 GeV – 10 TeV
• Interact with ordinary matter via the weak force (and possibly exchange of heavy mediator particles)
• Annihilate with anti-WIMPs to produce gamma rays, neutrinos, and particle-antiparticle pairs
• Scatter elastically off atomic nuclei with a cross section of ~10⁻⁴⁷–10⁻⁴⁵ cm²

Direct Detection: Underground Detectors

Direct detection searches for the nuclear recoils produced when a WIMP from the galactic dark matter halo collides with an atomic nucleus in a detector. The expected signal:

• Recoil energy: 1–100 keV (very low energy), making detection challenging
• Rate: Depends on WIMP mass, cross section, and local dark matter density (~0.3 GeV/cm³)
• Expected few events per year in a tonne of target material for optimistic WIMP parameters
• Background signals from natural radioactivity, cosmic rays, and solar neutrinos must be suppressed by many orders of magnitude

To reduce backgrounds, direct detection experiments are placed deep underground (to shield against cosmic ray muons), inside layers of radiation shielding, and use ultra-pure materials:

Liquid Noble Gas Detectors

LUX-ZEPLIN (LZ): Currently the world's most sensitive direct detection experiment, operating at the Sanford Underground Research Facility (SURF) in Lead, South Dakota, 1,478 meters underground. LZ contains 10 tonnes of liquid xenon (7 tonnes of active target). When a particle interacts in the xenon:

1. Prompt scintillation (S1): UV photons produced instantly by the interaction
2. Delayed electroluminescence (S2): Ionization electrons drift upward in an electric field to a gas phase, where they produce a delayed light signal proportional to the ionization

The ratio S2/S1 distinguishes nuclear recoils (from potential WIMPs or neutron backgrounds) from electron recoils (from gamma rays and betas, which are more common backgrounds). LZ's 2023 results (60 live-days) set the world's best limits on spin-independent WIMP-nucleon cross sections for WIMP masses above ~10 GeV: σ < 9.2 × 10⁻⁴⁸ cm² at 36 GeV WIMP mass. This rules out the most natural WIMP parameter space.

PandaX-4T: A 4-tonne liquid xenon detector at the China Jinping Underground Laboratory (CJPL), with similar sensitivity to LZ.

XENONnT: A 5.9-tonne liquid xenon detector at the Gran Sasso National Laboratory (LNGS) in Italy.

These three experiments form the leading frontier of WIMP direct detection. Their combined null results have essentially ruled out the "vanilla" WIMP with weak-scale cross sections.

DEAP-3600: A 3.3-tonne liquid argon detector at SNOLAB (Sudbury, Ontario). Argon offers different sensitivity than xenon — particularly for the spin-dependent channel — and different systematic uncertainties.

DarkSide-20k (planned): A 20-tonne liquid argon detector under construction, targeting even higher sensitivity through the larger mass.

Crystal Detectors: Germanium and Silicon

SuperCDMS (Cryogenic Dark Matter Search): Uses germanium and silicon crystals cooled to millikelvin temperatures. Phonon (heat) and ionization signals are measured simultaneously. The extraordinarily low energy threshold (~1 eV in some configurations) allows sensitivity to light dark matter (sub-GeV masses) that liquid xenon experiments cannot reach because xenon nuclei are too heavy to be significantly recoiled by light WIMPs.

COSINE-100 and ANAIS-112: These experiments test an anomalous annual modulation signal reported by the DAMA/LIBRA experiment (NaI crystals at Gran Sasso), which has claimed a WIMP dark matter signal at ~10σ statistical significance for over 20 years. The annual modulation (peaking in June, minimum in December) matches predictions for WIMPs: Earth's velocity relative to the galactic dark matter halo peaks in summer when Earth's orbital velocity adds to the Sun's motion through the halo. However, COSINE-100 and ANAIS-112 (using the same NaI material as DAMA but at different sites) do not confirm DAMA's signal. The DAMA anomaly remains unexplained and controversial.

The Neutrino Floor: the definitive Background

As direct detection experiments become more sensitive, they approach an irreducible background: the neutrino floor (or neutrino fog). Solar neutrinos, atmospheric neutrinos, and diffuse supernova background neutrinos scatter off nuclei via coherent elastic neutrino-nucleus scattering (CEνNS), producing nuclear recoils indistinguishable from WIMP signals at low energies. This neutrino background is irreducible — you cannot shield against it, only model it. Future experiments must use directional information or nuclear spin sensitivity to distinguish WIMP from neutrino recoils.

The first experiment to detect CEνNS (proving the concept) was COHERENT (at Oak Ridge), using the CsI detector to detect neutrinos from the Spallation Neutron Source. Direct detection experiments at tonne-scale will begin to see solar neutrino backgrounds imminently.

Indirect Detection: Looking for Dark Matter Annihilation

Indirect detection searches for the products of dark matter annihilation or decay in regions of high dark matter density (galactic center, dwarf galaxies, galaxy clusters, the Sun, Earth). If dark matter is a WIMP, pairs of WIMPs (and anti-WIMPs) in the galactic halo will occasionally annihilate, producing Standard Model particles — most usefully gamma rays, neutrinos, positrons, and antiprotons.

Gamma-Ray Searches

Fermi-LAT: The Fermi Large Area Telescope, in orbit since 2008, scans the entire sky at 20 MeV–300 GeV every 3 hours. It is the primary tool for gamma-ray dark matter searches:

• Galactic center excess: The galactic center shows a spatially extended excess of ~1–3 GeV gamma rays above expected astrophysical backgrounds (supernovae remnants, pulsars, diffuse emission). This excess was reported in 2014 and has morphological and spectral properties consistent with WIMP annihilation (mass ~30–50 GeV, annihilating to bb̄). However, an unresolved population of millisecond pulsars (which are numerous at the galactic center) would produce a statistically similar emission. Current analyses cannot definitively distinguish the two explanations; the statistical properties of the excess at small angular scales favor a population of point sources (pulsars) over truly diffuse emission (dark matter), but the debate continues.

• Dwarf spheroidal galaxies: The Milky Way's satellite dwarf galaxies are among the highest dark matter density environments known (mass-to-light ratios of 100–1000), and they are nearly free of astrophysical backgrounds (no star formation, no cosmic ray sources). Fermi-LAT searches for dark matter annihilation in ~15 dwarf galaxies set some of the strongest constraints on WIMP annihilation cross sections for masses below ~1 TeV — approaching the thermal relic cross section (~3 × 10⁻²⁶ cm³/s) for masses below ~30 GeV.

• Galaxy clusters: The Perseus, Virgo, and Fornax clusters contain enormous amounts of dark matter but also have complex astrophysical backgrounds. No convincing dark matter signal has been found.

Ground-based Cherenkov telescopes (MAGIC, HESS, VERITAS, CTA): Detect gamma rays above ~100 GeV from dark matter targets. HESS observations of the galactic center constrain WIMP masses above ~300 GeV. The Cherenkov Telescope Array (CTA), under construction with sites in Chile and La Palma, will improve sensitivity by a factor ~10 across the 30 GeV – 100 TeV range — potentially reaching the thermal relic sensitivity for multi-TeV WIMPs.

Positron and Antiproton Searches

AMS-02 (Alpha Magnetic Spectrometer, on the International Space Station since 2011): Measures cosmic ray positrons, antiprotons, and other anti-nuclei with extraordinary precision. A rising positron fraction above 10 GeV was confirmed by AMS-02 (following PAMELA's discovery), but is explained by nearby pulsars rather than dark matter. AMS-02 antiproton measurements are consistent with secondary production but have been used to constrain dark matter annihilation cross sections competitive with Fermi-LAT.

Antideuteron search: Dark matter annihilation could produce low-energy antideuterons (antihydrogen nuclei) that would be virtually impossible to produce by astrophysical secondaries at low energies. The GAPS experiment (General Antiparticle Spectrometer) flew a balloon-borne detector in Antarctica (2021) searching for this "smoking gun" signal.

Neutrino Searches from the Sun and Earth

WIMPs gravitationally captured by the Sun or Earth would accumulate in their cores and annihilate, producing neutrinos that escape and could be detected at neutrino telescopes. IceCube (South Pole) and ANTARES/KM3NeT (Mediterranean) search for an excess of high-energy neutrinos from the direction of the Sun and Earth's center. No signal has been found; limits on the WIMP-proton spin-dependent cross section from IceCube are competitive with direct detection experiments for heavy WIMPs.

Collider Production: Creating Dark Matter in the Lab

If WIMPs exist and interact via the weak force, they should be producible at particle colliders. The signature at the LHC (Large Hadron Collider): two protons collide, produce a Standard Model particle + a dark matter particle. The dark matter escapes undetected, but the recoiling Standard Model particle (a quark jet, gauge boson, Higgs, or photon) can be detected. These "mono-X" (monojet, mono-photon, mono-W/Z, mono-Higgs) searches look for events with large missing transverse momentum (energy apparently carried away by invisible particles).

Despite extensive searches, the LHC has found no evidence for WIMP production after Run 1 and Run 2. The limits on mediator masses (for simplified dark matter models) now exceed 2–3 TeV in many channels. The absence of new physics at the LHC has severely constrained supersymmetry and many WIMP models.

The HL-LHC (High-Luminosity LHC, operational ~2029) and proposed future colliders (100 TeV FCC-hh, ILC, CLIC) would extend the search substantially.

Alternative Dark Matter Candidates

The failure to find WIMPs despite the "WIMP miracle" has reinvigorated interest in alternative dark matter candidates spanning an enormous mass range:

Axions and Axion-Like Particles (ALPs)

The axion was originally proposed by Peccei and Quinn (1977) to solve the "strong CP problem" — the mystery of why the strong nuclear force does not violate CP symmetry even though the Standard Model allows it. The Peccei-Quinn mechanism introduces a new symmetry that is spontaneously broken, producing a pseudo-Goldstone boson — the axion.

Axions have:

• Mass: ~10⁻⁶–10⁻³ eV (microelectronvolt to millielectronvolt)
• Two-photon coupling: Axions couple to two photons, enabling them to convert to photons (and vice versa) in magnetic fields via the Primakoff effect: a + B → γ
• Cosmological production: Misalignment mechanism — if axions were produced non-thermally in the early universe, they could account for cold dark matter

The definitive axion search experiment is ADMX (Axion Dark Matter eXperiment) at the University of Washington: a high-Q microwave resonant cavity inside a strong superconducting magnet. When the cavity resonant frequency matches the axion mass, dark matter axions convert to microwave photons via the Primakoff effect. ADMX has excluded axion couplings over two decades of mass space in the 1–40 µeV range. Upgrades and next-generation experiments (HAYSTAC, ABRACADABRA, DMRadio, BREAD) are extending coverage across a much wider mass range.

Solar axion searches: Axions produced in the Sun (via Primakoff conversion of thermal photons) could be detected on Earth when they reconvert to photons in a strong laboratory magnet. The CAST experiment (CERN Axion Solar Telescope) and its successor IAXO (International Axion Observatory, planned) use long magnets pointed at the Sun to search for these solar axions.

Sterile Neutrinos

A natural extension of the Standard Model adds right-handed neutrinos that do not couple to the weak force (hence "sterile"). With masses of ~keV, they are viable warm dark matter candidates. They decay very slowly: ν_s → ν + γ, producing a faint X-ray line at energy E = m_s/2. The claimed detection of an unidentified line at 3.5 keV in X-ray spectra of galaxy clusters and M31 (Bulbul et al. 2014, Boyarsky et al. 2014) created enormous excitement as a potential sterile neutrino signal. However, subsequent high-resolution spectroscopy by XMM-Newton and Hitomi (before its loss in orbit) found no confirmation, and the XRISM observatory (launched 2023) has now resolved the 3.5 keV feature as atomic line emission from potassium, effectively ruling out the sterile neutrino interpretation.

Primordial Black Holes (PBHs)

If the early universe had density fluctuations large enough to locally exceed the critical density for gravitational collapse, regions might have collapsed into primordial black holes (PBHs) before nucleosynthesis began. PBHs are not particle physics dark matter — they are macroscopic objects — but they could contribute to (or constitute all of) dark matter.

Constraints from microlensing (MACHO, EROS surveys; Subaru HSC survey), gravitational wave observations (LIGO binary black hole mass spectrum), and CMB distortions rule out PBHs as all of dark matter except in a mass window of roughly 10¹⁷–10²³ g (asteroid mass range). In this window, PBHs have never emitted significant Hawking radiation (they are too massive) and are essentially transparent to light, making them very difficult to constrain. Current and future microlensing surveys (Roman Space Telescope) and gravitational wave detectors (LISA, Einstein Telescope) will probe this window.

Self-Interacting Dark Matter (SIDM)

Observations of small-scale structure — the "core-cusp problem" (galaxy halos show constant-density cores rather than the cuspy profile predicted by cold dark matter simulations), the "too-big-to-fail" problem (massive Milky Way satellite halos appear less concentrated than predicted), and the "missing satellites problem" (fewer small satellites than predicted) — have motivated dark matter models with self-interactions (dark matter-dark matter scattering via a dark force mediator).

SIDM with a cross section of ~1 cm²/g can thermalize galactic cores and create flat-density profiles. This is large compared to the electroweak-scale WIMP self-interaction, implying a dark sector mediator (dark photon, etc.). Constraints from the Bullet Cluster, galaxy cluster shapes, and N-body simulations bound the cross section. Multiple ongoing searches look for the dark sector mediator particle at fixed-target experiments and the LHC.

Ultra-Light Dark Matter (Fuzzy Dark Matter)

At the opposite extreme from heavy WIMPs, dark matter with a mass of ~10⁻²² eV would have a de Broglie wavelength of ~1 kpc — the scale of galactic cores. This fuzzy dark matter (also called ultra-light axion or wave dark matter) would form a Bose-Einstein condensate on galactic scales, with quantum pressure preventing core collapse and naturally producing flat-density cores in dwarf galaxies. The mass is fixed by the requirement that the quantum coherence scale equals observed dwarf galaxy core radii.

Evidence for or against fuzzy dark matter comes from:

• Lyman-alpha forest: The power spectrum of density fluctuations at z ~ 5 constrains the dark matter to have mass > ~10⁻²¹ eV — near the fuzzy dark matter sweet spot but ruling out the lightest models
• Stellar dynamics: The stellar kinematics of ultra-faint dwarf galaxies probe whether their halos have cores or cusps
• CMB constraints: Fuzzy dark matter changes the growth of structure in ways constrained by CMB observations

The Current Status and Future Direction

After decades of null results across direct detection, indirect detection, and collider searches, the WIMP paradigm is under stress but not dead. The parameter space has been reduced dramatically, but discovery potential remains. The future program includes:

• DARWIN/XLZD: A proposed 40–50-tonne liquid xenon experiment that would reach the neutrino fog — essentially completing the WIMP search down to the definitive limit set by neutrino backgrounds
• Next-generation CMB experiments (CMB-S4): Constrain light dark matter through CMB damping
• DUNE: A next-generation neutrino experiment also capable of detecting WIMPs via solar neutrino from WIMP annihilation
• Gravitational wave observatories: Probe compact dark matter (PBHs, boson stars) through their binary inspiral statistics

The discovery of dark matter would be one of the greatest achievements in the history of science. Its continued absence is forcing physicists to think more creatively about what dark matter might be — and whether the "WIMP miracle" was a coincidence we were seduced by, rather than a genuine hint from nature.

---

Related Reading

• What Is Dark Matter and Do We Know What It Is?
• What Is Dark Energy and Why Is the Universe Accelerating?
• What Are Cosmic Rays? High-Energy Particles from Space
• What Is Gravitational Lensing and How Does It Work?
• What Is the Cosmic Web? Large-Scale Structure of the Universe