Dark Matter: What It Is and How We Detect It

Dark matter is the dominant non-luminous component of the universe's mass-energy budget, making up approximately 27% of the total content of the cosmos according to the Planck Collaboration's 2018 results derived from observations of the cosmic microwave background. It neither emits, absorbs, nor reflects electromagnetic radiation, making it invisible to every optical, radio, X-ray, and gamma-ray instrument ever built — yet its gravitational influence shapes every galaxy, filament, and void in the observable universe. This page covers what dark matter is, how physicists characterize its behavior, what candidate particles or objects have been proposed, and how detection experiments attempt to identify something that refuses to interact with light.

Definition and scope

Dark matter is defined operationally as a form of matter that accounts for observed gravitational effects — galaxy rotation curves, gravitational lensing, large-scale structure formation — that cannot be explained by the mass of all visible (baryonic) matter combined. The term "dark" refers strictly to its electromagnetic invisibility, not to any exotic or negative property.

The quantitative case is compelling. Vera Rubin and Kent Ford's 1970 measurements of the Andromeda galaxy's rotation curve (published in the Astrophysical Journal, 1970) showed that stars at the galaxy's outer edges orbit at speeds that should be impossible if only visible mass were present. Stars far from galactic centers travel at roughly the same orbital velocity as stars near the core — a flat rotation curve — rather than declining as Newtonian gravity would predict for an isolated disk. This flatness implies a surrounding halo of unseen mass extending well beyond the luminous disk.

Dark matter is formally distinguished from dark energy, which is the separate phenomenon responsible for the accelerating expansion of the universe and constitutes approximately 68% of the total energy budget. Baryonic matter — protons, neutrons, electrons, and everything built from them — accounts for only about 5% of the total.

Within the broader framework described on the cosmology home resource at /index, dark matter sits at the intersection of particle physics, astrophysics, and observational cosmology, and remains one of the most active unresolved problems in fundamental science.

Core mechanics or structure

Dark matter behaves gravitationally like ordinary matter but does not interact via the electromagnetic force and almost certainly does not interact via the strong nuclear force. Whether it interacts via the weak nuclear force depends on which candidate model is considered — this distinction drives the entire experimental program.

The standard treatment embeds dark matter within the Lambda-CDM model, where "CDM" stands for Cold Dark Matter. "Cold" means the particles were moving non-relativistically at the epoch of matter-radiation decoupling (approximately 380,000 years after the Big Bang), which is required to match the observed pattern of large-scale structure. Hot dark matter — composed of relativistic particles like neutrinos — would have washed out small-scale structure through free-streaming, producing a universe with fewer small galaxies than observed.

Dark matter forms roughly spherical halos around galaxies and galaxy clusters. N-body simulations, such as the Millennium Simulation run by the Virgo Consortium (described in Springel et al. 2005, Nature), demonstrate that these halos have density profiles consistent with a Navarro-Frenk-White (NFW) profile: density scales as 1/(r × (r + r_s)²), where r_s is a characteristic scale radius. The NFW profile predicts a central density "cusp," a detail that remains contested against observational evidence of flatter "cores" in dwarf galaxies.

Causal relationships or drivers

The gravitational evidence for dark matter arises from three independent lines of observation, each probing different physical scales.

Galaxy rotation curves reveal excess centripetal acceleration at large galactic radii. Rubin and Ford's Andromeda data, later extended to over 200 spiral galaxies by subsequent surveys, established flatness as a universal feature rather than an anomaly.

Gravitational lensing provides a geometry-independent mass map. The Bullet Cluster (1E 0657-558), observed by the Chandra X-ray Observatory (Clowe et al. 2006, Astrophysical Journal Letters), showed two merging galaxy clusters where the hot X-ray-emitting gas (baryonic matter) was displaced from the gravitational mass center inferred by weak lensing. The mass followed the galaxies, not the gas — direct evidence that a non-baryonic, non-self-interacting mass component exists.

Cosmic structure formation provides the third pillar. The Planck satellite measurements of temperature anisotropies in the cosmic microwave background encode a precise ratio of dark matter to baryonic matter. Without dark matter's gravitational potential wells forming before recombination, baryonic matter would not have had sufficient time to collapse into the filaments and voids of the cosmic web seen today.

Dark matter does not arise from measurement error in these three domains simultaneously. Modified Newtonian Dynamics (MOND), proposed by Mordehai Milgrom in 1983, can reproduce rotation curves but fails to account for the Bullet Cluster lensing offset and CMB power spectrum without the addition of a separate dark matter-like component.

Classification boundaries

Dark matter candidates fall into distinct theoretical categories that differ in mass scale, interaction type, and detection strategy.

WIMPs (Weakly Interacting Massive Particles) are the historically dominant candidate class, predicted to have masses between approximately 10 GeV and 1 TeV and to interact via the weak force. They arise naturally in supersymmetric extensions of the Standard Model.

Axions are ultralight particles (mass range ~10⁻⁵ to 10⁻³ eV) proposed by Roberto Peccei and Helen Quinn in 1977 to resolve the strong CP problem in quantum chromodynamics. Their cosmological abundance could account for dark matter if produced non-thermally in the early universe.

Sterile neutrinos are hypothetical right-handed neutrinos that mix minimally with Standard Model neutrinos. Proposed mass ranges span keV to MeV scales.

Primordial black holes (PBHs) are a non-particle candidate — macroscopic remnants from density fluctuations in the early universe. Gravitational microlensing surveys constrain PBHs from accounting for all dark matter across most mass windows above ~10⁻¹⁰ solar masses.

Self-interacting dark matter (SIDM) is a class of models where dark matter particles interact with each other (but not baryons), proposed to resolve the core-cusp and too-big-to-fail problems identified in CDM simulations.

Tradeoffs and tensions

The CDM paradigm is highly successful on large scales but faces persistent small-scale tensions. The "missing satellites problem" — CDM simulations predict hundreds of subhalos around a Milky Way-sized galaxy, while only roughly 50 dwarf satellite galaxies have been confirmed in the Local Group — remains under active investigation. Baryonic feedback processes (supernovae, stellar winds) may suppress small halo formation, but the degree of required feedback is itself debated.

The "core-cusp problem" contrasts CDM's predicted central density cusps with the flat cores observed in dwarf and low-surface-brightness galaxies. SIDM models and baryonic feedback each offer partial resolution, but no single mechanism is universally accepted.

On the experimental side, WIMP searches have progressively excluded large regions of parameter space. The LUX-ZEPLIN (LZ) experiment, reporting in 2022, set the most stringent spin-independent WIMP-nucleon cross-section limit at approximately 9.2 × 10⁻⁴⁸ cm² for a 36 GeV WIMP mass (LZ Collaboration, 2022, arXiv:2207.03764). This null result pressures the WIMP paradigm without eliminating it, since lower cross-sections remain viable.

The Euclid mission, launched in 2023, and the Rubin Observatory LSST will map weak lensing across billions of galaxies, providing new constraints on dark matter's spatial distribution and self-interaction cross-section.

Common misconceptions

"Dark matter is just undetected ordinary matter (gas, dust, dim stars)."
Baryonic matter in all forms — including cold molecular clouds, brown dwarfs, and faint stars — has been accounted for through multiple independent inventories. Big Bang nucleosynthesis constrains the total baryon density to approximately 4–5% of the critical density, far below the ~27% attributed to dark matter (Particle Data Group, Review of Particle Physics, 2022).

"Dark matter is a theoretical placeholder with no real evidence."
The Bullet Cluster observation provides spatially resolved, direct evidence that mass is located where no gas or stars are present. This is not an inference from a single technique but a geometric measurement from X-ray emission maps and independent lensing mass reconstructions.

"MOND eliminates the need for dark matter."
MOND is an empirical modification of gravitational dynamics that succeeds at galaxy scales but requires a relativistic extension (TeVeS, developed by Jacob Bekenstein in 2004) to be applied cosmologically. Even TeVeS requires some form of dark matter to match CMB observations — it does not eliminate the substance, only modifies the force law.

"Neutrinos are dark matter."
Standard Model neutrinos are "hot" — they were relativistic at decoupling — and their combined mass is constrained below approximately 0.12 eV by Planck data. This mass is far too low to constitute the observed dark matter fraction, and their free-streaming would erase the small-scale structure that is observed.

Detection methods: a structured sequence

The experimental program to detect dark matter is organized into three logically distinct approaches, each operating independently.

  1. Direct detection — Underground detectors filled with target nuclei (xenon, germanium, argon) wait for a dark matter particle to scatter off a nucleus, depositing recoil energy as phonons, ionization, or scintillation. Experiments: LUX-ZEPLIN (LZ), XENONnT, PandaX-4T. Requires shielding from cosmic rays, hence deep placement (LZ operates at 4,850 feet underground at the Sanford Underground Research Facility in Lead, South Dakota).

  2. Indirect detection — Telescopes search for the products of dark matter self-annihilation or decay: gamma rays (Fermi-LAT satellite), antimatter (AMS-02 on the International Space Station), or neutrinos (IceCube Neutrino Observatory at the South Pole). The signal is an excess above known astrophysical backgrounds at characteristic energies or spatial distributions.

  3. Collider production — Particle accelerators attempt to produce dark matter particles directly. At the Large Hadron Collider (LHC), "missing transverse energy" signatures in collision events indicate a particle that escaped the detector undetected. No confirmed dark matter signal had been reported through LHC Run 3 as of the 2022 PDG Review.

  4. Axion-specific searches — Axions are targeted by dedicated experiments distinct from WIMP searches. The Axion Dark Matter Experiment (ADMX) at the University of Washington uses a resonant microwave cavity in a strong magnetic field to convert axions to detectable photons via the Primakoff effect. ADMX has excluded axion masses between approximately 2.66 and 3.31 μeV in its core sensitivity range (ADMX Collaboration, Physical Review Letters, 2020).

  5. Gravitational and astrophysical probes — Independently of particle physics, surveys using gravitational lensing and baryon acoustic oscillations measure the dark matter power spectrum. The Sloan Digital Sky Survey has contributed galaxy clustering data that constrains dark matter's clustering properties on scales from 1 Mpc to over 1,000 Mpc.

Reference table: dark matter candidates compared

Candidate Mass range Interaction type Primary detection method Status (2022 PDG)
WIMP 10 GeV – 1 TeV Weak + gravity Direct recoil, collider Not detected; large parameter space excluded
Axion 10⁻⁵ – 10⁻³ eV Electromagnetic (via coupling) Resonant cavity (ADMX) Not detected; narrow mass range probed
Sterile neutrino keV – MeV Gravity only (minimal mixing) X-ray line searches Not confirmed; 3.5 keV line disputed
Primordial black hole 10⁻¹⁰ – 10² M☉ Gravity Microlensing (MACHO, EROS surveys) Excluded as sole component in most mass windows
SIMP (Strongly Interacting Massive Particle) MeV scale Strong-sector dark force Collider, astrophysical cooling Theoretical; largely unconstrained
Fuzzy dark matter (ultralight axion) ~10⁻²² eV Gravity + quantum pressure Lyman-alpha forest, dwarf galaxies Active area; tensions at small scales

Sources: Particle Data Group Review of Particle Physics 2022; Bertone, Hooper & Silk, Physics Reports 2005.

References

The law belongs to the people. Georgia v. Public.Resource.Org, 590 U.S. (2020)