The phase diagram of the D1-D5 CFT and localized black holes

Ofer Aharony, Ronny Frumkin, and Jonathan Mehl · 2026

Read The phase diagram of the D1-D5 CFT and localized black holes by Ofer Aharony, Ronny Frumkin, and Jonathan Mehl (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-11181.txt.

Introduction

Ofer Aharony, Ronny Frumkin, and Jonathan Mehl map which phases dominate the D1-D5 conformal field theory at each energy in the microcanonical ensemble. As small localized black holes grow, they can begin to fill the three-sphere, the four-torus, or both. The authors analyze transitions among uniform and non-uniform black hole phases. When the torus is much larger than the AdS radius, they find evidence for a novel lattice phase of black holes. Its entropy grows linearly with energy.

Catalog record

CREATOROfer Aharony, Ronny Frumkin, and Jonathan Mehl
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE120836
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.11181

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

The phase diagram of the D1-D5 CFT and localized black holes Ofer Aharony, Ronny Frumkin, and Jonathan Mehl Abstract In this paper we analyze the phases that dominate the microcanonical ensemble at various energies in the D1-D5 CFT, which is dual to type II string theory on A ​ d ​ S 3 × S 3 × T 4 AdS_{3}\times S^{3}\times T^{4} . We focus on black hole solutions, and on the dependence of the phase structure on the ratio of the size of the torus to the AdS scale; as small localized black holes (with horizon topology S 8 S^{8} ) grow, they can start to fill the S 3 S^{3} or the T 4 T^{4} or both, and we analyze the general aspects of the transitions between the various phases of uniform and non-uniform black holes, incorporating known solutions and discussing the properties of additional unknown solutions. Some features of the transitions between these phases are similar to higher dimensional AdS spaces, while other features are different. We provide evidence that when the torus is much larger than the AdS radius, there is a large range of energies where the typical states are a novel phase, described by a lattice (in the T 4 T^{4} directions) of black holes with horizon topology S 5 × S 3 S^{5}\times S^{3} . In this phase the entropy is linear in the energy, with a coefficient that is of order the AdS radius. 1 Introduction and summary of results One of the general questions that can be asked about any physical theory is what a typical state at energy E looks like, and what is the density of states as a function of the energy, or in other words, what is the phase structure of the microcanonical ensemble as a function of the energy. In particular one can ask this question about d -dimensional conformal field theories (CFTs) on S^{d-1} (of radius L ), where this question is related by the state-operator correspondence to asking about the spectrum of local operators. At large enough energies compared to the scale 1/L of the S^{d-1} , scale-invariance implies that the entropy behaves as S(E)\propto(c_{T}EL)^{(d-1)/d} for some constant c_{T} related to the number of degrees of freedom in the theory. However, the behavior at lower energies is not universal. In “holographic CFTs” which are dual by the AdS/CFT correspondence 53 ; 37 ; 67 to some weakly coupled and weakly curved quantum gravity theory on AdS_{d+1}\times\mathcal{M} , one can answer this question using classical gravity. For example, consider the d=4 {\cal N}=4 SU(N) supersymmetric Yang-Mills theory in the ’t Hooft large N limit at large ’t Hooft coupling, dual to type IIB string theory on AdS_{5}\times S^{5} , where both the AdS_{5} and the S^{5} have a large radius of curvature R . As we review in section 3.1 (following the review 1 and references therein), at low energies the spectrum is dominated by a gas of gravitons. As the energy is increased this is replaced by a Hagedorn gas of weakly coupled strings. When the energy is further increased the strings start interacting and eventually (when the energy goes above the 10d Planck scale) become a small black hole with S^{8} horizon topology. As the energy is increased the horizon size becomes of order R , and then the solutions fill the S^{5} and undergo a topology-changing transition to black holes with horizon topology S^{3}\times S^{5} 1 1 1 As a matter of convention, throughout this paper we label horizon topologies so that any compact manifold wrapped by the black object is written on the right. that are non-uniform on the S^{5} , and this branch of solutions then connects to the branch of AdS_{5} -Schwarzschild black holes that are uniform on the S^{5} , at the point where the latter black holes have a Gregory-Laflamme-type instability 9 ; 55 ; 46 ; 25 ; 26 . As we review below, the actual transition between the dominant states in the microcanonical ensemble is a first order transition between the localized black holes with S^{8} topology and the uniform black holes. The latter black holes dominate for all higher energies, and have the expected universal high-energy behavior. In the canonical ensemble only the graviton gas and the uniform black holes appear, with a Hawking-Page transition between them at a temperature T\sim 1/L 44 ; 68 . The same theory at weak ’t Hooft coupling has a similar phase diagram 60 ; 2 but without the localized black holes, such that there is a direct transition between the Hagedorn phase and the uniform black hole phase. Other holographic theories with d\geq 3 have a similar phase diagram, except that the Hagedorn phase only appears when the gravitational description involves a weakly coupled string theory. In this paper we describe the answer to this question for the 1+1 -dimensional “D1-D5 CFT” that is holographically dual to type II string theory on AdS_{3}\times S^{3}\times T^{4} , when it has a large central charge c=6Q_{1}Q_{5}\gg 1 (here Q_{1} is the number of D1-branes and Q_{5} the number of D5-branes, when we view the CFT as arising from the low-energy limit of Q_{5} D5-branes wrapping T^{4} and Q_{1} D1-branes on top of them). When Q_{1} and Q_{5} are both large and mutually prime (this is the case we will focus on in this paper), there are many similarities between this case and the case discussed in the previous paragraph, but there are also some interesting differences. Like the {\cal N}=4 SYM theory, the D1-D5 CFT has a conformal moduli space of exactly marginal deformations (whose structure was recently reviewed in 3 ), such that in one region of this parameter space the holographic description is weakly coupled and weakly curved (so it is well-approximated by type II supergravity on AdS_{3}\times S^{3}\times T^{4} ), while in another region the CFT is a free orbifold (an {\cal N}=(4,4) supersymmetric sigma model on (T^{4N}/S_{N})\times T^{4} , where N=Q_{1}Q_{5} ). One minor difference from the previous case is that when N has several different decompositions into a product of relatively prime integers, there are different weakly coupled and weakly curved limits of the moduli space, with different values of the string scale (and thus different properties for the Hagedorn phase). A more important difference is that one of the exactly marginal deformations of the CFT controls the size of the T^{4} (for simplicity we will discuss only the case where the four circles have an equal radius R_{T} , the generalization to different radii is not trivial but is straightforward). From the holographic space-time point of view, this size R_{T} can be of order the string scale, it can be larger than the string scale but smaller than the radius R of AdS_{3} and S^{3} (which is of order N^{1/4} in 6d Planck units), or (for Q_{1}\gg Q_{5} ) it can be larger than R , and as we will discuss in detail, the microcanonical phase diagram looks quite different in these different regions of the parameter space. The phase diagram at very low and at very high energies is always similar to the AdS_{5}\times S^{5} case 2 2 2 More precisely, this

Read the complete text