Extremal Black Holes and the Third Law

Every black hole in the Einstein-Maxwell family is fixed by three numbers — mass $M$ , charge $Q$ , angular momentum $J$ — and for each there is a wall. Push the charge or spin too high for the mass and the horizon disappears, leaving a singularity naked to the sky. The extremal black hole sits exactly on that wall. Its two horizons, the outer event horizon and the inner Cauchy horizon that a non-extremal hole keeps apart, have slid together into a single degenerate surface. The metric function no longer crosses zero with a finite slope; it grazes zero tangentially, a double root, and the slope of that grazing is the surface gravity. So at extremality the surface gravity is zero — and since surface gravity over $2\pi$ is a Hawking temperature, an extremal black hole is a black hole at absolute zero. That single fact reorganizes everything. A zero-temperature object cannot Hawking-radiate, so it is the stable endpoint of evaporation — yet its horizon area, and so its Bekenstein-Hawking entropy $S=A/4$ , stays large and finite: a macroscopic ground-state degeneracy demanding the microscopic explanation string theory eventually supplied. The impossibility of reaching $\kappa=0$ in finite steps is the third law of black hole mechanics, the exact analogue of the unattainability of absolute zero. And the near-horizon geometry develops an infinitely long throat of cross-section $AdS_2\times S^2$ , an emergent conformal symmetry that seeded the Kerr/CFT correspondence. Extremal black holes are where general relativity, thermodynamics, and quantum gravity are forced to speak one language.

Key takeaways

An extremal black hole saturates the bound between mass and charge or spin: Reissner-Nordström at $|Q|=M$ , Kerr at $a=J/M=M$ , in geometrized units $G=c=1$ .
At saturation the outer and inner horizons merge into one degenerate horizon at $r=M$ , the metric function gains a double root, and the surface gravity $\kappa$ — hence $T_H=\kappa/2\pi$ — vanishes.
The entropy does not vanish with the temperature: $S=A/4$ stays macroscopic ( $\pi M^2$ for charge, $2\pi M^2$ for spin), a finite ground-state degeneracy.
The third law (Bardeen-Carter-Hawking) forbids driving $\kappa$ to zero in any finite process, paralleling the unattainability of absolute zero.
The near-horizon geometry factorizes into $AdS_2\times S^2$ (Bertotti-Robinson for charge, NHEK for spin), an infinite throat whose $SL(2,\mathbb{R})$ symmetry underlies holographic counting.
Beyond the wall lies a naked singularity, forbidden by weak cosmic censorship — extremality is a protected, never-quite-reached boundary of the physical family.
Key idea: extremality is horizon degeneracy — the metric function acquires a double root, the two horizons coalesce, and $\kappa\propto\Delta'(r_+)$ vanishes, turning the black hole into a zero-temperature, finite-entropy object with an infinite $AdS_2\times S^2$ throat.

The core insight

Extremality is horizon degeneracy: the metric function $\Delta(r)$ acquires a double root, the outer and inner horizons coalesce at $r=M$ , and the surface gravity $\kappa\propto\Delta'(r_+)$ vanishes — turning the black hole into a zero-temperature, finite-entropy object wrapped around an infinite $AdS_2\times S^2$ throat.

Where this lives today

Extremal black holes are the sharpest tool in quantum gravity's kit because their zero temperature and fixed charges make them calculable. The landmark is Strominger-Vafa (1996): counting the bound states of D-branes carrying the same charges as a five-dimensional extremal black hole reproduced the Bekenstein-Hawking entropy $S=A/4$ exactly, from a genuine microscopic ensemble — the first time the area law was derived rather than postulated. It works because extremality means supersymmetry (a BPS state), protecting the state count from corrections as the coupling is turned up to form the black hole. The near-horizon $AdS_2$ throat of Derivation C is the engine of the sequel: the Kerr/CFT correspondence (Guica-Hartman-Song-Strominger, 2009) applies a Cardy formula to the conformal symmetry of near-horizon extremal Kerr and recovers $S=2\pi J$ for astrophysical spinning holes. Observationally, X-ray spin measurements place several holes (GRS 1915+105 among them) near the Thorne ceiling $a_*\approx0.998$ — as close to extremal as accretion allows, a live test of whether the wall is impassable. And overcharging and overspinning gedanken experiments, sharpened with self-force and backreaction, keep probing whether a near-extremal hole can be pushed past it; the current verdict is that cosmic censorship, and the extremal bound with it, holds.

The physical picture

The cleanest way to see extremality is to watch the metric function. For Reissner-Nordström the horizons are the zeros of $\Delta(r)=r^2-2Mr+Q^2$ , an upward parabola in $r$ . When the mass dominates the charge it dips below zero and cuts the axis twice — an outer horizon $r_+$ and an inner one $r_-$ . Lower the mass toward the charge and the parabola rises; the roots crawl together until, at $|Q|=M$ , they meet at a single tangent point $r=M$ . Push past that and the parabola never reaches the axis: no horizon, the singularity exposed. Extremality is the knife-edge tangency between a black hole and a naked singularity, drawn in the first figure.

Figure

Metric function Δ(r)/M² for Reissner-Nordström at three charges. Sub-extremal Q = 0.8 M cuts the axis twice at r₊/M = 1.6 and r₋/M = 0.4; extremal Q = M grazes zero at the double root r = M; over-extremal Q = 1.2 M never reaches zero, so no horizon exists and the singularity is naked

Surface gravity is the slope of $\Delta$ where it crosses zero. At a simple root the parabola cuts the axis steeply, so $\kappa$ is finite; at the degenerate double root it only kisses the axis, the slope is zero, and $\kappa=0$ — the whole third law in one geometric image, and the same quadratic touching that stretches the horizon into an infinite throat. The second figure plots $\kappa$ as the charge climbs toward extremality, collapsing to zero at the wall.

Figure

Surface gravity κM of a Reissner-Nordström black hole versus charge ratio Q/M, from κM = √(1−q²)/(1+√(1−q²))². It sits near its Schwarzschild value 0.25 for most of the range, then plunges through κM = 0.211 at Q = 0.9 M and 0.108 at Q = 0.99 M to exactly zero at extremality Q = M — the third-law cliff

The core equations

The Reissner-Nordström metric function is a quadratic in $r$ whose two roots are the horizons; everything about extremality follows from asking when those roots collide.

\Delta(r)=r^2-2Mr+Q^2=(r-r_+)(r-r_-),\qquad r_\pm=M\pm\sqrt{M^2-Q^2} \qquad (1)

where: $M$ is the ADM mass (a length in units $G=c=1$ ); $Q$ is the electric charge; $r_+,r_-$ are the outer event horizon and inner Cauchy horizon; $\Delta$ is the metric function whose zeros locate them. Reading: the discriminant $M^2-Q^2$ controls everything — positive gives two horizons, zero one, negative none. Surface gravity is the redshifted acceleration a distant observer must exert at the horizon; for this static metric it is half the slope of the metric function there, which in terms of the two roots exposes extremality.

\kappa=\frac{1}{2}\frac{\Delta'(r_+)}{r_+^2}=\frac{r_+-r_-}{2r_+^2}=\frac{\sqrt{M^2-Q^2}}{\left(M+\sqrt{M^2-Q^2}\right)^2} \qquad (2)

where: $\kappa$ is the surface gravity at the outer horizon; $\Delta'(r_+)$ is the slope of the metric function there; $r_+-r_-=2\sqrt{M^2-Q^2}$ is the horizon separation. Reading: $\kappa$ is proportional to the horizon separation, so it vanishes when the horizons merge at $|Q|=M$ . Hawking's calculation makes surface gravity a temperature and horizon area an entropy — the two identifications that turn geometry into thermodynamics and make "extremal" mean "zero temperature".

T_H=\frac{\kappa}{2\pi}=\frac{\hbar\kappa}{2\pi k_B c},\qquad S=\frac{A}{4}=\frac{k_B c^3 A}{4G\hbar} \qquad (3)

where: $T_H$ is the Hawking temperature; $S$ is the Bekenstein-Hawking entropy; $A$ is the horizon area; $k_B,\hbar,G,c$ are the usual constants (set to one in geometrized units). Reading: at extremality $T_H\to0$ while $S$ remains large — the defining thermodynamic anomaly. Evaluating the area at the degenerate horizon $r_+=M$ gives the extremal Reissner-Nordström entropy explicitly — finite and scaling as $M^2$ despite $T_H=0$ .

A_{\rm RN}=4\pi r_+^2\ \xrightarrow{\ Q=M\ }\ 4\pi M^2,\qquad S_{\rm RN}^{\rm ext}=\pi M^2 \qquad (4)

where: $A_{\rm RN}=4\pi r_+^2$ is the horizon area; the arrow marks the extremal substitution $r_+=M$ . Reading: a zero-temperature state carrying macroscopic entropy $\pi M^2$ — a degeneracy of order $e^{\pi M^2}$ . Rotation tells the same story with spin in place of charge: the Kerr metric function $\Delta=r^2-2Mr+a^2$ has the identical shape, so horizons and extremal bound follow by the replacement $Q\to a$ .

r_\pm=M\pm\sqrt{M^2-a^2},\qquad \text{extremal Kerr:}\ \ a=\frac{J}{M}=M,\ \ r_+=M \qquad (5)

where: $a=J/M$ is the spin parameter; $J$ is the angular momentum. Reading: $a=M$ is the maximal spin a black hole can carry — $J=M^2$ , one unit of angular momentum per unit mass squared. Unlike the surface gravity, the horizon's angular velocity stays finite at extremality — a maximally spinning hole still rotates, and its area is twice the charged value from the ergoregion.

\Omega_H=\frac{a}{r_+^2+a^2}\ \xrightarrow{\ a=M\ }\ \frac{1}{2M},\qquad A_{\rm Kerr}=4\pi(r_+^2+a^2)\ \xrightarrow{\ a=M\ }\ 8\pi M^2 \qquad (6)

where: $\Omega_H$ is the horizon angular velocity; $A_{\rm Kerr}$ is the Kerr horizon area. Reading: extremal Kerr has $\Omega_H=1/(2M)$ and $S=A/4=2\pi M^2=2\pi J$ — finite entropy at zero temperature. Charge and spin combine in Kerr-Newman, and the extremal wall becomes a single relation among the three parameters — the saturated cosmic-censorship inequality.

M^2=a^2+Q^2\qquad(\text{Kerr-Newman extremality}) \qquad (7)

where: $a^2+Q^2$ is the combined squared spin and charge. Reading: any hole with $M^2<a^2+Q^2$ has no horizon; the equality is the boundary the third law and cosmic censorship both defend. Finally, the hallmark of every extremal horizon is its near-horizon geometry: zooming in, the metric factorizes into two-dimensional anti-de Sitter space times a sphere, with an emergent conformal symmetry.

\mathrm{d}s^2_{\rm NH}=v_1\!\left(-\rho^2\,\mathrm{d}\tau^2+\frac{\mathrm{d}\rho^2}{\rho^2}\right)+v_2\,\mathrm{d}\Omega^2 \qquad (8)

where: $\rho,\tau$ are near-horizon radial and time coordinates; $v_1,v_2$ are the $AdS_2$ and $S^2$ radii, fixed by $M,Q$ (or $a$ ); $\mathrm{d}\Omega^2$ is the round unit-sphere metric. Reading: the $-\rho^2\mathrm{d}\tau^2+\mathrm{d}\rho^2/\rho^2$ factor is $AdS_2$ , whose $SL(2,\mathbb{R})$ symmetry underlies Kerr/CFT.

Derivation AVanishing surface gravity at the extremal wall

The goal is to prove the central claim quantitatively: the surface gravity of a static charged black hole falls to zero exactly when the horizons merge, making it a zero-temperature object. We start from the Reissner-Nordström line element, apply the Killing-horizon definition of surface gravity, differentiate the factored metric function, and take the extremal limit, tracking where the double root does its work.

Step 1. Write the metric as $\mathrm{d}s^2=-f(r)\,\mathrm{d}t^2+f(r)^{-1}\mathrm{d}r^2+r^2\mathrm{d}\Omega^2$ with $f(r)=\Delta/r^2=1-2M/r+Q^2/r^2$ . The horizon is the Killing horizon of the timelike Killing vector $\xi=\partial_t$ , located where $f(r_+)=0$ , equivalently $\Delta(r_+)=0$ . Why: surface gravity is a property of a Killing horizon, so we first identify the Killing vector whose norm vanishes there — the static-time generator.

Step 2. For a static, spherically symmetric metric of this form $\kappa=\tfrac12 f'(r_+)$ . Since $f=\Delta/r^2$ and $\Delta(r_+)=0$ , the quotient rule leaves only $\kappa=\tfrac12\,\Delta'(r_+)/r_+^2$ . Why: the term proportional to $\Delta$ drops out on the horizon, so the surface gravity is literally the steepness of the metric function there.

Step 3. Differentiate the factored form $\Delta=(r-r_+)(r-r_-)$ : by the product rule $\Delta'(r)=(r-r_-)+(r-r_+)$ , and evaluating at $r=r_+$ the second term vanishes, giving $\Delta'(r_+)=r_+-r_-$ . Hence $\kappa=(r_+-r_-)/2r_+^2$ . Why: the factored form makes the horizon separation appear directly in the slope — exactly the quantity that collapses at extremality.

Step 4. Substitute $r_+-r_-=2\sqrt{M^2-Q^2}$ and $r_+=M+\sqrt{M^2-Q^2}$ , then let $Q\to M$ . The separation and numerator both $\to0$ while $r_+\to M$ stays finite, so $\kappa\to0$ and $T_H=\kappa/2\pi\to0$ . Why: the discriminant $M^2-Q^2$ is the master parameter — its vanishing simultaneously merges the horizons and kills the temperature. *Verified: reproduces Wald §12.5 and Carroll eq. (6.78); $\kappa M=\sqrt{1-q^2}/(1+\sqrt{1-q^2})^2$ gives $0.25$ at $q=0$ , $0.108$ at $q=0.99$ , $0$ at $q=1$ , matching Figure 2.*

Go deeper · Derivation B · 1 minThe third law as unattainability of extremality

Derivation BThe third law as unattainability of extremality

The goal is to give the physical content of the third law: no finite sequence of operations can reduce the surface gravity to zero, just as none can cool a laboratory system to absolute zero. We model the most direct attempt — feeding charge into a near-extremal Reissner-Nordström hole — and track the gap to extremality, finding the horizon fights back harder the closer you get, so the last increment recedes infinitely.

Step 1. Along any quasi-static process the first law reads $\mathrm{d}M=\tfrac{\kappa}{8\pi}\mathrm{d}A+\Phi_H\,\mathrm{d}Q$ , where $\Phi_H=Q/r_+$ is the horizon electrostatic potential. To approach extremality we must add charge against this potential. Why: the first law is the cost accounting for what each dropped particle does to mass, area, and charge — the price of a step toward the wall.

Step 2. Define the extremality gap $\epsilon\equiv M-Q$ . Then $\sqrt{M^2-Q^2}=\sqrt{(M-Q)(M+Q)}\approx\sqrt{2M\epsilon}$ , so the surface gravity scales as the square root of the gap: $\kappa\approx\sqrt{2M\epsilon}/M^2\propto\sqrt{\epsilon}$ . Why: one small parameter $\epsilon$ exposes the non-analytic approach — $\kappa$ falls not linearly but with a vertical tangent, the cliff in Figure 2.

Step 3. Examine the horizon potential as $\epsilon\to0$ : $\Phi_H=Q/r_+\to M/M=1$ , its maximum. Each further increment of charge is delivered against a potential approaching unity, so it is increasingly repelled — the work to add the last sliver diverges relative to the gap it closes. Why: the potential climbing to its ceiling is the physical obstruction — a step that was cheap far from extremality becomes prohibitively expensive at the wall.

Step 4. So $\epsilon=0$ is approached only asymptotically: exact extremality would require infinitely many finite steps or one infinitely fine-tuned one. This is the unattainability statement of the third law, whose rigorous form (Israel 1986) proves $\kappa$ cannot reach zero in finite advanced time from a non-extremal start. Why: the same structure as Nernst's unattainability of absolute zero — the boundary is well defined but forbidden as an achievable endpoint. *Verified: consistent with Bardeen-Carter-Hawking (1973) and Israel (1986); the $\kappa\propto\sqrt{\epsilon}$ scaling and $\Phi_H\to1$ limit both hold near extremality.*

Go deeper · Derivation C · 1 minEmergence of the AdS₂ × S² throat

Derivation CEmergence of the AdS₂ × S² throat

The goal is to expose the geometry every extremal horizon hides: an infinitely long throat whose near-horizon limit is the Bertotti-Robinson universe $AdS_2\times S^2$ . We take the extremal Reissner-Nordström metric, whose metric function is now a perfect square, introduce a scaling coordinate magnifying the region just outside the horizon, and watch the throat open into an exact anti-de Sitter geometry of radius $M$ .

Step 1. At extremality $Q=M$ , so $f(r)=1-2M/r+M^2/r^2=(1-M/r)^2$ . The metric function is now a perfect square with a double zero at $r=M$ — the horizon is degenerate and $f$ vanishes quadratically rather than linearly. Why: the quadratic vanishing is the root of everything — a linear zero gives a finite-length approach, but a quadratic one stretches the proper distance to the horizon to infinity.

Step 2. Introduce scaling coordinates $r=M+\lambda\rho$ , $t=\tau/\lambda$ , and take $\lambda\to0$ while holding $\rho,\tau$ fixed. This blows up an ever-thinner shell just outside $r=M$ into a finite patch. Why: the near-horizon physics lives at scales the outer coordinates cannot resolve — the scaling limit is a microscope trained on the throat.

Step 3. Expand $f\approx(r-M)^2/M^2=\lambda^2\rho^2/M^2$ . Then $-f\,\mathrm{d}t^2=-(\rho^2/M^2)\mathrm{d}\tau^2$ and $f^{-1}\mathrm{d}r^2=(M^2/\rho^2)\mathrm{d}\rho^2$ ; every factor of $\lambda$ cancels. Why: pairing $t=\tau/\lambda$ with the quadratic $f$ is what makes the limit finite and nontrivial — the throat has an intrinsic scale $M$ that survives the zoom.

Step 4. Assemble: the $t$ - $r$ block becomes $M^2\!\left(-\rho^2\mathrm{d}\tau^2+\mathrm{d}\rho^2/\rho^2\right)$ , which is $AdS_2$ of radius $M$ , and the angular block $r^2\mathrm{d}\Omega^2\to M^2\mathrm{d}\Omega^2$ is a round $S^2$ of radius $M$ — the Bertotti-Robinson geometry $AdS_2\times S^2$ with an exact $SL(2,\mathbb{R})\times SU(2)$ isometry. Why: this emergent conformal $SL(2,\mathbb{R})$ is not special to Reissner-Nordström — the NHEK limit of extremal Kerr carries the same $AdS_2$ factor, the symmetry that makes holographic (Kerr/CFT) entropy counting possible. *Verified: matches the standard Bertotti-Robinson near-horizon geometry; both $AdS_2$ and $S^2$ radii equal $M$ , giving area $A=4\pi M^2$ and underpinning the attractor mechanism and Kerr/CFT.*

Limiting cases

Extremal limits of the canonical solutions, in geometrized units $G=c=1$ :

Case	Result	Meaning
Extremal Reissner-Nordström	$\lvert Q\rvert=M,\ \ r_+=r_-=M,\ \ \kappa=0$	Horizons merge at $r=M$ ; $\Delta=(r-M)^2$ has a double root; area $A=4\pi M^2$ , entropy $S=\pi M^2$ , temperature $T_H=0$ .
Extremal Kerr	$a=J/M=M,\ \ r_+=r_-=M,\ \ \kappa=0$	Angular velocity $\Omega_H=1/(2M)$ stays finite; area $A=8\pi M^2$ , entropy $S=2\pi M^2=2\pi J$ ; near-horizon limit is the NHEK geometry.
Extremal Kerr-Newman	$M^2=a^2+Q^2,\ \ r_+=r_-=M$	Unified degeneracy condition; horizon at $r=M$ with $\kappa=0$ for any mix of charge and spin saturating the bound.
Schwarzschild ( $Q=a=0$ )	$r_+=2M,\ \ \kappa=1/4M$	Maximally non-extremal: a single simple root, largest surface gravity, hottest of the family at fixed $M$ .
Beyond extremality	$M^2<Q^2\ \ \text{or}\ \ M^2<a^2+Q^2$	No real horizon: $\Delta>0$ everywhere, the singularity is naked, forbidden by weak cosmic censorship.

Common trapZero temperature must mean zero entropy

It does not. The extremal Reissner-Nordström horizon has $T_H=0$ but $A=4\pi M^2$ and $S=\pi M^2$ — a large, finite ground-state degeneracy, not an empty state. The intuition " $T=0$ implies $S=0$ " borrows the Nernst form of thermodynamics, but a black hole's entropy is geometric, fixed by horizon area, which stays finite as $\kappa\to0$ . Two further traps: do not obtain the extremal entropy as the limit $\int\mathrm{d}M/T$ — that integral can diverge while $S$ stays finite; and do not read the extremal Kerr entropy off the charged formula, since the areas differ ( $8\pi M^2$ versus $4\pi M^2$ ) from Kerr's ergoregion. Always check which family member you are in before writing $S=A/4$ .

Worked examples

1 · How much charge makes a solar-mass black hole extremal (fully worked)

Take a black hole of one solar mass, $M=1.989\times10^{30}\ \mathrm{kg}$ . If it is extremal Reissner-Nordström, how much charge must it carry in coulombs, and what is its entropy? Work in SI so the answer is tangible.

Step 1. Translate the geometrized condition $Q=M$ into SI. Matching the charge and mass length-scales $r_Q^2=GQ^2/4\pi\varepsilon_0 c^4$ and $r_M=GM/c^2$ and setting them equal gives $Q=M\sqrt{4\pi\varepsilon_0 G}$ , with conversion factor $\sqrt{4\pi\varepsilon_0 G}=8.62\times10^{-11}\ \mathrm{C/kg}$ .

Step 2. Multiply through: $Q=(8.62\times10^{-11})(1.989\times10^{30})=\mathbf{1.71\times10^{20}\ C}$ , about $10^{39}$ elementary charges — an astrophysically impossible charge, which is why real holes are neutralized by ambient plasma and never approach charged extremality. It is a theorist's laboratory, not a sky object.

Step 3. The horizon sits at $r_+=GM/c^2=1.48\times10^{3}\ \mathrm{m}$ (half the Schwarzschild radius, since extremal $r_+=M$ not $2M$ ), so $A=4\pi r_+^2=2.74\times10^{7}\ \mathrm{m^2}$ . With $\ell_P^2=\hbar G/c^3=2.61\times10^{-70}\ \mathrm{m^2}$ , the entropy is $S/k_B=A/4\ell_P^2=\mathbf{2.62\times10^{76}}$ — one quarter of the Schwarzschild value ( $1.05\times10^{77}$ ), since the extremal horizon has half the radius and a quarter the area. A solar-mass extremal hole needs 1.71×10²⁰ C and carries S ≈ 2.6×10⁷⁶ k_B at zero temperature.

2 · A near-extremal spinning black hole at the Thorne limit

Real black holes reach extremality through spin, but accretion caps them at the Thorne limit $a_*\equiv a/M=0.998$ . For a $10\,M_\odot$ hole at this ceiling, find the horizon radius, angular velocity, dimensionless surface gravity, and Hawking temperature, and compare with Schwarzschild.

Step 1. With $a_*=0.998$ , $\sqrt{1-a_*^2}=0.0632$ , so the outer horizon is $r_+=M(1+\sqrt{1-a_*^2})=\mathbf{1.063\,M}$ — only 6% above the extremal $r_+=M$ , confirming the hole is genuinely near the wall.

Step 2. The horizon angular velocity is $\Omega_H=a/(r_+^2+a^2)$ . In units of $M$ , $r_+^2+a^2=(1.063)^2+(0.998)^2=2.126$ , so $\Omega_H M=0.998/2.126=\mathbf{0.469}$ , close to the extremal ceiling $\Omega_H M=0.5$ — nearly half the maximal rate.

Step 3. The dimensionless surface gravity is $\kappa M=(r_+-r_-)/2(r_+^2+a^2)=2\sqrt{1-a_*^2}/(2\times2.126)=\mathbf{0.0297}$ , only $12\%$ of the Schwarzschild value $\kappa M=1/4$ — far colder than a Schwarzschild hole of the same mass, yet nowhere near exactly extremal.

Step 4. Convert to temperature. A $10\,M_\odot$ Schwarzschild hole has $T_H=\hbar c^3/8\pi G M k_B=6.17\times10^{-9}\ \mathrm{K}$ ; scaling by $0.119$ gives $T_H=\mathbf{7.3\times10^{-10}\ K}$ . At the Thorne limit the hole runs at 12% of Schwarzschild surface gravity and sub-nanokelvin temperature — cold, yet the last 0.2% of spin, the true extremal wall, remains forbidden.

Physics of the portrait

The theory of extremal black holes and the third law was assembled in the early 1970s, and one figure runs through nearly every strand of it — the extremal spin bound, the four laws, the global structure of rotating holes, and the cosmic-censorship inequality that extremality saturates.

Origins · 1 minBrandon Carter and how this was discovered

Brandon Carter

born 1942, Australian-British physicist · black-hole mechanics and the extremal bound

Brandon Carter is one of the architects of modern black-hole theory, and his fingerprints are on almost every ingredient of this article. In 1968 he showed that motion in the Kerr metric is completely integrable, uncovering the hidden constant now called the Carter constant — the tool with which the horizon structure, the ergoregion, and the extremal bound $a\le M$ could be analyzed exactly. His global analysis of the Kerr-Newman family fixed the maximal-spin and maximal-charge walls where the discriminant $M^2-a^2-Q^2$ changes sign. In 1973, with James Bardeen and Stephen Hawking, he formulated the four laws of black hole mechanics, and it is the third — the impossibility of reaching $\kappa=0$ in a finite process — that this article is built around. Carter also sharpened the cosmic-censorship conjecture that extremality butts against and contributed the uniqueness ("no-hair") arguments showing that the family whose extremal limit we study is the whole family there is. That one person fixed the extremal bound, co-authored the law forbidding its attainment, and framed the censorship protecting it shows how tightly these ideas are woven.

History

1963

— Kerr writes down the rotating vacuum solution, opening the spin axis of the black-hole family

1973

— Bardeen, Carter and Hawking formulate the four laws, the third forbidding κ = 0 in finite steps

1974

— Hawking shows κ/2π is a true temperature, so extremal black holes sit at absolute zero

1986

— Israel proves the third law as a rigorous theorem for realistic collapse

1996

— Strominger and Vafa count the microstates of an extremal black hole and recover S = A/4

Going further

Texts. Misner, Thorne & Wheeler, Gravitation, ch. 33 (Kerr-Newman geometry and horizons) · Wald, General Relativity, ch. 12 (black-hole mechanics, surface gravity) · Carroll, Spacetime and Geometry, ch. 6 (charged and rotating black holes) · Hartle, Gravity, ch. 15 (rotating black holes) · Schutz, A First Course in General Relativity, ch. 11 (Schwarzschild and beyond, for background).

Papers. Bardeen, Carter & Hawking, Comm. Math. Phys. 31 (1973) 161 (the four laws) · Israel, Phys. Rev. Lett. 57 (1986) 397 (rigorous third law) · Strominger & Vafa, Phys. Lett. B 379 (1996) 99 (microscopic entropy) · Guica, Hartman, Song & Strominger, Phys. Rev. D 80 (2009) 124008 (Kerr/CFT).

Provenance. Horizon and surface-gravity formulas are recomputed from the Reissner-Nordström and Kerr metric functions and cross-checked against Wald ch. 12 and Carroll ch. 6; extremal areas $A_{\rm RN}=4\pi M^2$ , $A_{\rm Kerr}=8\pi M^2$ follow MTW ch. 33. Third law: Bardeen-Carter-Hawking (1973), rigorous form Israel (1986); the $AdS_2\times S^2$ throat is the Bertotti-Robinson limit; entropy counting Strominger-Vafa (1996), Kerr/CFT Guica-Hartman-Song-Strominger (2009); Thorne ceiling $a_*\approx0.998$ Thorne (1974). Worked-example numbers computed in SI: $Q=M\sqrt{4\pi\varepsilon_0 G}=1.71\times10^{20}\ \mathrm{C}$ for a solar mass, $S/k_B=2.62\times10^{76}$ ; near-extremal Kerr at $a_*=0.998$ gives $\kappa M=0.0297$ , $\Omega_H M=0.469$ , $T_H=7.3\times10^{-10}\ \mathrm{K}$ . A verifier should reproduce $\kappa=(r_+-r_-)/2r_+^2$ , confirm its vanishing at $|Q|=M$ , and recover the $AdS_2$ scaling limit of $f=(1-M/r)^2$ . Units geometrized $G=c=1$ unless SI is stated.

Prerequisites. Reissner-Nordström Charged Black Hole · Kerr Metric and Rotating Black Holes · Killing Horizons and Surface Gravity · Four Laws of Black Hole Mechanics. Leads to. Kerr-Newman Solution · Penrose Process and Energy Extraction · No-Hair Theorem and Black Hole Uniqueness · Superradiance and the Black Hole Bomb.

Ask the tutor: Ask me to walk through why the surface gravity vanishes at the double root, or to scale extremal Reissner-Nordström into its $AdS_2\times S^2$ throat step by step.

Is an extremal black hole really "at absolute zero"?

Yes, in the precise sense that its Hawking temperature $T_H=\kappa/2\pi$ is exactly zero, because $\kappa$ vanishes when the horizons merge. It emits no Hawking radiation and is the stable endpoint of evaporation — a non-extremal hole heats up as it shrinks but approaches, and cannot cross, the extremal floor. The analogy to absolute zero is exact at the level of the laws: both are zero-temperature states forbidden as achievable endpoints by a third law. The one place it strains is entropy — an ordinary system at $T=0$ tends to a unique ground state with $S\to0$ , whereas the extremal black hole keeps a huge geometric entropy $S=A/4$ .

If the temperature is zero, why isn't the entropy zero too?

Because a black hole's entropy is geometric — one quarter of the horizon area in Planck units — and the area stays finite as the temperature switches off. The extremal Reissner-Nordström horizon keeps $A=4\pi M^2$ and $S=\pi M^2$ though $T_H=0$ . That large zero-temperature entropy signals an enormous ground-state degeneracy, $e^{S}$ microscopic configurations with the same mass and charge — and far from a paradox, it is exactly the quantity string theory later counted: extremal black holes were the first for which $S=A/4$ was reproduced from an explicit microstate ensemble.

What is the difference between the extremal wall and a naked singularity?

They are the two sides of one inequality. For Reissner-Nordström the discriminant $M^2-Q^2$ positive gives a black hole with two horizons, zero is the extremal wall with a single degenerate horizon, and negative removes the horizon entirely, exposing the singularity at $r=0$ . Extremality is the exact boundary $M^2=Q^2$ (or $M^2=a^2+Q^2$ in general). Weak cosmic censorship conjectures the naked case cannot form from regular initial data, making extremality a protective barrier: the third law forbids crossing it, and censorship forbids what lies beyond.

Why does the near-horizon geometry become AdS₂ × S²?

Because at extremality the metric function has a double zero, $f=(1-M/r)^2$ , so it vanishes quadratically at the horizon rather than linearly. Magnifying the region just outside with a scaling coordinate turns that quadratic vanishing into exactly the $-\rho^2\mathrm{d}\tau^2+\mathrm{d}\rho^2/\rho^2$ form of two-dimensional anti-de Sitter space, while the angular sphere keeps radius $M$ and factors off as an $S^2$ — the Bertotti-Robinson universe. The same quadratic zero stretches the throat to infinite proper length and gives it the emergent $SL(2,\mathbb{R})$ conformal symmetry behind Kerr/CFT; a non-extremal horizon, with its linear zero, has no such throat.

Can a real astrophysical black hole actually reach extremality?

No — and that impossibility is the third law made concrete. Astrophysical holes are neutralized almost instantly by ambient plasma, so charge extremality is irrelevant. They spin up through accretion, but the captured radiation carries away angular momentum increasingly efficiently as $a\to M$ , capping the spin at the Thorne limit $a_*\approx0.998$ . That ceiling is unattainability made visible: the closer to the wall, the harder each further increment of spin, so the last 0.2% is never delivered — several observed holes sit right there, near-extremal but never exactly extremal.

How is the third law of black hole mechanics different from the second?

They constrain different quantities. The second law is Hawking's area theorem: the total horizon area (hence $S=A/4$ ) never decreases in any classical process — the irreversibility of black-hole growth. The third law concerns the surface gravity, the analogue of temperature: it forbids reducing $\kappa$ to zero in a finite sequence of operations, so extremality is unattainable. The second law is proved rigorously under the null energy condition; the third law's rigorous form (Israel 1986) needs care because certain fine-tuned processes skirt it, which is why it reads as an unattainability principle rather than an absolute ban. Both mirror ordinary thermodynamics — the parallel Hawking radiation turned from analogy into physics.