The Dark Ages and First Structures

1. What defines the Dark Ages?

Between neutral atoms and the first stars

The Dark Ages followed recombination, when electrons became bound in neutral atoms and the free-electron population fell sharply. Around 380,000 years after the Big Bang, most photons in what we now call the cosmic microwave background experienced their last scattering.

Ordinary matter was then predominantly hydrogen and helium gas. Helium accounted for roughly a quarter of this ordinary matter by mass—not a quarter of the universe’s total contents. Dark matter was a separate, major component of the matter distribution.^[1][5]

The starless interval ended with the emergence of the first stars. An age of roughly 100–200 million years is a useful illustrative range for early star formation in theoretical models, rather than a precisely observed starting date. Different regions developed at different rates; cosmic dawn was not a switch thrown everywhere simultaneously.^[7]

“Dark” does not mean without radiation

The background radiation did not vanish when atoms formed. It continued filling space, cooling and shifting toward longer wavelengths as the universe expanded. What was missing was a population of stars, galaxies, and other bright objects that could illuminate their surroundings.^[2]

2. Cooling of the universe after recombination

There was no single temperature for the whole era

Near the beginning of the Dark Ages, the background radiation was still about 3,000 K. It reached hundreds and then tens of kelvins only as expansion continued. Those lower temperatures describe later stages, not an abrupt change at recombination.^[1][3]

It also matters what is cooling. The temperature of the radiation background and the temperature of the gas are different quantities. They initially remained close because residual free electrons continued transferring energy between radiation and matter. Later, that exchange became too weak to keep them together.^[3]

A small electron population with an important role

Recombination left a small residue of free electrons. In standard calculations, a representative later residual abundance is around two free electrons per ten thousand hydrogen nuclei, although it varies with time and environment. This residue remained because recombination did not proceed to completion as the universe expanded—not because the gas had to stay hot.

These electrons helped drive reactions that produced small quantities of molecular hydrogen, H₂. That chemistry would become important inside the first collapsing clouds, where molecules could help the gas lose energy. A nearly neutral universe was not a chemically inactive one.^[14]

3. Growth of density fluctuations

The seeds were older than the Dark Ages

The early universe was very smooth on large scales, but its matter distribution contained small departures from the average. A slightly denser region is an overdensity; a less dense one is an underdensity. These differences existed before recombination. The Dark Ages gave them time to develop, rather than creating them from nothing.^[4][5]

In the inflationary picture, primordial fluctuations originated from quantum processes during an extremely early period of accelerated expansion. This is a theoretical explanation consistent with important observations, not a directly witnessed event. CMB temperature and polarization patterns constrain those early conditions; they are not photographs of already formed galaxies.^[8]

How a small difference becomes a large one

Gravity responds to differences in the distribution of mass. Regions with more matter attract surrounding material, allowing their density contrast to grow. Eventually, sufficiently strong overdensities can stop expanding with the general cosmic flow and become gravitationally bound.

By the Dark Ages, matter dominated the cosmic energy budget over radiation. That transition had begun earlier; it was not caused by the first atoms forming. Under these conditions, density fluctuations could grow into dark matter concentrations and gas clouds.^[4]

Growth was not equally easy for every component. Ordinary gas had previously been strongly coupled to radiation, whose pressure resisted compression. Dark matter was not coupled in the same way, so its structure could develop earlier. After decoupling, gas could increasingly collect in those pre-existing gravitational concentrations.^[1][5]

4. Dark matter and the first structures

Halos: gravitational concentrations, not solid shells

A dark matter halo is a gravitationally bound concentration of dark matter. It is not a solid surface around a cloud. Gas can gather within its gravitational field, and the halo can continue growing by attracting surrounding matter and merging with other halos.^[4]

In the ΛCDM framework, dark matter is “cold”: its particles moved slowly compared with light when the relevant structures developed. The term concerns particle motion, not a low gas temperature measured by a thermometer. This behavior allows small-scale concentrations to survive and grow. The identity of the dark matter particles is not established by this description.^[1][8]

From small halos toward a cosmic web

In this framework, structure grows hierarchically: smaller systems assemble into larger ones through accretion and mergers. Sheets, filaments, and dense intersections emerge from the evolving matter distribution, forming the developing cosmic web. The early network was not yet the mature collection of galaxies and clusters seen today.^[5]

The first small halos and protogalactic systems provided environments where ordinary gas could accumulate. But collecting gas was only part of the problem.

Why gravity needed help from cooling

Compression heats gas, and the resulting pressure can resist further collapse. To reach stellar densities, a cloud must shed enough energy. In many primordial star-forming halos, molecular hydrogen supplied an important cooling route: collisions excited the molecules, which could then release energy as radiation.^[7]

This distinguishes a halo from a star-forming cloud. A halo can contain gas without immediately producing stars. Gravity brings material together; the gas’s ability to cool helps determine what happens next.^[5]

5. Cosmic dawn: emergence of the first stars

Population III stars

When dense gas could collapse far enough, it formed the first stars, known as Population III stars. Their material was almost entirely primordial hydrogen and helium, with traces of other light nuclei, rather than gas enriched by previous generations of stars.

Many models favor a greater proportion of massive stars than in present-day star formation. However, it is too restrictive to picture every first star as an isolated giant of the same enormous mass. Primordial clouds can fragment, producing small groups of stars with a range of masses.^[7]

The first stars and reionization were different milestones

The appearance of the first stars marks cosmic dawn, the transition out of the starless Dark Ages. Their energetic radiation began changing nearby gas. Photons energetic enough to remove electrons from hydrogen created ionized regions around luminous sources.

As star formation and early galaxies developed, these regions expanded and increasingly overlapped. This extended transformation is reionization. It continued long after the first stars formed, with most diffuse intergalactic hydrogen ionized by roughly a billion years after the Big Bang.^[2]

6. Observational challenges and probes

Why ordinary images are not enough

A starless gas distribution cannot be surveyed in the same way as a sky full of bright galaxies. The challenge is not simply that everything was hidden behind an opaque screen: much of the universe was already transparent, but the objects that normally serve as astronomical beacons had not formed.

The CMB supplies an earlier view, while distant galaxies reveal a later stage. To investigate the intervening gas more directly, astronomers need a signal from the gas itself.^[6]

The 21-centimeter line of hydrogen

Neutral hydrogen has two closely spaced energy states associated with the relative spins of its proton and electron. Transitions between them absorb or emit radiation with a rest wavelength of about 21 centimeters—a frequency of approximately 1,420 MHz. This is the hydrogen hyperfine transition.

Against the CMB, hydrogen can produce an absorption or emission signal, depending on the populations of those two states relative to the radiation background. A quantity called the spin temperature describes that balance. It is related to, but not automatically identical to, the gas’s ordinary temperature.^[6]

A possible three-dimensional view

Combining position on the sky with frequency could reveal how the hydrogen signal varied across space and time—a form of cosmic tomography. But a brightness map is not simply a density map: temperature, ionization, and gas motions also affect the signal and must be modeled.^[15]

Different experiments target different periods

LOFAR and the Murchison Widefield Array (MWA) use arrays of antennas to investigate spatial fluctuations in the redshifted signal, especially during reionization and related early epochs. A 2025 LOFAR analysis, for example, reported upper limits at redshifts around 8–10, not a completed image of the starless Dark Ages.^[9][10]

EDGES takes a different approach: it measures the radio spectrum averaged over a broad area of sky, searching for the changing hydrogen signature associated with cosmic dawn and reionization. A sky-averaged spectrum and a spatial map answer different questions.^[11]

SKA-Low observing plans include statistical measurements and imaging of hydrogen during cosmic dawn and reionization. These goals should not be confused with a claim that the much earlier starless universe has already been mapped.^[12]

Why the earlier Dark Ages are harder still

The cosmological signal is faint compared with radio foregrounds, including emission from our own galaxy. Instrument calibration and human-made interference complicate the measurement. For the highest redshifts, the signal reaches particularly low radio frequencies, where Earth’s ionosphere introduces further obstacles.^[10][13]

These difficulties motivate concepts for instruments beyond Earth’s ionosphere, including on the far side of the Moon, where shielding from terrestrial radio interference is an advantage. Such proposals address a specific observational problem; they are not evidence that a definitive Dark Ages map already exists.^[13]

Indirect evidence still matters

Researchers also test simulations against early galaxies, the CMB, and later signs of reionization. Infrared observations, including those made with Webb, investigate galaxies whose stars had already formed. They help constrain the preceding history, but do not directly photograph an earlier starless cloud population.^[2][8]

7. Implications for modern cosmology

Testing how structure formed

The Dark Ages connect the small early variations recorded in the CMB with the later universe of stars and galaxies. A successful model must explain how matter evolved between those stages—not just describe either endpoint.

That makes the timing and distribution of the earliest structures valuable tests of dark matter behavior and primordial fluctuations. Different assumptions can change how much small-scale structure develops, how rapidly gas collects, and when star formation becomes possible.^[4][8]

Separating observations from predictions

There are several kinds of knowledge in this story. The CMB is measured. The growth of early halos and primordial star-forming clouds is reconstructed through physical models and simulations. Detailed properties of the first stars remain less certain. Prospective Dark Ages radio maps would add another observational test rather than replace that distinction.^[7][8][15]

Likewise, an observational upper limit is useful without being a detection. It tells researchers how strong a signal could be while remaining compatible with the data, allowing some predictions to be constrained. Turning a faint measurement into a cosmic history still requires careful treatment of foregrounds, instruments, and the underlying physics.^[9]

One connected history

Recombination changed the interaction between light and matter. Gravity amplified existing density variations. Gas accumulated and cooled. The first stars introduced new radiation sources, and reionization changed the surrounding hydrogen. Each stage inherited conditions from the one before it.^[4][5]

The essential takeaway

The cosmic Dark Ages were dark because stars had not yet formed—not because nothing existed or nothing was happening. Neutral gas cooled, dark matter concentrations grew, and gravity gathered the material from which the first luminous objects would emerge.

The first stars were the visible result of that long preparation. Understanding the preparation helps explain both why they appeared and why the universe developed structure instead of remaining almost uniform.^[1][5]