Reionization: Ending the Dark Ages

Reionization: Ending the Dark Ages

Knowledge Ark · The early universe

The Dark Ages & First Structures

Before the first stars, gravity was already reshaping the universe. Across a cooling sea of hydrogen and helium, small differences in density grew into the places where starlight would begin.

Neutral gasGrowing halosBefore starlight
Primordial matter taking shape · an artistic interpretation
≈380,000 yearsThe era begins after recombination and the main release of the CMB from frequent scattering.
Before starlightThe dark ages end as the first stars appear and cosmic dawn begins.
21 centimetersA hydrogen radio wavelength that could help reveal the starless universe.

The first stars did not appear everywhere at once. These stages describe a gradual, uneven cosmic transformation. [1] [3]

The quiet before cosmic dawn

A universe in motion, before stars.

A sky without stars might seem like a universe in which nothing was happening. In reality, the cosmic dark ages were a time of slow but consequential change: gas cooled, matter gathered, and simple chemistry helped prepare the first stellar nurseries.

This interval followed recombination, when ordinary matter became mostly neutral and the background radiation could travel much farther. It lasted until the emergence of the first stars—a transition called cosmic dawn.[1]

01
A sky before starlight

What made the dark ages dark?

The name refers to the absence of stars and other bright astronomical sources. It does not mean that all light disappeared. The thermal radiation left from the hot early universe was still everywhere.

At the beginning of this interval, that background was still around 3,000 K, with infrared and visible radiation. As space expanded, its wavelengths stretched and its glow cooled. Much later, we observe it as the cosmic microwave background. The dark ages therefore did not begin with a universe already chilled to only tens of kelvins.[11]

The ordinary gas was simple: roughly three-quarters hydrogen and one-quarter helium by mass, with traces of other light elements. These proportions describe ordinary matter; dark matter was a separate component. There were no rocky planets or mineral dust grains from earlier generations of stars.

02
Expansion changes the temperature

How the gas and radiation cooled

After recombination, it becomes useful to distinguish the temperature of the background radiation from that of the gas. Their histories were connected, but eventually diverged.

Expansion stretched the background photons to longer wavelengths, lowering the radiation temperature. At first, the gas stayed close to that temperature because a small population of leftover free electrons continued transferring energy from the radiation to the gas through Compton scattering.

As the universe became more dilute, this energy exchange weakened. The diffuse gas then cooled faster than the radiation. In the usual cosmological picture, this gradual separation took place several million years after the Big Bang.[2]

A closer look: two different cooling rates

The scale factor, written a, tracks cosmic expansion. The background radiation temperature falls approximately as 1/a. After thermal coupling becomes weak, freely expanding, unheated gas approaches a temperature proportional to 1/a2.

If the scale factor doubles, the radiation temperature halves, while the temperature of that idealized gas falls to one-quarter. Collapsing clouds behave differently: compression and shocks can heat them.[2]

The leftover electrons mattered

Recombination was not perfectly complete. Expansion spread the remaining electrons and ions farther apart until further recombination became slow. A residual fraction of order 10−4—roughly a few free electrons per ten thousand hydrogen nuclei during the later dark ages—remained. The precise fraction evolved with time.

These electrons were primarily leftovers from incomplete recombination. Besides exchanging heat with the radiation, they helped drive reactions that produced the first molecules.[4]

03
The beginnings of cosmic architecture

How small differences became structures

The universe was never perfectly smooth. Small density differences existed before recombination, and dark matter had already developed gravitational concentrations. The dark ages continued this earlier growth.

Regions containing slightly more matter exerted a stronger gravitational pull. On scales where gravity overcame the relevant opposing effects, their density contrast grew. Ordinary gas could increasingly gather in the gravitational wells of dark matter after its strong coupling to radiation weakened.[2]

Why dark matter helped lead the way

In the standard model, cold dark matter does not interact appreciably with light. It therefore avoided the radiation pressure that strongly affected the early gas. “Cold” describes its relatively slow particle motion during structure formation; its microscopic identity remains unknown.

Gravity assembled dark matter into bound concentrations called halos. Small halos grew by collecting material and merging with others. Gas accumulated within some of them, creating potential birthplaces for stars.[5] [6]

Cosmic expansion did not prevent all local collapse. A sufficiently overdense region could stop expanding with its surroundings and become gravitationally bound. The universe could grow more spacious overall while selected regions became denser.

Explore this foundation further in Gravitational Clumping and Density Fluctuations and Early Mini-Halos and Protogalaxies.

Why gas and dark matter followed different paths
Dark matter Ordinary gas
Gathers under gravity into extended halos. Falls into gravitational wells but also resists compression through gas pressure.
Does not radiate energy away like ordinary gas in the standard cold-dark-matter picture. Can lose energy through radiation and contract into much denser clouds.
Provides much of the gravitational mass of the host halo. Supplies the material from which conventional stars form.
Gravity gathers diffuse gas into a dark matter halo A schematic compares widespread neutral gas with gas gathering inside a dark matter halo. The second panel shows a denser gas core before stars form. The drawing has no physical scale. Small density differences Gravity gathers matter Gas is spread through space Gas collects inside a dark matter halo Time Denser gas core Dark matter halo Neutral gas Dark matter Schematic · no physical scale
Gravity gathers diffuse gas into a dark matter halo A schematic compares widespread neutral gas in the upper panel with gas gathering inside a dark matter halo below. The lower panel shows a denser gas core before stars form. The drawing has no physical scale. Small density differences Gas is spread through space Time Gravity gathers matter Gas gathers inside a dark matter halo Denser gas core Dark matter halo Neutral gas Dark matter Schematic · no physical scale
A schematic of growth over time: diffuse matter gathers into a halo, and gas can form a denser core. Colors identify components; the illustration has no physical scale.
04
Small molecules, enormous consequences

Why the first clouds needed molecules

Gathering gas into a halo was only part of the task. As gas fell inward, compression and shocks heated it. The resulting pressure could resist further contraction. To reach much higher densities, the gas needed a way to shed energy.

Today, interstellar clouds can cool through radiation from a rich mixture of atoms, molecules, and dust. The first clouds had very little chemical variety. At the temperatures reached in small primordial halos, atomic hydrogen alone was an inefficient coolant.

A small amount of molecular hydrogen, H2, made a major difference. Collisions transferred energy into molecular rotation and vibration. Radiation from those excited molecules could then carry energy out of the cloud, allowing contraction to continue.[5]

Models commonly place the earliest stellar nurseries in minihalos with total masses around 100,000 to one million Suns, mostly dark matter. This is a characteristic range, not a fixed threshold: gas motion, chemistry, and the surrounding environment affect which halos can form stars.[10]

How a leftover electron could help make a molecule

One important route used a negatively charged hydrogen ion as an intermediate:

H + e− → H− + photon

H− + H → H2 + e−

An electron helps two hydrogen atoms join, then is released and can participate again.

Other reactions also contributed, and their importance changed with density and temperature. This route shows how a tiny remaining charged population could influence the future of mostly neutral gas.[4]

05
The arrival of stellar light

How the first stars changed everything

As some clouds became dense enough, they developed protostars that gathered more gas. The first generation formed from material essentially untouched by earlier stars and is known as Population III.

Models place an early era of star formation roughly 100–200 million years after the Big Bang. This is a useful approximate timescale, not a directly measured date for the universe's very first star.[10]

Their masses and birth environments remain active research questions. Simulations allow fragmentation into multiple stars, and their growth can be limited by radiation and other feedback. A single enormous star in every halo is too simple a picture.[6]

The transition was gradual and uneven

The appearance of starlight marked cosmic dawn. Nearby gas began responding to the new radiation, while other regions remained starless. Some radiation broke molecules apart; ionizing ultraviolet light stripped electrons from hydrogen. The consequences could encourage or suppress further star formation depending on local conditions.

Reionization was the extended transformation that followed as ionized regions grew and connected. The birth of the first stars ended the starless dark ages, but it took much longer to ionize most hydrogen between galaxies.[1] [10]

These later chapters continue in Population III Stars: The Universe's First Generation and Reionization: Ending the Dark Ages.

06
Listening to ancient hydrogen

How to study a starless universe

Ordinary astronomical images rely on luminous objects or illuminated material. The dark ages offer neither a sky full of galaxies nor a population of stars to trace. Studying this interval requires signals from the gas itself, together with evidence from the periods before and after it.

Hydrogen's 21-centimeter signal

Neutral hydrogen can emit or absorb radio radiation through a tiny energy change associated with the relative spin states of its proton and electron. In the atom's rest frame, this transition corresponds to a wavelength of about 21 centimeters, or a frequency of about 1,420 MHz.

Expansion stretches radiation from the early universe to longer wavelengths and lower frequencies. Measuring its strength at different frequencies could therefore sample different stages of cosmic history.

The signal is a contrast against background radio radiation, particularly the CMB. Depending on the hydrogen's excitation state, it can appear in absorption or emission. Its strength depends on gas density, temperature, neutral fraction, and the processes setting that excitation state. Reading it requires a physical model.[3]

Why the measurement is so difficult

The ancient hydrogen signal is extremely faint compared with radio emission from our own galaxy and other foreground sources. Human transmissions, instrument effects, and Earth's ionosphere further complicate the measurement.[7]

Observatories such as SKA-Low are designed to investigate cosmic dawn and reionization through redshifted hydrogen. The earlier, starless dark ages require lower frequencies and different observing strategies; a telescope optimized for later epochs does not automatically cover that entire interval.[8]

Proposed radio arrays on the Moon's far side aim to reach some of those difficult frequencies while benefiting from shielding against terrestrial radio interference and the absence of Earth's ionosphere. Such instruments could provide information about the gas long before it produced stars.[9]

Reconstructing the missing interval

The CMB constrains early conditions. Observations of young galaxies reveal later outcomes. Researchers use gravity, gas dynamics, chemistry, and simulations to test how one could develop into the other. A simulation visualizes a calculation; matching several independent observations is what gives that calculation weight.

07
From simplicity to a structured universe

What this era can teach us

The dark ages connect a nearly smooth early universe with a cosmos containing stars and galaxies. They offer a way to study the emergence of structure before generations of stars substantially altered the gas.

How many small halos formed, how much gas they retained, and when they first produced stars depend on both cosmology and astrophysics. Different dark matter properties could alter small-scale structure, while gas chemistry and heating could change whether those structures become luminous. Comparing these effects is part of the challenge.[6] [10]

The broad sequence is well grounded: recombination, a mostly neutral starless interval, and the emergence of luminous sources. The detailed timing, masses, and environments of the first stars are less certain. Keeping that distinction clear makes the story more useful—and leaves room for observations to improve it.

The dark ages were full of change. Expansion cooled the diffuse universe while gravity gathered matter into selected regions. A few surviving electrons helped build molecules; molecules helped gas lose heat; dense clouds prepared the way for stars.

Those stars would begin the later production and dispersal of elements needed for planets, minerals, and life. The next chapter follows how their radiation changed the surrounding universe: Reionization: Ending the Dark Ages.

Follow the evidence

Sources and further reading

  1. NASA Webb — Early UniverseThe dark ages, first stars, and extended era of reionization.
  2. Naoz and Barkana — Growth of Linear Perturbations before the Era of the First Galaxies (2005)Gas temperature, residual coupling to radiation, and early density growth.
  3. Furlanetto, Oh, and Briggs — Cosmology at Low Frequencies (2006)The physics of the hydrogen 21-centimeter signal and its observational challenges.
  4. Galli and Palla — The Chemistry of the Early Universe (1998)Residual ionization, primordial molecules, and cooling reactions.
  5. Abel, Bryan, and Norman — The Formation of the First Star in the Universe (2002)A foundational simulation of gas collapse and molecular cooling in a primordial halo.
  6. Klessen — Formation of the First Stars (2018)Primordial stellar nurseries, fragmentation, feedback, and uncertainties.
  7. SKAO — Probing the Cosmic DawnAn accessible introduction to radio observations of early hydrogen.
  8. Overview of 21 cm Experiments at High Redshift with SKAO (2026)The observational scope and planned approaches of SKA-Low experiments.
  9. Burns and colleagues — A Lunar Farside Low Radio Frequency Array for Dark Ages 21-cm Cosmology (2021)A proposed method for investigating the earlier starless universe.
  10. Feathers, Kulkarni, and Visbal — From Dark Matter Minihalos to Large-Scale Radiative Feedback (2025)Modeling early star formation across halo and intergalactic scales.
  11. Kurki-Suonio — Physics of the Cosmic Microwave Background and the Planck Mission (2010)The radiation background, last scattering, and cooling through expansion.
Explore this chapter: The Grand Beginning

Chapter introduction: Why Study the Early Universe?

  1. The Singularity and Moment of Creation
  2. Quantum Fluctuations and Inflation
  3. Cooling and the Formation of Fundamental Particles
  4. Matter vs. Antimatter
  5. Big Bang Nucleosynthesis
  6. Dark Matter: Unveiling the Universe's Hidden Mass
  7. Recombination and the First Atoms
  8. The Cosmic Microwave Background
  9. The Dark Ages and First Structures — you are here
  10. Reionization: Ending the Dark Ages
  11. Dark Energy: The Enigma Driving Cosmic Acceleration
Back to blog