A Complementary Third Law for Black Hole Thermodynamics

Yuan Yao,1 Meng-Shi Hou,1 and Yen Chin Ong1, 2, 3, l 1Center for Gravitation and Cosmology, College of Physical Science and Technology, Yangzhou University, 180 Siwangting Road, Yangzhou City, Jiangsu Province 225002, China 2School of Aeronautics and Astronautics, Shanghai Jiao Tong University, Shanghai 200240, China 3Nordita, KTH Royal Institute of Technology & Stockholm University, Roslagstullsbacken 23, SE-106 91 Stockholm, Sweden There are some examples in the literature, in which despite the fact that the underlying theory or model does not impose a lower bound on the size of black holes, the final temperature under Hawking evaporation is nevertheless finite and nonzero. We show that under some loose conditions, the black hole is necessarily an effective remnant, in the sense that its evaporation time is infinite. That is, the final state that there is nonzero finite temperature despite having no black hole remaining cannot be realized. We discuss the limitations, subtleties, and the implications of this result, which is reminiscent of the third law of black hole thermodynamics, but with the roles of temperature and size interchanged. We therefore refer to our result as the “complemetary third law” for black hole thermodynamics.

I. INTRODUCTION: THE ISSUE WITH From here onwards, we set ~ = c = G = kB = 1, unless TEMPERATURE OF BLACK HOLE REMNANTS when explicitly restored for clarity. Note that if α ∼ O(1), as is usually considered in theoretical calculation, then In the usual picture of Hawking evaporation, an asymp- the correction term becomes important at Planck scale. totically flat Schwarzschild black hole evaporates com- It has been argued that this leads to a correction in the pletely in finite time, although the time scale is extremely Hawking temperature, resulting in a black hole remnant long for a stellar mass black hole1. Since the Hawking [12]. In this scenario however, as the evaporation stops at temperature is inversely proportional to the mass, the some finite mass, the temperature also stops at a finite, black hole becomes hotter as it shrinks. Eventually the nonzero value, see Fig.(1) below. This is somewhat pe- energy scale becomes so high that new physics could po- culiar: a positive temperature seems to suggest that the tentially enter and affect the subsequent evolution. In black hole continues to emit particle, how then does the particular, novel quantum gravity effect may put a stop evaporation completely stop? A possible interpretation is on Hawking evaporation, thus resulting in a black hole as follows: since the remnant heat capacity vanishes, there “remnant”. is no thermodynamical interaction with its environment. The idea of a black hole remnant can be traced back to Therefore, it is thermodynamically inert and behaves like the work of Aharonov, Casher and Nussinov [1]. It has an elementary particle [13]. The finite remnant “temper- been suggested that black hole remnants could help to ature” should therefore be interpreted as energy of the ameliorate the black hole information paradox, though remnant (via E = kBT ). there are arguments against the very existence of rem- If one takes α < 0 in Eq.(1), we would find that there nants. See [2] for a comprehensive discussion of the debate is no lower bound for black hole mass, so in principle the and a review of various remnant scenarios. black hole can evaporate completely. However the final A popular way of obtaining a black hole remnant is temperature is finite and nonzero [14, 15], also see Fig.(1). via the generalized uncertainty principle (GUP), which Such a choice of sign of α may seem unusual, but it is incorporates the effect of gravity into the Heisenberg’s consistent with some quantum gravity models in which uncertainty principle. Since GUP arises from various physics at the Planck scale “classicalized” and becomes arXiv:1812.03136v2 [gr-qc] 22 Jun 2019 general considerations involving gravity and quantum deterministic (as the RHS of the GUP equation goes to mechanics, as well as string theory [3–11], it is usually zero when ∆p · c is equal to the Planck energy) [13–16]. treated as a phenomenological approach to study various Specifically, we have seen that a lattice “spacetime crystal” properties of quantum gravity. The simplest form of GUP gives rise to such a GUP [13]. Negative GUP parameter is given by is also required if one accepts that Wick-rotation can be applied to obtain GUP-corrected black hole temperature " 2 2 # from a Schwarzschild-like black hole (with higher order 1 αLp∆p ∆x∆p > ~ + . (1) terms) [17]. More recently, it has also been shown that 2 ~ non-commutative geometry [18], as well as corpuscular gravity, give rise to negative GUP parameter [19]. In addition, even without GUP, the situation that ∆x∆p ∼ 0 is compatible with other scenarios that have lElectronic address: [email protected] been proposed in the literature, such as the possibility 1 67 A solar mass non-rotating neutral black hole takes 10 years that is a dynamical field that flows to zero in high energy to evaporate, which far exceeds the current age of the Universe ~ ∼ 1010 years. limit [20, 21]; and with Planck mass fixed in 4-dimensions, 2

evaporate completely, so there is no need to resort to the aforementioned interpretation; the black hole simply con- tinues to evaporate indefinitely (indeed its heat capacity is always negative and so it interacts thermodynamically with the environment), with temperature asymptotes to a finite nonzero value. This value is T ∗ = 1/(4πp|α|) [14, 15] (note the typo in [15]). Thus even though black hole mass is not bounded below, the black hole behaves effectively as a meta-stable remnant at late times. See Fig.(2). This behavior is due to the fact that dM/ dt, though always negative, is not monotonic, and tends to zero as M → 0.

FIG. 1: The Hawking temperature of an asymptotically flat Schwarzschild black hole. The middle dashed curve corresponds to the usual picture of Hawking evaporation, which diverges as M → 0. The divergence is removed with GUP correction. Specifically,√ if α > 0, the temperature curve terminates at around M ∼ αMp, as shown by the right-most curve. If α < 0, however, GUP correction no longer imposes a lower bound on the black hole size. This corresponds to the left- most curve: the temperature remains finite as the black hole appears to shrink down to zero size.

Asymptotically Safe Gravity with G → 0 is equivalent to ~ → 0 [22] (see also a similar proposal in f(R) gravity FIG. 2: The Hawking temperature of an asymptotically flat [23]). Incidentally, although some of the very first GUP Schwarzschild black hole with α = −1, here in black dash-dotted curve, as a function of time, shows that the temperature tends to a constant scenarios came from string theory, in which α is naturally value. The mass of the black hole, in red solid curve, tends to zero positive, it is not clear that negative α is incompatible asymptotically. In order to display both curves in the same diagram, we have multiplied the Hawking temperature by a factor of 120, so that with string theory. For example, low energy limit of string the temperature curve tends to 120T ∗ = 120/4π ≈ 9.549. theory gives rise to charged dilaton black hole, with string coupling being weak near the singularity. Though we have no right to trust this solution near the singularity, as Indeed, the fact that the black hole lifetime is infinite Horowitz put it [24], it is tempting to speculate whether can be shown analytically [15]. The evolution equation is “contrary to the usual picture of large quantum fluctuations (a minus sign in missing in [15]): and spacetime foam near the singularity, quantum effects r !4 might actually be suppressed.” This would then be, at dM M 6 |α| = − 1 − 1 + . (2) least naively, compatible with the ∆x∆p ∼ 0 scenario at dt (4|α|π)4 M 2 high energy. This can be seen as a virtue of GUP as a phenomenologi- so as M becomes sufficiently small, we have cal tool: by taking different signs of α, it can accommodate different kinds of quantum gravity models. The question dM M 2 ∼ − , (3) is: How does one make sense of a nonzero black hole dt (4π)4α2 temperature if the black hole has completely evaporated? A possible interpretation is that this is the temperature which leads to of the Hawking radiation at the final moment just before the black hole disappears [14]. This parallels the expla-  256π4α2  M = M0 4 2 , (4) nation for the α > 0 case, but instead of interpreting the 256π α + M0t temperature as energy of the remnant (since there is no radiation), one now takes it to be the temperature of the where M0 is the “initial” (small) mass. final emission of radiation (since there is no black hole). This leads to a natural question: how generic is this An alternative interpretation as “vacuum fluctuation” is behavior of having an infinite evaporation time when the discussed further Sec.(II). final temperature is finite and nonzero? Are there any However, in [15], a more detailed study of the evap- condition required to ensure this? Black holes are known oration process reveals that the α < 0 GUP-corrected to behave like thermodynamical systems, in particular the black hole actually takes an infinite amount of time to third law states that a black hole (of course of nonzero 3 mass) cannot reach zero temperature state in finite num- it is the geometric optic cross section that goes into the ber of steps. Here we are claiming a complementary result: Stefan-Boltzmann law, but we also absorb this correction a black hole cannot reach a state with nonzero tempera- into C, as it will not affect the qualitative behavior of the ture but zero mass. This is the complementary third law solution. of black hole thermodynamics. In the following, we will first illustrate the theorem with We found that this behavior is in fact rather general, another concrete example, before proving the theorem for and can be stated as the general case. Unlike the GUP-corrected black hole studied in [14] discussed above, the following example is Theorem: Consider an n-dimensional neu- obtained from classical modified gravity (the Hawking tral static black hole spacetime, with areal ra- radiation itself is of course semi-classical). dius r, and horizon at r = rh. Assume that the Hawking temperature T and the black hole mass M are analytic functions of rh. Sup- n II. ANOTHER EXAMPLE: A BLACK HOLE pose dM/ dt = −CAT , where C > 0 is a REMNANT IN MASSIVE GRAVITY ∗ constant, and T → T ∈ (0, ∞) as rh → 0, then r → 0 only if t → ∞, provided that the h Note that the theorem is quite generic: it does not need k-th derivative M (k), for k < n − 1, do not the underlying theory to be general relativity. Here for all vanish when r = 0. h explicitness we show an example in the context of dRGT We have assumed that there is no problem with conver- (de Rham-Gabadadze-Tolley) massive gravity [34–37], in gence of the series expansion. In particular, since T and which graviton has nonzero mass. M are defined only in the domain [0, ∞), all the associ- A dyonic black hole solution in this theory was found ated limits and differentiability at 0 are to be understood in [38], with metric coefficient given by + as being one-sided (rh → 0 ). Note that Hawking tem- 2 2 2 perature of the usual kind T ∝ 1/M is not differentiable r 2m0 qE + qM 2 cc1 2  −gtt = k+ 2 − + 2 +m r + c c2 , (5) at rh = 0. l r r 2 This result reminded us of the (“Nernst version” of) third law of black hole thermodynamics: zero tempera- where m0 is related to the physical mass of the black ture (extremal) black hole, which is of nonzero size, is hole. Likewise, qE and qM are related to the electric unattainable in finite number of steps. Here we have the and magnetic charges, respectively. In addition, m is opposite scenario, zero mass/size2 black hole is unattain- essentially the graviton mass, c is a positive constant, able in finite time3 under Hawking evaporation if the while c1, c2 and k (the sectional curvature of the horizon) temperature is nonzero. We therefore refer to this the- can be positive or negative. The Hawking temperature orem as a “complementary third law”. See Sec.(IV) for for this black hole is [38] more discussions. 1  k 3r q2 + q2  c2c  Here we assume that the Stefan-Boltzmann law for ar- T = + + − E M + m2 cc + 2 . 4π r l2 r3 1 r bitrary spacetime dimension n > 4 holds during the entire + + + evolution [25, 26], which of course need not be the case; (6) see, e.g., [27–30], for a different viewpoint. Note that We emphasize that these black holes are “physical”, in we assumed that the Hawking evaporation is governed the sense that they have well-behaved thermodynamical by the simple Stefan-Boltzmann law only. This means, properties, as shown in [38]. Of course, dRGT massive for example, we do not study asymptotically de-Sitter gravity has two metric tensors: a dynamical one, gµν , spacetimes, in which Gibbons-Hawking temperature from and a fixed background fiducial metric fµν . The first and the cosmological horizon would contribute. Likewise, in foremost requirement (and the entire purpose of dRGT an asymptotically locally anti-de Sitter spacetime, the theory) is that the theory should be ghost-free. This usual reflective boundary condition would complicate the is not guaranteed for an arbitrary choice of the fiducial situation (see, however, Sec.(IV) for more discussions). A metric. In most applications in the context of holography, denotes the horizon area, with the constant C incorporat- a degenerate fiducial metric is chosen (as opposed to a ing the Boltzmann constant and the greybody factor. The Minkowski metric). Such metric was also applied in [38] effect of greybody factor, as well as the discreteness of the to obtain the black hole solution above. Such a choice of Hawking radiation (the sparsity [31–33]) can be ignored fµν has been demonstrated to also lead to a ghost-free since their effects would result in an even longer evapo- massive gravity theory [39]. Thus, the theory (and hence ration time [15]. In the geometric optic approximation, the solution) is also physical in the sense that there is no ghost. In the absence of magnetic charge, the authors showed that by tuning the various parameters such that 2 Zero mass and zero size are not always interchangeable, see 2 2 2 Sec.(IV). m c2c + k − ΦE = 0, (7) 3 A finite time is equivalent to a finite number of steps, with each Hawking particle emission counted as one “step”. where ΦE is the electric potential, one could have a solu- 4 tion with Hawking temperature of the form which yields, with C˜ = 2C/(4π)4,

2 ∗ m cc1 dr Z ε dr Z t T = 2rhP + , (8) h ˜ h ˜ 4π = −Crh =⇒ = −C dt, (11) dt rh(0) rh 0 where P = −Λ/(8π) is the pressure term in the extended ∗ black hole thermodynamics in an asymptotically locally where rh(0) is the initial horizon size, and t is the time anti-de Sitter spacetime. In the limit of vanishing horizon at which the horizon has shrunk to ε. Integerating yields 2 rh → 0, we see that 0 < T = m cc1/(4π) < ∞. The h i authors interpreted this as a fluctuation in the tempera- ε = r (0) exp −Ct˜ ∗ . (12) ture of the background spacetime after the black hole has h evaporated. That is to say, there is a trace of the black hole left behind if we look at the energy fluctuation of Therefore, in order to shrink to zero size, ε → 0, one must the vacuum. This is of course tiny because graviton mass have an infinite evaporation time t∗ → ∞. m is miniscule, although m of order unity was studied in [38] in the context of holography. (Of course, large AdS black holes with the usual reflective boundary condition will not evaporate in the first place; but see below.) The simplest way to achieve Eq.(7) is by considering the neutral case and set ΦE = 0. Take also Λ = 0 so that the pressure term vanishes. We choose k = 1 (this is enforced by a topological theorem in general relativity once Λ = 0 [40], but this may not be the case for massive gravity). 2 2 Since c2 can be negative, one can set c2 = −1/(m c ). This yields, surprisingly, a constant Hawking temperature 2 regardless of the black hole size, namely T ≡ m cc1/(4π), independent of rh. To further simplify the calculation, we set the numeri- cal values m = c = 1, so c2 = −1. We remind the readers FIG. 3: The evolution of the massive gravity black hole event horizon that our purpose is only to illustrate the aforementioned radius as function of time. Here we choose rh(0) = 1000. The black hole parameters are m = k = c = 1, c2 = −1 The black hole asymptotes theorem. It is possible that with this choice of the pa- to zero size as time goes to infinity. rameter values the black hole becomes unstable or other problems might arise4. The readers are referred to [38] for detailed study of the black hole solutions. (Massive grav- The caveat here is that in this particular example, we ity also admits a more conventional black hole remnant have assumed that the area that appears in the Stefan- that tends to zero temperature with finite size [45].) Boltzmann law is the horizon area. However due to the The physical mass (the mass that appears in the first metric function being −gtt = r/2 − 2M/r ∼ r/2 at large law of thermodynamics) is [38] r, the asymptotic structure is not flat. We know that in AdS, the effective emitting area of black holes with r  r2 q2 + q2 cc  genus g > 1 (k = −1, 0) is a constant (essentially the AdS M = h k + h + E M + m2 1 r + c2c , 2 l2 r2 2 h 2 length scale) and independent of the black hole mass [46]. (9) Thus, for our example with unusual asymptotic structure, which, with our choice of the parameter values, reduces a similar study along the line of [46] should be carried to M = r2 /4. In Fig.(3), we set the initial condition out to determine its effective emitting area, which might h not simply be the horizon area up to some factors. In rh = 1000, and obtain a plot which shows that the horizon tends to zero size only asymptotically. other words, our example above may not be correct, but 2 we use it to illustrate how the theorem would work if the The analytic proof is straightforward: with M = rh/4 and T = 1/(4π), we have emission surface is indeed the horizon area. We can also choose other values for the various parame- ters so that the black hole is asymptotically AdS-like, and dM r dr 1 h h 2 then modify the boundary conditions to allow the black = = −Crh 4 , (10) dt 2 dt (4π) hole to evaporate (see Sec.(IV) for details). An example with m = 1 = c, c2 = −1, k = l = 1 are shown in Fig.(4). Its evolution time is also infinite, but the rate of Hawking evaporation is different from the previous example with- 4 There are indications that dRGT gravity is problematic, since it is plagued with superluminal propagation. In addition, there out the cosmological constant term (the “pressure” term). exist arbitrarily small closed causal curves that result in a lack of The difference will become clear in the next section after well-posed Cauchy problem [41–44]. the theorem is proved. 5

Since M increases with r, M 0 > 0. In particular, M 0(0) > 0. So K := M 0(0)−1C = const. Consequently, Z ε Z t∗ 2−n ∗ n rh drh = −K(T ) dt, (18) rh(t0) t0

where t0  1 is the initial condition at a sufficiently late time where the series approximation is valid. Integrating yields 1 r (t )3−n − ε3−n
= K(T ∗)n(t − t∗). (19) n − 3 h 0 0 It is now clear that if ε → 0, t∗ must tend to infinity (since n > 4). In particular, in 4-dimensions, we get, 1 1 FIG. 4: The evolution of the black hole event horizon radius as function − = K(T ∗)4(t − t∗). (20) of time. Here we choose initial radius r (0) = 10. The black hole 0 h rh(t0) ε parameters are m = 1 = c, c2 = −1, l = 1. The black hole also asymptotes to zero size as time goes to infinity. This is the case for the GUP-corrected black hole in [15] with negative GUP parameter α, since rh = 2M, the same as the usual Schwarzschild black hole; c.f. Eq.(4). III. PROOF OF THE THEOREM If M 0(0) = 0, Eq.(16) would lead to, to lowest order of rh, the differential equation We now proceed to prove the theorem. As in the dr previous example, we will work with rh in place of M, h 00 −1 n−3 ∗ n = −M (0) Crh (T ) . (21) since M is an increasing function of rh. (Though this does dt not imply that rh = 0 ⇐⇒ M = 0). The dimensionality This leads to of spacetime is n 4. We assume that r is the areal > Z ε Z t∗ radius (if not, change to an appropriate coordinate system 3−n ˜ ∗ n n−2 rh drh = −K(T ) dt, (22) under which this is true), thus A ∝ rh . Absorb the rh(0) t0 proportional constant into C. It suffices to consider the late stages of the evolution. Assuming analyticity of where K˜ := M 00(0)−1C. Integrating yields the Hawking temperature, we can Taylor expand around 1 r = 0 to obtain r (t )4−n − ε4−n
= K˜ (T ∗)n(t − t∗). (23) h n − 4 h 0 0 T 00(0) T (r ) = T (0) +T 0(0)r + r2 + O(r3 ), (13) The evaporation time is clearly infinite for n > 5. In h h h h ∗ |{z} 2 4-dimensions, ε is exponential in −t . Thus, we again 6=0 obtain an infinite evaporation time. This is the case for the massive gravity black hole example illustrated in where T (0) = T ∗ ∈ (0, ∞) in the statement of the theorem, Fig.(3). and prime denotes derivative with respect to r . Similarly, h However, if both M 0(0) and M 00(0) vanish, and M 00(0) M 000(0) 6= 0, then the integration gives M(r ) = M(0) + M 0(0)r + r2 + O(r3 ). (14) h h 2 h h 1 5−n 5−n
∗ rh(t0) − ε = const.(t0 − t ). (24) Taking the derivative yields n − 5 Note that in 4-dimensions, the corresponding result is dM(rh)  0 00 2  drh = M (0) + M (0)rh + O(rh) . (15) ∗ 4 ∗ dt dt ε − rh(t0) = K˜ (T ) (t0 − t ) (25) Then, if M 0(0) 6= 0, we have for some constant K˜ . Therefore t∗ is now finite as ε → 0:

dr −1 ˜ ∗ 4 h  0 00 2  ∗ K(T ) t0 + rh(t0) = M (0) + M (0)rh + O(rh) t [ε = 0] = < ∞. (26) dt K˜ (T ∗)4  n−2 ∗ n n−1  · −Crh (T ) + O(rh ) 0 −1 n−2 ∗ n n−1 In n > 5 the evaporation time is still infinite. = −M (0) Crh (T ) + O(rh ). (16) Indeed, in general, one can see that as long as the lowest order of nonzero M (k)(0) is k = n − 1, the evaporation To lowest order in r , the differential equation is h time will be finite. This completes the proof. In particu- lar, this implies that in 4-dimensions, the evaporation is drh n−2 = −M 0(0)−1Cr (T ∗)n. (17) 0 00 dt h infinite if M (0) and M (0) do not both vanish. 6

IV. THE COMPLEMENTARY THIRD LAW: gauge fields, more analysis would be required to study APPLICABILITY AND SUBTLETIES whether the complementary third law holds, since the evo- lution under Hawking evaporation would be considerably To summarize our findings so far: in this work, we inves- more complicated. Even for asymptotically flat Reissner- tigated the conditions for a black hole to have “left-over” Nordstr¨omblack holes, Hiscock and Weems showed that nonzero and finite temperature at the end of Hawking there are charge loss and mass loss regimes, so the ratio evaporation, at which point the black hole shrinks to zero Q/M is not necessarily monotonic in time (the evolution size. To our knowledge (and that of the authors of [38]), is governed by coupled differential equations in certain the massive gravity black hole discussed in Sec.(II) is the range of the parameters) [50]. only known example in classical modified gravity with However, one special charged black hole solution – the such a property. The GUP corrected black hole studied in GHS black hole (see below) – is worth a separate mention, [14–16] provided another example from a quantum grav- since it gives us the opportunity to point out that the itational correction. This result parallels the third law complementary third law is really stating that the final of black hole thermodynamics in which (nonzero mass) state with zero mass but nonzero temperature cannot be extremal black hole with zero temperature cannot be attained, not zero size. Usually zero mass also coincides attained in finite time, but with the role of mass and with zero size, but this is not always the case, and therefore temperature reversed, we therefore dubbed it the “com- the distinction is important. plementary third law of black hole thermodynamics”. The charged dilatonic “GHS” (Garfinkle-Horowitz- To be more specific, we found that if the first and Strominger) black hole [51–53], obtained in a low en- the second derivative of M(rh) do not both vanish at ergy limit of string theory, with metric tensor in a rh = 0, then the evaporation time in 4-dimensions is Schwarzschild-like coordinate system {t, r, θ, ϕ}: actually infinite (analogously in higher dimensions), and  2M   2M −1 so the black hole behaves as an effective, meta-stable ds2 = − 1 − dt2 + 1 − dr2 remnant. This result does not assume the underlying r r theory to be general relativity. It is simply a consequence  Q2  + r r − (dθ2 + sin2 θ dϕ2). (27) of the mathematical properties of the Stefan-Boltzman M differential equation. Our proof relies on the assumption that M(rh) and T (rh) are both real analytic functions, The Hawking temperature is always T = 1/(8πM) in- which are infinitely differentiable at rh = 0. Although dependent of the charge, even in the zero size limit. most calculations in physics literature make use of Taylor Note that the horizon stays fixed at r = 2M, but r expansion almost ubiquitously, this assumption might be is not an areal radius, so one has to transform via too strong. Perhaps one could relax it and the theorem r2 7→ R2 := r(r − Q2/M) to a coordinate {t, R, θ, ϕ} would still remain true. An alternative proof without the p 2 2 in which the areal radius is R = 4M −√2Q , which use of series expansion would be welcomed so the issues goes to zero in the extremal limit |Q| = 2M, with of convergence can be avoided altogether. Now we shall nonzero temperature (the temperature is not affected un- discuss a few more aspects of the complementary third der this change of coordinate). This is very different from law. extremal Reissner-Nordstr¨omblack hole which has zero temperature. A. Charged and Dilaton Black Holes: Despite dM/ dt being independent of Q, charge loss Nonzero Mass vs. Nonzero Size can occur from spontaneous charge particle emission `ala Schwinger, and as shown by Hiscock and Weems [50], this It is worth emphasizing that this phenomenon is some- may affect the evolution under Hawking evaporation just what opposite to that of the third law of black hole ther- like in the Reissner-Nordstr¨omcase. Thus our theorem modynamics, in which the black hole size remains finite does not strictly apply. (and nonzero) and T = 0 cannot be achieved; whereas Nevertheless, if we ignore these subtleties6, and consider here the temperature remains finite (and nonzero) and the mass loss of GHS black hole to be governed only rh = 0 cannot be achieved. The main difference is that our theorem only concerns neutral black holes, whereas for the third law, it applies to charged and rotating black holes (the only way to get zero temperature in general modified gravity theories, e.g., asymptotically safe gravity with relativity5). In the presence of electrical charges and other higher derivative terms [48] and in conformal (Weyl) gravity [49] (note that entropy vanishes does not always imply zero area for modified gravity black holes). The usual third law applies to these black holes – they have infinite lifetime. 6 However, since T does not depend on Q here, as opposed to the 5 One could have a zero temperature black hole in the zero size Reissner-Nordstr¨omcase, the Schwinger process is not expected limit, if, for example, higher order curvature terms are included to affect the result by much, at least for large enough M, in in the action [47]. It is also possible to obtain zero temperature the regime where it is suppressed [54]. This requires a further black hole at some nonzero mass without any gauge field in investigation beyond the scope of the current work, and will be addressed elsewhere. 7

n-dimensional spacetime, is given by [60]

1  2 2 T = 2 (n − 1)rh + (n − 3)kl , (28) 4πl rh

which either goes to zero (for k = 0) or diverges (for k = 1) as rh → 0. (For k = −1 case, the black hole tends to a minimum size as the temperature goes to zero.) [59] In addition, for k ∈ {−1, 0}, the effective emitting surface area of the black hole is independent of the mass, and is completely fixed by the cosmological constant [46]. The complementary third law therefore does not apply to AdS black holes, at least the simplest ones addressed here.

C. Comparison with Conventional FIG. 5: The evolution of the areal radius (blue dashed curve) and the mass (red solid curve) of a charged dilatonic GHS black hole as Thermodynamics functions of time, assuming that there is no charge loss. Here we choose initial mass to be M(0) =√ 10 and Q=1. The mass thus asymptotes to its lower bound M = Q/ 2 ≈ 0.7071. The evolution takes an infinite It came at a surprise when Bardeen, Carter and Hawk- amount of time. ing [61] discovered that black holes satisfy properties that are analogous to thermodynamics. Subsequent discovery that black holes do radiate when quantum mechanics is by Stefan-Boltzmann law, then our theorem does apply. taken into account, established that black holes are ther- Numerically the results are shown in Fig.(5). Note the modynamical system. In particular, black holes satisfy a peculiarity that R → 0 but M tends to a finite value, as form of third law: zero temperature configuration cannot the black hole approaches a null singularity of zero size, be reached in a finite number of step. This parallels the which somehow supports the mass. “Nernst version” of third law in conventional thermody- namics. What about the complementary third law? Is there an analogous phenomenon in other thermodynami- cal system? B. Non-Examples of Complementary Third Law: By definition, in classical thermodynamics, the temper- Anti-de Sitter Black Holes ature is related to the entropy by 1/T = (∂S/∂U)|V,N , where U is the internal energy of the system, and the It might be illuminating to also discuss here some non- volume V , as well as particle number N, are held fixed. examples that do not satisfy the premise of the com- It would seem that nonzero temperature (with nonzero plementary third law (note that this is not the same internal energy) always corresponds to nonzero entropy. as “violating” the law). In view of the popularity of In the context of black holes, nonzero entropy means holography, asymptotically locally AdS spacetimes are nonzero area. an important class of solutions. As we mentioned in the The complementary third law essentially says that black Introduction, we assumed that the evolution is governed holes with zero entropy yet with nonzero temperature by the simple Stefan-Boltzmann law. In asymptotically cannot be attained. This is therefore consistent with locally AdS spacetimes, our analysis can be carried over conventional thermodynamics. by replacing the usual reflective boundary condition with Note that if we consider quantizing the system, the an absorbing boundary condition, which allows large black entropy can be zero for small temperature if the first holes to evaporate (see, e.g., [55–57]). In holography, this excited state of the system has energy higher than kBT . can be done by coupling the boundary field theory with This does not concern us since black hole thermodynamics an auxiliary system (“AUX” [56]), such as another field is analogous to classical ordinary thermodynamics. theory. It was previously found that regardless of their horizon topologies, neutral AdS black holes in such spacetimes V. DISCUSSIONS: SOME REMAINING take about the same amount of time to evaporate down PUZZLES to the same size of order l, the AdS length scale [58, 59]. For positively curved (k = 1) case, the black hole take The question remains for the case M 0(0) = M 00(0) about the same amount of time to completely evaporate in 4-dimensions (and analogously in higher dimensions). regardless of its initial mass [58, 59]. For flat (k = 0) case, Black holes that satisfy this property will evaporate in a the evaporation time is infinite. finite time, and leave behind a nonzero finite temperature. Nevertheless these black holes do not satisfy the premise What is the correct interpretation for such a temperature? of the theorems. This is because their temperature, in Is it the temperature of the last bit of radiation, or an 8 energy fluctuation of the ambient spacetime (now without restricted than the standard third law in two ways: firstly, a black hole – so it would be a kind of “vacuum memory”)? we only study Hawking evaporation, not other physical Presumably if the final temperature is very low, the processes. We restricted our study to the static case in second interpretation (to our knowledge, first proposed in n > 4 spacetime dimensions, and only to neutral black [38]) is plausible, but what if the temperature is “high”, holes (though the result might hold in more general cases). say 1◦C (note that the theorem does not constrain the Secondly, if we expresse the black hole mass M as a ∗ value of T other that it is nonzero and finite), how can function of its horizon M(rh), the complementary third this be a “fluctuation” of the vacuum? It would be good law can be violated if the derivatives M (k)(0) = 0 for to have an explicit black hole solution of this type for a k < n − 1, but then the standard third law may also detailed study, if it exists. be violated under certain circumstances [67–70]. In a Let us speculate on a possibility: since black holes way, the shortcoming of the complementary third law is contain an enormous volume [62, 63] (which does not a virtue since it gives a clear condition for its violation. decrease even as the black hole evaporates [64–66]), the A further study into these conditions and their physical end state of the evolution could be a baby universe that interpretations could yield a deeper understanding into pinches off from the original universe. Perhaps such a black hole thermodynamics. pinch-off leaves a finite temperature signature behind in the parent universe. On the other hand, we could turn this around and say that, if the complementary third law is true generically, then maybe there is no 4-dimensional black hole solution that would satisfy M 0(0) = M 00(0), Acknowledgments and similarly for higher dimensions. Another issue concerns the singularity of the black YCO thanks the National Natural Science Foundation hole. As black hole evaporates, does it leave behind a of China (grant No.11705162) and the Natural Science naked singularity? In the usual Schwarzschild case, since Foundation of Jiangsu Province (No.BK20170479) for temperature becomes extremely high, one may invoke funding support. YCO also thanks Nordita, where the new physics and hope that the even if singularity wasn’t initial phase of this work was carried out, for hospitality already cured by quantum gravity in the first case, will during his summer visit while participating at the Lambda evaporate away together with the horizon. However, if Program. He thanks Ingemar Bengtsson for related discus- the final temperature remains mild, it leaves open the sion and Brett McInnes for some suggestions. The authors possibility that naked singularity may form, thus violating thank members of Center for Gravitation and Cosmology the cosmic censorship conjecture. (CGC) of Yangzhou University (http://www.cgc-yzu.cn) To summarize, the complementary third law is more for various discussions and supports.

[1] Yakir Aharonov, Aharon Casher, Shmuel Nussinov, “The “Can Spacetime be Probed Below the String Size?”, Phys. Unitarity Puzzle and Planck Mass Stable Particles”, Phys. Lett. B 216 (1989) 41. Lett. B 191 (1987) 51. [10] Kenichi Konishi, Giampiero Paffuti, Paolo Provero , “Min- [2] Pisin Chen, Yen Chin Ong, Dong-han Yeom, “Black Hole imum Physical Length and the Generalized Uncertainty Remnants and the Information Loss Paradox”, Phys. Rept. Principle in String Theory”, Phys. Lett. B 234 (1990) 603 (2015) 1,[arXiv:1412.8366 [gr-qc]]. 276. [3] Michele Maggiore, “A Generalized Uncertainty Princi- [11] Edward Witten, “Reflections on the Fate of Spacetime”, ple in Quantum Gravity”, Phys. Lett. B 304 (1993) 65, Phys. Today 49 (1996) 24. [arXiv:hep-th/9301067]. [12] Ronald J. Adler, Pisin Chen, David I. Santiago, “The Gen- [4] Michele Maggiore, “Quantum Groups, Gravity, and the eralized Uncertainty Principle and Black Hole Remnants”, Generalized Uncertainty Principle”, Phys. Rev. D 49 Gen. Rel. Grav. 33 (2001) 2101,[arXiv:gr-qc/0106080]. (1994) 5182,[arXiv:hep-th/9305163]. [13] Petr Jizba, Hagen Kleinert, Fabio Scardigli, “Uncer- [5] Fabio Scardigli, “Generalized Uncertainty Principle in tainty Relation on World Crystal and its Applications Quantum Gravity from Micro-Black Hole Gedanken Ex- to Micro Black Holes”, Phys. Rev. D 81 (2010) 084030, periment”, Phys. Lett. B 452 (1999) 39,[arXiv:hep- [arXiv:0912.2253 [hep-th]]. th/9904025]. [14] Yen Chin Ong, “Generalized Uncertainty Principle, Black [6] Ronald J. Adler, David I. Santiago, “On Gravity and the Holes, and White Dwarfs: A Tale of Two Infinities”, Uncertainty Principle”, Mod. Phys. Lett. A 14 (1999) JCAP 09 (2018) 015,[arXiv:1804.05176 [gr-qc]]. 1371,[arXiv:gr-qc/9904026]. [15] Yen Chin Ong, “An Effective Black Hole Remnant via [7] Gabriele Veneziano, “A Stringy Nature Needs Just Two Infinite Evaporation Time due to Generalized Uncertainty Constants”, Europhys. Lett. 2 (1986) 199. Principle”, JHEP 10 (2018) 195,[arXiv:1806.03691 [gr- [8] David J. Gross, Paul F. Mende, “String Theory Beyond qc]]. the Planck Scale”, Nucl. Phys. B 303 (1988) 407. [16] Bernard J. Carr, Jonas Mureika, Piero Nicolini, “Sub- [9] Daniele Amati, Marcello Ciafolini, Gabriele Veneziano, Planckian Black Holes and the Generalized Uncertainty 9

Principle”, JHEP 07 (2015) 052,[arXiv:1504.07637 [gr- [37] Claudia de Rham, “Massive Gravity”, Living Rev. Relativ. qc]]. 17 (2014) 7,[arXiv:1401.4173 [hep-th]]. [17] Fabio Scardigli, Roberto Casadio, Gravitational tests of [38] S. H. Hendi, N. Riazi, S. Panahiyan, “Holographical the Generalized Uncertainty Principle, Eur. Phys. J. C Aspects of Dyonic Black Holes: Massive Gravity Gen- 75 (2015) 425, arXiv:1407.0113 [hep-th]. eralization”, Ann. Phys. (Berlin) 530 (2018) 1700211, [18] T. Kanazawa, G. Lambiase, G. Vilasi, A. Yoshioka, “Non- [arXiv:1610.01505 [hep-th]]. commutative Schwarzschild Geometry and Generalized [39] Hongsheng Zhang, Xin-Zhou Li, “Ghost Free Massive Uncertainty Principle”, Eur. Phys. J. C 79 (2019) 95. Gravity with Singular Reference Metrics”, Phys. Rev. D [19] Luca Buoninfante, Giuseppe Gaetano Luciano, Luciano 93 (2016) 124039,[arXiv:1510.03204 [gr-qc]]. Petruzziello, “Generalized Uncertainty Principle and Cor- [40] Stephen W. Hawking, “Black Holes in General Relativity”, puscular Gravity”,[arXiv:1903.01382 [gr-qc]]. Commun. Math. Phys. 25 (1972) 152. [20] Sabine Hossenfelder, “Gravity Can be Neither Classical [41] Stanley Deser, Keisuke Izumi, Yen Chin Ong, Andrew nor Quantized”, In: Aguirre A., Foster B., Merali Z. (eds) Waldron, “Massive Gravity Acausality Redux”, Phys. Questioning the Foundations of Physics. The Frontiers Lett. B 726 (2013) 544,[arXiv:1306.5457 [hep-th]]. Collection. Springer, Cham,[arXiv:1212.0454 [gr-qc]]. [42] Stanley Deser, Keisuke Izumi, Yen Chin Ong, Andrew [21] Sabine Hossenfelder, “A Possibility to Solve the Problems Waldron, “Problems of Massive Gravities”, Mod. Phys. with Quantizing Gravity”, Phys. Lett. B 725 (2013) 473, Lett. A 30 (2015) 1540006,[arXiv:1410.2289 [hep-th]]. [arXiv:1208.5874 [gr-qc]]. [43] Pavel Motloch, Wayne Hu, Austin Joyce, Hayato Moto- [22] Max Niedermaier, “The Asymptotic Safety Scenario in hashi, “Self-Accelerating Massive Gravity: Superluminal- Quantum Gravity – An Introduction”, Class. Quant. Grav. ity, Cauchy Surfaces and Strong Coupling”, Phys. Rev. 24 (2007) R171,[arXiv:gr-qc/0610018]. D 92 (2015) 044024,[arXiv:1505.03518 [hep-th]]. [23] Brian Greene, Kurt Hinterbichler, Simon Judes, Maulik K. [44] Pavel Motloch, Wayne Hu, Hayato Motohashi, “Self- Parikh, “Smooth Initial Conditions from Weak Gravity”, accelerating Massive Gravity: Hidden Constraints and Phys. Lett. B 697 (2011) 178,[arXiv:0911.0693 [hep-th]]. Characteristics”, Phys. Rev. D 93 (2016) 104026, [24] Gary T. Horowitz, “The Dark Side of String Theory: [arXiv:1603.03423 [hep-th]]. Black Holes and Black Strings”,[arXiv:hep-th/9210119]. [45] Behzad Eslam Panah, Seyed Hossein Hendi, Yen [25] P. T. Landsberg, A De Vos, “The Stefan-Boltzmann Con- Chin Ong, “Black Hole Remnant in Massive Gravity”, stant in n-Dimensional Space”, Jour. of Phys. A 22 (1989) [arXiv:1808.07829 [gr-qc]]. 1073. [46] Dietmar Klemm, Luciano Vanzo, “Quantum Properties of [26] Tatiana R. Cardoso, Antonio S. de Castro, “The Black- Topological Black Holes”, Phys. Rev. D 58 (1998) 104025, body Radiation in D-Dimensional Universes”, Rev. Bras. [arXiv:gr-qc/9803061]. Ens. Fis. 27 (2005) 559,[arXiv:quant-ph/0510002]. [47] Pablo Bueno, Pablo A. Cano, “Universal Black Hole [27] Roberto Casadio, Benjamin Harms, “Black Hole Evapo- Stability in Four Dimensions”, Phys. Rev. D 96 (2017) ration and Large Extra Dimensions”, Phys. Lett. B 487 024034,[arXiv:1704.02967 [hep-th]]. (2000) 209,[arXiv:hep-th/0004004]. [48] Yi-Fu Cai, Damien A. Easson, “Black Holes in an Asymp- [28] Roberto Casadio, Benjamn Harms, Yvan Leblanc, “Mi- totically Safe Gravity Theory with Higher Derivatives”, crofield Dynamics of Black Holes”, Phys. Rev. D 58 (1998) JCAP 1009 (2010) 002,[arXiv:1007.1317 [hep-th]]. 044014,[arXiv:gr-qc/9712017]. [49] Hao Xu, Man-Hong Yung, “Black Hole Evaporation in [29] Roberto Casadio, Benjamin Harms, “Microcanonical De- Conformal (Weyl) Gravity”,[arXiv:1811.07309 [gr-qc]]. scription of (Micro) Black Holes”, Entropy 13 (2011) 502, [50] William A. Hiscock, Lance D. Weems, “Evolution of [arXiv:1101.1384 [hep-th]]. Charged Evaporating Black Holes”, Phys. Rev. D 41 [30] Sabine Hossenfelder, “What Black Holes Can Teach Us”, (1990) 1142. Focus on Black Hole Research, pp. 155-192, Nova Science [51] David Garfinkle, Gary T. Horowitz, Andrew Strominger, Publishers (2005),[arXiv:hep-ph/0412265]. “Charged Black Holes in String Theory”, Phys. Rev. D 43 [31] Finnian Gray, Sebastian Schuster, Alexander Van-Brunt, (1991) 31403143. Matt Visser, “The Hawking Cascade from a Black Hole Is [52] Gary W. Gibbons, “Antigravitating Black Hole Solutions Extremely Sparse”, Class. Quant. Grav. 33 (2016) 115003, with Scalar Hair in N = 4 Supergravity”, Nucl. Phys. B [arXiv:1506.03975 [gr-qc]]. 207 (1982) 337. [32] Matt Visser, Finnian Gray, Sebastian Schuster, Alexander [53] Gary W. Gibbons, Kei-ichi Maeda, “Black Holes and Van-Brunt, “Sparsity of the Hawking Flux”, in Proceed- Membranes in Higher Dimensional Theories with Dilaton ings of the MG14 Meeting on General Relativity (2017); Fields”, Nucl. Phys. B 298 (1988) 741. pp. 1724-1729,[arXiv:1512.05809 [gr-qc]]. [54] Gary W. Gibbons, “Vacuum Polarization and the Spon- [33] Wolfgang M¨uck, “Hawking Radiation is Corpuscular”, taneous Loss of Charge by Black Holes”, Comm. Math. Eur. Phys. J. C 76 (2016) 374,[arXiv:1606.01790 [hep- Phys. 44 (1975) 245. th]]. [55] Jorge V. Rocha, “Evaporation of Large Black Holes in [34] Claudia de Rham, Gregory Gabadadze, “Generalization of AdS: Coupling to the Evaporon”, JHEP 08 (2008) 075, the Fierz-Pauli Action”, Phys. Rev. D 82 (2010) 044020, [arXiv:0804.0055 [hep-th]]. [arXiv:1007.0443 [hep-th]]. [56] Ahmed Almheiri, Donald Marolf, Joseph Polchinski, Dou- [35] Claudia de Rham, Gregory Gabadadze, Andrew J. Tolley, glas Stanford, James Sully, “An Apologia for Firewalls”, “Resummation of Massive Gravity”, Phys. Rev. Lett. 106 JHEP 09 (2013) 018,[arXiv:1304.6483 [hep-th]]. (2011) 231101,[arXiv:1011.1232 [hep-th]]. [57] Mark Van Raamsdonk, “Evaporating Firewalls”, JHEP [36] S. F. Hassan, Rachel A. Rosen, “On Non-Linear Ac- 11 (2014) 038,[arXiv:1307.1796 [hep-th]]. tions for Massive Gravity”, JHEP 07 (2011) 009, [58] Don N. Page, “Finite Upper Bound for the Hawking [arXiv:1103.6055 [hep-th]]. Decay Time of an Arbitrarily Large Black Hole in Anti- 10

de Sitter Spacetime”, Phys. Rev. D 97 (2018) 024004, [65] Yen Chin Ong, “The Persistence of the Large Volumes [arXiv:1507.02682 [hep-th]]. in Black Holes”, Gen. Relativ. Gravit. 47 (2015) 88, [59] Yen Chin Ong, “Hawking Evaporation Time Scale of [arXiv:1503.08245 [gr-qc]]. Topological Black Holes in Anti-de Sitter Spacetime”, [66] Marios Christodoulou, Tommaso De Lorenzo, “On the Nucl. Phys. B 903 (2016) 387,[arXiv:1507.07845 [gr-qc]]. Volume Inside Old Black Holes”, Phys. Rev. D 94 (2016) [60] Danny Birmingham, “Topological Black Holes in Anti- 104002,[arXiv:1604.07222 [gr-qc]]. de Sitter Space”, Class. Quant. Grav. 16 (1999) 1197, [67] Paul C. W. Davies, “Thermodynamics of Black Holes”, [arXiv:hep-th/9808032]. Rep. Prog. Phys. 41 (1978) 1313. [61] James M. Bardeen, Brandon Carter, Stephen W. Hawking, [68] Charles J. Farrugia, Petr H´aj´ıˇcek, “The Third Law of “The Four Laws of Black Hole Mechanics”, Comm. Math. Black Hole Mechanics: A Counterexample”, Commun. Phys. 31 (1973) 161. Math. Phys. 68 (1979) 291. [62] Marios Christodoulou, Carlo Rovelli, “How Big is a Black [69] Mieczysaw Pr´oszy´nsk, “Thin Charged Shells and the Hole?”, Phys. Rev. D 91 (2015) 064046,[arXiv:1411.2854 Violation of the Third Law of Black Hole Mechanics”, [gr-qc]]. Gen Relat. Gravit. 15 (1983) 403. [63] Yen Chin Ong, “Never Judge a Black Hole by Its Area”, [70] Takashi Torii, “Violation of the Third Law of Black Hole JCAP 04 (2015) 003,[arXiv:1503.01092 [gr-qc]]. Thermodynamics in Higher Curvature Gravity”, Entropy [64] Renaud Parentani, Tsvi Piran, “The Internal Geometry 14 (2012) 229. of an Evaporating Black Hole”, Phys. Rev. Lett. 73 (1994) 2805