The setting is brane–antibrane inflation, one of the most concrete realizations of inflation within string theory. In the influential KKLMMT construction, our universe is a compact, curled-up space threaded with warped throats, deep funnels of strongly curved geometry. A mobile D3-brane slides down one such throat toward an antibrane parked at its tip, and the separation between them acts as the inflaton field driving the accelerated expansion. Unlike ordinary single-field models, where inflation ends with the decay of a particle-like field, this scenario terminates in an intrinsically stringy event: the open string stretched between brane and antibrane becomes tachyonic, the pair undergoes tachyon condensation, and they annihilate in a burst of energy.

That raises a question the authors take seriously: should the aftermath of this annihilation be described immediately as a bath of ordinary radiation, or does the universe first pass through a high-temperature string phase? Their answer is that a Hagedorn phase is not only possible but, under reasonable assumptions, natural. The key lies in the exponential growth of the string spectrum. The number of possible string states grows as the exponential of the energy, defining a maximum temperature, the Hagedorn temperature, set by the local string scale. Near this temperature, additional energy is stored not in hotter particles but in ever longer, more highly excited strings, so the temperature effectively pins itself just below the Hagedorn limit.

The study builds on recent progress in perturbatively stabilized brane–antibrane inflation, a framework in which the overall volume of the compact space is fixed by controlled perturbative corrections to the theory rather than by non-perturbative effects. This matters because conventional stabilization schemes generically spoil slow-roll inflation through the notorious eta-problem, an inflaton mass of order the Hubble scale. The perturbative construction sidesteps this difficulty, providing a controlled arena in which the end-of-inflation energy budget can be computed reliably. The authors show that the energy released at annihilation is of order the warped brane tension, which exceeds the local string scale in the throat, precisely the condition needed to push a gas of open strings into the Hagedorn regime.

The consequences depend on where the Standard Model lives. In the simplest case, the visible sector branes sit in the same warped throat where the annihilation occurs. The authors find the entry condition is remarkably mild: for weak string coupling, only a few percent of the annihilation energy, roughly 0.4 percent for a single visible brane stack and about 10 percent for a stack of five, must be deposited into surviving visible open strings for the Hagedorn threshold to be crossed. The resulting stringy epoch is brief, lasting roughly one to a few e-folds of expansion, but its effects can be profound and potentially observable.

The most striking consequence concerns dark radiation. String compactifications generically contain light hidden-sector fields, axions, hidden gauge sectors, and other weakly coupled species, which, if populated during reheating, contribute to the effective number of relativistic species, a quantity cosmologists denote Delta N_eff and which is tightly constrained by observations of the cosmic microwave background. The Hagedorn phase offers a natural suppression mechanism. Because the entropy of a gas near the Hagedorn temperature is enormous, the visible sector acquires a huge entropy reservoir, diluting the relative contribution of any decoupled dark radiation. The authors quantify this: if the visible sector just barely enters the Hagedorn regime, the allowed dark-radiation branching fraction is limited to the one-percent level or below, comfortably satisfying current bounds.

The analysis becomes richer when the Standard Model resides in a different throat from the annihilation. In that case, energy must travel between throats, carried by massive closed strings and throat-localized Kaluza–Klein modes that can tunnel through the bulk compactification. The authors parameterize this transfer by an effective rate and distinguish two regimes. In prompt transfer, the energy reaches the visible throat before cosmological expansion dilutes it, and Hagedorn reheating proceeds efficiently provided the visible throat is at least as strongly warped as the annihilation throat, meaning its local string scale is lower or comparable. Intriguingly, the more strongly warped the visible throat, the closer the reheating temperature approaches the Hagedorn limit, and the longer the stringy phase lasts, up to several e-folds in favorable cases.

In the delayed regime, the annihilation-throat energy lingers until the Hubble rate drops to the transfer rate, by which point the universe has expanded and the energy density has fallen. Hagedorn reheating then requires the transfer rate to exceed a minimum threshold before dilution becomes fatal, while remaining below the Hubble scale at the end of inflation. For benchmark parameters, this delayed window exists when the visible throat’s string scale is at most about 1.7 times that of the annihilation throat, and it widens rapidly as the visible throat becomes more strongly warped. Across both regimes, the conclusion is the same: a visible Hagedorn phase is most readily realized when the Standard Model’s throat is the deeper one.

The work connects to a broader wave of interest in stringy early-universe physics. Recent studies have explored gravitational waves produced during and after Hagedorn phases, as well as cosmic superstrings, the fundamental strings stretched to astronomical scales that brane–antibrane annihilation naturally produces. Together, these threads suggest that the aftermath of brane inflation could leave fingerprints in both the relic radiation content of the universe and the stochastic gravitational-wave background that observatories such as LISA may one day detect.

None of this is yet observation, and the authors are careful to frame their results as a controlled theoretical exploration, with inter-throat transfer rates left as geometry-dependent parameters rather than computed from first principles. But the message is provocative: the end of inflation in string theory need not look like a simple particle bath at all. Instead, the universe’s first breath after inflation may have been a brief, searing string phase, a fire at the tip of a warped throat, hot enough that temperature itself lost its meaning, and structured in just the right way to hide dark radiation from the eyes of modern cosmology.

Subject of Research: The post-inflationary open-string Hagedorn phase in perturbatively stabilized brane–antibrane inflation and its consequences for dark radiation

Article Title: Fire at the tip of the throat: Hagedorn phase after brane–antibrane inflation?

Article References: Fire at the tip of the throat: Hagedorn phase after brane–antibrane inflation?. (n.d.). https://doi.org/10.1140/epjc/s10052-026-16470-w

Image Credits: AI Generated

DOI: 10.1140/epjc/s10052-026-16470-w

Keywords: string theory, brane–antibrane inflation, Hagedorn phase, cosmic inflation, warped throats, dark radiation, Delta N_eff, tachyon condensation, reheating, D-branes, cosmology, gravitational waves