Coma cluster
Fritz Zwicky measures surprisingly large velocity dispersions, inferring a mass-to-light ratio near 500 and naming 'dunkle Materie'.
arXiv:1801.08128 · astro-ph.CO
The Search for Dark Matter
Laura Baudis · Physik Institut, University of Zurich. A century-old problem: our cosmos is dominated by non-baryonic matter that holds galaxies together and shapes structure on the largest scales — yet does not interact with light.
drag to orbit the atom
01 — Observation
Evidence for dark matter
From Zwicky's Coma cluster velocities to the acoustic peaks of the CMB, every probe points the same way: there is far more mass than light.
Flat rotation curves
Vera Rubin, Albert Bosma and 21 cm radio surveys show galactic rotation curves stay flat well beyond the optical radius.
Planck power spectrum
Fluctuations of one part in 100,000 fix the budget: a flat, accelerating Universe of 5% baryons, 27% cold dark matter, 68% dark energy.
Composition of the cosmos
- Baryonic matter5%
stars, planets, gas
- Cold dark matter27%
invisible scaffolding
- Dark energy68%
accelerating expansion
MOND and its relativistic descendants explain flat rotation curves elegantly, but fail on clusters and on the observed CMB pattern. They also predict photons and gravitational waves travelling on different geodesics — an ~800 day delay over 130 million light years, against the 1.7 s actually measured in GW170817.
Primordial nucleosynthesis fixes the abundance of light elements far below the total matter density, confirmed precisely by the CMB. Microlensing searches also rule out MACHOs — stellar remnants, faint stars, planets and black holes — as a substantial part of the Milky Way halo.
02 — Theory
Dark matter candidates
Allowed masses span 10⁻²² eV to 10¹⁵ GeV and interaction strengths spread over some 60 orders of magnitude. Two islands stand out, because neither was invented to solve this problem.
WIMPs
few GeV – 10⁵ GeVWeakly interacting massive particles. If they were in equilibrium in the early plasma, an annihilation cross section near 3 × 10⁻²⁶ cm³s⁻¹ reproduces the measured Ω_DM ≈ 0.27 — the 'WIMP miracle'.
Axions
µeV scaleBorn from the Peccei–Quinn solution to the strong CP problem. Couplings scale inversely with the decay constant f_a, so lighter axions are ever more weakly coupled: m_a f_a ≈ m_π f_π.
Neutrinos
< 2 eV — ruled outReal, dark and abundant, but far too light. As hot dark matter their velocities erase small-scale structure, contradicting the observed formation of galaxies.
03 — Experiment
Detecting WIMPs
Three complementary attacks: direct scattering off nuclei deep underground, indirect searches for annihilation products, and production at colliders.
Direct
Ultra-low-background detectors underground look for keV-scale nuclear recoils. Liquid xenon TPCs (XENON1T/nT, LUX/LZ, PandaX) lead above 6 GeV; cryogenic and CCD detectors (CRESST, SuperCDMS, EDELWEISS, DAMIC) probe below.
Indirect
Fermi-LAT, AMS-02, Cherenkov arrays and neutrino telescopes hunt gamma rays, positrons, antiprotons and neutrinos from annihilation — in the galactic centre, dwarf spheroidals or the Sun.
Colliders
At the LHC the signature is missing energy recoiling against a jet, photon or Z boson. Intense searches so far show no non-standard-model particle.
0.4 GeV/cm³
local dark matter density near the Sun
220 km/s
mean WIMP velocity through the halo
10⁵ cm⁻²s⁻¹
WIMP flux on Earth for a 100 GeV particle
< 1 / t / yr
expected events per tonne of detector
46 orders
of magnitude in allowed candidate mass
13.8 Gyr
minimum lifetime a candidate must survive
The Earth acts as a filter against cosmic rays. At Gran Sasso, 1400 m of Corno Grande rock leaves 1 muon per square metre per hour; at JinPing in China, 2400 m of overburden reduce that to 2 muons per square metre per 10 days.
As detectors reach the multi-tonne scale, material radioactivity fades and neutrinos become the irreducible background: pp and ⁷Be solar neutrinos scattering off electrons, ⁸B neutrinos mimicking 5–6 GeV WIMPs, atmospheric and diffuse supernova neutrinos mimicking heavier ones. Sensitivity must still improve by more than a factor 100 before they appear — and then they become a signal in their own right.
3.2 tonnes of liquid xenon, 2 t of them in the time projection chamber, watched by 248 photomultipliers tuned to 178 nm scintillation light. A uniform drift field extracts electrons into the vapour phase, producing a delayed secondary signal that yields 3D position reconstruction.
04 — Experiment
Axion detection
Nearly every axion search exploits the two-photon coupling: in a strong magnetic field, an axion can convert into a photon.
Light shining through walls
Laser photons convert to axions in a transverse magnetic field, cross an opaque barrier, and reconvert to photons beyond it. ALPS-II at DESY leads in sensitivity.
Helioscopes
CAST at CERN — using a refurbished LHC dipole — and the proposed IAXO track the Sun, converting solar axions of ~4.2 keV into detectable X-rays.
Haloscopes
Sikivie's 1983 scheme: a tunable microwave cavity in a strong field resonates at the axion rest mass. ADMX has been scanning QCD axion parameter space since 2016.
For a 0.5 µeV axion and a local density of 0.4 GeV cm⁻³, the flux on Earth is roughly 10²¹ cm⁻²s⁻¹.
05 — Outlook
A vast sea of the unknown
“We live in a vast sea of dark matter, but decades of intensive research have mostly established what it is not.”
Understanding its nature may reveal new symmetries, new fundamental particles and new forces — while opening a new field: dark matter astronomy. No particle has been identified yet, but sensitivities have reached the point where a discovery might be imminent, albeit certainly not guaranteed.