Hint of an unexpectedly cool early universe suggests interacting dark matter

Just before the first stars turned on, three-quarters of the conventional matter of the universe consisted of neutral hydrogen atoms. As in the case in the modern universe, the spin of a hydrogen atom’s electron was usually antiparallel to the spin of its proton. But some of the hydrogen was in a higher-energy state with parallel spins; the energy difference between the two states corresponds to a photon with a rest-frame wavelength of 21 cm. Because the gas of hydrogen atoms had adiabatically expanded since shortly after the Big Bang, its bulk kinetic temperature was less than that of the ambient cosmic radiation. The two-state spin system, however, was in thermal equilibrium with the cosmic radiation, so in the time just before star formation, neither net 21 cm emission nor absorption occurred.

Once stars began to shine, their UV radiation provided a pathway by which the spin system could couple to the lower-temperature hydrogen gas, cool down, and absorb 21 cm cosmic radiation. And that it did, until stellar output was so intense that interstellar hydrogen was heated and eventually ionized.

Researchers with the Experiment to Detect the Global Epoch of Reionization Signature (EDGES) have now combed through two years of observations and extracted the 21 cm signal from their sky-averaged radio spectrum. The absorption trough detected by EDGES’s two RF antennas (one of which is shown in the photo) occurred over a range of redshifts spanning the period from stellar ignition, which the EDGES researchers pin to 180 million years after the Big Bang, to cosmic ionization at least 90 million years later. And that trough is more than twice as intense as conventional models predict. Apparently, the early-universe hydrogen was cooler than expected or the background radiation was hotter than expected.

In a separate work, Rennan Barkana of Tel Aviv University argues that dark matter, the still mysterious stuff that makes up five-sixths of the universe’s material content, could interact feebly with hydrogen atoms via some nongravitational force and cool the gas. If he is right, his analysis of the EDGES observation suggests that a dark-matter particle is no heavier than several proton masses, much lighter than in most models. Less than a week after the publication of the EDGES and Barkana papers, however, Fermilab’s Dan Hooper and colleagues posted an article on the arXiv preprint server severely challenging, if not quite ruling out, the compatibility of the EDGES result and dark-matter cooling.

Several experiments currently under way should ultimately test the EDGES claim. Meanwhile, the EDGES researchers are working to devise more sophisticated models for properly subtracting out the unwanted signals, including FM radio, that mask the 21 cm absorption. (J. D. Bowman et al., Nature 555, 67, 2018 ; R. Barkana, Nature 555, 71, 2018 ; A. Berlin et al., https://arxiv.org/abs/1803.02804 .