1: \documentclass[prl,aps,floats,twocolumn,tightenlines,notitlepage,preprintnumbers,showpacs]{revtex4}
2: \usepackage{epsfig}
3: \usepackage{amsmath}
4: \usepackage{graphics}
5: \usepackage{epsfig}
6:
7: \newcommand{\bea}{\begin{eqnarray}}
8: \newcommand{\eea}{\end{eqnarray}}
9: \newcommand{\beq}{\begin{equation}}
10: \newcommand{\eeq}{\end{equation}}
11:
12: \begin{document}
13: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
14: \title{Escape of black holes from the brane}
15: \author{Antonino Flachi}
16: \email{flachi@yukawa.kyoto-u.ac.jp}
17: \affiliation{Yukawa Institute for Theoretical Physics, Kyoto University, Kyoto 606-8502, Japan}
18: \author{Takahiro Tanaka}
19: \email{tama@scphys.kyoto-u.ac.jp}
20: \affiliation{Department of Physics, Kyoto University, Kyoto 606-8502, Japan}
21: %\date{May 2005}
22: \preprint{YITP-05-28}
23: \preprint{KUNS-1974}
24: \pacs{04.70.Dy, 11.10.Kk}
25: \begin{abstract}
26: TeV-scale gravity theories allow the possibility of producing small
27: black holes at energies that soon will be explored at the LHC or at
28: the Auger observatory.
29: One of the expected signatures is the detection of Hawking radiation,
30: that might eventually terminate if the black hole, once perturbed,
31: leaves the brane.
32: Here, we study how the `black hole plus brane' system evolves once the
33: black hole is given an initial velocity, that mimics, for
34: instance, the recoil due to the
35: emission of a graviton. The results of our dynamical analysis
36: show that the brane bends around the black hole,
37: suggesting that the
38: black hole eventually escapes into the extra dimensions once two
39: portions of the brane come in contact and reconnect. This gives a
40: dynamical mechanism for the creation of baby
41: branes.
42: \end{abstract}
43: \maketitle
44: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
45: \noindent
46: {\it Introduction}:
47: The reason why gravity is so much weaker than all the other forces,
48: %$G_N/G_F \sim 10^{-33}$,
49: aside from being still a mystery ({\it the
50: hierarchy problem}), constitutes
51: the major obstacle in performing experiments in the realm of quantum
52: gravity, for the rather obvious reason that overcoming the fundamental
53: Planck scale requires center of
54: mass energies greater than $10^{18}$ GeV.
55:
56: This common belief was radically questioned a few years ago.
57: Although in four dimensions there is no hope to reach such an energy
58: in a terrestrial experiment, the situation
59: drastically changes when one assumes the existence of extra dimensions.
60: In the latter case it might even be possible to reach a transplanckian
61: regime. When geometry and scales of the extra (space-like) dimensions
62: are appropriately chosen, the higher dimensional Planck scale may turn
63: out to be much smaller than $10^{18}$ GeV, thus reformulating the
64: hierarchy problem in geometrical terms.
65:
66: The details depend upon the specific realisation of the model, two popular examples
67: being scenarios with large \cite{add} or warped extra dimensions \cite{rs}, but the
68: common features are a low fundamental Planck scale, $M\sim$TeV, the localisation of the
69: standard model on branes, representing our directly observable universe, and the
70: propagation of gravity throughout the higher dimensional space. This is, roughly, what
71: is known as the brane world.
72:
73: In the past few years many people started to use these simple ideas to
74: build models of various type ranging from cosmology to particle physics
75: and, more speculatively,
76: investigate how the brane world can be used as a tool to solve long
77: standing problems in physics. To date no complete solution to any
78: problem has yet been found and
79: many of the predictions of the brane world, relying on tunings of various
80: sorts, do not provide definitive and clear-cut answers. However, it is
81: widely believed that, if the
82: fundamental scale of gravity truly lies in the TeV range, a very
83: spectacular and relatively model independent prediction can be made,
84: that is the creation of small black
85: holes in the high energy collision of two particles \cite{dl,gt}. Thus
86: the LHC or the Auger Observatory might be able to perform quantum
87: gravity experiments and initiate the study of
88: black hole microphysics.
89:
90: A generic assumption is to consider the production of
91: black holes whose mass exceeds the fundamental Planck scale $M$. In this
92: regime a semiclassical treatment is possible and quantum
93: gravity corrections can be ignored. The size of the black holes produced
94: at colliders is typically assumed to be much smaller than the
95: characteristic length of the extra
96: dimensions. Then it seems reasonable to describe these objects by higher
97: dimensional asymptotically flat solutions \cite{tangherlini,mp}. In
98: this approximation, the Schwarzschild radius %of $R_s$
99: of a $d$-dimensional black hole of mass $m$ is given by
100: \beq
101: R_s = {1\over \sqrt{\pi}}\left({8\Gamma((d-1)/2)\over d-2}\right)^{1\over(d-3)}
102: \left({m\over M}\right)^{1\over(d-3)} {1\over M}~.
103: \nonumber
104: \eeq
105: We use units in which $c=\hbar=1$.
106: Given these assumptions, one considers two particles with center of mass energy
107: $\sqrt{s}$ moving in opposite directions at impact parameter $b$. When $b$ is smaller than
108: $R_s$ with $m$ replaced by $\sqrt{s}$,
109: semiclassical arguments show that a marginally trapped surface forms at the overlap
110: between the two colliding shock waves describing the two scattering
111: particles. This implies the formation of a common horizon unless naked
112: singularities are formed.
113: The black disk approximation, then, suggests the following formula for the cross section:
114: \beq
115: \sigma_{BH} \sim \pi R_s^2~.
116: \nonumber
117: \eeq
118: Assuming a higher dimensional Planck scale $M \leq 3$TeV,
119: the cross section will range between $10^{-2}$ to $10^2$ pb
120: at the LHC energy, $\sqrt{s} \sim 14$ TeV.
121: At the luminosity $L \sim 10^{34}$ cm$^{-2}$ s$^{-1}$, the LHC will be
122: able to produce about $10^{7}$ black holes per year.
123:
124: After the black hole is formed, it will decay by Hawking radiation at a temperature
125: \beq
126: T_H = {d-3\over 4\pi R_s}~.
127: \nonumber
128: \eeq
129: As it was argued in Ref.~\cite{ehm}, Hawking evaporation must emit comparable
130: amounts of energy into each low energy effective degree of freedom
131: in the bulk and on the brane. Then, if we assume that only gravity propagates in
132: the bulk, radiation on the brane will be the dominant component of the Hawking radiation.
133:
134: However, this is not the whole story. As noticed in Ref.~\cite{frolov}, the black hole
135: is expected to be rotating and thus exhibits superradiance; this would enhance the
136: emission of higher spin particles, possibly making the emission of gravitons a dominant
137: effect, thus strongly perturbing the system and eventually resulting in the black
138: hole leaving the brane with sudden termination of the Hawking
139: radiation. Ref.~\cite{frolov} discusses this on the basis of a field theory model,
140: where the black hole,
141: treated as a point radiator, is described as a massive scalar field with internal
142: degrees of freedom. Ref.~\cite{stojkovic} pushes this further and argues that it is
143: possible to distinguish between ${\mathcal Z}_2-$ and non ${\mathcal Z}_2-$symmetric
144: scenarios due to the fact that a black hole cannot recoil if the spacetime is
145: ${\mathcal Z}_2-$symmetric.
146:
147: In this paper we would like to reconsider this problem. Although the interaction
148: between defects and black holes is a relatively well studied subject \cite{cfl}, most
149: of the investigations performed so far focus on the static case.
150: Apart from this, it is assumed
151: that a black hole may leave the brane somehow, but
152: there is no direct example showing how such departure may occur. In the
153: present work we will look at the motion of a brane in the gravitational
154: field of a (small) black
155: hole, and see explicitly how a
156: black hole can escape from the brane.
157: % by solving the dynamical problem, and .
158: %As a result of our investigation all the quantities of interest can
159: %be directly evaluated and the various regimes studied directly.
160:
161: As we will see, due to the presence of the black hole, even in the most crude
162: approximation, any perturbation that will give the brane an initial velocity with
163: respect to the black hole, will cause deformations in the brane itself. This is
164: essential to understand whether or not a black hole can leave the brane. We will also
165: see that these deformations induced in the brane can be indeed simulated in a precise
166: way, allowing us to determine the time scale, $\tau_*$, at which this separation
167: eventually occurs and to compare it with the lifetime $\tau$ of the black hole.
168: This will tell us whether or not the escape occurs before the
169: black hole evaporation completes.
170:
171: \noindent
172: {\it Membrane dynamics}:
173: We shall describe the set-up and the limitation of our approach.
174: As we have mentioned, the question we would like to answer is whether or not a black
175: hole on a brane leaves the brane once the system is
176: perturbed and, if this is the case, work out the time scale at which this process occurs.
177:
178: We will consider a general $d-$dimensional bulk spacetime and assume that our universe is
179: a $(p+1)-$dimensional brane. We consider black holes whose gravitational
180: radius is much smaller than the size of the extra dimensions, and
181: this allows us to assume that our spacetime is adequately described by the
182: asymptotically flat solution \cite{tangherlini, mp}. This is of course a
183: reasonable assumption only if the self-gravity of the
184: brane is negligible, which is the case when the size of black
185: holes is sufficiently small.
186: %{\it i.e.} if the tension is very small and,
187: As long as this approximation is reasonable, our
188: results will give a model-independent prediction. We assume that the
189: black hole is formed out of matter on the brane, and symmetries require
190: that the brane initially lies on the equatorial plane of the black hole.
191: From the point of view of accelerator generated black holes, the interesting
192: possibility is to consider rotating black
193: holes, however here we start with the case of Schwarzschild black holes
194: as a first step, hoping to capture the essential features of the
195: problem, %and in order to compare
196: and deferring the rotating case to our forthcoming work
197: \cite{nino-takahiro-forth}. Thus the bulk space is
198: described by the following line element:
199: \beq
200: ds^2 = -f(r) dt^2 + f^{-1}(r) dr^2 + r^2 d \Omega^2_{d-2}~,
201: \nonumber
202: \eeq
203: where the function $f(r)=1-(1/r)^{d-3}$. We set the horizon
204: radius to unity by adjusting the unit of the length.
205: Then the area of the event horizon is equal to that of a unit $(d-2)-$sphere,
206: $\Omega_{d-2}$.
207:
208: The brane in the leading approximation is described by a Dirac-Nambu-Goto action:
209: \beq
210: S = - \sigma \int d^{p+1}\zeta \sqrt{\gamma}~,
211: \nonumber
212: \eeq
213: where $\sigma$ is the tension of the brane and $\gamma$ is
214: the determinant
215: of the induced metric $\gamma_{ij}$ on the brane.
216:
217: We use $\{\zeta^a\}\equiv
218: \{t,r,\mbox{\boldmath$\chi$}\}$ with $a= 0,1, ..., p$
219: as the coordinates on the brane, where
220: {\boldmath$\chi$} represent
221: coordinates of a $(p-1)$-dimensional sphere.
222: Trajectory of a spherically symmetric brane
223: is specified by
224: the azimuthal inclination angle $\theta(t,r)$.
225: %\begin{equation}
226: %x^A = x^A(\zeta^a)
227: %\end{equation}
228: %where $A= 0, 1, ..., d$ and
229: %${\zeta^a}$
230: %In the vicinity of the membrane a convenient choice of coordinates is $(\zeta^a, z)$ with the $z-$direction being orthogonal to the wall. In these coordinates the invariant line element has the form
231: %\begin{equation}
232: %ds^2 = g_{AB}dx^Adx^B = g_{AB}x^A_{,a} x^B_{,b} d\zeta^a d\zeta^b -dz^2~,
233: %\end{equation}
234: %and
235: The induced metric on the brane is then given by
236: \begin{eqnarray}
237: \gamma_{ab} d\zeta^a d\zeta^b &= &
238: - f(r) dt^2 + f^{-1}(r)dr^2 \cr
239: && +r^2
240: \left[(\theta_t dt +\theta_r dr)^2+\sin^2\theta d\Omega_{p-1}^2\right]~,
241: \nonumber
242: \end{eqnarray}
243: where
244: the notation $h_x \equiv \partial h /\partial x$ is used,
245: and we have
246: \beq
247: S_{\cal B} = - \sigma \Sigma_{p-1}\int dt dr {\cal L}~,
248: \nonumber
249: \eeq
250: with
251: \beq
252: {\cal L} = \left(r \sin\theta \right)^{p-1}\sqrt{1 + r^2 f(r)
253: \theta_{r}^2 - r^2 f^{-1}(r) \theta_{t}^2}~.
254: \nonumber
255: \eeq
256: The case with $d=5$ and $p=3$ corresponds to a co-dimension one brane in a five
257: dimensional bulk, whereas that with $d=4$ and $p=1$
258: corresponds to a string in a four
259: dimensional spacetime. From the above action the equation of motion
260: can be obtained and it has to be supplemented by appropriate boundary
261: conditions. As we have already mentioned we work in the center of mass
262: frame of the black hole, and the fact that
263: the intersection point between the black
264: hole and the brane is not allowed to move fixes the boundary conditions
265: at the horizon. The recoil of the black hole in this frame
266: is described by giving some initial asymptotically uniform
267: velocity to the brane.
268: Pictorially the situation is described in {\it
269: fig}.~\ref{motion}.
270: The explicit form of the initial velocity we gave in our simulation
271: is $u=v\left(1-{1\over r}\right)$, but the result does not depend much on our choice of
272: the initial velocity profile.
273:
274: \begin{figure}
275: \scalebox{0.5} {\includegraphics{motion_pict.eps}}
276: \caption{Schematic picture of the system in the center of mass frame of
277: the black hole. The intersection point is held fixed and the brane is
278: given an initial velocity.}
279: \label{motion}
280: \end{figure}
281:
282: The task at hand is to solve the dynamical evolution of the brane.
283: This has been done first by writing the equation of motion in a new radial coordinate $\tilde{r}$, defined via the relation $r = 1/(1-e^{-\tilde{r}})$, and then by
284: applying finite difference methods
285: along with the Newton-Raphson algorithm to solve the associated non-linear algebraic system. The grid spacing is chosen to be $\delta \tilde{r}= 0.025$ along the
286: radial direction
287: and $\delta t = 0.0025$ along the time direction.
288: The number of points of the grid is fixed to be $N_{\tilde{r}} = 134$ along $\tilde{r}$ and varied along $t$
289: depending on the initial velocity. The chosen values are $N_t = 3400, 2420, 1970, 1700$ corresponding to $v=0.25,~0.50,~0.75,~1$ respectively.
290: The results of the simulation are reported in {\it fig}.~\ref{simulation}.
291: We find that in all cases the brane is bent and eventually the radius
292: of the pinched part goes to zero.
293: This result does not contradict with the stability analysis
294: presented in Ref.~\cite{ishibashi}, in which it was shown
295: that there is no linear instability for the case with $d=4$ and $p=2$,
296: since the imposed asymptotic boundary conditions are different.
297: Roughly speaking,
298: the time scale of the escape of a
299: black hole is found to be given by $\tau_*\approx R_s/v$ where
300: $v$ is the initial recoil velocity.
301: In the limit $v\ll 1$ back reaction due to the tension of the brane,
302: which is neglected in our present calculation, may
303: affect the evolution significantly.
304: In this limit, however, the evolution will be well
305: approximated by a sequence of static configurations.
306: For the case with $d=4$ and $p=2$, static solutions are
307: given in Ref.~\cite{cfl}.
308: If we imagine the situation that there is an edge of the brane at
309: a large radius $r=r_{\rm edge}$, the force necessary
310: to sustain the static configuration will be estimated by
311: $F=\sigma\Sigma_{p-1}r^{p-1} d(r\theta)/dr|_{r=r_{\rm edge}}$.
312: For $p\geq 3$ this force becomes negative when $\theta$ is positive.
313: This means that the force between a black hole and the brane is
314: repulsive in this static limit. Therefore the effect of the brane
315: tension will not prevent the black hole escaping from the brane.
316:
317: \begin{figure}
318: \scalebox{0.5} {\includegraphics{025.eps}}
319: \scalebox{0.5} {\includegraphics{05.eps}}
320: \scalebox{0.5} {\includegraphics{075.eps}}
321: \scalebox{0.5} {\includegraphics{1.eps}}
322: \caption{Unstable deformations of the brane: results of the simulation.
323: $v$ labels the initial velocity. All the simulations reported refer to a five dimensional bulk and 3+1 dimensional brane and
324: the plots are drown in cylindrical coordinates $R=r \sin \theta$ and $Z=-r\cos \theta$.}
325: \label{simulation}
326: \end{figure}
327:
328: \noindent
329: {\it Concluding remarks}:
330: We have considered a system consisting of a brane plus a black hole.
331: We assumed that the size of the black hole is small compared to the extra
332: dimensions and that the tension of the brane is negligible. In this
333: approximation, the spacetime with a black hole is well described by
334: asymptotically flat solution as given in
335: Refs.~\cite{tangherlini,mp}.
336: If ${\mathcal Z}_2$-symmetry is not imposed,
337: it seems natural to consider the situation in
338: which the black hole acquires some initial velocity with respect to the
339: brane, say, due to anisotropic emission of particles.
340: We simulated this process by studying the dynamical evolution of a
341: brane in a fixed black hole spacetime.
342: The main result of our study is that, irrespectively of the initial
343: velocity $v$,
344: the brane tends to wrap around the black hole, suggesting that the black hole
345: might escape in the extra dimensions after two portions of the brane
346: come in contact and reconnect (see {\it
347: fig}.~\ref{brane_nucleation}). Such a set-up is of relevance in a number
348: of physical situations.
349:
350: \begin{figure}
351: \scalebox{0.5} {\includegraphics{reconnection.eps}}
352: \caption{Creation of a baby brane wrapped around the black hole.}
353: \label{brane_nucleation}
354: \end{figure}
355:
356: The first application we have in mind is that of accelerator generated
357: black holes: will it be possible to observe the black hole in its final
358: state and the
359: process of evaporation or will the black hole disappear in the extra
360: dimensions?
361: The results of our analysis strongly suggests
362: that the black hole will indeed escape from the brane.
363: In fact, this process is likely to occur before
364: the black hole evaporation completes if the initial mass of
365: the produced black hole is sufficiently large.
366: The time scale of Hawking evaporation will be estimated as
367: $\tau\approx (dM/Mdt)^{-1}\approx M^{-1}(m/M)^{n+3/n+1}$.
368: On the other hand,
369: the time scale of the escape of a
370: black hole was found to be given by $\tau_*\approx R_s/v$.
371: If the origin of the recoil
372: is the inhomogeneous emission of particles due to
373: Hawking radiation, we will have a rough estimate
374: $v\approx \alpha/\sqrt{N}$, where $\alpha$ is
375: the fraction of the evaporated mass and $N\approx \alpha M/T_H$
376: is the number of emitted particles.
377: Then we have $\tau_*\approx M^{-1}(m/M)^{n+4\over 2(n+1)}
378: /\sqrt{\alpha}$.
379: For $\alpha=O(1)$ and $m\agt M$,
380: the time scale of escape is shorter than the time scale of
381: evaporation.
382: Furthermore, $\alpha$ at $\tau_*$ is estimated as
383: $\alpha \approx \tau_*/\tau$. Using this relation,
384: we find $\tau_*\approx M^{-1}(m/M)^{2n+7\over 3(n+1)}$.
385:
386: Aside from more phenomenological application mentioned above, it is
387: amusing to speculate about other possible situations where the process
388: of baby brane nucleation may be relevant. An interesting application is
389: the possibility to generate global charge non-conserving processes on the
390: defect. This has already been noticed in Ref.~\cite{dg}, where the
391: quantum fluctuations of the geometry induce nucleation of baby branes
392: and this is used as a mechanism to generate baryon asymmetry in the
393: early universe. Our example, though, is different in that it is entirely
394: classical and provides an explicit example of how this nucleation takes
395: place.
396: All this is also reminiscent of the baryogenesis mechanism proposed in
397: Refs.~\cite{hawking,zeldovich}, where the evaporation of primordial black
398: hole is used as a means
399: to generate the asymmetry. Of course a more detailed study is necessary
400: to quantify whether or not such a possibility is feasible
401: \cite{nino-takahiro-forth}.
402:
403:
404: One can push things even further by relaxing the condition that the
405: fundamental Planck scale is in the TeV range, but still assume a higher
406: dimensional spacetime. This will not change the process of baby brane
407: creation but will reduce the size of such objects from $10^{-16}$ cm to the traditional
408: Planck length $l_P \sim 10^{-33}$ cm. It is, then, tempting to imagine the spacetime at
409: the Planck scale
410: as a foam of baby branes and the extra dimensions filled with such
411: bubbles \cite{nino-takahiro-forth}.
412:
413: In the present paper, we completely neglected the self-gravity of the
414: brane. When ${\mathcal Z}_2$-symmetry across the brane is imposed,
415: the brane bending is not allowed in our current treatment.
416: Once the self-gravity of the brane is turned on, however, it is not
417: completely clear whether the escape of black holes from the brane is
418: forbidden or not. In fact, such a process is
419: thought to be one possible mechanism to explain the
420: proposed conjecture of classical black hole evaporation
421: in the Randall-Sundrum II model~\cite{takahiro, kaloper}.
422: This is a challenging problem which requires development of
423: numerical relativity with a self-gravitating
424: singular hypersurface and a black hole.
425:
426: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
427: %\section*{Acknowledgements}
428: We wish to thank C. Germani, O. Pujol\`as and M. Sasaki for useful discussions.
429: This work is supported in part by
430: Grant-in-Aid for Scientific Research, No. 16740165 from Japan Society for Promotion of Science
431: and by that for the 21st Century COE
432: ``Center for Diversity and Universality in Physics'' at
433: Kyoto university, both from the Ministry of
434: Education, Culture, Sports, Science and Technology of Japan.
435: A.F. is supported by the JSPS under contract No. P047724.
436:
437: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
438: \begin{thebibliography}{99}
439:
440: \bibitem{add} N.~Arkani-Hamed, S.~Dimopoulos and G.~R.~Dvali, Phys.\ Lett.\ B {\bf 429}, 263 (1998).
441:
442: \bibitem{rs} L.~Randall and R.~Sundrum, Phys.\ Rev.\ Lett.\ {\bf 83}, 3370 (1999).
443:
444: \bibitem{dl}
445: S.~Dimopoulos and G.~Landsberg, Phys.\ Rev.\ Lett.\ {\bf 87} (2001) 161602.
446:
447: \bibitem{gt}
448: S.~B.~Giddings and S.~Thomas, Phys.\ Rev.\ {\bf D65} (2002) 056010.
449:
450: \bibitem{tangherlini} F.~R.~Tangherlini, Nuovo Cim.\ B {\bf 77}, (1963) 636.
451:
452: \bibitem{mp} R.~C.~Myers and M.~J.~Perry, Annals Phys.\ {\bf 172} (1986) 304.
453:
454: \bibitem{ehm} R.~Emparan, G.~T.~Horowitz, R.~C.~Myers, Phys.\ Rev.\ Lett.\ {\bf 85} (2000) 499.
455:
456: \bibitem{frolov} V.~P.~Frolov, D.~Stojkovic, Phys.\ Rev.\ Lett.\ {\bf 89} (2002) 151302.
457:
458: \bibitem{stojkovic} D.~Stojkovic, Phys.\ Rev.\ Lett.\ {\bf 94} (2005) 011603.
459:
460: \bibitem{cfl} M.~Christensen, V.~P.~Frolov, A.~L.~Larsen, Phys.\ Rev. {\bf D58} (1998) 085008.
461:
462: %\bibitem{hm} G.~T.~Horowitz and K.~Maeda, Phys.\ Rev.\ Lett.\ {\bf 87} (2001) 131301.
463:
464: \bibitem{nino-takahiro-forth} A.~Flachi and T.~Tanaka, work in progress.
465:
466: %\bibitem{hhr} S.~Hawking, T.~Hertog, H.~Reall, Phys.\ Rev.\ {\bf D62} (2000) 043501.
467:
468: \bibitem{ishibashi} S.~Higaki, A.~Ishibashi, D.~Ida,
469: Phys.\ Rev.\ {\bf D63} (2001) 025002.
470:
471: \bibitem{dg} G.~Dvali, G.~Gabadadze, Phys.\ Lett.\ {\bf B460} (1999) 47.
472:
473: \bibitem{hawking} S.~W.~Hawking, Nature {\bf 248} (1974) 30.
474:
475: \bibitem{zeldovich} Ya.~B.~Zeldovich, Zh. Eksp. Teor. Fiz. Pis'ma Red. {\bf 24} (1976) 29.
476:
477: \bibitem{takahiro} T.~Tanaka, Prog. Theor. Phys. {\bf 148} (2002) 307.
478:
479: \bibitem{kaloper}
480: R.~Emparan, A.~Fabbri and N.~Kaloper, JHEP {\bf 0208}, (2002) 043.
481:
482: \end{thebibliography}
483: %%%%%%%%%%%%%%%%%%%%%%%%%%%
484:
485: \end{document}
486:
487: