CALT-68-1774 DOE RESEARCH AND DEVELOPMENT REPORT Evaporation of Two Dimensional Black Holes S. W. Hawking California Institute of Technology, Pasadena, CA 91125 and Department of Applied Mathematics and Theoretical Physics University of Cambridge Silver Street Cambridge CB3 9EW, UK Abstract Callan, Giddings, Harvey and Strominger have proposed an interesting two di- mensional model theory that allows one to consider black hole evaporation in the semi-classical approximation. They originally hoped the black hole would evaporate completely without a singularity. However it has been shown that the semi-classical equations will give a singularity where the dilaton field reaches a certain critical value. Initially, it seems this singularity will be hidden inside a black hole. However, as the evaporation proceeds, the dilaton field on the horizon will approach the critical value but the temperature and rate of emission will remain finite. These results indicate either that there is a naked singularity, or (more likely) that the semi-classical ap- proximation breaks down when the dilaton field approaches the critical value. Work supported in part by the U.S. Dept. of Energy under Contract no. DEAC-03-81ER40050. arXiv:hep-th/9203052 18 Mar 92 February 1992 2 Introduction Callan, Giddings, Harvey and Strominger (CGHS) [1] have suggested an interest- ing two dimensional theory with a metric coupled to a dilaton field and N minimal scalar fields. The Lagrangian is N " 1 1 L = -g[e-2Ć(R + 4("Ć)2 + 42) - ("fi)2], 2Ä„ 2 i=1 If one writes the metric in the form ds2 = e2Ádx+dx- the classical field equations are "+"-fi = 0, 2 2"+"-Ć - 2"+Ć"-Ć - e2Á = "+"-Á, 2 2 "+"-Ć - 2"+Ć"-Ć - e2Á = 0. 2 These equations have a solution Ć = -b log(-x+x-) - c - log 1 2b Á = - log(-x+x-) + log 2 where b and c are constants and b can be taken to be positive without loss of generality. 1 A change of coordinates 2b 1 uÄ… = Ä… log(Ä…xÄ…) Ä… (c + log )
gives a flat metric and a linear dilaton field Á = 0
Ć = - (u+ - u-) 2 This solution is known as the linear dilaton. The solution is independent of the constants b and c which correspond to freedom in the choice of coordinates. Normally 1 b is taken to have the value . 2 These equations also admit a solution 1 Ć = Á - c = - log(M-1 - e2cx+x-) 2 . This represents a two dimensional black hole with horizons at xÄ… = 0 and singu- larities at x+x- = M-2e-2c. Note that there is still freedom to shift the Á field on the horizon by a constant and compensate by rescaling the coordinates xÄ…, but there s nothing corresponding to the freedom to choose the constant b. In terms of 1 the coordinates uÄ… defined as before with b = 2 1 +-u-) Á = - log(1 - M-1e-(u ) 2 1 +-u-) Ć = - (u+ - u-) - log(1 - M-1e-(u ) 2 2 This black hole solution is periodic in the imaginary time with period 2Ä„-1. One would therefore expect it to have a temperature
T = 2Ä„ and to emit thermal radiation [2]. This is confirmed by CGHS. They considered a black hole formed by sending in a thin shock wave of one of the fi fields from the weak 2 coupling region (large negative Ć) region of the linear dilaton. One can calculate the energy momentum tensors of the fi fields, using the conservation and trace anomaly equations. If one imposes the boundary condition that there is no incoming energy momentum apart from the shock wave, one finds that at late retarded times u- there a steady flow of energy in each fi field at the mass independent rate 2 48 If this radiation continued indefinitely, the black hole would radiate an infinite amount of energy, which seems absurd. One might therefore expect that the back reaction would modify the emission and cause it to stop when the black hole had radiated away its initial mass. A fully quantum treatment of the back reaction seem very difficult even in this two dimensional theory. But CGHS suggested that in the limit of a large number N of scalar fields fi, one could neglect the quantum fluctuations of the dilaton and the metric and treat the back reaction of the radiation in the fi fields semi-classically by adding to the action a trace anomaly term N "+"-Á. 12 The evolution equations that result from this action are N "+"-Ć = (1 - e2Ć)"+"-Á, 24 N N 2(1 - e2Ć)"+"-Ć = (1 - e2Ć)(4"+Ć"-Ć + 2e2Á). 12 24 In addition there are two equations that can be regarded as constraints on the data on characteristic surfaces of constant xÄ… N 2 2 ("+Ć - 2"+Á"+Ć) = e2Ć("+Á - "+Á"+Á - t+(x+)), 24 N 2 2 ("-Ć - 2"-Á"-Ć) = e2Ć("-Á - "-Á"-Á - t-(x-)), 24 where tÄ…(xÄ…) are determined by the boundary conditions in a manner that will be 3 explained later. Even these semi-classical equations seem too difficult to solve in closed form. CGHS suggested that a black hole formed from an f wave would evaporate completely without there being any singularity. The solution would approach the linear dilaton at late retarded times u- and there would be no horizons. They therefore claimed that there would be no loss of quantum coherence in the formation and evaporation of a two dimensional black hole: the radiation would be in a pure quantum state, rather than in a mixed state. In [3] and [4] it was shown that this scenario could not be correct. The solution would develop a singularity on the incoming f wave at the point where the dilaton field reached the critical value N Ć0 = -1 log 2 12 This singularity will be spacelike near the f wave [4]. Thus at least part of the final quantum state will end up on the singularity, which implies that the radiation at infinity in the weak coupling region will not be in a pure quantum state. The outstanding question is: How does the spacetime evolve to the future of the f wave? There seem to be two main possibilities: 1 The singularity remains hidden behind an event horizon. One can continue an infinite distance into the future on a line of constant Ć < Ć0 without ever seeing the singularity. If this were the case, the rate of radiation would have to go to zero. 2 The singularity is naked. That is, it is visible from a line of constant Ć at a finite time to the future of the f wave. Any evolution of the solution after this would not be uniquely determined by the semi- classical equations and the initial data. Indeed, it is likely that the point at which the singularity became visible was itself singular and that the solution could not be evolved to the future for more than a finite time. 4 In what follows I shall present evidence that suggests the semi-classical equations lead to possibility 2. This probably indicates that the semi- classical approximation breaks down as the dilaton field on the horizon approaches the critical value. Static Black Holes If the solution were to evolve without a naked singularity, it would presumably approach a static state in which a singularity was hidden behind an event horizon. This motivates a study a study of static black hole solutions of the semi-classical equations. One could look for solutions in which Ć and Á depended only on a radial variable à = x+ - x- but this has the disadvantage that the black hole horizon is at à = - inf. Instead it seems better to define the radial coordinate to be r2 = -x+x- The horizon is then at r = 0 and the field equations for a static solution are: 1 N 1 Ć + Ć = 1 - e2Ć Á + Á r 24 r N 1 N 1 - e2Ć Ć + Ć = 2 1 - e2Ć (Ć )2 - 2e2Á 12 r 24 The boundary conditions for a regular horizon are Ć = Á = 0 A static black hole solution is therefore determined by the values of Ć and Á on the horizon. The value of Á however can be changed by a constant by rescaling the coordinates xÄ…. The physical distinct static solutions with a horizon are therefore characterized simply by Ćh, the value of the dilaton on the horizon. If Ćh > Ć0, Ć would increase away from the horizon and would always be greater than its horizon value. This shows that to get a static black hole solution that is 5 asymptotic to the weak coupling region of the linear dilaton, Ćh must be less than the critical value Ć0. One can then show that both Ć and Á must decrease with increasing r. This means the back reaction terms proportional to N will become unimportant. For large r one can therefore approximate by putting N = 0. This gives Ć = Á - (2b - 1) log r - c 1 Ć + Ć = 2((Á - (2b - 1)r-1)2 - 2e2Á) r Asymptotically these have the solution 2b K + L log r Á = - log r + log - + ... r4b where b, c, K, L are parameters that determine the solution. The parameters band c correspond to the coordinate freedom in the linear dilaton that the solution ap- proaches at large r. The parameter L does not appear in the black hole solutions. If it is zero, the parameter K can be related to the ADM mass M of the solution. The effects of the back reaction terms proportional to N will affect only the higher order terms in r-1. For Ćh << Ć0, the back reaction terms will be small at all values of r and the solutions of the semi-classical equations will be almost the same as the classical black holes. So M Ć0 = -1 log 2
Consider a situation in which a black hole of large mass (M >> N/12) is created by sending in an f wave. One could approximate the subsequent evolution by a sequence of static black hole solutions with a steadily increasing value of Ć on the horizon. However, when the value of Ć on the horizon approaches the critical value 6 Ć0, the back reaction will become important and will change the black hole solutions solutions significantly. Let Å» Ć = Ć0 + Ć, Á = log + Á Å» Then N and disappear and the equations for static black holes become 1 1 1 Å» Å» Å» Ć + Ć = 2 - e2Ć Á + Á Å» Å» r 2 r 1 Å» Å» Å» Å» Å» Å» 1 - e2Ć Ć + Ć = 2 - e2Ć (Ć )2 - e2Á r As the dilaton field on the horizon approaches the critical value Ć0, the term Å» (1- e2Ć) will approach 2 , where = Ć0 -Ćh. This will cause the second derivative of Å»h Å» Å» Ć to be very large until Ć approaches -eÁ in a coordinate distance "r of order 4 . Å»h By the above equations, Á approaches -2eÁ in the same distance. A power series solution and numerical calculations carried out by Jonathan Brenchley confirm that Å» in the limit as tends to zero, the solution tends to a limiting form Ćc, Ác. Å» The limiting black hole is regular everywhere outside the horizon, but has a fairly mild singularity on the horizon with R diverging like r-1. At large values of r, the solution will tend to the linear dilaton in the manner of the asymptotic expansion given before. One or both of the constants K and L must be non zero, because the solution is not exactly the linear dilaton. Fitting to the asymptotic expansion gives a value bc H" 0.4 If the singularity inside the black hole were to remain hidden at all times, as in possibility (1) above, one might expect that the temperature and rate of evolution of the black hole would approach zero as the dilaton field on the horizon approached the critical value. However, this is not what happens. The fact that the black holes 7 Å» tend to the limiting solution Ćc, Ác means that the period in imaginary time will tend Å» 4Ä„bc to . Thus the temperature will be
Tc = 4Ä„bc The energy momentum tensor of one of the fi fields can be calculated from the conservation equations. In the xÄ… coordinates, they are: 1 f 2 T++ = - ("+Á"+Á - "+Á + t+(x+)), Å» Å» Å» 12 1 f 2 T-- = - ("-Á"-Á - "-Á + t-(x-)) Å» Å» Å» 12 where tÄ…(xÄ…) are chosen to satisfy the boundary conditions on the energy momentum tensor. In the case of a black hole formed by sending in an f wave, the boundary f condition is that the incoming flux T++ should be zero at large r. This would imply that 1 t+ = 4x2 + The energy momentum tensor would not be regular on the past horizon, but this does not matter as the physical spacetime would not have a past horizon but would be different before the f wave. On the other hand, the energy momentum tensor should be regular on the future horizon. This would imply that t-(x-) should be regular at x- = 0. Converting to the coordinates uÄ…, one then would obtain a steady rate 2 192b2 c of energy outflow in each f field at late retarded times u-. 8 Conclusions The fact that the temperature and rate of emission of the limiting black hole do not go to zero, establishes a contradiction with the idea that the black hole settles down to a stable state. Of course, this does not tell us what the semi-classical equations will predict, but it makes it very plausible that they will lead either to a naked singularity, or to a singularity that spreads out to infinity at some finite retarded time. The semi-classical evolution of these two dimensional black holes, is very similar to that of charged black holes in four dimensions with a dilaton field [5]. If one supposes that there are no fields in the theory that can carry away the charge, the steady loss of mass would suggest that the black hole would approach an extreme state. However, unlike the case of the Reissner-Nordstom solutions, the extreme black holes with a dilaton have a finite temperature and rate of emission. So one obtains a similar contradiction. If the solution where to evolve to a state of lower mass but the same charge, the singularity would become naked. There seems no way of avoiding naked singularity in the context of the semi- classical theory. If spacetime is described by a semi-classical Lorentz metric, a black hole can not disappear completely without there being some sort of naked singularity. But there seem to be zero temperature non radiating black holes only in a few cases. For example, charged black holes with no dilaton field and no fields to carry away the charge. What seems to happening is that the semi-classical approximation is breaking down in the strong coupling regime. In convential general relativity, this breakdown occurs only when the black hole gets down to the Planck mass. But in the two and four dimensional dilatonic theories, it can occur for macroscopic black holes when the dilaton field on the on the horizon approaches the critical value. When the coupling becomes strong, the semi-classical approximation will break down. Quantum fluctuations of the metric and the dilaton could no longer be neglected. One could imagine that this might lead to a tremendous explosion in which the remaining mass 9 energy of the black hole was released. Such explosions might be detected as gamma ray bursts. Even though the semi-classical equations seem to lead to a naked singularity, one would hope that this would not happen in a full quantum treatment. Quite what it means not to have naked singularities in a quantum theory of gravity is not immediately obvious. One possible interpretation is the no boundary condition [6]: spacetime is non singular and without boundary in the Euclidean regime. If this proposal is correct, some sort of Euclidean wormhole would have to occur, which would carry away the particles that went in to form the black hole, and bring in the particles to be emitted. These wormholes could be in a coherent state described by alpha parameters [7]. These parameters might be determined by the minimumization of the effective gravitational constant G [7,8,9]. In this case, there would be no loss of quantum coherence if a black hole were to evaporate and disappear completely. Or the alpha parameters might be different moments of a quantum field Ä… on superspace[10]. In this case there would be effective loss of quantum coherence, but it might be possible to measure all the alpha parameters involved in the evaporation of a black hole of a given mass. In that case, there would be no further loss of quantum coherence when black holes of up to that mass evaporated. I was greatly helped by talking to Giddings and Stominger who were working along similar lines. I also had useful discussions with Hayward, Horowitz and Preskill. This work was carried out during a visit to Cal Tech as a Sherman Fairchild Scholar. References 1. Callan, C.G., Giddings, S.B., Harvey, J.A., Strominger, A. Evanescent Black Holes UCSB-TH-91-54. 2. Hawking, S.W. Particle Creation by Black Holes, Commun. Math. Phys. 43,199 (1975). 3. Banks, T., Dabholkar, A., Douglas, M.R., O Loughlin, M. Are Horned Particles the Climax Of Hawking Evaporation? RU-91-54. 10 4. Russo, J.G., Susskind, L., Thorlacius, L. Black Hole Evaporation in 1+1 Di- mensions SU-ITP-92-4. 5. Garfinkle, D., Horowitz, G.T., Strominger, A. Charged Black Holes in String Theory, Phys. Rev D 43, 3140. 6. Hartle, J.B., Hawking, S.W. Wave Function of the Universe Phys. Rev. D28, 2960-2975 (1983). 7. Coleman, S. Why There Is Nothing Rather Than Something: A Theory Of The Cosmological Constant Nucl. Phys. B310 (1988), 643. 8. Preskill, J. Wormholes In Spacetime And The Constants Of Nature. Nucl. Phys. B323 (1989), 141. 9. Hawking, S.W. Do Wormholes Fix The Constants Of Nature? Nucl. Phys. B335,155-165 (1990). 10. Hawking, S.W. The Effective Action For Wormholes. Nucl. Phys. B363, 117- 131 (1991). 11