Dark Matter

Full definition

Imagine an ice rink in complete darkness. You slide a puck; it veers sharply, bounces back, zigzags. You infer invisible obstacles — walls, posts, other skaters — simply by watching their effect on its trajectory. Dark matter is exactly that, at cosmic scale: a mass we cannot see, but whose gravitational pull on everything luminous we can measure precisely.

More formally, dark matter is a form of matter that neither emits, absorbs nor reflects light — essentially no interaction with the electromagnetic field. It is therefore not 'dark' in a visual sense, nor absorbing like a blackbody: it is simply transparent. It is detected only through gravity. The most precise cosmological measurements (Planck satellite, 2018) give Ω_m ≈ 0.315 for total matter, of which only Ω_b ≈ 0.049 is ordinary baryonic matter. The rest — nearly 27 % of the Universe's energy budget — is non-baryonic dark matter.

The historical lineage begins as early as 1933, when Fritz Zwicky measured the velocity dispersion of galaxies in the Coma cluster and realised one needs ~400 times more mass than the luminous galaxies provide to keep the cluster from flying apart. The idea lay dormant for forty years. It was revived in the 1970s when Vera Rubin and Kent Ford measured rotation curves of spiral galaxies (M31 first): outer stars orbit as fast as inner ones, instead of slowing down like a Keplerian system. Conclusion: every galaxy is embedded in an invisible halo, several times more massive than its luminous component. Dark matter is now a pillar of the standard cosmological ΛCDM model (Lambda Cold Dark Matter).

Observational evidence

Four converging lines of evidence operate at all scales.

Galactic rotation curves (galaxy scale). In a spiral galaxy, stellar circular velocity should drop as 1/√r beyond the bulge. Instead we see a near-flat plateau out to the edges — a spherical invisible halo dominating the total mass.

Cluster dispersion (Mpc scale). Zwicky 1933 revisited: the virial theorem applied to cluster galaxy velocities demands 5-10 times more mass than the sum of visible galaxies + X-ray-detected hot gas.

Gravitational lensing. Bending of background-galaxy light by clusters (Einstein rings, arcs) maps the mass distribution directly — and reveals far more mass than luminous matter accounts for. The Bullet Cluster (Clowe et al. 2006) is the iconic image: two colliding clusters, the hot gas (ordinary matter) stalls at the center by ram pressure, but the gravitational lens shows the bulk of the mass sailed through unimpeded — so it can't be gas.

Cosmic Microwave Background and large-scale structure. The acoustic peaks of the CMB (measured by COBE 1992, WMAP 2003, Planck 2013-2018) are only consistent with ~27 % cold non-baryonic matter. Without dark matter, galaxies simply couldn't have formed in 13.8 billion years — N-body simulations (Millennium Simulation, 2005) confirm this quantitatively.

Theoretical candidates

We still don't know what dark matter is. Several families of candidates are actively explored.

WIMPs (Weakly Interacting Massive Particles). Massive particles (10 GeV to 10 TeV) interacting only via gravity and the weak force. Long-time favourite, motivated by supersymmetry (neutralino). Experiments XENONnT (Italy), LZ (South Dakota), PandaX (China) hunt for their collisions on liquid xenon — no convincing detection so far, sharply shrinking the available parameter space.

Axions. Ultra-light particles (µeV to meV) originally invented to solve the strong-CP problem in QCD. Detectable via photon conversion in intense magnetic fields (ADMX at Fermilab, HAYSTAC at Yale). A rising favourite since 2020.

Sterile neutrinos. Cousins of Standard-Model neutrinos, with no weak interaction, keV-scale mass. Would leave an X-ray signature. A 3.5 keV line detected in clusters in 2014 remains debated.

MACHOs (Massive Astrophysical Compact Halo Objects). Brown dwarfs, cold white dwarfs, primordial black holes. Microlensing surveys (EROS, MACHO, OGLE) have ruled out MACHOs as the dominant component of galactic dark matter. Primordial black holes (30 M☉) have returned to favour since LIGO's 2015 gravitational-wave detections.

Gravitational alternatives (MOND, TeVeS, modified gravity). Rather than postulating invisible matter, modify gravity at low accelerations. MOND (Milgrom, 1983) reproduces rotation curves well, but struggles with large-scale structure and the CMB.

How do we hunt for it?

Three complementary strategies are in play.

Direct detection. Ultra-pure detectors are buried under kilometres of rock to filter cosmic noise, hoping for a dark-matter particle to ricochet off an atomic nucleus. XENONnT (3.5 tonnes of liquid xenon at Gran Sasso, Italy), LZ (10 tonnes, South Dakota, 2022), DAMA/LIBRA (Italy, claims an unconfirmed annual modulation). No robust signal so far — constraints now exclude much of the WIMP parameter space.

Indirect detection. If dark matter annihilates somewhere (galactic centre, Sun, dwarf galaxies), it produces detectable γ-rays, neutrinos or antimatter. Instruments: Fermi-LAT (NASA, 2008) for γ-rays, IceCube (South Pole, km³ of ice) for neutrinos, AMS-02 (ISS, 2011) for antimatter. An excess of high-energy positrons is seen, but can be explained by nearby pulsars.

Collider production. At the LHC (CERN), physicists attempt to create dark-matter particle pairs in proton-proton collisions, signed by 'missing energy'. Nothing found yet.

Cosmological missions. The European Euclid space telescope, launched on July 1, 2023, is mapping dark-matter distribution via weak gravitational lensing over a third of the sky — by 2030 it will deliver the most precise 3D map ever. Its American counterpart, the Nancy Grace Roman Space Telescope (2027), will provide independent confirmation.

Not to be confused with

Dark matter is commonly confused with several cosmological neighbours.

Dark energy. Completely distinct. Dark matter slows cosmic expansion through its attraction — dark energy accelerates it through negative pressure. They cohabit the cosmic budget (27 % dark matter + 68 % dark energy + 5 % ordinary matter) but share neither nature, effects nor observational signatures.

Black hole. A black hole is a localized compact object, not a diffuse cosmological component. Only the primordial black hole hypothesis (formed in the first fractions of a second after the Big Bang) suggests they could make up a fraction of dark matter — but observational constraints cap this fraction tightly.

Antimatter. Antimatter is thoroughly identified (positrons, antiprotons, antihelium nuclei detected by AMS-02). It interacts normally with light and annihilates on contact with ordinary matter. It has nothing to do with dark matter.

Dark molecular clouds. Dark nebulae (Coalsack, Horsehead) absorb background starlight — they are optically dark but are ordinary baryonic matter, fully accounted for in cosmic inventories.

Standard-model neutrinos. The three known neutrino flavours exist, have been detected since 1956, and form hot (relativistic) matter — they can only make up a marginal fraction (< 1 %) of dark matter, which must be overwhelmingly cold (non-relativistic) to explain structure formation.