The 21 cm line is a specific wavelength of radio light emitted by neutral hydrogen atoms, and it has become one of the most powerful tools in modern astronomy for studying the universe’s history. Because hydrogen is the most abundant element in the cosmos, this faint signal acts like a fossil record stretching back to the earliest epochs after the Big Bang. Researchers use it to study everything from the magnetic fields threading through our galaxy to the formation of the very first stars, and a new generation of radio telescopes is being built specifically to exploit it.
Where the Signal Comes From
Every hydrogen atom consists of a single proton orbited by a single electron. Both the proton and the electron have a quantum property called spin, and when those spins are aligned in the same direction the atom holds a tiny bit more energy than when they point in opposite directions. When the atom flips from the higher-energy (parallel) state to the lower-energy (antiparallel) state, it releases a photon with a wavelength of about 21.1 centimeters, corresponding to a frequency near 1420 MHz. The transition is extraordinarily rare for any individual atom, but the sheer quantity of hydrogen in the universe means that the collective signal is detectable.
This emission sits in the radio part of the electromagnetic spectrum, which is convenient for astronomers. Radio waves pass through dust clouds that block visible light, and Earth’s atmosphere is largely transparent to them. The 21 cm line also has a precise, well-known rest frequency, so when astronomers observe it at a different frequency they can immediately tell how much the light has been stretched by the expansion of the universe. That redshift tells them both the distance and the cosmic epoch they are looking at. A signal received at, say, 142 MHz instead of 1420 MHz has been redshifted by a factor of ten, meaning it originated when the universe was roughly a tenth of its current size.
A Window Into the Dark Ages
The most ambitious use of the 21 cm line is as a probe of the so-called cosmic Dark Ages, the period roughly 10 to 200 million years after the Big Bang when no stars or galaxies had yet formed. During this era the universe was filled with a nearly uniform fog of neutral hydrogen and little else. Because no luminous objects existed, the only background radiation was the fading glow of the cosmic microwave background. The hydrogen gas was cooling faster than that background radiation, so it should have absorbed 21 cm photons from the microwave background rather than emitting them, leaving a distinctive absorption signature.
Modeling this absorption feature is a major area of active research. In the standard cosmological model, the first broad absorption trough is predicted to appear at a frequency around 16 MHz, corresponding to a redshift near 87, with a depth of roughly 39 millikelvins in brightness temperature and a width of about 23 MHz.1Monthly Notices of the Royal Astronomical Society. Formation of the hydrogen line 21-cm in Dark Ages and Cosmic Dawn: dependences on cosmology and first light The exact shape and depth of this trough depend on the thermal history of the intergalactic gas, which in turn depends on the properties of dark matter. If dark matter particles can annihilate or decay, even slowly, they would inject extra energy into the gas and change the absorption profile. Researchers have shown that the Dark Ages portion of the 21 cm signal is particularly sensitive to the parameters of self-annihilating and decaying dark matter, making it a potential diagnostic tool for particle physics that no laboratory experiment can replicate.2Physical Review D. Global signal in the redshifted hydrogen 21-cm line from the dark ages and cosmic dawn: Dependence on the nature of dark matter and modeling of first light
Cosmic Dawn and the First Stars
The Dark Ages ended when the first stars ignited, an era known as Cosmic Dawn. These early stars were massive, hot, and short-lived, and they pumped ultraviolet radiation into the surrounding hydrogen gas. A particular wavelength of that ultraviolet light, called Lyman-alpha, plays a critical role in making the 21 cm signal visible. When Lyman-alpha photons scatter repeatedly through the gas, they effectively force the spin temperature of the hydrogen to match the gas’s own kinetic temperature rather than the temperature of the cosmic microwave background. This process is known as the Wouthuysen-Field effect, and simulations confirm that it works even in gas with modest optical thickness.3The Astrophysical Journal Supplement Series. Lyα Radiative Transfer: Monte Carlo Simulation of the Wouthuysen–Field Effect
Before this coupling saturated everywhere, the first galaxies produced strong spatial variations in the 21 cm signal. Regions near early galaxies received more Lyman-alpha photons, so their hydrogen was coupled to the gas temperature earlier, while distant voids remained coupled to the microwave background. These fluctuations were driven by the uneven distribution of the earliest galaxies.4The Astrophysical Journal / IOPscience. Detecting the Earliest Galaxies through Two New Sources of 21 Centimeter Fluctuations For the dominant population of faint galaxies at redshifts as early as around 25, the Lyman-alpha-driven 21 cm absorption signal may be the only feasible method of detection we have.5Physics Reports. The rise of the first stars: Supersonic streaming, radiative feedback, and 21-cm cosmology No optical or infrared telescope, not even the James Webb Space Telescope, can see these earliest, faintest objects. Only their imprint on the surrounding hydrogen is potentially observable.
X-Ray Heating and the Epoch of Reionization
After the first stars formed, the universe entered a more complex phase. Remnants of massive stars formed binary systems where gas falling onto a compact object produced X-rays. These high-mass X-ray binaries gradually heated the intergalactic gas above the temperature of the microwave background, flipping the 21 cm signal from absorption to emission. The timing and strength of this heating epoch depend on the X-ray output of these early binaries, which in turn depends on the metal content of the host galaxies. Low-metallicity galaxies are more X-ray luminous per unit of star formation, and differences in the assumed relationship between metallicity and X-ray luminosity can shift the predicted mean temperature of the intergalactic gas and the large-scale 21 cm power by a factor of about three at a given cosmic epoch.6Monthly Notices of the Royal Astronomical Society. The 21-cm signal from the cosmic dawn: metallicity dependence of high-mass X-ray binaries
Adding to this complexity, the interstellar medium within the host galaxies absorbs some of those X-rays before they escape. Simulations indicate that the gas in early galaxies was relatively metal-poor and therefore more transparent to soft X-rays than local galaxies are, allowing photons with energies as low as 0.5 keV to reach the intergalactic medium. The resulting soft-band X-ray luminosity escaping these galaxies was roughly a factor of three larger than estimates that assumed more opaque conditions.7Monthly Notices of the Royal Astronomical Society. High-mass X-ray binaries and the cosmic 21-cm signal: impact of host galaxy absorption Getting these details right matters because they determine the shape of the 21 cm signal that telescopes will measure.
Eventually, the cumulative ultraviolet radiation from stars and galaxies began ionizing the hydrogen itself, carving out expanding bubbles of ionized gas in the neutral intergalactic medium. This Epoch of Reionization lasted roughly from a redshift of about 12 down to about 6, and it erased the 21 cm signal region by region as the neutral hydrogen disappeared. Tracking how the average neutral fraction evolved over time is one of the central goals of 21 cm cosmology. Researchers have developed fitting methods that can recover this redshift evolution from the 21 cm power spectrum to within an absolute deviation of roughly 0.1, even across different reionization scenarios.8EDP Sciences (Astronomy & Astrophysics). Probing the intergalactic medium during the Epoch of Reionization using 21 cm signal power spectra
The EDGES Controversy
In 2018, the EDGES experiment reported a detection of the global 21 cm absorption signal from Cosmic Dawn centered around a redshift of 17. The signal was roughly the right shape but about twice as deep as standard models predicted, a discrepancy estimated at 3.8 standard deviations from the expected value in the standard cosmological model.9The Astrophysical Journal. Constraining the Dark Matter Vacuum Energy Interaction Using the EDGES 21 cm Absorption Signal This triggered a flurry of theoretical papers proposing explanations.
Two broad categories of explanation emerged. One posited that the gas was colder than expected, perhaps cooled by interactions with dark matter particles. The other proposed that the radio background against which the absorption was measured was brighter than the cosmic microwave background alone, perhaps augmented by photons from dark matter decay. Detailed analysis has found the cooling-based explanations severely constrained by other observations. Among photon-producing dark matter candidates, most perturbative scenarios involving axion-like particles, massive spin-2 dark matter, or excited dark matter states are ruled out by existing experimental bounds. The remaining viable explanations, according to one thorough analysis, involve mini-clusters of axion-like or oscillating spin-2 dark matter where nonlinear instabilities enhance photon production.10Physics Letters B. The EDGES 21 cm anomaly and properties of dark matter
The EDGES result remains contentious. The SARAS 3 experiment, using a different instrument design, failed to confirm the signal, and many researchers suspect the original detection was contaminated by systematic errors in foreground subtraction. Whether or not the specific EDGES measurement holds up, the episode demonstrated how sensitive the 21 cm absorption profile is to new physics. Even a null result constrains the properties of dark matter and the thermal history of the early universe in ways that complement particle collider experiments.
Why the Signal Is So Hard to Detect
The 21 cm cosmological signal is extraordinarily faint. Astrophysical foregrounds, primarily synchrotron radiation from our own galaxy, are thousands of times brighter at the relevant frequencies. Fortunately, these foregrounds are spectrally smooth, meaning their brightness changes gradually with frequency, while the cosmological signal has sharp spectral structure. This difference allows researchers to separate the two, at least in principle. Observations have confirmed that foreground contamination can be isolated to a predictable “wedge-shaped” region in the data, leaving a cleaner window for cosmological analysis at other spatial scales.11The Astrophysical Journal Letters. Opening the 21 cm Epoch of Reionization Window: Measurements of Foreground Isolation with PAPER
Not all foregrounds behave so neatly, though. Polarized synchrotron emission from our galaxy undergoes Faraday rotation as it passes through the magnetized interstellar medium, introducing frequency-dependent structure that mimics the cosmological signal. When this polarized emission leaks into the total intensity measurement due to instrumental imperfections, it becomes a particularly dangerous contaminant because standard smooth-spectrum removal techniques cannot handle it.12Monthly Notices of the Royal Astronomical Society. Simulations of Galactic polarized synchrotron emission for Epoch of Reionization observations
Earth’s ionosphere adds another layer of difficulty. The ionosphere refracts and distorts incoming radio waves, and its effects worsen at lower frequencies, exactly the range needed for high-redshift 21 cm observations. Radio frequency interference from human technology is a persistent problem as well. Deep observations with the Murchison Widefield Array have required meticulous excision of ultrafaint interference and harsh quality cuts based on ionospheric conditions, removing a substantial fraction of observing time. Even after all that filtering, the best current upper limits on the 21 cm power spectrum remain several orders of magnitude above the expected signal level.13The Astrophysical Journal. Evidence of Ultrafaint Radio Frequency Interference in Deep 21 cm Epoch of Reionization Power Spectra with the Murchison Wide-field Array
The Telescopes Racing to Make a Detection
Several purpose-built radio arrays are either operating or under construction to target the cosmological 21 cm signal. The Murchison Widefield Array in Western Australia, the LOw Frequency ARray in the Netherlands, and the Precision Array for Probing the Epoch of Reionization in South Africa have been collecting data for over a decade, steadily pushing upper limits downward and refining calibration and foreground removal techniques.14IOPscience. Status of 21 cm Interferometric Experiments The Hydrogen Epoch of Reionization Array, also in South Africa, was designed from the ground up to maximize sensitivity to the 21 cm power spectrum during reionization, using a compact, highly redundant layout of 350 dishes.
The instrument expected to transform the field is the Square Kilometre Array, an international mega-project being built in Australia and South Africa. Its low-frequency component, SKA-Low, will consist of over 130,000 individual antennas spread across a remote Australian desert, providing sensitivity orders of magnitude beyond current arrays. Modeling studies have shown that instruments at this scale could detect the peaks of the cosmological 21 cm signal across multiple redshift ranges, from reionization back into Cosmic Dawn.15Monthly Notices of the Royal Astronomical Society. Reionization and beyond: detecting the peaks of the cosmological 21 cm signal
For the Dark Ages signal at frequencies below about 30 MHz, even the most remote terrestrial site is inadequate. Earth’s ionosphere becomes opaque at those frequencies, and human-generated radio interference is pervasive. This has motivated concepts for radio telescopes on the far side of the Moon, which is permanently shielded from Earth’s radio noise and has no ionosphere. FarView is one such concept: a large low-frequency array that would be manufactured directly on the lunar surface from metals extracted from the regolith. Its science case emphasizes the ability to produce both power spectra and tomographic maps tracing the universe from before the birth of the first stars through the beginning of Cosmic Dawn, potentially constraining dark matter properties, early dark energy, neutrino masses, and inflationary physics.16Advances in Space Research. FarView: An in-situ manufactured lunar far side radio array concept for 21-cm Dark Ages cosmology
Cosmological Parameters From 21 cm Data
Beyond reconstructing the history of reionization, the 21 cm signal carries information about fundamental cosmological parameters. Because 21 cm observations probe a three-dimensional volume of the universe (two angular dimensions on the sky plus frequency as a proxy for distance), they sample an enormous number of independent modes. This gives them statistical power that complements and in some cases surpasses galaxy surveys.
Neutrino masses are one target. Massive neutrinos subtly suppress the growth of structure on small scales, and because 21 cm experiments access smaller scales and larger volumes than galaxy surveys, they can provide tighter constraints. Fisher matrix forecasts have shown that 21 cm experiments could outperform even very large galaxy surveys in constraining the sum of neutrino masses.17Physical Review D. Constraining massive neutrinos using cosmological 21 cm observations Primordial non-Gaussianity, a signature of inflationary physics imprinted on the large-scale distribution of matter, is another target. Forecasts indicate that instruments like the SKA could measure the non-Gaussianity parameter to a precision of order 10, with purpose-built fast Fourier transform telescopes potentially improving that by an order of magnitude.18PubMed. Primordial non-Gaussianity from the 21 cm power spectrum during the epoch of reionization
Dark energy is accessible through the 21 cm line as well. The baryon acoustic oscillation pattern, a preferred scale imprinted on matter by sound waves in the early universe, acts as a standard ruler. Mapping that pattern in neutral hydrogen emission at redshifts of order one using a technique called intensity mapping can constrain the expansion history of the universe. A ground-based 21 cm intensity mapping survey offers a relatively low-cost way to measure baryon acoustic oscillations at these intermediate redshifts.19The Astrophysical Journal. A GROUND-BASED 21 cm BARYON ACOUSTIC OSCILLATION SURVEY Combining such measurements with cosmic microwave background data from Planck can yield improved constraints on the Hubble constant and dark energy equation-of-state parameters.20Modern Physics Letters A. Cosmological Forecasts from the Baryon Acoustic Oscillations in 21cm Intensity Mapping
Cross-Correlation and the 21 cm Forest
One way to build confidence in a faint detection is to cross-correlate the 21 cm signal with an independent tracer of the same cosmic structures. Lyman-alpha emitting galaxies are one such tracer: these are galaxies whose ultraviolet emission line is sensitive to the presence of surrounding neutral hydrogen, so their observed properties correlate with the reionization state of their environment. Simulations find that the cross-correlation between Lyman-alpha emitter maps and the 21 cm signal is a robust probe of reionization, relatively insensitive to assumptions about how galaxies populate their host halos. A 1,000-hour observation with LOFAR could distinguish a fully ionized universe from one that is still half neutral, while the SKA could do so with far less observing time even under pessimistic assumptions.21Monthly Notices of the Royal Astronomical Society. Cross-correlation of the cosmic 21-cm signal and Lyman Alpha Emitters during reionization
A conceptually different approach is the 21 cm forest. Just as the Lyman-alpha forest reveals intervening gas clouds along the line of sight to a distant quasar, the 21 cm forest would reveal neutral hydrogen structures along the line of sight to a bright background radio source at high redshift. Semianalytic modeling has shown that features from both small collapsed objects called minihalos and larger protogalactic disks should be observable by next-generation low-frequency arrays, provided sufficiently bright background radio sources exist at those redshifts.22The Astrophysical Journal. The 21 Centimeter Forest: Radio Absorption Spectra as Probes of Minihalos before Reionization Whether those background sources exist in sufficient numbers remains an open question, but even a handful of detections would provide uniquely detailed information about the small-scale structure of the intergalactic medium before reionization completed.
Closer to Home: Galaxy Rotation and Magnetic Fields
The 21 cm line is not only a cosmological probe. It has been a workhorse of galactic astronomy for decades. Because neutral hydrogen extends far beyond the visible disk of most spiral galaxies, 21 cm observations reveal rotation speeds at large distances from the galactic center where optical tracers have faded below detectability. These extended rotation curves were among the earliest and most compelling evidence for dark matter in galaxies. The baryonic Tully-Fisher relation, which connects a galaxy’s total mass in stars and gas to its rotation speed, has been calibrated using 21 cm rotation curves spanning velocities from about 20 to 300 kilometers per second.23The Astrophysical Journal. The Baryonic Tully-Fisher Relation of Galaxies with Extended Rotation Curves and the Stellar Mass of Rotating Galaxies This relation is a fundamental scaling law that constrains models of galaxy formation and dark matter halos.
The 21 cm line also provides the only direct measurement of magnetic field strength in neutral interstellar gas, through the Zeeman effect. When hydrogen atoms sit in a magnetic field, the 21 cm line splits into closely spaced components whose separation is proportional to the field strength along the line of sight. This splitting gives both the magnitude and the direction of the magnetic field on the scale of interstellar gas complexes.24Nature. Interstellar Magnetic Fields determined from Zeeman Effect Measurements Typical fields measured this way are a few microgauss, tiny by everyday standards but significant for the dynamics of interstellar gas, cosmic ray propagation, and star formation. No other observational technique gives this information for the cold, diffuse neutral medium.