Dark Matter: Hidden Mass
Linas JuozėnasShare
Knowledge Ark · Universe · Chapter 09 / Article 07
The universe has more to it.
Stars reveal where a galaxy shines. Their motions help reveal how much matter it contains. Across galaxies, colliding clusters, and the early universe, gravity points to a substantial component whose identity remains unknown.
The percentages are rounded, model-based cosmic averages—not the composition of every galaxy.[1]
How do you study something that does not shine?
Start with what it changes: the speed of an orbit, the path of light, or the pattern of galaxies across the sky.
Dark matter is a name for a physical explanation of those observations. Understanding the evidence and identifying its cause are related tasks, but they are not the same achievement. We can learn a great deal about an unseen component’s distribution before discovering what it is made of.
What does “dark matter” mean?
In the standard picture, dark matter contributes mass and gravity while producing no established electromagnetic signal of its own. It does not behave like an opaque cloud hiding the stars behind it. Its gravitational effects are how astronomers locate it.[2]
The word dark does not prove that gravity is its only interaction. Experiments search for weak interactions with ordinary particles or light, within existing observational limits.[3]
Could it simply be ordinary matter we have missed?
Cold gas, faint stars, planets, and stellar remnants all contribute mass. Astronomers include these when estimating the ordinary, or baryonic, component. But the primordial light-element abundances and the microwave background restrict how much baryonic matter the universe contains. Hiding more atoms does not by itself supply the much larger dark component required in ΛCDM.[4]
Dark matter and dark energy answer different questions. Dark matter clusters gravitationally in structures such as halos. Dark energy is the name for the component associated with accelerated cosmic expansion; in the simplest model, it is a cosmological constant.[2]
Why do rotation curves matter?
In 1933, Fritz Zwicky used the motions of galaxies in the Coma Cluster to argue for more mass than their light suggested. This early cluster-scale clue preceded the later rotation studies of individual galaxies.[5]
A rotation curve plots orbital speed against distance from a galaxy’s center. Optical measurements associated with Vera Rubin and colleagues, together with radio studies of hydrogen extending beyond bright stellar disks, helped establish that many outer rotation curves remain approximately flat.[6]
That matters because a compact concentration of mass predicts a declining orbital speed sufficiently far outside it. The bright part of a galaxy cannot automatically account for the measured speeds farther out. In Newtonian gravity, an extended dark halo provides the additional gravitational contribution.[7]
The orbital-speed argument, with its assumptions
For a circular orbit in a spherical Newtonian mass distribution, the speed depends on the mass enclosed within radius r. If that mass has effectively stopped increasing, vc falls as r−1/2. A flat curve instead implies M(<r) grows approximately in proportion to r across the measured range. A flattened galactic disk needs a more detailed calculation.[7]
The important comparison is quantitative: how much gravity should the measured ordinary matter produce, and how much is needed to explain the motion? A rotation curve establishes a discrepancy within a gravitational model; by itself, it does not identify a new particle.
What does the Bullet Cluster show?
Gravitational lensing lets astronomers infer mass from the distorted images of background objects. The reconstruction depends on gravity, source distances, and the observed distortions, rather than on the foreground material’s brightness. It measures the combined gravitational contribution along the line of sight.[8]
In the Bullet Cluster, two galaxy clusters have passed through one another. Their hot gas—the largest share of their ordinary matter—interacted strongly during the collision. The galaxies mostly passed through. The main lensing-inferred mass concentrations lie closer to the galaxies than to the displaced gas.[9]
This is a strong test of explanations based only on the observed ordinary matter. It does not reveal whether the extra component is a WIMP, an axion, or something else. Nor does one collision prove that dark matter has absolutely no interactions with itself.[10]
How does the early universe strengthen the case?
The microwave background measures a cosmic mixture
The cosmic microwave background records early fluctuations through patterns that include acoustic peaks. Ordinary matter interacted with radiation differently from dark matter, leaving different effects on those peaks. In the six-parameter ΛCDM model, Planck’s measurements favor roughly five times as much cold dark matter as baryonic matter.[1]
The two percentages at the beginning use different totals: matter alone versus all cosmic energy today. Dark energy contributes to the second total. Neither percentage is read directly from a photograph.[1]
From early fluctuations to galaxies
Dark matter that does not share the gas’s strong coupling to radiation can build gravitational structure differently in the early universe. Simulations follow its growth into halos and filaments, while gas falls in, cools, and forms stars. This connects the microwave background to the later cosmic web.[11]
Cold describes particles moving slowly compared with light when relevant structures grow. It does not mean today’s halo has an ordinary low thermometer temperature. Particles moving too rapidly can stream out of small density concentrations, suppressing structure on those scales.[12]
Small galaxies remain demanding tests. Star formation, gas outflows, and other ordinary-matter processes can change a halo’s central structure, so a mismatch with a dark-matter-only simulation is not automatically a failure of the whole model.[13]
What could dark matter be made of?
WIMPs: a possible relic of the hot early universe
Weakly interacting massive particles are a broad family of candidates. In a familiar thermal scenario, their creation and annihilation initially maintained equilibrium. As the universe expanded, reactions became too infrequent to maintain that equilibrium, leaving a relic population: freeze-out.[14]
The resulting abundance can be near the cosmologically inferred value for suitable masses and interaction strengths. That coincidence motivates searches; it is not a prediction that nature must contain a particular WIMP. Freeze-out is possible in more varied models than particles interacting through the familiar weak force alone.[14]
Axions: a different route to cold dark matter
The QCD axion was proposed to explain why the strong interaction appears to respect CP symmetry so closely—a symmetry linking a process to its mirror-image antimatter counterpart. Its possible role as dark matter is an additional consequence, depending on its properties and production history. Axions can be produced with sufficiently small momenta to act as cold dark matter despite their tiny mass.[15]
Axion-like particles are a broader category and need not solve the strong CP problem. Extremely light, wave-like or “fuzzy” dark-matter models belong to this wider discussion; they should not be treated as interchangeable with the conventional QCD axion.[15], [12]
Other particles—and primordial black holes
Proposals also include sterile neutrinos and particles in a hidden sector, with their own interactions. Different production histories can yield cold or warmer populations. A candidate must explain both how enough of it survived and how its properties fit astronomical and laboratory limits.[16]
Primordial black holes offer a different possibility: compact objects formed in the early universe, rather than from dying stars. Their contribution is constrained by searches including gravitational microlensing and Hawking evaporation. The limits depend strongly on mass and assumptions; the idea is neither an unrestricted solution nor excluded at every possible mass.[17]
How do scientists search for it?
| Approach | What is measured | What must be established |
|---|---|---|
| Direct detection | A possible interaction with a detector’s nuclei or electrons; axion searches also test conversion or other field effects. | The signal must be distinguished from radioactivity, neutrinos, and detector backgrounds. |
| Indirect detection | Possible products of annihilation or decay, such as gamma rays, cosmic rays, or neutrinos. | Known astrophysical sources can produce the same kinds of particles. |
| Accelerator searches | Signs that new particles were produced, including an imbalance in measured momentum or energy. | An unseen product must be distinguished from neutrinos and measurement effects; its cosmic abundance remains a further question. |
Each method tests particular interactions. A null result constrains the models an experiment can probe.[3]
Xenon experiments such as LZ, XENONnT, and PandaX look for small energy deposits. Other detector materials and lower thresholds probe different masses and interactions. An exclusion curve therefore answers a conditional question: which combinations of particle mass and interaction strength would have produced a signal under the analysis assumptions?[18]
For axions that couple to photons, a strong magnetic field can enable conversion into an electromagnetic signal. Resonant searches tune through narrow frequency ranges, testing particular masses and couplings rather than the entire axion idea at once.[19]
In its July 2026 report, ADMX stated that no candidate spectral line passed all tests in the reported search. The result constrains selected masses and couplings; it does not exclude every possible axion.[19]
Is the “neutrino floor” a hard limit?
Neutrinos can create recoils resembling some dark-matter signals. Uncertainty in that background can make further improvements much slower—the neutrino fog. It is not a universal wall beyond which detection becomes impossible. Greater exposure and information such as recoil direction, timing, or different target materials can help, depending on the signal.[20]
What do the latest results establish?
LZ has reported an unexplained candidate event
On 1 September 2026, the LZ collaboration announced a single unusual event in an analysis of 220 live days of data collected during 2023–2024. The search extended to higher-energy nuclear recoils and a broader set of possible WIMP interactions. The collaboration explicitly did not claim a dark-matter discovery.[21]
The reported global significance is 2.6σ, after accounting for the range of possibilities searched. That describes tension with the specified background model; it is not a probability that the event “is dark matter.” Additional data and independent scrutiny are needed to determine its origin.[22]
This result makes it especially important to separate three statements: an event was measured; the modeled backgrounds have difficulty explaining it; a dark-matter interaction caused it. The first two can motivate further investigation without establishing the third.
Mapping progress is a different kind of discovery
In January 2026, researchers published a detailed COSMOS-Web weak-lensing mass map using the James Webb Space Telescope. It reconstructed structure from background-galaxy shapes across roughly half a square degree. Such maps improve our picture of how matter is distributed; they do not photograph dark-matter particles or identify their nature.[8]
Accelerator searches also continue through analysis and detector development. CERN began Long Shutdown 3 in June 2026 to prepare the High-Luminosity LHC. Its future measurements will test additional interactions and parameter ranges, without guaranteeing that a dark-matter particle lies within reach.[23]
Could gravity work differently?
The inference of extra matter uses a theory of gravity. It is therefore reasonable to test that theory as well. Modified Newtonian Dynamics, or MOND, introduces different behavior at very small accelerations and captures important regularities in galaxy rotation data.[24]
One striking observation is the radial acceleration relation: measured accelerations in disk galaxies closely track those predicted by Newtonian gravity from their ordinary matter, even where the two differ greatly in magnitude. Any successful explanation of galaxies needs to account for that organization in the data.[25]
Explaining galactic orbits is only part of the task. A complete theory must also predict lensing, cluster dynamics, cosmic expansion, and the growth of fluctuations. The original nonrelativistic MOND prescription does not by itself provide all that machinery.[24]
It would be too broad to say every alternative fails the microwave background. For example, Skordis and Złośnik’s relativistic theory demonstrated agreement with CMB and matter-power spectra on linear cosmological scales. That result deserves assessment on its own terms; it does not establish success for every observation or every modified-gravity proposal.[26]
The productive comparison asks how a fully specified model performs across independent measurements, including its extra assumptions and unresolved problems. A good fit to one phenomenon is valuable evidence, but it does not settle the whole question.
What about holography or a simulated universe?
Holography is a precise theoretical idea
In examples such as AdS/CFT, a gravitational description in a volume is related to a quantum-field description with fewer spacetime dimensions. This is a proposed equivalence between descriptions of particular physical systems. It is not a literal image projected onto our surroundings, and it does not by itself establish what dark matter is.[27]
Verlinde’s emergent-gravity proposal draws on ideas about entropy and microscopic information to derive an additional gravitational response. It attempts to explain effects normally assigned to dark matter. Evaluating it requires testing those derivations and predictions; calling an explanation “holographic” cannot replace that work.[28]
Could dark matter be part of a simulation or cosmic experiment?
Such scenarios can be explored as philosophy or fiction. Bostrom’s simulation argument is conditional on assumptions about future civilizations and simulated minds; it is not an astronomical detection or a model of galactic mass.[29]
Imagining a programmed halo—or a universe built as someone’s experiment—does not yet explain its measured properties. To become a competing scientific explanation, the idea would need a concrete mechanism and predictions that distinguish it from other models. An unanswered question gives us room to investigate; it does not give every imagined answer equal evidential support.
The dark-matter problem is compelling precisely because there is so much to explain. Galaxy motions, lensing, and early-universe patterns give new ideas demanding targets. The next advance may identify a particle, reveal an unexpected interaction, or change part of our theoretical picture. Its strength will come from the evidence it connects.
Sources and further reading
Observational research, particle-physics studies, and original theoretical proposals. Sources checked in September 2026. All three illustrations are explanatory; the rotation curves are a calculated toy model.
- Planck Collaboration — Planck 2018 results. VI. Cosmological parametersBase-ΛCDM fits distinguish baryonic and cold-dark-matter densities through CMB acoustic patterns and lensing.
- NASA Science — Dark MatterAn introduction to gravitational evidence and the distinction between dark matter and dark energy.
- Boveia and Doglioni (2018) — Dark Matter Searches at CollidersExplains how collider, direct and indirect searches constrain different processes and require model-dependent comparisons.
- Fields, Molaro and Sarkar — Particle Data Group 2025: Big Bang NucleosynthesisPrimordial light-element abundances constrain ordinary matter independently of stellar brightness and complement the CMB matter census.
- Jo Bovy — Dynamics and Astrophysics of Galaxies, §6.1: Dark matter in the Coma clusterZwicky’s 1933 Coma-cluster mass argument preceded modern galaxy rotation studies; later work revised the numerical discrepancy.
- Carnegie Science — Vera Rubin: Opening doors to dark matter and women in STEMInstitutional history and publication bibliography establish the 1970 Andromeda study and 1980 three-author galaxy survey.
- Jo Bovy — Dynamics and Astrophysics of Galaxies, §8.2: Rotation curves of external galaxiesHow projected velocities and stellar light constrain rotation curves, with distinct spherical and disk mass calculations.
- Scognamiglio et al. (2026) — An ultra-high-resolution map of (dark) matterCOSMOS-Web weak-lensing measurements reconstruct projected mass from distorted background-galaxy shapes.
- Clowe et al. (2006) — A direct empirical proof of the existence of dark matterWeak-lensing peaks follow galaxies rather than the dominant hot-gas baryons in the merging Bullet Cluster.
- Markevitch et al. (2004) — Direct constraints on the dark matter self-interaction cross-sectionThe Bullet Cluster constrains self-interactions under modeling assumptions; it does not establish an exactly zero cross-section.
- ESA Planck — History of cosmic structure formationDark-matter perturbations precede baryonic collapse; particle free streaming shapes which cosmic structures can develop.
- Mikhail Shaposhnikov (2025) — Cosmology, CERN Yellow ReportsExplains cold, warm, and hot dark matter through early velocities, free streaming, and structure formation.
- Sales, Wetzel and Fattahi (2022) — Baryonic solutions and challenges for cosmological models of dwarf galaxiesDwarf-galaxy tests require gas and stellar feedback alongside dark-matter dynamics; several observational and modeling tensions remain.
- Particle Data Group (2026 edition) — Dark Matter, Baudis and ProfumoExplains thermal freeze-out, alternative production mechanisms and the limits of the weak-scale WIMP argument.
- Particle Data Group (2026 edition) — Axions and Other Similar ParticlesDistinguishes QCD axions from broader ALPs and explains nonthermal production of cold dark matter.
- Jonathan L. Feng (2010) — Dark Matter Candidates from Particle Physics and Methods of DetectionReviews sterile-neutrino and hidden-sector candidates, linking production histories to relic abundance and observational constraints.
- Gorton and Green (2024) — How open is the asteroid-mass primordial black hole window?Shows an asteroid-mass primordial-black-hole dark-matter window can survive constraints, depending on the mass distribution.
- Billard et al. (2021/2022) — Direct Detection of Dark Matter: APPEC Committee ReportExplains detector targets, thresholds, backgrounds and the model assumptions behind dark-matter exclusion curves.
- ADMX / Andrew Sonnenschein (23 July 2026) — Searching for Axion Dark Matter with ADMXThe collaboration's July 2026 report describes axion-photon conversion searches and states no candidate passed all tests.
- O'Hare (2021) — Fog on the horizon: a new definition of the neutrino floor for direct dark matter searchesShows that neutrino backgrounds complicate nuclear-recoil dark-matter searches without imposing an absolute discovery barrier.
- Berkeley Lab (1 September 2026) — LZ Sees Surprising Result in Search for Dark MatterOfficial announcement identifies an intriguing LZ event while explicitly declining to claim a dark-matter discovery.
- LZ Collaboration (2026) — Search for dark matter particle interactions in an extended nuclear recoil energy windowPrimary LZ analysis quantifies the candidate event and its 2.6-sigma global statistical significance.
- CERN (29 June 2026) — CERN bids farewell to the LHC and enters Long Shutdown 3CERN announced the shutdown and installation program preparing the High-Luminosity LHC.
- Famaey & McGaugh (2012) — Modified Newtonian Dynamics (MOND): Observational Phenomenology and Relativistic ExtensionsExplains residual cluster mass discrepancies and why galactic MOND phenomenology requires a broader relativistic theory.
- McGaugh, Lelli & Schombert (2016) — The Radial Acceleration Relation in Rotationally Supported GalaxiesGalaxy rotation data show tight baryon–acceleration correlations anticipated by MOND, without uniquely selecting a theory.
- Skordis and Złośnik (2021) — A New Relativistic Theory for Modified Newtonian DynamicsA relativistic MOND theory reproducing CMB and linear matter-power spectra illustrates why alternatives must be assessed individually.
- Maldacena (1998) — The Large N Limit of Superconformal Field Theories and SupergravityThe foundational AdS/CFT proposal relates particular gravitational and quantum-field descriptions.
- Verlinde (2017) — Emergent Gravity and the Dark UniverseAn information-inspired proposal for additional gravitational effects usually attributed to dark matter.
- Bostrom (2003) — Are You Living in a Computer Simulation?A conditional philosophical argument about simulated observers, rather than a particle or gravitational model.
The Nature of Space and Time
- Special Relativity: Time Dilation and Length Contraction
- General Relativity: Gravity as Curved Spacetime
- Quantum Mechanics: Wave-Particle Duality
- Quantum Field Theory and the Standard Model
- Black Holes and Event Horizons
- Wormholes and Time Travel
- Dark Matter: Hidden Mass · You are here
- Dark Energy: Accelerating Expansion
- Gravitational Waves
- Toward a Unified Theory