Universality in logarithmic temperature corrections to near-extremal rotating black hole thermodynamics in various dimensions

Published for SISSA by Springer
Received: February 16, 2024

Accepted: May 2, 2024
Published: June 6, 2024
Universality in logarithmic temperature corrections to

near-extremal rotating black hole thermodynamics in
various dimensions
JHEP06(2024)034
Sabyasachi Maulik ,a Leopoldo A. Pando Zayas ,b,c Augniva Ray d
and Jingchao Zhang b

a
Department of Physics, Indian Institute of Technology Kanpur,
Kalyanpur, Kanpur, Uttar Pradesh 208016, India
b
Leinweber Center for Theoretical Physics, University of Michigan,
Ann Arbor, MI 48109, U.S.A.
c
The Abdus Salam International Centre for Theoretical Physics,
34014 Trieste, Italy
d
Asia Pacific Center for Theoretical Physics,
Postech, Pohang 37673, Korea
E-mail: mauliks@iitk.ac.in, lpandoz@umich.edu, augniva.ray@apctp.org,
jingchaz@umich.edu, mauliks@iitk.ac.in, lpandoz@umich.edu,
augniva.ray@apctp.org, jingchaz@umich.edu
Abstract: The low-temperature thermodynamics of near-extremal rotating black holes has

recently been revisited to incorporate one-loop contributions that are dominant in this regime.
We discuss these quantum corrections to the gravitational path integral for asymptotically
Anti de-Sitter black holes in four and five dimensions. In four dimensions we explicitly consider
Kerr-AdS4 , Kerr-Newman-AdS4 and the rotating black hole in N = 4 gauged supergravity
with two scalars and two electric charges turned on. In five dimensions we explicitly address
the asymptotically flat Myers-Perry black hole and the Kerr-AdS5 black hole. In every case
we find that tensor modes contribute 32 log THawking to the low-temperature thermodynamics.
We identify the root cause of this universality in two facts: (i) the universal presence of a
SL(2, R) subgroup of isometries in the near-horizon geometry and (ii) a set of cancellations in
the Lichnerowicz operator. We show that these two conditions hold for near-extremal black
holes in asymptotically flat and asymptotically AdS spacetimes of various dimensions.
Keywords: Black Holes, Models of Quantum Gravity, Black Holes in String Theory
ArXiv ePrint: 2401.16507
Open Access, © The Authors.

https://doi.org/10.1007/JHEP06(2024)034
Article funded by SCOAP3 .
Contents
1 Introduction 2
2 Kerr-AdS4 black hole 3

2.1 Near-extremal, near-horizon geometry 4
2.2 Quantum corrections to the entropy and zero modes 7
2.3 Log T corrections from zero modes 10
3 Kerr-Newman-AdS4 11
JHEP06(2024)034
3.1 Ensemble choice and near-horizon limit 12
3.2 Lichnerowicz operator and the zero modes 13
3.3 The lifted eigenvalues and entropy correction 13
3.4 Cancellations within the Lichnerowicz operator 14
4 Kerr-Newmann-AdS black hole in N = 4 gauged supergravity 15

5 Myers-Perry black holes in five dimensions 19

5.1 Near horizon limit of the MP black hole 21
5.2 Zero modes and correction to black hole entropy 22
5.3 Comment on the case of two different angular momenta 23
6 Kerr-AdS5 black hole 25

6.3 The lifted eigenvalue and entropy correction 28
6.4 Results for different angular momenta 28
6.5 Comments on the rotating black hole in minimal gauged supergravity 29
3
7 On the origin of the universal 2
log T behavior 30
8 Conclusions 32
A Ensemble choice 34
A.1 A new ensemble choice and near-horizon limit 34
A.2 Lichnerowicz operator and the zero modes 36
A.3 The lifted eigenvalue and entropy correction 36
B SL(2, R) symmetries of the near-horizon geometries 37

B.1 Near-horizon symmetry of Kerr-AdS4 black hole 37
B.2 Near-horizon symmetry of Kerr-AdS black hole in various dimensions 39
C The structure of the Lichnerowicz operator 39
–1–
1 Introduction
The low temperature breakdown of black hole thermodynamics for extremal black holes was
pointed out more than thirty years ago [1]. The conundrum consists in that, at very low
temperature, the emission of one Hawking quantum can drastically change the temperature
of the near-extremal black hole. It is quite remarkable that the resolution to this puzzle did
not require knowledge of the full path integral (quantum gravity) and was first achieved by a
careful treatment of certain zero modes in the extremal solution [2–4]. The realization that
temperature effectively acts as a coupling constant whereby the high-temperature regime
is classical while the very low-temperature regime is quantum and strongly coupled was
JHEP06(2024)034
understood first in the context of two-dimensional Jackiw-Teitelboim (JT) gravity [5–7].
The key realization in [2–4] resides in understanding that the near-horizon region of the
higher-dimensional black hole contains a JT subsector that dominates the full path integral. In
this sense, a once interesting toy model of two-dimensional gravity finds impactful implications
for the physics of higher-dimensional black holes. These works emphasized the role of zero
modes in the full geometry and their connection with the Schwarzian action (see [8, 9] for
reviews), thus pointing to universality within high-dimensional extremal black holes. Any
higher-dimensional gravity theory admitting near-extremal solutions will admit such zero
modes and, consequently, the low-temperature thermodynamics will be accordingly corrected.
In practice, embedding quantum aspects of JT gravity in spherically symmetric near-
extremal black holes is almost immediate. Indeed, the relevance of JT for Reissner-Nordstrom
black holes is directly suggested by the fact that the near-horizon geometry of extremal
Reissner-Nordstrom is AdS2 × S p . Another recent spherically symmetric example is [10] which
relies on the properties of the modes to resolve an old puzzle concerning the entanglement
entropy of holographic quantum fields in Rindler space.
It is natural to ponder to what extent this relationship might hold true for rotating black
holes. Gravitational perturbations to near-extremal Kerr spacetime were studied in [11, 12],
where a Schwarzian sector was identified. Recently log T correction to the thermodynamics
of Kerr spacetime was addressed in [13, 14]. There it was explicitly shown that the near-
horizon extremal Kerr solution admits a family of normalizable zero modes corresponding to
reparametrizations of the boundary time, just as in JT gravity. The path integral over these
zero modes leads to an infrared divergence in the one-loop approximation to the Euclidean
partition function. This divergence can be regulated following recent suggestion of turning
on a small but finite temperature correction in the geometry [15–17]. The resulting finite-
temperature geometry lifts the eigenvalues of the zero modes, rendering the path integral
infrared finite and leading to the thermodynamic-altering correction to the near-extremal
black hole: 32 log THawking .
In this manuscript, largely motivated by applications to the AdS/CFT correspondence,
we perform a similar path integral treatment for asymptotically AdS black holes. We consider
Kerr-AdS4 , Kerr-Newman-AdS4 and a near-extremal black hole in N = 4 gauged supergravity
that contains two scalar fields and two Maxwell fields. We then turn to five dimensions
where we consider the asymptotically flat Myers-Perry five-dimensional black hole as well
as the Kerr-AdS5 black hole, explicitly. In every case we find that the contribution to the
path integral given by the gravitional tensor modes is 32 log THawking . Along the way, we
–2–
gain understanding into this universality as a consequence of symmetries preserved near the
horizon and particular properties of the kinetic (Lichnerowicz) operator for tensor modes.
An added motivation for our work are the positive results in the framework of viewing
certain black holes in the AdS/CFT correspondence through the lens of the Kerr/CFT
correspondence. In [18] aspects of the Kerr/CFT correspondence were applied to the
near-extremal black holes in AdS4,5,6,7 and yielded a microscopic explanation for their
entropy, besides the canonical AdSd+1 /CFTd one. Some sub-leading aspects, such as certain
logarithmic corrections to the entropy of AdS5 black holes and black strings, were also
successfully matched [19]. It is natural to expect similar and complementary quantum
insights to arise by championing the JT point of view for the near-horizon geometries. We
JHEP06(2024)034
hope that this synergy of two lower-dimensional approaches will shed light on important
quantum aspects of higher-dimensional black holes.
The rest of the manuscript is organized as follows. In section 2 we review the geometry of
Kerr-AdS4 , discuss its near-horizon extremal limit and explicitly derived the low-temperature
contribution to the thermodynamics. Section 3 considers Kerr-Newman-AdS4 black holes.
In section 4 we discuss the rotating black hole with two scalars and two electric charges
turned on in N = 4 four-dimensional gauged supergravity. Section 5 present details of the
tensor modes computation for the asymptoticaly flat five-dimensional Myers-Perry black
hole while in section 6 we discuss the Kerr-AdS5 black hole explicitly. We collect a number
of observations and remarks regarding the universality of the logarithmic in temperature
correction from tensor modes in section 7 and conclude in section 8. We relegate some
technical material and a few very long expressions to a series of appendices.
2 Kerr-AdS4 black hole
The Kerr-AdS4 black hole is a solution of the equations of motion of the Einstein-Hilbert
action in (3 + 1) spacetime dimensions with a negative cosmological constant (we always
work in units where the Newton’s constant GN = 1.)
1
Z
4 √ 1
Z √
Igrav = − d x −g (R − 2Λ) − d3 y h K. (2.1)
16π M 8π ∂M
The geometry is described by the metric ([20, 21], we will follow the notation of [22])
!2
∆r Σ2 2 Σ2 2 ∆θ r2 +a2
2
a

ds2 = − 2 dt− sin2 θdϕ + dr + dθ + 2 sin2 θ adt− dϕ , (2.2)
Σ Ξ ∆r ∆θ Σ Ξ
where
!
r2 a2
∆r = r + a
2 2
a+ 2 − 2M r , ∆θ = 1 − cos2 θ , (2.3)
L L2
a2 p
Ξ=1− , Σ= r2 + a2 cos2 θ . (2.4)
L2
The cosmological constant Λ is related to the length-scale L by Λ = − L32 . The black hole is
characterized by two conserved quantities, namely its energy E and angular momentum J:
M Ma
E= , J= . (2.5)
Ξ2 Ξ2
–3–
The thermodynamic properties of this black hole have been extensively studied in the
literature, see e.g. [22, 23]. The outer event horizon r+ is the largest real, positive root of
∆r = 0. The Bekenstein-Hawking entropy is related to the area of the event horizon in
the usual manner, and is given by
2 + a2

π r+
S= , (2.6)
Ξ
while the surface gravity κ, and the temperature are
2

a2 3r+ a2
r+ 1 + L2
+ 2 − 2
κ L r+
JHEP06(2024)034
T = = . (2.7)
2π 4π r+
2 + a2

The angular velocity of the black hole relative to a frame which is non-rotating at infinity,
is given by
2

r+
a 1+ L2
Ω= 2 + a2 . (2.8)
r+
Together, these quantitites satisfy the first law of black hole thermodynamics for the outer
event horizon
dE = T dS + ΩdJ. (2.9)
2.1 Near-extremal, near-horizon geometry

We are interested in the near horizon geometry of the Kerr-AdS4 black hole for small
deviations away from extremality. For this purpose, let us rewrite the function ∆r in the
black-hole metric (2.2) as
1 L 2 L 2 + r+2 + r2 + r r
!
− + −
∆r = 2 (r − r+ ) (r − r− ) r2 + (r+ + r− ) r + 2
, (2.10)
L L − r+ r−
where r± are the two real roots of ∆r = 0, they describe the outer and inner horizon radii
of the black hole. In terms of these two parameters, the mass parameter and the angular
momentum are expressed as
(r− + r+ ) L2 + r−
2 L 2 + r+
2

M= , (2.11)
2L2 (L2 − r+ r− )
r
√ L2 (L2 +r− + −)
2 +r 2 +r r
L r+ r− (r− + r+ ) L2 + 2 L2 L2 + 2
+
r− − r+ r− r+ L2 −r+ r−
J= . (2.12)
2 L4 − 2L2 r+ r− − r+ r− r−
2 +r r + 2

r+ 2
+ −
Note that a, which is related to J by equation (2.5), is given by

√
L r+ r − L 2 + r +
2 + r2 + r r

− + −
a= . (2.13)
L 2 − r+ r−
The merit of such parameterization is that the extremal limit can be easily defined as the
limit when the two horizons coalesce into one, and the temperature becomes zero. Let
–4–
us denote the radius of the event horizon at extremality as r0 . We find that r± have the
following small temperature (T ) expansion
4πL2 r02 L2 + r02 4π 2 L2 r03 L4 + 3r04 5L6 + 11L4 r02 − 17L2 r04 + 9r06 2

r± = r0 ± 4 T ± T +...,
L + 6L2 r02 − 3r04
3
L4 + 6L2 r02 − 3r04
(2.14)
and from equations (2.11) and (2.13), M , a and S receive corrections which are given by
2
r0 L2 +r2 4 π 2 L4 r03 +2π 2 L2 r05 +9π 2 r07 L2 +r02 2

M = 2 2 02 + T +. . . , (2.15)
JHEP06(2024)034
L2 −r02 L4 +6L2 r02 −3r04

L L −r0
Lr0
a= × (2.16)
(L−r0 )(L+r0 )
4π L2 +3r02 L10 +5L8 r02 −2L6 r04 −202L4 r06 +161L2 r08 −27r010
!
2 L2 r 2

L 2
+3r02 − 0
3 T 2
.
L4 +6L2 r02 −3r04
r2

2πr02 8π 2 r03 1− L02
S= 3r 2
+ 3r 2

6r2 3r4
T +. . . (2.17)
1− L20 1− L20 1+ L20 − L40
This immediately sets a temperature scale for us. To see this, we note that (2.17) is a series
expansion for entropy and therefore, in units where the Boltzmann constant is set to 1, the
terms in the series are dimensionless.
Therefore,
to make the terms dimensionless, a scale is
2
3r0 2
6r0 3r 4
1− 1+ − 40
L2 L2
introduced, given by Tq = , i.e., Tq is inverse of the coefficient sitting in
L
r2

8π 2 r03 1− 0
L2
front of T in the above series expansion. We will use this insight in all subsequent sections.
Note that while doing this expansion we choose to keep the angular momentum J fixed
at its extremal value, i.e., J receives no T corrections and it is given by
q
Lr02 L2 − r02 L4 + 3L2 r02

J= 2 . (2.18)
L2 − 3r02
Therefore, we have adopted the canonical ensemble here. However, we should add that
this is more of an aesthetic choice than a necessary condition; in fact in appendix A, we
show that the result remains unchanged if we choose to keep M fixed, and let J depend
on the temperature. Note that the leading correction to the mass occurs at O T 2 . This

observation is at the heart of the original discussion of the breakdown of thermodynamics at

low temperatures in [1]. Recent works [13, 14, 16] clarify important quantum corrections to
this dependence and resolve the breakdown of thermodynamics for rotating black holes.
Taking inspiration from [13, 14, 16] we introduce the following change of coordinates: we
do a ‘Bardeen-Horowitz’ like scaling transformation {(r, t, θ, ϕ) → (y, τ, θ, φ)} to obtain for the
near-horizon geometry of the near-extremal Kerr-AdS4 black hole [24]. The transformation
–5–
is given by
4πL2 r02 L2 + r02

r = r+ (T ) + 4 T (y − 1) , (2.19)
L + 6L2 r02 − 3r04
iτ
t=− , θ = θ, (2.20)
2πT r
 
L4 +3L2 r2
 L − 3r0
2 2

0
L2 −r02 1
ϕ = φ − iτ  − 1 (2.21)

.

4πL r0
3 T
JHEP06(2024)034
This allows us to simultaneously zoom into the near-horizon region of the black hole and also
take a small temperature limit. At leading order, the above transformation gives us
O(1) : ds2(0) = ḡµν dxµ dxν

!
dy 2
= g1 (θ) + y 2
− 1 dτ 2 + g2 (θ)dθ2 + g3 (θ) (dφ + ig4 (y) (y − 1) dτ )2 ,
y2 − 1
(2.22)
where
L2 r02 L2 − r02 + L2 + 3r02 cos2 θ

g1 (θ) = , (2.23)
L4 + 6L2 r02 − 3r04
2 L6 − L2 r04

g2 (θ) = − L2 , (2.24)
2L4 − 3L2 r02 − 3r04 − r02 L2 + 3r02 cos 2θ

4L2 r02 sin2 θ L4 − L2 r02 − r02 L2 + 3r02 cos2 θ

g3 (θ) = 2 , (2.25)
L2 − 3r02 L2 − r02 + L2 + 3r02 cos2 θ

r
L4 +3L2 r02
L5 + 6L3 r02 3Lr04 L4 + 4L2 r02 − 3r04 y

− − L2 −r02
g4 (y) = . (2.26)
L(y − 1) L4 + 6L2 r02 − 3r04

This geometry is the analogue of the near-horizon extreme Kerr (NHEK ) spacetime [13,
14, 24] and reduces to the same in the large AdS4 radius L → ∞ limit. It may further
be checked the geometry in equation (2.22) has a Ricci scalar R = − L122 . We shall call
this geometry NHEK-AdS.
By plugging in the expressions for M and a from equations (2.15) and (2.16), we also
obtain the O(T ) correction to the metric in (2.22) given by
O(T ) : ds2(1) = T gµν

(1)
dxµ dxν . (2.27)
(1)
The expressions for gµν are typically too lengthy to be presented and not particularly
illuminating.
This concludes our section on near horizon near extremal Kerr geometries and its
correction. We now turn to the problem of computing the log T corrections, arising from
the zero modes of the Lichnerowicz operator.
–6–
2.2 Quantum corrections to the entropy and zero modes
We are interested in the quantum corrections to the entropy which scale as the logarithm in
temperature. This requires a path integral over the massless modes propagating in the near
horizon throat. We perform this integral by doing a saddle point analysis, where the saddle
is given by the NHEK-AdS geometry in (2.22). Schematically, we consider gµν = ḡµν + hµν ,
for a normalizable perturbation h, and evaluate the 1-loop partition function around a
given saddle point ḡ
1
Z " Z #
Z = exp(−I(ḡ)) D[h] exp − d x ḡ h̃ ∆ hµν ,
4 µν Λ
(2.28)
p
64π
JHEP06(2024)034
with h̃µν = hµν − 12 ḡ µν hµν . Note that here we use the gauge fixed action which satisfies
the harmonic gauge condition given by
1 1 ¯µ α 1 ¯ν β

LGF = ¯ αµ
ḡµν ∇α h − ∇ h α ¯ βν
∇β h − ∇ h β . (2.29)
32π 2 2
For this gauge fixed action, the linearized kinetic operator for hµν , also known as the
Lichnerowicz operator, is given by [25]
1 ¯ 2 hµν +2R̄µρ hρ −2R̄µρνσ hρσ −4 R̄νσ − 1 ḡνσ R̄ hσ −2Λhµν .
!
∆ hµν =
Λ
−∇ (2.30)
32π ν
4 µ
As we shall see in section 2.3, the log T corrections arise from the zero modes of the
Lichnerowicz operator. In the remaining part of this section we will evaluate the zero modes
of the same, anticipating the arguments given in next section. Armed with the expressions for
the zero modes, we would then be ready to directly evaluate the log T correction in section 2.3.
Typically, the Lichnerowicz operator is quite intractable, more so in our case where the
metric is not Ricci flat. We would need a judicious ansatz for the vector field generating
the zero modes as diffeomorpshisms. Following the prescription as given in [14], we choose
an ansatz for the vector field as
∂ ∂ ∂

ξ =e
(n) inτ
f1 (y) + f2 (y) + f3 (y) (2.31)
∂y ∂τ ∂φ
(0)
and we generate diffeomorphisms given by hµν = Lξ gµν . The goal now is to find the functions
f1,2,3 (y) such that the hµν generated belongs to the kernel of the Lichnerowicz opertor. We
state the result first. The vector fields are given by (for |n| ≥ 2)
ξ (n) =
|n|−2
2
|n|
2
ieinτ y−1
n2 +|n|y+y 2 −1 einτ y−1
(|n|+y)

y+1 y+1
∂τ + ∂y +
2|n| (n2 −1) (y+1)2 2 (n2 −1)
 
|n| r
L4 +3L2 r02
2
einτ y−1 L5 n2 y+|n| −L4 n2 +|n|y+y 2 −1 −


 y+1 L2 −r 2 0


 r r !
 
L4 +3L2 r02 L4 +3L2 r02
6L2 r02 n2 +|n|y+y 2 −1 +3r04 n2 +|n|y+y 2 −1 −9Lr04 (n2 y+|n|)
 
L2 −r02 L2 −r02
r ∂φ
L4 +3L2 r02
2|n| (n2 −1) (y 2 −1) L2 −r02
(L4 +6L2 r02 −3r04 )
(2.32)
–7–
such that the diffeomorphism generated by this vector field (given in equation (2.39)) belongs
to the kernel of the Lichnerowicz operator. Upon introducing a new radial coordinate
y = cosh η, we may also write the above as
ξ (n) =
η
einτ csch2 η tanh|n|
2
i n2 +cosh η (cosh η+|n|)−1 ∂τ +(|n| sinh η (cosh η+|n|)) ∂η +
2|n| (n2 −1)
r
L4 +3L2 r02
L4 −9r04 n2 cosh η+|n| L4 +6L2 r02 −3r04 n2 +cosh η (cosh η+|n|)−1

L − L2 −r02
!
JHEP06(2024)034
r ∂φ .
L4 +3L2 r02
L4 +6L2 r02 −3r04

L2 −r02
(2.33)
Typically, it becomes an involved problem to solve directly for the functions fi (y). Borrowing
inspiration from the treatment of zero modes in [14], we illustrate below some of the steps
to solve for the vector field. We act the Lichnerowicz operator (∆Λ ) on the spin 2 field
hµν = 2∇(µ ξν) which represents the change of the NHEK-AdS metric under the diffeomorphism
generated by the vector field ξ. Specifically, we observe that
θ= π2
(∆Λ h)θθ ∼ inf2 (y) + f1′ (y) (2.34)

θ=0
where the proportionality factor is a complicated function of the metric functions, but is
independent of any other fi . This immediately sets
if1′ (y)
f2 (y) = . (2.35)
n
Next, we choose an ansatz similar to [14] for f3 (y) and proceed. That is, we choose
(y − 1)f1′ (y) − f1 (y)

f3 (y) = + δf3 (y) (2.36)
n
and solve for δf3 (y) from the equation
1
∂θ (∆Λ hθy ) + ∂θ (∆Λ hθy ) = 0. (2.37)
θ=0 2 θ= π2
Here, we see a difference with the Kerr case [14], and we do not obtain δf3 = 0. Rather, we
get a complicated function that we do not write here for brevity. We also do a consistency
check which shows that δf3 vanishes as L → ∞, as expected. We are ultimately left with a
single undetermined function f1 (y), which can be solved by brute force. Finally, we arrive at
|n|
y−1 (|n| + y)

2
f1 (y) = , |n| ≥ 2. (2.38)
y+1 2 (n2 − 1)
–8–
This is exactly the solution found by the authors in [14]. Once we have found the functions
fi ’s, the family of normalized zero modes of the Lichnerowicz operator is obtained as
s
3|n| (n2 − 1) L6 + 3L4 r02 − 21L2 r04 + 9r06 L2 − r02 + L2 + 3r02 cos2 θ

h(n) µ
µν dx dx
ν
=
64π 2 L L4 + 6L2 r02 − 3r04
|n|
y−1
!
|n| dτ dy dy 2

2
e inτ
−dτ + 2i
2
+ , |n| ≥ 2 .
y+1 n y 2 − 1 (y 2 − 1)2
(2.39)
One can immediately check that
JHEP06(2024)034
(∆Λ h)µν = 0 , µ, ν ∈ {y, τ, θ, φ}. (2.40)
(n)
The zero modes in equation (2.39) are generated by diffeomorphisms h µν ∝ Lξ ḡµν ,
where the vector fields are given by (2.32). Note that they reduce to the same set of vector
modes as is reported in [14] in the large L limit.
Large diffeomorphism. At the very onset, we should point that the diffeomorphisms
generated by the vector field ξ are large gauge transformations and they can not be gauged
away. To see this, we first note that the vector field ξ does not die off at the asymptotic
boundary of AdS2 . Specifically from (2.33), we see that
einτ
ξ (n) = (|n|∂η + i∂τ − ∂φ ) , (2.41)
η→∞ 2|n| (n2 − 1)
(n)
i.e., the vector fields themselves are O(1) near the boundary. However, the zero modes {hµν }
themselves are normalizable and are part of the physical spectrum and need to be integrated
over in the path integral. We use the normalization [26]
Z p ∗ µν
⟨h, h⟩ = d4 x ḡ h(n) h(n) (2.42)
µν
as would be appropriate for the space of L2 integrable functions necessary to define path
integrals. Setting the above inner product one, i.e.,
Z p ∗ µν
d4 x ḡ h(n) h(n) =1 (2.43)
µν
(n)
fixes the coefficient for hµν dxµ dxν in equation (2.39).
A bit more insight might be had from the behaviour of the diffeomorphism generating
vector fields (2.32) at the boundary. In the coordinates {η, τ, θ, φ}
∞
ξ ∼ ε′ (τ )∂η − ε(τ )∂τ − iε(τ )∂φ , ε(τ ) = fn exp(inτ ) . (2.44)
X
n=2
The vector field then generates the reparametrization η → η + ε′ (τ ) , τ → τ − ε, φ → φ − iε.

Following the arguments presented in [7, 15], we conclude that the vector fields generating
the large diffeomorphisms are the Schwarzian modes which act on the boundary via (2.44).
–9–
2.3 Log T corrections from zero modes
In this section, we motivate the claim made in the last section that only the zero modes of the
Lichnerowicz operator contribute to log T corrections. The claim is based on perturbation
theory. Then, building on that claim, we actually evaluate the log T correction, where the
final result is presented in equation (2.52).
First, we note that the correction to metric (2.22) induces a correction δ∆Λ to the
Lichnerowicz operator. In order to find the eigenvalue of the corrected Lichnerowicz operator
(∆Λ + δ∆Λ ), we take help of perturbation theory. At first order, we have

∆Λ + δ∆Λ (hn + δhn ) = λ0n + δλn (hn + δhn ) , (2.45)
JHEP06(2024)034
where {hn , λ0n } is the eigenspectrum and {δhn , δλn } is the respective correction induced.
Then, at first order, we have
Z p αβ,µν
δλn = d4 x ḡ h(n) δ∆Λ h(n) . (2.46)
αβ µν
Since, the partition function in (2.28) is a Gaussian integral over bosonic fields, ignoring
overall factor, which play little role, we have
1
log Z ∼ − log λ0n + δλn (T ) . (2.47)
Y X
Z∼ ,
n (λn + δλn (T ))
0
n
To proceed further, we need two crucial observations. We observe that λ0n is T -independent,
being the eigenvalue of the Lichnerowicz operator evaluated at O(1) and therefore, log T
corrections would necessarily come from the corrected eigenvalue δλn (T ). Furthermore, for
non-zero a, we note that the Maclaurin expansion of log(a + x(T )) is always a polynomial
series in small x (T ). Specifically, we mean that
∞
xn (T )
log(a + x(T )) = log a + (2.48)
X
.
n=1
an n
Therefore, corrections of the form log (x(T )) come only when a = 0. Adapting this observation
to our case, this implies that the log T corrections are obtained only from the zero modes of
the Lichnerowicz operator when λ0n = 0. This establishes our claim in the last section that
for the log T correction, it suffices to focus on the zero modes of the Lichnerowicz operator.
Therefore, we are left with the job of evaluating the corrected eigenvalue of the zero modes,
via equation (2.46). The integral is quite involved. However, it can be evaluated and we
quote the result here directly. We have
r02

3n 1 − L2
T
δλn = , n ≥ 2. (2.49)
64 r0
Therefore, setting λ0n = 0 in equation (2.47) we obtain the log T corrections as

 #  
r2

Y  3n 1 − L02 T 
" ∞ " ∞
64 r0
#
δ log Z = − log  = log (2.50)
Y
 .

64 r0 r02
 
log T n=2 n=2 3n 1 − L2
T
– 10 –
The infinite product above can be regularized using standard zeta function regularization [27]
given by
∞
x 3
∼ x− 2 . (2.51)
Y
n=2
n
Setting x = T −1 and ignoring constants irrelevant to the present pursuit, we obtain
3 64 r0
!
T
δ log Z = log + O(1), Tq ≡ . (2.52)
2 r02

Tq 3 1−
log T L2
JHEP06(2024)034
The constant indicated by O(1) depends on the scheme. In this manuscript we will use Tq as
the temperature scale below which the logarithmic correction become dominant.
Two final comments are in order here.
1. Note that on taking the large L limit in (2.50), one recovers the result for the Kerr
black hole in flat space presented in [13, 14].
2. The result agrees with the generic prediction of [14]. We have derived it explicitly,
and we confirm the independence of the log T correction with respect to the ensemble
(canonical versus grand canonical). For the grand canonical ensemble (where m is kept
fixed), see appendix A.
3 Kerr-Newman-AdS4
In this section, we provide the details of treating the path integral for the near-extremal
Kerr-Newman-AdS black hole in four dimensions. We highlight a cancellation pattern among
terms in the Lichnerowicz operator. Such cancellations are verified in all the explicit cases
we considered in this paper. We start with the Einstein-Maxwell action,
1 √
Z
I=− d4 x −g R − 2Λ − F 2 + Iboundary . (3.1)
16π M
The Kerr-Newman-AdS4 solution is given by [22]

!2
∆r Σ2 2 Σ2 2 ∆θ r2 +a2
2
a

ds = − 2 dt− sin2 θdϕ
2
+ dr + dθ + 2 sin2 θ adt− dϕ , (3.2)
where

∆r = r 2 + a 2 a + g 2 r2 − 2M r + q 2 , ∆θ = 1 − g 2 a2 cos2 θ , (3.3)
p
Ξ = 1 − g 2 a2 , Σ= r2 + a2 cos2 θ . (3.4)
The cosmological constant Λ is related to g ≡ L1 by Λ = −3g 2 . The gauge field is given by (we
consider only the electric case, and the magnetically charged case can be treated analogously):
qr qra sin2 θ
A=− dt + dϕ. (3.5)
Σ2 Σ2 Ξ
– 11 –
The temperature of the black hole is given by

a2 +q 2
r+ 1 + g 2 a2 + 3g 2 r+
2 −
2
r+
T = , (3.6)
4π(r+
2 + a2 )
where the r+ denotes radius of the outer horizon.
3.1 Ensemble choice and near-horizon limit

Under the extremal condition, we have
m0 = r0 + a2 g 2 r0 + 2g 2 r03 , (3.7)
JHEP06(2024)034
q
q0 = −a2 + r02 + a2 g 2 r02 + 3g 2 r04 , (3.8)
where r0 is the radius of the extreme black hole horizon. In our ensemble, we choose a and
r0 as the two free parameters, m and r+ depend on T :
m(T ) = m0 + c1 T + c2 T 2 + c3 T 3 + O(T 4 ), (3.9)
r+ (T ) = r0 + d1 T + d2 T 2 + d3 T 3 + O(T 4 ), (3.10)
where the coefficients ci , di can be obtained by solving (3.6) and ∆r = 0. We verify that
c1 = 0. Other coefficients are lengthy and not illuminating to be written here. The Hawking
entropy can be calculated using the above temperature dependence
π(a2 + r02 ) 4π 2 r0 (a2 + r02 )
S= + T + O(T 2 ). (3.11)
1−a g2 2 (1 − a g 2 )(1 + a2 g 2 + 6g 2 r02 )
2
To zoom into the near-horizon region, we conduct the following coordinate transformation:
−iτ −ik3 τ
r = r+ (T ) + T k1 (y − 1), t = , ϕ=φ+ + k4 iτ, (3.12)
k2 T T
where
2π(a2 + r02 )
k1 = ,
1 + a2 g 2 + 6g 2 r02
k2 = 2π,
−a(−1 + a2 g 2 ) (3.13)
k3 = ,
2π(a2 + r02 )
−2a(−1 + a2 g 2 )r0
k4 = 2 .
(a + r02 )(1 + a2 g 2 + 6g 2 r02 )
By taking the T → 0 limit, we obtain the near-horizon geometry at the zeroth order of T ,
!
dy 2 2
ḡµν dx dx = g1 (θ)
µ ν
+ dτ 2
y 2
− 1 + g 2 (θ)dθ 2
+ g3 (θ) dφ + ik(y − 1)dτ , (3.14)
y2 − 1
where
r02 + a2 cos2 θ
g1 (θ) = , (3.15)
1 + a2 g 2 + 6g 2 r02
r2 + a2 cos2 θ
g2 (θ) = 0 2 2 , (3.16)
1 − a g cos2 θ
– 12 –
(a2 + r02 )2 (−1 + a2 g 2 cos2 θ) sin2 θ
g3 (θ) = − , (3.17)
(−1 + a2 g 2 )2 (r02 + a2 cos2 θ)
k = −k4 . (3.18)
(3.19)
We can obtain the near-horizon gauge field at the same time,
q
Āµ dx = i r02 +3g 2 r04 +a2 (−1+g 2 r02 ) a2 +2r02 +a2 cos(2θ) (−a2 r02 (−2+y)+a4 y−2r04 y
µ

+a2 (r02 (−2+y)+a2 y) cos(2θ)) / 4(a2 +r02 )(1+a2 g 2 +6g 2 r02 )(r02 +a2 cos2 θ)2 dτ
JHEP06(2024)034
q
2ar0 r02 +3g 2 r04 +a2 (−1+g 2 r02 ) sin2 θ
− dφ.
(−1+a2 g 2 )(a2 +2r02 +a2 cos(2θ))
(3.20)
By keeping O(T ) when T → 0, we also obtain δgµν , δAµ , the near-horizon geometry and gauge
field at O(T ). These expressions are lengthy and not very illuminating to be written here.
3.2 Lichnerowicz operator and the zero modes

The Lichnerowicz operator of our Einstein-Maxwell action with a negative cosmological
constant is
hαβ ∆αβ,µν
L hµν = hαβ (∆αβ,µν
EH − 2Λ∆αβ,µν
1 − 2∆αβ,µν
F )hµν , (3.21)
where ∆αβ,µν
EH , ∆1
αβ,µν
, ∆αβ,µν
F can be found in appendix C. At zero temperature, i.e. when
g = ḡ, A = Ā, one can verify that the above operator has the following zero modes:
|n|
−1 + y
!
|n| dτ dy dy 2

2
h(n) µ
µν dx dx
ν
= cn e inτ
a +
2
2r02 + a cos(2θ)
2
dτ − 2i
2
−
1+y n y 2 − 1 (y 2 − 1)2
(3.22)
for |n| ≥ 2, where cn is the normalization constant.
3.3 The lifted eigenvalues and entropy correction

By turning on a small temperature, the “zero modes” will be lifted and gain a nonzero
eigenvalue. Now we substitute g = ḡ + δg into the Lichnerowicz operator (3.21) and apply
it to the zero modes (3.22). The lifted eigenvalue is
Z
(n)∗
δλn = dx4 ḡhαβ δ∆αβ,µν h(n) (3.23)
p
L µν .
The expression of the operator is intractable, but the integration is straight forward and
the result is simple.
3nr0 T
δλn = , |n| ≥ 2. (3.24)
8(a2 + 3r02 )
The contribution of the extremal zero modes to the low-temperature partition function
is therefore
 
3
!
Y π T
δ log Z = log   = log + O(1). (3.25)
n≥2
δλn 2 Tq
– 13 –
The O(1) term is not reliable. Tq is a temperature scale under which the quantum correction
should be considered. We use the inverse of the O(T ) coefficient of the classical entropy (3.11)
as this scale
(1 − a2 g 2 )(1 + a2 g 2 + 6g 2 r02 )
Tq = . (3.26)
4π 2 r0 (a2 + r02 )
This is the temperature we expect the classical low temperature thermodynamics to break down.
3.4 Cancellations within the Lichnerowicz operator

If calculating every term in the Lichnerowicz operator respectively, one can find a pattern
JHEP06(2024)034
that hold in all cases considered in this paper: only two terms really contribute to the lifted
eigenvalue. The other terms always cancel with each other. In this subsection, we use the
example of Kerr-Newman-AdS4 black hole to illustrate this cancellation. For convenience, we
first rename the terms that appear in the Lichnerowicz operator (3.21),
1 αµ βν

Term 1a : hαβ δ g g □ hµν
2
1

Term 1b : hαβ δ − g αβ g µν □ hµν = 0
4
Term 2 : hαβ δ(R αµβν
)hµν
Term 3a : hαβ δ(Rαµ g βν )hµν
Term 3b : hαβ δ(−Rαβ g µν )hµν = 0

1

Term 4a : hαβ δ − Rg αµ g βν hµν
2
1 αβ µν

Term 4b : hαβ δ Rg g hµν = 0
4
Term 5a : hαβ δ(Λg αµ g βν )hµν
1

Term 5b : hαβ δ − Λg αβ g µν hµν = 0
2
1 2 αµ βν

Term 6a : hαβ δ F g g hµν
2
1

Term 6b : hαβ δ − F 2 g αβ g µν hµν = 0
4
Term 7 : hαβ δ(−2F αµ F βν )hµν
Term 8a : hαβ δ(−4F αγ F µγ g βν )hµν
Term 8b : hαβ δ(2F αγ F βγ g µν )hµν = 0 , (3.27)
where δ means the above expressions are of order T . As hµν are zero modes, the O(1) results
will cancel with each other automatically. We find two patterns for the above terms,
• As indicated in (3.27), all terms containing g µν hµν are zero by themselves.
• Term 3a + Term 4a + Term 5a + Term 6a + Term 7 + Term 8a = 0.
– 14 –
Therefore, only two terms contribute non-trivially
1 αµ βν

hαβ ∆αβ,µν hµν = Term 1a + Term 2 = hαβ δ g g □ + Rαµβν hµν . (3.28)
L
2
It is noteworthy that the result of (3.28) can be verified for all cases included in this paper,
including the upcoming supersymmetric case and five-dimensional cases in the corresponding
next sections. For the supersymmetric case in section 4, we have two gauge fields and two addi-
tional scalar fields. There are more terms in the Lichnerowicz operator, but only the two terms
indicated in (3.28) contribute non-trivially. This cancellation naturally reduces to previous
studies of asymptotically flat Kerr and Kerr-Newman black holes in four dimensions [13, 14].
JHEP06(2024)034
4 Kerr-Newmann-AdS black hole in N = 4 gauged supergravity
The 4d N = 4 gauged supergravity can be obtained by the truncation of the 11d supergrav-
ity [28], here we exclusively focus on its bosonic sector which is given by
1 1 1 1
L4 = R ∗ 1 − ∗ dζ ∧ dζ − e2ζ ∗ dχ ∧ dχ − e−ζ ∗ F(2)2 ∧ F(2)2 − χF(2)2 ∧ F(2)2
2 2 2 2
1
− ζ 2ζ
e ∗ F(2)1 ∧ F(2)1 − e χF(2)1 ∧ F(2)1 (4.1)
2 (1 + χ2 e2ζ )

+ g 2 4 + 2 cosh ζ + eζ χ2 ∗ 1,
where ζ and χ are the dilaton and axion. The subscript in parenthesis denotes the degree
of the form. The solution has two pairwise equal charges and therefore two gauge potential
A(1)1 and A(1)2 .
The non-extremal rotating, electrically charged asymptotically AdS4 black hole solution
with gauge group U(1) × U(1) in 4d N = 4 gauged supergravity was constructed in [28].
The solution is characterized by four parameters (a, m, δ1 , δ2 ). The metric, the scalars and
the gauge fields are given by
!2 #2
∆r a sin2 θ ∆θ sin2 θ r1 r2 + a 2
! "
dr2 dθ2
ds2 = − dt − dϕ +W + + a dt − dϕ ,
W Ξ ∆r ∆θ W Ξ
(4.2)
r12 + a2 cos2 θ a(r2 − r1 ) cos θ

eζ = , χ= 2 ,
W r1 + a2 cos2 θ
√
2 2m sinh(δ1 ) cosh(δ1 ) r2 a sin2 θ
!
A1 = dt − dϕ + α41 dt , (4.3)
W Ξ
√
2 2m sinh(δ2 ) cosh(δ2 ) r1 a sin2 θ
!
A2 = dt − dϕ + α42 dt ,
W Ξ
where
ri ≡ r + 2m sinh2 (δi ) , (i = 1, 2)
∆r ≡ r2 + a2 − 2mr + g 2 r1 r2 (r1 r2 + a2 ) ,
– 15 –
∆θ ≡ 1 − g 2 a2 cos2 θ ,
W ≡ r1 r2 + a2 cos2 θ ,
Ξ ≡ 1 − a2 g 2 , (4.4)
and g ≡ L−1 is the inverse of the AdS4 radius. Note that we have added pure gauge terms
to the two gauge fields where α41 and α42 are constant.
The non-extremal asymptotically AdS4 black holes with four degenerate electric charges
(Q1 = Q2 , Q3 = Q4 ) and one angular momentum J have been found in [29], which are
characterized by four parameters (a, m, δ1 , δ2 ). The BPS limit imposes a condition
JHEP06(2024)034
2
e2δ1 +2δ2 = 1 + . (4.5)
ag
For the black hole solution to have a regular horizon, we impose an additional constraint
cosh2 (δ1 + δ2 )
(mg)2 = . (4.6)
eδ1 +δ2 sinh3 (δ1 + δ2 ) sinh(2δ1 ) sinh(2δ2 )
The two conditions (4.5) (4.6) in [29] have typos, which have been corrected in [30, 31], see
also [32]. With these constraints, there are only two independent parameters for asymptotically
AdS4 BPS black holes, which we choose to be (δ1 , δ2 ) for convenience. In the BPS limit,
the position of the outer horizon is
2m sinh(δ1 ) sinh(δ2 )
r+ = , (4.7)
cosh(δ1 + δ2 )
which coincides with the inner horizon.

The physical quantities of non-extremal AdS4 black holes can also be solved as functions
of (a, m, δ1 , δ2 ). In particular, the gravitational angular velocity Ω and the temperature
T are given by
a(1 + g 2 r1 r2 ) ∆′r
Ω= , T = . (4.8)
r1 r2 + a 2 4π(r1 r2 + a2 )
Moreover, the other thermodynamic quantities of asymptotically AdS 4 black holes are [29]
π(r1 r2 + a2 )
S= ,
Ξ
ma
J= (cosh(2δ1 ) + cosh(2δ2 )) ,
2Ξ2 (4.9)
m
Q1 = Q2 = sinh(2δ1 ) ,
4Ξ
m
Q3 = Q4 = sinh(2δ2 ) .
4Ξ
The 4d N = 4 gauged supergravity can be obtained by the truncation of the 11d
supergravity, it admits supersymmetric black holes with two scalar and two Maxwell fields [28].
A microscopic foundation for the entropy of these black holes was provided in [33, 34] via
the AdS/CFT correspondence. Inspired by the successful studies of logarithmic in area
corrections to the Bekenstein-Hawking entropy of asymptotically flat black holes pioneered
– 16 –
by Sen (a different one-loop contribution ∼ log(S0 ) [35–37], the question of such logarithmic
corrections for AdS black holes was addressed in [38–41]. In this case the treatment of zero
modes is more subtle that in asymptotically flat black holes. In the spherically symmetric
case there are two types of zero modes. There is a two-form zero mode in AdS4 , however,
in the AdS2 × S 2 near-horizon geometry there are “other zero modes” (graviton, one-form,
gravitino). Performing the one-loop computation in the near-horizon geometry [38, 39] does
not agree with the field theory answer. This circumstance has been crucial in successfully
matching the logarithmic corrections to the entropy of AdS4 × S 7 black holes to computations
on the field theory side (ABJM) [40] and various other AdS4 /CFT3 pairs [41–43].
A systematic, bottom-up treatment of logarithmic area corrections (∼ log S0 ) to AdS4
JHEP06(2024)034
black holes was presented in [19] including for minimal gauged supergravity. An important
conclusion relevant for extremal black holes was to demonstrate that the local contribution
to the one-loop effective action is the same whether it is computed from the full geometry or
from the near-horizon geometry; only the contribution of zero modes is different from the
full geometry and the near-horizon region. We hope to revisit the interplay of these log(S0 )
versus log(THawking ) in the supersymmetric setups elsewhere.
To obtain the near-horizon and low-temperature geometry, we first choose an ensemble.

As we expect the one-loop correction to the low-temperature entropy does not depend on
the ensemble choice, we choose the ensemble for calculation convenience. In this ensemble,
the relation (4.6) is always kept. The parameter a and the radius of the outer horizon r+
depend on the temperature,
2
a(T ) = + c1 T + c2 T 2 + c3 T 3 + O(T 4 ), (4.10)
− 1)
g(e2δ1 +2δ2
2m sinh(δ1 ) sinh(δ2 )
r+ (T ) = + d1 T + d2 T 2 + d3 T 3 + O(T 4 ), (4.11)
cosh(δ1 + δ2 )
where the coefficients ci , di can be obtained by solving the second equation of (4.8) and ∆r = 0.
We verify that c1 = 0. Other coefficients are lengthy and not illuminating to be written here.
−iτ −ik3 τ
r = r+ (T ) + T k1 (y − 1), t = , ϕ=φ+ + k4 iτ, (4.12)
k2 T T
where the precise expressions for the constants ki are as follows:
k2 = 2π, (4.13)
(−3 + e2δ1 +2δ2 )g
k3 = , (4.14)
(−1 + e2δ1 +2δ2 )k2
k1 = (4e2(δ1 +δ2 ) (−1 + e4δ1 )(−1 + e4δ2 )(1 + e2(δ1 +δ2 ) )π)/(−2e4δ1 − 2e4δ2
− 3e2(δ1 +δ2 ) + 14e4(δ1 +δ2 ) − 10e6(δ1 +δ2 ) + 6e8(δ1 +δ2 ) + e10(δ1 +δ2 ) + 5e6δ1 +2δ2
− 8e8δ1 +4δ2 + 5e2δ1 +6δ2 + e10δ1 +6δ2 − 8e4δ1 +8δ2 + e6δ1 +10δ2 g 2 ), (4.15)
– 17 –
k4 = (−1 + e4δ1 )(e2δ1 + e2δ2 )(−1 + e4δ2 )(−3 + e2(δ1 +δ2 ) )(−1 + e2(δ1 +δ2 ) )3
q
e−δ1 −δ2 coth2 (δ1 + δ2 ) csch(2δ1 ) csch(2δ2 ) csch(δ1 + δ2 )/(2(1 + e2(δ1 +δ2 ) ) (4.16)
(−2e4δ1 − 2e4δ2 − 3e2(δ1 +δ2 ) + 14e4(δ1 +δ2 ) − 10e6(δ1 +δ2 ) + 6e8(δ1 +δ2 ) + e10(δ1 +δ2 )
+ 5e6δ1 +2δ2 − 8e8δ1 +4δ2 + 5e2δ1 +6δ2 + e10δ1 +6δ2 − 8e4δ1 +8δ2 + e6δ1 +10δ2 )). (4.17)
By taking the T → 0 limit, we obtain the near-horizon geometry on the zeroth order of T :
!
dy 2 2
ḡµν dx dx = g1 (θ) 2
µ ν
+dτ 2 y 2 −1 +g2 (θ)dθ2 +g3 (θ) dφ+ik(y−1)dτ , (4.18)
y −1
JHEP06(2024)034
where
e2(δ1 +δ2 ) + cos(2θ)
g1 (θ) = k1 ,
2 1 + e2(δ1 +δ2 ) π

2 e2(δ1 +δ2 ) + cos(2θ)
g2 (θ) = − ,
g 2 1 + 3e2(δ1 +δ2 ) − 3e4(δ1 +δ2 ) + e6(δ1 +δ2 ) + 2 cos(2θ) (4.19)

2 −1 − 2e2(δ1 +δ2 ) + e4(δ1 +δ2 ) − 2 cos(2θ) sin2 θ
g3 (θ) = 2 ,
−3 + e2(δ1 +δ2 ) g 2 e2(δ1 +δ2 ) + cos(2θ)

k = − k4 .
We can also obtain the near-horizon gauge fields and scalar fields, Ā1 , Ā2 , ζ̄, χ̄, by taking the
same limit. By keeping O(T ) when T → 0, we can obtain the first order correction to the
above fields, δg, δF1 , δF2 , δζ, δχ. Again, these expressions are lengthy and not illuminating
to be provided here.

The Lichnerowicz operator of our action (4.1) is
1 2ζ αβ,µν eζ 1 1

hαβ ∆αβ,µν hµν = hαβ ∆αβ,µν − e ∆ − ∆αβ,µν − ∆αβ,µν − e2ζ ∆αβ,µν
L EH
2 F2
2 (1 + χ e )
2 2ζ F1
2 ζ
2 χ

+g 2 (4 + 2 cosh ζ + eζ χ2 )∆αβ,µν
1 hµν , (4.20)
where the different contributions to the Lichnerowicz operator can be found in C.

At zero temperature, i.e. when g = ḡ, A1,2 = Ā1,2 , ζ = ζ̄, χ = χ̄, one can verify that
the above operator has the following zero modes:
|n|
−1 + y
!
|n| dτ dy dy 2

2
h(n) µ
µν dx dx
ν
= cn e inτ
e 2(δ1 +δ2 )
+ cos(2θ) dτ − 2i
2
−
1+y n y 2 − 1 (y 2 − 1)2
(4.21)
for |n| ≥ 2, where cn is the normalization constant.
eigenvalue. Now we substitute g = ḡ + δg, A1,2 = Ā1,2 + δA1,2 , ζ = ζ̄ + δζ, χ = χ̄ + δχ into the
Lichnerowicz operator (4.20) and apply it to the zero modes (4.21). The lifted eigenvalue is
Z
(n)∗
p
L µν .
– 18 –
the result is simple:
δλn ∝ nT, |n| ≥ 2. (4.23)
Using Zeta function regularization, the contribution of the extremal zero modes to the
low-temperature partition function is therefore:
 
3
!
Y π T
δ log Z = log   = log + O(1), (4.24)
n≥2
δλn 2 Tq
JHEP06(2024)034
where the O(1) term is not reliable. The temperature scale Tq can be extracted from the
O(T ) coefficient of classical entropy, its precise form is

Tq = 2 −3 + e2(δ1 +δ2 ) 1 + e2(δ1 +δ2 ) (−2e4δ1 − 2e4δ2 − 3e2(δ1 +δ2 ) + 14e4(δ1 +δ2 )
−10e6(δ1 +δ2 ) + 6e8(δ1 +δ2 ) + e10(δ1 +δ2 ) + 5e6δ1 +2δ2 − 8e8δ1 +4δ2 + 5e2δ1 +6δ2

+e 10δ1 +6δ2
− 8e 4δ1 +8δ2
+e 6δ1 +10δ2
)g 3
/ 4π 2 −1 + e4δ1 e2δ1 + e2δ2 −1 + e4δ2
q
(−1 + e2(δ1 +δ2 ) )4 e−δ1 −δ2 coth2 (δ1 + δ2 ) csch 2δ1 csch 2δ2 csch (δ1 + δ2 ) . (4.25)
This quantity is extracted from the classical heat capacity and agrees (in the δ1 = δ2 = δ ̸= 0
limit) with the expression presented in [44, 45].
It is worth emphasizing that our investigation in this paper focuses strictly in the
contribution of the graviton zero modes to the path integral. A bona fide discussion of the
black hole in the supersymmetric BPS limit, requires a proper treatment of the gravitino
and other zero modes which is expected to changed the resulting logarithmic in temperature
contribution. Here we want to highlight only the universality of the graviton contribution
and that is the main purpose that this section serves. A systematic treatment, including the
effects of fermionic and vector zero modes will be presented elsewhere.
5 Myers-Perry black holes in five dimensions
In this section, we study a vacuum solution of the Einstein’s equations that describes spinning
black holes, known as the Myers-Perry (MP) black holes. Some aspects of the connection of
five-dimensional rotating black holes with JT gravity were discussed in [46–48] with emphasis
on classical aspects. Our goal in this manuscript is to focus on the quantum contribution
to the low-temperature thermodynamics.
The MP black hole [49, 50] in general odd d (≥ 5) dimensions with d = 2n + 1 is
described by the metric
n
!2 n
µr2 ΠF
ds = −dt + dt + + + + + (5.1)
X X
2 2
ai µ2i dϕi dr 2
r 2
a 2
dµ 2
µ 2
dϕ 2
i ,
ΠF i=1
Π − µr2 i=1
i i i
where the parameters in the metric are given by

n n
a2i µ2i
F =1− Π= r2 + a2i , (5.2)
X Y
,
i=1
r2 + a2i i=1
– 19 –
where µ is a free parameter in the theory proportional to the mass of the black hole, ai ’s
are related to the angular velocity of the rotating black hole and {µi } satisfy the constraint
i=1 µi = 1. The conserved charges of this class of black holes are given by
Pn 2
(d − 2) Ωd−2 2
M= µ, J yi xi = M ai , (5.3)
16π d−2
where Ωd is the volume of the d sphere. Naturally, setting ai to zero takes us to the
Schwarzschild metric whereas setting M to zero gives us the flat space vacuum limit.
In this section, we focus only on the d = 5 case, and the metric is obtained by setting
n = 2 in equations (5.1) and (5.2). Setting a1 = a, a2 = b and parametrizing µi as
JHEP06(2024)034
µ1 = sin θ , µ2 = cos θ, we obtain the metric as
µ Σ
2
ds2(5d) = −dt2 + dt + a sin2 θdϕ1 + b cos2 θdϕ2 + r2 dr2 +
Σ Π − µr2

Σdθ2 + r2 + a2 sin2 θdϕ21 + r2 + b2 cos2 θdϕ22 , (5.4)
where
Σ = r2 + a2 cos2 θ + b2 sin2 θ , Π = r 2 + a2 r 2 + b2 . (5.5)
To calculate log T corrections, as before, we need to move slightly away from extremality,
thereby leading to a perturbative series for the metric, with the perturbing parameter being
the temperature T associated with the outer horizon. Below, we proceed to calculate the
near-horizon limit of 5d Myers-Perry black hole.
Equal angular momenta. For ease of computation, we first focus on the case of equal
angular momenta, i.e., we set a = b. In the following we use 2πT in place of the temperature.
We see that the horizon is obtained as a root of the equation
Π − µr2 = 0 (5.6)
and this immediately gives us the inner (r− (T )) and outer horizon (r+ (T )) as
1
r q
r− (T ) = √ µ − µ (µ − 4a2 ) − 2a2 , (5.7)
2
1
r q
r+ (T ) = √ µ + µ (µ − 4a2 ) − 2a2 . (5.8)
2
The temperature itself is given by
∂r Π − 2µr 2µ (µ − 4a2 )
p
T = = qp . (5.9)
2µr2 µ µ (µ − 4a2 ) − 2a2 + µ
r=r+
This induces a T correction to the mass and angular velocity, and they are given by
4J 2/3 4J 4/3 T 2 8J 5/3 T 3

µ= + + + ... , (5.10)
π 2/3 3π 4/3 3π 5/3
J 1/3 JT 2 2J 4/3 T 3
a = 1/3 − − + ... . (5.11)
π 3π 3π 4/3
– 20 –
Note that the corrections are such that
Ω3
J= µa (5.12)
8π
is actually kept fixed at its extremal value, i.e., it receives no T corrections. This ensures
that we are manifestly working in the canonical ensemble where the angular momenta are
kept fixed. Substituting back in equations (5.7) and (5.8), we obtain the following series
for the inner and outer horizon radii
1 2
J 3 J 3 5J 2
r− (T ) = − T− T + ... , (5.13)
π π 6π
JHEP06(2024)034
1 2
J 3 J 3 7J 2
r+ (T ) = + T+ T + ... . (5.14)
π π 6π
5.1 Near horizon limit of the MP black hole
In this section we do a coordinate transformation to probe the near horizon geometry of
the MP black hole when it is slightly out of extremality.
Following [51, 52], we probe the near horizon geometry of 5d MP black hole as follows.
We do a coordinate transformation given by
iτ 1

t→− , r → r+ (T ) + r02 T (y − 1) , ϕ1,2 → φ1,2 + i − 1 τ. (5.15)
T 2r0 T
Here, r0 is the radius of the outer horizon at extremality which is related to the angular
momentum as
1
J 3
r0 = . (5.16)
π
This transformation yields the near horizon metric to be
2 2
2J 3
!
J3dy 2 2 2
O (1) : ds2(0) = ḡµν dx dx = 2
µ ν
2
+ y −1 dτ + 2 dθ2
2π 3 y −1 π3
2
2J 2/3 sin2 θ((y−1) cos 2θ+3y−5) i (y−2)

+ d1 + dτ
π 2/3 (y−2) 2
2
2J 2/3 cos2 θ ((y−1) cos 2θ+5y−7) i (y−2)

+ dφ2 + dτ
π 2/3 (3y−4) 2
2J 2/3 (y−2) sin2 θ cos2 θ 3y−4 2

− i (y−1) dτ 2 + dφ1 +dφ22 . (5.17)
π 2/3 (3y−4) y−2
We also obtain the corresponding correction to the near-horizon extreme Myers-Perry

(NHEMP) metric given by
O (T ) : ds2(1) = T gµν
(1)
dxµ dxν , (5.18)
where, as in section 2, for the sake of presentation, we do not explicitly type the corrections
here. With the corrected metric in hand, we now proceed to evaluate the zero modes of
the Lichnerowicz operator.
– 21 –
5.2 Zero modes and correction to black hole entropy
We begin by choosing an ansatz for the diffeomorphism generating vector field given by
ξ (n) = einτ (f2 (y) ∂τ + f1 (y) ∂y + f3 (y) ∂φ1 + f4 (y) ∂φ2 ) . (5.19)
With this ansatz, we construct the zero modes of the Lichnerowicz operator given by hµν =
Lξ ḡµν . We further demand that the hµν so generated belong to the kernel of the Lichnerowicz
operator. This fixes the functions fi (y) to be
|n|
JHEP06(2024)034
y−1 |n| + y

2
f1 (y) = , (5.20)
y+1 2 (n2 − 1)
|n|−2
y−1 2 + |n| + n2 (y − 2) − 2|n|y − 2y 2

2
f2 (y) = i , (5.21)
y+1 4|n| (n2 − 1) (y + 1)2
|n|−2
y−1 2 + |n| + n2 (y − 2) − 2|n|y − 2y 2

2
f3 (y) = , (5.22)
y+1 4|n| (n2 − 1) (y + 1)2
|n|−2
y−1 2 + |n| + n2 (y − 2) − 2|n|y − 2y 2

2
f4 (y) = = f3 (y) , (5.23)
y+1 4|n| (n2 − 1) (y + 1)2
and
|n|
|n| (n2 −1)einτ
!
dy 2
p
y−1 |n| dτ dy

2
h(n) µ
µν dx dx
ν
= −dτ +2i 2
+ 2 , (5.24)
8π 4/3 J 1/6 y+1 n y −1 (y −1)2
2
for |n| ≥ 2. Hereafter, the steps to be performed mimic section 2.3 and the corrected
eigenvalue of the Lichnerowicz operator at first order is given by
Z p αβ,µν
δλn = d4 x ḡ h(n) δ∆Λ h(n) . (5.25)
αβ µν
Explicit computation of (5.25) gives us
nπ 4/3 T
δλn = , (5.26)
8J 1/3
where we reverted back to the actual temperature T of the black hole by appropriately
inserting factors of 2π (see statement before equation (5.6)).
Thereafter, following arguments similar to the one in section 2.3 and the Zeta function
regularization, we have
" ∞ 4
3
!#
nπ 3 T
log Z = − log log T , (5.27)
Y
∼
n=2 8J
1
3 2
log T
where once again we have chosen to leave out O(1) constant pieces which are irrelevant
at the moment.
– 22 –
5.3 Comment on the case of two different angular momenta
Here, we summarize a few statements about the 5d Myers Perry black with different angular
momenta (J1 ̸= J2 ). The calculation proceeds exactly as we describe in the previous section
with added computational complexity. We find that the outer horizon now admits the
small-temperature expansion
√ √
22/3 J1 J2 (J1 +J2 )2/3 9J12 +10J1 J2 +9J22 J1 J2 2
r+ (T ) = 1/3 + T+ T +. . . . (5.28)
π (J1 +J2 )2/3 (2π)2/3 24πJ1 J2
As before, we start from the metric (5.4), and do the following coordinate transformations
√
JHEP06(2024)034
iτ 2 J1 J2 1

t → − , r → r+ (T )+r0 T (y−1) , ϕ1,2 → φ1,2 +
2
−1 iτ, (5.29)
T J1 +J2 2r0 T
to zoom into the near-horizon region. Upto linear order in temperature, we obtain the
near horizon extreme geometry
ds2(0) = ḡµν dxµ dxν = ḡ11 (y, θ) dτ 2 + 2ḡ14 (y, θ) dτ dϕ1 + 2ḡ15 (y, θ) dτ dϕ2 + ḡ22 (y, θ) dr2
+ ḡ33 (θ) dθ12 + ḡ44 (θ) dϕ21 + 2ḡ45 (θ) dϕ1 dϕ2 ,
(5.30)
and its first-order temperature correction (which we do not write explicitly)
ds2(1) = T gµν
(1)
dxµ dxν . (5.31)
Here the functions entering the near-horizon extreme metric (5.30) are given by

ḡ11 (y, θ) = − (J1 + J2 + (J1 − J2 ) cos 2θ)2 3J16 + 22J15 J2 − 192J14 J22 y + 253J14 J22 +
128J13 J23 y 2 − 640J13 J23 y + 596J13 J23 − 192J12 J24 y + 253J12 J24 +

4 cos 2θ (J1 + J2 )3 J13 − 15J12 J2 + 15J1 J22 − J23 + 22J1 J25 +

cos 4θ (J1 − J2 )2 (J1 + J2 )4 + 3J26 32 (2π)2/3 (J1 + J2 )4/3 ×
3 −1
J12 cos2 θ + J2 J1 + J2 sin2 θ , (5.32)

ḡ14 (y, θ) = 21/3 i sin2 θ J1 J2 − 5J13 + cos 2θ (J1 − J2 ) J12 + 2J1 J2 (2y − 1) + J22 +
p

3J12 J2 (4y − 7) + J1 J22 (4y − 7) + J23 π 2/3 (J1 + J2 )7/3 ×
−1
(cos 2θ (J1 − J2 ) + J1 + J2 ) , (5.33)

ḡ15 (y, θ) = 21/3 i sin2 θ J1 J2 − 5J23 + cos 2θ (J1 − J2 ) J12 + 2J1 J2 (2y − 1) + J22 +
p

3J22 J1 (4y − 7) + J2 J12 (4y − 7) + J13 π 2/3 (J1 + J2 )7/3 ×
−1
(cos 2θ (J1 − J2 ) + J1 + J2 ) , (5.34)
J1 J2 (J1 + J2 + (J1 − J2 ) cos 2θ)

ḡ22 (y, θ) = , (5.35)
(2π)2/3 (y − 1) (J1 + J2 )1/3 J12 + 2yJ1 J2 + J22
– 23 –
21/3 (J1 + J2 + (J1 − J2 ) cos 2θ)
ḡ33 (θ) = , (5.36)
π 2/3 (J1 + J2 )1/3
21/3 J1 sin2 θ(J1 + J2 )2/3 (2J1 − J2 cos 2θ + J2 )

ḡ44 (θ) = , (5.37)
π 2/3 J12 cos2 (θ) + J2 J1 + J2 sin2 θ

21/3 J2 cos2 (θ)(J1 + J2 )2/3 (J1 cos 2θ + J1 + 2J2 )

ḡ55 (θ) = , (5.38)
π 2/3 J12 cos2 θ + J2 J1 + J2 sin2 θ

J1 J2 sin2 2θ(J1 + J2 )2/3

ḡ45 (θ) = . (5.39)
(2π)2/3 J12 cos2 θ + J2 J1 + J2 sin2 θ
JHEP06(2024)034
The diffeomorphism generating vector fields are given by
ξ = einτ (f2 (y) ∂τ + f1 (y) ∂y + f3 (y) ∂φ1 + f4 (y) ∂φ2 ) , (5.40)
where
|n|
1 Θ(J1 , J2 ; y)−1

2
f1 = (|n|+Θ(J1 , J2 ; y)) , (5.41)
2 (n2 −1) Θ(J1 , J2 ; y)+1

(1,0) (2,0)
i f1′ (y)ḡ15 (y, θ)+f1 (y)ḡ15 (y, θ)
f2 = (1,0)
, (5.42)
|n| ḡ15 (y, θ)

(1,0) (1,0)
i f1 ḡ45 (θ)ḡ15 (y, θ)−f1 ḡ55 (θ)ḡ14 (y, θ) +|n| f2 ḡ55 (θ)ḡ14 (y, θ)−|n| f2 ḡ45 (θ)ḡ15 (y, θ)
f3 = ,
|n| (ḡ45 (θ)2 −ḡ44 (θ)ḡ55 (θ))
(5.43)
(1,0)
if1 ḡ15 (y, θ)−|n| f2 ḡ15 (y, θ)−|n| f3 ḡ45 (θ)
f4 = , (5.44)
|n| ḡ55 (θ)
with
J 2 + 2J1 J2 (2y − 1) + J22

Θ(J1 , J2 ; y) = 1 . (5.45)
(J1 + J2 )2
It can be checked that these diffeomorphism generating vectors are harmonic, i.e.,
¯ ξµ = 0
□ (5.46)
and, therefore, they satisfy the harmonic gauge fixing condition. The zero modes of the
Lichnerowicz operator are given by
1 n
|n|
J1 J2 (y−1)
p
|n| (n2
− 1) 2 3 + 2 einτ J12 +J22 +2yJ1 J2
2
(J1 + J2 + (J1 − J2 ) cos 2θ)
h(n)
µν = × ×
8(2π ) (J1 J2 )1/4
2 2/3
(J1 + J2 ) 3
2
 
2 J1 J2 (y−1)
|n| (J1 + J2 ) J12 +J22 +2yJ1 J2 (J1 + J2 ) 4
−dτ + i dτ dy +
2 2
dy  ,

2 2
n J1 J2 (y − 1) 4 (y − 1) J1 + J2 + 2yJ1 J2
2 2
(5.47)
which are generated by hµν = Lξ ḡµν , as usual. The zero modes also reduce to equation (5.17)
when J1,2 → J. The integration over these zero modes (as in equation (2.46)) is quite
– 24 –
intractable in this general case of different angular momenta. However, we see the n
dependence of hµν above is comparable with the equal angular momenta case. This similarity
owes its roots to the similar functional form1 of the diffemorphism generating vector field in
equation (5.41) (to be compared with the corresponding one in equation (5.24)). Therefore,
the corresponding log correction is anticipated to be
3
log Z ∼ log T , (5.48)
2
log T
where J1,2 simply alter the functional form of J1,2 dependent constants.
JHEP06(2024)034
6 Kerr-AdS5 black hole
In this section, we provide the story for the Kerr-AdS5 black hole with equal angular momenta
We start from the Einstein-Hilbert action with cosmology constant in 5 dimensions:
1 √
Z
I=− d5 x g(R − 2Λ) + Iboundary , (6.1)
16π
where Λ = −6g 2 . The Kerr-AdS5 solution is given by
∆θ1 + g 2 r2 ρ2 dt dt f ∆θ dt
2
ρ2 dr2 ρ2 dθ2

ds = −
2
+ 4 −ω + +
Ξa Ξb ρ2 ρ Ξa Ξb ∆r ∆θ
(6.2)
r 2 + a2 r 2 + b2
+ sin2 θdϕ21 + cos2 θdϕ22 ,
Ξa Ξb
where
ν ≡ b sin2 θdϕ1 + a cos2 θdϕ2 ,
dϕ1 dϕ2
ω ≡ a sin2 θ + b cos2 θ ,
Ξa Ξb
∆θ ≡ 1 − a2 g 2 cos2 θ − b2 g 2 sin2 θ,
r 2 + a2r 2 + b2 1 + g 2 r 2

∆r ≡ − 2m, (6.3)
r2
ρ2 ≡ r2 + a2 cos2 θ + b2 sin2 θ,
Ξa ≡ 1 − a2 g 2 ,
Ξ b ≡ 1 − b2 g 2 ,
f ≡ 2mρ2 .
For the convenience of our following discussion, we rewrite the above solution (6.2) in the
following form [18]:
∆r ∆θ r2 sin2 (2θ) 2 2 dr2 dθ2
!
ds2 = − dt +ρ + +Bϕ2 (dϕ2 +v1 dϕ1 +v2 dt)2 +Bϕ1 (dϕ1 +v3 dt)2 ,
4Ξa Ξb Bϕ1 Bϕ2
2 2 ∆ r ∆θ
(6.4)
where
2
g33 g44 − g34
Bϕ1 ≡ , Bϕ2 ≡ g44 ,
g44
(6.5)
g34 g04 g04 g34 − g03 g44
v1 ≡ , v2 ≡ , v3 ≡ 2 −g g ,
g44 g44 g34 33 44
1
Note that Θ(J, J; y) = y.
– 25 –
with the non-vanishing components of the metric (6.2)
∆θ 1 + g 2 r 2 ∆2 2mρ2

g00 =− + θ4 2 2 ,
Ξa Ξb ρ Ξa Ξb
∆θ a 2mρ2 sin2 θ

g03 = g30 =− ,
ρ4 Ξ2a Ξb
∆θ b 2mρ2 cos2 θ

g04 = g40 =− ,
ρ4 Ξ2b Ξa
ρ2 ρ2
g11 = , g22 = , (6.6)
∆r ∆θ
JHEP06(2024)034
r2 + a2 sin2 θ a2 2mρ2 sin4 θ

g33 = + ,
Ξa ρ4 Ξ2a
r2 + b2 cos2 θ b2 2mρ2 cos4 θ

g44 = + ,
Ξb ρ4 Ξ2b
ab 2mρ2 sin2 θ cos2 θ

g34 = g43 = .
ρ4 Ξ a Ξ b
The temperature is given by:
r4 1 + g 2 2r+2 + a2 + b2 − (ab)2

T = + . (6.7)
2πr+ r+ + a2 r+ + b2
2 2
For simplicity, we consider the a = b case first in the rest parts of this section.

Under the extreme condition, we have
3
m0 = 2r02 1 + g 2 r02 , (6.8)
q
a0 = r02 + 2g 2 r04 , (6.9)
where r0 is the radius of the extreme black hole horizon. In our ensemble, we choose a to be
a constant. m and r+ , the radius of the outer horizon, depend on T :
m(T ) = m0 + c1 T + c2 T 2 + c3 T 3 + O(T 4 ), (6.10)
r+ (T ) = r0 + d1 T + d2 T 2 + d3 T 3 + O(T 4 ), (6.11)
where the coefficients ci , di can be obtained by solving (6.7) and ∆r = 0. We can see c1 = 0.
Other coefficients are lengthy and not illuminating to be written here. The Hawking entropy
can be calculated using the above temperature dependence:
A 2π 2 r03 4π 3 r04 (1 − g 2 r02 )

S= = + T + O(T 2 ). (6.12)
4 (1 − 2g 2 r0 )2 1 − 12g 4 r04 + 16g 6 r06
2
−iτ −ik3 τ −ik5 τ
r = r+ (T )+T k1 (y−1), t = , ϕ 1 = φ1 + +k4 iτ, ϕ2 = φ2 + +k6 iτ, (6.13)
k2 T T T
– 26 –
where
2πr02 (1 + g 2 r02 )
k1 = ,
1 + 4g 2 r02
k2 = 2π,
q
r02 + 2g 2 r04
k3 = ,
4πr02 (6.14)
q 2
− −1 + 2g 2 r02 r0 + 2g 2 r04
k4 = ,
2 r0 + 4g 2 r0 3

k5 = k3 ,
JHEP06(2024)034
k6 = k4 .
!
dy 2
ḡµν dx dx = g1
µ ν
+dτ 2
y 2
−1 +g2 dθ 2
+g ij (θ) dx i
+ik i (y−1)dτ dx j
+ik j (y−1)dτ ,
y 2 −1
(6.15)
where
r02
g1 =
2 1 + 4g 2 r02

2r02
g2 =
1 − 2g 2 r02
r02 −3 + 2g 2 r02 + 1 + 2g 2 r02 cos(2θ) sin2 (θ)

gφ1 φ1 (θ) = − 2
1 − 2g 2 r02
2r2 1 + 2g 2 r02 cos2 (θ) sin2 (θ)
(6.16)
gφ1 φ2 (θ) = 0 2
1 − 2g 2 r02
r02 cos2 (θ) 3 − 2g 2 r02 + 1 + 2g 2 r02 cos(2θ)

gφ2 φ2 (θ) = 2
1 − 2g 2 r02
q 2
−1 + 2g 2 r02 r0 + 2g 2 r04
k φ1 = k φ2 =
2 r0 + 4g 2 r03

By keeping the first order of T , we can also obtain δgµν , the near-horizon geometry on O(T ).
Again, this expression is lengthy and not illuminating to be written here.
The Lichnerowicz operator of our Einstein-Hilbert action with cosmology constant is as follows:
1 1 αµ βν 1 1
hαβ ∆αβ,µν hµν = − hαβ g g □− g αβ g µν □+Rαµβν +Rαµ g βν −Rαβ g µν − Rg αµ g βν
L
16π 2 4 2
1 1
!
+ Rg αβ g µν +Λg αµ g βν − Λg αβ g µν hµν . (6.17)
4 2
– 27 –
At zero temperature when g = ḡ, one can verify that the above operator has the following
zero modes:
|n|
−1 + y
!
|n| dτ dy dy 2

2
h(n) µ
µν dx dx
ν
= cn e inτ
dτ − 2i
2
− |n| ≥ 2, (6.18)
1+y n y 2 − 1 (y 2 − 1)2
where cn is the normalization constant.
6.3 The lifted eigenvalue and entropy correction

eigenvalue. Now we substitute g = ḡ + δg into the Lichnerowicz operator (6.17) and apply
JHEP06(2024)034
it to the zero modes (6.18). The lifted eigenvalue is
Z
(n)∗
p
L µν .
the result is simple
1 − g 2 r02
δλn = nT , |n| ≥ 2. (6.20)
8r02
is therefore:
 
3
!
Y π T
δ log Z = log   = log + O(1). (6.21)
λ
n≥2 n
2 Tq
Again, the O(1) term is not reliable, and Tq can be extracted from the O(T ) term of the
classical entropy (6.12):
1 − 12g 4 r04 + 16g 6 r06
Tq = . (6.22)
4π 3 r04 (1 − g 2 r02 )
6.4 Results for different angular momenta
For completeness, we also do the calculation for different angular momenta. Central results
are presented in this subsection. The Kerr-AdS black hole in five dimensions is naturally
parameterized by (a, b, m). r0 and m0 depend on a, b. However, for convenience, we use
(r0 , b, m) instead of (a, b, m) as parameters in our calculation. The zeroth order near-horizon
metric is given by:
!
dy 2
ḡµν dx dx = g1
µ ν
+dτ 2
y 2
−1 +g2 dθ 2
+g ij (θ) dx i
+ik i (y−1)dτ dx j
+ik j (y−1)dτ ,
y 2 −1
(6.23)
where the coefficients are
(b2 + r02 ) −b2 − r02 + (b2 − r02 − 2g 2 r04 ) cos(2θ)

g1 = − , (6.24)
8 b4 g 2 − g 4 r06 + b2 (1 + 3g 2 r02 )

(b2 + r02 ) −b2 − r02 + (b2 − r02 − 2g 2 r04 ) cos(2θ)

g2 = , (6.25)
2 g 2 r04 (1 + b2 g 2 + 2g 2 r02 ) cos2 θ + (b2 − g 2 r04 )(−1 + b2 g 2 sin2 θ)
– 28 –
q q
2br02 (b2 + r02 )(1 + g 2 r02 ) 1 + b2 g 2 + 2g 2 r02 b2 − g 2 r04 cos2 θ sin2 θ
gφ1 φ2 (θ) = ,
(−1 + b2 g 2 )(2g 2 r04 + b2 (−1 + g 2 r02 ))(b2 + r02 + (−b2 + r02 + 2g 2 r04 ) cos(2θ))
(6.26)

gφ1 φ1 (θ) = − r02 (b2 + r02 )(−b4 + g 2 r06 (3 + 2g 2 r02 ) + b2 r02 (−2 + g 2 r02 + g 4 r04 ) + (b4 + g 2 r06
(6.27)
2
+ 2g 4 r08 + b2 g 2 r04 (−1 + g 2 r02 )) cos(2θ)) sin2 θ / 2g 2 r04 + b2 (−1 + g 2 r02 )
(6.28)
JHEP06(2024)034

b2 + r02 + (−b2 + r02 + 2g 2 r04 ) cos(2θ) , (6.29)
q q
2br02 (b2 + r02 )(1 + g 2 r02 ) 1 + b2 g 2 + 2g 2 r02 b2 − g 2 r04 cos2 θ sin2 θ
gφ2 φ2 (θ) = ,
(−1 + b2 g 2 )(2g 2 r04 + b2 (−1 + g 2 r02 ))(b2 + r02 + (−b2 + r02 + 2g 2 r04 ) cos(2θ))
(6.30)
q
1 + b2 g 2 + 2g 2 r02 2g 4 r08 + b4 (1 − g 2 r02 ) + b2 g 2 r04 (−3 + g 2 r02 )

k φ1 = − q , (6.31)
2r0 b2 − g 2 r04 b4 g 2 − g 4 r06 + b2 (1 + 3g 2 r02 )

b(−1 + b2 g 2 )r0 (1 + g 2 r02 )

k φ2 = . (6.32)
2 b4 g 2 − g 4 r06 + b2 (1 + 3g 2 r02 )
The metric (6.23) is written in the generic form shown to admit a SL(2, R) subgroup of
isometries via a lemma in [53] which we quote explicitly in appendix B.
The zero modes are given by
−1 + y |n|
2
µν dx dx = cn e b2 − r02 + b2 − r02 − 2g 2 r04 cos(2θ)
µ ν inτ
h(n)
1+y
!
|n| dτ dy dy 2
dτ 2 − 2i − , |n| ≥ 2. (6.33)
n y 2 − 1 (y 2 − 1)2
As final result we obtain

3
!
T
δ log Z = log + O(1). (6.34)
2 Tq
6.5 Comments on the rotating black hole in minimal gauged supergravity

The rotating, electrically charged black hole in minimal gauged supergravity in five dimensions
plays an important role in the context of the AdS/CFT correspondence. Importantly, there
is an explanation for its Bekenstein-Hawking entropy in terms of the superconformal index
of N = 4 SYM [54–56]. Some aspects of the logarithmic corrections to the entropy (log S0 )
have been discussed on the field theory side [57].Some of these logarithmic corrections to
the entropy have been matched by using generalizations of the Kerr/CFT correspondence at
one-loop level in [58]. We have performed some preliminary explorations of the logarithmic
– 29 –
in temperature corrections to the entropy in these models. The results are consistent with
other models studied in this manuscript but will be reported elsewhere with a discussions of
all the other zero modes arising from the gravitino and other supergravity fields.
3
7 On the origin of the universal 2
log T behavior
Having performed a number of explicit computations for various spacetimes in dimensions

four and five, we are ready to highlight aspects of the universality of the logarithmic in
temperature tensor modes contributions to the thermodynamics of generic near-extremal black
holes. The (mathematical) origin of 32 factor of 32 log THawking behavior can be understood as
JHEP06(2024)034
a consequence of the following universal facts and general arguments. They are elucidated
sequentially as follows.
1. Extremal black holes in their near horizon geometry always have an AdS2 factor,
possibly fibered over some compact directions. This is explicitly seen in (2.22) for the
Kerr-AdS case and in (3.14) for the Kerr-Newman AdS, similarly, it can be seen in
the five-dimensional cases we considered (5.17) and (6.15). This behavior is expected
following general arguments given in [24, 36, 53, 59–61], in the particular case of Kerr
black holes, the explicit form of the near-horizon isometry group was first explicitly
presented in [24]. This implies that there is a near horizon enhancement in the isometry
group to G = SL(2, R) × Gcompact where Gcompact depends on the black hole in question,
whose near horizon we are interested in. In what follows, the exact nature of Gcompact
will play no role and we focus on the fact that there exists an unbroken SL(2, R) at the
near horizon.
2. Technically, the log T corrections arise as corrections to the eigenspectrum of the

Lichnerowicz operator. This follows from perturbation theory. Namely, imagine that
¯ evaluated at the
we know the eigenspectrum {hn , λ0n } of the Lichnerowicz operator ∆
NHEK-AdS. This gives
∆¯ hn = λ 0 hn . (7.1)
n
Now we move slightly away from extremality, i.e., we turn on a small temperature T .
This induces a change in the metric ḡ → g = ḡ + T δg + O(T 2 ). This in turn induces
a change in the Lichnerowicz operator ∆ ¯ →∆=∆ ¯ + δ∆(T ) and the eigenspectrum
{hn + δhn (T ), λ0n + δλn (T )}. First order perturbation theory then tells us that the
corrected eigenvalue is the expectation value of the corrected operator. That is,
Z p
δλn (T ) = d4 x ḡ h(n) (δ∆)αβ,µν h(n) . (7.2)
αβ µν
Since it is a Gaussian integral over bosonic fields, we have, ignoring an overall factor
which play little role
1
log Z ∼ − log λ0n + δλn (T ) . (7.3)
Y X
Z∼ ,
n (λn + δλn (T ))
0
n
To proceed further, we need two crucial observations as argued in section 2. Since λ0n is
T -independent, log T corrections would necessarily come from the corrected eigenvalue
– 30 –
δλn (T ). Also the Maclaurin expansion of log(a + x(T ))
∞
xn (T )
log(a + x(T )) = log a + (7.4)
X
,
n=1
an n
is a series in small x (T ), and therefore, the log x(T ) corrections come only when a = 0.
Adapting this observation to our case, this implies that the log T corrections are obtained
only from the zero modes of the Lichnerowicz operator when λ0n = 0. In this situation,
we have the following schematic form of the log T correction, given by
JHEP06(2024)034
log Z ∼ log (7.5)
Y
n δλn (T )
where δλn (T ) = nf1 T + O T 2 . Note that, if δλn (T ) necessarily starts from O(T ) as

any constant value would not contribute to log T as per the discussion in this point. It
is crucial to observe that the corrected eigenvalue scales linearly with n. This is seen
in all our previous computations, for example, in (2.49) for the Kerr-AdS4 black hole,
in (4.23) for the Kerr-Newman AdS black hole in N = 4 gauged supergravity, in (5.26)
for the Myers-Perry black hole in 5d, and finally in (6.20) for the AdS5 Kerr black hole.
So far, the evidence has been empirical observation for all the sets of black holes that
we have considered in this note. Later on, in this section, we shall motivate this further,
but for the time being, let us accept that the corrected eigenvalues scale linearly with n.
3. We need a third ingredient to establish our claim that we have stated at the beginning
of this section. This involves realising the range for n in equation (7.3) and this is where
point (1) above comes in. As described there, the near horizon geometry necessarily has
an AdS2 factor. The eigenfunctions h0n of the Lichnerowicz operator in equation (7.1)
are generated by diffeomorphisms. Namely,
hn ∼ Lξ(n) ḡ . (7.6)
Now, due to the presence of the AdS2 throat, Lξ(n) ḡ vanish for n = ±1, 0 as they
correspond to the isometry of AdS2 . Note that this fact, again, is universal due to
the argument present in point (1). For example, this expectation has been borne out
in (2.22) for the 4d Kerr-AdS black hole, in (4.18) for the 4d Kerr-Newman black hole
in N = 4 gauged supergravity, in (5.17) for the 5d Myers-Perry black hole, in (6.15) for
the AdS5 Kerr black hole. The presence of an AdS2 throat in the near horizon of all
the black hole geometries we have studied implies n runs from |n| ≥ 2, as argued above.
This immediately gives us
∞
1
log Z ∼ log (7.7)
Y
.
n=2
nf1 T
To regularize the above result, we use Zeta function regularization [27] that gives
∞
α 3
∼ α− 2 . (7.8)
Y
n=2
n
– 31 –
Setting α = T,
1
this immediately gives us the log T correction to be
3
log Z ∼ log T + . . . (7.9)
2
Note, from the arguments above, the presence of the factor 32 in the log T correction
relies just on two universal facts, which where shown to be manifestly true in all the
cases we have analysed above, viz.,
(a) Presence of an AdS2 throat in the near horizon geometry, and

(b) Regularized value of an infinite product.
JHEP06(2024)034
Since the above two points are independent of the black hole geometry in question, the
presence of an universal 32 factor in the log T correction is strongly motivated.
Insensitive to matter sector. One could wonder if the presence of matter affects the
argument presented above. Certainly, it is true, that in presence of matter, there could be
some other contribution over and above the one discussed above. However, we can go beyond
and motivate that the log T receives no correction from the matter sector. This follows
from observing the terms that actually contribute to the corrected eigenvalue of the zero
mode (2.46). The contributions come from only two terms in hαβ δ∆αβ,µν hµν , viz., from

• hαβ δ 1 αµ βν
2 g g □hµν

• hαβ δ Rαµβν hµν
For example, this claim can be checked from (3.27) and we commented on this fact below (3.28).
This implies that the contribution comes only from the metric data and is independent of
the matter sector and the arguments presented above lead us to conclude that the 32 log T
correction is universal and linked solely to the presence of near horizon AdS 2 throat.
8 Conclusions
In this manuscript we have systematically discussed the low-temperature quantum contribu-

tions to the partition functions of near-extremal rotating black holes. In particular, we have
explicitly treated various black holes in asymptotically four-dimensional AdS spacetime: the
Kerr-AdS4 black hole in section 2, the Kerr-Newman-AdS4 black hole in section 3 and the
rotating black hole with two electric charges in N = 4 gauged supergravity in section 4. In
five dimensions we have addressed the asymptotically flat Myers-Perry black hole in section 5
and the Kerr-AdS5 in section 6. In every case we found that the contribution to the path
integral arising from the gravitational tensor modes is universal and equal to 32 log THawking .
With the above overwhelming amount of evidence at hand, we argue for this universality
based on two main facts: (i) The near-horizon geometry contains the SL(2, R) group of
isometries and (ii) The relevant contribution from the Lichnerowicz operator is matter-
independent; the answer, in all cases, comes from only two geometric terms: g αµ g βν □ and
Rαµβν . Our computations explicitly demonstrate a direct and universal embedding of an
important quantum aspect, first discussed in the context of JT gravity, into higher-dimensional
rotating black holes in both asymptotically flat and asymptotically AdS spacetimes.
– 32 –
Throughout this manuscript we have concentrated on the tensor zero modes associated
with the large gauge transformations of the near-horizon AdS2 geometry. On general grounds,
it is expected that an additional zero mode related to fluctuations of the angular velocity
should also be present in rotating black holes [14, 36]. Attempts to identify such a zero mode,
however, lead to some confusing results in [14], and we have refrained from considering it in
this paper. Nevertheless, it is safe to conclude that in any ensemble the 32 log T correction
coming from the tensor zero modes is sacrosanct.
There is a number of interesting problems that our result motivates. It would be
interesting to track our arguments in a more rigorous fashion. Black hole dynamics is highly
dependent on the dimensionality of spacetime, the universality discussed in this paper, being
JHEP06(2024)034
quantum, opens the door for other similar universal properties for all near-extremal black
holes in various dimensions.
Given recent advances in the understanding of the microscopic entropy of asymptotically
AdS black holes, it would be interesting to consider their partition functions in the near-
extremal context along the lines pursued in this manuscript. In section 4 we have discussed
only the tensor modes for the two-charge, rotating black hole in N = 4 gauged supergravity
whose field theory dual is well understood; a complete analysis will require a proper treatment
of the vector and fermionic zero modes. One expects that in these supersymmetric cases the
relevant intuition should arise from supersymmetric versions of JT gravity. Indeed, some
relevant work has already been presented in [4] and it would be interesting to extend it to
other rotating black holes and to more general configurations.
It remains a truly interesting problem to derive this low-temperature thermodynamics
directly from field theory in the context of the AdS/CFT correspondence. A natural starting
point is to consider the reduction of the dual field theory to one-dimension. For example,
in [62], the reduction of the three-dimensional theory dual to certain magnetically charged
asymptotically AdS4 black hole was performed but extracting the expected behavior might
prove a Herculean task. It is also possible to consider bottom-up toy models arising from
modifications of the Sachdev-Ye-Kitaev models which contain the Schwarzian mode that
figures so prominently on the gravity side. This direction holds the promise to address some
of the puzzles (including the factorization puzzle) that low-dimensional models of gravity
such as JT gravity, have brought upon us.
Finally, in this manuscript we have studied a particular quantum effect that takes place
at low temperatures and we have demonstrated its universality. It would be interesting to
clarify the interplay of the logarithmic contribution we discussed here with other phenomena
that takes place at very low temperatures. For example, Kerr and Kerr-Newman black holes
suffer from superradiance instability [63] as well as Schwinger pair production (in the case
of charged black holes). There have been a number of discussions regarding superradiance
instability in asymptotically AdS spacetimes [64, 65] and very recently within the context
of AdS/CFT in [66]. It is quite plaussible that the JT point of view spoused in this paper
might lead to insight into those phenomena. We hope to return to this issue elsewhere.
Acknowledgments
We are grateful to Alejandra Castro, Marina David, Matt Heydeman, Imtak Jeon, Alfredo
Gonzàlez Lezcano, Juan Maldacena, Jun Nian and Joaquín Turiaci for discussions on related
– 33 –
Black hole types Variables constrained as constants
Kerr-AdS4 J
Kerr-Newman-AdS4 a
Kerr-Newman-AdS4 in supergravity m
Myers-Perry J
Kerr-AdS5 a
Table 1. Ensemble choices.
topics. This work is supported in part by the U.S. Department of Energy under grant DE-
JHEP06(2024)034
SC0007859. LPZ gratefully acknowledges support from an IBM Einstein fellowship during
2022/2023 while at the Institute for Advanced Study. AR is supported by an appointment
to the JRG program at the APCTP through the Science and Technology Promotion Fund
and the Lottery Fund of the Korean Government, by the Korean Local Governments -
Gyeongsangbuk-do Province and Pohang City and by the National Research Foundation of
Korea (NRF) grant funded by the Korean government (MSIT) (No. 2021R1F1A1048531).
SM acknowledges the warm hospitality of APCTP, Postech, South Korea during the APCTP
Winter School on Fundamental Physics 2024, where a part of this work was carried out. SM
is financially supported by the institute post-doctoral fellowship of IITK.
A Ensemble choice
This appendix clarifies our ensemble choices in the main text and the potential problems
raised by those choices. Table 1 shows the variables chosen to be constant when we turn
on a small temperature.
For Kerr-AdS4 black hole in section 2 and Myers-Perry black hole in section 5, we keep
the angular momentum J as a constant, which means we work in the canonical ensemble.
For other black holes, we choose different ensembles for calculational convenience. For
these ensemble choices with constant m or a, there could be additional contributions from
the rotational zero modes. See, for instance, the discussion of this puzzle in section 3.2.2
of [14]. As this paper aims to show the universal 32 log T contribution from gravitational
tensor zero modes, we leave the discussion about additional contributions from rotational
zero modes to future work.
Within the discussion of tensor zero modes, we believe our results are ensemble-independent.
To show that, we provide the parallel story for the Kerr-AdS black hole in a different ensemble
from that in section 2. In the new ensemble, we keep the mass constant, m = m0 , while a
depending on the temperature, a = a(T ). As we will see in the following subsections, working
in this new ensemble does not influence the 32 log T result.
A.1 A new ensemble choice and near-horizon limit
In this subsection, we illustrate our new ensemble choice. We obtain the near-horizon
geometry on the zeroth and the first orders of temperature.
– 34 –
We start with the Einstein-Hilbert action:
1 √
Z
I=− d4 x −g (R − 2Λ) + Iboundary . (A.1)
16π M
The Kerr-Newman-AdS4 solution is given by
!2
∆r Σ2 2 Σ2 2 ∆ θ r2 +a2
2
a

ds2 = − 2 dt− sin2 θdϕ + dr + dθ + 2 sin2 θ adt− dϕ , (A.2)
where

∆r = r 2 + a 2 a + g 2 r2 − 2M r , ∆θ = 1 − g 2 a2 cos2 θ , (A.3)
JHEP06(2024)034
p
Ξ = 1 − g 2 a2 , Σ= r2 + a2 cos2 θ . (A.4)
The cosmological constant Λ is related g ≡ 1

L by Λ = −3g 2 . At extremality, we have
r0 (1 + g 2 r02 )2
m0 = , (A.5)
1 − g 2 r02
s
1 + 3g 2 r02
a 0 = r0 , (A.6)
1 − g 2 r02
where r0 is the radius of the extreme black hole horizon, g = L1 and L is the AdS radius. In
our new ensemble, we choose m to be constant. a and r+ depend on T ,
a(T ) = a0 + c1 T + c2 T 2 + c3 T 3 + O(T 4 ), (A.7)

r+ (T ) = r0 + d1 T + d2 T + d3 T + O(T ).
2 3 4
(A.8)
Where the coefficients ci , di can be obtained by solving the second equation of (4.8) and ∆r = 0.
We verify that c1 = 0. Other coefficients are lengthy and not illuminating to be written here.
The Hawking entropy can be calculated using the above temperature dependence,
2πr02 8π 2 r03 (1 − g 2 r02 )
S= + T + O(T 2 ). (A.9)
1 − 3g r0
2 2 1 + 3g 2 r02 − 21g 4 r04 + 9g 6 r06
−iτ −ik3 τ
r = r+ (T ) + T k1 (y − 1), t= , ϕ=φ+ + k4 iτ, (A.10)
k2 T T
where
4πr02 (1 + g 2 r02 )
k1 = , (A.11)
1 + 6g 2 r02 − 3g 4 r04
k2 = 2π, (A.12)
r
−1−3g 2 r02
g g (−1+g 2 r02 )
2 (−1 + 3g 2 r02 )
k3 = − , (A.13)
2k2 r0
r
1+3g 2 r02
g(−1 + g 2 r02 )(−1 + 3g 2 r02 ) g 2 −g 4 r02
k4 = − . (A.14)
−1 − 6g 2 r02 + 3g 4 r04
– 35 –
!
dy 2 2
ḡµν dxµ dxν = g1 (θ) 2 +dτ 2 y 2 −1 +g2 (θ)dθ2 +g3 (θ) dφ+ik(y−1)dτ , (A.15)
y −1
where
r2 3 + g 2 r02 + (1 + 3g 2 r02 ) cos(2θ)

g1 (θ) = − 0 ,
−2 − 12g 2 r02 + 6g 4 r04
r02 3 + g 2 r02 + (1 + 3g 2 r02 ) cos(2θ)

g2 (θ) = − ,
2 −1 + g 2 r02 + g 2 r02 (1 + 3g 2 r02 ) cos2 θ (A.16)
2r02 −2 + 3g 2 r02 + 3g 4 r04 + g 2 r02 1 + 3g 2 r02 cos(2θ) sin2 θ

JHEP06(2024)034
g3 (θ) = − 2 ,
1 − 3g 2 r02 1 − g 2 r02 + 1 + 3g 2 r02 cos2 θ

k = −k4 .
By keeping the first order in T , we obtain δgµν , the near-horizon geometry on O(T ). This
expression is lengthy and not illuminating to be written here.
A.2 Lichnerowicz operator and the zero modes

The Lichnerowicz operator of our Einstein-Hilbert action with cosmology constant is as follows:
1 1 1 1

hαβ ∆αβ,µν hµν = − hαβ g αµ g βν □− g αβ g µν □+Rαµβν +Rαµ g βν −Rαβ g µν − Rg αµ g βν
L
16π 2 4 2
1 1

+ Rg αβ g µν +Λg αµ g βν − Λg αβ g µν hµν . (A.17)
4 2
At zero temperature when g = ḡ, one can verify that the above operator has the following
zero modes:
|n|
y−1

2
h(n) µ
µν dx dx
ν
= cn e inτ
3 + g 2 r02 + 1 + 3g 2 r02 cos 2θ
y+1
! (A.18)
|n| dτ dy dy 2
dτ − 2i
2
− , |n| ≥ 2,
n y 2 − 1 (y 2 − 1)2
where cn is the normalization constant.
A.3 The lifted eigenvalue and entropy correction

eigenvalue. Now we substitute g = ḡ + δg into the Lichnerowicz operator (A.17) and apply
it to the zero modes (A.18). The lifted eigenvalue is
Z
(n)∗
δλn = dx4 ḡhαβ δ∆αβ,µν h(n) (A.19)
p
L µν .
the result is simple,
3n(1 − g 2 r02 )T
δλn = , |n| ≥ 2. (A.20)
32r0
– 36 –
is therefore:
   
Y π 32πr0
δ log Z = log   = log  (A.21)
Y
.
n≥2
λ n n≥2
3n(1 − g 2 r 2 )T
0
Using Zeta function regularization

Y α 1 1
=√ , (A.22)
n≥2
n 2π α 3/2
JHEP06(2024)034
the one-loop correction is
3
!
T
δ log Z = log + O(1), (A.23)
2 Tq
where the O(1) term is not reliable. The temperature scale Tq is the temperature under
which the above one-loop correction needs to be considered. Tq can be extracted from the
classical entropy (A.9). As indicated in [14],
1 + 3g 2 r02 − 21g 4 r04 + 9g 6 r06

Tq = . (A.24)
8π 2 r03 (1 − g 2 r02 )
B SL(2, R) symmetries of the near-horizon geometries
Let us see the action of SL(2, R) explicitly. We recall [24] showed that the near-horizon
extremal Kerr metric
1 + cos2 θ 2r02 sin2 θ

! 2
r2 2 r02 2 r

ds =
2
− dt + dr + dθ 2
+ dϕ + 2 dt , (B.1)
2 2
r0 r 2 1 + cos θ
2 r0
admits the following Killing vectors. The precise claim is that first, the metric (B.1) has
enhanced symmetry with respect to the full Kerr metric. Namely, in addition to ∂t and ∂ϕ of
the original Kerr, it is invariant under r → α r, t → t/α, for any constant α. To see the
full SL(2, R) × U(1) it is convenient to transform to global coordinate where SL(2, R) arises
as the isometries of AdS2 and U(1) is generated by ∂ϕ .
Every near-horizon metric we arrive at in this manuscript has a near horizon that is of
the form A.1 and thus admit the action of SL(2, R) × U(1) as isometries.
Further improvement on this observation were presented in, for example, [53] that
established that similar results more broadly for large classes of near-horizon backgrounds
in fairly general theories, including in higher dimensions. Some discussion of the above
geometry in the context of the rotating attractor mechanism was originally presented in [59]
and extended to rotating asymptotically AdS4 black holes in [67].
B.1 Near-horizon symmetry of Kerr-AdS4 black hole

Here, we establish the fact that our near horizon metric, as given in equation (2.22) is
diffeomorphic to the metric given in equation (2) of [60]. Once established, this guarantees
– 37 –
the existence of an isometry group G = SL(2, R) × U(1), as proven in [60]. The starting
point for us is the NHEK-AdS given in (2.22). The computation proceeds in two steps.
First we do a coordinate transformation such that the NHEK-AdS metric may be written
in the following manner
!
dy 2 2
2
ds2NHEK-AdS = g1 (θ) + dτ 2 y 2 − 1 + g3 (θ)−1 dθ̃ + g3 (θ) dϕ + ig4 (y)(y − 1)dτ
y −1
2
(B.2)
for some coordinate θ̃ obtained as θ → θ̃(θ). This transformation is required to match the form
given in [60]. To obtain the functional form of θ̃, we introduce a coordinate transformation as
JHEP06(2024)034
θ → θ̃(θ) : g2 (θ)dθ2 = g2 (θ)θ̃′ (θ)2 dθ̃2 (B.3)
and demand
g2 (θ)θ̃′ (θ)2 = g3 (θ)−1 . (B.4)
The above equation is solved to give

3r 2
1 1 − L20 θ
Z
θ̃(θ) = dθ p =− log cot . (B.5)
g2 (θ)g3 (θ) 2r0
2 2
For the second step, we see that the metric now has the general form which we write
succinctly as
ds2 = G11 dτ 2 + G22 dy 2 + G33 dθ̃2 + G44 dϕ2 + 2G14 dτ dϕ, (B.6)
where, for example, G33 = g3−1 (θ) and so on. To establish the desired diffeomorphism that
takes the metric in equation (B.6) to the one in equation (2) of Lucietti et al, we do a further
coordinate transformation given by
dτ = dv + ady , dϕ = dφ + bdy , (B.7)
crucially without involving θ̃. This induces the following change in the metric given by
ds2 = G11 dv 2 + 2 (aG11 + bG14 ) dvdy + 2G14 dvdφ + G33 dθ2 + G44 dφ2 +
G11 a2 + G44 b2 + 2G14 ab + G22 dy 2 + 2(G14 a + G44 b)dφdy . (B.8)

The next step is to choose a and b judiciously such that
(gφy :) G14 a + G44 b = 0 , (B.9)

(gyy :) G11 a + G44 b + 2G14 ab + G22 = 0 .
2 2
(B.10)
. The above pair of equations can simultaneously be solved for

√ √
G22 G44 G22
a=∓ r
2
, b = ±r . (B.11)
G G44 G11 G244
G14 G44 − 11 G2
G 44 − G2
14 14
This choice of a and b takes our metric to the form stated in [60] and guarantees the existence
of the isometry group SL(2, R) × U(1).
– 38 –
B.2 Near-horizon symmetry of Kerr-AdS black hole in various dimensions
The situation in dimensions five, six and seven can be addressed with the help of the following
Lemma [53]. The metric

ds = Γ(ρ) A0 r dv + 2dvdr + dρ2 + γij (ρ)(dxi + k i rdv)(dxj + k j rdv)
2 2 2
(B.12)
has isometry group Ĝ3 × U(1)D−3 where the three-dimensional group Ĝ3 is 2d Poincarè
if A0 = 0 or O(2, 1) if A0 ̸= 0. The orbits of Ĝ3 are three-dimensional if k i =
̸ 0 and
two-dimensional if k = 0.
i
JHEP06(2024)034
This powerful statement is central in arguing for the universality of the 32 log T contribution
coming from the tensor modes. Coupling the existence of the SL(2, R) subgroup of isometries
with the fact that the contribution to the Lichnerowicz operator is strictly metric-driven,
as the contributions from other matter fields cancel among themselves, we obtain necessary
ingredients towards the universality of the 32 log T contribution in various dimensions.
C The structure of the Lichnerowicz operator
The generalized Lichnerowicz operator is obtained by the second-order metric perturbation

of the action. In this subsection, we list the ingredients of the operator used in the previous
sections.
For L = R ∗ 1, we can see Einstein-Hilbert’s contribution to the Lichnerowicz operator is
1 αµ βν ¯ 1 αβ µν ¯

hαβ ∆αβ,µν
EH hµν = hαβ ḡ ḡ □ − ḡ ḡ □ + R̄αµβν
2 4
1 αµ βν 1 αβ µν

+R̄ ḡ − R̄ ḡ − R̄ḡ ḡ + R̄ḡ ḡ
αµ βν αβ µν
hµν . (C.1)
2 4
For L = ∗F ∧ F = 1
2 ∗ F 2 , the contribution is
1

hαβ ∆αβ,µν hµν = hαβ − F 2 2g αµ g βν − g αβ g µν + F αµ F βν
F
8

+2F αγ
F µγ g βν −F αγ
F βγ g µν hµν , (C.2)
which can be applied to F1 and F2 in section 4.

For L = ∗1, the kinetic term is
1 1

hαβ ∆αβ,µν hµν = hαβ − g αµ g βν + g αβ g µν hµν , (C.3)
1
2 4
which can be used to construct the contribution of cosmology constant.
For L = ∗dζ ∧ dζ, the kinetic term is
1 1

hαβ ∆αβ,µν hµν = hαβ ∇γ ζ∇γ ζ − g αµ g βν + g αβ g µν +2g αµ ∇β ζ∇ν ζ −g αβ ∇µ ζ∇ν ζ hµν ,
ζ
2 4
(C.4)
which can also be applied to χ by ζ ↔ χ.
– 39 –
Open Access. This article is distributed under the terms of the Creative Commons
Attribution License (CC-BY4.0), which permits any use, distribution and reproduction in
any medium, provided the original author(s) and source are credited.
References
[1] J. Preskill et al., Limitations on the statistical description of black holes, Mod. Phys. Lett. A 6
(1991) 2353 [INSPIRE].
[2] L.V. Iliesiu and G.J. Turiaci, The statistical mechanics of near-extremal black holes, JHEP 05
(2021) 145 [arXiv:2003.02860] [INSPIRE].
JHEP06(2024)034
[3] M. Heydeman, L.V. Iliesiu, G.J. Turiaci and W. Zhao, The statistical mechanics of near-BPS
black holes, J. Phys. A 55 (2022) 014004 [arXiv:2011.01953] [INSPIRE].
[4] J. Boruch, M.T. Heydeman, L.V. Iliesiu and G.J. Turiaci, BPS and near-BPS black holes in
AdS5 and their spectrum in N = 4 SYM, arXiv:2203.01331 [INSPIRE].
[5] A. Almheiri and J. Polchinski, Models of AdS2 backreaction and holography, JHEP 11 (2015) 014
[arXiv:1402.6334] [INSPIRE].
[6] K. Jensen, Chaos in AdS2 Holography, Phys. Rev. Lett. 117 (2016) 111601 [arXiv:1605.06098]
[INSPIRE].
[7] J. Maldacena, D. Stanford and Z. Yang, Conformal symmetry and its breaking in two
dimensional Nearly Anti-de-Sitter space, PTEP 2016 (2016) 12C104 [arXiv:1606.01857]
[INSPIRE].
[8] T.G. Mertens and G.J. Turiaci, Solvable models of quantum black holes: a review on
Jackiw-Teitelboim gravity, Living Rev. Rel. 26 (2023) 4 [arXiv:2210.10846] [INSPIRE].
[9] G.J. Turiaci, New insights on near-extremal black holes, arXiv:2307.10423 [INSPIRE].
[10] R. Emparan and J.M. Magan, Tearing down spacetime with quantum disentanglement, JHEP 03
(2024) 078 [arXiv:2312.06803] [INSPIRE].
[11] A. Castro and V. Godet, Breaking away from the near horizon of extreme Kerr, SciPost Phys. 8
(2020) 089 [arXiv:1906.09083] [INSPIRE].
[12] A. Castro et al., Gravitational perturbations from NHEK to Kerr, JHEP 07 (2021) 218
[13] D. Kapec, A. Sheta, A. Strominger and C. Toldo, Logarithmic Corrections to Kerr
Thermodynamics, arXiv:2310.00848 [INSPIRE].
[14] I. Rakic, M. Rangamani and G.J. Turiaci, Thermodynamics of the near-extremal Kerr spacetime,
arXiv:2310.04532 [INSPIRE].
[15] L.V. Iliesiu, S. Murthy and G.J. Turiaci, Revisiting the Logarithmic Corrections to the Black
Hole Entropy, arXiv:2209.13608 [INSPIRE].
[16] N. Banerjee and M. Saha, Revisiting leading quantum corrections to near extremal black hole
thermodynamics, JHEP 07 (2023) 010 [arXiv:2303.12415] [INSPIRE].
[17] N. Banerjee, M. Saha and S. Srinivasan, Logarithmic corrections for near-extremal black holes,
JHEP 02 (2024) 077 [arXiv:2311.09595] [INSPIRE].
[18] M. David, J. Nian and L.A. Pando Zayas, Gravitational Cardy Limit and AdS Black Hole
Entropy, JHEP 11 (2020) 041 [arXiv:2005.10251] [INSPIRE].
– 40 –
[19] M. David, V. Godet, Z. Liu and L.A. Pando Zayas, Non-topological logarithmic corrections in
minimal gauged supergravity, JHEP 08 (2022) 043 [arXiv:2112.09444] [INSPIRE].
[20] B. Carter, Hamilton-Jacobi and Schrodinger separable solutions of Einstein’s equations, Commun.
Math. Phys. 10 (1968) 280 [INSPIRE].
[21] J.F. Plebanski and M. Demianski, Rotating, charged, and uniformly accelerating mass in general
relativity, Annals Phys. 98 (1976) 98 [INSPIRE].
[22] M.M. Caldarelli, G. Cognola and D. Klemm, Thermodynamics of Kerr-Newman-AdS black holes
and conformal field theories, Class. Quant. Grav. 17 (2000) 399 [hep-th/9908022] [INSPIRE].
[23] G.W. Gibbons, M.J. Perry and C.N. Pope, The first law of thermodynamics for Kerr-anti-de
JHEP06(2024)034
Sitter black holes, Class. Quant. Grav. 22 (2005) 1503 [hep-th/0408217] [INSPIRE].
[24] J.M. Bardeen and G.T. Horowitz, The Extreme Kerr throat geometry: A vacuum analog of
AdS2 × S 2 , Phys. Rev. D 60 (1999) 104030 [hep-th/9905099] [INSPIRE].
[25] S.M. Christensen and M.J. Duff, Quantizing Gravity with a Cosmological Constant, Nucl. Phys.
B 170 (1980) 480 [INSPIRE].
[26] R. Camporesi and A. Higuchi, Spectral functions and zeta functions in hyperbolic spaces, J.
Math. Phys. 35 (1994) 4217 [INSPIRE].
[27] A. González Lezcano, I. Jeon and A. Ray, Supersymmetric localization: N = (2) theories on S 2
and AdS2 , JHEP 07 (2023) 056 [Erratum ibid. 09 (2023) 003] [arXiv:2302.10370] [INSPIRE].
[28] Z.-W. Chong, M. Cvetic, H. Lu and C.N. Pope, Charged rotating black holes in four-dimensional
gauged and ungauged supergravities, Nucl. Phys. B 717 (2005) 246 [hep-th/0411045] [INSPIRE].
[29] M. Cvetic, G.W. Gibbons, H. Lu and C.N. Pope, Rotating black holes in gauged supergravities:
Thermodynamics, supersymmetric limits, topological solitons and time machines,
hep-th/0504080 [INSPIRE].
[30] D.D.K. Chow and G. Compère, Dyonic AdS black holes in maximal gauged supergravity, Phys.
Rev. D 89 (2014) 065003 [arXiv:1311.1204] [INSPIRE].
[31] S. Choi, C. Hwang, S. Kim and J. Nahmgoong, Entropy Functions of BPS Black Holes in AdS4
and AdS6 , J. Korean Phys. Soc. 76 (2020) 101 [arXiv:1811.02158] [INSPIRE].
[32] D. Cassani and L. Papini, The BPS limit of rotating AdS black hole thermodynamics, JHEP 09
(2019) 079 [arXiv:1906.10148] [INSPIRE].
[33] S. Choi, C. Hwang and S. Kim, Quantum vortices, M2-branes and black holes,
[34] J. Nian and L.A. Pando Zayas, Microscopic entropy of rotating electrically charged AdS4 black
holes from field theory localization, JHEP 03 (2020) 081 [arXiv:1909.07943] [INSPIRE].
[35] A. Sen, Logarithmic Corrections to N = 2 Black Hole Entropy: An infrared Window into the
Microstates, Gen. Rel. Grav. 44 (2012) 1207 [arXiv:1108.3842] [INSPIRE].
[36] A. Sen, Logarithmic Corrections to Rotating Extremal Black Hole Entropy in Four and Five
Dimensions, Gen. Rel. Grav. 44 (2012) 1947 [arXiv:1109.3706] [INSPIRE].
[37] A. Sen, Logarithmic Corrections to Schwarzschild and Other Non-extremal Black Hole Entropy
in Different Dimensions, JHEP 04 (2013) 156 [arXiv:1205.0971] [INSPIRE].
[38] J.T. Liu, L.A. Pando Zayas, V. Rathee and W. Zhao, Toward Microstate Counting Beyond Large
N in Localization and the Dual One-loop Quantum Supergravity, JHEP 01 (2018) 026
– 41 –
[39] I. Jeon and S. Lal, Logarithmic Corrections to Entropy of Magnetically Charged AdS4 Black
Holes, Phys. Lett. B 774 (2017) 41 [arXiv:1707.04208] [INSPIRE].
[40] J.T. Liu, L.A. Pando Zayas, V. Rathee and W. Zhao, One-Loop Test of Quantum Black Holes in
anti-de Sitter Space, Phys. Rev. Lett. 120 (2018) 221602 [arXiv:1711.01076] [INSPIRE].
[41] L.A. Pando Zayas and Y. Xin, Universal logarithmic behavior in microstate counting and the
dual one-loop entropy of AdS4 black holes, Phys. Rev. D 103 (2021) 026003 [arXiv:2008.03239]
[INSPIRE].
[42] D. Gang, N. Kim and L.A. Pando Zayas, Precision Microstate Counting for the Entropy of
Wrapped M5-branes, JHEP 03 (2020) 164 [arXiv:1905.01559] [INSPIRE].
JHEP06(2024)034
[43] F. Benini, D. Gang and L.A. Pando Zayas, Rotating Black Hole Entropy from M5 Branes, JHEP
03 (2020) 057 [arXiv:1909.11612] [INSPIRE].
[44] M. David and J. Nian, Universal entropy and hawking radiation of near-extremal AdS4 black
holes, JHEP 04 (2021) 256 [arXiv:2009.12370] [INSPIRE].
[45] F. Larsen and S. Paranjape, Thermodynamics of near BPS black holes in AdS4 and AdS7 , JHEP
10 (2021) 198 [arXiv:2010.04359] [INSPIRE].
[46] A. Castro, F. Larsen and I. Papadimitriou, 5D rotating black holes and the nAdS2 /nCFT1
correspondence, JHEP 10 (2018) 042 [arXiv:1807.06988] [INSPIRE].
[47] U. Moitra, S.K. Sake, S.P. Trivedi and V. Vishal, Jackiw-Teitelboim Gravity and Rotating Black
Holes, JHEP 11 (2019) 047 [arXiv:1905.10378] [INSPIRE].
[48] A. Castro, J.F. Pedraza, C. Toldo and E. Verheijden, Rotating 5D Black Holes: Interactions and
deformations near extremality, SciPost Phys. 11 (2021) 102 [arXiv:2106.00649] [INSPIRE].
[49] R.C. Myers and M.J. Perry, Black Holes in Higher Dimensional Space-Times, Annals Phys. 172
(1986) 304 [INSPIRE].
[50] R.C. Myers, Myers-Perry black holes, in Black holes in higher dimensions, G.T. Horowitz ed.,
Cambridge University Press (2012), p. 101–133 [arXiv:1111.1903] [INSPIRE].
[51] A. Galajinsky, Near horizon black holes in diverse dimensions and integrable models, Phys. Rev.
D 87 (2013) 024023 [arXiv:1209.5034] [INSPIRE].
[52] A. Galajinsky, A. Nersessian and A. Saghatelian, Superintegrable models related to near horizon
extremal Myers-Perry black hole in arbitrary dimension, JHEP 06 (2013) 002
[53] H.K. Kunduri, J. Lucietti and H.S. Reall, Near-horizon symmetries of extremal black holes,
Class. Quant. Grav. 24 (2007) 4169 [arXiv:0705.4214] [INSPIRE].
[54] A. Cabo-Bizet, D. Cassani, D. Martelli and S. Murthy, Microscopic origin of the
Bekenstein-Hawking entropy of supersymmetric AdS5 black holes, JHEP 10 (2019) 062
[55] S. Choi, J. Kim, S. Kim and J. Nahmgoong, Large AdS black holes from QFT,
[56] F. Benini and E. Milan, Black Holes in 4D N = 4 Super-Yang-Mills Field Theory, Phys. Rev. X
10 (2020) 021037 [arXiv:1812.09613] [INSPIRE].
[57] A. González Lezcano, J. Hong, J.T. Liu and L.A. Pando Zayas, Sub-leading Structures in
Superconformal Indices: Subdominant Saddles and Logarithmic Contributions, JHEP 01 (2021)
001 [arXiv:2007.12604] [INSPIRE].
– 42 –
[58] M. David, A. Lezcano González, J. Nian and L.A. Pando Zayas, Logarithmic corrections to the
entropy of rotating black holes and black strings in AdS5 , JHEP 04 (2022) 160
[59] D. Astefanesei et al., Rotating attractors, JHEP 10 (2006) 058 [hep-th/0606244] [INSPIRE].
[60] H.K. Kunduri and J. Lucietti, A classification of near-horizon geometries of extremal vacuum
black holes, J. Math. Phys. 50 (2009) 082502 [arXiv:0806.2051] [INSPIRE].
[61] P. Figueras, H.K. Kunduri, J. Lucietti and M. Rangamani, Extremal vacuum black holes in
higher dimensions, Phys. Rev. D 78 (2008) 044042 [arXiv:0803.2998] [INSPIRE].
[62] F. Benini, S. Soltani and Z. Zhang, A quantum mechanics for magnetic horizons, JHEP 05
JHEP06(2024)034
(2023) 070 [arXiv:2212.00672] [INSPIRE].
[63] Y.B. Zel’Dovich, Generation of Waves by a Rotating Body, JETP Lett. 14 (1971) 180.
[64] V. Cardoso and O.J.C. Dias, Small Kerr-anti-de Sitter black holes are unstable, Phys. Rev. D 70
(2004) 084011 [hep-th/0405006] [INSPIRE].
[65] V. Cardoso, O.J.C. Dias, J.P.S. Lemos and S. Yoshida, The black hole bomb and superradiant
instabilities, Phys. Rev. D 70 (2004) 044039 [Erratum ibid. 70 (2004) 049903] [hep-th/0404096]
[INSPIRE].
[66] S. Kim et al., Grey Galaxies’ as an endpoint of the Kerr-AdS superradiant instability, JHEP 11
(2023) 024 [arXiv:2305.08922] [INSPIRE].
[67] J.K. Ghosh and L.A. Pando Zayas, Comments on Sen’s classical entropy function for static and
rotating AdS4 black holes, Nucl. Phys. B 993 (2023) 116270 [arXiv:2009.11147] [INSPIRE].
– 43 –

Universality in logarithmic temperature corrections to near-extremal rotating black hole thermodynamics in various dimensions

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Universality in logarithmic temperature corrections to near-extremal rotating black hole thermodynamics in various dimensions

Uploaded by

Copyright:

Available Formats

Published for SISSA by Springer

Received: February 16, 2024

Universality in logarithmic temperature corrections to

and Jingchao Zhang b

Abstract: The low-temperature thermodynamics of near-extremal rotating black holes has

ArXiv ePrint: 2401.16507

Open Access, © The Authors.

2 Kerr-AdS4 black hole 3

4 Kerr-Newmann-AdS black hole in N = 4 gauged supergravity 15

5 Myers-Perry black holes in five dimensions 19

6 Kerr-AdS5 black hole 25

B SL(2, R) symmetries of the near-horizon geometries 37

C The structure of the Lichnerowicz operator 39

2 Kerr-AdS4 black hole

2.1 Near-extremal, near-horizon geometry

Note that a, which is related to J by equation (2.5), is given by

observation is at the heart of the original discussion of the breakdown of thermodynamics at

4πL2 r02 L2 + r02

O(1) : ds2(0) = ḡµν dxµ dxν

4L2 r02 sin2 θ L4 − L2 r02 − r02 L2 + 3r02 cos2 θ

O(T ) : ds2(1) = T gµν

(y − 1)f1′ (y) − f1 (y)

and solve for δf3 (y) from the equation

One can immediately check that

The vector field then generates the reparametrization η → η + ε′ (τ ) , τ → τ − ε, φ → φ − iε.

Therefore, setting λ0n = 0 in equation (2.47) we obtain the log T corrections as

Setting x = T −1 and ignoring constants irrelevant to the present pursuit, we obtain

The Kerr-Newman-AdS4 solution is given by [22]

where the r+ denotes radius of the outer horizon.

3.1 Ensemble choice and near-horizon limit

m(T ) = m0 + c1 T + c2 T 2 + c3 T 3 + O(T 4 ), (3.9)

3.2 Lichnerowicz operator and the zero modes

3.3 The lifted eigenvalues and entropy correction

3.4 Cancellations within the Lichnerowicz operator

Term 3a : hαβ δ(Rαµ g βν )hµν

Term 3b : hαβ δ(−Rαβ g µν )hµν = 0

Term 8a : hαβ δ(−4F αγ F µγ g βν )hµν

Term 8b : hαβ δ(2F αγ F βγ g µν )hµν = 0 , (3.27)

• As indicated in (3.27), all terms containing g µν hµν are zero by themselves.

• Term 3a + Term 4a + Term 5a + Term 6a + Term 7 + Term 8a = 0.

r12 + a2 cos2 θ a(r2 − r1 ) cos θ

which coincides with the inner horizon.

4.1 Ensemble choice and near-horizon limit

To obtain the near-horizon and low-temperature geometry, we first choose an ensemble.

where the precise expressions for the constants ki are as follows:

4.2 Lichnerowicz operator and the zero modes

where the different contributions to the Lichnerowicz operator can be found in C.

δλn ∝ nT, |n| ≥ 2. (4.23)

5 Myers-Perry black holes in five dimensions

where the parameters in the metric are given by

4J 2/3 4J 4/3 T 2 8J 5/3 T 3

We also obtain the corresponding correction to the near-horizon extreme Myers-Perry

Explicit computation of (5.25) gives us

and its first-order temperature correction (which we do not write explicitly)

J1 J2 (J1 + J2 + (J1 − J2 ) cos 2θ)

21/3 J1 sin2 θ(J1 + J2 )2/3 (2J1 − J2 cos 2θ + J2 )

21/3 J2 cos2 (θ)(J1 + J2 )2/3 (J1 cos 2θ + J1 + 2J2 )

J1 J2 sin2 2θ(J1 + J2 )2/3

ξ = einτ (f2 (y) ∂τ + f1 (y) ∂y + f3 (y) ∂φ1 + f4 (y) ∂φ2 ) , (5.40)

6.1 Ensemble choice and near-horizon limit

m(T ) = m0 + c1 T + c2 T 2 + c3 T 3 + O(T 4 ), (6.10)

A 2π 2 r03 4π 3 r04 (1 − g 2 r02 )

6.2 Lichnerowicz operator and the zero modes

6.3 The lifted eigenvalue and entropy correction